diff --git "a/checkpoint-2592/trainer_state.json" "b/checkpoint-2592/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/checkpoint-2592/trainer_state.json"
@@ -0,0 +1,18177 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.9998071359691417,
+  "eval_steps": 500,
+  "global_step": 2592,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0003857280617164899,
+      "grad_norm": 1.3028969772220882,
+      "learning_rate": 2.5000000000000004e-07,
+      "loss": 2.4718,
+      "step": 1
+    },
+    {
+      "epoch": 0.0007714561234329798,
+      "grad_norm": 1.3665860686662479,
+      "learning_rate": 5.000000000000001e-07,
+      "loss": 2.3575,
+      "step": 2
+    },
+    {
+      "epoch": 0.0011571841851494697,
+      "grad_norm": 1.3998982983137798,
+      "learning_rate": 7.5e-07,
+      "loss": 2.5221,
+      "step": 3
+    },
+    {
+      "epoch": 0.0015429122468659595,
+      "grad_norm": 1.4627027358629756,
+      "learning_rate": 1.0000000000000002e-06,
+      "loss": 2.4767,
+      "step": 4
+    },
+    {
+      "epoch": 0.0019286403085824494,
+      "grad_norm": 1.4038200232517333,
+      "learning_rate": 1.25e-06,
+      "loss": 2.4538,
+      "step": 5
+    },
+    {
+      "epoch": 0.0023143683702989393,
+      "grad_norm": 1.3419044955414192,
+      "learning_rate": 1.5e-06,
+      "loss": 2.3874,
+      "step": 6
+    },
+    {
+      "epoch": 0.002700096432015429,
+      "grad_norm": 1.3860959919367515,
+      "learning_rate": 1.75e-06,
+      "loss": 2.4335,
+      "step": 7
+    },
+    {
+      "epoch": 0.003085824493731919,
+      "grad_norm": 1.35282104723839,
+      "learning_rate": 2.0000000000000003e-06,
+      "loss": 2.4681,
+      "step": 8
+    },
+    {
+      "epoch": 0.0034715525554484088,
+      "grad_norm": 1.281614923512164,
+      "learning_rate": 2.25e-06,
+      "loss": 2.4464,
+      "step": 9
+    },
+    {
+      "epoch": 0.003857280617164899,
+      "grad_norm": 1.4254055492712447,
+      "learning_rate": 2.5e-06,
+      "loss": 2.4931,
+      "step": 10
+    },
+    {
+      "epoch": 0.004243008678881389,
+      "grad_norm": 1.3481677623706914,
+      "learning_rate": 2.7500000000000004e-06,
+      "loss": 2.4665,
+      "step": 11
+    },
+    {
+      "epoch": 0.004628736740597879,
+      "grad_norm": 1.2672866006199177,
+      "learning_rate": 3e-06,
+      "loss": 2.4167,
+      "step": 12
+    },
+    {
+      "epoch": 0.005014464802314368,
+      "grad_norm": 1.390249615802629,
+      "learning_rate": 3.2500000000000002e-06,
+      "loss": 2.4674,
+      "step": 13
+    },
+    {
+      "epoch": 0.005400192864030858,
+      "grad_norm": 1.305099944034975,
+      "learning_rate": 3.5e-06,
+      "loss": 2.3945,
+      "step": 14
+    },
+    {
+      "epoch": 0.0057859209257473485,
+      "grad_norm": 1.2796827824031904,
+      "learning_rate": 3.7500000000000005e-06,
+      "loss": 2.4574,
+      "step": 15
+    },
+    {
+      "epoch": 0.006171648987463838,
+      "grad_norm": 1.281187496954195,
+      "learning_rate": 4.000000000000001e-06,
+      "loss": 2.4474,
+      "step": 16
+    },
+    {
+      "epoch": 0.006557377049180328,
+      "grad_norm": 1.2706764452637,
+      "learning_rate": 4.25e-06,
+      "loss": 2.4759,
+      "step": 17
+    },
+    {
+      "epoch": 0.0069431051108968175,
+      "grad_norm": 1.1124376648153766,
+      "learning_rate": 4.5e-06,
+      "loss": 2.4301,
+      "step": 18
+    },
+    {
+      "epoch": 0.007328833172613307,
+      "grad_norm": 1.1285940335012594,
+      "learning_rate": 4.75e-06,
+      "loss": 2.4671,
+      "step": 19
+    },
+    {
+      "epoch": 0.007714561234329798,
+      "grad_norm": 1.0971409205597384,
+      "learning_rate": 5e-06,
+      "loss": 2.4581,
+      "step": 20
+    },
+    {
+      "epoch": 0.008100289296046287,
+      "grad_norm": 1.0101028043054117,
+      "learning_rate": 5.2500000000000006e-06,
+      "loss": 2.4187,
+      "step": 21
+    },
+    {
+      "epoch": 0.008486017357762778,
+      "grad_norm": 1.0183351895502508,
+      "learning_rate": 5.500000000000001e-06,
+      "loss": 2.4967,
+      "step": 22
+    },
+    {
+      "epoch": 0.008871745419479268,
+      "grad_norm": 0.9615041393264294,
+      "learning_rate": 5.75e-06,
+      "loss": 2.4402,
+      "step": 23
+    },
+    {
+      "epoch": 0.009257473481195757,
+      "grad_norm": 0.8512923775098707,
+      "learning_rate": 6e-06,
+      "loss": 2.4582,
+      "step": 24
+    },
+    {
+      "epoch": 0.009643201542912247,
+      "grad_norm": 0.7092461493595851,
+      "learning_rate": 6.25e-06,
+      "loss": 2.4808,
+      "step": 25
+    },
+    {
+      "epoch": 0.010028929604628737,
+      "grad_norm": 0.6819046573220403,
+      "learning_rate": 6.5000000000000004e-06,
+      "loss": 2.3906,
+      "step": 26
+    },
+    {
+      "epoch": 0.010414657666345226,
+      "grad_norm": 0.7115437695404054,
+      "learning_rate": 6.750000000000001e-06,
+      "loss": 2.407,
+      "step": 27
+    },
+    {
+      "epoch": 0.010800385728061716,
+      "grad_norm": 0.76903326342247,
+      "learning_rate": 7e-06,
+      "loss": 2.4771,
+      "step": 28
+    },
+    {
+      "epoch": 0.011186113789778206,
+      "grad_norm": 0.6338741014214005,
+      "learning_rate": 7.25e-06,
+      "loss": 2.4336,
+      "step": 29
+    },
+    {
+      "epoch": 0.011571841851494697,
+      "grad_norm": 0.640435250241746,
+      "learning_rate": 7.500000000000001e-06,
+      "loss": 2.408,
+      "step": 30
+    },
+    {
+      "epoch": 0.011957569913211187,
+      "grad_norm": 0.6176702235125447,
+      "learning_rate": 7.75e-06,
+      "loss": 2.4004,
+      "step": 31
+    },
+    {
+      "epoch": 0.012343297974927676,
+      "grad_norm": 0.609835959019172,
+      "learning_rate": 8.000000000000001e-06,
+      "loss": 2.3579,
+      "step": 32
+    },
+    {
+      "epoch": 0.012729026036644166,
+      "grad_norm": 0.5619355195918498,
+      "learning_rate": 8.25e-06,
+      "loss": 2.4959,
+      "step": 33
+    },
+    {
+      "epoch": 0.013114754098360656,
+      "grad_norm": 0.5757497035883407,
+      "learning_rate": 8.5e-06,
+      "loss": 2.3962,
+      "step": 34
+    },
+    {
+      "epoch": 0.013500482160077145,
+      "grad_norm": 0.5810404938379216,
+      "learning_rate": 8.750000000000001e-06,
+      "loss": 2.4677,
+      "step": 35
+    },
+    {
+      "epoch": 0.013886210221793635,
+      "grad_norm": 0.5710345213527928,
+      "learning_rate": 9e-06,
+      "loss": 2.341,
+      "step": 36
+    },
+    {
+      "epoch": 0.014271938283510125,
+      "grad_norm": 0.6565216595062554,
+      "learning_rate": 9.250000000000001e-06,
+      "loss": 2.4151,
+      "step": 37
+    },
+    {
+      "epoch": 0.014657666345226614,
+      "grad_norm": 0.5518571592310421,
+      "learning_rate": 9.5e-06,
+      "loss": 2.3074,
+      "step": 38
+    },
+    {
+      "epoch": 0.015043394406943106,
+      "grad_norm": 0.5187339842543488,
+      "learning_rate": 9.75e-06,
+      "loss": 2.4212,
+      "step": 39
+    },
+    {
+      "epoch": 0.015429122468659595,
+      "grad_norm": 0.5563829699152197,
+      "learning_rate": 1e-05,
+      "loss": 2.4743,
+      "step": 40
+    },
+    {
+      "epoch": 0.015814850530376085,
+      "grad_norm": 0.5041845776833512,
+      "learning_rate": 9.999996211403454e-06,
+      "loss": 2.3575,
+      "step": 41
+    },
+    {
+      "epoch": 0.016200578592092573,
+      "grad_norm": 0.48969986698988993,
+      "learning_rate": 9.999984845619553e-06,
+      "loss": 2.4481,
+      "step": 42
+    },
+    {
+      "epoch": 0.016586306653809064,
+      "grad_norm": 0.4802875056014886,
+      "learning_rate": 9.999965902665524e-06,
+      "loss": 2.4324,
+      "step": 43
+    },
+    {
+      "epoch": 0.016972034715525556,
+      "grad_norm": 0.4966005227112398,
+      "learning_rate": 9.999939382570075e-06,
+      "loss": 2.4681,
+      "step": 44
+    },
+    {
+      "epoch": 0.017357762777242044,
+      "grad_norm": 0.504548661198233,
+      "learning_rate": 9.999905285373392e-06,
+      "loss": 2.3829,
+      "step": 45
+    },
+    {
+      "epoch": 0.017743490838958535,
+      "grad_norm": 0.49530725000256803,
+      "learning_rate": 9.999863611127149e-06,
+      "loss": 2.3726,
+      "step": 46
+    },
+    {
+      "epoch": 0.018129218900675023,
+      "grad_norm": 0.496879485860879,
+      "learning_rate": 9.999814359894501e-06,
+      "loss": 2.355,
+      "step": 47
+    },
+    {
+      "epoch": 0.018514946962391515,
+      "grad_norm": 0.5349294236673756,
+      "learning_rate": 9.999757531750086e-06,
+      "loss": 2.3913,
+      "step": 48
+    },
+    {
+      "epoch": 0.018900675024108003,
+      "grad_norm": 0.48220878118791843,
+      "learning_rate": 9.999693126780022e-06,
+      "loss": 2.3789,
+      "step": 49
+    },
+    {
+      "epoch": 0.019286403085824494,
+      "grad_norm": 0.4778862019205235,
+      "learning_rate": 9.99962114508191e-06,
+      "loss": 2.4296,
+      "step": 50
+    },
+    {
+      "epoch": 0.019672131147540985,
+      "grad_norm": 0.4973185957898752,
+      "learning_rate": 9.999541586764836e-06,
+      "loss": 2.4465,
+      "step": 51
+    },
+    {
+      "epoch": 0.020057859209257473,
+      "grad_norm": 0.49336298166277864,
+      "learning_rate": 9.999454451949364e-06,
+      "loss": 2.4485,
+      "step": 52
+    },
+    {
+      "epoch": 0.020443587270973965,
+      "grad_norm": 0.48791115925851264,
+      "learning_rate": 9.999359740767545e-06,
+      "loss": 2.4737,
+      "step": 53
+    },
+    {
+      "epoch": 0.020829315332690453,
+      "grad_norm": 0.49124507643462106,
+      "learning_rate": 9.999257453362903e-06,
+      "loss": 2.364,
+      "step": 54
+    },
+    {
+      "epoch": 0.021215043394406944,
+      "grad_norm": 0.5260600198155415,
+      "learning_rate": 9.999147589890452e-06,
+      "loss": 2.3577,
+      "step": 55
+    },
+    {
+      "epoch": 0.021600771456123432,
+      "grad_norm": 0.4646617505959831,
+      "learning_rate": 9.999030150516681e-06,
+      "loss": 2.4449,
+      "step": 56
+    },
+    {
+      "epoch": 0.021986499517839923,
+      "grad_norm": 0.4841418940374766,
+      "learning_rate": 9.998905135419564e-06,
+      "loss": 2.4372,
+      "step": 57
+    },
+    {
+      "epoch": 0.02237222757955641,
+      "grad_norm": 0.48770580532041086,
+      "learning_rate": 9.998772544788552e-06,
+      "loss": 2.4455,
+      "step": 58
+    },
+    {
+      "epoch": 0.022757955641272903,
+      "grad_norm": 0.5108619805179353,
+      "learning_rate": 9.99863237882458e-06,
+      "loss": 2.4247,
+      "step": 59
+    },
+    {
+      "epoch": 0.023143683702989394,
+      "grad_norm": 0.48302169372077075,
+      "learning_rate": 9.998484637740058e-06,
+      "loss": 2.3931,
+      "step": 60
+    },
+    {
+      "epoch": 0.023529411764705882,
+      "grad_norm": 0.491126215051955,
+      "learning_rate": 9.998329321758882e-06,
+      "loss": 2.402,
+      "step": 61
+    },
+    {
+      "epoch": 0.023915139826422373,
+      "grad_norm": 0.46744989065036346,
+      "learning_rate": 9.998166431116421e-06,
+      "loss": 2.354,
+      "step": 62
+    },
+    {
+      "epoch": 0.02430086788813886,
+      "grad_norm": 0.5088343527314104,
+      "learning_rate": 9.997995966059526e-06,
+      "loss": 2.4644,
+      "step": 63
+    },
+    {
+      "epoch": 0.024686595949855353,
+      "grad_norm": 0.4601911278579166,
+      "learning_rate": 9.997817926846528e-06,
+      "loss": 2.3642,
+      "step": 64
+    },
+    {
+      "epoch": 0.02507232401157184,
+      "grad_norm": 0.45558237724734235,
+      "learning_rate": 9.997632313747236e-06,
+      "loss": 2.3585,
+      "step": 65
+    },
+    {
+      "epoch": 0.025458052073288332,
+      "grad_norm": 0.5098804291523603,
+      "learning_rate": 9.99743912704293e-06,
+      "loss": 2.4082,
+      "step": 66
+    },
+    {
+      "epoch": 0.02584378013500482,
+      "grad_norm": 0.48981215111458465,
+      "learning_rate": 9.997238367026376e-06,
+      "loss": 2.3659,
+      "step": 67
+    },
+    {
+      "epoch": 0.02622950819672131,
+      "grad_norm": 0.46146276348442455,
+      "learning_rate": 9.997030034001815e-06,
+      "loss": 2.4289,
+      "step": 68
+    },
+    {
+      "epoch": 0.026615236258437803,
+      "grad_norm": 0.4866095040257039,
+      "learning_rate": 9.99681412828496e-06,
+      "loss": 2.4155,
+      "step": 69
+    },
+    {
+      "epoch": 0.02700096432015429,
+      "grad_norm": 0.44699856443081387,
+      "learning_rate": 9.996590650203003e-06,
+      "loss": 2.418,
+      "step": 70
+    },
+    {
+      "epoch": 0.027386692381870782,
+      "grad_norm": 0.45814399872315087,
+      "learning_rate": 9.996359600094612e-06,
+      "loss": 2.3706,
+      "step": 71
+    },
+    {
+      "epoch": 0.02777242044358727,
+      "grad_norm": 0.4581038253559308,
+      "learning_rate": 9.99612097830993e-06,
+      "loss": 2.3944,
+      "step": 72
+    },
+    {
+      "epoch": 0.02815814850530376,
+      "grad_norm": 0.4436937845767424,
+      "learning_rate": 9.995874785210573e-06,
+      "loss": 2.3962,
+      "step": 73
+    },
+    {
+      "epoch": 0.02854387656702025,
+      "grad_norm": 0.5058877161626817,
+      "learning_rate": 9.995621021169632e-06,
+      "loss": 2.4237,
+      "step": 74
+    },
+    {
+      "epoch": 0.02892960462873674,
+      "grad_norm": 0.4793468391613926,
+      "learning_rate": 9.99535968657167e-06,
+      "loss": 2.4446,
+      "step": 75
+    },
+    {
+      "epoch": 0.02931533269045323,
+      "grad_norm": 0.4796523581998484,
+      "learning_rate": 9.995090781812724e-06,
+      "loss": 2.4206,
+      "step": 76
+    },
+    {
+      "epoch": 0.02970106075216972,
+      "grad_norm": 0.49382659971704934,
+      "learning_rate": 9.994814307300302e-06,
+      "loss": 2.3688,
+      "step": 77
+    },
+    {
+      "epoch": 0.03008678881388621,
+      "grad_norm": 0.5288191844079528,
+      "learning_rate": 9.994530263453385e-06,
+      "loss": 2.4649,
+      "step": 78
+    },
+    {
+      "epoch": 0.0304725168756027,
+      "grad_norm": 0.46921741488140223,
+      "learning_rate": 9.994238650702425e-06,
+      "loss": 2.3567,
+      "step": 79
+    },
+    {
+      "epoch": 0.03085824493731919,
+      "grad_norm": 0.4840280310683048,
+      "learning_rate": 9.993939469489342e-06,
+      "loss": 2.4211,
+      "step": 80
+    },
+    {
+      "epoch": 0.03124397299903568,
+      "grad_norm": 0.5169497417227851,
+      "learning_rate": 9.993632720267526e-06,
+      "loss": 2.4632,
+      "step": 81
+    },
+    {
+      "epoch": 0.03162970106075217,
+      "grad_norm": 0.6302728557829981,
+      "learning_rate": 9.993318403501838e-06,
+      "loss": 2.525,
+      "step": 82
+    },
+    {
+      "epoch": 0.03201542912246866,
+      "grad_norm": 0.46673916252446257,
+      "learning_rate": 9.992996519668603e-06,
+      "loss": 2.3906,
+      "step": 83
+    },
+    {
+      "epoch": 0.032401157184185146,
+      "grad_norm": 0.4938456663759333,
+      "learning_rate": 9.99266706925562e-06,
+      "loss": 2.4276,
+      "step": 84
+    },
+    {
+      "epoch": 0.03278688524590164,
+      "grad_norm": 0.4561673092917819,
+      "learning_rate": 9.99233005276215e-06,
+      "loss": 2.4089,
+      "step": 85
+    },
+    {
+      "epoch": 0.03317261330761813,
+      "grad_norm": 0.4775094525872891,
+      "learning_rate": 9.991985470698918e-06,
+      "loss": 2.4591,
+      "step": 86
+    },
+    {
+      "epoch": 0.03355834136933462,
+      "grad_norm": 0.4450031576193394,
+      "learning_rate": 9.99163332358812e-06,
+      "loss": 2.3826,
+      "step": 87
+    },
+    {
+      "epoch": 0.03394406943105111,
+      "grad_norm": 0.4728950026349946,
+      "learning_rate": 9.991273611963413e-06,
+      "loss": 2.3936,
+      "step": 88
+    },
+    {
+      "epoch": 0.0343297974927676,
+      "grad_norm": 0.4504962222392687,
+      "learning_rate": 9.990906336369917e-06,
+      "loss": 2.4611,
+      "step": 89
+    },
+    {
+      "epoch": 0.03471552555448409,
+      "grad_norm": 0.45006031421593823,
+      "learning_rate": 9.990531497364215e-06,
+      "loss": 2.3998,
+      "step": 90
+    },
+    {
+      "epoch": 0.035101253616200576,
+      "grad_norm": 0.45430590871658183,
+      "learning_rate": 9.990149095514354e-06,
+      "loss": 2.4285,
+      "step": 91
+    },
+    {
+      "epoch": 0.03548698167791707,
+      "grad_norm": 0.48930021919773625,
+      "learning_rate": 9.98975913139984e-06,
+      "loss": 2.4687,
+      "step": 92
+    },
+    {
+      "epoch": 0.03587270973963356,
+      "grad_norm": 0.4458311534356548,
+      "learning_rate": 9.989361605611638e-06,
+      "loss": 2.439,
+      "step": 93
+    },
+    {
+      "epoch": 0.036258437801350046,
+      "grad_norm": 0.4829210847535642,
+      "learning_rate": 9.988956518752178e-06,
+      "loss": 2.3716,
+      "step": 94
+    },
+    {
+      "epoch": 0.03664416586306654,
+      "grad_norm": 0.4860495403190378,
+      "learning_rate": 9.988543871435342e-06,
+      "loss": 2.3246,
+      "step": 95
+    },
+    {
+      "epoch": 0.03702989392478303,
+      "grad_norm": 0.4590978918787668,
+      "learning_rate": 9.98812366428647e-06,
+      "loss": 2.3632,
+      "step": 96
+    },
+    {
+      "epoch": 0.03741562198649952,
+      "grad_norm": 0.5170375903870793,
+      "learning_rate": 9.98769589794236e-06,
+      "loss": 2.3509,
+      "step": 97
+    },
+    {
+      "epoch": 0.037801350048216005,
+      "grad_norm": 0.4930988306444351,
+      "learning_rate": 9.987260573051268e-06,
+      "loss": 2.4131,
+      "step": 98
+    },
+    {
+      "epoch": 0.0381870781099325,
+      "grad_norm": 0.5027035108292636,
+      "learning_rate": 9.986817690272902e-06,
+      "loss": 2.3941,
+      "step": 99
+    },
+    {
+      "epoch": 0.03857280617164899,
+      "grad_norm": 0.5253968168941702,
+      "learning_rate": 9.986367250278423e-06,
+      "loss": 2.3307,
+      "step": 100
+    },
+    {
+      "epoch": 0.038958534233365476,
+      "grad_norm": 0.4719637441013078,
+      "learning_rate": 9.985909253750446e-06,
+      "loss": 2.3666,
+      "step": 101
+    },
+    {
+      "epoch": 0.03934426229508197,
+      "grad_norm": 0.4712787856936807,
+      "learning_rate": 9.985443701383035e-06,
+      "loss": 2.3941,
+      "step": 102
+    },
+    {
+      "epoch": 0.03972999035679846,
+      "grad_norm": 0.46114806739089415,
+      "learning_rate": 9.984970593881706e-06,
+      "loss": 2.4337,
+      "step": 103
+    },
+    {
+      "epoch": 0.040115718418514947,
+      "grad_norm": 0.46850436431850573,
+      "learning_rate": 9.984489931963429e-06,
+      "loss": 2.4107,
+      "step": 104
+    },
+    {
+      "epoch": 0.040501446480231434,
+      "grad_norm": 0.49937095248671337,
+      "learning_rate": 9.984001716356611e-06,
+      "loss": 2.382,
+      "step": 105
+    },
+    {
+      "epoch": 0.04088717454194793,
+      "grad_norm": 0.4597593612632286,
+      "learning_rate": 9.983505947801115e-06,
+      "loss": 2.4465,
+      "step": 106
+    },
+    {
+      "epoch": 0.04127290260366442,
+      "grad_norm": 0.5102293641150926,
+      "learning_rate": 9.983002627048248e-06,
+      "loss": 2.4051,
+      "step": 107
+    },
+    {
+      "epoch": 0.041658630665380905,
+      "grad_norm": 0.48329948878166523,
+      "learning_rate": 9.982491754860763e-06,
+      "loss": 2.4089,
+      "step": 108
+    },
+    {
+      "epoch": 0.04204435872709739,
+      "grad_norm": 0.4985851779622545,
+      "learning_rate": 9.981973332012856e-06,
+      "loss": 2.3836,
+      "step": 109
+    },
+    {
+      "epoch": 0.04243008678881389,
+      "grad_norm": 0.46047059896927356,
+      "learning_rate": 9.981447359290162e-06,
+      "loss": 2.3619,
+      "step": 110
+    },
+    {
+      "epoch": 0.042815814850530376,
+      "grad_norm": 0.45941412104146895,
+      "learning_rate": 9.980913837489763e-06,
+      "loss": 2.4467,
+      "step": 111
+    },
+    {
+      "epoch": 0.043201542912246864,
+      "grad_norm": 0.5279912896817367,
+      "learning_rate": 9.980372767420179e-06,
+      "loss": 2.3873,
+      "step": 112
+    },
+    {
+      "epoch": 0.04358727097396336,
+      "grad_norm": 0.5173284181396758,
+      "learning_rate": 9.979824149901365e-06,
+      "loss": 2.3712,
+      "step": 113
+    },
+    {
+      "epoch": 0.04397299903567985,
+      "grad_norm": 0.46425124191934264,
+      "learning_rate": 9.979267985764717e-06,
+      "loss": 2.3972,
+      "step": 114
+    },
+    {
+      "epoch": 0.044358727097396335,
+      "grad_norm": 0.47210175255183745,
+      "learning_rate": 9.978704275853073e-06,
+      "loss": 2.4281,
+      "step": 115
+    },
+    {
+      "epoch": 0.04474445515911282,
+      "grad_norm": 0.5117270444340153,
+      "learning_rate": 9.978133021020697e-06,
+      "loss": 2.5032,
+      "step": 116
+    },
+    {
+      "epoch": 0.04513018322082932,
+      "grad_norm": 0.45237840748789165,
+      "learning_rate": 9.977554222133293e-06,
+      "loss": 2.2994,
+      "step": 117
+    },
+    {
+      "epoch": 0.045515911282545805,
+      "grad_norm": 0.4829428321396551,
+      "learning_rate": 9.97696788006799e-06,
+      "loss": 2.3987,
+      "step": 118
+    },
+    {
+      "epoch": 0.04590163934426229,
+      "grad_norm": 0.4576774127420072,
+      "learning_rate": 9.976373995713358e-06,
+      "loss": 2.3653,
+      "step": 119
+    },
+    {
+      "epoch": 0.04628736740597879,
+      "grad_norm": 0.4743705723346769,
+      "learning_rate": 9.97577256996939e-06,
+      "loss": 2.4184,
+      "step": 120
+    },
+    {
+      "epoch": 0.046673095467695276,
+      "grad_norm": 0.46090426942328017,
+      "learning_rate": 9.975163603747513e-06,
+      "loss": 2.3988,
+      "step": 121
+    },
+    {
+      "epoch": 0.047058823529411764,
+      "grad_norm": 0.452297730740023,
+      "learning_rate": 9.974547097970576e-06,
+      "loss": 2.4127,
+      "step": 122
+    },
+    {
+      "epoch": 0.04744455159112825,
+      "grad_norm": 0.45844755127239656,
+      "learning_rate": 9.973923053572854e-06,
+      "loss": 2.3684,
+      "step": 123
+    },
+    {
+      "epoch": 0.04783027965284475,
+      "grad_norm": 0.49492845581087924,
+      "learning_rate": 9.97329147150005e-06,
+      "loss": 2.4074,
+      "step": 124
+    },
+    {
+      "epoch": 0.048216007714561235,
+      "grad_norm": 0.484978985690606,
+      "learning_rate": 9.972652352709287e-06,
+      "loss": 2.468,
+      "step": 125
+    },
+    {
+      "epoch": 0.04860173577627772,
+      "grad_norm": 0.4611298393733667,
+      "learning_rate": 9.972005698169112e-06,
+      "loss": 2.4323,
+      "step": 126
+    },
+    {
+      "epoch": 0.04898746383799421,
+      "grad_norm": 0.4646255015376031,
+      "learning_rate": 9.971351508859488e-06,
+      "loss": 2.4634,
+      "step": 127
+    },
+    {
+      "epoch": 0.049373191899710706,
+      "grad_norm": 0.46220388970928555,
+      "learning_rate": 9.970689785771798e-06,
+      "loss": 2.3982,
+      "step": 128
+    },
+    {
+      "epoch": 0.049758919961427193,
+      "grad_norm": 0.47071092914994284,
+      "learning_rate": 9.970020529908846e-06,
+      "loss": 2.4697,
+      "step": 129
+    },
+    {
+      "epoch": 0.05014464802314368,
+      "grad_norm": 0.5604351159632398,
+      "learning_rate": 9.969343742284847e-06,
+      "loss": 2.3377,
+      "step": 130
+    },
+    {
+      "epoch": 0.050530376084860176,
+      "grad_norm": 0.47620334301858896,
+      "learning_rate": 9.968659423925429e-06,
+      "loss": 2.3535,
+      "step": 131
+    },
+    {
+      "epoch": 0.050916104146576664,
+      "grad_norm": 0.4828800048233955,
+      "learning_rate": 9.96796757586764e-06,
+      "loss": 2.4521,
+      "step": 132
+    },
+    {
+      "epoch": 0.05130183220829315,
+      "grad_norm": 0.4595892831688092,
+      "learning_rate": 9.967268199159926e-06,
+      "loss": 2.3431,
+      "step": 133
+    },
+    {
+      "epoch": 0.05168756027000964,
+      "grad_norm": 0.46535351807962866,
+      "learning_rate": 9.96656129486215e-06,
+      "loss": 2.336,
+      "step": 134
+    },
+    {
+      "epoch": 0.052073288331726135,
+      "grad_norm": 0.5597994767184472,
+      "learning_rate": 9.96584686404559e-06,
+      "loss": 2.399,
+      "step": 135
+    },
+    {
+      "epoch": 0.05245901639344262,
+      "grad_norm": 0.5182784238006419,
+      "learning_rate": 9.965124907792916e-06,
+      "loss": 2.419,
+      "step": 136
+    },
+    {
+      "epoch": 0.05284474445515911,
+      "grad_norm": 0.484806699314445,
+      "learning_rate": 9.964395427198208e-06,
+      "loss": 2.4725,
+      "step": 137
+    },
+    {
+      "epoch": 0.053230472516875606,
+      "grad_norm": 0.46748733607259835,
+      "learning_rate": 9.963658423366951e-06,
+      "loss": 2.3721,
+      "step": 138
+    },
+    {
+      "epoch": 0.053616200578592094,
+      "grad_norm": 0.45804476035104996,
+      "learning_rate": 9.962913897416029e-06,
+      "loss": 2.4362,
+      "step": 139
+    },
+    {
+      "epoch": 0.05400192864030858,
+      "grad_norm": 0.5644951235988059,
+      "learning_rate": 9.962161850473723e-06,
+      "loss": 2.4108,
+      "step": 140
+    },
+    {
+      "epoch": 0.05438765670202507,
+      "grad_norm": 0.4584390238284688,
+      "learning_rate": 9.961402283679718e-06,
+      "loss": 2.3989,
+      "step": 141
+    },
+    {
+      "epoch": 0.054773384763741564,
+      "grad_norm": 0.4539282651014262,
+      "learning_rate": 9.960635198185088e-06,
+      "loss": 2.3916,
+      "step": 142
+    },
+    {
+      "epoch": 0.05515911282545805,
+      "grad_norm": 0.4413326116727136,
+      "learning_rate": 9.959860595152305e-06,
+      "loss": 2.3971,
+      "step": 143
+    },
+    {
+      "epoch": 0.05554484088717454,
+      "grad_norm": 0.4883760828911208,
+      "learning_rate": 9.95907847575523e-06,
+      "loss": 2.3996,
+      "step": 144
+    },
+    {
+      "epoch": 0.055930568948891035,
+      "grad_norm": 0.4509831906670727,
+      "learning_rate": 9.958288841179121e-06,
+      "loss": 2.3887,
+      "step": 145
+    },
+    {
+      "epoch": 0.05631629701060752,
+      "grad_norm": 0.48204373539948087,
+      "learning_rate": 9.957491692620618e-06,
+      "loss": 2.3994,
+      "step": 146
+    },
+    {
+      "epoch": 0.05670202507232401,
+      "grad_norm": 0.5064694164808567,
+      "learning_rate": 9.956687031287752e-06,
+      "loss": 2.4212,
+      "step": 147
+    },
+    {
+      "epoch": 0.0570877531340405,
+      "grad_norm": 0.47286697840128156,
+      "learning_rate": 9.955874858399936e-06,
+      "loss": 2.4378,
+      "step": 148
+    },
+    {
+      "epoch": 0.057473481195756994,
+      "grad_norm": 0.45461008312404527,
+      "learning_rate": 9.955055175187971e-06,
+      "loss": 2.427,
+      "step": 149
+    },
+    {
+      "epoch": 0.05785920925747348,
+      "grad_norm": 0.5304097965230556,
+      "learning_rate": 9.954227982894034e-06,
+      "loss": 2.3624,
+      "step": 150
+    },
+    {
+      "epoch": 0.05824493731918997,
+      "grad_norm": 0.48513100057264075,
+      "learning_rate": 9.953393282771686e-06,
+      "loss": 2.3925,
+      "step": 151
+    },
+    {
+      "epoch": 0.05863066538090646,
+      "grad_norm": 0.5064590654892867,
+      "learning_rate": 9.952551076085864e-06,
+      "loss": 2.4262,
+      "step": 152
+    },
+    {
+      "epoch": 0.05901639344262295,
+      "grad_norm": 0.4579406686580277,
+      "learning_rate": 9.951701364112877e-06,
+      "loss": 2.311,
+      "step": 153
+    },
+    {
+      "epoch": 0.05940212150433944,
+      "grad_norm": 0.48021046645886806,
+      "learning_rate": 9.950844148140414e-06,
+      "loss": 2.414,
+      "step": 154
+    },
+    {
+      "epoch": 0.05978784956605593,
+      "grad_norm": 0.4384289042787429,
+      "learning_rate": 9.949979429467534e-06,
+      "loss": 2.4737,
+      "step": 155
+    },
+    {
+      "epoch": 0.06017357762777242,
+      "grad_norm": 0.4860862476128489,
+      "learning_rate": 9.949107209404664e-06,
+      "loss": 2.4112,
+      "step": 156
+    },
+    {
+      "epoch": 0.06055930568948891,
+      "grad_norm": 0.47596173584160284,
+      "learning_rate": 9.948227489273601e-06,
+      "loss": 2.4056,
+      "step": 157
+    },
+    {
+      "epoch": 0.0609450337512054,
+      "grad_norm": 0.5025215866684043,
+      "learning_rate": 9.947340270407504e-06,
+      "loss": 2.4447,
+      "step": 158
+    },
+    {
+      "epoch": 0.06133076181292189,
+      "grad_norm": 0.46084735188375686,
+      "learning_rate": 9.946445554150902e-06,
+      "loss": 2.462,
+      "step": 159
+    },
+    {
+      "epoch": 0.06171648987463838,
+      "grad_norm": 0.45982518950698714,
+      "learning_rate": 9.945543341859681e-06,
+      "loss": 2.3593,
+      "step": 160
+    },
+    {
+      "epoch": 0.06210221793635487,
+      "grad_norm": 0.4945780654463559,
+      "learning_rate": 9.94463363490109e-06,
+      "loss": 2.3348,
+      "step": 161
+    },
+    {
+      "epoch": 0.06248794599807136,
+      "grad_norm": 0.4872729237867549,
+      "learning_rate": 9.94371643465373e-06,
+      "loss": 2.392,
+      "step": 162
+    },
+    {
+      "epoch": 0.06287367405978785,
+      "grad_norm": 0.4605048563245808,
+      "learning_rate": 9.942791742507565e-06,
+      "loss": 2.4605,
+      "step": 163
+    },
+    {
+      "epoch": 0.06325940212150434,
+      "grad_norm": 0.4482486921811318,
+      "learning_rate": 9.94185955986391e-06,
+      "loss": 2.4488,
+      "step": 164
+    },
+    {
+      "epoch": 0.06364513018322084,
+      "grad_norm": 0.47037373849736897,
+      "learning_rate": 9.940919888135428e-06,
+      "loss": 2.3973,
+      "step": 165
+    },
+    {
+      "epoch": 0.06403085824493732,
+      "grad_norm": 0.45236768246003195,
+      "learning_rate": 9.939972728746134e-06,
+      "loss": 2.3402,
+      "step": 166
+    },
+    {
+      "epoch": 0.06441658630665381,
+      "grad_norm": 0.4779222209932742,
+      "learning_rate": 9.939018083131391e-06,
+      "loss": 2.363,
+      "step": 167
+    },
+    {
+      "epoch": 0.06480231436837029,
+      "grad_norm": 0.4901328716880792,
+      "learning_rate": 9.938055952737908e-06,
+      "loss": 2.3983,
+      "step": 168
+    },
+    {
+      "epoch": 0.06518804243008679,
+      "grad_norm": 0.4750880213070407,
+      "learning_rate": 9.937086339023731e-06,
+      "loss": 2.4079,
+      "step": 169
+    },
+    {
+      "epoch": 0.06557377049180328,
+      "grad_norm": 0.4706342992387472,
+      "learning_rate": 9.93610924345825e-06,
+      "loss": 2.3022,
+      "step": 170
+    },
+    {
+      "epoch": 0.06595949855351976,
+      "grad_norm": 0.50942295471628,
+      "learning_rate": 9.935124667522196e-06,
+      "loss": 2.4003,
+      "step": 171
+    },
+    {
+      "epoch": 0.06634522661523626,
+      "grad_norm": 0.4648678544986133,
+      "learning_rate": 9.934132612707631e-06,
+      "loss": 2.3952,
+      "step": 172
+    },
+    {
+      "epoch": 0.06673095467695275,
+      "grad_norm": 0.4849676435310516,
+      "learning_rate": 9.933133080517956e-06,
+      "loss": 2.3687,
+      "step": 173
+    },
+    {
+      "epoch": 0.06711668273866923,
+      "grad_norm": 0.45490408653760767,
+      "learning_rate": 9.932126072467897e-06,
+      "loss": 2.3959,
+      "step": 174
+    },
+    {
+      "epoch": 0.06750241080038573,
+      "grad_norm": 0.46281412003095096,
+      "learning_rate": 9.931111590083516e-06,
+      "loss": 2.4556,
+      "step": 175
+    },
+    {
+      "epoch": 0.06788813886210222,
+      "grad_norm": 0.5152142848860477,
+      "learning_rate": 9.930089634902197e-06,
+      "loss": 2.4501,
+      "step": 176
+    },
+    {
+      "epoch": 0.0682738669238187,
+      "grad_norm": 0.4688795671612781,
+      "learning_rate": 9.92906020847265e-06,
+      "loss": 2.4516,
+      "step": 177
+    },
+    {
+      "epoch": 0.0686595949855352,
+      "grad_norm": 0.45994214817232393,
+      "learning_rate": 9.92802331235491e-06,
+      "loss": 2.4451,
+      "step": 178
+    },
+    {
+      "epoch": 0.0690453230472517,
+      "grad_norm": 0.4621313038771384,
+      "learning_rate": 9.926978948120327e-06,
+      "loss": 2.4197,
+      "step": 179
+    },
+    {
+      "epoch": 0.06943105110896818,
+      "grad_norm": 0.4582006758782177,
+      "learning_rate": 9.925927117351573e-06,
+      "loss": 2.4572,
+      "step": 180
+    },
+    {
+      "epoch": 0.06981677917068467,
+      "grad_norm": 0.4779378215804943,
+      "learning_rate": 9.92486782164263e-06,
+      "loss": 2.3648,
+      "step": 181
+    },
+    {
+      "epoch": 0.07020250723240115,
+      "grad_norm": 0.44552167365942835,
+      "learning_rate": 9.923801062598799e-06,
+      "loss": 2.3999,
+      "step": 182
+    },
+    {
+      "epoch": 0.07058823529411765,
+      "grad_norm": 0.4962343901077161,
+      "learning_rate": 9.922726841836685e-06,
+      "loss": 2.3891,
+      "step": 183
+    },
+    {
+      "epoch": 0.07097396335583414,
+      "grad_norm": 0.4859300553836987,
+      "learning_rate": 9.921645160984205e-06,
+      "loss": 2.3188,
+      "step": 184
+    },
+    {
+      "epoch": 0.07135969141755062,
+      "grad_norm": 0.49044136281300504,
+      "learning_rate": 9.92055602168058e-06,
+      "loss": 2.4622,
+      "step": 185
+    },
+    {
+      "epoch": 0.07174541947926712,
+      "grad_norm": 0.5242947128137062,
+      "learning_rate": 9.919459425576334e-06,
+      "loss": 2.3797,
+      "step": 186
+    },
+    {
+      "epoch": 0.07213114754098361,
+      "grad_norm": 0.4982936128608122,
+      "learning_rate": 9.918355374333292e-06,
+      "loss": 2.4064,
+      "step": 187
+    },
+    {
+      "epoch": 0.07251687560270009,
+      "grad_norm": 0.47856529619520927,
+      "learning_rate": 9.917243869624573e-06,
+      "loss": 2.2904,
+      "step": 188
+    },
+    {
+      "epoch": 0.07290260366441659,
+      "grad_norm": 0.4760751948697406,
+      "learning_rate": 9.916124913134594e-06,
+      "loss": 2.3582,
+      "step": 189
+    },
+    {
+      "epoch": 0.07328833172613308,
+      "grad_norm": 0.4639298555212939,
+      "learning_rate": 9.91499850655907e-06,
+      "loss": 2.3503,
+      "step": 190
+    },
+    {
+      "epoch": 0.07367405978784956,
+      "grad_norm": 0.4807769790534427,
+      "learning_rate": 9.913864651604996e-06,
+      "loss": 2.399,
+      "step": 191
+    },
+    {
+      "epoch": 0.07405978784956606,
+      "grad_norm": 0.5044607391473807,
+      "learning_rate": 9.91272334999066e-06,
+      "loss": 2.3792,
+      "step": 192
+    },
+    {
+      "epoch": 0.07444551591128254,
+      "grad_norm": 0.4575543809543949,
+      "learning_rate": 9.911574603445637e-06,
+      "loss": 2.3502,
+      "step": 193
+    },
+    {
+      "epoch": 0.07483124397299903,
+      "grad_norm": 0.46442861911525574,
+      "learning_rate": 9.91041841371078e-06,
+      "loss": 2.3847,
+      "step": 194
+    },
+    {
+      "epoch": 0.07521697203471553,
+      "grad_norm": 0.4663103425438184,
+      "learning_rate": 9.909254782538225e-06,
+      "loss": 2.3995,
+      "step": 195
+    },
+    {
+      "epoch": 0.07560270009643201,
+      "grad_norm": 0.502061724200111,
+      "learning_rate": 9.908083711691383e-06,
+      "loss": 2.3949,
+      "step": 196
+    },
+    {
+      "epoch": 0.0759884281581485,
+      "grad_norm": 0.4772691500713791,
+      "learning_rate": 9.906905202944939e-06,
+      "loss": 2.3846,
+      "step": 197
+    },
+    {
+      "epoch": 0.076374156219865,
+      "grad_norm": 0.4568898979377072,
+      "learning_rate": 9.905719258084852e-06,
+      "loss": 2.3305,
+      "step": 198
+    },
+    {
+      "epoch": 0.07675988428158148,
+      "grad_norm": 0.4249948751433217,
+      "learning_rate": 9.904525878908347e-06,
+      "loss": 2.3695,
+      "step": 199
+    },
+    {
+      "epoch": 0.07714561234329798,
+      "grad_norm": 0.48786089598149546,
+      "learning_rate": 9.903325067223918e-06,
+      "loss": 2.35,
+      "step": 200
+    },
+    {
+      "epoch": 0.07753134040501447,
+      "grad_norm": 0.47807928221591933,
+      "learning_rate": 9.902116824851323e-06,
+      "loss": 2.3781,
+      "step": 201
+    },
+    {
+      "epoch": 0.07791706846673095,
+      "grad_norm": 0.46074562043721295,
+      "learning_rate": 9.900901153621576e-06,
+      "loss": 2.3326,
+      "step": 202
+    },
+    {
+      "epoch": 0.07830279652844745,
+      "grad_norm": 0.4785080056291243,
+      "learning_rate": 9.899678055376955e-06,
+      "loss": 2.352,
+      "step": 203
+    },
+    {
+      "epoch": 0.07868852459016394,
+      "grad_norm": 0.49822216149179815,
+      "learning_rate": 9.898447531970989e-06,
+      "loss": 2.329,
+      "step": 204
+    },
+    {
+      "epoch": 0.07907425265188042,
+      "grad_norm": 0.5074455882320913,
+      "learning_rate": 9.897209585268459e-06,
+      "loss": 2.2767,
+      "step": 205
+    },
+    {
+      "epoch": 0.07945998071359692,
+      "grad_norm": 0.48299990133600534,
+      "learning_rate": 9.8959642171454e-06,
+      "loss": 2.3657,
+      "step": 206
+    },
+    {
+      "epoch": 0.0798457087753134,
+      "grad_norm": 0.4606233819521811,
+      "learning_rate": 9.89471142948909e-06,
+      "loss": 2.4016,
+      "step": 207
+    },
+    {
+      "epoch": 0.08023143683702989,
+      "grad_norm": 0.44478840251929497,
+      "learning_rate": 9.893451224198051e-06,
+      "loss": 2.3426,
+      "step": 208
+    },
+    {
+      "epoch": 0.08061716489874639,
+      "grad_norm": 0.46445194930483236,
+      "learning_rate": 9.892183603182048e-06,
+      "loss": 2.3325,
+      "step": 209
+    },
+    {
+      "epoch": 0.08100289296046287,
+      "grad_norm": 0.5371784356002836,
+      "learning_rate": 9.890908568362083e-06,
+      "loss": 2.4046,
+      "step": 210
+    },
+    {
+      "epoch": 0.08138862102217936,
+      "grad_norm": 0.4971218959499928,
+      "learning_rate": 9.889626121670391e-06,
+      "loss": 2.3784,
+      "step": 211
+    },
+    {
+      "epoch": 0.08177434908389586,
+      "grad_norm": 0.4783572091434314,
+      "learning_rate": 9.888336265050443e-06,
+      "loss": 2.3865,
+      "step": 212
+    },
+    {
+      "epoch": 0.08216007714561234,
+      "grad_norm": 0.510470355835948,
+      "learning_rate": 9.887039000456937e-06,
+      "loss": 2.3624,
+      "step": 213
+    },
+    {
+      "epoch": 0.08254580520732883,
+      "grad_norm": 0.4613120488085547,
+      "learning_rate": 9.885734329855798e-06,
+      "loss": 2.4258,
+      "step": 214
+    },
+    {
+      "epoch": 0.08293153326904533,
+      "grad_norm": 0.43709866085432875,
+      "learning_rate": 9.884422255224175e-06,
+      "loss": 2.308,
+      "step": 215
+    },
+    {
+      "epoch": 0.08331726133076181,
+      "grad_norm": 0.4857474219872402,
+      "learning_rate": 9.883102778550434e-06,
+      "loss": 2.3864,
+      "step": 216
+    },
+    {
+      "epoch": 0.0837029893924783,
+      "grad_norm": 0.5003412853512436,
+      "learning_rate": 9.881775901834164e-06,
+      "loss": 2.4105,
+      "step": 217
+    },
+    {
+      "epoch": 0.08408871745419479,
+      "grad_norm": 0.48105392183158313,
+      "learning_rate": 9.880441627086163e-06,
+      "loss": 2.4384,
+      "step": 218
+    },
+    {
+      "epoch": 0.08447444551591128,
+      "grad_norm": 0.46812824003346093,
+      "learning_rate": 9.879099956328443e-06,
+      "loss": 2.4241,
+      "step": 219
+    },
+    {
+      "epoch": 0.08486017357762778,
+      "grad_norm": 0.5071249854343619,
+      "learning_rate": 9.877750891594224e-06,
+      "loss": 2.3748,
+      "step": 220
+    },
+    {
+      "epoch": 0.08524590163934426,
+      "grad_norm": 0.47055633644804373,
+      "learning_rate": 9.876394434927931e-06,
+      "loss": 2.3642,
+      "step": 221
+    },
+    {
+      "epoch": 0.08563162970106075,
+      "grad_norm": 0.4913503211525076,
+      "learning_rate": 9.875030588385192e-06,
+      "loss": 2.3467,
+      "step": 222
+    },
+    {
+      "epoch": 0.08601735776277725,
+      "grad_norm": 0.4742540637522417,
+      "learning_rate": 9.873659354032829e-06,
+      "loss": 2.345,
+      "step": 223
+    },
+    {
+      "epoch": 0.08640308582449373,
+      "grad_norm": 0.4627514802436143,
+      "learning_rate": 9.872280733948867e-06,
+      "loss": 2.4574,
+      "step": 224
+    },
+    {
+      "epoch": 0.08678881388621022,
+      "grad_norm": 0.45535902185916416,
+      "learning_rate": 9.87089473022252e-06,
+      "loss": 2.4138,
+      "step": 225
+    },
+    {
+      "epoch": 0.08717454194792672,
+      "grad_norm": 0.43825210670158116,
+      "learning_rate": 9.869501344954188e-06,
+      "loss": 2.4522,
+      "step": 226
+    },
+    {
+      "epoch": 0.0875602700096432,
+      "grad_norm": 0.5354119753902198,
+      "learning_rate": 9.868100580255466e-06,
+      "loss": 2.3617,
+      "step": 227
+    },
+    {
+      "epoch": 0.0879459980713597,
+      "grad_norm": 0.46706788377779374,
+      "learning_rate": 9.866692438249124e-06,
+      "loss": 2.4118,
+      "step": 228
+    },
+    {
+      "epoch": 0.08833172613307617,
+      "grad_norm": 0.4471121597083322,
+      "learning_rate": 9.865276921069113e-06,
+      "loss": 2.3571,
+      "step": 229
+    },
+    {
+      "epoch": 0.08871745419479267,
+      "grad_norm": 0.4572982101835507,
+      "learning_rate": 9.863854030860566e-06,
+      "loss": 2.3805,
+      "step": 230
+    },
+    {
+      "epoch": 0.08910318225650916,
+      "grad_norm": 0.45200242655619743,
+      "learning_rate": 9.862423769779784e-06,
+      "loss": 2.4114,
+      "step": 231
+    },
+    {
+      "epoch": 0.08948891031822565,
+      "grad_norm": 0.48459327556912984,
+      "learning_rate": 9.86098613999424e-06,
+      "loss": 2.4248,
+      "step": 232
+    },
+    {
+      "epoch": 0.08987463837994214,
+      "grad_norm": 0.4382958973086269,
+      "learning_rate": 9.859541143682573e-06,
+      "loss": 2.3683,
+      "step": 233
+    },
+    {
+      "epoch": 0.09026036644165863,
+      "grad_norm": 0.5249998518789483,
+      "learning_rate": 9.858088783034587e-06,
+      "loss": 2.3673,
+      "step": 234
+    },
+    {
+      "epoch": 0.09064609450337512,
+      "grad_norm": 0.49347712460532517,
+      "learning_rate": 9.856629060251247e-06,
+      "loss": 2.486,
+      "step": 235
+    },
+    {
+      "epoch": 0.09103182256509161,
+      "grad_norm": 0.5340059340122717,
+      "learning_rate": 9.855161977544672e-06,
+      "loss": 2.4404,
+      "step": 236
+    },
+    {
+      "epoch": 0.0914175506268081,
+      "grad_norm": 0.5052074730677552,
+      "learning_rate": 9.853687537138132e-06,
+      "loss": 2.3487,
+      "step": 237
+    },
+    {
+      "epoch": 0.09180327868852459,
+      "grad_norm": 0.4584681855122188,
+      "learning_rate": 9.852205741266058e-06,
+      "loss": 2.4361,
+      "step": 238
+    },
+    {
+      "epoch": 0.09218900675024108,
+      "grad_norm": 0.45006330044968135,
+      "learning_rate": 9.850716592174016e-06,
+      "loss": 2.4544,
+      "step": 239
+    },
+    {
+      "epoch": 0.09257473481195758,
+      "grad_norm": 0.45259872735656165,
+      "learning_rate": 9.849220092118721e-06,
+      "loss": 2.2847,
+      "step": 240
+    },
+    {
+      "epoch": 0.09296046287367406,
+      "grad_norm": 0.4868831359383384,
+      "learning_rate": 9.847716243368027e-06,
+      "loss": 2.3389,
+      "step": 241
+    },
+    {
+      "epoch": 0.09334619093539055,
+      "grad_norm": 0.4760449473372184,
+      "learning_rate": 9.846205048200926e-06,
+      "loss": 2.3317,
+      "step": 242
+    },
+    {
+      "epoch": 0.09373191899710703,
+      "grad_norm": 0.4438227065202477,
+      "learning_rate": 9.844686508907538e-06,
+      "loss": 2.3582,
+      "step": 243
+    },
+    {
+      "epoch": 0.09411764705882353,
+      "grad_norm": 0.43517075741433026,
+      "learning_rate": 9.84316062778912e-06,
+      "loss": 2.365,
+      "step": 244
+    },
+    {
+      "epoch": 0.09450337512054002,
+      "grad_norm": 0.5025767914527827,
+      "learning_rate": 9.841627407158048e-06,
+      "loss": 2.3178,
+      "step": 245
+    },
+    {
+      "epoch": 0.0948891031822565,
+      "grad_norm": 0.4864659521213612,
+      "learning_rate": 9.840086849337825e-06,
+      "loss": 2.3858,
+      "step": 246
+    },
+    {
+      "epoch": 0.095274831243973,
+      "grad_norm": 0.44385232721166673,
+      "learning_rate": 9.838538956663073e-06,
+      "loss": 2.3473,
+      "step": 247
+    },
+    {
+      "epoch": 0.0956605593056895,
+      "grad_norm": 0.4353216339544687,
+      "learning_rate": 9.836983731479526e-06,
+      "loss": 2.3667,
+      "step": 248
+    },
+    {
+      "epoch": 0.09604628736740597,
+      "grad_norm": 0.4719586700058661,
+      "learning_rate": 9.835421176144035e-06,
+      "loss": 2.3507,
+      "step": 249
+    },
+    {
+      "epoch": 0.09643201542912247,
+      "grad_norm": 0.4721523839328503,
+      "learning_rate": 9.833851293024555e-06,
+      "loss": 2.4036,
+      "step": 250
+    },
+    {
+      "epoch": 0.09681774349083896,
+      "grad_norm": 0.43112104586716343,
+      "learning_rate": 9.832274084500147e-06,
+      "loss": 2.3394,
+      "step": 251
+    },
+    {
+      "epoch": 0.09720347155255545,
+      "grad_norm": 0.4645093917726468,
+      "learning_rate": 9.830689552960974e-06,
+      "loss": 2.3434,
+      "step": 252
+    },
+    {
+      "epoch": 0.09758919961427194,
+      "grad_norm": 0.4464501744666932,
+      "learning_rate": 9.829097700808298e-06,
+      "loss": 2.268,
+      "step": 253
+    },
+    {
+      "epoch": 0.09797492767598842,
+      "grad_norm": 0.4412531990241394,
+      "learning_rate": 9.827498530454473e-06,
+      "loss": 2.3434,
+      "step": 254
+    },
+    {
+      "epoch": 0.09836065573770492,
+      "grad_norm": 0.48563901834289575,
+      "learning_rate": 9.825892044322942e-06,
+      "loss": 2.4006,
+      "step": 255
+    },
+    {
+      "epoch": 0.09874638379942141,
+      "grad_norm": 0.48283553769044785,
+      "learning_rate": 9.824278244848236e-06,
+      "loss": 2.4335,
+      "step": 256
+    },
+    {
+      "epoch": 0.09913211186113789,
+      "grad_norm": 0.43055299801280267,
+      "learning_rate": 9.82265713447597e-06,
+      "loss": 2.3679,
+      "step": 257
+    },
+    {
+      "epoch": 0.09951783992285439,
+      "grad_norm": 0.449769379434095,
+      "learning_rate": 9.821028715662838e-06,
+      "loss": 2.3132,
+      "step": 258
+    },
+    {
+      "epoch": 0.09990356798457088,
+      "grad_norm": 0.5217734914013734,
+      "learning_rate": 9.819392990876605e-06,
+      "loss": 2.3516,
+      "step": 259
+    },
+    {
+      "epoch": 0.10028929604628736,
+      "grad_norm": 0.4615898258764757,
+      "learning_rate": 9.817749962596115e-06,
+      "loss": 2.3833,
+      "step": 260
+    },
+    {
+      "epoch": 0.10067502410800386,
+      "grad_norm": 0.48582752923842243,
+      "learning_rate": 9.816099633311278e-06,
+      "loss": 2.3628,
+      "step": 261
+    },
+    {
+      "epoch": 0.10106075216972035,
+      "grad_norm": 0.4559771187990709,
+      "learning_rate": 9.814442005523062e-06,
+      "loss": 2.4158,
+      "step": 262
+    },
+    {
+      "epoch": 0.10144648023143683,
+      "grad_norm": 0.489757834124081,
+      "learning_rate": 9.812777081743505e-06,
+      "loss": 2.4279,
+      "step": 263
+    },
+    {
+      "epoch": 0.10183220829315333,
+      "grad_norm": 0.44309092173930786,
+      "learning_rate": 9.811104864495691e-06,
+      "loss": 2.3007,
+      "step": 264
+    },
+    {
+      "epoch": 0.10221793635486982,
+      "grad_norm": 0.47803882641272966,
+      "learning_rate": 9.809425356313769e-06,
+      "loss": 2.3382,
+      "step": 265
+    },
+    {
+      "epoch": 0.1026036644165863,
+      "grad_norm": 0.46596798590577343,
+      "learning_rate": 9.807738559742927e-06,
+      "loss": 2.3615,
+      "step": 266
+    },
+    {
+      "epoch": 0.1029893924783028,
+      "grad_norm": 0.4664501923411451,
+      "learning_rate": 9.806044477339403e-06,
+      "loss": 2.3507,
+      "step": 267
+    },
+    {
+      "epoch": 0.10337512054001928,
+      "grad_norm": 0.49023888589795483,
+      "learning_rate": 9.804343111670472e-06,
+      "loss": 2.3983,
+      "step": 268
+    },
+    {
+      "epoch": 0.10376084860173578,
+      "grad_norm": 0.45763647245985856,
+      "learning_rate": 9.802634465314454e-06,
+      "loss": 2.3629,
+      "step": 269
+    },
+    {
+      "epoch": 0.10414657666345227,
+      "grad_norm": 0.4723827861961394,
+      "learning_rate": 9.800918540860693e-06,
+      "loss": 2.3405,
+      "step": 270
+    },
+    {
+      "epoch": 0.10453230472516875,
+      "grad_norm": 0.44017857265774585,
+      "learning_rate": 9.799195340909569e-06,
+      "loss": 2.3689,
+      "step": 271
+    },
+    {
+      "epoch": 0.10491803278688525,
+      "grad_norm": 0.4656051971642151,
+      "learning_rate": 9.797464868072489e-06,
+      "loss": 2.3007,
+      "step": 272
+    },
+    {
+      "epoch": 0.10530376084860174,
+      "grad_norm": 0.47987588123571606,
+      "learning_rate": 9.795727124971872e-06,
+      "loss": 2.406,
+      "step": 273
+    },
+    {
+      "epoch": 0.10568948891031822,
+      "grad_norm": 0.5056657398354122,
+      "learning_rate": 9.793982114241165e-06,
+      "loss": 2.3804,
+      "step": 274
+    },
+    {
+      "epoch": 0.10607521697203472,
+      "grad_norm": 0.47845361621935306,
+      "learning_rate": 9.792229838524825e-06,
+      "loss": 2.343,
+      "step": 275
+    },
+    {
+      "epoch": 0.10646094503375121,
+      "grad_norm": 0.48207101891115595,
+      "learning_rate": 9.790470300478318e-06,
+      "loss": 2.3679,
+      "step": 276
+    },
+    {
+      "epoch": 0.10684667309546769,
+      "grad_norm": 0.44257531338996314,
+      "learning_rate": 9.788703502768115e-06,
+      "loss": 2.3744,
+      "step": 277
+    },
+    {
+      "epoch": 0.10723240115718419,
+      "grad_norm": 0.45301627449323617,
+      "learning_rate": 9.786929448071688e-06,
+      "loss": 2.3729,
+      "step": 278
+    },
+    {
+      "epoch": 0.10761812921890067,
+      "grad_norm": 0.4469254649273637,
+      "learning_rate": 9.785148139077511e-06,
+      "loss": 2.3602,
+      "step": 279
+    },
+    {
+      "epoch": 0.10800385728061716,
+      "grad_norm": 0.4484779739046592,
+      "learning_rate": 9.783359578485047e-06,
+      "loss": 2.4075,
+      "step": 280
+    },
+    {
+      "epoch": 0.10838958534233366,
+      "grad_norm": 0.4636230876300128,
+      "learning_rate": 9.78156376900475e-06,
+      "loss": 2.4094,
+      "step": 281
+    },
+    {
+      "epoch": 0.10877531340405014,
+      "grad_norm": 0.4746029120882621,
+      "learning_rate": 9.77976071335806e-06,
+      "loss": 2.3912,
+      "step": 282
+    },
+    {
+      "epoch": 0.10916104146576663,
+      "grad_norm": 0.4644868421831817,
+      "learning_rate": 9.777950414277394e-06,
+      "loss": 2.3936,
+      "step": 283
+    },
+    {
+      "epoch": 0.10954676952748313,
+      "grad_norm": 0.5069332456975906,
+      "learning_rate": 9.776132874506153e-06,
+      "loss": 2.3874,
+      "step": 284
+    },
+    {
+      "epoch": 0.10993249758919961,
+      "grad_norm": 0.448790481721612,
+      "learning_rate": 9.774308096798704e-06,
+      "loss": 2.4064,
+      "step": 285
+    },
+    {
+      "epoch": 0.1103182256509161,
+      "grad_norm": 0.5221704977963175,
+      "learning_rate": 9.772476083920388e-06,
+      "loss": 2.3179,
+      "step": 286
+    },
+    {
+      "epoch": 0.1107039537126326,
+      "grad_norm": 0.45028528688105157,
+      "learning_rate": 9.770636838647505e-06,
+      "loss": 2.333,
+      "step": 287
+    },
+    {
+      "epoch": 0.11108968177434908,
+      "grad_norm": 0.46213043493703676,
+      "learning_rate": 9.768790363767321e-06,
+      "loss": 2.3342,
+      "step": 288
+    },
+    {
+      "epoch": 0.11147540983606558,
+      "grad_norm": 0.48597786762721035,
+      "learning_rate": 9.766936662078056e-06,
+      "loss": 2.3701,
+      "step": 289
+    },
+    {
+      "epoch": 0.11186113789778207,
+      "grad_norm": 0.47811060819847545,
+      "learning_rate": 9.76507573638888e-06,
+      "loss": 2.407,
+      "step": 290
+    },
+    {
+      "epoch": 0.11224686595949855,
+      "grad_norm": 0.47693681660455023,
+      "learning_rate": 9.763207589519909e-06,
+      "loss": 2.4047,
+      "step": 291
+    },
+    {
+      "epoch": 0.11263259402121505,
+      "grad_norm": 0.44714431772174534,
+      "learning_rate": 9.761332224302209e-06,
+      "loss": 2.3458,
+      "step": 292
+    },
+    {
+      "epoch": 0.11301832208293153,
+      "grad_norm": 0.4706222187538263,
+      "learning_rate": 9.759449643577779e-06,
+      "loss": 2.3536,
+      "step": 293
+    },
+    {
+      "epoch": 0.11340405014464802,
+      "grad_norm": 0.4822820286599736,
+      "learning_rate": 9.757559850199554e-06,
+      "loss": 2.3624,
+      "step": 294
+    },
+    {
+      "epoch": 0.11378977820636452,
+      "grad_norm": 0.4527420181209796,
+      "learning_rate": 9.755662847031402e-06,
+      "loss": 2.3276,
+      "step": 295
+    },
+    {
+      "epoch": 0.114175506268081,
+      "grad_norm": 0.5009015927230964,
+      "learning_rate": 9.753758636948112e-06,
+      "loss": 2.3411,
+      "step": 296
+    },
+    {
+      "epoch": 0.11456123432979749,
+      "grad_norm": 0.45340300327958766,
+      "learning_rate": 9.751847222835399e-06,
+      "loss": 2.3366,
+      "step": 297
+    },
+    {
+      "epoch": 0.11494696239151399,
+      "grad_norm": 0.4521806041401524,
+      "learning_rate": 9.749928607589894e-06,
+      "loss": 2.3661,
+      "step": 298
+    },
+    {
+      "epoch": 0.11533269045323047,
+      "grad_norm": 0.43386796678166034,
+      "learning_rate": 9.74800279411914e-06,
+      "loss": 2.4343,
+      "step": 299
+    },
+    {
+      "epoch": 0.11571841851494696,
+      "grad_norm": 0.5073054891521527,
+      "learning_rate": 9.74606978534159e-06,
+      "loss": 2.3711,
+      "step": 300
+    },
+    {
+      "epoch": 0.11610414657666346,
+      "grad_norm": 0.48069295086138375,
+      "learning_rate": 9.744129584186599e-06,
+      "loss": 2.3452,
+      "step": 301
+    },
+    {
+      "epoch": 0.11648987463837994,
+      "grad_norm": 0.4983738490461878,
+      "learning_rate": 9.742182193594424e-06,
+      "loss": 2.3582,
+      "step": 302
+    },
+    {
+      "epoch": 0.11687560270009643,
+      "grad_norm": 0.5346958331811935,
+      "learning_rate": 9.740227616516215e-06,
+      "loss": 2.4356,
+      "step": 303
+    },
+    {
+      "epoch": 0.11726133076181292,
+      "grad_norm": 0.4880172814003539,
+      "learning_rate": 9.738265855914014e-06,
+      "loss": 2.3423,
+      "step": 304
+    },
+    {
+      "epoch": 0.11764705882352941,
+      "grad_norm": 0.4717538049526971,
+      "learning_rate": 9.736296914760747e-06,
+      "loss": 2.3923,
+      "step": 305
+    },
+    {
+      "epoch": 0.1180327868852459,
+      "grad_norm": 0.4971984961224734,
+      "learning_rate": 9.734320796040226e-06,
+      "loss": 2.4046,
+      "step": 306
+    },
+    {
+      "epoch": 0.11841851494696239,
+      "grad_norm": 0.511356699564154,
+      "learning_rate": 9.732337502747137e-06,
+      "loss": 2.3471,
+      "step": 307
+    },
+    {
+      "epoch": 0.11880424300867888,
+      "grad_norm": 0.5462753805792145,
+      "learning_rate": 9.730347037887041e-06,
+      "loss": 2.2914,
+      "step": 308
+    },
+    {
+      "epoch": 0.11918997107039538,
+      "grad_norm": 0.5658159223894998,
+      "learning_rate": 9.728349404476361e-06,
+      "loss": 2.3901,
+      "step": 309
+    },
+    {
+      "epoch": 0.11957569913211186,
+      "grad_norm": 0.48142167178073736,
+      "learning_rate": 9.726344605542388e-06,
+      "loss": 2.3021,
+      "step": 310
+    },
+    {
+      "epoch": 0.11996142719382835,
+      "grad_norm": 0.4797926831647081,
+      "learning_rate": 9.724332644123278e-06,
+      "loss": 2.371,
+      "step": 311
+    },
+    {
+      "epoch": 0.12034715525554485,
+      "grad_norm": 0.4786174999989658,
+      "learning_rate": 9.722313523268028e-06,
+      "loss": 2.3725,
+      "step": 312
+    },
+    {
+      "epoch": 0.12073288331726133,
+      "grad_norm": 0.4438762134466573,
+      "learning_rate": 9.720287246036495e-06,
+      "loss": 2.2856,
+      "step": 313
+    },
+    {
+      "epoch": 0.12111861137897782,
+      "grad_norm": 0.50049840606718,
+      "learning_rate": 9.718253815499379e-06,
+      "loss": 2.3902,
+      "step": 314
+    },
+    {
+      "epoch": 0.12150433944069432,
+      "grad_norm": 0.476365793733286,
+      "learning_rate": 9.716213234738216e-06,
+      "loss": 2.3777,
+      "step": 315
+    },
+    {
+      "epoch": 0.1218900675024108,
+      "grad_norm": 0.4488991949236611,
+      "learning_rate": 9.714165506845381e-06,
+      "loss": 2.3435,
+      "step": 316
+    },
+    {
+      "epoch": 0.1222757955641273,
+      "grad_norm": 0.4587270176274335,
+      "learning_rate": 9.712110634924083e-06,
+      "loss": 2.3591,
+      "step": 317
+    },
+    {
+      "epoch": 0.12266152362584377,
+      "grad_norm": 0.4538941606082488,
+      "learning_rate": 9.710048622088354e-06,
+      "loss": 2.237,
+      "step": 318
+    },
+    {
+      "epoch": 0.12304725168756027,
+      "grad_norm": 0.4704173813002971,
+      "learning_rate": 9.707979471463045e-06,
+      "loss": 2.3786,
+      "step": 319
+    },
+    {
+      "epoch": 0.12343297974927676,
+      "grad_norm": 0.5109877934613372,
+      "learning_rate": 9.705903186183828e-06,
+      "loss": 2.3326,
+      "step": 320
+    },
+    {
+      "epoch": 0.12381870781099324,
+      "grad_norm": 0.46072430320991625,
+      "learning_rate": 9.703819769397187e-06,
+      "loss": 2.4276,
+      "step": 321
+    },
+    {
+      "epoch": 0.12420443587270974,
+      "grad_norm": 0.5050060879426643,
+      "learning_rate": 9.70172922426041e-06,
+      "loss": 2.3332,
+      "step": 322
+    },
+    {
+      "epoch": 0.12459016393442623,
+      "grad_norm": 0.4821117791422475,
+      "learning_rate": 9.699631553941591e-06,
+      "loss": 2.364,
+      "step": 323
+    },
+    {
+      "epoch": 0.12497589199614272,
+      "grad_norm": 0.46553118925707043,
+      "learning_rate": 9.697526761619621e-06,
+      "loss": 2.3143,
+      "step": 324
+    },
+    {
+      "epoch": 0.1253616200578592,
+      "grad_norm": 0.5019839429179425,
+      "learning_rate": 9.695414850484187e-06,
+      "loss": 2.3849,
