diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,59299 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9984301412872841,
+  "eval_steps": 8435,
+  "global_step": 8435,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0012795905310300703,
+      "grad_norm": 0.0,
+      "learning_rate": 0.0,
+      "loss": 1.0854,
+      "step": 1
+    },
+    {
+      "epoch": 0.0025591810620601407,
+      "grad_norm": 12.814065883733651,
+      "learning_rate": 6.25e-08,
+      "loss": 1.1119,
+      "step": 2
+    },
+    {
+      "epoch": 0.003838771593090211,
+      "grad_norm": 12.814065883733651,
+      "learning_rate": 6.25e-08,
+      "loss": 1.1259,
+      "step": 3
+    },
+    {
+      "epoch": 0.005118362124120281,
+      "grad_norm": 11.692146286212495,
+      "learning_rate": 1.25e-07,
+      "loss": 1.2041,
+      "step": 4
+    },
+    {
+      "epoch": 0.006397952655150352,
+      "grad_norm": 20.58418180248586,
+      "learning_rate": 1.875e-07,
+      "loss": 1.1192,
+      "step": 5
+    },
+    {
+      "epoch": 0.007677543186180422,
+      "grad_norm": 14.937213110456936,
+      "learning_rate": 2.5e-07,
+      "loss": 1.1382,
+      "step": 6
+    },
+    {
+      "epoch": 0.008957133717210493,
+      "grad_norm": 22.81831781610188,
+      "learning_rate": 3.1249999999999997e-07,
+      "loss": 1.1557,
+      "step": 7
+    },
+    {
+      "epoch": 0.010236724248240563,
+      "grad_norm": 12.704842548106877,
+      "learning_rate": 3.75e-07,
+      "loss": 1.1161,
+      "step": 8
+    },
+    {
+      "epoch": 0.011516314779270634,
+      "grad_norm": 16.8869492025962,
+      "learning_rate": 4.375e-07,
+      "loss": 1.0873,
+      "step": 9
+    },
+    {
+      "epoch": 0.012795905310300703,
+      "grad_norm": 11.613500906997396,
+      "learning_rate": 5e-07,
+      "loss": 1.0701,
+      "step": 10
+    },
+    {
+      "epoch": 0.014075495841330775,
+      "grad_norm": 14.228422014308805,
+      "learning_rate": 5.625e-07,
+      "loss": 1.0735,
+      "step": 11
+    },
+    {
+      "epoch": 0.015355086372360844,
+      "grad_norm": 8.005557267033051,
+      "learning_rate": 6.249999999999999e-07,
+      "loss": 1.0007,
+      "step": 12
+    },
+    {
+      "epoch": 0.016634676903390915,
+      "grad_norm": 6.111155320976804,
+      "learning_rate": 6.875e-07,
+      "loss": 0.9478,
+      "step": 13
+    },
+    {
+      "epoch": 0.017914267434420986,
+      "grad_norm": 11.454622711115135,
+      "learning_rate": 7.5e-07,
+      "loss": 0.9891,
+      "step": 14
+    },
+    {
+      "epoch": 0.019193857965451054,
+      "grad_norm": 10.783961258267503,
+      "learning_rate": 8.125e-07,
+      "loss": 0.9233,
+      "step": 15
+    },
+    {
+      "epoch": 0.020473448496481125,
+      "grad_norm": 8.67980534265599,
+      "learning_rate": 8.75e-07,
+      "loss": 0.9221,
+      "step": 16
+    },
+    {
+      "epoch": 0.021753039027511197,
+      "grad_norm": 12.251575948508574,
+      "learning_rate": 9.374999999999999e-07,
+      "loss": 0.9929,
+      "step": 17
+    },
+    {
+      "epoch": 0.023032629558541268,
+      "grad_norm": 11.289043148876596,
+      "learning_rate": 1e-06,
+      "loss": 0.8957,
+      "step": 18
+    },
+    {
+      "epoch": 0.02431222008957134,
+      "grad_norm": 8.515749337990352,
+      "learning_rate": 1e-06,
+      "loss": 0.8998,
+      "step": 19
+    },
+    {
+      "epoch": 0.025591810620601407,
+      "grad_norm": 8.795231852711888,
+      "learning_rate": 1e-06,
+      "loss": 0.8795,
+      "step": 20
+    },
+    {
+      "epoch": 0.026871401151631478,
+      "grad_norm": 7.335592358778729,
+      "learning_rate": 1e-06,
+      "loss": 0.8861,
+      "step": 21
+    },
+    {
+      "epoch": 0.02815099168266155,
+      "grad_norm": 6.396246343205464,
+      "learning_rate": 1e-06,
+      "loss": 0.8855,
+      "step": 22
+    },
+    {
+      "epoch": 0.02943058221369162,
+      "grad_norm": 9.404965979933332,
+      "learning_rate": 1e-06,
+      "loss": 0.8899,
+      "step": 23
+    },
+    {
+      "epoch": 0.030710172744721688,
+      "grad_norm": 5.130401621788799,
+      "learning_rate": 1e-06,
+      "loss": 0.8412,
+      "step": 24
+    },
+    {
+      "epoch": 0.03198976327575176,
+      "grad_norm": 7.128916936378112,
+      "learning_rate": 1e-06,
+      "loss": 0.8219,
+      "step": 25
+    },
+    {
+      "epoch": 0.03326935380678183,
+      "grad_norm": 7.684370107510004,
+      "learning_rate": 1e-06,
+      "loss": 0.8907,
+      "step": 26
+    },
+    {
+      "epoch": 0.0345489443378119,
+      "grad_norm": 8.913643361737412,
+      "learning_rate": 1e-06,
+      "loss": 0.8293,
+      "step": 27
+    },
+    {
+      "epoch": 0.03582853486884197,
+      "grad_norm": 6.325911882206449,
+      "learning_rate": 1e-06,
+      "loss": 0.9362,
+      "step": 28
+    },
+    {
+      "epoch": 0.037108125399872044,
+      "grad_norm": 7.603003738222256,
+      "learning_rate": 1e-06,
+      "loss": 0.8313,
+      "step": 29
+    },
+    {
+      "epoch": 0.03838771593090211,
+      "grad_norm": 7.07594773195082,
+      "learning_rate": 1e-06,
+      "loss": 0.8901,
+      "step": 30
+    },
+    {
+      "epoch": 0.03966730646193218,
+      "grad_norm": 8.432274576237745,
+      "learning_rate": 1e-06,
+      "loss": 0.8413,
+      "step": 31
+    },
+    {
+      "epoch": 0.04094689699296225,
+      "grad_norm": 7.755240102548513,
+      "learning_rate": 1e-06,
+      "loss": 0.8957,
+      "step": 32
+    },
+    {
+      "epoch": 0.04222648752399232,
+      "grad_norm": 7.2655002111567875,
+      "learning_rate": 1e-06,
+      "loss": 0.8384,
+      "step": 33
+    },
+    {
+      "epoch": 0.04350607805502239,
+      "grad_norm": 9.752270597284719,
+      "learning_rate": 1e-06,
+      "loss": 0.804,
+      "step": 34
+    },
+    {
+      "epoch": 0.044785668586052464,
+      "grad_norm": 6.740566109302612,
+      "learning_rate": 1e-06,
+      "loss": 0.847,
+      "step": 35
+    },
+    {
+      "epoch": 0.046065259117082535,
+      "grad_norm": 7.866343496654274,
+      "learning_rate": 1e-06,
+      "loss": 0.7829,
+      "step": 36
+    },
+    {
+      "epoch": 0.04734484964811261,
+      "grad_norm": 6.363319060417899,
+      "learning_rate": 1e-06,
+      "loss": 0.8483,
+      "step": 37
+    },
+    {
+      "epoch": 0.04862444017914268,
+      "grad_norm": 9.56490908253171,
+      "learning_rate": 1e-06,
+      "loss": 0.8746,
+      "step": 38
+    },
+    {
+      "epoch": 0.04990403071017274,
+      "grad_norm": 4.819116687727287,
+      "learning_rate": 1e-06,
+      "loss": 0.8045,
+      "step": 39
+    },
+    {
+      "epoch": 0.05118362124120281,
+      "grad_norm": 5.084925305679592,
+      "learning_rate": 1e-06,
+      "loss": 0.8255,
+      "step": 40
+    },
+    {
+      "epoch": 0.052463211772232884,
+      "grad_norm": 7.466655307306172,
+      "learning_rate": 1e-06,
+      "loss": 0.8216,
+      "step": 41
+    },
+    {
+      "epoch": 0.053742802303262956,
+      "grad_norm": 5.611468711405401,
+      "learning_rate": 1e-06,
+      "loss": 0.8123,
+      "step": 42
+    },
+    {
+      "epoch": 0.05502239283429303,
+      "grad_norm": 7.680060722725615,
+      "learning_rate": 1e-06,
+      "loss": 0.8294,
+      "step": 43
+    },
+    {
+      "epoch": 0.0563019833653231,
+      "grad_norm": 6.71194850252695,
+      "learning_rate": 1e-06,
+      "loss": 0.7882,
+      "step": 44
+    },
+    {
+      "epoch": 0.05758157389635317,
+      "grad_norm": 4.97334178287351,
+      "learning_rate": 1e-06,
+      "loss": 0.8102,
+      "step": 45
+    },
+    {
+      "epoch": 0.05886116442738324,
+      "grad_norm": 6.801551748702033,
+      "learning_rate": 1e-06,
+      "loss": 0.8133,
+      "step": 46
+    },
+    {
+      "epoch": 0.060140754958413305,
+      "grad_norm": 8.207031832237497,
+      "learning_rate": 1e-06,
+      "loss": 0.8121,
+      "step": 47
+    },
+    {
+      "epoch": 0.061420345489443376,
+      "grad_norm": 4.654628623004874,
+      "learning_rate": 1e-06,
+      "loss": 0.7891,
+      "step": 48
+    },
+    {
+      "epoch": 0.06269993602047345,
+      "grad_norm": 6.189728621637134,
+      "learning_rate": 1e-06,
+      "loss": 0.7823,
+      "step": 49
+    },
+    {
+      "epoch": 0.06397952655150352,
+      "grad_norm": 5.941002734728932,
+      "learning_rate": 1e-06,
+      "loss": 0.8296,
+      "step": 50
+    },
+    {
+      "epoch": 0.06525911708253358,
+      "grad_norm": 6.06929561970064,
+      "learning_rate": 1e-06,
+      "loss": 0.8834,
+      "step": 51
+    },
+    {
+      "epoch": 0.06653870761356366,
+      "grad_norm": 8.213970780215423,
+      "learning_rate": 1e-06,
+      "loss": 0.8274,
+      "step": 52
+    },
+    {
+      "epoch": 0.06781829814459372,
+      "grad_norm": 7.794762162804007,
+      "learning_rate": 1e-06,
+      "loss": 0.8828,
+      "step": 53
+    },
+    {
+      "epoch": 0.0690978886756238,
+      "grad_norm": 9.812088037670044,
+      "learning_rate": 1e-06,
+      "loss": 0.8494,
+      "step": 54
+    },
+    {
+      "epoch": 0.07037747920665387,
+      "grad_norm": 7.314373514478439,
+      "learning_rate": 1e-06,
+      "loss": 0.855,
+      "step": 55
+    },
+    {
+      "epoch": 0.07165706973768395,
+      "grad_norm": 7.476365287911143,
+      "learning_rate": 1e-06,
+      "loss": 0.8592,
+      "step": 56
+    },
+    {
+      "epoch": 0.07293666026871401,
+      "grad_norm": 7.9128786576763295,
+      "learning_rate": 1e-06,
+      "loss": 0.8254,
+      "step": 57
+    },
+    {
+      "epoch": 0.07421625079974409,
+      "grad_norm": 6.726207670698307,
+      "learning_rate": 1e-06,
+      "loss": 0.8547,
+      "step": 58
+    },
+    {
+      "epoch": 0.07549584133077415,
+      "grad_norm": 5.61475641294302,
+      "learning_rate": 1e-06,
+      "loss": 0.8223,
+      "step": 59
+    },
+    {
+      "epoch": 0.07677543186180422,
+      "grad_norm": 8.17902257464014,
+      "learning_rate": 1e-06,
+      "loss": 0.7744,
+      "step": 60
+    },
+    {
+      "epoch": 0.0780550223928343,
+      "grad_norm": 6.606371121543283,
+      "learning_rate": 1e-06,
+      "loss": 0.8371,
+      "step": 61
+    },
+    {
+      "epoch": 0.07933461292386436,
+      "grad_norm": 6.8306985022146804,
+      "learning_rate": 1e-06,
+      "loss": 0.828,
+      "step": 62
+    },
+    {
+      "epoch": 0.08061420345489444,
+      "grad_norm": 5.212742827292372,
+      "learning_rate": 1e-06,
+      "loss": 0.8599,
+      "step": 63
+    },
+    {
+      "epoch": 0.0818937939859245,
+      "grad_norm": 5.622850966109007,
+      "learning_rate": 1e-06,
+      "loss": 0.788,
+      "step": 64
+    },
+    {
+      "epoch": 0.08317338451695458,
+      "grad_norm": 9.529617940821993,
+      "learning_rate": 1e-06,
+      "loss": 0.8103,
+      "step": 65
+    },
+    {
+      "epoch": 0.08445297504798464,
+      "grad_norm": 5.455455197247744,
+      "learning_rate": 1e-06,
+      "loss": 0.7755,
+      "step": 66
+    },
+    {
+      "epoch": 0.08573256557901472,
+      "grad_norm": 8.018750786419632,
+      "learning_rate": 1e-06,
+      "loss": 0.7498,
+      "step": 67
+    },
+    {
+      "epoch": 0.08701215611004479,
+      "grad_norm": 8.449729455793022,
+      "learning_rate": 1e-06,
+      "loss": 0.8448,
+      "step": 68
+    },
+    {
+      "epoch": 0.08829174664107485,
+      "grad_norm": 7.8137004713101375,
+      "learning_rate": 1e-06,
+      "loss": 0.7806,
+      "step": 69
+    },
+    {
+      "epoch": 0.08957133717210493,
+      "grad_norm": 9.822721590947728,
+      "learning_rate": 1e-06,
+      "loss": 0.7949,
+      "step": 70
+    },
+    {
+      "epoch": 0.09085092770313499,
+      "grad_norm": 6.947567524713278,
+      "learning_rate": 1e-06,
+      "loss": 0.7933,
+      "step": 71
+    },
+    {
+      "epoch": 0.09213051823416507,
+      "grad_norm": 7.980515698431891,
+      "learning_rate": 1e-06,
+      "loss": 0.7441,
+      "step": 72
+    },
+    {
+      "epoch": 0.09341010876519514,
+      "grad_norm": 7.325910029658637,
+      "learning_rate": 1e-06,
+      "loss": 0.7416,
+      "step": 73
+    },
+    {
+      "epoch": 0.09468969929622521,
+      "grad_norm": 6.89135596049996,
+      "learning_rate": 1e-06,
+      "loss": 0.8142,
+      "step": 74
+    },
+    {
+      "epoch": 0.09596928982725528,
+      "grad_norm": 6.2753522373579775,
+      "learning_rate": 1e-06,
+      "loss": 0.787,
+      "step": 75
+    },
+    {
+      "epoch": 0.09724888035828536,
+      "grad_norm": 4.982802320637717,
+      "learning_rate": 1e-06,
+      "loss": 0.8191,
+      "step": 76
+    },
+    {
+      "epoch": 0.09852847088931542,
+      "grad_norm": 7.835684340558416,
+      "learning_rate": 1e-06,
+      "loss": 0.8146,
+      "step": 77
+    },
+    {
+      "epoch": 0.09980806142034548,
+      "grad_norm": 4.907759995202286,
+      "learning_rate": 1e-06,
+      "loss": 0.7454,
+      "step": 78
+    },
+    {
+      "epoch": 0.10108765195137556,
+      "grad_norm": 10.288723075253866,
+      "learning_rate": 1e-06,
+      "loss": 0.7621,
+      "step": 79
+    },
+    {
+      "epoch": 0.10236724248240563,
+      "grad_norm": 5.62317987996513,
+      "learning_rate": 1e-06,
+      "loss": 0.8204,
+      "step": 80
+    },
+    {
+      "epoch": 0.1036468330134357,
+      "grad_norm": 4.995907815210038,
+      "learning_rate": 1e-06,
+      "loss": 0.8045,
+      "step": 81
+    },
+    {
+      "epoch": 0.10492642354446577,
+      "grad_norm": 10.522352516344274,
+      "learning_rate": 1e-06,
+      "loss": 0.805,
+      "step": 82
+    },
+    {
+      "epoch": 0.10620601407549585,
+      "grad_norm": 7.46755807435595,
+      "learning_rate": 1e-06,
+      "loss": 0.7453,
+      "step": 83
+    },
+    {
+      "epoch": 0.10748560460652591,
+      "grad_norm": 8.633484911919759,
+      "learning_rate": 1e-06,
+      "loss": 0.8208,
+      "step": 84
+    },
+    {
+      "epoch": 0.10876519513755598,
+      "grad_norm": 6.8324057400220095,
+      "learning_rate": 1e-06,
+      "loss": 0.8111,
+      "step": 85
+    },
+    {
+      "epoch": 0.11004478566858605,
+      "grad_norm": 6.8701793081881695,
+      "learning_rate": 1e-06,
+      "loss": 0.8145,
+      "step": 86
+    },
+    {
+      "epoch": 0.11132437619961612,
+      "grad_norm": 8.117987736637867,
+      "learning_rate": 1e-06,
+      "loss": 0.8156,
+      "step": 87
+    },
+    {
+      "epoch": 0.1126039667306462,
+      "grad_norm": 8.3839327799946,
+      "learning_rate": 1e-06,
+      "loss": 0.8166,
+      "step": 88
+    },
+    {
+      "epoch": 0.11388355726167626,
+      "grad_norm": 6.363789505259682,
+      "learning_rate": 1e-06,
+      "loss": 0.7668,
+      "step": 89
+    },
+    {
+      "epoch": 0.11516314779270634,
+      "grad_norm": 6.986172572258636,
+      "learning_rate": 1e-06,
+      "loss": 0.8175,
+      "step": 90
+    },
+    {
+      "epoch": 0.1164427383237364,
+      "grad_norm": 6.827182777059679,
+      "learning_rate": 1e-06,
+      "loss": 0.8067,
+      "step": 91
+    },
+    {
+      "epoch": 0.11772232885476648,
+      "grad_norm": 7.169862980822827,
+      "learning_rate": 1e-06,
+      "loss": 0.7934,
+      "step": 92
+    },
+    {
+      "epoch": 0.11900191938579655,
+      "grad_norm": 7.019166817832967,
+      "learning_rate": 1e-06,
+      "loss": 0.8251,
+      "step": 93
+    },
+    {
+      "epoch": 0.12028150991682661,
+      "grad_norm": 4.4498436110553525,
+      "learning_rate": 1e-06,
+      "loss": 0.7672,
+      "step": 94
+    },
+    {
+      "epoch": 0.12156110044785669,
+      "grad_norm": 5.6447641498186805,
+      "learning_rate": 1e-06,
+      "loss": 0.7806,
+      "step": 95
+    },
+    {
+      "epoch": 0.12284069097888675,
+      "grad_norm": 7.634792466964547,
+      "learning_rate": 1e-06,
+      "loss": 0.7951,
+      "step": 96
+    },
+    {
+      "epoch": 0.12412028150991683,
+      "grad_norm": 8.278290688129937,
+      "learning_rate": 1e-06,
+      "loss": 0.8001,
+      "step": 97
+    },
+    {
+      "epoch": 0.1253998720409469,
+      "grad_norm": 7.539458580302871,
+      "learning_rate": 1e-06,
+      "loss": 0.7795,
+      "step": 98
+    },
+    {
+      "epoch": 0.12667946257197696,
+      "grad_norm": 8.044471192568723,
+      "learning_rate": 1e-06,
+      "loss": 0.7834,
+      "step": 99
+    },
+    {
+      "epoch": 0.12795905310300704,
+      "grad_norm": 7.300758062129908,
+      "learning_rate": 1e-06,
+      "loss": 0.7861,
+      "step": 100
+    },
+    {
+      "epoch": 0.12923864363403711,
+      "grad_norm": 9.263316142132904,
+      "learning_rate": 1e-06,
+      "loss": 0.8373,
+      "step": 101
+    },
+    {
+      "epoch": 0.13051823416506717,
+      "grad_norm": 5.325027216463253,
+      "learning_rate": 1e-06,
+      "loss": 0.8164,
+      "step": 102
+    },
+    {
+      "epoch": 0.13179782469609724,
+      "grad_norm": 8.245222284340631,
+      "learning_rate": 1e-06,
+      "loss": 0.8552,
+      "step": 103
+    },
+    {
+      "epoch": 0.13307741522712732,
+      "grad_norm": 6.550681590613512,
+      "learning_rate": 1e-06,
+      "loss": 0.7507,
+      "step": 104
+    },
+    {
+      "epoch": 0.1343570057581574,
+      "grad_norm": 6.550681590613512,
+      "learning_rate": 1e-06,
+      "loss": 0.846,
+      "step": 105
+    },
+    {
+      "epoch": 0.13563659628918745,
+      "grad_norm": 7.90848861437755,
+      "learning_rate": 1e-06,
+      "loss": 0.7684,
+      "step": 106
+    },
+    {
+      "epoch": 0.13691618682021753,
+      "grad_norm": 6.444297457135872,
+      "learning_rate": 1e-06,
+      "loss": 0.7811,
+      "step": 107
+    },
+    {
+      "epoch": 0.1381957773512476,
+      "grad_norm": 7.134793533494606,
+      "learning_rate": 1e-06,
+      "loss": 0.7903,
+      "step": 108
+    },
+    {
+      "epoch": 0.13947536788227768,
+      "grad_norm": 7.037414300818179,
+      "learning_rate": 1e-06,
+      "loss": 0.8092,
+      "step": 109
+    },
+    {
+      "epoch": 0.14075495841330773,
+      "grad_norm": 7.438332279031611,
+      "learning_rate": 1e-06,
+      "loss": 0.8082,
+      "step": 110
+    },
+    {
+      "epoch": 0.1420345489443378,
+      "grad_norm": 5.256621761080708,
+      "learning_rate": 1e-06,
+      "loss": 0.7828,
+      "step": 111
+    },
+    {
+      "epoch": 0.1433141394753679,
+      "grad_norm": 9.216063856974321,
+      "learning_rate": 1e-06,
+      "loss": 0.7603,
+      "step": 112
+    },
+    {
+      "epoch": 0.14459373000639794,
+      "grad_norm": 4.568054662275053,
+      "learning_rate": 1e-06,
+      "loss": 0.8182,
+      "step": 113
+    },
+    {
+      "epoch": 0.14587332053742802,
+      "grad_norm": 6.174106514652569,
+      "learning_rate": 1e-06,
+      "loss": 0.8436,
+      "step": 114
+    },
+    {
+      "epoch": 0.1471529110684581,
+      "grad_norm": 6.4256394555983,
+      "learning_rate": 1e-06,
+      "loss": 0.7798,
+      "step": 115
+    },
+    {
+      "epoch": 0.14843250159948818,
+      "grad_norm": 4.73786791509776,
+      "learning_rate": 1e-06,
+      "loss": 0.8081,
+      "step": 116
+    },
+    {
+      "epoch": 0.14971209213051823,
+      "grad_norm": 7.741686218287495,
+      "learning_rate": 1e-06,
+      "loss": 0.7907,
+      "step": 117
+    },
+    {
+      "epoch": 0.1509916826615483,
+      "grad_norm": 7.365715010848928,
+      "learning_rate": 1e-06,
+      "loss": 0.7243,
+      "step": 118
+    },
+    {
+      "epoch": 0.15227127319257838,
+      "grad_norm": 7.587984974170548,
+      "learning_rate": 1e-06,
+      "loss": 0.8043,
+      "step": 119
+    },
+    {
+      "epoch": 0.15355086372360843,
+      "grad_norm": 5.771732630681212,
+      "learning_rate": 1e-06,
+      "loss": 0.7885,
+      "step": 120
+    },
+    {
+      "epoch": 0.1548304542546385,
+      "grad_norm": 6.401826450951003,
+      "learning_rate": 1e-06,
+      "loss": 0.8184,
+      "step": 121
+    },
+    {
+      "epoch": 0.1561100447856686,
+      "grad_norm": 8.535192165705647,
+      "learning_rate": 1e-06,
+      "loss": 0.7874,
+      "step": 122
+    },
+    {
+      "epoch": 0.15738963531669867,
+      "grad_norm": 7.843728770907306,
+      "learning_rate": 1e-06,
+      "loss": 0.7729,
+      "step": 123
+    },
+    {
+      "epoch": 0.15866922584772872,
+      "grad_norm": 7.095712787500253,
+      "learning_rate": 1e-06,
+      "loss": 0.7789,
+      "step": 124
+    },
+    {
+      "epoch": 0.1599488163787588,
+      "grad_norm": 7.752152352158962,
+      "learning_rate": 1e-06,
+      "loss": 0.8182,
+      "step": 125
+    },
+    {
+      "epoch": 0.16122840690978887,
+      "grad_norm": 6.2325208637624465,
+      "learning_rate": 1e-06,
+      "loss": 0.8247,
+      "step": 126
+    },
+    {
+      "epoch": 0.16250799744081892,
+      "grad_norm": 7.526472899608975,
+      "learning_rate": 1e-06,
+      "loss": 0.7844,
+      "step": 127
+    },
+    {
+      "epoch": 0.163787587971849,
+      "grad_norm": 7.202038141812327,
+      "learning_rate": 1e-06,
+      "loss": 0.7573,
+      "step": 128
+    },
+    {
+      "epoch": 0.16506717850287908,
+      "grad_norm": 7.192430075501812,
+      "learning_rate": 1e-06,
+      "loss": 0.7735,
+      "step": 129
+    },
+    {
+      "epoch": 0.16634676903390916,
+      "grad_norm": 8.720337013118497,
+      "learning_rate": 1e-06,
+      "loss": 0.7781,
+      "step": 130
+    },
+    {
+      "epoch": 0.1676263595649392,
+      "grad_norm": 6.058354804985836,
+      "learning_rate": 1e-06,
+      "loss": 0.7347,
+      "step": 131
+    },
+    {
+      "epoch": 0.1689059500959693,
+      "grad_norm": 7.5453944590015904,
+      "learning_rate": 1e-06,
+      "loss": 0.7436,
+      "step": 132
+    },
+    {
+      "epoch": 0.17018554062699937,
+      "grad_norm": 6.640381817037646,
+      "learning_rate": 1e-06,
+      "loss": 0.8335,
+      "step": 133
+    },
+    {
+      "epoch": 0.17146513115802944,
+      "grad_norm": 7.951553236514403,
+      "learning_rate": 1e-06,
+      "loss": 0.7894,
+      "step": 134
+    },
+    {
+      "epoch": 0.1727447216890595,
+      "grad_norm": 7.350739646056547,
+      "learning_rate": 1e-06,
+      "loss": 0.7792,
+      "step": 135
+    },
+    {
+      "epoch": 0.17402431222008957,
+      "grad_norm": 8.330334083061526,
+      "learning_rate": 1e-06,
+      "loss": 0.819,
+      "step": 136
+    },
+    {
+      "epoch": 0.17530390275111965,
+      "grad_norm": 7.334034791070249,
+      "learning_rate": 1e-06,
+      "loss": 0.8337,
+      "step": 137
+    },
+    {
+      "epoch": 0.1765834932821497,
+      "grad_norm": 6.599069675350379,
+      "learning_rate": 1e-06,
+      "loss": 0.7787,
+      "step": 138
+    },
+    {
+      "epoch": 0.17786308381317978,
+      "grad_norm": 6.456525852097127,
+      "learning_rate": 1e-06,
+      "loss": 0.8157,
+      "step": 139
+    },
+    {
+      "epoch": 0.17914267434420986,
+      "grad_norm": 6.788228100755149,
+      "learning_rate": 1e-06,
+      "loss": 0.7914,
+      "step": 140
+    },
+    {
+      "epoch": 0.18042226487523993,
+      "grad_norm": 6.888485293888954,
+      "learning_rate": 1e-06,
+      "loss": 0.7916,
+      "step": 141
+    },
+    {
+      "epoch": 0.18170185540626999,
+      "grad_norm": 7.545211025585791,
+      "learning_rate": 1e-06,
+      "loss": 0.7741,
+      "step": 142
+    },
+    {
+      "epoch": 0.18298144593730006,
+      "grad_norm": 9.288770916532101,
+      "learning_rate": 1e-06,
+      "loss": 0.8248,
+      "step": 143
+    },
+    {
+      "epoch": 0.18426103646833014,
+      "grad_norm": 7.640115412191106,
+      "learning_rate": 1e-06,
+      "loss": 0.7814,
+      "step": 144
+    },
+    {
+      "epoch": 0.1855406269993602,
+      "grad_norm": 7.670889764158031,
+      "learning_rate": 1e-06,
+      "loss": 0.7951,
+      "step": 145
+    },
+    {
+      "epoch": 0.18682021753039027,
+      "grad_norm": 7.866432469202212,
+      "learning_rate": 1e-06,
+      "loss": 0.7627,
+      "step": 146
+    },
+    {
+      "epoch": 0.18809980806142035,
+      "grad_norm": 5.802134336738597,
+      "learning_rate": 1e-06,
+      "loss": 0.7778,
+      "step": 147
+    },
+    {
+      "epoch": 0.18937939859245043,
+      "grad_norm": 7.725547687000397,
+      "learning_rate": 1e-06,
+      "loss": 0.8242,
+      "step": 148
+    },
+    {
+      "epoch": 0.19065898912348048,
+      "grad_norm": 8.073802472014968,
+      "learning_rate": 1e-06,
+      "loss": 0.7689,
+      "step": 149
+    },
+    {
+      "epoch": 0.19193857965451055,
+      "grad_norm": 6.877806989199121,
+      "learning_rate": 1e-06,
+      "loss": 0.7223,
+      "step": 150
+    },
+    {
+      "epoch": 0.19321817018554063,
+      "grad_norm": 7.0732279021828095,
+      "learning_rate": 1e-06,
+      "loss": 0.7584,
+      "step": 151
+    },
+    {
+      "epoch": 0.1944977607165707,
+      "grad_norm": 6.337488058479585,
+      "learning_rate": 1e-06,
+      "loss": 0.792,
+      "step": 152
+    },
+    {
+      "epoch": 0.19577735124760076,
+      "grad_norm": 6.626969802685368,
+      "learning_rate": 1e-06,
+      "loss": 0.8107,
+      "step": 153
+    },
+    {
+      "epoch": 0.19705694177863084,
+      "grad_norm": 6.066629155455074,
+      "learning_rate": 1e-06,
+      "loss": 0.7038,
+      "step": 154
+    },
+    {
+      "epoch": 0.19833653230966092,
+      "grad_norm": 6.445326142477278,
+      "learning_rate": 1e-06,
+      "loss": 0.8382,
+      "step": 155
+    },
+    {
+      "epoch": 0.19961612284069097,
+      "grad_norm": 7.90587988903203,
+      "learning_rate": 1e-06,
+      "loss": 0.7599,
+      "step": 156
+    },
+    {
+      "epoch": 0.20089571337172105,
+      "grad_norm": 8.27914376964039,
+      "learning_rate": 1e-06,
+      "loss": 0.8342,
+      "step": 157
+    },
+    {
+      "epoch": 0.20217530390275112,
+      "grad_norm": 6.529243859359478,
+      "learning_rate": 1e-06,
+      "loss": 0.7201,
+      "step": 158
+    },
+    {
+      "epoch": 0.2034548944337812,
+      "grad_norm": 4.822625148589539,
+      "learning_rate": 1e-06,
+      "loss": 0.8174,
+      "step": 159
+    },
+    {
+      "epoch": 0.20473448496481125,
+      "grad_norm": 8.121335108943036,
+      "learning_rate": 1e-06,
+      "loss": 0.8225,
+      "step": 160
+    },
+    {
+      "epoch": 0.20601407549584133,
+      "grad_norm": 9.579936292954068,
+      "learning_rate": 1e-06,
+      "loss": 0.7717,
+      "step": 161
+    },
+    {
+      "epoch": 0.2072936660268714,
+      "grad_norm": 6.311352322492659,
+      "learning_rate": 1e-06,
+      "loss": 0.8153,
+      "step": 162
+    },
+    {
+      "epoch": 0.20857325655790146,
+      "grad_norm": 7.037046382674988,
+      "learning_rate": 1e-06,
+      "loss": 0.8861,
+      "step": 163
+    },
+    {
+      "epoch": 0.20985284708893154,
+      "grad_norm": 7.326131521030794,
+      "learning_rate": 1e-06,
+      "loss": 0.7908,
+      "step": 164
+    },
+    {
+      "epoch": 0.21113243761996162,
+      "grad_norm": 7.355477087729125,
+      "learning_rate": 1e-06,
+      "loss": 0.7887,
+      "step": 165
+    },
+    {
+      "epoch": 0.2124120281509917,
+      "grad_norm": 6.337886462426987,
+      "learning_rate": 1e-06,
+      "loss": 0.802,
+      "step": 166
+    },
+    {
+      "epoch": 0.21369161868202174,
+      "grad_norm": 8.260497933750369,
+      "learning_rate": 1e-06,
+      "loss": 0.8322,
+      "step": 167
+    },
+    {
+      "epoch": 0.21497120921305182,
+      "grad_norm": 7.910338446585131,
+      "learning_rate": 1e-06,
+      "loss": 0.7926,
+      "step": 168
+    },
+    {
+      "epoch": 0.2162507997440819,
+      "grad_norm": 6.231993464284912,
+      "learning_rate": 1e-06,
+      "loss": 0.77,
+      "step": 169
+    },
+    {
+      "epoch": 0.21753039027511195,
+      "grad_norm": 5.924083589558338,
+      "learning_rate": 1e-06,
+      "loss": 0.7578,
+      "step": 170
+    },
+    {
+      "epoch": 0.21880998080614203,
+      "grad_norm": 8.470799887936016,
+      "learning_rate": 1e-06,
+      "loss": 0.8162,
+      "step": 171
+    },
+    {
+      "epoch": 0.2200895713371721,
+      "grad_norm": 6.6294804569369505,
+      "learning_rate": 1e-06,
+      "loss": 0.7851,
+      "step": 172
+    },
+    {
+      "epoch": 0.22136916186820219,
+      "grad_norm": 7.997180385319155,
+      "learning_rate": 1e-06,
+      "loss": 0.8114,
+      "step": 173
+    },
+    {
+      "epoch": 0.22264875239923224,
+      "grad_norm": 7.791632179439827,
+      "learning_rate": 1e-06,
+      "loss": 0.7401,
+      "step": 174
+    },
+    {
+      "epoch": 0.22392834293026231,
+      "grad_norm": 6.199668909735166,
+      "learning_rate": 1e-06,
+      "loss": 0.7735,
+      "step": 175
+    },
+    {
+      "epoch": 0.2252079334612924,
+      "grad_norm": 6.619234447014463,
+      "learning_rate": 1e-06,
+      "loss": 0.7793,
+      "step": 176
+    },
+    {
+      "epoch": 0.22648752399232247,
+      "grad_norm": 9.773393125605084,
+      "learning_rate": 1e-06,
+      "loss": 0.7642,
+      "step": 177
+    },
+    {
+      "epoch": 0.22776711452335252,
+      "grad_norm": 9.163385552636624,
+      "learning_rate": 1e-06,
+      "loss": 0.7928,
+      "step": 178
+    },
+    {
+      "epoch": 0.2290467050543826,
+      "grad_norm": 6.0165832914960555,
+      "learning_rate": 1e-06,
+      "loss": 0.7825,
+      "step": 179
+    },
+    {
+      "epoch": 0.23032629558541268,
+      "grad_norm": 5.246675507674416,
+      "learning_rate": 1e-06,
+      "loss": 0.8118,
+      "step": 180
+    },
+    {
+      "epoch": 0.23160588611644273,
+      "grad_norm": 7.016788642535557,
+      "learning_rate": 1e-06,
+      "loss": 0.7651,
+      "step": 181
+    },
+    {
+      "epoch": 0.2328854766474728,
+      "grad_norm": 9.019250028197792,
+      "learning_rate": 1e-06,
+      "loss": 0.8088,
+      "step": 182
+    },
+    {
+      "epoch": 0.23416506717850288,
+      "grad_norm": 7.375872542349399,
+      "learning_rate": 1e-06,
+      "loss": 0.8251,
+      "step": 183
+    },
+    {
+      "epoch": 0.23544465770953296,
+      "grad_norm": 5.818648406690168,
+      "learning_rate": 1e-06,
+      "loss": 0.806,
+      "step": 184
+    },
+    {
+      "epoch": 0.236724248240563,
+      "grad_norm": 7.606309270350813,
+      "learning_rate": 1e-06,
+      "loss": 0.7952,
+      "step": 185
+    },
+    {
+      "epoch": 0.2380038387715931,
+      "grad_norm": 7.797347524690805,
+      "learning_rate": 1e-06,
+      "loss": 0.7935,
+      "step": 186
+    },
+    {
+      "epoch": 0.23928342930262317,
+      "grad_norm": 5.874272353292162,
+      "learning_rate": 1e-06,
+      "loss": 0.7845,
+      "step": 187
+    },
+    {
+      "epoch": 0.24056301983365322,
+      "grad_norm": 8.320808739209538,
+      "learning_rate": 1e-06,
+      "loss": 0.7811,
+      "step": 188
+    },
+    {
+      "epoch": 0.2418426103646833,
+      "grad_norm": 5.079233982279212,
+      "learning_rate": 1e-06,
+      "loss": 0.7224,
+      "step": 189
+    },
+    {
+      "epoch": 0.24312220089571338,
+      "grad_norm": 6.550967523791025,
+      "learning_rate": 1e-06,
+      "loss": 0.7407,
+      "step": 190
+    },
+    {
+      "epoch": 0.24440179142674345,
+      "grad_norm": 7.768447074303632,
+      "learning_rate": 1e-06,
+      "loss": 0.829,
+      "step": 191
+    },
+    {
+      "epoch": 0.2456813819577735,
+      "grad_norm": 8.714832353496304,
+      "learning_rate": 1e-06,
+      "loss": 0.7588,
+      "step": 192
+    },
+    {
+      "epoch": 0.24696097248880358,
+      "grad_norm": 6.73092190059146,
+      "learning_rate": 1e-06,
+      "loss": 0.7788,
+      "step": 193
+    },
+    {
+      "epoch": 0.24824056301983366,
+      "grad_norm": 5.676271579988171,
+      "learning_rate": 1e-06,
+      "loss": 0.8037,
+      "step": 194
+    },
+    {
+      "epoch": 0.2495201535508637,
+      "grad_norm": 6.6188640201378615,
+      "learning_rate": 1e-06,
+      "loss": 0.8195,
+      "step": 195
+    },
+    {
+      "epoch": 0.2507997440818938,
+      "grad_norm": 5.405765781684467,
+      "learning_rate": 1e-06,
+      "loss": 0.8184,
+      "step": 196
+    },
+    {
+      "epoch": 0.25207933461292387,
+      "grad_norm": 9.10747742283157,
+      "learning_rate": 1e-06,
+      "loss": 0.7821,
+      "step": 197
+    },
+    {
+      "epoch": 0.2533589251439539,
+      "grad_norm": 6.844431032906677,
+      "learning_rate": 1e-06,
+      "loss": 0.8294,
+      "step": 198
+    },
+    {
+      "epoch": 0.254638515674984,
+      "grad_norm": 9.932158997750424,
+      "learning_rate": 1e-06,
+      "loss": 0.8153,
+      "step": 199
+    },
+    {
+      "epoch": 0.2559181062060141,
+      "grad_norm": 6.403898541067495,
+      "learning_rate": 1e-06,
+      "loss": 0.7031,
+      "step": 200
+    },
+    {
+      "epoch": 0.2571976967370441,
+      "grad_norm": 7.909532496178034,
+      "learning_rate": 1e-06,
+      "loss": 0.848,
+      "step": 201
+    },
+    {
+      "epoch": 0.25847728726807423,
+      "grad_norm": 7.673559416473428,
+      "learning_rate": 1e-06,
+      "loss": 0.7993,
+      "step": 202
+    },
+    {
+      "epoch": 0.2597568777991043,
+      "grad_norm": 6.178257371465311,
+      "learning_rate": 1e-06,
+      "loss": 0.8202,
+      "step": 203
+    },
+    {
+      "epoch": 0.26103646833013433,
+      "grad_norm": 7.059253776194904,
+      "learning_rate": 1e-06,
+      "loss": 0.8072,
+      "step": 204
+    },
+    {
+      "epoch": 0.26231605886116444,
+      "grad_norm": 6.547343334151184,
+      "learning_rate": 1e-06,
+      "loss": 0.7651,
+      "step": 205
+    },
+    {
+      "epoch": 0.2635956493921945,
+      "grad_norm": 8.253797926803523,
+      "learning_rate": 1e-06,
+      "loss": 0.8177,
+      "step": 206
+    },
+    {
+      "epoch": 0.2648752399232246,
+      "grad_norm": 7.551540387026114,
+      "learning_rate": 1e-06,
+      "loss": 0.7418,
+      "step": 207
+    },
+    {
+      "epoch": 0.26615483045425464,
+      "grad_norm": 7.636093231037009,
+      "learning_rate": 1e-06,
+      "loss": 0.8004,
+      "step": 208
+    },
+    {
+      "epoch": 0.2674344209852847,
+      "grad_norm": 5.400587004775909,
+      "learning_rate": 1e-06,
+      "loss": 0.8107,
+      "step": 209
+    },
+    {
+      "epoch": 0.2687140115163148,
+      "grad_norm": 5.151786220217282,
+      "learning_rate": 1e-06,
+      "loss": 0.7955,
+      "step": 210
+    },
+    {
+      "epoch": 0.26999360204734485,
+      "grad_norm": 8.30557779383527,
+      "learning_rate": 1e-06,
+      "loss": 0.8368,
+      "step": 211
+    },
+    {
+      "epoch": 0.2712731925783749,
+      "grad_norm": 6.114205069492278,
+      "learning_rate": 1e-06,
+      "loss": 0.771,
+      "step": 212
+    },
+    {
+      "epoch": 0.272552783109405,
+      "grad_norm": 10.696243957211696,
+      "learning_rate": 1e-06,
+      "loss": 0.7852,
+      "step": 213
+    },
+    {
+      "epoch": 0.27383237364043506,
+      "grad_norm": 6.9433294305091975,
+      "learning_rate": 1e-06,
+      "loss": 0.7744,
+      "step": 214
+    },
+    {
+      "epoch": 0.2751119641714651,
+      "grad_norm": 7.654829324955112,
+      "learning_rate": 1e-06,
+      "loss": 0.7404,
+      "step": 215
+    },
+    {
+      "epoch": 0.2763915547024952,
+      "grad_norm": 6.494955389442734,
+      "learning_rate": 1e-06,
+      "loss": 0.8223,
+      "step": 216
+    },
+    {
+      "epoch": 0.27767114523352526,
+      "grad_norm": 7.107284889364752,
+      "learning_rate": 1e-06,
+      "loss": 0.822,
+      "step": 217
+    },
+    {
+      "epoch": 0.27895073576455537,
+      "grad_norm": 7.907183553741368,
+      "learning_rate": 1e-06,
+      "loss": 0.7934,
+      "step": 218
+    },
+    {
+      "epoch": 0.2802303262955854,
+      "grad_norm": 5.52056775091328,
+      "learning_rate": 1e-06,
+      "loss": 0.8091,
+      "step": 219
+    },
+    {
+      "epoch": 0.28150991682661547,
+      "grad_norm": 7.963537730618335,
+      "learning_rate": 1e-06,
+      "loss": 0.7148,
+      "step": 220
+    },
+    {
+      "epoch": 0.2827895073576456,
+      "grad_norm": 7.215918423312781,
+      "learning_rate": 1e-06,
+      "loss": 0.7871,
+      "step": 221
+    },
+    {
+      "epoch": 0.2840690978886756,
+      "grad_norm": 5.8336415232592405,
+      "learning_rate": 1e-06,
+      "loss": 0.8033,
+      "step": 222
+    },
+    {
+      "epoch": 0.2853486884197057,
+      "grad_norm": 4.778862976421255,
+      "learning_rate": 1e-06,
+      "loss": 0.8063,
+      "step": 223
+    },
+    {
+      "epoch": 0.2866282789507358,
+      "grad_norm": 7.775393129547543,
+      "learning_rate": 1e-06,
+      "loss": 0.7901,
+      "step": 224
+    },
+    {
+      "epoch": 0.28790786948176583,
+      "grad_norm": 7.307295898402273,
+      "learning_rate": 1e-06,
+      "loss": 0.8098,
+      "step": 225
+    },
+    {
+      "epoch": 0.2891874600127959,
+      "grad_norm": 6.831298129177906,
+      "learning_rate": 1e-06,
+      "loss": 0.7996,
+      "step": 226
+    },
+    {
+      "epoch": 0.290467050543826,
+      "grad_norm": 8.306686778851839,
+      "learning_rate": 1e-06,
+      "loss": 0.8024,
+      "step": 227
+    },
+    {
+      "epoch": 0.29174664107485604,
+      "grad_norm": 7.570436186973826,
+      "learning_rate": 1e-06,
+      "loss": 0.8119,
+      "step": 228
+    },
+    {
+      "epoch": 0.2930262316058861,
+      "grad_norm": 7.185463935695015,
+      "learning_rate": 1e-06,
+      "loss": 0.8162,
+      "step": 229
+    },
+    {
+      "epoch": 0.2943058221369162,
+      "grad_norm": 6.760391977865229,
+      "learning_rate": 1e-06,
+      "loss": 0.8164,
+      "step": 230
+    },
+    {
+      "epoch": 0.29558541266794625,
+      "grad_norm": 6.782883581846931,
+      "learning_rate": 1e-06,
+      "loss": 0.7475,
+      "step": 231
+    },
+    {
+      "epoch": 0.29686500319897635,
+      "grad_norm": 9.46283507961913,
+      "learning_rate": 1e-06,
+      "loss": 0.799,
+      "step": 232
+    },
+    {
+      "epoch": 0.2981445937300064,
+      "grad_norm": 8.49811701492493,
+      "learning_rate": 1e-06,
+      "loss": 0.7839,
+      "step": 233
+    },
+    {
+      "epoch": 0.29942418426103645,
+      "grad_norm": 6.790587435998067,
+      "learning_rate": 1e-06,
+      "loss": 0.7885,
+      "step": 234
+    },
+    {
+      "epoch": 0.30070377479206656,
+      "grad_norm": 6.7489996452088805,
+      "learning_rate": 1e-06,
+      "loss": 0.7626,
+      "step": 235
+    },
+    {
+      "epoch": 0.3019833653230966,
+      "grad_norm": 6.852199644936773,
+      "learning_rate": 1e-06,
+      "loss": 0.7988,
+      "step": 236
+    },
+    {
+      "epoch": 0.30326295585412666,
+      "grad_norm": 7.0643792095969795,
+      "learning_rate": 1e-06,
+      "loss": 0.7971,
+      "step": 237
+    },
+    {
+      "epoch": 0.30454254638515676,
+      "grad_norm": 6.130903811478302,
+      "learning_rate": 1e-06,
+      "loss": 0.8152,
+      "step": 238
+    },
+    {
+      "epoch": 0.3058221369161868,
+      "grad_norm": 5.720469170642259,
+      "learning_rate": 1e-06,
+      "loss": 0.8251,
+      "step": 239
+    },
+    {
+      "epoch": 0.30710172744721687,
+      "grad_norm": 7.011222539895656,
+      "learning_rate": 1e-06,
+      "loss": 0.7603,
+      "step": 240
+    },
+    {
+      "epoch": 0.30838131797824697,
+      "grad_norm": 5.291882120263234,
+      "learning_rate": 1e-06,
+      "loss": 0.8287,
+      "step": 241
+    },
+    {
+      "epoch": 0.309660908509277,
+      "grad_norm": 6.372447951460974,
+      "learning_rate": 1e-06,
+      "loss": 0.7785,
+      "step": 242
+    },
+    {
+      "epoch": 0.31094049904030713,
+      "grad_norm": 7.849454858159149,
+      "learning_rate": 1e-06,
+      "loss": 0.7949,
+      "step": 243
+    },
+    {
+      "epoch": 0.3122200895713372,
+      "grad_norm": 8.889705215264089,
+      "learning_rate": 1e-06,
+      "loss": 0.8089,
+      "step": 244
+    },
+    {
+      "epoch": 0.31349968010236723,
+      "grad_norm": 8.803664220819146,
+      "learning_rate": 1e-06,
+      "loss": 0.8075,
+      "step": 245
+    },
+    {
+      "epoch": 0.31477927063339733,
+      "grad_norm": 9.614773330727509,
+      "learning_rate": 1e-06,
+      "loss": 0.7584,
+      "step": 246
+    },
+    {
+      "epoch": 0.3160588611644274,
+      "grad_norm": 6.428085926711957,
+      "learning_rate": 1e-06,
+      "loss": 0.7402,
+      "step": 247
+    },
+    {
+      "epoch": 0.31733845169545744,
+      "grad_norm": 4.98148145396834,
+      "learning_rate": 1e-06,
+      "loss": 0.8023,
+      "step": 248
+    },
+    {
+      "epoch": 0.31861804222648754,
+      "grad_norm": 6.6094138098767194,
+      "learning_rate": 1e-06,
+      "loss": 0.7614,
+      "step": 249
+    },
+    {
+      "epoch": 0.3198976327575176,
+      "grad_norm": 5.269633819493451,
+      "learning_rate": 1e-06,
+      "loss": 0.7965,
+      "step": 250
+    },
+    {
+      "epoch": 0.32117722328854764,
+      "grad_norm": 6.571757858941038,
+      "learning_rate": 1e-06,
+      "loss": 0.7932,
+      "step": 251
+    },
+    {
+      "epoch": 0.32245681381957775,
+      "grad_norm": 7.049344947110724,
+      "learning_rate": 1e-06,
+      "loss": 0.793,
+      "step": 252
+    },
+    {
+      "epoch": 0.3237364043506078,
+      "grad_norm": 7.157843136932305,
+      "learning_rate": 1e-06,
+      "loss": 0.7489,
+      "step": 253
+    },
+    {
+      "epoch": 0.32501599488163785,
+      "grad_norm": 6.16908844046291,
+      "learning_rate": 1e-06,
+      "loss": 0.7713,
+      "step": 254
+    },
+    {
+      "epoch": 0.32629558541266795,
+      "grad_norm": 6.428331496212471,
+      "learning_rate": 1e-06,
+      "loss": 0.8131,
+      "step": 255
+    },
+    {
+      "epoch": 0.327575175943698,
+      "grad_norm": 7.8337991305049055,
+      "learning_rate": 1e-06,
+      "loss": 0.8001,
+      "step": 256
+    },
+    {
+      "epoch": 0.3288547664747281,
+      "grad_norm": 6.691687484287362,
+      "learning_rate": 1e-06,
+      "loss": 0.8034,
+      "step": 257
+    },
+    {
+      "epoch": 0.33013435700575816,
+      "grad_norm": 5.929088206408611,
+      "learning_rate": 1e-06,
+      "loss": 0.82,
+      "step": 258
+    },
+    {
+      "epoch": 0.3314139475367882,
+      "grad_norm": 5.861805350801838,
+      "learning_rate": 1e-06,
+      "loss": 0.7529,
+      "step": 259
+    },
+    {
+      "epoch": 0.3326935380678183,
+      "grad_norm": 6.555900899466572,
+      "learning_rate": 1e-06,
+      "loss": 0.7447,
+      "step": 260
+    },
+    {
+      "epoch": 0.33397312859884837,
+      "grad_norm": 6.034254197669106,
+      "learning_rate": 1e-06,
+      "loss": 0.805,
+      "step": 261
+    },
+    {
+      "epoch": 0.3352527191298784,
+      "grad_norm": 7.92617483927141,
+      "learning_rate": 1e-06,
+      "loss": 0.733,
+      "step": 262
+    },
+    {
+      "epoch": 0.3365323096609085,
+      "grad_norm": 6.968943505797303,
+      "learning_rate": 1e-06,
+      "loss": 0.8505,
+      "step": 263
+    },
+    {
+      "epoch": 0.3378119001919386,
+      "grad_norm": 7.380648128833638,
+      "learning_rate": 1e-06,
+      "loss": 0.7778,
+      "step": 264
+    },
+    {
+      "epoch": 0.3390914907229686,
+      "grad_norm": 7.316088347130466,
+      "learning_rate": 1e-06,
+      "loss": 0.8073,
+      "step": 265
+    },
+    {
+      "epoch": 0.34037108125399873,
+      "grad_norm": 8.794207341761147,
+      "learning_rate": 1e-06,
+      "loss": 0.7877,
+      "step": 266
+    },
+    {
+      "epoch": 0.3416506717850288,
+      "grad_norm": 9.34322264342688,
+      "learning_rate": 1e-06,
+      "loss": 0.8173,
+      "step": 267
+    },
+    {
+      "epoch": 0.3429302623160589,
+      "grad_norm": 5.8813236086735,
+      "learning_rate": 1e-06,
+      "loss": 0.7405,
+      "step": 268
+    },
+    {
+      "epoch": 0.34420985284708894,
+      "grad_norm": 5.999202920109405,
+      "learning_rate": 1e-06,
+      "loss": 0.692,
+      "step": 269
+    },
+    {
+      "epoch": 0.345489443378119,
+      "grad_norm": 6.865101002809172,
+      "learning_rate": 1e-06,
+      "loss": 0.7634,
+      "step": 270
+    },
+    {
+      "epoch": 0.3467690339091491,
+      "grad_norm": 8.123026121672345,
+      "learning_rate": 1e-06,
+      "loss": 0.788,
+      "step": 271
+    },
+    {
+      "epoch": 0.34804862444017914,
+      "grad_norm": 6.39662919729574,
+      "learning_rate": 1e-06,
+      "loss": 0.816,
+      "step": 272
+    },
+    {
+      "epoch": 0.3493282149712092,
+      "grad_norm": 7.062001195330502,
+      "learning_rate": 1e-06,
+      "loss": 0.7813,
+      "step": 273
+    },
+    {
+      "epoch": 0.3506078055022393,
+      "grad_norm": 6.763534249865772,
+      "learning_rate": 1e-06,
+      "loss": 0.789,
+      "step": 274
+    },
+    {
+      "epoch": 0.35188739603326935,
+      "grad_norm": 7.739163698709465,
+      "learning_rate": 1e-06,
+      "loss": 0.7736,
+      "step": 275
+    },
+    {
+      "epoch": 0.3531669865642994,
+      "grad_norm": 5.81423260313367,
+      "learning_rate": 1e-06,
+      "loss": 0.8012,
+      "step": 276
+    },
+    {
+      "epoch": 0.3544465770953295,
+      "grad_norm": 7.9122831017547535,
+      "learning_rate": 1e-06,
+      "loss": 0.8255,
+      "step": 277
+    },
+    {
+      "epoch": 0.35572616762635956,
+      "grad_norm": 6.059998446203241,
+      "learning_rate": 1e-06,
+      "loss": 0.7475,
+      "step": 278
+    },
+    {
+      "epoch": 0.3570057581573896,
+      "grad_norm": 8.404605838917645,
+      "learning_rate": 1e-06,
+      "loss": 0.8124,
+      "step": 279
+    },
+    {
+      "epoch": 0.3582853486884197,
+      "grad_norm": 6.7210760611186595,
+      "learning_rate": 1e-06,
+      "loss": 0.7598,
+      "step": 280
+    },
+    {
+      "epoch": 0.35956493921944976,
+      "grad_norm": 4.82532519859772,
+      "learning_rate": 1e-06,
+      "loss": 0.817,
+      "step": 281
+    },
+    {
+      "epoch": 0.36084452975047987,
+      "grad_norm": 7.636512719230949,
+      "learning_rate": 1e-06,
+      "loss": 0.763,
+      "step": 282
+    },
+    {
+      "epoch": 0.3621241202815099,
+      "grad_norm": 7.8119054501544545,
+      "learning_rate": 1e-06,
+      "loss": 0.7812,
+      "step": 283
+    },
+    {
+      "epoch": 0.36340371081253997,
+      "grad_norm": 5.4164899252809375,
+      "learning_rate": 1e-06,
+      "loss": 0.7393,
+      "step": 284
+    },
+    {
+      "epoch": 0.3646833013435701,
+      "grad_norm": 5.87392878998198,
+      "learning_rate": 1e-06,
+      "loss": 0.8532,
+      "step": 285
+    },
+    {
+      "epoch": 0.3659628918746001,
+      "grad_norm": 6.284316989722169,
+      "learning_rate": 1e-06,
+      "loss": 0.761,
+      "step": 286
+    },
+    {
+      "epoch": 0.3672424824056302,
+      "grad_norm": 6.883977237795165,
+      "learning_rate": 1e-06,
+      "loss": 0.7836,
+      "step": 287
+    },
+    {
+      "epoch": 0.3685220729366603,
+      "grad_norm": 5.588418124671875,
+      "learning_rate": 1e-06,
+      "loss": 0.8011,
+      "step": 288
+    },
+    {
+      "epoch": 0.36980166346769033,
+      "grad_norm": 7.576103789041292,
+      "learning_rate": 1e-06,
+      "loss": 0.7773,
+      "step": 289
+    },
+    {
+      "epoch": 0.3710812539987204,
+      "grad_norm": 8.624795579392284,
+      "learning_rate": 1e-06,
+      "loss": 0.8399,
+      "step": 290
+    },
+    {
+      "epoch": 0.3723608445297505,
+      "grad_norm": 6.594057800655919,
+      "learning_rate": 1e-06,
+      "loss": 0.8456,
+      "step": 291
+    },
+    {
+      "epoch": 0.37364043506078054,
+      "grad_norm": 6.595240247256281,
+      "learning_rate": 1e-06,
+      "loss": 0.7627,
+      "step": 292
+    },
+    {
+      "epoch": 0.37492002559181065,
+      "grad_norm": 6.379438842306764,
+      "learning_rate": 1e-06,
+      "loss": 0.7959,
+      "step": 293
+    },
+    {
+      "epoch": 0.3761996161228407,
+      "grad_norm": 8.083408263688332,
+      "learning_rate": 1e-06,
+      "loss": 0.7597,
+      "step": 294
+    },
+    {
+      "epoch": 0.37747920665387075,
+      "grad_norm": 8.27539971058152,
+      "learning_rate": 1e-06,
+      "loss": 0.786,
+      "step": 295
+    },
+    {
+      "epoch": 0.37875879718490085,
+      "grad_norm": 10.057634209831933,
+      "learning_rate": 1e-06,
+      "loss": 0.8092,
+      "step": 296
+    },
+    {
+      "epoch": 0.3800383877159309,
+      "grad_norm": 6.312020344131469,
+      "learning_rate": 1e-06,
+      "loss": 0.7813,
+      "step": 297
+    },
+    {
+      "epoch": 0.38131797824696095,
+      "grad_norm": 8.143162748820128,
+      "learning_rate": 1e-06,
+      "loss": 0.7961,
+      "step": 298
+    },
+    {
+      "epoch": 0.38259756877799106,
+      "grad_norm": 4.749150587642253,
+      "learning_rate": 1e-06,
+      "loss": 0.7748,
+      "step": 299
+    },
+    {
+      "epoch": 0.3838771593090211,
+      "grad_norm": 5.923912585280056,
+      "learning_rate": 1e-06,
+      "loss": 0.7711,
+      "step": 300
+    },
+    {
+      "epoch": 0.38515674984005116,
+      "grad_norm": 4.2687507976173045,
+      "learning_rate": 1e-06,
+      "loss": 0.7561,
+      "step": 301
+    },
+    {
+      "epoch": 0.38643634037108127,
+      "grad_norm": 8.98567368711672,
+      "learning_rate": 1e-06,
+      "loss": 0.7659,
+      "step": 302
+    },
+    {
+      "epoch": 0.3877159309021113,
+      "grad_norm": 5.290175392311582,
+      "learning_rate": 1e-06,
+      "loss": 0.7721,
+      "step": 303
+    },
+    {
+      "epoch": 0.3889955214331414,
+      "grad_norm": 5.291107864881045,
+      "learning_rate": 1e-06,
+      "loss": 0.7969,
+      "step": 304
+    },
+    {
+      "epoch": 0.3902751119641715,
+      "grad_norm": 7.091021515980082,
+      "learning_rate": 1e-06,
+      "loss": 0.7929,
+      "step": 305
+    },
+    {
+      "epoch": 0.3915547024952015,
+      "grad_norm": 8.627290691534634,
+      "learning_rate": 1e-06,
+      "loss": 0.8328,
+      "step": 306
+    },
+    {
+      "epoch": 0.39283429302623163,
+      "grad_norm": 9.317984410614255,
+      "learning_rate": 1e-06,
+      "loss": 0.817,
+      "step": 307
+    },
+    {
+      "epoch": 0.3941138835572617,
+      "grad_norm": 5.628247376730313,
+      "learning_rate": 1e-06,
+      "loss": 0.7135,
+      "step": 308
+    },
+    {
+      "epoch": 0.39539347408829173,
+      "grad_norm": 6.333481671523346,
+      "learning_rate": 1e-06,
+      "loss": 0.8045,
+      "step": 309
+    },
+    {
+      "epoch": 0.39667306461932184,
+      "grad_norm": 7.457980862153026,
+      "learning_rate": 1e-06,
+      "loss": 0.8245,
+      "step": 310
+    },
+    {
+      "epoch": 0.3979526551503519,
+      "grad_norm": 5.9473606190607615,
+      "learning_rate": 1e-06,
+      "loss": 0.7651,
+      "step": 311
+    },
+    {
+      "epoch": 0.39923224568138194,
+      "grad_norm": 7.818764554679867,
+      "learning_rate": 1e-06,
+      "loss": 0.7465,
+      "step": 312
+    },
+    {
+      "epoch": 0.40051183621241204,
+      "grad_norm": 8.229002315074977,
+      "learning_rate": 1e-06,
+      "loss": 0.7923,
+      "step": 313
+    },
+    {
+      "epoch": 0.4017914267434421,
+      "grad_norm": 7.234175361673951,
+      "learning_rate": 1e-06,
+      "loss": 0.7644,
+      "step": 314
+    },
+    {
+      "epoch": 0.40307101727447214,
+      "grad_norm": 5.965671887146069,
+      "learning_rate": 1e-06,
+      "loss": 0.7337,
+      "step": 315
+    },
+    {
+      "epoch": 0.40435060780550225,
+      "grad_norm": 6.945331905802045,
+      "learning_rate": 1e-06,
+      "loss": 0.7925,
+      "step": 316
+    },
+    {
+      "epoch": 0.4056301983365323,
+      "grad_norm": 8.065671339249564,
+      "learning_rate": 1e-06,
+      "loss": 0.7658,
+      "step": 317
+    },
+    {
+      "epoch": 0.4069097888675624,
+      "grad_norm": 6.393769246209232,
+      "learning_rate": 1e-06,
+      "loss": 0.8614,
+      "step": 318
+    },
+    {
+      "epoch": 0.40818937939859246,
+      "grad_norm": 7.147423658936338,
+      "learning_rate": 1e-06,
+      "loss": 0.7123,
+      "step": 319
+    },
+    {
+      "epoch": 0.4094689699296225,
+      "grad_norm": 8.076214009517987,
+      "learning_rate": 1e-06,
+      "loss": 0.7581,
+      "step": 320
+    },
+    {
+      "epoch": 0.4107485604606526,
+      "grad_norm": 8.909521715423345,
+      "learning_rate": 1e-06,
+      "loss": 0.7946,
+      "step": 321
+    },
+    {
+      "epoch": 0.41202815099168266,
+      "grad_norm": 10.64788803529604,
+      "learning_rate": 1e-06,
+      "loss": 0.7665,
+      "step": 322
+    },
+    {
+      "epoch": 0.4133077415227127,
+      "grad_norm": 8.655179901402049,
+      "learning_rate": 1e-06,
+      "loss": 0.7581,
+      "step": 323
+    },
+    {
+      "epoch": 0.4145873320537428,
+      "grad_norm": 7.733264094313162,
+      "learning_rate": 1e-06,
+      "loss": 0.8251,
+      "step": 324
+    },
+    {
+      "epoch": 0.41586692258477287,
+      "grad_norm": 5.9694502855314395,
+      "learning_rate": 1e-06,
+      "loss": 0.78,
+      "step": 325
+    },
+    {
+      "epoch": 0.4171465131158029,
+      "grad_norm": 7.529668616359048,
+      "learning_rate": 1e-06,
+      "loss": 0.7508,
+      "step": 326
+    },
+    {
+      "epoch": 0.418426103646833,
+      "grad_norm": 7.577607198435007,
+      "learning_rate": 1e-06,
+      "loss": 0.7743,
+      "step": 327
+    },
+    {
+      "epoch": 0.4197056941778631,
+      "grad_norm": 8.844578188947295,
+      "learning_rate": 1e-06,
+      "loss": 0.7654,
+      "step": 328
+    },
+    {
+      "epoch": 0.4209852847088932,
+      "grad_norm": 4.741848582789355,
+      "learning_rate": 1e-06,
+      "loss": 0.7368,
+      "step": 329
+    },
+    {
+      "epoch": 0.42226487523992323,
+      "grad_norm": 7.875716037559371,
+      "learning_rate": 1e-06,
+      "loss": 0.7095,
+      "step": 330
+    },
+    {
+      "epoch": 0.4235444657709533,
+      "grad_norm": 5.961781924914345,
+      "learning_rate": 1e-06,
+      "loss": 0.7632,
+      "step": 331
+    },
+    {
+      "epoch": 0.4248240563019834,
+      "grad_norm": 6.733857061680772,
+      "learning_rate": 1e-06,
+      "loss": 0.7797,
+      "step": 332
+    },
+    {
+      "epoch": 0.42610364683301344,
+      "grad_norm": 6.039359845635748,
+      "learning_rate": 1e-06,
+      "loss": 0.7637,
+      "step": 333
+    },
+    {
+      "epoch": 0.4273832373640435,
+      "grad_norm": 6.068749010600414,
+      "learning_rate": 1e-06,
+      "loss": 0.7288,
+      "step": 334
+    },
+    {
+      "epoch": 0.4286628278950736,
+      "grad_norm": 5.1685501978308315,
+      "learning_rate": 1e-06,
+      "loss": 0.825,
+      "step": 335
+    },
+    {
+      "epoch": 0.42994241842610365,
+      "grad_norm": 9.0074866728123,
+      "learning_rate": 1e-06,
+      "loss": 0.8366,
+      "step": 336
+    },
+    {
+      "epoch": 0.4312220089571337,
+      "grad_norm": 8.893018414409717,
+      "learning_rate": 1e-06,
+      "loss": 0.8371,
+      "step": 337
+    },
+    {
+      "epoch": 0.4325015994881638,
+      "grad_norm": 6.645360183897912,
+      "learning_rate": 1e-06,
+      "loss": 0.7637,
+      "step": 338
+    },
+    {
+      "epoch": 0.43378119001919385,
+      "grad_norm": 8.544044203452453,
+      "learning_rate": 1e-06,
+      "loss": 0.8194,
+      "step": 339
+    },
+    {
+      "epoch": 0.4350607805502239,
+      "grad_norm": 6.5787800444737154,
+      "learning_rate": 1e-06,
+      "loss": 0.7949,
+      "step": 340
+    },
+    {
+      "epoch": 0.436340371081254,
+      "grad_norm": 5.792852209418329,
+      "learning_rate": 1e-06,
+      "loss": 0.8243,
+      "step": 341
+    },
+    {
+      "epoch": 0.43761996161228406,
+      "grad_norm": 6.220485442352441,
+      "learning_rate": 1e-06,
+      "loss": 0.7366,
+      "step": 342
+    },
+    {
+      "epoch": 0.43889955214331416,
+      "grad_norm": 7.130094322710521,
+      "learning_rate": 1e-06,
+      "loss": 0.8031,
+      "step": 343
+    },
+    {
+      "epoch": 0.4401791426743442,
+      "grad_norm": 5.440225608431488,
+      "learning_rate": 1e-06,
+      "loss": 0.7306,
+      "step": 344
+    },
+    {
+      "epoch": 0.44145873320537427,
+      "grad_norm": 8.804169996260807,
+      "learning_rate": 1e-06,
+      "loss": 0.8221,
+      "step": 345
+    },
+    {
+      "epoch": 0.44273832373640437,
+      "grad_norm": 8.09537578321888,
+      "learning_rate": 1e-06,
+      "loss": 0.7482,
+      "step": 346
+    },
+    {
+      "epoch": 0.4440179142674344,
+      "grad_norm": 6.423297012058344,
+      "learning_rate": 1e-06,
+      "loss": 0.8016,
+      "step": 347
+    },
+    {
+      "epoch": 0.44529750479846447,
+      "grad_norm": 6.673991148061816,
+      "learning_rate": 1e-06,
+      "loss": 0.8129,
+      "step": 348
+    },
+    {
+      "epoch": 0.4465770953294946,
+      "grad_norm": 5.901464139729259,
+      "learning_rate": 1e-06,
+      "loss": 0.7946,
+      "step": 349
+    },
+    {
+      "epoch": 0.44785668586052463,
+      "grad_norm": 6.5805627504325725,
+      "learning_rate": 1e-06,
+      "loss": 0.7997,
+      "step": 350
+    },
+    {
+      "epoch": 0.4491362763915547,
+      "grad_norm": 7.675399221463476,
+      "learning_rate": 1e-06,
+      "loss": 0.7595,
+      "step": 351
+    },
+    {
+      "epoch": 0.4504158669225848,
+      "grad_norm": 9.916609733980106,
+      "learning_rate": 1e-06,
+      "loss": 0.8285,
+      "step": 352
+    },
+    {
+      "epoch": 0.45169545745361483,
+      "grad_norm": 9.886041039317421,
+      "learning_rate": 1e-06,
+      "loss": 0.7649,
+      "step": 353
+    },
+    {
+      "epoch": 0.45297504798464494,
+      "grad_norm": 6.190855491103708,
+      "learning_rate": 1e-06,
+      "loss": 0.7377,
+      "step": 354
+    },
+    {
+      "epoch": 0.454254638515675,
+      "grad_norm": 6.293392128084022,
+      "learning_rate": 1e-06,
+      "loss": 0.7853,
+      "step": 355
+    },
+    {
+      "epoch": 0.45553422904670504,
+      "grad_norm": 5.433921429444888,
+      "learning_rate": 1e-06,
+      "loss": 0.7843,
+      "step": 356
+    },
+    {
+      "epoch": 0.45681381957773515,
+      "grad_norm": 8.248831349649747,
+      "learning_rate": 1e-06,
+      "loss": 0.8066,
+      "step": 357
+    },
+    {
+      "epoch": 0.4580934101087652,
+      "grad_norm": 7.748048783963599,
+      "learning_rate": 1e-06,
+      "loss": 0.8067,
+      "step": 358
+    },
+    {
+      "epoch": 0.45937300063979525,
+      "grad_norm": 7.852868139957132,
+      "learning_rate": 1e-06,
+      "loss": 0.8262,
+      "step": 359
+    },
+    {
+      "epoch": 0.46065259117082535,
+      "grad_norm": 9.365284874523109,
+      "learning_rate": 1e-06,
+      "loss": 0.827,
+      "step": 360
+    },
+    {
+      "epoch": 0.4619321817018554,
+      "grad_norm": 7.762780241155836,
+      "learning_rate": 1e-06,
+      "loss": 0.7338,
+      "step": 361
+    },
+    {
+      "epoch": 0.46321177223288545,
+      "grad_norm": 6.848645086702751,
+      "learning_rate": 1e-06,
+      "loss": 0.7962,
+      "step": 362
+    },
+    {
+      "epoch": 0.46449136276391556,
+      "grad_norm": 8.98061942310754,
+      "learning_rate": 1e-06,
+      "loss": 0.7859,
+      "step": 363
+    },
+    {
+      "epoch": 0.4657709532949456,
+      "grad_norm": 5.798556030258737,
+      "learning_rate": 1e-06,
+      "loss": 0.7295,
+      "step": 364
+    },
+    {
+      "epoch": 0.46705054382597566,
+      "grad_norm": 5.983125695897138,
+      "learning_rate": 1e-06,
+      "loss": 0.7835,
+      "step": 365
+    },
+    {
+      "epoch": 0.46833013435700577,
+      "grad_norm": 5.5248396747813295,
+      "learning_rate": 1e-06,
+      "loss": 0.7982,
+      "step": 366
+    },
+    {
+      "epoch": 0.4696097248880358,
+      "grad_norm": 8.35928664681611,
+      "learning_rate": 1e-06,
+      "loss": 0.8003,
+      "step": 367
+    },
+    {
+      "epoch": 0.4708893154190659,
+      "grad_norm": 7.493930739535596,
+      "learning_rate": 1e-06,
+      "loss": 0.7805,
+      "step": 368
+    },
+    {
+      "epoch": 0.472168905950096,
+      "grad_norm": 6.6698391067174825,
+      "learning_rate": 1e-06,
+      "loss": 0.7999,
+      "step": 369
+    },
+    {
+      "epoch": 0.473448496481126,
+      "grad_norm": 6.973229491938506,
+      "learning_rate": 1e-06,
+      "loss": 0.8023,
+      "step": 370
+    },
+    {
+      "epoch": 0.47472808701215613,
+      "grad_norm": 9.045697778329098,
+      "learning_rate": 1e-06,
+      "loss": 0.8176,
+      "step": 371
+    },
+    {
+      "epoch": 0.4760076775431862,
+      "grad_norm": 5.948249856746707,
+      "learning_rate": 1e-06,
+      "loss": 0.7522,
+      "step": 372
+    },
+    {
+      "epoch": 0.47728726807421623,
+      "grad_norm": 7.792940007997215,
+      "learning_rate": 1e-06,
+      "loss": 0.7776,
+      "step": 373
+    },
+    {
+      "epoch": 0.47856685860524634,
+      "grad_norm": 7.855543110276994,
+      "learning_rate": 1e-06,
+      "loss": 0.7866,
+      "step": 374
+    },
+    {
+      "epoch": 0.4798464491362764,
+      "grad_norm": 6.329540078094709,
+      "learning_rate": 1e-06,
+      "loss": 0.7318,
+      "step": 375
+    },
+    {
+      "epoch": 0.48112603966730644,
+      "grad_norm": 6.717563904808815,
+      "learning_rate": 1e-06,
+      "loss": 0.7797,
+      "step": 376
+    },
+    {
+      "epoch": 0.48240563019833654,
+      "grad_norm": 6.5471669209315015,
+      "learning_rate": 1e-06,
+      "loss": 0.8144,
+      "step": 377
+    },
+    {
+      "epoch": 0.4836852207293666,
+      "grad_norm": 5.207872745433186,
+      "learning_rate": 1e-06,
+      "loss": 0.7922,
+      "step": 378
+    },
+    {
+      "epoch": 0.4849648112603967,
+      "grad_norm": 7.301771887967449,
+      "learning_rate": 1e-06,
+      "loss": 0.7688,
+      "step": 379
+    },
+    {
+      "epoch": 0.48624440179142675,
+      "grad_norm": 8.838669728705156,
+      "learning_rate": 1e-06,
+      "loss": 0.8265,
+      "step": 380
+    },
+    {
+      "epoch": 0.4875239923224568,
+      "grad_norm": 5.86811851380072,
+      "learning_rate": 1e-06,
+      "loss": 0.751,
+      "step": 381
+    },
+    {
+      "epoch": 0.4888035828534869,
+      "grad_norm": 8.56195636457536,
+      "learning_rate": 1e-06,
+      "loss": 0.8054,
+      "step": 382
+    },
+    {
+      "epoch": 0.49008317338451696,
+      "grad_norm": 14.871163968700975,
+      "learning_rate": 1e-06,
+      "loss": 0.7945,
+      "step": 383
+    },
+    {
+      "epoch": 0.491362763915547,
+      "grad_norm": 6.516031580059405,
+      "learning_rate": 1e-06,
+      "loss": 0.7378,
+      "step": 384
+    },
+    {
+      "epoch": 0.4926423544465771,
+      "grad_norm": 6.595253942092295,
+      "learning_rate": 1e-06,
+      "loss": 0.8073,
+      "step": 385
+    },
+    {
+      "epoch": 0.49392194497760716,
+      "grad_norm": 6.460296858958053,
+      "learning_rate": 1e-06,
+      "loss": 0.7507,
+      "step": 386
+    },
+    {
+      "epoch": 0.4952015355086372,
+      "grad_norm": 7.153051130308482,
+      "learning_rate": 1e-06,
+      "loss": 0.7828,
+      "step": 387
+    },
+    {
+      "epoch": 0.4964811260396673,
+      "grad_norm": 6.904692605022299,
+      "learning_rate": 1e-06,
+      "loss": 0.747,
+      "step": 388
+    },
+    {
+      "epoch": 0.49776071657069737,
+      "grad_norm": 8.144792639066413,
+      "learning_rate": 1e-06,
+      "loss": 0.7926,
+      "step": 389
+    },
+    {
+      "epoch": 0.4990403071017274,
+      "grad_norm": 8.36313001057412,
+      "learning_rate": 1e-06,
+      "loss": 0.7594,
+      "step": 390
+    },
+    {
+      "epoch": 0.5003198976327575,
+      "grad_norm": 7.078371398738784,
+      "learning_rate": 1e-06,
+      "loss": 0.8028,
+      "step": 391
+    },
+    {
+      "epoch": 0.5015994881637876,
+      "grad_norm": 7.205935486683429,
+      "learning_rate": 1e-06,
+      "loss": 0.8492,
+      "step": 392
+    },
+    {
+      "epoch": 0.5028790786948176,
+      "grad_norm": 6.678111214729273,
+      "learning_rate": 1e-06,
+      "loss": 0.7912,
+      "step": 393
+    },
+    {
+      "epoch": 0.5041586692258477,
+      "grad_norm": 9.090417183915806,
+      "learning_rate": 1e-06,
+      "loss": 0.8078,
+      "step": 394
+    },
+    {
+      "epoch": 0.5054382597568778,
+      "grad_norm": 7.064723617844546,
+      "learning_rate": 1e-06,
+      "loss": 0.8588,
+      "step": 395
+    },
+    {
+      "epoch": 0.5067178502879078,
+      "grad_norm": 7.561405816290959,
+      "learning_rate": 1e-06,
+      "loss": 0.8215,
+      "step": 396
+    },
+    {
+      "epoch": 0.5079974408189379,
+      "grad_norm": 7.939910488683977,
+      "learning_rate": 1e-06,
+      "loss": 0.8073,
+      "step": 397
+    },
+    {
+      "epoch": 0.509277031349968,
+      "grad_norm": 5.6272155458736055,
+      "learning_rate": 1e-06,
+      "loss": 0.8082,
+      "step": 398
+    },
+    {
+      "epoch": 0.510556621880998,
+      "grad_norm": 5.30117982037291,
+      "learning_rate": 1e-06,
+      "loss": 0.7864,
+      "step": 399
+    },
+    {
+      "epoch": 0.5118362124120281,
+      "grad_norm": 7.554269745074449,
+      "learning_rate": 1e-06,
+      "loss": 0.8241,
+      "step": 400
+    },
+    {
+      "epoch": 0.5131158029430583,
+      "grad_norm": 8.231485491741653,
+      "learning_rate": 1e-06,
+      "loss": 0.807,
+      "step": 401
+    },
+    {
+      "epoch": 0.5143953934740882,
+      "grad_norm": 7.596597187664772,
+      "learning_rate": 1e-06,
+      "loss": 0.7594,
+      "step": 402
+    },
+    {
+      "epoch": 0.5156749840051184,
+      "grad_norm": 10.727806277662745,
+      "learning_rate": 1e-06,
+      "loss": 0.7835,
+      "step": 403
+    },
+    {
+      "epoch": 0.5169545745361485,
+      "grad_norm": 5.367732655198655,
+      "learning_rate": 1e-06,
+      "loss": 0.7975,
+      "step": 404
+    },
+    {
+      "epoch": 0.5182341650671785,
+      "grad_norm": 7.2356733301632765,
+      "learning_rate": 1e-06,
+      "loss": 0.8032,
+      "step": 405
+    },
+    {
+      "epoch": 0.5195137555982086,
+      "grad_norm": 6.923786587911698,
+      "learning_rate": 1e-06,
+      "loss": 0.7957,
+      "step": 406
+    },
+    {
+      "epoch": 0.5207933461292387,
+      "grad_norm": 7.51696614253174,
+      "learning_rate": 1e-06,
+      "loss": 0.8539,
+      "step": 407
+    },
+    {
+      "epoch": 0.5220729366602687,
+      "grad_norm": 6.850302731948679,
+      "learning_rate": 1e-06,
+      "loss": 0.744,
+      "step": 408
+    },
+    {
+      "epoch": 0.5233525271912988,
+      "grad_norm": 6.200227661656017,
+      "learning_rate": 1e-06,
+      "loss": 0.7739,
+      "step": 409
+    },
+    {
+      "epoch": 0.5246321177223289,
+      "grad_norm": 7.2028291747076825,
+      "learning_rate": 1e-06,
+      "loss": 0.8148,
+      "step": 410
+    },
+    {
+      "epoch": 0.525911708253359,
+      "grad_norm": 6.620362860990904,
+      "learning_rate": 1e-06,
+      "loss": 0.7502,
+      "step": 411
+    },
+    {
+      "epoch": 0.527191298784389,
+      "grad_norm": 6.513633637963036,
+      "learning_rate": 1e-06,
+      "loss": 0.7493,
+      "step": 412
+    },
+    {
+      "epoch": 0.5284708893154191,
+      "grad_norm": 11.57422451721062,
+      "learning_rate": 1e-06,
+      "loss": 0.7656,
+      "step": 413
+    },
+    {
+      "epoch": 0.5297504798464492,
+      "grad_norm": 7.191327066687489,
+      "learning_rate": 1e-06,
+      "loss": 0.7863,
+      "step": 414
+    },
+    {
+      "epoch": 0.5310300703774792,
+      "grad_norm": 6.076346758201362,
+      "learning_rate": 1e-06,
+      "loss": 0.8144,
+      "step": 415
+    },
+    {
+      "epoch": 0.5323096609085093,
+      "grad_norm": 7.7468240223881875,
+      "learning_rate": 1e-06,
+      "loss": 0.7726,
+      "step": 416
+    },
+    {
+      "epoch": 0.5335892514395394,
+      "grad_norm": 8.540728948358163,
+      "learning_rate": 1e-06,
+      "loss": 0.8088,
+      "step": 417
+    },
+    {
+      "epoch": 0.5348688419705694,
+      "grad_norm": 7.871084779621567,
+      "learning_rate": 1e-06,
+      "loss": 0.8036,
+      "step": 418
+    },
+    {
+      "epoch": 0.5361484325015995,
+      "grad_norm": 7.409198950153203,
+      "learning_rate": 1e-06,
+      "loss": 0.7921,
+      "step": 419
+    },
+    {
+      "epoch": 0.5374280230326296,
+      "grad_norm": 6.912830709151778,
+      "learning_rate": 1e-06,
+      "loss": 0.7604,
+      "step": 420
+    },
+    {
+      "epoch": 0.5387076135636596,
+      "grad_norm": 7.83112425014489,
+      "learning_rate": 1e-06,
+      "loss": 0.8328,
+      "step": 421
+    },
+    {
+      "epoch": 0.5399872040946897,
+      "grad_norm": 8.756260204129756,
+      "learning_rate": 1e-06,
+      "loss": 0.7963,
+      "step": 422
+    },
+    {
+      "epoch": 0.5412667946257198,
+      "grad_norm": 6.074968310345647,
+      "learning_rate": 1e-06,
+      "loss": 0.7769,
+      "step": 423
+    },
+    {
+      "epoch": 0.5425463851567498,
+      "grad_norm": 5.616314934070467,
+      "learning_rate": 1e-06,
+      "loss": 0.7395,
+      "step": 424
+    },
+    {
+      "epoch": 0.5438259756877799,
+      "grad_norm": 5.881643996160075,
+      "learning_rate": 1e-06,
+      "loss": 0.8111,
+      "step": 425
+    },
+    {
+      "epoch": 0.54510556621881,
+      "grad_norm": 9.178052498515173,
+      "learning_rate": 1e-06,
+      "loss": 0.7361,
+      "step": 426
+    },
+    {
+      "epoch": 0.54638515674984,
+      "grad_norm": 6.757778343575352,
+      "learning_rate": 1e-06,
+      "loss": 0.7841,
+      "step": 427
+    },
+    {
+      "epoch": 0.5476647472808701,
+      "grad_norm": 5.7211921563924895,
+      "learning_rate": 1e-06,
+      "loss": 0.8335,
+      "step": 428
+    },
+    {
+      "epoch": 0.5489443378119002,
+      "grad_norm": 6.329019464832709,
+      "learning_rate": 1e-06,
+      "loss": 0.7075,
+      "step": 429
+    },
+    {
+      "epoch": 0.5502239283429302,
+      "grad_norm": 7.380219432532545,
+      "learning_rate": 1e-06,
+      "loss": 0.8114,
+      "step": 430
+    },
+    {
+      "epoch": 0.5515035188739603,
+      "grad_norm": 6.47433229564834,
+      "learning_rate": 1e-06,
+      "loss": 0.8434,
+      "step": 431
+    },
+    {
+      "epoch": 0.5527831094049904,
+      "grad_norm": 8.352164806439534,
+      "learning_rate": 1e-06,
+      "loss": 0.7499,
+      "step": 432
+    },
+    {
+      "epoch": 0.5540626999360204,
+      "grad_norm": 6.665301065240502,
+      "learning_rate": 1e-06,
+      "loss": 0.8073,
+      "step": 433
+    },
+    {
+      "epoch": 0.5553422904670505,
+      "grad_norm": 9.243048193392237,
+      "learning_rate": 1e-06,
+      "loss": 0.7694,
+      "step": 434
+    },
+    {
+      "epoch": 0.5566218809980806,
+      "grad_norm": 6.919807685597431,
+      "learning_rate": 1e-06,
+      "loss": 0.7316,
+      "step": 435
+    },
+    {
+      "epoch": 0.5579014715291107,
+      "grad_norm": 7.577664754857968,
+      "learning_rate": 1e-06,
+      "loss": 0.743,
+      "step": 436
+    },
+    {
+      "epoch": 0.5591810620601407,
+      "grad_norm": 6.874903774934258,
+      "learning_rate": 1e-06,
+      "loss": 0.7914,
+      "step": 437
+    },
+    {
+      "epoch": 0.5604606525911708,
+      "grad_norm": 5.498823924337238,
+      "learning_rate": 1e-06,
+      "loss": 0.8721,
+      "step": 438
+    },
+    {
+      "epoch": 0.5617402431222009,
+      "grad_norm": 7.793463529553969,
+      "learning_rate": 1e-06,
+      "loss": 0.7947,
+      "step": 439
+    },
+    {
+      "epoch": 0.5630198336532309,
+      "grad_norm": 7.902602188906223,
+      "learning_rate": 1e-06,
+      "loss": 0.7892,
+      "step": 440
+    },
+    {
+      "epoch": 0.564299424184261,
+      "grad_norm": 7.1949939393541404,
+      "learning_rate": 1e-06,
+      "loss": 0.7998,
+      "step": 441
+    },
+    {
+      "epoch": 0.5655790147152912,
+      "grad_norm": 6.702220512033791,
+      "learning_rate": 1e-06,
+      "loss": 0.7855,
+      "step": 442
+    },
+    {
+      "epoch": 0.5668586052463211,
+      "grad_norm": 7.332063322095238,
+      "learning_rate": 1e-06,
+      "loss": 0.7838,
+      "step": 443
+    },
+    {
+      "epoch": 0.5681381957773513,
+      "grad_norm": 6.449997885936394,
+      "learning_rate": 1e-06,
+      "loss": 0.7549,
+      "step": 444
+    },
+    {
+      "epoch": 0.5694177863083814,
+      "grad_norm": 8.43202091261427,
+      "learning_rate": 1e-06,
+      "loss": 0.7911,
+      "step": 445
+    },
+    {
+      "epoch": 0.5706973768394114,
+      "grad_norm": 6.57541235311022,
+      "learning_rate": 1e-06,
+      "loss": 0.8078,
+      "step": 446
+    },
+    {
+      "epoch": 0.5719769673704415,
+      "grad_norm": 7.593983738868652,
+      "learning_rate": 1e-06,
+      "loss": 0.7921,
+      "step": 447
+    },
+    {
+      "epoch": 0.5732565579014716,
+      "grad_norm": 8.035937566563542,
+      "learning_rate": 1e-06,
+      "loss": 0.7963,
+      "step": 448
+    },
+    {
+      "epoch": 0.5745361484325016,
+      "grad_norm": 7.011843678844465,
+      "learning_rate": 1e-06,
+      "loss": 0.7821,
+      "step": 449
+    },
+    {
+      "epoch": 0.5758157389635317,
+      "grad_norm": 7.515815379401376,
+      "learning_rate": 1e-06,
+      "loss": 0.7542,
+      "step": 450
+    },
+    {
+      "epoch": 0.5770953294945618,
+      "grad_norm": 7.411635558623592,
+      "learning_rate": 1e-06,
+      "loss": 0.7713,
+      "step": 451
+    },
+    {
+      "epoch": 0.5783749200255918,
+      "grad_norm": 6.2239690608830545,
+      "learning_rate": 1e-06,
+      "loss": 0.7838,
+      "step": 452
+    },
+    {
+      "epoch": 0.5796545105566219,
+      "grad_norm": 6.643871418051003,
+      "learning_rate": 1e-06,
+      "loss": 0.822,
+      "step": 453
+    },
+    {
+      "epoch": 0.580934101087652,
+      "grad_norm": 5.414613344372611,
+      "learning_rate": 1e-06,
+      "loss": 0.7558,
+      "step": 454
+    },
+    {
+      "epoch": 0.582213691618682,
+      "grad_norm": 7.658600379099193,
+      "learning_rate": 1e-06,
+      "loss": 0.8168,
+      "step": 455
+    },
+    {
+      "epoch": 0.5834932821497121,
+      "grad_norm": 6.89107198395542,
+      "learning_rate": 1e-06,
+      "loss": 0.7282,
+      "step": 456
+    },
+    {
+      "epoch": 0.5847728726807422,
+      "grad_norm": 6.575052747072502,
+      "learning_rate": 1e-06,
+      "loss": 0.7986,
+      "step": 457
+    },
+    {
+      "epoch": 0.5860524632117722,
+      "grad_norm": 7.685555560641489,
+      "learning_rate": 1e-06,
+      "loss": 0.7973,
+      "step": 458
+    },
+    {
+      "epoch": 0.5873320537428023,
+      "grad_norm": 6.541884786860553,
+      "learning_rate": 1e-06,
+      "loss": 0.7628,
+      "step": 459
+    },
+    {
+      "epoch": 0.5886116442738324,
+      "grad_norm": 7.7326930727103065,
+      "learning_rate": 1e-06,
+      "loss": 0.7635,
+      "step": 460
+    },
+    {
+      "epoch": 0.5898912348048625,
+      "grad_norm": 7.839086278284893,
+      "learning_rate": 1e-06,
+      "loss": 0.7461,
+      "step": 461
+    },
+    {
+      "epoch": 0.5911708253358925,
+      "grad_norm": 8.335326636900563,
+      "learning_rate": 1e-06,
+      "loss": 0.7604,
+      "step": 462
+    },
+    {
+      "epoch": 0.5924504158669226,
+      "grad_norm": 8.00369443644871,
+      "learning_rate": 1e-06,
+      "loss": 0.7681,
+      "step": 463
+    },
+    {
+      "epoch": 0.5937300063979527,
+      "grad_norm": 7.1566794948957435,
+      "learning_rate": 1e-06,
+      "loss": 0.7162,
+      "step": 464
+    },
+    {
+      "epoch": 0.5950095969289827,
+      "grad_norm": 6.684407001519589,
+      "learning_rate": 1e-06,
+      "loss": 0.7663,
+      "step": 465
+    },
+    {
+      "epoch": 0.5962891874600128,
+      "grad_norm": 7.743819134098399,
+      "learning_rate": 1e-06,
+      "loss": 0.8146,
+      "step": 466
+    },
+    {
+      "epoch": 0.5975687779910429,
+      "grad_norm": 8.999737043929967,
+      "learning_rate": 1e-06,
+      "loss": 0.8147,
+      "step": 467
+    },
+    {
+      "epoch": 0.5988483685220729,
+      "grad_norm": 4.983920286224524,
+      "learning_rate": 1e-06,
+      "loss": 0.7689,
+      "step": 468
+    },
+    {
+      "epoch": 0.600127959053103,
+      "grad_norm": 5.99207716091729,
+      "learning_rate": 1e-06,
+      "loss": 0.7766,
+      "step": 469
+    },
+    {
+      "epoch": 0.6014075495841331,
+      "grad_norm": 9.281879065763764,
+      "learning_rate": 1e-06,
+      "loss": 0.8054,
+      "step": 470
+    },
+    {
+      "epoch": 0.6026871401151631,
+      "grad_norm": 7.259239601767904,
+      "learning_rate": 1e-06,
+      "loss": 0.7993,
+      "step": 471
+    },
+    {
+      "epoch": 0.6039667306461932,
+      "grad_norm": 6.815669564105977,
+      "learning_rate": 1e-06,
+      "loss": 0.7454,
+      "step": 472
+    },
+    {
+      "epoch": 0.6052463211772233,
+      "grad_norm": 9.24365258890641,
+      "learning_rate": 1e-06,
+      "loss": 0.7573,
+      "step": 473
+    },
+    {
+      "epoch": 0.6065259117082533,
+      "grad_norm": 6.1041583507874355,
+      "learning_rate": 1e-06,
+      "loss": 0.7717,
+      "step": 474
+    },
+    {
+      "epoch": 0.6078055022392834,
+      "grad_norm": 6.290945560802551,
+      "learning_rate": 1e-06,
+      "loss": 0.679,
+      "step": 475
+    },
+    {
+      "epoch": 0.6090850927703135,
+      "grad_norm": 7.494591404508937,
+      "learning_rate": 1e-06,
+      "loss": 0.7875,
+      "step": 476
+    },
+    {
+      "epoch": 0.6103646833013435,
+      "grad_norm": 5.587872866371001,
+      "learning_rate": 1e-06,
+      "loss": 0.7809,
+      "step": 477
+    },
+    {
+      "epoch": 0.6116442738323736,
+      "grad_norm": 7.23056518885902,
+      "learning_rate": 1e-06,
+      "loss": 0.7658,
+      "step": 478
+    },
+    {
+      "epoch": 0.6129238643634037,
+      "grad_norm": 6.4684072998480024,
+      "learning_rate": 1e-06,
+      "loss": 0.7435,
+      "step": 479
+    },
+    {
+      "epoch": 0.6142034548944337,
+      "grad_norm": 6.169615723193868,
+      "learning_rate": 1e-06,
+      "loss": 0.7277,
+      "step": 480
+    },
+    {
+      "epoch": 0.6154830454254638,
+      "grad_norm": 6.844378149525197,
+      "learning_rate": 1e-06,
+      "loss": 0.8073,
+      "step": 481
+    },
+    {
+      "epoch": 0.6167626359564939,
+      "grad_norm": 5.40913494877238,
+      "learning_rate": 1e-06,
+      "loss": 0.7823,
+      "step": 482
+    },
+    {
+      "epoch": 0.6180422264875239,
+      "grad_norm": 7.233899758826381,
+      "learning_rate": 1e-06,
+      "loss": 0.7802,
+      "step": 483
+    },
+    {
+      "epoch": 0.619321817018554,
+      "grad_norm": 6.835521104426315,
+      "learning_rate": 1e-06,
+      "loss": 0.7718,
+      "step": 484
+    },
+    {
+      "epoch": 0.6206014075495841,
+      "grad_norm": 6.421051013660347,
+      "learning_rate": 1e-06,
+      "loss": 0.7789,
+      "step": 485
+    },
+    {
+      "epoch": 0.6218809980806143,
+      "grad_norm": 8.331860765488772,
+      "learning_rate": 1e-06,
+      "loss": 0.8063,
+      "step": 486
+    },
+    {
+      "epoch": 0.6231605886116443,
+      "grad_norm": 7.141836269386319,
+      "learning_rate": 1e-06,
+      "loss": 0.7251,
+      "step": 487
+    },
+    {
+      "epoch": 0.6244401791426744,
+      "grad_norm": 9.12512204635997,
+      "learning_rate": 1e-06,
+      "loss": 0.7541,
+      "step": 488
+    },
+    {
+      "epoch": 0.6257197696737045,
+      "grad_norm": 4.914882633263556,
+      "learning_rate": 1e-06,
+      "loss": 0.8127,
+      "step": 489
+    },
+    {
+      "epoch": 0.6269993602047345,
+      "grad_norm": 5.777006206249883,
+      "learning_rate": 1e-06,
+      "loss": 0.8517,
+      "step": 490
+    },
+    {
+      "epoch": 0.6282789507357646,
+      "grad_norm": 7.018329561292823,
+      "learning_rate": 1e-06,
+      "loss": 0.8039,
+      "step": 491
+    },
+    {
+      "epoch": 0.6295585412667947,
+      "grad_norm": 6.824082666688617,
+      "learning_rate": 1e-06,
+      "loss": 0.7997,
+      "step": 492
+    },
+    {
+      "epoch": 0.6308381317978247,
+      "grad_norm": 5.696288459637765,
+      "learning_rate": 1e-06,
+      "loss": 0.7486,
+      "step": 493
+    },
+    {
+      "epoch": 0.6321177223288548,
+      "grad_norm": 6.995153013902109,
+      "learning_rate": 1e-06,
+      "loss": 0.8004,
+      "step": 494
+    },
+    {
+      "epoch": 0.6333973128598849,
+      "grad_norm": 7.114441868787617,
+      "learning_rate": 1e-06,
+      "loss": 0.7821,
+      "step": 495
+    },
+    {
+      "epoch": 0.6346769033909149,
+      "grad_norm": 6.980592142748543,
+      "learning_rate": 1e-06,
+      "loss": 0.8426,
+      "step": 496
+    },
+    {
+      "epoch": 0.635956493921945,
+      "grad_norm": 7.776594772413459,
+      "learning_rate": 1e-06,
+      "loss": 0.8168,
+      "step": 497
+    },
+    {
+      "epoch": 0.6372360844529751,
+      "grad_norm": 6.242527800683221,
+      "learning_rate": 1e-06,
+      "loss": 0.7701,
+      "step": 498
+    },
+    {
+      "epoch": 0.6385156749840051,
+      "grad_norm": 6.692862269247606,
+      "learning_rate": 1e-06,
+      "loss": 0.8124,
+      "step": 499
+    },
+    {
+      "epoch": 0.6397952655150352,
+      "grad_norm": 6.279783576228503,
+      "learning_rate": 1e-06,
+      "loss": 0.7699,
+      "step": 500
+    },
+    {
+      "epoch": 0.6410748560460653,
+      "grad_norm": 5.437812110390189,
+      "learning_rate": 1e-06,
+      "loss": 0.7424,
+      "step": 501
+    },
+    {
+      "epoch": 0.6423544465770953,
+      "grad_norm": 7.338168720335665,
+      "learning_rate": 1e-06,
+      "loss": 0.7925,
+      "step": 502
+    },
+    {
+      "epoch": 0.6436340371081254,
+      "grad_norm": 6.4091622506428605,
+      "learning_rate": 1e-06,
+      "loss": 0.7937,
+      "step": 503
+    },
+    {
+      "epoch": 0.6449136276391555,
+      "grad_norm": 7.206322252201436,
+      "learning_rate": 1e-06,
+      "loss": 0.789,
+      "step": 504
+    },
+    {
+      "epoch": 0.6461932181701855,
+      "grad_norm": 8.216411886040873,
+      "learning_rate": 1e-06,
+      "loss": 0.8101,
+      "step": 505
+    },
+    {
+      "epoch": 0.6474728087012156,
+      "grad_norm": 8.104553504043496,
+      "learning_rate": 1e-06,
+      "loss": 0.7552,
+      "step": 506
+    },
+    {
+      "epoch": 0.6487523992322457,
+      "grad_norm": 4.933406472733816,
+      "learning_rate": 1e-06,
+      "loss": 0.7571,
+      "step": 507
+    },
+    {
+      "epoch": 0.6500319897632757,
+      "grad_norm": 6.926212374550864,
+      "learning_rate": 1e-06,
+      "loss": 0.7367,
+      "step": 508
+    },
+    {
+      "epoch": 0.6513115802943058,
+      "grad_norm": 5.642592786485672,
+      "learning_rate": 1e-06,
+      "loss": 0.8341,
+      "step": 509
+    },
+    {
+      "epoch": 0.6525911708253359,
+      "grad_norm": 5.30755064895389,
+      "learning_rate": 1e-06,
+      "loss": 0.8007,
+      "step": 510
+    },
+    {
+      "epoch": 0.653870761356366,
+      "grad_norm": 9.752509131430628,
+      "learning_rate": 1e-06,
+      "loss": 0.7819,
+      "step": 511
+    },
+    {
+      "epoch": 0.655150351887396,
+      "grad_norm": 5.807409203195054,
+      "learning_rate": 1e-06,
+      "loss": 0.7602,
+      "step": 512
+    },
+    {
+      "epoch": 0.6564299424184261,
+      "grad_norm": 5.965288209317058,
+      "learning_rate": 1e-06,
+      "loss": 0.8102,
+      "step": 513
+    },
+    {
+      "epoch": 0.6577095329494562,
+      "grad_norm": 7.339219962154503,
+      "learning_rate": 1e-06,
+      "loss": 0.8083,
+      "step": 514
+    },
+    {
+      "epoch": 0.6589891234804862,
+      "grad_norm": 5.990959249847153,
+      "learning_rate": 1e-06,
+      "loss": 0.7196,
+      "step": 515
+    },
+    {
+      "epoch": 0.6602687140115163,
+      "grad_norm": 7.641484442767789,
+      "learning_rate": 1e-06,
+      "loss": 0.7289,
+      "step": 516
+    },
+    {
+      "epoch": 0.6615483045425464,
+      "grad_norm": 8.273354286438902,
+      "learning_rate": 1e-06,
+      "loss": 0.7993,
+      "step": 517
+    },
+    {
+      "epoch": 0.6628278950735764,
+      "grad_norm": 6.044998093617163,
+      "learning_rate": 1e-06,
+      "loss": 0.8108,
+      "step": 518
+    },
+    {
+      "epoch": 0.6641074856046065,
+      "grad_norm": 6.109442673191591,
+      "learning_rate": 1e-06,
+      "loss": 0.8439,
+      "step": 519
+    },
+    {
+      "epoch": 0.6653870761356366,
+      "grad_norm": 6.634248293364319,
+      "learning_rate": 1e-06,
+      "loss": 0.7679,
+      "step": 520
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 6.894491699604436,
+      "learning_rate": 1e-06,
+      "loss": 0.758,
+      "step": 521
+    },
+    {
+      "epoch": 0.6679462571976967,
+      "grad_norm": 6.1917612805423605,
+      "learning_rate": 1e-06,
+      "loss": 0.7449,
+      "step": 522
+    },
+    {
+      "epoch": 0.6692258477287268,
+      "grad_norm": 7.950016328331254,
+      "learning_rate": 1e-06,
+      "loss": 0.7561,
+      "step": 523
+    },
+    {
+      "epoch": 0.6705054382597568,
+      "grad_norm": 6.299790573758471,
+      "learning_rate": 1e-06,
+      "loss": 0.7525,
+      "step": 524
+    },
+    {
+      "epoch": 0.6717850287907869,
+      "grad_norm": 7.472344373069824,
+      "learning_rate": 1e-06,
+      "loss": 0.808,
+      "step": 525
+    },
+    {
+      "epoch": 0.673064619321817,
+      "grad_norm": 6.954829822732137,
+      "learning_rate": 1e-06,
+      "loss": 0.7794,
+      "step": 526
+    },
+    {
+      "epoch": 0.674344209852847,
+      "grad_norm": 6.405471486173751,
+      "learning_rate": 1e-06,
+      "loss": 0.7849,
+      "step": 527
+    },
+    {
+      "epoch": 0.6756238003838771,
+      "grad_norm": 6.490268375504329,
+      "learning_rate": 1e-06,
+      "loss": 0.8371,
+      "step": 528
+    },
+    {
+      "epoch": 0.6769033909149073,
+      "grad_norm": 6.846099340039869,
+      "learning_rate": 1e-06,
+      "loss": 0.7986,
+      "step": 529
+    },
+    {
+      "epoch": 0.6781829814459372,
+      "grad_norm": 7.803535423261094,
+      "learning_rate": 1e-06,
+      "loss": 0.7491,
+      "step": 530
+    },
+    {
+      "epoch": 0.6794625719769674,
+      "grad_norm": 7.817672789950872,
+      "learning_rate": 1e-06,
+      "loss": 0.8144,
+      "step": 531
+    },
+    {
+      "epoch": 0.6807421625079975,
+      "grad_norm": 6.61702651464866,
+      "learning_rate": 1e-06,
+      "loss": 0.7949,
+      "step": 532
+    },
+    {
+      "epoch": 0.6820217530390275,
+      "grad_norm": 8.395010485684727,
+      "learning_rate": 1e-06,
+      "loss": 0.7183,
+      "step": 533
+    },
+    {
+      "epoch": 0.6833013435700576,
+      "grad_norm": 7.702883930879132,
+      "learning_rate": 1e-06,
+      "loss": 0.7736,
+      "step": 534
+    },
+    {
+      "epoch": 0.6845809341010877,
+      "grad_norm": 6.359426129614325,
+      "learning_rate": 1e-06,
+      "loss": 0.7327,
+      "step": 535
+    },
+    {
+      "epoch": 0.6858605246321178,
+      "grad_norm": 5.734932401251934,
+      "learning_rate": 1e-06,
+      "loss": 0.7965,
+      "step": 536
+    },
+    {
+      "epoch": 0.6871401151631478,
+      "grad_norm": 6.940852196995998,
+      "learning_rate": 1e-06,
+      "loss": 0.8068,
+      "step": 537
+    },
+    {
+      "epoch": 0.6884197056941779,
+      "grad_norm": 8.111126214451016,
+      "learning_rate": 1e-06,
+      "loss": 0.8066,
+      "step": 538
+    },
+    {
+      "epoch": 0.689699296225208,
+      "grad_norm": 7.647204253515068,
+      "learning_rate": 1e-06,
+      "loss": 0.7806,
+      "step": 539
+    },
+    {
+      "epoch": 0.690978886756238,
+      "grad_norm": 7.74624658310495,
+      "learning_rate": 1e-06,
+      "loss": 0.7454,
+      "step": 540
+    },
+    {
+      "epoch": 0.6922584772872681,
+      "grad_norm": 7.437936230283719,
+      "learning_rate": 1e-06,
+      "loss": 0.7961,
+      "step": 541
+    },
+    {
+      "epoch": 0.6935380678182982,
+      "grad_norm": 7.741128988131192,
+      "learning_rate": 1e-06,
+      "loss": 0.7989,
+      "step": 542
+    },
+    {
+      "epoch": 0.6948176583493282,
+      "grad_norm": 8.151767237189496,
+      "learning_rate": 1e-06,
+      "loss": 0.7759,
+      "step": 543
+    },
+    {
+      "epoch": 0.6960972488803583,
+      "grad_norm": 7.8917348781631205,
+      "learning_rate": 1e-06,
+      "loss": 0.8038,
+      "step": 544
+    },
+    {
+      "epoch": 0.6973768394113884,
+      "grad_norm": 8.550848815124091,
+      "learning_rate": 1e-06,
+      "loss": 0.7551,
+      "step": 545
+    },
+    {
+      "epoch": 0.6986564299424184,
+      "grad_norm": 8.514957370603693,
+      "learning_rate": 1e-06,
+      "loss": 0.7808,
+      "step": 546
+    },
+    {
+      "epoch": 0.6999360204734485,
+      "grad_norm": 5.067506990226004,
+      "learning_rate": 1e-06,
+      "loss": 0.8113,
+      "step": 547
+    },
+    {
+      "epoch": 0.7012156110044786,
+      "grad_norm": 6.373716244488164,
+      "learning_rate": 1e-06,
+      "loss": 0.7566,
+      "step": 548
+    },
+    {
+      "epoch": 0.7024952015355086,
+      "grad_norm": 5.461386347665844,
+      "learning_rate": 1e-06,
+      "loss": 0.8028,
+      "step": 549
+    },
+    {
+      "epoch": 0.7037747920665387,
+      "grad_norm": 7.35101062381485,
+      "learning_rate": 1e-06,
+      "loss": 0.7456,
+      "step": 550
+    },
+    {
+      "epoch": 0.7050543825975688,
+      "grad_norm": 7.435431055634779,
+      "learning_rate": 1e-06,
+      "loss": 0.8365,
+      "step": 551
+    },
+    {
+      "epoch": 0.7063339731285988,
+      "grad_norm": 7.201726151120405,
+      "learning_rate": 1e-06,
+      "loss": 0.7584,
+      "step": 552
+    },
+    {
+      "epoch": 0.7076135636596289,
+      "grad_norm": 7.291031950882336,
+      "learning_rate": 1e-06,
+      "loss": 0.7659,
+      "step": 553
+    },
+    {
+      "epoch": 0.708893154190659,
+      "grad_norm": 7.669638565604909,
+      "learning_rate": 1e-06,
+      "loss": 0.7697,
+      "step": 554
+    },
+    {
+      "epoch": 0.710172744721689,
+      "grad_norm": 5.146245748854998,
+      "learning_rate": 1e-06,
+      "loss": 0.7864,
+      "step": 555
+    },
+    {
+      "epoch": 0.7114523352527191,
+      "grad_norm": 6.9973342067181745,
+      "learning_rate": 1e-06,
+      "loss": 0.7671,
+      "step": 556
+    },
+    {
+      "epoch": 0.7127319257837492,
+      "grad_norm": 6.444586373232263,
+      "learning_rate": 1e-06,
+      "loss": 0.7946,
+      "step": 557
+    },
+    {
+      "epoch": 0.7140115163147792,
+      "grad_norm": 7.71406465193111,
+      "learning_rate": 1e-06,
+      "loss": 0.7555,
+      "step": 558
+    },
+    {
+      "epoch": 0.7152911068458093,
+      "grad_norm": 8.24302744705605,
+      "learning_rate": 1e-06,
+      "loss": 0.8107,
+      "step": 559
+    },
+    {
+      "epoch": 0.7165706973768394,
+      "grad_norm": 7.0806477002710135,
+      "learning_rate": 1e-06,
+      "loss": 0.8044,
+      "step": 560
+    },
+    {
+      "epoch": 0.7178502879078695,
+      "grad_norm": 8.56844290184104,
+      "learning_rate": 1e-06,
+      "loss": 0.7308,
+      "step": 561
+    },
+    {
+      "epoch": 0.7191298784388995,
+      "grad_norm": 8.623158190822549,
+      "learning_rate": 1e-06,
+      "loss": 0.8282,
+      "step": 562
+    },
+    {
+      "epoch": 0.7204094689699296,
+      "grad_norm": 7.575374713248748,
+      "learning_rate": 1e-06,
+      "loss": 0.7257,
+      "step": 563
+    },
+    {
+      "epoch": 0.7216890595009597,
+      "grad_norm": 6.2637118337743605,
+      "learning_rate": 1e-06,
+      "loss": 0.7759,
+      "step": 564
+    },
+    {
+      "epoch": 0.7229686500319897,
+      "grad_norm": 7.689457714208068,
+      "learning_rate": 1e-06,
+      "loss": 0.7443,
+      "step": 565
+    },
+    {
+      "epoch": 0.7242482405630198,
+      "grad_norm": 6.261701011290504,
+      "learning_rate": 1e-06,
+      "loss": 0.8156,
+      "step": 566
+    },
+    {
+      "epoch": 0.72552783109405,
+      "grad_norm": 9.533167672907101,
+      "learning_rate": 1e-06,
+      "loss": 0.8305,
+      "step": 567
+    },
+    {
+      "epoch": 0.7268074216250799,
+      "grad_norm": 6.974775053514513,
+      "learning_rate": 1e-06,
+      "loss": 0.7804,
+      "step": 568
+    },
+    {
+      "epoch": 0.72808701215611,
+      "grad_norm": 7.679216269934546,
+      "learning_rate": 1e-06,
+      "loss": 0.7951,
+      "step": 569
+    },
+    {
+      "epoch": 0.7293666026871402,
+      "grad_norm": 5.988346585004729,
+      "learning_rate": 1e-06,
+      "loss": 0.8202,
+      "step": 570
+    },
+    {
+      "epoch": 0.7306461932181701,
+      "grad_norm": 10.45114266421651,
+      "learning_rate": 1e-06,
+      "loss": 0.7865,
+      "step": 571
+    },
+    {
+      "epoch": 0.7319257837492003,
+      "grad_norm": 9.192974884811715,
+      "learning_rate": 1e-06,
+      "loss": 0.7984,
+      "step": 572
+    },
+    {
+      "epoch": 0.7332053742802304,
+      "grad_norm": 6.252272547888634,
+      "learning_rate": 1e-06,
+      "loss": 0.7703,
+      "step": 573
+    },
+    {
+      "epoch": 0.7344849648112604,
+      "grad_norm": 8.07765618160131,
+      "learning_rate": 1e-06,
+      "loss": 0.7311,
+      "step": 574
+    },
+    {
+      "epoch": 0.7357645553422905,
+      "grad_norm": 7.758311836915258,
+      "learning_rate": 1e-06,
+      "loss": 0.8148,
+      "step": 575
+    },
+    {
+      "epoch": 0.7370441458733206,
+      "grad_norm": 8.720184011817095,
+      "learning_rate": 1e-06,
+      "loss": 0.8112,
+      "step": 576
+    },
+    {
+      "epoch": 0.7383237364043506,
+      "grad_norm": 6.145093199572889,
+      "learning_rate": 1e-06,
+      "loss": 0.7811,
+      "step": 577
+    },
+    {
+      "epoch": 0.7396033269353807,
+      "grad_norm": 8.320211478041763,
+      "learning_rate": 1e-06,
+      "loss": 0.7958,
+      "step": 578
+    },
+    {
+      "epoch": 0.7408829174664108,
+      "grad_norm": 4.421508449888181,
+      "learning_rate": 1e-06,
+      "loss": 0.7695,
+      "step": 579
+    },
+    {
+      "epoch": 0.7421625079974408,
+      "grad_norm": 6.255077858060669,
+      "learning_rate": 1e-06,
+      "loss": 0.756,
+      "step": 580
+    },
+    {
+      "epoch": 0.7434420985284709,
+      "grad_norm": 9.594519622708969,
+      "learning_rate": 1e-06,
+      "loss": 0.7461,
+      "step": 581
+    },
+    {
+      "epoch": 0.744721689059501,
+      "grad_norm": 7.153868687103066,
+      "learning_rate": 1e-06,
+      "loss": 0.7475,
+      "step": 582
+    },
+    {
+      "epoch": 0.746001279590531,
+      "grad_norm": 7.002179645521348,
+      "learning_rate": 1e-06,
+      "loss": 0.7719,
+      "step": 583
+    },
+    {
+      "epoch": 0.7472808701215611,
+      "grad_norm": 6.954253008962774,
+      "learning_rate": 1e-06,
+      "loss": 0.7703,
+      "step": 584
+    },
+    {
+      "epoch": 0.7485604606525912,
+      "grad_norm": 8.037624967550803,
+      "learning_rate": 1e-06,
+      "loss": 0.8256,
+      "step": 585
+    },
+    {
+      "epoch": 0.7498400511836213,
+      "grad_norm": 8.198704780900506,
+      "learning_rate": 1e-06,
+      "loss": 0.7433,
+      "step": 586
+    },
+    {
+      "epoch": 0.7511196417146513,
+      "grad_norm": 6.809247559192661,
+      "learning_rate": 1e-06,
+      "loss": 0.737,
+      "step": 587
+    },
+    {
+      "epoch": 0.7523992322456814,
+      "grad_norm": 7.602533677714186,
+      "learning_rate": 1e-06,
+      "loss": 0.786,
+      "step": 588
+    },
+    {
+      "epoch": 0.7536788227767115,
+      "grad_norm": 6.36084397455598,
+      "learning_rate": 1e-06,
+      "loss": 0.8235,
+      "step": 589
+    },
+    {
+      "epoch": 0.7549584133077415,
+      "grad_norm": 6.314512160497504,
+      "learning_rate": 1e-06,
+      "loss": 0.7715,
+      "step": 590
+    },
+    {
+      "epoch": 0.7562380038387716,
+      "grad_norm": 5.458748128189958,
+      "learning_rate": 1e-06,
+      "loss": 0.7859,
+      "step": 591
+    },
+    {
+      "epoch": 0.7575175943698017,
+      "grad_norm": 5.149325446021146,
+      "learning_rate": 1e-06,
+      "loss": 0.8173,
+      "step": 592
+    },
+    {
+      "epoch": 0.7587971849008317,
+      "grad_norm": 8.407896755954056,
+      "learning_rate": 1e-06,
+      "loss": 0.7373,
+      "step": 593
+    },
+    {
+      "epoch": 0.7600767754318618,
+      "grad_norm": 7.591831800795671,
+      "learning_rate": 1e-06,
+      "loss": 0.8061,
+      "step": 594
+    },
+    {
+      "epoch": 0.7613563659628919,
+      "grad_norm": 6.180732253152484,
+      "learning_rate": 1e-06,
+      "loss": 0.8345,
+      "step": 595
+    },
+    {
+      "epoch": 0.7626359564939219,
+      "grad_norm": 5.508610723123652,
+      "learning_rate": 1e-06,
+      "loss": 0.7807,
+      "step": 596
+    },
+    {
+      "epoch": 0.763915547024952,
+      "grad_norm": 6.964842368236831,
+      "learning_rate": 1e-06,
+      "loss": 0.7754,
+      "step": 597
+    },
+    {
+      "epoch": 0.7651951375559821,
+      "grad_norm": 6.330247352140575,
+      "learning_rate": 1e-06,
+      "loss": 0.7677,
+      "step": 598
+    },
+    {
+      "epoch": 0.7664747280870121,
+      "grad_norm": 7.372373506486871,
+      "learning_rate": 1e-06,
+      "loss": 0.7938,
+      "step": 599
+    },
+    {
+      "epoch": 0.7677543186180422,
+      "grad_norm": 5.600080467260605,
+      "learning_rate": 1e-06,
+      "loss": 0.7626,
+      "step": 600
+    },
+    {
+      "epoch": 0.7690339091490723,
+      "grad_norm": 7.094856895272238,
+      "learning_rate": 1e-06,
+      "loss": 0.7963,
+      "step": 601
+    },
+    {
+      "epoch": 0.7703134996801023,
+      "grad_norm": 8.013368233486496,
+      "learning_rate": 1e-06,
+      "loss": 0.8243,
+      "step": 602
+    },
+    {
+      "epoch": 0.7715930902111324,
+      "grad_norm": 5.300481671346095,
+      "learning_rate": 1e-06,
+      "loss": 0.7816,
+      "step": 603
+    },
+    {
+      "epoch": 0.7728726807421625,
+      "grad_norm": 6.277243956983349,
+      "learning_rate": 1e-06,
+      "loss": 0.8307,
+      "step": 604
+    },
+    {
+      "epoch": 0.7741522712731925,
+      "grad_norm": 6.920820053012162,
+      "learning_rate": 1e-06,
+      "loss": 0.7665,
+      "step": 605
+    },
+    {
+      "epoch": 0.7754318618042226,
+      "grad_norm": 7.018296959872455,
+      "learning_rate": 1e-06,
+      "loss": 0.7722,
+      "step": 606
+    },
+    {
+      "epoch": 0.7767114523352527,
+      "grad_norm": 6.167823983529555,
+      "learning_rate": 1e-06,
+      "loss": 0.7677,
+      "step": 607
+    },
+    {
+      "epoch": 0.7779910428662828,
+      "grad_norm": 7.939039655302789,
+      "learning_rate": 1e-06,
+      "loss": 0.7705,
+      "step": 608
+    },
+    {
+      "epoch": 0.7792706333973128,
+      "grad_norm": 7.478827734290663,
+      "learning_rate": 1e-06,
+      "loss": 0.7253,
+      "step": 609
+    },
+    {
+      "epoch": 0.780550223928343,
+      "grad_norm": 7.485807552371261,
+      "learning_rate": 1e-06,
+      "loss": 0.7504,
+      "step": 610
+    },
+    {
+      "epoch": 0.781829814459373,
+      "grad_norm": 8.042404289129731,
+      "learning_rate": 1e-06,
+      "loss": 0.7887,
+      "step": 611
+    },
+    {
+      "epoch": 0.783109404990403,
+      "grad_norm": 8.587435883296408,
+      "learning_rate": 1e-06,
+      "loss": 0.7755,
+      "step": 612
+    },
+    {
+      "epoch": 0.7843889955214332,
+      "grad_norm": 6.267218882989902,
+      "learning_rate": 1e-06,
+      "loss": 0.7934,
+      "step": 613
+    },
+    {
+      "epoch": 0.7856685860524633,
+      "grad_norm": 6.620722397356821,
+      "learning_rate": 1e-06,
+      "loss": 0.8228,
+      "step": 614
+    },
+    {
+      "epoch": 0.7869481765834933,
+      "grad_norm": 6.345824089868392,
+      "learning_rate": 1e-06,
+      "loss": 0.7728,
+      "step": 615
+    },
+    {
+      "epoch": 0.7882277671145234,
+      "grad_norm": 6.637518685531026,
+      "learning_rate": 1e-06,
+      "loss": 0.8004,
+      "step": 616
+    },
+    {
+      "epoch": 0.7895073576455535,
+      "grad_norm": 8.36907448250758,
+      "learning_rate": 1e-06,
+      "loss": 0.7645,
+      "step": 617
+    },
+    {
+      "epoch": 0.7907869481765835,
+      "grad_norm": 6.774681018388789,
+      "learning_rate": 1e-06,
+      "loss": 0.7891,
+      "step": 618
+    },
+    {
+      "epoch": 0.7920665387076136,
+      "grad_norm": 6.7144465041173715,
+      "learning_rate": 1e-06,
+      "loss": 0.725,
+      "step": 619
+    },
+    {
+      "epoch": 0.7933461292386437,
+      "grad_norm": 6.557185434902306,
+      "learning_rate": 1e-06,
+      "loss": 0.7689,
+      "step": 620
+    },
+    {
+      "epoch": 0.7946257197696737,
+      "grad_norm": 5.6706296465554695,
+      "learning_rate": 1e-06,
+      "loss": 0.789,
+      "step": 621
+    },
+    {
+      "epoch": 0.7959053103007038,
+      "grad_norm": 7.344972300620442,
+      "learning_rate": 1e-06,
+      "loss": 0.7805,
+      "step": 622
+    },
+    {
+      "epoch": 0.7971849008317339,
+      "grad_norm": 6.610389207304863,
+      "learning_rate": 1e-06,
+      "loss": 0.8067,
+      "step": 623
+    },
+    {
+      "epoch": 0.7984644913627639,
+      "grad_norm": 7.6887779095549575,
+      "learning_rate": 1e-06,
+      "loss": 0.7936,
+      "step": 624
+    },
+    {
+      "epoch": 0.799744081893794,
+      "grad_norm": 9.223951354642638,
+      "learning_rate": 1e-06,
+      "loss": 0.7634,
+      "step": 625
+    },
+    {
+      "epoch": 0.8010236724248241,
+      "grad_norm": 6.257973141887604,
+      "learning_rate": 1e-06,
+      "loss": 0.8116,
+      "step": 626
+    },
+    {
+      "epoch": 0.8023032629558541,
+      "grad_norm": 6.87195734574355,
+      "learning_rate": 1e-06,
+      "loss": 0.7599,
+      "step": 627
+    },
+    {
+      "epoch": 0.8035828534868842,
+      "grad_norm": 6.139568768639382,
+      "learning_rate": 1e-06,
+      "loss": 0.752,
+      "step": 628
+    },
+    {
+      "epoch": 0.8048624440179143,
+      "grad_norm": 5.9110715602025605,
+      "learning_rate": 1e-06,
+      "loss": 0.8368,
+      "step": 629
+    },
+    {
+      "epoch": 0.8061420345489443,
+      "grad_norm": 7.896956639039807,
+      "learning_rate": 1e-06,
+      "loss": 0.8004,
+      "step": 630
+    },
+    {
+      "epoch": 0.8074216250799744,
+      "grad_norm": 6.207534832488455,
+      "learning_rate": 1e-06,
+      "loss": 0.7855,
+      "step": 631
+    },
+    {
+      "epoch": 0.8087012156110045,
+      "grad_norm": 7.76435154868241,
+      "learning_rate": 1e-06,
+      "loss": 0.7589,
+      "step": 632
+    },
+    {
+      "epoch": 0.8099808061420346,
+      "grad_norm": 5.985541870890986,
+      "learning_rate": 1e-06,
+      "loss": 0.7676,
+      "step": 633
+    },
+    {
+      "epoch": 0.8112603966730646,
+      "grad_norm": 8.609625787893325,
+      "learning_rate": 1e-06,
+      "loss": 0.7795,
+      "step": 634
+    },
+    {
+      "epoch": 0.8125399872040947,
+      "grad_norm": 10.195384020324417,
+      "learning_rate": 1e-06,
+      "loss": 0.789,
+      "step": 635
+    },
+    {
+      "epoch": 0.8138195777351248,
+      "grad_norm": 6.151144052959048,
+      "learning_rate": 1e-06,
+      "loss": 0.7892,
+      "step": 636
+    },
+    {
+      "epoch": 0.8150991682661548,
+      "grad_norm": 5.609114847974993,
+      "learning_rate": 1e-06,
+      "loss": 0.7989,
+      "step": 637
+    },
+    {
+      "epoch": 0.8163787587971849,
+      "grad_norm": 6.2257344020456715,
+      "learning_rate": 1e-06,
+      "loss": 0.7957,
+      "step": 638
+    },
+    {
+      "epoch": 0.817658349328215,
+      "grad_norm": 5.440217253083261,
+      "learning_rate": 1e-06,
+      "loss": 0.8016,
+      "step": 639
+    },
+    {
+      "epoch": 0.818937939859245,
+      "grad_norm": 5.09477255685576,
+      "learning_rate": 1e-06,
+      "loss": 0.7655,
+      "step": 640
+    },
+    {
+      "epoch": 0.8202175303902751,
+      "grad_norm": 7.176821422280409,
+      "learning_rate": 1e-06,
+      "loss": 0.746,
+      "step": 641
+    },
+    {
+      "epoch": 0.8214971209213052,
+      "grad_norm": 5.7293169165636755,
+      "learning_rate": 1e-06,
+      "loss": 0.7809,
+      "step": 642
+    },
+    {
+      "epoch": 0.8227767114523352,
+      "grad_norm": 6.234406985391164,
+      "learning_rate": 1e-06,
+      "loss": 0.7927,
+      "step": 643
+    },
+    {
+      "epoch": 0.8240563019833653,
+      "grad_norm": 6.6852010248954326,
+      "learning_rate": 1e-06,
+      "loss": 0.8036,
+      "step": 644
+    },
+    {
+      "epoch": 0.8253358925143954,
+      "grad_norm": 7.864579576007929,
+      "learning_rate": 1e-06,
+      "loss": 0.7922,
+      "step": 645
+    },
+    {
+      "epoch": 0.8266154830454254,
+      "grad_norm": 7.475851850849958,
+      "learning_rate": 1e-06,
+      "loss": 0.7508,
+      "step": 646
+    },
+    {
+      "epoch": 0.8278950735764555,
+      "grad_norm": 7.099527265134232,
+      "learning_rate": 1e-06,
+      "loss": 0.7331,
+      "step": 647
+    },
+    {
+      "epoch": 0.8291746641074856,
+      "grad_norm": 6.311857827198018,
+      "learning_rate": 1e-06,
+      "loss": 0.7363,
+      "step": 648
+    },
+    {
+      "epoch": 0.8304542546385156,
+      "grad_norm": 8.057766559743813,
+      "learning_rate": 1e-06,
+      "loss": 0.787,
+      "step": 649
+    },
+    {
+      "epoch": 0.8317338451695457,
+      "grad_norm": 4.469013942608469,
+      "learning_rate": 1e-06,
+      "loss": 0.7239,
+      "step": 650
+    },
+    {
+      "epoch": 0.8330134357005758,
+      "grad_norm": 8.026385064366371,
+      "learning_rate": 1e-06,
+      "loss": 0.8064,
+      "step": 651
+    },
+    {
+      "epoch": 0.8342930262316058,
+      "grad_norm": 6.188954723846468,
+      "learning_rate": 1e-06,
+      "loss": 0.7873,
+      "step": 652
+    },
+    {
+      "epoch": 0.835572616762636,
+      "grad_norm": 6.406807777672047,
+      "learning_rate": 1e-06,
+      "loss": 0.8139,
+      "step": 653
+    },
+    {
+      "epoch": 0.836852207293666,
+      "grad_norm": 7.508011762009619,
+      "learning_rate": 1e-06,
+      "loss": 0.841,
+      "step": 654
+    },
+    {
+      "epoch": 0.838131797824696,
+      "grad_norm": 9.458040978706375,
+      "learning_rate": 1e-06,
+      "loss": 0.8243,
+      "step": 655
+    },
+    {
+      "epoch": 0.8394113883557262,
+      "grad_norm": 9.610322493474891,
+      "learning_rate": 1e-06,
+      "loss": 0.7695,
+      "step": 656
+    },
+    {
+      "epoch": 0.8406909788867563,
+      "grad_norm": 8.710011682697766,
+      "learning_rate": 1e-06,
+      "loss": 0.7754,
+      "step": 657
+    },
+    {
+      "epoch": 0.8419705694177864,
+      "grad_norm": 4.359533142247382,
+      "learning_rate": 1e-06,
+      "loss": 0.774,
+      "step": 658
+    },
+    {
+      "epoch": 0.8432501599488164,
+      "grad_norm": 6.920726269028371,
+      "learning_rate": 1e-06,
+      "loss": 0.8037,
+      "step": 659
+    },
+    {
+      "epoch": 0.8445297504798465,
+      "grad_norm": 5.104294227319665,
+      "learning_rate": 1e-06,
+      "loss": 0.8145,
+      "step": 660
+    },
+    {
+      "epoch": 0.8458093410108766,
+      "grad_norm": 5.149992898317781,
+      "learning_rate": 1e-06,
+      "loss": 0.6816,
+      "step": 661
+    },
+    {
+      "epoch": 0.8470889315419066,
+      "grad_norm": 8.369826072288385,
+      "learning_rate": 1e-06,
+      "loss": 0.6949,
+      "step": 662
+    },
+    {
+      "epoch": 0.8483685220729367,
+      "grad_norm": 6.133038341978874,
+      "learning_rate": 1e-06,
+      "loss": 0.7539,
+      "step": 663
+    },
+    {
+      "epoch": 0.8496481126039668,
+      "grad_norm": 4.616819495586585,
+      "learning_rate": 1e-06,
+      "loss": 0.8134,
+      "step": 664
+    },
+    {
+      "epoch": 0.8509277031349968,
+      "grad_norm": 7.129204644676411,
+      "learning_rate": 1e-06,
+      "loss": 0.7792,
+      "step": 665
+    },
+    {
+      "epoch": 0.8522072936660269,
+      "grad_norm": 8.240152738405646,
+      "learning_rate": 1e-06,
+      "loss": 0.7727,
+      "step": 666
+    },
+    {
+      "epoch": 0.853486884197057,
+      "grad_norm": 5.070317418758419,
+      "learning_rate": 1e-06,
+      "loss": 0.8289,
+      "step": 667
+    },
+    {
+      "epoch": 0.854766474728087,
+      "grad_norm": 4.717813627059127,
+      "learning_rate": 1e-06,
+      "loss": 0.7023,
+      "step": 668
+    },
+    {
+      "epoch": 0.8560460652591171,
+      "grad_norm": 5.230989985053747,
+      "learning_rate": 1e-06,
+      "loss": 0.7291,
+      "step": 669
+    },
+    {
+      "epoch": 0.8573256557901472,
+      "grad_norm": 5.660207816976507,
+      "learning_rate": 1e-06,
+      "loss": 0.7747,
+      "step": 670
+    },
+    {
+      "epoch": 0.8586052463211772,
+      "grad_norm": 7.528459746676139,
+      "learning_rate": 1e-06,
+      "loss": 0.7583,
+      "step": 671
+    },
+    {
+      "epoch": 0.8598848368522073,
+      "grad_norm": 6.131949501214144,
+      "learning_rate": 1e-06,
+      "loss": 0.8208,
+      "step": 672
+    },
+    {
+      "epoch": 0.8611644273832374,
+      "grad_norm": 9.07583103787078,
+      "learning_rate": 1e-06,
+      "loss": 0.7237,
+      "step": 673
+    },
+    {
+      "epoch": 0.8624440179142674,
+      "grad_norm": 6.466784964739768,
+      "learning_rate": 1e-06,
+      "loss": 0.8388,
+      "step": 674
+    },
+    {
+      "epoch": 0.8637236084452975,
+      "grad_norm": 5.141914072868644,
+      "learning_rate": 1e-06,
+      "loss": 0.7949,
+      "step": 675
+    },
+    {
+      "epoch": 0.8650031989763276,
+      "grad_norm": 6.8003554663274,
+      "learning_rate": 1e-06,
+      "loss": 0.7787,
+      "step": 676
+    },
+    {
+      "epoch": 0.8662827895073576,
+      "grad_norm": 6.702759154000229,
+      "learning_rate": 1e-06,
+      "loss": 0.7742,
+      "step": 677
+    },
+    {
+      "epoch": 0.8675623800383877,
+      "grad_norm": 7.597264287306959,
+      "learning_rate": 1e-06,
+      "loss": 0.7692,
+      "step": 678
+    },
+    {
+      "epoch": 0.8688419705694178,
+      "grad_norm": 8.688011159105901,
+      "learning_rate": 1e-06,
+      "loss": 0.7466,
+      "step": 679
+    },
+    {
+      "epoch": 0.8701215611004478,
+      "grad_norm": 8.477909100612505,
+      "learning_rate": 1e-06,
+      "loss": 0.8137,
+      "step": 680
+    },
+    {
+      "epoch": 0.8714011516314779,
+      "grad_norm": 5.513760107833012,
+      "learning_rate": 1e-06,
+      "loss": 0.7757,
+      "step": 681
+    },
+    {
+      "epoch": 0.872680742162508,
+      "grad_norm": 5.5189095663108025,
+      "learning_rate": 1e-06,
+      "loss": 0.7693,
+      "step": 682
+    },
+    {
+      "epoch": 0.8739603326935381,
+      "grad_norm": 4.905915096883278,
+      "learning_rate": 1e-06,
+      "loss": 0.7476,
+      "step": 683
+    },
+    {
+      "epoch": 0.8752399232245681,
+      "grad_norm": 8.71497798375582,
+      "learning_rate": 1e-06,
+      "loss": 0.7634,
+      "step": 684
+    },
+    {
+      "epoch": 0.8765195137555982,
+      "grad_norm": 7.3210085021081275,
+      "learning_rate": 1e-06,
+      "loss": 0.8214,
+      "step": 685
+    },
+    {
+      "epoch": 0.8777991042866283,
+      "grad_norm": 5.882368612054625,
+      "learning_rate": 1e-06,
+      "loss": 0.7777,
+      "step": 686
+    },
+    {
+      "epoch": 0.8790786948176583,
+      "grad_norm": 5.37585445751749,
+      "learning_rate": 1e-06,
+      "loss": 0.8242,
+      "step": 687
+    },
+    {
+      "epoch": 0.8803582853486884,
+      "grad_norm": 9.349219636951391,
+      "learning_rate": 1e-06,
+      "loss": 0.7296,
+      "step": 688
+    },
+    {
+      "epoch": 0.8816378758797185,
+      "grad_norm": 8.299612256754608,
+      "learning_rate": 1e-06,
+      "loss": 0.8253,
+      "step": 689
+    },
+    {
+      "epoch": 0.8829174664107485,
+      "grad_norm": 8.328601278933903,
+      "learning_rate": 1e-06,
+      "loss": 0.8006,
+      "step": 690
+    },
+    {
+      "epoch": 0.8841970569417786,
+      "grad_norm": 5.9674580491788936,
+      "learning_rate": 1e-06,
+      "loss": 0.7661,
+      "step": 691
+    },
+    {
+      "epoch": 0.8854766474728087,
+      "grad_norm": 7.402672918532203,
+      "learning_rate": 1e-06,
+      "loss": 0.7952,
+      "step": 692
+    },
+    {
+      "epoch": 0.8867562380038387,
+      "grad_norm": 6.193523812314037,
+      "learning_rate": 1e-06,
+      "loss": 0.7901,
+      "step": 693
+    },
+    {
+      "epoch": 0.8880358285348688,
+      "grad_norm": 6.651981463488995,
+      "learning_rate": 1e-06,
+      "loss": 0.7404,
+      "step": 694
+    },
+    {
+      "epoch": 0.889315419065899,
+      "grad_norm": 5.844332419098689,
+      "learning_rate": 1e-06,
+      "loss": 0.8016,
+      "step": 695
+    },
+    {
+      "epoch": 0.8905950095969289,
+      "grad_norm": 7.777492305225009,
+      "learning_rate": 1e-06,
+      "loss": 0.8062,
+      "step": 696
+    },
+    {
+      "epoch": 0.891874600127959,
+      "grad_norm": 6.19574847107412,
+      "learning_rate": 1e-06,
+      "loss": 0.8037,
+      "step": 697
+    },
+    {
+      "epoch": 0.8931541906589892,
+      "grad_norm": 7.016314396252926,
+      "learning_rate": 1e-06,
+      "loss": 0.7767,
+      "step": 698
+    },
+    {
+      "epoch": 0.8944337811900192,
+      "grad_norm": 8.575887880390992,
+      "learning_rate": 1e-06,
+      "loss": 0.7628,
+      "step": 699
+    },
+    {
+      "epoch": 0.8957133717210493,
+      "grad_norm": 4.8200793920453595,
+      "learning_rate": 1e-06,
+      "loss": 0.7459,
+      "step": 700
+    },
+    {
+      "epoch": 0.8969929622520794,
+      "grad_norm": 7.1585112134126865,
+      "learning_rate": 1e-06,
+      "loss": 0.837,
+      "step": 701
+    },
+    {
+      "epoch": 0.8982725527831094,
+      "grad_norm": 7.6122448379006356,
+      "learning_rate": 1e-06,
+      "loss": 0.7916,
+      "step": 702
+    },
+    {
+      "epoch": 0.8995521433141395,
+      "grad_norm": 7.35358081341658,
+      "learning_rate": 1e-06,
+      "loss": 0.8189,
+      "step": 703
+    },
+    {
+      "epoch": 0.9008317338451696,
+      "grad_norm": 6.161719920722362,
+      "learning_rate": 1e-06,
+      "loss": 0.7678,
+      "step": 704
+    },
+    {
+      "epoch": 0.9021113243761996,
+      "grad_norm": 7.745230147345831,
+      "learning_rate": 1e-06,
+      "loss": 0.7416,
+      "step": 705
+    },
+    {
+      "epoch": 0.9033909149072297,
+      "grad_norm": 7.083629196413595,
+      "learning_rate": 1e-06,
+      "loss": 0.7425,
+      "step": 706
+    },
+    {
+      "epoch": 0.9046705054382598,
+      "grad_norm": 6.273820898794386,
+      "learning_rate": 1e-06,
+      "loss": 0.7828,
+      "step": 707
+    },
+    {
+      "epoch": 0.9059500959692899,
+      "grad_norm": 7.662118869924387,
+      "learning_rate": 1e-06,
+      "loss": 0.7566,
+      "step": 708
+    },
+    {
+      "epoch": 0.9072296865003199,
+      "grad_norm": 6.564842926819231,
+      "learning_rate": 1e-06,
+      "loss": 0.8355,
+      "step": 709
+    },
+    {
+      "epoch": 0.90850927703135,
+      "grad_norm": 8.603419042701699,
+      "learning_rate": 1e-06,
+      "loss": 0.8116,
+      "step": 710
+    },
+    {
+      "epoch": 0.9097888675623801,
+      "grad_norm": 9.645169114447333,
+      "learning_rate": 1e-06,
+      "loss": 0.8007,
+      "step": 711
+    },
+    {
+      "epoch": 0.9110684580934101,
+      "grad_norm": 5.773047443525601,
+      "learning_rate": 1e-06,
+      "loss": 0.8312,
+      "step": 712
+    },
+    {
+      "epoch": 0.9123480486244402,
+      "grad_norm": 7.204970658316607,
+      "learning_rate": 1e-06,
+      "loss": 0.7655,
+      "step": 713
+    },
+    {
+      "epoch": 0.9136276391554703,
+      "grad_norm": 6.618998817506019,
+      "learning_rate": 1e-06,
+      "loss": 0.8115,
+      "step": 714
+    },
+    {
+      "epoch": 0.9149072296865003,
+      "grad_norm": 7.070806260034711,
+      "learning_rate": 1e-06,
+      "loss": 0.71,
+      "step": 715
+    },
+    {
+      "epoch": 0.9161868202175304,
+      "grad_norm": 6.644156427616485,
+      "learning_rate": 1e-06,
+      "loss": 0.7893,
+      "step": 716
+    },
+    {
+      "epoch": 0.9174664107485605,
+      "grad_norm": 8.033864216185428,
+      "learning_rate": 1e-06,
+      "loss": 0.853,
+      "step": 717
+    },
+    {
+      "epoch": 0.9187460012795905,
+      "grad_norm": 8.042480933283352,
+      "learning_rate": 1e-06,
+      "loss": 0.7772,
+      "step": 718
+    },
+    {
+      "epoch": 0.9200255918106206,
+      "grad_norm": 5.731099714278954,
+      "learning_rate": 1e-06,
+      "loss": 0.7847,
+      "step": 719
+    },
+    {
+      "epoch": 0.9213051823416507,
+      "grad_norm": 6.203450294460579,
+      "learning_rate": 1e-06,
+      "loss": 0.7934,
+      "step": 720
+    },
+    {
+      "epoch": 0.9225847728726807,
+      "grad_norm": 5.761852850063157,
+      "learning_rate": 1e-06,
+      "loss": 0.8002,
+      "step": 721
+    },
+    {
+      "epoch": 0.9238643634037108,
+      "grad_norm": 4.701804115537021,
+      "learning_rate": 1e-06,
+      "loss": 0.7923,
+      "step": 722
+    },
+    {
+      "epoch": 0.9251439539347409,
+      "grad_norm": 7.776112292618453,
+      "learning_rate": 1e-06,
+      "loss": 0.7618,
+      "step": 723
+    },
+    {
+      "epoch": 0.9264235444657709,
+      "grad_norm": 8.32995273618979,
+      "learning_rate": 1e-06,
+      "loss": 0.7685,
+      "step": 724
+    },
+    {
+      "epoch": 0.927703134996801,
+      "grad_norm": 8.343467479044074,
+      "learning_rate": 1e-06,
+      "loss": 0.7996,
+      "step": 725
+    },
+    {
+      "epoch": 0.9289827255278311,
+      "grad_norm": 7.878490579115052,
+      "learning_rate": 1e-06,
+      "loss": 0.7944,
+      "step": 726
+    },
+    {
+      "epoch": 0.9302623160588611,
+      "grad_norm": 6.05836977378853,
+      "learning_rate": 1e-06,
+      "loss": 0.7808,
+      "step": 727
+    },
+    {
+      "epoch": 0.9315419065898912,
+      "grad_norm": 7.770723169629369,
+      "learning_rate": 1e-06,
+      "loss": 0.7148,
+      "step": 728
+    },
+    {
+      "epoch": 0.9328214971209213,
+      "grad_norm": 8.36250332101991,
+      "learning_rate": 1e-06,
+      "loss": 0.7411,
+      "step": 729
+    },
+    {
+      "epoch": 0.9341010876519513,
+      "grad_norm": 10.421614477285154,
+      "learning_rate": 1e-06,
+      "loss": 0.7746,
+      "step": 730
+    },
+    {
+      "epoch": 0.9353806781829814,
+      "grad_norm": 8.094521997025975,
+      "learning_rate": 1e-06,
+      "loss": 0.7902,
+      "step": 731
+    },
+    {
+      "epoch": 0.9366602687140115,
+      "grad_norm": 6.55295452493918,
+      "learning_rate": 1e-06,
+      "loss": 0.7595,
+      "step": 732
+    },
+    {
+      "epoch": 0.9379398592450416,
+      "grad_norm": 4.12551668542858,
+      "learning_rate": 1e-06,
+      "loss": 0.6826,
+      "step": 733
+    },
+    {
+      "epoch": 0.9392194497760716,
+      "grad_norm": 5.189420059189927,
+      "learning_rate": 1e-06,
+      "loss": 0.7855,
+      "step": 734
+    },
+    {
+      "epoch": 0.9404990403071017,
+      "grad_norm": 6.4854035641085055,
+      "learning_rate": 1e-06,
+      "loss": 0.7941,
+      "step": 735
+    },
+    {
+      "epoch": 0.9417786308381318,
+      "grad_norm": 6.899683276976177,
+      "learning_rate": 1e-06,
+      "loss": 0.7746,
+      "step": 736
+    },
+    {
+      "epoch": 0.9430582213691618,
+      "grad_norm": 5.768665608099385,
+      "learning_rate": 1e-06,
+      "loss": 0.7555,
+      "step": 737
+    },
+    {
+      "epoch": 0.944337811900192,
+      "grad_norm": 6.724158168914129,
+      "learning_rate": 1e-06,
+      "loss": 0.7661,
+      "step": 738
+    },
+    {
+      "epoch": 0.945617402431222,
+      "grad_norm": 6.0751386477450415,
+      "learning_rate": 1e-06,
+      "loss": 0.7477,
+      "step": 739
+    },
+    {
+      "epoch": 0.946896992962252,
+      "grad_norm": 7.20902699369686,
+      "learning_rate": 1e-06,
+      "loss": 0.8243,
+      "step": 740
+    },
+    {
+      "epoch": 0.9481765834932822,
+      "grad_norm": 8.756779085963897,
+      "learning_rate": 1e-06,
+      "loss": 0.7879,
+      "step": 741
+    },
+    {
+      "epoch": 0.9494561740243123,
+      "grad_norm": 8.092048594807453,
+      "learning_rate": 1e-06,
+      "loss": 0.8067,
+      "step": 742
+    },
+    {
+      "epoch": 0.9507357645553423,
+      "grad_norm": 9.402027315496179,
+      "learning_rate": 1e-06,
+      "loss": 0.8906,
+      "step": 743
+    },
+    {
+      "epoch": 0.9520153550863724,
+      "grad_norm": 6.782151929864907,
+      "learning_rate": 1e-06,
+      "loss": 0.8151,
+      "step": 744
+    },
+    {
+      "epoch": 0.9532949456174025,
+      "grad_norm": 7.564321687759882,
+      "learning_rate": 1e-06,
+      "loss": 0.8328,
+      "step": 745
+    },
+    {
+      "epoch": 0.9545745361484325,
+      "grad_norm": 5.866660339034752,
+      "learning_rate": 1e-06,
+      "loss": 0.7983,
+      "step": 746
+    },
+    {
+      "epoch": 0.9558541266794626,
+      "grad_norm": 7.908405877808982,
+      "learning_rate": 1e-06,
+      "loss": 0.773,
+      "step": 747
+    },
+    {
+      "epoch": 0.9571337172104927,
+      "grad_norm": 5.0111898998629965,
+      "learning_rate": 1e-06,
+      "loss": 0.8028,
+      "step": 748
+    },
+    {
+      "epoch": 0.9584133077415227,
+      "grad_norm": 6.925983079845888,
+      "learning_rate": 1e-06,
+      "loss": 0.7447,
+      "step": 749
+    },
+    {
+      "epoch": 0.9596928982725528,
+      "grad_norm": 7.303158056213572,
+      "learning_rate": 1e-06,
+      "loss": 0.7683,
+      "step": 750
+    },
+    {
+      "epoch": 0.9609724888035829,
+      "grad_norm": 7.354824249679899,
+      "learning_rate": 1e-06,
+      "loss": 0.7794,
+      "step": 751
+    },
+    {
+      "epoch": 0.9622520793346129,
+      "grad_norm": 6.8034995901605795,
+      "learning_rate": 1e-06,
+      "loss": 0.7947,
+      "step": 752
+    },
+    {
+      "epoch": 0.963531669865643,
+      "grad_norm": 6.8034995901605795,
+      "learning_rate": 1e-06,
+      "loss": 0.7362,
+      "step": 753
+    },
+    {
+      "epoch": 0.9648112603966731,
+      "grad_norm": 8.6316040017734,
+      "learning_rate": 1e-06,
+      "loss": 0.8092,
+      "step": 754
+    },
+    {
+      "epoch": 0.9660908509277031,
+      "grad_norm": 6.667911248274814,
+      "learning_rate": 1e-06,
+      "loss": 0.8082,
+      "step": 755
+    },
+    {
+      "epoch": 0.9673704414587332,
+      "grad_norm": 7.706334080098833,
+      "learning_rate": 1e-06,
+      "loss": 0.8045,
+      "step": 756
+    },
+    {
+      "epoch": 0.9686500319897633,
+      "grad_norm": 6.47671662094834,
+      "learning_rate": 1e-06,
+      "loss": 0.8266,
+      "step": 757
+    },
+    {
+      "epoch": 0.9699296225207934,
+      "grad_norm": 6.9862152556031365,
+      "learning_rate": 1e-06,
+      "loss": 0.7693,
+      "step": 758
+    },
+    {
+      "epoch": 0.9712092130518234,
+      "grad_norm": 6.490539306984062,
+      "learning_rate": 1e-06,
+      "loss": 0.7502,
+      "step": 759
+    },
+    {
+      "epoch": 0.9724888035828535,
+      "grad_norm": 7.136521515643652,
+      "learning_rate": 1e-06,
+      "loss": 0.7554,
+      "step": 760
+    },
+    {
+      "epoch": 0.9737683941138836,
+      "grad_norm": 4.732697262512005,
+      "learning_rate": 1e-06,
+      "loss": 0.7819,
+      "step": 761
+    },
+    {
+      "epoch": 0.9750479846449136,
+      "grad_norm": 3.921344679597292,
+      "learning_rate": 1e-06,
+      "loss": 0.7673,
+      "step": 762
+    },
+    {
+      "epoch": 0.9763275751759437,
+      "grad_norm": 6.377618636840941,
+      "learning_rate": 1e-06,
+      "loss": 0.8128,
+      "step": 763
+    },
+    {
+      "epoch": 0.9776071657069738,
+      "grad_norm": 7.823276219572525,
+      "learning_rate": 1e-06,
+      "loss": 0.7625,
+      "step": 764
+    },
+    {
+      "epoch": 0.9788867562380038,
+      "grad_norm": 8.099224176863821,
+      "learning_rate": 1e-06,
+      "loss": 0.8149,
+      "step": 765
+    },
+    {
+      "epoch": 0.9801663467690339,
+      "grad_norm": 5.2924368059542095,
+      "learning_rate": 1e-06,
+      "loss": 0.8007,
+      "step": 766
+    },
+    {
+      "epoch": 0.981445937300064,
+      "grad_norm": 5.9343929460627605,
+      "learning_rate": 1e-06,
+      "loss": 0.7791,
+      "step": 767
+    },
+    {
+      "epoch": 0.982725527831094,
+      "grad_norm": 6.627801773169884,
+      "learning_rate": 1e-06,
+      "loss": 0.7858,
+      "step": 768
+    },
+    {
+      "epoch": 0.9840051183621241,
+      "grad_norm": 6.355689315999818,
+      "learning_rate": 1e-06,
+      "loss": 0.8397,
+      "step": 769
+    },
+    {
+      "epoch": 0.9852847088931542,
+      "grad_norm": 7.52253045458746,
+      "learning_rate": 1e-06,
+      "loss": 0.7802,
+      "step": 770
+    },
+    {
+      "epoch": 0.9865642994241842,
+      "grad_norm": 7.5731237833387155,
+      "learning_rate": 1e-06,
+      "loss": 0.7558,
+      "step": 771
+    },
+    {
+      "epoch": 0.9878438899552143,
+      "grad_norm": 8.675153614165298,
+      "learning_rate": 1e-06,
+      "loss": 0.807,
+      "step": 772
+    },
+    {
+      "epoch": 0.9891234804862444,
+      "grad_norm": 8.007408074002637,
+      "learning_rate": 1e-06,
+      "loss": 0.7532,
+      "step": 773
+    },
+    {
+      "epoch": 0.9904030710172744,
+      "grad_norm": 6.499843051562209,
+      "learning_rate": 1e-06,
+      "loss": 0.7916,
+      "step": 774
+    },
+    {
+      "epoch": 0.9916826615483045,
+      "grad_norm": 5.863708725183646,
+      "learning_rate": 1e-06,
+      "loss": 0.7684,
+      "step": 775
+    },
+    {
+      "epoch": 0.9929622520793346,
+      "grad_norm": 8.170105687900008,
+      "learning_rate": 1e-06,
+      "loss": 0.7933,
+      "step": 776
+    },
+    {
+      "epoch": 0.9942418426103646,
+      "grad_norm": 5.826961580016016,
+      "learning_rate": 1e-06,
+      "loss": 0.7485,
+      "step": 777
+    },
+    {
+      "epoch": 0.9955214331413947,
+      "grad_norm": 5.24706486414675,
+      "learning_rate": 1e-06,
+      "loss": 0.7765,
+      "step": 778
+    },
+    {
+      "epoch": 0.9968010236724248,
+      "grad_norm": 6.490480897093001,
+      "learning_rate": 1e-06,
+      "loss": 0.7737,
+      "step": 779
+    },
+    {
+      "epoch": 0.9980806142034548,
+      "grad_norm": 8.609506273280285,
+      "learning_rate": 1e-06,
+      "loss": 0.7794,
+      "step": 780
+    },
+    {
+      "epoch": 0.999360204734485,
+      "grad_norm": 4.623090523159483,
+      "learning_rate": 1e-06,
+      "loss": 0.7326,
+      "step": 781
+    },
+    {
+      "epoch": 0.999360204734485,
+      "eval_loss": 1.970609188079834,
+      "eval_runtime": 35.6019,
+      "eval_samples_per_second": 28.088,
+      "eval_steps_per_second": 3.511,
+      "step": 781
+    },
+    {
+      "epoch": 0.0027210884353741495,
+      "grad_norm": 8.95383509429665,
+      "learning_rate": 1e-06,
+      "loss": 0.3981,
+      "step": 782
+    },
+    {
+      "epoch": 0.005442176870748299,
+      "grad_norm": 8.502197202523664,
+      "learning_rate": 1e-06,
+      "loss": 0.4096,
+      "step": 783
+    },
+    {
+      "epoch": 0.00816326530612245,
+      "grad_norm": 8.934583763754567,
+      "learning_rate": 1e-06,
+      "loss": 0.3934,
+      "step": 784
+    },
+    {
+      "epoch": 0.010884353741496598,
+      "grad_norm": 6.603664736534862,
+      "learning_rate": 1e-06,
+      "loss": 0.3656,
+      "step": 785
+    },
+    {
+      "epoch": 0.013605442176870748,
+      "grad_norm": 7.088253382691928,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 786
+    },
+    {
+      "epoch": 0.0163265306122449,
+      "grad_norm": 9.833117826401432,
+      "learning_rate": 1e-06,
+      "loss": 0.3456,
+      "step": 787
+    },
+    {
+      "epoch": 0.01904761904761905,
+      "grad_norm": 6.1608479838315455,
+      "learning_rate": 1e-06,
+      "loss": 0.3379,
+      "step": 788
+    },
+    {
+      "epoch": 0.021768707482993196,
+      "grad_norm": 10.272116262416533,
+      "learning_rate": 1e-06,
+      "loss": 0.3303,
+      "step": 789
+    },
+    {
+      "epoch": 0.024489795918367346,
+      "grad_norm": 8.546021224679029,
+      "learning_rate": 1e-06,
+      "loss": 0.358,
+      "step": 790
+    },
+    {
+      "epoch": 0.027210884353741496,
+      "grad_norm": 8.763263657271041,
+      "learning_rate": 1e-06,
+      "loss": 0.3623,
+      "step": 791
+    },
+    {
+      "epoch": 0.029931972789115645,
+      "grad_norm": 10.705854841346273,
+      "learning_rate": 1e-06,
+      "loss": 0.3198,
+      "step": 792
+    },
+    {
+      "epoch": 0.0326530612244898,
+      "grad_norm": 6.459697282042882,
+      "learning_rate": 1e-06,
+      "loss": 0.3417,
+      "step": 793
+    },
+    {
+      "epoch": 0.03537414965986395,
+      "grad_norm": 10.359445926609911,
+      "learning_rate": 1e-06,
+      "loss": 0.3195,
+      "step": 794
+    },
+    {
+      "epoch": 0.0380952380952381,
+      "grad_norm": 7.602353085434091,
+      "learning_rate": 1e-06,
+      "loss": 0.3428,
+      "step": 795
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 10.269721245608784,
+      "learning_rate": 1e-06,
+      "loss": 0.332,
+      "step": 796
+    },
+    {
+      "epoch": 0.04353741496598639,
+      "grad_norm": 9.811167482053428,
+      "learning_rate": 1e-06,
+      "loss": 0.3001,
+      "step": 797
+    },
+    {
+      "epoch": 0.04625850340136054,
+      "grad_norm": 8.942471207163809,
+      "learning_rate": 1e-06,
+      "loss": 0.3338,
+      "step": 798
+    },
+    {
+      "epoch": 0.04897959183673469,
+      "grad_norm": 12.999776125610065,
+      "learning_rate": 1e-06,
+      "loss": 0.3172,
+      "step": 799
+    },
+    {
+      "epoch": 0.05170068027210884,
+      "grad_norm": 10.716147080680555,
+      "learning_rate": 1e-06,
+      "loss": 0.3448,
+      "step": 800
+    },
+    {
+      "epoch": 0.05442176870748299,
+      "grad_norm": 8.975671509374875,
+      "learning_rate": 1e-06,
+      "loss": 0.3301,
+      "step": 801
+    },
+    {
+      "epoch": 0.05714285714285714,
+      "grad_norm": 9.506793900917545,
+      "learning_rate": 1e-06,
+      "loss": 0.3377,
+      "step": 802
+    },
+    {
+      "epoch": 0.05986394557823129,
+      "grad_norm": 9.756545179908757,
+      "learning_rate": 1e-06,
+      "loss": 0.3406,
+      "step": 803
+    },
+    {
+      "epoch": 0.06258503401360545,
+      "grad_norm": 10.599887350493493,
+      "learning_rate": 1e-06,
+      "loss": 0.3324,
+      "step": 804
+    },
+    {
+      "epoch": 0.0653061224489796,
+      "grad_norm": 8.748813359814463,
+      "learning_rate": 1e-06,
+      "loss": 0.3191,
+      "step": 805
+    },
+    {
+      "epoch": 0.06802721088435375,
+      "grad_norm": 7.395951097430482,
+      "learning_rate": 1e-06,
+      "loss": 0.3422,
+      "step": 806
+    },
+    {
+      "epoch": 0.0707482993197279,
+      "grad_norm": 12.678342199533287,
+      "learning_rate": 1e-06,
+      "loss": 0.3227,
+      "step": 807
+    },
+    {
+      "epoch": 0.07346938775510205,
+      "grad_norm": 9.107579439582814,
+      "learning_rate": 1e-06,
+      "loss": 0.353,
+      "step": 808
+    },
+    {
+      "epoch": 0.0761904761904762,
+      "grad_norm": 8.611336479582576,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 809
+    },
+    {
+      "epoch": 0.07891156462585033,
+      "grad_norm": 7.160687070542016,
+      "learning_rate": 1e-06,
+      "loss": 0.3582,
+      "step": 810
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 7.500085491029069,
+      "learning_rate": 1e-06,
+      "loss": 0.3334,
+      "step": 811
+    },
+    {
+      "epoch": 0.08435374149659863,
+      "grad_norm": 10.66467497199079,
+      "learning_rate": 1e-06,
+      "loss": 0.3455,
+      "step": 812
+    },
+    {
+      "epoch": 0.08707482993197278,
+      "grad_norm": 7.517714145285407,
+      "learning_rate": 1e-06,
+      "loss": 0.3163,
+      "step": 813
+    },
+    {
+      "epoch": 0.08979591836734693,
+      "grad_norm": 8.296616979940483,
+      "learning_rate": 1e-06,
+      "loss": 0.3228,
+      "step": 814
+    },
+    {
+      "epoch": 0.09251700680272108,
+      "grad_norm": 6.096164008733242,
+      "learning_rate": 1e-06,
+      "loss": 0.329,
+      "step": 815
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 8.896837069223206,
+      "learning_rate": 1e-06,
+      "loss": 0.3107,
+      "step": 816
+    },
+    {
+      "epoch": 0.09795918367346938,
+      "grad_norm": 7.451142684431256,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 817
+    },
+    {
+      "epoch": 0.10068027210884353,
+      "grad_norm": 9.05185785990919,
+      "learning_rate": 1e-06,
+      "loss": 0.344,
+      "step": 818
+    },
+    {
+      "epoch": 0.10340136054421768,
+      "grad_norm": 8.702620537714257,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 819
+    },
+    {
+      "epoch": 0.10612244897959183,
+      "grad_norm": 7.780907955432281,
+      "learning_rate": 1e-06,
+      "loss": 0.3265,
+      "step": 820
+    },
+    {
+      "epoch": 0.10884353741496598,
+      "grad_norm": 6.60527931970439,
+      "learning_rate": 1e-06,
+      "loss": 0.327,
+      "step": 821
+    },
+    {
+      "epoch": 0.11156462585034013,
+      "grad_norm": 6.547438037236302,
+      "learning_rate": 1e-06,
+      "loss": 0.3285,
+      "step": 822
+    },
+    {
+      "epoch": 0.11428571428571428,
+      "grad_norm": 9.615594353449664,
+      "learning_rate": 1e-06,
+      "loss": 0.3264,
+      "step": 823
+    },
+    {
+      "epoch": 0.11700680272108843,
+      "grad_norm": 7.524320570830924,
+      "learning_rate": 1e-06,
+      "loss": 0.3168,
+      "step": 824
+    },
+    {
+      "epoch": 0.11972789115646258,
+      "grad_norm": 9.96514512730406,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 825
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 7.247418575798255,
+      "learning_rate": 1e-06,
+      "loss": 0.3078,
+      "step": 826
+    },
+    {
+      "epoch": 0.1251700680272109,
+      "grad_norm": 6.819746902787109,
+      "learning_rate": 1e-06,
+      "loss": 0.3106,
+      "step": 827
+    },
+    {
+      "epoch": 0.12789115646258503,
+      "grad_norm": 7.77124118379486,
+      "learning_rate": 1e-06,
+      "loss": 0.3025,
+      "step": 828
+    },
+    {
+      "epoch": 0.1306122448979592,
+      "grad_norm": 8.339803053803506,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 829
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 9.208228251682115,
+      "learning_rate": 1e-06,
+      "loss": 0.3259,
+      "step": 830
+    },
+    {
+      "epoch": 0.1360544217687075,
+      "grad_norm": 7.781076093800601,
+      "learning_rate": 1e-06,
+      "loss": 0.3279,
+      "step": 831
+    },
+    {
+      "epoch": 0.13877551020408163,
+      "grad_norm": 8.705581477529977,
+      "learning_rate": 1e-06,
+      "loss": 0.3159,
+      "step": 832
+    },
+    {
+      "epoch": 0.1414965986394558,
+      "grad_norm": 7.461943436416155,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 833
+    },
+    {
+      "epoch": 0.14421768707482993,
+      "grad_norm": 6.544302239235068,
+      "learning_rate": 1e-06,
+      "loss": 0.2908,
+      "step": 834
+    },
+    {
+      "epoch": 0.1469387755102041,
+      "grad_norm": 8.81102111352323,
+      "learning_rate": 1e-06,
+      "loss": 0.3341,
+      "step": 835
+    },
+    {
+      "epoch": 0.14965986394557823,
+      "grad_norm": 7.220706069407586,
+      "learning_rate": 1e-06,
+      "loss": 0.2967,
+      "step": 836
+    },
+    {
+      "epoch": 0.1523809523809524,
+      "grad_norm": 6.218388835122915,
+      "learning_rate": 1e-06,
+      "loss": 0.3371,
+      "step": 837
+    },
+    {
+      "epoch": 0.15510204081632653,
+      "grad_norm": 10.746454648568886,
+      "learning_rate": 1e-06,
+      "loss": 0.3517,
+      "step": 838
+    },
+    {
+      "epoch": 0.15782312925170067,
+      "grad_norm": 9.926366229081003,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 839
+    },
+    {
+      "epoch": 0.16054421768707483,
+      "grad_norm": 7.078064531454577,
+      "learning_rate": 1e-06,
+      "loss": 0.335,
+      "step": 840
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 10.090171857839058,
+      "learning_rate": 1e-06,
+      "loss": 0.3077,
+      "step": 841
+    },
+    {
+      "epoch": 0.16598639455782313,
+      "grad_norm": 10.329956663870153,
+      "learning_rate": 1e-06,
+      "loss": 0.321,
+      "step": 842
+    },
+    {
+      "epoch": 0.16870748299319727,
+      "grad_norm": 8.608332406423386,
+      "learning_rate": 1e-06,
+      "loss": 0.3323,
+      "step": 843
+    },
+    {
+      "epoch": 0.17142857142857143,
+      "grad_norm": 10.175955467119337,
+      "learning_rate": 1e-06,
+      "loss": 0.3327,
+      "step": 844
+    },
+    {
+      "epoch": 0.17414965986394557,
+      "grad_norm": 9.628834610741682,
+      "learning_rate": 1e-06,
+      "loss": 0.3308,
+      "step": 845
+    },
+    {
+      "epoch": 0.17687074829931973,
+      "grad_norm": 7.598132037269626,
+      "learning_rate": 1e-06,
+      "loss": 0.3126,
+      "step": 846
+    },
+    {
+      "epoch": 0.17959183673469387,
+      "grad_norm": 10.225233614391897,
+      "learning_rate": 1e-06,
+      "loss": 0.3168,
+      "step": 847
+    },
+    {
+      "epoch": 0.18231292517006803,
+      "grad_norm": 7.854325894813079,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 848
+    },
+    {
+      "epoch": 0.18503401360544217,
+      "grad_norm": 9.08458359270501,
+      "learning_rate": 1e-06,
+      "loss": 0.33,
+      "step": 849
+    },
+    {
+      "epoch": 0.18775510204081633,
+      "grad_norm": 8.47956817604547,
+      "learning_rate": 1e-06,
+      "loss": 0.3101,
+      "step": 850
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 8.895460417513716,
+      "learning_rate": 1e-06,
+      "loss": 0.3171,
+      "step": 851
+    },
+    {
+      "epoch": 0.19319727891156463,
+      "grad_norm": 7.812016619331411,
+      "learning_rate": 1e-06,
+      "loss": 0.3191,
+      "step": 852
+    },
+    {
+      "epoch": 0.19591836734693877,
+      "grad_norm": 9.09308169473296,
+      "learning_rate": 1e-06,
+      "loss": 0.3148,
+      "step": 853
+    },
+    {
+      "epoch": 0.19863945578231293,
+      "grad_norm": 8.488943996603188,
+      "learning_rate": 1e-06,
+      "loss": 0.3086,
+      "step": 854
+    },
+    {
+      "epoch": 0.20136054421768707,
+      "grad_norm": 8.405098060661587,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 855
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 11.769745153715757,
+      "learning_rate": 1e-06,
+      "loss": 0.3583,
+      "step": 856
+    },
+    {
+      "epoch": 0.20680272108843537,
+      "grad_norm": 9.709805028889127,
+      "learning_rate": 1e-06,
+      "loss": 0.3538,
+      "step": 857
+    },
+    {
+      "epoch": 0.20952380952380953,
+      "grad_norm": 8.503983657853516,
+      "learning_rate": 1e-06,
+      "loss": 0.3372,
+      "step": 858
+    },
+    {
+      "epoch": 0.21224489795918366,
+      "grad_norm": 10.412646748347678,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 859
+    },
+    {
+      "epoch": 0.21496598639455783,
+      "grad_norm": 6.734509804901044,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 860
+    },
+    {
+      "epoch": 0.21768707482993196,
+      "grad_norm": 9.174388686930875,
+      "learning_rate": 1e-06,
+      "loss": 0.3226,
+      "step": 861
+    },
+    {
+      "epoch": 0.22040816326530613,
+      "grad_norm": 5.810569583243387,
+      "learning_rate": 1e-06,
+      "loss": 0.3083,
+      "step": 862
+    },
+    {
+      "epoch": 0.22312925170068026,
+      "grad_norm": 7.979797269322362,
+      "learning_rate": 1e-06,
+      "loss": 0.3352,
+      "step": 863
+    },
+    {
+      "epoch": 0.22585034013605443,
+      "grad_norm": 7.833062319000555,
+      "learning_rate": 1e-06,
+      "loss": 0.3176,
+      "step": 864
+    },
+    {
+      "epoch": 0.22857142857142856,
+      "grad_norm": 7.449367209141244,
+      "learning_rate": 1e-06,
+      "loss": 0.3433,
+      "step": 865
+    },
+    {
+      "epoch": 0.23129251700680273,
+      "grad_norm": 7.384671998980106,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 866
+    },
+    {
+      "epoch": 0.23401360544217686,
+      "grad_norm": 8.057586616536833,
+      "learning_rate": 1e-06,
+      "loss": 0.3532,
+      "step": 867
+    },
+    {
+      "epoch": 0.23673469387755103,
+      "grad_norm": 11.045585935090763,
+      "learning_rate": 1e-06,
+      "loss": 0.3443,
+      "step": 868
+    },
+    {
+      "epoch": 0.23945578231292516,
+      "grad_norm": 8.830292323934293,
+      "learning_rate": 1e-06,
+      "loss": 0.3346,
+      "step": 869
+    },
+    {
+      "epoch": 0.24217687074829933,
+      "grad_norm": 9.236853379847435,
+      "learning_rate": 1e-06,
+      "loss": 0.3264,
+      "step": 870
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 9.647810072131708,
+      "learning_rate": 1e-06,
+      "loss": 0.329,
+      "step": 871
+    },
+    {
+      "epoch": 0.24761904761904763,
+      "grad_norm": 9.239616303600005,
+      "learning_rate": 1e-06,
+      "loss": 0.3461,
+      "step": 872
+    },
+    {
+      "epoch": 0.2503401360544218,
+      "grad_norm": 9.499596194409069,
+      "learning_rate": 1e-06,
+      "loss": 0.3019,
+      "step": 873
+    },
+    {
+      "epoch": 0.2530612244897959,
+      "grad_norm": 8.879218229049203,
+      "learning_rate": 1e-06,
+      "loss": 0.3479,
+      "step": 874
+    },
+    {
+      "epoch": 0.25578231292517006,
+      "grad_norm": 9.723903588602319,
+      "learning_rate": 1e-06,
+      "loss": 0.3,
+      "step": 875
+    },
+    {
+      "epoch": 0.2585034013605442,
+      "grad_norm": 9.796253868968215,
+      "learning_rate": 1e-06,
+      "loss": 0.32,
+      "step": 876
+    },
+    {
+      "epoch": 0.2612244897959184,
+      "grad_norm": 9.803503807419805,
+      "learning_rate": 1e-06,
+      "loss": 0.2947,
+      "step": 877
+    },
+    {
+      "epoch": 0.2639455782312925,
+      "grad_norm": 6.546586424722833,
+      "learning_rate": 1e-06,
+      "loss": 0.3131,
+      "step": 878
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 8.270870420559163,
+      "learning_rate": 1e-06,
+      "loss": 0.3293,
+      "step": 879
+    },
+    {
+      "epoch": 0.2693877551020408,
+      "grad_norm": 8.51031470937762,
+      "learning_rate": 1e-06,
+      "loss": 0.319,
+      "step": 880
+    },
+    {
+      "epoch": 0.272108843537415,
+      "grad_norm": 9.623582510739977,
+      "learning_rate": 1e-06,
+      "loss": 0.3303,
+      "step": 881
+    },
+    {
+      "epoch": 0.2748299319727891,
+      "grad_norm": 9.948988339668645,
+      "learning_rate": 1e-06,
+      "loss": 0.3386,
+      "step": 882
+    },
+    {
+      "epoch": 0.27755102040816326,
+      "grad_norm": 7.2258123464856885,
+      "learning_rate": 1e-06,
+      "loss": 0.2992,
+      "step": 883
+    },
+    {
+      "epoch": 0.2802721088435374,
+      "grad_norm": 6.850561997190994,
+      "learning_rate": 1e-06,
+      "loss": 0.3366,
+      "step": 884
+    },
+    {
+      "epoch": 0.2829931972789116,
+      "grad_norm": 8.369682301882367,
+      "learning_rate": 1e-06,
+      "loss": 0.3406,
+      "step": 885
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 9.990552789789048,
+      "learning_rate": 1e-06,
+      "loss": 0.3274,
+      "step": 886
+    },
+    {
+      "epoch": 0.28843537414965986,
+      "grad_norm": 8.685271598668963,
+      "learning_rate": 1e-06,
+      "loss": 0.3222,
+      "step": 887
+    },
+    {
+      "epoch": 0.291156462585034,
+      "grad_norm": 8.886061618306329,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 888
+    },
+    {
+      "epoch": 0.2938775510204082,
+      "grad_norm": 7.841117824015353,
+      "learning_rate": 1e-06,
+      "loss": 0.334,
+      "step": 889
+    },
+    {
+      "epoch": 0.2965986394557823,
+      "grad_norm": 5.724375087591104,
+      "learning_rate": 1e-06,
+      "loss": 0.3302,
+      "step": 890
+    },
+    {
+      "epoch": 0.29931972789115646,
+      "grad_norm": 7.03477311418178,
+      "learning_rate": 1e-06,
+      "loss": 0.3007,
+      "step": 891
+    },
+    {
+      "epoch": 0.3020408163265306,
+      "grad_norm": 8.528164395288826,
+      "learning_rate": 1e-06,
+      "loss": 0.3207,
+      "step": 892
+    },
+    {
+      "epoch": 0.3047619047619048,
+      "grad_norm": 9.225399349162448,
+      "learning_rate": 1e-06,
+      "loss": 0.324,
+      "step": 893
+    },
+    {
+      "epoch": 0.3074829931972789,
+      "grad_norm": 7.0528833175610215,
+      "learning_rate": 1e-06,
+      "loss": 0.3243,
+      "step": 894
+    },
+    {
+      "epoch": 0.31020408163265306,
+      "grad_norm": 9.045541333709416,
+      "learning_rate": 1e-06,
+      "loss": 0.3031,
+      "step": 895
+    },
+    {
+      "epoch": 0.3129251700680272,
+      "grad_norm": 9.840199367937336,
+      "learning_rate": 1e-06,
+      "loss": 0.3588,
+      "step": 896
+    },
+    {
+      "epoch": 0.31564625850340133,
+      "grad_norm": 7.800716970319289,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 897
+    },
+    {
+      "epoch": 0.3183673469387755,
+      "grad_norm": 8.496577485888524,
+      "learning_rate": 1e-06,
+      "loss": 0.3001,
+      "step": 898
+    },
+    {
+      "epoch": 0.32108843537414966,
+      "grad_norm": 8.214573838445242,
+      "learning_rate": 1e-06,
+      "loss": 0.327,
+      "step": 899
+    },
+    {
+      "epoch": 0.3238095238095238,
+      "grad_norm": 9.308346167622362,
+      "learning_rate": 1e-06,
+      "loss": 0.3262,
+      "step": 900
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 7.607314669988555,
+      "learning_rate": 1e-06,
+      "loss": 0.3403,
+      "step": 901
+    },
+    {
+      "epoch": 0.3292517006802721,
+      "grad_norm": 10.614861151175514,
+      "learning_rate": 1e-06,
+      "loss": 0.3383,
+      "step": 902
+    },
+    {
+      "epoch": 0.33197278911564626,
+      "grad_norm": 8.292190462433707,
+      "learning_rate": 1e-06,
+      "loss": 0.3352,
+      "step": 903
+    },
+    {
+      "epoch": 0.3346938775510204,
+      "grad_norm": 9.157269650371953,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 904
+    },
+    {
+      "epoch": 0.33741496598639453,
+      "grad_norm": 6.563995347742781,
+      "learning_rate": 1e-06,
+      "loss": 0.3313,
+      "step": 905
+    },
+    {
+      "epoch": 0.3401360544217687,
+      "grad_norm": 9.125216480073076,
+      "learning_rate": 1e-06,
+      "loss": 0.3237,
+      "step": 906
+    },
+    {
+      "epoch": 0.34285714285714286,
+      "grad_norm": 11.41516726054542,
+      "learning_rate": 1e-06,
+      "loss": 0.3479,
+      "step": 907
+    },
+    {
+      "epoch": 0.345578231292517,
+      "grad_norm": 9.20154856879593,
+      "learning_rate": 1e-06,
+      "loss": 0.3412,
+      "step": 908
+    },
+    {
+      "epoch": 0.34829931972789113,
+      "grad_norm": 9.625276812713846,
+      "learning_rate": 1e-06,
+      "loss": 0.3069,
+      "step": 909
+    },
+    {
+      "epoch": 0.3510204081632653,
+      "grad_norm": 6.934240511377328,
+      "learning_rate": 1e-06,
+      "loss": 0.3009,
+      "step": 910
+    },
+    {
+      "epoch": 0.35374149659863946,
+      "grad_norm": 11.244350796436656,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 911
+    },
+    {
+      "epoch": 0.3564625850340136,
+      "grad_norm": 7.558316304769799,
+      "learning_rate": 1e-06,
+      "loss": 0.3114,
+      "step": 912
+    },
+    {
+      "epoch": 0.35918367346938773,
+      "grad_norm": 9.7275737993799,
+      "learning_rate": 1e-06,
+      "loss": 0.3472,
+      "step": 913
+    },
+    {
+      "epoch": 0.3619047619047619,
+      "grad_norm": 9.580315626356995,
+      "learning_rate": 1e-06,
+      "loss": 0.3236,
+      "step": 914
+    },
+    {
+      "epoch": 0.36462585034013606,
+      "grad_norm": 9.99667951745668,
+      "learning_rate": 1e-06,
+      "loss": 0.3441,
+      "step": 915
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 9.417933187517475,
+      "learning_rate": 1e-06,
+      "loss": 0.3089,
+      "step": 916
+    },
+    {
+      "epoch": 0.37006802721088433,
+      "grad_norm": 9.932357839475847,
+      "learning_rate": 1e-06,
+      "loss": 0.3319,
+      "step": 917
+    },
+    {
+      "epoch": 0.3727891156462585,
+      "grad_norm": 5.353821317285504,
+      "learning_rate": 1e-06,
+      "loss": 0.3099,
+      "step": 918
+    },
+    {
+      "epoch": 0.37551020408163266,
+      "grad_norm": 7.131456409496499,
+      "learning_rate": 1e-06,
+      "loss": 0.3233,
+      "step": 919
+    },
+    {
+      "epoch": 0.3782312925170068,
+      "grad_norm": 8.01997088703517,
+      "learning_rate": 1e-06,
+      "loss": 0.3105,
+      "step": 920
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 6.466271890435131,
+      "learning_rate": 1e-06,
+      "loss": 0.3421,
+      "step": 921
+    },
+    {
+      "epoch": 0.3836734693877551,
+      "grad_norm": 6.926224960194497,
+      "learning_rate": 1e-06,
+      "loss": 0.3273,
+      "step": 922
+    },
+    {
+      "epoch": 0.38639455782312926,
+      "grad_norm": 8.130863441172583,
+      "learning_rate": 1e-06,
+      "loss": 0.3257,
+      "step": 923
+    },
+    {
+      "epoch": 0.3891156462585034,
+      "grad_norm": 10.493096218238168,
+      "learning_rate": 1e-06,
+      "loss": 0.3307,
+      "step": 924
+    },
+    {
+      "epoch": 0.39183673469387753,
+      "grad_norm": 7.582033901124161,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 925
+    },
+    {
+      "epoch": 0.3945578231292517,
+      "grad_norm": 8.125727086448974,
+      "learning_rate": 1e-06,
+      "loss": 0.3187,
+      "step": 926
+    },
+    {
+      "epoch": 0.39727891156462586,
+      "grad_norm": 7.1151568602179465,
+      "learning_rate": 1e-06,
+      "loss": 0.3174,
+      "step": 927
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 9.387691433763987,
+      "learning_rate": 1e-06,
+      "loss": 0.3287,
+      "step": 928
+    },
+    {
+      "epoch": 0.40272108843537413,
+      "grad_norm": 9.70702165084111,
+      "learning_rate": 1e-06,
+      "loss": 0.3168,
+      "step": 929
+    },
+    {
+      "epoch": 0.4054421768707483,
+      "grad_norm": 11.25319029748679,
+      "learning_rate": 1e-06,
+      "loss": 0.3395,
+      "step": 930
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 10.3045032602604,
+      "learning_rate": 1e-06,
+      "loss": 0.3255,
+      "step": 931
+    },
+    {
+      "epoch": 0.4108843537414966,
+      "grad_norm": 8.523139411781436,
+      "learning_rate": 1e-06,
+      "loss": 0.3182,
+      "step": 932
+    },
+    {
+      "epoch": 0.41360544217687073,
+      "grad_norm": 7.717897388522883,
+      "learning_rate": 1e-06,
+      "loss": 0.3296,
+      "step": 933
+    },
+    {
+      "epoch": 0.4163265306122449,
+      "grad_norm": 10.081067253834975,
+      "learning_rate": 1e-06,
+      "loss": 0.3044,
+      "step": 934
+    },
+    {
+      "epoch": 0.41904761904761906,
+      "grad_norm": 7.6057657160179755,
+      "learning_rate": 1e-06,
+      "loss": 0.3162,
+      "step": 935
+    },
+    {
+      "epoch": 0.4217687074829932,
+      "grad_norm": 8.077946405303079,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 936
+    },
+    {
+      "epoch": 0.42448979591836733,
+      "grad_norm": 7.368504830608405,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 937
+    },
+    {
+      "epoch": 0.4272108843537415,
+      "grad_norm": 8.037048218837434,
+      "learning_rate": 1e-06,
+      "loss": 0.3185,
+      "step": 938
+    },
+    {
+      "epoch": 0.42993197278911566,
+      "grad_norm": 7.507503289979073,
+      "learning_rate": 1e-06,
+      "loss": 0.3347,
+      "step": 939
+    },
+    {
+      "epoch": 0.4326530612244898,
+      "grad_norm": 10.000540667230853,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 940
+    },
+    {
+      "epoch": 0.43537414965986393,
+      "grad_norm": 11.782678631843511,
+      "learning_rate": 1e-06,
+      "loss": 0.3285,
+      "step": 941
+    },
+    {
+      "epoch": 0.4380952380952381,
+      "grad_norm": 9.140624123658847,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 942
+    },
+    {
+      "epoch": 0.44081632653061226,
+      "grad_norm": 9.48131862293576,
+      "learning_rate": 1e-06,
+      "loss": 0.3567,
+      "step": 943
+    },
+    {
+      "epoch": 0.4435374149659864,
+      "grad_norm": 8.180316653752396,
+      "learning_rate": 1e-06,
+      "loss": 0.3351,
+      "step": 944
+    },
+    {
+      "epoch": 0.44625850340136053,
+      "grad_norm": 9.793904940418019,
+      "learning_rate": 1e-06,
+      "loss": 0.3436,
+      "step": 945
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 8.774196628695602,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 946
+    },
+    {
+      "epoch": 0.45170068027210886,
+      "grad_norm": 8.987498436970643,
+      "learning_rate": 1e-06,
+      "loss": 0.3287,
+      "step": 947
+    },
+    {
+      "epoch": 0.454421768707483,
+      "grad_norm": 7.824307630257181,
+      "learning_rate": 1e-06,
+      "loss": 0.3224,
+      "step": 948
+    },
+    {
+      "epoch": 0.45714285714285713,
+      "grad_norm": 8.456835936676253,
+      "learning_rate": 1e-06,
+      "loss": 0.3249,
+      "step": 949
+    },
+    {
+      "epoch": 0.4598639455782313,
+      "grad_norm": 9.301102397398253,
+      "learning_rate": 1e-06,
+      "loss": 0.3313,
+      "step": 950
+    },
+    {
+      "epoch": 0.46258503401360546,
+      "grad_norm": 10.463768025683592,
+      "learning_rate": 1e-06,
+      "loss": 0.3165,
+      "step": 951
+    },
+    {
+      "epoch": 0.46530612244897956,
+      "grad_norm": 8.687247357007717,
+      "learning_rate": 1e-06,
+      "loss": 0.3044,
+      "step": 952
+    },
+    {
+      "epoch": 0.46802721088435373,
+      "grad_norm": 8.614602962710912,
+      "learning_rate": 1e-06,
+      "loss": 0.3356,
+      "step": 953
+    },
+    {
+      "epoch": 0.4707482993197279,
+      "grad_norm": 10.236647198770886,
+      "learning_rate": 1e-06,
+      "loss": 0.3166,
+      "step": 954
+    },
+    {
+      "epoch": 0.47346938775510206,
+      "grad_norm": 9.500407207963299,
+      "learning_rate": 1e-06,
+      "loss": 0.3132,
+      "step": 955
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 9.35094738045064,
+      "learning_rate": 1e-06,
+      "loss": 0.3233,
+      "step": 956
+    },
+    {
+      "epoch": 0.47891156462585033,
+      "grad_norm": 8.043675750872643,
+      "learning_rate": 1e-06,
+      "loss": 0.3027,
+      "step": 957
+    },
+    {
+      "epoch": 0.4816326530612245,
+      "grad_norm": 6.981161977101288,
+      "learning_rate": 1e-06,
+      "loss": 0.3275,
+      "step": 958
+    },
+    {
+      "epoch": 0.48435374149659866,
+      "grad_norm": 8.681421342803574,
+      "learning_rate": 1e-06,
+      "loss": 0.3383,
+      "step": 959
+    },
+    {
+      "epoch": 0.48707482993197276,
+      "grad_norm": 8.847240360211114,
+      "learning_rate": 1e-06,
+      "loss": 0.3148,
+      "step": 960
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 6.9864004521375955,
+      "learning_rate": 1e-06,
+      "loss": 0.2941,
+      "step": 961
+    },
+    {
+      "epoch": 0.4925170068027211,
+      "grad_norm": 8.005675437969956,
+      "learning_rate": 1e-06,
+      "loss": 0.3363,
+      "step": 962
+    },
+    {
+      "epoch": 0.49523809523809526,
+      "grad_norm": 9.028511174445178,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 963
+    },
+    {
+      "epoch": 0.49795918367346936,
+      "grad_norm": 7.926792085676764,
+      "learning_rate": 1e-06,
+      "loss": 0.3557,
+      "step": 964
+    },
+    {
+      "epoch": 0.5006802721088436,
+      "grad_norm": 7.595713731891252,
+      "learning_rate": 1e-06,
+      "loss": 0.3428,
+      "step": 965
+    },
+    {
+      "epoch": 0.5034013605442177,
+      "grad_norm": 8.157359665200664,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 966
+    },
+    {
+      "epoch": 0.5061224489795918,
+      "grad_norm": 8.007431067593538,
+      "learning_rate": 1e-06,
+      "loss": 0.3413,
+      "step": 967
+    },
+    {
+      "epoch": 0.508843537414966,
+      "grad_norm": 7.829058322775787,
+      "learning_rate": 1e-06,
+      "loss": 0.3131,
+      "step": 968
+    },
+    {
+      "epoch": 0.5115646258503401,
+      "grad_norm": 9.955261606694947,
+      "learning_rate": 1e-06,
+      "loss": 0.3302,
+      "step": 969
+    },
+    {
+      "epoch": 0.5142857142857142,
+      "grad_norm": 7.687781618186785,
+      "learning_rate": 1e-06,
+      "loss": 0.3089,
+      "step": 970
+    },
+    {
+      "epoch": 0.5170068027210885,
+      "grad_norm": 8.199714028128824,
+      "learning_rate": 1e-06,
+      "loss": 0.3409,
+      "step": 971
+    },
+    {
+      "epoch": 0.5197278911564626,
+      "grad_norm": 8.757153294184452,
+      "learning_rate": 1e-06,
+      "loss": 0.296,
+      "step": 972
+    },
+    {
+      "epoch": 0.5224489795918368,
+      "grad_norm": 10.215121753460357,
+      "learning_rate": 1e-06,
+      "loss": 0.3536,
+      "step": 973
+    },
+    {
+      "epoch": 0.5251700680272109,
+      "grad_norm": 7.695409751768623,
+      "learning_rate": 1e-06,
+      "loss": 0.3318,
+      "step": 974
+    },
+    {
+      "epoch": 0.527891156462585,
+      "grad_norm": 9.240982021385662,
+      "learning_rate": 1e-06,
+      "loss": 0.3453,
+      "step": 975
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 7.813258584387002,
+      "learning_rate": 1e-06,
+      "loss": 0.3149,
+      "step": 976
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 7.650070646271369,
+      "learning_rate": 1e-06,
+      "loss": 0.3065,
+      "step": 977
+    },
+    {
+      "epoch": 0.5360544217687074,
+      "grad_norm": 9.765500319257182,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 978
+    },
+    {
+      "epoch": 0.5387755102040817,
+      "grad_norm": 8.099512013856767,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 979
+    },
+    {
+      "epoch": 0.5414965986394558,
+      "grad_norm": 8.71568109978164,
+      "learning_rate": 1e-06,
+      "loss": 0.3408,
+      "step": 980
+    },
+    {
+      "epoch": 0.54421768707483,
+      "grad_norm": 8.538287363863668,
+      "learning_rate": 1e-06,
+      "loss": 0.3316,
+      "step": 981
+    },
+    {
+      "epoch": 0.5469387755102041,
+      "grad_norm": 9.054133558526395,
+      "learning_rate": 1e-06,
+      "loss": 0.3304,
+      "step": 982
+    },
+    {
+      "epoch": 0.5496598639455782,
+      "grad_norm": 7.684353479574705,
+      "learning_rate": 1e-06,
+      "loss": 0.3488,
+      "step": 983
+    },
+    {
+      "epoch": 0.5523809523809524,
+      "grad_norm": 7.779728934976665,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 984
+    },
+    {
+      "epoch": 0.5551020408163265,
+      "grad_norm": 7.997501220301693,
+      "learning_rate": 1e-06,
+      "loss": 0.3115,
+      "step": 985
+    },
+    {
+      "epoch": 0.5578231292517006,
+      "grad_norm": 9.729787725089544,
+      "learning_rate": 1e-06,
+      "loss": 0.327,
+      "step": 986
+    },
+    {
+      "epoch": 0.5605442176870749,
+      "grad_norm": 8.395248094530618,
+      "learning_rate": 1e-06,
+      "loss": 0.3265,
+      "step": 987
+    },
+    {
+      "epoch": 0.563265306122449,
+      "grad_norm": 9.602445762247065,
+      "learning_rate": 1e-06,
+      "loss": 0.3108,
+      "step": 988
+    },
+    {
+      "epoch": 0.5659863945578232,
+      "grad_norm": 9.153187267695792,
+      "learning_rate": 1e-06,
+      "loss": 0.2861,
+      "step": 989
+    },
+    {
+      "epoch": 0.5687074829931973,
+      "grad_norm": 8.95600082478285,
+      "learning_rate": 1e-06,
+      "loss": 0.3194,
+      "step": 990
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 7.910411478151801,
+      "learning_rate": 1e-06,
+      "loss": 0.3107,
+      "step": 991
+    },
+    {
+      "epoch": 0.5741496598639456,
+      "grad_norm": 7.108411702112771,
+      "learning_rate": 1e-06,
+      "loss": 0.3382,
+      "step": 992
+    },
+    {
+      "epoch": 0.5768707482993197,
+      "grad_norm": 10.760575820319392,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 993
+    },
+    {
+      "epoch": 0.5795918367346938,
+      "grad_norm": 9.054761792346445,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 994
+    },
+    {
+      "epoch": 0.582312925170068,
+      "grad_norm": 11.629941951428535,
+      "learning_rate": 1e-06,
+      "loss": 0.3096,
+      "step": 995
+    },
+    {
+      "epoch": 0.5850340136054422,
+      "grad_norm": 7.6429358904979665,
+      "learning_rate": 1e-06,
+      "loss": 0.3005,
+      "step": 996
+    },
+    {
+      "epoch": 0.5877551020408164,
+      "grad_norm": 9.7765110010126,
+      "learning_rate": 1e-06,
+      "loss": 0.3306,
+      "step": 997
+    },
+    {
+      "epoch": 0.5904761904761905,
+      "grad_norm": 10.888969000089608,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 998
+    },
+    {
+      "epoch": 0.5931972789115646,
+      "grad_norm": 9.55980197918004,
+      "learning_rate": 1e-06,
+      "loss": 0.3307,
+      "step": 999
+    },
+    {
+      "epoch": 0.5959183673469388,
+      "grad_norm": 11.012883278490515,
+      "learning_rate": 1e-06,
+      "loss": 0.2995,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5959183673469388,
+      "eval_loss": 2.6654717922210693,
+      "eval_runtime": 21.7613,
+      "eval_samples_per_second": 45.953,
+      "eval_steps_per_second": 5.744,
+      "step": 1000
+    },
+    {
+      "epoch": 0.5986394557823129,
+      "grad_norm": 9.121120027212239,
+      "learning_rate": 1e-06,
+      "loss": 0.3272,
+      "step": 1001
+    },
+    {
+      "epoch": 0.601360544217687,
+      "grad_norm": 8.702324207936027,
+      "learning_rate": 1e-06,
+      "loss": 0.3194,
+      "step": 1002
+    },
+    {
+      "epoch": 0.6040816326530613,
+      "grad_norm": 9.202197123694965,
+      "learning_rate": 1e-06,
+      "loss": 0.3267,
+      "step": 1003
+    },
+    {
+      "epoch": 0.6068027210884354,
+      "grad_norm": 11.099893120443632,
+      "learning_rate": 1e-06,
+      "loss": 0.3218,
+      "step": 1004
+    },
+    {
+      "epoch": 0.6095238095238096,
+      "grad_norm": 8.510247323684077,
+      "learning_rate": 1e-06,
+      "loss": 0.3174,
+      "step": 1005
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.293054579305477,
+      "learning_rate": 1e-06,
+      "loss": 0.2985,
+      "step": 1006
+    },
+    {
+      "epoch": 0.6149659863945578,
+      "grad_norm": 9.007384259435915,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 1007
+    },
+    {
+      "epoch": 0.617687074829932,
+      "grad_norm": 7.53413548604913,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 1008
+    },
+    {
+      "epoch": 0.6204081632653061,
+      "grad_norm": 8.323377076424652,
+      "learning_rate": 1e-06,
+      "loss": 0.3245,
+      "step": 1009
+    },
+    {
+      "epoch": 0.6231292517006802,
+      "grad_norm": 6.866046553551617,
+      "learning_rate": 1e-06,
+      "loss": 0.3255,
+      "step": 1010
+    },
+    {
+      "epoch": 0.6258503401360545,
+      "grad_norm": 9.218686169720568,
+      "learning_rate": 1e-06,
+      "loss": 0.3082,
+      "step": 1011
+    },
+    {
+      "epoch": 0.6285714285714286,
+      "grad_norm": 8.089620935135077,
+      "learning_rate": 1e-06,
+      "loss": 0.3134,
+      "step": 1012
+    },
+    {
+      "epoch": 0.6312925170068027,
+      "grad_norm": 7.105162168992683,
+      "learning_rate": 1e-06,
+      "loss": 0.3289,
+      "step": 1013
+    },
+    {
+      "epoch": 0.6340136054421769,
+      "grad_norm": 8.011991165825659,
+      "learning_rate": 1e-06,
+      "loss": 0.3099,
+      "step": 1014
+    },
+    {
+      "epoch": 0.636734693877551,
+      "grad_norm": 6.999734413777645,
+      "learning_rate": 1e-06,
+      "loss": 0.303,
+      "step": 1015
+    },
+    {
+      "epoch": 0.6394557823129252,
+      "grad_norm": 9.16106325049575,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 1016
+    },
+    {
+      "epoch": 0.6421768707482993,
+      "grad_norm": 9.878200790406378,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 1017
+    },
+    {
+      "epoch": 0.6448979591836734,
+      "grad_norm": 6.87654213143073,
+      "learning_rate": 1e-06,
+      "loss": 0.3106,
+      "step": 1018
+    },
+    {
