diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,15239 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998560778331078,
+  "eval_steps": 500,
+  "global_step": 2171,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00046055093405486313,
+      "grad_norm": 10.224479026613773,
+      "learning_rate": 3.0303030303030305e-08,
+      "loss": 1.084,
+      "step": 1
+    },
+    {
+      "epoch": 0.0009211018681097263,
+      "grad_norm": 10.903636883871654,
+      "learning_rate": 6.060606060606061e-08,
+      "loss": 0.9824,
+      "step": 2
+    },
+    {
+      "epoch": 0.0013816528021645895,
+      "grad_norm": 11.344940351126398,
+      "learning_rate": 9.09090909090909e-08,
+      "loss": 1.0421,
+      "step": 3
+    },
+    {
+      "epoch": 0.0018422037362194525,
+      "grad_norm": 13.25793073962023,
+      "learning_rate": 1.2121212121212122e-07,
+      "loss": 0.9886,
+      "step": 4
+    },
+    {
+      "epoch": 0.002302754670274316,
+      "grad_norm": 9.70140482967039,
+      "learning_rate": 1.5151515151515152e-07,
+      "loss": 0.9634,
+      "step": 5
+    },
+    {
+      "epoch": 0.002763305604329179,
+      "grad_norm": 12.03838558755089,
+      "learning_rate": 1.818181818181818e-07,
+      "loss": 0.9483,
+      "step": 6
+    },
+    {
+      "epoch": 0.003223856538384042,
+      "grad_norm": 12.588391333344033,
+      "learning_rate": 2.121212121212121e-07,
+      "loss": 0.9984,
+      "step": 7
+    },
+    {
+      "epoch": 0.003684407472438905,
+      "grad_norm": 11.836562687027973,
+      "learning_rate": 2.4242424242424244e-07,
+      "loss": 1.0708,
+      "step": 8
+    },
+    {
+      "epoch": 0.004144958406493768,
+      "grad_norm": 12.706113691995517,
+      "learning_rate": 2.727272727272727e-07,
+      "loss": 1.0863,
+      "step": 9
+    },
+    {
+      "epoch": 0.004605509340548632,
+      "grad_norm": 11.407622369949724,
+      "learning_rate": 3.0303030303030305e-07,
+      "loss": 1.2087,
+      "step": 10
+    },
+    {
+      "epoch": 0.005066060274603494,
+      "grad_norm": 12.237310594074764,
+      "learning_rate": 3.333333333333333e-07,
+      "loss": 1.0967,
+      "step": 11
+    },
+    {
+      "epoch": 0.005526611208658358,
+      "grad_norm": 10.99654178909918,
+      "learning_rate": 3.636363636363636e-07,
+      "loss": 1.1732,
+      "step": 12
+    },
+    {
+      "epoch": 0.0059871621427132204,
+      "grad_norm": 11.042210513697741,
+      "learning_rate": 3.939393939393939e-07,
+      "loss": 1.1777,
+      "step": 13
+    },
+    {
+      "epoch": 0.006447713076768084,
+      "grad_norm": 11.496495867822293,
+      "learning_rate": 4.242424242424242e-07,
+      "loss": 1.0627,
+      "step": 14
+    },
+    {
+      "epoch": 0.006908264010822947,
+      "grad_norm": 11.882894204754962,
+      "learning_rate": 4.545454545454545e-07,
+      "loss": 0.9844,
+      "step": 15
+    },
+    {
+      "epoch": 0.00736881494487781,
+      "grad_norm": 11.794166405407287,
+      "learning_rate": 4.848484848484849e-07,
+      "loss": 1.0168,
+      "step": 16
+    },
+    {
+      "epoch": 0.007829365878932673,
+      "grad_norm": 10.01774185514084,
+      "learning_rate": 5.151515151515151e-07,
+      "loss": 1.2505,
+      "step": 17
+    },
+    {
+      "epoch": 0.008289916812987536,
+      "grad_norm": 9.567696400885666,
+      "learning_rate": 5.454545454545454e-07,
+      "loss": 1.1344,
+      "step": 18
+    },
+    {
+      "epoch": 0.0087504677470424,
+      "grad_norm": 11.67769227316676,
+      "learning_rate": 5.757575757575758e-07,
+      "loss": 0.9009,
+      "step": 19
+    },
+    {
+      "epoch": 0.009211018681097263,
+      "grad_norm": 9.530138294227273,
+      "learning_rate": 6.060606060606061e-07,
+      "loss": 1.0289,
+      "step": 20
+    },
+    {
+      "epoch": 0.009671569615152125,
+      "grad_norm": 8.071657974205257,
+      "learning_rate": 6.363636363636363e-07,
+      "loss": 1.2151,
+      "step": 21
+    },
+    {
+      "epoch": 0.010132120549206989,
+      "grad_norm": 11.50961336906488,
+      "learning_rate": 6.666666666666666e-07,
+      "loss": 0.8917,
+      "step": 22
+    },
+    {
+      "epoch": 0.010592671483261852,
+      "grad_norm": 9.46367673420139,
+      "learning_rate": 6.96969696969697e-07,
+      "loss": 1.1334,
+      "step": 23
+    },
+    {
+      "epoch": 0.011053222417316716,
+      "grad_norm": 7.980897491793901,
+      "learning_rate": 7.272727272727272e-07,
+      "loss": 0.968,
+      "step": 24
+    },
+    {
+      "epoch": 0.011513773351371579,
+      "grad_norm": 6.49691041072656,
+      "learning_rate": 7.575757575757575e-07,
+      "loss": 1.0277,
+      "step": 25
+    },
+    {
+      "epoch": 0.011974324285426441,
+      "grad_norm": 6.041354612689945,
+      "learning_rate": 7.878787878787878e-07,
+      "loss": 0.9477,
+      "step": 26
+    },
+    {
+      "epoch": 0.012434875219481304,
+      "grad_norm": 6.34623067352923,
+      "learning_rate": 8.181818181818182e-07,
+      "loss": 0.9635,
+      "step": 27
+    },
+    {
+      "epoch": 0.012895426153536168,
+      "grad_norm": 5.768779273607869,
+      "learning_rate": 8.484848484848484e-07,
+      "loss": 0.9567,
+      "step": 28
+    },
+    {
+      "epoch": 0.013355977087591031,
+      "grad_norm": 5.159941543576065,
+      "learning_rate": 8.787878787878787e-07,
+      "loss": 0.8742,
+      "step": 29
+    },
+    {
+      "epoch": 0.013816528021645893,
+      "grad_norm": 4.459623755013343,
+      "learning_rate": 9.09090909090909e-07,
+      "loss": 0.905,
+      "step": 30
+    },
+    {
+      "epoch": 0.014277078955700757,
+      "grad_norm": 4.578386981237651,
+      "learning_rate": 9.393939393939395e-07,
+      "loss": 0.9856,
+      "step": 31
+    },
+    {
+      "epoch": 0.01473762988975562,
+      "grad_norm": 4.599340605830885,
+      "learning_rate": 9.696969696969698e-07,
+      "loss": 1.0623,
+      "step": 32
+    },
+    {
+      "epoch": 0.015198180823810484,
+      "grad_norm": 4.715495998419273,
+      "learning_rate": 1e-06,
+      "loss": 0.9893,
+      "step": 33
+    },
+    {
+      "epoch": 0.015658731757865346,
+      "grad_norm": 3.9725156691925405,
+      "learning_rate": 1.0303030303030302e-06,
+      "loss": 0.9039,
+      "step": 34
+    },
+    {
+      "epoch": 0.01611928269192021,
+      "grad_norm": 4.01091093069899,
+      "learning_rate": 1.0606060606060606e-06,
+      "loss": 0.7272,
+      "step": 35
+    },
+    {
+      "epoch": 0.016579833625975073,
+      "grad_norm": 4.427996895865591,
+      "learning_rate": 1.0909090909090908e-06,
+      "loss": 0.8846,
+      "step": 36
+    },
+    {
+      "epoch": 0.017040384560029934,
+      "grad_norm": 3.545754336884005,
+      "learning_rate": 1.121212121212121e-06,
+      "loss": 0.829,
+      "step": 37
+    },
+    {
+      "epoch": 0.0175009354940848,
+      "grad_norm": 5.525053440753492,
+      "learning_rate": 1.1515151515151516e-06,
+      "loss": 0.6791,
+      "step": 38
+    },
+    {
+      "epoch": 0.01796148642813966,
+      "grad_norm": 3.383199059764453,
+      "learning_rate": 1.1818181818181818e-06,
+      "loss": 0.8036,
+      "step": 39
+    },
+    {
+      "epoch": 0.018422037362194527,
+      "grad_norm": 4.1135930428353,
+      "learning_rate": 1.2121212121212122e-06,
+      "loss": 0.89,
+      "step": 40
+    },
+    {
+      "epoch": 0.01888258829624939,
+      "grad_norm": 4.200860213314929,
+      "learning_rate": 1.2424242424242424e-06,
+      "loss": 0.8742,
+      "step": 41
+    },
+    {
+      "epoch": 0.01934313923030425,
+      "grad_norm": 4.621032204910162,
+      "learning_rate": 1.2727272727272726e-06,
+      "loss": 0.9061,
+      "step": 42
+    },
+    {
+      "epoch": 0.019803690164359115,
+      "grad_norm": 4.05937719343738,
+      "learning_rate": 1.303030303030303e-06,
+      "loss": 0.8869,
+      "step": 43
+    },
+    {
+      "epoch": 0.020264241098413977,
+      "grad_norm": 3.908268562451142,
+      "learning_rate": 1.3333333333333332e-06,
+      "loss": 0.8652,
+      "step": 44
+    },
+    {
+      "epoch": 0.020724792032468842,
+      "grad_norm": 3.8127642183113104,
+      "learning_rate": 1.3636363636363634e-06,
+      "loss": 0.8547,
+      "step": 45
+    },
+    {
+      "epoch": 0.021185342966523704,
+      "grad_norm": 3.7006255903086016,
+      "learning_rate": 1.393939393939394e-06,
+      "loss": 0.8816,
+      "step": 46
+    },
+    {
+      "epoch": 0.021645893900578566,
+      "grad_norm": 4.278643366823466,
+      "learning_rate": 1.4242424242424242e-06,
+      "loss": 0.9507,
+      "step": 47
+    },
+    {
+      "epoch": 0.02210644483463343,
+      "grad_norm": 5.121316020450439,
+      "learning_rate": 1.4545454545454544e-06,
+      "loss": 0.8144,
+      "step": 48
+    },
+    {
+      "epoch": 0.022566995768688293,
+      "grad_norm": 3.048465483366566,
+      "learning_rate": 1.4848484848484848e-06,
+      "loss": 0.654,
+      "step": 49
+    },
+    {
+      "epoch": 0.023027546702743158,
+      "grad_norm": 3.9812784832535133,
+      "learning_rate": 1.515151515151515e-06,
+      "loss": 0.6679,
+      "step": 50
+    },
+    {
+      "epoch": 0.02348809763679802,
+      "grad_norm": 3.787536459284705,
+      "learning_rate": 1.5454545454545454e-06,
+      "loss": 0.9465,
+      "step": 51
+    },
+    {
+      "epoch": 0.023948648570852882,
+      "grad_norm": 3.6018385428584123,
+      "learning_rate": 1.5757575757575756e-06,
+      "loss": 0.7828,
+      "step": 52
+    },
+    {
+      "epoch": 0.024409199504907747,
+      "grad_norm": 3.5109451677027614,
+      "learning_rate": 1.6060606060606058e-06,
+      "loss": 0.8131,
+      "step": 53
+    },
+    {
+      "epoch": 0.02486975043896261,
+      "grad_norm": 3.0290570716615903,
+      "learning_rate": 1.6363636363636365e-06,
+      "loss": 0.6437,
+      "step": 54
+    },
+    {
+      "epoch": 0.02533030137301747,
+      "grad_norm": 3.682837755403324,
+      "learning_rate": 1.6666666666666667e-06,
+      "loss": 0.7215,
+      "step": 55
+    },
+    {
+      "epoch": 0.025790852307072336,
+      "grad_norm": 3.476439822446147,
+      "learning_rate": 1.6969696969696969e-06,
+      "loss": 0.7884,
+      "step": 56
+    },
+    {
+      "epoch": 0.026251403241127198,
+      "grad_norm": 3.9813216927132475,
+      "learning_rate": 1.7272727272727273e-06,
+      "loss": 0.7466,
+      "step": 57
+    },
+    {
+      "epoch": 0.026711954175182063,
+      "grad_norm": 3.6806123465375116,
+      "learning_rate": 1.7575757575757575e-06,
+      "loss": 0.7043,
+      "step": 58
+    },
+    {
+      "epoch": 0.027172505109236925,
+      "grad_norm": 3.1611663708183344,
+      "learning_rate": 1.7878787878787877e-06,
+      "loss": 0.7341,
+      "step": 59
+    },
+    {
+      "epoch": 0.027633056043291786,
+      "grad_norm": 3.444488689830893,
+      "learning_rate": 1.818181818181818e-06,
+      "loss": 0.8762,
+      "step": 60
+    },
+    {
+      "epoch": 0.02809360697734665,
+      "grad_norm": 3.549558394232681,
+      "learning_rate": 1.8484848484848483e-06,
+      "loss": 0.8806,
+      "step": 61
+    },
+    {
+      "epoch": 0.028554157911401513,
+      "grad_norm": 3.9711382750959747,
+      "learning_rate": 1.878787878787879e-06,
+      "loss": 0.8029,
+      "step": 62
+    },
+    {
+      "epoch": 0.02901470884545638,
+      "grad_norm": 3.5959357584947615,
+      "learning_rate": 1.909090909090909e-06,
+      "loss": 0.7292,
+      "step": 63
+    },
+    {
+      "epoch": 0.02947525977951124,
+      "grad_norm": 3.9415569403625024,
+      "learning_rate": 1.9393939393939395e-06,
+      "loss": 0.856,
+      "step": 64
+    },
+    {
+      "epoch": 0.029935810713566102,
+      "grad_norm": 3.406751239525263,
+      "learning_rate": 1.9696969696969695e-06,
+      "loss": 0.7986,
+      "step": 65
+    },
+    {
+      "epoch": 0.030396361647620967,
+      "grad_norm": 3.6931502231787183,
+      "learning_rate": 2e-06,
+      "loss": 0.6504,
+      "step": 66
+    },
+    {
+      "epoch": 0.03085691258167583,
+      "grad_norm": 3.940678039571864,
+      "learning_rate": 1.9999988863070544e-06,
+      "loss": 0.6272,
+      "step": 67
+    },
+    {
+      "epoch": 0.03131746351573069,
+      "grad_norm": 3.361818700666645,
+      "learning_rate": 1.999995545230698e-06,
+      "loss": 0.8018,
+      "step": 68
+    },
+    {
+      "epoch": 0.031778014449785556,
+      "grad_norm": 4.745930001970848,
+      "learning_rate": 1.9999899767783724e-06,
+      "loss": 0.7793,
+      "step": 69
+    },
+    {
+      "epoch": 0.03223856538384042,
+      "grad_norm": 4.566864649641224,
+      "learning_rate": 1.999982180962482e-06,
+      "loss": 0.7536,
+      "step": 70
+    },
+    {
+      "epoch": 0.03269911631789528,
+      "grad_norm": 3.4766274202922123,
+      "learning_rate": 1.999972157800389e-06,
+      "loss": 0.6682,
+      "step": 71
+    },
+    {
+      "epoch": 0.033159667251950145,
+      "grad_norm": 3.2407423987046564,
+      "learning_rate": 1.999959907314421e-06,
+      "loss": 0.5877,
+      "step": 72
+    },
+    {
+      "epoch": 0.03362021818600501,
+      "grad_norm": 3.854719893893145,
+      "learning_rate": 1.999945429531863e-06,
+      "loss": 0.8257,
+      "step": 73
+    },
+    {
+      "epoch": 0.03408076912005987,
+      "grad_norm": 3.384850197913963,
+      "learning_rate": 1.9999287244849633e-06,
+      "loss": 0.7341,
+      "step": 74
+    },
+    {
+      "epoch": 0.034541320054114734,
+      "grad_norm": 3.826326260538933,
+      "learning_rate": 1.9999097922109303e-06,
+      "loss": 0.7884,
+      "step": 75
+    },
+    {
+      "epoch": 0.0350018709881696,
+      "grad_norm": 3.5685218916656805,
+      "learning_rate": 1.9998886327519336e-06,
+      "loss": 0.7556,
+      "step": 76
+    },
+    {
+      "epoch": 0.035462421922224464,
+      "grad_norm": 3.314138350796672,
+      "learning_rate": 1.999865246155103e-06,
+      "loss": 0.7505,
+      "step": 77
+    },
+    {
+      "epoch": 0.03592297285627932,
+      "grad_norm": 3.218790293838154,
+      "learning_rate": 1.9998396324725305e-06,
+      "loss": 0.6178,
+      "step": 78
+    },
+    {
+      "epoch": 0.03638352379033419,
+      "grad_norm": 3.5441600733921055,
+      "learning_rate": 1.999811791761267e-06,
+      "loss": 0.711,
+      "step": 79
+    },
+    {
+      "epoch": 0.03684407472438905,
+      "grad_norm": 3.2797802004892773,
+      "learning_rate": 1.999781724083324e-06,
+      "loss": 0.6754,
+      "step": 80
+    },
+    {
+      "epoch": 0.03730462565844391,
+      "grad_norm": 3.514532124195952,
+      "learning_rate": 1.9997494295056746e-06,
+      "loss": 0.7056,
+      "step": 81
+    },
+    {
+      "epoch": 0.03776517659249878,
+      "grad_norm": 3.4677096105337433,
+      "learning_rate": 1.9997149081002514e-06,
+      "loss": 0.6394,
+      "step": 82
+    },
+    {
+      "epoch": 0.03822572752655364,
+      "grad_norm": 2.932267254260553,
+      "learning_rate": 1.9996781599439464e-06,
+      "loss": 0.7235,
+      "step": 83
+    },
+    {
+      "epoch": 0.0386862784606085,
+      "grad_norm": 3.096434205139558,
+      "learning_rate": 1.9996391851186118e-06,
+      "loss": 0.674,
+      "step": 84
+    },
+    {
+      "epoch": 0.039146829394663366,
+      "grad_norm": 3.858101988445306,
+      "learning_rate": 1.99959798371106e-06,
+      "loss": 0.6127,
+      "step": 85
+    },
+    {
+      "epoch": 0.03960738032871823,
+      "grad_norm": 3.361207098481145,
+      "learning_rate": 1.999554555813062e-06,
+      "loss": 0.6981,
+      "step": 86
+    },
+    {
+      "epoch": 0.04006793126277309,
+      "grad_norm": 3.4015641056770503,
+      "learning_rate": 1.9995089015213493e-06,
+      "loss": 0.8552,
+      "step": 87
+    },
+    {
+      "epoch": 0.040528482196827954,
+      "grad_norm": 3.1305710457612004,
+      "learning_rate": 1.999461020937611e-06,
+      "loss": 0.6014,
+      "step": 88
+    },
+    {
+      "epoch": 0.04098903313088282,
+      "grad_norm": 3.820781969106205,
+      "learning_rate": 1.999410914168495e-06,
+      "loss": 0.8311,
+      "step": 89
+    },
+    {
+      "epoch": 0.041449584064937685,
+      "grad_norm": 3.2079786714114906,
+      "learning_rate": 1.99935858132561e-06,
+      "loss": 0.682,
+      "step": 90
+    },
+    {
+      "epoch": 0.04191013499899254,
+      "grad_norm": 3.643489327427482,
+      "learning_rate": 1.99930402252552e-06,
+      "loss": 0.6605,
+      "step": 91
+    },
+    {
+      "epoch": 0.04237068593304741,
+      "grad_norm": 3.3738758807758766,
+      "learning_rate": 1.9992472378897497e-06,
+      "loss": 0.6143,
+      "step": 92
+    },
+    {
+      "epoch": 0.042831236867102274,
+      "grad_norm": 3.077610620836017,
+      "learning_rate": 1.9991882275447794e-06,
+      "loss": 0.6217,
+      "step": 93
+    },
+    {
+      "epoch": 0.04329178780115713,
+      "grad_norm": 3.162135055961725,
+      "learning_rate": 1.9991269916220485e-06,
+      "loss": 0.7265,
+      "step": 94
+    },
+    {
+      "epoch": 0.043752338735212,
+      "grad_norm": 2.9823928585506647,
+      "learning_rate": 1.999063530257952e-06,
+      "loss": 0.6289,
+      "step": 95
+    },
+    {
+      "epoch": 0.04421288966926686,
+      "grad_norm": 3.511247382327224,
+      "learning_rate": 1.998997843593845e-06,
+      "loss": 0.6145,
+      "step": 96
+    },
+    {
+      "epoch": 0.04467344060332172,
+      "grad_norm": 3.6517354310292265,
+      "learning_rate": 1.9989299317760345e-06,
+      "loss": 0.6997,
+      "step": 97
+    },
+    {
+      "epoch": 0.045133991537376586,
+      "grad_norm": 3.5728547634334746,
+      "learning_rate": 1.9988597949557883e-06,
+      "loss": 0.7399,
+      "step": 98
+    },
+    {
+      "epoch": 0.04559454247143145,
+      "grad_norm": 3.4027410732295653,
+      "learning_rate": 1.998787433289327e-06,
+      "loss": 0.695,
+      "step": 99
+    },
+    {
+      "epoch": 0.046055093405486316,
+      "grad_norm": 3.2197558124034984,
+      "learning_rate": 1.9987128469378284e-06,
+      "loss": 0.6144,
+      "step": 100
+    },
+    {
+      "epoch": 0.046515644339541175,
+      "grad_norm": 3.1601430739183245,
+      "learning_rate": 1.998636036067425e-06,
+      "loss": 0.6233,
+      "step": 101
+    },
+    {
+      "epoch": 0.04697619527359604,
+      "grad_norm": 3.0419954341335087,
+      "learning_rate": 1.9985570008492044e-06,
+      "loss": 0.5879,
+      "step": 102
+    },
+    {
+      "epoch": 0.047436746207650905,
+      "grad_norm": 3.4126431282945027,
+      "learning_rate": 1.9984757414592083e-06,
+      "loss": 0.7941,
+      "step": 103
+    },
+    {
+      "epoch": 0.047897297141705764,
+      "grad_norm": 2.973595724909931,
+      "learning_rate": 1.998392258078433e-06,
+      "loss": 0.6119,
+      "step": 104
+    },
+    {
+      "epoch": 0.04835784807576063,
+      "grad_norm": 4.503321929117909,
+      "learning_rate": 1.998306550892828e-06,
+      "loss": 0.6559,
+      "step": 105
+    },
+    {
+      "epoch": 0.048818399009815494,
+      "grad_norm": 3.615113444334991,
+      "learning_rate": 1.9982186200932964e-06,
+      "loss": 0.6638,
+      "step": 106
+    },
+    {
+      "epoch": 0.04927894994387035,
+      "grad_norm": 3.4410984720501094,
+      "learning_rate": 1.998128465875694e-06,
+      "loss": 0.6764,
+      "step": 107
+    },
+    {
+      "epoch": 0.04973950087792522,
+      "grad_norm": 3.1749314754075617,
+      "learning_rate": 1.9980360884408288e-06,
+      "loss": 0.701,
+      "step": 108
+    },
+    {
+      "epoch": 0.05020005181198008,
+      "grad_norm": 3.2473102824540083,
+      "learning_rate": 1.997941487994461e-06,
+      "loss": 0.6957,
+      "step": 109
+    },
+    {
+      "epoch": 0.05066060274603494,
+      "grad_norm": 3.6947631666546843,
+      "learning_rate": 1.9978446647473024e-06,
+      "loss": 0.7448,
+      "step": 110
+    },
+    {
+      "epoch": 0.051121153680089806,
+      "grad_norm": 3.261951870816807,
+      "learning_rate": 1.9977456189150163e-06,
+      "loss": 0.7727,
+      "step": 111
+    },
+    {
+      "epoch": 0.05158170461414467,
+      "grad_norm": 3.3464119040387925,
+      "learning_rate": 1.9976443507182152e-06,
+      "loss": 0.6483,
+      "step": 112
+    },
+    {
+      "epoch": 0.05204225554819954,
+      "grad_norm": 3.533060942648228,
+      "learning_rate": 1.997540860382463e-06,
+      "loss": 0.8126,
+      "step": 113
+    },
+    {
+      "epoch": 0.052502806482254395,
+      "grad_norm": 3.2761042459948877,
+      "learning_rate": 1.997435148138272e-06,
+      "loss": 0.5982,
+      "step": 114
+    },
+    {
+      "epoch": 0.05296335741630926,
+      "grad_norm": 3.1082981914368473,
+      "learning_rate": 1.9973272142211046e-06,
+      "loss": 0.6504,
+      "step": 115
+    },
+    {
+      "epoch": 0.053423908350364126,
+      "grad_norm": 3.3969089614282444,
+      "learning_rate": 1.997217058871371e-06,
+      "loss": 0.5928,
+      "step": 116
+    },
+    {
+      "epoch": 0.053884459284418984,
+      "grad_norm": 3.1930029822538173,
+      "learning_rate": 1.9971046823344304e-06,
+      "loss": 0.5868,
+      "step": 117
+    },
+    {
+      "epoch": 0.05434501021847385,
+      "grad_norm": 3.6974861112788457,
+      "learning_rate": 1.9969900848605877e-06,
+      "loss": 0.7963,
+      "step": 118
+    },
+    {
+      "epoch": 0.054805561152528715,
+      "grad_norm": 3.315165825861059,
+      "learning_rate": 1.9968732667050966e-06,
+      "loss": 0.6313,
+      "step": 119
+    },
+    {
+      "epoch": 0.05526611208658357,
+      "grad_norm": 3.539748098126486,
+      "learning_rate": 1.9967542281281557e-06,
+      "loss": 0.7429,
+      "step": 120
+    },
+    {
+      "epoch": 0.05572666302063844,
+      "grad_norm": 3.338078973425553,
+      "learning_rate": 1.9966329693949093e-06,
+      "loss": 0.7662,
+      "step": 121
+    },
+    {
+      "epoch": 0.0561872139546933,
+      "grad_norm": 3.472065054117364,
+      "learning_rate": 1.996509490775449e-06,
+      "loss": 0.6625,
+      "step": 122
+    },
+    {
+      "epoch": 0.05664776488874816,
+      "grad_norm": 3.1792157328292974,
+      "learning_rate": 1.996383792544808e-06,
+      "loss": 0.6792,
+      "step": 123
+    },
+    {
+      "epoch": 0.05710831582280303,
+      "grad_norm": 3.5210682757899145,
+      "learning_rate": 1.996255874982965e-06,
+      "loss": 0.6079,
+      "step": 124
+    },
+    {
+      "epoch": 0.05756886675685789,
+      "grad_norm": 3.2138238654727416,
+      "learning_rate": 1.996125738374842e-06,
+      "loss": 0.5784,
+      "step": 125
+    },
+    {
+      "epoch": 0.05802941769091276,
+      "grad_norm": 3.3633304567565196,
+      "learning_rate": 1.995993383010303e-06,
+      "loss": 0.6367,
+      "step": 126
+    },
+    {
+      "epoch": 0.058489968624967616,
+      "grad_norm": 3.781021254907743,
+      "learning_rate": 1.9958588091841553e-06,
+      "loss": 0.6771,
+      "step": 127
+    },
+    {
+      "epoch": 0.05895051955902248,
+      "grad_norm": 3.6201021154476516,
+      "learning_rate": 1.9957220171961465e-06,
+      "loss": 0.6707,
+      "step": 128
+    },
+    {
+      "epoch": 0.059411070493077346,
+      "grad_norm": 3.6881373018752885,
+      "learning_rate": 1.995583007350964e-06,
+      "loss": 0.7512,
+      "step": 129
+    },
+    {
+      "epoch": 0.059871621427132204,
+      "grad_norm": 3.3800092959402566,
+      "learning_rate": 1.9954417799582382e-06,
+      "loss": 0.5795,
+      "step": 130
+    },
+    {
+      "epoch": 0.06033217236118707,
+      "grad_norm": 3.6812949356112936,
+      "learning_rate": 1.9952983353325356e-06,
+      "loss": 0.59,
+      "step": 131
+    },
+    {
+      "epoch": 0.060792723295241935,
+      "grad_norm": 3.730872974671056,
+      "learning_rate": 1.9951526737933634e-06,
+      "loss": 0.6077,
+      "step": 132
+    },
+    {
+      "epoch": 0.06125327422929679,
+      "grad_norm": 3.5504209377161757,
+      "learning_rate": 1.9950047956651657e-06,
+      "loss": 0.606,
+      "step": 133
+    },
+    {
+      "epoch": 0.06171382516335166,
+      "grad_norm": 3.2843333588119727,
+      "learning_rate": 1.9948547012773246e-06,
+      "loss": 0.6067,
+      "step": 134
+    },
+    {
+      "epoch": 0.062174376097406524,
+      "grad_norm": 3.3876731358864403,
+      "learning_rate": 1.9947023909641574e-06,
+      "loss": 0.7097,
+      "step": 135
+    },
+    {
+      "epoch": 0.06263492703146138,
+      "grad_norm": 4.024239315384627,
+      "learning_rate": 1.994547865064919e-06,
+      "loss": 0.8154,
+      "step": 136
+    },
+    {
+      "epoch": 0.06309547796551625,
+      "grad_norm": 3.47895486959038,
+      "learning_rate": 1.9943911239237974e-06,
+      "loss": 0.5583,
+      "step": 137
+    },
+    {
+      "epoch": 0.06355602889957111,
+      "grad_norm": 3.409716745725618,
+      "learning_rate": 1.9942321678899163e-06,
+      "loss": 0.5774,
+      "step": 138
+    },
+    {
+      "epoch": 0.06401657983362598,
+      "grad_norm": 3.32970506539607,
+      "learning_rate": 1.9940709973173314e-06,
+      "loss": 0.6011,
+      "step": 139
+    },
+    {
+      "epoch": 0.06447713076768084,
+      "grad_norm": 3.1166430692289246,
+      "learning_rate": 1.993907612565032e-06,
+      "loss": 0.5851,
+      "step": 140
+    },
+    {
+      "epoch": 0.0649376817017357,
+      "grad_norm": 3.5411479372126426,
+      "learning_rate": 1.9937420139969395e-06,
+      "loss": 0.7496,
+      "step": 141
+    },
+    {
+      "epoch": 0.06539823263579056,
+      "grad_norm": 3.421729369489728,
+      "learning_rate": 1.993574201981905e-06,
+      "loss": 0.6842,
+      "step": 142
+    },
+    {
+      "epoch": 0.06585878356984542,
+      "grad_norm": 3.1449216041214205,
+      "learning_rate": 1.9934041768937114e-06,
+      "loss": 0.6461,
+      "step": 143
+    },
+    {
+      "epoch": 0.06631933450390029,
+      "grad_norm": 3.6494550010926328,
+      "learning_rate": 1.9932319391110695e-06,
+      "loss": 0.7231,
+      "step": 144
+    },
+    {
+      "epoch": 0.06677988543795516,
+      "grad_norm": 3.728480459329745,
+      "learning_rate": 1.99305748901762e-06,
+      "loss": 0.7552,
+      "step": 145
+    },
+    {
+      "epoch": 0.06724043637201002,
+      "grad_norm": 3.2822173129999137,
+      "learning_rate": 1.9928808270019296e-06,
+      "loss": 0.6228,
+      "step": 146
+    },
+    {
+      "epoch": 0.06770098730606489,
+      "grad_norm": 3.2867458531135196,
+      "learning_rate": 1.9927019534574937e-06,
+      "loss": 0.7294,
+      "step": 147
+    },
+    {
+      "epoch": 0.06816153824011974,
+      "grad_norm": 3.1013936687675714,
+      "learning_rate": 1.992520868782732e-06,
+      "loss": 0.5613,
+      "step": 148
+    },
+    {
+      "epoch": 0.0686220891741746,
+      "grad_norm": 3.159174332303596,
+      "learning_rate": 1.9923375733809905e-06,
+      "loss": 0.7149,
+      "step": 149
+    },
+    {
+      "epoch": 0.06908264010822947,
+      "grad_norm": 3.516091730035019,
+      "learning_rate": 1.992152067660539e-06,
+      "loss": 0.6315,
+      "step": 150
+    },
+    {
+      "epoch": 0.06954319104228433,
+      "grad_norm": 3.165146365156602,
+      "learning_rate": 1.9919643520345695e-06,
+      "loss": 0.5459,
+      "step": 151
+    },
+    {
+      "epoch": 0.0700037419763392,
+      "grad_norm": 4.0062232007227685,
+      "learning_rate": 1.991774426921198e-06,
+      "loss": 0.6454,
+      "step": 152
+    },
+    {
+      "epoch": 0.07046429291039406,
+      "grad_norm": 3.548301168807787,
+      "learning_rate": 1.99158229274346e-06,
+      "loss": 0.676,
+      "step": 153
+    },
+    {
+      "epoch": 0.07092484384444893,
+      "grad_norm": 3.4249899072112027,
+      "learning_rate": 1.9913879499293136e-06,
+      "loss": 0.6644,
+      "step": 154
+    },
+    {
+      "epoch": 0.07138539477850378,
+      "grad_norm": 3.707615974526274,
+      "learning_rate": 1.9911913989116345e-06,
+      "loss": 0.6739,
+      "step": 155
+    },
+    {
+      "epoch": 0.07184594571255865,
+      "grad_norm": 3.086896742538257,
+      "learning_rate": 1.990992640128218e-06,
+      "loss": 0.5981,
+      "step": 156
+    },
+    {
+      "epoch": 0.07230649664661351,
+      "grad_norm": 3.1043886733514983,
+      "learning_rate": 1.990791674021776e-06,
+      "loss": 0.5621,
+      "step": 157
+    },
+    {
+      "epoch": 0.07276704758066838,
+      "grad_norm": 2.7518305668279432,
+      "learning_rate": 1.9905885010399386e-06,
+      "loss": 0.5827,
+      "step": 158
+    },
+    {
+      "epoch": 0.07322759851472324,
+      "grad_norm": 3.1688929903728904,
+      "learning_rate": 1.9903831216352494e-06,
+      "loss": 0.5834,
+      "step": 159
+    },
+    {
+      "epoch": 0.0736881494487781,
+      "grad_norm": 3.3606584053832695,
+      "learning_rate": 1.9901755362651685e-06,
+      "loss": 0.6374,
+      "step": 160
+    },
+    {
+      "epoch": 0.07414870038283296,
+      "grad_norm": 3.8896133343206922,
+      "learning_rate": 1.9899657453920676e-06,
+      "loss": 0.7499,
+      "step": 161
+    },
+    {
+      "epoch": 0.07460925131688782,
+      "grad_norm": 3.18752926424928,
+      "learning_rate": 1.989753749483233e-06,
+      "loss": 0.686,
+      "step": 162
+    },
+    {
+      "epoch": 0.07506980225094269,
+      "grad_norm": 3.4277096993868557,
+      "learning_rate": 1.989539549010861e-06,
+      "loss": 0.6334,
+      "step": 163
+    },
+    {
+      "epoch": 0.07553035318499755,
+      "grad_norm": 3.1363256404631876,
+      "learning_rate": 1.9893231444520584e-06,
+      "loss": 0.7184,
+      "step": 164
+    },
+    {
+      "epoch": 0.07599090411905242,
+      "grad_norm": 3.773031496303488,
+      "learning_rate": 1.9891045362888413e-06,
+      "loss": 0.7071,
+      "step": 165
+    },
+    {
+      "epoch": 0.07645145505310728,
+      "grad_norm": 3.4894250417343957,
+      "learning_rate": 1.988883725008136e-06,
+      "loss": 0.8024,
+      "step": 166
+    },
+    {
+      "epoch": 0.07691200598716215,
+      "grad_norm": 3.5237596679440126,
+      "learning_rate": 1.9886607111017727e-06,
+      "loss": 0.5565,
+      "step": 167
+    },
+    {
+      "epoch": 0.077372556921217,
+      "grad_norm": 3.956330390896236,
+      "learning_rate": 1.988435495066491e-06,
+      "loss": 0.6834,
+      "step": 168
+    },
+    {
+      "epoch": 0.07783310785527187,
+      "grad_norm": 3.0581620476086027,
+      "learning_rate": 1.988208077403932e-06,
+      "loss": 0.5355,
+      "step": 169
+    },
+    {
+      "epoch": 0.07829365878932673,
+      "grad_norm": 3.359700225843598,
+      "learning_rate": 1.9879784586206446e-06,
+      "loss": 0.6266,
+      "step": 170
+    },
+    {
+      "epoch": 0.0787542097233816,
+      "grad_norm": 3.160066872012096,
+      "learning_rate": 1.987746639228077e-06,
+      "loss": 0.4693,
+      "step": 171
+    },
+    {
+      "epoch": 0.07921476065743646,
+      "grad_norm": 2.8930357386693037,
+      "learning_rate": 1.9875126197425812e-06,
+      "loss": 0.543,
+      "step": 172
+    },
+    {
+      "epoch": 0.07967531159149133,
+      "grad_norm": 3.2160039035976538,
+      "learning_rate": 1.987276400685409e-06,
+      "loss": 0.5285,
+      "step": 173
+    },
+    {
+      "epoch": 0.08013586252554618,
+      "grad_norm": 3.5154065293512837,
+      "learning_rate": 1.9870379825827105e-06,
+      "loss": 0.7303,
+      "step": 174
+    },
+    {
+      "epoch": 0.08059641345960104,
+      "grad_norm": 3.272929698816058,
+      "learning_rate": 1.9867973659655357e-06,
+      "loss": 0.596,
+      "step": 175
+    },
+    {
+      "epoch": 0.08105696439365591,
+      "grad_norm": 3.2230741681038837,
+      "learning_rate": 1.9865545513698304e-06,
+      "loss": 0.7758,
+      "step": 176
+    },
+    {
+      "epoch": 0.08151751532771077,
+      "grad_norm": 2.88861925930577,
+      "learning_rate": 1.9863095393364363e-06,
+      "loss": 0.5791,
+      "step": 177
+    },
+    {
+      "epoch": 0.08197806626176564,
+      "grad_norm": 3.6487253551379166,
+      "learning_rate": 1.9860623304110895e-06,
+      "loss": 0.8919,
+      "step": 178
+    },
+    {
+      "epoch": 0.0824386171958205,
+      "grad_norm": 3.3372588648412513,
+      "learning_rate": 1.9858129251444203e-06,
+      "loss": 0.6433,
+      "step": 179
+    },
+    {
+      "epoch": 0.08289916812987537,
+      "grad_norm": 3.069467629265819,
+      "learning_rate": 1.9855613240919496e-06,
+      "loss": 0.617,
+      "step": 180
+    },
+    {
+      "epoch": 0.08335971906393022,
+      "grad_norm": 3.489373026628576,
+      "learning_rate": 1.985307527814091e-06,
+      "loss": 0.768,
+      "step": 181
+    },
+    {
+      "epoch": 0.08382026999798509,
+      "grad_norm": 3.459352404519042,
+      "learning_rate": 1.9850515368761465e-06,
+      "loss": 0.7647,
+      "step": 182
+    },
+    {
+      "epoch": 0.08428082093203995,
+      "grad_norm": 3.4396136216484012,
+      "learning_rate": 1.9847933518483066e-06,
+      "loss": 0.6323,
+      "step": 183
+    },
+    {
+      "epoch": 0.08474137186609482,
+      "grad_norm": 3.369315529955375,
+      "learning_rate": 1.9845329733056488e-06,
+      "loss": 0.6724,
+      "step": 184
+    },
+    {
+      "epoch": 0.08520192280014968,
+      "grad_norm": 3.144962754808177,
+      "learning_rate": 1.9842704018281364e-06,
+      "loss": 0.6974,
+      "step": 185
+    },
+    {
+      "epoch": 0.08566247373420455,
+      "grad_norm": 3.3041054108917143,
+      "learning_rate": 1.984005638000618e-06,
+      "loss": 0.6231,
+      "step": 186
+    },
+    {
+      "epoch": 0.0861230246682594,
+      "grad_norm": 4.170198999419159,
+      "learning_rate": 1.983738682412824e-06,
+      "loss": 0.644,
+      "step": 187
+    },
+    {
+      "epoch": 0.08658357560231426,
+      "grad_norm": 3.22338168544018,
+      "learning_rate": 1.983469535659369e-06,
+      "loss": 0.5809,
+      "step": 188
+    },
+    {
+      "epoch": 0.08704412653636913,
+      "grad_norm": 3.87679312890683,
+      "learning_rate": 1.983198198339745e-06,
+      "loss": 0.6326,
+      "step": 189
+    },
+    {
+      "epoch": 0.087504677470424,
+      "grad_norm": 2.8932671924853026,
+      "learning_rate": 1.9829246710583258e-06,
+      "loss": 0.694,
+      "step": 190
+    },
+    {
+      "epoch": 0.08796522840447886,
+      "grad_norm": 3.0104693995609293,
+      "learning_rate": 1.982648954424362e-06,
+      "loss": 0.5702,
+      "step": 191
+    },
+    {
+      "epoch": 0.08842577933853372,
+      "grad_norm": 3.6696726695949597,
+      "learning_rate": 1.982371049051981e-06,
+      "loss": 0.6627,
+      "step": 192
+    },
+    {
+      "epoch": 0.08888633027258859,
+      "grad_norm": 3.483251606814318,
+      "learning_rate": 1.982090955560185e-06,
+      "loss": 0.6148,
+      "step": 193
+    },
+    {
+      "epoch": 0.08934688120664344,
+      "grad_norm": 3.059134143200745,
+      "learning_rate": 1.981808674572851e-06,
+      "loss": 0.641,
+      "step": 194
+    },
+    {
+      "epoch": 0.0898074321406983,
+      "grad_norm": 3.471348679245806,
+      "learning_rate": 1.9815242067187273e-06,
+      "loss": 0.7375,
+      "step": 195
+    },
+    {
+      "epoch": 0.09026798307475317,
+      "grad_norm": 3.2356019265585423,
+      "learning_rate": 1.9812375526314335e-06,
+      "loss": 0.6559,
+      "step": 196
+    },
+    {
+      "epoch": 0.09072853400880804,
+      "grad_norm": 3.5553418134929737,
+      "learning_rate": 1.9809487129494588e-06,
+      "loss": 0.6052,
+      "step": 197
+    },
+    {
+      "epoch": 0.0911890849428629,
+      "grad_norm": 3.5207928333701126,
+      "learning_rate": 1.9806576883161607e-06,
+      "loss": 0.7502,
+      "step": 198
+    },
+    {
+      "epoch": 0.09164963587691777,
+      "grad_norm": 3.6571981250699808,
+      "learning_rate": 1.9803644793797635e-06,
+      "loss": 0.6423,
+      "step": 199
+    },
+    {
+      "epoch": 0.09211018681097263,
+      "grad_norm": 3.3310862469732236,
+      "learning_rate": 1.9800690867933567e-06,
+      "loss": 0.6166,
+      "step": 200
+    },
+    {
+      "epoch": 0.09257073774502748,
+      "grad_norm": 3.599428672561491,
+      "learning_rate": 1.9797715112148933e-06,
+      "loss": 0.6751,
+      "step": 201
+    },
+    {
+      "epoch": 0.09303128867908235,
+      "grad_norm": 3.111228152420326,
+      "learning_rate": 1.979471753307189e-06,
+      "loss": 0.5873,
+      "step": 202
+    },
+    {
+      "epoch": 0.09349183961313721,
+      "grad_norm": 2.7796406722600433,
+      "learning_rate": 1.979169813737921e-06,
+      "loss": 0.5509,
+      "step": 203
+    },
+    {
+      "epoch": 0.09395239054719208,
+      "grad_norm": 3.3190228535283586,
+      "learning_rate": 1.9788656931796237e-06,
+      "loss": 0.5311,
+      "step": 204
+    },
+    {
+      "epoch": 0.09441294148124695,
+      "grad_norm": 3.4901850657341678,
+      "learning_rate": 1.9785593923096927e-06,
+      "loss": 0.6073,
+      "step": 205
+    },
+    {
+      "epoch": 0.09487349241530181,
+      "grad_norm": 3.8594657707637836,
+      "learning_rate": 1.978250911810377e-06,
+      "loss": 0.7903,
+      "step": 206
+    },
+    {
+      "epoch": 0.09533404334935666,
+      "grad_norm": 3.783029519442971,
+      "learning_rate": 1.9779402523687825e-06,
+      "loss": 0.7535,
+      "step": 207
+    },
+    {
+      "epoch": 0.09579459428341153,
+      "grad_norm": 3.541223829083605,
+      "learning_rate": 1.977627414676867e-06,
+      "loss": 0.6787,
+      "step": 208
+    },
+    {
+      "epoch": 0.09625514521746639,
+      "grad_norm": 3.300314419811215,
+      "learning_rate": 1.977312399431441e-06,
+      "loss": 0.688,
+      "step": 209
+    },
+    {
+      "epoch": 0.09671569615152126,
+      "grad_norm": 3.021536775472106,
+      "learning_rate": 1.9769952073341655e-06,
+      "loss": 0.5406,
+      "step": 210
+    },
+    {
+      "epoch": 0.09717624708557612,
+      "grad_norm": 3.1063359917852016,
+      "learning_rate": 1.976675839091549e-06,
+      "loss": 0.7334,
+      "step": 211
+    },
+    {
+      "epoch": 0.09763679801963099,
+      "grad_norm": 4.2418847264483555,
+      "learning_rate": 1.976354295414948e-06,
+      "loss": 0.7438,
+      "step": 212
+    },
+    {
+      "epoch": 0.09809734895368585,
+      "grad_norm": 3.2746417894451927,
+      "learning_rate": 1.9760305770205648e-06,
+      "loss": 0.6335,
+      "step": 213
+    },
+    {
+      "epoch": 0.0985578998877407,
+      "grad_norm": 3.217191492035501,
+      "learning_rate": 1.9757046846294446e-06,
+      "loss": 0.6167,
+      "step": 214
+    },
+    {
+      "epoch": 0.09901845082179557,
+      "grad_norm": 3.202923015353876,
+      "learning_rate": 1.975376618967476e-06,
+      "loss": 0.6576,
+      "step": 215
+    },
+    {
+      "epoch": 0.09947900175585044,
+      "grad_norm": 3.368536938836128,
+      "learning_rate": 1.975046380765387e-06,
+      "loss": 0.5377,
+      "step": 216
+    },
+    {
+      "epoch": 0.0999395526899053,
+      "grad_norm": 3.371798575684341,
+      "learning_rate": 1.9747139707587467e-06,
+      "loss": 0.5324,
+      "step": 217
+    },
+    {
+      "epoch": 0.10040010362396017,
+      "grad_norm": 2.7715577719122484,
+      "learning_rate": 1.9743793896879595e-06,
+      "loss": 0.6635,
+      "step": 218
+    },
+    {
+      "epoch": 0.10086065455801503,
+      "grad_norm": 2.829524830044529,
+      "learning_rate": 1.974042638298267e-06,
+      "loss": 0.5327,
+      "step": 219
+    },
+    {
+      "epoch": 0.10132120549206988,
+      "grad_norm": 2.891869373792989,
+      "learning_rate": 1.9737037173397446e-06,
+      "loss": 0.5856,
+      "step": 220
+    },
+    {
+      "epoch": 0.10178175642612475,
+      "grad_norm": 3.062890989327354,
+      "learning_rate": 1.9733626275672996e-06,
+      "loss": 0.7426,
+      "step": 221
+    },
+    {
+      "epoch": 0.10224230736017961,
+      "grad_norm": 3.116251286088085,
+      "learning_rate": 1.973019369740671e-06,
+      "loss": 0.5652,
+      "step": 222
+    },
+    {
+      "epoch": 0.10270285829423448,
+      "grad_norm": 3.4890492450754502,
+      "learning_rate": 1.972673944624426e-06,
+      "loss": 0.6855,
+      "step": 223
+    },
+    {
+      "epoch": 0.10316340922828934,
+      "grad_norm": 3.0888002182898444,
+      "learning_rate": 1.9723263529879598e-06,
+      "loss": 0.83,
+      "step": 224
+    },
+    {
+      "epoch": 0.10362396016234421,
+      "grad_norm": 3.3556762378140896,
+      "learning_rate": 1.9719765956054933e-06,
+      "loss": 0.6933,
+      "step": 225
+    },
+    {
+      "epoch": 0.10408451109639907,
+      "grad_norm": 3.195505438673579,
+      "learning_rate": 1.971624673256071e-06,
+      "loss": 0.5613,
+      "step": 226
+    },
+    {
+      "epoch": 0.10454506203045393,
+      "grad_norm": 2.6074576346797143,
+      "learning_rate": 1.9712705867235604e-06,
+      "loss": 0.4869,
+      "step": 227
+    },
+    {
+      "epoch": 0.10500561296450879,
+      "grad_norm": 3.421431138997975,
+      "learning_rate": 1.970914336796648e-06,
+      "loss": 0.5345,
+      "step": 228
+    },
+    {
+      "epoch": 0.10546616389856366,
+      "grad_norm": 3.558425101140546,
+      "learning_rate": 1.97055592426884e-06,
+      "loss": 0.6834,
+      "step": 229
+    },
+    {
+      "epoch": 0.10592671483261852,
+      "grad_norm": 3.662238091068311,
+      "learning_rate": 1.9701953499384593e-06,
+      "loss": 0.6788,
+      "step": 230
+    },
+    {
+      "epoch": 0.10638726576667339,
+      "grad_norm": 2.9575097174697387,
+      "learning_rate": 1.9698326146086445e-06,
+      "loss": 0.6197,
+      "step": 231
+    },
+    {
+      "epoch": 0.10684781670072825,
+      "grad_norm": 3.4475135509424293,
+      "learning_rate": 1.9694677190873467e-06,
+      "loss": 0.659,
+      "step": 232
+    },
+    {
+      "epoch": 0.1073083676347831,
+      "grad_norm": 3.108410036960475,
+      "learning_rate": 1.9691006641873296e-06,
+      "loss": 0.6831,
+      "step": 233
+    },
+    {
+      "epoch": 0.10776891856883797,
+      "grad_norm": 3.220407600024432,
+      "learning_rate": 1.968731450726166e-06,
+      "loss": 0.6389,
+      "step": 234
+    },
+    {
+      "epoch": 0.10822946950289283,
+      "grad_norm": 3.534132987080869,
+      "learning_rate": 1.9683600795262364e-06,
+      "loss": 0.6332,
+      "step": 235
+    },
+    {
+      "epoch": 0.1086900204369477,
+      "grad_norm": 3.5324339129255877,
+      "learning_rate": 1.9679865514147277e-06,
+      "loss": 0.7518,
+      "step": 236
+    },
+    {
+      "epoch": 0.10915057137100256,
+      "grad_norm": 3.0421407458400003,
+      "learning_rate": 1.9676108672236317e-06,
+      "loss": 0.5808,
+      "step": 237
+    },
+    {
+      "epoch": 0.10961112230505743,
+      "grad_norm": 3.5520947520869908,
+      "learning_rate": 1.9672330277897414e-06,
+      "loss": 0.5368,
+      "step": 238
+    },
+    {
+      "epoch": 0.1100716732391123,
+      "grad_norm": 3.451159362431226,
+      "learning_rate": 1.9668530339546514e-06,
+      "loss": 0.6296,
+      "step": 239
+    },
+    {
+      "epoch": 0.11053222417316715,
+      "grad_norm": 3.509910875733122,
+      "learning_rate": 1.966470886564755e-06,
+      "loss": 0.7006,
+      "step": 240
+    },
+    {
+      "epoch": 0.11099277510722201,
+      "grad_norm": 2.596656303513613,
+      "learning_rate": 1.9660865864712412e-06,
+      "loss": 0.4933,
+      "step": 241
+    },
+    {
+      "epoch": 0.11145332604127688,
+      "grad_norm": 3.4232091740027655,
+      "learning_rate": 1.965700134530095e-06,
+      "loss": 0.5211,
+      "step": 242
+    },
+    {
+      "epoch": 0.11191387697533174,
+      "grad_norm": 3.040611124293651,
+      "learning_rate": 1.9653115316020935e-06,
+      "loss": 0.6377,
+      "step": 243
+    },
+    {
+      "epoch": 0.1123744279093866,
+      "grad_norm": 3.4791503945160267,
+      "learning_rate": 1.9649207785528065e-06,
+      "loss": 0.6889,
+      "step": 244
+    },
+    {
+      "epoch": 0.11283497884344147,
+      "grad_norm": 3.0895910555628032,
+      "learning_rate": 1.96452787625259e-06,
+      "loss": 0.6833,
+      "step": 245
+    },
+    {
+      "epoch": 0.11329552977749632,
+      "grad_norm": 3.001038166095486,
+      "learning_rate": 1.9641328255765913e-06,
+      "loss": 0.5572,
+      "step": 246
+    },
+    {
+      "epoch": 0.11375608071155119,
+      "grad_norm": 3.7375507199090596,
+      "learning_rate": 1.963735627404739e-06,
+      "loss": 0.6776,
+      "step": 247
+    },
+    {
+      "epoch": 0.11421663164560605,
+      "grad_norm": 3.4861645782956723,
+      "learning_rate": 1.963336282621747e-06,
+      "loss": 0.6076,
+      "step": 248
+    },
+    {
+      "epoch": 0.11467718257966092,
+      "grad_norm": 3.9356402228698313,
+      "learning_rate": 1.962934792117111e-06,
+      "loss": 0.5698,
+      "step": 249
+    },
+    {
+      "epoch": 0.11513773351371578,
+      "grad_norm": 3.4680226962421496,
+      "learning_rate": 1.9625311567851045e-06,
+      "loss": 0.6332,
+      "step": 250
+    },
+    {
+      "epoch": 0.11559828444777065,
+      "grad_norm": 4.110638488505918,
+      "learning_rate": 1.9621253775247795e-06,
+      "loss": 0.7351,
+      "step": 251
+    },
+    {
+      "epoch": 0.11605883538182551,
+      "grad_norm": 3.736146712062541,
+      "learning_rate": 1.9617174552399633e-06,
+      "loss": 0.7494,
+      "step": 252
+    },
+    {
+      "epoch": 0.11651938631588037,
+      "grad_norm": 2.938611398017176,
+      "learning_rate": 1.961307390839255e-06,
+      "loss": 0.6439,
+      "step": 253
+    },
+    {
+      "epoch": 0.11697993724993523,
+      "grad_norm": 3.267375637056784,
+      "learning_rate": 1.960895185236028e-06,
+      "loss": 0.624,
+      "step": 254
+    },
+    {
+      "epoch": 0.1174404881839901,
+      "grad_norm": 3.394791456038148,
+      "learning_rate": 1.9604808393484217e-06,
+      "loss": 0.605,
+      "step": 255
+    },
+    {
+      "epoch": 0.11790103911804496,
+      "grad_norm": 3.1028647299114174,
+      "learning_rate": 1.960064354099345e-06,
+      "loss": 0.6428,
+      "step": 256
+    },
+    {
+      "epoch": 0.11836159005209983,
+      "grad_norm": 3.0414855513121357,
+      "learning_rate": 1.959645730416471e-06,
+      "loss": 0.4093,
+      "step": 257
+    },
+    {
+      "epoch": 0.11882214098615469,
+      "grad_norm": 3.4120022845796525,
+      "learning_rate": 1.959224969232237e-06,
+      "loss": 0.6576,
+      "step": 258
+    },
+    {
+      "epoch": 0.11928269192020956,
+      "grad_norm": 2.972667933535389,
+      "learning_rate": 1.9588020714838394e-06,
+      "loss": 0.5633,
+      "step": 259
+    },
+    {
+      "epoch": 0.11974324285426441,
+      "grad_norm": 3.2887632380841842,
+      "learning_rate": 1.9583770381132357e-06,
+      "loss": 0.6012,
+      "step": 260
+    },
+    {
+      "epoch": 0.12020379378831927,
+      "grad_norm": 3.5081309566331527,
+      "learning_rate": 1.9579498700671386e-06,
+      "loss": 0.6691,
+      "step": 261
+    },
+    {
+      "epoch": 0.12066434472237414,
+      "grad_norm": 3.3208918064597936,
+      "learning_rate": 1.9575205682970163e-06,
+      "loss": 0.6513,
+      "step": 262
+    },
+    {
+      "epoch": 0.121124895656429,
+      "grad_norm": 3.2712864716169943,
+      "learning_rate": 1.9570891337590895e-06,
+      "loss": 0.5478,
+      "step": 263
+    },
+    {
+      "epoch": 0.12158544659048387,
+      "grad_norm": 3.282868754227743,
+      "learning_rate": 1.956655567414329e-06,
+      "loss": 0.6925,
+      "step": 264
+    },
+    {
+      "epoch": 0.12204599752453874,
+      "grad_norm": 3.482367964725982,
+      "learning_rate": 1.9562198702284552e-06,
+      "loss": 0.7356,
+      "step": 265
+    },
+    {
+      "epoch": 0.12250654845859359,
+      "grad_norm": 3.084010149386528,
+      "learning_rate": 1.955782043171933e-06,
+      "loss": 0.5609,
+      "step": 266
+    },
+    {
+      "epoch": 0.12296709939264845,
+      "grad_norm": 3.0585298128518765,
+      "learning_rate": 1.9553420872199732e-06,
+      "loss": 0.6027,
+      "step": 267
+    },
+    {
+      "epoch": 0.12342765032670332,
+      "grad_norm": 3.386973557288646,
+      "learning_rate": 1.954900003352527e-06,
+      "loss": 0.5398,
+      "step": 268
+    },
+    {
+      "epoch": 0.12388820126075818,
+      "grad_norm": 3.2451705841602396,
+      "learning_rate": 1.954455792554285e-06,
+      "loss": 0.7089,
+      "step": 269
+    },
+    {
+      "epoch": 0.12434875219481305,
+      "grad_norm": 2.9403279279548,
+      "learning_rate": 1.9540094558146775e-06,
+      "loss": 0.4651,
+      "step": 270
+    },
+    {
+      "epoch": 0.12480930312886791,
+      "grad_norm": 3.422975954193148,
+      "learning_rate": 1.9535609941278677e-06,
+      "loss": 0.6637,