+      "step": 325
+    },
+    {
+      "epoch": 0.1257473481195757,
+      "grad_norm": 0.4665907597446247,
+      "learning_rate": 9.693295823735754e-06,
+      "loss": 2.3864,
+      "step": 326
+    },
+    {
+      "epoch": 0.1261330761812922,
+      "grad_norm": 0.47054309847305004,
+      "learning_rate": 9.69116968458558e-06,
+      "loss": 2.4386,
+      "step": 327
+    },
+    {
+      "epoch": 0.12651880424300868,
+      "grad_norm": 0.5694549688899382,
+      "learning_rate": 9.689036436255698e-06,
+      "loss": 2.3638,
+      "step": 328
+    },
+    {
+      "epoch": 0.12690453230472518,
+      "grad_norm": 0.4993630278897394,
+      "learning_rate": 9.686896081978916e-06,
+      "loss": 2.3018,
+      "step": 329
+    },
+    {
+      "epoch": 0.12729026036644167,
+      "grad_norm": 0.46646727753640355,
+      "learning_rate": 9.68474862499881e-06,
+      "loss": 2.3632,
+      "step": 330
+    },
+    {
+      "epoch": 0.12767598842815814,
+      "grad_norm": 0.5514409158628053,
+      "learning_rate": 9.682594068569717e-06,
+      "loss": 2.3586,
+      "step": 331
+    },
+    {
+      "epoch": 0.12806171648987463,
+      "grad_norm": 0.5365834653256637,
+      "learning_rate": 9.680432415956736e-06,
+      "loss": 2.283,
+      "step": 332
+    },
+    {
+      "epoch": 0.12844744455159113,
+      "grad_norm": 0.46818705457501164,
+      "learning_rate": 9.67826367043572e-06,
+      "loss": 2.3932,
+      "step": 333
+    },
+    {
+      "epoch": 0.12883317261330762,
+      "grad_norm": 0.4541066736608887,
+      "learning_rate": 9.676087835293267e-06,
+      "loss": 2.3231,
+      "step": 334
+    },
+    {
+      "epoch": 0.12921890067502412,
+      "grad_norm": 0.5282386505767401,
+      "learning_rate": 9.673904913826723e-06,
+      "loss": 2.2421,
+      "step": 335
+    },
+    {
+      "epoch": 0.12960462873674058,
+      "grad_norm": 0.468829947274691,
+      "learning_rate": 9.671714909344175e-06,
+      "loss": 2.3366,
+      "step": 336
+    },
+    {
+      "epoch": 0.12999035679845708,
+      "grad_norm": 0.44396215189823446,
+      "learning_rate": 9.669517825164435e-06,
+      "loss": 2.4577,
+      "step": 337
+    },
+    {
+      "epoch": 0.13037608486017357,
+      "grad_norm": 0.45822960137335506,
+      "learning_rate": 9.66731366461705e-06,
+      "loss": 2.3348,
+      "step": 338
+    },
+    {
+      "epoch": 0.13076181292189007,
+      "grad_norm": 0.49070250854985425,
+      "learning_rate": 9.665102431042294e-06,
+      "loss": 2.3269,
+      "step": 339
+    },
+    {
+      "epoch": 0.13114754098360656,
+      "grad_norm": 0.4750510730638908,
+      "learning_rate": 9.66288412779115e-06,
+      "loss": 2.3423,
+      "step": 340
+    },
+    {
+      "epoch": 0.13153326904532306,
+      "grad_norm": 0.4491421521169457,
+      "learning_rate": 9.660658758225328e-06,
+      "loss": 2.3735,
+      "step": 341
+    },
+    {
+      "epoch": 0.13191899710703953,
+      "grad_norm": 0.48144278294663,
+      "learning_rate": 9.658426325717231e-06,
+      "loss": 2.3237,
+      "step": 342
+    },
+    {
+      "epoch": 0.13230472516875602,
+      "grad_norm": 0.4458502716413071,
+      "learning_rate": 9.656186833649978e-06,
+      "loss": 2.3359,
+      "step": 343
+    },
+    {
+      "epoch": 0.13269045323047252,
+      "grad_norm": 0.47471916997622327,
+      "learning_rate": 9.653940285417381e-06,
+      "loss": 2.2852,
+      "step": 344
+    },
+    {
+      "epoch": 0.133076181292189,
+      "grad_norm": 0.44132656234673406,
+      "learning_rate": 9.651686684423946e-06,
+      "loss": 2.4675,
+      "step": 345
+    },
+    {
+      "epoch": 0.1334619093539055,
+      "grad_norm": 0.448204840397715,
+      "learning_rate": 9.649426034084866e-06,
+      "loss": 2.4185,
+      "step": 346
+    },
+    {
+      "epoch": 0.13384763741562197,
+      "grad_norm": 0.4783505363815993,
+      "learning_rate": 9.64715833782602e-06,
+      "loss": 2.3576,
+      "step": 347
+    },
+    {
+      "epoch": 0.13423336547733847,
+      "grad_norm": 0.5392006855126578,
+      "learning_rate": 9.644883599083959e-06,
+      "loss": 2.3052,
+      "step": 348
+    },
+    {
+      "epoch": 0.13461909353905496,
+      "grad_norm": 0.5398502095090069,
+      "learning_rate": 9.642601821305911e-06,
+      "loss": 2.3771,
+      "step": 349
+    },
+    {
+      "epoch": 0.13500482160077146,
+      "grad_norm": 0.4652744324669575,
+      "learning_rate": 9.640313007949774e-06,
+      "loss": 2.3416,
+      "step": 350
+    },
+    {
+      "epoch": 0.13539054966248795,
+      "grad_norm": 0.4874516740981523,
+      "learning_rate": 9.638017162484099e-06,
+      "loss": 2.3085,
+      "step": 351
+    },
+    {
+      "epoch": 0.13577627772420445,
+      "grad_norm": 0.5146225636471995,
+      "learning_rate": 9.635714288388103e-06,
+      "loss": 2.3684,
+      "step": 352
+    },
+    {
+      "epoch": 0.13616200578592091,
+      "grad_norm": 0.5324693443633871,
+      "learning_rate": 9.633404389151647e-06,
+      "loss": 2.3293,
+      "step": 353
+    },
+    {
+      "epoch": 0.1365477338476374,
+      "grad_norm": 0.5005897826018193,
+      "learning_rate": 9.631087468275242e-06,
+      "loss": 2.3957,
+      "step": 354
+    },
+    {
+      "epoch": 0.1369334619093539,
+      "grad_norm": 0.4794002915682946,
+      "learning_rate": 9.628763529270042e-06,
+      "loss": 2.3003,
+      "step": 355
+    },
+    {
+      "epoch": 0.1373191899710704,
+      "grad_norm": 0.4519655989539701,
+      "learning_rate": 9.626432575657834e-06,
+      "loss": 2.3054,
+      "step": 356
+    },
+    {
+      "epoch": 0.1377049180327869,
+      "grad_norm": 0.45990015572061077,
+      "learning_rate": 9.624094610971031e-06,
+      "loss": 2.3092,
+      "step": 357
+    },
+    {
+      "epoch": 0.1380906460945034,
+      "grad_norm": 0.4575290225254486,
+      "learning_rate": 9.621749638752677e-06,
+      "loss": 2.3802,
+      "step": 358
+    },
+    {
+      "epoch": 0.13847637415621986,
+      "grad_norm": 0.48239530007607073,
+      "learning_rate": 9.619397662556434e-06,
+      "loss": 2.4129,
+      "step": 359
+    },
+    {
+      "epoch": 0.13886210221793635,
+      "grad_norm": 0.4722242819640516,
+      "learning_rate": 9.617038685946578e-06,
+      "loss": 2.3015,
+      "step": 360
+    },
+    {
+      "epoch": 0.13924783027965285,
+      "grad_norm": 0.4717278397903987,
+      "learning_rate": 9.614672712497994e-06,
+      "loss": 2.3601,
+      "step": 361
+    },
+    {
+      "epoch": 0.13963355834136934,
+      "grad_norm": 0.4629947753534378,
+      "learning_rate": 9.612299745796166e-06,
+      "loss": 2.3506,
+      "step": 362
+    },
+    {
+      "epoch": 0.14001928640308584,
+      "grad_norm": 0.44161467357227746,
+      "learning_rate": 9.609919789437181e-06,
+      "loss": 2.3704,
+      "step": 363
+    },
+    {
+      "epoch": 0.1404050144648023,
+      "grad_norm": 0.48893954585393296,
+      "learning_rate": 9.60753284702772e-06,
+      "loss": 2.343,
+      "step": 364
+    },
+    {
+      "epoch": 0.1407907425265188,
+      "grad_norm": 0.48320802138822283,
+      "learning_rate": 9.605138922185044e-06,
+      "loss": 2.2883,
+      "step": 365
+    },
+    {
+      "epoch": 0.1411764705882353,
+      "grad_norm": 0.46256366347211225,
+      "learning_rate": 9.602738018536999e-06,
+      "loss": 2.3546,
+      "step": 366
+    },
+    {
+      "epoch": 0.1415621986499518,
+      "grad_norm": 0.47369112795205537,
+      "learning_rate": 9.600330139722009e-06,
+      "loss": 2.338,
+      "step": 367
+    },
+    {
+      "epoch": 0.14194792671166828,
+      "grad_norm": 0.44509196448459337,
+      "learning_rate": 9.597915289389067e-06,
+      "loss": 2.3689,
+      "step": 368
+    },
+    {
+      "epoch": 0.14233365477338478,
+      "grad_norm": 0.5206361761868962,
+      "learning_rate": 9.595493471197728e-06,
+      "loss": 2.3584,
+      "step": 369
+    },
+    {
+      "epoch": 0.14271938283510124,
+      "grad_norm": 0.4494183534270254,
+      "learning_rate": 9.59306468881811e-06,
+      "loss": 2.3922,
+      "step": 370
+    },
+    {
+      "epoch": 0.14310511089681774,
+      "grad_norm": 0.46343082392561685,
+      "learning_rate": 9.590628945930884e-06,
+      "loss": 2.3408,
+      "step": 371
+    },
+    {
+      "epoch": 0.14349083895853423,
+      "grad_norm": 0.5078625730082782,
+      "learning_rate": 9.58818624622727e-06,
+      "loss": 2.2884,
+      "step": 372
+    },
+    {
+      "epoch": 0.14387656702025073,
+      "grad_norm": 0.4584681472733906,
+      "learning_rate": 9.585736593409025e-06,
+      "loss": 2.3047,
+      "step": 373
+    },
+    {
+      "epoch": 0.14426229508196722,
+      "grad_norm": 0.4505415778870253,
+      "learning_rate": 9.583279991188452e-06,
+      "loss": 2.362,
+      "step": 374
+    },
+    {
+      "epoch": 0.1446480231436837,
+      "grad_norm": 0.4473492838799735,
+      "learning_rate": 9.58081644328838e-06,
+      "loss": 2.4563,
+      "step": 375
+    },
+    {
+      "epoch": 0.14503375120540019,
+      "grad_norm": 0.4564603087597555,
+      "learning_rate": 9.578345953442163e-06,
+      "loss": 2.2288,
+      "step": 376
+    },
+    {
+      "epoch": 0.14541947926711668,
+      "grad_norm": 0.43177932208894787,
+      "learning_rate": 9.575868525393678e-06,
+      "loss": 2.2975,
+      "step": 377
+    },
+    {
+      "epoch": 0.14580520732883318,
+      "grad_norm": 0.46636470820623344,
+      "learning_rate": 9.573384162897316e-06,
+      "loss": 2.4027,
+      "step": 378
+    },
+    {
+      "epoch": 0.14619093539054967,
+      "grad_norm": 0.4782220814127971,
+      "learning_rate": 9.570892869717973e-06,
+      "loss": 2.3228,
+      "step": 379
+    },
+    {
+      "epoch": 0.14657666345226616,
+      "grad_norm": 0.47904333240288893,
+      "learning_rate": 9.568394649631055e-06,
+      "loss": 2.345,
+      "step": 380
+    },
+    {
+      "epoch": 0.14696239151398263,
+      "grad_norm": 0.5044958437706698,
+      "learning_rate": 9.565889506422457e-06,
+      "loss": 2.3638,
+      "step": 381
+    },
+    {
+      "epoch": 0.14734811957569913,
+      "grad_norm": 0.49190167391071266,
+      "learning_rate": 9.56337744388857e-06,
+      "loss": 2.3272,
+      "step": 382
+    },
+    {
+      "epoch": 0.14773384763741562,
+      "grad_norm": 0.44996933951384155,
+      "learning_rate": 9.560858465836276e-06,
+      "loss": 2.3653,
+      "step": 383
+    },
+    {
+      "epoch": 0.14811957569913212,
+      "grad_norm": 0.4587857596998676,
+      "learning_rate": 9.558332576082925e-06,
+      "loss": 2.3414,
+      "step": 384
+    },
+    {
+      "epoch": 0.1485053037608486,
+      "grad_norm": 0.5012824701110264,
+      "learning_rate": 9.555799778456352e-06,
+      "loss": 2.3932,
+      "step": 385
+    },
+    {
+      "epoch": 0.14889103182256508,
+      "grad_norm": 0.5113006222659796,
+      "learning_rate": 9.553260076794854e-06,
+      "loss": 2.3353,
+      "step": 386
+    },
+    {
+      "epoch": 0.14927675988428157,
+      "grad_norm": 0.48958486992708833,
+      "learning_rate": 9.550713474947195e-06,
+      "loss": 2.3772,
+      "step": 387
+    },
+    {
+      "epoch": 0.14966248794599807,
+      "grad_norm": 0.43811850438162614,
+      "learning_rate": 9.548159976772593e-06,
+      "loss": 2.3487,
+      "step": 388
+    },
+    {
+      "epoch": 0.15004821600771456,
+      "grad_norm": 0.46168083383138425,
+      "learning_rate": 9.545599586140717e-06,
+      "loss": 2.3849,
+      "step": 389
+    },
+    {
+      "epoch": 0.15043394406943106,
+      "grad_norm": 0.49256260847551697,
+      "learning_rate": 9.543032306931683e-06,
+      "loss": 2.3543,
+      "step": 390
+    },
+    {
+      "epoch": 0.15081967213114755,
+      "grad_norm": 0.473167833158212,
+      "learning_rate": 9.540458143036043e-06,
+      "loss": 2.3816,
+      "step": 391
+    },
+    {
+      "epoch": 0.15120540019286402,
+      "grad_norm": 0.4742238894783086,
+      "learning_rate": 9.537877098354787e-06,
+      "loss": 2.3936,
+      "step": 392
+    },
+    {
+      "epoch": 0.15159112825458051,
+      "grad_norm": 0.5024024944905685,
+      "learning_rate": 9.535289176799327e-06,
+      "loss": 2.3507,
+      "step": 393
+    },
+    {
+      "epoch": 0.151976856316297,
+      "grad_norm": 0.4856201150929417,
+      "learning_rate": 9.532694382291502e-06,
+      "loss": 2.2789,
+      "step": 394
+    },
+    {
+      "epoch": 0.1523625843780135,
+      "grad_norm": 0.472014061842457,
+      "learning_rate": 9.530092718763563e-06,
+      "loss": 2.3341,
+      "step": 395
+    },
+    {
+      "epoch": 0.15274831243973,
+      "grad_norm": 0.42326984525803946,
+      "learning_rate": 9.527484190158171e-06,
+      "loss": 2.3542,
+      "step": 396
+    },
+    {
+      "epoch": 0.15313404050144647,
+      "grad_norm": 0.49125522387471887,
+      "learning_rate": 9.52486880042839e-06,
+      "loss": 2.312,
+      "step": 397
+    },
+    {
+      "epoch": 0.15351976856316296,
+      "grad_norm": 0.4598836914194489,
+      "learning_rate": 9.522246553537684e-06,
+      "loss": 2.3529,
+      "step": 398
+    },
+    {
+      "epoch": 0.15390549662487946,
+      "grad_norm": 0.4521553323843294,
+      "learning_rate": 9.51961745345991e-06,
+      "loss": 2.2969,
+      "step": 399
+    },
+    {
+      "epoch": 0.15429122468659595,
+      "grad_norm": 0.4796362607034985,
+      "learning_rate": 9.5169815041793e-06,
+      "loss": 2.2865,
+      "step": 400
+    },
+    {
+      "epoch": 0.15467695274831245,
+      "grad_norm": 0.45184163779831893,
+      "learning_rate": 9.514338709690479e-06,
+      "loss": 2.3817,
+      "step": 401
+    },
+    {
+      "epoch": 0.15506268081002894,
+      "grad_norm": 0.45974094003422233,
+      "learning_rate": 9.51168907399844e-06,
+      "loss": 2.3112,
+      "step": 402
+    },
+    {
+      "epoch": 0.1554484088717454,
+      "grad_norm": 0.4892632061678719,
+      "learning_rate": 9.509032601118541e-06,
+      "loss": 2.3575,
+      "step": 403
+    },
+    {
+      "epoch": 0.1558341369334619,
+      "grad_norm": 0.48317202108320073,
+      "learning_rate": 9.506369295076505e-06,
+      "loss": 2.3653,
+      "step": 404
+    },
+    {
+      "epoch": 0.1562198649951784,
+      "grad_norm": 0.460662253665198,
+      "learning_rate": 9.50369915990841e-06,
+      "loss": 2.3577,
+      "step": 405
+    },
+    {
+      "epoch": 0.1566055930568949,
+      "grad_norm": 0.47046930881397986,
+      "learning_rate": 9.50102219966068e-06,
+      "loss": 2.2641,
+      "step": 406
+    },
+    {
+      "epoch": 0.1569913211186114,
+      "grad_norm": 0.4667940344677699,
+      "learning_rate": 9.498338418390084e-06,
+      "loss": 2.3379,
+      "step": 407
+    },
+    {
+      "epoch": 0.15737704918032788,
+      "grad_norm": 0.4517494208806375,
+      "learning_rate": 9.495647820163725e-06,
+      "loss": 2.4353,
+      "step": 408
+    },
+    {
+      "epoch": 0.15776277724204435,
+      "grad_norm": 0.5185786383663151,
+      "learning_rate": 9.492950409059046e-06,
+      "loss": 2.3267,
+      "step": 409
+    },
+    {
+      "epoch": 0.15814850530376084,
+      "grad_norm": 0.4580602346169873,
+      "learning_rate": 9.490246189163804e-06,
+      "loss": 2.347,
+      "step": 410
+    },
+    {
+      "epoch": 0.15853423336547734,
+      "grad_norm": 0.46822479368086606,
+      "learning_rate": 9.487535164576078e-06,
+      "loss": 2.3668,
+      "step": 411
+    },
+    {
+      "epoch": 0.15891996142719383,
+      "grad_norm": 0.4400590003473538,
+      "learning_rate": 9.484817339404261e-06,
+      "loss": 2.3596,
+      "step": 412
+    },
+    {
+      "epoch": 0.15930568948891033,
+      "grad_norm": 0.5087509111696566,
+      "learning_rate": 9.482092717767051e-06,
+      "loss": 2.3396,
+      "step": 413
+    },
+    {
+      "epoch": 0.1596914175506268,
+      "grad_norm": 0.4679149494631726,
+      "learning_rate": 9.479361303793441e-06,
+      "loss": 2.317,
+      "step": 414
+    },
+    {
+      "epoch": 0.1600771456123433,
+      "grad_norm": 0.5143950303506667,
+      "learning_rate": 9.476623101622723e-06,
+      "loss": 2.3757,
+      "step": 415
+    },
+    {
+      "epoch": 0.16046287367405979,
+      "grad_norm": 0.4899194637705207,
+      "learning_rate": 9.473878115404477e-06,
+      "loss": 2.3438,
+      "step": 416
+    },
+    {
+      "epoch": 0.16084860173577628,
+      "grad_norm": 0.4324465415290763,
+      "learning_rate": 9.471126349298557e-06,
+      "loss": 2.3271,
+      "step": 417
+    },
+    {
+      "epoch": 0.16123432979749278,
+      "grad_norm": 0.4959757578143956,
+      "learning_rate": 9.468367807475098e-06,
+      "loss": 2.373,
+      "step": 418
+    },
+    {
+      "epoch": 0.16162005785920927,
+      "grad_norm": 0.45871196145856946,
+      "learning_rate": 9.465602494114501e-06,
+      "loss": 2.3363,
+      "step": 419
+    },
+    {
+      "epoch": 0.16200578592092574,
+      "grad_norm": 0.44081619492718654,
+      "learning_rate": 9.462830413407427e-06,
+      "loss": 2.388,
+      "step": 420
+    },
+    {
+      "epoch": 0.16239151398264223,
+      "grad_norm": 0.4834495637227687,
+      "learning_rate": 9.460051569554797e-06,
+      "loss": 2.3146,
+      "step": 421
+    },
+    {
+      "epoch": 0.16277724204435873,
+      "grad_norm": 0.4472145307722556,
+      "learning_rate": 9.457265966767774e-06,
+      "loss": 2.34,
+      "step": 422
+    },
+    {
+      "epoch": 0.16316297010607522,
+      "grad_norm": 0.45602302896948904,
+      "learning_rate": 9.454473609267774e-06,
+      "loss": 2.4182,
+      "step": 423
+    },
+    {
+      "epoch": 0.16354869816779172,
+      "grad_norm": 0.4630513057117173,
+      "learning_rate": 9.451674501286436e-06,
+      "loss": 2.3345,
+      "step": 424
+    },
+    {
+      "epoch": 0.16393442622950818,
+      "grad_norm": 0.5332933685367667,
+      "learning_rate": 9.448868647065644e-06,
+      "loss": 2.3917,
+      "step": 425
+    },
+    {
+      "epoch": 0.16432015429122468,
+      "grad_norm": 0.44731427171532107,
+      "learning_rate": 9.44605605085749e-06,
+      "loss": 2.3259,
+      "step": 426
+    },
+    {
+      "epoch": 0.16470588235294117,
+      "grad_norm": 0.45357761505856337,
+      "learning_rate": 9.443236716924297e-06,
+      "loss": 2.4065,
+      "step": 427
+    },
+    {
+      "epoch": 0.16509161041465767,
+      "grad_norm": 0.47800651336465766,
+      "learning_rate": 9.440410649538592e-06,
+      "loss": 2.2823,
+      "step": 428
+    },
+    {
+      "epoch": 0.16547733847637416,
+      "grad_norm": 0.5054964717603495,
+      "learning_rate": 9.437577852983103e-06,
+      "loss": 2.4017,
+      "step": 429
+    },
+    {
+      "epoch": 0.16586306653809066,
+      "grad_norm": 0.43849628461745427,
+      "learning_rate": 9.434738331550763e-06,
+      "loss": 2.3686,
+      "step": 430
+    },
+    {
+      "epoch": 0.16624879459980713,
+      "grad_norm": 0.4487380662190444,
+      "learning_rate": 9.43189208954469e-06,
+      "loss": 2.2904,
+      "step": 431
+    },
+    {
+      "epoch": 0.16663452266152362,
+      "grad_norm": 0.49527336714809694,
+      "learning_rate": 9.42903913127819e-06,
+      "loss": 2.3772,
+      "step": 432
+    },
+    {
+      "epoch": 0.16702025072324012,
+      "grad_norm": 0.43855103001961837,
+      "learning_rate": 9.426179461074745e-06,
+      "loss": 2.3655,
+      "step": 433
+    },
+    {
+      "epoch": 0.1674059787849566,
+      "grad_norm": 0.4441094939078043,
+      "learning_rate": 9.423313083268013e-06,
+      "loss": 2.3571,
+      "step": 434
+    },
+    {
+      "epoch": 0.1677917068466731,
+      "grad_norm": 0.45172643249146127,
+      "learning_rate": 9.42044000220181e-06,
+      "loss": 2.3865,
+      "step": 435
+    },
+    {
+      "epoch": 0.16817743490838957,
+      "grad_norm": 0.45725013139103915,
+      "learning_rate": 9.417560222230115e-06,
+      "loss": 2.3225,
+      "step": 436
+    },
+    {
+      "epoch": 0.16856316297010607,
+      "grad_norm": 0.4495590545821153,
+      "learning_rate": 9.41467374771706e-06,
+      "loss": 2.384,
+      "step": 437
+    },
+    {
+      "epoch": 0.16894889103182256,
+      "grad_norm": 0.4348734627267952,
+      "learning_rate": 9.411780583036915e-06,
+      "loss": 2.2426,
+      "step": 438
+    },
+    {
+      "epoch": 0.16933461909353906,
+      "grad_norm": 0.44935831762246375,
+      "learning_rate": 9.4088807325741e-06,
+      "loss": 2.3759,
+      "step": 439
+    },
+    {
+      "epoch": 0.16972034715525555,
+      "grad_norm": 0.46552305380180337,
+      "learning_rate": 9.405974200723156e-06,
+      "loss": 2.3293,
+      "step": 440
+    },
+    {
+      "epoch": 0.17010607521697205,
+      "grad_norm": 0.45321579183130734,
+      "learning_rate": 9.403060991888753e-06,
+      "loss": 2.42,
+      "step": 441
+    },
+    {
+      "epoch": 0.17049180327868851,
+      "grad_norm": 0.47121176667252546,
+      "learning_rate": 9.400141110485684e-06,
+      "loss": 2.3055,
+      "step": 442
+    },
+    {
+      "epoch": 0.170877531340405,
+      "grad_norm": 0.4628982145702491,
+      "learning_rate": 9.397214560938845e-06,
+      "loss": 2.3334,
+      "step": 443
+    },
+    {
+      "epoch": 0.1712632594021215,
+      "grad_norm": 0.4520120405725651,
+      "learning_rate": 9.394281347683247e-06,
+      "loss": 2.3796,
+      "step": 444
+    },
+    {
+      "epoch": 0.171648987463838,
+      "grad_norm": 0.4614216676717978,
+      "learning_rate": 9.391341475163992e-06,
+      "loss": 2.3848,
+      "step": 445
+    },
+    {
+      "epoch": 0.1720347155255545,
+      "grad_norm": 0.44564394784034217,
+      "learning_rate": 9.388394947836278e-06,
+      "loss": 2.3898,
+      "step": 446
+    },
+    {
+      "epoch": 0.17242044358727096,
+      "grad_norm": 0.47489566440126735,
+      "learning_rate": 9.385441770165385e-06,
+      "loss": 2.2923,
+      "step": 447
+    },
+    {
+      "epoch": 0.17280617164898746,
+      "grad_norm": 0.4749715130955814,
+      "learning_rate": 9.382481946626673e-06,
+      "loss": 2.3554,
+      "step": 448
+    },
+    {
+      "epoch": 0.17319189971070395,
+      "grad_norm": 0.4561615049418411,
+      "learning_rate": 9.379515481705572e-06,
+      "loss": 2.2604,
+      "step": 449
+    },
+    {
+      "epoch": 0.17357762777242045,
+      "grad_norm": 0.45020708135658016,
+      "learning_rate": 9.37654237989758e-06,
+      "loss": 2.3612,
+      "step": 450
+    },
+    {
+      "epoch": 0.17396335583413694,
+      "grad_norm": 0.4388756109677834,
+      "learning_rate": 9.373562645708244e-06,
+      "loss": 2.3052,
+      "step": 451
+    },
+    {
+      "epoch": 0.17434908389585344,
+      "grad_norm": 0.47173029790030807,
+      "learning_rate": 9.370576283653178e-06,
+      "loss": 2.3531,
+      "step": 452
+    },
+    {
+      "epoch": 0.1747348119575699,
+      "grad_norm": 0.4551478961521674,
+      "learning_rate": 9.367583298258022e-06,
+      "loss": 2.32,
+      "step": 453
+    },
+    {
+      "epoch": 0.1751205400192864,
+      "grad_norm": 0.5107186970646902,
+      "learning_rate": 9.364583694058467e-06,
+      "loss": 2.3383,
+      "step": 454
+    },
+    {
+      "epoch": 0.1755062680810029,
+      "grad_norm": 0.49295732309731005,
+      "learning_rate": 9.361577475600225e-06,
+      "loss": 2.3149,
+      "step": 455
+    },
+    {
+      "epoch": 0.1758919961427194,
+      "grad_norm": 0.45013631985174496,
+      "learning_rate": 9.358564647439037e-06,
+      "loss": 2.3459,
+      "step": 456
+    },
+    {
+      "epoch": 0.17627772420443588,
+      "grad_norm": 0.4400225469735529,
+      "learning_rate": 9.355545214140661e-06,
+      "loss": 2.2976,
+      "step": 457
+    },
+    {
+      "epoch": 0.17666345226615235,
+      "grad_norm": 0.44432846198913595,
+      "learning_rate": 9.352519180280862e-06,
+      "loss": 2.3229,
+      "step": 458
+    },
+    {
+      "epoch": 0.17704918032786884,
+      "grad_norm": 0.4568845745655981,
+      "learning_rate": 9.349486550445405e-06,
+      "loss": 2.3313,
+      "step": 459
+    },
+    {
+      "epoch": 0.17743490838958534,
+      "grad_norm": 0.4922752298047154,
+      "learning_rate": 9.34644732923006e-06,
+      "loss": 2.2953,
+      "step": 460
+    },
+    {
+      "epoch": 0.17782063645130183,
+      "grad_norm": 0.4478173878901112,
+      "learning_rate": 9.343401521240576e-06,
+      "loss": 2.3652,
+      "step": 461
+    },
+    {
+      "epoch": 0.17820636451301833,
+      "grad_norm": 0.4392171343396795,
+      "learning_rate": 9.34034913109269e-06,
+      "loss": 2.3811,
+      "step": 462
+    },
+    {
+      "epoch": 0.17859209257473482,
+      "grad_norm": 0.43781933893230074,
+      "learning_rate": 9.337290163412112e-06,
+      "loss": 2.3914,
+      "step": 463
+    },
+    {
+      "epoch": 0.1789778206364513,
+      "grad_norm": 0.46351158158141825,
+      "learning_rate": 9.33422462283452e-06,
+      "loss": 2.3988,
+      "step": 464
+    },
+    {
+      "epoch": 0.17936354869816779,
+      "grad_norm": 0.5066065947328378,
+      "learning_rate": 9.33115251400555e-06,
+      "loss": 2.3164,
+      "step": 465
+    },
+    {
+      "epoch": 0.17974927675988428,
+      "grad_norm": 0.4179985733238741,
+      "learning_rate": 9.328073841580797e-06,
+      "loss": 2.382,
+      "step": 466
+    },
+    {
+      "epoch": 0.18013500482160077,
+      "grad_norm": 0.4633632643057099,
+      "learning_rate": 9.3249886102258e-06,
+      "loss": 2.4012,
+      "step": 467
+    },
+    {
+      "epoch": 0.18052073288331727,
+      "grad_norm": 0.43953321607847123,
+      "learning_rate": 9.321896824616036e-06,
+      "loss": 2.3646,
+      "step": 468
+    },
+    {
+      "epoch": 0.18090646094503376,
+      "grad_norm": 0.47262124365709346,
+      "learning_rate": 9.318798489436917e-06,
+      "loss": 2.3455,
+      "step": 469
+    },
+    {
+      "epoch": 0.18129218900675023,
+      "grad_norm": 0.4734091829521179,
+      "learning_rate": 9.315693609383782e-06,
+      "loss": 2.4445,
+      "step": 470
+    },
+    {
+      "epoch": 0.18167791706846673,
+      "grad_norm": 0.5357820111129492,
+      "learning_rate": 9.312582189161882e-06,
+      "loss": 2.3196,
+      "step": 471
+    },
+    {
+      "epoch": 0.18206364513018322,
+      "grad_norm": 0.4531939491222502,
+      "learning_rate": 9.309464233486386e-06,
+      "loss": 2.3352,
+      "step": 472
+    },
+    {
+      "epoch": 0.18244937319189972,
+      "grad_norm": 0.46164565898099374,
+      "learning_rate": 9.306339747082364e-06,
+      "loss": 2.3742,
+      "step": 473
+    },
+    {
+      "epoch": 0.1828351012536162,
+      "grad_norm": 0.46841595728918883,
+      "learning_rate": 9.303208734684785e-06,
+      "loss": 2.3534,
+      "step": 474
+    },
+    {
+      "epoch": 0.18322082931533268,
+      "grad_norm": 0.47295947245811865,
+      "learning_rate": 9.300071201038503e-06,
+      "loss": 2.3432,
+      "step": 475
+    },
+    {
+      "epoch": 0.18360655737704917,
+      "grad_norm": 0.4343256543024035,
+      "learning_rate": 9.29692715089826e-06,
+      "loss": 2.3627,
+      "step": 476
+    },
+    {
+      "epoch": 0.18399228543876567,
+      "grad_norm": 0.4689889558698709,
+      "learning_rate": 9.29377658902867e-06,
+      "loss": 2.3754,
+      "step": 477
+    },
+    {
+      "epoch": 0.18437801350048216,
+      "grad_norm": 0.47896804482222666,
+      "learning_rate": 9.290619520204216e-06,
+      "loss": 2.3702,
+      "step": 478
+    },
+    {
+      "epoch": 0.18476374156219866,
+      "grad_norm": 0.4482334033726198,
+      "learning_rate": 9.287455949209243e-06,
+      "loss": 2.2704,
+      "step": 479
+    },
+    {
+      "epoch": 0.18514946962391515,
+      "grad_norm": 0.47595219738664213,
+      "learning_rate": 9.284285880837947e-06,
+      "loss": 2.3589,
+      "step": 480
+    },
+    {
+      "epoch": 0.18553519768563162,
+      "grad_norm": 0.5474764319097681,
+      "learning_rate": 9.281109319894374e-06,
+      "loss": 2.3885,
+      "step": 481
+    },
+    {
+      "epoch": 0.18592092574734811,
+      "grad_norm": 0.45726139926742926,
+      "learning_rate": 9.277926271192405e-06,
+      "loss": 2.3444,
+      "step": 482
+    },
+    {
+      "epoch": 0.1863066538090646,
+      "grad_norm": 0.45950159400239354,
+      "learning_rate": 9.274736739555757e-06,
+      "loss": 2.3134,
+      "step": 483
+    },
+    {
+      "epoch": 0.1866923818707811,
+      "grad_norm": 0.4510519833063498,
+      "learning_rate": 9.271540729817969e-06,
+      "loss": 2.3257,
+      "step": 484
+    },
+    {
+      "epoch": 0.1870781099324976,
+      "grad_norm": 0.47307586476649705,
+      "learning_rate": 9.268338246822395e-06,
+      "loss": 2.3307,
+      "step": 485
+    },
+    {
+      "epoch": 0.18746383799421407,
+      "grad_norm": 0.4863114816770395,
+      "learning_rate": 9.265129295422205e-06,
+      "loss": 2.3995,
+      "step": 486
+    },
+    {
+      "epoch": 0.18784956605593056,
+      "grad_norm": 0.4962829228709082,
+      "learning_rate": 9.261913880480367e-06,
+      "loss": 2.3765,
+      "step": 487
+    },
+    {
+      "epoch": 0.18823529411764706,
+      "grad_norm": 0.43479392629654684,
+      "learning_rate": 9.258692006869644e-06,
+      "loss": 2.3637,
+      "step": 488
+    },
+    {
+      "epoch": 0.18862102217936355,
+      "grad_norm": 0.460315396760295,
+      "learning_rate": 9.255463679472587e-06,
+      "loss": 2.3175,
+      "step": 489
+    },
+    {
+      "epoch": 0.18900675024108005,
+      "grad_norm": 0.45600398012066895,
+      "learning_rate": 9.252228903181529e-06,
+      "loss": 2.3246,
+      "step": 490
+    },
+    {
+      "epoch": 0.18939247830279654,
+      "grad_norm": 0.4473255725474842,
+      "learning_rate": 9.248987682898576e-06,
+      "loss": 2.3491,
+      "step": 491
+    },
+    {
+      "epoch": 0.189778206364513,
+      "grad_norm": 0.46269118307905355,
+      "learning_rate": 9.245740023535596e-06,
+      "loss": 2.2622,
+      "step": 492
+    },
+    {
+      "epoch": 0.1901639344262295,
+      "grad_norm": 0.4925059596550583,
+      "learning_rate": 9.24248593001422e-06,
+      "loss": 2.2895,
+      "step": 493
+    },
+    {
+      "epoch": 0.190549662487946,
+      "grad_norm": 0.4773842445251738,
+      "learning_rate": 9.239225407265824e-06,
+      "loss": 2.3874,
+      "step": 494
+    },
+    {
+      "epoch": 0.1909353905496625,
+      "grad_norm": 0.4662931552622624,
+      "learning_rate": 9.235958460231533e-06,
+      "loss": 2.298,
+      "step": 495
+    },
+    {
+      "epoch": 0.191321118611379,
+      "grad_norm": 0.4619381336604932,
+      "learning_rate": 9.232685093862206e-06,
+      "loss": 2.381,
+      "step": 496
+    },
+    {
+      "epoch": 0.19170684667309545,
+      "grad_norm": 0.5298956071333619,
+      "learning_rate": 9.229405313118423e-06,
+      "loss": 2.3675,
+      "step": 497
+    },
+    {
+      "epoch": 0.19209257473481195,
+      "grad_norm": 0.4743693788571214,
+      "learning_rate": 9.226119122970495e-06,
+      "loss": 2.3474,
+      "step": 498
+    },
+    {
+      "epoch": 0.19247830279652844,
+      "grad_norm": 0.5161243598264611,
+      "learning_rate": 9.22282652839844e-06,
+      "loss": 2.3182,
+      "step": 499
+    },
+    {
+      "epoch": 0.19286403085824494,
+      "grad_norm": 0.46825367215183233,
+      "learning_rate": 9.219527534391983e-06,
+      "loss": 2.3798,
+      "step": 500
+    },
+    {
+      "epoch": 0.19324975891996143,
+      "grad_norm": 0.49111756652037414,
+      "learning_rate": 9.216222145950548e-06,
+      "loss": 2.3427,
+      "step": 501
+    },
+    {
+      "epoch": 0.19363548698167793,
+      "grad_norm": 0.4372455755844022,
+      "learning_rate": 9.212910368083246e-06,
+      "loss": 2.352,
+      "step": 502
+    },
+    {
+      "epoch": 0.1940212150433944,
+      "grad_norm": 0.444514189504708,
+      "learning_rate": 9.209592205808874e-06,
+      "loss": 2.314,
+      "step": 503
+    },
+    {
+      "epoch": 0.1944069431051109,
+      "grad_norm": 0.4494673362799293,
+      "learning_rate": 9.206267664155906e-06,
+      "loss": 2.3621,
+      "step": 504
+    },
+    {
+      "epoch": 0.19479267116682739,
+      "grad_norm": 0.4854587757083868,
+      "learning_rate": 9.202936748162479e-06,
+      "loss": 2.3142,
+      "step": 505
+    },
+    {
+      "epoch": 0.19517839922854388,
+      "grad_norm": 0.42915591124069197,
+      "learning_rate": 9.19959946287639e-06,
+      "loss": 2.3916,
+      "step": 506
+    },
+    {
+      "epoch": 0.19556412729026038,
+      "grad_norm": 0.4490181754160567,
+      "learning_rate": 9.19625581335509e-06,
+      "loss": 2.3163,
+      "step": 507
+    },
+    {
+      "epoch": 0.19594985535197684,
+      "grad_norm": 0.4415330391865188,
+      "learning_rate": 9.192905804665677e-06,
+      "loss": 2.3438,
+      "step": 508
+    },
+    {
+      "epoch": 0.19633558341369334,
+      "grad_norm": 0.4649406985335145,
+      "learning_rate": 9.189549441884883e-06,
+      "loss": 2.3441,
+      "step": 509
+    },
+    {
+      "epoch": 0.19672131147540983,
+      "grad_norm": 0.4619548519344354,
+      "learning_rate": 9.18618673009907e-06,
+      "loss": 2.3707,
+      "step": 510
+    },
+    {
+      "epoch": 0.19710703953712633,
+      "grad_norm": 0.4378739732903529,
+      "learning_rate": 9.182817674404218e-06,
+      "loss": 2.2881,
+      "step": 511
+    },
+    {
+      "epoch": 0.19749276759884282,
+      "grad_norm": 0.43622883307093985,
+      "learning_rate": 9.179442279905927e-06,
+      "loss": 2.3662,
+      "step": 512
+    },
+    {
+      "epoch": 0.19787849566055932,
+      "grad_norm": 0.41856212113307684,
+      "learning_rate": 9.176060551719402e-06,
+      "loss": 2.4293,
+      "step": 513
+    },
+    {
+      "epoch": 0.19826422372227578,
+      "grad_norm": 0.45985586879947066,
+      "learning_rate": 9.17267249496944e-06,
+      "loss": 2.3028,
+      "step": 514
+    },
+    {
+      "epoch": 0.19864995178399228,
+      "grad_norm": 0.47187581156802855,
+      "learning_rate": 9.169278114790437e-06,
+      "loss": 2.3083,
+      "step": 515
+    },
+    {
+      "epoch": 0.19903567984570877,
+      "grad_norm": 0.46813467893089566,
+      "learning_rate": 9.165877416326365e-06,
+      "loss": 2.4425,
+      "step": 516
+    },
+    {
+      "epoch": 0.19942140790742527,
+      "grad_norm": 0.430401839388525,
+      "learning_rate": 9.162470404730776e-06,
+      "loss": 2.3625,
+      "step": 517
+    },
+    {
+      "epoch": 0.19980713596914176,
+      "grad_norm": 0.45099082208073454,
+      "learning_rate": 9.159057085166785e-06,
+      "loss": 2.3849,
+      "step": 518
+    },
+    {
+      "epoch": 0.20019286403085826,
+      "grad_norm": 0.45837731641693436,
+      "learning_rate": 9.15563746280707e-06,
+      "loss": 2.3482,
+      "step": 519
+    },
+    {
+      "epoch": 0.20057859209257473,
+      "grad_norm": 0.4453193647370109,
+      "learning_rate": 9.152211542833856e-06,
+      "loss": 2.3383,
+      "step": 520
+    },
+    {
+      "epoch": 0.20096432015429122,
+      "grad_norm": 0.5345378341413923,
+      "learning_rate": 9.148779330438919e-06,
+      "loss": 2.3119,
+      "step": 521
+    },
+    {
+      "epoch": 0.20135004821600772,
+      "grad_norm": 0.48502988720256984,
+      "learning_rate": 9.145340830823562e-06,
+      "loss": 2.2941,
+      "step": 522
+    },
+    {
+      "epoch": 0.2017357762777242,
+      "grad_norm": 0.45826176855332257,
+      "learning_rate": 9.141896049198622e-06,
+      "loss": 2.3359,
+      "step": 523
+    },
+    {
+      "epoch": 0.2021215043394407,
+      "grad_norm": 0.458602324657938,
+      "learning_rate": 9.138444990784455e-06,
+      "loss": 2.2403,
+      "step": 524
+    },
+    {
+      "epoch": 0.20250723240115717,
+      "grad_norm": 0.4370961113251191,
+      "learning_rate": 9.134987660810925e-06,
+      "loss": 2.3572,
+      "step": 525
+    },
+    {
+      "epoch": 0.20289296046287367,
+      "grad_norm": 0.4644845727347313,
+      "learning_rate": 9.131524064517405e-06,
+      "loss": 2.3709,
+      "step": 526
+    },
+    {
+      "epoch": 0.20327868852459016,
+      "grad_norm": 0.4649486219929489,
+      "learning_rate": 9.128054207152765e-06,
+      "loss": 2.3325,
+      "step": 527
+    },
+    {
+      "epoch": 0.20366441658630666,
+      "grad_norm": 0.4215258238013722,
+      "learning_rate": 9.124578093975358e-06,
+      "loss": 2.2895,
+      "step": 528
+    },
+    {
+      "epoch": 0.20405014464802315,
+      "grad_norm": 0.44795826208091805,
+      "learning_rate": 9.12109573025302e-06,
+      "loss": 2.422,
+      "step": 529
+    },
+    {
+      "epoch": 0.20443587270973965,
+      "grad_norm": 0.471764689731237,
+      "learning_rate": 9.117607121263063e-06,
+      "loss": 2.288,
+      "step": 530
+    },
+    {
+      "epoch": 0.2048216007714561,
+      "grad_norm": 0.45638319685158746,
+      "learning_rate": 9.114112272292255e-06,
+      "loss": 2.3147,
+      "step": 531
+    },
+    {
+      "epoch": 0.2052073288331726,
+      "grad_norm": 0.4922252882046052,
+      "learning_rate": 9.110611188636828e-06,
+      "loss": 2.2945,
+      "step": 532
+    },
+    {
+      "epoch": 0.2055930568948891,
+      "grad_norm": 0.4522653884951574,
+      "learning_rate": 9.107103875602458e-06,
+      "loss": 2.3377,
+      "step": 533
+    },
+    {
+      "epoch": 0.2059787849566056,
+      "grad_norm": 0.41728328170335205,
+      "learning_rate": 9.103590338504264e-06,
+      "loss": 2.3368,
+      "step": 534
+    },
+    {
+      "epoch": 0.2063645130183221,
+      "grad_norm": 0.46950092641132146,
+      "learning_rate": 9.100070582666796e-06,
+      "loss": 2.3087,
+      "step": 535
+    },
+    {
+      "epoch": 0.20675024108003856,
+      "grad_norm": 0.4641571541323398,
+      "learning_rate": 9.096544613424026e-06,
+      "loss": 2.3273,
+      "step": 536
+    },
+    {
+      "epoch": 0.20713596914175506,
+      "grad_norm": 0.449176581685245,
+      "learning_rate": 9.093012436119345e-06,
+      "loss": 2.2831,
+      "step": 537
+    },
+    {
+      "epoch": 0.20752169720347155,
+      "grad_norm": 0.44605553704842194,
+      "learning_rate": 9.089474056105552e-06,
+      "loss": 2.3976,
+      "step": 538
+    },
+    {
+      "epoch": 0.20790742526518805,
+      "grad_norm": 0.47316784820593766,
+      "learning_rate": 9.085929478744841e-06,
+      "loss": 2.3183,
+      "step": 539
+    },
+    {
+      "epoch": 0.20829315332690454,
+      "grad_norm": 0.46460066029125063,
+      "learning_rate": 9.082378709408805e-06,
+      "loss": 2.3535,
+      "step": 540
+    },
+    {
+      "epoch": 0.20867888138862103,
+      "grad_norm": 0.4811002373146888,
+      "learning_rate": 9.078821753478417e-06,
+      "loss": 2.3269,
+      "step": 541
+    },
+    {
+      "epoch": 0.2090646094503375,
+      "grad_norm": 0.49989065114093767,
+      "learning_rate": 9.075258616344025e-06,
+      "loss": 2.4009,
+      "step": 542
+    },
+    {
+      "epoch": 0.209450337512054,
+      "grad_norm": 0.4984731432556212,
+      "learning_rate": 9.071689303405343e-06,
+      "loss": 2.3227,
+      "step": 543
+    },
+    {
+      "epoch": 0.2098360655737705,
+      "grad_norm": 0.49437252214019445,
+      "learning_rate": 9.068113820071447e-06,
+      "loss": 2.3462,
+      "step": 544
+    },
+    {
+      "epoch": 0.210221793635487,
+      "grad_norm": 0.4515422763419102,
+      "learning_rate": 9.064532171760762e-06,
+      "loss": 2.4018,
+      "step": 545
+    },
+    {
+      "epoch": 0.21060752169720348,
+      "grad_norm": 0.4713555212878107,
+      "learning_rate": 9.060944363901057e-06,
+      "loss": 2.327,
+      "step": 546
+    },
+    {
+      "epoch": 0.21099324975891995,
+      "grad_norm": 0.4506689586670132,
+      "learning_rate": 9.057350401929433e-06,
+      "loss": 2.3214,
+      "step": 547
+    },
+    {
+      "epoch": 0.21137897782063644,
+      "grad_norm": 0.43907614616137236,
+      "learning_rate": 9.053750291292321e-06,
+      "loss": 2.323,
+      "step": 548
+    },
+    {
+      "epoch": 0.21176470588235294,
+      "grad_norm": 0.5266396975771747,
+      "learning_rate": 9.050144037445465e-06,
+      "loss": 2.2813,
+      "step": 549
+    },
+    {
+      "epoch": 0.21215043394406943,
+      "grad_norm": 0.4664813996758015,
+      "learning_rate": 9.046531645853924e-06,
+      "loss": 2.3815,
+      "step": 550
+    },
+    {
+      "epoch": 0.21253616200578593,
+      "grad_norm": 0.4201578229894165,
+      "learning_rate": 9.042913121992053e-06,
+      "loss": 2.3655,
+      "step": 551
+    },
+    {
+      "epoch": 0.21292189006750242,
+      "grad_norm": 0.4435148066995547,
+      "learning_rate": 9.039288471343505e-06,
+      "loss": 2.3591,
+      "step": 552
+    },
+    {
+      "epoch": 0.2133076181292189,
+      "grad_norm": 0.47388724861237486,
+      "learning_rate": 9.035657699401215e-06,
+      "loss": 2.3069,
+      "step": 553
+    },
+    {
+      "epoch": 0.21369334619093538,
+      "grad_norm": 0.4435532778834711,
+      "learning_rate": 9.032020811667395e-06,
+      "loss": 2.3913,
+      "step": 554
+    },
+    {
+      "epoch": 0.21407907425265188,
+      "grad_norm": 0.46191975386537937,
+      "learning_rate": 9.028377813653525e-06,
+      "loss": 2.3414,
+      "step": 555
+    },
+    {
+      "epoch": 0.21446480231436837,
+      "grad_norm": 0.4808909866476146,
+      "learning_rate": 9.024728710880345e-06,
+      "loss": 2.3831,
+      "step": 556
+    },
+    {
+      "epoch": 0.21485053037608487,
+      "grad_norm": 0.4586450639319388,
+      "learning_rate": 9.021073508877845e-06,
+      "loss": 2.3432,
+      "step": 557
+    },
+    {
+      "epoch": 0.21523625843780134,
+      "grad_norm": 0.42942847956441976,
+      "learning_rate": 9.017412213185261e-06,
+      "loss": 2.3106,
+      "step": 558
+    },
+    {
+      "epoch": 0.21562198649951783,
+      "grad_norm": 0.4130686749756181,
+      "learning_rate": 9.013744829351063e-06,
+      "loss": 2.3833,
+      "step": 559
+    },
+    {
+      "epoch": 0.21600771456123433,
+      "grad_norm": 0.45046846934747004,
+      "learning_rate": 9.010071362932945e-06,
+      "loss": 2.3834,
+      "step": 560
+    },
+    {
+      "epoch": 0.21639344262295082,
+      "grad_norm": 0.4973014254267896,
+      "learning_rate": 9.006391819497815e-06,
+      "loss": 2.3622,
+      "step": 561
+    },
+    {
+      "epoch": 0.21677917068466732,
+      "grad_norm": 0.448815959873419,
+      "learning_rate": 9.002706204621802e-06,
+      "loss": 2.2579,
+      "step": 562
+    },
+    {
+      "epoch": 0.2171648987463838,
+      "grad_norm": 0.48175635182232734,
+      "learning_rate": 8.999014523890228e-06,
+      "loss": 2.371,
+      "step": 563
+    },
+    {
+      "epoch": 0.21755062680810028,
+      "grad_norm": 0.44560435917971575,
+      "learning_rate": 8.995316782897605e-06,
+      "loss": 2.3395,
+      "step": 564
+    },
+    {
+      "epoch": 0.21793635486981677,
+      "grad_norm": 0.4655634371173414,
+      "learning_rate": 8.991612987247635e-06,
+      "loss": 2.3734,
+      "step": 565
+    },
+    {
+      "epoch": 0.21832208293153327,
+      "grad_norm": 0.40798212196098665,
+      "learning_rate": 8.987903142553194e-06,
+      "loss": 2.2733,
+      "step": 566
+    },
+    {
+      "epoch": 0.21870781099324976,
+      "grad_norm": 0.46181147562124997,
+      "learning_rate": 8.984187254436321e-06,
+      "loss": 2.357,
+      "step": 567
+    },
+    {
+      "epoch": 0.21909353905496626,
+      "grad_norm": 0.460719819698626,
+      "learning_rate": 8.98046532852822e-06,
+      "loss": 2.378,
+      "step": 568
+    },
+    {
+      "epoch": 0.21947926711668275,
+      "grad_norm": 0.4549844145982879,
+      "learning_rate": 8.976737370469237e-06,
+      "loss": 2.3645,
+      "step": 569
+    },
+    {
+      "epoch": 0.21986499517839922,
+      "grad_norm": 0.4472909074531031,
+      "learning_rate": 8.973003385908867e-06,
+      "loss": 2.343,
+      "step": 570
+    },
+    {
+      "epoch": 0.22025072324011571,
+      "grad_norm": 0.49221278008360814,
+      "learning_rate": 8.969263380505732e-06,
+      "loss": 2.3563,
+      "step": 571
+    },
+    {
+      "epoch": 0.2206364513018322,
+      "grad_norm": 0.4921528083374167,
+      "learning_rate": 8.965517359927583e-06,
+      "loss": 2.3788,
+      "step": 572
+    },
+    {
+      "epoch": 0.2210221793635487,
+      "grad_norm": 0.43757484012513737,
+      "learning_rate": 8.961765329851284e-06,
+      "loss": 2.2875,
+      "step": 573
+    },
+    {
+      "epoch": 0.2214079074252652,
+      "grad_norm": 0.4839031556215558,
+      "learning_rate": 8.958007295962802e-06,
+      "loss": 2.3777,
+      "step": 574
+    },
+    {
+      "epoch": 0.22179363548698167,
+      "grad_norm": 0.4596815016724665,
+      "learning_rate": 8.954243263957214e-06,
+      "loss": 2.2989,
+      "step": 575
+    },
+    {
+      "epoch": 0.22217936354869816,
+      "grad_norm": 0.43764815076903374,
+      "learning_rate": 8.950473239538672e-06,
+      "loss": 2.2998,
+      "step": 576
+    },
+    {
+      "epoch": 0.22256509161041466,
+      "grad_norm": 0.4579255766738418,
+      "learning_rate": 8.946697228420422e-06,
+      "loss": 2.3002,
+      "step": 577
+    },
+    {
+      "epoch": 0.22295081967213115,
+      "grad_norm": 0.4588483179120956,
+      "learning_rate": 8.942915236324775e-06,
+      "loss": 2.3287,
+      "step": 578
+    },
+    {
+      "epoch": 0.22333654773384765,
+      "grad_norm": 0.46612698165947597,
+      "learning_rate": 8.93912726898311e-06,
+      "loss": 2.3349,
+      "step": 579
+    },
+    {
+      "epoch": 0.22372227579556414,
+      "grad_norm": 0.457613146049226,
+      "learning_rate": 8.935333332135853e-06,
+      "loss": 2.3414,
+      "step": 580
+    },
+    {
+      "epoch": 0.2241080038572806,
+      "grad_norm": 0.44860534787088,
+      "learning_rate": 8.93153343153249e-06,
+      "loss": 2.3232,
+      "step": 581
+    },
+    {
+      "epoch": 0.2244937319189971,
+      "grad_norm": 0.4937219548706247,
+      "learning_rate": 8.927727572931532e-06,
+      "loss": 2.3366,
+      "step": 582
+    },
+    {
+      "epoch": 0.2248794599807136,
+      "grad_norm": 0.4932265294129161,
+      "learning_rate": 8.923915762100525e-06,
+      "loss": 2.3102,
+      "step": 583
+    },
+    {
+      "epoch": 0.2252651880424301,
+      "grad_norm": 0.44934181403039386,
+      "learning_rate": 8.920098004816035e-06,
+      "loss": 2.3293,
+      "step": 584
+    },
+    {
+      "epoch": 0.2256509161041466,
+      "grad_norm": 0.45126599523522487,
+      "learning_rate": 8.916274306863642e-06,
+      "loss": 2.3543,
+      "step": 585
+    },
+    {
+      "epoch": 0.22603664416586305,
+      "grad_norm": 0.5065377259132144,
+      "learning_rate": 8.91244467403792e-06,
+      "loss": 2.3546,
+      "step": 586
+    },
+    {
+      "epoch": 0.22642237222757955,
+      "grad_norm": 0.4778584518441631,
+      "learning_rate": 8.908609112142444e-06,
+      "loss": 2.3189,
+      "step": 587
+    },
+    {
+      "epoch": 0.22680810028929604,
+      "grad_norm": 0.4822383520476654,
+      "learning_rate": 8.904767626989774e-06,
+      "loss": 2.3384,
+      "step": 588
+    },
+    {
+      "epoch": 0.22719382835101254,
+      "grad_norm": 0.4856551848734807,
+      "learning_rate": 8.900920224401446e-06,
+      "loss": 2.3957,
+      "step": 589
+    },
+    {
+      "epoch": 0.22757955641272903,
+      "grad_norm": 0.4153806184547439,
+      "learning_rate": 8.897066910207958e-06,
+      "loss": 2.307,
+      "step": 590
+    },
+    {
+      "epoch": 0.22796528447444553,
+      "grad_norm": 0.46317031367156986,
+      "learning_rate": 8.893207690248776e-06,
+      "loss": 2.3562,
+      "step": 591
+    },
+    {
+      "epoch": 0.228351012536162,
+      "grad_norm": 0.48489527142084543,
+      "learning_rate": 8.88934257037231e-06,
+      "loss": 2.3807,
+      "step": 592
+    },
+    {
+      "epoch": 0.2287367405978785,
+      "grad_norm": 0.4528141570294088,
+      "learning_rate": 8.88547155643591e-06,
+      "loss": 2.2668,
+      "step": 593
+    },
+    {
+      "epoch": 0.22912246865959499,
+      "grad_norm": 0.4251865157625066,
+      "learning_rate": 8.88159465430586e-06,
+      "loss": 2.2951,
+      "step": 594
+    },
+    {
+      "epoch": 0.22950819672131148,
+      "grad_norm": 0.46093546704278265,
+      "learning_rate": 8.877711869857368e-06,
+      "loss": 2.3992,
+      "step": 595
+    },
+    {
+      "epoch": 0.22989392478302798,
+      "grad_norm": 0.4433096544674703,
+      "learning_rate": 8.873823208974557e-06,
+      "loss": 2.3377,
+      "step": 596
+    },
+    {
+      "epoch": 0.23027965284474444,
+      "grad_norm": 0.46530553084201687,
+      "learning_rate": 8.869928677550453e-06,
+      "loss": 2.3484,
+      "step": 597
+    },
+    {
+      "epoch": 0.23066538090646094,
+      "grad_norm": 0.46108298845071083,
+      "learning_rate": 8.866028281486978e-06,
+      "loss": 2.3731,
+      "step": 598
+    },
+    {
+      "epoch": 0.23105110896817743,
+      "grad_norm": 0.43532198423080554,
+      "learning_rate": 8.862122026694944e-06,
+      "loss": 2.3332,
+      "step": 599
+    },
+    {
+      "epoch": 0.23143683702989393,
+      "grad_norm": 0.44454741853143076,
+      "learning_rate": 8.85820991909404e-06,
+      "loss": 2.2882,
+      "step": 600
+    },
+    {
+      "epoch": 0.23182256509161042,
+      "grad_norm": 0.46404386426637784,
+      "learning_rate": 8.854291964612824e-06,
+      "loss": 2.3363,
+      "step": 601
+    },
+    {
+      "epoch": 0.23220829315332692,
+      "grad_norm": 0.4398707058122081,
+      "learning_rate": 8.850368169188717e-06,
+      "loss": 2.3552,
+      "step": 602
+    },
+    {
+      "epoch": 0.23259402121504338,
+      "grad_norm": 0.45305515998890616,
+      "learning_rate": 8.84643853876799e-06,
+      "loss": 2.3186,
+      "step": 603
+    },
+    {
+      "epoch": 0.23297974927675988,
+      "grad_norm": 0.4821684943054939,
+      "learning_rate": 8.842503079305757e-06,
+      "loss": 2.3453,
+      "step": 604
+    },
+    {
+      "epoch": 0.23336547733847637,
+      "grad_norm": 0.4470144221350872,
+      "learning_rate": 8.838561796765964e-06,
+      "loss": 2.3314,
+      "step": 605
+    },
+    {
+      "epoch": 0.23375120540019287,
+      "grad_norm": 0.4798478803514554,
+      "learning_rate": 8.834614697121384e-06,
+      "loss": 2.299,
+      "step": 606
+    },
+    {
+      "epoch": 0.23413693346190936,
+      "grad_norm": 0.45620806995059565,
+      "learning_rate": 8.830661786353602e-06,
+      "loss": 2.3667,
+      "step": 607
+    },
+    {
+      "epoch": 0.23452266152362583,
+      "grad_norm": 0.4363843604095868,
+      "learning_rate": 8.826703070453014e-06,
+      "loss": 2.3131,
+      "step": 608
+    },
+    {
+      "epoch": 0.23490838958534233,
+      "grad_norm": 0.41484049221741226,
+      "learning_rate": 8.82273855541881e-06,
+      "loss": 2.3871,
+      "step": 609
+    },
+    {
+      "epoch": 0.23529411764705882,
+      "grad_norm": 0.43654986338390206,
+      "learning_rate": 8.81876824725897e-06,
+      "loss": 2.4467,
+      "step": 610
+    },
+    {
+      "epoch": 0.23567984570877532,
+      "grad_norm": 0.43244037227389426,
+      "learning_rate": 8.814792151990253e-06,
+      "loss": 2.3314,
+      "step": 611
+    },
+    {
+      "epoch": 0.2360655737704918,
+      "grad_norm": 0.4471834289259313,
+      "learning_rate": 8.810810275638183e-06,
+      "loss": 2.2777,
+      "step": 612
+    },
+    {
+      "epoch": 0.2364513018322083,
+      "grad_norm": 0.4610781584149346,
+      "learning_rate": 8.806822624237055e-06,
+      "loss": 2.2824,
+      "step": 613
+    },
+    {
+      "epoch": 0.23683702989392477,
+      "grad_norm": 0.5055359126910268,
+      "learning_rate": 8.802829203829904e-06,
+      "loss": 2.3409,
+      "step": 614
+    },
+    {
+      "epoch": 0.23722275795564127,
+      "grad_norm": 0.431466867002683,
+      "learning_rate": 8.798830020468519e-06,
+      "loss": 2.3704,
+      "step": 615
+    },
+    {
+      "epoch": 0.23760848601735776,
+      "grad_norm": 0.41289024443297107,
+      "learning_rate": 8.794825080213415e-06,
+      "loss": 2.2887,
+      "step": 616
+    },
+    {
+      "epoch": 0.23799421407907426,
+      "grad_norm": 0.42480354215245325,
+      "learning_rate": 8.79081438913383e-06,
+      "loss": 2.2875,
+      "step": 617
+    },
+    {
+      "epoch": 0.23837994214079075,
+      "grad_norm": 0.4368624737440716,
+      "learning_rate": 8.786797953307725e-06,
+      "loss": 2.3376,
+      "step": 618
+    },
+    {
+      "epoch": 0.23876567020250722,
+      "grad_norm": 0.4443608049799916,
+      "learning_rate": 8.78277577882176e-06,
+      "loss": 2.2969,
+      "step": 619
+    },
+    {
+      "epoch": 0.2391513982642237,
+      "grad_norm": 0.45500903232533707,
+      "learning_rate": 8.778747871771293e-06,
+      "loss": 2.3506,
+      "step": 620
+    },
+    {
+      "epoch": 0.2395371263259402,
+      "grad_norm": 0.4491929565912108,
+      "learning_rate": 8.774714238260369e-06,
+      "loss": 2.4093,
+      "step": 621
+    },
+    {
+      "epoch": 0.2399228543876567,
+      "grad_norm": 0.4826105022745694,
+      "learning_rate": 8.770674884401714e-06,
+      "loss": 2.3438,
+      "step": 622
+    },
+    {
+      "epoch": 0.2403085824493732,
+      "grad_norm": 0.4541889182729078,
+      "learning_rate": 8.766629816316722e-06,
+      "loss": 2.2042,
+      "step": 623
+    },
+    {
+      "epoch": 0.2406943105110897,
+      "grad_norm": 0.4576938178983255,
+      "learning_rate": 8.76257904013544e-06,
+      "loss": 2.3415,
+      "step": 624
+    },
+    {
+      "epoch": 0.24108003857280616,
+      "grad_norm": 0.4234180305215837,
+      "learning_rate": 8.758522561996577e-06,
+      "loss": 2.3312,
+      "step": 625
+    },
+    {
+      "epoch": 0.24146576663452265,
+      "grad_norm": 0.44449556958449665,
+      "learning_rate": 8.754460388047472e-06,
+      "loss": 2.381,
+      "step": 626
+    },
+    {
+      "epoch": 0.24185149469623915,
+      "grad_norm": 0.4432541863472157,
+      "learning_rate": 8.750392524444102e-06,
+      "loss": 2.3367,
+      "step": 627
+    },
+    {
+      "epoch": 0.24223722275795564,
+      "grad_norm": 0.44148252274322086,
+      "learning_rate": 8.746318977351066e-06,
+      "loss": 2.3565,
+      "step": 628
+    },
+    {
+      "epoch": 0.24262295081967214,
+      "grad_norm": 0.5134089752365603,
+      "learning_rate": 8.742239752941572e-06,
+      "loss": 2.303,
+      "step": 629
+    },
+    {
+      "epoch": 0.24300867888138863,
+      "grad_norm": 0.4443950399081667,
+      "learning_rate": 8.738154857397437e-06,
+      "loss": 2.3333,
+      "step": 630
+    },
+    {
+      "epoch": 0.2433944069431051,
+      "grad_norm": 0.42103159116790523,
+      "learning_rate": 8.734064296909066e-06,
+      "loss": 2.3859,
+      "step": 631
+    },
+    {
+      "epoch": 0.2437801350048216,
+      "grad_norm": 0.42843587691048385,
+      "learning_rate": 8.729968077675454e-06,
+      "loss": 2.3206,
+      "step": 632
+    },
+    {
+      "epoch": 0.2441658630665381,
+      "grad_norm": 0.4830071126848421,
+      "learning_rate": 8.725866205904173e-06,
+      "loss": 2.3747,
+      "step": 633
+    },
+    {
+      "epoch": 0.2445515911282546,
+      "grad_norm": 0.49859578340636046,
+      "learning_rate": 8.721758687811353e-06,
+      "loss": 2.388,
+      "step": 634
+    },
+    {
+      "epoch": 0.24493731918997108,
+      "grad_norm": 0.4439955166387394,
+      "learning_rate": 8.717645529621686e-06,
+      "loss": 2.3674,
+      "step": 635
+    },
+    {
+      "epoch": 0.24532304725168755,
+      "grad_norm": 0.4229037183311655,
+      "learning_rate": 8.713526737568415e-06,
+      "loss": 2.3056,
+      "step": 636
+    },
+    {
+      "epoch": 0.24570877531340404,
+      "grad_norm": 0.4383644039158399,
+      "learning_rate": 8.709402317893312e-06,
+      "loss": 2.3957,
+      "step": 637
+    },
+    {
+      "epoch": 0.24609450337512054,
+      "grad_norm": 0.42455327869988235,
+      "learning_rate": 8.705272276846684e-06,
+      "loss": 2.3569,
+      "step": 638
+    },
+    {
+      "epoch": 0.24648023143683703,
+      "grad_norm": 0.45921277553192413,
+      "learning_rate": 8.701136620687355e-06,
+      "loss": 2.3386,
+      "step": 639
+    },
+    {
+      "epoch": 0.24686595949855353,
+      "grad_norm": 0.43121730879012,
+      "learning_rate": 8.696995355682656e-06,
+      "loss": 2.3687,
+      "step": 640
+    },
+    {
+      "epoch": 0.24725168756027002,
+      "grad_norm": 0.4462369019493956,
+      "learning_rate": 8.692848488108423e-06,
+      "loss": 2.3777,
+      "step": 641
+    },
+    {
+      "epoch": 0.2476374156219865,
+      "grad_norm": 0.43436360192004747,
+      "learning_rate": 8.688696024248977e-06,
+      "loss": 2.3914,
+      "step": 642
+    },
+    {
+      "epoch": 0.24802314368370298,
+      "grad_norm": 0.4137768504976057,
+      "learning_rate": 8.684537970397122e-06,
+      "loss": 2.4054,
+      "step": 643
+    },
+    {
+      "epoch": 0.24840887174541948,
+      "grad_norm": 0.4415096763888693,
+      "learning_rate": 8.680374332854134e-06,
+      "loss": 2.3903,
+      "step": 644
+    },
+    {
+      "epoch": 0.24879459980713597,
+      "grad_norm": 0.4556463786464995,
+      "learning_rate": 8.676205117929752e-06,
+      "loss": 2.3369,
+      "step": 645
+    },
+    {
+      "epoch": 0.24918032786885247,
+      "grad_norm": 0.4182380195172059,
+      "learning_rate": 8.672030331942163e-06,
+      "loss": 2.3101,
+      "step": 646
+    },
+    {
+      "epoch": 0.24956605593056894,
+      "grad_norm": 0.43948879831910515,
+      "learning_rate": 8.667849981217997e-06,
+      "loss": 2.3495,
+      "step": 647
+    },
+    {
+      "epoch": 0.24995178399228543,
+      "grad_norm": 0.431206271755024,
+      "learning_rate": 8.663664072092324e-06,
+      "loss": 2.3322,
+      "step": 648
+    },
+    {
+      "epoch": 0.25033751205400195,
+      "grad_norm": 0.4628389056590261,
+      "learning_rate": 8.659472610908628e-06,
+      "loss": 2.2407,
+      "step": 649
+    },
+    {
+      "epoch": 0.2507232401157184,
+      "grad_norm": 0.42026399292897054,
+      "learning_rate": 8.655275604018813e-06,
+      "loss": 2.3461,