+      "epoch": 0.6476190476190476,
+      "grad_norm": 8.703705874856047,
+      "learning_rate": 1e-06,
+      "loss": 0.287,
+      "step": 1019
+    },
+    {
+      "epoch": 0.6503401360544218,
+      "grad_norm": 8.49730606491065,
+      "learning_rate": 1e-06,
+      "loss": 0.3309,
+      "step": 1020
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 8.727971335584828,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 1021
+    },
+    {
+      "epoch": 0.6557823129251701,
+      "grad_norm": 9.133994190553446,
+      "learning_rate": 1e-06,
+      "loss": 0.3381,
+      "step": 1022
+    },
+    {
+      "epoch": 0.6585034013605442,
+      "grad_norm": 7.6766604384864685,
+      "learning_rate": 1e-06,
+      "loss": 0.3082,
+      "step": 1023
+    },
+    {
+      "epoch": 0.6612244897959184,
+      "grad_norm": 7.9156025824778125,
+      "learning_rate": 1e-06,
+      "loss": 0.3272,
+      "step": 1024
+    },
+    {
+      "epoch": 0.6639455782312925,
+      "grad_norm": 7.3468347697329595,
+      "learning_rate": 1e-06,
+      "loss": 0.3143,
+      "step": 1025
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 8.664914147299818,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 1026
+    },
+    {
+      "epoch": 0.6693877551020408,
+      "grad_norm": 10.340458556458668,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 1027
+    },
+    {
+      "epoch": 0.672108843537415,
+      "grad_norm": 7.598306609017855,
+      "learning_rate": 1e-06,
+      "loss": 0.3098,
+      "step": 1028
+    },
+    {
+      "epoch": 0.6748299319727891,
+      "grad_norm": 6.829542688834379,
+      "learning_rate": 1e-06,
+      "loss": 0.326,
+      "step": 1029
+    },
+    {
+      "epoch": 0.6775510204081633,
+      "grad_norm": 7.155545939914754,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 1030
+    },
+    {
+      "epoch": 0.6802721088435374,
+      "grad_norm": 10.237470733106003,
+      "learning_rate": 1e-06,
+      "loss": 0.3129,
+      "step": 1031
+    },
+    {
+      "epoch": 0.6829931972789116,
+      "grad_norm": 6.488315257895372,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 1032
+    },
+    {
+      "epoch": 0.6857142857142857,
+      "grad_norm": 9.014874422504478,
+      "learning_rate": 1e-06,
+      "loss": 0.3069,
+      "step": 1033
+    },
+    {
+      "epoch": 0.6884353741496598,
+      "grad_norm": 7.394591141877351,
+      "learning_rate": 1e-06,
+      "loss": 0.3351,
+      "step": 1034
+    },
+    {
+      "epoch": 0.691156462585034,
+      "grad_norm": 10.027882904389081,
+      "learning_rate": 1e-06,
+      "loss": 0.3285,
+      "step": 1035
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 8.627014035428644,
+      "learning_rate": 1e-06,
+      "loss": 0.3132,
+      "step": 1036
+    },
+    {
+      "epoch": 0.6965986394557823,
+      "grad_norm": 7.890371508114607,
+      "learning_rate": 1e-06,
+      "loss": 0.3169,
+      "step": 1037
+    },
+    {
+      "epoch": 0.6993197278911565,
+      "grad_norm": 7.668873438965283,
+      "learning_rate": 1e-06,
+      "loss": 0.3178,
+      "step": 1038
+    },
+    {
+      "epoch": 0.7020408163265306,
+      "grad_norm": 7.91236639199808,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 1039
+    },
+    {
+      "epoch": 0.7047619047619048,
+      "grad_norm": 7.994635585156435,
+      "learning_rate": 1e-06,
+      "loss": 0.3222,
+      "step": 1040
+    },
+    {
+      "epoch": 0.7074829931972789,
+      "grad_norm": 7.159686250714685,
+      "learning_rate": 1e-06,
+      "loss": 0.3203,
+      "step": 1041
+    },
+    {
+      "epoch": 0.710204081632653,
+      "grad_norm": 6.989808745341601,
+      "learning_rate": 1e-06,
+      "loss": 0.2943,
+      "step": 1042
+    },
+    {
+      "epoch": 0.7129251700680272,
+      "grad_norm": 6.263330716712718,
+      "learning_rate": 1e-06,
+      "loss": 0.3235,
+      "step": 1043
+    },
+    {
+      "epoch": 0.7156462585034014,
+      "grad_norm": 8.694140993888391,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 1044
+    },
+    {
+      "epoch": 0.7183673469387755,
+      "grad_norm": 7.024840486956401,
+      "learning_rate": 1e-06,
+      "loss": 0.3364,
+      "step": 1045
+    },
+    {
+      "epoch": 0.7210884353741497,
+      "grad_norm": 7.6778273536950605,
+      "learning_rate": 1e-06,
+      "loss": 0.3254,
+      "step": 1046
+    },
+    {
+      "epoch": 0.7238095238095238,
+      "grad_norm": 7.617518374099135,
+      "learning_rate": 1e-06,
+      "loss": 0.3375,
+      "step": 1047
+    },
+    {
+      "epoch": 0.726530612244898,
+      "grad_norm": 10.78179655779082,
+      "learning_rate": 1e-06,
+      "loss": 0.3187,
+      "step": 1048
+    },
+    {
+      "epoch": 0.7292517006802721,
+      "grad_norm": 6.730192995137268,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 1049
+    },
+    {
+      "epoch": 0.7319727891156462,
+      "grad_norm": 7.628686995817748,
+      "learning_rate": 1e-06,
+      "loss": 0.3463,
+      "step": 1050
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 7.048497847694406,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 1051
+    },
+    {
+      "epoch": 0.7374149659863946,
+      "grad_norm": 7.488613198072443,
+      "learning_rate": 1e-06,
+      "loss": 0.3295,
+      "step": 1052
+    },
+    {
+      "epoch": 0.7401360544217687,
+      "grad_norm": 8.736566837754694,
+      "learning_rate": 1e-06,
+      "loss": 0.3289,
+      "step": 1053
+    },
+    {
+      "epoch": 0.7428571428571429,
+      "grad_norm": 11.177384730643242,
+      "learning_rate": 1e-06,
+      "loss": 0.3382,
+      "step": 1054
+    },
+    {
+      "epoch": 0.745578231292517,
+      "grad_norm": 7.58182434896046,
+      "learning_rate": 1e-06,
+      "loss": 0.3089,
+      "step": 1055
+    },
+    {
+      "epoch": 0.7482993197278912,
+      "grad_norm": 7.834883201632613,
+      "learning_rate": 1e-06,
+      "loss": 0.3358,
+      "step": 1056
+    },
+    {
+      "epoch": 0.7510204081632653,
+      "grad_norm": 9.27592130340726,
+      "learning_rate": 1e-06,
+      "loss": 0.3396,
+      "step": 1057
+    },
+    {
+      "epoch": 0.7537414965986394,
+      "grad_norm": 7.062067896755278,
+      "learning_rate": 1e-06,
+      "loss": 0.2757,
+      "step": 1058
+    },
+    {
+      "epoch": 0.7564625850340136,
+      "grad_norm": 6.940158687423098,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 1059
+    },
+    {
+      "epoch": 0.7591836734693878,
+      "grad_norm": 6.158032157535268,
+      "learning_rate": 1e-06,
+      "loss": 0.3414,
+      "step": 1060
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 7.872707382772136,
+      "learning_rate": 1e-06,
+      "loss": 0.3342,
+      "step": 1061
+    },
+    {
+      "epoch": 0.7646258503401361,
+      "grad_norm": 9.802437862501364,
+      "learning_rate": 1e-06,
+      "loss": 0.3168,
+      "step": 1062
+    },
+    {
+      "epoch": 0.7673469387755102,
+      "grad_norm": 8.579270173311855,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 1063
+    },
+    {
+      "epoch": 0.7700680272108843,
+      "grad_norm": 8.773615676543434,
+      "learning_rate": 1e-06,
+      "loss": 0.3023,
+      "step": 1064
+    },
+    {
+      "epoch": 0.7727891156462585,
+      "grad_norm": 7.985379589608822,
+      "learning_rate": 1e-06,
+      "loss": 0.3258,
+      "step": 1065
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 8.247706067103946,
+      "learning_rate": 1e-06,
+      "loss": 0.3153,
+      "step": 1066
+    },
+    {
+      "epoch": 0.7782312925170068,
+      "grad_norm": 7.144117583353914,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 1067
+    },
+    {
+      "epoch": 0.780952380952381,
+      "grad_norm": 8.456850050379195,
+      "learning_rate": 1e-06,
+      "loss": 0.3357,
+      "step": 1068
+    },
+    {
+      "epoch": 0.7836734693877551,
+      "grad_norm": 10.740680402936311,
+      "learning_rate": 1e-06,
+      "loss": 0.3256,
+      "step": 1069
+    },
+    {
+      "epoch": 0.7863945578231293,
+      "grad_norm": 6.780669098139295,
+      "learning_rate": 1e-06,
+      "loss": 0.3002,
+      "step": 1070
+    },
+    {
+      "epoch": 0.7891156462585034,
+      "grad_norm": 9.445976941777388,
+      "learning_rate": 1e-06,
+      "loss": 0.3097,
+      "step": 1071
+    },
+    {
+      "epoch": 0.7918367346938775,
+      "grad_norm": 8.017942664978193,
+      "learning_rate": 1e-06,
+      "loss": 0.3252,
+      "step": 1072
+    },
+    {
+      "epoch": 0.7945578231292517,
+      "grad_norm": 10.012890799595358,
+      "learning_rate": 1e-06,
+      "loss": 0.3353,
+      "step": 1073
+    },
+    {
+      "epoch": 0.7972789115646258,
+      "grad_norm": 10.121783009790892,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 1074
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 10.844887677794592,
+      "learning_rate": 1e-06,
+      "loss": 0.3195,
+      "step": 1075
+    },
+    {
+      "epoch": 0.8027210884353742,
+      "grad_norm": 7.3329327716066395,
+      "learning_rate": 1e-06,
+      "loss": 0.306,
+      "step": 1076
+    },
+    {
+      "epoch": 0.8054421768707483,
+      "grad_norm": 9.764817321553252,
+      "learning_rate": 1e-06,
+      "loss": 0.3098,
+      "step": 1077
+    },
+    {
+      "epoch": 0.8081632653061225,
+      "grad_norm": 8.79758352673661,
+      "learning_rate": 1e-06,
+      "loss": 0.3469,
+      "step": 1078
+    },
+    {
+      "epoch": 0.8108843537414966,
+      "grad_norm": 6.775159433727832,
+      "learning_rate": 1e-06,
+      "loss": 0.3312,
+      "step": 1079
+    },
+    {
+      "epoch": 0.8136054421768707,
+      "grad_norm": 8.334098425744775,
+      "learning_rate": 1e-06,
+      "loss": 0.3315,
+      "step": 1080
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 6.179748894815101,
+      "learning_rate": 1e-06,
+      "loss": 0.3016,
+      "step": 1081
+    },
+    {
+      "epoch": 0.819047619047619,
+      "grad_norm": 5.050454609937429,
+      "learning_rate": 1e-06,
+      "loss": 0.3238,
+      "step": 1082
+    },
+    {
+      "epoch": 0.8217687074829932,
+      "grad_norm": 7.364071588114907,
+      "learning_rate": 1e-06,
+      "loss": 0.3202,
+      "step": 1083
+    },
+    {
+      "epoch": 0.8244897959183674,
+      "grad_norm": 7.345744040075277,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 1084
+    },
+    {
+      "epoch": 0.8272108843537415,
+      "grad_norm": 9.763737201716557,
+      "learning_rate": 1e-06,
+      "loss": 0.3142,
+      "step": 1085
+    },
+    {
+      "epoch": 0.8299319727891157,
+      "grad_norm": 8.7106021344788,
+      "learning_rate": 1e-06,
+      "loss": 0.3049,
+      "step": 1086
+    },
+    {
+      "epoch": 0.8326530612244898,
+      "grad_norm": 8.607345168085638,
+      "learning_rate": 1e-06,
+      "loss": 0.3195,
+      "step": 1087
+    },
+    {
+      "epoch": 0.8353741496598639,
+      "grad_norm": 9.04526585039096,
+      "learning_rate": 1e-06,
+      "loss": 0.3193,
+      "step": 1088
+    },
+    {
+      "epoch": 0.8380952380952381,
+      "grad_norm": 7.341716131272211,
+      "learning_rate": 1e-06,
+      "loss": 0.3155,
+      "step": 1089
+    },
+    {
+      "epoch": 0.8408163265306122,
+      "grad_norm": 10.05957865205487,
+      "learning_rate": 1e-06,
+      "loss": 0.3236,
+      "step": 1090
+    },
+    {
+      "epoch": 0.8435374149659864,
+      "grad_norm": 8.907245085155065,
+      "learning_rate": 1e-06,
+      "loss": 0.3032,
+      "step": 1091
+    },
+    {
+      "epoch": 0.8462585034013606,
+      "grad_norm": 7.610234897380251,
+      "learning_rate": 1e-06,
+      "loss": 0.3188,
+      "step": 1092
+    },
+    {
+      "epoch": 0.8489795918367347,
+      "grad_norm": 9.29558824941524,
+      "learning_rate": 1e-06,
+      "loss": 0.3443,
+      "step": 1093
+    },
+    {
+      "epoch": 0.8517006802721089,
+      "grad_norm": 8.645312110751389,
+      "learning_rate": 1e-06,
+      "loss": 0.3267,
+      "step": 1094
+    },
+    {
+      "epoch": 0.854421768707483,
+      "grad_norm": 7.972329121784028,
+      "learning_rate": 1e-06,
+      "loss": 0.3063,
+      "step": 1095
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 7.081096833670109,
+      "learning_rate": 1e-06,
+      "loss": 0.3162,
+      "step": 1096
+    },
+    {
+      "epoch": 0.8598639455782313,
+      "grad_norm": 10.382289641169832,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 1097
+    },
+    {
+      "epoch": 0.8625850340136054,
+      "grad_norm": 7.978211204132755,
+      "learning_rate": 1e-06,
+      "loss": 0.331,
+      "step": 1098
+    },
+    {
+      "epoch": 0.8653061224489796,
+      "grad_norm": 8.647863811889165,
+      "learning_rate": 1e-06,
+      "loss": 0.2968,
+      "step": 1099
+    },
+    {
+      "epoch": 0.8680272108843538,
+      "grad_norm": 6.71797658443923,
+      "learning_rate": 1e-06,
+      "loss": 0.3136,
+      "step": 1100
+    },
+    {
+      "epoch": 0.8707482993197279,
+      "grad_norm": 10.935364261408203,
+      "learning_rate": 1e-06,
+      "loss": 0.3226,
+      "step": 1101
+    },
+    {
+      "epoch": 0.8734693877551021,
+      "grad_norm": 8.216608114513559,
+      "learning_rate": 1e-06,
+      "loss": 0.3256,
+      "step": 1102
+    },
+    {
+      "epoch": 0.8761904761904762,
+      "grad_norm": 7.197913639103016,
+      "learning_rate": 1e-06,
+      "loss": 0.3292,
+      "step": 1103
+    },
+    {
+      "epoch": 0.8789115646258503,
+      "grad_norm": 8.361528527253666,
+      "learning_rate": 1e-06,
+      "loss": 0.3131,
+      "step": 1104
+    },
+    {
+      "epoch": 0.8816326530612245,
+      "grad_norm": 8.834806409502773,
+      "learning_rate": 1e-06,
+      "loss": 0.305,
+      "step": 1105
+    },
+    {
+      "epoch": 0.8843537414965986,
+      "grad_norm": 8.058499872224347,
+      "learning_rate": 1e-06,
+      "loss": 0.3117,
+      "step": 1106
+    },
+    {
+      "epoch": 0.8870748299319728,
+      "grad_norm": 10.512019109660894,
+      "learning_rate": 1e-06,
+      "loss": 0.3185,
+      "step": 1107
+    },
+    {
+      "epoch": 0.889795918367347,
+      "grad_norm": 8.158048209808793,
+      "learning_rate": 1e-06,
+      "loss": 0.3143,
+      "step": 1108
+    },
+    {
+      "epoch": 0.8925170068027211,
+      "grad_norm": 8.501626778993263,
+      "learning_rate": 1e-06,
+      "loss": 0.3133,
+      "step": 1109
+    },
+    {
+      "epoch": 0.8952380952380953,
+      "grad_norm": 9.486594573630693,
+      "learning_rate": 1e-06,
+      "loss": 0.325,
+      "step": 1110
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 9.365791353823768,
+      "learning_rate": 1e-06,
+      "loss": 0.3076,
+      "step": 1111
+    },
+    {
+      "epoch": 0.9006802721088435,
+      "grad_norm": 6.348696519817804,
+      "learning_rate": 1e-06,
+      "loss": 0.3243,
+      "step": 1112
+    },
+    {
+      "epoch": 0.9034013605442177,
+      "grad_norm": 7.939505881445669,
+      "learning_rate": 1e-06,
+      "loss": 0.3169,
+      "step": 1113
+    },
+    {
+      "epoch": 0.9061224489795918,
+      "grad_norm": 9.789408828112952,
+      "learning_rate": 1e-06,
+      "loss": 0.3207,
+      "step": 1114
+    },
+    {
+      "epoch": 0.908843537414966,
+      "grad_norm": 7.253212765388169,
+      "learning_rate": 1e-06,
+      "loss": 0.3308,
+      "step": 1115
+    },
+    {
+      "epoch": 0.9115646258503401,
+      "grad_norm": 7.938551715491752,
+      "learning_rate": 1e-06,
+      "loss": 0.2894,
+      "step": 1116
+    },
+    {
+      "epoch": 0.9142857142857143,
+      "grad_norm": 7.998951242843783,
+      "learning_rate": 1e-06,
+      "loss": 0.2951,
+      "step": 1117
+    },
+    {
+      "epoch": 0.9170068027210885,
+      "grad_norm": 7.013929305715802,
+      "learning_rate": 1e-06,
+      "loss": 0.3165,
+      "step": 1118
+    },
+    {
+      "epoch": 0.9197278911564626,
+      "grad_norm": 8.244856642691825,
+      "learning_rate": 1e-06,
+      "loss": 0.3002,
+      "step": 1119
+    },
+    {
+      "epoch": 0.9224489795918367,
+      "grad_norm": 8.649319407625757,
+      "learning_rate": 1e-06,
+      "loss": 0.3319,
+      "step": 1120
+    },
+    {
+      "epoch": 0.9251700680272109,
+      "grad_norm": 6.434344003334588,
+      "learning_rate": 1e-06,
+      "loss": 0.3362,
+      "step": 1121
+    },
+    {
+      "epoch": 0.927891156462585,
+      "grad_norm": 7.986760897279392,
+      "learning_rate": 1e-06,
+      "loss": 0.3323,
+      "step": 1122
+    },
+    {
+      "epoch": 0.9306122448979591,
+      "grad_norm": 8.208262312615274,
+      "learning_rate": 1e-06,
+      "loss": 0.3237,
+      "step": 1123
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 9.99436829706517,
+      "learning_rate": 1e-06,
+      "loss": 0.3379,
+      "step": 1124
+    },
+    {
+      "epoch": 0.9360544217687075,
+      "grad_norm": 8.348458305330084,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 1125
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 9.09283050096805,
+      "learning_rate": 1e-06,
+      "loss": 0.2975,
+      "step": 1126
+    },
+    {
+      "epoch": 0.9414965986394558,
+      "grad_norm": 8.475697194604672,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 1127
+    },
+    {
+      "epoch": 0.9442176870748299,
+      "grad_norm": 7.422483226799972,
+      "learning_rate": 1e-06,
+      "loss": 0.339,
+      "step": 1128
+    },
+    {
+      "epoch": 0.9469387755102041,
+      "grad_norm": 9.157380492537378,
+      "learning_rate": 1e-06,
+      "loss": 0.3273,
+      "step": 1129
+    },
+    {
+      "epoch": 0.9496598639455782,
+      "grad_norm": 7.6216100979177694,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 1130
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 8.632690529209786,
+      "learning_rate": 1e-06,
+      "loss": 0.3013,
+      "step": 1131
+    },
+    {
+      "epoch": 0.9551020408163265,
+      "grad_norm": 7.613282274701332,
+      "learning_rate": 1e-06,
+      "loss": 0.3213,
+      "step": 1132
+    },
+    {
+      "epoch": 0.9578231292517007,
+      "grad_norm": 8.598393697909668,
+      "learning_rate": 1e-06,
+      "loss": 0.3218,
+      "step": 1133
+    },
+    {
+      "epoch": 0.9605442176870749,
+      "grad_norm": 10.294283633828698,
+      "learning_rate": 1e-06,
+      "loss": 0.3296,
+      "step": 1134
+    },
+    {
+      "epoch": 0.963265306122449,
+      "grad_norm": 7.916360129409138,
+      "learning_rate": 1e-06,
+      "loss": 0.3115,
+      "step": 1135
+    },
+    {
+      "epoch": 0.9659863945578231,
+      "grad_norm": 8.63830256620055,
+      "learning_rate": 1e-06,
+      "loss": 0.3222,
+      "step": 1136
+    },
+    {
+      "epoch": 0.9687074829931973,
+      "grad_norm": 6.039379787854105,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 1137
+    },
+    {
+      "epoch": 0.9714285714285714,
+      "grad_norm": 8.308960252670737,
+      "learning_rate": 1e-06,
+      "loss": 0.3229,
+      "step": 1138
+    },
+    {
+      "epoch": 0.9741496598639455,
+      "grad_norm": 9.445670876801511,
+      "learning_rate": 1e-06,
+      "loss": 0.3355,
+      "step": 1139
+    },
+    {
+      "epoch": 0.9768707482993197,
+      "grad_norm": 8.172807921784118,
+      "learning_rate": 1e-06,
+      "loss": 0.3061,
+      "step": 1140
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 11.84477753905021,
+      "learning_rate": 1e-06,
+      "loss": 0.2952,
+      "step": 1141
+    },
+    {
+      "epoch": 0.9823129251700681,
+      "grad_norm": 7.828991660815482,
+      "learning_rate": 1e-06,
+      "loss": 0.3066,
+      "step": 1142
+    },
+    {
+      "epoch": 0.9850340136054422,
+      "grad_norm": 10.09221731979383,
+      "learning_rate": 1e-06,
+      "loss": 0.3141,
+      "step": 1143
+    },
+    {
+      "epoch": 0.9877551020408163,
+      "grad_norm": 8.463698075047468,
+      "learning_rate": 1e-06,
+      "loss": 0.3319,
+      "step": 1144
+    },
+    {
+      "epoch": 0.9904761904761905,
+      "grad_norm": 10.058306855084675,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 1145
+    },
+    {
+      "epoch": 0.9931972789115646,
+      "grad_norm": 8.148610507431059,
+      "learning_rate": 1e-06,
+      "loss": 0.3102,
+      "step": 1146
+    },
+    {
+      "epoch": 0.9959183673469387,
+      "grad_norm": 6.482236778708309,
+      "learning_rate": 1e-06,
+      "loss": 0.3269,
+      "step": 1147
+    },
+    {
+      "epoch": 0.998639455782313,
+      "grad_norm": 8.500003243097416,
+      "learning_rate": 1e-06,
+      "loss": 0.2998,
+      "step": 1148
+    },
+    {
+      "epoch": 0.998639455782313,
+      "eval_loss": 2.727992296218872,
+      "eval_runtime": 21.421,
+      "eval_samples_per_second": 46.683,
+      "eval_steps_per_second": 5.835,
+      "step": 1148
+    },
+    {
+      "epoch": 0.0028011204481792717,
+      "grad_norm": 6.3763214558323265,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 1149
+    },
+    {
+      "epoch": 0.0056022408963585435,
+      "grad_norm": 7.908490771048352,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 1150
+    },
+    {
+      "epoch": 0.008403361344537815,
+      "grad_norm": 7.724350262993959,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 1151
+    },
+    {
+      "epoch": 0.011204481792717087,
+      "grad_norm": 6.880785314806721,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 1152
+    },
+    {
+      "epoch": 0.014005602240896359,
+      "grad_norm": 9.190046057741924,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 1153
+    },
+    {
+      "epoch": 0.01680672268907563,
+      "grad_norm": 8.0283238663745,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 1154
+    },
+    {
+      "epoch": 0.0196078431372549,
+      "grad_norm": 7.830894635574845,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 1155
+    },
+    {
+      "epoch": 0.022408963585434174,
+      "grad_norm": 8.040445637742431,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 1156
+    },
+    {
+      "epoch": 0.025210084033613446,
+      "grad_norm": 8.789679009179983,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 1157
+    },
+    {
+      "epoch": 0.028011204481792718,
+      "grad_norm": 8.770690635992432,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 1158
+    },
+    {
+      "epoch": 0.03081232492997199,
+      "grad_norm": 9.42848335651522,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 1159
+    },
+    {
+      "epoch": 0.03361344537815126,
+      "grad_norm": 10.310502222725566,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 1160
+    },
+    {
+      "epoch": 0.036414565826330535,
+      "grad_norm": 11.726153773185464,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 1161
+    },
+    {
+      "epoch": 0.0392156862745098,
+      "grad_norm": 9.02602198616084,
+      "learning_rate": 1e-06,
+      "loss": 0.1961,
+      "step": 1162
+    },
+    {
+      "epoch": 0.04201680672268908,
+      "grad_norm": 10.342943044533717,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 1163
+    },
+    {
+      "epoch": 0.04481792717086835,
+      "grad_norm": 9.923811095653315,
+      "learning_rate": 1e-06,
+      "loss": 0.1805,
+      "step": 1164
+    },
+    {
+      "epoch": 0.047619047619047616,
+      "grad_norm": 11.385621057968345,
+      "learning_rate": 1e-06,
+      "loss": 0.1908,
+      "step": 1165
+    },
+    {
+      "epoch": 0.05042016806722689,
+      "grad_norm": 9.371547370220073,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 1166
+    },
+    {
+      "epoch": 0.05322128851540616,
+      "grad_norm": 11.188903786806657,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 1167
+    },
+    {
+      "epoch": 0.056022408963585436,
+      "grad_norm": 12.523786502316732,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 1168
+    },
+    {
+      "epoch": 0.058823529411764705,
+      "grad_norm": 10.801947568764506,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 1169
+    },
+    {
+      "epoch": 0.06162464985994398,
+      "grad_norm": 15.827277970304726,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 1170
+    },
+    {
+      "epoch": 0.06442577030812324,
+      "grad_norm": 12.597793008150722,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 1171
+    },
+    {
+      "epoch": 0.06722689075630252,
+      "grad_norm": 7.3677272853293685,
+      "learning_rate": 1e-06,
+      "loss": 0.1792,
+      "step": 1172
+    },
+    {
+      "epoch": 0.0700280112044818,
+      "grad_norm": 10.839974343978072,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 1173
+    },
+    {
+      "epoch": 0.07282913165266107,
+      "grad_norm": 9.779879692051066,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 1174
+    },
+    {
+      "epoch": 0.07563025210084033,
+      "grad_norm": 10.687766250740673,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 1175
+    },
+    {
+      "epoch": 0.0784313725490196,
+      "grad_norm": 8.155065192157387,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 1176
+    },
+    {
+      "epoch": 0.08123249299719888,
+      "grad_norm": 9.854227287667177,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 1177
+    },
+    {
+      "epoch": 0.08403361344537816,
+      "grad_norm": 7.162536325412529,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 1178
+    },
+    {
+      "epoch": 0.08683473389355742,
+      "grad_norm": 10.740172165634181,
+      "learning_rate": 1e-06,
+      "loss": 0.1992,
+      "step": 1179
+    },
+    {
+      "epoch": 0.0896358543417367,
+      "grad_norm": 11.265071793869666,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1180
+    },
+    {
+      "epoch": 0.09243697478991597,
+      "grad_norm": 9.82839791550283,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 1181
+    },
+    {
+      "epoch": 0.09523809523809523,
+      "grad_norm": 9.865422806688365,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 1182
+    },
+    {
+      "epoch": 0.09803921568627451,
+      "grad_norm": 8.721224881857536,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 1183
+    },
+    {
+      "epoch": 0.10084033613445378,
+      "grad_norm": 7.893937414814106,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 1184
+    },
+    {
+      "epoch": 0.10364145658263306,
+      "grad_norm": 7.796399628731324,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1185
+    },
+    {
+      "epoch": 0.10644257703081232,
+      "grad_norm": 10.712752673785303,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 1186
+    },
+    {
+      "epoch": 0.1092436974789916,
+      "grad_norm": 10.229461934584354,
+      "learning_rate": 1e-06,
+      "loss": 0.1942,
+      "step": 1187
+    },
+    {
+      "epoch": 0.11204481792717087,
+      "grad_norm": 7.801598280667384,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 1188
+    },
+    {
+      "epoch": 0.11484593837535013,
+      "grad_norm": 9.974255369344881,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 1189
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 9.286844779186636,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 1190
+    },
+    {
+      "epoch": 0.12044817927170869,
+      "grad_norm": 8.599089004831896,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 1191
+    },
+    {
+      "epoch": 0.12324929971988796,
+      "grad_norm": 8.02887990963132,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 1192
+    },
+    {
+      "epoch": 0.12605042016806722,
+      "grad_norm": 9.319945994425108,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 1193
+    },
+    {
+      "epoch": 0.12885154061624648,
+      "grad_norm": 7.882573926465686,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 1194
+    },
+    {
+      "epoch": 0.13165266106442577,
+      "grad_norm": 11.337624204981694,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 1195
+    },
+    {
+      "epoch": 0.13445378151260504,
+      "grad_norm": 10.130586981869005,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 1196
+    },
+    {
+      "epoch": 0.13725490196078433,
+      "grad_norm": 10.575831598221685,
+      "learning_rate": 1e-06,
+      "loss": 0.1789,
+      "step": 1197
+    },
+    {
+      "epoch": 0.1400560224089636,
+      "grad_norm": 10.445682908347411,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 1198
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 13.73901235664324,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 1199
+    },
+    {
+      "epoch": 0.14565826330532214,
+      "grad_norm": 7.9491633478168335,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 1200
+    },
+    {
+      "epoch": 0.1484593837535014,
+      "grad_norm": 7.79082146426804,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 1201
+    },
+    {
+      "epoch": 0.15126050420168066,
+      "grad_norm": 10.930010623548943,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 1202
+    },
+    {
+      "epoch": 0.15406162464985995,
+      "grad_norm": 10.03897446217369,
+      "learning_rate": 1e-06,
+      "loss": 0.1865,
+      "step": 1203
+    },
+    {
+      "epoch": 0.1568627450980392,
+      "grad_norm": 9.36681995961309,
+      "learning_rate": 1e-06,
+      "loss": 0.1836,
+      "step": 1204
+    },
+    {
+      "epoch": 0.15966386554621848,
+      "grad_norm": 11.602959341972761,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 1205
+    },
+    {
+      "epoch": 0.16246498599439776,
+      "grad_norm": 8.06890018088231,
+      "learning_rate": 1e-06,
+      "loss": 0.1792,
+      "step": 1206
+    },
+    {
+      "epoch": 0.16526610644257703,
+      "grad_norm": 9.832480536350069,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 1207
+    },
+    {
+      "epoch": 0.16806722689075632,
+      "grad_norm": 8.974841482590193,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 1208
+    },
+    {
+      "epoch": 0.17086834733893558,
+      "grad_norm": 9.43419504445943,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 1209
+    },
+    {
+      "epoch": 0.17366946778711484,
+      "grad_norm": 11.12812909143882,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 1210
+    },
+    {
+      "epoch": 0.17647058823529413,
+      "grad_norm": 9.592277688266005,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 1211
+    },
+    {
+      "epoch": 0.1792717086834734,
+      "grad_norm": 10.829170519640591,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 1212
+    },
+    {
+      "epoch": 0.18207282913165265,
+      "grad_norm": 8.875840183519912,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 1213
+    },
+    {
+      "epoch": 0.18487394957983194,
+      "grad_norm": 12.573079191892061,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 1214
+    },
+    {
+      "epoch": 0.1876750700280112,
+      "grad_norm": 8.780389182367665,
+      "learning_rate": 1e-06,
+      "loss": 0.1885,
+      "step": 1215
+    },
+    {
+      "epoch": 0.19047619047619047,
+      "grad_norm": 11.124554279226135,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 1216
+    },
+    {
+      "epoch": 0.19327731092436976,
+      "grad_norm": 7.724009953750721,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 1217
+    },
+    {
+      "epoch": 0.19607843137254902,
+      "grad_norm": 8.678681908896214,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 1218
+    },
+    {
+      "epoch": 0.19887955182072828,
+      "grad_norm": 8.601822006818285,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 1219
+    },
+    {
+      "epoch": 0.20168067226890757,
+      "grad_norm": 7.624460648968144,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 1220
+    },
+    {
+      "epoch": 0.20448179271708683,
+      "grad_norm": 10.340435047856234,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 1221
+    },
+    {
+      "epoch": 0.20728291316526612,
+      "grad_norm": 13.813205288289456,
+      "learning_rate": 1e-06,
+      "loss": 0.1936,
+      "step": 1222
+    },
+    {
+      "epoch": 0.21008403361344538,
+      "grad_norm": 12.508645313698064,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 1223
+    },
+    {
+      "epoch": 0.21288515406162464,
+      "grad_norm": 9.731454467458098,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 1224
+    },
+    {
+      "epoch": 0.21568627450980393,
+      "grad_norm": 10.171097811566083,
+      "learning_rate": 1e-06,
+      "loss": 0.184,
+      "step": 1225
+    },
+    {
+      "epoch": 0.2184873949579832,
+      "grad_norm": 10.96121875262816,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 1226
+    },
+    {
+      "epoch": 0.22128851540616246,
+      "grad_norm": 10.376473714328721,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 1227
+    },
+    {
+      "epoch": 0.22408963585434175,
+      "grad_norm": 9.460404873405627,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 1228
+    },
+    {
+      "epoch": 0.226890756302521,
+      "grad_norm": 10.597127984265079,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 1229
+    },
+    {
+      "epoch": 0.22969187675070027,
+      "grad_norm": 7.415990581745192,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 1230
+    },
+    {
+      "epoch": 0.23249299719887956,
+      "grad_norm": 10.422887578330768,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 1231
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 13.803036349870526,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 1232
+    },
+    {
+      "epoch": 0.23809523809523808,
+      "grad_norm": 12.950698659713554,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 1233
+    },
+    {
+      "epoch": 0.24089635854341737,
+      "grad_norm": 9.061777185931938,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 1234
+    },
+    {
+      "epoch": 0.24369747899159663,
+      "grad_norm": 10.115981045118303,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 1235
+    },
+    {
+      "epoch": 0.24649859943977592,
+      "grad_norm": 10.571715274240255,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 1236
+    },
+    {
+      "epoch": 0.24929971988795518,
+      "grad_norm": 9.447277478738446,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 1237
+    },
+    {
+      "epoch": 0.25210084033613445,
+      "grad_norm": 8.706533636765567,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 1238
+    },
+    {
+      "epoch": 0.2549019607843137,
+      "grad_norm": 12.334835537504908,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 1239
+    },
+    {
+      "epoch": 0.25770308123249297,
+      "grad_norm": 8.216518917532714,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 1240
+    },
+    {
+      "epoch": 0.2605042016806723,
+      "grad_norm": 12.540138694658744,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 1241
+    },
+    {
+      "epoch": 0.26330532212885155,
+      "grad_norm": 10.104346829810652,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 1242
+    },
+    {
+      "epoch": 0.2661064425770308,
+      "grad_norm": 10.19950935378434,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 1243
+    },
+    {
+      "epoch": 0.2689075630252101,
+      "grad_norm": 10.119118543577592,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 1244
+    },
+    {
+      "epoch": 0.27170868347338933,
+      "grad_norm": 9.566519433305958,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 1245
+    },
+    {
+      "epoch": 0.27450980392156865,
+      "grad_norm": 7.830068028823046,
+      "learning_rate": 1e-06,
+      "loss": 0.1769,
+      "step": 1246
+    },
+    {
+      "epoch": 0.2773109243697479,
+      "grad_norm": 9.529588508783243,
+      "learning_rate": 1e-06,
+      "loss": 0.1738,
+      "step": 1247
+    },
+    {
+      "epoch": 0.2801120448179272,
+      "grad_norm": 9.008821624125842,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1248
+    },
+    {
+      "epoch": 0.28291316526610644,
+      "grad_norm": 10.506686959334864,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 1249
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 9.1405322419151,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 1250
+    },
+    {
+      "epoch": 0.28851540616246496,
+      "grad_norm": 10.775586910274638,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 1251
+    },
+    {
+      "epoch": 0.2913165266106443,
+      "grad_norm": 9.905821421828268,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 1252
+    },
+    {
+      "epoch": 0.29411764705882354,
+      "grad_norm": 9.519156807834904,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 1253
+    },
+    {
+      "epoch": 0.2969187675070028,
+      "grad_norm": 8.748636402614776,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 1254
+    },
+    {
+      "epoch": 0.29971988795518206,
+      "grad_norm": 10.532829400231362,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 1255
+    },
+    {
+      "epoch": 0.3025210084033613,
+      "grad_norm": 8.593968311863888,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 1256
+    },
+    {
+      "epoch": 0.30532212885154064,
+      "grad_norm": 11.541386948579666,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 1257
+    },
+    {
+      "epoch": 0.3081232492997199,
+      "grad_norm": 7.6907940788830595,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 1258
+    },
+    {
+      "epoch": 0.31092436974789917,
+      "grad_norm": 9.524011587765084,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 1259
+    },
+    {
+      "epoch": 0.3137254901960784,
+      "grad_norm": 10.307414763530717,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 1260
+    },
+    {
+      "epoch": 0.3165266106442577,
+      "grad_norm": 8.29354573708552,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 1261
+    },
+    {
+      "epoch": 0.31932773109243695,
+      "grad_norm": 11.517166437669927,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 1262
+    },
+    {
+      "epoch": 0.32212885154061627,
+      "grad_norm": 9.10117397430712,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 1263
+    },
+    {
+      "epoch": 0.32492997198879553,
+      "grad_norm": 9.91347063126054,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 1264
+    },
+    {
+      "epoch": 0.3277310924369748,
+      "grad_norm": 11.885892942454877,
+      "learning_rate": 1e-06,
+      "loss": 0.1835,
+      "step": 1265
+    },
+    {
+      "epoch": 0.33053221288515405,
+      "grad_norm": 8.702622889096967,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 1266
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 7.705755827988987,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 1267
+    },
+    {
+      "epoch": 0.33613445378151263,
+      "grad_norm": 11.902627942570662,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 1268
+    },
+    {
+      "epoch": 0.3389355742296919,
+      "grad_norm": 12.393608290628748,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 1269
+    },
+    {
+      "epoch": 0.34173669467787116,
+      "grad_norm": 10.523490598809861,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 1270
+    },
+    {
+      "epoch": 0.3445378151260504,
+      "grad_norm": 9.406930690995988,
+      "learning_rate": 1e-06,
+      "loss": 0.1782,
+      "step": 1271
+    },
+    {
+      "epoch": 0.3473389355742297,
+      "grad_norm": 11.505963044716578,
+      "learning_rate": 1e-06,
+      "loss": 0.1897,
+      "step": 1272
+    },
+    {
+      "epoch": 0.35014005602240894,
+      "grad_norm": 7.864676713829887,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 1273
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 9.643045440011969,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 1274
+    },
+    {
+      "epoch": 0.3557422969187675,
+      "grad_norm": 11.715339465221017,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 1275
+    },
+    {
+      "epoch": 0.3585434173669468,
+      "grad_norm": 9.59943200941472,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 1276
+    },
+    {
+      "epoch": 0.36134453781512604,
+      "grad_norm": 7.5413152055481785,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 1277
+    },
+    {
+      "epoch": 0.3641456582633053,
+      "grad_norm": 10.856385541996532,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 1278
+    },
+    {
+      "epoch": 0.36694677871148457,
+      "grad_norm": 9.973287937949564,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 1279
+    },
+    {
+      "epoch": 0.3697478991596639,
+      "grad_norm": 12.167000195062533,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 1280
+    },
+    {
+      "epoch": 0.37254901960784315,
+      "grad_norm": 11.044577163216978,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 1281
+    },
+    {
+      "epoch": 0.3753501400560224,
+      "grad_norm": 10.719474309667152,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 1282
+    },
+    {
+      "epoch": 0.37815126050420167,
+      "grad_norm": 9.139163064008489,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 1283
+    },
+    {
+      "epoch": 0.38095238095238093,
+      "grad_norm": 11.663563203461097,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 1284
+    },
+    {
+      "epoch": 0.38375350140056025,
+      "grad_norm": 9.919655736726217,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 1285
+    },
+    {
+      "epoch": 0.3865546218487395,
+      "grad_norm": 10.409220335428566,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 1286
+    },
+    {
+      "epoch": 0.38935574229691877,
+      "grad_norm": 10.390813702468835,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 1287
+    },
+    {
+      "epoch": 0.39215686274509803,
+      "grad_norm": 10.454431457940181,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 1288
+    },
+    {
+      "epoch": 0.3949579831932773,
+      "grad_norm": 10.087724580822297,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 1289
+    },
+    {
+      "epoch": 0.39775910364145656,
+      "grad_norm": 12.625101567584368,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 1290
+    },
+    {
+      "epoch": 0.4005602240896359,
+      "grad_norm": 7.4922805526211,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 1291
+    },
+    {
+      "epoch": 0.40336134453781514,
+      "grad_norm": 7.71536830507057,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 1292
+    },
+    {
+      "epoch": 0.4061624649859944,
+      "grad_norm": 6.681096500424577,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 1293
+    },
+    {
+      "epoch": 0.40896358543417366,
+      "grad_norm": 9.173077551237682,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 1294
+    },
+    {
+      "epoch": 0.4117647058823529,
+      "grad_norm": 11.406555917247253,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 1295
+    },
+    {
+      "epoch": 0.41456582633053224,
+      "grad_norm": 10.926818306140754,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 1296
+    },
+    {
+      "epoch": 0.4173669467787115,
+      "grad_norm": 10.880649636505444,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 1297
+    },
+    {
+      "epoch": 0.42016806722689076,
+      "grad_norm": 8.070906696923357,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 1298
+    },
+    {
+      "epoch": 0.42296918767507,
+      "grad_norm": 7.059656182704983,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 1299
+    },
+    {
+      "epoch": 0.4257703081232493,
+      "grad_norm": 9.025143207900616,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 1300
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 10.082480463272509,
+      "learning_rate": 1e-06,
+      "loss": 0.1827,
+      "step": 1301
+    },
+    {
+      "epoch": 0.43137254901960786,
+      "grad_norm": 10.192984919043827,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 1302
+    },
+    {
+      "epoch": 0.4341736694677871,
+      "grad_norm": 7.492540863614779,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 1303
+    },
+    {
+      "epoch": 0.4369747899159664,
+      "grad_norm": 10.097825546356889,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 1304
+    },
+    {
+      "epoch": 0.43977591036414565,
+      "grad_norm": 9.716251480213224,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 1305
+    },
+    {
+      "epoch": 0.4425770308123249,
+      "grad_norm": 13.826859564366579,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1306
+    },
+    {
+      "epoch": 0.44537815126050423,
+      "grad_norm": 10.742462783797462,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 1307
+    },
+    {
+      "epoch": 0.4481792717086835,
+      "grad_norm": 8.809987730648151,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 1308
+    },
+    {
+      "epoch": 0.45098039215686275,
+      "grad_norm": 10.073902654445165,
+      "learning_rate": 1e-06,
+      "loss": 0.1982,
+      "step": 1309
+    },
+    {
+      "epoch": 0.453781512605042,
+      "grad_norm": 10.86587387400025,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 1310
+    },
+    {
+      "epoch": 0.4565826330532213,
+      "grad_norm": 9.761184761674917,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 1311
+    },
+    {
+      "epoch": 0.45938375350140054,
+      "grad_norm": 6.566720251816652,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 1312
+    },
+    {
+      "epoch": 0.46218487394957986,
+      "grad_norm": 8.225222147108173,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 1313
+    },
+    {
+      "epoch": 0.4649859943977591,
+      "grad_norm": 10.156090632790749,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 1314
+    },
+    {
+      "epoch": 0.4677871148459384,
+      "grad_norm": 9.682891648866095,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 1315
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 6.996790554083489,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 1316
+    },
+    {
+      "epoch": 0.4733893557422969,
+      "grad_norm": 9.395826252151286,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 1317
+    },
+    {
+      "epoch": 0.47619047619047616,
+      "grad_norm": 11.436851751824143,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 1318
+    },
+    {
+      "epoch": 0.4789915966386555,
+      "grad_norm": 8.773748719588726,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 1319
+    },
+    {
+      "epoch": 0.48179271708683474,
+      "grad_norm": 11.400579213700338,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 1320
+    },
+    {
+      "epoch": 0.484593837535014,
+      "grad_norm": 12.049281885302314,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 1321
+    },
+    {
+      "epoch": 0.48739495798319327,
+      "grad_norm": 5.1406097570859295,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 1322
+    },
+    {
+      "epoch": 0.49019607843137253,
+      "grad_norm": 8.063702354957726,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 1323
+    },
+    {
+      "epoch": 0.49299719887955185,
+      "grad_norm": 8.803853514926942,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 1324
+    },
+    {
+      "epoch": 0.4957983193277311,
+      "grad_norm": 7.382641518611728,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 1325
+    },
+    {
+      "epoch": 0.49859943977591037,
+      "grad_norm": 8.849369802654548,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5014005602240896,
+      "grad_norm": 10.914535806530674,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5042016806722689,
+      "grad_norm": 13.876541434192127,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5070028011204482,
+      "grad_norm": 9.27549832445796,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5098039215686274,
+      "grad_norm": 8.991900460500112,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5126050420168067,
+      "grad_norm": 9.547642626919279,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5154061624649859,
+      "grad_norm": 9.319703948635471,
+      "learning_rate": 1e-06,
+      "loss": 0.1883,
+      "step": 1332
+    },
+    {
+      "epoch": 0.5182072829131653,
+      "grad_norm": 10.160999275839622,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5210084033613446,
+      "grad_norm": 8.235417346031172,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5238095238095238,
+      "grad_norm": 10.596109213804045,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5266106442577031,
+      "grad_norm": 8.534242979267981,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5294117647058824,
+      "grad_norm": 10.863348986663283,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5322128851540616,
+      "grad_norm": 10.856237134336236,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5350140056022409,
+      "grad_norm": 11.413003008935645,
+      "learning_rate": 1e-06,
+      "loss": 0.1801,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5378151260504201,
+      "grad_norm": 7.9048070121741985,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5406162464985994,
+      "grad_norm": 8.96740372956126,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5434173669467787,
+      "grad_norm": 9.940579654137812,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5462184873949579,
+      "grad_norm": 9.236461012467226,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5490196078431373,
+      "grad_norm": 7.747196555777026,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5518207282913166,
+      "grad_norm": 9.219978511609519,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5546218487394958,
+      "grad_norm": 7.448088511261046,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5574229691876751,
+      "grad_norm": 8.659971328930354,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5602240896358543,
+      "grad_norm": 9.391143415609275,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5630252100840336,
+      "grad_norm": 9.800096543014094,
+      "learning_rate": 1e-06,
+      "loss": 0.1879,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5658263305322129,
+      "grad_norm": 10.132256459310646,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5686274509803921,
+      "grad_norm": 7.8994457812149355,
+      "learning_rate": 1e-06,
+      "loss": 0.1816,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 7.56861976289431,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5742296918767507,
+      "grad_norm": 10.912545981511618,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5770308123249299,
+      "grad_norm": 9.385321217796621,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5798319327731093,
+      "grad_norm": 10.296820300405383,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5826330532212886,
+      "grad_norm": 9.200038619570199,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5854341736694678,
+      "grad_norm": 11.936907225720747,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 9.424881480969065,
+      "learning_rate": 1e-06,
+      "loss": 0.1827,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5910364145658263,
+      "grad_norm": 10.096131295624213,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5938375350140056,
+      "grad_norm": 8.13067612775405,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5966386554621849,
+      "grad_norm": 9.247425073950554,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5994397759103641,
+      "grad_norm": 8.03734416063969,
+      "learning_rate": 1e-06,
+      "loss": 0.1834,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6022408963585434,
+      "grad_norm": 9.933033638081115,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6050420168067226,
+      "grad_norm": 9.565289279127278,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6078431372549019,
+      "grad_norm": 10.403384180843334,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 1365
+    },
+    {
+      "epoch": 0.6106442577030813,
+      "grad_norm": 12.001138530115181,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6134453781512605,
+      "grad_norm": 9.516055078116592,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6162464985994398,
+      "grad_norm": 7.957835074521483,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6190476190476191,
+      "grad_norm": 8.562261842114987,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6218487394957983,
+      "grad_norm": 8.46959600567353,
+      "learning_rate": 1e-06,
+      "loss": 0.1957,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6246498599439776,
+      "grad_norm": 10.903459337042605,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6274509803921569,
+      "grad_norm": 9.023962675330557,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6302521008403361,
+      "grad_norm": 12.348305808745838,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6330532212885154,
+      "grad_norm": 12.052566733389801,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6358543417366946,
+      "grad_norm": 10.058572156637588,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6386554621848739,
+      "grad_norm": 9.555427393713835,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6414565826330533,
+      "grad_norm": 7.343220355205719,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6442577030812325,
+      "grad_norm": 9.361105028610387,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6470588235294118,
+      "grad_norm": 8.399025892976486,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6498599439775911,
+      "grad_norm": 11.135499118412547,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 1380
+    },
+    {
+      "epoch": 0.6526610644257703,
+      "grad_norm": 7.572167164005233,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6554621848739496,
+      "grad_norm": 9.175133632804854,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6582633053221288,
+      "grad_norm": 9.189669285308497,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6610644257703081,
+      "grad_norm": 9.151199783204154,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6638655462184874,
+      "grad_norm": 6.680197600147881,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 7.207210738010096,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6694677871148459,
+      "grad_norm": 8.005942093314129,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6722689075630253,
+      "grad_norm": 8.151078512944926,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6750700280112045,
+      "grad_norm": 11.749826785217977,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6778711484593838,
+      "grad_norm": 8.761747787485293,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 1390
+    },
+    {
+      "epoch": 0.680672268907563,
+      "grad_norm": 9.082207806099063,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6834733893557423,
+      "grad_norm": 7.560779933775972,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6862745098039216,
+      "grad_norm": 12.502296493601479,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6890756302521008,
+      "grad_norm": 8.995959050420522,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 1394
+    },
+    {
+      "epoch": 0.6918767507002801,
+      "grad_norm": 8.829498124174947,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6946778711484594,
+      "grad_norm": 9.995766597694084,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6974789915966386,
+      "grad_norm": 10.604886175171405,
+      "learning_rate": 1e-06,
+      "loss": 0.1654,
+      "step": 1397
+    },
+    {
+      "epoch": 0.7002801120448179,
+      "grad_norm": 8.541107739045167,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 1398
+    },
+    {
+      "epoch": 0.7030812324929971,
+      "grad_norm": 10.079994767288214,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 1399
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 8.98296828367011,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1400
+    },
+    {
+      "epoch": 0.7086834733893558,
+      "grad_norm": 11.727331501973289,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 1401
+    },
+    {
+      "epoch": 0.711484593837535,
+      "grad_norm": 7.954698385793363,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 1402
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 9.783660677608243,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 1403
+    },
+    {
+      "epoch": 0.7170868347338936,
+      "grad_norm": 9.681952530853435,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 1404
+    },
+    {
+      "epoch": 0.7198879551820728,
+      "grad_norm": 8.496957964612577,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 1405
+    },
+    {
+      "epoch": 0.7226890756302521,
+      "grad_norm": 10.819138155483541,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 1406
+    },
+    {
+      "epoch": 0.7254901960784313,
+      "grad_norm": 9.165952724496606,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1407
+    },
+    {
+      "epoch": 0.7282913165266106,
+      "grad_norm": 9.146582860320208,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 1408
+    },
+    {
+      "epoch": 0.7310924369747899,
+      "grad_norm": 9.666088499996757,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 1409
+    },
+    {
+      "epoch": 0.7338935574229691,
+      "grad_norm": 8.182283385511294,
+      "learning_rate": 1e-06,
+      "loss": 0.1734,
+      "step": 1410
+    },
+    {
+      "epoch": 0.7366946778711485,
+      "grad_norm": 9.108240160518308,
+      "learning_rate": 1e-06,
+      "loss": 0.1722,
+      "step": 1411
+    },
+    {
+      "epoch": 0.7394957983193278,
+      "grad_norm": 10.363901945870039,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 1412
+    },
+    {
+      "epoch": 0.742296918767507,
+      "grad_norm": 9.299560145561147,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 1413
+    },
+    {
+      "epoch": 0.7450980392156863,
+      "grad_norm": 9.356603699643909,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 1414
+    },
+    {
+      "epoch": 0.7478991596638656,
+      "grad_norm": 13.28934838680159,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 1415
+    },
+    {
+      "epoch": 0.7507002801120448,
+      "grad_norm": 8.427185963538218,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 1416
+    },
+    {
+      "epoch": 0.7535014005602241,
+      "grad_norm": 11.481531605516288,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 1417
+    },
+    {
+      "epoch": 0.7563025210084033,
+      "grad_norm": 9.670696827920953,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 1418
+    },
+    {
+      "epoch": 0.7591036414565826,
+      "grad_norm": 8.45423788455618,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 1419
+    },
+    {
+      "epoch": 0.7619047619047619,
+      "grad_norm": 9.47478340602613,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 1420
+    },
+    {
+      "epoch": 0.7647058823529411,
+      "grad_norm": 11.638954236891609,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 1421
+    },
+    {
+      "epoch": 0.7675070028011205,
+      "grad_norm": 8.136882305747713,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 1422
+    },
+    {
+      "epoch": 0.7703081232492998,
+      "grad_norm": 8.54505590204305,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 1423
+    },
+    {
+      "epoch": 0.773109243697479,
+      "grad_norm": 12.808676945707575,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 1424
+    },
+    {
+      "epoch": 0.7759103641456583,
+      "grad_norm": 9.160923361976266,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 1425
+    },
+    {
+      "epoch": 0.7787114845938375,
+      "grad_norm": 9.144728525930077,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 1426
+    },
+    {
+      "epoch": 0.7815126050420168,
+      "grad_norm": 11.746997527184124,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 1427
+    },
+    {
+      "epoch": 0.7843137254901961,
+      "grad_norm": 11.71467006642207,
+      "learning_rate": 1e-06,
+      "loss": 0.1782,
+      "step": 1428
+    },
+    {
+      "epoch": 0.7871148459383753,
+      "grad_norm": 10.555854990337343,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 1429
+    },
+    {
+      "epoch": 0.7899159663865546,
+      "grad_norm": 12.180941690794542,
+      "learning_rate": 1e-06,
+      "loss": 0.196,
+      "step": 1430
+    },
+    {
+      "epoch": 0.7927170868347339,
+      "grad_norm": 10.5324074074305,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 1431
+    },
+    {
+      "epoch": 0.7955182072829131,
+      "grad_norm": 11.106710867036968,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 1432
+    },
+    {
+      "epoch": 0.7983193277310925,
+      "grad_norm": 10.547845445343231,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 1433
+    },
+    {
+      "epoch": 0.8011204481792717,
+      "grad_norm": 8.082773136036357,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 1434
+    },
+    {
+      "epoch": 0.803921568627451,
+      "grad_norm": 8.259175243939062,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 1435
+    },
+    {
+      "epoch": 0.8067226890756303,
+      "grad_norm": 9.010079122559867,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 1436
+    },
+    {
+      "epoch": 0.8095238095238095,
+      "grad_norm": 8.636982439942415,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 1437
+    },
+    {
+      "epoch": 0.8123249299719888,
+      "grad_norm": 9.735552761450867,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 1438
+    },
+    {
+      "epoch": 0.8151260504201681,
+      "grad_norm": 10.142650373181983,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 1439
+    },
+    {
+      "epoch": 0.8179271708683473,
+      "grad_norm": 10.706758614155223,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 1440
+    },
+    {
+      "epoch": 0.8207282913165266,
+      "grad_norm": 9.504037030926717,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 1441
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 8.911868245072814,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 1442
+    },
+    {
+      "epoch": 0.8263305322128851,
+      "grad_norm": 8.146447896046888,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 1443
+    },
+    {
+      "epoch": 0.8291316526610645,
+      "grad_norm": 13.019427882259729,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 1444
+    },
+    {
+      "epoch": 0.8319327731092437,
+      "grad_norm": 9.98323321119452,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 1445
+    },
+    {
+      "epoch": 0.834733893557423,
+      "grad_norm": 8.4701563390894,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 1446
+    },
+    {
+      "epoch": 0.8375350140056023,
+      "grad_norm": 10.7946480092881,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 1447
+    },
+    {
+      "epoch": 0.8403361344537815,
+      "grad_norm": 11.089000093618122,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 1448
+    },
+    {
+      "epoch": 0.8431372549019608,
+      "grad_norm": 8.273726402355692,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1449
+    },
+    {
+      "epoch": 0.84593837535014,
+      "grad_norm": 9.05460300243619,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 1450
+    },
+    {
+      "epoch": 0.8487394957983193,
+      "grad_norm": 9.80781089853407,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 1451
+    },
+    {
+      "epoch": 0.8515406162464986,
+      "grad_norm": 9.8346536107,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 1452
+    },
+    {
+      "epoch": 0.8543417366946778,
+      "grad_norm": 11.593610235664139,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 1453
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 9.372153314938581,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 1454
+    },
+    {
+      "epoch": 0.8599439775910365,
+      "grad_norm": 12.46990272454077,
+      "learning_rate": 1e-06,
+      "loss": 0.1962,
+      "step": 1455
+    },
+    {
+      "epoch": 0.8627450980392157,
+      "grad_norm": 10.004037448400432,
+      "learning_rate": 1e-06,
+      "loss": 0.1962,
+      "step": 1456
+    },
+    {
+      "epoch": 0.865546218487395,
+      "grad_norm": 9.19628095113786,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 1457
+    },
+    {
+      "epoch": 0.8683473389355743,
+      "grad_norm": 8.746808749912597,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 1458
+    },
+    {
+      "epoch": 0.8711484593837535,
+      "grad_norm": 10.29229806143627,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 1459
+    },
+    {
+      "epoch": 0.8739495798319328,
+      "grad_norm": 9.157838353682079,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 1460
+    },
+    {
+      "epoch": 0.876750700280112,
+      "grad_norm": 8.988249780453927,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 1461
+    },
+    {
+      "epoch": 0.8795518207282913,
+      "grad_norm": 10.160538415076497,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 1462
+    },
+    {
+      "epoch": 0.8823529411764706,
+      "grad_norm": 12.691657039770476,
+      "learning_rate": 1e-06,
+      "loss": 0.1962,
+      "step": 1463
+    },
+    {
+      "epoch": 0.8851540616246498,
+      "grad_norm": 12.396139880571932,
+      "learning_rate": 1e-06,
+      "loss": 0.1824,
+      "step": 1464
+    },
+    {
+      "epoch": 0.8879551820728291,
+      "grad_norm": 8.566046030344355,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 1465
+    },
+    {
+      "epoch": 0.8907563025210085,
+      "grad_norm": 11.36843645257593,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 1466
+    },
+    {
+      "epoch": 0.8935574229691877,
+      "grad_norm": 9.658069165398137,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 1467
+    },
+    {
+      "epoch": 0.896358543417367,
+      "grad_norm": 10.759269790525579,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 1468
+    },
+    {
+      "epoch": 0.8991596638655462,
+      "grad_norm": 10.231442660789877,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 1469
+    },
+    {
+      "epoch": 0.9019607843137255,
+      "grad_norm": 7.9505542454572975,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 1470
+    },
+    {
+      "epoch": 0.9047619047619048,
+      "grad_norm": 10.345761258843453,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 1471
+    },
+    {
+      "epoch": 0.907563025210084,
+      "grad_norm": 10.571614066021924,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 1472
+    },
+    {
+      "epoch": 0.9103641456582633,
+      "grad_norm": 9.20912491154155,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 1473
+    },
+    {
+      "epoch": 0.9131652661064426,
+      "grad_norm": 9.36139509924034,
+      "learning_rate": 1e-06,
+      "loss": 0.1723,
+      "step": 1474
+    },
+    {
+      "epoch": 0.9159663865546218,
+      "grad_norm": 8.271535888481065,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 1475
+    },
+    {
+      "epoch": 0.9187675070028011,
+      "grad_norm": 10.824578263071254,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 1476
+    },
+    {
+      "epoch": 0.9215686274509803,
+      "grad_norm": 11.873525233220864,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 1477
+    },
+    {
+      "epoch": 0.9243697478991597,
+      "grad_norm": 11.940890133180272,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 1478
+    },
+    {
+      "epoch": 0.927170868347339,
+      "grad_norm": 7.498646715163616,
+      "learning_rate": 1e-06,
+      "loss": 0.183,
+      "step": 1479
+    },
+    {
+      "epoch": 0.9299719887955182,
+      "grad_norm": 10.363372991903871,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 1480
+    },
+    {
+      "epoch": 0.9327731092436975,
+      "grad_norm": 11.025314423899198,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 1481
+    },
+    {
+      "epoch": 0.9355742296918768,
+      "grad_norm": 9.43781014106129,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 1482
+    },
+    {
+      "epoch": 0.938375350140056,
+      "grad_norm": 13.783275690495165,
+      "learning_rate": 1e-06,
+      "loss": 0.1957,
+      "step": 1483
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 9.718427909860743,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 1484
+    },
+    {
+      "epoch": 0.9439775910364145,
+      "grad_norm": 10.719986729637165,
+      "learning_rate": 1e-06,
+      "loss": 0.196,
+      "step": 1485
+    },
+    {
+      "epoch": 0.9467787114845938,
+      "grad_norm": 8.301559716465986,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 1486
+    },
+    {
+      "epoch": 0.9495798319327731,
+      "grad_norm": 9.286080071735979,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 1487
+    },
+    {
+      "epoch": 0.9523809523809523,
+      "grad_norm": 14.65317961418066,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 1488
+    },
+    {
+      "epoch": 0.9551820728291317,
+      "grad_norm": 12.490645581169384,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 1489
+    },
+    {
+      "epoch": 0.957983193277311,
+      "grad_norm": 10.826348408023309,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 1490
+    },
+    {
+      "epoch": 0.9607843137254902,
+      "grad_norm": 8.644933960573793,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 1491
+    },
+    {
+      "epoch": 0.9635854341736695,
+      "grad_norm": 10.908368557285558,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 1492
+    },
+    {
+      "epoch": 0.9663865546218487,
+      "grad_norm": 10.50757074504513,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 1493
+    },
+    {
+      "epoch": 0.969187675070028,
+      "grad_norm": 10.654525909958128,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 1494
+    },
+    {
+      "epoch": 0.9719887955182073,
+      "grad_norm": 9.692798947258003,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 1495
+    },
+    {
+      "epoch": 0.9747899159663865,
+      "grad_norm": 9.693352389700616,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 1496
+    },
+    {
+      "epoch": 0.9775910364145658,
+      "grad_norm": 12.065611827547318,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 1497
+    },
+    {
+      "epoch": 0.9803921568627451,
+      "grad_norm": 8.354691040887468,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 1498
+    },
+    {
+      "epoch": 0.9831932773109243,
+      "grad_norm": 9.26741214569398,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 1499
+    },
+    {
+      "epoch": 0.9859943977591037,
+      "grad_norm": 13.540228990308819,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 1500
+    },
+    {
+      "epoch": 0.988795518207283,
+      "grad_norm": 7.934504539466558,
+      "learning_rate": 1e-06,
+      "loss": 0.1848,
+      "step": 1501
+    },
+    {
+      "epoch": 0.9915966386554622,
+      "grad_norm": 9.35194676724591,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 1502
+    },
+    {
+      "epoch": 0.9943977591036415,
+      "grad_norm": 13.673825053338437,
+      "learning_rate": 1e-06,
+      "loss": 0.1963,
+      "step": 1503
+    },
+    {
+      "epoch": 0.9971988795518207,
+      "grad_norm": 8.142949579650573,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 1504
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 11.170031274751139,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 1505
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.242586374282837,
+      "eval_runtime": 22.0129,
+      "eval_samples_per_second": 45.428,
+      "eval_steps_per_second": 5.678,
+      "step": 1505
+    },
+    {
+      "epoch": 0.0027359781121751026,
+      "grad_norm": 10.366737693233963,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 1506
+    },
+    {
+      "epoch": 0.005471956224350205,
+      "grad_norm": 10.330871489537895,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 1507
+    },
+    {
+      "epoch": 0.008207934336525308,
+      "grad_norm": 8.94952664873047,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 1508
+    },
+    {
+      "epoch": 0.01094391244870041,
+      "grad_norm": 7.137243075800746,
+      "learning_rate": 1e-06,
+      "loss": 0.1603,
+      "step": 1509
+    },
+    {
+      "epoch": 0.013679890560875513,
+      "grad_norm": 9.57708389717625,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 1510
+    },
+    {
+      "epoch": 0.016415868673050615,
+      "grad_norm": 9.158026355239226,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 1511
+    },
+    {
+      "epoch": 0.019151846785225718,
+      "grad_norm": 8.850569580774714,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 1512
+    },
+    {
+      "epoch": 0.02188782489740082,
+      "grad_norm": 10.404920613059081,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 1513
+    },
+    {
+      "epoch": 0.024623803009575923,
+      "grad_norm": 7.299191275550601,
+      "learning_rate": 1e-06,
+      "loss": 0.1615,
+      "step": 1514
+    },
+    {
+      "epoch": 0.027359781121751026,
+      "grad_norm": 8.509156307037234,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 1515
+    },
+    {
+      "epoch": 0.030095759233926128,
+      "grad_norm": 11.329524201868521,
+      "learning_rate": 1e-06,
+      "loss": 0.1707,
+      "step": 1516
+    },
+    {
+      "epoch": 0.03283173734610123,
+      "grad_norm": 10.329504892950668,
+      "learning_rate": 1e-06,
+      "loss": 0.1623,
+      "step": 1517
+    },
+    {
+      "epoch": 0.03556771545827633,
+      "grad_norm": 10.29956600108497,
+      "learning_rate": 1e-06,
+      "loss": 0.1456,
+      "step": 1518
+    },
+    {
+      "epoch": 0.038303693570451436,
+      "grad_norm": 11.898665792483342,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 1519
+    },
+    {
+      "epoch": 0.04103967168262654,
+      "grad_norm": 10.350961958496297,
+      "learning_rate": 1e-06,
+      "loss": 0.1668,
+      "step": 1520
+    },
+    {
+      "epoch": 0.04377564979480164,
+      "grad_norm": 10.918051251185458,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 1521
+    },
+    {
+      "epoch": 0.046511627906976744,
+      "grad_norm": 9.800239063884625,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 1522
+    },
+    {
+      "epoch": 0.049247606019151846,
+      "grad_norm": 15.263782657461059,
+      "learning_rate": 1e-06,
+      "loss": 0.159,
+      "step": 1523
+    },
+    {
+      "epoch": 0.05198358413132695,
+      "grad_norm": 10.969455417476626,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 1524
+    },
+    {
+      "epoch": 0.05471956224350205,
+      "grad_norm": 13.843329899828792,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 1525
+    },
+    {
+      "epoch": 0.057455540355677154,
+      "grad_norm": 12.740463027778386,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 1526
+    },
+    {
+      "epoch": 0.060191518467852256,
+      "grad_norm": 7.805488608659641,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 1527
+    },
+    {
+      "epoch": 0.06292749658002736,
+      "grad_norm": 9.589720572070934,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 1528
+    },
+    {
+      "epoch": 0.06566347469220246,
+      "grad_norm": 12.061761635364913,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 1529
+    },
+    {
+      "epoch": 0.06839945280437756,
+      "grad_norm": 9.580392857605956,
+      "learning_rate": 1e-06,
+      "loss": 0.1719,
+      "step": 1530
+    },
+    {
+      "epoch": 0.07113543091655267,
+      "grad_norm": 9.459348329983229,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 1531
+    },
+    {
+      "epoch": 0.07387140902872777,
+      "grad_norm": 11.684013004289186,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 1532
+    },
+    {
+      "epoch": 0.07660738714090287,
+      "grad_norm": 9.17173793225653,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 1533
+    },
+    {
+      "epoch": 0.07934336525307797,
+      "grad_norm": 9.813912134265278,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 1534
+    },
+    {
+      "epoch": 0.08207934336525308,
+      "grad_norm": 11.42646539342435,
+      "learning_rate": 1e-06,
+      "loss": 0.1707,
+      "step": 1535
+    },
+    {
+      "epoch": 0.08481532147742818,
+      "grad_norm": 9.147119489658197,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 1536
+    },
+    {
+      "epoch": 0.08755129958960328,
+      "grad_norm": 11.628029575805497,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 1537
+    },
+    {
+      "epoch": 0.09028727770177838,
+      "grad_norm": 10.386946426577603,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 1538
+    },
+    {
+      "epoch": 0.09302325581395349,
+      "grad_norm": 9.81232688368699,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 1539
+    },
+    {
+      "epoch": 0.09575923392612859,
+      "grad_norm": 9.362206885472192,
+      "learning_rate": 1e-06,
+      "loss": 0.1624,
+      "step": 1540
+    },
+    {
+      "epoch": 0.09849521203830369,
+      "grad_norm": 10.440559304112242,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 1541
+    },
+    {
+      "epoch": 0.1012311901504788,
+      "grad_norm": 11.820535819633337,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 1542
+    },
+    {
+      "epoch": 0.1039671682626539,
+      "grad_norm": 12.300618194716149,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 1543
+    },
+    {
+      "epoch": 0.106703146374829,
+      "grad_norm": 12.065049461477718,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 1544
+    },
+    {
+      "epoch": 0.1094391244870041,
+      "grad_norm": 10.703584667066806,
+      "learning_rate": 1e-06,
+      "loss": 0.1505,
+      "step": 1545
+    },
+    {
+      "epoch": 0.1121751025991792,
+      "grad_norm": 10.496273779971082,
+      "learning_rate": 1e-06,
+      "loss": 0.1646,
+      "step": 1546
+    },
+    {
+      "epoch": 0.11491108071135431,
+      "grad_norm": 10.512529036739117,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 1547
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 9.786409059576314,
+      "learning_rate": 1e-06,
+      "loss": 0.1483,
+      "step": 1548
+    },
+    {
+      "epoch": 0.12038303693570451,
+      "grad_norm": 9.221730932455696,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 1549
+    },
+    {
+      "epoch": 0.12311901504787962,
+      "grad_norm": 11.911847757988607,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 1550
+    },
+    {
+      "epoch": 0.12585499316005472,
+      "grad_norm": 9.121851227519182,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 1551
+    },
+    {
+      "epoch": 0.12859097127222982,
+      "grad_norm": 10.514024958338982,
+      "learning_rate": 1e-06,
+      "loss": 0.1648,
+      "step": 1552
+    },
+    {
+      "epoch": 0.13132694938440492,
+      "grad_norm": 11.047751495136929,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 1553
+    },
+    {
+      "epoch": 0.13406292749658003,
+      "grad_norm": 8.493893473986125,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 1554
+    },
+    {
+      "epoch": 0.13679890560875513,
+      "grad_norm": 11.385667084581913,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 1555
+    },
+    {
+      "epoch": 0.13953488372093023,
+      "grad_norm": 14.232771150726224,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 1556
+    },
+    {
+      "epoch": 0.14227086183310533,
+      "grad_norm": 10.466292662137922,
+      "learning_rate": 1e-06,
+      "loss": 0.1474,
+      "step": 1557
+    },
+    {
+      "epoch": 0.14500683994528044,
+      "grad_norm": 8.721557255510174,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 1558
+    },
+    {
+      "epoch": 0.14774281805745554,
+      "grad_norm": 10.099566954293293,
+      "learning_rate": 1e-06,
+      "loss": 0.1534,
+      "step": 1559
+    },
+    {
+      "epoch": 0.15047879616963064,
+      "grad_norm": 10.879872451109524,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 1560
+    },
+    {
+      "epoch": 0.15321477428180574,
+      "grad_norm": 10.299667893135002,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 1561
+    },
+    {
+      "epoch": 0.15595075239398085,
+      "grad_norm": 10.128067850825355,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 1562
+    },
+    {
+      "epoch": 0.15868673050615595,
+      "grad_norm": 7.935592167828565,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 1563
+    },
+    {
+      "epoch": 0.16142270861833105,
+      "grad_norm": 9.187205252407027,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 1564
+    },
+    {
+      "epoch": 0.16415868673050615,
+      "grad_norm": 11.570875863209432,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 1565
+    },
+    {
+      "epoch": 0.16689466484268126,
+      "grad_norm": 9.90710301605386,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 1566
+    },
+    {
+      "epoch": 0.16963064295485636,
+      "grad_norm": 13.147323799384994,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 1567
+    },
+    {
+      "epoch": 0.17236662106703146,
+      "grad_norm": 8.787002742584137,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 1568
+    },
+    {
+      "epoch": 0.17510259917920656,
+      "grad_norm": 10.554266751305164,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 1569
+    },
+    {
+      "epoch": 0.17783857729138167,
+      "grad_norm": 11.203096796680107,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 1570
+    },
+    {
+      "epoch": 0.18057455540355677,
+      "grad_norm": 12.373179850036879,
+      "learning_rate": 1e-06,
+      "loss": 0.1624,
+      "step": 1571
+    },
+    {
+      "epoch": 0.18331053351573187,
+      "grad_norm": 13.156843958292411,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 1572
+    },
+    {
+      "epoch": 0.18604651162790697,
+      "grad_norm": 10.744812348857781,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 1573
+    },
+    {
+      "epoch": 0.18878248974008208,
+      "grad_norm": 10.131377344622514,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 1574
+    },
+    {
+      "epoch": 0.19151846785225718,
+      "grad_norm": 8.919834904051477,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 1575
+    },
+    {
+      "epoch": 0.19425444596443228,
+      "grad_norm": 7.387661650997607,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 1576
+    },
+    {
+      "epoch": 0.19699042407660738,
+      "grad_norm": 11.399976181868384,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 1577
+    },
+    {
+      "epoch": 0.1997264021887825,
+      "grad_norm": 12.052056510204471,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 1578
+    },
+    {
+      "epoch": 0.2024623803009576,
+      "grad_norm": 8.899070425911898,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 1579
+    },
+    {
+      "epoch": 0.2051983584131327,
+      "grad_norm": 10.401362543988787,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1580
+    },
+    {
+      "epoch": 0.2079343365253078,
+      "grad_norm": 10.177835371029335,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 1581
+    },
+    {
+      "epoch": 0.2106703146374829,
+      "grad_norm": 10.12372996929794,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 1582
+    },
+    {
+      "epoch": 0.213406292749658,
+      "grad_norm": 10.174025908132423,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 1583
+    },
+    {
+      "epoch": 0.2161422708618331,
+      "grad_norm": 11.397931583748623,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 1584
+    },
+    {
+      "epoch": 0.2188782489740082,
+      "grad_norm": 9.186725965210126,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 1585
+    },
+    {
+      "epoch": 0.2216142270861833,
+      "grad_norm": 7.629737710726509,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 1586
+    },
+    {
+      "epoch": 0.2243502051983584,
+      "grad_norm": 8.672412000951738,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 1587
+    },
+    {
+      "epoch": 0.2270861833105335,
+      "grad_norm": 11.70443662669903,
+      "learning_rate": 1e-06,
+      "loss": 0.1703,
+      "step": 1588
+    },
+    {
+      "epoch": 0.22982216142270862,
+      "grad_norm": 12.51916672055767,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 1589
+    },
+    {
+      "epoch": 0.23255813953488372,
+      "grad_norm": 10.325621016321735,
+      "learning_rate": 1e-06,
+      "loss": 0.1591,
+      "step": 1590
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 14.05906776908692,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 1591
+    },
+    {
+      "epoch": 0.23803009575923392,
+      "grad_norm": 10.943232084141963,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 1592
+    },
+    {
+      "epoch": 0.24076607387140903,
+      "grad_norm": 10.314451192978634,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 1593
+    },
+    {
+      "epoch": 0.24350205198358413,
+      "grad_norm": 9.41641363627807,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 1594
+    },
+    {
+      "epoch": 0.24623803009575923,
+      "grad_norm": 11.646436212428252,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 1595
+    },
+    {
+      "epoch": 0.24897400820793433,
+      "grad_norm": 9.432070667904416,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 1596
+    },
+    {
+      "epoch": 0.25170998632010944,
+      "grad_norm": 8.149933651730354,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 1597
+    },
+    {
+      "epoch": 0.25444596443228457,
+      "grad_norm": 9.349074413325274,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 1598
+    },
+    {
+      "epoch": 0.25718194254445964,
+      "grad_norm": 10.740337984303665,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 1599
+    },
+    {
+      "epoch": 0.25991792065663477,
+      "grad_norm": 8.35717701670238,
+      "learning_rate": 1e-06,
+      "loss": 0.1606,
+      "step": 1600
+    },
+    {
+      "epoch": 0.26265389876880985,
+      "grad_norm": 12.496195737239443,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 1601
+    },
+    {
+      "epoch": 0.265389876880985,
+      "grad_norm": 10.923607099512163,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 1602
+    },
+    {
+      "epoch": 0.26812585499316005,
+      "grad_norm": 11.51043612070502,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 1603
+    },
+    {
+      "epoch": 0.2708618331053352,
+      "grad_norm": 7.719861190371166,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 1604
+    },
+    {
+      "epoch": 0.27359781121751026,
+      "grad_norm": 10.891532919601666,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1605
+    },
+    {
+      "epoch": 0.2763337893296854,
+      "grad_norm": 10.39120313823709,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 1606
+    },
+    {
+      "epoch": 0.27906976744186046,
+      "grad_norm": 10.369590277069296,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 1607
+    },
+    {
+      "epoch": 0.2818057455540356,
+      "grad_norm": 8.313590882148375,
+      "learning_rate": 1e-06,
+      "loss": 0.1795,
+      "step": 1608
+    },
+    {
+      "epoch": 0.28454172366621067,
+      "grad_norm": 10.252215166528,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 1609
+    },
+    {
+      "epoch": 0.2872777017783858,
+      "grad_norm": 11.944289031432547,
+      "learning_rate": 1e-06,
+      "loss": 0.1567,
+      "step": 1610
+    },
+    {
+      "epoch": 0.29001367989056087,
+      "grad_norm": 10.38080102887538,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 1611
+    },
+    {
+      "epoch": 0.292749658002736,
+      "grad_norm": 10.85758270273052,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 1612
+    },
+    {
+      "epoch": 0.2954856361149111,
+      "grad_norm": 7.9394666731755175,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 1613
+    },
+    {
+      "epoch": 0.2982216142270862,
+      "grad_norm": 9.520268174478655,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 1614
+    },
+    {
+      "epoch": 0.3009575923392613,
+      "grad_norm": 8.370817370659585,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 1615
+    },
+    {
+      "epoch": 0.3036935704514364,
+      "grad_norm": 8.472543589519852,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 1616
+    },
+    {
+      "epoch": 0.3064295485636115,
+      "grad_norm": 12.089992368935137,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 1617
+    },
+    {
+      "epoch": 0.3091655266757866,
+      "grad_norm": 10.367569520128464,
+      "learning_rate": 1e-06,
+      "loss": 0.1707,
+      "step": 1618
+    },
+    {
+      "epoch": 0.3119015047879617,
+      "grad_norm": 7.891032513514995,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 1619
+    },
+    {
+      "epoch": 0.3146374829001368,
+      "grad_norm": 9.593078099349361,
+      "learning_rate": 1e-06,
+      "loss": 0.1759,
+      "step": 1620
+    },
+    {
+      "epoch": 0.3173734610123119,
+      "grad_norm": 11.256276745647934,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 1621
+    },
+    {
+      "epoch": 0.320109439124487,
+      "grad_norm": 6.826942680154672,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 1622
+    },
+    {
+      "epoch": 0.3228454172366621,
+      "grad_norm": 11.46506395309239,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 1623
+    },
+    {
+      "epoch": 0.32558139534883723,
+      "grad_norm": 9.131754358079395,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 1624
+    },
+    {
+      "epoch": 0.3283173734610123,
+      "grad_norm": 9.639892091385171,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 1625
+    },
+    {
+      "epoch": 0.33105335157318744,
+      "grad_norm": 10.896333061463555,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1626
+    },
+    {
+      "epoch": 0.3337893296853625,
+      "grad_norm": 9.166058299387775,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 1627
+    },
+    {
+      "epoch": 0.33652530779753764,
+      "grad_norm": 11.008321394322586,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 1628
+    },
+    {
+      "epoch": 0.3392612859097127,
+      "grad_norm": 10.31111452418774,
+      "learning_rate": 1e-06,
+      "loss": 0.163,
+      "step": 1629
+    },
+    {
+      "epoch": 0.34199726402188785,
+      "grad_norm": 9.957409253096897,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 1630
+    },
+    {
+      "epoch": 0.3447332421340629,
+      "grad_norm": 8.477408092678179,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 1631
+    },
+    {
+      "epoch": 0.34746922024623805,
+      "grad_norm": 10.146873803462844,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 1632
+    },
+    {
+      "epoch": 0.35020519835841313,
+      "grad_norm": 11.897255006195595,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 1633
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 11.698425672235238,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 1634
+    },
+    {
+      "epoch": 0.35567715458276333,
+      "grad_norm": 8.560942371226556,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 1635
+    },
+    {
+      "epoch": 0.35841313269493846,
+      "grad_norm": 10.30042607760991,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 1636
+    },
+    {
+      "epoch": 0.36114911080711354,
+      "grad_norm": 7.872964432546622,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 1637
+    },
+    {
+      "epoch": 0.36388508891928867,
+      "grad_norm": 10.989662345576109,
+      "learning_rate": 1e-06,
+      "loss": 0.1676,
+      "step": 1638
+    },
+    {
+      "epoch": 0.36662106703146374,
+      "grad_norm": 7.295757526658234,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 1639
+    },
+    {
+      "epoch": 0.3693570451436389,
+      "grad_norm": 10.005907278311108,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 1640
+    },
+    {
+      "epoch": 0.37209302325581395,
+      "grad_norm": 12.19315042409113,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 1641
+    },
+    {
+      "epoch": 0.3748290013679891,
+      "grad_norm": 9.023981073085347,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 1642
+    },
+    {
+      "epoch": 0.37756497948016415,
+      "grad_norm": 11.803970591807449,
+      "learning_rate": 1e-06,
+      "loss": 0.1668,
+      "step": 1643
+    },
+    {
+      "epoch": 0.3803009575923393,
+      "grad_norm": 10.137508962093193,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 1644
+    },
+    {
+      "epoch": 0.38303693570451436,
+      "grad_norm": 9.802027540110092,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 1645
+    },
+    {
+      "epoch": 0.3857729138166895,
+      "grad_norm": 8.7133760300891,
+      "learning_rate": 1e-06,
+      "loss": 0.1734,
+      "step": 1646
+    },
+    {
+      "epoch": 0.38850889192886456,
+      "grad_norm": 9.609305343506305,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 1647
+    },
+    {
+      "epoch": 0.3912448700410397,
+      "grad_norm": 12.849886517156829,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 1648
+    },
+    {
+      "epoch": 0.39398084815321477,
+      "grad_norm": 8.692079340650269,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 1649
+    },
+    {
+      "epoch": 0.3967168262653899,
+      "grad_norm": 9.942010958612183,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 1650
+    },
+    {
+      "epoch": 0.399452804377565,
+      "grad_norm": 9.523929672709796,
+      "learning_rate": 1e-06,
+      "loss": 0.1608,
+      "step": 1651
+    },
+    {
+      "epoch": 0.4021887824897401,
+      "grad_norm": 10.249884022288809,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 1652
+    },
+    {
+      "epoch": 0.4049247606019152,
+      "grad_norm": 8.689835169673778,
+      "learning_rate": 1e-06,
+      "loss": 0.1532,
+      "step": 1653
+    },
+    {
+      "epoch": 0.4076607387140903,
+      "grad_norm": 9.365288162375322,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 1654
+    },
+    {
+      "epoch": 0.4103967168262654,
+      "grad_norm": 8.803429599914427,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1655
+    },
+    {
+      "epoch": 0.4131326949384405,
+      "grad_norm": 10.867052255630242,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 1656
+    },
+    {
+      "epoch": 0.4158686730506156,
+      "grad_norm": 10.483827468584185,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 1657
+    },
+    {
+      "epoch": 0.4186046511627907,
+      "grad_norm": 11.529888990753973,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 1658
+    },
+    {
+      "epoch": 0.4213406292749658,
+      "grad_norm": 10.350280111205308,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 1659
+    },
+    {
+      "epoch": 0.4240766073871409,
+      "grad_norm": 9.424992350352397,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 1660
+    },
+    {
+      "epoch": 0.426812585499316,
+      "grad_norm": 11.958701708132914,
+      "learning_rate": 1e-06,
+      "loss": 0.1723,
+      "step": 1661
+    },
+    {
+      "epoch": 0.42954856361149113,
+      "grad_norm": 11.984882181432061,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 1662
+    },
+    {
+      "epoch": 0.4322845417236662,
+      "grad_norm": 11.046727117209741,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 1663
+    },
+    {
+      "epoch": 0.43502051983584133,
+      "grad_norm": 8.191139380678822,
+      "learning_rate": 1e-06,
+      "loss": 0.1732,
+      "step": 1664
+    },
+    {
+      "epoch": 0.4377564979480164,
+      "grad_norm": 10.901288264385395,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 1665
+    },
+    {
+      "epoch": 0.44049247606019154,
+      "grad_norm": 12.055596002330224,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 1666
+    },
+    {
+      "epoch": 0.4432284541723666,
+      "grad_norm": 11.527986432325676,
+      "learning_rate": 1e-06,
+      "loss": 0.1595,
+      "step": 1667
+    },
+    {
+      "epoch": 0.44596443228454175,
+      "grad_norm": 10.694394560790734,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 1668
+    },
+    {
+      "epoch": 0.4487004103967168,
+      "grad_norm": 7.40639389550052,
+      "learning_rate": 1e-06,
+      "loss": 0.1643,
+      "step": 1669
+    },
+    {
+      "epoch": 0.45143638850889195,
+      "grad_norm": 8.031180843326176,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1670
+    },
+    {
+      "epoch": 0.454172366621067,
+      "grad_norm": 11.83173625840976,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 1671
+    },
+    {
+      "epoch": 0.45690834473324216,
+      "grad_norm": 11.702363976913231,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 1672
+    },
+    {
+      "epoch": 0.45964432284541723,
+      "grad_norm": 11.566611988628617,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 1673
+    },
+    {
+      "epoch": 0.46238030095759236,
+      "grad_norm": 10.080648368161334,
+      "learning_rate": 1e-06,
+      "loss": 0.1719,
+      "step": 1674
+    },
+    {
+      "epoch": 0.46511627906976744,
+      "grad_norm": 11.765756737539391,
+      "learning_rate": 1e-06,
+      "loss": 0.1704,
+      "step": 1675
+    },
+    {
+      "epoch": 0.46785225718194257,
+      "grad_norm": 9.47280376693067,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 1676
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 9.964819612542978,
+      "learning_rate": 1e-06,
+      "loss": 0.1752,
+      "step": 1677
+    },
+    {
+      "epoch": 0.47332421340629277,
+      "grad_norm": 9.591121414149427,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 1678
+    },
+    {
+      "epoch": 0.47606019151846785,
+      "grad_norm": 10.996638070843199,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 1679
+    },
+    {
+      "epoch": 0.478796169630643,
+      "grad_norm": 12.820290144860758,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 1680
+    },
+    {
+      "epoch": 0.48153214774281805,
+      "grad_norm": 10.245055271569191,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 1681
+    },
+    {
+      "epoch": 0.4842681258549932,
+      "grad_norm": 9.375312384924273,
+      "learning_rate": 1e-06,
+      "loss": 0.1694,
+      "step": 1682
+    },
+    {
+      "epoch": 0.48700410396716826,
+      "grad_norm": 11.208573541849356,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 1683
+    },
+    {
+      "epoch": 0.4897400820793434,
+      "grad_norm": 13.855473414655018,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 1684
+    },
+    {
+      "epoch": 0.49247606019151846,
+      "grad_norm": 12.352547848360054,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 1685
+    },
+    {
+      "epoch": 0.4952120383036936,
+      "grad_norm": 9.500905535504298,
+      "learning_rate": 1e-06,
+      "loss": 0.1709,
+      "step": 1686
+    },
+    {
+      "epoch": 0.49794801641586867,
+      "grad_norm": 8.80433174674803,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 1687
+    },
+    {
+      "epoch": 0.5006839945280438,
+      "grad_norm": 8.8635465332481,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 1688
+    },
+    {
+      "epoch": 0.5034199726402189,
+      "grad_norm": 11.513102654387312,
+      "learning_rate": 1e-06,
+      "loss": 0.1673,
+      "step": 1689
+    },
+    {
+      "epoch": 0.506155950752394,
+      "grad_norm": 12.037319645816112,
+      "learning_rate": 1e-06,
+      "loss": 0.1739,
+      "step": 1690
+    },
+    {
+      "epoch": 0.5088919288645691,
+      "grad_norm": 9.371128010309537,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 1691
+    },
+    {
+      "epoch": 0.5116279069767442,
+      "grad_norm": 11.078896544540463,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 1692
+    },
+    {
+      "epoch": 0.5143638850889193,
+      "grad_norm": 12.081307908031919,
+      "learning_rate": 1e-06,
+      "loss": 0.1765,
+      "step": 1693
+    },
+    {
+      "epoch": 0.5170998632010944,
+      "grad_norm": 11.39086097866553,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 1694
+    },
+    {
+      "epoch": 0.5198358413132695,
+      "grad_norm": 10.275044392564325,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 1695
+    },
+    {
+      "epoch": 0.5225718194254446,
+      "grad_norm": 10.910880005515452,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 1696
+    },
+    {
+      "epoch": 0.5253077975376197,
+      "grad_norm": 7.440411295575803,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 1697
+    },
+    {
+      "epoch": 0.5280437756497948,
+      "grad_norm": 8.981169690768029,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 1698
+    },
+    {
+      "epoch": 0.53077975376197,
+      "grad_norm": 13.554759915258744,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 1699
+    },
+    {
+      "epoch": 0.533515731874145,
+      "grad_norm": 8.795579781471961,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 1700
+    },
+    {
+      "epoch": 0.5362517099863201,
+      "grad_norm": 10.830939636843864,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 1701
+    },
+    {
+      "epoch": 0.5389876880984952,
+      "grad_norm": 9.561181829367897,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 1702
+    },
+    {
+      "epoch": 0.5417236662106704,
+      "grad_norm": 10.986202782445725,
+      "learning_rate": 1e-06,
+      "loss": 0.183,
+      "step": 1703
+    },
+    {
+      "epoch": 0.5444596443228454,
+      "grad_norm": 9.33293062064203,
+      "learning_rate": 1e-06,
+      "loss": 0.1712,
+      "step": 1704
+    },
+    {
+      "epoch": 0.5471956224350205,
+      "grad_norm": 7.5715776746977275,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 1705
+    },
+    {
+      "epoch": 0.5499316005471956,
+      "grad_norm": 11.62151619900682,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 1706
+    },
+    {
+      "epoch": 0.5526675786593708,
+      "grad_norm": 9.869600428700238,
+      "learning_rate": 1e-06,
+      "loss": 0.1678,
+      "step": 1707
+    },
+    {
+      "epoch": 0.5554035567715458,
+      "grad_norm": 8.557712767704007,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 1708
+    },
+    {
+      "epoch": 0.5581395348837209,
+      "grad_norm": 10.6856244291534,
+      "learning_rate": 1e-06,
+      "loss": 0.1606,
+      "step": 1709
+    },
+    {
+      "epoch": 0.560875512995896,
+      "grad_norm": 11.278893547113046,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 1710
+    },
+    {
+      "epoch": 0.5636114911080712,
+      "grad_norm": 9.481631555853674,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 1711
+    },
+    {
+      "epoch": 0.5663474692202463,
+      "grad_norm": 10.340483494783378,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 1712
+    },
+    {
+      "epoch": 0.5690834473324213,
+      "grad_norm": 11.914767150150087,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 1713
+    },
+    {
+      "epoch": 0.5718194254445964,
+      "grad_norm": 11.426445278726096,
+      "learning_rate": 1e-06,
+      "loss": 0.1673,
+      "step": 1714
+    },
+    {
+      "epoch": 0.5745554035567716,
+      "grad_norm": 9.866319447822821,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 1715
+    },
+    {
+      "epoch": 0.5772913816689467,
+      "grad_norm": 9.455005976041749,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 1716
+    },
+    {
+      "epoch": 0.5800273597811217,
+      "grad_norm": 10.523012590319286,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 1717
+    },
+    {
+      "epoch": 0.5827633378932968,
+      "grad_norm": 9.921873970693436,
+      "learning_rate": 1e-06,
+      "loss": 0.1694,
+      "step": 1718
+    },
+    {
+      "epoch": 0.585499316005472,
+      "grad_norm": 11.068991984970737,
+      "learning_rate": 1e-06,
+      "loss": 0.1726,
+      "step": 1719
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 11.799061608666374,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 1720
+    },
+    {
+      "epoch": 0.5909712722298222,
+      "grad_norm": 11.744400313727603,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 1721
+    },
+    {
+      "epoch": 0.5937072503419972,
+      "grad_norm": 12.186497440771998,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1722
+    },
+    {
+      "epoch": 0.5964432284541724,
+      "grad_norm": 11.831670256902074,
+      "learning_rate": 1e-06,
+      "loss": 0.1719,
+      "step": 1723
+    },
+    {
+      "epoch": 0.5991792065663475,
+      "grad_norm": 6.938570457644639,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 1724
+    },
+    {
+      "epoch": 0.6019151846785226,
+      "grad_norm": 10.022516502023587,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6046511627906976,
+      "grad_norm": 10.391199498029296,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 1726
+    },
+    {
+      "epoch": 0.6073871409028728,
+      "grad_norm": 12.30608953121559,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6101231190150479,
+      "grad_norm": 9.955794606219074,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 1728
+    },
+    {
+      "epoch": 0.612859097127223,
+      "grad_norm": 10.111434797558697,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 1729
+    },
+    {
+      "epoch": 0.615595075239398,
+      "grad_norm": 11.6708575984067,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 1730
+    },
+    {
+      "epoch": 0.6183310533515732,
+      "grad_norm": 7.828437405801201,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6210670314637483,
+      "grad_norm": 10.678997366392775,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6238030095759234,
+      "grad_norm": 8.566109010150933,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6265389876880985,
+      "grad_norm": 11.151503747769027,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6292749658002736,
+      "grad_norm": 9.609395824035222,
+      "learning_rate": 1e-06,
+      "loss": 0.1633,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6320109439124487,
+      "grad_norm": 12.033532705845156,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6347469220246238,
+      "grad_norm": 11.16605557895303,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6374829001367989,
+      "grad_norm": 13.084718976775546,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 1738
+    },
+    {
+      "epoch": 0.640218878248974,
+      "grad_norm": 12.094803368498185,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6429548563611491,
+      "grad_norm": 8.62718781024542,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6456908344733242,
+      "grad_norm": 10.590933227178965,
+      "learning_rate": 1e-06,
+      "loss": 0.1707,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6484268125854993,
+      "grad_norm": 8.787427459028173,
+      "learning_rate": 1e-06,
+      "loss": 0.1769,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6511627906976745,
+      "grad_norm": 8.561076862306722,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6538987688098495,
+      "grad_norm": 10.428363721508484,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6566347469220246,
+      "grad_norm": 8.76455678650161,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6593707250341997,
+      "grad_norm": 10.718048958745369,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6621067031463749,
+      "grad_norm": 10.122653173225894,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 1747
+    },
+    {
+      "epoch": 0.66484268125855,
+      "grad_norm": 12.709196331453066,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 1748
+    },
+    {
+      "epoch": 0.667578659370725,
+      "grad_norm": 11.509080328490532,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6703146374829001,
+      "grad_norm": 10.007808844294315,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6730506155950753,
+      "grad_norm": 9.145162685369526,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6757865937072504,
+      "grad_norm": 9.65598233386882,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6785225718194254,
+      "grad_norm": 11.002899783438236,
+      "learning_rate": 1e-06,
+      "loss": 0.1742,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6812585499316005,
+      "grad_norm": 12.305954938980868,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6839945280437757,
+      "grad_norm": 7.080482376988474,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6867305061559508,
+      "grad_norm": 9.810503551927464,
+      "learning_rate": 1e-06,
+      "loss": 0.1756,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6894664842681258,
+      "grad_norm": 11.590943347866228,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6922024623803009,
+      "grad_norm": 12.423684071329525,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6949384404924761,
+      "grad_norm": 11.123571071233549,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6976744186046512,
+      "grad_norm": 9.717636689813979,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 1760
+    },
+    {
+      "epoch": 0.7004103967168263,
+      "grad_norm": 8.159127119807858,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 1761
+    },
+    {
+      "epoch": 0.7031463748290013,
+      "grad_norm": 10.107329292610483,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1762
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 10.565590955502811,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 1763
+    },
+    {
+      "epoch": 0.7086183310533516,
+      "grad_norm": 10.65637052532875,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 1764
+    },
+    {
+      "epoch": 0.7113543091655267,
+      "grad_norm": 10.406884180293853,
+      "learning_rate": 1e-06,
+      "loss": 0.1836,
+      "step": 1765
+    },
+    {
+      "epoch": 0.7140902872777017,
+      "grad_norm": 13.084588377569998,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 1766
+    },
+    {
+      "epoch": 0.7168262653898769,
+      "grad_norm": 12.232284090983212,
+      "learning_rate": 1e-06,
+      "loss": 0.1795,
+      "step": 1767
+    },
+    {
+      "epoch": 0.719562243502052,
+      "grad_norm": 7.859642846178349,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 1768
+    },
+    {
+      "epoch": 0.7222982216142271,
+      "grad_norm": 8.03171018375466,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1769
+    },
+    {
+      "epoch": 0.7250341997264022,
+      "grad_norm": 9.569479678066937,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1770
+    },
+    {
+      "epoch": 0.7277701778385773,
+      "grad_norm": 11.469528812812467,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 1771
+    },
+    {
+      "epoch": 0.7305061559507524,
+      "grad_norm": 11.885702528018413,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 1772
+    },
+    {
+      "epoch": 0.7332421340629275,
+      "grad_norm": 11.534412735430967,
+      "learning_rate": 1e-06,
+      "loss": 0.1647,
+      "step": 1773
+    },
+    {
+      "epoch": 0.7359781121751026,
+      "grad_norm": 8.891686578191779,
+      "learning_rate": 1e-06,
+      "loss": 0.1689,
+      "step": 1774
+    },
+    {
+      "epoch": 0.7387140902872777,
+      "grad_norm": 9.691283080014736,
+      "learning_rate": 1e-06,
+      "loss": 0.1739,
+      "step": 1775
+    },
+    {
+      "epoch": 0.7414500683994528,
+      "grad_norm": 9.44966647100449,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 1776
+    },
+    {
+      "epoch": 0.7441860465116279,
+      "grad_norm": 11.217450544690381,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 1777
+    },
+    {
+      "epoch": 0.746922024623803,
+      "grad_norm": 8.487744698197666,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 1778
+    },
+    {
+      "epoch": 0.7496580027359782,
+      "grad_norm": 11.713215388068216,
+      "learning_rate": 1e-06,
+      "loss": 0.1867,
+      "step": 1779
+    },
+    {
+      "epoch": 0.7523939808481532,
+      "grad_norm": 9.268094803239713,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 1780
+    },
+    {
+      "epoch": 0.7551299589603283,
+      "grad_norm": 10.352070921387133,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 1781
+    },
+    {
+      "epoch": 0.7578659370725034,
+      "grad_norm": 12.878334172830462,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 1782
+    },
+    {
+      "epoch": 0.7606019151846786,
+      "grad_norm": 7.499235403182706,
+      "learning_rate": 1e-06,
+      "loss": 0.1639,
+      "step": 1783
+    },
+    {
+      "epoch": 0.7633378932968536,
+      "grad_norm": 13.591212282084513,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 1784
+    },
+    {
+      "epoch": 0.7660738714090287,
+      "grad_norm": 9.260835205024508,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 1785
+    },
+    {
+      "epoch": 0.7688098495212038,
+      "grad_norm": 10.961710651630192,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 1786
+    },
+    {
+      "epoch": 0.771545827633379,
+      "grad_norm": 10.11769597353802,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 1787
+    },
+    {
+      "epoch": 0.774281805745554,
+      "grad_norm": 12.141676818440233,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 1788
+    },
+    {
+      "epoch": 0.7770177838577291,
+      "grad_norm": 8.62216881477987,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 1789
+    },
+    {
+      "epoch": 0.7797537619699042,
+      "grad_norm": 7.4015507367642375,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 1790
+    },
+    {
+      "epoch": 0.7824897400820794,
+      "grad_norm": 9.850670362109618,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 1791
+    },
+    {
+      "epoch": 0.7852257181942545,
+      "grad_norm": 10.512414661903444,
+      "learning_rate": 1e-06,
+      "loss": 0.1801,
+      "step": 1792
+    },
+    {
+      "epoch": 0.7879616963064295,
+      "grad_norm": 9.58542842465836,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 1793
+    },
+    {
+      "epoch": 0.7906976744186046,
+      "grad_norm": 11.222727195402308,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 1794
+    },
+    {
+      "epoch": 0.7934336525307798,
+      "grad_norm": 11.096703100439946,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 1795
+    },
+    {
+      "epoch": 0.7961696306429549,
+      "grad_norm": 8.8668701436214,
+      "learning_rate": 1e-06,
+      "loss": 0.1657,
+      "step": 1796
+    },
+    {
+      "epoch": 0.79890560875513,
+      "grad_norm": 9.228127485191411,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 1797
+    },
+    {
+      "epoch": 0.801641586867305,
+      "grad_norm": 11.306065003999462,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8043775649794802,
+      "grad_norm": 9.268480895513036,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8071135430916553,
+      "grad_norm": 10.482614868110002,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8098495212038304,
+      "grad_norm": 10.418977307915801,
+      "learning_rate": 1e-06,
+      "loss": 0.1701,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8125854993160054,
+      "grad_norm": 10.336257475216547,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8153214774281806,
+      "grad_norm": 11.99158254168856,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8180574555403557,
+      "grad_norm": 12.184051216481551,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 1804
+    },
+    {
+      "epoch": 0.8207934336525308,
+      "grad_norm": 10.860797102180822,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 10.574636371748314,
+      "learning_rate": 1e-06,
+      "loss": 0.1748,
+      "step": 1806
+    },
+    {
+      "epoch": 0.826265389876881,
+      "grad_norm": 6.367091238209832,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8290013679890561,
+      "grad_norm": 11.024720834790303,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8317373461012312,
+      "grad_norm": 9.885656448952629,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8344733242134063,
+      "grad_norm": 8.889716955194528,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8372093023255814,
+      "grad_norm": 9.36387165920665,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8399452804377565,
+      "grad_norm": 10.83812413524357,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8426812585499316,
+      "grad_norm": 10.467121177146431,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8454172366621067,
+      "grad_norm": 9.990050953137233,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8481532147742818,
+      "grad_norm": 10.477325863010835,
+      "learning_rate": 1e-06,
+      "loss": 0.1722,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8508891928864569,
+      "grad_norm": 9.406955752580341,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 1816
+    },
+    {
+      "epoch": 0.853625170998632,
+      "grad_norm": 9.018261704051476,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8563611491108071,
+      "grad_norm": 8.41754560958752,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 1818
+    },
+    {
+      "epoch": 0.8590971272229823,
+      "grad_norm": 10.46492307559433,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8618331053351573,
+      "grad_norm": 8.80208762076963,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8645690834473324,
+      "grad_norm": 11.99084310188209,
+      "learning_rate": 1e-06,
+      "loss": 0.1606,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8673050615595075,
+      "grad_norm": 11.809690922821758,
+      "learning_rate": 1e-06,
+      "loss": 0.1752,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8700410396716827,
+      "grad_norm": 11.039374412605136,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8727770177838577,
+      "grad_norm": 11.846095720776333,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8755129958960328,
+      "grad_norm": 11.764000874233375,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8782489740082079,
+      "grad_norm": 11.329887878199697,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 1826
+    },
+    {
+      "epoch": 0.8809849521203831,
+      "grad_norm": 7.93840704946836,
+      "learning_rate": 1e-06,
+      "loss": 0.1694,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8837209302325582,
+      "grad_norm": 11.2500490050006,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8864569083447332,
+      "grad_norm": 8.892054331132885,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8891928864569083,
+      "grad_norm": 10.896976059833065,
+      "learning_rate": 1e-06,
+      "loss": 0.1739,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8919288645690835,
+      "grad_norm": 8.590134141929482,
+      "learning_rate": 1e-06,
+      "loss": 0.1648,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8946648426812586,
+      "grad_norm": 8.903345558630297,
+      "learning_rate": 1e-06,
+      "loss": 0.1554,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8974008207934336,
+      "grad_norm": 11.433286168016378,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 1833
+    },
+    {
+      "epoch": 0.9001367989056087,
+      "grad_norm": 11.668767909152358,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 1834
+    },
+    {
+      "epoch": 0.9028727770177839,
+      "grad_norm": 10.49487448871191,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 1835
+    },
+    {
+      "epoch": 0.905608755129959,
+      "grad_norm": 8.79902393256028,
+      "learning_rate": 1e-06,
+      "loss": 0.164,
+      "step": 1836
+    },
+    {
+      "epoch": 0.908344733242134,
+      "grad_norm": 11.81188614093435,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 1837
+    },
+    {
+      "epoch": 0.9110807113543091,
+      "grad_norm": 10.224753006565704,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 1838
+    },
+    {
+      "epoch": 0.9138166894664843,
+      "grad_norm": 12.134398510573456,
+      "learning_rate": 1e-06,
+      "loss": 0.1661,
+      "step": 1839
+    },
+    {
+      "epoch": 0.9165526675786594,
+      "grad_norm": 10.462280541957787,
+      "learning_rate": 1e-06,
+      "loss": 0.1675,
+      "step": 1840
+    },
+    {
+      "epoch": 0.9192886456908345,
+      "grad_norm": 9.500818579274652,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 1841
+    },
+    {
+      "epoch": 0.9220246238030095,
+      "grad_norm": 11.193475195866746,
+      "learning_rate": 1e-06,
+      "loss": 0.1834,
+      "step": 1842
+    },
+    {
+      "epoch": 0.9247606019151847,
+      "grad_norm": 11.728034746109023,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 1843
+    },
+    {
+      "epoch": 0.9274965800273598,
+      "grad_norm": 10.161627673028336,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 1844
+    },
+    {
+      "epoch": 0.9302325581395349,
+      "grad_norm": 12.334400294956064,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 1845
+    },
+    {
+      "epoch": 0.93296853625171,
+      "grad_norm": 13.974065596677473,
+      "learning_rate": 1e-06,
+      "loss": 0.1639,
+      "step": 1846
+    },
+    {
+      "epoch": 0.9357045143638851,
+      "grad_norm": 7.249628417835854,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 1847
+    },
+    {
+      "epoch": 0.9384404924760602,
+      "grad_norm": 9.316744516237632,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 1848
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 8.089899313339634,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 1849
+    },
+    {
+      "epoch": 0.9439124487004104,
+      "grad_norm": 10.16066214376708,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 1850
+    },
+    {
+      "epoch": 0.9466484268125855,
+      "grad_norm": 10.099640625213908,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 1851
+    },
+    {
+      "epoch": 0.9493844049247606,
+      "grad_norm": 10.551373145917683,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 1852
+    },
+    {
+      "epoch": 0.9521203830369357,
+      "grad_norm": 9.494408464692782,
+      "learning_rate": 1e-06,
+      "loss": 0.1569,
+      "step": 1853
+    },
+    {
+      "epoch": 0.9548563611491108,
+      "grad_norm": 9.690003376909404,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 1854
+    },
+    {
+      "epoch": 0.957592339261286,
+      "grad_norm": 10.524955979011247,
+      "learning_rate": 1e-06,
+      "loss": 0.1795,
+      "step": 1855
+    },
+    {
+      "epoch": 0.960328317373461,
+      "grad_norm": 10.185821975488512,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 1856
+    },
+    {
+      "epoch": 0.9630642954856361,
+      "grad_norm": 9.034850342175256,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 1857
+    },
+    {
+      "epoch": 0.9658002735978112,
+      "grad_norm": 10.860883184051612,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 1858
+    },
+    {
+      "epoch": 0.9685362517099864,
+      "grad_norm": 7.688797526078269,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 1859
+    },
+    {
+      "epoch": 0.9712722298221614,
+      "grad_norm": 10.006342546275283,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 1860
+    },
+    {
+      "epoch": 0.9740082079343365,
+      "grad_norm": 9.854827167057353,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 1861
+    },
+    {
+      "epoch": 0.9767441860465116,
+      "grad_norm": 8.865701282301288,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 1862
+    },
+    {
+      "epoch": 0.9794801641586868,
+      "grad_norm": 10.411136584034791,
+      "learning_rate": 1e-06,
+      "loss": 0.1722,
+      "step": 1863
+    },
+    {
+      "epoch": 0.9822161422708618,
+      "grad_norm": 10.637155884678215,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 1864
+    },
+    {
+      "epoch": 0.9849521203830369,
+      "grad_norm": 9.990218400219948,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 1865
+    },
+    {
+      "epoch": 0.987688098495212,
+      "grad_norm": 10.493117798143645,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 1866
+    },
+    {
+      "epoch": 0.9904240766073872,
+      "grad_norm": 12.8558187205912,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 1867
+    },
+    {
+      "epoch": 0.9931600547195623,
+      "grad_norm": 9.406112353241634,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 1868
+    },
+    {
+      "epoch": 0.9958960328317373,
+      "grad_norm": 10.418757507314934,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 1869
+    },
+    {
+      "epoch": 0.9986320109439124,
+      "grad_norm": 7.922605358563275,
+      "learning_rate": 1e-06,
+      "loss": 0.1739,
+      "step": 1870
+    },
+    {
+      "epoch": 0.9986320109439124,
+      "eval_loss": 3.2260634899139404,
+      "eval_runtime": 21.8035,
+      "eval_samples_per_second": 45.864,
+      "eval_steps_per_second": 5.733,
+      "step": 1870
+    },
+    {
+      "epoch": 0.0036036036036036037,
+      "grad_norm": 9.415705090733116,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 1871
+    },
+    {
+      "epoch": 0.007207207207207207,
+      "grad_norm": 10.910665929299158,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 1872
+    },
+    {
+      "epoch": 0.010810810810810811,
+      "grad_norm": 8.680756058110509,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 1873
+    },
+    {
+      "epoch": 0.014414414414414415,
+      "grad_norm": 8.418378859794737,
+      "learning_rate": 1e-06,
+      "loss": 0.1406,
+      "step": 1874
+    },
+    {
+      "epoch": 0.018018018018018018,
+      "grad_norm": 12.51094768812987,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 1875
+    },
+    {
+      "epoch": 0.021621621621621623,
+      "grad_norm": 9.125814517864903,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 1876
+    },
+    {
+      "epoch": 0.025225225225225224,
+      "grad_norm": 8.465740762773123,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 1877
+    },
+    {
+      "epoch": 0.02882882882882883,
+      "grad_norm": 8.251706338865826,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 1878
+    },
+    {
+      "epoch": 0.032432432432432434,
+      "grad_norm": 9.396244010564741,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 1879
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 8.624444028165643,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 1880
+    },
+    {
+      "epoch": 0.03963963963963964,
+      "grad_norm": 8.022526906503222,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 1881
+    },
+    {
+      "epoch": 0.043243243243243246,
+      "grad_norm": 13.25284581027488,
+      "learning_rate": 1e-06,
+      "loss": 0.1417,
+      "step": 1882
+    },
+    {
+      "epoch": 0.04684684684684685,
+      "grad_norm": 9.74139995673603,
+      "learning_rate": 1e-06,
+      "loss": 0.1404,
+      "step": 1883
+    },
+    {
+      "epoch": 0.05045045045045045,
+      "grad_norm": 10.350791907058465,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 1884
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 10.16578153364529,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 1885
+    },
+    {
+      "epoch": 0.05765765765765766,
+      "grad_norm": 8.104417458401272,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 1886
+    },
+    {
+      "epoch": 0.06126126126126126,
+      "grad_norm": 9.513527323742228,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 1887
+    },
+    {
+      "epoch": 0.06486486486486487,
+      "grad_norm": 9.190723428846953,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 1888
+    },
+    {
+      "epoch": 0.06846846846846846,
+      "grad_norm": 11.809530548756495,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 1889
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 11.102978452787656,
+      "learning_rate": 1e-06,
+      "loss": 0.1455,
+      "step": 1890
+    },
+    {
+      "epoch": 0.07567567567567568,
+      "grad_norm": 13.13461717383435,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 1891
+    },
+    {
+      "epoch": 0.07927927927927927,
+      "grad_norm": 7.663170705910772,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 1892
+    },
+    {
+      "epoch": 0.08288288288288288,
+      "grad_norm": 13.406145750410996,
+      "learning_rate": 1e-06,
+      "loss": 0.1376,
+      "step": 1893
+    },
+    {
+      "epoch": 0.08648648648648649,
+      "grad_norm": 8.706376365772472,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 1894
+    },
+    {
+      "epoch": 0.09009009009009009,
+      "grad_norm": 10.338138758354214,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 1895
+    },
+    {
+      "epoch": 0.0936936936936937,
+      "grad_norm": 11.31536845229905,
+      "learning_rate": 1e-06,
+      "loss": 0.1416,
+      "step": 1896
+    },
+    {
+      "epoch": 0.0972972972972973,
+      "grad_norm": 10.156847036825306,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 1897
+    },
+    {
+      "epoch": 0.1009009009009009,
+      "grad_norm": 9.22510350597727,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 1898
+    },
+    {
+      "epoch": 0.1045045045045045,
+      "grad_norm": 10.690250861550576,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 1899
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 7.356020425630322,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 1900
+    },
+    {
+      "epoch": 0.11171171171171171,
+      "grad_norm": 13.937995369709435,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 1901
+    },
+    {
+      "epoch": 0.11531531531531532,
+      "grad_norm": 13.788866943118844,
+      "learning_rate": 1e-06,
+      "loss": 0.1581,
+      "step": 1902
+    },
+    {
+      "epoch": 0.11891891891891893,
+      "grad_norm": 10.967594594754479,
+      "learning_rate": 1e-06,
+      "loss": 0.1474,
+      "step": 1903
+    },
+    {
+      "epoch": 0.12252252252252252,
+      "grad_norm": 11.868237975580236,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 1904
+    },
+    {
+      "epoch": 0.12612612612612611,
+      "grad_norm": 7.524189145293153,
+      "learning_rate": 1e-06,
+      "loss": 0.1483,
+      "step": 1905
+    },
+    {
+      "epoch": 0.12972972972972974,
+      "grad_norm": 10.446916448648096,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 1906
+    },
+    {
+      "epoch": 0.13333333333333333,
+      "grad_norm": 11.07437116911299,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 1907
+    },
+    {
+      "epoch": 0.13693693693693693,
+      "grad_norm": 11.502776102727529,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 1908
+    },
+    {
+      "epoch": 0.14054054054054055,
+      "grad_norm": 10.636076053931356,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1909
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 13.313061762641416,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 1910
+    },
+    {
+      "epoch": 0.14774774774774774,
+      "grad_norm": 10.711721570490898,
+      "learning_rate": 1e-06,
+      "loss": 0.1429,
+      "step": 1911
+    },
+    {
+      "epoch": 0.15135135135135136,
+      "grad_norm": 10.828009551251293,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 1912
+    },
+    {
+      "epoch": 0.15495495495495495,
+      "grad_norm": 12.22388831469904,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 1913
+    },
+    {
+      "epoch": 0.15855855855855855,
+      "grad_norm": 13.576709794693059,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 1914
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 8.68387982452374,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 1915
+    },
+    {
+      "epoch": 0.16576576576576577,
+      "grad_norm": 10.861684726372193,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 1916
+    },
+    {
+      "epoch": 0.16936936936936936,
+      "grad_norm": 12.218357864505894,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 1917
+    },
+    {
+      "epoch": 0.17297297297297298,
+      "grad_norm": 9.154282911529902,
+      "learning_rate": 1e-06,
+      "loss": 0.1505,
+      "step": 1918
+    },
+    {
+      "epoch": 0.17657657657657658,
+      "grad_norm": 10.6292890113222,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 1919
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 10.377170599405192,
+      "learning_rate": 1e-06,
+      "loss": 0.142,
+      "step": 1920
+    },
+    {
+      "epoch": 0.1837837837837838,
+      "grad_norm": 11.974709722941228,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 1921
+    },
+    {
+      "epoch": 0.1873873873873874,
+      "grad_norm": 9.84343186869173,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 1922
+    },
+    {
+      "epoch": 0.19099099099099098,
+      "grad_norm": 9.490595950201161,
+      "learning_rate": 1e-06,
+      "loss": 0.1505,
+      "step": 1923
+    },
+    {
+      "epoch": 0.1945945945945946,
+      "grad_norm": 12.941112698976609,
+      "learning_rate": 1e-06,
+      "loss": 0.153,
+      "step": 1924
+    },
+    {
+      "epoch": 0.1981981981981982,
+      "grad_norm": 11.739355676145662,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 1925
+    },
+    {
+      "epoch": 0.2018018018018018,
+      "grad_norm": 11.095558211564704,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 1926
+    },
+    {
+      "epoch": 0.20540540540540542,
+      "grad_norm": 10.64967348826955,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 1927
+    },
+    {
+      "epoch": 0.209009009009009,
+      "grad_norm": 9.279194333894555,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 1928
+    },
+    {
+      "epoch": 0.2126126126126126,
+      "grad_norm": 12.341169379719748,
+      "learning_rate": 1e-06,
+      "loss": 0.1594,
+      "step": 1929
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 11.02582681754316,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 1930
+    },
+    {
+      "epoch": 0.21981981981981982,
+      "grad_norm": 11.500587937172645,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 1931
+    },
+    {
+      "epoch": 0.22342342342342342,
+      "grad_norm": 12.846590563806776,
+      "learning_rate": 1e-06,
+      "loss": 0.1472,
+      "step": 1932
+    },
+    {
+      "epoch": 0.22702702702702704,
+      "grad_norm": 8.90172078136038,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 1933
+    },
+    {
+      "epoch": 0.23063063063063063,
+      "grad_norm": 10.481086524671698,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 1934
+    },
+    {
+      "epoch": 0.23423423423423423,
+      "grad_norm": 11.887887592072113,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 1935
+    },
+    {
+      "epoch": 0.23783783783783785,
+      "grad_norm": 13.571869042974212,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 1936
+    },
+    {
+      "epoch": 0.24144144144144145,
+      "grad_norm": 10.343099286046947,
+      "learning_rate": 1e-06,
+      "loss": 0.1455,
+      "step": 1937
+    },
+    {
+      "epoch": 0.24504504504504504,
+      "grad_norm": 13.221439917610072,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 1938
+    },
+    {
+      "epoch": 0.24864864864864866,
+      "grad_norm": 11.533388039998536,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 1939
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 7.454296276126189,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 1940
+    },
+    {
+      "epoch": 0.25585585585585585,
+      "grad_norm": 8.48636962004724,
+      "learning_rate": 1e-06,
+      "loss": 0.1405,
+      "step": 1941
+    },
+    {
+      "epoch": 0.2594594594594595,
+      "grad_norm": 10.150378748151118,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 1942
+    },
+    {
+      "epoch": 0.26306306306306304,
+      "grad_norm": 9.560428677515995,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 1943
+    },
+    {
+      "epoch": 0.26666666666666666,
+      "grad_norm": 15.275867412001006,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 1944
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 10.8273853909978,
+      "learning_rate": 1e-06,
+      "loss": 0.1536,
+      "step": 1945
+    },
+    {
+      "epoch": 0.27387387387387385,
+      "grad_norm": 9.561043398338128,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 1946
+    },
+    {
+      "epoch": 0.2774774774774775,
+      "grad_norm": 11.255252455477116,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 1947
+    },
+    {
+      "epoch": 0.2810810810810811,
+      "grad_norm": 8.397919144491466,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 1948
+    },
+    {
+      "epoch": 0.28468468468468466,
+      "grad_norm": 9.879532113637584,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 1949
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 11.277753938133769,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 1950
+    },
+    {
+      "epoch": 0.2918918918918919,
+      "grad_norm": 14.89768115327551,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 1951
+    },
+    {
+      "epoch": 0.2954954954954955,
+      "grad_norm": 14.89768115327551,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 1952
+    },
+    {
+      "epoch": 0.2990990990990991,
+      "grad_norm": 11.997672350030074,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1953
+    },
+    {
+      "epoch": 0.3027027027027027,
+      "grad_norm": 10.978556527749713,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 1954
+    },
+    {
+      "epoch": 0.3063063063063063,
+      "grad_norm": 10.299506599968048,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 1955
+    },
+    {
+      "epoch": 0.3099099099099099,
+      "grad_norm": 11.463331524661283,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 1956
+    },
+    {
+      "epoch": 0.31351351351351353,
+      "grad_norm": 12.553622779273283,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 1957
+    },
+    {
+      "epoch": 0.3171171171171171,
+      "grad_norm": 12.724850030352105,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 1958
+    },
+    {
+      "epoch": 0.3207207207207207,
+      "grad_norm": 12.279336494248701,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 1959
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 11.314397235645403,
+      "learning_rate": 1e-06,
+      "loss": 0.1494,
+      "step": 1960
+    },
+    {
+      "epoch": 0.3279279279279279,
+      "grad_norm": 11.842180328426581,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 1961
+    },
+    {
+      "epoch": 0.33153153153153153,
+      "grad_norm": 11.842180328426581,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 1962
+    },
+    {
+      "epoch": 0.33513513513513515,
+      "grad_norm": 11.003951541523389,
+      "learning_rate": 1e-06,
+      "loss": 0.1486,
+      "step": 1963
+    },
+    {
+      "epoch": 0.3387387387387387,
+      "grad_norm": 9.425785534617038,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 1964
+    },
+    {
+      "epoch": 0.34234234234234234,
+      "grad_norm": 13.06363869225882,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 1965
+    },
+    {
+      "epoch": 0.34594594594594597,
+      "grad_norm": 13.448098312103047,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 1966
+    },
+    {
+      "epoch": 0.34954954954954953,
+      "grad_norm": 8.968581517477919,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 1967
+    },
+    {
+      "epoch": 0.35315315315315315,
+      "grad_norm": 14.85710087545447,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 1968
+    },
+    {
+      "epoch": 0.3567567567567568,
+      "grad_norm": 8.419446005361598,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 1969
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 10.907214907146765,
+      "learning_rate": 1e-06,
+      "loss": 0.1356,
+      "step": 1970
+    },
+    {
+      "epoch": 0.36396396396396397,
+      "grad_norm": 9.378954343042414,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 1971
+    },
+    {
+      "epoch": 0.3675675675675676,
+      "grad_norm": 10.264776479228052,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 1972
+    },
+    {
+      "epoch": 0.37117117117117115,
+      "grad_norm": 8.623981422946812,
+      "learning_rate": 1e-06,
+      "loss": 0.146,
+      "step": 1973
+    },
+    {
+      "epoch": 0.3747747747747748,
+      "grad_norm": 10.399139266624363,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 1974
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 9.525754156141014,
+      "learning_rate": 1e-06,
+      "loss": 0.1463,
+      "step": 1975
+    },
+    {
+      "epoch": 0.38198198198198197,
+      "grad_norm": 13.623988396845142,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 1976
+    },
+    {
+      "epoch": 0.3855855855855856,
+      "grad_norm": 12.673666406251648,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 1977
+    },
+    {
+      "epoch": 0.3891891891891892,
+      "grad_norm": 12.161202990680868,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 1978
+    },
+    {
+      "epoch": 0.3927927927927928,
+      "grad_norm": 12.224469041510243,
+      "learning_rate": 1e-06,
+      "loss": 0.1602,
+      "step": 1979
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 8.311875651650398,
+      "learning_rate": 1e-06,
+      "loss": 0.1483,
+      "step": 1980
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 10.733949929084956,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 1981
+    },
+    {
+      "epoch": 0.4036036036036036,
+      "grad_norm": 9.304534376509553,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 1982
+    },
+    {
+      "epoch": 0.4072072072072072,
+      "grad_norm": 10.745251972003533,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 1983
+    },
+    {
+      "epoch": 0.41081081081081083,
+      "grad_norm": 11.51804729231904,
+      "learning_rate": 1e-06,
+      "loss": 0.1647,
+      "step": 1984
+    },
+    {
+      "epoch": 0.4144144144144144,
+      "grad_norm": 13.988106945325246,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 1985
+    },
+    {
+      "epoch": 0.418018018018018,
+      "grad_norm": 10.727265824687143,
+      "learning_rate": 1e-06,
+      "loss": 0.1351,
+      "step": 1986
+    },
+    {
+      "epoch": 0.42162162162162165,
+      "grad_norm": 9.681016451348983,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 1987
+    },
+    {
+      "epoch": 0.4252252252252252,
+      "grad_norm": 9.370076771102818,
+      "learning_rate": 1e-06,
+      "loss": 0.1461,
+      "step": 1988
+    },
+    {
+      "epoch": 0.42882882882882883,
+      "grad_norm": 10.780564680855512,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 1989
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 11.169032304903142,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 1990
+    },
+    {
+      "epoch": 0.436036036036036,
+      "grad_norm": 8.70338055694096,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 1991
+    },
+    {
+      "epoch": 0.43963963963963965,
+      "grad_norm": 10.317336005625092,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 1992
+    },
+    {
+      "epoch": 0.44324324324324327,
+      "grad_norm": 11.632172313368137,
+      "learning_rate": 1e-06,
+      "loss": 0.1451,
+      "step": 1993
+    },
+    {
+      "epoch": 0.44684684684684683,
+      "grad_norm": 9.128407396451712,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 1994
+    },
+    {
+      "epoch": 0.45045045045045046,
+      "grad_norm": 9.7013761003716,
+      "learning_rate": 1e-06,
+      "loss": 0.1623,
+      "step": 1995
+    },
+    {
+      "epoch": 0.4540540540540541,
+      "grad_norm": 10.873090989414118,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 1996
+    },
+    {
+      "epoch": 0.45765765765765765,
+      "grad_norm": 11.261520355443144,
+      "learning_rate": 1e-06,
+      "loss": 0.1417,
+      "step": 1997
+    },
+    {
+      "epoch": 0.46126126126126127,
+      "grad_norm": 9.478737880094457,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 1998
+    },
+    {
+      "epoch": 0.4648648648648649,
+      "grad_norm": 11.395074914166784,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 1999
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 10.334028173179725,
+      "learning_rate": 1e-06,
+      "loss": 0.1431,
+      "step": 2000
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "eval_loss": 3.2478280067443848,
+      "eval_runtime": 21.5531,
+      "eval_samples_per_second": 46.397,
+      "eval_steps_per_second": 5.8,
+      "step": 2000
+    },
+    {
+      "epoch": 0.4720720720720721,
+      "grad_norm": 11.88854017669651,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 2001
+    },
+    {
+      "epoch": 0.4756756756756757,
+      "grad_norm": 9.733316287265454,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 2002
+    },
+    {
+      "epoch": 0.47927927927927927,
+      "grad_norm": 11.931035916066937,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 2003
+    },
+    {
+      "epoch": 0.4828828828828829,
+      "grad_norm": 8.376264933480154,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 2004
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 13.008140546726022,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 2005
+    },
+    {
+      "epoch": 0.4900900900900901,
+      "grad_norm": 10.028117712936684,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 2006
+    },
+    {
+      "epoch": 0.4936936936936937,
+      "grad_norm": 8.377628765250057,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2007
+    },
+    {
+      "epoch": 0.4972972972972973,
+      "grad_norm": 9.976913060239506,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 2008
+    },
+    {
+      "epoch": 0.5009009009009009,
+      "grad_norm": 12.784711030398324,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 2009
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 11.141619361564308,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 2010
+    },
+    {
+      "epoch": 0.5081081081081081,
+      "grad_norm": 9.633561812329239,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 2011
+    },
+    {
+      "epoch": 0.5117117117117117,
+      "grad_norm": 10.58858181693258,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 2012
+    },
+    {
+      "epoch": 0.5153153153153153,
+      "grad_norm": 10.572758462176687,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 2013
+    },
+    {
+      "epoch": 0.518918918918919,
+      "grad_norm": 8.064650541251083,
+      "learning_rate": 1e-06,
+      "loss": 0.1435,
+      "step": 2014
+    },
+    {
+      "epoch": 0.5225225225225225,
+      "grad_norm": 9.331619732255424,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 2015
+    },
+    {
+      "epoch": 0.5261261261261261,
+      "grad_norm": 11.777463946037527,
+      "learning_rate": 1e-06,
+      "loss": 0.1471,
+      "step": 2016
+    },
+    {
+      "epoch": 0.5297297297297298,
+      "grad_norm": 9.151459563407526,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 2017
+    },
+    {
+      "epoch": 0.5333333333333333,
+      "grad_norm": 10.957987337836537,
+      "learning_rate": 1e-06,
+      "loss": 0.1534,
+      "step": 2018
+    },
+    {
+      "epoch": 0.5369369369369369,
+      "grad_norm": 13.5206242798445,
+      "learning_rate": 1e-06,
+      "loss": 0.1499,
+      "step": 2019
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 11.292559973865343,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 2020
+    },
+    {
+      "epoch": 0.5441441441441441,
+      "grad_norm": 11.5641066553748,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 2021
+    },
+    {
+      "epoch": 0.5477477477477477,
+      "grad_norm": 10.038725365435512,
+      "learning_rate": 1e-06,
+      "loss": 0.1466,
+      "step": 2022
+    },
+    {
+      "epoch": 0.5513513513513514,
+      "grad_norm": 10.179199340435309,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 2023
+    },
+    {
+      "epoch": 0.554954954954955,
+      "grad_norm": 13.141920837640475,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 2024
+    },
+    {
+      "epoch": 0.5585585585585585,
+      "grad_norm": 11.51598438508629,
+      "learning_rate": 1e-06,
+      "loss": 0.1701,
+      "step": 2025
+    },
+    {
+      "epoch": 0.5621621621621622,
+      "grad_norm": 14.036162089634002,
+      "learning_rate": 1e-06,
+      "loss": 0.1702,
+      "step": 2026
+    },
+    {
+      "epoch": 0.5657657657657658,
+      "grad_norm": 9.46013168550245,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 2027
+    },
+    {
+      "epoch": 0.5693693693693693,
+      "grad_norm": 9.846021997768858,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 2028
+    },
+    {
+      "epoch": 0.572972972972973,
+      "grad_norm": 12.788049916871747,
+      "learning_rate": 1e-06,
+      "loss": 0.1442,
+      "step": 2029
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 12.286545935869633,
+      "learning_rate": 1e-06,
+      "loss": 0.1612,
+      "step": 2030
+    },
+    {
+      "epoch": 0.5801801801801801,
+      "grad_norm": 10.472501784435494,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 2031
+    },
+    {
+      "epoch": 0.5837837837837838,
+      "grad_norm": 8.899586087756349,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 2032
+    },
+    {
+      "epoch": 0.5873873873873874,
+      "grad_norm": 13.707644347016084,
+      "learning_rate": 1e-06,
+      "loss": 0.153,
+      "step": 2033
+    },
+    {
+      "epoch": 0.590990990990991,
+      "grad_norm": 6.825144320173454,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 2034
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 11.689861290819973,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 2035
+    },
+    {
+      "epoch": 0.5981981981981982,
+      "grad_norm": 11.479512527111925,
+      "learning_rate": 1e-06,
+      "loss": 0.16,
+      "step": 2036
+    },
+    {
+      "epoch": 0.6018018018018018,
+      "grad_norm": 12.553012754446772,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 2037
+    },
+    {
+      "epoch": 0.6054054054054054,
+      "grad_norm": 11.222776326835604,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 2038
+    },
+    {
+      "epoch": 0.609009009009009,
+      "grad_norm": 12.549986624593705,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 2039
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 11.315947963238647,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 2040
+    },
+    {
+      "epoch": 0.6162162162162163,
+      "grad_norm": 8.086771564209371,
+      "learning_rate": 1e-06,
+      "loss": 0.1449,
+      "step": 2041
+    },
+    {
+      "epoch": 0.6198198198198198,
+      "grad_norm": 11.630489837647772,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 2042
+    },
+    {
+      "epoch": 0.6234234234234234,
+      "grad_norm": 10.234690724232156,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 2043
+    },
+    {
+      "epoch": 0.6270270270270271,
+      "grad_norm": 12.04915357245808,
+      "learning_rate": 1e-06,
+      "loss": 0.1507,
+      "step": 2044
+    },
+    {
+      "epoch": 0.6306306306306306,
+      "grad_norm": 9.700541339680314,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 2045
+    },
+    {
+      "epoch": 0.6342342342342342,
+      "grad_norm": 9.989465448573304,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 2046
+    },
+    {
+      "epoch": 0.6378378378378379,
+      "grad_norm": 13.935511871466352,
+      "learning_rate": 1e-06,
+      "loss": 0.1502,
+      "step": 2047
+    },
+    {
+      "epoch": 0.6414414414414414,
+      "grad_norm": 9.848316286643225,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 2048
+    },
+    {
+      "epoch": 0.645045045045045,
+      "grad_norm": 12.689971409929933,
+      "learning_rate": 1e-06,
+      "loss": 0.1608,
+      "step": 2049
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 10.575972392388318,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 2050
+    },
+    {
+      "epoch": 0.6522522522522523,
+      "grad_norm": 12.552697067083255,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 2051
+    },
+    {
+      "epoch": 0.6558558558558558,
+      "grad_norm": 9.12167872072162,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 2052
+    },
+    {
+      "epoch": 0.6594594594594595,
+      "grad_norm": 11.600426857207028,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 2053
+    },
+    {
+      "epoch": 0.6630630630630631,
+      "grad_norm": 13.625212055476041,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 2054
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 10.454815548143682,
+      "learning_rate": 1e-06,
+      "loss": 0.1502,
+      "step": 2055
+    },
+    {
+      "epoch": 0.6702702702702703,
+      "grad_norm": 12.237966173432794,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 2056
+    },
+    {
+      "epoch": 0.6738738738738739,
+      "grad_norm": 12.22242169736704,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 2057
+    },
+    {
+      "epoch": 0.6774774774774774,
+      "grad_norm": 8.717727313825593,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 2058
+    },
+    {
+      "epoch": 0.6810810810810811,
+      "grad_norm": 11.030565579726318,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 2059
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 8.742085161661576,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 2060
+    },
+    {
+      "epoch": 0.6882882882882883,
+      "grad_norm": 12.210825353622349,
+      "learning_rate": 1e-06,
+      "loss": 0.1566,
+      "step": 2061
+    },
+    {
+      "epoch": 0.6918918918918919,
+      "grad_norm": 8.003284918949483,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 2062
+    },
+    {
+      "epoch": 0.6954954954954955,
+      "grad_norm": 9.581248695221966,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 2063
+    },
+    {
+      "epoch": 0.6990990990990991,
+      "grad_norm": 11.029649576653085,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 12.997787489298412,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7063063063063063,
+      "grad_norm": 10.96460750671751,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7099099099099099,
+      "grad_norm": 11.697363729056775,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7135135135135136,
+      "grad_norm": 10.25422817624052,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7171171171171171,
+      "grad_norm": 10.678224470734175,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 10.33749939815884,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7243243243243244,
+      "grad_norm": 9.64338908398559,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 2071
+    },
+    {
+      "epoch": 0.7279279279279279,
+      "grad_norm": 12.749104386642088,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7315315315315315,
+      "grad_norm": 9.460118107372683,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 2073
+    },
+    {
+      "epoch": 0.7351351351351352,
+      "grad_norm": 14.877879802002386,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 2074
+    },
+    {
+      "epoch": 0.7387387387387387,
+      "grad_norm": 9.670876182648081,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 2075
+    },
+    {
+      "epoch": 0.7423423423423423,
+      "grad_norm": 11.302329015256642,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 2076
+    },
+    {
+      "epoch": 0.745945945945946,
+      "grad_norm": 9.211099307469722,
+      "learning_rate": 1e-06,
+      "loss": 0.1563,
+      "step": 2077
+    },
+    {
+      "epoch": 0.7495495495495496,
+      "grad_norm": 8.905670898167768,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 2078
+    },
+    {
+      "epoch": 0.7531531531531531,
+      "grad_norm": 10.045890728560122,
+      "learning_rate": 1e-06,
+      "loss": 0.1591,
+      "step": 2079
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 13.623659745140897,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 2080
+    },
+    {
+      "epoch": 0.7603603603603604,
+      "grad_norm": 6.91532044150687,
+      "learning_rate": 1e-06,
+      "loss": 0.1452,
+      "step": 2081
+    },
+    {
+      "epoch": 0.7639639639639639,
+      "grad_norm": 11.723192289048507,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 2082
+    },
+    {
+      "epoch": 0.7675675675675676,
+      "grad_norm": 10.756882201773626,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 2083
+    },
+    {
+      "epoch": 0.7711711711711712,
+      "grad_norm": 9.961013037462951,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 2084
+    },
+    {
+      "epoch": 0.7747747747747747,
+      "grad_norm": 11.189576594201206,
+      "learning_rate": 1e-06,
+      "loss": 0.1538,
+      "step": 2085
+    },
+    {
+      "epoch": 0.7783783783783784,
+      "grad_norm": 15.006427958683245,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 2086
+    },
+    {
+      "epoch": 0.781981981981982,
+      "grad_norm": 13.521903721138429,
+      "learning_rate": 1e-06,
+      "loss": 0.1602,
+      "step": 2087
+    },
+    {
+      "epoch": 0.7855855855855856,
+      "grad_norm": 13.371229094929664,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 2088
+    },
+    {
+      "epoch": 0.7891891891891892,
+      "grad_norm": 10.780528034737552,
+      "learning_rate": 1e-06,
+      "loss": 0.156,
+      "step": 2089
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 8.392435978106699,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 2090
+    },
+    {
+      "epoch": 0.7963963963963964,
+      "grad_norm": 13.990255631133525,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 15.823498295068228,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8036036036036036,
+      "grad_norm": 10.621366961813925,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8072072072072072,
+      "grad_norm": 10.228799925723774,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 10.398968576843492,
+      "learning_rate": 1e-06,
+      "loss": 0.1486,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8144144144144144,
+      "grad_norm": 11.928622569117161,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 2096
+    },
+    {
+      "epoch": 0.818018018018018,
+      "grad_norm": 10.616576761313992,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8216216216216217,
+      "grad_norm": 10.686027413379147,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8252252252252252,
+      "grad_norm": 9.534904085882872,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 9.121983860584487,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8324324324324325,
+      "grad_norm": 11.385528536927827,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 2101
+    },
+    {
+      "epoch": 0.836036036036036,
+      "grad_norm": 11.00690802458408,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8396396396396396,
+      "grad_norm": 13.991292245404557,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8432432432432433,
+      "grad_norm": 13.412903712013488,
+      "learning_rate": 1e-06,
+      "loss": 0.1556,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8468468468468469,
+      "grad_norm": 10.143202144213316,
+      "learning_rate": 1e-06,
+      "loss": 0.161,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8504504504504504,
+      "grad_norm": 12.756492549018605,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8540540540540541,
+      "grad_norm": 10.01499599705061,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8576576576576577,
+      "grad_norm": 9.551553948252671,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8612612612612612,
+      "grad_norm": 10.662267398375088,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 2109
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 14.164386021867857,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8684684684684685,
+      "grad_norm": 10.03081989454832,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 2111
+    },
+    {
+      "epoch": 0.872072072072072,
+      "grad_norm": 12.704536485415167,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8756756756756757,
+      "grad_norm": 11.364447054027345,
+      "learning_rate": 1e-06,
+      "loss": 0.1578,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8792792792792793,
+      "grad_norm": 11.066182980593581,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8828828828828829,
+      "grad_norm": 11.79957804045468,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8864864864864865,
+      "grad_norm": 14.945861266672521,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 2116
+    },
+    {
+      "epoch": 0.8900900900900901,
+      "grad_norm": 12.826083153138553,
+      "learning_rate": 1e-06,
+      "loss": 0.152,
+      "step": 2117
+    },
+    {
+      "epoch": 0.8936936936936937,
+      "grad_norm": 11.746837974646667,
+      "learning_rate": 1e-06,
+      "loss": 0.1578,
+      "step": 2118
+    },
+    {
+      "epoch": 0.8972972972972973,
+      "grad_norm": 12.352136489533713,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 11.726986426154118,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9045045045045045,
+      "grad_norm": 11.456473876030389,
+      "learning_rate": 1e-06,
+      "loss": 0.1553,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9081081081081082,
+      "grad_norm": 12.771148939508818,
+      "learning_rate": 1e-06,
+      "loss": 0.1534,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9117117117117117,
+      "grad_norm": 12.62916760044422,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9153153153153153,
+      "grad_norm": 9.278082776938705,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 2124
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 7.386602942247581,
+      "learning_rate": 1e-06,
+      "loss": 0.161,
+      "step": 2125
+    },
+    {
+      "epoch": 0.9225225225225225,
+      "grad_norm": 9.236977618767634,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9261261261261261,
+      "grad_norm": 9.735661406632168,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9297297297297298,
+      "grad_norm": 11.337102449709123,
+      "learning_rate": 1e-06,
+      "loss": 0.1567,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9333333333333333,
+      "grad_norm": 8.330352498451843,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 12.832919155135663,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9405405405405406,
+      "grad_norm": 9.411138061294565,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9441441441441442,
+      "grad_norm": 10.689227454619084,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9477477477477477,
+      "grad_norm": 8.528791055368094,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9513513513513514,
+      "grad_norm": 8.528901592882052,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 2134
+    },
+    {
+      "epoch": 0.954954954954955,
+      "grad_norm": 9.375540402068717,
+      "learning_rate": 1e-06,
+      "loss": 0.153,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9585585585585585,
+      "grad_norm": 10.339812470653325,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9621621621621622,
+      "grad_norm": 9.230263167330593,
+      "learning_rate": 1e-06,
+      "loss": 0.1463,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9657657657657658,
+      "grad_norm": 11.659014332169308,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9693693693693693,
+      "grad_norm": 12.213630756065621,
+      "learning_rate": 1e-06,
+      "loss": 0.1531,
+      "step": 2139
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 10.75932020598007,
+      "learning_rate": 1e-06,
+      "loss": 0.1509,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9765765765765766,
+      "grad_norm": 11.128969897515494,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9801801801801802,
+      "grad_norm": 8.339566649712399,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9837837837837838,
+      "grad_norm": 9.524657758115271,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9873873873873874,
+      "grad_norm": 14.501963316579891,
+      "learning_rate": 1e-06,
+      "loss": 0.1606,
+      "step": 2144
+    },
+    {
+      "epoch": 0.990990990990991,
