diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7034 @@
+{
+  "best_global_step": null,
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.13995801259622112,
+  "eval_steps": 500,
+  "global_step": 1000,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.00013995801259622114,
+      "grad_norm": 1.6205339948497404,
+      "learning_rate": 0.0,
+      "loss": 1.8649,
+      "step": 1
+    },
+    {
+      "epoch": 0.00027991602519244227,
+      "grad_norm": 1.8823543727463894,
+      "learning_rate": 2.3255813953488374e-07,
+      "loss": 1.8156,
+      "step": 2
+    },
+    {
+      "epoch": 0.0004198740377886634,
+      "grad_norm": 1.5734521895417912,
+      "learning_rate": 4.651162790697675e-07,
+      "loss": 1.8395,
+      "step": 3
+    },
+    {
+      "epoch": 0.0005598320503848845,
+      "grad_norm": 1.884881935408653,
+      "learning_rate": 6.976744186046511e-07,
+      "loss": 1.8379,
+      "step": 4
+    },
+    {
+      "epoch": 0.0006997900629811056,
+      "grad_norm": 1.570942303353113,
+      "learning_rate": 9.30232558139535e-07,
+      "loss": 1.7683,
+      "step": 5
+    },
+    {
+      "epoch": 0.0008397480755773268,
+      "grad_norm": 2.1232694447925367,
+      "learning_rate": 1.1627906976744186e-06,
+      "loss": 1.834,
+      "step": 6
+    },
+    {
+      "epoch": 0.000979706088173548,
+      "grad_norm": 1.5460826364849427,
+      "learning_rate": 1.3953488372093023e-06,
+      "loss": 1.8431,
+      "step": 7
+    },
+    {
+      "epoch": 0.001119664100769769,
+      "grad_norm": 1.5383810682727819,
+      "learning_rate": 1.627906976744186e-06,
+      "loss": 1.7739,
+      "step": 8
+    },
+    {
+      "epoch": 0.0012596221133659902,
+      "grad_norm": 2.321401184160743,
+      "learning_rate": 1.86046511627907e-06,
+      "loss": 1.7974,
+      "step": 9
+    },
+    {
+      "epoch": 0.0013995801259622112,
+      "grad_norm": 1.5959823871905185,
+      "learning_rate": 2.0930232558139536e-06,
+      "loss": 1.802,
+      "step": 10
+    },
+    {
+      "epoch": 0.0015395381385584325,
+      "grad_norm": 1.4312201376848954,
+      "learning_rate": 2.325581395348837e-06,
+      "loss": 1.8292,
+      "step": 11
+    },
+    {
+      "epoch": 0.0016794961511546536,
+      "grad_norm": 1.3848325251343343,
+      "learning_rate": 2.558139534883721e-06,
+      "loss": 1.7843,
+      "step": 12
+    },
+    {
+      "epoch": 0.0018194541637508747,
+      "grad_norm": 1.5003345129280181,
+      "learning_rate": 2.7906976744186046e-06,
+      "loss": 1.8619,
+      "step": 13
+    },
+    {
+      "epoch": 0.001959412176347096,
+      "grad_norm": 1.5397334321812373,
+      "learning_rate": 3.0232558139534885e-06,
+      "loss": 1.8203,
+      "step": 14
+    },
+    {
+      "epoch": 0.002099370188943317,
+      "grad_norm": 1.5024026903634549,
+      "learning_rate": 3.255813953488372e-06,
+      "loss": 1.7945,
+      "step": 15
+    },
+    {
+      "epoch": 0.002239328201539538,
+      "grad_norm": 1.4537582898628276,
+      "learning_rate": 3.488372093023256e-06,
+      "loss": 1.7954,
+      "step": 16
+    },
+    {
+      "epoch": 0.0023792862141357595,
+      "grad_norm": 1.4392959782369958,
+      "learning_rate": 3.72093023255814e-06,
+      "loss": 1.7539,
+      "step": 17
+    },
+    {
+      "epoch": 0.0025192442267319803,
+      "grad_norm": 1.4318028272838452,
+      "learning_rate": 3.953488372093024e-06,
+      "loss": 1.7287,
+      "step": 18
+    },
+    {
+      "epoch": 0.0026592022393282016,
+      "grad_norm": 1.5567916289389137,
+      "learning_rate": 4.186046511627907e-06,
+      "loss": 1.7651,
+      "step": 19
+    },
+    {
+      "epoch": 0.0027991602519244225,
+      "grad_norm": 2.133786571708112,
+      "learning_rate": 4.418604651162791e-06,
+      "loss": 1.7431,
+      "step": 20
+    },
+    {
+      "epoch": 0.0029391182645206438,
+      "grad_norm": 1.3887907218945514,
+      "learning_rate": 4.651162790697674e-06,
+      "loss": 1.769,
+      "step": 21
+    },
+    {
+      "epoch": 0.003079076277116865,
+      "grad_norm": 1.4440046752202067,
+      "learning_rate": 4.883720930232559e-06,
+      "loss": 1.6852,
+      "step": 22
+    },
+    {
+      "epoch": 0.003219034289713086,
+      "grad_norm": 1.7560763608783,
+      "learning_rate": 5.116279069767442e-06,
+      "loss": 1.7515,
+      "step": 23
+    },
+    {
+      "epoch": 0.0033589923023093072,
+      "grad_norm": 1.4866717560719882,
+      "learning_rate": 5.348837209302326e-06,
+      "loss": 1.7103,
+      "step": 24
+    },
+    {
+      "epoch": 0.0034989503149055285,
+      "grad_norm": 1.3191066550690622,
+      "learning_rate": 5.581395348837209e-06,
+      "loss": 1.6115,
+      "step": 25
+    },
+    {
+      "epoch": 0.0036389083275017494,
+      "grad_norm": 3.5279348783477205,
+      "learning_rate": 5.8139534883720935e-06,
+      "loss": 1.639,
+      "step": 26
+    },
+    {
+      "epoch": 0.0037788663400979707,
+      "grad_norm": 1.4814448441008254,
+      "learning_rate": 6.046511627906977e-06,
+      "loss": 1.6249,
+      "step": 27
+    },
+    {
+      "epoch": 0.003918824352694192,
+      "grad_norm": 1.2456228376106218,
+      "learning_rate": 6.279069767441861e-06,
+      "loss": 1.5937,
+      "step": 28
+    },
+    {
+      "epoch": 0.004058782365290413,
+      "grad_norm": 1.2032083110394567,
+      "learning_rate": 6.511627906976744e-06,
+      "loss": 1.57,
+      "step": 29
+    },
+    {
+      "epoch": 0.004198740377886634,
+      "grad_norm": 1.1873685890041599,
+      "learning_rate": 6.744186046511629e-06,
+      "loss": 1.5443,
+      "step": 30
+    },
+    {
+      "epoch": 0.0043386983904828555,
+      "grad_norm": 1.1206050573785102,
+      "learning_rate": 6.976744186046512e-06,
+      "loss": 1.4658,
+      "step": 31
+    },
+    {
+      "epoch": 0.004478656403079076,
+      "grad_norm": 1.613173262015054,
+      "learning_rate": 7.209302325581396e-06,
+      "loss": 1.5061,
+      "step": 32
+    },
+    {
+      "epoch": 0.004618614415675297,
+      "grad_norm": 1.0997443478827273,
+      "learning_rate": 7.44186046511628e-06,
+      "loss": 1.4946,
+      "step": 33
+    },
+    {
+      "epoch": 0.004758572428271519,
+      "grad_norm": 1.0976998793927069,
+      "learning_rate": 7.674418604651164e-06,
+      "loss": 1.4857,
+      "step": 34
+    },
+    {
+      "epoch": 0.00489853044086774,
+      "grad_norm": 0.9524795998625615,
+      "learning_rate": 7.906976744186048e-06,
+      "loss": 1.4493,
+      "step": 35
+    },
+    {
+      "epoch": 0.005038488453463961,
+      "grad_norm": 0.9642671767042398,
+      "learning_rate": 8.139534883720931e-06,
+      "loss": 1.4039,
+      "step": 36
+    },
+    {
+      "epoch": 0.005178446466060182,
+      "grad_norm": 0.8572170901816704,
+      "learning_rate": 8.372093023255815e-06,
+      "loss": 1.3529,
+      "step": 37
+    },
+    {
+      "epoch": 0.005318404478656403,
+      "grad_norm": 0.8008007221794986,
+      "learning_rate": 8.604651162790698e-06,
+      "loss": 1.3514,
+      "step": 38
+    },
+    {
+      "epoch": 0.005458362491252624,
+      "grad_norm": 1.484907272279964,
+      "learning_rate": 8.837209302325582e-06,
+      "loss": 1.3637,
+      "step": 39
+    },
+    {
+      "epoch": 0.005598320503848845,
+      "grad_norm": 1.3633541730503271,
+      "learning_rate": 9.069767441860467e-06,
+      "loss": 1.3263,
+      "step": 40
+    },
+    {
+      "epoch": 0.005738278516445067,
+      "grad_norm": 0.6081019616299234,
+      "learning_rate": 9.302325581395349e-06,
+      "loss": 1.333,
+      "step": 41
+    },
+    {
+      "epoch": 0.0058782365290412876,
+      "grad_norm": 0.6853488905437116,
+      "learning_rate": 9.534883720930234e-06,
+      "loss": 1.2903,
+      "step": 42
+    },
+    {
+      "epoch": 0.006018194541637508,
+      "grad_norm": 0.8466578955626367,
+      "learning_rate": 9.767441860465117e-06,
+      "loss": 1.3079,
+      "step": 43
+    },
+    {
+      "epoch": 0.00615815255423373,
+      "grad_norm": 0.567927473680037,
+      "learning_rate": 1e-05,
+      "loss": 1.2735,
+      "step": 44
+    },
+    {
+      "epoch": 0.006298110566829951,
+      "grad_norm": 0.6150202452547671,
+      "learning_rate": 1.0232558139534884e-05,
+      "loss": 1.2115,
+      "step": 45
+    },
+    {
+      "epoch": 0.006438068579426172,
+      "grad_norm": 0.4989582290513981,
+      "learning_rate": 1.0465116279069768e-05,
+      "loss": 1.2315,
+      "step": 46
+    },
+    {
+      "epoch": 0.006578026592022394,
+      "grad_norm": 0.671517817392481,
+      "learning_rate": 1.0697674418604651e-05,
+      "loss": 1.211,
+      "step": 47
+    },
+    {
+      "epoch": 0.0067179846046186145,
+      "grad_norm": 0.4743826379981072,
+      "learning_rate": 1.0930232558139537e-05,
+      "loss": 1.2323,
+      "step": 48
+    },
+    {
+      "epoch": 0.006857942617214835,
+      "grad_norm": 0.6130428027193264,
+      "learning_rate": 1.1162790697674418e-05,
+      "loss": 1.1871,
+      "step": 49
+    },
+    {
+      "epoch": 0.006997900629811057,
+      "grad_norm": 0.5556426649020297,
+      "learning_rate": 1.1395348837209304e-05,
+      "loss": 1.1869,
+      "step": 50
+    },
+    {
+      "epoch": 0.007137858642407278,
+      "grad_norm": 1.896863162731097,
+      "learning_rate": 1.1627906976744187e-05,
+      "loss": 1.1683,
+      "step": 51
+    },
+    {
+      "epoch": 0.007277816655003499,
+      "grad_norm": 0.8412820077594921,
+      "learning_rate": 1.186046511627907e-05,
+      "loss": 1.1453,
+      "step": 52
+    },
+    {
+      "epoch": 0.00741777466759972,
+      "grad_norm": 0.5453872617764336,
+      "learning_rate": 1.2093023255813954e-05,
+      "loss": 1.1449,
+      "step": 53
+    },
+    {
+      "epoch": 0.007557732680195941,
+      "grad_norm": 0.5447374444716608,
+      "learning_rate": 1.2325581395348838e-05,
+      "loss": 1.0974,
+      "step": 54
+    },
+    {
+      "epoch": 0.007697690692792162,
+      "grad_norm": 0.4102706328599184,
+      "learning_rate": 1.2558139534883723e-05,
+      "loss": 1.1077,
+      "step": 55
+    },
+    {
+      "epoch": 0.007837648705388384,
+      "grad_norm": 3.5443873887390946,
+      "learning_rate": 1.2790697674418606e-05,
+      "loss": 1.1119,
+      "step": 56
+    },
+    {
+      "epoch": 0.007977606717984604,
+      "grad_norm": 0.46404403603707844,
+      "learning_rate": 1.3023255813953488e-05,
+      "loss": 1.074,
+      "step": 57
+    },
+    {
+      "epoch": 0.008117564730580826,
+      "grad_norm": 0.39045250619813876,
+      "learning_rate": 1.3255813953488372e-05,
+      "loss": 1.1102,
+      "step": 58
+    },
+    {
+      "epoch": 0.008257522743177047,
+      "grad_norm": 1.343556937733491,
+      "learning_rate": 1.3488372093023258e-05,
+      "loss": 1.1127,
+      "step": 59
+    },
+    {
+      "epoch": 0.008397480755773267,
+      "grad_norm": 0.7610889883379984,
+      "learning_rate": 1.372093023255814e-05,
+      "loss": 1.0093,
+      "step": 60
+    },
+    {
+      "epoch": 0.00853743876836949,
+      "grad_norm": 0.4556554186064198,
+      "learning_rate": 1.3953488372093024e-05,
+      "loss": 1.0192,
+      "step": 61
+    },
+    {
+      "epoch": 0.008677396780965711,
+      "grad_norm": 0.501444946545551,
+      "learning_rate": 1.4186046511627907e-05,
+      "loss": 1.0403,
+      "step": 62
+    },
+    {
+      "epoch": 0.008817354793561931,
+      "grad_norm": 0.35988075895590804,
+      "learning_rate": 1.4418604651162792e-05,
+      "loss": 1.1192,
+      "step": 63
+    },
+    {
+      "epoch": 0.008957312806158153,
+      "grad_norm": 0.6921457564720671,
+      "learning_rate": 1.4651162790697676e-05,
+      "loss": 1.056,
+      "step": 64
+    },
+    {
+      "epoch": 0.009097270818754374,
+      "grad_norm": 0.37626208734508315,
+      "learning_rate": 1.488372093023256e-05,
+      "loss": 1.0427,
+      "step": 65
+    },
+    {
+      "epoch": 0.009237228831350594,