+      "step": 650
+    },
+    {
+      "epoch": 0.2511089681774349,
+      "grad_norm": 0.46049227256605096,
+      "learning_rate": 8.651073057783185e-06,
+      "loss": 2.2959,
+      "step": 651
+    },
+    {
+      "epoch": 0.2514946962391514,
+      "grad_norm": 0.44202962984695643,
+      "learning_rate": 8.646864978570445e-06,
+      "loss": 2.3031,
+      "step": 652
+    },
+    {
+      "epoch": 0.2518804243008679,
+      "grad_norm": 0.4360453070528761,
+      "learning_rate": 8.64265137275768e-06,
+      "loss": 2.3151,
+      "step": 653
+    },
+    {
+      "epoch": 0.2522661523625844,
+      "grad_norm": 0.444762011719982,
+      "learning_rate": 8.638432246730351e-06,
+      "loss": 2.2721,
+      "step": 654
+    },
+    {
+      "epoch": 0.25265188042430087,
+      "grad_norm": 0.4147806152584227,
+      "learning_rate": 8.634207606882282e-06,
+      "loss": 2.2823,
+      "step": 655
+    },
+    {
+      "epoch": 0.25303760848601736,
+      "grad_norm": 0.4529975286905433,
+      "learning_rate": 8.629977459615655e-06,
+      "loss": 2.3199,
+      "step": 656
+    },
+    {
+      "epoch": 0.25342333654773386,
+      "grad_norm": 0.44515481137345064,
+      "learning_rate": 8.625741811341001e-06,
+      "loss": 2.328,
+      "step": 657
+    },
+    {
+      "epoch": 0.25380906460945035,
+      "grad_norm": 0.44598812762748385,
+      "learning_rate": 8.621500668477184e-06,
+      "loss": 2.3782,
+      "step": 658
+    },
+    {
+      "epoch": 0.25419479267116685,
+      "grad_norm": 0.45731308991486175,
+      "learning_rate": 8.617254037451396e-06,
+      "loss": 2.3809,
+      "step": 659
+    },
+    {
+      "epoch": 0.25458052073288334,
+      "grad_norm": 0.45291014192182655,
+      "learning_rate": 8.613001924699146e-06,
+      "loss": 2.3475,
+      "step": 660
+    },
+    {
+      "epoch": 0.2549662487945998,
+      "grad_norm": 0.4594502645503529,
+      "learning_rate": 8.60874433666425e-06,
+      "loss": 2.2756,
+      "step": 661
+    },
+    {
+      "epoch": 0.2553519768563163,
+      "grad_norm": 0.45670632233299724,
+      "learning_rate": 8.60448127979882e-06,
+      "loss": 2.3647,
+      "step": 662
+    },
+    {
+      "epoch": 0.25573770491803277,
+      "grad_norm": 0.44423047286891265,
+      "learning_rate": 8.600212760563257e-06,
+      "loss": 2.3758,
+      "step": 663
+    },
+    {
+      "epoch": 0.25612343297974927,
+      "grad_norm": 0.42893572139376235,
+      "learning_rate": 8.595938785426241e-06,
+      "loss": 2.3226,
+      "step": 664
+    },
+    {
+      "epoch": 0.25650916104146576,
+      "grad_norm": 0.47474284251148635,
+      "learning_rate": 8.591659360864718e-06,
+      "loss": 2.3082,
+      "step": 665
+    },
+    {
+      "epoch": 0.25689488910318226,
+      "grad_norm": 0.47126129734527794,
+      "learning_rate": 8.587374493363895e-06,
+      "loss": 2.3365,
+      "step": 666
+    },
+    {
+      "epoch": 0.25728061716489875,
+      "grad_norm": 0.4788897233983171,
+      "learning_rate": 8.583084189417225e-06,
+      "loss": 2.4082,
+      "step": 667
+    },
+    {
+      "epoch": 0.25766634522661525,
+      "grad_norm": 0.44412197119577956,
+      "learning_rate": 8.578788455526398e-06,
+      "loss": 2.3528,
+      "step": 668
+    },
+    {
+      "epoch": 0.25805207328833174,
+      "grad_norm": 0.4333346854250443,
+      "learning_rate": 8.574487298201337e-06,
+      "loss": 2.3591,
+      "step": 669
+    },
+    {
+      "epoch": 0.25843780135004824,
+      "grad_norm": 0.45321598100218463,
+      "learning_rate": 8.570180723960181e-06,
+      "loss": 2.3196,
+      "step": 670
+    },
+    {
+      "epoch": 0.25882352941176473,
+      "grad_norm": 0.4720265187911564,
+      "learning_rate": 8.565868739329282e-06,
+      "loss": 2.2847,
+      "step": 671
+    },
+    {
+      "epoch": 0.25920925747348117,
+      "grad_norm": 0.4266778768608411,
+      "learning_rate": 8.561551350843185e-06,
+      "loss": 2.311,
+      "step": 672
+    },
+    {
+      "epoch": 0.25959498553519766,
+      "grad_norm": 0.41875169594932343,
+      "learning_rate": 8.557228565044629e-06,
+      "loss": 2.3319,
+      "step": 673
+    },
+    {
+      "epoch": 0.25998071359691416,
+      "grad_norm": 0.4576486818375762,
+      "learning_rate": 8.552900388484527e-06,
+      "loss": 2.3906,
+      "step": 674
+    },
+    {
+      "epoch": 0.26036644165863065,
+      "grad_norm": 0.45405522032831847,
+      "learning_rate": 8.548566827721968e-06,
+      "loss": 2.3113,
+      "step": 675
+    },
+    {
+      "epoch": 0.26075216972034715,
+      "grad_norm": 0.45848135978953214,
+      "learning_rate": 8.544227889324199e-06,
+      "loss": 2.3568,
+      "step": 676
+    },
+    {
+      "epoch": 0.26113789778206364,
+      "grad_norm": 0.5325117292327657,
+      "learning_rate": 8.53988357986661e-06,
+      "loss": 2.3044,
+      "step": 677
+    },
+    {
+      "epoch": 0.26152362584378014,
+      "grad_norm": 0.460540990758648,
+      "learning_rate": 8.535533905932739e-06,
+      "loss": 2.335,
+      "step": 678
+    },
+    {
+      "epoch": 0.26190935390549663,
+      "grad_norm": 0.4297960673737978,
+      "learning_rate": 8.531178874114248e-06,
+      "loss": 2.3894,
+      "step": 679
+    },
+    {
+      "epoch": 0.26229508196721313,
+      "grad_norm": 0.43060572349815474,
+      "learning_rate": 8.526818491010922e-06,
+      "loss": 2.2778,
+      "step": 680
+    },
+    {
+      "epoch": 0.2626808100289296,
+      "grad_norm": 0.4052028043437063,
+      "learning_rate": 8.52245276323065e-06,
+      "loss": 2.271,
+      "step": 681
+    },
+    {
+      "epoch": 0.2630665380906461,
+      "grad_norm": 0.467624997249077,
+      "learning_rate": 8.518081697389433e-06,
+      "loss": 2.3156,
+      "step": 682
+    },
+    {
+      "epoch": 0.26345226615236256,
+      "grad_norm": 0.456634740268385,
+      "learning_rate": 8.513705300111344e-06,
+      "loss": 2.3432,
+      "step": 683
+    },
+    {
+      "epoch": 0.26383799421407905,
+      "grad_norm": 0.4565601445461014,
+      "learning_rate": 8.509323578028547e-06,
+      "loss": 2.3478,
+      "step": 684
+    },
+    {
+      "epoch": 0.26422372227579555,
+      "grad_norm": 0.43377812392761256,
+      "learning_rate": 8.504936537781276e-06,
+      "loss": 2.337,
+      "step": 685
+    },
+    {
+      "epoch": 0.26460945033751204,
+      "grad_norm": 0.4379959990068602,
+      "learning_rate": 8.500544186017818e-06,
+      "loss": 2.3597,
+      "step": 686
+    },
+    {
+      "epoch": 0.26499517839922854,
+      "grad_norm": 0.4393854424171466,
+      "learning_rate": 8.496146529394515e-06,
+      "loss": 2.3475,
+      "step": 687
+    },
+    {
+      "epoch": 0.26538090646094503,
+      "grad_norm": 0.46530108150002264,
+      "learning_rate": 8.491743574575743e-06,
+      "loss": 2.3816,
+      "step": 688
+    },
+    {
+      "epoch": 0.2657666345226615,
+      "grad_norm": 0.42409285117482814,
+      "learning_rate": 8.487335328233912e-06,
+      "loss": 2.2594,
+      "step": 689
+    },
+    {
+      "epoch": 0.266152362584378,
+      "grad_norm": 0.46669962561508027,
+      "learning_rate": 8.482921797049445e-06,
+      "loss": 2.3251,
+      "step": 690
+    },
+    {
+      "epoch": 0.2665380906460945,
+      "grad_norm": 0.45599266566706936,
+      "learning_rate": 8.478502987710784e-06,
+      "loss": 2.2822,
+      "step": 691
+    },
+    {
+      "epoch": 0.266923818707811,
+      "grad_norm": 0.44453326279872246,
+      "learning_rate": 8.474078906914359e-06,
+      "loss": 2.2975,
+      "step": 692
+    },
+    {
+      "epoch": 0.2673095467695275,
+      "grad_norm": 0.4421890363738117,
+      "learning_rate": 8.469649561364592e-06,
+      "loss": 2.4006,
+      "step": 693
+    },
+    {
+      "epoch": 0.26769527483124395,
+      "grad_norm": 0.42181612986205536,
+      "learning_rate": 8.465214957773886e-06,
+      "loss": 2.339,
+      "step": 694
+    },
+    {
+      "epoch": 0.26808100289296044,
+      "grad_norm": 0.4781947811050221,
+      "learning_rate": 8.46077510286261e-06,
+      "loss": 2.3552,
+      "step": 695
+    },
+    {
+      "epoch": 0.26846673095467694,
+      "grad_norm": 0.44218204028140373,
+      "learning_rate": 8.456330003359093e-06,
+      "loss": 2.3174,
+      "step": 696
+    },
+    {
+      "epoch": 0.26885245901639343,
+      "grad_norm": 0.4335540427536611,
+      "learning_rate": 8.45187966599961e-06,
+      "loss": 2.3073,
+      "step": 697
+    },
+    {
+      "epoch": 0.2692381870781099,
+      "grad_norm": 0.4255526529918983,
+      "learning_rate": 8.447424097528374e-06,
+      "loss": 2.3241,
+      "step": 698
+    },
+    {
+      "epoch": 0.2696239151398264,
+      "grad_norm": 0.4519888234246226,
+      "learning_rate": 8.442963304697522e-06,
+      "loss": 2.2504,
+      "step": 699
+    },
+    {
+      "epoch": 0.2700096432015429,
+      "grad_norm": 0.4285496344210807,
+      "learning_rate": 8.438497294267117e-06,
+      "loss": 2.3666,
+      "step": 700
+    },
+    {
+      "epoch": 0.2703953712632594,
+      "grad_norm": 0.46756712739850437,
+      "learning_rate": 8.434026073005121e-06,
+      "loss": 2.3016,
+      "step": 701
+    },
+    {
+      "epoch": 0.2707810993249759,
+      "grad_norm": 0.43224586635772977,
+      "learning_rate": 8.429549647687396e-06,
+      "loss": 2.2907,
+      "step": 702
+    },
+    {
+      "epoch": 0.2711668273866924,
+      "grad_norm": 0.4279270332396626,
+      "learning_rate": 8.42506802509769e-06,
+      "loss": 2.327,
+      "step": 703
+    },
+    {
+      "epoch": 0.2715525554484089,
+      "grad_norm": 0.4806493766376229,
+      "learning_rate": 8.420581212027625e-06,
+      "loss": 2.3157,
+      "step": 704
+    },
+    {
+      "epoch": 0.2719382835101254,
+      "grad_norm": 0.43953840194209576,
+      "learning_rate": 8.416089215276695e-06,
+      "loss": 2.2619,
+      "step": 705
+    },
+    {
+      "epoch": 0.27232401157184183,
+      "grad_norm": 0.4582687236589324,
+      "learning_rate": 8.411592041652241e-06,
+      "loss": 2.3504,
+      "step": 706
+    },
+    {
+      "epoch": 0.2727097396335583,
+      "grad_norm": 0.44503957981450315,
+      "learning_rate": 8.407089697969458e-06,
+      "loss": 2.3481,
+      "step": 707
+    },
+    {
+      "epoch": 0.2730954676952748,
+      "grad_norm": 0.42107672742502356,
+      "learning_rate": 8.402582191051365e-06,
+      "loss": 2.3103,
+      "step": 708
+    },
+    {
+      "epoch": 0.2734811957569913,
+      "grad_norm": 0.45027516032651865,
+      "learning_rate": 8.398069527728818e-06,
+      "loss": 2.3194,
+      "step": 709
+    },
+    {
+      "epoch": 0.2738669238187078,
+      "grad_norm": 0.46235879544669245,
+      "learning_rate": 8.393551714840477e-06,
+      "loss": 2.2534,
+      "step": 710
+    },
+    {
+      "epoch": 0.2742526518804243,
+      "grad_norm": 0.4241828703032586,
+      "learning_rate": 8.389028759232816e-06,
+      "loss": 2.3229,
+      "step": 711
+    },
+    {
+      "epoch": 0.2746383799421408,
+      "grad_norm": 0.3991820998003778,
+      "learning_rate": 8.38450066776009e-06,
+      "loss": 2.3235,
+      "step": 712
+    },
+    {
+      "epoch": 0.2750241080038573,
+      "grad_norm": 0.45056071729226377,
+      "learning_rate": 8.379967447284348e-06,
+      "loss": 2.4087,
+      "step": 713
+    },
+    {
+      "epoch": 0.2754098360655738,
+      "grad_norm": 0.44409452817733797,
+      "learning_rate": 8.375429104675404e-06,
+      "loss": 2.3609,
+      "step": 714
+    },
+    {
+      "epoch": 0.2757955641272903,
+      "grad_norm": 0.45902422826903083,
+      "learning_rate": 8.370885646810842e-06,
+      "loss": 2.3387,
+      "step": 715
+    },
+    {
+      "epoch": 0.2761812921890068,
+      "grad_norm": 0.4678718330955543,
+      "learning_rate": 8.36633708057599e-06,
+      "loss": 2.3463,
+      "step": 716
+    },
+    {
+      "epoch": 0.2765670202507232,
+      "grad_norm": 0.42990641040563493,
+      "learning_rate": 8.361783412863922e-06,
+      "loss": 2.3658,
+      "step": 717
+    },
+    {
+      "epoch": 0.2769527483124397,
+      "grad_norm": 0.44293983886938715,
+      "learning_rate": 8.357224650575442e-06,
+      "loss": 2.2432,
+      "step": 718
+    },
+    {
+      "epoch": 0.2773384763741562,
+      "grad_norm": 0.4748980516116094,
+      "learning_rate": 8.352660800619075e-06,
+      "loss": 2.3018,
+      "step": 719
+    },
+    {
+      "epoch": 0.2777242044358727,
+      "grad_norm": 0.45821392252427534,
+      "learning_rate": 8.348091869911054e-06,
+      "loss": 2.378,
+      "step": 720
+    },
+    {
+      "epoch": 0.2781099324975892,
+      "grad_norm": 0.43505219416127844,
+      "learning_rate": 8.343517865375314e-06,
+      "loss": 2.3017,
+      "step": 721
+    },
+    {
+      "epoch": 0.2784956605593057,
+      "grad_norm": 0.47054578385085105,
+      "learning_rate": 8.338938793943478e-06,
+      "loss": 2.3064,
+      "step": 722
+    },
+    {
+      "epoch": 0.2788813886210222,
+      "grad_norm": 0.5068953478792466,
+      "learning_rate": 8.334354662554848e-06,
+      "loss": 2.3934,
+      "step": 723
+    },
+    {
+      "epoch": 0.2792671166827387,
+      "grad_norm": 0.4274771286480794,
+      "learning_rate": 8.329765478156394e-06,
+      "loss": 2.3296,
+      "step": 724
+    },
+    {
+      "epoch": 0.2796528447444552,
+      "grad_norm": 0.4540017236562187,
+      "learning_rate": 8.325171247702742e-06,
+      "loss": 2.4178,
+      "step": 725
+    },
+    {
+      "epoch": 0.28003857280617167,
+      "grad_norm": 0.43197431133254655,
+      "learning_rate": 8.320571978156169e-06,
+      "loss": 2.306,
+      "step": 726
+    },
+    {
+      "epoch": 0.28042430086788817,
+      "grad_norm": 0.5257629218842567,
+      "learning_rate": 8.315967676486581e-06,
+      "loss": 2.3461,
+      "step": 727
+    },
+    {
+      "epoch": 0.2808100289296046,
+      "grad_norm": 0.42206142604158375,
+      "learning_rate": 8.311358349671516e-06,
+      "loss": 2.3429,
+      "step": 728
+    },
+    {
+      "epoch": 0.2811957569913211,
+      "grad_norm": 0.4310166461530932,
+      "learning_rate": 8.30674400469613e-06,
+      "loss": 2.3838,
+      "step": 729
+    },
+    {
+      "epoch": 0.2815814850530376,
+      "grad_norm": 0.47512136364839236,
+      "learning_rate": 8.302124648553175e-06,
+      "loss": 2.2788,
+      "step": 730
+    },
+    {
+      "epoch": 0.2819672131147541,
+      "grad_norm": 0.44485823670509395,
+      "learning_rate": 8.297500288243006e-06,
+      "loss": 2.3716,
+      "step": 731
+    },
+    {
+      "epoch": 0.2823529411764706,
+      "grad_norm": 0.42194008296317503,
+      "learning_rate": 8.292870930773551e-06,
+      "loss": 2.2852,
+      "step": 732
+    },
+    {
+      "epoch": 0.2827386692381871,
+      "grad_norm": 0.4727245134331024,
+      "learning_rate": 8.288236583160322e-06,
+      "loss": 2.4187,
+      "step": 733
+    },
+    {
+      "epoch": 0.2831243972999036,
+      "grad_norm": 0.4656778918218644,
+      "learning_rate": 8.283597252426389e-06,
+      "loss": 2.3504,
+      "step": 734
+    },
+    {
+      "epoch": 0.28351012536162007,
+      "grad_norm": 0.4431636979476998,
+      "learning_rate": 8.27895294560237e-06,
+      "loss": 2.3422,
+      "step": 735
+    },
+    {
+      "epoch": 0.28389585342333656,
+      "grad_norm": 0.41930968524089257,
+      "learning_rate": 8.274303669726427e-06,
+      "loss": 2.308,
+      "step": 736
+    },
+    {
+      "epoch": 0.28428158148505306,
+      "grad_norm": 0.4722213099735777,
+      "learning_rate": 8.269649431844253e-06,
+      "loss": 2.359,
+      "step": 737
+    },
+    {
+      "epoch": 0.28466730954676955,
+      "grad_norm": 0.4884260822490044,
+      "learning_rate": 8.26499023900906e-06,
+      "loss": 2.3165,
+      "step": 738
+    },
+    {
+      "epoch": 0.285053037608486,
+      "grad_norm": 0.453666419747345,
+      "learning_rate": 8.260326098281567e-06,
+      "loss": 2.2771,
+      "step": 739
+    },
+    {
+      "epoch": 0.2854387656702025,
+      "grad_norm": 0.434870597831758,
+      "learning_rate": 8.255657016729997e-06,
+      "loss": 2.3319,
+      "step": 740
+    },
+    {
+      "epoch": 0.285824493731919,
+      "grad_norm": 0.4431025745177239,
+      "learning_rate": 8.250983001430055e-06,
+      "loss": 2.3395,
+      "step": 741
+    },
+    {
+      "epoch": 0.2862102217936355,
+      "grad_norm": 0.47702025242587814,
+      "learning_rate": 8.24630405946492e-06,
+      "loss": 2.2912,
+      "step": 742
+    },
+    {
+      "epoch": 0.286595949855352,
+      "grad_norm": 0.4285271915769268,
+      "learning_rate": 8.241620197925247e-06,
+      "loss": 2.3997,
+      "step": 743
+    },
+    {
+      "epoch": 0.28698167791706847,
+      "grad_norm": 0.4519377530435274,
+      "learning_rate": 8.23693142390914e-06,
+      "loss": 2.385,
+      "step": 744
+    },
+    {
+      "epoch": 0.28736740597878496,
+      "grad_norm": 0.42967581483794326,
+      "learning_rate": 8.232237744522145e-06,
+      "loss": 2.3602,
+      "step": 745
+    },
+    {
+      "epoch": 0.28775313404050146,
+      "grad_norm": 0.4380475395781914,
+      "learning_rate": 8.227539166877244e-06,
+      "loss": 2.356,
+      "step": 746
+    },
+    {
+      "epoch": 0.28813886210221795,
+      "grad_norm": 0.428033800976379,
+      "learning_rate": 8.222835698094849e-06,
+      "loss": 2.3886,
+      "step": 747
+    },
+    {
+      "epoch": 0.28852459016393445,
+      "grad_norm": 0.43342082942139437,
+      "learning_rate": 8.218127345302775e-06,
+      "loss": 2.3613,
+      "step": 748
+    },
+    {
+      "epoch": 0.28891031822565094,
+      "grad_norm": 0.43796959158042176,
+      "learning_rate": 8.21341411563624e-06,
+      "loss": 2.2804,
+      "step": 749
+    },
+    {
+      "epoch": 0.2892960462873674,
+      "grad_norm": 0.4258383538912118,
+      "learning_rate": 8.208696016237858e-06,
+      "loss": 2.3471,
+      "step": 750
+    },
+    {
+      "epoch": 0.2896817743490839,
+      "grad_norm": 0.4527951901212638,
+      "learning_rate": 8.203973054257614e-06,
+      "loss": 2.3433,
+      "step": 751
+    },
+    {
+      "epoch": 0.29006750241080037,
+      "grad_norm": 0.4746672206815161,
+      "learning_rate": 8.199245236852871e-06,
+      "loss": 2.3056,
+      "step": 752
+    },
+    {
+      "epoch": 0.29045323047251687,
+      "grad_norm": 0.446204159927227,
+      "learning_rate": 8.194512571188347e-06,
+      "loss": 2.3905,
+      "step": 753
+    },
+    {
+      "epoch": 0.29083895853423336,
+      "grad_norm": 0.4390420515188638,
+      "learning_rate": 8.189775064436101e-06,
+      "loss": 2.2572,
+      "step": 754
+    },
+    {
+      "epoch": 0.29122468659594986,
+      "grad_norm": 0.4249036856189722,
+      "learning_rate": 8.18503272377554e-06,
+      "loss": 2.3421,
+      "step": 755
+    },
+    {
+      "epoch": 0.29161041465766635,
+      "grad_norm": 0.4477391836717647,
+      "learning_rate": 8.180285556393384e-06,
+      "loss": 2.241,
+      "step": 756
+    },
+    {
+      "epoch": 0.29199614271938285,
+      "grad_norm": 0.45573037235185965,
+      "learning_rate": 8.175533569483678e-06,
+      "loss": 2.2949,
+      "step": 757
+    },
+    {
+      "epoch": 0.29238187078109934,
+      "grad_norm": 0.45332611581297916,
+      "learning_rate": 8.170776770247766e-06,
+      "loss": 2.2902,
+      "step": 758
+    },
+    {
+      "epoch": 0.29276759884281583,
+      "grad_norm": 0.42911111909344035,
+      "learning_rate": 8.166015165894285e-06,
+      "loss": 2.3321,
+      "step": 759
+    },
+    {
+      "epoch": 0.29315332690453233,
+      "grad_norm": 0.45635018516913833,
+      "learning_rate": 8.161248763639154e-06,
+      "loss": 2.3595,
+      "step": 760
+    },
+    {
+      "epoch": 0.29353905496624877,
+      "grad_norm": 0.4353529159629687,
+      "learning_rate": 8.156477570705561e-06,
+      "loss": 2.3467,
+      "step": 761
+    },
+    {
+      "epoch": 0.29392478302796526,
+      "grad_norm": 0.43159025437557735,
+      "learning_rate": 8.151701594323957e-06,
+      "loss": 2.3273,
+      "step": 762
+    },
+    {
+      "epoch": 0.29431051108968176,
+      "grad_norm": 0.45204397895329773,
+      "learning_rate": 8.146920841732045e-06,
+      "loss": 2.3402,
+      "step": 763
+    },
+    {
+      "epoch": 0.29469623915139825,
+      "grad_norm": 0.48590496149018464,
+      "learning_rate": 8.142135320174758e-06,
+      "loss": 2.3194,
+      "step": 764
+    },
+    {
+      "epoch": 0.29508196721311475,
+      "grad_norm": 0.44858443291686595,
+      "learning_rate": 8.13734503690426e-06,
+      "loss": 2.4044,
+      "step": 765
+    },
+    {
+      "epoch": 0.29546769527483124,
+      "grad_norm": 0.4780747442685094,
+      "learning_rate": 8.132549999179934e-06,
+      "loss": 2.3557,
+      "step": 766
+    },
+    {
+      "epoch": 0.29585342333654774,
+      "grad_norm": 0.46334226284408603,
+      "learning_rate": 8.127750214268363e-06,
+      "loss": 2.3217,
+      "step": 767
+    },
+    {
+      "epoch": 0.29623915139826423,
+      "grad_norm": 0.428819680186002,
+      "learning_rate": 8.122945689443328e-06,
+      "loss": 2.374,
+      "step": 768
+    },
+    {
+      "epoch": 0.29662487945998073,
+      "grad_norm": 0.49990179942528384,
+      "learning_rate": 8.11813643198579e-06,
+      "loss": 2.2795,
+      "step": 769
+    },
+    {
+      "epoch": 0.2970106075216972,
+      "grad_norm": 0.4869602981354855,
+      "learning_rate": 8.113322449183884e-06,
+      "loss": 2.3162,
+      "step": 770
+    },
+    {
+      "epoch": 0.2973963355834137,
+      "grad_norm": 0.4359106823661919,
+      "learning_rate": 8.108503748332906e-06,
+      "loss": 2.2967,
+      "step": 771
+    },
+    {
+      "epoch": 0.29778206364513016,
+      "grad_norm": 0.46852221277067796,
+      "learning_rate": 8.1036803367353e-06,
+      "loss": 2.39,
+      "step": 772
+    },
+    {
+      "epoch": 0.29816779170684665,
+      "grad_norm": 0.4376965119507577,
+      "learning_rate": 8.098852221700652e-06,
+      "loss": 2.3758,
+      "step": 773
+    },
+    {
+      "epoch": 0.29855351976856315,
+      "grad_norm": 0.42922571975329926,
+      "learning_rate": 8.094019410545673e-06,
+      "loss": 2.2821,
+      "step": 774
+    },
+    {
+      "epoch": 0.29893924783027964,
+      "grad_norm": 0.43239232104124353,
+      "learning_rate": 8.089181910594191e-06,
+      "loss": 2.4046,
+      "step": 775
+    },
+    {
+      "epoch": 0.29932497589199614,
+      "grad_norm": 0.4361179145248521,
+      "learning_rate": 8.084339729177142e-06,
+      "loss": 2.3665,
+      "step": 776
+    },
+    {
+      "epoch": 0.29971070395371263,
+      "grad_norm": 0.4620195317271602,
+      "learning_rate": 8.079492873632554e-06,
+      "loss": 2.3647,
+      "step": 777
+    },
+    {
+      "epoch": 0.3000964320154291,
+      "grad_norm": 0.4292740072549495,
+      "learning_rate": 8.074641351305539e-06,
+      "loss": 2.3595,
+      "step": 778
+    },
+    {
+      "epoch": 0.3004821600771456,
+      "grad_norm": 0.4702921848552218,
+      "learning_rate": 8.069785169548279e-06,
+      "loss": 2.3734,
+      "step": 779
+    },
+    {
+      "epoch": 0.3008678881388621,
+      "grad_norm": 0.47865810368594297,
+      "learning_rate": 8.064924335720023e-06,
+      "loss": 2.324,
+      "step": 780
+    },
+    {
+      "epoch": 0.3012536162005786,
+      "grad_norm": 0.4235903947368299,
+      "learning_rate": 8.060058857187066e-06,
+      "loss": 2.3847,
+      "step": 781
+    },
+    {
+      "epoch": 0.3016393442622951,
+      "grad_norm": 0.4284741604176135,
+      "learning_rate": 8.05518874132274e-06,
+      "loss": 2.3455,
+      "step": 782
+    },
+    {
+      "epoch": 0.30202507232401155,
+      "grad_norm": 0.4345349921182491,
+      "learning_rate": 8.050313995507406e-06,
+      "loss": 2.3149,
+      "step": 783
+    },
+    {
+      "epoch": 0.30241080038572804,
+      "grad_norm": 0.4687050619045661,
+      "learning_rate": 8.045434627128446e-06,
+      "loss": 2.3534,
+      "step": 784
+    },
+    {
+      "epoch": 0.30279652844744454,
+      "grad_norm": 0.44461581794714267,
+      "learning_rate": 8.04055064358024e-06,
+      "loss": 2.3629,
+      "step": 785
+    },
+    {
+      "epoch": 0.30318225650916103,
+      "grad_norm": 0.4460439925749081,
+      "learning_rate": 8.035662052264167e-06,
+      "loss": 2.2856,
+      "step": 786
+    },
+    {
+      "epoch": 0.3035679845708775,
+      "grad_norm": 0.42462245180475355,
+      "learning_rate": 8.030768860588585e-06,
+      "loss": 2.3613,
+      "step": 787
+    },
+    {
+      "epoch": 0.303953712632594,
+      "grad_norm": 0.45679951273446,
+      "learning_rate": 8.025871075968828e-06,
+      "loss": 2.303,
+      "step": 788
+    },
+    {
+      "epoch": 0.3043394406943105,
+      "grad_norm": 0.42176465264134017,
+      "learning_rate": 8.020968705827184e-06,
+      "loss": 2.4177,
+      "step": 789
+    },
+    {
+      "epoch": 0.304725168756027,
+      "grad_norm": 0.45242146039478087,
+      "learning_rate": 8.0160617575929e-06,
+      "loss": 2.3975,
+      "step": 790
+    },
+    {
+      "epoch": 0.3051108968177435,
+      "grad_norm": 0.45600534039329604,
+      "learning_rate": 8.01115023870215e-06,
+      "loss": 2.347,
+      "step": 791
+    },
+    {
+      "epoch": 0.30549662487946,
+      "grad_norm": 0.444106058641205,
+      "learning_rate": 8.006234156598043e-06,
+      "loss": 2.3369,
+      "step": 792
+    },
+    {
+      "epoch": 0.3058823529411765,
+      "grad_norm": 0.42845035166525575,
+      "learning_rate": 8.001313518730596e-06,
+      "loss": 2.3673,
+      "step": 793
+    },
+    {
+      "epoch": 0.30626808100289293,
+      "grad_norm": 0.4232424401622947,
+      "learning_rate": 7.996388332556735e-06,
+      "loss": 2.3429,
+      "step": 794
+    },
+    {
+      "epoch": 0.30665380906460943,
+      "grad_norm": 0.4488939460799838,
+      "learning_rate": 7.99145860554028e-06,
+      "loss": 2.248,
+      "step": 795
+    },
+    {
+      "epoch": 0.3070395371263259,
+      "grad_norm": 0.4455748705862296,
+      "learning_rate": 7.986524345151924e-06,
+      "loss": 2.3008,
+      "step": 796
+    },
+    {
+      "epoch": 0.3074252651880424,
+      "grad_norm": 0.4425635266461437,
+      "learning_rate": 7.981585558869244e-06,
+      "loss": 2.3849,
+      "step": 797
+    },
+    {
+      "epoch": 0.3078109932497589,
+      "grad_norm": 0.4777966028201153,
+      "learning_rate": 7.976642254176658e-06,
+      "loss": 2.3225,
+      "step": 798
+    },
+    {
+      "epoch": 0.3081967213114754,
+      "grad_norm": 0.47200162397794504,
+      "learning_rate": 7.97169443856545e-06,
+      "loss": 2.2937,
+      "step": 799
+    },
+    {
+      "epoch": 0.3085824493731919,
+      "grad_norm": 0.42909077888030944,
+      "learning_rate": 7.966742119533724e-06,
+      "loss": 2.3257,
+      "step": 800
+    },
+    {
+      "epoch": 0.3089681774349084,
+      "grad_norm": 0.4096126054850406,
+      "learning_rate": 7.961785304586418e-06,
+      "loss": 2.3265,
+      "step": 801
+    },
+    {
+      "epoch": 0.3093539054966249,
+      "grad_norm": 0.4347992134327031,
+      "learning_rate": 7.956824001235281e-06,
+      "loss": 2.3526,
+      "step": 802
+    },
+    {
+      "epoch": 0.3097396335583414,
+      "grad_norm": 0.43862916899554255,
+      "learning_rate": 7.951858216998863e-06,
+      "loss": 2.2812,
+      "step": 803
+    },
+    {
+      "epoch": 0.3101253616200579,
+      "grad_norm": 0.4600817148233885,
+      "learning_rate": 7.946887959402504e-06,
+      "loss": 2.33,
+      "step": 804
+    },
+    {
+      "epoch": 0.3105110896817743,
+      "grad_norm": 0.48101554196914814,
+      "learning_rate": 7.941913235978329e-06,
+      "loss": 2.3809,
+      "step": 805
+    },
+    {
+      "epoch": 0.3108968177434908,
+      "grad_norm": 0.4665225951071249,
+      "learning_rate": 7.936934054265222e-06,
+      "loss": 2.3546,
+      "step": 806
+    },
+    {
+      "epoch": 0.3112825458052073,
+      "grad_norm": 0.43534765876665066,
+      "learning_rate": 7.931950421808828e-06,
+      "loss": 2.3582,
+      "step": 807
+    },
+    {
+      "epoch": 0.3116682738669238,
+      "grad_norm": 0.42892045412622404,
+      "learning_rate": 7.926962346161535e-06,
+      "loss": 2.3411,
+      "step": 808
+    },
+    {
+      "epoch": 0.3120540019286403,
+      "grad_norm": 0.42918015973344015,
+      "learning_rate": 7.921969834882468e-06,
+      "loss": 2.2586,
+      "step": 809
+    },
+    {
+      "epoch": 0.3124397299903568,
+      "grad_norm": 0.4571111610981652,
+      "learning_rate": 7.916972895537471e-06,
+      "loss": 2.2982,
+      "step": 810
+    },
+    {
+      "epoch": 0.3128254580520733,
+      "grad_norm": 0.42461745307941356,
+      "learning_rate": 7.911971535699097e-06,
+      "loss": 2.3182,
+      "step": 811
+    },
+    {
+      "epoch": 0.3132111861137898,
+      "grad_norm": 0.46143612577410686,
+      "learning_rate": 7.9069657629466e-06,
+      "loss": 2.4201,
+      "step": 812
+    },
+    {
+      "epoch": 0.3135969141755063,
+      "grad_norm": 0.4515370603709611,
+      "learning_rate": 7.901955584865923e-06,
+      "loss": 2.3354,
+      "step": 813
+    },
+    {
+      "epoch": 0.3139826422372228,
+      "grad_norm": 0.4408669003589643,
+      "learning_rate": 7.896941009049682e-06,
+      "loss": 2.3545,
+      "step": 814
+    },
+    {
+      "epoch": 0.31436837029893927,
+      "grad_norm": 0.43867232073734636,
+      "learning_rate": 7.891922043097162e-06,
+      "loss": 2.3471,
+      "step": 815
+    },
+    {
+      "epoch": 0.31475409836065577,
+      "grad_norm": 0.41494825233536436,
+      "learning_rate": 7.886898694614292e-06,
+      "loss": 2.3706,
+      "step": 816
+    },
+    {
+      "epoch": 0.3151398264223722,
+      "grad_norm": 0.4654886458710088,
+      "learning_rate": 7.881870971213652e-06,
+      "loss": 2.3748,
+      "step": 817
+    },
+    {
+      "epoch": 0.3155255544840887,
+      "grad_norm": 0.42965298008792185,
+      "learning_rate": 7.876838880514448e-06,
+      "loss": 2.3053,
+      "step": 818
+    },
+    {
+      "epoch": 0.3159112825458052,
+      "grad_norm": 0.4838698164022045,
+      "learning_rate": 7.871802430142506e-06,
+      "loss": 2.3315,
+      "step": 819
+    },
+    {
+      "epoch": 0.3162970106075217,
+      "grad_norm": 0.42798835508259303,
+      "learning_rate": 7.866761627730253e-06,
+      "loss": 2.3773,
+      "step": 820
+    },
+    {
+      "epoch": 0.3166827386692382,
+      "grad_norm": 0.47092082650843975,
+      "learning_rate": 7.86171648091672e-06,
+      "loss": 2.3291,
+      "step": 821
+    },
+    {
+      "epoch": 0.3170684667309547,
+      "grad_norm": 0.458184182466206,
+      "learning_rate": 7.856666997347515e-06,
+      "loss": 2.3083,
+      "step": 822
+    },
+    {
+      "epoch": 0.3174541947926712,
+      "grad_norm": 0.5946076914449817,
+      "learning_rate": 7.851613184674821e-06,
+      "loss": 2.4127,
+      "step": 823
+    },
+    {
+      "epoch": 0.31783992285438767,
+      "grad_norm": 0.5024372758017904,
+      "learning_rate": 7.846555050557381e-06,
+      "loss": 2.3815,
+      "step": 824
+    },
+    {
+      "epoch": 0.31822565091610416,
+      "grad_norm": 0.4388051557059938,
+      "learning_rate": 7.841492602660487e-06,
+      "loss": 2.3495,
+      "step": 825
+    },
+    {
+      "epoch": 0.31861137897782066,
+      "grad_norm": 0.4467754067676991,
+      "learning_rate": 7.836425848655968e-06,
+      "loss": 2.3293,
+      "step": 826
+    },
+    {
+      "epoch": 0.31899710703953715,
+      "grad_norm": 0.4586352536542521,
+      "learning_rate": 7.831354796222178e-06,
+      "loss": 2.372,
+      "step": 827
+    },
+    {
+      "epoch": 0.3193828351012536,
+      "grad_norm": 0.45899021692487624,
+      "learning_rate": 7.826279453043985e-06,
+      "loss": 2.3055,
+      "step": 828
+    },
+    {
+      "epoch": 0.3197685631629701,
+      "grad_norm": 0.4625457915914088,
+      "learning_rate": 7.821199826812764e-06,
+      "loss": 2.3704,
+      "step": 829
+    },
+    {
+      "epoch": 0.3201542912246866,
+      "grad_norm": 0.48422010360163636,
+      "learning_rate": 7.816115925226373e-06,
+      "loss": 2.4121,
+      "step": 830
+    },
+    {
+      "epoch": 0.3205400192864031,
+      "grad_norm": 0.4441602028094247,
+      "learning_rate": 7.811027755989153e-06,
+      "loss": 2.3545,
+      "step": 831
+    },
+    {
+      "epoch": 0.32092574734811957,
+      "grad_norm": 0.47419398203385676,
+      "learning_rate": 7.805935326811913e-06,
+      "loss": 2.3258,
+      "step": 832
+    },
+    {
+      "epoch": 0.32131147540983607,
+      "grad_norm": 0.42641661954889865,
+      "learning_rate": 7.800838645411917e-06,
+      "loss": 2.3169,
+      "step": 833
+    },
+    {
+      "epoch": 0.32169720347155256,
+      "grad_norm": 0.4545956560840867,
+      "learning_rate": 7.795737719512872e-06,
+      "loss": 2.3415,
+      "step": 834
+    },
+    {
+      "epoch": 0.32208293153326906,
+      "grad_norm": 0.4596332641427211,
+      "learning_rate": 7.79063255684492e-06,
+      "loss": 2.3337,
+      "step": 835
+    },
+    {
+      "epoch": 0.32246865959498555,
+      "grad_norm": 0.43100947353655417,
+      "learning_rate": 7.78552316514462e-06,
+      "loss": 2.3427,
+      "step": 836
+    },
+    {
+      "epoch": 0.32285438765670205,
+      "grad_norm": 0.4581439366747127,
+      "learning_rate": 7.78040955215494e-06,
+      "loss": 2.3594,
+      "step": 837
+    },
+    {
+      "epoch": 0.32324011571841854,
+      "grad_norm": 0.4395366404437616,
+      "learning_rate": 7.775291725625252e-06,
+      "loss": 2.3767,
+      "step": 838
+    },
+    {
+      "epoch": 0.323625843780135,
+      "grad_norm": 0.44458456150956316,
+      "learning_rate": 7.7701696933113e-06,
+      "loss": 2.3406,
+      "step": 839
+    },
+    {
+      "epoch": 0.3240115718418515,
+      "grad_norm": 0.47111729135319297,
+      "learning_rate": 7.765043462975217e-06,
+      "loss": 2.3604,
+      "step": 840
+    },
+    {
+      "epoch": 0.32439729990356797,
+      "grad_norm": 0.4224299456693926,
+      "learning_rate": 7.759913042385487e-06,
+      "loss": 2.3147,
+      "step": 841
+    },
+    {
+      "epoch": 0.32478302796528447,
+      "grad_norm": 0.4255486761004747,
+      "learning_rate": 7.754778439316947e-06,
+      "loss": 2.2993,
+      "step": 842
+    },
+    {
+      "epoch": 0.32516875602700096,
+      "grad_norm": 0.4250357635068921,
+      "learning_rate": 7.749639661550775e-06,
+      "loss": 2.3983,
+      "step": 843
+    },
+    {
+      "epoch": 0.32555448408871746,
+      "grad_norm": 0.44438113155091435,
+      "learning_rate": 7.744496716874472e-06,
+      "loss": 2.263,
+      "step": 844
+    },
+    {
+      "epoch": 0.32594021215043395,
+      "grad_norm": 0.46758813922495346,
+      "learning_rate": 7.739349613081854e-06,
+      "loss": 2.2812,
+      "step": 845
+    },
+    {
+      "epoch": 0.32632594021215044,
+      "grad_norm": 0.45898502931127505,
+      "learning_rate": 7.734198357973041e-06,
+      "loss": 2.3454,
+      "step": 846
+    },
+    {
+      "epoch": 0.32671166827386694,
+      "grad_norm": 0.44360162517764584,
+      "learning_rate": 7.729042959354447e-06,
+      "loss": 2.3242,
+      "step": 847
+    },
+    {
+      "epoch": 0.32709739633558343,
+      "grad_norm": 0.42832329145176784,
+      "learning_rate": 7.723883425038759e-06,
+      "loss": 2.3197,
+      "step": 848
+    },
+    {
+      "epoch": 0.32748312439729993,
+      "grad_norm": 0.44317044594073346,
+      "learning_rate": 7.718719762844935e-06,
+      "loss": 2.2811,
+      "step": 849
+    },
+    {
+      "epoch": 0.32786885245901637,
+      "grad_norm": 0.47226771226988595,
+      "learning_rate": 7.713551980598189e-06,
+      "loss": 2.3576,
+      "step": 850
+    },
+    {
+      "epoch": 0.32825458052073286,
+      "grad_norm": 0.460364931979272,
+      "learning_rate": 7.708380086129977e-06,
+      "loss": 2.3019,
+      "step": 851
+    },
+    {
+      "epoch": 0.32864030858244936,
+      "grad_norm": 0.5167468336564306,
+      "learning_rate": 7.703204087277989e-06,
+      "loss": 2.3055,
+      "step": 852
+    },
+    {
+      "epoch": 0.32902603664416585,
+      "grad_norm": 0.4936892938530032,
+      "learning_rate": 7.698023991886133e-06,
+      "loss": 2.3918,
+      "step": 853
+    },
+    {
+      "epoch": 0.32941176470588235,
+      "grad_norm": 0.4659719100731399,
+      "learning_rate": 7.692839807804522e-06,
+      "loss": 2.3606,
+      "step": 854
+    },
+    {
+      "epoch": 0.32979749276759884,
+      "grad_norm": 0.41721873745637655,
+      "learning_rate": 7.687651542889474e-06,
+      "loss": 2.3027,
+      "step": 855
+    },
+    {
+      "epoch": 0.33018322082931534,
+      "grad_norm": 0.47568708565046847,
+      "learning_rate": 7.682459205003484e-06,
+      "loss": 2.1596,
+      "step": 856
+    },
+    {
+      "epoch": 0.33056894889103183,
+      "grad_norm": 0.4212915311019565,
+      "learning_rate": 7.677262802015223e-06,
+      "loss": 2.3223,
+      "step": 857
+    },
+    {
+      "epoch": 0.33095467695274833,
+      "grad_norm": 0.41267202978176354,
+      "learning_rate": 7.672062341799516e-06,
+      "loss": 2.3424,
+      "step": 858
+    },
+    {
+      "epoch": 0.3313404050144648,
+      "grad_norm": 0.43703247499268455,
+      "learning_rate": 7.666857832237343e-06,
+      "loss": 2.351,
+      "step": 859
+    },
+    {
+      "epoch": 0.3317261330761813,
+      "grad_norm": 0.4339373605530483,
+      "learning_rate": 7.661649281215823e-06,
+      "loss": 2.2267,
+      "step": 860
+    },
+    {
+      "epoch": 0.33211186113789776,
+      "grad_norm": 0.4095000678783189,
+      "learning_rate": 7.656436696628194e-06,
+      "loss": 2.3461,
+      "step": 861
+    },
+    {
+      "epoch": 0.33249758919961425,
+      "grad_norm": 0.4052235350798568,
+      "learning_rate": 7.651220086373803e-06,
+      "loss": 2.371,
+      "step": 862
+    },
+    {
+      "epoch": 0.33288331726133075,
+      "grad_norm": 0.42506673786742166,
+      "learning_rate": 7.645999458358107e-06,
+      "loss": 2.3749,
+      "step": 863
+    },
+    {
+      "epoch": 0.33326904532304724,
+      "grad_norm": 0.4308040407483772,
+      "learning_rate": 7.640774820492647e-06,
+      "loss": 2.3551,
+      "step": 864
+    },
+    {
+      "epoch": 0.33365477338476374,
+      "grad_norm": 0.44619888704058536,
+      "learning_rate": 7.635546180695039e-06,
+      "loss": 2.3325,
+      "step": 865
+    },
+    {
+      "epoch": 0.33404050144648023,
+      "grad_norm": 0.458016446786189,
+      "learning_rate": 7.630313546888968e-06,
+      "loss": 2.3347,
+      "step": 866
+    },
+    {
+      "epoch": 0.3344262295081967,
+      "grad_norm": 0.4123412862810054,
+      "learning_rate": 7.625076927004169e-06,
+      "loss": 2.3084,
+      "step": 867
+    },
+    {
+      "epoch": 0.3348119575699132,
+      "grad_norm": 0.4705401473635554,
+      "learning_rate": 7.619836328976416e-06,
+      "loss": 2.3727,
+      "step": 868
+    },
+    {
+      "epoch": 0.3351976856316297,
+      "grad_norm": 0.4362543045922105,
+      "learning_rate": 7.614591760747516e-06,
+      "loss": 2.3146,
+      "step": 869
+    },
+    {
+      "epoch": 0.3355834136933462,
+      "grad_norm": 0.4445694236054695,
+      "learning_rate": 7.6093432302652895e-06,
+      "loss": 2.308,
+      "step": 870
+    },
+    {
+      "epoch": 0.3359691417550627,
+      "grad_norm": 0.43128765692864757,
+      "learning_rate": 7.604090745483562e-06,
+      "loss": 2.3179,
+      "step": 871
+    },
+    {
+      "epoch": 0.33635486981677915,
+      "grad_norm": 0.45804986585748414,
+      "learning_rate": 7.598834314362151e-06,
+      "loss": 2.3559,
+      "step": 872
+    },
+    {
+      "epoch": 0.33674059787849564,
+      "grad_norm": 0.43170738334665876,
+      "learning_rate": 7.593573944866857e-06,
+      "loss": 2.3143,
+      "step": 873
+    },
+    {
+      "epoch": 0.33712632594021213,
+      "grad_norm": 0.5067556063547982,
+      "learning_rate": 7.588309644969445e-06,
+      "loss": 2.3403,
+      "step": 874
+    },
+    {
+      "epoch": 0.33751205400192863,
+      "grad_norm": 0.45811233476707985,
+      "learning_rate": 7.58304142264764e-06,
+      "loss": 2.3567,
+      "step": 875
+    },
+    {
+      "epoch": 0.3378977820636451,
+      "grad_norm": 0.5042358345345198,
+      "learning_rate": 7.57776928588511e-06,
+      "loss": 2.289,
+      "step": 876
+    },
+    {
+      "epoch": 0.3382835101253616,
+      "grad_norm": 0.47536789766317805,
+      "learning_rate": 7.572493242671453e-06,
+      "loss": 2.3533,
+      "step": 877
+    },
+    {
+      "epoch": 0.3386692381870781,
+      "grad_norm": 0.4330481840781185,
+      "learning_rate": 7.567213301002189e-06,
+      "loss": 2.2871,
+      "step": 878
+    },
+    {
+      "epoch": 0.3390549662487946,
+      "grad_norm": 0.43206128149569756,
+      "learning_rate": 7.561929468878746e-06,
+      "loss": 2.3191,
+      "step": 879
+    },
+    {
+      "epoch": 0.3394406943105111,
+      "grad_norm": 0.43227398982142884,
+      "learning_rate": 7.556641754308447e-06,
+      "loss": 2.2718,
+      "step": 880
+    },
+    {
+      "epoch": 0.3398264223722276,
+      "grad_norm": 0.49571664458087944,
+      "learning_rate": 7.5513501653045e-06,
+      "loss": 2.4013,
+      "step": 881
+    },
+    {
+      "epoch": 0.3402121504339441,
+      "grad_norm": 0.45840248989677757,
+      "learning_rate": 7.546054709885981e-06,
+      "loss": 2.2785,
+      "step": 882
+    },
+    {
+      "epoch": 0.34059787849566053,
+      "grad_norm": 0.5225076487174943,
+      "learning_rate": 7.540755396077828e-06,
+      "loss": 2.3079,
+      "step": 883
+    },
+    {
+      "epoch": 0.34098360655737703,
+      "grad_norm": 0.41229766105030147,
+      "learning_rate": 7.535452231910829e-06,
+      "loss": 2.3381,
+      "step": 884
+    },
+    {
+      "epoch": 0.3413693346190935,
+      "grad_norm": 0.4652528286296387,
+      "learning_rate": 7.5301452254216e-06,
+      "loss": 2.361,
+      "step": 885
+    },
+    {
+      "epoch": 0.34175506268081,
+      "grad_norm": 0.4670802361126979,
+      "learning_rate": 7.524834384652586e-06,
+      "loss": 2.3392,
+      "step": 886
+    },
+    {
+      "epoch": 0.3421407907425265,
+      "grad_norm": 0.4408580017883797,
+      "learning_rate": 7.519519717652039e-06,
+      "loss": 2.322,
+      "step": 887
+    },
+    {
+      "epoch": 0.342526518804243,
+      "grad_norm": 0.4193533845351374,
+      "learning_rate": 7.514201232474012e-06,
+      "loss": 2.3312,
+      "step": 888
+    },
+    {
+      "epoch": 0.3429122468659595,
+      "grad_norm": 0.4159448991368155,
+      "learning_rate": 7.50887893717834e-06,
+      "loss": 2.3291,
+      "step": 889
+    },
+    {
+      "epoch": 0.343297974927676,
+      "grad_norm": 0.41481103356640897,
+      "learning_rate": 7.503552839830638e-06,
+      "loss": 2.3178,
+      "step": 890
+    },
+    {
+      "epoch": 0.3436837029893925,
+      "grad_norm": 0.4491536254248892,
+      "learning_rate": 7.498222948502277e-06,
+      "loss": 2.3412,
+      "step": 891
+    },
+    {
+      "epoch": 0.344069431051109,
+      "grad_norm": 0.43446834098269727,
+      "learning_rate": 7.492889271270382e-06,
+      "loss": 2.3726,
+      "step": 892
+    },
+    {
+      "epoch": 0.3444551591128255,
+      "grad_norm": 0.44546096046509476,
+      "learning_rate": 7.487551816217813e-06,
+      "loss": 2.2741,
+      "step": 893
+    },
+    {
+      "epoch": 0.3448408871745419,
+      "grad_norm": 0.44732584277374626,
+      "learning_rate": 7.482210591433156e-06,
+      "loss": 2.3504,
+      "step": 894
+    },
+    {
+      "epoch": 0.3452266152362584,
+      "grad_norm": 0.44113243248302264,
+      "learning_rate": 7.4768656050107065e-06,
+      "loss": 2.305,
+      "step": 895
+    },
+    {
+      "epoch": 0.3456123432979749,
+      "grad_norm": 0.48335465507229275,
+      "learning_rate": 7.471516865050468e-06,
+      "loss": 2.3396,
+      "step": 896
+    },
+    {
+      "epoch": 0.3459980713596914,
+      "grad_norm": 0.4466799510147099,
+      "learning_rate": 7.466164379658123e-06,
+      "loss": 2.3965,
+      "step": 897
+    },
+    {
+      "epoch": 0.3463837994214079,
+      "grad_norm": 0.43440164723782954,
+      "learning_rate": 7.4608081569450365e-06,
+      "loss": 2.3035,
+      "step": 898
+    },
+    {
+      "epoch": 0.3467695274831244,
+      "grad_norm": 0.4400112286896313,
+      "learning_rate": 7.455448205028238e-06,
+      "loss": 2.3103,
+      "step": 899
+    },
+    {
+      "epoch": 0.3471552555448409,
+      "grad_norm": 0.46608894618304125,
+      "learning_rate": 7.450084532030402e-06,
+      "loss": 2.2765,
+      "step": 900
+    },
+    {
+      "epoch": 0.3475409836065574,
+      "grad_norm": 0.45086681159300906,
+      "learning_rate": 7.444717146079845e-06,
+      "loss": 2.4157,
+      "step": 901
+    },
+    {
+      "epoch": 0.3479267116682739,
+      "grad_norm": 0.4478101138542285,
+      "learning_rate": 7.439346055310514e-06,
+      "loss": 2.3622,
+      "step": 902
+    },
+    {
+      "epoch": 0.3483124397299904,
+      "grad_norm": 0.41300538182631347,
+      "learning_rate": 7.433971267861966e-06,
+      "loss": 2.3724,
+      "step": 903
+    },
+    {
+      "epoch": 0.34869816779170687,
+      "grad_norm": 0.4158739751317302,
+      "learning_rate": 7.428592791879361e-06,
+      "loss": 2.3306,
+      "step": 904
+    },
+    {
+      "epoch": 0.3490838958534233,
+      "grad_norm": 0.4609002769648654,
+      "learning_rate": 7.42321063551345e-06,
+      "loss": 2.2837,
+      "step": 905
+    },
+    {
+      "epoch": 0.3494696239151398,
+      "grad_norm": 0.4252440776041672,
+      "learning_rate": 7.41782480692056e-06,
+      "loss": 2.3061,
+      "step": 906
+    },
+    {
+      "epoch": 0.3498553519768563,
+      "grad_norm": 0.48234747553660545,
+      "learning_rate": 7.412435314262585e-06,
+      "loss": 2.3373,
+      "step": 907
+    },
+    {
+      "epoch": 0.3502410800385728,
+      "grad_norm": 0.4716302401394498,
+      "learning_rate": 7.407042165706969e-06,
+      "loss": 2.32,
+      "step": 908
+    },
+    {
+      "epoch": 0.3506268081002893,
+      "grad_norm": 0.49459850070567807,
+      "learning_rate": 7.401645369426697e-06,
+      "loss": 2.3398,
+      "step": 909
+    },
+    {
+      "epoch": 0.3510125361620058,
+      "grad_norm": 0.41483151543671365,
+      "learning_rate": 7.396244933600285e-06,
+      "loss": 2.3372,
+      "step": 910
+    },
+    {
+      "epoch": 0.3513982642237223,
+      "grad_norm": 0.4368914093246457,
+      "learning_rate": 7.390840866411759e-06,
+      "loss": 2.3382,
+      "step": 911
+    },
+    {
+      "epoch": 0.3517839922854388,
+      "grad_norm": 0.42396785049832025,
+      "learning_rate": 7.385433176050654e-06,
+      "loss": 2.3367,
+      "step": 912
+    },
+    {
+      "epoch": 0.35216972034715527,
+      "grad_norm": 0.44575975227453146,
+      "learning_rate": 7.380021870711991e-06,
+      "loss": 2.392,
+      "step": 913
+    },
+    {
+      "epoch": 0.35255544840887176,
+      "grad_norm": 0.4987776280098204,
+      "learning_rate": 7.37460695859627e-06,
+      "loss": 2.3795,
+      "step": 914
+    },
+    {
+      "epoch": 0.35294117647058826,
+      "grad_norm": 0.4463947309244009,
+      "learning_rate": 7.369188447909459e-06,
+      "loss": 2.3268,
+      "step": 915
+    },
+    {
+      "epoch": 0.3533269045323047,
+      "grad_norm": 0.41673276875071985,
+      "learning_rate": 7.36376634686298e-06,
+      "loss": 2.3676,
+      "step": 916
+    },
+    {
+      "epoch": 0.3537126325940212,
+      "grad_norm": 0.44534856324225514,
+      "learning_rate": 7.358340663673695e-06,
+      "loss": 2.3705,
+      "step": 917
+    },
+    {
+      "epoch": 0.3540983606557377,
+      "grad_norm": 0.460255688689065,
+      "learning_rate": 7.352911406563888e-06,
+      "loss": 2.322,
+      "step": 918
+    },
+    {
+      "epoch": 0.3544840887174542,
+      "grad_norm": 0.43611753646963614,
+      "learning_rate": 7.347478583761272e-06,
+      "loss": 2.2735,
+      "step": 919
+    },
+    {
+      "epoch": 0.3548698167791707,
+      "grad_norm": 0.43948537974215407,
+      "learning_rate": 7.342042203498952e-06,
+      "loss": 2.319,
+      "step": 920
+    },
+    {
+      "epoch": 0.35525554484088717,
+      "grad_norm": 0.45940409785850245,
+      "learning_rate": 7.3366022740154285e-06,
+      "loss": 2.3538,
+      "step": 921
+    },
+    {
+      "epoch": 0.35564127290260367,
+      "grad_norm": 0.46162875647189644,
+      "learning_rate": 7.331158803554585e-06,
+      "loss": 2.3226,
+      "step": 922
+    },
+    {
+      "epoch": 0.35602700096432016,
+      "grad_norm": 0.4368969469152973,
+      "learning_rate": 7.325711800365662e-06,
+      "loss": 2.3425,
+      "step": 923
+    },
+    {
+      "epoch": 0.35641272902603666,
+      "grad_norm": 0.4591835143995513,
+      "learning_rate": 7.320261272703259e-06,
+      "loss": 2.3363,
+      "step": 924
+    },
+    {
+      "epoch": 0.35679845708775315,
+      "grad_norm": 0.4760097133592221,
+      "learning_rate": 7.31480722882732e-06,
+      "loss": 2.3522,
+      "step": 925
+    },
+    {
+      "epoch": 0.35718418514946965,
+      "grad_norm": 0.46169731625643,
+      "learning_rate": 7.309349677003111e-06,
+      "loss": 2.3759,
+      "step": 926
+    },
+    {
+      "epoch": 0.35756991321118614,
+      "grad_norm": 0.47075847741995125,
+      "learning_rate": 7.303888625501217e-06,
+      "loss": 2.3803,
+      "step": 927
+    },
+    {
+      "epoch": 0.3579556412729026,
+      "grad_norm": 0.44926789399658806,
+      "learning_rate": 7.298424082597526e-06,
+      "loss": 2.3828,
+      "step": 928
+    },
+    {
+      "epoch": 0.3583413693346191,
+      "grad_norm": 0.47082834920460115,
+      "learning_rate": 7.292956056573217e-06,
+      "loss": 2.326,
+      "step": 929
+    },
+    {
+      "epoch": 0.35872709739633557,
+      "grad_norm": 0.43121536673404864,
+      "learning_rate": 7.28748455571475e-06,
+      "loss": 2.3573,
+      "step": 930
+    },
+    {
+      "epoch": 0.35911282545805207,
+      "grad_norm": 0.49565820410327954,
+      "learning_rate": 7.2820095883138456e-06,
+      "loss": 2.3946,
+      "step": 931
+    },
+    {
+      "epoch": 0.35949855351976856,
+      "grad_norm": 0.4462732434716897,
+      "learning_rate": 7.276531162667484e-06,
+      "loss": 2.3244,
+      "step": 932
+    },
+    {
+      "epoch": 0.35988428158148505,
+      "grad_norm": 0.5351326609547389,
+      "learning_rate": 7.271049287077881e-06,
+      "loss": 2.2503,
+      "step": 933
+    },
+    {
+      "epoch": 0.36027000964320155,
+      "grad_norm": 0.4590061951597883,
+      "learning_rate": 7.265563969852482e-06,
+      "loss": 2.318,
+      "step": 934
+    },
+    {
+      "epoch": 0.36065573770491804,
+      "grad_norm": 0.4319454825999862,
+      "learning_rate": 7.260075219303951e-06,
+      "loss": 2.3753,
+      "step": 935
+    },
+    {
+      "epoch": 0.36104146576663454,
+      "grad_norm": 0.41309154995577047,
+      "learning_rate": 7.254583043750152e-06,
+      "loss": 2.3311,
+      "step": 936
+    },
+    {
+      "epoch": 0.36142719382835103,
+      "grad_norm": 0.5407552089493747,
+      "learning_rate": 7.249087451514137e-06,
+      "loss": 2.2955,
+      "step": 937
+    },
+    {
+      "epoch": 0.36181292189006753,
+      "grad_norm": 0.4492520741093502,
+      "learning_rate": 7.243588450924142e-06,
+      "loss": 2.2689,
+      "step": 938
+    },
+    {
+      "epoch": 0.36219864995178397,
+      "grad_norm": 0.4493553262474411,
+      "learning_rate": 7.238086050313563e-06,
+      "loss": 2.2745,
+      "step": 939
+    },
+    {
+      "epoch": 0.36258437801350046,
+      "grad_norm": 0.42785233193380856,
+      "learning_rate": 7.232580258020952e-06,
+      "loss": 2.277,
+      "step": 940
+    },
+    {
+      "epoch": 0.36297010607521696,
+      "grad_norm": 0.43432059382700255,
+      "learning_rate": 7.227071082389998e-06,
+      "loss": 2.3192,
+      "step": 941
+    },
+    {
+      "epoch": 0.36335583413693345,
+      "grad_norm": 0.4330307978504909,
+      "learning_rate": 7.221558531769519e-06,
+      "loss": 2.2972,
+      "step": 942
+    },
+    {
+      "epoch": 0.36374156219864995,
+      "grad_norm": 0.41452292523779316,
+      "learning_rate": 7.216042614513446e-06,
+      "loss": 2.302,
+      "step": 943
+    },
+    {
+      "epoch": 0.36412729026036644,
+      "grad_norm": 0.4538549860302695,
+      "learning_rate": 7.210523338980814e-06,
+      "loss": 2.3121,
+      "step": 944
+    },
+    {
+      "epoch": 0.36451301832208294,
+      "grad_norm": 0.4138770978200875,
+      "learning_rate": 7.205000713535748e-06,
+      "loss": 2.3459,
+      "step": 945
+    },
+    {
+      "epoch": 0.36489874638379943,
+      "grad_norm": 0.4120360593592124,
+      "learning_rate": 7.199474746547445e-06,
+      "loss": 2.2938,
+      "step": 946
+    },
+    {
+      "epoch": 0.3652844744455159,
+      "grad_norm": 0.4247085637801656,
+      "learning_rate": 7.193945446390169e-06,
+      "loss": 2.3401,
+      "step": 947
+    },
+    {
+      "epoch": 0.3656702025072324,
+      "grad_norm": 0.41971540037244914,
+      "learning_rate": 7.1884128214432366e-06,
+      "loss": 2.3172,
+      "step": 948
+    },
+    {
+      "epoch": 0.3660559305689489,
+      "grad_norm": 0.47931566387475316,
+      "learning_rate": 7.182876880091001e-06,
+      "loss": 2.3714,
+      "step": 949
+    },
+    {
+      "epoch": 0.36644165863066536,
+      "grad_norm": 0.4609317408192823,
+      "learning_rate": 7.17733763072284e-06,
+      "loss": 2.3339,
+      "step": 950
+    },
+    {
+      "epoch": 0.36682738669238185,
+      "grad_norm": 0.43813545970177914,
+      "learning_rate": 7.171795081733149e-06,
+      "loss": 2.3266,
+      "step": 951
+    },
+    {
+      "epoch": 0.36721311475409835,
+      "grad_norm": 0.4354158690725303,
+      "learning_rate": 7.1662492415213194e-06,
+      "loss": 2.3486,
+      "step": 952
+    },
+    {
+      "epoch": 0.36759884281581484,
+      "grad_norm": 0.4553878583964571,
+      "learning_rate": 7.160700118491729e-06,
+      "loss": 2.3382,
+      "step": 953
+    },
+    {
+      "epoch": 0.36798457087753134,
+      "grad_norm": 0.3973964395326932,
+      "learning_rate": 7.155147721053736e-06,
+      "loss": 2.2795,
+      "step": 954
+    },
+    {
+      "epoch": 0.36837029893924783,
+      "grad_norm": 0.43994093319418587,
+      "learning_rate": 7.149592057621657e-06,
+      "loss": 2.3168,
+      "step": 955
+    },
+    {
+      "epoch": 0.3687560270009643,
+      "grad_norm": 0.4318978015975414,
+      "learning_rate": 7.14403313661476e-06,
+      "loss": 2.3598,
+      "step": 956
+    },
+    {
+      "epoch": 0.3691417550626808,
+      "grad_norm": 0.44106364168307033,
+      "learning_rate": 7.138470966457247e-06,
+      "loss": 2.3031,
+      "step": 957
+    },
+    {
+      "epoch": 0.3695274831243973,
+      "grad_norm": 0.5289909050360543,
+      "learning_rate": 7.1329055555782455e-06,
+      "loss": 2.3867,
+      "step": 958
+    },
+    {
+      "epoch": 0.3699132111861138,
+      "grad_norm": 0.44511966109041345,
+      "learning_rate": 7.127336912411796e-06,
+      "loss": 2.3054,
+      "step": 959
+    },
+    {
+      "epoch": 0.3702989392478303,
+      "grad_norm": 0.4484647822924042,
+      "learning_rate": 7.1217650453968335e-06,
+      "loss": 2.327,
+      "step": 960
+    },
+    {
+      "epoch": 0.37068466730954674,
+      "grad_norm": 0.43351059716778734,
+      "learning_rate": 7.116189962977182e-06,
+      "loss": 2.3268,
+      "step": 961
+    },
+    {
+      "epoch": 0.37107039537126324,
+      "grad_norm": 0.4477952358364692,
+      "learning_rate": 7.110611673601534e-06,
+      "loss": 2.2614,
+      "step": 962
+    },
+    {
+      "epoch": 0.37145612343297973,
+      "grad_norm": 0.41218857307769957,
+      "learning_rate": 7.105030185723447e-06,
+      "loss": 2.2849,
+      "step": 963
+    },
+    {
+      "epoch": 0.37184185149469623,
+      "grad_norm": 0.45707003023428944,
+      "learning_rate": 7.099445507801324e-06,
+      "loss": 2.4215,
+      "step": 964
+    },
+    {
+      "epoch": 0.3722275795564127,
+      "grad_norm": 0.4653349176520788,
+      "learning_rate": 7.093857648298399e-06,
+      "loss": 2.3859,
+      "step": 965
+    },
+    {
+      "epoch": 0.3726133076181292,
+      "grad_norm": 0.49872774030691847,
+      "learning_rate": 7.0882666156827315e-06,
+      "loss": 2.3693,
+      "step": 966
+    },
+    {
+      "epoch": 0.3729990356798457,
+      "grad_norm": 0.424303126549995,
+      "learning_rate": 7.082672418427189e-06,
+      "loss": 2.224,
+      "step": 967
+    },
+    {
+      "epoch": 0.3733847637415622,
+      "grad_norm": 0.45802316947265115,
+      "learning_rate": 7.0770750650094335e-06,
+      "loss": 2.3169,
+      "step": 968
+    },
+    {
+      "epoch": 0.3737704918032787,
+      "grad_norm": 0.45558049307985826,
+      "learning_rate": 7.07147456391191e-06,
+      "loss": 2.3528,
+      "step": 969
+    },
+    {
+      "epoch": 0.3741562198649952,
+      "grad_norm": 0.41730404131935284,
+      "learning_rate": 7.065870923621832e-06,
+      "loss": 2.3625,
+      "step": 970
+    },
+    {
+      "epoch": 0.3745419479267117,
+      "grad_norm": 0.44765780350037143,
+      "learning_rate": 7.060264152631178e-06,
+      "loss": 2.3049,
+      "step": 971
+    },
+    {
+      "epoch": 0.37492767598842813,
+      "grad_norm": 0.427541289342369,
+      "learning_rate": 7.0546542594366605e-06,
+      "loss": 2.3661,
+      "step": 972
+    },
+    {
+      "epoch": 0.37531340405014463,
+      "grad_norm": 0.457471287514202,
+      "learning_rate": 7.04904125253973e-06,
+      "loss": 2.3248,
+      "step": 973
+    },
+    {
+      "epoch": 0.3756991321118611,
+      "grad_norm": 0.42521854187642844,
+      "learning_rate": 7.0434251404465536e-06,
+      "loss": 2.3568,
+      "step": 974
+    },
+    {
+      "epoch": 0.3760848601735776,
+      "grad_norm": 0.46678355766392204,
+      "learning_rate": 7.037805931668006e-06,
+      "loss": 2.2357,
+      "step": 975
+    },
+    {