+      "grad_norm": 10.132482187411922,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9945945945945946,
+      "grad_norm": 12.840391221986925,
+      "learning_rate": 1e-06,
+      "loss": 0.1664,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9981981981981982,
+      "grad_norm": 9.87794246038704,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 2147
+    },
+    {
+      "epoch": 0.9981981981981982,
+      "eval_loss": 3.264464855194092,
+      "eval_runtime": 21.4226,
+      "eval_samples_per_second": 46.68,
+      "eval_steps_per_second": 5.835,
+      "step": 2147
+    },
+    {
+      "epoch": 0.002628120893561104,
+      "grad_norm": 8.66023502603848,
+      "learning_rate": 1e-06,
+      "loss": 0.119,
+      "step": 2148
+    },
+    {
+      "epoch": 0.005256241787122208,
+      "grad_norm": 10.042821520266711,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 2149
+    },
+    {
+      "epoch": 0.00788436268068331,
+      "grad_norm": 10.600456304583386,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 2150
+    },
+    {
+      "epoch": 0.010512483574244415,
+      "grad_norm": 8.666755555804418,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 2151
+    },
+    {
+      "epoch": 0.013140604467805518,
+      "grad_norm": 9.30674708647994,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 2152
+    },
+    {
+      "epoch": 0.01576872536136662,
+      "grad_norm": 10.382446083129931,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 2153
+    },
+    {
+      "epoch": 0.018396846254927726,
+      "grad_norm": 10.174021034410611,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 2154
+    },
+    {
+      "epoch": 0.02102496714848883,
+      "grad_norm": 8.292979622829035,
+      "learning_rate": 1e-06,
+      "loss": 0.1277,
+      "step": 2155
+    },
+    {
+      "epoch": 0.023653088042049936,
+      "grad_norm": 8.495434760401114,
+      "learning_rate": 1e-06,
+      "loss": 0.1239,
+      "step": 2156
+    },
+    {
+      "epoch": 0.026281208935611037,
+      "grad_norm": 12.186717195924661,
+      "learning_rate": 1e-06,
+      "loss": 0.1369,
+      "step": 2157
+    },
+    {
+      "epoch": 0.02890932982917214,
+      "grad_norm": 7.476116182271973,
+      "learning_rate": 1e-06,
+      "loss": 0.118,
+      "step": 2158
+    },
+    {
+      "epoch": 0.03153745072273324,
+      "grad_norm": 10.013712824179157,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 2159
+    },
+    {
+      "epoch": 0.03416557161629435,
+      "grad_norm": 11.370506839039402,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2160
+    },
+    {
+      "epoch": 0.03679369250985545,
+      "grad_norm": 8.651656915558723,
+      "learning_rate": 1e-06,
+      "loss": 0.1263,
+      "step": 2161
+    },
+    {
+      "epoch": 0.03942181340341656,
+      "grad_norm": 7.611028422693545,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 2162
+    },
+    {
+      "epoch": 0.04204993429697766,
+      "grad_norm": 10.610442220157491,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 2163
+    },
+    {
+      "epoch": 0.04467805519053877,
+      "grad_norm": 8.931158631410513,
+      "learning_rate": 1e-06,
+      "loss": 0.1316,
+      "step": 2164
+    },
+    {
+      "epoch": 0.04730617608409987,
+      "grad_norm": 8.766585001755097,
+      "learning_rate": 1e-06,
+      "loss": 0.1282,
+      "step": 2165
+    },
+    {
+      "epoch": 0.04993429697766097,
+      "grad_norm": 12.648185715730204,
+      "learning_rate": 1e-06,
+      "loss": 0.138,
+      "step": 2166
+    },
+    {
+      "epoch": 0.052562417871222074,
+      "grad_norm": 14.45687688741498,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 2167
+    },
+    {
+      "epoch": 0.05519053876478318,
+      "grad_norm": 13.477939831711796,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 2168
+    },
+    {
+      "epoch": 0.05781865965834428,
+      "grad_norm": 12.383057543796154,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2169
+    },
+    {
+      "epoch": 0.06044678055190539,
+      "grad_norm": 11.375623377664084,
+      "learning_rate": 1e-06,
+      "loss": 0.1397,
+      "step": 2170
+    },
+    {
+      "epoch": 0.06307490144546649,
+      "grad_norm": 12.582985670684929,
+      "learning_rate": 1e-06,
+      "loss": 0.1265,
+      "step": 2171
+    },
+    {
+      "epoch": 0.0657030223390276,
+      "grad_norm": 9.089668186230083,
+      "learning_rate": 1e-06,
+      "loss": 0.137,
+      "step": 2172
+    },
+    {
+      "epoch": 0.0683311432325887,
+      "grad_norm": 11.15237578894327,
+      "learning_rate": 1e-06,
+      "loss": 0.1245,
+      "step": 2173
+    },
+    {
+      "epoch": 0.07095926412614981,
+      "grad_norm": 8.565138627406764,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 2174
+    },
+    {
+      "epoch": 0.0735873850197109,
+      "grad_norm": 8.58789543341393,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2175
+    },
+    {
+      "epoch": 0.07621550591327202,
+      "grad_norm": 10.216426658750372,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 2176
+    },
+    {
+      "epoch": 0.07884362680683311,
+      "grad_norm": 11.662378389540462,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 2177
+    },
+    {
+      "epoch": 0.08147174770039421,
+      "grad_norm": 9.5791670534219,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 2178
+    },
+    {
+      "epoch": 0.08409986859395532,
+      "grad_norm": 9.416715381291011,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 2179
+    },
+    {
+      "epoch": 0.08672798948751642,
+      "grad_norm": 12.73909355005036,
+      "learning_rate": 1e-06,
+      "loss": 0.1305,
+      "step": 2180
+    },
+    {
+      "epoch": 0.08935611038107753,
+      "grad_norm": 10.892661776960177,
+      "learning_rate": 1e-06,
+      "loss": 0.1409,
+      "step": 2181
+    },
+    {
+      "epoch": 0.09198423127463863,
+      "grad_norm": 10.754774304719529,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 2182
+    },
+    {
+      "epoch": 0.09461235216819974,
+      "grad_norm": 11.823842658657302,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 2183
+    },
+    {
+      "epoch": 0.09724047306176084,
+      "grad_norm": 11.864557356377038,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 2184
+    },
+    {
+      "epoch": 0.09986859395532194,
+      "grad_norm": 9.135837004713384,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 2185
+    },
+    {
+      "epoch": 0.10249671484888305,
+      "grad_norm": 12.224193815171576,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2186
+    },
+    {
+      "epoch": 0.10512483574244415,
+      "grad_norm": 6.3962379693116445,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 2187
+    },
+    {
+      "epoch": 0.10775295663600526,
+      "grad_norm": 11.568312391087982,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 2188
+    },
+    {
+      "epoch": 0.11038107752956636,
+      "grad_norm": 12.43689507773304,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 2189
+    },
+    {
+      "epoch": 0.11300919842312747,
+      "grad_norm": 12.188708384664745,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 2190
+    },
+    {
+      "epoch": 0.11563731931668857,
+      "grad_norm": 10.744077467034973,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 2191
+    },
+    {
+      "epoch": 0.11826544021024968,
+      "grad_norm": 14.060636194383648,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 2192
+    },
+    {
+      "epoch": 0.12089356110381078,
+      "grad_norm": 8.05810062724633,
+      "learning_rate": 1e-06,
+      "loss": 0.1257,
+      "step": 2193
+    },
+    {
+      "epoch": 0.12352168199737187,
+      "grad_norm": 13.485802542063306,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 2194
+    },
+    {
+      "epoch": 0.12614980289093297,
+      "grad_norm": 10.405539936898105,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2195
+    },
+    {
+      "epoch": 0.1287779237844941,
+      "grad_norm": 11.562479811627412,
+      "learning_rate": 1e-06,
+      "loss": 0.1432,
+      "step": 2196
+    },
+    {
+      "epoch": 0.1314060446780552,
+      "grad_norm": 12.12533722861644,
+      "learning_rate": 1e-06,
+      "loss": 0.1417,
+      "step": 2197
+    },
+    {
+      "epoch": 0.1340341655716163,
+      "grad_norm": 9.614662275384383,
+      "learning_rate": 1e-06,
+      "loss": 0.1245,
+      "step": 2198
+    },
+    {
+      "epoch": 0.1366622864651774,
+      "grad_norm": 8.561206040574946,
+      "learning_rate": 1e-06,
+      "loss": 0.1334,
+      "step": 2199
+    },
+    {
+      "epoch": 0.1392904073587385,
+      "grad_norm": 10.9111294465895,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 2200
+    },
+    {
+      "epoch": 0.14191852825229961,
+      "grad_norm": 8.914927276914831,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2201
+    },
+    {
+      "epoch": 0.1445466491458607,
+      "grad_norm": 8.985031284743238,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 2202
+    },
+    {
+      "epoch": 0.1471747700394218,
+      "grad_norm": 10.503874634731156,
+      "learning_rate": 1e-06,
+      "loss": 0.1291,
+      "step": 2203
+    },
+    {
+      "epoch": 0.1498028909329829,
+      "grad_norm": 12.075230665961133,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 2204
+    },
+    {
+      "epoch": 0.15243101182654403,
+      "grad_norm": 7.42931113439299,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 2205
+    },
+    {
+      "epoch": 0.15505913272010513,
+      "grad_norm": 10.214715929261907,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 2206
+    },
+    {
+      "epoch": 0.15768725361366623,
+      "grad_norm": 11.319739024559087,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 2207
+    },
+    {
+      "epoch": 0.16031537450722733,
+      "grad_norm": 10.596667513424208,
+      "learning_rate": 1e-06,
+      "loss": 0.1348,
+      "step": 2208
+    },
+    {
+      "epoch": 0.16294349540078842,
+      "grad_norm": 11.01996733081042,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2209
+    },
+    {
+      "epoch": 0.16557161629434955,
+      "grad_norm": 11.185018560901858,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 2210
+    },
+    {
+      "epoch": 0.16819973718791065,
+      "grad_norm": 12.860518841791054,
+      "learning_rate": 1e-06,
+      "loss": 0.1408,
+      "step": 2211
+    },
+    {
+      "epoch": 0.17082785808147175,
+      "grad_norm": 11.153638793340187,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 2212
+    },
+    {
+      "epoch": 0.17345597897503284,
+      "grad_norm": 11.169580187497397,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 2213
+    },
+    {
+      "epoch": 0.17608409986859397,
+      "grad_norm": 9.454295258236618,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 2214
+    },
+    {
+      "epoch": 0.17871222076215507,
+      "grad_norm": 9.753484403625075,
+      "learning_rate": 1e-06,
+      "loss": 0.1381,
+      "step": 2215
+    },
+    {
+      "epoch": 0.18134034165571616,
+      "grad_norm": 8.676549261076449,
+      "learning_rate": 1e-06,
+      "loss": 0.127,
+      "step": 2216
+    },
+    {
+      "epoch": 0.18396846254927726,
+      "grad_norm": 8.564808895546795,
+      "learning_rate": 1e-06,
+      "loss": 0.1381,
+      "step": 2217
+    },
+    {
+      "epoch": 0.18659658344283836,
+      "grad_norm": 10.692209050064168,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 2218
+    },
+    {
+      "epoch": 0.18922470433639949,
+      "grad_norm": 8.122617143380403,
+      "learning_rate": 1e-06,
+      "loss": 0.1285,
+      "step": 2219
+    },
+    {
+      "epoch": 0.19185282522996058,
+      "grad_norm": 13.350401734167566,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 2220
+    },
+    {
+      "epoch": 0.19448094612352168,
+      "grad_norm": 9.48894286025611,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 2221
+    },
+    {
+      "epoch": 0.19710906701708278,
+      "grad_norm": 10.612484712623571,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 2222
+    },
+    {
+      "epoch": 0.19973718791064388,
+      "grad_norm": 10.652476340067771,
+      "learning_rate": 1e-06,
+      "loss": 0.1282,
+      "step": 2223
+    },
+    {
+      "epoch": 0.202365308804205,
+      "grad_norm": 9.267080009427493,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2224
+    },
+    {
+      "epoch": 0.2049934296977661,
+      "grad_norm": 10.491123031382683,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 2225
+    },
+    {
+      "epoch": 0.2076215505913272,
+      "grad_norm": 10.246741631278121,
+      "learning_rate": 1e-06,
+      "loss": 0.1261,
+      "step": 2226
+    },
+    {
+      "epoch": 0.2102496714848883,
+      "grad_norm": 9.6101471165028,
+      "learning_rate": 1e-06,
+      "loss": 0.1305,
+      "step": 2227
+    },
+    {
+      "epoch": 0.21287779237844942,
+      "grad_norm": 11.36257597162339,
+      "learning_rate": 1e-06,
+      "loss": 0.1435,
+      "step": 2228
+    },
+    {
+      "epoch": 0.21550591327201052,
+      "grad_norm": 7.972851214642841,
+      "learning_rate": 1e-06,
+      "loss": 0.1332,
+      "step": 2229
+    },
+    {
+      "epoch": 0.21813403416557162,
+      "grad_norm": 10.15177104682382,
+      "learning_rate": 1e-06,
+      "loss": 0.1337,
+      "step": 2230
+    },
+    {
+      "epoch": 0.22076215505913271,
+      "grad_norm": 13.449581333912228,
+      "learning_rate": 1e-06,
+      "loss": 0.1432,
+      "step": 2231
+    },
+    {
+      "epoch": 0.2233902759526938,
+      "grad_norm": 8.941342057715419,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 2232
+    },
+    {
+      "epoch": 0.22601839684625494,
+      "grad_norm": 9.198564783835408,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 2233
+    },
+    {
+      "epoch": 0.22864651773981604,
+      "grad_norm": 9.472213248163277,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 2234
+    },
+    {
+      "epoch": 0.23127463863337713,
+      "grad_norm": 9.934766971687562,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2235
+    },
+    {
+      "epoch": 0.23390275952693823,
+      "grad_norm": 13.193124876472542,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 2236
+    },
+    {
+      "epoch": 0.23653088042049936,
+      "grad_norm": 13.193124876472542,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 2237
+    },
+    {
+      "epoch": 0.23915900131406045,
+      "grad_norm": 11.773769956349469,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 2238
+    },
+    {
+      "epoch": 0.24178712220762155,
+      "grad_norm": 12.729282171940811,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 2239
+    },
+    {
+      "epoch": 0.24441524310118265,
+      "grad_norm": 9.414726324019217,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 2240
+    },
+    {
+      "epoch": 0.24704336399474375,
+      "grad_norm": 8.927555193424665,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 2241
+    },
+    {
+      "epoch": 0.24967148488830487,
+      "grad_norm": 9.09634353451201,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 2242
+    },
+    {
+      "epoch": 0.25229960578186594,
+      "grad_norm": 10.151417042915778,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 2243
+    },
+    {
+      "epoch": 0.25492772667542707,
+      "grad_norm": 13.694608468077014,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 2244
+    },
+    {
+      "epoch": 0.2575558475689882,
+      "grad_norm": 12.563952821956944,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 2245
+    },
+    {
+      "epoch": 0.26018396846254926,
+      "grad_norm": 10.503409850245665,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 2246
+    },
+    {
+      "epoch": 0.2628120893561104,
+      "grad_norm": 10.384933515903219,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 2247
+    },
+    {
+      "epoch": 0.26544021024967146,
+      "grad_norm": 9.64348113758674,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 2248
+    },
+    {
+      "epoch": 0.2680683311432326,
+      "grad_norm": 10.743830098148102,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2249
+    },
+    {
+      "epoch": 0.2706964520367937,
+      "grad_norm": 9.613982211626528,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 2250
+    },
+    {
+      "epoch": 0.2733245729303548,
+      "grad_norm": 15.512590185084756,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2251
+    },
+    {
+      "epoch": 0.2759526938239159,
+      "grad_norm": 9.740682633230877,
+      "learning_rate": 1e-06,
+      "loss": 0.1212,
+      "step": 2252
+    },
+    {
+      "epoch": 0.278580814717477,
+      "grad_norm": 13.961420201180154,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2253
+    },
+    {
+      "epoch": 0.2812089356110381,
+      "grad_norm": 11.62377671847275,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2254
+    },
+    {
+      "epoch": 0.28383705650459923,
+      "grad_norm": 10.347350594802348,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 2255
+    },
+    {
+      "epoch": 0.2864651773981603,
+      "grad_norm": 11.60198724923247,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 2256
+    },
+    {
+      "epoch": 0.2890932982917214,
+      "grad_norm": 11.437443514932625,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 2257
+    },
+    {
+      "epoch": 0.29172141918528255,
+      "grad_norm": 11.225641222494513,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2258
+    },
+    {
+      "epoch": 0.2943495400788436,
+      "grad_norm": 10.112640306147389,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 2259
+    },
+    {
+      "epoch": 0.29697766097240474,
+      "grad_norm": 9.360911161985683,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 2260
+    },
+    {
+      "epoch": 0.2996057818659658,
+      "grad_norm": 9.785011443298721,
+      "learning_rate": 1e-06,
+      "loss": 0.1288,
+      "step": 2261
+    },
+    {
+      "epoch": 0.30223390275952694,
+      "grad_norm": 12.537247058503421,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 2262
+    },
+    {
+      "epoch": 0.30486202365308807,
+      "grad_norm": 8.044309697854088,
+      "learning_rate": 1e-06,
+      "loss": 0.1232,
+      "step": 2263
+    },
+    {
+      "epoch": 0.30749014454664914,
+      "grad_norm": 9.652939701705147,
+      "learning_rate": 1e-06,
+      "loss": 0.1287,
+      "step": 2264
+    },
+    {
+      "epoch": 0.31011826544021026,
+      "grad_norm": 10.848495060607336,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 2265
+    },
+    {
+      "epoch": 0.31274638633377133,
+      "grad_norm": 10.450002593650165,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 2266
+    },
+    {
+      "epoch": 0.31537450722733246,
+      "grad_norm": 10.613262025192162,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 2267
+    },
+    {
+      "epoch": 0.3180026281208936,
+      "grad_norm": 10.947143330112132,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 2268
+    },
+    {
+      "epoch": 0.32063074901445465,
+      "grad_norm": 9.506623783742052,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2269
+    },
+    {
+      "epoch": 0.3232588699080158,
+      "grad_norm": 8.093625855316503,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 2270
+    },
+    {
+      "epoch": 0.32588699080157685,
+      "grad_norm": 10.09629749108699,
+      "learning_rate": 1e-06,
+      "loss": 0.1323,
+      "step": 2271
+    },
+    {
+      "epoch": 0.328515111695138,
+      "grad_norm": 11.66455649547995,
+      "learning_rate": 1e-06,
+      "loss": 0.1408,
+      "step": 2272
+    },
+    {
+      "epoch": 0.3311432325886991,
+      "grad_norm": 9.370236751256742,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 2273
+    },
+    {
+      "epoch": 0.33377135348226017,
+      "grad_norm": 11.390763700880935,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 2274
+    },
+    {
+      "epoch": 0.3363994743758213,
+      "grad_norm": 10.284156382402742,
+      "learning_rate": 1e-06,
+      "loss": 0.1399,
+      "step": 2275
+    },
+    {
+      "epoch": 0.33902759526938236,
+      "grad_norm": 11.343283386135061,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 2276
+    },
+    {
+      "epoch": 0.3416557161629435,
+      "grad_norm": 12.11905246473164,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 2277
+    },
+    {
+      "epoch": 0.3442838370565046,
+      "grad_norm": 9.494252818475086,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 2278
+    },
+    {
+      "epoch": 0.3469119579500657,
+      "grad_norm": 9.814520820483848,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2279
+    },
+    {
+      "epoch": 0.3495400788436268,
+      "grad_norm": 14.220822494797224,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 2280
+    },
+    {
+      "epoch": 0.35216819973718794,
+      "grad_norm": 10.661083823473497,
+      "learning_rate": 1e-06,
+      "loss": 0.1391,
+      "step": 2281
+    },
+    {
+      "epoch": 0.354796320630749,
+      "grad_norm": 11.589076051805208,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 2282
+    },
+    {
+      "epoch": 0.35742444152431013,
+      "grad_norm": 9.97998698164963,
+      "learning_rate": 1e-06,
+      "loss": 0.1296,
+      "step": 2283
+    },
+    {
+      "epoch": 0.3600525624178712,
+      "grad_norm": 11.97183009925302,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 2284
+    },
+    {
+      "epoch": 0.36268068331143233,
+      "grad_norm": 12.13196513575946,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 2285
+    },
+    {
+      "epoch": 0.36530880420499345,
+      "grad_norm": 10.901166607668754,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 2286
+    },
+    {
+      "epoch": 0.3679369250985545,
+      "grad_norm": 8.23719418838252,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2287
+    },
+    {
+      "epoch": 0.37056504599211565,
+      "grad_norm": 9.777629506743795,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 2288
+    },
+    {
+      "epoch": 0.3731931668856767,
+      "grad_norm": 14.413479349392164,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 2289
+    },
+    {
+      "epoch": 0.37582128777923784,
+      "grad_norm": 13.66405276123145,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2290
+    },
+    {
+      "epoch": 0.37844940867279897,
+      "grad_norm": 11.753333593121297,
+      "learning_rate": 1e-06,
+      "loss": 0.1451,
+      "step": 2291
+    },
+    {
+      "epoch": 0.38107752956636004,
+      "grad_norm": 11.070204706346967,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2292
+    },
+    {
+      "epoch": 0.38370565045992117,
+      "grad_norm": 13.032828543655024,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 2293
+    },
+    {
+      "epoch": 0.38633377135348224,
+      "grad_norm": 9.521561504516228,
+      "learning_rate": 1e-06,
+      "loss": 0.1461,
+      "step": 2294
+    },
+    {
+      "epoch": 0.38896189224704336,
+      "grad_norm": 12.064134590958243,
+      "learning_rate": 1e-06,
+      "loss": 0.1488,
+      "step": 2295
+    },
+    {
+      "epoch": 0.3915900131406045,
+      "grad_norm": 11.199488671796828,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 2296
+    },
+    {
+      "epoch": 0.39421813403416556,
+      "grad_norm": 17.15588122268286,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 2297
+    },
+    {
+      "epoch": 0.3968462549277267,
+      "grad_norm": 8.397428253213143,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 2298
+    },
+    {
+      "epoch": 0.39947437582128775,
+      "grad_norm": 10.850700938567497,
+      "learning_rate": 1e-06,
+      "loss": 0.1288,
+      "step": 2299
+    },
+    {
+      "epoch": 0.4021024967148489,
+      "grad_norm": 12.053508717926714,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2300
+    },
+    {
+      "epoch": 0.40473061760841,
+      "grad_norm": 9.042559802283957,
+      "learning_rate": 1e-06,
+      "loss": 0.1306,
+      "step": 2301
+    },
+    {
+      "epoch": 0.4073587385019711,
+      "grad_norm": 10.864613737919637,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 2302
+    },
+    {
+      "epoch": 0.4099868593955322,
+      "grad_norm": 10.69736602016788,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 2303
+    },
+    {
+      "epoch": 0.4126149802890933,
+      "grad_norm": 9.97460799994981,
+      "learning_rate": 1e-06,
+      "loss": 0.1449,
+      "step": 2304
+    },
+    {
+      "epoch": 0.4152431011826544,
+      "grad_norm": 10.578889401994457,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 2305
+    },
+    {
+      "epoch": 0.4178712220762155,
+      "grad_norm": 9.433652910176368,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2306
+    },
+    {
+      "epoch": 0.4204993429697766,
+      "grad_norm": 11.488216580014672,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2307
+    },
+    {
+      "epoch": 0.4231274638633377,
+      "grad_norm": 7.645115566265672,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 2308
+    },
+    {
+      "epoch": 0.42575558475689884,
+      "grad_norm": 11.23108587919389,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 2309
+    },
+    {
+      "epoch": 0.4283837056504599,
+      "grad_norm": 9.350142985701256,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 2310
+    },
+    {
+      "epoch": 0.43101182654402104,
+      "grad_norm": 10.788288935553709,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 2311
+    },
+    {
+      "epoch": 0.4336399474375821,
+      "grad_norm": 8.837035354359816,
+      "learning_rate": 1e-06,
+      "loss": 0.1407,
+      "step": 2312
+    },
+    {
+      "epoch": 0.43626806833114323,
+      "grad_norm": 9.0126306643432,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 2313
+    },
+    {
+      "epoch": 0.43889618922470436,
+      "grad_norm": 14.322215317079594,
+      "learning_rate": 1e-06,
+      "loss": 0.1349,
+      "step": 2314
+    },
+    {
+      "epoch": 0.44152431011826543,
+      "grad_norm": 7.954795160050442,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 2315
+    },
+    {
+      "epoch": 0.44415243101182655,
+      "grad_norm": 13.66173036612331,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 2316
+    },
+    {
+      "epoch": 0.4467805519053876,
+      "grad_norm": 11.46384679640742,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 2317
+    },
+    {
+      "epoch": 0.44940867279894875,
+      "grad_norm": 16.90914904718049,
+      "learning_rate": 1e-06,
+      "loss": 0.1427,
+      "step": 2318
+    },
+    {
+      "epoch": 0.4520367936925099,
+      "grad_norm": 14.561980818772387,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 2319
+    },
+    {
+      "epoch": 0.45466491458607095,
+      "grad_norm": 11.49467517264891,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 2320
+    },
+    {
+      "epoch": 0.45729303547963207,
+      "grad_norm": 10.639670966314034,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 2321
+    },
+    {
+      "epoch": 0.45992115637319314,
+      "grad_norm": 16.166419889078576,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2322
+    },
+    {
+      "epoch": 0.46254927726675427,
+      "grad_norm": 13.255201575923367,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 2323
+    },
+    {
+      "epoch": 0.4651773981603154,
+      "grad_norm": 9.760417689482106,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 2324
+    },
+    {
+      "epoch": 0.46780551905387646,
+      "grad_norm": 10.572350787075333,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 2325
+    },
+    {
+      "epoch": 0.4704336399474376,
+      "grad_norm": 11.38255060516214,
+      "learning_rate": 1e-06,
+      "loss": 0.1356,
+      "step": 2326
+    },
+    {
+      "epoch": 0.4730617608409987,
+      "grad_norm": 11.475789686167829,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2327
+    },
+    {
+      "epoch": 0.4756898817345598,
+      "grad_norm": 10.806052800943826,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 2328
+    },
+    {
+      "epoch": 0.4783180026281209,
+      "grad_norm": 10.241189869114853,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 2329
+    },
+    {
+      "epoch": 0.480946123521682,
+      "grad_norm": 9.854429174917238,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 2330
+    },
+    {
+      "epoch": 0.4835742444152431,
+      "grad_norm": 8.208550490305235,
+      "learning_rate": 1e-06,
+      "loss": 0.1322,
+      "step": 2331
+    },
+    {
+      "epoch": 0.48620236530880423,
+      "grad_norm": 11.271018869647376,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 2332
+    },
+    {
+      "epoch": 0.4888304862023653,
+      "grad_norm": 8.611947986940509,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 2333
+    },
+    {
+      "epoch": 0.4914586070959264,
+      "grad_norm": 12.000479470895952,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2334
+    },
+    {
+      "epoch": 0.4940867279894875,
+      "grad_norm": 10.683666579370515,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2335
+    },
+    {
+      "epoch": 0.4967148488830486,
+      "grad_norm": 12.417258661467603,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 2336
+    },
+    {
+      "epoch": 0.49934296977660975,
+      "grad_norm": 10.215761462404389,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 2337
+    },
+    {
+      "epoch": 0.5019710906701709,
+      "grad_norm": 8.02653206543822,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 2338
+    },
+    {
+      "epoch": 0.5045992115637319,
+      "grad_norm": 13.219137859831372,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 2339
+    },
+    {
+      "epoch": 0.507227332457293,
+      "grad_norm": 8.36978046929735,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2340
+    },
+    {
+      "epoch": 0.5098554533508541,
+      "grad_norm": 10.547008163369801,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 2341
+    },
+    {
+      "epoch": 0.5124835742444153,
+      "grad_norm": 12.609804653980495,
+      "learning_rate": 1e-06,
+      "loss": 0.1372,
+      "step": 2342
+    },
+    {
+      "epoch": 0.5151116951379764,
+      "grad_norm": 9.734728041213305,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 2343
+    },
+    {
+      "epoch": 0.5177398160315374,
+      "grad_norm": 11.243018763029148,
+      "learning_rate": 1e-06,
+      "loss": 0.1351,
+      "step": 2344
+    },
+    {
+      "epoch": 0.5203679369250985,
+      "grad_norm": 11.847878641145405,
+      "learning_rate": 1e-06,
+      "loss": 0.1376,
+      "step": 2345
+    },
+    {
+      "epoch": 0.5229960578186597,
+      "grad_norm": 12.656455376476053,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 2346
+    },
+    {
+      "epoch": 0.5256241787122208,
+      "grad_norm": 9.922032747366332,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2347
+    },
+    {
+      "epoch": 0.5282522996057819,
+      "grad_norm": 9.981132616795273,
+      "learning_rate": 1e-06,
+      "loss": 0.1456,
+      "step": 2348
+    },
+    {
+      "epoch": 0.5308804204993429,
+      "grad_norm": 10.387919997144357,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 2349
+    },
+    {
+      "epoch": 0.533508541392904,
+      "grad_norm": 14.448500173292045,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 2350
+    },
+    {
+      "epoch": 0.5361366622864652,
+      "grad_norm": 12.227087305824933,
+      "learning_rate": 1e-06,
+      "loss": 0.1379,
+      "step": 2351
+    },
+    {
+      "epoch": 0.5387647831800263,
+      "grad_norm": 10.467610659057884,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 2352
+    },
+    {
+      "epoch": 0.5413929040735874,
+      "grad_norm": 11.210813308735723,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 2353
+    },
+    {
+      "epoch": 0.5440210249671484,
+      "grad_norm": 9.981521032003435,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 2354
+    },
+    {
+      "epoch": 0.5466491458607096,
+      "grad_norm": 10.714673159279478,
+      "learning_rate": 1e-06,
+      "loss": 0.1146,
+      "step": 2355
+    },
+    {
+      "epoch": 0.5492772667542707,
+      "grad_norm": 10.709807019559461,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2356
+    },
+    {
+      "epoch": 0.5519053876478318,
+      "grad_norm": 10.748107705185888,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 2357
+    },
+    {
+      "epoch": 0.5545335085413929,
+      "grad_norm": 8.622750344380945,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 2358
+    },
+    {
+      "epoch": 0.557161629434954,
+      "grad_norm": 12.660965683761203,
+      "learning_rate": 1e-06,
+      "loss": 0.1279,
+      "step": 2359
+    },
+    {
+      "epoch": 0.5597897503285151,
+      "grad_norm": 9.169299268101755,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 2360
+    },
+    {
+      "epoch": 0.5624178712220762,
+      "grad_norm": 9.852267755058115,
+      "learning_rate": 1e-06,
+      "loss": 0.1309,
+      "step": 2361
+    },
+    {
+      "epoch": 0.5650459921156373,
+      "grad_norm": 13.0555507008127,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 2362
+    },
+    {
+      "epoch": 0.5676741130091985,
+      "grad_norm": 10.820231814182609,
+      "learning_rate": 1e-06,
+      "loss": 0.1401,
+      "step": 2363
+    },
+    {
+      "epoch": 0.5703022339027596,
+      "grad_norm": 11.435149453214,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2364
+    },
+    {
+      "epoch": 0.5729303547963206,
+      "grad_norm": 10.266979447836816,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 2365
+    },
+    {
+      "epoch": 0.5755584756898817,
+      "grad_norm": 10.681506123574842,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 2366
+    },
+    {
+      "epoch": 0.5781865965834428,
+      "grad_norm": 10.598206820400492,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 2367
+    },
+    {
+      "epoch": 0.580814717477004,
+      "grad_norm": 9.44838605746558,
+      "learning_rate": 1e-06,
+      "loss": 0.1388,
+      "step": 2368
+    },
+    {
+      "epoch": 0.5834428383705651,
+      "grad_norm": 12.32385753001722,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 2369
+    },
+    {
+      "epoch": 0.5860709592641261,
+      "grad_norm": 10.662463203067576,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 2370
+    },
+    {
+      "epoch": 0.5886990801576872,
+      "grad_norm": 11.706049948444969,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2371
+    },
+    {
+      "epoch": 0.5913272010512484,
+      "grad_norm": 11.22198422028852,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 2372
+    },
+    {
+      "epoch": 0.5939553219448095,
+      "grad_norm": 12.036866484327966,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 2373
+    },
+    {
+      "epoch": 0.5965834428383706,
+      "grad_norm": 9.602931005031325,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 2374
+    },
+    {
+      "epoch": 0.5992115637319316,
+      "grad_norm": 11.120617030406766,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2375
+    },
+    {
+      "epoch": 0.6018396846254928,
+      "grad_norm": 10.499560373324199,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 2376
+    },
+    {
+      "epoch": 0.6044678055190539,
+      "grad_norm": 13.740203958822015,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 2377
+    },
+    {
+      "epoch": 0.607095926412615,
+      "grad_norm": 11.028657921129904,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2378
+    },
+    {
+      "epoch": 0.6097240473061761,
+      "grad_norm": 10.971593617503599,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 2379
+    },
+    {
+      "epoch": 0.6123521681997371,
+      "grad_norm": 10.446214100839864,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 2380
+    },
+    {
+      "epoch": 0.6149802890932983,
+      "grad_norm": 11.827918426252754,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 2381
+    },
+    {
+      "epoch": 0.6176084099868594,
+      "grad_norm": 11.139376959933449,
+      "learning_rate": 1e-06,
+      "loss": 0.1475,
+      "step": 2382
+    },
+    {
+      "epoch": 0.6202365308804205,
+      "grad_norm": 10.875099990421184,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 2383
+    },
+    {
+      "epoch": 0.6228646517739816,
+      "grad_norm": 7.2192368368380215,
+      "learning_rate": 1e-06,
+      "loss": 0.1307,
+      "step": 2384
+    },
+    {
+      "epoch": 0.6254927726675427,
+      "grad_norm": 9.142090941827748,
+      "learning_rate": 1e-06,
+      "loss": 0.1446,
+      "step": 2385
+    },
+    {
+      "epoch": 0.6281208935611038,
+      "grad_norm": 12.273649932217202,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 2386
+    },
+    {
+      "epoch": 0.6307490144546649,
+      "grad_norm": 13.980447637016308,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 2387
+    },
+    {
+      "epoch": 0.633377135348226,
+      "grad_norm": 10.268343181059858,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 2388
+    },
+    {
+      "epoch": 0.6360052562417872,
+      "grad_norm": 10.815453357390773,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 2389
+    },
+    {
+      "epoch": 0.6386333771353482,
+      "grad_norm": 11.142681245426962,
+      "learning_rate": 1e-06,
+      "loss": 0.1395,
+      "step": 2390
+    },
+    {
+      "epoch": 0.6412614980289093,
+      "grad_norm": 10.471017873519706,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 2391
+    },
+    {
+      "epoch": 0.6438896189224704,
+      "grad_norm": 8.081983541829054,
+      "learning_rate": 1e-06,
+      "loss": 0.152,
+      "step": 2392
+    },
+    {
+      "epoch": 0.6465177398160316,
+      "grad_norm": 10.216749956841468,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 2393
+    },
+    {
+      "epoch": 0.6491458607095927,
+      "grad_norm": 10.262609521881327,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 2394
+    },
+    {
+      "epoch": 0.6517739816031537,
+      "grad_norm": 9.848798204833736,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2395
+    },
+    {
+      "epoch": 0.6544021024967148,
+      "grad_norm": 10.777609536558279,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 2396
+    },
+    {
+      "epoch": 0.657030223390276,
+      "grad_norm": 12.767767428933158,
+      "learning_rate": 1e-06,
+      "loss": 0.1565,
+      "step": 2397
+    },
+    {
+      "epoch": 0.6596583442838371,
+      "grad_norm": 10.037066105228597,
+      "learning_rate": 1e-06,
+      "loss": 0.1316,
+      "step": 2398
+    },
+    {
+      "epoch": 0.6622864651773982,
+      "grad_norm": 9.695075552863901,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 2399
+    },
+    {
+      "epoch": 0.6649145860709592,
+      "grad_norm": 9.514780708029537,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 2400
+    },
+    {
+      "epoch": 0.6675427069645203,
+      "grad_norm": 11.083833157557098,
+      "learning_rate": 1e-06,
+      "loss": 0.1352,
+      "step": 2401
+    },
+    {
+      "epoch": 0.6701708278580815,
+      "grad_norm": 8.759974909212557,
+      "learning_rate": 1e-06,
+      "loss": 0.1303,
+      "step": 2402
+    },
+    {
+      "epoch": 0.6727989487516426,
+      "grad_norm": 10.548623904333379,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 2403
+    },
+    {
+      "epoch": 0.6754270696452037,
+      "grad_norm": 8.590791366465396,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2404
+    },
+    {
+      "epoch": 0.6780551905387647,
+      "grad_norm": 9.979393877474594,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 2405
+    },
+    {
+      "epoch": 0.6806833114323259,
+      "grad_norm": 10.17688617683912,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2406
+    },
+    {
+      "epoch": 0.683311432325887,
+      "grad_norm": 11.360449948237743,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 2407
+    },
+    {
+      "epoch": 0.6859395532194481,
+      "grad_norm": 12.9901651036432,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 2408
+    },
+    {
+      "epoch": 0.6885676741130092,
+      "grad_norm": 10.174939299091701,
+      "learning_rate": 1e-06,
+      "loss": 0.1456,
+      "step": 2409
+    },
+    {
+      "epoch": 0.6911957950065704,
+      "grad_norm": 10.84402660925017,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2410
+    },
+    {
+      "epoch": 0.6938239159001314,
+      "grad_norm": 11.444968988415253,
+      "learning_rate": 1e-06,
+      "loss": 0.1381,
+      "step": 2411
+    },
+    {
+      "epoch": 0.6964520367936925,
+      "grad_norm": 12.562802277723735,
+      "learning_rate": 1e-06,
+      "loss": 0.1381,
+      "step": 2412
+    },
+    {
+      "epoch": 0.6990801576872536,
+      "grad_norm": 11.8953264228649,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 2413
+    },
+    {
+      "epoch": 0.7017082785808147,
+      "grad_norm": 13.333062984275148,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 2414
+    },
+    {
+      "epoch": 0.7043363994743759,
+      "grad_norm": 11.689367020002251,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 2415
+    },
+    {
+      "epoch": 0.7069645203679369,
+      "grad_norm": 10.70306491767507,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 2416
+    },
+    {
+      "epoch": 0.709592641261498,
+      "grad_norm": 10.13368582376551,
+      "learning_rate": 1e-06,
+      "loss": 0.1363,
+      "step": 2417
+    },
+    {
+      "epoch": 0.7122207621550591,
+      "grad_norm": 10.093144467846342,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 2418
+    },
+    {
+      "epoch": 0.7148488830486203,
+      "grad_norm": 11.656177997366726,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 2419
+    },
+    {
+      "epoch": 0.7174770039421814,
+      "grad_norm": 10.084216433172559,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 2420
+    },
+    {
+      "epoch": 0.7201051248357424,
+      "grad_norm": 9.580285435731646,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 2421
+    },
+    {
+      "epoch": 0.7227332457293035,
+      "grad_norm": 10.807827558363233,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2422
+    },
+    {
+      "epoch": 0.7253613666228647,
+      "grad_norm": 12.334872638039263,
+      "learning_rate": 1e-06,
+      "loss": 0.1349,
+      "step": 2423
+    },
+    {
+      "epoch": 0.7279894875164258,
+      "grad_norm": 8.404238196979755,
+      "learning_rate": 1e-06,
+      "loss": 0.1532,
+      "step": 2424
+    },
+    {
+      "epoch": 0.7306176084099869,
+      "grad_norm": 10.686841970606633,
+      "learning_rate": 1e-06,
+      "loss": 0.1222,
+      "step": 2425
+    },
+    {
+      "epoch": 0.7332457293035479,
+      "grad_norm": 11.80501462623101,
+      "learning_rate": 1e-06,
+      "loss": 0.1399,
+      "step": 2426
+    },
+    {
+      "epoch": 0.735873850197109,
+      "grad_norm": 12.089261114404854,
+      "learning_rate": 1e-06,
+      "loss": 0.1374,
+      "step": 2427
+    },
+    {
+      "epoch": 0.7385019710906702,
+      "grad_norm": 8.385679953869964,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 2428
+    },
+    {
+      "epoch": 0.7411300919842313,
+      "grad_norm": 12.245312952801351,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2429
+    },
+    {
+      "epoch": 0.7437582128777924,
+      "grad_norm": 12.37808457794941,
+      "learning_rate": 1e-06,
+      "loss": 0.137,
+      "step": 2430
+    },
+    {
+      "epoch": 0.7463863337713534,
+      "grad_norm": 15.239739312397182,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 2431
+    },
+    {
+      "epoch": 0.7490144546649146,
+      "grad_norm": 10.46480787333769,
+      "learning_rate": 1e-06,
+      "loss": 0.1232,
+      "step": 2432
+    },
+    {
+      "epoch": 0.7516425755584757,
+      "grad_norm": 10.130215745553816,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 2433
+    },
+    {
+      "epoch": 0.7542706964520368,
+      "grad_norm": 11.413706919699354,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2434
+    },
+    {
+      "epoch": 0.7568988173455979,
+      "grad_norm": 12.78039121173657,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 2435
+    },
+    {
+      "epoch": 0.759526938239159,
+      "grad_norm": 10.856867491512512,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 2436
+    },
+    {
+      "epoch": 0.7621550591327201,
+      "grad_norm": 9.198277449487104,
+      "learning_rate": 1e-06,
+      "loss": 0.1337,
+      "step": 2437
+    },
+    {
+      "epoch": 0.7647831800262812,
+      "grad_norm": 12.892211625033614,
+      "learning_rate": 1e-06,
+      "loss": 0.146,
+      "step": 2438
+    },
+    {
+      "epoch": 0.7674113009198423,
+      "grad_norm": 10.750360450642159,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 2439
+    },
+    {
+      "epoch": 0.7700394218134035,
+      "grad_norm": 8.729448295126687,
+      "learning_rate": 1e-06,
+      "loss": 0.1305,
+      "step": 2440
+    },
+    {
+      "epoch": 0.7726675427069645,
+      "grad_norm": 11.011306845941949,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 2441
+    },
+    {
+      "epoch": 0.7752956636005256,
+      "grad_norm": 9.565287083684431,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 2442
+    },
+    {
+      "epoch": 0.7779237844940867,
+      "grad_norm": 9.037200150266806,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 2443
+    },
+    {
+      "epoch": 0.7805519053876478,
+      "grad_norm": 9.866609586549364,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 2444
+    },
+    {
+      "epoch": 0.783180026281209,
+      "grad_norm": 11.718283205967555,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 2445
+    },
+    {
+      "epoch": 0.78580814717477,
+      "grad_norm": 10.928058388278936,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 2446
+    },
+    {
+      "epoch": 0.7884362680683311,
+      "grad_norm": 11.210717406961848,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 2447
+    },
+    {
+      "epoch": 0.7910643889618922,
+      "grad_norm": 10.40459025148217,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 2448
+    },
+    {
+      "epoch": 0.7936925098554534,
+      "grad_norm": 9.268831023526202,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 2449
+    },
+    {
+      "epoch": 0.7963206307490145,
+      "grad_norm": 9.614043789860027,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 2450
+    },
+    {
+      "epoch": 0.7989487516425755,
+      "grad_norm": 10.390287759894822,
+      "learning_rate": 1e-06,
+      "loss": 0.1393,
+      "step": 2451
+    },
+    {
+      "epoch": 0.8015768725361366,
+      "grad_norm": 10.310269787856724,
+      "learning_rate": 1e-06,
+      "loss": 0.1404,
+      "step": 2452
+    },
+    {
+      "epoch": 0.8042049934296978,
+      "grad_norm": 11.859152108063954,
+      "learning_rate": 1e-06,
+      "loss": 0.1452,
+      "step": 2453
+    },
+    {
+      "epoch": 0.8068331143232589,
+      "grad_norm": 11.840258945440649,
+      "learning_rate": 1e-06,
+      "loss": 0.1425,
+      "step": 2454
+    },
+    {
+      "epoch": 0.80946123521682,
+      "grad_norm": 12.617029308585563,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 2455
+    },
+    {
+      "epoch": 0.812089356110381,
+      "grad_norm": 10.74567693248533,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 2456
+    },
+    {
+      "epoch": 0.8147174770039421,
+      "grad_norm": 9.889021974419027,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 2457
+    },
+    {
+      "epoch": 0.8173455978975033,
+      "grad_norm": 9.554030152284392,
+      "learning_rate": 1e-06,
+      "loss": 0.1374,
+      "step": 2458
+    },
+    {
+      "epoch": 0.8199737187910644,
+      "grad_norm": 12.319196511359214,
+      "learning_rate": 1e-06,
+      "loss": 0.1311,
+      "step": 2459
+    },
+    {
+      "epoch": 0.8226018396846255,
+      "grad_norm": 8.232492283155748,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 2460
+    },
+    {
+      "epoch": 0.8252299605781866,
+      "grad_norm": 8.622498858152198,
+      "learning_rate": 1e-06,
+      "loss": 0.1391,
+      "step": 2461
+    },
+    {
+      "epoch": 0.8278580814717477,
+      "grad_norm": 10.521084107425516,
+      "learning_rate": 1e-06,
+      "loss": 0.1435,
+      "step": 2462
+    },
+    {
+      "epoch": 0.8304862023653088,
+      "grad_norm": 8.980468746655502,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 2463
+    },
+    {
+      "epoch": 0.8331143232588699,
+      "grad_norm": 10.616578197039482,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 2464
+    },
+    {
+      "epoch": 0.835742444152431,
+      "grad_norm": 14.035430805850368,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 2465
+    },
+    {
+      "epoch": 0.8383705650459922,
+      "grad_norm": 10.362963869525924,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 2466
+    },
+    {
+      "epoch": 0.8409986859395532,
+      "grad_norm": 7.940366041242493,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 2467
+    },
+    {
+      "epoch": 0.8436268068331143,
+      "grad_norm": 12.23866470814424,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 2468
+    },
+    {
+      "epoch": 0.8462549277266754,
+      "grad_norm": 13.02008804940414,
+      "learning_rate": 1e-06,
+      "loss": 0.1397,
+      "step": 2469
+    },
+    {
+      "epoch": 0.8488830486202366,
+      "grad_norm": 11.18271068639763,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2470
+    },
+    {
+      "epoch": 0.8515111695137977,
+      "grad_norm": 12.818585227061071,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2471
+    },
+    {
+      "epoch": 0.8541392904073587,
+      "grad_norm": 9.992040882170153,
+      "learning_rate": 1e-06,
+      "loss": 0.1374,
+      "step": 2472
+    },
+    {
+      "epoch": 0.8567674113009198,
+      "grad_norm": 9.311942448362634,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2473
+    },
+    {
+      "epoch": 0.859395532194481,
+      "grad_norm": 10.692248038586095,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 2474
+    },
+    {
+      "epoch": 0.8620236530880421,
+      "grad_norm": 8.16373276010391,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2475
+    },
+    {
+      "epoch": 0.8646517739816032,
+      "grad_norm": 11.815950579546818,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 2476
+    },
+    {
+      "epoch": 0.8672798948751642,
+      "grad_norm": 12.41214423382949,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 2477
+    },
+    {
+      "epoch": 0.8699080157687253,
+      "grad_norm": 10.95507250050492,
+      "learning_rate": 1e-06,
+      "loss": 0.1382,
+      "step": 2478
+    },
+    {
+      "epoch": 0.8725361366622865,
+      "grad_norm": 11.062892135028147,
+      "learning_rate": 1e-06,
+      "loss": 0.1376,
+      "step": 2479
+    },
+    {
+      "epoch": 0.8751642575558476,
+      "grad_norm": 8.982455386796868,
+      "learning_rate": 1e-06,
+      "loss": 0.1371,
+      "step": 2480
+    },
+    {
+      "epoch": 0.8777923784494087,
+      "grad_norm": 11.502352767939149,
+      "learning_rate": 1e-06,
+      "loss": 0.1426,
+      "step": 2481
+    },
+    {
+      "epoch": 0.8804204993429697,
+      "grad_norm": 11.223517175968437,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 2482
+    },
+    {
+      "epoch": 0.8830486202365309,
+      "grad_norm": 9.550120842027745,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2483
+    },
+    {
+      "epoch": 0.885676741130092,
+      "grad_norm": 8.383274381665883,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 2484
+    },
+    {
+      "epoch": 0.8883048620236531,
+      "grad_norm": 11.002675465859719,
+      "learning_rate": 1e-06,
+      "loss": 0.1386,
+      "step": 2485
+    },
+    {
+      "epoch": 0.8909329829172142,
+      "grad_norm": 10.138111430735657,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 2486
+    },
+    {
+      "epoch": 0.8935611038107752,
+      "grad_norm": 8.35065745130803,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2487
+    },
+    {
+      "epoch": 0.8961892247043364,
+      "grad_norm": 9.352239766788415,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 2488
+    },
+    {
+      "epoch": 0.8988173455978975,
+      "grad_norm": 12.00330711043625,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9014454664914586,
+      "grad_norm": 11.032503655531812,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9040735873850198,
+      "grad_norm": 8.368707606070105,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9067017082785808,
+      "grad_norm": 11.270194113996661,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9093298291721419,
+      "grad_norm": 8.857523191460334,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 2493
+    },
+    {
+      "epoch": 0.911957950065703,
+      "grad_norm": 12.38950165042773,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9145860709592641,
+      "grad_norm": 16.167292536255914,
+      "learning_rate": 1e-06,
+      "loss": 0.1512,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9172141918528253,
+      "grad_norm": 9.21450650356582,
+      "learning_rate": 1e-06,
+      "loss": 0.1355,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9198423127463863,
+      "grad_norm": 11.811502734864325,
+      "learning_rate": 1e-06,
+      "loss": 0.1346,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9224704336399474,
+      "grad_norm": 8.211512306677557,
+      "learning_rate": 1e-06,
+      "loss": 0.1472,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9250985545335085,
+      "grad_norm": 10.994414258418459,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9277266754270697,
+      "grad_norm": 12.047410637794883,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9303547963206308,
+      "grad_norm": 9.796529117314206,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9329829172141918,
+      "grad_norm": 11.144598936300495,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9356110381077529,
+      "grad_norm": 9.824709888054205,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 2503
+    },
+    {
+      "epoch": 0.938239159001314,
+      "grad_norm": 10.384790965182612,
+      "learning_rate": 1e-06,
+      "loss": 0.1477,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9408672798948752,
+      "grad_norm": 10.441729916132797,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 2505
+    },
+    {
+      "epoch": 0.9434954007884363,
+      "grad_norm": 12.472200640854513,
+      "learning_rate": 1e-06,
+      "loss": 0.1507,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9461235216819974,
+      "grad_norm": 10.531631488794494,
+      "learning_rate": 1e-06,
+      "loss": 0.1418,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9487516425755584,
+      "grad_norm": 12.639792659360097,
+      "learning_rate": 1e-06,
+      "loss": 0.1337,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9513797634691196,
+      "grad_norm": 11.827314052582931,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9540078843626807,
+      "grad_norm": 10.633207667948083,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2510
+    },
+    {
+      "epoch": 0.9566360052562418,
+      "grad_norm": 9.956144438652887,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9592641261498029,
+      "grad_norm": 10.77070851769781,
+      "learning_rate": 1e-06,
+      "loss": 0.1444,
+      "step": 2512
+    },
+    {
+      "epoch": 0.961892247043364,
+      "grad_norm": 11.855176804956649,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9645203679369251,
+      "grad_norm": 12.349794172261486,
+      "learning_rate": 1e-06,
+      "loss": 0.1352,
+      "step": 2514
+    },
+    {
+      "epoch": 0.9671484888304862,
+      "grad_norm": 10.135047987279426,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9697766097240473,
+      "grad_norm": 11.32009064104941,
+      "learning_rate": 1e-06,
+      "loss": 0.1329,
+      "step": 2516
+    },
+    {
+      "epoch": 0.9724047306176085,
+      "grad_norm": 9.742963953320409,
+      "learning_rate": 1e-06,
+      "loss": 0.1423,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9750328515111695,
+      "grad_norm": 11.795829753109674,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 2518
+    },
+    {
+      "epoch": 0.9776609724047306,
+      "grad_norm": 10.418843110452386,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9802890932982917,
+      "grad_norm": 12.686253903947206,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 2520
+    },
+    {
+      "epoch": 0.9829172141918529,
+      "grad_norm": 10.232747425316791,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 2521
+    },
+    {
+      "epoch": 0.985545335085414,
+      "grad_norm": 10.606376692695523,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 2522
+    },
+    {
+      "epoch": 0.988173455978975,
+      "grad_norm": 14.196827658927363,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9908015768725361,
+      "grad_norm": 8.632364601481735,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2524
+    },
+    {
+      "epoch": 0.9934296977660972,
+      "grad_norm": 9.086006113143444,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9960578186596584,
+      "grad_norm": 12.47853586308814,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 2526
+    },
+    {
+      "epoch": 0.9986859395532195,
+      "grad_norm": 12.463563594813436,
+      "learning_rate": 1e-06,
+      "loss": 0.1298,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9986859395532195,
+      "eval_loss": 3.3509135246276855,
+      "eval_runtime": 21.89,
+      "eval_samples_per_second": 45.683,
+      "eval_steps_per_second": 5.71,
+      "step": 2527
+    },
+    {
+      "epoch": 0.0026246719160104987,
+      "grad_norm": 10.853525056666708,
+      "learning_rate": 1e-06,
+      "loss": 0.1244,
+      "step": 2528
+    },
+    {
+      "epoch": 0.005249343832020997,
+      "grad_norm": 9.620053291253818,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 2529
+    },
+    {
+      "epoch": 0.007874015748031496,
+      "grad_norm": 8.564828225303499,
+      "learning_rate": 1e-06,
+      "loss": 0.115,
+      "step": 2530
+    },
+    {
+      "epoch": 0.010498687664041995,
+      "grad_norm": 7.808537109042193,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 2531
+    },
+    {
+      "epoch": 0.013123359580052493,
+      "grad_norm": 11.959066718713357,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 2532
+    },
+    {
+      "epoch": 0.015748031496062992,
+      "grad_norm": 9.207021343257633,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 2533
+    },
+    {
+      "epoch": 0.01837270341207349,
+      "grad_norm": 10.858602596354041,
+      "learning_rate": 1e-06,
+      "loss": 0.118,
+      "step": 2534
+    },
+    {
+      "epoch": 0.02099737532808399,
+      "grad_norm": 7.706265604278983,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 2535
+    },
+    {
+      "epoch": 0.023622047244094488,
+      "grad_norm": 6.1897920574103455,
+      "learning_rate": 1e-06,
+      "loss": 0.1264,
+      "step": 2536
+    },
+    {
+      "epoch": 0.026246719160104987,
+      "grad_norm": 8.437432784057263,
+      "learning_rate": 1e-06,
+      "loss": 0.1127,
+      "step": 2537
+    },
+    {
+      "epoch": 0.028871391076115485,
+      "grad_norm": 11.17518843448044,
+      "learning_rate": 1e-06,
+      "loss": 0.1183,
+      "step": 2538
+    },
+    {
+      "epoch": 0.031496062992125984,
+      "grad_norm": 12.22564295863918,
+      "learning_rate": 1e-06,
+      "loss": 0.1197,
+      "step": 2539
+    },
+    {
+      "epoch": 0.03412073490813648,
+      "grad_norm": 8.711642258327066,
+      "learning_rate": 1e-06,
+      "loss": 0.1172,
+      "step": 2540
+    },
+    {
+      "epoch": 0.03674540682414698,
+      "grad_norm": 7.899255773864468,
+      "learning_rate": 1e-06,
+      "loss": 0.1137,
+      "step": 2541
+    },
+    {
+      "epoch": 0.03937007874015748,
+      "grad_norm": 10.256043744856875,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2542
+    },
+    {
+      "epoch": 0.04199475065616798,
+      "grad_norm": 10.732425284743092,
+      "learning_rate": 1e-06,
+      "loss": 0.1185,
+      "step": 2543
+    },
+    {
+      "epoch": 0.04461942257217848,
+      "grad_norm": 9.60834057852342,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 2544
+    },
+    {
+      "epoch": 0.047244094488188976,
+      "grad_norm": 11.990444463114073,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 2545
+    },
+    {
+      "epoch": 0.049868766404199474,
+      "grad_norm": 9.980394498736434,
+      "learning_rate": 1e-06,
+      "loss": 0.1219,
+      "step": 2546
+    },
+    {
+      "epoch": 0.05249343832020997,
+      "grad_norm": 11.863834834318464,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 2547
+    },
+    {
+      "epoch": 0.05511811023622047,
+      "grad_norm": 13.011202096863418,
+      "learning_rate": 1e-06,
+      "loss": 0.1061,
+      "step": 2548
+    },
+    {
+      "epoch": 0.05774278215223097,
+      "grad_norm": 13.041934522692681,
+      "learning_rate": 1e-06,
+      "loss": 0.1173,
+      "step": 2549
+    },
+    {
+      "epoch": 0.06036745406824147,
+      "grad_norm": 8.127228193259471,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 2550
+    },
+    {
+      "epoch": 0.06299212598425197,
+      "grad_norm": 10.016116997547917,
+      "learning_rate": 1e-06,
+      "loss": 0.1203,
+      "step": 2551
+    },
+    {
+      "epoch": 0.06561679790026247,
+      "grad_norm": 8.117874285113714,
+      "learning_rate": 1e-06,
+      "loss": 0.1248,
+      "step": 2552
+    },
+    {
+      "epoch": 0.06824146981627296,
+      "grad_norm": 12.361437675836267,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 2553
+    },
+    {
+      "epoch": 0.07086614173228346,
+      "grad_norm": 8.267962578721002,
+      "learning_rate": 1e-06,
+      "loss": 0.1199,
+      "step": 2554
+    },
+    {
+      "epoch": 0.07349081364829396,
+      "grad_norm": 10.167744907227267,
+      "learning_rate": 1e-06,
+      "loss": 0.1138,
+      "step": 2555
+    },
+    {
+      "epoch": 0.07611548556430446,
+      "grad_norm": 10.006634148189722,
+      "learning_rate": 1e-06,
+      "loss": 0.1163,
+      "step": 2556
+    },
+    {
+      "epoch": 0.07874015748031496,
+      "grad_norm": 10.443375671768885,
+      "learning_rate": 1e-06,
+      "loss": 0.1285,
+      "step": 2557
+    },
+    {
+      "epoch": 0.08136482939632546,
+      "grad_norm": 11.58940020275945,
+      "learning_rate": 1e-06,
+      "loss": 0.1169,
+      "step": 2558
+    },
+    {
+      "epoch": 0.08398950131233596,
+      "grad_norm": 12.210486568864862,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 2559
+    },
+    {
+      "epoch": 0.08661417322834646,
+      "grad_norm": 8.839589827695402,
+      "learning_rate": 1e-06,
+      "loss": 0.1222,
+      "step": 2560
+    },
+    {
+      "epoch": 0.08923884514435695,
+      "grad_norm": 11.101973458470598,
+      "learning_rate": 1e-06,
+      "loss": 0.1093,
+      "step": 2561
+    },
+    {
+      "epoch": 0.09186351706036745,
+      "grad_norm": 10.277640273981957,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 2562
+    },
+    {
+      "epoch": 0.09448818897637795,
+      "grad_norm": 13.887185946652036,
+      "learning_rate": 1e-06,
+      "loss": 0.1329,
+      "step": 2563
+    },
+    {
+      "epoch": 0.09711286089238845,
+      "grad_norm": 10.98177778010855,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 2564
+    },
+    {
+      "epoch": 0.09973753280839895,
+      "grad_norm": 11.73838744482427,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 2565
+    },
+    {
+      "epoch": 0.10236220472440945,
+      "grad_norm": 8.805815100167768,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 2566
+    },
+    {
+      "epoch": 0.10498687664041995,
+      "grad_norm": 10.641214815813354,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 2567
+    },
+    {
+      "epoch": 0.10761154855643044,
+      "grad_norm": 13.0498648897384,
+      "learning_rate": 1e-06,
+      "loss": 0.1277,
+      "step": 2568
+    },
+    {
+      "epoch": 0.11023622047244094,
+      "grad_norm": 8.351446958579881,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 2569
+    },
+    {
+      "epoch": 0.11286089238845144,
+      "grad_norm": 7.3759601985959975,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 2570
+    },
+    {
+      "epoch": 0.11548556430446194,
+      "grad_norm": 12.10191719301966,
+      "learning_rate": 1e-06,
+      "loss": 0.1158,
+      "step": 2571
+    },
+    {
+      "epoch": 0.11811023622047244,
+      "grad_norm": 11.347603015130526,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2572
+    },
+    {
+      "epoch": 0.12073490813648294,
+      "grad_norm": 9.811800858167505,
+      "learning_rate": 1e-06,
+      "loss": 0.1306,
+      "step": 2573
+    },
+    {
+      "epoch": 0.12335958005249344,
+      "grad_norm": 10.129213943961082,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 2574
+    },
+    {
+      "epoch": 0.12598425196850394,
+      "grad_norm": 12.751473892806873,
+      "learning_rate": 1e-06,
+      "loss": 0.1263,
+      "step": 2575
+    },
+    {
+      "epoch": 0.12860892388451445,
+      "grad_norm": 12.674407907823538,
+      "learning_rate": 1e-06,
+      "loss": 0.1165,
+      "step": 2576
+    },
+    {
+      "epoch": 0.13123359580052493,
+      "grad_norm": 8.92701715230463,
+      "learning_rate": 1e-06,
+      "loss": 0.1245,
+      "step": 2577
+    },
+    {
+      "epoch": 0.13385826771653545,
+      "grad_norm": 9.830048744520427,
+      "learning_rate": 1e-06,
+      "loss": 0.1355,
+      "step": 2578
+    },
+    {
+      "epoch": 0.13648293963254593,
+      "grad_norm": 9.917002959294857,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 2579
+    },
+    {
+      "epoch": 0.13910761154855644,
+      "grad_norm": 10.563816660588241,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 2580
+    },
+    {
+      "epoch": 0.14173228346456693,
+      "grad_norm": 11.081697133826768,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 2581
+    },
+    {
+      "epoch": 0.14435695538057744,
+      "grad_norm": 10.216848643286072,
+      "learning_rate": 1e-06,
+      "loss": 0.1124,
+      "step": 2582
+    },
+    {
+      "epoch": 0.14698162729658792,
+      "grad_norm": 12.321382815010464,
+      "learning_rate": 1e-06,
+      "loss": 0.1216,
+      "step": 2583
+    },
+    {
+      "epoch": 0.14960629921259844,
+      "grad_norm": 10.265350823670671,
+      "learning_rate": 1e-06,
+      "loss": 0.1417,
+      "step": 2584
+    },
+    {
+      "epoch": 0.15223097112860892,
+      "grad_norm": 12.242787874135638,
+      "learning_rate": 1e-06,
+      "loss": 0.1234,
+      "step": 2585
+    },
+    {
+      "epoch": 0.15485564304461943,
+      "grad_norm": 11.767631474449187,
+      "learning_rate": 1e-06,
+      "loss": 0.1173,
+      "step": 2586
+    },
+    {
+      "epoch": 0.15748031496062992,
+      "grad_norm": 12.701412130453313,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 2587
+    },
+    {
+      "epoch": 0.16010498687664043,
+      "grad_norm": 11.825968803667257,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 2588
+    },
+    {
+      "epoch": 0.16272965879265092,
+      "grad_norm": 12.516317299033641,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 2589
+    },
+    {
+      "epoch": 0.16535433070866143,
+      "grad_norm": 9.577108221595894,
+      "learning_rate": 1e-06,
+      "loss": 0.1196,
+      "step": 2590
+    },
+    {
+      "epoch": 0.1679790026246719,
+      "grad_norm": 8.645921081152505,
+      "learning_rate": 1e-06,
+      "loss": 0.125,
+      "step": 2591
+    },
+    {
+      "epoch": 0.17060367454068243,
+      "grad_norm": 12.042510565106594,
+      "learning_rate": 1e-06,
+      "loss": 0.1367,
+      "step": 2592
+    },
+    {
+      "epoch": 0.1732283464566929,
+      "grad_norm": 10.174564675648933,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 2593
+    },
+    {
+      "epoch": 0.17585301837270342,
+      "grad_norm": 12.970315940274304,
+      "learning_rate": 1e-06,
+      "loss": 0.1181,
+      "step": 2594
+    },
+    {
+      "epoch": 0.1784776902887139,
+      "grad_norm": 11.890673652426736,
+      "learning_rate": 1e-06,
+      "loss": 0.1289,
+      "step": 2595
+    },
+    {
+      "epoch": 0.18110236220472442,
+      "grad_norm": 10.649277956891714,
+      "learning_rate": 1e-06,
+      "loss": 0.1322,
+      "step": 2596
+    },
+    {
+      "epoch": 0.1837270341207349,
+      "grad_norm": 11.09614691890112,
+      "learning_rate": 1e-06,
+      "loss": 0.1256,
+      "step": 2597
+    },
+    {
+      "epoch": 0.18635170603674542,
+      "grad_norm": 10.675851441577281,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2598
+    },
+    {
+      "epoch": 0.1889763779527559,
+      "grad_norm": 8.205527375879507,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 2599
+    },
+    {
+      "epoch": 0.19160104986876642,
+      "grad_norm": 11.996543960413062,
+      "learning_rate": 1e-06,
+      "loss": 0.1172,
+      "step": 2600
+    },
+    {
+      "epoch": 0.1942257217847769,
+      "grad_norm": 11.0308799240873,
+      "learning_rate": 1e-06,
+      "loss": 0.1378,
+      "step": 2601
+    },
+    {
+      "epoch": 0.1968503937007874,
+      "grad_norm": 11.326611250622028,
+      "learning_rate": 1e-06,
+      "loss": 0.123,
+      "step": 2602
+    },
+    {
+      "epoch": 0.1994750656167979,
+      "grad_norm": 13.43703078898254,
+      "learning_rate": 1e-06,
+      "loss": 0.1328,
+      "step": 2603
+    },
+    {
+      "epoch": 0.2020997375328084,
+      "grad_norm": 9.537503684421736,
+      "learning_rate": 1e-06,
+      "loss": 0.133,
+      "step": 2604
+    },
+    {
+      "epoch": 0.2047244094488189,
+      "grad_norm": 12.576030396308333,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2605
+    },
+    {
+      "epoch": 0.2073490813648294,
+      "grad_norm": 8.552755821919389,
+      "learning_rate": 1e-06,
+      "loss": 0.1259,
+      "step": 2606
+    },
+    {
+      "epoch": 0.2099737532808399,
+      "grad_norm": 8.510651348921046,
+      "learning_rate": 1e-06,
+      "loss": 0.1281,
+      "step": 2607
+    },
+    {
+      "epoch": 0.2125984251968504,
+      "grad_norm": 9.703573741488208,
+      "learning_rate": 1e-06,
+      "loss": 0.1277,
+      "step": 2608
+    },
+    {
+      "epoch": 0.2152230971128609,
+      "grad_norm": 11.951342797027317,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2609
+    },
+    {
+      "epoch": 0.2178477690288714,
+      "grad_norm": 13.19718633959311,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2610
+    },
+    {
+      "epoch": 0.2204724409448819,
+      "grad_norm": 10.690497437560634,
+      "learning_rate": 1e-06,
+      "loss": 0.1286,
+      "step": 2611
+    },
+    {
+      "epoch": 0.2230971128608924,
+      "grad_norm": 7.568210735430382,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 2612
+    },
+    {
+      "epoch": 0.22572178477690288,
+      "grad_norm": 10.806581282159712,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2613
+    },
+    {
+      "epoch": 0.2283464566929134,
+      "grad_norm": 11.549737381499119,
+      "learning_rate": 1e-06,
+      "loss": 0.127,
+      "step": 2614
+    },
+    {
+      "epoch": 0.23097112860892388,
+      "grad_norm": 11.981747376617738,
+      "learning_rate": 1e-06,
+      "loss": 0.1168,
+      "step": 2615
+    },
+    {
+      "epoch": 0.2335958005249344,
+      "grad_norm": 9.648708724626285,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 2616
+    },
+    {
+      "epoch": 0.23622047244094488,
+      "grad_norm": 12.665617711927833,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 2617
+    },
+    {
+      "epoch": 0.2388451443569554,
+      "grad_norm": 9.116470019087,
+      "learning_rate": 1e-06,
+      "loss": 0.1187,
+      "step": 2618
+    },
+    {
+      "epoch": 0.24146981627296588,
+      "grad_norm": 9.182039809282843,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 2619
+    },
+    {
+      "epoch": 0.2440944881889764,
+      "grad_norm": 9.961206559259224,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 2620
+    },
+    {
+      "epoch": 0.24671916010498687,
+      "grad_norm": 9.105662470309587,
+      "learning_rate": 1e-06,
+      "loss": 0.1287,
+      "step": 2621
+    },
+    {
+      "epoch": 0.24934383202099739,
+      "grad_norm": 9.069336483344264,
+      "learning_rate": 1e-06,
+      "loss": 0.1271,
+      "step": 2622
+    },
+    {
+      "epoch": 0.25196850393700787,
+      "grad_norm": 10.185906351595456,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 2623
+    },
+    {
+      "epoch": 0.2545931758530184,
+      "grad_norm": 11.916651753651628,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 2624
+    },
+    {
+      "epoch": 0.2572178477690289,
+      "grad_norm": 10.730178996604957,
+      "learning_rate": 1e-06,
+      "loss": 0.1291,
+      "step": 2625
+    },
+    {
+      "epoch": 0.25984251968503935,
+      "grad_norm": 11.155747715956918,
+      "learning_rate": 1e-06,
+      "loss": 0.1156,
+      "step": 2626
+    },
+    {
+      "epoch": 0.26246719160104987,
+      "grad_norm": 9.445387093870744,
+      "learning_rate": 1e-06,
+      "loss": 0.1185,
+      "step": 2627
+    },
+    {
+      "epoch": 0.2650918635170604,
+      "grad_norm": 11.12677963631987,
+      "learning_rate": 1e-06,
+      "loss": 0.1412,
+      "step": 2628
+    },
+    {
+      "epoch": 0.2677165354330709,
+      "grad_norm": 9.541371354091634,
+      "learning_rate": 1e-06,
+      "loss": 0.1287,
+      "step": 2629
+    },
+    {
+      "epoch": 0.27034120734908135,
+      "grad_norm": 9.518626352083256,
+      "learning_rate": 1e-06,
+      "loss": 0.1431,
+      "step": 2630
+    },
+    {
+      "epoch": 0.27296587926509186,
+      "grad_norm": 11.11374627278674,
+      "learning_rate": 1e-06,
+      "loss": 0.1216,
+      "step": 2631
+    },
+    {
+      "epoch": 0.2755905511811024,
+      "grad_norm": 11.883364298936367,
+      "learning_rate": 1e-06,
+      "loss": 0.1257,
+      "step": 2632
+    },
+    {
+      "epoch": 0.2782152230971129,
+      "grad_norm": 9.400263054120007,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 2633
+    },
+    {
+      "epoch": 0.28083989501312334,
+      "grad_norm": 10.504335368831258,
+      "learning_rate": 1e-06,
+      "loss": 0.1329,
+      "step": 2634
+    },
+    {
+      "epoch": 0.28346456692913385,
+      "grad_norm": 10.44314348831234,
+      "learning_rate": 1e-06,
+      "loss": 0.1254,
+      "step": 2635
+    },
+    {
+      "epoch": 0.28608923884514437,
+      "grad_norm": 8.613563004116385,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 2636
+    },
+    {
+      "epoch": 0.2887139107611549,
+      "grad_norm": 10.541820588797977,
+      "learning_rate": 1e-06,
+      "loss": 0.1187,
+      "step": 2637
+    },
+    {
+      "epoch": 0.29133858267716534,
+      "grad_norm": 8.869348866824433,
+      "learning_rate": 1e-06,
+      "loss": 0.1353,
+      "step": 2638
+    },
+    {
+      "epoch": 0.29396325459317585,
+      "grad_norm": 11.726180657732282,
+      "learning_rate": 1e-06,
+      "loss": 0.1297,
+      "step": 2639
+    },
+    {
+      "epoch": 0.29658792650918636,
+      "grad_norm": 7.949818437550837,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 2640
+    },
+    {
+      "epoch": 0.2992125984251969,
+      "grad_norm": 10.77901551999534,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 2641
+    },
+    {
+      "epoch": 0.30183727034120733,
+      "grad_norm": 11.12742905750198,
+      "learning_rate": 1e-06,
+      "loss": 0.1341,
+      "step": 2642
+    },
+    {
+      "epoch": 0.30446194225721784,
+      "grad_norm": 12.6680665357383,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 2643
+    },
+    {
+      "epoch": 0.30708661417322836,
+      "grad_norm": 12.164392810541209,
+      "learning_rate": 1e-06,
+      "loss": 0.1297,
+      "step": 2644
+    },
+    {
+      "epoch": 0.30971128608923887,
+      "grad_norm": 14.920359289559052,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 2645
+    },
+    {
+      "epoch": 0.3123359580052493,
+      "grad_norm": 12.001062468360502,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2646
+    },
+    {
+      "epoch": 0.31496062992125984,
+      "grad_norm": 10.144008171834377,
+      "learning_rate": 1e-06,
+      "loss": 0.1217,
+      "step": 2647
+    },
+    {
+      "epoch": 0.31758530183727035,
+      "grad_norm": 11.092886603605812,
+      "learning_rate": 1e-06,
+      "loss": 0.1239,
+      "step": 2648
+    },
+    {
+      "epoch": 0.32020997375328086,
+      "grad_norm": 11.035546579715746,
+      "learning_rate": 1e-06,
+      "loss": 0.1384,
+      "step": 2649
+    },
+    {
+      "epoch": 0.3228346456692913,
+      "grad_norm": 12.267691213421788,
+      "learning_rate": 1e-06,
+      "loss": 0.1211,
+      "step": 2650
+    },
+    {
+      "epoch": 0.32545931758530183,
+      "grad_norm": 11.632883324393095,
+      "learning_rate": 1e-06,
+      "loss": 0.1253,
+      "step": 2651
+    },
+    {
+      "epoch": 0.32808398950131235,
+      "grad_norm": 12.006317459634573,
+      "learning_rate": 1e-06,
+      "loss": 0.1328,
+      "step": 2652
+    },
+    {
+      "epoch": 0.33070866141732286,
+      "grad_norm": 9.947755873176751,
+      "learning_rate": 1e-06,
+      "loss": 0.1342,
+      "step": 2653
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 8.921264557275823,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 2654
+    },
+    {
+      "epoch": 0.3359580052493438,
+      "grad_norm": 8.667275614026334,
+      "learning_rate": 1e-06,
+      "loss": 0.124,
+      "step": 2655
+    },
+    {
+      "epoch": 0.33858267716535434,
+      "grad_norm": 8.531902936132965,
+      "learning_rate": 1e-06,
+      "loss": 0.1187,
+      "step": 2656
+    },
+    {
+      "epoch": 0.34120734908136485,
+      "grad_norm": 10.005946461719093,
+      "learning_rate": 1e-06,
+      "loss": 0.1288,
+      "step": 2657
+    },
+    {
+      "epoch": 0.3438320209973753,
+      "grad_norm": 12.291761661119068,
+      "learning_rate": 1e-06,
+      "loss": 0.1352,
+      "step": 2658
+    },
+    {
+      "epoch": 0.3464566929133858,
+      "grad_norm": 14.502196782245482,
+      "learning_rate": 1e-06,
+      "loss": 0.1349,
+      "step": 2659
+    },
+    {
+      "epoch": 0.34908136482939633,
+      "grad_norm": 11.992645667880774,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 2660
+    },
+    {
+      "epoch": 0.35170603674540685,
+      "grad_norm": 12.382841445147442,
+      "learning_rate": 1e-06,
+      "loss": 0.1298,
+      "step": 2661
+    },
+    {
+      "epoch": 0.3543307086614173,
+      "grad_norm": 12.530085506468847,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2662
+    },
+    {
+      "epoch": 0.3569553805774278,
+      "grad_norm": 7.752650491223479,
+      "learning_rate": 1e-06,
+      "loss": 0.1269,
+      "step": 2663
+    },
+    {
+      "epoch": 0.35958005249343833,
+      "grad_norm": 10.349576577253218,
+      "learning_rate": 1e-06,
+      "loss": 0.1224,
+      "step": 2664
+    },
+    {
+      "epoch": 0.36220472440944884,
+      "grad_norm": 12.922065981154486,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 2665
+    },
+    {
+      "epoch": 0.3648293963254593,
+      "grad_norm": 10.709015459771816,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 2666
+    },
+    {
+      "epoch": 0.3674540682414698,
+      "grad_norm": 7.455727143890579,
+      "learning_rate": 1e-06,
+      "loss": 0.1436,
+      "step": 2667
+    },
+    {
+      "epoch": 0.3700787401574803,
+      "grad_norm": 11.94205876827365,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 2668
+    },
+    {
+      "epoch": 0.37270341207349084,
+      "grad_norm": 12.309488909659144,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 2669
+    },
+    {
+      "epoch": 0.3753280839895013,
+      "grad_norm": 11.452540853872087,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 2670
+    },
+    {
+      "epoch": 0.3779527559055118,
+      "grad_norm": 9.359943584733172,
+      "learning_rate": 1e-06,
+      "loss": 0.1322,
+      "step": 2671
+    },
+    {
+      "epoch": 0.3805774278215223,
+      "grad_norm": 7.642868834111331,
+      "learning_rate": 1e-06,
+      "loss": 0.1256,
+      "step": 2672
+    },
+    {
+      "epoch": 0.38320209973753283,
+      "grad_norm": 11.330417153469256,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 2673
+    },
+    {
+      "epoch": 0.3858267716535433,
+      "grad_norm": 11.08107761400604,
+      "learning_rate": 1e-06,
+      "loss": 0.1161,
+      "step": 2674
+    },
+    {
+      "epoch": 0.3884514435695538,
+      "grad_norm": 11.038352675878103,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 2675
+    },
+    {
+      "epoch": 0.3910761154855643,
+      "grad_norm": 11.801705387047964,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 2676
+    },
+    {
+      "epoch": 0.3937007874015748,
+      "grad_norm": 12.908492153714409,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 2677
+    },
+    {
+      "epoch": 0.3963254593175853,
+      "grad_norm": 12.284647064211288,
+      "learning_rate": 1e-06,
+      "loss": 0.1262,
+      "step": 2678
+    },
+    {
+      "epoch": 0.3989501312335958,
+      "grad_norm": 8.664663993340897,
+      "learning_rate": 1e-06,
+      "loss": 0.1229,
+      "step": 2679
+    },
+    {
+      "epoch": 0.4015748031496063,
+      "grad_norm": 11.110986588384337,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 2680
+    },
+    {
+      "epoch": 0.4041994750656168,
+      "grad_norm": 10.524879140591601,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 2681
+    },
+    {
+      "epoch": 0.4068241469816273,
+      "grad_norm": 10.378937002361855,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 2682
+    },
+    {
+      "epoch": 0.4094488188976378,
+      "grad_norm": 10.549255950673285,
+      "learning_rate": 1e-06,
+      "loss": 0.1254,
+      "step": 2683
+    },
+    {
+      "epoch": 0.4120734908136483,
+      "grad_norm": 8.371957155963825,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 2684
+    },
+    {
+      "epoch": 0.4146981627296588,
+      "grad_norm": 10.637905347717465,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 2685
+    },
+    {
+      "epoch": 0.41732283464566927,
+      "grad_norm": 10.895235400001118,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 2686
+    },
+    {
+      "epoch": 0.4199475065616798,
+      "grad_norm": 8.552757991827637,
+      "learning_rate": 1e-06,
+      "loss": 0.1296,
+      "step": 2687
+    },
+    {
+      "epoch": 0.4225721784776903,
+      "grad_norm": 10.692350786580798,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 2688
+    },
+    {
+      "epoch": 0.4251968503937008,
+      "grad_norm": 14.9164782615388,
+      "learning_rate": 1e-06,
+      "loss": 0.1301,
+      "step": 2689
+    },
+    {
+      "epoch": 0.42782152230971127,
+      "grad_norm": 12.044104910304759,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 2690
+    },
+    {
+      "epoch": 0.4304461942257218,
+      "grad_norm": 13.9192179695436,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 2691
+    },
+    {
+      "epoch": 0.4330708661417323,
+      "grad_norm": 9.439340506278864,
+      "learning_rate": 1e-06,
+      "loss": 0.1348,
+      "step": 2692
+    },
+    {
+      "epoch": 0.4356955380577428,
+      "grad_norm": 10.938624881732942,
+      "learning_rate": 1e-06,
+      "loss": 0.1245,
+      "step": 2693
+    },
+    {
+      "epoch": 0.43832020997375326,
+      "grad_norm": 8.831689385939185,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2694
+    },
+    {
+      "epoch": 0.4409448818897638,
+      "grad_norm": 8.43410477819861,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 2695
+    },
+    {
+      "epoch": 0.4435695538057743,
+      "grad_norm": 11.646731271544537,
+      "learning_rate": 1e-06,
+      "loss": 0.1234,
+      "step": 2696
+    },
+    {
+      "epoch": 0.4461942257217848,
+      "grad_norm": 9.135953820097523,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 2697
+    },
+    {
+      "epoch": 0.44881889763779526,
+      "grad_norm": 10.189050290864623,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 2698
+    },
+    {
+      "epoch": 0.45144356955380577,
+      "grad_norm": 6.829931040629058,
+      "learning_rate": 1e-06,
+      "loss": 0.1232,
+      "step": 2699
+    },
+    {
+      "epoch": 0.4540682414698163,
+      "grad_norm": 10.529347926260016,
+      "learning_rate": 1e-06,
+      "loss": 0.1281,
+      "step": 2700
+    },
+    {
+      "epoch": 0.4566929133858268,
+      "grad_norm": 10.99788344050654,
+      "learning_rate": 1e-06,
+      "loss": 0.124,
+      "step": 2701
+    },
+    {
+      "epoch": 0.45931758530183725,
+      "grad_norm": 11.41079173317968,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 2702
+    },
+    {
+      "epoch": 0.46194225721784776,
+      "grad_norm": 11.26899127560127,
+      "learning_rate": 1e-06,
+      "loss": 0.1302,
+      "step": 2703
+    },
+    {
+      "epoch": 0.4645669291338583,
+      "grad_norm": 9.186221745071714,
+      "learning_rate": 1e-06,
+      "loss": 0.1364,
+      "step": 2704
+    },
+    {
+      "epoch": 0.4671916010498688,
+      "grad_norm": 10.377758377749007,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 2705
+    },
+    {
+      "epoch": 0.46981627296587924,
+      "grad_norm": 9.1429138072054,
+      "learning_rate": 1e-06,
+      "loss": 0.1257,
+      "step": 2706
+    },
+    {
+      "epoch": 0.47244094488188976,
+      "grad_norm": 11.58515679358197,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 2707
+    },
+    {
+      "epoch": 0.47506561679790027,
+      "grad_norm": 9.5149516438888,
+      "learning_rate": 1e-06,
+      "loss": 0.1315,
+      "step": 2708
+    },
+    {
+      "epoch": 0.4776902887139108,
+      "grad_norm": 11.587583093029957,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 2709
+    },
+    {
+      "epoch": 0.48031496062992124,
+      "grad_norm": 12.05474904052806,
+      "learning_rate": 1e-06,
+      "loss": 0.1408,
+      "step": 2710
+    },
+    {
+      "epoch": 0.48293963254593175,
+      "grad_norm": 10.75376266974449,
+      "learning_rate": 1e-06,
+      "loss": 0.1425,
+      "step": 2711
+    },
+    {
+      "epoch": 0.48556430446194226,
+      "grad_norm": 10.991910342850721,
+      "learning_rate": 1e-06,
+      "loss": 0.1228,
+      "step": 2712
+    },
+    {
+      "epoch": 0.4881889763779528,
+      "grad_norm": 12.691914440667084,
+      "learning_rate": 1e-06,
+      "loss": 0.1344,
+      "step": 2713
+    },
+    {
+      "epoch": 0.49081364829396323,
+      "grad_norm": 9.809045712920607,
+      "learning_rate": 1e-06,
+      "loss": 0.1307,
+      "step": 2714
+    },
+    {
+      "epoch": 0.49343832020997375,
+      "grad_norm": 12.909865303681874,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2715
+    },
+    {
+      "epoch": 0.49606299212598426,
+      "grad_norm": 12.569456419031313,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 2716
+    },
+    {
+      "epoch": 0.49868766404199477,
+      "grad_norm": 11.995692742190524,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 2717
+    },
+    {
+      "epoch": 0.5013123359580053,
+      "grad_norm": 10.812862406446085,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 2718
+    },
+    {
+      "epoch": 0.5039370078740157,
+      "grad_norm": 10.761123307357153,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 2719
+    },
+    {
+      "epoch": 0.5065616797900262,
+      "grad_norm": 8.632483605378436,
+      "learning_rate": 1e-06,
+      "loss": 0.1234,
+      "step": 2720
+    },
+    {
+      "epoch": 0.5091863517060368,
+      "grad_norm": 7.3089611994158945,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 2721
+    },
+    {
+      "epoch": 0.5118110236220472,
+      "grad_norm": 8.269814671240209,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 2722
+    },
+    {
+      "epoch": 0.5144356955380578,
+      "grad_norm": 11.08555755636647,
+      "learning_rate": 1e-06,
+      "loss": 0.1364,
+      "step": 2723
+    },
+    {
+      "epoch": 0.5170603674540682,
+      "grad_norm": 11.243042987144896,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2724
+    },
+    {
+      "epoch": 0.5196850393700787,
+      "grad_norm": 12.182631489624924,
+      "learning_rate": 1e-06,
+      "loss": 0.1369,
+      "step": 2725
+    },
+    {
+      "epoch": 0.5223097112860893,
+      "grad_norm": 10.655475290904816,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 2726
+    },
+    {
+      "epoch": 0.5249343832020997,
+      "grad_norm": 10.457646631269624,
+      "learning_rate": 1e-06,
+      "loss": 0.1389,
+      "step": 2727
+    },
+    {
+      "epoch": 0.5275590551181102,
+      "grad_norm": 7.917941036041188,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 2728
+    },
+    {
+      "epoch": 0.5301837270341208,
+      "grad_norm": 8.736077436046994,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 2729
+    },
+    {
+      "epoch": 0.5328083989501312,
+      "grad_norm": 9.293196102258953,
+      "learning_rate": 1e-06,
+      "loss": 0.1409,
+      "step": 2730
+    },
+    {
+      "epoch": 0.5354330708661418,
+      "grad_norm": 10.681133887618332,
+      "learning_rate": 1e-06,
+      "loss": 0.1307,
+      "step": 2731
+    },
+    {
+      "epoch": 0.5380577427821522,
+      "grad_norm": 8.364147596807793,
+      "learning_rate": 1e-06,
+      "loss": 0.1287,
+      "step": 2732
+    },
+    {
+      "epoch": 0.5406824146981627,
+      "grad_norm": 10.274599381566595,
+      "learning_rate": 1e-06,
+      "loss": 0.1293,
+      "step": 2733
+    },
+    {
+      "epoch": 0.5433070866141733,
+      "grad_norm": 8.253064072331512,
+      "learning_rate": 1e-06,
+      "loss": 0.137,
+      "step": 2734
+    },
+    {
+      "epoch": 0.5459317585301837,
+      "grad_norm": 9.476591412090531,
+      "learning_rate": 1e-06,
+      "loss": 0.1237,
+      "step": 2735
+    },
+    {
+      "epoch": 0.5485564304461942,
+      "grad_norm": 8.819166851117702,
+      "learning_rate": 1e-06,
+      "loss": 0.1292,
+      "step": 2736
+    },
+    {
+      "epoch": 0.5511811023622047,
+      "grad_norm": 9.793630596393005,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 2737
+    },
+    {
+      "epoch": 0.5538057742782152,
+      "grad_norm": 10.127574878760466,
+      "learning_rate": 1e-06,
+      "loss": 0.1235,
+      "step": 2738
+    },
+    {
+      "epoch": 0.5564304461942258,
+      "grad_norm": 11.295838448725975,
+      "learning_rate": 1e-06,
+      "loss": 0.1326,
+      "step": 2739
+    },
+    {
+      "epoch": 0.5590551181102362,
+      "grad_norm": 10.550076537113199,
+      "learning_rate": 1e-06,
+      "loss": 0.1252,
+      "step": 2740
+    },
+    {
+      "epoch": 0.5616797900262467,
+      "grad_norm": 9.24004337918305,
+      "learning_rate": 1e-06,
+      "loss": 0.1226,
+      "step": 2741
+    },
+    {
+      "epoch": 0.5643044619422573,
+      "grad_norm": 10.704410182897014,
+      "learning_rate": 1e-06,
+      "loss": 0.129,
+      "step": 2742
+    },
+    {
+      "epoch": 0.5669291338582677,
+      "grad_norm": 10.590804084481363,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 2743
+    },
+    {
+      "epoch": 0.5695538057742782,
+      "grad_norm": 10.280097017924218,
+      "learning_rate": 1e-06,
+      "loss": 0.1218,
+      "step": 2744
+    },
+    {
+      "epoch": 0.5721784776902887,
+      "grad_norm": 9.795473205014943,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 2745
+    },
+    {
+      "epoch": 0.5748031496062992,
+      "grad_norm": 10.694856439595995,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 2746
+    },
+    {
+      "epoch": 0.5774278215223098,
+      "grad_norm": 12.943802457822128,
+      "learning_rate": 1e-06,
+      "loss": 0.1394,
+      "step": 2747
+    },
+    {
+      "epoch": 0.5800524934383202,
+      "grad_norm": 11.681293291682179,
+      "learning_rate": 1e-06,
+      "loss": 0.1402,
+      "step": 2748
+    },
+    {
+      "epoch": 0.5826771653543307,
+      "grad_norm": 10.111605870344654,
+      "learning_rate": 1e-06,
+      "loss": 0.1279,
+      "step": 2749
+    },
+    {
+      "epoch": 0.5853018372703412,
+      "grad_norm": 9.430501005890871,
+      "learning_rate": 1e-06,
+      "loss": 0.1219,
+      "step": 2750
+    },
+    {
+      "epoch": 0.5879265091863517,
+      "grad_norm": 11.942234974928772,
+      "learning_rate": 1e-06,
+      "loss": 0.135,
+      "step": 2751
+    },
+    {
+      "epoch": 0.5905511811023622,
+      "grad_norm": 13.16141004264911,
+      "learning_rate": 1e-06,
+      "loss": 0.1375,
+      "step": 2752
+    },
+    {
+      "epoch": 0.5931758530183727,
+      "grad_norm": 10.626348931373146,
+      "learning_rate": 1e-06,
+      "loss": 0.1262,
+      "step": 2753
+    },
+    {
+      "epoch": 0.5958005249343832,
+      "grad_norm": 11.288815975673641,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 2754
+    },
+    {
+      "epoch": 0.5984251968503937,
+      "grad_norm": 9.923618456003828,
+      "learning_rate": 1e-06,
+      "loss": 0.1238,
+      "step": 2755
+    },
+    {
+      "epoch": 0.6010498687664042,
+      "grad_norm": 10.102527579279826,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 2756
+    },
+    {
+      "epoch": 0.6036745406824147,
+      "grad_norm": 8.781034713630044,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 2757
+    },
+    {
+      "epoch": 0.6062992125984252,
+      "grad_norm": 12.373071619417788,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2758
+    },
+    {
+      "epoch": 0.6089238845144357,
+      "grad_norm": 11.228596723539516,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 2759
+    },
+    {
+      "epoch": 0.6115485564304461,
+      "grad_norm": 9.124026758849086,
+      "learning_rate": 1e-06,
+      "loss": 0.1206,
+      "step": 2760
+    },
+    {
+      "epoch": 0.6141732283464567,
+      "grad_norm": 8.825747390716986,
+      "learning_rate": 1e-06,
+      "loss": 0.1334,
+      "step": 2761
+    },
+    {
+      "epoch": 0.6167979002624672,
+      "grad_norm": 9.085682381184208,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2762
+    },
+    {
+      "epoch": 0.6194225721784777,
+      "grad_norm": 10.4208028831137,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 2763
+    },
+    {
+      "epoch": 0.6220472440944882,
+      "grad_norm": 10.846886619629924,
+      "learning_rate": 1e-06,
+      "loss": 0.1357,
+      "step": 2764
+    },
+    {
+      "epoch": 0.6246719160104987,
+      "grad_norm": 7.866812238259621,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2765
+    },
+    {
+      "epoch": 0.6272965879265092,
+      "grad_norm": 9.892044333582225,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 2766
+    },
+    {
+      "epoch": 0.6299212598425197,
+      "grad_norm": 11.262490253517534,
+      "learning_rate": 1e-06,
+      "loss": 0.1205,
+      "step": 2767
+    },
+    {
+      "epoch": 0.6325459317585301,
+      "grad_norm": 12.090139281873764,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2768
+    },
+    {
+      "epoch": 0.6351706036745407,
+      "grad_norm": 9.25819321671449,
+      "learning_rate": 1e-06,
+      "loss": 0.1275,
+      "step": 2769
+    },
+    {
+      "epoch": 0.6377952755905512,
+      "grad_norm": 9.213608611657651,
+      "learning_rate": 1e-06,
+      "loss": 0.138,
+      "step": 2770
+    },
+    {
+      "epoch": 0.6404199475065617,
+      "grad_norm": 8.483788835742999,
+      "learning_rate": 1e-06,
+      "loss": 0.1278,
+      "step": 2771
+    },
+    {
+      "epoch": 0.6430446194225722,
+      "grad_norm": 10.536120994277324,
+      "learning_rate": 1e-06,
+      "loss": 0.1396,
+      "step": 2772
+    },
+    {
+      "epoch": 0.6456692913385826,
+      "grad_norm": 11.433830816413225,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 2773
+    },
+    {
+      "epoch": 0.6482939632545932,
+      "grad_norm": 8.07683344599695,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 2774
+    },
+    {
+      "epoch": 0.6509186351706037,
+      "grad_norm": 11.87320941146899,
+      "learning_rate": 1e-06,
+      "loss": 0.1276,
+      "step": 2775
+    },
+    {
+      "epoch": 0.6535433070866141,
+      "grad_norm": 10.669450738815485,
+      "learning_rate": 1e-06,
+      "loss": 0.1317,
+      "step": 2776
+    },
+    {
+      "epoch": 0.6561679790026247,
+      "grad_norm": 10.15710675053186,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 2777
+    },
+    {
+      "epoch": 0.6587926509186351,
+      "grad_norm": 9.557649586110966,
+      "learning_rate": 1e-06,
+      "loss": 0.1273,
+      "step": 2778
+    },
+    {
+      "epoch": 0.6614173228346457,
+      "grad_norm": 10.977657803048611,
+      "learning_rate": 1e-06,
+      "loss": 0.1227,
+      "step": 2779
+    },
+    {
+      "epoch": 0.6640419947506562,
+      "grad_norm": 9.033166782563795,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 2780
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 12.026391856024592,
+      "learning_rate": 1e-06,
+      "loss": 0.1355,
+      "step": 2781
+    },
+    {
+      "epoch": 0.6692913385826772,
+      "grad_norm": 10.011196154614362,
+      "learning_rate": 1e-06,
+      "loss": 0.1318,
+      "step": 2782
+    },
+    {
+      "epoch": 0.6719160104986877,
+      "grad_norm": 9.925232286080545,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 2783
+    },
+    {
+      "epoch": 0.6745406824146981,
+      "grad_norm": 11.485669047956351,
+      "learning_rate": 1e-06,
+      "loss": 0.1343,
+      "step": 2784
+    },
+    {
+      "epoch": 0.6771653543307087,
+      "grad_norm": 9.883451619362871,
+      "learning_rate": 1e-06,
+      "loss": 0.1428,
+      "step": 2785
+    },
+    {
+      "epoch": 0.6797900262467191,
+      "grad_norm": 12.493351030914981,
+      "learning_rate": 1e-06,
+      "loss": 0.1295,
+      "step": 2786
+    },
+    {
+      "epoch": 0.6824146981627297,
+      "grad_norm": 9.077951487605079,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 2787
+    },
+    {
+      "epoch": 0.6850393700787402,
+      "grad_norm": 10.352188358393105,
+      "learning_rate": 1e-06,
+      "loss": 0.1277,
+      "step": 2788
+    },
+    {
+      "epoch": 0.6876640419947506,
+      "grad_norm": 7.874861147028414,
+      "learning_rate": 1e-06,
+      "loss": 0.1243,
+      "step": 2789
+    },
+    {
+      "epoch": 0.6902887139107612,
+      "grad_norm": 7.9675706741640235,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 2790
+    },
+    {
+      "epoch": 0.6929133858267716,
+      "grad_norm": 8.887958224391618,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 2791
+    },
+    {
+      "epoch": 0.6955380577427821,
+      "grad_norm": 9.167791655321894,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 2792
+    },
+    {
+      "epoch": 0.6981627296587927,
+      "grad_norm": 10.669297176577649,
+      "learning_rate": 1e-06,
+      "loss": 0.1339,
+      "step": 2793
+    },
+    {
+      "epoch": 0.7007874015748031,
+      "grad_norm": 10.8722903248789,
+      "learning_rate": 1e-06,
+      "loss": 0.1257,
+      "step": 2794
+    },
+    {
+      "epoch": 0.7034120734908137,
+      "grad_norm": 10.188919247028768,
+      "learning_rate": 1e-06,
+      "loss": 0.1244,
+      "step": 2795
+    },
+    {
+      "epoch": 0.7060367454068242,
+      "grad_norm": 11.755938920105109,
+      "learning_rate": 1e-06,
+      "loss": 0.1333,
+      "step": 2796
+    },
+    {
+      "epoch": 0.7086614173228346,
+      "grad_norm": 13.50543432080585,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 2797
+    },
+    {
+      "epoch": 0.7112860892388452,
+      "grad_norm": 9.817918423690198,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 2798
+    },
+    {
+      "epoch": 0.7139107611548556,
+      "grad_norm": 11.448196335547536,
+      "learning_rate": 1e-06,
+      "loss": 0.1266,
+      "step": 2799
+    },
+    {
+      "epoch": 0.7165354330708661,
+      "grad_norm": 10.999845429645182,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2800
+    },
+    {
+      "epoch": 0.7191601049868767,
+      "grad_norm": 7.956103546214499,
+      "learning_rate": 1e-06,
+      "loss": 0.1336,
+      "step": 2801
+    },
+    {
+      "epoch": 0.7217847769028871,
+      "grad_norm": 9.438466389730017,
+      "learning_rate": 1e-06,
+      "loss": 0.1306,
+      "step": 2802
+    },
+    {
+      "epoch": 0.7244094488188977,
+      "grad_norm": 9.214932002958431,
+      "learning_rate": 1e-06,
+      "loss": 0.118,
+      "step": 2803
+    },
+    {
+      "epoch": 0.7270341207349081,
+      "grad_norm": 10.813766340413029,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 2804
+    },
+    {
+      "epoch": 0.7296587926509186,
+      "grad_norm": 7.619758413907678,
+      "learning_rate": 1e-06,
+      "loss": 0.13,
+      "step": 2805
+    },
+    {
+      "epoch": 0.7322834645669292,
+      "grad_norm": 8.789196573652093,
+      "learning_rate": 1e-06,
+      "loss": 0.1202,
+      "step": 2806
+    },
+    {
+      "epoch": 0.7349081364829396,
+      "grad_norm": 8.830151069590734,
+      "learning_rate": 1e-06,
+      "loss": 0.1284,
+      "step": 2807
+    },
+    {
+      "epoch": 0.7375328083989501,
+      "grad_norm": 11.61604370942473,
+      "learning_rate": 1e-06,
+      "loss": 0.1259,
+      "step": 2808
+    },
+    {
+      "epoch": 0.7401574803149606,
+      "grad_norm": 11.497580267757414,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2809
+    },
+    {
+      "epoch": 0.7427821522309711,
+      "grad_norm": 14.325759795730054,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 2810
+    },
+    {
+      "epoch": 0.7454068241469817,
+      "grad_norm": 7.522260093149601,
+      "learning_rate": 1e-06,
+      "loss": 0.1299,
+      "step": 2811
+    },
+    {
+      "epoch": 0.7480314960629921,
+      "grad_norm": 8.45918677938265,
+      "learning_rate": 1e-06,
+      "loss": 0.136,
+      "step": 2812
+    },
+    {
+      "epoch": 0.7506561679790026,
+      "grad_norm": 10.900709969427904,
+      "learning_rate": 1e-06,
+      "loss": 0.1398,
+      "step": 2813
+    },
+    {
+      "epoch": 0.7532808398950132,
+      "grad_norm": 7.486313899489735,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 2814
+    },
+    {
+      "epoch": 0.7559055118110236,
+      "grad_norm": 10.94047055700859,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2815
+    },
+    {
+      "epoch": 0.7585301837270341,
+      "grad_norm": 9.341124161929523,
+      "learning_rate": 1e-06,
+      "loss": 0.1231,
+      "step": 2816
+    },
+    {
+      "epoch": 0.7611548556430446,
+      "grad_norm": 7.562384809482976,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 2817
+    },
+    {
+      "epoch": 0.7637795275590551,
+      "grad_norm": 9.896954466950646,
+      "learning_rate": 1e-06,
+      "loss": 0.1321,
+      "step": 2818
+    },
+    {
+      "epoch": 0.7664041994750657,
+      "grad_norm": 8.909504756090067,
+      "learning_rate": 1e-06,
+      "loss": 0.1176,
+      "step": 2819
+    },
+    {
+      "epoch": 0.7690288713910761,
+      "grad_norm": 10.879895929258963,
+      "learning_rate": 1e-06,
+      "loss": 0.1291,
+      "step": 2820
+    },
+    {
+      "epoch": 0.7716535433070866,
+      "grad_norm": 12.338454498757056,
+      "learning_rate": 1e-06,
+      "loss": 0.1369,
+      "step": 2821
+    },
+    {
+      "epoch": 0.7742782152230971,
+      "grad_norm": 9.574642589332681,
+      "learning_rate": 1e-06,
+      "loss": 0.127,
+      "step": 2822
+    },
+    {
+      "epoch": 0.7769028871391076,
+      "grad_norm": 11.768773470213159,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 2823
+    },
+    {
+      "epoch": 0.7795275590551181,
+      "grad_norm": 10.75035629217642,
+      "learning_rate": 1e-06,
+      "loss": 0.1392,
+      "step": 2824
+    },
+    {
+      "epoch": 0.7821522309711286,
+      "grad_norm": 10.282378702702918,
+      "learning_rate": 1e-06,
+      "loss": 0.12,
+      "step": 2825
+    },
+    {
+      "epoch": 0.7847769028871391,
+      "grad_norm": 10.535621051308688,
+      "learning_rate": 1e-06,
+      "loss": 0.1372,
+      "step": 2826
+    },
+    {
+      "epoch": 0.7874015748031497,
+      "grad_norm": 9.596826580363945,
+      "learning_rate": 1e-06,
+      "loss": 0.1209,
+      "step": 2827
+    },
+    {
+      "epoch": 0.7900262467191601,
+      "grad_norm": 14.09626615893552,
+      "learning_rate": 1e-06,
+      "loss": 0.1412,
+      "step": 2828
+    },
+    {
+      "epoch": 0.7926509186351706,
+      "grad_norm": 10.04440963165168,
+      "learning_rate": 1e-06,
+      "loss": 0.1256,
+      "step": 2829
+    },
+    {
+      "epoch": 0.7952755905511811,
+      "grad_norm": 7.852686372103919,
+      "learning_rate": 1e-06,
+      "loss": 0.1291,
+      "step": 2830
+    },
+    {
+      "epoch": 0.7979002624671916,
+      "grad_norm": 11.393502794547123,
+      "learning_rate": 1e-06,
+      "loss": 0.1308,
+      "step": 2831
+    },
+    {
+      "epoch": 0.800524934383202,
+      "grad_norm": 9.775844483398066,
+      "learning_rate": 1e-06,
+      "loss": 0.1369,
+      "step": 2832
+    },
+    {
+      "epoch": 0.8031496062992126,
+      "grad_norm": 9.199283489856986,
+      "learning_rate": 1e-06,
+      "loss": 0.1305,
+      "step": 2833
+    },
+    {
+      "epoch": 0.8057742782152231,
+      "grad_norm": 10.847389832360632,
+      "learning_rate": 1e-06,
+      "loss": 0.1258,
+      "step": 2834
+    },
+    {
+      "epoch": 0.8083989501312336,
+      "grad_norm": 10.987707651257825,
+      "learning_rate": 1e-06,
+      "loss": 0.1351,
+      "step": 2835
+    },
+    {
+      "epoch": 0.8110236220472441,
+      "grad_norm": 11.37256432545772,
+      "learning_rate": 1e-06,
+      "loss": 0.1359,
+      "step": 2836
+    },
+    {
+      "epoch": 0.8136482939632546,
+      "grad_norm": 12.528306792789603,
+      "learning_rate": 1e-06,
+      "loss": 0.1352,
+      "step": 2837
+    },
+    {
+      "epoch": 0.8162729658792651,
+      "grad_norm": 12.80873401793862,
+      "learning_rate": 1e-06,
+      "loss": 0.1373,
+      "step": 2838
+    },
+    {
+      "epoch": 0.8188976377952756,
+      "grad_norm": 8.330890767538751,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 2839
+    },
+    {
+      "epoch": 0.821522309711286,
+      "grad_norm": 12.755415243468061,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 2840
+    },
+    {
+      "epoch": 0.8241469816272966,
+      "grad_norm": 6.784636303051179,
+      "learning_rate": 1e-06,
+      "loss": 0.1283,
+      "step": 2841
+    },
+    {
+      "epoch": 0.8267716535433071,
+      "grad_norm": 9.472502407835307,
+      "learning_rate": 1e-06,
+      "loss": 0.1413,
+      "step": 2842
+    },
+    {
+      "epoch": 0.8293963254593176,
+      "grad_norm": 11.762144589467228,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 2843
+    },
+    {
+      "epoch": 0.8320209973753281,
+      "grad_norm": 10.760229645733869,
+      "learning_rate": 1e-06,
+      "loss": 0.1331,
+      "step": 2844
+    },
+    {
+      "epoch": 0.8346456692913385,
+      "grad_norm": 8.978022148591192,
+      "learning_rate": 1e-06,
+      "loss": 0.1282,
+      "step": 2845
+    },
+    {
+      "epoch": 0.8372703412073491,
+      "grad_norm": 12.218062921133676,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 2846
+    },
+    {
+      "epoch": 0.8398950131233596,
+      "grad_norm": 7.733265582625983,
+      "learning_rate": 1e-06,
+      "loss": 0.1383,
+      "step": 2847
+    },
+    {
+      "epoch": 0.84251968503937,
+      "grad_norm": 10.924195186252929,
+      "learning_rate": 1e-06,
+      "loss": 0.1265,
+      "step": 2848
+    },
+    {
+      "epoch": 0.8451443569553806,
+      "grad_norm": 10.956730055430445,
+      "learning_rate": 1e-06,
+      "loss": 0.1272,
+      "step": 2849
+    },
+    {
+      "epoch": 0.847769028871391,
+      "grad_norm": 7.763571645101701,
+      "learning_rate": 1e-06,
+      "loss": 0.1304,
+      "step": 2850
+    },
+    {
+      "epoch": 0.8503937007874016,
+      "grad_norm": 11.386070247311196,
+      "learning_rate": 1e-06,
+      "loss": 0.1356,
+      "step": 2851
+    },
+    {
+      "epoch": 0.8530183727034121,
+      "grad_norm": 12.877297065804491,
+      "learning_rate": 1e-06,
+      "loss": 0.1439,
+      "step": 2852
+    },
+    {
+      "epoch": 0.8556430446194225,
+      "grad_norm": 11.196010104379598,
+      "learning_rate": 1e-06,
+      "loss": 0.1215,
+      "step": 2853
+    },
+    {
+      "epoch": 0.8582677165354331,
+      "grad_norm": 10.783785956047923,
+      "learning_rate": 1e-06,
+      "loss": 0.1221,
+      "step": 2854
+    },
+    {
+      "epoch": 0.8608923884514436,
+      "grad_norm": 10.823557669431768,
+      "learning_rate": 1e-06,
+      "loss": 0.1424,
+      "step": 2855
+    },
+    {
+      "epoch": 0.863517060367454,
+      "grad_norm": 9.281444490527065,
+      "learning_rate": 1e-06,
+      "loss": 0.1361,
+      "step": 2856
+    },
+    {
+      "epoch": 0.8661417322834646,
+      "grad_norm": 8.572265119698178,
+      "learning_rate": 1e-06,
+      "loss": 0.1334,
+      "step": 2857
+    },
+    {
+      "epoch": 0.868766404199475,
+      "grad_norm": 12.749932717080473,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 2858
+    },
+    {
+      "epoch": 0.8713910761154856,
+      "grad_norm": 12.253827594607094,
+      "learning_rate": 1e-06,
+      "loss": 0.1391,
+      "step": 2859
+    },
+    {
+      "epoch": 0.8740157480314961,
+      "grad_norm": 13.011174315904285,
+      "learning_rate": 1e-06,
+      "loss": 0.1411,
+      "step": 2860
+    },
+    {
+      "epoch": 0.8766404199475065,
+      "grad_norm": 12.36151610901074,
+      "learning_rate": 1e-06,
+      "loss": 0.1268,
+      "step": 2861
+    },
+    {
+      "epoch": 0.8792650918635171,
+      "grad_norm": 15.768131214551813,
+      "learning_rate": 1e-06,
+      "loss": 0.1385,
+      "step": 2862
+    },
+    {
+      "epoch": 0.8818897637795275,
+      "grad_norm": 10.318938512519013,
+      "learning_rate": 1e-06,
+      "loss": 0.124,
+      "step": 2863
+    },
+    {
+      "epoch": 0.884514435695538,
+      "grad_norm": 11.642282439521,
+      "learning_rate": 1e-06,
+      "loss": 0.1288,
+      "step": 2864
+    },
+    {
+      "epoch": 0.8871391076115486,
+      "grad_norm": 10.72934882048341,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 2865
+    },
+    {
+      "epoch": 0.889763779527559,
+      "grad_norm": 12.598566697529476,
+      "learning_rate": 1e-06,
+      "loss": 0.1365,
+      "step": 2866
+    },
+    {
+      "epoch": 0.8923884514435696,
+      "grad_norm": 10.6425063690354,
+      "learning_rate": 1e-06,
+      "loss": 0.1297,
+      "step": 2867
+    },
+    {
+      "epoch": 0.89501312335958,
+      "grad_norm": 11.40392029827406,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 2868
+    },
+    {
+      "epoch": 0.8976377952755905,
+      "grad_norm": 15.091570676956303,
+      "learning_rate": 1e-06,
+      "loss": 0.14,
+      "step": 2869
+    },
+    {
+      "epoch": 0.9002624671916011,
+      "grad_norm": 10.940067492792574,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 2870
+    },
+    {
+      "epoch": 0.9028871391076115,
+      "grad_norm": 9.35007015274826,
+      "learning_rate": 1e-06,
+      "loss": 0.1296,
+      "step": 2871
+    },
+    {
+      "epoch": 0.905511811023622,
+      "grad_norm": 12.795875092054747,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 2872
+    },
+    {
+      "epoch": 0.9081364829396326,
+      "grad_norm": 10.568794446905095,
+      "learning_rate": 1e-06,
+      "loss": 0.1296,
+      "step": 2873
+    },
+    {
+      "epoch": 0.910761154855643,
+      "grad_norm": 9.992155191806969,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2874
+    },
+    {
+      "epoch": 0.9133858267716536,
+      "grad_norm": 10.520276046384552,
+      "learning_rate": 1e-06,
+      "loss": 0.1178,
+      "step": 2875
+    },
+    {
+      "epoch": 0.916010498687664,
+      "grad_norm": 8.860886791235417,
+      "learning_rate": 1e-06,
+      "loss": 0.1242,
+      "step": 2876
+    },
+    {
+      "epoch": 0.9186351706036745,
+      "grad_norm": 9.959275773742153,
+      "learning_rate": 1e-06,
+      "loss": 0.1295,
+      "step": 2877
+    },
+    {
+      "epoch": 0.9212598425196851,
+      "grad_norm": 8.467827784058445,
+      "learning_rate": 1e-06,
+      "loss": 0.132,
+      "step": 2878
+    },
+    {
+      "epoch": 0.9238845144356955,
+      "grad_norm": 10.961915280739222,
+      "learning_rate": 1e-06,
+      "loss": 0.1338,
+      "step": 2879
+    },
+    {
+      "epoch": 0.926509186351706,
+      "grad_norm": 10.254039295014596,
+      "learning_rate": 1e-06,
+      "loss": 0.1335,
+      "step": 2880
+    },
+    {
+      "epoch": 0.9291338582677166,
+      "grad_norm": 12.658122219819074,
+      "learning_rate": 1e-06,
+      "loss": 0.1442,
+      "step": 2881
+    },
+    {
+      "epoch": 0.931758530183727,
+      "grad_norm": 9.401190975473895,
+      "learning_rate": 1e-06,
+      "loss": 0.1294,
+      "step": 2882
+    },
+    {
+      "epoch": 0.9343832020997376,
+      "grad_norm": 9.995535315091841,
+      "learning_rate": 1e-06,
+      "loss": 0.1274,
+      "step": 2883
+    },
+    {
+      "epoch": 0.937007874015748,
+      "grad_norm": 9.688417167674249,
+      "learning_rate": 1e-06,
+      "loss": 0.1237,
+      "step": 2884
+    },
+    {
+      "epoch": 0.9396325459317585,
+      "grad_norm": 9.44987581753076,
+      "learning_rate": 1e-06,
+      "loss": 0.1267,
+      "step": 2885
+    },
+    {
+      "epoch": 0.9422572178477691,
+      "grad_norm": 9.795646670115833,
+      "learning_rate": 1e-06,
+      "loss": 0.1354,
+      "step": 2886
+    },
+    {
+      "epoch": 0.9448818897637795,
+      "grad_norm": 9.991207317945973,
+      "learning_rate": 1e-06,
+      "loss": 0.131,
+      "step": 2887
+    },
+    {
+      "epoch": 0.94750656167979,
+      "grad_norm": 13.320795198499647,
+      "learning_rate": 1e-06,
+      "loss": 0.139,
+      "step": 2888
+    },
+    {
+      "epoch": 0.9501312335958005,
+      "grad_norm": 11.802530998491159,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 2889
+    },
+    {
+      "epoch": 0.952755905511811,
+      "grad_norm": 7.565204360368046,
+      "learning_rate": 1e-06,
+      "loss": 0.134,
+      "step": 2890
+    },
+    {
+      "epoch": 0.9553805774278216,
+      "grad_norm": 9.891092893091997,
+      "learning_rate": 1e-06,
+      "loss": 0.1314,
+      "step": 2891
+    },
+    {
+      "epoch": 0.958005249343832,
+      "grad_norm": 8.491193836454332,
+      "learning_rate": 1e-06,
+      "loss": 0.122,
+      "step": 2892
+    },
+    {
+      "epoch": 0.9606299212598425,
+      "grad_norm": 10.260093909604468,
+      "learning_rate": 1e-06,
+      "loss": 0.1387,
+      "step": 2893
+    },
+    {
+      "epoch": 0.963254593175853,
+      "grad_norm": 8.216169352946142,
+      "learning_rate": 1e-06,
+      "loss": 0.1247,
+      "step": 2894
+    },
+    {
+      "epoch": 0.9658792650918635,
+      "grad_norm": 7.90140116064533,
+      "learning_rate": 1e-06,
+      "loss": 0.1263,
+      "step": 2895
+    },
+    {
+      "epoch": 0.968503937007874,
+      "grad_norm": 10.310316525777711,
+      "learning_rate": 1e-06,
+      "loss": 0.1246,
+      "step": 2896
+    },
+    {
+      "epoch": 0.9711286089238845,
+      "grad_norm": 10.482796621993808,
+      "learning_rate": 1e-06,
+      "loss": 0.1327,
+      "step": 2897
+    },
+    {
+      "epoch": 0.973753280839895,
+      "grad_norm": 11.740845097924447,
+      "learning_rate": 1e-06,
+      "loss": 0.1362,
+      "step": 2898
+    },
+    {
+      "epoch": 0.9763779527559056,
+      "grad_norm": 11.29846413851584,
+      "learning_rate": 1e-06,
+      "loss": 0.1313,
+      "step": 2899
+    },
+    {
+      "epoch": 0.979002624671916,
+      "grad_norm": 11.644850750066743,
+      "learning_rate": 1e-06,
+      "loss": 0.128,
+      "step": 2900
+    },
+    {
+      "epoch": 0.9816272965879265,
+      "grad_norm": 8.16125006823714,
+      "learning_rate": 1e-06,
+      "loss": 0.1265,
+      "step": 2901
+    },
+    {
+      "epoch": 0.984251968503937,
+      "grad_norm": 9.991791002991667,
+      "learning_rate": 1e-06,
+      "loss": 0.1249,
+      "step": 2902
+    },
+    {
+      "epoch": 0.9868766404199475,
+      "grad_norm": 14.452220286175518,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 2903
+    },
+    {
+      "epoch": 0.989501312335958,
+      "grad_norm": 12.31180203214049,
+      "learning_rate": 1e-06,
+      "loss": 0.1355,
+      "step": 2904
+    },
+    {
+      "epoch": 0.9921259842519685,
+      "grad_norm": 7.794949917448917,
+      "learning_rate": 1e-06,
+      "loss": 0.1333,
+      "step": 2905
+    },
+    {
+      "epoch": 0.994750656167979,
+      "grad_norm": 12.544923164067537,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 2906
+    },
+    {
+      "epoch": 0.9973753280839895,
+      "grad_norm": 11.364043786610958,
+      "learning_rate": 1e-06,
+      "loss": 0.1348,
+      "step": 2907
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.567962930340413,
+      "learning_rate": 1e-06,
+      "loss": 0.1193,
+      "step": 2908
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.467581033706665,
+      "eval_runtime": 21.924,
+      "eval_samples_per_second": 45.612,
+      "eval_steps_per_second": 5.702,
+      "step": 2908
+    },
+    {
+      "epoch": 0.0025031289111389237,
+      "grad_norm": 9.592962299926755,
+      "learning_rate": 1e-06,
+      "loss": 0.173,
+      "step": 2909
+    },
+    {
+      "epoch": 0.0050062578222778474,
+      "grad_norm": 9.657263474948872,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 2910
+    },
+    {
+      "epoch": 0.007509386733416771,
+      "grad_norm": 7.422357134060183,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 2911
+    },
+    {
+      "epoch": 0.010012515644555695,
+      "grad_norm": 10.33172162739967,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 2912
+    },
+    {
+      "epoch": 0.012515644555694618,
+      "grad_norm": 8.657527686288937,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 2913
+    },
+    {
+      "epoch": 0.015018773466833541,
+      "grad_norm": 9.629305482653518,
+      "learning_rate": 1e-06,
+      "loss": 0.1722,
+      "step": 2914
+    },
+    {
+      "epoch": 0.017521902377972465,
+      "grad_norm": 11.157142173022232,
+      "learning_rate": 1e-06,
+      "loss": 0.166,
+      "step": 2915
+    },
+    {
+      "epoch": 0.02002503128911139,
+      "grad_norm": 10.21745620096707,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 2916
+    },
+    {
+      "epoch": 0.02252816020025031,
+      "grad_norm": 11.465710406409809,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 2917
+    },
+    {
+      "epoch": 0.025031289111389236,
+      "grad_norm": 12.22331329773291,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 2918
+    },
+    {
+      "epoch": 0.02753441802252816,
+      "grad_norm": 11.08402405070039,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 2919
+    },
+    {
+      "epoch": 0.030037546933667083,
+      "grad_norm": 11.430239439607407,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 2920
+    },
+    {
+      "epoch": 0.03254067584480601,
+      "grad_norm": 11.717000075636303,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 2921
+    },
+    {
+      "epoch": 0.03504380475594493,
+      "grad_norm": 9.657825099701537,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 2922
+    },
+    {
+      "epoch": 0.03754693366708386,
+      "grad_norm": 10.307243988942096,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 2923
+    },
+    {
+      "epoch": 0.04005006257822278,
+      "grad_norm": 9.325483934059841,
+      "learning_rate": 1e-06,
+      "loss": 0.1759,
+      "step": 2924
+    },
+    {
+      "epoch": 0.0425531914893617,
+      "grad_norm": 8.039846492348365,
+      "learning_rate": 1e-06,
+      "loss": 0.1747,
+      "step": 2925
+    },
+    {
+      "epoch": 0.04505632040050062,
+      "grad_norm": 10.132770923492073,
+      "learning_rate": 1e-06,
+      "loss": 0.1766,
+      "step": 2926
+    },
+    {
+      "epoch": 0.04755944931163955,
+      "grad_norm": 11.751728670803725,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 2927
+    },
+    {
+      "epoch": 0.05006257822277847,
+      "grad_norm": 11.688832236643721,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 2928
+    },
+    {
+      "epoch": 0.052565707133917394,
+      "grad_norm": 10.486836471623677,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 2929
+    },
+    {
+      "epoch": 0.05506883604505632,
+      "grad_norm": 9.2967039955314,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 2930
+    },
+    {
+      "epoch": 0.057571964956195244,
+      "grad_norm": 7.799487468965689,
+      "learning_rate": 1e-06,
+      "loss": 0.1719,
+      "step": 2931
+    },
+    {
+      "epoch": 0.060075093867334166,
+      "grad_norm": 12.488875287306904,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 2932
+    },
+    {
+      "epoch": 0.0625782227784731,
+      "grad_norm": 11.46949632366066,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 2933
+    },
+    {
+      "epoch": 0.06508135168961202,
+      "grad_norm": 9.627553625232,
+      "learning_rate": 1e-06,
+      "loss": 0.1597,
+      "step": 2934
+    },
+    {
+      "epoch": 0.06758448060075094,
+      "grad_norm": 10.795422449147644,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 2935
+    },
+    {
+      "epoch": 0.07008760951188986,
+      "grad_norm": 11.129831621484598,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 2936
+    },
+    {
+      "epoch": 0.07259073842302878,
+      "grad_norm": 11.129927097726819,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 2937
+    },
+    {
+      "epoch": 0.07509386733416772,
+      "grad_norm": 11.354526427879193,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 2938
+    },
+    {
+      "epoch": 0.07759699624530664,
+      "grad_norm": 11.117976061149145,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 2939
+    },
+    {
+      "epoch": 0.08010012515644556,
+      "grad_norm": 8.5871206132048,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 2940
+    },
+    {
+      "epoch": 0.08260325406758448,
+      "grad_norm": 13.900941958588097,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 2941
+    },
+    {
+      "epoch": 0.0851063829787234,
+      "grad_norm": 8.999796335247133,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 2942
+    },
+    {
+      "epoch": 0.08760951188986232,
+      "grad_norm": 9.49999368351086,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 2943
+    },
+    {
+      "epoch": 0.09011264080100125,
+      "grad_norm": 8.481205258804545,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 2944
+    },
+    {
+      "epoch": 0.09261576971214018,
+      "grad_norm": 10.896015271668478,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 2945
+    },
+    {
+      "epoch": 0.0951188986232791,
+      "grad_norm": 11.795651338480772,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 2946
+    },
+    {
+      "epoch": 0.09762202753441802,
+      "grad_norm": 11.962828952085378,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 2947
+    },
+    {
+      "epoch": 0.10012515644555695,
+      "grad_norm": 8.236930627207306,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 2948
+    },
+    {
+      "epoch": 0.10262828535669587,
+      "grad_norm": 9.275421143306081,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 2949
+    },
+    {
+      "epoch": 0.10513141426783479,
+      "grad_norm": 8.08848095332067,
+      "learning_rate": 1e-06,
+      "loss": 0.1707,
+      "step": 2950
+    },
+    {
+      "epoch": 0.10763454317897372,
+      "grad_norm": 8.370590811369853,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 2951
+    },
+    {
+      "epoch": 0.11013767209011265,
+      "grad_norm": 10.817675119618784,
+      "learning_rate": 1e-06,
+      "loss": 0.1654,
+      "step": 2952
+    },
+    {
+      "epoch": 0.11264080100125157,
+      "grad_norm": 8.898519599210738,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 2953
+    },
+    {
+      "epoch": 0.11514392991239049,
+      "grad_norm": 10.926600500358889,
+      "learning_rate": 1e-06,
+      "loss": 0.1686,
+      "step": 2954
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 10.306200073182,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 2955
+    },
+    {
+      "epoch": 0.12015018773466833,
+      "grad_norm": 9.0534125521556,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 2956
+    },
+    {
+      "epoch": 0.12265331664580725,
+      "grad_norm": 8.0959343128734,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 2957
+    },
+    {
+      "epoch": 0.1251564455569462,
+      "grad_norm": 9.810357684864963,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 2958
+    },
+    {
+      "epoch": 0.1276595744680851,
+      "grad_norm": 9.221609611940817,
+      "learning_rate": 1e-06,
+      "loss": 0.1697,
+      "step": 2959
+    },
+    {
+      "epoch": 0.13016270337922403,
+      "grad_norm": 7.650321055517909,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 2960
+    },
+    {
+      "epoch": 0.13266583229036297,
+      "grad_norm": 8.479438965526718,
+      "learning_rate": 1e-06,
+      "loss": 0.1569,
+      "step": 2961
+    },
+    {
+      "epoch": 0.13516896120150187,
+      "grad_norm": 9.031471810341504,
+      "learning_rate": 1e-06,
+      "loss": 0.1759,
+      "step": 2962
+    },
+    {
+      "epoch": 0.1376720901126408,
+      "grad_norm": 9.43617641144616,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 2963
+    },
+    {
+      "epoch": 0.14017521902377972,
+      "grad_norm": 9.853485060541406,
+      "learning_rate": 1e-06,
+      "loss": 0.1703,
+      "step": 2964
+    },
+    {
+      "epoch": 0.14267834793491865,
+      "grad_norm": 9.9449656812837,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 2965
+    },
+    {
+      "epoch": 0.14518147684605756,
+      "grad_norm": 7.809681524569133,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 2966
+    },
+    {
+      "epoch": 0.1476846057571965,
+      "grad_norm": 10.034770976880809,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 2967
+    },
+    {
+      "epoch": 0.15018773466833543,
+      "grad_norm": 8.841187746332121,
+      "learning_rate": 1e-06,
+      "loss": 0.1726,
+      "step": 2968
+    },
+    {
+      "epoch": 0.15269086357947434,
+      "grad_norm": 9.548461346644492,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 2969
+    },
+    {
+      "epoch": 0.15519399249061328,
+      "grad_norm": 7.59251044152144,
+      "learning_rate": 1e-06,
+      "loss": 0.1705,
+      "step": 2970
+    },
+    {
+      "epoch": 0.15769712140175218,
+      "grad_norm": 9.077578028372773,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 2971
+    },
+    {
+      "epoch": 0.16020025031289112,
+      "grad_norm": 9.042301408586178,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 2972
+    },
+    {
+      "epoch": 0.16270337922403003,
+      "grad_norm": 9.275230483169553,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 2973
+    },
+    {
+      "epoch": 0.16520650813516896,
+      "grad_norm": 11.906297018624441,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 2974
+    },
+    {
+      "epoch": 0.1677096370463079,
+      "grad_norm": 10.262650161981519,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 2975
+    },
+    {
+      "epoch": 0.1702127659574468,
+      "grad_norm": 11.766802020191406,
+      "learning_rate": 1e-06,
+      "loss": 0.1668,
+      "step": 2976
+    },
+    {
+      "epoch": 0.17271589486858574,
+      "grad_norm": 8.975226091903405,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 2977
+    },
+    {
+      "epoch": 0.17521902377972465,
+      "grad_norm": 8.867637388830914,
+      "learning_rate": 1e-06,
+      "loss": 0.1885,
+      "step": 2978
+    },
+    {
+      "epoch": 0.17772215269086358,
+      "grad_norm": 8.655650685562756,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 2979
+    },
+    {
+      "epoch": 0.1802252816020025,
+      "grad_norm": 10.217773085671373,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 2980
+    },
+    {
+      "epoch": 0.18272841051314143,
+      "grad_norm": 7.97747113306852,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 2981
+    },
+    {
+      "epoch": 0.18523153942428036,
+      "grad_norm": 10.435703384869484,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 2982
+    },
+    {
+      "epoch": 0.18773466833541927,
+      "grad_norm": 7.251091673576298,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 2983
+    },
+    {
+      "epoch": 0.1902377972465582,
+      "grad_norm": 11.10039792806063,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 2984
+    },
+    {
+      "epoch": 0.1927409261576971,
+      "grad_norm": 9.073480050179619,
+      "learning_rate": 1e-06,
+      "loss": 0.165,
+      "step": 2985
+    },
+    {
+      "epoch": 0.19524405506883605,
+      "grad_norm": 10.589846724031707,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 2986
+    },
+    {
+      "epoch": 0.19774718397997496,
+      "grad_norm": 9.718371899421088,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 2987
+    },
+    {
+      "epoch": 0.2002503128911139,
+      "grad_norm": 9.995560543849988,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 2988
+    },
+    {
+      "epoch": 0.20275344180225283,
+      "grad_norm": 9.63568616071078,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 2989
+    },
+    {
+      "epoch": 0.20525657071339173,
+      "grad_norm": 9.831023668535531,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 2990
+    },
+    {
+      "epoch": 0.20775969962453067,
+      "grad_norm": 11.52745955119458,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 2991
+    },
+    {
+      "epoch": 0.21026282853566958,
+      "grad_norm": 7.638386870386563,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 2992
+    },
+    {
+      "epoch": 0.2127659574468085,
+      "grad_norm": 10.189372644415995,
+      "learning_rate": 1e-06,
+      "loss": 0.1712,
+      "step": 2993
+    },
+    {
+      "epoch": 0.21526908635794745,
+      "grad_norm": 10.447829360872628,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 2994
+    },
+    {
+      "epoch": 0.21777221526908636,
+      "grad_norm": 9.071955014707543,
+      "learning_rate": 1e-06,
+      "loss": 0.1681,
+      "step": 2995
+    },
+    {
+      "epoch": 0.2202753441802253,
+      "grad_norm": 9.971556750408219,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 2996
+    },
+    {
+      "epoch": 0.2227784730913642,
+      "grad_norm": 10.558850742089179,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 2997
+    },
+    {
+      "epoch": 0.22528160200250313,
+      "grad_norm": 10.256026572804453,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 2998
+    },
+    {
+      "epoch": 0.22778473091364204,
+      "grad_norm": 9.960218205952796,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 2999
+    },
+    {
+      "epoch": 0.23028785982478098,
+      "grad_norm": 9.015298752515822,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 3000
+    },
+    {
+      "epoch": 0.23028785982478098,
+      "eval_loss": 3.112135171890259,
+      "eval_runtime": 21.5781,
+      "eval_samples_per_second": 46.343,
+      "eval_steps_per_second": 5.793,
+      "step": 3000
+    },
+    {
+      "epoch": 0.2327909887359199,
+      "grad_norm": 8.623220047394518,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 3001
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 9.655557787377745,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 3002
+    },
+    {
+      "epoch": 0.23779724655819776,
+      "grad_norm": 9.776115872392133,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 3003
+    },
+    {
+      "epoch": 0.24030037546933666,
+      "grad_norm": 8.22700795313,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 3004
+    },
+    {
+      "epoch": 0.2428035043804756,
+      "grad_norm": 10.055507695233635,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 3005
+    },
+    {
+      "epoch": 0.2453066332916145,
+      "grad_norm": 9.866391939033818,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 3006
+    },
+    {
+      "epoch": 0.24780976220275344,
+      "grad_norm": 9.868376188697432,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 3007
+    },
+    {
+      "epoch": 0.2503128911138924,
+      "grad_norm": 11.101830304926999,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 3008
+    },
+    {
+      "epoch": 0.2528160200250313,
+      "grad_norm": 8.40451751110084,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 3009
+    },
+    {
+      "epoch": 0.2553191489361702,
+      "grad_norm": 10.572409958539538,
+      "learning_rate": 1e-06,
+      "loss": 0.1569,
+      "step": 3010
+    },
+    {
+      "epoch": 0.25782227784730916,
+      "grad_norm": 12.440654240963788,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3011
+    },
+    {
+      "epoch": 0.26032540675844806,
+      "grad_norm": 11.84762119200142,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3012
+    },
+    {
+      "epoch": 0.26282853566958697,
+      "grad_norm": 9.183649779564226,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 3013
+    },
+    {
+      "epoch": 0.26533166458072593,
+      "grad_norm": 7.807487763816631,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 3014
+    },
+    {
+      "epoch": 0.26783479349186484,
+      "grad_norm": 8.857737254528704,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 3015
+    },
+    {
+      "epoch": 0.27033792240300375,
+      "grad_norm": 8.733356467278073,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 3016
+    },
+    {
+      "epoch": 0.27284105131414266,
+      "grad_norm": 10.170963257822507,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 3017
+    },
+    {
+      "epoch": 0.2753441802252816,
+      "grad_norm": 11.562022220892837,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 3018
+    },
+    {
+      "epoch": 0.27784730913642053,
+      "grad_norm": 9.78947039381912,
+      "learning_rate": 1e-06,
+      "loss": 0.1735,
+      "step": 3019
+    },
+    {
+      "epoch": 0.28035043804755944,
+      "grad_norm": 9.733926842624516,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 3020
+    },
+    {
+      "epoch": 0.2828535669586984,
+      "grad_norm": 7.879783399391874,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 3021
+    },
+    {
+      "epoch": 0.2853566958698373,
+      "grad_norm": 9.541346610122979,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 3022
+    },
+    {
+      "epoch": 0.2878598247809762,
+      "grad_norm": 8.290576550704843,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 3023
+    },
+    {
+      "epoch": 0.2903629536921151,
+      "grad_norm": 8.326766385093324,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 3024
+    },
+    {
+      "epoch": 0.2928660826032541,
+      "grad_norm": 8.55352737815791,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 3025
+    },
+    {
+      "epoch": 0.295369211514393,
+      "grad_norm": 12.451934838190683,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 3026
+    },
+    {
+      "epoch": 0.2978723404255319,
+      "grad_norm": 9.249980411182081,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 3027
+    },
+    {
+      "epoch": 0.30037546933667086,
+      "grad_norm": 8.903214193319236,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 3028
+    },
+    {
+      "epoch": 0.30287859824780977,
+      "grad_norm": 11.821277267382415,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3029
+    },
+    {
+      "epoch": 0.3053817271589487,
+      "grad_norm": 12.775009098245917,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 3030
+    },
+    {
+      "epoch": 0.3078848560700876,
+      "grad_norm": 8.423972388812636,
+      "learning_rate": 1e-06,
+      "loss": 0.1689,
+      "step": 3031
+    },
+    {
+      "epoch": 0.31038798498122655,
+      "grad_norm": 6.500678364895759,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 3032
+    },
+    {
+      "epoch": 0.31289111389236546,
+      "grad_norm": 11.082539352212374,
+      "learning_rate": 1e-06,
+      "loss": 0.1694,
+      "step": 3033
+    },
+    {
+      "epoch": 0.31539424280350437,
+      "grad_norm": 8.209730866331338,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 3034
+    },
+    {
+      "epoch": 0.31789737171464333,
+      "grad_norm": 10.877601098220927,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 3035
+    },
+    {
+      "epoch": 0.32040050062578224,
+      "grad_norm": 10.121673599281488,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 3036
+    },
+    {
+      "epoch": 0.32290362953692114,
+      "grad_norm": 9.140227734458863,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 3037
+    },
+    {
+      "epoch": 0.32540675844806005,
+      "grad_norm": 11.174108385338023,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 3038
+    },
+    {
+      "epoch": 0.327909887359199,
+      "grad_norm": 10.019810948525247,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 3039
+    },
+    {
+      "epoch": 0.3304130162703379,
+      "grad_norm": 7.349548680320085,
+      "learning_rate": 1e-06,
+      "loss": 0.1714,
+      "step": 3040
+    },
+    {
+      "epoch": 0.33291614518147683,
+      "grad_norm": 10.695855548246488,
+      "learning_rate": 1e-06,
+      "loss": 0.1732,
+      "step": 3041
+    },
+    {
+      "epoch": 0.3354192740926158,
+      "grad_norm": 10.529880171567585,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 3042
+    },
+    {
+      "epoch": 0.3379224030037547,
+      "grad_norm": 10.81095847500232,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 3043
+    },
+    {
+      "epoch": 0.3404255319148936,
+      "grad_norm": 8.801821897212681,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 3044
+    },
+    {
+      "epoch": 0.3429286608260325,
+      "grad_norm": 10.8804628817339,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 3045
+    },
+    {
+      "epoch": 0.3454317897371715,
+      "grad_norm": 9.882722387355349,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 3046
+    },
+    {
+      "epoch": 0.3479349186483104,
+      "grad_norm": 19.86084128820019,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 3047
+    },
+    {
+      "epoch": 0.3504380475594493,
+      "grad_norm": 11.398326299475782,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 3048
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 9.147848193820016,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 3049
+    },
+    {
+      "epoch": 0.35544430538172717,
+      "grad_norm": 11.578864704450073,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 3050
+    },
+    {
+      "epoch": 0.3579474342928661,
+      "grad_norm": 7.189838249651743,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 3051
+    },
+    {
+      "epoch": 0.360450563204005,
+      "grad_norm": 8.81138339814454,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 3052
+    },
+    {
+      "epoch": 0.36295369211514394,
+      "grad_norm": 10.00562979896289,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 3053
+    },
+    {
+      "epoch": 0.36545682102628285,
+      "grad_norm": 9.299216204534298,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 3054
+    },
+    {
+      "epoch": 0.36795994993742176,
+      "grad_norm": 10.51231830059759,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 3055
+    },
+    {
+      "epoch": 0.3704630788485607,
+      "grad_norm": 9.538085464446477,
+      "learning_rate": 1e-06,
+      "loss": 0.1761,
+      "step": 3056
+    },
+    {
+      "epoch": 0.37296620775969963,
+      "grad_norm": 8.635703335402532,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 3057
+    },
+    {
+      "epoch": 0.37546933667083854,
+      "grad_norm": 8.808765520525832,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 3058
+    },
+    {
+      "epoch": 0.37797246558197745,
+      "grad_norm": 8.69438904922041,
+      "learning_rate": 1e-06,
+      "loss": 0.1676,
+      "step": 3059
+    },
+    {
+      "epoch": 0.3804755944931164,
+      "grad_norm": 10.702856188171173,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 3060
+    },
+    {
+      "epoch": 0.3829787234042553,
+      "grad_norm": 7.727777339764176,
+      "learning_rate": 1e-06,
+      "loss": 0.1801,
+      "step": 3061
+    },
+    {
+      "epoch": 0.3854818523153942,
+      "grad_norm": 8.06118622154845,
+      "learning_rate": 1e-06,
+      "loss": 0.1795,
+      "step": 3062
+    },
+    {
+      "epoch": 0.3879849812265332,
+      "grad_norm": 10.55606357645305,
+      "learning_rate": 1e-06,
+      "loss": 0.1836,
+      "step": 3063
+    },
+    {
+      "epoch": 0.3904881101376721,
+      "grad_norm": 11.16261833164057,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 3064
+    },
+    {
+      "epoch": 0.392991239048811,
+      "grad_norm": 10.163607169553082,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 3065
+    },
+    {
+      "epoch": 0.3954943679599499,
+      "grad_norm": 12.083748121824541,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 3066
+    },
+    {
+      "epoch": 0.3979974968710889,
+      "grad_norm": 10.677609151917437,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 3067
+    },
+    {
+      "epoch": 0.4005006257822278,
+      "grad_norm": 12.903684083582059,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 3068
+    },
+    {
+      "epoch": 0.4030037546933667,
+      "grad_norm": 8.012211184532827,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 3069
+    },
+    {
+      "epoch": 0.40550688360450565,
+      "grad_norm": 7.67808398952415,
+      "learning_rate": 1e-06,
+      "loss": 0.1795,
+      "step": 3070
+    },
+    {
+      "epoch": 0.40801001251564456,
+      "grad_norm": 9.228450316520988,
+      "learning_rate": 1e-06,
+      "loss": 0.1961,
+      "step": 3071
+    },
+    {
+      "epoch": 0.41051314142678347,
+      "grad_norm": 10.231563985894041,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 3072
+    },
+    {
+      "epoch": 0.4130162703379224,
+      "grad_norm": 10.020724815866773,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 3073
+    },
+    {
+      "epoch": 0.41551939924906134,
+      "grad_norm": 10.300260406213463,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 3074
+    },
+    {
+      "epoch": 0.41802252816020025,
+      "grad_norm": 7.767739147099826,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 3075
+    },
+    {
+      "epoch": 0.42052565707133915,
+      "grad_norm": 8.896373796979672,
+      "learning_rate": 1e-06,
+      "loss": 0.1843,
+      "step": 3076
+    },
+    {
+      "epoch": 0.4230287859824781,
+      "grad_norm": 10.053370645529656,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 3077
+    },
+    {
+      "epoch": 0.425531914893617,
+      "grad_norm": 8.181886627186355,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 3078
+    },
+    {
+      "epoch": 0.42803504380475593,
+      "grad_norm": 8.793059672801107,
+      "learning_rate": 1e-06,
+      "loss": 0.1632,
+      "step": 3079
+    },
+    {
+      "epoch": 0.4305381727158949,
+      "grad_norm": 8.961277847724546,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 3080
+    },
+    {
+      "epoch": 0.4330413016270338,
+      "grad_norm": 8.443167572161505,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 3081
+    },
+    {
+      "epoch": 0.4355444305381727,
+      "grad_norm": 9.896608100858886,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 3082
+    },
+    {
+      "epoch": 0.4380475594493116,
+      "grad_norm": 10.552038419095679,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3083
+    },
+    {
+      "epoch": 0.4405506883604506,
+      "grad_norm": 9.738622433514266,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 3084
+    },
+    {
+      "epoch": 0.4430538172715895,
+      "grad_norm": 9.810851991623434,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 3085
+    },
+    {
+      "epoch": 0.4455569461827284,
+      "grad_norm": 10.431154106092043,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 3086
+    },
+    {
+      "epoch": 0.44806007509386736,
+      "grad_norm": 9.93356587394905,
+      "learning_rate": 1e-06,
+      "loss": 0.1832,
+      "step": 3087
+    },
+    {
+      "epoch": 0.45056320400500627,
+      "grad_norm": 11.789445806385698,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 3088
+    },
+    {
+      "epoch": 0.4530663329161452,
+      "grad_norm": 7.946223346836491,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 3089
+    },
+    {
+      "epoch": 0.4555694618272841,
+      "grad_norm": 8.224152032023587,
+      "learning_rate": 1e-06,
+      "loss": 0.1644,
+      "step": 3090
+    },
+    {
+      "epoch": 0.45807259073842305,
+      "grad_norm": 11.422489229193275,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 3091
+    },
+    {
+      "epoch": 0.46057571964956195,
+      "grad_norm": 8.620668864320924,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 3092
+    },
+    {
+      "epoch": 0.46307884856070086,
+      "grad_norm": 6.841725430080166,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 3093
+    },
+    {
+      "epoch": 0.4655819774718398,
+      "grad_norm": 9.577222435684709,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 3094
+    },
+    {
+      "epoch": 0.46808510638297873,
+      "grad_norm": 11.5590038552628,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 3095
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 7.526987386766864,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 3096
+    },
+    {
+      "epoch": 0.47309136420525655,
+      "grad_norm": 10.207288709289797,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 3097
+    },
+    {
+      "epoch": 0.4755944931163955,
+      "grad_norm": 10.667230459314709,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 3098
+    },
+    {
+      "epoch": 0.4780976220275344,
+      "grad_norm": 9.814259947991765,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 3099
+    },
+    {
+      "epoch": 0.4806007509386733,
+      "grad_norm": 7.810526004850661,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 3100
+    },
+    {
+      "epoch": 0.4831038798498123,
+      "grad_norm": 6.670348121082551,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 3101
+    },
+    {
+      "epoch": 0.4856070087609512,
+      "grad_norm": 10.911632102426541,
+      "learning_rate": 1e-06,
+      "loss": 0.1678,
+      "step": 3102
+    },
+    {
+      "epoch": 0.4881101376720901,
+      "grad_norm": 11.381299020290236,
+      "learning_rate": 1e-06,
+      "loss": 0.1765,
+      "step": 3103
+    },
+    {
+      "epoch": 0.490613266583229,
+      "grad_norm": 10.102031964752824,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 3104
+    },
+    {
+      "epoch": 0.493116395494368,
+      "grad_norm": 10.740973170330916,
+      "learning_rate": 1e-06,
+      "loss": 0.183,
+      "step": 3105
+    },
+    {
+      "epoch": 0.4956195244055069,
+      "grad_norm": 12.235614219124948,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 3106
+    },
+    {
+      "epoch": 0.4981226533166458,
+      "grad_norm": 7.565222708628387,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 3107
+    },
+    {
+      "epoch": 0.5006257822277848,
+      "grad_norm": 10.08975394204979,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 3108
+    },
+    {
+      "epoch": 0.5031289111389237,
+      "grad_norm": 9.171281501969917,
+      "learning_rate": 1e-06,
+      "loss": 0.1704,
+      "step": 3109
+    },
+    {
+      "epoch": 0.5056320400500626,
+      "grad_norm": 9.170214391751243,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 3110
+    },
+    {
+      "epoch": 0.5081351689612015,
+      "grad_norm": 9.534486000806407,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 3111
+    },
+    {
+      "epoch": 0.5106382978723404,
+      "grad_norm": 9.66952321220942,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 3112
+    },
+    {
+      "epoch": 0.5131414267834794,
+      "grad_norm": 8.92747204045063,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 3113
+    },
+    {
+      "epoch": 0.5156445556946183,
+      "grad_norm": 9.25891448584172,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 3114
+    },
+    {
+      "epoch": 0.5181476846057572,
+      "grad_norm": 8.356843127806627,
+      "learning_rate": 1e-06,
+      "loss": 0.1834,
+      "step": 3115
+    },
+    {
+      "epoch": 0.5206508135168961,
+      "grad_norm": 11.729811566614034,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 3116
+    },
+    {
+      "epoch": 0.523153942428035,
+      "grad_norm": 11.393595123235611,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 3117
+    },
+    {
+      "epoch": 0.5256570713391739,
+      "grad_norm": 9.212647466102226,
+      "learning_rate": 1e-06,
+      "loss": 0.1738,
+      "step": 3118
+    },
+    {
+      "epoch": 0.5281602002503129,
+      "grad_norm": 11.617159744835526,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 3119
+    },
+    {
+      "epoch": 0.5306633291614519,
+      "grad_norm": 7.105384461257896,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 3120
+    },
+    {
+      "epoch": 0.5331664580725908,
+      "grad_norm": 12.226770219813165,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 3121
+    },
+    {
+      "epoch": 0.5356695869837297,
+      "grad_norm": 7.671719826053543,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 3122
+    },
+    {
+      "epoch": 0.5381727158948686,
+      "grad_norm": 10.200900340720796,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 3123
+    },
+    {
+      "epoch": 0.5406758448060075,
+      "grad_norm": 8.1200908479963,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 3124
+    },
+    {
+      "epoch": 0.5431789737171464,
+      "grad_norm": 8.137374664380287,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 3125
+    },
+    {
+      "epoch": 0.5456821026282853,
+      "grad_norm": 10.561164472399273,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 3126
+    },
+    {
+      "epoch": 0.5481852315394243,
+      "grad_norm": 11.698348028470209,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 3127
+    },
+    {
+      "epoch": 0.5506883604505632,
+      "grad_norm": 9.797887555108261,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 3128
+    },
+    {
+      "epoch": 0.5531914893617021,
+      "grad_norm": 10.150253285151793,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 3129
+    },
+    {
+      "epoch": 0.5556946182728411,
+      "grad_norm": 9.489091554385364,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 3130
+    },
+    {
+      "epoch": 0.55819774718398,
+      "grad_norm": 10.480721796720607,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 3131
+    },
+    {
+      "epoch": 0.5607008760951189,
+      "grad_norm": 8.979080707548421,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 3132
+    },
+    {
+      "epoch": 0.5632040050062578,
+      "grad_norm": 8.953515729462724,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 3133
+    },
+    {
+      "epoch": 0.5657071339173968,
+      "grad_norm": 11.41190128952382,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 3134
+    },
+    {
+      "epoch": 0.5682102628285357,
+      "grad_norm": 6.857179536872529,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 3135
+    },
+    {
+      "epoch": 0.5707133917396746,
+      "grad_norm": 11.328783342843378,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 3136
+    },
+    {
+      "epoch": 0.5732165206508135,
+      "grad_norm": 8.081912788486989,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 3137
+    },
+    {
+      "epoch": 0.5757196495619524,
+      "grad_norm": 10.636420777457143,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3138
+    },
+    {
+      "epoch": 0.5782227784730913,
+      "grad_norm": 7.122324543520396,
+      "learning_rate": 1e-06,
+      "loss": 0.184,
+      "step": 3139
+    },
+    {
+      "epoch": 0.5807259073842302,
+      "grad_norm": 8.928513259268206,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 3140
+    },
+    {
+      "epoch": 0.5832290362953693,
+      "grad_norm": 10.556379582552502,
+      "learning_rate": 1e-06,
+      "loss": 0.1797,
+      "step": 3141
+    },
+    {
+      "epoch": 0.5857321652065082,
+      "grad_norm": 7.95816106819265,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 3142
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 9.251155154815676,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 3143
+    },
+    {
+      "epoch": 0.590738423028786,
+      "grad_norm": 9.134423455050442,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 3144
+    },
+    {
+      "epoch": 0.5932415519399249,
+      "grad_norm": 9.594115618699604,
+      "learning_rate": 1e-06,
+      "loss": 0.1655,
+      "step": 3145
+    },
+    {
+      "epoch": 0.5957446808510638,
+      "grad_norm": 9.01460809239391,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 3146
+    },
+    {
+      "epoch": 0.5982478097622027,
+      "grad_norm": 10.916719535953352,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 3147
+    },
+    {
+      "epoch": 0.6007509386733417,
+      "grad_norm": 9.053487557597578,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 3148
+    },
+    {
+      "epoch": 0.6032540675844806,
+      "grad_norm": 10.007051267156415,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 3149
+    },
+    {
+      "epoch": 0.6057571964956195,
+      "grad_norm": 10.153304111328914,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 3150
+    },
+    {
+      "epoch": 0.6082603254067585,
+      "grad_norm": 9.176704371087556,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 3151
+    },
+    {
+      "epoch": 0.6107634543178974,
+      "grad_norm": 10.064900965598497,
+      "learning_rate": 1e-06,
+      "loss": 0.1835,
+      "step": 3152
+    },
+    {
+      "epoch": 0.6132665832290363,
+      "grad_norm": 11.45239191011179,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 3153
+    },
+    {
+      "epoch": 0.6157697121401752,
+      "grad_norm": 10.288167985820708,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 3154
+    },
+    {
+      "epoch": 0.6182728410513142,
+      "grad_norm": 11.682698509327002,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 3155
+    },
+    {
+      "epoch": 0.6207759699624531,
+      "grad_norm": 7.9086770580517465,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 3156
+    },
+    {
+      "epoch": 0.623279098873592,
+      "grad_norm": 9.312233766823061,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 3157
+    },
+    {
+      "epoch": 0.6257822277847309,
+      "grad_norm": 10.224299530104746,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 3158
+    },
+    {
+      "epoch": 0.6282853566958698,
+      "grad_norm": 8.563101523738661,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 3159
+    },
+    {
+      "epoch": 0.6307884856070087,
+      "grad_norm": 11.003759563176407,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 3160
+    },
+    {
+      "epoch": 0.6332916145181476,
+      "grad_norm": 11.048434766453608,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 3161
+    },
+    {
+      "epoch": 0.6357947434292867,
+      "grad_norm": 6.849339626905496,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 3162
+    },
+    {
+      "epoch": 0.6382978723404256,
+      "grad_norm": 8.205051262798392,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 3163
+    },
+    {
+      "epoch": 0.6408010012515645,
+      "grad_norm": 9.72760977259561,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 3164
+    },
+    {
+      "epoch": 0.6433041301627034,
+      "grad_norm": 11.215506881499495,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 3165
+    },
+    {
+      "epoch": 0.6458072590738423,
+      "grad_norm": 12.216770153050735,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 3166
+    },
+    {
+      "epoch": 0.6483103879849812,
+      "grad_norm": 7.259417590059672,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 3167
+    },
+    {
+      "epoch": 0.6508135168961201,
+      "grad_norm": 9.061647074489615,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 3168
+    },
+    {
+      "epoch": 0.6533166458072591,
+      "grad_norm": 10.235554140207869,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 3169
+    },
+    {
+      "epoch": 0.655819774718398,
+      "grad_norm": 8.11820752073032,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 3170
+    },
+    {
+      "epoch": 0.6583229036295369,
+      "grad_norm": 10.463521144990851,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 3171
+    },
+    {
+      "epoch": 0.6608260325406758,
+      "grad_norm": 9.037494711380033,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 3172
+    },
+    {
+      "epoch": 0.6633291614518148,
+      "grad_norm": 7.7496637755534366,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 3173
+    },
+    {
+      "epoch": 0.6658322903629537,
+      "grad_norm": 9.883028993348566,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 3174
+    },
+    {
+      "epoch": 0.6683354192740926,
+      "grad_norm": 9.62591919811024,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 3175
+    },
+    {
+      "epoch": 0.6708385481852316,
+      "grad_norm": 9.066129632025941,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 3176
+    },
+    {
+      "epoch": 0.6733416770963705,
+      "grad_norm": 9.562114875653688,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 3177
+    },
+    {
+      "epoch": 0.6758448060075094,
+      "grad_norm": 9.457315491723936,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 3178
+    },
+    {
+      "epoch": 0.6783479349186483,
+      "grad_norm": 8.194683629212335,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 3179
+    },
+    {
+      "epoch": 0.6808510638297872,
+      "grad_norm": 9.402221440293859,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3180
+    },
+    {
+      "epoch": 0.6833541927409261,
+      "grad_norm": 8.647595058192117,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 3181
+    },
+    {
+      "epoch": 0.685857321652065,
+      "grad_norm": 8.13534805152251,
+      "learning_rate": 1e-06,
+      "loss": 0.1816,
+      "step": 3182
+    },
+    {
+      "epoch": 0.688360450563204,
+      "grad_norm": 9.221969249949424,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 3183
+    },
+    {
+      "epoch": 0.690863579474343,
+      "grad_norm": 7.674308111517979,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 3184
+    },
+    {
+      "epoch": 0.6933667083854819,
+      "grad_norm": 7.988442335568194,
+      "learning_rate": 1e-06,
+      "loss": 0.2042,
+      "step": 3185
+    },
+    {
+      "epoch": 0.6958698372966208,
+      "grad_norm": 9.202375711188965,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 3186
+    },
+    {
+      "epoch": 0.6983729662077597,
+      "grad_norm": 9.393348110500627,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 3187
+    },
+    {
+      "epoch": 0.7008760951188986,
+      "grad_norm": 9.928977921423112,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 3188
+    },
+    {
+      "epoch": 0.7033792240300375,
+      "grad_norm": 9.024515697154408,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 3189
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 9.357967864303241,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 3190
+    },
+    {
+      "epoch": 0.7083854818523154,
+      "grad_norm": 8.349910389872786,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 3191
+    },
+    {
+      "epoch": 0.7108886107634543,
+      "grad_norm": 8.293285778799719,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 3192
+    },
+    {
+      "epoch": 0.7133917396745932,
+      "grad_norm": 10.040441818723629,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 3193
+    },
+    {
+      "epoch": 0.7158948685857321,
+      "grad_norm": 8.490402360864133,
+      "learning_rate": 1e-06,
+      "loss": 0.1662,
+      "step": 3194
+    },
+    {
+      "epoch": 0.718397997496871,
+      "grad_norm": 10.627374464992364,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 3195
+    },
+    {
+      "epoch": 0.72090112640801,
+      "grad_norm": 8.988790542702935,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 3196
+    },
+    {
+      "epoch": 0.723404255319149,
+      "grad_norm": 10.247285040251825,
+      "learning_rate": 1e-06,
+      "loss": 0.1667,
+      "step": 3197
+    },
+    {
+      "epoch": 0.7259073842302879,
+      "grad_norm": 11.306369114685936,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 3198
+    },
+    {
+      "epoch": 0.7284105131414268,
+      "grad_norm": 9.848336581695714,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 3199
+    },
+    {
+      "epoch": 0.7309136420525657,
+      "grad_norm": 7.69971336712777,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 3200
+    },
+    {
+      "epoch": 0.7334167709637046,
+      "grad_norm": 7.814046588342119,
+      "learning_rate": 1e-06,
+      "loss": 0.1721,
+      "step": 3201
+    },
+    {
+      "epoch": 0.7359198998748435,
+      "grad_norm": 8.797270875151598,
+      "learning_rate": 1e-06,
+      "loss": 0.1703,
+      "step": 3202
+    },
+    {
+      "epoch": 0.7384230287859824,
+      "grad_norm": 10.229731287972863,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 3203
+    },
+    {
+      "epoch": 0.7409261576971214,
+      "grad_norm": 12.858018093382578,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 3204
+    },
+    {
+      "epoch": 0.7434292866082604,
+      "grad_norm": 8.376448358079259,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 3205
+    },
+    {
+      "epoch": 0.7459324155193993,
+      "grad_norm": 9.252656387027088,
+      "learning_rate": 1e-06,
+      "loss": 0.1699,
+      "step": 3206
+    },
+    {
+      "epoch": 0.7484355444305382,
+      "grad_norm": 8.491960942437403,
+      "learning_rate": 1e-06,
+      "loss": 0.1789,
+      "step": 3207
+    },
+    {
+      "epoch": 0.7509386733416771,
+      "grad_norm": 9.862477614763739,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3208
+    },
+    {
+      "epoch": 0.753441802252816,
+      "grad_norm": 9.8162104641277,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 3209
+    },
+    {
+      "epoch": 0.7559449311639549,
+      "grad_norm": 9.26744329384079,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 3210
+    },
+    {
+      "epoch": 0.7584480600750939,
+      "grad_norm": 11.720549593474376,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 3211
+    },
+    {
+      "epoch": 0.7609511889862328,
+      "grad_norm": 10.30884162469122,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 3212
+    },
+    {
+      "epoch": 0.7634543178973717,
+      "grad_norm": 10.946479310768872,
+      "learning_rate": 1e-06,
+      "loss": 0.1657,
+      "step": 3213
+    },
+    {
+      "epoch": 0.7659574468085106,
+      "grad_norm": 9.094701915379574,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3214
+    },
+    {
+      "epoch": 0.7684605757196495,
+      "grad_norm": 8.033044123681726,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 3215
+    },
+    {
+      "epoch": 0.7709637046307884,