+      "step": 271
+    },
+    {
+      "epoch": 0.12526985406292276,
+      "grad_norm": 2.8483103524833906,
+      "learning_rate": 1.9531104084927526e-06,
+      "loss": 0.5443,
+      "step": 272
+    },
+    {
+      "epoch": 0.12573040499697763,
+      "grad_norm": 3.0046696813378477,
+      "learning_rate": 1.9526576999129613e-06,
+      "loss": 0.6132,
+      "step": 273
+    },
+    {
+      "epoch": 0.1261909559310325,
+      "grad_norm": 3.030506496717706,
+      "learning_rate": 1.9522028693968496e-06,
+      "loss": 0.6696,
+      "step": 274
+    },
+    {
+      "epoch": 0.12665150686508736,
+      "grad_norm": 3.4189036410010076,
+      "learning_rate": 1.951745917957501e-06,
+      "loss": 0.6398,
+      "step": 275
+    },
+    {
+      "epoch": 0.12711205779914223,
+      "grad_norm": 3.4281845642915854,
+      "learning_rate": 1.951286846612723e-06,
+      "loss": 0.613,
+      "step": 276
+    },
+    {
+      "epoch": 0.1275726087331971,
+      "grad_norm": 3.5114510461027573,
+      "learning_rate": 1.9508256563850437e-06,
+      "loss": 0.5248,
+      "step": 277
+    },
+    {
+      "epoch": 0.12803315966725196,
+      "grad_norm": 3.2175363479882853,
+      "learning_rate": 1.9503623483017125e-06,
+      "loss": 0.6153,
+      "step": 278
+    },
+    {
+      "epoch": 0.12849371060130682,
+      "grad_norm": 3.257507884907778,
+      "learning_rate": 1.949896923394695e-06,
+      "loss": 0.6326,
+      "step": 279
+    },
+    {
+      "epoch": 0.12895426153536169,
+      "grad_norm": 3.4234091810352894,
+      "learning_rate": 1.9494293827006724e-06,
+      "loss": 0.6533,
+      "step": 280
+    },
+    {
+      "epoch": 0.12941481246941655,
+      "grad_norm": 2.9859571949464803,
+      "learning_rate": 1.9489597272610374e-06,
+      "loss": 0.7496,
+      "step": 281
+    },
+    {
+      "epoch": 0.1298753634034714,
+      "grad_norm": 3.6355830400808125,
+      "learning_rate": 1.948487958121895e-06,
+      "loss": 0.6794,
+      "step": 282
+    },
+    {
+      "epoch": 0.13033591433752625,
+      "grad_norm": 3.2842113677863733,
+      "learning_rate": 1.9480140763340563e-06,
+      "loss": 0.6123,
+      "step": 283
+    },
+    {
+      "epoch": 0.13079646527158112,
+      "grad_norm": 3.094397720174347,
+      "learning_rate": 1.9475380829530394e-06,
+      "loss": 0.5419,
+      "step": 284
+    },
+    {
+      "epoch": 0.13125701620563598,
+      "grad_norm": 3.527829614510662,
+      "learning_rate": 1.947059979039065e-06,
+      "loss": 0.7257,
+      "step": 285
+    },
+    {
+      "epoch": 0.13171756713969085,
+      "grad_norm": 3.4106857915169475,
+      "learning_rate": 1.9465797656570544e-06,
+      "loss": 0.6017,
+      "step": 286
+    },
+    {
+      "epoch": 0.13217811807374572,
+      "grad_norm": 3.319543434532599,
+      "learning_rate": 1.946097443876629e-06,
+      "loss": 0.6015,
+      "step": 287
+    },
+    {
+      "epoch": 0.13263866900780058,
+      "grad_norm": 3.141575156484543,
+      "learning_rate": 1.9456130147721057e-06,
+      "loss": 0.6703,
+      "step": 288
+    },
+    {
+      "epoch": 0.13309921994185545,
+      "grad_norm": 2.8399129741558635,
+      "learning_rate": 1.9451264794224948e-06,
+      "loss": 0.4491,
+      "step": 289
+    },
+    {
+      "epoch": 0.1335597708759103,
+      "grad_norm": 3.0162883821205484,
+      "learning_rate": 1.944637838911498e-06,
+      "loss": 0.4653,
+      "step": 290
+    },
+    {
+      "epoch": 0.13402032180996518,
+      "grad_norm": 3.0158920954224593,
+      "learning_rate": 1.944147094327506e-06,
+      "loss": 0.5747,
+      "step": 291
+    },
+    {
+      "epoch": 0.13448087274402004,
+      "grad_norm": 2.8892050309237054,
+      "learning_rate": 1.9436542467635968e-06,
+      "loss": 0.4266,
+      "step": 292
+    },
+    {
+      "epoch": 0.1349414236780749,
+      "grad_norm": 3.643489464124624,
+      "learning_rate": 1.943159297317532e-06,
+      "loss": 0.6408,
+      "step": 293
+    },
+    {
+      "epoch": 0.13540197461212977,
+      "grad_norm": 3.1781209929973815,
+      "learning_rate": 1.9426622470917553e-06,
+      "loss": 0.588,
+      "step": 294
+    },
+    {
+      "epoch": 0.13586252554618464,
+      "grad_norm": 3.2919873800040245,
+      "learning_rate": 1.942163097193389e-06,
+      "loss": 0.6894,
+      "step": 295
+    },
+    {
+      "epoch": 0.13632307648023947,
+      "grad_norm": 2.858947646506684,
+      "learning_rate": 1.941661848734233e-06,
+      "loss": 0.6467,
+      "step": 296
+    },
+    {
+      "epoch": 0.13678362741429434,
+      "grad_norm": 3.246171280951807,
+      "learning_rate": 1.9411585028307604e-06,
+      "loss": 0.5943,
+      "step": 297
+    },
+    {
+      "epoch": 0.1372441783483492,
+      "grad_norm": 2.7358113053763953,
+      "learning_rate": 1.9406530606041173e-06,
+      "loss": 0.524,
+      "step": 298
+    },
+    {
+      "epoch": 0.13770472928240407,
+      "grad_norm": 3.136480105540605,
+      "learning_rate": 1.940145523180118e-06,
+      "loss": 0.6102,
+      "step": 299
+    },
+    {
+      "epoch": 0.13816528021645894,
+      "grad_norm": 3.155533214531188,
+      "learning_rate": 1.939635891689245e-06,
+      "loss": 0.5248,
+      "step": 300
+    },
+    {
+      "epoch": 0.1386258311505138,
+      "grad_norm": 3.647288084061138,
+      "learning_rate": 1.9391241672666437e-06,
+      "loss": 0.7047,
+      "step": 301
+    },
+    {
+      "epoch": 0.13908638208456867,
+      "grad_norm": 2.979265708987498,
+      "learning_rate": 1.938610351052122e-06,
+      "loss": 0.5515,
+      "step": 302
+    },
+    {
+      "epoch": 0.13954693301862353,
+      "grad_norm": 3.119300289126859,
+      "learning_rate": 1.938094444190147e-06,
+      "loss": 0.6248,
+      "step": 303
+    },
+    {
+      "epoch": 0.1400074839526784,
+      "grad_norm": 3.663446793597039,
+      "learning_rate": 1.937576447829842e-06,
+      "loss": 0.5661,
+      "step": 304
+    },
+    {
+      "epoch": 0.14046803488673326,
+      "grad_norm": 3.5292734921724236,
+      "learning_rate": 1.937056363124985e-06,
+      "loss": 0.5915,
+      "step": 305
+    },
+    {
+      "epoch": 0.14092858582078813,
+      "grad_norm": 3.037749619620941,
+      "learning_rate": 1.936534191234006e-06,
+      "loss": 0.676,
+      "step": 306
+    },
+    {
+      "epoch": 0.141389136754843,
+      "grad_norm": 3.387669461542685,
+      "learning_rate": 1.9360099333199825e-06,
+      "loss": 0.6448,
+      "step": 307
+    },
+    {
+      "epoch": 0.14184968768889786,
+      "grad_norm": 3.5739322009118166,
+      "learning_rate": 1.935483590550639e-06,
+      "loss": 0.7227,
+      "step": 308
+    },
+    {
+      "epoch": 0.1423102386229527,
+      "grad_norm": 3.1735367015087115,
+      "learning_rate": 1.9349551640983444e-06,
+      "loss": 0.5428,
+      "step": 309
+    },
+    {
+      "epoch": 0.14277078955700756,
+      "grad_norm": 3.048510928639603,
+      "learning_rate": 1.934424655140109e-06,
+      "loss": 0.5981,
+      "step": 310
+    },
+    {
+      "epoch": 0.14323134049106243,
+      "grad_norm": 3.586272313938004,
+      "learning_rate": 1.933892064857579e-06,
+      "loss": 0.5659,
+      "step": 311
+    },
+    {
+      "epoch": 0.1436918914251173,
+      "grad_norm": 3.3251317073628393,
+      "learning_rate": 1.933357394437041e-06,
+      "loss": 0.6554,
+      "step": 312
+    },
+    {
+      "epoch": 0.14415244235917216,
+      "grad_norm": 3.2915172640337915,
+      "learning_rate": 1.93282064506941e-06,
+      "loss": 0.5574,
+      "step": 313
+    },
+    {
+      "epoch": 0.14461299329322702,
+      "grad_norm": 3.1215547918258952,
+      "learning_rate": 1.9322818179502356e-06,
+      "loss": 0.5984,
+      "step": 314
+    },
+    {
+      "epoch": 0.1450735442272819,
+      "grad_norm": 3.1067732908414336,
+      "learning_rate": 1.931740914279693e-06,
+      "loss": 0.5418,
+      "step": 315
+    },
+    {
+      "epoch": 0.14553409516133675,
+      "grad_norm": 3.0231982850922363,
+      "learning_rate": 1.9311979352625832e-06,
+      "loss": 0.6471,
+      "step": 316
+    },
+    {
+      "epoch": 0.14599464609539162,
+      "grad_norm": 3.2286421391211326,
+      "learning_rate": 1.930652882108331e-06,
+      "loss": 0.667,
+      "step": 317
+    },
+    {
+      "epoch": 0.14645519702944648,
+      "grad_norm": 3.051273886454387,
+      "learning_rate": 1.930105756030979e-06,
+      "loss": 0.6372,
+      "step": 318
+    },
+    {
+      "epoch": 0.14691574796350135,
+      "grad_norm": 3.357978841575257,
+      "learning_rate": 1.929556558249189e-06,
+      "loss": 0.5968,
+      "step": 319
+    },
+    {
+      "epoch": 0.1473762988975562,
+      "grad_norm": 3.408657581264013,
+      "learning_rate": 1.9290052899862353e-06,
+      "loss": 0.6864,
+      "step": 320
+    },
+    {
+      "epoch": 0.14783684983161108,
+      "grad_norm": 2.965828703883621,
+      "learning_rate": 1.9284519524700063e-06,
+      "loss": 0.4714,
+      "step": 321
+    },
+    {
+      "epoch": 0.14829740076566592,
+      "grad_norm": 2.936230948042832,
+      "learning_rate": 1.9278965469329976e-06,
+      "loss": 0.5051,
+      "step": 322
+    },
+    {
+      "epoch": 0.14875795169972078,
+      "grad_norm": 2.9304142245949727,
+      "learning_rate": 1.9273390746123115e-06,
+      "loss": 0.5798,
+      "step": 323
+    },
+    {
+      "epoch": 0.14921850263377565,
+      "grad_norm": 2.6044770470699885,
+      "learning_rate": 1.926779536749654e-06,
+      "loss": 0.4478,
+      "step": 324
+    },
+    {
+      "epoch": 0.1496790535678305,
+      "grad_norm": 3.346102851664222,
+      "learning_rate": 1.9262179345913323e-06,
+      "loss": 0.6081,
+      "step": 325
+    },
+    {
+      "epoch": 0.15013960450188538,
+      "grad_norm": 3.599494713042751,
+      "learning_rate": 1.9256542693882503e-06,
+      "loss": 0.66,
+      "step": 326
+    },
+    {
+      "epoch": 0.15060015543594024,
+      "grad_norm": 3.3559078451878883,
+      "learning_rate": 1.925088542395909e-06,
+      "loss": 0.5762,
+      "step": 327
+    },
+    {
+      "epoch": 0.1510607063699951,
+      "grad_norm": 3.305256183172175,
+      "learning_rate": 1.9245207548743994e-06,
+      "loss": 0.5205,
+      "step": 328
+    },
+    {
+      "epoch": 0.15152125730404997,
+      "grad_norm": 3.189694287544651,
+      "learning_rate": 1.9239509080884043e-06,
+      "loss": 0.5483,
+      "step": 329
+    },
+    {
+      "epoch": 0.15198180823810484,
+      "grad_norm": 3.8482914338019802,
+      "learning_rate": 1.923379003307193e-06,
+      "loss": 0.5602,
+      "step": 330
+    },
+    {
+      "epoch": 0.1524423591721597,
+      "grad_norm": 2.804743136804608,
+      "learning_rate": 1.9228050418046165e-06,
+      "loss": 0.4328,
+      "step": 331
+    },
+    {
+      "epoch": 0.15290291010621457,
+      "grad_norm": 3.675849703736383,
+      "learning_rate": 1.92222902485911e-06,
+      "loss": 0.5611,
+      "step": 332
+    },
+    {
+      "epoch": 0.15336346104026943,
+      "grad_norm": 3.145151001423423,
+      "learning_rate": 1.921650953753685e-06,
+      "loss": 0.44,
+      "step": 333
+    },
+    {
+      "epoch": 0.1538240119743243,
+      "grad_norm": 3.580270585290245,
+      "learning_rate": 1.9210708297759284e-06,
+      "loss": 0.6488,
+      "step": 334
+    },
+    {
+      "epoch": 0.15428456290837914,
+      "grad_norm": 3.003304167462429,
+      "learning_rate": 1.9204886542180007e-06,
+      "loss": 0.5877,
+      "step": 335
+    },
+    {
+      "epoch": 0.154745113842434,
+      "grad_norm": 2.8636522642130275,
+      "learning_rate": 1.9199044283766315e-06,
+      "loss": 0.5932,
+      "step": 336
+    },
+    {
+      "epoch": 0.15520566477648887,
+      "grad_norm": 3.3796119616806743,
+      "learning_rate": 1.9193181535531177e-06,
+      "loss": 0.6361,
+      "step": 337
+    },
+    {
+      "epoch": 0.15566621571054373,
+      "grad_norm": 3.185129118132095,
+      "learning_rate": 1.9187298310533184e-06,
+      "loss": 0.5942,
+      "step": 338
+    },
+    {
+      "epoch": 0.1561267666445986,
+      "grad_norm": 3.337715563794658,
+      "learning_rate": 1.9181394621876556e-06,
+      "loss": 0.6108,
+      "step": 339
+    },
+    {
+      "epoch": 0.15658731757865346,
+      "grad_norm": 2.8640792419997907,
+      "learning_rate": 1.917547048271109e-06,
+      "loss": 0.633,
+      "step": 340
+    },
+    {
+      "epoch": 0.15704786851270833,
+      "grad_norm": 2.7926946129227943,
+      "learning_rate": 1.916952590623212e-06,
+      "loss": 0.4712,
+      "step": 341
+    },
+    {
+      "epoch": 0.1575084194467632,
+      "grad_norm": 3.4005962235241203,
+      "learning_rate": 1.9163560905680514e-06,
+      "loss": 0.5317,
+      "step": 342
+    },
+    {
+      "epoch": 0.15796897038081806,
+      "grad_norm": 3.1869189227339416,
+      "learning_rate": 1.9157575494342636e-06,
+      "loss": 0.5394,
+      "step": 343
+    },
+    {
+      "epoch": 0.15842952131487292,
+      "grad_norm": 3.4152466358703713,
+      "learning_rate": 1.91515696855503e-06,
+      "loss": 0.6375,
+      "step": 344
+    },
+    {
+      "epoch": 0.1588900722489278,
+      "grad_norm": 3.0926055262112615,
+      "learning_rate": 1.9145543492680763e-06,
+      "loss": 0.6227,
+      "step": 345
+    },
+    {
+      "epoch": 0.15935062318298265,
+      "grad_norm": 3.22498936596375,
+      "learning_rate": 1.9139496929156683e-06,
+      "loss": 0.6416,
+      "step": 346
+    },
+    {
+      "epoch": 0.15981117411703752,
+      "grad_norm": 2.9706888314628466,
+      "learning_rate": 1.913343000844609e-06,
+      "loss": 0.5891,
+      "step": 347
+    },
+    {
+      "epoch": 0.16027172505109236,
+      "grad_norm": 3.19859477403754,
+      "learning_rate": 1.9127342744062357e-06,
+      "loss": 0.5765,
+      "step": 348
+    },
+    {
+      "epoch": 0.16073227598514722,
+      "grad_norm": 3.115476706969415,
+      "learning_rate": 1.912123514956417e-06,
+      "loss": 0.5554,
+      "step": 349
+    },
+    {
+      "epoch": 0.1611928269192021,
+      "grad_norm": 2.961966146622805,
+      "learning_rate": 1.9115107238555497e-06,
+      "loss": 0.5382,
+      "step": 350
+    },
+    {
+      "epoch": 0.16165337785325695,
+      "grad_norm": 3.671673821035005,
+      "learning_rate": 1.9108959024685566e-06,
+      "loss": 0.6642,
+      "step": 351
+    },
+    {
+      "epoch": 0.16211392878731182,
+      "grad_norm": 3.3302938456428106,
+      "learning_rate": 1.9102790521648817e-06,
+      "loss": 0.5453,
+      "step": 352
+    },
+    {
+      "epoch": 0.16257447972136668,
+      "grad_norm": 3.4915603642383752,
+      "learning_rate": 1.909660174318489e-06,
+      "loss": 0.5795,
+      "step": 353
+    },
+    {
+      "epoch": 0.16303503065542155,
+      "grad_norm": 3.211148339907401,
+      "learning_rate": 1.909039270307858e-06,
+      "loss": 0.5538,
+      "step": 354
+    },
+    {
+      "epoch": 0.1634955815894764,
+      "grad_norm": 2.998522836200678,
+      "learning_rate": 1.9084163415159817e-06,
+      "loss": 0.6143,
+      "step": 355
+    },
+    {
+      "epoch": 0.16395613252353128,
+      "grad_norm": 3.088249307549935,
+      "learning_rate": 1.907791389330363e-06,
+      "loss": 0.5506,
+      "step": 356
+    },
+    {
+      "epoch": 0.16441668345758614,
+      "grad_norm": 3.0463921020716604,
+      "learning_rate": 1.9071644151430108e-06,
+      "loss": 0.5874,
+      "step": 357
+    },
+    {
+      "epoch": 0.164877234391641,
+      "grad_norm": 3.345306076824923,
+      "learning_rate": 1.9065354203504398e-06,
+      "loss": 0.7046,
+      "step": 358
+    },
+    {
+      "epoch": 0.16533778532569587,
+      "grad_norm": 3.228814822006655,
+      "learning_rate": 1.9059044063536633e-06,
+      "loss": 0.6248,
+      "step": 359
+    },
+    {
+      "epoch": 0.16579833625975074,
+      "grad_norm": 3.1464755508670637,
+      "learning_rate": 1.9052713745581931e-06,
+      "loss": 0.7453,
+      "step": 360
+    },
+    {
+      "epoch": 0.16625888719380558,
+      "grad_norm": 3.2839840109926697,
+      "learning_rate": 1.9046363263740358e-06,
+      "loss": 0.5185,
+      "step": 361
+    },
+    {
+      "epoch": 0.16671943812786044,
+      "grad_norm": 3.2693787327619046,
+      "learning_rate": 1.9039992632156881e-06,
+      "loss": 0.6118,
+      "step": 362
+    },
+    {
+      "epoch": 0.1671799890619153,
+      "grad_norm": 2.666213429284101,
+      "learning_rate": 1.9033601865021356e-06,
+      "loss": 0.6548,
+      "step": 363
+    },
+    {
+      "epoch": 0.16764053999597017,
+      "grad_norm": 3.814240106647521,
+      "learning_rate": 1.902719097656849e-06,
+      "loss": 0.6145,
+      "step": 364
+    },
+    {
+      "epoch": 0.16810109093002504,
+      "grad_norm": 3.5057882856050515,
+      "learning_rate": 1.9020759981077804e-06,
+      "loss": 0.5805,
+      "step": 365
+    },
+    {
+      "epoch": 0.1685616418640799,
+      "grad_norm": 3.039639804053503,
+      "learning_rate": 1.9014308892873608e-06,
+      "loss": 0.5629,
+      "step": 366
+    },
+    {
+      "epoch": 0.16902219279813477,
+      "grad_norm": 3.492100318826316,
+      "learning_rate": 1.9007837726324965e-06,
+      "loss": 0.5946,
+      "step": 367
+    },
+    {
+      "epoch": 0.16948274373218963,
+      "grad_norm": 3.5705400131693494,
+      "learning_rate": 1.9001346495845656e-06,
+      "loss": 0.6548,
+      "step": 368
+    },
+    {
+      "epoch": 0.1699432946662445,
+      "grad_norm": 4.12268491301927,
+      "learning_rate": 1.899483521589416e-06,
+      "loss": 0.7214,
+      "step": 369
+    },
+    {
+      "epoch": 0.17040384560029936,
+      "grad_norm": 3.4249008502834934,
+      "learning_rate": 1.8988303900973612e-06,
+      "loss": 0.5911,
+      "step": 370
+    },
+    {
+      "epoch": 0.17086439653435423,
+      "grad_norm": 3.656574957105641,
+      "learning_rate": 1.8981752565631767e-06,
+      "loss": 0.6093,
+      "step": 371
+    },
+    {
+      "epoch": 0.1713249474684091,
+      "grad_norm": 3.079763893362344,
+      "learning_rate": 1.8975181224460974e-06,
+      "loss": 0.6065,
+      "step": 372
+    },
+    {
+      "epoch": 0.17178549840246396,
+      "grad_norm": 3.2800007239600197,
+      "learning_rate": 1.8968589892098153e-06,
+      "loss": 0.4789,
+      "step": 373
+    },
+    {
+      "epoch": 0.1722460493365188,
+      "grad_norm": 3.2117733850824313,
+      "learning_rate": 1.8961978583224743e-06,
+      "loss": 0.5964,
+      "step": 374
+    },
+    {
+      "epoch": 0.17270660027057366,
+      "grad_norm": 3.4590302152792876,
+      "learning_rate": 1.8955347312566675e-06,
+      "loss": 0.6101,
+      "step": 375
+    },
+    {
+      "epoch": 0.17316715120462853,
+      "grad_norm": 4.060553328835282,
+      "learning_rate": 1.8948696094894352e-06,
+      "loss": 0.5648,
+      "step": 376
+    },
+    {
+      "epoch": 0.1736277021386834,
+      "grad_norm": 3.567361103730285,
+      "learning_rate": 1.8942024945022598e-06,
+      "loss": 0.7206,
+      "step": 377
+    },
+    {
+      "epoch": 0.17408825307273826,
+      "grad_norm": 3.11756724456868,
+      "learning_rate": 1.8935333877810646e-06,
+      "loss": 0.5549,
+      "step": 378
+    },
+    {
+      "epoch": 0.17454880400679312,
+      "grad_norm": 3.0493230161140357,
+      "learning_rate": 1.892862290816208e-06,
+      "loss": 0.5972,
+      "step": 379
+    },
+    {
+      "epoch": 0.175009354940848,
+      "grad_norm": 3.0362303456982156,
+      "learning_rate": 1.8921892051024816e-06,
+      "loss": 0.5893,
+      "step": 380
+    },
+    {
+      "epoch": 0.17546990587490285,
+      "grad_norm": 3.2964073384518082,
+      "learning_rate": 1.8915141321391078e-06,
+      "loss": 0.6622,
+      "step": 381
+    },
+    {
+      "epoch": 0.17593045680895772,
+      "grad_norm": 3.177252523691639,
+      "learning_rate": 1.8908370734297338e-06,
+      "loss": 0.7608,
+      "step": 382
+    },
+    {
+      "epoch": 0.17639100774301258,
+      "grad_norm": 2.9299509906121077,
+      "learning_rate": 1.8901580304824311e-06,
+      "loss": 0.4753,
+      "step": 383
+    },
+    {
+      "epoch": 0.17685155867706745,
+      "grad_norm": 3.66473331747743,
+      "learning_rate": 1.8894770048096903e-06,
+      "loss": 0.551,
+      "step": 384
+    },
+    {
+      "epoch": 0.17731210961112231,
+      "grad_norm": 2.857955898303891,
+      "learning_rate": 1.8887939979284182e-06,
+      "loss": 0.6346,
+      "step": 385
+    },
+    {
+      "epoch": 0.17777266054517718,
+      "grad_norm": 3.1162479244718666,
+      "learning_rate": 1.8881090113599352e-06,
+      "loss": 0.5842,
+      "step": 386
+    },
+    {
+      "epoch": 0.17823321147923202,
+      "grad_norm": 3.329339301935554,
+      "learning_rate": 1.88742204662997e-06,
+      "loss": 0.6312,
+      "step": 387
+    },
+    {
+      "epoch": 0.17869376241328688,
+      "grad_norm": 3.334334327924235,
+      "learning_rate": 1.8867331052686583e-06,
+      "loss": 0.625,
+      "step": 388
+    },
+    {
+      "epoch": 0.17915431334734175,
+      "grad_norm": 3.2174795958697033,
+      "learning_rate": 1.886042188810539e-06,
+      "loss": 0.5223,
+      "step": 389
+    },
+    {
+      "epoch": 0.1796148642813966,
+      "grad_norm": 3.185576064014635,
+      "learning_rate": 1.8853492987945487e-06,
+      "loss": 0.5531,
+      "step": 390
+    },
+    {
+      "epoch": 0.18007541521545148,
+      "grad_norm": 3.463451373745552,
+      "learning_rate": 1.8846544367640216e-06,
+      "loss": 0.6159,
+      "step": 391
+    },
+    {
+      "epoch": 0.18053596614950634,
+      "grad_norm": 3.2172789875643604,
+      "learning_rate": 1.8839576042666833e-06,
+      "loss": 0.545,
+      "step": 392
+    },
+    {
+      "epoch": 0.1809965170835612,
+      "grad_norm": 3.050074228085725,
+      "learning_rate": 1.883258802854649e-06,
+      "loss": 0.6598,
+      "step": 393
+    },
+    {
+      "epoch": 0.18145706801761607,
+      "grad_norm": 3.3861305570759974,
+      "learning_rate": 1.8825580340844187e-06,
+      "loss": 0.6044,
+      "step": 394
+    },
+    {
+      "epoch": 0.18191761895167094,
+      "grad_norm": 3.4136999849562772,
+      "learning_rate": 1.8818552995168748e-06,
+      "loss": 0.6544,
+      "step": 395
+    },
+    {
+      "epoch": 0.1823781698857258,
+      "grad_norm": 3.3360169966873547,
+      "learning_rate": 1.8811506007172788e-06,
+      "loss": 0.6911,
+      "step": 396
+    },
+    {
+      "epoch": 0.18283872081978067,
+      "grad_norm": 3.0761653456275875,
+      "learning_rate": 1.8804439392552664e-06,
+      "loss": 0.5872,
+      "step": 397
+    },
+    {
+      "epoch": 0.18329927175383554,
+      "grad_norm": 3.466796092903531,
+      "learning_rate": 1.8797353167048457e-06,
+      "loss": 0.545,
+      "step": 398
+    },
+    {
+      "epoch": 0.1837598226878904,
+      "grad_norm": 3.0584194265671636,
+      "learning_rate": 1.8790247346443927e-06,
+      "loss": 0.5631,
+      "step": 399
+    },
+    {
+      "epoch": 0.18422037362194527,
+      "grad_norm": 3.149641541487599,
+      "learning_rate": 1.8783121946566473e-06,
+      "loss": 0.6112,
+      "step": 400
+    },
+    {
+      "epoch": 0.1846809245560001,
+      "grad_norm": 3.555627278211036,
+      "learning_rate": 1.8775976983287114e-06,
+      "loss": 0.571,
+      "step": 401
+    },
+    {
+      "epoch": 0.18514147549005497,
+      "grad_norm": 3.1940216915097412,
+      "learning_rate": 1.876881247252044e-06,
+      "loss": 0.6587,
+      "step": 402
+    },
+    {
+      "epoch": 0.18560202642410983,
+      "grad_norm": 3.488393433022656,
+      "learning_rate": 1.8761628430224582e-06,
+      "loss": 0.6844,
+      "step": 403
+    },
+    {
+      "epoch": 0.1860625773581647,
+      "grad_norm": 3.1438026269472075,
+      "learning_rate": 1.875442487240117e-06,
+      "loss": 0.5829,
+      "step": 404
+    },
+    {
+      "epoch": 0.18652312829221956,
+      "grad_norm": 3.441265497700973,
+      "learning_rate": 1.8747201815095313e-06,
+      "loss": 0.5813,
+      "step": 405
+    },
+    {
+      "epoch": 0.18698367922627443,
+      "grad_norm": 3.125275659245325,
+      "learning_rate": 1.8739959274395547e-06,
+      "loss": 0.5402,
+      "step": 406
+    },
+    {
+      "epoch": 0.1874442301603293,
+      "grad_norm": 3.182859255843592,
+      "learning_rate": 1.87326972664338e-06,
+      "loss": 0.5338,
+      "step": 407
+    },
+    {
+      "epoch": 0.18790478109438416,
+      "grad_norm": 3.3508341677793867,
+      "learning_rate": 1.8725415807385368e-06,
+      "loss": 0.596,
+      "step": 408
+    },
+    {
+      "epoch": 0.18836533202843903,
+      "grad_norm": 3.5167971779519616,
+      "learning_rate": 1.871811491346887e-06,
+      "loss": 0.738,
+      "step": 409
+    },
+    {
+      "epoch": 0.1888258829624939,
+      "grad_norm": 3.2498092239498844,
+      "learning_rate": 1.8710794600946216e-06,
+      "loss": 0.5168,
+      "step": 410
+    },
+    {
+      "epoch": 0.18928643389654876,
+      "grad_norm": 3.260212416401363,
+      "learning_rate": 1.8703454886122565e-06,
+      "loss": 0.5984,
+      "step": 411
+    },
+    {
+      "epoch": 0.18974698483060362,
+      "grad_norm": 3.507545427256196,
+      "learning_rate": 1.8696095785346295e-06,
+      "loss": 0.5627,
+      "step": 412
+    },
+    {
+      "epoch": 0.1902075357646585,
+      "grad_norm": 3.3634449621572466,
+      "learning_rate": 1.8688717315008962e-06,
+      "loss": 0.541,
+      "step": 413
+    },
+    {
+      "epoch": 0.19066808669871332,
+      "grad_norm": 3.2112093920249105,
+      "learning_rate": 1.8681319491545269e-06,
+      "loss": 0.6084,
+      "step": 414
+    },
+    {
+      "epoch": 0.1911286376327682,
+      "grad_norm": 3.5404353894665532,
+      "learning_rate": 1.8673902331433022e-06,
+      "loss": 0.5919,
+      "step": 415
+    },
+    {
+      "epoch": 0.19158918856682305,
+      "grad_norm": 3.104412575312322,
+      "learning_rate": 1.86664658511931e-06,
+      "loss": 0.6024,
+      "step": 416
+    },
+    {
+      "epoch": 0.19204973950087792,
+      "grad_norm": 3.1299398856732945,
+      "learning_rate": 1.8659010067389414e-06,
+      "loss": 0.55,
+      "step": 417
+    },
+    {
+      "epoch": 0.19251029043493278,
+      "grad_norm": 3.289621978009311,
+      "learning_rate": 1.8651534996628869e-06,
+      "loss": 0.6006,
+      "step": 418
+    },
+    {
+      "epoch": 0.19297084136898765,
+      "grad_norm": 3.388227682628728,
+      "learning_rate": 1.8644040655561334e-06,
+      "loss": 0.6185,
+      "step": 419
+    },
+    {
+      "epoch": 0.19343139230304252,
+      "grad_norm": 3.2963874382804432,
+      "learning_rate": 1.8636527060879601e-06,
+      "loss": 0.5446,
+      "step": 420
+    },
+    {
+      "epoch": 0.19389194323709738,
+      "grad_norm": 2.951087777995391,
+      "learning_rate": 1.8628994229319338e-06,
+      "loss": 0.5546,
+      "step": 421
+    },
+    {
+      "epoch": 0.19435249417115225,
+      "grad_norm": 2.970629687314148,
+      "learning_rate": 1.8621442177659076e-06,
+      "loss": 0.7266,
+      "step": 422
+    },
+    {
+      "epoch": 0.1948130451052071,
+      "grad_norm": 3.3425017993739696,
+      "learning_rate": 1.8613870922720145e-06,
+      "loss": 0.6919,
+      "step": 423
+    },
+    {
+      "epoch": 0.19527359603926198,
+      "grad_norm": 3.138471966309496,
+      "learning_rate": 1.8606280481366649e-06,
+      "loss": 0.6644,
+      "step": 424
+    },
+    {
+      "epoch": 0.19573414697331684,
+      "grad_norm": 2.931479609326468,
+      "learning_rate": 1.8598670870505434e-06,
+      "loss": 0.5852,
+      "step": 425
+    },
+    {
+      "epoch": 0.1961946979073717,
+      "grad_norm": 3.1399142652071563,
+      "learning_rate": 1.8591042107086038e-06,
+      "loss": 0.6798,
+      "step": 426
+    },
+    {
+      "epoch": 0.19665524884142654,
+      "grad_norm": 3.31831713584121,
+      "learning_rate": 1.8583394208100658e-06,
+      "loss": 0.6821,
+      "step": 427
+    },
+    {
+      "epoch": 0.1971157997754814,
+      "grad_norm": 3.570916313275249,
+      "learning_rate": 1.857572719058412e-06,
+      "loss": 0.6067,
+      "step": 428
+    },
+    {
+      "epoch": 0.19757635070953627,
+      "grad_norm": 3.3374646506424983,
+      "learning_rate": 1.8568041071613832e-06,
+      "loss": 0.5824,
+      "step": 429
+    },
+    {
+      "epoch": 0.19803690164359114,
+      "grad_norm": 3.45107101342791,
+      "learning_rate": 1.8560335868309742e-06,
+      "loss": 0.6396,
+      "step": 430
+    },
+    {
+      "epoch": 0.198497452577646,
+      "grad_norm": 2.832158757011958,
+      "learning_rate": 1.8552611597834317e-06,
+      "loss": 0.4881,
+      "step": 431
+    },
+    {
+      "epoch": 0.19895800351170087,
+      "grad_norm": 3.526958026283652,
+      "learning_rate": 1.8544868277392482e-06,
+      "loss": 0.5955,
+      "step": 432
+    },
+    {
+      "epoch": 0.19941855444575574,
+      "grad_norm": 3.368802966721466,
+      "learning_rate": 1.8537105924231601e-06,
+      "loss": 0.568,
+      "step": 433
+    },
+    {
+      "epoch": 0.1998791053798106,
+      "grad_norm": 2.9552863120111,
+      "learning_rate": 1.8529324555641436e-06,
+      "loss": 0.6433,
+      "step": 434
+    },
+    {
+      "epoch": 0.20033965631386547,
+      "grad_norm": 2.6981629403620615,
+      "learning_rate": 1.8521524188954091e-06,
+      "loss": 0.5556,
+      "step": 435
+    },
+    {
+      "epoch": 0.20080020724792033,
+      "grad_norm": 2.817878440174924,
+      "learning_rate": 1.8513704841543995e-06,
+      "loss": 0.5547,
+      "step": 436
+    },
+    {
+      "epoch": 0.2012607581819752,
+      "grad_norm": 3.2229669697907553,
+      "learning_rate": 1.8505866530827855e-06,
+      "loss": 0.5965,
+      "step": 437
+    },
+    {
+      "epoch": 0.20172130911603006,
+      "grad_norm": 3.1101407413877538,
+      "learning_rate": 1.8498009274264605e-06,
+      "loss": 0.5596,
+      "step": 438
+    },
+    {
+      "epoch": 0.20218186005008493,
+      "grad_norm": 3.3912236595059975,
+      "learning_rate": 1.8490133089355398e-06,
+      "loss": 0.7065,
+      "step": 439
+    },
+    {
+      "epoch": 0.20264241098413976,
+      "grad_norm": 3.0636194517862965,
+      "learning_rate": 1.848223799364353e-06,
+      "loss": 0.6438,
+      "step": 440
+    },
+    {
+      "epoch": 0.20310296191819463,
+      "grad_norm": 3.0364026600968277,
+      "learning_rate": 1.8474324004714428e-06,
+      "loss": 0.6763,
+      "step": 441
+    },
+    {
+      "epoch": 0.2035635128522495,
+      "grad_norm": 3.401042051162378,
+      "learning_rate": 1.8466391140195601e-06,
+      "loss": 0.6028,
+      "step": 442
+    },
+    {
+      "epoch": 0.20402406378630436,
+      "grad_norm": 3.324755759973617,
+      "learning_rate": 1.8458439417756594e-06,
+      "loss": 0.75,
+      "step": 443
+    },
+    {
+      "epoch": 0.20448461472035923,
+      "grad_norm": 3.452759888109864,
+      "learning_rate": 1.8450468855108969e-06,
+      "loss": 0.7129,
+      "step": 444
+    },
+    {
+      "epoch": 0.2049451656544141,
+      "grad_norm": 3.567128370695001,
+      "learning_rate": 1.8442479470006239e-06,
+      "loss": 0.6197,
+      "step": 445
+    },
+    {
+      "epoch": 0.20540571658846896,
+      "grad_norm": 3.3653450254819237,
+      "learning_rate": 1.843447128024385e-06,
+      "loss": 0.663,
+      "step": 446
+    },
+    {
+      "epoch": 0.20586626752252382,
+      "grad_norm": 3.3553923088948117,
+      "learning_rate": 1.842644430365913e-06,
+      "loss": 0.5946,
+      "step": 447
+    },
+    {
+      "epoch": 0.2063268184565787,
+      "grad_norm": 3.0905391763237806,
+      "learning_rate": 1.8418398558131257e-06,
+      "loss": 0.5987,
+      "step": 448
+    },
+    {
+      "epoch": 0.20678736939063355,
+      "grad_norm": 2.8613755616545875,
+      "learning_rate": 1.8410334061581206e-06,
+      "loss": 0.5291,
+      "step": 449
+    },
+    {
+      "epoch": 0.20724792032468842,
+      "grad_norm": 3.510631784297277,
+      "learning_rate": 1.8402250831971723e-06,
+      "loss": 0.5732,
+      "step": 450
+    },
+    {
+      "epoch": 0.20770847125874328,
+      "grad_norm": 3.5041502693832203,
+      "learning_rate": 1.8394148887307285e-06,
+      "loss": 0.585,
+      "step": 451
+    },
+    {
+      "epoch": 0.20816902219279815,
+      "grad_norm": 3.1445431808415556,
+      "learning_rate": 1.8386028245634041e-06,
+      "loss": 0.5904,
+      "step": 452
+    },
+    {
+      "epoch": 0.20862957312685299,
+      "grad_norm": 3.257773233991117,
+      "learning_rate": 1.83778889250398e-06,
+      "loss": 0.6259,
+      "step": 453
+    },
+    {
+      "epoch": 0.20909012406090785,
+      "grad_norm": 2.7341574429872764,
+      "learning_rate": 1.836973094365397e-06,
+      "loss": 0.4938,
+      "step": 454
+    },
+    {
+      "epoch": 0.20955067499496272,
+      "grad_norm": 3.6461025551874497,
+      "learning_rate": 1.8361554319647522e-06,
+      "loss": 0.6574,
+      "step": 455
+    },
+    {
+      "epoch": 0.21001122592901758,
+      "grad_norm": 3.278582866514977,
+      "learning_rate": 1.8353359071232951e-06,
+      "loss": 0.5728,
+      "step": 456
+    },
+    {
+      "epoch": 0.21047177686307245,
+      "grad_norm": 3.9462317493517025,
+      "learning_rate": 1.8345145216664242e-06,
+      "loss": 0.755,
+      "step": 457
+    },
+    {
+      "epoch": 0.2109323277971273,
+      "grad_norm": 3.481064251792427,
+      "learning_rate": 1.8336912774236818e-06,
+      "loss": 0.6502,
+      "step": 458
+    },
+    {
+      "epoch": 0.21139287873118218,
+      "grad_norm": 3.248277112137531,
+      "learning_rate": 1.83286617622875e-06,
+      "loss": 0.5746,
+      "step": 459
+    },
+    {
+      "epoch": 0.21185342966523704,
+      "grad_norm": 3.3146346341451918,
+      "learning_rate": 1.8320392199194476e-06,
+      "loss": 0.605,
+      "step": 460
+    },
+    {
+      "epoch": 0.2123139805992919,
+      "grad_norm": 3.441495638998992,
+      "learning_rate": 1.8312104103377262e-06,
+      "loss": 0.6416,
+      "step": 461
+    },
+    {
+      "epoch": 0.21277453153334677,
+      "grad_norm": 2.9904476105113065,
+      "learning_rate": 1.8303797493296637e-06,
+      "loss": 0.6347,
+      "step": 462
+    },
+    {
+      "epoch": 0.21323508246740164,
+      "grad_norm": 3.63776364142898,
+      "learning_rate": 1.8295472387454636e-06,
+      "loss": 0.4796,
+      "step": 463
+    },
+    {
+      "epoch": 0.2136956334014565,
+      "grad_norm": 3.0013856019668763,
+      "learning_rate": 1.8287128804394474e-06,
+      "loss": 0.6319,
+      "step": 464
+    },
+    {
+      "epoch": 0.21415618433551137,
+      "grad_norm": 3.0880145601287783,
+      "learning_rate": 1.8278766762700534e-06,
+      "loss": 0.5423,
+      "step": 465
+    },
+    {
+      "epoch": 0.2146167352695662,
+      "grad_norm": 3.133296021110922,
+      "learning_rate": 1.8270386280998309e-06,
+      "loss": 0.5982,
+      "step": 466
+    },
+    {
+      "epoch": 0.21507728620362107,
+      "grad_norm": 3.4704322121668914,
+      "learning_rate": 1.8261987377954365e-06,
+      "loss": 0.6409,
+      "step": 467
+    },
+    {
+      "epoch": 0.21553783713767594,
+      "grad_norm": 3.0683920370476203,
+      "learning_rate": 1.8253570072276303e-06,
+      "loss": 0.6111,
+      "step": 468
+    },
+    {
+      "epoch": 0.2159983880717308,
+      "grad_norm": 3.2983986131430663,
+      "learning_rate": 1.8245134382712709e-06,
+      "loss": 0.6993,
+      "step": 469
+    },
+    {
+      "epoch": 0.21645893900578567,
+      "grad_norm": 3.2695193504256754,
+      "learning_rate": 1.8236680328053116e-06,
+      "loss": 0.5836,
+      "step": 470
+    },
+    {
+      "epoch": 0.21691948993984053,
+      "grad_norm": 3.1255673883544612,
+      "learning_rate": 1.822820792712797e-06,
+      "loss": 0.5577,
+      "step": 471
+    },
+    {
+      "epoch": 0.2173800408738954,
+      "grad_norm": 3.6062462823524735,
+      "learning_rate": 1.8219717198808578e-06,
+      "loss": 0.6389,
+      "step": 472
+    },
+    {
+      "epoch": 0.21784059180795026,
+      "grad_norm": 3.393965634932339,
+      "learning_rate": 1.8211208162007065e-06,
+      "loss": 0.5364,
+      "step": 473
+    },
+    {
+      "epoch": 0.21830114274200513,
+      "grad_norm": 3.244656999193985,
+      "learning_rate": 1.820268083567634e-06,
+      "loss": 0.525,
+      "step": 474
+    },
+    {
+      "epoch": 0.21876169367606,
+      "grad_norm": 3.2505550753334664,
+      "learning_rate": 1.819413523881005e-06,
+      "loss": 0.6237,
+      "step": 475
+    },
+    {
+      "epoch": 0.21922224461011486,
+      "grad_norm": 3.0423831974765894,
+      "learning_rate": 1.818557139044254e-06,
+      "loss": 0.5472,
+      "step": 476
+    },
+    {
+      "epoch": 0.21968279554416972,
+      "grad_norm": 2.9954662337432656,
+      "learning_rate": 1.8176989309648803e-06,
+      "loss": 0.5291,
+      "step": 477
+    },
+    {
+      "epoch": 0.2201433464782246,
+      "grad_norm": 2.918815185530408,
+      "learning_rate": 1.8168389015544444e-06,
+      "loss": 0.5379,
+      "step": 478
+    },
+    {
+      "epoch": 0.22060389741227943,
+      "grad_norm": 3.265440208483845,
+      "learning_rate": 1.8159770527285634e-06,
+      "loss": 0.5796,
+      "step": 479
+    },
+    {
+      "epoch": 0.2210644483463343,
+      "grad_norm": 3.3450099140811806,
+      "learning_rate": 1.815113386406908e-06,
+      "loss": 0.5663,
+      "step": 480
+    },
+    {
+      "epoch": 0.22152499928038916,
+      "grad_norm": 2.8483938675266547,
+      "learning_rate": 1.8142479045131953e-06,
+      "loss": 0.6591,
+      "step": 481
+    },
+    {
+      "epoch": 0.22198555021444402,
+      "grad_norm": 3.1336266713837744,
+      "learning_rate": 1.8133806089751884e-06,
+      "loss": 0.4984,
+      "step": 482
+    },
+    {
+      "epoch": 0.2224461011484989,
+      "grad_norm": 3.080985904456354,
+      "learning_rate": 1.8125115017246887e-06,
+      "loss": 0.4414,
+      "step": 483
+    },
+    {
+      "epoch": 0.22290665208255375,
+      "grad_norm": 3.371599831924596,
+      "learning_rate": 1.8116405846975335e-06,
+      "loss": 0.5807,
+      "step": 484
+    },
+    {
+      "epoch": 0.22336720301660862,
+      "grad_norm": 3.078985566820636,
+      "learning_rate": 1.8107678598335912e-06,
+      "loss": 0.6484,
+      "step": 485
+    },
+    {
+      "epoch": 0.22382775395066348,
+      "grad_norm": 2.9841159659031526,
+      "learning_rate": 1.8098933290767567e-06,
+      "loss": 0.569,
+      "step": 486
+    },
+    {
+      "epoch": 0.22428830488471835,
+      "grad_norm": 3.110058890834877,
+      "learning_rate": 1.8090169943749474e-06,
+      "loss": 0.6511,
+      "step": 487
+    },
+    {
+      "epoch": 0.2247488558187732,
+      "grad_norm": 3.153179145491399,
+      "learning_rate": 1.808138857680099e-06,
+      "loss": 0.4959,
+      "step": 488
+    },
+    {
+      "epoch": 0.22520940675282808,
+      "grad_norm": 3.004599141952458,
+      "learning_rate": 1.8072589209481607e-06,
+      "loss": 0.5327,
+      "step": 489
+    },
+    {
+      "epoch": 0.22566995768688294,
+      "grad_norm": 3.3051175185768273,
+      "learning_rate": 1.8063771861390915e-06,
+      "loss": 0.6466,
+      "step": 490
+    },
+    {
+      "epoch": 0.2261305086209378,
+      "grad_norm": 3.096130029682352,
+      "learning_rate": 1.8054936552168547e-06,
+      "loss": 0.4745,
+      "step": 491
+    },
+    {
+      "epoch": 0.22659105955499265,
+      "grad_norm": 3.756055051745282,
+      "learning_rate": 1.804608330149415e-06,
+      "loss": 0.6128,
+      "step": 492
+    },
+    {
+      "epoch": 0.2270516104890475,
+      "grad_norm": 3.342218840047437,
+      "learning_rate": 1.8037212129087322e-06,
+      "loss": 0.561,
+      "step": 493
+    },
+    {
+      "epoch": 0.22751216142310238,
+      "grad_norm": 3.244488624559804,
+      "learning_rate": 1.8028323054707592e-06,
+      "loss": 0.6411,
+      "step": 494
+    },
+    {
+      "epoch": 0.22797271235715724,
+      "grad_norm": 3.3116315535429055,
+      "learning_rate": 1.801941609815436e-06,
+      "loss": 0.6216,
+      "step": 495
+    },
+    {
+      "epoch": 0.2284332632912121,
+      "grad_norm": 3.3075500193801775,
+      "learning_rate": 1.8010491279266858e-06,
+      "loss": 0.6051,
+      "step": 496
+    },
+    {
+      "epoch": 0.22889381422526697,
+      "grad_norm": 3.0538650883239,
+      "learning_rate": 1.8001548617924095e-06,
+      "loss": 0.4452,
+      "step": 497
+    },
+    {
+      "epoch": 0.22935436515932184,
+      "grad_norm": 3.5424554842795835,
+      "learning_rate": 1.799258813404483e-06,
+      "loss": 0.6483,
+      "step": 498
+    },
+    {
+      "epoch": 0.2298149160933767,
+      "grad_norm": 3.3078686719248593,
+      "learning_rate": 1.7983609847587521e-06,
+      "loss": 0.585,
+      "step": 499
+    },
+    {
+      "epoch": 0.23027546702743157,
+      "grad_norm": 3.3972513361686185,
+      "learning_rate": 1.7974613778550278e-06,
+      "loss": 0.6655,
+      "step": 500
+    },
+    {
+      "epoch": 0.23073601796148643,
+      "grad_norm": 3.465463703248316,
+      "learning_rate": 1.7965599946970812e-06,
+      "loss": 0.6034,
+      "step": 501
+    },
+    {
+      "epoch": 0.2311965688955413,
+      "grad_norm": 3.325451562972118,
+      "learning_rate": 1.795656837292641e-06,
+      "loss": 0.6398,
+      "step": 502
+    },
+    {
+      "epoch": 0.23165711982959616,
+      "grad_norm": 3.5458965409990593,
+      "learning_rate": 1.7947519076533873e-06,
+      "loss": 0.5217,
+      "step": 503
+    },
+    {
+      "epoch": 0.23211767076365103,
+      "grad_norm": 3.3055864844064904,
+      "learning_rate": 1.793845207794947e-06,
+      "loss": 0.633,
+      "step": 504
+    },
+    {
+      "epoch": 0.23257822169770587,
+      "grad_norm": 3.353646102717693,
+      "learning_rate": 1.7929367397368913e-06,
+      "loss": 0.5797,
+      "step": 505
+    },
+    {
+      "epoch": 0.23303877263176073,
+      "grad_norm": 3.254837375875168,
+      "learning_rate": 1.7920265055027283e-06,
+      "loss": 0.5433,
+      "step": 506
+    },
+    {
+      "epoch": 0.2334993235658156,
+      "grad_norm": 3.5539641885768214,
+      "learning_rate": 1.7911145071199018e-06,
+      "loss": 0.6674,
+      "step": 507
+    },
+    {
+      "epoch": 0.23395987449987046,
+      "grad_norm": 3.4968687641517584,
+      "learning_rate": 1.7902007466197837e-06,
+      "loss": 0.5679,
+      "step": 508
+    },
+    {
+      "epoch": 0.23442042543392533,
+      "grad_norm": 3.316075449684041,
+      "learning_rate": 1.789285226037671e-06,
+      "loss": 0.6063,
+      "step": 509
+    },
+    {
+      "epoch": 0.2348809763679802,
+      "grad_norm": 3.4021632966156696,
+      "learning_rate": 1.788367947412782e-06,
+      "loss": 0.5539,
+      "step": 510
+    },
+    {
+      "epoch": 0.23534152730203506,
+      "grad_norm": 3.1412284614498103,
+      "learning_rate": 1.7874489127882493e-06,
+      "loss": 0.5244,
+      "step": 511
+    },
+    {
+      "epoch": 0.23580207823608992,
+      "grad_norm": 3.6411365029508143,
+      "learning_rate": 1.7865281242111182e-06,
+      "loss": 0.6874,
+      "step": 512
+    },
+    {
+      "epoch": 0.2362626291701448,
+      "grad_norm": 2.939080873360146,
+      "learning_rate": 1.7856055837323406e-06,
+      "loss": 0.6462,
+      "step": 513
+    },
+    {
+      "epoch": 0.23672318010419965,
+      "grad_norm": 3.5060804854877334,
+      "learning_rate": 1.784681293406769e-06,
+      "loss": 0.6256,
+      "step": 514
+    },
+    {
+      "epoch": 0.23718373103825452,
+      "grad_norm": 3.454373638780033,
+      "learning_rate": 1.7837552552931555e-06,
+      "loss": 0.5419,
+      "step": 515
+    },
+    {
+      "epoch": 0.23764428197230938,
+      "grad_norm": 2.981684278208563,
+      "learning_rate": 1.7828274714541443e-06,
+      "loss": 0.6395,
+      "step": 516
+    },
+    {
+      "epoch": 0.23810483290636425,
+      "grad_norm": 2.6776702286309413,
+      "learning_rate": 1.7818979439562677e-06,
+      "loss": 0.4575,
+      "step": 517
+    },
+    {
+      "epoch": 0.23856538384041912,
+      "grad_norm": 3.0713454976431342,
+      "learning_rate": 1.7809666748699424e-06,
+      "loss": 0.6557,
+      "step": 518
+    },
+    {
+      "epoch": 0.23902593477447395,
+      "grad_norm": 3.332095530691328,
+      "learning_rate": 1.7800336662694635e-06,
+      "loss": 0.5726,
+      "step": 519
+    },
+    {
+      "epoch": 0.23948648570852882,
+      "grad_norm": 3.3759244307287375,
+      "learning_rate": 1.7790989202330018e-06,
+      "loss": 0.5942,
+      "step": 520
+    },
+    {
+      "epoch": 0.23994703664258368,
+      "grad_norm": 3.3882812703718583,
+      "learning_rate": 1.7781624388425973e-06,
+      "loss": 0.6916,
+      "step": 521
+    },
+    {
+      "epoch": 0.24040758757663855,
+      "grad_norm": 3.4802417994319,
+      "learning_rate": 1.7772242241841552e-06,
+      "loss": 0.5549,
+      "step": 522
+    },
+    {
+      "epoch": 0.2408681385106934,
+      "grad_norm": 2.8321037056813925,
+      "learning_rate": 1.7762842783474419e-06,
+      "loss": 0.5675,
+      "step": 523
+    },
+    {
+      "epoch": 0.24132868944474828,
+      "grad_norm": 3.0286366892013024,
+      "learning_rate": 1.7753426034260794e-06,
+      "loss": 0.6203,
+      "step": 524
+    },
+    {
+      "epoch": 0.24178924037880314,
+      "grad_norm": 3.2252778331423255,
+      "learning_rate": 1.774399201517541e-06,
+      "loss": 0.6389,
+      "step": 525
+    },
+    {
+      "epoch": 0.242249791312858,
+      "grad_norm": 3.0971473717588087,
+      "learning_rate": 1.7734540747231469e-06,
+      "loss": 0.6017,
+      "step": 526
+    },
+    {
+      "epoch": 0.24271034224691287,
+      "grad_norm": 3.4994719951042104,
+      "learning_rate": 1.772507225148059e-06,
+      "loss": 0.4603,
+      "step": 527
+    },
+    {
+      "epoch": 0.24317089318096774,
+      "grad_norm": 2.713673764054212,
+      "learning_rate": 1.7715586549012768e-06,
+      "loss": 0.5765,
+      "step": 528
+    },
+    {
+      "epoch": 0.2436314441150226,
+      "grad_norm": 3.221284406276941,
+      "learning_rate": 1.7706083660956327e-06,
+      "loss": 0.6266,
+      "step": 529
+    },
+    {
+      "epoch": 0.24409199504907747,
+      "grad_norm": 3.306672434466321,
+      "learning_rate": 1.7696563608477862e-06,
+      "loss": 0.6451,
+      "step": 530
+    },
+    {
+      "epoch": 0.24455254598313234,
+      "grad_norm": 3.065434289834053,
+      "learning_rate": 1.76870264127822e-06,
+      "loss": 0.6552,
+      "step": 531
+    },
+    {
+      "epoch": 0.24501309691718717,
+      "grad_norm": 3.356746881862158,
+      "learning_rate": 1.7677472095112363e-06,
+      "loss": 0.4904,
+      "step": 532
+    },
+    {
+      "epoch": 0.24547364785124204,
+      "grad_norm": 3.253784280023295,
+      "learning_rate": 1.7667900676749498e-06,
+      "loss": 0.5361,
+      "step": 533
+    },
+    {
+      "epoch": 0.2459341987852969,
+      "grad_norm": 3.0192332980381344,
+      "learning_rate": 1.7658312179012854e-06,
+      "loss": 0.5099,
+      "step": 534
+    },
+    {
+      "epoch": 0.24639474971935177,
+      "grad_norm": 3.0809296396824384,
+      "learning_rate": 1.7648706623259706e-06,
+      "loss": 0.6658,
+      "step": 535
+    },
+    {
+      "epoch": 0.24685530065340663,
+      "grad_norm": 3.6830695514949077,
+      "learning_rate": 1.7639084030885338e-06,
+      "loss": 0.5973,
+      "step": 536
+    },
+    {
+      "epoch": 0.2473158515874615,
+      "grad_norm": 3.330551105763686,
+      "learning_rate": 1.7629444423322982e-06,
+      "loss": 0.7364,
+      "step": 537
+    },
+    {
+      "epoch": 0.24777640252151636,
+      "grad_norm": 2.9911270084180366,
+      "learning_rate": 1.7619787822043754e-06,
+      "loss": 0.462,
+      "step": 538
+    },
+    {
+      "epoch": 0.24823695345557123,
+      "grad_norm": 3.149958771441102,
+      "learning_rate": 1.7610114248556639e-06,
+      "loss": 0.611,
+      "step": 539
+    },
+    {
+      "epoch": 0.2486975043896261,
+      "grad_norm": 3.30756958264156,
+      "learning_rate": 1.7600423724408415e-06,
+      "loss": 0.5285,
+      "step": 540
+    },
+    {
+      "epoch": 0.24915805532368096,
+      "grad_norm": 3.3029379864351,
+      "learning_rate": 1.759071627118362e-06,
+      "loss": 0.5643,
+      "step": 541
+    },
+    {
+      "epoch": 0.24961860625773583,
+      "grad_norm": 3.454862601842869,
+      "learning_rate": 1.7580991910504494e-06,
+      "loss": 0.6229,
+      "step": 542
+    },
+    {
+      "epoch": 0.2500791571917907,
+      "grad_norm": 3.2773777585364616,
+      "learning_rate": 1.7571250664030947e-06,