+      "epoch": 0.3764705882352941,
+      "grad_norm": 0.4351826329091472,
+      "learning_rate": 7.03218363471965e-06,
+      "loss": 2.2827,
+      "step": 976
+    },
+    {
+      "epoch": 0.3768563162970106,
+      "grad_norm": 0.4214350713148253,
+      "learning_rate": 7.026558258121734e-06,
+      "loss": 2.3179,
+      "step": 977
+    },
+    {
+      "epoch": 0.3772420443587271,
+      "grad_norm": 0.4403774879476315,
+      "learning_rate": 7.0209298103991705e-06,
+      "loss": 2.3061,
+      "step": 978
+    },
+    {
+      "epoch": 0.3776277724204436,
+      "grad_norm": 0.45153347431122515,
+      "learning_rate": 7.015298300081527e-06,
+      "loss": 2.3657,
+      "step": 979
+    },
+    {
+      "epoch": 0.3780135004821601,
+      "grad_norm": 0.4554861232665788,
+      "learning_rate": 7.0096637357030105e-06,
+      "loss": 2.3103,
+      "step": 980
+    },
+    {
+      "epoch": 0.3783992285438766,
+      "grad_norm": 0.433555203257244,
+      "learning_rate": 7.004026125802458e-06,
+      "loss": 2.3119,
+      "step": 981
+    },
+    {
+      "epoch": 0.3787849566055931,
+      "grad_norm": 0.4376671806296453,
+      "learning_rate": 6.998385478923322e-06,
+      "loss": 2.3454,
+      "step": 982
+    },
+    {
+      "epoch": 0.3791706846673095,
+      "grad_norm": 0.46881557305357713,
+      "learning_rate": 6.992741803613654e-06,
+      "loss": 2.2832,
+      "step": 983
+    },
+    {
+      "epoch": 0.379556412729026,
+      "grad_norm": 0.399577766685155,
+      "learning_rate": 6.987095108426102e-06,
+      "loss": 2.3893,
+      "step": 984
+    },
+    {
+      "epoch": 0.3799421407907425,
+      "grad_norm": 0.42203561978643744,
+      "learning_rate": 6.981445401917883e-06,
+      "loss": 2.3288,
+      "step": 985
+    },
+    {
+      "epoch": 0.380327868852459,
+      "grad_norm": 0.4515287192066342,
+      "learning_rate": 6.975792692650778e-06,
+      "loss": 2.3583,
+      "step": 986
+    },
+    {
+      "epoch": 0.3807135969141755,
+      "grad_norm": 0.444049448296168,
+      "learning_rate": 6.970136989191125e-06,
+      "loss": 2.3105,
+      "step": 987
+    },
+    {
+      "epoch": 0.381099324975892,
+      "grad_norm": 0.46122786470791066,
+      "learning_rate": 6.964478300109796e-06,
+      "loss": 2.3619,
+      "step": 988
+    },
+    {
+      "epoch": 0.3814850530376085,
+      "grad_norm": 0.4158567496051892,
+      "learning_rate": 6.958816633982183e-06,
+      "loss": 2.3948,
+      "step": 989
+    },
+    {
+      "epoch": 0.381870781099325,
+      "grad_norm": 0.4177324575146256,
+      "learning_rate": 6.953151999388196e-06,
+      "loss": 2.4152,
+      "step": 990
+    },
+    {
+      "epoch": 0.3822565091610415,
+      "grad_norm": 0.41288361870688683,
+      "learning_rate": 6.9474844049122415e-06,
+      "loss": 2.3977,
+      "step": 991
+    },
+    {
+      "epoch": 0.382642237222758,
+      "grad_norm": 0.4491521001025632,
+      "learning_rate": 6.94181385914321e-06,
+      "loss": 2.2437,
+      "step": 992
+    },
+    {
+      "epoch": 0.38302796528447447,
+      "grad_norm": 0.4394758332260508,
+      "learning_rate": 6.936140370674465e-06,
+      "loss": 2.411,
+      "step": 993
+    },
+    {
+      "epoch": 0.3834136933461909,
+      "grad_norm": 0.4387644067430558,
+      "learning_rate": 6.930463948103833e-06,
+      "loss": 2.2535,
+      "step": 994
+    },
+    {
+      "epoch": 0.3837994214079074,
+      "grad_norm": 0.45901455967029586,
+      "learning_rate": 6.924784600033579e-06,
+      "loss": 2.2672,
+      "step": 995
+    },
+    {
+      "epoch": 0.3841851494696239,
+      "grad_norm": 0.4231729373706068,
+      "learning_rate": 6.91910233507041e-06,
+      "loss": 2.3228,
+      "step": 996
+    },
+    {
+      "epoch": 0.3845708775313404,
+      "grad_norm": 0.43524885209076375,
+      "learning_rate": 6.913417161825449e-06,
+      "loss": 2.2992,
+      "step": 997
+    },
+    {
+      "epoch": 0.3849566055930569,
+      "grad_norm": 0.4405235231336523,
+      "learning_rate": 6.907729088914228e-06,
+      "loss": 2.2838,
+      "step": 998
+    },
+    {
+      "epoch": 0.3853423336547734,
+      "grad_norm": 0.44791515209255905,
+      "learning_rate": 6.90203812495667e-06,
+      "loss": 2.3547,
+      "step": 999
+    },
+    {
+      "epoch": 0.3857280617164899,
+      "grad_norm": 0.4365039922356492,
+      "learning_rate": 6.896344278577083e-06,
+      "loss": 2.3111,
+      "step": 1000
+    },
+    {
+      "epoch": 0.3861137897782064,
+      "grad_norm": 0.43431408111788405,
+      "learning_rate": 6.890647558404144e-06,
+      "loss": 2.3078,
+      "step": 1001
+    },
+    {
+      "epoch": 0.38649951783992287,
+      "grad_norm": 0.4569136157900356,
+      "learning_rate": 6.8849479730708765e-06,
+      "loss": 2.3569,
+      "step": 1002
+    },
+    {
+      "epoch": 0.38688524590163936,
+      "grad_norm": 0.39957289809295965,
+      "learning_rate": 6.87924553121466e-06,
+      "loss": 2.3248,
+      "step": 1003
+    },
+    {
+      "epoch": 0.38727097396335586,
+      "grad_norm": 0.4372886094285002,
+      "learning_rate": 6.873540241477189e-06,
+      "loss": 2.323,
+      "step": 1004
+    },
+    {
+      "epoch": 0.3876567020250723,
+      "grad_norm": 0.43480340806009493,
+      "learning_rate": 6.867832112504482e-06,
+      "loss": 2.3657,
+      "step": 1005
+    },
+    {
+      "epoch": 0.3880424300867888,
+      "grad_norm": 0.4658185278735477,
+      "learning_rate": 6.862121152946858e-06,
+      "loss": 2.3572,
+      "step": 1006
+    },
+    {
+      "epoch": 0.3884281581485053,
+      "grad_norm": 0.4359512135855857,
+      "learning_rate": 6.856407371458927e-06,
+      "loss": 2.2816,
+      "step": 1007
+    },
+    {
+      "epoch": 0.3888138862102218,
+      "grad_norm": 0.4386396770138102,
+      "learning_rate": 6.850690776699574e-06,
+      "loss": 2.307,
+      "step": 1008
+    },
+    {
+      "epoch": 0.3891996142719383,
+      "grad_norm": 0.4324053729484432,
+      "learning_rate": 6.844971377331942e-06,
+      "loss": 2.3195,
+      "step": 1009
+    },
+    {
+      "epoch": 0.38958534233365477,
+      "grad_norm": 0.4274709230781966,
+      "learning_rate": 6.839249182023439e-06,
+      "loss": 2.2916,
+      "step": 1010
+    },
+    {
+      "epoch": 0.38997107039537127,
+      "grad_norm": 0.49161756609994206,
+      "learning_rate": 6.833524199445694e-06,
+      "loss": 2.33,
+      "step": 1011
+    },
+    {
+      "epoch": 0.39035679845708776,
+      "grad_norm": 0.4280522034863406,
+      "learning_rate": 6.8277964382745675e-06,
+      "loss": 2.286,
+      "step": 1012
+    },
+    {
+      "epoch": 0.39074252651880426,
+      "grad_norm": 0.4415005657982364,
+      "learning_rate": 6.822065907190133e-06,
+      "loss": 2.3205,
+      "step": 1013
+    },
+    {
+      "epoch": 0.39112825458052075,
+      "grad_norm": 0.4381340413937925,
+      "learning_rate": 6.816332614876655e-06,
+      "loss": 2.3975,
+      "step": 1014
+    },
+    {
+      "epoch": 0.39151398264223725,
+      "grad_norm": 0.42718477861729054,
+      "learning_rate": 6.810596570022589e-06,
+      "loss": 2.3751,
+      "step": 1015
+    },
+    {
+      "epoch": 0.3918997107039537,
+      "grad_norm": 0.41276047458090415,
+      "learning_rate": 6.804857781320558e-06,
+      "loss": 2.2694,
+      "step": 1016
+    },
+    {
+      "epoch": 0.3922854387656702,
+      "grad_norm": 0.4520249379157355,
+      "learning_rate": 6.799116257467342e-06,
+      "loss": 2.3873,
+      "step": 1017
+    },
+    {
+      "epoch": 0.3926711668273867,
+      "grad_norm": 0.5172398184768827,
+      "learning_rate": 6.79337200716387e-06,
+      "loss": 2.2735,
+      "step": 1018
+    },
+    {
+      "epoch": 0.39305689488910317,
+      "grad_norm": 0.46114562885130483,
+      "learning_rate": 6.7876250391152e-06,
+      "loss": 2.391,
+      "step": 1019
+    },
+    {
+      "epoch": 0.39344262295081966,
+      "grad_norm": 0.42465065549191094,
+      "learning_rate": 6.781875362030512e-06,
+      "loss": 2.3423,
+      "step": 1020
+    },
+    {
+      "epoch": 0.39382835101253616,
+      "grad_norm": 0.4704726760079544,
+      "learning_rate": 6.776122984623086e-06,
+      "loss": 2.3638,
+      "step": 1021
+    },
+    {
+      "epoch": 0.39421407907425265,
+      "grad_norm": 0.4369827657610331,
+      "learning_rate": 6.770367915610295e-06,
+      "loss": 2.2782,
+      "step": 1022
+    },
+    {
+      "epoch": 0.39459980713596915,
+      "grad_norm": 0.47649969787580804,
+      "learning_rate": 6.764610163713597e-06,
+      "loss": 2.3465,
+      "step": 1023
+    },
+    {
+      "epoch": 0.39498553519768564,
+      "grad_norm": 0.5111935072885732,
+      "learning_rate": 6.758849737658508e-06,
+      "loss": 2.3016,
+      "step": 1024
+    },
+    {
+      "epoch": 0.39537126325940214,
+      "grad_norm": 0.4909865341420742,
+      "learning_rate": 6.753086646174602e-06,
+      "loss": 2.2941,
+      "step": 1025
+    },
+    {
+      "epoch": 0.39575699132111863,
+      "grad_norm": 0.4606179194384059,
+      "learning_rate": 6.747320897995493e-06,
+      "loss": 2.3839,
+      "step": 1026
+    },
+    {
+      "epoch": 0.39614271938283513,
+      "grad_norm": 0.44722228600393227,
+      "learning_rate": 6.741552501858814e-06,
+      "loss": 2.3077,
+      "step": 1027
+    },
+    {
+      "epoch": 0.39652844744455157,
+      "grad_norm": 0.4364105456771137,
+      "learning_rate": 6.735781466506216e-06,
+      "loss": 2.3102,
+      "step": 1028
+    },
+    {
+      "epoch": 0.39691417550626806,
+      "grad_norm": 0.4790670655719006,
+      "learning_rate": 6.73000780068335e-06,
+      "loss": 2.4034,
+      "step": 1029
+    },
+    {
+      "epoch": 0.39729990356798456,
+      "grad_norm": 0.45100353119796904,
+      "learning_rate": 6.724231513139853e-06,
+      "loss": 2.321,
+      "step": 1030
+    },
+    {
+      "epoch": 0.39768563162970105,
+      "grad_norm": 0.44416370251594317,
+      "learning_rate": 6.718452612629333e-06,
+      "loss": 2.2442,
+      "step": 1031
+    },
+    {
+      "epoch": 0.39807135969141755,
+      "grad_norm": 0.44156998375007717,
+      "learning_rate": 6.712671107909359e-06,
+      "loss": 2.3522,
+      "step": 1032
+    },
+    {
+      "epoch": 0.39845708775313404,
+      "grad_norm": 0.4444707116798702,
+      "learning_rate": 6.706887007741445e-06,
+      "loss": 2.3001,
+      "step": 1033
+    },
+    {
+      "epoch": 0.39884281581485054,
+      "grad_norm": 0.46496092995888816,
+      "learning_rate": 6.701100320891044e-06,
+      "loss": 2.3374,
+      "step": 1034
+    },
+    {
+      "epoch": 0.39922854387656703,
+      "grad_norm": 0.4268218167377186,
+      "learning_rate": 6.69531105612752e-06,
+      "loss": 2.2871,
+      "step": 1035
+    },
+    {
+      "epoch": 0.3996142719382835,
+      "grad_norm": 0.44627263198522205,
+      "learning_rate": 6.6895192222241534e-06,
+      "loss": 2.3167,
+      "step": 1036
+    },
+    {
+      "epoch": 0.4,
+      "grad_norm": 0.4394902467570602,
+      "learning_rate": 6.683724827958108e-06,
+      "loss": 2.3015,
+      "step": 1037
+    },
+    {
+      "epoch": 0.4003857280617165,
+      "grad_norm": 0.44809256036451395,
+      "learning_rate": 6.677927882110435e-06,
+      "loss": 2.2743,
+      "step": 1038
+    },
+    {
+      "epoch": 0.40077145612343296,
+      "grad_norm": 0.4480462093293534,
+      "learning_rate": 6.672128393466051e-06,
+      "loss": 2.3038,
+      "step": 1039
+    },
+    {
+      "epoch": 0.40115718418514945,
+      "grad_norm": 0.40716720447609844,
+      "learning_rate": 6.666326370813722e-06,
+      "loss": 2.3595,
+      "step": 1040
+    },
+    {
+      "epoch": 0.40154291224686595,
+      "grad_norm": 0.3871021637643824,
+      "learning_rate": 6.66052182294606e-06,
+      "loss": 2.3098,
+      "step": 1041
+    },
+    {
+      "epoch": 0.40192864030858244,
+      "grad_norm": 0.4255994407428415,
+      "learning_rate": 6.654714758659499e-06,
+      "loss": 2.3425,
+      "step": 1042
+    },
+    {
+      "epoch": 0.40231436837029894,
+      "grad_norm": 0.41962801248829623,
+      "learning_rate": 6.648905186754292e-06,
+      "loss": 2.3371,
+      "step": 1043
+    },
+    {
+      "epoch": 0.40270009643201543,
+      "grad_norm": 0.43733012177467506,
+      "learning_rate": 6.643093116034486e-06,
+      "loss": 2.3944,
+      "step": 1044
+    },
+    {
+      "epoch": 0.4030858244937319,
+      "grad_norm": 0.4558494652602206,
+      "learning_rate": 6.637278555307915e-06,
+      "loss": 2.477,
+      "step": 1045
+    },
+    {
+      "epoch": 0.4034715525554484,
+      "grad_norm": 0.48233116424295136,
+      "learning_rate": 6.631461513386195e-06,
+      "loss": 2.337,
+      "step": 1046
+    },
+    {
+      "epoch": 0.4038572806171649,
+      "grad_norm": 0.451424760143431,
+      "learning_rate": 6.625641999084689e-06,
+      "loss": 2.3773,
+      "step": 1047
+    },
+    {
+      "epoch": 0.4042430086788814,
+      "grad_norm": 0.4144779902411331,
+      "learning_rate": 6.619820021222518e-06,
+      "loss": 2.3218,
+      "step": 1048
+    },
+    {
+      "epoch": 0.4046287367405979,
+      "grad_norm": 0.4327313055254783,
+      "learning_rate": 6.613995588622533e-06,
+      "loss": 2.3059,
+      "step": 1049
+    },
+    {
+      "epoch": 0.40501446480231434,
+      "grad_norm": 0.44106603363727004,
+      "learning_rate": 6.608168710111301e-06,
+      "loss": 2.2968,
+      "step": 1050
+    },
+    {
+      "epoch": 0.40540019286403084,
+      "grad_norm": 0.4297979184946325,
+      "learning_rate": 6.602339394519101e-06,
+      "loss": 2.3781,
+      "step": 1051
+    },
+    {
+      "epoch": 0.40578592092574733,
+      "grad_norm": 0.47658296834396163,
+      "learning_rate": 6.5965076506799e-06,
+      "loss": 2.2953,
+      "step": 1052
+    },
+    {
+      "epoch": 0.40617164898746383,
+      "grad_norm": 0.41497031845158716,
+      "learning_rate": 6.590673487431352e-06,
+      "loss": 2.3416,
+      "step": 1053
+    },
+    {
+      "epoch": 0.4065573770491803,
+      "grad_norm": 0.4179010618522049,
+      "learning_rate": 6.584836913614769e-06,
+      "loss": 2.3061,
+      "step": 1054
+    },
+    {
+      "epoch": 0.4069431051108968,
+      "grad_norm": 0.4500351290200825,
+      "learning_rate": 6.578997938075126e-06,
+      "loss": 2.2873,
+      "step": 1055
+    },
+    {
+      "epoch": 0.4073288331726133,
+      "grad_norm": 0.4432168262062113,
+      "learning_rate": 6.573156569661026e-06,
+      "loss": 2.3566,
+      "step": 1056
+    },
+    {
+      "epoch": 0.4077145612343298,
+      "grad_norm": 0.4326865514231597,
+      "learning_rate": 6.567312817224707e-06,
+      "loss": 2.2927,
+      "step": 1057
+    },
+    {
+      "epoch": 0.4081002892960463,
+      "grad_norm": 0.439714361044198,
+      "learning_rate": 6.561466689622018e-06,
+      "loss": 2.3334,
+      "step": 1058
+    },
+    {
+      "epoch": 0.4084860173577628,
+      "grad_norm": 0.4333011032218123,
+      "learning_rate": 6.555618195712405e-06,
+      "loss": 2.2632,
+      "step": 1059
+    },
+    {
+      "epoch": 0.4088717454194793,
+      "grad_norm": 0.45387751813644217,
+      "learning_rate": 6.549767344358903e-06,
+      "loss": 2.3249,
+      "step": 1060
+    },
+    {
+      "epoch": 0.40925747348119573,
+      "grad_norm": 0.4434742842228654,
+      "learning_rate": 6.543914144428114e-06,
+      "loss": 2.337,
+      "step": 1061
+    },
+    {
+      "epoch": 0.4096432015429122,
+      "grad_norm": 0.447638566703459,
+      "learning_rate": 6.538058604790209e-06,
+      "loss": 2.3462,
+      "step": 1062
+    },
+    {
+      "epoch": 0.4100289296046287,
+      "grad_norm": 0.4790848677092544,
+      "learning_rate": 6.532200734318896e-06,
+      "loss": 2.3278,
+      "step": 1063
+    },
+    {
+      "epoch": 0.4104146576663452,
+      "grad_norm": 0.4827492802460279,
+      "learning_rate": 6.526340541891418e-06,
+      "loss": 2.2715,
+      "step": 1064
+    },
+    {
+      "epoch": 0.4108003857280617,
+      "grad_norm": 0.4257176094164772,
+      "learning_rate": 6.5204780363885374e-06,
+      "loss": 2.3327,
+      "step": 1065
+    },
+    {
+      "epoch": 0.4111861137897782,
+      "grad_norm": 0.43133813252324593,
+      "learning_rate": 6.514613226694522e-06,
+      "loss": 2.4307,
+      "step": 1066
+    },
+    {
+      "epoch": 0.4115718418514947,
+      "grad_norm": 0.48643274699614625,
+      "learning_rate": 6.508746121697129e-06,
+      "loss": 2.3788,
+      "step": 1067
+    },
+    {
+      "epoch": 0.4119575699132112,
+      "grad_norm": 0.4402660949629646,
+      "learning_rate": 6.5028767302875974e-06,
+      "loss": 2.3179,
+      "step": 1068
+    },
+    {
+      "epoch": 0.4123432979749277,
+      "grad_norm": 0.44133891806353254,
+      "learning_rate": 6.4970050613606305e-06,
+      "loss": 2.2797,
+      "step": 1069
+    },
+    {
+      "epoch": 0.4127290260366442,
+      "grad_norm": 0.4394619756368342,
+      "learning_rate": 6.491131123814379e-06,
+      "loss": 2.3056,
+      "step": 1070
+    },
+    {
+      "epoch": 0.4131147540983607,
+      "grad_norm": 0.4615846772059102,
+      "learning_rate": 6.485254926550438e-06,
+      "loss": 2.3564,
+      "step": 1071
+    },
+    {
+      "epoch": 0.4135004821600771,
+      "grad_norm": 0.4450207637638467,
+      "learning_rate": 6.479376478473822e-06,
+      "loss": 2.3143,
+      "step": 1072
+    },
+    {
+      "epoch": 0.4138862102217936,
+      "grad_norm": 0.4340102860306911,
+      "learning_rate": 6.473495788492961e-06,
+      "loss": 2.3714,
+      "step": 1073
+    },
+    {
+      "epoch": 0.4142719382835101,
+      "grad_norm": 0.443586757260675,
+      "learning_rate": 6.467612865519674e-06,
+      "loss": 2.3184,
+      "step": 1074
+    },
+    {
+      "epoch": 0.4146576663452266,
+      "grad_norm": 0.42694356262634586,
+      "learning_rate": 6.461727718469175e-06,
+      "loss": 2.2255,
+      "step": 1075
+    },
+    {
+      "epoch": 0.4150433944069431,
+      "grad_norm": 0.4516396805934383,
+      "learning_rate": 6.455840356260041e-06,
+      "loss": 2.315,
+      "step": 1076
+    },
+    {
+      "epoch": 0.4154291224686596,
+      "grad_norm": 0.43394672178411253,
+      "learning_rate": 6.449950787814207e-06,
+      "loss": 2.3639,
+      "step": 1077
+    },
+    {
+      "epoch": 0.4158148505303761,
+      "grad_norm": 0.4443342157979516,
+      "learning_rate": 6.444059022056957e-06,
+      "loss": 2.3522,
+      "step": 1078
+    },
+    {
+      "epoch": 0.4162005785920926,
+      "grad_norm": 0.4204045067858496,
+      "learning_rate": 6.438165067916895e-06,
+      "loss": 2.3528,
+      "step": 1079
+    },
+    {
+      "epoch": 0.4165863066538091,
+      "grad_norm": 0.43346000237735166,
+      "learning_rate": 6.432268934325947e-06,
+      "loss": 2.3276,
+      "step": 1080
+    },
+    {
+      "epoch": 0.4169720347155256,
+      "grad_norm": 0.4203595622350257,
+      "learning_rate": 6.4263706302193455e-06,
+      "loss": 2.3096,
+      "step": 1081
+    },
+    {
+      "epoch": 0.41735776277724207,
+      "grad_norm": 0.442932032351111,
+      "learning_rate": 6.420470164535606e-06,
+      "loss": 2.3486,
+      "step": 1082
+    },
+    {
+      "epoch": 0.4177434908389585,
+      "grad_norm": 0.43755358770487474,
+      "learning_rate": 6.414567546216522e-06,
+      "loss": 2.3114,
+      "step": 1083
+    },
+    {
+      "epoch": 0.418129218900675,
+      "grad_norm": 0.45867286880559605,
+      "learning_rate": 6.408662784207149e-06,
+      "loss": 2.3757,
+      "step": 1084
+    },
+    {
+      "epoch": 0.4185149469623915,
+      "grad_norm": 0.4549111483968304,
+      "learning_rate": 6.402755887455792e-06,
+      "loss": 2.3097,
+      "step": 1085
+    },
+    {
+      "epoch": 0.418900675024108,
+      "grad_norm": 0.4190428005930495,
+      "learning_rate": 6.396846864913992e-06,
+      "loss": 2.374,
+      "step": 1086
+    },
+    {
+      "epoch": 0.4192864030858245,
+      "grad_norm": 0.4388628337078911,
+      "learning_rate": 6.390935725536506e-06,
+      "loss": 2.2961,
+      "step": 1087
+    },
+    {
+      "epoch": 0.419672131147541,
+      "grad_norm": 0.44117997063904274,
+      "learning_rate": 6.385022478281307e-06,
+      "loss": 2.3508,
+      "step": 1088
+    },
+    {
+      "epoch": 0.4200578592092575,
+      "grad_norm": 0.42126577551344346,
+      "learning_rate": 6.379107132109556e-06,
+      "loss": 2.3741,
+      "step": 1089
+    },
+    {
+      "epoch": 0.420443587270974,
+      "grad_norm": 0.43608915227968786,
+      "learning_rate": 6.3731896959855955e-06,
+      "loss": 2.3676,
+      "step": 1090
+    },
+    {
+      "epoch": 0.42082931533269047,
+      "grad_norm": 0.4468702780714682,
+      "learning_rate": 6.367270178876941e-06,
+      "loss": 2.3011,
+      "step": 1091
+    },
+    {
+      "epoch": 0.42121504339440696,
+      "grad_norm": 0.4604216720014819,
+      "learning_rate": 6.361348589754255e-06,
+      "loss": 2.2769,
+      "step": 1092
+    },
+    {
+      "epoch": 0.42160077145612346,
+      "grad_norm": 0.4332754651757432,
+      "learning_rate": 6.355424937591341e-06,
+      "loss": 2.3619,
+      "step": 1093
+    },
+    {
+      "epoch": 0.4219864995178399,
+      "grad_norm": 0.4590925298486195,
+      "learning_rate": 6.349499231365132e-06,
+      "loss": 2.2536,
+      "step": 1094
+    },
+    {
+      "epoch": 0.4223722275795564,
+      "grad_norm": 0.4497533415984344,
+      "learning_rate": 6.3435714800556725e-06,
+      "loss": 2.2562,
+      "step": 1095
+    },
+    {
+      "epoch": 0.4227579556412729,
+      "grad_norm": 0.4656021835189469,
+      "learning_rate": 6.337641692646106e-06,
+      "loss": 2.2969,
+      "step": 1096
+    },
+    {
+      "epoch": 0.4231436837029894,
+      "grad_norm": 0.4642208333188446,
+      "learning_rate": 6.331709878122658e-06,
+      "loss": 2.3359,
+      "step": 1097
+    },
+    {
+      "epoch": 0.4235294117647059,
+      "grad_norm": 0.4324865356121325,
+      "learning_rate": 6.325776045474632e-06,
+      "loss": 2.309,
+      "step": 1098
+    },
+    {
+      "epoch": 0.42391513982642237,
+      "grad_norm": 0.4431490537166105,
+      "learning_rate": 6.319840203694388e-06,
+      "loss": 2.3479,
+      "step": 1099
+    },
+    {
+      "epoch": 0.42430086788813887,
+      "grad_norm": 0.44518768669815395,
+      "learning_rate": 6.313902361777327e-06,
+      "loss": 2.2904,
+      "step": 1100
+    },
+    {
+      "epoch": 0.42468659594985536,
+      "grad_norm": 0.49563019133481434,
+      "learning_rate": 6.307962528721887e-06,
+      "loss": 2.3937,
+      "step": 1101
+    },
+    {
+      "epoch": 0.42507232401157186,
+      "grad_norm": 0.47713133946510067,
+      "learning_rate": 6.3020207135295185e-06,
+      "loss": 2.3491,
+      "step": 1102
+    },
+    {
+      "epoch": 0.42545805207328835,
+      "grad_norm": 0.4011094627347639,
+      "learning_rate": 6.296076925204677e-06,
+      "loss": 2.3313,
+      "step": 1103
+    },
+    {
+      "epoch": 0.42584378013500485,
+      "grad_norm": 0.4040592127888121,
+      "learning_rate": 6.290131172754811e-06,
+      "loss": 2.3553,
+      "step": 1104
+    },
+    {
+      "epoch": 0.4262295081967213,
+      "grad_norm": 0.4473439863544359,
+      "learning_rate": 6.284183465190343e-06,
+      "loss": 2.3219,
+      "step": 1105
+    },
+    {
+      "epoch": 0.4266152362584378,
+      "grad_norm": 0.46805268291151503,
+      "learning_rate": 6.278233811524657e-06,
+      "loss": 2.3188,
+      "step": 1106
+    },
+    {
+      "epoch": 0.4270009643201543,
+      "grad_norm": 0.4597217826013756,
+      "learning_rate": 6.272282220774091e-06,
+      "loss": 2.3134,
+      "step": 1107
+    },
+    {
+      "epoch": 0.42738669238187077,
+      "grad_norm": 0.42617259659875806,
+      "learning_rate": 6.266328701957911e-06,
+      "loss": 2.3124,
+      "step": 1108
+    },
+    {
+      "epoch": 0.42777242044358726,
+      "grad_norm": 0.44021464574232927,
+      "learning_rate": 6.260373264098314e-06,
+      "loss": 2.2858,
+      "step": 1109
+    },
+    {
+      "epoch": 0.42815814850530376,
+      "grad_norm": 0.47478062965458695,
+      "learning_rate": 6.2544159162203975e-06,
+      "loss": 2.3755,
+      "step": 1110
+    },
+    {
+      "epoch": 0.42854387656702025,
+      "grad_norm": 0.4740157113934357,
+      "learning_rate": 6.248456667352158e-06,
+      "loss": 2.2967,
+      "step": 1111
+    },
+    {
+      "epoch": 0.42892960462873675,
+      "grad_norm": 0.47571505212355597,
+      "learning_rate": 6.24249552652447e-06,
+      "loss": 2.4558,
+      "step": 1112
+    },
+    {
+      "epoch": 0.42931533269045324,
+      "grad_norm": 0.5414223860738109,
+      "learning_rate": 6.236532502771078e-06,
+      "loss": 2.3475,
+      "step": 1113
+    },
+    {
+      "epoch": 0.42970106075216974,
+      "grad_norm": 0.45421679017277083,
+      "learning_rate": 6.230567605128578e-06,
+      "loss": 2.3586,
+      "step": 1114
+    },
+    {
+      "epoch": 0.43008678881388623,
+      "grad_norm": 0.4554354798531583,
+      "learning_rate": 6.2246008426364055e-06,
+      "loss": 2.4012,
+      "step": 1115
+    },
+    {
+      "epoch": 0.4304725168756027,
+      "grad_norm": 0.47175937847091437,
+      "learning_rate": 6.2186322243368236e-06,
+      "loss": 2.3741,
+      "step": 1116
+    },
+    {
+      "epoch": 0.43085824493731917,
+      "grad_norm": 0.43380626429841324,
+      "learning_rate": 6.212661759274908e-06,
+      "loss": 2.3263,
+      "step": 1117
+    },
+    {
+      "epoch": 0.43124397299903566,
+      "grad_norm": 0.4472251530815354,
+      "learning_rate": 6.206689456498529e-06,
+      "loss": 2.3582,
+      "step": 1118
+    },
+    {
+      "epoch": 0.43162970106075216,
+      "grad_norm": 0.4768444445219848,
+      "learning_rate": 6.200715325058349e-06,
+      "loss": 2.3286,
+      "step": 1119
+    },
+    {
+      "epoch": 0.43201542912246865,
+      "grad_norm": 0.4296130917301774,
+      "learning_rate": 6.194739374007792e-06,
+      "loss": 2.2701,
+      "step": 1120
+    },
+    {
+      "epoch": 0.43240115718418515,
+      "grad_norm": 0.43050510001729003,
+      "learning_rate": 6.1887616124030505e-06,
+      "loss": 2.3901,
+      "step": 1121
+    },
+    {
+      "epoch": 0.43278688524590164,
+      "grad_norm": 0.42942575073230405,
+      "learning_rate": 6.182782049303051e-06,
+      "loss": 2.3512,
+      "step": 1122
+    },
+    {
+      "epoch": 0.43317261330761814,
+      "grad_norm": 0.47879872044883814,
+      "learning_rate": 6.176800693769457e-06,
+      "loss": 2.2915,
+      "step": 1123
+    },
+    {
+      "epoch": 0.43355834136933463,
+      "grad_norm": 0.448815528010462,
+      "learning_rate": 6.170817554866646e-06,
+      "loss": 2.2555,
+      "step": 1124
+    },
+    {
+      "epoch": 0.4339440694310511,
+      "grad_norm": 0.4227739666249548,
+      "learning_rate": 6.164832641661698e-06,
+      "loss": 2.3306,
+      "step": 1125
+    },
+    {
+      "epoch": 0.4343297974927676,
+      "grad_norm": 0.4485330017422935,
+      "learning_rate": 6.158845963224377e-06,
+      "loss": 2.3105,
+      "step": 1126
+    },
+    {
+      "epoch": 0.43471552555448406,
+      "grad_norm": 0.4584878233301734,
+      "learning_rate": 6.1528575286271306e-06,
+      "loss": 2.2823,
+      "step": 1127
+    },
+    {
+      "epoch": 0.43510125361620056,
+      "grad_norm": 0.4488282278195123,
+      "learning_rate": 6.1468673469450655e-06,
+      "loss": 2.4241,
+      "step": 1128
+    },
+    {
+      "epoch": 0.43548698167791705,
+      "grad_norm": 0.4466795628436329,
+      "learning_rate": 6.14087542725593e-06,
+      "loss": 2.3984,
+      "step": 1129
+    },
+    {
+      "epoch": 0.43587270973963355,
+      "grad_norm": 0.46895294939856697,
+      "learning_rate": 6.134881778640115e-06,
+      "loss": 2.4015,
+      "step": 1130
+    },
+    {
+      "epoch": 0.43625843780135004,
+      "grad_norm": 0.4389431896841497,
+      "learning_rate": 6.1288864101806225e-06,
+      "loss": 2.3532,
+      "step": 1131
+    },
+    {
+      "epoch": 0.43664416586306654,
+      "grad_norm": 0.4506338770546297,
+      "learning_rate": 6.122889330963069e-06,
+      "loss": 2.312,
+      "step": 1132
+    },
+    {
+      "epoch": 0.43702989392478303,
+      "grad_norm": 0.43919560541539265,
+      "learning_rate": 6.116890550075658e-06,
+      "loss": 2.3842,
+      "step": 1133
+    },
+    {
+      "epoch": 0.4374156219864995,
+      "grad_norm": 0.4520050253783449,
+      "learning_rate": 6.110890076609175e-06,
+      "loss": 2.3763,
+      "step": 1134
+    },
+    {
+      "epoch": 0.437801350048216,
+      "grad_norm": 0.44824681576411424,
+      "learning_rate": 6.10488791965697e-06,
+      "loss": 2.3638,
+      "step": 1135
+    },
+    {
+      "epoch": 0.4381870781099325,
+      "grad_norm": 0.4277658180056835,
+      "learning_rate": 6.098884088314938e-06,
+      "loss": 2.339,
+      "step": 1136
+    },
+    {
+      "epoch": 0.438572806171649,
+      "grad_norm": 0.4594464904112313,
+      "learning_rate": 6.092878591681525e-06,
+      "loss": 2.3469,
+      "step": 1137
+    },
+    {
+      "epoch": 0.4389585342333655,
+      "grad_norm": 0.4274399567143741,
+      "learning_rate": 6.086871438857687e-06,
+      "loss": 2.3536,
+      "step": 1138
+    },
+    {
+      "epoch": 0.43934426229508194,
+      "grad_norm": 0.4737894356640477,
+      "learning_rate": 6.080862638946896e-06,
+      "loss": 2.2823,
+      "step": 1139
+    },
+    {
+      "epoch": 0.43972999035679844,
+      "grad_norm": 0.43143990698958135,
+      "learning_rate": 6.074852201055121e-06,
+      "loss": 2.371,
+      "step": 1140
+    },
+    {
+      "epoch": 0.44011571841851493,
+      "grad_norm": 0.49600949126416094,
+      "learning_rate": 6.068840134290811e-06,
+      "loss": 2.273,
+      "step": 1141
+    },
+    {
+      "epoch": 0.44050144648023143,
+      "grad_norm": 0.422300559886733,
+      "learning_rate": 6.062826447764883e-06,
+      "loss": 2.2668,
+      "step": 1142
+    },
+    {
+      "epoch": 0.4408871745419479,
+      "grad_norm": 0.4844578511809956,
+      "learning_rate": 6.056811150590713e-06,
+      "loss": 2.3362,
+      "step": 1143
+    },
+    {
+      "epoch": 0.4412729026036644,
+      "grad_norm": 0.434657304710951,
+      "learning_rate": 6.050794251884112e-06,
+      "loss": 2.2864,
+      "step": 1144
+    },
+    {
+      "epoch": 0.4416586306653809,
+      "grad_norm": 0.44665209883883367,
+      "learning_rate": 6.044775760763321e-06,
+      "loss": 2.2931,
+      "step": 1145
+    },
+    {
+      "epoch": 0.4420443587270974,
+      "grad_norm": 0.4371263489152066,
+      "learning_rate": 6.038755686348993e-06,
+      "loss": 2.3158,
+      "step": 1146
+    },
+    {
+      "epoch": 0.4424300867888139,
+      "grad_norm": 0.4391629731269026,
+      "learning_rate": 6.032734037764184e-06,
+      "loss": 2.3269,
+      "step": 1147
+    },
+    {
+      "epoch": 0.4428158148505304,
+      "grad_norm": 0.4942247415917742,
+      "learning_rate": 6.026710824134331e-06,
+      "loss": 2.3198,
+      "step": 1148
+    },
+    {
+      "epoch": 0.4432015429122469,
+      "grad_norm": 0.4050150430809796,
+      "learning_rate": 6.020686054587244e-06,
+      "loss": 2.2747,
+      "step": 1149
+    },
+    {
+      "epoch": 0.44358727097396333,
+      "grad_norm": 0.43967979539812374,
+      "learning_rate": 6.014659738253091e-06,
+      "loss": 2.3176,
+      "step": 1150
+    },
+    {
+      "epoch": 0.4439729990356798,
+      "grad_norm": 0.4412842004726475,
+      "learning_rate": 6.008631884264387e-06,
+      "loss": 2.2611,
+      "step": 1151
+    },
+    {
+      "epoch": 0.4443587270973963,
+      "grad_norm": 0.441535243961793,
+      "learning_rate": 6.002602501755974e-06,
+      "loss": 2.2964,
+      "step": 1152
+    },
+    {
+      "epoch": 0.4447444551591128,
+      "grad_norm": 0.43134500859104097,
+      "learning_rate": 5.996571599865011e-06,
+      "loss": 2.343,
+      "step": 1153
+    },
+    {
+      "epoch": 0.4451301832208293,
+      "grad_norm": 0.43813050565346195,
+      "learning_rate": 5.9905391877309585e-06,
+      "loss": 2.346,
+      "step": 1154
+    },
+    {
+      "epoch": 0.4455159112825458,
+      "grad_norm": 0.44953540142040926,
+      "learning_rate": 5.9845052744955654e-06,
+      "loss": 2.3625,
+      "step": 1155
+    },
+    {
+      "epoch": 0.4459016393442623,
+      "grad_norm": 0.45320640228396486,
+      "learning_rate": 5.978469869302861e-06,
+      "loss": 2.355,
+      "step": 1156
+    },
+    {
+      "epoch": 0.4462873674059788,
+      "grad_norm": 0.4289413336000678,
+      "learning_rate": 5.972432981299129e-06,
+      "loss": 2.2561,
+      "step": 1157
+    },
+    {
+      "epoch": 0.4466730954676953,
+      "grad_norm": 0.42557121234458123,
+      "learning_rate": 5.9663946196329016e-06,
+      "loss": 2.3525,
+      "step": 1158
+    },
+    {
+      "epoch": 0.4470588235294118,
+      "grad_norm": 0.45419794091593085,
+      "learning_rate": 5.960354793454948e-06,
+      "loss": 2.3306,
+      "step": 1159
+    },
+    {
+      "epoch": 0.4474445515911283,
+      "grad_norm": 0.4776081570201772,
+      "learning_rate": 5.954313511918252e-06,
+      "loss": 2.379,
+      "step": 1160
+    },
+    {
+      "epoch": 0.4478302796528447,
+      "grad_norm": 0.40265412301439,
+      "learning_rate": 5.948270784178007e-06,
+      "loss": 2.3548,
+      "step": 1161
+    },
+    {
+      "epoch": 0.4482160077145612,
+      "grad_norm": 0.43486616171542647,
+      "learning_rate": 5.942226619391592e-06,
+      "loss": 2.2836,
+      "step": 1162
+    },
+    {
+      "epoch": 0.4486017357762777,
+      "grad_norm": 0.4870799701993185,
+      "learning_rate": 5.936181026718572e-06,
+      "loss": 2.3652,
+      "step": 1163
+    },
+    {
+      "epoch": 0.4489874638379942,
+      "grad_norm": 0.4255862868250421,
+      "learning_rate": 5.9301340153206685e-06,
+      "loss": 2.3503,
+      "step": 1164
+    },
+    {
+      "epoch": 0.4493731918997107,
+      "grad_norm": 0.40623416645588406,
+      "learning_rate": 5.924085594361758e-06,
+      "loss": 2.377,
+      "step": 1165
+    },
+    {
+      "epoch": 0.4497589199614272,
+      "grad_norm": 0.40894215034833015,
+      "learning_rate": 5.918035773007852e-06,
+      "loss": 2.3663,
+      "step": 1166
+    },
+    {
+      "epoch": 0.4501446480231437,
+      "grad_norm": 0.4534195425524319,
+      "learning_rate": 5.911984560427082e-06,
+      "loss": 2.3498,
+      "step": 1167
+    },
+    {
+      "epoch": 0.4505303760848602,
+      "grad_norm": 0.4173435915739408,
+      "learning_rate": 5.905931965789688e-06,
+      "loss": 2.3478,
+      "step": 1168
+    },
+    {
+      "epoch": 0.4509161041465767,
+      "grad_norm": 0.4153750621998993,
+      "learning_rate": 5.899877998268006e-06,
+      "loss": 2.393,
+      "step": 1169
+    },
+    {
+      "epoch": 0.4513018322082932,
+      "grad_norm": 0.47144396929105753,
+      "learning_rate": 5.893822667036456e-06,
+      "loss": 2.3126,
+      "step": 1170
+    },
+    {
+      "epoch": 0.45168756027000967,
+      "grad_norm": 0.4697653920570496,
+      "learning_rate": 5.887765981271518e-06,
+      "loss": 2.3625,
+      "step": 1171
+    },
+    {
+      "epoch": 0.4520732883317261,
+      "grad_norm": 0.4243931648314425,
+      "learning_rate": 5.881707950151725e-06,
+      "loss": 2.3607,
+      "step": 1172
+    },
+    {
+      "epoch": 0.4524590163934426,
+      "grad_norm": 0.41514890362349877,
+      "learning_rate": 5.875648582857655e-06,
+      "loss": 2.3081,
+      "step": 1173
+    },
+    {
+      "epoch": 0.4528447444551591,
+      "grad_norm": 0.45353219387179644,
+      "learning_rate": 5.869587888571906e-06,
+      "loss": 2.4277,
+      "step": 1174
+    },
+    {
+      "epoch": 0.4532304725168756,
+      "grad_norm": 0.3951554178998431,
+      "learning_rate": 5.863525876479088e-06,
+      "loss": 2.346,
+      "step": 1175
+    },
+    {
+      "epoch": 0.4536162005785921,
+      "grad_norm": 0.4298758994384629,
+      "learning_rate": 5.857462555765809e-06,
+      "loss": 2.3002,
+      "step": 1176
+    },
+    {
+      "epoch": 0.4540019286403086,
+      "grad_norm": 0.4437585353827277,
+      "learning_rate": 5.851397935620659e-06,
+      "loss": 2.3046,
+      "step": 1177
+    },
+    {
+      "epoch": 0.4543876567020251,
+      "grad_norm": 0.44068266378563453,
+      "learning_rate": 5.845332025234195e-06,
+      "loss": 2.3419,
+      "step": 1178
+    },
+    {
+      "epoch": 0.4547733847637416,
+      "grad_norm": 0.45015328025310364,
+      "learning_rate": 5.839264833798937e-06,
+      "loss": 2.2871,
+      "step": 1179
+    },
+    {
+      "epoch": 0.45515911282545807,
+      "grad_norm": 0.41377865869977437,
+      "learning_rate": 5.8331963705093375e-06,
+      "loss": 2.2378,
+      "step": 1180
+    },
+    {
+      "epoch": 0.45554484088717456,
+      "grad_norm": 0.4224676221449054,
+      "learning_rate": 5.82712664456178e-06,
+      "loss": 2.3075,
+      "step": 1181
+    },
+    {
+      "epoch": 0.45593056894889106,
+      "grad_norm": 0.43244460378263555,
+      "learning_rate": 5.8210556651545645e-06,
+      "loss": 2.3632,
+      "step": 1182
+    },
+    {
+      "epoch": 0.4563162970106075,
+      "grad_norm": 0.42085739790821947,
+      "learning_rate": 5.814983441487885e-06,
+      "loss": 2.3397,
+      "step": 1183
+    },
+    {
+      "epoch": 0.456702025072324,
+      "grad_norm": 0.4439530858820655,
+      "learning_rate": 5.808909982763825e-06,
+      "loss": 2.1945,
+      "step": 1184
+    },
+    {
+      "epoch": 0.4570877531340405,
+      "grad_norm": 0.4462098611435156,
+      "learning_rate": 5.802835298186337e-06,
+      "loss": 2.4003,
+      "step": 1185
+    },
+    {
+      "epoch": 0.457473481195757,
+      "grad_norm": 0.4367403521099576,
+      "learning_rate": 5.796759396961235e-06,
+      "loss": 2.3101,
+      "step": 1186
+    },
+    {
+      "epoch": 0.4578592092574735,
+      "grad_norm": 0.43024092109324674,
+      "learning_rate": 5.79068228829617e-06,
+      "loss": 2.2912,
+      "step": 1187
+    },
+    {
+      "epoch": 0.45824493731918997,
+      "grad_norm": 0.45455438629206824,
+      "learning_rate": 5.784603981400632e-06,
+      "loss": 2.394,
+      "step": 1188
+    },
+    {
+      "epoch": 0.45863066538090647,
+      "grad_norm": 0.45733608566425527,
+      "learning_rate": 5.77852448548592e-06,
+      "loss": 2.3055,
+      "step": 1189
+    },
+    {
+      "epoch": 0.45901639344262296,
+      "grad_norm": 0.4484306633332954,
+      "learning_rate": 5.772443809765138e-06,
+      "loss": 2.2913,
+      "step": 1190
+    },
+    {
+      "epoch": 0.45940212150433946,
+      "grad_norm": 0.4697565911997198,
+      "learning_rate": 5.766361963453174e-06,
+      "loss": 2.35,
+      "step": 1191
+    },
+    {
+      "epoch": 0.45978784956605595,
+      "grad_norm": 0.45054113892803993,
+      "learning_rate": 5.760278955766695e-06,
+      "loss": 2.3265,
+      "step": 1192
+    },
+    {
+      "epoch": 0.46017357762777245,
+      "grad_norm": 0.4221795155143893,
+      "learning_rate": 5.754194795924126e-06,
+      "loss": 2.2774,
+      "step": 1193
+    },
+    {
+      "epoch": 0.4605593056894889,
+      "grad_norm": 0.43959673814629013,
+      "learning_rate": 5.748109493145637e-06,
+      "loss": 2.3489,
+      "step": 1194
+    },
+    {
+      "epoch": 0.4609450337512054,
+      "grad_norm": 0.434043983494275,
+      "learning_rate": 5.742023056653131e-06,
+      "loss": 2.3185,
+      "step": 1195
+    },
+    {
+      "epoch": 0.4613307618129219,
+      "grad_norm": 0.4499995133774159,
+      "learning_rate": 5.735935495670229e-06,
+      "loss": 2.3143,
+      "step": 1196
+    },
+    {
+      "epoch": 0.46171648987463837,
+      "grad_norm": 0.43859434091634275,
+      "learning_rate": 5.7298468194222555e-06,
+      "loss": 2.2815,
+      "step": 1197
+    },
+    {
+      "epoch": 0.46210221793635486,
+      "grad_norm": 0.45098456142101523,
+      "learning_rate": 5.723757037136226e-06,
+      "loss": 2.3214,
+      "step": 1198
+    },
+    {
+      "epoch": 0.46248794599807136,
+      "grad_norm": 0.44782039911925825,
+      "learning_rate": 5.717666158040832e-06,
+      "loss": 2.3265,
+      "step": 1199
+    },
+    {
+      "epoch": 0.46287367405978785,
+      "grad_norm": 0.42669035365800917,
+      "learning_rate": 5.711574191366427e-06,
+      "loss": 2.3586,
+      "step": 1200
+    },
+    {
+      "epoch": 0.46325940212150435,
+      "grad_norm": 0.4312352059791168,
+      "learning_rate": 5.705481146345011e-06,
+      "loss": 2.3525,
+      "step": 1201
+    },
+    {
+      "epoch": 0.46364513018322084,
+      "grad_norm": 0.39976774767851975,
+      "learning_rate": 5.699387032210222e-06,
+      "loss": 2.3207,
+      "step": 1202
+    },
+    {
+      "epoch": 0.46403085824493734,
+      "grad_norm": 0.42260370567419747,
+      "learning_rate": 5.693291858197315e-06,
+      "loss": 2.3909,
+      "step": 1203
+    },
+    {
+      "epoch": 0.46441658630665383,
+      "grad_norm": 0.45382296447308357,
+      "learning_rate": 5.687195633543151e-06,
+      "loss": 2.3194,
+      "step": 1204
+    },
+    {
+      "epoch": 0.4648023143683703,
+      "grad_norm": 0.41319552901533824,
+      "learning_rate": 5.681098367486186e-06,
+      "loss": 2.3009,
+      "step": 1205
+    },
+    {
+      "epoch": 0.46518804243008677,
+      "grad_norm": 0.4328173372360925,
+      "learning_rate": 5.675000069266451e-06,
+      "loss": 2.326,
+      "step": 1206
+    },
+    {
+      "epoch": 0.46557377049180326,
+      "grad_norm": 0.4581633461649314,
+      "learning_rate": 5.6689007481255445e-06,
+      "loss": 2.3827,
+      "step": 1207
+    },
+    {
+      "epoch": 0.46595949855351976,
+      "grad_norm": 0.41910554221151053,
+      "learning_rate": 5.662800413306611e-06,
+      "loss": 2.3139,
+      "step": 1208
+    },
+    {
+      "epoch": 0.46634522661523625,
+      "grad_norm": 0.4682743739464939,
+      "learning_rate": 5.656699074054335e-06,
+      "loss": 2.2777,
+      "step": 1209
+    },
+    {
+      "epoch": 0.46673095467695275,
+      "grad_norm": 0.4471323728277224,
+      "learning_rate": 5.650596739614921e-06,
+      "loss": 2.2842,
+      "step": 1210
+    },
+    {
+      "epoch": 0.46711668273866924,
+      "grad_norm": 0.4385459014905679,
+      "learning_rate": 5.644493419236082e-06,
+      "loss": 2.3688,
+      "step": 1211
+    },
+    {
+      "epoch": 0.46750241080038574,
+      "grad_norm": 0.4661959535855007,
+      "learning_rate": 5.6383891221670275e-06,
+      "loss": 2.2435,
+      "step": 1212
+    },
+    {
+      "epoch": 0.46788813886210223,
+      "grad_norm": 0.4187724967770663,
+      "learning_rate": 5.632283857658442e-06,
+      "loss": 2.3224,
+      "step": 1213
+    },
+    {
+      "epoch": 0.4682738669238187,
+      "grad_norm": 0.4309626015977979,
+      "learning_rate": 5.626177634962482e-06,
+      "loss": 2.36,
+      "step": 1214
+    },
+    {
+      "epoch": 0.4686595949855352,
+      "grad_norm": 0.45728438707785074,
+      "learning_rate": 5.620070463332751e-06,
+      "loss": 2.3238,
+      "step": 1215
+    },
+    {
+      "epoch": 0.46904532304725166,
+      "grad_norm": 0.45588519636584973,
+      "learning_rate": 5.613962352024293e-06,
+      "loss": 2.3423,
+      "step": 1216
+    },
+    {
+      "epoch": 0.46943105110896816,
+      "grad_norm": 0.4203095541797606,
+      "learning_rate": 5.607853310293575e-06,
+      "loss": 2.3643,
+      "step": 1217
+    },
+    {
+      "epoch": 0.46981677917068465,
+      "grad_norm": 0.4428037199827808,
+      "learning_rate": 5.601743347398478e-06,
+      "loss": 2.3693,
+      "step": 1218
+    },
+    {
+      "epoch": 0.47020250723240115,
+      "grad_norm": 0.42730369062409596,
+      "learning_rate": 5.595632472598273e-06,
+      "loss": 2.2706,
+      "step": 1219
+    },
+    {
+      "epoch": 0.47058823529411764,
+      "grad_norm": 0.4847039689722017,
+      "learning_rate": 5.589520695153618e-06,
+      "loss": 2.3504,
+      "step": 1220
+    },
+    {
+      "epoch": 0.47097396335583414,
+      "grad_norm": 0.4466785201608283,
+      "learning_rate": 5.583408024326532e-06,
+      "loss": 2.3141,
+      "step": 1221
+    },
+    {
+      "epoch": 0.47135969141755063,
+      "grad_norm": 0.4644702092382401,
+      "learning_rate": 5.5772944693803975e-06,
+      "loss": 2.3732,
+      "step": 1222
+    },
+    {
+      "epoch": 0.4717454194792671,
+      "grad_norm": 0.42841728929128997,
+      "learning_rate": 5.57118003957993e-06,
+      "loss": 2.3068,
+      "step": 1223
+    },
+    {
+      "epoch": 0.4721311475409836,
+      "grad_norm": 0.43004816739382506,
+      "learning_rate": 5.5650647441911706e-06,
+      "loss": 2.2982,
+      "step": 1224
+    },
+    {
+      "epoch": 0.4725168756027001,
+      "grad_norm": 0.4377588605123363,
+      "learning_rate": 5.558948592481476e-06,
+      "loss": 2.3656,
+      "step": 1225
+    },
+    {
+      "epoch": 0.4729026036644166,
+      "grad_norm": 0.4078263685690202,
+      "learning_rate": 5.552831593719499e-06,
+      "loss": 2.3767,
+      "step": 1226
+    },
+    {
+      "epoch": 0.47328833172613305,
+      "grad_norm": 0.4374828944134357,
+      "learning_rate": 5.546713757175174e-06,
+      "loss": 2.3627,
+      "step": 1227
+    },
+    {
+      "epoch": 0.47367405978784954,
+      "grad_norm": 0.4083844779519322,
+      "learning_rate": 5.540595092119709e-06,
+      "loss": 2.2955,
+      "step": 1228
+    },
+    {
+      "epoch": 0.47405978784956604,
+      "grad_norm": 0.427796250465258,
+      "learning_rate": 5.534475607825566e-06,
+      "loss": 2.3696,
+      "step": 1229
+    },
+    {
+      "epoch": 0.47444551591128253,
+      "grad_norm": 0.4883224066551824,
+      "learning_rate": 5.5283553135664426e-06,
+      "loss": 2.3231,
+      "step": 1230
+    },
+    {
+      "epoch": 0.47483124397299903,
+      "grad_norm": 0.44997208402273214,
+      "learning_rate": 5.522234218617274e-06,
+      "loss": 2.3012,
+      "step": 1231
+    },
+    {
+      "epoch": 0.4752169720347155,
+      "grad_norm": 0.5034216061553358,
+      "learning_rate": 5.516112332254203e-06,
+      "loss": 2.3678,
+      "step": 1232
+    },
+    {
+      "epoch": 0.475602700096432,
+      "grad_norm": 0.43877172254018615,
+      "learning_rate": 5.509989663754572e-06,
+      "loss": 2.3391,
+      "step": 1233
+    },
+    {
+      "epoch": 0.4759884281581485,
+      "grad_norm": 0.4543499788909993,
+      "learning_rate": 5.503866222396907e-06,
+      "loss": 2.2994,
+      "step": 1234
+    },
+    {
+      "epoch": 0.476374156219865,
+      "grad_norm": 0.45789077129692657,
+      "learning_rate": 5.497742017460913e-06,
+      "loss": 2.3188,
+      "step": 1235
+    },
+    {
+      "epoch": 0.4767598842815815,
+      "grad_norm": 0.4560334384462107,
+      "learning_rate": 5.491617058227443e-06,
+      "loss": 2.3405,
+      "step": 1236
+    },
+    {
+      "epoch": 0.477145612343298,
+      "grad_norm": 0.43544502546982145,
+      "learning_rate": 5.485491353978496e-06,
+      "loss": 2.3168,
+      "step": 1237
+    },
+    {
+      "epoch": 0.47753134040501444,
+      "grad_norm": 0.4295577460446279,
+      "learning_rate": 5.479364913997203e-06,
+      "loss": 2.3379,
+      "step": 1238
+    },
+    {
+      "epoch": 0.47791706846673093,
+      "grad_norm": 0.46521640185997026,
+      "learning_rate": 5.473237747567805e-06,
+      "loss": 2.3228,
+      "step": 1239
+    },
+    {
+      "epoch": 0.4783027965284474,
+      "grad_norm": 0.43381577321765624,
+      "learning_rate": 5.46710986397565e-06,
+      "loss": 2.3475,
+      "step": 1240
+    },
+    {
+      "epoch": 0.4786885245901639,
+      "grad_norm": 0.4866871687241284,
+      "learning_rate": 5.460981272507168e-06,
+      "loss": 2.3004,
+      "step": 1241
+    },
+    {
+      "epoch": 0.4790742526518804,
+      "grad_norm": 0.424201002696731,
+      "learning_rate": 5.454851982449865e-06,
+      "loss": 2.2405,
+      "step": 1242
+    },
+    {
+      "epoch": 0.4794599807135969,
+      "grad_norm": 0.42804105550903804,
+      "learning_rate": 5.448722003092298e-06,
+      "loss": 2.337,
+      "step": 1243
+    },
+    {
+      "epoch": 0.4798457087753134,
+      "grad_norm": 0.4459269229444017,
+      "learning_rate": 5.442591343724081e-06,
+      "loss": 2.393,
+      "step": 1244
+    },
+    {
+      "epoch": 0.4802314368370299,
+      "grad_norm": 0.4278328610127386,
+      "learning_rate": 5.436460013635848e-06,
+      "loss": 2.2744,
+      "step": 1245
+    },
+    {
+      "epoch": 0.4806171648987464,
+      "grad_norm": 0.4463499915836363,
+      "learning_rate": 5.430328022119255e-06,
+      "loss": 2.3151,
+      "step": 1246
+    },
+    {
+      "epoch": 0.4810028929604629,
+      "grad_norm": 0.4356287393133619,
+      "learning_rate": 5.42419537846696e-06,
+      "loss": 2.2979,
+      "step": 1247
+    },
+    {
+      "epoch": 0.4813886210221794,
+      "grad_norm": 0.4507469977004841,
+      "learning_rate": 5.418062091972604e-06,
+      "loss": 2.3654,
+      "step": 1248
+    },
+    {
+      "epoch": 0.4817743490838959,
+      "grad_norm": 0.43202360474966794,
+      "learning_rate": 5.411928171930812e-06,
+      "loss": 2.3442,
+      "step": 1249
+    },
+    {
+      "epoch": 0.4821600771456123,
+      "grad_norm": 0.4556106440128594,
+      "learning_rate": 5.405793627637157e-06,
+      "loss": 2.3677,
+      "step": 1250
+    },
+    {
+      "epoch": 0.4825458052073288,
+      "grad_norm": 0.4481360098922905,
+      "learning_rate": 5.399658468388169e-06,
+      "loss": 2.3548,
+      "step": 1251
+    },
+    {
+      "epoch": 0.4829315332690453,
+      "grad_norm": 0.4255551311824385,
+      "learning_rate": 5.393522703481303e-06,
+      "loss": 2.3924,
+      "step": 1252
+    },
+    {
+      "epoch": 0.4833172613307618,
+      "grad_norm": 0.43423503978903033,
+      "learning_rate": 5.387386342214934e-06,
+      "loss": 2.302,
+      "step": 1253
+    },
+    {
+      "epoch": 0.4837029893924783,
+      "grad_norm": 0.45181001802914283,
+      "learning_rate": 5.381249393888344e-06,
+      "loss": 2.3078,
+      "step": 1254
+    },
+    {
+      "epoch": 0.4840887174541948,
+      "grad_norm": 0.4141462355372266,
+      "learning_rate": 5.375111867801698e-06,
+      "loss": 2.3217,
+      "step": 1255
+    },
+    {
+      "epoch": 0.4844744455159113,
+      "grad_norm": 0.41247263365108455,
+      "learning_rate": 5.36897377325604e-06,
+      "loss": 2.3347,
+      "step": 1256
+    },
+    {
+      "epoch": 0.4848601735776278,
+      "grad_norm": 0.4523684838820164,
+      "learning_rate": 5.362835119553278e-06,
+      "loss": 2.3655,
+      "step": 1257
+    },
+    {
+      "epoch": 0.4852459016393443,
+      "grad_norm": 0.45940229024333956,
+      "learning_rate": 5.356695915996162e-06,
+      "loss": 2.3997,
+      "step": 1258
+    },
+    {
+      "epoch": 0.4856316297010608,
+      "grad_norm": 0.45153234599778547,
+      "learning_rate": 5.350556171888281e-06,
+      "loss": 2.3613,
+      "step": 1259
+    },
+    {
+      "epoch": 0.48601735776277727,
+      "grad_norm": 0.46289662017840116,
+      "learning_rate": 5.344415896534039e-06,
+      "loss": 2.3734,
+      "step": 1260
+    },
+    {
+      "epoch": 0.4864030858244937,
+      "grad_norm": 0.4176290117254934,
+      "learning_rate": 5.338275099238647e-06,
+      "loss": 2.3401,
+      "step": 1261
+    },
+    {
+      "epoch": 0.4867888138862102,
+      "grad_norm": 0.45879286317120854,
+      "learning_rate": 5.332133789308104e-06,
+      "loss": 2.3718,
+      "step": 1262
+    },
+    {
+      "epoch": 0.4871745419479267,
+      "grad_norm": 0.44584372814567164,
+      "learning_rate": 5.325991976049191e-06,
+      "loss": 2.2913,
+      "step": 1263
+    },
+    {
+      "epoch": 0.4875602700096432,
+      "grad_norm": 0.474275042923422,
+      "learning_rate": 5.319849668769449e-06,
+      "loss": 2.3786,
+      "step": 1264
+    },
+    {
+      "epoch": 0.4879459980713597,
+      "grad_norm": 0.4273666670416358,
+      "learning_rate": 5.313706876777166e-06,
+      "loss": 2.3495,
+      "step": 1265
+    },
+    {
+      "epoch": 0.4883317261330762,
+      "grad_norm": 0.4153343181353523,
+      "learning_rate": 5.307563609381367e-06,
+      "loss": 2.3742,
+      "step": 1266
+    },
+    {
+      "epoch": 0.4887174541947927,
+      "grad_norm": 0.4274131690236131,
+      "learning_rate": 5.301419875891795e-06,
+      "loss": 2.3406,
+      "step": 1267
+    },
+    {
+      "epoch": 0.4891031822565092,
+      "grad_norm": 0.40845967490965335,
+      "learning_rate": 5.295275685618905e-06,
+      "loss": 2.2525,
+      "step": 1268
+    },
+    {
+      "epoch": 0.48948891031822567,
+      "grad_norm": 0.4400933708424202,
+      "learning_rate": 5.289131047873837e-06,
+      "loss": 2.3325,
+      "step": 1269
+    },
+    {
+      "epoch": 0.48987463837994216,
+      "grad_norm": 0.43694952970669976,
+      "learning_rate": 5.282985971968413e-06,
+      "loss": 2.2801,
+      "step": 1270
+    },
+    {
+      "epoch": 0.49026036644165866,
+      "grad_norm": 0.42276340759115294,
+      "learning_rate": 5.276840467215119e-06,
+      "loss": 2.3553,
+      "step": 1271
+    },
+    {
+      "epoch": 0.4906460945033751,
+      "grad_norm": 0.42298935637179624,
+      "learning_rate": 5.270694542927089e-06,
+      "loss": 2.3527,
+      "step": 1272
+    },
+    {
+      "epoch": 0.4910318225650916,
+      "grad_norm": 0.4718174194659704,
+      "learning_rate": 5.264548208418094e-06,
+      "loss": 2.3241,
+      "step": 1273
+    },
+    {
+      "epoch": 0.4914175506268081,
+      "grad_norm": 0.4507159267224731,
+      "learning_rate": 5.258401473002529e-06,
+      "loss": 2.3285,
+      "step": 1274
+    },
+    {
+      "epoch": 0.4918032786885246,
+      "grad_norm": 0.4421805779778448,
+      "learning_rate": 5.252254345995392e-06,
+      "loss": 2.3861,
+      "step": 1275
+    },
+    {
+      "epoch": 0.4921890067502411,
+      "grad_norm": 0.440561261668261,
+      "learning_rate": 5.246106836712277e-06,
+      "loss": 2.3358,
+      "step": 1276
+    },
+    {
+      "epoch": 0.49257473481195757,
+      "grad_norm": 0.43081093028252854,
+      "learning_rate": 5.239958954469358e-06,
+      "loss": 2.3077,
+      "step": 1277
+    },
+    {
+      "epoch": 0.49296046287367407,
+      "grad_norm": 0.4127708092181289,
+      "learning_rate": 5.233810708583372e-06,
+      "loss": 2.2792,
+      "step": 1278
+    },
+    {
+      "epoch": 0.49334619093539056,
+      "grad_norm": 0.4159049601401717,
+      "learning_rate": 5.227662108371609e-06,
+      "loss": 2.283,
+      "step": 1279
+    },
+    {
+      "epoch": 0.49373191899710706,
+      "grad_norm": 0.489563289357969,
+      "learning_rate": 5.2215131631518945e-06,
+      "loss": 2.3171,
+      "step": 1280
+    },
+    {
+      "epoch": 0.49411764705882355,
+      "grad_norm": 0.442079206163343,
+      "learning_rate": 5.215363882242578e-06,
+      "loss": 2.3156,
+      "step": 1281
+    },
+    {
+      "epoch": 0.49450337512054005,
+      "grad_norm": 0.4133470673765935,
+      "learning_rate": 5.2092142749625165e-06,
+      "loss": 2.2269,
+      "step": 1282
+    },
+    {
+      "epoch": 0.4948891031822565,
+      "grad_norm": 0.43427602878160826,
+      "learning_rate": 5.203064350631064e-06,
+      "loss": 2.353,
+      "step": 1283
+    },
+    {
+      "epoch": 0.495274831243973,
+      "grad_norm": 0.4193435283723235,
+      "learning_rate": 5.196914118568054e-06,
+      "loss": 2.3391,
+      "step": 1284
+    },
+    {
+      "epoch": 0.4956605593056895,
+      "grad_norm": 0.4386644430997478,
+      "learning_rate": 5.190763588093781e-06,
+      "loss": 2.3025,
+      "step": 1285
+    },
+    {
+      "epoch": 0.49604628736740597,
+      "grad_norm": 0.40588838928948484,
+      "learning_rate": 5.184612768529002e-06,
+      "loss": 2.3091,
+      "step": 1286
+    },
+    {
+      "epoch": 0.49643201542912246,
+      "grad_norm": 0.4816780898609368,
+      "learning_rate": 5.178461669194903e-06,
+      "loss": 2.4023,
+      "step": 1287
+    },
+    {
+      "epoch": 0.49681774349083896,
+      "grad_norm": 0.420314726568252,
+      "learning_rate": 5.1723102994130994e-06,
+      "loss": 2.3143,
+      "step": 1288
+    },
+    {
+      "epoch": 0.49720347155255545,
+      "grad_norm": 0.46688921048479026,
+      "learning_rate": 5.166158668505612e-06,
+      "loss": 2.3528,
+      "step": 1289
+    },
+    {
+      "epoch": 0.49758919961427195,
+      "grad_norm": 0.42807252967185566,
+      "learning_rate": 5.1600067857948634e-06,
+      "loss": 2.3267,
+      "step": 1290
+    },
+    {
+      "epoch": 0.49797492767598844,
+      "grad_norm": 0.41117819104535525,
+      "learning_rate": 5.153854660603651e-06,
+      "loss": 2.2806,
+      "step": 1291
+    },
+    {
+      "epoch": 0.49836065573770494,
+      "grad_norm": 0.4073016141616718,
+      "learning_rate": 5.147702302255143e-06,
+      "loss": 2.2486,
+      "step": 1292
+    },
+    {
+      "epoch": 0.49874638379942143,
+      "grad_norm": 0.4333639612761763,
+      "learning_rate": 5.141549720072865e-06,
+      "loss": 2.2796,
+      "step": 1293
+    },
+    {
+      "epoch": 0.4991321118611379,
+      "grad_norm": 0.43911100915033385,
+      "learning_rate": 5.1353969233806735e-06,
+      "loss": 2.3437,
+      "step": 1294
+    },
+    {
+      "epoch": 0.49951783992285437,
+      "grad_norm": 0.43606136330998496,
+      "learning_rate": 5.129243921502756e-06,
+      "loss": 2.3163,
+      "step": 1295
+    },
+    {
+      "epoch": 0.49990356798457086,
+      "grad_norm": 0.4643969346558146,
+      "learning_rate": 5.123090723763607e-06,
+      "loss": 2.3239,
+      "step": 1296
+    },
+    {
+      "epoch": 0.5002892960462874,
+      "grad_norm": 0.43901031034161053,
+      "learning_rate": 5.116937339488023e-06,
+      "loss": 2.2926,
+      "step": 1297
+    },
+    {
+      "epoch": 0.5006750241080039,
+      "grad_norm": 0.4297549565643895,
+      "learning_rate": 5.110783778001077e-06,
+      "loss": 2.3467,
+      "step": 1298
+    },
+    {
+      "epoch": 0.5010607521697203,
+      "grad_norm": 0.442465634264455,
+      "learning_rate": 5.104630048628117e-06,