+      "grad_norm": 8.261354776505557,
+      "learning_rate": 1e-06,
+      "loss": 0.1643,
+      "step": 3216
+    },
+    {
+      "epoch": 0.7734668335419274,
+      "grad_norm": 9.81519439531196,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 3217
+    },
+    {
+      "epoch": 0.7759699624530664,
+      "grad_norm": 11.62145908069977,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 3218
+    },
+    {
+      "epoch": 0.7784730913642053,
+      "grad_norm": 13.182977373681748,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 3219
+    },
+    {
+      "epoch": 0.7809762202753442,
+      "grad_norm": 8.686039859617773,
+      "learning_rate": 1e-06,
+      "loss": 0.1824,
+      "step": 3220
+    },
+    {
+      "epoch": 0.7834793491864831,
+      "grad_norm": 13.244354280639481,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 3221
+    },
+    {
+      "epoch": 0.785982478097622,
+      "grad_norm": 8.59285889985264,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 3222
+    },
+    {
+      "epoch": 0.7884856070087609,
+      "grad_norm": 8.455692212948303,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 3223
+    },
+    {
+      "epoch": 0.7909887359198998,
+      "grad_norm": 8.126395876123706,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 3224
+    },
+    {
+      "epoch": 0.7934918648310388,
+      "grad_norm": 7.60343581925934,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 3225
+    },
+    {
+      "epoch": 0.7959949937421777,
+      "grad_norm": 7.322796407747814,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 3226
+    },
+    {
+      "epoch": 0.7984981226533167,
+      "grad_norm": 10.037557667887652,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 3227
+    },
+    {
+      "epoch": 0.8010012515644556,
+      "grad_norm": 9.822556489202762,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 3228
+    },
+    {
+      "epoch": 0.8035043804755945,
+      "grad_norm": 11.593035623352204,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 3229
+    },
+    {
+      "epoch": 0.8060075093867334,
+      "grad_norm": 7.473420922297409,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 3230
+    },
+    {
+      "epoch": 0.8085106382978723,
+      "grad_norm": 9.157460513507546,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 3231
+    },
+    {
+      "epoch": 0.8110137672090113,
+      "grad_norm": 10.496603657479755,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 3232
+    },
+    {
+      "epoch": 0.8135168961201502,
+      "grad_norm": 9.096328902007842,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 3233
+    },
+    {
+      "epoch": 0.8160200250312891,
+      "grad_norm": 8.785961258600826,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 3234
+    },
+    {
+      "epoch": 0.818523153942428,
+      "grad_norm": 9.41264206582,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 3235
+    },
+    {
+      "epoch": 0.8210262828535669,
+      "grad_norm": 7.021477843090568,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 3236
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 6.911367174697023,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 3237
+    },
+    {
+      "epoch": 0.8260325406758448,
+      "grad_norm": 9.61674801618471,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 3238
+    },
+    {
+      "epoch": 0.8285356695869838,
+      "grad_norm": 10.121977737213198,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 3239
+    },
+    {
+      "epoch": 0.8310387984981227,
+      "grad_norm": 8.069665728360228,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 3240
+    },
+    {
+      "epoch": 0.8335419274092616,
+      "grad_norm": 10.175923651112548,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 3241
+    },
+    {
+      "epoch": 0.8360450563204005,
+      "grad_norm": 9.252288748426801,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 3242
+    },
+    {
+      "epoch": 0.8385481852315394,
+      "grad_norm": 9.510163190443,
+      "learning_rate": 1e-06,
+      "loss": 0.1941,
+      "step": 3243
+    },
+    {
+      "epoch": 0.8410513141426783,
+      "grad_norm": 10.042538465666823,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 3244
+    },
+    {
+      "epoch": 0.8435544430538173,
+      "grad_norm": 9.335193975620435,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 3245
+    },
+    {
+      "epoch": 0.8460575719649562,
+      "grad_norm": 7.566471550760737,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3246
+    },
+    {
+      "epoch": 0.8485607008760951,
+      "grad_norm": 10.342744628251836,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 3247
+    },
+    {
+      "epoch": 0.851063829787234,
+      "grad_norm": 8.688292015663034,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 3248
+    },
+    {
+      "epoch": 0.853566958698373,
+      "grad_norm": 11.514049513907125,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 3249
+    },
+    {
+      "epoch": 0.8560700876095119,
+      "grad_norm": 10.632703285467466,
+      "learning_rate": 1e-06,
+      "loss": 0.167,
+      "step": 3250
+    },
+    {
+      "epoch": 0.8585732165206508,
+      "grad_norm": 10.365613314570192,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3251
+    },
+    {
+      "epoch": 0.8610763454317898,
+      "grad_norm": 11.70342534038006,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3252
+    },
+    {
+      "epoch": 0.8635794743429287,
+      "grad_norm": 9.72891277596022,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 3253
+    },
+    {
+      "epoch": 0.8660826032540676,
+      "grad_norm": 9.574940037278928,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 3254
+    },
+    {
+      "epoch": 0.8685857321652065,
+      "grad_norm": 7.489893878113175,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 3255
+    },
+    {
+      "epoch": 0.8710888610763454,
+      "grad_norm": 8.959059934867122,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 3256
+    },
+    {
+      "epoch": 0.8735919899874843,
+      "grad_norm": 10.584482996849287,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 3257
+    },
+    {
+      "epoch": 0.8760951188986232,
+      "grad_norm": 10.56249120720908,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 3258
+    },
+    {
+      "epoch": 0.8785982478097623,
+      "grad_norm": 10.30882407845294,
+      "learning_rate": 1e-06,
+      "loss": 0.1738,
+      "step": 3259
+    },
+    {
+      "epoch": 0.8811013767209012,
+      "grad_norm": 9.517548745087678,
+      "learning_rate": 1e-06,
+      "loss": 0.1824,
+      "step": 3260
+    },
+    {
+      "epoch": 0.8836045056320401,
+      "grad_norm": 9.368933335470981,
+      "learning_rate": 1e-06,
+      "loss": 0.1689,
+      "step": 3261
+    },
+    {
+      "epoch": 0.886107634543179,
+      "grad_norm": 8.413082602005428,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 3262
+    },
+    {
+      "epoch": 0.8886107634543179,
+      "grad_norm": 9.599210222057417,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 3263
+    },
+    {
+      "epoch": 0.8911138923654568,
+      "grad_norm": 10.428043989862328,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 3264
+    },
+    {
+      "epoch": 0.8936170212765957,
+      "grad_norm": 9.33863282340912,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 3265
+    },
+    {
+      "epoch": 0.8961201501877347,
+      "grad_norm": 10.107454145651822,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 3266
+    },
+    {
+      "epoch": 0.8986232790988736,
+      "grad_norm": 8.816700732340118,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 3267
+    },
+    {
+      "epoch": 0.9011264080100125,
+      "grad_norm": 9.764059484705095,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 3268
+    },
+    {
+      "epoch": 0.9036295369211514,
+      "grad_norm": 11.948193544944333,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 3269
+    },
+    {
+      "epoch": 0.9061326658322904,
+      "grad_norm": 8.626163864145518,
+      "learning_rate": 1e-06,
+      "loss": 0.1729,
+      "step": 3270
+    },
+    {
+      "epoch": 0.9086357947434293,
+      "grad_norm": 8.318870302415679,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 3271
+    },
+    {
+      "epoch": 0.9111389236545682,
+      "grad_norm": 9.272632383594075,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 3272
+    },
+    {
+      "epoch": 0.9136420525657072,
+      "grad_norm": 10.286220937875571,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 3273
+    },
+    {
+      "epoch": 0.9161451814768461,
+      "grad_norm": 8.059560183416222,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 3274
+    },
+    {
+      "epoch": 0.918648310387985,
+      "grad_norm": 8.914215659531678,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 3275
+    },
+    {
+      "epoch": 0.9211514392991239,
+      "grad_norm": 9.826665759008007,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 3276
+    },
+    {
+      "epoch": 0.9236545682102628,
+      "grad_norm": 8.816299929963337,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 3277
+    },
+    {
+      "epoch": 0.9261576971214017,
+      "grad_norm": 9.76216963458088,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 3278
+    },
+    {
+      "epoch": 0.9286608260325406,
+      "grad_norm": 8.972220620026798,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 3279
+    },
+    {
+      "epoch": 0.9311639549436797,
+      "grad_norm": 10.719493049179562,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 3280
+    },
+    {
+      "epoch": 0.9336670838548186,
+      "grad_norm": 9.728808032771934,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 3281
+    },
+    {
+      "epoch": 0.9361702127659575,
+      "grad_norm": 10.030865262976222,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3282
+    },
+    {
+      "epoch": 0.9386733416770964,
+      "grad_norm": 10.656218562141401,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 3283
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 8.174370741721981,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 3284
+    },
+    {
+      "epoch": 0.9436795994993742,
+      "grad_norm": 9.534037434456318,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 3285
+    },
+    {
+      "epoch": 0.9461827284105131,
+      "grad_norm": 8.31395249307873,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 3286
+    },
+    {
+      "epoch": 0.9486858573216521,
+      "grad_norm": 10.639233612794555,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 3287
+    },
+    {
+      "epoch": 0.951188986232791,
+      "grad_norm": 11.31330635831206,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 3288
+    },
+    {
+      "epoch": 0.9536921151439299,
+      "grad_norm": 9.674597003315613,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 3289
+    },
+    {
+      "epoch": 0.9561952440550688,
+      "grad_norm": 10.076045403016787,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 3290
+    },
+    {
+      "epoch": 0.9586983729662077,
+      "grad_norm": 10.702690851345363,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 3291
+    },
+    {
+      "epoch": 0.9612015018773467,
+      "grad_norm": 9.591866715174426,
+      "learning_rate": 1e-06,
+      "loss": 0.1756,
+      "step": 3292
+    },
+    {
+      "epoch": 0.9637046307884856,
+      "grad_norm": 10.05619909245997,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 3293
+    },
+    {
+      "epoch": 0.9662077596996246,
+      "grad_norm": 11.376211181195549,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 3294
+    },
+    {
+      "epoch": 0.9687108886107635,
+      "grad_norm": 9.201541794308929,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 3295
+    },
+    {
+      "epoch": 0.9712140175219024,
+      "grad_norm": 11.302378629064831,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 3296
+    },
+    {
+      "epoch": 0.9737171464330413,
+      "grad_norm": 10.352317990970791,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 3297
+    },
+    {
+      "epoch": 0.9762202753441802,
+      "grad_norm": 11.236170731575308,
+      "learning_rate": 1e-06,
+      "loss": 0.1687,
+      "step": 3298
+    },
+    {
+      "epoch": 0.9787234042553191,
+      "grad_norm": 8.378142734232783,
+      "learning_rate": 1e-06,
+      "loss": 0.1885,
+      "step": 3299
+    },
+    {
+      "epoch": 0.981226533166458,
+      "grad_norm": 10.771437348140351,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 3300
+    },
+    {
+      "epoch": 0.983729662077597,
+      "grad_norm": 9.033670674383444,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 3301
+    },
+    {
+      "epoch": 0.986232790988736,
+      "grad_norm": 8.286357280903074,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 3302
+    },
+    {
+      "epoch": 0.9887359198998749,
+      "grad_norm": 10.357993413688295,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 3303
+    },
+    {
+      "epoch": 0.9912390488110138,
+      "grad_norm": 7.367052334135115,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 3304
+    },
+    {
+      "epoch": 0.9937421777221527,
+      "grad_norm": 10.647835350840335,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 3305
+    },
+    {
+      "epoch": 0.9962453066332916,
+      "grad_norm": 8.655673617599838,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 3306
+    },
+    {
+      "epoch": 0.9987484355444305,
+      "grad_norm": 8.753826659005444,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 3307
+    },
+    {
+      "epoch": 0.9987484355444305,
+      "eval_loss": 3.163262128829956,
+      "eval_runtime": 21.7691,
+      "eval_samples_per_second": 45.937,
+      "eval_steps_per_second": 5.742,
+      "step": 3307
+    },
+    {
+      "epoch": 0.003484320557491289,
+      "grad_norm": 9.718756948933635,
+      "learning_rate": 1e-06,
+      "loss": 0.2109,
+      "step": 3308
+    },
+    {
+      "epoch": 0.006968641114982578,
+      "grad_norm": 9.68438968772975,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 3309
+    },
+    {
+      "epoch": 0.010452961672473868,
+      "grad_norm": 10.19812369325723,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 3310
+    },
+    {
+      "epoch": 0.013937282229965157,
+      "grad_norm": 10.587803227088624,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 3311
+    },
+    {
+      "epoch": 0.017421602787456445,
+      "grad_norm": 10.089565843300935,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 3312
+    },
+    {
+      "epoch": 0.020905923344947737,
+      "grad_norm": 9.639178723485065,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 3313
+    },
+    {
+      "epoch": 0.024390243902439025,
+      "grad_norm": 8.482289915905152,
+      "learning_rate": 1e-06,
+      "loss": 0.2264,
+      "step": 3314
+    },
+    {
+      "epoch": 0.027874564459930314,
+      "grad_norm": 9.995524679546675,
+      "learning_rate": 1e-06,
+      "loss": 0.2223,
+      "step": 3315
+    },
+    {
+      "epoch": 0.0313588850174216,
+      "grad_norm": 8.14021876072815,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 3316
+    },
+    {
+      "epoch": 0.03484320557491289,
+      "grad_norm": 7.886822223737351,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 3317
+    },
+    {
+      "epoch": 0.03832752613240418,
+      "grad_norm": 9.814236543173019,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 3318
+    },
+    {
+      "epoch": 0.041811846689895474,
+      "grad_norm": 7.551154371022341,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 3319
+    },
+    {
+      "epoch": 0.04529616724738676,
+      "grad_norm": 6.6675886619265485,
+      "learning_rate": 1e-06,
+      "loss": 0.2155,
+      "step": 3320
+    },
+    {
+      "epoch": 0.04878048780487805,
+      "grad_norm": 10.150313624097203,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 3321
+    },
+    {
+      "epoch": 0.05226480836236934,
+      "grad_norm": 8.812684847454745,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 3322
+    },
+    {
+      "epoch": 0.05574912891986063,
+      "grad_norm": 11.266933213030718,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 3323
+    },
+    {
+      "epoch": 0.059233449477351915,
+      "grad_norm": 10.227297403668057,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 3324
+    },
+    {
+      "epoch": 0.0627177700348432,
+      "grad_norm": 7.694107390228623,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 3325
+    },
+    {
+      "epoch": 0.06620209059233449,
+      "grad_norm": 7.324070283688375,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 3326
+    },
+    {
+      "epoch": 0.06968641114982578,
+      "grad_norm": 9.38009972209204,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 3327
+    },
+    {
+      "epoch": 0.07317073170731707,
+      "grad_norm": 8.234378185567927,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 3328
+    },
+    {
+      "epoch": 0.07665505226480836,
+      "grad_norm": 9.231320083092706,
+      "learning_rate": 1e-06,
+      "loss": 0.2092,
+      "step": 3329
+    },
+    {
+      "epoch": 0.08013937282229965,
+      "grad_norm": 7.523902077587705,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 3330
+    },
+    {
+      "epoch": 0.08362369337979095,
+      "grad_norm": 8.246851086173036,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3331
+    },
+    {
+      "epoch": 0.08710801393728224,
+      "grad_norm": 7.648269410931393,
+      "learning_rate": 1e-06,
+      "loss": 0.234,
+      "step": 3332
+    },
+    {
+      "epoch": 0.09059233449477352,
+      "grad_norm": 8.384709560965538,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 3333
+    },
+    {
+      "epoch": 0.09407665505226481,
+      "grad_norm": 10.407840174726388,
+      "learning_rate": 1e-06,
+      "loss": 0.2245,
+      "step": 3334
+    },
+    {
+      "epoch": 0.0975609756097561,
+      "grad_norm": 9.294611269568012,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 3335
+    },
+    {
+      "epoch": 0.10104529616724739,
+      "grad_norm": 8.966074920541717,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 3336
+    },
+    {
+      "epoch": 0.10452961672473868,
+      "grad_norm": 6.762890590454576,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 3337
+    },
+    {
+      "epoch": 0.10801393728222997,
+      "grad_norm": 11.240224785504447,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 3338
+    },
+    {
+      "epoch": 0.11149825783972125,
+      "grad_norm": 9.012430009580688,
+      "learning_rate": 1e-06,
+      "loss": 0.2074,
+      "step": 3339
+    },
+    {
+      "epoch": 0.11498257839721254,
+      "grad_norm": 8.189171374697796,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3340
+    },
+    {
+      "epoch": 0.11846689895470383,
+      "grad_norm": 7.246088345447232,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 3341
+    },
+    {
+      "epoch": 0.12195121951219512,
+      "grad_norm": 8.919247091695647,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 3342
+    },
+    {
+      "epoch": 0.1254355400696864,
+      "grad_norm": 10.293336617401218,
+      "learning_rate": 1e-06,
+      "loss": 0.2196,
+      "step": 3343
+    },
+    {
+      "epoch": 0.1289198606271777,
+      "grad_norm": 10.732256982597923,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 3344
+    },
+    {
+      "epoch": 0.13240418118466898,
+      "grad_norm": 9.753063102420601,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3345
+    },
+    {
+      "epoch": 0.13588850174216027,
+      "grad_norm": 10.715662628715993,
+      "learning_rate": 1e-06,
+      "loss": 0.2186,
+      "step": 3346
+    },
+    {
+      "epoch": 0.13937282229965156,
+      "grad_norm": 10.93905099365442,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 3347
+    },
+    {
+      "epoch": 0.14285714285714285,
+      "grad_norm": 8.415212449013708,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 3348
+    },
+    {
+      "epoch": 0.14634146341463414,
+      "grad_norm": 12.821143420099935,
+      "learning_rate": 1e-06,
+      "loss": 0.232,
+      "step": 3349
+    },
+    {
+      "epoch": 0.14982578397212543,
+      "grad_norm": 7.561465287545369,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 3350
+    },
+    {
+      "epoch": 0.15331010452961671,
+      "grad_norm": 9.525165048874722,
+      "learning_rate": 1e-06,
+      "loss": 0.2183,
+      "step": 3351
+    },
+    {
+      "epoch": 0.156794425087108,
+      "grad_norm": 8.69408618695653,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 3352
+    },
+    {
+      "epoch": 0.1602787456445993,
+      "grad_norm": 8.058155468667898,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 3353
+    },
+    {
+      "epoch": 0.16376306620209058,
+      "grad_norm": 8.062083974894172,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 3354
+    },
+    {
+      "epoch": 0.1672473867595819,
+      "grad_norm": 10.042200894794409,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 3355
+    },
+    {
+      "epoch": 0.17073170731707318,
+      "grad_norm": 7.5437703598130135,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 3356
+    },
+    {
+      "epoch": 0.17421602787456447,
+      "grad_norm": 9.890006632386047,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 3357
+    },
+    {
+      "epoch": 0.17770034843205576,
+      "grad_norm": 9.010851670083065,
+      "learning_rate": 1e-06,
+      "loss": 0.2217,
+      "step": 3358
+    },
+    {
+      "epoch": 0.18118466898954705,
+      "grad_norm": 8.692554963942362,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 3359
+    },
+    {
+      "epoch": 0.18466898954703834,
+      "grad_norm": 10.02168883406165,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 3360
+    },
+    {
+      "epoch": 0.18815331010452963,
+      "grad_norm": 8.114876541822435,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 3361
+    },
+    {
+      "epoch": 0.1916376306620209,
+      "grad_norm": 10.258628351322635,
+      "learning_rate": 1e-06,
+      "loss": 0.2163,
+      "step": 3362
+    },
+    {
+      "epoch": 0.1951219512195122,
+      "grad_norm": 8.361371154903885,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 3363
+    },
+    {
+      "epoch": 0.1986062717770035,
+      "grad_norm": 9.07224446913754,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 3364
+    },
+    {
+      "epoch": 0.20209059233449478,
+      "grad_norm": 9.327876179537926,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 3365
+    },
+    {
+      "epoch": 0.20557491289198607,
+      "grad_norm": 7.955000217542925,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 3366
+    },
+    {
+      "epoch": 0.20905923344947736,
+      "grad_norm": 10.541345068793634,
+      "learning_rate": 1e-06,
+      "loss": 0.2249,
+      "step": 3367
+    },
+    {
+      "epoch": 0.21254355400696864,
+      "grad_norm": 9.817763394102615,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 3368
+    },
+    {
+      "epoch": 0.21602787456445993,
+      "grad_norm": 8.288320642182388,
+      "learning_rate": 1e-06,
+      "loss": 0.2067,
+      "step": 3369
+    },
+    {
+      "epoch": 0.21951219512195122,
+      "grad_norm": 8.01639991358054,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 3370
+    },
+    {
+      "epoch": 0.2229965156794425,
+      "grad_norm": 9.584130232468567,
+      "learning_rate": 1e-06,
+      "loss": 0.222,
+      "step": 3371
+    },
+    {
+      "epoch": 0.2264808362369338,
+      "grad_norm": 8.405266638585793,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3372
+    },
+    {
+      "epoch": 0.22996515679442509,
+      "grad_norm": 8.187384434309381,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 3373
+    },
+    {
+      "epoch": 0.23344947735191637,
+      "grad_norm": 8.112379451398926,
+      "learning_rate": 1e-06,
+      "loss": 0.2281,
+      "step": 3374
+    },
+    {
+      "epoch": 0.23693379790940766,
+      "grad_norm": 9.408565613000105,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 3375
+    },
+    {
+      "epoch": 0.24041811846689895,
+      "grad_norm": 9.012143538648669,
+      "learning_rate": 1e-06,
+      "loss": 0.2317,
+      "step": 3376
+    },
+    {
+      "epoch": 0.24390243902439024,
+      "grad_norm": 12.301625905166663,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 3377
+    },
+    {
+      "epoch": 0.24738675958188153,
+      "grad_norm": 8.558985503838068,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3378
+    },
+    {
+      "epoch": 0.2508710801393728,
+      "grad_norm": 8.106542811038913,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3379
+    },
+    {
+      "epoch": 0.25435540069686413,
+      "grad_norm": 7.7713610730883955,
+      "learning_rate": 1e-06,
+      "loss": 0.211,
+      "step": 3380
+    },
+    {
+      "epoch": 0.2578397212543554,
+      "grad_norm": 8.356654259545936,
+      "learning_rate": 1e-06,
+      "loss": 0.2115,
+      "step": 3381
+    },
+    {
+      "epoch": 0.2613240418118467,
+      "grad_norm": 8.096957288487742,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 3382
+    },
+    {
+      "epoch": 0.26480836236933797,
+      "grad_norm": 8.763570564531197,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 3383
+    },
+    {
+      "epoch": 0.2682926829268293,
+      "grad_norm": 10.003191833683783,
+      "learning_rate": 1e-06,
+      "loss": 0.2101,
+      "step": 3384
+    },
+    {
+      "epoch": 0.27177700348432055,
+      "grad_norm": 8.285962624184585,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 3385
+    },
+    {
+      "epoch": 0.27526132404181186,
+      "grad_norm": 9.078024708860507,
+      "learning_rate": 1e-06,
+      "loss": 0.2102,
+      "step": 3386
+    },
+    {
+      "epoch": 0.2787456445993031,
+      "grad_norm": 12.047660608334823,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 3387
+    },
+    {
+      "epoch": 0.28222996515679444,
+      "grad_norm": 7.2116111561492895,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 3388
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 7.416113582428283,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 3389
+    },
+    {
+      "epoch": 0.289198606271777,
+      "grad_norm": 9.901412239432165,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3390
+    },
+    {
+      "epoch": 0.2926829268292683,
+      "grad_norm": 9.855705642842818,
+      "learning_rate": 1e-06,
+      "loss": 0.2201,
+      "step": 3391
+    },
+    {
+      "epoch": 0.2961672473867596,
+      "grad_norm": 8.893010526927776,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 3392
+    },
+    {
+      "epoch": 0.29965156794425085,
+      "grad_norm": 10.211392159591997,
+      "learning_rate": 1e-06,
+      "loss": 0.2243,
+      "step": 3393
+    },
+    {
+      "epoch": 0.30313588850174217,
+      "grad_norm": 7.758846935250837,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 3394
+    },
+    {
+      "epoch": 0.30662020905923343,
+      "grad_norm": 7.193182142669817,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 3395
+    },
+    {
+      "epoch": 0.31010452961672474,
+      "grad_norm": 10.492282544412653,
+      "learning_rate": 1e-06,
+      "loss": 0.2108,
+      "step": 3396
+    },
+    {
+      "epoch": 0.313588850174216,
+      "grad_norm": 9.354769702161622,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3397
+    },
+    {
+      "epoch": 0.3170731707317073,
+      "grad_norm": 9.619862011421828,
+      "learning_rate": 1e-06,
+      "loss": 0.215,
+      "step": 3398
+    },
+    {
+      "epoch": 0.3205574912891986,
+      "grad_norm": 7.643298382453477,
+      "learning_rate": 1e-06,
+      "loss": 0.2109,
+      "step": 3399
+    },
+    {
+      "epoch": 0.3240418118466899,
+      "grad_norm": 10.57319156789065,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 3400
+    },
+    {
+      "epoch": 0.32752613240418116,
+      "grad_norm": 6.913890082861755,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 3401
+    },
+    {
+      "epoch": 0.3310104529616725,
+      "grad_norm": 9.352562832817068,
+      "learning_rate": 1e-06,
+      "loss": 0.2136,
+      "step": 3402
+    },
+    {
+      "epoch": 0.3344947735191638,
+      "grad_norm": 8.14679048458566,
+      "learning_rate": 1e-06,
+      "loss": 0.2293,
+      "step": 3403
+    },
+    {
+      "epoch": 0.33797909407665505,
+      "grad_norm": 11.831654830802272,
+      "learning_rate": 1e-06,
+      "loss": 0.2152,
+      "step": 3404
+    },
+    {
+      "epoch": 0.34146341463414637,
+      "grad_norm": 7.9370828921557015,
+      "learning_rate": 1e-06,
+      "loss": 0.2066,
+      "step": 3405
+    },
+    {
+      "epoch": 0.34494773519163763,
+      "grad_norm": 8.97954132389193,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 3406
+    },
+    {
+      "epoch": 0.34843205574912894,
+      "grad_norm": 9.81370242479509,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 3407
+    },
+    {
+      "epoch": 0.3519163763066202,
+      "grad_norm": 10.895271739221178,
+      "learning_rate": 1e-06,
+      "loss": 0.2238,
+      "step": 3408
+    },
+    {
+      "epoch": 0.3554006968641115,
+      "grad_norm": 8.971373019793528,
+      "learning_rate": 1e-06,
+      "loss": 0.2304,
+      "step": 3409
+    },
+    {
+      "epoch": 0.3588850174216028,
+      "grad_norm": 8.107641976974563,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 3410
+    },
+    {
+      "epoch": 0.3623693379790941,
+      "grad_norm": 8.55027850907507,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 3411
+    },
+    {
+      "epoch": 0.36585365853658536,
+      "grad_norm": 8.909203783058308,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 3412
+    },
+    {
+      "epoch": 0.3693379790940767,
+      "grad_norm": 9.940703631567956,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 3413
+    },
+    {
+      "epoch": 0.37282229965156793,
+      "grad_norm": 9.048744402163523,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 3414
+    },
+    {
+      "epoch": 0.37630662020905925,
+      "grad_norm": 9.454948335986353,
+      "learning_rate": 1e-06,
+      "loss": 0.2272,
+      "step": 3415
+    },
+    {
+      "epoch": 0.3797909407665505,
+      "grad_norm": 8.957858634938802,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 3416
+    },
+    {
+      "epoch": 0.3832752613240418,
+      "grad_norm": 6.503528252097664,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 3417
+    },
+    {
+      "epoch": 0.3867595818815331,
+      "grad_norm": 7.7613122910251136,
+      "learning_rate": 1e-06,
+      "loss": 0.2268,
+      "step": 3418
+    },
+    {
+      "epoch": 0.3902439024390244,
+      "grad_norm": 8.252591523102428,
+      "learning_rate": 1e-06,
+      "loss": 0.2241,
+      "step": 3419
+    },
+    {
+      "epoch": 0.39372822299651566,
+      "grad_norm": 7.959271979840039,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 3420
+    },
+    {
+      "epoch": 0.397212543554007,
+      "grad_norm": 8.97732401606567,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 3421
+    },
+    {
+      "epoch": 0.40069686411149824,
+      "grad_norm": 9.675509925973124,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3422
+    },
+    {
+      "epoch": 0.40418118466898956,
+      "grad_norm": 8.629620718121837,
+      "learning_rate": 1e-06,
+      "loss": 0.2092,
+      "step": 3423
+    },
+    {
+      "epoch": 0.4076655052264808,
+      "grad_norm": 10.796413114827818,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 3424
+    },
+    {
+      "epoch": 0.41114982578397213,
+      "grad_norm": 9.683395602501717,
+      "learning_rate": 1e-06,
+      "loss": 0.215,
+      "step": 3425
+    },
+    {
+      "epoch": 0.4146341463414634,
+      "grad_norm": 9.392559755683251,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 3426
+    },
+    {
+      "epoch": 0.4181184668989547,
+      "grad_norm": 9.552460482015256,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 3427
+    },
+    {
+      "epoch": 0.42160278745644597,
+      "grad_norm": 10.249591758567796,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 3428
+    },
+    {
+      "epoch": 0.4250871080139373,
+      "grad_norm": 8.371248288592762,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 3429
+    },
+    {
+      "epoch": 0.42857142857142855,
+      "grad_norm": 6.751771325213432,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 3430
+    },
+    {
+      "epoch": 0.43205574912891986,
+      "grad_norm": 7.851711673647982,
+      "learning_rate": 1e-06,
+      "loss": 0.2294,
+      "step": 3431
+    },
+    {
+      "epoch": 0.4355400696864111,
+      "grad_norm": 8.948029524056771,
+      "learning_rate": 1e-06,
+      "loss": 0.2319,
+      "step": 3432
+    },
+    {
+      "epoch": 0.43902439024390244,
+      "grad_norm": 9.43344024467658,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 3433
+    },
+    {
+      "epoch": 0.4425087108013937,
+      "grad_norm": 11.736923465850682,
+      "learning_rate": 1e-06,
+      "loss": 0.2163,
+      "step": 3434
+    },
+    {
+      "epoch": 0.445993031358885,
+      "grad_norm": 7.572912692073809,
+      "learning_rate": 1e-06,
+      "loss": 0.2128,
+      "step": 3435
+    },
+    {
+      "epoch": 0.44947735191637633,
+      "grad_norm": 7.764847529948859,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 3436
+    },
+    {
+      "epoch": 0.4529616724738676,
+      "grad_norm": 8.089955598630654,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 3437
+    },
+    {
+      "epoch": 0.4564459930313589,
+      "grad_norm": 9.448748722064206,
+      "learning_rate": 1e-06,
+      "loss": 0.2148,
+      "step": 3438
+    },
+    {
+      "epoch": 0.45993031358885017,
+      "grad_norm": 8.329421593299136,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 3439
+    },
+    {
+      "epoch": 0.4634146341463415,
+      "grad_norm": 9.226362517587512,
+      "learning_rate": 1e-06,
+      "loss": 0.2272,
+      "step": 3440
+    },
+    {
+      "epoch": 0.46689895470383275,
+      "grad_norm": 7.100074163758349,
+      "learning_rate": 1e-06,
+      "loss": 0.2203,
+      "step": 3441
+    },
+    {
+      "epoch": 0.47038327526132406,
+      "grad_norm": 9.667678165885375,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3442
+    },
+    {
+      "epoch": 0.4738675958188153,
+      "grad_norm": 7.160774426008832,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 3443
+    },
+    {
+      "epoch": 0.47735191637630664,
+      "grad_norm": 9.724206036166285,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 3444
+    },
+    {
+      "epoch": 0.4808362369337979,
+      "grad_norm": 9.089268352501943,
+      "learning_rate": 1e-06,
+      "loss": 0.2217,
+      "step": 3445
+    },
+    {
+      "epoch": 0.4843205574912892,
+      "grad_norm": 7.6857628631794155,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 3446
+    },
+    {
+      "epoch": 0.4878048780487805,
+      "grad_norm": 8.432916765325391,
+      "learning_rate": 1e-06,
+      "loss": 0.2284,
+      "step": 3447
+    },
+    {
+      "epoch": 0.4912891986062718,
+      "grad_norm": 9.837969814226632,
+      "learning_rate": 1e-06,
+      "loss": 0.2082,
+      "step": 3448
+    },
+    {
+      "epoch": 0.49477351916376305,
+      "grad_norm": 9.801519478794221,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 3449
+    },
+    {
+      "epoch": 0.49825783972125437,
+      "grad_norm": 8.553329627111877,
+      "learning_rate": 1e-06,
+      "loss": 0.2045,
+      "step": 3450
+    },
+    {
+      "epoch": 0.5017421602787456,
+      "grad_norm": 9.57763704922987,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 3451
+    },
+    {
+      "epoch": 0.5052264808362369,
+      "grad_norm": 8.538182000190076,
+      "learning_rate": 1e-06,
+      "loss": 0.222,
+      "step": 3452
+    },
+    {
+      "epoch": 0.5087108013937283,
+      "grad_norm": 7.155044278042497,
+      "learning_rate": 1e-06,
+      "loss": 0.2188,
+      "step": 3453
+    },
+    {
+      "epoch": 0.5121951219512195,
+      "grad_norm": 10.209130720538651,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 3454
+    },
+    {
+      "epoch": 0.5156794425087108,
+      "grad_norm": 11.617861151504018,
+      "learning_rate": 1e-06,
+      "loss": 0.2303,
+      "step": 3455
+    },
+    {
+      "epoch": 0.519163763066202,
+      "grad_norm": 11.738923240390495,
+      "learning_rate": 1e-06,
+      "loss": 0.2305,
+      "step": 3456
+    },
+    {
+      "epoch": 0.5226480836236934,
+      "grad_norm": 7.445711148114206,
+      "learning_rate": 1e-06,
+      "loss": 0.2351,
+      "step": 3457
+    },
+    {
+      "epoch": 0.5261324041811847,
+      "grad_norm": 7.402664089348543,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 3458
+    },
+    {
+      "epoch": 0.5296167247386759,
+      "grad_norm": 7.087541548149982,
+      "learning_rate": 1e-06,
+      "loss": 0.2085,
+      "step": 3459
+    },
+    {
+      "epoch": 0.5331010452961672,
+      "grad_norm": 8.98274517701605,
+      "learning_rate": 1e-06,
+      "loss": 0.2266,
+      "step": 3460
+    },
+    {
+      "epoch": 0.5365853658536586,
+      "grad_norm": 6.901048699132092,
+      "learning_rate": 1e-06,
+      "loss": 0.2166,
+      "step": 3461
+    },
+    {
+      "epoch": 0.5400696864111498,
+      "grad_norm": 9.915910205013953,
+      "learning_rate": 1e-06,
+      "loss": 0.2134,
+      "step": 3462
+    },
+    {
+      "epoch": 0.5435540069686411,
+      "grad_norm": 8.648202019841586,
+      "learning_rate": 1e-06,
+      "loss": 0.2117,
+      "step": 3463
+    },
+    {
+      "epoch": 0.5470383275261324,
+      "grad_norm": 9.416782351989124,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 3464
+    },
+    {
+      "epoch": 0.5505226480836237,
+      "grad_norm": 9.734148251746946,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 3465
+    },
+    {
+      "epoch": 0.554006968641115,
+      "grad_norm": 10.497105571507682,
+      "learning_rate": 1e-06,
+      "loss": 0.2227,
+      "step": 3466
+    },
+    {
+      "epoch": 0.5574912891986062,
+      "grad_norm": 8.041090044666808,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 3467
+    },
+    {
+      "epoch": 0.5609756097560976,
+      "grad_norm": 10.070460944239041,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 3468
+    },
+    {
+      "epoch": 0.5644599303135889,
+      "grad_norm": 9.108317440032122,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 3469
+    },
+    {
+      "epoch": 0.5679442508710801,
+      "grad_norm": 10.576644840566638,
+      "learning_rate": 1e-06,
+      "loss": 0.2117,
+      "step": 3470
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 8.49944319004083,
+      "learning_rate": 1e-06,
+      "loss": 0.2131,
+      "step": 3471
+    },
+    {
+      "epoch": 0.5749128919860628,
+      "grad_norm": 9.54008429524662,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 3472
+    },
+    {
+      "epoch": 0.578397212543554,
+      "grad_norm": 10.298554732589954,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 3473
+    },
+    {
+      "epoch": 0.5818815331010453,
+      "grad_norm": 8.385715222612436,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 3474
+    },
+    {
+      "epoch": 0.5853658536585366,
+      "grad_norm": 9.058264828377649,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 3475
+    },
+    {
+      "epoch": 0.5888501742160279,
+      "grad_norm": 7.178375649821451,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 3476
+    },
+    {
+      "epoch": 0.5923344947735192,
+      "grad_norm": 9.157040443487547,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 3477
+    },
+    {
+      "epoch": 0.5958188153310104,
+      "grad_norm": 7.955776190278973,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 3478
+    },
+    {
+      "epoch": 0.5993031358885017,
+      "grad_norm": 5.551611289873011,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 3479
+    },
+    {
+      "epoch": 0.6027874564459931,
+      "grad_norm": 9.05760820092735,
+      "learning_rate": 1e-06,
+      "loss": 0.2201,
+      "step": 3480
+    },
+    {
+      "epoch": 0.6062717770034843,
+      "grad_norm": 8.711385236620316,
+      "learning_rate": 1e-06,
+      "loss": 0.2163,
+      "step": 3481
+    },
+    {
+      "epoch": 0.6097560975609756,
+      "grad_norm": 7.899356807672673,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 3482
+    },
+    {
+      "epoch": 0.6132404181184669,
+      "grad_norm": 9.421515695405004,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 3483
+    },
+    {
+      "epoch": 0.6167247386759582,
+      "grad_norm": 10.932213443236884,
+      "learning_rate": 1e-06,
+      "loss": 0.2404,
+      "step": 3484
+    },
+    {
+      "epoch": 0.6202090592334495,
+      "grad_norm": 10.122676823672515,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 3485
+    },
+    {
+      "epoch": 0.6236933797909407,
+      "grad_norm": 7.231865493575346,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 3486
+    },
+    {
+      "epoch": 0.627177700348432,
+      "grad_norm": 7.634569914414338,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 3487
+    },
+    {
+      "epoch": 0.6306620209059234,
+      "grad_norm": 8.439313510813914,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 3488
+    },
+    {
+      "epoch": 0.6341463414634146,
+      "grad_norm": 8.76670587620923,
+      "learning_rate": 1e-06,
+      "loss": 0.2157,
+      "step": 3489
+    },
+    {
+      "epoch": 0.6376306620209059,
+      "grad_norm": 6.5594153946405225,
+      "learning_rate": 1e-06,
+      "loss": 0.2221,
+      "step": 3490
+    },
+    {
+      "epoch": 0.6411149825783972,
+      "grad_norm": 8.522073381842084,
+      "learning_rate": 1e-06,
+      "loss": 0.2215,
+      "step": 3491
+    },
+    {
+      "epoch": 0.6445993031358885,
+      "grad_norm": 7.355210065534485,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3492
+    },
+    {
+      "epoch": 0.6480836236933798,
+      "grad_norm": 8.158418420072314,
+      "learning_rate": 1e-06,
+      "loss": 0.2193,
+      "step": 3493
+    },
+    {
+      "epoch": 0.6515679442508711,
+      "grad_norm": 7.796809031862253,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 3494
+    },
+    {
+      "epoch": 0.6550522648083623,
+      "grad_norm": 7.560459951434701,
+      "learning_rate": 1e-06,
+      "loss": 0.2134,
+      "step": 3495
+    },
+    {
+      "epoch": 0.6585365853658537,
+      "grad_norm": 8.5946539154275,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 3496
+    },
+    {
+      "epoch": 0.662020905923345,
+      "grad_norm": 6.503897544923747,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 3497
+    },
+    {
+      "epoch": 0.6655052264808362,
+      "grad_norm": 9.314073070407295,
+      "learning_rate": 1e-06,
+      "loss": 0.213,
+      "step": 3498
+    },
+    {
+      "epoch": 0.6689895470383276,
+      "grad_norm": 10.974751328520904,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 3499
+    },
+    {
+      "epoch": 0.6724738675958188,
+      "grad_norm": 10.454334706299582,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3500
+    },
+    {
+      "epoch": 0.6759581881533101,
+      "grad_norm": 9.931385743166476,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 3501
+    },
+    {
+      "epoch": 0.6794425087108014,
+      "grad_norm": 9.089542267031584,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 3502
+    },
+    {
+      "epoch": 0.6829268292682927,
+      "grad_norm": 10.14384262567664,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 3503
+    },
+    {
+      "epoch": 0.686411149825784,
+      "grad_norm": 11.097175613395148,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 3504
+    },
+    {
+      "epoch": 0.6898954703832753,
+      "grad_norm": 11.188061816517388,
+      "learning_rate": 1e-06,
+      "loss": 0.2125,
+      "step": 3505
+    },
+    {
+      "epoch": 0.6933797909407665,
+      "grad_norm": 9.540600032977064,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 3506
+    },
+    {
+      "epoch": 0.6968641114982579,
+      "grad_norm": 8.9986702370786,
+      "learning_rate": 1e-06,
+      "loss": 0.2345,
+      "step": 3507
+    },
+    {
+      "epoch": 0.7003484320557491,
+      "grad_norm": 9.790140844800595,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 3508
+    },
+    {
+      "epoch": 0.7038327526132404,
+      "grad_norm": 8.86441465041589,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 3509
+    },
+    {
+      "epoch": 0.7073170731707317,
+      "grad_norm": 6.886906429926021,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 3510
+    },
+    {
+      "epoch": 0.710801393728223,
+      "grad_norm": 8.751259033139403,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3511
+    },
+    {
+      "epoch": 0.7142857142857143,
+      "grad_norm": 9.247977299186836,
+      "learning_rate": 1e-06,
+      "loss": 0.2225,
+      "step": 3512
+    },
+    {
+      "epoch": 0.7177700348432056,
+      "grad_norm": 6.411944990980544,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 3513
+    },
+    {
+      "epoch": 0.7212543554006968,
+      "grad_norm": 8.774089133567555,
+      "learning_rate": 1e-06,
+      "loss": 0.2209,
+      "step": 3514
+    },
+    {
+      "epoch": 0.7247386759581882,
+      "grad_norm": 11.787864170523717,
+      "learning_rate": 1e-06,
+      "loss": 0.2301,
+      "step": 3515
+    },
+    {
+      "epoch": 0.7282229965156795,
+      "grad_norm": 9.666886618866581,
+      "learning_rate": 1e-06,
+      "loss": 0.2108,
+      "step": 3516
+    },
+    {
+      "epoch": 0.7317073170731707,
+      "grad_norm": 8.322128482437053,
+      "learning_rate": 1e-06,
+      "loss": 0.2216,
+      "step": 3517
+    },
+    {
+      "epoch": 0.735191637630662,
+      "grad_norm": 8.21040927003539,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 3518
+    },
+    {
+      "epoch": 0.7386759581881533,
+      "grad_norm": 7.764918653899849,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 3519
+    },
+    {
+      "epoch": 0.7421602787456446,
+      "grad_norm": 7.322742149551277,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 3520
+    },
+    {
+      "epoch": 0.7456445993031359,
+      "grad_norm": 9.767173320444718,
+      "learning_rate": 1e-06,
+      "loss": 0.227,
+      "step": 3521
+    },
+    {
+      "epoch": 0.7491289198606271,
+      "grad_norm": 7.482781506579077,
+      "learning_rate": 1e-06,
+      "loss": 0.2277,
+      "step": 3522
+    },
+    {
+      "epoch": 0.7526132404181185,
+      "grad_norm": 9.969651830553847,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 3523
+    },
+    {
+      "epoch": 0.7560975609756098,
+      "grad_norm": 11.245492222739847,
+      "learning_rate": 1e-06,
+      "loss": 0.2289,
+      "step": 3524
+    },
+    {
+      "epoch": 0.759581881533101,
+      "grad_norm": 7.273941676695488,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 3525
+    },
+    {
+      "epoch": 0.7630662020905923,
+      "grad_norm": 9.780467852658912,
+      "learning_rate": 1e-06,
+      "loss": 0.2189,
+      "step": 3526
+    },
+    {
+      "epoch": 0.7665505226480837,
+      "grad_norm": 7.251454286010235,
+      "learning_rate": 1e-06,
+      "loss": 0.2201,
+      "step": 3527
+    },
+    {
+      "epoch": 0.7700348432055749,
+      "grad_norm": 7.017755441228058,
+      "learning_rate": 1e-06,
+      "loss": 0.2102,
+      "step": 3528
+    },
+    {
+      "epoch": 0.7735191637630662,
+      "grad_norm": 8.287998636230247,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 3529
+    },
+    {
+      "epoch": 0.7770034843205574,
+      "grad_norm": 11.38101758598068,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 3530
+    },
+    {
+      "epoch": 0.7804878048780488,
+      "grad_norm": 9.80315729875092,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 3531
+    },
+    {
+      "epoch": 0.7839721254355401,
+      "grad_norm": 7.22072262697821,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 3532
+    },
+    {
+      "epoch": 0.7874564459930313,
+      "grad_norm": 9.622580042953693,
+      "learning_rate": 1e-06,
+      "loss": 0.2346,
+      "step": 3533
+    },
+    {
+      "epoch": 0.7909407665505227,
+      "grad_norm": 10.575266307717909,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 3534
+    },
+    {
+      "epoch": 0.794425087108014,
+      "grad_norm": 10.147914982688269,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 3535
+    },
+    {
+      "epoch": 0.7979094076655052,
+      "grad_norm": 9.77144567743874,
+      "learning_rate": 1e-06,
+      "loss": 0.2243,
+      "step": 3536
+    },
+    {
+      "epoch": 0.8013937282229965,
+      "grad_norm": 9.761158150584917,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 3537
+    },
+    {
+      "epoch": 0.8048780487804879,
+      "grad_norm": 7.7993769082561295,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 3538
+    },
+    {
+      "epoch": 0.8083623693379791,
+      "grad_norm": 6.741603513085161,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 3539
+    },
+    {
+      "epoch": 0.8118466898954704,
+      "grad_norm": 8.29768300442933,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 3540
+    },
+    {
+      "epoch": 0.8153310104529616,
+      "grad_norm": 8.265525422798957,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 3541
+    },
+    {
+      "epoch": 0.818815331010453,
+      "grad_norm": 8.192202040666379,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 3542
+    },
+    {
+      "epoch": 0.8222996515679443,
+      "grad_norm": 8.46734346305818,
+      "learning_rate": 1e-06,
+      "loss": 0.2103,
+      "step": 3543
+    },
+    {
+      "epoch": 0.8257839721254355,
+      "grad_norm": 7.696630638475367,
+      "learning_rate": 1e-06,
+      "loss": 0.2298,
+      "step": 3544
+    },
+    {
+      "epoch": 0.8292682926829268,
+      "grad_norm": 8.689115367677807,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 3545
+    },
+    {
+      "epoch": 0.8327526132404182,
+      "grad_norm": 6.783527646833595,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 3546
+    },
+    {
+      "epoch": 0.8362369337979094,
+      "grad_norm": 12.196074324689718,
+      "learning_rate": 1e-06,
+      "loss": 0.2067,
+      "step": 3547
+    },
+    {
+      "epoch": 0.8397212543554007,
+      "grad_norm": 9.389762387772013,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 3548
+    },
+    {
+      "epoch": 0.8432055749128919,
+      "grad_norm": 7.7230260224694165,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 3549
+    },
+    {
+      "epoch": 0.8466898954703833,
+      "grad_norm": 7.290261874830889,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 3550
+    },
+    {
+      "epoch": 0.8501742160278746,
+      "grad_norm": 7.87858767465581,
+      "learning_rate": 1e-06,
+      "loss": 0.213,
+      "step": 3551
+    },
+    {
+      "epoch": 0.8536585365853658,
+      "grad_norm": 5.866181248973454,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 3552
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 7.694212250959694,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 3553
+    },
+    {
+      "epoch": 0.8606271777003485,
+      "grad_norm": 8.12217215457778,
+      "learning_rate": 1e-06,
+      "loss": 0.2261,
+      "step": 3554
+    },
+    {
+      "epoch": 0.8641114982578397,
+      "grad_norm": 12.169386036827426,
+      "learning_rate": 1e-06,
+      "loss": 0.2125,
+      "step": 3555
+    },
+    {
+      "epoch": 0.867595818815331,
+      "grad_norm": 7.223204732633192,
+      "learning_rate": 1e-06,
+      "loss": 0.2157,
+      "step": 3556
+    },
+    {
+      "epoch": 0.8710801393728222,
+      "grad_norm": 9.208299991135352,
+      "learning_rate": 1e-06,
+      "loss": 0.2304,
+      "step": 3557
+    },
+    {
+      "epoch": 0.8745644599303136,
+      "grad_norm": 7.471573113354612,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 3558
+    },
+    {
+      "epoch": 0.8780487804878049,
+      "grad_norm": 8.211060636629911,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 3559
+    },
+    {
+      "epoch": 0.8815331010452961,
+      "grad_norm": 9.892342126490574,
+      "learning_rate": 1e-06,
+      "loss": 0.2246,
+      "step": 3560
+    },
+    {
+      "epoch": 0.8850174216027874,
+      "grad_norm": 7.316883526295152,
+      "learning_rate": 1e-06,
+      "loss": 0.2283,
+      "step": 3561
+    },
+    {
+      "epoch": 0.8885017421602788,
+      "grad_norm": 8.800797141528417,
+      "learning_rate": 1e-06,
+      "loss": 0.2036,
+      "step": 3562
+    },
+    {
+      "epoch": 0.89198606271777,
+      "grad_norm": 7.053189106029893,
+      "learning_rate": 1e-06,
+      "loss": 0.2194,
+      "step": 3563
+    },
+    {
+      "epoch": 0.8954703832752613,
+      "grad_norm": 7.082221310404558,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 3564
+    },
+    {
+      "epoch": 0.8989547038327527,
+      "grad_norm": 10.34909796129509,
+      "learning_rate": 1e-06,
+      "loss": 0.2254,
+      "step": 3565
+    },
+    {
+      "epoch": 0.9024390243902439,
+      "grad_norm": 6.2254963511491255,
+      "learning_rate": 1e-06,
+      "loss": 0.2108,
+      "step": 3566
+    },
+    {
+      "epoch": 0.9059233449477352,
+      "grad_norm": 7.544301006363569,
+      "learning_rate": 1e-06,
+      "loss": 0.2152,
+      "step": 3567
+    },
+    {
+      "epoch": 0.9094076655052264,
+      "grad_norm": 9.135179269702801,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 3568
+    },
+    {
+      "epoch": 0.9128919860627178,
+      "grad_norm": 9.467530540249419,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 3569
+    },
+    {
+      "epoch": 0.9163763066202091,
+      "grad_norm": 8.15828066236896,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 3570
+    },
+    {
+      "epoch": 0.9198606271777003,
+      "grad_norm": 10.964112460816258,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 3571
+    },
+    {
+      "epoch": 0.9233449477351916,
+      "grad_norm": 9.364321069280477,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 3572
+    },
+    {
+      "epoch": 0.926829268292683,
+      "grad_norm": 10.080890425472555,
+      "learning_rate": 1e-06,
+      "loss": 0.2326,
+      "step": 3573
+    },
+    {
+      "epoch": 0.9303135888501742,
+      "grad_norm": 10.252686950896056,
+      "learning_rate": 1e-06,
+      "loss": 0.2066,
+      "step": 3574
+    },
+    {
+      "epoch": 0.9337979094076655,
+      "grad_norm": 9.018155923752591,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 3575
+    },
+    {
+      "epoch": 0.9372822299651568,
+      "grad_norm": 7.709630548254519,
+      "learning_rate": 1e-06,
+      "loss": 0.2125,
+      "step": 3576
+    },
+    {
+      "epoch": 0.9407665505226481,
+      "grad_norm": 6.244293949606123,
+      "learning_rate": 1e-06,
+      "loss": 0.2068,
+      "step": 3577
+    },
+    {
+      "epoch": 0.9442508710801394,
+      "grad_norm": 8.842771191862633,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 3578
+    },
+    {
+      "epoch": 0.9477351916376306,
+      "grad_norm": 8.08383333739928,
+      "learning_rate": 1e-06,
+      "loss": 0.228,
+      "step": 3579
+    },
+    {
+      "epoch": 0.9512195121951219,
+      "grad_norm": 7.713611918443035,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 3580
+    },
+    {
+      "epoch": 0.9547038327526133,
+      "grad_norm": 8.613141617432472,
+      "learning_rate": 1e-06,
+      "loss": 0.2045,
+      "step": 3581
+    },
+    {
+      "epoch": 0.9581881533101045,
+      "grad_norm": 9.0981259998838,
+      "learning_rate": 1e-06,
+      "loss": 0.2229,
+      "step": 3582
+    },
+    {
+      "epoch": 0.9616724738675958,
+      "grad_norm": 10.518771598812107,
+      "learning_rate": 1e-06,
+      "loss": 0.2175,
+      "step": 3583
+    },
+    {
+      "epoch": 0.9651567944250871,
+      "grad_norm": 9.418589912040607,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 3584
+    },
+    {
+      "epoch": 0.9686411149825784,
+      "grad_norm": 7.939529347171159,
+      "learning_rate": 1e-06,
+      "loss": 0.2216,
+      "step": 3585
+    },
+    {
+      "epoch": 0.9721254355400697,
+      "grad_norm": 11.200740784467227,
+      "learning_rate": 1e-06,
+      "loss": 0.2125,
+      "step": 3586
+    },
+    {
+      "epoch": 0.975609756097561,
+      "grad_norm": 6.827800702692437,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 3587
+    },
+    {
+      "epoch": 0.9790940766550522,
+      "grad_norm": 8.23245638856237,
+      "learning_rate": 1e-06,
+      "loss": 0.2131,
+      "step": 3588
+    },
+    {
+      "epoch": 0.9825783972125436,
+      "grad_norm": 10.74661175721978,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 3589
+    },
+    {
+      "epoch": 0.9860627177700348,
+      "grad_norm": 8.649264564992903,
+      "learning_rate": 1e-06,
+      "loss": 0.2198,
+      "step": 3590
+    },
+    {
+      "epoch": 0.9895470383275261,
+      "grad_norm": 7.928845880430522,
+      "learning_rate": 1e-06,
+      "loss": 0.234,
+      "step": 3591
+    },
+    {
+      "epoch": 0.9930313588850174,
+      "grad_norm": 8.366681913807941,
+      "learning_rate": 1e-06,
+      "loss": 0.2123,
+      "step": 3592
+    },
+    {
+      "epoch": 0.9965156794425087,
+      "grad_norm": 8.728308455281768,
+      "learning_rate": 1e-06,
+      "loss": 0.2306,
+      "step": 3593
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 10.173974389272825,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 3594
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.9281303882598877,
+      "eval_runtime": 21.7371,
+      "eval_samples_per_second": 46.004,
+      "eval_steps_per_second": 5.751,
+      "step": 3594
+    },
+    {
+      "epoch": 0.0023894862604540022,
+      "grad_norm": 6.312387775111674,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3595
+    },
+    {
+      "epoch": 0.0047789725209080045,
+      "grad_norm": 6.074521747369287,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 3596
+    },
+    {
+      "epoch": 0.007168458781362007,
+      "grad_norm": 6.856707862053721,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 3597
+    },
+    {
+      "epoch": 0.009557945041816009,
+      "grad_norm": 7.995262727766026,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 3598
+    },
+    {
+      "epoch": 0.011947431302270013,
+      "grad_norm": 6.976429518801098,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 3599
+    },
+    {
+      "epoch": 0.014336917562724014,
+      "grad_norm": 7.849062100095326,
+      "learning_rate": 1e-06,
+      "loss": 0.1897,
+      "step": 3600
+    },
+    {
+      "epoch": 0.016726403823178016,
+      "grad_norm": 7.853087454235319,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3601
+    },
+    {
+      "epoch": 0.019115890083632018,
+      "grad_norm": 10.537496749547163,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 3602
+    },
+    {
+      "epoch": 0.021505376344086023,
+      "grad_norm": 8.098271118630347,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 3603
+    },
+    {
+      "epoch": 0.023894862604540025,
+      "grad_norm": 8.168125678428986,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 3604
+    },
+    {
+      "epoch": 0.026284348864994027,
+      "grad_norm": 7.83201504408231,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 3605
+    },
+    {
+      "epoch": 0.02867383512544803,
+      "grad_norm": 9.17642845548922,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 3606
+    },
+    {
+      "epoch": 0.03106332138590203,
+      "grad_norm": 8.770738233875438,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 3607
+    },
+    {
+      "epoch": 0.03345280764635603,
+      "grad_norm": 7.710824609244599,
+      "learning_rate": 1e-06,
+      "loss": 0.2018,
+      "step": 3608
+    },
+    {
+      "epoch": 0.035842293906810034,
+      "grad_norm": 8.082251890016101,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 3609
+    },
+    {
+      "epoch": 0.038231780167264036,
+      "grad_norm": 8.173965498551494,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 3610
+    },
+    {
+      "epoch": 0.04062126642771804,
+      "grad_norm": 7.358225397968691,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 3611
+    },
+    {
+      "epoch": 0.043010752688172046,
+      "grad_norm": 7.960880377458544,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 3612
+    },
+    {
+      "epoch": 0.04540023894862605,
+      "grad_norm": 8.815788405318589,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 3613
+    },
+    {
+      "epoch": 0.04778972520908005,
+      "grad_norm": 9.882116275013951,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 3614
+    },
+    {
+      "epoch": 0.05017921146953405,
+      "grad_norm": 10.788613406880827,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 3615
+    },
+    {
+      "epoch": 0.052568697729988054,
+      "grad_norm": 7.596319565246984,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 3616
+    },
+    {
+      "epoch": 0.054958183990442055,
+      "grad_norm": 8.46430817489901,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 3617
+    },
+    {
+      "epoch": 0.05734767025089606,
+      "grad_norm": 9.564429653113656,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3618
+    },
+    {
+      "epoch": 0.05973715651135006,
+      "grad_norm": 9.519338400043342,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 3619
+    },
+    {
+      "epoch": 0.06212664277180406,
+      "grad_norm": 10.64012729622826,
+      "learning_rate": 1e-06,
+      "loss": 0.1974,
+      "step": 3620
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 10.86686647616832,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3621
+    },
+    {
+      "epoch": 0.06690561529271206,
+      "grad_norm": 6.11020044453817,
+      "learning_rate": 1e-06,
+      "loss": 0.2042,
+      "step": 3622
+    },
+    {
+      "epoch": 0.06929510155316607,
+      "grad_norm": 7.182750768865081,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 3623
+    },
+    {
+      "epoch": 0.07168458781362007,
+      "grad_norm": 6.66745564926861,
+      "learning_rate": 1e-06,
+      "loss": 0.1903,
+      "step": 3624
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 7.7528103098518,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 3625
+    },
+    {
+      "epoch": 0.07646356033452807,
+      "grad_norm": 7.35048642233049,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 3626
+    },
+    {
+      "epoch": 0.07885304659498207,
+      "grad_norm": 8.116260687367259,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 3627
+    },
+    {
+      "epoch": 0.08124253285543608,
+      "grad_norm": 7.027485598710405,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3628
+    },
+    {
+      "epoch": 0.08363201911589008,
+      "grad_norm": 8.890665473949566,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 3629
+    },
+    {
+      "epoch": 0.08602150537634409,
+      "grad_norm": 8.37133203684156,
+      "learning_rate": 1e-06,
+      "loss": 0.1865,
+      "step": 3630
+    },
+    {
+      "epoch": 0.0884109916367981,
+      "grad_norm": 9.381217592609273,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 3631
+    },
+    {
+      "epoch": 0.0908004778972521,
+      "grad_norm": 7.629883456913686,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 3632
+    },
+    {
+      "epoch": 0.0931899641577061,
+      "grad_norm": 10.081079225137069,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 3633
+    },
+    {
+      "epoch": 0.0955794504181601,
+      "grad_norm": 7.739136100151934,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 3634
+    },
+    {
+      "epoch": 0.0979689366786141,
+      "grad_norm": 10.461619870000089,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 3635
+    },
+    {
+      "epoch": 0.1003584229390681,
+      "grad_norm": 10.025221175337146,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 3636
+    },
+    {
+      "epoch": 0.1027479091995221,
+      "grad_norm": 9.922044467347016,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 3637
+    },
+    {
+      "epoch": 0.10513739545997611,
+      "grad_norm": 9.330566726753238,
+      "learning_rate": 1e-06,
+      "loss": 0.2223,
+      "step": 3638
+    },
+    {
+      "epoch": 0.10752688172043011,
+      "grad_norm": 8.44476284152023,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3639
+    },
+    {
+      "epoch": 0.10991636798088411,
+      "grad_norm": 10.057207228624433,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 3640
+    },
+    {
+      "epoch": 0.11230585424133811,
+      "grad_norm": 10.503480737132936,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 3641
+    },
+    {
+      "epoch": 0.11469534050179211,
+      "grad_norm": 9.267722409562733,
+      "learning_rate": 1e-06,
+      "loss": 0.2068,
+      "step": 3642
+    },
+    {
+      "epoch": 0.11708482676224612,
+      "grad_norm": 7.240461652207892,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 3643
+    },
+    {
+      "epoch": 0.11947431302270012,
+      "grad_norm": 7.6051950870956215,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 3644
+    },
+    {
+      "epoch": 0.12186379928315412,
+      "grad_norm": 10.64028793627033,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 3645
+    },
+    {
+      "epoch": 0.12425328554360812,
+      "grad_norm": 6.775739454635101,
+      "learning_rate": 1e-06,
+      "loss": 0.2074,
+      "step": 3646
+    },
+    {
+      "epoch": 0.12664277180406214,
+      "grad_norm": 8.41611509863008,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 3647
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 8.402775166944954,
+      "learning_rate": 1e-06,
+      "loss": 0.1936,
+      "step": 3648
+    },
+    {
+      "epoch": 0.13142174432497014,
+      "grad_norm": 7.9425215422314075,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3649
+    },
+    {
+      "epoch": 0.13381123058542413,
+      "grad_norm": 8.479431477175263,
+      "learning_rate": 1e-06,
+      "loss": 0.21,
+      "step": 3650
+    },
+    {
+      "epoch": 0.13620071684587814,
+      "grad_norm": 9.473701060672246,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 3651
+    },
+    {
+      "epoch": 0.13859020310633213,
+      "grad_norm": 9.000982744696548,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 3652
+    },
+    {
+      "epoch": 0.14097968936678615,
+      "grad_norm": 8.752830974575941,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 3653
+    },
+    {
+      "epoch": 0.14336917562724014,
+      "grad_norm": 8.087101581000995,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 3654
+    },
+    {
+      "epoch": 0.14575866188769415,
+      "grad_norm": 7.785000509161607,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 3655
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 8.417676044534119,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 3656
+    },
+    {
+      "epoch": 0.15053763440860216,
+      "grad_norm": 7.828436643463196,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 3657
+    },
+    {
+      "epoch": 0.15292712066905614,
+      "grad_norm": 6.304845311514582,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 3658
+    },
+    {
+      "epoch": 0.15531660692951016,
+      "grad_norm": 7.741880803368095,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 3659
+    },
+    {
+      "epoch": 0.15770609318996415,
+      "grad_norm": 7.227422032265995,
+      "learning_rate": 1e-06,
+      "loss": 0.181,
+      "step": 3660
+    },
+    {
+      "epoch": 0.16009557945041816,
+      "grad_norm": 8.077377804893555,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 3661
+    },
+    {
+      "epoch": 0.16248506571087215,
+      "grad_norm": 8.833553135991721,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 3662
+    },
+    {
+      "epoch": 0.16487455197132617,
+      "grad_norm": 7.925023951871433,
+      "learning_rate": 1e-06,
+      "loss": 0.1982,
+      "step": 3663
+    },
+    {
+      "epoch": 0.16726403823178015,
+      "grad_norm": 8.136134322266324,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 3664
+    },
+    {
+      "epoch": 0.16965352449223417,
+      "grad_norm": 7.425414551164528,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 3665
+    },
+    {
+      "epoch": 0.17204301075268819,
+      "grad_norm": 6.72435446675906,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 3666
+    },
+    {
+      "epoch": 0.17443249701314217,
+      "grad_norm": 9.639743560705819,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 3667
+    },
+    {
+      "epoch": 0.1768219832735962,
+      "grad_norm": 7.8986293073634934,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 3668
+    },
+    {
+      "epoch": 0.17921146953405018,
+      "grad_norm": 8.307577806007584,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3669
+    },
+    {
+      "epoch": 0.1816009557945042,
+      "grad_norm": 8.854977527221672,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 3670
+    },
+    {
+      "epoch": 0.18399044205495818,
+      "grad_norm": 9.052027132026188,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 3671
+    },
+    {
+      "epoch": 0.1863799283154122,
+      "grad_norm": 8.852851466885685,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 3672
+    },
+    {
+      "epoch": 0.18876941457586618,
+      "grad_norm": 7.414097497736115,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 3673
+    },
+    {
+      "epoch": 0.1911589008363202,
+      "grad_norm": 9.535720584752054,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 3674
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 8.933347142972869,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 3675
+    },
+    {
+      "epoch": 0.1959378733572282,
+      "grad_norm": 9.172909822571382,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 3676
+    },
+    {
+      "epoch": 0.1983273596176822,
+      "grad_norm": 8.060114834022428,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 3677
+    },
+    {
+      "epoch": 0.2007168458781362,
+      "grad_norm": 8.871279220898572,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 3678
+    },
+    {
+      "epoch": 0.2031063321385902,
+      "grad_norm": 8.4627961628722,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 3679
+    },
+    {
+      "epoch": 0.2054958183990442,
+      "grad_norm": 10.012986459661944,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 3680
+    },
+    {
+      "epoch": 0.2078853046594982,
+      "grad_norm": 7.416252309681986,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3681
+    },
+    {
+      "epoch": 0.21027479091995221,
+      "grad_norm": 10.348114335576712,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 3682
+    },
+    {
+      "epoch": 0.2126642771804062,
+      "grad_norm": 8.388673733970826,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 3683
+    },
+    {
+      "epoch": 0.21505376344086022,
+      "grad_norm": 9.843790230025478,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 3684
+    },
+    {
+      "epoch": 0.2174432497013142,
+      "grad_norm": 11.219818183995233,
+      "learning_rate": 1e-06,
+      "loss": 0.21,
+      "step": 3685
+    },
+    {
+      "epoch": 0.21983273596176822,
+      "grad_norm": 9.219768307128652,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 3686
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 5.844695588319013,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3687
+    },
+    {
+      "epoch": 0.22461170848267623,
+      "grad_norm": 8.827325980364666,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 3688
+    },
+    {
+      "epoch": 0.2270011947431302,
+      "grad_norm": 6.598182034793477,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 3689
+    },
+    {
+      "epoch": 0.22939068100358423,
+      "grad_norm": 9.282307312765656,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3690
+    },
+    {
+      "epoch": 0.23178016726403824,
+      "grad_norm": 9.016328112927045,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 3691
+    },
+    {
+      "epoch": 0.23416965352449223,
+      "grad_norm": 7.401304147207503,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 3692
+    },
+    {
+      "epoch": 0.23655913978494625,
+      "grad_norm": 8.48351275009509,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 3693
+    },
+    {
+      "epoch": 0.23894862604540024,
+      "grad_norm": 7.620015666830012,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 3694
+    },
+    {
+      "epoch": 0.24133811230585425,
+      "grad_norm": 9.728393121187198,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 3695
+    },
+    {
+      "epoch": 0.24372759856630824,
+      "grad_norm": 11.083548877054616,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 3696
+    },
+    {
+      "epoch": 0.24611708482676226,
+      "grad_norm": 5.940196343598266,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 3697
+    },
+    {
+      "epoch": 0.24850657108721624,
+      "grad_norm": 8.765954193698118,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 3698
+    },
+    {
+      "epoch": 0.25089605734767023,
+      "grad_norm": 7.796815043066098,
+      "learning_rate": 1e-06,
+      "loss": 0.2135,
+      "step": 3699
+    },
+    {
+      "epoch": 0.2532855436081243,
+      "grad_norm": 8.487756131566309,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 3700
+    },
+    {
+      "epoch": 0.25567502986857826,
+      "grad_norm": 7.583914875588996,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 3701
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 7.549894053353748,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 3702
+    },
+    {
+      "epoch": 0.26045400238948624,
+      "grad_norm": 9.919087343465682,
+      "learning_rate": 1e-06,
+      "loss": 0.2112,
+      "step": 3703
+    },
+    {
+      "epoch": 0.2628434886499403,
+      "grad_norm": 8.395722197069185,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 3704
+    },
+    {
+      "epoch": 0.26523297491039427,
+      "grad_norm": 6.756869021020593,
+      "learning_rate": 1e-06,
+      "loss": 0.1897,
+      "step": 3705
+    },
+    {
+      "epoch": 0.26762246117084826,
+      "grad_norm": 7.678477986977018,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 3706
+    },
+    {
+      "epoch": 0.27001194743130225,
+      "grad_norm": 7.113677019084856,
+      "learning_rate": 1e-06,
+      "loss": 0.2018,
+      "step": 3707
+    },
+    {
+      "epoch": 0.2724014336917563,
+      "grad_norm": 7.117583639061138,
+      "learning_rate": 1e-06,
+      "loss": 0.1961,
+      "step": 3708
+    },
+    {
+      "epoch": 0.2747909199522103,
+      "grad_norm": 9.973260168353002,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 3709
+    },
+    {
+      "epoch": 0.27718040621266427,
+      "grad_norm": 7.301130882314901,
+      "learning_rate": 1e-06,
+      "loss": 0.1785,
+      "step": 3710
+    },
+    {
+      "epoch": 0.27956989247311825,
+      "grad_norm": 9.371454437324195,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 3711
+    },
+    {
+      "epoch": 0.2819593787335723,
+      "grad_norm": 10.216957055102627,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 3712
+    },
+    {
+      "epoch": 0.2843488649940263,
+      "grad_norm": 9.618373194775396,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3713
+    },
+    {
+      "epoch": 0.2867383512544803,
+      "grad_norm": 8.216610052183427,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 3714
+    },
+    {
+      "epoch": 0.2891278375149343,
+      "grad_norm": 7.841046523169203,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 3715
+    },
+    {
+      "epoch": 0.2915173237753883,
+      "grad_norm": 7.295561239889686,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 3716
+    },
+    {
+      "epoch": 0.2939068100358423,
+      "grad_norm": 10.503905982865234,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 3717
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 8.756531730316297,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 3718
+    },
+    {
+      "epoch": 0.2986857825567503,
+      "grad_norm": 10.185670322317495,
+      "learning_rate": 1e-06,
+      "loss": 0.221,
+      "step": 3719
+    },
+    {
+      "epoch": 0.3010752688172043,
+      "grad_norm": 9.66699195718606,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 3720
+    },
+    {
+      "epoch": 0.3034647550776583,
+      "grad_norm": 8.363048560920197,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 3721
+    },
+    {
+      "epoch": 0.3058542413381123,
+      "grad_norm": 8.074700830547986,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 3722
+    },
+    {
+      "epoch": 0.30824372759856633,
+      "grad_norm": 9.912601243814194,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3723
+    },
+    {
+      "epoch": 0.3106332138590203,
+      "grad_norm": 8.427527365236232,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 3724
+    },
+    {
+      "epoch": 0.3130227001194743,
+      "grad_norm": 6.387128714814097,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 3725
+    },
+    {
+      "epoch": 0.3154121863799283,
+      "grad_norm": 10.958472892908938,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 3726
+    },
+    {
+      "epoch": 0.31780167264038234,
+      "grad_norm": 9.447755908638086,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 3727
+    },
+    {
+      "epoch": 0.3201911589008363,
+      "grad_norm": 8.600442472952004,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 3728
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 5.815508690352668,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 3729
+    },
+    {
+      "epoch": 0.3249701314217443,
+      "grad_norm": 8.203761190128668,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 3730
+    },
+    {
+      "epoch": 0.32735961768219835,
+      "grad_norm": 6.981879265949544,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 3731
+    },
+    {
+      "epoch": 0.32974910394265233,
+      "grad_norm": 7.41031605222569,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 3732
+    },
+    {
+      "epoch": 0.3321385902031063,
+      "grad_norm": 8.463628917057907,
+      "learning_rate": 1e-06,
+      "loss": 0.2075,
+      "step": 3733
+    },
+    {
+      "epoch": 0.3345280764635603,
+      "grad_norm": 10.077583727254156,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3734
+    },
+    {
+      "epoch": 0.33691756272401435,
+      "grad_norm": 8.918214134769281,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3735
+    },
+    {
+      "epoch": 0.33930704898446834,
+      "grad_norm": 8.62233302738196,
+      "learning_rate": 1e-06,
+      "loss": 0.1885,
+      "step": 3736
+    },
+    {
+      "epoch": 0.34169653524492233,
+      "grad_norm": 8.213900644778246,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 3737
+    },
+    {
+      "epoch": 0.34408602150537637,
+      "grad_norm": 11.15295530685688,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 3738
+    },
+    {
+      "epoch": 0.34647550776583036,
+      "grad_norm": 9.143002603082897,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 3739
+    },
+    {
+      "epoch": 0.34886499402628435,
+      "grad_norm": 10.680143982428515,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3740
+    },
+    {
+      "epoch": 0.35125448028673834,
+      "grad_norm": 8.396924989798162,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 3741
+    },
+    {
+      "epoch": 0.3536439665471924,
+      "grad_norm": 7.467171749096495,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 3742
+    },
+    {
+      "epoch": 0.35603345280764637,
+      "grad_norm": 11.57925657492074,
+      "learning_rate": 1e-06,
+      "loss": 0.2278,
+      "step": 3743
+    },
+    {
+      "epoch": 0.35842293906810035,
+      "grad_norm": 9.548848274128355,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3744
+    },
+    {
+      "epoch": 0.36081242532855434,
+      "grad_norm": 8.237115291793547,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 3745
+    },
+    {
+      "epoch": 0.3632019115890084,
+      "grad_norm": 10.60055424552847,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 3746
+    },
+    {
+      "epoch": 0.3655913978494624,
+      "grad_norm": 9.108576971427343,
+      "learning_rate": 1e-06,
+      "loss": 0.2123,
+      "step": 3747
+    },
+    {
+      "epoch": 0.36798088410991636,
+      "grad_norm": 9.163630928390635,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 3748
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 8.998910558777084,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3749
+    },
+    {
+      "epoch": 0.3727598566308244,
+      "grad_norm": 9.739178740596293,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3750
+    },
+    {
+      "epoch": 0.3751493428912784,
+      "grad_norm": 10.089086472803181,
+      "learning_rate": 1e-06,
+      "loss": 0.2123,
+      "step": 3751
+    },
+    {
+      "epoch": 0.37753882915173237,
+      "grad_norm": 7.74713021136752,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 3752
+    },
+    {
+      "epoch": 0.37992831541218636,
+      "grad_norm": 7.6526059729129265,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 3753
+    },
+    {
+      "epoch": 0.3823178016726404,
+      "grad_norm": 5.3133583303270235,
+      "learning_rate": 1e-06,
+      "loss": 0.1867,
+      "step": 3754
+    },
+    {
+      "epoch": 0.3847072879330944,
+      "grad_norm": 8.44878287997158,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3755
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 9.01918333967781,
+      "learning_rate": 1e-06,
+      "loss": 0.2235,
+      "step": 3756
+    },
+    {
+      "epoch": 0.38948626045400236,
+      "grad_norm": 9.519855608089888,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 3757
+    },
+    {
+      "epoch": 0.3918757467144564,
+      "grad_norm": 8.626442575834874,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 3758
+    },
+    {
+      "epoch": 0.3942652329749104,
+      "grad_norm": 9.078976654436966,
+      "learning_rate": 1e-06,
+      "loss": 0.1942,
+      "step": 3759
+    },
+    {
+      "epoch": 0.3966547192353644,
+      "grad_norm": 7.790546082396701,
+      "learning_rate": 1e-06,
+      "loss": 0.2188,
+      "step": 3760
+    },
+    {
+      "epoch": 0.39904420549581837,
+      "grad_norm": 8.63168066723789,
+      "learning_rate": 1e-06,
+      "loss": 0.2101,
+      "step": 3761
+    },
+    {
+      "epoch": 0.4014336917562724,
+      "grad_norm": 8.589345411440389,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 3762
+    },
+    {
+      "epoch": 0.4038231780167264,
+      "grad_norm": 10.681706043235076,
+      "learning_rate": 1e-06,
+      "loss": 0.2159,
+      "step": 3763
+    },
+    {
+      "epoch": 0.4062126642771804,
+      "grad_norm": 8.42844122414905,
+      "learning_rate": 1e-06,
+      "loss": 0.2286,
+      "step": 3764
+    },
+    {
+      "epoch": 0.40860215053763443,
+      "grad_norm": 6.894817643002186,
+      "learning_rate": 1e-06,
+      "loss": 0.1908,
+      "step": 3765
+    },
+    {
+      "epoch": 0.4109916367980884,
+      "grad_norm": 8.548105740886523,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 3766
+    },
+    {
+      "epoch": 0.4133811230585424,
+      "grad_norm": 7.533543883074144,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 3767
+    },
+    {
+      "epoch": 0.4157706093189964,
+      "grad_norm": 7.090917304494287,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 3768
+    },
+    {
+      "epoch": 0.41816009557945044,
+      "grad_norm": 7.321583931210473,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3769
+    },
+    {
+      "epoch": 0.42054958183990443,
+      "grad_norm": 7.8331727855626685,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 3770
+    },
+    {
+      "epoch": 0.4229390681003584,
+      "grad_norm": 9.591116878256516,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 3771
+    },
+    {
+      "epoch": 0.4253285543608124,
+      "grad_norm": 7.350763742965476,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 3772
+    },
+    {
+      "epoch": 0.42771804062126645,
+      "grad_norm": 8.338111896928632,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 3773
+    },
+    {
+      "epoch": 0.43010752688172044,
+      "grad_norm": 8.248299866598613,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 3774
+    },
+    {
+      "epoch": 0.4324970131421744,
+      "grad_norm": 9.746641351716532,
+      "learning_rate": 1e-06,
+      "loss": 0.1835,
+      "step": 3775
+    },
+    {
+      "epoch": 0.4348864994026284,
+      "grad_norm": 7.358263783719501,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 3776
+    },
+    {
+      "epoch": 0.43727598566308246,
+      "grad_norm": 7.284483100994395,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 3777
+    },
+    {
+      "epoch": 0.43966547192353644,
+      "grad_norm": 9.629450017280945,
+      "learning_rate": 1e-06,
+      "loss": 0.2078,
+      "step": 3778
+    },
+    {
+      "epoch": 0.44205495818399043,
+      "grad_norm": 8.991802067638426,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 3779
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 6.24556516719407,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 3780
+    },
+    {
+      "epoch": 0.44683393070489846,
+      "grad_norm": 10.062112326068592,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 3781
+    },
+    {
+      "epoch": 0.44922341696535245,
+      "grad_norm": 8.378530149553905,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 3782
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 8.158910561955121,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 3783
+    },
+    {
+      "epoch": 0.4540023894862604,
+      "grad_norm": 8.420368007182244,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 3784
+    },
+    {
+      "epoch": 0.45639187574671447,
+      "grad_norm": 8.224280623632819,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 3785
+    },
+    {
+      "epoch": 0.45878136200716846,
+      "grad_norm": 8.978335925012956,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 3786
+    },
+    {
+      "epoch": 0.46117084826762245,
+      "grad_norm": 8.954470799479202,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3787
+    },
+    {
+      "epoch": 0.4635603345280765,
+      "grad_norm": 9.689473409661273,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 3788
+    },
+    {
+      "epoch": 0.4659498207885305,
+      "grad_norm": 9.577824559923464,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 3789
+    },
+    {
+      "epoch": 0.46833930704898447,
+      "grad_norm": 8.779258707145736,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 3790
+    },
+    {
+      "epoch": 0.47072879330943845,
+      "grad_norm": 10.990838698214718,
+      "learning_rate": 1e-06,
+      "loss": 0.2075,
+      "step": 3791
+    },
+    {
+      "epoch": 0.4731182795698925,
+      "grad_norm": 6.032869886199108,
+      "learning_rate": 1e-06,
+      "loss": 0.1897,
+      "step": 3792
+    },
+    {
+      "epoch": 0.4755077658303465,
+      "grad_norm": 9.712723903708468,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 3793
+    },
+    {
+      "epoch": 0.4778972520908005,
+      "grad_norm": 10.620208406001808,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 3794
+    },
+    {
+      "epoch": 0.48028673835125446,
+      "grad_norm": 6.81486149280884,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 3795
+    },
+    {
+      "epoch": 0.4826762246117085,
+      "grad_norm": 7.020603539294711,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 3796
+    },
+    {
+      "epoch": 0.4850657108721625,
+      "grad_norm": 9.08744967378207,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 3797
+    },
+    {
+      "epoch": 0.4874551971326165,
+      "grad_norm": 9.524745656819263,
+      "learning_rate": 1e-06,
+      "loss": 0.2205,
+      "step": 3798
+    },
+    {
+      "epoch": 0.48984468339307047,
+      "grad_norm": 8.424966875642353,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3799
+    },
+    {
+      "epoch": 0.4922341696535245,
+      "grad_norm": 9.61663990496366,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 3800
+    },
+    {
+      "epoch": 0.4946236559139785,
+      "grad_norm": 6.948043826006782,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 3801
+    },
+    {
+      "epoch": 0.4970131421744325,
+      "grad_norm": 8.89355578211924,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 3802
+    },
+    {
+      "epoch": 0.4994026284348865,
+      "grad_norm": 8.743134388457326,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 3803
+    },
+    {
+      "epoch": 0.5017921146953405,
+      "grad_norm": 8.542660476819085,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 3804
+    },
+    {
+      "epoch": 0.5041816009557945,
+      "grad_norm": 7.671850244579028,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 3805
+    },
+    {
+      "epoch": 0.5065710872162486,
+      "grad_norm": 9.94029271081964,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 3806
+    },
+    {
+      "epoch": 0.5089605734767025,
+      "grad_norm": 10.053363449471522,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 3807
+    },
+    {
+      "epoch": 0.5113500597371565,
+      "grad_norm": 9.83368015922066,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 3808
+    },
+    {
+      "epoch": 0.5137395459976105,
+      "grad_norm": 7.70980165025909,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 3809
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 6.546802767369563,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 3810
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 8.19967340727562,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 3811
+    },
+    {
+      "epoch": 0.5209080047789725,
+      "grad_norm": 7.514641809553528,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 3812
+    },
+    {
+      "epoch": 0.5232974910394266,
+      "grad_norm": 10.006166895750248,
+      "learning_rate": 1e-06,
+      "loss": 0.218,
+      "step": 3813
+    },
+    {
+      "epoch": 0.5256869772998806,
+      "grad_norm": 8.000503111570325,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 3814
+    },
+    {
+      "epoch": 0.5280764635603346,
+      "grad_norm": 9.592591648321985,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 3815
+    },
+    {
+      "epoch": 0.5304659498207885,
+      "grad_norm": 8.874153049533009,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3816
+    },
+    {
+      "epoch": 0.5328554360812425,
+      "grad_norm": 8.625887117456022,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 3817
+    },
+    {
+      "epoch": 0.5352449223416965,
+      "grad_norm": 8.389755833628543,
+      "learning_rate": 1e-06,
+      "loss": 0.2045,
+      "step": 3818
+    },
+    {
+      "epoch": 0.5376344086021505,
+      "grad_norm": 5.772443705560726,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 3819
+    },
+    {
+      "epoch": 0.5400238948626045,
+      "grad_norm": 11.245169548321895,
+      "learning_rate": 1e-06,
+      "loss": 0.1897,
+      "step": 3820
+    },
+    {
+      "epoch": 0.5424133811230586,
+      "grad_norm": 7.238900423474743,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 3821
+    },
+    {
+      "epoch": 0.5448028673835126,
+      "grad_norm": 9.018334097932051,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 3822
+    },
+    {
+      "epoch": 0.5471923536439666,
+      "grad_norm": 8.792122212306737,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 3823
+    },
+    {
+      "epoch": 0.5495818399044206,
+      "grad_norm": 7.51870485966665,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 3824
+    },
+    {
+      "epoch": 0.5519713261648745,
+      "grad_norm": 10.231969126953905,
+      "learning_rate": 1e-06,
+      "loss": 0.2043,
+      "step": 3825
+    },
+    {
+      "epoch": 0.5543608124253285,
+      "grad_norm": 7.154869587510789,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 3826
+    },
+    {
+      "epoch": 0.5567502986857825,
+      "grad_norm": 6.757657891181742,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 3827
+    },
+    {
+      "epoch": 0.5591397849462365,
+      "grad_norm": 10.67735707843371,
+      "learning_rate": 1e-06,
+      "loss": 0.2102,
+      "step": 3828
+    },
+    {
+      "epoch": 0.5615292712066906,
+      "grad_norm": 8.0031795492379,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 3829
+    },
+    {
+      "epoch": 0.5639187574671446,
+      "grad_norm": 8.742986642065453,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 3830
+    },
+    {
+      "epoch": 0.5663082437275986,
+      "grad_norm": 9.46237987558654,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 3831
+    },
+    {
+      "epoch": 0.5686977299880526,
+      "grad_norm": 9.329021557861493,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 3832
+    },
+    {
+      "epoch": 0.5710872162485066,
+      "grad_norm": 6.494455984407059,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 3833
+    },
+    {
+      "epoch": 0.5734767025089605,
+      "grad_norm": 9.115811903017228,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 3834
+    },
+    {
+      "epoch": 0.5758661887694145,
+      "grad_norm": 7.87703081305669,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 3835
+    },
+    {
+      "epoch": 0.5782556750298686,
+      "grad_norm": 7.521416756159578,
+      "learning_rate": 1e-06,
+      "loss": 0.2163,
+      "step": 3836
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 7.861971488121043,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 3837
+    },
+    {
+      "epoch": 0.5830346475507766,
+      "grad_norm": 7.747691202480335,
+      "learning_rate": 1e-06,
+      "loss": 0.1748,
+      "step": 3838
+    },
+    {
+      "epoch": 0.5854241338112306,
+      "grad_norm": 7.080879533168128,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 3839
+    },
+    {
+      "epoch": 0.5878136200716846,
+      "grad_norm": 7.456270114669404,
+      "learning_rate": 1e-06,
+      "loss": 0.2012,
+      "step": 3840
+    },
+    {
+      "epoch": 0.5902031063321386,
+      "grad_norm": 8.967561451623796,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 3841
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 7.998584161371586,
+      "learning_rate": 1e-06,
+      "loss": 0.2064,
+      "step": 3842
+    },
+    {
+      "epoch": 0.5949820788530465,
+      "grad_norm": 9.609867930641338,
+      "learning_rate": 1e-06,
+      "loss": 0.2149,
+      "step": 3843
+    },
+    {
+      "epoch": 0.5973715651135006,
+      "grad_norm": 10.021441658330124,
+      "learning_rate": 1e-06,
+      "loss": 0.2103,
+      "step": 3844
+    },
+    {
+      "epoch": 0.5997610513739546,
+      "grad_norm": 7.750495099773531,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 3845
+    },
+    {
+      "epoch": 0.6021505376344086,
+      "grad_norm": 8.006162091088806,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 3846
+    },
+    {
+      "epoch": 0.6045400238948626,
+      "grad_norm": 9.926898350837252,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 3847
+    },
+    {
+      "epoch": 0.6069295101553166,
+      "grad_norm": 9.246614410688716,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 3848
+    },
+    {
+      "epoch": 0.6093189964157706,
+      "grad_norm": 10.318445344475538,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 3849
+    },
+    {
+      "epoch": 0.6117084826762246,
+      "grad_norm": 9.188377707908767,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 3850
+    },
+    {
+      "epoch": 0.6140979689366786,
+      "grad_norm": 9.069725308373108,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 3851
+    },
+    {
+      "epoch": 0.6164874551971327,
+      "grad_norm": 10.77719830595121,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3852
+    },
+    {
+      "epoch": 0.6188769414575866,
+      "grad_norm": 9.338510854324932,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 3853
+    },
+    {
+      "epoch": 0.6212664277180406,
+      "grad_norm": 11.125408813798817,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 3854
+    },
+    {
+      "epoch": 0.6236559139784946,
+      "grad_norm": 9.580112757228912,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 3855
+    },
+    {
+      "epoch": 0.6260454002389486,
+      "grad_norm": 9.487245677257947,
+      "learning_rate": 1e-06,
+      "loss": 0.2183,
+      "step": 3856
+    },
+    {
+      "epoch": 0.6284348864994026,
+      "grad_norm": 9.754827682003937,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 3857
+    },
+    {
+      "epoch": 0.6308243727598566,
+      "grad_norm": 8.763404524111268,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 3858
+    },
+    {
+      "epoch": 0.6332138590203107,
+      "grad_norm": 9.235767492144024,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 3859
+    },
+    {
+      "epoch": 0.6356033452807647,
+      "grad_norm": 7.537599665779962,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 3860
+    },
+    {
+      "epoch": 0.6379928315412187,
+      "grad_norm": 7.742507807532268,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 3861
+    },
+    {
+      "epoch": 0.6403823178016727,
+      "grad_norm": 7.69953049473014,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 3862
+    },
+    {
+      "epoch": 0.6427718040621266,
+      "grad_norm": 9.415402185614859,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 3863
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 9.96049559274456,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 3864
+    },
+    {
+      "epoch": 0.6475507765830346,
+      "grad_norm": 6.885381442768574,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 3865
+    },
+    {
+      "epoch": 0.6499402628434886,
+      "grad_norm": 7.937305073038304,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 3866
+    },
+    {
+      "epoch": 0.6523297491039427,
+      "grad_norm": 9.120981471771548,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 3867
+    },
+    {
+      "epoch": 0.6547192353643967,
+      "grad_norm": 9.705118485993165,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 3868
+    },
+    {
+      "epoch": 0.6571087216248507,
+      "grad_norm": 7.714333228198665,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 3869
+    },
+    {
+      "epoch": 0.6594982078853047,
+      "grad_norm": 7.807215035109162,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 3870
+    },
+    {
+      "epoch": 0.6618876941457587,
+      "grad_norm": 8.209473332565755,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 3871
+    },
+    {
+      "epoch": 0.6642771804062126,
+      "grad_norm": 6.813656792636022,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 3872
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 8.240830512050731,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 3873
+    },
+    {
+      "epoch": 0.6690561529271206,
+      "grad_norm": 7.853068272536695,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 3874
+    },
+    {
+      "epoch": 0.6714456391875747,
+      "grad_norm": 10.359505541598018,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 3875
+    },
+    {
+      "epoch": 0.6738351254480287,
+      "grad_norm": 9.711723720834307,
+      "learning_rate": 1e-06,
+      "loss": 0.2125,
+      "step": 3876
+    },
+    {
+      "epoch": 0.6762246117084827,
+      "grad_norm": 8.048905526386436,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 3877
+    },
+    {
+      "epoch": 0.6786140979689367,
+      "grad_norm": 6.784405116174484,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 3878
+    },
+    {
+      "epoch": 0.6810035842293907,
+      "grad_norm": 9.333990175369514,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 3879
+    },
+    {
+      "epoch": 0.6833930704898447,
+      "grad_norm": 9.610329360098497,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 3880
+    },
+    {
+      "epoch": 0.6857825567502986,
+      "grad_norm": 8.705028590524007,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 3881
+    },
+    {
+      "epoch": 0.6881720430107527,
+      "grad_norm": 9.589704898809396,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 3882
+    },
+    {
+      "epoch": 0.6905615292712067,
+      "grad_norm": 8.181301429008283,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 3883
+    },
+    {
+      "epoch": 0.6929510155316607,
+      "grad_norm": 11.178611936948956,
+      "learning_rate": 1e-06,
+      "loss": 0.2136,
+      "step": 3884
+    },
+    {
+      "epoch": 0.6953405017921147,
+      "grad_norm": 7.266511170921256,
+      "learning_rate": 1e-06,
+      "loss": 0.1923,
+      "step": 3885
+    },
+    {
+      "epoch": 0.6977299880525687,
+      "grad_norm": 8.91149178979861,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 3886
+    },
+    {
+      "epoch": 0.7001194743130227,
+      "grad_norm": 7.79422852472028,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 3887
+    },
+    {
+      "epoch": 0.7025089605734767,
+      "grad_norm": 8.507010152477113,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 3888
+    },
+    {
+      "epoch": 0.7048984468339307,
+      "grad_norm": 7.861430736665809,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 3889
+    },
+    {
+      "epoch": 0.7072879330943848,
+      "grad_norm": 7.630116777703454,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 3890
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 8.830241348853304,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 3891
+    },
+    {
+      "epoch": 0.7120669056152927,
+      "grad_norm": 8.118786289148073,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 3892
+    },
+    {
+      "epoch": 0.7144563918757467,
+      "grad_norm": 9.311364033074245,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 3893
+    },
+    {
+      "epoch": 0.7168458781362007,
+      "grad_norm": 7.966328352553812,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 3894
+    },
+    {
+      "epoch": 0.7192353643966547,
+      "grad_norm": 9.018692364994596,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 3895
+    },
+    {
+      "epoch": 0.7216248506571087,
+      "grad_norm": 8.941181890406781,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 3896
+    },
+    {
+      "epoch": 0.7240143369175627,
+      "grad_norm": 10.227815293567755,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 3897
+    },
+    {
+      "epoch": 0.7264038231780168,
+      "grad_norm": 7.882585392561184,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 3898
+    },
+    {
+      "epoch": 0.7287933094384708,
+      "grad_norm": 7.398785536056563,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 3899
+    },
+    {
+      "epoch": 0.7311827956989247,
+      "grad_norm": 8.913622877295865,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 3900
+    },
+    {
+      "epoch": 0.7335722819593787,
+      "grad_norm": 9.586512757684794,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 3901
+    },
+    {
+      "epoch": 0.7359617682198327,
+      "grad_norm": 10.101311105858079,
+      "learning_rate": 1e-06,
+      "loss": 0.2135,
+      "step": 3902
+    },
+    {
+      "epoch": 0.7383512544802867,
+      "grad_norm": 7.606531433400806,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3903
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 12.974971362838577,
+      "learning_rate": 1e-06,
+      "loss": 0.2143,
+      "step": 3904
+    },
+    {
+      "epoch": 0.7431302270011948,
+      "grad_norm": 8.970693749974794,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 3905
+    },
+    {
+      "epoch": 0.7455197132616488,
+      "grad_norm": 10.183730183797875,
+      "learning_rate": 1e-06,
+      "loss": 0.2124,
+      "step": 3906
+    },
+    {
+      "epoch": 0.7479091995221028,
+      "grad_norm": 7.936253520489404,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 3907
+    },
+    {
+      "epoch": 0.7502986857825568,
+      "grad_norm": 8.85764823632083,
+      "learning_rate": 1e-06,
+      "loss": 0.2126,
+      "step": 3908
+    },
+    {
+      "epoch": 0.7526881720430108,
+      "grad_norm": 9.041713824902015,
+      "learning_rate": 1e-06,
+      "loss": 0.2186,
+      "step": 3909
+    },
+    {
+      "epoch": 0.7550776583034647,
+      "grad_norm": 10.317132152211704,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 3910
+    },
+    {
+      "epoch": 0.7574671445639187,
+      "grad_norm": 9.365982990853977,
+      "learning_rate": 1e-06,
+      "loss": 0.2012,
+      "step": 3911
+    },
+    {
+      "epoch": 0.7598566308243727,
+      "grad_norm": 7.7924550627357565,
+      "learning_rate": 1e-06,
+      "loss": 0.2095,
+      "step": 3912
+    },
+    {
+      "epoch": 0.7622461170848268,
+      "grad_norm": 8.69927265922315,
+      "learning_rate": 1e-06,
+      "loss": 0.2212,
+      "step": 3913
+    },
+    {
+      "epoch": 0.7646356033452808,
+      "grad_norm": 8.797826729128202,
+      "learning_rate": 1e-06,
+      "loss": 0.2106,
+      "step": 3914
+    },
+    {
+      "epoch": 0.7670250896057348,
+      "grad_norm": 6.881773808288041,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 3915
+    },
+    {
+      "epoch": 0.7694145758661888,
+      "grad_norm": 10.190786009934817,
+      "learning_rate": 1e-06,
+      "loss": 0.2216,
+      "step": 3916
+    },
+    {
+      "epoch": 0.7718040621266428,
+      "grad_norm": 8.032975479419523,
+      "learning_rate": 1e-06,
+      "loss": 0.2064,
+      "step": 3917
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 9.255106082736907,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 3918
+    },
+    {
+      "epoch": 0.7765830346475507,
+      "grad_norm": 9.714624714727266,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 3919
+    },
+    {
+      "epoch": 0.7789725209080047,
+      "grad_norm": 8.296277151990395,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3920
+    },
+    {
+      "epoch": 0.7813620071684588,
+      "grad_norm": 9.11972629445385,
+      "learning_rate": 1e-06,
+      "loss": 0.1862,
+      "step": 3921
+    },
+    {
+      "epoch": 0.7837514934289128,
+      "grad_norm": 7.724699743680573,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 3922
+    },
+    {
+      "epoch": 0.7861409796893668,
+      "grad_norm": 7.722929664117139,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 3923
+    },
+    {
+      "epoch": 0.7885304659498208,
+      "grad_norm": 7.630950693010527,
+      "learning_rate": 1e-06,
+      "loss": 0.2355,
+      "step": 3924
+    },
+    {
+      "epoch": 0.7909199522102748,
+      "grad_norm": 7.770032241618276,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 3925
+    },
+    {
+      "epoch": 0.7933094384707288,
+      "grad_norm": 8.122819782428486,
+      "learning_rate": 1e-06,
+      "loss": 0.201,
+      "step": 3926
+    },
+    {
+      "epoch": 0.7956989247311828,
+      "grad_norm": 7.584581259102992,
+      "learning_rate": 1e-06,
+      "loss": 0.2092,
+      "step": 3927
+    },
+    {
+      "epoch": 0.7980884109916367,
+      "grad_norm": 6.240386325836056,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 3928
+    },
+    {
+      "epoch": 0.8004778972520908,
+      "grad_norm": 7.164251880400724,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 3929
+    },
+    {
+      "epoch": 0.8028673835125448,
+      "grad_norm": 8.18777849543532,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 3930
+    },
+    {
+      "epoch": 0.8052568697729988,
+      "grad_norm": 9.223861383856763,
+      "learning_rate": 1e-06,
+      "loss": 0.2174,
+      "step": 3931
+    },
+    {
+      "epoch": 0.8076463560334528,
+      "grad_norm": 5.363201334498833,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 3932
+    },
+    {
+      "epoch": 0.8100358422939068,
+      "grad_norm": 7.792049434858007,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 3933
+    },
+    {
+      "epoch": 0.8124253285543608,
+      "grad_norm": 9.567110282298911,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 3934
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 8.333831641105354,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 3935
+    },
+    {
+      "epoch": 0.8172043010752689,
+      "grad_norm": 8.348414430328939,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 3936
+    },
+    {
+      "epoch": 0.8195937873357229,
+      "grad_norm": 6.659483572299661,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 3937
+    },
+    {
+      "epoch": 0.8219832735961768,
+      "grad_norm": 7.7516092677816895,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 3938
+    },
+    {
+      "epoch": 0.8243727598566308,
+      "grad_norm": 7.792547737785921,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 3939
+    },
+    {
+      "epoch": 0.8267622461170848,
+      "grad_norm": 8.001738860353147,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 3940
+    },
+    {
+      "epoch": 0.8291517323775388,
+      "grad_norm": 9.081086254921354,
+      "learning_rate": 1e-06,
+      "loss": 0.1881,
+      "step": 3941
+    },
+    {
+      "epoch": 0.8315412186379928,
+      "grad_norm": 7.754176155777894,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 3942
+    },
+    {
+      "epoch": 0.8339307048984468,
+      "grad_norm": 7.739861358971692,
+      "learning_rate": 1e-06,
+      "loss": 0.2009,
+      "step": 3943
+    },
+    {
+      "epoch": 0.8363201911589009,
+      "grad_norm": 9.329541020246404,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 3944
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 8.087181905661343,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 3945
+    },
+    {
+      "epoch": 0.8410991636798089,
+      "grad_norm": 8.539620739905304,
+      "learning_rate": 1e-06,
+      "loss": 0.2067,
+      "step": 3946
+    },
+    {
+      "epoch": 0.8434886499402628,
+      "grad_norm": 9.8143384380899,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 3947
+    },
+    {
+      "epoch": 0.8458781362007168,
+      "grad_norm": 7.398556772385823,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3948
+    },
+    {
+      "epoch": 0.8482676224611708,
+      "grad_norm": 6.96449858422067,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 3949
+    },
+    {
+      "epoch": 0.8506571087216248,
+      "grad_norm": 6.056296437742272,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 3950
+    },
+    {
+      "epoch": 0.8530465949820788,
+      "grad_norm": 10.214661998116503,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 3951
+    },
+    {
+      "epoch": 0.8554360812425329,
+      "grad_norm": 11.049849633314246,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 3952
+    },
+    {
+      "epoch": 0.8578255675029869,
+      "grad_norm": 9.523261749425805,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 3953
+    },
+    {
+      "epoch": 0.8602150537634409,
+      "grad_norm": 7.927924521580743,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 3954
+    },
+    {
+      "epoch": 0.8626045400238949,
+      "grad_norm": 5.763004282574792,
+      "learning_rate": 1e-06,
+      "loss": 0.1834,
+      "step": 3955
+    },
+    {
+      "epoch": 0.8649940262843488,
+      "grad_norm": 7.09513603861317,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 3956
+    },
+    {
+      "epoch": 0.8673835125448028,
+      "grad_norm": 10.415477295757496,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 3957
+    },
+    {
+      "epoch": 0.8697729988052568,
+      "grad_norm": 9.373527933065931,
+      "learning_rate": 1e-06,
+      "loss": 0.1935,
+      "step": 3958
+    },
+    {
+      "epoch": 0.8721624850657109,
+      "grad_norm": 10.40226961793056,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 3959
+    },
+    {
+      "epoch": 0.8745519713261649,
+      "grad_norm": 8.270486889209241,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 3960
+    },
+    {
+      "epoch": 0.8769414575866189,
+      "grad_norm": 8.254800464842049,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 3961
+    },
+    {
+      "epoch": 0.8793309438470729,
+      "grad_norm": 6.577145093302454,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 3962
+    },
+    {
+      "epoch": 0.8817204301075269,
+      "grad_norm": 7.933891750759702,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 3963
+    },
+    {
+      "epoch": 0.8841099163679809,
+      "grad_norm": 10.417982207857428,
+      "learning_rate": 1e-06,
+      "loss": 0.2179,
+      "step": 3964
+    },
+    {
+      "epoch": 0.8864994026284349,
+      "grad_norm": 9.285796115484677,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 3965
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 7.677617437665883,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 3966
+    },
+    {
+      "epoch": 0.8912783751493429,
+      "grad_norm": 7.956133061654264,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 3967
+    },
+    {
+      "epoch": 0.8936678614097969,
+      "grad_norm": 8.939640846900545,
+      "learning_rate": 1e-06,
+      "loss": 0.1992,
+      "step": 3968
+    },
+    {
+      "epoch": 0.8960573476702509,
+      "grad_norm": 8.419414540719869,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 3969
+    },
+    {
+      "epoch": 0.8984468339307049,
+      "grad_norm": 8.227910687093484,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 3970
+    },
+    {
+      "epoch": 0.9008363201911589,
+      "grad_norm": 8.372153263218957,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 3971
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 10.321335869945656,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 3972
+    },
+    {
+      "epoch": 0.9056152927120669,
+      "grad_norm": 7.570442998785368,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 3973
+    },
+    {
+      "epoch": 0.9080047789725209,
+      "grad_norm": 8.240462958990515,
+      "learning_rate": 1e-06,
+      "loss": 0.1812,
+      "step": 3974
+    },
+    {
+      "epoch": 0.910394265232975,
+      "grad_norm": 7.365622070787061,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 3975
+    },
+    {
+      "epoch": 0.9127837514934289,
+      "grad_norm": 7.774118369951355,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 3976
+    },
+    {
+      "epoch": 0.9151732377538829,
+      "grad_norm": 6.330645491745627,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 3977
+    },
+    {
+      "epoch": 0.9175627240143369,
+      "grad_norm": 8.522465930902069,
+      "learning_rate": 1e-06,
+      "loss": 0.1974,
+      "step": 3978
+    },
+    {
+      "epoch": 0.9199522102747909,
+      "grad_norm": 6.78619657649288,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 3979
+    },
+    {
+      "epoch": 0.9223416965352449,
+      "grad_norm": 10.176544682304245,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 3980
+    },
+    {
+      "epoch": 0.9247311827956989,
+      "grad_norm": 10.351174161324964,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 3981
+    },
+    {
+      "epoch": 0.927120669056153,
+      "grad_norm": 13.224947079848285,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 3982
+    },
+    {
+      "epoch": 0.929510155316607,
+      "grad_norm": 8.85630813084483,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 3983
+    },
+    {
+      "epoch": 0.931899641577061,
+      "grad_norm": 8.418501684742374,
+      "learning_rate": 1e-06,
+      "loss": 0.2109,
+      "step": 3984
+    },
+    {
+      "epoch": 0.9342891278375149,
+      "grad_norm": 7.145052534020787,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 3985
+    },
+    {
+      "epoch": 0.9366786140979689,
+      "grad_norm": 9.632607419770258,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 3986
+    },
+    {
+      "epoch": 0.9390681003584229,
+      "grad_norm": 8.833652282724863,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 3987
+    },
+    {
+      "epoch": 0.9414575866188769,
+      "grad_norm": 8.792160845642702,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 3988
+    },
+    {
+      "epoch": 0.9438470728793309,
+      "grad_norm": 7.666196627029027,
+      "learning_rate": 1e-06,
+      "loss": 0.1828,
+      "step": 3989
+    },
+    {
+      "epoch": 0.946236559139785,
+      "grad_norm": 8.72991796904069,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 3990
+    },
+    {
+      "epoch": 0.948626045400239,
+      "grad_norm": 10.65150563124372,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 3991
+    },
+    {
+      "epoch": 0.951015531660693,
+      "grad_norm": 8.039348358859119,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 3992
+    },
+    {
+      "epoch": 0.953405017921147,
+      "grad_norm": 8.88783400917979,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 3993
+    },
+    {
+      "epoch": 0.955794504181601,
+      "grad_norm": 7.433860690732034,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 3994
+    },
+    {
+      "epoch": 0.9581839904420549,
+      "grad_norm": 7.102364364915397,
+      "learning_rate": 1e-06,
+      "loss": 0.2179,
+      "step": 3995
+    },
+    {
+      "epoch": 0.9605734767025089,
+      "grad_norm": 9.979397214054648,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 3996
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 8.114502709835135,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 3997
+    },
+    {
+      "epoch": 0.965352449223417,
+      "grad_norm": 7.9271123414288445,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 3998
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 9.505873411558944,
+      "learning_rate": 1e-06,
+      "loss": 0.2075,
+      "step": 3999
+    },
+    {
+      "epoch": 0.970131421744325,
+      "grad_norm": 8.332004002787496,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 4000
+    },
+    {
+      "epoch": 0.970131421744325,
+      "eval_loss": 3.0996291637420654,
+      "eval_runtime": 21.9045,
+      "eval_samples_per_second": 45.653,
+      "eval_steps_per_second": 5.707,
+      "step": 4000
+    },
+    {
+      "epoch": 0.972520908004779,
+      "grad_norm": 8.024551147587205,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 4001
+    },
+    {
+      "epoch": 0.974910394265233,
+      "grad_norm": 4.936267554177057,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 4002
+    },
+    {
+      "epoch": 0.977299880525687,
+      "grad_norm": 7.334987041748554,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 4003
+    },
+    {
+      "epoch": 0.9796893667861409,
+      "grad_norm": 8.044567409756928,
+      "learning_rate": 1e-06,
+      "loss": 0.2207,
+      "step": 4004
+    },
+    {
+      "epoch": 0.982078853046595,
+      "grad_norm": 10.098901575921436,
+      "learning_rate": 1e-06,
+      "loss": 0.1941,
+      "step": 4005
+    },
+    {
+      "epoch": 0.984468339307049,
+      "grad_norm": 9.75457414204856,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 4006
+    },
+    {
+      "epoch": 0.986857825567503,
+      "grad_norm": 8.166860625062789,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 4007
+    },
+    {
+      "epoch": 0.989247311827957,
+      "grad_norm": 6.5932030417573,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 4008
+    },
+    {
+      "epoch": 0.991636798088411,
+      "grad_norm": 7.745652967618301,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 4009
+    },
+    {
+      "epoch": 0.994026284348865,
+      "grad_norm": 7.166488228793955,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 4010
+    },
+    {
+      "epoch": 0.996415770609319,
+      "grad_norm": 7.41984061629122,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 4011
+    },
+    {
+      "epoch": 0.998805256869773,
+      "grad_norm": 7.104350993795116,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4012
+    },
+    {
+      "epoch": 0.998805256869773,
+      "eval_loss": 3.036895275115967,
+      "eval_runtime": 21.6638,
+      "eval_samples_per_second": 46.16,
+      "eval_steps_per_second": 5.77,
+      "step": 4012
+    },
+    {
+      "epoch": 0.002398081534772182,
+      "grad_norm": 7.7196938652152,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 4013
+    },
+    {
+      "epoch": 0.004796163069544364,
+      "grad_norm": 7.215788965696631,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 4014
+    },
+    {
+      "epoch": 0.007194244604316547,
+      "grad_norm": 6.8973231052553245,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 4015
+    },
+    {
+      "epoch": 0.009592326139088728,
+      "grad_norm": 6.7037894200432815,
+      "learning_rate": 1e-06,
+      "loss": 0.1726,
+      "step": 4016
+    },
+    {
+      "epoch": 0.011990407673860911,
+      "grad_norm": 8.092799726781038,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 4017
+    },
+    {
+      "epoch": 0.014388489208633094,
+      "grad_norm": 6.623953529872254,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 4018
+    },
+    {
+      "epoch": 0.016786570743405275,
+      "grad_norm": 9.079537343121904,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 4019
+    },
+    {
+      "epoch": 0.019184652278177457,
+      "grad_norm": 7.552935672030848,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 4020
+    },
+    {
+      "epoch": 0.02158273381294964,
+      "grad_norm": 7.118026170475696,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 4021
+    },
+    {
+      "epoch": 0.023980815347721823,
+      "grad_norm": 7.198553219908629,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 4022
+    },
+    {
+      "epoch": 0.026378896882494004,
+      "grad_norm": 7.74909809452316,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 4023
+    },
+    {
+      "epoch": 0.02877697841726619,
+      "grad_norm": 8.201961003950743,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 4024
+    },
+    {
+      "epoch": 0.03117505995203837,
+      "grad_norm": 8.748982012047401,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 4025
+    },
+    {
+      "epoch": 0.03357314148681055,
+      "grad_norm": 8.28835353987331,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 4026
+    },
+    {
+      "epoch": 0.03597122302158273,
+      "grad_norm": 7.567315153441481,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 4027
+    },
+    {
+      "epoch": 0.03836930455635491,
+      "grad_norm": 11.692322655900753,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 4028
+    },
+    {
+      "epoch": 0.0407673860911271,
+      "grad_norm": 10.0577796238405,
+      "learning_rate": 1e-06,
+      "loss": 0.1883,
+      "step": 4029
+    },
+    {
+      "epoch": 0.04316546762589928,
+      "grad_norm": 8.877725066703995,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 4030
+    },
+    {
+      "epoch": 0.045563549160671464,
+      "grad_norm": 9.8905083004086,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 4031
+    },
+    {
+      "epoch": 0.047961630695443645,
+      "grad_norm": 9.094355696169814,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 4032
+    },
+    {
+      "epoch": 0.050359712230215826,
+      "grad_norm": 10.075684514749318,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 4033
+    },
+    {
+      "epoch": 0.05275779376498801,
+      "grad_norm": 7.2013048634468895,
+      "learning_rate": 1e-06,
+      "loss": 0.1754,
+      "step": 4034
+    },
+    {
+      "epoch": 0.05515587529976019,
+      "grad_norm": 7.566094215066042,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 4035
+    },
+    {
+      "epoch": 0.05755395683453238,
+      "grad_norm": 8.689619520122157,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 4036
+    },
+    {
+      "epoch": 0.05995203836930456,
+      "grad_norm": 8.001378372470874,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 4037
+    },
+    {
+      "epoch": 0.06235011990407674,
+      "grad_norm": 8.38363516190156,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 4038
+    },
+    {
+      "epoch": 0.06474820143884892,
+      "grad_norm": 6.451483172813302,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 4039
+    },
+    {
+      "epoch": 0.0671462829736211,
+      "grad_norm": 8.010532701501116,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 4040
+    },
+    {
+      "epoch": 0.06954436450839328,
+      "grad_norm": 7.3787279523993945,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 4041
+    },
+    {
+      "epoch": 0.07194244604316546,
+      "grad_norm": 8.162238692906044,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 4042
+    },
+    {
+      "epoch": 0.07434052757793765,
+      "grad_norm": 9.67442348444359,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 4043
+    },
+    {
+      "epoch": 0.07673860911270983,
+      "grad_norm": 6.521773255927714,
+      "learning_rate": 1e-06,
+      "loss": 0.2042,
+      "step": 4044
+    },
+    {
+      "epoch": 0.07913669064748201,
+      "grad_norm": 6.626893949708232,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 4045
+    },
+    {
+      "epoch": 0.0815347721822542,
+      "grad_norm": 9.361067988329518,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 4046
+    },
+    {
+      "epoch": 0.08393285371702638,
+      "grad_norm": 7.705341397081286,
+      "learning_rate": 1e-06,
+      "loss": 0.2078,
+      "step": 4047
+    },
+    {
+      "epoch": 0.08633093525179857,
+      "grad_norm": 6.935434290949853,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 4048
+    },
+    {
+      "epoch": 0.08872901678657075,
+      "grad_norm": 7.041199763230758,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 4049
+    },
+    {
+      "epoch": 0.09112709832134293,
+      "grad_norm": 7.074979986938193,
+      "learning_rate": 1e-06,
+      "loss": 0.2009,
+      "step": 4050
+    },
+    {
+      "epoch": 0.09352517985611511,
+      "grad_norm": 7.362350002260175,
+      "learning_rate": 1e-06,
+      "loss": 0.1865,
+      "step": 4051
+    },
+    {
+      "epoch": 0.09592326139088729,
+      "grad_norm": 7.814731192167318,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 4052
+    },
+    {
+      "epoch": 0.09832134292565947,
+      "grad_norm": 7.825542533384117,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 4053
+    },
+    {
+      "epoch": 0.10071942446043165,
+      "grad_norm": 8.492407832414582,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 4054
+    },
+    {
+      "epoch": 0.10311750599520383,
+      "grad_norm": 9.375784174103288,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 4055
+    },
+    {
+      "epoch": 0.10551558752997602,
+      "grad_norm": 10.14845663962301,
+      "learning_rate": 1e-06,
+      "loss": 0.2111,
+      "step": 4056
+    },
+    {
+      "epoch": 0.1079136690647482,
+      "grad_norm": 7.488843952858037,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 4057
+    },
+    {
+      "epoch": 0.11031175059952038,
+      "grad_norm": 10.094248483237315,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 4058
+    },
+    {
+      "epoch": 0.11270983213429256,
+      "grad_norm": 6.608557547272705,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 4059
+    },
+    {
+      "epoch": 0.11510791366906475,
+      "grad_norm": 9.034769540498338,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 4060
+    },
+    {
+      "epoch": 0.11750599520383694,
+      "grad_norm": 7.095775435483761,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 4061
+    },
+    {
+      "epoch": 0.11990407673860912,
+      "grad_norm": 7.841899110432263,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 4062
+    },
+    {
+      "epoch": 0.1223021582733813,
+      "grad_norm": 11.918591992099442,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 4063
+    },
+    {
+      "epoch": 0.12470023980815348,
+      "grad_norm": 7.843980040634473,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 4064
+    },
+    {
+      "epoch": 0.12709832134292565,
+      "grad_norm": 9.897400432102028,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 4065
+    },
+    {
+      "epoch": 0.12949640287769784,
+      "grad_norm": 8.283820446205578,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 4066
+    },
+    {
+      "epoch": 0.13189448441247004,
+      "grad_norm": 12.076618820983484,
+      "learning_rate": 1e-06,
+      "loss": 0.2144,
+      "step": 4067
+    },
+    {
+      "epoch": 0.1342925659472422,
+      "grad_norm": 7.6415303270764285,
+      "learning_rate": 1e-06,
+      "loss": 0.2188,
+      "step": 4068
+    },
+    {
+      "epoch": 0.1366906474820144,
+      "grad_norm": 9.19798175343296,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 4069
+    },
+    {
+      "epoch": 0.13908872901678657,
+      "grad_norm": 9.13026540599072,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 4070
+    },
+    {
+      "epoch": 0.14148681055155876,
+      "grad_norm": 7.564075278865214,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 4071
+    },
+    {
+      "epoch": 0.14388489208633093,
+      "grad_norm": 6.679248416022126,
+      "learning_rate": 1e-06,
+      "loss": 0.1859,
+      "step": 4072
+    },
+    {
+      "epoch": 0.14628297362110312,
+      "grad_norm": 9.397333613163147,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 4073
+    },
+    {
+      "epoch": 0.1486810551558753,
+      "grad_norm": 6.931917474867233,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 4074
+    },
+    {
+      "epoch": 0.1510791366906475,
+      "grad_norm": 8.212268730844405,
+      "learning_rate": 1e-06,
+      "loss": 0.1789,
+      "step": 4075
+    },
+    {
+      "epoch": 0.15347721822541965,
+      "grad_norm": 6.865729260141344,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 4076
+    },
+    {
+      "epoch": 0.15587529976019185,
+      "grad_norm": 7.9872046843595275,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 4077
+    },
+    {
+      "epoch": 0.15827338129496402,
+      "grad_norm": 6.195103424762274,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 4078
+    },
+    {
+      "epoch": 0.1606714628297362,
+      "grad_norm": 8.427663041875999,
+      "learning_rate": 1e-06,
+      "loss": 0.2062,
+      "step": 4079
+    },
+    {
+      "epoch": 0.1630695443645084,
+      "grad_norm": 8.798440765567832,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 4080
+    },
+    {
+      "epoch": 0.16546762589928057,
+      "grad_norm": 10.288943305588756,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 4081
+    },
+    {
+      "epoch": 0.16786570743405277,
+      "grad_norm": 8.279937295966374,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 4082
+    },
+    {
+      "epoch": 0.17026378896882494,
+      "grad_norm": 7.905039414596362,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 4083
+    },
+    {
+      "epoch": 0.17266187050359713,
+      "grad_norm": 7.591331160547937,
+      "learning_rate": 1e-06,
+      "loss": 0.2068,
+      "step": 4084
+    },
+    {
+      "epoch": 0.1750599520383693,
+      "grad_norm": 8.86685911357719,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 4085
+    },
+    {
+      "epoch": 0.1774580335731415,
+      "grad_norm": 11.008201205420203,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 4086
+    },
+    {
+      "epoch": 0.17985611510791366,
+      "grad_norm": 9.906655944510451,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 4087
+    },
+    {
+      "epoch": 0.18225419664268586,
+      "grad_norm": 8.63548511805487,
+      "learning_rate": 1e-06,
+      "loss": 0.1782,
+      "step": 4088
+    },
+    {
+      "epoch": 0.18465227817745802,
+      "grad_norm": 7.658125128618825,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 4089
+    },
+    {
+      "epoch": 0.18705035971223022,
+      "grad_norm": 8.61568481074069,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 4090
+    },
+    {
+      "epoch": 0.18944844124700239,
+      "grad_norm": 11.494688142029396,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 4091
+    },
+    {
+      "epoch": 0.19184652278177458,
+      "grad_norm": 9.007051709183967,
+      "learning_rate": 1e-06,
+      "loss": 0.2077,
+      "step": 4092
+    },
+    {
+      "epoch": 0.19424460431654678,
+      "grad_norm": 7.954416003431306,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 4093
+    },
+    {
+      "epoch": 0.19664268585131894,
+      "grad_norm": 10.515139251112604,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 4094
+    },
+    {
+      "epoch": 0.19904076738609114,
+      "grad_norm": 7.0909624755488805,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 4095
+    },
+    {
+      "epoch": 0.2014388489208633,
+      "grad_norm": 8.121023995673362,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 4096
+    },
+    {
+      "epoch": 0.2038369304556355,
+      "grad_norm": 7.721611101614933,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 4097
+    },
+    {
+      "epoch": 0.20623501199040767,
+      "grad_norm": 7.765596138465908,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 4098
+    },
+    {
+      "epoch": 0.20863309352517986,
+      "grad_norm": 6.972190513603777,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 4099
+    },
+    {
+      "epoch": 0.21103117505995203,
+      "grad_norm": 9.594576449287663,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 4100
+    },
+    {
+      "epoch": 0.21342925659472423,
+      "grad_norm": 10.222005580446353,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 4101
+    },
+    {
+      "epoch": 0.2158273381294964,
+      "grad_norm": 8.494913181298378,
+      "learning_rate": 1e-06,
+      "loss": 0.1837,
+      "step": 4102
+    },
+    {
+      "epoch": 0.2182254196642686,
+      "grad_norm": 7.983665883137081,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 4103
+    },
+    {
+      "epoch": 0.22062350119904076,
+      "grad_norm": 8.345794193035209,
+      "learning_rate": 1e-06,
+      "loss": 0.1873,
+      "step": 4104
+    },
+    {
+      "epoch": 0.22302158273381295,
+      "grad_norm": 8.114186903537783,
+      "learning_rate": 1e-06,
+      "loss": 0.2115,
+      "step": 4105
+    },
+    {
+      "epoch": 0.22541966426858512,
+      "grad_norm": 8.599797288536218,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 4106
+    },
+    {
+      "epoch": 0.2278177458033573,
+      "grad_norm": 9.467396251659268,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 4107
+    },
+    {
+      "epoch": 0.2302158273381295,
+      "grad_norm": 9.128509892658647,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 4108
+    },
+    {
+      "epoch": 0.23261390887290168,
+      "grad_norm": 8.312314319570282,
+      "learning_rate": 1e-06,
+      "loss": 0.1982,
+      "step": 4109
+    },
+    {
+      "epoch": 0.23501199040767387,
+      "grad_norm": 9.714613946277497,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 4110
+    },
+    {
+      "epoch": 0.23741007194244604,
+      "grad_norm": 9.572820625848546,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 4111
+    },
+    {
+      "epoch": 0.23980815347721823,
+      "grad_norm": 10.030016322666432,
+      "learning_rate": 1e-06,
+      "loss": 0.2136,
+      "step": 4112
+    },
+    {
+      "epoch": 0.2422062350119904,
+      "grad_norm": 8.54198573269,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 4113
+    },
+    {
+      "epoch": 0.2446043165467626,
+      "grad_norm": 8.578400384373955,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 4114
+    },
+    {
+      "epoch": 0.24700239808153476,
+      "grad_norm": 9.510019974375865,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 4115
+    },
+    {
+      "epoch": 0.24940047961630696,
+      "grad_norm": 9.099686893395265,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 4116
+    },
+    {
+      "epoch": 0.2517985611510791,
+      "grad_norm": 5.37210750584295,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 4117
+    },
+    {
+      "epoch": 0.2541966426858513,
+      "grad_norm": 8.364264651394267,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 4118
+    },
+    {
+      "epoch": 0.2565947242206235,
+      "grad_norm": 9.482506098002455,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 4119
+    },
+    {
+      "epoch": 0.2589928057553957,
+      "grad_norm": 7.572208297828945,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 4120
+    },
+    {
+      "epoch": 0.26139088729016785,
+      "grad_norm": 8.1226725590235,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 4121
+    },
+    {
+      "epoch": 0.2637889688249401,
+      "grad_norm": 6.926743098582435,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 4122
+    },
+    {
+      "epoch": 0.26618705035971224,
+      "grad_norm": 6.752659096178659,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 4123
+    },
+    {
+      "epoch": 0.2685851318944844,
+      "grad_norm": 9.530658652109237,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 4124
+    },
+    {
+      "epoch": 0.2709832134292566,
+      "grad_norm": 8.541812351109163,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 4125
+    },
+    {
+      "epoch": 0.2733812949640288,
+      "grad_norm": 9.602676011720575,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 4126
+    },
+    {
+      "epoch": 0.27577937649880097,
+      "grad_norm": 9.914317994659712,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 4127
+    },
+    {
+      "epoch": 0.27817745803357313,
+      "grad_norm": 7.9594453209081815,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 4128
+    },
+    {
+      "epoch": 0.2805755395683453,
+      "grad_norm": 9.06003000544769,
+      "learning_rate": 1e-06,
+      "loss": 0.1957,
+      "step": 4129
+    },
+    {
+      "epoch": 0.2829736211031175,
+      "grad_norm": 9.731660655950822,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 4130
+    },
+    {
+      "epoch": 0.2853717026378897,
+      "grad_norm": 8.188275587743888,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 4131
+    },
+    {
+      "epoch": 0.28776978417266186,
+      "grad_norm": 7.436238386109959,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 4132
+    },
+    {
+      "epoch": 0.290167865707434,
+      "grad_norm": 8.513916585031966,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 4133
+    },
+    {
+      "epoch": 0.29256594724220625,
+      "grad_norm": 8.956004737771492,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 4134
+    },
+    {
+      "epoch": 0.2949640287769784,
+      "grad_norm": 6.400787053892165,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 4135
+    },
+    {
+      "epoch": 0.2973621103117506,
+      "grad_norm": 9.194441320085833,
+      "learning_rate": 1e-06,
+      "loss": 0.212,
+      "step": 4136
+    },
+    {
+      "epoch": 0.2997601918465228,
+      "grad_norm": 6.496421727953734,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 4137
+    },
+    {
+      "epoch": 0.302158273381295,
+      "grad_norm": 6.910827258995419,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 4138
+    },
+    {
+      "epoch": 0.30455635491606714,
+      "grad_norm": 7.87346524960142,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 4139
+    },
+    {
+      "epoch": 0.3069544364508393,
+      "grad_norm": 8.586690279547748,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 4140
+    },
+    {
+      "epoch": 0.30935251798561153,
+      "grad_norm": 6.786960243535198,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 4141
+    },
+    {
+      "epoch": 0.3117505995203837,
+      "grad_norm": 8.077528275529923,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 4142
+    },
+    {
+      "epoch": 0.31414868105515587,
+      "grad_norm": 6.6362463389775215,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 4143
+    },
+    {
+      "epoch": 0.31654676258992803,
+      "grad_norm": 8.335703301150788,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4144
+    },
+    {
+      "epoch": 0.31894484412470026,
+      "grad_norm": 8.955915410867348,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 4145
+    },
+    {
+      "epoch": 0.3213429256594724,
+      "grad_norm": 7.628312466357048,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 4146
+    },
+    {
+      "epoch": 0.3237410071942446,
+      "grad_norm": 8.375524687589218,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 4147
+    },
+    {
+      "epoch": 0.3261390887290168,
+      "grad_norm": 8.590304956462333,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 4148
+    },
+    {
+      "epoch": 0.328537170263789,
+      "grad_norm": 10.351994982110622,
+      "learning_rate": 1e-06,
+      "loss": 0.2098,
+      "step": 4149
+    },
+    {
+      "epoch": 0.33093525179856115,
+      "grad_norm": 8.33185269069266,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 4150
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 8.376499889665148,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 4151
+    },
+    {
+      "epoch": 0.33573141486810554,
+      "grad_norm": 5.9183194636849805,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4152
+    },
+    {
+      "epoch": 0.3381294964028777,
+      "grad_norm": 7.999160127909184,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 4153
+    },
+    {
+      "epoch": 0.3405275779376499,
+      "grad_norm": 8.375437939988005,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 4154
+    },
+    {
+      "epoch": 0.34292565947242204,
+      "grad_norm": 9.346374750095288,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 4155
+    },
+    {
+      "epoch": 0.34532374100719426,
+      "grad_norm": 8.257570479222741,
+      "learning_rate": 1e-06,
+      "loss": 0.2074,
+      "step": 4156
+    },
+    {
+      "epoch": 0.34772182254196643,
+      "grad_norm": 8.0010436101039,
+      "learning_rate": 1e-06,
+      "loss": 0.1963,
+      "step": 4157
+    },
+    {
+      "epoch": 0.3501199040767386,
+      "grad_norm": 8.748575324385591,
+      "learning_rate": 1e-06,
+      "loss": 0.2104,
+      "step": 4158
+    },
+    {
+      "epoch": 0.35251798561151076,
+      "grad_norm": 7.595141646088576,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 4159
+    },
+    {
+      "epoch": 0.354916067146283,
+      "grad_norm": 6.781960464165493,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 4160
+    },
+    {
+      "epoch": 0.35731414868105515,
+      "grad_norm": 11.2314623423545,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 4161
+    },
+    {
+      "epoch": 0.3597122302158273,
+      "grad_norm": 8.083869321030415,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 4162
+    },
+    {
+      "epoch": 0.36211031175059955,
+      "grad_norm": 7.277486755782745,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 4163
+    },
+    {
+      "epoch": 0.3645083932853717,
+      "grad_norm": 7.105079719985905,
+      "learning_rate": 1e-06,
+      "loss": 0.2136,
+      "step": 4164
+    },
+    {
+      "epoch": 0.3669064748201439,
+      "grad_norm": 7.39632528959402,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 4165
+    },
+    {
+      "epoch": 0.36930455635491605,
+      "grad_norm": 8.291032007485414,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 4166
+    },
+    {
+      "epoch": 0.37170263788968827,
+      "grad_norm": 9.685424303672823,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 4167
+    },
+    {
+      "epoch": 0.37410071942446044,
+      "grad_norm": 8.68687193084643,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 4168
+    },
+    {
+      "epoch": 0.3764988009592326,
+      "grad_norm": 7.406463822037652,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 4169
+    },
+    {
+      "epoch": 0.37889688249400477,
+      "grad_norm": 8.720560366702491,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 4170
+    },
+    {
+      "epoch": 0.381294964028777,
+      "grad_norm": 8.249682316112626,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 4171
+    },
+    {
+      "epoch": 0.38369304556354916,
+      "grad_norm": 7.055951741023097,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 4172
+    },
+    {
+      "epoch": 0.38609112709832133,
+      "grad_norm": 7.950525163018366,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 4173
+    },
+    {
+      "epoch": 0.38848920863309355,
+      "grad_norm": 7.894713208157477,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 4174
+    },
+    {
+      "epoch": 0.3908872901678657,
+      "grad_norm": 8.825710313461082,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 4175
+    },
+    {
+      "epoch": 0.3932853717026379,
+      "grad_norm": 7.683454969861378,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 4176
+    },
+    {
+      "epoch": 0.39568345323741005,
+      "grad_norm": 10.869660325791283,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 4177
+    },
+    {
+      "epoch": 0.3980815347721823,
+      "grad_norm": 9.58216485992029,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 4178
+    },
+    {
+      "epoch": 0.40047961630695444,
+      "grad_norm": 8.397731226143069,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 4179
+    },
+    {
+      "epoch": 0.4028776978417266,
+      "grad_norm": 9.306925900616227,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 4180
+    },
+    {
+      "epoch": 0.4052757793764988,
+      "grad_norm": 8.631979521555108,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 4181
+    },
+    {
+      "epoch": 0.407673860911271,
+      "grad_norm": 8.04323005292332,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 4182
+    },
+    {
+      "epoch": 0.41007194244604317,
+      "grad_norm": 8.092722226856964,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 4183
+    },
+    {
+      "epoch": 0.41247002398081534,
+      "grad_norm": 9.367232586704029,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 4184
+    },
+    {
+      "epoch": 0.4148681055155875,
+      "grad_norm": 9.270470243954703,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 4185
+    },
+    {
+      "epoch": 0.4172661870503597,
+      "grad_norm": 9.163090531407894,
+      "learning_rate": 1e-06,
+      "loss": 0.2047,
+      "step": 4186
+    },
+    {
+      "epoch": 0.4196642685851319,
+      "grad_norm": 7.632815731377891,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4187
+    },
+    {
+      "epoch": 0.42206235011990406,
+      "grad_norm": 8.402613990204044,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 4188
+    },
+    {
+      "epoch": 0.4244604316546763,
+      "grad_norm": 7.292232392050231,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 4189
+    },
+    {
+      "epoch": 0.42685851318944845,
+      "grad_norm": 7.0850525691201645,
+      "learning_rate": 1e-06,
+      "loss": 0.183,
+      "step": 4190
+    },
+    {
+      "epoch": 0.4292565947242206,
+      "grad_norm": 8.515253545212447,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 4191
+    },
+    {
+      "epoch": 0.4316546762589928,
+      "grad_norm": 6.807283005168344,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 4192
+    },
+    {
+      "epoch": 0.434052757793765,
+      "grad_norm": 8.335319789766299,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 4193
+    },
+    {
+      "epoch": 0.4364508393285372,
+      "grad_norm": 9.00094010203796,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 4194
+    },
+    {
+      "epoch": 0.43884892086330934,
+      "grad_norm": 7.145184366639372,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 4195
+    },
+    {
+      "epoch": 0.4412470023980815,
+      "grad_norm": 7.52090371729049,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 4196
+    },
+    {
+      "epoch": 0.44364508393285373,
+      "grad_norm": 9.28750713913807,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 4197
+    },
+    {
+      "epoch": 0.4460431654676259,
+      "grad_norm": 7.567147339683438,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 4198
+    },
+    {
+      "epoch": 0.44844124700239807,
+      "grad_norm": 7.81001083769929,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 4199
+    },
+    {
+      "epoch": 0.45083932853717024,
+      "grad_norm": 8.362015304162929,
+      "learning_rate": 1e-06,
+      "loss": 0.2151,
+      "step": 4200
+    },
+    {
+      "epoch": 0.45323741007194246,
+      "grad_norm": 6.9567417143043295,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 4201
+    },
+    {
+      "epoch": 0.4556354916067146,
+      "grad_norm": 9.320656926095472,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 4202
+    },
+    {
+      "epoch": 0.4580335731414868,
+      "grad_norm": 8.26272002341036,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 4203
+    },
+    {
+      "epoch": 0.460431654676259,
+      "grad_norm": 9.785208803904718,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 4204
+    },
+    {
+      "epoch": 0.4628297362110312,
+      "grad_norm": 8.871133016775927,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 4205
+    },
+    {
+      "epoch": 0.46522781774580335,
+      "grad_norm": 6.96361143927504,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 4206
+    },
+    {
+      "epoch": 0.4676258992805755,
+      "grad_norm": 7.143527766210891,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 4207
+    },
+    {
+      "epoch": 0.47002398081534774,
+      "grad_norm": 6.997338115627772,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 4208
+    },
+    {
+      "epoch": 0.4724220623501199,
+      "grad_norm": 11.898788006780238,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 4209
+    },
+    {
+      "epoch": 0.4748201438848921,
+      "grad_norm": 9.020800126377198,
+      "learning_rate": 1e-06,
+      "loss": 0.2133,
+      "step": 4210
+    },
+    {
+      "epoch": 0.47721822541966424,
+      "grad_norm": 8.138277974867156,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 4211
+    },
+    {
+      "epoch": 0.47961630695443647,
+      "grad_norm": 9.61056584854218,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 4212
+    },
+    {
+      "epoch": 0.48201438848920863,
+      "grad_norm": 8.215919224016597,
+      "learning_rate": 1e-06,
+      "loss": 0.198,
+      "step": 4213
+    },
+    {
+      "epoch": 0.4844124700239808,
+      "grad_norm": 9.376957731621,
+      "learning_rate": 1e-06,
+      "loss": 0.2145,
+      "step": 4214
+    },
+    {
+      "epoch": 0.486810551558753,
+      "grad_norm": 10.124792918135393,
+      "learning_rate": 1e-06,
+      "loss": 0.2092,
+      "step": 4215
+    },
+    {
+      "epoch": 0.4892086330935252,
+      "grad_norm": 8.343933179219507,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 4216
+    },
+    {
+      "epoch": 0.49160671462829736,
+      "grad_norm": 9.66396902122693,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 4217
+    },
+    {
+      "epoch": 0.4940047961630695,
+      "grad_norm": 8.60186976579433,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 4218
+    },
+    {
+      "epoch": 0.49640287769784175,
+      "grad_norm": 7.073025449931336,
+      "learning_rate": 1e-06,
+      "loss": 0.1952,
+      "step": 4219
+    },
+    {
+      "epoch": 0.4988009592326139,
+      "grad_norm": 6.688644857739274,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 4220
+    },
+    {
+      "epoch": 0.5011990407673861,
+      "grad_norm": 10.459434297868043,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 4221
+    },
+    {
+      "epoch": 0.5035971223021583,
+      "grad_norm": 7.298715582659848,
+      "learning_rate": 1e-06,
+      "loss": 0.22,
+      "step": 4222
+    },
+    {
+      "epoch": 0.5059952038369304,
+      "grad_norm": 7.523513691742642,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 4223
+    },
+    {
+      "epoch": 0.5083932853717026,
+      "grad_norm": 8.929078859710833,
+      "learning_rate": 1e-06,
+      "loss": 0.2129,
+      "step": 4224
+    },
+    {
+      "epoch": 0.5107913669064749,
+      "grad_norm": 6.871709665713609,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 4225
+    },
+    {
+      "epoch": 0.513189448441247,
+      "grad_norm": 10.417895411690786,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 4226
+    },
+    {
+      "epoch": 0.5155875299760192,
+      "grad_norm": 9.28389438609543,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 4227
+    },
+    {
+      "epoch": 0.5179856115107914,
+      "grad_norm": 5.587105820885126,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 4228
+    },
+    {
+      "epoch": 0.5203836930455635,
+      "grad_norm": 7.391194175898859,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 4229
+    },
+    {
+      "epoch": 0.5227817745803357,
+      "grad_norm": 10.19987764546362,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 4230
+    },
+    {
+      "epoch": 0.5251798561151079,
+      "grad_norm": 9.385258711559665,
+      "learning_rate": 1e-06,
+      "loss": 0.2227,
+      "step": 4231
+    },
+    {
+      "epoch": 0.5275779376498801,
+      "grad_norm": 7.771442559899836,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 4232
+    },
+    {
+      "epoch": 0.5299760191846523,
+      "grad_norm": 8.515704787196618,
+      "learning_rate": 1e-06,
+      "loss": 0.2086,
+      "step": 4233
+    },
+    {
+      "epoch": 0.5323741007194245,
+      "grad_norm": 7.766291701711594,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 4234
+    },
+    {
+      "epoch": 0.5347721822541966,
+      "grad_norm": 7.61364849159865,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 4235
+    },
+    {
+      "epoch": 0.5371702637889688,
+      "grad_norm": 7.480578203278269,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 4236
+    },
+    {
+      "epoch": 0.539568345323741,
+      "grad_norm": 8.571233406453382,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 4237
+    },
+    {
+      "epoch": 0.5419664268585132,
+      "grad_norm": 6.996700678437078,
+      "learning_rate": 1e-06,
+      "loss": 0.2175,
+      "step": 4238
+    },
+    {
+      "epoch": 0.5443645083932853,
+      "grad_norm": 9.12943326698859,
+      "learning_rate": 1e-06,
+      "loss": 0.2145,
+      "step": 4239
+    },
+    {
+      "epoch": 0.5467625899280576,
+      "grad_norm": 7.073135531018459,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 4240
+    },
+    {
+      "epoch": 0.5491606714628298,
+      "grad_norm": 6.723273300533814,
+      "learning_rate": 1e-06,
+      "loss": 0.2103,
+      "step": 4241
+    },
+    {
+      "epoch": 0.5515587529976019,
+      "grad_norm": 6.682228431580702,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 4242
+    },
+    {
+      "epoch": 0.5539568345323741,
+      "grad_norm": 9.38398938816736,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 4243
+    },
+    {
+      "epoch": 0.5563549160671463,
+      "grad_norm": 7.241961542548498,
+      "learning_rate": 1e-06,
+      "loss": 0.1785,
+      "step": 4244
+    },
+    {
+      "epoch": 0.5587529976019184,
+      "grad_norm": 8.484098536612128,
+      "learning_rate": 1e-06,
+      "loss": 0.2134,
+      "step": 4245
+    },
+    {
+      "epoch": 0.5611510791366906,
+      "grad_norm": 8.783823901263478,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 4246
+    },
+    {
+      "epoch": 0.5635491606714629,
+      "grad_norm": 10.82760727381654,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 4247
+    },
+    {
+      "epoch": 0.565947242206235,
+      "grad_norm": 7.174533974305626,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 4248
+    },
+    {
+      "epoch": 0.5683453237410072,
+      "grad_norm": 7.231430207854266,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 4249
+    },
+    {
+      "epoch": 0.5707434052757794,
+      "grad_norm": 9.767689341652027,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4250
+    },
+    {
+      "epoch": 0.5731414868105515,
+      "grad_norm": 8.981756936119742,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 4251
+    },
+    {
+      "epoch": 0.5755395683453237,
+      "grad_norm": 7.555266430591786,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 4252
+    },
+    {
+      "epoch": 0.5779376498800959,
+      "grad_norm": 9.604741783273383,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 4253
+    },
+    {
+      "epoch": 0.580335731414868,
+      "grad_norm": 9.801928663600942,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 4254
+    },
+    {
+      "epoch": 0.5827338129496403,
+      "grad_norm": 8.859308206068437,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 4255
+    },
+    {
+      "epoch": 0.5851318944844125,
+      "grad_norm": 8.043623589153029,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 4256
+    },
+    {
+      "epoch": 0.5875299760191847,
+      "grad_norm": 8.578888482686473,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 4257
+    },
+    {
+      "epoch": 0.5899280575539568,
+      "grad_norm": 5.2469275315522985,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 4258
+    },
+    {
+      "epoch": 0.592326139088729,
+      "grad_norm": 8.942885696241438,
+      "learning_rate": 1e-06,
+      "loss": 0.2043,
+      "step": 4259
+    },
+    {
+      "epoch": 0.5947242206235012,
+      "grad_norm": 7.024862147533621,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 4260
+    },
+    {
+      "epoch": 0.5971223021582733,
+      "grad_norm": 8.306444924851306,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 4261
+    },
+    {
+      "epoch": 0.5995203836930456,
+      "grad_norm": 7.625717567012919,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 4262
+    },
+    {
+      "epoch": 0.6019184652278178,
+      "grad_norm": 6.572657827438896,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 4263
+    },
+    {
+      "epoch": 0.60431654676259,
+      "grad_norm": 8.98662579886661,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 4264
+    },
+    {
+      "epoch": 0.6067146282973621,
+      "grad_norm": 8.485498819583677,
+      "learning_rate": 1e-06,
+      "loss": 0.215,
+      "step": 4265
+    },
+    {
+      "epoch": 0.6091127098321343,
+      "grad_norm": 5.999090184783211,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 4266
+    },
+    {
+      "epoch": 0.6115107913669064,
+      "grad_norm": 10.708206736666883,
+      "learning_rate": 1e-06,
+      "loss": 0.2077,
+      "step": 4267
+    },
+    {
+      "epoch": 0.6139088729016786,
+      "grad_norm": 7.799981079865822,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 4268
+    },
+    {
+      "epoch": 0.6163069544364509,
+      "grad_norm": 9.896217096129233,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 4269
+    },
+    {
+      "epoch": 0.6187050359712231,
+      "grad_norm": 7.564765645426475,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 4270
+    },
+    {
+      "epoch": 0.6211031175059952,
+      "grad_norm": 8.968216320463872,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 4271
+    },
+    {
+      "epoch": 0.6235011990407674,
+      "grad_norm": 9.001051337752143,
+      "learning_rate": 1e-06,
+      "loss": 0.1923,
+      "step": 4272
+    },
+    {
+      "epoch": 0.6258992805755396,
+      "grad_norm": 10.007762715259476,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 4273
+    },
+    {
+      "epoch": 0.6282973621103117,
+      "grad_norm": 6.543591460835502,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 4274
+    },
+    {
+      "epoch": 0.6306954436450839,
+      "grad_norm": 8.534884877481852,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 4275
+    },
+    {
+      "epoch": 0.6330935251798561,
+      "grad_norm": 7.4631621889757795,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 4276
+    },
+    {
+      "epoch": 0.6354916067146283,
+      "grad_norm": 8.67241103686598,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 4277
+    },
+    {
+      "epoch": 0.6378896882494005,
+      "grad_norm": 7.541007999028764,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 4278
+    },
+    {
+      "epoch": 0.6402877697841727,
+      "grad_norm": 8.98047108573565,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 4279
+    },
+    {
+      "epoch": 0.6426858513189448,
+      "grad_norm": 9.906755634855822,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 4280
+    },
+    {
+      "epoch": 0.645083932853717,
+      "grad_norm": 9.181327269603617,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 4281
+    },
+    {
+      "epoch": 0.6474820143884892,
+      "grad_norm": 11.461582656655882,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 4282
+    },
+    {
+      "epoch": 0.6498800959232613,
+      "grad_norm": 7.481191077572718,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 4283
+    },
+    {
+      "epoch": 0.6522781774580336,
+      "grad_norm": 8.246796819332902,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 4284
+    },
+    {
+      "epoch": 0.6546762589928058,
+      "grad_norm": 9.077253152593155,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 4285
+    },
+    {
+      "epoch": 0.657074340527578,
+      "grad_norm": 7.459780425984989,
+      "learning_rate": 1e-06,
+      "loss": 0.2079,
+      "step": 4286
+    },
+    {
+      "epoch": 0.6594724220623501,
+      "grad_norm": 10.48458770205836,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 4287
+    },
+    {
+      "epoch": 0.6618705035971223,
+      "grad_norm": 6.5801616989211285,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 4288
+    },
+    {
+      "epoch": 0.6642685851318945,
+      "grad_norm": 7.802820599274112,
+      "learning_rate": 1e-06,
+      "loss": 0.1962,
+      "step": 4289
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 9.039936560087483,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 4290
+    },
+    {
+      "epoch": 0.6690647482014388,
+      "grad_norm": 8.119269525236799,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 4291
+    },
+    {
+      "epoch": 0.6714628297362111,
+      "grad_norm": 8.119269525236799,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 4292
+    },
+    {
+      "epoch": 0.6738609112709832,
+      "grad_norm": 9.498606701200869,
+      "learning_rate": 1e-06,
+      "loss": 0.2221,
+      "step": 4293
+    },
+    {
+      "epoch": 0.6762589928057554,
+      "grad_norm": 5.812290520708289,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 4294
+    },
+    {
+      "epoch": 0.6786570743405276,
+      "grad_norm": 9.264845009258735,
+      "learning_rate": 1e-06,
+      "loss": 0.2152,
+      "step": 4295
+    },
+    {
+      "epoch": 0.6810551558752997,
+      "grad_norm": 8.696924365985883,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 4296
+    },
+    {
+      "epoch": 0.6834532374100719,
+      "grad_norm": 7.616902591927943,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 4297
+    },
+    {
+      "epoch": 0.6858513189448441,
+      "grad_norm": 7.398256617911574,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 4298
+    },
+    {
+      "epoch": 0.6882494004796164,
+      "grad_norm": 6.446240167182421,
+      "learning_rate": 1e-06,
+      "loss": 0.2072,
+      "step": 4299
+    },
+    {
+      "epoch": 0.6906474820143885,
+      "grad_norm": 7.513211625506093,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 4300
+    },
+    {
+      "epoch": 0.6930455635491607,
+      "grad_norm": 6.798665436807346,
+      "learning_rate": 1e-06,
+      "loss": 0.1895,
+      "step": 4301
+    },
+    {
+      "epoch": 0.6954436450839329,
+      "grad_norm": 7.967610062744614,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 4302
+    },
+    {
+      "epoch": 0.697841726618705,
+      "grad_norm": 7.681039567615027,
+      "learning_rate": 1e-06,
+      "loss": 0.1992,
+      "step": 4303
+    },
+    {
+      "epoch": 0.7002398081534772,
+      "grad_norm": 7.782333871153293,