+      "loss": 0.6066,
+      "step": 543
+    },
+    {
+      "epoch": 0.25053970812584553,
+      "grad_norm": 2.955729624984104,
+      "learning_rate": 1.7561492553460488e-06,
+      "loss": 0.5717,
+      "step": 544
+    },
+    {
+      "epoch": 0.2510002590599004,
+      "grad_norm": 3.103695274249137,
+      "learning_rate": 1.7551717600528203e-06,
+      "loss": 0.5664,
+      "step": 545
+    },
+    {
+      "epoch": 0.25146080999395526,
+      "grad_norm": 3.1965303115665034,
+      "learning_rate": 1.7541925827006678e-06,
+      "loss": 0.5587,
+      "step": 546
+    },
+    {
+      "epoch": 0.25192136092801015,
+      "grad_norm": 2.9103203193249545,
+      "learning_rate": 1.7532117254705972e-06,
+      "loss": 0.6035,
+      "step": 547
+    },
+    {
+      "epoch": 0.252381911862065,
+      "grad_norm": 2.9173381955892976,
+      "learning_rate": 1.7522291905473558e-06,
+      "loss": 0.4366,
+      "step": 548
+    },
+    {
+      "epoch": 0.2528424627961199,
+      "grad_norm": 3.222972546291656,
+      "learning_rate": 1.7512449801194286e-06,
+      "loss": 0.5781,
+      "step": 549
+    },
+    {
+      "epoch": 0.2533030137301747,
+      "grad_norm": 3.1103988583748663,
+      "learning_rate": 1.7502590963790316e-06,
+      "loss": 0.5862,
+      "step": 550
+    },
+    {
+      "epoch": 0.25376356466422956,
+      "grad_norm": 3.4396946239347908,
+      "learning_rate": 1.7492715415221087e-06,
+      "loss": 0.6496,
+      "step": 551
+    },
+    {
+      "epoch": 0.25422411559828445,
+      "grad_norm": 2.9802260714638398,
+      "learning_rate": 1.7482823177483252e-06,
+      "loss": 0.6773,
+      "step": 552
+    },
+    {
+      "epoch": 0.2546846665323393,
+      "grad_norm": 3.082234293267444,
+      "learning_rate": 1.7472914272610646e-06,
+      "loss": 0.5246,
+      "step": 553
+    },
+    {
+      "epoch": 0.2551452174663942,
+      "grad_norm": 3.3567221598735992,
+      "learning_rate": 1.7462988722674221e-06,
+      "loss": 0.6232,
+      "step": 554
+    },
+    {
+      "epoch": 0.255605768400449,
+      "grad_norm": 3.2626603235826264,
+      "learning_rate": 1.7453046549782009e-06,
+      "loss": 0.5204,
+      "step": 555
+    },
+    {
+      "epoch": 0.2560663193345039,
+      "grad_norm": 3.2339294619977434,
+      "learning_rate": 1.7443087776079064e-06,
+      "loss": 0.5647,
+      "step": 556
+    },
+    {
+      "epoch": 0.25652687026855875,
+      "grad_norm": 3.3723328031579722,
+      "learning_rate": 1.7433112423747418e-06,
+      "loss": 0.6539,
+      "step": 557
+    },
+    {
+      "epoch": 0.25698742120261364,
+      "grad_norm": 3.184022192143854,
+      "learning_rate": 1.742312051500603e-06,
+      "loss": 0.6068,
+      "step": 558
+    },
+    {
+      "epoch": 0.2574479721366685,
+      "grad_norm": 3.5355428533513695,
+      "learning_rate": 1.741311207211074e-06,
+      "loss": 0.6395,
+      "step": 559
+    },
+    {
+      "epoch": 0.25790852307072337,
+      "grad_norm": 3.303032162182857,
+      "learning_rate": 1.740308711735421e-06,
+      "loss": 0.5797,
+      "step": 560
+    },
+    {
+      "epoch": 0.2583690740047782,
+      "grad_norm": 3.2151549157821235,
+      "learning_rate": 1.7393045673065878e-06,
+      "loss": 0.5189,
+      "step": 561
+    },
+    {
+      "epoch": 0.2588296249388331,
+      "grad_norm": 3.2889581937560513,
+      "learning_rate": 1.7382987761611924e-06,
+      "loss": 0.5853,
+      "step": 562
+    },
+    {
+      "epoch": 0.25929017587288794,
+      "grad_norm": 3.589638793664337,
+      "learning_rate": 1.7372913405395191e-06,
+      "loss": 0.591,
+      "step": 563
+    },
+    {
+      "epoch": 0.2597507268069428,
+      "grad_norm": 3.1711968406048734,
+      "learning_rate": 1.7362822626855165e-06,
+      "loss": 0.5466,
+      "step": 564
+    },
+    {
+      "epoch": 0.26021127774099767,
+      "grad_norm": 2.816516792416395,
+      "learning_rate": 1.7352715448467895e-06,
+      "loss": 0.5831,
+      "step": 565
+    },
+    {
+      "epoch": 0.2606718286750525,
+      "grad_norm": 3.154750601104403,
+      "learning_rate": 1.7342591892745974e-06,
+      "loss": 0.6726,
+      "step": 566
+    },
+    {
+      "epoch": 0.2611323796091074,
+      "grad_norm": 3.3840134771829637,
+      "learning_rate": 1.7332451982238465e-06,
+      "loss": 0.5483,
+      "step": 567
+    },
+    {
+      "epoch": 0.26159293054316224,
+      "grad_norm": 3.1406500671773645,
+      "learning_rate": 1.732229573953086e-06,
+      "loss": 0.6034,
+      "step": 568
+    },
+    {
+      "epoch": 0.26205348147721713,
+      "grad_norm": 2.862334788278051,
+      "learning_rate": 1.7312123187245037e-06,
+      "loss": 0.6242,
+      "step": 569
+    },
+    {
+      "epoch": 0.26251403241127197,
+      "grad_norm": 2.8647709222697095,
+      "learning_rate": 1.7301934348039188e-06,
+      "loss": 0.5623,
+      "step": 570
+    },
+    {
+      "epoch": 0.26297458334532686,
+      "grad_norm": 2.8749482072040458,
+      "learning_rate": 1.7291729244607795e-06,
+      "loss": 0.568,
+      "step": 571
+    },
+    {
+      "epoch": 0.2634351342793817,
+      "grad_norm": 3.1262185261082935,
+      "learning_rate": 1.7281507899681556e-06,
+      "loss": 0.5939,
+      "step": 572
+    },
+    {
+      "epoch": 0.2638956852134366,
+      "grad_norm": 3.2312398719111606,
+      "learning_rate": 1.7271270336027358e-06,
+      "loss": 0.5729,
+      "step": 573
+    },
+    {
+      "epoch": 0.26435623614749143,
+      "grad_norm": 3.035067327482627,
+      "learning_rate": 1.7261016576448198e-06,
+      "loss": 0.5952,
+      "step": 574
+    },
+    {
+      "epoch": 0.2648167870815463,
+      "grad_norm": 3.386614804112364,
+      "learning_rate": 1.7250746643783158e-06,
+      "loss": 0.7062,
+      "step": 575
+    },
+    {
+      "epoch": 0.26527733801560116,
+      "grad_norm": 3.3687551890454315,
+      "learning_rate": 1.7240460560907343e-06,
+      "loss": 0.4614,
+      "step": 576
+    },
+    {
+      "epoch": 0.26573788894965605,
+      "grad_norm": 2.9333951618662675,
+      "learning_rate": 1.7230158350731832e-06,
+      "loss": 0.5462,
+      "step": 577
+    },
+    {
+      "epoch": 0.2661984398837109,
+      "grad_norm": 3.2883977513315714,
+      "learning_rate": 1.7219840036203613e-06,
+      "loss": 0.668,
+      "step": 578
+    },
+    {
+      "epoch": 0.26665899081776573,
+      "grad_norm": 3.319840681654177,
+      "learning_rate": 1.7209505640305562e-06,
+      "loss": 0.5502,
+      "step": 579
+    },
+    {
+      "epoch": 0.2671195417518206,
+      "grad_norm": 3.344576294088726,
+      "learning_rate": 1.7199155186056364e-06,
+      "loss": 0.7453,
+      "step": 580
+    },
+    {
+      "epoch": 0.26758009268587546,
+      "grad_norm": 3.3507561207198266,
+      "learning_rate": 1.7188788696510474e-06,
+      "loss": 0.5591,
+      "step": 581
+    },
+    {
+      "epoch": 0.26804064361993035,
+      "grad_norm": 3.6763295819772,
+      "learning_rate": 1.7178406194758067e-06,
+      "loss": 0.5894,
+      "step": 582
+    },
+    {
+      "epoch": 0.2685011945539852,
+      "grad_norm": 2.7982715860658836,
+      "learning_rate": 1.7168007703924978e-06,
+      "loss": 0.6436,
+      "step": 583
+    },
+    {
+      "epoch": 0.2689617454880401,
+      "grad_norm": 3.1119980909697724,
+      "learning_rate": 1.7157593247172664e-06,
+      "loss": 0.557,
+      "step": 584
+    },
+    {
+      "epoch": 0.2694222964220949,
+      "grad_norm": 3.69421493980708,
+      "learning_rate": 1.714716284769813e-06,
+      "loss": 0.6445,
+      "step": 585
+    },
+    {
+      "epoch": 0.2698828473561498,
+      "grad_norm": 3.786080820523942,
+      "learning_rate": 1.7136716528733912e-06,
+      "loss": 0.417,
+      "step": 586
+    },
+    {
+      "epoch": 0.27034339829020465,
+      "grad_norm": 4.270736109571739,
+      "learning_rate": 1.7126254313547985e-06,
+      "loss": 0.6734,
+      "step": 587
+    },
+    {
+      "epoch": 0.27080394922425954,
+      "grad_norm": 3.600772785847931,
+      "learning_rate": 1.7115776225443739e-06,
+      "loss": 0.8145,
+      "step": 588
+    },
+    {
+      "epoch": 0.2712645001583144,
+      "grad_norm": 2.8943620464394293,
+      "learning_rate": 1.7105282287759926e-06,
+      "loss": 0.528,
+      "step": 589
+    },
+    {
+      "epoch": 0.2717250510923693,
+      "grad_norm": 3.0063394751214094,
+      "learning_rate": 1.7094772523870589e-06,
+      "loss": 0.5868,
+      "step": 590
+    },
+    {
+      "epoch": 0.2721856020264241,
+      "grad_norm": 3.3735726594180613,
+      "learning_rate": 1.7084246957185032e-06,
+      "loss": 0.6339,
+      "step": 591
+    },
+    {
+      "epoch": 0.27264615296047895,
+      "grad_norm": 3.3141049773629283,
+      "learning_rate": 1.707370561114775e-06,
+      "loss": 0.6053,
+      "step": 592
+    },
+    {
+      "epoch": 0.27310670389453384,
+      "grad_norm": 2.80960222995646,
+      "learning_rate": 1.7063148509238393e-06,
+      "loss": 0.4687,
+      "step": 593
+    },
+    {
+      "epoch": 0.2735672548285887,
+      "grad_norm": 3.313745439020705,
+      "learning_rate": 1.70525756749717e-06,
+      "loss": 0.6093,
+      "step": 594
+    },
+    {
+      "epoch": 0.2740278057626436,
+      "grad_norm": 3.7937269183858318,
+      "learning_rate": 1.7041987131897445e-06,
+      "loss": 0.5746,
+      "step": 595
+    },
+    {
+      "epoch": 0.2744883566966984,
+      "grad_norm": 2.9125080429436365,
+      "learning_rate": 1.703138290360041e-06,
+      "loss": 0.525,
+      "step": 596
+    },
+    {
+      "epoch": 0.2749489076307533,
+      "grad_norm": 3.2588529493155285,
+      "learning_rate": 1.7020763013700296e-06,
+      "loss": 0.5566,
+      "step": 597
+    },
+    {
+      "epoch": 0.27540945856480814,
+      "grad_norm": 3.0202214315810396,
+      "learning_rate": 1.70101274858517e-06,
+      "loss": 0.5711,
+      "step": 598
+    },
+    {
+      "epoch": 0.27587000949886303,
+      "grad_norm": 3.3568365962069664,
+      "learning_rate": 1.6999476343744047e-06,
+      "loss": 0.5359,
+      "step": 599
+    },
+    {
+      "epoch": 0.27633056043291787,
+      "grad_norm": 3.0103484901671638,
+      "learning_rate": 1.6988809611101535e-06,
+      "loss": 0.6112,
+      "step": 600
+    },
+    {
+      "epoch": 0.27679111136697276,
+      "grad_norm": 3.4029649629281673,
+      "learning_rate": 1.69781273116831e-06,
+      "loss": 0.5336,
+      "step": 601
+    },
+    {
+      "epoch": 0.2772516623010276,
+      "grad_norm": 2.8767133982019377,
+      "learning_rate": 1.6967429469282345e-06,
+      "loss": 0.6359,
+      "step": 602
+    },
+    {
+      "epoch": 0.2777122132350825,
+      "grad_norm": 3.2835587382748024,
+      "learning_rate": 1.695671610772749e-06,
+      "loss": 0.5618,
+      "step": 603
+    },
+    {
+      "epoch": 0.27817276416913733,
+      "grad_norm": 3.5991470633371283,
+      "learning_rate": 1.694598725088133e-06,
+      "loss": 0.5286,
+      "step": 604
+    },
+    {
+      "epoch": 0.27863331510319217,
+      "grad_norm": 3.1120545309975958,
+      "learning_rate": 1.6935242922641159e-06,
+      "loss": 0.6178,
+      "step": 605
+    },
+    {
+      "epoch": 0.27909386603724706,
+      "grad_norm": 3.1244820282575674,
+      "learning_rate": 1.6924483146938754e-06,
+      "loss": 0.5822,
+      "step": 606
+    },
+    {
+      "epoch": 0.2795544169713019,
+      "grad_norm": 2.714150777824839,
+      "learning_rate": 1.6913707947740284e-06,
+      "loss": 0.5298,
+      "step": 607
+    },
+    {
+      "epoch": 0.2800149679053568,
+      "grad_norm": 3.3837453901982486,
+      "learning_rate": 1.690291734904627e-06,
+      "loss": 0.5653,
+      "step": 608
+    },
+    {
+      "epoch": 0.28047551883941163,
+      "grad_norm": 3.1917062613137785,
+      "learning_rate": 1.6892111374891547e-06,
+      "loss": 0.5825,
+      "step": 609
+    },
+    {
+      "epoch": 0.2809360697734665,
+      "grad_norm": 3.3940845730675226,
+      "learning_rate": 1.6881290049345185e-06,
+      "loss": 0.5283,
+      "step": 610
+    },
+    {
+      "epoch": 0.28139662070752136,
+      "grad_norm": 3.3197149723223127,
+      "learning_rate": 1.6870453396510453e-06,
+      "loss": 0.5643,
+      "step": 611
+    },
+    {
+      "epoch": 0.28185717164157625,
+      "grad_norm": 3.368272451697648,
+      "learning_rate": 1.6859601440524757e-06,
+      "loss": 0.5448,
+      "step": 612
+    },
+    {
+      "epoch": 0.2823177225756311,
+      "grad_norm": 3.66743262346132,
+      "learning_rate": 1.6848734205559593e-06,
+      "loss": 0.5931,
+      "step": 613
+    },
+    {
+      "epoch": 0.282778273509686,
+      "grad_norm": 3.399278333660865,
+      "learning_rate": 1.6837851715820488e-06,
+      "loss": 0.6151,
+      "step": 614
+    },
+    {
+      "epoch": 0.2832388244437408,
+      "grad_norm": 3.57422302010664,
+      "learning_rate": 1.6826953995546945e-06,
+      "loss": 0.5642,
+      "step": 615
+    },
+    {
+      "epoch": 0.2836993753777957,
+      "grad_norm": 3.0447865419845286,
+      "learning_rate": 1.6816041069012388e-06,
+      "loss": 0.5781,
+      "step": 616
+    },
+    {
+      "epoch": 0.28415992631185055,
+      "grad_norm": 3.072556146750414,
+      "learning_rate": 1.680511296052412e-06,
+      "loss": 0.5839,
+      "step": 617
+    },
+    {
+      "epoch": 0.2846204772459054,
+      "grad_norm": 3.7455049180528484,
+      "learning_rate": 1.6794169694423257e-06,
+      "loss": 0.6444,
+      "step": 618
+    },
+    {
+      "epoch": 0.2850810281799603,
+      "grad_norm": 3.0929934631543574,
+      "learning_rate": 1.6783211295084669e-06,
+      "loss": 0.6054,
+      "step": 619
+    },
+    {
+      "epoch": 0.2855415791140151,
+      "grad_norm": 2.7963129880952415,
+      "learning_rate": 1.677223778691695e-06,
+      "loss": 0.586,
+      "step": 620
+    },
+    {
+      "epoch": 0.28600213004807,
+      "grad_norm": 3.0409334266150263,
+      "learning_rate": 1.6761249194362328e-06,
+      "loss": 0.6534,
+      "step": 621
+    },
+    {
+      "epoch": 0.28646268098212485,
+      "grad_norm": 3.541465069260458,
+      "learning_rate": 1.6750245541896644e-06,
+      "loss": 0.7103,
+      "step": 622
+    },
+    {
+      "epoch": 0.28692323191617974,
+      "grad_norm": 3.2927125901227012,
+      "learning_rate": 1.6739226854029276e-06,
+      "loss": 0.493,
+      "step": 623
+    },
+    {
+      "epoch": 0.2873837828502346,
+      "grad_norm": 2.761658471122107,
+      "learning_rate": 1.6728193155303097e-06,
+      "loss": 0.537,
+      "step": 624
+    },
+    {
+      "epoch": 0.2878443337842895,
+      "grad_norm": 3.362612929819794,
+      "learning_rate": 1.6717144470294406e-06,
+      "loss": 0.6486,
+      "step": 625
+    },
+    {
+      "epoch": 0.2883048847183443,
+      "grad_norm": 3.2064609899615877,
+      "learning_rate": 1.6706080823612894e-06,
+      "loss": 0.5923,
+      "step": 626
+    },
+    {
+      "epoch": 0.2887654356523992,
+      "grad_norm": 2.9733446787329227,
+      "learning_rate": 1.6695002239901569e-06,
+      "loss": 0.6514,
+      "step": 627
+    },
+    {
+      "epoch": 0.28922598658645404,
+      "grad_norm": 3.272788188514796,
+      "learning_rate": 1.6683908743836711e-06,
+      "loss": 0.599,
+      "step": 628
+    },
+    {
+      "epoch": 0.28968653752050894,
+      "grad_norm": 3.073673873179751,
+      "learning_rate": 1.6672800360127823e-06,
+      "loss": 0.5548,
+      "step": 629
+    },
+    {
+      "epoch": 0.2901470884545638,
+      "grad_norm": 3.1359157628557353,
+      "learning_rate": 1.6661677113517553e-06,
+      "loss": 0.6053,
+      "step": 630
+    },
+    {
+      "epoch": 0.2906076393886186,
+      "grad_norm": 3.2988300051982056,
+      "learning_rate": 1.6650539028781667e-06,
+      "loss": 0.7002,
+      "step": 631
+    },
+    {
+      "epoch": 0.2910681903226735,
+      "grad_norm": 3.025853139733737,
+      "learning_rate": 1.663938613072898e-06,
+      "loss": 0.5597,
+      "step": 632
+    },
+    {
+      "epoch": 0.29152874125672834,
+      "grad_norm": 2.797667473613527,
+      "learning_rate": 1.6628218444201299e-06,
+      "loss": 0.579,
+      "step": 633
+    },
+    {
+      "epoch": 0.29198929219078323,
+      "grad_norm": 2.482842591113322,
+      "learning_rate": 1.6617035994073372e-06,
+      "loss": 0.4768,
+      "step": 634
+    },
+    {
+      "epoch": 0.29244984312483807,
+      "grad_norm": 3.2875064889505423,
+      "learning_rate": 1.6605838805252828e-06,
+      "loss": 0.6799,
+      "step": 635
+    },
+    {
+      "epoch": 0.29291039405889296,
+      "grad_norm": 3.160691937520528,
+      "learning_rate": 1.6594626902680126e-06,
+      "loss": 0.4986,
+      "step": 636
+    },
+    {
+      "epoch": 0.2933709449929478,
+      "grad_norm": 3.2855289808671384,
+      "learning_rate": 1.6583400311328505e-06,
+      "loss": 0.5377,
+      "step": 637
+    },
+    {
+      "epoch": 0.2938314959270027,
+      "grad_norm": 2.9638072454998725,
+      "learning_rate": 1.6572159056203915e-06,
+      "loss": 0.6512,
+      "step": 638
+    },
+    {
+      "epoch": 0.29429204686105753,
+      "grad_norm": 3.371604241327082,
+      "learning_rate": 1.6560903162344966e-06,
+      "loss": 0.5783,
+      "step": 639
+    },
+    {
+      "epoch": 0.2947525977951124,
+      "grad_norm": 3.628939173234319,
+      "learning_rate": 1.6549632654822875e-06,
+      "loss": 0.6427,
+      "step": 640
+    },
+    {
+      "epoch": 0.29521314872916726,
+      "grad_norm": 2.945921755545379,
+      "learning_rate": 1.6538347558741422e-06,
+      "loss": 0.4827,
+      "step": 641
+    },
+    {
+      "epoch": 0.29567369966322216,
+      "grad_norm": 2.9596989147777513,
+      "learning_rate": 1.652704789923686e-06,
+      "loss": 0.7239,
+      "step": 642
+    },
+    {
+      "epoch": 0.296134250597277,
+      "grad_norm": 3.597968564106779,
+      "learning_rate": 1.6515733701477896e-06,
+      "loss": 0.666,
+      "step": 643
+    },
+    {
+      "epoch": 0.29659480153133183,
+      "grad_norm": 3.1426064389176087,
+      "learning_rate": 1.6504404990665615e-06,
+      "loss": 0.5744,
+      "step": 644
+    },
+    {
+      "epoch": 0.2970553524653867,
+      "grad_norm": 3.120168720987635,
+      "learning_rate": 1.6493061792033424e-06,
+      "loss": 0.5737,
+      "step": 645
+    },
+    {
+      "epoch": 0.29751590339944156,
+      "grad_norm": 2.7520653399387904,
+      "learning_rate": 1.648170413084701e-06,
+      "loss": 0.6586,
+      "step": 646
+    },
+    {
+      "epoch": 0.29797645433349645,
+      "grad_norm": 2.7598009601994575,
+      "learning_rate": 1.6470332032404258e-06,
+      "loss": 0.5887,
+      "step": 647
+    },
+    {
+      "epoch": 0.2984370052675513,
+      "grad_norm": 2.9219392147630456,
+      "learning_rate": 1.6458945522035227e-06,
+      "loss": 0.4776,
+      "step": 648
+    },
+    {
+      "epoch": 0.2988975562016062,
+      "grad_norm": 2.918413331344107,
+      "learning_rate": 1.6447544625102068e-06,
+      "loss": 0.5586,
+      "step": 649
+    },
+    {
+      "epoch": 0.299358107135661,
+      "grad_norm": 3.0539033522988275,
+      "learning_rate": 1.6436129366998973e-06,
+      "loss": 0.6355,
+      "step": 650
+    },
+    {
+      "epoch": 0.2998186580697159,
+      "grad_norm": 3.0936916830976475,
+      "learning_rate": 1.6424699773152138e-06,
+      "loss": 0.6304,
+      "step": 651
+    },
+    {
+      "epoch": 0.30027920900377075,
+      "grad_norm": 3.3790034098133384,
+      "learning_rate": 1.6413255869019666e-06,
+      "loss": 0.6053,
+      "step": 652
+    },
+    {
+      "epoch": 0.30073975993782565,
+      "grad_norm": 3.244406880170614,
+      "learning_rate": 1.640179768009156e-06,
+      "loss": 0.5883,
+      "step": 653
+    },
+    {
+      "epoch": 0.3012003108718805,
+      "grad_norm": 3.145131450305741,
+      "learning_rate": 1.6390325231889616e-06,
+      "loss": 0.4706,
+      "step": 654
+    },
+    {
+      "epoch": 0.3016608618059354,
+      "grad_norm": 2.7110601957589404,
+      "learning_rate": 1.6378838549967415e-06,
+      "loss": 0.6071,
+      "step": 655
+    },
+    {
+      "epoch": 0.3021214127399902,
+      "grad_norm": 3.343566724286675,
+      "learning_rate": 1.6367337659910221e-06,
+      "loss": 0.6244,
+      "step": 656
+    },
+    {
+      "epoch": 0.30258196367404505,
+      "grad_norm": 3.209028090230382,
+      "learning_rate": 1.6355822587334959e-06,
+      "loss": 0.543,
+      "step": 657
+    },
+    {
+      "epoch": 0.30304251460809994,
+      "grad_norm": 3.5185896956330422,
+      "learning_rate": 1.6344293357890137e-06,
+      "loss": 0.5821,
+      "step": 658
+    },
+    {
+      "epoch": 0.3035030655421548,
+      "grad_norm": 2.8601115189735604,
+      "learning_rate": 1.6332749997255804e-06,
+      "loss": 0.6433,
+      "step": 659
+    },
+    {
+      "epoch": 0.3039636164762097,
+      "grad_norm": 3.476073663868734,
+      "learning_rate": 1.632119253114347e-06,
+      "loss": 0.4601,
+      "step": 660
+    },
+    {
+      "epoch": 0.3044241674102645,
+      "grad_norm": 3.2412899564729005,
+      "learning_rate": 1.6309620985296072e-06,
+      "loss": 0.7235,
+      "step": 661
+    },
+    {
+      "epoch": 0.3048847183443194,
+      "grad_norm": 2.713918449445168,
+      "learning_rate": 1.6298035385487918e-06,
+      "loss": 0.6167,
+      "step": 662
+    },
+    {
+      "epoch": 0.30534526927837424,
+      "grad_norm": 3.4033868200489623,
+      "learning_rate": 1.6286435757524602e-06,
+      "loss": 0.5718,
+      "step": 663
+    },
+    {
+      "epoch": 0.30580582021242914,
+      "grad_norm": 3.0202609458048038,
+      "learning_rate": 1.6274822127242974e-06,
+      "loss": 0.4536,
+      "step": 664
+    },
+    {
+      "epoch": 0.306266371146484,
+      "grad_norm": 3.001878947353794,
+      "learning_rate": 1.6263194520511064e-06,
+      "loss": 0.57,
+      "step": 665
+    },
+    {
+      "epoch": 0.30672692208053887,
+      "grad_norm": 2.998809917811107,
+      "learning_rate": 1.6251552963228048e-06,
+      "loss": 0.5362,
+      "step": 666
+    },
+    {
+      "epoch": 0.3071874730145937,
+      "grad_norm": 2.977451183975403,
+      "learning_rate": 1.6239897481324164e-06,
+      "loss": 0.4483,
+      "step": 667
+    },
+    {
+      "epoch": 0.3076480239486486,
+      "grad_norm": 3.0423908095090733,
+      "learning_rate": 1.6228228100760664e-06,
+      "loss": 0.5557,
+      "step": 668
+    },
+    {
+      "epoch": 0.30810857488270343,
+      "grad_norm": 3.043152280285196,
+      "learning_rate": 1.6216544847529764e-06,
+      "loss": 0.6267,
+      "step": 669
+    },
+    {
+      "epoch": 0.30856912581675827,
+      "grad_norm": 2.742916110544047,
+      "learning_rate": 1.620484774765458e-06,
+      "loss": 0.5141,
+      "step": 670
+    },
+    {
+      "epoch": 0.30902967675081316,
+      "grad_norm": 3.535942753468968,
+      "learning_rate": 1.6193136827189065e-06,
+      "loss": 0.7185,
+      "step": 671
+    },
+    {
+      "epoch": 0.309490227684868,
+      "grad_norm": 3.4294982638238163,
+      "learning_rate": 1.6181412112217957e-06,
+      "loss": 0.6178,
+      "step": 672
+    },
+    {
+      "epoch": 0.3099507786189229,
+      "grad_norm": 3.3208662012013104,
+      "learning_rate": 1.6169673628856722e-06,
+      "loss": 0.5777,
+      "step": 673
+    },
+    {
+      "epoch": 0.31041132955297773,
+      "grad_norm": 3.6764674298301214,
+      "learning_rate": 1.6157921403251492e-06,
+      "loss": 0.5752,
+      "step": 674
+    },
+    {
+      "epoch": 0.3108718804870326,
+      "grad_norm": 3.252440774762675,
+      "learning_rate": 1.6146155461579007e-06,
+      "loss": 0.7095,
+      "step": 675
+    },
+    {
+      "epoch": 0.31133243142108746,
+      "grad_norm": 3.2165238232059084,
+      "learning_rate": 1.6134375830046563e-06,
+      "loss": 0.6169,
+      "step": 676
+    },
+    {
+      "epoch": 0.31179298235514236,
+      "grad_norm": 3.1854273045665287,
+      "learning_rate": 1.6122582534891942e-06,
+      "loss": 0.6611,
+      "step": 677
+    },
+    {
+      "epoch": 0.3122535332891972,
+      "grad_norm": 3.4248301426471297,
+      "learning_rate": 1.6110775602383365e-06,
+      "loss": 0.6427,
+      "step": 678
+    },
+    {
+      "epoch": 0.3127140842232521,
+      "grad_norm": 3.113032011879882,
+      "learning_rate": 1.6098955058819423e-06,
+      "loss": 0.5213,
+      "step": 679
+    },
+    {
+      "epoch": 0.3131746351573069,
+      "grad_norm": 3.3490627264511175,
+      "learning_rate": 1.6087120930529036e-06,
+      "loss": 0.5633,
+      "step": 680
+    },
+    {
+      "epoch": 0.3136351860913618,
+      "grad_norm": 3.2500116220033375,
+      "learning_rate": 1.6075273243871367e-06,
+      "loss": 0.5387,
+      "step": 681
+    },
+    {
+      "epoch": 0.31409573702541665,
+      "grad_norm": 3.6728777135144837,
+      "learning_rate": 1.606341202523579e-06,
+      "loss": 0.5113,
+      "step": 682
+    },
+    {
+      "epoch": 0.3145562879594715,
+      "grad_norm": 2.683705848328992,
+      "learning_rate": 1.6051537301041812e-06,
+      "loss": 0.5355,
+      "step": 683
+    },
+    {
+      "epoch": 0.3150168388935264,
+      "grad_norm": 3.4297858809617363,
+      "learning_rate": 1.6039649097739032e-06,
+      "loss": 0.6449,
+      "step": 684
+    },
+    {
+      "epoch": 0.3154773898275812,
+      "grad_norm": 3.292570386362054,
+      "learning_rate": 1.602774744180706e-06,
+      "loss": 0.6552,
+      "step": 685
+    },
+    {
+      "epoch": 0.3159379407616361,
+      "grad_norm": 3.123629844208179,
+      "learning_rate": 1.601583235975548e-06,
+      "loss": 0.4843,
+      "step": 686
+    },
+    {
+      "epoch": 0.31639849169569095,
+      "grad_norm": 3.1402599959393473,
+      "learning_rate": 1.6003903878123782e-06,
+      "loss": 0.5134,
+      "step": 687
+    },
+    {
+      "epoch": 0.31685904262974585,
+      "grad_norm": 3.3270904677521873,
+      "learning_rate": 1.599196202348129e-06,
+      "loss": 0.5501,
+      "step": 688
+    },
+    {
+      "epoch": 0.3173195935638007,
+      "grad_norm": 3.5015088481238577,
+      "learning_rate": 1.5980006822427123e-06,
+      "loss": 0.5634,
+      "step": 689
+    },
+    {
+      "epoch": 0.3177801444978556,
+      "grad_norm": 3.08180457972679,
+      "learning_rate": 1.5968038301590133e-06,
+      "loss": 0.6374,
+      "step": 690
+    },
+    {
+      "epoch": 0.3182406954319104,
+      "grad_norm": 3.2445020121060715,
+      "learning_rate": 1.5956056487628827e-06,
+      "loss": 0.6407,
+      "step": 691
+    },
+    {
+      "epoch": 0.3187012463659653,
+      "grad_norm": 3.233197428352453,
+      "learning_rate": 1.5944061407231338e-06,
+      "loss": 0.706,
+      "step": 692
+    },
+    {
+      "epoch": 0.31916179730002014,
+      "grad_norm": 3.280658309577868,
+      "learning_rate": 1.593205308711533e-06,
+      "loss": 0.5266,
+      "step": 693
+    },
+    {
+      "epoch": 0.31962234823407504,
+      "grad_norm": 3.5140387469250736,
+      "learning_rate": 1.5920031554027969e-06,
+      "loss": 0.5574,
+      "step": 694
+    },
+    {
+      "epoch": 0.3200828991681299,
+      "grad_norm": 3.42036786677391,
+      "learning_rate": 1.590799683474585e-06,
+      "loss": 0.6599,
+      "step": 695
+    },
+    {
+      "epoch": 0.3205434501021847,
+      "grad_norm": 3.1846836901173075,
+      "learning_rate": 1.5895948956074933e-06,
+      "loss": 0.668,
+      "step": 696
+    },
+    {
+      "epoch": 0.3210040010362396,
+      "grad_norm": 3.3566033229272967,
+      "learning_rate": 1.5883887944850495e-06,
+      "loss": 0.6669,
+      "step": 697
+    },
+    {
+      "epoch": 0.32146455197029444,
+      "grad_norm": 3.2045770323594382,
+      "learning_rate": 1.5871813827937063e-06,
+      "loss": 0.5538,
+      "step": 698
+    },
+    {
+      "epoch": 0.32192510290434934,
+      "grad_norm": 2.9902993069464565,
+      "learning_rate": 1.5859726632228357e-06,
+      "loss": 0.5938,
+      "step": 699
+    },
+    {
+      "epoch": 0.3223856538384042,
+      "grad_norm": 3.0057175558399485,
+      "learning_rate": 1.5847626384647221e-06,
+      "loss": 0.5181,
+      "step": 700
+    },
+    {
+      "epoch": 0.32284620477245907,
+      "grad_norm": 3.5222478856728254,
+      "learning_rate": 1.583551311214558e-06,
+      "loss": 0.5791,
+      "step": 701
+    },
+    {
+      "epoch": 0.3233067557065139,
+      "grad_norm": 3.998127982109179,
+      "learning_rate": 1.5823386841704362e-06,
+      "loss": 0.4773,
+      "step": 702
+    },
+    {
+      "epoch": 0.3237673066405688,
+      "grad_norm": 2.995949094547311,
+      "learning_rate": 1.5811247600333456e-06,
+      "loss": 0.5852,
+      "step": 703
+    },
+    {
+      "epoch": 0.32422785757462363,
+      "grad_norm": 2.9705371769732425,
+      "learning_rate": 1.5799095415071628e-06,
+      "loss": 0.5957,
+      "step": 704
+    },
+    {
+      "epoch": 0.3246884085086785,
+      "grad_norm": 2.6316896525372537,
+      "learning_rate": 1.5786930312986495e-06,
+      "loss": 0.4241,
+      "step": 705
+    },
+    {
+      "epoch": 0.32514895944273337,
+      "grad_norm": 3.200708018638179,
+      "learning_rate": 1.5774752321174427e-06,
+      "loss": 0.5434,
+      "step": 706
+    },
+    {
+      "epoch": 0.32560951037678826,
+      "grad_norm": 3.452466534810513,
+      "learning_rate": 1.576256146676051e-06,
+      "loss": 0.5766,
+      "step": 707
+    },
+    {
+      "epoch": 0.3260700613108431,
+      "grad_norm": 2.8456835798060816,
+      "learning_rate": 1.575035777689849e-06,
+      "loss": 0.4522,
+      "step": 708
+    },
+    {
+      "epoch": 0.32653061224489793,
+      "grad_norm": 3.593887134081324,
+      "learning_rate": 1.5738141278770685e-06,
+      "loss": 0.628,
+      "step": 709
+    },
+    {
+      "epoch": 0.3269911631789528,
+      "grad_norm": 3.217566834694963,
+      "learning_rate": 1.5725911999587953e-06,
+      "loss": 0.594,
+      "step": 710
+    },
+    {
+      "epoch": 0.32745171411300766,
+      "grad_norm": 3.0506131303862234,
+      "learning_rate": 1.5713669966589618e-06,
+      "loss": 0.6114,
+      "step": 711
+    },
+    {
+      "epoch": 0.32791226504706256,
+      "grad_norm": 3.2672150522195254,
+      "learning_rate": 1.5701415207043414e-06,
+      "loss": 0.5329,
+      "step": 712
+    },
+    {
+      "epoch": 0.3283728159811174,
+      "grad_norm": 3.0741601299943957,
+      "learning_rate": 1.5689147748245412e-06,
+      "loss": 0.5823,
+      "step": 713
+    },
+    {
+      "epoch": 0.3288333669151723,
+      "grad_norm": 2.7741610627962974,
+      "learning_rate": 1.5676867617519986e-06,
+      "loss": 0.5868,
+      "step": 714
+    },
+    {
+      "epoch": 0.3292939178492271,
+      "grad_norm": 3.3153530179766766,
+      "learning_rate": 1.566457484221972e-06,
+      "loss": 0.5851,
+      "step": 715
+    },
+    {
+      "epoch": 0.329754468783282,
+      "grad_norm": 3.317026645580635,
+      "learning_rate": 1.5652269449725374e-06,
+      "loss": 0.6181,
+      "step": 716
+    },
+    {
+      "epoch": 0.33021501971733686,
+      "grad_norm": 3.404975554795291,
+      "learning_rate": 1.5639951467445798e-06,
+      "loss": 0.6734,
+      "step": 717
+    },
+    {
+      "epoch": 0.33067557065139175,
+      "grad_norm": 2.8122628566375933,
+      "learning_rate": 1.5627620922817895e-06,
+      "loss": 0.5022,
+      "step": 718
+    },
+    {
+      "epoch": 0.3311361215854466,
+      "grad_norm": 2.8689456040300954,
+      "learning_rate": 1.561527784330655e-06,
+      "loss": 0.4525,
+      "step": 719
+    },
+    {
+      "epoch": 0.3315966725195015,
+      "grad_norm": 3.1909349124276734,
+      "learning_rate": 1.5602922256404556e-06,
+      "loss": 0.6127,
+      "step": 720
+    },
+    {
+      "epoch": 0.3320572234535563,
+      "grad_norm": 3.0736929753426776,
+      "learning_rate": 1.559055418963258e-06,
+      "loss": 0.5143,
+      "step": 721
+    },
+    {
+      "epoch": 0.33251777438761115,
+      "grad_norm": 2.9374948837179518,
+      "learning_rate": 1.557817367053908e-06,
+      "loss": 0.4802,
+      "step": 722
+    },
+    {
+      "epoch": 0.33297832532166605,
+      "grad_norm": 2.8350541529140316,
+      "learning_rate": 1.5565780726700244e-06,
+      "loss": 0.4546,
+      "step": 723
+    },
+    {
+      "epoch": 0.3334388762557209,
+      "grad_norm": 2.9599261081932027,
+      "learning_rate": 1.5553375385719943e-06,
+      "loss": 0.5198,
+      "step": 724
+    },
+    {
+      "epoch": 0.3338994271897758,
+      "grad_norm": 3.2317683594863715,
+      "learning_rate": 1.5540957675229663e-06,
+      "loss": 0.5195,
+      "step": 725
+    },
+    {
+      "epoch": 0.3343599781238306,
+      "grad_norm": 3.296064700074342,
+      "learning_rate": 1.5528527622888428e-06,
+      "loss": 0.6044,
+      "step": 726
+    },
+    {
+      "epoch": 0.3348205290578855,
+      "grad_norm": 2.863524627128321,
+      "learning_rate": 1.5516085256382764e-06,
+      "loss": 0.6144,
+      "step": 727
+    },
+    {
+      "epoch": 0.33528107999194035,
+      "grad_norm": 3.0821132303925594,
+      "learning_rate": 1.5503630603426628e-06,
+      "loss": 0.576,
+      "step": 728
+    },
+    {
+      "epoch": 0.33574163092599524,
+      "grad_norm": 2.9688429474981026,
+      "learning_rate": 1.5491163691761334e-06,
+      "loss": 0.5294,
+      "step": 729
+    },
+    {
+      "epoch": 0.3362021818600501,
+      "grad_norm": 3.268176435044111,
+      "learning_rate": 1.5478684549155507e-06,
+      "loss": 0.523,
+      "step": 730
+    },
+    {
+      "epoch": 0.33666273279410497,
+      "grad_norm": 3.3700516164655374,
+      "learning_rate": 1.5466193203405015e-06,
+      "loss": 0.6837,
+      "step": 731
+    },
+    {
+      "epoch": 0.3371232837281598,
+      "grad_norm": 3.2319186451474646,
+      "learning_rate": 1.5453689682332898e-06,
+      "loss": 0.6385,
+      "step": 732
+    },
+    {
+      "epoch": 0.3375838346622147,
+      "grad_norm": 3.4936427230146987,
+      "learning_rate": 1.5441174013789326e-06,
+      "loss": 0.6258,
+      "step": 733
+    },
+    {
+      "epoch": 0.33804438559626954,
+      "grad_norm": 3.305223919072284,
+      "learning_rate": 1.5428646225651525e-06,
+      "loss": 0.5428,
+      "step": 734
+    },
+    {
+      "epoch": 0.3385049365303244,
+      "grad_norm": 2.9963569062928777,
+      "learning_rate": 1.5416106345823714e-06,
+      "loss": 0.5679,
+      "step": 735
+    },
+    {
+      "epoch": 0.33896548746437927,
+      "grad_norm": 3.3209457502261026,
+      "learning_rate": 1.540355440223704e-06,
+      "loss": 0.6125,
+      "step": 736
+    },
+    {
+      "epoch": 0.3394260383984341,
+      "grad_norm": 3.304460716728442,
+      "learning_rate": 1.5390990422849528e-06,
+      "loss": 0.5128,
+      "step": 737
+    },
+    {
+      "epoch": 0.339886589332489,
+      "grad_norm": 3.177691233283076,
+      "learning_rate": 1.5378414435646008e-06,
+      "loss": 0.6157,
+      "step": 738
+    },
+    {
+      "epoch": 0.34034714026654384,
+      "grad_norm": 3.2484901797414687,
+      "learning_rate": 1.5365826468638057e-06,
+      "loss": 0.4804,
+      "step": 739
+    },
+    {
+      "epoch": 0.34080769120059873,
+      "grad_norm": 2.97003236725624,
+      "learning_rate": 1.5353226549863933e-06,
+      "loss": 0.5174,
+      "step": 740
+    },
+    {
+      "epoch": 0.34126824213465357,
+      "grad_norm": 3.0081534150970985,
+      "learning_rate": 1.5340614707388516e-06,
+      "loss": 0.5972,
+      "step": 741
+    },
+    {
+      "epoch": 0.34172879306870846,
+      "grad_norm": 3.0537773804963044,
+      "learning_rate": 1.5327990969303256e-06,
+      "loss": 0.6268,
+      "step": 742
+    },
+    {
+      "epoch": 0.3421893440027633,
+      "grad_norm": 2.4616414555847537,
+      "learning_rate": 1.531535536372608e-06,
+      "loss": 0.5428,
+      "step": 743
+    },
+    {
+      "epoch": 0.3426498949368182,
+      "grad_norm": 3.459066350919476,
+      "learning_rate": 1.5302707918801354e-06,
+      "loss": 0.6306,
+      "step": 744
+    },
+    {
+      "epoch": 0.343110445870873,
+      "grad_norm": 3.352692171370741,
+      "learning_rate": 1.5290048662699828e-06,
+      "loss": 0.5567,
+      "step": 745
+    },
+    {
+      "epoch": 0.3435709968049279,
+      "grad_norm": 3.444376949244577,
+      "learning_rate": 1.5277377623618546e-06,
+      "loss": 0.587,
+      "step": 746
+    },
+    {
+      "epoch": 0.34403154773898276,
+      "grad_norm": 3.018488954664052,
+      "learning_rate": 1.5264694829780801e-06,
+      "loss": 0.6234,
+      "step": 747
+    },
+    {
+      "epoch": 0.3444920986730376,
+      "grad_norm": 4.059892549594848,
+      "learning_rate": 1.5252000309436071e-06,
+      "loss": 0.5284,
+      "step": 748
+    },
+    {
+      "epoch": 0.3449526496070925,
+      "grad_norm": 3.1529595103252848,
+      "learning_rate": 1.5239294090859948e-06,
+      "loss": 0.5514,
+      "step": 749
+    },
+    {
+      "epoch": 0.3454132005411473,
+      "grad_norm": 3.2294187071291938,
+      "learning_rate": 1.522657620235409e-06,
+      "loss": 0.5978,
+      "step": 750
+    },
+    {
+      "epoch": 0.3458737514752022,
+      "grad_norm": 3.1490886609776685,
+      "learning_rate": 1.5213846672246138e-06,
+      "loss": 0.596,
+      "step": 751
+    },
+    {
+      "epoch": 0.34633430240925706,
+      "grad_norm": 3.1168755025387362,
+      "learning_rate": 1.5201105528889666e-06,
+      "loss": 0.5168,
+      "step": 752
+    },
+    {
+      "epoch": 0.34679485334331195,
+      "grad_norm": 3.054962204132803,
+      "learning_rate": 1.5188352800664119e-06,
+      "loss": 0.5459,
+      "step": 753
+    },
+    {
+      "epoch": 0.3472554042773668,
+      "grad_norm": 3.575951938973164,
+      "learning_rate": 1.5175588515974748e-06,
+      "loss": 0.6461,
+      "step": 754
+    },
+    {
+      "epoch": 0.3477159552114217,
+      "grad_norm": 2.8785648294557085,
+      "learning_rate": 1.5162812703252537e-06,
+      "loss": 0.4715,
+      "step": 755
+    },
+    {
+      "epoch": 0.3481765061454765,
+      "grad_norm": 3.2534706235609585,
+      "learning_rate": 1.5150025390954152e-06,
+      "loss": 0.5632,
+      "step": 756
+    },
+    {
+      "epoch": 0.3486370570795314,
+      "grad_norm": 3.0206088523549974,
+      "learning_rate": 1.513722660756187e-06,
+      "loss": 0.4888,
+      "step": 757
+    },
+    {
+      "epoch": 0.34909760801358625,
+      "grad_norm": 2.7017314694335086,
+      "learning_rate": 1.5124416381583517e-06,
+      "loss": 0.4697,
+      "step": 758
+    },
+    {
+      "epoch": 0.34955815894764114,
+      "grad_norm": 3.3514313878368625,
+      "learning_rate": 1.5111594741552423e-06,
+      "loss": 0.5044,
+      "step": 759
+    },
+    {
+      "epoch": 0.350018709881696,
+      "grad_norm": 3.338106474247453,
+      "learning_rate": 1.5098761716027315e-06,
+      "loss": 0.5476,
+      "step": 760
+    },
+    {
+      "epoch": 0.3504792608157508,
+      "grad_norm": 3.3207826162715666,
+      "learning_rate": 1.5085917333592297e-06,
+      "loss": 0.6685,
+      "step": 761
+    },
+    {
+      "epoch": 0.3509398117498057,
+      "grad_norm": 2.882204620095301,
+      "learning_rate": 1.5073061622856765e-06,
+      "loss": 0.4214,
+      "step": 762
+    },
+    {
+      "epoch": 0.35140036268386055,
+      "grad_norm": 3.2702426268023856,
+      "learning_rate": 1.506019461245535e-06,
+      "loss": 0.4785,
+      "step": 763
+    },
+    {
+      "epoch": 0.35186091361791544,
+      "grad_norm": 3.393774716059176,
+      "learning_rate": 1.5047316331047846e-06,
+      "loss": 0.5249,
+      "step": 764
+    },
+    {
+      "epoch": 0.3523214645519703,
+      "grad_norm": 2.626311070223108,
+      "learning_rate": 1.5034426807319162e-06,
+      "loss": 0.5107,
+      "step": 765
+    },
+    {
+      "epoch": 0.35278201548602517,
+      "grad_norm": 3.0508858155806586,
+      "learning_rate": 1.5021526069979232e-06,
+      "loss": 0.6297,
+      "step": 766
+    },
+    {
+      "epoch": 0.35324256642008,
+      "grad_norm": 3.2942945723575336,
+      "learning_rate": 1.5008614147762982e-06,
+      "loss": 0.6742,
+      "step": 767
+    },
+    {
+      "epoch": 0.3537031173541349,
+      "grad_norm": 3.375703871690601,
+      "learning_rate": 1.4995691069430244e-06,
+      "loss": 0.6203,
+      "step": 768
+    },
+    {
+      "epoch": 0.35416366828818974,
+      "grad_norm": 3.11233555593855,
+      "learning_rate": 1.49827568637657e-06,
+      "loss": 0.5992,
+      "step": 769
+    },
+    {
+      "epoch": 0.35462421922224463,
+      "grad_norm": 2.8563952523571965,
+      "learning_rate": 1.4969811559578818e-06,
+      "loss": 0.4969,
+      "step": 770
+    },
+    {
+      "epoch": 0.35508477015629947,
+      "grad_norm": 3.2356844981316235,
+      "learning_rate": 1.4956855185703786e-06,
+      "loss": 0.4214,
+      "step": 771
+    },
+    {
+      "epoch": 0.35554532109035436,
+      "grad_norm": 2.8462493199493024,
+      "learning_rate": 1.4943887770999447e-06,
+      "loss": 0.5523,
+      "step": 772
+    },
+    {
+      "epoch": 0.3560058720244092,
+      "grad_norm": 3.5665113209504105,
+      "learning_rate": 1.493090934434924e-06,
+      "loss": 0.5731,
+      "step": 773
+    },
+    {
+      "epoch": 0.35646642295846404,
+      "grad_norm": 3.169353405800153,
+      "learning_rate": 1.4917919934661128e-06,
+      "loss": 0.5358,
+      "step": 774
+    },
+    {
+      "epoch": 0.35692697389251893,
+      "grad_norm": 3.121133974442328,
+      "learning_rate": 1.4904919570867539e-06,
+      "loss": 0.5471,
+      "step": 775
+    },
+    {
+      "epoch": 0.35738752482657377,
+      "grad_norm": 3.0928538679959243,
+      "learning_rate": 1.4891908281925298e-06,
+      "loss": 0.6502,
+      "step": 776
+    },
+    {
+      "epoch": 0.35784807576062866,
+      "grad_norm": 3.2886205307157823,
+      "learning_rate": 1.4878886096815569e-06,
+      "loss": 0.5099,
+      "step": 777
+    },
+    {
+      "epoch": 0.3583086266946835,
+      "grad_norm": 3.041026939521522,
+      "learning_rate": 1.486585304454378e-06,
+      "loss": 0.5957,
+      "step": 778
+    },
+    {
+      "epoch": 0.3587691776287384,
+      "grad_norm": 2.8217805539112732,
+      "learning_rate": 1.4852809154139576e-06,
+      "loss": 0.5313,
+      "step": 779
+    },
+    {
+      "epoch": 0.3592297285627932,
+      "grad_norm": 2.6440360065601145,
+      "learning_rate": 1.4839754454656723e-06,
+      "loss": 0.5048,
+      "step": 780
+    },
+    {
+      "epoch": 0.3596902794968481,
+      "grad_norm": 3.4469745235083393,
+      "learning_rate": 1.4826688975173084e-06,
+      "loss": 0.6932,
+      "step": 781
+    },
+    {
+      "epoch": 0.36015083043090296,
+      "grad_norm": 4.096492089448597,
+      "learning_rate": 1.481361274479052e-06,
+      "loss": 0.5383,
+      "step": 782
+    },
+    {
+      "epoch": 0.36061138136495785,
+      "grad_norm": 2.9081019095353606,
+      "learning_rate": 1.4800525792634838e-06,
+      "loss": 0.5196,
+      "step": 783
+    },
+    {
+      "epoch": 0.3610719322990127,
+      "grad_norm": 3.0363832332393073,
+      "learning_rate": 1.4787428147855737e-06,
+      "loss": 0.5606,
+      "step": 784
+    },
+    {
+      "epoch": 0.3615324832330676,
+      "grad_norm": 3.404973081364874,
+      "learning_rate": 1.4774319839626725e-06,
+      "loss": 0.6067,
+      "step": 785
+    },
+    {
+      "epoch": 0.3619930341671224,
+      "grad_norm": 2.9657594105232916,
+      "learning_rate": 1.476120089714506e-06,
+      "loss": 0.5427,
+      "step": 786
+    },
+    {
+      "epoch": 0.36245358510117726,
+      "grad_norm": 3.6192624322683256,
+      "learning_rate": 1.4748071349631693e-06,
+      "loss": 0.6706,
+      "step": 787
+    },
+    {
+      "epoch": 0.36291413603523215,
+      "grad_norm": 3.526460171368003,
+      "learning_rate": 1.4734931226331188e-06,
+      "loss": 0.5103,
+      "step": 788
+    },
+    {
+      "epoch": 0.363374686969287,
+      "grad_norm": 2.8430807766055577,
+      "learning_rate": 1.4721780556511674e-06,
+      "loss": 0.4935,
+      "step": 789
+    },
+    {
+      "epoch": 0.3638352379033419,
+      "grad_norm": 3.264151914199905,
+      "learning_rate": 1.4708619369464765e-06,
+      "loss": 0.6703,
+      "step": 790
+    },
+    {
+      "epoch": 0.3642957888373967,
+      "grad_norm": 3.5948426136873843,
+      "learning_rate": 1.469544769450551e-06,
+      "loss": 0.5244,
+      "step": 791
+    },
+    {
+      "epoch": 0.3647563397714516,
+      "grad_norm": 3.2156355781796018,
+      "learning_rate": 1.46822655609723e-06,
+      "loss": 0.5689,
+      "step": 792
+    },
+    {
+      "epoch": 0.36521689070550645,
+      "grad_norm": 3.4654337540005002,
+      "learning_rate": 1.4669072998226843e-06,
+      "loss": 0.5605,
+      "step": 793
+    },
+    {
+      "epoch": 0.36567744163956134,
+      "grad_norm": 3.0725623075599184,
+      "learning_rate": 1.4655870035654065e-06,
+      "loss": 0.6267,
+      "step": 794
+    },
+    {
+      "epoch": 0.3661379925736162,
+      "grad_norm": 3.029931817827877,
+      "learning_rate": 1.4642656702662058e-06,
+      "loss": 0.6285,
+      "step": 795
+    },
+    {
+      "epoch": 0.36659854350767107,
+      "grad_norm": 2.9315413167488558,
+      "learning_rate": 1.4629433028682013e-06,
+      "loss": 0.5949,
+      "step": 796
+    },
+    {
+      "epoch": 0.3670590944417259,
+      "grad_norm": 3.151885439468786,
+      "learning_rate": 1.4616199043168154e-06,
+      "loss": 0.6758,
+      "step": 797
+    },
+    {
+      "epoch": 0.3675196453757808,
+      "grad_norm": 3.0346115082411655,
+      "learning_rate": 1.4602954775597673e-06,
+      "loss": 0.5696,
+      "step": 798
+    },
+    {
+      "epoch": 0.36798019630983564,
+      "grad_norm": 3.2157712771710827,
+      "learning_rate": 1.458970025547067e-06,
+      "loss": 0.6722,
+      "step": 799
+    },
+    {
+      "epoch": 0.36844074724389053,
+      "grad_norm": 3.2513066677812748,
+      "learning_rate": 1.457643551231007e-06,
+      "loss": 0.6715,
+      "step": 800
+    },
+    {
+      "epoch": 0.36890129817794537,
+      "grad_norm": 3.5891769122485653,
+      "learning_rate": 1.456316057566158e-06,
+      "loss": 0.6068,
+      "step": 801
+    },
+    {
+      "epoch": 0.3693618491120002,
+      "grad_norm": 2.9833048436720655,
+      "learning_rate": 1.45498754750936e-06,
+      "loss": 0.5265,
+      "step": 802
+    },
+    {
+      "epoch": 0.3698224000460551,
+      "grad_norm": 3.0143882511843754,
+      "learning_rate": 1.453658024019718e-06,
+      "loss": 0.6277,
+      "step": 803
+    },
+    {
+      "epoch": 0.37028295098010994,
+      "grad_norm": 3.303125029561216,
+      "learning_rate": 1.4523274900585942e-06,
+      "loss": 0.5178,
+      "step": 804
+    },
+    {
+      "epoch": 0.37074350191416483,
+      "grad_norm": 2.96600210169327,
+      "learning_rate": 1.4509959485896004e-06,
+      "loss": 0.6883,
+      "step": 805
+    },
+    {
+      "epoch": 0.37120405284821967,
+      "grad_norm": 3.222533442523701,
+      "learning_rate": 1.4496634025785937e-06,
+      "loss": 0.5035,
+      "step": 806
+    },
+    {
+      "epoch": 0.37166460378227456,
+      "grad_norm": 3.3343786199048027,
+      "learning_rate": 1.4483298549936684e-06,
+      "loss": 0.5683,
+      "step": 807
+    },
+    {
+      "epoch": 0.3721251547163294,
+      "grad_norm": 3.194169003790268,
+      "learning_rate": 1.4469953088051497e-06,
+      "loss": 0.6087,
+      "step": 808
+    },
+    {
+      "epoch": 0.3725857056503843,
+      "grad_norm": 3.1623966174805083,
+      "learning_rate": 1.445659766985586e-06,
+      "loss": 0.5334,
+      "step": 809
+    },
+    {
+      "epoch": 0.37304625658443913,
+      "grad_norm": 2.9573679310798138,
+      "learning_rate": 1.4443232325097454e-06,
+      "loss": 0.4965,
+      "step": 810
+    },
+    {
+      "epoch": 0.373506807518494,
+      "grad_norm": 3.1354387245216766,
+      "learning_rate": 1.4429857083546053e-06,
+      "loss": 0.4966,
+      "step": 811
+    },
+    {
+      "epoch": 0.37396735845254886,
+      "grad_norm": 3.7718550261376045,
+      "learning_rate": 1.4416471974993487e-06,
+      "loss": 0.4793,
+      "step": 812
+    },
+    {
+      "epoch": 0.37442790938660375,
+      "grad_norm": 3.0917131083646194,
+      "learning_rate": 1.4403077029253553e-06,
+      "loss": 0.6511,
+      "step": 813
+    },
+    {
+      "epoch": 0.3748884603206586,
+      "grad_norm": 3.060066893764434,
+      "learning_rate": 1.4389672276161963e-06,
+      "loss": 0.4503,
+      "step": 814
+    },
+    {
+      "epoch": 0.3753490112547134,
+      "grad_norm": 3.0880275565306006,