+      "loss": 2.3235,
+      "step": 1299
+    },
+    {
+      "epoch": 0.5014464802314368,
+      "grad_norm": 0.4118556545526013,
+      "learning_rate": 5.098476160694741e-06,
+      "loss": 2.3466,
+      "step": 1300
+    },
+    {
+      "epoch": 0.5018322082931533,
+      "grad_norm": 0.40736736064570456,
+      "learning_rate": 5.092322123526787e-06,
+      "loss": 2.3379,
+      "step": 1301
+    },
+    {
+      "epoch": 0.5022179363548698,
+      "grad_norm": 0.4059480335225931,
+      "learning_rate": 5.086167946450323e-06,
+      "loss": 2.342,
+      "step": 1302
+    },
+    {
+      "epoch": 0.5026036644165863,
+      "grad_norm": 0.4375340280301096,
+      "learning_rate": 5.080013638791624e-06,
+      "loss": 2.3221,
+      "step": 1303
+    },
+    {
+      "epoch": 0.5029893924783028,
+      "grad_norm": 0.4306039764256995,
+      "learning_rate": 5.073859209877167e-06,
+      "loss": 2.2592,
+      "step": 1304
+    },
+    {
+      "epoch": 0.5033751205400193,
+      "grad_norm": 0.4298973973686701,
+      "learning_rate": 5.06770466903361e-06,
+      "loss": 2.2559,
+      "step": 1305
+    },
+    {
+      "epoch": 0.5037608486017358,
+      "grad_norm": 0.4183112371095239,
+      "learning_rate": 5.0615500255877835e-06,
+      "loss": 2.3756,
+      "step": 1306
+    },
+    {
+      "epoch": 0.5041465766634523,
+      "grad_norm": 0.4493544264927047,
+      "learning_rate": 5.055395288866672e-06,
+      "loss": 2.3288,
+      "step": 1307
+    },
+    {
+      "epoch": 0.5045323047251687,
+      "grad_norm": 0.40468569502294427,
+      "learning_rate": 5.049240468197401e-06,
+      "loss": 2.3034,
+      "step": 1308
+    },
+    {
+      "epoch": 0.5049180327868853,
+      "grad_norm": 0.4137130754137583,
+      "learning_rate": 5.04308557290722e-06,
+      "loss": 2.3024,
+      "step": 1309
+    },
+    {
+      "epoch": 0.5053037608486017,
+      "grad_norm": 0.4400960341336459,
+      "learning_rate": 5.036930612323501e-06,
+      "loss": 2.3114,
+      "step": 1310
+    },
+    {
+      "epoch": 0.5056894889103182,
+      "grad_norm": 0.418064368559306,
+      "learning_rate": 5.030775595773706e-06,
+      "loss": 2.3444,
+      "step": 1311
+    },
+    {
+      "epoch": 0.5060752169720347,
+      "grad_norm": 0.4143452375728194,
+      "learning_rate": 5.0246205325853824e-06,
+      "loss": 2.3348,
+      "step": 1312
+    },
+    {
+      "epoch": 0.5064609450337512,
+      "grad_norm": 0.4520396890728923,
+      "learning_rate": 5.0184654320861545e-06,
+      "loss": 2.2867,
+      "step": 1313
+    },
+    {
+      "epoch": 0.5068466730954677,
+      "grad_norm": 0.42533132932115353,
+      "learning_rate": 5.0123103036036965e-06,
+      "loss": 2.3015,
+      "step": 1314
+    },
+    {
+      "epoch": 0.5072324011571842,
+      "grad_norm": 0.3916690241337248,
+      "learning_rate": 5.006155156465728e-06,
+      "loss": 2.3024,
+      "step": 1315
+    },
+    {
+      "epoch": 0.5076181292189007,
+      "grad_norm": 0.4776736612144727,
+      "learning_rate": 5e-06,
+      "loss": 2.3367,
+      "step": 1316
+    },
+    {
+      "epoch": 0.5080038572806171,
+      "grad_norm": 0.4229583495828416,
+      "learning_rate": 4.9938448435342725e-06,
+      "loss": 2.3766,
+      "step": 1317
+    },
+    {
+      "epoch": 0.5083895853423337,
+      "grad_norm": 0.45185543617759183,
+      "learning_rate": 4.987689696396305e-06,
+      "loss": 2.2442,
+      "step": 1318
+    },
+    {
+      "epoch": 0.5087753134040501,
+      "grad_norm": 0.43226886814619625,
+      "learning_rate": 4.981534567913848e-06,
+      "loss": 2.2452,
+      "step": 1319
+    },
+    {
+      "epoch": 0.5091610414657667,
+      "grad_norm": 0.4527728589734573,
+      "learning_rate": 4.975379467414621e-06,
+      "loss": 2.3577,
+      "step": 1320
+    },
+    {
+      "epoch": 0.5095467695274831,
+      "grad_norm": 0.4792812570785096,
+      "learning_rate": 4.969224404226296e-06,
+      "loss": 2.3183,
+      "step": 1321
+    },
+    {
+      "epoch": 0.5099324975891996,
+      "grad_norm": 0.42420738308445527,
+      "learning_rate": 4.963069387676499e-06,
+      "loss": 2.3004,
+      "step": 1322
+    },
+    {
+      "epoch": 0.5103182256509161,
+      "grad_norm": 0.4599343183213432,
+      "learning_rate": 4.95691442709278e-06,
+      "loss": 2.2649,
+      "step": 1323
+    },
+    {
+      "epoch": 0.5107039537126326,
+      "grad_norm": 0.41604459324505855,
+      "learning_rate": 4.950759531802602e-06,
+      "loss": 2.3898,
+      "step": 1324
+    },
+    {
+      "epoch": 0.5110896817743491,
+      "grad_norm": 0.4228763687924707,
+      "learning_rate": 4.944604711133329e-06,
+      "loss": 2.321,
+      "step": 1325
+    },
+    {
+      "epoch": 0.5114754098360655,
+      "grad_norm": 0.430884235363054,
+      "learning_rate": 4.938449974412217e-06,
+      "loss": 2.3314,
+      "step": 1326
+    },
+    {
+      "epoch": 0.5118611378977821,
+      "grad_norm": 0.46295425261968765,
+      "learning_rate": 4.932295330966392e-06,
+      "loss": 2.2935,
+      "step": 1327
+    },
+    {
+      "epoch": 0.5122468659594985,
+      "grad_norm": 0.4213530809499717,
+      "learning_rate": 4.926140790122835e-06,
+      "loss": 2.3326,
+      "step": 1328
+    },
+    {
+      "epoch": 0.5126325940212151,
+      "grad_norm": 0.4665793372541669,
+      "learning_rate": 4.919986361208379e-06,
+      "loss": 2.346,
+      "step": 1329
+    },
+    {
+      "epoch": 0.5130183220829315,
+      "grad_norm": 0.4456751472800828,
+      "learning_rate": 4.91383205354968e-06,
+      "loss": 2.3559,
+      "step": 1330
+    },
+    {
+      "epoch": 0.5134040501446481,
+      "grad_norm": 0.42134895981722886,
+      "learning_rate": 4.907677876473214e-06,
+      "loss": 2.2444,
+      "step": 1331
+    },
+    {
+      "epoch": 0.5137897782063645,
+      "grad_norm": 0.41934135489864954,
+      "learning_rate": 4.90152383930526e-06,
+      "loss": 2.2932,
+      "step": 1332
+    },
+    {
+      "epoch": 0.514175506268081,
+      "grad_norm": 0.45310057387043157,
+      "learning_rate": 4.895369951371884e-06,
+      "loss": 2.3435,
+      "step": 1333
+    },
+    {
+      "epoch": 0.5145612343297975,
+      "grad_norm": 0.5258865623441964,
+      "learning_rate": 4.889216221998925e-06,
+      "loss": 2.3424,
+      "step": 1334
+    },
+    {
+      "epoch": 0.5149469623915139,
+      "grad_norm": 0.48039153232624293,
+      "learning_rate": 4.883062660511979e-06,
+      "loss": 2.3399,
+      "step": 1335
+    },
+    {
+      "epoch": 0.5153326904532305,
+      "grad_norm": 0.47444154350583756,
+      "learning_rate": 4.876909276236395e-06,
+      "loss": 2.3539,
+      "step": 1336
+    },
+    {
+      "epoch": 0.5157184185149469,
+      "grad_norm": 0.5069800163012861,
+      "learning_rate": 4.870756078497247e-06,
+      "loss": 2.379,
+      "step": 1337
+    },
+    {
+      "epoch": 0.5161041465766635,
+      "grad_norm": 0.49748181116335205,
+      "learning_rate": 4.864603076619329e-06,
+      "loss": 2.2928,
+      "step": 1338
+    },
+    {
+      "epoch": 0.5164898746383799,
+      "grad_norm": 0.42706937479473217,
+      "learning_rate": 4.858450279927138e-06,
+      "loss": 2.3086,
+      "step": 1339
+    },
+    {
+      "epoch": 0.5168756027000965,
+      "grad_norm": 0.41046138972851537,
+      "learning_rate": 4.852297697744857e-06,
+      "loss": 2.2277,
+      "step": 1340
+    },
+    {
+      "epoch": 0.5172613307618129,
+      "grad_norm": 0.4330642896329234,
+      "learning_rate": 4.84614533939635e-06,
+      "loss": 2.2658,
+      "step": 1341
+    },
+    {
+      "epoch": 0.5176470588235295,
+      "grad_norm": 0.43557729756163627,
+      "learning_rate": 4.839993214205138e-06,
+      "loss": 2.2969,
+      "step": 1342
+    },
+    {
+      "epoch": 0.5180327868852459,
+      "grad_norm": 0.4132315515601232,
+      "learning_rate": 4.83384133149439e-06,
+      "loss": 2.306,
+      "step": 1343
+    },
+    {
+      "epoch": 0.5184185149469623,
+      "grad_norm": 0.4725680753577124,
+      "learning_rate": 4.827689700586902e-06,
+      "loss": 2.3319,
+      "step": 1344
+    },
+    {
+      "epoch": 0.5188042430086789,
+      "grad_norm": 0.4024975847853517,
+      "learning_rate": 4.821538330805098e-06,
+      "loss": 2.3366,
+      "step": 1345
+    },
+    {
+      "epoch": 0.5191899710703953,
+      "grad_norm": 0.4771435011033568,
+      "learning_rate": 4.815387231471001e-06,
+      "loss": 2.3304,
+      "step": 1346
+    },
+    {
+      "epoch": 0.5195756991321119,
+      "grad_norm": 0.45140386154895684,
+      "learning_rate": 4.80923641190622e-06,
+      "loss": 2.3414,
+      "step": 1347
+    },
+    {
+      "epoch": 0.5199614271938283,
+      "grad_norm": 0.42681886917358797,
+      "learning_rate": 4.803085881431949e-06,
+      "loss": 2.3342,
+      "step": 1348
+    },
+    {
+      "epoch": 0.5203471552555449,
+      "grad_norm": 0.4437892746591648,
+      "learning_rate": 4.796935649368936e-06,
+      "loss": 2.2996,
+      "step": 1349
+    },
+    {
+      "epoch": 0.5207328833172613,
+      "grad_norm": 0.5069638790201045,
+      "learning_rate": 4.790785725037484e-06,
+      "loss": 2.3107,
+      "step": 1350
+    },
+    {
+      "epoch": 0.5211186113789779,
+      "grad_norm": 0.4366235525915503,
+      "learning_rate": 4.784636117757423e-06,
+      "loss": 2.2616,
+      "step": 1351
+    },
+    {
+      "epoch": 0.5215043394406943,
+      "grad_norm": 0.4652015925814833,
+      "learning_rate": 4.778486836848107e-06,
+      "loss": 2.3107,
+      "step": 1352
+    },
+    {
+      "epoch": 0.5218900675024108,
+      "grad_norm": 0.46175852384207916,
+      "learning_rate": 4.772337891628394e-06,
+      "loss": 2.2378,
+      "step": 1353
+    },
+    {
+      "epoch": 0.5222757955641273,
+      "grad_norm": 0.43883478665108466,
+      "learning_rate": 4.76618929141663e-06,
+      "loss": 2.3241,
+      "step": 1354
+    },
+    {
+      "epoch": 0.5226615236258437,
+      "grad_norm": 0.4417539282200209,
+      "learning_rate": 4.760041045530645e-06,
+      "loss": 2.2594,
+      "step": 1355
+    },
+    {
+      "epoch": 0.5230472516875603,
+      "grad_norm": 0.4063010167186865,
+      "learning_rate": 4.7538931632877254e-06,
+      "loss": 2.3434,
+      "step": 1356
+    },
+    {
+      "epoch": 0.5234329797492767,
+      "grad_norm": 0.4384298510879702,
+      "learning_rate": 4.7477456540046105e-06,
+      "loss": 2.3348,
+      "step": 1357
+    },
+    {
+      "epoch": 0.5238187078109933,
+      "grad_norm": 0.45915359256206345,
+      "learning_rate": 4.741598526997474e-06,
+      "loss": 2.3056,
+      "step": 1358
+    },
+    {
+      "epoch": 0.5242044358727097,
+      "grad_norm": 0.43968549360910814,
+      "learning_rate": 4.7354517915819065e-06,
+      "loss": 2.336,
+      "step": 1359
+    },
+    {
+      "epoch": 0.5245901639344263,
+      "grad_norm": 0.41796295540394873,
+      "learning_rate": 4.729305457072913e-06,
+      "loss": 2.2748,
+      "step": 1360
+    },
+    {
+      "epoch": 0.5249758919961427,
+      "grad_norm": 0.4362828750825012,
+      "learning_rate": 4.723159532784883e-06,
+      "loss": 2.2561,
+      "step": 1361
+    },
+    {
+      "epoch": 0.5253616200578592,
+      "grad_norm": 0.4371254165879112,
+      "learning_rate": 4.717014028031589e-06,
+      "loss": 2.3335,
+      "step": 1362
+    },
+    {
+      "epoch": 0.5257473481195757,
+      "grad_norm": 0.45665829594986745,
+      "learning_rate": 4.710868952126166e-06,
+      "loss": 2.3818,
+      "step": 1363
+    },
+    {
+      "epoch": 0.5261330761812922,
+      "grad_norm": 0.44081750659388047,
+      "learning_rate": 4.704724314381097e-06,
+      "loss": 2.3263,
+      "step": 1364
+    },
+    {
+      "epoch": 0.5265188042430087,
+      "grad_norm": 0.4437002618568672,
+      "learning_rate": 4.6985801241082065e-06,
+      "loss": 2.2988,
+      "step": 1365
+    },
+    {
+      "epoch": 0.5269045323047251,
+      "grad_norm": 0.4459940408363005,
+      "learning_rate": 4.692436390618635e-06,
+      "loss": 2.279,
+      "step": 1366
+    },
+    {
+      "epoch": 0.5272902603664417,
+      "grad_norm": 0.4433554462879378,
+      "learning_rate": 4.686293123222837e-06,
+      "loss": 2.3205,
+      "step": 1367
+    },
+    {
+      "epoch": 0.5276759884281581,
+      "grad_norm": 0.4391895686696009,
+      "learning_rate": 4.680150331230552e-06,
+      "loss": 2.3565,
+      "step": 1368
+    },
+    {
+      "epoch": 0.5280617164898747,
+      "grad_norm": 0.44910921831985756,
+      "learning_rate": 4.674008023950809e-06,
+      "loss": 2.3278,
+      "step": 1369
+    },
+    {
+      "epoch": 0.5284474445515911,
+      "grad_norm": 0.4183548874298808,
+      "learning_rate": 4.667866210691897e-06,
+      "loss": 2.41,
+      "step": 1370
+    },
+    {
+      "epoch": 0.5288331726133076,
+      "grad_norm": 0.42557100108547463,
+      "learning_rate": 4.661724900761355e-06,
+      "loss": 2.2916,
+      "step": 1371
+    },
+    {
+      "epoch": 0.5292189006750241,
+      "grad_norm": 0.45979466957641174,
+      "learning_rate": 4.6555841034659625e-06,
+      "loss": 2.2926,
+      "step": 1372
+    },
+    {
+      "epoch": 0.5296046287367406,
+      "grad_norm": 0.40343968385412204,
+      "learning_rate": 4.6494438281117195e-06,
+      "loss": 2.3008,
+      "step": 1373
+    },
+    {
+      "epoch": 0.5299903567984571,
+      "grad_norm": 0.4418053528829375,
+      "learning_rate": 4.643304084003839e-06,
+      "loss": 2.2997,
+      "step": 1374
+    },
+    {
+      "epoch": 0.5303760848601736,
+      "grad_norm": 0.42781136201724485,
+      "learning_rate": 4.637164880446725e-06,
+      "loss": 2.2972,
+      "step": 1375
+    },
+    {
+      "epoch": 0.5307618129218901,
+      "grad_norm": 0.41523373826769044,
+      "learning_rate": 4.631026226743962e-06,
+      "loss": 2.3737,
+      "step": 1376
+    },
+    {
+      "epoch": 0.5311475409836065,
+      "grad_norm": 0.43181206158587115,
+      "learning_rate": 4.624888132198303e-06,
+      "loss": 2.3483,
+      "step": 1377
+    },
+    {
+      "epoch": 0.531533269045323,
+      "grad_norm": 0.4883659943209646,
+      "learning_rate": 4.618750606111657e-06,
+      "loss": 2.3484,
+      "step": 1378
+    },
+    {
+      "epoch": 0.5319189971070395,
+      "grad_norm": 0.4562404426597322,
+      "learning_rate": 4.612613657785066e-06,
+      "loss": 2.2604,
+      "step": 1379
+    },
+    {
+      "epoch": 0.532304725168756,
+      "grad_norm": 0.43283921590840496,
+      "learning_rate": 4.606477296518698e-06,
+      "loss": 2.3077,
+      "step": 1380
+    },
+    {
+      "epoch": 0.5326904532304725,
+      "grad_norm": 0.46474300575051714,
+      "learning_rate": 4.600341531611833e-06,
+      "loss": 2.3756,
+      "step": 1381
+    },
+    {
+      "epoch": 0.533076181292189,
+      "grad_norm": 0.4918375192419344,
+      "learning_rate": 4.594206372362845e-06,
+      "loss": 2.3844,
+      "step": 1382
+    },
+    {
+      "epoch": 0.5334619093539055,
+      "grad_norm": 0.43756529480102446,
+      "learning_rate": 4.588071828069191e-06,
+      "loss": 2.2807,
+      "step": 1383
+    },
+    {
+      "epoch": 0.533847637415622,
+      "grad_norm": 0.43556271744351255,
+      "learning_rate": 4.581937908027397e-06,
+      "loss": 2.2884,
+      "step": 1384
+    },
+    {
+      "epoch": 0.5342333654773385,
+      "grad_norm": 0.41233607827108854,
+      "learning_rate": 4.575804621533043e-06,
+      "loss": 2.3097,
+      "step": 1385
+    },
+    {
+      "epoch": 0.534619093539055,
+      "grad_norm": 0.4323643545291776,
+      "learning_rate": 4.5696719778807465e-06,
+      "loss": 2.3127,
+      "step": 1386
+    },
+    {
+      "epoch": 0.5350048216007715,
+      "grad_norm": 0.44083524101705623,
+      "learning_rate": 4.563539986364152e-06,
+      "loss": 2.3313,
+      "step": 1387
+    },
+    {
+      "epoch": 0.5353905496624879,
+      "grad_norm": 0.4144429426000985,
+      "learning_rate": 4.55740865627592e-06,
+      "loss": 2.3078,
+      "step": 1388
+    },
+    {
+      "epoch": 0.5357762777242044,
+      "grad_norm": 0.4984500916383888,
+      "learning_rate": 4.551277996907703e-06,
+      "loss": 2.2869,
+      "step": 1389
+    },
+    {
+      "epoch": 0.5361620057859209,
+      "grad_norm": 0.44955445867207805,
+      "learning_rate": 4.545148017550138e-06,
+      "loss": 2.3013,
+      "step": 1390
+    },
+    {
+      "epoch": 0.5365477338476374,
+      "grad_norm": 0.45229936484140115,
+      "learning_rate": 4.5390187274928325e-06,
+      "loss": 2.2827,
+      "step": 1391
+    },
+    {
+      "epoch": 0.5369334619093539,
+      "grad_norm": 0.41324968774465565,
+      "learning_rate": 4.532890136024351e-06,
+      "loss": 2.3624,
+      "step": 1392
+    },
+    {
+      "epoch": 0.5373191899710704,
+      "grad_norm": 0.47895598575081966,
+      "learning_rate": 4.5267622524321955e-06,
+      "loss": 2.2684,
+      "step": 1393
+    },
+    {
+      "epoch": 0.5377049180327869,
+      "grad_norm": 0.5188073056811957,
+      "learning_rate": 4.520635086002799e-06,
+      "loss": 2.3686,
+      "step": 1394
+    },
+    {
+      "epoch": 0.5380906460945034,
+      "grad_norm": 0.431259848225407,
+      "learning_rate": 4.514508646021506e-06,
+      "loss": 2.3401,
+      "step": 1395
+    },
+    {
+      "epoch": 0.5384763741562199,
+      "grad_norm": 0.47217102884984974,
+      "learning_rate": 4.508382941772558e-06,
+      "loss": 2.3236,
+      "step": 1396
+    },
+    {
+      "epoch": 0.5388621022179364,
+      "grad_norm": 0.44966054172796366,
+      "learning_rate": 4.502257982539087e-06,
+      "loss": 2.3164,
+      "step": 1397
+    },
+    {
+      "epoch": 0.5392478302796528,
+      "grad_norm": 0.47812913336760227,
+      "learning_rate": 4.496133777603093e-06,
+      "loss": 2.3211,
+      "step": 1398
+    },
+    {
+      "epoch": 0.5396335583413693,
+      "grad_norm": 0.465887586248686,
+      "learning_rate": 4.49001033624543e-06,
+      "loss": 2.3873,
+      "step": 1399
+    },
+    {
+      "epoch": 0.5400192864030858,
+      "grad_norm": 0.4611561152316998,
+      "learning_rate": 4.483887667745798e-06,
+      "loss": 2.3814,
+      "step": 1400
+    },
+    {
+      "epoch": 0.5404050144648023,
+      "grad_norm": 0.43480217765151674,
+      "learning_rate": 4.477765781382728e-06,
+      "loss": 2.3513,
+      "step": 1401
+    },
+    {
+      "epoch": 0.5407907425265188,
+      "grad_norm": 0.443621190936497,
+      "learning_rate": 4.471644686433559e-06,
+      "loss": 2.3188,
+      "step": 1402
+    },
+    {
+      "epoch": 0.5411764705882353,
+      "grad_norm": 0.47335316164044927,
+      "learning_rate": 4.465524392174437e-06,
+      "loss": 2.3548,
+      "step": 1403
+    },
+    {
+      "epoch": 0.5415621986499518,
+      "grad_norm": 0.4568037576128477,
+      "learning_rate": 4.459404907880293e-06,
+      "loss": 2.2584,
+      "step": 1404
+    },
+    {
+      "epoch": 0.5419479267116682,
+      "grad_norm": 0.48574365511330586,
+      "learning_rate": 4.453286242824827e-06,
+      "loss": 2.271,
+      "step": 1405
+    },
+    {
+      "epoch": 0.5423336547733848,
+      "grad_norm": 0.41630044315682513,
+      "learning_rate": 4.447168406280503e-06,
+      "loss": 2.3783,
+      "step": 1406
+    },
+    {
+      "epoch": 0.5427193828351012,
+      "grad_norm": 0.4388643867308547,
+      "learning_rate": 4.441051407518525e-06,
+      "loss": 2.3014,
+      "step": 1407
+    },
+    {
+      "epoch": 0.5431051108968178,
+      "grad_norm": 0.46351987282241525,
+      "learning_rate": 4.434935255808831e-06,
+      "loss": 2.3795,
+      "step": 1408
+    },
+    {
+      "epoch": 0.5434908389585342,
+      "grad_norm": 0.4067052982054639,
+      "learning_rate": 4.428819960420072e-06,
+      "loss": 2.3166,
+      "step": 1409
+    },
+    {
+      "epoch": 0.5438765670202508,
+      "grad_norm": 0.4424432749709641,
+      "learning_rate": 4.422705530619604e-06,
+      "loss": 2.2886,
+      "step": 1410
+    },
+    {
+      "epoch": 0.5442622950819672,
+      "grad_norm": 0.440011691897373,
+      "learning_rate": 4.4165919756734695e-06,
+      "loss": 2.354,
+      "step": 1411
+    },
+    {
+      "epoch": 0.5446480231436837,
+      "grad_norm": 0.4651265995172307,
+      "learning_rate": 4.410479304846385e-06,
+      "loss": 2.3595,
+      "step": 1412
+    },
+    {
+      "epoch": 0.5450337512054002,
+      "grad_norm": 0.4127500619791685,
+      "learning_rate": 4.4043675274017284e-06,
+      "loss": 2.2397,
+      "step": 1413
+    },
+    {
+      "epoch": 0.5454194792671166,
+      "grad_norm": 0.4228985803007235,
+      "learning_rate": 4.3982566526015244e-06,
+      "loss": 2.3301,
+      "step": 1414
+    },
+    {
+      "epoch": 0.5458052073288332,
+      "grad_norm": 0.4257754524806136,
+      "learning_rate": 4.392146689706426e-06,
+      "loss": 2.2894,
+      "step": 1415
+    },
+    {
+      "epoch": 0.5461909353905496,
+      "grad_norm": 0.4464196651405697,
+      "learning_rate": 4.386037647975708e-06,
+      "loss": 2.3799,
+      "step": 1416
+    },
+    {
+      "epoch": 0.5465766634522662,
+      "grad_norm": 0.4621712141445011,
+      "learning_rate": 4.379929536667251e-06,
+      "loss": 2.3379,
+      "step": 1417
+    },
+    {
+      "epoch": 0.5469623915139826,
+      "grad_norm": 0.4509721789709156,
+      "learning_rate": 4.37382236503752e-06,
+      "loss": 2.371,
+      "step": 1418
+    },
+    {
+      "epoch": 0.5473481195756992,
+      "grad_norm": 0.4314830170510386,
+      "learning_rate": 4.3677161423415584e-06,
+      "loss": 2.3325,
+      "step": 1419
+    },
+    {
+      "epoch": 0.5477338476374156,
+      "grad_norm": 0.42464418592417985,
+      "learning_rate": 4.361610877832974e-06,
+      "loss": 2.3283,
+      "step": 1420
+    },
+    {
+      "epoch": 0.5481195756991322,
+      "grad_norm": 0.4094192326677035,
+      "learning_rate": 4.355506580763919e-06,
+      "loss": 2.311,
+      "step": 1421
+    },
+    {
+      "epoch": 0.5485053037608486,
+      "grad_norm": 0.4677564370459633,
+      "learning_rate": 4.3494032603850804e-06,
+      "loss": 2.2784,
+      "step": 1422
+    },
+    {
+      "epoch": 0.548891031822565,
+      "grad_norm": 0.4205875476635024,
+      "learning_rate": 4.343300925945667e-06,
+      "loss": 2.3176,
+      "step": 1423
+    },
+    {
+      "epoch": 0.5492767598842816,
+      "grad_norm": 0.46917697693113575,
+      "learning_rate": 4.337199586693389e-06,
+      "loss": 2.3322,
+      "step": 1424
+    },
+    {
+      "epoch": 0.549662487945998,
+      "grad_norm": 0.4301394790054798,
+      "learning_rate": 4.331099251874457e-06,
+      "loss": 2.3428,
+      "step": 1425
+    },
+    {
+      "epoch": 0.5500482160077146,
+      "grad_norm": 0.43302569982748224,
+      "learning_rate": 4.32499993073355e-06,
+      "loss": 2.3217,
+      "step": 1426
+    },
+    {
+      "epoch": 0.550433944069431,
+      "grad_norm": 0.4603363514517138,
+      "learning_rate": 4.3189016325138155e-06,
+      "loss": 2.2692,
+      "step": 1427
+    },
+    {
+      "epoch": 0.5508196721311476,
+      "grad_norm": 0.41660294234922207,
+      "learning_rate": 4.312804366456851e-06,
+      "loss": 2.3322,
+      "step": 1428
+    },
+    {
+      "epoch": 0.551205400192864,
+      "grad_norm": 0.43506495250256916,
+      "learning_rate": 4.306708141802687e-06,
+      "loss": 2.3301,
+      "step": 1429
+    },
+    {
+      "epoch": 0.5515911282545806,
+      "grad_norm": 0.4453736574375762,
+      "learning_rate": 4.30061296778978e-06,
+      "loss": 2.2675,
+      "step": 1430
+    },
+    {
+      "epoch": 0.551976856316297,
+      "grad_norm": 0.4904730112987836,
+      "learning_rate": 4.294518853654991e-06,
+      "loss": 2.3282,
+      "step": 1431
+    },
+    {
+      "epoch": 0.5523625843780136,
+      "grad_norm": 0.45455508963296065,
+      "learning_rate": 4.2884258086335755e-06,
+      "loss": 2.3528,
+      "step": 1432
+    },
+    {
+      "epoch": 0.55274831243973,
+      "grad_norm": 0.4180507969561658,
+      "learning_rate": 4.282333841959171e-06,
+      "loss": 2.2981,
+      "step": 1433
+    },
+    {
+      "epoch": 0.5531340405014464,
+      "grad_norm": 0.45184851358234884,
+      "learning_rate": 4.276242962863775e-06,
+      "loss": 2.3093,
+      "step": 1434
+    },
+    {
+      "epoch": 0.553519768563163,
+      "grad_norm": 0.4554037815490411,
+      "learning_rate": 4.270153180577746e-06,
+      "loss": 2.3091,
+      "step": 1435
+    },
+    {
+      "epoch": 0.5539054966248794,
+      "grad_norm": 0.4346303129201663,
+      "learning_rate": 4.2640645043297715e-06,
+      "loss": 2.3912,
+      "step": 1436
+    },
+    {
+      "epoch": 0.554291224686596,
+      "grad_norm": 0.45225269018483133,
+      "learning_rate": 4.25797694334687e-06,
+      "loss": 2.2834,
+      "step": 1437
+    },
+    {
+      "epoch": 0.5546769527483124,
+      "grad_norm": 0.4084486155678106,
+      "learning_rate": 4.251890506854365e-06,
+      "loss": 2.3305,
+      "step": 1438
+    },
+    {
+      "epoch": 0.555062680810029,
+      "grad_norm": 0.4575803527285325,
+      "learning_rate": 4.2458052040758746e-06,
+      "loss": 2.3479,
+      "step": 1439
+    },
+    {
+      "epoch": 0.5554484088717454,
+      "grad_norm": 0.44545457389858556,
+      "learning_rate": 4.239721044233306e-06,
+      "loss": 2.3287,
+      "step": 1440
+    },
+    {
+      "epoch": 0.555834136933462,
+      "grad_norm": 0.409063828267486,
+      "learning_rate": 4.2336380365468274e-06,
+      "loss": 2.2689,
+      "step": 1441
+    },
+    {
+      "epoch": 0.5562198649951784,
+      "grad_norm": 0.43152638314173347,
+      "learning_rate": 4.227556190234864e-06,
+      "loss": 2.3285,
+      "step": 1442
+    },
+    {
+      "epoch": 0.5566055930568949,
+      "grad_norm": 0.43376742895032044,
+      "learning_rate": 4.22147551451408e-06,
+      "loss": 2.2717,
+      "step": 1443
+    },
+    {
+      "epoch": 0.5569913211186114,
+      "grad_norm": 0.42210336592990916,
+      "learning_rate": 4.215396018599369e-06,
+      "loss": 2.3696,
+      "step": 1444
+    },
+    {
+      "epoch": 0.5573770491803278,
+      "grad_norm": 0.4182398682850248,
+      "learning_rate": 4.209317711703829e-06,
+      "loss": 2.3391,
+      "step": 1445
+    },
+    {
+      "epoch": 0.5577627772420444,
+      "grad_norm": 0.4387139655900613,
+      "learning_rate": 4.203240603038768e-06,
+      "loss": 2.3068,
+      "step": 1446
+    },
+    {
+      "epoch": 0.5581485053037608,
+      "grad_norm": 0.4550086921214388,
+      "learning_rate": 4.197164701813665e-06,
+      "loss": 2.2963,
+      "step": 1447
+    },
+    {
+      "epoch": 0.5585342333654774,
+      "grad_norm": 0.44569448498905484,
+      "learning_rate": 4.191090017236177e-06,
+      "loss": 2.3024,
+      "step": 1448
+    },
+    {
+      "epoch": 0.5589199614271938,
+      "grad_norm": 0.4494128519533429,
+      "learning_rate": 4.185016558512117e-06,
+      "loss": 2.271,
+      "step": 1449
+    },
+    {
+      "epoch": 0.5593056894889104,
+      "grad_norm": 0.4208880799036616,
+      "learning_rate": 4.178944334845438e-06,
+      "loss": 2.3638,
+      "step": 1450
+    },
+    {
+      "epoch": 0.5596914175506268,
+      "grad_norm": 0.43348786281976237,
+      "learning_rate": 4.1728733554382204e-06,
+      "loss": 2.3346,
+      "step": 1451
+    },
+    {
+      "epoch": 0.5600771456123433,
+      "grad_norm": 0.4418399784800928,
+      "learning_rate": 4.166803629490664e-06,
+      "loss": 2.2924,
+      "step": 1452
+    },
+    {
+      "epoch": 0.5604628736740598,
+      "grad_norm": 0.44668162646159176,
+      "learning_rate": 4.160735166201063e-06,
+      "loss": 2.3616,
+      "step": 1453
+    },
+    {
+      "epoch": 0.5608486017357763,
+      "grad_norm": 0.4354460375268765,
+      "learning_rate": 4.1546679747658045e-06,
+      "loss": 2.3326,
+      "step": 1454
+    },
+    {
+      "epoch": 0.5612343297974928,
+      "grad_norm": 0.4142044774335118,
+      "learning_rate": 4.148602064379342e-06,
+      "loss": 2.3641,
+      "step": 1455
+    },
+    {
+      "epoch": 0.5616200578592092,
+      "grad_norm": 0.4640451050436532,
+      "learning_rate": 4.142537444234192e-06,
+      "loss": 2.324,
+      "step": 1456
+    },
+    {
+      "epoch": 0.5620057859209258,
+      "grad_norm": 0.43337984727739914,
+      "learning_rate": 4.136474123520913e-06,
+      "loss": 2.3159,
+      "step": 1457
+    },
+    {
+      "epoch": 0.5623915139826422,
+      "grad_norm": 0.42751653282762,
+      "learning_rate": 4.1304121114280946e-06,
+      "loss": 2.3692,
+      "step": 1458
+    },
+    {
+      "epoch": 0.5627772420443587,
+      "grad_norm": 0.4462906076451016,
+      "learning_rate": 4.1243514171423465e-06,
+      "loss": 2.3026,
+      "step": 1459
+    },
+    {
+      "epoch": 0.5631629701060752,
+      "grad_norm": 0.4535928576994934,
+      "learning_rate": 4.118292049848277e-06,
+      "loss": 2.2959,
+      "step": 1460
+    },
+    {
+      "epoch": 0.5635486981677917,
+      "grad_norm": 0.4306687463240508,
+      "learning_rate": 4.1122340187284845e-06,
+      "loss": 2.31,
+      "step": 1461
+    },
+    {
+      "epoch": 0.5639344262295082,
+      "grad_norm": 0.43039521461340857,
+      "learning_rate": 4.106177332963544e-06,
+      "loss": 2.281,
+      "step": 1462
+    },
+    {
+      "epoch": 0.5643201542912247,
+      "grad_norm": 0.4026555682467883,
+      "learning_rate": 4.100122001731993e-06,
+      "loss": 2.3083,
+      "step": 1463
+    },
+    {
+      "epoch": 0.5647058823529412,
+      "grad_norm": 0.429036499023987,
+      "learning_rate": 4.094068034210313e-06,
+      "loss": 2.3539,
+      "step": 1464
+    },
+    {
+      "epoch": 0.5650916104146577,
+      "grad_norm": 0.4221609577666498,
+      "learning_rate": 4.088015439572919e-06,
+      "loss": 2.379,
+      "step": 1465
+    },
+    {
+      "epoch": 0.5654773384763742,
+      "grad_norm": 0.4298721214257128,
+      "learning_rate": 4.081964226992149e-06,
+      "loss": 2.2893,
+      "step": 1466
+    },
+    {
+      "epoch": 0.5658630665380906,
+      "grad_norm": 0.4518357335488598,
+      "learning_rate": 4.0759144056382426e-06,
+      "loss": 2.4006,
+      "step": 1467
+    },
+    {
+      "epoch": 0.5662487945998071,
+      "grad_norm": 0.4362484857989554,
+      "learning_rate": 4.069865984679332e-06,
+      "loss": 2.2697,
+      "step": 1468
+    },
+    {
+      "epoch": 0.5666345226615236,
+      "grad_norm": 0.47563962927912806,
+      "learning_rate": 4.06381897328143e-06,
+      "loss": 2.3475,
+      "step": 1469
+    },
+    {
+      "epoch": 0.5670202507232401,
+      "grad_norm": 0.4090665480564414,
+      "learning_rate": 4.057773380608411e-06,
+      "loss": 2.3633,
+      "step": 1470
+    },
+    {
+      "epoch": 0.5674059787849566,
+      "grad_norm": 0.4041220890931108,
+      "learning_rate": 4.051729215821995e-06,
+      "loss": 2.2962,
+      "step": 1471
+    },
+    {
+      "epoch": 0.5677917068466731,
+      "grad_norm": 0.43674579923533235,
+      "learning_rate": 4.045686488081748e-06,
+      "loss": 2.3277,
+      "step": 1472
+    },
+    {
+      "epoch": 0.5681774349083896,
+      "grad_norm": 0.44877559752623075,
+      "learning_rate": 4.039645206545053e-06,
+      "loss": 2.3456,
+      "step": 1473
+    },
+    {
+      "epoch": 0.5685631629701061,
+      "grad_norm": 0.433832854221165,
+      "learning_rate": 4.033605380367099e-06,
+      "loss": 2.3341,
+      "step": 1474
+    },
+    {
+      "epoch": 0.5689488910318226,
+      "grad_norm": 0.46945090750470925,
+      "learning_rate": 4.027567018700873e-06,
+      "loss": 2.3741,
+      "step": 1475
+    },
+    {
+      "epoch": 0.5693346190935391,
+      "grad_norm": 0.4276102055477122,
+      "learning_rate": 4.021530130697141e-06,
+      "loss": 2.3246,
+      "step": 1476
+    },
+    {
+      "epoch": 0.5697203471552555,
+      "grad_norm": 0.42626648975286313,
+      "learning_rate": 4.015494725504435e-06,
+      "loss": 2.2405,
+      "step": 1477
+    },
+    {
+      "epoch": 0.570106075216972,
+      "grad_norm": 0.44367775635689727,
+      "learning_rate": 4.009460812269045e-06,
+      "loss": 2.3604,
+      "step": 1478
+    },
+    {
+      "epoch": 0.5704918032786885,
+      "grad_norm": 0.4082484041938663,
+      "learning_rate": 4.003428400134992e-06,
+      "loss": 2.3743,
+      "step": 1479
+    },
+    {
+      "epoch": 0.570877531340405,
+      "grad_norm": 0.43933866170044744,
+      "learning_rate": 3.997397498244028e-06,
+      "loss": 2.3001,
+      "step": 1480
+    },
+    {
+      "epoch": 0.5712632594021215,
+      "grad_norm": 0.43375151507001325,
+      "learning_rate": 3.991368115735612e-06,
+      "loss": 2.3359,
+      "step": 1481
+    },
+    {
+      "epoch": 0.571648987463838,
+      "grad_norm": 0.43964106653478796,
+      "learning_rate": 3.985340261746909e-06,
+      "loss": 2.3816,
+      "step": 1482
+    },
+    {
+      "epoch": 0.5720347155255545,
+      "grad_norm": 0.4363234931572726,
+      "learning_rate": 3.979313945412758e-06,
+      "loss": 2.3066,
+      "step": 1483
+    },
+    {
+      "epoch": 0.572420443587271,
+      "grad_norm": 0.4252315319429752,
+      "learning_rate": 3.97328917586567e-06,
+      "loss": 2.2941,
+      "step": 1484
+    },
+    {
+      "epoch": 0.5728061716489875,
+      "grad_norm": 0.4611132439778421,
+      "learning_rate": 3.9672659622358175e-06,
+      "loss": 2.2804,
+      "step": 1485
+    },
+    {
+      "epoch": 0.573191899710704,
+      "grad_norm": 0.47487041414747433,
+      "learning_rate": 3.961244313651008e-06,
+      "loss": 2.3275,
+      "step": 1486
+    },
+    {
+      "epoch": 0.5735776277724205,
+      "grad_norm": 0.44979611706443356,
+      "learning_rate": 3.955224239236681e-06,
+      "loss": 2.3187,
+      "step": 1487
+    },
+    {
+      "epoch": 0.5739633558341369,
+      "grad_norm": 0.4524099895055111,
+      "learning_rate": 3.9492057481158905e-06,
+      "loss": 2.3436,
+      "step": 1488
+    },
+    {
+      "epoch": 0.5743490838958534,
+      "grad_norm": 0.40135274768366935,
+      "learning_rate": 3.943188849409289e-06,
+      "loss": 2.3494,
+      "step": 1489
+    },
+    {
+      "epoch": 0.5747348119575699,
+      "grad_norm": 0.43440989096021043,
+      "learning_rate": 3.937173552235117e-06,
+      "loss": 2.2697,
+      "step": 1490
+    },
+    {
+      "epoch": 0.5751205400192864,
+      "grad_norm": 0.4704618668381495,
+      "learning_rate": 3.9311598657091895e-06,
+      "loss": 2.321,
+      "step": 1491
+    },
+    {
+      "epoch": 0.5755062680810029,
+      "grad_norm": 0.42035450557088433,
+      "learning_rate": 3.92514779894488e-06,
+      "loss": 2.3282,
+      "step": 1492
+    },
+    {
+      "epoch": 0.5758919961427194,
+      "grad_norm": 0.44099448828389753,
+      "learning_rate": 3.919137361053105e-06,
+      "loss": 2.3271,
+      "step": 1493
+    },
+    {
+      "epoch": 0.5762777242044359,
+      "grad_norm": 0.43312368312817023,
+      "learning_rate": 3.913128561142315e-06,
+      "loss": 2.2885,
+      "step": 1494
+    },
+    {
+      "epoch": 0.5766634522661523,
+      "grad_norm": 0.42701973541052424,
+      "learning_rate": 3.907121408318478e-06,
+      "loss": 2.3509,
+      "step": 1495
+    },
+    {
+      "epoch": 0.5770491803278689,
+      "grad_norm": 0.43036889157362906,
+      "learning_rate": 3.901115911685063e-06,
+      "loss": 2.3325,
+      "step": 1496
+    },
+    {
+      "epoch": 0.5774349083895853,
+      "grad_norm": 0.422898379339551,
+      "learning_rate": 3.895112080343033e-06,
+      "loss": 2.3412,
+      "step": 1497
+    },
+    {
+      "epoch": 0.5778206364513019,
+      "grad_norm": 0.42172624395411484,
+      "learning_rate": 3.889109923390827e-06,
+      "loss": 2.3041,
+      "step": 1498
+    },
+    {
+      "epoch": 0.5782063645130183,
+      "grad_norm": 0.413562412674589,
+      "learning_rate": 3.8831094499243425e-06,
+      "loss": 2.2498,
+      "step": 1499
+    },
+    {
+      "epoch": 0.5785920925747348,
+      "grad_norm": 0.439426523547599,
+      "learning_rate": 3.877110669036932e-06,
+      "loss": 2.295,
+      "step": 1500
+    },
+    {
+      "epoch": 0.5789778206364513,
+      "grad_norm": 0.45232269736121317,
+      "learning_rate": 3.8711135898193775e-06,
+      "loss": 2.2968,
+      "step": 1501
+    },
+    {
+      "epoch": 0.5793635486981678,
+      "grad_norm": 0.43144290461141577,
+      "learning_rate": 3.865118221359887e-06,
+      "loss": 2.2932,
+      "step": 1502
+    },
+    {
+      "epoch": 0.5797492767598843,
+      "grad_norm": 0.4824244549953793,
+      "learning_rate": 3.859124572744072e-06,
+      "loss": 2.3413,
+      "step": 1503
+    },
+    {
+      "epoch": 0.5801350048216007,
+      "grad_norm": 0.44213543325579097,
+      "learning_rate": 3.853132653054936e-06,
+      "loss": 2.4002,
+      "step": 1504
+    },
+    {
+      "epoch": 0.5805207328833173,
+      "grad_norm": 0.4425691507108305,
+      "learning_rate": 3.84714247137287e-06,
+      "loss": 2.3816,
+      "step": 1505
+    },
+    {
+      "epoch": 0.5809064609450337,
+      "grad_norm": 0.42064681797782016,
+      "learning_rate": 3.841154036775626e-06,
+      "loss": 2.3163,
+      "step": 1506
+    },
+    {
+      "epoch": 0.5812921890067503,
+      "grad_norm": 0.4354270633194001,
+      "learning_rate": 3.8351673583383055e-06,
+      "loss": 2.3095,
+      "step": 1507
+    },
+    {
+      "epoch": 0.5816779170684667,
+      "grad_norm": 0.45282359045318643,
+      "learning_rate": 3.829182445133356e-06,
+      "loss": 2.2697,
+      "step": 1508
+    },
+    {
+      "epoch": 0.5820636451301833,
+      "grad_norm": 0.4414981823819402,
+      "learning_rate": 3.823199306230543e-06,
+      "loss": 2.3051,
+      "step": 1509
+    },
+    {
+      "epoch": 0.5824493731918997,
+      "grad_norm": 0.458287159992059,
+      "learning_rate": 3.8172179506969495e-06,
+      "loss": 2.3139,
+      "step": 1510
+    },
+    {
+      "epoch": 0.5828351012536162,
+      "grad_norm": 0.49019278650383025,
+      "learning_rate": 3.811238387596951e-06,
+      "loss": 2.3708,
+      "step": 1511
+    },
+    {
+      "epoch": 0.5832208293153327,
+      "grad_norm": 0.42237821459993796,
+      "learning_rate": 3.8052606259922097e-06,
+      "loss": 2.3203,
+      "step": 1512
+    },
+    {
+      "epoch": 0.5836065573770491,
+      "grad_norm": 0.4739834488460313,
+      "learning_rate": 3.7992846749416536e-06,
+      "loss": 2.3406,
+      "step": 1513
+    },
+    {
+      "epoch": 0.5839922854387657,
+      "grad_norm": 0.4307938570749805,
+      "learning_rate": 3.7933105435014727e-06,
+      "loss": 2.2845,
+      "step": 1514
+    },
+    {
+      "epoch": 0.5843780135004821,
+      "grad_norm": 0.45217825822154756,
+      "learning_rate": 3.787338240725095e-06,
+      "loss": 2.4072,
+      "step": 1515
+    },
+    {
+      "epoch": 0.5847637415621987,
+      "grad_norm": 0.44028592257351734,
+      "learning_rate": 3.7813677756631773e-06,
+      "loss": 2.3072,
+      "step": 1516
+    },
+    {
+      "epoch": 0.5851494696239151,
+      "grad_norm": 0.43745662531069157,
+      "learning_rate": 3.775399157363596e-06,
+      "loss": 2.3843,
+      "step": 1517
+    },
+    {
+      "epoch": 0.5855351976856317,
+      "grad_norm": 0.4136627009639538,
+      "learning_rate": 3.7694323948714223e-06,
+      "loss": 2.3225,
+      "step": 1518
+    },
+    {
+      "epoch": 0.5859209257473481,
+      "grad_norm": 0.4502576623862565,
+      "learning_rate": 3.7634674972289227e-06,
+      "loss": 2.2701,
+      "step": 1519
+    },
+    {
+      "epoch": 0.5863066538090647,
+      "grad_norm": 0.46431712852149626,
+      "learning_rate": 3.75750447347553e-06,
+      "loss": 2.2981,
+      "step": 1520
+    },
+    {
+      "epoch": 0.5866923818707811,
+      "grad_norm": 0.568350545022941,
+      "learning_rate": 3.7515433326478435e-06,
+      "loss": 2.291,
+      "step": 1521
+    },
+    {
+      "epoch": 0.5870781099324975,
+      "grad_norm": 0.45444458005389265,
+      "learning_rate": 3.745584083779604e-06,
+      "loss": 2.2699,
+      "step": 1522
+    },
+    {
+      "epoch": 0.5874638379942141,
+      "grad_norm": 0.4199450739154832,
+      "learning_rate": 3.7396267359016867e-06,
+      "loss": 2.3459,
+      "step": 1523
+    },
+    {
+      "epoch": 0.5878495660559305,
+      "grad_norm": 0.4202023743038374,
+      "learning_rate": 3.7336712980420897e-06,
+      "loss": 2.2784,
+      "step": 1524
+    },
+    {
+      "epoch": 0.5882352941176471,
+      "grad_norm": 0.46221305351333675,
+      "learning_rate": 3.727717779225912e-06,
+      "loss": 2.3683,
+      "step": 1525
+    },
+    {
+      "epoch": 0.5886210221793635,
+      "grad_norm": 0.4351128959783611,
+      "learning_rate": 3.721766188475344e-06,
+      "loss": 2.3124,
+      "step": 1526
+    },
+    {
+      "epoch": 0.5890067502410801,
+      "grad_norm": 0.4209813229158257,
+      "learning_rate": 3.71581653480966e-06,
+      "loss": 2.3296,
+      "step": 1527
+    },
+    {
+      "epoch": 0.5893924783027965,
+      "grad_norm": 0.42243311573698344,
+      "learning_rate": 3.7098688272451893e-06,
+      "loss": 2.3514,
+      "step": 1528
+    },
+    {
+      "epoch": 0.5897782063645131,
+      "grad_norm": 0.43099090382210853,
+      "learning_rate": 3.7039230747953236e-06,
+      "loss": 2.3592,
+      "step": 1529
+    },
+    {
+      "epoch": 0.5901639344262295,
+      "grad_norm": 0.4354140278443831,
+      "learning_rate": 3.6979792864704832e-06,
+      "loss": 2.2796,
+      "step": 1530
+    },
+    {
+      "epoch": 0.590549662487946,
+      "grad_norm": 0.4195720074531869,
+      "learning_rate": 3.692037471278115e-06,
+      "loss": 2.2625,
+      "step": 1531
+    },
+    {
+      "epoch": 0.5909353905496625,
+      "grad_norm": 0.4213217187229761,
+      "learning_rate": 3.6860976382226747e-06,
+      "loss": 2.2571,
+      "step": 1532
+    },
+    {
+      "epoch": 0.5913211186113789,
+      "grad_norm": 0.4359726710019312,
+      "learning_rate": 3.680159796305614e-06,
+      "loss": 2.3521,
+      "step": 1533
+    },
+    {
+      "epoch": 0.5917068466730955,
+      "grad_norm": 0.4327228331680235,
+      "learning_rate": 3.67422395452537e-06,
+      "loss": 2.4329,
+      "step": 1534
+    },
+    {
+      "epoch": 0.5920925747348119,
+      "grad_norm": 0.4576120431790627,
+      "learning_rate": 3.6682901218773446e-06,
+      "loss": 2.2853,
+      "step": 1535
+    },
+    {
+      "epoch": 0.5924783027965285,
+      "grad_norm": 0.4444798965167098,
+      "learning_rate": 3.662358307353897e-06,
+      "loss": 2.3421,
+      "step": 1536
+    },
+    {
+      "epoch": 0.5928640308582449,
+      "grad_norm": 0.4699482709499635,
+      "learning_rate": 3.6564285199443274e-06,
+      "loss": 2.3009,
+      "step": 1537
+    },
+    {
+      "epoch": 0.5932497589199615,
+      "grad_norm": 0.4306699759632948,
+      "learning_rate": 3.650500768634868e-06,
+      "loss": 2.299,
+      "step": 1538
+    },
+    {
+      "epoch": 0.5936354869816779,
+      "grad_norm": 0.45327480781505264,
+      "learning_rate": 3.64457506240866e-06,
+      "loss": 2.3126,
+      "step": 1539
+    },
+    {
+      "epoch": 0.5940212150433944,
+      "grad_norm": 0.451769483636293,
+      "learning_rate": 3.638651410245746e-06,
+      "loss": 2.3094,
+      "step": 1540
+    },
+    {
+      "epoch": 0.5944069431051109,
+      "grad_norm": 0.4721999596669721,
+      "learning_rate": 3.63272982112306e-06,
+      "loss": 2.3099,
+      "step": 1541
+    },
+    {
+      "epoch": 0.5947926711668274,
+      "grad_norm": 0.3961455522079624,
+      "learning_rate": 3.6268103040144053e-06,
+      "loss": 2.3102,
+      "step": 1542
+    },
+    {
+      "epoch": 0.5951783992285439,
+      "grad_norm": 0.4156035408585732,
+      "learning_rate": 3.6208928678904463e-06,
+      "loss": 2.353,
+      "step": 1543
+    },
+    {
+      "epoch": 0.5955641272902603,
+      "grad_norm": 0.42120161369983333,
+      "learning_rate": 3.6149775217186954e-06,
+      "loss": 2.2909,
+      "step": 1544
+    },
+    {
+      "epoch": 0.5959498553519769,
+      "grad_norm": 0.41454445097440046,
+      "learning_rate": 3.6090642744634953e-06,
+      "loss": 2.284,
+      "step": 1545
+    },
+    {
+      "epoch": 0.5963355834136933,
+      "grad_norm": 0.3927789206783352,
+      "learning_rate": 3.6031531350860094e-06,
+      "loss": 2.3265,
+      "step": 1546
+    },
+    {
+      "epoch": 0.5967213114754099,
+      "grad_norm": 0.4091783102629233,
+      "learning_rate": 3.597244112544208e-06,
+      "loss": 2.2583,
+      "step": 1547
+    },
+    {
+      "epoch": 0.5971070395371263,
+      "grad_norm": 0.4148339546574294,
+      "learning_rate": 3.5913372157928515e-06,
+      "loss": 2.2539,
+      "step": 1548
+    },
+    {
+      "epoch": 0.5974927675988428,
+      "grad_norm": 0.41068301438559474,
+      "learning_rate": 3.585432453783479e-06,
+      "loss": 2.3277,
+      "step": 1549
+    },
+    {
+      "epoch": 0.5978784956605593,
+      "grad_norm": 0.46476593432998486,
+      "learning_rate": 3.5795298354643952e-06,
+      "loss": 2.3157,
+      "step": 1550
+    },
+    {
+      "epoch": 0.5982642237222758,
+      "grad_norm": 0.45193568556648994,
+      "learning_rate": 3.573629369780656e-06,
+      "loss": 2.3242,
+      "step": 1551
+    },
+    {
+      "epoch": 0.5986499517839923,
+      "grad_norm": 0.4066513892913827,
+      "learning_rate": 3.5677310656740537e-06,
+      "loss": 2.3037,
+      "step": 1552
+    },
+    {
+      "epoch": 0.5990356798457088,
+      "grad_norm": 0.39607445972972627,
+      "learning_rate": 3.561834932083108e-06,
+      "loss": 2.2809,
+      "step": 1553
+    },
+    {
+      "epoch": 0.5994214079074253,
+      "grad_norm": 0.4259688436253283,
+      "learning_rate": 3.5559409779430467e-06,
+      "loss": 2.3362,
+      "step": 1554
+    },
+    {
+      "epoch": 0.5998071359691417,
+      "grad_norm": 0.44233653847468657,
+      "learning_rate": 3.550049212185794e-06,
+      "loss": 2.3335,
+      "step": 1555
+    },
+    {
+      "epoch": 0.6001928640308583,
+      "grad_norm": 0.40783594510679305,
+      "learning_rate": 3.5441596437399596e-06,
+      "loss": 2.3427,
+      "step": 1556
+    },
+    {
+      "epoch": 0.6005785920925747,
+      "grad_norm": 0.5203177021711052,
+      "learning_rate": 3.5382722815308256e-06,
+      "loss": 2.3203,
+      "step": 1557
+    },
+    {
+      "epoch": 0.6009643201542912,
+      "grad_norm": 0.4106556534023462,
+      "learning_rate": 3.532387134480327e-06,
+      "loss": 2.2962,
+      "step": 1558
+    },
+    {
+      "epoch": 0.6013500482160077,
+      "grad_norm": 0.4222949937445898,
+      "learning_rate": 3.526504211507041e-06,
+      "loss": 2.2522,
+      "step": 1559
+    },
+    {
+      "epoch": 0.6017357762777242,
+      "grad_norm": 0.4389117767073239,
+      "learning_rate": 3.5206235215261785e-06,
+      "loss": 2.2762,
+      "step": 1560
+    },
+    {
+      "epoch": 0.6021215043394407,
+      "grad_norm": 0.41557848167566935,
+      "learning_rate": 3.5147450734495635e-06,
+      "loss": 2.2869,
+      "step": 1561
+    },
+    {
+      "epoch": 0.6025072324011572,
+      "grad_norm": 0.42102431393771766,
+      "learning_rate": 3.5088688761856215e-06,
+      "loss": 2.2785,
+      "step": 1562
+    },
+    {
+      "epoch": 0.6028929604628737,
+      "grad_norm": 0.42418048176304096,
+      "learning_rate": 3.5029949386393725e-06,
+      "loss": 2.307,
+      "step": 1563
+    },
+    {
+      "epoch": 0.6032786885245902,
+      "grad_norm": 0.4796721034987809,
+      "learning_rate": 3.4971232697124046e-06,
+      "loss": 2.3422,
+      "step": 1564
+    },
+    {
+      "epoch": 0.6036644165863067,
+      "grad_norm": 0.4742154173234965,
+      "learning_rate": 3.491253878302873e-06,
+      "loss": 2.3337,
+      "step": 1565
+    },
+    {
+      "epoch": 0.6040501446480231,
+      "grad_norm": 0.436248194948228,
+      "learning_rate": 3.485386773305479e-06,
+      "loss": 2.3014,
+      "step": 1566
+    },
+    {
+      "epoch": 0.6044358727097396,
+      "grad_norm": 0.42815065224421,
+      "learning_rate": 3.4795219636114642e-06,
+      "loss": 2.3166,
+      "step": 1567
+    },
+    {
+      "epoch": 0.6048216007714561,
+      "grad_norm": 0.42212387633519377,
+      "learning_rate": 3.4736594581085837e-06,
+      "loss": 2.3011,
+      "step": 1568
+    },
+    {
+      "epoch": 0.6052073288331726,
+      "grad_norm": 0.45334078792920063,
+      "learning_rate": 3.4677992656811054e-06,
+      "loss": 2.3444,
+      "step": 1569
+    },
+    {
+      "epoch": 0.6055930568948891,
+      "grad_norm": 0.4549394704936667,
+      "learning_rate": 3.4619413952097925e-06,
+      "loss": 2.4325,
+      "step": 1570
+    },
+    {
+      "epoch": 0.6059787849566056,
+      "grad_norm": 0.419176742644126,
+      "learning_rate": 3.4560858555718877e-06,
+      "loss": 2.2612,
+      "step": 1571
+    },
+    {
+      "epoch": 0.6063645130183221,
+      "grad_norm": 0.40484748565623707,
+      "learning_rate": 3.4502326556411e-06,
+      "loss": 2.3178,
+      "step": 1572
+    },
+    {
+      "epoch": 0.6067502410800386,
+      "grad_norm": 0.4314659069011491,
+      "learning_rate": 3.4443818042875974e-06,
+      "loss": 2.3171,
+      "step": 1573
+    },
+    {
+      "epoch": 0.607135969141755,
+      "grad_norm": 0.4904836199860373,
+      "learning_rate": 3.438533310377985e-06,
+      "loss": 2.2796,
+      "step": 1574
+    },
+    {
+      "epoch": 0.6075216972034716,
+      "grad_norm": 0.4249998138728461,
+      "learning_rate": 3.432687182775294e-06,
+      "loss": 2.2727,
+      "step": 1575
+    },
+    {
+      "epoch": 0.607907425265188,
+      "grad_norm": 0.4707558912746351,
+      "learning_rate": 3.4268434303389747e-06,
+      "loss": 2.2528,
+      "step": 1576
+    },
+    {
+      "epoch": 0.6082931533269045,
+      "grad_norm": 0.40487788077121967,
+      "learning_rate": 3.4210020619248762e-06,
+      "loss": 2.3378,
+      "step": 1577
+    },
+    {
+      "epoch": 0.608678881388621,
+      "grad_norm": 0.3928419411927058,
+      "learning_rate": 3.4151630863852315e-06,
+      "loss": 2.3446,
+      "step": 1578
+    },
+    {
+      "epoch": 0.6090646094503375,
+      "grad_norm": 0.4121647963429338,
+      "learning_rate": 3.4093265125686494e-06,
+      "loss": 2.3299,
+      "step": 1579
+    },
+    {
+      "epoch": 0.609450337512054,
+      "grad_norm": 0.4479085650162451,
+      "learning_rate": 3.403492349320101e-06,
+      "loss": 2.3532,
+      "step": 1580
+    },
+    {
+      "epoch": 0.6098360655737705,
+      "grad_norm": 0.4142417050318523,
+      "learning_rate": 3.3976606054809015e-06,
+      "loss": 2.3442,
+      "step": 1581
+    },
+    {
+      "epoch": 0.610221793635487,
+      "grad_norm": 0.42194842190876514,
+      "learning_rate": 3.391831289888701e-06,
+      "loss": 2.3086,
+      "step": 1582
+    },
+    {
+      "epoch": 0.6106075216972034,
+      "grad_norm": 0.42552892598640235,
+      "learning_rate": 3.38600441137747e-06,
+      "loss": 2.3675,
+      "step": 1583
+    },
+    {
+      "epoch": 0.61099324975892,
+      "grad_norm": 0.4451008561962209,
+      "learning_rate": 3.380179978777482e-06,
+      "loss": 2.3745,
+      "step": 1584
+    },
+    {
+      "epoch": 0.6113789778206364,
+      "grad_norm": 0.4307186268643373,
+      "learning_rate": 3.3743580009153122e-06,
+      "loss": 2.3006,
+      "step": 1585
+    },
+    {
+      "epoch": 0.611764705882353,
+      "grad_norm": 0.41560471224439033,
+      "learning_rate": 3.368538486613807e-06,
+      "loss": 2.3105,
+      "step": 1586
+    },
+    {
+      "epoch": 0.6121504339440694,
+      "grad_norm": 0.4508243452795475,
+      "learning_rate": 3.362721444692086e-06,
+      "loss": 2.3381,
+      "step": 1587
+    },
+    {
+      "epoch": 0.6125361620057859,
+      "grad_norm": 0.3995941055290941,
+      "learning_rate": 3.356906883965516e-06,
+      "loss": 2.3195,
+      "step": 1588
+    },
+    {
+      "epoch": 0.6129218900675024,
+      "grad_norm": 0.41518630090251907,
+      "learning_rate": 3.3510948132457087e-06,
+      "loss": 2.2516,
+      "step": 1589
+    },
+    {
+      "epoch": 0.6133076181292189,
+      "grad_norm": 0.40103288319382635,
+      "learning_rate": 3.3452852413405014e-06,
+      "loss": 2.2795,
+      "step": 1590
+    },
+    {
+      "epoch": 0.6136933461909354,
+      "grad_norm": 0.4185829392452625,
+      "learning_rate": 3.3394781770539406e-06,
+      "loss": 2.2732,
+      "step": 1591
+    },
+    {
+      "epoch": 0.6140790742526518,
+      "grad_norm": 0.4689714856626146,
+      "learning_rate": 3.33367362918628e-06,
+      "loss": 2.3501,
+      "step": 1592
+    },
+    {
+      "epoch": 0.6144648023143684,
+      "grad_norm": 0.4360872226256897,
+      "learning_rate": 3.32787160653395e-06,
+      "loss": 2.4018,
+      "step": 1593
+    },
+    {
+      "epoch": 0.6148505303760848,
+      "grad_norm": 0.447666636095158,
+      "learning_rate": 3.3220721178895658e-06,
+      "loss": 2.3115,
+      "step": 1594
+    },
+    {
+      "epoch": 0.6152362584378014,
+      "grad_norm": 0.4214603591205362,
+      "learning_rate": 3.316275172041893e-06,
+      "loss": 2.2318,
+      "step": 1595
+    },
+    {
+      "epoch": 0.6156219864995178,
+      "grad_norm": 0.41757506919655035,
+      "learning_rate": 3.3104807777758487e-06,
+      "loss": 2.3058,
+      "step": 1596
+    },
+    {
+      "epoch": 0.6160077145612344,
+      "grad_norm": 0.41339639782568854,
+      "learning_rate": 3.3046889438724805e-06,
+      "loss": 2.2834,
+      "step": 1597
+    },
+    {
+      "epoch": 0.6163934426229508,
+      "grad_norm": 0.4464343403324317,
+      "learning_rate": 3.2988996791089573e-06,
+      "loss": 2.2788,
+      "step": 1598
+    },
+    {
+      "epoch": 0.6167791706846673,
+      "grad_norm": 0.40466299123420946,
+      "learning_rate": 3.293112992258556e-06,
+      "loss": 2.2974,
+      "step": 1599
+    },
+    {
+      "epoch": 0.6171648987463838,
+      "grad_norm": 0.40325913870382823,
+      "learning_rate": 3.2873288920906436e-06,
+      "loss": 2.242,
+      "step": 1600
+    },
+    {
+      "epoch": 0.6175506268081002,
+      "grad_norm": 0.42192819850604385,
+      "learning_rate": 3.2815473873706696e-06,
+      "loss": 2.3584,
+      "step": 1601
+    },
+    {
+      "epoch": 0.6179363548698168,
+      "grad_norm": 0.42561536777423264,
+      "learning_rate": 3.275768486860149e-06,
+      "loss": 2.3795,
+      "step": 1602
+    },
+    {
+      "epoch": 0.6183220829315332,
+      "grad_norm": 0.42857073166383797,
+      "learning_rate": 3.2699921993166508e-06,
+      "loss": 2.3295,
+      "step": 1603
+    },
+    {
+      "epoch": 0.6187078109932498,
+      "grad_norm": 0.4066147316249197,
+      "learning_rate": 3.2642185334937853e-06,
+      "loss": 2.2482,
+      "step": 1604
+    },
+    {
+      "epoch": 0.6190935390549662,
+      "grad_norm": 0.44300196333196484,
+      "learning_rate": 3.2584474981411874e-06,
+      "loss": 2.3715,
+      "step": 1605
+    },
+    {
+      "epoch": 0.6194792671166828,
+      "grad_norm": 0.4148539593484898,
+      "learning_rate": 3.252679102004509e-06,
+      "loss": 2.299,
+      "step": 1606
+    },
+    {
+      "epoch": 0.6198649951783992,
+      "grad_norm": 0.43327244941421494,
+      "learning_rate": 3.2469133538253983e-06,
+      "loss": 2.3321,
+      "step": 1607
+    },
+    {
+      "epoch": 0.6202507232401158,
+      "grad_norm": 0.4449484874248776,
+      "learning_rate": 3.2411502623414925e-06,
+      "loss": 2.2933,
+      "step": 1608
+    },
+    {
+      "epoch": 0.6206364513018322,
+      "grad_norm": 0.4342425826361147,
+      "learning_rate": 3.2353898362864055e-06,
+      "loss": 2.3152,
+      "step": 1609
+    },
+    {
+      "epoch": 0.6210221793635486,
+      "grad_norm": 0.4821905887515583,
+      "learning_rate": 3.229632084389708e-06,
+      "loss": 2.3511,
+      "step": 1610
+    },
+    {
+      "epoch": 0.6214079074252652,
+      "grad_norm": 0.45075055098796424,
+      "learning_rate": 3.2238770153769173e-06,
+      "loss": 2.2534,
+      "step": 1611
+    },
+    {
+      "epoch": 0.6217936354869816,
+      "grad_norm": 0.43457470407406457,
+      "learning_rate": 3.2181246379694886e-06,
+      "loss": 2.3579,
+      "step": 1612
+    },
+    {
+      "epoch": 0.6221793635486982,
+      "grad_norm": 0.4398445075795343,
+      "learning_rate": 3.2123749608847998e-06,
+      "loss": 2.3447,
+      "step": 1613
+    },
+    {
+      "epoch": 0.6225650916104146,
+      "grad_norm": 0.4357463212834921,
+      "learning_rate": 3.206627992836131e-06,
+      "loss": 2.2671,
+      "step": 1614
+    },
+    {
+      "epoch": 0.6229508196721312,
+      "grad_norm": 0.4602686135319776,
+      "learning_rate": 3.200883742532659e-06,
+      "loss": 2.3401,
+      "step": 1615
+    },
+    {
+      "epoch": 0.6233365477338476,
+      "grad_norm": 0.4498777315680553,
+      "learning_rate": 3.1951422186794447e-06,
+      "loss": 2.3381,
+      "step": 1616
+    },
+    {
+      "epoch": 0.6237222757955642,
+      "grad_norm": 0.45243899257172177,
+      "learning_rate": 3.1894034299774125e-06,
+      "loss": 2.3527,
+      "step": 1617
+    },
+    {
+      "epoch": 0.6241080038572806,
+      "grad_norm": 0.4125076055492351,
+      "learning_rate": 3.183667385123346e-06,
+      "loss": 2.3594,
+      "step": 1618
+    },
+    {
+      "epoch": 0.6244937319189972,
+      "grad_norm": 0.41444805052369155,
+      "learning_rate": 3.1779340928098695e-06,
+      "loss": 2.2522,
+      "step": 1619
+    },
+    {
+      "epoch": 0.6248794599807136,
+      "grad_norm": 0.4402574577616683,
+      "learning_rate": 3.1722035617254333e-06,
+      "loss": 2.3001,
+      "step": 1620
+    },
+    {
+      "epoch": 0.6252651880424301,
+      "grad_norm": 0.39364272113708715,
+      "learning_rate": 3.1664758005543072e-06,
+      "loss": 2.2133,
+      "step": 1621
+    },
+    {
+      "epoch": 0.6256509161041466,
+      "grad_norm": 0.4447946352808919,
+      "learning_rate": 3.160750817976562e-06,
+      "loss": 2.3014,
+      "step": 1622
+    },
+    {
+      "epoch": 0.626036644165863,
+      "grad_norm": 0.41953913275683713,
+      "learning_rate": 3.1550286226680576e-06,