+      "learning_rate": 1e-06,
+      "loss": 0.1957,
+      "step": 4304
+    },
+    {
+      "epoch": 0.7026378896882494,
+      "grad_norm": 6.6049491769556035,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 4305
+    },
+    {
+      "epoch": 0.7050359712230215,
+      "grad_norm": 8.4282528685993,
+      "learning_rate": 1e-06,
+      "loss": 0.2018,
+      "step": 4306
+    },
+    {
+      "epoch": 0.7074340527577938,
+      "grad_norm": 8.4511395745242,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 4307
+    },
+    {
+      "epoch": 0.709832134292566,
+      "grad_norm": 8.244547389624069,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 4308
+    },
+    {
+      "epoch": 0.7122302158273381,
+      "grad_norm": 9.467278199107476,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 4309
+    },
+    {
+      "epoch": 0.7146282973621103,
+      "grad_norm": 8.513524742965268,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 4310
+    },
+    {
+      "epoch": 0.7170263788968825,
+      "grad_norm": 6.648270688390476,
+      "learning_rate": 1e-06,
+      "loss": 0.2077,
+      "step": 4311
+    },
+    {
+      "epoch": 0.7194244604316546,
+      "grad_norm": 8.424540575268944,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 4312
+    },
+    {
+      "epoch": 0.7218225419664268,
+      "grad_norm": 7.464581186735013,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 4313
+    },
+    {
+      "epoch": 0.7242206235011991,
+      "grad_norm": 7.79818328981586,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 4314
+    },
+    {
+      "epoch": 0.7266187050359713,
+      "grad_norm": 9.096990761093853,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 4315
+    },
+    {
+      "epoch": 0.7290167865707434,
+      "grad_norm": 9.110699371996022,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 4316
+    },
+    {
+      "epoch": 0.7314148681055156,
+      "grad_norm": 7.3148279187785015,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 4317
+    },
+    {
+      "epoch": 0.7338129496402878,
+      "grad_norm": 7.8822923988811135,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 4318
+    },
+    {
+      "epoch": 0.7362110311750599,
+      "grad_norm": 9.242827983242417,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 4319
+    },
+    {
+      "epoch": 0.7386091127098321,
+      "grad_norm": 8.289584712173477,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 4320
+    },
+    {
+      "epoch": 0.7410071942446043,
+      "grad_norm": 5.8435665136930375,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 4321
+    },
+    {
+      "epoch": 0.7434052757793765,
+      "grad_norm": 6.93849271368728,
+      "learning_rate": 1e-06,
+      "loss": 0.2042,
+      "step": 4322
+    },
+    {
+      "epoch": 0.7458033573141487,
+      "grad_norm": 8.074411523876924,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 4323
+    },
+    {
+      "epoch": 0.7482014388489209,
+      "grad_norm": 7.656186430226837,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 4324
+    },
+    {
+      "epoch": 0.750599520383693,
+      "grad_norm": 8.356030625013835,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 4325
+    },
+    {
+      "epoch": 0.7529976019184652,
+      "grad_norm": 7.865012452261279,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 4326
+    },
+    {
+      "epoch": 0.7553956834532374,
+      "grad_norm": 6.911302419954089,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 4327
+    },
+    {
+      "epoch": 0.7577937649880095,
+      "grad_norm": 8.589965293446353,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 4328
+    },
+    {
+      "epoch": 0.7601918465227818,
+      "grad_norm": 9.735254344889809,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 4329
+    },
+    {
+      "epoch": 0.762589928057554,
+      "grad_norm": 8.661747965468695,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 4330
+    },
+    {
+      "epoch": 0.7649880095923262,
+      "grad_norm": 7.993652880967289,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 4331
+    },
+    {
+      "epoch": 0.7673860911270983,
+      "grad_norm": 9.076569061067401,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 4332
+    },
+    {
+      "epoch": 0.7697841726618705,
+      "grad_norm": 6.476571795190289,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 4333
+    },
+    {
+      "epoch": 0.7721822541966427,
+      "grad_norm": 8.684893087040699,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4334
+    },
+    {
+      "epoch": 0.7745803357314148,
+      "grad_norm": 5.941472323144792,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 4335
+    },
+    {
+      "epoch": 0.7769784172661871,
+      "grad_norm": 9.12551721367951,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 4336
+    },
+    {
+      "epoch": 0.7793764988009593,
+      "grad_norm": 8.01217578527844,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 4337
+    },
+    {
+      "epoch": 0.7817745803357314,
+      "grad_norm": 8.592093776661962,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 4338
+    },
+    {
+      "epoch": 0.7841726618705036,
+      "grad_norm": 10.399073130634834,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 4339
+    },
+    {
+      "epoch": 0.7865707434052758,
+      "grad_norm": 8.811268534235868,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 4340
+    },
+    {
+      "epoch": 0.7889688249400479,
+      "grad_norm": 7.472605929177846,
+      "learning_rate": 1e-06,
+      "loss": 0.1924,
+      "step": 4341
+    },
+    {
+      "epoch": 0.7913669064748201,
+      "grad_norm": 8.308697528604604,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 4342
+    },
+    {
+      "epoch": 0.7937649880095923,
+      "grad_norm": 8.279290258488079,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 4343
+    },
+    {
+      "epoch": 0.7961630695443646,
+      "grad_norm": 8.125479279192154,
+      "learning_rate": 1e-06,
+      "loss": 0.2078,
+      "step": 4344
+    },
+    {
+      "epoch": 0.7985611510791367,
+      "grad_norm": 7.61477222321954,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 4345
+    },
+    {
+      "epoch": 0.8009592326139089,
+      "grad_norm": 8.91917397338739,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 4346
+    },
+    {
+      "epoch": 0.8033573141486811,
+      "grad_norm": 6.830281911825465,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 4347
+    },
+    {
+      "epoch": 0.8057553956834532,
+      "grad_norm": 8.875957232657894,
+      "learning_rate": 1e-06,
+      "loss": 0.1922,
+      "step": 4348
+    },
+    {
+      "epoch": 0.8081534772182254,
+      "grad_norm": 8.815803200561797,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 4349
+    },
+    {
+      "epoch": 0.8105515587529976,
+      "grad_norm": 9.301069473369205,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 4350
+    },
+    {
+      "epoch": 0.8129496402877698,
+      "grad_norm": 8.986855090931654,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 4351
+    },
+    {
+      "epoch": 0.815347721822542,
+      "grad_norm": 7.320371937993917,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 4352
+    },
+    {
+      "epoch": 0.8177458033573142,
+      "grad_norm": 5.842071744698994,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 4353
+    },
+    {
+      "epoch": 0.8201438848920863,
+      "grad_norm": 6.376311297524768,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 4354
+    },
+    {
+      "epoch": 0.8225419664268585,
+      "grad_norm": 5.695010778109702,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 4355
+    },
+    {
+      "epoch": 0.8249400479616307,
+      "grad_norm": 10.432766109492924,
+      "learning_rate": 1e-06,
+      "loss": 0.206,
+      "step": 4356
+    },
+    {
+      "epoch": 0.8273381294964028,
+      "grad_norm": 8.990017439290309,
+      "learning_rate": 1e-06,
+      "loss": 0.1822,
+      "step": 4357
+    },
+    {
+      "epoch": 0.829736211031175,
+      "grad_norm": 10.98576683172512,
+      "learning_rate": 1e-06,
+      "loss": 0.2144,
+      "step": 4358
+    },
+    {
+      "epoch": 0.8321342925659473,
+      "grad_norm": 6.220707766993032,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 4359
+    },
+    {
+      "epoch": 0.8345323741007195,
+      "grad_norm": 9.515276119880392,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 4360
+    },
+    {
+      "epoch": 0.8369304556354916,
+      "grad_norm": 9.694554498771984,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 4361
+    },
+    {
+      "epoch": 0.8393285371702638,
+      "grad_norm": 6.831290976283748,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 4362
+    },
+    {
+      "epoch": 0.841726618705036,
+      "grad_norm": 8.883441834159987,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 4363
+    },
+    {
+      "epoch": 0.8441247002398081,
+      "grad_norm": 7.991514052428756,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 4364
+    },
+    {
+      "epoch": 0.8465227817745803,
+      "grad_norm": 7.418720451876266,
+      "learning_rate": 1e-06,
+      "loss": 0.2027,
+      "step": 4365
+    },
+    {
+      "epoch": 0.8489208633093526,
+      "grad_norm": 8.20603027998942,
+      "learning_rate": 1e-06,
+      "loss": 0.1922,
+      "step": 4366
+    },
+    {
+      "epoch": 0.8513189448441247,
+      "grad_norm": 6.089065710902285,
+      "learning_rate": 1e-06,
+      "loss": 0.2069,
+      "step": 4367
+    },
+    {
+      "epoch": 0.8537170263788969,
+      "grad_norm": 8.463928750587138,
+      "learning_rate": 1e-06,
+      "loss": 0.2072,
+      "step": 4368
+    },
+    {
+      "epoch": 0.8561151079136691,
+      "grad_norm": 6.834500874036416,
+      "learning_rate": 1e-06,
+      "loss": 0.1947,
+      "step": 4369
+    },
+    {
+      "epoch": 0.8585131894484412,
+      "grad_norm": 6.911421686248241,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 4370
+    },
+    {
+      "epoch": 0.8609112709832134,
+      "grad_norm": 9.630298948184546,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 4371
+    },
+    {
+      "epoch": 0.8633093525179856,
+      "grad_norm": 7.233056673243568,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 4372
+    },
+    {
+      "epoch": 0.8657074340527577,
+      "grad_norm": 7.604492560589703,
+      "learning_rate": 1e-06,
+      "loss": 0.2146,
+      "step": 4373
+    },
+    {
+      "epoch": 0.86810551558753,
+      "grad_norm": 9.042127574983308,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 4374
+    },
+    {
+      "epoch": 0.8705035971223022,
+      "grad_norm": 9.100987193058234,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 4375
+    },
+    {
+      "epoch": 0.8729016786570744,
+      "grad_norm": 8.487348279472359,
+      "learning_rate": 1e-06,
+      "loss": 0.2147,
+      "step": 4376
+    },
+    {
+      "epoch": 0.8752997601918465,
+      "grad_norm": 7.533571505181183,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 4377
+    },
+    {
+      "epoch": 0.8776978417266187,
+      "grad_norm": 7.316377649722969,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 4378
+    },
+    {
+      "epoch": 0.8800959232613909,
+      "grad_norm": 6.519238708839852,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 4379
+    },
+    {
+      "epoch": 0.882494004796163,
+      "grad_norm": 6.612927240301263,
+      "learning_rate": 1e-06,
+      "loss": 0.1979,
+      "step": 4380
+    },
+    {
+      "epoch": 0.8848920863309353,
+      "grad_norm": 7.222500521983273,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 4381
+    },
+    {
+      "epoch": 0.8872901678657075,
+      "grad_norm": 7.862888088222771,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 4382
+    },
+    {
+      "epoch": 0.8896882494004796,
+      "grad_norm": 10.126560465461784,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 4383
+    },
+    {
+      "epoch": 0.8920863309352518,
+      "grad_norm": 7.088729066287488,
+      "learning_rate": 1e-06,
+      "loss": 0.2058,
+      "step": 4384
+    },
+    {
+      "epoch": 0.894484412470024,
+      "grad_norm": 7.159149289505676,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 4385
+    },
+    {
+      "epoch": 0.8968824940047961,
+      "grad_norm": 8.632879565699097,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 4386
+    },
+    {
+      "epoch": 0.8992805755395683,
+      "grad_norm": 10.879742548700285,
+      "learning_rate": 1e-06,
+      "loss": 0.2126,
+      "step": 4387
+    },
+    {
+      "epoch": 0.9016786570743405,
+      "grad_norm": 7.959310154712489,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 4388
+    },
+    {
+      "epoch": 0.9040767386091128,
+      "grad_norm": 8.118428950698663,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 4389
+    },
+    {
+      "epoch": 0.9064748201438849,
+      "grad_norm": 7.994454292097547,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 4390
+    },
+    {
+      "epoch": 0.9088729016786571,
+      "grad_norm": 8.871621649390327,
+      "learning_rate": 1e-06,
+      "loss": 0.2006,
+      "step": 4391
+    },
+    {
+      "epoch": 0.9112709832134293,
+      "grad_norm": 8.493548490558787,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 4392
+    },
+    {
+      "epoch": 0.9136690647482014,
+      "grad_norm": 11.666963274806093,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 4393
+    },
+    {
+      "epoch": 0.9160671462829736,
+      "grad_norm": 7.426374355457724,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 4394
+    },
+    {
+      "epoch": 0.9184652278177458,
+      "grad_norm": 7.787537167790965,
+      "learning_rate": 1e-06,
+      "loss": 0.1987,
+      "step": 4395
+    },
+    {
+      "epoch": 0.920863309352518,
+      "grad_norm": 8.185235744282915,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 4396
+    },
+    {
+      "epoch": 0.9232613908872902,
+      "grad_norm": 8.552520705809075,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 4397
+    },
+    {
+      "epoch": 0.9256594724220624,
+      "grad_norm": 6.71597096042761,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 4398
+    },
+    {
+      "epoch": 0.9280575539568345,
+      "grad_norm": 5.810315855747072,
+      "learning_rate": 1e-06,
+      "loss": 0.1736,
+      "step": 4399
+    },
+    {
+      "epoch": 0.9304556354916067,
+      "grad_norm": 6.269877853423797,
+      "learning_rate": 1e-06,
+      "loss": 0.2106,
+      "step": 4400
+    },
+    {
+      "epoch": 0.9328537170263789,
+      "grad_norm": 8.164266271363905,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 4401
+    },
+    {
+      "epoch": 0.935251798561151,
+      "grad_norm": 10.756932486201695,
+      "learning_rate": 1e-06,
+      "loss": 0.2118,
+      "step": 4402
+    },
+    {
+      "epoch": 0.9376498800959233,
+      "grad_norm": 7.970268647278667,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 4403
+    },
+    {
+      "epoch": 0.9400479616306955,
+      "grad_norm": 9.387207834221417,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 4404
+    },
+    {
+      "epoch": 0.9424460431654677,
+      "grad_norm": 6.027316793632892,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 4405
+    },
+    {
+      "epoch": 0.9448441247002398,
+      "grad_norm": 7.810113152345897,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 4406
+    },
+    {
+      "epoch": 0.947242206235012,
+      "grad_norm": 9.277038867985194,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 4407
+    },
+    {
+      "epoch": 0.9496402877697842,
+      "grad_norm": 7.152017893554699,
+      "learning_rate": 1e-06,
+      "loss": 0.1908,
+      "step": 4408
+    },
+    {
+      "epoch": 0.9520383693045563,
+      "grad_norm": 8.942355850031086,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 4409
+    },
+    {
+      "epoch": 0.9544364508393285,
+      "grad_norm": 8.488328251842136,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 4410
+    },
+    {
+      "epoch": 0.9568345323741008,
+      "grad_norm": 6.583245827572811,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 4411
+    },
+    {
+      "epoch": 0.9592326139088729,
+      "grad_norm": 10.538654895845138,
+      "learning_rate": 1e-06,
+      "loss": 0.2123,
+      "step": 4412
+    },
+    {
+      "epoch": 0.9616306954436451,
+      "grad_norm": 8.06297255071957,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 4413
+    },
+    {
+      "epoch": 0.9640287769784173,
+      "grad_norm": 6.819959913316409,
+      "learning_rate": 1e-06,
+      "loss": 0.2045,
+      "step": 4414
+    },
+    {
+      "epoch": 0.9664268585131894,
+      "grad_norm": 7.498632160840945,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 4415
+    },
+    {
+      "epoch": 0.9688249400479616,
+      "grad_norm": 7.780243440901565,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 4416
+    },
+    {
+      "epoch": 0.9712230215827338,
+      "grad_norm": 8.34617225437308,
+      "learning_rate": 1e-06,
+      "loss": 0.1881,
+      "step": 4417
+    },
+    {
+      "epoch": 0.973621103117506,
+      "grad_norm": 8.697013848605545,
+      "learning_rate": 1e-06,
+      "loss": 0.2124,
+      "step": 4418
+    },
+    {
+      "epoch": 0.9760191846522782,
+      "grad_norm": 7.394495443894971,
+      "learning_rate": 1e-06,
+      "loss": 0.1923,
+      "step": 4419
+    },
+    {
+      "epoch": 0.9784172661870504,
+      "grad_norm": 8.154305177330313,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 4420
+    },
+    {
+      "epoch": 0.9808153477218226,
+      "grad_norm": 6.665833636856415,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 4421
+    },
+    {
+      "epoch": 0.9832134292565947,
+      "grad_norm": 7.846206827312938,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 4422
+    },
+    {
+      "epoch": 0.9856115107913669,
+      "grad_norm": 7.703251168624964,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 4423
+    },
+    {
+      "epoch": 0.988009592326139,
+      "grad_norm": 6.995166082209988,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 4424
+    },
+    {
+      "epoch": 0.9904076738609112,
+      "grad_norm": 6.779356644680722,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 4425
+    },
+    {
+      "epoch": 0.9928057553956835,
+      "grad_norm": 7.8575172904635044,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 4426
+    },
+    {
+      "epoch": 0.9952038369304557,
+      "grad_norm": 5.765049583521772,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 4427
+    },
+    {
+      "epoch": 0.9976019184652278,
+      "grad_norm": 8.705912586380013,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 4428
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 9.203983689370464,
+      "learning_rate": 1e-06,
+      "loss": 0.2024,
+      "step": 4429
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.0759408473968506,
+      "eval_runtime": 21.7635,
+      "eval_samples_per_second": 45.948,
+      "eval_steps_per_second": 5.744,
+      "step": 4429
+    },
+    {
+      "epoch": 0.002288329519450801,
+      "grad_norm": 9.049964750752936,
+      "learning_rate": 1e-06,
+      "loss": 0.4193,
+      "step": 4430
+    },
+    {
+      "epoch": 0.004576659038901602,
+      "grad_norm": 9.948729978005568,
+      "learning_rate": 1e-06,
+      "loss": 0.347,
+      "step": 4431
+    },
+    {
+      "epoch": 0.006864988558352402,
+      "grad_norm": 12.00870363637527,
+      "learning_rate": 1e-06,
+      "loss": 0.3828,
+      "step": 4432
+    },
+    {
+      "epoch": 0.009153318077803204,
+      "grad_norm": 8.717720088088285,
+      "learning_rate": 1e-06,
+      "loss": 0.3444,
+      "step": 4433
+    },
+    {
+      "epoch": 0.011441647597254004,
+      "grad_norm": 9.211510072555514,
+      "learning_rate": 1e-06,
+      "loss": 0.3822,
+      "step": 4434
+    },
+    {
+      "epoch": 0.013729977116704805,
+      "grad_norm": 10.005465196472558,
+      "learning_rate": 1e-06,
+      "loss": 0.3744,
+      "step": 4435
+    },
+    {
+      "epoch": 0.016018306636155607,
+      "grad_norm": 10.191078321329375,
+      "learning_rate": 1e-06,
+      "loss": 0.3729,
+      "step": 4436
+    },
+    {
+      "epoch": 0.018306636155606407,
+      "grad_norm": 7.496301281617085,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 4437
+    },
+    {
+      "epoch": 0.020594965675057208,
+      "grad_norm": 7.690219245226319,
+      "learning_rate": 1e-06,
+      "loss": 0.3626,
+      "step": 4438
+    },
+    {
+      "epoch": 0.02288329519450801,
+      "grad_norm": 8.252291523777766,
+      "learning_rate": 1e-06,
+      "loss": 0.3564,
+      "step": 4439
+    },
+    {
+      "epoch": 0.02517162471395881,
+      "grad_norm": 6.025401102115667,
+      "learning_rate": 1e-06,
+      "loss": 0.3616,
+      "step": 4440
+    },
+    {
+      "epoch": 0.02745995423340961,
+      "grad_norm": 8.097733834001266,
+      "learning_rate": 1e-06,
+      "loss": 0.3957,
+      "step": 4441
+    },
+    {
+      "epoch": 0.029748283752860413,
+      "grad_norm": 8.203025300265889,
+      "learning_rate": 1e-06,
+      "loss": 0.3789,
+      "step": 4442
+    },
+    {
+      "epoch": 0.032036613272311214,
+      "grad_norm": 7.4511145567970445,
+      "learning_rate": 1e-06,
+      "loss": 0.3412,
+      "step": 4443
+    },
+    {
+      "epoch": 0.034324942791762014,
+      "grad_norm": 5.255529292009054,
+      "learning_rate": 1e-06,
+      "loss": 0.3645,
+      "step": 4444
+    },
+    {
+      "epoch": 0.036613272311212815,
+      "grad_norm": 8.837372080728285,
+      "learning_rate": 1e-06,
+      "loss": 0.3408,
+      "step": 4445
+    },
+    {
+      "epoch": 0.038901601830663615,
+      "grad_norm": 8.03342509728929,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 4446
+    },
+    {
+      "epoch": 0.041189931350114416,
+      "grad_norm": 7.884583399838572,
+      "learning_rate": 1e-06,
+      "loss": 0.3429,
+      "step": 4447
+    },
+    {
+      "epoch": 0.043478260869565216,
+      "grad_norm": 6.6839723185108975,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 4448
+    },
+    {
+      "epoch": 0.04576659038901602,
+      "grad_norm": 7.260854061838674,
+      "learning_rate": 1e-06,
+      "loss": 0.3451,
+      "step": 4449
+    },
+    {
+      "epoch": 0.04805491990846682,
+      "grad_norm": 7.623074825170965,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 4450
+    },
+    {
+      "epoch": 0.05034324942791762,
+      "grad_norm": 8.215498520919843,
+      "learning_rate": 1e-06,
+      "loss": 0.3646,
+      "step": 4451
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 6.266302543189828,
+      "learning_rate": 1e-06,
+      "loss": 0.3578,
+      "step": 4452
+    },
+    {
+      "epoch": 0.05491990846681922,
+      "grad_norm": 6.270729623442539,
+      "learning_rate": 1e-06,
+      "loss": 0.3231,
+      "step": 4453
+    },
+    {
+      "epoch": 0.057208237986270026,
+      "grad_norm": 6.207553742224537,
+      "learning_rate": 1e-06,
+      "loss": 0.3498,
+      "step": 4454
+    },
+    {
+      "epoch": 0.059496567505720827,
+      "grad_norm": 7.505972889605595,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 4455
+    },
+    {
+      "epoch": 0.06178489702517163,
+      "grad_norm": 7.186151891353009,
+      "learning_rate": 1e-06,
+      "loss": 0.3507,
+      "step": 4456
+    },
+    {
+      "epoch": 0.06407322654462243,
+      "grad_norm": 7.216164950296839,
+      "learning_rate": 1e-06,
+      "loss": 0.3529,
+      "step": 4457
+    },
+    {
+      "epoch": 0.06636155606407322,
+      "grad_norm": 6.766041362549465,
+      "learning_rate": 1e-06,
+      "loss": 0.3481,
+      "step": 4458
+    },
+    {
+      "epoch": 0.06864988558352403,
+      "grad_norm": 8.246306701700222,
+      "learning_rate": 1e-06,
+      "loss": 0.352,
+      "step": 4459
+    },
+    {
+      "epoch": 0.07093821510297482,
+      "grad_norm": 7.2615958529516815,
+      "learning_rate": 1e-06,
+      "loss": 0.3509,
+      "step": 4460
+    },
+    {
+      "epoch": 0.07322654462242563,
+      "grad_norm": 7.667994915863042,
+      "learning_rate": 1e-06,
+      "loss": 0.3497,
+      "step": 4461
+    },
+    {
+      "epoch": 0.07551487414187644,
+      "grad_norm": 8.019504725656947,
+      "learning_rate": 1e-06,
+      "loss": 0.3489,
+      "step": 4462
+    },
+    {
+      "epoch": 0.07780320366132723,
+      "grad_norm": 9.484799616979256,
+      "learning_rate": 1e-06,
+      "loss": 0.3599,
+      "step": 4463
+    },
+    {
+      "epoch": 0.08009153318077804,
+      "grad_norm": 7.987723925876434,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 4464
+    },
+    {
+      "epoch": 0.08237986270022883,
+      "grad_norm": 7.556425032032225,
+      "learning_rate": 1e-06,
+      "loss": 0.3695,
+      "step": 4465
+    },
+    {
+      "epoch": 0.08466819221967964,
+      "grad_norm": 7.092417486830778,
+      "learning_rate": 1e-06,
+      "loss": 0.3363,
+      "step": 4466
+    },
+    {
+      "epoch": 0.08695652173913043,
+      "grad_norm": 8.154138038196132,
+      "learning_rate": 1e-06,
+      "loss": 0.3694,
+      "step": 4467
+    },
+    {
+      "epoch": 0.08924485125858124,
+      "grad_norm": 7.490890937127408,
+      "learning_rate": 1e-06,
+      "loss": 0.3605,
+      "step": 4468
+    },
+    {
+      "epoch": 0.09153318077803203,
+      "grad_norm": 7.47441519210925,
+      "learning_rate": 1e-06,
+      "loss": 0.3883,
+      "step": 4469
+    },
+    {
+      "epoch": 0.09382151029748284,
+      "grad_norm": 7.154526742095325,
+      "learning_rate": 1e-06,
+      "loss": 0.3426,
+      "step": 4470
+    },
+    {
+      "epoch": 0.09610983981693363,
+      "grad_norm": 7.3431116885176975,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 4471
+    },
+    {
+      "epoch": 0.09839816933638444,
+      "grad_norm": 6.575277236214657,
+      "learning_rate": 1e-06,
+      "loss": 0.3296,
+      "step": 4472
+    },
+    {
+      "epoch": 0.10068649885583524,
+      "grad_norm": 7.005057629021649,
+      "learning_rate": 1e-06,
+      "loss": 0.339,
+      "step": 4473
+    },
+    {
+      "epoch": 0.10297482837528604,
+      "grad_norm": 5.502947151590897,
+      "learning_rate": 1e-06,
+      "loss": 0.3485,
+      "step": 4474
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 7.069717015218924,
+      "learning_rate": 1e-06,
+      "loss": 0.3452,
+      "step": 4475
+    },
+    {
+      "epoch": 0.10755148741418764,
+      "grad_norm": 6.456007208560856,
+      "learning_rate": 1e-06,
+      "loss": 0.3502,
+      "step": 4476
+    },
+    {
+      "epoch": 0.10983981693363844,
+      "grad_norm": 6.896663599499755,
+      "learning_rate": 1e-06,
+      "loss": 0.3384,
+      "step": 4477
+    },
+    {
+      "epoch": 0.11212814645308924,
+      "grad_norm": 4.722933609826658,
+      "learning_rate": 1e-06,
+      "loss": 0.3491,
+      "step": 4478
+    },
+    {
+      "epoch": 0.11441647597254005,
+      "grad_norm": 7.026159769867784,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 4479
+    },
+    {
+      "epoch": 0.11670480549199085,
+      "grad_norm": 8.283506128025058,
+      "learning_rate": 1e-06,
+      "loss": 0.3748,
+      "step": 4480
+    },
+    {
+      "epoch": 0.11899313501144165,
+      "grad_norm": 5.612626918860273,
+      "learning_rate": 1e-06,
+      "loss": 0.3304,
+      "step": 4481
+    },
+    {
+      "epoch": 0.12128146453089245,
+      "grad_norm": 8.55363625509232,
+      "learning_rate": 1e-06,
+      "loss": 0.3588,
+      "step": 4482
+    },
+    {
+      "epoch": 0.12356979405034325,
+      "grad_norm": 6.458288987266286,
+      "learning_rate": 1e-06,
+      "loss": 0.3541,
+      "step": 4483
+    },
+    {
+      "epoch": 0.12585812356979406,
+      "grad_norm": 8.108718948269072,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 4484
+    },
+    {
+      "epoch": 0.12814645308924486,
+      "grad_norm": 7.260098317122242,
+      "learning_rate": 1e-06,
+      "loss": 0.3336,
+      "step": 4485
+    },
+    {
+      "epoch": 0.13043478260869565,
+      "grad_norm": 6.231243900003414,
+      "learning_rate": 1e-06,
+      "loss": 0.3532,
+      "step": 4486
+    },
+    {
+      "epoch": 0.13272311212814644,
+      "grad_norm": 7.700106143026918,
+      "learning_rate": 1e-06,
+      "loss": 0.3517,
+      "step": 4487
+    },
+    {
+      "epoch": 0.13501144164759726,
+      "grad_norm": 7.5043422202102015,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 4488
+    },
+    {
+      "epoch": 0.13729977116704806,
+      "grad_norm": 7.4851731721718915,
+      "learning_rate": 1e-06,
+      "loss": 0.3326,
+      "step": 4489
+    },
+    {
+      "epoch": 0.13958810068649885,
+      "grad_norm": 7.302847195319356,
+      "learning_rate": 1e-06,
+      "loss": 0.3275,
+      "step": 4490
+    },
+    {
+      "epoch": 0.14187643020594964,
+      "grad_norm": 8.899443428083089,
+      "learning_rate": 1e-06,
+      "loss": 0.3507,
+      "step": 4491
+    },
+    {
+      "epoch": 0.14416475972540047,
+      "grad_norm": 7.798416284811466,
+      "learning_rate": 1e-06,
+      "loss": 0.3659,
+      "step": 4492
+    },
+    {
+      "epoch": 0.14645308924485126,
+      "grad_norm": 7.524140483331105,
+      "learning_rate": 1e-06,
+      "loss": 0.3512,
+      "step": 4493
+    },
+    {
+      "epoch": 0.14874141876430205,
+      "grad_norm": 8.00614004327026,
+      "learning_rate": 1e-06,
+      "loss": 0.3703,
+      "step": 4494
+    },
+    {
+      "epoch": 0.15102974828375287,
+      "grad_norm": 8.07733432726576,
+      "learning_rate": 1e-06,
+      "loss": 0.3613,
+      "step": 4495
+    },
+    {
+      "epoch": 0.15331807780320367,
+      "grad_norm": 6.901679230050025,
+      "learning_rate": 1e-06,
+      "loss": 0.3412,
+      "step": 4496
+    },
+    {
+      "epoch": 0.15560640732265446,
+      "grad_norm": 8.218710122497846,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 4497
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 6.103428515232263,
+      "learning_rate": 1e-06,
+      "loss": 0.3371,
+      "step": 4498
+    },
+    {
+      "epoch": 0.16018306636155608,
+      "grad_norm": 5.158590028116917,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 4499
+    },
+    {
+      "epoch": 0.16247139588100687,
+      "grad_norm": 9.638256054716539,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 4500
+    },
+    {
+      "epoch": 0.16475972540045766,
+      "grad_norm": 8.487617108798561,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 4501
+    },
+    {
+      "epoch": 0.16704805491990846,
+      "grad_norm": 8.267273149231888,
+      "learning_rate": 1e-06,
+      "loss": 0.3735,
+      "step": 4502
+    },
+    {
+      "epoch": 0.16933638443935928,
+      "grad_norm": 7.8354997179363215,
+      "learning_rate": 1e-06,
+      "loss": 0.3624,
+      "step": 4503
+    },
+    {
+      "epoch": 0.17162471395881007,
+      "grad_norm": 5.860495010222886,
+      "learning_rate": 1e-06,
+      "loss": 0.3604,
+      "step": 4504
+    },
+    {
+      "epoch": 0.17391304347826086,
+      "grad_norm": 7.90790373522701,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 4505
+    },
+    {
+      "epoch": 0.17620137299771166,
+      "grad_norm": 6.843133769810857,
+      "learning_rate": 1e-06,
+      "loss": 0.3439,
+      "step": 4506
+    },
+    {
+      "epoch": 0.17848970251716248,
+      "grad_norm": 6.665502297098539,
+      "learning_rate": 1e-06,
+      "loss": 0.3768,
+      "step": 4507
+    },
+    {
+      "epoch": 0.18077803203661327,
+      "grad_norm": 7.266069671707825,
+      "learning_rate": 1e-06,
+      "loss": 0.3355,
+      "step": 4508
+    },
+    {
+      "epoch": 0.18306636155606407,
+      "grad_norm": 6.748055635117422,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 4509
+    },
+    {
+      "epoch": 0.1853546910755149,
+      "grad_norm": 7.740435826083541,
+      "learning_rate": 1e-06,
+      "loss": 0.3585,
+      "step": 4510
+    },
+    {
+      "epoch": 0.18764302059496568,
+      "grad_norm": 6.423565772054938,
+      "learning_rate": 1e-06,
+      "loss": 0.3219,
+      "step": 4511
+    },
+    {
+      "epoch": 0.18993135011441648,
+      "grad_norm": 7.907479696717127,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 4512
+    },
+    {
+      "epoch": 0.19221967963386727,
+      "grad_norm": 7.269816130118411,
+      "learning_rate": 1e-06,
+      "loss": 0.3402,
+      "step": 4513
+    },
+    {
+      "epoch": 0.1945080091533181,
+      "grad_norm": 6.080388348462885,
+      "learning_rate": 1e-06,
+      "loss": 0.3593,
+      "step": 4514
+    },
+    {
+      "epoch": 0.19679633867276888,
+      "grad_norm": 8.955090672204433,
+      "learning_rate": 1e-06,
+      "loss": 0.3839,
+      "step": 4515
+    },
+    {
+      "epoch": 0.19908466819221968,
+      "grad_norm": 6.262518507849864,
+      "learning_rate": 1e-06,
+      "loss": 0.359,
+      "step": 4516
+    },
+    {
+      "epoch": 0.20137299771167047,
+      "grad_norm": 5.729133239523498,
+      "learning_rate": 1e-06,
+      "loss": 0.3742,
+      "step": 4517
+    },
+    {
+      "epoch": 0.2036613272311213,
+      "grad_norm": 6.394539018242784,
+      "learning_rate": 1e-06,
+      "loss": 0.3444,
+      "step": 4518
+    },
+    {
+      "epoch": 0.20594965675057209,
+      "grad_norm": 5.4151099733967385,
+      "learning_rate": 1e-06,
+      "loss": 0.3639,
+      "step": 4519
+    },
+    {
+      "epoch": 0.20823798627002288,
+      "grad_norm": 8.687953739055251,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 4520
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 7.58345920805256,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 4521
+    },
+    {
+      "epoch": 0.2128146453089245,
+      "grad_norm": 7.815009277520351,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 4522
+    },
+    {
+      "epoch": 0.2151029748283753,
+      "grad_norm": 6.634317067426688,
+      "learning_rate": 1e-06,
+      "loss": 0.3332,
+      "step": 4523
+    },
+    {
+      "epoch": 0.21739130434782608,
+      "grad_norm": 7.40469635919002,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 4524
+    },
+    {
+      "epoch": 0.21967963386727687,
+      "grad_norm": 6.485205633465764,
+      "learning_rate": 1e-06,
+      "loss": 0.3686,
+      "step": 4525
+    },
+    {
+      "epoch": 0.2219679633867277,
+      "grad_norm": 9.272892209266425,
+      "learning_rate": 1e-06,
+      "loss": 0.3723,
+      "step": 4526
+    },
+    {
+      "epoch": 0.2242562929061785,
+      "grad_norm": 6.302423885076749,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 4527
+    },
+    {
+      "epoch": 0.22654462242562928,
+      "grad_norm": 6.564191149537946,
+      "learning_rate": 1e-06,
+      "loss": 0.3372,
+      "step": 4528
+    },
+    {
+      "epoch": 0.2288329519450801,
+      "grad_norm": 7.800447882231838,
+      "learning_rate": 1e-06,
+      "loss": 0.3474,
+      "step": 4529
+    },
+    {
+      "epoch": 0.2311212814645309,
+      "grad_norm": 6.424683209331323,
+      "learning_rate": 1e-06,
+      "loss": 0.3662,
+      "step": 4530
+    },
+    {
+      "epoch": 0.2334096109839817,
+      "grad_norm": 7.744774721475848,
+      "learning_rate": 1e-06,
+      "loss": 0.3561,
+      "step": 4531
+    },
+    {
+      "epoch": 0.23569794050343248,
+      "grad_norm": 7.9506758010328005,
+      "learning_rate": 1e-06,
+      "loss": 0.3505,
+      "step": 4532
+    },
+    {
+      "epoch": 0.2379862700228833,
+      "grad_norm": 6.170735454261454,
+      "learning_rate": 1e-06,
+      "loss": 0.3259,
+      "step": 4533
+    },
+    {
+      "epoch": 0.2402745995423341,
+      "grad_norm": 6.272330791899602,
+      "learning_rate": 1e-06,
+      "loss": 0.3333,
+      "step": 4534
+    },
+    {
+      "epoch": 0.2425629290617849,
+      "grad_norm": 8.286497249250177,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 4535
+    },
+    {
+      "epoch": 0.2448512585812357,
+      "grad_norm": 7.1570576178621215,
+      "learning_rate": 1e-06,
+      "loss": 0.3632,
+      "step": 4536
+    },
+    {
+      "epoch": 0.2471395881006865,
+      "grad_norm": 7.791907895797384,
+      "learning_rate": 1e-06,
+      "loss": 0.3642,
+      "step": 4537
+    },
+    {
+      "epoch": 0.2494279176201373,
+      "grad_norm": 8.100598742198711,
+      "learning_rate": 1e-06,
+      "loss": 0.3324,
+      "step": 4538
+    },
+    {
+      "epoch": 0.2517162471395881,
+      "grad_norm": 8.457848430456108,
+      "learning_rate": 1e-06,
+      "loss": 0.3647,
+      "step": 4539
+    },
+    {
+      "epoch": 0.2540045766590389,
+      "grad_norm": 8.30233924193465,
+      "learning_rate": 1e-06,
+      "loss": 0.3714,
+      "step": 4540
+    },
+    {
+      "epoch": 0.2562929061784897,
+      "grad_norm": 6.908882222242276,
+      "learning_rate": 1e-06,
+      "loss": 0.3552,
+      "step": 4541
+    },
+    {
+      "epoch": 0.2585812356979405,
+      "grad_norm": 8.838540246466764,
+      "learning_rate": 1e-06,
+      "loss": 0.3654,
+      "step": 4542
+    },
+    {
+      "epoch": 0.2608695652173913,
+      "grad_norm": 7.596625589169397,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4543
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 6.39291810404921,
+      "learning_rate": 1e-06,
+      "loss": 0.3508,
+      "step": 4544
+    },
+    {
+      "epoch": 0.2654462242562929,
+      "grad_norm": 7.041458172421818,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 4545
+    },
+    {
+      "epoch": 0.26773455377574373,
+      "grad_norm": 6.804513107268906,
+      "learning_rate": 1e-06,
+      "loss": 0.3621,
+      "step": 4546
+    },
+    {
+      "epoch": 0.2700228832951945,
+      "grad_norm": 7.293262972289611,
+      "learning_rate": 1e-06,
+      "loss": 0.3356,
+      "step": 4547
+    },
+    {
+      "epoch": 0.2723112128146453,
+      "grad_norm": 6.902684279059504,
+      "learning_rate": 1e-06,
+      "loss": 0.3771,
+      "step": 4548
+    },
+    {
+      "epoch": 0.2745995423340961,
+      "grad_norm": 8.076831341440414,
+      "learning_rate": 1e-06,
+      "loss": 0.3456,
+      "step": 4549
+    },
+    {
+      "epoch": 0.2768878718535469,
+      "grad_norm": 7.054121356179121,
+      "learning_rate": 1e-06,
+      "loss": 0.3244,
+      "step": 4550
+    },
+    {
+      "epoch": 0.2791762013729977,
+      "grad_norm": 6.4495718305607,
+      "learning_rate": 1e-06,
+      "loss": 0.3607,
+      "step": 4551
+    },
+    {
+      "epoch": 0.2814645308924485,
+      "grad_norm": 7.224647786167758,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 4552
+    },
+    {
+      "epoch": 0.2837528604118993,
+      "grad_norm": 7.716500898183419,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 4553
+    },
+    {
+      "epoch": 0.28604118993135014,
+      "grad_norm": 7.766367524337637,
+      "learning_rate": 1e-06,
+      "loss": 0.3587,
+      "step": 4554
+    },
+    {
+      "epoch": 0.28832951945080093,
+      "grad_norm": 5.838837733639465,
+      "learning_rate": 1e-06,
+      "loss": 0.3752,
+      "step": 4555
+    },
+    {
+      "epoch": 0.2906178489702517,
+      "grad_norm": 4.422744551732463,
+      "learning_rate": 1e-06,
+      "loss": 0.3598,
+      "step": 4556
+    },
+    {
+      "epoch": 0.2929061784897025,
+      "grad_norm": 8.234107073752545,
+      "learning_rate": 1e-06,
+      "loss": 0.3523,
+      "step": 4557
+    },
+    {
+      "epoch": 0.2951945080091533,
+      "grad_norm": 7.160580432554881,
+      "learning_rate": 1e-06,
+      "loss": 0.3213,
+      "step": 4558
+    },
+    {
+      "epoch": 0.2974828375286041,
+      "grad_norm": 6.317322683298812,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 4559
+    },
+    {
+      "epoch": 0.2997711670480549,
+      "grad_norm": 7.905909920814693,
+      "learning_rate": 1e-06,
+      "loss": 0.3465,
+      "step": 4560
+    },
+    {
+      "epoch": 0.30205949656750575,
+      "grad_norm": 9.178733695880993,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 4561
+    },
+    {
+      "epoch": 0.30434782608695654,
+      "grad_norm": 7.699073052234349,
+      "learning_rate": 1e-06,
+      "loss": 0.3361,
+      "step": 4562
+    },
+    {
+      "epoch": 0.30663615560640733,
+      "grad_norm": 8.619616250580595,
+      "learning_rate": 1e-06,
+      "loss": 0.3661,
+      "step": 4563
+    },
+    {
+      "epoch": 0.30892448512585813,
+      "grad_norm": 5.837065343510285,
+      "learning_rate": 1e-06,
+      "loss": 0.3407,
+      "step": 4564
+    },
+    {
+      "epoch": 0.3112128146453089,
+      "grad_norm": 7.1784931530952045,
+      "learning_rate": 1e-06,
+      "loss": 0.3865,
+      "step": 4565
+    },
+    {
+      "epoch": 0.3135011441647597,
+      "grad_norm": 8.022907503370867,
+      "learning_rate": 1e-06,
+      "loss": 0.3527,
+      "step": 4566
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 7.360054138581774,
+      "learning_rate": 1e-06,
+      "loss": 0.368,
+      "step": 4567
+    },
+    {
+      "epoch": 0.3180778032036613,
+      "grad_norm": 6.797459795389073,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 4568
+    },
+    {
+      "epoch": 0.32036613272311215,
+      "grad_norm": 8.243501554326613,
+      "learning_rate": 1e-06,
+      "loss": 0.3716,
+      "step": 4569
+    },
+    {
+      "epoch": 0.32265446224256294,
+      "grad_norm": 5.973997097965646,
+      "learning_rate": 1e-06,
+      "loss": 0.3725,
+      "step": 4570
+    },
+    {
+      "epoch": 0.32494279176201374,
+      "grad_norm": 7.6014978297726685,
+      "learning_rate": 1e-06,
+      "loss": 0.3359,
+      "step": 4571
+    },
+    {
+      "epoch": 0.32723112128146453,
+      "grad_norm": 7.466273394627606,
+      "learning_rate": 1e-06,
+      "loss": 0.3605,
+      "step": 4572
+    },
+    {
+      "epoch": 0.3295194508009153,
+      "grad_norm": 8.607056666036563,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 4573
+    },
+    {
+      "epoch": 0.3318077803203661,
+      "grad_norm": 8.399432682052863,
+      "learning_rate": 1e-06,
+      "loss": 0.3528,
+      "step": 4574
+    },
+    {
+      "epoch": 0.3340961098398169,
+      "grad_norm": 8.941929841319459,
+      "learning_rate": 1e-06,
+      "loss": 0.3351,
+      "step": 4575
+    },
+    {
+      "epoch": 0.33638443935926776,
+      "grad_norm": 6.180423717056848,
+      "learning_rate": 1e-06,
+      "loss": 0.3446,
+      "step": 4576
+    },
+    {
+      "epoch": 0.33867276887871856,
+      "grad_norm": 7.702101786918552,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 4577
+    },
+    {
+      "epoch": 0.34096109839816935,
+      "grad_norm": 5.922733995972156,
+      "learning_rate": 1e-06,
+      "loss": 0.3469,
+      "step": 4578
+    },
+    {
+      "epoch": 0.34324942791762014,
+      "grad_norm": 6.789062149008197,
+      "learning_rate": 1e-06,
+      "loss": 0.3349,
+      "step": 4579
+    },
+    {
+      "epoch": 0.34553775743707094,
+      "grad_norm": 6.602156726688394,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 4580
+    },
+    {
+      "epoch": 0.34782608695652173,
+      "grad_norm": 6.503221957996044,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 4581
+    },
+    {
+      "epoch": 0.3501144164759725,
+      "grad_norm": 7.682885858053606,
+      "learning_rate": 1e-06,
+      "loss": 0.3436,
+      "step": 4582
+    },
+    {
+      "epoch": 0.3524027459954233,
+      "grad_norm": 7.710417939198685,
+      "learning_rate": 1e-06,
+      "loss": 0.3389,
+      "step": 4583
+    },
+    {
+      "epoch": 0.35469107551487417,
+      "grad_norm": 6.427819619128609,
+      "learning_rate": 1e-06,
+      "loss": 0.4004,
+      "step": 4584
+    },
+    {
+      "epoch": 0.35697940503432496,
+      "grad_norm": 7.427351260755735,
+      "learning_rate": 1e-06,
+      "loss": 0.3692,
+      "step": 4585
+    },
+    {
+      "epoch": 0.35926773455377575,
+      "grad_norm": 8.199358170808777,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4586
+    },
+    {
+      "epoch": 0.36155606407322655,
+      "grad_norm": 8.075688993765395,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 4587
+    },
+    {
+      "epoch": 0.36384439359267734,
+      "grad_norm": 6.550001879269343,
+      "learning_rate": 1e-06,
+      "loss": 0.3611,
+      "step": 4588
+    },
+    {
+      "epoch": 0.36613272311212813,
+      "grad_norm": 7.455714964278161,
+      "learning_rate": 1e-06,
+      "loss": 0.3587,
+      "step": 4589
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 8.391099095374173,
+      "learning_rate": 1e-06,
+      "loss": 0.3621,
+      "step": 4590
+    },
+    {
+      "epoch": 0.3707093821510298,
+      "grad_norm": 6.053817650796254,
+      "learning_rate": 1e-06,
+      "loss": 0.3967,
+      "step": 4591
+    },
+    {
+      "epoch": 0.37299771167048057,
+      "grad_norm": 6.724122560991037,
+      "learning_rate": 1e-06,
+      "loss": 0.3657,
+      "step": 4592
+    },
+    {
+      "epoch": 0.37528604118993136,
+      "grad_norm": 8.22427510421274,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 4593
+    },
+    {
+      "epoch": 0.37757437070938216,
+      "grad_norm": 6.9542139144142805,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 4594
+    },
+    {
+      "epoch": 0.37986270022883295,
+      "grad_norm": 7.558326372310931,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 4595
+    },
+    {
+      "epoch": 0.38215102974828374,
+      "grad_norm": 4.557244306585046,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 4596
+    },
+    {
+      "epoch": 0.38443935926773454,
+      "grad_norm": 7.062431555058038,
+      "learning_rate": 1e-06,
+      "loss": 0.3439,
+      "step": 4597
+    },
+    {
+      "epoch": 0.38672768878718533,
+      "grad_norm": 5.154633835897386,
+      "learning_rate": 1e-06,
+      "loss": 0.3413,
+      "step": 4598
+    },
+    {
+      "epoch": 0.3890160183066362,
+      "grad_norm": 6.288897335452252,
+      "learning_rate": 1e-06,
+      "loss": 0.3428,
+      "step": 4599
+    },
+    {
+      "epoch": 0.391304347826087,
+      "grad_norm": 6.78577962630092,
+      "learning_rate": 1e-06,
+      "loss": 0.3483,
+      "step": 4600
+    },
+    {
+      "epoch": 0.39359267734553777,
+      "grad_norm": 6.83281609094651,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 4601
+    },
+    {
+      "epoch": 0.39588100686498856,
+      "grad_norm": 6.376314962674605,
+      "learning_rate": 1e-06,
+      "loss": 0.3667,
+      "step": 4602
+    },
+    {
+      "epoch": 0.39816933638443935,
+      "grad_norm": 7.028656744070296,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 4603
+    },
+    {
+      "epoch": 0.40045766590389015,
+      "grad_norm": 7.72186799951693,
+      "learning_rate": 1e-06,
+      "loss": 0.3532,
+      "step": 4604
+    },
+    {
+      "epoch": 0.40274599542334094,
+      "grad_norm": 6.3088433426127475,
+      "learning_rate": 1e-06,
+      "loss": 0.3463,
+      "step": 4605
+    },
+    {
+      "epoch": 0.40503432494279173,
+      "grad_norm": 8.774570155937836,
+      "learning_rate": 1e-06,
+      "loss": 0.3748,
+      "step": 4606
+    },
+    {
+      "epoch": 0.4073226544622426,
+      "grad_norm": 7.288888713191752,
+      "learning_rate": 1e-06,
+      "loss": 0.3441,
+      "step": 4607
+    },
+    {
+      "epoch": 0.4096109839816934,
+      "grad_norm": 6.426480258515243,
+      "learning_rate": 1e-06,
+      "loss": 0.3415,
+      "step": 4608
+    },
+    {
+      "epoch": 0.41189931350114417,
+      "grad_norm": 6.532755942446694,
+      "learning_rate": 1e-06,
+      "loss": 0.3186,
+      "step": 4609
+    },
+    {
+      "epoch": 0.41418764302059496,
+      "grad_norm": 6.884786099291856,
+      "learning_rate": 1e-06,
+      "loss": 0.3338,
+      "step": 4610
+    },
+    {
+      "epoch": 0.41647597254004576,
+      "grad_norm": 7.51634429439888,
+      "learning_rate": 1e-06,
+      "loss": 0.3418,
+      "step": 4611
+    },
+    {
+      "epoch": 0.41876430205949655,
+      "grad_norm": 9.240901695731756,
+      "learning_rate": 1e-06,
+      "loss": 0.3629,
+      "step": 4612
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 7.982754374124543,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 4613
+    },
+    {
+      "epoch": 0.4233409610983982,
+      "grad_norm": 7.455448393201876,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 4614
+    },
+    {
+      "epoch": 0.425629290617849,
+      "grad_norm": 8.323010530275292,
+      "learning_rate": 1e-06,
+      "loss": 0.3613,
+      "step": 4615
+    },
+    {
+      "epoch": 0.4279176201372998,
+      "grad_norm": 7.437931209151298,
+      "learning_rate": 1e-06,
+      "loss": 0.3673,
+      "step": 4616
+    },
+    {
+      "epoch": 0.4302059496567506,
+      "grad_norm": 6.723506923385575,
+      "learning_rate": 1e-06,
+      "loss": 0.336,
+      "step": 4617
+    },
+    {
+      "epoch": 0.43249427917620137,
+      "grad_norm": 5.727188541645272,
+      "learning_rate": 1e-06,
+      "loss": 0.3242,
+      "step": 4618
+    },
+    {
+      "epoch": 0.43478260869565216,
+      "grad_norm": 8.39332328291412,
+      "learning_rate": 1e-06,
+      "loss": 0.3646,
+      "step": 4619
+    },
+    {
+      "epoch": 0.43707093821510296,
+      "grad_norm": 7.088724099649811,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 4620
+    },
+    {
+      "epoch": 0.43935926773455375,
+      "grad_norm": 7.818576148578594,
+      "learning_rate": 1e-06,
+      "loss": 0.3422,
+      "step": 4621
+    },
+    {
+      "epoch": 0.4416475972540046,
+      "grad_norm": 7.260324582824068,
+      "learning_rate": 1e-06,
+      "loss": 0.3774,
+      "step": 4622
+    },
+    {
+      "epoch": 0.4439359267734554,
+      "grad_norm": 7.893005036921673,
+      "learning_rate": 1e-06,
+      "loss": 0.3386,
+      "step": 4623
+    },
+    {
+      "epoch": 0.4462242562929062,
+      "grad_norm": 7.429458202469956,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 4624
+    },
+    {
+      "epoch": 0.448512585812357,
+      "grad_norm": 8.852662281019418,
+      "learning_rate": 1e-06,
+      "loss": 0.3397,
+      "step": 4625
+    },
+    {
+      "epoch": 0.45080091533180777,
+      "grad_norm": 7.334027889468023,
+      "learning_rate": 1e-06,
+      "loss": 0.3364,
+      "step": 4626
+    },
+    {
+      "epoch": 0.45308924485125857,
+      "grad_norm": 11.434535506836742,
+      "learning_rate": 1e-06,
+      "loss": 0.3705,
+      "step": 4627
+    },
+    {
+      "epoch": 0.45537757437070936,
+      "grad_norm": 8.350069703719567,
+      "learning_rate": 1e-06,
+      "loss": 0.3564,
+      "step": 4628
+    },
+    {
+      "epoch": 0.4576659038901602,
+      "grad_norm": 7.075678657540942,
+      "learning_rate": 1e-06,
+      "loss": 0.3644,
+      "step": 4629
+    },
+    {
+      "epoch": 0.459954233409611,
+      "grad_norm": 8.46386993630732,
+      "learning_rate": 1e-06,
+      "loss": 0.3441,
+      "step": 4630
+    },
+    {
+      "epoch": 0.4622425629290618,
+      "grad_norm": 8.198311395371482,
+      "learning_rate": 1e-06,
+      "loss": 0.3765,
+      "step": 4631
+    },
+    {
+      "epoch": 0.4645308924485126,
+      "grad_norm": 6.423382840425872,
+      "learning_rate": 1e-06,
+      "loss": 0.3361,
+      "step": 4632
+    },
+    {
+      "epoch": 0.4668192219679634,
+      "grad_norm": 7.9466582103301855,
+      "learning_rate": 1e-06,
+      "loss": 0.3806,
+      "step": 4633
+    },
+    {
+      "epoch": 0.4691075514874142,
+      "grad_norm": 6.055535616732826,
+      "learning_rate": 1e-06,
+      "loss": 0.3697,
+      "step": 4634
+    },
+    {
+      "epoch": 0.47139588100686497,
+      "grad_norm": 6.807311105569473,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 4635
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 7.317962969698412,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 4636
+    },
+    {
+      "epoch": 0.4759725400457666,
+      "grad_norm": 6.201412322233737,
+      "learning_rate": 1e-06,
+      "loss": 0.3684,
+      "step": 4637
+    },
+    {
+      "epoch": 0.4782608695652174,
+      "grad_norm": 6.191396759278637,
+      "learning_rate": 1e-06,
+      "loss": 0.3755,
+      "step": 4638
+    },
+    {
+      "epoch": 0.4805491990846682,
+      "grad_norm": 7.527104060658067,
+      "learning_rate": 1e-06,
+      "loss": 0.3594,
+      "step": 4639
+    },
+    {
+      "epoch": 0.482837528604119,
+      "grad_norm": 5.037107823707413,
+      "learning_rate": 1e-06,
+      "loss": 0.35,
+      "step": 4640
+    },
+    {
+      "epoch": 0.4851258581235698,
+      "grad_norm": 8.265103801814261,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 4641
+    },
+    {
+      "epoch": 0.4874141876430206,
+      "grad_norm": 7.21430389814292,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 4642
+    },
+    {
+      "epoch": 0.4897025171624714,
+      "grad_norm": 5.3709545079627405,
+      "learning_rate": 1e-06,
+      "loss": 0.3506,
+      "step": 4643
+    },
+    {
+      "epoch": 0.4919908466819222,
+      "grad_norm": 5.447960327269153,
+      "learning_rate": 1e-06,
+      "loss": 0.3591,
+      "step": 4644
+    },
+    {
+      "epoch": 0.494279176201373,
+      "grad_norm": 5.709122531949933,
+      "learning_rate": 1e-06,
+      "loss": 0.3325,
+      "step": 4645
+    },
+    {
+      "epoch": 0.4965675057208238,
+      "grad_norm": 6.525413687701847,
+      "learning_rate": 1e-06,
+      "loss": 0.3547,
+      "step": 4646
+    },
+    {
+      "epoch": 0.4988558352402746,
+      "grad_norm": 6.5062993168779695,
+      "learning_rate": 1e-06,
+      "loss": 0.3375,
+      "step": 4647
+    },
+    {
+      "epoch": 0.5011441647597255,
+      "grad_norm": 5.53292750042982,
+      "learning_rate": 1e-06,
+      "loss": 0.3438,
+      "step": 4648
+    },
+    {
+      "epoch": 0.5034324942791762,
+      "grad_norm": 6.676271064780165,
+      "learning_rate": 1e-06,
+      "loss": 0.3473,
+      "step": 4649
+    },
+    {
+      "epoch": 0.505720823798627,
+      "grad_norm": 7.529310501552625,
+      "learning_rate": 1e-06,
+      "loss": 0.3896,
+      "step": 4650
+    },
+    {
+      "epoch": 0.5080091533180778,
+      "grad_norm": 6.2203152977127845,
+      "learning_rate": 1e-06,
+      "loss": 0.3697,
+      "step": 4651
+    },
+    {
+      "epoch": 0.5102974828375286,
+      "grad_norm": 7.876830605125825,
+      "learning_rate": 1e-06,
+      "loss": 0.3579,
+      "step": 4652
+    },
+    {
+      "epoch": 0.5125858123569794,
+      "grad_norm": 7.183938686081819,
+      "learning_rate": 1e-06,
+      "loss": 0.3362,
+      "step": 4653
+    },
+    {
+      "epoch": 0.5148741418764302,
+      "grad_norm": 7.828998233014079,
+      "learning_rate": 1e-06,
+      "loss": 0.3425,
+      "step": 4654
+    },
+    {
+      "epoch": 0.517162471395881,
+      "grad_norm": 6.527378246105927,
+      "learning_rate": 1e-06,
+      "loss": 0.3386,
+      "step": 4655
+    },
+    {
+      "epoch": 0.5194508009153318,
+      "grad_norm": 7.053982454677281,
+      "learning_rate": 1e-06,
+      "loss": 0.3542,
+      "step": 4656
+    },
+    {
+      "epoch": 0.5217391304347826,
+      "grad_norm": 8.26977615619967,
+      "learning_rate": 1e-06,
+      "loss": 0.3989,
+      "step": 4657
+    },
+    {
+      "epoch": 0.5240274599542334,
+      "grad_norm": 9.03891859688553,
+      "learning_rate": 1e-06,
+      "loss": 0.3395,
+      "step": 4658
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 8.975362045219434,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 4659
+    },
+    {
+      "epoch": 0.528604118993135,
+      "grad_norm": 7.126400647571779,
+      "learning_rate": 1e-06,
+      "loss": 0.3599,
+      "step": 4660
+    },
+    {
+      "epoch": 0.5308924485125858,
+      "grad_norm": 8.348146696093764,
+      "learning_rate": 1e-06,
+      "loss": 0.3329,
+      "step": 4661
+    },
+    {
+      "epoch": 0.5331807780320366,
+      "grad_norm": 6.932622320239702,
+      "learning_rate": 1e-06,
+      "loss": 0.3072,
+      "step": 4662
+    },
+    {
+      "epoch": 0.5354691075514875,
+      "grad_norm": 5.890776804779739,
+      "learning_rate": 1e-06,
+      "loss": 0.3622,
+      "step": 4663
+    },
+    {
+      "epoch": 0.5377574370709383,
+      "grad_norm": 6.2320098046243695,
+      "learning_rate": 1e-06,
+      "loss": 0.3467,
+      "step": 4664
+    },
+    {
+      "epoch": 0.540045766590389,
+      "grad_norm": 7.511943954777533,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4665
+    },
+    {
+      "epoch": 0.5423340961098398,
+      "grad_norm": 7.198299806548221,
+      "learning_rate": 1e-06,
+      "loss": 0.3491,
+      "step": 4666
+    },
+    {
+      "epoch": 0.5446224256292906,
+      "grad_norm": 8.14465561815829,
+      "learning_rate": 1e-06,
+      "loss": 0.3706,
+      "step": 4667
+    },
+    {
+      "epoch": 0.5469107551487414,
+      "grad_norm": 8.395351818214468,
+      "learning_rate": 1e-06,
+      "loss": 0.3407,
+      "step": 4668
+    },
+    {
+      "epoch": 0.5491990846681922,
+      "grad_norm": 8.070236843774133,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4669
+    },
+    {
+      "epoch": 0.551487414187643,
+      "grad_norm": 7.422556732200889,
+      "learning_rate": 1e-06,
+      "loss": 0.3481,
+      "step": 4670
+    },
+    {
+      "epoch": 0.5537757437070938,
+      "grad_norm": 6.345437790147031,
+      "learning_rate": 1e-06,
+      "loss": 0.3742,
+      "step": 4671
+    },
+    {
+      "epoch": 0.5560640732265446,
+      "grad_norm": 8.51916902229746,
+      "learning_rate": 1e-06,
+      "loss": 0.3443,
+      "step": 4672
+    },
+    {
+      "epoch": 0.5583524027459954,
+      "grad_norm": 8.04337839911422,
+      "learning_rate": 1e-06,
+      "loss": 0.3561,
+      "step": 4673
+    },
+    {
+      "epoch": 0.5606407322654462,
+      "grad_norm": 7.134584327549627,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 4674
+    },
+    {
+      "epoch": 0.562929061784897,
+      "grad_norm": 8.471362496204017,
+      "learning_rate": 1e-06,
+      "loss": 0.362,
+      "step": 4675
+    },
+    {
+      "epoch": 0.5652173913043478,
+      "grad_norm": 8.921348462736034,
+      "learning_rate": 1e-06,
+      "loss": 0.3457,
+      "step": 4676
+    },
+    {
+      "epoch": 0.5675057208237986,
+      "grad_norm": 7.73584410741282,
+      "learning_rate": 1e-06,
+      "loss": 0.3546,
+      "step": 4677
+    },
+    {
+      "epoch": 0.5697940503432495,
+      "grad_norm": 4.9563609370851776,
+      "learning_rate": 1e-06,
+      "loss": 0.3455,
+      "step": 4678
+    },
+    {
+      "epoch": 0.5720823798627003,
+      "grad_norm": 7.051902241568968,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 4679
+    },
+    {
+      "epoch": 0.5743707093821511,
+      "grad_norm": 6.8438039211022605,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 4680
+    },
+    {
+      "epoch": 0.5766590389016019,
+      "grad_norm": 8.791666022880717,
+      "learning_rate": 1e-06,
+      "loss": 0.3599,
+      "step": 4681
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 7.295654901925049,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 4682
+    },
+    {
+      "epoch": 0.5812356979405034,
+      "grad_norm": 6.214883356954196,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 4683
+    },
+    {
+      "epoch": 0.5835240274599542,
+      "grad_norm": 6.2218898302818,
+      "learning_rate": 1e-06,
+      "loss": 0.3062,
+      "step": 4684
+    },
+    {
+      "epoch": 0.585812356979405,
+      "grad_norm": 6.711659604086461,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 4685
+    },
+    {
+      "epoch": 0.5881006864988558,
+      "grad_norm": 8.31408031143103,
+      "learning_rate": 1e-06,
+      "loss": 0.3569,
+      "step": 4686
+    },
+    {
+      "epoch": 0.5903890160183066,
+      "grad_norm": 7.094281344849522,
+      "learning_rate": 1e-06,
+      "loss": 0.3465,
+      "step": 4687
+    },
+    {
+      "epoch": 0.5926773455377574,
+      "grad_norm": 6.749108344683659,
+      "learning_rate": 1e-06,
+      "loss": 0.3755,
+      "step": 4688
+    },
+    {
+      "epoch": 0.5949656750572082,
+      "grad_norm": 8.850896902655203,
+      "learning_rate": 1e-06,
+      "loss": 0.3252,
+      "step": 4689
+    },
+    {
+      "epoch": 0.597254004576659,
+      "grad_norm": 7.488862740154608,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 4690
+    },
+    {
+      "epoch": 0.5995423340961098,
+      "grad_norm": 9.222953767137586,
+      "learning_rate": 1e-06,
+      "loss": 0.3427,
+      "step": 4691
+    },
+    {
+      "epoch": 0.6018306636155606,
+      "grad_norm": 7.024469358388853,
+      "learning_rate": 1e-06,
+      "loss": 0.3188,
+      "step": 4692
+    },
+    {
+      "epoch": 0.6041189931350115,
+      "grad_norm": 7.035061033897658,
+      "learning_rate": 1e-06,
+      "loss": 0.3439,
+      "step": 4693
+    },
+    {
+      "epoch": 0.6064073226544623,
+      "grad_norm": 7.486679066807913,
+      "learning_rate": 1e-06,
+      "loss": 0.3461,
+      "step": 4694
+    },
+    {
+      "epoch": 0.6086956521739131,
+      "grad_norm": 7.212244322135983,
+      "learning_rate": 1e-06,
+      "loss": 0.3628,
+      "step": 4695
+    },
+    {
+      "epoch": 0.6109839816933639,
+      "grad_norm": 7.375733399319294,
+      "learning_rate": 1e-06,
+      "loss": 0.3319,
+      "step": 4696
+    },
+    {
+      "epoch": 0.6132723112128147,
+      "grad_norm": 8.39212551895627,
+      "learning_rate": 1e-06,
+      "loss": 0.3483,
+      "step": 4697
+    },
+    {
+      "epoch": 0.6155606407322655,
+      "grad_norm": 6.40119151425221,
+      "learning_rate": 1e-06,
+      "loss": 0.3604,
+      "step": 4698
+    },
+    {
+      "epoch": 0.6178489702517163,
+      "grad_norm": 7.464575743742351,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 4699
+    },
+    {
+      "epoch": 0.620137299771167,
+      "grad_norm": 7.390306462235757,
+      "learning_rate": 1e-06,
+      "loss": 0.3498,
+      "step": 4700
+    },
+    {
+      "epoch": 0.6224256292906178,
+      "grad_norm": 9.640370890469988,
+      "learning_rate": 1e-06,
+      "loss": 0.3468,
+      "step": 4701
+    },
+    {
+      "epoch": 0.6247139588100686,
+      "grad_norm": 6.810893784410284,
+      "learning_rate": 1e-06,
+      "loss": 0.3348,
+      "step": 4702
+    },
+    {
+      "epoch": 0.6270022883295194,
+      "grad_norm": 7.999516567182361,
+      "learning_rate": 1e-06,
+      "loss": 0.3552,
+      "step": 4703
+    },
+    {
+      "epoch": 0.6292906178489702,
+      "grad_norm": 6.8885707633180004,
+      "learning_rate": 1e-06,
+      "loss": 0.329,
+      "step": 4704
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 5.2839870635658075,
+      "learning_rate": 1e-06,
+      "loss": 0.3342,
+      "step": 4705
+    },
+    {
+      "epoch": 0.6338672768878718,
+      "grad_norm": 7.994294177628593,
+      "learning_rate": 1e-06,
+      "loss": 0.3392,
+      "step": 4706
+    },
+    {
+      "epoch": 0.6361556064073226,
+      "grad_norm": 7.3608307845343015,
+      "learning_rate": 1e-06,
+      "loss": 0.3633,
+      "step": 4707
+    },
+    {
+      "epoch": 0.6384439359267735,
+      "grad_norm": 8.201002653694545,
+      "learning_rate": 1e-06,
+      "loss": 0.3441,
+      "step": 4708
+    },
+    {
+      "epoch": 0.6407322654462243,
+      "grad_norm": 7.626215749188774,
+      "learning_rate": 1e-06,
+      "loss": 0.3661,
+      "step": 4709
+    },
+    {
+      "epoch": 0.6430205949656751,
+      "grad_norm": 5.671749865194866,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 4710
+    },
+    {
+      "epoch": 0.6453089244851259,
+      "grad_norm": 6.4129965094329835,
+      "learning_rate": 1e-06,
+      "loss": 0.3738,
+      "step": 4711
+    },
+    {
+      "epoch": 0.6475972540045767,
+      "grad_norm": 7.81754124654807,
+      "learning_rate": 1e-06,
+      "loss": 0.344,
+      "step": 4712
+    },
+    {
+      "epoch": 0.6498855835240275,
+      "grad_norm": 8.869501250537791,
+      "learning_rate": 1e-06,
+      "loss": 0.3675,
+      "step": 4713
+    },
+    {
+      "epoch": 0.6521739130434783,
+      "grad_norm": 6.185323034556774,
+      "learning_rate": 1e-06,
+      "loss": 0.3334,
+      "step": 4714
+    },
+    {
+      "epoch": 0.6544622425629291,
+      "grad_norm": 5.952804672911564,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 4715
+    },
+    {
+      "epoch": 0.6567505720823799,
+      "grad_norm": 5.908579894103454,
+      "learning_rate": 1e-06,
+      "loss": 0.365,
+      "step": 4716
+    },
+    {
+      "epoch": 0.6590389016018307,
+      "grad_norm": 7.891059135166329,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 4717
+    },
+    {
+      "epoch": 0.6613272311212814,
+      "grad_norm": 8.112898741770728,
+      "learning_rate": 1e-06,
+      "loss": 0.3601,
+      "step": 4718
+    },
+    {
+      "epoch": 0.6636155606407322,
+      "grad_norm": 7.698813351475166,
+      "learning_rate": 1e-06,
+      "loss": 0.409,
+      "step": 4719
+    },
+    {
+      "epoch": 0.665903890160183,
+      "grad_norm": 6.772364226515706,
+      "learning_rate": 1e-06,
+      "loss": 0.3347,
+      "step": 4720
+    },
+    {
+      "epoch": 0.6681922196796338,
+      "grad_norm": 6.7271846732095755,
+      "learning_rate": 1e-06,
+      "loss": 0.3465,
+      "step": 4721
+    },
+    {
+      "epoch": 0.6704805491990846,
+      "grad_norm": 6.847184764224222,
+      "learning_rate": 1e-06,
+      "loss": 0.3329,
+      "step": 4722
+    },
+    {
+      "epoch": 0.6727688787185355,
+      "grad_norm": 6.085375204683007,
+      "learning_rate": 1e-06,
+      "loss": 0.3442,
+      "step": 4723
+    },
+    {
+      "epoch": 0.6750572082379863,
+      "grad_norm": 5.914960224484479,
+      "learning_rate": 1e-06,
+      "loss": 0.3228,
+      "step": 4724
+    },
+    {
+      "epoch": 0.6773455377574371,
+      "grad_norm": 6.909675427392637,
+      "learning_rate": 1e-06,
+      "loss": 0.3423,
+      "step": 4725
+    },
+    {
+      "epoch": 0.6796338672768879,
+      "grad_norm": 7.192587574613113,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 4726
+    },
+    {
+      "epoch": 0.6819221967963387,
+      "grad_norm": 7.39776374129484,
+      "learning_rate": 1e-06,
+      "loss": 0.3712,
+      "step": 4727
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 5.937410937258123,
+      "learning_rate": 1e-06,
+      "loss": 0.3509,
+      "step": 4728
+    },
+    {
+      "epoch": 0.6864988558352403,
+      "grad_norm": 7.7602962288002155,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 4729
+    },
+    {
+      "epoch": 0.6887871853546911,
+      "grad_norm": 5.9033020739632525,
+      "learning_rate": 1e-06,
+      "loss": 0.3591,
+      "step": 4730
+    },
+    {
+      "epoch": 0.6910755148741419,
+      "grad_norm": 8.212904687658511,
+      "learning_rate": 1e-06,
+      "loss": 0.367,
+      "step": 4731
+    },
+    {
+      "epoch": 0.6933638443935927,
+      "grad_norm": 6.0764620113942485,
+      "learning_rate": 1e-06,
+      "loss": 0.3922,
+      "step": 4732
+    },
+    {
+      "epoch": 0.6956521739130435,
+      "grad_norm": 6.0450387820153635,
+      "learning_rate": 1e-06,
+      "loss": 0.3795,
+      "step": 4733
+    },
+    {
+      "epoch": 0.6979405034324943,
+      "grad_norm": 8.08425040156844,
+      "learning_rate": 1e-06,
+      "loss": 0.3686,
+      "step": 4734
+    },
+    {
+      "epoch": 0.700228832951945,
+      "grad_norm": 6.6369647348102125,
+      "learning_rate": 1e-06,
+      "loss": 0.348,
+      "step": 4735
+    },
+    {
+      "epoch": 0.7025171624713958,
+      "grad_norm": 7.65271561805521,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 4736
+    },
+    {
+      "epoch": 0.7048054919908466,
+      "grad_norm": 6.611192186476246,
+      "learning_rate": 1e-06,
+      "loss": 0.3562,
+      "step": 4737
+    },
+    {
+      "epoch": 0.7070938215102975,
+      "grad_norm": 4.976262904516351,
+      "learning_rate": 1e-06,
+      "loss": 0.365,
+      "step": 4738
+    },
+    {
+      "epoch": 0.7093821510297483,
+      "grad_norm": 7.000416937332017,
+      "learning_rate": 1e-06,
+      "loss": 0.3678,
+      "step": 4739
+    },
+    {
+      "epoch": 0.7116704805491991,
+      "grad_norm": 5.258452098130029,
+      "learning_rate": 1e-06,
+      "loss": 0.3496,
+      "step": 4740
+    },
+    {
+      "epoch": 0.7139588100686499,
+      "grad_norm": 4.527318046503956,
+      "learning_rate": 1e-06,
+      "loss": 0.3384,
+      "step": 4741
+    },
+    {
+      "epoch": 0.7162471395881007,
+      "grad_norm": 8.666369358249614,
+      "learning_rate": 1e-06,
+      "loss": 0.3923,
+      "step": 4742
+    },
+    {
+      "epoch": 0.7185354691075515,
+      "grad_norm": 7.887706064463628,
+      "learning_rate": 1e-06,
+      "loss": 0.3587,
+      "step": 4743
+    },
+    {
+      "epoch": 0.7208237986270023,
+      "grad_norm": 6.573076006002632,
+      "learning_rate": 1e-06,
+      "loss": 0.3424,
+      "step": 4744
+    },
+    {
+      "epoch": 0.7231121281464531,
+      "grad_norm": 7.517509295272846,
+      "learning_rate": 1e-06,
+      "loss": 0.3529,
+      "step": 4745
+    },
+    {
+      "epoch": 0.7254004576659039,
+      "grad_norm": 6.041923212393632,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 4746
+    },
+    {
+      "epoch": 0.7276887871853547,
+      "grad_norm": 6.998867399749206,
+      "learning_rate": 1e-06,
+      "loss": 0.3504,
+      "step": 4747
+    },
+    {
+      "epoch": 0.7299771167048055,
+      "grad_norm": 6.825658793311907,
+      "learning_rate": 1e-06,
+      "loss": 0.3948,
+      "step": 4748
+    },
+    {
+      "epoch": 0.7322654462242563,
+      "grad_norm": 4.95599323704399,
+      "learning_rate": 1e-06,
+      "loss": 0.3533,
+      "step": 4749
+    },
+    {
+      "epoch": 0.7345537757437071,
+      "grad_norm": 7.159996512810418,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 4750
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 5.498889503300652,
+      "learning_rate": 1e-06,
+      "loss": 0.3239,
+      "step": 4751
+    },
+    {
+      "epoch": 0.7391304347826086,
+      "grad_norm": 6.882640887360658,
+      "learning_rate": 1e-06,
+      "loss": 0.3767,
+      "step": 4752
+    },
+    {
+      "epoch": 0.7414187643020596,
+      "grad_norm": 6.24216355492567,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 4753
+    },
+    {
+      "epoch": 0.7437070938215103,
+      "grad_norm": 7.3657187099275525,
+      "learning_rate": 1e-06,
+      "loss": 0.3527,
+      "step": 4754
+    },
+    {
+      "epoch": 0.7459954233409611,
+      "grad_norm": 6.0944803512585235,
+      "learning_rate": 1e-06,
+      "loss": 0.3503,
+      "step": 4755
+    },
+    {
+      "epoch": 0.7482837528604119,
+      "grad_norm": 7.335140916337724,
+      "learning_rate": 1e-06,
+      "loss": 0.3965,
+      "step": 4756
+    },
+    {
+      "epoch": 0.7505720823798627,
+      "grad_norm": 6.686808363046258,
+      "learning_rate": 1e-06,
+      "loss": 0.3699,
+      "step": 4757
+    },
+    {
+      "epoch": 0.7528604118993135,
+      "grad_norm": 6.954321202172538,
+      "learning_rate": 1e-06,
+      "loss": 0.3546,
+      "step": 4758
+    },
+    {
+      "epoch": 0.7551487414187643,
+      "grad_norm": 8.233048849225506,
+      "learning_rate": 1e-06,
+      "loss": 0.3509,
+      "step": 4759
+    },
+    {
+      "epoch": 0.7574370709382151,
+      "grad_norm": 6.913344571874717,
+      "learning_rate": 1e-06,
+      "loss": 0.3341,
+      "step": 4760
+    },
+    {
+      "epoch": 0.7597254004576659,
+      "grad_norm": 5.897081743279878,
+      "learning_rate": 1e-06,
+      "loss": 0.3412,
+      "step": 4761
+    },
+    {
+      "epoch": 0.7620137299771167,
+      "grad_norm": 7.259495664675403,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 4762
+    },
+    {
+      "epoch": 0.7643020594965675,
+      "grad_norm": 6.607979741445736,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 4763
+    },
+    {
+      "epoch": 0.7665903890160183,
+      "grad_norm": 7.4266493216295615,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 4764
+    },
+    {
+      "epoch": 0.7688787185354691,
+      "grad_norm": 8.229199818698206,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 4765
+    },
+    {
+      "epoch": 0.7711670480549199,
+      "grad_norm": 7.1221402771148945,
+      "learning_rate": 1e-06,
+      "loss": 0.3431,
+      "step": 4766
+    },
+    {
+      "epoch": 0.7734553775743707,
+      "grad_norm": 5.758999251184279,
+      "learning_rate": 1e-06,
+      "loss": 0.3294,
+      "step": 4767
+    },
+    {
+      "epoch": 0.7757437070938215,
+      "grad_norm": 8.66017840429491,
+      "learning_rate": 1e-06,
+      "loss": 0.3628,
+      "step": 4768
+    },
+    {
+      "epoch": 0.7780320366132724,
+      "grad_norm": 8.759187602099852,
+      "learning_rate": 1e-06,
+      "loss": 0.3786,
+      "step": 4769
+    },
+    {
+      "epoch": 0.7803203661327232,
+      "grad_norm": 7.22778526039397,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 4770
+    },
+    {
+      "epoch": 0.782608695652174,
+      "grad_norm": 6.758098761488354,
+      "learning_rate": 1e-06,
+      "loss": 0.3413,
+      "step": 4771
+    },
+    {
+      "epoch": 0.7848970251716247,
+      "grad_norm": 8.663912736378826,
+      "learning_rate": 1e-06,
+      "loss": 0.3663,
+      "step": 4772
+    },
+    {
+      "epoch": 0.7871853546910755,
+      "grad_norm": 7.657668617755174,
+      "learning_rate": 1e-06,
+      "loss": 0.3425,
+      "step": 4773
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 5.365597851025376,
+      "learning_rate": 1e-06,
+      "loss": 0.3556,
+      "step": 4774
+    },
+    {
+      "epoch": 0.7917620137299771,
+      "grad_norm": 8.39409230719484,
+      "learning_rate": 1e-06,
+      "loss": 0.3485,
+      "step": 4775
+    },
+    {
+      "epoch": 0.7940503432494279,
+      "grad_norm": 8.30559853297212,
+      "learning_rate": 1e-06,
+      "loss": 0.3403,
+      "step": 4776
+    },
+    {
+      "epoch": 0.7963386727688787,
+      "grad_norm": 6.206991369188037,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 4777
+    },
+    {
+      "epoch": 0.7986270022883295,
+      "grad_norm": 5.96962836830053,
+      "learning_rate": 1e-06,
+      "loss": 0.3377,
+      "step": 4778
+    },
+    {
+      "epoch": 0.8009153318077803,
+      "grad_norm": 7.231775563469396,
+      "learning_rate": 1e-06,
+      "loss": 0.3731,
+      "step": 4779
+    },
+    {
+      "epoch": 0.8032036613272311,
+      "grad_norm": 8.550154474604087,
+      "learning_rate": 1e-06,
+      "loss": 0.339,
+      "step": 4780
+    },
+    {
+      "epoch": 0.8054919908466819,
+      "grad_norm": 6.907638441076727,
+      "learning_rate": 1e-06,
+      "loss": 0.3594,
+      "step": 4781
+    },
+    {
+      "epoch": 0.8077803203661327,
+      "grad_norm": 5.457093118801232,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 4782
+    },
+    {
+      "epoch": 0.8100686498855835,
+      "grad_norm": 6.097741158497243,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 4783
+    },
+    {
+      "epoch": 0.8123569794050344,
+      "grad_norm": 8.016116384626569,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 4784
+    },
+    {
+      "epoch": 0.8146453089244852,
+      "grad_norm": 6.775662794221159,
+      "learning_rate": 1e-06,
+      "loss": 0.3381,
+      "step": 4785
+    },
+    {
+      "epoch": 0.816933638443936,
+      "grad_norm": 8.182093529292672,
+      "learning_rate": 1e-06,
+      "loss": 0.344,
+      "step": 4786
+    },
+    {
+      "epoch": 0.8192219679633868,
+      "grad_norm": 5.961795781342375,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 4787
+    },
+    {
+      "epoch": 0.8215102974828375,
+      "grad_norm": 5.320292396260607,
+      "learning_rate": 1e-06,
+      "loss": 0.3449,
+      "step": 4788
+    },
+    {
+      "epoch": 0.8237986270022883,
+      "grad_norm": 7.527081811771644,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 4789
+    },
+    {
+      "epoch": 0.8260869565217391,
+      "grad_norm": 5.452733278850019,
+      "learning_rate": 1e-06,
+      "loss": 0.3597,
+      "step": 4790
+    },
+    {
+      "epoch": 0.8283752860411899,
+      "grad_norm": 6.667014079323754,
+      "learning_rate": 1e-06,
+      "loss": 0.3768,
+      "step": 4791
+    },
+    {
+      "epoch": 0.8306636155606407,
+      "grad_norm": 6.6925351912355335,
+      "learning_rate": 1e-06,
+      "loss": 0.3409,
+      "step": 4792
+    },
+    {
+      "epoch": 0.8329519450800915,
+      "grad_norm": 6.756560735863721,
+      "learning_rate": 1e-06,
+      "loss": 0.3354,
+      "step": 4793
+    },
+    {
+      "epoch": 0.8352402745995423,
+      "grad_norm": 7.149362082371233,
+      "learning_rate": 1e-06,
+      "loss": 0.3862,
+      "step": 4794
+    },
+    {
+      "epoch": 0.8375286041189931,
+      "grad_norm": 8.145146612249428,
+      "learning_rate": 1e-06,
+      "loss": 0.3788,
+      "step": 4795
+    },
+    {
+      "epoch": 0.8398169336384439,
+      "grad_norm": 8.75909610105236,
+      "learning_rate": 1e-06,
+      "loss": 0.3429,
+      "step": 4796
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 6.919330146955886,
+      "learning_rate": 1e-06,
+      "loss": 0.3757,
+      "step": 4797
+    },
+    {
+      "epoch": 0.8443935926773455,
+      "grad_norm": 6.133201653013371,
+      "learning_rate": 1e-06,
+      "loss": 0.3449,
+      "step": 4798
+    },
+    {
+      "epoch": 0.8466819221967964,
+      "grad_norm": 6.210200130550881,
+      "learning_rate": 1e-06,
+      "loss": 0.3391,
+      "step": 4799
+    },
+    {
+      "epoch": 0.8489702517162472,
+      "grad_norm": 6.133844581322985,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 4800
+    },
+    {
+      "epoch": 0.851258581235698,
+      "grad_norm": 7.7234883890577475,
+      "learning_rate": 1e-06,
+      "loss": 0.3402,
+      "step": 4801
+    },
+    {
+      "epoch": 0.8535469107551488,
+      "grad_norm": 9.15873309136272,
+      "learning_rate": 1e-06,
+      "loss": 0.3586,
+      "step": 4802
+    },
+    {
+      "epoch": 0.8558352402745996,
+      "grad_norm": 9.390024021525308,
+      "learning_rate": 1e-06,
+      "loss": 0.3599,
+      "step": 4803
+    },
+    {
+      "epoch": 0.8581235697940504,
+      "grad_norm": 8.189345454948048,
+      "learning_rate": 1e-06,
+      "loss": 0.3516,
+      "step": 4804
+    },
+    {
+      "epoch": 0.8604118993135011,
+      "grad_norm": 6.847789128077376,
+      "learning_rate": 1e-06,
+      "loss": 0.3604,
+      "step": 4805
+    },
+    {
+      "epoch": 0.8627002288329519,
+      "grad_norm": 6.527073182858159,
+      "learning_rate": 1e-06,
+      "loss": 0.3775,
+      "step": 4806
+    },
+    {
+      "epoch": 0.8649885583524027,
+      "grad_norm": 7.145221504939826,
+      "learning_rate": 1e-06,
+      "loss": 0.3358,
+      "step": 4807
+    },
+    {
+      "epoch": 0.8672768878718535,
+      "grad_norm": 8.832549650955553,
+      "learning_rate": 1e-06,
+      "loss": 0.3686,
+      "step": 4808
+    },
+    {
+      "epoch": 0.8695652173913043,
+      "grad_norm": 7.5598858243156295,
+      "learning_rate": 1e-06,
+      "loss": 0.3404,
+      "step": 4809
+    },
+    {
+      "epoch": 0.8718535469107551,
+      "grad_norm": 5.555184281524914,
+      "learning_rate": 1e-06,
+      "loss": 0.3233,
+      "step": 4810
+    },
+    {
+      "epoch": 0.8741418764302059,
+      "grad_norm": 8.298761381767802,
+      "learning_rate": 1e-06,
+      "loss": 0.3642,
+      "step": 4811
+    },
+    {
+      "epoch": 0.8764302059496567,
+      "grad_norm": 5.704898475424967,
+      "learning_rate": 1e-06,
+      "loss": 0.3361,
+      "step": 4812
+    },
+    {
+      "epoch": 0.8787185354691075,
+      "grad_norm": 6.787295851255482,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 4813
+    },
+    {
+      "epoch": 0.8810068649885584,
+      "grad_norm": 8.315519556706775,
+      "learning_rate": 1e-06,
+      "loss": 0.3346,
+      "step": 4814
+    },
+    {
+      "epoch": 0.8832951945080092,
+      "grad_norm": 8.20839098493246,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 4815
+    },
+    {
+      "epoch": 0.88558352402746,
+      "grad_norm": 8.152236849289766,
+      "learning_rate": 1e-06,
+      "loss": 0.3566,
+      "step": 4816
+    },
+    {
+      "epoch": 0.8878718535469108,
+      "grad_norm": 7.763532176530299,
+      "learning_rate": 1e-06,
+      "loss": 0.3627,
+      "step": 4817
+    },
+    {
+      "epoch": 0.8901601830663616,
+      "grad_norm": 7.290130560205567,
+      "learning_rate": 1e-06,
+      "loss": 0.3454,
+      "step": 4818
+    },
+    {
+      "epoch": 0.8924485125858124,
+      "grad_norm": 8.023770567197198,
+      "learning_rate": 1e-06,
+      "loss": 0.3648,
+      "step": 4819
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 5.891650670673806,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 4820
+    },
+    {
+      "epoch": 0.897025171624714,
+      "grad_norm": 7.049973341840916,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 4821
+    },
+    {
+      "epoch": 0.8993135011441648,
+      "grad_norm": 5.718064956131607,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 4822
+    },
+    {
+      "epoch": 0.9016018306636155,
+      "grad_norm": 7.142045599362644,
+      "learning_rate": 1e-06,
+      "loss": 0.3134,
+      "step": 4823
+    },
+    {
+      "epoch": 0.9038901601830663,
+      "grad_norm": 5.982959890065492,
+      "learning_rate": 1e-06,
+      "loss": 0.3398,
+      "step": 4824
+    },
+    {
+      "epoch": 0.9061784897025171,
+      "grad_norm": 7.783479270963185,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 4825
+    },
+    {
+      "epoch": 0.9084668192219679,
+      "grad_norm": 6.15329207029583,
+      "learning_rate": 1e-06,
+      "loss": 0.3589,
+      "step": 4826
+    },
+    {
+      "epoch": 0.9107551487414187,
+      "grad_norm": 8.34387128257159,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 4827
+    },
+    {
+      "epoch": 0.9130434782608695,
+      "grad_norm": 7.165028434115515,
+      "learning_rate": 1e-06,
+      "loss": 0.3598,
+      "step": 4828
+    },
+    {
+      "epoch": 0.9153318077803204,
+      "grad_norm": 7.6997478488138,
+      "learning_rate": 1e-06,
+      "loss": 0.3603,
+      "step": 4829
+    },
+    {
+      "epoch": 0.9176201372997712,
+      "grad_norm": 7.687468426464388,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 4830
+    },
+    {
+      "epoch": 0.919908466819222,
+      "grad_norm": 6.565828433541215,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 4831
+    },
+    {
+      "epoch": 0.9221967963386728,
+      "grad_norm": 6.764835219890548,
+      "learning_rate": 1e-06,
+      "loss": 0.3684,
+      "step": 4832
+    },
+    {
+      "epoch": 0.9244851258581236,
+      "grad_norm": 7.363836723562641,
+      "learning_rate": 1e-06,
+      "loss": 0.3606,
+      "step": 4833
+    },
+    {
+      "epoch": 0.9267734553775744,
+      "grad_norm": 6.756799560322972,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 4834
+    },
+    {
+      "epoch": 0.9290617848970252,
+      "grad_norm": 7.291330286841248,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 4835
+    },
+    {
+      "epoch": 0.931350114416476,
+      "grad_norm": 7.803202750831844,
+      "learning_rate": 1e-06,
+      "loss": 0.3475,
+      "step": 4836
+    },
+    {
+      "epoch": 0.9336384439359268,
+      "grad_norm": 8.817376018142205,
+      "learning_rate": 1e-06,
+      "loss": 0.3645,
+      "step": 4837
+    },
+    {
+      "epoch": 0.9359267734553776,
+      "grad_norm": 8.082162011392256,
+      "learning_rate": 1e-06,
+      "loss": 0.351,
+      "step": 4838
+    },
+    {
+      "epoch": 0.9382151029748284,
+      "grad_norm": 6.156888576925938,
+      "learning_rate": 1e-06,
+      "loss": 0.3418,
+      "step": 4839
+    },
+    {
+      "epoch": 0.9405034324942791,
+      "grad_norm": 5.9419188372681235,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 4840
+    },
+    {
+      "epoch": 0.9427917620137299,
+      "grad_norm": 8.233472500543927,
+      "learning_rate": 1e-06,
+      "loss": 0.3502,
+      "step": 4841
+    },
+    {
+      "epoch": 0.9450800915331807,
+      "grad_norm": 9.600024884926828,
+      "learning_rate": 1e-06,
+      "loss": 0.3602,
+      "step": 4842
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 6.604797832508421,
+      "learning_rate": 1e-06,
+      "loss": 0.3391,
+      "step": 4843
+    },
+    {
+      "epoch": 0.9496567505720824,
+      "grad_norm": 6.538715706387555,
+      "learning_rate": 1e-06,
+      "loss": 0.3394,
+      "step": 4844
+    },
+    {
+      "epoch": 0.9519450800915332,
+      "grad_norm": 6.5794533195910825,
+      "learning_rate": 1e-06,
+      "loss": 0.3557,
+      "step": 4845
+    },
+    {
+      "epoch": 0.954233409610984,
+      "grad_norm": 6.412548263828761,
+      "learning_rate": 1e-06,
+      "loss": 0.378,
+      "step": 4846
+    },
+    {
+      "epoch": 0.9565217391304348,
+      "grad_norm": 8.243098790325602,
+      "learning_rate": 1e-06,
+      "loss": 0.3537,
+      "step": 4847
+    },
+    {
+      "epoch": 0.9588100686498856,
+      "grad_norm": 7.20151870760334,
+      "learning_rate": 1e-06,
+      "loss": 0.3475,
+      "step": 4848
+    },
+    {
+      "epoch": 0.9610983981693364,
+      "grad_norm": 7.368501611047199,
+      "learning_rate": 1e-06,
+      "loss": 0.3719,
+      "step": 4849
+    },
+    {
+      "epoch": 0.9633867276887872,
+      "grad_norm": 8.507851584520317,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 4850
+    },
+    {
+      "epoch": 0.965675057208238,
+      "grad_norm": 5.805812953635095,
+      "learning_rate": 1e-06,
+      "loss": 0.3279,
+      "step": 4851
+    },
+    {
+      "epoch": 0.9679633867276888,
+      "grad_norm": 6.095101230441509,
+      "learning_rate": 1e-06,
+      "loss": 0.351,
+      "step": 4852
+    },
+    {
+      "epoch": 0.9702517162471396,
+      "grad_norm": 8.034497439381001,
+      "learning_rate": 1e-06,
+      "loss": 0.3552,
+      "step": 4853
+    },
+    {
+      "epoch": 0.9725400457665904,
+      "grad_norm": 8.704467626281138,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 4854
+    },
+    {
+      "epoch": 0.9748283752860412,
+      "grad_norm": 5.986617184640766,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 4855
+    },
+    {
+      "epoch": 0.977116704805492,
+      "grad_norm": 5.73036395233407,
+      "learning_rate": 1e-06,
+      "loss": 0.3694,
+      "step": 4856
+    },
+    {
+      "epoch": 0.9794050343249427,
+      "grad_norm": 5.883667773405379,
+      "learning_rate": 1e-06,
+      "loss": 0.3506,
+      "step": 4857
+    },
+    {
+      "epoch": 0.9816933638443935,
+      "grad_norm": 9.214084649425777,
+      "learning_rate": 1e-06,
+      "loss": 0.3612,
+      "step": 4858
+    },
+    {
+      "epoch": 0.9839816933638444,
+      "grad_norm": 6.866502017204064,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 4859
+    },
+    {
+      "epoch": 0.9862700228832952,
+      "grad_norm": 8.008718542503226,
+      "learning_rate": 1e-06,
+      "loss": 0.3861,
+      "step": 4860
+    },
+    {
+      "epoch": 0.988558352402746,
+      "grad_norm": 8.470645551472668,
+      "learning_rate": 1e-06,
+      "loss": 0.3557,
+      "step": 4861
+    },
+    {
+      "epoch": 0.9908466819221968,
+      "grad_norm": 7.547056196500224,
+      "learning_rate": 1e-06,
+      "loss": 0.3309,
+      "step": 4862
+    },
+    {
+      "epoch": 0.9931350114416476,
+      "grad_norm": 5.931046249261669,
+      "learning_rate": 1e-06,
+      "loss": 0.3548,
+      "step": 4863
+    },
+    {
+      "epoch": 0.9954233409610984,
+      "grad_norm": 5.198062385966086,
+      "learning_rate": 1e-06,
+      "loss": 0.3532,
+      "step": 4864
+    },
+    {
+      "epoch": 0.9977116704805492,
+      "grad_norm": 7.587097723262857,
+      "learning_rate": 1e-06,
+      "loss": 0.3606,
+      "step": 4865
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 6.981536063546113,
+      "learning_rate": 1e-06,
+      "loss": 0.361,
+      "step": 4866
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 2.630324363708496,
+      "eval_runtime": 21.7998,
+      "eval_samples_per_second": 45.872,
+      "eval_steps_per_second": 5.734,
+      "step": 4866
+    },
+    {
+      "epoch": 0.0034423407917383822,
+      "grad_norm": 8.399436597058736,
+      "learning_rate": 1e-06,
+      "loss": 0.7743,
+      "step": 4867
+    },
+    {
+      "epoch": 0.0068846815834767644,
+      "grad_norm": 10.161157688223156,
+      "learning_rate": 1e-06,
+      "loss": 0.6737,
+      "step": 4868
+    },
+    {
+      "epoch": 0.010327022375215147,
+      "grad_norm": 9.763662856106027,
+      "learning_rate": 1e-06,
+      "loss": 0.7361,
+      "step": 4869
+    },
+    {
+      "epoch": 0.013769363166953529,
+      "grad_norm": 8.702675387607961,
+      "learning_rate": 1e-06,
+      "loss": 0.6234,
+      "step": 4870
+    },
+    {
+      "epoch": 0.01721170395869191,
+      "grad_norm": 9.121919469553047,
+      "learning_rate": 1e-06,
+      "loss": 0.6401,
+      "step": 4871
+    },
+    {
+      "epoch": 0.020654044750430294,
+      "grad_norm": 8.72592552804516,
+      "learning_rate": 1e-06,
+      "loss": 0.6849,
+      "step": 4872
+    },
+    {
+      "epoch": 0.024096385542168676,
+      "grad_norm": 9.975130376517516,
+      "learning_rate": 1e-06,
+      "loss": 0.6658,
+      "step": 4873
+    },
+    {
+      "epoch": 0.027538726333907058,
+      "grad_norm": 7.498989982818483,
+      "learning_rate": 1e-06,
+      "loss": 0.6334,
+      "step": 4874
+    },
+    {
+      "epoch": 0.03098106712564544,
+      "grad_norm": 8.389514230938643,
+      "learning_rate": 1e-06,
+      "loss": 0.6032,
+      "step": 4875
+    },
+    {
+      "epoch": 0.03442340791738382,
+      "grad_norm": 7.921518463046678,
+      "learning_rate": 1e-06,
+      "loss": 0.6683,
+      "step": 4876
+    },
+    {
+      "epoch": 0.0378657487091222,
+      "grad_norm": 6.977696157286023,
+      "learning_rate": 1e-06,
+      "loss": 0.6206,
+      "step": 4877
+    },
+    {
+      "epoch": 0.04130808950086059,
+      "grad_norm": 7.464238954098718,
+      "learning_rate": 1e-06,
+      "loss": 0.6633,
+      "step": 4878
+    },
+    {
+      "epoch": 0.04475043029259897,
+      "grad_norm": 7.659748880379759,
+      "learning_rate": 1e-06,
+      "loss": 0.6995,
+      "step": 4879
+    },
+    {
+      "epoch": 0.04819277108433735,
+      "grad_norm": 6.247308025053432,
+      "learning_rate": 1e-06,
+      "loss": 0.655,
+      "step": 4880
+    },
+    {
+      "epoch": 0.05163511187607573,
+      "grad_norm": 7.604569368559685,
+      "learning_rate": 1e-06,
+      "loss": 0.6174,
+      "step": 4881
+    },
+    {
+      "epoch": 0.055077452667814115,
+      "grad_norm": 6.71425522360885,
+      "learning_rate": 1e-06,
+      "loss": 0.6535,
+      "step": 4882
+    },
+    {
+      "epoch": 0.058519793459552494,
+      "grad_norm": 8.187028386768288,
+      "learning_rate": 1e-06,
+      "loss": 0.6446,
+      "step": 4883
+    },
+    {
+      "epoch": 0.06196213425129088,
+      "grad_norm": 8.436287716451634,
+      "learning_rate": 1e-06,
+      "loss": 0.6624,
+      "step": 4884
+    },
+    {
+      "epoch": 0.06540447504302926,
+      "grad_norm": 7.771337543165457,
+      "learning_rate": 1e-06,
+      "loss": 0.6397,
+      "step": 4885
+    },
+    {
+      "epoch": 0.06884681583476764,
+      "grad_norm": 6.918527925187201,
+      "learning_rate": 1e-06,
+      "loss": 0.7029,
+      "step": 4886
+    },
+    {
+      "epoch": 0.07228915662650602,
+      "grad_norm": 6.1505155281229955,
+      "learning_rate": 1e-06,
+      "loss": 0.6447,
+      "step": 4887
+    },
+    {
+      "epoch": 0.0757314974182444,
+      "grad_norm": 7.908605067275009,
+      "learning_rate": 1e-06,
+      "loss": 0.6772,
+      "step": 4888
+    },
+    {
+      "epoch": 0.07917383820998279,
+      "grad_norm": 6.6563119990006205,
+      "learning_rate": 1e-06,
+      "loss": 0.6514,
+      "step": 4889
+    },
+    {
+      "epoch": 0.08261617900172118,
+      "grad_norm": 6.028004522610612,
+      "learning_rate": 1e-06,
+      "loss": 0.6779,
+      "step": 4890
+    },
+    {
+      "epoch": 0.08605851979345955,
+      "grad_norm": 5.233800598699478,
+      "learning_rate": 1e-06,
+      "loss": 0.5903,
+      "step": 4891
+    },
+    {
+      "epoch": 0.08950086058519793,
+      "grad_norm": 6.474110845806838,
+      "learning_rate": 1e-06,
+      "loss": 0.6779,
+      "step": 4892
+    },
+    {
+      "epoch": 0.09294320137693632,
+      "grad_norm": 5.646675445640911,
+      "learning_rate": 1e-06,
+      "loss": 0.6862,
+      "step": 4893
+    },
+    {
+      "epoch": 0.0963855421686747,
+      "grad_norm": 6.848458525442593,
+      "learning_rate": 1e-06,
+      "loss": 0.7048,
+      "step": 4894
+    },
+    {
+      "epoch": 0.09982788296041308,
+      "grad_norm": 8.121797350832546,
+      "learning_rate": 1e-06,
+      "loss": 0.6079,
+      "step": 4895
+    },
+    {
+      "epoch": 0.10327022375215146,
+      "grad_norm": 7.9300857469926225,
+      "learning_rate": 1e-06,
+      "loss": 0.6606,
+      "step": 4896
+    },
+    {
+      "epoch": 0.10671256454388985,
+      "grad_norm": 6.323326319324455,
+      "learning_rate": 1e-06,
+      "loss": 0.6916,
+      "step": 4897
+    },
+    {
+      "epoch": 0.11015490533562823,
+      "grad_norm": 6.0512839706010535,
+      "learning_rate": 1e-06,
+      "loss": 0.6422,
+      "step": 4898
+    },
+    {
+      "epoch": 0.11359724612736662,
+      "grad_norm": 6.514847775693534,
+      "learning_rate": 1e-06,
+      "loss": 0.7382,
+      "step": 4899
+    },
+    {
+      "epoch": 0.11703958691910499,
+      "grad_norm": 5.617460000292631,
+      "learning_rate": 1e-06,
+      "loss": 0.7478,
+      "step": 4900
+    },
+    {
+      "epoch": 0.12048192771084337,
+      "grad_norm": 8.01138297977037,
+      "learning_rate": 1e-06,
+      "loss": 0.6432,
+      "step": 4901
+    },
+    {
+      "epoch": 0.12392426850258176,
+      "grad_norm": 5.400682675917501,
+      "learning_rate": 1e-06,
+      "loss": 0.603,
+      "step": 4902
+    },
+    {
+      "epoch": 0.12736660929432014,
+      "grad_norm": 8.280476664195115,
+      "learning_rate": 1e-06,
+      "loss": 0.6088,
+      "step": 4903
+    },
+    {
+      "epoch": 0.13080895008605853,
+      "grad_norm": 7.652605780372027,
+      "learning_rate": 1e-06,
+      "loss": 0.6574,
+      "step": 4904
+    },
+    {
+      "epoch": 0.1342512908777969,
+      "grad_norm": 7.761442175298503,
+      "learning_rate": 1e-06,
+      "loss": 0.7435,
+      "step": 4905
+    },
+    {
+      "epoch": 0.13769363166953527,
+      "grad_norm": 7.1671306820655705,
+      "learning_rate": 1e-06,
+      "loss": 0.6623,
+      "step": 4906
+    },
+    {
+      "epoch": 0.14113597246127366,
+      "grad_norm": 6.944707153771651,
+      "learning_rate": 1e-06,
+      "loss": 0.6358,
+      "step": 4907
+    },
+    {
+      "epoch": 0.14457831325301204,
+      "grad_norm": 5.534431052511391,
+      "learning_rate": 1e-06,
+      "loss": 0.6903,
+      "step": 4908
+    },
+    {
+      "epoch": 0.14802065404475043,
+      "grad_norm": 7.676488398356956,
+      "learning_rate": 1e-06,
+      "loss": 0.6143,
+      "step": 4909
+    },
+    {
+      "epoch": 0.1514629948364888,
+      "grad_norm": 7.598239936524427,
+      "learning_rate": 1e-06,
+      "loss": 0.6188,
+      "step": 4910
+    },
+    {
+      "epoch": 0.1549053356282272,
+      "grad_norm": 5.221953900191017,
+      "learning_rate": 1e-06,
+      "loss": 0.6002,
+      "step": 4911
+    },
+    {
+      "epoch": 0.15834767641996558,
+      "grad_norm": 6.04976439982155,
+      "learning_rate": 1e-06,
+      "loss": 0.6651,
+      "step": 4912
+    },
+    {
+      "epoch": 0.16179001721170397,
+      "grad_norm": 9.252679329041683,
+      "learning_rate": 1e-06,
+      "loss": 0.661,
+      "step": 4913
+    },
+    {
+      "epoch": 0.16523235800344235,
+      "grad_norm": 7.216422392933027,
+      "learning_rate": 1e-06,
+      "loss": 0.6713,
+      "step": 4914
+    },
+    {
+      "epoch": 0.1686746987951807,
+      "grad_norm": 4.846790339875952,
+      "learning_rate": 1e-06,
+      "loss": 0.6778,
+      "step": 4915
+    },
+    {
+      "epoch": 0.1721170395869191,
+      "grad_norm": 8.283500721492313,
+      "learning_rate": 1e-06,
+      "loss": 0.6571,
+      "step": 4916
+    },
+    {
+      "epoch": 0.17555938037865748,
+      "grad_norm": 6.381715055837104,
+      "learning_rate": 1e-06,
+      "loss": 0.6746,
+      "step": 4917
+    },
+    {
+      "epoch": 0.17900172117039587,
+      "grad_norm": 8.768924198361102,
+      "learning_rate": 1e-06,
+      "loss": 0.6981,
+      "step": 4918
+    },
+    {
+      "epoch": 0.18244406196213425,
+      "grad_norm": 7.466662834566129,
+      "learning_rate": 1e-06,
+      "loss": 0.6683,
+      "step": 4919
+    },
+    {
+      "epoch": 0.18588640275387264,
+      "grad_norm": 5.673242179793482,
+      "learning_rate": 1e-06,
+      "loss": 0.6174,
+      "step": 4920
+    },
+    {
+      "epoch": 0.18932874354561102,
+      "grad_norm": 6.9740461535120835,
+      "learning_rate": 1e-06,
+      "loss": 0.6301,
+      "step": 4921
+    },
+    {
+      "epoch": 0.1927710843373494,
+      "grad_norm": 5.686708976009216,
+      "learning_rate": 1e-06,
+      "loss": 0.69,
+      "step": 4922
+    },
+    {
+      "epoch": 0.1962134251290878,
+      "grad_norm": 7.254689113950109,
+      "learning_rate": 1e-06,
+      "loss": 0.612,
+      "step": 4923
+    },
+    {
+      "epoch": 0.19965576592082615,
+      "grad_norm": 6.611068280937196,
+      "learning_rate": 1e-06,
+      "loss": 0.6913,
+      "step": 4924
+    },
+    {
+      "epoch": 0.20309810671256454,
+      "grad_norm": 6.637441095575761,
+      "learning_rate": 1e-06,
+      "loss": 0.6765,
+      "step": 4925
+    },
+    {
+      "epoch": 0.20654044750430292,
+      "grad_norm": 6.916689203007378,
+      "learning_rate": 1e-06,
+      "loss": 0.6639,
+      "step": 4926
+    },
+    {
+      "epoch": 0.2099827882960413,
+      "grad_norm": 5.940060050753901,
+      "learning_rate": 1e-06,
+      "loss": 0.6411,
+      "step": 4927
+    },
+    {
+      "epoch": 0.2134251290877797,
+      "grad_norm": 7.745289859524665,
+      "learning_rate": 1e-06,
+      "loss": 0.6338,
+      "step": 4928
+    },
+    {
+      "epoch": 0.21686746987951808,
+      "grad_norm": 6.9616256581524185,
+      "learning_rate": 1e-06,
+      "loss": 0.6538,
+      "step": 4929
+    },
+    {
+      "epoch": 0.22030981067125646,
+      "grad_norm": 8.143326062402252,
+      "learning_rate": 1e-06,
+      "loss": 0.6933,
+      "step": 4930
+    },
+    {
+      "epoch": 0.22375215146299485,
+      "grad_norm": 6.210738988956374,
+      "learning_rate": 1e-06,
+      "loss": 0.6451,
+      "step": 4931
+    },
+    {
+      "epoch": 0.22719449225473323,
+      "grad_norm": 8.199306009610405,
+      "learning_rate": 1e-06,
+      "loss": 0.6969,
+      "step": 4932
+    },
+    {
+      "epoch": 0.2306368330464716,
+      "grad_norm": 6.370014116483212,
+      "learning_rate": 1e-06,
+      "loss": 0.6507,
+      "step": 4933
+    },
+    {
+      "epoch": 0.23407917383820998,
+      "grad_norm": 6.466381981342019,
+      "learning_rate": 1e-06,
+      "loss": 0.6992,
+      "step": 4934
+    },
+    {
+      "epoch": 0.23752151462994836,
+      "grad_norm": 7.794003307349674,
+      "learning_rate": 1e-06,
+      "loss": 0.715,
+      "step": 4935
+    },
+    {
+      "epoch": 0.24096385542168675,
+      "grad_norm": 6.860140154472764,
+      "learning_rate": 1e-06,
+      "loss": 0.6595,
+      "step": 4936
+    },
+    {
+      "epoch": 0.24440619621342513,
+      "grad_norm": 7.809426328666326,
+      "learning_rate": 1e-06,
+      "loss": 0.6713,
+      "step": 4937
+    },
+    {
+      "epoch": 0.24784853700516352,
+      "grad_norm": 6.735827979602437,
+      "learning_rate": 1e-06,
+      "loss": 0.5822,
+      "step": 4938
+    },
+    {
+      "epoch": 0.2512908777969019,
+      "grad_norm": 7.466841243338734,
+      "learning_rate": 1e-06,
+      "loss": 0.6503,
+      "step": 4939
+    },
+    {
+      "epoch": 0.2547332185886403,
+      "grad_norm": 6.248830010885907,
+      "learning_rate": 1e-06,
+      "loss": 0.6302,
+      "step": 4940
+    },
+    {
+      "epoch": 0.25817555938037867,
+      "grad_norm": 7.3914730364753085,
+      "learning_rate": 1e-06,
+      "loss": 0.6732,
+      "step": 4941
+    },
+    {
+      "epoch": 0.26161790017211706,
+      "grad_norm": 6.7590199395294706,
+      "learning_rate": 1e-06,
+      "loss": 0.6553,
+      "step": 4942
+    },
+    {
+      "epoch": 0.26506024096385544,
+      "grad_norm": 6.452809115559429,
+      "learning_rate": 1e-06,
+      "loss": 0.6376,
+      "step": 4943
+    },
+    {
+      "epoch": 0.2685025817555938,
+      "grad_norm": 6.412845774571026,
+      "learning_rate": 1e-06,
+      "loss": 0.6608,
+      "step": 4944
+    },
+    {
+      "epoch": 0.2719449225473322,
+      "grad_norm": 7.20350858895672,
+      "learning_rate": 1e-06,
+      "loss": 0.6416,
+      "step": 4945
+    },
+    {
+      "epoch": 0.27538726333907054,
+      "grad_norm": 7.220261841192338,
+      "learning_rate": 1e-06,
+      "loss": 0.6929,
+      "step": 4946
+    },
+    {
+      "epoch": 0.27882960413080893,
+      "grad_norm": 6.41250276737705,
+      "learning_rate": 1e-06,
+      "loss": 0.6991,
+      "step": 4947
+    },
+    {
+      "epoch": 0.2822719449225473,
+      "grad_norm": 7.593134635179562,
+      "learning_rate": 1e-06,
+      "loss": 0.7026,
+      "step": 4948
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 6.093354616879217,
+      "learning_rate": 1e-06,
+      "loss": 0.6494,
+      "step": 4949
+    },
+    {
+      "epoch": 0.2891566265060241,
+      "grad_norm": 6.481254091907865,
+      "learning_rate": 1e-06,
+      "loss": 0.7031,
+      "step": 4950
+    },
+    {
+      "epoch": 0.29259896729776247,
+      "grad_norm": 6.502724404730937,
+      "learning_rate": 1e-06,
+      "loss": 0.6696,
+      "step": 4951
+    },
+    {
+      "epoch": 0.29604130808950085,
+      "grad_norm": 7.61944198033823,
+      "learning_rate": 1e-06,
+      "loss": 0.6927,
+      "step": 4952
+    },
+    {
+      "epoch": 0.29948364888123924,
+      "grad_norm": 7.210785188246807,
+      "learning_rate": 1e-06,
+      "loss": 0.6469,
+      "step": 4953
+    },
+    {
+      "epoch": 0.3029259896729776,
+      "grad_norm": 5.876808391968143,
+      "learning_rate": 1e-06,
+      "loss": 0.6567,
+      "step": 4954
+    },
+    {
+      "epoch": 0.306368330464716,
+      "grad_norm": 6.14372052611385,
+      "learning_rate": 1e-06,
+      "loss": 0.6777,
+      "step": 4955
+    },
+    {
+      "epoch": 0.3098106712564544,
+      "grad_norm": 6.989092787718136,
+      "learning_rate": 1e-06,
+      "loss": 0.6679,
+      "step": 4956
+    },
+    {
+      "epoch": 0.3132530120481928,
+      "grad_norm": 6.257706600588234,
+      "learning_rate": 1e-06,
+      "loss": 0.6461,
+      "step": 4957
+    },
+    {
+      "epoch": 0.31669535283993117,
+      "grad_norm": 7.814330974463549,
+      "learning_rate": 1e-06,
+      "loss": 0.6758,
+      "step": 4958
+    },
+    {
+      "epoch": 0.32013769363166955,
+      "grad_norm": 6.1419061107813535,
+      "learning_rate": 1e-06,
+      "loss": 0.6802,
+      "step": 4959
+    },
+    {
+      "epoch": 0.32358003442340794,
+      "grad_norm": 8.330703161051863,
+      "learning_rate": 1e-06,
+      "loss": 0.6286,
+      "step": 4960
+    },
+    {
+      "epoch": 0.3270223752151463,
+      "grad_norm": 6.834600686619839,
+      "learning_rate": 1e-06,
+      "loss": 0.661,
+      "step": 4961
+    },
+    {
+      "epoch": 0.3304647160068847,
+      "grad_norm": 8.101302338683958,
+      "learning_rate": 1e-06,
+      "loss": 0.6851,
+      "step": 4962
+    },
+    {
+      "epoch": 0.33390705679862304,
+      "grad_norm": 7.058367291614693,
+      "learning_rate": 1e-06,
+      "loss": 0.6842,
+      "step": 4963
+    },
+    {
+      "epoch": 0.3373493975903614,
+      "grad_norm": 7.764903507021077,
+      "learning_rate": 1e-06,
+      "loss": 0.701,
+      "step": 4964
+    },
+    {
+      "epoch": 0.3407917383820998,
+      "grad_norm": 6.475143593419591,
+      "learning_rate": 1e-06,
+      "loss": 0.6399,
+      "step": 4965
+    },
+    {
+      "epoch": 0.3442340791738382,
+      "grad_norm": 6.744064726593709,
+      "learning_rate": 1e-06,
+      "loss": 0.6532,
+      "step": 4966
+    },
+    {
+      "epoch": 0.3476764199655766,
+      "grad_norm": 7.091584409575436,
+      "learning_rate": 1e-06,
+      "loss": 0.6486,
+      "step": 4967
+    },
+    {
+      "epoch": 0.35111876075731496,
+      "grad_norm": 7.049409608546178,
+      "learning_rate": 1e-06,
+      "loss": 0.6233,
+      "step": 4968
+    },
+    {
+      "epoch": 0.35456110154905335,
+      "grad_norm": 7.48585649564644,
+      "learning_rate": 1e-06,
+      "loss": 0.6692,
+      "step": 4969
+    },
+    {
+      "epoch": 0.35800344234079173,
+      "grad_norm": 7.379508570867265,
+      "learning_rate": 1e-06,
+      "loss": 0.6368,
+      "step": 4970
+    },
+    {
+      "epoch": 0.3614457831325301,
+      "grad_norm": 6.114897146855558,
+      "learning_rate": 1e-06,
+      "loss": 0.628,
+      "step": 4971
+    },
+    {
+      "epoch": 0.3648881239242685,
+      "grad_norm": 8.593850826923646,
+      "learning_rate": 1e-06,
+      "loss": 0.6208,
+      "step": 4972
+    },
+    {
+      "epoch": 0.3683304647160069,
+      "grad_norm": 7.083905113706674,
+      "learning_rate": 1e-06,
+      "loss": 0.6298,
+      "step": 4973
+    },
+    {
+      "epoch": 0.3717728055077453,
+      "grad_norm": 7.041568863828795,
+      "learning_rate": 1e-06,
+      "loss": 0.694,
+      "step": 4974
+    },
+    {
+      "epoch": 0.37521514629948366,
+      "grad_norm": 6.468250496059003,
+      "learning_rate": 1e-06,
+      "loss": 0.6597,
+      "step": 4975
+    },
+    {
+      "epoch": 0.37865748709122204,
+      "grad_norm": 7.742102855994714,
+      "learning_rate": 1e-06,
+      "loss": 0.6734,
+      "step": 4976
+    },
+    {
+      "epoch": 0.38209982788296043,
+      "grad_norm": 5.994504908983731,
+      "learning_rate": 1e-06,
+      "loss": 0.6375,
+      "step": 4977
+    },
+    {
+      "epoch": 0.3855421686746988,
+      "grad_norm": 6.975150574275447,
+      "learning_rate": 1e-06,
+      "loss": 0.6689,
+      "step": 4978
+    },
+    {
+      "epoch": 0.3889845094664372,
+      "grad_norm": 6.694453613894155,
+      "learning_rate": 1e-06,
+      "loss": 0.6246,
+      "step": 4979
+    },
+    {
+      "epoch": 0.3924268502581756,
+      "grad_norm": 5.0854407647726925,
+      "learning_rate": 1e-06,
+      "loss": 0.6394,
+      "step": 4980
+    },
+    {
+      "epoch": 0.3958691910499139,
+      "grad_norm": 7.33871620327319,
+      "learning_rate": 1e-06,
+      "loss": 0.6827,
+      "step": 4981
+    },
+    {
+      "epoch": 0.3993115318416523,
+      "grad_norm": 6.620767645766533,
+      "learning_rate": 1e-06,
+      "loss": 0.674,
+      "step": 4982
+    },
+    {
+      "epoch": 0.4027538726333907,
+      "grad_norm": 5.912241238025216,
+      "learning_rate": 1e-06,
+      "loss": 0.6761,
+      "step": 4983
+    },
+    {
+      "epoch": 0.40619621342512907,
+      "grad_norm": 5.447009306387902,
+      "learning_rate": 1e-06,
+      "loss": 0.6408,
+      "step": 4984
+    },
+    {
+      "epoch": 0.40963855421686746,
+      "grad_norm": 5.631325997819541,
+      "learning_rate": 1e-06,
+      "loss": 0.599,
+      "step": 4985
+    },
+    {
+      "epoch": 0.41308089500860584,
+      "grad_norm": 5.93213298921337,
+      "learning_rate": 1e-06,
+      "loss": 0.6567,
+      "step": 4986
+    },
+    {
+      "epoch": 0.4165232358003442,
+      "grad_norm": 7.86633391594967,
+      "learning_rate": 1e-06,
+      "loss": 0.6459,
+      "step": 4987
+    },
+    {
+      "epoch": 0.4199655765920826,
+      "grad_norm": 7.873752322757984,
+      "learning_rate": 1e-06,
+      "loss": 0.619,
+      "step": 4988
+    },
+    {
+      "epoch": 0.423407917383821,
+      "grad_norm": 6.86254592458152,
+      "learning_rate": 1e-06,
+      "loss": 0.6739,
+      "step": 4989
+    },
+    {
+      "epoch": 0.4268502581755594,
+      "grad_norm": 5.972260439862256,
+      "learning_rate": 1e-06,
+      "loss": 0.6407,
+      "step": 4990
+    },
+    {
+      "epoch": 0.43029259896729777,
+      "grad_norm": 7.3065077063811605,
+      "learning_rate": 1e-06,
+      "loss": 0.6497,
+      "step": 4991
+    },
+    {
+      "epoch": 0.43373493975903615,
+      "grad_norm": 6.151172398996421,
+      "learning_rate": 1e-06,
+      "loss": 0.6075,
+      "step": 4992
+    },
+    {
+      "epoch": 0.43717728055077454,
+      "grad_norm": 8.01527979156912,
+      "learning_rate": 1e-06,
+      "loss": 0.7141,
+      "step": 4993
+    },
+    {
+      "epoch": 0.4406196213425129,
+      "grad_norm": 8.01527979156912,
+      "learning_rate": 1e-06,
+      "loss": 0.6522,
+      "step": 4994
+    },
+    {
+      "epoch": 0.4440619621342513,
+      "grad_norm": 7.168438608482569,
+      "learning_rate": 1e-06,
+      "loss": 0.6579,
+      "step": 4995
+    },
+    {
+      "epoch": 0.4475043029259897,
+      "grad_norm": 6.521812086589511,
+      "learning_rate": 1e-06,
+      "loss": 0.6402,
+      "step": 4996
+    },
+    {
+      "epoch": 0.4509466437177281,
+      "grad_norm": 6.421839811337142,
+      "learning_rate": 1e-06,
+      "loss": 0.732,
+      "step": 4997
+    },
+    {
+      "epoch": 0.45438898450946646,
+      "grad_norm": 8.082526171830526,
+      "learning_rate": 1e-06,
+      "loss": 0.6908,
+      "step": 4998
+    },
+    {
+      "epoch": 0.4578313253012048,
+      "grad_norm": 5.24398707290673,
+      "learning_rate": 1e-06,
+      "loss": 0.6294,
+      "step": 4999
+    },
+    {
+      "epoch": 0.4612736660929432,
+      "grad_norm": 6.185001205188784,
+      "learning_rate": 1e-06,
+      "loss": 0.6289,
+      "step": 5000
+    },
+    {
+      "epoch": 0.4612736660929432,
+      "eval_loss": 2.5163192749023438,
+      "eval_runtime": 21.7098,
+      "eval_samples_per_second": 46.062,
+      "eval_steps_per_second": 5.758,
+      "step": 5000
+    },
+    {
+      "epoch": 0.46471600688468157,
+      "grad_norm": 7.185363403412653,
+      "learning_rate": 1e-06,
+      "loss": 0.6284,
+      "step": 5001
+    },
+    {
+      "epoch": 0.46815834767641995,
+      "grad_norm": 5.864288974221786,
+      "learning_rate": 1e-06,
+      "loss": 0.6231,
+      "step": 5002
+    },
+    {
+      "epoch": 0.47160068846815834,
+      "grad_norm": 6.4563617434206675,
+      "learning_rate": 1e-06,
+      "loss": 0.6607,
+      "step": 5003
+    },
+    {
+      "epoch": 0.4750430292598967,
+      "grad_norm": 6.30189277982893,
+      "learning_rate": 1e-06,
+      "loss": 0.6741,
+      "step": 5004
+    },
+    {
+      "epoch": 0.4784853700516351,
+      "grad_norm": 6.756874783903635,
+      "learning_rate": 1e-06,
+      "loss": 0.6646,
+      "step": 5005
+    },
+    {
+      "epoch": 0.4819277108433735,
+      "grad_norm": 7.784546376403721,
+      "learning_rate": 1e-06,
+      "loss": 0.6644,
+      "step": 5006
+    },
+    {
+      "epoch": 0.4853700516351119,
+      "grad_norm": 6.0258188006670155,
+      "learning_rate": 1e-06,
+      "loss": 0.681,
+      "step": 5007
+    },
+    {
+      "epoch": 0.48881239242685026,
+      "grad_norm": 7.933490685878364,
+      "learning_rate": 1e-06,
+      "loss": 0.6575,
+      "step": 5008
+    },
+    {
+      "epoch": 0.49225473321858865,
+      "grad_norm": 6.019955941397438,
+      "learning_rate": 1e-06,
+      "loss": 0.6823,
+      "step": 5009
+    },
+    {
+      "epoch": 0.49569707401032703,
+      "grad_norm": 7.504231260465043,
+      "learning_rate": 1e-06,
+      "loss": 0.669,
+      "step": 5010
+    },
+    {
+      "epoch": 0.4991394148020654,
+      "grad_norm": 6.003307949784288,
+      "learning_rate": 1e-06,
+      "loss": 0.6592,
+      "step": 5011
+    },
+    {
+      "epoch": 0.5025817555938038,
+      "grad_norm": 6.474010368150935,
+      "learning_rate": 1e-06,
+      "loss": 0.6556,
+      "step": 5012
+    },
+    {
+      "epoch": 0.5060240963855421,
+      "grad_norm": 7.709369109832686,
+      "learning_rate": 1e-06,
+      "loss": 0.6571,
+      "step": 5013
+    },
+    {
+      "epoch": 0.5094664371772806,
+      "grad_norm": 6.034151592347818,
+      "learning_rate": 1e-06,
+      "loss": 0.6762,
+      "step": 5014
+    },
+    {
+      "epoch": 0.5129087779690189,
+      "grad_norm": 5.031408776100482,
+      "learning_rate": 1e-06,
+      "loss": 0.634,
+      "step": 5015
+    },
+    {
+      "epoch": 0.5163511187607573,
+      "grad_norm": 7.434121468240524,
+      "learning_rate": 1e-06,
+      "loss": 0.6556,
+      "step": 5016
+    },
+    {
+      "epoch": 0.5197934595524957,
+      "grad_norm": 6.858454846737196,
+      "learning_rate": 1e-06,
+      "loss": 0.682,
+      "step": 5017
+    },
+    {
+      "epoch": 0.5232358003442341,
+      "grad_norm": 7.141180068249861,
+      "learning_rate": 1e-06,
+      "loss": 0.6743,
+      "step": 5018
+    },
+    {
+      "epoch": 0.5266781411359724,
+      "grad_norm": 7.836376306230385,
+      "learning_rate": 1e-06,
+      "loss": 0.6655,
+      "step": 5019
+    },
+    {
+      "epoch": 0.5301204819277109,
+      "grad_norm": 7.775621588130858,
+      "learning_rate": 1e-06,
+      "loss": 0.6947,
+      "step": 5020
+    },
+    {
+      "epoch": 0.5335628227194492,
+      "grad_norm": 8.39012924561531,
+      "learning_rate": 1e-06,
+      "loss": 0.6811,
+      "step": 5021
+    },
+    {
+      "epoch": 0.5370051635111877,
+      "grad_norm": 7.1610569193154445,
+      "learning_rate": 1e-06,
+      "loss": 0.6766,
+      "step": 5022
+    },
+    {
+      "epoch": 0.540447504302926,
+      "grad_norm": 7.618659492499224,
+      "learning_rate": 1e-06,
+      "loss": 0.6297,
+      "step": 5023
+    },
+    {
+      "epoch": 0.5438898450946644,
+      "grad_norm": 8.331662064008338,
+      "learning_rate": 1e-06,
+      "loss": 0.6802,
+      "step": 5024
+    },
+    {
+      "epoch": 0.5473321858864028,
+      "grad_norm": 7.244027621168721,
+      "learning_rate": 1e-06,
+      "loss": 0.6905,
+      "step": 5025
+    },
+    {
+      "epoch": 0.5507745266781411,
+      "grad_norm": 8.387813861025874,
+      "learning_rate": 1e-06,
+      "loss": 0.6346,
+      "step": 5026
+    },
+    {
+      "epoch": 0.5542168674698795,
+      "grad_norm": 6.214160525392823,
+      "learning_rate": 1e-06,
+      "loss": 0.6446,
+      "step": 5027
+    },
+    {
+      "epoch": 0.5576592082616179,
+      "grad_norm": 9.087105041256779,
+      "learning_rate": 1e-06,
+      "loss": 0.6599,
+      "step": 5028
+    },
+    {
+      "epoch": 0.5611015490533563,
+      "grad_norm": 6.60515936332129,
+      "learning_rate": 1e-06,
+      "loss": 0.686,
+      "step": 5029
+    },
+    {
+      "epoch": 0.5645438898450946,
+      "grad_norm": 4.7867305437241265,
+      "learning_rate": 1e-06,
+      "loss": 0.6708,
+      "step": 5030
+    },
+    {
+      "epoch": 0.5679862306368331,
+      "grad_norm": 7.590763289389431,
+      "learning_rate": 1e-06,
+      "loss": 0.6602,
+      "step": 5031
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 8.33169422604525,
+      "learning_rate": 1e-06,
+      "loss": 0.639,
+      "step": 5032
+    },
+    {
+      "epoch": 0.5748709122203098,
+      "grad_norm": 6.313284761179127,
+      "learning_rate": 1e-06,
+      "loss": 0.7141,
+      "step": 5033
+    },
+    {
+      "epoch": 0.5783132530120482,
+      "grad_norm": 5.363493311362969,
+      "learning_rate": 1e-06,
+      "loss": 0.6432,
+      "step": 5034
+    },
+    {
+      "epoch": 0.5817555938037866,
+      "grad_norm": 6.752053391147588,
+      "learning_rate": 1e-06,
+      "loss": 0.6868,
+      "step": 5035
+    },
+    {
+      "epoch": 0.5851979345955249,
+      "grad_norm": 7.13043572286772,
+      "learning_rate": 1e-06,
+      "loss": 0.6728,
+      "step": 5036
+    },
+    {
+      "epoch": 0.5886402753872634,
+      "grad_norm": 7.783064836732784,
+      "learning_rate": 1e-06,
+      "loss": 0.6473,
+      "step": 5037
+    },
+    {
+      "epoch": 0.5920826161790017,
+      "grad_norm": 7.24549295956451,
+      "learning_rate": 1e-06,
+      "loss": 0.6628,
+      "step": 5038
+    },
+    {
+      "epoch": 0.5955249569707401,
+      "grad_norm": 7.190392313145377,
+      "learning_rate": 1e-06,
+      "loss": 0.6394,
+      "step": 5039
+    },
+    {
+      "epoch": 0.5989672977624785,
+      "grad_norm": 7.07369734157358,
+      "learning_rate": 1e-06,
+      "loss": 0.6409,
+      "step": 5040
+    },
+    {
+      "epoch": 0.6024096385542169,
+      "grad_norm": 4.98947386120797,
+      "learning_rate": 1e-06,
+      "loss": 0.6737,
+      "step": 5041
+    },
+    {
+      "epoch": 0.6058519793459552,
+      "grad_norm": 6.275255955198839,
+      "learning_rate": 1e-06,
+      "loss": 0.6585,
+      "step": 5042
+    },
+    {
+      "epoch": 0.6092943201376936,
+      "grad_norm": 6.707993275250671,
+      "learning_rate": 1e-06,
+      "loss": 0.6064,
+      "step": 5043
+    },
+    {
+      "epoch": 0.612736660929432,
+      "grad_norm": 6.688603629240446,
+      "learning_rate": 1e-06,
+      "loss": 0.6589,
+      "step": 5044
+    },
+    {
+      "epoch": 0.6161790017211703,
+      "grad_norm": 6.324423451494869,
+      "learning_rate": 1e-06,
+      "loss": 0.6532,
+      "step": 5045
+    },
+    {
+      "epoch": 0.6196213425129088,
+      "grad_norm": 5.8071796575715116,
+      "learning_rate": 1e-06,
+      "loss": 0.6701,
+      "step": 5046
+    },
+    {
+      "epoch": 0.6230636833046471,
+      "grad_norm": 9.530776900965279,
+      "learning_rate": 1e-06,
+      "loss": 0.6517,
+      "step": 5047
+    },
+    {
+      "epoch": 0.6265060240963856,
+      "grad_norm": 6.33770155223311,
+      "learning_rate": 1e-06,
+      "loss": 0.6235,
+      "step": 5048
+    },
+    {
+      "epoch": 0.6299483648881239,
+      "grad_norm": 8.635748627603208,
+      "learning_rate": 1e-06,
+      "loss": 0.5851,
+      "step": 5049
+    },
+    {
+      "epoch": 0.6333907056798623,
+      "grad_norm": 7.85770077222489,
+      "learning_rate": 1e-06,
+      "loss": 0.6858,
+      "step": 5050
+    },
+    {
+      "epoch": 0.6368330464716007,
+      "grad_norm": 7.412775523799055,
+      "learning_rate": 1e-06,
+      "loss": 0.6456,
+      "step": 5051
+    },
+    {
+      "epoch": 0.6402753872633391,
+      "grad_norm": 6.34245209823316,
+      "learning_rate": 1e-06,
+      "loss": 0.6456,
+      "step": 5052
+    },
+    {
+      "epoch": 0.6437177280550774,
+      "grad_norm": 8.22078556793561,
+      "learning_rate": 1e-06,
+      "loss": 0.6165,
+      "step": 5053
+    },
+    {
+      "epoch": 0.6471600688468159,
+      "grad_norm": 7.93429997513278,
+      "learning_rate": 1e-06,
+      "loss": 0.701,
+      "step": 5054
+    },
+    {
+      "epoch": 0.6506024096385542,
+      "grad_norm": 6.658679438533953,
+      "learning_rate": 1e-06,
+      "loss": 0.6459,
+      "step": 5055
+    },
+    {
+      "epoch": 0.6540447504302926,
+      "grad_norm": 7.60205183349453,
+      "learning_rate": 1e-06,
+      "loss": 0.5658,
+      "step": 5056
+    },
+    {
+      "epoch": 0.657487091222031,
+      "grad_norm": 8.149749649632488,
+      "learning_rate": 1e-06,
+      "loss": 0.6544,
+      "step": 5057
+    },
+    {
+      "epoch": 0.6609294320137694,
+      "grad_norm": 6.4224032421405255,
+      "learning_rate": 1e-06,
+      "loss": 0.6555,
+      "step": 5058
+    },
+    {
+      "epoch": 0.6643717728055077,
+      "grad_norm": 7.186849450462647,
+      "learning_rate": 1e-06,
+      "loss": 0.6385,
+      "step": 5059
+    },
+    {
+      "epoch": 0.6678141135972461,
+      "grad_norm": 6.838115036886858,
+      "learning_rate": 1e-06,
+      "loss": 0.6676,
+      "step": 5060
+    },
+    {
+      "epoch": 0.6712564543889845,
+      "grad_norm": 7.485995021323868,
+      "learning_rate": 1e-06,
+      "loss": 0.5867,
+      "step": 5061
+    },
+    {
+      "epoch": 0.6746987951807228,
+      "grad_norm": 6.770173037542965,
+      "learning_rate": 1e-06,
+      "loss": 0.6511,
+      "step": 5062
+    },
+    {
+      "epoch": 0.6781411359724613,
+      "grad_norm": 7.214354572812459,
+      "learning_rate": 1e-06,
+      "loss": 0.6591,
+      "step": 5063
+    },
+    {
+      "epoch": 0.6815834767641996,
+      "grad_norm": 7.120116867101523,
+      "learning_rate": 1e-06,
+      "loss": 0.655,
+      "step": 5064
+    },
+    {
+      "epoch": 0.685025817555938,
+      "grad_norm": 6.434124842658215,
+      "learning_rate": 1e-06,
+      "loss": 0.6693,
+      "step": 5065
+    },
+    {
+      "epoch": 0.6884681583476764,
+      "grad_norm": 8.577837363111517,
+      "learning_rate": 1e-06,
+      "loss": 0.665,
+      "step": 5066
+    },
+    {
+      "epoch": 0.6919104991394148,
+      "grad_norm": 5.914862873446635,
+      "learning_rate": 1e-06,
+      "loss": 0.6544,
+      "step": 5067
+    },
+    {
+      "epoch": 0.6953528399311532,
+      "grad_norm": 6.226728555540829,
+      "learning_rate": 1e-06,
+      "loss": 0.6587,
+      "step": 5068
+    },
+    {
+      "epoch": 0.6987951807228916,
+      "grad_norm": 6.034306698240086,
+      "learning_rate": 1e-06,
+      "loss": 0.663,
+      "step": 5069
+    },
+    {
+      "epoch": 0.7022375215146299,
+      "grad_norm": 6.219906170984133,
+      "learning_rate": 1e-06,
+      "loss": 0.6421,
+      "step": 5070
+    },
+    {
+      "epoch": 0.7056798623063684,
+      "grad_norm": 6.2546354969305025,
+      "learning_rate": 1e-06,
+      "loss": 0.632,
+      "step": 5071
+    },
+    {
+      "epoch": 0.7091222030981067,
+      "grad_norm": 7.948456589494676,
+      "learning_rate": 1e-06,
+      "loss": 0.6722,
+      "step": 5072
+    },
+    {
+      "epoch": 0.7125645438898451,
+      "grad_norm": 7.590904724614749,
+      "learning_rate": 1e-06,
+      "loss": 0.692,
+      "step": 5073
+    },
+    {
+      "epoch": 0.7160068846815835,
+      "grad_norm": 7.579846271050383,
+      "learning_rate": 1e-06,
+      "loss": 0.6174,
+      "step": 5074
+    },
+    {
+      "epoch": 0.7194492254733219,
+      "grad_norm": 7.2663612559408115,
+      "learning_rate": 1e-06,
+      "loss": 0.6719,
+      "step": 5075
+    },
+    {
+      "epoch": 0.7228915662650602,
+      "grad_norm": 6.389602774704022,
+      "learning_rate": 1e-06,
+      "loss": 0.6602,
+      "step": 5076
+    },
+    {
+      "epoch": 0.7263339070567987,
+      "grad_norm": 8.865778621036679,
+      "learning_rate": 1e-06,
+      "loss": 0.6751,
+      "step": 5077
+    },
+    {
+      "epoch": 0.729776247848537,
+      "grad_norm": 6.137522118893293,
+      "learning_rate": 1e-06,
+      "loss": 0.6659,
+      "step": 5078
+    },
+    {
+      "epoch": 0.7332185886402753,
+      "grad_norm": 7.528493920403535,
+      "learning_rate": 1e-06,
+      "loss": 0.6746,
+      "step": 5079
+    },
+    {
+      "epoch": 0.7366609294320138,
+      "grad_norm": 6.973588531412749,
+      "learning_rate": 1e-06,
+      "loss": 0.6719,
+      "step": 5080
+    },
+    {
+      "epoch": 0.7401032702237521,
+      "grad_norm": 6.834460481003578,
+      "learning_rate": 1e-06,
+      "loss": 0.6244,
+      "step": 5081
+    },
+    {
+      "epoch": 0.7435456110154905,
+      "grad_norm": 6.423048033840628,
+      "learning_rate": 1e-06,
+      "loss": 0.6953,
+      "step": 5082
+    },
+    {
+      "epoch": 0.7469879518072289,
+      "grad_norm": 5.358185494496734,
+      "learning_rate": 1e-06,
+      "loss": 0.6645,
+      "step": 5083
+    },
+    {
+      "epoch": 0.7504302925989673,
+      "grad_norm": 7.537240931651455,
+      "learning_rate": 1e-06,
+      "loss": 0.6505,
+      "step": 5084
+    },
+    {
+      "epoch": 0.7538726333907056,
+      "grad_norm": 6.774409050707278,
+      "learning_rate": 1e-06,
+      "loss": 0.6722,
+      "step": 5085
+    },
+    {
+      "epoch": 0.7573149741824441,
+      "grad_norm": 7.219486081415032,
+      "learning_rate": 1e-06,
+      "loss": 0.6759,
+      "step": 5086
+    },
+    {
+      "epoch": 0.7607573149741824,
+      "grad_norm": 9.235752420934327,
+      "learning_rate": 1e-06,
+      "loss": 0.6914,
+      "step": 5087
+    },
+    {
+      "epoch": 0.7641996557659209,
+      "grad_norm": 6.205200524519897,
+      "learning_rate": 1e-06,
+      "loss": 0.6828,
+      "step": 5088
+    },
+    {
+      "epoch": 0.7676419965576592,
+      "grad_norm": 6.738601707745915,
+      "learning_rate": 1e-06,
+      "loss": 0.7192,
+      "step": 5089
+    },
+    {
+      "epoch": 0.7710843373493976,
+      "grad_norm": 7.006315364739306,
+      "learning_rate": 1e-06,
+      "loss": 0.6648,
+      "step": 5090
+    },
+    {
+      "epoch": 0.774526678141136,
+      "grad_norm": 7.004975013340174,
+      "learning_rate": 1e-06,
+      "loss": 0.6447,
+      "step": 5091
+    },
+    {
+      "epoch": 0.7779690189328744,
+      "grad_norm": 6.4402018554479845,
+      "learning_rate": 1e-06,
+      "loss": 0.6775,
+      "step": 5092
+    },
+    {
+      "epoch": 0.7814113597246127,
+      "grad_norm": 6.519354150552599,
+      "learning_rate": 1e-06,
+      "loss": 0.6539,
+      "step": 5093
+    },
+    {
+      "epoch": 0.7848537005163512,
+      "grad_norm": 7.857394555105048,
+      "learning_rate": 1e-06,
+      "loss": 0.6467,
+      "step": 5094
+    },
+    {
+      "epoch": 0.7882960413080895,
+      "grad_norm": 7.423167481535682,
+      "learning_rate": 1e-06,
+      "loss": 0.664,
+      "step": 5095
+    },
+    {
+      "epoch": 0.7917383820998278,
+      "grad_norm": 7.904184094416325,
+      "learning_rate": 1e-06,
+      "loss": 0.6738,
+      "step": 5096
+    },
+    {
+      "epoch": 0.7951807228915663,
+      "grad_norm": 6.883529521125306,
+      "learning_rate": 1e-06,
+      "loss": 0.7122,
+      "step": 5097
+    },
+    {
+      "epoch": 0.7986230636833046,
+      "grad_norm": 6.278938386859969,
+      "learning_rate": 1e-06,
+      "loss": 0.7261,
+      "step": 5098
+    },
+    {
+      "epoch": 0.802065404475043,
+      "grad_norm": 6.355650314035015,
+      "learning_rate": 1e-06,
+      "loss": 0.6423,
+      "step": 5099
+    },
+    {
+      "epoch": 0.8055077452667814,
+      "grad_norm": 5.368099320795738,
+      "learning_rate": 1e-06,
+      "loss": 0.6602,
+      "step": 5100
+    },
+    {
+      "epoch": 0.8089500860585198,
+      "grad_norm": 7.012430113820998,
+      "learning_rate": 1e-06,
+      "loss": 0.6725,
+      "step": 5101
+    },
+    {
+      "epoch": 0.8123924268502581,
+      "grad_norm": 5.930268332028906,
+      "learning_rate": 1e-06,
+      "loss": 0.6348,
+      "step": 5102
+    },
+    {
+      "epoch": 0.8158347676419966,
+      "grad_norm": 6.330027370190921,
+      "learning_rate": 1e-06,
+      "loss": 0.6595,
+      "step": 5103
+    },
+    {
+      "epoch": 0.8192771084337349,
+      "grad_norm": 7.420155679815863,
+      "learning_rate": 1e-06,
+      "loss": 0.6547,
+      "step": 5104
+    },
+    {
+      "epoch": 0.8227194492254734,
+      "grad_norm": 7.4646277457008745,
+      "learning_rate": 1e-06,
+      "loss": 0.6894,
+      "step": 5105
+    },
+    {
+      "epoch": 0.8261617900172117,
+      "grad_norm": 7.0235540246221415,
+      "learning_rate": 1e-06,
+      "loss": 0.6338,
+      "step": 5106
+    },
+    {
+      "epoch": 0.8296041308089501,
+      "grad_norm": 5.810057979259591,
+      "learning_rate": 1e-06,
+      "loss": 0.6625,
+      "step": 5107
+    },
+    {
+      "epoch": 0.8330464716006885,
+      "grad_norm": 6.504015323795398,
+      "learning_rate": 1e-06,
+      "loss": 0.6696,
+      "step": 5108
+    },
+    {
+      "epoch": 0.8364888123924269,
+      "grad_norm": 6.700485648278997,
+      "learning_rate": 1e-06,
+      "loss": 0.6603,
+      "step": 5109
+    },
+    {
+      "epoch": 0.8399311531841652,
+      "grad_norm": 5.910163165828733,
+      "learning_rate": 1e-06,
+      "loss": 0.622,
+      "step": 5110
+    },
+    {
+      "epoch": 0.8433734939759037,
+      "grad_norm": 5.871889993993969,
+      "learning_rate": 1e-06,
+      "loss": 0.6539,
+      "step": 5111
+    },
+    {
+      "epoch": 0.846815834767642,
+      "grad_norm": 6.845368672277225,
+      "learning_rate": 1e-06,
+      "loss": 0.6859,
+      "step": 5112
+    },
+    {
+      "epoch": 0.8502581755593803,
+      "grad_norm": 8.230728617254746,
+      "learning_rate": 1e-06,
+      "loss": 0.6526,
+      "step": 5113
+    },
+    {
+      "epoch": 0.8537005163511188,
+      "grad_norm": 6.274812450786174,
+      "learning_rate": 1e-06,
+      "loss": 0.6527,
+      "step": 5114
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 5.398978917561708,
+      "learning_rate": 1e-06,
+      "loss": 0.6333,
+      "step": 5115
+    },
+    {
+      "epoch": 0.8605851979345955,
+      "grad_norm": 6.9502700171784655,
+      "learning_rate": 1e-06,
+      "loss": 0.6782,
+      "step": 5116
+    },
+    {
+      "epoch": 0.8640275387263339,
+      "grad_norm": 5.392708238394533,
+      "learning_rate": 1e-06,
+      "loss": 0.6214,
+      "step": 5117
+    },
+    {
+      "epoch": 0.8674698795180723,
+      "grad_norm": 7.1382657034768355,
+      "learning_rate": 1e-06,
+      "loss": 0.6164,
+      "step": 5118
+    },
+    {
+      "epoch": 0.8709122203098106,
+      "grad_norm": 5.594006762519265,
+      "learning_rate": 1e-06,
+      "loss": 0.6478,
+      "step": 5119
+    },
+    {
+      "epoch": 0.8743545611015491,
+      "grad_norm": 7.07686321409344,
+      "learning_rate": 1e-06,
+      "loss": 0.6033,
+      "step": 5120
+    },
+    {
+      "epoch": 0.8777969018932874,
+      "grad_norm": 7.540609408898925,
+      "learning_rate": 1e-06,
+      "loss": 0.6192,
+      "step": 5121
+    },
+    {
+      "epoch": 0.8812392426850258,
+      "grad_norm": 7.865631397438872,
+      "learning_rate": 1e-06,
+      "loss": 0.6462,
+      "step": 5122
+    },
+    {
+      "epoch": 0.8846815834767642,
+      "grad_norm": 6.384764719662087,
+      "learning_rate": 1e-06,
+      "loss": 0.6763,
+      "step": 5123
+    },
+    {
+      "epoch": 0.8881239242685026,
+      "grad_norm": 7.6059105946780665,
+      "learning_rate": 1e-06,
+      "loss": 0.6399,
+      "step": 5124
+    },
+    {
+      "epoch": 0.891566265060241,
+      "grad_norm": 7.86532226092865,
+      "learning_rate": 1e-06,
+      "loss": 0.622,
+      "step": 5125
+    },
+    {
+      "epoch": 0.8950086058519794,
+      "grad_norm": 7.481905190610532,
+      "learning_rate": 1e-06,
+      "loss": 0.6189,
+      "step": 5126
+    },
+    {
+      "epoch": 0.8984509466437177,
+      "grad_norm": 7.175598036172626,
+      "learning_rate": 1e-06,
+      "loss": 0.6832,
+      "step": 5127
+    },
+    {
+      "epoch": 0.9018932874354562,
+      "grad_norm": 6.617686031341074,
+      "learning_rate": 1e-06,
+      "loss": 0.6915,
+      "step": 5128
+    },
+    {
+      "epoch": 0.9053356282271945,
+      "grad_norm": 6.500031118681312,
+      "learning_rate": 1e-06,
+      "loss": 0.6474,
+      "step": 5129
+    },
+    {
+      "epoch": 0.9087779690189329,
+      "grad_norm": 5.7817054849407565,
+      "learning_rate": 1e-06,
+      "loss": 0.5957,
+      "step": 5130
+    },
+    {
+      "epoch": 0.9122203098106713,
+      "grad_norm": 7.04517563981561,
+      "learning_rate": 1e-06,
+      "loss": 0.6683,
+      "step": 5131
+    },
+    {
+      "epoch": 0.9156626506024096,
+      "grad_norm": 7.550252631604601,
+      "learning_rate": 1e-06,
+      "loss": 0.6726,
+      "step": 5132
+    },
+    {
+      "epoch": 0.919104991394148,
+      "grad_norm": 6.3321350016516575,
+      "learning_rate": 1e-06,
+      "loss": 0.6559,
+      "step": 5133
+    },
+    {
+      "epoch": 0.9225473321858864,
+      "grad_norm": 6.053630615309342,
+      "learning_rate": 1e-06,
+      "loss": 0.6622,
+      "step": 5134
+    },
+    {
+      "epoch": 0.9259896729776248,
+      "grad_norm": 6.7714031700269945,
+      "learning_rate": 1e-06,
+      "loss": 0.6827,
+      "step": 5135
+    },
+    {
+      "epoch": 0.9294320137693631,
+      "grad_norm": 8.349817829457914,
+      "learning_rate": 1e-06,
+      "loss": 0.636,
+      "step": 5136
+    },
+    {
+      "epoch": 0.9328743545611016,
+      "grad_norm": 7.188441033201691,
+      "learning_rate": 1e-06,
+      "loss": 0.6567,
+      "step": 5137
+    },
+    {
+      "epoch": 0.9363166953528399,
+      "grad_norm": 6.7710226430260345,
+      "learning_rate": 1e-06,
+      "loss": 0.6433,
+      "step": 5138
+    },
+    {
+      "epoch": 0.9397590361445783,
+      "grad_norm": 8.120469859528134,
+      "learning_rate": 1e-06,
+      "loss": 0.6822,
+      "step": 5139
+    },
+    {
+      "epoch": 0.9432013769363167,
+      "grad_norm": 6.812770290406592,
+      "learning_rate": 1e-06,
+      "loss": 0.6572,
+      "step": 5140
+    },
+    {
+      "epoch": 0.9466437177280551,
+      "grad_norm": 7.263767740713461,
+      "learning_rate": 1e-06,
+      "loss": 0.6469,
+      "step": 5141
+    },
+    {
+      "epoch": 0.9500860585197934,
+      "grad_norm": 6.205822621218068,
+      "learning_rate": 1e-06,
+      "loss": 0.6272,
+      "step": 5142
+    },
+    {
+      "epoch": 0.9535283993115319,
+      "grad_norm": 8.789525611007884,
+      "learning_rate": 1e-06,
+      "loss": 0.7007,
+      "step": 5143
+    },
+    {
+      "epoch": 0.9569707401032702,
+      "grad_norm": 8.191509099336198,
+      "learning_rate": 1e-06,
+      "loss": 0.6747,
+      "step": 5144
+    },
+    {
+      "epoch": 0.9604130808950087,
+      "grad_norm": 7.400404007438004,
+      "learning_rate": 1e-06,
+      "loss": 0.6395,
+      "step": 5145
+    },
+    {
+      "epoch": 0.963855421686747,
+      "grad_norm": 5.356429489904784,
+      "learning_rate": 1e-06,
+      "loss": 0.7301,
+      "step": 5146
+    },
+    {
+      "epoch": 0.9672977624784854,
+      "grad_norm": 6.502019202142468,
+      "learning_rate": 1e-06,
+      "loss": 0.5926,
+      "step": 5147
+    },
+    {
+      "epoch": 0.9707401032702238,
+      "grad_norm": 5.237367360714358,
+      "learning_rate": 1e-06,
+      "loss": 0.6283,
+      "step": 5148
+    },
+    {
+      "epoch": 0.9741824440619621,
+      "grad_norm": 6.42383567186059,
+      "learning_rate": 1e-06,
+      "loss": 0.6513,
+      "step": 5149
+    },
+    {
+      "epoch": 0.9776247848537005,
+      "grad_norm": 7.215279664323738,
+      "learning_rate": 1e-06,
+      "loss": 0.6444,
+      "step": 5150
+    },
+    {
+      "epoch": 0.9810671256454389,
+      "grad_norm": 8.655218883294271,
+      "learning_rate": 1e-06,
+      "loss": 0.6438,
+      "step": 5151
+    },
+    {
+      "epoch": 0.9845094664371773,
+      "grad_norm": 5.675937050998256,
+      "learning_rate": 1e-06,
+      "loss": 0.6577,
+      "step": 5152
+    },
+    {
+      "epoch": 0.9879518072289156,
+      "grad_norm": 8.63763987599962,
+      "learning_rate": 1e-06,
+      "loss": 0.6987,
+      "step": 5153
+    },
+    {
+      "epoch": 0.9913941480206541,
+      "grad_norm": 8.774060117726597,
+      "learning_rate": 1e-06,
+      "loss": 0.6925,
+      "step": 5154
+    },
+    {
+      "epoch": 0.9948364888123924,
+      "grad_norm": 6.493969085227579,
+      "learning_rate": 1e-06,
+      "loss": 0.6641,
+      "step": 5155
+    },
+    {
+      "epoch": 0.9982788296041308,
+      "grad_norm": 7.822799448515648,
+      "learning_rate": 1e-06,
+      "loss": 0.6382,
+      "step": 5156
+    },
+    {
+      "epoch": 0.9982788296041308,
+      "eval_loss": 2.6471035480499268,
+      "eval_runtime": 21.4172,
+      "eval_samples_per_second": 46.691,
+      "eval_steps_per_second": 5.836,
+      "step": 5156
+    },
+    {
+      "epoch": 0.0022522522522522522,
+      "grad_norm": 6.176686188005488,
+      "learning_rate": 1e-06,
+      "loss": 0.3541,
+      "step": 5157
+    },
+    {
+      "epoch": 0.0045045045045045045,
+      "grad_norm": 6.253146461227654,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 5158
+    },
+    {
+      "epoch": 0.006756756756756757,
+      "grad_norm": 4.871658556638364,
+      "learning_rate": 1e-06,
+      "loss": 0.3603,
+      "step": 5159
+    },
+    {
+      "epoch": 0.009009009009009009,
+      "grad_norm": 6.962327855411496,
+      "learning_rate": 1e-06,
+      "loss": 0.3269,
+      "step": 5160
+    },
+    {
+      "epoch": 0.01126126126126126,
+      "grad_norm": 5.85832243250038,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 5161
+    },
+    {
+      "epoch": 0.013513513513513514,
+      "grad_norm": 7.975281875938298,
+      "learning_rate": 1e-06,
+      "loss": 0.3161,
+      "step": 5162
+    },
+    {
+      "epoch": 0.015765765765765764,
+      "grad_norm": 5.83367382361181,
+      "learning_rate": 1e-06,
+      "loss": 0.3192,
+      "step": 5163
+    },
+    {
+      "epoch": 0.018018018018018018,
+      "grad_norm": 8.078740482548014,
+      "learning_rate": 1e-06,
+      "loss": 0.3342,
+      "step": 5164
+    },
+    {
+      "epoch": 0.02027027027027027,
+      "grad_norm": 7.473579044757994,
+      "learning_rate": 1e-06,
+      "loss": 0.3468,
+      "step": 5165
+    },
+    {
+      "epoch": 0.02252252252252252,
+      "grad_norm": 9.654159814056829,
+      "learning_rate": 1e-06,
+      "loss": 0.3383,
+      "step": 5166
+    },
+    {
+      "epoch": 0.024774774774774775,
+      "grad_norm": 9.313158911993831,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 5167
+    },
+    {
+      "epoch": 0.02702702702702703,
+      "grad_norm": 11.217747147913624,
+      "learning_rate": 1e-06,
+      "loss": 0.3418,
+      "step": 5168
+    },
+    {
+      "epoch": 0.02927927927927928,
+      "grad_norm": 8.458533719388216,
+      "learning_rate": 1e-06,
+      "loss": 0.3079,
+      "step": 5169
+    },
+    {
+      "epoch": 0.03153153153153153,
+      "grad_norm": 8.067671252957467,
+      "learning_rate": 1e-06,
+      "loss": 0.3348,
+      "step": 5170
+    },
+    {
+      "epoch": 0.033783783783783786,
+      "grad_norm": 8.335459045483468,
+      "learning_rate": 1e-06,
+      "loss": 0.3219,
+      "step": 5171
+    },
+    {
+      "epoch": 0.036036036036036036,
+      "grad_norm": 10.731180860426202,
+      "learning_rate": 1e-06,
+      "loss": 0.3433,
+      "step": 5172
+    },
+    {
+      "epoch": 0.038288288288288286,
+      "grad_norm": 7.936039267366813,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 5173
+    },
+    {
+      "epoch": 0.04054054054054054,
+      "grad_norm": 9.467097295781274,
+      "learning_rate": 1e-06,
+      "loss": 0.3428,
+      "step": 5174
+    },
+    {
+      "epoch": 0.04279279279279279,
+      "grad_norm": 7.479784602902723,
+      "learning_rate": 1e-06,
+      "loss": 0.3029,
+      "step": 5175
+    },
+    {
+      "epoch": 0.04504504504504504,
+      "grad_norm": 9.924984404105189,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 5176
+    },
+    {
+      "epoch": 0.0472972972972973,
+      "grad_norm": 8.25808341688057,
+      "learning_rate": 1e-06,
+      "loss": 0.3261,
+      "step": 5177
+    },
+    {
+      "epoch": 0.04954954954954955,
+      "grad_norm": 5.87132111762747,
+      "learning_rate": 1e-06,
+      "loss": 0.3338,
+      "step": 5178
+    },
+    {
+      "epoch": 0.0518018018018018,
+      "grad_norm": 6.8389979780840315,
+      "learning_rate": 1e-06,
+      "loss": 0.3298,
+      "step": 5179
+    },
+    {
+      "epoch": 0.05405405405405406,
+      "grad_norm": 7.604368488462541,
+      "learning_rate": 1e-06,
+      "loss": 0.3244,
+      "step": 5180
+    },
+    {
+      "epoch": 0.05630630630630631,
+      "grad_norm": 6.572098999865782,
+      "learning_rate": 1e-06,
+      "loss": 0.321,
+      "step": 5181
+    },
+    {
+      "epoch": 0.05855855855855856,
+      "grad_norm": 6.679028441106784,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 5182
+    },
+    {
+      "epoch": 0.060810810810810814,
+      "grad_norm": 6.281723278513657,
+      "learning_rate": 1e-06,
+      "loss": 0.3383,
+      "step": 5183
+    },
+    {
+      "epoch": 0.06306306306306306,
+      "grad_norm": 7.677422050983386,
+      "learning_rate": 1e-06,
+      "loss": 0.3074,
+      "step": 5184
+    },
+    {
+      "epoch": 0.06531531531531531,
+      "grad_norm": 5.9728302293407065,
+      "learning_rate": 1e-06,
+      "loss": 0.3262,
+      "step": 5185
+    },
+    {
+      "epoch": 0.06756756756756757,
+      "grad_norm": 6.304804626802949,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 5186
+    },
+    {
+      "epoch": 0.06981981981981981,
+      "grad_norm": 5.986883407776628,
+      "learning_rate": 1e-06,
+      "loss": 0.316,
+      "step": 5187
+    },
+    {
+      "epoch": 0.07207207207207207,
+      "grad_norm": 7.898261040998382,
+      "learning_rate": 1e-06,
+      "loss": 0.3212,
+      "step": 5188
+    },
+    {
+      "epoch": 0.07432432432432433,
+      "grad_norm": 7.0426512015425935,
+      "learning_rate": 1e-06,
+      "loss": 0.3299,
+      "step": 5189
+    },
+    {
+      "epoch": 0.07657657657657657,
+      "grad_norm": 6.57198657444034,
+      "learning_rate": 1e-06,
+      "loss": 0.2842,
+      "step": 5190
+    },
+    {
+      "epoch": 0.07882882882882883,
+      "grad_norm": 7.485780654715777,
+      "learning_rate": 1e-06,
+      "loss": 0.3184,
+      "step": 5191
+    },
+    {
+      "epoch": 0.08108108108108109,
+      "grad_norm": 6.999384525215833,
+      "learning_rate": 1e-06,
+      "loss": 0.306,
+      "step": 5192
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 6.992809082369173,
+      "learning_rate": 1e-06,
+      "loss": 0.3138,
+      "step": 5193
+    },
+    {
+      "epoch": 0.08558558558558559,
+      "grad_norm": 8.473342315075941,
+      "learning_rate": 1e-06,
+      "loss": 0.3307,
+      "step": 5194
+    },
+    {
+      "epoch": 0.08783783783783784,
+      "grad_norm": 7.548347771275406,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 5195
+    },
+    {
+      "epoch": 0.09009009009009009,
+      "grad_norm": 7.265457820154639,
+      "learning_rate": 1e-06,
+      "loss": 0.308,
+      "step": 5196
+    },
+    {
+      "epoch": 0.09234234234234234,
+      "grad_norm": 8.849313167819085,
+      "learning_rate": 1e-06,
+      "loss": 0.3481,
+      "step": 5197
+    },
+    {
+      "epoch": 0.0945945945945946,
+      "grad_norm": 7.600557756390086,
+      "learning_rate": 1e-06,
+      "loss": 0.335,
+      "step": 5198
+    },
+    {
+      "epoch": 0.09684684684684684,
+      "grad_norm": 8.078128306702457,
+      "learning_rate": 1e-06,
+      "loss": 0.3338,
+      "step": 5199
+    },
+    {
+      "epoch": 0.0990990990990991,
+      "grad_norm": 10.38825611190609,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 5200
+    },
+    {
+      "epoch": 0.10135135135135136,
+      "grad_norm": 6.451902709907623,
+      "learning_rate": 1e-06,
+      "loss": 0.3102,
+      "step": 5201
+    },
+    {
+      "epoch": 0.1036036036036036,
+      "grad_norm": 7.038926270670157,
+      "learning_rate": 1e-06,
+      "loss": 0.3136,
+      "step": 5202
+    },
+    {
+      "epoch": 0.10585585585585586,
+      "grad_norm": 8.31143415948738,
+      "learning_rate": 1e-06,
+      "loss": 0.3246,
+      "step": 5203
+    },
+    {
+      "epoch": 0.10810810810810811,
+      "grad_norm": 9.552786956589797,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 5204
+    },
+    {
+      "epoch": 0.11036036036036036,
+      "grad_norm": 8.865973024813853,
+      "learning_rate": 1e-06,
+      "loss": 0.3216,
+      "step": 5205
+    },
+    {
+      "epoch": 0.11261261261261261,
+      "grad_norm": 5.405009345686311,
+      "learning_rate": 1e-06,
+      "loss": 0.333,
+      "step": 5206
+    },
+    {
+      "epoch": 0.11486486486486487,
+      "grad_norm": 7.879799401648092,
+      "learning_rate": 1e-06,
+      "loss": 0.3121,
+      "step": 5207
+    },
+    {
+      "epoch": 0.11711711711711711,
+      "grad_norm": 6.8835961440748346,
+      "learning_rate": 1e-06,
+      "loss": 0.3085,
+      "step": 5208
+    },
+    {
+      "epoch": 0.11936936936936937,
+      "grad_norm": 7.147925808406203,
+      "learning_rate": 1e-06,
+      "loss": 0.3109,
+      "step": 5209
+    },
+    {
+      "epoch": 0.12162162162162163,
+      "grad_norm": 7.930099124425205,
+      "learning_rate": 1e-06,
+      "loss": 0.3146,
+      "step": 5210
+    },
+    {
+      "epoch": 0.12387387387387387,
+      "grad_norm": 6.690008727027918,
+      "learning_rate": 1e-06,
+      "loss": 0.3371,
+      "step": 5211
+    },
+    {
+      "epoch": 0.12612612612612611,
+      "grad_norm": 7.51047184788796,
+      "learning_rate": 1e-06,
+      "loss": 0.33,
+      "step": 5212
+    },
+    {
+      "epoch": 0.12837837837837837,
+      "grad_norm": 7.928736661146186,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 5213
+    },
+    {
+      "epoch": 0.13063063063063063,
+      "grad_norm": 6.910105312673439,
+      "learning_rate": 1e-06,
+      "loss": 0.3083,
+      "step": 5214
+    },
+    {
+      "epoch": 0.13288288288288289,
+      "grad_norm": 8.194493823305688,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 5215
+    },
+    {
+      "epoch": 0.13513513513513514,
+      "grad_norm": 6.341638763374736,
+      "learning_rate": 1e-06,
+      "loss": 0.2945,
+      "step": 5216
+    },
+    {
+      "epoch": 0.1373873873873874,
+      "grad_norm": 6.9698546544772615,
+      "learning_rate": 1e-06,
+      "loss": 0.3227,
+      "step": 5217
+    },
+    {
+      "epoch": 0.13963963963963963,
+      "grad_norm": 7.037487856772894,
+      "learning_rate": 1e-06,
+      "loss": 0.3257,
+      "step": 5218
+    },
+    {
+      "epoch": 0.14189189189189189,
+      "grad_norm": 8.449629050600118,
+      "learning_rate": 1e-06,
+      "loss": 0.3166,
+      "step": 5219
+    },
+    {
+      "epoch": 0.14414414414414414,
+      "grad_norm": 6.2960878909206,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 5220
+    },
+    {
+      "epoch": 0.1463963963963964,
+      "grad_norm": 7.936707427901606,
+      "learning_rate": 1e-06,
+      "loss": 0.3263,
+      "step": 5221
+    },
+    {
+      "epoch": 0.14864864864864866,
+      "grad_norm": 6.6680069740837595,
+      "learning_rate": 1e-06,
+      "loss": 0.3146,
+      "step": 5222
+    },
+    {
+      "epoch": 0.15090090090090091,
+      "grad_norm": 6.341822977521908,
+      "learning_rate": 1e-06,
+      "loss": 0.3116,
+      "step": 5223
+    },
+    {
+      "epoch": 0.15315315315315314,
+      "grad_norm": 7.1864114902847716,
+      "learning_rate": 1e-06,
+      "loss": 0.3208,
+      "step": 5224
+    },
+    {
+      "epoch": 0.1554054054054054,
+      "grad_norm": 6.972848062421419,
+      "learning_rate": 1e-06,
+      "loss": 0.3011,
+      "step": 5225
+    },
+    {
+      "epoch": 0.15765765765765766,
+      "grad_norm": 7.3058143883771125,
+      "learning_rate": 1e-06,
+      "loss": 0.3425,
+      "step": 5226
+    },
+    {
+      "epoch": 0.15990990990990991,
+      "grad_norm": 7.926242314825811,
+      "learning_rate": 1e-06,
+      "loss": 0.3263,
+      "step": 5227
+    },
+    {
+      "epoch": 0.16216216216216217,
+      "grad_norm": 6.952461568133972,
+      "learning_rate": 1e-06,
+      "loss": 0.3019,
+      "step": 5228
+    },
+    {
+      "epoch": 0.16441441441441443,
+      "grad_norm": 6.762997253921209,
+      "learning_rate": 1e-06,
+      "loss": 0.3219,
+      "step": 5229
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 8.105902462215159,
+      "learning_rate": 1e-06,
+      "loss": 0.3176,
+      "step": 5230
+    },
+    {
+      "epoch": 0.16891891891891891,
+      "grad_norm": 7.30240918699225,
+      "learning_rate": 1e-06,
+      "loss": 0.3283,
+      "step": 5231
+    },
+    {
+      "epoch": 0.17117117117117117,
+      "grad_norm": 6.386953684121523,
+      "learning_rate": 1e-06,
+      "loss": 0.3132,
+      "step": 5232
+    },
+    {
+      "epoch": 0.17342342342342343,
+      "grad_norm": 7.928252181282579,
+      "learning_rate": 1e-06,
+      "loss": 0.3283,
+      "step": 5233
+    },
+    {
+      "epoch": 0.17567567567567569,
+      "grad_norm": 6.979641576566025,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 5234
+    },
+    {
+      "epoch": 0.17792792792792791,
+      "grad_norm": 7.253824996409714,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 5235
+    },
+    {
+      "epoch": 0.18018018018018017,
+      "grad_norm": 10.456596106365941,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 5236
+    },
+    {
+      "epoch": 0.18243243243243243,
+      "grad_norm": 7.694118158117361,
+      "learning_rate": 1e-06,
+      "loss": 0.3158,
+      "step": 5237
+    },
+    {
+      "epoch": 0.18468468468468469,
+      "grad_norm": 6.801194093380545,
+      "learning_rate": 1e-06,
+      "loss": 0.3134,
+      "step": 5238
+    },
+    {
+      "epoch": 0.18693693693693694,
+      "grad_norm": 7.200857433700844,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 5239
+    },
+    {
+      "epoch": 0.1891891891891892,
+      "grad_norm": 7.493926389811327,
+      "learning_rate": 1e-06,
+      "loss": 0.3363,
+      "step": 5240
+    },
+    {
+      "epoch": 0.19144144144144143,
+      "grad_norm": 6.819271718779478,
+      "learning_rate": 1e-06,
+      "loss": 0.3062,
+      "step": 5241
+    },
+    {
+      "epoch": 0.19369369369369369,
+      "grad_norm": 7.400967430561009,
+      "learning_rate": 1e-06,
+      "loss": 0.3286,
+      "step": 5242
+    },
+    {
+      "epoch": 0.19594594594594594,
+      "grad_norm": 5.727944845602254,
+      "learning_rate": 1e-06,
+      "loss": 0.3341,
+      "step": 5243
+    },
+    {
+      "epoch": 0.1981981981981982,
+      "grad_norm": 5.756660683536945,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 5244
+    },
+    {
+      "epoch": 0.20045045045045046,
+      "grad_norm": 7.794455821094258,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 5245
+    },
+    {
+      "epoch": 0.20270270270270271,
+      "grad_norm": 8.932294447977384,
+      "learning_rate": 1e-06,
+      "loss": 0.3286,
+      "step": 5246
+    },
+    {
+      "epoch": 0.20495495495495494,
+      "grad_norm": 6.6374851341362975,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 5247
+    },
+    {
+      "epoch": 0.2072072072072072,
+      "grad_norm": 6.769573983162398,
+      "learning_rate": 1e-06,
+      "loss": 0.3341,
+      "step": 5248
+    },
+    {
+      "epoch": 0.20945945945945946,
+      "grad_norm": 7.234027618235902,
+      "learning_rate": 1e-06,
+      "loss": 0.3012,
+      "step": 5249
+    },
+    {
+      "epoch": 0.21171171171171171,
+      "grad_norm": 7.431229892031314,
+      "learning_rate": 1e-06,
+      "loss": 0.2974,
+      "step": 5250
+    },
+    {
+      "epoch": 0.21396396396396397,
+      "grad_norm": 8.39596871484994,
+      "learning_rate": 1e-06,
+      "loss": 0.3427,
+      "step": 5251
+    },
+    {
+      "epoch": 0.21621621621621623,
+      "grad_norm": 6.740964269116039,
+      "learning_rate": 1e-06,
+      "loss": 0.3206,
+      "step": 5252
+    },
+    {
+      "epoch": 0.21846846846846846,
+      "grad_norm": 7.272772733350043,
+      "learning_rate": 1e-06,
+      "loss": 0.3448,
+      "step": 5253
+    },
+    {
+      "epoch": 0.22072072072072071,
+      "grad_norm": 6.832233374386195,
+      "learning_rate": 1e-06,
+      "loss": 0.3109,
+      "step": 5254
+    },
+    {
+      "epoch": 0.22297297297297297,
+      "grad_norm": 7.7756283885298805,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 5255
+    },
+    {
+      "epoch": 0.22522522522522523,
+      "grad_norm": 8.985869518891858,
+      "learning_rate": 1e-06,
+      "loss": 0.3292,
+      "step": 5256
+    },
+    {
+      "epoch": 0.22747747747747749,
+      "grad_norm": 6.403432343817598,
+      "learning_rate": 1e-06,
+      "loss": 0.3385,
+      "step": 5257
+    },
+    {
+      "epoch": 0.22972972972972974,
+      "grad_norm": 7.316276522411472,
+      "learning_rate": 1e-06,
+      "loss": 0.3072,
+      "step": 5258
+    },
+    {
+      "epoch": 0.23198198198198197,
+      "grad_norm": 5.645097117785438,
+      "learning_rate": 1e-06,
+      "loss": 0.299,
+      "step": 5259
+    },
+    {
+      "epoch": 0.23423423423423423,
+      "grad_norm": 6.36340071749768,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 5260
+    },
+    {
+      "epoch": 0.23648648648648649,
+      "grad_norm": 6.1547595481120405,
+      "learning_rate": 1e-06,
+      "loss": 0.3364,
+      "step": 5261
+    },
+    {
+      "epoch": 0.23873873873873874,
+      "grad_norm": 7.374429621950195,
+      "learning_rate": 1e-06,
+      "loss": 0.3466,
+      "step": 5262
+    },
+    {
+      "epoch": 0.240990990990991,
+      "grad_norm": 6.890248784066347,
+      "learning_rate": 1e-06,
+      "loss": 0.3342,
+      "step": 5263
+    },
+    {
+      "epoch": 0.24324324324324326,
+      "grad_norm": 7.403286169566154,
+      "learning_rate": 1e-06,
+      "loss": 0.2961,
+      "step": 5264
+    },
+    {
+      "epoch": 0.24549549549549549,
+      "grad_norm": 7.61185507432192,
+      "learning_rate": 1e-06,
+      "loss": 0.3198,
+      "step": 5265
+    },
+    {
+      "epoch": 0.24774774774774774,
+      "grad_norm": 7.093804822637177,
+      "learning_rate": 1e-06,
+      "loss": 0.3147,
+      "step": 5266
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 5.6855655958544125,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 5267
+    },
+    {
+      "epoch": 0.25225225225225223,
+      "grad_norm": 7.708903081703649,
+      "learning_rate": 1e-06,
+      "loss": 0.2973,
+      "step": 5268
+    },
+    {
+      "epoch": 0.2545045045045045,
+      "grad_norm": 6.855693133277391,
+      "learning_rate": 1e-06,
+      "loss": 0.3159,
+      "step": 5269
+    },
+    {
+      "epoch": 0.25675675675675674,
+      "grad_norm": 8.398384975730695,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 5270
+    },
+    {
+      "epoch": 0.25900900900900903,
+      "grad_norm": 8.496900136678478,
+      "learning_rate": 1e-06,
+      "loss": 0.318,
+      "step": 5271
+    },
+    {
+      "epoch": 0.26126126126126126,
+      "grad_norm": 7.793743700112725,
+      "learning_rate": 1e-06,
+      "loss": 0.3335,
+      "step": 5272
+    },
+    {
+      "epoch": 0.2635135135135135,
+      "grad_norm": 9.218200238136129,
+      "learning_rate": 1e-06,
+      "loss": 0.3419,
+      "step": 5273
+    },
+    {
+      "epoch": 0.26576576576576577,
+      "grad_norm": 6.900176046123427,
+      "learning_rate": 1e-06,
+      "loss": 0.3516,
+      "step": 5274
+    },
+    {
+      "epoch": 0.268018018018018,
+      "grad_norm": 8.28025492127777,
+      "learning_rate": 1e-06,
+      "loss": 0.3297,
+      "step": 5275
+    },
+    {
+      "epoch": 0.2702702702702703,
+      "grad_norm": 6.147015557751888,
+      "learning_rate": 1e-06,
+      "loss": 0.3439,
+      "step": 5276
+    },
+    {
+      "epoch": 0.2725225225225225,
+      "grad_norm": 7.056357132653111,
+      "learning_rate": 1e-06,
+      "loss": 0.3205,
+      "step": 5277
+    },
+    {
+      "epoch": 0.2747747747747748,
+      "grad_norm": 7.7034042240660865,
+      "learning_rate": 1e-06,
+      "loss": 0.3202,
+      "step": 5278
+    },
+    {
+      "epoch": 0.27702702702702703,
+      "grad_norm": 6.776411324617189,
+      "learning_rate": 1e-06,
+      "loss": 0.334,
+      "step": 5279
+    },
+    {
+      "epoch": 0.27927927927927926,
+      "grad_norm": 6.1835250973241545,
+      "learning_rate": 1e-06,
+      "loss": 0.3045,
+      "step": 5280
+    },
+    {
+      "epoch": 0.28153153153153154,
+      "grad_norm": 6.67342703071086,
+      "learning_rate": 1e-06,
+      "loss": 0.3144,
+      "step": 5281
+    },
+    {
+      "epoch": 0.28378378378378377,
+      "grad_norm": 5.842468592724617,
+      "learning_rate": 1e-06,
+      "loss": 0.3035,
+      "step": 5282
+    },
+    {
+      "epoch": 0.28603603603603606,
+      "grad_norm": 6.624652033495355,
+      "learning_rate": 1e-06,
+      "loss": 0.329,
+      "step": 5283
+    },
+    {
+      "epoch": 0.2882882882882883,
+      "grad_norm": 8.047380873457953,
+      "learning_rate": 1e-06,
+      "loss": 0.3247,
+      "step": 5284
+    },
+    {
+      "epoch": 0.2905405405405405,
+      "grad_norm": 8.189796907723114,
+      "learning_rate": 1e-06,
+      "loss": 0.3096,
+      "step": 5285
+    },
+    {
+      "epoch": 0.2927927927927928,
+      "grad_norm": 9.114505985593146,
+      "learning_rate": 1e-06,
+      "loss": 0.3256,
+      "step": 5286
+    },
+    {
+      "epoch": 0.29504504504504503,
+      "grad_norm": 7.0262301219066465,
+      "learning_rate": 1e-06,
+      "loss": 0.3185,
+      "step": 5287
+    },
+    {
+      "epoch": 0.2972972972972973,
+      "grad_norm": 6.701419341888639,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 5288
+    },
+    {
+      "epoch": 0.29954954954954954,
+      "grad_norm": 6.45831757443887,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 5289
+    },
+    {
+      "epoch": 0.30180180180180183,
+      "grad_norm": 6.706811150397045,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 5290
+    },
+    {
+      "epoch": 0.30405405405405406,
+      "grad_norm": 7.320480831942273,
+      "learning_rate": 1e-06,
+      "loss": 0.3486,
+      "step": 5291
+    },
+    {
+      "epoch": 0.3063063063063063,
+      "grad_norm": 7.650508602270824,
+      "learning_rate": 1e-06,
+      "loss": 0.319,
+      "step": 5292
+    },
+    {
+      "epoch": 0.30855855855855857,
+      "grad_norm": 5.574941014064892,
+      "learning_rate": 1e-06,
+      "loss": 0.3183,
+      "step": 5293
+    },
+    {
+      "epoch": 0.3108108108108108,
+      "grad_norm": 7.24462182002278,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 5294
+    },
+    {
+      "epoch": 0.3130630630630631,
+      "grad_norm": 6.659059885491798,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 5295
+    },
+    {
+      "epoch": 0.3153153153153153,
+      "grad_norm": 6.386876151714684,
+      "learning_rate": 1e-06,
+      "loss": 0.3273,
+      "step": 5296
+    },
+    {
+      "epoch": 0.31756756756756754,
+      "grad_norm": 7.345952132376607,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 5297
+    },
+    {
+      "epoch": 0.31981981981981983,
+      "grad_norm": 8.011779592641396,
+      "learning_rate": 1e-06,
+      "loss": 0.3285,
+      "step": 5298
+    },
+    {
+      "epoch": 0.32207207207207206,
+      "grad_norm": 5.764039755225807,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 5299
+    },
+    {
+      "epoch": 0.32432432432432434,
+      "grad_norm": 6.728748676249665,
+      "learning_rate": 1e-06,
+      "loss": 0.3097,
+      "step": 5300
+    },
+    {
+      "epoch": 0.32657657657657657,
+      "grad_norm": 5.813466605071482,
+      "learning_rate": 1e-06,
+      "loss": 0.3077,
+      "step": 5301
+    },
+    {
+      "epoch": 0.32882882882882886,
+      "grad_norm": 7.723844949744793,
+      "learning_rate": 1e-06,
+      "loss": 0.3176,
+      "step": 5302
+    },
+    {
+      "epoch": 0.3310810810810811,
+      "grad_norm": 7.416691868090636,
+      "learning_rate": 1e-06,
+      "loss": 0.3384,
+      "step": 5303
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 5.990935862958518,
+      "learning_rate": 1e-06,
+      "loss": 0.2933,
+      "step": 5304
+    },
+    {
+      "epoch": 0.3355855855855856,
+      "grad_norm": 6.924194988208131,
+      "learning_rate": 1e-06,
+      "loss": 0.3163,
+      "step": 5305
+    },
+    {
+      "epoch": 0.33783783783783783,
+      "grad_norm": 7.569139978853263,
+      "learning_rate": 1e-06,
+      "loss": 0.3125,
+      "step": 5306
+    },
+    {
+      "epoch": 0.3400900900900901,
+      "grad_norm": 5.846742599819281,
+      "learning_rate": 1e-06,
+      "loss": 0.3332,
+      "step": 5307
+    },
+    {
+      "epoch": 0.34234234234234234,
+      "grad_norm": 7.889042494748713,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 5308
+    },
+    {
+      "epoch": 0.34459459459459457,
+      "grad_norm": 6.949288908182527,
+      "learning_rate": 1e-06,
+      "loss": 0.2837,
+      "step": 5309
+    },
+    {
+      "epoch": 0.34684684684684686,
+      "grad_norm": 6.423513079887645,
+      "learning_rate": 1e-06,
+      "loss": 0.3227,
+      "step": 5310
+    },
+    {
+      "epoch": 0.3490990990990991,
+      "grad_norm": 6.312037525124616,
+      "learning_rate": 1e-06,
+      "loss": 0.3076,
+      "step": 5311
+    },
+    {
+      "epoch": 0.35135135135135137,
+      "grad_norm": 6.701541427036546,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 5312
+    },
+    {
+      "epoch": 0.3536036036036036,
+      "grad_norm": 7.643290239063676,
+      "learning_rate": 1e-06,
+      "loss": 0.3124,
+      "step": 5313
+    },
+    {
+      "epoch": 0.35585585585585583,
+      "grad_norm": 8.031531522588313,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 5314
+    },
+    {
+      "epoch": 0.3581081081081081,
+      "grad_norm": 7.8898336650408165,
+      "learning_rate": 1e-06,
+      "loss": 0.3212,
+      "step": 5315
+    },
+    {
+      "epoch": 0.36036036036036034,
+      "grad_norm": 8.370091018551285,
+      "learning_rate": 1e-06,
+      "loss": 0.3249,
+      "step": 5316
+    },
+    {
+      "epoch": 0.36261261261261263,
+      "grad_norm": 7.054478640921099,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 5317
+    },
+    {
+      "epoch": 0.36486486486486486,
+      "grad_norm": 6.760727827044356,
+      "learning_rate": 1e-06,
+      "loss": 0.3288,
+      "step": 5318
+    },
+    {
+      "epoch": 0.36711711711711714,
+      "grad_norm": 7.996954169491507,
+      "learning_rate": 1e-06,
+      "loss": 0.3475,
+      "step": 5319
+    },
+    {
+      "epoch": 0.36936936936936937,
+      "grad_norm": 6.0752249801149265,
+      "learning_rate": 1e-06,
+      "loss": 0.3,
+      "step": 5320
+    },
+    {
+      "epoch": 0.3716216216216216,
+      "grad_norm": 5.288289300498874,
+      "learning_rate": 1e-06,
+      "loss": 0.3179,
+      "step": 5321
+    },
+    {
+      "epoch": 0.3738738738738739,
+      "grad_norm": 6.915885116528909,
+      "learning_rate": 1e-06,
+      "loss": 0.3321,
+      "step": 5322
+    },
+    {
+      "epoch": 0.3761261261261261,
+      "grad_norm": 7.804815151021324,
+      "learning_rate": 1e-06,
+      "loss": 0.3408,
+      "step": 5323
+    },
+    {
+      "epoch": 0.3783783783783784,
+      "grad_norm": 7.502781942244273,
+      "learning_rate": 1e-06,
+      "loss": 0.3331,
+      "step": 5324
+    },
+    {
+      "epoch": 0.38063063063063063,
+      "grad_norm": 10.077090091783173,
+      "learning_rate": 1e-06,
+      "loss": 0.3074,
+      "step": 5325
+    },
+    {
+      "epoch": 0.38288288288288286,
+      "grad_norm": 6.502223354579169,
+      "learning_rate": 1e-06,
+      "loss": 0.3292,
+      "step": 5326
+    },
+    {
+      "epoch": 0.38513513513513514,
+      "grad_norm": 6.574616024362753,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 5327
+    },
+    {
+      "epoch": 0.38738738738738737,
+      "grad_norm": 7.266653771735807,
+      "learning_rate": 1e-06,
+      "loss": 0.3163,
+      "step": 5328
+    },
+    {
+      "epoch": 0.38963963963963966,
+      "grad_norm": 8.306836071908746,
+      "learning_rate": 1e-06,
+      "loss": 0.31,
+      "step": 5329
+    },
+    {
+      "epoch": 0.3918918918918919,
+      "grad_norm": 6.624689385604465,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 5330
+    },
+    {
+      "epoch": 0.39414414414414417,
+      "grad_norm": 8.049079093292672,
+      "learning_rate": 1e-06,
+      "loss": 0.32,
+      "step": 5331
+    },
+    {
+      "epoch": 0.3963963963963964,
+      "grad_norm": 8.318988708247792,
+      "learning_rate": 1e-06,
+      "loss": 0.3086,
+      "step": 5332
+    },
+    {
+      "epoch": 0.39864864864864863,
+      "grad_norm": 6.818692676727219,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 5333
+    },
+    {