+      "grad_norm": 0.4227916388896408,
+      "learning_rate": 1.5116279069767441e-05,
+      "loss": 1.0649,
+      "step": 66
+    },
+    {
+      "epoch": 0.009377186843946816,
+      "grad_norm": 0.43865399337074507,
+      "learning_rate": 1.5348837209302328e-05,
+      "loss": 1.0342,
+      "step": 67
+    },
+    {
+      "epoch": 0.009517144856543038,
+      "grad_norm": 1.1110606201155102,
+      "learning_rate": 1.558139534883721e-05,
+      "loss": 1.079,
+      "step": 68
+    },
+    {
+      "epoch": 0.009657102869139258,
+      "grad_norm": 0.3429950483555357,
+      "learning_rate": 1.5813953488372095e-05,
+      "loss": 1.0354,
+      "step": 69
+    },
+    {
+      "epoch": 0.00979706088173548,
+      "grad_norm": 0.500911837092983,
+      "learning_rate": 1.6046511627906977e-05,
+      "loss": 1.0117,
+      "step": 70
+    },
+    {
+      "epoch": 0.009937018894331701,
+      "grad_norm": 0.5054192697417574,
+      "learning_rate": 1.6279069767441862e-05,
+      "loss": 1.0435,
+      "step": 71
+    },
+    {
+      "epoch": 0.010076976906927921,
+      "grad_norm": 0.4024563344236819,
+      "learning_rate": 1.6511627906976744e-05,
+      "loss": 1.045,
+      "step": 72
+    },
+    {
+      "epoch": 0.010216934919524143,
+      "grad_norm": 0.4790998578083277,
+      "learning_rate": 1.674418604651163e-05,
+      "loss": 1.0253,
+      "step": 73
+    },
+    {
+      "epoch": 0.010356892932120365,
+      "grad_norm": 0.45731467968480566,
+      "learning_rate": 1.697674418604651e-05,
+      "loss": 1.0088,
+      "step": 74
+    },
+    {
+      "epoch": 0.010496850944716585,
+      "grad_norm": 0.4617808389918637,
+      "learning_rate": 1.7209302325581396e-05,
+      "loss": 0.9946,
+      "step": 75
+    },
+    {
+      "epoch": 0.010636808957312806,
+      "grad_norm": 0.4274246905900338,
+      "learning_rate": 1.744186046511628e-05,
+      "loss": 1.0078,
+      "step": 76
+    },
+    {
+      "epoch": 0.010776766969909026,
+      "grad_norm": 0.4859617233595146,
+      "learning_rate": 1.7674418604651163e-05,
+      "loss": 1.0292,
+      "step": 77
+    },
+    {
+      "epoch": 0.010916724982505248,
+      "grad_norm": 0.5837878294485921,
+      "learning_rate": 1.7906976744186045e-05,
+      "loss": 1.0338,
+      "step": 78
+    },
+    {
+      "epoch": 0.01105668299510147,
+      "grad_norm": 1.7173283109289441,
+      "learning_rate": 1.8139534883720934e-05,
+      "loss": 1.0126,
+      "step": 79
+    },
+    {
+      "epoch": 0.01119664100769769,
+      "grad_norm": 0.5256113477743695,
+      "learning_rate": 1.8372093023255815e-05,
+      "loss": 0.9907,
+      "step": 80
+    },
+    {
+      "epoch": 0.011336599020293912,
+      "grad_norm": 0.4483569419385721,
+      "learning_rate": 1.8604651162790697e-05,
+      "loss": 0.9643,
+      "step": 81
+    },
+    {
+      "epoch": 0.011476557032890133,
+      "grad_norm": 0.4539441530526643,
+      "learning_rate": 1.8837209302325582e-05,
+      "loss": 1.0462,
+      "step": 82
+    },
+    {
+      "epoch": 0.011616515045486353,
+      "grad_norm": 0.33426240052710565,
+      "learning_rate": 1.9069767441860468e-05,
+      "loss": 0.954,
+      "step": 83
+    },
+    {
+      "epoch": 0.011756473058082575,
+      "grad_norm": 0.3396314689877589,
+      "learning_rate": 1.930232558139535e-05,
+      "loss": 0.9857,
+      "step": 84
+    },
+    {
+      "epoch": 0.011896431070678797,
+      "grad_norm": 0.395155279481088,
+      "learning_rate": 1.9534883720930235e-05,
+      "loss": 1.004,
+      "step": 85
+    },
+    {
+      "epoch": 0.012036389083275017,
+      "grad_norm": 0.38928902258671755,
+      "learning_rate": 1.9767441860465116e-05,
+      "loss": 0.9253,
+      "step": 86
+    },
+    {
+      "epoch": 0.012176347095871239,
+      "grad_norm": 0.3964184151155477,
+      "learning_rate": 2e-05,
+      "loss": 0.9562,
+      "step": 87
+    },
+    {
+      "epoch": 0.01231630510846746,
+      "grad_norm": 0.41507841516147836,
+      "learning_rate": 2.0232558139534883e-05,
+      "loss": 0.9857,
+      "step": 88
+    },
+    {
+      "epoch": 0.01245626312106368,
+      "grad_norm": 0.3303740429955767,
+      "learning_rate": 2.046511627906977e-05,
+      "loss": 0.96,
+      "step": 89
+    },
+    {
+      "epoch": 0.012596221133659902,
+      "grad_norm": 0.35729090571512173,
+      "learning_rate": 2.069767441860465e-05,
+      "loss": 0.9453,
+      "step": 90
+    },
+    {
+      "epoch": 0.012736179146256124,
+      "grad_norm": 0.34652306094213037,
+      "learning_rate": 2.0930232558139536e-05,
+      "loss": 0.9478,
+      "step": 91
+    },
+    {
+      "epoch": 0.012876137158852344,
+      "grad_norm": 0.3781323029481633,
+      "learning_rate": 2.116279069767442e-05,
+      "loss": 0.9232,
+      "step": 92
+    },
+    {
+      "epoch": 0.013016095171448566,
+      "grad_norm": 0.3615380548631367,
+      "learning_rate": 2.1395348837209303e-05,
+      "loss": 0.903,
+      "step": 93
+    },
+    {
+      "epoch": 0.013156053184044787,
+      "grad_norm": 0.4096150963666029,
+      "learning_rate": 2.1627906976744184e-05,
+      "loss": 0.9599,
+      "step": 94
+    },
+    {
+      "epoch": 0.013296011196641007,
+      "grad_norm": 0.38313805223400516,
+      "learning_rate": 2.1860465116279073e-05,
+      "loss": 0.9512,
+      "step": 95
+    },
+    {
+      "epoch": 0.013435969209237229,
+      "grad_norm": 0.4249487976331709,
+      "learning_rate": 2.2093023255813955e-05,
+      "loss": 0.888,
+      "step": 96
+    },
+    {
+      "epoch": 0.01357592722183345,
+      "grad_norm": 0.3958713012909362,
+      "learning_rate": 2.2325581395348837e-05,
+      "loss": 0.9366,
+      "step": 97
+    },
+    {
+      "epoch": 0.01371588523442967,
+      "grad_norm": 0.3666415905158996,
+      "learning_rate": 2.2558139534883722e-05,
+      "loss": 0.9161,
+      "step": 98
+    },
+    {
+      "epoch": 0.013855843247025892,
+      "grad_norm": 0.4821438335731706,
+      "learning_rate": 2.2790697674418607e-05,
+      "loss": 0.9352,
+      "step": 99
+    },
+    {
+      "epoch": 0.013995801259622114,
+      "grad_norm": 0.3878615433598532,
+      "learning_rate": 2.302325581395349e-05,
+      "loss": 0.9557,
+      "step": 100
+    },
+    {
+      "epoch": 0.014135759272218334,
+      "grad_norm": 5.323533163270388,
+      "learning_rate": 2.3255813953488374e-05,
+      "loss": 0.9419,
+      "step": 101
+    },
+    {
+      "epoch": 0.014275717284814556,
+      "grad_norm": 0.6042685685685957,
+      "learning_rate": 2.3488372093023256e-05,
+      "loss": 0.9249,
+      "step": 102
+    },
+    {
+      "epoch": 0.014415675297410778,
+      "grad_norm": 0.4137056327381694,
+      "learning_rate": 2.372093023255814e-05,
+      "loss": 0.9179,
+      "step": 103
+    },
+    {
+      "epoch": 0.014555633310006998,
+      "grad_norm": 0.3772048144992787,
+      "learning_rate": 2.3953488372093026e-05,
+      "loss": 0.9144,
+      "step": 104
+    },
+    {
+      "epoch": 0.01469559132260322,
+      "grad_norm": 0.4175730821083398,
+      "learning_rate": 2.4186046511627908e-05,
+      "loss": 0.9227,
+      "step": 105
+    },
+    {
+      "epoch": 0.01483554933519944,
+      "grad_norm": 0.40061188528050645,
+      "learning_rate": 2.441860465116279e-05,
+      "loss": 0.9278,
+      "step": 106
+    },
+    {
+      "epoch": 0.014975507347795661,
+      "grad_norm": 0.3978558565443655,
+      "learning_rate": 2.4651162790697675e-05,
+      "loss": 0.8924,
+      "step": 107
+    },
+    {
+      "epoch": 0.015115465360391883,
+      "grad_norm": 0.8747382632018756,
+      "learning_rate": 2.488372093023256e-05,
+      "loss": 0.9015,
+      "step": 108
+    },
+    {
+      "epoch": 0.015255423372988103,
+      "grad_norm": 0.39737512145413006,
+      "learning_rate": 2.5116279069767445e-05,
+      "loss": 0.9201,
+      "step": 109
+    },
+    {
+      "epoch": 0.015395381385584325,
+      "grad_norm": 0.41650482599802,
+      "learning_rate": 2.5348837209302327e-05,
+      "loss": 0.8626,
+      "step": 110
+    },
+    {
+      "epoch": 0.015535339398180546,
+      "grad_norm": 0.42011360836146605,
+      "learning_rate": 2.5581395348837212e-05,
+      "loss": 0.8862,
+      "step": 111
+    },
+    {
+      "epoch": 0.015675297410776768,
+      "grad_norm": 0.4680549878413964,
+      "learning_rate": 2.5813953488372094e-05,
+      "loss": 0.9105,
+      "step": 112
+    },
+    {
+      "epoch": 0.01581525542337299,
+      "grad_norm": 0.49968103218957055,
+      "learning_rate": 2.6046511627906976e-05,
+      "loss": 0.8913,
+      "step": 113
+    },
+    {
+      "epoch": 0.015955213435969208,
+      "grad_norm": 0.49556276645971464,
+      "learning_rate": 2.627906976744186e-05,
+      "loss": 0.9266,
+      "step": 114
+    },
+    {
+      "epoch": 0.01609517144856543,
+      "grad_norm": 0.40655693565880857,
+      "learning_rate": 2.6511627906976743e-05,
+      "loss": 0.8941,
+      "step": 115
+    },
+    {
+      "epoch": 0.01623512946116165,
+      "grad_norm": 0.39000307235707055,
+      "learning_rate": 2.674418604651163e-05,
+      "loss": 0.8655,
+      "step": 116
+    },
+    {
+      "epoch": 0.016375087473757873,
+      "grad_norm": 0.4264245544215563,
+      "learning_rate": 2.6976744186046517e-05,
+      "loss": 0.837,
+      "step": 117
+    },
+    {
+      "epoch": 0.016515045486354095,
+      "grad_norm": 0.42674207968546857,
+      "learning_rate": 2.72093023255814e-05,
+      "loss": 0.8974,
+      "step": 118
+    },
+    {
+      "epoch": 0.016655003498950313,
+      "grad_norm": 0.46006604179365407,
+      "learning_rate": 2.744186046511628e-05,
+      "loss": 0.9413,
+      "step": 119
+    },
+    {
+      "epoch": 0.016794961511546535,
+      "grad_norm": 0.41235716420572693,
+      "learning_rate": 2.7674418604651166e-05,
+      "loss": 0.8682,
+      "step": 120
+    },
+    {
+      "epoch": 0.016934919524142757,
+      "grad_norm": 0.41352483122836786,
+      "learning_rate": 2.7906976744186048e-05,
+      "loss": 0.8926,
+      "step": 121
+    },
+    {
+      "epoch": 0.01707487753673898,
+      "grad_norm": 0.4135656272062978,
+      "learning_rate": 2.813953488372093e-05,
+      "loss": 0.8664,
+      "step": 122
+    },
+    {
+      "epoch": 0.0172148355493352,
+      "grad_norm": 0.40087350807045763,
+      "learning_rate": 2.8372093023255815e-05,
+      "loss": 0.8399,
+      "step": 123
+    },
+    {
+      "epoch": 0.017354793561931422,
+      "grad_norm": 0.4308300834648659,
+      "learning_rate": 2.8604651162790696e-05,
+      "loss": 0.9084,
+      "step": 124
+    },
+    {
+      "epoch": 0.01749475157452764,
+      "grad_norm": 0.48776022790516643,
+      "learning_rate": 2.8837209302325585e-05,
+      "loss": 0.911,
+      "step": 125
+    },
+    {
+      "epoch": 0.017634709587123862,
+      "grad_norm": 0.4111163096942089,
+      "learning_rate": 2.9069767441860467e-05,
+      "loss": 0.8848,
+      "step": 126
+    },
+    {
+      "epoch": 0.017774667599720084,
+      "grad_norm": 0.4381509329328657,
+      "learning_rate": 2.9302325581395352e-05,
+      "loss": 0.8876,
+      "step": 127
+    },
+    {
+      "epoch": 0.017914625612316305,
+      "grad_norm": 0.4459710226095744,
+      "learning_rate": 2.9534883720930234e-05,
+      "loss": 0.914,
+      "step": 128
+    },
+    {
+      "epoch": 0.018054583624912527,
+      "grad_norm": 0.7227835408304066,
+      "learning_rate": 2.976744186046512e-05,
+      "loss": 0.8587,
+      "step": 129
+    },
+    {
+      "epoch": 0.01819454163750875,
+      "grad_norm": 0.43926461194413513,
+      "learning_rate": 3e-05,
+      "loss": 0.8779,
+      "step": 130
+    },
+    {
+      "epoch": 0.018334499650104967,
+      "grad_norm": 0.44123229572799033,
+      "learning_rate": 3.0232558139534883e-05,
+      "loss": 0.8371,
+      "step": 131
+    },
+    {
+      "epoch": 0.01847445766270119,
+      "grad_norm": 0.5044627843039927,