+      "loss": 2.3584,
+      "step": 1623
+    },
+    {
+      "epoch": 0.6264223722275796,
+      "grad_norm": 0.40785748215954465,
+      "learning_rate": 3.149309223300428e-06,
+      "loss": 2.3599,
+      "step": 1624
+    },
+    {
+      "epoch": 0.626808100289296,
+      "grad_norm": 0.4490579214152913,
+      "learning_rate": 3.1435926285410747e-06,
+      "loss": 2.4041,
+      "step": 1625
+    },
+    {
+      "epoch": 0.6271938283510126,
+      "grad_norm": 0.4127887227261846,
+      "learning_rate": 3.137878847053143e-06,
+      "loss": 2.3671,
+      "step": 1626
+    },
+    {
+      "epoch": 0.627579556412729,
+      "grad_norm": 0.4481331305889832,
+      "learning_rate": 3.1321678874955193e-06,
+      "loss": 2.3999,
+      "step": 1627
+    },
+    {
+      "epoch": 0.6279652844744456,
+      "grad_norm": 0.4150063152135012,
+      "learning_rate": 3.126459758522813e-06,
+      "loss": 2.2777,
+      "step": 1628
+    },
+    {
+      "epoch": 0.628351012536162,
+      "grad_norm": 0.449950999952686,
+      "learning_rate": 3.120754468785343e-06,
+      "loss": 2.2863,
+      "step": 1629
+    },
+    {
+      "epoch": 0.6287367405978785,
+      "grad_norm": 0.4104039989836355,
+      "learning_rate": 3.115052026929124e-06,
+      "loss": 2.274,
+      "step": 1630
+    },
+    {
+      "epoch": 0.629122468659595,
+      "grad_norm": 0.4253221247881812,
+      "learning_rate": 3.1093524415958576e-06,
+      "loss": 2.3231,
+      "step": 1631
+    },
+    {
+      "epoch": 0.6295081967213115,
+      "grad_norm": 0.441253477244187,
+      "learning_rate": 3.103655721422917e-06,
+      "loss": 2.2865,
+      "step": 1632
+    },
+    {
+      "epoch": 0.629893924783028,
+      "grad_norm": 0.434903455606142,
+      "learning_rate": 3.097961875043331e-06,
+      "loss": 2.3141,
+      "step": 1633
+    },
+    {
+      "epoch": 0.6302796528447444,
+      "grad_norm": 0.402970101599088,
+      "learning_rate": 3.0922709110857727e-06,
+      "loss": 2.3473,
+      "step": 1634
+    },
+    {
+      "epoch": 0.630665380906461,
+      "grad_norm": 0.5255032149638728,
+      "learning_rate": 3.0865828381745515e-06,
+      "loss": 2.3338,
+      "step": 1635
+    },
+    {
+      "epoch": 0.6310511089681774,
+      "grad_norm": 0.5375959149946876,
+      "learning_rate": 3.080897664929592e-06,
+      "loss": 2.3365,
+      "step": 1636
+    },
+    {
+      "epoch": 0.631436837029894,
+      "grad_norm": 0.4437236135709563,
+      "learning_rate": 3.0752153999664225e-06,
+      "loss": 2.3017,
+      "step": 1637
+    },
+    {
+      "epoch": 0.6318225650916104,
+      "grad_norm": 0.4140603788140724,
+      "learning_rate": 3.06953605189617e-06,
+      "loss": 2.2803,
+      "step": 1638
+    },
+    {
+      "epoch": 0.6322082931533269,
+      "grad_norm": 0.40732890731601284,
+      "learning_rate": 3.0638596293255368e-06,
+      "loss": 2.3403,
+      "step": 1639
+    },
+    {
+      "epoch": 0.6325940212150434,
+      "grad_norm": 0.4480598498633505,
+      "learning_rate": 3.0581861408567907e-06,
+      "loss": 2.3276,
+      "step": 1640
+    },
+    {
+      "epoch": 0.6329797492767599,
+      "grad_norm": 0.4449957529966464,
+      "learning_rate": 3.052515595087759e-06,
+      "loss": 2.3904,
+      "step": 1641
+    },
+    {
+      "epoch": 0.6333654773384764,
+      "grad_norm": 0.44263014545792057,
+      "learning_rate": 3.0468480006118045e-06,
+      "loss": 2.3563,
+      "step": 1642
+    },
+    {
+      "epoch": 0.6337512054001929,
+      "grad_norm": 0.40308659294775584,
+      "learning_rate": 3.041183366017818e-06,
+      "loss": 2.3224,
+      "step": 1643
+    },
+    {
+      "epoch": 0.6341369334619094,
+      "grad_norm": 0.47183380223311655,
+      "learning_rate": 3.035521699890206e-06,
+      "loss": 2.3157,
+      "step": 1644
+    },
+    {
+      "epoch": 0.6345226615236258,
+      "grad_norm": 0.41576466733189876,
+      "learning_rate": 3.029863010808876e-06,
+      "loss": 2.3774,
+      "step": 1645
+    },
+    {
+      "epoch": 0.6349083895853423,
+      "grad_norm": 0.4361736521701304,
+      "learning_rate": 3.0242073073492238e-06,
+      "loss": 2.3852,
+      "step": 1646
+    },
+    {
+      "epoch": 0.6352941176470588,
+      "grad_norm": 0.43135315720350514,
+      "learning_rate": 3.0185545980821207e-06,
+      "loss": 2.3221,
+      "step": 1647
+    },
+    {
+      "epoch": 0.6356798457087753,
+      "grad_norm": 0.4385687708863605,
+      "learning_rate": 3.0129048915739013e-06,
+      "loss": 2.3825,
+      "step": 1648
+    },
+    {
+      "epoch": 0.6360655737704918,
+      "grad_norm": 0.4416969053883667,
+      "learning_rate": 3.007258196386347e-06,
+      "loss": 2.2408,
+      "step": 1649
+    },
+    {
+      "epoch": 0.6364513018322083,
+      "grad_norm": 0.4128777848509827,
+      "learning_rate": 3.00161452107668e-06,
+      "loss": 2.3332,
+      "step": 1650
+    },
+    {
+      "epoch": 0.6368370298939248,
+      "grad_norm": 0.453816183858808,
+      "learning_rate": 2.9959738741975426e-06,
+      "loss": 2.3545,
+      "step": 1651
+    },
+    {
+      "epoch": 0.6372227579556413,
+      "grad_norm": 0.40964967583845385,
+      "learning_rate": 2.9903362642969903e-06,
+      "loss": 2.2743,
+      "step": 1652
+    },
+    {
+      "epoch": 0.6376084860173578,
+      "grad_norm": 0.4417719523403045,
+      "learning_rate": 2.9847016999184746e-06,
+      "loss": 2.3381,
+      "step": 1653
+    },
+    {
+      "epoch": 0.6379942140790743,
+      "grad_norm": 0.41947075176552073,
+      "learning_rate": 2.97907018960083e-06,
+      "loss": 2.3772,
+      "step": 1654
+    },
+    {
+      "epoch": 0.6383799421407907,
+      "grad_norm": 0.4365891844676327,
+      "learning_rate": 2.9734417418782667e-06,
+      "loss": 2.3256,
+      "step": 1655
+    },
+    {
+      "epoch": 0.6387656702025072,
+      "grad_norm": 0.4303482492434974,
+      "learning_rate": 2.967816365280351e-06,
+      "loss": 2.3813,
+      "step": 1656
+    },
+    {
+      "epoch": 0.6391513982642237,
+      "grad_norm": 0.4363287258527234,
+      "learning_rate": 2.962194068331996e-06,
+      "loss": 2.3038,
+      "step": 1657
+    },
+    {
+      "epoch": 0.6395371263259402,
+      "grad_norm": 0.432977695465348,
+      "learning_rate": 2.956574859553448e-06,
+      "loss": 2.3224,
+      "step": 1658
+    },
+    {
+      "epoch": 0.6399228543876567,
+      "grad_norm": 0.47593063120339796,
+      "learning_rate": 2.9509587474602707e-06,
+      "loss": 2.3044,
+      "step": 1659
+    },
+    {
+      "epoch": 0.6403085824493732,
+      "grad_norm": 0.44002954822773555,
+      "learning_rate": 2.94534574056334e-06,
+      "loss": 2.2886,
+      "step": 1660
+    },
+    {
+      "epoch": 0.6406943105110897,
+      "grad_norm": 0.40756469467426576,
+      "learning_rate": 2.9397358473688232e-06,
+      "loss": 2.318,
+      "step": 1661
+    },
+    {
+      "epoch": 0.6410800385728062,
+      "grad_norm": 0.44199679711135886,
+      "learning_rate": 2.934129076378168e-06,
+      "loss": 2.3477,
+      "step": 1662
+    },
+    {
+      "epoch": 0.6414657666345227,
+      "grad_norm": 0.4359941554828112,
+      "learning_rate": 2.9285254360880922e-06,
+      "loss": 2.2384,
+      "step": 1663
+    },
+    {
+      "epoch": 0.6418514946962391,
+      "grad_norm": 0.38720988764246606,
+      "learning_rate": 2.9229249349905686e-06,
+      "loss": 2.311,
+      "step": 1664
+    },
+    {
+      "epoch": 0.6422372227579557,
+      "grad_norm": 0.46557718087737565,
+      "learning_rate": 2.917327581572812e-06,
+      "loss": 2.3703,
+      "step": 1665
+    },
+    {
+      "epoch": 0.6426229508196721,
+      "grad_norm": 0.3984082593840511,
+      "learning_rate": 2.911733384317269e-06,
+      "loss": 2.3391,
+      "step": 1666
+    },
+    {
+      "epoch": 0.6430086788813886,
+      "grad_norm": 0.42206119674386167,
+      "learning_rate": 2.906142351701603e-06,
+      "loss": 2.3009,
+      "step": 1667
+    },
+    {
+      "epoch": 0.6433944069431051,
+      "grad_norm": 0.4217609862574355,
+      "learning_rate": 2.9005544921986774e-06,
+      "loss": 2.3729,
+      "step": 1668
+    },
+    {
+      "epoch": 0.6437801350048216,
+      "grad_norm": 0.39429987610744055,
+      "learning_rate": 2.8949698142765535e-06,
+      "loss": 2.268,
+      "step": 1669
+    },
+    {
+      "epoch": 0.6441658630665381,
+      "grad_norm": 0.4164481989825121,
+      "learning_rate": 2.889388326398468e-06,
+      "loss": 2.3098,
+      "step": 1670
+    },
+    {
+      "epoch": 0.6445515911282546,
+      "grad_norm": 0.40129608423507873,
+      "learning_rate": 2.8838100370228213e-06,
+      "loss": 2.3361,
+      "step": 1671
+    },
+    {
+      "epoch": 0.6449373191899711,
+      "grad_norm": 0.4089424621249286,
+      "learning_rate": 2.8782349546031673e-06,
+      "loss": 2.3635,
+      "step": 1672
+    },
+    {
+      "epoch": 0.6453230472516875,
+      "grad_norm": 0.4240294298016162,
+      "learning_rate": 2.8726630875882056e-06,
+      "loss": 2.2852,
+      "step": 1673
+    },
+    {
+      "epoch": 0.6457087753134041,
+      "grad_norm": 0.4290293267118307,
+      "learning_rate": 2.867094444421756e-06,
+      "loss": 2.2841,
+      "step": 1674
+    },
+    {
+      "epoch": 0.6460945033751205,
+      "grad_norm": 0.4097212990786284,
+      "learning_rate": 2.861529033542756e-06,
+      "loss": 2.3201,
+      "step": 1675
+    },
+    {
+      "epoch": 0.6464802314368371,
+      "grad_norm": 0.40823257837484095,
+      "learning_rate": 2.8559668633852433e-06,
+      "loss": 2.3387,
+      "step": 1676
+    },
+    {
+      "epoch": 0.6468659594985535,
+      "grad_norm": 0.4291351890757635,
+      "learning_rate": 2.8504079423783443e-06,
+      "loss": 2.3246,
+      "step": 1677
+    },
+    {
+      "epoch": 0.64725168756027,
+      "grad_norm": 0.4506378274791692,
+      "learning_rate": 2.844852278946264e-06,
+      "loss": 2.3413,
+      "step": 1678
+    },
+    {
+      "epoch": 0.6476374156219865,
+      "grad_norm": 0.4858719061199603,
+      "learning_rate": 2.839299881508272e-06,
+      "loss": 2.3755,
+      "step": 1679
+    },
+    {
+      "epoch": 0.648023143683703,
+      "grad_norm": 0.4387548340054685,
+      "learning_rate": 2.8337507584786826e-06,
+      "loss": 2.2976,
+      "step": 1680
+    },
+    {
+      "epoch": 0.6484088717454195,
+      "grad_norm": 0.40231666835311036,
+      "learning_rate": 2.828204918266852e-06,
+      "loss": 2.3032,
+      "step": 1681
+    },
+    {
+      "epoch": 0.6487945998071359,
+      "grad_norm": 0.4303129513109538,
+      "learning_rate": 2.8226623692771605e-06,
+      "loss": 2.3042,
+      "step": 1682
+    },
+    {
+      "epoch": 0.6491803278688525,
+      "grad_norm": 0.44567862759054966,
+      "learning_rate": 2.817123119909001e-06,
+      "loss": 2.3626,
+      "step": 1683
+    },
+    {
+      "epoch": 0.6495660559305689,
+      "grad_norm": 0.41045229346005024,
+      "learning_rate": 2.811587178556764e-06,
+      "loss": 2.3123,
+      "step": 1684
+    },
+    {
+      "epoch": 0.6499517839922855,
+      "grad_norm": 0.42060750551913056,
+      "learning_rate": 2.8060545536098314e-06,
+      "loss": 2.3444,
+      "step": 1685
+    },
+    {
+      "epoch": 0.6503375120540019,
+      "grad_norm": 0.4603147009817396,
+      "learning_rate": 2.800525253452557e-06,
+      "loss": 2.3029,
+      "step": 1686
+    },
+    {
+      "epoch": 0.6507232401157185,
+      "grad_norm": 0.4452232135375676,
+      "learning_rate": 2.794999286464253e-06,
+      "loss": 2.344,
+      "step": 1687
+    },
+    {
+      "epoch": 0.6511089681774349,
+      "grad_norm": 0.4166776719874048,
+      "learning_rate": 2.789476661019186e-06,
+      "loss": 2.3088,
+      "step": 1688
+    },
+    {
+      "epoch": 0.6514946962391513,
+      "grad_norm": 0.4005437799726485,
+      "learning_rate": 2.7839573854865555e-06,
+      "loss": 2.3429,
+      "step": 1689
+    },
+    {
+      "epoch": 0.6518804243008679,
+      "grad_norm": 0.44743394893607225,
+      "learning_rate": 2.778441468230483e-06,
+      "loss": 2.3008,
+      "step": 1690
+    },
+    {
+      "epoch": 0.6522661523625843,
+      "grad_norm": 0.42389007114177785,
+      "learning_rate": 2.7729289176100026e-06,
+      "loss": 2.2756,
+      "step": 1691
+    },
+    {
+      "epoch": 0.6526518804243009,
+      "grad_norm": 0.4551381436945548,
+      "learning_rate": 2.7674197419790493e-06,
+      "loss": 2.3439,
+      "step": 1692
+    },
+    {
+      "epoch": 0.6530376084860173,
+      "grad_norm": 0.4186949915013729,
+      "learning_rate": 2.761913949686438e-06,
+      "loss": 2.3103,
+      "step": 1693
+    },
+    {
+      "epoch": 0.6534233365477339,
+      "grad_norm": 0.4509372293751877,
+      "learning_rate": 2.75641154907586e-06,
+      "loss": 2.2772,
+      "step": 1694
+    },
+    {
+      "epoch": 0.6538090646094503,
+      "grad_norm": 0.46075922985731843,
+      "learning_rate": 2.7509125484858657e-06,
+      "loss": 2.2995,
+      "step": 1695
+    },
+    {
+      "epoch": 0.6541947926711669,
+      "grad_norm": 0.4351647353450828,
+      "learning_rate": 2.7454169562498503e-06,
+      "loss": 2.3232,
+      "step": 1696
+    },
+    {
+      "epoch": 0.6545805207328833,
+      "grad_norm": 0.44116773382830016,
+      "learning_rate": 2.73992478069605e-06,
+      "loss": 2.3232,
+      "step": 1697
+    },
+    {
+      "epoch": 0.6549662487945999,
+      "grad_norm": 0.49166533573126286,
+      "learning_rate": 2.734436030147517e-06,
+      "loss": 2.3415,
+      "step": 1698
+    },
+    {
+      "epoch": 0.6553519768563163,
+      "grad_norm": 0.4109900880916673,
+      "learning_rate": 2.72895071292212e-06,
+      "loss": 2.2507,
+      "step": 1699
+    },
+    {
+      "epoch": 0.6557377049180327,
+      "grad_norm": 0.4523684680383627,
+      "learning_rate": 2.723468837332517e-06,
+      "loss": 2.3552,
+      "step": 1700
+    },
+    {
+      "epoch": 0.6561234329797493,
+      "grad_norm": 0.4515322104742252,
+      "learning_rate": 2.7179904116861557e-06,
+      "loss": 2.3195,
+      "step": 1701
+    },
+    {
+      "epoch": 0.6565091610414657,
+      "grad_norm": 0.524366845606669,
+      "learning_rate": 2.712515444285253e-06,
+      "loss": 2.3239,
+      "step": 1702
+    },
+    {
+      "epoch": 0.6568948891031823,
+      "grad_norm": 0.4187010538839707,
+      "learning_rate": 2.707043943426786e-06,
+      "loss": 2.3485,
+      "step": 1703
+    },
+    {
+      "epoch": 0.6572806171648987,
+      "grad_norm": 0.43658345945129773,
+      "learning_rate": 2.7015759174024756e-06,
+      "loss": 2.3283,
+      "step": 1704
+    },
+    {
+      "epoch": 0.6576663452266153,
+      "grad_norm": 0.4146289618702742,
+      "learning_rate": 2.6961113744987854e-06,
+      "loss": 2.2828,
+      "step": 1705
+    },
+    {
+      "epoch": 0.6580520732883317,
+      "grad_norm": 0.43820230295987506,
+      "learning_rate": 2.6906503229968895e-06,
+      "loss": 2.2593,
+      "step": 1706
+    },
+    {
+      "epoch": 0.6584378013500483,
+      "grad_norm": 0.42122366510118364,
+      "learning_rate": 2.6851927711726807e-06,
+      "loss": 2.3358,
+      "step": 1707
+    },
+    {
+      "epoch": 0.6588235294117647,
+      "grad_norm": 0.4117770533764119,
+      "learning_rate": 2.6797387272967414e-06,
+      "loss": 2.3342,
+      "step": 1708
+    },
+    {
+      "epoch": 0.6592092574734812,
+      "grad_norm": 0.4716978041525038,
+      "learning_rate": 2.6742881996343405e-06,
+      "loss": 2.3389,
+      "step": 1709
+    },
+    {
+      "epoch": 0.6595949855351977,
+      "grad_norm": 0.4340611250787719,
+      "learning_rate": 2.668841196445416e-06,
+      "loss": 2.3282,
+      "step": 1710
+    },
+    {
+      "epoch": 0.6599807135969141,
+      "grad_norm": 0.4396181233847713,
+      "learning_rate": 2.6633977259845715e-06,
+      "loss": 2.2852,
+      "step": 1711
+    },
+    {
+      "epoch": 0.6603664416586307,
+      "grad_norm": 0.41139067151993575,
+      "learning_rate": 2.65795779650105e-06,
+      "loss": 2.3336,
+      "step": 1712
+    },
+    {
+      "epoch": 0.6607521697203471,
+      "grad_norm": 0.4140739732396573,
+      "learning_rate": 2.65252141623873e-06,
+      "loss": 2.3266,
+      "step": 1713
+    },
+    {
+      "epoch": 0.6611378977820637,
+      "grad_norm": 0.4433489124410752,
+      "learning_rate": 2.6470885934361136e-06,
+      "loss": 2.2896,
+      "step": 1714
+    },
+    {
+      "epoch": 0.6615236258437801,
+      "grad_norm": 0.4471073261840135,
+      "learning_rate": 2.6416593363263067e-06,
+      "loss": 2.32,
+      "step": 1715
+    },
+    {
+      "epoch": 0.6619093539054967,
+      "grad_norm": 0.4188281593563816,
+      "learning_rate": 2.63623365313702e-06,
+      "loss": 2.2587,
+      "step": 1716
+    },
+    {
+      "epoch": 0.6622950819672131,
+      "grad_norm": 0.4196333926973484,
+      "learning_rate": 2.6308115520905396e-06,
+      "loss": 2.3181,
+      "step": 1717
+    },
+    {
+      "epoch": 0.6626808100289296,
+      "grad_norm": 0.43507817332036064,
+      "learning_rate": 2.625393041403731e-06,
+      "loss": 2.3033,
+      "step": 1718
+    },
+    {
+      "epoch": 0.6630665380906461,
+      "grad_norm": 0.4013952383118774,
+      "learning_rate": 2.619978129288011e-06,
+      "loss": 2.3137,
+      "step": 1719
+    },
+    {
+      "epoch": 0.6634522661523626,
+      "grad_norm": 0.4352226375058755,
+      "learning_rate": 2.614566823949348e-06,
+      "loss": 2.3434,
+      "step": 1720
+    },
+    {
+      "epoch": 0.6638379942140791,
+      "grad_norm": 0.46138167395279756,
+      "learning_rate": 2.6091591335882426e-06,
+      "loss": 2.2565,
+      "step": 1721
+    },
+    {
+      "epoch": 0.6642237222757955,
+      "grad_norm": 0.4315324035523396,
+      "learning_rate": 2.603755066399718e-06,
+      "loss": 2.2905,
+      "step": 1722
+    },
+    {
+      "epoch": 0.6646094503375121,
+      "grad_norm": 0.4248660288503237,
+      "learning_rate": 2.598354630573303e-06,
+      "loss": 2.3645,
+      "step": 1723
+    },
+    {
+      "epoch": 0.6649951783992285,
+      "grad_norm": 0.4627762735762227,
+      "learning_rate": 2.592957834293033e-06,
+      "loss": 2.3146,
+      "step": 1724
+    },
+    {
+      "epoch": 0.665380906460945,
+      "grad_norm": 0.42172579815965766,
+      "learning_rate": 2.5875646857374147e-06,
+      "loss": 2.3384,
+      "step": 1725
+    },
+    {
+      "epoch": 0.6657666345226615,
+      "grad_norm": 0.43837623448336754,
+      "learning_rate": 2.5821751930794404e-06,
+      "loss": 2.3123,
+      "step": 1726
+    },
+    {
+      "epoch": 0.666152362584378,
+      "grad_norm": 0.44426184885159165,
+      "learning_rate": 2.576789364486551e-06,
+      "loss": 2.3464,
+      "step": 1727
+    },
+    {
+      "epoch": 0.6665380906460945,
+      "grad_norm": 0.4087896213363994,
+      "learning_rate": 2.5714072081206407e-06,
+      "loss": 2.3194,
+      "step": 1728
+    },
+    {
+      "epoch": 0.666923818707811,
+      "grad_norm": 0.4309466855847297,
+      "learning_rate": 2.566028732138037e-06,
+      "loss": 2.3609,
+      "step": 1729
+    },
+    {
+      "epoch": 0.6673095467695275,
+      "grad_norm": 0.41812542408527814,
+      "learning_rate": 2.5606539446894875e-06,
+      "loss": 2.3093,
+      "step": 1730
+    },
+    {
+      "epoch": 0.667695274831244,
+      "grad_norm": 0.4197042581795082,
+      "learning_rate": 2.5552828539201563e-06,
+      "loss": 2.3305,
+      "step": 1731
+    },
+    {
+      "epoch": 0.6680810028929605,
+      "grad_norm": 0.44633494094061993,
+      "learning_rate": 2.5499154679696014e-06,
+      "loss": 2.3182,
+      "step": 1732
+    },
+    {
+      "epoch": 0.6684667309546769,
+      "grad_norm": 0.4093822342538184,
+      "learning_rate": 2.5445517949717645e-06,
+      "loss": 2.3515,
+      "step": 1733
+    },
+    {
+      "epoch": 0.6688524590163935,
+      "grad_norm": 0.4749740056226905,
+      "learning_rate": 2.5391918430549635e-06,
+      "loss": 2.3825,
+      "step": 1734
+    },
+    {
+      "epoch": 0.6692381870781099,
+      "grad_norm": 0.40362186487722646,
+      "learning_rate": 2.5338356203418784e-06,
+      "loss": 2.2753,
+      "step": 1735
+    },
+    {
+      "epoch": 0.6696239151398264,
+      "grad_norm": 0.40874498875645543,
+      "learning_rate": 2.528483134949535e-06,
+      "loss": 2.3529,
+      "step": 1736
+    },
+    {
+      "epoch": 0.6700096432015429,
+      "grad_norm": 0.45349066049912623,
+      "learning_rate": 2.523134394989294e-06,
+      "loss": 2.3471,
+      "step": 1737
+    },
+    {
+      "epoch": 0.6703953712632594,
+      "grad_norm": 0.43166929648623414,
+      "learning_rate": 2.517789408566846e-06,
+      "loss": 2.3572,
+      "step": 1738
+    },
+    {
+      "epoch": 0.6707810993249759,
+      "grad_norm": 0.4240149403537371,
+      "learning_rate": 2.5124481837821886e-06,
+      "loss": 2.3522,
+      "step": 1739
+    },
+    {
+      "epoch": 0.6711668273866924,
+      "grad_norm": 0.4235013301612557,
+      "learning_rate": 2.50711072872962e-06,
+      "loss": 2.2818,
+      "step": 1740
+    },
+    {
+      "epoch": 0.6715525554484089,
+      "grad_norm": 0.46681872751640835,
+      "learning_rate": 2.5017770514977252e-06,
+      "loss": 2.3037,
+      "step": 1741
+    },
+    {
+      "epoch": 0.6719382835101254,
+      "grad_norm": 0.4312151667630878,
+      "learning_rate": 2.4964471601693633e-06,
+      "loss": 2.3269,
+      "step": 1742
+    },
+    {
+      "epoch": 0.6723240115718419,
+      "grad_norm": 0.480321202635658,
+      "learning_rate": 2.4911210628216615e-06,
+      "loss": 2.3542,
+      "step": 1743
+    },
+    {
+      "epoch": 0.6727097396335583,
+      "grad_norm": 0.4161546691919782,
+      "learning_rate": 2.4857987675259887e-06,
+      "loss": 2.3684,
+      "step": 1744
+    },
+    {
+      "epoch": 0.6730954676952748,
+      "grad_norm": 0.4334867788444831,
+      "learning_rate": 2.480480282347961e-06,
+      "loss": 2.232,
+      "step": 1745
+    },
+    {
+      "epoch": 0.6734811957569913,
+      "grad_norm": 0.4205179198468628,
+      "learning_rate": 2.4751656153474147e-06,
+      "loss": 2.3038,
+      "step": 1746
+    },
+    {
+      "epoch": 0.6738669238187078,
+      "grad_norm": 0.4369725705928331,
+      "learning_rate": 2.4698547745784014e-06,
+      "loss": 2.3228,
+      "step": 1747
+    },
+    {
+      "epoch": 0.6742526518804243,
+      "grad_norm": 0.4527242369582889,
+      "learning_rate": 2.4645477680891734e-06,
+      "loss": 2.296,
+      "step": 1748
+    },
+    {
+      "epoch": 0.6746383799421408,
+      "grad_norm": 0.4206894542970612,
+      "learning_rate": 2.4592446039221718e-06,
+      "loss": 2.3163,
+      "step": 1749
+    },
+    {
+      "epoch": 0.6750241080038573,
+      "grad_norm": 0.4421546566249675,
+      "learning_rate": 2.453945290114021e-06,
+      "loss": 2.3798,
+      "step": 1750
+    },
+    {
+      "epoch": 0.6754098360655738,
+      "grad_norm": 0.4452265180248534,
+      "learning_rate": 2.448649834695503e-06,
+      "loss": 2.3544,
+      "step": 1751
+    },
+    {
+      "epoch": 0.6757955641272902,
+      "grad_norm": 0.4306368748720817,
+      "learning_rate": 2.4433582456915556e-06,
+      "loss": 2.276,
+      "step": 1752
+    },
+    {
+      "epoch": 0.6761812921890068,
+      "grad_norm": 0.4123619858403055,
+      "learning_rate": 2.4380705311212557e-06,
+      "loss": 2.4109,
+      "step": 1753
+    },
+    {
+      "epoch": 0.6765670202507232,
+      "grad_norm": 0.39794428972608337,
+      "learning_rate": 2.432786698997813e-06,
+      "loss": 2.3193,
+      "step": 1754
+    },
+    {
+      "epoch": 0.6769527483124397,
+      "grad_norm": 0.41602795068016346,
+      "learning_rate": 2.427506757328549e-06,
+      "loss": 2.3405,
+      "step": 1755
+    },
+    {
+      "epoch": 0.6773384763741562,
+      "grad_norm": 0.44838899638141694,
+      "learning_rate": 2.422230714114891e-06,
+      "loss": 2.3606,
+      "step": 1756
+    },
+    {
+      "epoch": 0.6777242044358727,
+      "grad_norm": 0.4234780319399461,
+      "learning_rate": 2.416958577352361e-06,
+      "loss": 2.2062,
+      "step": 1757
+    },
+    {
+      "epoch": 0.6781099324975892,
+      "grad_norm": 0.43427324024035047,
+      "learning_rate": 2.411690355030556e-06,
+      "loss": 2.3539,
+      "step": 1758
+    },
+    {
+      "epoch": 0.6784956605593057,
+      "grad_norm": 0.4471318387959939,
+      "learning_rate": 2.4064260551331452e-06,
+      "loss": 2.292,
+      "step": 1759
+    },
+    {
+      "epoch": 0.6788813886210222,
+      "grad_norm": 0.4152680657270176,
+      "learning_rate": 2.4011656856378513e-06,
+      "loss": 2.3433,
+      "step": 1760
+    },
+    {
+      "epoch": 0.6792671166827386,
+      "grad_norm": 0.4221820618812638,
+      "learning_rate": 2.3959092545164408e-06,
+      "loss": 2.3259,
+      "step": 1761
+    },
+    {
+      "epoch": 0.6796528447444552,
+      "grad_norm": 0.4226470871183536,
+      "learning_rate": 2.3906567697347117e-06,
+      "loss": 2.3538,
+      "step": 1762
+    },
+    {
+      "epoch": 0.6800385728061716,
+      "grad_norm": 0.3946388601121077,
+      "learning_rate": 2.3854082392524836e-06,
+      "loss": 2.2842,
+      "step": 1763
+    },
+    {
+      "epoch": 0.6804243008678882,
+      "grad_norm": 0.4023774836632626,
+      "learning_rate": 2.3801636710235836e-06,
+      "loss": 2.2965,
+      "step": 1764
+    },
+    {
+      "epoch": 0.6808100289296046,
+      "grad_norm": 0.39928186380628516,
+      "learning_rate": 2.3749230729958322e-06,
+      "loss": 2.256,
+      "step": 1765
+    },
+    {
+      "epoch": 0.6811957569913211,
+      "grad_norm": 0.40275495361852215,
+      "learning_rate": 2.369686453111033e-06,
+      "loss": 2.3468,
+      "step": 1766
+    },
+    {
+      "epoch": 0.6815814850530376,
+      "grad_norm": 0.4120092812454993,
+      "learning_rate": 2.3644538193049626e-06,
+      "loss": 2.3906,
+      "step": 1767
+    },
+    {
+      "epoch": 0.6819672131147541,
+      "grad_norm": 0.42677358596189907,
+      "learning_rate": 2.3592251795073564e-06,
+      "loss": 2.2787,
+      "step": 1768
+    },
+    {
+      "epoch": 0.6823529411764706,
+      "grad_norm": 0.4296128371192793,
+      "learning_rate": 2.3540005416418943e-06,
+      "loss": 2.3418,
+      "step": 1769
+    },
+    {
+      "epoch": 0.682738669238187,
+      "grad_norm": 0.45496302018867923,
+      "learning_rate": 2.348779913626199e-06,
+      "loss": 2.3438,
+      "step": 1770
+    },
+    {
+      "epoch": 0.6831243972999036,
+      "grad_norm": 0.43775959209446863,
+      "learning_rate": 2.3435633033718096e-06,
+      "loss": 2.3501,
+      "step": 1771
+    },
+    {
+      "epoch": 0.68351012536162,
+      "grad_norm": 0.46247162736100794,
+      "learning_rate": 2.338350718784177e-06,
+      "loss": 2.3259,
+      "step": 1772
+    },
+    {
+      "epoch": 0.6838958534233366,
+      "grad_norm": 0.45384808041325736,
+      "learning_rate": 2.333142167762657e-06,
+      "loss": 2.3733,
+      "step": 1773
+    },
+    {
+      "epoch": 0.684281581485053,
+      "grad_norm": 0.4430161348080049,
+      "learning_rate": 2.327937658200487e-06,
+      "loss": 2.2967,
+      "step": 1774
+    },
+    {
+      "epoch": 0.6846673095467696,
+      "grad_norm": 0.42186373207951944,
+      "learning_rate": 2.322737197984781e-06,
+      "loss": 2.3502,
+      "step": 1775
+    },
+    {
+      "epoch": 0.685053037608486,
+      "grad_norm": 0.4463067104849964,
+      "learning_rate": 2.3175407949965167e-06,
+      "loss": 2.2719,
+      "step": 1776
+    },
+    {
+      "epoch": 0.6854387656702025,
+      "grad_norm": 0.4328966014517113,
+      "learning_rate": 2.312348457110527e-06,
+      "loss": 2.2898,
+      "step": 1777
+    },
+    {
+      "epoch": 0.685824493731919,
+      "grad_norm": 0.4125922387553286,
+      "learning_rate": 2.3071601921954797e-06,
+      "loss": 2.3335,
+      "step": 1778
+    },
+    {
+      "epoch": 0.6862102217936354,
+      "grad_norm": 0.4264072177678742,
+      "learning_rate": 2.301976008113871e-06,
+      "loss": 2.3068,
+      "step": 1779
+    },
+    {
+      "epoch": 0.686595949855352,
+      "grad_norm": 0.4216162007809989,
+      "learning_rate": 2.296795912722014e-06,
+      "loss": 2.2289,
+      "step": 1780
+    },
+    {
+      "epoch": 0.6869816779170684,
+      "grad_norm": 0.4093559337331704,
+      "learning_rate": 2.291619913870024e-06,
+      "loss": 2.2638,
+      "step": 1781
+    },
+    {
+      "epoch": 0.687367405978785,
+      "grad_norm": 0.40958554700715705,
+      "learning_rate": 2.286448019401811e-06,
+      "loss": 2.3726,
+      "step": 1782
+    },
+    {
+      "epoch": 0.6877531340405014,
+      "grad_norm": 0.4060960866832944,
+      "learning_rate": 2.2812802371550653e-06,
+      "loss": 2.3531,
+      "step": 1783
+    },
+    {
+      "epoch": 0.688138862102218,
+      "grad_norm": 0.45165307711606734,
+      "learning_rate": 2.2761165749612417e-06,
+      "loss": 2.314,
+      "step": 1784
+    },
+    {
+      "epoch": 0.6885245901639344,
+      "grad_norm": 0.4020078234302099,
+      "learning_rate": 2.2709570406455543e-06,
+      "loss": 2.3234,
+      "step": 1785
+    },
+    {
+      "epoch": 0.688910318225651,
+      "grad_norm": 0.43220274620477667,
+      "learning_rate": 2.2658016420269596e-06,
+      "loss": 2.3538,
+      "step": 1786
+    },
+    {
+      "epoch": 0.6892960462873674,
+      "grad_norm": 0.4400411393426206,
+      "learning_rate": 2.2606503869181486e-06,
+      "loss": 2.2693,
+      "step": 1787
+    },
+    {
+      "epoch": 0.6896817743490838,
+      "grad_norm": 0.5031429505901274,
+      "learning_rate": 2.25550328312553e-06,
+      "loss": 2.3507,
+      "step": 1788
+    },
+    {
+      "epoch": 0.6900675024108004,
+      "grad_norm": 0.4292461378623028,
+      "learning_rate": 2.250360338449226e-06,
+      "loss": 2.3647,
+      "step": 1789
+    },
+    {
+      "epoch": 0.6904532304725168,
+      "grad_norm": 0.4382398684826801,
+      "learning_rate": 2.2452215606830524e-06,
+      "loss": 2.3459,
+      "step": 1790
+    },
+    {
+      "epoch": 0.6908389585342334,
+      "grad_norm": 0.417063718826344,
+      "learning_rate": 2.2400869576145135e-06,
+      "loss": 2.297,
+      "step": 1791
+    },
+    {
+      "epoch": 0.6912246865959498,
+      "grad_norm": 0.4262177150266158,
+      "learning_rate": 2.2349565370247837e-06,
+      "loss": 2.3026,
+      "step": 1792
+    },
+    {
+      "epoch": 0.6916104146576664,
+      "grad_norm": 0.4198489437872795,
+      "learning_rate": 2.2298303066887007e-06,
+      "loss": 2.3022,
+      "step": 1793
+    },
+    {
+      "epoch": 0.6919961427193828,
+      "grad_norm": 0.4382512325320904,
+      "learning_rate": 2.2247082743747517e-06,
+      "loss": 2.3303,
+      "step": 1794
+    },
+    {
+      "epoch": 0.6923818707810994,
+      "grad_norm": 0.3966081503331144,
+      "learning_rate": 2.2195904478450603e-06,
+      "loss": 2.2817,
+      "step": 1795
+    },
+    {
+      "epoch": 0.6927675988428158,
+      "grad_norm": 0.4501733901208891,
+      "learning_rate": 2.214476834855382e-06,
+      "loss": 2.3482,
+      "step": 1796
+    },
+    {
+      "epoch": 0.6931533269045324,
+      "grad_norm": 0.4034321531880344,
+      "learning_rate": 2.209367443155082e-06,
+      "loss": 2.3428,
+      "step": 1797
+    },
+    {
+      "epoch": 0.6935390549662488,
+      "grad_norm": 0.42956892349144943,
+      "learning_rate": 2.20426228048713e-06,
+      "loss": 2.3275,
+      "step": 1798
+    },
+    {
+      "epoch": 0.6939247830279652,
+      "grad_norm": 0.4554530017723556,
+      "learning_rate": 2.199161354588086e-06,
+      "loss": 2.3162,
+      "step": 1799
+    },
+    {
+      "epoch": 0.6943105110896818,
+      "grad_norm": 0.41340986165183063,
+      "learning_rate": 2.1940646731880887e-06,
+      "loss": 2.2701,
+      "step": 1800
+    },
+    {
+      "epoch": 0.6946962391513982,
+      "grad_norm": 0.47516419252309366,
+      "learning_rate": 2.188972244010849e-06,
+      "loss": 2.3452,
+      "step": 1801
+    },
+    {
+      "epoch": 0.6950819672131148,
+      "grad_norm": 0.4484595753194459,
+      "learning_rate": 2.183884074773628e-06,
+      "loss": 2.2771,
+      "step": 1802
+    },
+    {
+      "epoch": 0.6954676952748312,
+      "grad_norm": 0.4045747362266517,
+      "learning_rate": 2.178800173187237e-06,
+      "loss": 2.3176,
+      "step": 1803
+    },
+    {
+      "epoch": 0.6958534233365478,
+      "grad_norm": 0.42374833281973634,
+      "learning_rate": 2.173720546956015e-06,
+      "loss": 2.3732,
+      "step": 1804
+    },
+    {
+      "epoch": 0.6962391513982642,
+      "grad_norm": 0.43674250725646085,
+      "learning_rate": 2.1686452037778236e-06,
+      "loss": 2.3584,
+      "step": 1805
+    },
+    {
+      "epoch": 0.6966248794599808,
+      "grad_norm": 0.4525712743157628,
+      "learning_rate": 2.1635741513440346e-06,
+      "loss": 2.4265,
+      "step": 1806
+    },
+    {
+      "epoch": 0.6970106075216972,
+      "grad_norm": 0.4176078402911036,
+      "learning_rate": 2.1585073973395156e-06,
+      "loss": 2.2231,
+      "step": 1807
+    },
+    {
+      "epoch": 0.6973963355834137,
+      "grad_norm": 0.42853072370017875,
+      "learning_rate": 2.1534449494426203e-06,
+      "loss": 2.302,
+      "step": 1808
+    },
+    {
+      "epoch": 0.6977820636451302,
+      "grad_norm": 0.41296489653639573,
+      "learning_rate": 2.148386815325179e-06,
+      "loss": 2.3283,
+      "step": 1809
+    },
+    {
+      "epoch": 0.6981677917068466,
+      "grad_norm": 0.4514682175260553,
+      "learning_rate": 2.1433330026524855e-06,
+      "loss": 2.3262,
+      "step": 1810
+    },
+    {
+      "epoch": 0.6985535197685632,
+      "grad_norm": 0.45052354679809675,
+      "learning_rate": 2.138283519083281e-06,
+      "loss": 2.3511,
+      "step": 1811
+    },
+    {
+      "epoch": 0.6989392478302796,
+      "grad_norm": 0.4503982963335606,
+      "learning_rate": 2.1332383722697483e-06,
+      "loss": 2.2479,
+      "step": 1812
+    },
+    {
+      "epoch": 0.6993249758919962,
+      "grad_norm": 0.40719473821595903,
+      "learning_rate": 2.128197569857497e-06,
+      "loss": 2.3032,
+      "step": 1813
+    },
+    {
+      "epoch": 0.6997107039537126,
+      "grad_norm": 0.41991110805498033,
+      "learning_rate": 2.1231611194855523e-06,
+      "loss": 2.3421,
+      "step": 1814
+    },
+    {
+      "epoch": 0.7000964320154291,
+      "grad_norm": 0.44901149374013327,
+      "learning_rate": 2.118129028786349e-06,
+      "loss": 2.3204,
+      "step": 1815
+    },
+    {
+      "epoch": 0.7004821600771456,
+      "grad_norm": 0.41440766728142986,
+      "learning_rate": 2.1131013053857097e-06,
+      "loss": 2.3755,
+      "step": 1816
+    },
+    {
+      "epoch": 0.7008678881388621,
+      "grad_norm": 0.42158767575961104,
+      "learning_rate": 2.1080779569028413e-06,
+      "loss": 2.3608,
+      "step": 1817
+    },
+    {
+      "epoch": 0.7012536162005786,
+      "grad_norm": 0.4544717459877259,
+      "learning_rate": 2.103058990950318e-06,
+      "loss": 2.3146,
+      "step": 1818
+    },
+    {
+      "epoch": 0.7016393442622951,
+      "grad_norm": 0.424649340632615,
+      "learning_rate": 2.098044415134078e-06,
+      "loss": 2.33,
+      "step": 1819
+    },
+    {
+      "epoch": 0.7020250723240116,
+      "grad_norm": 0.40885864827974017,
+      "learning_rate": 2.0930342370534013e-06,
+      "loss": 2.3,
+      "step": 1820
+    },
+    {
+      "epoch": 0.702410800385728,
+      "grad_norm": 0.4552603406642715,
+      "learning_rate": 2.0880284643009035e-06,
+      "loss": 2.3369,
+      "step": 1821
+    },
+    {
+      "epoch": 0.7027965284474446,
+      "grad_norm": 0.4513056650519724,
+      "learning_rate": 2.08302710446253e-06,
+      "loss": 2.3225,
+      "step": 1822
+    },
+    {
+      "epoch": 0.703182256509161,
+      "grad_norm": 0.4244786335106065,
+      "learning_rate": 2.078030165117533e-06,
+      "loss": 2.3048,
+      "step": 1823
+    },
+    {
+      "epoch": 0.7035679845708775,
+      "grad_norm": 0.43298567730858045,
+      "learning_rate": 2.073037653838466e-06,
+      "loss": 2.2702,
+      "step": 1824
+    },
+    {
+      "epoch": 0.703953712632594,
+      "grad_norm": 0.45316993438511183,
+      "learning_rate": 2.0680495781911745e-06,
+      "loss": 2.2087,
+      "step": 1825
+    },
+    {
+      "epoch": 0.7043394406943105,
+      "grad_norm": 0.44461260348149667,
+      "learning_rate": 2.0630659457347806e-06,
+      "loss": 2.3887,
+      "step": 1826
+    },
+    {
+      "epoch": 0.704725168756027,
+      "grad_norm": 0.4384648003703453,
+      "learning_rate": 2.0580867640216723e-06,
+      "loss": 2.2269,
+      "step": 1827
+    },
+    {
+      "epoch": 0.7051108968177435,
+      "grad_norm": 0.44502413743025215,
+      "learning_rate": 2.053112040597495e-06,
+      "loss": 2.3219,
+      "step": 1828
+    },
+    {
+      "epoch": 0.70549662487946,
+      "grad_norm": 0.43791680419882656,
+      "learning_rate": 2.048141783001138e-06,
+      "loss": 2.3794,
+      "step": 1829
+    },
+    {
+      "epoch": 0.7058823529411765,
+      "grad_norm": 0.4527043653356696,
+      "learning_rate": 2.0431759987647206e-06,
+      "loss": 2.3024,
+      "step": 1830
+    },
+    {
+      "epoch": 0.706268081002893,
+      "grad_norm": 0.43486023816632846,
+      "learning_rate": 2.038214695413584e-06,
+      "loss": 2.3632,
+      "step": 1831
+    },
+    {
+      "epoch": 0.7066538090646094,
+      "grad_norm": 0.4340772873446721,
+      "learning_rate": 2.0332578804662783e-06,
+      "loss": 2.3251,
+      "step": 1832
+    },
+    {
+      "epoch": 0.707039537126326,
+      "grad_norm": 0.42095115265847993,
+      "learning_rate": 2.028305561434553e-06,
+      "loss": 2.3547,
+      "step": 1833
+    },
+    {
+      "epoch": 0.7074252651880424,
+      "grad_norm": 0.4153578236352047,
+      "learning_rate": 2.0233577458233418e-06,
+      "loss": 2.3595,
+      "step": 1834
+    },
+    {
+      "epoch": 0.7078109932497589,
+      "grad_norm": 0.45286628297659587,
+      "learning_rate": 2.0184144411307584e-06,
+      "loss": 2.2989,
+      "step": 1835
+    },
+    {
+      "epoch": 0.7081967213114754,
+      "grad_norm": 0.415210757441709,
+      "learning_rate": 2.013475654848076e-06,
+      "loss": 2.2673,
+      "step": 1836
+    },
+    {
+      "epoch": 0.7085824493731919,
+      "grad_norm": 0.43402706304210137,
+      "learning_rate": 2.008541394459721e-06,
+      "loss": 2.3889,
+      "step": 1837
+    },
+    {
+      "epoch": 0.7089681774349084,
+      "grad_norm": 0.4501553541223613,
+      "learning_rate": 2.0036116674432653e-06,
+      "loss": 2.3073,
+      "step": 1838
+    },
+    {
+      "epoch": 0.7093539054966249,
+      "grad_norm": 0.4137364602460989,
+      "learning_rate": 1.998686481269406e-06,
+      "loss": 2.3535,
+      "step": 1839
+    },
+    {
+      "epoch": 0.7097396335583414,
+      "grad_norm": 0.43657698866139233,
+      "learning_rate": 1.99376584340196e-06,
+      "loss": 2.3303,
+      "step": 1840
+    },
+    {
+      "epoch": 0.7101253616200579,
+      "grad_norm": 0.42596895550642816,
+      "learning_rate": 1.98884976129785e-06,
+      "loss": 2.4018,
+      "step": 1841
+    },
+    {
+      "epoch": 0.7105110896817743,
+      "grad_norm": 0.4200682470564326,
+      "learning_rate": 1.983938242407101e-06,
+      "loss": 2.3129,
+      "step": 1842
+    },
+    {
+      "epoch": 0.7108968177434909,
+      "grad_norm": 0.4324219822945481,
+      "learning_rate": 1.979031294172817e-06,
+      "loss": 2.2912,
+      "step": 1843
+    },
+    {
+      "epoch": 0.7112825458052073,
+      "grad_norm": 0.41765298347102997,
+      "learning_rate": 1.9741289240311757e-06,
+      "loss": 2.3259,
+      "step": 1844
+    },
+    {
+      "epoch": 0.7116682738669238,
+      "grad_norm": 0.42394530159768007,
+      "learning_rate": 1.9692311394114176e-06,
+      "loss": 2.3176,
+      "step": 1845
+    },
+    {
+      "epoch": 0.7120540019286403,
+      "grad_norm": 0.43116865679038874,
+      "learning_rate": 1.964337947735835e-06,
+      "loss": 2.337,
+      "step": 1846
+    },
+    {
+      "epoch": 0.7124397299903568,
+      "grad_norm": 0.4502038632023192,
+      "learning_rate": 1.9594493564197613e-06,
+      "loss": 2.2922,
+      "step": 1847
+    },
+    {
+      "epoch": 0.7128254580520733,
+      "grad_norm": 0.4447897770422271,
+      "learning_rate": 1.954565372871554e-06,
+      "loss": 2.3829,
+      "step": 1848
+    },
+    {
+      "epoch": 0.7132111861137898,
+      "grad_norm": 0.43691135598900194,
+      "learning_rate": 1.9496860044925935e-06,
+      "loss": 2.3216,
+      "step": 1849
+    },
+    {
+      "epoch": 0.7135969141755063,
+      "grad_norm": 0.4701764832621497,
+      "learning_rate": 1.9448112586772617e-06,
+      "loss": 2.3048,
+      "step": 1850
+    },
+    {
+      "epoch": 0.7139826422372227,
+      "grad_norm": 0.43593029585434484,
+      "learning_rate": 1.9399411428129354e-06,
+      "loss": 2.3111,
+      "step": 1851
+    },
+    {
+      "epoch": 0.7143683702989393,
+      "grad_norm": 0.4428201780077991,
+      "learning_rate": 1.935075664279978e-06,
+      "loss": 2.3719,
+      "step": 1852
+    },
+    {
+      "epoch": 0.7147540983606557,
+      "grad_norm": 0.43311839685942927,
+      "learning_rate": 1.930214830451721e-06,
+      "loss": 2.2691,
+      "step": 1853
+    },
+    {
+      "epoch": 0.7151398264223723,
+      "grad_norm": 0.42447069689000166,
+      "learning_rate": 1.925358648694463e-06,
+      "loss": 2.2861,
+      "step": 1854
+    },
+    {
+      "epoch": 0.7155255544840887,
+      "grad_norm": 0.40031901393760644,
+      "learning_rate": 1.920507126367448e-06,
+      "loss": 2.3643,
+      "step": 1855
+    },
+    {
+      "epoch": 0.7159112825458052,
+      "grad_norm": 0.3900072469818421,
+      "learning_rate": 1.9156602708228584e-06,
+      "loss": 2.3331,
+      "step": 1856
+    },
+    {
+      "epoch": 0.7162970106075217,
+      "grad_norm": 0.48597677530097333,
+      "learning_rate": 1.910818089405809e-06,
+      "loss": 2.3775,
+      "step": 1857
+    },
+    {
+      "epoch": 0.7166827386692382,
+      "grad_norm": 0.46915085296113956,
+      "learning_rate": 1.9059805894543288e-06,
+      "loss": 2.333,
+      "step": 1858
+    },
+    {
+      "epoch": 0.7170684667309547,
+      "grad_norm": 0.45610828678227044,
+      "learning_rate": 1.9011477782993503e-06,
+      "loss": 2.3094,
+      "step": 1859
+    },
+    {
+      "epoch": 0.7174541947926711,
+      "grad_norm": 0.4468466992999948,
+      "learning_rate": 1.8963196632647008e-06,
+      "loss": 2.3186,
+      "step": 1860
+    },
+    {
+      "epoch": 0.7178399228543877,
+      "grad_norm": 0.4498292884554748,
+      "learning_rate": 1.891496251667096e-06,
+      "loss": 2.3005,
+      "step": 1861
+    },
+    {
+      "epoch": 0.7182256509161041,
+      "grad_norm": 0.4089911948744338,
+      "learning_rate": 1.886677550816118e-06,
+      "loss": 2.2811,
+      "step": 1862
+    },
+    {
+      "epoch": 0.7186113789778207,
+      "grad_norm": 0.4000617709919224,
+      "learning_rate": 1.8818635680142127e-06,
+      "loss": 2.3181,
+      "step": 1863
+    },
+    {
+      "epoch": 0.7189971070395371,
+      "grad_norm": 0.4580902252341343,
+      "learning_rate": 1.8770543105566752e-06,
+      "loss": 2.2947,
+      "step": 1864
+    },
+    {
+      "epoch": 0.7193828351012537,
+      "grad_norm": 0.42253237694802087,
+      "learning_rate": 1.872249785731638e-06,
+      "loss": 2.3342,
+      "step": 1865
+    },
+    {
+      "epoch": 0.7197685631629701,
+      "grad_norm": 0.4496865269170129,
+      "learning_rate": 1.8674500008200675e-06,
+      "loss": 2.3673,
+      "step": 1866
+    },
+    {
+      "epoch": 0.7201542912246865,
+      "grad_norm": 0.4073635146420168,
+      "learning_rate": 1.8626549630957397e-06,
+      "loss": 2.3325,
+      "step": 1867
+    },
+    {
+      "epoch": 0.7205400192864031,
+      "grad_norm": 0.3997870993505134,
+      "learning_rate": 1.8578646798252432e-06,
+      "loss": 2.2871,
+      "step": 1868
+    },
+    {
+      "epoch": 0.7209257473481195,
+      "grad_norm": 0.40751796677737057,
+      "learning_rate": 1.8530791582679558e-06,
+      "loss": 2.3511,
+      "step": 1869
+    },
+    {
+      "epoch": 0.7213114754098361,
+      "grad_norm": 0.44983615888224715,
+      "learning_rate": 1.8482984056760434e-06,
+      "loss": 2.3554,
+      "step": 1870
+    },
+    {
+      "epoch": 0.7216972034715525,
+      "grad_norm": 0.4192126012438455,
+      "learning_rate": 1.8435224292944415e-06,
+      "loss": 2.3417,
+      "step": 1871
+    },
+    {
+      "epoch": 0.7220829315332691,
+      "grad_norm": 0.4506643100816781,
+      "learning_rate": 1.8387512363608496e-06,
+      "loss": 2.3664,
+      "step": 1872
+    },
+    {
+      "epoch": 0.7224686595949855,
+      "grad_norm": 0.4165920944006413,
+      "learning_rate": 1.8339848341057165e-06,
+      "loss": 2.3338,
+      "step": 1873
+    },
+    {
+      "epoch": 0.7228543876567021,
+      "grad_norm": 0.4315717005524774,
+      "learning_rate": 1.8292232297522356e-06,
+      "loss": 2.3309,
+      "step": 1874
+    },
+    {
+      "epoch": 0.7232401157184185,
+      "grad_norm": 0.4104338792616358,
+      "learning_rate": 1.8244664305163217e-06,
+      "loss": 2.3175,
+      "step": 1875
+    },
+    {
+      "epoch": 0.7236258437801351,
+      "grad_norm": 0.4250811881361986,
+      "learning_rate": 1.8197144436066167e-06,
+      "loss": 2.3429,
+      "step": 1876
+    },
+    {
+      "epoch": 0.7240115718418515,
+      "grad_norm": 0.4319063440906805,
+      "learning_rate": 1.8149672762244625e-06,
+      "loss": 2.3144,
+      "step": 1877
+    },
+    {
+      "epoch": 0.7243972999035679,
+      "grad_norm": 0.4631219421778302,
+      "learning_rate": 1.8102249355639007e-06,
+      "loss": 2.3249,
+      "step": 1878
+    },
+    {
+      "epoch": 0.7247830279652845,
+      "grad_norm": 0.4096130626344659,
+      "learning_rate": 1.8054874288116564e-06,
+      "loss": 2.3437,
+      "step": 1879
+    },
+    {
+      "epoch": 0.7251687560270009,
+      "grad_norm": 0.43221946726888827,
+      "learning_rate": 1.8007547631471289e-06,
+      "loss": 2.3126,
+      "step": 1880
+    },
+    {
+      "epoch": 0.7255544840887175,
+      "grad_norm": 0.41203528881787477,
+      "learning_rate": 1.7960269457423867e-06,
+      "loss": 2.3676,
+      "step": 1881
+    },
+    {
+      "epoch": 0.7259402121504339,
+      "grad_norm": 0.456724826662013,
+      "learning_rate": 1.7913039837621448e-06,
+      "loss": 2.3162,
+      "step": 1882
+    },
+    {
+      "epoch": 0.7263259402121505,
+      "grad_norm": 0.4622864968839015,
+      "learning_rate": 1.7865858843637617e-06,
+      "loss": 2.3036,
+      "step": 1883
+    },
+    {
+      "epoch": 0.7267116682738669,
+      "grad_norm": 0.3977901985670626,
+      "learning_rate": 1.781872654697226e-06,
+      "loss": 2.3802,
+      "step": 1884
+    },
+    {
+      "epoch": 0.7270973963355835,
+      "grad_norm": 0.44693389963733726,
+      "learning_rate": 1.7771643019051516e-06,
+      "loss": 2.3529,
+      "step": 1885
+    },
+    {
+      "epoch": 0.7274831243972999,
+      "grad_norm": 0.42913602906379306,
+      "learning_rate": 1.772460833122755e-06,
+      "loss": 2.3534,
+      "step": 1886
+    },
+    {
+      "epoch": 0.7278688524590164,
+      "grad_norm": 0.42282407024026286,
+      "learning_rate": 1.7677622554778568e-06,
+      "loss": 2.3453,
+      "step": 1887
+    },
+    {
+      "epoch": 0.7282545805207329,
+      "grad_norm": 0.4246761734320408,
+      "learning_rate": 1.7630685760908623e-06,
+      "loss": 2.2622,
+      "step": 1888
+    },
+    {
+      "epoch": 0.7286403085824493,
+      "grad_norm": 0.4085341091618752,
+      "learning_rate": 1.7583798020747538e-06,
+      "loss": 2.3744,
+      "step": 1889
+    },
+    {
+      "epoch": 0.7290260366441659,
+      "grad_norm": 0.3963714532359338,
+      "learning_rate": 1.7536959405350806e-06,
+      "loss": 2.2303,
+      "step": 1890
+    },
+    {
+      "epoch": 0.7294117647058823,
+      "grad_norm": 0.4256967560310133,
+      "learning_rate": 1.7490169985699485e-06,
+      "loss": 2.2885,
+      "step": 1891
+    },
+    {
+      "epoch": 0.7297974927675989,
+      "grad_norm": 0.4493754498912692,
+      "learning_rate": 1.7443429832700038e-06,
+      "loss": 2.3275,
+      "step": 1892
+    },
+    {
+      "epoch": 0.7301832208293153,
+      "grad_norm": 0.4290343544231461,
+      "learning_rate": 1.7396739017184334e-06,
+      "loss": 2.3917,
+      "step": 1893
+    },
+    {
+      "epoch": 0.7305689488910319,
+      "grad_norm": 0.4209821488017083,
+      "learning_rate": 1.735009760990941e-06,
+      "loss": 2.3115,
+      "step": 1894
+    },
+    {
+      "epoch": 0.7309546769527483,
+      "grad_norm": 0.44112705296346094,
+      "learning_rate": 1.7303505681557486e-06,
+      "loss": 2.2876,
+      "step": 1895
+    },
+    {
+      "epoch": 0.7313404050144648,
+      "grad_norm": 0.43928075906138714,
+      "learning_rate": 1.7256963302735752e-06,
+      "loss": 2.3287,
+      "step": 1896
+    },
+    {
+      "epoch": 0.7317261330761813,
+      "grad_norm": 0.4355951809171198,
+      "learning_rate": 1.7210470543976326e-06,
+      "loss": 2.3864,
+      "step": 1897
+    },
+    {
+      "epoch": 0.7321118611378978,
+      "grad_norm": 0.4313611024308735,
+      "learning_rate": 1.7164027475736134e-06,
+      "loss": 2.3034,
+      "step": 1898
+    },
+    {
+      "epoch": 0.7324975891996143,
+      "grad_norm": 0.4353284629392325,
+      "learning_rate": 1.7117634168396774e-06,
+      "loss": 2.2431,
+      "step": 1899
+    },
+    {
+      "epoch": 0.7328833172613307,
+      "grad_norm": 0.4382032196577667,
+      "learning_rate": 1.7071290692264492e-06,
+      "loss": 2.3491,
+      "step": 1900
+    },
+    {
+      "epoch": 0.7332690453230473,
+      "grad_norm": 0.43415950786337915,
+      "learning_rate": 1.7024997117569964e-06,
+      "loss": 2.3355,
+      "step": 1901
+    },
+    {
+      "epoch": 0.7336547733847637,
+      "grad_norm": 0.4314052449468606,
+      "learning_rate": 1.6978753514468255e-06,
+      "loss": 2.3296,
+      "step": 1902
+    },
+    {
+      "epoch": 0.7340405014464803,
+      "grad_norm": 0.44774770885005954,
+      "learning_rate": 1.6932559953038702e-06,
+      "loss": 2.3172,
+      "step": 1903
+    },
+    {
+      "epoch": 0.7344262295081967,
+      "grad_norm": 0.4312655490717796,
+      "learning_rate": 1.6886416503284835e-06,
+      "loss": 2.3042,
+      "step": 1904
+    },
+    {
+      "epoch": 0.7348119575699132,
+      "grad_norm": 0.43758573603925244,
+      "learning_rate": 1.684032323513421e-06,
+      "loss": 2.3183,
+      "step": 1905
+    },
+    {
+      "epoch": 0.7351976856316297,
+      "grad_norm": 0.48022398401192334,
+      "learning_rate": 1.679428021843833e-06,
+      "loss": 2.2606,
+      "step": 1906
+    },
+    {
+      "epoch": 0.7355834136933462,
+      "grad_norm": 0.43618605575066427,
+      "learning_rate": 1.6748287522972583e-06,
+      "loss": 2.2778,
+      "step": 1907
+    },
+    {
+      "epoch": 0.7359691417550627,
+      "grad_norm": 0.4445544662179184,
+      "learning_rate": 1.6702345218436066e-06,
+      "loss": 2.3342,
+      "step": 1908
+    },
+    {
+      "epoch": 0.7363548698167792,
+      "grad_norm": 0.4383728743068401,
+      "learning_rate": 1.665645337445153e-06,
+      "loss": 2.3321,
+      "step": 1909
+    },
+    {
+      "epoch": 0.7367405978784957,
+      "grad_norm": 0.42465807370261877,
+      "learning_rate": 1.6610612060565235e-06,
+      "loss": 2.3491,
+      "step": 1910
+    },
+    {
+      "epoch": 0.7371263259402121,
+      "grad_norm": 0.42171228959818263,
+      "learning_rate": 1.6564821346246878e-06,
+      "loss": 2.2639,
+      "step": 1911
+    },
+    {
+      "epoch": 0.7375120540019287,
+      "grad_norm": 0.4240188509741186,
+      "learning_rate": 1.6519081300889472e-06,
+      "loss": 2.3131,
+      "step": 1912
+    },
+    {
+      "epoch": 0.7378977820636451,
+      "grad_norm": 0.47621727140695824,
+      "learning_rate": 1.6473391993809252e-06,
+      "loss": 2.3753,
+      "step": 1913
+    },
+    {
+      "epoch": 0.7382835101253616,
+      "grad_norm": 0.40631055589112214,
+      "learning_rate": 1.6427753494245585e-06,
+      "loss": 2.3181,
+      "step": 1914
+    },
+    {
+      "epoch": 0.7386692381870781,
+      "grad_norm": 0.42382359906353934,
+      "learning_rate": 1.638216587136079e-06,
+      "loss": 2.3913,
+      "step": 1915
+    },
+    {
+      "epoch": 0.7390549662487946,
+      "grad_norm": 0.41456911250815726,
+      "learning_rate": 1.6336629194240118e-06,
+      "loss": 2.3084,
+      "step": 1916
+    },
+    {
+      "epoch": 0.7394406943105111,
+      "grad_norm": 0.42685946694878363,
+      "learning_rate": 1.6291143531891601e-06,
+      "loss": 2.3688,
+      "step": 1917
+    },
+    {
+      "epoch": 0.7398264223722276,
+      "grad_norm": 0.41571596018624085,
+      "learning_rate": 1.6245708953245958e-06,
+      "loss": 2.3355,
+      "step": 1918
+    },
+    {
+      "epoch": 0.7402121504339441,
+      "grad_norm": 0.4126446757420263,
+      "learning_rate": 1.6200325527156536e-06,
+      "loss": 2.3458,
+      "step": 1919
+    },
+    {
+      "epoch": 0.7405978784956606,
+      "grad_norm": 0.4203400452330613,
+      "learning_rate": 1.6154993322399114e-06,
+      "loss": 2.3521,
+      "step": 1920
+    },
+    {
+      "epoch": 0.740983606557377,
+      "grad_norm": 0.4283850137050537,
+      "learning_rate": 1.6109712407671867e-06,
+      "loss": 2.3439,
+      "step": 1921
+    },
+    {
+      "epoch": 0.7413693346190935,
+      "grad_norm": 0.4319712256454236,
+      "learning_rate": 1.6064482851595225e-06,
+      "loss": 2.3353,
+      "step": 1922
+    },
+    {
+      "epoch": 0.74175506268081,
+      "grad_norm": 0.43119097151201774,
+      "learning_rate": 1.6019304722711836e-06,
+      "loss": 2.2801,
+      "step": 1923
+    },
+    {
+      "epoch": 0.7421407907425265,
+      "grad_norm": 0.45487209978859533,
+      "learning_rate": 1.5974178089486364e-06,
+      "loss": 2.332,
+      "step": 1924
+    },
+    {
+      "epoch": 0.742526518804243,
+      "grad_norm": 0.44591444583471357,
+      "learning_rate": 1.5929103020305441e-06,
+      "loss": 2.2509,
+      "step": 1925
+    },
+    {
+      "epoch": 0.7429122468659595,
+      "grad_norm": 0.4381066815679619,
+      "learning_rate": 1.588407958347759e-06,
+      "loss": 2.235,
+      "step": 1926
+    },
+    {
+      "epoch": 0.743297974927676,
+      "grad_norm": 0.4454021513173899,
+      "learning_rate": 1.583910784723306e-06,
+      "loss": 2.3541,
+      "step": 1927
+    },
+    {
+      "epoch": 0.7436837029893925,
+      "grad_norm": 0.47553223454478355,
+      "learning_rate": 1.5794187879723755e-06,
+      "loss": 2.3544,
+      "step": 1928
+    },
+    {
+      "epoch": 0.744069431051109,
+      "grad_norm": 0.41385399136871154,
+      "learning_rate": 1.5749319749023117e-06,
+      "loss": 2.3134,
+      "step": 1929
+    },
+    {
+      "epoch": 0.7444551591128254,
+      "grad_norm": 0.3985543677381964,
+      "learning_rate": 1.5704503523126057e-06,
+      "loss": 2.3468,
+      "step": 1930
+    },
+    {
+      "epoch": 0.744840887174542,
+      "grad_norm": 0.40023562887529873,
+      "learning_rate": 1.5659739269948798e-06,
+      "loss": 2.3317,
+      "step": 1931
+    },
+    {
+      "epoch": 0.7452266152362584,
+      "grad_norm": 0.43607247084987116,
+      "learning_rate": 1.561502705732883e-06,
+      "loss": 2.3172,
+      "step": 1932
+    },
+    {
+      "epoch": 0.7456123432979749,
+      "grad_norm": 0.4112898127350544,
+      "learning_rate": 1.557036695302478e-06,
+      "loss": 2.3366,
+      "step": 1933
+    },
+    {
+      "epoch": 0.7459980713596914,
+      "grad_norm": 0.4225869777300909,
+      "learning_rate": 1.552575902471628e-06,
+      "loss": 2.3598,
+      "step": 1934
+    },
+    {
+      "epoch": 0.7463837994214079,
+      "grad_norm": 0.3947844663663247,
+      "learning_rate": 1.5481203340003915e-06,
+      "loss": 2.3811,
+      "step": 1935
+    },
+    {
+      "epoch": 0.7467695274831244,
+      "grad_norm": 0.4323476933757536,
+      "learning_rate": 1.543669996640908e-06,
+      "loss": 2.355,
+      "step": 1936
+    },
+    {
+      "epoch": 0.7471552555448409,
+      "grad_norm": 0.4260625784537816,
+      "learning_rate": 1.5392248971373913e-06,
+      "loss": 2.4,
+      "step": 1937
+    },
+    {
+      "epoch": 0.7475409836065574,
+      "grad_norm": 0.44096850417415434,
+      "learning_rate": 1.534785042226115e-06,
+      "loss": 2.3021,
+      "step": 1938
+    },
+    {
+      "epoch": 0.7479267116682738,
+      "grad_norm": 0.44503728345754445,
+      "learning_rate": 1.5303504386354096e-06,
+      "loss": 2.3117,
+      "step": 1939
+    },
+    {
+      "epoch": 0.7483124397299904,
+      "grad_norm": 0.43620387877127675,
+      "learning_rate": 1.5259210930856423e-06,
+      "loss": 2.3713,
+      "step": 1940
+    },
+    {
+      "epoch": 0.7486981677917068,
+      "grad_norm": 0.4136199715831025,
+      "learning_rate": 1.5214970122892164e-06,
+      "loss": 2.1578,
+      "step": 1941
+    },
+    {
+      "epoch": 0.7490838958534234,
+      "grad_norm": 0.3970309676881285,
+      "learning_rate": 1.5170782029505543e-06,
+      "loss": 2.3259,
+      "step": 1942
+    },
+    {
+      "epoch": 0.7494696239151398,
+      "grad_norm": 0.4624348047529903,
+      "learning_rate": 1.5126646717660898e-06,
+      "loss": 2.2813,
+      "step": 1943
+    },
+    {
+      "epoch": 0.7498553519768563,
+      "grad_norm": 0.4169276477348904,
+      "learning_rate": 1.5082564254242583e-06,
+      "loss": 2.3011,
+      "step": 1944
+    },
+    {
+      "epoch": 0.7502410800385728,
+      "grad_norm": 0.4220108001492482,
+      "learning_rate": 1.5038534706054857e-06,
+      "loss": 2.3509,
+      "step": 1945
+    },
+    {
+      "epoch": 0.7506268081002893,
+      "grad_norm": 0.4311426982017881,
+      "learning_rate": 1.4994558139821818e-06,
+      "loss": 2.3794,
+      "step": 1946
+    },
+    {