+      "learning_rate": 1.4376257745576282e-06,
+      "loss": 0.5234,
+      "step": 815
+    },
+    {
+      "epoch": 0.3758095621887683,
+      "grad_norm": 3.3246972054582233,
+      "learning_rate": 1.4362833467375836e-06,
+      "loss": 0.5849,
+      "step": 816
+    },
+    {
+      "epoch": 0.37627011312282316,
+      "grad_norm": 3.5727318566464263,
+      "learning_rate": 1.4349399471461684e-06,
+      "loss": 0.6084,
+      "step": 817
+    },
+    {
+      "epoch": 0.37673066405687805,
+      "grad_norm": 3.688093227741573,
+      "learning_rate": 1.4335955787756513e-06,
+      "loss": 0.6864,
+      "step": 818
+    },
+    {
+      "epoch": 0.3771912149909329,
+      "grad_norm": 3.6380575782207702,
+      "learning_rate": 1.4322502446204592e-06,
+      "loss": 0.6678,
+      "step": 819
+    },
+    {
+      "epoch": 0.3776517659249878,
+      "grad_norm": 3.1559208653988633,
+      "learning_rate": 1.4309039476771706e-06,
+      "loss": 0.4839,
+      "step": 820
+    },
+    {
+      "epoch": 0.3781123168590426,
+      "grad_norm": 3.7163228600713745,
+      "learning_rate": 1.429556690944509e-06,
+      "loss": 0.6348,
+      "step": 821
+    },
+    {
+      "epoch": 0.3785728677930975,
+      "grad_norm": 2.9842451520879876,
+      "learning_rate": 1.4282084774233338e-06,
+      "loss": 0.4819,
+      "step": 822
+    },
+    {
+      "epoch": 0.37903341872715235,
+      "grad_norm": 3.069211559831523,
+      "learning_rate": 1.4268593101166378e-06,
+      "loss": 0.6175,
+      "step": 823
+    },
+    {
+      "epoch": 0.37949396966120724,
+      "grad_norm": 3.115770022996151,
+      "learning_rate": 1.4255091920295367e-06,
+      "loss": 0.6113,
+      "step": 824
+    },
+    {
+      "epoch": 0.3799545205952621,
+      "grad_norm": 3.2472503472001484,
+      "learning_rate": 1.4241581261692647e-06,
+      "loss": 0.6686,
+      "step": 825
+    },
+    {
+      "epoch": 0.380415071529317,
+      "grad_norm": 3.1742136266716314,
+      "learning_rate": 1.422806115545167e-06,
+      "loss": 0.6194,
+      "step": 826
+    },
+    {
+      "epoch": 0.3808756224633718,
+      "grad_norm": 2.7416191430113788,
+      "learning_rate": 1.4214531631686929e-06,
+      "loss": 0.575,
+      "step": 827
+    },
+    {
+      "epoch": 0.38133617339742665,
+      "grad_norm": 2.991715260469191,
+      "learning_rate": 1.4200992720533886e-06,
+      "loss": 0.6348,
+      "step": 828
+    },
+    {
+      "epoch": 0.38179672433148154,
+      "grad_norm": 3.3536018044430143,
+      "learning_rate": 1.4187444452148934e-06,
+      "loss": 0.5707,
+      "step": 829
+    },
+    {
+      "epoch": 0.3822572752655364,
+      "grad_norm": 2.9504178057835957,
+      "learning_rate": 1.4173886856709288e-06,
+      "loss": 0.5539,
+      "step": 830
+    },
+    {
+      "epoch": 0.38271782619959127,
+      "grad_norm": 3.4467632526859826,
+      "learning_rate": 1.416031996441294e-06,
+      "loss": 0.5473,
+      "step": 831
+    },
+    {
+      "epoch": 0.3831783771336461,
+      "grad_norm": 3.1022606765939877,
+      "learning_rate": 1.4146743805478605e-06,
+      "loss": 0.5181,
+      "step": 832
+    },
+    {
+      "epoch": 0.383638928067701,
+      "grad_norm": 2.8177854685524575,
+      "learning_rate": 1.413315841014562e-06,
+      "loss": 0.5749,
+      "step": 833
+    },
+    {
+      "epoch": 0.38409947900175584,
+      "grad_norm": 3.6263375137296614,
+      "learning_rate": 1.4119563808673905e-06,
+      "loss": 0.5913,
+      "step": 834
+    },
+    {
+      "epoch": 0.38456002993581073,
+      "grad_norm": 3.0015961938916806,
+      "learning_rate": 1.4105960031343889e-06,
+      "loss": 0.5724,
+      "step": 835
+    },
+    {
+      "epoch": 0.38502058086986557,
+      "grad_norm": 3.0678893744276357,
+      "learning_rate": 1.4092347108456424e-06,
+      "loss": 0.6605,
+      "step": 836
+    },
+    {
+      "epoch": 0.38548113180392046,
+      "grad_norm": 2.899485221765189,
+      "learning_rate": 1.4078725070332746e-06,
+      "loss": 0.3908,
+      "step": 837
+    },
+    {
+      "epoch": 0.3859416827379753,
+      "grad_norm": 3.085157334581907,
+      "learning_rate": 1.4065093947314396e-06,
+      "loss": 0.4573,
+      "step": 838
+    },
+    {
+      "epoch": 0.3864022336720302,
+      "grad_norm": 3.4352911831634887,
+      "learning_rate": 1.4051453769763143e-06,
+      "loss": 0.6433,
+      "step": 839
+    },
+    {
+      "epoch": 0.38686278460608503,
+      "grad_norm": 3.742434977201632,
+      "learning_rate": 1.4037804568060919e-06,
+      "loss": 0.5888,
+      "step": 840
+    },
+    {
+      "epoch": 0.38732333554013987,
+      "grad_norm": 2.9975870813637586,
+      "learning_rate": 1.402414637260977e-06,
+      "loss": 0.5201,
+      "step": 841
+    },
+    {
+      "epoch": 0.38778388647419476,
+      "grad_norm": 3.0225885866635367,
+      "learning_rate": 1.4010479213831762e-06,
+      "loss": 0.4794,
+      "step": 842
+    },
+    {
+      "epoch": 0.3882444374082496,
+      "grad_norm": 3.2296706627603857,
+      "learning_rate": 1.399680312216894e-06,
+      "loss": 0.5355,
+      "step": 843
+    },
+    {
+      "epoch": 0.3887049883423045,
+      "grad_norm": 3.0042503059260883,
+      "learning_rate": 1.3983118128083234e-06,
+      "loss": 0.5446,
+      "step": 844
+    },
+    {
+      "epoch": 0.38916553927635933,
+      "grad_norm": 3.5187839319359684,
+      "learning_rate": 1.3969424262056402e-06,
+      "loss": 0.598,
+      "step": 845
+    },
+    {
+      "epoch": 0.3896260902104142,
+      "grad_norm": 3.059136381267136,
+      "learning_rate": 1.3955721554589975e-06,
+      "loss": 0.621,
+      "step": 846
+    },
+    {
+      "epoch": 0.39008664114446906,
+      "grad_norm": 3.186481313726005,
+      "learning_rate": 1.3942010036205165e-06,
+      "loss": 0.5483,
+      "step": 847
+    },
+    {
+      "epoch": 0.39054719207852395,
+      "grad_norm": 3.3209014633626133,
+      "learning_rate": 1.392828973744282e-06,
+      "loss": 0.5836,
+      "step": 848
+    },
+    {
+      "epoch": 0.3910077430125788,
+      "grad_norm": 3.3185866705706184,
+      "learning_rate": 1.3914560688863336e-06,
+      "loss": 0.4829,
+      "step": 849
+    },
+    {
+      "epoch": 0.3914682939466337,
+      "grad_norm": 3.1927851100882862,
+      "learning_rate": 1.39008229210466e-06,
+      "loss": 0.5053,
+      "step": 850
+    },
+    {
+      "epoch": 0.3919288448806885,
+      "grad_norm": 3.3595068103015473,
+      "learning_rate": 1.3887076464591928e-06,
+      "loss": 0.5338,
+      "step": 851
+    },
+    {
+      "epoch": 0.3923893958147434,
+      "grad_norm": 2.7905012085301837,
+      "learning_rate": 1.3873321350117981e-06,
+      "loss": 0.4979,
+      "step": 852
+    },
+    {
+      "epoch": 0.39284994674879825,
+      "grad_norm": 3.090347369854456,
+      "learning_rate": 1.3859557608262705e-06,
+      "loss": 0.5797,
+      "step": 853
+    },
+    {
+      "epoch": 0.3933104976828531,
+      "grad_norm": 3.060994610172803,
+      "learning_rate": 1.384578526968326e-06,
+      "loss": 0.5184,
+      "step": 854
+    },
+    {
+      "epoch": 0.393771048616908,
+      "grad_norm": 3.6607469509314914,
+      "learning_rate": 1.3832004365055974e-06,
+      "loss": 0.5485,
+      "step": 855
+    },
+    {
+      "epoch": 0.3942315995509628,
+      "grad_norm": 2.7397062923925053,
+      "learning_rate": 1.3818214925076223e-06,
+      "loss": 0.4594,
+      "step": 856
+    },
+    {
+      "epoch": 0.3946921504850177,
+      "grad_norm": 3.0413655649505262,
+      "learning_rate": 1.380441698045842e-06,
+      "loss": 0.495,
+      "step": 857
+    },
+    {
+      "epoch": 0.39515270141907255,
+      "grad_norm": 3.4287955433322423,
+      "learning_rate": 1.3790610561935911e-06,
+      "loss": 0.645,
+      "step": 858
+    },
+    {
+      "epoch": 0.39561325235312744,
+      "grad_norm": 3.3357893051930816,
+      "learning_rate": 1.3776795700260915e-06,
+      "loss": 0.6037,
+      "step": 859
+    },
+    {
+      "epoch": 0.3960738032871823,
+      "grad_norm": 3.6164730160728413,
+      "learning_rate": 1.3762972426204461e-06,
+      "loss": 0.622,
+      "step": 860
+    },
+    {
+      "epoch": 0.3965343542212372,
+      "grad_norm": 3.386533311875332,
+      "learning_rate": 1.374914077055632e-06,
+      "loss": 0.6871,
+      "step": 861
+    },
+    {
+      "epoch": 0.396994905155292,
+      "grad_norm": 2.9410360813121352,
+      "learning_rate": 1.3735300764124916e-06,
+      "loss": 0.5102,
+      "step": 862
+    },
+    {
+      "epoch": 0.3974554560893469,
+      "grad_norm": 3.103887613925395,
+      "learning_rate": 1.3721452437737293e-06,
+      "loss": 0.5432,
+      "step": 863
+    },
+    {
+      "epoch": 0.39791600702340174,
+      "grad_norm": 3.517765292291276,
+      "learning_rate": 1.3707595822239015e-06,
+      "loss": 0.5159,
+      "step": 864
+    },
+    {
+      "epoch": 0.39837655795745663,
+      "grad_norm": 3.043417914703796,
+      "learning_rate": 1.3693730948494114e-06,
+      "loss": 0.5317,
+      "step": 865
+    },
+    {
+      "epoch": 0.39883710889151147,
+      "grad_norm": 3.338449707312485,
+      "learning_rate": 1.3679857847385009e-06,
+      "loss": 0.632,
+      "step": 866
+    },
+    {
+      "epoch": 0.3992976598255663,
+      "grad_norm": 3.253152412711078,
+      "learning_rate": 1.3665976549812452e-06,
+      "loss": 0.5134,
+      "step": 867
+    },
+    {
+      "epoch": 0.3997582107596212,
+      "grad_norm": 3.072410142653441,
+      "learning_rate": 1.365208708669545e-06,
+      "loss": 0.4643,
+      "step": 868
+    },
+    {
+      "epoch": 0.40021876169367604,
+      "grad_norm": 3.525403495130597,
+      "learning_rate": 1.36381894889712e-06,
+      "loss": 0.5517,
+      "step": 869
+    },
+    {
+      "epoch": 0.40067931262773093,
+      "grad_norm": 3.556527192054287,
+      "learning_rate": 1.362428378759501e-06,
+      "loss": 0.6111,
+      "step": 870
+    },
+    {
+      "epoch": 0.40113986356178577,
+      "grad_norm": 3.1449463321315236,
+      "learning_rate": 1.3610370013540247e-06,
+      "loss": 0.5445,
+      "step": 871
+    },
+    {
+      "epoch": 0.40160041449584066,
+      "grad_norm": 3.0202852656779577,
+      "learning_rate": 1.3596448197798253e-06,
+      "loss": 0.576,
+      "step": 872
+    },
+    {
+      "epoch": 0.4020609654298955,
+      "grad_norm": 3.08760829256673,
+      "learning_rate": 1.3582518371378282e-06,
+      "loss": 0.6271,
+      "step": 873
+    },
+    {
+      "epoch": 0.4025215163639504,
+      "grad_norm": 3.328960142744962,
+      "learning_rate": 1.3568580565307436e-06,
+      "loss": 0.5746,
+      "step": 874
+    },
+    {
+      "epoch": 0.40298206729800523,
+      "grad_norm": 3.0333254368255136,
+      "learning_rate": 1.355463481063059e-06,
+      "loss": 0.5229,
+      "step": 875
+    },
+    {
+      "epoch": 0.4034426182320601,
+      "grad_norm": 2.933573447144409,
+      "learning_rate": 1.3540681138410314e-06,
+      "loss": 0.5778,
+      "step": 876
+    },
+    {
+      "epoch": 0.40390316916611496,
+      "grad_norm": 2.882788964082942,
+      "learning_rate": 1.3526719579726829e-06,
+      "loss": 0.5073,
+      "step": 877
+    },
+    {
+      "epoch": 0.40436372010016985,
+      "grad_norm": 3.2365144628176568,
+      "learning_rate": 1.3512750165677906e-06,
+      "loss": 0.5211,
+      "step": 878
+    },
+    {
+      "epoch": 0.4048242710342247,
+      "grad_norm": 3.38193990207538,
+      "learning_rate": 1.3498772927378824e-06,
+      "loss": 0.606,
+      "step": 879
+    },
+    {
+      "epoch": 0.40528482196827953,
+      "grad_norm": 3.56597317135757,
+      "learning_rate": 1.348478789596229e-06,
+      "loss": 0.6701,
+      "step": 880
+    },
+    {
+      "epoch": 0.4057453729023344,
+      "grad_norm": 3.324345046149201,
+      "learning_rate": 1.3470795102578355e-06,
+      "loss": 0.6522,
+      "step": 881
+    },
+    {
+      "epoch": 0.40620592383638926,
+      "grad_norm": 3.340866287319074,
+      "learning_rate": 1.3456794578394382e-06,
+      "loss": 0.6136,
+      "step": 882
+    },
+    {
+      "epoch": 0.40666647477044415,
+      "grad_norm": 3.131575960928237,
+      "learning_rate": 1.3442786354594937e-06,
+      "loss": 0.5138,
+      "step": 883
+    },
+    {
+      "epoch": 0.407127025704499,
+      "grad_norm": 3.1323338349582306,
+      "learning_rate": 1.3428770462381739e-06,
+      "loss": 0.5023,
+      "step": 884
+    },
+    {
+      "epoch": 0.4075875766385539,
+      "grad_norm": 2.8727768076877664,
+      "learning_rate": 1.3414746932973583e-06,
+      "loss": 0.5462,
+      "step": 885
+    },
+    {
+      "epoch": 0.4080481275726087,
+      "grad_norm": 2.818312617541394,
+      "learning_rate": 1.340071579760629e-06,
+      "loss": 0.5003,
+      "step": 886
+    },
+    {
+      "epoch": 0.4085086785066636,
+      "grad_norm": 2.8944571031714674,
+      "learning_rate": 1.338667708753261e-06,
+      "loss": 0.4793,
+      "step": 887
+    },
+    {
+      "epoch": 0.40896922944071845,
+      "grad_norm": 3.0617046790599582,
+      "learning_rate": 1.3372630834022165e-06,
+      "loss": 0.5714,
+      "step": 888
+    },
+    {
+      "epoch": 0.40942978037477334,
+      "grad_norm": 3.3240252994477477,
+      "learning_rate": 1.3358577068361383e-06,
+      "loss": 0.7543,
+      "step": 889
+    },
+    {
+      "epoch": 0.4098903313088282,
+      "grad_norm": 3.387874616156958,
+      "learning_rate": 1.3344515821853427e-06,
+      "loss": 0.5331,
+      "step": 890
+    },
+    {
+      "epoch": 0.4103508822428831,
+      "grad_norm": 2.9538858616900296,
+      "learning_rate": 1.3330447125818114e-06,
+      "loss": 0.5126,
+      "step": 891
+    },
+    {
+      "epoch": 0.4108114331769379,
+      "grad_norm": 2.6905912340286045,
+      "learning_rate": 1.331637101159186e-06,
+      "loss": 0.6335,
+      "step": 892
+    },
+    {
+      "epoch": 0.41127198411099275,
+      "grad_norm": 2.8234464497330842,
+      "learning_rate": 1.3302287510527606e-06,
+      "loss": 0.485,
+      "step": 893
+    },
+    {
+      "epoch": 0.41173253504504764,
+      "grad_norm": 3.364736286801205,
+      "learning_rate": 1.3288196653994742e-06,
+      "loss": 0.6351,
+      "step": 894
+    },
+    {
+      "epoch": 0.4121930859791025,
+      "grad_norm": 3.148255912278157,
+      "learning_rate": 1.3274098473379041e-06,
+      "loss": 0.501,
+      "step": 895
+    },
+    {
+      "epoch": 0.4126536369131574,
+      "grad_norm": 2.9927281051396397,
+      "learning_rate": 1.3259993000082597e-06,
+      "loss": 0.6576,
+      "step": 896
+    },
+    {
+      "epoch": 0.4131141878472122,
+      "grad_norm": 2.6304206915731387,
+      "learning_rate": 1.3245880265523737e-06,
+      "loss": 0.5005,
+      "step": 897
+    },
+    {
+      "epoch": 0.4135747387812671,
+      "grad_norm": 3.033315201466835,
+      "learning_rate": 1.3231760301136968e-06,
+      "loss": 0.5032,
+      "step": 898
+    },
+    {
+      "epoch": 0.41403528971532194,
+      "grad_norm": 3.153932562017399,
+      "learning_rate": 1.32176331383729e-06,
+      "loss": 0.6438,
+      "step": 899
+    },
+    {
+      "epoch": 0.41449584064937683,
+      "grad_norm": 3.0569467140610564,
+      "learning_rate": 1.3203498808698177e-06,
+      "loss": 0.4582,
+      "step": 900
+    },
+    {
+      "epoch": 0.41495639158343167,
+      "grad_norm": 3.4798136677368525,
+      "learning_rate": 1.3189357343595405e-06,
+      "loss": 0.4739,
+      "step": 901
+    },
+    {
+      "epoch": 0.41541694251748656,
+      "grad_norm": 3.267066000558948,
+      "learning_rate": 1.317520877456308e-06,
+      "loss": 0.5333,
+      "step": 902
+    },
+    {
+      "epoch": 0.4158774934515414,
+      "grad_norm": 3.0973720845505057,
+      "learning_rate": 1.3161053133115534e-06,
+      "loss": 0.6643,
+      "step": 903
+    },
+    {
+      "epoch": 0.4163380443855963,
+      "grad_norm": 3.124046522546217,
+      "learning_rate": 1.3146890450782833e-06,
+      "loss": 0.6385,
+      "step": 904
+    },
+    {
+      "epoch": 0.41679859531965113,
+      "grad_norm": 3.2564321361920348,
+      "learning_rate": 1.3132720759110742e-06,
+      "loss": 0.5914,
+      "step": 905
+    },
+    {
+      "epoch": 0.41725914625370597,
+      "grad_norm": 3.5529506389797207,
+      "learning_rate": 1.3118544089660632e-06,
+      "loss": 0.5549,
+      "step": 906
+    },
+    {
+      "epoch": 0.41771969718776086,
+      "grad_norm": 3.1088968115091458,
+      "learning_rate": 1.3104360474009413e-06,
+      "loss": 0.5016,
+      "step": 907
+    },
+    {
+      "epoch": 0.4181802481218157,
+      "grad_norm": 2.961528786967664,
+      "learning_rate": 1.3090169943749473e-06,
+      "loss": 0.6016,
+      "step": 908
+    },
+    {
+      "epoch": 0.4186407990558706,
+      "grad_norm": 2.9590563505493557,
+      "learning_rate": 1.3075972530488601e-06,
+      "loss": 0.6245,
+      "step": 909
+    },
+    {
+      "epoch": 0.41910134998992543,
+      "grad_norm": 3.2817559385376502,
+      "learning_rate": 1.306176826584991e-06,
+      "loss": 0.5741,
+      "step": 910
+    },
+    {
+      "epoch": 0.4195619009239803,
+      "grad_norm": 3.1851733205126456,
+      "learning_rate": 1.3047557181471782e-06,
+      "loss": 0.6102,
+      "step": 911
+    },
+    {
+      "epoch": 0.42002245185803516,
+      "grad_norm": 3.3709380563412856,
+      "learning_rate": 1.3033339309007782e-06,
+      "loss": 0.6029,
+      "step": 912
+    },
+    {
+      "epoch": 0.42048300279209005,
+      "grad_norm": 2.978439007428691,
+      "learning_rate": 1.3019114680126607e-06,
+      "loss": 0.6095,
+      "step": 913
+    },
+    {
+      "epoch": 0.4209435537261449,
+      "grad_norm": 3.3731454043582545,
+      "learning_rate": 1.3004883326511986e-06,
+      "loss": 0.6838,
+      "step": 914
+    },
+    {
+      "epoch": 0.4214041046601998,
+      "grad_norm": 3.3633712629594723,
+      "learning_rate": 1.2990645279862637e-06,
+      "loss": 0.637,
+      "step": 915
+    },
+    {
+      "epoch": 0.4218646555942546,
+      "grad_norm": 3.3342371154036936,
+      "learning_rate": 1.2976400571892187e-06,
+      "loss": 0.5637,
+      "step": 916
+    },
+    {
+      "epoch": 0.4223252065283095,
+      "grad_norm": 3.2152650834374885,
+      "learning_rate": 1.2962149234329096e-06,
+      "loss": 0.5651,
+      "step": 917
+    },
+    {
+      "epoch": 0.42278575746236435,
+      "grad_norm": 3.728918503268583,
+      "learning_rate": 1.2947891298916597e-06,
+      "loss": 0.5591,
+      "step": 918
+    },
+    {
+      "epoch": 0.4232463083964192,
+      "grad_norm": 3.4867704670014814,
+      "learning_rate": 1.2933626797412601e-06,
+      "loss": 0.4446,
+      "step": 919
+    },
+    {
+      "epoch": 0.4237068593304741,
+      "grad_norm": 3.3645940418700717,
+      "learning_rate": 1.2919355761589673e-06,
+      "loss": 0.5977,
+      "step": 920
+    },
+    {
+      "epoch": 0.4241674102645289,
+      "grad_norm": 3.165145021423808,
+      "learning_rate": 1.2905078223234907e-06,
+      "loss": 0.6094,
+      "step": 921
+    },
+    {
+      "epoch": 0.4246279611985838,
+      "grad_norm": 3.1311304590118083,
+      "learning_rate": 1.2890794214149895e-06,
+      "loss": 0.5793,
+      "step": 922
+    },
+    {
+      "epoch": 0.42508851213263865,
+      "grad_norm": 2.8231203010209462,
+      "learning_rate": 1.2876503766150634e-06,
+      "loss": 0.5225,
+      "step": 923
+    },
+    {
+      "epoch": 0.42554906306669354,
+      "grad_norm": 3.268397607752156,
+      "learning_rate": 1.2862206911067467e-06,
+      "loss": 0.5686,
+      "step": 924
+    },
+    {
+      "epoch": 0.4260096140007484,
+      "grad_norm": 3.305099301303672,
+      "learning_rate": 1.2847903680745012e-06,
+      "loss": 0.5445,
+      "step": 925
+    },
+    {
+      "epoch": 0.4264701649348033,
+      "grad_norm": 3.3965795137188675,
+      "learning_rate": 1.2833594107042075e-06,
+      "loss": 0.6591,
+      "step": 926
+    },
+    {
+      "epoch": 0.4269307158688581,
+      "grad_norm": 3.483867364296341,
+      "learning_rate": 1.2819278221831604e-06,
+      "loss": 0.6056,
+      "step": 927
+    },
+    {
+      "epoch": 0.427391266802913,
+      "grad_norm": 3.4347389707625107,
+      "learning_rate": 1.2804956057000597e-06,
+      "loss": 0.5861,
+      "step": 928
+    },
+    {
+      "epoch": 0.42785181773696784,
+      "grad_norm": 3.1912886523840545,
+      "learning_rate": 1.2790627644450042e-06,
+      "loss": 0.5413,
+      "step": 929
+    },
+    {
+      "epoch": 0.42831236867102274,
+      "grad_norm": 3.277432951067786,
+      "learning_rate": 1.2776293016094848e-06,
+      "loss": 0.5728,
+      "step": 930
+    },
+    {
+      "epoch": 0.4287729196050776,
+      "grad_norm": 3.1216436794666897,
+      "learning_rate": 1.2761952203863758e-06,
+      "loss": 0.5291,
+      "step": 931
+    },
+    {
+      "epoch": 0.4292334705391324,
+      "grad_norm": 3.2701087953996315,
+      "learning_rate": 1.2747605239699293e-06,
+      "loss": 0.5777,
+      "step": 932
+    },
+    {
+      "epoch": 0.4296940214731873,
+      "grad_norm": 2.703298534665295,
+      "learning_rate": 1.2733252155557686e-06,
+      "loss": 0.4752,
+      "step": 933
+    },
+    {
+      "epoch": 0.43015457240724214,
+      "grad_norm": 3.5331916544052584,
+      "learning_rate": 1.2718892983408787e-06,
+      "loss": 0.5741,
+      "step": 934
+    },
+    {
+      "epoch": 0.43061512334129703,
+      "grad_norm": 3.2776132030927196,
+      "learning_rate": 1.270452775523602e-06,
+      "loss": 0.6765,
+      "step": 935
+    },
+    {
+      "epoch": 0.43107567427535187,
+      "grad_norm": 3.556572826939555,
+      "learning_rate": 1.2690156503036288e-06,
+      "loss": 0.5738,
+      "step": 936
+    },
+    {
+      "epoch": 0.43153622520940677,
+      "grad_norm": 2.6845624221473003,
+      "learning_rate": 1.2675779258819913e-06,
+      "loss": 0.437,
+      "step": 937
+    },
+    {
+      "epoch": 0.4319967761434616,
+      "grad_norm": 3.0987030708733965,
+      "learning_rate": 1.2661396054610568e-06,
+      "loss": 0.5322,
+      "step": 938
+    },
+    {
+      "epoch": 0.4324573270775165,
+      "grad_norm": 2.69969210409678,
+      "learning_rate": 1.2647006922445203e-06,
+      "loss": 0.563,
+      "step": 939
+    },
+    {
+      "epoch": 0.43291787801157133,
+      "grad_norm": 3.1443985858808112,
+      "learning_rate": 1.2632611894373963e-06,
+      "loss": 0.5224,
+      "step": 940
+    },
+    {
+      "epoch": 0.4333784289456262,
+      "grad_norm": 3.041880247975588,
+      "learning_rate": 1.2618211002460133e-06,
+      "loss": 0.6286,
+      "step": 941
+    },
+    {
+      "epoch": 0.43383897987968106,
+      "grad_norm": 3.1243004909346612,
+      "learning_rate": 1.2603804278780054e-06,
+      "loss": 0.4417,
+      "step": 942
+    },
+    {
+      "epoch": 0.43429953081373596,
+      "grad_norm": 3.0394710034585044,
+      "learning_rate": 1.2589391755423061e-06,
+      "loss": 0.3974,
+      "step": 943
+    },
+    {
+      "epoch": 0.4347600817477908,
+      "grad_norm": 2.9343561758404855,
+      "learning_rate": 1.2574973464491406e-06,
+      "loss": 0.5109,
+      "step": 944
+    },
+    {
+      "epoch": 0.43522063268184563,
+      "grad_norm": 3.5301907527590837,
+      "learning_rate": 1.2560549438100187e-06,
+      "loss": 0.6818,
+      "step": 945
+    },
+    {
+      "epoch": 0.4356811836159005,
+      "grad_norm": 3.098431035830371,
+      "learning_rate": 1.2546119708377273e-06,
+      "loss": 0.6054,
+      "step": 946
+    },
+    {
+      "epoch": 0.43614173454995536,
+      "grad_norm": 3.0568020174713344,
+      "learning_rate": 1.2531684307463243e-06,
+      "loss": 0.5611,
+      "step": 947
+    },
+    {
+      "epoch": 0.43660228548401026,
+      "grad_norm": 3.133249513429743,
+      "learning_rate": 1.2517243267511308e-06,
+      "loss": 0.5266,
+      "step": 948
+    },
+    {
+      "epoch": 0.4370628364180651,
+      "grad_norm": 3.5046506132656656,
+      "learning_rate": 1.2502796620687232e-06,
+      "loss": 0.7523,
+      "step": 949
+    },
+    {
+      "epoch": 0.43752338735212,
+      "grad_norm": 3.8415980756490913,
+      "learning_rate": 1.2488344399169275e-06,
+      "loss": 0.6561,
+      "step": 950
+    },
+    {
+      "epoch": 0.4379839382861748,
+      "grad_norm": 3.04800030433528,
+      "learning_rate": 1.2473886635148107e-06,
+      "loss": 0.3969,
+      "step": 951
+    },
+    {
+      "epoch": 0.4384444892202297,
+      "grad_norm": 2.9634130405670542,
+      "learning_rate": 1.2459423360826753e-06,
+      "loss": 0.5083,
+      "step": 952
+    },
+    {
+      "epoch": 0.43890504015428455,
+      "grad_norm": 3.2286597169487496,
+      "learning_rate": 1.2444954608420509e-06,
+      "loss": 0.5204,
+      "step": 953
+    },
+    {
+      "epoch": 0.43936559108833945,
+      "grad_norm": 3.072141673231512,
+      "learning_rate": 1.2430480410156859e-06,
+      "loss": 0.5266,
+      "step": 954
+    },
+    {
+      "epoch": 0.4398261420223943,
+      "grad_norm": 3.1772864308017503,
+      "learning_rate": 1.2416000798275434e-06,
+      "loss": 0.5762,
+      "step": 955
+    },
+    {
+      "epoch": 0.4402866929564492,
+      "grad_norm": 3.1044222947486766,
+      "learning_rate": 1.2401515805027923e-06,
+      "loss": 0.6046,
+      "step": 956
+    },
+    {
+      "epoch": 0.440747243890504,
+      "grad_norm": 2.9714963515506674,
+      "learning_rate": 1.2387025462677986e-06,
+      "loss": 0.5226,
+      "step": 957
+    },
+    {
+      "epoch": 0.44120779482455885,
+      "grad_norm": 3.6632065071396602,
+      "learning_rate": 1.2372529803501212e-06,
+      "loss": 0.6285,
+      "step": 958
+    },
+    {
+      "epoch": 0.44166834575861375,
+      "grad_norm": 3.3335254391441236,
+      "learning_rate": 1.2358028859785027e-06,
+      "loss": 0.6491,
+      "step": 959
+    },
+    {
+      "epoch": 0.4421288966926686,
+      "grad_norm": 3.2741227211236277,
+      "learning_rate": 1.234352266382863e-06,
+      "loss": 0.472,
+      "step": 960
+    },
+    {
+      "epoch": 0.4425894476267235,
+      "grad_norm": 2.734950400129683,
+      "learning_rate": 1.2329011247942913e-06,
+      "loss": 0.4678,
+      "step": 961
+    },
+    {
+      "epoch": 0.4430499985607783,
+      "grad_norm": 3.706836559033099,
+      "learning_rate": 1.2314494644450405e-06,
+      "loss": 0.572,
+      "step": 962
+    },
+    {
+      "epoch": 0.4435105494948332,
+      "grad_norm": 3.191997067089114,
+      "learning_rate": 1.2299972885685175e-06,
+      "loss": 0.526,
+      "step": 963
+    },
+    {
+      "epoch": 0.44397110042888804,
+      "grad_norm": 3.750204622573935,
+      "learning_rate": 1.2285446003992794e-06,
+      "loss": 0.6557,
+      "step": 964
+    },
+    {
+      "epoch": 0.44443165136294294,
+      "grad_norm": 3.3390534313214086,
+      "learning_rate": 1.2270914031730227e-06,
+      "loss": 0.6844,
+      "step": 965
+    },
+    {
+      "epoch": 0.4448922022969978,
+      "grad_norm": 3.4357388932326103,
+      "learning_rate": 1.2256377001265782e-06,
+      "loss": 0.589,
+      "step": 966
+    },
+    {
+      "epoch": 0.44535275323105267,
+      "grad_norm": 2.9239348458649497,
+      "learning_rate": 1.2241834944979043e-06,
+      "loss": 0.5562,
+      "step": 967
+    },
+    {
+      "epoch": 0.4458133041651075,
+      "grad_norm": 3.0247685003568656,
+      "learning_rate": 1.2227287895260774e-06,
+      "loss": 0.5027,
+      "step": 968
+    },
+    {
+      "epoch": 0.4462738550991624,
+      "grad_norm": 3.4607078630570274,
+      "learning_rate": 1.2212735884512873e-06,
+      "loss": 0.5647,
+      "step": 969
+    },
+    {
+      "epoch": 0.44673440603321724,
+      "grad_norm": 3.134258529274002,
+      "learning_rate": 1.2198178945148284e-06,
+      "loss": 0.6152,
+      "step": 970
+    },
+    {
+      "epoch": 0.4471949569672721,
+      "grad_norm": 3.2437674513922143,
+      "learning_rate": 1.2183617109590923e-06,
+      "loss": 0.4777,
+      "step": 971
+    },
+    {
+      "epoch": 0.44765550790132697,
+      "grad_norm": 2.9881905608566846,
+      "learning_rate": 1.2169050410275617e-06,
+      "loss": 0.5202,
+      "step": 972
+    },
+    {
+      "epoch": 0.4481160588353818,
+      "grad_norm": 3.186446601937435,
+      "learning_rate": 1.2154478879648034e-06,
+      "loss": 0.5035,
+      "step": 973
+    },
+    {
+      "epoch": 0.4485766097694367,
+      "grad_norm": 3.5679383322300495,
+      "learning_rate": 1.213990255016459e-06,
+      "loss": 0.6799,
+      "step": 974
+    },
+    {
+      "epoch": 0.44903716070349153,
+      "grad_norm": 3.3889885780961526,
+      "learning_rate": 1.2125321454292397e-06,
+      "loss": 0.638,
+      "step": 975
+    },
+    {
+      "epoch": 0.4494977116375464,
+      "grad_norm": 3.0589827390964848,
+      "learning_rate": 1.2110735624509184e-06,
+      "loss": 0.6329,
+      "step": 976
+    },
+    {
+      "epoch": 0.44995826257160126,
+      "grad_norm": 3.2178815155136067,
+      "learning_rate": 1.2096145093303215e-06,
+      "loss": 0.5451,
+      "step": 977
+    },
+    {
+      "epoch": 0.45041881350565616,
+      "grad_norm": 3.1650611316618398,
+      "learning_rate": 1.2081549893173244e-06,
+      "loss": 0.5188,
+      "step": 978
+    },
+    {
+      "epoch": 0.450879364439711,
+      "grad_norm": 2.7477870679696226,
+      "learning_rate": 1.206695005662841e-06,
+      "loss": 0.4523,
+      "step": 979
+    },
+    {
+      "epoch": 0.4513399153737659,
+      "grad_norm": 2.9102863855372747,
+      "learning_rate": 1.2052345616188177e-06,
+      "loss": 0.5282,
+      "step": 980
+    },
+    {
+      "epoch": 0.4518004663078207,
+      "grad_norm": 3.1966574178250116,
+      "learning_rate": 1.2037736604382277e-06,
+      "loss": 0.5015,
+      "step": 981
+    },
+    {
+      "epoch": 0.4522610172418756,
+      "grad_norm": 2.89336588474616,
+      "learning_rate": 1.2023123053750613e-06,
+      "loss": 0.5461,
+      "step": 982
+    },
+    {
+      "epoch": 0.45272156817593046,
+      "grad_norm": 3.832342329803752,
+      "learning_rate": 1.2008504996843206e-06,
+      "loss": 0.6394,
+      "step": 983
+    },
+    {
+      "epoch": 0.4531821191099853,
+      "grad_norm": 3.4482291994471086,
+      "learning_rate": 1.1993882466220102e-06,
+      "loss": 0.6259,
+      "step": 984
+    },
+    {
+      "epoch": 0.4536426700440402,
+      "grad_norm": 3.611548614335293,
+      "learning_rate": 1.1979255494451326e-06,
+      "loss": 0.6066,
+      "step": 985
+    },
+    {
+      "epoch": 0.454103220978095,
+      "grad_norm": 2.7453352434114713,
+      "learning_rate": 1.1964624114116784e-06,
+      "loss": 0.6955,
+      "step": 986
+    },
+    {
+      "epoch": 0.4545637719121499,
+      "grad_norm": 3.075608198924997,
+      "learning_rate": 1.194998835780621e-06,
+      "loss": 0.6155,
+      "step": 987
+    },
+    {
+      "epoch": 0.45502432284620475,
+      "grad_norm": 3.1080551036437654,
+      "learning_rate": 1.1935348258119083e-06,
+      "loss": 0.578,
+      "step": 988
+    },
+    {
+      "epoch": 0.45548487378025965,
+      "grad_norm": 3.2505311952038234,
+      "learning_rate": 1.1920703847664546e-06,
+      "loss": 0.535,
+      "step": 989
+    },
+    {
+      "epoch": 0.4559454247143145,
+      "grad_norm": 3.0544071466547336,
+      "learning_rate": 1.190605515906136e-06,
+      "loss": 0.4539,
+      "step": 990
+    },
+    {
+      "epoch": 0.4564059756483694,
+      "grad_norm": 2.9834516516156366,
+      "learning_rate": 1.1891402224937804e-06,
+      "loss": 0.5145,
+      "step": 991
+    },
+    {
+      "epoch": 0.4568665265824242,
+      "grad_norm": 3.2098618470017963,
+      "learning_rate": 1.1876745077931617e-06,
+      "loss": 0.4956,
+      "step": 992
+    },
+    {
+      "epoch": 0.4573270775164791,
+      "grad_norm": 3.00024716203184,
+      "learning_rate": 1.1862083750689923e-06,
+      "loss": 0.5351,
+      "step": 993
+    },
+    {
+      "epoch": 0.45778762845053395,
+      "grad_norm": 3.084747400551803,
+      "learning_rate": 1.1847418275869151e-06,
+      "loss": 0.5766,
+      "step": 994
+    },
+    {
+      "epoch": 0.45824817938458884,
+      "grad_norm": 3.04028326868723,
+      "learning_rate": 1.183274868613498e-06,
+      "loss": 0.6411,
+      "step": 995
+    },
+    {
+      "epoch": 0.4587087303186437,
+      "grad_norm": 2.5418869985884354,
+      "learning_rate": 1.181807501416224e-06,
+      "loss": 0.4326,
+      "step": 996
+    },
+    {
+      "epoch": 0.4591692812526985,
+      "grad_norm": 3.5975504622299534,
+      "learning_rate": 1.1803397292634867e-06,
+      "loss": 0.6207,
+      "step": 997
+    },
+    {
+      "epoch": 0.4596298321867534,
+      "grad_norm": 3.385303810357531,
+      "learning_rate": 1.1788715554245807e-06,
+      "loss": 0.6074,
+      "step": 998
+    },
+    {
+      "epoch": 0.46009038312080824,
+      "grad_norm": 3.0211329657757005,
+      "learning_rate": 1.1774029831696955e-06,
+      "loss": 0.6681,
+      "step": 999
+    },
+    {
+      "epoch": 0.46055093405486314,
+      "grad_norm": 3.055209977773643,
+      "learning_rate": 1.1759340157699088e-06,
+      "loss": 0.5646,
+      "step": 1000
+    },
+    {
+      "epoch": 0.461011484988918,
+      "grad_norm": 3.1014875109105753,
+      "learning_rate": 1.1744646564971777e-06,
+      "loss": 0.5598,
+      "step": 1001
+    },
+    {
+      "epoch": 0.46147203592297287,
+      "grad_norm": 3.256357038090698,
+      "learning_rate": 1.1729949086243319e-06,
+      "loss": 0.6722,
+      "step": 1002
+    },
+    {
+      "epoch": 0.4619325868570277,
+      "grad_norm": 2.917952076658023,
+      "learning_rate": 1.1715247754250673e-06,
+      "loss": 0.4466,
+      "step": 1003
+    },
+    {
+      "epoch": 0.4623931377910826,
+      "grad_norm": 2.7874434854895207,
+      "learning_rate": 1.1700542601739381e-06,
+      "loss": 0.5989,
+      "step": 1004
+    },
+    {
+      "epoch": 0.46285368872513744,
+      "grad_norm": 3.0402556453953546,
+      "learning_rate": 1.1685833661463488e-06,
+      "loss": 0.4839,
+      "step": 1005
+    },
+    {
+      "epoch": 0.46331423965919233,
+      "grad_norm": 3.544520931422804,
+      "learning_rate": 1.1671120966185484e-06,
+      "loss": 0.5576,
+      "step": 1006
+    },
+    {
+      "epoch": 0.46377479059324717,
+      "grad_norm": 3.3715770705932204,
+      "learning_rate": 1.1656404548676219e-06,
+      "loss": 0.595,
+      "step": 1007
+    },
+    {
+      "epoch": 0.46423534152730206,
+      "grad_norm": 3.07379805336765,
+      "learning_rate": 1.1641684441714828e-06,
+      "loss": 0.5866,
+      "step": 1008
+    },
+    {
+      "epoch": 0.4646958924613569,
+      "grad_norm": 3.2769110964863413,
+      "learning_rate": 1.1626960678088677e-06,
+      "loss": 0.4229,
+      "step": 1009
+    },
+    {
+      "epoch": 0.46515644339541173,
+      "grad_norm": 3.231165225029422,
+      "learning_rate": 1.1612233290593264e-06,
+      "loss": 0.4908,
+      "step": 1010
+    },
+    {
+      "epoch": 0.4656169943294666,
+      "grad_norm": 2.6717664603336164,
+      "learning_rate": 1.1597502312032168e-06,
+      "loss": 0.5189,
+      "step": 1011
+    },
+    {
+      "epoch": 0.46607754526352146,
+      "grad_norm": 3.5412588764816575,
+      "learning_rate": 1.158276777521696e-06,
+      "loss": 0.5227,
+      "step": 1012
+    },
+    {
+      "epoch": 0.46653809619757636,
+      "grad_norm": 3.367681094901886,
+      "learning_rate": 1.1568029712967137e-06,
+      "loss": 0.4772,
+      "step": 1013
+    },
+    {
+      "epoch": 0.4669986471316312,
+      "grad_norm": 2.9896273143544185,
+      "learning_rate": 1.1553288158110057e-06,
+      "loss": 0.6199,
+      "step": 1014
+    },
+    {
+      "epoch": 0.4674591980656861,
+      "grad_norm": 3.5056640416959346,
+      "learning_rate": 1.153854314348085e-06,
+      "loss": 0.507,
+      "step": 1015
+    },
+    {
+      "epoch": 0.4679197489997409,
+      "grad_norm": 2.8729584735763694,
+      "learning_rate": 1.152379470192235e-06,
+      "loss": 0.5681,
+      "step": 1016
+    },
+    {
+      "epoch": 0.4683802999337958,
+      "grad_norm": 3.1311855689183608,
+      "learning_rate": 1.1509042866285028e-06,
+      "loss": 0.5204,
+      "step": 1017
+    },
+    {
+      "epoch": 0.46884085086785066,
+      "grad_norm": 3.6303555087991493,
+      "learning_rate": 1.149428766942692e-06,
+      "loss": 0.5597,
+      "step": 1018
+    },
+    {
+      "epoch": 0.46930140180190555,
+      "grad_norm": 3.0256164571238795,
+      "learning_rate": 1.1479529144213537e-06,
+      "loss": 0.5227,
+      "step": 1019
+    },
+    {
+      "epoch": 0.4697619527359604,
+      "grad_norm": 2.7202389363979016,
+      "learning_rate": 1.1464767323517813e-06,
+      "loss": 0.3788,
+      "step": 1020
+    },
+    {
+      "epoch": 0.4702225036700153,
+      "grad_norm": 2.9201338750532893,
+      "learning_rate": 1.145000224022002e-06,
+      "loss": 0.4937,
+      "step": 1021
+    },
+    {
+      "epoch": 0.4706830546040701,
+      "grad_norm": 3.387209727218353,
+      "learning_rate": 1.143523392720769e-06,
+      "loss": 0.5268,
+      "step": 1022
+    },
+    {
+      "epoch": 0.47114360553812495,
+      "grad_norm": 2.8811446548900124,
+      "learning_rate": 1.1420462417375562e-06,
+      "loss": 0.4288,
+      "step": 1023
+    },
+    {
+      "epoch": 0.47160415647217985,
+      "grad_norm": 3.416259827843531,
+      "learning_rate": 1.140568774362549e-06,
+      "loss": 0.5689,
+      "step": 1024
+    },
+    {
+      "epoch": 0.4720647074062347,
+      "grad_norm": 3.1462397361747043,
+      "learning_rate": 1.1390909938866367e-06,
+      "loss": 0.5057,
+      "step": 1025
+    },
+    {
+      "epoch": 0.4725252583402896,
+      "grad_norm": 3.064526836533319,
+      "learning_rate": 1.137612903601407e-06,
+      "loss": 0.4729,
+      "step": 1026
+    },
+    {
+      "epoch": 0.4729858092743444,
+      "grad_norm": 3.451553189665636,
+      "learning_rate": 1.1361345067991375e-06,
+      "loss": 0.7637,
+      "step": 1027
+    },
+    {
+      "epoch": 0.4734463602083993,
+      "grad_norm": 2.8877618039147515,
+      "learning_rate": 1.134655806772788e-06,
+      "loss": 0.4614,
+      "step": 1028
+    },
+    {
+      "epoch": 0.47390691114245415,
+      "grad_norm": 3.253794710465501,
+      "learning_rate": 1.1331768068159946e-06,
+      "loss": 0.6915,
+      "step": 1029
+    },
+    {
+      "epoch": 0.47436746207650904,
+      "grad_norm": 3.6020230395836617,
+      "learning_rate": 1.1316975102230604e-06,
+      "loss": 0.5978,
+      "step": 1030
+    },
+    {
+      "epoch": 0.4748280130105639,
+      "grad_norm": 3.2619527720299444,
+      "learning_rate": 1.1302179202889505e-06,
+      "loss": 0.5066,
+      "step": 1031
+    },
+    {
+      "epoch": 0.47528856394461877,
+      "grad_norm": 3.2521154822577687,
+      "learning_rate": 1.1287380403092816e-06,
+      "loss": 0.5423,
+      "step": 1032
+    },
+    {
+      "epoch": 0.4757491148786736,
+      "grad_norm": 2.851595035378972,
+      "learning_rate": 1.127257873580318e-06,
+      "loss": 0.493,
+      "step": 1033
+    },
+    {
+      "epoch": 0.4762096658127285,
+      "grad_norm": 2.813869134619398,
+      "learning_rate": 1.1257774233989623e-06,
+      "loss": 0.4491,
+      "step": 1034
+    },
+    {
+      "epoch": 0.47667021674678334,
+      "grad_norm": 2.8702386818101977,
+      "learning_rate": 1.1242966930627484e-06,
+      "loss": 0.4677,
+      "step": 1035
+    },
+    {
+      "epoch": 0.47713076768083823,
+      "grad_norm": 2.9773792194983235,
+      "learning_rate": 1.1228156858698343e-06,
+      "loss": 0.4339,
+      "step": 1036
+    },
+    {
+      "epoch": 0.47759131861489307,
+      "grad_norm": 2.936870179141437,
+      "learning_rate": 1.1213344051189939e-06,
+      "loss": 0.5943,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4780518695489479,
+      "grad_norm": 3.305259733776011,
+      "learning_rate": 1.1198528541096115e-06,
+      "loss": 0.4593,
+      "step": 1038
+    },
+    {
+      "epoch": 0.4785124204830028,
+      "grad_norm": 3.36680958646396,
+      "learning_rate": 1.1183710361416727e-06,
+      "loss": 0.7228,
+      "step": 1039
+    },
+    {
+      "epoch": 0.47897297141705764,
+      "grad_norm": 3.2573238811104486,
+      "learning_rate": 1.1168889545157582e-06,
+      "loss": 0.5007,
+      "step": 1040
+    },
+    {
+      "epoch": 0.47943352235111253,
+      "grad_norm": 2.9011246681530043,
+      "learning_rate": 1.1154066125330357e-06,
+      "loss": 0.5315,
+      "step": 1041
+    },
+    {
+      "epoch": 0.47989407328516737,
+      "grad_norm": 2.981185948614369,
+      "learning_rate": 1.1139240134952523e-06,
+      "loss": 0.5441,
+      "step": 1042
+    },
+    {
+      "epoch": 0.48035462421922226,
+      "grad_norm": 3.255149727332359,
+      "learning_rate": 1.1124411607047288e-06,
+      "loss": 0.5446,
+      "step": 1043
+    },
+    {
+      "epoch": 0.4808151751532771,
+      "grad_norm": 3.1280648749420132,
+      "learning_rate": 1.1109580574643503e-06,
+      "loss": 0.5637,
+      "step": 1044
+    },
+    {
+      "epoch": 0.481275726087332,
+      "grad_norm": 3.124825859884878,
+      "learning_rate": 1.10947470707756e-06,
+      "loss": 0.4994,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4817362770213868,
+      "grad_norm": 3.66113588450651,
+      "learning_rate": 1.107991112848352e-06,
+      "loss": 0.5917,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4821968279554417,
+      "grad_norm": 3.1021700535523298,
+      "learning_rate": 1.1065072780812625e-06,
+      "loss": 0.5356,
+      "step": 1047
+    },
+    {
+      "epoch": 0.48265737888949656,
+      "grad_norm": 3.3696723546575194,
+      "learning_rate": 1.1050232060813644e-06,
+      "loss": 0.5811,
+      "step": 1048
+    },
+    {
+      "epoch": 0.48311792982355145,
+      "grad_norm": 3.0818897055953434,
+      "learning_rate": 1.1035389001542595e-06,
+      "loss": 0.6459,
+      "step": 1049
+    },
+    {
+      "epoch": 0.4835784807576063,
+      "grad_norm": 3.024745793339092,
+      "learning_rate": 1.1020543636060683e-06,
+      "loss": 0.5107,
+      "step": 1050
+    },
+    {
+      "epoch": 0.4840390316916611,
+      "grad_norm": 3.2367103732823774,
+      "learning_rate": 1.100569599743428e-06,
+      "loss": 0.5783,
+      "step": 1051
+    },
+    {
+      "epoch": 0.484499582625716,
+      "grad_norm": 3.2724898116979797,
+      "learning_rate": 1.09908461187348e-06,
+      "loss": 0.6343,
+      "step": 1052
+    },
+    {
+      "epoch": 0.48496013355977086,
+      "grad_norm": 2.6886278432787836,
+      "learning_rate": 1.0975994033038656e-06,
+      "loss": 0.5134,
+      "step": 1053
+    },
+    {
+      "epoch": 0.48542068449382575,
+      "grad_norm": 3.5676405092063175,
+      "learning_rate": 1.0961139773427171e-06,
+      "loss": 0.6208,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4858812354278806,
+      "grad_norm": 2.8882955010908598,
+      "learning_rate": 1.0946283372986516e-06,
+      "loss": 0.5888,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4863417863619355,
+      "grad_norm": 3.5114513574479087,
+      "learning_rate": 1.0931424864807623e-06,
+      "loss": 0.57,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4868023372959903,
+      "grad_norm": 3.0588259590216236,
+      "learning_rate": 1.0916564281986133e-06,
+      "loss": 0.5229,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4872628882300452,
+      "grad_norm": 3.0749972740879348,
+      "learning_rate": 1.0901701657622291e-06,
+      "loss": 0.5475,
+      "step": 1058
+    },
+    {
+      "epoch": 0.48772343916410005,
+      "grad_norm": 3.6869592539698215,
+      "learning_rate": 1.0886837024820897e-06,
+      "loss": 0.6199,
+      "step": 1059
+    },
+    {
+      "epoch": 0.48818399009815494,
+      "grad_norm": 3.3785618023616797,
+      "learning_rate": 1.0871970416691227e-06,
+      "loss": 0.6568,
+      "step": 1060
+    },
+    {
+      "epoch": 0.4886445410322098,
+      "grad_norm": 3.2577480072609193,
+      "learning_rate": 1.085710186634695e-06,
+      "loss": 0.6063,
+      "step": 1061
+    },
+    {
+      "epoch": 0.48910509196626467,
+      "grad_norm": 2.930851386132042,
+      "learning_rate": 1.0842231406906076e-06,
+      "loss": 0.5489,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4895656429003195,
+      "grad_norm": 3.056596065641448,
+      "learning_rate": 1.0827359071490845e-06,
+      "loss": 0.5766,
+      "step": 1063
+    },
+    {
+      "epoch": 0.49002619383437435,
+      "grad_norm": 3.3438308914593686,
+      "learning_rate": 1.0812484893227688e-06,
+      "loss": 0.5343,
+      "step": 1064
+    },
+    {
+      "epoch": 0.49048674476842924,
+      "grad_norm": 3.2170156199513684,
+      "learning_rate": 1.079760890524715e-06,
+      "loss": 0.4873,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4909472957024841,
+      "grad_norm": 3.3949566731113276,
+      "learning_rate": 1.0782731140683784e-06,
+      "loss": 0.6256,
+      "step": 1066
+    },
+    {
+      "epoch": 0.49140784663653897,
+      "grad_norm": 3.8201701220980815,
+      "learning_rate": 1.0767851632676119e-06,
+      "loss": 0.5645,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4918683975705938,
+      "grad_norm": 3.321842231119507,
+      "learning_rate": 1.0752970414366561e-06,
+      "loss": 0.5429,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4923289485046487,
+      "grad_norm": 3.1614648927797493,
+      "learning_rate": 1.0738087518901326e-06,
+      "loss": 0.578,
+      "step": 1069
+    },
+    {
+      "epoch": 0.49278949943870354,
+      "grad_norm": 3.1744346557606957,
+      "learning_rate": 1.0723202979430364e-06,
+      "loss": 0.5857,
+      "step": 1070
+    },
+    {
+      "epoch": 0.49325005037275843,
+      "grad_norm": 3.0752420415036132,
+      "learning_rate": 1.0708316829107293e-06,
+      "loss": 0.4874,
+      "step": 1071
+    },
+    {
+      "epoch": 0.49371060130681327,
+      "grad_norm": 3.139920581664928,
+      "learning_rate": 1.0693429101089306e-06,
+      "loss": 0.5561,
+      "step": 1072
+    },
+    {
+      "epoch": 0.49417115224086816,
+      "grad_norm": 3.3471269540196786,
+      "learning_rate": 1.0678539828537123e-06,
+      "loss": 0.4969,
+      "step": 1073
+    },
+    {
+      "epoch": 0.494631703174923,
+      "grad_norm": 3.026522963165782,
+      "learning_rate": 1.06636490446149e-06,
+      "loss": 0.5516,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4950922541089779,
+      "grad_norm": 2.803007358195874,
+      "learning_rate": 1.064875678249016e-06,
+      "loss": 0.5154,
+      "step": 1075
+    },
+    {
+      "epoch": 0.49555280504303273,
+      "grad_norm": 3.5445642741425165,
+      "learning_rate": 1.0633863075333712e-06,
+      "loss": 0.6616,
+      "step": 1076
+    },
+    {
+      "epoch": 0.49601335597708757,
+      "grad_norm": 3.3956547555358725,
+      "learning_rate": 1.0618967956319595e-06,
+      "loss": 0.6653,
+      "step": 1077
+    },
+    {
+      "epoch": 0.49647390691114246,
+      "grad_norm": 3.236674834459769,
+      "learning_rate": 1.0604071458624985e-06,
+      "loss": 0.5443,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4969344578451973,
+      "grad_norm": 2.943577612886821,
+      "learning_rate": 1.058917361543013e-06,
+      "loss": 0.6144,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4973950087792522,
+      "grad_norm": 3.789273872300937,
+      "learning_rate": 1.0574274459918279e-06,
+      "loss": 0.5742,
+      "step": 1080
+    },
+    {
+      "epoch": 0.497855559713307,
+      "grad_norm": 3.4217054874372423,
+      "learning_rate": 1.0559374025275595e-06,
+      "loss": 0.5328,
+      "step": 1081
+    },
+    {
+      "epoch": 0.4983161106473619,
+      "grad_norm": 3.1895741231305252,
+      "learning_rate": 1.0544472344691102e-06,
+      "loss": 0.6774,
+      "step": 1082
+    },
+    {
+      "epoch": 0.49877666158141676,
+      "grad_norm": 3.2720121537966347,
+      "learning_rate": 1.0529569451356586e-06,
+      "loss": 0.5828,
+      "step": 1083
+    },
+    {
+      "epoch": 0.49923721251547165,
+      "grad_norm": 3.404120717189837,
+      "learning_rate": 1.051466537846655e-06,
+      "loss": 0.6939,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4996977634495265,
+      "grad_norm": 3.3671005760273323,
+      "learning_rate": 1.049976015921811e-06,
+      "loss": 0.6451,
+      "step": 1085
+    },
+    {
+      "epoch": 0.5001583143835814,
+      "grad_norm": 3.529477019557084,
+      "learning_rate": 1.048485382681094e-06,