+      "epoch": 0.4009009009009009,
+      "grad_norm": 6.232991643134649,
+      "learning_rate": 1e-06,
+      "loss": 0.3092,
+      "step": 5334
+    },
+    {
+      "epoch": 0.40315315315315314,
+      "grad_norm": 8.450220757107994,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 5335
+    },
+    {
+      "epoch": 0.40540540540540543,
+      "grad_norm": 6.463455079510501,
+      "learning_rate": 1e-06,
+      "loss": 0.3182,
+      "step": 5336
+    },
+    {
+      "epoch": 0.40765765765765766,
+      "grad_norm": 5.822598778959468,
+      "learning_rate": 1e-06,
+      "loss": 0.2981,
+      "step": 5337
+    },
+    {
+      "epoch": 0.4099099099099099,
+      "grad_norm": 8.209002179397201,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 5338
+    },
+    {
+      "epoch": 0.41216216216216217,
+      "grad_norm": 8.67660796830226,
+      "learning_rate": 1e-06,
+      "loss": 0.3277,
+      "step": 5339
+    },
+    {
+      "epoch": 0.4144144144144144,
+      "grad_norm": 7.452992612597149,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 5340
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 6.766376607545276,
+      "learning_rate": 1e-06,
+      "loss": 0.3056,
+      "step": 5341
+    },
+    {
+      "epoch": 0.4189189189189189,
+      "grad_norm": 6.562287514977561,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 5342
+    },
+    {
+      "epoch": 0.42117117117117114,
+      "grad_norm": 9.611403731759212,
+      "learning_rate": 1e-06,
+      "loss": 0.332,
+      "step": 5343
+    },
+    {
+      "epoch": 0.42342342342342343,
+      "grad_norm": 7.529991441416387,
+      "learning_rate": 1e-06,
+      "loss": 0.3245,
+      "step": 5344
+    },
+    {
+      "epoch": 0.42567567567567566,
+      "grad_norm": 6.124216266059375,
+      "learning_rate": 1e-06,
+      "loss": 0.3098,
+      "step": 5345
+    },
+    {
+      "epoch": 0.42792792792792794,
+      "grad_norm": 6.332480240443934,
+      "learning_rate": 1e-06,
+      "loss": 0.3175,
+      "step": 5346
+    },
+    {
+      "epoch": 0.43018018018018017,
+      "grad_norm": 9.254224235366847,
+      "learning_rate": 1e-06,
+      "loss": 0.343,
+      "step": 5347
+    },
+    {
+      "epoch": 0.43243243243243246,
+      "grad_norm": 6.692152598998611,
+      "learning_rate": 1e-06,
+      "loss": 0.3115,
+      "step": 5348
+    },
+    {
+      "epoch": 0.4346846846846847,
+      "grad_norm": 6.222591882263219,
+      "learning_rate": 1e-06,
+      "loss": 0.3221,
+      "step": 5349
+    },
+    {
+      "epoch": 0.4369369369369369,
+      "grad_norm": 7.206835789359219,
+      "learning_rate": 1e-06,
+      "loss": 0.3244,
+      "step": 5350
+    },
+    {
+      "epoch": 0.4391891891891892,
+      "grad_norm": 7.06692576087257,
+      "learning_rate": 1e-06,
+      "loss": 0.2953,
+      "step": 5351
+    },
+    {
+      "epoch": 0.44144144144144143,
+      "grad_norm": 6.511793296784286,
+      "learning_rate": 1e-06,
+      "loss": 0.3027,
+      "step": 5352
+    },
+    {
+      "epoch": 0.4436936936936937,
+      "grad_norm": 7.207748898985032,
+      "learning_rate": 1e-06,
+      "loss": 0.3232,
+      "step": 5353
+    },
+    {
+      "epoch": 0.44594594594594594,
+      "grad_norm": 6.273703383114765,
+      "learning_rate": 1e-06,
+      "loss": 0.3488,
+      "step": 5354
+    },
+    {
+      "epoch": 0.44819819819819817,
+      "grad_norm": 5.466727854057295,
+      "learning_rate": 1e-06,
+      "loss": 0.3107,
+      "step": 5355
+    },
+    {
+      "epoch": 0.45045045045045046,
+      "grad_norm": 8.80211897549044,
+      "learning_rate": 1e-06,
+      "loss": 0.3206,
+      "step": 5356
+    },
+    {
+      "epoch": 0.4527027027027027,
+      "grad_norm": 8.112947810173672,
+      "learning_rate": 1e-06,
+      "loss": 0.3223,
+      "step": 5357
+    },
+    {
+      "epoch": 0.45495495495495497,
+      "grad_norm": 7.342585243612855,
+      "learning_rate": 1e-06,
+      "loss": 0.304,
+      "step": 5358
+    },
+    {
+      "epoch": 0.4572072072072072,
+      "grad_norm": 8.84116894145625,
+      "learning_rate": 1e-06,
+      "loss": 0.3169,
+      "step": 5359
+    },
+    {
+      "epoch": 0.4594594594594595,
+      "grad_norm": 6.819602370229979,
+      "learning_rate": 1e-06,
+      "loss": 0.3218,
+      "step": 5360
+    },
+    {
+      "epoch": 0.4617117117117117,
+      "grad_norm": 7.193224055307329,
+      "learning_rate": 1e-06,
+      "loss": 0.3331,
+      "step": 5361
+    },
+    {
+      "epoch": 0.46396396396396394,
+      "grad_norm": 6.477654721292751,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 5362
+    },
+    {
+      "epoch": 0.46621621621621623,
+      "grad_norm": 6.324639709256162,
+      "learning_rate": 1e-06,
+      "loss": 0.323,
+      "step": 5363
+    },
+    {
+      "epoch": 0.46846846846846846,
+      "grad_norm": 7.818049073166432,
+      "learning_rate": 1e-06,
+      "loss": 0.3311,
+      "step": 5364
+    },
+    {
+      "epoch": 0.47072072072072074,
+      "grad_norm": 9.81941495457653,
+      "learning_rate": 1e-06,
+      "loss": 0.3188,
+      "step": 5365
+    },
+    {
+      "epoch": 0.47297297297297297,
+      "grad_norm": 5.57928294388273,
+      "learning_rate": 1e-06,
+      "loss": 0.3102,
+      "step": 5366
+    },
+    {
+      "epoch": 0.4752252252252252,
+      "grad_norm": 6.307197498719491,
+      "learning_rate": 1e-06,
+      "loss": 0.3183,
+      "step": 5367
+    },
+    {
+      "epoch": 0.4774774774774775,
+      "grad_norm": 8.229089698181317,
+      "learning_rate": 1e-06,
+      "loss": 0.3314,
+      "step": 5368
+    },
+    {
+      "epoch": 0.4797297297297297,
+      "grad_norm": 8.086068719230227,
+      "learning_rate": 1e-06,
+      "loss": 0.3226,
+      "step": 5369
+    },
+    {
+      "epoch": 0.481981981981982,
+      "grad_norm": 8.388534234880606,
+      "learning_rate": 1e-06,
+      "loss": 0.3381,
+      "step": 5370
+    },
+    {
+      "epoch": 0.48423423423423423,
+      "grad_norm": 6.650621017223249,
+      "learning_rate": 1e-06,
+      "loss": 0.3272,
+      "step": 5371
+    },
+    {
+      "epoch": 0.4864864864864865,
+      "grad_norm": 7.750660993316189,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 5372
+    },
+    {
+      "epoch": 0.48873873873873874,
+      "grad_norm": 6.601194443402548,
+      "learning_rate": 1e-06,
+      "loss": 0.3042,
+      "step": 5373
+    },
+    {
+      "epoch": 0.49099099099099097,
+      "grad_norm": 9.172915933278926,
+      "learning_rate": 1e-06,
+      "loss": 0.3121,
+      "step": 5374
+    },
+    {
+      "epoch": 0.49324324324324326,
+      "grad_norm": 6.603692979103642,
+      "learning_rate": 1e-06,
+      "loss": 0.3007,
+      "step": 5375
+    },
+    {
+      "epoch": 0.4954954954954955,
+      "grad_norm": 7.892959921164577,
+      "learning_rate": 1e-06,
+      "loss": 0.3325,
+      "step": 5376
+    },
+    {
+      "epoch": 0.49774774774774777,
+      "grad_norm": 6.794432971865564,
+      "learning_rate": 1e-06,
+      "loss": 0.3272,
+      "step": 5377
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.934772826937557,
+      "learning_rate": 1e-06,
+      "loss": 0.3432,
+      "step": 5378
+    },
+    {
+      "epoch": 0.5022522522522522,
+      "grad_norm": 7.36652703383297,
+      "learning_rate": 1e-06,
+      "loss": 0.2952,
+      "step": 5379
+    },
+    {
+      "epoch": 0.5045045045045045,
+      "grad_norm": 6.537646345633481,
+      "learning_rate": 1e-06,
+      "loss": 0.3327,
+      "step": 5380
+    },
+    {
+      "epoch": 0.5067567567567568,
+      "grad_norm": 6.522983621287693,
+      "learning_rate": 1e-06,
+      "loss": 0.3071,
+      "step": 5381
+    },
+    {
+      "epoch": 0.509009009009009,
+      "grad_norm": 8.578787268391912,
+      "learning_rate": 1e-06,
+      "loss": 0.3415,
+      "step": 5382
+    },
+    {
+      "epoch": 0.5112612612612613,
+      "grad_norm": 7.634361074706386,
+      "learning_rate": 1e-06,
+      "loss": 0.3261,
+      "step": 5383
+    },
+    {
+      "epoch": 0.5135135135135135,
+      "grad_norm": 6.846434846557052,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 5384
+    },
+    {
+      "epoch": 0.5157657657657657,
+      "grad_norm": 6.761929340759996,
+      "learning_rate": 1e-06,
+      "loss": 0.3364,
+      "step": 5385
+    },
+    {
+      "epoch": 0.5180180180180181,
+      "grad_norm": 8.822120492764133,
+      "learning_rate": 1e-06,
+      "loss": 0.3144,
+      "step": 5386
+    },
+    {
+      "epoch": 0.5202702702702703,
+      "grad_norm": 7.154055121334878,
+      "learning_rate": 1e-06,
+      "loss": 0.3201,
+      "step": 5387
+    },
+    {
+      "epoch": 0.5225225225225225,
+      "grad_norm": 7.764316924756586,
+      "learning_rate": 1e-06,
+      "loss": 0.3149,
+      "step": 5388
+    },
+    {
+      "epoch": 0.5247747747747747,
+      "grad_norm": 6.502119592849681,
+      "learning_rate": 1e-06,
+      "loss": 0.31,
+      "step": 5389
+    },
+    {
+      "epoch": 0.527027027027027,
+      "grad_norm": 6.511889803438759,
+      "learning_rate": 1e-06,
+      "loss": 0.305,
+      "step": 5390
+    },
+    {
+      "epoch": 0.5292792792792793,
+      "grad_norm": 5.329047226445315,
+      "learning_rate": 1e-06,
+      "loss": 0.3301,
+      "step": 5391
+    },
+    {
+      "epoch": 0.5315315315315315,
+      "grad_norm": 7.477826226606451,
+      "learning_rate": 1e-06,
+      "loss": 0.3323,
+      "step": 5392
+    },
+    {
+      "epoch": 0.5337837837837838,
+      "grad_norm": 9.14864229761287,
+      "learning_rate": 1e-06,
+      "loss": 0.346,
+      "step": 5393
+    },
+    {
+      "epoch": 0.536036036036036,
+      "grad_norm": 6.810966749173878,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 5394
+    },
+    {
+      "epoch": 0.5382882882882883,
+      "grad_norm": 6.421293234128329,
+      "learning_rate": 1e-06,
+      "loss": 0.3522,
+      "step": 5395
+    },
+    {
+      "epoch": 0.5405405405405406,
+      "grad_norm": 8.740454276790953,
+      "learning_rate": 1e-06,
+      "loss": 0.3444,
+      "step": 5396
+    },
+    {
+      "epoch": 0.5427927927927928,
+      "grad_norm": 6.553450679393297,
+      "learning_rate": 1e-06,
+      "loss": 0.3089,
+      "step": 5397
+    },
+    {
+      "epoch": 0.545045045045045,
+      "grad_norm": 6.721473749717357,
+      "learning_rate": 1e-06,
+      "loss": 0.3215,
+      "step": 5398
+    },
+    {
+      "epoch": 0.5472972972972973,
+      "grad_norm": 8.158941154371897,
+      "learning_rate": 1e-06,
+      "loss": 0.3545,
+      "step": 5399
+    },
+    {
+      "epoch": 0.5495495495495496,
+      "grad_norm": 6.382272033578016,
+      "learning_rate": 1e-06,
+      "loss": 0.3099,
+      "step": 5400
+    },
+    {
+      "epoch": 0.5518018018018018,
+      "grad_norm": 8.944121959771305,
+      "learning_rate": 1e-06,
+      "loss": 0.3136,
+      "step": 5401
+    },
+    {
+      "epoch": 0.5540540540540541,
+      "grad_norm": 7.328239442740871,
+      "learning_rate": 1e-06,
+      "loss": 0.341,
+      "step": 5402
+    },
+    {
+      "epoch": 0.5563063063063063,
+      "grad_norm": 7.763775711759702,
+      "learning_rate": 1e-06,
+      "loss": 0.3165,
+      "step": 5403
+    },
+    {
+      "epoch": 0.5585585585585585,
+      "grad_norm": 7.344217764415733,
+      "learning_rate": 1e-06,
+      "loss": 0.3068,
+      "step": 5404
+    },
+    {
+      "epoch": 0.5608108108108109,
+      "grad_norm": 6.766891919235008,
+      "learning_rate": 1e-06,
+      "loss": 0.3303,
+      "step": 5405
+    },
+    {
+      "epoch": 0.5630630630630631,
+      "grad_norm": 6.664495533633409,
+      "learning_rate": 1e-06,
+      "loss": 0.3244,
+      "step": 5406
+    },
+    {
+      "epoch": 0.5653153153153153,
+      "grad_norm": 7.637748058930405,
+      "learning_rate": 1e-06,
+      "loss": 0.3552,
+      "step": 5407
+    },
+    {
+      "epoch": 0.5675675675675675,
+      "grad_norm": 7.487955657586506,
+      "learning_rate": 1e-06,
+      "loss": 0.3077,
+      "step": 5408
+    },
+    {
+      "epoch": 0.5698198198198198,
+      "grad_norm": 7.2306108287207085,
+      "learning_rate": 1e-06,
+      "loss": 0.346,
+      "step": 5409
+    },
+    {
+      "epoch": 0.5720720720720721,
+      "grad_norm": 10.283722517308728,
+      "learning_rate": 1e-06,
+      "loss": 0.3487,
+      "step": 5410
+    },
+    {
+      "epoch": 0.5743243243243243,
+      "grad_norm": 6.765029449655265,
+      "learning_rate": 1e-06,
+      "loss": 0.2953,
+      "step": 5411
+    },
+    {
+      "epoch": 0.5765765765765766,
+      "grad_norm": 6.332889977975074,
+      "learning_rate": 1e-06,
+      "loss": 0.3193,
+      "step": 5412
+    },
+    {
+      "epoch": 0.5788288288288288,
+      "grad_norm": 6.604253542160518,
+      "learning_rate": 1e-06,
+      "loss": 0.3177,
+      "step": 5413
+    },
+    {
+      "epoch": 0.581081081081081,
+      "grad_norm": 8.274908246967714,
+      "learning_rate": 1e-06,
+      "loss": 0.3226,
+      "step": 5414
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 6.480148500180884,
+      "learning_rate": 1e-06,
+      "loss": 0.3127,
+      "step": 5415
+    },
+    {
+      "epoch": 0.5855855855855856,
+      "grad_norm": 6.738967881860603,
+      "learning_rate": 1e-06,
+      "loss": 0.3067,
+      "step": 5416
+    },
+    {
+      "epoch": 0.5878378378378378,
+      "grad_norm": 4.694863861170037,
+      "learning_rate": 1e-06,
+      "loss": 0.3252,
+      "step": 5417
+    },
+    {
+      "epoch": 0.5900900900900901,
+      "grad_norm": 5.774401965637477,
+      "learning_rate": 1e-06,
+      "loss": 0.3082,
+      "step": 5418
+    },
+    {
+      "epoch": 0.5923423423423423,
+      "grad_norm": 7.259356112494466,
+      "learning_rate": 1e-06,
+      "loss": 0.31,
+      "step": 5419
+    },
+    {
+      "epoch": 0.5945945945945946,
+      "grad_norm": 7.885436634139152,
+      "learning_rate": 1e-06,
+      "loss": 0.341,
+      "step": 5420
+    },
+    {
+      "epoch": 0.5968468468468469,
+      "grad_norm": 7.333766706861398,
+      "learning_rate": 1e-06,
+      "loss": 0.3494,
+      "step": 5421
+    },
+    {
+      "epoch": 0.5990990990990991,
+      "grad_norm": 7.524551546075724,
+      "learning_rate": 1e-06,
+      "loss": 0.3101,
+      "step": 5422
+    },
+    {
+      "epoch": 0.6013513513513513,
+      "grad_norm": 7.8514530442846935,
+      "learning_rate": 1e-06,
+      "loss": 0.2988,
+      "step": 5423
+    },
+    {
+      "epoch": 0.6036036036036037,
+      "grad_norm": 8.678893733892284,
+      "learning_rate": 1e-06,
+      "loss": 0.3403,
+      "step": 5424
+    },
+    {
+      "epoch": 0.6058558558558559,
+      "grad_norm": 7.230935300496467,
+      "learning_rate": 1e-06,
+      "loss": 0.3355,
+      "step": 5425
+    },
+    {
+      "epoch": 0.6081081081081081,
+      "grad_norm": 7.319346722043944,
+      "learning_rate": 1e-06,
+      "loss": 0.3233,
+      "step": 5426
+    },
+    {
+      "epoch": 0.6103603603603603,
+      "grad_norm": 6.524120256275736,
+      "learning_rate": 1e-06,
+      "loss": 0.3131,
+      "step": 5427
+    },
+    {
+      "epoch": 0.6126126126126126,
+      "grad_norm": 5.620542474266848,
+      "learning_rate": 1e-06,
+      "loss": 0.3049,
+      "step": 5428
+    },
+    {
+      "epoch": 0.6148648648648649,
+      "grad_norm": 6.935087337523633,
+      "learning_rate": 1e-06,
+      "loss": 0.3118,
+      "step": 5429
+    },
+    {
+      "epoch": 0.6171171171171171,
+      "grad_norm": 8.708631269017031,
+      "learning_rate": 1e-06,
+      "loss": 0.3151,
+      "step": 5430
+    },
+    {
+      "epoch": 0.6193693693693694,
+      "grad_norm": 5.116056305514849,
+      "learning_rate": 1e-06,
+      "loss": 0.3019,
+      "step": 5431
+    },
+    {
+      "epoch": 0.6216216216216216,
+      "grad_norm": 7.765983241886376,
+      "learning_rate": 1e-06,
+      "loss": 0.3126,
+      "step": 5432
+    },
+    {
+      "epoch": 0.6238738738738738,
+      "grad_norm": 10.062621186226187,
+      "learning_rate": 1e-06,
+      "loss": 0.3352,
+      "step": 5433
+    },
+    {
+      "epoch": 0.6261261261261262,
+      "grad_norm": 7.937985370319497,
+      "learning_rate": 1e-06,
+      "loss": 0.3038,
+      "step": 5434
+    },
+    {
+      "epoch": 0.6283783783783784,
+      "grad_norm": 6.0095656249986,
+      "learning_rate": 1e-06,
+      "loss": 0.3273,
+      "step": 5435
+    },
+    {
+      "epoch": 0.6306306306306306,
+      "grad_norm": 8.003398668443836,
+      "learning_rate": 1e-06,
+      "loss": 0.3393,
+      "step": 5436
+    },
+    {
+      "epoch": 0.6328828828828829,
+      "grad_norm": 6.198337659563322,
+      "learning_rate": 1e-06,
+      "loss": 0.322,
+      "step": 5437
+    },
+    {
+      "epoch": 0.6351351351351351,
+      "grad_norm": 6.0534235195029105,
+      "learning_rate": 1e-06,
+      "loss": 0.3277,
+      "step": 5438
+    },
+    {
+      "epoch": 0.6373873873873874,
+      "grad_norm": 7.072162519159402,
+      "learning_rate": 1e-06,
+      "loss": 0.3213,
+      "step": 5439
+    },
+    {
+      "epoch": 0.6396396396396397,
+      "grad_norm": 8.13101613754172,
+      "learning_rate": 1e-06,
+      "loss": 0.3256,
+      "step": 5440
+    },
+    {
+      "epoch": 0.6418918918918919,
+      "grad_norm": 8.54616141650256,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 5441
+    },
+    {
+      "epoch": 0.6441441441441441,
+      "grad_norm": 6.686733521830167,
+      "learning_rate": 1e-06,
+      "loss": 0.3183,
+      "step": 5442
+    },
+    {
+      "epoch": 0.6463963963963963,
+      "grad_norm": 6.839676983327748,
+      "learning_rate": 1e-06,
+      "loss": 0.317,
+      "step": 5443
+    },
+    {
+      "epoch": 0.6486486486486487,
+      "grad_norm": 5.188435368739653,
+      "learning_rate": 1e-06,
+      "loss": 0.3116,
+      "step": 5444
+    },
+    {
+      "epoch": 0.6509009009009009,
+      "grad_norm": 7.9285460095710345,
+      "learning_rate": 1e-06,
+      "loss": 0.3451,
+      "step": 5445
+    },
+    {
+      "epoch": 0.6531531531531531,
+      "grad_norm": 6.707808371543336,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 5446
+    },
+    {
+      "epoch": 0.6554054054054054,
+      "grad_norm": 7.918029222371571,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 5447
+    },
+    {
+      "epoch": 0.6576576576576577,
+      "grad_norm": 8.327985084888867,
+      "learning_rate": 1e-06,
+      "loss": 0.3285,
+      "step": 5448
+    },
+    {
+      "epoch": 0.6599099099099099,
+      "grad_norm": 8.597511754453068,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 5449
+    },
+    {
+      "epoch": 0.6621621621621622,
+      "grad_norm": 6.723830164097339,
+      "learning_rate": 1e-06,
+      "loss": 0.3239,
+      "step": 5450
+    },
+    {
+      "epoch": 0.6644144144144144,
+      "grad_norm": 7.913747980390913,
+      "learning_rate": 1e-06,
+      "loss": 0.312,
+      "step": 5451
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 6.32911323096526,
+      "learning_rate": 1e-06,
+      "loss": 0.3184,
+      "step": 5452
+    },
+    {
+      "epoch": 0.668918918918919,
+      "grad_norm": 6.473286591123673,
+      "learning_rate": 1e-06,
+      "loss": 0.3207,
+      "step": 5453
+    },
+    {
+      "epoch": 0.6711711711711712,
+      "grad_norm": 5.280324579624947,
+      "learning_rate": 1e-06,
+      "loss": 0.3227,
+      "step": 5454
+    },
+    {
+      "epoch": 0.6734234234234234,
+      "grad_norm": 6.140282185729232,
+      "learning_rate": 1e-06,
+      "loss": 0.3305,
+      "step": 5455
+    },
+    {
+      "epoch": 0.6756756756756757,
+      "grad_norm": 8.559245878025136,
+      "learning_rate": 1e-06,
+      "loss": 0.3143,
+      "step": 5456
+    },
+    {
+      "epoch": 0.6779279279279279,
+      "grad_norm": 5.781403593168831,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 5457
+    },
+    {
+      "epoch": 0.6801801801801802,
+      "grad_norm": 6.492601509322805,
+      "learning_rate": 1e-06,
+      "loss": 0.3034,
+      "step": 5458
+    },
+    {
+      "epoch": 0.6824324324324325,
+      "grad_norm": 8.889710052983968,
+      "learning_rate": 1e-06,
+      "loss": 0.3063,
+      "step": 5459
+    },
+    {
+      "epoch": 0.6846846846846847,
+      "grad_norm": 6.852961850099681,
+      "learning_rate": 1e-06,
+      "loss": 0.3227,
+      "step": 5460
+    },
+    {
+      "epoch": 0.6869369369369369,
+      "grad_norm": 7.647690839853147,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 5461
+    },
+    {
+      "epoch": 0.6891891891891891,
+      "grad_norm": 7.218714981267028,
+      "learning_rate": 1e-06,
+      "loss": 0.3442,
+      "step": 5462
+    },
+    {
+      "epoch": 0.6914414414414415,
+      "grad_norm": 9.34915401262728,
+      "learning_rate": 1e-06,
+      "loss": 0.3509,
+      "step": 5463
+    },
+    {
+      "epoch": 0.6936936936936937,
+      "grad_norm": 5.842200512358564,
+      "learning_rate": 1e-06,
+      "loss": 0.2942,
+      "step": 5464
+    },
+    {
+      "epoch": 0.6959459459459459,
+      "grad_norm": 6.058731082406452,
+      "learning_rate": 1e-06,
+      "loss": 0.2709,
+      "step": 5465
+    },
+    {
+      "epoch": 0.6981981981981982,
+      "grad_norm": 6.775316737449217,
+      "learning_rate": 1e-06,
+      "loss": 0.351,
+      "step": 5466
+    },
+    {
+      "epoch": 0.7004504504504504,
+      "grad_norm": 8.254137534678167,
+      "learning_rate": 1e-06,
+      "loss": 0.3344,
+      "step": 5467
+    },
+    {
+      "epoch": 0.7027027027027027,
+      "grad_norm": 7.2486984790866655,
+      "learning_rate": 1e-06,
+      "loss": 0.3274,
+      "step": 5468
+    },
+    {
+      "epoch": 0.704954954954955,
+      "grad_norm": 7.259950732617692,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 5469
+    },
+    {
+      "epoch": 0.7072072072072072,
+      "grad_norm": 7.492961957863386,
+      "learning_rate": 1e-06,
+      "loss": 0.3173,
+      "step": 5470
+    },
+    {
+      "epoch": 0.7094594594594594,
+      "grad_norm": 8.60214254617289,
+      "learning_rate": 1e-06,
+      "loss": 0.3278,
+      "step": 5471
+    },
+    {
+      "epoch": 0.7117117117117117,
+      "grad_norm": 7.318302653714911,
+      "learning_rate": 1e-06,
+      "loss": 0.3358,
+      "step": 5472
+    },
+    {
+      "epoch": 0.713963963963964,
+      "grad_norm": 6.253968914448552,
+      "learning_rate": 1e-06,
+      "loss": 0.3323,
+      "step": 5473
+    },
+    {
+      "epoch": 0.7162162162162162,
+      "grad_norm": 7.825969044741898,
+      "learning_rate": 1e-06,
+      "loss": 0.3125,
+      "step": 5474
+    },
+    {
+      "epoch": 0.7184684684684685,
+      "grad_norm": 7.93952949058017,
+      "learning_rate": 1e-06,
+      "loss": 0.3269,
+      "step": 5475
+    },
+    {
+      "epoch": 0.7207207207207207,
+      "grad_norm": 7.621179638021929,
+      "learning_rate": 1e-06,
+      "loss": 0.3288,
+      "step": 5476
+    },
+    {
+      "epoch": 0.722972972972973,
+      "grad_norm": 7.089772854395029,
+      "learning_rate": 1e-06,
+      "loss": 0.3003,
+      "step": 5477
+    },
+    {
+      "epoch": 0.7252252252252253,
+      "grad_norm": 7.414297386714026,
+      "learning_rate": 1e-06,
+      "loss": 0.3418,
+      "step": 5478
+    },
+    {
+      "epoch": 0.7274774774774775,
+      "grad_norm": 6.700315695582439,
+      "learning_rate": 1e-06,
+      "loss": 0.3129,
+      "step": 5479
+    },
+    {
+      "epoch": 0.7297297297297297,
+      "grad_norm": 9.035948086470816,
+      "learning_rate": 1e-06,
+      "loss": 0.3144,
+      "step": 5480
+    },
+    {
+      "epoch": 0.7319819819819819,
+      "grad_norm": 8.532019574909159,
+      "learning_rate": 1e-06,
+      "loss": 0.2999,
+      "step": 5481
+    },
+    {
+      "epoch": 0.7342342342342343,
+      "grad_norm": 7.009399448616399,
+      "learning_rate": 1e-06,
+      "loss": 0.3366,
+      "step": 5482
+    },
+    {
+      "epoch": 0.7364864864864865,
+      "grad_norm": 6.518763566188763,
+      "learning_rate": 1e-06,
+      "loss": 0.3145,
+      "step": 5483
+    },
+    {
+      "epoch": 0.7387387387387387,
+      "grad_norm": 6.87509593168568,
+      "learning_rate": 1e-06,
+      "loss": 0.2841,
+      "step": 5484
+    },
+    {
+      "epoch": 0.740990990990991,
+      "grad_norm": 6.973645721296594,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 5485
+    },
+    {
+      "epoch": 0.7432432432432432,
+      "grad_norm": 6.986885597098479,
+      "learning_rate": 1e-06,
+      "loss": 0.3217,
+      "step": 5486
+    },
+    {
+      "epoch": 0.7454954954954955,
+      "grad_norm": 6.3776498961291,
+      "learning_rate": 1e-06,
+      "loss": 0.3398,
+      "step": 5487
+    },
+    {
+      "epoch": 0.7477477477477478,
+      "grad_norm": 6.884005590465922,
+      "learning_rate": 1e-06,
+      "loss": 0.309,
+      "step": 5488
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.110109254021525,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 5489
+    },
+    {
+      "epoch": 0.7522522522522522,
+      "grad_norm": 9.444983897673477,
+      "learning_rate": 1e-06,
+      "loss": 0.3317,
+      "step": 5490
+    },
+    {
+      "epoch": 0.7545045045045045,
+      "grad_norm": 6.779761472263322,
+      "learning_rate": 1e-06,
+      "loss": 0.3061,
+      "step": 5491
+    },
+    {
+      "epoch": 0.7567567567567568,
+      "grad_norm": 6.869045504395302,
+      "learning_rate": 1e-06,
+      "loss": 0.352,
+      "step": 5492
+    },
+    {
+      "epoch": 0.759009009009009,
+      "grad_norm": 5.712326322708846,
+      "learning_rate": 1e-06,
+      "loss": 0.3289,
+      "step": 5493
+    },
+    {
+      "epoch": 0.7612612612612613,
+      "grad_norm": 8.298532073675108,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 5494
+    },
+    {
+      "epoch": 0.7635135135135135,
+      "grad_norm": 7.425458699796488,
+      "learning_rate": 1e-06,
+      "loss": 0.3443,
+      "step": 5495
+    },
+    {
+      "epoch": 0.7657657657657657,
+      "grad_norm": 6.9469700074890275,
+      "learning_rate": 1e-06,
+      "loss": 0.3233,
+      "step": 5496
+    },
+    {
+      "epoch": 0.7680180180180181,
+      "grad_norm": 4.2809140975185125,
+      "learning_rate": 1e-06,
+      "loss": 0.3307,
+      "step": 5497
+    },
+    {
+      "epoch": 0.7702702702702703,
+      "grad_norm": 8.248586872956057,
+      "learning_rate": 1e-06,
+      "loss": 0.3501,
+      "step": 5498
+    },
+    {
+      "epoch": 0.7725225225225225,
+      "grad_norm": 6.9044676034725265,
+      "learning_rate": 1e-06,
+      "loss": 0.3422,
+      "step": 5499
+    },
+    {
+      "epoch": 0.7747747747747747,
+      "grad_norm": 7.01994530679417,
+      "learning_rate": 1e-06,
+      "loss": 0.3295,
+      "step": 5500
+    },
+    {
+      "epoch": 0.777027027027027,
+      "grad_norm": 7.672773338079101,
+      "learning_rate": 1e-06,
+      "loss": 0.3299,
+      "step": 5501
+    },
+    {
+      "epoch": 0.7792792792792793,
+      "grad_norm": 6.398699284467016,
+      "learning_rate": 1e-06,
+      "loss": 0.3368,
+      "step": 5502
+    },
+    {
+      "epoch": 0.7815315315315315,
+      "grad_norm": 6.249991784444267,
+      "learning_rate": 1e-06,
+      "loss": 0.3231,
+      "step": 5503
+    },
+    {
+      "epoch": 0.7837837837837838,
+      "grad_norm": 7.972611380790875,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 5504
+    },
+    {
+      "epoch": 0.786036036036036,
+      "grad_norm": 6.039499501974334,
+      "learning_rate": 1e-06,
+      "loss": 0.3364,
+      "step": 5505
+    },
+    {
+      "epoch": 0.7882882882882883,
+      "grad_norm": 6.594903891524761,
+      "learning_rate": 1e-06,
+      "loss": 0.3208,
+      "step": 5506
+    },
+    {
+      "epoch": 0.7905405405405406,
+      "grad_norm": 7.328023643078021,
+      "learning_rate": 1e-06,
+      "loss": 0.3234,
+      "step": 5507
+    },
+    {
+      "epoch": 0.7927927927927928,
+      "grad_norm": 6.859469497472776,
+      "learning_rate": 1e-06,
+      "loss": 0.3138,
+      "step": 5508
+    },
+    {
+      "epoch": 0.795045045045045,
+      "grad_norm": 7.399960368366918,
+      "learning_rate": 1e-06,
+      "loss": 0.3185,
+      "step": 5509
+    },
+    {
+      "epoch": 0.7972972972972973,
+      "grad_norm": 8.966524957512618,
+      "learning_rate": 1e-06,
+      "loss": 0.3374,
+      "step": 5510
+    },
+    {
+      "epoch": 0.7995495495495496,
+      "grad_norm": 6.697641716993483,
+      "learning_rate": 1e-06,
+      "loss": 0.3281,
+      "step": 5511
+    },
+    {
+      "epoch": 0.8018018018018018,
+      "grad_norm": 8.300646137133672,
+      "learning_rate": 1e-06,
+      "loss": 0.3306,
+      "step": 5512
+    },
+    {
+      "epoch": 0.8040540540540541,
+      "grad_norm": 8.14272246382396,
+      "learning_rate": 1e-06,
+      "loss": 0.3245,
+      "step": 5513
+    },
+    {
+      "epoch": 0.8063063063063063,
+      "grad_norm": 8.089241840523876,
+      "learning_rate": 1e-06,
+      "loss": 0.3062,
+      "step": 5514
+    },
+    {
+      "epoch": 0.8085585585585585,
+      "grad_norm": 6.689231448042,
+      "learning_rate": 1e-06,
+      "loss": 0.3286,
+      "step": 5515
+    },
+    {
+      "epoch": 0.8108108108108109,
+      "grad_norm": 7.193497405587711,
+      "learning_rate": 1e-06,
+      "loss": 0.3415,
+      "step": 5516
+    },
+    {
+      "epoch": 0.8130630630630631,
+      "grad_norm": 7.565743899658709,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 5517
+    },
+    {
+      "epoch": 0.8153153153153153,
+      "grad_norm": 4.998543240431273,
+      "learning_rate": 1e-06,
+      "loss": 0.2945,
+      "step": 5518
+    },
+    {
+      "epoch": 0.8175675675675675,
+      "grad_norm": 7.456655039151803,
+      "learning_rate": 1e-06,
+      "loss": 0.3309,
+      "step": 5519
+    },
+    {
+      "epoch": 0.8198198198198198,
+      "grad_norm": 8.148301112673357,
+      "learning_rate": 1e-06,
+      "loss": 0.2993,
+      "step": 5520
+    },
+    {
+      "epoch": 0.8220720720720721,
+      "grad_norm": 6.966460442754697,
+      "learning_rate": 1e-06,
+      "loss": 0.3244,
+      "step": 5521
+    },
+    {
+      "epoch": 0.8243243243243243,
+      "grad_norm": 7.861421783093714,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 5522
+    },
+    {
+      "epoch": 0.8265765765765766,
+      "grad_norm": 7.729087319060444,
+      "learning_rate": 1e-06,
+      "loss": 0.3196,
+      "step": 5523
+    },
+    {
+      "epoch": 0.8288288288288288,
+      "grad_norm": 8.254588067762546,
+      "learning_rate": 1e-06,
+      "loss": 0.3473,
+      "step": 5524
+    },
+    {
+      "epoch": 0.831081081081081,
+      "grad_norm": 7.110869819441001,
+      "learning_rate": 1e-06,
+      "loss": 0.3316,
+      "step": 5525
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 7.358635829628693,
+      "learning_rate": 1e-06,
+      "loss": 0.3434,
+      "step": 5526
+    },
+    {
+      "epoch": 0.8355855855855856,
+      "grad_norm": 7.855232400052965,
+      "learning_rate": 1e-06,
+      "loss": 0.3206,
+      "step": 5527
+    },
+    {
+      "epoch": 0.8378378378378378,
+      "grad_norm": 6.4011728840115865,
+      "learning_rate": 1e-06,
+      "loss": 0.3303,
+      "step": 5528
+    },
+    {
+      "epoch": 0.8400900900900901,
+      "grad_norm": 8.970655395529091,
+      "learning_rate": 1e-06,
+      "loss": 0.3091,
+      "step": 5529
+    },
+    {
+      "epoch": 0.8423423423423423,
+      "grad_norm": 7.834052362503054,
+      "learning_rate": 1e-06,
+      "loss": 0.3316,
+      "step": 5530
+    },
+    {
+      "epoch": 0.8445945945945946,
+      "grad_norm": 5.8064564201787245,
+      "learning_rate": 1e-06,
+      "loss": 0.3187,
+      "step": 5531
+    },
+    {
+      "epoch": 0.8468468468468469,
+      "grad_norm": 8.738518249581647,
+      "learning_rate": 1e-06,
+      "loss": 0.3098,
+      "step": 5532
+    },
+    {
+      "epoch": 0.8490990990990991,
+      "grad_norm": 9.136594714899294,
+      "learning_rate": 1e-06,
+      "loss": 0.3287,
+      "step": 5533
+    },
+    {
+      "epoch": 0.8513513513513513,
+      "grad_norm": 6.849608408085804,
+      "learning_rate": 1e-06,
+      "loss": 0.3229,
+      "step": 5534
+    },
+    {
+      "epoch": 0.8536036036036037,
+      "grad_norm": 7.0715114182265575,
+      "learning_rate": 1e-06,
+      "loss": 0.341,
+      "step": 5535
+    },
+    {
+      "epoch": 0.8558558558558559,
+      "grad_norm": 7.77813043732286,
+      "learning_rate": 1e-06,
+      "loss": 0.3435,
+      "step": 5536
+    },
+    {
+      "epoch": 0.8581081081081081,
+      "grad_norm": 6.600698406864426,
+      "learning_rate": 1e-06,
+      "loss": 0.3155,
+      "step": 5537
+    },
+    {
+      "epoch": 0.8603603603603603,
+      "grad_norm": 8.018120871258368,
+      "learning_rate": 1e-06,
+      "loss": 0.3374,
+      "step": 5538
+    },
+    {
+      "epoch": 0.8626126126126126,
+      "grad_norm": 7.290229894895965,
+      "learning_rate": 1e-06,
+      "loss": 0.3292,
+      "step": 5539
+    },
+    {
+      "epoch": 0.8648648648648649,
+      "grad_norm": 7.311527569747864,
+      "learning_rate": 1e-06,
+      "loss": 0.3163,
+      "step": 5540
+    },
+    {
+      "epoch": 0.8671171171171171,
+      "grad_norm": 7.318342402529882,
+      "learning_rate": 1e-06,
+      "loss": 0.3131,
+      "step": 5541
+    },
+    {
+      "epoch": 0.8693693693693694,
+      "grad_norm": 6.276644299642851,
+      "learning_rate": 1e-06,
+      "loss": 0.3052,
+      "step": 5542
+    },
+    {
+      "epoch": 0.8716216216216216,
+      "grad_norm": 9.741875856850305,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 5543
+    },
+    {
+      "epoch": 0.8738738738738738,
+      "grad_norm": 7.469196682134166,
+      "learning_rate": 1e-06,
+      "loss": 0.3369,
+      "step": 5544
+    },
+    {
+      "epoch": 0.8761261261261262,
+      "grad_norm": 5.5257631165914995,
+      "learning_rate": 1e-06,
+      "loss": 0.3482,
+      "step": 5545
+    },
+    {
+      "epoch": 0.8783783783783784,
+      "grad_norm": 7.203168736439987,
+      "learning_rate": 1e-06,
+      "loss": 0.3039,
+      "step": 5546
+    },
+    {
+      "epoch": 0.8806306306306306,
+      "grad_norm": 6.526473321594657,
+      "learning_rate": 1e-06,
+      "loss": 0.3232,
+      "step": 5547
+    },
+    {
+      "epoch": 0.8828828828828829,
+      "grad_norm": 7.1752659216160435,
+      "learning_rate": 1e-06,
+      "loss": 0.3238,
+      "step": 5548
+    },
+    {
+      "epoch": 0.8851351351351351,
+      "grad_norm": 7.577840627663471,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 5549
+    },
+    {
+      "epoch": 0.8873873873873874,
+      "grad_norm": 7.1070705689325235,
+      "learning_rate": 1e-06,
+      "loss": 0.3204,
+      "step": 5550
+    },
+    {
+      "epoch": 0.8896396396396397,
+      "grad_norm": 6.786163428984164,
+      "learning_rate": 1e-06,
+      "loss": 0.2999,
+      "step": 5551
+    },
+    {
+      "epoch": 0.8918918918918919,
+      "grad_norm": 7.216127589963638,
+      "learning_rate": 1e-06,
+      "loss": 0.3113,
+      "step": 5552
+    },
+    {
+      "epoch": 0.8941441441441441,
+      "grad_norm": 7.626758088904114,
+      "learning_rate": 1e-06,
+      "loss": 0.3186,
+      "step": 5553
+    },
+    {
+      "epoch": 0.8963963963963963,
+      "grad_norm": 7.936911959433355,
+      "learning_rate": 1e-06,
+      "loss": 0.3374,
+      "step": 5554
+    },
+    {
+      "epoch": 0.8986486486486487,
+      "grad_norm": 7.4234272255817135,
+      "learning_rate": 1e-06,
+      "loss": 0.3293,
+      "step": 5555
+    },
+    {
+      "epoch": 0.9009009009009009,
+      "grad_norm": 7.310144916277256,
+      "learning_rate": 1e-06,
+      "loss": 0.3235,
+      "step": 5556
+    },
+    {
+      "epoch": 0.9031531531531531,
+      "grad_norm": 6.815288869862716,
+      "learning_rate": 1e-06,
+      "loss": 0.3212,
+      "step": 5557
+    },
+    {
+      "epoch": 0.9054054054054054,
+      "grad_norm": 8.012227446532233,
+      "learning_rate": 1e-06,
+      "loss": 0.3235,
+      "step": 5558
+    },
+    {
+      "epoch": 0.9076576576576577,
+      "grad_norm": 6.497451873084041,
+      "learning_rate": 1e-06,
+      "loss": 0.3414,
+      "step": 5559
+    },
+    {
+      "epoch": 0.9099099099099099,
+      "grad_norm": 7.131475053920634,
+      "learning_rate": 1e-06,
+      "loss": 0.3237,
+      "step": 5560
+    },
+    {
+      "epoch": 0.9121621621621622,
+      "grad_norm": 7.444580966340294,
+      "learning_rate": 1e-06,
+      "loss": 0.3214,
+      "step": 5561
+    },
+    {
+      "epoch": 0.9144144144144144,
+      "grad_norm": 6.902032656734674,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 5562
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 7.061329437979711,
+      "learning_rate": 1e-06,
+      "loss": 0.3255,
+      "step": 5563
+    },
+    {
+      "epoch": 0.918918918918919,
+      "grad_norm": 7.566323767715221,
+      "learning_rate": 1e-06,
+      "loss": 0.332,
+      "step": 5564
+    },
+    {
+      "epoch": 0.9211711711711712,
+      "grad_norm": 6.354406600772665,
+      "learning_rate": 1e-06,
+      "loss": 0.3259,
+      "step": 5565
+    },
+    {
+      "epoch": 0.9234234234234234,
+      "grad_norm": 6.101264840461013,
+      "learning_rate": 1e-06,
+      "loss": 0.3648,
+      "step": 5566
+    },
+    {
+      "epoch": 0.9256756756756757,
+      "grad_norm": 8.176091070497137,
+      "learning_rate": 1e-06,
+      "loss": 0.3126,
+      "step": 5567
+    },
+    {
+      "epoch": 0.9279279279279279,
+      "grad_norm": 6.897870391348852,
+      "learning_rate": 1e-06,
+      "loss": 0.3333,
+      "step": 5568
+    },
+    {
+      "epoch": 0.9301801801801802,
+      "grad_norm": 7.483869145194099,
+      "learning_rate": 1e-06,
+      "loss": 0.3295,
+      "step": 5569
+    },
+    {
+      "epoch": 0.9324324324324325,
+      "grad_norm": 6.149974518432305,
+      "learning_rate": 1e-06,
+      "loss": 0.3176,
+      "step": 5570
+    },
+    {
+      "epoch": 0.9346846846846847,
+      "grad_norm": 8.280241389555064,
+      "learning_rate": 1e-06,
+      "loss": 0.3278,
+      "step": 5571
+    },
+    {
+      "epoch": 0.9369369369369369,
+      "grad_norm": 7.453245179223433,
+      "learning_rate": 1e-06,
+      "loss": 0.3138,
+      "step": 5572
+    },
+    {
+      "epoch": 0.9391891891891891,
+      "grad_norm": 5.818610939778544,
+      "learning_rate": 1e-06,
+      "loss": 0.3181,
+      "step": 5573
+    },
+    {
+      "epoch": 0.9414414414414415,
+      "grad_norm": 6.11706190380727,
+      "learning_rate": 1e-06,
+      "loss": 0.3396,
+      "step": 5574
+    },
+    {
+      "epoch": 0.9436936936936937,
+      "grad_norm": 6.712196200173086,
+      "learning_rate": 1e-06,
+      "loss": 0.3066,
+      "step": 5575
+    },
+    {
+      "epoch": 0.9459459459459459,
+      "grad_norm": 5.502070413037745,
+      "learning_rate": 1e-06,
+      "loss": 0.318,
+      "step": 5576
+    },
+    {
+      "epoch": 0.9481981981981982,
+      "grad_norm": 7.301926082995069,
+      "learning_rate": 1e-06,
+      "loss": 0.311,
+      "step": 5577
+    },
+    {
+      "epoch": 0.9504504504504504,
+      "grad_norm": 6.059522775510438,
+      "learning_rate": 1e-06,
+      "loss": 0.3303,
+      "step": 5578
+    },
+    {
+      "epoch": 0.9527027027027027,
+      "grad_norm": 6.7888067581795655,
+      "learning_rate": 1e-06,
+      "loss": 0.3252,
+      "step": 5579
+    },
+    {
+      "epoch": 0.954954954954955,
+      "grad_norm": 6.815461541568879,
+      "learning_rate": 1e-06,
+      "loss": 0.3159,
+      "step": 5580
+    },
+    {
+      "epoch": 0.9572072072072072,
+      "grad_norm": 8.241721686679453,
+      "learning_rate": 1e-06,
+      "loss": 0.3146,
+      "step": 5581
+    },
+    {
+      "epoch": 0.9594594594594594,
+      "grad_norm": 6.65883115746924,
+      "learning_rate": 1e-06,
+      "loss": 0.3222,
+      "step": 5582
+    },
+    {
+      "epoch": 0.9617117117117117,
+      "grad_norm": 6.907081779758289,
+      "learning_rate": 1e-06,
+      "loss": 0.3211,
+      "step": 5583
+    },
+    {
+      "epoch": 0.963963963963964,
+      "grad_norm": 7.245270420752542,
+      "learning_rate": 1e-06,
+      "loss": 0.3368,
+      "step": 5584
+    },
+    {
+      "epoch": 0.9662162162162162,
+      "grad_norm": 6.975926557362084,
+      "learning_rate": 1e-06,
+      "loss": 0.3163,
+      "step": 5585
+    },
+    {
+      "epoch": 0.9684684684684685,
+      "grad_norm": 5.888462219867215,
+      "learning_rate": 1e-06,
+      "loss": 0.3209,
+      "step": 5586
+    },
+    {
+      "epoch": 0.9707207207207207,
+      "grad_norm": 8.179243973428564,
+      "learning_rate": 1e-06,
+      "loss": 0.3154,
+      "step": 5587
+    },
+    {
+      "epoch": 0.972972972972973,
+      "grad_norm": 6.557468938785026,
+      "learning_rate": 1e-06,
+      "loss": 0.3215,
+      "step": 5588
+    },
+    {
+      "epoch": 0.9752252252252253,
+      "grad_norm": 8.094217587348076,
+      "learning_rate": 1e-06,
+      "loss": 0.3232,
+      "step": 5589
+    },
+    {
+      "epoch": 0.9774774774774775,
+      "grad_norm": 7.53569478436751,
+      "learning_rate": 1e-06,
+      "loss": 0.3456,
+      "step": 5590
+    },
+    {
+      "epoch": 0.9797297297297297,
+      "grad_norm": 7.154488400360359,
+      "learning_rate": 1e-06,
+      "loss": 0.3271,
+      "step": 5591
+    },
+    {
+      "epoch": 0.9819819819819819,
+      "grad_norm": 5.005835375247366,
+      "learning_rate": 1e-06,
+      "loss": 0.3248,
+      "step": 5592
+    },
+    {
+      "epoch": 0.9842342342342343,
+      "grad_norm": 7.4757913302598435,
+      "learning_rate": 1e-06,
+      "loss": 0.3345,
+      "step": 5593
+    },
+    {
+      "epoch": 0.9864864864864865,
+      "grad_norm": 8.065145562778516,
+      "learning_rate": 1e-06,
+      "loss": 0.3405,
+      "step": 5594
+    },
+    {
+      "epoch": 0.9887387387387387,
+      "grad_norm": 7.681714183053637,
+      "learning_rate": 1e-06,
+      "loss": 0.3253,
+      "step": 5595
+    },
+    {
+      "epoch": 0.990990990990991,
+      "grad_norm": 8.57095513236474,
+      "learning_rate": 1e-06,
+      "loss": 0.3062,
+      "step": 5596
+    },
+    {
+      "epoch": 0.9932432432432432,
+      "grad_norm": 7.077888070252476,
+      "learning_rate": 1e-06,
+      "loss": 0.3033,
+      "step": 5597
+    },
+    {
+      "epoch": 0.9954954954954955,
+      "grad_norm": 6.90855312200592,
+      "learning_rate": 1e-06,
+      "loss": 0.3182,
+      "step": 5598
+    },
+    {
+      "epoch": 0.9977477477477478,
+      "grad_norm": 5.393928615668058,
+      "learning_rate": 1e-06,
+      "loss": 0.3225,
+      "step": 5599
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.115082811797784,
+      "learning_rate": 1e-06,
+      "loss": 0.3135,
+      "step": 5600
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.384183168411255,
+      "eval_runtime": 21.9166,
+      "eval_samples_per_second": 45.628,
+      "eval_steps_per_second": 5.703,
+      "step": 5600
+    },
+    {
+      "epoch": 0.0023121387283236996,
+      "grad_norm": 5.436765910789929,
+      "learning_rate": 1e-06,
+      "loss": 0.2256,
+      "step": 5601
+    },
+    {
+      "epoch": 0.004624277456647399,
+      "grad_norm": 6.3234989584887265,
+      "learning_rate": 1e-06,
+      "loss": 0.218,
+      "step": 5602
+    },
+    {
+      "epoch": 0.006936416184971098,
+      "grad_norm": 7.00692016524355,
+      "learning_rate": 1e-06,
+      "loss": 0.2326,
+      "step": 5603
+    },
+    {
+      "epoch": 0.009248554913294798,
+      "grad_norm": 5.146215546461571,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 5604
+    },
+    {
+      "epoch": 0.011560693641618497,
+      "grad_norm": 7.568203187269179,
+      "learning_rate": 1e-06,
+      "loss": 0.233,
+      "step": 5605
+    },
+    {
+      "epoch": 0.013872832369942197,
+      "grad_norm": 8.08888645960954,
+      "learning_rate": 1e-06,
+      "loss": 0.2143,
+      "step": 5606
+    },
+    {
+      "epoch": 0.016184971098265895,
+      "grad_norm": 5.660156109277093,
+      "learning_rate": 1e-06,
+      "loss": 0.2186,
+      "step": 5607
+    },
+    {
+      "epoch": 0.018497109826589597,
+      "grad_norm": 7.305744323772735,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 5608
+    },
+    {
+      "epoch": 0.020809248554913295,
+      "grad_norm": 8.418494082862125,
+      "learning_rate": 1e-06,
+      "loss": 0.243,
+      "step": 5609
+    },
+    {
+      "epoch": 0.023121387283236993,
+      "grad_norm": 7.628262917741139,
+      "learning_rate": 1e-06,
+      "loss": 0.2361,
+      "step": 5610
+    },
+    {
+      "epoch": 0.025433526011560695,
+      "grad_norm": 6.673384674596028,
+      "learning_rate": 1e-06,
+      "loss": 0.2314,
+      "step": 5611
+    },
+    {
+      "epoch": 0.027745664739884393,
+      "grad_norm": 8.52265196483748,
+      "learning_rate": 1e-06,
+      "loss": 0.2251,
+      "step": 5612
+    },
+    {
+      "epoch": 0.03005780346820809,
+      "grad_norm": 8.07683052906729,
+      "learning_rate": 1e-06,
+      "loss": 0.2197,
+      "step": 5613
+    },
+    {
+      "epoch": 0.03236994219653179,
+      "grad_norm": 8.551358716119333,
+      "learning_rate": 1e-06,
+      "loss": 0.2171,
+      "step": 5614
+    },
+    {
+      "epoch": 0.03468208092485549,
+      "grad_norm": 6.99096865153567,
+      "learning_rate": 1e-06,
+      "loss": 0.2213,
+      "step": 5615
+    },
+    {
+      "epoch": 0.03699421965317919,
+      "grad_norm": 6.54813490950779,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 5616
+    },
+    {
+      "epoch": 0.03930635838150289,
+      "grad_norm": 8.245764342057072,
+      "learning_rate": 1e-06,
+      "loss": 0.2322,
+      "step": 5617
+    },
+    {
+      "epoch": 0.04161849710982659,
+      "grad_norm": 8.414050287458771,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 5618
+    },
+    {
+      "epoch": 0.04393063583815029,
+      "grad_norm": 9.417831779038321,
+      "learning_rate": 1e-06,
+      "loss": 0.2219,
+      "step": 5619
+    },
+    {
+      "epoch": 0.046242774566473986,
+      "grad_norm": 6.412183722712215,
+      "learning_rate": 1e-06,
+      "loss": 0.2269,
+      "step": 5620
+    },
+    {
+      "epoch": 0.048554913294797684,
+      "grad_norm": 8.30602204642093,
+      "learning_rate": 1e-06,
+      "loss": 0.244,
+      "step": 5621
+    },
+    {
+      "epoch": 0.05086705202312139,
+      "grad_norm": 8.61100315180324,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 5622
+    },
+    {
+      "epoch": 0.05317919075144509,
+      "grad_norm": 7.981699328401414,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 5623
+    },
+    {
+      "epoch": 0.055491329479768786,
+      "grad_norm": 6.555102934441452,
+      "learning_rate": 1e-06,
+      "loss": 0.2193,
+      "step": 5624
+    },
+    {
+      "epoch": 0.057803468208092484,
+      "grad_norm": 6.65090104985722,
+      "learning_rate": 1e-06,
+      "loss": 0.243,
+      "step": 5625
+    },
+    {
+      "epoch": 0.06011560693641618,
+      "grad_norm": 9.084827749070396,
+      "learning_rate": 1e-06,
+      "loss": 0.2237,
+      "step": 5626
+    },
+    {
+      "epoch": 0.06242774566473988,
+      "grad_norm": 6.7628276291101,
+      "learning_rate": 1e-06,
+      "loss": 0.2346,
+      "step": 5627
+    },
+    {
+      "epoch": 0.06473988439306358,
+      "grad_norm": 7.2531956871169685,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 5628
+    },
+    {
+      "epoch": 0.06705202312138728,
+      "grad_norm": 6.083599058857174,
+      "learning_rate": 1e-06,
+      "loss": 0.2112,
+      "step": 5629
+    },
+    {
+      "epoch": 0.06936416184971098,
+      "grad_norm": 7.075318207012412,
+      "learning_rate": 1e-06,
+      "loss": 0.2361,
+      "step": 5630
+    },
+    {
+      "epoch": 0.07167630057803469,
+      "grad_norm": 6.811436074937457,
+      "learning_rate": 1e-06,
+      "loss": 0.2116,
+      "step": 5631
+    },
+    {
+      "epoch": 0.07398843930635839,
+      "grad_norm": 7.7724414653219025,
+      "learning_rate": 1e-06,
+      "loss": 0.2208,
+      "step": 5632
+    },
+    {
+      "epoch": 0.07630057803468208,
+      "grad_norm": 4.482264893485006,
+      "learning_rate": 1e-06,
+      "loss": 0.2268,
+      "step": 5633
+    },
+    {
+      "epoch": 0.07861271676300578,
+      "grad_norm": 7.6708590299036015,
+      "learning_rate": 1e-06,
+      "loss": 0.2273,
+      "step": 5634
+    },
+    {
+      "epoch": 0.08092485549132948,
+      "grad_norm": 8.131441523112889,
+      "learning_rate": 1e-06,
+      "loss": 0.2465,
+      "step": 5635
+    },
+    {
+      "epoch": 0.08323699421965318,
+      "grad_norm": 9.086752951478477,
+      "learning_rate": 1e-06,
+      "loss": 0.2234,
+      "step": 5636
+    },
+    {
+      "epoch": 0.08554913294797688,
+      "grad_norm": 7.7567936665031425,
+      "learning_rate": 1e-06,
+      "loss": 0.2154,
+      "step": 5637
+    },
+    {
+      "epoch": 0.08786127167630058,
+      "grad_norm": 7.97164876632758,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 5638
+    },
+    {
+      "epoch": 0.09017341040462427,
+      "grad_norm": 8.142150481254196,
+      "learning_rate": 1e-06,
+      "loss": 0.2253,
+      "step": 5639
+    },
+    {
+      "epoch": 0.09248554913294797,
+      "grad_norm": 8.611151082663053,
+      "learning_rate": 1e-06,
+      "loss": 0.2248,
+      "step": 5640
+    },
+    {
+      "epoch": 0.09479768786127167,
+      "grad_norm": 7.93441141311288,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 5641
+    },
+    {
+      "epoch": 0.09710982658959537,
+      "grad_norm": 8.45144385161806,
+      "learning_rate": 1e-06,
+      "loss": 0.2313,
+      "step": 5642
+    },
+    {
+      "epoch": 0.09942196531791908,
+      "grad_norm": 8.061607380157902,
+      "learning_rate": 1e-06,
+      "loss": 0.2332,
+      "step": 5643
+    },
+    {
+      "epoch": 0.10173410404624278,
+      "grad_norm": 8.29992643229555,
+      "learning_rate": 1e-06,
+      "loss": 0.2203,
+      "step": 5644
+    },
+    {
+      "epoch": 0.10404624277456648,
+      "grad_norm": 6.458979788739076,
+      "learning_rate": 1e-06,
+      "loss": 0.2365,
+      "step": 5645
+    },
+    {
+      "epoch": 0.10635838150289018,
+      "grad_norm": 8.536344791983462,
+      "learning_rate": 1e-06,
+      "loss": 0.2175,
+      "step": 5646
+    },
+    {
+      "epoch": 0.10867052023121387,
+      "grad_norm": 9.194442838654108,
+      "learning_rate": 1e-06,
+      "loss": 0.2366,
+      "step": 5647
+    },
+    {
+      "epoch": 0.11098265895953757,
+      "grad_norm": 9.197019233768145,
+      "learning_rate": 1e-06,
+      "loss": 0.213,
+      "step": 5648
+    },
+    {
+      "epoch": 0.11329479768786127,
+      "grad_norm": 8.550222299290402,
+      "learning_rate": 1e-06,
+      "loss": 0.2505,
+      "step": 5649
+    },
+    {
+      "epoch": 0.11560693641618497,
+      "grad_norm": 9.20245776141855,
+      "learning_rate": 1e-06,
+      "loss": 0.2411,
+      "step": 5650
+    },
+    {
+      "epoch": 0.11791907514450867,
+      "grad_norm": 9.26008265598692,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 5651
+    },
+    {
+      "epoch": 0.12023121387283237,
+      "grad_norm": 6.713280685504522,
+      "learning_rate": 1e-06,
+      "loss": 0.238,
+      "step": 5652
+    },
+    {
+      "epoch": 0.12254335260115606,
+      "grad_norm": 6.589401711050442,
+      "learning_rate": 1e-06,
+      "loss": 0.2364,
+      "step": 5653
+    },
+    {
+      "epoch": 0.12485549132947976,
+      "grad_norm": 7.598548809672375,
+      "learning_rate": 1e-06,
+      "loss": 0.237,
+      "step": 5654
+    },
+    {
+      "epoch": 0.12716763005780346,
+      "grad_norm": 7.032896859321714,
+      "learning_rate": 1e-06,
+      "loss": 0.2293,
+      "step": 5655
+    },
+    {
+      "epoch": 0.12947976878612716,
+      "grad_norm": 8.458528267576204,
+      "learning_rate": 1e-06,
+      "loss": 0.2228,
+      "step": 5656
+    },
+    {
+      "epoch": 0.13179190751445086,
+      "grad_norm": 9.862521883536814,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5657
+    },
+    {
+      "epoch": 0.13410404624277455,
+      "grad_norm": 7.240907788024173,
+      "learning_rate": 1e-06,
+      "loss": 0.242,
+      "step": 5658
+    },
+    {
+      "epoch": 0.13641618497109825,
+      "grad_norm": 6.381664883686039,
+      "learning_rate": 1e-06,
+      "loss": 0.23,
+      "step": 5659
+    },
+    {
+      "epoch": 0.13872832369942195,
+      "grad_norm": 9.025944784474497,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 5660
+    },
+    {
+      "epoch": 0.14104046242774568,
+      "grad_norm": 8.734728954651246,
+      "learning_rate": 1e-06,
+      "loss": 0.2074,
+      "step": 5661
+    },
+    {
+      "epoch": 0.14335260115606938,
+      "grad_norm": 7.632644880358296,
+      "learning_rate": 1e-06,
+      "loss": 0.2304,
+      "step": 5662
+    },
+    {
+      "epoch": 0.14566473988439307,
+      "grad_norm": 7.41393481167453,
+      "learning_rate": 1e-06,
+      "loss": 0.2316,
+      "step": 5663
+    },
+    {
+      "epoch": 0.14797687861271677,
+      "grad_norm": 5.8836132128344785,
+      "learning_rate": 1e-06,
+      "loss": 0.214,
+      "step": 5664
+    },
+    {
+      "epoch": 0.15028901734104047,
+      "grad_norm": 7.569104309214419,
+      "learning_rate": 1e-06,
+      "loss": 0.2159,
+      "step": 5665
+    },
+    {
+      "epoch": 0.15260115606936417,
+      "grad_norm": 7.075756670370005,
+      "learning_rate": 1e-06,
+      "loss": 0.2209,
+      "step": 5666
+    },
+    {
+      "epoch": 0.15491329479768787,
+      "grad_norm": 8.731942561419688,
+      "learning_rate": 1e-06,
+      "loss": 0.2479,
+      "step": 5667
+    },
+    {
+      "epoch": 0.15722543352601157,
+      "grad_norm": 7.173969890212261,
+      "learning_rate": 1e-06,
+      "loss": 0.2194,
+      "step": 5668
+    },
+    {
+      "epoch": 0.15953757225433526,
+      "grad_norm": 8.849374885234411,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 5669
+    },
+    {
+      "epoch": 0.16184971098265896,
+      "grad_norm": 8.725237063145402,
+      "learning_rate": 1e-06,
+      "loss": 0.2095,
+      "step": 5670
+    },
+    {
+      "epoch": 0.16416184971098266,
+      "grad_norm": 9.396794008718418,
+      "learning_rate": 1e-06,
+      "loss": 0.2252,
+      "step": 5671
+    },
+    {
+      "epoch": 0.16647398843930636,
+      "grad_norm": 8.095728090377754,
+      "learning_rate": 1e-06,
+      "loss": 0.2394,
+      "step": 5672
+    },
+    {
+      "epoch": 0.16878612716763006,
+      "grad_norm": 6.768028068103643,
+      "learning_rate": 1e-06,
+      "loss": 0.225,
+      "step": 5673
+    },
+    {
+      "epoch": 0.17109826589595376,
+      "grad_norm": 7.450618970519344,
+      "learning_rate": 1e-06,
+      "loss": 0.2124,
+      "step": 5674
+    },
+    {
+      "epoch": 0.17341040462427745,
+      "grad_norm": 7.253747962497706,
+      "learning_rate": 1e-06,
+      "loss": 0.2346,
+      "step": 5675
+    },
+    {
+      "epoch": 0.17572254335260115,
+      "grad_norm": 8.586325419809919,
+      "learning_rate": 1e-06,
+      "loss": 0.2281,
+      "step": 5676
+    },
+    {
+      "epoch": 0.17803468208092485,
+      "grad_norm": 6.827723706279262,
+      "learning_rate": 1e-06,
+      "loss": 0.2295,
+      "step": 5677
+    },
+    {
+      "epoch": 0.18034682080924855,
+      "grad_norm": 7.4786588798035485,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 5678
+    },
+    {
+      "epoch": 0.18265895953757225,
+      "grad_norm": 8.511326060898478,
+      "learning_rate": 1e-06,
+      "loss": 0.2296,
+      "step": 5679
+    },
+    {
+      "epoch": 0.18497109826589594,
+      "grad_norm": 5.694331542925839,
+      "learning_rate": 1e-06,
+      "loss": 0.2157,
+      "step": 5680
+    },
+    {
+      "epoch": 0.18728323699421964,
+      "grad_norm": 8.774972413371913,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 5681
+    },
+    {
+      "epoch": 0.18959537572254334,
+      "grad_norm": 7.479250509057153,
+      "learning_rate": 1e-06,
+      "loss": 0.2341,
+      "step": 5682
+    },
+    {
+      "epoch": 0.19190751445086704,
+      "grad_norm": 8.272805088221672,
+      "learning_rate": 1e-06,
+      "loss": 0.2336,
+      "step": 5683
+    },
+    {
+      "epoch": 0.19421965317919074,
+      "grad_norm": 9.140706520972815,
+      "learning_rate": 1e-06,
+      "loss": 0.2339,
+      "step": 5684
+    },
+    {
+      "epoch": 0.19653179190751446,
+      "grad_norm": 10.209168009135462,
+      "learning_rate": 1e-06,
+      "loss": 0.2375,
+      "step": 5685
+    },
+    {
+      "epoch": 0.19884393063583816,
+      "grad_norm": 7.19198451836344,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 5686
+    },
+    {
+      "epoch": 0.20115606936416186,
+      "grad_norm": 7.594652287207816,
+      "learning_rate": 1e-06,
+      "loss": 0.2356,
+      "step": 5687
+    },
+    {
+      "epoch": 0.20346820809248556,
+      "grad_norm": 7.591425111344706,
+      "learning_rate": 1e-06,
+      "loss": 0.2112,
+      "step": 5688
+    },
+    {
+      "epoch": 0.20578034682080926,
+      "grad_norm": 8.649130214211622,
+      "learning_rate": 1e-06,
+      "loss": 0.2227,
+      "step": 5689
+    },
+    {
+      "epoch": 0.20809248554913296,
+      "grad_norm": 10.26096044242295,
+      "learning_rate": 1e-06,
+      "loss": 0.2299,
+      "step": 5690
+    },
+    {
+      "epoch": 0.21040462427745665,
+      "grad_norm": 10.125078877333854,
+      "learning_rate": 1e-06,
+      "loss": 0.2421,
+      "step": 5691
+    },
+    {
+      "epoch": 0.21271676300578035,
+      "grad_norm": 6.318243653945329,
+      "learning_rate": 1e-06,
+      "loss": 0.2454,
+      "step": 5692
+    },
+    {
+      "epoch": 0.21502890173410405,
+      "grad_norm": 8.730785272448154,
+      "learning_rate": 1e-06,
+      "loss": 0.2182,
+      "step": 5693
+    },
+    {
+      "epoch": 0.21734104046242775,
+      "grad_norm": 6.417093868862165,
+      "learning_rate": 1e-06,
+      "loss": 0.2164,
+      "step": 5694
+    },
+    {
+      "epoch": 0.21965317919075145,
+      "grad_norm": 7.845117593637822,
+      "learning_rate": 1e-06,
+      "loss": 0.2349,
+      "step": 5695
+    },
+    {
+      "epoch": 0.22196531791907514,
+      "grad_norm": 8.665391087098925,
+      "learning_rate": 1e-06,
+      "loss": 0.2307,
+      "step": 5696
+    },
+    {
+      "epoch": 0.22427745664739884,
+      "grad_norm": 8.415896698521337,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5697
+    },
+    {
+      "epoch": 0.22658959537572254,
+      "grad_norm": 9.29534238818007,
+      "learning_rate": 1e-06,
+      "loss": 0.2211,
+      "step": 5698
+    },
+    {
+      "epoch": 0.22890173410404624,
+      "grad_norm": 10.576477787380917,
+      "learning_rate": 1e-06,
+      "loss": 0.2317,
+      "step": 5699
+    },
+    {
+      "epoch": 0.23121387283236994,
+      "grad_norm": 5.626230817826263,
+      "learning_rate": 1e-06,
+      "loss": 0.2266,
+      "step": 5700
+    },
+    {
+      "epoch": 0.23352601156069364,
+      "grad_norm": 6.60045204832468,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 5701
+    },
+    {
+      "epoch": 0.23583815028901733,
+      "grad_norm": 6.528089921940309,
+      "learning_rate": 1e-06,
+      "loss": 0.2248,
+      "step": 5702
+    },
+    {
+      "epoch": 0.23815028901734103,
+      "grad_norm": 7.270280172176922,
+      "learning_rate": 1e-06,
+      "loss": 0.2161,
+      "step": 5703
+    },
+    {
+      "epoch": 0.24046242774566473,
+      "grad_norm": 5.664156575839857,
+      "learning_rate": 1e-06,
+      "loss": 0.2369,
+      "step": 5704
+    },
+    {
+      "epoch": 0.24277456647398843,
+      "grad_norm": 7.867685429938776,
+      "learning_rate": 1e-06,
+      "loss": 0.2308,
+      "step": 5705
+    },
+    {
+      "epoch": 0.24508670520231213,
+      "grad_norm": 7.002979809814808,
+      "learning_rate": 1e-06,
+      "loss": 0.22,
+      "step": 5706
+    },
+    {
+      "epoch": 0.24739884393063583,
+      "grad_norm": 8.102592132428354,
+      "learning_rate": 1e-06,
+      "loss": 0.2126,
+      "step": 5707
+    },
+    {
+      "epoch": 0.24971098265895952,
+      "grad_norm": 7.688005265992132,
+      "learning_rate": 1e-06,
+      "loss": 0.2389,
+      "step": 5708
+    },
+    {
+      "epoch": 0.2520231213872832,
+      "grad_norm": 7.691310900306751,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 5709
+    },
+    {
+      "epoch": 0.2543352601156069,
+      "grad_norm": 6.413047433836749,
+      "learning_rate": 1e-06,
+      "loss": 0.2257,
+      "step": 5710
+    },
+    {
+      "epoch": 0.2566473988439306,
+      "grad_norm": 8.748007704294082,
+      "learning_rate": 1e-06,
+      "loss": 0.2215,
+      "step": 5711
+    },
+    {
+      "epoch": 0.2589595375722543,
+      "grad_norm": 8.5149266171097,
+      "learning_rate": 1e-06,
+      "loss": 0.2415,
+      "step": 5712
+    },
+    {
+      "epoch": 0.261271676300578,
+      "grad_norm": 9.426834584087482,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 5713
+    },
+    {
+      "epoch": 0.2635838150289017,
+      "grad_norm": 7.063689680970438,
+      "learning_rate": 1e-06,
+      "loss": 0.2306,
+      "step": 5714
+    },
+    {
+      "epoch": 0.2658959537572254,
+      "grad_norm": 9.37598903608441,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 5715
+    },
+    {
+      "epoch": 0.2682080924855491,
+      "grad_norm": 8.848973474253741,
+      "learning_rate": 1e-06,
+      "loss": 0.2301,
+      "step": 5716
+    },
+    {
+      "epoch": 0.2705202312138728,
+      "grad_norm": 7.7144265270286505,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 5717
+    },
+    {
+      "epoch": 0.2728323699421965,
+      "grad_norm": 6.93429107894354,
+      "learning_rate": 1e-06,
+      "loss": 0.2373,
+      "step": 5718
+    },
+    {
+      "epoch": 0.2751445086705202,
+      "grad_norm": 11.23250456140754,
+      "learning_rate": 1e-06,
+      "loss": 0.2351,
+      "step": 5719
+    },
+    {
+      "epoch": 0.2774566473988439,
+      "grad_norm": 5.186834845417571,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 5720
+    },
+    {
+      "epoch": 0.27976878612716766,
+      "grad_norm": 7.638677716816844,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 5721
+    },
+    {
+      "epoch": 0.28208092485549136,
+      "grad_norm": 7.094775910757211,
+      "learning_rate": 1e-06,
+      "loss": 0.2185,
+      "step": 5722
+    },
+    {
+      "epoch": 0.28439306358381505,
+      "grad_norm": 7.133056950715215,
+      "learning_rate": 1e-06,
+      "loss": 0.2264,
+      "step": 5723
+    },
+    {
+      "epoch": 0.28670520231213875,
+      "grad_norm": 6.490697458399165,
+      "learning_rate": 1e-06,
+      "loss": 0.2381,
+      "step": 5724
+    },
+    {
+      "epoch": 0.28901734104046245,
+      "grad_norm": 9.562208391199782,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 5725
+    },
+    {
+      "epoch": 0.29132947976878615,
+      "grad_norm": 7.464812338582423,
+      "learning_rate": 1e-06,
+      "loss": 0.2421,
+      "step": 5726
+    },
+    {
+      "epoch": 0.29364161849710985,
+      "grad_norm": 8.179199970670155,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 5727
+    },
+    {
+      "epoch": 0.29595375722543354,
+      "grad_norm": 7.601059626682718,
+      "learning_rate": 1e-06,
+      "loss": 0.2167,
+      "step": 5728
+    },
+    {
+      "epoch": 0.29826589595375724,
+      "grad_norm": 7.8227350207851405,
+      "learning_rate": 1e-06,
+      "loss": 0.2336,
+      "step": 5729
+    },
+    {
+      "epoch": 0.30057803468208094,
+      "grad_norm": 6.653419335584946,
+      "learning_rate": 1e-06,
+      "loss": 0.2219,
+      "step": 5730
+    },
+    {
+      "epoch": 0.30289017341040464,
+      "grad_norm": 9.387634273425059,
+      "learning_rate": 1e-06,
+      "loss": 0.2201,
+      "step": 5731
+    },
+    {
+      "epoch": 0.30520231213872834,
+      "grad_norm": 7.68370380097653,
+      "learning_rate": 1e-06,
+      "loss": 0.2156,
+      "step": 5732
+    },
+    {
+      "epoch": 0.30751445086705204,
+      "grad_norm": 8.040177812025435,
+      "learning_rate": 1e-06,
+      "loss": 0.2152,
+      "step": 5733
+    },
+    {
+      "epoch": 0.30982658959537573,
+      "grad_norm": 7.85700825583057,
+      "learning_rate": 1e-06,
+      "loss": 0.2308,
+      "step": 5734
+    },
+    {
+      "epoch": 0.31213872832369943,
+      "grad_norm": 7.206804437285996,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 5735
+    },
+    {
+      "epoch": 0.31445086705202313,
+      "grad_norm": 6.2784926669265255,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 5736
+    },
+    {
+      "epoch": 0.31676300578034683,
+      "grad_norm": 6.802203515816805,
+      "learning_rate": 1e-06,
+      "loss": 0.216,
+      "step": 5737
+    },
+    {
+      "epoch": 0.3190751445086705,
+      "grad_norm": 8.364389283039296,
+      "learning_rate": 1e-06,
+      "loss": 0.2171,
+      "step": 5738
+    },
+    {
+      "epoch": 0.3213872832369942,
+      "grad_norm": 7.338520566717229,
+      "learning_rate": 1e-06,
+      "loss": 0.2212,
+      "step": 5739
+    },
+    {
+      "epoch": 0.3236994219653179,
+      "grad_norm": 6.345398139507792,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 5740
+    },
+    {
+      "epoch": 0.3260115606936416,
+      "grad_norm": 8.908815876605297,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 5741
+    },
+    {
+      "epoch": 0.3283236994219653,
+      "grad_norm": 6.254871023220592,
+      "learning_rate": 1e-06,
+      "loss": 0.2285,
+      "step": 5742
+    },
+    {
+      "epoch": 0.330635838150289,
+      "grad_norm": 8.270390100983734,
+      "learning_rate": 1e-06,
+      "loss": 0.2251,
+      "step": 5743
+    },
+    {
+      "epoch": 0.3329479768786127,
+      "grad_norm": 10.218342507295873,
+      "learning_rate": 1e-06,
+      "loss": 0.2469,
+      "step": 5744
+    },
+    {
+      "epoch": 0.3352601156069364,
+      "grad_norm": 8.272720706210631,
+      "learning_rate": 1e-06,
+      "loss": 0.2398,
+      "step": 5745
+    },
+    {
+      "epoch": 0.3375722543352601,
+      "grad_norm": 6.764009367562462,
+      "learning_rate": 1e-06,
+      "loss": 0.2379,
+      "step": 5746
+    },
+    {
+      "epoch": 0.3398843930635838,
+      "grad_norm": 8.302794381419268,
+      "learning_rate": 1e-06,
+      "loss": 0.2112,
+      "step": 5747
+    },
+    {
+      "epoch": 0.3421965317919075,
+      "grad_norm": 7.9343705103440065,
+      "learning_rate": 1e-06,
+      "loss": 0.211,
+      "step": 5748
+    },
+    {
+      "epoch": 0.3445086705202312,
+      "grad_norm": 8.248632626377356,
+      "learning_rate": 1e-06,
+      "loss": 0.2239,
+      "step": 5749
+    },
+    {
+      "epoch": 0.3468208092485549,
+      "grad_norm": 6.760953267110688,
+      "learning_rate": 1e-06,
+      "loss": 0.2327,
+      "step": 5750
+    },
+    {
+      "epoch": 0.3491329479768786,
+      "grad_norm": 5.871583839543117,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 5751
+    },
+    {
+      "epoch": 0.3514450867052023,
+      "grad_norm": 9.32214179230614,
+      "learning_rate": 1e-06,
+      "loss": 0.2201,
+      "step": 5752
+    },
+    {
+      "epoch": 0.353757225433526,
+      "grad_norm": 8.017140606780405,
+      "learning_rate": 1e-06,
+      "loss": 0.2393,
+      "step": 5753
+    },
+    {
+      "epoch": 0.3560693641618497,
+      "grad_norm": 6.86538542629482,
+      "learning_rate": 1e-06,
+      "loss": 0.2308,
+      "step": 5754
+    },
+    {
+      "epoch": 0.3583815028901734,
+      "grad_norm": 7.507631738977984,
+      "learning_rate": 1e-06,
+      "loss": 0.2249,
+      "step": 5755
+    },
+    {
+      "epoch": 0.3606936416184971,
+      "grad_norm": 9.499857866722817,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 5756
+    },
+    {
+      "epoch": 0.3630057803468208,
+      "grad_norm": 7.092496861106624,
+      "learning_rate": 1e-06,
+      "loss": 0.2211,
+      "step": 5757
+    },
+    {
+      "epoch": 0.3653179190751445,
+      "grad_norm": 6.154618886864094,
+      "learning_rate": 1e-06,
+      "loss": 0.218,
+      "step": 5758
+    },
+    {
+      "epoch": 0.3676300578034682,
+      "grad_norm": 8.642198730172147,
+      "learning_rate": 1e-06,
+      "loss": 0.2263,
+      "step": 5759
+    },
+    {
+      "epoch": 0.3699421965317919,
+      "grad_norm": 5.944033138153203,
+      "learning_rate": 1e-06,
+      "loss": 0.2158,
+      "step": 5760
+    },
+    {
+      "epoch": 0.3722543352601156,
+      "grad_norm": 7.133959939585759,
+      "learning_rate": 1e-06,
+      "loss": 0.2472,
+      "step": 5761
+    },
+    {
+      "epoch": 0.3745664739884393,
+      "grad_norm": 6.627744112027572,
+      "learning_rate": 1e-06,
+      "loss": 0.2413,
+      "step": 5762
+    },
+    {
+      "epoch": 0.376878612716763,
+      "grad_norm": 8.091440913929159,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 5763
+    },
+    {
+      "epoch": 0.3791907514450867,
+      "grad_norm": 8.417616386690678,
+      "learning_rate": 1e-06,
+      "loss": 0.2398,
+      "step": 5764
+    },
+    {
+      "epoch": 0.3815028901734104,
+      "grad_norm": 6.7889149295475795,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 5765
+    },
+    {
+      "epoch": 0.3838150289017341,
+      "grad_norm": 8.058102917874502,
+      "learning_rate": 1e-06,
+      "loss": 0.2297,
+      "step": 5766
+    },
+    {
+      "epoch": 0.3861271676300578,
+      "grad_norm": 6.936947768775758,
+      "learning_rate": 1e-06,
+      "loss": 0.2192,
+      "step": 5767
+    },
+    {
+      "epoch": 0.3884393063583815,
+      "grad_norm": 8.846423354679347,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 5768
+    },
+    {
+      "epoch": 0.39075144508670523,
+      "grad_norm": 7.939781627510063,
+      "learning_rate": 1e-06,
+      "loss": 0.2254,
+      "step": 5769
+    },
+    {
+      "epoch": 0.3930635838150289,
+      "grad_norm": 9.629385009994923,
+      "learning_rate": 1e-06,
+      "loss": 0.2429,
+      "step": 5770
+    },
+    {
+      "epoch": 0.3953757225433526,
+      "grad_norm": 7.951365112432929,
+      "learning_rate": 1e-06,
+      "loss": 0.2113,
+      "step": 5771
+    },
+    {
+      "epoch": 0.3976878612716763,
+      "grad_norm": 8.929501423428599,
+      "learning_rate": 1e-06,
+      "loss": 0.2429,
+      "step": 5772
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 7.707693994430751,
+      "learning_rate": 1e-06,
+      "loss": 0.2228,
+      "step": 5773
+    },
+    {
+      "epoch": 0.4023121387283237,
+      "grad_norm": 7.748996354559339,
+      "learning_rate": 1e-06,
+      "loss": 0.2395,
+      "step": 5774
+    },
+    {
+      "epoch": 0.4046242774566474,
+      "grad_norm": 8.221608392998434,
+      "learning_rate": 1e-06,
+      "loss": 0.2101,
+      "step": 5775
+    },
+    {
+      "epoch": 0.4069364161849711,
+      "grad_norm": 5.2690726614235555,
+      "learning_rate": 1e-06,
+      "loss": 0.2083,
+      "step": 5776
+    },
+    {
+      "epoch": 0.4092485549132948,
+      "grad_norm": 8.57616678535095,
+      "learning_rate": 1e-06,
+      "loss": 0.2269,
+      "step": 5777
+    },
+    {
+      "epoch": 0.4115606936416185,
+      "grad_norm": 6.642532881573142,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 5778
+    },
+    {
+      "epoch": 0.4138728323699422,
+      "grad_norm": 8.453681341169517,
+      "learning_rate": 1e-06,
+      "loss": 0.2191,
+      "step": 5779
+    },
+    {
+      "epoch": 0.4161849710982659,
+      "grad_norm": 8.55204649459406,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 5780
+    },
+    {
+      "epoch": 0.4184971098265896,
+      "grad_norm": 7.5689014706588225,
+      "learning_rate": 1e-06,
+      "loss": 0.2236,
+      "step": 5781
+    },
+    {
+      "epoch": 0.4208092485549133,
+      "grad_norm": 9.390332033561613,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 5782
+    },
+    {
+      "epoch": 0.423121387283237,
+      "grad_norm": 6.789681055272474,
+      "learning_rate": 1e-06,
+      "loss": 0.2228,
+      "step": 5783
+    },
+    {
+      "epoch": 0.4254335260115607,
+      "grad_norm": 7.288099095490994,
+      "learning_rate": 1e-06,
+      "loss": 0.2219,
+      "step": 5784
+    },
+    {
+      "epoch": 0.4277456647398844,
+      "grad_norm": 8.60971031378669,
+      "learning_rate": 1e-06,
+      "loss": 0.23,
+      "step": 5785
+    },
+    {
+      "epoch": 0.4300578034682081,
+      "grad_norm": 7.555854215240798,
+      "learning_rate": 1e-06,
+      "loss": 0.2397,
+      "step": 5786
+    },
+    {
+      "epoch": 0.4323699421965318,
+      "grad_norm": 7.453139726386014,
+      "learning_rate": 1e-06,
+      "loss": 0.2177,
+      "step": 5787
+    },
+    {
+      "epoch": 0.4346820809248555,
+      "grad_norm": 8.37069268140302,
+      "learning_rate": 1e-06,
+      "loss": 0.2264,
+      "step": 5788
+    },
+    {
+      "epoch": 0.4369942196531792,
+      "grad_norm": 8.083944290913818,
+      "learning_rate": 1e-06,
+      "loss": 0.2322,
+      "step": 5789
+    },
+    {
+      "epoch": 0.4393063583815029,
+      "grad_norm": 7.461114379274073,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 5790
+    },
+    {
+      "epoch": 0.4416184971098266,
+      "grad_norm": 6.939543219360501,
+      "learning_rate": 1e-06,
+      "loss": 0.2313,
+      "step": 5791
+    },
+    {
+      "epoch": 0.4439306358381503,
+      "grad_norm": 6.045121240448099,
+      "learning_rate": 1e-06,
+      "loss": 0.2173,
+      "step": 5792
+    },
+    {
+      "epoch": 0.446242774566474,
+      "grad_norm": 6.973618935847128,
+      "learning_rate": 1e-06,
+      "loss": 0.2194,
+      "step": 5793
+    },
+    {
+      "epoch": 0.4485549132947977,
+      "grad_norm": 7.723896927627919,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 5794
+    },
+    {
+      "epoch": 0.4508670520231214,
+      "grad_norm": 8.550551227490068,
+      "learning_rate": 1e-06,
+      "loss": 0.234,
+      "step": 5795
+    },
+    {
+      "epoch": 0.4531791907514451,
+      "grad_norm": 8.043010458321646,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 5796
+    },
+    {
+      "epoch": 0.4554913294797688,
+      "grad_norm": 8.361408058164026,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 5797
+    },
+    {
+      "epoch": 0.4578034682080925,
+      "grad_norm": 7.734648526236644,
+      "learning_rate": 1e-06,
+      "loss": 0.2233,
+      "step": 5798
+    },
+    {
+      "epoch": 0.4601156069364162,
+      "grad_norm": 9.47406933177143,
+      "learning_rate": 1e-06,
+      "loss": 0.2256,
+      "step": 5799
+    },
+    {
+      "epoch": 0.4624277456647399,
+      "grad_norm": 9.082439707651416,
+      "learning_rate": 1e-06,
+      "loss": 0.2337,
+      "step": 5800
+    },
+    {
+      "epoch": 0.4647398843930636,
+      "grad_norm": 6.359572754647228,
+      "learning_rate": 1e-06,
+      "loss": 0.2273,
+      "step": 5801
+    },
+    {
+      "epoch": 0.46705202312138727,
+      "grad_norm": 5.815618776342935,
+      "learning_rate": 1e-06,
+      "loss": 0.225,
+      "step": 5802
+    },
+    {
+      "epoch": 0.46936416184971097,
+      "grad_norm": 7.013711296050164,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 5803
+    },
+    {
+      "epoch": 0.47167630057803467,
+      "grad_norm": 8.008568387540402,
+      "learning_rate": 1e-06,
+      "loss": 0.2247,
+      "step": 5804
+    },
+    {
+      "epoch": 0.47398843930635837,
+      "grad_norm": 7.215445646819842,
+      "learning_rate": 1e-06,
+      "loss": 0.22,
+      "step": 5805
+    },
+    {
+      "epoch": 0.47630057803468207,
+      "grad_norm": 7.8578912584622405,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 5806
+    },
+    {
+      "epoch": 0.47861271676300576,
+      "grad_norm": 9.776645480763232,
+      "learning_rate": 1e-06,
+      "loss": 0.2333,
+      "step": 5807
+    },
+    {
+      "epoch": 0.48092485549132946,
+      "grad_norm": 8.635053038694275,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 5808
+    },
+    {
+      "epoch": 0.48323699421965316,
+      "grad_norm": 8.845395213949857,
+      "learning_rate": 1e-06,
+      "loss": 0.2325,
+      "step": 5809
+    },
+    {
+      "epoch": 0.48554913294797686,
+      "grad_norm": 11.410412321581937,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5810
+    },
+    {
+      "epoch": 0.48786127167630056,
+      "grad_norm": 8.417148241192987,
+      "learning_rate": 1e-06,
+      "loss": 0.2249,
+      "step": 5811
+    },
+    {
+      "epoch": 0.49017341040462425,
+      "grad_norm": 8.234954039683966,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 5812
+    },
+    {
+      "epoch": 0.49248554913294795,
+      "grad_norm": 7.88898924606854,
+      "learning_rate": 1e-06,
+      "loss": 0.2204,
+      "step": 5813
+    },
+    {
+      "epoch": 0.49479768786127165,
+      "grad_norm": 7.490148753565223,
+      "learning_rate": 1e-06,
+      "loss": 0.2143,
+      "step": 5814
+    },
+    {
+      "epoch": 0.49710982658959535,
+      "grad_norm": 8.213640260121359,
+      "learning_rate": 1e-06,
+      "loss": 0.2153,
+      "step": 5815
+    },
+    {
+      "epoch": 0.49942196531791905,
+      "grad_norm": 6.36014094841205,
+      "learning_rate": 1e-06,
+      "loss": 0.2357,
+      "step": 5816
+    },
+    {
+      "epoch": 0.5017341040462427,
+      "grad_norm": 7.870414033667676,
+      "learning_rate": 1e-06,
+      "loss": 0.2378,
+      "step": 5817
+    },
+    {
+      "epoch": 0.5040462427745664,
+      "grad_norm": 7.6266089752954995,
+      "learning_rate": 1e-06,
+      "loss": 0.2418,
+      "step": 5818
+    },
+    {
+      "epoch": 0.5063583815028901,
+      "grad_norm": 8.354187322161943,
+      "learning_rate": 1e-06,
+      "loss": 0.2269,
+      "step": 5819
+    },
+    {
+      "epoch": 0.5086705202312138,
+      "grad_norm": 7.7299276309851805,
+      "learning_rate": 1e-06,
+      "loss": 0.2256,
+      "step": 5820
+    },
+    {
+      "epoch": 0.5109826589595375,
+      "grad_norm": 5.6532526536880505,
+      "learning_rate": 1e-06,
+      "loss": 0.2362,
+      "step": 5821
+    },
+    {
+      "epoch": 0.5132947976878612,
+      "grad_norm": 6.545833027473516,
+      "learning_rate": 1e-06,
+      "loss": 0.2229,
+      "step": 5822
+    },
+    {
+      "epoch": 0.5156069364161849,
+      "grad_norm": 6.941685985111952,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5823
+    },
+    {
+      "epoch": 0.5179190751445086,
+      "grad_norm": 6.8049148864065625,
+      "learning_rate": 1e-06,
+      "loss": 0.2323,
+      "step": 5824
+    },
+    {
+      "epoch": 0.5202312138728323,
+      "grad_norm": 8.740078655699158,
+      "learning_rate": 1e-06,
+      "loss": 0.2384,
+      "step": 5825
+    },
+    {
+      "epoch": 0.522543352601156,
+      "grad_norm": 6.10281002589539,
+      "learning_rate": 1e-06,
+      "loss": 0.2155,
+      "step": 5826
+    },
+    {
+      "epoch": 0.5248554913294797,
+      "grad_norm": 7.316992208136257,
+      "learning_rate": 1e-06,
+      "loss": 0.232,
+      "step": 5827
+    },
+    {
+      "epoch": 0.5271676300578034,
+      "grad_norm": 9.968131123163095,
+      "learning_rate": 1e-06,
+      "loss": 0.2378,
+      "step": 5828
+    },
+    {
+      "epoch": 0.5294797687861271,
+      "grad_norm": 5.5386677160251825,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 5829
+    },
+    {
+      "epoch": 0.5317919075144508,
+      "grad_norm": 8.43665050764315,
+      "learning_rate": 1e-06,
+      "loss": 0.2226,
+      "step": 5830
+    },
+    {
+      "epoch": 0.5341040462427745,
+      "grad_norm": 7.31250669319736,
+      "learning_rate": 1e-06,
+      "loss": 0.2477,
+      "step": 5831
+    },
+    {
+      "epoch": 0.5364161849710982,
+      "grad_norm": 9.619536715185262,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 5832
+    },
+    {
+      "epoch": 0.5387283236994219,
+      "grad_norm": 7.531979515626846,
+      "learning_rate": 1e-06,
+      "loss": 0.2507,
+      "step": 5833
+    },
+    {
+      "epoch": 0.5410404624277456,
+      "grad_norm": 7.953322608664794,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 5834
+    },
+    {
+      "epoch": 0.5433526011560693,
+      "grad_norm": 9.326137428182628,
+      "learning_rate": 1e-06,
+      "loss": 0.2329,
+      "step": 5835
+    },
+    {
+      "epoch": 0.545664739884393,
+      "grad_norm": 6.994573231936536,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 5836
+    },
+    {
+      "epoch": 0.5479768786127167,
+      "grad_norm": 10.11283213673357,
+      "learning_rate": 1e-06,
+      "loss": 0.2358,
+      "step": 5837
+    },
+    {
+      "epoch": 0.5502890173410404,
+      "grad_norm": 8.523006809256405,
+      "learning_rate": 1e-06,
+      "loss": 0.2323,
+      "step": 5838
+    },
+    {
+      "epoch": 0.5526011560693641,
+      "grad_norm": 8.016251385928385,
+      "learning_rate": 1e-06,
+      "loss": 0.2391,
+      "step": 5839
+    },
+    {
+      "epoch": 0.5549132947976878,
+      "grad_norm": 7.146768463514273,
+      "learning_rate": 1e-06,
+      "loss": 0.2254,
+      "step": 5840
+    },
+    {
+      "epoch": 0.5572254335260116,
+      "grad_norm": 6.406649967099804,
+      "learning_rate": 1e-06,
+      "loss": 0.223,
+      "step": 5841
+    },
+    {
+      "epoch": 0.5595375722543353,
+      "grad_norm": 6.310864597041233,
+      "learning_rate": 1e-06,
+      "loss": 0.2203,
+      "step": 5842
+    },
+    {
+      "epoch": 0.561849710982659,
+      "grad_norm": 8.35347920440617,
+      "learning_rate": 1e-06,
+      "loss": 0.2383,
+      "step": 5843
+    },
+    {
+      "epoch": 0.5641618497109827,
+      "grad_norm": 7.508553593716911,
+      "learning_rate": 1e-06,
+      "loss": 0.2259,
+      "step": 5844
+    },
+    {
+      "epoch": 0.5664739884393064,
+      "grad_norm": 7.790039109895294,
+      "learning_rate": 1e-06,
+      "loss": 0.2353,
+      "step": 5845
+    },
+    {
+      "epoch": 0.5687861271676301,
+      "grad_norm": 8.97104315329526,
+      "learning_rate": 1e-06,
+      "loss": 0.234,
+      "step": 5846
+    },
+    {
+      "epoch": 0.5710982658959538,
+      "grad_norm": 7.337510526306941,
+      "learning_rate": 1e-06,
+      "loss": 0.2182,
+      "step": 5847
+    },
+    {
+      "epoch": 0.5734104046242775,
+      "grad_norm": 7.789748672306867,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 5848
+    },
+    {
+      "epoch": 0.5757225433526012,
+      "grad_norm": 6.58750630267285,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 5849
+    },
+    {
+      "epoch": 0.5780346820809249,
+      "grad_norm": 8.404225248874985,
+      "learning_rate": 1e-06,
+      "loss": 0.2251,
+      "step": 5850
+    },
+    {
+      "epoch": 0.5803468208092486,
+      "grad_norm": 7.279010549365793,
+      "learning_rate": 1e-06,
+      "loss": 0.2241,
+      "step": 5851
+    },
+    {
+      "epoch": 0.5826589595375723,
+      "grad_norm": 8.267254045279822,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 5852
+    },
+    {
+      "epoch": 0.584971098265896,
+      "grad_norm": 7.609871327861318,
+      "learning_rate": 1e-06,
+      "loss": 0.2127,
+      "step": 5853
+    },
+    {
+      "epoch": 0.5872832369942197,
+      "grad_norm": 6.966339045536449,
+      "learning_rate": 1e-06,
+      "loss": 0.2165,
+      "step": 5854
+    },
+    {
+      "epoch": 0.5895953757225434,
+      "grad_norm": 7.0381904689215595,
+      "learning_rate": 1e-06,
+      "loss": 0.2091,
+      "step": 5855
+    },
+    {
+      "epoch": 0.5919075144508671,
+      "grad_norm": 7.65169102843695,
+      "learning_rate": 1e-06,
+      "loss": 0.2183,
+      "step": 5856
+    },
+    {
+      "epoch": 0.5942196531791908,
+      "grad_norm": 7.703933862171829,
+      "learning_rate": 1e-06,
+      "loss": 0.224,
+      "step": 5857
+    },
+    {
+      "epoch": 0.5965317919075145,
+      "grad_norm": 6.997430220662777,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 5858
+    },
+    {
+      "epoch": 0.5988439306358382,
+      "grad_norm": 8.476145401525603,
+      "learning_rate": 1e-06,
+      "loss": 0.2155,
+      "step": 5859
+    },
+    {
+      "epoch": 0.6011560693641619,
+      "grad_norm": 7.398941548658004,
+      "learning_rate": 1e-06,
+      "loss": 0.2016,
+      "step": 5860
+    },
+    {
+      "epoch": 0.6034682080924856,
+      "grad_norm": 6.37551091877726,
+      "learning_rate": 1e-06,
+      "loss": 0.2266,
+      "step": 5861
+    },
+    {
+      "epoch": 0.6057803468208093,
+      "grad_norm": 7.051444192216577,
+      "learning_rate": 1e-06,
+      "loss": 0.2426,
+      "step": 5862
+    },
+    {
+      "epoch": 0.608092485549133,
+      "grad_norm": 8.975368013950172,
+      "learning_rate": 1e-06,
+      "loss": 0.2122,
+      "step": 5863
+    },
+    {
+      "epoch": 0.6104046242774567,
+      "grad_norm": 7.719334226811065,
+      "learning_rate": 1e-06,
+      "loss": 0.2381,
+      "step": 5864
+    },
+    {
+      "epoch": 0.6127167630057804,
+      "grad_norm": 8.819735022163892,
+      "learning_rate": 1e-06,
+      "loss": 0.2146,
+      "step": 5865
+    },
+    {
+      "epoch": 0.6150289017341041,
+      "grad_norm": 9.115422212166445,
+      "learning_rate": 1e-06,
+      "loss": 0.2361,
+      "step": 5866
+    },
+    {
+      "epoch": 0.6173410404624278,
+      "grad_norm": 8.322530653805167,
+      "learning_rate": 1e-06,
+      "loss": 0.2302,
+      "step": 5867
+    },
+    {
+      "epoch": 0.6196531791907515,
+      "grad_norm": 7.934483014609534,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 5868
+    },
+    {
+      "epoch": 0.6219653179190752,
+      "grad_norm": 6.812730541584774,
+      "learning_rate": 1e-06,
+      "loss": 0.2212,
+      "step": 5869
+    },
+    {
+      "epoch": 0.6242774566473989,
+      "grad_norm": 7.423971378402763,
+      "learning_rate": 1e-06,
+      "loss": 0.2178,
+      "step": 5870
+    },
+    {
+      "epoch": 0.6265895953757226,
+      "grad_norm": 6.889547209127817,
+      "learning_rate": 1e-06,
+      "loss": 0.2249,
+      "step": 5871
+    },
+    {
+      "epoch": 0.6289017341040463,
+      "grad_norm": 8.422646444960566,
+      "learning_rate": 1e-06,
+      "loss": 0.2378,
+      "step": 5872
+    },
+    {
+      "epoch": 0.63121387283237,
+      "grad_norm": 8.268841002257147,
+      "learning_rate": 1e-06,
+      "loss": 0.2378,
+      "step": 5873
+    },
+    {
+      "epoch": 0.6335260115606937,
+      "grad_norm": 6.370860620813672,
+      "learning_rate": 1e-06,
+      "loss": 0.2169,
+      "step": 5874
+    },
+    {
+      "epoch": 0.6358381502890174,
+      "grad_norm": 7.563066612645879,
+      "learning_rate": 1e-06,
+      "loss": 0.2197,
+      "step": 5875
+    },
+    {
+      "epoch": 0.638150289017341,
+      "grad_norm": 7.062869406608608,
+      "learning_rate": 1e-06,
+      "loss": 0.2273,
+      "step": 5876
+    },
+    {
+      "epoch": 0.6404624277456648,
+      "grad_norm": 9.520759971547388,
+      "learning_rate": 1e-06,
+      "loss": 0.2295,
+      "step": 5877
+    },
+    {
+      "epoch": 0.6427745664739885,
+      "grad_norm": 8.348862426346493,
+      "learning_rate": 1e-06,
+      "loss": 0.2301,
+      "step": 5878
+    },
+    {
+      "epoch": 0.6450867052023121,
+      "grad_norm": 8.274658973509679,
+      "learning_rate": 1e-06,
+      "loss": 0.2363,
+      "step": 5879
+    },
+    {
+      "epoch": 0.6473988439306358,
+      "grad_norm": 7.703047374200983,
+      "learning_rate": 1e-06,
+      "loss": 0.2301,
+      "step": 5880
+    },
+    {
+      "epoch": 0.6497109826589595,
+      "grad_norm": 7.154046729654559,
+      "learning_rate": 1e-06,
+      "loss": 0.2389,
+      "step": 5881
+    },
+    {
+      "epoch": 0.6520231213872832,
+      "grad_norm": 8.319434372571674,
+      "learning_rate": 1e-06,
+      "loss": 0.2111,
+      "step": 5882
+    },
+    {
+      "epoch": 0.6543352601156069,
+      "grad_norm": 7.750085111420353,
+      "learning_rate": 1e-06,
+      "loss": 0.2345,
+      "step": 5883
+    },
+    {
+      "epoch": 0.6566473988439306,
+      "grad_norm": 6.6416375393314455,
+      "learning_rate": 1e-06,
+      "loss": 0.2318,
+      "step": 5884
+    },
+    {
+      "epoch": 0.6589595375722543,
+      "grad_norm": 6.842057460574584,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 5885
+    },
+    {
+      "epoch": 0.661271676300578,
+      "grad_norm": 5.936074591536496,
+      "learning_rate": 1e-06,
+      "loss": 0.2173,
+      "step": 5886
+    },
+    {
+      "epoch": 0.6635838150289017,
+      "grad_norm": 7.9442238249660875,
+      "learning_rate": 1e-06,
+      "loss": 0.2287,
+      "step": 5887
+    },
+    {
+      "epoch": 0.6658959537572254,
+      "grad_norm": 5.7976641583418465,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 5888
+    },
+    {
+      "epoch": 0.6682080924855491,
+      "grad_norm": 7.713464953136145,
+      "learning_rate": 1e-06,
+      "loss": 0.2338,
+      "step": 5889
+    },
+    {
+      "epoch": 0.6705202312138728,
+      "grad_norm": 8.4185938178997,
+      "learning_rate": 1e-06,
+      "loss": 0.2355,
+      "step": 5890
+    },
+    {
+      "epoch": 0.6728323699421965,
+      "grad_norm": 7.381048635664173,
+      "learning_rate": 1e-06,
+      "loss": 0.2261,
+      "step": 5891
+    },
+    {
+      "epoch": 0.6751445086705202,
+      "grad_norm": 7.054637957367755,
+      "learning_rate": 1e-06,
+      "loss": 0.2233,
+      "step": 5892
+    },
+    {
+      "epoch": 0.6774566473988439,
+      "grad_norm": 6.952028739299363,
+      "learning_rate": 1e-06,
+      "loss": 0.2218,
+      "step": 5893
+    },
+    {
+      "epoch": 0.6797687861271676,
+      "grad_norm": 8.70959604152863,
+      "learning_rate": 1e-06,
+      "loss": 0.2214,
+      "step": 5894
+    },
+    {
+      "epoch": 0.6820809248554913,
+      "grad_norm": 8.161552171932492,
+      "learning_rate": 1e-06,
+      "loss": 0.2299,
+      "step": 5895
+    },
+    {
+      "epoch": 0.684393063583815,
+      "grad_norm": 8.706712504898316,
+      "learning_rate": 1e-06,
+      "loss": 0.2386,
+      "step": 5896
+    },
+    {
+      "epoch": 0.6867052023121387,
+      "grad_norm": 7.846256974542348,
+      "learning_rate": 1e-06,
+      "loss": 0.2531,
+      "step": 5897
+    },
+    {
+      "epoch": 0.6890173410404624,
+      "grad_norm": 8.069820554457428,
+      "learning_rate": 1e-06,
+      "loss": 0.2349,
+      "step": 5898
+    },
+    {
+      "epoch": 0.6913294797687861,
+      "grad_norm": 6.071883842238314,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 5899
+    },
+    {
+      "epoch": 0.6936416184971098,
+      "grad_norm": 6.579516237091357,
+      "learning_rate": 1e-06,
+      "loss": 0.2184,
+      "step": 5900
+    },
+    {
+      "epoch": 0.6959537572254335,
+      "grad_norm": 6.833711462351095,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 5901
+    },
+    {
+      "epoch": 0.6982658959537572,
+      "grad_norm": 6.139731885142104,
+      "learning_rate": 1e-06,
+      "loss": 0.2278,
+      "step": 5902
+    },
+    {
+      "epoch": 0.7005780346820809,
+      "grad_norm": 7.6193089588502,
+      "learning_rate": 1e-06,
+      "loss": 0.2281,
+      "step": 5903
+    },
+    {
+      "epoch": 0.7028901734104046,
+      "grad_norm": 7.6076696305566145,
+      "learning_rate": 1e-06,
+      "loss": 0.2424,
+      "step": 5904
+    },
+    {
+      "epoch": 0.7052023121387283,
+      "grad_norm": 9.161043002486466,
+      "learning_rate": 1e-06,
+      "loss": 0.2332,
+      "step": 5905
+    },
+    {
+      "epoch": 0.707514450867052,
+      "grad_norm": 6.802696217669632,
+      "learning_rate": 1e-06,
+      "loss": 0.2299,
+      "step": 5906
+    },
+    {
+      "epoch": 0.7098265895953757,
+      "grad_norm": 9.455571685364045,
+      "learning_rate": 1e-06,
+      "loss": 0.2242,
+      "step": 5907
+    },
+    {
+      "epoch": 0.7121387283236994,
+      "grad_norm": 11.025584514506827,
+      "learning_rate": 1e-06,
+      "loss": 0.2262,
+      "step": 5908
+    },
+    {
+      "epoch": 0.7144508670520231,
+      "grad_norm": 10.111663389892529,
+      "learning_rate": 1e-06,
+      "loss": 0.2289,
+      "step": 5909
+    },
+    {
+      "epoch": 0.7167630057803468,
+      "grad_norm": 7.079485706668731,
+      "learning_rate": 1e-06,
+      "loss": 0.2332,
+      "step": 5910
+    },
+    {
+      "epoch": 0.7190751445086705,
+      "grad_norm": 7.3047622925134625,
+      "learning_rate": 1e-06,
+      "loss": 0.234,
+      "step": 5911
+    },
+    {
+      "epoch": 0.7213872832369942,
+      "grad_norm": 7.090298501745801,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 5912
+    },
+    {
+      "epoch": 0.7236994219653179,
+      "grad_norm": 8.221353476822014,
+      "learning_rate": 1e-06,
+      "loss": 0.2406,
+      "step": 5913
+    },
+    {
+      "epoch": 0.7260115606936416,
+      "grad_norm": 7.62882316372469,
+      "learning_rate": 1e-06,
+      "loss": 0.2097,
+      "step": 5914
+    },
+    {
+      "epoch": 0.7283236994219653,
+      "grad_norm": 7.199555530923558,
+      "learning_rate": 1e-06,
+      "loss": 0.2374,
+      "step": 5915
+    },
+    {
+      "epoch": 0.730635838150289,
+      "grad_norm": 6.124162156914387,
+      "learning_rate": 1e-06,
+      "loss": 0.2438,
+      "step": 5916
+    },
+    {
+      "epoch": 0.7329479768786127,
+      "grad_norm": 6.510471669565751,
+      "learning_rate": 1e-06,
+      "loss": 0.2387,
+      "step": 5917
+    },
+    {
+      "epoch": 0.7352601156069364,
+      "grad_norm": 7.491400062818381,
+      "learning_rate": 1e-06,
+      "loss": 0.2334,
+      "step": 5918
+    },
+    {
+      "epoch": 0.7375722543352601,
+      "grad_norm": 7.428341609618867,
+      "learning_rate": 1e-06,
+      "loss": 0.2319,
+      "step": 5919
+    },
+    {
+      "epoch": 0.7398843930635838,
+      "grad_norm": 8.388782787713863,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 5920
+    },
+    {
+      "epoch": 0.7421965317919075,
+      "grad_norm": 8.252697288122944,
+      "learning_rate": 1e-06,
+      "loss": 0.2232,
+      "step": 5921
+    },
+    {
+      "epoch": 0.7445086705202312,
+      "grad_norm": 6.879665471377449,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 5922
+    },
+    {
+      "epoch": 0.7468208092485549,
+      "grad_norm": 7.398963849312697,
+      "learning_rate": 1e-06,
+      "loss": 0.2383,
+      "step": 5923
+    },
+    {
+      "epoch": 0.7491329479768786,
+      "grad_norm": 7.151144565090044,
+      "learning_rate": 1e-06,
+      "loss": 0.2425,
+      "step": 5924
+    },
+    {
+      "epoch": 0.7514450867052023,
+      "grad_norm": 7.760793110325313,
+      "learning_rate": 1e-06,
+      "loss": 0.2195,
+      "step": 5925
+    },
+    {
+      "epoch": 0.753757225433526,
+      "grad_norm": 7.886475466421215,
+      "learning_rate": 1e-06,
+      "loss": 0.2155,
+      "step": 5926
+    },
+    {
+      "epoch": 0.7560693641618497,
+      "grad_norm": 8.552379267974981,
+      "learning_rate": 1e-06,
+      "loss": 0.2235,
+      "step": 5927
+    },
+    {
+      "epoch": 0.7583815028901734,
+      "grad_norm": 7.992537898746797,
+      "learning_rate": 1e-06,
+      "loss": 0.2228,
+      "step": 5928
+    },
+    {
+      "epoch": 0.7606936416184971,
+      "grad_norm": 6.904195812334142,
+      "learning_rate": 1e-06,
+      "loss": 0.2312,
+      "step": 5929
+    },
+    {
+      "epoch": 0.7630057803468208,
+      "grad_norm": 5.556951302833563,
+      "learning_rate": 1e-06,
+      "loss": 0.2142,
+      "step": 5930
+    },
+    {
+      "epoch": 0.7653179190751445,
+      "grad_norm": 8.520640706690092,
+      "learning_rate": 1e-06,
+      "loss": 0.2261,
+      "step": 5931
+    },
+    {
+      "epoch": 0.7676300578034682,
+      "grad_norm": 8.315190447226225,
+      "learning_rate": 1e-06,
+      "loss": 0.2224,
+      "step": 5932
+    },
+    {
+      "epoch": 0.7699421965317919,
+      "grad_norm": 5.775116872862745,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 5933
+    },
+    {
+      "epoch": 0.7722543352601156,
+      "grad_norm": 7.59386166625805,
+      "learning_rate": 1e-06,
+      "loss": 0.2372,
+      "step": 5934
+    },
+    {
+      "epoch": 0.7745664739884393,
+      "grad_norm": 8.170547497704723,
+      "learning_rate": 1e-06,
+      "loss": 0.2229,
+      "step": 5935
+    },
+    {
+      "epoch": 0.776878612716763,
+      "grad_norm": 7.373081003545884,
+      "learning_rate": 1e-06,
+      "loss": 0.2194,
+      "step": 5936
+    },
+    {
+      "epoch": 0.7791907514450868,
+      "grad_norm": 8.636741850866606,
+      "learning_rate": 1e-06,
+      "loss": 0.2251,
+      "step": 5937
+    },
+    {
+      "epoch": 0.7815028901734105,
+      "grad_norm": 9.461028435714125,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 5938
+    },
+    {
+      "epoch": 0.7838150289017342,
+      "grad_norm": 7.649072791487915,
+      "learning_rate": 1e-06,
+      "loss": 0.2182,
+      "step": 5939
+    },
+    {
+      "epoch": 0.7861271676300579,
+      "grad_norm": 5.929677229527722,
+      "learning_rate": 1e-06,
+      "loss": 0.229,
+      "step": 5940
+    },
+    {
+      "epoch": 0.7884393063583816,
+      "grad_norm": 7.7945602340150755,
+      "learning_rate": 1e-06,
+      "loss": 0.244,
+      "step": 5941
+    },
+    {
+      "epoch": 0.7907514450867053,
+      "grad_norm": 8.736519220518137,
+      "learning_rate": 1e-06,
+      "loss": 0.2302,
+      "step": 5942
+    },
+    {
+      "epoch": 0.793063583815029,
+      "grad_norm": 8.53082920702337,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 5943
+    },
+    {
+      "epoch": 0.7953757225433526,
+      "grad_norm": 7.2396585006750955,
+      "learning_rate": 1e-06,
+      "loss": 0.2313,
+      "step": 5944
+    },
+    {
+      "epoch": 0.7976878612716763,
+      "grad_norm": 7.21702731081427,
+      "learning_rate": 1e-06,
+      "loss": 0.2266,
+      "step": 5945
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 5.889036310619998,
+      "learning_rate": 1e-06,
+      "loss": 0.2208,
+      "step": 5946
+    },
+    {
+      "epoch": 0.8023121387283237,
+      "grad_norm": 7.096010203732887,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5947
+    },
+    {
+      "epoch": 0.8046242774566474,
+      "grad_norm": 8.435568184689336,
+      "learning_rate": 1e-06,
+      "loss": 0.2187,
+      "step": 5948
+    },
+    {
+      "epoch": 0.8069364161849711,
+      "grad_norm": 7.99529954084769,
+      "learning_rate": 1e-06,
+      "loss": 0.2342,
+      "step": 5949
+    },
+    {
+      "epoch": 0.8092485549132948,
+      "grad_norm": 9.019985816918165,
+      "learning_rate": 1e-06,
+      "loss": 0.2361,
+      "step": 5950
+    },
+    {
+      "epoch": 0.8115606936416185,
+      "grad_norm": 7.988539384248091,
+      "learning_rate": 1e-06,
+      "loss": 0.2559,
+      "step": 5951
+    },
+    {
+      "epoch": 0.8138728323699422,
+      "grad_norm": 7.87869632723919,
+      "learning_rate": 1e-06,
+      "loss": 0.2148,
+      "step": 5952
+    },
+    {
+      "epoch": 0.8161849710982659,
+      "grad_norm": 7.982736959117944,
+      "learning_rate": 1e-06,
+      "loss": 0.2396,
+      "step": 5953
+    },
+    {
+      "epoch": 0.8184971098265896,
+      "grad_norm": 9.484601415691582,
+      "learning_rate": 1e-06,
+      "loss": 0.2461,
+      "step": 5954
+    },
+    {
+      "epoch": 0.8208092485549133,
+      "grad_norm": 8.353030529739081,
+      "learning_rate": 1e-06,
+      "loss": 0.2366,
+      "step": 5955
+    },
+    {
+      "epoch": 0.823121387283237,
+      "grad_norm": 7.8695287273828605,
+      "learning_rate": 1e-06,
+      "loss": 0.2199,
+      "step": 5956
+    },
+    {
+      "epoch": 0.8254335260115607,
+      "grad_norm": 6.5961006756819245,
+      "learning_rate": 1e-06,
+      "loss": 0.2279,
+      "step": 5957
+    },
+    {
+      "epoch": 0.8277456647398844,
+      "grad_norm": 9.965442678012597,
+      "learning_rate": 1e-06,
+      "loss": 0.2267,
+      "step": 5958
+    },
+    {
+      "epoch": 0.8300578034682081,
+      "grad_norm": 8.694035834790887,
+      "learning_rate": 1e-06,
+      "loss": 0.238,
+      "step": 5959
+    },
+    {
+      "epoch": 0.8323699421965318,
+      "grad_norm": 7.42478464220277,
+      "learning_rate": 1e-06,
+      "loss": 0.2211,
+      "step": 5960
+    },
+    {
+      "epoch": 0.8346820809248555,
+      "grad_norm": 5.239317606516322,
+      "learning_rate": 1e-06,
+      "loss": 0.2397,
+      "step": 5961
+    },
+    {
+      "epoch": 0.8369942196531792,
+      "grad_norm": 7.935716053968509,
+      "learning_rate": 1e-06,
+      "loss": 0.2281,
+      "step": 5962
+    },
+    {
+      "epoch": 0.8393063583815029,
+      "grad_norm": 7.933699146629049,
+      "learning_rate": 1e-06,
+      "loss": 0.227,
+      "step": 5963
+    },
+    {
+      "epoch": 0.8416184971098266,
+      "grad_norm": 6.807850936247645,
+      "learning_rate": 1e-06,
+      "loss": 0.2256,
+      "step": 5964
+    },
+    {
+      "epoch": 0.8439306358381503,
+      "grad_norm": 6.923279906263713,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 5965
+    },
+    {
+      "epoch": 0.846242774566474,
+      "grad_norm": 6.936255553009109,
+      "learning_rate": 1e-06,
+      "loss": 0.2232,
+      "step": 5966
+    },
+    {
+      "epoch": 0.8485549132947977,
+      "grad_norm": 8.365795676947421,
+      "learning_rate": 1e-06,
+      "loss": 0.25,
+      "step": 5967
+    },
+    {
+      "epoch": 0.8508670520231214,
+      "grad_norm": 7.3951568960074265,
+      "learning_rate": 1e-06,
+      "loss": 0.23,
+      "step": 5968
+    },
+    {
+      "epoch": 0.8531791907514451,
+      "grad_norm": 7.4301095463244184,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 5969
+    },
+    {
+      "epoch": 0.8554913294797688,
+      "grad_norm": 7.554696429473192,
+      "learning_rate": 1e-06,
+      "loss": 0.2247,
+      "step": 5970
+    },
+    {
+      "epoch": 0.8578034682080925,
+      "grad_norm": 7.280541621960601,
+      "learning_rate": 1e-06,
+      "loss": 0.2223,
+      "step": 5971
+    },
+    {
+      "epoch": 0.8601156069364162,
+      "grad_norm": 9.296360844694277,
+      "learning_rate": 1e-06,
+      "loss": 0.236,
+      "step": 5972
+    },
+    {
+      "epoch": 0.8624277456647399,
+      "grad_norm": 8.475786684994693,
+      "learning_rate": 1e-06,
+      "loss": 0.2426,
+      "step": 5973
+    },
+    {
+      "epoch": 0.8647398843930636,
+      "grad_norm": 7.515603765319628,
+      "learning_rate": 1e-06,
+      "loss": 0.2203,
+      "step": 5974
+    },
+    {
+      "epoch": 0.8670520231213873,
+      "grad_norm": 8.55904049125805,
+      "learning_rate": 1e-06,
+      "loss": 0.2327,
+      "step": 5975
+    },
+    {
+      "epoch": 0.869364161849711,
+      "grad_norm": 8.032028952431645,
+      "learning_rate": 1e-06,
+      "loss": 0.2296,
+      "step": 5976
+    },
+    {
+      "epoch": 0.8716763005780347,
+      "grad_norm": 6.570089768338027,
+      "learning_rate": 1e-06,
+      "loss": 0.2124,
+      "step": 5977
+    },
+    {
+      "epoch": 0.8739884393063584,
+      "grad_norm": 7.99475944194107,
+      "learning_rate": 1e-06,
+      "loss": 0.2368,
+      "step": 5978
+    },
+    {
+      "epoch": 0.8763005780346821,
+      "grad_norm": 7.899067842870029,
+      "learning_rate": 1e-06,
+      "loss": 0.2365,
+      "step": 5979
+    },
+    {
+      "epoch": 0.8786127167630058,
+      "grad_norm": 8.810455059492424,
+      "learning_rate": 1e-06,
+      "loss": 0.2237,
+      "step": 5980
+    },
+    {
+      "epoch": 0.8809248554913295,
+      "grad_norm": 6.028443802065488,
+      "learning_rate": 1e-06,
+      "loss": 0.2195,
+      "step": 5981
+    },
+    {
+      "epoch": 0.8832369942196532,
+      "grad_norm": 7.027845754095178,
+      "learning_rate": 1e-06,
+      "loss": 0.235,
+      "step": 5982
+    },
+    {
+      "epoch": 0.8855491329479769,
+      "grad_norm": 5.831817002319046,
+      "learning_rate": 1e-06,
+      "loss": 0.2179,
+      "step": 5983
+    },
+    {
+      "epoch": 0.8878612716763006,
+      "grad_norm": 7.569474344256366,
+      "learning_rate": 1e-06,
+      "loss": 0.2296,
+      "step": 5984
+    },
+    {
+      "epoch": 0.8901734104046243,
+      "grad_norm": 8.8488901542884,
+      "learning_rate": 1e-06,
+      "loss": 0.2404,
+      "step": 5985
+    },
+    {
+      "epoch": 0.892485549132948,
+      "grad_norm": 7.48621027223855,
+      "learning_rate": 1e-06,
+      "loss": 0.2379,
+      "step": 5986
+    },
+    {
+      "epoch": 0.8947976878612717,
+      "grad_norm": 8.457381713892392,
+      "learning_rate": 1e-06,
+      "loss": 0.2309,
+      "step": 5987
+    },
+    {
+      "epoch": 0.8971098265895954,
+      "grad_norm": 6.837329172223216,
+      "learning_rate": 1e-06,
+      "loss": 0.2239,
+      "step": 5988
+    },
+    {
+      "epoch": 0.8994219653179191,
+      "grad_norm": 6.054201772966406,
+      "learning_rate": 1e-06,
+      "loss": 0.2246,
+      "step": 5989
+    },
+    {
+      "epoch": 0.9017341040462428,
+      "grad_norm": 6.620796952568709,
+      "learning_rate": 1e-06,
+      "loss": 0.225,
+      "step": 5990
+    },
+    {
+      "epoch": 0.9040462427745665,
+      "grad_norm": 6.012711066344197,
+      "learning_rate": 1e-06,
+      "loss": 0.2354,
+      "step": 5991
+    },
+    {
+      "epoch": 0.9063583815028902,
+      "grad_norm": 10.249628838541506,
+      "learning_rate": 1e-06,
+      "loss": 0.2385,
+      "step": 5992
+    },
+    {
+      "epoch": 0.9086705202312139,
+      "grad_norm": 6.484258327238793,
+      "learning_rate": 1e-06,
+      "loss": 0.2422,
+      "step": 5993
+    },
+    {
+      "epoch": 0.9109826589595376,
+      "grad_norm": 9.123586218004297,
+      "learning_rate": 1e-06,
+      "loss": 0.2342,
+      "step": 5994
+    },
+    {
+      "epoch": 0.9132947976878613,
+      "grad_norm": 6.153327589484033,
+      "learning_rate": 1e-06,
+      "loss": 0.2166,
+      "step": 5995
+    },
+    {
+      "epoch": 0.915606936416185,
+      "grad_norm": 6.636088549279357,
+      "learning_rate": 1e-06,
+      "loss": 0.2315,
+      "step": 5996
+    },
+    {
+      "epoch": 0.9179190751445087,
+      "grad_norm": 8.548534222953302,
+      "learning_rate": 1e-06,
+      "loss": 0.2316,
+      "step": 5997
+    },
+    {
+      "epoch": 0.9202312138728324,
+      "grad_norm": 7.875800385828306,
+      "learning_rate": 1e-06,
+      "loss": 0.2239,
+      "step": 5998
+    },
+    {
+      "epoch": 0.922543352601156,
+      "grad_norm": 6.79208715339543,
+      "learning_rate": 1e-06,
+      "loss": 0.2314,
+      "step": 5999
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "grad_norm": 7.7221146543425,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9248554913294798,
+      "eval_loss": 3.949756622314453,
+      "eval_runtime": 21.6887,
+      "eval_samples_per_second": 46.107,
+      "eval_steps_per_second": 5.763,
+      "step": 6000
+    },
+    {
+      "epoch": 0.9271676300578034,
+      "grad_norm": 6.440151969201464,
+      "learning_rate": 1e-06,
+      "loss": 0.2134,
+      "step": 6001
+    },
+    {
+      "epoch": 0.9294797687861271,
+      "grad_norm": 6.992015558073699,
+      "learning_rate": 1e-06,
+      "loss": 0.2217,
+      "step": 6002
+    },
+    {
+      "epoch": 0.9317919075144508,
+      "grad_norm": 6.8467059468705855,
+      "learning_rate": 1e-06,
+      "loss": 0.2349,
+      "step": 6003
+    },
+    {
+      "epoch": 0.9341040462427745,
+      "grad_norm": 7.500748664849668,
+      "learning_rate": 1e-06,
+      "loss": 0.2222,
+      "step": 6004
+    },
+    {
+      "epoch": 0.9364161849710982,
+      "grad_norm": 7.508593845099518,
+      "learning_rate": 1e-06,
+      "loss": 0.2321,
+      "step": 6005
+    },
+    {
+      "epoch": 0.9387283236994219,
+      "grad_norm": 8.20277187387753,
+      "learning_rate": 1e-06,
+      "loss": 0.2375,
+      "step": 6006
+    },
+    {
+      "epoch": 0.9410404624277456,
+      "grad_norm": 8.43665748336918,
+      "learning_rate": 1e-06,
+      "loss": 0.2205,
+      "step": 6007
+    },
+    {
+      "epoch": 0.9433526011560693,
+      "grad_norm": 8.555932804090803,
+      "learning_rate": 1e-06,
+      "loss": 0.2265,
+      "step": 6008
+    },
+    {
+      "epoch": 0.945664739884393,
+      "grad_norm": 7.677946849419459,
+      "learning_rate": 1e-06,
+      "loss": 0.2137,
+      "step": 6009
+    },
+    {
+      "epoch": 0.9479768786127167,
+      "grad_norm": 8.950226994331965,
+      "learning_rate": 1e-06,
+      "loss": 0.2359,
+      "step": 6010
+    },
+    {
+      "epoch": 0.9502890173410404,
+      "grad_norm": 6.110764954491366,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 6011
+    },
+    {
+      "epoch": 0.9526011560693641,
+      "grad_norm": 6.072134294647192,
+      "learning_rate": 1e-06,
+      "loss": 0.2312,
+      "step": 6012
+    },
+    {
+      "epoch": 0.9549132947976878,
+      "grad_norm": 7.773515376163373,
+      "learning_rate": 1e-06,
+      "loss": 0.2197,
+      "step": 6013
+    },
+    {
+      "epoch": 0.9572254335260115,
+      "grad_norm": 10.003189148419946,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 6014
+    },
+    {
+      "epoch": 0.9595375722543352,
+      "grad_norm": 8.400269902733722,
+      "learning_rate": 1e-06,
+      "loss": 0.2294,
+      "step": 6015
+    },
+    {
+      "epoch": 0.9618497109826589,
+      "grad_norm": 6.145347777902678,
+      "learning_rate": 1e-06,
+      "loss": 0.217,
+      "step": 6016
+    },
+    {
+      "epoch": 0.9641618497109826,
+      "grad_norm": 8.02201190512869,
+      "learning_rate": 1e-06,
+      "loss": 0.2358,
+      "step": 6017
+    },
+    {
+      "epoch": 0.9664739884393063,
+      "grad_norm": 8.901856548126592,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 6018
+    },
+    {
+      "epoch": 0.96878612716763,
+      "grad_norm": 7.790878317221408,
+      "learning_rate": 1e-06,
+      "loss": 0.2311,
+      "step": 6019
+    },
+    {
+      "epoch": 0.9710982658959537,
+      "grad_norm": 7.008303096366611,
+      "learning_rate": 1e-06,
+      "loss": 0.2399,
+      "step": 6020
+    },
+    {
+      "epoch": 0.9734104046242774,
+      "grad_norm": 7.784839556080759,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 6021
+    },
+    {
+      "epoch": 0.9757225433526011,
+      "grad_norm": 7.913242531953787,
+      "learning_rate": 1e-06,
+      "loss": 0.236,
+      "step": 6022
+    },
+    {
+      "epoch": 0.9780346820809248,
+      "grad_norm": 8.916547354545374,
+      "learning_rate": 1e-06,
+      "loss": 0.2389,
+      "step": 6023
+    },
+    {
+      "epoch": 0.9803468208092485,
+      "grad_norm": 7.267299042012616,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 6024
+    },
+    {
+      "epoch": 0.9826589595375722,
+      "grad_norm": 7.4786812447198585,
+      "learning_rate": 1e-06,
+      "loss": 0.2186,
+      "step": 6025
+    },
+    {
+      "epoch": 0.9849710982658959,
+      "grad_norm": 7.831419486722893,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 6026
+    },
+    {
+      "epoch": 0.9872832369942196,
+      "grad_norm": 7.742206926548804,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 6027
+    },
+    {
+      "epoch": 0.9895953757225433,
+      "grad_norm": 7.923536143748387,
+      "learning_rate": 1e-06,
+      "loss": 0.2467,
+      "step": 6028
+    },
+    {
+      "epoch": 0.991907514450867,
+      "grad_norm": 7.153057777807927,
+      "learning_rate": 1e-06,
+      "loss": 0.2285,
+      "step": 6029
+    },
+    {
+      "epoch": 0.9942196531791907,
+      "grad_norm": 8.742175289961349,
+      "learning_rate": 1e-06,
+      "loss": 0.247,
+      "step": 6030
+    },
+    {
+      "epoch": 0.9965317919075144,
+      "grad_norm": 7.953969772457746,
+      "learning_rate": 1e-06,
+      "loss": 0.2258,
+      "step": 6031
+    },
+    {
+      "epoch": 0.9988439306358381,
+      "grad_norm": 7.218740277501831,
+      "learning_rate": 1e-06,
+      "loss": 0.2232,
+      "step": 6032
+    },
+    {
+      "epoch": 0.9988439306358381,
+      "eval_loss": 3.6173949241638184,
+      "eval_runtime": 22.0333,
+      "eval_samples_per_second": 45.386,
+      "eval_steps_per_second": 5.673,
+      "step": 6032
+    },
+    {
+      "epoch": 0.0023894862604540022,
+      "grad_norm": 6.225016952187947,
+      "learning_rate": 1e-06,
+      "loss": 0.1941,
+      "step": 6033
+    },
+    {
+      "epoch": 0.0047789725209080045,
+      "grad_norm": 6.451225632625604,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 6034
+    },
+    {
+      "epoch": 0.007168458781362007,
+      "grad_norm": 5.703749542892216,
+      "learning_rate": 1e-06,
+      "loss": 0.1919,
+      "step": 6035
+    },
+    {
+      "epoch": 0.009557945041816009,
+      "grad_norm": 6.698608366911937,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 6036
+    },
+    {
+      "epoch": 0.011947431302270013,
+      "grad_norm": 8.654768395643051,
+      "learning_rate": 1e-06,
+      "loss": 0.1869,
+      "step": 6037
+    },
+    {
+      "epoch": 0.014336917562724014,
+      "grad_norm": 7.083526699532923,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 6038
+    },
+    {
+      "epoch": 0.016726403823178016,
+      "grad_norm": 8.548140171652692,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 6039
+    },
+    {
+      "epoch": 0.019115890083632018,
+      "grad_norm": 7.862797624729008,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6040
+    },
+    {
+      "epoch": 0.021505376344086023,
+      "grad_norm": 8.459658622761857,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 6041
+    },
+    {
+      "epoch": 0.023894862604540025,
+      "grad_norm": 8.239662686215445,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 6042
+    },
+    {
+      "epoch": 0.026284348864994027,
+      "grad_norm": 10.664913167698142,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 6043
+    },
+    {
+      "epoch": 0.02867383512544803,
+      "grad_norm": 8.03128549840124,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 6044
+    },
+    {
+      "epoch": 0.03106332138590203,
+      "grad_norm": 8.56679614993886,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 6045
+    },
+    {
+      "epoch": 0.03345280764635603,
+      "grad_norm": 9.594132891820717,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 6046
+    },
+    {
+      "epoch": 0.035842293906810034,
+      "grad_norm": 8.323978480531009,
+      "learning_rate": 1e-06,
+      "loss": 0.1754,
+      "step": 6047
+    },
+    {
+      "epoch": 0.038231780167264036,
+      "grad_norm": 6.553515773710691,
+      "learning_rate": 1e-06,
+      "loss": 0.193,
+      "step": 6048
+    },
+    {
+      "epoch": 0.04062126642771804,
+      "grad_norm": 8.337205962050696,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 6049
+    },
+    {
+      "epoch": 0.043010752688172046,
+      "grad_norm": 7.7995784498531,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 6050
+    },
+    {
+      "epoch": 0.04540023894862605,
+      "grad_norm": 7.213269568207067,
+      "learning_rate": 1e-06,
+      "loss": 0.1989,
+      "step": 6051
+    },
+    {
+      "epoch": 0.04778972520908005,
+      "grad_norm": 8.128558382660577,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6052
+    },
+    {
+      "epoch": 0.05017921146953405,
+      "grad_norm": 9.655271620013322,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 6053
+    },
+    {
+      "epoch": 0.052568697729988054,
+      "grad_norm": 8.278155738340628,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 6054
+    },
+    {
+      "epoch": 0.054958183990442055,
+      "grad_norm": 8.888556115935943,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 6055
+    },
+    {
+      "epoch": 0.05734767025089606,
+      "grad_norm": 8.753410601290959,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 6056
+    },
+    {
+      "epoch": 0.05973715651135006,
+      "grad_norm": 7.596263364864642,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 6057
+    },
+    {
+      "epoch": 0.06212664277180406,
+      "grad_norm": 6.568376410518868,
+      "learning_rate": 1e-06,
+      "loss": 0.197,
+      "step": 6058
+    },
+    {
+      "epoch": 0.06451612903225806,
+      "grad_norm": 7.705998918251214,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 6059
+    },
+    {
+      "epoch": 0.06690561529271206,
+      "grad_norm": 9.11323831144599,
+      "learning_rate": 1e-06,
+      "loss": 0.1972,
+      "step": 6060
+    },
+    {
+      "epoch": 0.06929510155316607,
+      "grad_norm": 6.9891985947305715,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 6061
+    },
+    {
+      "epoch": 0.07168458781362007,
+      "grad_norm": 6.912554592806263,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 6062
+    },
+    {
+      "epoch": 0.07407407407407407,
+      "grad_norm": 7.760573908735408,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 6063
+    },
+    {
+      "epoch": 0.07646356033452807,
+      "grad_norm": 5.634285433095045,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 6064
+    },
+    {
+      "epoch": 0.07885304659498207,
+      "grad_norm": 7.797458663655402,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 6065
+    },
+    {
+      "epoch": 0.08124253285543608,
+      "grad_norm": 8.676428840896806,
+      "learning_rate": 1e-06,
+      "loss": 0.2098,
+      "step": 6066
+    },
+    {
+      "epoch": 0.08363201911589008,
+      "grad_norm": 8.23238898619432,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 6067
+    },
+    {
+      "epoch": 0.08602150537634409,
+      "grad_norm": 8.566614491855699,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 6068
+    },
+    {
+      "epoch": 0.0884109916367981,
+      "grad_norm": 7.347557657298775,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 6069
+    },
+    {
+      "epoch": 0.0908004778972521,
+      "grad_norm": 7.801144450043827,
+      "learning_rate": 1e-06,
+      "loss": 0.1872,
+      "step": 6070
+    },
+    {
+      "epoch": 0.0931899641577061,
+      "grad_norm": 7.815016267984591,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 6071
+    },
+    {
+      "epoch": 0.0955794504181601,
+      "grad_norm": 9.836435297064062,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 6072
+    },
+    {
+      "epoch": 0.0979689366786141,
+      "grad_norm": 8.176127776727927,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6073
+    },
+    {
+      "epoch": 0.1003584229390681,
+      "grad_norm": 6.483966929883785,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 6074
+    },
+    {
+      "epoch": 0.1027479091995221,
+      "grad_norm": 8.46075684908595,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 6075
+    },
+    {
+      "epoch": 0.10513739545997611,
+      "grad_norm": 7.205921356483948,
+      "learning_rate": 1e-06,
+      "loss": 0.1917,
+      "step": 6076
+    },
+    {
+      "epoch": 0.10752688172043011,
+      "grad_norm": 8.375271764302981,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 6077
+    },
+    {
+      "epoch": 0.10991636798088411,
+      "grad_norm": 8.619727709311494,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 6078
+    },
+    {
+      "epoch": 0.11230585424133811,
+      "grad_norm": 6.939061660294519,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6079
+    },
+    {
+      "epoch": 0.11469534050179211,
+      "grad_norm": 7.8804470469930035,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 6080
+    },
+    {
+      "epoch": 0.11708482676224612,
+      "grad_norm": 7.364557899402964,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 6081
+    },
+    {
+      "epoch": 0.11947431302270012,
+      "grad_norm": 9.958306092167918,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 6082
+    },
+    {
+      "epoch": 0.12186379928315412,
+      "grad_norm": 7.607521581317743,
+      "learning_rate": 1e-06,
+      "loss": 0.1895,
+      "step": 6083
+    },
+    {
+      "epoch": 0.12425328554360812,
+      "grad_norm": 9.272315623175244,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 6084
+    },
+    {
+      "epoch": 0.12664277180406214,
+      "grad_norm": 6.194543982538839,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 6085
+    },
+    {
+      "epoch": 0.12903225806451613,
+      "grad_norm": 7.2926075481856,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 6086
+    },
+    {
+      "epoch": 0.13142174432497014,
+      "grad_norm": 8.746925137977572,
+      "learning_rate": 1e-06,
+      "loss": 0.205,
+      "step": 6087
+    },
+    {
+      "epoch": 0.13381123058542413,
+      "grad_norm": 8.783033115453202,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 6088
+    },
+    {
+      "epoch": 0.13620071684587814,
+      "grad_norm": 6.248784799828615,
+      "learning_rate": 1e-06,
+      "loss": 0.2031,
+      "step": 6089
+    },
+    {
+      "epoch": 0.13859020310633213,
+      "grad_norm": 5.751966265269135,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 6090
+    },
+    {
+      "epoch": 0.14097968936678615,
+      "grad_norm": 6.6402981631405495,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 6091
+    },
+    {
+      "epoch": 0.14336917562724014,
+      "grad_norm": 8.625350583629634,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 6092
+    },
+    {
+      "epoch": 0.14575866188769415,
+      "grad_norm": 8.600916193166318,
+      "learning_rate": 1e-06,
+      "loss": 0.2126,
+      "step": 6093
+    },
+    {
+      "epoch": 0.14814814814814814,
+      "grad_norm": 7.771591974974606,
+      "learning_rate": 1e-06,
+      "loss": 0.2028,
+      "step": 6094
+    },
+    {
+      "epoch": 0.15053763440860216,
+      "grad_norm": 7.103887887731394,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 6095
+    },
+    {
+      "epoch": 0.15292712066905614,
+      "grad_norm": 6.587415331019801,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 6096
+    },
+    {
+      "epoch": 0.15531660692951016,
+      "grad_norm": 8.169938551615926,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 6097
+    },
+    {
+      "epoch": 0.15770609318996415,
+      "grad_norm": 8.77254490336821,
+      "learning_rate": 1e-06,
+      "loss": 0.2059,
+      "step": 6098
+    },
+    {
+      "epoch": 0.16009557945041816,
+      "grad_norm": 7.229688139296822,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 6099
+    },
+    {
+      "epoch": 0.16248506571087215,
+      "grad_norm": 17.810153197703322,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 6100
+    },
+    {
+      "epoch": 0.16487455197132617,
+      "grad_norm": 6.069754851598662,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6101
+    },
+    {
+      "epoch": 0.16726403823178015,
+      "grad_norm": 5.772542749200917,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 6102
+    },
+    {
+      "epoch": 0.16965352449223417,
+      "grad_norm": 8.162415302293663,
+      "learning_rate": 1e-06,
+      "loss": 0.1883,
+      "step": 6103
+    },
+    {
+      "epoch": 0.17204301075268819,
+      "grad_norm": 7.6947689763209155,
+      "learning_rate": 1e-06,
+      "loss": 0.1973,
+      "step": 6104
+    },
+    {
+      "epoch": 0.17443249701314217,
+      "grad_norm": 6.517023137598556,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 6105
+    },
+    {
+      "epoch": 0.1768219832735962,
+      "grad_norm": 7.857200855349952,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 6106
+    },
+    {
+      "epoch": 0.17921146953405018,
+      "grad_norm": 9.06105815980428,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 6107
+    },
+    {
+      "epoch": 0.1816009557945042,
+      "grad_norm": 10.236877916398265,
+      "learning_rate": 1e-06,
+      "loss": 0.2217,
+      "step": 6108
+    },
+    {
+      "epoch": 0.18399044205495818,
+      "grad_norm": 7.586024532149726,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6109
+    },
+    {
+      "epoch": 0.1863799283154122,
+      "grad_norm": 10.564936638516446,
+      "learning_rate": 1e-06,
+      "loss": 0.2181,
+      "step": 6110
+    },
+    {
+      "epoch": 0.18876941457586618,
+      "grad_norm": 7.433016477492031,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 6111
+    },
+    {
+      "epoch": 0.1911589008363202,
+      "grad_norm": 8.890079352019384,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 6112
+    },
+    {
+      "epoch": 0.1935483870967742,
+      "grad_norm": 7.319972391590128,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 6113
+    },
+    {
+      "epoch": 0.1959378733572282,
+      "grad_norm": 7.217161273628625,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 6114
+    },
+    {
+      "epoch": 0.1983273596176822,
+      "grad_norm": 7.833185775397633,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 6115
+    },
+    {
+      "epoch": 0.2007168458781362,
+      "grad_norm": 5.984131945200286,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 6116
+    },
+    {
+      "epoch": 0.2031063321385902,
+      "grad_norm": 8.363980562157408,
+      "learning_rate": 1e-06,
+      "loss": 0.187,
+      "step": 6117
+    },
+    {
+      "epoch": 0.2054958183990442,
+      "grad_norm": 8.227484280661386,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 6118
+    },
+    {
+      "epoch": 0.2078853046594982,
+      "grad_norm": 8.678681569993005,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 6119
+    },
+    {
+      "epoch": 0.21027479091995221,
+      "grad_norm": 8.27998658539977,
+      "learning_rate": 1e-06,
+      "loss": 0.2141,
+      "step": 6120
+    },
+    {
+      "epoch": 0.2126642771804062,
+      "grad_norm": 9.192443470882417,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 6121
+    },
+    {
+      "epoch": 0.21505376344086022,
+      "grad_norm": 7.363184852753288,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6122
+    },
+    {
+      "epoch": 0.2174432497013142,
+      "grad_norm": 6.456016776706843,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 6123
+    },
+    {
+      "epoch": 0.21983273596176822,
+      "grad_norm": 8.724587719906488,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 6124
+    },
+    {
+      "epoch": 0.2222222222222222,
+      "grad_norm": 8.035678617443136,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 6125
+    },
+    {
+      "epoch": 0.22461170848267623,
+      "grad_norm": 8.211816354616694,
+      "learning_rate": 1e-06,
+      "loss": 0.1936,
+      "step": 6126
+    },
+    {
+      "epoch": 0.2270011947431302,
+      "grad_norm": 6.556482525688671,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 6127
+    },
+    {
+      "epoch": 0.22939068100358423,
+      "grad_norm": 8.262526032757624,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 6128
+    },
+    {
+      "epoch": 0.23178016726403824,
+      "grad_norm": 6.9681673476569115,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 6129
+    },
+    {
+      "epoch": 0.23416965352449223,
+      "grad_norm": 6.760206525604888,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 6130
+    },
+    {
+      "epoch": 0.23655913978494625,
+      "grad_norm": 8.05098443105192,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6131
+    },
+    {
+      "epoch": 0.23894862604540024,
+      "grad_norm": 10.87402996151681,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 6132
+    },
+    {
+      "epoch": 0.24133811230585425,
+      "grad_norm": 8.367505701345744,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 6133
+    },
+    {
+      "epoch": 0.24372759856630824,
+      "grad_norm": 7.588698917741454,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 6134
+    },
+    {
+      "epoch": 0.24611708482676226,
+      "grad_norm": 6.9947370966521305,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 6135
+    },
+    {
+      "epoch": 0.24850657108721624,
+      "grad_norm": 8.656125646899323,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 6136
+    },
+    {
+      "epoch": 0.25089605734767023,
+      "grad_norm": 6.679705343097145,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 6137
+    },
+    {
+      "epoch": 0.2532855436081243,
+      "grad_norm": 8.131988406312413,
+      "learning_rate": 1e-06,
+      "loss": 0.2085,
+      "step": 6138
+    },
+    {
+      "epoch": 0.25567502986857826,
+      "grad_norm": 8.304368094182125,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 6139
+    },
+    {
+      "epoch": 0.25806451612903225,
+      "grad_norm": 7.7003052546708215,
+      "learning_rate": 1e-06,
+      "loss": 0.1883,
+      "step": 6140
+    },
+    {
+      "epoch": 0.26045400238948624,
+      "grad_norm": 6.753595157601772,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 6141
+    },
+    {
+      "epoch": 0.2628434886499403,
+      "grad_norm": 7.201393550589847,
+      "learning_rate": 1e-06,
+      "loss": 0.1784,
+      "step": 6142
+    },
+    {
+      "epoch": 0.26523297491039427,
+      "grad_norm": 7.955437111812129,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 6143
+    },
+    {
+      "epoch": 0.26762246117084826,
+      "grad_norm": 8.232642399418133,
+      "learning_rate": 1e-06,
+      "loss": 0.2021,
+      "step": 6144
+    },
+    {
+      "epoch": 0.27001194743130225,
+      "grad_norm": 8.76197473824464,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 6145
+    },
+    {
+      "epoch": 0.2724014336917563,
+      "grad_norm": 10.188348391631749,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 6146
+    },
+    {
+      "epoch": 0.2747909199522103,
+      "grad_norm": 9.539075657337936,
+      "learning_rate": 1e-06,
+      "loss": 0.1865,
+      "step": 6147
+    },
+    {
+      "epoch": 0.27718040621266427,
+      "grad_norm": 6.595896800292524,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 6148
+    },
+    {
+      "epoch": 0.27956989247311825,
+      "grad_norm": 6.351246492542711,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 6149
+    },
+    {
+      "epoch": 0.2819593787335723,
+      "grad_norm": 8.210076796441996,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 6150
+    },
+    {
+      "epoch": 0.2843488649940263,
+      "grad_norm": 8.608924104016435,
+      "learning_rate": 1e-06,
+      "loss": 0.2008,
+      "step": 6151
+    },
+    {
+      "epoch": 0.2867383512544803,
+      "grad_norm": 9.36782068387673,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 6152
+    },
+    {
+      "epoch": 0.2891278375149343,
+      "grad_norm": 8.730342445577326,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 6153
+    },
+    {
+      "epoch": 0.2915173237753883,
+      "grad_norm": 6.753258058105709,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6154
+    },
+    {
+      "epoch": 0.2939068100358423,
+      "grad_norm": 6.759389189849049,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 6155
+    },
+    {
+      "epoch": 0.2962962962962963,
+      "grad_norm": 7.680866316729064,
+      "learning_rate": 1e-06,
+      "loss": 0.2064,
+      "step": 6156
+    },
+    {
+      "epoch": 0.2986857825567503,
+      "grad_norm": 6.0512727494478655,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 6157
+    },
+    {
+      "epoch": 0.3010752688172043,
+      "grad_norm": 8.743247750748942,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 6158
+    },
+    {
+      "epoch": 0.3034647550776583,
+      "grad_norm": 8.202747983755215,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 6159
+    },
+    {
+      "epoch": 0.3058542413381123,
+      "grad_norm": 6.4006414717067175,
+      "learning_rate": 1e-06,
+      "loss": 0.2088,
+      "step": 6160
+    },
+    {
+      "epoch": 0.30824372759856633,
+      "grad_norm": 7.3672768172716,
+      "learning_rate": 1e-06,
+      "loss": 0.2121,
+      "step": 6161
+    },
+    {
+      "epoch": 0.3106332138590203,
+      "grad_norm": 8.949114984372187,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 6162
+    },
+    {
+      "epoch": 0.3130227001194743,
+      "grad_norm": 8.119517729123272,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 6163
+    },
+    {
+      "epoch": 0.3154121863799283,
+      "grad_norm": 8.912534187366825,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 6164
+    },
+    {
+      "epoch": 0.31780167264038234,
+      "grad_norm": 7.507154136219722,
+      "learning_rate": 1e-06,
+      "loss": 0.2068,
+      "step": 6165
+    },
+    {
+      "epoch": 0.3201911589008363,
+      "grad_norm": 5.9422629531975035,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 6166
+    },
+    {
+      "epoch": 0.3225806451612903,
+      "grad_norm": 7.859800671159057,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 6167
+    },
+    {
+      "epoch": 0.3249701314217443,
+      "grad_norm": 7.22571164315821,
+      "learning_rate": 1e-06,
+      "loss": 0.2056,
+      "step": 6168
+    },
+    {
+      "epoch": 0.32735961768219835,
+      "grad_norm": 9.44052862515501,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 6169
+    },
+    {
+      "epoch": 0.32974910394265233,
+      "grad_norm": 8.621425858057394,
+      "learning_rate": 1e-06,
+      "loss": 0.2047,
+      "step": 6170
+    },
+    {
+      "epoch": 0.3321385902031063,
+      "grad_norm": 7.217195807733579,
+      "learning_rate": 1e-06,
+      "loss": 0.2019,
+      "step": 6171
+    },
+    {
+      "epoch": 0.3345280764635603,
+      "grad_norm": 7.243868694089476,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 6172
+    },
+    {
+      "epoch": 0.33691756272401435,
+      "grad_norm": 7.49199354835725,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 6173
+    },
+    {
+      "epoch": 0.33930704898446834,
+      "grad_norm": 8.408000022045952,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 6174
+    },
+    {
+      "epoch": 0.34169653524492233,
+      "grad_norm": 7.32650491102878,
+      "learning_rate": 1e-06,
+      "loss": 0.204,
+      "step": 6175
+    },
+    {
+      "epoch": 0.34408602150537637,
+      "grad_norm": 9.173206877086145,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 6176
+    },
+    {
+      "epoch": 0.34647550776583036,
+      "grad_norm": 7.1414525731334155,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 6177
+    },
+    {
+      "epoch": 0.34886499402628435,
+      "grad_norm": 9.33921819034445,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6178
+    },
+    {
+      "epoch": 0.35125448028673834,
+      "grad_norm": 7.291244328107447,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 6179
+    },
+    {
+      "epoch": 0.3536439665471924,
+      "grad_norm": 7.854445988480568,
+      "learning_rate": 1e-06,
+      "loss": 0.1786,
+      "step": 6180
+    },
+    {
+      "epoch": 0.35603345280764637,
+      "grad_norm": 10.153794982441966,
+      "learning_rate": 1e-06,
+      "loss": 0.2119,
+      "step": 6181
+    },
+    {
+      "epoch": 0.35842293906810035,
+      "grad_norm": 7.486923307445408,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 6182
+    },
+    {
+      "epoch": 0.36081242532855434,
+      "grad_norm": 11.137141768532802,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 6183
+    },
+    {
+      "epoch": 0.3632019115890084,
+      "grad_norm": 7.718901660658505,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 6184
+    },
+    {
+      "epoch": 0.3655913978494624,
+      "grad_norm": 7.070662109282803,
+      "learning_rate": 1e-06,
+      "loss": 0.189,
+      "step": 6185
+    },
+    {
+      "epoch": 0.36798088410991636,
+      "grad_norm": 7.38845431301579,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 6186
+    },
+    {
+      "epoch": 0.37037037037037035,
+      "grad_norm": 9.304118823082229,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 6187
+    },
+    {
+      "epoch": 0.3727598566308244,
+      "grad_norm": 7.8331330109926345,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 6188
+    },
+    {
+      "epoch": 0.3751493428912784,
+      "grad_norm": 8.332240587587277,
+      "learning_rate": 1e-06,
+      "loss": 0.1921,
+      "step": 6189
+    },
+    {
+      "epoch": 0.37753882915173237,
+      "grad_norm": 5.906645198647377,
+      "learning_rate": 1e-06,
+      "loss": 0.1857,
+      "step": 6190
+    },
+    {
+      "epoch": 0.37992831541218636,
+      "grad_norm": 7.069884236317658,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 6191
+    },
+    {
+      "epoch": 0.3823178016726404,
+      "grad_norm": 7.409808923521254,
+      "learning_rate": 1e-06,
+      "loss": 0.191,
+      "step": 6192
+    },
+    {
+      "epoch": 0.3847072879330944,
+      "grad_norm": 7.8258249003411,
+      "learning_rate": 1e-06,
+      "loss": 0.2073,
+      "step": 6193
+    },
+    {
+      "epoch": 0.3870967741935484,
+      "grad_norm": 7.491167940119686,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 6194
+    },
+    {
+      "epoch": 0.38948626045400236,
+      "grad_norm": 7.033483163878459,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 6195
+    },
+    {
+      "epoch": 0.3918757467144564,
+      "grad_norm": 7.975479368215356,
+      "learning_rate": 1e-06,
+      "loss": 0.2093,
+      "step": 6196
+    },
+    {
+      "epoch": 0.3942652329749104,
+      "grad_norm": 7.255578681864382,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 6197
+    },
+    {
+      "epoch": 0.3966547192353644,
+      "grad_norm": 8.054641599313639,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6198
+    },
+    {
+      "epoch": 0.39904420549581837,
+      "grad_norm": 8.578116160024065,
+      "learning_rate": 1e-06,
+      "loss": 0.1999,
+      "step": 6199
+    },
+    {
+      "epoch": 0.4014336917562724,
+      "grad_norm": 6.364442747779351,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 6200
+    },
+    {
+      "epoch": 0.4038231780167264,
+      "grad_norm": 7.725530961305735,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 6201
+    },
+    {
+      "epoch": 0.4062126642771804,
+      "grad_norm": 6.65980737746375,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 6202
+    },
+    {
+      "epoch": 0.40860215053763443,
+      "grad_norm": 6.866069877975742,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6203
+    },
+    {
+      "epoch": 0.4109916367980884,
+      "grad_norm": 6.3637065840982485,
+      "learning_rate": 1e-06,
+      "loss": 0.1962,
+      "step": 6204
+    },
+    {
+      "epoch": 0.4133811230585424,
+      "grad_norm": 6.380837926906663,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 6205
+    },
+    {
+      "epoch": 0.4157706093189964,
+      "grad_norm": 6.531676883819588,
+      "learning_rate": 1e-06,
+      "loss": 0.2053,
+      "step": 6206
+    },
+    {
+      "epoch": 0.41816009557945044,
+      "grad_norm": 8.740878336894887,
+      "learning_rate": 1e-06,
+      "loss": 0.2239,
+      "step": 6207
+    },
+    {
+      "epoch": 0.42054958183990443,
+      "grad_norm": 7.773853091441648,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 6208
+    },
+    {
+      "epoch": 0.4229390681003584,
+      "grad_norm": 7.254848059563167,
+      "learning_rate": 1e-06,
+      "loss": 0.1936,
+      "step": 6209
+    },
+    {
+      "epoch": 0.4253285543608124,
+      "grad_norm": 5.792148471563922,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 6210
+    },
+    {
+      "epoch": 0.42771804062126645,
+      "grad_norm": 8.900572349826147,
+      "learning_rate": 1e-06,
+      "loss": 0.2034,
+      "step": 6211
+    },
+    {
+      "epoch": 0.43010752688172044,
+      "grad_norm": 8.105678274910822,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 6212
+    },
+    {
+      "epoch": 0.4324970131421744,
+      "grad_norm": 6.090441216193753,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 6213
+    },
+    {
+      "epoch": 0.4348864994026284,
+      "grad_norm": 9.422952555342444,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 6214
+    },
+    {
+      "epoch": 0.43727598566308246,
+      "grad_norm": 9.480138178399432,
+      "learning_rate": 1e-06,
+      "loss": 0.2122,
+      "step": 6215
+    },
+    {
+      "epoch": 0.43966547192353644,
+      "grad_norm": 7.400535362492091,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 6216
+    },
+    {
+      "epoch": 0.44205495818399043,
+      "grad_norm": 6.762645397243069,
+      "learning_rate": 1e-06,
+      "loss": 0.196,
+      "step": 6217
+    },
+    {
+      "epoch": 0.4444444444444444,
+      "grad_norm": 6.806204000200093,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 6218
+    },
+    {
+      "epoch": 0.44683393070489846,
+      "grad_norm": 5.549992848716701,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 6219
+    },
+    {
+      "epoch": 0.44922341696535245,
+      "grad_norm": 8.996020451646505,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 6220
+    },
+    {
+      "epoch": 0.45161290322580644,
+      "grad_norm": 6.94523975439615,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 6221
+    },
+    {
+      "epoch": 0.4540023894862604,
+      "grad_norm": 8.029553895698328,
+      "learning_rate": 1e-06,
+      "loss": 0.2098,
+      "step": 6222
+    },
+    {
+      "epoch": 0.45639187574671447,
+      "grad_norm": 8.051920678994001,
+      "learning_rate": 1e-06,
+      "loss": 0.2039,
+      "step": 6223
+    },
+    {
+      "epoch": 0.45878136200716846,
+      "grad_norm": 9.91373269323134,
+      "learning_rate": 1e-06,
+      "loss": 0.2035,
+      "step": 6224
+    },
+    {
+      "epoch": 0.46117084826762245,
+      "grad_norm": 6.306569767000794,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 6225
+    },
+    {
+      "epoch": 0.4635603345280765,
+      "grad_norm": 9.408361590452401,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 6226
+    },
+    {
+      "epoch": 0.4659498207885305,
+      "grad_norm": 8.703469535515369,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 6227
+    },
+    {
+      "epoch": 0.46833930704898447,
+      "grad_norm": 9.152525303006012,
+      "learning_rate": 1e-06,
+      "loss": 0.2055,
+      "step": 6228
+    },
+    {
+      "epoch": 0.47072879330943845,
+      "grad_norm": 6.350346601909033,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 6229
+    },
+    {
+      "epoch": 0.4731182795698925,
+      "grad_norm": 7.505462466807896,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 6230
+    },
+    {
+      "epoch": 0.4755077658303465,
+      "grad_norm": 8.731704872103867,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 6231
+    },
+    {
+      "epoch": 0.4778972520908005,
+      "grad_norm": 7.857367880740251,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 6232
+    },
+    {
+      "epoch": 0.48028673835125446,
+      "grad_norm": 7.851299604553452,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 6233
+    },
+    {
+      "epoch": 0.4826762246117085,
+      "grad_norm": 8.454316375488997,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 6234
+    },
+    {
+      "epoch": 0.4850657108721625,
+      "grad_norm": 8.287372661882557,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 6235
+    },
+    {
+      "epoch": 0.4874551971326165,
+      "grad_norm": 6.753312801290767,
+      "learning_rate": 1e-06,
+      "loss": 0.181,
+      "step": 6236
+    },
+    {
+      "epoch": 0.48984468339307047,
+      "grad_norm": 10.190067288905627,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 6237
+    },
+    {
+      "epoch": 0.4922341696535245,
+      "grad_norm": 9.284367432847208,
+      "learning_rate": 1e-06,
+      "loss": 0.2139,
+      "step": 6238
+    },
+    {
+      "epoch": 0.4946236559139785,
+      "grad_norm": 8.822116524271642,
+      "learning_rate": 1e-06,
+      "loss": 0.1928,
+      "step": 6239
+    },
+    {
+      "epoch": 0.4970131421744325,
+      "grad_norm": 8.005566373940454,
+      "learning_rate": 1e-06,
+      "loss": 0.2029,
+      "step": 6240
+    },
+    {
+      "epoch": 0.4994026284348865,
+      "grad_norm": 7.758063979032401,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 6241
+    },
+    {
+      "epoch": 0.5017921146953405,
+      "grad_norm": 9.187810397500318,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6242
+    },
+    {
+      "epoch": 0.5041816009557945,
+      "grad_norm": 7.3891315701647065,
+      "learning_rate": 1e-06,
+      "loss": 0.2106,
+      "step": 6243
+    },
+    {
+      "epoch": 0.5065710872162486,
+      "grad_norm": 7.708948675087596,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 6244
+    },
+    {
+      "epoch": 0.5089605734767025,
+      "grad_norm": 9.132450238371892,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 6245
+    },
+    {
+      "epoch": 0.5113500597371565,
+      "grad_norm": 8.103140378596356,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 6246
+    },
+    {
+      "epoch": 0.5137395459976105,
+      "grad_norm": 8.224015912519365,
+      "learning_rate": 1e-06,
+      "loss": 0.1938,
+      "step": 6247
+    },
+    {
+      "epoch": 0.5161290322580645,
+      "grad_norm": 9.61715424239066,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 6248
+    },
+    {
+      "epoch": 0.5185185185185185,
+      "grad_norm": 5.833968536605606,
+      "learning_rate": 1e-06,
+      "loss": 0.1817,
+      "step": 6249
+    },
+    {
+      "epoch": 0.5209080047789725,
+      "grad_norm": 9.52893844130447,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6250
+    },
+    {
+      "epoch": 0.5232974910394266,
+      "grad_norm": 7.622078041412423,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6251
+    },
+    {
+      "epoch": 0.5256869772998806,
+      "grad_norm": 6.210261334194704,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 6252
+    },
+    {
+      "epoch": 0.5280764635603346,
+      "grad_norm": 8.922654549093284,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 6253
+    },
+    {
+      "epoch": 0.5304659498207885,
+      "grad_norm": 8.654833688398123,
+      "learning_rate": 1e-06,
+      "loss": 0.1926,
+      "step": 6254
+    },
+    {
+      "epoch": 0.5328554360812425,
+      "grad_norm": 9.210072218522546,
+      "learning_rate": 1e-06,
+      "loss": 0.203,
+      "step": 6255
+    },
+    {
+      "epoch": 0.5352449223416965,
+      "grad_norm": 7.221713091977428,
+      "learning_rate": 1e-06,
+      "loss": 0.2002,
+      "step": 6256
+    },
+    {
+      "epoch": 0.5376344086021505,
+      "grad_norm": 9.85829587497377,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 6257
+    },
+    {
+      "epoch": 0.5400238948626045,
+      "grad_norm": 5.78161378526215,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 6258
+    },
+    {
+      "epoch": 0.5424133811230586,
+      "grad_norm": 8.149693805459913,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 6259
+    },
+    {
+      "epoch": 0.5448028673835126,
+      "grad_norm": 7.44695756784942,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 6260
+    },
+    {
+      "epoch": 0.5471923536439666,
+      "grad_norm": 6.4248497358603505,
+      "learning_rate": 1e-06,
+      "loss": 0.2077,
+      "step": 6261
+    },
+    {
+      "epoch": 0.5495818399044206,
+      "grad_norm": 7.901362157178732,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 6262
+    },
+    {
+      "epoch": 0.5519713261648745,
+      "grad_norm": 7.870362593221789,
+      "learning_rate": 1e-06,
+      "loss": 0.2271,
+      "step": 6263
+    },
+    {
+      "epoch": 0.5543608124253285,
+      "grad_norm": 8.752794760789604,
+      "learning_rate": 1e-06,
+      "loss": 0.1942,
+      "step": 6264
+    },
+    {
+      "epoch": 0.5567502986857825,
+      "grad_norm": 7.757553227880324,
+      "learning_rate": 1e-06,
+      "loss": 0.2026,
+      "step": 6265
+    },
+    {
+      "epoch": 0.5591397849462365,
+      "grad_norm": 8.740891929484626,
+      "learning_rate": 1e-06,
+      "loss": 0.2043,
+      "step": 6266
+    },
+    {
+      "epoch": 0.5615292712066906,
+      "grad_norm": 9.885207848748964,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 6267
+    },
+    {
+      "epoch": 0.5639187574671446,
+      "grad_norm": 7.644581011789953,
+      "learning_rate": 1e-06,
+      "loss": 0.2099,
+      "step": 6268
+    },
+    {
+      "epoch": 0.5663082437275986,
+      "grad_norm": 7.1214499109695835,
+      "learning_rate": 1e-06,
+      "loss": 0.2076,
+      "step": 6269
+    },
+    {
+      "epoch": 0.5686977299880526,
+      "grad_norm": 9.502979612552531,
+      "learning_rate": 1e-06,
+      "loss": 0.209,
+      "step": 6270
+    },
+    {
+      "epoch": 0.5710872162485066,
+      "grad_norm": 7.549660628614885,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 6271
+    },
+    {
+      "epoch": 0.5734767025089605,
+      "grad_norm": 9.269729497349307,
+      "learning_rate": 1e-06,
+      "loss": 0.2084,
+      "step": 6272
+    },
+    {
+      "epoch": 0.5758661887694145,
+      "grad_norm": 7.6289316047452465,
+      "learning_rate": 1e-06,
+      "loss": 0.1956,
+      "step": 6273
+    },
+    {
+      "epoch": 0.5782556750298686,
+      "grad_norm": 9.218924222326292,
+      "learning_rate": 1e-06,
+      "loss": 0.1983,
+      "step": 6274
+    },
+    {
+      "epoch": 0.5806451612903226,
+      "grad_norm": 8.851291893735478,
+      "learning_rate": 1e-06,
+      "loss": 0.2078,
+      "step": 6275
+    },
+    {
+      "epoch": 0.5830346475507766,
+      "grad_norm": 7.427063639931895,
+      "learning_rate": 1e-06,
+      "loss": 0.2136,
+      "step": 6276
+    },
+    {
+      "epoch": 0.5854241338112306,
+      "grad_norm": 8.82285962131716,
+      "learning_rate": 1e-06,
+      "loss": 0.1993,
+      "step": 6277
+    },
+    {
+      "epoch": 0.5878136200716846,
+      "grad_norm": 9.032054361112827,
+      "learning_rate": 1e-06,
+      "loss": 0.1799,
+      "step": 6278
+    },
+    {
+      "epoch": 0.5902031063321386,
+      "grad_norm": 8.110067507381695,
+      "learning_rate": 1e-06,
+      "loss": 0.1957,
+      "step": 6279
+    },
+    {
+      "epoch": 0.5925925925925926,
+      "grad_norm": 10.470184313585454,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 6280
+    },
+    {
+      "epoch": 0.5949820788530465,
+      "grad_norm": 6.9782925249017485,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6281
+    },
+    {
+      "epoch": 0.5973715651135006,
+      "grad_norm": 8.460803118307389,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 6282
+    },
+    {
+      "epoch": 0.5997610513739546,
+      "grad_norm": 9.072907135464375,
+      "learning_rate": 1e-06,
+      "loss": 0.2041,
+      "step": 6283
+    },
+    {
+      "epoch": 0.6021505376344086,
+      "grad_norm": 10.14562252916881,
+      "learning_rate": 1e-06,
+      "loss": 0.2108,
+      "step": 6284
+    },
+    {
+      "epoch": 0.6045400238948626,
+      "grad_norm": 6.4746235665177325,
+      "learning_rate": 1e-06,
+      "loss": 0.1996,
+      "step": 6285
+    },
+    {
+      "epoch": 0.6069295101553166,
+      "grad_norm": 9.991569593585425,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 6286
+    },
+    {
+      "epoch": 0.6093189964157706,
+      "grad_norm": 7.259521188955652,
+      "learning_rate": 1e-06,
+      "loss": 0.2003,
+      "step": 6287
+    },
+    {
+      "epoch": 0.6117084826762246,
+      "grad_norm": 6.530104982573435,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 6288
+    },
+    {
+      "epoch": 0.6140979689366786,
+      "grad_norm": 7.158247677235371,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 6289
+    },
+    {
+      "epoch": 0.6164874551971327,
+      "grad_norm": 8.374196324968798,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 6290
+    },
+    {
+      "epoch": 0.6188769414575866,
+      "grad_norm": 7.704221813399698,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 6291
+    },
+    {
+      "epoch": 0.6212664277180406,
+      "grad_norm": 9.788452190693787,
+      "learning_rate": 1e-06,
+      "loss": 0.2005,
+      "step": 6292
+    },
+    {
+      "epoch": 0.6236559139784946,
+      "grad_norm": 8.058321945563513,
+      "learning_rate": 1e-06,
+      "loss": 0.2057,
+      "step": 6293
+    },
+    {
+      "epoch": 0.6260454002389486,
+      "grad_norm": 7.495284757123306,
+      "learning_rate": 1e-06,
+      "loss": 0.2014,
+      "step": 6294
+    },
+    {
+      "epoch": 0.6284348864994026,
+      "grad_norm": 4.952648613855383,
+      "learning_rate": 1e-06,
+      "loss": 0.2049,
+      "step": 6295
+    },
+    {
+      "epoch": 0.6308243727598566,
+      "grad_norm": 7.676396148063367,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 6296
+    },
+    {
+      "epoch": 0.6332138590203107,
+      "grad_norm": 8.98817199724069,
+      "learning_rate": 1e-06,
+      "loss": 0.2145,
+      "step": 6297
+    },
+    {
+      "epoch": 0.6356033452807647,
+      "grad_norm": 6.759988875582247,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 6298
+    },
+    {
+      "epoch": 0.6379928315412187,
+      "grad_norm": 8.89917668126906,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 6299
+    },
+    {
+      "epoch": 0.6403823178016727,
+      "grad_norm": 7.993981793106322,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 6300
+    },
+    {
+      "epoch": 0.6427718040621266,
+      "grad_norm": 8.81479542201741,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 6301
+    },
+    {
+      "epoch": 0.6451612903225806,
+      "grad_norm": 8.844271018022905,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 6302
+    },
+    {
+      "epoch": 0.6475507765830346,
+      "grad_norm": 6.926257434702635,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6303
+    },
+    {
+      "epoch": 0.6499402628434886,
+      "grad_norm": 9.216620053655827,
+      "learning_rate": 1e-06,
+      "loss": 0.1968,
+      "step": 6304
+    },
+    {
+      "epoch": 0.6523297491039427,
+      "grad_norm": 8.26176354441775,
+      "learning_rate": 1e-06,
+      "loss": 0.2221,
+      "step": 6305
+    },
+    {
+      "epoch": 0.6547192353643967,
+      "grad_norm": 8.413883235011223,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 6306
+    },
+    {
+      "epoch": 0.6571087216248507,
+      "grad_norm": 7.031213249981393,
+      "learning_rate": 1e-06,
+      "loss": 0.1929,
+      "step": 6307
+    },
+    {
+      "epoch": 0.6594982078853047,
+      "grad_norm": 7.911527688839324,
+      "learning_rate": 1e-06,
+      "loss": 0.1834,
+      "step": 6308
+    },
+    {
+      "epoch": 0.6618876941457587,
+      "grad_norm": 8.003421265997831,
+      "learning_rate": 1e-06,
+      "loss": 0.2094,
+      "step": 6309
+    },
+    {
+      "epoch": 0.6642771804062126,
+      "grad_norm": 7.612522512222637,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 6310
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 7.477728628368986,
+      "learning_rate": 1e-06,
+      "loss": 0.1984,
+      "step": 6311
+    },
+    {
+      "epoch": 0.6690561529271206,
+      "grad_norm": 10.525817837626253,
+      "learning_rate": 1e-06,
+      "loss": 0.2206,
+      "step": 6312
+    },
+    {
+      "epoch": 0.6714456391875747,
+      "grad_norm": 7.342824149676446,
+      "learning_rate": 1e-06,
+      "loss": 0.1991,
+      "step": 6313
+    },
+    {
+      "epoch": 0.6738351254480287,
+      "grad_norm": 7.278583501435414,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 6314
+    },
+    {
+      "epoch": 0.6762246117084827,
+      "grad_norm": 7.112087879244279,
+      "learning_rate": 1e-06,
+      "loss": 0.1986,
+      "step": 6315
+    },
+    {
+      "epoch": 0.6786140979689367,
+      "grad_norm": 7.267288556823423,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 6316
+    },
+    {
+      "epoch": 0.6810035842293907,
+      "grad_norm": 7.568221690556007,
+      "learning_rate": 1e-06,
+      "loss": 0.1915,
+      "step": 6317
+    },
+    {
+      "epoch": 0.6833930704898447,
+      "grad_norm": 7.561064574443383,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 6318
+    },
+    {
+      "epoch": 0.6857825567502986,
+      "grad_norm": 7.68773272493295,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 6319
+    },
+    {
+      "epoch": 0.6881720430107527,
+      "grad_norm": 8.996256155752866,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 6320
+    },
+    {
+      "epoch": 0.6905615292712067,
+      "grad_norm": 6.417442036818004,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6321
+    },
+    {
+      "epoch": 0.6929510155316607,
+      "grad_norm": 7.542642056040704,
+      "learning_rate": 1e-06,
+      "loss": 0.1998,
+      "step": 6322
+    },
+    {
+      "epoch": 0.6953405017921147,
+      "grad_norm": 7.8515034814014335,
+      "learning_rate": 1e-06,
+      "loss": 0.2061,
+      "step": 6323
+    },
+    {
+      "epoch": 0.6977299880525687,
+      "grad_norm": 7.608744429095273,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 6324
+    },
+    {
+      "epoch": 0.7001194743130227,
+      "grad_norm": 7.335839672383411,
+      "learning_rate": 1e-06,
+      "loss": 0.1954,
+      "step": 6325
+    },
+    {
+      "epoch": 0.7025089605734767,
+      "grad_norm": 7.638759988909906,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 6326
+    },
+    {
+      "epoch": 0.7048984468339307,
+      "grad_norm": 8.019514368004527,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 6327
+    },
+    {
+      "epoch": 0.7072879330943848,
+      "grad_norm": 10.418754402610164,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 6328
+    },
+    {
+      "epoch": 0.7096774193548387,
+      "grad_norm": 7.015833652779731,
+      "learning_rate": 1e-06,
+      "loss": 0.2063,
+      "step": 6329
+    },
+    {
+      "epoch": 0.7120669056152927,
+      "grad_norm": 7.194410451803741,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 6330
+    },
+    {
+      "epoch": 0.7144563918757467,
+      "grad_norm": 9.747671786048395,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 6331
+    },
+    {
+      "epoch": 0.7168458781362007,
+      "grad_norm": 8.036987904726338,
+      "learning_rate": 1e-06,
+      "loss": 0.2023,
+      "step": 6332
+    },
+    {
+      "epoch": 0.7192353643966547,
+      "grad_norm": 9.095307034558338,
+      "learning_rate": 1e-06,
+      "loss": 0.199,
+      "step": 6333
+    },
+    {
+      "epoch": 0.7216248506571087,
+      "grad_norm": 8.537776528675366,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 6334
+    },
+    {
+      "epoch": 0.7240143369175627,
+      "grad_norm": 7.616580782609059,
+      "learning_rate": 1e-06,
+      "loss": 0.2048,
+      "step": 6335
+    },
+    {
+      "epoch": 0.7264038231780168,
+      "grad_norm": 8.315387533266376,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 6336
+    },
+    {
+      "epoch": 0.7287933094384708,
+      "grad_norm": 8.609702413635723,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 6337
+    },
+    {
+      "epoch": 0.7311827956989247,
+      "grad_norm": 9.339359221626209,
+      "learning_rate": 1e-06,
+      "loss": 0.2044,
+      "step": 6338
+    },
+    {
+      "epoch": 0.7335722819593787,
+      "grad_norm": 8.451179267818574,
+      "learning_rate": 1e-06,
+      "loss": 0.1934,
+      "step": 6339
+    },
+    {
+      "epoch": 0.7359617682198327,
+      "grad_norm": 7.313954774947176,
+      "learning_rate": 1e-06,
+      "loss": 0.1895,
+      "step": 6340
+    },
+    {
+      "epoch": 0.7383512544802867,
+      "grad_norm": 9.42498845865525,
+      "learning_rate": 1e-06,
+      "loss": 0.2051,
+      "step": 6341
+    },
+    {
+      "epoch": 0.7407407407407407,
+      "grad_norm": 6.60760647281618,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 6342
+    },
+    {
+      "epoch": 0.7431302270011948,
+      "grad_norm": 8.372573833462274,
+      "learning_rate": 1e-06,
+      "loss": 0.1964,
+      "step": 6343
+    },
+    {
+      "epoch": 0.7455197132616488,
+      "grad_norm": 7.134139513700308,
+      "learning_rate": 1e-06,
+      "loss": 0.1907,
+      "step": 6344
+    },
+    {
+      "epoch": 0.7479091995221028,
+      "grad_norm": 8.221556164353029,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6345
+    },
+    {
+      "epoch": 0.7502986857825568,
+      "grad_norm": 7.7886075841263045,
+      "learning_rate": 1e-06,
+      "loss": 0.2117,
+      "step": 6346
+    },
+    {
+      "epoch": 0.7526881720430108,
+      "grad_norm": 9.043089660123032,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 6347
+    },
+    {
+      "epoch": 0.7550776583034647,
+      "grad_norm": 6.459611592015846,
+      "learning_rate": 1e-06,
+      "loss": 0.219,
+      "step": 6348
+    },
+    {
+      "epoch": 0.7574671445639187,
+      "grad_norm": 7.089174059037784,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 6349
+    },
+    {
+      "epoch": 0.7598566308243727,
+      "grad_norm": 6.987126203439306,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 6350
+    },
+    {
+      "epoch": 0.7622461170848268,
+      "grad_norm": 6.593370060908807,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 6351
+    },
+    {
+      "epoch": 0.7646356033452808,
+      "grad_norm": 9.365939816650664,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 6352
+    },
+    {
+      "epoch": 0.7670250896057348,
+      "grad_norm": 8.86186747194302,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 6353
+    },
+    {
+      "epoch": 0.7694145758661888,
+      "grad_norm": 8.940833454126901,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6354
+    },
+    {
+      "epoch": 0.7718040621266428,
+      "grad_norm": 8.809288683815417,
+      "learning_rate": 1e-06,
+      "loss": 0.1988,
+      "step": 6355
+    },
+    {
+      "epoch": 0.7741935483870968,
+      "grad_norm": 8.17980901712758,
+      "learning_rate": 1e-06,
+      "loss": 0.2001,
+      "step": 6356
+    },
+    {
+      "epoch": 0.7765830346475507,
+      "grad_norm": 7.651189156952092,
+      "learning_rate": 1e-06,
+      "loss": 0.2082,
+      "step": 6357
+    },
+    {
+      "epoch": 0.7789725209080047,
+      "grad_norm": 8.631342794651433,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 6358
+    },
+    {
+      "epoch": 0.7813620071684588,
+      "grad_norm": 7.8686591559199215,
+      "learning_rate": 1e-06,
+      "loss": 0.2081,
+      "step": 6359
+    },
+    {
+      "epoch": 0.7837514934289128,
+      "grad_norm": 6.618356336850333,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 6360
+    },
+    {
+      "epoch": 0.7861409796893668,
+      "grad_norm": 7.0251593395616,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 6361
+    },
+    {
+      "epoch": 0.7885304659498208,
+      "grad_norm": 9.134990995155729,
+      "learning_rate": 1e-06,
+      "loss": 0.2065,
+      "step": 6362
+    },
+    {
+      "epoch": 0.7909199522102748,
+      "grad_norm": 7.005358421542621,
+      "learning_rate": 1e-06,
+      "loss": 0.2007,
+      "step": 6363
+    },
+    {
+      "epoch": 0.7933094384707288,
+      "grad_norm": 7.471748411440643,
+      "learning_rate": 1e-06,
+      "loss": 0.2176,
+      "step": 6364
+    },
+    {
+      "epoch": 0.7956989247311828,
+      "grad_norm": 8.187663507131331,
+      "learning_rate": 1e-06,
+      "loss": 0.2089,
+      "step": 6365
+    },
+    {
+      "epoch": 0.7980884109916367,
+      "grad_norm": 8.162716673574748,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 6366
+    },
+    {
+      "epoch": 0.8004778972520908,
+      "grad_norm": 7.931856470164117,
+      "learning_rate": 1e-06,
+      "loss": 0.1977,
+      "step": 6367
+    },
+    {
+      "epoch": 0.8028673835125448,
+      "grad_norm": 6.322343576788603,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6368
+    },
+    {
+      "epoch": 0.8052568697729988,
+      "grad_norm": 7.518729082954533,
+      "learning_rate": 1e-06,
+      "loss": 0.2231,
+      "step": 6369
+    },
+    {
+      "epoch": 0.8076463560334528,
+      "grad_norm": 7.30906504365564,
+      "learning_rate": 1e-06,
+      "loss": 0.2196,
+      "step": 6370
+    },
+    {
+      "epoch": 0.8100358422939068,
+      "grad_norm": 6.681993682661756,
+      "learning_rate": 1e-06,
+      "loss": 0.2015,
+      "step": 6371
+    },
+    {
+      "epoch": 0.8124253285543608,