+      "epoch": 0.7510125361620058,
+      "grad_norm": 0.4443090002313747,
+      "learning_rate": 1.495063462218725e-06,
+      "loss": 2.3786,
+      "step": 1947
+    },
+    {
+      "epoch": 0.7513982642237222,
+      "grad_norm": 0.42055047265269285,
+      "learning_rate": 1.4906764219714537e-06,
+      "loss": 2.3093,
+      "step": 1948
+    },
+    {
+      "epoch": 0.7517839922854388,
+      "grad_norm": 0.4271206972448656,
+      "learning_rate": 1.4862946998886591e-06,
+      "loss": 2.3313,
+      "step": 1949
+    },
+    {
+      "epoch": 0.7521697203471552,
+      "grad_norm": 0.4523623802186264,
+      "learning_rate": 1.4819183026105694e-06,
+      "loss": 2.3825,
+      "step": 1950
+    },
+    {
+      "epoch": 0.7525554484088718,
+      "grad_norm": 0.44065059445820953,
+      "learning_rate": 1.47754723676935e-06,
+      "loss": 2.3994,
+      "step": 1951
+    },
+    {
+      "epoch": 0.7529411764705882,
+      "grad_norm": 0.4418740367804237,
+      "learning_rate": 1.4731815089890795e-06,
+      "loss": 2.3543,
+      "step": 1952
+    },
+    {
+      "epoch": 0.7533269045323048,
+      "grad_norm": 0.4172566227457047,
+      "learning_rate": 1.4688211258857533e-06,
+      "loss": 2.3196,
+      "step": 1953
+    },
+    {
+      "epoch": 0.7537126325940212,
+      "grad_norm": 0.4550251205994799,
+      "learning_rate": 1.4644660940672628e-06,
+      "loss": 2.4147,
+      "step": 1954
+    },
+    {
+      "epoch": 0.7540983606557377,
+      "grad_norm": 0.4176140241551424,
+      "learning_rate": 1.4601164201333917e-06,
+      "loss": 2.2887,
+      "step": 1955
+    },
+    {
+      "epoch": 0.7544840887174542,
+      "grad_norm": 0.4414057227206095,
+      "learning_rate": 1.455772110675804e-06,
+      "loss": 2.342,
+      "step": 1956
+    },
+    {
+      "epoch": 0.7548698167791706,
+      "grad_norm": 0.41491981206533557,
+      "learning_rate": 1.4514331722780323e-06,
+      "loss": 2.3814,
+      "step": 1957
+    },
+    {
+      "epoch": 0.7552555448408872,
+      "grad_norm": 0.42835423357893915,
+      "learning_rate": 1.447099611515474e-06,
+      "loss": 2.3606,
+      "step": 1958
+    },
+    {
+      "epoch": 0.7556412729026036,
+      "grad_norm": 0.41793133773717467,
+      "learning_rate": 1.4427714349553718e-06,
+      "loss": 2.2872,
+      "step": 1959
+    },
+    {
+      "epoch": 0.7560270009643202,
+      "grad_norm": 0.4126520103044043,
+      "learning_rate": 1.438448649156815e-06,
+      "loss": 2.3769,
+      "step": 1960
+    },
+    {
+      "epoch": 0.7564127290260366,
+      "grad_norm": 0.42809154465845917,
+      "learning_rate": 1.434131260670718e-06,
+      "loss": 2.3048,
+      "step": 1961
+    },
+    {
+      "epoch": 0.7567984570877532,
+      "grad_norm": 0.4159494516222891,
+      "learning_rate": 1.4298192760398183e-06,
+      "loss": 2.3678,
+      "step": 1962
+    },
+    {
+      "epoch": 0.7571841851494696,
+      "grad_norm": 0.42294985526580336,
+      "learning_rate": 1.4255127017986642e-06,
+      "loss": 2.3726,
+      "step": 1963
+    },
+    {
+      "epoch": 0.7575699132111862,
+      "grad_norm": 0.4547618222932602,
+      "learning_rate": 1.4212115444736024e-06,
+      "loss": 2.3227,
+      "step": 1964
+    },
+    {
+      "epoch": 0.7579556412729026,
+      "grad_norm": 0.4539738123807865,
+      "learning_rate": 1.4169158105827768e-06,
+      "loss": 2.3494,
+      "step": 1965
+    },
+    {
+      "epoch": 0.758341369334619,
+      "grad_norm": 0.4447504726542575,
+      "learning_rate": 1.412625506636106e-06,
+      "loss": 2.3716,
+      "step": 1966
+    },
+    {
+      "epoch": 0.7587270973963356,
+      "grad_norm": 0.4441122610676579,
+      "learning_rate": 1.4083406391352827e-06,
+      "loss": 2.3429,
+      "step": 1967
+    },
+    {
+      "epoch": 0.759112825458052,
+      "grad_norm": 0.4361873669315586,
+      "learning_rate": 1.4040612145737608e-06,
+      "loss": 2.3435,
+      "step": 1968
+    },
+    {
+      "epoch": 0.7594985535197686,
+      "grad_norm": 0.4501748064507914,
+      "learning_rate": 1.399787239436744e-06,
+      "loss": 2.3487,
+      "step": 1969
+    },
+    {
+      "epoch": 0.759884281581485,
+      "grad_norm": 0.3961270353598128,
+      "learning_rate": 1.3955187202011817e-06,
+      "loss": 2.2895,
+      "step": 1970
+    },
+    {
+      "epoch": 0.7602700096432016,
+      "grad_norm": 0.40673706251633657,
+      "learning_rate": 1.3912556633357504e-06,
+      "loss": 2.2987,
+      "step": 1971
+    },
+    {
+      "epoch": 0.760655737704918,
+      "grad_norm": 0.42338880344973645,
+      "learning_rate": 1.3869980753008537e-06,
+      "loss": 2.3168,
+      "step": 1972
+    },
+    {
+      "epoch": 0.7610414657666346,
+      "grad_norm": 0.40140883912164094,
+      "learning_rate": 1.382745962548604e-06,
+      "loss": 2.3646,
+      "step": 1973
+    },
+    {
+      "epoch": 0.761427193828351,
+      "grad_norm": 0.41867937012538364,
+      "learning_rate": 1.3784993315228167e-06,
+      "loss": 2.3299,
+      "step": 1974
+    },
+    {
+      "epoch": 0.7618129218900676,
+      "grad_norm": 0.4303118053950361,
+      "learning_rate": 1.3742581886590006e-06,
+      "loss": 2.3334,
+      "step": 1975
+    },
+    {
+      "epoch": 0.762198649951784,
+      "grad_norm": 0.3981641920452056,
+      "learning_rate": 1.370022540384347e-06,
+      "loss": 2.3107,
+      "step": 1976
+    },
+    {
+      "epoch": 0.7625843780135004,
+      "grad_norm": 0.4612849602908465,
+      "learning_rate": 1.3657923931177204e-06,
+      "loss": 2.2303,
+      "step": 1977
+    },
+    {
+      "epoch": 0.762970106075217,
+      "grad_norm": 0.4180792084943692,
+      "learning_rate": 1.3615677532696498e-06,
+      "loss": 2.3706,
+      "step": 1978
+    },
+    {
+      "epoch": 0.7633558341369334,
+      "grad_norm": 0.38250956972257466,
+      "learning_rate": 1.3573486272423192e-06,
+      "loss": 2.2321,
+      "step": 1979
+    },
+    {
+      "epoch": 0.76374156219865,
+      "grad_norm": 0.4027643412011808,
+      "learning_rate": 1.353135021429554e-06,
+      "loss": 2.3544,
+      "step": 1980
+    },
+    {
+      "epoch": 0.7641272902603664,
+      "grad_norm": 0.44615568380452053,
+      "learning_rate": 1.348926942216815e-06,
+      "loss": 2.2998,
+      "step": 1981
+    },
+    {
+      "epoch": 0.764513018322083,
+      "grad_norm": 0.4386786384305421,
+      "learning_rate": 1.3447243959811885e-06,
+      "loss": 2.3685,
+      "step": 1982
+    },
+    {
+      "epoch": 0.7648987463837994,
+      "grad_norm": 0.40691005347999887,
+      "learning_rate": 1.340527389091374e-06,
+      "loss": 2.3685,
+      "step": 1983
+    },
+    {
+      "epoch": 0.765284474445516,
+      "grad_norm": 0.444251477937014,
+      "learning_rate": 1.3363359279076776e-06,
+      "loss": 2.3357,
+      "step": 1984
+    },
+    {
+      "epoch": 0.7656702025072324,
+      "grad_norm": 0.4260239073946116,
+      "learning_rate": 1.3321500187820042e-06,
+      "loss": 2.3069,
+      "step": 1985
+    },
+    {
+      "epoch": 0.7660559305689489,
+      "grad_norm": 0.40422071633890755,
+      "learning_rate": 1.3279696680578402e-06,
+      "loss": 2.3677,
+      "step": 1986
+    },
+    {
+      "epoch": 0.7664416586306654,
+      "grad_norm": 0.41280437028129924,
+      "learning_rate": 1.3237948820702495e-06,
+      "loss": 2.331,
+      "step": 1987
+    },
+    {
+      "epoch": 0.7668273866923818,
+      "grad_norm": 0.4082483834119175,
+      "learning_rate": 1.3196256671458663e-06,
+      "loss": 2.3281,
+      "step": 1988
+    },
+    {
+      "epoch": 0.7672131147540984,
+      "grad_norm": 0.41258302839639005,
+      "learning_rate": 1.3154620296028793e-06,
+      "loss": 2.2833,
+      "step": 1989
+    },
+    {
+      "epoch": 0.7675988428158148,
+      "grad_norm": 0.4344894821948446,
+      "learning_rate": 1.3113039757510253e-06,
+      "loss": 2.3197,
+      "step": 1990
+    },
+    {
+      "epoch": 0.7679845708775314,
+      "grad_norm": 0.42892943095676617,
+      "learning_rate": 1.307151511891578e-06,
+      "loss": 2.3273,
+      "step": 1991
+    },
+    {
+      "epoch": 0.7683702989392478,
+      "grad_norm": 0.44491385707895625,
+      "learning_rate": 1.3030046443173445e-06,
+      "loss": 2.3597,
+      "step": 1992
+    },
+    {
+      "epoch": 0.7687560270009643,
+      "grad_norm": 0.41228660786813054,
+      "learning_rate": 1.298863379312647e-06,
+      "loss": 2.3733,
+      "step": 1993
+    },
+    {
+      "epoch": 0.7691417550626808,
+      "grad_norm": 0.4383023013387142,
+      "learning_rate": 1.2947277231533178e-06,
+      "loss": 2.2713,
+      "step": 1994
+    },
+    {
+      "epoch": 0.7695274831243973,
+      "grad_norm": 0.4036438771963324,
+      "learning_rate": 1.2905976821066902e-06,
+      "loss": 2.295,
+      "step": 1995
+    },
+    {
+      "epoch": 0.7699132111861138,
+      "grad_norm": 0.45641979064711524,
+      "learning_rate": 1.2864732624315867e-06,
+      "loss": 2.3163,
+      "step": 1996
+    },
+    {
+      "epoch": 0.7702989392478303,
+      "grad_norm": 0.44286916639484614,
+      "learning_rate": 1.282354470378313e-06,
+      "loss": 2.3142,
+      "step": 1997
+    },
+    {
+      "epoch": 0.7706846673095468,
+      "grad_norm": 0.45764787143026897,
+      "learning_rate": 1.2782413121886483e-06,
+      "loss": 2.3925,
+      "step": 1998
+    },
+    {
+      "epoch": 0.7710703953712632,
+      "grad_norm": 0.41820084025876386,
+      "learning_rate": 1.2741337940958286e-06,
+      "loss": 2.3918,
+      "step": 1999
+    },
+    {
+      "epoch": 0.7714561234329798,
+      "grad_norm": 0.43730016656356097,
+      "learning_rate": 1.270031922324546e-06,
+      "loss": 2.2286,
+      "step": 2000
+    },
+    {
+      "epoch": 0.7718418514946962,
+      "grad_norm": 0.45294857793904747,
+      "learning_rate": 1.2659357030909352e-06,
+      "loss": 2.3372,
+      "step": 2001
+    },
+    {
+      "epoch": 0.7722275795564127,
+      "grad_norm": 0.44175153237351367,
+      "learning_rate": 1.2618451426025657e-06,
+      "loss": 2.3348,
+      "step": 2002
+    },
+    {
+      "epoch": 0.7726133076181292,
+      "grad_norm": 0.4199874401953492,
+      "learning_rate": 1.2577602470584287e-06,
+      "loss": 2.2427,
+      "step": 2003
+    },
+    {
+      "epoch": 0.7729990356798457,
+      "grad_norm": 0.42520445167328846,
+      "learning_rate": 1.2536810226489354e-06,
+      "loss": 2.3594,
+      "step": 2004
+    },
+    {
+      "epoch": 0.7733847637415622,
+      "grad_norm": 0.396497578004788,
+      "learning_rate": 1.249607475555899e-06,
+      "loss": 2.2683,
+      "step": 2005
+    },
+    {
+      "epoch": 0.7737704918032787,
+      "grad_norm": 0.4298494247154455,
+      "learning_rate": 1.2455396119525288e-06,
+      "loss": 2.2133,
+      "step": 2006
+    },
+    {
+      "epoch": 0.7741562198649952,
+      "grad_norm": 0.5036877150521499,
+      "learning_rate": 1.2414774380034245e-06,
+      "loss": 2.2628,
+      "step": 2007
+    },
+    {
+      "epoch": 0.7745419479267117,
+      "grad_norm": 0.4205072735039371,
+      "learning_rate": 1.237420959864561e-06,
+      "loss": 2.3491,
+      "step": 2008
+    },
+    {
+      "epoch": 0.7749276759884282,
+      "grad_norm": 0.43258037370084457,
+      "learning_rate": 1.2333701836832812e-06,
+      "loss": 2.3784,
+      "step": 2009
+    },
+    {
+      "epoch": 0.7753134040501446,
+      "grad_norm": 0.41445328672967857,
+      "learning_rate": 1.229325115598286e-06,
+      "loss": 2.2929,
+      "step": 2010
+    },
+    {
+      "epoch": 0.7756991321118611,
+      "grad_norm": 0.42352117823562174,
+      "learning_rate": 1.2252857617396318e-06,
+      "loss": 2.3134,
+      "step": 2011
+    },
+    {
+      "epoch": 0.7760848601735776,
+      "grad_norm": 0.4106272439078515,
+      "learning_rate": 1.2212521282287093e-06,
+      "loss": 2.3272,
+      "step": 2012
+    },
+    {
+      "epoch": 0.7764705882352941,
+      "grad_norm": 0.40093498606113714,
+      "learning_rate": 1.217224221178242e-06,
+      "loss": 2.2491,
+      "step": 2013
+    },
+    {
+      "epoch": 0.7768563162970106,
+      "grad_norm": 0.43709998605681294,
+      "learning_rate": 1.2132020466922767e-06,
+      "loss": 2.3315,
+      "step": 2014
+    },
+    {
+      "epoch": 0.7772420443587271,
+      "grad_norm": 0.4356135695882619,
+      "learning_rate": 1.2091856108661703e-06,
+      "loss": 2.3562,
+      "step": 2015
+    },
+    {
+      "epoch": 0.7776277724204436,
+      "grad_norm": 0.47240650389883526,
+      "learning_rate": 1.2051749197865875e-06,
+      "loss": 2.3961,
+      "step": 2016
+    },
+    {
+      "epoch": 0.7780135004821601,
+      "grad_norm": 0.4060173420056584,
+      "learning_rate": 1.2011699795314813e-06,
+      "loss": 2.3648,
+      "step": 2017
+    },
+    {
+      "epoch": 0.7783992285438766,
+      "grad_norm": 0.41026562263462857,
+      "learning_rate": 1.1971707961700962e-06,
+      "loss": 2.2899,
+      "step": 2018
+    },
+    {
+      "epoch": 0.7787849566055931,
+      "grad_norm": 0.42500261896680547,
+      "learning_rate": 1.1931773757629472e-06,
+      "loss": 2.365,
+      "step": 2019
+    },
+    {
+      "epoch": 0.7791706846673095,
+      "grad_norm": 0.4110188581895219,
+      "learning_rate": 1.1891897243618184e-06,
+      "loss": 2.3169,
+      "step": 2020
+    },
+    {
+      "epoch": 0.779556412729026,
+      "grad_norm": 0.4412302630768088,
+      "learning_rate": 1.1852078480097502e-06,
+      "loss": 2.3269,
+      "step": 2021
+    },
+    {
+      "epoch": 0.7799421407907425,
+      "grad_norm": 0.42791139175603443,
+      "learning_rate": 1.1812317527410316e-06,
+      "loss": 2.4182,
+      "step": 2022
+    },
+    {
+      "epoch": 0.780327868852459,
+      "grad_norm": 0.43155330381976453,
+      "learning_rate": 1.1772614445811902e-06,
+      "loss": 2.3056,
+      "step": 2023
+    },
+    {
+      "epoch": 0.7807135969141755,
+      "grad_norm": 0.4248080667352216,
+      "learning_rate": 1.173296929546987e-06,
+      "loss": 2.3511,
+      "step": 2024
+    },
+    {
+      "epoch": 0.781099324975892,
+      "grad_norm": 0.47188007333530674,
+      "learning_rate": 1.1693382136463981e-06,
+      "loss": 2.3643,
+      "step": 2025
+    },
+    {
+      "epoch": 0.7814850530376085,
+      "grad_norm": 0.4554633115058257,
+      "learning_rate": 1.1653853028786177e-06,
+      "loss": 2.319,
+      "step": 2026
+    },
+    {
+      "epoch": 0.781870781099325,
+      "grad_norm": 0.41320717435656057,
+      "learning_rate": 1.161438203234037e-06,
+      "loss": 2.324,
+      "step": 2027
+    },
+    {
+      "epoch": 0.7822565091610415,
+      "grad_norm": 0.44992969789800136,
+      "learning_rate": 1.1574969206942443e-06,
+      "loss": 2.3228,
+      "step": 2028
+    },
+    {
+      "epoch": 0.7826422372227579,
+      "grad_norm": 0.42752178483527925,
+      "learning_rate": 1.15356146123201e-06,
+      "loss": 2.3232,
+      "step": 2029
+    },
+    {
+      "epoch": 0.7830279652844745,
+      "grad_norm": 0.42090346482240726,
+      "learning_rate": 1.149631830811283e-06,
+      "loss": 2.3241,
+      "step": 2030
+    },
+    {
+      "epoch": 0.7834136933461909,
+      "grad_norm": 0.40632873911064693,
+      "learning_rate": 1.145708035387177e-06,
+      "loss": 2.3086,
+      "step": 2031
+    },
+    {
+      "epoch": 0.7837994214079074,
+      "grad_norm": 0.43403847555527236,
+      "learning_rate": 1.1417900809059623e-06,
+      "loss": 2.3284,
+      "step": 2032
+    },
+    {
+      "epoch": 0.7841851494696239,
+      "grad_norm": 0.45825087452891977,
+      "learning_rate": 1.1378779733050583e-06,
+      "loss": 2.2471,
+      "step": 2033
+    },
+    {
+      "epoch": 0.7845708775313404,
+      "grad_norm": 0.44605854512315274,
+      "learning_rate": 1.1339717185130228e-06,
+      "loss": 2.3399,
+      "step": 2034
+    },
+    {
+      "epoch": 0.7849566055930569,
+      "grad_norm": 0.445422337924896,
+      "learning_rate": 1.1300713224495485e-06,
+      "loss": 2.3489,
+      "step": 2035
+    },
+    {
+      "epoch": 0.7853423336547734,
+      "grad_norm": 0.42565645651135264,
+      "learning_rate": 1.1261767910254422e-06,
+      "loss": 2.2588,
+      "step": 2036
+    },
+    {
+      "epoch": 0.7857280617164899,
+      "grad_norm": 0.4181093795070767,
+      "learning_rate": 1.1222881301426314e-06,
+      "loss": 2.3327,
+      "step": 2037
+    },
+    {
+      "epoch": 0.7861137897782063,
+      "grad_norm": 0.42210516058715325,
+      "learning_rate": 1.1184053456941407e-06,
+      "loss": 2.2581,
+      "step": 2038
+    },
+    {
+      "epoch": 0.7864995178399229,
+      "grad_norm": 0.40726524748024123,
+      "learning_rate": 1.1145284435640918e-06,
+      "loss": 2.3303,
+      "step": 2039
+    },
+    {
+      "epoch": 0.7868852459016393,
+      "grad_norm": 0.42613476920779403,
+      "learning_rate": 1.1106574296276923e-06,
+      "loss": 2.3447,
+      "step": 2040
+    },
+    {
+      "epoch": 0.7872709739633559,
+      "grad_norm": 0.4229952696715132,
+      "learning_rate": 1.1067923097512256e-06,
+      "loss": 2.3759,
+      "step": 2041
+    },
+    {
+      "epoch": 0.7876567020250723,
+      "grad_norm": 0.3964771073110038,
+      "learning_rate": 1.102933089792042e-06,
+      "loss": 2.3604,
+      "step": 2042
+    },
+    {
+      "epoch": 0.7880424300867888,
+      "grad_norm": 0.39912612226497574,
+      "learning_rate": 1.0990797755985567e-06,
+      "loss": 2.2094,
+      "step": 2043
+    },
+    {
+      "epoch": 0.7884281581485053,
+      "grad_norm": 0.41715700699022745,
+      "learning_rate": 1.095232373010226e-06,
+      "loss": 2.3472,
+      "step": 2044
+    },
+    {
+      "epoch": 0.7888138862102217,
+      "grad_norm": 0.4661905801110712,
+      "learning_rate": 1.0913908878575568e-06,
+      "loss": 2.372,
+      "step": 2045
+    },
+    {
+      "epoch": 0.7891996142719383,
+      "grad_norm": 0.4475875561296712,
+      "learning_rate": 1.0875553259620825e-06,
+      "loss": 2.3541,
+      "step": 2046
+    },
+    {
+      "epoch": 0.7895853423336547,
+      "grad_norm": 0.4137122903983067,
+      "learning_rate": 1.0837256931363605e-06,
+      "loss": 2.3271,
+      "step": 2047
+    },
+    {
+      "epoch": 0.7899710703953713,
+      "grad_norm": 0.43475606513803694,
+      "learning_rate": 1.0799019951839656e-06,
+      "loss": 2.3564,
+      "step": 2048
+    },
+    {
+      "epoch": 0.7903567984570877,
+      "grad_norm": 0.42735929912744547,
+      "learning_rate": 1.0760842378994758e-06,
+      "loss": 2.2648,
+      "step": 2049
+    },
+    {
+      "epoch": 0.7907425265188043,
+      "grad_norm": 0.4054186470588228,
+      "learning_rate": 1.0722724270684698e-06,
+      "loss": 2.2956,
+      "step": 2050
+    },
+    {
+      "epoch": 0.7911282545805207,
+      "grad_norm": 0.4164187575257658,
+      "learning_rate": 1.068466568467512e-06,
+      "loss": 2.3812,
+      "step": 2051
+    },
+    {
+      "epoch": 0.7915139826422373,
+      "grad_norm": 0.4209567437960228,
+      "learning_rate": 1.0646666678641477e-06,
+      "loss": 2.3218,
+      "step": 2052
+    },
+    {
+      "epoch": 0.7918997107039537,
+      "grad_norm": 0.4200255182068855,
+      "learning_rate": 1.0608727310168921e-06,
+      "loss": 2.2764,
+      "step": 2053
+    },
+    {
+      "epoch": 0.7922854387656703,
+      "grad_norm": 0.4555452108402252,
+      "learning_rate": 1.0570847636752251e-06,
+      "loss": 2.297,
+      "step": 2054
+    },
+    {
+      "epoch": 0.7926711668273867,
+      "grad_norm": 0.43866423150867556,
+      "learning_rate": 1.0533027715795784e-06,
+      "loss": 2.3602,
+      "step": 2055
+    },
+    {
+      "epoch": 0.7930568948891031,
+      "grad_norm": 0.4040195252185418,
+      "learning_rate": 1.0495267604613273e-06,
+      "loss": 2.3524,
+      "step": 2056
+    },
+    {
+      "epoch": 0.7934426229508197,
+      "grad_norm": 0.41119784078419164,
+      "learning_rate": 1.0457567360427872e-06,
+      "loss": 2.3815,
+      "step": 2057
+    },
+    {
+      "epoch": 0.7938283510125361,
+      "grad_norm": 0.4082097704919852,
+      "learning_rate": 1.041992704037198e-06,
+      "loss": 2.2881,
+      "step": 2058
+    },
+    {
+      "epoch": 0.7942140790742527,
+      "grad_norm": 0.44778947055140056,
+      "learning_rate": 1.0382346701487183e-06,
+      "loss": 2.4268,
+      "step": 2059
+    },
+    {
+      "epoch": 0.7945998071359691,
+      "grad_norm": 0.41670178757828186,
+      "learning_rate": 1.0344826400724185e-06,
+      "loss": 2.2601,
+      "step": 2060
+    },
+    {
+      "epoch": 0.7949855351976857,
+      "grad_norm": 0.4050164213793503,
+      "learning_rate": 1.030736619494268e-06,
+      "loss": 2.2754,
+      "step": 2061
+    },
+    {
+      "epoch": 0.7953712632594021,
+      "grad_norm": 0.41930209822290115,
+      "learning_rate": 1.0269966140911343e-06,
+      "loss": 2.3653,
+      "step": 2062
+    },
+    {
+      "epoch": 0.7957569913211187,
+      "grad_norm": 0.43621583234469957,
+      "learning_rate": 1.023262629530763e-06,
+      "loss": 2.2771,
+      "step": 2063
+    },
+    {
+      "epoch": 0.7961427193828351,
+      "grad_norm": 0.44674323122071313,
+      "learning_rate": 1.0195346714717813e-06,
+      "loss": 2.429,
+      "step": 2064
+    },
+    {
+      "epoch": 0.7965284474445516,
+      "grad_norm": 0.4344073799130207,
+      "learning_rate": 1.015812745563679e-06,
+      "loss": 2.2308,
+      "step": 2065
+    },
+    {
+      "epoch": 0.7969141755062681,
+      "grad_norm": 0.4339314134771839,
+      "learning_rate": 1.012096857446807e-06,
+      "loss": 2.3263,
+      "step": 2066
+    },
+    {
+      "epoch": 0.7972999035679845,
+      "grad_norm": 0.41138510874721523,
+      "learning_rate": 1.0083870127523659e-06,
+      "loss": 2.2738,
+      "step": 2067
+    },
+    {
+      "epoch": 0.7976856316297011,
+      "grad_norm": 0.44137290067576496,
+      "learning_rate": 1.0046832171023952e-06,
+      "loss": 2.4554,
+      "step": 2068
+    },
+    {
+      "epoch": 0.7980713596914175,
+      "grad_norm": 0.4232345697274198,
+      "learning_rate": 1.0009854761097736e-06,
+      "loss": 2.3486,
+      "step": 2069
+    },
+    {
+      "epoch": 0.7984570877531341,
+      "grad_norm": 0.42589530418445515,
+      "learning_rate": 9.972937953781985e-07,
+      "loss": 2.2941,
+      "step": 2070
+    },
+    {
+      "epoch": 0.7988428158148505,
+      "grad_norm": 0.4431955037608249,
+      "learning_rate": 9.936081805021859e-07,
+      "loss": 2.2909,
+      "step": 2071
+    },
+    {
+      "epoch": 0.799228543876567,
+      "grad_norm": 0.41161461847207664,
+      "learning_rate": 9.899286370670575e-07,
+      "loss": 2.3233,
+      "step": 2072
+    },
+    {
+      "epoch": 0.7996142719382835,
+      "grad_norm": 0.43207424360876473,
+      "learning_rate": 9.862551706489382e-07,
+      "loss": 2.2991,
+      "step": 2073
+    },
+    {
+      "epoch": 0.8,
+      "grad_norm": 0.4216834706882256,
+      "learning_rate": 9.825877868147393e-07,
+      "loss": 2.3383,
+      "step": 2074
+    },
+    {
+      "epoch": 0.8003857280617165,
+      "grad_norm": 0.3959572897055696,
+      "learning_rate": 9.789264911221546e-07,
+      "loss": 2.2559,
+      "step": 2075
+    },
+    {
+      "epoch": 0.800771456123433,
+      "grad_norm": 0.4385171900226881,
+      "learning_rate": 9.752712891196558e-07,
+      "loss": 2.3079,
+      "step": 2076
+    },
+    {
+      "epoch": 0.8011571841851495,
+      "grad_norm": 0.452752952242241,
+      "learning_rate": 9.716221863464764e-07,
+      "loss": 2.3363,
+      "step": 2077
+    },
+    {
+      "epoch": 0.8015429122468659,
+      "grad_norm": 0.4393729464245923,
+      "learning_rate": 9.679791883326067e-07,
+      "loss": 2.2671,
+      "step": 2078
+    },
+    {
+      "epoch": 0.8019286403085825,
+      "grad_norm": 0.4390517173709909,
+      "learning_rate": 9.643423005987868e-07,
+      "loss": 2.2943,
+      "step": 2079
+    },
+    {
+      "epoch": 0.8023143683702989,
+      "grad_norm": 0.45152023580011974,
+      "learning_rate": 9.607115286564972e-07,
+      "loss": 2.3131,
+      "step": 2080
+    },
+    {
+      "epoch": 0.8027000964320155,
+      "grad_norm": 0.41921933819791163,
+      "learning_rate": 9.570868780079485e-07,
+      "loss": 2.2806,
+      "step": 2081
+    },
+    {
+      "epoch": 0.8030858244937319,
+      "grad_norm": 0.4239070139144562,
+      "learning_rate": 9.534683541460771e-07,
+      "loss": 2.2803,
+      "step": 2082
+    },
+    {
+      "epoch": 0.8034715525554484,
+      "grad_norm": 0.42556425223869837,
+      "learning_rate": 9.498559625545362e-07,
+      "loss": 2.3287,
+      "step": 2083
+    },
+    {
+      "epoch": 0.8038572806171649,
+      "grad_norm": 0.42612821228367675,
+      "learning_rate": 9.46249708707681e-07,
+      "loss": 2.3217,
+      "step": 2084
+    },
+    {
+      "epoch": 0.8042430086788814,
+      "grad_norm": 0.4124648062571102,
+      "learning_rate": 9.426495980705685e-07,
+      "loss": 2.3432,
+      "step": 2085
+    },
+    {
+      "epoch": 0.8046287367405979,
+      "grad_norm": 0.39315266190910947,
+      "learning_rate": 9.39055636098945e-07,
+      "loss": 2.3182,
+      "step": 2086
+    },
+    {
+      "epoch": 0.8050144648023144,
+      "grad_norm": 0.40497942402672604,
+      "learning_rate": 9.354678282392399e-07,
+      "loss": 2.3232,
+      "step": 2087
+    },
+    {
+      "epoch": 0.8054001928640309,
+      "grad_norm": 0.43575039141039523,
+      "learning_rate": 9.318861799285539e-07,
+      "loss": 2.2737,
+      "step": 2088
+    },
+    {
+      "epoch": 0.8057859209257473,
+      "grad_norm": 0.4195391658604251,
+      "learning_rate": 9.283106965946581e-07,
+      "loss": 2.3175,
+      "step": 2089
+    },
+    {
+      "epoch": 0.8061716489874639,
+      "grad_norm": 0.42360172858140893,
+      "learning_rate": 9.247413836559765e-07,
+      "loss": 2.254,
+      "step": 2090
+    },
+    {
+      "epoch": 0.8065573770491803,
+      "grad_norm": 0.4252842232896849,
+      "learning_rate": 9.211782465215829e-07,
+      "loss": 2.3785,
+      "step": 2091
+    },
+    {
+      "epoch": 0.8069431051108968,
+      "grad_norm": 0.4354684092231028,
+      "learning_rate": 9.176212905911946e-07,
+      "loss": 2.3159,
+      "step": 2092
+    },
+    {
+      "epoch": 0.8073288331726133,
+      "grad_norm": 0.4120198398311047,
+      "learning_rate": 9.140705212551599e-07,
+      "loss": 2.3067,
+      "step": 2093
+    },
+    {
+      "epoch": 0.8077145612343298,
+      "grad_norm": 0.4062676518671219,
+      "learning_rate": 9.105259438944508e-07,
+      "loss": 2.3309,
+      "step": 2094
+    },
+    {
+      "epoch": 0.8081002892960463,
+      "grad_norm": 0.4218432803347716,
+      "learning_rate": 9.069875638806558e-07,
+      "loss": 2.3381,
+      "step": 2095
+    },
+    {
+      "epoch": 0.8084860173577628,
+      "grad_norm": 0.44741866515440026,
+      "learning_rate": 9.034553865759754e-07,
+      "loss": 2.2959,
+      "step": 2096
+    },
+    {
+      "epoch": 0.8088717454194793,
+      "grad_norm": 0.42302813421063773,
+      "learning_rate": 8.999294173332058e-07,
+      "loss": 2.258,
+      "step": 2097
+    },
+    {
+      "epoch": 0.8092574734811958,
+      "grad_norm": 0.40136791403407035,
+      "learning_rate": 8.964096614957374e-07,
+      "loss": 2.3537,
+      "step": 2098
+    },
+    {
+      "epoch": 0.8096432015429122,
+      "grad_norm": 0.41843971603799174,
+      "learning_rate": 8.928961243975437e-07,
+      "loss": 2.3205,
+      "step": 2099
+    },
+    {
+      "epoch": 0.8100289296046287,
+      "grad_norm": 0.41636229509726935,
+      "learning_rate": 8.893888113631732e-07,
+      "loss": 2.3583,
+      "step": 2100
+    },
+    {
+      "epoch": 0.8104146576663452,
+      "grad_norm": 0.41726692500260676,
+      "learning_rate": 8.858877277077455e-07,
+      "loss": 2.3512,
+      "step": 2101
+    },
+    {
+      "epoch": 0.8108003857280617,
+      "grad_norm": 0.41826271820553335,
+      "learning_rate": 8.823928787369379e-07,
+      "loss": 2.2571,
+      "step": 2102
+    },
+    {
+      "epoch": 0.8111861137897782,
+      "grad_norm": 0.4342375756262929,
+      "learning_rate": 8.789042697469796e-07,
+      "loss": 2.3969,
+      "step": 2103
+    },
+    {
+      "epoch": 0.8115718418514947,
+      "grad_norm": 0.4768657159728018,
+      "learning_rate": 8.754219060246432e-07,
+      "loss": 2.2958,
+      "step": 2104
+    },
+    {
+      "epoch": 0.8119575699132112,
+      "grad_norm": 0.4123373053535353,
+      "learning_rate": 8.719457928472364e-07,
+      "loss": 2.264,
+      "step": 2105
+    },
+    {
+      "epoch": 0.8123432979749277,
+      "grad_norm": 0.44858485725351716,
+      "learning_rate": 8.684759354825962e-07,
+      "loss": 2.2995,
+      "step": 2106
+    },
+    {
+      "epoch": 0.8127290260366442,
+      "grad_norm": 0.41609733353522543,
+      "learning_rate": 8.650123391890763e-07,
+      "loss": 2.3769,
+      "step": 2107
+    },
+    {
+      "epoch": 0.8131147540983606,
+      "grad_norm": 0.40205999326820274,
+      "learning_rate": 8.615550092155478e-07,
+      "loss": 2.3102,
+      "step": 2108
+    },
+    {
+      "epoch": 0.8135004821600772,
+      "grad_norm": 0.4545386030204492,
+      "learning_rate": 8.581039508013788e-07,
+      "loss": 2.2828,
+      "step": 2109
+    },
+    {
+      "epoch": 0.8138862102217936,
+      "grad_norm": 0.40903450628581595,
+      "learning_rate": 8.546591691764388e-07,
+      "loss": 2.3142,
+      "step": 2110
+    },
+    {
+      "epoch": 0.8142719382835101,
+      "grad_norm": 0.43073974568345547,
+      "learning_rate": 8.512206695610825e-07,
+      "loss": 2.3096,
+      "step": 2111
+    },
+    {
+      "epoch": 0.8146576663452266,
+      "grad_norm": 0.4310644884162425,
+      "learning_rate": 8.477884571661449e-07,
+      "loss": 2.3036,
+      "step": 2112
+    },
+    {
+      "epoch": 0.8150433944069431,
+      "grad_norm": 0.43803695575052093,
+      "learning_rate": 8.443625371929326e-07,
+      "loss": 2.3054,
+      "step": 2113
+    },
+    {
+      "epoch": 0.8154291224686596,
+      "grad_norm": 0.41544065559345766,
+      "learning_rate": 8.40942914833216e-07,
+      "loss": 2.3236,
+      "step": 2114
+    },
+    {
+      "epoch": 0.8158148505303761,
+      "grad_norm": 0.42351460560781906,
+      "learning_rate": 8.375295952692258e-07,
+      "loss": 2.383,
+      "step": 2115
+    },
+    {
+      "epoch": 0.8162005785920926,
+      "grad_norm": 0.4622928943201081,
+      "learning_rate": 8.341225836736367e-07,
+      "loss": 2.3541,
+      "step": 2116
+    },
+    {
+      "epoch": 0.816586306653809,
+      "grad_norm": 0.3875682392366773,
+      "learning_rate": 8.30721885209565e-07,
+      "loss": 2.26,
+      "step": 2117
+    },
+    {
+      "epoch": 0.8169720347155256,
+      "grad_norm": 0.4227521615438232,
+      "learning_rate": 8.273275050305618e-07,
+      "loss": 2.2635,
+      "step": 2118
+    },
+    {
+      "epoch": 0.817357762777242,
+      "grad_norm": 0.42449931643301214,
+      "learning_rate": 8.239394482805996e-07,
+      "loss": 2.3466,
+      "step": 2119
+    },
+    {
+      "epoch": 0.8177434908389586,
+      "grad_norm": 0.4285021268387793,
+      "learning_rate": 8.20557720094074e-07,
+      "loss": 2.2994,
+      "step": 2120
+    },
+    {
+      "epoch": 0.818129218900675,
+      "grad_norm": 0.43324914742483134,
+      "learning_rate": 8.171823255957828e-07,
+      "loss": 2.3021,
+      "step": 2121
+    },
+    {
+      "epoch": 0.8185149469623915,
+      "grad_norm": 0.43250542249302915,
+      "learning_rate": 8.138132699009321e-07,
+      "loss": 2.3227,
+      "step": 2122
+    },
+    {
+      "epoch": 0.818900675024108,
+      "grad_norm": 0.4718675403886274,
+      "learning_rate": 8.104505581151184e-07,
+      "loss": 2.3697,
+      "step": 2123
+    },
+    {
+      "epoch": 0.8192864030858245,
+      "grad_norm": 0.4216607077105276,
+      "learning_rate": 8.070941953343242e-07,
+      "loss": 2.3251,
+      "step": 2124
+    },
+    {
+      "epoch": 0.819672131147541,
+      "grad_norm": 0.40915512928375525,
+      "learning_rate": 8.037441866449114e-07,
+      "loss": 2.3603,
+      "step": 2125
+    },
+    {
+      "epoch": 0.8200578592092574,
+      "grad_norm": 0.46439779886388954,
+      "learning_rate": 8.004005371236128e-07,
+      "loss": 2.2923,
+      "step": 2126
+    },
+    {
+      "epoch": 0.820443587270974,
+      "grad_norm": 0.4024444417990968,
+      "learning_rate": 7.970632518375232e-07,
+      "loss": 2.3659,
+      "step": 2127
+    },
+    {
+      "epoch": 0.8208293153326904,
+      "grad_norm": 0.44130633810097636,
+      "learning_rate": 7.937323358440935e-07,
+      "loss": 2.3072,
+      "step": 2128
+    },
+    {
+      "epoch": 0.821215043394407,
+      "grad_norm": 0.41030222383464565,
+      "learning_rate": 7.904077941911248e-07,
+      "loss": 2.2652,
+      "step": 2129
+    },
+    {
+      "epoch": 0.8216007714561234,
+      "grad_norm": 0.4239855735926553,
+      "learning_rate": 7.870896319167548e-07,
+      "loss": 2.2922,
+      "step": 2130
+    },
+    {
+      "epoch": 0.82198649951784,
+      "grad_norm": 0.4264318179022017,
+      "learning_rate": 7.83777854049454e-07,
+      "loss": 2.461,
+      "step": 2131
+    },
+    {
+      "epoch": 0.8223722275795564,
+      "grad_norm": 0.39931907861667765,
+      "learning_rate": 7.804724656080182e-07,
+      "loss": 2.2963,
+      "step": 2132
+    },
+    {
+      "epoch": 0.8227579556412729,
+      "grad_norm": 0.44069287415695574,
+      "learning_rate": 7.771734716015611e-07,
+      "loss": 2.3215,
+      "step": 2133
+    },
+    {
+      "epoch": 0.8231436837029894,
+      "grad_norm": 0.4764007174469988,
+      "learning_rate": 7.738808770295064e-07,
+      "loss": 2.3178,
+      "step": 2134
+    },
+    {
+      "epoch": 0.8235294117647058,
+      "grad_norm": 0.40896327860723114,
+      "learning_rate": 7.705946868815783e-07,
+      "loss": 2.3397,
+      "step": 2135
+    },
+    {
+      "epoch": 0.8239151398264224,
+      "grad_norm": 0.4133582523462277,
+      "learning_rate": 7.673149061377966e-07,
+      "loss": 2.295,
+      "step": 2136
+    },
+    {
+      "epoch": 0.8243008678881388,
+      "grad_norm": 0.44252932709212506,
+      "learning_rate": 7.64041539768467e-07,
+      "loss": 2.3135,
+      "step": 2137
+    },
+    {
+      "epoch": 0.8246865959498554,
+      "grad_norm": 0.4315479104904479,
+      "learning_rate": 7.607745927341764e-07,
+      "loss": 2.364,
+      "step": 2138
+    },
+    {
+      "epoch": 0.8250723240115718,
+      "grad_norm": 0.4230358525826942,
+      "learning_rate": 7.575140699857819e-07,
+      "loss": 2.2812,
+      "step": 2139
+    },
+    {
+      "epoch": 0.8254580520732884,
+      "grad_norm": 0.4407423704703889,
+      "learning_rate": 7.542599764644049e-07,
+      "loss": 2.2873,
+      "step": 2140
+    },
+    {
+      "epoch": 0.8258437801350048,
+      "grad_norm": 0.4315046975539552,
+      "learning_rate": 7.510123171014255e-07,
+      "loss": 2.2976,
+      "step": 2141
+    },
+    {
+      "epoch": 0.8262295081967214,
+      "grad_norm": 0.440416523656838,
+      "learning_rate": 7.477710968184726e-07,
+      "loss": 2.3353,
+      "step": 2142
+    },
+    {
+      "epoch": 0.8266152362584378,
+      "grad_norm": 0.4283525113622728,
+      "learning_rate": 7.445363205274153e-07,
+      "loss": 2.3893,
+      "step": 2143
+    },
+    {
+      "epoch": 0.8270009643201542,
+      "grad_norm": 0.42482838905355197,
+      "learning_rate": 7.413079931303591e-07,
+      "loss": 2.3279,
+      "step": 2144
+    },
+    {
+      "epoch": 0.8273866923818708,
+      "grad_norm": 0.42004246146651786,
+      "learning_rate": 7.380861195196359e-07,
+      "loss": 2.2928,
+      "step": 2145
+    },
+    {
+      "epoch": 0.8277724204435872,
+      "grad_norm": 0.4646102318061459,
+      "learning_rate": 7.348707045777959e-07,
+      "loss": 2.3576,
+      "step": 2146
+    },
+    {
+      "epoch": 0.8281581485053038,
+      "grad_norm": 0.43910211059318865,
+      "learning_rate": 7.316617531776049e-07,
+      "loss": 2.36,
+      "step": 2147
+    },
+    {
+      "epoch": 0.8285438765670202,
+      "grad_norm": 0.4473140935403787,
+      "learning_rate": 7.284592701820325e-07,
+      "loss": 2.2544,
+      "step": 2148
+    },
+    {
+      "epoch": 0.8289296046287368,
+      "grad_norm": 0.4477860585093107,
+      "learning_rate": 7.252632604442439e-07,
+      "loss": 2.2522,
+      "step": 2149
+    },
+    {
+      "epoch": 0.8293153326904532,
+      "grad_norm": 0.4139567407677305,
+      "learning_rate": 7.220737288075958e-07,
+      "loss": 2.3147,
+      "step": 2150
+    },
+    {
+      "epoch": 0.8297010607521698,
+      "grad_norm": 0.4053269153531082,
+      "learning_rate": 7.188906801056277e-07,
+      "loss": 2.2956,
+      "step": 2151
+    },
+    {
+      "epoch": 0.8300867888138862,
+      "grad_norm": 0.4441494340410561,
+      "learning_rate": 7.157141191620548e-07,
+      "loss": 2.3576,
+      "step": 2152
+    },
+    {
+      "epoch": 0.8304725168756028,
+      "grad_norm": 0.40325041084116087,
+      "learning_rate": 7.125440507907583e-07,
+      "loss": 2.3164,
+      "step": 2153
+    },
+    {
+      "epoch": 0.8308582449373192,
+      "grad_norm": 0.40994622003056547,
+      "learning_rate": 7.093804797957849e-07,
+      "loss": 2.2081,
+      "step": 2154
+    },
+    {
+      "epoch": 0.8312439729990356,
+      "grad_norm": 0.4436215541711604,
+      "learning_rate": 7.062234109713318e-07,
+      "loss": 2.3287,
+      "step": 2155
+    },
+    {
+      "epoch": 0.8316297010607522,
+      "grad_norm": 0.41634391743990945,
+      "learning_rate": 7.030728491017408e-07,
+      "loss": 2.2857,
+      "step": 2156
+    },
+    {
+      "epoch": 0.8320154291224686,
+      "grad_norm": 0.40361544747904327,
+      "learning_rate": 6.999287989614972e-07,
+      "loss": 2.377,
+      "step": 2157
+    },
+    {
+      "epoch": 0.8324011571841852,
+      "grad_norm": 0.40753496942994805,
+      "learning_rate": 6.967912653152164e-07,
+      "loss": 2.2995,
+      "step": 2158
+    },
+    {
+      "epoch": 0.8327868852459016,
+      "grad_norm": 0.42439267943337,
+      "learning_rate": 6.936602529176367e-07,
+      "loss": 2.3756,
+      "step": 2159
+    },
+    {
+      "epoch": 0.8331726133076182,
+      "grad_norm": 0.4113830283805619,
+      "learning_rate": 6.905357665136142e-07,
+      "loss": 2.245,
+      "step": 2160
+    },
+    {
+      "epoch": 0.8335583413693346,
+      "grad_norm": 0.42612328379407305,
+      "learning_rate": 6.874178108381191e-07,
+      "loss": 2.3246,
+      "step": 2161
+    },
+    {
+      "epoch": 0.8339440694310511,
+      "grad_norm": 0.46288626517979153,
+      "learning_rate": 6.8430639061622e-07,
+      "loss": 2.2819,
+      "step": 2162
+    },
+    {
+      "epoch": 0.8343297974927676,
+      "grad_norm": 0.41650229120385046,
+      "learning_rate": 6.812015105630842e-07,
+      "loss": 2.412,
+      "step": 2163
+    },
+    {
+      "epoch": 0.8347155255544841,
+      "grad_norm": 0.4217207294993816,
+      "learning_rate": 6.781031753839662e-07,
+      "loss": 2.2941,
+      "step": 2164
+    },
+    {
+      "epoch": 0.8351012536162006,
+      "grad_norm": 0.4456538316970643,
+      "learning_rate": 6.750113897742017e-07,
+      "loss": 2.2193,
+      "step": 2165
+    },
+    {
+      "epoch": 0.835486981677917,
+      "grad_norm": 0.41758648776543356,
+      "learning_rate": 6.719261584192038e-07,
+      "loss": 2.2988,
+      "step": 2166
+    },
+    {
+      "epoch": 0.8358727097396336,
+      "grad_norm": 0.4186724447342165,
+      "learning_rate": 6.6884748599445e-07,
+      "loss": 2.3763,
+      "step": 2167
+    },
+    {
+      "epoch": 0.83625843780135,
+      "grad_norm": 0.4092927778532757,
+      "learning_rate": 6.657753771654812e-07,
+      "loss": 2.2815,
+      "step": 2168
+    },
+    {
+      "epoch": 0.8366441658630666,
+      "grad_norm": 0.41925614674502665,
+      "learning_rate": 6.627098365878886e-07,
+      "loss": 2.3579,
+      "step": 2169
+    },
+    {
+      "epoch": 0.837029893924783,
+      "grad_norm": 0.4447516307021346,
+      "learning_rate": 6.596508689073105e-07,
+      "loss": 2.3064,
+      "step": 2170
+    },
+    {
+      "epoch": 0.8374156219864995,
+      "grad_norm": 0.4357105661743486,
+      "learning_rate": 6.565984787594248e-07,
+      "loss": 2.2484,
+      "step": 2171
+    },
+    {
+      "epoch": 0.837801350048216,
+      "grad_norm": 0.4454408567281427,
+      "learning_rate": 6.535526707699408e-07,
+      "loss": 2.3461,
+      "step": 2172
+    },
+    {
+      "epoch": 0.8381870781099325,
+      "grad_norm": 0.421682585445223,
+      "learning_rate": 6.505134495545951e-07,
+      "loss": 2.364,
+      "step": 2173
+    },
+    {
+      "epoch": 0.838572806171649,
+      "grad_norm": 0.44971672903581555,
+      "learning_rate": 6.474808197191401e-07,
+      "loss": 2.3122,
+      "step": 2174
+    },
+    {
+      "epoch": 0.8389585342333655,
+      "grad_norm": 0.4430760452217771,
+      "learning_rate": 6.444547858593392e-07,
+      "loss": 2.3761,
+      "step": 2175
+    },
+    {
+      "epoch": 0.839344262295082,
+      "grad_norm": 0.4320460359114813,
+      "learning_rate": 6.414353525609628e-07,
+      "loss": 2.3304,
+      "step": 2176
+    },
+    {
+      "epoch": 0.8397299903567984,
+      "grad_norm": 0.40090412609242976,
+      "learning_rate": 6.384225243997765e-07,
+      "loss": 2.3369,
+      "step": 2177
+    },
+    {
+      "epoch": 0.840115718418515,
+      "grad_norm": 0.4307273849934172,
+      "learning_rate": 6.354163059415353e-07,
+      "loss": 2.3445,
+      "step": 2178
+    },
+    {
+      "epoch": 0.8405014464802314,
+      "grad_norm": 0.4314592802294269,
+      "learning_rate": 6.32416701741978e-07,
+      "loss": 2.2853,
+      "step": 2179
+    },
+    {
+      "epoch": 0.840887174541948,
+      "grad_norm": 0.40742818362166616,
+      "learning_rate": 6.294237163468231e-07,
+      "loss": 2.3268,
+      "step": 2180
+    },
+    {
+      "epoch": 0.8412729026036644,
+      "grad_norm": 0.46955868732988343,
+      "learning_rate": 6.264373542917551e-07,
+      "loss": 2.3619,
+      "step": 2181
+    },
+    {
+      "epoch": 0.8416586306653809,
+      "grad_norm": 0.42557126770219617,
+      "learning_rate": 6.234576201024223e-07,
+      "loss": 2.3603,
+      "step": 2182
+    },
+    {
+      "epoch": 0.8420443587270974,
+      "grad_norm": 0.42459882470639426,
+      "learning_rate": 6.204845182944292e-07,
+      "loss": 2.3542,
+      "step": 2183
+    },
+    {
+      "epoch": 0.8424300867888139,
+      "grad_norm": 0.4319083557644331,
+      "learning_rate": 6.175180533733277e-07,
+      "loss": 2.3457,
+      "step": 2184
+    },
+    {
+      "epoch": 0.8428158148505304,
+      "grad_norm": 0.40950629424784474,
+      "learning_rate": 6.145582298346153e-07,
+      "loss": 2.3043,
+      "step": 2185
+    },
+    {
+      "epoch": 0.8432015429122469,
+      "grad_norm": 0.40940256778256134,
+      "learning_rate": 6.116050521637218e-07,
+      "loss": 2.3052,
+      "step": 2186
+    },
+    {
+      "epoch": 0.8435872709739634,
+      "grad_norm": 0.46231089574854295,
+      "learning_rate": 6.086585248360072e-07,
+      "loss": 2.2459,
+      "step": 2187
+    },
+    {
+      "epoch": 0.8439729990356798,
+      "grad_norm": 0.43997580961979005,
+      "learning_rate": 6.057186523167529e-07,
+      "loss": 2.4005,
+      "step": 2188
+    },
+    {
+      "epoch": 0.8443587270973963,
+      "grad_norm": 0.4443737132363861,
+      "learning_rate": 6.027854390611548e-07,
+      "loss": 2.3286,
+      "step": 2189
+    },
+    {
+      "epoch": 0.8447444551591128,
+      "grad_norm": 0.45487195535732927,
+      "learning_rate": 5.998588895143181e-07,
+      "loss": 2.2987,
+      "step": 2190
+    },
+    {
+      "epoch": 0.8451301832208293,
+      "grad_norm": 0.4104730105003813,
+      "learning_rate": 5.96939008111248e-07,
+      "loss": 2.3644,
+      "step": 2191
+    },
+    {
+      "epoch": 0.8455159112825458,
+      "grad_norm": 0.3998158167331894,
+      "learning_rate": 5.940257992768456e-07,
+      "loss": 2.3193,
+      "step": 2192
+    },
+    {
+      "epoch": 0.8459016393442623,
+      "grad_norm": 0.44317647366113244,
+      "learning_rate": 5.911192674259015e-07,
+      "loss": 2.3473,
+      "step": 2193
+    },
+    {
+      "epoch": 0.8462873674059788,
+      "grad_norm": 0.40494619795811615,
+      "learning_rate": 5.882194169630845e-07,
+      "loss": 2.2956,
+      "step": 2194
+    },
+    {
+      "epoch": 0.8466730954676953,
+      "grad_norm": 0.42962309814899585,
+      "learning_rate": 5.853262522829417e-07,
+      "loss": 2.3451,
+      "step": 2195
+    },
+    {
+      "epoch": 0.8470588235294118,
+      "grad_norm": 0.4168087040989416,
+      "learning_rate": 5.824397777698859e-07,
+      "loss": 2.2744,
+      "step": 2196
+    },
+    {
+      "epoch": 0.8474445515911283,
+      "grad_norm": 0.4589074894558346,
+      "learning_rate": 5.795599977981914e-07,
+      "loss": 2.286,
+      "step": 2197
+    },
+    {
+      "epoch": 0.8478302796528447,
+      "grad_norm": 0.4270457108660484,
+      "learning_rate": 5.766869167319893e-07,
+      "loss": 2.2972,
+      "step": 2198
+    },
+    {
+      "epoch": 0.8482160077145612,
+      "grad_norm": 0.41266450983386743,
+      "learning_rate": 5.738205389252555e-07,
+      "loss": 2.2554,
+      "step": 2199
+    },
+    {
+      "epoch": 0.8486017357762777,
+      "grad_norm": 0.4257329026347416,
+      "learning_rate": 5.709608687218116e-07,
+      "loss": 2.3224,
+      "step": 2200
+    },
+    {
+      "epoch": 0.8489874638379942,
+      "grad_norm": 0.44714776911423726,
+      "learning_rate": 5.681079104553122e-07,
+      "loss": 2.3076,
+      "step": 2201
+    },
+    {
+      "epoch": 0.8493731918997107,
+      "grad_norm": 0.4411737147369575,
+      "learning_rate": 5.652616684492396e-07,
+      "loss": 2.3045,
+      "step": 2202
+    },
+    {
+      "epoch": 0.8497589199614272,
+      "grad_norm": 0.3930964058758994,
+      "learning_rate": 5.624221470168978e-07,
+      "loss": 2.3267,
+      "step": 2203
+    },
+    {
+      "epoch": 0.8501446480231437,
+      "grad_norm": 0.42756363784629275,
+      "learning_rate": 5.595893504614097e-07,
+      "loss": 2.2043,
+      "step": 2204
+    },
+    {
+      "epoch": 0.8505303760848602,
+      "grad_norm": 0.43391519530192535,
+      "learning_rate": 5.567632830757025e-07,
+      "loss": 2.2901,
+      "step": 2205
+    },
+    {
+      "epoch": 0.8509161041465767,
+      "grad_norm": 0.4307181250385132,
+      "learning_rate": 5.539439491425097e-07,
+      "loss": 2.3412,
+      "step": 2206
+    },
+    {
+      "epoch": 0.8513018322082931,
+      "grad_norm": 0.4112928115331509,
+      "learning_rate": 5.511313529343581e-07,
+      "loss": 2.2968,
+      "step": 2207
+    },
+    {
+      "epoch": 0.8516875602700097,
+      "grad_norm": 0.4083314717986259,
+      "learning_rate": 5.483254987135644e-07,
+      "loss": 2.3146,
+      "step": 2208
+    },
+    {
+      "epoch": 0.8520732883317261,
+      "grad_norm": 0.42163950916500303,
+      "learning_rate": 5.455263907322283e-07,
+      "loss": 2.3131,
+      "step": 2209
+    },
+    {
+      "epoch": 0.8524590163934426,
+      "grad_norm": 0.43750119275808214,
+      "learning_rate": 5.427340332322267e-07,
+      "loss": 2.3545,
+      "step": 2210
+    },
+    {
+      "epoch": 0.8528447444551591,
+      "grad_norm": 0.4227491273975129,
+      "learning_rate": 5.39948430445204e-07,
+      "loss": 2.3194,
+      "step": 2211
+    },
+    {
+      "epoch": 0.8532304725168756,
+      "grad_norm": 0.47559103478413123,
+      "learning_rate": 5.371695865925736e-07,
+      "loss": 2.3153,
+      "step": 2212
+    },
+    {
+      "epoch": 0.8536162005785921,
+      "grad_norm": 0.45948833709331677,
+      "learning_rate": 5.343975058854994e-07,
+      "loss": 2.3079,
+      "step": 2213
+    },
+    {
+      "epoch": 0.8540019286403085,
+      "grad_norm": 0.4213448822063752,
+      "learning_rate": 5.316321925249024e-07,
+      "loss": 2.1981,
+      "step": 2214
+    },
+    {
+      "epoch": 0.8543876567020251,
+      "grad_norm": 0.42214244984468713,
+      "learning_rate": 5.288736507014436e-07,
+      "loss": 2.3274,
+      "step": 2215
+    },
+    {
+      "epoch": 0.8547733847637415,
+      "grad_norm": 0.41575191369539755,
+      "learning_rate": 5.261218845955246e-07,
+      "loss": 2.3627,
+      "step": 2216
+    },
+    {
+      "epoch": 0.8551591128254581,
+      "grad_norm": 0.45330059585872606,
+      "learning_rate": 5.23376898377278e-07,
+      "loss": 2.3306,
+      "step": 2217
+    },
+    {
+      "epoch": 0.8555448408871745,
+      "grad_norm": 0.4033291703126306,
+      "learning_rate": 5.206386962065601e-07,
+      "loss": 2.4107,
+      "step": 2218
+    },
+    {
+      "epoch": 0.8559305689488911,
+      "grad_norm": 0.4298766207997605,
+      "learning_rate": 5.179072822329512e-07,
+      "loss": 2.3519,
+      "step": 2219
+    },
+    {
+      "epoch": 0.8563162970106075,
+      "grad_norm": 0.45446292904760033,
+      "learning_rate": 5.151826605957394e-07,
+      "loss": 2.3882,
+      "step": 2220
+    },
+    {
+      "epoch": 0.856702025072324,
+      "grad_norm": 0.43566691330033935,
+      "learning_rate": 5.124648354239225e-07,
+      "loss": 2.3187,
+      "step": 2221
+    },
+    {
+      "epoch": 0.8570877531340405,
+      "grad_norm": 0.4330203954906256,
+      "learning_rate": 5.097538108361966e-07,
+      "loss": 2.3087,
+      "step": 2222
+    },
+    {
+      "epoch": 0.857473481195757,
+      "grad_norm": 0.41939583997928825,
+      "learning_rate": 5.070495909409551e-07,
+      "loss": 2.2939,
+      "step": 2223
+    },
+    {
+      "epoch": 0.8578592092574735,
+      "grad_norm": 0.4736228920931165,
+      "learning_rate": 5.043521798362755e-07,
+      "loss": 2.2919,
+      "step": 2224
+    },
+    {
+      "epoch": 0.8582449373191899,
+      "grad_norm": 0.43631394319711947,
+      "learning_rate": 5.016615816099185e-07,
+      "loss": 2.2537,
+      "step": 2225
+    },
+    {
+      "epoch": 0.8586306653809065,
+      "grad_norm": 0.40377370202039525,
+      "learning_rate": 4.98977800339322e-07,
+      "loss": 2.3361,
+      "step": 2226
+    },
+    {
+      "epoch": 0.8590163934426229,
+      "grad_norm": 0.42579008936762563,
+      "learning_rate": 4.963008400915914e-07,
+      "loss": 2.2608,
+      "step": 2227
+    },
+    {
+      "epoch": 0.8594021215043395,
+      "grad_norm": 0.4133094110710469,
+      "learning_rate": 4.936307049234956e-07,
+      "loss": 2.291,
+      "step": 2228
+    },
+    {
+      "epoch": 0.8597878495660559,
+      "grad_norm": 0.40409474331202133,
+      "learning_rate": 4.9096739888146e-07,
+      "loss": 2.3848,
+      "step": 2229
+    },
+    {
+      "epoch": 0.8601735776277725,
+      "grad_norm": 0.4167750604678949,
+      "learning_rate": 4.883109260015617e-07,
+      "loss": 2.3461,
+      "step": 2230
+    },
+    {
+      "epoch": 0.8605593056894889,
+      "grad_norm": 0.4138540442699837,
+      "learning_rate": 4.85661290309522e-07,
+      "loss": 2.2979,
+      "step": 2231
+    },
+    {
+      "epoch": 0.8609450337512053,
+      "grad_norm": 0.4011426011687145,
+      "learning_rate": 4.830184958207007e-07,
+      "loss": 2.3469,
+      "step": 2232
+    },
+    {
+      "epoch": 0.8613307618129219,
+      "grad_norm": 0.47545862045215365,
+      "learning_rate": 4.80382546540093e-07,
+      "loss": 2.3112,
+      "step": 2233
+    },
+    {
+      "epoch": 0.8617164898746383,
+      "grad_norm": 0.41395417512831273,
+      "learning_rate": 4.777534464623162e-07,
+      "loss": 2.3335,
+      "step": 2234
+    },
+    {
+      "epoch": 0.8621022179363549,
+      "grad_norm": 0.43386901834187463,
+      "learning_rate": 4.7513119957161124e-07,
+      "loss": 2.3024,
+      "step": 2235
+    },
+    {
+      "epoch": 0.8624879459980713,
+      "grad_norm": 0.44148333849080224,
+      "learning_rate": 4.725158098418309e-07,
+      "loss": 2.3392,
+      "step": 2236
+    },
+    {
+      "epoch": 0.8628736740597879,
+      "grad_norm": 0.40902133683113945,
+      "learning_rate": 4.69907281236438e-07,
+      "loss": 2.3232,
+      "step": 2237
+    },
+    {
+      "epoch": 0.8632594021215043,
+      "grad_norm": 0.4303489752981223,
+      "learning_rate": 4.673056177084989e-07,
+      "loss": 2.3198,
+      "step": 2238
+    },
+    {
+      "epoch": 0.8636451301832209,
+      "grad_norm": 0.4156111488028808,
+      "learning_rate": 4.647108232006742e-07,
+      "loss": 2.3342,
+      "step": 2239
+    },
+    {
+      "epoch": 0.8640308582449373,
+      "grad_norm": 0.39961453276398634,
+      "learning_rate": 4.6212290164521554e-07,
+      "loss": 2.3079,
+      "step": 2240
+    },
+    {
+      "epoch": 0.8644165863066539,
+      "grad_norm": 0.4141830467341266,
+      "learning_rate": 4.595418569639581e-07,
+      "loss": 2.2747,
+      "step": 2241
+    },
+    {
+      "epoch": 0.8648023143683703,
+      "grad_norm": 0.42352865639027676,
+      "learning_rate": 4.5696769306831923e-07,
+      "loss": 2.2992,
+      "step": 2242
+    },
+    {
+      "epoch": 0.8651880424300867,
+      "grad_norm": 0.39859439958207843,
+      "learning_rate": 4.5440041385928444e-07,
+      "loss": 2.3091,
+      "step": 2243
+    },
+    {
+      "epoch": 0.8655737704918033,
+      "grad_norm": 0.4484657124226204,
+      "learning_rate": 4.5184002322740784e-07,
+      "loss": 2.2968,
+      "step": 2244
+    },
+    {
+      "epoch": 0.8659594985535197,
+      "grad_norm": 0.44190808883333643,
+      "learning_rate": 4.492865250528056e-07,
+      "loss": 2.3755,
+      "step": 2245
+    },
+    {
+      "epoch": 0.8663452266152363,
+      "grad_norm": 0.5020943728416454,
+      "learning_rate": 4.4673992320514617e-07,
+      "loss": 2.3072,
+      "step": 2246
+    },
+    {
+      "epoch": 0.8667309546769527,
+      "grad_norm": 0.4294192462710892,
+      "learning_rate": 4.4420022154364917e-07,
+      "loss": 2.3103,
+      "step": 2247
+    },
+    {
+      "epoch": 0.8671166827386693,
+      "grad_norm": 0.4203332570620652,
+      "learning_rate": 4.4166742391707593e-07,
+      "loss": 2.3168,
+      "step": 2248
+    },
+    {
+      "epoch": 0.8675024108003857,
+      "grad_norm": 0.4479748088729882,
+      "learning_rate": 4.391415341637262e-07,
+      "loss": 2.323,
+      "step": 2249
+    },
+    {
+      "epoch": 0.8678881388621023,
+      "grad_norm": 0.3912067407465435,
+      "learning_rate": 4.366225561114296e-07,
+      "loss": 2.3569,
+      "step": 2250
+    },
+    {
+      "epoch": 0.8682738669238187,
+      "grad_norm": 0.38938757566727544,
+      "learning_rate": 4.341104935775442e-07,
+      "loss": 2.322,
+      "step": 2251
+    },
+    {
+      "epoch": 0.8686595949855352,
+      "grad_norm": 0.4240750439778621,
+      "learning_rate": 4.316053503689466e-07,
+      "loss": 2.2382,
+      "step": 2252
+    },
+    {
+      "epoch": 0.8690453230472517,
+      "grad_norm": 0.4023975911484655,
+      "learning_rate": 4.291071302820271e-07,
+      "loss": 2.3566,
+      "step": 2253
+    },
+    {
+      "epoch": 0.8694310511089681,
+      "grad_norm": 0.42667166936274664,
+      "learning_rate": 4.2661583710268573e-07,
+      "loss": 2.2845,
+      "step": 2254
+    },
+    {
+      "epoch": 0.8698167791706847,
+      "grad_norm": 0.4160071160989676,
+      "learning_rate": 4.24131474606323e-07,
+      "loss": 2.327,
+      "step": 2255
+    },
+    {
+      "epoch": 0.8702025072324011,
+      "grad_norm": 0.4209819473631007,
+      "learning_rate": 4.2165404655783836e-07,
+      "loss": 2.3523,
+      "step": 2256
+    },
+    {
+      "epoch": 0.8705882352941177,
+      "grad_norm": 0.4440843418215145,
+      "learning_rate": 4.1918355671162145e-07,
+      "loss": 2.3495,
+      "step": 2257
+    },
+    {
+      "epoch": 0.8709739633558341,
+      "grad_norm": 0.41164003016046363,
+      "learning_rate": 4.1672000881154917e-07,
+      "loss": 2.3301,
+      "step": 2258
+    },
+    {
+      "epoch": 0.8713596914175507,
+      "grad_norm": 0.4469948251850552,
+      "learning_rate": 4.1426340659097565e-07,
+      "loss": 2.3737,
+      "step": 2259
+    },
+    {
+      "epoch": 0.8717454194792671,
+      "grad_norm": 0.3987525236231152,
+      "learning_rate": 4.1181375377273237e-07,
+      "loss": 2.3774,
+      "step": 2260
+    },
+    {
+      "epoch": 0.8721311475409836,
+      "grad_norm": 0.43640589294749244,
+      "learning_rate": 4.09371054069117e-07,
+      "loss": 2.2801,
+      "step": 2261
+    },
+    {
+      "epoch": 0.8725168756027001,
+      "grad_norm": 0.4329967109483763,
+      "learning_rate": 4.069353111818913e-07,
+      "loss": 2.3759,
+      "step": 2262
+    },
+    {
+      "epoch": 0.8729026036644166,
+      "grad_norm": 0.4204653148417705,
+      "learning_rate": 4.0450652880227426e-07,
+      "loss": 2.2806,
+      "step": 2263
+    },
+    {
+      "epoch": 0.8732883317261331,
+      "grad_norm": 0.46488748061246604,
+      "learning_rate": 4.020847106109349e-07,
+      "loss": 2.2773,
+      "step": 2264
+    },
+    {
+      "epoch": 0.8736740597878495,
+      "grad_norm": 0.4126593478938596,
+      "learning_rate": 3.996698602779919e-07,
+      "loss": 2.4213,
+      "step": 2265
+    },
+    {
+      "epoch": 0.8740597878495661,
+      "grad_norm": 0.41842747786828716,
+      "learning_rate": 3.9726198146300185e-07,
+      "loss": 2.3888,
+      "step": 2266
+    },
+    {
+      "epoch": 0.8744455159112825,
+      "grad_norm": 0.40822133226916885,
+      "learning_rate": 3.948610778149581e-07,
+      "loss": 2.3844,
+      "step": 2267
+    },
+    {
+      "epoch": 0.874831243972999,
+      "grad_norm": 0.4237314541249901,
+      "learning_rate": 3.9246715297228176e-07,
+      "loss": 2.306,
+      "step": 2268
+    },
+    {
+      "epoch": 0.8752169720347155,
+      "grad_norm": 0.42912593943031246,
+      "learning_rate": 3.9008021056281875e-07,
+      "loss": 2.295,
+      "step": 2269
+    },
+    {
+      "epoch": 0.875602700096432,
+      "grad_norm": 0.4105781495982416,
+      "learning_rate": 3.877002542038355e-07,