+      "learning_rate": 3.0465116279069768e-05,
+      "loss": 0.8745,
+      "step": 132
+    },
+    {
+      "epoch": 0.01861441567529741,
+      "grad_norm": 0.4802286671204574,
+      "learning_rate": 3.0697674418604656e-05,
+      "loss": 0.8163,
+      "step": 133
+    },
+    {
+      "epoch": 0.018754373687893632,
+      "grad_norm": 0.4753543186959038,
+      "learning_rate": 3.093023255813954e-05,
+      "loss": 0.8499,
+      "step": 134
+    },
+    {
+      "epoch": 0.018894331700489854,
+      "grad_norm": 0.4957303847669158,
+      "learning_rate": 3.116279069767442e-05,
+      "loss": 0.8485,
+      "step": 135
+    },
+    {
+      "epoch": 0.019034289713086076,
+      "grad_norm": 0.4550945161127889,
+      "learning_rate": 3.13953488372093e-05,
+      "loss": 0.7769,
+      "step": 136
+    },
+    {
+      "epoch": 0.019174247725682294,
+      "grad_norm": 0.6087163747209936,
+      "learning_rate": 3.162790697674419e-05,
+      "loss": 0.8363,
+      "step": 137
+    },
+    {
+      "epoch": 0.019314205738278516,
+      "grad_norm": 0.500988900697958,
+      "learning_rate": 3.186046511627907e-05,
+      "loss": 0.8602,
+      "step": 138
+    },
+    {
+      "epoch": 0.019454163750874737,
+      "grad_norm": 0.5444272083147511,
+      "learning_rate": 3.2093023255813954e-05,
+      "loss": 0.8407,
+      "step": 139
+    },
+    {
+      "epoch": 0.01959412176347096,
+      "grad_norm": 0.5954883412818768,
+      "learning_rate": 3.2325581395348836e-05,
+      "loss": 0.8497,
+      "step": 140
+    },
+    {
+      "epoch": 0.01973407977606718,
+      "grad_norm": 0.5375949124547408,
+      "learning_rate": 3.2558139534883724e-05,
+      "loss": 0.8361,
+      "step": 141
+    },
+    {
+      "epoch": 0.019874037788663403,
+      "grad_norm": 0.5552189636185006,
+      "learning_rate": 3.2790697674418606e-05,
+      "loss": 0.8342,
+      "step": 142
+    },
+    {
+      "epoch": 0.02001399580125962,
+      "grad_norm": 0.5148019418462507,
+      "learning_rate": 3.302325581395349e-05,
+      "loss": 0.8261,
+      "step": 143
+    },
+    {
+      "epoch": 0.020153953813855843,
+      "grad_norm": 0.5700734210055411,
+      "learning_rate": 3.3255813953488377e-05,
+      "loss": 0.8323,
+      "step": 144
+    },
+    {
+      "epoch": 0.020293911826452064,
+      "grad_norm": 0.48554968620911604,
+      "learning_rate": 3.348837209302326e-05,
+      "loss": 0.8401,
+      "step": 145
+    },
+    {
+      "epoch": 0.020433869839048286,
+      "grad_norm": 0.46690708567053424,
+      "learning_rate": 3.372093023255814e-05,
+      "loss": 0.832,
+      "step": 146
+    },
+    {
+      "epoch": 0.020573827851644508,
+      "grad_norm": 0.4830729701490898,
+      "learning_rate": 3.395348837209302e-05,
+      "loss": 0.8425,
+      "step": 147
+    },
+    {
+      "epoch": 0.02071378586424073,
+      "grad_norm": 0.48191860044948637,
+      "learning_rate": 3.4186046511627904e-05,
+      "loss": 0.8274,
+      "step": 148
+    },
+    {
+      "epoch": 0.020853743876836948,
+      "grad_norm": 0.5142113223620266,
+      "learning_rate": 3.441860465116279e-05,
+      "loss": 0.8972,
+      "step": 149
+    },
+    {
+      "epoch": 0.02099370188943317,
+      "grad_norm": 0.4790668173486759,
+      "learning_rate": 3.465116279069768e-05,
+      "loss": 0.8315,
+      "step": 150
+    },
+    {
+      "epoch": 0.02113365990202939,
+      "grad_norm": 0.46716122441432334,
+      "learning_rate": 3.488372093023256e-05,
+      "loss": 0.8027,
+      "step": 151
+    },
+    {
+      "epoch": 0.021273617914625613,
+      "grad_norm": 0.5138958015599178,
+      "learning_rate": 3.5116279069767445e-05,
+      "loss": 0.7866,
+      "step": 152
+    },
+    {
+      "epoch": 0.021413575927221835,
+      "grad_norm": 0.5846032834190681,
+      "learning_rate": 3.5348837209302326e-05,
+      "loss": 0.7656,
+      "step": 153
+    },
+    {
+      "epoch": 0.021553533939818053,
+      "grad_norm": 0.4931365055890448,
+      "learning_rate": 3.558139534883721e-05,
+      "loss": 0.7875,
+      "step": 154
+    },
+    {
+      "epoch": 0.021693491952414275,
+      "grad_norm": 0.49665243607020065,
+      "learning_rate": 3.581395348837209e-05,
+      "loss": 0.8285,
+      "step": 155
+    },
+    {
+      "epoch": 0.021833449965010496,
+      "grad_norm": 0.4851409756713546,
+      "learning_rate": 3.604651162790698e-05,
+      "loss": 0.8395,
+      "step": 156
+    },
+    {
+      "epoch": 0.021973407977606718,
+      "grad_norm": 0.4915230656047339,
+      "learning_rate": 3.627906976744187e-05,
+      "loss": 0.8381,
+      "step": 157
+    },
+    {
+      "epoch": 0.02211336599020294,
+      "grad_norm": 0.4892154916147055,
+      "learning_rate": 3.651162790697675e-05,
+      "loss": 0.8372,
+      "step": 158
+    },
+    {
+      "epoch": 0.02225332400279916,
+      "grad_norm": 0.48541705485781506,
+      "learning_rate": 3.674418604651163e-05,
+      "loss": 0.8512,
+      "step": 159
+    },
+    {
+      "epoch": 0.02239328201539538,
+      "grad_norm": 0.490142048396627,
+      "learning_rate": 3.697674418604651e-05,
+      "loss": 0.8185,
+      "step": 160
+    },
+    {
+      "epoch": 0.0225332400279916,
+      "grad_norm": 0.5674374097426846,
+      "learning_rate": 3.7209302325581394e-05,
+      "loss": 0.8101,
+      "step": 161
+    },
+    {
+      "epoch": 0.022673198040587823,
+      "grad_norm": 0.617554041890815,
+      "learning_rate": 3.7441860465116276e-05,
+      "loss": 0.8273,
+      "step": 162
+    },
+    {
+      "epoch": 0.022813156053184045,
+      "grad_norm": 0.5446295352802258,
+      "learning_rate": 3.7674418604651165e-05,
+      "loss": 0.867,
+      "step": 163
+    },
+    {
+      "epoch": 0.022953114065780267,
+      "grad_norm": 0.5370506605400496,
+      "learning_rate": 3.790697674418605e-05,
+      "loss": 0.7857,
+      "step": 164
+    },
+    {
+      "epoch": 0.02309307207837649,
+      "grad_norm": 0.5011901313523426,
+      "learning_rate": 3.8139534883720935e-05,
+      "loss": 0.7731,
+      "step": 165
+    },
+    {
+      "epoch": 0.023233030090972707,
+      "grad_norm": 0.5329852231835671,
+      "learning_rate": 3.837209302325582e-05,
+      "loss": 0.7844,
+      "step": 166
+    },
+    {
+      "epoch": 0.02337298810356893,
+      "grad_norm": 0.5409808244274257,
+      "learning_rate": 3.86046511627907e-05,
+      "loss": 0.8565,
+      "step": 167
+    },
+    {
+      "epoch": 0.02351294611616515,
+      "grad_norm": 0.4698618846754603,
+      "learning_rate": 3.883720930232558e-05,
+      "loss": 0.8027,
+      "step": 168
+    },
+    {
+      "epoch": 0.023652904128761372,
+      "grad_norm": 0.5265079885633662,
+      "learning_rate": 3.906976744186047e-05,
+      "loss": 0.7934,
+      "step": 169
+    },
+    {
+      "epoch": 0.023792862141357594,
+      "grad_norm": 0.5066786460553115,
+      "learning_rate": 3.930232558139535e-05,
+      "loss": 0.8472,
+      "step": 170
+    },
+    {
+      "epoch": 0.023932820153953815,
+      "grad_norm": 0.4895093560469156,
+      "learning_rate": 3.953488372093023e-05,
+      "loss": 0.8038,
+      "step": 171
+    },
+    {
+      "epoch": 0.024072778166550034,
+      "grad_norm": 0.4722527785641036,
+      "learning_rate": 3.9767441860465115e-05,
+      "loss": 0.7926,
+      "step": 172
+    },
+    {
+      "epoch": 0.024212736179146255,
+      "grad_norm": 0.5273651829904968,
+      "learning_rate": 4e-05,
+      "loss": 0.8107,
+      "step": 173
+    },
+    {
+      "epoch": 0.024352694191742477,
+      "grad_norm": 0.5381629057966655,
+      "learning_rate": 4.0232558139534885e-05,
+      "loss": 0.7478,
+      "step": 174
+    },
+    {
+      "epoch": 0.0244926522043387,
+      "grad_norm": 0.5109213734104476,
+      "learning_rate": 4.046511627906977e-05,
+      "loss": 0.7436,
+      "step": 175
+    },
+    {
+      "epoch": 0.02463261021693492,
+      "grad_norm": 0.5144280803181732,
+      "learning_rate": 4.0697674418604655e-05,
+      "loss": 0.8024,
+      "step": 176
+    },
+    {
+      "epoch": 0.024772568229531142,
+      "grad_norm": 0.5119285784380162,
+      "learning_rate": 4.093023255813954e-05,
+      "loss": 0.8339,
+      "step": 177
+    },
+    {
+      "epoch": 0.02491252624212736,
+      "grad_norm": 0.5343750927263191,
+      "learning_rate": 4.116279069767442e-05,
+      "loss": 0.7744,
+      "step": 178
+    },
+    {
+      "epoch": 0.025052484254723582,
+      "grad_norm": 0.534909362676503,
+      "learning_rate": 4.13953488372093e-05,
+      "loss": 0.7942,
+      "step": 179
+    },
+    {
+      "epoch": 0.025192442267319804,
+      "grad_norm": 0.5393369737565247,
+      "learning_rate": 4.162790697674418e-05,
+      "loss": 0.7677,
+      "step": 180
+    },
+    {
+      "epoch": 0.025332400279916026,
+      "grad_norm": 0.4985423936875566,
+      "learning_rate": 4.186046511627907e-05,
+      "loss": 0.8088,
+      "step": 181
+    },
+    {
+      "epoch": 0.025472358292512248,
+      "grad_norm": 0.5167443499358275,
+      "learning_rate": 4.209302325581396e-05,
+      "loss": 0.7999,
+      "step": 182
+    },
+    {
+      "epoch": 0.025612316305108466,
+      "grad_norm": 0.5170409826954023,
+      "learning_rate": 4.232558139534884e-05,
+      "loss": 0.7601,
+      "step": 183
+    },
+    {
+      "epoch": 0.025752274317704688,
+      "grad_norm": 0.5219678512296154,
+      "learning_rate": 4.2558139534883724e-05,
+      "loss": 0.8032,
+      "step": 184
+    },
+    {
+      "epoch": 0.02589223233030091,
+      "grad_norm": 0.5652005338899992,
+      "learning_rate": 4.2790697674418605e-05,
+      "loss": 0.8096,
+      "step": 185
+    },
+    {
+      "epoch": 0.02603219034289713,
+      "grad_norm": 0.5199050303840674,
+      "learning_rate": 4.302325581395349e-05,
+      "loss": 0.7953,
+      "step": 186
+    },
+    {
+      "epoch": 0.026172148355493353,
+      "grad_norm": 0.5389317228680723,
+      "learning_rate": 4.325581395348837e-05,
+      "loss": 0.7972,
+      "step": 187
+    },
+    {
+      "epoch": 0.026312106368089574,
+      "grad_norm": 0.5496576941977979,
+      "learning_rate": 4.348837209302326e-05,
+      "loss": 0.7948,
+      "step": 188
+    },
+    {
+      "epoch": 0.026452064380685793,
+      "grad_norm": 0.4991575270339626,
+      "learning_rate": 4.3720930232558146e-05,
+      "loss": 0.7805,
+      "step": 189
+    },
+    {
+      "epoch": 0.026592022393282014,
+      "grad_norm": 0.5848638237399637,
+      "learning_rate": 4.395348837209303e-05,
+      "loss": 0.8075,
+      "step": 190
+    },
+    {
+      "epoch": 0.026731980405878236,
+      "grad_norm": 0.5521242621533461,
+      "learning_rate": 4.418604651162791e-05,
+      "loss": 0.8142,
+      "step": 191
+    },
+    {
+      "epoch": 0.026871938418474458,
+      "grad_norm": 0.5435127966309655,
+      "learning_rate": 4.441860465116279e-05,
+      "loss": 0.7825,
+      "step": 192
+    },
+    {
+      "epoch": 0.02701189643107068,
+      "grad_norm": 0.5470711522807221,
+      "learning_rate": 4.465116279069767e-05,
+      "loss": 0.7823,
+      "step": 193
+    },
+    {
+      "epoch": 0.0271518544436669,
+      "grad_norm": 0.5527490475528084,
+      "learning_rate": 4.488372093023256e-05,
+      "loss": 0.7346,
+      "step": 194
+    },
+    {
+      "epoch": 0.02729181245626312,
+      "grad_norm": 0.5440674129311732,
+      "learning_rate": 4.5116279069767444e-05,
+      "loss": 0.7794,
+      "step": 195
+    },
+    {
+      "epoch": 0.02743177046885934,
+      "grad_norm": 0.5470033960615739,
+      "learning_rate": 4.5348837209302326e-05,
+      "loss": 0.7778,
+      "step": 196
+    },
+    {
+      "epoch": 0.027571728481455563,
+      "grad_norm": 0.5297715369487911,
+      "learning_rate": 4.5581395348837214e-05,
+      "loss": 0.7817,
+      "step": 197
+    },
+    {
+      "epoch": 0.027711686494051785,
+      "grad_norm": 0.5904204687588617,
+      "learning_rate": 4.5813953488372096e-05,