+      "loss": 0.7398,
+      "step": 1086
+    },
+    {
+      "epoch": 0.5006188653176362,
+      "grad_norm": 2.985478155983292,
+      "learning_rate": 1.0469946414447196e-06,
+      "loss": 0.4825,
+      "step": 1087
+    },
+    {
+      "epoch": 0.5010794162516911,
+      "grad_norm": 3.257171883150608,
+      "learning_rate": 1.0455037955331447e-06,
+      "loss": 0.4787,
+      "step": 1088
+    },
+    {
+      "epoch": 0.5015399671857459,
+      "grad_norm": 2.9340259115329226,
+      "learning_rate": 1.0440128482670569e-06,
+      "loss": 0.5687,
+      "step": 1089
+    },
+    {
+      "epoch": 0.5020005181198008,
+      "grad_norm": 3.7349893867163773,
+      "learning_rate": 1.0425218029673718e-06,
+      "loss": 0.502,
+      "step": 1090
+    },
+    {
+      "epoch": 0.5024610690538557,
+      "grad_norm": 3.1463328721295714,
+      "learning_rate": 1.0410306629552231e-06,
+      "loss": 0.5196,
+      "step": 1091
+    },
+    {
+      "epoch": 0.5029216199879105,
+      "grad_norm": 3.2092568851256007,
+      "learning_rate": 1.0395394315519541e-06,
+      "loss": 0.5733,
+      "step": 1092
+    },
+    {
+      "epoch": 0.5033821709219654,
+      "grad_norm": 3.079164349910014,
+      "learning_rate": 1.0380481120791136e-06,
+      "loss": 0.5044,
+      "step": 1093
+    },
+    {
+      "epoch": 0.5038427218560203,
+      "grad_norm": 2.701558803955672,
+      "learning_rate": 1.036556707858445e-06,
+      "loss": 0.4957,
+      "step": 1094
+    },
+    {
+      "epoch": 0.5043032727900751,
+      "grad_norm": 3.1632512432778226,
+      "learning_rate": 1.0350652222118807e-06,
+      "loss": 0.5624,
+      "step": 1095
+    },
+    {
+      "epoch": 0.50476382372413,
+      "grad_norm": 2.872004805465866,
+      "learning_rate": 1.0335736584615356e-06,
+      "loss": 0.5077,
+      "step": 1096
+    },
+    {
+      "epoch": 0.5052243746581848,
+      "grad_norm": 3.038715190759613,
+      "learning_rate": 1.0320820199296974e-06,
+      "loss": 0.487,
+      "step": 1097
+    },
+    {
+      "epoch": 0.5056849255922398,
+      "grad_norm": 3.1237869179473594,
+      "learning_rate": 1.0305903099388202e-06,
+      "loss": 0.418,
+      "step": 1098
+    },
+    {
+      "epoch": 0.5061454765262946,
+      "grad_norm": 2.9307402726415757,
+      "learning_rate": 1.0290985318115184e-06,
+      "loss": 0.5496,
+      "step": 1099
+    },
+    {
+      "epoch": 0.5066060274603494,
+      "grad_norm": 3.3590679344076193,
+      "learning_rate": 1.0276066888705574e-06,
+      "loss": 0.6662,
+      "step": 1100
+    },
+    {
+      "epoch": 0.5070665783944043,
+      "grad_norm": 3.0656098416878033,
+      "learning_rate": 1.0261147844388472e-06,
+      "loss": 0.5917,
+      "step": 1101
+    },
+    {
+      "epoch": 0.5075271293284591,
+      "grad_norm": 3.082634007004262,
+      "learning_rate": 1.0246228218394346e-06,
+      "loss": 0.5372,
+      "step": 1102
+    },
+    {
+      "epoch": 0.5079876802625141,
+      "grad_norm": 3.5524616274647567,
+      "learning_rate": 1.023130804395496e-06,
+      "loss": 0.5913,
+      "step": 1103
+    },
+    {
+      "epoch": 0.5084482311965689,
+      "grad_norm": 2.9986863183328696,
+      "learning_rate": 1.0216387354303295e-06,
+      "loss": 0.4397,
+      "step": 1104
+    },
+    {
+      "epoch": 0.5089087821306237,
+      "grad_norm": 2.613344675102279,
+      "learning_rate": 1.0201466182673498e-06,
+      "loss": 0.4783,
+      "step": 1105
+    },
+    {
+      "epoch": 0.5093693330646786,
+      "grad_norm": 3.2217947984483,
+      "learning_rate": 1.0186544562300764e-06,
+      "loss": 0.5062,
+      "step": 1106
+    },
+    {
+      "epoch": 0.5098298839987335,
+      "grad_norm": 2.9440216394401904,
+      "learning_rate": 1.0171622526421304e-06,
+      "loss": 0.6022,
+      "step": 1107
+    },
+    {
+      "epoch": 0.5102904349327884,
+      "grad_norm": 2.7773735259088257,
+      "learning_rate": 1.0156700108272252e-06,
+      "loss": 0.6232,
+      "step": 1108
+    },
+    {
+      "epoch": 0.5107509858668432,
+      "grad_norm": 2.941806533194381,
+      "learning_rate": 1.0141777341091587e-06,
+      "loss": 0.4926,
+      "step": 1109
+    },
+    {
+      "epoch": 0.511211536800898,
+      "grad_norm": 3.548025228777574,
+      "learning_rate": 1.0126854258118074e-06,
+      "loss": 0.5843,
+      "step": 1110
+    },
+    {
+      "epoch": 0.511672087734953,
+      "grad_norm": 3.275863911485487,
+      "learning_rate": 1.011193089259118e-06,
+      "loss": 0.4883,
+      "step": 1111
+    },
+    {
+      "epoch": 0.5121326386690078,
+      "grad_norm": 3.0059327106969653,
+      "learning_rate": 1.009700727775099e-06,
+      "loss": 0.5342,
+      "step": 1112
+    },
+    {
+      "epoch": 0.5125931896030627,
+      "grad_norm": 3.0101617328590278,
+      "learning_rate": 1.008208344683816e-06,
+      "loss": 0.4742,
+      "step": 1113
+    },
+    {
+      "epoch": 0.5130537405371175,
+      "grad_norm": 3.752871827480606,
+      "learning_rate": 1.0067159433093815e-06,
+      "loss": 0.605,
+      "step": 1114
+    },
+    {
+      "epoch": 0.5135142914711723,
+      "grad_norm": 3.3469797895498843,
+      "learning_rate": 1.00522352697595e-06,
+      "loss": 0.681,
+      "step": 1115
+    },
+    {
+      "epoch": 0.5139748424052273,
+      "grad_norm": 2.6263493290364246,
+      "learning_rate": 1.003731099007708e-06,
+      "loss": 0.4897,
+      "step": 1116
+    },
+    {
+      "epoch": 0.5144353933392821,
+      "grad_norm": 2.9725521268454003,
+      "learning_rate": 1.002238662728869e-06,
+      "loss": 0.5346,
+      "step": 1117
+    },
+    {
+      "epoch": 0.514895944273337,
+      "grad_norm": 3.049153105281882,
+      "learning_rate": 1.000746221463664e-06,
+      "loss": 0.5959,
+      "step": 1118
+    },
+    {
+      "epoch": 0.5153564952073918,
+      "grad_norm": 2.7288089533635334,
+      "learning_rate": 9.992537785363361e-07,
+      "loss": 0.4807,
+      "step": 1119
+    },
+    {
+      "epoch": 0.5158170461414467,
+      "grad_norm": 3.0318291312947667,
+      "learning_rate": 9.977613372711308e-07,
+      "loss": 0.5651,
+      "step": 1120
+    },
+    {
+      "epoch": 0.5162775970755016,
+      "grad_norm": 2.6295894746203583,
+      "learning_rate": 9.962689009922918e-07,
+      "loss": 0.5055,
+      "step": 1121
+    },
+    {
+      "epoch": 0.5167381480095564,
+      "grad_norm": 3.1101382519689285,
+      "learning_rate": 9.947764730240501e-07,
+      "loss": 0.5315,
+      "step": 1122
+    },
+    {
+      "epoch": 0.5171986989436113,
+      "grad_norm": 3.019675975546636,
+      "learning_rate": 9.932840566906184e-07,
+      "loss": 0.5095,
+      "step": 1123
+    },
+    {
+      "epoch": 0.5176592498776662,
+      "grad_norm": 2.8010710502034897,
+      "learning_rate": 9.917916553161841e-07,
+      "loss": 0.4929,
+      "step": 1124
+    },
+    {
+      "epoch": 0.518119800811721,
+      "grad_norm": 3.0678462211009436,
+      "learning_rate": 9.90299272224901e-07,
+      "loss": 0.5167,
+      "step": 1125
+    },
+    {
+      "epoch": 0.5185803517457759,
+      "grad_norm": 3.134942537587258,
+      "learning_rate": 9.888069107408824e-07,
+      "loss": 0.5602,
+      "step": 1126
+    },
+    {
+      "epoch": 0.5190409026798307,
+      "grad_norm": 2.971903282768269,
+      "learning_rate": 9.873145741881927e-07,
+      "loss": 0.6232,
+      "step": 1127
+    },
+    {
+      "epoch": 0.5195014536138856,
+      "grad_norm": 3.402808011964365,
+      "learning_rate": 9.858222658908412e-07,
+      "loss": 0.6225,
+      "step": 1128
+    },
+    {
+      "epoch": 0.5199620045479405,
+      "grad_norm": 3.0789459137502644,
+      "learning_rate": 9.84329989172775e-07,
+      "loss": 0.6397,
+      "step": 1129
+    },
+    {
+      "epoch": 0.5204225554819953,
+      "grad_norm": 2.9406870784619428,
+      "learning_rate": 9.828377473578697e-07,
+      "loss": 0.5466,
+      "step": 1130
+    },
+    {
+      "epoch": 0.5208831064160502,
+      "grad_norm": 3.364937021577735,
+      "learning_rate": 9.813455437699237e-07,
+      "loss": 0.6092,
+      "step": 1131
+    },
+    {
+      "epoch": 0.521343657350105,
+      "grad_norm": 2.9557279281094724,
+      "learning_rate": 9.798533817326504e-07,
+      "loss": 0.5889,
+      "step": 1132
+    },
+    {
+      "epoch": 0.52180420828416,
+      "grad_norm": 3.7473302217626654,
+      "learning_rate": 9.783612645696702e-07,
+      "loss": 0.5108,
+      "step": 1133
+    },
+    {
+      "epoch": 0.5222647592182148,
+      "grad_norm": 2.973166241915579,
+      "learning_rate": 9.768691956045042e-07,
+      "loss": 0.5232,
+      "step": 1134
+    },
+    {
+      "epoch": 0.5227253101522696,
+      "grad_norm": 3.21594378369249,
+      "learning_rate": 9.753771781605657e-07,
+      "loss": 0.5568,
+      "step": 1135
+    },
+    {
+      "epoch": 0.5231858610863245,
+      "grad_norm": 3.144947147889769,
+      "learning_rate": 9.73885215561153e-07,
+      "loss": 0.5373,
+      "step": 1136
+    },
+    {
+      "epoch": 0.5236464120203794,
+      "grad_norm": 3.2253884546962137,
+      "learning_rate": 9.723933111294427e-07,
+      "loss": 0.5413,
+      "step": 1137
+    },
+    {
+      "epoch": 0.5241069629544343,
+      "grad_norm": 3.223035357344408,
+      "learning_rate": 9.709014681884815e-07,
+      "loss": 0.4688,
+      "step": 1138
+    },
+    {
+      "epoch": 0.5245675138884891,
+      "grad_norm": 3.4567627154136322,
+      "learning_rate": 9.6940969006118e-07,
+      "loss": 0.6319,
+      "step": 1139
+    },
+    {
+      "epoch": 0.5250280648225439,
+      "grad_norm": 3.2739054341505596,
+      "learning_rate": 9.67917980070303e-07,
+      "loss": 0.6505,
+      "step": 1140
+    },
+    {
+      "epoch": 0.5254886157565988,
+      "grad_norm": 3.0010401161709943,
+      "learning_rate": 9.664263415384643e-07,
+      "loss": 0.4819,
+      "step": 1141
+    },
+    {
+      "epoch": 0.5259491666906537,
+      "grad_norm": 3.213300971817815,
+      "learning_rate": 9.649347777881192e-07,
+      "loss": 0.4948,
+      "step": 1142
+    },
+    {
+      "epoch": 0.5264097176247086,
+      "grad_norm": 3.834117009480357,
+      "learning_rate": 9.634432921415554e-07,
+      "loss": 0.5202,
+      "step": 1143
+    },
+    {
+      "epoch": 0.5268702685587634,
+      "grad_norm": 3.0405631156566524,
+      "learning_rate": 9.619518879208865e-07,
+      "loss": 0.5498,
+      "step": 1144
+    },
+    {
+      "epoch": 0.5273308194928182,
+      "grad_norm": 2.736642170245626,
+      "learning_rate": 9.604605684480458e-07,
+      "loss": 0.5615,
+      "step": 1145
+    },
+    {
+      "epoch": 0.5277913704268732,
+      "grad_norm": 3.4155336627808577,
+      "learning_rate": 9.589693370447768e-07,
+      "loss": 0.6218,
+      "step": 1146
+    },
+    {
+      "epoch": 0.528251921360928,
+      "grad_norm": 3.197761407312923,
+      "learning_rate": 9.574781970326283e-07,
+      "loss": 0.5547,
+      "step": 1147
+    },
+    {
+      "epoch": 0.5287124722949829,
+      "grad_norm": 3.025078798504295,
+      "learning_rate": 9.559871517329434e-07,
+      "loss": 0.4866,
+      "step": 1148
+    },
+    {
+      "epoch": 0.5291730232290377,
+      "grad_norm": 3.2517856164626027,
+      "learning_rate": 9.544962044668555e-07,
+      "loss": 0.5777,
+      "step": 1149
+    },
+    {
+      "epoch": 0.5296335741630926,
+      "grad_norm": 2.996348096347719,
+      "learning_rate": 9.530053585552802e-07,
+      "loss": 0.5648,
+      "step": 1150
+    },
+    {
+      "epoch": 0.5300941250971475,
+      "grad_norm": 3.119156079432281,
+      "learning_rate": 9.515146173189057e-07,
+      "loss": 0.5534,
+      "step": 1151
+    },
+    {
+      "epoch": 0.5305546760312023,
+      "grad_norm": 2.984885291180427,
+      "learning_rate": 9.50023984078189e-07,
+      "loss": 0.4584,
+      "step": 1152
+    },
+    {
+      "epoch": 0.5310152269652572,
+      "grad_norm": 3.2174005544489765,
+      "learning_rate": 9.485334621533453e-07,
+      "loss": 0.4978,
+      "step": 1153
+    },
+    {
+      "epoch": 0.5314757778993121,
+      "grad_norm": 2.838123600964648,
+      "learning_rate": 9.470430548643411e-07,
+      "loss": 0.5271,
+      "step": 1154
+    },
+    {
+      "epoch": 0.531936328833367,
+      "grad_norm": 3.2298666663636766,
+      "learning_rate": 9.455527655308899e-07,
+      "loss": 0.5232,
+      "step": 1155
+    },
+    {
+      "epoch": 0.5323968797674218,
+      "grad_norm": 3.7741680239909963,
+      "learning_rate": 9.440625974724407e-07,
+      "loss": 0.6279,
+      "step": 1156
+    },
+    {
+      "epoch": 0.5328574307014766,
+      "grad_norm": 2.9599505842958873,
+      "learning_rate": 9.425725540081721e-07,
+      "loss": 0.4482,
+      "step": 1157
+    },
+    {
+      "epoch": 0.5333179816355315,
+      "grad_norm": 3.0963847317641533,
+      "learning_rate": 9.410826384569869e-07,
+      "loss": 0.5946,
+      "step": 1158
+    },
+    {
+      "epoch": 0.5337785325695864,
+      "grad_norm": 3.0756573597480825,
+      "learning_rate": 9.395928541375013e-07,
+      "loss": 0.4987,
+      "step": 1159
+    },
+    {
+      "epoch": 0.5342390835036412,
+      "grad_norm": 2.96523606047648,
+      "learning_rate": 9.381032043680405e-07,
+      "loss": 0.5864,
+      "step": 1160
+    },
+    {
+      "epoch": 0.5346996344376961,
+      "grad_norm": 2.8034523910128377,
+      "learning_rate": 9.366136924666288e-07,
+      "loss": 0.6241,
+      "step": 1161
+    },
+    {
+      "epoch": 0.5351601853717509,
+      "grad_norm": 3.0938529680087794,
+      "learning_rate": 9.351243217509842e-07,
+      "loss": 0.613,
+      "step": 1162
+    },
+    {
+      "epoch": 0.5356207363058059,
+      "grad_norm": 3.3153575829450728,
+      "learning_rate": 9.336350955385101e-07,
+      "loss": 0.5514,
+      "step": 1163
+    },
+    {
+      "epoch": 0.5360812872398607,
+      "grad_norm": 3.1688595006317124,
+      "learning_rate": 9.321460171462876e-07,
+      "loss": 0.5431,
+      "step": 1164
+    },
+    {
+      "epoch": 0.5365418381739155,
+      "grad_norm": 3.1725759909285123,
+      "learning_rate": 9.306570898910695e-07,
+      "loss": 0.4865,
+      "step": 1165
+    },
+    {
+      "epoch": 0.5370023891079704,
+      "grad_norm": 2.9853416997922406,
+      "learning_rate": 9.29168317089271e-07,
+      "loss": 0.5308,
+      "step": 1166
+    },
+    {
+      "epoch": 0.5374629400420253,
+      "grad_norm": 3.2282727253852053,
+      "learning_rate": 9.276797020569635e-07,
+      "loss": 0.5703,
+      "step": 1167
+    },
+    {
+      "epoch": 0.5379234909760802,
+      "grad_norm": 2.7961799624127734,
+      "learning_rate": 9.261912481098675e-07,
+      "loss": 0.4944,
+      "step": 1168
+    },
+    {
+      "epoch": 0.538384041910135,
+      "grad_norm": 3.3641211633805126,
+      "learning_rate": 9.24702958563344e-07,
+      "loss": 0.5362,
+      "step": 1169
+    },
+    {
+      "epoch": 0.5388445928441898,
+      "grad_norm": 3.101865055004353,
+      "learning_rate": 9.232148367323882e-07,
+      "loss": 0.6308,
+      "step": 1170
+    },
+    {
+      "epoch": 0.5393051437782447,
+      "grad_norm": 3.189825939308207,
+      "learning_rate": 9.217268859316218e-07,
+      "loss": 0.5148,
+      "step": 1171
+    },
+    {
+      "epoch": 0.5397656947122996,
+      "grad_norm": 3.82610831328444,
+      "learning_rate": 9.202391094752853e-07,
+      "loss": 0.5346,
+      "step": 1172
+    },
+    {
+      "epoch": 0.5402262456463545,
+      "grad_norm": 3.2076777255853672,
+      "learning_rate": 9.187515106772311e-07,
+      "loss": 0.5662,
+      "step": 1173
+    },
+    {
+      "epoch": 0.5406867965804093,
+      "grad_norm": 3.4891483553752343,
+      "learning_rate": 9.172640928509158e-07,
+      "loss": 0.5638,
+      "step": 1174
+    },
+    {
+      "epoch": 0.5411473475144641,
+      "grad_norm": 3.5063754666432865,
+      "learning_rate": 9.157768593093925e-07,
+      "loss": 0.4947,
+      "step": 1175
+    },
+    {
+      "epoch": 0.5416078984485191,
+      "grad_norm": 2.803505276485617,
+      "learning_rate": 9.142898133653047e-07,
+      "loss": 0.5989,
+      "step": 1176
+    },
+    {
+      "epoch": 0.5420684493825739,
+      "grad_norm": 3.4189599088321745,
+      "learning_rate": 9.128029583308773e-07,
+      "loss": 0.5396,
+      "step": 1177
+    },
+    {
+      "epoch": 0.5425290003166288,
+      "grad_norm": 3.4562786815564723,
+      "learning_rate": 9.113162975179104e-07,
+      "loss": 0.5468,
+      "step": 1178
+    },
+    {
+      "epoch": 0.5429895512506836,
+      "grad_norm": 3.1130591716501272,
+      "learning_rate": 9.098298342377711e-07,
+      "loss": 0.6484,
+      "step": 1179
+    },
+    {
+      "epoch": 0.5434501021847385,
+      "grad_norm": 3.218340063155536,
+      "learning_rate": 9.083435718013868e-07,
+      "loss": 0.5841,
+      "step": 1180
+    },
+    {
+      "epoch": 0.5439106531187934,
+      "grad_norm": 2.871969590049317,
+      "learning_rate": 9.068575135192376e-07,
+      "loss": 0.5347,
+      "step": 1181
+    },
+    {
+      "epoch": 0.5443712040528482,
+      "grad_norm": 3.1118275872712906,
+      "learning_rate": 9.053716627013487e-07,
+      "loss": 0.528,
+      "step": 1182
+    },
+    {
+      "epoch": 0.5448317549869031,
+      "grad_norm": 3.262925860586302,
+      "learning_rate": 9.038860226572831e-07,
+      "loss": 0.6184,
+      "step": 1183
+    },
+    {
+      "epoch": 0.5452923059209579,
+      "grad_norm": 3.379255799830584,
+      "learning_rate": 9.024005966961346e-07,
+      "loss": 0.6056,
+      "step": 1184
+    },
+    {
+      "epoch": 0.5457528568550128,
+      "grad_norm": 3.1526864635641103,
+      "learning_rate": 9.009153881265198e-07,
+      "loss": 0.6153,
+      "step": 1185
+    },
+    {
+      "epoch": 0.5462134077890677,
+      "grad_norm": 3.2354021477420547,
+      "learning_rate": 8.994304002565722e-07,
+      "loss": 0.6598,
+      "step": 1186
+    },
+    {
+      "epoch": 0.5466739587231225,
+      "grad_norm": 3.3399627148509077,
+      "learning_rate": 8.979456363939317e-07,
+      "loss": 0.5406,
+      "step": 1187
+    },
+    {
+      "epoch": 0.5471345096571774,
+      "grad_norm": 3.2227858379890106,
+      "learning_rate": 8.964610998457407e-07,
+      "loss": 0.4731,
+      "step": 1188
+    },
+    {
+      "epoch": 0.5475950605912323,
+      "grad_norm": 3.128161422108759,
+      "learning_rate": 8.949767939186356e-07,
+      "loss": 0.5487,
+      "step": 1189
+    },
+    {
+      "epoch": 0.5480556115252871,
+      "grad_norm": 3.2758864860806076,
+      "learning_rate": 8.934927219187373e-07,
+      "loss": 0.5796,
+      "step": 1190
+    },
+    {
+      "epoch": 0.548516162459342,
+      "grad_norm": 3.0508736285456903,
+      "learning_rate": 8.920088871516481e-07,
+      "loss": 0.5156,
+      "step": 1191
+    },
+    {
+      "epoch": 0.5489767133933968,
+      "grad_norm": 3.454863882802425,
+      "learning_rate": 8.905252929224402e-07,
+      "loss": 0.5471,
+      "step": 1192
+    },
+    {
+      "epoch": 0.5494372643274518,
+      "grad_norm": 3.241696417453311,
+      "learning_rate": 8.890419425356495e-07,
+      "loss": 0.6282,
+      "step": 1193
+    },
+    {
+      "epoch": 0.5498978152615066,
+      "grad_norm": 3.168159151579064,
+      "learning_rate": 8.875588392952712e-07,
+      "loss": 0.5145,
+      "step": 1194
+    },
+    {
+      "epoch": 0.5503583661955614,
+      "grad_norm": 3.496337240225602,
+      "learning_rate": 8.860759865047475e-07,
+      "loss": 0.5628,
+      "step": 1195
+    },
+    {
+      "epoch": 0.5508189171296163,
+      "grad_norm": 3.2304073520392165,
+      "learning_rate": 8.845933874669644e-07,
+      "loss": 0.4703,
+      "step": 1196
+    },
+    {
+      "epoch": 0.5512794680636711,
+      "grad_norm": 3.1834073949679214,
+      "learning_rate": 8.831110454842418e-07,
+      "loss": 0.5432,
+      "step": 1197
+    },
+    {
+      "epoch": 0.5517400189977261,
+      "grad_norm": 3.4899372677451823,
+      "learning_rate": 8.816289638583272e-07,
+      "loss": 0.6382,
+      "step": 1198
+    },
+    {
+      "epoch": 0.5522005699317809,
+      "grad_norm": 2.9457383672792754,
+      "learning_rate": 8.801471458903885e-07,
+      "loss": 0.4662,
+      "step": 1199
+    },
+    {
+      "epoch": 0.5526611208658357,
+      "grad_norm": 3.2935543447014104,
+      "learning_rate": 8.786655948810062e-07,
+      "loss": 0.4872,
+      "step": 1200
+    },
+    {
+      "epoch": 0.5531216717998906,
+      "grad_norm": 3.2241377818141057,
+      "learning_rate": 8.771843141301658e-07,
+      "loss": 0.568,
+      "step": 1201
+    },
+    {
+      "epoch": 0.5535822227339455,
+      "grad_norm": 3.1610954571795853,
+      "learning_rate": 8.757033069372514e-07,
+      "loss": 0.5381,
+      "step": 1202
+    },
+    {
+      "epoch": 0.5540427736680004,
+      "grad_norm": 2.8053997458619193,
+      "learning_rate": 8.742225766010375e-07,
+      "loss": 0.6084,
+      "step": 1203
+    },
+    {
+      "epoch": 0.5545033246020552,
+      "grad_norm": 3.2358176609289977,
+      "learning_rate": 8.727421264196819e-07,
+      "loss": 0.5904,
+      "step": 1204
+    },
+    {
+      "epoch": 0.55496387553611,
+      "grad_norm": 3.0926449818062127,
+      "learning_rate": 8.712619596907187e-07,
+      "loss": 0.5746,
+      "step": 1205
+    },
+    {
+      "epoch": 0.555424426470165,
+      "grad_norm": 2.941504999556666,
+      "learning_rate": 8.697820797110498e-07,
+      "loss": 0.5781,
+      "step": 1206
+    },
+    {
+      "epoch": 0.5558849774042198,
+      "grad_norm": 2.871854054187517,
+      "learning_rate": 8.683024897769395e-07,
+      "loss": 0.4819,
+      "step": 1207
+    },
+    {
+      "epoch": 0.5563455283382747,
+      "grad_norm": 3.357101022208657,
+      "learning_rate": 8.668231931840053e-07,
+      "loss": 0.5112,
+      "step": 1208
+    },
+    {
+      "epoch": 0.5568060792723295,
+      "grad_norm": 2.926617988170589,
+      "learning_rate": 8.653441932272118e-07,
+      "loss": 0.5429,
+      "step": 1209
+    },
+    {
+      "epoch": 0.5572666302063843,
+      "grad_norm": 2.8511154209063703,
+      "learning_rate": 8.638654932008626e-07,
+      "loss": 0.589,
+      "step": 1210
+    },
+    {
+      "epoch": 0.5577271811404393,
+      "grad_norm": 3.0744117053038984,
+      "learning_rate": 8.623870963985929e-07,
+      "loss": 0.4968,
+      "step": 1211
+    },
+    {
+      "epoch": 0.5581877320744941,
+      "grad_norm": 3.2017070331025663,
+      "learning_rate": 8.609090061133633e-07,
+      "loss": 0.483,
+      "step": 1212
+    },
+    {
+      "epoch": 0.558648283008549,
+      "grad_norm": 3.4242826485145024,
+      "learning_rate": 8.594312256374512e-07,
+      "loss": 0.6987,
+      "step": 1213
+    },
+    {
+      "epoch": 0.5591088339426038,
+      "grad_norm": 3.215000736037071,
+      "learning_rate": 8.579537582624437e-07,
+      "loss": 0.569,
+      "step": 1214
+    },
+    {
+      "epoch": 0.5595693848766587,
+      "grad_norm": 3.21779173610873,
+      "learning_rate": 8.564766072792311e-07,
+      "loss": 0.5391,
+      "step": 1215
+    },
+    {
+      "epoch": 0.5600299358107136,
+      "grad_norm": 3.131067041749219,
+      "learning_rate": 8.54999775977998e-07,
+      "loss": 0.5869,
+      "step": 1216
+    },
+    {
+      "epoch": 0.5604904867447684,
+      "grad_norm": 3.4662747526743347,
+      "learning_rate": 8.535232676482189e-07,
+      "loss": 0.5868,
+      "step": 1217
+    },
+    {
+      "epoch": 0.5609510376788233,
+      "grad_norm": 3.428017059923847,
+      "learning_rate": 8.520470855786466e-07,
+      "loss": 0.6251,
+      "step": 1218
+    },
+    {
+      "epoch": 0.5614115886128782,
+      "grad_norm": 3.3290755200586872,
+      "learning_rate": 8.505712330573079e-07,
+      "loss": 0.6237,
+      "step": 1219
+    },
+    {
+      "epoch": 0.561872139546933,
+      "grad_norm": 2.9893892056090734,
+      "learning_rate": 8.490957133714973e-07,
+      "loss": 0.4902,
+      "step": 1220
+    },
+    {
+      "epoch": 0.5623326904809879,
+      "grad_norm": 3.5120837728765664,
+      "learning_rate": 8.476205298077649e-07,
+      "loss": 0.6707,
+      "step": 1221
+    },
+    {
+      "epoch": 0.5627932414150427,
+      "grad_norm": 3.1658388926282517,
+      "learning_rate": 8.46145685651915e-07,
+      "loss": 0.514,
+      "step": 1222
+    },
+    {
+      "epoch": 0.5632537923490976,
+      "grad_norm": 3.117691401780372,
+      "learning_rate": 8.446711841889945e-07,
+      "loss": 0.6758,
+      "step": 1223
+    },
+    {
+      "epoch": 0.5637143432831525,
+      "grad_norm": 3.000516000189048,
+      "learning_rate": 8.431970287032861e-07,
+      "loss": 0.5931,
+      "step": 1224
+    },
+    {
+      "epoch": 0.5641748942172073,
+      "grad_norm": 2.796362177103071,
+      "learning_rate": 8.417232224783041e-07,
+      "loss": 0.475,
+      "step": 1225
+    },
+    {
+      "epoch": 0.5646354451512622,
+      "grad_norm": 3.1187898436905743,
+      "learning_rate": 8.402497687967836e-07,
+      "loss": 0.4477,
+      "step": 1226
+    },
+    {
+      "epoch": 0.565095996085317,
+      "grad_norm": 3.7014134238719962,
+      "learning_rate": 8.387766709406735e-07,
+      "loss": 0.5918,
+      "step": 1227
+    },
+    {
+      "epoch": 0.565556547019372,
+      "grad_norm": 3.8201360952855854,
+      "learning_rate": 8.373039321911323e-07,
+      "loss": 0.6437,
+      "step": 1228
+    },
+    {
+      "epoch": 0.5660170979534268,
+      "grad_norm": 3.4961060613658907,
+      "learning_rate": 8.358315558285169e-07,
+      "loss": 0.6191,
+      "step": 1229
+    },
+    {
+      "epoch": 0.5664776488874816,
+      "grad_norm": 3.374429880061983,
+      "learning_rate": 8.343595451323781e-07,
+      "loss": 0.5717,
+      "step": 1230
+    },
+    {
+      "epoch": 0.5669381998215365,
+      "grad_norm": 3.5990525265114863,
+      "learning_rate": 8.328879033814515e-07,
+      "loss": 0.7007,
+      "step": 1231
+    },
+    {
+      "epoch": 0.5673987507555914,
+      "grad_norm": 3.43846448225756,
+      "learning_rate": 8.31416633853651e-07,
+      "loss": 0.5106,
+      "step": 1232
+    },
+    {
+      "epoch": 0.5678593016896463,
+      "grad_norm": 3.240170428540241,
+      "learning_rate": 8.29945739826062e-07,
+      "loss": 0.6552,
+      "step": 1233
+    },
+    {
+      "epoch": 0.5683198526237011,
+      "grad_norm": 3.268819388844623,
+      "learning_rate": 8.284752245749327e-07,
+      "loss": 0.563,
+      "step": 1234
+    },
+    {
+      "epoch": 0.5687804035577559,
+      "grad_norm": 3.322362445609891,
+      "learning_rate": 8.270050913756683e-07,
+      "loss": 0.6392,
+      "step": 1235
+    },
+    {
+      "epoch": 0.5692409544918108,
+      "grad_norm": 3.4868505757465855,
+      "learning_rate": 8.255353435028226e-07,
+      "loss": 0.5105,
+      "step": 1236
+    },
+    {
+      "epoch": 0.5697015054258657,
+      "grad_norm": 3.7205750052168556,
+      "learning_rate": 8.240659842300912e-07,
+      "loss": 0.555,
+      "step": 1237
+    },
+    {
+      "epoch": 0.5701620563599206,
+      "grad_norm": 3.3997467433257222,
+      "learning_rate": 8.225970168303045e-07,
+      "loss": 0.5784,
+      "step": 1238
+    },
+    {
+      "epoch": 0.5706226072939754,
+      "grad_norm": 3.666481631082462,
+      "learning_rate": 8.211284445754197e-07,
+      "loss": 0.6415,
+      "step": 1239
+    },
+    {
+      "epoch": 0.5710831582280302,
+      "grad_norm": 3.030350707526125,
+      "learning_rate": 8.196602707365134e-07,
+      "loss": 0.5301,
+      "step": 1240
+    },
+    {
+      "epoch": 0.5715437091620852,
+      "grad_norm": 3.4386931144286104,
+      "learning_rate": 8.18192498583776e-07,
+      "loss": 0.4771,
+      "step": 1241
+    },
+    {
+      "epoch": 0.57200426009614,
+      "grad_norm": 3.6879705378930017,
+      "learning_rate": 8.16725131386502e-07,
+      "loss": 0.4923,
+      "step": 1242
+    },
+    {
+      "epoch": 0.5724648110301949,
+      "grad_norm": 3.258657524610071,
+      "learning_rate": 8.152581724130849e-07,
+      "loss": 0.4746,
+      "step": 1243
+    },
+    {
+      "epoch": 0.5729253619642497,
+      "grad_norm": 3.207585199421819,
+      "learning_rate": 8.13791624931008e-07,
+      "loss": 0.6232,
+      "step": 1244
+    },
+    {
+      "epoch": 0.5733859128983047,
+      "grad_norm": 3.1524462601648535,
+      "learning_rate": 8.123254922068383e-07,
+      "loss": 0.6256,
+      "step": 1245
+    },
+    {
+      "epoch": 0.5738464638323595,
+      "grad_norm": 2.829223353488783,
+      "learning_rate": 8.108597775062199e-07,
+      "loss": 0.5638,
+      "step": 1246
+    },
+    {
+      "epoch": 0.5743070147664143,
+      "grad_norm": 2.991108516511504,
+      "learning_rate": 8.093944840938638e-07,
+      "loss": 0.528,
+      "step": 1247
+    },
+    {
+      "epoch": 0.5747675657004692,
+      "grad_norm": 3.122195079526398,
+      "learning_rate": 8.079296152335454e-07,
+      "loss": 0.5076,
+      "step": 1248
+    },
+    {
+      "epoch": 0.575228116634524,
+      "grad_norm": 2.8771471658342227,
+      "learning_rate": 8.06465174188092e-07,
+      "loss": 0.4779,
+      "step": 1249
+    },
+    {
+      "epoch": 0.575688667568579,
+      "grad_norm": 2.9430839805871236,
+      "learning_rate": 8.050011642193787e-07,
+      "loss": 0.5621,
+      "step": 1250
+    },
+    {
+      "epoch": 0.5761492185026338,
+      "grad_norm": 3.2638073071160147,
+      "learning_rate": 8.035375885883217e-07,
+      "loss": 0.581,
+      "step": 1251
+    },
+    {
+      "epoch": 0.5766097694366886,
+      "grad_norm": 2.8037502705877397,
+      "learning_rate": 8.020744505548678e-07,
+      "loss": 0.4642,
+      "step": 1252
+    },
+    {
+      "epoch": 0.5770703203707435,
+      "grad_norm": 3.009407280040476,
+      "learning_rate": 8.006117533779897e-07,
+      "loss": 0.5578,
+      "step": 1253
+    },
+    {
+      "epoch": 0.5775308713047984,
+      "grad_norm": 2.7699074199330163,
+      "learning_rate": 7.991495003156799e-07,
+      "loss": 0.548,
+      "step": 1254
+    },
+    {
+      "epoch": 0.5779914222388532,
+      "grad_norm": 3.105688826369827,
+      "learning_rate": 7.976876946249385e-07,
+      "loss": 0.4663,
+      "step": 1255
+    },
+    {
+      "epoch": 0.5784519731729081,
+      "grad_norm": 3.111402743097655,
+      "learning_rate": 7.962263395617723e-07,
+      "loss": 0.6285,
+      "step": 1256
+    },
+    {
+      "epoch": 0.5789125241069629,
+      "grad_norm": 2.740304619765724,
+      "learning_rate": 7.947654383811826e-07,
+      "loss": 0.6369,
+      "step": 1257
+    },
+    {
+      "epoch": 0.5793730750410179,
+      "grad_norm": 3.321527884757525,
+      "learning_rate": 7.933049943371591e-07,
+      "loss": 0.5833,
+      "step": 1258
+    },
+    {
+      "epoch": 0.5798336259750727,
+      "grad_norm": 3.3544814320822325,
+      "learning_rate": 7.918450106826756e-07,
+      "loss": 0.4738,
+      "step": 1259
+    },
+    {
+      "epoch": 0.5802941769091275,
+      "grad_norm": 2.8103939697101104,
+      "learning_rate": 7.903854906696783e-07,
+      "loss": 0.4384,
+      "step": 1260
+    },
+    {
+      "epoch": 0.5807547278431824,
+      "grad_norm": 2.9663453333883867,
+      "learning_rate": 7.889264375490819e-07,
+      "loss": 0.5188,
+      "step": 1261
+    },
+    {
+      "epoch": 0.5812152787772372,
+      "grad_norm": 3.2684264262146296,
+      "learning_rate": 7.874678545707605e-07,
+      "loss": 0.4919,
+      "step": 1262
+    },
+    {
+      "epoch": 0.5816758297112922,
+      "grad_norm": 3.2619536382755183,
+      "learning_rate": 7.86009744983541e-07,
+      "loss": 0.5307,
+      "step": 1263
+    },
+    {
+      "epoch": 0.582136380645347,
+      "grad_norm": 3.4192324602148094,
+      "learning_rate": 7.845521120351967e-07,
+      "loss": 0.5289,
+      "step": 1264
+    },
+    {
+      "epoch": 0.5825969315794018,
+      "grad_norm": 3.359910314026855,
+      "learning_rate": 7.830949589724381e-07,
+      "loss": 0.6354,
+      "step": 1265
+    },
+    {
+      "epoch": 0.5830574825134567,
+      "grad_norm": 2.824340063484425,
+      "learning_rate": 7.816382890409079e-07,
+      "loss": 0.4714,
+      "step": 1266
+    },
+    {
+      "epoch": 0.5835180334475116,
+      "grad_norm": 3.0782272088831473,
+      "learning_rate": 7.80182105485172e-07,
+      "loss": 0.606,
+      "step": 1267
+    },
+    {
+      "epoch": 0.5839785843815665,
+      "grad_norm": 3.4682328165851817,
+      "learning_rate": 7.787264115487125e-07,
+      "loss": 0.6244,
+      "step": 1268
+    },
+    {
+      "epoch": 0.5844391353156213,
+      "grad_norm": 3.094844298682695,
+      "learning_rate": 7.772712104739225e-07,
+      "loss": 0.6137,
+      "step": 1269
+    },
+    {
+      "epoch": 0.5848996862496761,
+      "grad_norm": 3.1487313141421818,
+      "learning_rate": 7.758165055020959e-07,
+      "loss": 0.4985,
+      "step": 1270
+    },
+    {
+      "epoch": 0.5853602371837311,
+      "grad_norm": 3.0555631645161974,
+      "learning_rate": 7.743622998734216e-07,
+      "loss": 0.5003,
+      "step": 1271
+    },
+    {
+      "epoch": 0.5858207881177859,
+      "grad_norm": 2.89659469677659,
+      "learning_rate": 7.729085968269775e-07,
+      "loss": 0.5633,
+      "step": 1272
+    },
+    {
+      "epoch": 0.5862813390518408,
+      "grad_norm": 3.0731709449259954,
+      "learning_rate": 7.714553996007207e-07,
+      "loss": 0.6242,
+      "step": 1273
+    },
+    {
+      "epoch": 0.5867418899858956,
+      "grad_norm": 3.5726138177591227,
+      "learning_rate": 7.700027114314824e-07,
+      "loss": 0.5221,
+      "step": 1274
+    },
+    {
+      "epoch": 0.5872024409199504,
+      "grad_norm": 2.7589540594281448,
+      "learning_rate": 7.685505355549599e-07,
+      "loss": 0.5246,
+      "step": 1275
+    },
+    {
+      "epoch": 0.5876629918540054,
+      "grad_norm": 2.7404624681182486,
+      "learning_rate": 7.670988752057087e-07,
+      "loss": 0.4816,
+      "step": 1276
+    },
+    {
+      "epoch": 0.5881235427880602,
+      "grad_norm": 3.6053355624856906,
+      "learning_rate": 7.656477336171372e-07,
+      "loss": 0.5095,
+      "step": 1277
+    },
+    {
+      "epoch": 0.5885840937221151,
+      "grad_norm": 3.127209652870214,
+      "learning_rate": 7.64197114021497e-07,
+      "loss": 0.5685,
+      "step": 1278
+    },
+    {
+      "epoch": 0.5890446446561699,
+      "grad_norm": 3.3101979407630506,
+      "learning_rate": 7.627470196498788e-07,
+      "loss": 0.6193,
+      "step": 1279
+    },
+    {
+      "epoch": 0.5895051955902249,
+      "grad_norm": 3.0195380063665938,
+      "learning_rate": 7.612974537322015e-07,
+      "loss": 0.6205,
+      "step": 1280
+    },
+    {
+      "epoch": 0.5899657465242797,
+      "grad_norm": 3.130560991980397,
+      "learning_rate": 7.598484194972076e-07,
+      "loss": 0.5309,
+      "step": 1281
+    },
+    {
+      "epoch": 0.5904262974583345,
+      "grad_norm": 2.875522406498749,
+      "learning_rate": 7.583999201724565e-07,
+      "loss": 0.5542,
+      "step": 1282
+    },
+    {
+      "epoch": 0.5908868483923894,
+      "grad_norm": 3.0420530994023336,
+      "learning_rate": 7.569519589843144e-07,
+      "loss": 0.5288,
+      "step": 1283
+    },
+    {
+      "epoch": 0.5913473993264443,
+      "grad_norm": 2.9257495629106565,
+      "learning_rate": 7.555045391579492e-07,
+      "loss": 0.5173,
+      "step": 1284
+    },
+    {
+      "epoch": 0.5918079502604991,
+      "grad_norm": 3.0015108226030085,
+      "learning_rate": 7.540576639173247e-07,
+      "loss": 0.4839,
+      "step": 1285
+    },
+    {
+      "epoch": 0.592268501194554,
+      "grad_norm": 2.7862911331282447,
+      "learning_rate": 7.526113364851891e-07,
+      "loss": 0.5727,
+      "step": 1286
+    },
+    {
+      "epoch": 0.5927290521286088,
+      "grad_norm": 3.101011663748682,
+      "learning_rate": 7.511655600830727e-07,
+      "loss": 0.4828,
+      "step": 1287
+    },
+    {
+      "epoch": 0.5931896030626637,
+      "grad_norm": 2.8268391388305663,
+      "learning_rate": 7.497203379312771e-07,
+      "loss": 0.471,
+      "step": 1288
+    },
+    {
+      "epoch": 0.5936501539967186,
+      "grad_norm": 2.9519498437477503,
+      "learning_rate": 7.482756732488691e-07,
+      "loss": 0.5866,
+      "step": 1289
+    },
+    {
+      "epoch": 0.5941107049307734,
+      "grad_norm": 2.8204629128276255,
+      "learning_rate": 7.468315692536755e-07,
+      "loss": 0.5342,
+      "step": 1290
+    },
+    {
+      "epoch": 0.5945712558648283,
+      "grad_norm": 3.1769123313998913,
+      "learning_rate": 7.453880291622725e-07,
+      "loss": 0.4381,
+      "step": 1291
+    },
+    {
+      "epoch": 0.5950318067988831,
+      "grad_norm": 3.014532693904851,
+      "learning_rate": 7.439450561899813e-07,
+      "loss": 0.4928,
+      "step": 1292
+    },
+    {
+      "epoch": 0.5954923577329381,
+      "grad_norm": 3.0757999291111813,
+      "learning_rate": 7.425026535508593e-07,
+      "loss": 0.5666,
+      "step": 1293
+    },
+    {
+      "epoch": 0.5959529086669929,
+      "grad_norm": 3.3979586211632555,
+      "learning_rate": 7.410608244576937e-07,
+      "loss": 0.5643,
+      "step": 1294
+    },
+    {
+      "epoch": 0.5964134596010477,
+      "grad_norm": 3.259635269308452,
+      "learning_rate": 7.396195721219945e-07,
+      "loss": 0.5211,
+      "step": 1295
+    },
+    {
+      "epoch": 0.5968740105351026,
+      "grad_norm": 2.9795483102185014,
+      "learning_rate": 7.381788997539868e-07,
+      "loss": 0.5127,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5973345614691575,
+      "grad_norm": 3.296329788058233,
+      "learning_rate": 7.367388105626036e-07,
+      "loss": 0.5219,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5977951124032124,
+      "grad_norm": 3.2181878786157205,
+      "learning_rate": 7.352993077554798e-07,
+      "loss": 0.4597,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5982556633372672,
+      "grad_norm": 2.952432694766975,
+      "learning_rate": 7.33860394538943e-07,
+      "loss": 0.5352,
+      "step": 1299
+    },
+    {
+      "epoch": 0.598716214271322,
+      "grad_norm": 2.9260033657087505,
+      "learning_rate": 7.324220741180088e-07,
+      "loss": 0.5421,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5991767652053769,
+      "grad_norm": 3.211788941840772,
+      "learning_rate": 7.309843496963715e-07,
+      "loss": 0.5918,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5996373161394318,
+      "grad_norm": 2.6095863519936033,
+      "learning_rate": 7.295472244763981e-07,
+      "loss": 0.4312,
+      "step": 1302
+    },
+    {
+      "epoch": 0.6000978670734867,
+      "grad_norm": 2.8768992985690374,
+      "learning_rate": 7.281107016591213e-07,
+      "loss": 0.4502,
+      "step": 1303
+    },
+    {
+      "epoch": 0.6005584180075415,
+      "grad_norm": 3.0788536260420947,
+      "learning_rate": 7.266747844442315e-07,
+      "loss": 0.6204,
+      "step": 1304
+    },
+    {
+      "epoch": 0.6010189689415963,
+      "grad_norm": 3.326645359641706,
+      "learning_rate": 7.252394760300707e-07,
+      "loss": 0.6226,
+      "step": 1305
+    },
+    {
+      "epoch": 0.6014795198756513,
+      "grad_norm": 3.211498413419631,
+      "learning_rate": 7.238047796136246e-07,
+      "loss": 0.5597,
+      "step": 1306
+    },
+    {
+      "epoch": 0.6019400708097061,
+      "grad_norm": 3.2233758236272863,
+      "learning_rate": 7.223706983905153e-07,
+      "loss": 0.5445,
+      "step": 1307
+    },
+    {
+      "epoch": 0.602400621743761,
+      "grad_norm": 3.47488384908334,
+      "learning_rate": 7.209372355549956e-07,
+      "loss": 0.5055,
+      "step": 1308
+    },
+    {
+      "epoch": 0.6028611726778158,
+      "grad_norm": 3.2688153234505664,
+      "learning_rate": 7.195043942999404e-07,
+      "loss": 0.6033,
+      "step": 1309
+    },
+    {
+      "epoch": 0.6033217236118708,
+      "grad_norm": 2.95887248920406,
+      "learning_rate": 7.180721778168397e-07,
+      "loss": 0.4693,
+      "step": 1310
+    },
+    {
+      "epoch": 0.6037822745459256,
+      "grad_norm": 3.132075461669666,
+      "learning_rate": 7.166405892957925e-07,
+      "loss": 0.602,
+      "step": 1311
+    },
+    {
+      "epoch": 0.6042428254799804,
+      "grad_norm": 3.177087477555439,
+      "learning_rate": 7.152096319254988e-07,
+      "loss": 0.5589,
+      "step": 1312
+    },
+    {
+      "epoch": 0.6047033764140353,
+      "grad_norm": 3.189488580952695,
+      "learning_rate": 7.137793088932533e-07,
+      "loss": 0.6641,
+      "step": 1313
+    },
+    {
+      "epoch": 0.6051639273480901,
+      "grad_norm": 2.9722614592308263,
+      "learning_rate": 7.123496233849367e-07,
+      "loss": 0.5488,
+      "step": 1314
+    },
+    {
+      "epoch": 0.605624478282145,
+      "grad_norm": 2.7544767985522443,
+      "learning_rate": 7.109205785850106e-07,
+      "loss": 0.4281,
+      "step": 1315
+    },
+    {
+      "epoch": 0.6060850292161999,
+      "grad_norm": 3.3634419751440388,
+      "learning_rate": 7.094921776765094e-07,
+      "loss": 0.509,
+      "step": 1316
+    },
+    {
+      "epoch": 0.6065455801502547,
+      "grad_norm": 2.9636798760844845,
+      "learning_rate": 7.080644238410325e-07,
+      "loss": 0.5577,
+      "step": 1317
+    },
+    {
+      "epoch": 0.6070061310843096,
+      "grad_norm": 3.4072070822619063,
+      "learning_rate": 7.066373202587397e-07,
+      "loss": 0.6024,
+      "step": 1318
+    },
+    {
+      "epoch": 0.6074666820183645,
+      "grad_norm": 3.107976483566616,
+      "learning_rate": 7.052108701083407e-07,
+      "loss": 0.5571,
+      "step": 1319
+    },
+    {
+      "epoch": 0.6079272329524193,
+      "grad_norm": 2.965152268025164,
+      "learning_rate": 7.0378507656709e-07,
+      "loss": 0.5274,
+      "step": 1320
+    },
+    {
+      "epoch": 0.6083877838864742,
+      "grad_norm": 2.910489708230388,
+      "learning_rate": 7.023599428107814e-07,
+      "loss": 0.4719,
+      "step": 1321
+    },
+    {
+      "epoch": 0.608848334820529,
+      "grad_norm": 3.1173838425844553,
+      "learning_rate": 7.009354720137364e-07,
+      "loss": 0.5094,
+      "step": 1322
+    },
+    {
+      "epoch": 0.609308885754584,
+      "grad_norm": 2.797005334303556,
+      "learning_rate": 6.995116673488014e-07,
+      "loss": 0.6097,
+      "step": 1323
+    },
+    {
+      "epoch": 0.6097694366886388,
+      "grad_norm": 3.0227555124071928,
+      "learning_rate": 6.980885319873397e-07,
+      "loss": 0.5937,
+      "step": 1324
+    },
+    {
+      "epoch": 0.6102299876226936,
+      "grad_norm": 2.777519608456516,
+      "learning_rate": 6.966660690992214e-07,
+      "loss": 0.4724,
+      "step": 1325
+    },
+    {
+      "epoch": 0.6106905385567485,
+      "grad_norm": 3.095622108896745,
+      "learning_rate": 6.952442818528219e-07,
+      "loss": 0.5058,
+      "step": 1326
+    },
+    {
+      "epoch": 0.6111510894908033,
+      "grad_norm": 3.334941539522611,
+      "learning_rate": 6.938231734150093e-07,
+      "loss": 0.6349,
+      "step": 1327
+    },
+    {
+      "epoch": 0.6116116404248583,
+      "grad_norm": 3.059763555645584,
+      "learning_rate": 6.9240274695114e-07,
+      "loss": 0.4978,
+      "step": 1328
+    },
+    {
+      "epoch": 0.6120721913589131,
+      "grad_norm": 3.115604189349727,
+      "learning_rate": 6.909830056250526e-07,
+      "loss": 0.6158,
+      "step": 1329
+    },
+    {
+      "epoch": 0.612532742292968,
+      "grad_norm": 3.5110850133149913,
+      "learning_rate": 6.895639525990586e-07,
+      "loss": 0.6059,
+      "step": 1330
+    },
+    {
+      "epoch": 0.6129932932270228,
+      "grad_norm": 3.1973315575132717,
+      "learning_rate": 6.881455910339369e-07,
+      "loss": 0.5804,
+      "step": 1331
+    },
+    {
+      "epoch": 0.6134538441610777,
+      "grad_norm": 3.2520702824243335,
+      "learning_rate": 6.867279240889259e-07,
+      "loss": 0.5976,
+      "step": 1332
+    },
+    {
+      "epoch": 0.6139143950951326,
+      "grad_norm": 3.447876365367868,
+      "learning_rate": 6.853109549217166e-07,
+      "loss": 0.6177,
+      "step": 1333
+    },
+    {
+      "epoch": 0.6143749460291874,
+      "grad_norm": 3.1913415774468676,
+      "learning_rate": 6.838946866884467e-07,
+      "loss": 0.6005,
+      "step": 1334
+    },
+    {
+      "epoch": 0.6148354969632422,
+      "grad_norm": 3.445057324348534,
+      "learning_rate": 6.824791225436918e-07,
+      "loss": 0.5425,
+      "step": 1335
+    },
+    {
+      "epoch": 0.6152960478972972,
+      "grad_norm": 3.7985479498424786,
+      "learning_rate": 6.810642656404596e-07,
+      "loss": 0.5883,
+      "step": 1336
+    },
+    {
+      "epoch": 0.615756598831352,
+      "grad_norm": 2.9721621801480067,
+      "learning_rate": 6.796501191301824e-07,
+      "loss": 0.5032,
+      "step": 1337
+    },
+    {
+      "epoch": 0.6162171497654069,
+      "grad_norm": 2.709696747780453,
+      "learning_rate": 6.782366861627101e-07,
+      "loss": 0.408,
+      "step": 1338
+    },
+    {
+      "epoch": 0.6166777006994617,
+      "grad_norm": 2.797272959212533,
+      "learning_rate": 6.768239698863033e-07,
+      "loss": 0.4895,
+      "step": 1339
+    },
+    {
+      "epoch": 0.6171382516335165,
+      "grad_norm": 2.687895200110118,
+      "learning_rate": 6.754119734476266e-07,
+      "loss": 0.4482,
+      "step": 1340
+    },
+    {
+      "epoch": 0.6175988025675715,
+      "grad_norm": 3.0721198321588887,
+      "learning_rate": 6.740006999917405e-07,
+      "loss": 0.6072,
+      "step": 1341
+    },
+    {
+      "epoch": 0.6180593535016263,
+      "grad_norm": 2.9971206089309823,
+      "learning_rate": 6.725901526620959e-07,
+      "loss": 0.596,
+      "step": 1342
+    },
+    {
+      "epoch": 0.6185199044356812,
+      "grad_norm": 2.9204887920199987,
+      "learning_rate": 6.711803346005258e-07,
+      "loss": 0.4545,
+      "step": 1343
+    },
+    {
+      "epoch": 0.618980455369736,
+      "grad_norm": 3.208202279508134,
+      "learning_rate": 6.697712489472395e-07,
+      "loss": 0.6018,
+      "step": 1344
+    },
+    {
+      "epoch": 0.619441006303791,
+      "grad_norm": 2.698911787184527,
+      "learning_rate": 6.68362898840814e-07,
+      "loss": 0.5294,
+      "step": 1345
+    },
+    {
+      "epoch": 0.6199015572378458,
+      "grad_norm": 2.628724441436368,
+      "learning_rate": 6.669552874181888e-07,
+      "loss": 0.4935,
+      "step": 1346
+    },
+    {
+      "epoch": 0.6203621081719006,
+      "grad_norm": 3.6333914395043996,
+      "learning_rate": 6.655484178146576e-07,
+      "loss": 0.5121,
+      "step": 1347
+    },
+    {
+      "epoch": 0.6208226591059555,
+      "grad_norm": 2.840757981463247,
+      "learning_rate": 6.641422931638614e-07,
+      "loss": 0.5378,
+      "step": 1348
+    },
+    {
+      "epoch": 0.6212832100400104,
+      "grad_norm": 3.119838224826943,
+      "learning_rate": 6.627369165977837e-07,
+      "loss": 0.5759,
+      "step": 1349
+    },
+    {
+      "epoch": 0.6217437609740653,
+      "grad_norm": 2.854001213576483,
+      "learning_rate": 6.613322912467392e-07,
+      "loss": 0.4695,
+      "step": 1350
+    },
+    {
+      "epoch": 0.6222043119081201,
+      "grad_norm": 3.0516655650363695,
+      "learning_rate": 6.599284202393708e-07,
+      "loss": 0.5992,
+      "step": 1351
+    },
+    {
+      "epoch": 0.6226648628421749,
+      "grad_norm": 3.1765489105582927,
+      "learning_rate": 6.585253067026417e-07,
+      "loss": 0.5748,
+      "step": 1352
+    },
+    {
+      "epoch": 0.6231254137762298,
+      "grad_norm": 3.012215519926939,
+      "learning_rate": 6.571229537618266e-07,
+      "loss": 0.5621,
+      "step": 1353
+    },
+    {
+      "epoch": 0.6235859647102847,
+      "grad_norm": 3.277543278020275,
+      "learning_rate": 6.557213645405064e-07,
+      "loss": 0.4244,
+      "step": 1354
+    },
+    {
+      "epoch": 0.6240465156443396,
+      "grad_norm": 3.070612553633971,
+      "learning_rate": 6.54320542160562e-07,
+      "loss": 0.6149,
+      "step": 1355
+    },
+    {
+      "epoch": 0.6245070665783944,
+      "grad_norm": 3.3615068454127046,
+      "learning_rate": 6.529204897421643e-07,
+      "loss": 0.5554,
+      "step": 1356
+    },
+    {
+      "epoch": 0.6249676175124492,
+      "grad_norm": 3.3561738697585266,
+      "learning_rate": 6.515212104037713e-07,
+      "loss": 0.5889,
+      "step": 1357
+    },
+    {
+      "epoch": 0.6254281684465042,
+      "grad_norm": 3.0933916171873483,
+      "learning_rate": 6.50122707262118e-07,
+      "loss": 0.4735,
+      "step": 1358
+    },
+    {
+      "epoch": 0.625888719380559,