+      "loss": 2.2801,
+      "step": 2270
+    },
+    {
+      "epoch": 0.8759884281581485,
+      "grad_norm": 0.39721896942295326,
+      "learning_rate": 3.8532728750200755e-07,
+      "loss": 2.3438,
+      "step": 2271
+    },
+    {
+      "epoch": 0.876374156219865,
+      "grad_norm": 0.41985251769629356,
+      "learning_rate": 3.829613140534222e-07,
+      "loss": 2.3071,
+      "step": 2272
+    },
+    {
+      "epoch": 0.8767598842815815,
+      "grad_norm": 0.44265077565168853,
+      "learning_rate": 3.8060233744356634e-07,
+      "loss": 2.2983,
+      "step": 2273
+    },
+    {
+      "epoch": 0.877145612343298,
+      "grad_norm": 0.418111678423349,
+      "learning_rate": 3.782503612473243e-07,
+      "loss": 2.3021,
+      "step": 2274
+    },
+    {
+      "epoch": 0.8775313404050145,
+      "grad_norm": 0.4250751660834272,
+      "learning_rate": 3.75905389028971e-07,
+      "loss": 2.3847,
+      "step": 2275
+    },
+    {
+      "epoch": 0.877917068466731,
+      "grad_norm": 0.41628911378730005,
+      "learning_rate": 3.7356742434216775e-07,
+      "loss": 2.2959,
+      "step": 2276
+    },
+    {
+      "epoch": 0.8783027965284474,
+      "grad_norm": 0.4187293548369921,
+      "learning_rate": 3.7123647072995816e-07,
+      "loss": 2.3418,
+      "step": 2277
+    },
+    {
+      "epoch": 0.8786885245901639,
+      "grad_norm": 0.43655945789184647,
+      "learning_rate": 3.689125317247572e-07,
+      "loss": 2.3148,
+      "step": 2278
+    },
+    {
+      "epoch": 0.8790742526518804,
+      "grad_norm": 0.43266462241172693,
+      "learning_rate": 3.665956108483543e-07,
+      "loss": 2.2817,
+      "step": 2279
+    },
+    {
+      "epoch": 0.8794599807135969,
+      "grad_norm": 0.42782770811801585,
+      "learning_rate": 3.642857116118986e-07,
+      "loss": 2.3549,
+      "step": 2280
+    },
+    {
+      "epoch": 0.8798457087753134,
+      "grad_norm": 0.4321683221918526,
+      "learning_rate": 3.619828375159018e-07,
+      "loss": 2.3334,
+      "step": 2281
+    },
+    {
+      "epoch": 0.8802314368370299,
+      "grad_norm": 0.4396171200268804,
+      "learning_rate": 3.5968699205022827e-07,
+      "loss": 2.2895,
+      "step": 2282
+    },
+    {
+      "epoch": 0.8806171648987464,
+      "grad_norm": 0.4467943875881692,
+      "learning_rate": 3.573981786940889e-07,
+      "loss": 2.2695,
+      "step": 2283
+    },
+    {
+      "epoch": 0.8810028929604629,
+      "grad_norm": 0.47767710110174755,
+      "learning_rate": 3.5511640091604293e-07,
+      "loss": 2.3356,
+      "step": 2284
+    },
+    {
+      "epoch": 0.8813886210221794,
+      "grad_norm": 0.41802680830792355,
+      "learning_rate": 3.5284166217398276e-07,
+      "loss": 2.279,
+      "step": 2285
+    },
+    {
+      "epoch": 0.8817743490838958,
+      "grad_norm": 0.438723155619965,
+      "learning_rate": 3.505739659151358e-07,
+      "loss": 2.2767,
+      "step": 2286
+    },
+    {
+      "epoch": 0.8821600771456124,
+      "grad_norm": 0.45784513304782515,
+      "learning_rate": 3.4831331557605597e-07,
+      "loss": 2.3033,
+      "step": 2287
+    },
+    {
+      "epoch": 0.8825458052073288,
+      "grad_norm": 0.40870966872961656,
+      "learning_rate": 3.4605971458262e-07,
+      "loss": 2.3508,
+      "step": 2288
+    },
+    {
+      "epoch": 0.8829315332690453,
+      "grad_norm": 0.4424833444084585,
+      "learning_rate": 3.4381316635002324e-07,
+      "loss": 2.3019,
+      "step": 2289
+    },
+    {
+      "epoch": 0.8833172613307618,
+      "grad_norm": 0.4292800347530429,
+      "learning_rate": 3.4157367428276966e-07,
+      "loss": 2.2879,
+      "step": 2290
+    },
+    {
+      "epoch": 0.8837029893924783,
+      "grad_norm": 0.4461628943725614,
+      "learning_rate": 3.3934124177467386e-07,
+      "loss": 2.311,
+      "step": 2291
+    },
+    {
+      "epoch": 0.8840887174541948,
+      "grad_norm": 0.41792428163644263,
+      "learning_rate": 3.371158722088497e-07,
+      "loss": 2.3708,
+      "step": 2292
+    },
+    {
+      "epoch": 0.8844744455159113,
+      "grad_norm": 0.45556965914486197,
+      "learning_rate": 3.3489756895770773e-07,
+      "loss": 2.3435,
+      "step": 2293
+    },
+    {
+      "epoch": 0.8848601735776278,
+      "grad_norm": 0.42326123853106545,
+      "learning_rate": 3.326863353829507e-07,
+      "loss": 2.2452,
+      "step": 2294
+    },
+    {
+      "epoch": 0.8852459016393442,
+      "grad_norm": 0.43115329010814496,
+      "learning_rate": 3.3048217483556743e-07,
+      "loss": 2.3017,
+      "step": 2295
+    },
+    {
+      "epoch": 0.8856316297010608,
+      "grad_norm": 0.4017502606904032,
+      "learning_rate": 3.2828509065582713e-07,
+      "loss": 2.2709,
+      "step": 2296
+    },
+    {
+      "epoch": 0.8860173577627772,
+      "grad_norm": 0.43269418962686174,
+      "learning_rate": 3.260950861732765e-07,
+      "loss": 2.2945,
+      "step": 2297
+    },
+    {
+      "epoch": 0.8864030858244938,
+      "grad_norm": 0.4442870353620615,
+      "learning_rate": 3.239121647067339e-07,
+      "loss": 2.3491,
+      "step": 2298
+    },
+    {
+      "epoch": 0.8867888138862102,
+      "grad_norm": 0.41736126954832276,
+      "learning_rate": 3.217363295642817e-07,
+      "loss": 2.3845,
+      "step": 2299
+    },
+    {
+      "epoch": 0.8871745419479267,
+      "grad_norm": 0.40746274890345213,
+      "learning_rate": 3.195675840432655e-07,
+      "loss": 2.3505,
+      "step": 2300
+    },
+    {
+      "epoch": 0.8875602700096432,
+      "grad_norm": 0.43921816246720496,
+      "learning_rate": 3.1740593143028463e-07,
+      "loss": 2.3695,
+      "step": 2301
+    },
+    {
+      "epoch": 0.8879459980713597,
+      "grad_norm": 0.4417907633038376,
+      "learning_rate": 3.1525137500119207e-07,
+      "loss": 2.3178,
+      "step": 2302
+    },
+    {
+      "epoch": 0.8883317261330762,
+      "grad_norm": 0.40236247950052306,
+      "learning_rate": 3.1310391802108433e-07,
+      "loss": 2.3592,
+      "step": 2303
+    },
+    {
+      "epoch": 0.8887174541947926,
+      "grad_norm": 0.4389765459305632,
+      "learning_rate": 3.109635637443026e-07,
+      "loss": 2.2578,
+      "step": 2304
+    },
+    {
+      "epoch": 0.8891031822565092,
+      "grad_norm": 0.44248147502648144,
+      "learning_rate": 3.0883031541442175e-07,
+      "loss": 2.2892,
+      "step": 2305
+    },
+    {
+      "epoch": 0.8894889103182256,
+      "grad_norm": 0.4212530551412743,
+      "learning_rate": 3.067041762642475e-07,
+      "loss": 2.275,
+      "step": 2306
+    },
+    {
+      "epoch": 0.8898746383799422,
+      "grad_norm": 0.39390096118261103,
+      "learning_rate": 3.045851495158147e-07,
+      "loss": 2.3076,
+      "step": 2307
+    },
+    {
+      "epoch": 0.8902603664416586,
+      "grad_norm": 0.44065450958834507,
+      "learning_rate": 3.02473238380378e-07,
+      "loss": 2.3761,
+      "step": 2308
+    },
+    {
+      "epoch": 0.8906460945033752,
+      "grad_norm": 0.39869216936050716,
+      "learning_rate": 3.0036844605840944e-07,
+      "loss": 2.3692,
+      "step": 2309
+    },
+    {
+      "epoch": 0.8910318225650916,
+      "grad_norm": 0.47639377826582363,
+      "learning_rate": 2.9827077573959083e-07,
+      "loss": 2.2737,
+      "step": 2310
+    },
+    {
+      "epoch": 0.891417550626808,
+      "grad_norm": 0.3949086111875954,
+      "learning_rate": 2.9618023060281443e-07,
+      "loss": 2.3125,
+      "step": 2311
+    },
+    {
+      "epoch": 0.8918032786885246,
+      "grad_norm": 0.43709075645674,
+      "learning_rate": 2.9409681381617315e-07,
+      "loss": 2.3073,
+      "step": 2312
+    },
+    {
+      "epoch": 0.892189006750241,
+      "grad_norm": 0.44877858308367297,
+      "learning_rate": 2.920205285369565e-07,
+      "loss": 2.3552,
+      "step": 2313
+    },
+    {
+      "epoch": 0.8925747348119576,
+      "grad_norm": 0.448956433956615,
+      "learning_rate": 2.899513779116475e-07,
+      "loss": 2.3559,
+      "step": 2314
+    },
+    {
+      "epoch": 0.892960462873674,
+      "grad_norm": 0.4040751591496477,
+      "learning_rate": 2.878893650759168e-07,
+      "loss": 2.3402,
+      "step": 2315
+    },
+    {
+      "epoch": 0.8933461909353906,
+      "grad_norm": 0.42941973754766005,
+      "learning_rate": 2.858344931546181e-07,
+      "loss": 2.2931,
+      "step": 2316
+    },
+    {
+      "epoch": 0.893731918997107,
+      "grad_norm": 0.4093804426084204,
+      "learning_rate": 2.8378676526178484e-07,
+      "loss": 2.281,
+      "step": 2317
+    },
+    {
+      "epoch": 0.8941176470588236,
+      "grad_norm": 0.4056051720053737,
+      "learning_rate": 2.8174618450062254e-07,
+      "loss": 2.2842,
+      "step": 2318
+    },
+    {
+      "epoch": 0.89450337512054,
+      "grad_norm": 0.44303445355744625,
+      "learning_rate": 2.7971275396350526e-07,
+      "loss": 2.2731,
+      "step": 2319
+    },
+    {
+      "epoch": 0.8948891031822566,
+      "grad_norm": 0.4262011309842942,
+      "learning_rate": 2.776864767319731e-07,
+      "loss": 2.3174,
+      "step": 2320
+    },
+    {
+      "epoch": 0.895274831243973,
+      "grad_norm": 0.4343001065859361,
+      "learning_rate": 2.756673558767242e-07,
+      "loss": 2.2447,
+      "step": 2321
+    },
+    {
+      "epoch": 0.8956605593056894,
+      "grad_norm": 0.4042191076748588,
+      "learning_rate": 2.7365539445761204e-07,
+      "loss": 2.3615,
+      "step": 2322
+    },
+    {
+      "epoch": 0.896046287367406,
+      "grad_norm": 0.4324732347462319,
+      "learning_rate": 2.716505955236415e-07,
+      "loss": 2.3086,
+      "step": 2323
+    },
+    {
+      "epoch": 0.8964320154291224,
+      "grad_norm": 0.44877695307282456,
+      "learning_rate": 2.696529621129618e-07,
+      "loss": 2.3156,
+      "step": 2324
+    },
+    {
+      "epoch": 0.896817743490839,
+      "grad_norm": 0.4130849621531806,
+      "learning_rate": 2.6766249725286295e-07,
+      "loss": 2.3451,
+      "step": 2325
+    },
+    {
+      "epoch": 0.8972034715525554,
+      "grad_norm": 0.4215410840108579,
+      "learning_rate": 2.656792039597744e-07,
+      "loss": 2.3394,
+      "step": 2326
+    },
+    {
+      "epoch": 0.897589199614272,
+      "grad_norm": 0.4267289938196295,
+      "learning_rate": 2.637030852392536e-07,
+      "loss": 2.3131,
+      "step": 2327
+    },
+    {
+      "epoch": 0.8979749276759884,
+      "grad_norm": 0.43553479832965797,
+      "learning_rate": 2.617341440859883e-07,
+      "loss": 2.3712,
+      "step": 2328
+    },
+    {
+      "epoch": 0.898360655737705,
+      "grad_norm": 0.40774833683590433,
+      "learning_rate": 2.597723834837862e-07,
+      "loss": 2.4135,
+      "step": 2329
+    },
+    {
+      "epoch": 0.8987463837994214,
+      "grad_norm": 0.41728549772051965,
+      "learning_rate": 2.5781780640557753e-07,
+      "loss": 2.3074,
+      "step": 2330
+    },
+    {
+      "epoch": 0.899132111861138,
+      "grad_norm": 0.4429978263158119,
+      "learning_rate": 2.5587041581340235e-07,
+      "loss": 2.3504,
+      "step": 2331
+    },
+    {
+      "epoch": 0.8995178399228544,
+      "grad_norm": 0.3935302876461795,
+      "learning_rate": 2.539302146584116e-07,
+      "loss": 2.343,
+      "step": 2332
+    },
+    {
+      "epoch": 0.8999035679845708,
+      "grad_norm": 0.39486222802808546,
+      "learning_rate": 2.5199720588086117e-07,
+      "loss": 2.3279,
+      "step": 2333
+    },
+    {
+      "epoch": 0.9002892960462874,
+      "grad_norm": 0.39773603659851414,
+      "learning_rate": 2.5007139241010724e-07,
+      "loss": 2.2666,
+      "step": 2334
+    },
+    {
+      "epoch": 0.9006750241080038,
+      "grad_norm": 0.4314496361740582,
+      "learning_rate": 2.48152777164602e-07,
+      "loss": 2.3723,
+      "step": 2335
+    },
+    {
+      "epoch": 0.9010607521697204,
+      "grad_norm": 0.39779368566168694,
+      "learning_rate": 2.4624136305188895e-07,
+      "loss": 2.3041,
+      "step": 2336
+    },
+    {
+      "epoch": 0.9014464802314368,
+      "grad_norm": 0.43539426008635035,
+      "learning_rate": 2.443371529685995e-07,
+      "loss": 2.2757,
+      "step": 2337
+    },
+    {
+      "epoch": 0.9018322082931534,
+      "grad_norm": 0.41172077652364375,
+      "learning_rate": 2.424401498004464e-07,
+      "loss": 2.34,
+      "step": 2338
+    },
+    {
+      "epoch": 0.9022179363548698,
+      "grad_norm": 0.4311143648496032,
+      "learning_rate": 2.4055035642222225e-07,
+      "loss": 2.292,
+      "step": 2339
+    },
+    {
+      "epoch": 0.9026036644165863,
+      "grad_norm": 0.43199604546203846,
+      "learning_rate": 2.3866777569779234e-07,
+      "loss": 2.2811,
+      "step": 2340
+    },
+    {
+      "epoch": 0.9029893924783028,
+      "grad_norm": 0.3996744267613919,
+      "learning_rate": 2.367924104800917e-07,
+      "loss": 2.2995,
+      "step": 2341
+    },
+    {
+      "epoch": 0.9033751205400193,
+      "grad_norm": 0.467895433250459,
+      "learning_rate": 2.3492426361112153e-07,
+      "loss": 2.3085,
+      "step": 2342
+    },
+    {
+      "epoch": 0.9037608486017358,
+      "grad_norm": 0.42536228011701505,
+      "learning_rate": 2.3306333792194492e-07,
+      "loss": 2.3072,
+      "step": 2343
+    },
+    {
+      "epoch": 0.9041465766634522,
+      "grad_norm": 0.4360625771288912,
+      "learning_rate": 2.3120963623267822e-07,
+      "loss": 2.3197,
+      "step": 2344
+    },
+    {
+      "epoch": 0.9045323047251688,
+      "grad_norm": 0.4301196858620419,
+      "learning_rate": 2.2936316135249492e-07,
+      "loss": 2.2895,
+      "step": 2345
+    },
+    {
+      "epoch": 0.9049180327868852,
+      "grad_norm": 0.4154437321377232,
+      "learning_rate": 2.2752391607961388e-07,
+      "loss": 2.323,
+      "step": 2346
+    },
+    {
+      "epoch": 0.9053037608486018,
+      "grad_norm": 0.41515933007956946,
+      "learning_rate": 2.256919032012972e-07,
+      "loss": 2.2567,
+      "step": 2347
+    },
+    {
+      "epoch": 0.9056894889103182,
+      "grad_norm": 0.39822301614572697,
+      "learning_rate": 2.2386712549384848e-07,
+      "loss": 2.308,
+      "step": 2348
+    },
+    {
+      "epoch": 0.9060752169720347,
+      "grad_norm": 0.4183729867262524,
+      "learning_rate": 2.220495857226068e-07,
+      "loss": 2.3475,
+      "step": 2349
+    },
+    {
+      "epoch": 0.9064609450337512,
+      "grad_norm": 0.40238712744809574,
+      "learning_rate": 2.2023928664194229e-07,
+      "loss": 2.2622,
+      "step": 2350
+    },
+    {
+      "epoch": 0.9068466730954677,
+      "grad_norm": 0.3984804803845149,
+      "learning_rate": 2.1843623099525146e-07,
+      "loss": 2.2536,
+      "step": 2351
+    },
+    {
+      "epoch": 0.9072324011571842,
+      "grad_norm": 0.4396943702451173,
+      "learning_rate": 2.1664042151495424e-07,
+      "loss": 2.4127,
+      "step": 2352
+    },
+    {
+      "epoch": 0.9076181292189007,
+      "grad_norm": 0.4272040571822996,
+      "learning_rate": 2.1485186092248978e-07,
+      "loss": 2.2906,
+      "step": 2353
+    },
+    {
+      "epoch": 0.9080038572806172,
+      "grad_norm": 0.46163668035869504,
+      "learning_rate": 2.1307055192831272e-07,
+      "loss": 2.3236,
+      "step": 2354
+    },
+    {
+      "epoch": 0.9083895853423336,
+      "grad_norm": 0.41300006472431816,
+      "learning_rate": 2.112964972318865e-07,
+      "loss": 2.2955,
+      "step": 2355
+    },
+    {
+      "epoch": 0.9087753134040502,
+      "grad_norm": 0.40025540439999374,
+      "learning_rate": 2.095296995216828e-07,
+      "loss": 2.3027,
+      "step": 2356
+    },
+    {
+      "epoch": 0.9091610414657666,
+      "grad_norm": 0.4245241449879845,
+      "learning_rate": 2.0777016147517536e-07,
+      "loss": 2.3115,
+      "step": 2357
+    },
+    {
+      "epoch": 0.9095467695274831,
+      "grad_norm": 0.40282234698213815,
+      "learning_rate": 2.0601788575883518e-07,
+      "loss": 2.2349,
+      "step": 2358
+    },
+    {
+      "epoch": 0.9099324975891996,
+      "grad_norm": 0.46274752117782375,
+      "learning_rate": 2.042728750281292e-07,
+      "loss": 2.37,
+      "step": 2359
+    },
+    {
+      "epoch": 0.9103182256509161,
+      "grad_norm": 0.39501138729671303,
+      "learning_rate": 2.0253513192751374e-07,
+      "loss": 2.3541,
+      "step": 2360
+    },
+    {
+      "epoch": 0.9107039537126326,
+      "grad_norm": 0.4223437499567797,
+      "learning_rate": 2.0080465909043113e-07,
+      "loss": 2.352,
+      "step": 2361
+    },
+    {
+      "epoch": 0.9110896817743491,
+      "grad_norm": 0.4551690474658373,
+      "learning_rate": 1.990814591393081e-07,
+      "loss": 2.2622,
+      "step": 2362
+    },
+    {
+      "epoch": 0.9114754098360656,
+      "grad_norm": 0.4505583736302094,
+      "learning_rate": 1.973655346855474e-07,
+      "loss": 2.287,
+      "step": 2363
+    },
+    {
+      "epoch": 0.9118611378977821,
+      "grad_norm": 0.43738325452276416,
+      "learning_rate": 1.9565688832952846e-07,
+      "loss": 2.3309,
+      "step": 2364
+    },
+    {
+      "epoch": 0.9122468659594986,
+      "grad_norm": 0.4098843311547293,
+      "learning_rate": 1.939555226605988e-07,
+      "loss": 2.3374,
+      "step": 2365
+    },
+    {
+      "epoch": 0.912632594021215,
+      "grad_norm": 0.4324462884840351,
+      "learning_rate": 1.9226144025707382e-07,
+      "loss": 2.3175,
+      "step": 2366
+    },
+    {
+      "epoch": 0.9130183220829315,
+      "grad_norm": 0.4348134837758005,
+      "learning_rate": 1.9057464368623213e-07,
+      "loss": 2.3331,
+      "step": 2367
+    },
+    {
+      "epoch": 0.913404050144648,
+      "grad_norm": 0.40591563134425135,
+      "learning_rate": 1.8889513550430892e-07,
+      "loss": 2.2424,
+      "step": 2368
+    },
+    {
+      "epoch": 0.9137897782063645,
+      "grad_norm": 0.42326917689373966,
+      "learning_rate": 1.872229182564972e-07,
+      "loss": 2.3636,
+      "step": 2369
+    },
+    {
+      "epoch": 0.914175506268081,
+      "grad_norm": 0.4330510921583434,
+      "learning_rate": 1.855579944769387e-07,
+      "loss": 2.3307,
+      "step": 2370
+    },
+    {
+      "epoch": 0.9145612343297975,
+      "grad_norm": 0.47084888462293784,
+      "learning_rate": 1.8390036668872403e-07,
+      "loss": 2.3363,
+      "step": 2371
+    },
+    {
+      "epoch": 0.914946962391514,
+      "grad_norm": 0.42684759215732565,
+      "learning_rate": 1.8225003740388546e-07,
+      "loss": 2.3209,
+      "step": 2372
+    },
+    {
+      "epoch": 0.9153326904532305,
+      "grad_norm": 0.4514543622337268,
+      "learning_rate": 1.8060700912339635e-07,
+      "loss": 2.3375,
+      "step": 2373
+    },
+    {
+      "epoch": 0.915718418514947,
+      "grad_norm": 0.4186512484287575,
+      "learning_rate": 1.7897128433716493e-07,
+      "loss": 2.393,
+      "step": 2374
+    },
+    {
+      "epoch": 0.9161041465766635,
+      "grad_norm": 0.39577712124974357,
+      "learning_rate": 1.7734286552403114e-07,
+      "loss": 2.2997,
+      "step": 2375
+    },
+    {
+      "epoch": 0.9164898746383799,
+      "grad_norm": 0.4402007170416471,
+      "learning_rate": 1.7572175515176538e-07,
+      "loss": 2.4085,
+      "step": 2376
+    },
+    {
+      "epoch": 0.9168756027000964,
+      "grad_norm": 0.4066003985671325,
+      "learning_rate": 1.7410795567705973e-07,
+      "loss": 2.3095,
+      "step": 2377
+    },
+    {
+      "epoch": 0.9172613307618129,
+      "grad_norm": 0.4060197004309517,
+      "learning_rate": 1.725014695455285e-07,
+      "loss": 2.4112,
+      "step": 2378
+    },
+    {
+      "epoch": 0.9176470588235294,
+      "grad_norm": 0.4236957028335664,
+      "learning_rate": 1.7090229919170254e-07,
+      "loss": 2.3272,
+      "step": 2379
+    },
+    {
+      "epoch": 0.9180327868852459,
+      "grad_norm": 0.4587135748056596,
+      "learning_rate": 1.693104470390261e-07,
+      "loss": 2.3457,
+      "step": 2380
+    },
+    {
+      "epoch": 0.9184185149469624,
+      "grad_norm": 0.445833825057682,
+      "learning_rate": 1.6772591549985395e-07,
+      "loss": 2.3792,
+      "step": 2381
+    },
+    {
+      "epoch": 0.9188042430086789,
+      "grad_norm": 0.41916605257667877,
+      "learning_rate": 1.6614870697544638e-07,
+      "loss": 2.297,
+      "step": 2382
+    },
+    {
+      "epoch": 0.9191899710703954,
+      "grad_norm": 0.40163545031193365,
+      "learning_rate": 1.6457882385596647e-07,
+      "loss": 2.3092,
+      "step": 2383
+    },
+    {
+      "epoch": 0.9195756991321119,
+      "grad_norm": 0.4533615432359611,
+      "learning_rate": 1.6301626852047504e-07,
+      "loss": 2.3314,
+      "step": 2384
+    },
+    {
+      "epoch": 0.9199614271938283,
+      "grad_norm": 0.4179243811649184,
+      "learning_rate": 1.6146104333692902e-07,
+      "loss": 2.3185,
+      "step": 2385
+    },
+    {
+      "epoch": 0.9203471552555449,
+      "grad_norm": 0.41413676747662054,
+      "learning_rate": 1.599131506621765e-07,
+      "loss": 2.3773,
+      "step": 2386
+    },
+    {
+      "epoch": 0.9207328833172613,
+      "grad_norm": 0.4375738973653733,
+      "learning_rate": 1.5837259284195383e-07,
+      "loss": 2.2568,
+      "step": 2387
+    },
+    {
+      "epoch": 0.9211186113789778,
+      "grad_norm": 0.4246892931477109,
+      "learning_rate": 1.5683937221088242e-07,
+      "loss": 2.3448,
+      "step": 2388
+    },
+    {
+      "epoch": 0.9215043394406943,
+      "grad_norm": 0.4298415692395747,
+      "learning_rate": 1.5531349109246364e-07,
+      "loss": 2.3201,
+      "step": 2389
+    },
+    {
+      "epoch": 0.9218900675024108,
+      "grad_norm": 0.4422975749981783,
+      "learning_rate": 1.5379495179907666e-07,
+      "loss": 2.3471,
+      "step": 2390
+    },
+    {
+      "epoch": 0.9222757955641273,
+      "grad_norm": 0.41086563016151534,
+      "learning_rate": 1.5228375663197404e-07,
+      "loss": 2.3569,
+      "step": 2391
+    },
+    {
+      "epoch": 0.9226615236258437,
+      "grad_norm": 0.4168003402699989,
+      "learning_rate": 1.5077990788127993e-07,
+      "loss": 2.3138,
+      "step": 2392
+    },
+    {
+      "epoch": 0.9230472516875603,
+      "grad_norm": 0.43143785619098457,
+      "learning_rate": 1.4928340782598526e-07,
+      "loss": 2.3009,
+      "step": 2393
+    },
+    {
+      "epoch": 0.9234329797492767,
+      "grad_norm": 0.4182799340654419,
+      "learning_rate": 1.477942587339426e-07,
+      "loss": 2.2615,
+      "step": 2394
+    },
+    {
+      "epoch": 0.9238187078109933,
+      "grad_norm": 0.4193441615282384,
+      "learning_rate": 1.4631246286186783e-07,
+      "loss": 2.2258,
+      "step": 2395
+    },
+    {
+      "epoch": 0.9242044358727097,
+      "grad_norm": 0.4320029682004851,
+      "learning_rate": 1.448380224553303e-07,
+      "loss": 2.3066,
+      "step": 2396
+    },
+    {
+      "epoch": 0.9245901639344263,
+      "grad_norm": 0.4049944879632156,
+      "learning_rate": 1.4337093974875427e-07,
+      "loss": 2.3107,
+      "step": 2397
+    },
+    {
+      "epoch": 0.9249758919961427,
+      "grad_norm": 0.41431123640753376,
+      "learning_rate": 1.41911216965413e-07,
+      "loss": 2.3093,
+      "step": 2398
+    },
+    {
+      "epoch": 0.9253616200578592,
+      "grad_norm": 0.40260499302202707,
+      "learning_rate": 1.4045885631742807e-07,
+      "loss": 2.3547,
+      "step": 2399
+    },
+    {
+      "epoch": 0.9257473481195757,
+      "grad_norm": 0.40863834105476954,
+      "learning_rate": 1.3901386000576112e-07,
+      "loss": 2.3927,
+      "step": 2400
+    },
+    {
+      "epoch": 0.9261330761812921,
+      "grad_norm": 0.41884773263249914,
+      "learning_rate": 1.375762302202166e-07,
+      "loss": 2.2981,
+      "step": 2401
+    },
+    {
+      "epoch": 0.9265188042430087,
+      "grad_norm": 0.42266939838517653,
+      "learning_rate": 1.3614596913943457e-07,
+      "loss": 2.3489,
+      "step": 2402
+    },
+    {
+      "epoch": 0.9269045323047251,
+      "grad_norm": 0.4097322608020759,
+      "learning_rate": 1.3472307893088733e-07,
+      "loss": 2.3568,
+      "step": 2403
+    },
+    {
+      "epoch": 0.9272902603664417,
+      "grad_norm": 0.4108441704482335,
+      "learning_rate": 1.3330756175087778e-07,
+      "loss": 2.249,
+      "step": 2404
+    },
+    {
+      "epoch": 0.9276759884281581,
+      "grad_norm": 0.42904792387415436,
+      "learning_rate": 1.3189941974453502e-07,
+      "loss": 2.324,
+      "step": 2405
+    },
+    {
+      "epoch": 0.9280617164898747,
+      "grad_norm": 0.41881412719687494,
+      "learning_rate": 1.3049865504581204e-07,
+      "loss": 2.3073,
+      "step": 2406
+    },
+    {
+      "epoch": 0.9284474445515911,
+      "grad_norm": 0.4447067217596798,
+      "learning_rate": 1.2910526977748084e-07,
+      "loss": 2.361,
+      "step": 2407
+    },
+    {
+      "epoch": 0.9288331726133077,
+      "grad_norm": 0.4361601708399571,
+      "learning_rate": 1.2771926605113283e-07,
+      "loss": 2.355,
+      "step": 2408
+    },
+    {
+      "epoch": 0.9292189006750241,
+      "grad_norm": 0.411206644148289,
+      "learning_rate": 1.2634064596717122e-07,
+      "loss": 2.3693,
+      "step": 2409
+    },
+    {
+      "epoch": 0.9296046287367405,
+      "grad_norm": 0.41399257441418463,
+      "learning_rate": 1.249694116148087e-07,
+      "loss": 2.3556,
+      "step": 2410
+    },
+    {
+      "epoch": 0.9299903567984571,
+      "grad_norm": 0.42099022046753215,
+      "learning_rate": 1.2360556507206912e-07,
+      "loss": 2.3021,
+      "step": 2411
+    },
+    {
+      "epoch": 0.9303760848601735,
+      "grad_norm": 0.4407611306603906,
+      "learning_rate": 1.2224910840577642e-07,
+      "loss": 2.3072,
+      "step": 2412
+    },
+    {
+      "epoch": 0.9307618129218901,
+      "grad_norm": 0.4307686712814946,
+      "learning_rate": 1.2090004367155795e-07,
+      "loss": 2.3553,
+      "step": 2413
+    },
+    {
+      "epoch": 0.9311475409836065,
+      "grad_norm": 0.4170762789838413,
+      "learning_rate": 1.1955837291383776e-07,
+      "loss": 2.2864,
+      "step": 2414
+    },
+    {
+      "epoch": 0.9315332690453231,
+      "grad_norm": 0.4208399466609305,
+      "learning_rate": 1.1822409816583724e-07,
+      "loss": 2.3465,
+      "step": 2415
+    },
+    {
+      "epoch": 0.9319189971070395,
+      "grad_norm": 0.41854091983297737,
+      "learning_rate": 1.1689722144956672e-07,
+      "loss": 2.3208,
+      "step": 2416
+    },
+    {
+      "epoch": 0.9323047251687561,
+      "grad_norm": 0.43824070200599957,
+      "learning_rate": 1.1557774477582662e-07,
+      "loss": 2.2214,
+      "step": 2417
+    },
+    {
+      "epoch": 0.9326904532304725,
+      "grad_norm": 0.39789756046365826,
+      "learning_rate": 1.1426567014420297e-07,
+      "loss": 2.2473,
+      "step": 2418
+    },
+    {
+      "epoch": 0.933076181292189,
+      "grad_norm": 0.4116644410048528,
+      "learning_rate": 1.129609995430636e-07,
+      "loss": 2.3905,
+      "step": 2419
+    },
+    {
+      "epoch": 0.9334619093539055,
+      "grad_norm": 0.42436502034051177,
+      "learning_rate": 1.1166373494955696e-07,
+      "loss": 2.3822,
+      "step": 2420
+    },
+    {
+      "epoch": 0.9338476374156219,
+      "grad_norm": 0.39280611047179315,
+      "learning_rate": 1.1037387832960933e-07,
+      "loss": 2.3411,
+      "step": 2421
+    },
+    {
+      "epoch": 0.9342333654773385,
+      "grad_norm": 0.4510042871704501,
+      "learning_rate": 1.0909143163791769e-07,
+      "loss": 2.2484,
+      "step": 2422
+    },
+    {
+      "epoch": 0.9346190935390549,
+      "grad_norm": 0.42283535322397575,
+      "learning_rate": 1.0781639681795187e-07,
+      "loss": 2.3367,
+      "step": 2423
+    },
+    {
+      "epoch": 0.9350048216007715,
+      "grad_norm": 0.41401166297073777,
+      "learning_rate": 1.06548775801949e-07,
+      "loss": 2.335,
+      "step": 2424
+    },
+    {
+      "epoch": 0.9353905496624879,
+      "grad_norm": 0.3991091673458918,
+      "learning_rate": 1.0528857051091079e-07,
+      "loss": 2.3217,
+      "step": 2425
+    },
+    {
+      "epoch": 0.9357762777242045,
+      "grad_norm": 0.4206636984941167,
+      "learning_rate": 1.0403578285460014e-07,
+      "loss": 2.345,
+      "step": 2426
+    },
+    {
+      "epoch": 0.9361620057859209,
+      "grad_norm": 0.40537932353771305,
+      "learning_rate": 1.0279041473154117e-07,
+      "loss": 2.3187,
+      "step": 2427
+    },
+    {
+      "epoch": 0.9365477338476375,
+      "grad_norm": 0.406306268082581,
+      "learning_rate": 1.0155246802901198e-07,
+      "loss": 2.3173,
+      "step": 2428
+    },
+    {
+      "epoch": 0.9369334619093539,
+      "grad_norm": 0.4400571600499307,
+      "learning_rate": 1.0032194462304523e-07,
+      "loss": 2.338,
+      "step": 2429
+    },
+    {
+      "epoch": 0.9373191899710704,
+      "grad_norm": 0.421442152372355,
+      "learning_rate": 9.909884637842371e-08,
+      "loss": 2.3363,
+      "step": 2430
+    },
+    {
+      "epoch": 0.9377049180327869,
+      "grad_norm": 0.4456464895379789,
+      "learning_rate": 9.788317514867751e-08,
+      "loss": 2.4074,
+      "step": 2431
+    },
+    {
+      "epoch": 0.9380906460945033,
+      "grad_norm": 0.44272456024096185,
+      "learning_rate": 9.667493277608187e-08,
+      "loss": 2.3709,
+      "step": 2432
+    },
+    {
+      "epoch": 0.9384763741562199,
+      "grad_norm": 0.4287548757129891,
+      "learning_rate": 9.547412109165321e-08,
+      "loss": 2.3216,
+      "step": 2433
+    },
+    {
+      "epoch": 0.9388621022179363,
+      "grad_norm": 0.4120997550840361,
+      "learning_rate": 9.428074191514924e-08,
+      "loss": 2.2839,
+      "step": 2434
+    },
+    {
+      "epoch": 0.9392478302796529,
+      "grad_norm": 0.4163304656784519,
+      "learning_rate": 9.309479705506219e-08,
+      "loss": 2.3909,
+      "step": 2435
+    },
+    {
+      "epoch": 0.9396335583413693,
+      "grad_norm": 0.39798785466770226,
+      "learning_rate": 9.191628830861832e-08,
+      "loss": 2.347,
+      "step": 2436
+    },
+    {
+      "epoch": 0.9400192864030859,
+      "grad_norm": 0.43253527714333934,
+      "learning_rate": 9.074521746177567e-08,
+      "loss": 2.3688,
+      "step": 2437
+    },
+    {
+      "epoch": 0.9404050144648023,
+      "grad_norm": 0.4532266241093941,
+      "learning_rate": 8.95815862892202e-08,
+      "loss": 2.3301,
+      "step": 2438
+    },
+    {
+      "epoch": 0.9407907425265188,
+      "grad_norm": 0.4439214571057158,
+      "learning_rate": 8.842539655436355e-08,
+      "loss": 2.3264,
+      "step": 2439
+    },
+    {
+      "epoch": 0.9411764705882353,
+      "grad_norm": 0.43604973208520453,
+      "learning_rate": 8.727665000934027e-08,
+      "loss": 2.3637,
+      "step": 2440
+    },
+    {
+      "epoch": 0.9415621986499518,
+      "grad_norm": 0.4153258000523079,
+      "learning_rate": 8.61353483950056e-08,
+      "loss": 2.4173,
+      "step": 2441
+    },
+    {
+      "epoch": 0.9419479267116683,
+      "grad_norm": 0.46124634253916963,
+      "learning_rate": 8.500149344093156e-08,
+      "loss": 2.2817,
+      "step": 2442
+    },
+    {
+      "epoch": 0.9423336547733847,
+      "grad_norm": 0.4402039690213307,
+      "learning_rate": 8.387508686540591e-08,
+      "loss": 2.3774,
+      "step": 2443
+    },
+    {
+      "epoch": 0.9427193828351013,
+      "grad_norm": 0.4658228652575852,
+      "learning_rate": 8.275613037542873e-08,
+      "loss": 2.3657,
+      "step": 2444
+    },
+    {
+      "epoch": 0.9431051108968177,
+      "grad_norm": 0.41699101927893295,
+      "learning_rate": 8.164462566670972e-08,
+      "loss": 2.2583,
+      "step": 2445
+    },
+    {
+      "epoch": 0.9434908389585343,
+      "grad_norm": 0.4962189996072224,
+      "learning_rate": 8.054057442366592e-08,
+      "loss": 2.3248,
+      "step": 2446
+    },
+    {
+      "epoch": 0.9438765670202507,
+      "grad_norm": 0.41155828648085263,
+      "learning_rate": 7.944397831941952e-08,
+      "loss": 2.292,
+      "step": 2447
+    },
+    {
+      "epoch": 0.9442622950819672,
+      "grad_norm": 0.42135807540720943,
+      "learning_rate": 7.835483901579454e-08,
+      "loss": 2.3418,
+      "step": 2448
+    },
+    {
+      "epoch": 0.9446480231436837,
+      "grad_norm": 0.41326626703751895,
+      "learning_rate": 7.727315816331515e-08,
+      "loss": 2.3092,
+      "step": 2449
+    },
+    {
+      "epoch": 0.9450337512054002,
+      "grad_norm": 0.42688792733304276,
+      "learning_rate": 7.619893740120176e-08,
+      "loss": 2.3105,
+      "step": 2450
+    },
+    {
+      "epoch": 0.9454194792671167,
+      "grad_norm": 0.4504376421920826,
+      "learning_rate": 7.513217835737052e-08,
+      "loss": 2.3061,
+      "step": 2451
+    },
+    {
+      "epoch": 0.9458052073288332,
+      "grad_norm": 0.42387316716416107,
+      "learning_rate": 7.407288264842772e-08,
+      "loss": 2.3136,
+      "step": 2452
+    },
+    {
+      "epoch": 0.9461909353905497,
+      "grad_norm": 0.42731506298397026,
+      "learning_rate": 7.302105187967313e-08,
+      "loss": 2.3327,
+      "step": 2453
+    },
+    {
+      "epoch": 0.9465766634522661,
+      "grad_norm": 0.3897487344371067,
+      "learning_rate": 7.197668764509058e-08,
+      "loss": 2.3351,
+      "step": 2454
+    },
+    {
+      "epoch": 0.9469623915139826,
+      "grad_norm": 0.40963650978286614,
+      "learning_rate": 7.09397915273502e-08,
+      "loss": 2.3007,
+      "step": 2455
+    },
+    {
+      "epoch": 0.9473481195756991,
+      "grad_norm": 0.40256147687936505,
+      "learning_rate": 6.991036509780391e-08,
+      "loss": 2.3398,
+      "step": 2456
+    },
+    {
+      "epoch": 0.9477338476374156,
+      "grad_norm": 0.4131690467739318,
+      "learning_rate": 6.888840991648493e-08,
+      "loss": 2.3185,
+      "step": 2457
+    },
+    {
+      "epoch": 0.9481195756991321,
+      "grad_norm": 0.41776805663834055,
+      "learning_rate": 6.787392753210386e-08,
+      "loss": 2.3094,
+      "step": 2458
+    },
+    {
+      "epoch": 0.9485053037608486,
+      "grad_norm": 0.41453002864252014,
+      "learning_rate": 6.686691948204537e-08,
+      "loss": 2.2026,
+      "step": 2459
+    },
+    {
+      "epoch": 0.9488910318225651,
+      "grad_norm": 0.43591139239000615,
+      "learning_rate": 6.58673872923693e-08,
+      "loss": 2.3234,
+      "step": 2460
+    },
+    {
+      "epoch": 0.9492767598842816,
+      "grad_norm": 0.400008500109623,
+      "learning_rate": 6.487533247780508e-08,
+      "loss": 2.3007,
+      "step": 2461
+    },
+    {
+      "epoch": 0.9496624879459981,
+      "grad_norm": 0.4025190134433101,
+      "learning_rate": 6.38907565417507e-08,
+      "loss": 2.3597,
+      "step": 2462
+    },
+    {
+      "epoch": 0.9500482160077146,
+      "grad_norm": 0.40652255242436164,
+      "learning_rate": 6.291366097627095e-08,
+      "loss": 2.3401,
+      "step": 2463
+    },
+    {
+      "epoch": 0.950433944069431,
+      "grad_norm": 0.420449026168815,
+      "learning_rate": 6.194404726209358e-08,
+      "loss": 2.2636,
+      "step": 2464
+    },
+    {
+      "epoch": 0.9508196721311475,
+      "grad_norm": 0.44154819359621317,
+      "learning_rate": 6.098191686860877e-08,
+      "loss": 2.4091,
+      "step": 2465
+    },
+    {
+      "epoch": 0.951205400192864,
+      "grad_norm": 0.407960358547453,
+      "learning_rate": 6.002727125386631e-08,
+      "loss": 2.2806,
+      "step": 2466
+    },
+    {
+      "epoch": 0.9515911282545805,
+      "grad_norm": 0.40314387291041454,
+      "learning_rate": 5.908011186457341e-08,
+      "loss": 2.3228,
+      "step": 2467
+    },
+    {
+      "epoch": 0.951976856316297,
+      "grad_norm": 0.4091025733020704,
+      "learning_rate": 5.8140440136091326e-08,
+      "loss": 2.3326,
+      "step": 2468
+    },
+    {
+      "epoch": 0.9523625843780135,
+      "grad_norm": 0.4194135293166133,
+      "learning_rate": 5.720825749243541e-08,
+      "loss": 2.3769,
+      "step": 2469
+    },
+    {
+      "epoch": 0.95274831243973,
+      "grad_norm": 0.41061934123931143,
+      "learning_rate": 5.628356534627122e-08,
+      "loss": 2.3163,
+      "step": 2470
+    },
+    {
+      "epoch": 0.9531340405014465,
+      "grad_norm": 0.40610058709846136,
+      "learning_rate": 5.536636509891225e-08,
+      "loss": 2.3115,
+      "step": 2471
+    },
+    {
+      "epoch": 0.953519768563163,
+      "grad_norm": 0.41017568899600615,
+      "learning_rate": 5.445665814031942e-08,
+      "loss": 2.2737,
+      "step": 2472
+    },
+    {
+      "epoch": 0.9539054966248794,
+      "grad_norm": 0.42044391166451395,
+      "learning_rate": 5.355444584909886e-08,
+      "loss": 2.2323,
+      "step": 2473
+    },
+    {
+      "epoch": 0.954291224686596,
+      "grad_norm": 0.405416280132108,
+      "learning_rate": 5.265972959249632e-08,
+      "loss": 2.3236,
+      "step": 2474
+    },
+    {
+      "epoch": 0.9546769527483124,
+      "grad_norm": 0.4367778478379245,
+      "learning_rate": 5.1772510726399996e-08,
+      "loss": 2.3579,
+      "step": 2475
+    },
+    {
+      "epoch": 0.9550626808100289,
+      "grad_norm": 0.38365766889537234,
+      "learning_rate": 5.089279059533658e-08,
+      "loss": 2.3188,
+      "step": 2476
+    },
+    {
+      "epoch": 0.9554484088717454,
+      "grad_norm": 0.49310282815831163,
+      "learning_rate": 5.002057053246634e-08,
+      "loss": 2.3614,
+      "step": 2477
+    },
+    {
+      "epoch": 0.9558341369334619,
+      "grad_norm": 0.4340207617363217,
+      "learning_rate": 4.915585185958638e-08,
+      "loss": 2.304,
+      "step": 2478
+    },
+    {
+      "epoch": 0.9562198649951784,
+      "grad_norm": 0.41879047081622034,
+      "learning_rate": 4.829863588712402e-08,
+      "loss": 2.3394,
+      "step": 2479
+    },
+    {
+      "epoch": 0.9566055930568949,
+      "grad_norm": 0.4297196789279753,
+      "learning_rate": 4.744892391413791e-08,
+      "loss": 2.3564,
+      "step": 2480
+    },
+    {
+      "epoch": 0.9569913211186114,
+      "grad_norm": 0.3926456841142491,
+      "learning_rate": 4.660671722831467e-08,
+      "loss": 2.3058,
+      "step": 2481
+    },
+    {
+      "epoch": 0.9573770491803278,
+      "grad_norm": 0.39642546104848503,
+      "learning_rate": 4.577201710596613e-08,
+      "loss": 2.3557,
+      "step": 2482
+    },
+    {
+      "epoch": 0.9577627772420444,
+      "grad_norm": 0.4438348675032907,
+      "learning_rate": 4.4944824812029886e-08,
+      "loss": 2.3176,
+      "step": 2483
+    },
+    {
+      "epoch": 0.9581485053037608,
+      "grad_norm": 0.4157969799084524,
+      "learning_rate": 4.412514160006376e-08,
+      "loss": 2.3298,
+      "step": 2484
+    },
+    {
+      "epoch": 0.9585342333654774,
+      "grad_norm": 0.41079706672034005,
+      "learning_rate": 4.33129687122491e-08,
+      "loss": 2.3701,
+      "step": 2485
+    },
+    {
+      "epoch": 0.9589199614271938,
+      "grad_norm": 0.40202699652356505,
+      "learning_rate": 4.25083073793825e-08,
+      "loss": 2.3547,
+      "step": 2486
+    },
+    {
+      "epoch": 0.9593056894889104,
+      "grad_norm": 0.40564148246556647,
+      "learning_rate": 4.1711158820879613e-08,
+      "loss": 2.3365,
+      "step": 2487
+    },
+    {
+      "epoch": 0.9596914175506268,
+      "grad_norm": 0.4127156328756135,
+      "learning_rate": 4.092152424477025e-08,
+      "loss": 2.274,
+      "step": 2488
+    },
+    {
+      "epoch": 0.9600771456123433,
+      "grad_norm": 0.44242938132635207,
+      "learning_rate": 4.013940484769718e-08,
+      "loss": 2.3907,
+      "step": 2489
+    },
+    {
+      "epoch": 0.9604628736740598,
+      "grad_norm": 0.41195702427983877,
+      "learning_rate": 3.936480181491342e-08,
+      "loss": 2.289,
+      "step": 2490
+    },
+    {
+      "epoch": 0.9608486017357762,
+      "grad_norm": 0.40946865056306514,
+      "learning_rate": 3.859771632028331e-08,
+      "loss": 2.3897,
+      "step": 2491
+    },
+    {
+      "epoch": 0.9612343297974928,
+      "grad_norm": 0.4244008294049646,
+      "learning_rate": 3.7838149526277514e-08,
+      "loss": 2.286,
+      "step": 2492
+    },
+    {
+      "epoch": 0.9616200578592092,
+      "grad_norm": 0.41393559740833624,
+      "learning_rate": 3.70861025839725e-08,
+      "loss": 2.3189,
+      "step": 2493
+    },
+    {
+      "epoch": 0.9620057859209258,
+      "grad_norm": 0.42191742412885425,
+      "learning_rate": 3.634157663304994e-08,
+      "loss": 2.3266,
+      "step": 2494
+    },
+    {
+      "epoch": 0.9623915139826422,
+      "grad_norm": 0.42793463279971766,
+      "learning_rate": 3.560457280179286e-08,
+      "loss": 2.3268,
+      "step": 2495
+    },
+    {
+      "epoch": 0.9627772420443588,
+      "grad_norm": 0.4345097407966401,
+      "learning_rate": 3.487509220708563e-08,
+      "loss": 2.327,
+      "step": 2496
+    },
+    {
+      "epoch": 0.9631629701060752,
+      "grad_norm": 0.40205889662274236,
+      "learning_rate": 3.415313595441116e-08,
+      "loss": 2.2177,
+      "step": 2497
+    },
+    {
+      "epoch": 0.9635486981677918,
+      "grad_norm": 0.42168734366980054,
+      "learning_rate": 3.343870513784875e-08,
+      "loss": 2.3378,
+      "step": 2498
+    },
+    {
+      "epoch": 0.9639344262295082,
+      "grad_norm": 0.4102487120890495,
+      "learning_rate": 3.2731800840076213e-08,
+      "loss": 2.3205,
+      "step": 2499
+    },
+    {
+      "epoch": 0.9643201542912246,
+      "grad_norm": 0.4299063950757636,
+      "learning_rate": 3.2032424132362736e-08,
+      "loss": 2.3536,
+      "step": 2500
+    },
+    {
+      "epoch": 0.9647058823529412,
+      "grad_norm": 0.40180559664214427,
+      "learning_rate": 3.134057607457108e-08,
+      "loss": 2.2852,
+      "step": 2501
+    },
+    {
+      "epoch": 0.9650916104146576,
+      "grad_norm": 0.4291300595198783,
+      "learning_rate": 3.065625771515424e-08,
+      "loss": 2.3647,
+      "step": 2502
+    },
+    {
+      "epoch": 0.9654773384763742,
+      "grad_norm": 0.4030551949343693,
+      "learning_rate": 2.9979470091154315e-08,
+      "loss": 2.3201,
+      "step": 2503
+    },
+    {
+      "epoch": 0.9658630665380906,
+      "grad_norm": 0.4075491282522468,
+      "learning_rate": 2.9310214228202016e-08,
+      "loss": 2.2759,
+      "step": 2504
+    },
+    {
+      "epoch": 0.9662487945998072,
+      "grad_norm": 0.41618422691287965,
+      "learning_rate": 2.8648491140513267e-08,
+      "loss": 2.3569,
+      "step": 2505
+    },
+    {
+      "epoch": 0.9666345226615236,
+      "grad_norm": 0.4083096699215891,
+      "learning_rate": 2.799430183088925e-08,
+      "loss": 2.2653,
+      "step": 2506
+    },
+    {
+      "epoch": 0.9670202507232402,
+      "grad_norm": 0.40400059592556226,
+      "learning_rate": 2.734764729071304e-08,
+      "loss": 2.2471,
+      "step": 2507
+    },
+    {
+      "epoch": 0.9674059787849566,
+      "grad_norm": 0.41087315233608723,
+      "learning_rate": 2.6708528499950758e-08,
+      "loss": 2.2883,
+      "step": 2508
+    },
+    {
+      "epoch": 0.9677917068466731,
+      "grad_norm": 0.4186037693686628,
+      "learning_rate": 2.607694642714653e-08,
+      "loss": 2.2833,
+      "step": 2509
+    },
+    {
+      "epoch": 0.9681774349083896,
+      "grad_norm": 0.43098089391725436,
+      "learning_rate": 2.5452902029425297e-08,
+      "loss": 2.3275,
+      "step": 2510
+    },
+    {
+      "epoch": 0.968563162970106,
+      "grad_norm": 0.44530166646693153,
+      "learning_rate": 2.483639625248724e-08,
+      "loss": 2.2917,
+      "step": 2511
+    },
+    {
+      "epoch": 0.9689488910318226,
+      "grad_norm": 0.4331736474354493,
+      "learning_rate": 2.4227430030609455e-08,
+      "loss": 2.2545,
+      "step": 2512
+    },
+    {
+      "epoch": 0.969334619093539,
+      "grad_norm": 0.4609065384992006,
+      "learning_rate": 2.3626004286642634e-08,
+      "loss": 2.3039,
+      "step": 2513
+    },
+    {
+      "epoch": 0.9697203471552556,
+      "grad_norm": 0.4136615874241027,
+      "learning_rate": 2.3032119932010488e-08,
+      "loss": 2.384,
+      "step": 2514
+    },
+    {
+      "epoch": 0.970106075216972,
+      "grad_norm": 0.40736978504633947,
+      "learning_rate": 2.2445777866709208e-08,
+      "loss": 2.278,
+      "step": 2515
+    },
+    {
+      "epoch": 0.9704918032786886,
+      "grad_norm": 0.4603048542520577,
+      "learning_rate": 2.1866978979303567e-08,
+      "loss": 2.3386,
+      "step": 2516
+    },
+    {
+      "epoch": 0.970877531340405,
+      "grad_norm": 0.39110840096324556,
+      "learning_rate": 2.1295724146926933e-08,
+      "loss": 2.3721,
+      "step": 2517
+    },
+    {
+      "epoch": 0.9712632594021215,
+      "grad_norm": 0.4331393875353548,
+      "learning_rate": 2.073201423528237e-08,
+      "loss": 2.3596,
+      "step": 2518
+    },
+    {
+      "epoch": 0.971648987463838,
+      "grad_norm": 0.42332854514733886,
+      "learning_rate": 2.017585009863654e-08,
+      "loss": 2.3586,
+      "step": 2519
+    },
+    {
+      "epoch": 0.9720347155255545,
+      "grad_norm": 0.43115704440923985,
+      "learning_rate": 1.962723257982302e-08,
+      "loss": 2.3228,
+      "step": 2520
+    },
+    {
+      "epoch": 0.972420443587271,
+      "grad_norm": 0.449521995644178,
+      "learning_rate": 1.9086162510237316e-08,
+      "loss": 2.3004,
+      "step": 2521
+    },
+    {
+      "epoch": 0.9728061716489874,
+      "grad_norm": 0.40298981836234643,
+      "learning_rate": 1.8552640709837977e-08,
+      "loss": 2.3239,
+      "step": 2522
+    },
+    {
+      "epoch": 0.973191899710704,
+      "grad_norm": 0.4490677597059497,
+      "learning_rate": 1.8026667987144363e-08,
+      "loss": 2.3012,
+      "step": 2523
+    },
+    {
+      "epoch": 0.9735776277724204,
+      "grad_norm": 0.4054054120261914,
+      "learning_rate": 1.7508245139236658e-08,
+      "loss": 2.3121,
+      "step": 2524
+    },
+    {
+      "epoch": 0.973963355834137,
+      "grad_norm": 0.43090912391038244,
+      "learning_rate": 1.6997372951751967e-08,
+      "loss": 2.3489,
+      "step": 2525
+    },
+    {
+      "epoch": 0.9743490838958534,
+      "grad_norm": 0.4233189860834805,
+      "learning_rate": 1.6494052198886557e-08,
+      "loss": 2.3873,
+      "step": 2526
+    },
+    {
+      "epoch": 0.97473481195757,
+      "grad_norm": 0.4058108388557626,
+      "learning_rate": 1.59982836433914e-08,
+      "loss": 2.3247,
+      "step": 2527
+    },
+    {
+      "epoch": 0.9751205400192864,
+      "grad_norm": 0.4258335088907587,
+      "learning_rate": 1.5510068036573288e-08,
+      "loss": 2.3609,
+      "step": 2528
+    },
+    {
+      "epoch": 0.9755062680810029,
+      "grad_norm": 0.4207761216385338,
+      "learning_rate": 1.5029406118293732e-08,
+      "loss": 2.3278,
+      "step": 2529
+    },
+    {
+      "epoch": 0.9758919961427194,
+      "grad_norm": 0.4232939417597577,
+      "learning_rate": 1.4556298616965614e-08,
+      "loss": 2.3444,
+      "step": 2530
+    },
+    {
+      "epoch": 0.9762777242044359,
+      "grad_norm": 0.44772513815545695,
+      "learning_rate": 1.4090746249554866e-08,
+      "loss": 2.3005,
+      "step": 2531
+    },
+    {
+      "epoch": 0.9766634522661524,
+      "grad_norm": 0.4501684978470362,
+      "learning_rate": 1.3632749721577132e-08,
+      "loss": 2.2306,
+      "step": 2532
+    },
+    {
+      "epoch": 0.9770491803278688,
+      "grad_norm": 0.44218898437816934,
+      "learning_rate": 1.318230972709833e-08,
+      "loss": 2.3319,
+      "step": 2533
+    },
+    {
+      "epoch": 0.9774349083895854,
+      "grad_norm": 0.4158586144659339,
+      "learning_rate": 1.2739426948732426e-08,
+      "loss": 2.3159,
+      "step": 2534
+    },
+    {
+      "epoch": 0.9778206364513018,
+      "grad_norm": 0.4196125452677132,
+      "learning_rate": 1.2304102057640877e-08,
+      "loss": 2.2893,
+      "step": 2535
+    },
+    {
+      "epoch": 0.9782063645130183,
+      "grad_norm": 0.40891874679918727,
+      "learning_rate": 1.1876335713532638e-08,
+      "loss": 2.2473,
+      "step": 2536
+    },
+    {
+      "epoch": 0.9785920925747348,
+      "grad_norm": 0.4406733934886187,
+      "learning_rate": 1.1456128564660273e-08,
+      "loss": 2.3896,
+      "step": 2537
+    },
+    {
+      "epoch": 0.9789778206364513,
+      "grad_norm": 0.42079275376767034,
+      "learning_rate": 1.1043481247823285e-08,
+      "loss": 2.2683,
+      "step": 2538
+    },
+    {
+      "epoch": 0.9793635486981678,
+      "grad_norm": 0.4008125060938278,
+      "learning_rate": 1.0638394388362006e-08,
+      "loss": 2.2474,
+      "step": 2539
+    },
+    {
+      "epoch": 0.9797492767598843,
+      "grad_norm": 0.4272944856472343,
+      "learning_rate": 1.024086860016149e-08,
+      "loss": 2.3548,
+      "step": 2540
+    },
+    {
+      "epoch": 0.9801350048216008,
+      "grad_norm": 0.40159087051803666,
+      "learning_rate": 9.850904485647072e-09,
+      "loss": 2.3237,
+      "step": 2541
+    },
+    {
+      "epoch": 0.9805207328833173,
+      "grad_norm": 0.4156694315192405,
+      "learning_rate": 9.468502635786026e-09,
+      "loss": 2.3672,
+      "step": 2542
+    },
+    {
+      "epoch": 0.9809064609450338,
+      "grad_norm": 0.422406372438295,
+      "learning_rate": 9.093663630084237e-09,
+      "loss": 2.3174,
+      "step": 2543
+    },
+    {
+      "epoch": 0.9812921890067502,
+      "grad_norm": 0.44066709303635726,
+      "learning_rate": 8.726388036587874e-09,
+      "loss": 2.3132,
+      "step": 2544
+    },
+    {
+      "epoch": 0.9816779170684667,
+      "grad_norm": 0.41356414626582866,
+      "learning_rate": 8.366676411880602e-09,
+      "loss": 2.2772,
+      "step": 2545
+    },
+    {
+      "epoch": 0.9820636451301832,
+      "grad_norm": 0.42263596386774394,
+      "learning_rate": 8.014529301082485e-09,
+      "loss": 2.2357,
+      "step": 2546
+    },
+    {
+      "epoch": 0.9824493731918997,
+      "grad_norm": 0.43398917077284627,
+      "learning_rate": 7.669947237851637e-09,
+      "loss": 2.3422,
+      "step": 2547
+    },
+    {
+      "epoch": 0.9828351012536162,
+      "grad_norm": 0.4001274687897164,
+      "learning_rate": 7.332930744380906e-09,
+      "loss": 2.3663,
+      "step": 2548
+    },
+    {
+      "epoch": 0.9832208293153327,
+      "grad_norm": 0.4070491519301213,
+      "learning_rate": 7.00348033139786e-09,
+      "loss": 2.3069,
+      "step": 2549
+    },
+    {
+      "epoch": 0.9836065573770492,
+      "grad_norm": 0.42079521174878126,
+      "learning_rate": 6.681596498164244e-09,
+      "loss": 2.2963,
+      "step": 2550
+    },
+    {
+      "epoch": 0.9839922854387657,
+      "grad_norm": 0.43723928010333774,
+      "learning_rate": 6.367279732475418e-09,
+      "loss": 2.358,
+      "step": 2551
+    },
+    {
+      "epoch": 0.9843780135004822,
+      "grad_norm": 0.4108275482871391,
+      "learning_rate": 6.060530510659246e-09,
+      "loss": 2.326,
+      "step": 2552
+    },
+    {
+      "epoch": 0.9847637415621987,
+      "grad_norm": 0.4350703058011143,
+      "learning_rate": 5.761349297575547e-09,
+      "loss": 2.3581,
+      "step": 2553
+    },
+    {
+      "epoch": 0.9851494696239151,
+      "grad_norm": 0.43235359013039043,
+      "learning_rate": 5.469736546614979e-09,
+      "loss": 2.2757,
+      "step": 2554
+    },
+    {
+      "epoch": 0.9855351976856316,
+      "grad_norm": 0.40961132656186466,
+      "learning_rate": 5.185692699697931e-09,
+      "loss": 2.36,
+      "step": 2555
+    },
+    {
+      "epoch": 0.9859209257473481,
+      "grad_norm": 0.4430751567999117,
+      "learning_rate": 4.909218187276743e-09,
+      "loss": 2.3624,
+      "step": 2556
+    },
+    {
+      "epoch": 0.9863066538090646,
+      "grad_norm": 0.4289839532627191,
+      "learning_rate": 4.640313428330711e-09,
+      "loss": 2.2914,
+      "step": 2557
+    },
+    {
+      "epoch": 0.9866923818707811,
+      "grad_norm": 0.44036969688797795,
+      "learning_rate": 4.378978830368863e-09,
+      "loss": 2.2774,
+      "step": 2558
+    },
+    {
+      "epoch": 0.9870781099324976,
+      "grad_norm": 0.4143311437290899,
+      "learning_rate": 4.125214789427734e-09,
+      "loss": 2.2661,
+      "step": 2559
+    },
+    {
+      "epoch": 0.9874638379942141,
+      "grad_norm": 0.4240774884679542,
+      "learning_rate": 3.8790216900702615e-09,
+      "loss": 2.2581,
+      "step": 2560
+    },
+    {
+      "epoch": 0.9878495660559306,
+      "grad_norm": 0.41615959808469105,
+      "learning_rate": 3.6403999053885584e-09,
+      "loss": 2.3018,
+      "step": 2561
+    },
+    {
+      "epoch": 0.9882352941176471,
+      "grad_norm": 0.41624863195931333,
+      "learning_rate": 3.4093497969983625e-09,
+      "loss": 2.4007,
+      "step": 2562
+    },
+    {
+      "epoch": 0.9886210221793635,
+      "grad_norm": 0.4130284474102173,
+      "learning_rate": 3.1858717150412554e-09,
+      "loss": 2.3366,
+      "step": 2563
+    },
+    {
+      "epoch": 0.9890067502410801,
+      "grad_norm": 0.4268181991955994,
+      "learning_rate": 2.9699659981863306e-09,
+      "loss": 2.2902,
+      "step": 2564
+    },
+    {
+      "epoch": 0.9893924783027965,
+      "grad_norm": 0.42065001359224774,
+      "learning_rate": 2.761632973624084e-09,
+      "loss": 2.3251,
+      "step": 2565
+    },
+    {
+      "epoch": 0.989778206364513,
+      "grad_norm": 0.43868310422689855,
+      "learning_rate": 2.5608729570703037e-09,
+      "loss": 2.3727,
+      "step": 2566
+    },
+    {
+      "epoch": 0.9901639344262295,
+      "grad_norm": 0.40977965981613995,
+      "learning_rate": 2.367686252765511e-09,
+      "loss": 2.3281,
+      "step": 2567
+    },
+    {
+      "epoch": 0.990549662487946,
+      "grad_norm": 0.4121241906272853,
+      "learning_rate": 2.182073153471631e-09,
+      "loss": 2.2966,
+      "step": 2568
+    },
+    {
+      "epoch": 0.9909353905496625,
+      "grad_norm": 0.42911358142367184,
+      "learning_rate": 2.0040339404742147e-09,
+      "loss": 2.357,
+      "step": 2569
+    },
+    {
+      "epoch": 0.991321118611379,
+      "grad_norm": 0.43132519983787015,
+      "learning_rate": 1.8335688835802169e-09,
+      "loss": 2.3263,
+      "step": 2570
+    },
+    {
+      "epoch": 0.9917068466730955,
+      "grad_norm": 0.4363798226276988,
+      "learning_rate": 1.670678241119661e-09,
+      "loss": 2.3435,
+      "step": 2571
+    },
+    {
+      "epoch": 0.9920925747348119,
+      "grad_norm": 0.4289536892138021,
+      "learning_rate": 1.5153622599428652e-09,
+      "loss": 2.2977,
+      "step": 2572
+    },
+    {
+      "epoch": 0.9924783027965285,
+      "grad_norm": 0.4062667678567385,
+      "learning_rate": 1.3676211754215518e-09,
+      "loss": 2.2861,
+      "step": 2573
+    },
+    {
+      "epoch": 0.9928640308582449,
+      "grad_norm": 0.4275697437950462,
+      "learning_rate": 1.227455211448847e-09,
+      "loss": 2.3045,
+      "step": 2574
+    },
+    {
+      "epoch": 0.9932497589199615,
+      "grad_norm": 0.41249327348223197,
+      "learning_rate": 1.0948645804370605e-09,
+      "loss": 2.3402,
+      "step": 2575
+    },
+    {
+      "epoch": 0.9936354869816779,
+      "grad_norm": 0.4091376478413197,
+      "learning_rate": 9.698494833199068e-10,
+      "loss": 2.3213,
+      "step": 2576
+    },
+    {
+      "epoch": 0.9940212150433944,
+      "grad_norm": 0.44573230043842443,
+      "learning_rate": 8.524101095491733e-10,
+      "loss": 2.2555,
+      "step": 2577
+    },
+    {
+      "epoch": 0.9944069431051109,
+      "grad_norm": 0.4352921106893979,
+      "learning_rate": 7.425466370974965e-10,
+      "loss": 2.3701,
+      "step": 2578
+    },
+    {
+      "epoch": 0.9947926711668273,
+      "grad_norm": 0.43617276647783876,
+      "learning_rate": 6.402592324561418e-10,
+      "loss": 2.3061,
+      "step": 2579
+    },
+    {
+      "epoch": 0.9951783992285439,
+      "grad_norm": 0.4551060450487974,
+      "learning_rate": 5.455480506355582e-10,
+      "loss": 2.3767,
+      "step": 2580
+    },
+    {
+      "epoch": 0.9955641272902603,
+      "grad_norm": 0.4117741397937645,
+      "learning_rate": 4.5841323516426784e-10,
+      "loss": 2.3894,
+      "step": 2581
+    },
+    {
+      "epoch": 0.9959498553519769,
+      "grad_norm": 0.39894957552658944,
+      "learning_rate": 3.7885491809053207e-10,
+      "loss": 2.2485,
+      "step": 2582
+    },
+    {
+      "epoch": 0.9963355834136933,
+      "grad_norm": 0.41464108147053547,
+      "learning_rate": 3.0687321997957543e-10,
+      "loss": 2.324,
+      "step": 2583
+    },
+    {
+      "epoch": 0.9967213114754099,
+      "grad_norm": 0.4423602065102726,
+      "learning_rate": 2.4246824991525085e-10,
+      "loss": 2.3225,
+      "step": 2584
+    },
+    {
+      "epoch": 0.9971070395371263,
+      "grad_norm": 0.4362643537518209,
+      "learning_rate": 1.8564010549948497e-10,
+      "loss": 2.2874,
+      "step": 2585
+    },
+    {
+      "epoch": 0.9974927675988429,
+      "grad_norm": 0.4346366544157702,
+      "learning_rate": 1.3638887285116753e-10,
+      "loss": 2.3311,
+      "step": 2586
+    },
+    {
+      "epoch": 0.9978784956605593,
+      "grad_norm": 0.4214231323688105,
+      "learning_rate": 9.471462660892716e-11,
+      "loss": 2.3397,
+      "step": 2587
+    },
+    {
+      "epoch": 0.9982642237222757,
+      "grad_norm": 0.4105430433409002,
+      "learning_rate": 6.061742992613529e-11,
+      "loss": 2.3279,
+      "step": 2588
+    },
+    {
+      "epoch": 0.9986499517839923,
+      "grad_norm": 0.42157880275509585,
+      "learning_rate": 3.4097334475902135e-11,
+      "loss": 2.3044,
+      "step": 2589
+    },
+    {
+      "epoch": 0.9990356798457087,
+      "grad_norm": 0.4295909376753097,
+      "learning_rate": 1.5154380447190976e-11,
+      "loss": 2.3718,
+      "step": 2590
+    },
+    {
+      "epoch": 0.9994214079074253,
+      "grad_norm": 0.42028982333761167,
+      "learning_rate": 3.788596547038559e-12,
+      "loss": 2.2899,
+      "step": 2591
+    },
+    {
+      "epoch": 0.9998071359691417,
+      "grad_norm": 0.392346457170918,
+      "learning_rate": 0.0,
+      "loss": 2.3139,
+      "step": 2592
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 2592,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 1296,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.591619149639975e+18,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}