+      "loss": 0.741,
+      "step": 198
+    },
+    {
+      "epoch": 0.027851644506648007,
+      "grad_norm": 0.5344813639779119,
+      "learning_rate": 4.604651162790698e-05,
+      "loss": 0.7695,
+      "step": 199
+    },
+    {
+      "epoch": 0.02799160251924423,
+      "grad_norm": 0.5332713262025963,
+      "learning_rate": 4.627906976744186e-05,
+      "loss": 0.7648,
+      "step": 200
+    },
+    {
+      "epoch": 0.028131560531840447,
+      "grad_norm": 0.5503858696857928,
+      "learning_rate": 4.651162790697675e-05,
+      "loss": 0.7835,
+      "step": 201
+    },
+    {
+      "epoch": 0.02827151854443667,
+      "grad_norm": 0.5358241040816709,
+      "learning_rate": 4.674418604651163e-05,
+      "loss": 0.8045,
+      "step": 202
+    },
+    {
+      "epoch": 0.02841147655703289,
+      "grad_norm": 0.5486918473034686,
+      "learning_rate": 4.697674418604651e-05,
+      "loss": 0.7938,
+      "step": 203
+    },
+    {
+      "epoch": 0.028551434569629112,
+      "grad_norm": 0.5435329407519798,
+      "learning_rate": 4.7209302325581394e-05,
+      "loss": 0.8122,
+      "step": 204
+    },
+    {
+      "epoch": 0.028691392582225334,
+      "grad_norm": 0.5353685402018145,
+      "learning_rate": 4.744186046511628e-05,
+      "loss": 0.7831,
+      "step": 205
+    },
+    {
+      "epoch": 0.028831350594821555,
+      "grad_norm": 0.5707995107234539,
+      "learning_rate": 4.7674418604651164e-05,
+      "loss": 0.7409,
+      "step": 206
+    },
+    {
+      "epoch": 0.028971308607417774,
+      "grad_norm": 0.5604394624011331,
+      "learning_rate": 4.790697674418605e-05,
+      "loss": 0.7517,
+      "step": 207
+    },
+    {
+      "epoch": 0.029111266620013995,
+      "grad_norm": 0.5515428854045646,
+      "learning_rate": 4.8139534883720934e-05,
+      "loss": 0.7812,
+      "step": 208
+    },
+    {
+      "epoch": 0.029251224632610217,
+      "grad_norm": 0.4865993670607851,
+      "learning_rate": 4.8372093023255816e-05,
+      "loss": 0.7767,
+      "step": 209
+    },
+    {
+      "epoch": 0.02939118264520644,
+      "grad_norm": 0.5336012601695919,
+      "learning_rate": 4.86046511627907e-05,
+      "loss": 0.7463,
+      "step": 210
+    },
+    {
+      "epoch": 0.02953114065780266,
+      "grad_norm": 0.5759753656135316,
+      "learning_rate": 4.883720930232558e-05,
+      "loss": 0.7899,
+      "step": 211
+    },
+    {
+      "epoch": 0.02967109867039888,
+      "grad_norm": 0.6037700025251694,
+      "learning_rate": 4.906976744186046e-05,
+      "loss": 0.7493,
+      "step": 212
+    },
+    {
+      "epoch": 0.0298110566829951,
+      "grad_norm": 0.5451464405404893,
+      "learning_rate": 4.930232558139535e-05,
+      "loss": 0.7501,
+      "step": 213
+    },
+    {
+      "epoch": 0.029951014695591322,
+      "grad_norm": 0.5084395290000155,
+      "learning_rate": 4.953488372093024e-05,
+      "loss": 0.8,
+      "step": 214
+    },
+    {
+      "epoch": 0.030090972708187544,
+      "grad_norm": 0.5375968031425802,
+      "learning_rate": 4.976744186046512e-05,
+      "loss": 0.7644,
+      "step": 215
+    },
+    {
+      "epoch": 0.030230930720783766,
+      "grad_norm": 0.5232124591965054,
+      "learning_rate": 5e-05,
+      "loss": 0.7661,
+      "step": 216
+    },
+    {
+      "epoch": 0.030370888733379987,
+      "grad_norm": 0.53168165514151,
+      "learning_rate": 4.999999743112317e-05,
+      "loss": 0.7469,
+      "step": 217
+    },
+    {
+      "epoch": 0.030510846745976206,
+      "grad_norm": 0.5682427706170695,
+      "learning_rate": 4.9999989724493205e-05,
+      "loss": 0.723,
+      "step": 218
+    },
+    {
+      "epoch": 0.030650804758572427,
+      "grad_norm": 0.5709911717094754,
+      "learning_rate": 4.99999768801117e-05,
+      "loss": 0.8116,
+      "step": 219
+    },
+    {
+      "epoch": 0.03079076277116865,
+      "grad_norm": 0.5259086859276223,
+      "learning_rate": 4.999995889798127e-05,
+      "loss": 0.7802,
+      "step": 220
+    },
+    {
+      "epoch": 0.03093072078376487,
+      "grad_norm": 0.5855912789813319,
+      "learning_rate": 4.999993577810563e-05,
+      "loss": 0.7523,
+      "step": 221
+    },
+    {
+      "epoch": 0.031070678796361093,
+      "grad_norm": 0.5703052221422155,
+      "learning_rate": 4.999990752048953e-05,
+      "loss": 0.7753,
+      "step": 222
+    },
+    {
+      "epoch": 0.031210636808957314,
+      "grad_norm": 0.5108767700185519,
+      "learning_rate": 4.999987412513878e-05,
+      "loss": 0.7634,
+      "step": 223
+    },
+    {
+      "epoch": 0.031350594821553536,
+      "grad_norm": 0.5527712966540788,
+      "learning_rate": 4.999983559206023e-05,
+      "loss": 0.7776,
+      "step": 224
+    },
+    {
+      "epoch": 0.031490552834149754,
+      "grad_norm": 0.5490103596581906,
+      "learning_rate": 4.999979192126181e-05,
+      "loss": 0.8043,
+      "step": 225
+    },
+    {
+      "epoch": 0.03163051084674598,
+      "grad_norm": 0.5298825659589008,
+      "learning_rate": 4.9999743112752485e-05,
+      "loss": 0.7873,
+      "step": 226
+    },
+    {
+      "epoch": 0.0317704688593422,
+      "grad_norm": 0.5312004706999939,
+      "learning_rate": 4.9999689166542295e-05,
+      "loss": 0.7243,
+      "step": 227
+    },
+    {
+      "epoch": 0.031910426871938416,
+      "grad_norm": 0.5282314986009607,
+      "learning_rate": 4.9999630082642325e-05,
+      "loss": 0.7747,
+      "step": 228
+    },
+    {
+      "epoch": 0.03205038488453464,
+      "grad_norm": 0.5605656711144292,
+      "learning_rate": 4.999956586106472e-05,
+      "loss": 0.8079,
+      "step": 229
+    },
+    {
+      "epoch": 0.03219034289713086,
+      "grad_norm": 0.5457865205370973,
+      "learning_rate": 4.999949650182266e-05,
+      "loss": 0.7413,
+      "step": 230
+    },
+    {
+      "epoch": 0.032330300909727085,
+      "grad_norm": 0.5870565643730205,
+      "learning_rate": 4.999942200493043e-05,
+      "loss": 0.696,
+      "step": 231
+    },
+    {
+      "epoch": 0.0324702589223233,
+      "grad_norm": 0.5334906454266579,
+      "learning_rate": 4.9999342370403316e-05,
+      "loss": 0.7259,
+      "step": 232
+    },
+    {
+      "epoch": 0.03261021693491952,
+      "grad_norm": 0.5740266159557987,
+      "learning_rate": 4.999925759825768e-05,
+      "loss": 0.7221,
+      "step": 233
+    },
+    {
+      "epoch": 0.032750174947515746,
+      "grad_norm": 0.613668330354887,
+      "learning_rate": 4.999916768851096e-05,
+      "loss": 0.7756,
+      "step": 234
+    },
+    {
+      "epoch": 0.032890132960111965,
+      "grad_norm": 0.5903235415127636,
+      "learning_rate": 4.999907264118163e-05,
+      "loss": 0.7386,
+      "step": 235
+    },
+    {
+      "epoch": 0.03303009097270819,
+      "grad_norm": 0.534428463431133,
+      "learning_rate": 4.9998972456289226e-05,
+      "loss": 0.7299,
+      "step": 236
+    },
+    {
+      "epoch": 0.03317004898530441,
+      "grad_norm": 0.5238837539032761,
+      "learning_rate": 4.999886713385432e-05,
+      "loss": 0.7432,
+      "step": 237
+    },
+    {
+      "epoch": 0.033310006997900626,
+      "grad_norm": 0.5371791485677779,
+      "learning_rate": 4.999875667389858e-05,
+      "loss": 0.7918,
+      "step": 238
+    },
+    {
+      "epoch": 0.03344996501049685,
+      "grad_norm": 0.5835373561208981,
+      "learning_rate": 4.999864107644469e-05,
+      "loss": 0.7534,
+      "step": 239
+    },
+    {
+      "epoch": 0.03358992302309307,
+      "grad_norm": 0.552016602574312,
+      "learning_rate": 4.999852034151641e-05,
+      "loss": 0.6981,
+      "step": 240
+    },
+    {
+      "epoch": 0.033729881035689295,
+      "grad_norm": 0.5154017381634809,
+      "learning_rate": 4.999839446913855e-05,
+      "loss": 0.6855,
+      "step": 241
+    },
+    {
+      "epoch": 0.03386983904828551,
+      "grad_norm": 0.567904246402228,
+      "learning_rate": 4.999826345933699e-05,
+      "loss": 0.7559,
+      "step": 242
+    },
+    {
+      "epoch": 0.03400979706088174,
+      "grad_norm": 0.5334580095149974,
+      "learning_rate": 4.999812731213864e-05,
+      "loss": 0.7427,
+      "step": 243
+    },
+    {
+      "epoch": 0.03414975507347796,
+      "grad_norm": 0.5705684901629905,
+      "learning_rate": 4.9997986027571485e-05,
+      "loss": 0.7097,
+      "step": 244
+    },
+    {
+      "epoch": 0.034289713086074175,
+      "grad_norm": 0.5308751498286732,
+      "learning_rate": 4.9997839605664564e-05,
+      "loss": 0.7885,
+      "step": 245
+    },
+    {
+      "epoch": 0.0344296710986704,
+      "grad_norm": 0.5655649609764111,
+      "learning_rate": 4.999768804644796e-05,
+      "loss": 0.7829,
+      "step": 246
+    },
+    {
+      "epoch": 0.03456962911126662,
+      "grad_norm": 0.5251341000071817,
+      "learning_rate": 4.999753134995283e-05,
+      "loss": 0.7406,
+      "step": 247
+    },
+    {
+      "epoch": 0.034709587123862844,
+      "grad_norm": 0.5176591391474799,
+      "learning_rate": 4.999736951621137e-05,
+      "loss": 0.7466,
+      "step": 248
+    },
+    {
+      "epoch": 0.03484954513645906,
+      "grad_norm": 0.510003241623571,
+      "learning_rate": 4.999720254525684e-05,
+      "loss": 0.7625,
+      "step": 249
+    },
+    {
+      "epoch": 0.03498950314905528,
+      "grad_norm": 0.5472293188653408,
+      "learning_rate": 4.999703043712355e-05,
+      "loss": 0.7486,
+      "step": 250
+    },
+    {
+      "epoch": 0.035129461161651505,
+      "grad_norm": 0.513397158699985,
+      "learning_rate": 4.9996853191846885e-05,
+      "loss": 0.7676,
+      "step": 251
+    },
+    {
+      "epoch": 0.035269419174247724,
+      "grad_norm": 0.5227735996669705,
+      "learning_rate": 4.999667080946324e-05,
+      "loss": 0.7421,
+      "step": 252
+    },
+    {
+      "epoch": 0.03540937718684395,
+      "grad_norm": 0.5706154309444754,
+      "learning_rate": 4.999648329001013e-05,
+      "loss": 0.7591,
+      "step": 253
+    },
+    {
+      "epoch": 0.03554933519944017,
+      "grad_norm": 0.5258959198504635,
+      "learning_rate": 4.999629063352608e-05,
+      "loss": 0.7478,
+      "step": 254
+    },
+    {
+      "epoch": 0.03568929321203639,
+      "grad_norm": 0.5487756790231113,
+      "learning_rate": 4.999609284005068e-05,
+      "loss": 0.7605,
+      "step": 255
+    },
+    {
+      "epoch": 0.03582925122463261,
+      "grad_norm": 0.5536825461384619,
+      "learning_rate": 4.999588990962458e-05,
+      "loss": 0.743,
+      "step": 256
+    },
+    {
+      "epoch": 0.03596920923722883,
+      "grad_norm": 0.5272600123907261,
+      "learning_rate": 4.9995681842289476e-05,
+      "loss": 0.6928,
+      "step": 257
+    },
+    {
+      "epoch": 0.036109167249825054,
+      "grad_norm": 0.5932574229764146,
+      "learning_rate": 4.999546863808815e-05,
+      "loss": 0.7398,
+      "step": 258
+    },
+    {
+      "epoch": 0.03624912526242127,
+      "grad_norm": 0.5318314385969402,
+      "learning_rate": 4.999525029706439e-05,
+      "loss": 0.7392,
+      "step": 259
+    },
+    {
+      "epoch": 0.0363890832750175,
+      "grad_norm": 0.548901725278328,
+      "learning_rate": 4.999502681926309e-05,
+      "loss": 0.7501,
+      "step": 260
+    },
+    {
+      "epoch": 0.036529041287613716,
+      "grad_norm": 0.5288753436227192,
+      "learning_rate": 4.9994798204730166e-05,
+      "loss": 0.6955,
+      "step": 261
+    },
+    {
+      "epoch": 0.036668999300209934,
+      "grad_norm": 0.5694503048772611,
+      "learning_rate": 4.99945644535126e-05,
+      "loss": 0.723,
+      "step": 262
+    },
+    {
+      "epoch": 0.03680895731280616,
+      "grad_norm": 0.5665031858281824,
+      "learning_rate": 4.999432556565843e-05,
+      "loss": 0.6754,
+      "step": 263
+    },
+    {
+      "epoch": 0.03694891532540238,
+      "grad_norm": 0.5647585870166252,
+      "learning_rate": 4.999408154121676e-05,
+      "loss": 0.7434,
+      "step": 264
+    },
+    {