+      "grad_norm": 3.5136171225226622,
+      "learning_rate": 6.487249834322095e-07,
+      "loss": 0.6822,
+      "step": 1359
+    },
+    {
+      "epoch": 0.6263492703146138,
+      "grad_norm": 3.133863907558626,
+      "learning_rate": 6.473280420273172e-07,
+      "loss": 0.5195,
+      "step": 1360
+    },
+    {
+      "epoch": 0.6268098212486687,
+      "grad_norm": 3.302281968664348,
+      "learning_rate": 6.459318861589685e-07,
+      "loss": 0.6776,
+      "step": 1361
+    },
+    {
+      "epoch": 0.6272703721827236,
+      "grad_norm": 3.1331202449009643,
+      "learning_rate": 6.445365189369411e-07,
+      "loss": 0.5329,
+      "step": 1362
+    },
+    {
+      "epoch": 0.6277309231167785,
+      "grad_norm": 3.2107222157004673,
+      "learning_rate": 6.431419434692563e-07,
+      "loss": 0.4996,
+      "step": 1363
+    },
+    {
+      "epoch": 0.6281914740508333,
+      "grad_norm": 3.3986716761760984,
+      "learning_rate": 6.417481628621717e-07,
+      "loss": 0.5303,
+      "step": 1364
+    },
+    {
+      "epoch": 0.6286520249848881,
+      "grad_norm": 3.066320791566163,
+      "learning_rate": 6.403551802201748e-07,
+      "loss": 0.5218,
+      "step": 1365
+    },
+    {
+      "epoch": 0.629112575918943,
+      "grad_norm": 3.2972142393202772,
+      "learning_rate": 6.389629986459755e-07,
+      "loss": 0.6122,
+      "step": 1366
+    },
+    {
+      "epoch": 0.6295731268529979,
+      "grad_norm": 3.1560358559235415,
+      "learning_rate": 6.375716212404989e-07,
+      "loss": 0.539,
+      "step": 1367
+    },
+    {
+      "epoch": 0.6300336777870528,
+      "grad_norm": 3.0343679900830036,
+      "learning_rate": 6.3618105110288e-07,
+      "loss": 0.4605,
+      "step": 1368
+    },
+    {
+      "epoch": 0.6304942287211076,
+      "grad_norm": 3.050511631671474,
+      "learning_rate": 6.347912913304548e-07,
+      "loss": 0.6179,
+      "step": 1369
+    },
+    {
+      "epoch": 0.6309547796551624,
+      "grad_norm": 3.2629876129159627,
+      "learning_rate": 6.334023450187549e-07,
+      "loss": 0.5378,
+      "step": 1370
+    },
+    {
+      "epoch": 0.6314153305892174,
+      "grad_norm": 3.0550921504844113,
+      "learning_rate": 6.320142152614993e-07,
+      "loss": 0.5586,
+      "step": 1371
+    },
+    {
+      "epoch": 0.6318758815232722,
+      "grad_norm": 2.87904745693612,
+      "learning_rate": 6.306269051505888e-07,
+      "loss": 0.5824,
+      "step": 1372
+    },
+    {
+      "epoch": 0.6323364324573271,
+      "grad_norm": 4.120605427481767,
+      "learning_rate": 6.292404177760986e-07,
+      "loss": 0.5787,
+      "step": 1373
+    },
+    {
+      "epoch": 0.6327969833913819,
+      "grad_norm": 3.281112660782104,
+      "learning_rate": 6.278547562262706e-07,
+      "loss": 0.5811,
+      "step": 1374
+    },
+    {
+      "epoch": 0.6332575343254369,
+      "grad_norm": 3.154859004430421,
+      "learning_rate": 6.264699235875084e-07,
+      "loss": 0.6182,
+      "step": 1375
+    },
+    {
+      "epoch": 0.6337180852594917,
+      "grad_norm": 3.0668232663377295,
+      "learning_rate": 6.250859229443684e-07,
+      "loss": 0.5067,
+      "step": 1376
+    },
+    {
+      "epoch": 0.6341786361935465,
+      "grad_norm": 3.060022701793419,
+      "learning_rate": 6.237027573795538e-07,
+      "loss": 0.5955,
+      "step": 1377
+    },
+    {
+      "epoch": 0.6346391871276014,
+      "grad_norm": 3.158763791575397,
+      "learning_rate": 6.223204299739087e-07,
+      "loss": 0.6123,
+      "step": 1378
+    },
+    {
+      "epoch": 0.6350997380616562,
+      "grad_norm": 3.075043672464968,
+      "learning_rate": 6.209389438064092e-07,
+      "loss": 0.5822,
+      "step": 1379
+    },
+    {
+      "epoch": 0.6355602889957112,
+      "grad_norm": 2.7672680741174247,
+      "learning_rate": 6.19558301954158e-07,
+      "loss": 0.5106,
+      "step": 1380
+    },
+    {
+      "epoch": 0.636020839929766,
+      "grad_norm": 2.9893686560154107,
+      "learning_rate": 6.181785074923777e-07,
+      "loss": 0.523,
+      "step": 1381
+    },
+    {
+      "epoch": 0.6364813908638208,
+      "grad_norm": 3.6442869591374722,
+      "learning_rate": 6.167995634944025e-07,
+      "loss": 0.6623,
+      "step": 1382
+    },
+    {
+      "epoch": 0.6369419417978757,
+      "grad_norm": 2.9356555531231527,
+      "learning_rate": 6.154214730316738e-07,
+      "loss": 0.4692,
+      "step": 1383
+    },
+    {
+      "epoch": 0.6374024927319306,
+      "grad_norm": 3.201063109690519,
+      "learning_rate": 6.1404423917373e-07,
+      "loss": 0.4439,
+      "step": 1384
+    },
+    {
+      "epoch": 0.6378630436659855,
+      "grad_norm": 3.0104867074963315,
+      "learning_rate": 6.126678649882019e-07,
+      "loss": 0.4768,
+      "step": 1385
+    },
+    {
+      "epoch": 0.6383235946000403,
+      "grad_norm": 3.327881299870761,
+      "learning_rate": 6.112923535408073e-07,
+      "loss": 0.4477,
+      "step": 1386
+    },
+    {
+      "epoch": 0.6387841455340951,
+      "grad_norm": 3.0269280110830863,
+      "learning_rate": 6.099177078953397e-07,
+      "loss": 0.5409,
+      "step": 1387
+    },
+    {
+      "epoch": 0.6392446964681501,
+      "grad_norm": 3.277885482165906,
+      "learning_rate": 6.085439311136664e-07,
+      "loss": 0.6291,
+      "step": 1388
+    },
+    {
+      "epoch": 0.6397052474022049,
+      "grad_norm": 3.180671239132894,
+      "learning_rate": 6.071710262557181e-07,
+      "loss": 0.5016,
+      "step": 1389
+    },
+    {
+      "epoch": 0.6401657983362598,
+      "grad_norm": 3.056915861723639,
+      "learning_rate": 6.057989963794832e-07,
+      "loss": 0.4639,
+      "step": 1390
+    },
+    {
+      "epoch": 0.6406263492703146,
+      "grad_norm": 2.977693194989121,
+      "learning_rate": 6.044278445410025e-07,
+      "loss": 0.5895,
+      "step": 1391
+    },
+    {
+      "epoch": 0.6410869002043694,
+      "grad_norm": 2.9786799583494266,
+      "learning_rate": 6.030575737943595e-07,
+      "loss": 0.4599,
+      "step": 1392
+    },
+    {
+      "epoch": 0.6415474511384244,
+      "grad_norm": 2.8694400733669227,
+      "learning_rate": 6.016881871916766e-07,
+      "loss": 0.6239,
+      "step": 1393
+    },
+    {
+      "epoch": 0.6420080020724792,
+      "grad_norm": 3.509458526598221,
+      "learning_rate": 6.003196877831059e-07,
+      "loss": 0.6965,
+      "step": 1394
+    },
+    {
+      "epoch": 0.642468553006534,
+      "grad_norm": 3.131981687735861,
+      "learning_rate": 5.989520786168235e-07,
+      "loss": 0.5088,
+      "step": 1395
+    },
+    {
+      "epoch": 0.6429291039405889,
+      "grad_norm": 3.532782457274475,
+      "learning_rate": 5.975853627390232e-07,
+      "loss": 0.5129,
+      "step": 1396
+    },
+    {
+      "epoch": 0.6433896548746438,
+      "grad_norm": 2.5625501669909307,
+      "learning_rate": 5.962195431939084e-07,
+      "loss": 0.4677,
+      "step": 1397
+    },
+    {
+      "epoch": 0.6438502058086987,
+      "grad_norm": 3.27442037003512,
+      "learning_rate": 5.94854623023686e-07,
+      "loss": 0.5125,
+      "step": 1398
+    },
+    {
+      "epoch": 0.6443107567427535,
+      "grad_norm": 3.2218724725996184,
+      "learning_rate": 5.934906052685603e-07,
+      "loss": 0.6094,
+      "step": 1399
+    },
+    {
+      "epoch": 0.6447713076768083,
+      "grad_norm": 3.5948100953063102,
+      "learning_rate": 5.921274929667251e-07,
+      "loss": 0.6102,
+      "step": 1400
+    },
+    {
+      "epoch": 0.6452318586108633,
+      "grad_norm": 3.1588508612291766,
+      "learning_rate": 5.907652891543576e-07,
+      "loss": 0.5734,
+      "step": 1401
+    },
+    {
+      "epoch": 0.6456924095449181,
+      "grad_norm": 3.700691639948302,
+      "learning_rate": 5.894039968656114e-07,
+      "loss": 0.5963,
+      "step": 1402
+    },
+    {
+      "epoch": 0.646152960478973,
+      "grad_norm": 3.2297651679426806,
+      "learning_rate": 5.880436191326092e-07,
+      "loss": 0.6396,
+      "step": 1403
+    },
+    {
+      "epoch": 0.6466135114130278,
+      "grad_norm": 3.241229586568738,
+      "learning_rate": 5.866841589854381e-07,
+      "loss": 0.5148,
+      "step": 1404
+    },
+    {
+      "epoch": 0.6470740623470826,
+      "grad_norm": 3.2127854047817186,
+      "learning_rate": 5.853256194521395e-07,
+      "loss": 0.5669,
+      "step": 1405
+    },
+    {
+      "epoch": 0.6475346132811376,
+      "grad_norm": 3.1150319104457234,
+      "learning_rate": 5.83968003558706e-07,
+      "loss": 0.5669,
+      "step": 1406
+    },
+    {
+      "epoch": 0.6479951642151924,
+      "grad_norm": 3.1285530927395095,
+      "learning_rate": 5.826113143290717e-07,
+      "loss": 0.5897,
+      "step": 1407
+    },
+    {
+      "epoch": 0.6484557151492473,
+      "grad_norm": 3.5323185152016388,
+      "learning_rate": 5.812555547851068e-07,
+      "loss": 0.5374,
+      "step": 1408
+    },
+    {
+      "epoch": 0.6489162660833021,
+      "grad_norm": 3.4748049321150383,
+      "learning_rate": 5.799007279466111e-07,
+      "loss": 0.5503,
+      "step": 1409
+    },
+    {
+      "epoch": 0.649376817017357,
+      "grad_norm": 3.404681472569695,
+      "learning_rate": 5.785468368313076e-07,
+      "loss": 0.5303,
+      "step": 1410
+    },
+    {
+      "epoch": 0.6498373679514119,
+      "grad_norm": 2.9485744210160107,
+      "learning_rate": 5.77193884454833e-07,
+      "loss": 0.5113,
+      "step": 1411
+    },
+    {
+      "epoch": 0.6502979188854667,
+      "grad_norm": 2.7092540440014807,
+      "learning_rate": 5.758418738307351e-07,
+      "loss": 0.5,
+      "step": 1412
+    },
+    {
+      "epoch": 0.6507584698195216,
+      "grad_norm": 2.853364404240714,
+      "learning_rate": 5.74490807970463e-07,
+      "loss": 0.5665,
+      "step": 1413
+    },
+    {
+      "epoch": 0.6512190207535765,
+      "grad_norm": 3.116516600142004,
+      "learning_rate": 5.731406898833623e-07,
+      "loss": 0.6661,
+      "step": 1414
+    },
+    {
+      "epoch": 0.6516795716876314,
+      "grad_norm": 2.8236292161466294,
+      "learning_rate": 5.717915225766661e-07,
+      "loss": 0.4524,
+      "step": 1415
+    },
+    {
+      "epoch": 0.6521401226216862,
+      "grad_norm": 3.0310468470234246,
+      "learning_rate": 5.704433090554911e-07,
+      "loss": 0.462,
+      "step": 1416
+    },
+    {
+      "epoch": 0.652600673555741,
+      "grad_norm": 3.24038506868089,
+      "learning_rate": 5.690960523228294e-07,
+      "loss": 0.6213,
+      "step": 1417
+    },
+    {
+      "epoch": 0.6530612244897959,
+      "grad_norm": 3.0890141559456157,
+      "learning_rate": 5.677497553795409e-07,
+      "loss": 0.4549,
+      "step": 1418
+    },
+    {
+      "epoch": 0.6535217754238508,
+      "grad_norm": 2.7880931209315842,
+      "learning_rate": 5.664044212243489e-07,
+      "loss": 0.4132,
+      "step": 1419
+    },
+    {
+      "epoch": 0.6539823263579057,
+      "grad_norm": 2.9558254074326427,
+      "learning_rate": 5.650600528538318e-07,
+      "loss": 0.494,
+      "step": 1420
+    },
+    {
+      "epoch": 0.6544428772919605,
+      "grad_norm": 2.914896993211941,
+      "learning_rate": 5.637166532624163e-07,
+      "loss": 0.5937,
+      "step": 1421
+    },
+    {
+      "epoch": 0.6549034282260153,
+      "grad_norm": 3.292628338829398,
+      "learning_rate": 5.623742254423718e-07,
+      "loss": 0.5269,
+      "step": 1422
+    },
+    {
+      "epoch": 0.6553639791600703,
+      "grad_norm": 3.2242310456562673,
+      "learning_rate": 5.610327723838037e-07,
+      "loss": 0.7023,
+      "step": 1423
+    },
+    {
+      "epoch": 0.6558245300941251,
+      "grad_norm": 3.240542534935448,
+      "learning_rate": 5.596922970746449e-07,
+      "loss": 0.5462,
+      "step": 1424
+    },
+    {
+      "epoch": 0.65628508102818,
+      "grad_norm": 2.949873633276392,
+      "learning_rate": 5.583528025006513e-07,
+      "loss": 0.4527,
+      "step": 1425
+    },
+    {
+      "epoch": 0.6567456319622348,
+      "grad_norm": 3.45439820707799,
+      "learning_rate": 5.570142916453944e-07,
+      "loss": 0.5727,
+      "step": 1426
+    },
+    {
+      "epoch": 0.6572061828962897,
+      "grad_norm": 3.0037709983691574,
+      "learning_rate": 5.556767674902548e-07,
+      "loss": 0.5799,
+      "step": 1427
+    },
+    {
+      "epoch": 0.6576667338303446,
+      "grad_norm": 2.846902934654531,
+      "learning_rate": 5.54340233014414e-07,
+      "loss": 0.5848,
+      "step": 1428
+    },
+    {
+      "epoch": 0.6581272847643994,
+      "grad_norm": 3.193107289231858,
+      "learning_rate": 5.530046911948505e-07,
+      "loss": 0.4999,
+      "step": 1429
+    },
+    {
+      "epoch": 0.6585878356984542,
+      "grad_norm": 3.396826464978981,
+      "learning_rate": 5.516701450063316e-07,
+      "loss": 0.4476,
+      "step": 1430
+    },
+    {
+      "epoch": 0.6590483866325091,
+      "grad_norm": 3.8083097104308505,
+      "learning_rate": 5.503365974214058e-07,
+      "loss": 0.6308,
+      "step": 1431
+    },
+    {
+      "epoch": 0.659508937566564,
+      "grad_norm": 2.9191556934998615,
+      "learning_rate": 5.490040514103995e-07,
+      "loss": 0.4147,
+      "step": 1432
+    },
+    {
+      "epoch": 0.6599694885006189,
+      "grad_norm": 2.8784243722308553,
+      "learning_rate": 5.476725099414062e-07,
+      "loss": 0.5174,
+      "step": 1433
+    },
+    {
+      "epoch": 0.6604300394346737,
+      "grad_norm": 3.350266406134782,
+      "learning_rate": 5.463419759802817e-07,
+      "loss": 0.6058,
+      "step": 1434
+    },
+    {
+      "epoch": 0.6608905903687285,
+      "grad_norm": 3.1483929420773555,
+      "learning_rate": 5.450124524906401e-07,
+      "loss": 0.518,
+      "step": 1435
+    },
+    {
+      "epoch": 0.6613511413027835,
+      "grad_norm": 3.4936502869433137,
+      "learning_rate": 5.436839424338425e-07,
+      "loss": 0.5507,
+      "step": 1436
+    },
+    {
+      "epoch": 0.6618116922368383,
+      "grad_norm": 3.1219573705478125,
+      "learning_rate": 5.423564487689929e-07,
+      "loss": 0.4988,
+      "step": 1437
+    },
+    {
+      "epoch": 0.6622722431708932,
+      "grad_norm": 3.582117403088983,
+      "learning_rate": 5.410299744529332e-07,
+      "loss": 0.5612,
+      "step": 1438
+    },
+    {
+      "epoch": 0.662732794104948,
+      "grad_norm": 3.5187886191292295,
+      "learning_rate": 5.397045224402326e-07,
+      "loss": 0.5867,
+      "step": 1439
+    },
+    {
+      "epoch": 0.663193345039003,
+      "grad_norm": 2.963457737411872,
+      "learning_rate": 5.383800956831846e-07,
+      "loss": 0.4929,
+      "step": 1440
+    },
+    {
+      "epoch": 0.6636538959730578,
+      "grad_norm": 3.4468724975782883,
+      "learning_rate": 5.370566971317989e-07,
+      "loss": 0.6108,
+      "step": 1441
+    },
+    {
+      "epoch": 0.6641144469071126,
+      "grad_norm": 3.1533494269546267,
+      "learning_rate": 5.357343297337943e-07,
+      "loss": 0.5902,
+      "step": 1442
+    },
+    {
+      "epoch": 0.6645749978411675,
+      "grad_norm": 3.3732802906601522,
+      "learning_rate": 5.344129964345934e-07,
+      "loss": 0.4952,
+      "step": 1443
+    },
+    {
+      "epoch": 0.6650355487752223,
+      "grad_norm": 2.9687293866646747,
+      "learning_rate": 5.330927001773154e-07,
+      "loss": 0.5435,
+      "step": 1444
+    },
+    {
+      "epoch": 0.6654960997092773,
+      "grad_norm": 3.480755832372405,
+      "learning_rate": 5.317734439027699e-07,
+      "loss": 0.564,
+      "step": 1445
+    },
+    {
+      "epoch": 0.6659566506433321,
+      "grad_norm": 2.9931832336561564,
+      "learning_rate": 5.304552305494492e-07,
+      "loss": 0.4201,
+      "step": 1446
+    },
+    {
+      "epoch": 0.6664172015773869,
+      "grad_norm": 2.85584265239967,
+      "learning_rate": 5.291380630535231e-07,
+      "loss": 0.5615,
+      "step": 1447
+    },
+    {
+      "epoch": 0.6668777525114418,
+      "grad_norm": 2.8281402193294056,
+      "learning_rate": 5.278219443488328e-07,
+      "loss": 0.5043,
+      "step": 1448
+    },
+    {
+      "epoch": 0.6673383034454967,
+      "grad_norm": 3.0403775949677923,
+      "learning_rate": 5.265068773668812e-07,
+      "loss": 0.5079,
+      "step": 1449
+    },
+    {
+      "epoch": 0.6677988543795516,
+      "grad_norm": 3.3174054936344897,
+      "learning_rate": 5.251928650368307e-07,
+      "loss": 0.5716,
+      "step": 1450
+    },
+    {
+      "epoch": 0.6682594053136064,
+      "grad_norm": 3.081361444525388,
+      "learning_rate": 5.238799102854941e-07,
+      "loss": 0.4981,
+      "step": 1451
+    },
+    {
+      "epoch": 0.6687199562476612,
+      "grad_norm": 3.1883940258418684,
+      "learning_rate": 5.225680160373275e-07,
+      "loss": 0.4924,
+      "step": 1452
+    },
+    {
+      "epoch": 0.6691805071817162,
+      "grad_norm": 3.311273891600756,
+      "learning_rate": 5.212571852144261e-07,
+      "loss": 0.5837,
+      "step": 1453
+    },
+    {
+      "epoch": 0.669641058115771,
+      "grad_norm": 3.138972108234792,
+      "learning_rate": 5.199474207365162e-07,
+      "loss": 0.5111,
+      "step": 1454
+    },
+    {
+      "epoch": 0.6701016090498259,
+      "grad_norm": 2.749902984155162,
+      "learning_rate": 5.186387255209481e-07,
+      "loss": 0.4912,
+      "step": 1455
+    },
+    {
+      "epoch": 0.6705621599838807,
+      "grad_norm": 3.201073779062438,
+      "learning_rate": 5.173311024826915e-07,
+      "loss": 0.5993,
+      "step": 1456
+    },
+    {
+      "epoch": 0.6710227109179355,
+      "grad_norm": 3.054277406726679,
+      "learning_rate": 5.160245545343274e-07,
+      "loss": 0.5429,
+      "step": 1457
+    },
+    {
+      "epoch": 0.6714832618519905,
+      "grad_norm": 3.262974854228882,
+      "learning_rate": 5.147190845860426e-07,
+      "loss": 0.67,
+      "step": 1458
+    },
+    {
+      "epoch": 0.6719438127860453,
+      "grad_norm": 3.2252159866634584,
+      "learning_rate": 5.134146955456218e-07,
+      "loss": 0.4783,
+      "step": 1459
+    },
+    {
+      "epoch": 0.6724043637201002,
+      "grad_norm": 3.3051119530232946,
+      "learning_rate": 5.121113903184431e-07,
+      "loss": 0.4966,
+      "step": 1460
+    },
+    {
+      "epoch": 0.672864914654155,
+      "grad_norm": 3.064671697196766,
+      "learning_rate": 5.108091718074705e-07,
+      "loss": 0.6525,
+      "step": 1461
+    },
+    {
+      "epoch": 0.6733254655882099,
+      "grad_norm": 2.857963094780648,
+      "learning_rate": 5.095080429132459e-07,
+      "loss": 0.5369,
+      "step": 1462
+    },
+    {
+      "epoch": 0.6737860165222648,
+      "grad_norm": 3.310426657500392,
+      "learning_rate": 5.082080065338872e-07,
+      "loss": 0.5112,
+      "step": 1463
+    },
+    {
+      "epoch": 0.6742465674563196,
+      "grad_norm": 3.225512724763038,
+      "learning_rate": 5.069090655650762e-07,
+      "loss": 0.581,
+      "step": 1464
+    },
+    {
+      "epoch": 0.6747071183903744,
+      "grad_norm": 3.2679785563098784,
+      "learning_rate": 5.05611222900055e-07,
+      "loss": 0.5522,
+      "step": 1465
+    },
+    {
+      "epoch": 0.6751676693244294,
+      "grad_norm": 3.446697533304695,
+      "learning_rate": 5.043144814296215e-07,
+      "loss": 0.5458,
+      "step": 1466
+    },
+    {
+      "epoch": 0.6756282202584842,
+      "grad_norm": 3.180139270812277,
+      "learning_rate": 5.030188440421185e-07,
+      "loss": 0.4763,
+      "step": 1467
+    },
+    {
+      "epoch": 0.6760887711925391,
+      "grad_norm": 3.448958257550766,
+      "learning_rate": 5.017243136234298e-07,
+      "loss": 0.5669,
+      "step": 1468
+    },
+    {
+      "epoch": 0.6765493221265939,
+      "grad_norm": 3.5331462163297243,
+      "learning_rate": 5.004308930569757e-07,
+      "loss": 0.6389,
+      "step": 1469
+    },
+    {
+      "epoch": 0.6770098730606487,
+      "grad_norm": 2.9967238157665124,
+      "learning_rate": 4.991385852237017e-07,
+      "loss": 0.5617,
+      "step": 1470
+    },
+    {
+      "epoch": 0.6774704239947037,
+      "grad_norm": 2.7688203871377235,
+      "learning_rate": 4.978473930020767e-07,
+      "loss": 0.4754,
+      "step": 1471
+    },
+    {
+      "epoch": 0.6779309749287585,
+      "grad_norm": 3.266647952459185,
+      "learning_rate": 4.965573192680841e-07,
+      "loss": 0.5109,
+      "step": 1472
+    },
+    {
+      "epoch": 0.6783915258628134,
+      "grad_norm": 2.825024520007373,
+      "learning_rate": 4.952683668952152e-07,
+      "loss": 0.4731,
+      "step": 1473
+    },
+    {
+      "epoch": 0.6788520767968682,
+      "grad_norm": 3.426822674267336,
+      "learning_rate": 4.939805387544649e-07,
+      "loss": 0.5494,
+      "step": 1474
+    },
+    {
+      "epoch": 0.6793126277309232,
+      "grad_norm": 3.4379075787612448,
+      "learning_rate": 4.926938377143232e-07,
+      "loss": 0.6234,
+      "step": 1475
+    },
+    {
+      "epoch": 0.679773178664978,
+      "grad_norm": 3.037669101666441,
+      "learning_rate": 4.914082666407704e-07,
+      "loss": 0.4952,
+      "step": 1476
+    },
+    {
+      "epoch": 0.6802337295990328,
+      "grad_norm": 3.1953172155925977,
+      "learning_rate": 4.901238283972685e-07,
+      "loss": 0.4285,
+      "step": 1477
+    },
+    {
+      "epoch": 0.6806942805330877,
+      "grad_norm": 3.2567976107387646,
+      "learning_rate": 4.888405258447576e-07,
+      "loss": 0.5178,
+      "step": 1478
+    },
+    {
+      "epoch": 0.6811548314671426,
+      "grad_norm": 3.0243656981433,
+      "learning_rate": 4.875583618416481e-07,
+      "loss": 0.5382,
+      "step": 1479
+    },
+    {
+      "epoch": 0.6816153824011975,
+      "grad_norm": 2.995186764899891,
+      "learning_rate": 4.862773392438131e-07,
+      "loss": 0.5426,
+      "step": 1480
+    },
+    {
+      "epoch": 0.6820759333352523,
+      "grad_norm": 3.6636504352304686,
+      "learning_rate": 4.849974609045848e-07,
+      "loss": 0.6404,
+      "step": 1481
+    },
+    {
+      "epoch": 0.6825364842693071,
+      "grad_norm": 2.692736667622777,
+      "learning_rate": 4.837187296747463e-07,
+      "loss": 0.5281,
+      "step": 1482
+    },
+    {
+      "epoch": 0.682997035203362,
+      "grad_norm": 2.7042483128746064,
+      "learning_rate": 4.82441148402525e-07,
+      "loss": 0.575,
+      "step": 1483
+    },
+    {
+      "epoch": 0.6834575861374169,
+      "grad_norm": 3.0352564028625952,
+      "learning_rate": 4.811647199335877e-07,
+      "loss": 0.5098,
+      "step": 1484
+    },
+    {
+      "epoch": 0.6839181370714718,
+      "grad_norm": 3.127898271916526,
+      "learning_rate": 4.798894471110336e-07,
+      "loss": 0.5613,
+      "step": 1485
+    },
+    {
+      "epoch": 0.6843786880055266,
+      "grad_norm": 2.950470950095703,
+      "learning_rate": 4.786153327753864e-07,
+      "loss": 0.5638,
+      "step": 1486
+    },
+    {
+      "epoch": 0.6848392389395814,
+      "grad_norm": 2.9327112511286777,
+      "learning_rate": 4.773423797645911e-07,
+      "loss": 0.5308,
+      "step": 1487
+    },
+    {
+      "epoch": 0.6852997898736364,
+      "grad_norm": 3.2221808265567917,
+      "learning_rate": 4.76070590914005e-07,
+      "loss": 0.5875,
+      "step": 1488
+    },
+    {
+      "epoch": 0.6857603408076912,
+      "grad_norm": 2.9032838510402383,
+      "learning_rate": 4.747999690563932e-07,
+      "loss": 0.4657,
+      "step": 1489
+    },
+    {
+      "epoch": 0.686220891741746,
+      "grad_norm": 3.253618175545866,
+      "learning_rate": 4.7353051702191994e-07,
+      "loss": 0.5498,
+      "step": 1490
+    },
+    {
+      "epoch": 0.6866814426758009,
+      "grad_norm": 3.3055447616805296,
+      "learning_rate": 4.7226223763814545e-07,
+      "loss": 0.5192,
+      "step": 1491
+    },
+    {
+      "epoch": 0.6871419936098558,
+      "grad_norm": 3.662312307552012,
+      "learning_rate": 4.709951337300174e-07,
+      "loss": 0.5508,
+      "step": 1492
+    },
+    {
+      "epoch": 0.6876025445439107,
+      "grad_norm": 3.0362093035408795,
+      "learning_rate": 4.697292081198646e-07,
+      "loss": 0.4648,
+      "step": 1493
+    },
+    {
+      "epoch": 0.6880630954779655,
+      "grad_norm": 3.1492972902080627,
+      "learning_rate": 4.684644636273922e-07,
+      "loss": 0.4502,
+      "step": 1494
+    },
+    {
+      "epoch": 0.6885236464120204,
+      "grad_norm": 2.9996769191548416,
+      "learning_rate": 4.6720090306967465e-07,
+      "loss": 0.6546,
+      "step": 1495
+    },
+    {
+      "epoch": 0.6889841973460752,
+      "grad_norm": 2.9045793925682104,
+      "learning_rate": 4.6593852926114784e-07,
+      "loss": 0.4884,
+      "step": 1496
+    },
+    {
+      "epoch": 0.6894447482801301,
+      "grad_norm": 3.1298670395653962,
+      "learning_rate": 4.646773450136067e-07,
+      "loss": 0.5642,
+      "step": 1497
+    },
+    {
+      "epoch": 0.689905299214185,
+      "grad_norm": 3.0495111305757163,
+      "learning_rate": 4.634173531361947e-07,
+      "loss": 0.5585,
+      "step": 1498
+    },
+    {
+      "epoch": 0.6903658501482398,
+      "grad_norm": 3.020709936806421,
+      "learning_rate": 4.6215855643539903e-07,
+      "loss": 0.4491,
+      "step": 1499
+    },
+    {
+      "epoch": 0.6908264010822946,
+      "grad_norm": 3.1842354060638467,
+      "learning_rate": 4.609009577150472e-07,
+      "loss": 0.5734,
+      "step": 1500
+    },
+    {
+      "epoch": 0.6912869520163496,
+      "grad_norm": 3.4495373317520914,
+      "learning_rate": 4.5964455977629593e-07,
+      "loss": 0.5147,
+      "step": 1501
+    },
+    {
+      "epoch": 0.6917475029504044,
+      "grad_norm": 2.8702835074826494,
+      "learning_rate": 4.583893654176285e-07,
+      "loss": 0.4863,
+      "step": 1502
+    },
+    {
+      "epoch": 0.6922080538844593,
+      "grad_norm": 2.929782205551764,
+      "learning_rate": 4.5713537743484754e-07,
+      "loss": 0.4934,
+      "step": 1503
+    },
+    {
+      "epoch": 0.6926686048185141,
+      "grad_norm": 3.0017796045087017,
+      "learning_rate": 4.5588259862106725e-07,
+      "loss": 0.5107,
+      "step": 1504
+    },
+    {
+      "epoch": 0.6931291557525691,
+      "grad_norm": 3.484258951507888,
+      "learning_rate": 4.5463103176671016e-07,
+      "loss": 0.6145,
+      "step": 1505
+    },
+    {
+      "epoch": 0.6935897066866239,
+      "grad_norm": 2.7555413280578143,
+      "learning_rate": 4.533806796594989e-07,
+      "loss": 0.5649,
+      "step": 1506
+    },
+    {
+      "epoch": 0.6940502576206787,
+      "grad_norm": 3.3509703674324425,
+      "learning_rate": 4.521315450844492e-07,
+      "loss": 0.5738,
+      "step": 1507
+    },
+    {
+      "epoch": 0.6945108085547336,
+      "grad_norm": 3.075909066062986,
+      "learning_rate": 4.508836308238664e-07,
+      "loss": 0.5215,
+      "step": 1508
+    },
+    {
+      "epoch": 0.6949713594887884,
+      "grad_norm": 3.116185512270017,
+      "learning_rate": 4.4963693965733686e-07,
+      "loss": 0.5707,
+      "step": 1509
+    },
+    {
+      "epoch": 0.6954319104228434,
+      "grad_norm": 3.301765717057608,
+      "learning_rate": 4.483914743617235e-07,
+      "loss": 0.6423,
+      "step": 1510
+    },
+    {
+      "epoch": 0.6958924613568982,
+      "grad_norm": 3.517582404076343,
+      "learning_rate": 4.471472377111574e-07,
+      "loss": 0.6392,
+      "step": 1511
+    },
+    {
+      "epoch": 0.696353012290953,
+      "grad_norm": 3.4096020565867193,
+      "learning_rate": 4.459042324770338e-07,
+      "loss": 0.59,
+      "step": 1512
+    },
+    {
+      "epoch": 0.6968135632250079,
+      "grad_norm": 3.2420031543237977,
+      "learning_rate": 4.446624614280058e-07,
+      "loss": 0.5858,
+      "step": 1513
+    },
+    {
+      "epoch": 0.6972741141590628,
+      "grad_norm": 3.3459346730805,
+      "learning_rate": 4.4342192732997565e-07,
+      "loss": 0.5173,
+      "step": 1514
+    },
+    {
+      "epoch": 0.6977346650931177,
+      "grad_norm": 3.2862422773543147,
+      "learning_rate": 4.4218263294609205e-07,
+      "loss": 0.5255,
+      "step": 1515
+    },
+    {
+      "epoch": 0.6981952160271725,
+      "grad_norm": 3.1669321223815614,
+      "learning_rate": 4.4094458103674204e-07,
+      "loss": 0.5333,
+      "step": 1516
+    },
+    {
+      "epoch": 0.6986557669612273,
+      "grad_norm": 2.9341763845722686,
+      "learning_rate": 4.397077743595444e-07,
+      "loss": 0.5475,
+      "step": 1517
+    },
+    {
+      "epoch": 0.6991163178952823,
+      "grad_norm": 3.298479093779986,
+      "learning_rate": 4.384722156693451e-07,
+      "loss": 0.6168,
+      "step": 1518
+    },
+    {
+      "epoch": 0.6995768688293371,
+      "grad_norm": 3.774693547188722,
+      "learning_rate": 4.3723790771821067e-07,
+      "loss": 0.4971,
+      "step": 1519
+    },
+    {
+      "epoch": 0.700037419763392,
+      "grad_norm": 3.2701266862488003,
+      "learning_rate": 4.3600485325542047e-07,
+      "loss": 0.5021,
+      "step": 1520
+    },
+    {
+      "epoch": 0.7004979706974468,
+      "grad_norm": 2.8544733630249457,
+      "learning_rate": 4.3477305502746275e-07,
+      "loss": 0.5578,
+      "step": 1521
+    },
+    {
+      "epoch": 0.7009585216315016,
+      "grad_norm": 2.859027758778164,
+      "learning_rate": 4.335425157780277e-07,
+      "loss": 0.5669,
+      "step": 1522
+    },
+    {
+      "epoch": 0.7014190725655566,
+      "grad_norm": 3.1340569464788226,
+      "learning_rate": 4.323132382480015e-07,
+      "loss": 0.4886,
+      "step": 1523
+    },
+    {
+      "epoch": 0.7018796234996114,
+      "grad_norm": 3.533977315795044,
+      "learning_rate": 4.3108522517545866e-07,
+      "loss": 0.6171,
+      "step": 1524
+    },
+    {
+      "epoch": 0.7023401744336663,
+      "grad_norm": 3.1550549229148186,
+      "learning_rate": 4.2985847929565865e-07,
+      "loss": 0.5375,
+      "step": 1525
+    },
+    {
+      "epoch": 0.7028007253677211,
+      "grad_norm": 2.7951674621480063,
+      "learning_rate": 4.2863300334103837e-07,
+      "loss": 0.4927,
+      "step": 1526
+    },
+    {
+      "epoch": 0.703261276301776,
+      "grad_norm": 3.404007775142097,
+      "learning_rate": 4.2740880004120474e-07,
+      "loss": 0.5759,
+      "step": 1527
+    },
+    {
+      "epoch": 0.7037218272358309,
+      "grad_norm": 3.0714696704935154,
+      "learning_rate": 4.2618587212293147e-07,
+      "loss": 0.5976,
+      "step": 1528
+    },
+    {
+      "epoch": 0.7041823781698857,
+      "grad_norm": 3.6804392641687826,
+      "learning_rate": 4.2496422231015115e-07,
+      "loss": 0.6249,
+      "step": 1529
+    },
+    {
+      "epoch": 0.7046429291039406,
+      "grad_norm": 3.2731509390374627,
+      "learning_rate": 4.237438533239488e-07,
+      "loss": 0.5356,
+      "step": 1530
+    },
+    {
+      "epoch": 0.7051034800379955,
+      "grad_norm": 3.6052417982765563,
+      "learning_rate": 4.2252476788255733e-07,
+      "loss": 0.5792,
+      "step": 1531
+    },
+    {
+      "epoch": 0.7055640309720503,
+      "grad_norm": 3.3733022802458446,
+      "learning_rate": 4.213069687013505e-07,
+      "loss": 0.6865,
+      "step": 1532
+    },
+    {
+      "epoch": 0.7060245819061052,
+      "grad_norm": 2.6172790916082618,
+      "learning_rate": 4.200904584928373e-07,
+      "loss": 0.4629,
+      "step": 1533
+    },
+    {
+      "epoch": 0.70648513284016,
+      "grad_norm": 3.3735377620802764,
+      "learning_rate": 4.1887523996665474e-07,
+      "loss": 0.6296,
+      "step": 1534
+    },
+    {
+      "epoch": 0.7069456837742149,
+      "grad_norm": 3.5155816418818584,
+      "learning_rate": 4.176613158295639e-07,
+      "loss": 0.6252,
+      "step": 1535
+    },
+    {
+      "epoch": 0.7074062347082698,
+      "grad_norm": 2.9319016244934795,
+      "learning_rate": 4.164486887854424e-07,
+      "loss": 0.6537,
+      "step": 1536
+    },
+    {
+      "epoch": 0.7078667856423246,
+      "grad_norm": 3.0752439651363703,
+      "learning_rate": 4.15237361535278e-07,
+      "loss": 0.6219,
+      "step": 1537
+    },
+    {
+      "epoch": 0.7083273365763795,
+      "grad_norm": 3.772619082368496,
+      "learning_rate": 4.140273367771643e-07,
+      "loss": 0.5692,
+      "step": 1538
+    },
+    {
+      "epoch": 0.7087878875104343,
+      "grad_norm": 3.0735890994766457,
+      "learning_rate": 4.1281861720629374e-07,
+      "loss": 0.5757,
+      "step": 1539
+    },
+    {
+      "epoch": 0.7092484384444893,
+      "grad_norm": 3.449289367509567,
+      "learning_rate": 4.1161120551495023e-07,
+      "loss": 0.6328,
+      "step": 1540
+    },
+    {
+      "epoch": 0.7097089893785441,
+      "grad_norm": 3.4333027638235,
+      "learning_rate": 4.1040510439250676e-07,
+      "loss": 0.453,
+      "step": 1541
+    },
+    {
+      "epoch": 0.7101695403125989,
+      "grad_norm": 3.4665469813554695,
+      "learning_rate": 4.092003165254154e-07,
+      "loss": 0.5686,
+      "step": 1542
+    },
+    {
+      "epoch": 0.7106300912466538,
+      "grad_norm": 2.979020490366814,
+      "learning_rate": 4.0799684459720295e-07,
+      "loss": 0.459,
+      "step": 1543
+    },
+    {
+      "epoch": 0.7110906421807087,
+      "grad_norm": 3.180076549332751,
+      "learning_rate": 4.067946912884672e-07,
+      "loss": 0.6623,
+      "step": 1544
+    },
+    {
+      "epoch": 0.7115511931147636,
+      "grad_norm": 3.345275079523344,
+      "learning_rate": 4.055938592768663e-07,
+      "loss": 0.5705,
+      "step": 1545
+    },
+    {
+      "epoch": 0.7120117440488184,
+      "grad_norm": 2.9898569750138875,
+      "learning_rate": 4.0439435123711707e-07,
+      "loss": 0.4446,
+      "step": 1546
+    },
+    {
+      "epoch": 0.7124722949828732,
+      "grad_norm": 2.892301015122262,
+      "learning_rate": 4.031961698409869e-07,
+      "loss": 0.4983,
+      "step": 1547
+    },
+    {
+      "epoch": 0.7129328459169281,
+      "grad_norm": 3.291501709727223,
+      "learning_rate": 4.0199931775728767e-07,
+      "loss": 0.6132,
+      "step": 1548
+    },
+    {
+      "epoch": 0.713393396850983,
+      "grad_norm": 3.1031247090774974,
+      "learning_rate": 4.008037976518711e-07,
+      "loss": 0.5285,
+      "step": 1549
+    },
+    {
+      "epoch": 0.7138539477850379,
+      "grad_norm": 3.5470619388484623,
+      "learning_rate": 3.996096121876221e-07,
+      "loss": 0.4594,
+      "step": 1550
+    },
+    {
+      "epoch": 0.7143144987190927,
+      "grad_norm": 3.095992564328872,
+      "learning_rate": 3.984167640244518e-07,
+      "loss": 0.5343,
+      "step": 1551
+    },
+    {
+      "epoch": 0.7147750496531475,
+      "grad_norm": 3.528528857914403,
+      "learning_rate": 3.972252558192938e-07,
+      "loss": 0.5564,
+      "step": 1552
+    },
+    {
+      "epoch": 0.7152356005872025,
+      "grad_norm": 3.0649641218559673,
+      "learning_rate": 3.960350902260967e-07,
+      "loss": 0.4632,
+      "step": 1553
+    },
+    {
+      "epoch": 0.7156961515212573,
+      "grad_norm": 3.4153270827868747,
+      "learning_rate": 3.948462698958188e-07,
+      "loss": 0.4675,
+      "step": 1554
+    },
+    {
+      "epoch": 0.7161567024553122,
+      "grad_norm": 3.1991732501247507,
+      "learning_rate": 3.9365879747642106e-07,
+      "loss": 0.5517,
+      "step": 1555
+    },
+    {
+      "epoch": 0.716617253389367,
+      "grad_norm": 2.9871782936902713,
+      "learning_rate": 3.924726756128631e-07,
+      "loss": 0.5189,
+      "step": 1556
+    },
+    {
+      "epoch": 0.7170778043234219,
+      "grad_norm": 2.9744051796884334,
+      "learning_rate": 3.912879069470966e-07,
+      "loss": 0.537,
+      "step": 1557
+    },
+    {
+      "epoch": 0.7175383552574768,
+      "grad_norm": 2.9363741803915433,
+      "learning_rate": 3.9010449411805747e-07,
+      "loss": 0.6124,
+      "step": 1558
+    },
+    {
+      "epoch": 0.7179989061915316,
+      "grad_norm": 2.9230129555443094,
+      "learning_rate": 3.889224397616635e-07,
+      "loss": 0.4577,
+      "step": 1559
+    },
+    {
+      "epoch": 0.7184594571255865,
+      "grad_norm": 2.9997803472056392,
+      "learning_rate": 3.8774174651080596e-07,
+      "loss": 0.5885,
+      "step": 1560
+    },
+    {
+      "epoch": 0.7189200080596413,
+      "grad_norm": 3.121077182785591,
+      "learning_rate": 3.865624169953439e-07,
+      "loss": 0.617,
+      "step": 1561
+    },
+    {
+      "epoch": 0.7193805589936962,
+      "grad_norm": 3.3670297246006378,
+      "learning_rate": 3.853844538420993e-07,
+      "loss": 0.5977,
+      "step": 1562
+    },
+    {
+      "epoch": 0.7198411099277511,
+      "grad_norm": 2.842166340649367,
+      "learning_rate": 3.8420785967485115e-07,
+      "loss": 0.5409,
+      "step": 1563
+    },
+    {
+      "epoch": 0.7203016608618059,
+      "grad_norm": 3.5406635193421656,
+      "learning_rate": 3.83032637114328e-07,
+      "loss": 0.6372,
+      "step": 1564
+    },
+    {
+      "epoch": 0.7207622117958608,
+      "grad_norm": 3.194617429657409,
+      "learning_rate": 3.8185878877820443e-07,
+      "loss": 0.5469,
+      "step": 1565
+    },
+    {
+      "epoch": 0.7212227627299157,
+      "grad_norm": 3.417157379310719,
+      "learning_rate": 3.806863172810936e-07,
+      "loss": 0.4731,
+      "step": 1566
+    },
+    {
+      "epoch": 0.7216833136639705,
+      "grad_norm": 3.231247160093252,
+      "learning_rate": 3.7951522523454214e-07,
+      "loss": 0.6959,
+      "step": 1567
+    },
+    {
+      "epoch": 0.7221438645980254,
+      "grad_norm": 3.358643739910038,
+      "learning_rate": 3.7834551524702364e-07,
+      "loss": 0.616,
+      "step": 1568
+    },
+    {
+      "epoch": 0.7226044155320802,
+      "grad_norm": 3.159395399638094,
+      "learning_rate": 3.7717718992393365e-07,
+      "loss": 0.539,
+      "step": 1569
+    },
+    {
+      "epoch": 0.7230649664661352,
+      "grad_norm": 3.288938398533548,
+      "learning_rate": 3.760102518675839e-07,
+      "loss": 0.6766,
+      "step": 1570
+    },
+    {
+      "epoch": 0.72352551740019,
+      "grad_norm": 3.1929425292220484,
+      "learning_rate": 3.748447036771949e-07,
+      "loss": 0.5312,
+      "step": 1571
+    },
+    {
+      "epoch": 0.7239860683342448,
+      "grad_norm": 3.1532440738029455,
+      "learning_rate": 3.736805479488936e-07,
+      "loss": 0.5694,
+      "step": 1572
+    },
+    {
+      "epoch": 0.7244466192682997,
+      "grad_norm": 3.1699423213216726,
+      "learning_rate": 3.7251778727570305e-07,
+      "loss": 0.5722,
+      "step": 1573
+    },
+    {
+      "epoch": 0.7249071702023545,
+      "grad_norm": 2.9111701929100198,
+      "learning_rate": 3.7135642424753967e-07,
+      "loss": 0.5001,
+      "step": 1574
+    },
+    {
+      "epoch": 0.7253677211364095,
+      "grad_norm": 3.1237883324779623,
+      "learning_rate": 3.701964614512082e-07,
+      "loss": 0.5708,
+      "step": 1575
+    },
+    {
+      "epoch": 0.7258282720704643,
+      "grad_norm": 2.7652944931913623,
+      "learning_rate": 3.690379014703928e-07,
+      "loss": 0.4368,
+      "step": 1576
+    },
+    {
+      "epoch": 0.7262888230045191,
+      "grad_norm": 3.209037928138226,
+      "learning_rate": 3.67880746885653e-07,
+      "loss": 0.6249,
+      "step": 1577
+    },
+    {
+      "epoch": 0.726749373938574,
+      "grad_norm": 3.0792001259828323,
+      "learning_rate": 3.667250002744199e-07,
+      "loss": 0.5269,
+      "step": 1578
+    },
+    {
+      "epoch": 0.7272099248726289,
+      "grad_norm": 3.3870656223423428,
+      "learning_rate": 3.6557066421098604e-07,
+      "loss": 0.4438,
+      "step": 1579
+    },
+    {
+      "epoch": 0.7276704758066838,
+      "grad_norm": 3.187784911852215,
+      "learning_rate": 3.644177412665039e-07,
+      "loss": 0.4357,
+      "step": 1580
+    },
+    {
+      "epoch": 0.7281310267407386,
+      "grad_norm": 3.3629689440356927,
+      "learning_rate": 3.6326623400897796e-07,
+      "loss": 0.5516,
+      "step": 1581
+    },
+    {
+      "epoch": 0.7285915776747934,
+      "grad_norm": 3.579677642223135,
+      "learning_rate": 3.621161450032586e-07,
+      "loss": 0.6288,
+      "step": 1582
+    },
+    {
+      "epoch": 0.7290521286088484,
+      "grad_norm": 2.918537780420392,
+      "learning_rate": 3.609674768110381e-07,
+      "loss": 0.5887,
+      "step": 1583
+    },
+    {
+      "epoch": 0.7295126795429032,
+      "grad_norm": 3.2728257749235903,
+      "learning_rate": 3.59820231990844e-07,
+      "loss": 0.6692,
+      "step": 1584
+    },
+    {
+      "epoch": 0.7299732304769581,
+      "grad_norm": 3.1155337342093614,
+      "learning_rate": 3.5867441309803325e-07,
+      "loss": 0.4757,
+      "step": 1585
+    },
+    {
+      "epoch": 0.7304337814110129,
+      "grad_norm": 3.017811596425641,
+      "learning_rate": 3.5753002268478625e-07,
+      "loss": 0.5903,
+      "step": 1586
+    },
+    {
+      "epoch": 0.7308943323450677,
+      "grad_norm": 3.5070818468579708,
+      "learning_rate": 3.5638706330010236e-07,
+      "loss": 0.5974,
+      "step": 1587
+    },
+    {
+      "epoch": 0.7313548832791227,
+      "grad_norm": 3.013090840123994,
+      "learning_rate": 3.552455374897935e-07,
+      "loss": 0.536,
+      "step": 1588
+    },
+    {
+      "epoch": 0.7318154342131775,
+      "grad_norm": 2.7740669709707517,
+      "learning_rate": 3.5410544779647735e-07,
+      "loss": 0.4466,
+      "step": 1589
+    },
+    {
+      "epoch": 0.7322759851472324,
+      "grad_norm": 3.437946984823077,
+      "learning_rate": 3.529667967595742e-07,
+      "loss": 0.5002,
+      "step": 1590
+    },
+    {
+      "epoch": 0.7327365360812872,
+      "grad_norm": 2.813883495540624,
+      "learning_rate": 3.518295869152994e-07,
+      "loss": 0.4955,
+      "step": 1591
+    },
+    {
+      "epoch": 0.7331970870153421,
+      "grad_norm": 2.8502290667648342,
+      "learning_rate": 3.5069382079665763e-07,
+      "loss": 0.4805,
+      "step": 1592
+    },
+    {
+      "epoch": 0.733657637949397,
+      "grad_norm": 3.234497100576854,
+      "learning_rate": 3.4955950093343857e-07,
+      "loss": 0.7048,
+      "step": 1593
+    },
+    {
+      "epoch": 0.7341181888834518,
+      "grad_norm": 3.131622303838718,
+      "learning_rate": 3.484266298522106e-07,
+      "loss": 0.5671,
+      "step": 1594
+    },
+    {
+      "epoch": 0.7345787398175067,
+      "grad_norm": 2.875003807206262,
+      "learning_rate": 3.472952100763141e-07,
+      "loss": 0.5951,
+      "step": 1595
+    },
+    {
+      "epoch": 0.7350392907515616,
+      "grad_norm": 2.967122454191587,
+      "learning_rate": 3.461652441258579e-07,
+      "loss": 0.5307,
+      "step": 1596
+    },
+    {
+      "epoch": 0.7354998416856164,
+      "grad_norm": 3.501420492607783,
+      "learning_rate": 3.450367345177122e-07,
+      "loss": 0.5531,
+      "step": 1597
+    },
+    {
+      "epoch": 0.7359603926196713,
+      "grad_norm": 3.0527813124759535,
+      "learning_rate": 3.4390968376550367e-07,
+      "loss": 0.5332,
+      "step": 1598
+    },
+    {
+      "epoch": 0.7364209435537261,
+      "grad_norm": 2.8619147841769266,
+      "learning_rate": 3.4278409437960865e-07,
+      "loss": 0.5634,
+      "step": 1599
+    },
+    {
+      "epoch": 0.7368814944877811,
+      "grad_norm": 3.1275902113895264,
+      "learning_rate": 3.4165996886714944e-07,
+      "loss": 0.5385,
+      "step": 1600
+    },
+    {
+      "epoch": 0.7373420454218359,
+      "grad_norm": 2.93151492063357,
+      "learning_rate": 3.405373097319875e-07,
+      "loss": 0.489,
+      "step": 1601
+    },
+    {
+      "epoch": 0.7378025963558907,
+      "grad_norm": 3.2966345212103354,
+      "learning_rate": 3.3941611947471703e-07,
+      "loss": 0.5367,
+      "step": 1602
+    },
+    {
+      "epoch": 0.7382631472899456,
+      "grad_norm": 2.7068239388271516,
+      "learning_rate": 3.3829640059266283e-07,
+      "loss": 0.5216,
+      "step": 1603
+    },
+    {
+      "epoch": 0.7387236982240004,
+      "grad_norm": 3.435870012306948,
+      "learning_rate": 3.3717815557987027e-07,
+      "loss": 0.5401,
+      "step": 1604
+    },
+    {
+      "epoch": 0.7391842491580554,
+      "grad_norm": 3.774078430992497,
+      "learning_rate": 3.360613869271016e-07,
+      "loss": 0.5301,
+      "step": 1605
+    },
+    {
+      "epoch": 0.7396448000921102,
+      "grad_norm": 2.974006922477261,
+      "learning_rate": 3.349460971218332e-07,
+      "loss": 0.5254,
+      "step": 1606
+    },
+    {
+      "epoch": 0.740105351026165,
+      "grad_norm": 3.1026779731207106,
+      "learning_rate": 3.3383228864824496e-07,
+      "loss": 0.4061,
+      "step": 1607
+    },
+    {
+      "epoch": 0.7405659019602199,
+      "grad_norm": 3.1384179646636836,
+      "learning_rate": 3.327199639872177e-07,
+      "loss": 0.5428,
+      "step": 1608
+    },
+    {
+      "epoch": 0.7410264528942748,
+      "grad_norm": 2.860698890001966,
+      "learning_rate": 3.316091256163288e-07,
+      "loss": 0.5852,
+      "step": 1609
+    },
+    {
+      "epoch": 0.7414870038283297,
+      "grad_norm": 2.8765636527594154,
+      "learning_rate": 3.3049977600984304e-07,
+      "loss": 0.4724,
+      "step": 1610
+    },
+    {
+      "epoch": 0.7419475547623845,
+      "grad_norm": 3.3500780381811746,
+      "learning_rate": 3.293919176387104e-07,
+      "loss": 0.6346,
+      "step": 1611
+    },
+    {
+      "epoch": 0.7424081056964393,
+      "grad_norm": 3.7026885508653193,
+      "learning_rate": 3.2828555297055946e-07,
+      "loss": 0.5841,
+      "step": 1612
+    },
+    {
+      "epoch": 0.7428686566304943,
+      "grad_norm": 3.0274792259951737,
+      "learning_rate": 3.271806844696905e-07,
+      "loss": 0.5425,
+      "step": 1613
+    },
+    {
+      "epoch": 0.7433292075645491,
+      "grad_norm": 3.746883296992905,
+      "learning_rate": 3.260773145970723e-07,
+      "loss": 0.6408,
+      "step": 1614
+    },
+    {
+      "epoch": 0.743789758498604,
+      "grad_norm": 3.357176161663444,
+      "learning_rate": 3.2497544581033555e-07,
+      "loss": 0.6193,
+      "step": 1615
+    },
+    {
+      "epoch": 0.7442503094326588,
+      "grad_norm": 3.59282723815335,
+      "learning_rate": 3.2387508056376724e-07,
+      "loss": 0.5914,
+      "step": 1616
+    },
+    {
+      "epoch": 0.7447108603667136,
+      "grad_norm": 3.3394681551065184,
+      "learning_rate": 3.2277622130830505e-07,
+      "loss": 0.5148,
+      "step": 1617
+    },
+    {
+      "epoch": 0.7451714113007686,
+      "grad_norm": 2.8985388546730535,
+      "learning_rate": 3.216788704915327e-07,
+      "loss": 0.5079,
+      "step": 1618
+    },
+    {
+      "epoch": 0.7456319622348234,
+      "grad_norm": 3.069882169066366,
+      "learning_rate": 3.2058303055767443e-07,
+      "loss": 0.4949,
+      "step": 1619
+    },
+    {
+      "epoch": 0.7460925131688783,
+      "grad_norm": 2.911216557138714,
+      "learning_rate": 3.19488703947588e-07,
+      "loss": 0.585,
+      "step": 1620
+    },
+    {
+      "epoch": 0.7465530641029331,
+      "grad_norm": 3.442946480558847,
+      "learning_rate": 3.1839589309876115e-07,
+      "loss": 0.544,
+      "step": 1621
+    },
+    {
+      "epoch": 0.747013615036988,
+      "grad_norm": 2.7674990139149847,
+      "learning_rate": 3.1730460044530573e-07,
+      "loss": 0.5363,
+      "step": 1622
+    },
+    {
+      "epoch": 0.7474741659710429,
+      "grad_norm": 3.229630271750196,
+      "learning_rate": 3.1621482841795124e-07,
+      "loss": 0.6189,
+      "step": 1623
+    },
+    {
+      "epoch": 0.7479347169050977,
+      "grad_norm": 3.1155759443087945,
+      "learning_rate": 3.151265794440404e-07,
+      "loss": 0.4911,
+      "step": 1624
+    },
+    {
+      "epoch": 0.7483952678391526,
+      "grad_norm": 3.4937874237026114,
+      "learning_rate": 3.140398559475244e-07,
+      "loss": 0.4647,
+      "step": 1625
+    },
+    {
+      "epoch": 0.7488558187732075,
+      "grad_norm": 3.2934599527181034,
+      "learning_rate": 3.129546603489548e-07,
+      "loss": 0.4893,
+      "step": 1626
+    },
+    {
+      "epoch": 0.7493163697072623,
+      "grad_norm": 3.0506555877752595,
+      "learning_rate": 3.1187099506548153e-07,
+      "loss": 0.5267,
+      "step": 1627
+    },
+    {
+      "epoch": 0.7497769206413172,
+      "grad_norm": 3.5830543718291015,
+      "learning_rate": 3.1078886251084525e-07,
+      "loss": 0.5325,
+      "step": 1628
+    },
+    {
+      "epoch": 0.750237471575372,
+      "grad_norm": 3.088278820265878,
+      "learning_rate": 3.0970826509537304e-07,
+      "loss": 0.5628,
+      "step": 1629
+    },
+    {
+      "epoch": 0.7506980225094269,
+      "grad_norm": 3.1594316317748294,
+      "learning_rate": 3.0862920522597167e-07,
+      "loss": 0.4871,
+      "step": 1630
+    },
+    {
+      "epoch": 0.7511585734434818,