+      "grad_norm": 6.316935964953277,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 6372
+    },
+    {
+      "epoch": 0.8148148148148148,
+      "grad_norm": 9.343935167419902,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 6373
+    },
+    {
+      "epoch": 0.8172043010752689,
+      "grad_norm": 9.497166988559837,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 6374
+    },
+    {
+      "epoch": 0.8195937873357229,
+      "grad_norm": 6.964018895089961,
+      "learning_rate": 1e-06,
+      "loss": 0.1912,
+      "step": 6375
+    },
+    {
+      "epoch": 0.8219832735961768,
+      "grad_norm": 7.538715176259326,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 6376
+    },
+    {
+      "epoch": 0.8243727598566308,
+      "grad_norm": 7.804377727640994,
+      "learning_rate": 1e-06,
+      "loss": 0.2017,
+      "step": 6377
+    },
+    {
+      "epoch": 0.8267622461170848,
+      "grad_norm": 8.05762061000939,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 6378
+    },
+    {
+      "epoch": 0.8291517323775388,
+      "grad_norm": 8.82992467763268,
+      "learning_rate": 1e-06,
+      "loss": 0.2105,
+      "step": 6379
+    },
+    {
+      "epoch": 0.8315412186379928,
+      "grad_norm": 7.481941159346659,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 6380
+    },
+    {
+      "epoch": 0.8339307048984468,
+      "grad_norm": 8.310962492319966,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 6381
+    },
+    {
+      "epoch": 0.8363201911589009,
+      "grad_norm": 7.125925573006734,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 6382
+    },
+    {
+      "epoch": 0.8387096774193549,
+      "grad_norm": 6.249501577699998,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 6383
+    },
+    {
+      "epoch": 0.8410991636798089,
+      "grad_norm": 7.2147953666951645,
+      "learning_rate": 1e-06,
+      "loss": 0.1975,
+      "step": 6384
+    },
+    {
+      "epoch": 0.8434886499402628,
+      "grad_norm": 7.4103516729865655,
+      "learning_rate": 1e-06,
+      "loss": 0.1804,
+      "step": 6385
+    },
+    {
+      "epoch": 0.8458781362007168,
+      "grad_norm": 7.5326515955960955,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 6386
+    },
+    {
+      "epoch": 0.8482676224611708,
+      "grad_norm": 8.11106174204325,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 6387
+    },
+    {
+      "epoch": 0.8506571087216248,
+      "grad_norm": 7.47258094611564,
+      "learning_rate": 1e-06,
+      "loss": 0.1994,
+      "step": 6388
+    },
+    {
+      "epoch": 0.8530465949820788,
+      "grad_norm": 9.471084858198223,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 6389
+    },
+    {
+      "epoch": 0.8554360812425329,
+      "grad_norm": 8.997645833385434,
+      "learning_rate": 1e-06,
+      "loss": 0.2114,
+      "step": 6390
+    },
+    {
+      "epoch": 0.8578255675029869,
+      "grad_norm": 6.861322741721173,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6391
+    },
+    {
+      "epoch": 0.8602150537634409,
+      "grad_norm": 8.947740208093585,
+      "learning_rate": 1e-06,
+      "loss": 0.2107,
+      "step": 6392
+    },
+    {
+      "epoch": 0.8626045400238949,
+      "grad_norm": 8.301701026453777,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 6393
+    },
+    {
+      "epoch": 0.8649940262843488,
+      "grad_norm": 8.927067505096216,
+      "learning_rate": 1e-06,
+      "loss": 0.2,
+      "step": 6394
+    },
+    {
+      "epoch": 0.8673835125448028,
+      "grad_norm": 7.53318267452126,
+      "learning_rate": 1e-06,
+      "loss": 0.1943,
+      "step": 6395
+    },
+    {
+      "epoch": 0.8697729988052568,
+      "grad_norm": 8.995345663832765,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 6396
+    },
+    {
+      "epoch": 0.8721624850657109,
+      "grad_norm": 8.096405239944,
+      "learning_rate": 1e-06,
+      "loss": 0.1974,
+      "step": 6397
+    },
+    {
+      "epoch": 0.8745519713261649,
+      "grad_norm": 7.756289835018832,
+      "learning_rate": 1e-06,
+      "loss": 0.1995,
+      "step": 6398
+    },
+    {
+      "epoch": 0.8769414575866189,
+      "grad_norm": 7.190144479973087,
+      "learning_rate": 1e-06,
+      "loss": 0.2033,
+      "step": 6399
+    },
+    {
+      "epoch": 0.8793309438470729,
+      "grad_norm": 8.054523621352889,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 6400
+    },
+    {
+      "epoch": 0.8817204301075269,
+      "grad_norm": 9.646675327312147,
+      "learning_rate": 1e-06,
+      "loss": 0.2032,
+      "step": 6401
+    },
+    {
+      "epoch": 0.8841099163679809,
+      "grad_norm": 5.917694478448302,
+      "learning_rate": 1e-06,
+      "loss": 0.1966,
+      "step": 6402
+    },
+    {
+      "epoch": 0.8864994026284349,
+      "grad_norm": 7.463459523052286,
+      "learning_rate": 1e-06,
+      "loss": 0.2046,
+      "step": 6403
+    },
+    {
+      "epoch": 0.8888888888888888,
+      "grad_norm": 8.132065409310092,
+      "learning_rate": 1e-06,
+      "loss": 0.1997,
+      "step": 6404
+    },
+    {
+      "epoch": 0.8912783751493429,
+      "grad_norm": 7.199430243885,
+      "learning_rate": 1e-06,
+      "loss": 0.209,
+      "step": 6405
+    },
+    {
+      "epoch": 0.8936678614097969,
+      "grad_norm": 8.042117447440695,
+      "learning_rate": 1e-06,
+      "loss": 0.2025,
+      "step": 6406
+    },
+    {
+      "epoch": 0.8960573476702509,
+      "grad_norm": 7.283248777874938,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6407
+    },
+    {
+      "epoch": 0.8984468339307049,
+      "grad_norm": 7.831672411212561,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 6408
+    },
+    {
+      "epoch": 0.9008363201911589,
+      "grad_norm": 6.990589547906707,
+      "learning_rate": 1e-06,
+      "loss": 0.1965,
+      "step": 6409
+    },
+    {
+      "epoch": 0.9032258064516129,
+      "grad_norm": 8.914652724622627,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 6410
+    },
+    {
+      "epoch": 0.9056152927120669,
+      "grad_norm": 6.07718202208154,
+      "learning_rate": 1e-06,
+      "loss": 0.1881,
+      "step": 6411
+    },
+    {
+      "epoch": 0.9080047789725209,
+      "grad_norm": 5.923250129309763,
+      "learning_rate": 1e-06,
+      "loss": 0.1821,
+      "step": 6412
+    },
+    {
+      "epoch": 0.910394265232975,
+      "grad_norm": 9.091867465307047,
+      "learning_rate": 1e-06,
+      "loss": 0.201,
+      "step": 6413
+    },
+    {
+      "epoch": 0.9127837514934289,
+      "grad_norm": 8.134491650276976,
+      "learning_rate": 1e-06,
+      "loss": 0.194,
+      "step": 6414
+    },
+    {
+      "epoch": 0.9151732377538829,
+      "grad_norm": 7.270198421383376,
+      "learning_rate": 1e-06,
+      "loss": 0.2071,
+      "step": 6415
+    },
+    {
+      "epoch": 0.9175627240143369,
+      "grad_norm": 7.652793282397852,
+      "learning_rate": 1e-06,
+      "loss": 0.1958,
+      "step": 6416
+    },
+    {
+      "epoch": 0.9199522102747909,
+      "grad_norm": 7.609883817192814,
+      "learning_rate": 1e-06,
+      "loss": 0.2112,
+      "step": 6417
+    },
+    {
+      "epoch": 0.9223416965352449,
+      "grad_norm": 10.02167069466345,
+      "learning_rate": 1e-06,
+      "loss": 0.2064,
+      "step": 6418
+    },
+    {
+      "epoch": 0.9247311827956989,
+      "grad_norm": 8.69303939171949,
+      "learning_rate": 1e-06,
+      "loss": 0.2189,
+      "step": 6419
+    },
+    {
+      "epoch": 0.927120669056153,
+      "grad_norm": 7.327550404329366,
+      "learning_rate": 1e-06,
+      "loss": 0.2087,
+      "step": 6420
+    },
+    {
+      "epoch": 0.929510155316607,
+      "grad_norm": 9.226699166886632,
+      "learning_rate": 1e-06,
+      "loss": 0.1953,
+      "step": 6421
+    },
+    {
+      "epoch": 0.931899641577061,
+      "grad_norm": 7.682341015343721,
+      "learning_rate": 1e-06,
+      "loss": 0.1974,
+      "step": 6422
+    },
+    {
+      "epoch": 0.9342891278375149,
+      "grad_norm": 9.547583048334555,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 6423
+    },
+    {
+      "epoch": 0.9366786140979689,
+      "grad_norm": 7.91652934899933,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 6424
+    },
+    {
+      "epoch": 0.9390681003584229,
+      "grad_norm": 9.983984541715769,
+      "learning_rate": 1e-06,
+      "loss": 0.2162,
+      "step": 6425
+    },
+    {
+      "epoch": 0.9414575866188769,
+      "grad_norm": 8.05761875850231,
+      "learning_rate": 1e-06,
+      "loss": 0.2022,
+      "step": 6426
+    },
+    {
+      "epoch": 0.9438470728793309,
+      "grad_norm": 6.817259477822241,
+      "learning_rate": 1e-06,
+      "loss": 0.1889,
+      "step": 6427
+    },
+    {
+      "epoch": 0.946236559139785,
+      "grad_norm": 5.110270617475638,
+      "learning_rate": 1e-06,
+      "loss": 0.208,
+      "step": 6428
+    },
+    {
+      "epoch": 0.948626045400239,
+      "grad_norm": 6.132327042943566,
+      "learning_rate": 1e-06,
+      "loss": 0.1922,
+      "step": 6429
+    },
+    {
+      "epoch": 0.951015531660693,
+      "grad_norm": 7.5501926309741805,
+      "learning_rate": 1e-06,
+      "loss": 0.2037,
+      "step": 6430
+    },
+    {
+      "epoch": 0.953405017921147,
+      "grad_norm": 7.989137079592959,
+      "learning_rate": 1e-06,
+      "loss": 0.1949,
+      "step": 6431
+    },
+    {
+      "epoch": 0.955794504181601,
+      "grad_norm": 8.483561641639994,
+      "learning_rate": 1e-06,
+      "loss": 0.2138,
+      "step": 6432
+    },
+    {
+      "epoch": 0.9581839904420549,
+      "grad_norm": 8.350372936013496,
+      "learning_rate": 1e-06,
+      "loss": 0.1866,
+      "step": 6433
+    },
+    {
+      "epoch": 0.9605734767025089,
+      "grad_norm": 8.080020141330118,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 6434
+    },
+    {
+      "epoch": 0.9629629629629629,
+      "grad_norm": 8.954777693310891,
+      "learning_rate": 1e-06,
+      "loss": 0.1933,
+      "step": 6435
+    },
+    {
+      "epoch": 0.965352449223417,
+      "grad_norm": 7.01497775546441,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 6436
+    },
+    {
+      "epoch": 0.967741935483871,
+      "grad_norm": 7.511162618728521,
+      "learning_rate": 1e-06,
+      "loss": 0.1771,
+      "step": 6437
+    },
+    {
+      "epoch": 0.970131421744325,
+      "grad_norm": 9.4739487018235,
+      "learning_rate": 1e-06,
+      "loss": 0.2172,
+      "step": 6438
+    },
+    {
+      "epoch": 0.972520908004779,
+      "grad_norm": 9.190249713810198,
+      "learning_rate": 1e-06,
+      "loss": 0.2054,
+      "step": 6439
+    },
+    {
+      "epoch": 0.974910394265233,
+      "grad_norm": 8.266261719234432,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 6440
+    },
+    {
+      "epoch": 0.977299880525687,
+      "grad_norm": 7.203438304104611,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 6441
+    },
+    {
+      "epoch": 0.9796893667861409,
+      "grad_norm": 7.458389210473639,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 6442
+    },
+    {
+      "epoch": 0.982078853046595,
+      "grad_norm": 8.305534023812093,
+      "learning_rate": 1e-06,
+      "loss": 0.1931,
+      "step": 6443
+    },
+    {
+      "epoch": 0.984468339307049,
+      "grad_norm": 7.396515836867875,
+      "learning_rate": 1e-06,
+      "loss": 0.2221,
+      "step": 6444
+    },
+    {
+      "epoch": 0.986857825567503,
+      "grad_norm": 7.138858391250131,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 6445
+    },
+    {
+      "epoch": 0.989247311827957,
+      "grad_norm": 8.072235759369043,
+      "learning_rate": 1e-06,
+      "loss": 0.2052,
+      "step": 6446
+    },
+    {
+      "epoch": 0.991636798088411,
+      "grad_norm": 8.391955961494357,
+      "learning_rate": 1e-06,
+      "loss": 0.2096,
+      "step": 6447
+    },
+    {
+      "epoch": 0.994026284348865,
+      "grad_norm": 9.001228630421249,
+      "learning_rate": 1e-06,
+      "loss": 0.1936,
+      "step": 6448
+    },
+    {
+      "epoch": 0.996415770609319,
+      "grad_norm": 6.5384913900157215,
+      "learning_rate": 1e-06,
+      "loss": 0.2004,
+      "step": 6449
+    },
+    {
+      "epoch": 0.998805256869773,
+      "grad_norm": 9.09436486410846,
+      "learning_rate": 1e-06,
+      "loss": 0.1925,
+      "step": 6450
+    },
+    {
+      "epoch": 0.998805256869773,
+      "eval_loss": 3.705120325088501,
+      "eval_runtime": 22.0352,
+      "eval_samples_per_second": 45.382,
+      "eval_steps_per_second": 5.673,
+      "step": 6450
+    },
+    {
+      "epoch": 0.0030911901081916537,
+      "grad_norm": 8.022866622030836,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 6451
+    },
+    {
+      "epoch": 0.0061823802163833074,
+      "grad_norm": 6.3020029415834316,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 6452
+    },
+    {
+      "epoch": 0.00927357032457496,
+      "grad_norm": 6.604746499785408,
+      "learning_rate": 1e-06,
+      "loss": 0.1726,
+      "step": 6453
+    },
+    {
+      "epoch": 0.012364760432766615,
+      "grad_norm": 6.535544858556494,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 6454
+    },
+    {
+      "epoch": 0.015455950540958269,
+      "grad_norm": 6.586894391117457,
+      "learning_rate": 1e-06,
+      "loss": 0.1653,
+      "step": 6455
+    },
+    {
+      "epoch": 0.01854714064914992,
+      "grad_norm": 7.663447318647278,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 6456
+    },
+    {
+      "epoch": 0.021638330757341576,
+      "grad_norm": 8.944050187253215,
+      "learning_rate": 1e-06,
+      "loss": 0.181,
+      "step": 6457
+    },
+    {
+      "epoch": 0.02472952086553323,
+      "grad_norm": 7.013041130628615,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 6458
+    },
+    {
+      "epoch": 0.027820710973724884,
+      "grad_norm": 8.928461145027505,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 6459
+    },
+    {
+      "epoch": 0.030911901081916538,
+      "grad_norm": 7.851985625878916,
+      "learning_rate": 1e-06,
+      "loss": 0.162,
+      "step": 6460
+    },
+    {
+      "epoch": 0.03400309119010819,
+      "grad_norm": 8.485484017941207,
+      "learning_rate": 1e-06,
+      "loss": 0.1798,
+      "step": 6461
+    },
+    {
+      "epoch": 0.03709428129829984,
+      "grad_norm": 5.796841876662399,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 6462
+    },
+    {
+      "epoch": 0.0401854714064915,
+      "grad_norm": 7.25941855904635,
+      "learning_rate": 1e-06,
+      "loss": 0.164,
+      "step": 6463
+    },
+    {
+      "epoch": 0.04327666151468315,
+      "grad_norm": 10.072954904386267,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 6464
+    },
+    {
+      "epoch": 0.04636785162287481,
+      "grad_norm": 10.31207163884674,
+      "learning_rate": 1e-06,
+      "loss": 0.1775,
+      "step": 6465
+    },
+    {
+      "epoch": 0.04945904173106646,
+      "grad_norm": 7.774294130096605,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 6466
+    },
+    {
+      "epoch": 0.05255023183925812,
+      "grad_norm": 8.974170570430692,
+      "learning_rate": 1e-06,
+      "loss": 0.1846,
+      "step": 6467
+    },
+    {
+      "epoch": 0.05564142194744977,
+      "grad_norm": 8.61220066518907,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 6468
+    },
+    {
+      "epoch": 0.05873261205564142,
+      "grad_norm": 7.1251007527701375,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 6469
+    },
+    {
+      "epoch": 0.061823802163833076,
+      "grad_norm": 10.047681934672228,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 6470
+    },
+    {
+      "epoch": 0.06491499227202473,
+      "grad_norm": 8.274602300521885,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 6471
+    },
+    {
+      "epoch": 0.06800618238021638,
+      "grad_norm": 8.202725936423015,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 6472
+    },
+    {
+      "epoch": 0.07109737248840804,
+      "grad_norm": 7.301457897945461,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 6473
+    },
+    {
+      "epoch": 0.07418856259659969,
+      "grad_norm": 7.347725096781325,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 6474
+    },
+    {
+      "epoch": 0.07727975270479134,
+      "grad_norm": 8.531483844443027,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 6475
+    },
+    {
+      "epoch": 0.080370942812983,
+      "grad_norm": 7.510667888528431,
+      "learning_rate": 1e-06,
+      "loss": 0.1685,
+      "step": 6476
+    },
+    {
+      "epoch": 0.08346213292117466,
+      "grad_norm": 7.192454240587355,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6477
+    },
+    {
+      "epoch": 0.0865533230293663,
+      "grad_norm": 8.409442810112479,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6478
+    },
+    {
+      "epoch": 0.08964451313755796,
+      "grad_norm": 9.006598431799675,
+      "learning_rate": 1e-06,
+      "loss": 0.1742,
+      "step": 6479
+    },
+    {
+      "epoch": 0.09273570324574962,
+      "grad_norm": 7.731627926838696,
+      "learning_rate": 1e-06,
+      "loss": 0.1691,
+      "step": 6480
+    },
+    {
+      "epoch": 0.09582689335394126,
+      "grad_norm": 8.114520788673019,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 6481
+    },
+    {
+      "epoch": 0.09891808346213292,
+      "grad_norm": 10.46312564015366,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 6482
+    },
+    {
+      "epoch": 0.10200927357032458,
+      "grad_norm": 6.072153393765615,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 6483
+    },
+    {
+      "epoch": 0.10510046367851623,
+      "grad_norm": 6.85767750893739,
+      "learning_rate": 1e-06,
+      "loss": 0.1855,
+      "step": 6484
+    },
+    {
+      "epoch": 0.10819165378670788,
+      "grad_norm": 10.672424389261106,
+      "learning_rate": 1e-06,
+      "loss": 0.174,
+      "step": 6485
+    },
+    {
+      "epoch": 0.11128284389489954,
+      "grad_norm": 8.30253782535488,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 6486
+    },
+    {
+      "epoch": 0.1143740340030912,
+      "grad_norm": 8.540059724757388,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 6487
+    },
+    {
+      "epoch": 0.11746522411128284,
+      "grad_norm": 7.2543407352368545,
+      "learning_rate": 1e-06,
+      "loss": 0.1748,
+      "step": 6488
+    },
+    {
+      "epoch": 0.1205564142194745,
+      "grad_norm": 6.842302590813948,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 6489
+    },
+    {
+      "epoch": 0.12364760432766615,
+      "grad_norm": 7.184103174651009,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 6490
+    },
+    {
+      "epoch": 0.1267387944358578,
+      "grad_norm": 8.760066041451752,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 6491
+    },
+    {
+      "epoch": 0.12982998454404945,
+      "grad_norm": 6.850039184629926,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 6492
+    },
+    {
+      "epoch": 0.13292117465224113,
+      "grad_norm": 8.96028800653891,
+      "learning_rate": 1e-06,
+      "loss": 0.1875,
+      "step": 6493
+    },
+    {
+      "epoch": 0.13601236476043277,
+      "grad_norm": 6.837026878867207,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 6494
+    },
+    {
+      "epoch": 0.1391035548686244,
+      "grad_norm": 7.232955198196647,
+      "learning_rate": 1e-06,
+      "loss": 0.1765,
+      "step": 6495
+    },
+    {
+      "epoch": 0.14219474497681608,
+      "grad_norm": 8.933150415220798,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 6496
+    },
+    {
+      "epoch": 0.14528593508500773,
+      "grad_norm": 7.906634788519629,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6497
+    },
+    {
+      "epoch": 0.14837712519319937,
+      "grad_norm": 8.536164284096746,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 6498
+    },
+    {
+      "epoch": 0.15146831530139104,
+      "grad_norm": 8.333989914109342,
+      "learning_rate": 1e-06,
+      "loss": 0.1879,
+      "step": 6499
+    },
+    {
+      "epoch": 0.1545595054095827,
+      "grad_norm": 8.520726879099156,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 6500
+    },
+    {
+      "epoch": 0.15765069551777433,
+      "grad_norm": 7.903880279486001,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6501
+    },
+    {
+      "epoch": 0.160741885625966,
+      "grad_norm": 5.750023221671661,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 6502
+    },
+    {
+      "epoch": 0.16383307573415765,
+      "grad_norm": 5.719319697429699,
+      "learning_rate": 1e-06,
+      "loss": 0.1694,
+      "step": 6503
+    },
+    {
+      "epoch": 0.16692426584234932,
+      "grad_norm": 8.414544622479276,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6504
+    },
+    {
+      "epoch": 0.17001545595054096,
+      "grad_norm": 8.834202859323927,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 6505
+    },
+    {
+      "epoch": 0.1731066460587326,
+      "grad_norm": 8.818512837868987,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6506
+    },
+    {
+      "epoch": 0.17619783616692428,
+      "grad_norm": 8.29927463072674,
+      "learning_rate": 1e-06,
+      "loss": 0.1832,
+      "step": 6507
+    },
+    {
+      "epoch": 0.17928902627511592,
+      "grad_norm": 9.162322790229597,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 6508
+    },
+    {
+      "epoch": 0.18238021638330756,
+      "grad_norm": 8.620612468638317,
+      "learning_rate": 1e-06,
+      "loss": 0.1746,
+      "step": 6509
+    },
+    {
+      "epoch": 0.18547140649149924,
+      "grad_norm": 7.901552940735184,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 6510
+    },
+    {
+      "epoch": 0.18856259659969088,
+      "grad_norm": 7.005297025445319,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 6511
+    },
+    {
+      "epoch": 0.19165378670788252,
+      "grad_norm": 6.18430965333544,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 6512
+    },
+    {
+      "epoch": 0.1947449768160742,
+      "grad_norm": 7.13185986545599,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 6513
+    },
+    {
+      "epoch": 0.19783616692426584,
+      "grad_norm": 7.376302816857971,
+      "learning_rate": 1e-06,
+      "loss": 0.1944,
+      "step": 6514
+    },
+    {
+      "epoch": 0.2009273570324575,
+      "grad_norm": 8.149671609664727,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 6515
+    },
+    {
+      "epoch": 0.20401854714064915,
+      "grad_norm": 8.89761941397276,
+      "learning_rate": 1e-06,
+      "loss": 0.1948,
+      "step": 6516
+    },
+    {
+      "epoch": 0.2071097372488408,
+      "grad_norm": 9.129521243335693,
+      "learning_rate": 1e-06,
+      "loss": 0.1765,
+      "step": 6517
+    },
+    {
+      "epoch": 0.21020092735703247,
+      "grad_norm": 9.58187306373798,
+      "learning_rate": 1e-06,
+      "loss": 0.1819,
+      "step": 6518
+    },
+    {
+      "epoch": 0.2132921174652241,
+      "grad_norm": 7.858102470770091,
+      "learning_rate": 1e-06,
+      "loss": 0.1737,
+      "step": 6519
+    },
+    {
+      "epoch": 0.21638330757341576,
+      "grad_norm": 7.2999516505499225,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 6520
+    },
+    {
+      "epoch": 0.21947449768160743,
+      "grad_norm": 8.029560546670886,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 6521
+    },
+    {
+      "epoch": 0.22256568778979907,
+      "grad_norm": 7.319958142128086,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 6522
+    },
+    {
+      "epoch": 0.22565687789799072,
+      "grad_norm": 8.82956279723494,
+      "learning_rate": 1e-06,
+      "loss": 0.1696,
+      "step": 6523
+    },
+    {
+      "epoch": 0.2287480680061824,
+      "grad_norm": 8.215509555650069,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 6524
+    },
+    {
+      "epoch": 0.23183925811437403,
+      "grad_norm": 7.169618843128498,
+      "learning_rate": 1e-06,
+      "loss": 0.188,
+      "step": 6525
+    },
+    {
+      "epoch": 0.23493044822256567,
+      "grad_norm": 10.124223758386782,
+      "learning_rate": 1e-06,
+      "loss": 0.1911,
+      "step": 6526
+    },
+    {
+      "epoch": 0.23802163833075735,
+      "grad_norm": 7.277905539757565,
+      "learning_rate": 1e-06,
+      "loss": 0.1743,
+      "step": 6527
+    },
+    {
+      "epoch": 0.241112828438949,
+      "grad_norm": 7.778403579136919,
+      "learning_rate": 1e-06,
+      "loss": 0.1757,
+      "step": 6528
+    },
+    {
+      "epoch": 0.24420401854714066,
+      "grad_norm": 8.685119704209908,
+      "learning_rate": 1e-06,
+      "loss": 0.1797,
+      "step": 6529
+    },
+    {
+      "epoch": 0.2472952086553323,
+      "grad_norm": 8.092205193697431,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 6530
+    },
+    {
+      "epoch": 0.250386398763524,
+      "grad_norm": 9.80722661723471,
+      "learning_rate": 1e-06,
+      "loss": 0.1742,
+      "step": 6531
+    },
+    {
+      "epoch": 0.2534775888717156,
+      "grad_norm": 7.80738985079281,
+      "learning_rate": 1e-06,
+      "loss": 0.1683,
+      "step": 6532
+    },
+    {
+      "epoch": 0.25656877897990726,
+      "grad_norm": 8.094316059727683,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 6533
+    },
+    {
+      "epoch": 0.2596599690880989,
+      "grad_norm": 7.7045836425828575,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 6534
+    },
+    {
+      "epoch": 0.26275115919629055,
+      "grad_norm": 5.836258356686844,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 6535
+    },
+    {
+      "epoch": 0.26584234930448225,
+      "grad_norm": 9.44267447853166,
+      "learning_rate": 1e-06,
+      "loss": 0.1818,
+      "step": 6536
+    },
+    {
+      "epoch": 0.2689335394126739,
+      "grad_norm": 8.127510030862963,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 6537
+    },
+    {
+      "epoch": 0.27202472952086554,
+      "grad_norm": 6.5984041910034525,
+      "learning_rate": 1e-06,
+      "loss": 0.1642,
+      "step": 6538
+    },
+    {
+      "epoch": 0.2751159196290572,
+      "grad_norm": 7.730390705666554,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6539
+    },
+    {
+      "epoch": 0.2782071097372488,
+      "grad_norm": 6.772355644152555,
+      "learning_rate": 1e-06,
+      "loss": 0.1755,
+      "step": 6540
+    },
+    {
+      "epoch": 0.28129829984544047,
+      "grad_norm": 7.856506704036645,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 6541
+    },
+    {
+      "epoch": 0.28438948995363217,
+      "grad_norm": 7.348576738422421,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 6542
+    },
+    {
+      "epoch": 0.2874806800618238,
+      "grad_norm": 5.872250159916791,
+      "learning_rate": 1e-06,
+      "loss": 0.1698,
+      "step": 6543
+    },
+    {
+      "epoch": 0.29057187017001546,
+      "grad_norm": 7.429090644627223,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 6544
+    },
+    {
+      "epoch": 0.2936630602782071,
+      "grad_norm": 6.468648867331983,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 6545
+    },
+    {
+      "epoch": 0.29675425038639874,
+      "grad_norm": 8.765863958178478,
+      "learning_rate": 1e-06,
+      "loss": 0.1939,
+      "step": 6546
+    },
+    {
+      "epoch": 0.29984544049459044,
+      "grad_norm": 7.969440184461612,
+      "learning_rate": 1e-06,
+      "loss": 0.1801,
+      "step": 6547
+    },
+    {
+      "epoch": 0.3029366306027821,
+      "grad_norm": 8.148064942537404,
+      "learning_rate": 1e-06,
+      "loss": 0.1945,
+      "step": 6548
+    },
+    {
+      "epoch": 0.30602782071097373,
+      "grad_norm": 6.919299036337801,
+      "learning_rate": 1e-06,
+      "loss": 0.1689,
+      "step": 6549
+    },
+    {
+      "epoch": 0.3091190108191654,
+      "grad_norm": 8.504611758929668,
+      "learning_rate": 1e-06,
+      "loss": 0.1718,
+      "step": 6550
+    },
+    {
+      "epoch": 0.312210200927357,
+      "grad_norm": 7.696470403331865,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 6551
+    },
+    {
+      "epoch": 0.31530139103554866,
+      "grad_norm": 9.7532220331523,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 6552
+    },
+    {
+      "epoch": 0.31839258114374036,
+      "grad_norm": 7.331712254022098,
+      "learning_rate": 1e-06,
+      "loss": 0.1769,
+      "step": 6553
+    },
+    {
+      "epoch": 0.321483771251932,
+      "grad_norm": 6.815329028407151,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 6554
+    },
+    {
+      "epoch": 0.32457496136012365,
+      "grad_norm": 8.949879502256865,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 6555
+    },
+    {
+      "epoch": 0.3276661514683153,
+      "grad_norm": 7.499778311465562,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 6556
+    },
+    {
+      "epoch": 0.33075734157650694,
+      "grad_norm": 9.527796536109326,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 6557
+    },
+    {
+      "epoch": 0.33384853168469864,
+      "grad_norm": 7.471056691222567,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 6558
+    },
+    {
+      "epoch": 0.3369397217928903,
+      "grad_norm": 7.506825297594253,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 6559
+    },
+    {
+      "epoch": 0.3400309119010819,
+      "grad_norm": 7.916612741546534,
+      "learning_rate": 1e-06,
+      "loss": 0.1751,
+      "step": 6560
+    },
+    {
+      "epoch": 0.34312210200927357,
+      "grad_norm": 7.581264990024492,
+      "learning_rate": 1e-06,
+      "loss": 0.1706,
+      "step": 6561
+    },
+    {
+      "epoch": 0.3462132921174652,
+      "grad_norm": 8.830091421532444,
+      "learning_rate": 1e-06,
+      "loss": 0.1981,
+      "step": 6562
+    },
+    {
+      "epoch": 0.34930448222565685,
+      "grad_norm": 6.024437055839159,
+      "learning_rate": 1e-06,
+      "loss": 0.1708,
+      "step": 6563
+    },
+    {
+      "epoch": 0.35239567233384855,
+      "grad_norm": 7.806039823939993,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 6564
+    },
+    {
+      "epoch": 0.3554868624420402,
+      "grad_norm": 9.680611734158397,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 6565
+    },
+    {
+      "epoch": 0.35857805255023184,
+      "grad_norm": 8.124391990402213,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 6566
+    },
+    {
+      "epoch": 0.3616692426584235,
+      "grad_norm": 8.1272709862505,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 6567
+    },
+    {
+      "epoch": 0.36476043276661513,
+      "grad_norm": 6.805796976669091,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 6568
+    },
+    {
+      "epoch": 0.3678516228748068,
+      "grad_norm": 8.623800675913968,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 6569
+    },
+    {
+      "epoch": 0.37094281298299847,
+      "grad_norm": 8.13666103495829,
+      "learning_rate": 1e-06,
+      "loss": 0.1666,
+      "step": 6570
+    },
+    {
+      "epoch": 0.3740340030911901,
+      "grad_norm": 7.076468257783597,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 6571
+    },
+    {
+      "epoch": 0.37712519319938176,
+      "grad_norm": 8.643628186576692,
+      "learning_rate": 1e-06,
+      "loss": 0.1901,
+      "step": 6572
+    },
+    {
+      "epoch": 0.3802163833075734,
+      "grad_norm": 6.331453401774583,
+      "learning_rate": 1e-06,
+      "loss": 0.1772,
+      "step": 6573
+    },
+    {
+      "epoch": 0.38330757341576505,
+      "grad_norm": 6.171610369669614,
+      "learning_rate": 1e-06,
+      "loss": 0.1656,
+      "step": 6574
+    },
+    {
+      "epoch": 0.38639876352395675,
+      "grad_norm": 6.171610369669614,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 6575
+    },
+    {
+      "epoch": 0.3894899536321484,
+      "grad_norm": 9.229097146515905,
+      "learning_rate": 1e-06,
+      "loss": 0.1867,
+      "step": 6576
+    },
+    {
+      "epoch": 0.39258114374034003,
+      "grad_norm": 8.732091038518023,
+      "learning_rate": 1e-06,
+      "loss": 0.1909,
+      "step": 6577
+    },
+    {
+      "epoch": 0.3956723338485317,
+      "grad_norm": 10.34186272216336,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6578
+    },
+    {
+      "epoch": 0.3987635239567233,
+      "grad_norm": 6.235973957621764,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 6579
+    },
+    {
+      "epoch": 0.401854714064915,
+      "grad_norm": 8.255009825643302,
+      "learning_rate": 1e-06,
+      "loss": 0.1773,
+      "step": 6580
+    },
+    {
+      "epoch": 0.40494590417310666,
+      "grad_norm": 6.360920817922092,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 6581
+    },
+    {
+      "epoch": 0.4080370942812983,
+      "grad_norm": 6.702239831875602,
+      "learning_rate": 1e-06,
+      "loss": 0.1692,
+      "step": 6582
+    },
+    {
+      "epoch": 0.41112828438948995,
+      "grad_norm": 7.38231334951037,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 6583
+    },
+    {
+      "epoch": 0.4142194744976816,
+      "grad_norm": 5.64656069848952,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 6584
+    },
+    {
+      "epoch": 0.41731066460587324,
+      "grad_norm": 5.807032723874409,
+      "learning_rate": 1e-06,
+      "loss": 0.1832,
+      "step": 6585
+    },
+    {
+      "epoch": 0.42040185471406494,
+      "grad_norm": 6.003711394697606,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 6586
+    },
+    {
+      "epoch": 0.4234930448222566,
+      "grad_norm": 8.975959170640252,
+      "learning_rate": 1e-06,
+      "loss": 0.1946,
+      "step": 6587
+    },
+    {
+      "epoch": 0.4265842349304482,
+      "grad_norm": 9.366238975405874,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 6588
+    },
+    {
+      "epoch": 0.42967542503863987,
+      "grad_norm": 10.468725663589856,
+      "learning_rate": 1e-06,
+      "loss": 0.2013,
+      "step": 6589
+    },
+    {
+      "epoch": 0.4327666151468315,
+      "grad_norm": 9.403415496049089,
+      "learning_rate": 1e-06,
+      "loss": 0.1791,
+      "step": 6590
+    },
+    {
+      "epoch": 0.43585780525502316,
+      "grad_norm": 6.895220982838442,
+      "learning_rate": 1e-06,
+      "loss": 0.1631,
+      "step": 6591
+    },
+    {
+      "epoch": 0.43894899536321486,
+      "grad_norm": 9.62864386847146,
+      "learning_rate": 1e-06,
+      "loss": 0.1703,
+      "step": 6592
+    },
+    {
+      "epoch": 0.4420401854714065,
+      "grad_norm": 5.589949579627442,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 6593
+    },
+    {
+      "epoch": 0.44513137557959814,
+      "grad_norm": 8.228539638888307,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 6594
+    },
+    {
+      "epoch": 0.4482225656877898,
+      "grad_norm": 9.266464031000952,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 6595
+    },
+    {
+      "epoch": 0.45131375579598143,
+      "grad_norm": 7.769629465607218,
+      "learning_rate": 1e-06,
+      "loss": 0.1741,
+      "step": 6596
+    },
+    {
+      "epoch": 0.45440494590417313,
+      "grad_norm": 7.859282125297894,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 6597
+    },
+    {
+      "epoch": 0.4574961360123648,
+      "grad_norm": 8.788029895390345,
+      "learning_rate": 1e-06,
+      "loss": 0.1967,
+      "step": 6598
+    },
+    {
+      "epoch": 0.4605873261205564,
+      "grad_norm": 7.766135773345866,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6599
+    },
+    {
+      "epoch": 0.46367851622874806,
+      "grad_norm": 8.757020732775118,
+      "learning_rate": 1e-06,
+      "loss": 0.1852,
+      "step": 6600
+    },
+    {
+      "epoch": 0.4667697063369397,
+      "grad_norm": 7.91498910595768,
+      "learning_rate": 1e-06,
+      "loss": 0.1849,
+      "step": 6601
+    },
+    {
+      "epoch": 0.46986089644513135,
+      "grad_norm": 9.567125551403269,
+      "learning_rate": 1e-06,
+      "loss": 0.1955,
+      "step": 6602
+    },
+    {
+      "epoch": 0.47295208655332305,
+      "grad_norm": 8.198307331063813,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 6603
+    },
+    {
+      "epoch": 0.4760432766615147,
+      "grad_norm": 8.077875467353321,
+      "learning_rate": 1e-06,
+      "loss": 0.1878,
+      "step": 6604
+    },
+    {
+      "epoch": 0.47913446676970634,
+      "grad_norm": 7.404804243814033,
+      "learning_rate": 1e-06,
+      "loss": 0.1727,
+      "step": 6605
+    },
+    {
+      "epoch": 0.482225656877898,
+      "grad_norm": 7.8339835581741015,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 6606
+    },
+    {
+      "epoch": 0.4853168469860896,
+      "grad_norm": 7.242664947815419,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 6607
+    },
+    {
+      "epoch": 0.4884080370942813,
+      "grad_norm": 11.366342979841297,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 6608
+    },
+    {
+      "epoch": 0.49149922720247297,
+      "grad_norm": 9.90709383496323,
+      "learning_rate": 1e-06,
+      "loss": 0.1891,
+      "step": 6609
+    },
+    {
+      "epoch": 0.4945904173106646,
+      "grad_norm": 5.39908277817307,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6610
+    },
+    {
+      "epoch": 0.49768160741885625,
+      "grad_norm": 6.935933235846953,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6611
+    },
+    {
+      "epoch": 0.500772797527048,
+      "grad_norm": 7.782305755046552,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 6612
+    },
+    {
+      "epoch": 0.5038639876352395,
+      "grad_norm": 8.196001488515215,
+      "learning_rate": 1e-06,
+      "loss": 0.1853,
+      "step": 6613
+    },
+    {
+      "epoch": 0.5069551777434312,
+      "grad_norm": 7.673225388931027,
+      "learning_rate": 1e-06,
+      "loss": 0.1742,
+      "step": 6614
+    },
+    {
+      "epoch": 0.5100463678516228,
+      "grad_norm": 9.596897898648711,
+      "learning_rate": 1e-06,
+      "loss": 0.1753,
+      "step": 6615
+    },
+    {
+      "epoch": 0.5131375579598145,
+      "grad_norm": 8.12903197348307,
+      "learning_rate": 1e-06,
+      "loss": 0.1971,
+      "step": 6616
+    },
+    {
+      "epoch": 0.5162287480680062,
+      "grad_norm": 7.94020300384907,
+      "learning_rate": 1e-06,
+      "loss": 0.1651,
+      "step": 6617
+    },
+    {
+      "epoch": 0.5193199381761978,
+      "grad_norm": 9.211419543660517,
+      "learning_rate": 1e-06,
+      "loss": 0.1721,
+      "step": 6618
+    },
+    {
+      "epoch": 0.5224111282843895,
+      "grad_norm": 8.826364212859078,
+      "learning_rate": 1e-06,
+      "loss": 0.1868,
+      "step": 6619
+    },
+    {
+      "epoch": 0.5255023183925811,
+      "grad_norm": 9.172056780142174,
+      "learning_rate": 1e-06,
+      "loss": 0.1823,
+      "step": 6620
+    },
+    {
+      "epoch": 0.5285935085007728,
+      "grad_norm": 9.945414017920884,
+      "learning_rate": 1e-06,
+      "loss": 0.1886,
+      "step": 6621
+    },
+    {
+      "epoch": 0.5316846986089645,
+      "grad_norm": 7.5597706455456395,
+      "learning_rate": 1e-06,
+      "loss": 0.175,
+      "step": 6622
+    },
+    {
+      "epoch": 0.5347758887171561,
+      "grad_norm": 7.746456706400544,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 6623
+    },
+    {
+      "epoch": 0.5378670788253478,
+      "grad_norm": 10.03602542675447,
+      "learning_rate": 1e-06,
+      "loss": 0.1713,
+      "step": 6624
+    },
+    {
+      "epoch": 0.5409582689335394,
+      "grad_norm": 7.174289117745116,
+      "learning_rate": 1e-06,
+      "loss": 0.1831,
+      "step": 6625
+    },
+    {
+      "epoch": 0.5440494590417311,
+      "grad_norm": 7.564299871533804,
+      "learning_rate": 1e-06,
+      "loss": 0.1728,
+      "step": 6626
+    },
+    {
+      "epoch": 0.5471406491499228,
+      "grad_norm": 9.690969321515652,
+      "learning_rate": 1e-06,
+      "loss": 0.1842,
+      "step": 6627
+    },
+    {
+      "epoch": 0.5502318392581144,
+      "grad_norm": 8.305787939034259,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 6628
+    },
+    {
+      "epoch": 0.5533230293663061,
+      "grad_norm": 10.131186453626896,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 6629
+    },
+    {
+      "epoch": 0.5564142194744977,
+      "grad_norm": 9.6269449805642,
+      "learning_rate": 1e-06,
+      "loss": 0.1839,
+      "step": 6630
+    },
+    {
+      "epoch": 0.5595054095826894,
+      "grad_norm": 10.086724334970752,
+      "learning_rate": 1e-06,
+      "loss": 0.185,
+      "step": 6631
+    },
+    {
+      "epoch": 0.5625965996908809,
+      "grad_norm": 6.500578341914664,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 6632
+    },
+    {
+      "epoch": 0.5656877897990726,
+      "grad_norm": 8.546755241525037,
+      "learning_rate": 1e-06,
+      "loss": 0.1805,
+      "step": 6633
+    },
+    {
+      "epoch": 0.5687789799072643,
+      "grad_norm": 8.725844128433211,
+      "learning_rate": 1e-06,
+      "loss": 0.1896,
+      "step": 6634
+    },
+    {
+      "epoch": 0.5718701700154559,
+      "grad_norm": 8.239615292775442,
+      "learning_rate": 1e-06,
+      "loss": 0.1978,
+      "step": 6635
+    },
+    {
+      "epoch": 0.5749613601236476,
+      "grad_norm": 7.785561775399897,
+      "learning_rate": 1e-06,
+      "loss": 0.1736,
+      "step": 6636
+    },
+    {
+      "epoch": 0.5780525502318392,
+      "grad_norm": 7.680374016972156,
+      "learning_rate": 1e-06,
+      "loss": 0.1817,
+      "step": 6637
+    },
+    {
+      "epoch": 0.5811437403400309,
+      "grad_norm": 6.36878027389339,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 6638
+    },
+    {
+      "epoch": 0.5842349304482226,
+      "grad_norm": 8.9420901881958,
+      "learning_rate": 1e-06,
+      "loss": 0.1767,
+      "step": 6639
+    },
+    {
+      "epoch": 0.5873261205564142,
+      "grad_norm": 8.730567640892847,
+      "learning_rate": 1e-06,
+      "loss": 0.186,
+      "step": 6640
+    },
+    {
+      "epoch": 0.5904173106646059,
+      "grad_norm": 11.053955470130486,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 6641
+    },
+    {
+      "epoch": 0.5935085007727975,
+      "grad_norm": 7.63399710162199,
+      "learning_rate": 1e-06,
+      "loss": 0.1813,
+      "step": 6642
+    },
+    {
+      "epoch": 0.5965996908809892,
+      "grad_norm": 9.03895158062643,
+      "learning_rate": 1e-06,
+      "loss": 0.1779,
+      "step": 6643
+    },
+    {
+      "epoch": 0.5996908809891809,
+      "grad_norm": 8.437761293500166,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6644
+    },
+    {
+      "epoch": 0.6027820710973725,
+      "grad_norm": 7.966919960833477,
+      "learning_rate": 1e-06,
+      "loss": 0.1899,
+      "step": 6645
+    },
+    {
+      "epoch": 0.6058732612055642,
+      "grad_norm": 6.6303063189736555,
+      "learning_rate": 1e-06,
+      "loss": 0.1916,
+      "step": 6646
+    },
+    {
+      "epoch": 0.6089644513137558,
+      "grad_norm": 7.281264754293063,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 6647
+    },
+    {
+      "epoch": 0.6120556414219475,
+      "grad_norm": 7.547950211436911,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 6648
+    },
+    {
+      "epoch": 0.615146831530139,
+      "grad_norm": 8.701030625361907,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 6649
+    },
+    {
+      "epoch": 0.6182380216383307,
+      "grad_norm": 8.114704083294864,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 6650
+    },
+    {
+      "epoch": 0.6213292117465224,
+      "grad_norm": 6.9701133757051785,
+      "learning_rate": 1e-06,
+      "loss": 0.1745,
+      "step": 6651
+    },
+    {
+      "epoch": 0.624420401854714,
+      "grad_norm": 7.709478171993677,
+      "learning_rate": 1e-06,
+      "loss": 0.1904,
+      "step": 6652
+    },
+    {
+      "epoch": 0.6275115919629057,
+      "grad_norm": 8.564809460333397,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 6653
+    },
+    {
+      "epoch": 0.6306027820710973,
+      "grad_norm": 5.692463737137694,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 6654
+    },
+    {
+      "epoch": 0.633693972179289,
+      "grad_norm": 8.421010628986362,
+      "learning_rate": 1e-06,
+      "loss": 0.1914,
+      "step": 6655
+    },
+    {
+      "epoch": 0.6367851622874807,
+      "grad_norm": 7.986468660240707,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 6656
+    },
+    {
+      "epoch": 0.6398763523956723,
+      "grad_norm": 7.846809953078294,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 6657
+    },
+    {
+      "epoch": 0.642967542503864,
+      "grad_norm": 6.15450335776385,
+      "learning_rate": 1e-06,
+      "loss": 0.1803,
+      "step": 6658
+    },
+    {
+      "epoch": 0.6460587326120556,
+      "grad_norm": 7.11696284672819,
+      "learning_rate": 1e-06,
+      "loss": 0.1932,
+      "step": 6659
+    },
+    {
+      "epoch": 0.6491499227202473,
+      "grad_norm": 8.558530467860184,
+      "learning_rate": 1e-06,
+      "loss": 0.1825,
+      "step": 6660
+    },
+    {
+      "epoch": 0.652241112828439,
+      "grad_norm": 8.710589892648198,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6661
+    },
+    {
+      "epoch": 0.6553323029366306,
+      "grad_norm": 7.969013457115489,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 6662
+    },
+    {
+      "epoch": 0.6584234930448223,
+      "grad_norm": 8.59749558456472,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 6663
+    },
+    {
+      "epoch": 0.6615146831530139,
+      "grad_norm": 8.79616902316514,
+      "learning_rate": 1e-06,
+      "loss": 0.1845,
+      "step": 6664
+    },
+    {
+      "epoch": 0.6646058732612056,
+      "grad_norm": 8.22105543046577,
+      "learning_rate": 1e-06,
+      "loss": 0.1765,
+      "step": 6665
+    },
+    {
+      "epoch": 0.6676970633693973,
+      "grad_norm": 8.806996492966189,
+      "learning_rate": 1e-06,
+      "loss": 0.1793,
+      "step": 6666
+    },
+    {
+      "epoch": 0.6707882534775889,
+      "grad_norm": 8.061899937710828,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 6667
+    },
+    {
+      "epoch": 0.6738794435857806,
+      "grad_norm": 7.390728065525513,
+      "learning_rate": 1e-06,
+      "loss": 0.1844,
+      "step": 6668
+    },
+    {
+      "epoch": 0.6769706336939721,
+      "grad_norm": 6.401733006778724,
+      "learning_rate": 1e-06,
+      "loss": 0.1809,
+      "step": 6669
+    },
+    {
+      "epoch": 0.6800618238021638,
+      "grad_norm": 10.029473474544016,
+      "learning_rate": 1e-06,
+      "loss": 0.1811,
+      "step": 6670
+    },
+    {
+      "epoch": 0.6831530139103554,
+      "grad_norm": 7.387666299423233,
+      "learning_rate": 1e-06,
+      "loss": 0.1777,
+      "step": 6671
+    },
+    {
+      "epoch": 0.6862442040185471,
+      "grad_norm": 8.62828026166174,
+      "learning_rate": 1e-06,
+      "loss": 0.1887,
+      "step": 6672
+    },
+    {
+      "epoch": 0.6893353941267388,
+      "grad_norm": 7.172351130624503,
+      "learning_rate": 1e-06,
+      "loss": 0.1758,
+      "step": 6673
+    },
+    {
+      "epoch": 0.6924265842349304,
+      "grad_norm": 7.2572081564741255,
+      "learning_rate": 1e-06,
+      "loss": 0.1752,
+      "step": 6674
+    },
+    {
+      "epoch": 0.6955177743431221,
+      "grad_norm": 10.280685984704654,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 6675
+    },
+    {
+      "epoch": 0.6986089644513137,
+      "grad_norm": 7.71325443841677,
+      "learning_rate": 1e-06,
+      "loss": 0.1807,
+      "step": 6676
+    },
+    {
+      "epoch": 0.7017001545595054,
+      "grad_norm": 8.036929800957996,
+      "learning_rate": 1e-06,
+      "loss": 0.1854,
+      "step": 6677
+    },
+    {
+      "epoch": 0.7047913446676971,
+      "grad_norm": 10.298904595904121,
+      "learning_rate": 1e-06,
+      "loss": 0.1898,
+      "step": 6678
+    },
+    {
+      "epoch": 0.7078825347758887,
+      "grad_norm": 7.932703717078685,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 6679
+    },
+    {
+      "epoch": 0.7109737248840804,
+      "grad_norm": 9.405267234158938,
+      "learning_rate": 1e-06,
+      "loss": 0.1815,
+      "step": 6680
+    },
+    {
+      "epoch": 0.714064914992272,
+      "grad_norm": 7.26163513188098,
+      "learning_rate": 1e-06,
+      "loss": 0.1951,
+      "step": 6681
+    },
+    {
+      "epoch": 0.7171561051004637,
+      "grad_norm": 7.814227003627742,
+      "learning_rate": 1e-06,
+      "loss": 0.167,
+      "step": 6682
+    },
+    {
+      "epoch": 0.7202472952086554,
+      "grad_norm": 7.629935666874913,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 6683
+    },
+    {
+      "epoch": 0.723338485316847,
+      "grad_norm": 8.985237860411099,
+      "learning_rate": 1e-06,
+      "loss": 0.1802,
+      "step": 6684
+    },
+    {
+      "epoch": 0.7264296754250387,
+      "grad_norm": 8.703216576998669,
+      "learning_rate": 1e-06,
+      "loss": 0.1762,
+      "step": 6685
+    },
+    {
+      "epoch": 0.7295208655332303,
+      "grad_norm": 8.577217354294556,
+      "learning_rate": 1e-06,
+      "loss": 0.1794,
+      "step": 6686
+    },
+    {
+      "epoch": 0.732612055641422,
+      "grad_norm": 7.944471255450067,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 6687
+    },
+    {
+      "epoch": 0.7357032457496137,
+      "grad_norm": 7.963300831640238,
+      "learning_rate": 1e-06,
+      "loss": 0.184,
+      "step": 6688
+    },
+    {
+      "epoch": 0.7387944358578052,
+      "grad_norm": 10.127074808812354,
+      "learning_rate": 1e-06,
+      "loss": 0.202,
+      "step": 6689
+    },
+    {
+      "epoch": 0.7418856259659969,
+      "grad_norm": 6.964680253777274,
+      "learning_rate": 1e-06,
+      "loss": 0.1861,
+      "step": 6690
+    },
+    {
+      "epoch": 0.7449768160741885,
+      "grad_norm": 7.458986169959559,
+      "learning_rate": 1e-06,
+      "loss": 0.1731,
+      "step": 6691
+    },
+    {
+      "epoch": 0.7480680061823802,
+      "grad_norm": 7.786918029495548,
+      "learning_rate": 1e-06,
+      "loss": 0.1829,
+      "step": 6692
+    },
+    {
+      "epoch": 0.7511591962905718,
+      "grad_norm": 7.028134591414406,
+      "learning_rate": 1e-06,
+      "loss": 0.1768,
+      "step": 6693
+    },
+    {
+      "epoch": 0.7542503863987635,
+      "grad_norm": 8.546181139677133,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6694
+    },
+    {
+      "epoch": 0.7573415765069552,
+      "grad_norm": 8.168094314172073,
+      "learning_rate": 1e-06,
+      "loss": 0.1847,
+      "step": 6695
+    },
+    {
+      "epoch": 0.7604327666151468,
+      "grad_norm": 8.226643093398762,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 6696
+    },
+    {
+      "epoch": 0.7635239567233385,
+      "grad_norm": 8.110730270647792,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 6697
+    },
+    {
+      "epoch": 0.7666151468315301,
+      "grad_norm": 8.376263087112958,
+      "learning_rate": 1e-06,
+      "loss": 0.1976,
+      "step": 6698
+    },
+    {
+      "epoch": 0.7697063369397218,
+      "grad_norm": 7.539864232909179,
+      "learning_rate": 1e-06,
+      "loss": 0.1985,
+      "step": 6699
+    },
+    {
+      "epoch": 0.7727975270479135,
+      "grad_norm": 7.72685196661432,
+      "learning_rate": 1e-06,
+      "loss": 0.1763,
+      "step": 6700
+    },
+    {
+      "epoch": 0.7758887171561051,
+      "grad_norm": 9.016319178397639,
+      "learning_rate": 1e-06,
+      "loss": 0.1918,
+      "step": 6701
+    },
+    {
+      "epoch": 0.7789799072642968,
+      "grad_norm": 8.510345834165733,
+      "learning_rate": 1e-06,
+      "loss": 0.1838,
+      "step": 6702
+    },
+    {
+      "epoch": 0.7820710973724884,
+      "grad_norm": 7.490771528511289,
+      "learning_rate": 1e-06,
+      "loss": 0.1806,
+      "step": 6703
+    },
+    {
+      "epoch": 0.7851622874806801,
+      "grad_norm": 7.246960138713434,
+      "learning_rate": 1e-06,
+      "loss": 0.195,
+      "step": 6704
+    },
+    {
+      "epoch": 0.7882534775888718,
+      "grad_norm": 10.304057104287784,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 6705
+    },
+    {
+      "epoch": 0.7913446676970634,
+      "grad_norm": 7.0678821120233914,
+      "learning_rate": 1e-06,
+      "loss": 0.182,
+      "step": 6706
+    },
+    {
+      "epoch": 0.794435857805255,
+      "grad_norm": 7.648400519444676,
+      "learning_rate": 1e-06,
+      "loss": 0.1783,
+      "step": 6707
+    },
+    {
+      "epoch": 0.7975270479134466,
+      "grad_norm": 6.50877083804805,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 6708
+    },
+    {
+      "epoch": 0.8006182380216383,
+      "grad_norm": 9.145805688101763,
+      "learning_rate": 1e-06,
+      "loss": 0.1716,
+      "step": 6709
+    },
+    {
+      "epoch": 0.80370942812983,
+      "grad_norm": 6.453019830878496,
+      "learning_rate": 1e-06,
+      "loss": 0.2011,
+      "step": 6710
+    },
+    {
+      "epoch": 0.8068006182380216,
+      "grad_norm": 7.3828888561038,
+      "learning_rate": 1e-06,
+      "loss": 0.1778,
+      "step": 6711
+    },
+    {
+      "epoch": 0.8098918083462133,
+      "grad_norm": 9.125482951421455,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 6712
+    },
+    {
+      "epoch": 0.8129829984544049,
+      "grad_norm": 8.561276651690926,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 6713
+    },
+    {
+      "epoch": 0.8160741885625966,
+      "grad_norm": 9.676461111200037,
+      "learning_rate": 1e-06,
+      "loss": 0.1858,
+      "step": 6714
+    },
+    {
+      "epoch": 0.8191653786707882,
+      "grad_norm": 6.145712949964476,
+      "learning_rate": 1e-06,
+      "loss": 0.1874,
+      "step": 6715
+    },
+    {
+      "epoch": 0.8222565687789799,
+      "grad_norm": 7.671675982760035,
+      "learning_rate": 1e-06,
+      "loss": 0.169,
+      "step": 6716
+    },
+    {
+      "epoch": 0.8253477588871716,
+      "grad_norm": 6.965161214151917,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 6717
+    },
+    {
+      "epoch": 0.8284389489953632,
+      "grad_norm": 7.6592136265284125,
+      "learning_rate": 1e-06,
+      "loss": 0.1826,
+      "step": 6718
+    },
+    {
+      "epoch": 0.8315301391035549,
+      "grad_norm": 8.474452073352163,
+      "learning_rate": 1e-06,
+      "loss": 0.1893,
+      "step": 6719
+    },
+    {
+      "epoch": 0.8346213292117465,
+      "grad_norm": 7.610085269581773,
+      "learning_rate": 1e-06,
+      "loss": 0.1715,
+      "step": 6720
+    },
+    {
+      "epoch": 0.8377125193199382,
+      "grad_norm": 9.049695267722814,
+      "learning_rate": 1e-06,
+      "loss": 0.1808,
+      "step": 6721
+    },
+    {
+      "epoch": 0.8408037094281299,
+      "grad_norm": 8.424653559787128,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 6722
+    },
+    {
+      "epoch": 0.8438948995363215,
+      "grad_norm": 8.96607802040107,
+      "learning_rate": 1e-06,
+      "loss": 0.1749,
+      "step": 6723
+    },
+    {
+      "epoch": 0.8469860896445132,
+      "grad_norm": 7.835875336179628,
+      "learning_rate": 1e-06,
+      "loss": 0.1902,
+      "step": 6724
+    },
+    {
+      "epoch": 0.8500772797527048,
+      "grad_norm": 8.223763223838146,
+      "learning_rate": 1e-06,
+      "loss": 0.1682,
+      "step": 6725
+    },
+    {
+      "epoch": 0.8531684698608965,
+      "grad_norm": 7.591427515475169,
+      "learning_rate": 1e-06,
+      "loss": 0.1814,
+      "step": 6726
+    },
+    {
+      "epoch": 0.8562596599690881,
+      "grad_norm": 9.323413875066272,
+      "learning_rate": 1e-06,
+      "loss": 0.1863,
+      "step": 6727
+    },
+    {
+      "epoch": 0.8593508500772797,
+      "grad_norm": 8.197925803774657,
+      "learning_rate": 1e-06,
+      "loss": 0.2038,
+      "step": 6728
+    },
+    {
+      "epoch": 0.8624420401854714,
+      "grad_norm": 8.955600259193396,
+      "learning_rate": 1e-06,
+      "loss": 0.1851,
+      "step": 6729
+    },
+    {
+      "epoch": 0.865533230293663,
+      "grad_norm": 7.599490010024607,
+      "learning_rate": 1e-06,
+      "loss": 0.1695,
+      "step": 6730
+    },
+    {
+      "epoch": 0.8686244204018547,
+      "grad_norm": 10.03200829009792,
+      "learning_rate": 1e-06,
+      "loss": 0.1892,
+      "step": 6731
+    },
+    {
+      "epoch": 0.8717156105100463,
+      "grad_norm": 7.101812521895584,
+      "learning_rate": 1e-06,
+      "loss": 0.1781,
+      "step": 6732
+    },
+    {
+      "epoch": 0.874806800618238,
+      "grad_norm": 10.554380523067245,
+      "learning_rate": 1e-06,
+      "loss": 0.1776,
+      "step": 6733
+    },
+    {
+      "epoch": 0.8778979907264297,
+      "grad_norm": 8.314481727843475,
+      "learning_rate": 1e-06,
+      "loss": 0.1748,
+      "step": 6734
+    },
+    {
+      "epoch": 0.8809891808346213,
+      "grad_norm": 8.910566186101917,
+      "learning_rate": 1e-06,
+      "loss": 0.1906,
+      "step": 6735
+    },
+    {
+      "epoch": 0.884080370942813,
+      "grad_norm": 7.811319837809256,
+      "learning_rate": 1e-06,
+      "loss": 0.1726,
+      "step": 6736
+    },
+    {
+      "epoch": 0.8871715610510046,
+      "grad_norm": 7.721221220923507,
+      "learning_rate": 1e-06,
+      "loss": 0.1788,
+      "step": 6737
+    },
+    {
+      "epoch": 0.8902627511591963,
+      "grad_norm": 7.580894548881891,
+      "learning_rate": 1e-06,
+      "loss": 0.1796,
+      "step": 6738
+    },
+    {
+      "epoch": 0.893353941267388,
+      "grad_norm": 8.936899534412449,
+      "learning_rate": 1e-06,
+      "loss": 0.1864,
+      "step": 6739
+    },
+    {
+      "epoch": 0.8964451313755796,
+      "grad_norm": 10.420319453723646,
+      "learning_rate": 1e-06,
+      "loss": 0.181,
+      "step": 6740
+    },
+    {
+      "epoch": 0.8995363214837713,
+      "grad_norm": 11.203407997138449,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 6741
+    },
+    {
+      "epoch": 0.9026275115919629,
+      "grad_norm": 7.858979612182452,
+      "learning_rate": 1e-06,
+      "loss": 0.1871,
+      "step": 6742
+    },
+    {
+      "epoch": 0.9057187017001546,
+      "grad_norm": 9.223247834927479,
+      "learning_rate": 1e-06,
+      "loss": 0.18,
+      "step": 6743
+    },
+    {
+      "epoch": 0.9088098918083463,
+      "grad_norm": 8.792557196193107,
+      "learning_rate": 1e-06,
+      "loss": 0.19,
+      "step": 6744
+    },
+    {
+      "epoch": 0.9119010819165378,
+      "grad_norm": 8.046590859887717,
+      "learning_rate": 1e-06,
+      "loss": 0.177,
+      "step": 6745
+    },
+    {
+      "epoch": 0.9149922720247295,
+      "grad_norm": 8.35949499715928,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 6746
+    },
+    {
+      "epoch": 0.9180834621329211,
+      "grad_norm": 7.714179171445471,
+      "learning_rate": 1e-06,
+      "loss": 0.1927,
+      "step": 6747
+    },
+    {
+      "epoch": 0.9211746522411128,
+      "grad_norm": 9.13296500598022,
+      "learning_rate": 1e-06,
+      "loss": 0.1841,
+      "step": 6748
+    },
+    {
+      "epoch": 0.9242658423493045,
+      "grad_norm": 8.101996083262602,
+      "learning_rate": 1e-06,
+      "loss": 0.1928,
+      "step": 6749
+    },
+    {
+      "epoch": 0.9273570324574961,
+      "grad_norm": 11.082058325361057,
+      "learning_rate": 1e-06,
+      "loss": 0.192,
+      "step": 6750
+    },
+    {
+      "epoch": 0.9304482225656878,
+      "grad_norm": 6.779010013395291,
+      "learning_rate": 1e-06,
+      "loss": 0.1913,
+      "step": 6751
+    },
+    {
+      "epoch": 0.9335394126738794,
+      "grad_norm": 8.914134173448314,
+      "learning_rate": 1e-06,
+      "loss": 0.1833,
+      "step": 6752
+    },
+    {
+      "epoch": 0.9366306027820711,
+      "grad_norm": 8.121103224409355,
+      "learning_rate": 1e-06,
+      "loss": 0.1856,
+      "step": 6753
+    },
+    {
+      "epoch": 0.9397217928902627,
+      "grad_norm": 6.414404245334248,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 6754
+    },
+    {
+      "epoch": 0.9428129829984544,
+      "grad_norm": 7.3591666889106255,
+      "learning_rate": 1e-06,
+      "loss": 0.1764,
+      "step": 6755
+    },
+    {
+      "epoch": 0.9459041731066461,
+      "grad_norm": 6.740891000720759,
+      "learning_rate": 1e-06,
+      "loss": 0.1882,
+      "step": 6756
+    },
+    {
+      "epoch": 0.9489953632148377,
+      "grad_norm": 8.999082967162321,
+      "learning_rate": 1e-06,
+      "loss": 0.1905,
+      "step": 6757
+    },
+    {
+      "epoch": 0.9520865533230294,
+      "grad_norm": 8.541633241128164,
+      "learning_rate": 1e-06,
+      "loss": 0.1937,
+      "step": 6758
+    },
+    {
+      "epoch": 0.955177743431221,
+      "grad_norm": 7.320534691309373,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 6759
+    },
+    {
+      "epoch": 0.9582689335394127,
+      "grad_norm": 8.290725592080095,
+      "learning_rate": 1e-06,
+      "loss": 0.1894,
+      "step": 6760
+    },
+    {
+      "epoch": 0.9613601236476044,
+      "grad_norm": 6.931575119706512,
+      "learning_rate": 1e-06,
+      "loss": 0.1759,
+      "step": 6761
+    },
+    {
+      "epoch": 0.964451313755796,
+      "grad_norm": 7.5733372376273635,
+      "learning_rate": 1e-06,
+      "loss": 0.171,
+      "step": 6762
+    },
+    {
+      "epoch": 0.9675425038639877,
+      "grad_norm": 7.983397352498532,
+      "learning_rate": 1e-06,
+      "loss": 0.1774,
+      "step": 6763
+    },
+    {
+      "epoch": 0.9706336939721792,
+      "grad_norm": 8.696978421976441,
+      "learning_rate": 1e-06,
+      "loss": 0.179,
+      "step": 6764
+    },
+    {
+      "epoch": 0.973724884080371,
+      "grad_norm": 8.972344874396116,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 6765
+    },
+    {
+      "epoch": 0.9768160741885626,
+      "grad_norm": 7.472409582653311,
+      "learning_rate": 1e-06,
+      "loss": 0.1959,
+      "step": 6766
+    },
+    {
+      "epoch": 0.9799072642967542,
+      "grad_norm": 8.403472234742447,
+      "learning_rate": 1e-06,
+      "loss": 0.1969,
+      "step": 6767
+    },
+    {
+      "epoch": 0.9829984544049459,
+      "grad_norm": 6.101917162313505,
+      "learning_rate": 1e-06,
+      "loss": 0.1734,
+      "step": 6768
+    },
+    {
+      "epoch": 0.9860896445131375,
+      "grad_norm": 10.00768625042411,
+      "learning_rate": 1e-06,
+      "loss": 0.1884,
+      "step": 6769
+    },
+    {
+      "epoch": 0.9891808346213292,
+      "grad_norm": 8.972901540026541,
+      "learning_rate": 1e-06,
+      "loss": 0.1876,
+      "step": 6770
+    },
+    {
+      "epoch": 0.9922720247295209,
+      "grad_norm": 7.222763945272477,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 6771
+    },
+    {
+      "epoch": 0.9953632148377125,
+      "grad_norm": 6.660289009667922,
+      "learning_rate": 1e-06,
+      "loss": 0.1877,
+      "step": 6772
+    },
+    {
+      "epoch": 0.9984544049459042,
+      "grad_norm": 7.7811315295554895,
+      "learning_rate": 1e-06,
+      "loss": 0.1885,
+      "step": 6773
+    },
+    {
+      "epoch": 0.9984544049459042,
+      "eval_loss": 3.7303085327148438,
+      "eval_runtime": 21.875,
+      "eval_samples_per_second": 45.714,
+      "eval_steps_per_second": 5.714,
+      "step": 6773
+    },
+    {
+      "epoch": 0.002207505518763797,
+      "grad_norm": 6.9541483891943106,
+      "learning_rate": 1e-06,
+      "loss": 0.2652,
+      "step": 6774
+    },
+    {
+      "epoch": 0.004415011037527594,
+      "grad_norm": 9.467537071399246,
+      "learning_rate": 1e-06,
+      "loss": 0.2703,
+      "step": 6775
+    },
+    {
+      "epoch": 0.006622516556291391,
+      "grad_norm": 9.722196693530288,
+      "learning_rate": 1e-06,
+      "loss": 0.2456,
+      "step": 6776
+    },
+    {
+      "epoch": 0.008830022075055188,
+      "grad_norm": 8.687278154093326,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 6777
+    },
+    {
+      "epoch": 0.011037527593818985,
+      "grad_norm": 9.087151233761842,
+      "learning_rate": 1e-06,
+      "loss": 0.2652,
+      "step": 6778
+    },
+    {
+      "epoch": 0.013245033112582781,
+      "grad_norm": 7.7311814488729675,
+      "learning_rate": 1e-06,
+      "loss": 0.2586,
+      "step": 6779
+    },
+    {
+      "epoch": 0.01545253863134658,
+      "grad_norm": 5.430174481031411,
+      "learning_rate": 1e-06,
+      "loss": 0.2599,
+      "step": 6780
+    },
+    {
+      "epoch": 0.017660044150110375,
+      "grad_norm": 6.608075352318515,
+      "learning_rate": 1e-06,
+      "loss": 0.2489,
+      "step": 6781
+    },
+    {
+      "epoch": 0.019867549668874173,
+      "grad_norm": 7.22437782840186,
+      "learning_rate": 1e-06,
+      "loss": 0.265,
+      "step": 6782
+    },
+    {
+      "epoch": 0.02207505518763797,
+      "grad_norm": 6.219340235120875,
+      "learning_rate": 1e-06,
+      "loss": 0.2482,
+      "step": 6783
+    },
+    {
+      "epoch": 0.024282560706401765,
+      "grad_norm": 6.843024172591394,
+      "learning_rate": 1e-06,
+      "loss": 0.245,
+      "step": 6784
+    },
+    {
+      "epoch": 0.026490066225165563,
+      "grad_norm": 6.875614507261631,
+      "learning_rate": 1e-06,
+      "loss": 0.2623,
+      "step": 6785
+    },
+    {
+      "epoch": 0.02869757174392936,
+      "grad_norm": 9.574149485200465,
+      "learning_rate": 1e-06,
+      "loss": 0.2704,
+      "step": 6786
+    },
+    {
+      "epoch": 0.03090507726269316,
+      "grad_norm": 8.840563028167834,
+      "learning_rate": 1e-06,
+      "loss": 0.2679,
+      "step": 6787
+    },
+    {
+      "epoch": 0.033112582781456956,
+      "grad_norm": 7.998710891269987,
+      "learning_rate": 1e-06,
+      "loss": 0.289,
+      "step": 6788
+    },
+    {
+      "epoch": 0.03532008830022075,
+      "grad_norm": 6.4613134364227,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 6789
+    },
+    {
+      "epoch": 0.037527593818984545,
+      "grad_norm": 6.841001127739274,
+      "learning_rate": 1e-06,
+      "loss": 0.2517,
+      "step": 6790
+    },
+    {
+      "epoch": 0.039735099337748346,
+      "grad_norm": 5.743038273977431,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 6791
+    },
+    {
+      "epoch": 0.04194260485651214,
+      "grad_norm": 5.904760952288504,
+      "learning_rate": 1e-06,
+      "loss": 0.2549,
+      "step": 6792
+    },
+    {
+      "epoch": 0.04415011037527594,
+      "grad_norm": 7.018379945256303,
+      "learning_rate": 1e-06,
+      "loss": 0.2658,
+      "step": 6793
+    },
+    {
+      "epoch": 0.046357615894039736,
+      "grad_norm": 6.850152297512068,
+      "learning_rate": 1e-06,
+      "loss": 0.2794,
+      "step": 6794
+    },
+    {
+      "epoch": 0.04856512141280353,
+      "grad_norm": 7.276017812650683,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 6795
+    },
+    {
+      "epoch": 0.05077262693156733,
+      "grad_norm": 7.772091700354643,
+      "learning_rate": 1e-06,
+      "loss": 0.2618,
+      "step": 6796
+    },
+    {
+      "epoch": 0.052980132450331126,
+      "grad_norm": 7.09640872020233,
+      "learning_rate": 1e-06,
+      "loss": 0.2788,
+      "step": 6797
+    },
+    {
+      "epoch": 0.05518763796909492,
+      "grad_norm": 7.341466758662004,
+      "learning_rate": 1e-06,
+      "loss": 0.2395,
+      "step": 6798
+    },
+    {
+      "epoch": 0.05739514348785872,
+      "grad_norm": 7.266109715256356,
+      "learning_rate": 1e-06,
+      "loss": 0.2731,
+      "step": 6799
+    },
+    {
+      "epoch": 0.059602649006622516,
+      "grad_norm": 8.465985323012811,
+      "learning_rate": 1e-06,
+      "loss": 0.2592,
+      "step": 6800
+    },
+    {
+      "epoch": 0.06181015452538632,
+      "grad_norm": 6.6517189688801786,
+      "learning_rate": 1e-06,
+      "loss": 0.2799,
+      "step": 6801
+    },
+    {
+      "epoch": 0.0640176600441501,
+      "grad_norm": 6.28729626711318,
+      "learning_rate": 1e-06,
+      "loss": 0.2535,
+      "step": 6802
+    },
+    {
+      "epoch": 0.06622516556291391,
+      "grad_norm": 7.71632842314092,
+      "learning_rate": 1e-06,
+      "loss": 0.271,
+      "step": 6803
+    },
+    {
+      "epoch": 0.0684326710816777,
+      "grad_norm": 6.686199394981638,
+      "learning_rate": 1e-06,
+      "loss": 0.2512,
+      "step": 6804
+    },
+    {
+      "epoch": 0.0706401766004415,
+      "grad_norm": 7.997573793746047,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 6805
+    },
+    {
+      "epoch": 0.0728476821192053,
+      "grad_norm": 7.018075906825007,
+      "learning_rate": 1e-06,
+      "loss": 0.2747,
+      "step": 6806
+    },
+    {
+      "epoch": 0.07505518763796909,
+      "grad_norm": 8.845693076406604,
+      "learning_rate": 1e-06,
+      "loss": 0.2568,
+      "step": 6807
+    },
+    {
+      "epoch": 0.0772626931567329,
+      "grad_norm": 7.475095900106726,
+      "learning_rate": 1e-06,
+      "loss": 0.2498,
+      "step": 6808
+    },
+    {
+      "epoch": 0.07947019867549669,
+      "grad_norm": 6.133139022113882,
+      "learning_rate": 1e-06,
+      "loss": 0.2454,
+      "step": 6809
+    },
+    {
+      "epoch": 0.08167770419426049,
+      "grad_norm": 7.423304350630173,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 6810
+    },
+    {
+      "epoch": 0.08388520971302428,
+      "grad_norm": 9.674804871554233,
+      "learning_rate": 1e-06,
+      "loss": 0.2457,
+      "step": 6811
+    },
+    {
+      "epoch": 0.08609271523178808,
+      "grad_norm": 7.351974862868904,
+      "learning_rate": 1e-06,
+      "loss": 0.2634,
+      "step": 6812
+    },
+    {
+      "epoch": 0.08830022075055188,
+      "grad_norm": 10.124112420076491,
+      "learning_rate": 1e-06,
+      "loss": 0.2579,
+      "step": 6813
+    },
+    {
+      "epoch": 0.09050772626931568,
+      "grad_norm": 8.254033279595168,
+      "learning_rate": 1e-06,
+      "loss": 0.267,
+      "step": 6814
+    },
+    {
+      "epoch": 0.09271523178807947,
+      "grad_norm": 8.301998208062923,
+      "learning_rate": 1e-06,
+      "loss": 0.2783,
+      "step": 6815
+    },
+    {
+      "epoch": 0.09492273730684327,
+      "grad_norm": 7.2344375511513945,
+      "learning_rate": 1e-06,
+      "loss": 0.2652,
+      "step": 6816
+    },
+    {
+      "epoch": 0.09713024282560706,
+      "grad_norm": 7.354517375171986,
+      "learning_rate": 1e-06,
+      "loss": 0.2881,
+      "step": 6817
+    },
+    {
+      "epoch": 0.09933774834437085,
+      "grad_norm": 7.208145351752179,
+      "learning_rate": 1e-06,
+      "loss": 0.2359,
+      "step": 6818
+    },
+    {
+      "epoch": 0.10154525386313466,
+      "grad_norm": 7.046089745787376,
+      "learning_rate": 1e-06,
+      "loss": 0.2495,
+      "step": 6819
+    },
+    {
+      "epoch": 0.10375275938189846,
+      "grad_norm": 7.364577932564958,
+      "learning_rate": 1e-06,
+      "loss": 0.2601,
+      "step": 6820
+    },
+    {
+      "epoch": 0.10596026490066225,
+      "grad_norm": 7.896400592022992,
+      "learning_rate": 1e-06,
+      "loss": 0.2813,
+      "step": 6821
+    },
+    {
+      "epoch": 0.10816777041942605,
+      "grad_norm": 6.830067557286022,
+      "learning_rate": 1e-06,
+      "loss": 0.2489,
+      "step": 6822
+    },
+    {
+      "epoch": 0.11037527593818984,
+      "grad_norm": 7.10720294740419,
+      "learning_rate": 1e-06,
+      "loss": 0.2755,
+      "step": 6823
+    },
+    {
+      "epoch": 0.11258278145695365,
+      "grad_norm": 7.054039825077189,
+      "learning_rate": 1e-06,
+      "loss": 0.2476,
+      "step": 6824
+    },
+    {
+      "epoch": 0.11479028697571744,
+      "grad_norm": 7.652916575444127,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 6825
+    },
+    {
+      "epoch": 0.11699779249448124,
+      "grad_norm": 8.07412512030686,
+      "learning_rate": 1e-06,
+      "loss": 0.2833,
+      "step": 6826
+    },
+    {
+      "epoch": 0.11920529801324503,
+      "grad_norm": 7.023504097034574,
+      "learning_rate": 1e-06,
+      "loss": 0.2696,
+      "step": 6827
+    },
+    {
+      "epoch": 0.12141280353200883,
+      "grad_norm": 8.616398720106647,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 6828
+    },
+    {
+      "epoch": 0.12362030905077263,
+      "grad_norm": 6.543282977803763,
+      "learning_rate": 1e-06,
+      "loss": 0.2487,
+      "step": 6829
+    },
+    {
+      "epoch": 0.12582781456953643,
+      "grad_norm": 6.140153763294262,
+      "learning_rate": 1e-06,
+      "loss": 0.2922,
+      "step": 6830
+    },
+    {
+      "epoch": 0.1280353200883002,
+      "grad_norm": 7.590296350387899,
+      "learning_rate": 1e-06,
+      "loss": 0.2611,
+      "step": 6831
+    },
+    {
+      "epoch": 0.13024282560706402,
+      "grad_norm": 7.355832859511199,
+      "learning_rate": 1e-06,
+      "loss": 0.2709,
+      "step": 6832
+    },
+    {
+      "epoch": 0.13245033112582782,
+      "grad_norm": 7.188167842176103,
+      "learning_rate": 1e-06,
+      "loss": 0.2753,
+      "step": 6833
+    },
+    {
+      "epoch": 0.1346578366445916,
+      "grad_norm": 7.736145781513598,
+      "learning_rate": 1e-06,
+      "loss": 0.2511,
+      "step": 6834
+    },
+    {
+      "epoch": 0.1368653421633554,
+      "grad_norm": 9.07640787146278,
+      "learning_rate": 1e-06,
+      "loss": 0.2743,
+      "step": 6835
+    },
+    {
+      "epoch": 0.1390728476821192,
+      "grad_norm": 7.485829622761397,
+      "learning_rate": 1e-06,
+      "loss": 0.2553,
+      "step": 6836
+    },
+    {
+      "epoch": 0.141280353200883,
+      "grad_norm": 8.916999271670125,
+      "learning_rate": 1e-06,
+      "loss": 0.267,
+      "step": 6837
+    },
+    {
+      "epoch": 0.1434878587196468,
+      "grad_norm": 9.305539028798373,
+      "learning_rate": 1e-06,
+      "loss": 0.2591,
+      "step": 6838
+    },
+    {
+      "epoch": 0.1456953642384106,
+      "grad_norm": 6.635176425075537,
+      "learning_rate": 1e-06,
+      "loss": 0.2547,
+      "step": 6839
+    },
+    {
+      "epoch": 0.1479028697571744,
+      "grad_norm": 5.966500435651503,
+      "learning_rate": 1e-06,
+      "loss": 0.2728,
+      "step": 6840
+    },
+    {
+      "epoch": 0.15011037527593818,
+      "grad_norm": 9.073761560568158,
+      "learning_rate": 1e-06,
+      "loss": 0.252,
+      "step": 6841
+    },
+    {
+      "epoch": 0.152317880794702,
+      "grad_norm": 7.50713797935282,
+      "learning_rate": 1e-06,
+      "loss": 0.2602,
+      "step": 6842
+    },
+    {
+      "epoch": 0.1545253863134658,
+      "grad_norm": 7.223669793046364,
+      "learning_rate": 1e-06,
+      "loss": 0.2671,
+      "step": 6843
+    },
+    {
+      "epoch": 0.15673289183222958,
+      "grad_norm": 7.193941158220929,
+      "learning_rate": 1e-06,
+      "loss": 0.2403,
+      "step": 6844
+    },
+    {
+      "epoch": 0.15894039735099338,
+      "grad_norm": 7.223157509908992,
+      "learning_rate": 1e-06,
+      "loss": 0.2375,
+      "step": 6845
+    },
+    {
+      "epoch": 0.16114790286975716,
+      "grad_norm": 7.762646421780506,
+      "learning_rate": 1e-06,
+      "loss": 0.2519,
+      "step": 6846
+    },
+    {
+      "epoch": 0.16335540838852097,
+      "grad_norm": 8.77716024305362,
+      "learning_rate": 1e-06,
+      "loss": 0.2701,
+      "step": 6847
+    },
+    {
+      "epoch": 0.16556291390728478,
+      "grad_norm": 6.253997677458488,
+      "learning_rate": 1e-06,
+      "loss": 0.2797,
+      "step": 6848
+    },
+    {
+      "epoch": 0.16777041942604856,
+      "grad_norm": 8.635964269218691,
+      "learning_rate": 1e-06,
+      "loss": 0.268,
+      "step": 6849
+    },
+    {
+      "epoch": 0.16997792494481237,
+      "grad_norm": 6.1026761882311575,
+      "learning_rate": 1e-06,
+      "loss": 0.2782,
+      "step": 6850
+    },
+    {
+      "epoch": 0.17218543046357615,
+      "grad_norm": 10.015635670492678,
+      "learning_rate": 1e-06,
+      "loss": 0.2772,
+      "step": 6851
+    },
+    {
+      "epoch": 0.17439293598233996,
+      "grad_norm": 9.891069031371275,
+      "learning_rate": 1e-06,
+      "loss": 0.2628,
+      "step": 6852
+    },
+    {
+      "epoch": 0.17660044150110377,
+      "grad_norm": 6.703631606321109,
+      "learning_rate": 1e-06,
+      "loss": 0.2558,
+      "step": 6853
+    },
+    {
+      "epoch": 0.17880794701986755,
+      "grad_norm": 8.784115936936965,
+      "learning_rate": 1e-06,
+      "loss": 0.2648,
+      "step": 6854
+    },
+    {
+      "epoch": 0.18101545253863136,
+      "grad_norm": 8.789107952013639,
+      "learning_rate": 1e-06,
+      "loss": 0.3055,
+      "step": 6855
+    },
+    {
+      "epoch": 0.18322295805739514,
+      "grad_norm": 8.221459107828966,
+      "learning_rate": 1e-06,
+      "loss": 0.2673,
+      "step": 6856
+    },
+    {
+      "epoch": 0.18543046357615894,
+      "grad_norm": 7.202512847369432,
+      "learning_rate": 1e-06,
+      "loss": 0.263,
+      "step": 6857
+    },
+    {
+      "epoch": 0.18763796909492272,
+      "grad_norm": 7.2019646507590735,
+      "learning_rate": 1e-06,
+      "loss": 0.2603,
+      "step": 6858
+    },
+    {
+      "epoch": 0.18984547461368653,
+      "grad_norm": 7.385451056016387,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 6859
+    },
+    {
+      "epoch": 0.19205298013245034,
+      "grad_norm": 6.326999287993438,
+      "learning_rate": 1e-06,
+      "loss": 0.2765,
+      "step": 6860
+    },
+    {
+      "epoch": 0.19426048565121412,
+      "grad_norm": 9.113620786215623,
+      "learning_rate": 1e-06,
+      "loss": 0.2804,
+      "step": 6861
+    },
+    {
+      "epoch": 0.19646799116997793,
+      "grad_norm": 7.2981816040686365,
+      "learning_rate": 1e-06,
+      "loss": 0.2584,
+      "step": 6862
+    },
+    {
+      "epoch": 0.1986754966887417,
+      "grad_norm": 6.801985469439551,
+      "learning_rate": 1e-06,
+      "loss": 0.2673,
+      "step": 6863
+    },
+    {
+      "epoch": 0.20088300220750552,
+      "grad_norm": 7.0966344809690085,
+      "learning_rate": 1e-06,
+      "loss": 0.2612,
+      "step": 6864
+    },
+    {
+      "epoch": 0.20309050772626933,
+      "grad_norm": 8.254216747194354,
+      "learning_rate": 1e-06,
+      "loss": 0.2544,
+      "step": 6865
+    },
+    {
+      "epoch": 0.2052980132450331,
+      "grad_norm": 7.348771779577696,
+      "learning_rate": 1e-06,
+      "loss": 0.2496,
+      "step": 6866
+    },
+    {
+      "epoch": 0.20750551876379691,
+      "grad_norm": 6.687390359290394,
+      "learning_rate": 1e-06,
+      "loss": 0.2464,
+      "step": 6867
+    },
+    {
+      "epoch": 0.2097130242825607,
+      "grad_norm": 5.077322263497198,
+      "learning_rate": 1e-06,
+      "loss": 0.2844,
+      "step": 6868
+    },
+    {
+      "epoch": 0.2119205298013245,
+      "grad_norm": 8.02255345369268,
+      "learning_rate": 1e-06,
+      "loss": 0.2581,
+      "step": 6869
+    },
+    {
+      "epoch": 0.2141280353200883,
+      "grad_norm": 8.612192759347227,
+      "learning_rate": 1e-06,
+      "loss": 0.2846,
+      "step": 6870
+    },
+    {
+      "epoch": 0.2163355408388521,
+      "grad_norm": 7.003958054754921,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 6871
+    },
+    {
+      "epoch": 0.2185430463576159,
+      "grad_norm": 7.151022146290699,
+      "learning_rate": 1e-06,
+      "loss": 0.2702,
+      "step": 6872
+    },
+    {
+      "epoch": 0.22075055187637968,
+      "grad_norm": 5.787258835327641,
+      "learning_rate": 1e-06,
+      "loss": 0.2618,
+      "step": 6873
+    },
+    {
+      "epoch": 0.2229580573951435,
+      "grad_norm": 6.868573587643637,
+      "learning_rate": 1e-06,
+      "loss": 0.2606,
+      "step": 6874
+    },
+    {
+      "epoch": 0.2251655629139073,
+      "grad_norm": 6.218060336448435,
+      "learning_rate": 1e-06,
+      "loss": 0.2756,
+      "step": 6875
+    },
+    {
+      "epoch": 0.22737306843267108,
+      "grad_norm": 7.695533082827407,
+      "learning_rate": 1e-06,
+      "loss": 0.276,
+      "step": 6876
+    },
+    {
+      "epoch": 0.22958057395143489,
+      "grad_norm": 7.849729552467853,
+      "learning_rate": 1e-06,
+      "loss": 0.2606,
+      "step": 6877
+    },
+    {
+      "epoch": 0.23178807947019867,
+      "grad_norm": 7.08062252263539,
+      "learning_rate": 1e-06,
+      "loss": 0.2958,
+      "step": 6878
+    },
+    {
+      "epoch": 0.23399558498896247,
+      "grad_norm": 8.630095834544296,
+      "learning_rate": 1e-06,
+      "loss": 0.2702,
+      "step": 6879
+    },
+    {
+      "epoch": 0.23620309050772628,
+      "grad_norm": 7.193892865743503,
+      "learning_rate": 1e-06,
+      "loss": 0.2659,
+      "step": 6880
+    },
+    {
+      "epoch": 0.23841059602649006,
+      "grad_norm": 7.376330177482858,
+      "learning_rate": 1e-06,
+      "loss": 0.2687,
+      "step": 6881
+    },
+    {
+      "epoch": 0.24061810154525387,
+      "grad_norm": 7.76238440392513,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 6882
+    },
+    {
+      "epoch": 0.24282560706401765,
+      "grad_norm": 6.594503460615002,
+      "learning_rate": 1e-06,
+      "loss": 0.2633,
+      "step": 6883
+    },
+    {
+      "epoch": 0.24503311258278146,
+      "grad_norm": 5.185732875781994,
+      "learning_rate": 1e-06,
+      "loss": 0.2561,
+      "step": 6884
+    },
+    {
+      "epoch": 0.24724061810154527,
+      "grad_norm": 7.998930113851942,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 6885
+    },
+    {
+      "epoch": 0.24944812362030905,
+      "grad_norm": 6.465096024401925,
+      "learning_rate": 1e-06,
+      "loss": 0.2806,
+      "step": 6886
+    },
+    {
+      "epoch": 0.25165562913907286,
+      "grad_norm": 6.894497070746153,
+      "learning_rate": 1e-06,
+      "loss": 0.2544,
+      "step": 6887
+    },
+    {
+      "epoch": 0.25386313465783666,
+      "grad_norm": 5.562975508555851,
+      "learning_rate": 1e-06,
+      "loss": 0.245,
+      "step": 6888
+    },
+    {
+      "epoch": 0.2560706401766004,
+      "grad_norm": 8.500682560426007,
+      "learning_rate": 1e-06,
+      "loss": 0.2934,
+      "step": 6889
+    },
+    {
+      "epoch": 0.2582781456953642,
+      "grad_norm": 6.8324984987988575,
+      "learning_rate": 1e-06,
+      "loss": 0.2504,
+      "step": 6890
+    },
+    {
+      "epoch": 0.26048565121412803,
+      "grad_norm": 7.281109953623135,
+      "learning_rate": 1e-06,
+      "loss": 0.2714,
+      "step": 6891
+    },
+    {
+      "epoch": 0.26269315673289184,
+      "grad_norm": 8.849305575487495,
+      "learning_rate": 1e-06,
+      "loss": 0.2432,
+      "step": 6892
+    },
+    {
+      "epoch": 0.26490066225165565,
+      "grad_norm": 8.270390181657717,
+      "learning_rate": 1e-06,
+      "loss": 0.2702,
+      "step": 6893
+    },
+    {
+      "epoch": 0.2671081677704194,
+      "grad_norm": 8.444504708195252,
+      "learning_rate": 1e-06,
+      "loss": 0.2626,
+      "step": 6894
+    },
+    {
+      "epoch": 0.2693156732891832,
+      "grad_norm": 6.187358048310921,
+      "learning_rate": 1e-06,
+      "loss": 0.2657,
+      "step": 6895
+    },
+    {
+      "epoch": 0.271523178807947,
+      "grad_norm": 5.451083522346802,
+      "learning_rate": 1e-06,
+      "loss": 0.266,
+      "step": 6896
+    },
+    {
+      "epoch": 0.2737306843267108,
+      "grad_norm": 8.66100060523019,
+      "learning_rate": 1e-06,
+      "loss": 0.2653,
+      "step": 6897
+    },
+    {
+      "epoch": 0.27593818984547464,
+      "grad_norm": 6.738800315197182,
+      "learning_rate": 1e-06,
+      "loss": 0.2618,
+      "step": 6898
+    },
+    {
+      "epoch": 0.2781456953642384,
+      "grad_norm": 8.447137142937999,
+      "learning_rate": 1e-06,
+      "loss": 0.2613,
+      "step": 6899
+    },
+    {
+      "epoch": 0.2803532008830022,
+      "grad_norm": 7.675092824595699,
+      "learning_rate": 1e-06,
+      "loss": 0.2703,
+      "step": 6900
+    },
+    {
+      "epoch": 0.282560706401766,
+      "grad_norm": 8.424042623589633,
+      "learning_rate": 1e-06,
+      "loss": 0.269,
+      "step": 6901
+    },
+    {
+      "epoch": 0.2847682119205298,
+      "grad_norm": 7.127366620317775,
+      "learning_rate": 1e-06,
+      "loss": 0.2772,
+      "step": 6902
+    },
+    {
+      "epoch": 0.2869757174392936,
+      "grad_norm": 7.988906311740312,
+      "learning_rate": 1e-06,
+      "loss": 0.2482,
+      "step": 6903
+    },
+    {
+      "epoch": 0.2891832229580574,
+      "grad_norm": 5.9847497315398375,
+      "learning_rate": 1e-06,
+      "loss": 0.2852,
+      "step": 6904
+    },
+    {
+      "epoch": 0.2913907284768212,
+      "grad_norm": 6.310738662685346,
+      "learning_rate": 1e-06,
+      "loss": 0.2573,
+      "step": 6905
+    },
+    {
+      "epoch": 0.293598233995585,
+      "grad_norm": 7.6342813884320595,
+      "learning_rate": 1e-06,
+      "loss": 0.2528,
+      "step": 6906
+    },
+    {
+      "epoch": 0.2958057395143488,
+      "grad_norm": 8.543651587370586,
+      "learning_rate": 1e-06,
+      "loss": 0.2762,
+      "step": 6907
+    },
+    {
+      "epoch": 0.2980132450331126,
+      "grad_norm": 8.25971501396179,
+      "learning_rate": 1e-06,
+      "loss": 0.2728,
+      "step": 6908
+    },
+    {
+      "epoch": 0.30022075055187636,
+      "grad_norm": 6.29456022967607,
+      "learning_rate": 1e-06,
+      "loss": 0.257,
+      "step": 6909
+    },
+    {
+      "epoch": 0.30242825607064017,
+      "grad_norm": 6.4447365015571885,
+      "learning_rate": 1e-06,
+      "loss": 0.2627,
+      "step": 6910
+    },
+    {
+      "epoch": 0.304635761589404,
+      "grad_norm": 7.054166312338007,
+      "learning_rate": 1e-06,
+      "loss": 0.2549,
+      "step": 6911
+    },
+    {
+      "epoch": 0.3068432671081678,
+      "grad_norm": 7.626300033176331,
+      "learning_rate": 1e-06,
+      "loss": 0.2527,
+      "step": 6912
+    },
+    {
+      "epoch": 0.3090507726269316,
+      "grad_norm": 7.454968983099517,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 6913
+    },
+    {
+      "epoch": 0.31125827814569534,
+      "grad_norm": 9.002018166106067,
+      "learning_rate": 1e-06,
+      "loss": 0.2613,
+      "step": 6914
+    },
+    {
+      "epoch": 0.31346578366445915,
+      "grad_norm": 8.904621116252654,
+      "learning_rate": 1e-06,
+      "loss": 0.2444,
+      "step": 6915
+    },
+    {
+      "epoch": 0.31567328918322296,
+      "grad_norm": 8.059316153304996,
+      "learning_rate": 1e-06,
+      "loss": 0.254,
+      "step": 6916
+    },
+    {
+      "epoch": 0.31788079470198677,
+      "grad_norm": 8.343588924810712,
+      "learning_rate": 1e-06,
+      "loss": 0.2595,
+      "step": 6917
+    },
+    {
+      "epoch": 0.3200883002207506,
+      "grad_norm": 8.443049340146365,
+      "learning_rate": 1e-06,
+      "loss": 0.2648,
+      "step": 6918
+    },
+    {
+      "epoch": 0.32229580573951433,
+      "grad_norm": 8.18194257142921,
+      "learning_rate": 1e-06,
+      "loss": 0.2472,
+      "step": 6919
+    },
+    {
+      "epoch": 0.32450331125827814,
+      "grad_norm": 6.59085791266125,
+      "learning_rate": 1e-06,
+      "loss": 0.2578,
+      "step": 6920
+    },
+    {
+      "epoch": 0.32671081677704195,
+      "grad_norm": 9.16510323768497,
+      "learning_rate": 1e-06,
+      "loss": 0.2484,
+      "step": 6921
+    },
+    {
+      "epoch": 0.32891832229580575,
+      "grad_norm": 8.269008526323466,
+      "learning_rate": 1e-06,
+      "loss": 0.2751,
+      "step": 6922
+    },
+    {
+      "epoch": 0.33112582781456956,
+      "grad_norm": 8.835296913105989,
+      "learning_rate": 1e-06,
+      "loss": 0.2542,
+      "step": 6923
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 5.1508316421998215,
+      "learning_rate": 1e-06,
+      "loss": 0.255,
+      "step": 6924
+    },
+    {
+      "epoch": 0.3355408388520971,
+      "grad_norm": 8.03654732164176,
+      "learning_rate": 1e-06,
+      "loss": 0.2584,
+      "step": 6925
+    },
+    {
+      "epoch": 0.33774834437086093,
+      "grad_norm": 7.941430450614318,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 6926
+    },
+    {
+      "epoch": 0.33995584988962474,
+      "grad_norm": 8.725626520917427,
+      "learning_rate": 1e-06,
+      "loss": 0.2748,
+      "step": 6927
+    },
+    {
+      "epoch": 0.34216335540838855,
+      "grad_norm": 9.252254948705605,
+      "learning_rate": 1e-06,
+      "loss": 0.2701,
+      "step": 6928
+    },
+    {
+      "epoch": 0.3443708609271523,
+      "grad_norm": 7.714609669849294,
+      "learning_rate": 1e-06,
+      "loss": 0.2856,
+      "step": 6929
+    },
+    {
+      "epoch": 0.3465783664459161,
+      "grad_norm": 8.465594818064448,
+      "learning_rate": 1e-06,
+      "loss": 0.2858,
+      "step": 6930
+    },
+    {
+      "epoch": 0.3487858719646799,
+      "grad_norm": 7.673761455345068,
+      "learning_rate": 1e-06,
+      "loss": 0.2581,
+      "step": 6931
+    },
+    {
+      "epoch": 0.3509933774834437,
+      "grad_norm": 7.063263176362585,
+      "learning_rate": 1e-06,
+      "loss": 0.2522,
+      "step": 6932
+    },
+    {
+      "epoch": 0.35320088300220753,
+      "grad_norm": 6.4824541346990605,
+      "learning_rate": 1e-06,
+      "loss": 0.2483,
+      "step": 6933
+    },
+    {
+      "epoch": 0.3554083885209713,
+      "grad_norm": 5.393994775673176,
+      "learning_rate": 1e-06,
+      "loss": 0.2842,
+      "step": 6934
+    },
+    {
+      "epoch": 0.3576158940397351,
+      "grad_norm": 9.149417546263267,
+      "learning_rate": 1e-06,
+      "loss": 0.2705,
+      "step": 6935
+    },
+    {
+      "epoch": 0.3598233995584989,
+      "grad_norm": 6.801441536654397,
+      "learning_rate": 1e-06,
+      "loss": 0.2634,
+      "step": 6936
+    },
+    {
+      "epoch": 0.3620309050772627,
+      "grad_norm": 6.609170250523191,
+      "learning_rate": 1e-06,
+      "loss": 0.2556,
+      "step": 6937
+    },
+    {
+      "epoch": 0.36423841059602646,
+      "grad_norm": 8.173281704932709,
+      "learning_rate": 1e-06,
+      "loss": 0.2593,
+      "step": 6938
+    },
+    {
+      "epoch": 0.36644591611479027,
+      "grad_norm": 6.641326359901027,
+      "learning_rate": 1e-06,
+      "loss": 0.2554,
+      "step": 6939
+    },
+    {
+      "epoch": 0.3686534216335541,
+      "grad_norm": 7.392872624920044,
+      "learning_rate": 1e-06,
+      "loss": 0.2402,
+      "step": 6940
+    },
+    {
+      "epoch": 0.3708609271523179,
+      "grad_norm": 8.241079803850907,
+      "learning_rate": 1e-06,
+      "loss": 0.2609,
+      "step": 6941
+    },
+    {
+      "epoch": 0.3730684326710817,
+      "grad_norm": 8.79963851340694,
+      "learning_rate": 1e-06,
+      "loss": 0.2697,
+      "step": 6942
+    },
+    {
+      "epoch": 0.37527593818984545,
+      "grad_norm": 7.755923885809896,
+      "learning_rate": 1e-06,
+      "loss": 0.2547,
+      "step": 6943
+    },
+    {
+      "epoch": 0.37748344370860926,
+      "grad_norm": 5.94499843611395,
+      "learning_rate": 1e-06,
+      "loss": 0.2723,
+      "step": 6944
+    },
+    {
+      "epoch": 0.37969094922737306,
+      "grad_norm": 8.71204181014534,
+      "learning_rate": 1e-06,
+      "loss": 0.2928,
+      "step": 6945
+    },
+    {
+      "epoch": 0.3818984547461369,
+      "grad_norm": 6.677260218889544,
+      "learning_rate": 1e-06,
+      "loss": 0.2515,
+      "step": 6946
+    },
+    {
+      "epoch": 0.3841059602649007,
+      "grad_norm": 7.756043335056953,
+      "learning_rate": 1e-06,
+      "loss": 0.2567,
+      "step": 6947
+    },
+    {
+      "epoch": 0.38631346578366443,
+      "grad_norm": 6.277050652468283,
+      "learning_rate": 1e-06,
+      "loss": 0.2317,
+      "step": 6948
+    },
+    {
+      "epoch": 0.38852097130242824,
+      "grad_norm": 8.798091595327463,
+      "learning_rate": 1e-06,
+      "loss": 0.2711,
+      "step": 6949
+    },
+    {
+      "epoch": 0.39072847682119205,
+      "grad_norm": 8.143975318358635,
+      "learning_rate": 1e-06,
+      "loss": 0.2523,
+      "step": 6950
+    },
+    {
+      "epoch": 0.39293598233995586,
+      "grad_norm": 6.322997064607684,
+      "learning_rate": 1e-06,
+      "loss": 0.257,
+      "step": 6951
+    },
+    {
+      "epoch": 0.39514348785871967,
+      "grad_norm": 9.000540695765755,
+      "learning_rate": 1e-06,
+      "loss": 0.2675,
+      "step": 6952
+    },
+    {
+      "epoch": 0.3973509933774834,
+      "grad_norm": 7.496272646283537,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 6953
+    },
+    {
+      "epoch": 0.3995584988962472,
+      "grad_norm": 5.9491903718302375,
+      "learning_rate": 1e-06,
+      "loss": 0.2724,
+      "step": 6954
+    },
+    {
+      "epoch": 0.40176600441501104,
+      "grad_norm": 6.117737237369057,
+      "learning_rate": 1e-06,
+      "loss": 0.2585,
+      "step": 6955
+    },
+    {
+      "epoch": 0.40397350993377484,
+      "grad_norm": 8.783325199633346,
+      "learning_rate": 1e-06,
+      "loss": 0.2642,
+      "step": 6956
+    },
+    {
+      "epoch": 0.40618101545253865,
+      "grad_norm": 7.424117571963376,
+      "learning_rate": 1e-06,
+      "loss": 0.2599,
+      "step": 6957
+    },
+    {
+      "epoch": 0.4083885209713024,
+      "grad_norm": 7.808681728375264,
+      "learning_rate": 1e-06,
+      "loss": 0.2518,
+      "step": 6958
+    },
+    {
+      "epoch": 0.4105960264900662,
+      "grad_norm": 6.115860144312745,
+      "learning_rate": 1e-06,
+      "loss": 0.2423,
+      "step": 6959
+    },
+    {
+      "epoch": 0.41280353200883,
+      "grad_norm": 8.225576713048016,
+      "learning_rate": 1e-06,
+      "loss": 0.2805,
+      "step": 6960
+    },
+    {
+      "epoch": 0.41501103752759383,
+      "grad_norm": 8.466780086657986,
+      "learning_rate": 1e-06,
+      "loss": 0.2678,
+      "step": 6961
+    },
+    {
+      "epoch": 0.41721854304635764,
+      "grad_norm": 9.373239685367198,
+      "learning_rate": 1e-06,
+      "loss": 0.2713,
+      "step": 6962
+    },
+    {
+      "epoch": 0.4194260485651214,
+      "grad_norm": 6.127658489958951,
+      "learning_rate": 1e-06,
+      "loss": 0.2753,
+      "step": 6963
+    },
+    {
+      "epoch": 0.4216335540838852,
+      "grad_norm": 8.607331597483073,
+      "learning_rate": 1e-06,
+      "loss": 0.271,
+      "step": 6964
+    },
+    {
+      "epoch": 0.423841059602649,
+      "grad_norm": 7.551451709186536,
+      "learning_rate": 1e-06,
+      "loss": 0.245,
+      "step": 6965
+    },
+    {
+      "epoch": 0.4260485651214128,
+      "grad_norm": 7.38499933926397,
+      "learning_rate": 1e-06,
+      "loss": 0.2697,
+      "step": 6966
+    },
+    {
+      "epoch": 0.4282560706401766,
+      "grad_norm": 6.667081762315985,
+      "learning_rate": 1e-06,
+      "loss": 0.251,
+      "step": 6967
+    },
+    {
+      "epoch": 0.4304635761589404,
+      "grad_norm": 7.686365830639501,
+      "learning_rate": 1e-06,
+      "loss": 0.2913,
+      "step": 6968
+    },
+    {
+      "epoch": 0.4326710816777042,
+      "grad_norm": 8.651823897895927,
+      "learning_rate": 1e-06,
+      "loss": 0.2499,
+      "step": 6969
+    },
+    {
+      "epoch": 0.434878587196468,
+      "grad_norm": 5.4316712303771375,
+      "learning_rate": 1e-06,
+      "loss": 0.2588,
+      "step": 6970
+    },
+    {
+      "epoch": 0.4370860927152318,
+      "grad_norm": 8.36625219658636,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 6971
+    },
+    {
+      "epoch": 0.4392935982339956,
+      "grad_norm": 9.304936518499074,
+      "learning_rate": 1e-06,
+      "loss": 0.2612,
+      "step": 6972
+    },
+    {
+      "epoch": 0.44150110375275936,
+      "grad_norm": 7.857625298263448,
+      "learning_rate": 1e-06,
+      "loss": 0.2425,
+      "step": 6973
+    },
+    {
+      "epoch": 0.44370860927152317,
+      "grad_norm": 8.076547454161009,
+      "learning_rate": 1e-06,
+      "loss": 0.2606,
+      "step": 6974
+    },
+    {
+      "epoch": 0.445916114790287,
+      "grad_norm": 6.771459471868864,
+      "learning_rate": 1e-06,
+      "loss": 0.2536,
+      "step": 6975
+    },
+    {
+      "epoch": 0.4481236203090508,
+      "grad_norm": 8.715297420862191,
+      "learning_rate": 1e-06,
+      "loss": 0.2661,
+      "step": 6976
+    },
+    {
+      "epoch": 0.4503311258278146,
+      "grad_norm": 7.821776066508571,
+      "learning_rate": 1e-06,
+      "loss": 0.2833,
+      "step": 6977
+    },
+    {
+      "epoch": 0.45253863134657835,
+      "grad_norm": 7.853692230850414,
+      "learning_rate": 1e-06,
+      "loss": 0.2768,
+      "step": 6978
+    },
+    {
+      "epoch": 0.45474613686534215,
+      "grad_norm": 8.456005235256017,
+      "learning_rate": 1e-06,
+      "loss": 0.2825,
+      "step": 6979
+    },
+    {
+      "epoch": 0.45695364238410596,
+      "grad_norm": 8.334972094518305,
+      "learning_rate": 1e-06,
+      "loss": 0.2743,
+      "step": 6980
+    },
+    {
+      "epoch": 0.45916114790286977,
+      "grad_norm": 6.924350157738124,
+      "learning_rate": 1e-06,
+      "loss": 0.2996,
+      "step": 6981
+    },
+    {
+      "epoch": 0.4613686534216336,
+      "grad_norm": 6.696456473217004,
+      "learning_rate": 1e-06,
+      "loss": 0.2614,
+      "step": 6982
+    },
+    {
+      "epoch": 0.46357615894039733,
+      "grad_norm": 7.279163512487779,
+      "learning_rate": 1e-06,
+      "loss": 0.2733,
+      "step": 6983
+    },
+    {
+      "epoch": 0.46578366445916114,
+      "grad_norm": 7.105756776694156,
+      "learning_rate": 1e-06,
+      "loss": 0.271,
+      "step": 6984
+    },
+    {
+      "epoch": 0.46799116997792495,
+      "grad_norm": 5.81396769084598,
+      "learning_rate": 1e-06,
+      "loss": 0.2558,
+      "step": 6985
+    },
+    {
+      "epoch": 0.47019867549668876,
+      "grad_norm": 7.061825888730111,
+      "learning_rate": 1e-06,
+      "loss": 0.2689,
+      "step": 6986
+    },
+    {
+      "epoch": 0.47240618101545256,
+      "grad_norm": 7.940189080055314,
+      "learning_rate": 1e-06,
+      "loss": 0.2785,
+      "step": 6987
+    },
+    {
+      "epoch": 0.4746136865342163,
+      "grad_norm": 6.51362924905903,
+      "learning_rate": 1e-06,
+      "loss": 0.2615,
+      "step": 6988
+    },
+    {
+      "epoch": 0.4768211920529801,
+      "grad_norm": 6.95696732106413,
+      "learning_rate": 1e-06,
+      "loss": 0.2577,
+      "step": 6989
+    },
+    {
+      "epoch": 0.47902869757174393,
+      "grad_norm": 7.304601257713644,
+      "learning_rate": 1e-06,
+      "loss": 0.2561,
+      "step": 6990
+    },
+    {
+      "epoch": 0.48123620309050774,
+      "grad_norm": 7.3931732653115985,
+      "learning_rate": 1e-06,
+      "loss": 0.2836,
+      "step": 6991
+    },
+    {
+      "epoch": 0.48344370860927155,
+      "grad_norm": 7.931170346159731,
+      "learning_rate": 1e-06,
+      "loss": 0.2784,
+      "step": 6992
+    },
+    {
+      "epoch": 0.4856512141280353,
+      "grad_norm": 8.06624729932778,
+      "learning_rate": 1e-06,
+      "loss": 0.286,
+      "step": 6993
+    },
+    {
+      "epoch": 0.4878587196467991,
+      "grad_norm": 8.112412901697951,
+      "learning_rate": 1e-06,
+      "loss": 0.2626,
+      "step": 6994
+    },
+    {
+      "epoch": 0.4900662251655629,
+      "grad_norm": 8.466001269978774,
+      "learning_rate": 1e-06,
+      "loss": 0.2507,
+      "step": 6995
+    },
+    {
+      "epoch": 0.4922737306843267,
+      "grad_norm": 6.773886874639259,
+      "learning_rate": 1e-06,
+      "loss": 0.2636,
+      "step": 6996
+    },
+    {
+      "epoch": 0.49448123620309054,
+      "grad_norm": 4.38931017573247,
+      "learning_rate": 1e-06,
+      "loss": 0.26,
+      "step": 6997
+    },
+    {
+      "epoch": 0.4966887417218543,
+      "grad_norm": 8.767566977025595,
+      "learning_rate": 1e-06,
+      "loss": 0.2709,
+      "step": 6998
+    },
+    {
+      "epoch": 0.4988962472406181,
+      "grad_norm": 6.96388092153003,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 6999
+    },
+    {
+      "epoch": 0.5011037527593819,
+      "grad_norm": 7.543200472137415,
+      "learning_rate": 1e-06,
+      "loss": 0.2719,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5011037527593819,
+      "eval_loss": 3.5172221660614014,
+      "eval_runtime": 21.5524,
+      "eval_samples_per_second": 46.399,
+      "eval_steps_per_second": 5.8,
+      "step": 7000
+    },
+    {
+      "epoch": 0.5033112582781457,
+      "grad_norm": 6.403791080204397,
+      "learning_rate": 1e-06,
+      "loss": 0.2535,
+      "step": 7001
+    },
+    {
+      "epoch": 0.5055187637969095,
+      "grad_norm": 7.6281211427098325,
+      "learning_rate": 1e-06,
+      "loss": 0.2859,
+      "step": 7002
+    },
+    {
+      "epoch": 0.5077262693156733,
+      "grad_norm": 6.394174072881324,
+      "learning_rate": 1e-06,
+      "loss": 0.2769,
+      "step": 7003
+    },
+    {
+      "epoch": 0.5099337748344371,
+      "grad_norm": 7.500857016078618,
+      "learning_rate": 1e-06,
+      "loss": 0.2814,
+      "step": 7004
+    },
+    {
+      "epoch": 0.5121412803532008,
+      "grad_norm": 6.504258913702573,
+      "learning_rate": 1e-06,
+      "loss": 0.2882,
+      "step": 7005
+    },
+    {
+      "epoch": 0.5143487858719646,
+      "grad_norm": 6.363826594111154,
+      "learning_rate": 1e-06,
+      "loss": 0.2632,
+      "step": 7006
+    },
+    {
+      "epoch": 0.5165562913907285,
+      "grad_norm": 7.539404505230728,
+      "learning_rate": 1e-06,
+      "loss": 0.2493,
+      "step": 7007
+    },
+    {
+      "epoch": 0.5187637969094923,
+      "grad_norm": 5.252774451119778,
+      "learning_rate": 1e-06,
+      "loss": 0.2604,
+      "step": 7008
+    },
+    {
+      "epoch": 0.5209713024282561,
+      "grad_norm": 6.5722371909722455,
+      "learning_rate": 1e-06,
+      "loss": 0.2638,
+      "step": 7009
+    },
+    {
+      "epoch": 0.5231788079470199,
+      "grad_norm": 8.20678368963692,
+      "learning_rate": 1e-06,
+      "loss": 0.2862,
+      "step": 7010
+    },
+    {
+      "epoch": 0.5253863134657837,
+      "grad_norm": 6.469012484261908,
+      "learning_rate": 1e-06,
+      "loss": 0.2867,
+      "step": 7011
+    },
+    {
+      "epoch": 0.5275938189845475,
+      "grad_norm": 7.007462851087211,
+      "learning_rate": 1e-06,
+      "loss": 0.2523,
+      "step": 7012
+    },
+    {
+      "epoch": 0.5298013245033113,
+      "grad_norm": 7.443770295720464,
+      "learning_rate": 1e-06,
+      "loss": 0.2753,
+      "step": 7013
+    },
+    {
+      "epoch": 0.5320088300220751,
+      "grad_norm": 6.664467813426193,
+      "learning_rate": 1e-06,
+      "loss": 0.2422,
+      "step": 7014
+    },
+    {
+      "epoch": 0.5342163355408388,
+      "grad_norm": 7.744601539654017,
+      "learning_rate": 1e-06,
+      "loss": 0.2746,
+      "step": 7015
+    },
+    {
+      "epoch": 0.5364238410596026,
+      "grad_norm": 7.859120542482086,
+      "learning_rate": 1e-06,
+      "loss": 0.2592,
+      "step": 7016
+    },
+    {
+      "epoch": 0.5386313465783664,
+      "grad_norm": 7.780440564847086,
+      "learning_rate": 1e-06,
+      "loss": 0.2672,
+      "step": 7017
+    },
+    {
+      "epoch": 0.5408388520971302,
+      "grad_norm": 7.770211795573546,
+      "learning_rate": 1e-06,
+      "loss": 0.2698,
+      "step": 7018
+    },
+    {
+      "epoch": 0.543046357615894,
+      "grad_norm": 6.73071865255332,
+      "learning_rate": 1e-06,
+      "loss": 0.254,
+      "step": 7019
+    },
+    {
+      "epoch": 0.5452538631346578,
+      "grad_norm": 8.789165527402169,
+      "learning_rate": 1e-06,
+      "loss": 0.2702,
+      "step": 7020
+    },
+    {
+      "epoch": 0.5474613686534217,
+      "grad_norm": 7.539042518257556,
+      "learning_rate": 1e-06,
+      "loss": 0.2767,
+      "step": 7021
+    },
+    {
+      "epoch": 0.5496688741721855,
+      "grad_norm": 8.048331400403924,
+      "learning_rate": 1e-06,
+      "loss": 0.2879,
+      "step": 7022
+    },
+    {
+      "epoch": 0.5518763796909493,
+      "grad_norm": 6.5227295102500245,
+      "learning_rate": 1e-06,
+      "loss": 0.2644,
+      "step": 7023
+    },
+    {
+      "epoch": 0.5540838852097131,
+      "grad_norm": 6.1009954849378145,
+      "learning_rate": 1e-06,
+      "loss": 0.2496,
+      "step": 7024
+    },
+    {
+      "epoch": 0.5562913907284768,
+      "grad_norm": 7.545827919746044,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 7025
+    },
+    {
+      "epoch": 0.5584988962472406,
+      "grad_norm": 7.90336193487432,
+      "learning_rate": 1e-06,
+      "loss": 0.2708,
+      "step": 7026
+    },
+    {
+      "epoch": 0.5607064017660044,
+      "grad_norm": 7.1001199361615415,
+      "learning_rate": 1e-06,
+      "loss": 0.2491,
+      "step": 7027
+    },
+    {
+      "epoch": 0.5629139072847682,
+      "grad_norm": 8.426750186535036,
+      "learning_rate": 1e-06,
+      "loss": 0.2845,
+      "step": 7028
+    },
+    {
+      "epoch": 0.565121412803532,
+      "grad_norm": 7.32830928985897,
+      "learning_rate": 1e-06,
+      "loss": 0.278,
+      "step": 7029
+    },
+    {
+      "epoch": 0.5673289183222958,
+      "grad_norm": 8.708889143006157,
+      "learning_rate": 1e-06,
+      "loss": 0.2833,
+      "step": 7030
+    },
+    {
+      "epoch": 0.5695364238410596,
+      "grad_norm": 7.431074821056651,
+      "learning_rate": 1e-06,
+      "loss": 0.241,
+      "step": 7031
+    },
+    {
+      "epoch": 0.5717439293598234,
+      "grad_norm": 8.262133514770346,
+      "learning_rate": 1e-06,
+      "loss": 0.2635,
+      "step": 7032
+    },
+    {
+      "epoch": 0.5739514348785872,
+      "grad_norm": 6.104768682370094,
+      "learning_rate": 1e-06,
+      "loss": 0.2672,
+      "step": 7033
+    },
+    {
+      "epoch": 0.5761589403973509,
+      "grad_norm": 7.290487118268256,
+      "learning_rate": 1e-06,
+      "loss": 0.2577,
+      "step": 7034
+    },
+    {
+      "epoch": 0.5783664459161147,
+      "grad_norm": 8.702490363332151,
+      "learning_rate": 1e-06,
+      "loss": 0.2929,
+      "step": 7035
+    },
+    {
+      "epoch": 0.5805739514348786,
+      "grad_norm": 7.962524449049126,
+      "learning_rate": 1e-06,
+      "loss": 0.2755,
+      "step": 7036
+    },
+    {
+      "epoch": 0.5827814569536424,
+      "grad_norm": 6.556995995599669,
+      "learning_rate": 1e-06,
+      "loss": 0.2713,
+      "step": 7037
+    },
+    {
+      "epoch": 0.5849889624724062,
+      "grad_norm": 8.071506849063438,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 7038
+    },
+    {
+      "epoch": 0.58719646799117,
+      "grad_norm": 7.676418979344512,
+      "learning_rate": 1e-06,
+      "loss": 0.2633,
+      "step": 7039
+    },
+    {
+      "epoch": 0.5894039735099338,
+      "grad_norm": 8.086585510676418,
+      "learning_rate": 1e-06,
+      "loss": 0.2729,
+      "step": 7040
+    },
+    {
+      "epoch": 0.5916114790286976,
+      "grad_norm": 7.284832636140395,
+      "learning_rate": 1e-06,
+      "loss": 0.2553,
+      "step": 7041
+    },
+    {
+      "epoch": 0.5938189845474614,
+      "grad_norm": 9.24060407277746,
+      "learning_rate": 1e-06,
+      "loss": 0.2765,
+      "step": 7042
+    },
+    {
+      "epoch": 0.5960264900662252,
+      "grad_norm": 7.276112745949919,
+      "learning_rate": 1e-06,
+      "loss": 0.2914,
+      "step": 7043
+    },
+    {
+      "epoch": 0.5982339955849889,
+      "grad_norm": 10.195053612034194,
+      "learning_rate": 1e-06,
+      "loss": 0.2654,
+      "step": 7044
+    },
+    {
+      "epoch": 0.6004415011037527,
+      "grad_norm": 8.112995984903424,
+      "learning_rate": 1e-06,
+      "loss": 0.2738,
+      "step": 7045
+    },
+    {
+      "epoch": 0.6026490066225165,
+      "grad_norm": 6.549860076474938,
+      "learning_rate": 1e-06,
+      "loss": 0.2483,
+      "step": 7046
+    },
+    {
+      "epoch": 0.6048565121412803,
+      "grad_norm": 6.565131764296074,
+      "learning_rate": 1e-06,
+      "loss": 0.2573,
+      "step": 7047
+    },
+    {
+      "epoch": 0.6070640176600441,
+      "grad_norm": 7.331777229275749,
+      "learning_rate": 1e-06,
+      "loss": 0.2824,
+      "step": 7048
+    },
+    {
+      "epoch": 0.609271523178808,
+      "grad_norm": 8.030696609122032,
+      "learning_rate": 1e-06,
+      "loss": 0.2873,
+      "step": 7049
+    },
+    {
+      "epoch": 0.6114790286975718,
+      "grad_norm": 6.712241055675814,
+      "learning_rate": 1e-06,
+      "loss": 0.2555,
+      "step": 7050
+    },
+    {
+      "epoch": 0.6136865342163356,
+      "grad_norm": 7.2302539975549465,
+      "learning_rate": 1e-06,
+      "loss": 0.2739,
+      "step": 7051
+    },
+    {
+      "epoch": 0.6158940397350994,
+      "grad_norm": 7.185483134378557,
+      "learning_rate": 1e-06,
+      "loss": 0.2861,
+      "step": 7052
+    },
+    {
+      "epoch": 0.6181015452538632,
+      "grad_norm": 6.468242923614848,
+      "learning_rate": 1e-06,
+      "loss": 0.2694,
+      "step": 7053
+    },
+    {
+      "epoch": 0.6203090507726269,
+      "grad_norm": 6.141197428176475,
+      "learning_rate": 1e-06,
+      "loss": 0.2678,
+      "step": 7054
+    },
+    {
+      "epoch": 0.6225165562913907,
+      "grad_norm": 6.6031192609125915,
+      "learning_rate": 1e-06,
+      "loss": 0.2602,
+      "step": 7055
+    },
+    {
+      "epoch": 0.6247240618101545,
+      "grad_norm": 6.038808243748644,
+      "learning_rate": 1e-06,
+      "loss": 0.2507,
+      "step": 7056
+    },
+    {
+      "epoch": 0.6269315673289183,
+      "grad_norm": 6.624806352116815,
+      "learning_rate": 1e-06,
+      "loss": 0.2571,
+      "step": 7057
+    },
+    {
+      "epoch": 0.6291390728476821,
+      "grad_norm": 6.125919308726288,
+      "learning_rate": 1e-06,
+      "loss": 0.2921,
+      "step": 7058
+    },
+    {
+      "epoch": 0.6313465783664459,
+      "grad_norm": 6.88612645185359,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7059
+    },
+    {
+      "epoch": 0.6335540838852097,
+      "grad_norm": 7.270104129945196,
+      "learning_rate": 1e-06,
+      "loss": 0.2823,
+      "step": 7060
+    },
+    {
+      "epoch": 0.6357615894039735,
+      "grad_norm": 6.846691497626053,
+      "learning_rate": 1e-06,
+      "loss": 0.2759,
+      "step": 7061
+    },
+    {
+      "epoch": 0.6379690949227373,
+      "grad_norm": 6.275933766193909,
+      "learning_rate": 1e-06,
+      "loss": 0.2618,
+      "step": 7062
+    },
+    {
+      "epoch": 0.6401766004415012,
+      "grad_norm": 6.08874680409467,
+      "learning_rate": 1e-06,
+      "loss": 0.2539,
+      "step": 7063
+    },
+    {
+      "epoch": 0.6423841059602649,
+      "grad_norm": 6.587200887653168,
+      "learning_rate": 1e-06,
+      "loss": 0.2534,
+      "step": 7064
+    },
+    {
+      "epoch": 0.6445916114790287,
+      "grad_norm": 9.3742359414965,
+      "learning_rate": 1e-06,
+      "loss": 0.2794,
+      "step": 7065
+    },
+    {
+      "epoch": 0.6467991169977925,
+      "grad_norm": 5.833379370473224,
+      "learning_rate": 1e-06,
+      "loss": 0.2703,
+      "step": 7066
+    },
+    {
+      "epoch": 0.6490066225165563,
+      "grad_norm": 6.174069128952458,
+      "learning_rate": 1e-06,
+      "loss": 0.2691,
+      "step": 7067
+    },
+    {
+      "epoch": 0.6512141280353201,
+      "grad_norm": 7.382048085053144,
+      "learning_rate": 1e-06,
+      "loss": 0.2581,
+      "step": 7068
+    },
+    {
+      "epoch": 0.6534216335540839,
+      "grad_norm": 7.1683250916366,
+      "learning_rate": 1e-06,
+      "loss": 0.2655,
+      "step": 7069
+    },
+    {
+      "epoch": 0.6556291390728477,
+      "grad_norm": 9.44642539997161,
+      "learning_rate": 1e-06,
+      "loss": 0.2734,
+      "step": 7070
+    },
+    {
+      "epoch": 0.6578366445916115,
+      "grad_norm": 7.365733019690641,
+      "learning_rate": 1e-06,
+      "loss": 0.2425,
+      "step": 7071
+    },
+    {
+      "epoch": 0.6600441501103753,
+      "grad_norm": 7.476410038102276,
+      "learning_rate": 1e-06,
+      "loss": 0.2572,
+      "step": 7072
+    },
+    {
+      "epoch": 0.6622516556291391,
+      "grad_norm": 6.705183412774674,
+      "learning_rate": 1e-06,
+      "loss": 0.277,
+      "step": 7073
+    },
+    {
+      "epoch": 0.6644591611479028,
+      "grad_norm": 7.77582633473068,
+      "learning_rate": 1e-06,
+      "loss": 0.2576,
+      "step": 7074
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 8.12925849797069,
+      "learning_rate": 1e-06,
+      "loss": 0.2871,
+      "step": 7075
+    },
+    {
+      "epoch": 0.6688741721854304,
+      "grad_norm": 6.238224529811016,
+      "learning_rate": 1e-06,
+      "loss": 0.2604,
+      "step": 7076
+    },
+    {
+      "epoch": 0.6710816777041942,
+      "grad_norm": 5.965573938009095,
+      "learning_rate": 1e-06,
+      "loss": 0.261,
+      "step": 7077
+    },
+    {
+      "epoch": 0.673289183222958,
+      "grad_norm": 7.861589860024249,
+      "learning_rate": 1e-06,
+      "loss": 0.2771,
+      "step": 7078
+    },
+    {
+      "epoch": 0.6754966887417219,
+      "grad_norm": 6.589372026515254,
+      "learning_rate": 1e-06,
+      "loss": 0.2843,
+      "step": 7079
+    },
+    {
+      "epoch": 0.6777041942604857,
+      "grad_norm": 5.634892459674078,
+      "learning_rate": 1e-06,
+      "loss": 0.264,
+      "step": 7080
+    },
+    {
+      "epoch": 0.6799116997792495,
+      "grad_norm": 8.310193402427826,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 7081
+    },
+    {
+      "epoch": 0.6821192052980133,
+      "grad_norm": 7.262413787501925,
+      "learning_rate": 1e-06,
+      "loss": 0.2661,
+      "step": 7082
+    },
+    {
+      "epoch": 0.6843267108167771,
+      "grad_norm": 5.896792326167022,
+      "learning_rate": 1e-06,
+      "loss": 0.2742,
+      "step": 7083
+    },
+    {
+      "epoch": 0.6865342163355408,
+      "grad_norm": 7.113252367252198,
+      "learning_rate": 1e-06,
+      "loss": 0.2471,
+      "step": 7084
+    },
+    {
+      "epoch": 0.6887417218543046,
+      "grad_norm": 7.718948299044285,
+      "learning_rate": 1e-06,
+      "loss": 0.2767,
+      "step": 7085
+    },
+    {
+      "epoch": 0.6909492273730684,
+      "grad_norm": 5.925414520092801,
+      "learning_rate": 1e-06,
+      "loss": 0.2651,
+      "step": 7086
+    },
+    {
+      "epoch": 0.6931567328918322,
+      "grad_norm": 5.56178488737482,
+      "learning_rate": 1e-06,
+      "loss": 0.2691,
+      "step": 7087
+    },
+    {
+      "epoch": 0.695364238410596,
+      "grad_norm": 5.83432173112729,
+      "learning_rate": 1e-06,
+      "loss": 0.2773,
+      "step": 7088
+    },
+    {
+      "epoch": 0.6975717439293598,
+      "grad_norm": 6.704315126630016,
+      "learning_rate": 1e-06,
+      "loss": 0.2725,
+      "step": 7089
+    },
+    {
+      "epoch": 0.6997792494481236,
+      "grad_norm": 7.662724041879456,
+      "learning_rate": 1e-06,
+      "loss": 0.268,
+      "step": 7090
+    },
+    {
+      "epoch": 0.7019867549668874,
+      "grad_norm": 6.1477208903475065,
+      "learning_rate": 1e-06,
+      "loss": 0.2766,
+      "step": 7091
+    },
+    {
+      "epoch": 0.7041942604856513,
+      "grad_norm": 5.611293618451789,
+      "learning_rate": 1e-06,
+      "loss": 0.2689,
+      "step": 7092
+    },
+    {
+      "epoch": 0.7064017660044151,
+      "grad_norm": 8.066828033031063,
+      "learning_rate": 1e-06,
+      "loss": 0.2735,
+      "step": 7093
+    },
+    {
+      "epoch": 0.7086092715231788,
+      "grad_norm": 6.972407780397189,
+      "learning_rate": 1e-06,
+      "loss": 0.2616,
+      "step": 7094
+    },
+    {
+      "epoch": 0.7108167770419426,
+      "grad_norm": 6.752724488630619,
+      "learning_rate": 1e-06,
+      "loss": 0.2755,
+      "step": 7095
+    },
+    {
+      "epoch": 0.7130242825607064,
+      "grad_norm": 7.718632325960298,
+      "learning_rate": 1e-06,
+      "loss": 0.254,
+      "step": 7096
+    },
+    {
+      "epoch": 0.7152317880794702,
+      "grad_norm": 5.839187969072113,
+      "learning_rate": 1e-06,
+      "loss": 0.251,
+      "step": 7097
+    },
+    {
+      "epoch": 0.717439293598234,
+      "grad_norm": 7.548607671465412,
+      "learning_rate": 1e-06,
+      "loss": 0.2786,
+      "step": 7098
+    },
+    {
+      "epoch": 0.7196467991169978,
+      "grad_norm": 5.923088505259694,
+      "learning_rate": 1e-06,
+      "loss": 0.262,
+      "step": 7099
+    },
+    {
+      "epoch": 0.7218543046357616,
+      "grad_norm": 5.698598945638354,
+      "learning_rate": 1e-06,
+      "loss": 0.2433,
+      "step": 7100
+    },
+    {
+      "epoch": 0.7240618101545254,
+      "grad_norm": 7.385831181214352,
+      "learning_rate": 1e-06,
+      "loss": 0.2874,
+      "step": 7101
+    },
+    {
+      "epoch": 0.7262693156732892,
+      "grad_norm": 9.509985238562832,
+      "learning_rate": 1e-06,
+      "loss": 0.2901,
+      "step": 7102
+    },
+    {
+      "epoch": 0.7284768211920529,
+      "grad_norm": 6.508992460471817,
+      "learning_rate": 1e-06,
+      "loss": 0.2882,
+      "step": 7103
+    },
+    {
+      "epoch": 0.7306843267108167,
+      "grad_norm": 5.077625273355657,
+      "learning_rate": 1e-06,
+      "loss": 0.2746,
+      "step": 7104
+    },
+    {
+      "epoch": 0.7328918322295805,
+      "grad_norm": 7.652617634298475,
+      "learning_rate": 1e-06,
+      "loss": 0.2634,
+      "step": 7105
+    },
+    {
+      "epoch": 0.7350993377483444,
+      "grad_norm": 9.229668856725953,
+      "learning_rate": 1e-06,
+      "loss": 0.283,
+      "step": 7106
+    },
+    {
+      "epoch": 0.7373068432671082,
+      "grad_norm": 7.468674792587929,
+      "learning_rate": 1e-06,
+      "loss": 0.262,
+      "step": 7107
+    },
+    {
+      "epoch": 0.739514348785872,
+      "grad_norm": 7.891804724975607,
+      "learning_rate": 1e-06,
+      "loss": 0.2682,
+      "step": 7108
+    },
+    {
+      "epoch": 0.7417218543046358,
+      "grad_norm": 7.104492224499942,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7109
+    },
+    {
+      "epoch": 0.7439293598233996,
+      "grad_norm": 6.2875338706988275,
+      "learning_rate": 1e-06,
+      "loss": 0.2761,
+      "step": 7110
+    },
+    {
+      "epoch": 0.7461368653421634,
+      "grad_norm": 6.276070701615142,
+      "learning_rate": 1e-06,
+      "loss": 0.2711,
+      "step": 7111
+    },
+    {
+      "epoch": 0.7483443708609272,
+      "grad_norm": 8.93517509299343,
+      "learning_rate": 1e-06,
+      "loss": 0.2553,
+      "step": 7112
+    },
+    {
+      "epoch": 0.7505518763796909,
+      "grad_norm": 8.163098577753606,
+      "learning_rate": 1e-06,
+      "loss": 0.2686,
+      "step": 7113
+    },
+    {
+      "epoch": 0.7527593818984547,
+      "grad_norm": 6.149940811913736,
+      "learning_rate": 1e-06,
+      "loss": 0.2721,
+      "step": 7114
+    },
+    {
+      "epoch": 0.7549668874172185,
+      "grad_norm": 7.898877390781494,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 7115
+    },
+    {
+      "epoch": 0.7571743929359823,
+      "grad_norm": 8.083661309749807,
+      "learning_rate": 1e-06,
+      "loss": 0.262,
+      "step": 7116
+    },
+    {
+      "epoch": 0.7593818984547461,
+      "grad_norm": 7.422422722460569,
+      "learning_rate": 1e-06,
+      "loss": 0.2664,
+      "step": 7117
+    },
+    {
+      "epoch": 0.7615894039735099,
+      "grad_norm": 7.421406926904935,
+      "learning_rate": 1e-06,
+      "loss": 0.2709,
+      "step": 7118
+    },
+    {
+      "epoch": 0.7637969094922737,
+      "grad_norm": 7.106697372968231,
+      "learning_rate": 1e-06,
+      "loss": 0.268,
+      "step": 7119
+    },
+    {
+      "epoch": 0.7660044150110376,
+      "grad_norm": 7.702316005395933,
+      "learning_rate": 1e-06,
+      "loss": 0.2924,
+      "step": 7120
+    },
+    {
+      "epoch": 0.7682119205298014,
+      "grad_norm": 7.823554312547061,
+      "learning_rate": 1e-06,
+      "loss": 0.2816,
+      "step": 7121
+    },
+    {
+      "epoch": 0.7704194260485652,
+      "grad_norm": 9.123702562185601,
+      "learning_rate": 1e-06,
+      "loss": 0.278,
+      "step": 7122
+    },
+    {
+      "epoch": 0.7726269315673289,
+      "grad_norm": 6.994362585679771,
+      "learning_rate": 1e-06,
+      "loss": 0.2629,
+      "step": 7123
+    },
+    {
+      "epoch": 0.7748344370860927,
+      "grad_norm": 8.916440106791086,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7124
+    },
+    {
+      "epoch": 0.7770419426048565,
+      "grad_norm": 7.243380713046636,
+      "learning_rate": 1e-06,
+      "loss": 0.2964,
+      "step": 7125
+    },
+    {
+      "epoch": 0.7792494481236203,
+      "grad_norm": 8.152934703065036,
+      "learning_rate": 1e-06,
+      "loss": 0.2531,
+      "step": 7126
+    },
+    {
+      "epoch": 0.7814569536423841,
+      "grad_norm": 8.340805370472408,
+      "learning_rate": 1e-06,
+      "loss": 0.286,
+      "step": 7127
+    },
+    {
+      "epoch": 0.7836644591611479,
+      "grad_norm": 8.807740870756849,
+      "learning_rate": 1e-06,
+      "loss": 0.2882,
+      "step": 7128
+    },
+    {
+      "epoch": 0.7858719646799117,
+      "grad_norm": 6.733203750026787,
+      "learning_rate": 1e-06,
+      "loss": 0.2452,
+      "step": 7129
+    },
+    {
+      "epoch": 0.7880794701986755,
+      "grad_norm": 6.673380880612749,
+      "learning_rate": 1e-06,
+      "loss": 0.2778,
+      "step": 7130
+    },
+    {
+      "epoch": 0.7902869757174393,
+      "grad_norm": 5.810371286239545,
+      "learning_rate": 1e-06,
+      "loss": 0.257,
+      "step": 7131
+    },
+    {
+      "epoch": 0.7924944812362031,
+      "grad_norm": 8.33341852859739,
+      "learning_rate": 1e-06,
+      "loss": 0.2872,
+      "step": 7132
+    },
+    {
+      "epoch": 0.7947019867549668,
+      "grad_norm": 7.065317457397844,
+      "learning_rate": 1e-06,
+      "loss": 0.2655,
+      "step": 7133
+    },
+    {
+      "epoch": 0.7969094922737306,
+      "grad_norm": 6.939900052304305,
+      "learning_rate": 1e-06,
+      "loss": 0.2635,
+      "step": 7134
+    },
+    {
+      "epoch": 0.7991169977924945,
+      "grad_norm": 6.78191131494407,
+      "learning_rate": 1e-06,
+      "loss": 0.2583,
+      "step": 7135
+    },
+    {
+      "epoch": 0.8013245033112583,
+      "grad_norm": 9.781170469311025,
+      "learning_rate": 1e-06,
+      "loss": 0.2966,
+      "step": 7136
+    },
+    {
+      "epoch": 0.8035320088300221,
+      "grad_norm": 6.603762773546745,
+      "learning_rate": 1e-06,
+      "loss": 0.2541,
+      "step": 7137
+    },
+    {
+      "epoch": 0.8057395143487859,
+      "grad_norm": 6.366890099518922,
+      "learning_rate": 1e-06,
+      "loss": 0.2705,
+      "step": 7138
+    },
+    {
+      "epoch": 0.8079470198675497,
+      "grad_norm": 6.499002052607989,
+      "learning_rate": 1e-06,
+      "loss": 0.2705,
+      "step": 7139
+    },
+    {
+      "epoch": 0.8101545253863135,
+      "grad_norm": 5.793392889093349,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7140
+    },
+    {
+      "epoch": 0.8123620309050773,
+      "grad_norm": 6.810467199943779,
+      "learning_rate": 1e-06,
+      "loss": 0.2578,
+      "step": 7141
+    },
+    {
+      "epoch": 0.8145695364238411,
+      "grad_norm": 9.569872417318356,
+      "learning_rate": 1e-06,
+      "loss": 0.269,
+      "step": 7142
+    },
+    {
+      "epoch": 0.8167770419426048,
+      "grad_norm": 5.609092139192416,
+      "learning_rate": 1e-06,
+      "loss": 0.2606,
+      "step": 7143
+    },
+    {
+      "epoch": 0.8189845474613686,
+      "grad_norm": 8.743904897967226,
+      "learning_rate": 1e-06,
+      "loss": 0.2476,
+      "step": 7144
+    },
+    {
+      "epoch": 0.8211920529801324,
+      "grad_norm": 5.494888399433796,
+      "learning_rate": 1e-06,
+      "loss": 0.27,
+      "step": 7145
+    },
+    {
+      "epoch": 0.8233995584988962,
+      "grad_norm": 6.047104628367362,
+      "learning_rate": 1e-06,
+      "loss": 0.2687,
+      "step": 7146
+    },
+    {
+      "epoch": 0.82560706401766,
+      "grad_norm": 7.294623355019005,
+      "learning_rate": 1e-06,
+      "loss": 0.2607,
+      "step": 7147
+    },
+    {
+      "epoch": 0.8278145695364238,
+      "grad_norm": 8.518981912111737,
+      "learning_rate": 1e-06,
+      "loss": 0.2662,
+      "step": 7148
+    },
+    {
+      "epoch": 0.8300220750551877,
+      "grad_norm": 8.8704700575649,
+      "learning_rate": 1e-06,
+      "loss": 0.2754,
+      "step": 7149
+    },
+    {
+      "epoch": 0.8322295805739515,
+      "grad_norm": 9.185549489834989,
+      "learning_rate": 1e-06,
+      "loss": 0.2732,
+      "step": 7150
+    },
+    {
+      "epoch": 0.8344370860927153,
+      "grad_norm": 9.285487958861859,
+      "learning_rate": 1e-06,
+      "loss": 0.2653,
+      "step": 7151
+    },
+    {
+      "epoch": 0.8366445916114791,
+      "grad_norm": 7.861166785810314,
+      "learning_rate": 1e-06,
+      "loss": 0.2545,
+      "step": 7152
+    },
+    {
+      "epoch": 0.8388520971302428,
+      "grad_norm": 7.940772186230171,
+      "learning_rate": 1e-06,
+      "loss": 0.2672,
+      "step": 7153
+    },
+    {
+      "epoch": 0.8410596026490066,
+      "grad_norm": 7.146557055848447,
+      "learning_rate": 1e-06,
+      "loss": 0.281,
+      "step": 7154
+    },
+    {
+      "epoch": 0.8432671081677704,
+      "grad_norm": 7.505048832192539,
+      "learning_rate": 1e-06,
+      "loss": 0.2739,
+      "step": 7155
+    },
+    {
+      "epoch": 0.8454746136865342,
+      "grad_norm": 6.525963360967428,
+      "learning_rate": 1e-06,
+      "loss": 0.2639,
+      "step": 7156
+    },
+    {
+      "epoch": 0.847682119205298,
+      "grad_norm": 7.7585390752346415,
+      "learning_rate": 1e-06,
+      "loss": 0.2559,
+      "step": 7157
+    },
+    {
+      "epoch": 0.8498896247240618,
+      "grad_norm": 7.398244874241306,
+      "learning_rate": 1e-06,
+      "loss": 0.2608,
+      "step": 7158
+    },
+    {
+      "epoch": 0.8520971302428256,
+      "grad_norm": 8.577393524565204,
+      "learning_rate": 1e-06,
+      "loss": 0.2599,
+      "step": 7159
+    },
+    {
+      "epoch": 0.8543046357615894,
+      "grad_norm": 7.853459981695362,
+      "learning_rate": 1e-06,
+      "loss": 0.277,
+      "step": 7160
+    },
+    {
+      "epoch": 0.8565121412803532,
+      "grad_norm": 6.769189269110699,
+      "learning_rate": 1e-06,
+      "loss": 0.2624,
+      "step": 7161
+    },
+    {
+      "epoch": 0.8587196467991169,
+      "grad_norm": 7.267365891498084,
+      "learning_rate": 1e-06,
+      "loss": 0.2546,
+      "step": 7162
+    },
+    {
+      "epoch": 0.8609271523178808,
+      "grad_norm": 7.562178094539485,
+      "learning_rate": 1e-06,
+      "loss": 0.2719,
+      "step": 7163
+    },
+    {
+      "epoch": 0.8631346578366446,
+      "grad_norm": 7.679952372410113,
+      "learning_rate": 1e-06,
+      "loss": 0.2572,
+      "step": 7164
+    },
+    {
+      "epoch": 0.8653421633554084,
+      "grad_norm": 7.855612338203512,
+      "learning_rate": 1e-06,
+      "loss": 0.2824,
+      "step": 7165
+    },
+    {
+      "epoch": 0.8675496688741722,
+      "grad_norm": 8.124900360288805,
+      "learning_rate": 1e-06,
+      "loss": 0.2611,
+      "step": 7166
+    },
+    {
+      "epoch": 0.869757174392936,
+      "grad_norm": 6.439388897800479,
+      "learning_rate": 1e-06,
+      "loss": 0.2663,
+      "step": 7167
+    },
+    {
+      "epoch": 0.8719646799116998,
+      "grad_norm": 5.570368733312825,
+      "learning_rate": 1e-06,
+      "loss": 0.2434,
+      "step": 7168
+    },
+    {
+      "epoch": 0.8741721854304636,
+      "grad_norm": 8.024277335200189,
+      "learning_rate": 1e-06,
+      "loss": 0.274,
+      "step": 7169
+    },
+    {
+      "epoch": 0.8763796909492274,
+      "grad_norm": 6.887400197369902,
+      "learning_rate": 1e-06,
+      "loss": 0.2517,
+      "step": 7170
+    },
+    {
+      "epoch": 0.8785871964679912,
+      "grad_norm": 7.775069689002338,
+      "learning_rate": 1e-06,
+      "loss": 0.2562,
+      "step": 7171
+    },
+    {
+      "epoch": 0.8807947019867549,
+      "grad_norm": 7.46442221521989,
+      "learning_rate": 1e-06,
+      "loss": 0.264,
+      "step": 7172
+    },
+    {
+      "epoch": 0.8830022075055187,
+      "grad_norm": 6.900050142814722,
+      "learning_rate": 1e-06,
+      "loss": 0.252,
+      "step": 7173
+    },
+    {
+      "epoch": 0.8852097130242825,
+      "grad_norm": 7.206517201307857,
+      "learning_rate": 1e-06,
+      "loss": 0.2463,
+      "step": 7174
+    },
+    {
+      "epoch": 0.8874172185430463,
+      "grad_norm": 7.8553121539738076,
+      "learning_rate": 1e-06,
+      "loss": 0.2601,
+      "step": 7175
+    },
+    {
+      "epoch": 0.8896247240618101,
+      "grad_norm": 7.048794237192395,
+      "learning_rate": 1e-06,
+      "loss": 0.245,
+      "step": 7176
+    },
+    {
+      "epoch": 0.891832229580574,
+      "grad_norm": 6.227694856493409,
+      "learning_rate": 1e-06,
+      "loss": 0.2529,
+      "step": 7177
+    },
+    {
+      "epoch": 0.8940397350993378,
+      "grad_norm": 7.033142905740049,
+      "learning_rate": 1e-06,
+      "loss": 0.2418,
+      "step": 7178
+    },
+    {
+      "epoch": 0.8962472406181016,
+      "grad_norm": 8.395554359057298,
+      "learning_rate": 1e-06,
+      "loss": 0.2598,
+      "step": 7179
+    },
+    {
+      "epoch": 0.8984547461368654,
+      "grad_norm": 7.317295661972007,
+      "learning_rate": 1e-06,
+      "loss": 0.2678,
+      "step": 7180
+    },
+    {
+      "epoch": 0.9006622516556292,
+      "grad_norm": 7.984978188862027,
+      "learning_rate": 1e-06,
+      "loss": 0.2684,
+      "step": 7181
+    },
+    {
+      "epoch": 0.9028697571743929,
+      "grad_norm": 6.8059223322826545,
+      "learning_rate": 1e-06,
+      "loss": 0.2823,
+      "step": 7182
+    },
+    {
+      "epoch": 0.9050772626931567,
+      "grad_norm": 8.499337678870102,
+      "learning_rate": 1e-06,
+      "loss": 0.2643,
+      "step": 7183
+    },
+    {
+      "epoch": 0.9072847682119205,
+      "grad_norm": 7.733132473862889,
+      "learning_rate": 1e-06,
+      "loss": 0.2811,
+      "step": 7184
+    },
+    {
+      "epoch": 0.9094922737306843,
+      "grad_norm": 6.720699160484485,
+      "learning_rate": 1e-06,
+      "loss": 0.266,
+      "step": 7185
+    },
+    {
+      "epoch": 0.9116997792494481,
+      "grad_norm": 6.89081465921751,
+      "learning_rate": 1e-06,
+      "loss": 0.2675,
+      "step": 7186
+    },
+    {
+      "epoch": 0.9139072847682119,
+      "grad_norm": 10.042100102946693,
+      "learning_rate": 1e-06,
+      "loss": 0.2669,
+      "step": 7187
+    },
+    {
+      "epoch": 0.9161147902869757,
+      "grad_norm": 7.597887736066307,
+      "learning_rate": 1e-06,
+      "loss": 0.2792,
+      "step": 7188
+    },
+    {
+      "epoch": 0.9183222958057395,
+      "grad_norm": 6.464609922359196,
+      "learning_rate": 1e-06,
+      "loss": 0.2528,
+      "step": 7189
+    },
+    {
+      "epoch": 0.9205298013245033,
+      "grad_norm": 8.13742759417398,
+      "learning_rate": 1e-06,
+      "loss": 0.2704,
+      "step": 7190
+    },
+    {
+      "epoch": 0.9227373068432672,
+      "grad_norm": 5.444344314602962,
+      "learning_rate": 1e-06,
+      "loss": 0.2573,
+      "step": 7191
+    },
+    {
+      "epoch": 0.9249448123620309,
+      "grad_norm": 5.287757597514947,
+      "learning_rate": 1e-06,
+      "loss": 0.2864,
+      "step": 7192
+    },
+    {
+      "epoch": 0.9271523178807947,
+      "grad_norm": 7.558841702056245,
+      "learning_rate": 1e-06,
+      "loss": 0.2742,
+      "step": 7193
+    },
+    {
+      "epoch": 0.9293598233995585,
+      "grad_norm": 6.4847930978438715,
+      "learning_rate": 1e-06,
+      "loss": 0.269,
+      "step": 7194
+    },
+    {
+      "epoch": 0.9315673289183223,
+      "grad_norm": 5.6501248222908,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7195
+    },
+    {
+      "epoch": 0.9337748344370861,
+      "grad_norm": 5.885897336108135,
+      "learning_rate": 1e-06,
+      "loss": 0.2446,
+      "step": 7196
+    },
+    {
+      "epoch": 0.9359823399558499,
+      "grad_norm": 7.916879947923287,
+      "learning_rate": 1e-06,
+      "loss": 0.2826,
+      "step": 7197
+    },
+    {
+      "epoch": 0.9381898454746137,
+      "grad_norm": 7.992787010775927,
+      "learning_rate": 1e-06,
+      "loss": 0.2482,
+      "step": 7198
+    },
+    {
+      "epoch": 0.9403973509933775,
+      "grad_norm": 6.289762891025908,
+      "learning_rate": 1e-06,
+      "loss": 0.2624,
+      "step": 7199
+    },
+    {
+      "epoch": 0.9426048565121413,
+      "grad_norm": 7.281118435787698,
+      "learning_rate": 1e-06,
+      "loss": 0.2618,
+      "step": 7200
+    },
+    {
+      "epoch": 0.9448123620309051,
+      "grad_norm": 6.221511765430705,
+      "learning_rate": 1e-06,
+      "loss": 0.2466,
+      "step": 7201
+    },
+    {
+      "epoch": 0.9470198675496688,
+      "grad_norm": 7.7314010426777795,
+      "learning_rate": 1e-06,
+      "loss": 0.2619,
+      "step": 7202
+    },
+    {
+      "epoch": 0.9492273730684326,
+      "grad_norm": 6.004317864229244,
+      "learning_rate": 1e-06,
+      "loss": 0.2573,
+      "step": 7203
+    },
+    {
+      "epoch": 0.9514348785871964,
+      "grad_norm": 6.026628099325086,
+      "learning_rate": 1e-06,
+      "loss": 0.2436,
+      "step": 7204
+    },
+    {
+      "epoch": 0.9536423841059603,
+      "grad_norm": 5.3987954667707,
+      "learning_rate": 1e-06,
+      "loss": 0.2647,
+      "step": 7205
+    },
+    {
+      "epoch": 0.9558498896247241,
+      "grad_norm": 7.953453822220395,
+      "learning_rate": 1e-06,
+      "loss": 0.2804,
+      "step": 7206
+    },
+    {
+      "epoch": 0.9580573951434879,
+      "grad_norm": 8.912661720945433,
+      "learning_rate": 1e-06,
+      "loss": 0.2749,
+      "step": 7207
+    },
+    {
+      "epoch": 0.9602649006622517,
+      "grad_norm": 6.9265078098898485,
+      "learning_rate": 1e-06,
+      "loss": 0.2659,
+      "step": 7208
+    },
+    {
+      "epoch": 0.9624724061810155,
+      "grad_norm": 6.8122890599207375,
+      "learning_rate": 1e-06,
+      "loss": 0.2857,
+      "step": 7209
+    },
+    {
+      "epoch": 0.9646799116997793,
+      "grad_norm": 7.144911460796272,
+      "learning_rate": 1e-06,
+      "loss": 0.2663,
+      "step": 7210
+    },
+    {
+      "epoch": 0.9668874172185431,
+      "grad_norm": 5.581105808740851,
+      "learning_rate": 1e-06,
+      "loss": 0.2656,
+      "step": 7211
+    },
+    {
+      "epoch": 0.9690949227373068,
+      "grad_norm": 7.259078186552697,
+      "learning_rate": 1e-06,
+      "loss": 0.2382,
+      "step": 7212
+    },
+    {
+      "epoch": 0.9713024282560706,
+      "grad_norm": 7.643379663009667,
+      "learning_rate": 1e-06,
+      "loss": 0.2676,
+      "step": 7213
+    },
+    {
+      "epoch": 0.9735099337748344,
+      "grad_norm": 7.470746207346763,
+      "learning_rate": 1e-06,
+      "loss": 0.281,
+      "step": 7214
+    },
+    {
+      "epoch": 0.9757174392935982,
+      "grad_norm": 7.367549189962075,
+      "learning_rate": 1e-06,
+      "loss": 0.2731,
+      "step": 7215
+    },
+    {
+      "epoch": 0.977924944812362,
+      "grad_norm": 8.351256023657994,
+      "learning_rate": 1e-06,
+      "loss": 0.2521,
+      "step": 7216
+    },
+    {
+      "epoch": 0.9801324503311258,
+      "grad_norm": 8.539665161112282,
+      "learning_rate": 1e-06,
+      "loss": 0.2852,
+      "step": 7217
+    },
+    {
+      "epoch": 0.9823399558498896,
+      "grad_norm": 7.457267430250184,
+      "learning_rate": 1e-06,
+      "loss": 0.259,
+      "step": 7218
+    },
+    {
+      "epoch": 0.9845474613686535,
+      "grad_norm": 6.947071583751046,
+      "learning_rate": 1e-06,
+      "loss": 0.2621,
+      "step": 7219
+    },
+    {
+      "epoch": 0.9867549668874173,
+      "grad_norm": 8.606105759424453,
+      "learning_rate": 1e-06,
+      "loss": 0.2694,
+      "step": 7220
+    },
+    {
+      "epoch": 0.9889624724061811,
+      "grad_norm": 8.531367409357467,
+      "learning_rate": 1e-06,
+      "loss": 0.27,
+      "step": 7221
+    },
+    {
+      "epoch": 0.9911699779249448,
+      "grad_norm": 8.997347065671349,
+      "learning_rate": 1e-06,
+      "loss": 0.2831,
+      "step": 7222
+    },
+    {
+      "epoch": 0.9933774834437086,
+      "grad_norm": 7.503155044739067,
+      "learning_rate": 1e-06,
+      "loss": 0.269,
+      "step": 7223
+    },
+    {
+      "epoch": 0.9955849889624724,
+      "grad_norm": 10.06307618147763,
+      "learning_rate": 1e-06,
+      "loss": 0.2834,
+      "step": 7224
+    },
+    {
+      "epoch": 0.9977924944812362,
+      "grad_norm": 5.1456178044620895,
+      "learning_rate": 1e-06,
+      "loss": 0.2585,
+      "step": 7225
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 5.665498374349392,
+      "learning_rate": 1e-06,
+      "loss": 0.2568,
+      "step": 7226
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.567829132080078,
+      "eval_runtime": 21.5751,
+      "eval_samples_per_second": 46.35,
+      "eval_steps_per_second": 5.794,
+      "step": 7226
+    },
+    {
+      "epoch": 0.0021929824561403508,
+      "grad_norm": 8.399354578704186,
+      "learning_rate": 1e-06,
+      "loss": 0.3665,
+      "step": 7227
+    },
+    {
+      "epoch": 0.0043859649122807015,
+      "grad_norm": 6.391744454937992,
+      "learning_rate": 1e-06,
+      "loss": 0.3374,
+      "step": 7228
+    },
+    {
+      "epoch": 0.006578947368421052,
+      "grad_norm": 7.422752468148163,
+      "learning_rate": 1e-06,
+      "loss": 0.3737,
+      "step": 7229
+    },
+    {
+      "epoch": 0.008771929824561403,
+      "grad_norm": 8.47096121555006,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 7230
+    },
+    {
+      "epoch": 0.010964912280701754,
+      "grad_norm": 5.979250122813078,
+      "learning_rate": 1e-06,
+      "loss": 0.3416,
+      "step": 7231
+    },
+    {
+      "epoch": 0.013157894736842105,
+      "grad_norm": 7.206493568845566,
+      "learning_rate": 1e-06,
+      "loss": 0.3633,
+      "step": 7232
+    },
+    {
+      "epoch": 0.015350877192982455,
+      "grad_norm": 3.3626354637471136,
+      "learning_rate": 1e-06,
+      "loss": 0.3714,
+      "step": 7233
+    },
+    {
+      "epoch": 0.017543859649122806,
+      "grad_norm": 6.671073878511276,
+      "learning_rate": 1e-06,
+      "loss": 0.3742,
+      "step": 7234
+    },
+    {
+      "epoch": 0.019736842105263157,
+      "grad_norm": 5.847703979319459,
+      "learning_rate": 1e-06,
+      "loss": 0.3534,
+      "step": 7235
+    },
+    {
+      "epoch": 0.021929824561403508,
+      "grad_norm": 6.697987946870648,
+      "learning_rate": 1e-06,
+      "loss": 0.3739,
+      "step": 7236
+    },
+    {
+      "epoch": 0.02412280701754386,
+      "grad_norm": 6.89001326227006,
+      "learning_rate": 1e-06,
+      "loss": 0.3873,
+      "step": 7237
+    },
+    {
+      "epoch": 0.02631578947368421,
+      "grad_norm": 8.146204084301917,
+      "learning_rate": 1e-06,
+      "loss": 0.3686,
+      "step": 7238
+    },
+    {
+      "epoch": 0.02850877192982456,
+      "grad_norm": 6.377751930036124,
+      "learning_rate": 1e-06,
+      "loss": 0.3641,
+      "step": 7239
+    },
+    {
+      "epoch": 0.03070175438596491,
+      "grad_norm": 7.4243081778330495,
+      "learning_rate": 1e-06,
+      "loss": 0.4161,
+      "step": 7240
+    },
+    {
+      "epoch": 0.03289473684210526,
+      "grad_norm": 7.74178317019827,
+      "learning_rate": 1e-06,
+      "loss": 0.3506,
+      "step": 7241
+    },
+    {
+      "epoch": 0.03508771929824561,
+      "grad_norm": 5.08511285205115,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 7242
+    },
+    {
+      "epoch": 0.03728070175438596,
+      "grad_norm": 8.608476631305187,
+      "learning_rate": 1e-06,
+      "loss": 0.3388,
+      "step": 7243
+    },
+    {
+      "epoch": 0.039473684210526314,
+      "grad_norm": 7.189982415116332,
+      "learning_rate": 1e-06,
+      "loss": 0.3551,
+      "step": 7244
+    },
+    {
+      "epoch": 0.041666666666666664,
+      "grad_norm": 7.244045116147084,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7245
+    },
+    {
+      "epoch": 0.043859649122807015,
+      "grad_norm": 5.5883852527427385,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 7246
+    },
+    {
+      "epoch": 0.046052631578947366,
+      "grad_norm": 6.406912980404517,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 7247
+    },
+    {
+      "epoch": 0.04824561403508772,
+      "grad_norm": 8.677606271109294,
+      "learning_rate": 1e-06,
+      "loss": 0.3758,
+      "step": 7248
+    },
+    {
+      "epoch": 0.05043859649122807,
+      "grad_norm": 6.517168409504246,
+      "learning_rate": 1e-06,
+      "loss": 0.3367,
+      "step": 7249
+    },
+    {
+      "epoch": 0.05263157894736842,
+      "grad_norm": 6.9646260444267,
+      "learning_rate": 1e-06,
+      "loss": 0.3653,
+      "step": 7250
+    },
+    {
+      "epoch": 0.05482456140350877,
+      "grad_norm": 6.859465884482189,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 7251
+    },
+    {
+      "epoch": 0.05701754385964912,
+      "grad_norm": 6.272998717022487,
+      "learning_rate": 1e-06,
+      "loss": 0.3523,
+      "step": 7252
+    },
+    {
+      "epoch": 0.05921052631578947,
+      "grad_norm": 7.089088305262431,
+      "learning_rate": 1e-06,
+      "loss": 0.3583,
+      "step": 7253
+    },
+    {
+      "epoch": 0.06140350877192982,
+      "grad_norm": 6.087010556691974,
+      "learning_rate": 1e-06,
+      "loss": 0.3545,
+      "step": 7254
+    },
+    {
+      "epoch": 0.06359649122807018,
+      "grad_norm": 6.611733518351989,
+      "learning_rate": 1e-06,
+      "loss": 0.3974,
+      "step": 7255
+    },
+    {
+      "epoch": 0.06578947368421052,
+      "grad_norm": 5.914591100758919,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 7256
+    },
+    {
+      "epoch": 0.06798245614035088,
+      "grad_norm": 7.773587272479548,
+      "learning_rate": 1e-06,
+      "loss": 0.3682,
+      "step": 7257
+    },
+    {
+      "epoch": 0.07017543859649122,
+      "grad_norm": 5.2125054281011405,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 7258
+    },
+    {
+      "epoch": 0.07236842105263158,
+      "grad_norm": 7.39029492573411,
+      "learning_rate": 1e-06,
+      "loss": 0.3982,
+      "step": 7259
+    },
+    {
+      "epoch": 0.07456140350877193,
+      "grad_norm": 8.414253064233632,
+      "learning_rate": 1e-06,
+      "loss": 0.3836,
+      "step": 7260
+    },
+    {
+      "epoch": 0.07675438596491228,
+      "grad_norm": 6.668192220392629,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 7261
+    },
+    {
+      "epoch": 0.07894736842105263,
+      "grad_norm": 7.4499250131816925,
+      "learning_rate": 1e-06,
+      "loss": 0.3666,
+      "step": 7262
+    },
+    {
+      "epoch": 0.08114035087719298,
+      "grad_norm": 6.310851773331443,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 7263
+    },
+    {
+      "epoch": 0.08333333333333333,
+      "grad_norm": 6.767705271702761,
+      "learning_rate": 1e-06,
+      "loss": 0.355,
+      "step": 7264
+    },
+    {
+      "epoch": 0.08552631578947369,
+      "grad_norm": 7.051781021954138,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 7265
+    },
+    {
+      "epoch": 0.08771929824561403,
+      "grad_norm": 7.2033101964910395,
+      "learning_rate": 1e-06,
+      "loss": 0.3507,
+      "step": 7266
+    },
+    {
+      "epoch": 0.08991228070175439,
+      "grad_norm": 7.460441405228069,
+      "learning_rate": 1e-06,
+      "loss": 0.3724,
+      "step": 7267
+    },
+    {
+      "epoch": 0.09210526315789473,
+      "grad_norm": 5.4656409195855185,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 7268
+    },
+    {
+      "epoch": 0.09429824561403509,
+      "grad_norm": 7.443852629445643,
+      "learning_rate": 1e-06,
+      "loss": 0.3706,
+      "step": 7269
+    },
+    {
+      "epoch": 0.09649122807017543,
+      "grad_norm": 5.5942168697972265,
+      "learning_rate": 1e-06,
+      "loss": 0.3795,
+      "step": 7270
+    },
+    {
+      "epoch": 0.09868421052631579,
+      "grad_norm": 5.926293931829736,
+      "learning_rate": 1e-06,
+      "loss": 0.3851,
+      "step": 7271
+    },
+    {
+      "epoch": 0.10087719298245613,
+      "grad_norm": 7.273680797273144,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 7272
+    },
+    {
+      "epoch": 0.10307017543859649,
+      "grad_norm": 7.41707167957173,
+      "learning_rate": 1e-06,
+      "loss": 0.3519,
+      "step": 7273
+    },
+    {
+      "epoch": 0.10526315789473684,
+      "grad_norm": 6.112543787746375,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 7274
+    },
+    {
+      "epoch": 0.1074561403508772,
+      "grad_norm": 6.6410493331561025,
+      "learning_rate": 1e-06,
+      "loss": 0.3641,
+      "step": 7275
+    },
+    {
+      "epoch": 0.10964912280701754,
+      "grad_norm": 6.946011115159454,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 7276
+    },
+    {
+      "epoch": 0.1118421052631579,
+      "grad_norm": 6.233037471283428,
+      "learning_rate": 1e-06,
+      "loss": 0.3422,
+      "step": 7277
+    },
+    {
+      "epoch": 0.11403508771929824,
+      "grad_norm": 5.4743444763316225,
+      "learning_rate": 1e-06,
+      "loss": 0.3637,
+      "step": 7278
+    },
+    {
+      "epoch": 0.1162280701754386,
+      "grad_norm": 5.683057850139127,
+      "learning_rate": 1e-06,
+      "loss": 0.3676,
+      "step": 7279
+    },
+    {
+      "epoch": 0.11842105263157894,
+      "grad_norm": 7.1002279079734105,
+      "learning_rate": 1e-06,
+      "loss": 0.3647,
+      "step": 7280
+    },
+    {
+      "epoch": 0.1206140350877193,
+      "grad_norm": 6.875655797714209,
+      "learning_rate": 1e-06,
+      "loss": 0.3615,
+      "step": 7281
+    },
+    {
+      "epoch": 0.12280701754385964,
+      "grad_norm": 6.017617155488323,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 7282
+    },
+    {
+      "epoch": 0.125,
+      "grad_norm": 7.147916868472657,
+      "learning_rate": 1e-06,
+      "loss": 0.3626,
+      "step": 7283
+    },
+    {
+      "epoch": 0.12719298245614036,
+      "grad_norm": 6.359024687597897,
+      "learning_rate": 1e-06,
+      "loss": 0.3645,
+      "step": 7284
+    },
+    {
+      "epoch": 0.12938596491228072,
+      "grad_norm": 6.328385517875031,
+      "learning_rate": 1e-06,
+      "loss": 0.3578,
+      "step": 7285
+    },
+    {
+      "epoch": 0.13157894736842105,
+      "grad_norm": 8.561312298750954,
+      "learning_rate": 1e-06,
+      "loss": 0.3495,
+      "step": 7286
+    },
+    {
+      "epoch": 0.1337719298245614,
+      "grad_norm": 6.901682126365142,
+      "learning_rate": 1e-06,
+      "loss": 0.3459,
+      "step": 7287
+    },
+    {
+      "epoch": 0.13596491228070176,
+      "grad_norm": 8.960013714382848,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 7288
+    },
+    {
+      "epoch": 0.13815789473684212,
+      "grad_norm": 7.321869519576636,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 7289
+    },
+    {
+      "epoch": 0.14035087719298245,
+      "grad_norm": 6.110221549869078,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 7290
+    },
+    {
+      "epoch": 0.1425438596491228,
+      "grad_norm": 6.794656518609233,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 7291
+    },
+    {
+      "epoch": 0.14473684210526316,
+      "grad_norm": 6.953578556767948,
+      "learning_rate": 1e-06,
+      "loss": 0.3593,
+      "step": 7292
+    },
+    {
+      "epoch": 0.14692982456140352,
+      "grad_norm": 7.906636785972883,
+      "learning_rate": 1e-06,
+      "loss": 0.3651,
+      "step": 7293
+    },
+    {
+      "epoch": 0.14912280701754385,
+      "grad_norm": 6.768389771727818,
+      "learning_rate": 1e-06,
+      "loss": 0.3736,
+      "step": 7294
+    },
+    {
+      "epoch": 0.1513157894736842,
+      "grad_norm": 7.549854659511062,
+      "learning_rate": 1e-06,
+      "loss": 0.3607,
+      "step": 7295
+    },
+    {
+      "epoch": 0.15350877192982457,
+      "grad_norm": 5.7102942760226085,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 7296
+    },
+    {
+      "epoch": 0.15570175438596492,
+      "grad_norm": 7.124709204587776,
+      "learning_rate": 1e-06,
+      "loss": 0.3735,
+      "step": 7297
+    },
+    {
+      "epoch": 0.15789473684210525,
+      "grad_norm": 7.203801983116346,
+      "learning_rate": 1e-06,
+      "loss": 0.375,
+      "step": 7298
+    },
+    {
+      "epoch": 0.1600877192982456,
+      "grad_norm": 5.59513302410192,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 7299
+    },
+    {
+      "epoch": 0.16228070175438597,
+      "grad_norm": 5.463931068369983,
+      "learning_rate": 1e-06,
+      "loss": 0.3756,
+      "step": 7300
+    },
+    {
+      "epoch": 0.16447368421052633,
+      "grad_norm": 5.453793360254799,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 7301
+    },
+    {
+      "epoch": 0.16666666666666666,
+      "grad_norm": 5.967888806856309,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 7302
+    },
+    {
+      "epoch": 0.16885964912280702,
+      "grad_norm": 6.466087194598642,
+      "learning_rate": 1e-06,
+      "loss": 0.3617,
+      "step": 7303
+    },
+    {
+      "epoch": 0.17105263157894737,
+      "grad_norm": 6.671775597033122,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 7304
+    },
+    {
+      "epoch": 0.17324561403508773,
+      "grad_norm": 6.990033780642126,
+      "learning_rate": 1e-06,
+      "loss": 0.3523,
+      "step": 7305
+    },
+    {
+      "epoch": 0.17543859649122806,
+      "grad_norm": 7.909052587393246,
+      "learning_rate": 1e-06,
+      "loss": 0.3703,
+      "step": 7306
+    },
+    {
+      "epoch": 0.17763157894736842,
+      "grad_norm": 5.963006199218983,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 7307
+    },
+    {
+      "epoch": 0.17982456140350878,
+      "grad_norm": 6.862341252761283,
+      "learning_rate": 1e-06,
+      "loss": 0.3611,
+      "step": 7308
+    },
+    {
+      "epoch": 0.18201754385964913,
+      "grad_norm": 9.854957316545828,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 7309
+    },
+    {
+      "epoch": 0.18421052631578946,
+      "grad_norm": 6.830736847791499,
+      "learning_rate": 1e-06,
+      "loss": 0.3472,
+      "step": 7310
+    },
+    {
+      "epoch": 0.18640350877192982,
+      "grad_norm": 7.103214826552931,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 7311
+    },
+    {
+      "epoch": 0.18859649122807018,
+      "grad_norm": 8.982997755956406,
+      "learning_rate": 1e-06,
+      "loss": 0.3392,
+      "step": 7312
+    },
+    {
+      "epoch": 0.19078947368421054,
+      "grad_norm": 7.168328266570389,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 7313
+    },
+    {
+      "epoch": 0.19298245614035087,
+      "grad_norm": 6.935299883540508,
+      "learning_rate": 1e-06,
+      "loss": 0.3579,
+      "step": 7314
+    },
+    {
+      "epoch": 0.19517543859649122,
+      "grad_norm": 5.975548862029615,
+      "learning_rate": 1e-06,
+      "loss": 0.368,
+      "step": 7315
+    },
+    {
+      "epoch": 0.19736842105263158,
+      "grad_norm": 6.885376872218259,
+      "learning_rate": 1e-06,
+      "loss": 0.358,
+      "step": 7316
+    },
+    {
+      "epoch": 0.19956140350877194,
+      "grad_norm": 5.950606961373902,
+      "learning_rate": 1e-06,
+      "loss": 0.362,
+      "step": 7317
+    },
+    {
+      "epoch": 0.20175438596491227,
+      "grad_norm": 7.1833850617809825,
+      "learning_rate": 1e-06,
+      "loss": 0.3879,
+      "step": 7318
+    },
+    {
+      "epoch": 0.20394736842105263,
+      "grad_norm": 5.243869602214475,
+      "learning_rate": 1e-06,
+      "loss": 0.3393,
+      "step": 7319
+    },
+    {
+      "epoch": 0.20614035087719298,
+      "grad_norm": 6.279504991392965,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 7320
+    },
+    {
+      "epoch": 0.20833333333333334,
+      "grad_norm": 4.978681857848636,
+      "learning_rate": 1e-06,
+      "loss": 0.355,
+      "step": 7321
+    },
+    {
+      "epoch": 0.21052631578947367,
+      "grad_norm": 6.374984851381183,
+      "learning_rate": 1e-06,
+      "loss": 0.3787,
+      "step": 7322
+    },
+    {
+      "epoch": 0.21271929824561403,
+      "grad_norm": 7.929030586339701,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 7323
+    },
+    {
+      "epoch": 0.2149122807017544,
+      "grad_norm": 6.303614665447935,
+      "learning_rate": 1e-06,
+      "loss": 0.3664,
+      "step": 7324
+    },
+    {
+      "epoch": 0.21710526315789475,
+      "grad_norm": 8.171739573855492,
+      "learning_rate": 1e-06,
+      "loss": 0.3824,
+      "step": 7325
+    },
+    {
+      "epoch": 0.21929824561403508,
+      "grad_norm": 6.100087976973779,
+      "learning_rate": 1e-06,
+      "loss": 0.3606,
+      "step": 7326
+    },
+    {
+      "epoch": 0.22149122807017543,
+      "grad_norm": 6.428657184126131,
+      "learning_rate": 1e-06,
+      "loss": 0.3429,
+      "step": 7327
+    },
+    {
+      "epoch": 0.2236842105263158,
+      "grad_norm": 6.866427979004686,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 7328
+    },
+    {
+      "epoch": 0.22587719298245615,
+      "grad_norm": 6.767478797181616,
+      "learning_rate": 1e-06,
+      "loss": 0.348,
+      "step": 7329
+    },
+    {
+      "epoch": 0.22807017543859648,
+      "grad_norm": 8.305825077599449,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 7330
+    },
+    {
+      "epoch": 0.23026315789473684,
+      "grad_norm": 5.71854743675516,
+      "learning_rate": 1e-06,
+      "loss": 0.3917,
+      "step": 7331
+    },
+    {
+      "epoch": 0.2324561403508772,
+      "grad_norm": 7.951910671917778,
+      "learning_rate": 1e-06,
+      "loss": 0.3723,
+      "step": 7332
+    },
+    {
+      "epoch": 0.23464912280701755,
+      "grad_norm": 5.509645131519667,
+      "learning_rate": 1e-06,
+      "loss": 0.361,
+      "step": 7333
+    },
+    {
+      "epoch": 0.23684210526315788,
+      "grad_norm": 6.860622564406902,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 7334
+    },
+    {
+      "epoch": 0.23903508771929824,
+      "grad_norm": 6.717282996591877,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 7335
+    },
+    {
+      "epoch": 0.2412280701754386,
+      "grad_norm": 6.970427296159852,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 7336
+    },
+    {
+      "epoch": 0.24342105263157895,
+      "grad_norm": 6.936460566776181,
+      "learning_rate": 1e-06,
+      "loss": 0.3621,
+      "step": 7337
+    },
+    {
+      "epoch": 0.24561403508771928,
+      "grad_norm": 5.579838099135995,
+      "learning_rate": 1e-06,
+      "loss": 0.347,
+      "step": 7338
+    },
+    {
+      "epoch": 0.24780701754385964,
+      "grad_norm": 7.188475274129415,
+      "learning_rate": 1e-06,
+      "loss": 0.3591,
+      "step": 7339
+    },
+    {
+      "epoch": 0.25,
+      "grad_norm": 6.7260575468839825,
+      "learning_rate": 1e-06,
+      "loss": 0.337,
+      "step": 7340
+    },
+    {
+      "epoch": 0.25219298245614036,
+      "grad_norm": 6.226367030314179,
+      "learning_rate": 1e-06,
+      "loss": 0.368,
+      "step": 7341
+    },
+    {
+      "epoch": 0.2543859649122807,
+      "grad_norm": 5.943718420458673,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 7342
+    },
+    {
+      "epoch": 0.2565789473684211,
+      "grad_norm": 7.607536232101633,
+      "learning_rate": 1e-06,
+      "loss": 0.3572,
+      "step": 7343
+    },
+    {
+      "epoch": 0.25877192982456143,
+      "grad_norm": 7.300756548090427,
+      "learning_rate": 1e-06,
+      "loss": 0.335,
+      "step": 7344
+    },
+    {
+      "epoch": 0.26096491228070173,
+      "grad_norm": 6.586000358171558,
+      "learning_rate": 1e-06,
+      "loss": 0.3543,
+      "step": 7345
+    },
+    {
+      "epoch": 0.2631578947368421,
+      "grad_norm": 6.528375987494568,
+      "learning_rate": 1e-06,
+      "loss": 0.3732,
+      "step": 7346
+    },
+    {
+      "epoch": 0.26535087719298245,
+      "grad_norm": 6.769858165586273,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 7347
+    },
+    {
+      "epoch": 0.2675438596491228,
+      "grad_norm": 6.878865992529417,
+      "learning_rate": 1e-06,
+      "loss": 0.3709,
+      "step": 7348
+    },
+    {
+      "epoch": 0.26973684210526316,
+      "grad_norm": 6.598477471607985,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 7349
+    },
+    {
+      "epoch": 0.2719298245614035,
+      "grad_norm": 6.204555982027651,
+      "learning_rate": 1e-06,
+      "loss": 0.366,
+      "step": 7350
+    },
+    {
+      "epoch": 0.2741228070175439,
+      "grad_norm": 8.082023964805325,
+      "learning_rate": 1e-06,
+      "loss": 0.3716,
+      "step": 7351
+    },
+    {
+      "epoch": 0.27631578947368424,
+      "grad_norm": 6.766779869471377,
+      "learning_rate": 1e-06,
+      "loss": 0.3571,
+      "step": 7352
+    },
+    {
+      "epoch": 0.27850877192982454,
+      "grad_norm": 7.307522134606836,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 7353
+    },
+    {
+      "epoch": 0.2807017543859649,
+      "grad_norm": 6.556696377224533,
+      "learning_rate": 1e-06,
+      "loss": 0.3885,
+      "step": 7354
+    },
+    {
+      "epoch": 0.28289473684210525,
+      "grad_norm": 6.898490454515678,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 7355
+    },
+    {
+      "epoch": 0.2850877192982456,
+      "grad_norm": 6.897531545561997,
+      "learning_rate": 1e-06,
+      "loss": 0.369,
+      "step": 7356
+    },
+    {
+      "epoch": 0.28728070175438597,
+      "grad_norm": 7.548132044353999,
+      "learning_rate": 1e-06,
+      "loss": 0.3779,
+      "step": 7357
+    },
+    {
+      "epoch": 0.2894736842105263,
+      "grad_norm": 5.717588596345156,
+      "learning_rate": 1e-06,
+      "loss": 0.3486,
+      "step": 7358
+    },
+    {
+      "epoch": 0.2916666666666667,
+      "grad_norm": 6.385591213603696,
+      "learning_rate": 1e-06,
+      "loss": 0.3648,
+      "step": 7359
+    },
+    {
+      "epoch": 0.29385964912280704,
+      "grad_norm": 8.05728930608768,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 7360
+    },
+    {
+      "epoch": 0.29605263157894735,
+      "grad_norm": 6.458637693961394,
+      "learning_rate": 1e-06,
+      "loss": 0.37,
+      "step": 7361
+    },
+    {
+      "epoch": 0.2982456140350877,
+      "grad_norm": 7.939638928187078,
+      "learning_rate": 1e-06,
+      "loss": 0.3455,
+      "step": 7362
+    },
+    {
+      "epoch": 0.30043859649122806,
+      "grad_norm": 6.445215946037831,
+      "learning_rate": 1e-06,
+      "loss": 0.3508,
+      "step": 7363
+    },
+    {
+      "epoch": 0.3026315789473684,
+      "grad_norm": 7.945870126631332,
+      "learning_rate": 1e-06,
+      "loss": 0.355,
+      "step": 7364
+    },
+    {
+      "epoch": 0.3048245614035088,
+      "grad_norm": 5.9582547776395876,
+      "learning_rate": 1e-06,
+      "loss": 0.3507,
+      "step": 7365
+    },
+    {
+      "epoch": 0.30701754385964913,
+      "grad_norm": 6.394311936810813,
+      "learning_rate": 1e-06,
+      "loss": 0.3781,
+      "step": 7366
+    },
+    {
+      "epoch": 0.3092105263157895,
+      "grad_norm": 8.04485675077243,
+      "learning_rate": 1e-06,
+      "loss": 0.3664,
+      "step": 7367
+    },
+    {
+      "epoch": 0.31140350877192985,
+      "grad_norm": 6.231237976353527,
+      "learning_rate": 1e-06,
+      "loss": 0.3505,
+      "step": 7368
+    },
+    {
+      "epoch": 0.31359649122807015,
+      "grad_norm": 7.708614164833223,
+      "learning_rate": 1e-06,
+      "loss": 0.3827,
+      "step": 7369
+    },
+    {
+      "epoch": 0.3157894736842105,
+      "grad_norm": 6.5166350979448024,
+      "learning_rate": 1e-06,
+      "loss": 0.3577,
+      "step": 7370
+    },
+    {
+      "epoch": 0.31798245614035087,
+      "grad_norm": 6.957259500050806,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 7371
+    },
+    {
+      "epoch": 0.3201754385964912,
+      "grad_norm": 6.141975734854891,
+      "learning_rate": 1e-06,
+      "loss": 0.3647,
+      "step": 7372
+    },
+    {
+      "epoch": 0.3223684210526316,
+      "grad_norm": 7.316620785897282,
+      "learning_rate": 1e-06,
+      "loss": 0.3859,
+      "step": 7373
+    },
+    {
+      "epoch": 0.32456140350877194,
+      "grad_norm": 9.297455091182407,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7374
+    },
+    {
+      "epoch": 0.3267543859649123,
+      "grad_norm": 7.251162685500487,
+      "learning_rate": 1e-06,
+      "loss": 0.3565,
+      "step": 7375
+    },
+    {
+      "epoch": 0.32894736842105265,
+      "grad_norm": 5.68981405348859,
+      "learning_rate": 1e-06,
+      "loss": 0.3576,
+      "step": 7376
+    },
+    {
+      "epoch": 0.33114035087719296,
+      "grad_norm": 5.146795732509725,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 7377
+    },
+    {
+      "epoch": 0.3333333333333333,
+      "grad_norm": 7.850354395449812,
+      "learning_rate": 1e-06,
+      "loss": 0.3518,
+      "step": 7378
+    },
+    {
+      "epoch": 0.3355263157894737,
+      "grad_norm": 7.427764973497797,
+      "learning_rate": 1e-06,
+      "loss": 0.3822,
+      "step": 7379
+    },
+    {
+      "epoch": 0.33771929824561403,
+      "grad_norm": 7.579655043808997,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 7380
+    },
+    {
+      "epoch": 0.3399122807017544,
+      "grad_norm": 5.971459567988351,
+      "learning_rate": 1e-06,
+      "loss": 0.3662,
+      "step": 7381
+    },
+    {
+      "epoch": 0.34210526315789475,
+      "grad_norm": 6.459474903645241,
+      "learning_rate": 1e-06,
+      "loss": 0.3958,
+      "step": 7382
+    },
+    {
+      "epoch": 0.3442982456140351,
+      "grad_norm": 8.08826125890157,
+      "learning_rate": 1e-06,
+      "loss": 0.3911,
+      "step": 7383
+    },
+    {
+      "epoch": 0.34649122807017546,
+      "grad_norm": 5.707002078990044,
+      "learning_rate": 1e-06,
+      "loss": 0.3642,
+      "step": 7384
+    },
+    {
+      "epoch": 0.34868421052631576,
+      "grad_norm": 6.228828794397183,
+      "learning_rate": 1e-06,
+      "loss": 0.3524,
+      "step": 7385
+    },
+    {
+      "epoch": 0.3508771929824561,
+      "grad_norm": 6.122139482458061,
+      "learning_rate": 1e-06,
+      "loss": 0.3997,
+      "step": 7386
+    },
+    {
+      "epoch": 0.3530701754385965,
+      "grad_norm": 4.953199267512203,
+      "learning_rate": 1e-06,
+      "loss": 0.3733,
+      "step": 7387
+    },
+    {
+      "epoch": 0.35526315789473684,
+      "grad_norm": 8.908710341709154,
+      "learning_rate": 1e-06,
+      "loss": 0.3639,
+      "step": 7388
+    },
+    {
+      "epoch": 0.3574561403508772,
+      "grad_norm": 9.064856744926196,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 7389
+    },
+    {
+      "epoch": 0.35964912280701755,
+      "grad_norm": 5.663165279905027,
+      "learning_rate": 1e-06,
+      "loss": 0.3638,
+      "step": 7390
+    },
+    {
+      "epoch": 0.3618421052631579,
+      "grad_norm": 7.223195375279547,
+      "learning_rate": 1e-06,
+      "loss": 0.3794,
+      "step": 7391
+    },
+    {
+      "epoch": 0.36403508771929827,
+      "grad_norm": 5.243028823623023,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 7392
+    },
+    {
+      "epoch": 0.36622807017543857,
+      "grad_norm": 6.126848629129393,
+      "learning_rate": 1e-06,
+      "loss": 0.383,
+      "step": 7393
+    },
+    {
+      "epoch": 0.3684210526315789,
+      "grad_norm": 6.9695936909551905,
+      "learning_rate": 1e-06,
+      "loss": 0.3758,
+      "step": 7394
+    },
+    {
+      "epoch": 0.3706140350877193,
+      "grad_norm": 6.979820209321579,
+      "learning_rate": 1e-06,
+      "loss": 0.3681,
+      "step": 7395
+    },
+    {
+      "epoch": 0.37280701754385964,
+      "grad_norm": 7.497320231704673,
+      "learning_rate": 1e-06,
+      "loss": 0.3752,
+      "step": 7396
+    },
+    {
+      "epoch": 0.375,
+      "grad_norm": 6.184110036860732,
+      "learning_rate": 1e-06,
+      "loss": 0.3726,
+      "step": 7397
+    },
+    {
+      "epoch": 0.37719298245614036,
+      "grad_norm": 7.519204540888368,
+      "learning_rate": 1e-06,
+      "loss": 0.3647,
+      "step": 7398
+    },
+    {
+      "epoch": 0.3793859649122807,
+      "grad_norm": 7.180273702732639,
+      "learning_rate": 1e-06,
+      "loss": 0.3702,
+      "step": 7399
+    },
+    {
+      "epoch": 0.3815789473684211,
+      "grad_norm": 6.771240936469755,
+      "learning_rate": 1e-06,