+      "epoch": 0.0370888733379986,
+      "grad_norm": 0.5198420793446112,
+      "learning_rate": 4.9993832380237735e-05,
+      "loss": 0.7398,
+      "step": 265
+    },
+    {
+      "epoch": 0.03722883135059482,
+      "grad_norm": 0.5250695619803318,
+      "learning_rate": 4.999357808277255e-05,
+      "loss": 0.7355,
+      "step": 266
+    },
+    {
+      "epoch": 0.03736878936319104,
+      "grad_norm": 0.582781926595669,
+      "learning_rate": 4.999331864887347e-05,
+      "loss": 0.794,
+      "step": 267
+    },
+    {
+      "epoch": 0.037508747375787264,
+      "grad_norm": 0.5564844235496481,
+      "learning_rate": 4.9993054078593824e-05,
+      "loss": 0.7205,
+      "step": 268
+    },
+    {
+      "epoch": 0.03764870538838348,
+      "grad_norm": 0.5165771944533392,
+      "learning_rate": 4.9992784371987966e-05,
+      "loss": 0.723,
+      "step": 269
+    },
+    {
+      "epoch": 0.03778866340097971,
+      "grad_norm": 0.528823368400942,
+      "learning_rate": 4.999250952911133e-05,
+      "loss": 0.752,
+      "step": 270
+    },
+    {
+      "epoch": 0.037928621413575926,
+      "grad_norm": 0.5429179446888814,
+      "learning_rate": 4.999222955002041e-05,
+      "loss": 0.723,
+      "step": 271
+    },
+    {
+      "epoch": 0.03806857942617215,
+      "grad_norm": 0.5587199303859207,
+      "learning_rate": 4.9991944434772734e-05,
+      "loss": 0.7417,
+      "step": 272
+    },
+    {
+      "epoch": 0.03820853743876837,
+      "grad_norm": 0.5640839472568651,
+      "learning_rate": 4.999165418342689e-05,
+      "loss": 0.7231,
+      "step": 273
+    },
+    {
+      "epoch": 0.03834849545136459,
+      "grad_norm": 0.5679111247364553,
+      "learning_rate": 4.9991358796042535e-05,
+      "loss": 0.7469,
+      "step": 274
+    },
+    {
+      "epoch": 0.03848845346396081,
+      "grad_norm": 0.5333389832026874,
+      "learning_rate": 4.999105827268038e-05,
+      "loss": 0.7593,
+      "step": 275
+    },
+    {
+      "epoch": 0.03862841147655703,
+      "grad_norm": 0.576446977047569,
+      "learning_rate": 4.999075261340218e-05,
+      "loss": 0.7442,
+      "step": 276
+    },
+    {
+      "epoch": 0.03876836948915326,
+      "grad_norm": 0.5832905654806,
+      "learning_rate": 4.9990441818270745e-05,
+      "loss": 0.7631,
+      "step": 277
+    },
+    {
+      "epoch": 0.038908327501749475,
+      "grad_norm": 0.5606985999591908,
+      "learning_rate": 4.9990125887349956e-05,
+      "loss": 0.7525,
+      "step": 278
+    },
+    {
+      "epoch": 0.03904828551434569,
+      "grad_norm": 0.5397877770051253,
+      "learning_rate": 4.9989804820704735e-05,
+      "loss": 0.677,
+      "step": 279
+    },
+    {
+      "epoch": 0.03918824352694192,
+      "grad_norm": 0.5715063431844626,
+      "learning_rate": 4.998947861840106e-05,
+      "loss": 0.7539,
+      "step": 280
+    },
+    {
+      "epoch": 0.039328201539538137,
+      "grad_norm": 0.542818827148529,
+      "learning_rate": 4.998914728050598e-05,
+      "loss": 0.7505,
+      "step": 281
+    },
+    {
+      "epoch": 0.03946815955213436,
+      "grad_norm": 0.5465408569180555,
+      "learning_rate": 4.9988810807087584e-05,
+      "loss": 0.7119,
+      "step": 282
+    },
+    {
+      "epoch": 0.03960811756473058,
+      "grad_norm": 0.5093766924471657,
+      "learning_rate": 4.998846919821502e-05,
+      "loss": 0.7062,
+      "step": 283
+    },
+    {
+      "epoch": 0.039748075577326805,
+      "grad_norm": 0.5138282632852486,
+      "learning_rate": 4.998812245395849e-05,
+      "loss": 0.7453,
+      "step": 284
+    },
+    {
+      "epoch": 0.03988803358992302,
+      "grad_norm": 0.49293443848284324,
+      "learning_rate": 4.998777057438926e-05,
+      "loss": 0.7277,
+      "step": 285
+    },
+    {
+      "epoch": 0.04002799160251924,
+      "grad_norm": 0.5162138996089223,
+      "learning_rate": 4.9987413559579636e-05,
+      "loss": 0.7804,
+      "step": 286
+    },
+    {
+      "epoch": 0.04016794961511547,
+      "grad_norm": 0.5357676179971859,
+      "learning_rate": 4.998705140960299e-05,
+      "loss": 0.7399,
+      "step": 287
+    },
+    {
+      "epoch": 0.040307907627711685,
+      "grad_norm": 0.518163900018162,
+      "learning_rate": 4.998668412453374e-05,
+      "loss": 0.6914,
+      "step": 288
+    },
+    {
+      "epoch": 0.04044786564030791,
+      "grad_norm": 0.5334432804565435,
+      "learning_rate": 4.9986311704447395e-05,
+      "loss": 0.7186,
+      "step": 289
+    },
+    {
+      "epoch": 0.04058782365290413,
+      "grad_norm": 0.5466256273934306,
+      "learning_rate": 4.9985934149420466e-05,
+      "loss": 0.7361,
+      "step": 290
+    },
+    {
+      "epoch": 0.04072778166550035,
+      "grad_norm": 0.5299994895086962,
+      "learning_rate": 4.998555145953054e-05,
+      "loss": 0.7223,
+      "step": 291
+    },
+    {
+      "epoch": 0.04086773967809657,
+      "grad_norm": 0.5350513478260928,
+      "learning_rate": 4.998516363485629e-05,
+      "loss": 0.725,
+      "step": 292
+    },
+    {
+      "epoch": 0.04100769769069279,
+      "grad_norm": 0.47179695277188055,
+      "learning_rate": 4.99847706754774e-05,
+      "loss": 0.7441,
+      "step": 293
+    },
+    {
+      "epoch": 0.041147655703289016,
+      "grad_norm": 0.53626273737559,
+      "learning_rate": 4.998437258147462e-05,
+      "loss": 0.7643,
+      "step": 294
+    },
+    {
+      "epoch": 0.041287613715885234,
+      "grad_norm": 0.5542158039437199,
+      "learning_rate": 4.9983969352929786e-05,
+      "loss": 0.6996,
+      "step": 295
+    },
+    {
+      "epoch": 0.04142757172848146,
+      "grad_norm": 0.513139127485963,
+      "learning_rate": 4.9983560989925736e-05,
+      "loss": 0.7176,
+      "step": 296
+    },
+    {
+      "epoch": 0.04156752974107768,
+      "grad_norm": 0.5464627329044695,
+      "learning_rate": 4.9983147492546414e-05,
+      "loss": 0.6829,
+      "step": 297
+    },
+    {
+      "epoch": 0.041707487753673896,
+      "grad_norm": 0.4970498064260649,
+      "learning_rate": 4.9982728860876794e-05,
+      "loss": 0.6962,
+      "step": 298
+    },
+    {
+      "epoch": 0.04184744576627012,
+      "grad_norm": 0.5892603998956238,
+      "learning_rate": 4.998230509500291e-05,
+      "loss": 0.7526,
+      "step": 299
+    },
+    {
+      "epoch": 0.04198740377886634,
+      "grad_norm": 0.5501443462930937,
+      "learning_rate": 4.9981876195011844e-05,
+      "loss": 0.7133,
+      "step": 300
+    },
+    {
+      "epoch": 0.042127361791462564,
+      "grad_norm": 0.5498252845378556,
+      "learning_rate": 4.998144216099174e-05,
+      "loss": 0.7257,
+      "step": 301
+    },
+    {
+      "epoch": 0.04226731980405878,
+      "grad_norm": 0.5513802243572777,
+      "learning_rate": 4.99810029930318e-05,
+      "loss": 0.7403,
+      "step": 302
+    },
+    {
+      "epoch": 0.042407277816655,
+      "grad_norm": 0.5188265748765649,
+      "learning_rate": 4.998055869122228e-05,
+      "loss": 0.7302,
+      "step": 303
+    },
+    {
+      "epoch": 0.042547235829251226,
+      "grad_norm": 0.5540179540190183,
+      "learning_rate": 4.998010925565448e-05,
+      "loss": 0.7517,
+      "step": 304
+    },
+    {
+      "epoch": 0.042687193841847444,
+      "grad_norm": 0.5281059249590544,
+      "learning_rate": 4.9979654686420775e-05,
+      "loss": 0.7317,
+      "step": 305
+    },
+    {
+      "epoch": 0.04282715185444367,
+      "grad_norm": 0.5359394569524711,
+      "learning_rate": 4.997919498361457e-05,
+      "loss": 0.7344,
+      "step": 306
+    },
+    {
+      "epoch": 0.04296710986703989,
+      "grad_norm": 0.5074397919192178,
+      "learning_rate": 4.9978730147330355e-05,
+      "loss": 0.7245,
+      "step": 307
+    },
+    {
+      "epoch": 0.043107067879636106,
+      "grad_norm": 0.49809746671209315,
+      "learning_rate": 4.997826017766364e-05,
+      "loss": 0.7181,
+      "step": 308
+    },
+    {
+      "epoch": 0.04324702589223233,
+      "grad_norm": 0.5593359787877635,
+      "learning_rate": 4.997778507471102e-05,
+      "loss": 0.6895,
+      "step": 309
+    },
+    {
+      "epoch": 0.04338698390482855,
+      "grad_norm": 0.5259196612830013,
+      "learning_rate": 4.997730483857014e-05,
+      "loss": 0.7075,
+      "step": 310
+    },
+    {
+      "epoch": 0.043526941917424775,
+      "grad_norm": 0.5158600291227199,
+      "learning_rate": 4.997681946933967e-05,
+      "loss": 0.7071,
+      "step": 311
+    },
+    {
+      "epoch": 0.04366689993002099,
+      "grad_norm": 0.534021017035884,
+      "learning_rate": 4.997632896711939e-05,
+      "loss": 0.6599,
+      "step": 312
+    },
+    {
+      "epoch": 0.04380685794261722,
+      "grad_norm": 0.5669098102151614,
+      "learning_rate": 4.997583333201008e-05,
+      "loss": 0.6998,
+      "step": 313
+    },
+    {
+      "epoch": 0.043946815955213436,
+      "grad_norm": 0.5277846431661678,
+      "learning_rate": 4.99753325641136e-05,
+      "loss": 0.7232,
+      "step": 314
+    },
+    {
+      "epoch": 0.044086773967809655,
+      "grad_norm": 0.5834444960242834,
+      "learning_rate": 4.997482666353287e-05,
+      "loss": 0.7281,
+      "step": 315
+    },
+    {
+      "epoch": 0.04422673198040588,
+      "grad_norm": 0.5050930819325137,
+      "learning_rate": 4.9974315630371855e-05,
+      "loss": 0.7042,
+      "step": 316
+    },
+    {
+      "epoch": 0.0443666899930021,
+      "grad_norm": 0.5077463810437512,
+      "learning_rate": 4.997379946473557e-05,
+      "loss": 0.7043,
+      "step": 317
+    },
+    {
+      "epoch": 0.04450664800559832,
+      "grad_norm": 0.552209585789916,
+      "learning_rate": 4.9973278166730106e-05,
+      "loss": 0.7197,
+      "step": 318
+    },
+    {
+      "epoch": 0.04464660601819454,
+      "grad_norm": 0.6653079515801595,
+      "learning_rate": 4.99727517364626e-05,
+      "loss": 0.7359,
+      "step": 319
+    },
+    {
+      "epoch": 0.04478656403079076,
+      "grad_norm": 0.5255861840491078,
+      "learning_rate": 4.9972220174041205e-05,
+      "loss": 0.6736,
+      "step": 320
+    },
+    {
+      "epoch": 0.044926522043386985,
+      "grad_norm": 0.5400381711080283,
+      "learning_rate": 4.99716834795752e-05,
+      "loss": 0.7683,
+      "step": 321
+    },
+    {
+      "epoch": 0.0450664800559832,
+      "grad_norm": 0.5297848029054303,
+      "learning_rate": 4.997114165317486e-05,
+      "loss": 0.7252,
+      "step": 322
+    },
+    {
+      "epoch": 0.04520643806857943,
+      "grad_norm": 0.5344509508263574,
+      "learning_rate": 4.997059469495155e-05,
+      "loss": 0.6991,
+      "step": 323
+    },
+    {
+      "epoch": 0.04534639608117565,
+      "grad_norm": 0.5037923428143048,
+      "learning_rate": 4.997004260501766e-05,
+      "loss": 0.7393,
+      "step": 324
+    },
+    {
+      "epoch": 0.04548635409377187,
+      "grad_norm": 0.5338819883439406,
+      "learning_rate": 4.996948538348666e-05,
+      "loss": 0.711,
+      "step": 325
+    },
+    {
+      "epoch": 0.04562631210636809,
+      "grad_norm": 0.542939872517023,
+      "learning_rate": 4.996892303047306e-05,
+      "loss": 0.7,
+      "step": 326
+    },
+    {
+      "epoch": 0.04576627011896431,
+      "grad_norm": 0.5211829437872917,
+      "learning_rate": 4.996835554609244e-05,
+      "loss": 0.7185,
+      "step": 327
+    },
+    {
+      "epoch": 0.045906228131560534,
+      "grad_norm": 0.6756691366488846,
+      "learning_rate": 4.996778293046141e-05,
+      "loss": 0.7205,
+      "step": 328
+    },
+    {
+      "epoch": 0.04604618614415675,
+      "grad_norm": 0.5016052820529759,
+      "learning_rate": 4.996720518369764e-05,
+      "loss": 0.6912,
+      "step": 329
+    },
+    {
+      "epoch": 0.04618614415675298,
+      "grad_norm": 0.5201008063930296,
+      "learning_rate": 4.996662230591989e-05,
+      "loss": 0.7341,
+      "step": 330
+    },
+    {
+      "epoch": 0.046326102169349195,
+      "grad_norm": 0.567491502617686,