+      "grad_norm": 2.8769280379230735,
+      "learning_rate": 3.075516853061244e-07,
+      "loss": 0.4936,
+      "step": 1631
+    },
+    {
+      "epoch": 0.7516191243775366,
+      "grad_norm": 3.4881331778008176,
+      "learning_rate": 3.0647570773588403e-07,
+      "loss": 0.5471,
+      "step": 1632
+    },
+    {
+      "epoch": 0.7520796753115915,
+      "grad_norm": 3.006655258177075,
+      "learning_rate": 3.0540127491186727e-07,
+      "loss": 0.4765,
+      "step": 1633
+    },
+    {
+      "epoch": 0.7525402262456463,
+      "grad_norm": 3.4773351804629296,
+      "learning_rate": 3.043283892272508e-07,
+      "loss": 0.5993,
+      "step": 1634
+    },
+    {
+      "epoch": 0.7530007771797013,
+      "grad_norm": 2.9508457470911256,
+      "learning_rate": 3.0325705307176564e-07,
+      "loss": 0.444,
+      "step": 1635
+    },
+    {
+      "epoch": 0.7534613281137561,
+      "grad_norm": 3.593179054371365,
+      "learning_rate": 3.0218726883168955e-07,
+      "loss": 0.5524,
+      "step": 1636
+    },
+    {
+      "epoch": 0.7539218790478109,
+      "grad_norm": 2.985359000550235,
+      "learning_rate": 3.011190388898464e-07,
+      "loss": 0.4384,
+      "step": 1637
+    },
+    {
+      "epoch": 0.7543824299818658,
+      "grad_norm": 3.0881433691468723,
+      "learning_rate": 3.0005236562559566e-07,
+      "loss": 0.5832,
+      "step": 1638
+    },
+    {
+      "epoch": 0.7548429809159207,
+      "grad_norm": 2.64976112453638,
+      "learning_rate": 2.989872514148298e-07,
+      "loss": 0.5254,
+      "step": 1639
+    },
+    {
+      "epoch": 0.7553035318499756,
+      "grad_norm": 3.178227072350045,
+      "learning_rate": 2.9792369862997046e-07,
+      "loss": 0.4896,
+      "step": 1640
+    },
+    {
+      "epoch": 0.7557640827840304,
+      "grad_norm": 3.1094527462428543,
+      "learning_rate": 2.9686170963995915e-07,
+      "loss": 0.5024,
+      "step": 1641
+    },
+    {
+      "epoch": 0.7562246337180852,
+      "grad_norm": 2.830372620158942,
+      "learning_rate": 2.958012868102553e-07,
+      "loss": 0.5466,
+      "step": 1642
+    },
+    {
+      "epoch": 0.7566851846521401,
+      "grad_norm": 3.5892656567533057,
+      "learning_rate": 2.9474243250283035e-07,
+      "loss": 0.5726,
+      "step": 1643
+    },
+    {
+      "epoch": 0.757145735586195,
+      "grad_norm": 3.1400502525766463,
+      "learning_rate": 2.936851490761606e-07,
+      "loss": 0.6686,
+      "step": 1644
+    },
+    {
+      "epoch": 0.7576062865202499,
+      "grad_norm": 3.611911685117865,
+      "learning_rate": 2.926294388852246e-07,
+      "loss": 0.531,
+      "step": 1645
+    },
+    {
+      "epoch": 0.7580668374543047,
+      "grad_norm": 2.9071023212444076,
+      "learning_rate": 2.9157530428149677e-07,
+      "loss": 0.446,
+      "step": 1646
+    },
+    {
+      "epoch": 0.7585273883883595,
+      "grad_norm": 3.573059215096367,
+      "learning_rate": 2.9052274761294094e-07,
+      "loss": 0.6017,
+      "step": 1647
+    },
+    {
+      "epoch": 0.7589879393224145,
+      "grad_norm": 2.926410495931059,
+      "learning_rate": 2.8947177122400737e-07,
+      "loss": 0.4499,
+      "step": 1648
+    },
+    {
+      "epoch": 0.7594484902564693,
+      "grad_norm": 2.9077271228177906,
+      "learning_rate": 2.8842237745562583e-07,
+      "loss": 0.4628,
+      "step": 1649
+    },
+    {
+      "epoch": 0.7599090411905242,
+      "grad_norm": 3.128424023474016,
+      "learning_rate": 2.873745686452017e-07,
+      "loss": 0.5119,
+      "step": 1650
+    },
+    {
+      "epoch": 0.760369592124579,
+      "grad_norm": 3.2061268862774464,
+      "learning_rate": 2.863283471266088e-07,
+      "loss": 0.5419,
+      "step": 1651
+    },
+    {
+      "epoch": 0.760830143058634,
+      "grad_norm": 3.176712368087586,
+      "learning_rate": 2.852837152301867e-07,
+      "loss": 0.575,
+      "step": 1652
+    },
+    {
+      "epoch": 0.7612906939926888,
+      "grad_norm": 3.002968313659636,
+      "learning_rate": 2.8424067528273374e-07,
+      "loss": 0.5412,
+      "step": 1653
+    },
+    {
+      "epoch": 0.7617512449267436,
+      "grad_norm": 3.3131721641797522,
+      "learning_rate": 2.83199229607502e-07,
+      "loss": 0.5517,
+      "step": 1654
+    },
+    {
+      "epoch": 0.7622117958607985,
+      "grad_norm": 3.0397564537569806,
+      "learning_rate": 2.821593805241932e-07,
+      "loss": 0.4976,
+      "step": 1655
+    },
+    {
+      "epoch": 0.7626723467948533,
+      "grad_norm": 3.3953414466743714,
+      "learning_rate": 2.811211303489527e-07,
+      "loss": 0.5132,
+      "step": 1656
+    },
+    {
+      "epoch": 0.7631328977289082,
+      "grad_norm": 3.2736894751575525,
+      "learning_rate": 2.8008448139436367e-07,
+      "loss": 0.5464,
+      "step": 1657
+    },
+    {
+      "epoch": 0.7635934486629631,
+      "grad_norm": 3.1651926004830915,
+      "learning_rate": 2.7904943596944373e-07,
+      "loss": 0.5585,
+      "step": 1658
+    },
+    {
+      "epoch": 0.7640539995970179,
+      "grad_norm": 3.036581859418946,
+      "learning_rate": 2.7801599637963893e-07,
+      "loss": 0.5441,
+      "step": 1659
+    },
+    {
+      "epoch": 0.7645145505310728,
+      "grad_norm": 2.625544113481615,
+      "learning_rate": 2.769841649268171e-07,
+      "loss": 0.5043,
+      "step": 1660
+    },
+    {
+      "epoch": 0.7649751014651277,
+      "grad_norm": 3.165210415457858,
+      "learning_rate": 2.759539439092655e-07,
+      "loss": 0.5323,
+      "step": 1661
+    },
+    {
+      "epoch": 0.7654356523991825,
+      "grad_norm": 2.968389660757054,
+      "learning_rate": 2.7492533562168407e-07,
+      "loss": 0.5542,
+      "step": 1662
+    },
+    {
+      "epoch": 0.7658962033332374,
+      "grad_norm": 3.28366116605155,
+      "learning_rate": 2.738983423551804e-07,
+      "loss": 0.4676,
+      "step": 1663
+    },
+    {
+      "epoch": 0.7663567542672922,
+      "grad_norm": 3.0790653400040715,
+      "learning_rate": 2.7287296639726443e-07,
+      "loss": 0.5762,
+      "step": 1664
+    },
+    {
+      "epoch": 0.7668173052013472,
+      "grad_norm": 2.8472233247954795,
+      "learning_rate": 2.7184921003184424e-07,
+      "loss": 0.5132,
+      "step": 1665
+    },
+    {
+      "epoch": 0.767277856135402,
+      "grad_norm": 3.173379604510871,
+      "learning_rate": 2.7082707553922067e-07,
+      "loss": 0.5762,
+      "step": 1666
+    },
+    {
+      "epoch": 0.7677384070694568,
+      "grad_norm": 3.3239774867727268,
+      "learning_rate": 2.698065651960809e-07,
+      "loss": 0.5437,
+      "step": 1667
+    },
+    {
+      "epoch": 0.7681989580035117,
+      "grad_norm": 3.004069248804067,
+      "learning_rate": 2.687876812754963e-07,
+      "loss": 0.5013,
+      "step": 1668
+    },
+    {
+      "epoch": 0.7686595089375665,
+      "grad_norm": 2.848925751252902,
+      "learning_rate": 2.67770426046914e-07,
+      "loss": 0.3852,
+      "step": 1669
+    },
+    {
+      "epoch": 0.7691200598716215,
+      "grad_norm": 3.47280346861256,
+      "learning_rate": 2.6675480177615326e-07,
+      "loss": 0.6775,
+      "step": 1670
+    },
+    {
+      "epoch": 0.7695806108056763,
+      "grad_norm": 3.2787012091514938,
+      "learning_rate": 2.6574081072540264e-07,
+      "loss": 0.575,
+      "step": 1671
+    },
+    {
+      "epoch": 0.7700411617397311,
+      "grad_norm": 2.7278525194039727,
+      "learning_rate": 2.647284551532104e-07,
+      "loss": 0.4094,
+      "step": 1672
+    },
+    {
+      "epoch": 0.770501712673786,
+      "grad_norm": 3.4300818176528063,
+      "learning_rate": 2.6371773731448357e-07,
+      "loss": 0.5578,
+      "step": 1673
+    },
+    {
+      "epoch": 0.7709622636078409,
+      "grad_norm": 2.968395762450879,
+      "learning_rate": 2.6270865946048084e-07,
+      "loss": 0.5684,
+      "step": 1674
+    },
+    {
+      "epoch": 0.7714228145418958,
+      "grad_norm": 2.8747901281611172,
+      "learning_rate": 2.617012238388077e-07,
+      "loss": 0.4909,
+      "step": 1675
+    },
+    {
+      "epoch": 0.7718833654759506,
+      "grad_norm": 3.1871301141813384,
+      "learning_rate": 2.60695432693412e-07,
+      "loss": 0.5484,
+      "step": 1676
+    },
+    {
+      "epoch": 0.7723439164100054,
+      "grad_norm": 3.2847684133576855,
+      "learning_rate": 2.596912882645792e-07,
+      "loss": 0.5747,
+      "step": 1677
+    },
+    {
+      "epoch": 0.7728044673440604,
+      "grad_norm": 2.6955957814746707,
+      "learning_rate": 2.5868879278892597e-07,
+      "loss": 0.4021,
+      "step": 1678
+    },
+    {
+      "epoch": 0.7732650182781152,
+      "grad_norm": 2.5964611847982786,
+      "learning_rate": 2.576879484993968e-07,
+      "loss": 0.5846,
+      "step": 1679
+    },
+    {
+      "epoch": 0.7737255692121701,
+      "grad_norm": 3.0346764723083868,
+      "learning_rate": 2.56688757625258e-07,
+      "loss": 0.4852,
+      "step": 1680
+    },
+    {
+      "epoch": 0.7741861201462249,
+      "grad_norm": 3.172623244649948,
+      "learning_rate": 2.5569122239209364e-07,
+      "loss": 0.5507,
+      "step": 1681
+    },
+    {
+      "epoch": 0.7746466710802797,
+      "grad_norm": 3.513693571783645,
+      "learning_rate": 2.54695345021799e-07,
+      "loss": 0.6611,
+      "step": 1682
+    },
+    {
+      "epoch": 0.7751072220143347,
+      "grad_norm": 2.9204689205282337,
+      "learning_rate": 2.537011277325777e-07,
+      "loss": 0.4998,
+      "step": 1683
+    },
+    {
+      "epoch": 0.7755677729483895,
+      "grad_norm": 2.8688540667247437,
+      "learning_rate": 2.527085727389354e-07,
+      "loss": 0.5582,
+      "step": 1684
+    },
+    {
+      "epoch": 0.7760283238824444,
+      "grad_norm": 3.254940468026647,
+      "learning_rate": 2.5171768225167465e-07,
+      "loss": 0.5264,
+      "step": 1685
+    },
+    {
+      "epoch": 0.7764888748164992,
+      "grad_norm": 3.319619529120737,
+      "learning_rate": 2.5072845847789126e-07,
+      "loss": 0.5844,
+      "step": 1686
+    },
+    {
+      "epoch": 0.7769494257505541,
+      "grad_norm": 2.711029642886499,
+      "learning_rate": 2.4974090362096843e-07,
+      "loss": 0.4466,
+      "step": 1687
+    },
+    {
+      "epoch": 0.777409976684609,
+      "grad_norm": 2.9892514572487947,
+      "learning_rate": 2.487550198805715e-07,
+      "loss": 0.5246,
+      "step": 1688
+    },
+    {
+      "epoch": 0.7778705276186638,
+      "grad_norm": 3.2361104141741213,
+      "learning_rate": 2.4777080945264416e-07,
+      "loss": 0.5495,
+      "step": 1689
+    },
+    {
+      "epoch": 0.7783310785527187,
+      "grad_norm": 2.778603000630699,
+      "learning_rate": 2.467882745294031e-07,
+      "loss": 0.43,
+      "step": 1690
+    },
+    {
+      "epoch": 0.7787916294867736,
+      "grad_norm": 2.9641786432879442,
+      "learning_rate": 2.458074172993324e-07,
+      "loss": 0.5534,
+      "step": 1691
+    },
+    {
+      "epoch": 0.7792521804208284,
+      "grad_norm": 3.0337502563475276,
+      "learning_rate": 2.4482823994717974e-07,
+      "loss": 0.556,
+      "step": 1692
+    },
+    {
+      "epoch": 0.7797127313548833,
+      "grad_norm": 3.0752220435659616,
+      "learning_rate": 2.4385074465395084e-07,
+      "loss": 0.536,
+      "step": 1693
+    },
+    {
+      "epoch": 0.7801732822889381,
+      "grad_norm": 3.1565462424883246,
+      "learning_rate": 2.4287493359690534e-07,
+      "loss": 0.4882,
+      "step": 1694
+    },
+    {
+      "epoch": 0.780633833222993,
+      "grad_norm": 3.0427946667754777,
+      "learning_rate": 2.4190080894955054e-07,
+      "loss": 0.4546,
+      "step": 1695
+    },
+    {
+      "epoch": 0.7810943841570479,
+      "grad_norm": 3.2264986971076155,
+      "learning_rate": 2.4092837288163805e-07,
+      "loss": 0.4948,
+      "step": 1696
+    },
+    {
+      "epoch": 0.7815549350911027,
+      "grad_norm": 3.35844053135339,
+      "learning_rate": 2.399576275591586e-07,
+      "loss": 0.5087,
+      "step": 1697
+    },
+    {
+      "epoch": 0.7820154860251576,
+      "grad_norm": 3.0112899189235116,
+      "learning_rate": 2.389885751443358e-07,
+      "loss": 0.5081,
+      "step": 1698
+    },
+    {
+      "epoch": 0.7824760369592124,
+      "grad_norm": 3.2352717237579722,
+      "learning_rate": 2.3802121779562446e-07,
+      "loss": 0.6003,
+      "step": 1699
+    },
+    {
+      "epoch": 0.7829365878932674,
+      "grad_norm": 2.671473345514583,
+      "learning_rate": 2.3705555766770203e-07,
+      "loss": 0.466,
+      "step": 1700
+    },
+    {
+      "epoch": 0.7833971388273222,
+      "grad_norm": 3.574193258639437,
+      "learning_rate": 2.3609159691146575e-07,
+      "loss": 0.5839,
+      "step": 1701
+    },
+    {
+      "epoch": 0.783857689761377,
+      "grad_norm": 3.051925522352705,
+      "learning_rate": 2.3512933767402942e-07,
+      "loss": 0.6309,
+      "step": 1702
+    },
+    {
+      "epoch": 0.7843182406954319,
+      "grad_norm": 3.5580346934856686,
+      "learning_rate": 2.3416878209871493e-07,
+      "loss": 0.5154,
+      "step": 1703
+    },
+    {
+      "epoch": 0.7847787916294868,
+      "grad_norm": 3.2002268979972905,
+      "learning_rate": 2.3320993232504993e-07,
+      "loss": 0.4721,
+      "step": 1704
+    },
+    {
+      "epoch": 0.7852393425635417,
+      "grad_norm": 2.9878864928285007,
+      "learning_rate": 2.3225279048876367e-07,
+      "loss": 0.5072,
+      "step": 1705
+    },
+    {
+      "epoch": 0.7856998934975965,
+      "grad_norm": 3.1381323142253685,
+      "learning_rate": 2.312973587217798e-07,
+      "loss": 0.6361,
+      "step": 1706
+    },
+    {
+      "epoch": 0.7861604444316513,
+      "grad_norm": 2.833292804782264,
+      "learning_rate": 2.3034363915221378e-07,
+      "loss": 0.4536,
+      "step": 1707
+    },
+    {
+      "epoch": 0.7866209953657062,
+      "grad_norm": 3.1215024593062015,
+      "learning_rate": 2.2939163390436732e-07,
+      "loss": 0.4223,
+      "step": 1708
+    },
+    {
+      "epoch": 0.7870815462997611,
+      "grad_norm": 3.311510890413361,
+      "learning_rate": 2.2844134509872292e-07,
+      "loss": 0.5819,
+      "step": 1709
+    },
+    {
+      "epoch": 0.787542097233816,
+      "grad_norm": 3.040556067928825,
+      "learning_rate": 2.2749277485194085e-07,
+      "loss": 0.5059,
+      "step": 1710
+    },
+    {
+      "epoch": 0.7880026481678708,
+      "grad_norm": 2.9822721827674985,
+      "learning_rate": 2.26545925276853e-07,
+      "loss": 0.4199,
+      "step": 1711
+    },
+    {
+      "epoch": 0.7884631991019256,
+      "grad_norm": 3.1686788844095637,
+      "learning_rate": 2.2560079848245905e-07,
+      "loss": 0.4875,
+      "step": 1712
+    },
+    {
+      "epoch": 0.7889237500359806,
+      "grad_norm": 3.3382646126754154,
+      "learning_rate": 2.2465739657392057e-07,
+      "loss": 0.5694,
+      "step": 1713
+    },
+    {
+      "epoch": 0.7893843009700354,
+      "grad_norm": 2.9076693356746084,
+      "learning_rate": 2.2371572165255792e-07,
+      "loss": 0.4706,
+      "step": 1714
+    },
+    {
+      "epoch": 0.7898448519040903,
+      "grad_norm": 3.925072525348147,
+      "learning_rate": 2.2277577581584473e-07,
+      "loss": 0.5798,
+      "step": 1715
+    },
+    {
+      "epoch": 0.7903054028381451,
+      "grad_norm": 3.484998071545751,
+      "learning_rate": 2.218375611574027e-07,
+      "loss": 0.5484,
+      "step": 1716
+    },
+    {
+      "epoch": 0.7907659537722,
+      "grad_norm": 3.301699935752327,
+      "learning_rate": 2.2090107976699802e-07,
+      "loss": 0.5115,
+      "step": 1717
+    },
+    {
+      "epoch": 0.7912265047062549,
+      "grad_norm": 3.042551513495106,
+      "learning_rate": 2.1996633373053653e-07,
+      "loss": 0.579,
+      "step": 1718
+    },
+    {
+      "epoch": 0.7916870556403097,
+      "grad_norm": 4.004445646052093,
+      "learning_rate": 2.190333251300578e-07,
+      "loss": 0.5797,
+      "step": 1719
+    },
+    {
+      "epoch": 0.7921476065743646,
+      "grad_norm": 3.0220642732410457,
+      "learning_rate": 2.1810205604373233e-07,
+      "loss": 0.5784,
+      "step": 1720
+    },
+    {
+      "epoch": 0.7926081575084194,
+      "grad_norm": 3.1050063154048204,
+      "learning_rate": 2.171725285458559e-07,
+      "loss": 0.4233,
+      "step": 1721
+    },
+    {
+      "epoch": 0.7930687084424743,
+      "grad_norm": 2.8115340027844966,
+      "learning_rate": 2.162447447068444e-07,
+      "loss": 0.5435,
+      "step": 1722
+    },
+    {
+      "epoch": 0.7935292593765292,
+      "grad_norm": 2.9604984251953526,
+      "learning_rate": 2.1531870659323082e-07,
+      "loss": 0.5504,
+      "step": 1723
+    },
+    {
+      "epoch": 0.793989810310584,
+      "grad_norm": 3.2405218813980166,
+      "learning_rate": 2.1439441626765943e-07,
+      "loss": 0.5433,
+      "step": 1724
+    },
+    {
+      "epoch": 0.7944503612446389,
+      "grad_norm": 3.1770006678350655,
+      "learning_rate": 2.1347187578888158e-07,
+      "loss": 0.6119,
+      "step": 1725
+    },
+    {
+      "epoch": 0.7949109121786938,
+      "grad_norm": 3.5617997893715945,
+      "learning_rate": 2.1255108721175065e-07,
+      "loss": 0.5371,
+      "step": 1726
+    },
+    {
+      "epoch": 0.7953714631127486,
+      "grad_norm": 3.6514613591140317,
+      "learning_rate": 2.1163205258721806e-07,
+      "loss": 0.6407,
+      "step": 1727
+    },
+    {
+      "epoch": 0.7958320140468035,
+      "grad_norm": 3.0047824295231127,
+      "learning_rate": 2.1071477396232894e-07,
+      "loss": 0.488,
+      "step": 1728
+    },
+    {
+      "epoch": 0.7962925649808583,
+      "grad_norm": 2.8092385249228293,
+      "learning_rate": 2.097992533802163e-07,
+      "loss": 0.5019,
+      "step": 1729
+    },
+    {
+      "epoch": 0.7967531159149133,
+      "grad_norm": 3.173996181660635,
+      "learning_rate": 2.0888549288009804e-07,
+      "loss": 0.6845,
+      "step": 1730
+    },
+    {
+      "epoch": 0.7972136668489681,
+      "grad_norm": 2.8459418296608145,
+      "learning_rate": 2.0797349449727163e-07,
+      "loss": 0.5215,
+      "step": 1731
+    },
+    {
+      "epoch": 0.7976742177830229,
+      "grad_norm": 2.867902063042974,
+      "learning_rate": 2.070632602631086e-07,
+      "loss": 0.5121,
+      "step": 1732
+    },
+    {
+      "epoch": 0.7981347687170778,
+      "grad_norm": 3.218011225763629,
+      "learning_rate": 2.0615479220505293e-07,
+      "loss": 0.5774,
+      "step": 1733
+    },
+    {
+      "epoch": 0.7985953196511326,
+      "grad_norm": 3.037535439229652,
+      "learning_rate": 2.05248092346613e-07,
+      "loss": 0.5455,
+      "step": 1734
+    },
+    {
+      "epoch": 0.7990558705851876,
+      "grad_norm": 3.44155141665967,
+      "learning_rate": 2.0434316270735875e-07,
+      "loss": 0.3846,
+      "step": 1735
+    },
+    {
+      "epoch": 0.7995164215192424,
+      "grad_norm": 3.251900855648779,
+      "learning_rate": 2.0344000530291872e-07,
+      "loss": 0.4974,
+      "step": 1736
+    },
+    {
+      "epoch": 0.7999769724532972,
+      "grad_norm": 3.3912040917359167,
+      "learning_rate": 2.025386221449722e-07,
+      "loss": 0.5486,
+      "step": 1737
+    },
+    {
+      "epoch": 0.8004375233873521,
+      "grad_norm": 3.289387489843882,
+      "learning_rate": 2.0163901524124771e-07,
+      "loss": 0.5502,
+      "step": 1738
+    },
+    {
+      "epoch": 0.800898074321407,
+      "grad_norm": 3.3531542710146884,
+      "learning_rate": 2.0074118659551697e-07,
+      "loss": 0.4939,
+      "step": 1739
+    },
+    {
+      "epoch": 0.8013586252554619,
+      "grad_norm": 3.0019863613387643,
+      "learning_rate": 1.9984513820759052e-07,
+      "loss": 0.4646,
+      "step": 1740
+    },
+    {
+      "epoch": 0.8018191761895167,
+      "grad_norm": 2.9999237180628726,
+      "learning_rate": 1.9895087207331417e-07,
+      "loss": 0.5051,
+      "step": 1741
+    },
+    {
+      "epoch": 0.8022797271235715,
+      "grad_norm": 3.440811419447699,
+      "learning_rate": 1.980583901845636e-07,
+      "loss": 0.5719,
+      "step": 1742
+    },
+    {
+      "epoch": 0.8027402780576265,
+      "grad_norm": 2.954781327967473,
+      "learning_rate": 1.9716769452924065e-07,
+      "loss": 0.4992,
+      "step": 1743
+    },
+    {
+      "epoch": 0.8032008289916813,
+      "grad_norm": 2.9449328433082433,
+      "learning_rate": 1.9627878709126778e-07,
+      "loss": 0.5873,
+      "step": 1744
+    },
+    {
+      "epoch": 0.8036613799257362,
+      "grad_norm": 2.9550948833942914,
+      "learning_rate": 1.9539166985058508e-07,
+      "loss": 0.5085,
+      "step": 1745
+    },
+    {
+      "epoch": 0.804121930859791,
+      "grad_norm": 3.0968335366037256,
+      "learning_rate": 1.945063447831452e-07,
+      "loss": 0.4837,
+      "step": 1746
+    },
+    {
+      "epoch": 0.8045824817938458,
+      "grad_norm": 2.933117571611147,
+      "learning_rate": 1.936228138609084e-07,
+      "loss": 0.5462,
+      "step": 1747
+    },
+    {
+      "epoch": 0.8050430327279008,
+      "grad_norm": 3.652807035479544,
+      "learning_rate": 1.92741079051839e-07,
+      "loss": 0.5419,
+      "step": 1748
+    },
+    {
+      "epoch": 0.8055035836619556,
+      "grad_norm": 2.906573757054981,
+      "learning_rate": 1.9186114231990104e-07,
+      "loss": 0.5128,
+      "step": 1749
+    },
+    {
+      "epoch": 0.8059641345960105,
+      "grad_norm": 2.947586128595587,
+      "learning_rate": 1.9098300562505264e-07,
+      "loss": 0.4349,
+      "step": 1750
+    },
+    {
+      "epoch": 0.8064246855300653,
+      "grad_norm": 3.0454260234507924,
+      "learning_rate": 1.901066709232434e-07,
+      "loss": 0.6054,
+      "step": 1751
+    },
+    {
+      "epoch": 0.8068852364641202,
+      "grad_norm": 3.411169435801757,
+      "learning_rate": 1.8923214016640898e-07,
+      "loss": 0.4776,
+      "step": 1752
+    },
+    {
+      "epoch": 0.8073457873981751,
+      "grad_norm": 3.2706182841611278,
+      "learning_rate": 1.8835941530246657e-07,
+      "loss": 0.5874,
+      "step": 1753
+    },
+    {
+      "epoch": 0.8078063383322299,
+      "grad_norm": 3.0129377224539513,
+      "learning_rate": 1.8748849827531133e-07,
+      "loss": 0.4817,
+      "step": 1754
+    },
+    {
+      "epoch": 0.8082668892662848,
+      "grad_norm": 3.185149212645243,
+      "learning_rate": 1.866193910248115e-07,
+      "loss": 0.4825,
+      "step": 1755
+    },
+    {
+      "epoch": 0.8087274402003397,
+      "grad_norm": 3.343736216544324,
+      "learning_rate": 1.857520954868047e-07,
+      "loss": 0.5353,
+      "step": 1756
+    },
+    {
+      "epoch": 0.8091879911343945,
+      "grad_norm": 2.7806290534872526,
+      "learning_rate": 1.848866135930922e-07,
+      "loss": 0.5085,
+      "step": 1757
+    },
+    {
+      "epoch": 0.8096485420684494,
+      "grad_norm": 3.0440711729049656,
+      "learning_rate": 1.8402294727143642e-07,
+      "loss": 0.4974,
+      "step": 1758
+    },
+    {
+      "epoch": 0.8101090930025042,
+      "grad_norm": 3.084951995273267,
+      "learning_rate": 1.831610984455557e-07,
+      "loss": 0.5661,
+      "step": 1759
+    },
+    {
+      "epoch": 0.8105696439365591,
+      "grad_norm": 3.0500734438430106,
+      "learning_rate": 1.8230106903511965e-07,
+      "loss": 0.4579,
+      "step": 1760
+    },
+    {
+      "epoch": 0.811030194870614,
+      "grad_norm": 3.2959846911138304,
+      "learning_rate": 1.814428609557458e-07,
+      "loss": 0.396,
+      "step": 1761
+    },
+    {
+      "epoch": 0.8114907458046688,
+      "grad_norm": 3.4228830494829583,
+      "learning_rate": 1.805864761189949e-07,
+      "loss": 0.6473,
+      "step": 1762
+    },
+    {
+      "epoch": 0.8119512967387237,
+      "grad_norm": 3.5275669816599247,
+      "learning_rate": 1.7973191643236574e-07,
+      "loss": 0.5014,
+      "step": 1763
+    },
+    {
+      "epoch": 0.8124118476727785,
+      "grad_norm": 2.655456203502326,
+      "learning_rate": 1.7887918379929356e-07,
+      "loss": 0.4284,
+      "step": 1764
+    },
+    {
+      "epoch": 0.8128723986068335,
+      "grad_norm": 3.119374485962208,
+      "learning_rate": 1.780282801191425e-07,
+      "loss": 0.5117,
+      "step": 1765
+    },
+    {
+      "epoch": 0.8133329495408883,
+      "grad_norm": 3.603722648362608,
+      "learning_rate": 1.771792072872028e-07,
+      "loss": 0.5251,
+      "step": 1766
+    },
+    {
+      "epoch": 0.8137935004749431,
+      "grad_norm": 3.636072681634815,
+      "learning_rate": 1.7633196719468846e-07,
+      "loss": 0.5528,
+      "step": 1767
+    },
+    {
+      "epoch": 0.814254051408998,
+      "grad_norm": 2.9215076629877563,
+      "learning_rate": 1.7548656172872922e-07,
+      "loss": 0.6344,
+      "step": 1768
+    },
+    {
+      "epoch": 0.8147146023430529,
+      "grad_norm": 3.093587532258315,
+      "learning_rate": 1.746429927723696e-07,
+      "loss": 0.5363,
+      "step": 1769
+    },
+    {
+      "epoch": 0.8151751532771078,
+      "grad_norm": 3.202770682033625,
+      "learning_rate": 1.738012622045635e-07,
+      "loss": 0.4939,
+      "step": 1770
+    },
+    {
+      "epoch": 0.8156357042111626,
+      "grad_norm": 3.0790051489322945,
+      "learning_rate": 1.7296137190016913e-07,
+      "loss": 0.5565,
+      "step": 1771
+    },
+    {
+      "epoch": 0.8160962551452174,
+      "grad_norm": 3.522068919897294,
+      "learning_rate": 1.7212332372994654e-07,
+      "loss": 0.6042,
+      "step": 1772
+    },
+    {
+      "epoch": 0.8165568060792723,
+      "grad_norm": 2.947521873551182,
+      "learning_rate": 1.7128711956055274e-07,
+      "loss": 0.5351,
+      "step": 1773
+    },
+    {
+      "epoch": 0.8170173570133272,
+      "grad_norm": 3.197903959240655,
+      "learning_rate": 1.7045276125453645e-07,
+      "loss": 0.5098,
+      "step": 1774
+    },
+    {
+      "epoch": 0.8174779079473821,
+      "grad_norm": 3.3416585809701833,
+      "learning_rate": 1.6962025067033604e-07,
+      "loss": 0.5499,
+      "step": 1775
+    },
+    {
+      "epoch": 0.8179384588814369,
+      "grad_norm": 2.6718382039429756,
+      "learning_rate": 1.6878958966227363e-07,
+      "loss": 0.4988,
+      "step": 1776
+    },
+    {
+      "epoch": 0.8183990098154917,
+      "grad_norm": 2.79487789143777,
+      "learning_rate": 1.6796078008055225e-07,
+      "loss": 0.4255,
+      "step": 1777
+    },
+    {
+      "epoch": 0.8188595607495467,
+      "grad_norm": 2.693027511901247,
+      "learning_rate": 1.671338237712502e-07,
+      "loss": 0.5501,
+      "step": 1778
+    },
+    {
+      "epoch": 0.8193201116836015,
+      "grad_norm": 2.834140007806073,
+      "learning_rate": 1.6630872257631834e-07,
+      "loss": 0.3992,
+      "step": 1779
+    },
+    {
+      "epoch": 0.8197806626176564,
+      "grad_norm": 2.9518201333896883,
+      "learning_rate": 1.6548547833357573e-07,
+      "loss": 0.4642,
+      "step": 1780
+    },
+    {
+      "epoch": 0.8202412135517112,
+      "grad_norm": 3.1159470952195636,
+      "learning_rate": 1.6466409287670468e-07,
+      "loss": 0.5313,
+      "step": 1781
+    },
+    {
+      "epoch": 0.8207017644857661,
+      "grad_norm": 3.257480140561922,
+      "learning_rate": 1.638445680352476e-07,
+      "loss": 0.5634,
+      "step": 1782
+    },
+    {
+      "epoch": 0.821162315419821,
+      "grad_norm": 3.423203304235,
+      "learning_rate": 1.6302690563460288e-07,
+      "loss": 0.5939,
+      "step": 1783
+    },
+    {
+      "epoch": 0.8216228663538758,
+      "grad_norm": 2.954374172187009,
+      "learning_rate": 1.6221110749601973e-07,
+      "loss": 0.3948,
+      "step": 1784
+    },
+    {
+      "epoch": 0.8220834172879307,
+      "grad_norm": 2.9485459890560994,
+      "learning_rate": 1.613971754365957e-07,
+      "loss": 0.483,
+      "step": 1785
+    },
+    {
+      "epoch": 0.8225439682219855,
+      "grad_norm": 3.052068636325435,
+      "learning_rate": 1.6058511126927176e-07,
+      "loss": 0.6306,
+      "step": 1786
+    },
+    {
+      "epoch": 0.8230045191560404,
+      "grad_norm": 3.276425368352071,
+      "learning_rate": 1.5977491680282762e-07,
+      "loss": 0.5366,
+      "step": 1787
+    },
+    {
+      "epoch": 0.8234650700900953,
+      "grad_norm": 3.322933684622075,
+      "learning_rate": 1.589665938418795e-07,
+      "loss": 0.5081,
+      "step": 1788
+    },
+    {
+      "epoch": 0.8239256210241501,
+      "grad_norm": 3.119742625969975,
+      "learning_rate": 1.581601441868743e-07,
+      "loss": 0.5368,
+      "step": 1789
+    },
+    {
+      "epoch": 0.824386171958205,
+      "grad_norm": 2.9653697872519036,
+      "learning_rate": 1.5735556963408693e-07,
+      "loss": 0.6565,
+      "step": 1790
+    },
+    {
+      "epoch": 0.8248467228922599,
+      "grad_norm": 3.373344219441705,
+      "learning_rate": 1.5655287197561495e-07,
+      "loss": 0.6218,
+      "step": 1791
+    },
+    {
+      "epoch": 0.8253072738263147,
+      "grad_norm": 3.1421457828079795,
+      "learning_rate": 1.5575205299937599e-07,
+      "loss": 0.5257,
+      "step": 1792
+    },
+    {
+      "epoch": 0.8257678247603696,
+      "grad_norm": 2.9754341538820426,
+      "learning_rate": 1.549531144891032e-07,
+      "loss": 0.5216,
+      "step": 1793
+    },
+    {
+      "epoch": 0.8262283756944244,
+      "grad_norm": 3.744224931739949,
+      "learning_rate": 1.5415605822434053e-07,
+      "loss": 0.5341,
+      "step": 1794
+    },
+    {
+      "epoch": 0.8266889266284794,
+      "grad_norm": 3.312132421784843,
+      "learning_rate": 1.5336088598043995e-07,
+      "loss": 0.4812,
+      "step": 1795
+    },
+    {
+      "epoch": 0.8271494775625342,
+      "grad_norm": 3.1823045221187805,
+      "learning_rate": 1.5256759952855737e-07,
+      "loss": 0.6088,
+      "step": 1796
+    },
+    {
+      "epoch": 0.827610028496589,
+      "grad_norm": 3.144738416475895,
+      "learning_rate": 1.5177620063564712e-07,
+      "loss": 0.626,
+      "step": 1797
+    },
+    {
+      "epoch": 0.8280705794306439,
+      "grad_norm": 3.09856360625291,
+      "learning_rate": 1.5098669106446026e-07,
+      "loss": 0.5281,
+      "step": 1798
+    },
+    {
+      "epoch": 0.8285311303646987,
+      "grad_norm": 2.8488512325058206,
+      "learning_rate": 1.5019907257353925e-07,
+      "loss": 0.4859,
+      "step": 1799
+    },
+    {
+      "epoch": 0.8289916812987537,
+      "grad_norm": 3.1008236444549166,
+      "learning_rate": 1.4941334691721474e-07,
+      "loss": 0.5706,
+      "step": 1800
+    },
+    {
+      "epoch": 0.8294522322328085,
+      "grad_norm": 3.5681942314685804,
+      "learning_rate": 1.4862951584560034e-07,
+      "loss": 0.5725,
+      "step": 1801
+    },
+    {
+      "epoch": 0.8299127831668633,
+      "grad_norm": 3.479007930611103,
+      "learning_rate": 1.4784758110459073e-07,
+      "loss": 0.455,
+      "step": 1802
+    },
+    {
+      "epoch": 0.8303733341009182,
+      "grad_norm": 3.6942391140034454,
+      "learning_rate": 1.4706754443585644e-07,
+      "loss": 0.5998,
+      "step": 1803
+    },
+    {
+      "epoch": 0.8308338850349731,
+      "grad_norm": 3.0399020070035414,
+      "learning_rate": 1.4628940757683972e-07,
+      "loss": 0.531,
+      "step": 1804
+    },
+    {
+      "epoch": 0.831294435969028,
+      "grad_norm": 3.3824973071742668,
+      "learning_rate": 1.4551317226075176e-07,
+      "loss": 0.632,
+      "step": 1805
+    },
+    {
+      "epoch": 0.8317549869030828,
+      "grad_norm": 3.0111947391524474,
+      "learning_rate": 1.4473884021656858e-07,
+      "loss": 0.4523,
+      "step": 1806
+    },
+    {
+      "epoch": 0.8322155378371376,
+      "grad_norm": 3.245242080771524,
+      "learning_rate": 1.4396641316902558e-07,
+      "loss": 0.5705,
+      "step": 1807
+    },
+    {
+      "epoch": 0.8326760887711926,
+      "grad_norm": 3.101099899015713,
+      "learning_rate": 1.431958928386169e-07,
+      "loss": 0.5403,
+      "step": 1808
+    },
+    {
+      "epoch": 0.8331366397052474,
+      "grad_norm": 3.4188190961039213,
+      "learning_rate": 1.4242728094158807e-07,
+      "loss": 0.5086,
+      "step": 1809
+    },
+    {
+      "epoch": 0.8335971906393023,
+      "grad_norm": 3.179675044847406,
+      "learning_rate": 1.41660579189934e-07,
+      "loss": 0.4884,
+      "step": 1810
+    },
+    {
+      "epoch": 0.8340577415733571,
+      "grad_norm": 3.225968066840041,
+      "learning_rate": 1.4089578929139635e-07,
+      "loss": 0.558,
+      "step": 1811
+    },
+    {
+      "epoch": 0.8345182925074119,
+      "grad_norm": 2.9209325285695433,
+      "learning_rate": 1.4013291294945652e-07,
+      "loss": 0.5264,
+      "step": 1812
+    },
+    {
+      "epoch": 0.8349788434414669,
+      "grad_norm": 3.1770764013810244,
+      "learning_rate": 1.3937195186333483e-07,
+      "loss": 0.4391,
+      "step": 1813
+    },
+    {
+      "epoch": 0.8354393943755217,
+      "grad_norm": 3.5252276429059015,
+      "learning_rate": 1.3861290772798552e-07,
+      "loss": 0.6889,
+      "step": 1814
+    },
+    {
+      "epoch": 0.8358999453095766,
+      "grad_norm": 3.334270364364526,
+      "learning_rate": 1.378557822340922e-07,
+      "loss": 0.5445,
+      "step": 1815
+    },
+    {
+      "epoch": 0.8363604962436314,
+      "grad_norm": 2.8274203446918955,
+      "learning_rate": 1.3710057706806588e-07,
+      "loss": 0.4528,
+      "step": 1816
+    },
+    {
+      "epoch": 0.8368210471776864,
+      "grad_norm": 2.970687258829064,
+      "learning_rate": 1.3634729391204003e-07,
+      "loss": 0.4576,
+      "step": 1817
+    },
+    {
+      "epoch": 0.8372815981117412,
+      "grad_norm": 3.0526123989623404,
+      "learning_rate": 1.355959344438665e-07,
+      "loss": 0.4776,
+      "step": 1818
+    },
+    {
+      "epoch": 0.837742149045796,
+      "grad_norm": 3.108639242698878,
+      "learning_rate": 1.3484650033711308e-07,
+      "loss": 0.505,
+      "step": 1819
+    },
+    {
+      "epoch": 0.8382026999798509,
+      "grad_norm": 3.568732239359027,
+      "learning_rate": 1.3409899326105856e-07,
+      "loss": 0.6024,
+      "step": 1820
+    },
+    {
+      "epoch": 0.8386632509139058,
+      "grad_norm": 3.6924487796854875,
+      "learning_rate": 1.3335341488068997e-07,
+      "loss": 0.5641,
+      "step": 1821
+    },
+    {
+      "epoch": 0.8391238018479606,
+      "grad_norm": 2.7392248909998833,
+      "learning_rate": 1.3260976685669767e-07,
+      "loss": 0.4318,
+      "step": 1822
+    },
+    {
+      "epoch": 0.8395843527820155,
+      "grad_norm": 3.0029706301333032,
+      "learning_rate": 1.3186805084547292e-07,
+      "loss": 0.5296,
+      "step": 1823
+    },
+    {
+      "epoch": 0.8400449037160703,
+      "grad_norm": 2.837543312679349,
+      "learning_rate": 1.3112826849910374e-07,
+      "loss": 0.6362,
+      "step": 1824
+    },
+    {
+      "epoch": 0.8405054546501252,
+      "grad_norm": 3.625903809155881,
+      "learning_rate": 1.303904214653705e-07,
+      "loss": 0.637,
+      "step": 1825
+    },
+    {
+      "epoch": 0.8409660055841801,
+      "grad_norm": 3.0363996469171393,
+      "learning_rate": 1.2965451138774342e-07,
+      "loss": 0.4819,
+      "step": 1826
+    },
+    {
+      "epoch": 0.841426556518235,
+      "grad_norm": 2.98767711754336,
+      "learning_rate": 1.2892053990537855e-07,
+      "loss": 0.416,
+      "step": 1827
+    },
+    {
+      "epoch": 0.8418871074522898,
+      "grad_norm": 3.368814353904181,
+      "learning_rate": 1.2818850865311304e-07,
+      "loss": 0.4946,
+      "step": 1828
+    },
+    {
+      "epoch": 0.8423476583863446,
+      "grad_norm": 2.866982470674382,
+      "learning_rate": 1.2745841926146328e-07,
+      "loss": 0.6318,
+      "step": 1829
+    },
+    {
+      "epoch": 0.8428082093203996,
+      "grad_norm": 3.1759522349648397,
+      "learning_rate": 1.2673027335662023e-07,
+      "loss": 0.5292,
+      "step": 1830
+    },
+    {
+      "epoch": 0.8432687602544544,
+      "grad_norm": 3.6119453846451184,
+      "learning_rate": 1.2600407256044543e-07,
+      "loss": 0.5958,
+      "step": 1831
+    },
+    {
+      "epoch": 0.8437293111885092,
+      "grad_norm": 3.0612740647541714,
+      "learning_rate": 1.2527981849046855e-07,
+      "loss": 0.53,
+      "step": 1832
+    },
+    {
+      "epoch": 0.8441898621225641,
+      "grad_norm": 3.0373834058484523,
+      "learning_rate": 1.245575127598828e-07,
+      "loss": 0.4711,
+      "step": 1833
+    },
+    {
+      "epoch": 0.844650413056619,
+      "grad_norm": 3.368821133890806,
+      "learning_rate": 1.2383715697754194e-07,
+      "loss": 0.5848,
+      "step": 1834
+    },
+    {
+      "epoch": 0.8451109639906739,
+      "grad_norm": 2.6277881136703476,
+      "learning_rate": 1.23118752747956e-07,
+      "loss": 0.4662,
+      "step": 1835
+    },
+    {
+      "epoch": 0.8455715149247287,
+      "grad_norm": 3.3152498659397733,
+      "learning_rate": 1.224023016712886e-07,
+      "loss": 0.5397,
+      "step": 1836
+    },
+    {
+      "epoch": 0.8460320658587835,
+      "grad_norm": 2.820968906592374,
+      "learning_rate": 1.2168780534335288e-07,
+      "loss": 0.5026,
+      "step": 1837
+    },
+    {
+      "epoch": 0.8464926167928384,
+      "grad_norm": 3.677672748447081,
+      "learning_rate": 1.2097526535560732e-07,
+      "loss": 0.6399,
+      "step": 1838
+    },
+    {
+      "epoch": 0.8469531677268933,
+      "grad_norm": 3.009389680786758,
+      "learning_rate": 1.2026468329515415e-07,
+      "loss": 0.4806,
+      "step": 1839
+    },
+    {
+      "epoch": 0.8474137186609482,
+      "grad_norm": 3.3016771762707444,
+      "learning_rate": 1.1955606074473368e-07,
+      "loss": 0.5922,
+      "step": 1840
+    },
+    {
+      "epoch": 0.847874269595003,
+      "grad_norm": 3.161579889786646,
+      "learning_rate": 1.1884939928272108e-07,
+      "loss": 0.587,
+      "step": 1841
+    },
+    {
+      "epoch": 0.8483348205290578,
+      "grad_norm": 3.490485543641311,
+      "learning_rate": 1.1814470048312508e-07,
+      "loss": 0.4875,
+      "step": 1842
+    },
+    {
+      "epoch": 0.8487953714631128,
+      "grad_norm": 3.106076702165179,
+      "learning_rate": 1.1744196591558153e-07,
+      "loss": 0.4083,
+      "step": 1843
+    },
+    {
+      "epoch": 0.8492559223971676,
+      "grad_norm": 3.138124379295683,
+      "learning_rate": 1.167411971453509e-07,
+      "loss": 0.5066,
+      "step": 1844
+    },
+    {
+      "epoch": 0.8497164733312225,
+      "grad_norm": 2.9058322118896642,
+      "learning_rate": 1.1604239573331653e-07,
+      "loss": 0.4712,
+      "step": 1845
+    },
+    {
+      "epoch": 0.8501770242652773,
+      "grad_norm": 3.2851181268981398,
+      "learning_rate": 1.1534556323597821e-07,
+      "loss": 0.5974,
+      "step": 1846
+    },
+    {
+      "epoch": 0.8506375751993323,
+      "grad_norm": 3.000290790230382,
+      "learning_rate": 1.1465070120545106e-07,
+      "loss": 0.4775,
+      "step": 1847
+    },
+    {
+      "epoch": 0.8510981261333871,
+      "grad_norm": 3.3447028114933377,
+      "learning_rate": 1.1395781118946124e-07,
+      "loss": 0.59,
+      "step": 1848
+    },
+    {
+      "epoch": 0.8515586770674419,
+      "grad_norm": 2.6080519147762393,
+      "learning_rate": 1.1326689473134166e-07,
+      "loss": 0.4667,
+      "step": 1849
+    },
+    {
+      "epoch": 0.8520192280014968,
+      "grad_norm": 2.9008718617685556,
+      "learning_rate": 1.1257795337003007e-07,
+      "loss": 0.5282,
+      "step": 1850
+    },
+    {
+      "epoch": 0.8524797789355516,
+      "grad_norm": 3.169497217996663,
+      "learning_rate": 1.1189098864006486e-07,
+      "loss": 0.5051,
+      "step": 1851
+    },
+    {
+      "epoch": 0.8529403298696066,
+      "grad_norm": 3.215162842301655,
+      "learning_rate": 1.112060020715817e-07,
+      "loss": 0.6173,
+      "step": 1852
+    },
+    {
+      "epoch": 0.8534008808036614,
+      "grad_norm": 3.42236331926524,
+      "learning_rate": 1.1052299519030961e-07,
+      "loss": 0.48,
+      "step": 1853
+    },
+    {
+      "epoch": 0.8538614317377162,
+      "grad_norm": 3.045553200483245,
+      "learning_rate": 1.0984196951756863e-07,
+      "loss": 0.5809,
+      "step": 1854
+    },
+    {
+      "epoch": 0.8543219826717711,
+      "grad_norm": 3.0183531859561543,
+      "learning_rate": 1.0916292657026616e-07,
+      "loss": 0.5366,
+      "step": 1855
+    },
+    {
+      "epoch": 0.854782533605826,
+      "grad_norm": 3.233752941923723,
+      "learning_rate": 1.084858678608922e-07,
+      "loss": 0.5023,
+      "step": 1856
+    },
+    {
+      "epoch": 0.8552430845398808,
+      "grad_norm": 3.7031466518660467,
+      "learning_rate": 1.078107948975181e-07,
+      "loss": 0.5871,
+      "step": 1857
+    },
+    {
+      "epoch": 0.8557036354739357,
+      "grad_norm": 3.231664196992226,
+      "learning_rate": 1.0713770918379206e-07,
+      "loss": 0.5641,
+      "step": 1858
+    },
+    {
+      "epoch": 0.8561641864079905,
+      "grad_norm": 2.8655131047972344,
+      "learning_rate": 1.0646661221893538e-07,
+      "loss": 0.5098,
+      "step": 1859
+    },
+    {
+      "epoch": 0.8566247373420455,
+      "grad_norm": 3.01957958338486,
+      "learning_rate": 1.0579750549773992e-07,
+      "loss": 0.5884,
+      "step": 1860
+    },
+    {
+      "epoch": 0.8570852882761003,
+      "grad_norm": 2.8146657597507017,
+      "learning_rate": 1.0513039051056504e-07,
+      "loss": 0.5192,
+      "step": 1861
+    },
+    {
+      "epoch": 0.8575458392101551,
+      "grad_norm": 3.579182844425862,
+      "learning_rate": 1.0446526874333262e-07,
+      "loss": 0.4323,
+      "step": 1862
+    },
+    {
+      "epoch": 0.85800639014421,
+      "grad_norm": 3.1443912809179,
+      "learning_rate": 1.0380214167752588e-07,
+      "loss": 0.5651,
+      "step": 1863
+    },
+    {
+      "epoch": 0.8584669410782648,
+      "grad_norm": 3.1366442156691607,
+      "learning_rate": 1.0314101079018456e-07,
+      "loss": 0.597,
+      "step": 1864
+    },
+    {
+      "epoch": 0.8589274920123198,
+      "grad_norm": 3.0110739023780178,
+      "learning_rate": 1.0248187755390247e-07,
+      "loss": 0.4957,
+      "step": 1865
+    },
+    {
+      "epoch": 0.8593880429463746,
+      "grad_norm": 3.1211140110824656,
+      "learning_rate": 1.0182474343682346e-07,
+      "loss": 0.5807,
+      "step": 1866
+    },
+    {
+      "epoch": 0.8598485938804294,
+      "grad_norm": 3.196850276682387,
+      "learning_rate": 1.0116960990263879e-07,
+      "loss": 0.432,
+      "step": 1867
+    },
+    {
+      "epoch": 0.8603091448144843,
+      "grad_norm": 3.0649211283399143,
+      "learning_rate": 1.0051647841058385e-07,
+      "loss": 0.4683,
+      "step": 1868
+    },
+    {
+      "epoch": 0.8607696957485392,
+      "grad_norm": 3.0704976261787276,
+      "learning_rate": 9.986535041543409e-08,
+      "loss": 0.5084,
+      "step": 1869
+    },
+    {
+      "epoch": 0.8612302466825941,
+      "grad_norm": 2.8430066136777223,
+      "learning_rate": 9.921622736750345e-08,
+      "loss": 0.4893,
+      "step": 1870
+    },
+    {
+      "epoch": 0.8616907976166489,
+      "grad_norm": 3.1871422194117915,
+      "learning_rate": 9.856911071263918e-08,
+      "loss": 0.5582,
+      "step": 1871
+    },
+    {
+      "epoch": 0.8621513485507037,
+      "grad_norm": 3.1695517066788335,
+      "learning_rate": 9.792400189221927e-08,
+      "loss": 0.5491,
+      "step": 1872
+    },
+    {
+      "epoch": 0.8626118994847587,
+      "grad_norm": 3.192178213736622,
+      "learning_rate": 9.72809023431509e-08,
+      "loss": 0.5684,
+      "step": 1873
+    },
+    {
+      "epoch": 0.8630724504188135,
+      "grad_norm": 3.5963603776479833,
+      "learning_rate": 9.663981349786443e-08,
+      "loss": 0.5006,
+      "step": 1874
+    },
+    {
+      "epoch": 0.8635330013528684,
+      "grad_norm": 2.9762857754260996,
+      "learning_rate": 9.600073678431186e-08,
+      "loss": 0.5183,
+      "step": 1875
+    },
+    {
+      "epoch": 0.8639935522869232,
+      "grad_norm": 3.512211825383684,
+      "learning_rate": 9.53636736259642e-08,
+      "loss": 0.5459,
+      "step": 1876
+    },
+    {
+      "epoch": 0.864454103220978,
+      "grad_norm": 2.959882910299774,
+      "learning_rate": 9.472862544180659e-08,
+      "loss": 0.538,
+      "step": 1877
+    },
+    {
+      "epoch": 0.864914654155033,
+      "grad_norm": 3.201929324826627,
+      "learning_rate": 9.409559364633646e-08,
+      "loss": 0.4949,
+      "step": 1878
+    },
+    {
+      "epoch": 0.8653752050890878,
+      "grad_norm": 3.2818565007848703,
+      "learning_rate": 9.346457964956023e-08,
+      "loss": 0.5904,
+      "step": 1879
+    },
+    {
+      "epoch": 0.8658357560231427,
+      "grad_norm": 3.1114338599754374,
+      "learning_rate": 9.283558485698894e-08,
+      "loss": 0.4755,
+      "step": 1880
+    },
+    {
+      "epoch": 0.8662963069571975,
+      "grad_norm": 2.923497203537618,
+      "learning_rate": 9.220861066963715e-08,
+      "loss": 0.4812,
+      "step": 1881
+    },
+    {
+      "epoch": 0.8667568578912525,
+      "grad_norm": 3.199024742719346,
+      "learning_rate": 9.158365848401817e-08,
+      "loss": 0.546,
+      "step": 1882
+    },
+    {
+      "epoch": 0.8672174088253073,
+      "grad_norm": 2.9843247776927257,
+      "learning_rate": 9.096072969214197e-08,
+      "loss": 0.543,
+      "step": 1883
+    },
+    {
+      "epoch": 0.8676779597593621,
+      "grad_norm": 3.0691692564049653,
+      "learning_rate": 9.0339825681511e-08,
+      "loss": 0.5088,
+      "step": 1884
+    },
+    {
+      "epoch": 0.868138510693417,
+      "grad_norm": 3.383333358394925,
+      "learning_rate": 8.972094783511807e-08,
+      "loss": 0.6416,
+      "step": 1885
+    },
+    {
+      "epoch": 0.8685990616274719,
+      "grad_norm": 3.0949849503575773,
+      "learning_rate": 8.910409753144344e-08,
+      "loss": 0.4999,
+      "step": 1886
+    },
+    {
+      "epoch": 0.8690596125615268,
+      "grad_norm": 3.273519298979524,
+      "learning_rate": 8.848927614445011e-08,
+      "loss": 0.5933,
+      "step": 1887
+    },
+    {
+      "epoch": 0.8695201634955816,
+      "grad_norm": 3.0038213358417183,
+      "learning_rate": 8.787648504358291e-08,
+      "loss": 0.4551,
+      "step": 1888
+    },
+    {
+      "epoch": 0.8699807144296364,
+      "grad_norm": 3.1769567170597615,
+      "learning_rate": 8.726572559376433e-08,
+      "loss": 0.6628,
+      "step": 1889
+    },
+    {
+      "epoch": 0.8704412653636913,
+      "grad_norm": 3.3453327234543067,
+      "learning_rate": 8.665699915539094e-08,
+      "loss": 0.656,
+      "step": 1890
+    },
+    {
+      "epoch": 0.8709018162977462,
+      "grad_norm": 3.0704691251867513,
+      "learning_rate": 8.605030708433147e-08,
+      "loss": 0.5198,
+      "step": 1891
+    },
+    {
+      "epoch": 0.871362367231801,
+      "grad_norm": 3.019643595841549,
+      "learning_rate": 8.544565073192367e-08,
+      "loss": 0.4624,
+      "step": 1892
+    },
+    {
+      "epoch": 0.8718229181658559,
+      "grad_norm": 3.08664418112486,
+      "learning_rate": 8.484303144497007e-08,
+      "loss": 0.49,
+      "step": 1893
+    },
+    {
+      "epoch": 0.8722834690999107,
+      "grad_norm": 3.262560025269264,
+      "learning_rate": 8.424245056573653e-08,
+      "loss": 0.5118,
+      "step": 1894
+    },
+    {
+      "epoch": 0.8727440200339657,
+      "grad_norm": 3.0452114320652086,
+      "learning_rate": 8.364390943194855e-08,
+      "loss": 0.5284,
+      "step": 1895
+    },
+    {
+      "epoch": 0.8732045709680205,
+      "grad_norm": 3.2005415403238366,
+      "learning_rate": 8.304740937678833e-08,
+      "loss": 0.6092,
+      "step": 1896
+    },
+    {
+      "epoch": 0.8736651219020753,
+      "grad_norm": 3.1426226729757794,
+      "learning_rate": 8.245295172889121e-08,
+      "loss": 0.4615,
+      "step": 1897
+    },
+    {
+      "epoch": 0.8741256728361302,
+      "grad_norm": 2.9393983445977403,
+      "learning_rate": 8.186053781234414e-08,
+      "loss": 0.4892,
+      "step": 1898
+    },
+    {
+      "epoch": 0.8745862237701851,
+      "grad_norm": 2.745952741647922,
+      "learning_rate": 8.12701689466816e-08,
+      "loss": 0.5493,
+      "step": 1899
+    },
+    {
+      "epoch": 0.87504677470424,
+      "grad_norm": 2.948466311290946,
+      "learning_rate": 8.068184644688248e-08,
+      "loss": 0.474,
+      "step": 1900
+    },
+    {
+      "epoch": 0.8755073256382948,
+      "grad_norm": 3.172245629853397,
+      "learning_rate": 8.009557162336822e-08,
+      "loss": 0.5008,
+      "step": 1901
+    },
+    {
+      "epoch": 0.8759678765723496,
+      "grad_norm": 3.2352168665011853,
+      "learning_rate": 7.951134578199925e-08,
+      "loss": 0.4891,