+      "loss": 0.3826,
+      "step": 7400
+    },
+    {
+      "epoch": 0.38377192982456143,
+      "grad_norm": 6.45725838170993,
+      "learning_rate": 1e-06,
+      "loss": 0.3663,
+      "step": 7401
+    },
+    {
+      "epoch": 0.38596491228070173,
+      "grad_norm": 7.915478114579846,
+      "learning_rate": 1e-06,
+      "loss": 0.3693,
+      "step": 7402
+    },
+    {
+      "epoch": 0.3881578947368421,
+      "grad_norm": 6.339807388767104,
+      "learning_rate": 1e-06,
+      "loss": 0.3755,
+      "step": 7403
+    },
+    {
+      "epoch": 0.39035087719298245,
+      "grad_norm": 6.698877281058327,
+      "learning_rate": 1e-06,
+      "loss": 0.3737,
+      "step": 7404
+    },
+    {
+      "epoch": 0.3925438596491228,
+      "grad_norm": 5.27886300062907,
+      "learning_rate": 1e-06,
+      "loss": 0.3724,
+      "step": 7405
+    },
+    {
+      "epoch": 0.39473684210526316,
+      "grad_norm": 5.532708255978353,
+      "learning_rate": 1e-06,
+      "loss": 0.3462,
+      "step": 7406
+    },
+    {
+      "epoch": 0.3969298245614035,
+      "grad_norm": 7.510068704239833,
+      "learning_rate": 1e-06,
+      "loss": 0.3775,
+      "step": 7407
+    },
+    {
+      "epoch": 0.3991228070175439,
+      "grad_norm": 5.595679589437584,
+      "learning_rate": 1e-06,
+      "loss": 0.3426,
+      "step": 7408
+    },
+    {
+      "epoch": 0.40131578947368424,
+      "grad_norm": 5.005719693554534,
+      "learning_rate": 1e-06,
+      "loss": 0.3437,
+      "step": 7409
+    },
+    {
+      "epoch": 0.40350877192982454,
+      "grad_norm": 6.643728395190395,
+      "learning_rate": 1e-06,
+      "loss": 0.3823,
+      "step": 7410
+    },
+    {
+      "epoch": 0.4057017543859649,
+      "grad_norm": 6.318887759947985,
+      "learning_rate": 1e-06,
+      "loss": 0.3627,
+      "step": 7411
+    },
+    {
+      "epoch": 0.40789473684210525,
+      "grad_norm": 8.119510559876062,
+      "learning_rate": 1e-06,
+      "loss": 0.3531,
+      "step": 7412
+    },
+    {
+      "epoch": 0.4100877192982456,
+      "grad_norm": 5.86833017092401,
+      "learning_rate": 1e-06,
+      "loss": 0.3822,
+      "step": 7413
+    },
+    {
+      "epoch": 0.41228070175438597,
+      "grad_norm": 7.769752913612583,
+      "learning_rate": 1e-06,
+      "loss": 0.3882,
+      "step": 7414
+    },
+    {
+      "epoch": 0.4144736842105263,
+      "grad_norm": 5.816598910637293,
+      "learning_rate": 1e-06,
+      "loss": 0.3679,
+      "step": 7415
+    },
+    {
+      "epoch": 0.4166666666666667,
+      "grad_norm": 9.334410319803855,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 7416
+    },
+    {
+      "epoch": 0.41885964912280704,
+      "grad_norm": 5.6308308153299835,
+      "learning_rate": 1e-06,
+      "loss": 0.3463,
+      "step": 7417
+    },
+    {
+      "epoch": 0.42105263157894735,
+      "grad_norm": 5.921927178997208,
+      "learning_rate": 1e-06,
+      "loss": 0.3578,
+      "step": 7418
+    },
+    {
+      "epoch": 0.4232456140350877,
+      "grad_norm": 6.310660968816208,
+      "learning_rate": 1e-06,
+      "loss": 0.3707,
+      "step": 7419
+    },
+    {
+      "epoch": 0.42543859649122806,
+      "grad_norm": 5.620922988797356,
+      "learning_rate": 1e-06,
+      "loss": 0.3519,
+      "step": 7420
+    },
+    {
+      "epoch": 0.4276315789473684,
+      "grad_norm": 6.874156171680931,
+      "learning_rate": 1e-06,
+      "loss": 0.3551,
+      "step": 7421
+    },
+    {
+      "epoch": 0.4298245614035088,
+      "grad_norm": 8.48076012663136,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 7422
+    },
+    {
+      "epoch": 0.43201754385964913,
+      "grad_norm": 6.277523518483997,
+      "learning_rate": 1e-06,
+      "loss": 0.3782,
+      "step": 7423
+    },
+    {
+      "epoch": 0.4342105263157895,
+      "grad_norm": 7.186248306540009,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 7424
+    },
+    {
+      "epoch": 0.43640350877192985,
+      "grad_norm": 6.7708721757846,
+      "learning_rate": 1e-06,
+      "loss": 0.3861,
+      "step": 7425
+    },
+    {
+      "epoch": 0.43859649122807015,
+      "grad_norm": 5.818767567743086,
+      "learning_rate": 1e-06,
+      "loss": 0.3835,
+      "step": 7426
+    },
+    {
+      "epoch": 0.4407894736842105,
+      "grad_norm": 6.719511271786504,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 7427
+    },
+    {
+      "epoch": 0.44298245614035087,
+      "grad_norm": 6.804543672198414,
+      "learning_rate": 1e-06,
+      "loss": 0.3705,
+      "step": 7428
+    },
+    {
+      "epoch": 0.4451754385964912,
+      "grad_norm": 6.678392459561648,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 7429
+    },
+    {
+      "epoch": 0.4473684210526316,
+      "grad_norm": 5.815051793364725,
+      "learning_rate": 1e-06,
+      "loss": 0.3393,
+      "step": 7430
+    },
+    {
+      "epoch": 0.44956140350877194,
+      "grad_norm": 6.052651349199429,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7431
+    },
+    {
+      "epoch": 0.4517543859649123,
+      "grad_norm": 6.013855059758517,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 7432
+    },
+    {
+      "epoch": 0.45394736842105265,
+      "grad_norm": 6.849500811808389,
+      "learning_rate": 1e-06,
+      "loss": 0.362,
+      "step": 7433
+    },
+    {
+      "epoch": 0.45614035087719296,
+      "grad_norm": 5.793122252594787,
+      "learning_rate": 1e-06,
+      "loss": 0.3575,
+      "step": 7434
+    },
+    {
+      "epoch": 0.4583333333333333,
+      "grad_norm": 6.4442245373799505,
+      "learning_rate": 1e-06,
+      "loss": 0.359,
+      "step": 7435
+    },
+    {
+      "epoch": 0.4605263157894737,
+      "grad_norm": 6.765092069367693,
+      "learning_rate": 1e-06,
+      "loss": 0.3841,
+      "step": 7436
+    },
+    {
+      "epoch": 0.46271929824561403,
+      "grad_norm": 7.444538513213365,
+      "learning_rate": 1e-06,
+      "loss": 0.3584,
+      "step": 7437
+    },
+    {
+      "epoch": 0.4649122807017544,
+      "grad_norm": 6.977902736297657,
+      "learning_rate": 1e-06,
+      "loss": 0.3598,
+      "step": 7438
+    },
+    {
+      "epoch": 0.46710526315789475,
+      "grad_norm": 7.354253244656211,
+      "learning_rate": 1e-06,
+      "loss": 0.385,
+      "step": 7439
+    },
+    {
+      "epoch": 0.4692982456140351,
+      "grad_norm": 6.004991037720651,
+      "learning_rate": 1e-06,
+      "loss": 0.3594,
+      "step": 7440
+    },
+    {
+      "epoch": 0.47149122807017546,
+      "grad_norm": 7.60452504920644,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 7441
+    },
+    {
+      "epoch": 0.47368421052631576,
+      "grad_norm": 7.840017699400008,
+      "learning_rate": 1e-06,
+      "loss": 0.3768,
+      "step": 7442
+    },
+    {
+      "epoch": 0.4758771929824561,
+      "grad_norm": 5.42549962932565,
+      "learning_rate": 1e-06,
+      "loss": 0.3704,
+      "step": 7443
+    },
+    {
+      "epoch": 0.4780701754385965,
+      "grad_norm": 5.3868902922241375,
+      "learning_rate": 1e-06,
+      "loss": 0.3636,
+      "step": 7444
+    },
+    {
+      "epoch": 0.48026315789473684,
+      "grad_norm": 6.133299379997361,
+      "learning_rate": 1e-06,
+      "loss": 0.3337,
+      "step": 7445
+    },
+    {
+      "epoch": 0.4824561403508772,
+      "grad_norm": 6.026647344570084,
+      "learning_rate": 1e-06,
+      "loss": 0.3753,
+      "step": 7446
+    },
+    {
+      "epoch": 0.48464912280701755,
+      "grad_norm": 7.464227427275389,
+      "learning_rate": 1e-06,
+      "loss": 0.3693,
+      "step": 7447
+    },
+    {
+      "epoch": 0.4868421052631579,
+      "grad_norm": 6.723793654484921,
+      "learning_rate": 1e-06,
+      "loss": 0.3587,
+      "step": 7448
+    },
+    {
+      "epoch": 0.48903508771929827,
+      "grad_norm": 9.338299370794738,
+      "learning_rate": 1e-06,
+      "loss": 0.3837,
+      "step": 7449
+    },
+    {
+      "epoch": 0.49122807017543857,
+      "grad_norm": 6.7308132666495215,
+      "learning_rate": 1e-06,
+      "loss": 0.3959,
+      "step": 7450
+    },
+    {
+      "epoch": 0.4934210526315789,
+      "grad_norm": 5.4595394914390685,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7451
+    },
+    {
+      "epoch": 0.4956140350877193,
+      "grad_norm": 6.304427283167244,
+      "learning_rate": 1e-06,
+      "loss": 0.3728,
+      "step": 7452
+    },
+    {
+      "epoch": 0.49780701754385964,
+      "grad_norm": 6.5017224024452664,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 7453
+    },
+    {
+      "epoch": 0.5,
+      "grad_norm": 6.974948356804394,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 7454
+    },
+    {
+      "epoch": 0.5021929824561403,
+      "grad_norm": 7.546111138100996,
+      "learning_rate": 1e-06,
+      "loss": 0.3721,
+      "step": 7455
+    },
+    {
+      "epoch": 0.5043859649122807,
+      "grad_norm": 6.0804183228549675,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 7456
+    },
+    {
+      "epoch": 0.506578947368421,
+      "grad_norm": 5.034459806552062,
+      "learning_rate": 1e-06,
+      "loss": 0.3745,
+      "step": 7457
+    },
+    {
+      "epoch": 0.5087719298245614,
+      "grad_norm": 7.193979565342462,
+      "learning_rate": 1e-06,
+      "loss": 0.3438,
+      "step": 7458
+    },
+    {
+      "epoch": 0.5109649122807017,
+      "grad_norm": 6.541164499717168,
+      "learning_rate": 1e-06,
+      "loss": 0.3602,
+      "step": 7459
+    },
+    {
+      "epoch": 0.5131578947368421,
+      "grad_norm": 6.955504529842805,
+      "learning_rate": 1e-06,
+      "loss": 0.3649,
+      "step": 7460
+    },
+    {
+      "epoch": 0.5153508771929824,
+      "grad_norm": 6.010539043535702,
+      "learning_rate": 1e-06,
+      "loss": 0.3601,
+      "step": 7461
+    },
+    {
+      "epoch": 0.5175438596491229,
+      "grad_norm": 6.71927994633116,
+      "learning_rate": 1e-06,
+      "loss": 0.3567,
+      "step": 7462
+    },
+    {
+      "epoch": 0.5197368421052632,
+      "grad_norm": 4.121239017200561,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 7463
+    },
+    {
+      "epoch": 0.5219298245614035,
+      "grad_norm": 7.812999394574864,
+      "learning_rate": 1e-06,
+      "loss": 0.372,
+      "step": 7464
+    },
+    {
+      "epoch": 0.5241228070175439,
+      "grad_norm": 5.88704475243401,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 7465
+    },
+    {
+      "epoch": 0.5263157894736842,
+      "grad_norm": 7.202795711259781,
+      "learning_rate": 1e-06,
+      "loss": 0.3869,
+      "step": 7466
+    },
+    {
+      "epoch": 0.5285087719298246,
+      "grad_norm": 7.155483664167044,
+      "learning_rate": 1e-06,
+      "loss": 0.3634,
+      "step": 7467
+    },
+    {
+      "epoch": 0.5307017543859649,
+      "grad_norm": 8.163340658977386,
+      "learning_rate": 1e-06,
+      "loss": 0.3739,
+      "step": 7468
+    },
+    {
+      "epoch": 0.5328947368421053,
+      "grad_norm": 7.2522617720550855,
+      "learning_rate": 1e-06,
+      "loss": 0.3753,
+      "step": 7469
+    },
+    {
+      "epoch": 0.5350877192982456,
+      "grad_norm": 5.621447652569409,
+      "learning_rate": 1e-06,
+      "loss": 0.3868,
+      "step": 7470
+    },
+    {
+      "epoch": 0.5372807017543859,
+      "grad_norm": 5.8457993076643975,
+      "learning_rate": 1e-06,
+      "loss": 0.3783,
+      "step": 7471
+    },
+    {
+      "epoch": 0.5394736842105263,
+      "grad_norm": 6.59031039578389,
+      "learning_rate": 1e-06,
+      "loss": 0.3996,
+      "step": 7472
+    },
+    {
+      "epoch": 0.5416666666666666,
+      "grad_norm": 7.800938138652031,
+      "learning_rate": 1e-06,
+      "loss": 0.3946,
+      "step": 7473
+    },
+    {
+      "epoch": 0.543859649122807,
+      "grad_norm": 7.948743397781104,
+      "learning_rate": 1e-06,
+      "loss": 0.387,
+      "step": 7474
+    },
+    {
+      "epoch": 0.5460526315789473,
+      "grad_norm": 5.064452998343001,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 7475
+    },
+    {
+      "epoch": 0.5482456140350878,
+      "grad_norm": 7.085991070917747,
+      "learning_rate": 1e-06,
+      "loss": 0.3767,
+      "step": 7476
+    },
+    {
+      "epoch": 0.5504385964912281,
+      "grad_norm": 7.56206615930517,
+      "learning_rate": 1e-06,
+      "loss": 0.3526,
+      "step": 7477
+    },
+    {
+      "epoch": 0.5526315789473685,
+      "grad_norm": 6.660154524283626,
+      "learning_rate": 1e-06,
+      "loss": 0.3476,
+      "step": 7478
+    },
+    {
+      "epoch": 0.5548245614035088,
+      "grad_norm": 6.135112190393495,
+      "learning_rate": 1e-06,
+      "loss": 0.3639,
+      "step": 7479
+    },
+    {
+      "epoch": 0.5570175438596491,
+      "grad_norm": 6.772262444497539,
+      "learning_rate": 1e-06,
+      "loss": 0.3748,
+      "step": 7480
+    },
+    {
+      "epoch": 0.5592105263157895,
+      "grad_norm": 7.051186509787276,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 7481
+    },
+    {
+      "epoch": 0.5614035087719298,
+      "grad_norm": 9.031745496563861,
+      "learning_rate": 1e-06,
+      "loss": 0.3807,
+      "step": 7482
+    },
+    {
+      "epoch": 0.5635964912280702,
+      "grad_norm": 6.893583387123159,
+      "learning_rate": 1e-06,
+      "loss": 0.3572,
+      "step": 7483
+    },
+    {
+      "epoch": 0.5657894736842105,
+      "grad_norm": 8.621010836042984,
+      "learning_rate": 1e-06,
+      "loss": 0.3688,
+      "step": 7484
+    },
+    {
+      "epoch": 0.5679824561403509,
+      "grad_norm": 5.369413845046128,
+      "learning_rate": 1e-06,
+      "loss": 0.3405,
+      "step": 7485
+    },
+    {
+      "epoch": 0.5701754385964912,
+      "grad_norm": 7.593736096927932,
+      "learning_rate": 1e-06,
+      "loss": 0.3962,
+      "step": 7486
+    },
+    {
+      "epoch": 0.5723684210526315,
+      "grad_norm": 6.693237018103414,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 7487
+    },
+    {
+      "epoch": 0.5745614035087719,
+      "grad_norm": 6.789186296038058,
+      "learning_rate": 1e-06,
+      "loss": 0.3706,
+      "step": 7488
+    },
+    {
+      "epoch": 0.5767543859649122,
+      "grad_norm": 6.251648536728357,
+      "learning_rate": 1e-06,
+      "loss": 0.3582,
+      "step": 7489
+    },
+    {
+      "epoch": 0.5789473684210527,
+      "grad_norm": 5.833238224198108,
+      "learning_rate": 1e-06,
+      "loss": 0.3574,
+      "step": 7490
+    },
+    {
+      "epoch": 0.581140350877193,
+      "grad_norm": 7.159996566985842,
+      "learning_rate": 1e-06,
+      "loss": 0.3678,
+      "step": 7491
+    },
+    {
+      "epoch": 0.5833333333333334,
+      "grad_norm": 5.886447222323071,
+      "learning_rate": 1e-06,
+      "loss": 0.3299,
+      "step": 7492
+    },
+    {
+      "epoch": 0.5855263157894737,
+      "grad_norm": 6.315003692543136,
+      "learning_rate": 1e-06,
+      "loss": 0.3541,
+      "step": 7493
+    },
+    {
+      "epoch": 0.5877192982456141,
+      "grad_norm": 6.094757703607746,
+      "learning_rate": 1e-06,
+      "loss": 0.391,
+      "step": 7494
+    },
+    {
+      "epoch": 0.5899122807017544,
+      "grad_norm": 7.030149999245711,
+      "learning_rate": 1e-06,
+      "loss": 0.3813,
+      "step": 7495
+    },
+    {
+      "epoch": 0.5921052631578947,
+      "grad_norm": 7.436326658470419,
+      "learning_rate": 1e-06,
+      "loss": 0.382,
+      "step": 7496
+    },
+    {
+      "epoch": 0.5942982456140351,
+      "grad_norm": 6.902949544401232,
+      "learning_rate": 1e-06,
+      "loss": 0.3876,
+      "step": 7497
+    },
+    {
+      "epoch": 0.5964912280701754,
+      "grad_norm": 6.861192908410106,
+      "learning_rate": 1e-06,
+      "loss": 0.3566,
+      "step": 7498
+    },
+    {
+      "epoch": 0.5986842105263158,
+      "grad_norm": 6.393251727301545,
+      "learning_rate": 1e-06,
+      "loss": 0.3734,
+      "step": 7499
+    },
+    {
+      "epoch": 0.6008771929824561,
+      "grad_norm": 7.532891247223321,
+      "learning_rate": 1e-06,
+      "loss": 0.3666,
+      "step": 7500
+    },
+    {
+      "epoch": 0.6030701754385965,
+      "grad_norm": 7.146778489317739,
+      "learning_rate": 1e-06,
+      "loss": 0.3737,
+      "step": 7501
+    },
+    {
+      "epoch": 0.6052631578947368,
+      "grad_norm": 6.565016871733574,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 7502
+    },
+    {
+      "epoch": 0.6074561403508771,
+      "grad_norm": 5.473841128885625,
+      "learning_rate": 1e-06,
+      "loss": 0.34,
+      "step": 7503
+    },
+    {
+      "epoch": 0.6096491228070176,
+      "grad_norm": 6.960929196184998,
+      "learning_rate": 1e-06,
+      "loss": 0.3569,
+      "step": 7504
+    },
+    {
+      "epoch": 0.6118421052631579,
+      "grad_norm": 5.873708167755924,
+      "learning_rate": 1e-06,
+      "loss": 0.3518,
+      "step": 7505
+    },
+    {
+      "epoch": 0.6140350877192983,
+      "grad_norm": 7.946584856563132,
+      "learning_rate": 1e-06,
+      "loss": 0.3502,
+      "step": 7506
+    },
+    {
+      "epoch": 0.6162280701754386,
+      "grad_norm": 8.389484540498788,
+      "learning_rate": 1e-06,
+      "loss": 0.3918,
+      "step": 7507
+    },
+    {
+      "epoch": 0.618421052631579,
+      "grad_norm": 7.184077064135135,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 7508
+    },
+    {
+      "epoch": 0.6206140350877193,
+      "grad_norm": 7.786095550768631,
+      "learning_rate": 1e-06,
+      "loss": 0.3658,
+      "step": 7509
+    },
+    {
+      "epoch": 0.6228070175438597,
+      "grad_norm": 5.827630269399014,
+      "learning_rate": 1e-06,
+      "loss": 0.3743,
+      "step": 7510
+    },
+    {
+      "epoch": 0.625,
+      "grad_norm": 6.614195293101189,
+      "learning_rate": 1e-06,
+      "loss": 0.356,
+      "step": 7511
+    },
+    {
+      "epoch": 0.6271929824561403,
+      "grad_norm": 6.4298559648353875,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 7512
+    },
+    {
+      "epoch": 0.6293859649122807,
+      "grad_norm": 7.862633930001,
+      "learning_rate": 1e-06,
+      "loss": 0.3677,
+      "step": 7513
+    },
+    {
+      "epoch": 0.631578947368421,
+      "grad_norm": 7.862633930001,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7514
+    },
+    {
+      "epoch": 0.6337719298245614,
+      "grad_norm": 6.791920199516645,
+      "learning_rate": 1e-06,
+      "loss": 0.38,
+      "step": 7515
+    },
+    {
+      "epoch": 0.6359649122807017,
+      "grad_norm": 5.779323378474881,
+      "learning_rate": 1e-06,
+      "loss": 0.346,
+      "step": 7516
+    },
+    {
+      "epoch": 0.6381578947368421,
+      "grad_norm": 6.737447625628418,
+      "learning_rate": 1e-06,
+      "loss": 0.3649,
+      "step": 7517
+    },
+    {
+      "epoch": 0.6403508771929824,
+      "grad_norm": 7.347964832887293,
+      "learning_rate": 1e-06,
+      "loss": 0.3578,
+      "step": 7518
+    },
+    {
+      "epoch": 0.6425438596491229,
+      "grad_norm": 7.425077987350616,
+      "learning_rate": 1e-06,
+      "loss": 0.3573,
+      "step": 7519
+    },
+    {
+      "epoch": 0.6447368421052632,
+      "grad_norm": 4.993469146134579,
+      "learning_rate": 1e-06,
+      "loss": 0.3773,
+      "step": 7520
+    },
+    {
+      "epoch": 0.6469298245614035,
+      "grad_norm": 6.837699648702226,
+      "learning_rate": 1e-06,
+      "loss": 0.3781,
+      "step": 7521
+    },
+    {
+      "epoch": 0.6491228070175439,
+      "grad_norm": 6.684254004474366,
+      "learning_rate": 1e-06,
+      "loss": 0.3678,
+      "step": 7522
+    },
+    {
+      "epoch": 0.6513157894736842,
+      "grad_norm": 5.955334025162885,
+      "learning_rate": 1e-06,
+      "loss": 0.3558,
+      "step": 7523
+    },
+    {
+      "epoch": 0.6535087719298246,
+      "grad_norm": 7.2885891078928,
+      "learning_rate": 1e-06,
+      "loss": 0.346,
+      "step": 7524
+    },
+    {
+      "epoch": 0.6557017543859649,
+      "grad_norm": 7.033586081815567,
+      "learning_rate": 1e-06,
+      "loss": 0.3877,
+      "step": 7525
+    },
+    {
+      "epoch": 0.6578947368421053,
+      "grad_norm": 8.829914576369157,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 7526
+    },
+    {
+      "epoch": 0.6600877192982456,
+      "grad_norm": 5.011847412384148,
+      "learning_rate": 1e-06,
+      "loss": 0.3718,
+      "step": 7527
+    },
+    {
+      "epoch": 0.6622807017543859,
+      "grad_norm": 6.812653568286395,
+      "learning_rate": 1e-06,
+      "loss": 0.3695,
+      "step": 7528
+    },
+    {
+      "epoch": 0.6644736842105263,
+      "grad_norm": 7.1485925282574705,
+      "learning_rate": 1e-06,
+      "loss": 0.3672,
+      "step": 7529
+    },
+    {
+      "epoch": 0.6666666666666666,
+      "grad_norm": 7.242003529502546,
+      "learning_rate": 1e-06,
+      "loss": 0.3895,
+      "step": 7530
+    },
+    {
+      "epoch": 0.668859649122807,
+      "grad_norm": 5.336443802546755,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 7531
+    },
+    {
+      "epoch": 0.6710526315789473,
+      "grad_norm": 6.923447244879061,
+      "learning_rate": 1e-06,
+      "loss": 0.3654,
+      "step": 7532
+    },
+    {
+      "epoch": 0.6732456140350878,
+      "grad_norm": 6.452124933957357,
+      "learning_rate": 1e-06,
+      "loss": 0.3553,
+      "step": 7533
+    },
+    {
+      "epoch": 0.6754385964912281,
+      "grad_norm": 5.58393414644655,
+      "learning_rate": 1e-06,
+      "loss": 0.3867,
+      "step": 7534
+    },
+    {
+      "epoch": 0.6776315789473685,
+      "grad_norm": 6.291529639416827,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 7535
+    },
+    {
+      "epoch": 0.6798245614035088,
+      "grad_norm": 7.0630525515849865,
+      "learning_rate": 1e-06,
+      "loss": 0.3748,
+      "step": 7536
+    },
+    {
+      "epoch": 0.6820175438596491,
+      "grad_norm": 6.742147470061322,
+      "learning_rate": 1e-06,
+      "loss": 0.3791,
+      "step": 7537
+    },
+    {
+      "epoch": 0.6842105263157895,
+      "grad_norm": 6.75856062683986,
+      "learning_rate": 1e-06,
+      "loss": 0.3683,
+      "step": 7538
+    },
+    {
+      "epoch": 0.6864035087719298,
+      "grad_norm": 7.332836345609679,
+      "learning_rate": 1e-06,
+      "loss": 0.3634,
+      "step": 7539
+    },
+    {
+      "epoch": 0.6885964912280702,
+      "grad_norm": 8.966724491507536,
+      "learning_rate": 1e-06,
+      "loss": 0.3513,
+      "step": 7540
+    },
+    {
+      "epoch": 0.6907894736842105,
+      "grad_norm": 7.1017304219336905,
+      "learning_rate": 1e-06,
+      "loss": 0.381,
+      "step": 7541
+    },
+    {
+      "epoch": 0.6929824561403509,
+      "grad_norm": 7.965506429674132,
+      "learning_rate": 1e-06,
+      "loss": 0.3867,
+      "step": 7542
+    },
+    {
+      "epoch": 0.6951754385964912,
+      "grad_norm": 4.315081795817021,
+      "learning_rate": 1e-06,
+      "loss": 0.3739,
+      "step": 7543
+    },
+    {
+      "epoch": 0.6973684210526315,
+      "grad_norm": 7.119216868653921,
+      "learning_rate": 1e-06,
+      "loss": 0.3326,
+      "step": 7544
+    },
+    {
+      "epoch": 0.6995614035087719,
+      "grad_norm": 8.37058414220666,
+      "learning_rate": 1e-06,
+      "loss": 0.3643,
+      "step": 7545
+    },
+    {
+      "epoch": 0.7017543859649122,
+      "grad_norm": 6.220574746901329,
+      "learning_rate": 1e-06,
+      "loss": 0.3626,
+      "step": 7546
+    },
+    {
+      "epoch": 0.7039473684210527,
+      "grad_norm": 5.822379218347921,
+      "learning_rate": 1e-06,
+      "loss": 0.3436,
+      "step": 7547
+    },
+    {
+      "epoch": 0.706140350877193,
+      "grad_norm": 6.6728048903123245,
+      "learning_rate": 1e-06,
+      "loss": 0.3873,
+      "step": 7548
+    },
+    {
+      "epoch": 0.7083333333333334,
+      "grad_norm": 7.463592036743008,
+      "learning_rate": 1e-06,
+      "loss": 0.3608,
+      "step": 7549
+    },
+    {
+      "epoch": 0.7105263157894737,
+      "grad_norm": 6.352690873175673,
+      "learning_rate": 1e-06,
+      "loss": 0.3747,
+      "step": 7550
+    },
+    {
+      "epoch": 0.7127192982456141,
+      "grad_norm": 5.466396073565239,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 7551
+    },
+    {
+      "epoch": 0.7149122807017544,
+      "grad_norm": 7.4485297066997305,
+      "learning_rate": 1e-06,
+      "loss": 0.3772,
+      "step": 7552
+    },
+    {
+      "epoch": 0.7171052631578947,
+      "grad_norm": 6.0094742767685085,
+      "learning_rate": 1e-06,
+      "loss": 0.3682,
+      "step": 7553
+    },
+    {
+      "epoch": 0.7192982456140351,
+      "grad_norm": 6.6203456482184775,
+      "learning_rate": 1e-06,
+      "loss": 0.3514,
+      "step": 7554
+    },
+    {
+      "epoch": 0.7214912280701754,
+      "grad_norm": 5.375396726025063,
+      "learning_rate": 1e-06,
+      "loss": 0.3765,
+      "step": 7555
+    },
+    {
+      "epoch": 0.7236842105263158,
+      "grad_norm": 9.356474404865372,
+      "learning_rate": 1e-06,
+      "loss": 0.3652,
+      "step": 7556
+    },
+    {
+      "epoch": 0.7258771929824561,
+      "grad_norm": 5.941202677703259,
+      "learning_rate": 1e-06,
+      "loss": 0.3689,
+      "step": 7557
+    },
+    {
+      "epoch": 0.7280701754385965,
+      "grad_norm": 6.28472753089701,
+      "learning_rate": 1e-06,
+      "loss": 0.3243,
+      "step": 7558
+    },
+    {
+      "epoch": 0.7302631578947368,
+      "grad_norm": 6.667812477866891,
+      "learning_rate": 1e-06,
+      "loss": 0.3582,
+      "step": 7559
+    },
+    {
+      "epoch": 0.7324561403508771,
+      "grad_norm": 6.449524463786067,
+      "learning_rate": 1e-06,
+      "loss": 0.3614,
+      "step": 7560
+    },
+    {
+      "epoch": 0.7346491228070176,
+      "grad_norm": 7.575519880207911,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 7561
+    },
+    {
+      "epoch": 0.7368421052631579,
+      "grad_norm": 5.582331777559372,
+      "learning_rate": 1e-06,
+      "loss": 0.3645,
+      "step": 7562
+    },
+    {
+      "epoch": 0.7390350877192983,
+      "grad_norm": 6.384240198136722,
+      "learning_rate": 1e-06,
+      "loss": 0.3744,
+      "step": 7563
+    },
+    {
+      "epoch": 0.7412280701754386,
+      "grad_norm": 8.051516323598982,
+      "learning_rate": 1e-06,
+      "loss": 0.3477,
+      "step": 7564
+    },
+    {
+      "epoch": 0.743421052631579,
+      "grad_norm": 8.429363439910386,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 7565
+    },
+    {
+      "epoch": 0.7456140350877193,
+      "grad_norm": 7.161593769599014,
+      "learning_rate": 1e-06,
+      "loss": 0.4042,
+      "step": 7566
+    },
+    {
+      "epoch": 0.7478070175438597,
+      "grad_norm": 6.762003935046035,
+      "learning_rate": 1e-06,
+      "loss": 0.371,
+      "step": 7567
+    },
+    {
+      "epoch": 0.75,
+      "grad_norm": 6.548231454310887,
+      "learning_rate": 1e-06,
+      "loss": 0.3581,
+      "step": 7568
+    },
+    {
+      "epoch": 0.7521929824561403,
+      "grad_norm": 5.67768861435938,
+      "learning_rate": 1e-06,
+      "loss": 0.39,
+      "step": 7569
+    },
+    {
+      "epoch": 0.7543859649122807,
+      "grad_norm": 4.712836747996818,
+      "learning_rate": 1e-06,
+      "loss": 0.3582,
+      "step": 7570
+    },
+    {
+      "epoch": 0.756578947368421,
+      "grad_norm": 6.780178181457338,
+      "learning_rate": 1e-06,
+      "loss": 0.3417,
+      "step": 7571
+    },
+    {
+      "epoch": 0.7587719298245614,
+      "grad_norm": 9.126249366260204,
+      "learning_rate": 1e-06,
+      "loss": 0.3556,
+      "step": 7572
+    },
+    {
+      "epoch": 0.7609649122807017,
+      "grad_norm": 7.689066423610702,
+      "learning_rate": 1e-06,
+      "loss": 0.3852,
+      "step": 7573
+    },
+    {
+      "epoch": 0.7631578947368421,
+      "grad_norm": 7.9252512368103565,
+      "learning_rate": 1e-06,
+      "loss": 0.363,
+      "step": 7574
+    },
+    {
+      "epoch": 0.7653508771929824,
+      "grad_norm": 7.005169256686047,
+      "learning_rate": 1e-06,
+      "loss": 0.3668,
+      "step": 7575
+    },
+    {
+      "epoch": 0.7675438596491229,
+      "grad_norm": 6.078403915173341,
+      "learning_rate": 1e-06,
+      "loss": 0.3813,
+      "step": 7576
+    },
+    {
+      "epoch": 0.7697368421052632,
+      "grad_norm": 6.677728225838402,
+      "learning_rate": 1e-06,
+      "loss": 0.3597,
+      "step": 7577
+    },
+    {
+      "epoch": 0.7719298245614035,
+      "grad_norm": 7.430367058620288,
+      "learning_rate": 1e-06,
+      "loss": 0.3559,
+      "step": 7578
+    },
+    {
+      "epoch": 0.7741228070175439,
+      "grad_norm": 7.605551846556423,
+      "learning_rate": 1e-06,
+      "loss": 0.3655,
+      "step": 7579
+    },
+    {
+      "epoch": 0.7763157894736842,
+      "grad_norm": 6.447980784804229,
+      "learning_rate": 1e-06,
+      "loss": 0.3398,
+      "step": 7580
+    },
+    {
+      "epoch": 0.7785087719298246,
+      "grad_norm": 8.344984629008593,
+      "learning_rate": 1e-06,
+      "loss": 0.3904,
+      "step": 7581
+    },
+    {
+      "epoch": 0.7807017543859649,
+      "grad_norm": 6.851512865088732,
+      "learning_rate": 1e-06,
+      "loss": 0.3555,
+      "step": 7582
+    },
+    {
+      "epoch": 0.7828947368421053,
+      "grad_norm": 6.970628796751602,
+      "learning_rate": 1e-06,
+      "loss": 0.3328,
+      "step": 7583
+    },
+    {
+      "epoch": 0.7850877192982456,
+      "grad_norm": 6.555640410760705,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 7584
+    },
+    {
+      "epoch": 0.7872807017543859,
+      "grad_norm": 7.019781129542681,
+      "learning_rate": 1e-06,
+      "loss": 0.3592,
+      "step": 7585
+    },
+    {
+      "epoch": 0.7894736842105263,
+      "grad_norm": 6.545585974962807,
+      "learning_rate": 1e-06,
+      "loss": 0.3508,
+      "step": 7586
+    },
+    {
+      "epoch": 0.7916666666666666,
+      "grad_norm": 7.230106005355199,
+      "learning_rate": 1e-06,
+      "loss": 0.3445,
+      "step": 7587
+    },
+    {
+      "epoch": 0.793859649122807,
+      "grad_norm": 5.773939398432959,
+      "learning_rate": 1e-06,
+      "loss": 0.3448,
+      "step": 7588
+    },
+    {
+      "epoch": 0.7960526315789473,
+      "grad_norm": 6.3280982018053455,
+      "learning_rate": 1e-06,
+      "loss": 0.3672,
+      "step": 7589
+    },
+    {
+      "epoch": 0.7982456140350878,
+      "grad_norm": 8.525073784717277,
+      "learning_rate": 1e-06,
+      "loss": 0.398,
+      "step": 7590
+    },
+    {
+      "epoch": 0.8004385964912281,
+      "grad_norm": 6.839783048688898,
+      "learning_rate": 1e-06,
+      "loss": 0.3674,
+      "step": 7591
+    },
+    {
+      "epoch": 0.8026315789473685,
+      "grad_norm": 5.371358860333069,
+      "learning_rate": 1e-06,
+      "loss": 0.3741,
+      "step": 7592
+    },
+    {
+      "epoch": 0.8048245614035088,
+      "grad_norm": 8.249444905507522,
+      "learning_rate": 1e-06,
+      "loss": 0.3866,
+      "step": 7593
+    },
+    {
+      "epoch": 0.8070175438596491,
+      "grad_norm": 5.377098968139984,
+      "learning_rate": 1e-06,
+      "loss": 0.3715,
+      "step": 7594
+    },
+    {
+      "epoch": 0.8092105263157895,
+      "grad_norm": 6.42901072402725,
+      "learning_rate": 1e-06,
+      "loss": 0.3475,
+      "step": 7595
+    },
+    {
+      "epoch": 0.8114035087719298,
+      "grad_norm": 7.796593094228552,
+      "learning_rate": 1e-06,
+      "loss": 0.3778,
+      "step": 7596
+    },
+    {
+      "epoch": 0.8135964912280702,
+      "grad_norm": 6.670640993675697,
+      "learning_rate": 1e-06,
+      "loss": 0.3588,
+      "step": 7597
+    },
+    {
+      "epoch": 0.8157894736842105,
+      "grad_norm": 7.581704988865563,
+      "learning_rate": 1e-06,
+      "loss": 0.3642,
+      "step": 7598
+    },
+    {
+      "epoch": 0.8179824561403509,
+      "grad_norm": 7.030675022653214,
+      "learning_rate": 1e-06,
+      "loss": 0.3618,
+      "step": 7599
+    },
+    {
+      "epoch": 0.8201754385964912,
+      "grad_norm": 5.670892489226362,
+      "learning_rate": 1e-06,
+      "loss": 0.3671,
+      "step": 7600
+    },
+    {
+      "epoch": 0.8223684210526315,
+      "grad_norm": 6.592786380727288,
+      "learning_rate": 1e-06,
+      "loss": 0.3663,
+      "step": 7601
+    },
+    {
+      "epoch": 0.8245614035087719,
+      "grad_norm": 7.987972689616043,
+      "learning_rate": 1e-06,
+      "loss": 0.3759,
+      "step": 7602
+    },
+    {
+      "epoch": 0.8267543859649122,
+      "grad_norm": 7.018007764522847,
+      "learning_rate": 1e-06,
+      "loss": 0.3499,
+      "step": 7603
+    },
+    {
+      "epoch": 0.8289473684210527,
+      "grad_norm": 5.58739962780794,
+      "learning_rate": 1e-06,
+      "loss": 0.3483,
+      "step": 7604
+    },
+    {
+      "epoch": 0.831140350877193,
+      "grad_norm": 5.901577979393532,
+      "learning_rate": 1e-06,
+      "loss": 0.3639,
+      "step": 7605
+    },
+    {
+      "epoch": 0.8333333333333334,
+      "grad_norm": 6.5751496538999445,
+      "learning_rate": 1e-06,
+      "loss": 0.3238,
+      "step": 7606
+    },
+    {
+      "epoch": 0.8355263157894737,
+      "grad_norm": 6.97493837461438,
+      "learning_rate": 1e-06,
+      "loss": 0.354,
+      "step": 7607
+    },
+    {
+      "epoch": 0.8377192982456141,
+      "grad_norm": 6.879303577607576,
+      "learning_rate": 1e-06,
+      "loss": 0.374,
+      "step": 7608
+    },
+    {
+      "epoch": 0.8399122807017544,
+      "grad_norm": 8.219890713313136,
+      "learning_rate": 1e-06,
+      "loss": 0.3701,
+      "step": 7609
+    },
+    {
+      "epoch": 0.8421052631578947,
+      "grad_norm": 8.161212368243941,
+      "learning_rate": 1e-06,
+      "loss": 0.3803,
+      "step": 7610
+    },
+    {
+      "epoch": 0.8442982456140351,
+      "grad_norm": 6.163887907751214,
+      "learning_rate": 1e-06,
+      "loss": 0.364,
+      "step": 7611
+    },
+    {
+      "epoch": 0.8464912280701754,
+      "grad_norm": 6.47432708438602,
+      "learning_rate": 1e-06,
+      "loss": 0.3479,
+      "step": 7612
+    },
+    {
+      "epoch": 0.8486842105263158,
+      "grad_norm": 9.115113007334367,
+      "learning_rate": 1e-06,
+      "loss": 0.3519,
+      "step": 7613
+    },
+    {
+      "epoch": 0.8508771929824561,
+      "grad_norm": 7.805145647820362,
+      "learning_rate": 1e-06,
+      "loss": 0.3871,
+      "step": 7614
+    },
+    {
+      "epoch": 0.8530701754385965,
+      "grad_norm": 7.703560668798051,
+      "learning_rate": 1e-06,
+      "loss": 0.381,
+      "step": 7615
+    },
+    {
+      "epoch": 0.8552631578947368,
+      "grad_norm": 7.6084600454560185,
+      "learning_rate": 1e-06,
+      "loss": 0.3864,
+      "step": 7616
+    },
+    {
+      "epoch": 0.8574561403508771,
+      "grad_norm": 6.4616913541086065,
+      "learning_rate": 1e-06,
+      "loss": 0.3712,
+      "step": 7617
+    },
+    {
+      "epoch": 0.8596491228070176,
+      "grad_norm": 8.168489325993866,
+      "learning_rate": 1e-06,
+      "loss": 0.3786,
+      "step": 7618
+    },
+    {
+      "epoch": 0.8618421052631579,
+      "grad_norm": 5.9533731281242925,
+      "learning_rate": 1e-06,
+      "loss": 0.3385,
+      "step": 7619
+    },
+    {
+      "epoch": 0.8640350877192983,
+      "grad_norm": 4.6439634595780745,
+      "learning_rate": 1e-06,
+      "loss": 0.3612,
+      "step": 7620
+    },
+    {
+      "epoch": 0.8662280701754386,
+      "grad_norm": 6.617906699785013,
+      "learning_rate": 1e-06,
+      "loss": 0.3517,
+      "step": 7621
+    },
+    {
+      "epoch": 0.868421052631579,
+      "grad_norm": 6.5760090087614715,
+      "learning_rate": 1e-06,
+      "loss": 0.3713,
+      "step": 7622
+    },
+    {
+      "epoch": 0.8706140350877193,
+      "grad_norm": 5.955674675820295,
+      "learning_rate": 1e-06,
+      "loss": 0.3568,
+      "step": 7623
+    },
+    {
+      "epoch": 0.8728070175438597,
+      "grad_norm": 6.530125560632848,
+      "learning_rate": 1e-06,
+      "loss": 0.3595,
+      "step": 7624
+    },
+    {
+      "epoch": 0.875,
+      "grad_norm": 7.15231240355239,
+      "learning_rate": 1e-06,
+      "loss": 0.3746,
+      "step": 7625
+    },
+    {
+      "epoch": 0.8771929824561403,
+      "grad_norm": 8.332344145449117,
+      "learning_rate": 1e-06,
+      "loss": 0.3482,
+      "step": 7626
+    },
+    {
+      "epoch": 0.8793859649122807,
+      "grad_norm": 5.626496308283103,
+      "learning_rate": 1e-06,
+      "loss": 0.3468,
+      "step": 7627
+    },
+    {
+      "epoch": 0.881578947368421,
+      "grad_norm": 5.626496308283103,
+      "learning_rate": 1e-06,
+      "loss": 0.3491,
+      "step": 7628
+    },
+    {
+      "epoch": 0.8837719298245614,
+      "grad_norm": 7.253836795563802,
+      "learning_rate": 1e-06,
+      "loss": 0.3978,
+      "step": 7629
+    },
+    {
+      "epoch": 0.8859649122807017,
+      "grad_norm": 5.191720998751648,
+      "learning_rate": 1e-06,
+      "loss": 0.3667,
+      "step": 7630
+    },
+    {
+      "epoch": 0.8881578947368421,
+      "grad_norm": 6.245338775537678,
+      "learning_rate": 1e-06,
+      "loss": 0.3722,
+      "step": 7631
+    },
+    {
+      "epoch": 0.8903508771929824,
+      "grad_norm": 7.4135035590717235,
+      "learning_rate": 1e-06,
+      "loss": 0.3797,
+      "step": 7632
+    },
+    {
+      "epoch": 0.8925438596491229,
+      "grad_norm": 8.517420341098365,
+      "learning_rate": 1e-06,
+      "loss": 0.342,
+      "step": 7633
+    },
+    {
+      "epoch": 0.8947368421052632,
+      "grad_norm": 7.0590603947101735,
+      "learning_rate": 1e-06,
+      "loss": 0.3745,
+      "step": 7634
+    },
+    {
+      "epoch": 0.8969298245614035,
+      "grad_norm": 7.18163888745226,
+      "learning_rate": 1e-06,
+      "loss": 0.3808,
+      "step": 7635
+    },
+    {
+      "epoch": 0.8991228070175439,
+      "grad_norm": 6.673902671347109,
+      "learning_rate": 1e-06,
+      "loss": 0.359,
+      "step": 7636
+    },
+    {
+      "epoch": 0.9013157894736842,
+      "grad_norm": 5.542360363548584,
+      "learning_rate": 1e-06,
+      "loss": 0.3569,
+      "step": 7637
+    },
+    {
+      "epoch": 0.9035087719298246,
+      "grad_norm": 6.6608832175813415,
+      "learning_rate": 1e-06,
+      "loss": 0.3793,
+      "step": 7638
+    },
+    {
+      "epoch": 0.9057017543859649,
+      "grad_norm": 7.407126514866222,
+      "learning_rate": 1e-06,
+      "loss": 0.3471,
+      "step": 7639
+    },
+    {
+      "epoch": 0.9078947368421053,
+      "grad_norm": 6.303322069041416,
+      "learning_rate": 1e-06,
+      "loss": 0.3842,
+      "step": 7640
+    },
+    {
+      "epoch": 0.9100877192982456,
+      "grad_norm": 6.968051596850506,
+      "learning_rate": 1e-06,
+      "loss": 0.3785,
+      "step": 7641
+    },
+    {
+      "epoch": 0.9122807017543859,
+      "grad_norm": 4.942785955339007,
+      "learning_rate": 1e-06,
+      "loss": 0.3593,
+      "step": 7642
+    },
+    {
+      "epoch": 0.9144736842105263,
+      "grad_norm": 5.694970916325372,
+      "learning_rate": 1e-06,
+      "loss": 0.376,
+      "step": 7643
+    },
+    {
+      "epoch": 0.9166666666666666,
+      "grad_norm": 6.4783846360358135,
+      "learning_rate": 1e-06,
+      "loss": 0.3857,
+      "step": 7644
+    },
+    {
+      "epoch": 0.918859649122807,
+      "grad_norm": 5.412923049609409,
+      "learning_rate": 1e-06,
+      "loss": 0.3677,
+      "step": 7645
+    },
+    {
+      "epoch": 0.9210526315789473,
+      "grad_norm": 8.202455355500673,
+      "learning_rate": 1e-06,
+      "loss": 0.3404,
+      "step": 7646
+    },
+    {
+      "epoch": 0.9232456140350878,
+      "grad_norm": 8.788475964989015,
+      "learning_rate": 1e-06,
+      "loss": 0.3872,
+      "step": 7647
+    },
+    {
+      "epoch": 0.9254385964912281,
+      "grad_norm": 6.295509470849562,
+      "learning_rate": 1e-06,
+      "loss": 0.3489,
+      "step": 7648
+    },
+    {
+      "epoch": 0.9276315789473685,
+      "grad_norm": 9.429803478407045,
+      "learning_rate": 1e-06,
+      "loss": 0.3696,
+      "step": 7649
+    },
+    {
+      "epoch": 0.9298245614035088,
+      "grad_norm": 6.436162644551775,
+      "learning_rate": 1e-06,
+      "loss": 0.3789,
+      "step": 7650
+    },
+    {
+      "epoch": 0.9320175438596491,
+      "grad_norm": 5.660590348399415,
+      "learning_rate": 1e-06,
+      "loss": 0.3554,
+      "step": 7651
+    },
+    {
+      "epoch": 0.9342105263157895,
+      "grad_norm": 8.778769707665234,
+      "learning_rate": 1e-06,
+      "loss": 0.3593,
+      "step": 7652
+    },
+    {
+      "epoch": 0.9364035087719298,
+      "grad_norm": 7.518408375114575,
+      "learning_rate": 1e-06,
+      "loss": 0.3597,
+      "step": 7653
+    },
+    {
+      "epoch": 0.9385964912280702,
+      "grad_norm": 6.417355430057249,
+      "learning_rate": 1e-06,
+      "loss": 0.3731,
+      "step": 7654
+    },
+    {
+      "epoch": 0.9407894736842105,
+      "grad_norm": 7.38211929038243,
+      "learning_rate": 1e-06,
+      "loss": 0.3348,
+      "step": 7655
+    },
+    {
+      "epoch": 0.9429824561403509,
+      "grad_norm": 6.306539047068658,
+      "learning_rate": 1e-06,
+      "loss": 0.3807,
+      "step": 7656
+    },
+    {
+      "epoch": 0.9451754385964912,
+      "grad_norm": 6.274386874485113,
+      "learning_rate": 1e-06,
+      "loss": 0.3636,
+      "step": 7657
+    },
+    {
+      "epoch": 0.9473684210526315,
+      "grad_norm": 6.751785771833849,
+      "learning_rate": 1e-06,
+      "loss": 0.3956,
+      "step": 7658
+    },
+    {
+      "epoch": 0.9495614035087719,
+      "grad_norm": 6.169366246709903,
+      "learning_rate": 1e-06,
+      "loss": 0.3749,
+      "step": 7659
+    },
+    {
+      "epoch": 0.9517543859649122,
+      "grad_norm": 7.725991593884762,
+      "learning_rate": 1e-06,
+      "loss": 0.3745,
+      "step": 7660
+    },
+    {
+      "epoch": 0.9539473684210527,
+      "grad_norm": 8.014099564654742,
+      "learning_rate": 1e-06,
+      "loss": 0.3612,
+      "step": 7661
+    },
+    {
+      "epoch": 0.956140350877193,
+      "grad_norm": 5.659788338049941,
+      "learning_rate": 1e-06,
+      "loss": 0.3587,
+      "step": 7662
+    },
+    {
+      "epoch": 0.9583333333333334,
+      "grad_norm": 7.129223071498234,
+      "learning_rate": 1e-06,
+      "loss": 0.3677,
+      "step": 7663
+    },
+    {
+      "epoch": 0.9605263157894737,
+      "grad_norm": 6.513993718154325,
+      "learning_rate": 1e-06,
+      "loss": 0.3691,
+      "step": 7664
+    },
+    {
+      "epoch": 0.9627192982456141,
+      "grad_norm": 7.272061812093391,
+      "learning_rate": 1e-06,
+      "loss": 0.3535,
+      "step": 7665
+    },
+    {
+      "epoch": 0.9649122807017544,
+      "grad_norm": 6.256242738467903,
+      "learning_rate": 1e-06,
+      "loss": 0.37,
+      "step": 7666
+    },
+    {
+      "epoch": 0.9671052631578947,
+      "grad_norm": 7.272507187587852,
+      "learning_rate": 1e-06,
+      "loss": 0.3678,
+      "step": 7667
+    },
+    {
+      "epoch": 0.9692982456140351,
+      "grad_norm": 5.922000675187737,
+      "learning_rate": 1e-06,
+      "loss": 0.3534,
+      "step": 7668
+    },
+    {
+      "epoch": 0.9714912280701754,
+      "grad_norm": 7.664888926930641,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 7669
+    },
+    {
+      "epoch": 0.9736842105263158,
+      "grad_norm": 6.347767178703265,
+      "learning_rate": 1e-06,
+      "loss": 0.3697,
+      "step": 7670
+    },
+    {
+      "epoch": 0.9758771929824561,
+      "grad_norm": 6.828213931818434,
+      "learning_rate": 1e-06,
+      "loss": 0.3385,
+      "step": 7671
+    },
+    {
+      "epoch": 0.9780701754385965,
+      "grad_norm": 7.224835310938079,
+      "learning_rate": 1e-06,
+      "loss": 0.4021,
+      "step": 7672
+    },
+    {
+      "epoch": 0.9802631578947368,
+      "grad_norm": 6.790830217190092,
+      "learning_rate": 1e-06,
+      "loss": 0.3761,
+      "step": 7673
+    },
+    {
+      "epoch": 0.9824561403508771,
+      "grad_norm": 4.670624314661031,
+      "learning_rate": 1e-06,
+      "loss": 0.3608,
+      "step": 7674
+    },
+    {
+      "epoch": 0.9846491228070176,
+      "grad_norm": 5.044286763979725,
+      "learning_rate": 1e-06,
+      "loss": 0.352,
+      "step": 7675
+    },
+    {
+      "epoch": 0.9868421052631579,
+      "grad_norm": 6.245179195744917,
+      "learning_rate": 1e-06,
+      "loss": 0.3829,
+      "step": 7676
+    },
+    {
+      "epoch": 0.9890350877192983,
+      "grad_norm": 5.023702517356843,
+      "learning_rate": 1e-06,
+      "loss": 0.3586,
+      "step": 7677
+    },
+    {
+      "epoch": 0.9912280701754386,
+      "grad_norm": 5.917719495004594,
+      "learning_rate": 1e-06,
+      "loss": 0.3747,
+      "step": 7678
+    },
+    {
+      "epoch": 0.993421052631579,
+      "grad_norm": 6.198873789064963,
+      "learning_rate": 1e-06,
+      "loss": 0.3544,
+      "step": 7679
+    },
+    {
+      "epoch": 0.9956140350877193,
+      "grad_norm": 8.428801939869373,
+      "learning_rate": 1e-06,
+      "loss": 0.3526,
+      "step": 7680
+    },
+    {
+      "epoch": 0.9978070175438597,
+      "grad_norm": 6.56758838675707,
+      "learning_rate": 1e-06,
+      "loss": 0.3669,
+      "step": 7681
+    },
+    {
+      "epoch": 1.0,
+      "grad_norm": 7.200610597499089,
+      "learning_rate": 1e-06,
+      "loss": 0.3376,
+      "step": 7682
+    },
+    {
+      "epoch": 1.0,
+      "eval_loss": 3.5755255222320557,
+      "eval_runtime": 21.8571,
+      "eval_samples_per_second": 45.752,
+      "eval_steps_per_second": 5.719,
+      "step": 7682
+    },
+    {
+      "epoch": 0.002296211251435132,
+      "grad_norm": 4.662707903022065,
+      "learning_rate": 1e-06,
+      "loss": 0.178,
+      "step": 7683
+    },
+    {
+      "epoch": 0.004592422502870264,
+      "grad_norm": 5.5626550118783715,
+      "learning_rate": 1e-06,
+      "loss": 0.176,
+      "step": 7684
+    },
+    {
+      "epoch": 0.006888633754305396,
+      "grad_norm": 4.92737814782817,
+      "learning_rate": 1e-06,
+      "loss": 0.1733,
+      "step": 7685
+    },
+    {
+      "epoch": 0.009184845005740528,
+      "grad_norm": 5.196438005909256,
+      "learning_rate": 1e-06,
+      "loss": 0.1589,
+      "step": 7686
+    },
+    {
+      "epoch": 0.011481056257175661,
+      "grad_norm": 7.455577261368645,
+      "learning_rate": 1e-06,
+      "loss": 0.1662,
+      "step": 7687
+    },
+    {
+      "epoch": 0.013777267508610792,
+      "grad_norm": 6.403906767332956,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 7688
+    },
+    {
+      "epoch": 0.016073478760045924,
+      "grad_norm": 7.000030820531041,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 7689
+    },
+    {
+      "epoch": 0.018369690011481057,
+      "grad_norm": 6.86515698810693,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 7690
+    },
+    {
+      "epoch": 0.02066590126291619,
+      "grad_norm": 9.803616650834549,
+      "learning_rate": 1e-06,
+      "loss": 0.1553,
+      "step": 7691
+    },
+    {
+      "epoch": 0.022962112514351322,
+      "grad_norm": 8.228365214360139,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 7692
+    },
+    {
+      "epoch": 0.02525832376578645,
+      "grad_norm": 10.96665501345685,
+      "learning_rate": 1e-06,
+      "loss": 0.1583,
+      "step": 7693
+    },
+    {
+      "epoch": 0.027554535017221583,
+      "grad_norm": 8.229274671454174,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 7694
+    },
+    {
+      "epoch": 0.029850746268656716,
+      "grad_norm": 8.944381076408545,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7695
+    },
+    {
+      "epoch": 0.03214695752009185,
+      "grad_norm": 10.53413178650815,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 7696
+    },
+    {
+      "epoch": 0.03444316877152698,
+      "grad_norm": 9.014055613455557,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 7697
+    },
+    {
+      "epoch": 0.03673938002296211,
+      "grad_norm": 10.665756234863576,
+      "learning_rate": 1e-06,
+      "loss": 0.166,
+      "step": 7698
+    },
+    {
+      "epoch": 0.03903559127439724,
+      "grad_norm": 6.172616876079602,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7699
+    },
+    {
+      "epoch": 0.04133180252583238,
+      "grad_norm": 9.297392342577469,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 7700
+    },
+    {
+      "epoch": 0.04362801377726751,
+      "grad_norm": 8.278781689516055,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 7701
+    },
+    {
+      "epoch": 0.045924225028702644,
+      "grad_norm": 6.897399698995845,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 7702
+    },
+    {
+      "epoch": 0.04822043628013777,
+      "grad_norm": 6.183560929441737,
+      "learning_rate": 1e-06,
+      "loss": 0.1587,
+      "step": 7703
+    },
+    {
+      "epoch": 0.0505166475315729,
+      "grad_norm": 7.751960048852055,
+      "learning_rate": 1e-06,
+      "loss": 0.1629,
+      "step": 7704
+    },
+    {
+      "epoch": 0.05281285878300804,
+      "grad_norm": 8.11856564979211,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 7705
+    },
+    {
+      "epoch": 0.05510907003444317,
+      "grad_norm": 7.175803455956716,
+      "learning_rate": 1e-06,
+      "loss": 0.141,
+      "step": 7706
+    },
+    {
+      "epoch": 0.0574052812858783,
+      "grad_norm": 8.022622892053347,
+      "learning_rate": 1e-06,
+      "loss": 0.1603,
+      "step": 7707
+    },
+    {
+      "epoch": 0.05970149253731343,
+      "grad_norm": 4.995266637356303,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 7708
+    },
+    {
+      "epoch": 0.06199770378874857,
+      "grad_norm": 6.1360802814296225,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 7709
+    },
+    {
+      "epoch": 0.0642939150401837,
+      "grad_norm": 8.059198799460324,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 7710
+    },
+    {
+      "epoch": 0.06659012629161883,
+      "grad_norm": 8.739430689982719,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 7711
+    },
+    {
+      "epoch": 0.06888633754305395,
+      "grad_norm": 9.22080860769983,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 7712
+    },
+    {
+      "epoch": 0.0711825487944891,
+      "grad_norm": 8.306301312469964,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 7713
+    },
+    {
+      "epoch": 0.07347876004592423,
+      "grad_norm": 5.574510404103377,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 7714
+    },
+    {
+      "epoch": 0.07577497129735936,
+      "grad_norm": 8.136003414040959,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7715
+    },
+    {
+      "epoch": 0.07807118254879448,
+      "grad_norm": 5.613728071563841,
+      "learning_rate": 1e-06,
+      "loss": 0.1445,
+      "step": 7716
+    },
+    {
+      "epoch": 0.08036739380022963,
+      "grad_norm": 5.539570379159489,
+      "learning_rate": 1e-06,
+      "loss": 0.1507,
+      "step": 7717
+    },
+    {
+      "epoch": 0.08266360505166476,
+      "grad_norm": 6.363419502423012,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 7718
+    },
+    {
+      "epoch": 0.08495981630309989,
+      "grad_norm": 7.128193351147782,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 7719
+    },
+    {
+      "epoch": 0.08725602755453502,
+      "grad_norm": 7.157006633702301,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 7720
+    },
+    {
+      "epoch": 0.08955223880597014,
+      "grad_norm": 6.200888546700024,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 7721
+    },
+    {
+      "epoch": 0.09184845005740529,
+      "grad_norm": 6.442435594857567,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 7722
+    },
+    {
+      "epoch": 0.09414466130884042,
+      "grad_norm": 8.178093333852859,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 7723
+    },
+    {
+      "epoch": 0.09644087256027555,
+      "grad_norm": 7.67582492947132,
+      "learning_rate": 1e-06,
+      "loss": 0.1633,
+      "step": 7724
+    },
+    {
+      "epoch": 0.09873708381171067,
+      "grad_norm": 6.705987326604643,
+      "learning_rate": 1e-06,
+      "loss": 0.1672,
+      "step": 7725
+    },
+    {
+      "epoch": 0.1010332950631458,
+      "grad_norm": 5.8397061196187705,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 7726
+    },
+    {
+      "epoch": 0.10332950631458095,
+      "grad_norm": 7.725270959205483,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 7727
+    },
+    {
+      "epoch": 0.10562571756601608,
+      "grad_norm": 8.749830114017131,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 7728
+    },
+    {
+      "epoch": 0.1079219288174512,
+      "grad_norm": 6.974707202677967,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7729
+    },
+    {
+      "epoch": 0.11021814006888633,
+      "grad_norm": 7.466797821120385,
+      "learning_rate": 1e-06,
+      "loss": 0.1512,
+      "step": 7730
+    },
+    {
+      "epoch": 0.11251435132032148,
+      "grad_norm": 7.884730700387932,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7731
+    },
+    {
+      "epoch": 0.1148105625717566,
+      "grad_norm": 7.6658291592365035,
+      "learning_rate": 1e-06,
+      "loss": 0.1516,
+      "step": 7732
+    },
+    {
+      "epoch": 0.11710677382319173,
+      "grad_norm": 7.026946183277749,
+      "learning_rate": 1e-06,
+      "loss": 0.1538,
+      "step": 7733
+    },
+    {
+      "epoch": 0.11940298507462686,
+      "grad_norm": 7.2255697318658525,
+      "learning_rate": 1e-06,
+      "loss": 0.1603,
+      "step": 7734
+    },
+    {
+      "epoch": 0.12169919632606199,
+      "grad_norm": 5.959540919151646,
+      "learning_rate": 1e-06,
+      "loss": 0.1366,
+      "step": 7735
+    },
+    {
+      "epoch": 0.12399540757749714,
+      "grad_norm": 5.385159737329753,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 7736
+    },
+    {
+      "epoch": 0.12629161882893225,
+      "grad_norm": 4.927228613485907,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 7737
+    },
+    {
+      "epoch": 0.1285878300803674,
+      "grad_norm": 9.384764225252107,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 7738
+    },
+    {
+      "epoch": 0.13088404133180254,
+      "grad_norm": 6.418079346408692,
+      "learning_rate": 1e-06,
+      "loss": 0.1472,
+      "step": 7739
+    },
+    {
+      "epoch": 0.13318025258323765,
+      "grad_norm": 5.933409981512082,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 7740
+    },
+    {
+      "epoch": 0.1354764638346728,
+      "grad_norm": 7.616717108414114,
+      "learning_rate": 1e-06,
+      "loss": 0.1547,
+      "step": 7741
+    },
+    {
+      "epoch": 0.1377726750861079,
+      "grad_norm": 5.738154276267275,
+      "learning_rate": 1e-06,
+      "loss": 0.1527,
+      "step": 7742
+    },
+    {
+      "epoch": 0.14006888633754305,
+      "grad_norm": 6.745030769410637,
+      "learning_rate": 1e-06,
+      "loss": 0.1515,
+      "step": 7743
+    },
+    {
+      "epoch": 0.1423650975889782,
+      "grad_norm": 6.1677766148134925,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 7744
+    },
+    {
+      "epoch": 0.1446613088404133,
+      "grad_norm": 7.935945600999973,
+      "learning_rate": 1e-06,
+      "loss": 0.1377,
+      "step": 7745
+    },
+    {
+      "epoch": 0.14695752009184845,
+      "grad_norm": 7.244225385651613,
+      "learning_rate": 1e-06,
+      "loss": 0.1618,
+      "step": 7746
+    },
+    {
+      "epoch": 0.14925373134328357,
+      "grad_norm": 8.415894516260288,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 7747
+    },
+    {
+      "epoch": 0.1515499425947187,
+      "grad_norm": 6.311517116822205,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 7748
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 8.131212864266372,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 7749
+    },
+    {
+      "epoch": 0.15614236509758897,
+      "grad_norm": 7.303030729689006,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 7750
+    },
+    {
+      "epoch": 0.1584385763490241,
+      "grad_norm": 7.238689181724591,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 7751
+    },
+    {
+      "epoch": 0.16073478760045926,
+      "grad_norm": 8.851319182315272,
+      "learning_rate": 1e-06,
+      "loss": 0.1623,
+      "step": 7752
+    },
+    {
+      "epoch": 0.16303099885189437,
+      "grad_norm": 7.426883531198577,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 7753
+    },
+    {
+      "epoch": 0.1653272101033295,
+      "grad_norm": 6.7156792267805185,
+      "learning_rate": 1e-06,
+      "loss": 0.1438,
+      "step": 7754
+    },
+    {
+      "epoch": 0.16762342135476463,
+      "grad_norm": 5.99499246264355,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 7755
+    },
+    {
+      "epoch": 0.16991963260619977,
+      "grad_norm": 6.6624371973734755,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 7756
+    },
+    {
+      "epoch": 0.17221584385763491,
+      "grad_norm": 6.869878445864526,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 7757
+    },
+    {
+      "epoch": 0.17451205510907003,
+      "grad_norm": 6.777685442739428,
+      "learning_rate": 1e-06,
+      "loss": 0.1463,
+      "step": 7758
+    },
+    {
+      "epoch": 0.17680826636050517,
+      "grad_norm": 9.650622120294324,
+      "learning_rate": 1e-06,
+      "loss": 0.1559,
+      "step": 7759
+    },
+    {
+      "epoch": 0.1791044776119403,
+      "grad_norm": 7.128496248460968,
+      "learning_rate": 1e-06,
+      "loss": 0.1554,
+      "step": 7760
+    },
+    {
+      "epoch": 0.18140068886337543,
+      "grad_norm": 8.561917327233285,
+      "learning_rate": 1e-06,
+      "loss": 0.1619,
+      "step": 7761
+    },
+    {
+      "epoch": 0.18369690011481057,
+      "grad_norm": 6.121383111710097,
+      "learning_rate": 1e-06,
+      "loss": 0.161,
+      "step": 7762
+    },
+    {
+      "epoch": 0.1859931113662457,
+      "grad_norm": 7.074668291539274,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 7763
+    },
+    {
+      "epoch": 0.18828932261768083,
+      "grad_norm": 5.661806008303572,
+      "learning_rate": 1e-06,
+      "loss": 0.1454,
+      "step": 7764
+    },
+    {
+      "epoch": 0.19058553386911595,
+      "grad_norm": 7.920963662937405,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 7765
+    },
+    {
+      "epoch": 0.1928817451205511,
+      "grad_norm": 8.832369797946468,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 7766
+    },
+    {
+      "epoch": 0.19517795637198623,
+      "grad_norm": 8.102431359143639,
+      "learning_rate": 1e-06,
+      "loss": 0.159,
+      "step": 7767
+    },
+    {
+      "epoch": 0.19747416762342135,
+      "grad_norm": 7.716257465038426,
+      "learning_rate": 1e-06,
+      "loss": 0.1499,
+      "step": 7768
+    },
+    {
+      "epoch": 0.1997703788748565,
+      "grad_norm": 6.550928399633227,
+      "learning_rate": 1e-06,
+      "loss": 0.1661,
+      "step": 7769
+    },
+    {
+      "epoch": 0.2020665901262916,
+      "grad_norm": 7.818930714942751,
+      "learning_rate": 1e-06,
+      "loss": 0.154,
+      "step": 7770
+    },
+    {
+      "epoch": 0.20436280137772675,
+      "grad_norm": 7.434046520076304,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 7771
+    },
+    {
+      "epoch": 0.2066590126291619,
+      "grad_norm": 9.22303705232602,
+      "learning_rate": 1e-06,
+      "loss": 0.1485,
+      "step": 7772
+    },
+    {
+      "epoch": 0.208955223880597,
+      "grad_norm": 5.279068761789632,
+      "learning_rate": 1e-06,
+      "loss": 0.1614,
+      "step": 7773
+    },
+    {
+      "epoch": 0.21125143513203215,
+      "grad_norm": 5.222217728799955,
+      "learning_rate": 1e-06,
+      "loss": 0.1658,
+      "step": 7774
+    },
+    {
+      "epoch": 0.21354764638346727,
+      "grad_norm": 7.027670435017313,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 7775
+    },
+    {
+      "epoch": 0.2158438576349024,
+      "grad_norm": 7.003243371948342,
+      "learning_rate": 1e-06,
+      "loss": 0.1608,
+      "step": 7776
+    },
+    {
+      "epoch": 0.21814006888633755,
+      "grad_norm": 7.667508556002907,
+      "learning_rate": 1e-06,
+      "loss": 0.1511,
+      "step": 7777
+    },
+    {
+      "epoch": 0.22043628013777267,
+      "grad_norm": 8.269209326156695,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 7778
+    },
+    {
+      "epoch": 0.2227324913892078,
+      "grad_norm": 6.751838095917989,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 7779
+    },
+    {
+      "epoch": 0.22502870264064295,
+      "grad_norm": 6.049312156811157,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 7780
+    },
+    {
+      "epoch": 0.22732491389207807,
+      "grad_norm": 7.399570853480395,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7781
+    },
+    {
+      "epoch": 0.2296211251435132,
+      "grad_norm": 8.366411364883726,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 7782
+    },
+    {
+      "epoch": 0.23191733639494833,
+      "grad_norm": 9.209725056899812,
+      "learning_rate": 1e-06,
+      "loss": 0.159,
+      "step": 7783
+    },
+    {
+      "epoch": 0.23421354764638347,
+      "grad_norm": 6.7816853869146465,
+      "learning_rate": 1e-06,
+      "loss": 0.1491,
+      "step": 7784
+    },
+    {
+      "epoch": 0.2365097588978186,
+      "grad_norm": 7.143470451274866,
+      "learning_rate": 1e-06,
+      "loss": 0.1368,
+      "step": 7785
+    },
+    {
+      "epoch": 0.23880597014925373,
+      "grad_norm": 7.416597928027529,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 7786
+    },
+    {
+      "epoch": 0.24110218140068887,
+      "grad_norm": 8.653591719530748,
+      "learning_rate": 1e-06,
+      "loss": 0.1566,
+      "step": 7787
+    },
+    {
+      "epoch": 0.24339839265212398,
+      "grad_norm": 8.60725937812581,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 7788
+    },
+    {
+      "epoch": 0.24569460390355913,
+      "grad_norm": 7.326581208116097,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 7789
+    },
+    {
+      "epoch": 0.24799081515499427,
+      "grad_norm": 7.6096145791787855,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 7790
+    },
+    {
+      "epoch": 0.2502870264064294,
+      "grad_norm": 8.13931056739048,
+      "learning_rate": 1e-06,
+      "loss": 0.1555,
+      "step": 7791
+    },
+    {
+      "epoch": 0.2525832376578645,
+      "grad_norm": 7.442519562686174,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 7792
+    },
+    {
+      "epoch": 0.25487944890929964,
+      "grad_norm": 7.602800853279926,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 7793
+    },
+    {
+      "epoch": 0.2571756601607348,
+      "grad_norm": 7.839048643116093,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 7794
+    },
+    {
+      "epoch": 0.25947187141216993,
+      "grad_norm": 7.110286453439429,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7795
+    },
+    {
+      "epoch": 0.2617680826636051,
+      "grad_norm": 8.970765076242085,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 7796
+    },
+    {
+      "epoch": 0.26406429391504016,
+      "grad_norm": 6.546329699472341,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 7797
+    },
+    {
+      "epoch": 0.2663605051664753,
+      "grad_norm": 6.466263796934826,
+      "learning_rate": 1e-06,
+      "loss": 0.1537,
+      "step": 7798
+    },
+    {
+      "epoch": 0.26865671641791045,
+      "grad_norm": 7.0819334907052,
+      "learning_rate": 1e-06,
+      "loss": 0.1515,
+      "step": 7799
+    },
+    {
+      "epoch": 0.2709529276693456,
+      "grad_norm": 7.2165005439648935,
+      "learning_rate": 1e-06,
+      "loss": 0.1423,
+      "step": 7800
+    },
+    {
+      "epoch": 0.27324913892078073,
+      "grad_norm": 7.4980596823488534,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 7801
+    },
+    {
+      "epoch": 0.2755453501722158,
+      "grad_norm": 5.587140113996311,
+      "learning_rate": 1e-06,
+      "loss": 0.1546,
+      "step": 7802
+    },
+    {
+      "epoch": 0.27784156142365096,
+      "grad_norm": 5.616375699950889,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 7803
+    },
+    {
+      "epoch": 0.2801377726750861,
+      "grad_norm": 7.618098411440994,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 7804
+    },
+    {
+      "epoch": 0.28243398392652125,
+      "grad_norm": 7.57086369267822,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 7805
+    },
+    {
+      "epoch": 0.2847301951779564,
+      "grad_norm": 7.626523692735426,
+      "learning_rate": 1e-06,
+      "loss": 0.1421,
+      "step": 7806
+    },
+    {
+      "epoch": 0.2870264064293915,
+      "grad_norm": 6.280655416138292,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 7807
+    },
+    {
+      "epoch": 0.2893226176808266,
+      "grad_norm": 7.502191640763075,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7808
+    },
+    {
+      "epoch": 0.29161882893226176,
+      "grad_norm": 9.425448082368991,
+      "learning_rate": 1e-06,
+      "loss": 0.1665,
+      "step": 7809
+    },
+    {
+      "epoch": 0.2939150401836969,
+      "grad_norm": 7.106855396841733,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7810
+    },
+    {
+      "epoch": 0.29621125143513205,
+      "grad_norm": 6.042773113735745,
+      "learning_rate": 1e-06,
+      "loss": 0.1564,
+      "step": 7811
+    },
+    {
+      "epoch": 0.29850746268656714,
+      "grad_norm": 8.528471207738923,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 7812
+    },
+    {
+      "epoch": 0.3008036739380023,
+      "grad_norm": 8.28478612727414,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7813
+    },
+    {
+      "epoch": 0.3030998851894374,
+      "grad_norm": 7.749014086851393,
+      "learning_rate": 1e-06,
+      "loss": 0.1437,
+      "step": 7814
+    },
+    {
+      "epoch": 0.30539609644087257,
+      "grad_norm": 6.316492939513686,
+      "learning_rate": 1e-06,
+      "loss": 0.1442,
+      "step": 7815
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 8.138046924112139,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 7816
+    },
+    {
+      "epoch": 0.3099885189437428,
+      "grad_norm": 6.405992056715333,
+      "learning_rate": 1e-06,
+      "loss": 0.153,
+      "step": 7817
+    },
+    {
+      "epoch": 0.31228473019517794,
+      "grad_norm": 5.6537544274826095,
+      "learning_rate": 1e-06,
+      "loss": 0.1502,
+      "step": 7818
+    },
+    {
+      "epoch": 0.3145809414466131,
+      "grad_norm": 9.847904487874944,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 7819
+    },
+    {
+      "epoch": 0.3168771526980482,
+      "grad_norm": 6.430799206343351,
+      "learning_rate": 1e-06,
+      "loss": 0.1476,
+      "step": 7820
+    },
+    {
+      "epoch": 0.31917336394948337,
+      "grad_norm": 8.125806579965495,
+      "learning_rate": 1e-06,
+      "loss": 0.1588,
+      "step": 7821
+    },
+    {
+      "epoch": 0.3214695752009185,
+      "grad_norm": 8.11960376495829,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 7822
+    },
+    {
+      "epoch": 0.3237657864523536,
+      "grad_norm": 7.0211202715237455,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 7823
+    },
+    {
+      "epoch": 0.32606199770378874,
+      "grad_norm": 7.068871953493892,
+      "learning_rate": 1e-06,
+      "loss": 0.1431,
+      "step": 7824
+    },
+    {
+      "epoch": 0.3283582089552239,
+      "grad_norm": 8.275376242880494,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 7825
+    },
+    {
+      "epoch": 0.330654420206659,
+      "grad_norm": 6.740494703440385,
+      "learning_rate": 1e-06,
+      "loss": 0.1453,
+      "step": 7826
+    },
+    {
+      "epoch": 0.33295063145809417,
+      "grad_norm": 6.718282761829523,
+      "learning_rate": 1e-06,
+      "loss": 0.1623,
+      "step": 7827
+    },
+    {
+      "epoch": 0.33524684270952926,
+      "grad_norm": 6.687311235953894,
+      "learning_rate": 1e-06,
+      "loss": 0.17,
+      "step": 7828
+    },
+    {
+      "epoch": 0.3375430539609644,
+      "grad_norm": 8.252647186271169,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 7829
+    },
+    {
+      "epoch": 0.33983926521239954,
+      "grad_norm": 9.295459426795992,
+      "learning_rate": 1e-06,
+      "loss": 0.1626,
+      "step": 7830
+    },
+    {
+      "epoch": 0.3421354764638347,
+      "grad_norm": 8.587353188563327,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 7831
+    },
+    {
+      "epoch": 0.34443168771526983,
+      "grad_norm": 5.589119341224708,
+      "learning_rate": 1e-06,
+      "loss": 0.1419,
+      "step": 7832
+    },
+    {
+      "epoch": 0.3467278989667049,
+      "grad_norm": 8.099979312560421,
+      "learning_rate": 1e-06,
+      "loss": 0.1515,
+      "step": 7833
+    },
+    {
+      "epoch": 0.34902411021814006,
+      "grad_norm": 6.480680434908623,
+      "learning_rate": 1e-06,
+      "loss": 0.1688,
+      "step": 7834
+    },
+    {
+      "epoch": 0.3513203214695752,
+      "grad_norm": 6.663261698473068,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 7835
+    },
+    {
+      "epoch": 0.35361653272101035,
+      "grad_norm": 9.88837767724838,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 7836
+    },
+    {
+      "epoch": 0.3559127439724455,
+      "grad_norm": 8.160158620090332,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 7837
+    },
+    {
+      "epoch": 0.3582089552238806,
+      "grad_norm": 6.476229524795883,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 7838
+    },
+    {
+      "epoch": 0.3605051664753157,
+      "grad_norm": 7.412269321152703,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 7839
+    },
+    {
+      "epoch": 0.36280137772675086,
+      "grad_norm": 6.84090848861817,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 7840
+    },
+    {
+      "epoch": 0.365097588978186,
+      "grad_norm": 8.25848692042268,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 7841
+    },
+    {
+      "epoch": 0.36739380022962115,
+      "grad_norm": 7.088276176512312,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 7842
+    },
+    {
+      "epoch": 0.36969001148105624,
+      "grad_norm": 8.948145331158361,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 7843
+    },
+    {
+      "epoch": 0.3719862227324914,
+      "grad_norm": 5.850932495731783,
+      "learning_rate": 1e-06,
+      "loss": 0.143,
+      "step": 7844
+    },
+    {
+      "epoch": 0.3742824339839265,
+      "grad_norm": 8.300458555723052,
+      "learning_rate": 1e-06,
+      "loss": 0.1514,
+      "step": 7845
+    },
+    {
+      "epoch": 0.37657864523536166,
+      "grad_norm": 6.4708862868513854,
+      "learning_rate": 1e-06,
+      "loss": 0.1565,
+      "step": 7846
+    },
+    {
+      "epoch": 0.3788748564867968,
+      "grad_norm": 7.7487224521530225,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 7847
+    },
+    {
+      "epoch": 0.3811710677382319,
+      "grad_norm": 6.906940692851268,
+      "learning_rate": 1e-06,
+      "loss": 0.1534,
+      "step": 7848
+    },
+    {
+      "epoch": 0.38346727898966704,
+      "grad_norm": 7.487098676654583,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 7849
+    },
+    {
+      "epoch": 0.3857634902411022,
+      "grad_norm": 9.92227989535963,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 7850
+    },
+    {
+      "epoch": 0.3880597014925373,
+      "grad_norm": 8.107290395880428,
+      "learning_rate": 1e-06,
+      "loss": 0.1519,
+      "step": 7851
+    },
+    {
+      "epoch": 0.39035591274397247,
+      "grad_norm": 6.279738323740598,
+      "learning_rate": 1e-06,
+      "loss": 0.1506,
+      "step": 7852
+    },
+    {
+      "epoch": 0.39265212399540755,
+      "grad_norm": 6.633417942878358,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7853
+    },
+    {
+      "epoch": 0.3949483352468427,
+      "grad_norm": 5.705810208828845,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 7854
+    },
+    {
+      "epoch": 0.39724454649827784,
+      "grad_norm": 5.578864743139153,
+      "learning_rate": 1e-06,
+      "loss": 0.1422,
+      "step": 7855
+    },
+    {
+      "epoch": 0.399540757749713,
+      "grad_norm": 6.757720470146357,
+      "learning_rate": 1e-06,
+      "loss": 0.1513,
+      "step": 7856
+    },
+    {
+      "epoch": 0.4018369690011481,
+      "grad_norm": 6.916180680091545,
+      "learning_rate": 1e-06,
+      "loss": 0.1621,
+      "step": 7857
+    },
+    {
+      "epoch": 0.4041331802525832,
+      "grad_norm": 7.548208183139974,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 7858
+    },
+    {
+      "epoch": 0.40642939150401836,
+      "grad_norm": 7.302769872357102,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 7859
+    },
+    {
+      "epoch": 0.4087256027554535,
+      "grad_norm": 7.429576246846864,
+      "learning_rate": 1e-06,
+      "loss": 0.1509,
+      "step": 7860
+    },
+    {
+      "epoch": 0.41102181400688864,
+      "grad_norm": 6.934177387667251,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 7861
+    },
+    {
+      "epoch": 0.4133180252583238,
+      "grad_norm": 8.110702333316059,
+      "learning_rate": 1e-06,
+      "loss": 0.1677,
+      "step": 7862
+    },
+    {
+      "epoch": 0.41561423650975887,
+      "grad_norm": 8.640178310107437,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 7863
+    },
+    {
+      "epoch": 0.417910447761194,
+      "grad_norm": 8.453362879045006,
+      "learning_rate": 1e-06,
+      "loss": 0.1459,
+      "step": 7864
+    },
+    {
+      "epoch": 0.42020665901262916,
+      "grad_norm": 5.309284935014146,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 7865
+    },
+    {
+      "epoch": 0.4225028702640643,
+      "grad_norm": 5.9616226668207135,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 7866
+    },
+    {
+      "epoch": 0.42479908151549944,
+      "grad_norm": 6.9369687618134,
+      "learning_rate": 1e-06,
+      "loss": 0.1438,
+      "step": 7867
+    },
+    {
+      "epoch": 0.42709529276693453,
+      "grad_norm": 7.390047198104045,
+      "learning_rate": 1e-06,
+      "loss": 0.1634,
+      "step": 7868
+    },
+    {
+      "epoch": 0.4293915040183697,
+      "grad_norm": 6.943872918916648,
+      "learning_rate": 1e-06,
+      "loss": 0.1522,
+      "step": 7869
+    },
+    {
+      "epoch": 0.4316877152698048,
+      "grad_norm": 6.316917684098999,
+      "learning_rate": 1e-06,
+      "loss": 0.1532,
+      "step": 7870
+    },
+    {
+      "epoch": 0.43398392652123996,
+      "grad_norm": 7.414920003475991,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 7871
+    },
+    {
+      "epoch": 0.4362801377726751,
+      "grad_norm": 6.498088823760817,
+      "learning_rate": 1e-06,
+      "loss": 0.1725,
+      "step": 7872
+    },
+    {
+      "epoch": 0.43857634902411025,
+      "grad_norm": 6.586170208684158,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 7873
+    },
+    {
+      "epoch": 0.44087256027554533,
+      "grad_norm": 7.102871598059924,
+      "learning_rate": 1e-06,
+      "loss": 0.1496,
+      "step": 7874
+    },
+    {
+      "epoch": 0.4431687715269805,
+      "grad_norm": 8.650817397920287,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 7875
+    },
+    {
+      "epoch": 0.4454649827784156,
+      "grad_norm": 7.806720980550269,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 7876
+    },
+    {
+      "epoch": 0.44776119402985076,
+      "grad_norm": 6.766582606166289,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 7877
+    },
+    {
+      "epoch": 0.4500574052812859,
+      "grad_norm": 7.812291652658006,
+      "learning_rate": 1e-06,
+      "loss": 0.1455,
+      "step": 7878
+    },
+    {
+      "epoch": 0.452353616532721,
+      "grad_norm": 7.348172276535951,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 7879
+    },
+    {
+      "epoch": 0.45464982778415614,
+      "grad_norm": 8.975456311023889,
+      "learning_rate": 1e-06,
+      "loss": 0.1567,
+      "step": 7880
+    },
+    {
+      "epoch": 0.4569460390355913,
+      "grad_norm": 7.359621356371232,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 7881
+    },
+    {
+      "epoch": 0.4592422502870264,
+      "grad_norm": 6.894507322374569,
+      "learning_rate": 1e-06,
+      "loss": 0.1672,
+      "step": 7882
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 6.801604401713015,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 7883
+    },
+    {
+      "epoch": 0.46383467278989665,
+      "grad_norm": 6.839037812923213,
+      "learning_rate": 1e-06,
+      "loss": 0.1469,
+      "step": 7884
+    },
+    {
+      "epoch": 0.4661308840413318,
+      "grad_norm": 6.914581394104219,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 7885
+    },
+    {
+      "epoch": 0.46842709529276694,
+      "grad_norm": 5.169819435960932,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 7886
+    },
+    {
+      "epoch": 0.4707233065442021,
+      "grad_norm": 6.472339436257362,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 7887
+    },
+    {
+      "epoch": 0.4730195177956372,
+      "grad_norm": 7.127826942528319,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 7888
+    },
+    {
+      "epoch": 0.4753157290470723,
+      "grad_norm": 5.698497403377875,
+      "learning_rate": 1e-06,
+      "loss": 0.1431,
+      "step": 7889
+    },
+    {
+      "epoch": 0.47761194029850745,
+      "grad_norm": 4.078342157732486,
+      "learning_rate": 1e-06,
+      "loss": 0.1484,
+      "step": 7890
+    },
+    {
+      "epoch": 0.4799081515499426,
+      "grad_norm": 8.042916041894408,
+      "learning_rate": 1e-06,
+      "loss": 0.1503,
+      "step": 7891
+    },
+    {
+      "epoch": 0.48220436280137774,
+      "grad_norm": 7.866072950464121,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 7892
+    },
+    {
+      "epoch": 0.4845005740528129,
+      "grad_norm": 8.658975333935173,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 7893
+    },
+    {
+      "epoch": 0.48679678530424797,
+      "grad_norm": 5.890590783494607,
+      "learning_rate": 1e-06,
+      "loss": 0.1523,
+      "step": 7894
+    },
+    {
+      "epoch": 0.4890929965556831,
+      "grad_norm": 5.993158370977292,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 7895
+    },
+    {
+      "epoch": 0.49138920780711826,
+      "grad_norm": 7.526479218820255,
+      "learning_rate": 1e-06,
+      "loss": 0.1518,
+      "step": 7896
+    },
+    {
+      "epoch": 0.4936854190585534,
+      "grad_norm": 7.153768319155287,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 7897
+    },
+    {
+      "epoch": 0.49598163030998854,
+      "grad_norm": 9.539815871922876,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 7898
+    },
+    {
+      "epoch": 0.49827784156142363,
+      "grad_norm": 6.383694288131782,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 7899
+    },
+    {
+      "epoch": 0.5005740528128588,
+      "grad_norm": 8.62803919938021,
+      "learning_rate": 1e-06,
+      "loss": 0.1546,
+      "step": 7900
+    },
+    {
+      "epoch": 0.5028702640642939,
+      "grad_norm": 8.164440442022897,
+      "learning_rate": 1e-06,
+      "loss": 0.1641,
+      "step": 7901
+    },
+    {
+      "epoch": 0.505166475315729,
+      "grad_norm": 8.448445966684385,
+      "learning_rate": 1e-06,
+      "loss": 0.1395,
+      "step": 7902
+    },
+    {
+      "epoch": 0.5074626865671642,
+      "grad_norm": 6.088299924463753,
+      "learning_rate": 1e-06,
+      "loss": 0.1671,
+      "step": 7903
+    },
+    {
+      "epoch": 0.5097588978185993,
+      "grad_norm": 5.495346391782884,
+      "learning_rate": 1e-06,
+      "loss": 0.1414,
+      "step": 7904
+    },
+    {
+      "epoch": 0.5120551090700345,
+      "grad_norm": 7.286995092589581,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 7905
+    },
+    {
+      "epoch": 0.5143513203214696,
+      "grad_norm": 7.947669487403078,
+      "learning_rate": 1e-06,
+      "loss": 0.1456,
+      "step": 7906
+    },
+    {
+      "epoch": 0.5166475315729047,
+      "grad_norm": 7.351856602975402,
+      "learning_rate": 1e-06,
+      "loss": 0.1633,
+      "step": 7907
+    },
+    {
+      "epoch": 0.5189437428243399,
+      "grad_norm": 8.125473380884628,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 7908
+    },
+    {
+      "epoch": 0.521239954075775,
+      "grad_norm": 9.695755717668492,
+      "learning_rate": 1e-06,
+      "loss": 0.1673,
+      "step": 7909
+    },
+    {
+      "epoch": 0.5235361653272101,
+      "grad_norm": 7.924682399713497,
+      "learning_rate": 1e-06,
+      "loss": 0.1521,
+      "step": 7910
+    },
+    {
+      "epoch": 0.5258323765786452,
+      "grad_norm": 8.828532906995784,
+      "learning_rate": 1e-06,
+      "loss": 0.1579,
+      "step": 7911
+    },
+    {
+      "epoch": 0.5281285878300803,
+      "grad_norm": 7.893281587138929,
+      "learning_rate": 1e-06,
+      "loss": 0.1553,
+      "step": 7912
+    },
+    {
+      "epoch": 0.5304247990815155,
+      "grad_norm": 7.368904198870412,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 7913
+    },
+    {
+      "epoch": 0.5327210103329506,
+      "grad_norm": 7.6044648700306485,
+      "learning_rate": 1e-06,
+      "loss": 0.1744,
+      "step": 7914
+    },
+    {
+      "epoch": 0.5350172215843858,
+      "grad_norm": 5.658565817329762,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7915
+    },
+    {
+      "epoch": 0.5373134328358209,
+      "grad_norm": 8.200426243160653,
+      "learning_rate": 1e-06,
+      "loss": 0.1465,
+      "step": 7916
+    },
+    {
+      "epoch": 0.539609644087256,
+      "grad_norm": 6.514727399871735,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 7917
+    },
+    {
+      "epoch": 0.5419058553386912,
+      "grad_norm": 6.209180127432922,
+      "learning_rate": 1e-06,
+      "loss": 0.1347,
+      "step": 7918
+    },
+    {
+      "epoch": 0.5442020665901263,
+      "grad_norm": 8.03369592632619,
+      "learning_rate": 1e-06,
+      "loss": 0.1717,
+      "step": 7919
+    },
+    {
+      "epoch": 0.5464982778415615,
+      "grad_norm": 8.051049315713481,
+      "learning_rate": 1e-06,
+      "loss": 0.1556,
+      "step": 7920
+    },
+    {
+      "epoch": 0.5487944890929966,
+      "grad_norm": 7.241149015517175,
+      "learning_rate": 1e-06,
+      "loss": 0.1477,
+      "step": 7921
+    },
+    {
+      "epoch": 0.5510907003444316,
+      "grad_norm": 8.664553486027406,
+      "learning_rate": 1e-06,
+      "loss": 0.1701,
+      "step": 7922
+    },
+    {
+      "epoch": 0.5533869115958668,
+      "grad_norm": 7.030463295559554,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7923
+    },
+    {
+      "epoch": 0.5556831228473019,
+      "grad_norm": 7.023796580751946,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 7924
+    },
+    {
+      "epoch": 0.5579793340987371,
+      "grad_norm": 6.6148947431247205,
+      "learning_rate": 1e-06,
+      "loss": 0.1511,
+      "step": 7925
+    },
+    {
+      "epoch": 0.5602755453501722,
+      "grad_norm": 8.192452026806459,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 7926
+    },
+    {
+      "epoch": 0.5625717566016073,
+      "grad_norm": 6.291373802226855,
+      "learning_rate": 1e-06,
+      "loss": 0.1462,
+      "step": 7927
+    },
+    {
+      "epoch": 0.5648679678530425,
+      "grad_norm": 6.596446936741643,
+      "learning_rate": 1e-06,
+      "loss": 0.1674,
+      "step": 7928
+    },
+    {
+      "epoch": 0.5671641791044776,
+      "grad_norm": 6.466243592305493,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 7929
+    },
+    {
+      "epoch": 0.5694603903559128,
+      "grad_norm": 6.572554025691939,
+      "learning_rate": 1e-06,
+      "loss": 0.1551,
+      "step": 7930
+    },
+    {
+      "epoch": 0.5717566016073479,
+      "grad_norm": 7.189838667582423,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 7931
+    },
+    {
+      "epoch": 0.574052812858783,
+      "grad_norm": 8.904862949498826,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 7932
+    },
+    {
+      "epoch": 0.5763490241102182,
+      "grad_norm": 7.5384732688618445,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7933
+    },
+    {
+      "epoch": 0.5786452353616532,
+      "grad_norm": 8.066781158800332,
+      "learning_rate": 1e-06,
+      "loss": 0.1563,
+      "step": 7934
+    },
+    {
+      "epoch": 0.5809414466130884,
+      "grad_norm": 7.820140430312395,
+      "learning_rate": 1e-06,
+      "loss": 0.1487,
+      "step": 7935
+    },
+    {
+      "epoch": 0.5832376578645235,
+      "grad_norm": 6.18284874863018,
+      "learning_rate": 1e-06,
+      "loss": 0.1554,
+      "step": 7936
+    },
+    {
+      "epoch": 0.5855338691159586,
+      "grad_norm": 6.536945722442396,
+      "learning_rate": 1e-06,
+      "loss": 0.1593,
+      "step": 7937
+    },
+    {
+      "epoch": 0.5878300803673938,
+      "grad_norm": 6.288535636603857,
+      "learning_rate": 1e-06,
+      "loss": 0.1607,
+      "step": 7938
+    },
+    {
+      "epoch": 0.5901262916188289,
+      "grad_norm": 6.1321095364509235,
+      "learning_rate": 1e-06,
+      "loss": 0.1502,
+      "step": 7939
+    },
+    {
+      "epoch": 0.5924225028702641,
+      "grad_norm": 8.788213734996276,
+      "learning_rate": 1e-06,
+      "loss": 0.1605,
+      "step": 7940
+    },
+    {
+      "epoch": 0.5947187141216992,
+      "grad_norm": 6.0361038324419205,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 7941
+    },
+    {
+      "epoch": 0.5970149253731343,
+      "grad_norm": 7.82924251050645,
+      "learning_rate": 1e-06,
+      "loss": 0.144,
+      "step": 7942
+    },
+    {
+      "epoch": 0.5993111366245695,
+      "grad_norm": 7.77903320607827,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 7943
+    },
+    {
+      "epoch": 0.6016073478760046,
+      "grad_norm": 6.021407713562156,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 7944
+    },
+    {
+      "epoch": 0.6039035591274398,
+      "grad_norm": 7.3255533684059335,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 7945
+    },
+    {
+      "epoch": 0.6061997703788748,
+      "grad_norm": 10.039779123180626,
+      "learning_rate": 1e-06,
+      "loss": 0.1583,
+      "step": 7946
+    },
+    {
+      "epoch": 0.6084959816303099,
+      "grad_norm": 6.63702967006575,
+      "learning_rate": 1e-06,
+      "loss": 0.1724,
+      "step": 7947
+    },
+    {
+      "epoch": 0.6107921928817451,
+      "grad_norm": 5.800799485681269,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 7948
+    },
+    {
+      "epoch": 0.6130884041331802,
+      "grad_norm": 7.863184554830368,
+      "learning_rate": 1e-06,
+      "loss": 0.1507,
+      "step": 7949
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 9.235658516252137,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 7950
+    },
+    {
+      "epoch": 0.6176808266360505,
+      "grad_norm": 7.50542829165141,
+      "learning_rate": 1e-06,
+      "loss": 0.1536,
+      "step": 7951
+    },
+    {
+      "epoch": 0.6199770378874856,
+      "grad_norm": 8.270067170127767,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7952
+    },
+    {
+      "epoch": 0.6222732491389208,
+      "grad_norm": 6.856028495492457,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 7953
+    },
+    {
+      "epoch": 0.6245694603903559,
+      "grad_norm": 6.264071931889724,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 7954
+    },
+    {
+      "epoch": 0.6268656716417911,
+      "grad_norm": 7.835631469455996,
+      "learning_rate": 1e-06,
+      "loss": 0.1536,
+      "step": 7955
+    },
+    {
+      "epoch": 0.6291618828932262,
+      "grad_norm": 7.2620078938619566,
+      "learning_rate": 1e-06,
+      "loss": 0.158,
+      "step": 7956
+    },
+    {
+      "epoch": 0.6314580941446614,
+      "grad_norm": 8.27804029698383,
+      "learning_rate": 1e-06,
+      "loss": 0.1592,
+      "step": 7957
+    },
+    {
+      "epoch": 0.6337543053960965,
+      "grad_norm": 8.65448515661955,
+      "learning_rate": 1e-06,
+      "loss": 0.1531,
+      "step": 7958
+    },
+    {
+      "epoch": 0.6360505166475315,
+      "grad_norm": 7.048634378149625,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 7959
+    },
+    {
+      "epoch": 0.6383467278989667,
+      "grad_norm": 5.823384123441346,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 7960
+    },
+    {
+      "epoch": 0.6406429391504018,
+      "grad_norm": 7.766516962010919,
+      "learning_rate": 1e-06,
+      "loss": 0.1582,
+      "step": 7961
+    },
+    {
+      "epoch": 0.642939150401837,
+      "grad_norm": 8.218499389215047,
+      "learning_rate": 1e-06,
+      "loss": 0.1615,
+      "step": 7962
+    },
+    {
+      "epoch": 0.6452353616532721,
+      "grad_norm": 6.2334589058807115,
+      "learning_rate": 1e-06,
+      "loss": 0.1732,
+      "step": 7963
+    },
+    {
+      "epoch": 0.6475315729047072,
+      "grad_norm": 8.506829980290446,
+      "learning_rate": 1e-06,
+      "loss": 0.1473,
+      "step": 7964
+    },
+    {
+      "epoch": 0.6498277841561424,
+      "grad_norm": 6.088846797190086,
+      "learning_rate": 1e-06,
+      "loss": 0.1585,
+      "step": 7965
+    },
+    {
+      "epoch": 0.6521239954075775,
+      "grad_norm": 9.195716398383984,
+      "learning_rate": 1e-06,
+      "loss": 0.1612,
+      "step": 7966
+    },
+    {
+      "epoch": 0.6544202066590127,
+      "grad_norm": 7.744224928110508,
+      "learning_rate": 1e-06,
+      "loss": 0.1598,
+      "step": 7967
+    },
+    {
+      "epoch": 0.6567164179104478,
+      "grad_norm": 6.217365772565875,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 7968
+    },
+    {
+      "epoch": 0.6590126291618829,
+      "grad_norm": 8.945851384913487,
+      "learning_rate": 1e-06,
+      "loss": 0.1478,
+      "step": 7969
+    },
+    {
+      "epoch": 0.661308840413318,
+      "grad_norm": 6.499817190950913,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 7970
+    },
+    {
+      "epoch": 0.6636050516647531,
+      "grad_norm": 9.718804796044896,
+      "learning_rate": 1e-06,
+      "loss": 0.1541,
+      "step": 7971
+    },
+    {
+      "epoch": 0.6659012629161883,
+      "grad_norm": 7.8143405593129565,
+      "learning_rate": 1e-06,
+      "loss": 0.1616,
+      "step": 7972
+    },
+    {
+      "epoch": 0.6681974741676234,
+      "grad_norm": 7.497852020956264,
+      "learning_rate": 1e-06,
+      "loss": 0.1508,
+      "step": 7973
+    },
+    {
+      "epoch": 0.6704936854190585,
+      "grad_norm": 7.486296075219998,
+      "learning_rate": 1e-06,
+      "loss": 0.1544,
+      "step": 7974
+    },
+    {
+      "epoch": 0.6727898966704937,
+      "grad_norm": 6.564502121138618,
+      "learning_rate": 1e-06,
+      "loss": 0.1565,
+      "step": 7975
+    },
+    {
+      "epoch": 0.6750861079219288,
+      "grad_norm": 6.249662303919117,
+      "learning_rate": 1e-06,
+      "loss": 0.1533,
+      "step": 7976
+    },
+    {
+      "epoch": 0.677382319173364,
+      "grad_norm": 7.143818102074975,
+      "learning_rate": 1e-06,
+      "loss": 0.145,
+      "step": 7977
+    },
+    {
+      "epoch": 0.6796785304247991,
+      "grad_norm": 7.599722525372654,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 7978
+    },
+    {
+      "epoch": 0.6819747416762342,
+      "grad_norm": 6.904029139300998,
+      "learning_rate": 1e-06,
+      "loss": 0.1566,
+      "step": 7979
+    },
+    {
+      "epoch": 0.6842709529276694,
+      "grad_norm": 7.614139925540853,
+      "learning_rate": 1e-06,
+      "loss": 0.151,
+      "step": 7980
+    },
+    {
+      "epoch": 0.6865671641791045,
+      "grad_norm": 7.123782738570024,
+      "learning_rate": 1e-06,
+      "loss": 0.1673,
+      "step": 7981
+    },
+    {
+      "epoch": 0.6888633754305397,
+      "grad_norm": 7.246344491702357,
+      "learning_rate": 1e-06,
+      "loss": 0.1595,
+      "step": 7982
+    },
+    {
+      "epoch": 0.6911595866819747,
+      "grad_norm": 8.160429987907232,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 7983
+    },
+    {
+      "epoch": 0.6934557979334098,
+      "grad_norm": 6.998930091566931,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 7984
+    },
+    {
+      "epoch": 0.695752009184845,
+      "grad_norm": 8.18567375768506,
+      "learning_rate": 1e-06,
+      "loss": 0.1643,
+      "step": 7985
+    },
+    {
+      "epoch": 0.6980482204362801,
+      "grad_norm": 7.234990162733502,
+      "learning_rate": 1e-06,
+      "loss": 0.1443,
+      "step": 7986
+    },
+    {
+      "epoch": 0.7003444316877153,
+      "grad_norm": 6.95317384255324,
+      "learning_rate": 1e-06,
+      "loss": 0.1638,
+      "step": 7987
+    },
+    {
+      "epoch": 0.7026406429391504,
+      "grad_norm": 5.570916524682004,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 7988
+    },
+    {
+      "epoch": 0.7049368541905855,
+      "grad_norm": 6.161256180493802,
+      "learning_rate": 1e-06,
+      "loss": 0.1608,
+      "step": 7989
+    },
+    {
+      "epoch": 0.7072330654420207,
+      "grad_norm": 6.707071371373673,
+      "learning_rate": 1e-06,
+      "loss": 0.1432,
+      "step": 7990
+    },
+    {
+      "epoch": 0.7095292766934558,
+      "grad_norm": 7.4324005813622405,
+      "learning_rate": 1e-06,
+      "loss": 0.1669,
+      "step": 7991
+    },
+    {
+      "epoch": 0.711825487944891,
+      "grad_norm": 7.2953053070288165,
+      "learning_rate": 1e-06,
+      "loss": 0.1511,
+      "step": 7992
+    },
+    {
+      "epoch": 0.7141216991963261,
+      "grad_norm": 7.958100750826989,
+      "learning_rate": 1e-06,
+      "loss": 0.1535,
+      "step": 7993
+    },
+    {
+      "epoch": 0.7164179104477612,
+      "grad_norm": 8.861065148413568,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 7994
+    },
+    {
+      "epoch": 0.7187141216991964,
+      "grad_norm": 5.8404457684622,
+      "learning_rate": 1e-06,
+      "loss": 0.1434,
+      "step": 7995
+    },
+    {
+      "epoch": 0.7210103329506314,
+      "grad_norm": 7.29672151897822,
+      "learning_rate": 1e-06,
+      "loss": 0.1645,
+      "step": 7996
+    },
+    {
+      "epoch": 0.7233065442020666,
+      "grad_norm": 8.281511164865023,
+      "learning_rate": 1e-06,
+      "loss": 0.1578,
+      "step": 7997
+    },
+    {
+      "epoch": 0.7256027554535017,
+      "grad_norm": 8.366168878486759,
+      "learning_rate": 1e-06,
+      "loss": 0.1577,
+      "step": 7998
+    },
+    {
+      "epoch": 0.7278989667049368,
+      "grad_norm": 6.493877695110145,
+      "learning_rate": 1e-06,
+      "loss": 0.15,
+      "step": 7999
+    },
+    {
+      "epoch": 0.730195177956372,
+      "grad_norm": 6.94702330845494,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 8000
+    },
+    {
+      "epoch": 0.730195177956372,
+      "eval_loss": 4.28629732131958,
+      "eval_runtime": 21.839,
+      "eval_samples_per_second": 45.79,
+      "eval_steps_per_second": 5.724,
+      "step": 8000
+    },
+    {
+      "epoch": 0.7324913892078071,
+      "grad_norm": 8.097260667118404,
+      "learning_rate": 1e-06,
+      "loss": 0.1464,
+      "step": 8001
+    },
+    {
+      "epoch": 0.7347876004592423,
+      "grad_norm": 5.8511234014131706,
+      "learning_rate": 1e-06,
+      "loss": 0.1628,
+      "step": 8002
+    },
+    {
+      "epoch": 0.7370838117106774,
+      "grad_norm": 6.491465990852245,
+      "learning_rate": 1e-06,
+      "loss": 0.1601,
+      "step": 8003
+    },
+    {
+      "epoch": 0.7393800229621125,
+      "grad_norm": 7.883374623579385,
+      "learning_rate": 1e-06,
+      "loss": 0.1594,
+      "step": 8004
+    },
+    {
+      "epoch": 0.7416762342135477,
+      "grad_norm": 8.169255241196858,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 8005
+    },
+    {
+      "epoch": 0.7439724454649828,
+      "grad_norm": 6.55965256172657,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 8006
+    },
+    {
+      "epoch": 0.746268656716418,
+      "grad_norm": 8.745338550931033,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 8007
+    },
+    {
+      "epoch": 0.748564867967853,
+      "grad_norm": 7.177496685071046,
+      "learning_rate": 1e-06,
+      "loss": 0.1506,
+      "step": 8008
+    },
+    {
+      "epoch": 0.7508610792192881,
+      "grad_norm": 7.746891235642046,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 8009
+    },
+    {
+      "epoch": 0.7531572904707233,
+      "grad_norm": 8.23356391566588,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 8010
+    },
+    {
+      "epoch": 0.7554535017221584,
+      "grad_norm": 6.370590385106818,
+      "learning_rate": 1e-06,
+      "loss": 0.1454,
+      "step": 8011
+    },
+    {
+      "epoch": 0.7577497129735936,
+      "grad_norm": 11.035271390677131,
+      "learning_rate": 1e-06,
+      "loss": 0.1648,
+      "step": 8012
+    },
+    {
+      "epoch": 0.7600459242250287,
+      "grad_norm": 6.446257862855823,
+      "learning_rate": 1e-06,
+      "loss": 0.1529,
+      "step": 8013
+    },
+    {
+      "epoch": 0.7623421354764638,
+      "grad_norm": 5.800281870624475,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 8014
+    },
+    {
+      "epoch": 0.764638346727899,
+      "grad_norm": 6.567274378675515,
+      "learning_rate": 1e-06,
+      "loss": 0.172,
+      "step": 8015
+    },
+    {
+      "epoch": 0.7669345579793341,
+      "grad_norm": 6.709432815035508,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 8016
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 8.738608603188792,
+      "learning_rate": 1e-06,
+      "loss": 0.152,
+      "step": 8017
+    },
+    {
+      "epoch": 0.7715269804822044,
+      "grad_norm": 9.441580364825633,
+      "learning_rate": 1e-06,
+      "loss": 0.1558,
+      "step": 8018
+    },
+    {
+      "epoch": 0.7738231917336394,
+      "grad_norm": 7.1753303540811135,
+      "learning_rate": 1e-06,
+      "loss": 0.1474,
+      "step": 8019
+    },
+    {
+      "epoch": 0.7761194029850746,
+      "grad_norm": 6.993110660771053,
+      "learning_rate": 1e-06,
+      "loss": 0.1599,
+      "step": 8020
+    },
+    {
+      "epoch": 0.7784156142365097,
+      "grad_norm": 8.126514513538627,
+      "learning_rate": 1e-06,
+      "loss": 0.1573,
+      "step": 8021
+    },
+    {
+      "epoch": 0.7807118254879449,
+      "grad_norm": 6.304055917121788,
+      "learning_rate": 1e-06,
+      "loss": 0.1549,
+      "step": 8022
+    },
+    {
+      "epoch": 0.78300803673938,
+      "grad_norm": 6.511019327572196,
+      "learning_rate": 1e-06,
+      "loss": 0.1415,
+      "step": 8023
+    },
+    {
+      "epoch": 0.7853042479908151,
+      "grad_norm": 9.08925029278556,
+      "learning_rate": 1e-06,
+      "loss": 0.1548,
+      "step": 8024
+    },
+    {
+      "epoch": 0.7876004592422503,
+      "grad_norm": 7.196172453580945,
+      "learning_rate": 1e-06,
+      "loss": 0.1458,
+      "step": 8025
+    },
+    {
+      "epoch": 0.7898966704936854,
+      "grad_norm": 5.912210940009095,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 8026
+    },
+    {
+      "epoch": 0.7921928817451206,
+      "grad_norm": 7.957659948869398,
+      "learning_rate": 1e-06,
+      "loss": 0.1572,
+      "step": 8027
+    },
+    {
+      "epoch": 0.7944890929965557,
+      "grad_norm": 7.644589005243513,
+      "learning_rate": 1e-06,
+      "loss": 0.1467,
+      "step": 8028
+    },
+    {
+      "epoch": 0.7967853042479908,
+      "grad_norm": 6.867340318099123,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 8029
+    },
+    {
+      "epoch": 0.799081515499426,
+      "grad_norm": 8.04747190175616,
+      "learning_rate": 1e-06,
+      "loss": 0.1576,
+      "step": 8030
+    },
+    {
+      "epoch": 0.801377726750861,
+      "grad_norm": 8.568965032313127,
+      "learning_rate": 1e-06,
+      "loss": 0.1657,
+      "step": 8031
+    },
+    {
+      "epoch": 0.8036739380022963,
+      "grad_norm": 6.345771810092971,
+      "learning_rate": 1e-06,
+      "loss": 0.1525,
+      "step": 8032
+    },
+    {
+      "epoch": 0.8059701492537313,
+      "grad_norm": 7.729412671347076,
+      "learning_rate": 1e-06,
+      "loss": 0.147,
+      "step": 8033
+    },
+    {
+      "epoch": 0.8082663605051664,
+      "grad_norm": 8.260524074814466,
+      "learning_rate": 1e-06,
+      "loss": 0.1568,
+      "step": 8034
+    },
+    {
+      "epoch": 0.8105625717566016,
+      "grad_norm": 8.213812026429773,
+      "learning_rate": 1e-06,
+      "loss": 0.1635,
+      "step": 8035
+    },
+    {
+      "epoch": 0.8128587830080367,
+      "grad_norm": 6.5537205739590645,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 8036
+    },
+    {
+      "epoch": 0.8151549942594719,
+      "grad_norm": 8.163308133942385,
+      "learning_rate": 1e-06,
+      "loss": 0.1447,
+      "step": 8037
+    },
+    {
+      "epoch": 0.817451205510907,
+      "grad_norm": 7.322732354328141,
+      "learning_rate": 1e-06,
+      "loss": 0.1528,
+      "step": 8038
+    },
+    {
+      "epoch": 0.8197474167623421,
+      "grad_norm": 6.526008890782628,
+      "learning_rate": 1e-06,
+      "loss": 0.1611,
+      "step": 8039
+    },
+    {
+      "epoch": 0.8220436280137773,
+      "grad_norm": 7.36849541183724,
+      "learning_rate": 1e-06,
+      "loss": 0.1545,
+      "step": 8040
+    },
+    {
+      "epoch": 0.8243398392652124,
+      "grad_norm": 6.8999033926836875,
+      "learning_rate": 1e-06,
+      "loss": 0.1433,
+      "step": 8041
+    },
+    {
+      "epoch": 0.8266360505166476,
+      "grad_norm": 6.728834671220053,
+      "learning_rate": 1e-06,
+      "loss": 0.1654,
+      "step": 8042
+    },
+    {
+      "epoch": 0.8289322617680827,
+      "grad_norm": 6.417742610328875,
+      "learning_rate": 1e-06,
+      "loss": 0.1441,
+      "step": 8043
+    },
+    {
+      "epoch": 0.8312284730195177,
+      "grad_norm": 7.002180242662402,
+      "learning_rate": 1e-06,
+      "loss": 0.1539,
+      "step": 8044
+    },
+    {
+      "epoch": 0.8335246842709529,
+      "grad_norm": 8.026181666678676,
+      "learning_rate": 1e-06,
+      "loss": 0.1489,
+      "step": 8045
+    },
+    {
+      "epoch": 0.835820895522388,
+      "grad_norm": 6.69771735934747,
+      "learning_rate": 1e-06,
+      "loss": 0.1659,
+      "step": 8046
+    },
+    {
+      "epoch": 0.8381171067738232,
+      "grad_norm": 8.491746431131942,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 8047
+    },
+    {
+      "epoch": 0.8404133180252583,
+      "grad_norm": 8.14673119396639,
+      "learning_rate": 1e-06,
+      "loss": 0.1578,
+      "step": 8048
+    },
+    {
+      "epoch": 0.8427095292766934,
+      "grad_norm": 4.281396189832617,
+      "learning_rate": 1e-06,
+      "loss": 0.1679,
+      "step": 8049
+    },
+    {
+      "epoch": 0.8450057405281286,
+      "grad_norm": 6.434891459504339,
+      "learning_rate": 1e-06,
+      "loss": 0.1458,
+      "step": 8050
+    },
+    {
+      "epoch": 0.8473019517795637,
+      "grad_norm": 8.783554850864506,
+      "learning_rate": 1e-06,
+      "loss": 0.173,
+      "step": 8051
+    },
+    {
+      "epoch": 0.8495981630309989,
+      "grad_norm": 7.4625815274585845,
+      "learning_rate": 1e-06,
+      "loss": 0.1449,
+      "step": 8052
+    },
+    {
+      "epoch": 0.851894374282434,
+      "grad_norm": 6.932053560514666,
+      "learning_rate": 1e-06,
+      "loss": 0.1571,
+      "step": 8053
+    },
+    {
+      "epoch": 0.8541905855338691,
+      "grad_norm": 6.258529191660508,
+      "learning_rate": 1e-06,
+      "loss": 0.1504,
+      "step": 8054
+    },
+    {
+      "epoch": 0.8564867967853043,
+      "grad_norm": 7.66453807150301,
+      "learning_rate": 1e-06,
+      "loss": 0.1684,
+      "step": 8055
+    },
+    {
+      "epoch": 0.8587830080367393,
+      "grad_norm": 7.56317522521938,
+      "learning_rate": 1e-06,
+      "loss": 0.1627,
+      "step": 8056
+    },
+    {
+      "epoch": 0.8610792192881745,
+      "grad_norm": 8.020678955066185,
+      "learning_rate": 1e-06,
+      "loss": 0.1492,
+      "step": 8057
+    },
+    {
+      "epoch": 0.8633754305396096,
+      "grad_norm": 10.14730804922406,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 8058
+    },
+    {
+      "epoch": 0.8656716417910447,
+      "grad_norm": 5.188171327761318,
+      "learning_rate": 1e-06,
+      "loss": 0.1325,
+      "step": 8059
+    },
+    {
+      "epoch": 0.8679678530424799,
+      "grad_norm": 5.757928049359244,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 8060
+    },
+    {
+      "epoch": 0.870264064293915,
+      "grad_norm": 6.270428645317131,
+      "learning_rate": 1e-06,
+      "loss": 0.1502,
+      "step": 8061
+    },
+    {
+      "epoch": 0.8725602755453502,
+      "grad_norm": 4.870050024805222,
+      "learning_rate": 1e-06,
+      "loss": 0.1603,
+      "step": 8062
+    },
+    {
+      "epoch": 0.8748564867967853,
+      "grad_norm": 7.435115412107751,
+      "learning_rate": 1e-06,
+      "loss": 0.1497,
+      "step": 8063
+    },
+    {
+      "epoch": 0.8771526980482205,
+      "grad_norm": 6.767497225824215,
+      "learning_rate": 1e-06,
+      "loss": 0.1501,
+      "step": 8064
+    },
+    {
+      "epoch": 0.8794489092996556,
+      "grad_norm": 5.3509202656297,
+      "learning_rate": 1e-06,
+      "loss": 0.1595,
+      "step": 8065
+    },
+    {
+      "epoch": 0.8817451205510907,
+      "grad_norm": 8.267609491789765,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 8066
+    },
+    {
+      "epoch": 0.8840413318025259,
+      "grad_norm": 7.327413835970671,
+      "learning_rate": 1e-06,
+      "loss": 0.1596,
+      "step": 8067
+    },
+    {
+      "epoch": 0.886337543053961,
+      "grad_norm": 6.858824981332983,
+      "learning_rate": 1e-06,
+      "loss": 0.1589,
+      "step": 8068
+    },
+    {
+      "epoch": 0.8886337543053962,
+      "grad_norm": 7.393491583350042,
+      "learning_rate": 1e-06,
+      "loss": 0.1625,
+      "step": 8069
+    },
+    {
+      "epoch": 0.8909299655568312,
+      "grad_norm": 7.73013593985009,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 8070
+    },
+    {
+      "epoch": 0.8932261768082663,
+      "grad_norm": 8.723704424854756,
+      "learning_rate": 1e-06,
+      "loss": 0.148,
+      "step": 8071
+    },
+    {
+      "epoch": 0.8955223880597015,
+      "grad_norm": 8.547770633588694,
+      "learning_rate": 1e-06,
+      "loss": 0.1485,
+      "step": 8072
+    },
+    {
+      "epoch": 0.8978185993111366,
+      "grad_norm": 8.976965460074851,
+      "learning_rate": 1e-06,
+      "loss": 0.157,
+      "step": 8073
+    },
+    {
+      "epoch": 0.9001148105625718,
+      "grad_norm": 6.134065044244521,
+      "learning_rate": 1e-06,
+      "loss": 0.1613,
+      "step": 8074
+    },
+    {
+      "epoch": 0.9024110218140069,
+      "grad_norm": 7.966591309736967,
+      "learning_rate": 1e-06,
+      "loss": 0.1526,
+      "step": 8075
+    },
+    {
+      "epoch": 0.904707233065442,
+      "grad_norm": 5.674845853399636,
+      "learning_rate": 1e-06,
+      "loss": 0.1498,
+      "step": 8076
+    },
+    {
+      "epoch": 0.9070034443168772,
+      "grad_norm": 9.211443439588548,
+      "learning_rate": 1e-06,
+      "loss": 0.1787,
+      "step": 8077
+    },
+    {
+      "epoch": 0.9092996555683123,
+      "grad_norm": 6.320588370651728,
+      "learning_rate": 1e-06,
+      "loss": 0.1457,
+      "step": 8078
+    },
+    {
+      "epoch": 0.9115958668197475,
+      "grad_norm": 8.94758900948424,
+      "learning_rate": 1e-06,
+      "loss": 0.1546,
+      "step": 8079
+    },
+    {
+      "epoch": 0.9138920780711826,
+      "grad_norm": 7.5681657620091425,
+      "learning_rate": 1e-06,
+      "loss": 0.1479,
+      "step": 8080
+    },
+    {
+      "epoch": 0.9161882893226176,
+      "grad_norm": 6.630668622996598,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 8081
+    },
+    {
+      "epoch": 0.9184845005740528,
+      "grad_norm": 6.993026404297085,
+      "learning_rate": 1e-06,
+      "loss": 0.152,
+      "step": 8082
+    },
+    {
+      "epoch": 0.9207807118254879,
+      "grad_norm": 5.263256858756267,
+      "learning_rate": 1e-06,
+      "loss": 0.1652,
+      "step": 8083
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 6.70811862858535,
+      "learning_rate": 1e-06,
+      "loss": 0.1482,
+      "step": 8084
+    },
+    {
+      "epoch": 0.9253731343283582,
+      "grad_norm": 7.231453465155128,
+      "learning_rate": 1e-06,
+      "loss": 0.149,
+      "step": 8085
+    },
+    {
+      "epoch": 0.9276693455797933,
+      "grad_norm": 5.879188218020214,
+      "learning_rate": 1e-06,
+      "loss": 0.1448,
+      "step": 8086
+    },
+    {
+      "epoch": 0.9299655568312285,
+      "grad_norm": 7.001188307679316,
+      "learning_rate": 1e-06,
+      "loss": 0.1552,
+      "step": 8087
+    },
+    {
+      "epoch": 0.9322617680826636,
+      "grad_norm": 7.707404703838211,
+      "learning_rate": 1e-06,
+      "loss": 0.155,
+      "step": 8088
+    },
+    {
+      "epoch": 0.9345579793340988,
+      "grad_norm": 6.61281995117146,
+      "learning_rate": 1e-06,
+      "loss": 0.1604,
+      "step": 8089
+    },
+    {
+      "epoch": 0.9368541905855339,
+      "grad_norm": 5.580937342628284,
+      "learning_rate": 1e-06,
+      "loss": 0.1637,
+      "step": 8090
+    },
+    {
+      "epoch": 0.939150401836969,
+      "grad_norm": 9.178236233573653,
+      "learning_rate": 1e-06,
+      "loss": 0.1649,
+      "step": 8091
+    },
+    {
+      "epoch": 0.9414466130884042,
+      "grad_norm": 7.337285429547349,
+      "learning_rate": 1e-06,
+      "loss": 0.1476,
+      "step": 8092
+    },
+    {
+      "epoch": 0.9437428243398392,
+      "grad_norm": 8.277619049937337,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 8093
+    },
+    {
+      "epoch": 0.9460390355912744,
+      "grad_norm": 8.17349086550506,
+      "learning_rate": 1e-06,
+      "loss": 0.1556,
+      "step": 8094
+    },
+    {
+      "epoch": 0.9483352468427095,
+      "grad_norm": 6.203857268790211,
+      "learning_rate": 1e-06,
+      "loss": 0.1561,
+      "step": 8095
+    },
+    {
+      "epoch": 0.9506314580941446,
+      "grad_norm": 8.864764250609594,
+      "learning_rate": 1e-06,
+      "loss": 0.1609,
+      "step": 8096
+    },
+    {
+      "epoch": 0.9529276693455798,
+      "grad_norm": 9.56940062346086,
+      "learning_rate": 1e-06,
+      "loss": 0.1617,
+      "step": 8097
+    },
+    {
+      "epoch": 0.9552238805970149,
+      "grad_norm": 6.701172612005919,
+      "learning_rate": 1e-06,
+      "loss": 0.1485,
+      "step": 8098
+    },
+    {
+      "epoch": 0.9575200918484501,
+      "grad_norm": 7.038817442731542,
+      "learning_rate": 1e-06,
+      "loss": 0.1633,
+      "step": 8099
+    },
+    {
+      "epoch": 0.9598163030998852,
+      "grad_norm": 7.39199663452204,
+      "learning_rate": 1e-06,
+      "loss": 0.168,
+      "step": 8100
+    },
+    {
+      "epoch": 0.9621125143513203,
+      "grad_norm": 8.939815567077272,
+      "learning_rate": 1e-06,
+      "loss": 0.1663,
+      "step": 8101
+    },
+    {
+      "epoch": 0.9644087256027555,
+      "grad_norm": 6.280079907032182,
+      "learning_rate": 1e-06,
+      "loss": 0.1589,
+      "step": 8102
+    },
+    {
+      "epoch": 0.9667049368541906,
+      "grad_norm": 7.773851918198832,
+      "learning_rate": 1e-06,
+      "loss": 0.1574,
+      "step": 8103
+    },
+    {
+      "epoch": 0.9690011481056258,
+      "grad_norm": 6.168963785836043,
+      "learning_rate": 1e-06,
+      "loss": 0.1584,
+      "step": 8104
+    },
+    {
+      "epoch": 0.9712973593570609,
+      "grad_norm": 6.772293871776498,
+      "learning_rate": 1e-06,
+      "loss": 0.1493,
+      "step": 8105
+    },
+    {
+      "epoch": 0.9735935706084959,
+      "grad_norm": 5.813379417301357,
+      "learning_rate": 1e-06,
+      "loss": 0.1461,
+      "step": 8106
+    },
+    {
+      "epoch": 0.9758897818599311,
+      "grad_norm": 6.303626204223311,
+      "learning_rate": 1e-06,
+      "loss": 0.1542,
+      "step": 8107
+    },
+    {
+      "epoch": 0.9781859931113662,
+      "grad_norm": 7.0444158435991096,
+      "learning_rate": 1e-06,
+      "loss": 0.1575,
+      "step": 8108
+    },
+    {
+      "epoch": 0.9804822043628014,
+      "grad_norm": 8.018551031217704,
+      "learning_rate": 1e-06,
+      "loss": 0.163,
+      "step": 8109
+    },
+    {
+      "epoch": 0.9827784156142365,
+      "grad_norm": 7.818242194092758,
+      "learning_rate": 1e-06,
+      "loss": 0.1622,
+      "step": 8110
+    },
+    {
+      "epoch": 0.9850746268656716,
+      "grad_norm": 7.655022999747668,
+      "learning_rate": 1e-06,
+      "loss": 0.1636,
+      "step": 8111
+    },
+    {
+      "epoch": 0.9873708381171068,
+      "grad_norm": 6.309786078418231,
+      "learning_rate": 1e-06,
+      "loss": 0.1524,
+      "step": 8112
+    },
+    {
+      "epoch": 0.9896670493685419,
+      "grad_norm": 6.12269256114612,
+      "learning_rate": 1e-06,
+      "loss": 0.1557,
+      "step": 8113
+    },
+    {
+      "epoch": 0.9919632606199771,
+      "grad_norm": 6.0336153772658,
+      "learning_rate": 1e-06,
+      "loss": 0.1543,
+      "step": 8114
+    },
+    {
+      "epoch": 0.9942594718714122,
+      "grad_norm": 6.607039611603088,
+      "learning_rate": 1e-06,
+      "loss": 0.1495,
+      "step": 8115
+    },
+    {
+      "epoch": 0.9965556831228473,
+      "grad_norm": 6.298348494887612,
+      "learning_rate": 1e-06,
+      "loss": 0.1517,
+      "step": 8116
+    },
+    {
+      "epoch": 0.9988518943742825,
+      "grad_norm": 7.771254646162162,
+      "learning_rate": 1e-06,
+      "loss": 0.1468,
+      "step": 8117
+    },
+    {
+      "epoch": 0.9988518943742825,
+      "eval_loss": 4.419665813446045,
+      "eval_runtime": 21.5422,
+      "eval_samples_per_second": 46.42,
+      "eval_steps_per_second": 5.803,
+      "step": 8117
+    },
+    {
+      "epoch": 0.0031397174254317113,
+      "grad_norm": 6.331689268813883,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 8118
+    },
+    {
+      "epoch": 0.006279434850863423,
+      "grad_norm": 5.298195012986347,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 8119
+    },
+    {
+      "epoch": 0.009419152276295133,
+      "grad_norm": 4.198508846413118,
+      "learning_rate": 1e-06,
+      "loss": 0.0959,
+      "step": 8120
+    },
+    {
+      "epoch": 0.012558869701726845,
+      "grad_norm": 5.976187524006599,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 8121
+    },
+    {
+      "epoch": 0.015698587127158554,
+      "grad_norm": 7.124714070177216,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 8122
+    },
+    {
+      "epoch": 0.018838304552590265,
+      "grad_norm": 5.83090126179824,
+      "learning_rate": 1e-06,
+      "loss": 0.1009,
+      "step": 8123
+    },
+    {
+      "epoch": 0.02197802197802198,
+      "grad_norm": 6.90893091226531,
+      "learning_rate": 1e-06,
+      "loss": 0.1001,
+      "step": 8124
+    },
+    {
+      "epoch": 0.02511773940345369,
+      "grad_norm": 8.027825966836733,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 8125
+    },
+    {
+      "epoch": 0.0282574568288854,
+      "grad_norm": 7.052593695708038,
+      "learning_rate": 1e-06,
+      "loss": 0.1063,
+      "step": 8126
+    },
+    {
+      "epoch": 0.03139717425431711,
+      "grad_norm": 6.522063817215499,
+      "learning_rate": 1e-06,
+      "loss": 0.1103,
+      "step": 8127
+    },
+    {
+      "epoch": 0.03453689167974882,
+      "grad_norm": 8.806398453458034,
+      "learning_rate": 1e-06,
+      "loss": 0.1,
+      "step": 8128
+    },
+    {
+      "epoch": 0.03767660910518053,
+      "grad_norm": 8.727920351386357,
+      "learning_rate": 1e-06,
+      "loss": 0.0994,
+      "step": 8129
+    },
+    {
+      "epoch": 0.04081632653061224,
+      "grad_norm": 9.292947491436696,
+      "learning_rate": 1e-06,
+      "loss": 0.1081,
+      "step": 8130
+    },
+    {
+      "epoch": 0.04395604395604396,
+      "grad_norm": 6.891754624345574,
+      "learning_rate": 1e-06,
+      "loss": 0.1016,
+      "step": 8131
+    },
+    {
+      "epoch": 0.04709576138147567,
+      "grad_norm": 9.582022244543088,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 8132
+    },
+    {
+      "epoch": 0.05023547880690738,
+      "grad_norm": 8.421181597036018,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8133
+    },
+    {
+      "epoch": 0.05337519623233909,
+      "grad_norm": 8.576144583820906,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 8134
+    },
+    {
+      "epoch": 0.0565149136577708,
+      "grad_norm": 9.288401303348493,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 8135
+    },
+    {
+      "epoch": 0.059654631083202514,
+      "grad_norm": 8.982679822592223,
+      "learning_rate": 1e-06,
+      "loss": 0.1015,
+      "step": 8136
+    },
+    {
+      "epoch": 0.06279434850863422,
+      "grad_norm": 11.630814886168201,
+      "learning_rate": 1e-06,
+      "loss": 0.1117,
+      "step": 8137
+    },
+    {
+      "epoch": 0.06593406593406594,
+      "grad_norm": 8.95020297678907,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8138
+    },
+    {
+      "epoch": 0.06907378335949764,
+      "grad_norm": 8.425454304056021,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 8139
+    },
+    {
+      "epoch": 0.07221350078492936,
+      "grad_norm": 6.721930272749378,
+      "learning_rate": 1e-06,
+      "loss": 0.0969,
+      "step": 8140
+    },
+    {
+      "epoch": 0.07535321821036106,
+      "grad_norm": 7.637066363930332,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8141
+    },
+    {
+      "epoch": 0.07849293563579278,
+      "grad_norm": 8.413238427091445,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 8142
+    },
+    {
+      "epoch": 0.08163265306122448,
+      "grad_norm": 8.174070337952772,
+      "learning_rate": 1e-06,
+      "loss": 0.0999,
+      "step": 8143
+    },
+    {
+      "epoch": 0.0847723704866562,
+      "grad_norm": 8.358799961122736,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 8144
+    },
+    {
+      "epoch": 0.08791208791208792,
+      "grad_norm": 5.2449127691931325,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 8145
+    },
+    {
+      "epoch": 0.09105180533751962,
+      "grad_norm": 8.258155567284327,
+      "learning_rate": 1e-06,
+      "loss": 0.1087,
+      "step": 8146
+    },
+    {
+      "epoch": 0.09419152276295134,
+      "grad_norm": 8.067861079211115,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 8147
+    },
+    {
+      "epoch": 0.09733124018838304,
+      "grad_norm": 7.847731362144061,
+      "learning_rate": 1e-06,
+      "loss": 0.1161,
+      "step": 8148
+    },
+    {
+      "epoch": 0.10047095761381476,
+      "grad_norm": 7.153982930976583,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 8149
+    },
+    {
+      "epoch": 0.10361067503924647,
+      "grad_norm": 9.359172333057876,
+      "learning_rate": 1e-06,
+      "loss": 0.1081,
+      "step": 8150
+    },
+    {
+      "epoch": 0.10675039246467818,
+      "grad_norm": 6.699929664960811,
+      "learning_rate": 1e-06,
+      "loss": 0.102,
+      "step": 8151
+    },
+    {
+      "epoch": 0.10989010989010989,
+      "grad_norm": 7.699245218590015,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 8152
+    },
+    {
+      "epoch": 0.1130298273155416,
+      "grad_norm": 9.549090393587957,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 8153
+    },
+    {
+      "epoch": 0.11616954474097331,
+      "grad_norm": 8.520786208602571,
+      "learning_rate": 1e-06,
+      "loss": 0.1011,
+      "step": 8154
+    },
+    {
+      "epoch": 0.11930926216640503,
+      "grad_norm": 9.255040518744414,
+      "learning_rate": 1e-06,
+      "loss": 0.0995,
+      "step": 8155
+    },
+    {
+      "epoch": 0.12244897959183673,
+      "grad_norm": 6.852025865299379,
+      "learning_rate": 1e-06,
+      "loss": 0.1064,
+      "step": 8156
+    },
+    {
+      "epoch": 0.12558869701726844,
+      "grad_norm": 7.7507522468873296,
+      "learning_rate": 1e-06,
+      "loss": 0.1012,
+      "step": 8157
+    },
+    {
+      "epoch": 0.12872841444270017,
+      "grad_norm": 7.686438004137731,
+      "learning_rate": 1e-06,
+      "loss": 0.1042,
+      "step": 8158
+    },
+    {
+      "epoch": 0.13186813186813187,
+      "grad_norm": 6.584449335689215,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 8159
+    },
+    {
+      "epoch": 0.13500784929356358,
+      "grad_norm": 7.244065626219772,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 8160
+    },
+    {
+      "epoch": 0.13814756671899528,
+      "grad_norm": 7.427038517029295,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 8161
+    },
+    {
+      "epoch": 0.141287284144427,
+      "grad_norm": 9.13976137084808,
+      "learning_rate": 1e-06,
+      "loss": 0.099,
+      "step": 8162
+    },
+    {
+      "epoch": 0.14442700156985872,
+      "grad_norm": 6.891120224594232,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 8163
+    },
+    {
+      "epoch": 0.14756671899529042,
+      "grad_norm": 7.654844384780352,
+      "learning_rate": 1e-06,
+      "loss": 0.1061,
+      "step": 8164
+    },
+    {
+      "epoch": 0.15070643642072212,
+      "grad_norm": 10.884162761323324,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 8165
+    },
+    {
+      "epoch": 0.15384615384615385,
+      "grad_norm": 7.444450376224758,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 8166
+    },
+    {
+      "epoch": 0.15698587127158556,
+      "grad_norm": 8.949997479476684,
+      "learning_rate": 1e-06,
+      "loss": 0.1019,
+      "step": 8167
+    },
+    {
+      "epoch": 0.16012558869701726,
+      "grad_norm": 7.01512426524724,
+      "learning_rate": 1e-06,
+      "loss": 0.1031,
+      "step": 8168
+    },
+    {
+      "epoch": 0.16326530612244897,
+      "grad_norm": 6.80345994209998,
+      "learning_rate": 1e-06,
+      "loss": 0.1046,
+      "step": 8169
+    },
+    {
+      "epoch": 0.1664050235478807,
+      "grad_norm": 8.466868525103449,
+      "learning_rate": 1e-06,
+      "loss": 0.1039,
+      "step": 8170
+    },
+    {
+      "epoch": 0.1695447409733124,
+      "grad_norm": 7.423619808893166,
+      "learning_rate": 1e-06,
+      "loss": 0.1055,
+      "step": 8171
+    },
+    {
+      "epoch": 0.1726844583987441,
+      "grad_norm": 8.342119258557334,
+      "learning_rate": 1e-06,
+      "loss": 0.113,
+      "step": 8172
+    },
+    {
+      "epoch": 0.17582417582417584,
+      "grad_norm": 8.143563492314236,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 8173
+    },
+    {
+      "epoch": 0.17896389324960754,
+      "grad_norm": 7.551692386511135,
+      "learning_rate": 1e-06,
+      "loss": 0.102,
+      "step": 8174
+    },
+    {
+      "epoch": 0.18210361067503925,
+      "grad_norm": 6.904973057873363,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 8175
+    },
+    {
+      "epoch": 0.18524332810047095,
+      "grad_norm": 9.602730196308821,
+      "learning_rate": 1e-06,
+      "loss": 0.121,
+      "step": 8176
+    },
+    {
+      "epoch": 0.18838304552590268,
+      "grad_norm": 7.545514571016006,
+      "learning_rate": 1e-06,
+      "loss": 0.0972,
+      "step": 8177
+    },
+    {
+      "epoch": 0.19152276295133439,
+      "grad_norm": 7.318680150234032,
+      "learning_rate": 1e-06,
+      "loss": 0.1039,
+      "step": 8178
+    },
+    {
+      "epoch": 0.1946624803767661,
+      "grad_norm": 7.6311099321232385,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 8179
+    },
+    {
+      "epoch": 0.1978021978021978,
+      "grad_norm": 7.161986033706798,
+      "learning_rate": 1e-06,
+      "loss": 0.1177,
+      "step": 8180
+    },
+    {
+      "epoch": 0.20094191522762953,
+      "grad_norm": 7.499606844047395,
+      "learning_rate": 1e-06,
+      "loss": 0.1108,
+      "step": 8181
+    },
+    {
+      "epoch": 0.20408163265306123,
+      "grad_norm": 9.305613070233273,
+      "learning_rate": 1e-06,
+      "loss": 0.1117,
+      "step": 8182
+    },
+    {
+      "epoch": 0.20722135007849293,
+      "grad_norm": 7.629007172722696,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 8183
+    },
+    {
+      "epoch": 0.21036106750392464,
+      "grad_norm": 9.950851184530924,
+      "learning_rate": 1e-06,
+      "loss": 0.1135,
+      "step": 8184
+    },
+    {
+      "epoch": 0.21350078492935637,
+      "grad_norm": 8.005352663907182,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 8185
+    },
+    {
+      "epoch": 0.21664050235478807,
+      "grad_norm": 7.654874639054399,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 8186
+    },
+    {
+      "epoch": 0.21978021978021978,
+      "grad_norm": 7.549483553750064,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 8187
+    },
+    {
+      "epoch": 0.22291993720565148,
+      "grad_norm": 7.776529450384181,
+      "learning_rate": 1e-06,
+      "loss": 0.1033,
+      "step": 8188
+    },
+    {
+      "epoch": 0.2260596546310832,
+      "grad_norm": 9.24460319831194,
+      "learning_rate": 1e-06,
+      "loss": 0.1038,
+      "step": 8189
+    },
+    {
+      "epoch": 0.22919937205651492,
+      "grad_norm": 7.4650390967235865,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 8190
+    },
+    {
+      "epoch": 0.23233908948194662,
+      "grad_norm": 8.270626493005143,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 8191
+    },
+    {
+      "epoch": 0.23547880690737832,
+      "grad_norm": 8.218111695701792,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 8192
+    },
+    {
+      "epoch": 0.23861852433281006,
+      "grad_norm": 6.121242342499773,
+      "learning_rate": 1e-06,
+      "loss": 0.0986,
+      "step": 8193
+    },
+    {
+      "epoch": 0.24175824175824176,
+      "grad_norm": 6.1072470031430255,
+      "learning_rate": 1e-06,
+      "loss": 0.1039,
+      "step": 8194
+    },
+    {
+      "epoch": 0.24489795918367346,
+      "grad_norm": 7.387421399741105,
+      "learning_rate": 1e-06,
+      "loss": 0.1092,
+      "step": 8195
+    },
+    {
+      "epoch": 0.24803767660910517,
+      "grad_norm": 9.494958822173812,
+      "learning_rate": 1e-06,
+      "loss": 0.1113,
+      "step": 8196
+    },
+    {
+      "epoch": 0.25117739403453687,
+      "grad_norm": 7.030636928689296,
+      "learning_rate": 1e-06,
+      "loss": 0.1048,
+      "step": 8197
+    },
+    {
+      "epoch": 0.2543171114599686,
+      "grad_norm": 7.21117890952622,
+      "learning_rate": 1e-06,
+      "loss": 0.1083,
+      "step": 8198
+    },
+    {
+      "epoch": 0.25745682888540034,
+      "grad_norm": 6.353648064150902,
+      "learning_rate": 1e-06,
+      "loss": 0.1059,
+      "step": 8199
+    },
+    {
+      "epoch": 0.260596546310832,
+      "grad_norm": 8.218965053917643,
+      "learning_rate": 1e-06,
+      "loss": 0.105,
+      "step": 8200
+    },
+    {
+      "epoch": 0.26373626373626374,
+      "grad_norm": 7.9288194656931825,
+      "learning_rate": 1e-06,
+      "loss": 0.0993,
+      "step": 8201
+    },
+    {
+      "epoch": 0.2668759811616955,
+      "grad_norm": 8.1596343303744,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 8202
+    },
+    {
+      "epoch": 0.27001569858712715,
+      "grad_norm": 7.871916773682,
+      "learning_rate": 1e-06,
+      "loss": 0.1126,
+      "step": 8203
+    },
+    {
+      "epoch": 0.2731554160125589,
+      "grad_norm": 8.396523730181238,
+      "learning_rate": 1e-06,
+      "loss": 0.104,
+      "step": 8204
+    },
+    {
+      "epoch": 0.27629513343799056,
+      "grad_norm": 8.93422697495349,
+      "learning_rate": 1e-06,
+      "loss": 0.1103,
+      "step": 8205
+    },
+    {
+      "epoch": 0.2794348508634223,
+      "grad_norm": 7.744304731910375,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8206
+    },
+    {
+      "epoch": 0.282574568288854,
+      "grad_norm": 9.941787036559397,
+      "learning_rate": 1e-06,
+      "loss": 0.1068,
+      "step": 8207
+    },
+    {
+      "epoch": 0.2857142857142857,
+      "grad_norm": 6.747325805746025,
+      "learning_rate": 1e-06,
+      "loss": 0.1069,
+      "step": 8208
+    },
+    {
+      "epoch": 0.28885400313971743,
+      "grad_norm": 6.981826774906655,
+      "learning_rate": 1e-06,
+      "loss": 0.1033,
+      "step": 8209
+    },
+    {
+      "epoch": 0.29199372056514916,
+      "grad_norm": 7.12458987972899,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8210
+    },
+    {
+      "epoch": 0.29513343799058084,
+      "grad_norm": 6.338620769615809,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 8211
+    },
+    {
+      "epoch": 0.29827315541601257,
+      "grad_norm": 8.899925912494755,
+      "learning_rate": 1e-06,
+      "loss": 0.1039,
+      "step": 8212
+    },
+    {
+      "epoch": 0.30141287284144425,
+      "grad_norm": 10.28616578087231,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 8213
+    },
+    {
+      "epoch": 0.304552590266876,
+      "grad_norm": 5.485387551909454,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 8214
+    },
+    {
+      "epoch": 0.3076923076923077,
+      "grad_norm": 8.185295120977221,
+      "learning_rate": 1e-06,
+      "loss": 0.1126,
+      "step": 8215
+    },
+    {
+      "epoch": 0.3108320251177394,
+      "grad_norm": 6.207025715243007,
+      "learning_rate": 1e-06,
+      "loss": 0.1064,
+      "step": 8216
+    },
+    {
+      "epoch": 0.3139717425431711,
+      "grad_norm": 6.983603828551381,
+      "learning_rate": 1e-06,
+      "loss": 0.0965,
+      "step": 8217
+    },
+    {
+      "epoch": 0.31711145996860285,
+      "grad_norm": 10.118288701884614,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 8218
+    },
+    {
+      "epoch": 0.3202511773940345,
+      "grad_norm": 6.693489034482936,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 8219
+    },
+    {
+      "epoch": 0.32339089481946626,
+      "grad_norm": 9.771023889616183,
+      "learning_rate": 1e-06,
+      "loss": 0.1155,
+      "step": 8220
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 6.486696197431825,
+      "learning_rate": 1e-06,
+      "loss": 0.0982,
+      "step": 8221
+    },
+    {
+      "epoch": 0.32967032967032966,
+      "grad_norm": 7.247306493851486,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 8222
+    },
+    {
+      "epoch": 0.3328100470957614,
+      "grad_norm": 10.543418121708857,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8223
+    },
+    {
+      "epoch": 0.3359497645211931,
+      "grad_norm": 7.89154755164855,
+      "learning_rate": 1e-06,
+      "loss": 0.0933,
+      "step": 8224
+    },
+    {
+      "epoch": 0.3390894819466248,
+      "grad_norm": 5.553749595133255,
+      "learning_rate": 1e-06,
+      "loss": 0.1058,
+      "step": 8225
+    },
+    {
+      "epoch": 0.34222919937205654,
+      "grad_norm": 8.236140718034283,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8226
+    },
+    {
+      "epoch": 0.3453689167974882,
+      "grad_norm": 7.102372348376214,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 8227
+    },
+    {
+      "epoch": 0.34850863422291994,
+      "grad_norm": 7.324736898245772,
+      "learning_rate": 1e-06,
+      "loss": 0.1016,
+      "step": 8228
+    },
+    {
+      "epoch": 0.3516483516483517,
+      "grad_norm": 9.20260678252701,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 8229
+    },
+    {
+      "epoch": 0.35478806907378335,
+      "grad_norm": 8.513232788652696,
+      "learning_rate": 1e-06,
+      "loss": 0.1062,
+      "step": 8230
+    },
+    {
+      "epoch": 0.3579277864992151,
+      "grad_norm": 7.243898725120383,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 8231
+    },
+    {
+      "epoch": 0.36106750392464676,
+      "grad_norm": 9.806059243138202,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 8232
+    },
+    {
+      "epoch": 0.3642072213500785,
+      "grad_norm": 10.234578775002014,
+      "learning_rate": 1e-06,
+      "loss": 0.1036,
+      "step": 8233
+    },
+    {
+      "epoch": 0.3673469387755102,
+      "grad_norm": 7.5465937456767955,
+      "learning_rate": 1e-06,
+      "loss": 0.1063,
+      "step": 8234
+    },
+    {
+      "epoch": 0.3704866562009419,
+      "grad_norm": 8.397942777245602,
+      "learning_rate": 1e-06,
+      "loss": 0.1189,
+      "step": 8235
+    },
+    {
+      "epoch": 0.37362637362637363,
+      "grad_norm": 8.797818436942935,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 8236
+    },
+    {
+      "epoch": 0.37676609105180536,
+      "grad_norm": 10.125325537020451,
+      "learning_rate": 1e-06,
+      "loss": 0.1167,
+      "step": 8237
+    },
+    {
+      "epoch": 0.37990580847723704,
+      "grad_norm": 8.369910254514117,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 8238
+    },
+    {
+      "epoch": 0.38304552590266877,
+      "grad_norm": 8.71145604379958,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8239
+    },
+    {
+      "epoch": 0.38618524332810045,
+      "grad_norm": 6.864744341587744,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 8240
+    },
+    {
+      "epoch": 0.3893249607535322,
+      "grad_norm": 8.725867502337142,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 8241
+    },
+    {
+      "epoch": 0.3924646781789639,
+      "grad_norm": 6.064003263210968,
+      "learning_rate": 1e-06,
+      "loss": 0.1051,
+      "step": 8242
+    },
+    {
+      "epoch": 0.3956043956043956,
+      "grad_norm": 7.8755904505020125,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 8243
+    },
+    {
+      "epoch": 0.3987441130298273,
+      "grad_norm": 8.227180694435857,
+      "learning_rate": 1e-06,
+      "loss": 0.1076,
+      "step": 8244
+    },
+    {
+      "epoch": 0.40188383045525905,
+      "grad_norm": 8.94110377753358,
+      "learning_rate": 1e-06,
+      "loss": 0.0983,
+      "step": 8245
+    },
+    {
+      "epoch": 0.4050235478806907,
+      "grad_norm": 8.883977410276689,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 8246
+    },
+    {
+      "epoch": 0.40816326530612246,
+      "grad_norm": 9.47513665466102,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 8247
+    },
+    {
+      "epoch": 0.41130298273155413,
+      "grad_norm": 7.671893363041628,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 8248
+    },
+    {
+      "epoch": 0.41444270015698587,
+      "grad_norm": 6.827119645727252,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8249
+    },
+    {
+      "epoch": 0.4175824175824176,
+      "grad_norm": 9.851114446198627,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 8250
+    },
+    {
+      "epoch": 0.4207221350078493,
+      "grad_norm": 8.166459160141889,
+      "learning_rate": 1e-06,
+      "loss": 0.1171,
+      "step": 8251
+    },
+    {
+      "epoch": 0.423861852433281,
+      "grad_norm": 8.57227748283951,
+      "learning_rate": 1e-06,
+      "loss": 0.1049,
+      "step": 8252
+    },
+    {
+      "epoch": 0.42700156985871274,
+      "grad_norm": 9.88701390578166,
+      "learning_rate": 1e-06,
+      "loss": 0.106,
+      "step": 8253
+    },
+    {
+      "epoch": 0.4301412872841444,
+      "grad_norm": 8.368478417522796,
+      "learning_rate": 1e-06,
+      "loss": 0.1028,
+      "step": 8254
+    },
+    {
+      "epoch": 0.43328100470957615,
+      "grad_norm": 7.836036834938156,
+      "learning_rate": 1e-06,
+      "loss": 0.1007,
+      "step": 8255
+    },
+    {
+      "epoch": 0.4364207221350079,
+      "grad_norm": 8.137924714547596,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 8256
+    },
+    {
+      "epoch": 0.43956043956043955,
+      "grad_norm": 7.601505266124318,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 8257
+    },
+    {
+      "epoch": 0.4427001569858713,
+      "grad_norm": 7.43688288162495,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 8258
+    },
+    {
+      "epoch": 0.44583987441130296,
+      "grad_norm": 6.691672991970506,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 8259
+    },
+    {
+      "epoch": 0.4489795918367347,
+      "grad_norm": 8.094891786197833,
+      "learning_rate": 1e-06,
+      "loss": 0.1138,
+      "step": 8260
+    },
+    {
+      "epoch": 0.4521193092621664,
+      "grad_norm": 6.7786768841603235,
+      "learning_rate": 1e-06,
+      "loss": 0.1048,
+      "step": 8261
+    },
+    {
+      "epoch": 0.4552590266875981,
+      "grad_norm": 9.194907364596714,
+      "learning_rate": 1e-06,
+      "loss": 0.0976,
+      "step": 8262
+    },
+    {
+      "epoch": 0.45839874411302983,
+      "grad_norm": 9.159440540828273,
+      "learning_rate": 1e-06,
+      "loss": 0.1034,
+      "step": 8263
+    },
+    {
+      "epoch": 0.46153846153846156,
+      "grad_norm": 7.140769655583965,
+      "learning_rate": 1e-06,
+      "loss": 0.1041,
+      "step": 8264
+    },
+    {
+      "epoch": 0.46467817896389324,
+      "grad_norm": 8.442579155680155,
+      "learning_rate": 1e-06,
+      "loss": 0.1045,
+      "step": 8265
+    },
+    {
+      "epoch": 0.46781789638932497,
+      "grad_norm": 7.54629468978974,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 8266
+    },
+    {
+      "epoch": 0.47095761381475665,
+      "grad_norm": 8.884648300638613,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 8267
+    },
+    {
+      "epoch": 0.4740973312401884,
+      "grad_norm": 9.924912522396498,
+      "learning_rate": 1e-06,
+      "loss": 0.1025,
+      "step": 8268
+    },
+    {
+      "epoch": 0.4772370486656201,
+      "grad_norm": 8.132514659476353,
+      "learning_rate": 1e-06,
+      "loss": 0.0983,
+      "step": 8269
+    },
+    {
+      "epoch": 0.4803767660910518,
+      "grad_norm": 9.152839812989995,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 8270
+    },
+    {
+      "epoch": 0.4835164835164835,
+      "grad_norm": 8.661018563305332,
+      "learning_rate": 1e-06,
+      "loss": 0.1,
+      "step": 8271
+    },
+    {
+      "epoch": 0.48665620094191525,
+      "grad_norm": 8.195091958781493,
+      "learning_rate": 1e-06,
+      "loss": 0.1011,
+      "step": 8272
+    },
+    {
+      "epoch": 0.4897959183673469,
+      "grad_norm": 8.434460168780793,
+      "learning_rate": 1e-06,
+      "loss": 0.1075,
+      "step": 8273
+    },
+    {
+      "epoch": 0.49293563579277866,
+      "grad_norm": 8.019518526677334,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 8274
+    },
+    {
+      "epoch": 0.49607535321821034,
+      "grad_norm": 6.603880332330038,
+      "learning_rate": 1e-06,
+      "loss": 0.1099,
+      "step": 8275
+    },
+    {
+      "epoch": 0.49921507064364207,
+      "grad_norm": 6.5124861351082854,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8276
+    },
+    {
+      "epoch": 0.5023547880690737,
+      "grad_norm": 6.80918599183059,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8277
+    },
+    {
+      "epoch": 0.5054945054945055,
+      "grad_norm": 6.35054473765149,
+      "learning_rate": 1e-06,
+      "loss": 0.1046,
+      "step": 8278
+    },
+    {
+      "epoch": 0.5086342229199372,
+      "grad_norm": 11.119995575415986,
+      "learning_rate": 1e-06,
+      "loss": 0.1101,
+      "step": 8279
+    },
+    {
+      "epoch": 0.5117739403453689,
+      "grad_norm": 6.308189973701426,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 8280
+    },
+    {
+      "epoch": 0.5149136577708007,
+      "grad_norm": 6.888180000924563,
+      "learning_rate": 1e-06,
+      "loss": 0.1058,
+      "step": 8281
+    },
+    {
+      "epoch": 0.5180533751962323,
+      "grad_norm": 9.214309843020237,
+      "learning_rate": 1e-06,
+      "loss": 0.1074,
+      "step": 8282
+    },
+    {
+      "epoch": 0.521193092621664,
+      "grad_norm": 7.4115298862626755,
+      "learning_rate": 1e-06,
+      "loss": 0.1137,
+      "step": 8283
+    },
+    {
+      "epoch": 0.5243328100470958,
+      "grad_norm": 7.93281710720701,
+      "learning_rate": 1e-06,
+      "loss": 0.1106,
+      "step": 8284
+    },
+    {
+      "epoch": 0.5274725274725275,
+      "grad_norm": 8.421684680108198,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8285
+    },
+    {
+      "epoch": 0.5306122448979592,
+      "grad_norm": 5.982854899312608,
+      "learning_rate": 1e-06,
+      "loss": 0.1025,
+      "step": 8286
+    },
+    {
+      "epoch": 0.533751962323391,
+      "grad_norm": 7.461758935949888,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 8287
+    },
+    {
+      "epoch": 0.5368916797488226,
+      "grad_norm": 8.732761897987373,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 8288
+    },
+    {
+      "epoch": 0.5400313971742543,
+      "grad_norm": 8.840159346232637,
+      "learning_rate": 1e-06,
+      "loss": 0.1058,
+      "step": 8289
+    },
+    {
+      "epoch": 0.543171114599686,
+      "grad_norm": 6.582734594028684,
+      "learning_rate": 1e-06,
+      "loss": 0.1021,
+      "step": 8290
+    },
+    {
+      "epoch": 0.5463108320251178,
+      "grad_norm": 7.754065097723962,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 8291
+    },
+    {
+      "epoch": 0.5494505494505495,
+      "grad_norm": 9.31468579303689,
+      "learning_rate": 1e-06,
+      "loss": 0.1043,
+      "step": 8292
+    },
+    {
+      "epoch": 0.5525902668759811,
+      "grad_norm": 5.763926035084116,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 8293
+    },
+    {
+      "epoch": 0.5557299843014128,
+      "grad_norm": 7.044221116905321,
+      "learning_rate": 1e-06,
+      "loss": 0.1009,
+      "step": 8294
+    },
+    {
+      "epoch": 0.5588697017268446,
+      "grad_norm": 7.10002590107703,
+      "learning_rate": 1e-06,
+      "loss": 0.1024,
+      "step": 8295
+    },
+    {
+      "epoch": 0.5620094191522763,
+      "grad_norm": 8.719915480862996,
+      "learning_rate": 1e-06,
+      "loss": 0.1084,
+      "step": 8296
+    },
+    {
+      "epoch": 0.565149136577708,
+      "grad_norm": 7.710603618584245,
+      "learning_rate": 1e-06,
+      "loss": 0.118,
+      "step": 8297
+    },
+    {
+      "epoch": 0.5682888540031397,
+      "grad_norm": 7.280064225308938,
+      "learning_rate": 1e-06,
+      "loss": 0.1029,
+      "step": 8298
+    },
+    {
+      "epoch": 0.5714285714285714,
+      "grad_norm": 5.9054163031276286,
+      "learning_rate": 1e-06,
+      "loss": 0.1087,
+      "step": 8299
+    },
+    {
+      "epoch": 0.5745682888540031,
+      "grad_norm": 9.683882708934464,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 8300
+    },
+    {
+      "epoch": 0.5777080062794349,
+      "grad_norm": 8.238423551582002,
+      "learning_rate": 1e-06,
+      "loss": 0.1123,
+      "step": 8301
+    },
+    {
+      "epoch": 0.5808477237048666,
+      "grad_norm": 6.523576708820898,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 8302
+    },
+    {
+      "epoch": 0.5839874411302983,
+      "grad_norm": 8.496573853068167,
+      "learning_rate": 1e-06,
+      "loss": 0.1176,
+      "step": 8303
+    },
+    {
+      "epoch": 0.5871271585557299,
+      "grad_norm": 9.760707379923591,
+      "learning_rate": 1e-06,
+      "loss": 0.1075,
+      "step": 8304
+    },
+    {
+      "epoch": 0.5902668759811617,
+      "grad_norm": 8.61834502665042,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 8305
+    },
+    {
+      "epoch": 0.5934065934065934,
+      "grad_norm": 8.059411627455749,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 8306
+    },
+    {
+      "epoch": 0.5965463108320251,
+      "grad_norm": 7.33076148437234,
+      "learning_rate": 1e-06,
+      "loss": 0.1086,
+      "step": 8307
+    },
+    {
+      "epoch": 0.5996860282574569,
+      "grad_norm": 8.861827729456907,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 8308
+    },
+    {
+      "epoch": 0.6028257456828885,
+      "grad_norm": 7.671766854986045,
+      "learning_rate": 1e-06,
+      "loss": 0.0996,
+      "step": 8309
+    },
+    {
+      "epoch": 0.6059654631083202,
+      "grad_norm": 6.110903927660954,
+      "learning_rate": 1e-06,
+      "loss": 0.1072,
+      "step": 8310
+    },
+    {
+      "epoch": 0.609105180533752,
+      "grad_norm": 6.790446516096176,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 8311
+    },
+    {
+      "epoch": 0.6122448979591837,
+      "grad_norm": 7.780999016936065,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 8312
+    },
+    {
+      "epoch": 0.6153846153846154,
+      "grad_norm": 10.175670129243409,
+      "learning_rate": 1e-06,
+      "loss": 0.112,
+      "step": 8313
+    },
+    {
+      "epoch": 0.6185243328100472,
+      "grad_norm": 7.234137606912939,
+      "learning_rate": 1e-06,
+      "loss": 0.1078,
+      "step": 8314
+    },
+    {
+      "epoch": 0.6216640502354788,
+      "grad_norm": 8.41402746700655,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 8315
+    },
+    {
+      "epoch": 0.6248037676609105,
+      "grad_norm": 8.049076677159366,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 8316
+    },
+    {
+      "epoch": 0.6279434850863422,
+      "grad_norm": 9.140302788448325,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 8317
+    },
+    {
+      "epoch": 0.631083202511774,
+      "grad_norm": 6.278115197022767,
+      "learning_rate": 1e-06,
+      "loss": 0.0973,
+      "step": 8318
+    },
+    {
+      "epoch": 0.6342229199372057,
+      "grad_norm": 7.234302528885519,
+      "learning_rate": 1e-06,
+      "loss": 0.115,
+      "step": 8319
+    },
+    {
+      "epoch": 0.6373626373626373,
+      "grad_norm": 8.102559631959936,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 8320
+    },
+    {
+      "epoch": 0.640502354788069,
+      "grad_norm": 9.43973603074764,
+      "learning_rate": 1e-06,
+      "loss": 0.1151,
+      "step": 8321
+    },
+    {
+      "epoch": 0.6436420722135008,
+      "grad_norm": 6.677102978870787,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8322
+    },
+    {
+      "epoch": 0.6467817896389325,
+      "grad_norm": 6.956223693668223,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 8323
+    },
+    {
+      "epoch": 0.6499215070643642,
+      "grad_norm": 8.477279070852633,
+      "learning_rate": 1e-06,
+      "loss": 0.0971,
+      "step": 8324
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 8.245694684052808,
+      "learning_rate": 1e-06,
+      "loss": 0.0969,
+      "step": 8325
+    },
+    {
+      "epoch": 0.6562009419152276,
+      "grad_norm": 8.093707715284419,
+      "learning_rate": 1e-06,
+      "loss": 0.1056,
+      "step": 8326
+    },
+    {
+      "epoch": 0.6593406593406593,
+      "grad_norm": 7.00628771720725,
+      "learning_rate": 1e-06,
+      "loss": 0.1128,
+      "step": 8327
+    },
+    {
+      "epoch": 0.6624803767660911,
+      "grad_norm": 5.762665145784075,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 8328
+    },
+    {
+      "epoch": 0.6656200941915228,
+      "grad_norm": 6.13072780421507,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 8329
+    },
+    {
+      "epoch": 0.6687598116169545,
+      "grad_norm": 9.31753645435285,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 8330
+    },
+    {
+      "epoch": 0.6718995290423861,
+      "grad_norm": 7.525586754211829,
+      "learning_rate": 1e-06,
+      "loss": 0.1066,
+      "step": 8331
+    },
+    {
+      "epoch": 0.6750392464678179,
+      "grad_norm": 8.766640270243206,
+      "learning_rate": 1e-06,
+      "loss": 0.1043,
+      "step": 8332
+    },
+    {
+      "epoch": 0.6781789638932496,
+      "grad_norm": 6.851072283142944,
+      "learning_rate": 1e-06,
+      "loss": 0.1104,
+      "step": 8333
+    },
+    {
+      "epoch": 0.6813186813186813,
+      "grad_norm": 8.786497035983126,
+      "learning_rate": 1e-06,
+      "loss": 0.1082,
+      "step": 8334
+    },
+    {
+      "epoch": 0.6844583987441131,
+      "grad_norm": 9.516649138594461,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 8335
+    },
+    {
+      "epoch": 0.6875981161695447,
+      "grad_norm": 6.886593704726748,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8336
+    },
+    {
+      "epoch": 0.6907378335949764,
+      "grad_norm": 7.497647825172497,
+      "learning_rate": 1e-06,
+      "loss": 0.1071,
+      "step": 8337
+    },
+    {
+      "epoch": 0.6938775510204082,
+      "grad_norm": 6.067440181339619,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 8338
+    },
+    {
+      "epoch": 0.6970172684458399,
+      "grad_norm": 7.68920095803338,
+      "learning_rate": 1e-06,
+      "loss": 0.1037,
+      "step": 8339
+    },
+    {
+      "epoch": 0.7001569858712716,
+      "grad_norm": 7.095336446848208,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 8340
+    },
+    {
+      "epoch": 0.7032967032967034,
+      "grad_norm": 7.207204654333947,
+      "learning_rate": 1e-06,
+      "loss": 0.1047,
+      "step": 8341
+    },
+    {
+      "epoch": 0.706436420722135,
+      "grad_norm": 7.377519360034569,
+      "learning_rate": 1e-06,
+      "loss": 0.1123,
+      "step": 8342
+    },
+    {
+      "epoch": 0.7095761381475667,
+      "grad_norm": 7.968424316932979,
+      "learning_rate": 1e-06,
+      "loss": 0.1042,
+      "step": 8343
+    },
+    {
+      "epoch": 0.7127158555729984,
+      "grad_norm": 7.591744991731952,
+      "learning_rate": 1e-06,
+      "loss": 0.1093,
+      "step": 8344
+    },
+    {
+      "epoch": 0.7158555729984302,
+      "grad_norm": 7.091818682928733,
+      "learning_rate": 1e-06,
+      "loss": 0.1025,
+      "step": 8345
+    },
+    {
+      "epoch": 0.7189952904238619,
+      "grad_norm": 9.148050947593406,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 8346
+    },
+    {
+      "epoch": 0.7221350078492935,
+      "grad_norm": 7.925865868109195,
+      "learning_rate": 1e-06,
+      "loss": 0.1017,
+      "step": 8347
+    },
+    {
+      "epoch": 0.7252747252747253,
+      "grad_norm": 8.990769379527578,
+      "learning_rate": 1e-06,
+      "loss": 0.1122,
+      "step": 8348
+    },
+    {
+      "epoch": 0.728414442700157,
+      "grad_norm": 7.287111911288224,
+      "learning_rate": 1e-06,
+      "loss": 0.1064,
+      "step": 8349
+    },
+    {
+      "epoch": 0.7315541601255887,
+      "grad_norm": 6.3198362629477325,
+      "learning_rate": 1e-06,
+      "loss": 0.1103,
+      "step": 8350
+    },
+    {
+      "epoch": 0.7346938775510204,
+      "grad_norm": 7.659927360299115,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 8351
+    },
+    {
+      "epoch": 0.7378335949764521,
+      "grad_norm": 9.429695380278892,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 8352
+    },
+    {
+      "epoch": 0.7409733124018838,
+      "grad_norm": 6.696302401879303,
+      "learning_rate": 1e-06,
+      "loss": 0.101,
+      "step": 8353
+    },
+    {
+      "epoch": 0.7441130298273155,
+      "grad_norm": 8.358732292492887,
+      "learning_rate": 1e-06,
+      "loss": 0.1183,
+      "step": 8354
+    },
+    {
+      "epoch": 0.7472527472527473,
+      "grad_norm": 7.438410246813748,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8355
+    },
+    {
+      "epoch": 0.750392464678179,
+      "grad_norm": 6.659943431493011,
+      "learning_rate": 1e-06,
+      "loss": 0.0998,
+      "step": 8356
+    },
+    {
+      "epoch": 0.7535321821036107,
+      "grad_norm": 8.977563016539332,
+      "learning_rate": 1e-06,
+      "loss": 0.1121,
+      "step": 8357
+    },
+    {
+      "epoch": 0.7566718995290423,
+      "grad_norm": 7.526950996490582,
+      "learning_rate": 1e-06,
+      "loss": 0.1143,
+      "step": 8358
+    },
+    {
+      "epoch": 0.7598116169544741,
+      "grad_norm": 9.203087491664265,
+      "learning_rate": 1e-06,
+      "loss": 0.1119,
+      "step": 8359
+    },
+    {
+      "epoch": 0.7629513343799058,
+      "grad_norm": 8.813496475732187,
+      "learning_rate": 1e-06,
+      "loss": 0.1015,
+      "step": 8360
+    },
+    {
+      "epoch": 0.7660910518053375,
+      "grad_norm": 8.631425314440635,
+      "learning_rate": 1e-06,
+      "loss": 0.1159,
+      "step": 8361
+    },
+    {
+      "epoch": 0.7692307692307693,
+      "grad_norm": 7.74020179284874,
+      "learning_rate": 1e-06,
+      "loss": 0.1152,
+      "step": 8362
+    },
+    {
+      "epoch": 0.7723704866562009,
+      "grad_norm": 8.184682687442645,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8363
+    },
+    {
+      "epoch": 0.7755102040816326,
+      "grad_norm": 7.17227398374944,
+      "learning_rate": 1e-06,
+      "loss": 0.1059,
+      "step": 8364
+    },
+    {
+      "epoch": 0.7786499215070644,
+      "grad_norm": 6.452937124146372,
+      "learning_rate": 1e-06,
+      "loss": 0.105,
+      "step": 8365
+    },
+    {
+      "epoch": 0.7817896389324961,
+      "grad_norm": 7.134723350986149,
+      "learning_rate": 1e-06,
+      "loss": 0.1013,
+      "step": 8366
+    },
+    {
+      "epoch": 0.7849293563579278,
+      "grad_norm": 6.680476079840052,
+      "learning_rate": 1e-06,
+      "loss": 0.1149,
+      "step": 8367
+    },
+    {
+      "epoch": 0.7880690737833596,
+      "grad_norm": 8.154880988349557,
+      "learning_rate": 1e-06,
+      "loss": 0.1198,
+      "step": 8368
+    },
+    {
+      "epoch": 0.7912087912087912,
+      "grad_norm": 7.970718749682879,
+      "learning_rate": 1e-06,
+      "loss": 0.1112,
+      "step": 8369
+    },
+    {
+      "epoch": 0.7943485086342229,
+      "grad_norm": 8.327527881096735,
+      "learning_rate": 1e-06,
+      "loss": 0.1095,
+      "step": 8370
+    },
+    {
+      "epoch": 0.7974882260596546,
+      "grad_norm": 7.686971800473386,
+      "learning_rate": 1e-06,
+      "loss": 0.11,
+      "step": 8371
+    },
+    {
+      "epoch": 0.8006279434850864,
+      "grad_norm": 7.665097605352539,
+      "learning_rate": 1e-06,
+      "loss": 0.114,
+      "step": 8372
+    },
+    {
+      "epoch": 0.8037676609105181,
+      "grad_norm": 7.394446343483226,
+      "learning_rate": 1e-06,
+      "loss": 0.1115,
+      "step": 8373
+    },
+    {
+      "epoch": 0.8069073783359497,
+      "grad_norm": 7.479984567637229,
+      "learning_rate": 1e-06,
+      "loss": 0.1139,
+      "step": 8374
+    },
+    {
+      "epoch": 0.8100470957613815,
+      "grad_norm": 5.467134300386294,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 8375
+    },
+    {
+      "epoch": 0.8131868131868132,
+      "grad_norm": 7.753906024906133,
+      "learning_rate": 1e-06,
+      "loss": 0.1024,
+      "step": 8376
+    },
+    {
+      "epoch": 0.8163265306122449,
+      "grad_norm": 8.499342449406232,
+      "learning_rate": 1e-06,
+      "loss": 0.1129,
+      "step": 8377
+    },
+    {
+      "epoch": 0.8194662480376766,
+      "grad_norm": 9.087619900027235,
+      "learning_rate": 1e-06,
+      "loss": 0.1095,
+      "step": 8378
+    },
+    {
+      "epoch": 0.8226059654631083,
+      "grad_norm": 8.021058366954907,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 8379
+    },
+    {
+      "epoch": 0.82574568288854,
+      "grad_norm": 9.314238100418079,
+      "learning_rate": 1e-06,
+      "loss": 0.1173,
+      "step": 8380
+    },
+    {
+      "epoch": 0.8288854003139717,
+      "grad_norm": 7.143819099175362,
+      "learning_rate": 1e-06,
+      "loss": 0.1191,
+      "step": 8381
+    },
+    {
+      "epoch": 0.8320251177394035,
+      "grad_norm": 8.671815732455917,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 8382
+    },
+    {
+      "epoch": 0.8351648351648352,
+      "grad_norm": 7.992187332547653,
+      "learning_rate": 1e-06,
+      "loss": 0.1108,
+      "step": 8383
+    },
+    {
+      "epoch": 0.8383045525902669,
+      "grad_norm": 7.377115649717169,
+      "learning_rate": 1e-06,
+      "loss": 0.0987,
+      "step": 8384
+    },
+    {
+      "epoch": 0.8414442700156985,
+      "grad_norm": 7.367053213232793,
+      "learning_rate": 1e-06,
+      "loss": 0.1066,
+      "step": 8385
+    },
+    {
+      "epoch": 0.8445839874411303,
+      "grad_norm": 7.399027814006133,
+      "learning_rate": 1e-06,
+      "loss": 0.1042,
+      "step": 8386
+    },
+    {
+      "epoch": 0.847723704866562,
+      "grad_norm": 6.857089626450962,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 8387
+    },
+    {
+      "epoch": 0.8508634222919937,
+      "grad_norm": 7.118427402528536,
+      "learning_rate": 1e-06,
+      "loss": 0.1014,
+      "step": 8388
+    },
+    {
+      "epoch": 0.8540031397174255,
+      "grad_norm": 7.639113336579951,
+      "learning_rate": 1e-06,
+      "loss": 0.1085,
+      "step": 8389
+    },
+    {
+      "epoch": 0.8571428571428571,
+      "grad_norm": 5.819744992711068,
+      "learning_rate": 1e-06,
+      "loss": 0.1073,
+      "step": 8390
+    },
+    {
+      "epoch": 0.8602825745682888,
+      "grad_norm": 6.386512550135361,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 8391
+    },
+    {
+      "epoch": 0.8634222919937206,
+      "grad_norm": 9.644772918446414,
+      "learning_rate": 1e-06,
+      "loss": 0.1111,
+      "step": 8392
+    },
+    {
+      "epoch": 0.8665620094191523,
+      "grad_norm": 8.408646535313432,
+      "learning_rate": 1e-06,
+      "loss": 0.1017,
+      "step": 8393
+    },
+    {
+      "epoch": 0.869701726844584,
+      "grad_norm": 7.561870677119412,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 8394
+    },
+    {
+      "epoch": 0.8728414442700158,
+      "grad_norm": 8.412548867411362,
+      "learning_rate": 1e-06,
+      "loss": 0.1147,
+      "step": 8395
+    },
+    {
+      "epoch": 0.8759811616954474,
+      "grad_norm": 9.954985514001503,
+      "learning_rate": 1e-06,
+      "loss": 0.1173,
+      "step": 8396
+    },
+    {
+      "epoch": 0.8791208791208791,
+      "grad_norm": 8.708399785994192,
+      "learning_rate": 1e-06,
+      "loss": 0.1079,
+      "step": 8397
+    },
+    {
+      "epoch": 0.8822605965463108,
+      "grad_norm": 5.871774941287088,
+      "learning_rate": 1e-06,
+      "loss": 0.1118,
+      "step": 8398
+    },
+    {
+      "epoch": 0.8854003139717426,
+      "grad_norm": 10.723438102566204,
+      "learning_rate": 1e-06,
+      "loss": 0.1105,
+      "step": 8399
+    },
+    {
+      "epoch": 0.8885400313971743,
+      "grad_norm": 8.130942770161216,
+      "learning_rate": 1e-06,
+      "loss": 0.1098,
+      "step": 8400
+    },
+    {
+      "epoch": 0.8916797488226059,
+      "grad_norm": 6.203295122291504,
+      "learning_rate": 1e-06,
+      "loss": 0.1117,
+      "step": 8401
+    },
+    {
+      "epoch": 0.8948194662480377,
+      "grad_norm": 7.251211832545261,
+      "learning_rate": 1e-06,
+      "loss": 0.1154,
+      "step": 8402
+    },
+    {
+      "epoch": 0.8979591836734694,
+      "grad_norm": 6.807991024595893,
+      "learning_rate": 1e-06,
+      "loss": 0.1019,
+      "step": 8403
+    },
+    {
+      "epoch": 0.9010989010989011,
+      "grad_norm": 7.238205790994883,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8404
+    },
+    {
+      "epoch": 0.9042386185243328,
+      "grad_norm": 7.991979216618561,
+      "learning_rate": 1e-06,
+      "loss": 0.1192,
+      "step": 8405
+    },
+    {
+      "epoch": 0.9073783359497645,
+      "grad_norm": 8.409936524337823,
+      "learning_rate": 1e-06,
+      "loss": 0.1148,
+      "step": 8406
+    },
+    {
+      "epoch": 0.9105180533751962,
+      "grad_norm": 8.334771679353631,
+      "learning_rate": 1e-06,
+      "loss": 0.1182,
+      "step": 8407
+    },
+    {
+      "epoch": 0.9136577708006279,
+      "grad_norm": 7.705214022542738,
+      "learning_rate": 1e-06,
+      "loss": 0.108,
+      "step": 8408
+    },
+    {
+      "epoch": 0.9167974882260597,
+      "grad_norm": 7.060123651118234,
+      "learning_rate": 1e-06,
+      "loss": 0.1174,
+      "step": 8409
+    },
+    {
+      "epoch": 0.9199372056514914,
+      "grad_norm": 6.556393903079437,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 8410
+    },
+    {
+      "epoch": 0.9230769230769231,
+      "grad_norm": 8.585970030949467,
+      "learning_rate": 1e-06,
+      "loss": 0.1126,
+      "step": 8411
+    },
+    {
+      "epoch": 0.9262166405023547,
+      "grad_norm": 6.66818825068754,
+      "learning_rate": 1e-06,
+      "loss": 0.1051,
+      "step": 8412
+    },
+    {
+      "epoch": 0.9293563579277865,
+      "grad_norm": 7.267307101190777,
+      "learning_rate": 1e-06,
+      "loss": 0.0997,
+      "step": 8413
+    },
+    {
+      "epoch": 0.9324960753532182,
+      "grad_norm": 8.873357322139178,
+      "learning_rate": 1e-06,
+      "loss": 0.1089,
+      "step": 8414
+    },
+    {
+      "epoch": 0.9356357927786499,
+      "grad_norm": 7.795096423948594,
+      "learning_rate": 1e-06,
+      "loss": 0.109,
+      "step": 8415
+    },
+    {
+      "epoch": 0.9387755102040817,
+      "grad_norm": 6.688046858856093,
+      "learning_rate": 1e-06,
+      "loss": 0.1077,
+      "step": 8416
+    },
+    {
+      "epoch": 0.9419152276295133,
+      "grad_norm": 6.757413888211312,
+      "learning_rate": 1e-06,
+      "loss": 0.1018,
+      "step": 8417
+    },
+    {
+      "epoch": 0.945054945054945,
+      "grad_norm": 7.710971428797476,
+      "learning_rate": 1e-06,
+      "loss": 0.1133,
+      "step": 8418
+    },
+    {
+      "epoch": 0.9481946624803768,
+      "grad_norm": 8.130481067037364,
+      "learning_rate": 1e-06,
+      "loss": 0.1128,
+      "step": 8419
+    },
+    {
+      "epoch": 0.9513343799058085,
+      "grad_norm": 6.931469714043536,
+      "learning_rate": 1e-06,
+      "loss": 0.0946,
+      "step": 8420
+    },
+    {
+      "epoch": 0.9544740973312402,
+      "grad_norm": 6.132651047627706,
+      "learning_rate": 1e-06,
+      "loss": 0.1169,
+      "step": 8421
+    },
+    {
+      "epoch": 0.957613814756672,
+      "grad_norm": 8.34009181858081,
+      "learning_rate": 1e-06,
+      "loss": 0.1127,
+      "step": 8422
+    },
+    {
+      "epoch": 0.9607535321821036,
+      "grad_norm": 7.78322011420479,
+      "learning_rate": 1e-06,
+      "loss": 0.1102,
+      "step": 8423
+    },
+    {
+      "epoch": 0.9638932496075353,
+      "grad_norm": 7.810877757326631,
+      "learning_rate": 1e-06,
+      "loss": 0.1052,
+      "step": 8424
+    },
+    {
+      "epoch": 0.967032967032967,
+      "grad_norm": 8.701390147816248,
+      "learning_rate": 1e-06,
+      "loss": 0.1162,
+      "step": 8425
+    },
+    {
+      "epoch": 0.9701726844583988,
+      "grad_norm": 7.669627936324654,
+      "learning_rate": 1e-06,
+      "loss": 0.0988,
+      "step": 8426
+    },
+    {
+      "epoch": 0.9733124018838305,
+      "grad_norm": 7.916744182385857,
+      "learning_rate": 1e-06,
+      "loss": 0.1088,
+      "step": 8427
+    },
+    {
+      "epoch": 0.9764521193092621,
+      "grad_norm": 6.226958825835985,
+      "learning_rate": 1e-06,
+      "loss": 0.1141,
+      "step": 8428
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 8.991812113116108,
+      "learning_rate": 1e-06,
+      "loss": 0.1114,
+      "step": 8429
+    },
+    {
+      "epoch": 0.9827315541601256,
+      "grad_norm": 7.491242105590485,
+      "learning_rate": 1e-06,
+      "loss": 0.1166,
+      "step": 8430
+    },
+    {
+      "epoch": 0.9858712715855573,
+      "grad_norm": 9.242334556450086,
+      "learning_rate": 1e-06,
+      "loss": 0.1143,
+      "step": 8431
+    },
+    {
+      "epoch": 0.989010989010989,
+      "grad_norm": 6.118840541834246,
+      "learning_rate": 1e-06,
+      "loss": 0.1096,
+      "step": 8432
+    },
+    {
+      "epoch": 0.9921507064364207,
+      "grad_norm": 6.0880425433160195,
+      "learning_rate": 1e-06,
+      "loss": 0.1057,
+      "step": 8433
+    },
+    {
+      "epoch": 0.9952904238618524,
+      "grad_norm": 9.359123045691804,
+      "learning_rate": 1e-06,
+      "loss": 0.105,
+      "step": 8434
+    },
+    {
+      "epoch": 0.9984301412872841,
+      "grad_norm": 6.38197313770872,
+      "learning_rate": 1e-06,
+      "loss": 0.1067,
+      "step": 8435
+    },
+    {
+      "epoch": 0.9984301412872841,
+      "eval_loss": 4.861870288848877,
+      "eval_runtime": 21.82,
+      "eval_samples_per_second": 45.83,
+      "eval_steps_per_second": 5.729,
+      "step": 8435
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 8435,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 8435,
+  "stateful_callbacks": {},
+  "total_flos": 4.380855757721253e+19,
+  "train_batch_size": 8,
+  "trial_name": null,
+  "trial_params": null
+}