+      "learning_rate": 4.996603429724793e-05,
+      "loss": 0.7365,
+      "step": 331
+    },
+    {
+      "epoch": 0.046466060181945414,
+      "grad_norm": 0.5449583987974782,
+      "learning_rate": 4.996544115780261e-05,
+      "loss": 0.7075,
+      "step": 332
+    },
+    {
+      "epoch": 0.04660601819454164,
+      "grad_norm": 0.5492024616709544,
+      "learning_rate": 4.996484288770582e-05,
+      "loss": 0.7896,
+      "step": 333
+    },
+    {
+      "epoch": 0.04674597620713786,
+      "grad_norm": 0.5076286169172769,
+      "learning_rate": 4.996423948708051e-05,
+      "loss": 0.7394,
+      "step": 334
+    },
+    {
+      "epoch": 0.04688593421973408,
+      "grad_norm": 0.550162273407423,
+      "learning_rate": 4.996363095605069e-05,
+      "loss": 0.7049,
+      "step": 335
+    },
+    {
+      "epoch": 0.0470258922323303,
+      "grad_norm": 0.6075217637750909,
+      "learning_rate": 4.9963017294741407e-05,
+      "loss": 0.656,
+      "step": 336
+    },
+    {
+      "epoch": 0.04716585024492652,
+      "grad_norm": 0.5456579494737005,
+      "learning_rate": 4.99623985032788e-05,
+      "loss": 0.6872,
+      "step": 337
+    },
+    {
+      "epoch": 0.047305808257522744,
+      "grad_norm": 0.661719363460849,
+      "learning_rate": 4.996177458179001e-05,
+      "loss": 0.7,
+      "step": 338
+    },
+    {
+      "epoch": 0.04744576627011896,
+      "grad_norm": 0.6073637233523712,
+      "learning_rate": 4.996114553040328e-05,
+      "loss": 0.7034,
+      "step": 339
+    },
+    {
+      "epoch": 0.04758572428271519,
+      "grad_norm": 0.627821201647191,
+      "learning_rate": 4.996051134924786e-05,
+      "loss": 0.6996,
+      "step": 340
+    },
+    {
+      "epoch": 0.047725682295311406,
+      "grad_norm": 0.5557264017341775,
+      "learning_rate": 4.99598720384541e-05,
+      "loss": 0.7071,
+      "step": 341
+    },
+    {
+      "epoch": 0.04786564030790763,
+      "grad_norm": 0.5980687216959104,
+      "learning_rate": 4.995922759815339e-05,
+      "loss": 0.7101,
+      "step": 342
+    },
+    {
+      "epoch": 0.04800559832050385,
+      "grad_norm": 0.6227145204853264,
+      "learning_rate": 4.995857802847816e-05,
+      "loss": 0.7248,
+      "step": 343
+    },
+    {
+      "epoch": 0.04814555633310007,
+      "grad_norm": 0.9190740737396743,
+      "learning_rate": 4.9957923329561907e-05,
+      "loss": 0.709,
+      "step": 344
+    },
+    {
+      "epoch": 0.04828551434569629,
+      "grad_norm": 0.5382237071595833,
+      "learning_rate": 4.9957263501539174e-05,
+      "loss": 0.6756,
+      "step": 345
+    },
+    {
+      "epoch": 0.04842547235829251,
+      "grad_norm": 0.7390793234122391,
+      "learning_rate": 4.9956598544545566e-05,
+      "loss": 0.7184,
+      "step": 346
+    },
+    {
+      "epoch": 0.048565430370888736,
+      "grad_norm": 0.5896661276064007,
+      "learning_rate": 4.9955928458717723e-05,
+      "loss": 0.7022,
+      "step": 347
+    },
+    {
+      "epoch": 0.048705388383484954,
+      "grad_norm": 0.5284578098126113,
+      "learning_rate": 4.9955253244193375e-05,
+      "loss": 0.7238,
+      "step": 348
+    },
+    {
+      "epoch": 0.04884534639608117,
+      "grad_norm": 0.5539013371779543,
+      "learning_rate": 4.9954572901111286e-05,
+      "loss": 0.693,
+      "step": 349
+    },
+    {
+      "epoch": 0.0489853044086774,
+      "grad_norm": 0.7749019171598228,
+      "learning_rate": 4.9953887429611256e-05,
+      "loss": 0.6873,
+      "step": 350
+    },
+    {
+      "epoch": 0.049125262421273616,
+      "grad_norm": 0.5416373773112516,
+      "learning_rate": 4.995319682983418e-05,
+      "loss": 0.7418,
+      "step": 351
+    },
+    {
+      "epoch": 0.04926522043386984,
+      "grad_norm": 0.5962345555842343,
+      "learning_rate": 4.995250110192195e-05,
+      "loss": 0.7048,
+      "step": 352
+    },
+    {
+      "epoch": 0.04940517844646606,
+      "grad_norm": 0.6171440633433287,
+      "learning_rate": 4.995180024601758e-05,
+      "loss": 0.719,
+      "step": 353
+    },
+    {
+      "epoch": 0.049545136459062285,
+      "grad_norm": 0.5613359206707593,
+      "learning_rate": 4.995109426226508e-05,
+      "loss": 0.7374,
+      "step": 354
+    },
+    {
+      "epoch": 0.0496850944716585,
+      "grad_norm": 0.5557877380798822,
+      "learning_rate": 4.995038315080954e-05,
+      "loss": 0.7149,
+      "step": 355
+    },
+    {
+      "epoch": 0.04982505248425472,
+      "grad_norm": 0.5681067506858839,
+      "learning_rate": 4.994966691179711e-05,
+      "loss": 0.74,
+      "step": 356
+    },
+    {
+      "epoch": 0.049965010496850946,
+      "grad_norm": 0.5671390578619896,
+      "learning_rate": 4.994894554537498e-05,
+      "loss": 0.743,
+      "step": 357
+    },
+    {
+      "epoch": 0.050104968509447165,
+      "grad_norm": 0.5040150906914191,
+      "learning_rate": 4.9948219051691394e-05,
+      "loss": 0.7411,
+      "step": 358
+    },
+    {
+      "epoch": 0.05024492652204339,
+      "grad_norm": 0.5000433753838786,
+      "learning_rate": 4.994748743089566e-05,
+      "loss": 0.709,
+      "step": 359
+    },
+    {
+      "epoch": 0.05038488453463961,
+      "grad_norm": 0.5669390690482634,
+      "learning_rate": 4.9946750683138134e-05,
+      "loss": 0.7157,
+      "step": 360
+    },
+    {
+      "epoch": 0.050524842547235826,
+      "grad_norm": 0.5348574517172033,
+      "learning_rate": 4.994600880857022e-05,
+      "loss": 0.6982,
+      "step": 361
+    },
+    {
+      "epoch": 0.05066480055983205,
+      "grad_norm": 0.5519610052382385,
+      "learning_rate": 4.9945261807344376e-05,
+      "loss": 0.701,
+      "step": 362
+    },
+    {
+      "epoch": 0.05080475857242827,
+      "grad_norm": 0.5267779469247078,
+      "learning_rate": 4.994450967961413e-05,
+      "loss": 0.7164,
+      "step": 363
+    },
+    {
+      "epoch": 0.050944716585024495,
+      "grad_norm": 0.5168765165006579,
+      "learning_rate": 4.994375242553405e-05,
+      "loss": 0.7321,
+      "step": 364
+    },
+    {
+      "epoch": 0.05108467459762071,
+      "grad_norm": 0.6298046212170411,
+      "learning_rate": 4.994299004525975e-05,
+      "loss": 0.6847,
+      "step": 365
+    },
+    {
+      "epoch": 0.05122463261021693,
+      "grad_norm": 0.557578502569354,
+      "learning_rate": 4.994222253894791e-05,
+      "loss": 0.7152,
+      "step": 366
+    },
+    {
+      "epoch": 0.05136459062281316,
+      "grad_norm": 0.5335254526570578,
+      "learning_rate": 4.994144990675627e-05,
+      "loss": 0.7247,
+      "step": 367
+    },
+    {
+      "epoch": 0.051504548635409375,
+      "grad_norm": 0.5400373413925982,
+      "learning_rate": 4.99406721488436e-05,
+      "loss": 0.6837,
+      "step": 368
+    },
+    {
+      "epoch": 0.0516445066480056,
+      "grad_norm": 0.4935375590306368,
+      "learning_rate": 4.993988926536975e-05,
+      "loss": 0.6861,
+      "step": 369
+    },
+    {
+      "epoch": 0.05178446466060182,
+      "grad_norm": 0.6275855125475005,
+      "learning_rate": 4.993910125649561e-05,
+      "loss": 0.6829,
+      "step": 370
+    },
+    {
+      "epoch": 0.051924422673198044,
+      "grad_norm": 1.019947219371485,
+      "learning_rate": 4.993830812238311e-05,
+      "loss": 0.7568,
+      "step": 371
+    },
+    {
+      "epoch": 0.05206438068579426,
+      "grad_norm": 0.5311851634325208,
+      "learning_rate": 4.9937509863195256e-05,
+      "loss": 0.6854,
+      "step": 372
+    },
+    {
+      "epoch": 0.05220433869839048,
+      "grad_norm": 0.5202083034030205,
+      "learning_rate": 4.993670647909611e-05,
+      "loss": 0.7297,
+      "step": 373
+    },
+    {
+      "epoch": 0.052344296710986706,
+      "grad_norm": 0.5639019664915923,
+      "learning_rate": 4.9935897970250745e-05,
+      "loss": 0.7471,
+      "step": 374
+    },
+    {
+      "epoch": 0.052484254723582924,
+      "grad_norm": 0.5482982154303914,
+      "learning_rate": 4.993508433682535e-05,
+      "loss": 0.6802,
+      "step": 375
+    },
+    {
+      "epoch": 0.05262421273617915,
+      "grad_norm": 0.576536365667398,
+      "learning_rate": 4.993426557898711e-05,
+      "loss": 0.6996,
+      "step": 376
+    },
+    {
+      "epoch": 0.05276417074877537,
+      "grad_norm": 0.4997211624617034,
+      "learning_rate": 4.993344169690431e-05,
+      "loss": 0.6907,
+      "step": 377
+    },
+    {
+      "epoch": 0.052904128761371585,
+      "grad_norm": 0.4948774132141246,
+      "learning_rate": 4.993261269074625e-05,
+      "loss": 0.6869,
+      "step": 378
+    },
+    {
+      "epoch": 0.05304408677396781,
+      "grad_norm": 0.49620741460648987,
+      "learning_rate": 4.9931778560683304e-05,
+      "loss": 0.7045,
+      "step": 379
+    },
+    {
+      "epoch": 0.05318404478656403,
+      "grad_norm": 0.5687949523084871,
+      "learning_rate": 4.99309393068869e-05,
+      "loss": 0.6876,
+      "step": 380
+    },
+    {
+      "epoch": 0.053324002799160254,
+      "grad_norm": 1.3449476384226542,
+      "learning_rate": 4.9930094929529506e-05,
+      "loss": 0.6932,
+      "step": 381
+    },
+    {
+      "epoch": 0.05346396081175647,
+      "grad_norm": 0.5421720343729313,
+      "learning_rate": 4.992924542878465e-05,
+      "loss": 0.6636,
+      "step": 382
+    },
+    {
+      "epoch": 0.0536039188243527,
+      "grad_norm": 0.5574373007750801,
+      "learning_rate": 4.9928390804826916e-05,
+      "loss": 0.7171,
+      "step": 383
+    },
+    {
+      "epoch": 0.053743876836948916,
+      "grad_norm": 0.5165807349158404,
+      "learning_rate": 4.992753105783194e-05,
+      "loss": 0.6979,
+      "step": 384
+    },
+    {
+      "epoch": 0.053883834849545134,
+      "grad_norm": 0.5207307553346301,
+      "learning_rate": 4.99266661879764e-05,
+      "loss": 0.6768,
+      "step": 385
+    },
+    {
+      "epoch": 0.05402379286214136,
+      "grad_norm": 0.7774982969038746,
+      "learning_rate": 4.9925796195438044e-05,
+      "loss": 0.6667,
+      "step": 386
+    },
+    {
+      "epoch": 0.05416375087473758,
+      "grad_norm": 0.7031427692671837,
+      "learning_rate": 4.992492108039566e-05,
+      "loss": 0.7051,
+      "step": 387
+    },
+    {
+      "epoch": 0.0543037088873338,
+      "grad_norm": 0.5217628043633614,
+      "learning_rate": 4.99240408430291e-05,
+      "loss": 0.6427,
+      "step": 388
+    },
+    {
+      "epoch": 0.05444366689993002,
+      "grad_norm": 0.8333810612085015,
+      "learning_rate": 4.992315548351925e-05,
+      "loss": 0.7245,
+      "step": 389
+    },
+    {
+      "epoch": 0.05458362491252624,
+      "grad_norm": 0.629384361310055,
+      "learning_rate": 4.992226500204808e-05,
+      "loss": 0.6932,
+      "step": 390
+    },
+    {
+      "epoch": 0.054723582925122465,
+      "grad_norm": 1.215348456810311,
+      "learning_rate": 4.992136939879856e-05,
+      "loss": 0.6962,
+      "step": 391
+    },
+    {
+      "epoch": 0.05486354093771868,
+      "grad_norm": 0.5807472413206434,
+      "learning_rate": 4.992046867395478e-05,
+      "loss": 0.7176,
+      "step": 392
+    },
+    {
+      "epoch": 0.05500349895031491,
+      "grad_norm": 0.5439919445778956,
+      "learning_rate": 4.9919562827701824e-05,
+      "loss": 0.716,
+      "step": 393
+    },
+    {
+      "epoch": 0.055143456962911126,
+      "grad_norm": 0.5602331962806683,
+      "learning_rate": 4.9918651860225864e-05,
+      "loss": 0.6776,
+      "step": 394
+    },
+    {
+      "epoch": 0.055283414975507345,
+      "grad_norm": 0.522684609371295,
+      "learning_rate": 4.9917735771714114e-05,
+      "loss": 0.6646,
+      "step": 395
+    },
+    {
+      "epoch": 0.05542337298810357,
+      "grad_norm": 0.6622073701270422,
+      "learning_rate": 4.991681456235483e-05,
+      "loss": 0.6714,
+      "step": 396
+    },
+    {
+      "epoch": 0.05556333100069979,
+      "grad_norm": 0.6427418148993167,
+      "learning_rate": 4.991588823233735e-05,
+      "loss": 0.7056,
+      "step": 397