+      "step": 1902
+    },
+    {
+      "epoch": 0.8764284275064045,
+      "grad_norm": 3.2067263063797866,
+      "learning_rate": 7.892917022407153e-08,
+      "loss": 0.5522,
+      "step": 1903
+    },
+    {
+      "epoch": 0.8768889784404594,
+      "grad_norm": 3.389480208236002,
+      "learning_rate": 7.834904624631523e-08,
+      "loss": 0.6624,
+      "step": 1904
+    },
+    {
+      "epoch": 0.8773495293745143,
+      "grad_norm": 3.2052256238249903,
+      "learning_rate": 7.777097514089014e-08,
+      "loss": 0.6098,
+      "step": 1905
+    },
+    {
+      "epoch": 0.8778100803085691,
+      "grad_norm": 3.2221846053421337,
+      "learning_rate": 7.719495819538324e-08,
+      "loss": 0.5657,
+      "step": 1906
+    },
+    {
+      "epoch": 0.878270631242624,
+      "grad_norm": 3.018523302700844,
+      "learning_rate": 7.66209966928072e-08,
+      "loss": 0.4995,
+      "step": 1907
+    },
+    {
+      "epoch": 0.8787311821766789,
+      "grad_norm": 3.5183165482279417,
+      "learning_rate": 7.604909191159537e-08,
+      "loss": 0.5429,
+      "step": 1908
+    },
+    {
+      "epoch": 0.8791917331107337,
+      "grad_norm": 3.1284052208629416,
+      "learning_rate": 7.547924512560044e-08,
+      "loss": 0.6085,
+      "step": 1909
+    },
+    {
+      "epoch": 0.8796522840447886,
+      "grad_norm": 2.985628461533933,
+      "learning_rate": 7.491145760409134e-08,
+      "loss": 0.6259,
+      "step": 1910
+    },
+    {
+      "epoch": 0.8801128349788434,
+      "grad_norm": 3.340901063995122,
+      "learning_rate": 7.434573061174965e-08,
+      "loss": 0.6032,
+      "step": 1911
+    },
+    {
+      "epoch": 0.8805733859128984,
+      "grad_norm": 3.0741105448726027,
+      "learning_rate": 7.378206540866783e-08,
+      "loss": 0.5554,
+      "step": 1912
+    },
+    {
+      "epoch": 0.8810339368469532,
+      "grad_norm": 2.9332744072817922,
+      "learning_rate": 7.322046325034603e-08,
+      "loss": 0.4567,
+      "step": 1913
+    },
+    {
+      "epoch": 0.881494487781008,
+      "grad_norm": 2.8092291752801857,
+      "learning_rate": 7.266092538768853e-08,
+      "loss": 0.5092,
+      "step": 1914
+    },
+    {
+      "epoch": 0.8819550387150629,
+      "grad_norm": 3.438214393413268,
+      "learning_rate": 7.210345306700238e-08,
+      "loss": 0.563,
+      "step": 1915
+    },
+    {
+      "epoch": 0.8824155896491177,
+      "grad_norm": 3.2745821627508565,
+      "learning_rate": 7.154804752999344e-08,
+      "loss": 0.486,
+      "step": 1916
+    },
+    {
+      "epoch": 0.8828761405831727,
+      "grad_norm": 2.892400504571004,
+      "learning_rate": 7.099471001376434e-08,
+      "loss": 0.5935,
+      "step": 1917
+    },
+    {
+      "epoch": 0.8833366915172275,
+      "grad_norm": 3.2933601567721387,
+      "learning_rate": 7.044344175081107e-08,
+      "loss": 0.6022,
+      "step": 1918
+    },
+    {
+      "epoch": 0.8837972424512823,
+      "grad_norm": 3.4349342024607052,
+      "learning_rate": 6.989424396902078e-08,
+      "loss": 0.6374,
+      "step": 1919
+    },
+    {
+      "epoch": 0.8842577933853372,
+      "grad_norm": 3.1507821922545203,
+      "learning_rate": 6.934711789166902e-08,
+      "loss": 0.5632,
+      "step": 1920
+    },
+    {
+      "epoch": 0.8847183443193921,
+      "grad_norm": 3.089402889899754,
+      "learning_rate": 6.880206473741646e-08,
+      "loss": 0.5506,
+      "step": 1921
+    },
+    {
+      "epoch": 0.885178895253447,
+      "grad_norm": 3.181790361740718,
+      "learning_rate": 6.825908572030703e-08,
+      "loss": 0.6158,
+      "step": 1922
+    },
+    {
+      "epoch": 0.8856394461875018,
+      "grad_norm": 3.0742747552763077,
+      "learning_rate": 6.771818204976453e-08,
+      "loss": 0.5871,
+      "step": 1923
+    },
+    {
+      "epoch": 0.8860999971215566,
+      "grad_norm": 3.010007045848609,
+      "learning_rate": 6.71793549305899e-08,
+      "loss": 0.5538,
+      "step": 1924
+    },
+    {
+      "epoch": 0.8865605480556116,
+      "grad_norm": 3.3014591830340225,
+      "learning_rate": 6.66426055629593e-08,
+      "loss": 0.5739,
+      "step": 1925
+    },
+    {
+      "epoch": 0.8870210989896664,
+      "grad_norm": 2.8280186580117337,
+      "learning_rate": 6.610793514242074e-08,
+      "loss": 0.5761,
+      "step": 1926
+    },
+    {
+      "epoch": 0.8874816499237212,
+      "grad_norm": 2.8955154098471696,
+      "learning_rate": 6.557534485989135e-08,
+      "loss": 0.4836,
+      "step": 1927
+    },
+    {
+      "epoch": 0.8879422008577761,
+      "grad_norm": 3.2863599262582546,
+      "learning_rate": 6.504483590165533e-08,
+      "loss": 0.6659,
+      "step": 1928
+    },
+    {
+      "epoch": 0.8884027517918309,
+      "grad_norm": 2.7810076021075325,
+      "learning_rate": 6.451640944936087e-08,
+      "loss": 0.5125,
+      "step": 1929
+    },
+    {
+      "epoch": 0.8888633027258859,
+      "grad_norm": 2.829017960020371,
+      "learning_rate": 6.399006668001772e-08,
+      "loss": 0.4712,
+      "step": 1930
+    },
+    {
+      "epoch": 0.8893238536599407,
+      "grad_norm": 3.1200160087326485,
+      "learning_rate": 6.346580876599394e-08,
+      "loss": 0.5241,
+      "step": 1931
+    },
+    {
+      "epoch": 0.8897844045939955,
+      "grad_norm": 3.284501857717466,
+      "learning_rate": 6.294363687501459e-08,
+      "loss": 0.4956,
+      "step": 1932
+    },
+    {
+      "epoch": 0.8902449555280504,
+      "grad_norm": 3.0368507746622497,
+      "learning_rate": 6.242355217015793e-08,
+      "loss": 0.5335,
+      "step": 1933
+    },
+    {
+      "epoch": 0.8907055064621053,
+      "grad_norm": 2.707473642304596,
+      "learning_rate": 6.190555580985291e-08,
+      "loss": 0.4174,
+      "step": 1934
+    },
+    {
+      "epoch": 0.8911660573961602,
+      "grad_norm": 3.31398125775273,
+      "learning_rate": 6.138964894787802e-08,
+      "loss": 0.5595,
+      "step": 1935
+    },
+    {
+      "epoch": 0.891626608330215,
+      "grad_norm": 3.113081540988843,
+      "learning_rate": 6.08758327333564e-08,
+      "loss": 0.5791,
+      "step": 1936
+    },
+    {
+      "epoch": 0.8920871592642698,
+      "grad_norm": 2.8145565837317608,
+      "learning_rate": 6.036410831075489e-08,
+      "loss": 0.5162,
+      "step": 1937
+    },
+    {
+      "epoch": 0.8925477101983248,
+      "grad_norm": 3.196959966502691,
+      "learning_rate": 5.985447681988187e-08,
+      "loss": 0.656,
+      "step": 1938
+    },
+    {
+      "epoch": 0.8930082611323796,
+      "grad_norm": 3.2061347516327476,
+      "learning_rate": 5.934693939588276e-08,
+      "loss": 0.5659,
+      "step": 1939
+    },
+    {
+      "epoch": 0.8934688120664345,
+      "grad_norm": 2.7407334976039937,
+      "learning_rate": 5.884149716923947e-08,
+      "loss": 0.5263,
+      "step": 1940
+    },
+    {
+      "epoch": 0.8939293630004893,
+      "grad_norm": 3.0639659095757925,
+      "learning_rate": 5.833815126576713e-08,
+      "loss": 0.4792,
+      "step": 1941
+    },
+    {
+      "epoch": 0.8943899139345441,
+      "grad_norm": 3.705402334304555,
+      "learning_rate": 5.78369028066108e-08,
+      "loss": 0.5598,
+      "step": 1942
+    },
+    {
+      "epoch": 0.8948504648685991,
+      "grad_norm": 3.314179143806477,
+      "learning_rate": 5.7337752908244604e-08,
+      "loss": 0.6057,
+      "step": 1943
+    },
+    {
+      "epoch": 0.8953110158026539,
+      "grad_norm": 3.093901085439443,
+      "learning_rate": 5.684070268246799e-08,
+      "loss": 0.5189,
+      "step": 1944
+    },
+    {
+      "epoch": 0.8957715667367088,
+      "grad_norm": 2.9306316685724814,
+      "learning_rate": 5.634575323640334e-08,
+      "loss": 0.518,
+      "step": 1945
+    },
+    {
+      "epoch": 0.8962321176707636,
+      "grad_norm": 2.6131420700596744,
+      "learning_rate": 5.5852905672494235e-08,
+      "loss": 0.4514,
+      "step": 1946
+    },
+    {
+      "epoch": 0.8966926686048186,
+      "grad_norm": 3.4838526888557224,
+      "learning_rate": 5.5362161088502335e-08,
+      "loss": 0.6153,
+      "step": 1947
+    },
+    {
+      "epoch": 0.8971532195388734,
+      "grad_norm": 2.7410462665447732,
+      "learning_rate": 5.487352057750538e-08,
+      "loss": 0.5487,
+      "step": 1948
+    },
+    {
+      "epoch": 0.8976137704729282,
+      "grad_norm": 3.323249823556552,
+      "learning_rate": 5.438698522789409e-08,
+      "loss": 0.4776,
+      "step": 1949
+    },
+    {
+      "epoch": 0.8980743214069831,
+      "grad_norm": 3.3914954871956207,
+      "learning_rate": 5.390255612337058e-08,
+      "loss": 0.5644,
+      "step": 1950
+    },
+    {
+      "epoch": 0.898534872341038,
+      "grad_norm": 3.453540431554986,
+      "learning_rate": 5.3420234342945515e-08,
+      "loss": 0.5669,
+      "step": 1951
+    },
+    {
+      "epoch": 0.8989954232750929,
+      "grad_norm": 3.2848747343127696,
+      "learning_rate": 5.2940020960935416e-08,
+      "loss": 0.6555,
+      "step": 1952
+    },
+    {
+      "epoch": 0.8994559742091477,
+      "grad_norm": 3.335247069769459,
+      "learning_rate": 5.246191704696079e-08,
+      "loss": 0.5593,
+      "step": 1953
+    },
+    {
+      "epoch": 0.8999165251432025,
+      "grad_norm": 2.373713227217355,
+      "learning_rate": 5.1985923665943787e-08,
+      "loss": 0.3408,
+      "step": 1954
+    },
+    {
+      "epoch": 0.9003770760772574,
+      "grad_norm": 3.2307700903116334,
+      "learning_rate": 5.1512041878105095e-08,
+      "loss": 0.5363,
+      "step": 1955
+    },
+    {
+      "epoch": 0.9008376270113123,
+      "grad_norm": 3.2016297093263817,
+      "learning_rate": 5.104027273896239e-08,
+      "loss": 0.5384,
+      "step": 1956
+    },
+    {
+      "epoch": 0.9012981779453672,
+      "grad_norm": 3.20013985926728,
+      "learning_rate": 5.057061729932777e-08,
+      "loss": 0.4193,
+      "step": 1957
+    },
+    {
+      "epoch": 0.901758728879422,
+      "grad_norm": 2.7276925258485436,
+      "learning_rate": 5.0103076605304885e-08,
+      "loss": 0.5143,
+      "step": 1958
+    },
+    {
+      "epoch": 0.9022192798134768,
+      "grad_norm": 2.9954612745734903,
+      "learning_rate": 4.963765169828737e-08,
+      "loss": 0.4886,
+      "step": 1959
+    },
+    {
+      "epoch": 0.9026798307475318,
+      "grad_norm": 3.1663609715145102,
+      "learning_rate": 4.917434361495609e-08,
+      "loss": 0.529,
+      "step": 1960
+    },
+    {
+      "epoch": 0.9031403816815866,
+      "grad_norm": 3.1213827237745333,
+      "learning_rate": 4.871315338727711e-08,
+      "loss": 0.5155,
+      "step": 1961
+    },
+    {
+      "epoch": 0.9036009326156415,
+      "grad_norm": 2.7436550688979837,
+      "learning_rate": 4.825408204249881e-08,
+      "loss": 0.4606,
+      "step": 1962
+    },
+    {
+      "epoch": 0.9040614835496963,
+      "grad_norm": 3.249120899149753,
+      "learning_rate": 4.779713060315016e-08,
+      "loss": 0.5054,
+      "step": 1963
+    },
+    {
+      "epoch": 0.9045220344837512,
+      "grad_norm": 2.90283494064835,
+      "learning_rate": 4.734230008703877e-08,
+      "loss": 0.5579,
+      "step": 1964
+    },
+    {
+      "epoch": 0.9049825854178061,
+      "grad_norm": 3.4042002585801865,
+      "learning_rate": 4.688959150724703e-08,
+      "loss": 0.5775,
+      "step": 1965
+    },
+    {
+      "epoch": 0.9054431363518609,
+      "grad_norm": 3.1250557810610182,
+      "learning_rate": 4.6439005872132454e-08,
+      "loss": 0.6137,
+      "step": 1966
+    },
+    {
+      "epoch": 0.9059036872859157,
+      "grad_norm": 3.1195028338428226,
+      "learning_rate": 4.599054418532267e-08,
+      "loss": 0.561,
+      "step": 1967
+    },
+    {
+      "epoch": 0.9063642382199706,
+      "grad_norm": 2.8851519193897643,
+      "learning_rate": 4.554420744571463e-08,
+      "loss": 0.5235,
+      "step": 1968
+    },
+    {
+      "epoch": 0.9068247891540255,
+      "grad_norm": 3.6481537119175997,
+      "learning_rate": 4.5099996647473215e-08,
+      "loss": 0.4759,
+      "step": 1969
+    },
+    {
+      "epoch": 0.9072853400880804,
+      "grad_norm": 2.790296368178869,
+      "learning_rate": 4.465791278002684e-08,
+      "loss": 0.4618,
+      "step": 1970
+    },
+    {
+      "epoch": 0.9077458910221352,
+      "grad_norm": 3.556677402896051,
+      "learning_rate": 4.4217956828066614e-08,
+      "loss": 0.5706,
+      "step": 1971
+    },
+    {
+      "epoch": 0.90820644195619,
+      "grad_norm": 2.978589496267026,
+      "learning_rate": 4.3780129771544885e-08,
+      "loss": 0.538,
+      "step": 1972
+    },
+    {
+      "epoch": 0.908666992890245,
+      "grad_norm": 3.6725000383644386,
+      "learning_rate": 4.3344432585670886e-08,
+      "loss": 0.6891,
+      "step": 1973
+    },
+    {
+      "epoch": 0.9091275438242998,
+      "grad_norm": 3.2917964842156717,
+      "learning_rate": 4.291086624091067e-08,
+      "loss": 0.5737,
+      "step": 1974
+    },
+    {
+      "epoch": 0.9095880947583547,
+      "grad_norm": 3.0743766947515088,
+      "learning_rate": 4.2479431702983845e-08,
+      "loss": 0.5731,
+      "step": 1975
+    },
+    {
+      "epoch": 0.9100486456924095,
+      "grad_norm": 3.0506298198871993,
+      "learning_rate": 4.205012993286139e-08,
+      "loss": 0.4587,
+      "step": 1976
+    },
+    {
+      "epoch": 0.9105091966264645,
+      "grad_norm": 3.0219223404896907,
+      "learning_rate": 4.162296188676417e-08,
+      "loss": 0.527,
+      "step": 1977
+    },
+    {
+      "epoch": 0.9109697475605193,
+      "grad_norm": 2.79095841731694,
+      "learning_rate": 4.119792851616022e-08,
+      "loss": 0.4888,
+      "step": 1978
+    },
+    {
+      "epoch": 0.9114302984945741,
+      "grad_norm": 3.4172455118807283,
+      "learning_rate": 4.0775030767762895e-08,
+      "loss": 0.5758,
+      "step": 1979
+    },
+    {
+      "epoch": 0.911890849428629,
+      "grad_norm": 3.239863460054807,
+      "learning_rate": 4.035426958352861e-08,
+      "loss": 0.5547,
+      "step": 1980
+    },
+    {
+      "epoch": 0.9123514003626838,
+      "grad_norm": 2.8654349056505017,
+      "learning_rate": 3.99356459006549e-08,
+      "loss": 0.438,
+      "step": 1981
+    },
+    {
+      "epoch": 0.9128119512967388,
+      "grad_norm": 2.9491143905443065,
+      "learning_rate": 3.9519160651578456e-08,
+      "loss": 0.5215,
+      "step": 1982
+    },
+    {
+      "epoch": 0.9132725022307936,
+      "grad_norm": 3.590224655674636,
+      "learning_rate": 3.910481476397231e-08,
+      "loss": 0.5205,
+      "step": 1983
+    },
+    {
+      "epoch": 0.9137330531648484,
+      "grad_norm": 3.0897242283731816,
+      "learning_rate": 3.8692609160744796e-08,
+      "loss": 0.4683,
+      "step": 1984
+    },
+    {
+      "epoch": 0.9141936040989033,
+      "grad_norm": 2.6174259753352773,
+      "learning_rate": 3.8282544760037005e-08,
+      "loss": 0.5923,
+      "step": 1985
+    },
+    {
+      "epoch": 0.9146541550329582,
+      "grad_norm": 3.281785734259912,
+      "learning_rate": 3.787462247522033e-08,
+      "loss": 0.5454,
+      "step": 1986
+    },
+    {
+      "epoch": 0.915114705967013,
+      "grad_norm": 3.0957714777119003,
+      "learning_rate": 3.74688432148953e-08,
+      "loss": 0.4839,
+      "step": 1987
+    },
+    {
+      "epoch": 0.9155752569010679,
+      "grad_norm": 3.2146157875749735,
+      "learning_rate": 3.7065207882888915e-08,
+      "loss": 0.5671,
+      "step": 1988
+    },
+    {
+      "epoch": 0.9160358078351227,
+      "grad_norm": 3.2632112706777097,
+      "learning_rate": 3.666371737825269e-08,
+      "loss": 0.5403,
+      "step": 1989
+    },
+    {
+      "epoch": 0.9164963587691777,
+      "grad_norm": 2.588003078548097,
+      "learning_rate": 3.626437259526094e-08,
+      "loss": 0.4528,
+      "step": 1990
+    },
+    {
+      "epoch": 0.9169569097032325,
+      "grad_norm": 3.5583915478306025,
+      "learning_rate": 3.58671744234087e-08,
+      "loss": 0.5301,
+      "step": 1991
+    },
+    {
+      "epoch": 0.9174174606372874,
+      "grad_norm": 2.97658613622478,
+      "learning_rate": 3.54721237474096e-08,
+      "loss": 0.5364,
+      "step": 1992
+    },
+    {
+      "epoch": 0.9178780115713422,
+      "grad_norm": 3.005308580137038,
+      "learning_rate": 3.5079221447193665e-08,
+      "loss": 0.5161,
+      "step": 1993
+    },
+    {
+      "epoch": 0.918338562505397,
+      "grad_norm": 3.4338709756555335,
+      "learning_rate": 3.468846839790629e-08,
+      "loss": 0.5028,
+      "step": 1994
+    },
+    {
+      "epoch": 0.918799113439452,
+      "grad_norm": 3.022995712791708,
+      "learning_rate": 3.4299865469905156e-08,
+      "loss": 0.4699,
+      "step": 1995
+    },
+    {
+      "epoch": 0.9192596643735068,
+      "grad_norm": 3.0589636260878654,
+      "learning_rate": 3.391341352875887e-08,
+      "loss": 0.5397,
+      "step": 1996
+    },
+    {
+      "epoch": 0.9197202153075617,
+      "grad_norm": 3.029785621200751,
+      "learning_rate": 3.3529113435245e-08,
+      "loss": 0.5563,
+      "step": 1997
+    },
+    {
+      "epoch": 0.9201807662416165,
+      "grad_norm": 3.12537296154988,
+      "learning_rate": 3.314696604534839e-08,
+      "loss": 0.559,
+      "step": 1998
+    },
+    {
+      "epoch": 0.9206413171756714,
+      "grad_norm": 3.416953532596617,
+      "learning_rate": 3.276697221025848e-08,
+      "loss": 0.5028,
+      "step": 1999
+    },
+    {
+      "epoch": 0.9211018681097263,
+      "grad_norm": 3.7639797946349485,
+      "learning_rate": 3.238913277636846e-08,
+      "loss": 0.4885,
+      "step": 2000
+    },
+    {
+      "epoch": 0.9215624190437811,
+      "grad_norm": 3.355422997580863,
+      "learning_rate": 3.201344858527233e-08,
+      "loss": 0.5052,
+      "step": 2001
+    },
+    {
+      "epoch": 0.922022969977836,
+      "grad_norm": 3.1382941450648425,
+      "learning_rate": 3.163992047376374e-08,
+      "loss": 0.5372,
+      "step": 2002
+    },
+    {
+      "epoch": 0.9224835209118909,
+      "grad_norm": 3.4395472891340377,
+      "learning_rate": 3.126854927383416e-08,
+      "loss": 0.5008,
+      "step": 2003
+    },
+    {
+      "epoch": 0.9229440718459457,
+      "grad_norm": 3.1440121431768078,
+      "learning_rate": 3.089933581267024e-08,
+      "loss": 0.6216,
+      "step": 2004
+    },
+    {
+      "epoch": 0.9234046227800006,
+      "grad_norm": 3.519897694068012,
+      "learning_rate": 3.053228091265314e-08,
+      "loss": 0.5959,
+      "step": 2005
+    },
+    {
+      "epoch": 0.9238651737140554,
+      "grad_norm": 3.1572535401512756,
+      "learning_rate": 3.016738539135566e-08,
+      "loss": 0.4811,
+      "step": 2006
+    },
+    {
+      "epoch": 0.9243257246481102,
+      "grad_norm": 3.2199243277505616,
+      "learning_rate": 2.980465006154076e-08,
+      "loss": 0.5451,
+      "step": 2007
+    },
+    {
+      "epoch": 0.9247862755821652,
+      "grad_norm": 3.3241287834796096,
+      "learning_rate": 2.9444075731160256e-08,
+      "loss": 0.6162,
+      "step": 2008
+    },
+    {
+      "epoch": 0.92524682651622,
+      "grad_norm": 3.164226428742799,
+      "learning_rate": 2.908566320335215e-08,
+      "loss": 0.5484,
+      "step": 2009
+    },
+    {
+      "epoch": 0.9257073774502749,
+      "grad_norm": 3.4187086885851774,
+      "learning_rate": 2.872941327643963e-08,
+      "loss": 0.5215,
+      "step": 2010
+    },
+    {
+      "epoch": 0.9261679283843297,
+      "grad_norm": 3.33801664810882,
+      "learning_rate": 2.837532674392862e-08,
+      "loss": 0.4978,
+      "step": 2011
+    },
+    {
+      "epoch": 0.9266284793183847,
+      "grad_norm": 3.1540649821799573,
+      "learning_rate": 2.8023404394506345e-08,
+      "loss": 0.5169,
+      "step": 2012
+    },
+    {
+      "epoch": 0.9270890302524395,
+      "grad_norm": 3.332203820351019,
+      "learning_rate": 2.767364701204e-08,
+      "loss": 0.4781,
+      "step": 2013
+    },
+    {
+      "epoch": 0.9275495811864943,
+      "grad_norm": 3.0643310804507253,
+      "learning_rate": 2.7326055375573976e-08,
+      "loss": 0.5149,
+      "step": 2014
+    },
+    {
+      "epoch": 0.9280101321205492,
+      "grad_norm": 2.841374181617664,
+      "learning_rate": 2.6980630259329063e-08,
+      "loss": 0.5501,
+      "step": 2015
+    },
+    {
+      "epoch": 0.9284706830546041,
+      "grad_norm": 3.3194075100355835,
+      "learning_rate": 2.6637372432700476e-08,
+      "loss": 0.6163,
+      "step": 2016
+    },
+    {
+      "epoch": 0.928931233988659,
+      "grad_norm": 3.0237095704445633,
+      "learning_rate": 2.629628266025552e-08,
+      "loss": 0.5245,
+      "step": 2017
+    },
+    {
+      "epoch": 0.9293917849227138,
+      "grad_norm": 3.0048824665623566,
+      "learning_rate": 2.5957361701732904e-08,
+      "loss": 0.7035,
+      "step": 2018
+    },
+    {
+      "epoch": 0.9298523358567686,
+      "grad_norm": 2.850832918602004,
+      "learning_rate": 2.5620610312040436e-08,
+      "loss": 0.4749,
+      "step": 2019
+    },
+    {
+      "epoch": 0.9303128867908235,
+      "grad_norm": 2.889924325240167,
+      "learning_rate": 2.528602924125334e-08,
+      "loss": 0.4982,
+      "step": 2020
+    },
+    {
+      "epoch": 0.9307734377248784,
+      "grad_norm": 3.097179425048999,
+      "learning_rate": 2.495361923461281e-08,
+      "loss": 0.57,
+      "step": 2021
+    },
+    {
+      "epoch": 0.9312339886589333,
+      "grad_norm": 3.195072157476586,
+      "learning_rate": 2.462338103252415e-08,
+      "loss": 0.6917,
+      "step": 2022
+    },
+    {
+      "epoch": 0.9316945395929881,
+      "grad_norm": 2.964114672020939,
+      "learning_rate": 2.4295315370555402e-08,
+      "loss": 0.4675,
+      "step": 2023
+    },
+    {
+      "epoch": 0.9321550905270429,
+      "grad_norm": 2.698298386130625,
+      "learning_rate": 2.3969422979435162e-08,
+      "loss": 0.4485,
+      "step": 2024
+    },
+    {
+      "epoch": 0.9326156414610979,
+      "grad_norm": 3.05208848367821,
+      "learning_rate": 2.3645704585051775e-08,
+      "loss": 0.6052,
+      "step": 2025
+    },
+    {
+      "epoch": 0.9330761923951527,
+      "grad_norm": 3.626805603045707,
+      "learning_rate": 2.3324160908451017e-08,
+      "loss": 0.458,
+      "step": 2026
+    },
+    {
+      "epoch": 0.9335367433292076,
+      "grad_norm": 3.2241633697370413,
+      "learning_rate": 2.300479266583455e-08,
+      "loss": 0.5033,
+      "step": 2027
+    },
+    {
+      "epoch": 0.9339972942632624,
+      "grad_norm": 3.5953424649745735,
+      "learning_rate": 2.2687600568558785e-08,
+      "loss": 0.7304,
+      "step": 2028
+    },
+    {
+      "epoch": 0.9344578451973173,
+      "grad_norm": 3.8158227802153952,
+      "learning_rate": 2.237258532313302e-08,
+      "loss": 0.5633,
+      "step": 2029
+    },
+    {
+      "epoch": 0.9349183961313722,
+      "grad_norm": 3.040372697625964,
+      "learning_rate": 2.205974763121754e-08,
+      "loss": 0.5436,
+      "step": 2030
+    },
+    {
+      "epoch": 0.935378947065427,
+      "grad_norm": 2.9246000563345866,
+      "learning_rate": 2.1749088189622844e-08,
+      "loss": 0.5782,
+      "step": 2031
+    },
+    {
+      "epoch": 0.9358394979994819,
+      "grad_norm": 3.1558727218247244,
+      "learning_rate": 2.144060769030742e-08,
+      "loss": 0.5921,
+      "step": 2032
+    },
+    {
+      "epoch": 0.9363000489335367,
+      "grad_norm": 2.9122079524337754,
+      "learning_rate": 2.113430682037598e-08,
+      "loss": 0.5348,
+      "step": 2033
+    },
+    {
+      "epoch": 0.9367605998675916,
+      "grad_norm": 2.5503478146770244,
+      "learning_rate": 2.083018626207933e-08,
+      "loss": 0.4812,
+      "step": 2034
+    },
+    {
+      "epoch": 0.9372211508016465,
+      "grad_norm": 3.2463091586540958,
+      "learning_rate": 2.0528246692810835e-08,
+      "loss": 0.5133,
+      "step": 2035
+    },
+    {
+      "epoch": 0.9376817017357013,
+      "grad_norm": 2.7033666955582225,
+      "learning_rate": 2.0228488785106634e-08,
+      "loss": 0.5074,
+      "step": 2036
+    },
+    {
+      "epoch": 0.9381422526697561,
+      "grad_norm": 3.3458142554070736,
+      "learning_rate": 1.9930913206643306e-08,
+      "loss": 0.5898,
+      "step": 2037
+    },
+    {
+      "epoch": 0.9386028036038111,
+      "grad_norm": 3.0937605106073947,
+      "learning_rate": 1.9635520620236323e-08,
+      "loss": 0.4866,
+      "step": 2038
+    },
+    {
+      "epoch": 0.9390633545378659,
+      "grad_norm": 3.3805419118935545,
+      "learning_rate": 1.934231168383915e-08,
+      "loss": 0.6963,
+      "step": 2039
+    },
+    {
+      "epoch": 0.9395239054719208,
+      "grad_norm": 2.9416660248540536,
+      "learning_rate": 1.9051287050541263e-08,
+      "loss": 0.5596,
+      "step": 2040
+    },
+    {
+      "epoch": 0.9399844564059756,
+      "grad_norm": 2.5583360386888883,
+      "learning_rate": 1.876244736856658e-08,
+      "loss": 0.3893,
+      "step": 2041
+    },
+    {
+      "epoch": 0.9404450073400306,
+      "grad_norm": 3.602394844208884,
+      "learning_rate": 1.847579328127269e-08,
+      "loss": 0.6501,
+      "step": 2042
+    },
+    {
+      "epoch": 0.9409055582740854,
+      "grad_norm": 3.278596111429843,
+      "learning_rate": 1.819132542714874e-08,
+      "loss": 0.5125,
+      "step": 2043
+    },
+    {
+      "epoch": 0.9413661092081402,
+      "grad_norm": 3.2792307923445794,
+      "learning_rate": 1.790904443981478e-08,
+      "loss": 0.546,
+      "step": 2044
+    },
+    {
+      "epoch": 0.9418266601421951,
+      "grad_norm": 3.354526716736895,
+      "learning_rate": 1.7628950948018974e-08,
+      "loss": 0.5228,
+      "step": 2045
+    },
+    {
+      "epoch": 0.9422872110762499,
+      "grad_norm": 3.2620023435160395,
+      "learning_rate": 1.7351045575638044e-08,
+      "loss": 0.4842,
+      "step": 2046
+    },
+    {
+      "epoch": 0.9427477620103049,
+      "grad_norm": 3.509391862715867,
+      "learning_rate": 1.7075328941674295e-08,
+      "loss": 0.5415,
+      "step": 2047
+    },
+    {
+      "epoch": 0.9432083129443597,
+      "grad_norm": 2.9644652855209945,
+      "learning_rate": 1.680180166025513e-08,
+      "loss": 0.5368,
+      "step": 2048
+    },
+    {
+      "epoch": 0.9436688638784145,
+      "grad_norm": 3.103421227871461,
+      "learning_rate": 1.653046434063121e-08,
+      "loss": 0.6085,
+      "step": 2049
+    },
+    {
+      "epoch": 0.9441294148124694,
+      "grad_norm": 3.2771987160259877,
+      "learning_rate": 1.626131758717575e-08,
+      "loss": 0.5967,
+      "step": 2050
+    },
+    {
+      "epoch": 0.9445899657465243,
+      "grad_norm": 2.840695547956422,
+      "learning_rate": 1.59943619993822e-08,
+      "loss": 0.5215,
+      "step": 2051
+    },
+    {
+      "epoch": 0.9450505166805792,
+      "grad_norm": 3.5326470386345545,
+      "learning_rate": 1.572959817186359e-08,
+      "loss": 0.5683,
+      "step": 2052
+    },
+    {
+      "epoch": 0.945511067614634,
+      "grad_norm": 3.4768749960533127,
+      "learning_rate": 1.5467026694351404e-08,
+      "loss": 0.5466,
+      "step": 2053
+    },
+    {
+      "epoch": 0.9459716185486888,
+      "grad_norm": 3.3583093067528313,
+      "learning_rate": 1.5206648151693478e-08,
+      "loss": 0.553,
+      "step": 2054
+    },
+    {
+      "epoch": 0.9464321694827438,
+      "grad_norm": 3.009812369106128,
+      "learning_rate": 1.4948463123853337e-08,
+      "loss": 0.4992,
+      "step": 2055
+    },
+    {
+      "epoch": 0.9468927204167986,
+      "grad_norm": 2.9099681097218286,
+      "learning_rate": 1.4692472185908633e-08,
+      "loss": 0.4736,
+      "step": 2056
+    },
+    {
+      "epoch": 0.9473532713508535,
+      "grad_norm": 3.3355322307095108,
+      "learning_rate": 1.4438675908050036e-08,
+      "loss": 0.5518,
+      "step": 2057
+    },
+    {
+      "epoch": 0.9478138222849083,
+      "grad_norm": 2.5897747211717586,
+      "learning_rate": 1.4187074855579795e-08,
+      "loss": 0.3858,
+      "step": 2058
+    },
+    {
+      "epoch": 0.9482743732189632,
+      "grad_norm": 3.3391284256863103,
+      "learning_rate": 1.3937669588910406e-08,
+      "loss": 0.6045,
+      "step": 2059
+    },
+    {
+      "epoch": 0.9487349241530181,
+      "grad_norm": 3.0970613079517957,
+      "learning_rate": 1.3690460663563829e-08,
+      "loss": 0.5936,
+      "step": 2060
+    },
+    {
+      "epoch": 0.9491954750870729,
+      "grad_norm": 3.3727344553879384,
+      "learning_rate": 1.344544863016961e-08,
+      "loss": 0.5295,
+      "step": 2061
+    },
+    {
+      "epoch": 0.9496560260211278,
+      "grad_norm": 2.9033018743602312,
+      "learning_rate": 1.3202634034464199e-08,
+      "loss": 0.549,
+      "step": 2062
+    },
+    {
+      "epoch": 0.9501165769551826,
+      "grad_norm": 2.958202980684059,
+      "learning_rate": 1.2962017417289418e-08,
+      "loss": 0.5569,
+      "step": 2063
+    },
+    {
+      "epoch": 0.9505771278892375,
+      "grad_norm": 2.976503730921731,
+      "learning_rate": 1.2723599314591105e-08,
+      "loss": 0.5678,
+      "step": 2064
+    },
+    {
+      "epoch": 0.9510376788232924,
+      "grad_norm": 3.1304654491952872,
+      "learning_rate": 1.2487380257418578e-08,
+      "loss": 0.5418,
+      "step": 2065
+    },
+    {
+      "epoch": 0.9514982297573472,
+      "grad_norm": 3.25198019264722,
+      "learning_rate": 1.2253360771922739e-08,
+      "loss": 0.5825,
+      "step": 2066
+    },
+    {
+      "epoch": 0.951958780691402,
+      "grad_norm": 3.5969220214188997,
+      "learning_rate": 1.2021541379355404e-08,
+      "loss": 0.5422,
+      "step": 2067
+    },
+    {
+      "epoch": 0.952419331625457,
+      "grad_norm": 2.84240161611037,
+      "learning_rate": 1.1791922596067649e-08,
+      "loss": 0.5475,
+      "step": 2068
+    },
+    {
+      "epoch": 0.9528798825595118,
+      "grad_norm": 2.7264361506813537,
+      "learning_rate": 1.1564504933509244e-08,
+      "loss": 0.4494,
+      "step": 2069
+    },
+    {
+      "epoch": 0.9533404334935667,
+      "grad_norm": 2.9820660647231296,
+      "learning_rate": 1.1339288898227106e-08,
+      "loss": 0.5436,
+      "step": 2070
+    },
+    {
+      "epoch": 0.9538009844276215,
+      "grad_norm": 3.139054611664098,
+      "learning_rate": 1.1116274991864072e-08,
+      "loss": 0.4871,
+      "step": 2071
+    },
+    {
+      "epoch": 0.9542615353616765,
+      "grad_norm": 2.8848562462423377,
+      "learning_rate": 1.0895463711158349e-08,
+      "loss": 0.4775,
+      "step": 2072
+    },
+    {
+      "epoch": 0.9547220862957313,
+      "grad_norm": 2.940726909982554,
+      "learning_rate": 1.0676855547941844e-08,
+      "loss": 0.4509,
+      "step": 2073
+    },
+    {
+      "epoch": 0.9551826372297861,
+      "grad_norm": 3.011334549424482,
+      "learning_rate": 1.0460450989139169e-08,
+      "loss": 0.5029,
+      "step": 2074
+    },
+    {
+      "epoch": 0.955643188163841,
+      "grad_norm": 2.997675838063664,
+      "learning_rate": 1.0246250516766863e-08,
+      "loss": 0.5425,
+      "step": 2075
+    },
+    {
+      "epoch": 0.9561037390978958,
+      "grad_norm": 2.985898615725367,
+      "learning_rate": 1.0034254607932168e-08,
+      "loss": 0.5514,
+      "step": 2076
+    },
+    {
+      "epoch": 0.9565642900319508,
+      "grad_norm": 3.0381025030705233,
+      "learning_rate": 9.82446373483159e-09,
+      "loss": 0.4691,
+      "step": 2077
+    },
+    {
+      "epoch": 0.9570248409660056,
+      "grad_norm": 2.7713452807820937,
+      "learning_rate": 9.616878364750446e-09,
+      "loss": 0.4628,
+      "step": 2078
+    },
+    {
+      "epoch": 0.9574853919000604,
+      "grad_norm": 2.9033280044478853,
+      "learning_rate": 9.411498960061436e-09,
+      "loss": 0.4301,
+      "step": 2079
+    },
+    {
+      "epoch": 0.9579459428341153,
+      "grad_norm": 3.2313767447057917,
+      "learning_rate": 9.208325978223741e-09,
+      "loss": 0.5824,
+      "step": 2080
+    },
+    {
+      "epoch": 0.9584064937681702,
+      "grad_norm": 3.0409141727255915,
+      "learning_rate": 9.00735987178214e-09,
+      "loss": 0.5694,
+      "step": 2081
+    },
+    {
+      "epoch": 0.9588670447022251,
+      "grad_norm": 3.0094026886557526,
+      "learning_rate": 8.808601088365453e-09,
+      "loss": 0.5971,
+      "step": 2082
+    },
+    {
+      "epoch": 0.9593275956362799,
+      "grad_norm": 3.4474240649260555,
+      "learning_rate": 8.612050070686217e-09,
+      "loss": 0.5686,
+      "step": 2083
+    },
+    {
+      "epoch": 0.9597881465703347,
+      "grad_norm": 3.264238061980219,
+      "learning_rate": 8.417707256539675e-09,
+      "loss": 0.4905,
+      "step": 2084
+    },
+    {
+      "epoch": 0.9602486975043897,
+      "grad_norm": 3.445851440079108,
+      "learning_rate": 8.225573078802006e-09,
+      "loss": 0.536,
+      "step": 2085
+    },
+    {
+      "epoch": 0.9607092484384445,
+      "grad_norm": 3.2368359736420684,
+      "learning_rate": 8.035647965430215e-09,
+      "loss": 0.5559,
+      "step": 2086
+    },
+    {
+      "epoch": 0.9611697993724994,
+      "grad_norm": 2.7042062125888835,
+      "learning_rate": 7.847932339460906e-09,
+      "loss": 0.537,
+      "step": 2087
+    },
+    {
+      "epoch": 0.9616303503065542,
+      "grad_norm": 3.352414822995779,
+      "learning_rate": 7.662426619009178e-09,
+      "loss": 0.5649,
+      "step": 2088
+    },
+    {
+      "epoch": 0.962090901240609,
+      "grad_norm": 3.04805145690995,
+      "learning_rate": 7.479131217267732e-09,
+      "loss": 0.6832,
+      "step": 2089
+    },
+    {
+      "epoch": 0.962551452174664,
+      "grad_norm": 3.0121047224737474,
+      "learning_rate": 7.2980465425063196e-09,
+      "loss": 0.5337,
+      "step": 2090
+    },
+    {
+      "epoch": 0.9630120031087188,
+      "grad_norm": 3.4643442345702815,
+      "learning_rate": 7.119172998070411e-09,
+      "loss": 0.5652,
+      "step": 2091
+    },
+    {
+      "epoch": 0.9634725540427737,
+      "grad_norm": 3.1880017551075035,
+      "learning_rate": 6.9425109823803e-09,
+      "loss": 0.4508,
+      "step": 2092
+    },
+    {
+      "epoch": 0.9639331049768285,
+      "grad_norm": 3.5879986405315836,
+      "learning_rate": 6.768060888930449e-09,
+      "loss": 0.611,
+      "step": 2093
+    },
+    {
+      "epoch": 0.9643936559108834,
+      "grad_norm": 3.1074545754597493,
+      "learning_rate": 6.595823106288589e-09,
+      "loss": 0.6024,
+      "step": 2094
+    },
+    {
+      "epoch": 0.9648542068449383,
+      "grad_norm": 3.1449480649098573,
+      "learning_rate": 6.4257980180948415e-09,
+      "loss": 0.5402,
+      "step": 2095
+    },
+    {
+      "epoch": 0.9653147577789931,
+      "grad_norm": 3.240649850034883,
+      "learning_rate": 6.257986003060489e-09,
+      "loss": 0.5048,
+      "step": 2096
+    },
+    {
+      "epoch": 0.965775308713048,
+      "grad_norm": 2.958071414094661,
+      "learning_rate": 6.09238743496776e-09,
+      "loss": 0.5455,
+      "step": 2097
+    },
+    {
+      "epoch": 0.9662358596471029,
+      "grad_norm": 3.3522897979936244,
+      "learning_rate": 5.929002682668494e-09,
+      "loss": 0.5016,
+      "step": 2098
+    },
+    {
+      "epoch": 0.9666964105811577,
+      "grad_norm": 3.3646238303023464,
+      "learning_rate": 5.7678321100836925e-09,
+      "loss": 0.6399,
+      "step": 2099
+    },
+    {
+      "epoch": 0.9671569615152126,
+      "grad_norm": 3.146114846622207,
+      "learning_rate": 5.608876076202307e-09,
+      "loss": 0.4167,
+      "step": 2100
+    },
+    {
+      "epoch": 0.9676175124492674,
+      "grad_norm": 3.2240830087243104,
+      "learning_rate": 5.452134935080899e-09,
+      "loss": 0.519,
+      "step": 2101
+    },
+    {
+      "epoch": 0.9680780633833223,
+      "grad_norm": 2.9509149534120898,
+      "learning_rate": 5.29760903584231e-09,
+      "loss": 0.4709,
+      "step": 2102
+    },
+    {
+      "epoch": 0.9685386143173772,
+      "grad_norm": 2.9753664130176345,
+      "learning_rate": 5.145298722675439e-09,
+      "loss": 0.5777,
+      "step": 2103
+    },
+    {
+      "epoch": 0.968999165251432,
+      "grad_norm": 3.026540264362648,
+      "learning_rate": 4.9952043348342465e-09,
+      "loss": 0.4863,
+      "step": 2104
+    },
+    {
+      "epoch": 0.9694597161854869,
+      "grad_norm": 3.3658355981602197,
+      "learning_rate": 4.847326206636526e-09,
+      "loss": 0.5415,
+      "step": 2105
+    },
+    {
+      "epoch": 0.9699202671195417,
+      "grad_norm": 3.370998012399039,
+      "learning_rate": 4.701664667464245e-09,
+      "loss": 0.5083,
+      "step": 2106
+    },
+    {
+      "epoch": 0.9703808180535967,
+      "grad_norm": 3.0673456063667293,
+      "learning_rate": 4.5582200417617625e-09,
+      "loss": 0.5203,
+      "step": 2107
+    },
+    {
+      "epoch": 0.9708413689876515,
+      "grad_norm": 3.05103189149114,
+      "learning_rate": 4.416992649035612e-09,
+      "loss": 0.5176,
+      "step": 2108
+    },
+    {
+      "epoch": 0.9713019199217063,
+      "grad_norm": 2.88875246748082,
+      "learning_rate": 4.2779828038536085e-09,
+      "loss": 0.5542,
+      "step": 2109
+    },
+    {
+      "epoch": 0.9717624708557612,
+      "grad_norm": 3.03592366750373,
+      "learning_rate": 4.14119081584452e-09,
+      "loss": 0.5563,
+      "step": 2110
+    },
+    {
+      "epoch": 0.9722230217898161,
+      "grad_norm": 3.6405619640934184,
+      "learning_rate": 4.00661698969662e-09,
+      "loss": 0.5303,
+      "step": 2111
+    },
+    {
+      "epoch": 0.972683572723871,
+      "grad_norm": 2.8429713783566926,
+      "learning_rate": 3.874261625157915e-09,
+      "loss": 0.5358,
+      "step": 2112
+    },
+    {
+      "epoch": 0.9731441236579258,
+      "grad_norm": 2.7464922162048238,
+      "learning_rate": 3.744125017034916e-09,
+      "loss": 0.5143,
+      "step": 2113
+    },
+    {
+      "epoch": 0.9736046745919806,
+      "grad_norm": 2.9563946004018575,
+      "learning_rate": 3.6162074551919772e-09,
+      "loss": 0.4829,
+      "step": 2114
+    },
+    {
+      "epoch": 0.9740652255260355,
+      "grad_norm": 3.4014494942852282,
+      "learning_rate": 3.4905092245509637e-09,
+      "loss": 0.5424,
+      "step": 2115
+    },
+    {
+      "epoch": 0.9745257764600904,
+      "grad_norm": 3.4672435378031845,
+      "learning_rate": 3.3670306050902485e-09,
+      "loss": 0.5683,
+      "step": 2116
+    },
+    {
+      "epoch": 0.9749863273941453,
+      "grad_norm": 2.852268606152846,
+      "learning_rate": 3.2457718718443827e-09,
+      "loss": 0.5653,
+      "step": 2117
+    },
+    {
+      "epoch": 0.9754468783282001,
+      "grad_norm": 3.00608356499513,
+      "learning_rate": 3.1267332949033166e-09,
+      "loss": 0.5951,
+      "step": 2118
+    },
+    {
+      "epoch": 0.9759074292622549,
+      "grad_norm": 3.279145796994075,
+      "learning_rate": 3.009915139412067e-09,
+      "loss": 0.5203,
+      "step": 2119
+    },
+    {
+      "epoch": 0.9763679801963099,
+      "grad_norm": 3.034428779025666,
+      "learning_rate": 2.8953176655696075e-09,
+      "loss": 0.6216,
+      "step": 2120
+    },
+    {
+      "epoch": 0.9768285311303647,
+      "grad_norm": 3.400829045951133,
+      "learning_rate": 2.7829411286287572e-09,
+      "loss": 0.5613,
+      "step": 2121
+    },
+    {
+      "epoch": 0.9772890820644196,
+      "grad_norm": 3.1672048746314547,
+      "learning_rate": 2.6727857788954033e-09,
+      "loss": 0.5935,
+      "step": 2122
+    },
+    {
+      "epoch": 0.9777496329984744,
+      "grad_norm": 3.26875632935082,
+      "learning_rate": 2.5648518617280567e-09,
+      "loss": 0.5012,
+      "step": 2123
+    },
+    {
+      "epoch": 0.9782101839325293,
+      "grad_norm": 3.271789631236318,
+      "learning_rate": 2.459139617537187e-09,
+      "loss": 0.4801,
+      "step": 2124
+    },
+    {
+      "epoch": 0.9786707348665842,
+      "grad_norm": 3.2181552684453583,
+      "learning_rate": 2.3556492817847773e-09,
+      "loss": 0.5425,
+      "step": 2125
+    },
+    {
+      "epoch": 0.979131285800639,
+      "grad_norm": 3.0926183100897493,
+      "learning_rate": 2.2543810849836586e-09,
+      "loss": 0.5107,
+      "step": 2126
+    },
+    {
+      "epoch": 0.9795918367346939,
+      "grad_norm": 2.8517011860409887,
+      "learning_rate": 2.1553352526972878e-09,
+      "loss": 0.5229,
+      "step": 2127
+    },
+    {
+      "epoch": 0.9800523876687487,
+      "grad_norm": 2.8804770422411283,
+      "learning_rate": 2.0585120055389705e-09,
+      "loss": 0.4706,
+      "step": 2128
+    },
+    {
+      "epoch": 0.9805129386028036,
+      "grad_norm": 3.202711757414141,
+      "learning_rate": 1.963911559171416e-09,
+      "loss": 0.5939,
+      "step": 2129
+    },
+    {
+      "epoch": 0.9809734895368585,
+      "grad_norm": 3.2187761279662706,
+      "learning_rate": 1.8715341243061846e-09,
+      "loss": 0.5934,
+      "step": 2130
+    },
+    {
+      "epoch": 0.9814340404709133,
+      "grad_norm": 3.881243006981297,
+      "learning_rate": 1.7813799067035729e-09,
+      "loss": 0.5152,
+      "step": 2131
+    },
+    {
+      "epoch": 0.9818945914049682,
+      "grad_norm": 3.1208611268878124,
+      "learning_rate": 1.6934491071719515e-09,
+      "loss": 0.5928,
+      "step": 2132
+    },
+    {
+      "epoch": 0.9823551423390231,
+      "grad_norm": 3.6853248210429426,
+      "learning_rate": 1.6077419215668742e-09,
+      "loss": 0.5364,
+      "step": 2133
+    },
+    {
+      "epoch": 0.9828156932730779,
+      "grad_norm": 3.672076871339007,
+      "learning_rate": 1.5242585407915231e-09,
+      "loss": 0.4995,
+      "step": 2134
+    },
+    {
+      "epoch": 0.9832762442071328,
+      "grad_norm": 3.308203790005742,
+      "learning_rate": 1.4429991507954874e-09,
+      "loss": 0.5985,
+      "step": 2135
+    },
+    {
+      "epoch": 0.9837367951411876,
+      "grad_norm": 3.4864392711956063,
+      "learning_rate": 1.3639639325748741e-09,
+      "loss": 0.5583,
+      "step": 2136
+    },
+    {
+      "epoch": 0.9841973460752426,
+      "grad_norm": 2.865215830877929,
+      "learning_rate": 1.287153062171642e-09,
+      "loss": 0.5195,
+      "step": 2137
+    },
+    {
+      "epoch": 0.9846578970092974,
+      "grad_norm": 3.1310255182151754,
+      "learning_rate": 1.2125667106730464e-09,
+      "loss": 0.5164,
+      "step": 2138
+    },
+    {
+      "epoch": 0.9851184479433522,
+      "grad_norm": 3.112056106943904,
+      "learning_rate": 1.1402050442118616e-09,
+      "loss": 0.4333,
+      "step": 2139
+    },
+    {
+      "epoch": 0.9855789988774071,
+      "grad_norm": 3.038352446915218,
+      "learning_rate": 1.0700682239653814e-09,
+      "loss": 0.5689,
+      "step": 2140
+    },
+    {
+      "epoch": 0.9860395498114619,
+      "grad_norm": 3.0348900204468903,
+      "learning_rate": 1.002156406155419e-09,
+      "loss": 0.5682,
+      "step": 2141
+    },
+    {
+      "epoch": 0.9865001007455169,
+      "grad_norm": 3.115185553653833,
+      "learning_rate": 9.364697420476408e-10,
+      "loss": 0.4725,
+      "step": 2142
+    },
+    {
+      "epoch": 0.9869606516795717,
+      "grad_norm": 3.087394317229768,
+      "learning_rate": 8.730083779516784e-10,
+      "loss": 0.5686,
+      "step": 2143
+    },
+    {
+      "epoch": 0.9874212026136265,
+      "grad_norm": 2.8838196694727647,
+      "learning_rate": 8.117724552205718e-10,
+      "loss": 0.4791,
+      "step": 2144
+    },
+    {
+      "epoch": 0.9878817535476814,
+      "grad_norm": 2.9019511674207417,
+      "learning_rate": 7.527621102503268e-10,
+      "loss": 0.5693,
+      "step": 2145
+    },
+    {
+      "epoch": 0.9883423044817363,
+      "grad_norm": 3.153315732843879,
+      "learning_rate": 6.959774744796921e-10,
+      "loss": 0.6183,
+      "step": 2146
+    },
+    {
+      "epoch": 0.9888028554157912,
+      "grad_norm": 3.2841453445938162,
+      "learning_rate": 6.414186743899375e-10,
+      "loss": 0.4677,
+      "step": 2147
+    },
+    {
+      "epoch": 0.989263406349846,
+      "grad_norm": 3.205902530946899,
+      "learning_rate": 5.890858315046321e-10,
+      "loss": 0.5202,
+      "step": 2148
+    },
+    {
+      "epoch": 0.9897239572839008,
+      "grad_norm": 2.911194242328273,
+      "learning_rate": 5.389790623891999e-10,
+      "loss": 0.4913,
+      "step": 2149
+    },
+    {
+      "epoch": 0.9901845082179558,
+      "grad_norm": 3.087952168592988,
+      "learning_rate": 4.910984786506978e-10,
+      "loss": 0.5976,
+      "step": 2150
+    },
+    {
+      "epoch": 0.9906450591520106,
+      "grad_norm": 3.2125702072705162,
+      "learning_rate": 4.454441869377046e-10,
+      "loss": 0.5693,
+      "step": 2151
+    },
+    {
+      "epoch": 0.9911056100860655,
+      "grad_norm": 3.1911572214123365,
+      "learning_rate": 4.020162889399881e-10,
+      "loss": 0.5079,
+      "step": 2152
+    },
+    {
+      "epoch": 0.9915661610201203,
+      "grad_norm": 2.842869367824956,
+      "learning_rate": 3.6081488138817176e-10,
+      "loss": 0.5438,
+      "step": 2153
+    },
+    {
+      "epoch": 0.9920267119541751,
+      "grad_norm": 3.59761551281296,
+      "learning_rate": 3.2184005605373487e-10,
+      "loss": 0.6356,
+      "step": 2154
+    },
+    {
+      "epoch": 0.9924872628882301,
+      "grad_norm": 3.266624178638357,
+      "learning_rate": 2.850918997485685e-10,
+      "loss": 0.6384,
+      "step": 2155
+    },
+    {
+      "epoch": 0.9929478138222849,
+      "grad_norm": 2.847984241902026,
+      "learning_rate": 2.505704943251974e-10,
+      "loss": 0.4189,
+      "step": 2156
+    },
+    {
+      "epoch": 0.9934083647563398,
+      "grad_norm": 2.8940228158585684,
+      "learning_rate": 2.1827591667578083e-10,
+      "loss": 0.543,
+      "step": 2157
+    },
+    {
+      "epoch": 0.9938689156903946,
+      "grad_norm": 3.2268790504480798,
+      "learning_rate": 1.8820823873311187e-10,
+      "loss": 0.5865,
+      "step": 2158
+    },
+    {
+      "epoch": 0.9943294666244495,
+      "grad_norm": 2.816014447163771,
+      "learning_rate": 1.6036752746939608e-10,
+      "loss": 0.6104,
+      "step": 2159
+    },
+    {
+      "epoch": 0.9947900175585044,
+      "grad_norm": 2.834508001501944,
+      "learning_rate": 1.347538448966956e-10,
+      "loss": 0.5525,
+      "step": 2160
+    },
+    {
+      "epoch": 0.9952505684925592,
+      "grad_norm": 3.296408037769054,
+      "learning_rate": 1.113672480663741e-10,
+      "loss": 0.527,
+      "step": 2161
+    },
+    {
+      "epoch": 0.995711119426614,
+      "grad_norm": 3.1020643793075795,
+      "learning_rate": 9.020778906965176e-11,
+      "loss": 0.5439,
+      "step": 2162
+    },
+    {
+      "epoch": 0.996171670360669,
+      "grad_norm": 2.9887893240557477,
+      "learning_rate": 7.127551503671724e-11,
+      "loss": 0.5408,
+      "step": 2163
+    },
+    {
+      "epoch": 0.9966322212947238,
+      "grad_norm": 3.311987992840308,
+      "learning_rate": 5.4570468136949655e-11,
+      "loss": 0.5569,
+      "step": 2164
+    },
+    {
+      "epoch": 0.9970927722287787,
+      "grad_norm": 3.3309536253152436,
+      "learning_rate": 4.009268557902956e-11,
+      "loss": 0.5253,
+      "step": 2165
+    },
+    {
+      "epoch": 0.9975533231628335,
+      "grad_norm": 3.3046896055168546,
+      "learning_rate": 2.7842199610605965e-11,
+      "loss": 0.5117,
+      "step": 2166
+    },
+    {
+      "epoch": 0.9980138740968884,
+      "grad_norm": 3.1272281406984206,
+      "learning_rate": 1.7819037518185252e-11,
+      "loss": 0.5905,
+      "step": 2167
+    },
+    {
+      "epoch": 0.9984744250309433,
+      "grad_norm": 3.614597151505851,
+      "learning_rate": 1.0023221627242229e-11,
+      "loss": 0.4945,
+      "step": 2168
+    },
+    {
+      "epoch": 0.9989349759649981,
+      "grad_norm": 3.190348932550531,
+      "learning_rate": 4.454769301998063e-12,
+      "loss": 0.6752,
+      "step": 2169
+    },
+    {
+      "epoch": 0.999395526899053,
+      "grad_norm": 3.393120014197588,
+      "learning_rate": 1.1136929456423416e-12,
+      "loss": 0.5265,
+      "step": 2170
+    },
+    {
+      "epoch": 0.9998560778331078,
+      "grad_norm": 3.1321707930638376,
+      "learning_rate": 0.0,
+      "loss": 0.5797,
+      "step": 2171
+    },
+    {
+      "epoch": 0.9998560778331078,
+      "step": 2171,
+      "total_flos": 1478472491958272.0,
+      "train_loss": 0.577474653254122,
+      "train_runtime": 98261.8902,
+      "train_samples_per_second": 1.414,
+      "train_steps_per_second": 0.022
+    }
+  ],
+  "logging_steps": 1.0,
+  "max_steps": 2171,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 50000,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1478472491958272.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}