+    },
+    {
+      "epoch": 0.05570328901329601,
+      "grad_norm": 0.5725098170561623,
+      "learning_rate": 4.991495678185202e-05,
+      "loss": 0.6672,
+      "step": 398
+    },
+    {
+      "epoch": 0.05584324702589223,
+      "grad_norm": 0.5585281717216761,
+      "learning_rate": 4.991402021109027e-05,
+      "loss": 0.7064,
+      "step": 399
+    },
+    {
+      "epoch": 0.05598320503848846,
+      "grad_norm": 1.7745368005124815,
+      "learning_rate": 4.991307852024458e-05,
+      "loss": 0.7246,
+      "step": 400
+    },
+    {
+      "epoch": 0.056123163051084675,
+      "grad_norm": 0.5282756777124027,
+      "learning_rate": 4.991213170950848e-05,
+      "loss": 0.6971,
+      "step": 401
+    },
+    {
+      "epoch": 0.05626312106368089,
+      "grad_norm": 0.7218574920206103,
+      "learning_rate": 4.9911179779076544e-05,
+      "loss": 0.7173,
+      "step": 402
+    },
+    {
+      "epoch": 0.05640307907627712,
+      "grad_norm": 0.5451854604938866,
+      "learning_rate": 4.99102227291444e-05,
+      "loss": 0.7146,
+      "step": 403
+    },
+    {
+      "epoch": 0.05654303708887334,
+      "grad_norm": 0.5389963653898296,
+      "learning_rate": 4.990926055990873e-05,
+      "loss": 0.6745,
+      "step": 404
+    },
+    {
+      "epoch": 0.05668299510146956,
+      "grad_norm": 0.5101195028542534,
+      "learning_rate": 4.9908293271567286e-05,
+      "loss": 0.7158,
+      "step": 405
+    },
+    {
+      "epoch": 0.05682295311406578,
+      "grad_norm": 0.5324266228237025,
+      "learning_rate": 4.990732086431884e-05,
+      "loss": 0.6766,
+      "step": 406
+    },
+    {
+      "epoch": 0.056962911126662,
+      "grad_norm": 0.5056404230280876,
+      "learning_rate": 4.990634333836324e-05,
+      "loss": 0.6859,
+      "step": 407
+    },
+    {
+      "epoch": 0.057102869139258224,
+      "grad_norm": 0.553015466117812,
+      "learning_rate": 4.990536069390136e-05,
+      "loss": 0.7476,
+      "step": 408
+    },
+    {
+      "epoch": 0.05724282715185444,
+      "grad_norm": 0.6468042464648912,
+      "learning_rate": 4.9904372931135167e-05,
+      "loss": 0.6926,
+      "step": 409
+    },
+    {
+      "epoch": 0.05738278516445067,
+      "grad_norm": 0.5169803890591222,
+      "learning_rate": 4.990338005026764e-05,
+      "loss": 0.6948,
+      "step": 410
+    },
+    {
+      "epoch": 0.057522743177046885,
+      "grad_norm": 0.7826976195286862,
+      "learning_rate": 4.990238205150284e-05,
+      "loss": 0.693,
+      "step": 411
+    },
+    {
+      "epoch": 0.05766270118964311,
+      "grad_norm": 0.5898380237885452,
+      "learning_rate": 4.990137893504585e-05,
+      "loss": 0.7025,
+      "step": 412
+    },
+    {
+      "epoch": 0.05780265920223933,
+      "grad_norm": 0.6044914970349515,
+      "learning_rate": 4.990037070110283e-05,
+      "loss": 0.6642,
+      "step": 413
+    },
+    {
+      "epoch": 0.05794261721483555,
+      "grad_norm": 0.6647502597764803,
+      "learning_rate": 4.989935734988098e-05,
+      "loss": 0.7141,
+      "step": 414
+    },
+    {
+      "epoch": 0.05808257522743177,
+      "grad_norm": 0.578032490799264,
+      "learning_rate": 4.989833888158856e-05,
+      "loss": 0.7525,
+      "step": 415
+    },
+    {
+      "epoch": 0.05822253324002799,
+      "grad_norm": 0.5446056431076586,
+      "learning_rate": 4.989731529643486e-05,
+      "loss": 0.7152,
+      "step": 416
+    },
+    {
+      "epoch": 0.058362491252624216,
+      "grad_norm": 0.5520058248905833,
+      "learning_rate": 4.9896286594630255e-05,
+      "loss": 0.7104,
+      "step": 417
+    },
+    {
+      "epoch": 0.058502449265220434,
+      "grad_norm": 0.5358722455002061,
+      "learning_rate": 4.989525277638614e-05,
+      "loss": 0.6877,
+      "step": 418
+    },
+    {
+      "epoch": 0.05864240727781665,
+      "grad_norm": 0.5211574683086443,
+      "learning_rate": 4.989421384191499e-05,
+      "loss": 0.7235,
+      "step": 419
+    },
+    {
+      "epoch": 0.05878236529041288,
+      "grad_norm": 0.5620604980498491,
+      "learning_rate": 4.98931697914303e-05,
+      "loss": 0.6579,
+      "step": 420
+    },
+    {
+      "epoch": 0.058922323303009096,
+      "grad_norm": 0.5591196420226805,
+      "learning_rate": 4.989212062514664e-05,
+      "loss": 0.6848,
+      "step": 421
+    },
+    {
+      "epoch": 0.05906228131560532,
+      "grad_norm": 0.593542823435001,
+      "learning_rate": 4.989106634327963e-05,
+      "loss": 0.6769,
+      "step": 422
+    },
+    {
+      "epoch": 0.05920223932820154,
+      "grad_norm": 0.4993385226927883,
+      "learning_rate": 4.989000694604593e-05,
+      "loss": 0.7045,
+      "step": 423
+    },
+    {
+      "epoch": 0.05934219734079776,
+      "grad_norm": 0.8610343379419788,
+      "learning_rate": 4.9888942433663255e-05,
+      "loss": 0.6875,
+      "step": 424
+    },
+    {
+      "epoch": 0.05948215535339398,
+      "grad_norm": 0.48976014126417133,
+      "learning_rate": 4.988787280635038e-05,
+      "loss": 0.7148,
+      "step": 425
+    },
+    {
+      "epoch": 0.0596221133659902,
+      "grad_norm": 0.4952300036598977,
+      "learning_rate": 4.988679806432712e-05,
+      "loss": 0.6947,
+      "step": 426
+    },
+    {
+      "epoch": 0.059762071378586426,
+      "grad_norm": 0.46474336456699533,
+      "learning_rate": 4.9885718207814335e-05,
+      "loss": 0.6794,
+      "step": 427
+    },
+    {
+      "epoch": 0.059902029391182644,
+      "grad_norm": 0.5147006168979844,
+      "learning_rate": 4.988463323703397e-05,
+      "loss": 0.6711,
+      "step": 428
+    },
+    {
+      "epoch": 0.06004198740377887,
+      "grad_norm": 0.637107834123775,
+      "learning_rate": 4.988354315220898e-05,
+      "loss": 0.6715,
+      "step": 429
+    },
+    {
+      "epoch": 0.06018194541637509,
+      "grad_norm": 0.5511914486791841,
+      "learning_rate": 4.988244795356339e-05,
+      "loss": 0.649,
+      "step": 430
+    },
+    {
+      "epoch": 0.060321903428971306,
+      "grad_norm": 0.4830610595446907,
+      "learning_rate": 4.9881347641322277e-05,
+      "loss": 0.6591,
+      "step": 431
+    },
+    {
+      "epoch": 0.06046186144156753,
+      "grad_norm": 0.5705178740705275,
+      "learning_rate": 4.988024221571177e-05,
+      "loss": 0.6686,
+      "step": 432
+    },
+    {
+      "epoch": 0.06060181945416375,
+      "grad_norm": 0.7215515180250442,
+      "learning_rate": 4.987913167695904e-05,
+      "loss": 0.7433,
+      "step": 433
+    },
+    {
+      "epoch": 0.060741777466759975,
+      "grad_norm": 0.7060924726627863,
+      "learning_rate": 4.9878016025292305e-05,
+      "loss": 0.6936,
+      "step": 434
+    },
+    {
+      "epoch": 0.06088173547935619,
+      "grad_norm": 0.5381947781056515,
+      "learning_rate": 4.987689526094087e-05,
+      "loss": 0.6751,
+      "step": 435
+    },
+    {
+      "epoch": 0.06102169349195241,
+      "grad_norm": 0.4984642961815355,
+      "learning_rate": 4.987576938413504e-05,
+      "loss": 0.6551,
+      "step": 436
+    },
+    {
+      "epoch": 0.061161651504548636,
+      "grad_norm": 0.6995103632612955,
+      "learning_rate": 4.98746383951062e-05,
+      "loss": 0.6852,
+      "step": 437
+    },
+    {
+      "epoch": 0.061301609517144855,
+      "grad_norm": 0.5321506444103643,
+      "learning_rate": 4.9873502294086785e-05,
+      "loss": 0.7047,
+      "step": 438
+    },
+    {
+      "epoch": 0.06144156752974108,
+      "grad_norm": 0.5373668549791867,
+      "learning_rate": 4.987236108131026e-05,
+      "loss": 0.6476,
+      "step": 439
+    },
+    {
+      "epoch": 0.0615815255423373,
+      "grad_norm": 0.5531880070215052,
+      "learning_rate": 4.9871214757011176e-05,
+      "loss": 0.6695,
+      "step": 440
+    },
+    {
+      "epoch": 0.06172148355493352,
+      "grad_norm": 0.5377302222338248,
+      "learning_rate": 4.9870063321425105e-05,
+      "loss": 0.6522,
+      "step": 441
+    },
+    {
+      "epoch": 0.06186144156752974,
+      "grad_norm": 0.4885102130170266,
+      "learning_rate": 4.986890677478867e-05,
+      "loss": 0.6693,
+      "step": 442
+    },
+    {
+      "epoch": 0.06200139958012596,
+      "grad_norm": 0.5111003799643065,
+      "learning_rate": 4.986774511733957e-05,
+      "loss": 0.6578,
+      "step": 443
+    },
+    {
+      "epoch": 0.062141357592722185,
+      "grad_norm": 0.45757845171169315,
+      "learning_rate": 4.986657834931653e-05,
+      "loss": 0.701,
+      "step": 444
+    },
+    {
+      "epoch": 0.0622813156053184,
+      "grad_norm": 0.6187591152576597,
+      "learning_rate": 4.986540647095933e-05,
+      "loss": 0.671,
+      "step": 445
+    },
+    {
+      "epoch": 0.06242127361791463,
+      "grad_norm": 0.5054481649682342,
+      "learning_rate": 4.9864229482508804e-05,
+      "loss": 0.7186,
+      "step": 446
+    },
+    {
+      "epoch": 0.06256123163051085,
+      "grad_norm": 0.5370871184063281,
+      "learning_rate": 4.9863047384206835e-05,
+      "loss": 0.6889,
+      "step": 447
+    },
+    {
+      "epoch": 0.06270118964310707,
+      "grad_norm": 0.5046075184933217,
+      "learning_rate": 4.986186017629636e-05,
+      "loss": 0.6968,
+      "step": 448
+    },
+    {
+      "epoch": 0.06284114765570328,
+      "grad_norm": 0.5017707016018389,
+      "learning_rate": 4.986066785902136e-05,
+      "loss": 0.7472,
+      "step": 449
+    },
+    {
+      "epoch": 0.06298110566829951,
+      "grad_norm": 0.488992352862207,
+      "learning_rate": 4.985947043262686e-05,
+      "loss": 0.6916,
+      "step": 450
+    },
+    {
+      "epoch": 0.06312106368089573,
+      "grad_norm": 0.5153973631708659,
+      "learning_rate": 4.9858267897358956e-05,
+      "loss": 0.7105,
+      "step": 451
+    },
+    {
+      "epoch": 0.06326102169349196,
+      "grad_norm": 0.5351052897634555,
+      "learning_rate": 4.985706025346477e-05,
+      "loss": 0.7257,
+      "step": 452
+    },
+    {
+      "epoch": 0.06340097970608817,
+      "grad_norm": 0.5382494705510661,
+      "learning_rate": 4.98558475011925e-05,
+      "loss": 0.6587,
+      "step": 453
+    },
+    {
+      "epoch": 0.0635409377186844,
+      "grad_norm": 0.5164087894936398,
+      "learning_rate": 4.985462964079137e-05,
+      "loss": 0.6732,
+      "step": 454
+    },
+    {
+      "epoch": 0.06368089573128062,
+      "grad_norm": 0.5018878284351228,
+      "learning_rate": 4.985340667251166e-05,
+      "loss": 0.6436,
+      "step": 455
+    },
+    {
+      "epoch": 0.06382085374387683,
+      "grad_norm": 0.6387432517988891,
+      "learning_rate": 4.9852178596604705e-05,
+      "loss": 0.6874,
+      "step": 456
+    },
+    {
+      "epoch": 0.06396081175647306,
+      "grad_norm": 0.525361815670854,
+      "learning_rate": 4.985094541332288e-05,
+      "loss": 0.6899,
+      "step": 457
+    },
+    {
+      "epoch": 0.06410076976906928,
+      "grad_norm": 0.5252701657432063,
+      "learning_rate": 4.984970712291963e-05,
+      "loss": 0.6788,
+      "step": 458
+    },
+    {
+      "epoch": 0.0642407277816655,
+      "grad_norm": 0.5268887795524961,
+      "learning_rate": 4.984846372564943e-05,
+      "loss": 0.6915,
+      "step": 459
+    },
+    {
+      "epoch": 0.06438068579426172,
+      "grad_norm": 0.5762015280618955,
+      "learning_rate": 4.9847215221767815e-05,
+      "loss": 0.7023,
+      "step": 460
+    },
+    {
+      "epoch": 0.06452064380685794,
+      "grad_norm": 0.5228367157993173,
+      "learning_rate": 4.984596161153136e-05,
+      "loss": 0.7088,
+      "step": 461
+    },
+    {
+      "epoch": 0.06466060181945417,
+      "grad_norm": 0.5157276830814348,
+      "learning_rate": 4.984470289519769e-05,
+      "loss": 0.7164,
+      "step": 462
+    },
+    {
+      "epoch": 0.06480055983205038,
+      "grad_norm": 0.5405166337013064,
+      "learning_rate": 4.9843439073025486e-05,
+      "loss": 0.7111,
+      "step": 463
+    },
+    {
+      "epoch": 0.0649405178446466,
+      "grad_norm": 0.5812948989732744,
+      "learning_rate": 4.984217014527449e-05,
+      "loss": 0.7043,
+      "step": 464
+    },
+    {
+      "epoch": 0.06508047585724283,
+      "grad_norm": 0.49548874056024383,
+      "learning_rate": 4.984089611220547e-05,
+      "loss": 0.6428,
+      "step": 465
+    },
+    {
+      "epoch": 0.06522043386983904,
+      "grad_norm": 0.5197419283605628,
+      "learning_rate": 4.9839616974080246e-05,
+      "loss": 0.7102,
+      "step": 466
+    },
+    {
+      "epoch": 0.06536039188243527,
+      "grad_norm": 0.6462028524249613,
+      "learning_rate": 4.9838332731161694e-05,
+      "loss": 0.664,
+      "step": 467
+    },
+    {
+      "epoch": 0.06550034989503149,
+      "grad_norm": 0.4883003485107369,
+      "learning_rate": 4.9837043383713753e-05,
+      "loss": 0.6494,
+      "step": 468
+    },
+    {
+      "epoch": 0.06564030790762772,
+      "grad_norm": 0.510996545798241,
+      "learning_rate": 4.983574893200139e-05,
+      "loss": 0.6762,
+      "step": 469
+    },
+    {
+      "epoch": 0.06578026592022393,
+      "grad_norm": 0.5132612521238953,
+      "learning_rate": 4.9834449376290625e-05,
+      "loss": 0.6646,
+      "step": 470
+    },
+    {
+      "epoch": 0.06592022393282015,
+      "grad_norm": 0.5184693525295171,
+      "learning_rate": 4.983314471684853e-05,
+      "loss": 0.6709,
+      "step": 471
+    },
+    {
+      "epoch": 0.06606018194541638,
+      "grad_norm": 0.48530849719789637,
+      "learning_rate": 4.9831834953943236e-05,
+      "loss": 0.6798,
+      "step": 472
+    },
+    {
+      "epoch": 0.06620013995801259,
+      "grad_norm": 0.5169999057943723,
+      "learning_rate": 4.9830520087843894e-05,
+      "loss": 0.6949,
+      "step": 473
+    },
+    {
+      "epoch": 0.06634009797060882,
+      "grad_norm": 0.563401180326856,
+      "learning_rate": 4.982920011882074e-05,
+      "loss": 0.7226,
+      "step": 474
+    },
+    {
+      "epoch": 0.06648005598320504,
+      "grad_norm": 0.5359095901003381,
+      "learning_rate": 4.982787504714503e-05,
+      "loss": 0.6834,
+      "step": 475
+    },
+    {
+      "epoch": 0.06662001399580125,
+      "grad_norm": 0.7237453631577152,
+      "learning_rate": 4.982654487308908e-05,
+      "loss": 0.6794,
+      "step": 476
+    },
+    {
+      "epoch": 0.06675997200839748,
+      "grad_norm": 0.5209505752877471,
+      "learning_rate": 4.982520959692626e-05,
+      "loss": 0.6533,
+      "step": 477
+    },
+    {
+      "epoch": 0.0668999300209937,
+      "grad_norm": 0.5103023397815947,
+      "learning_rate": 4.982386921893098e-05,
+      "loss": 0.6633,
+      "step": 478
+    },
+    {
+      "epoch": 0.06703988803358993,
+      "grad_norm": 0.5060683439354244,
+      "learning_rate": 4.98225237393787e-05,
+      "loss": 0.6369,
+      "step": 479
+    },
+    {
+      "epoch": 0.06717984604618614,
+      "grad_norm": 0.5296294502024173,
+      "learning_rate": 4.9821173158545936e-05,
+      "loss": 0.6394,
+      "step": 480
+    },
+    {
+      "epoch": 0.06731980405878236,
+      "grad_norm": 0.5333731273318062,
+      "learning_rate": 4.981981747671024e-05,
+      "loss": 0.6792,
+      "step": 481
+    },
+    {
+      "epoch": 0.06745976207137859,
+      "grad_norm": 0.5350960162399097,
+      "learning_rate": 4.981845669415022e-05,
+      "loss": 0.6378,
+      "step": 482
+    },
+    {
+      "epoch": 0.0675997200839748,
+      "grad_norm": 0.4966051310703327,
+      "learning_rate": 4.9817090811145524e-05,
+      "loss": 0.6664,
+      "step": 483
+    },
+    {
+      "epoch": 0.06773967809657103,
+      "grad_norm": 0.5117571057914638,
+      "learning_rate": 4.9815719827976864e-05,
+      "loss": 0.674,
+      "step": 484
+    },
+    {
+      "epoch": 0.06787963610916725,
+      "grad_norm": 0.5325656306807487,
+      "learning_rate": 4.9814343744925984e-05,
+      "loss": 0.7459,
+      "step": 485
+    },
+    {
+      "epoch": 0.06801959412176348,
+      "grad_norm": 0.5363735435396061,
+      "learning_rate": 4.981296256227569e-05,
+      "loss": 0.6356,
+      "step": 486
+    },
+    {
+      "epoch": 0.06815955213435969,
+      "grad_norm": 0.5880213444397324,
+      "learning_rate": 4.981157628030984e-05,
+      "loss": 0.7152,
+      "step": 487
+    },
+    {
+      "epoch": 0.06829951014695591,
+      "grad_norm": 0.5622620490135608,
+      "learning_rate": 4.9810184899313294e-05,
+      "loss": 0.6904,
+      "step": 488
+    },
+    {
+      "epoch": 0.06843946815955214,
+      "grad_norm": 0.5387407294746895,
+      "learning_rate": 4.980878841957203e-05,
+      "loss": 0.6417,
+      "step": 489
+    },
+    {
+      "epoch": 0.06857942617214835,
+      "grad_norm": 0.5810694092783647,
+      "learning_rate": 4.9807386841373014e-05,
+      "loss": 0.7077,
+      "step": 490
+    },
+    {
+      "epoch": 0.06871938418474458,
+      "grad_norm": 0.5468672188592382,
+      "learning_rate": 4.9805980165004304e-05,
+      "loss": 0.6831,
+      "step": 491
+    },
+    {
+      "epoch": 0.0688593421973408,
+      "grad_norm": 0.5009452260143993,
+      "learning_rate": 4.9804568390754974e-05,
+      "loss": 0.6618,
+      "step": 492
+    },
+    {
+      "epoch": 0.06899930020993703,
+      "grad_norm": 0.5198589285355217,
+      "learning_rate": 4.980315151891516e-05,
+      "loss": 0.7322,
+      "step": 493
+    },
+    {
+      "epoch": 0.06913925822253324,
+      "grad_norm": 0.5255753639976765,
+      "learning_rate": 4.980172954977605e-05,
+      "loss": 0.667,
+      "step": 494
+    },
+    {
+      "epoch": 0.06927921623512946,
+      "grad_norm": 0.514064283357064,
+      "learning_rate": 4.980030248362987e-05,
+      "loss": 0.7107,
+      "step": 495
+    },
+    {
+      "epoch": 0.06941917424772569,
+      "grad_norm": 0.5532838372693374,
+      "learning_rate": 4.9798870320769886e-05,
+      "loss": 0.6811,
+      "step": 496
+    },
+    {
+      "epoch": 0.0695591322603219,
+      "grad_norm": 0.602307520111878,
+      "learning_rate": 4.9797433061490434e-05,
+      "loss": 0.655,
+      "step": 497
+    },
+    {
+      "epoch": 0.06969909027291812,
+      "grad_norm": 1.1081876267570516,
+      "learning_rate": 4.979599070608688e-05,
+      "loss": 0.7107,
+      "step": 498
+    },
+    {
+      "epoch": 0.06983904828551435,
+      "grad_norm": 0.529722398927933,
+      "learning_rate": 4.979454325485565e-05,
+      "loss": 0.7186,
+      "step": 499
+    },
+    {
+      "epoch": 0.06997900629811056,
+      "grad_norm": 0.688076194733635,
+      "learning_rate": 4.97930907080942e-05,
+      "loss": 0.6966,
+      "step": 500
+    },
+    {
+      "epoch": 0.07011896431070679,
+      "grad_norm": 0.5241126214650118,
+      "learning_rate": 4.979163306610105e-05,
+      "loss": 0.6648,
+      "step": 501
+    },
+    {
+      "epoch": 0.07025892232330301,
+      "grad_norm": 0.6545221839203675,
+      "learning_rate": 4.9790170329175754e-05,
+      "loss": 0.6634,
+      "step": 502
+    },
+    {
+      "epoch": 0.07039888033589924,
+      "grad_norm": 0.5292755961917179,
+      "learning_rate": 4.978870249761893e-05,
+      "loss": 0.6554,
+      "step": 503
+    },
+    {
+      "epoch": 0.07053883834849545,
+      "grad_norm": 0.5071567724521334,
+      "learning_rate": 4.978722957173222e-05,
+      "loss": 0.6741,
+      "step": 504
+    },
+    {
+      "epoch": 0.07067879636109167,
+      "grad_norm": 0.6987987933626845,
+      "learning_rate": 4.9785751551818336e-05,
+      "loss": 0.6845,
+      "step": 505
+    },
+    {
+      "epoch": 0.0708187543736879,
+      "grad_norm": 0.6505466788008574,
+      "learning_rate": 4.9784268438181016e-05,
+      "loss": 0.6632,
+      "step": 506
+    },
+    {
+      "epoch": 0.07095871238628411,
+      "grad_norm": 0.5780869067311343,
+      "learning_rate": 4.978278023112506e-05,
+      "loss": 0.7101,
+      "step": 507
+    },
+    {
+      "epoch": 0.07109867039888033,
+      "grad_norm": 0.5668906286656329,
+      "learning_rate": 4.978128693095632e-05,
+      "loss": 0.6727,
+      "step": 508
+    },
+    {
+      "epoch": 0.07123862841147656,
+      "grad_norm": 0.5631385798045004,
+      "learning_rate": 4.977978853798166e-05,
+      "loss": 0.6936,
+      "step": 509
+    },
+    {
+      "epoch": 0.07137858642407278,
+      "grad_norm": 0.563836937443064,
+      "learning_rate": 4.977828505250903e-05,
+      "loss": 0.6646,
+      "step": 510
+    },
+    {
+      "epoch": 0.071518544436669,
+      "grad_norm": 0.49771256088584315,
+      "learning_rate": 4.977677647484741e-05,
+      "loss": 0.6543,
+      "step": 511
+    },
+    {
+      "epoch": 0.07165850244926522,
+      "grad_norm": 0.813746974710708,
+      "learning_rate": 4.977526280530684e-05,
+      "loss": 0.6579,
+      "step": 512
+    },
+    {
+      "epoch": 0.07179846046186145,
+      "grad_norm": 0.6705561238466075,
+      "learning_rate": 4.977374404419837e-05,
+      "loss": 0.6773,
+      "step": 513
+    },
+    {
+      "epoch": 0.07193841847445766,
+      "grad_norm": 0.6410722179382549,
+      "learning_rate": 4.977222019183414e-05,
+      "loss": 0.7036,
+      "step": 514
+    },
+    {
+      "epoch": 0.07207837648705388,
+      "grad_norm": 0.49081412125581536,
+      "learning_rate": 4.977069124852731e-05,
+      "loss": 0.6944,
+      "step": 515
+    },
+    {
+      "epoch": 0.07221833449965011,
+      "grad_norm": 4.072584572211915,
+      "learning_rate": 4.976915721459209e-05,
+      "loss": 0.6678,
+      "step": 516
+    },
+    {
+      "epoch": 0.07235829251224632,
+      "grad_norm": 0.5026288646761555,
+      "learning_rate": 4.9767618090343745e-05,
+      "loss": 0.6564,
+      "step": 517
+    },
+    {
+      "epoch": 0.07249825052484254,
+      "grad_norm": 0.5165366931023583,
+      "learning_rate": 4.976607387609858e-05,
+      "loss": 0.6766,
+      "step": 518
+    },
+    {
+      "epoch": 0.07263820853743877,
+      "grad_norm": 0.5086677754272171,
+      "learning_rate": 4.976452457217394e-05,
+      "loss": 0.6468,
+      "step": 519
+    },
+    {
+      "epoch": 0.072778166550035,
+      "grad_norm": 0.5430265340871115,
+      "learning_rate": 4.976297017888824e-05,
+      "loss": 0.6543,
+      "step": 520
+    },
+    {
+      "epoch": 0.0729181245626312,
+      "grad_norm": 0.5436880201922443,
+      "learning_rate": 4.976141069656091e-05,
+      "loss": 0.6712,
+      "step": 521
+    },
+    {
+      "epoch": 0.07305808257522743,
+      "grad_norm": 0.5178999941373105,
+      "learning_rate": 4.975984612551243e-05,
+      "loss": 0.6676,
+      "step": 522
+    },
+    {
+      "epoch": 0.07319804058782366,
+      "grad_norm": 0.5814901447061906,
+      "learning_rate": 4.975827646606436e-05,
+      "loss": 0.6959,
+      "step": 523
+    },
+    {
+      "epoch": 0.07333799860041987,
+      "grad_norm": 0.5577926798263833,
+      "learning_rate": 4.975670171853926e-05,
+      "loss": 0.6691,
+      "step": 524
+    },
+    {
+      "epoch": 0.0734779566130161,
+      "grad_norm": 0.5368192741447849,
+      "learning_rate": 4.975512188326077e-05,
+      "loss": 0.7067,
+      "step": 525
+    },
+    {
+      "epoch": 0.07361791462561232,
+      "grad_norm": 0.8965061390425338,
+      "learning_rate": 4.9753536960553545e-05,
+      "loss": 0.6601,
+      "step": 526
+    },
+    {
+      "epoch": 0.07375787263820854,
+      "grad_norm": 0.8372680909100517,
+      "learning_rate": 4.975194695074333e-05,
+      "loss": 0.7079,
+      "step": 527
+    },
+    {
+      "epoch": 0.07389783065080475,
+      "grad_norm": 0.610185754465145,
+      "learning_rate": 4.9750351854156864e-05,
+      "loss": 0.6345,
+      "step": 528
+    },
+    {
+      "epoch": 0.07403778866340098,
+      "grad_norm": 0.5906986023369948,
+      "learning_rate": 4.9748751671121964e-05,
+      "loss": 0.6875,
+      "step": 529
+    },
+    {
+      "epoch": 0.0741777466759972,
+      "grad_norm": 0.5382974215866725,
+      "learning_rate": 4.9747146401967484e-05,
+      "loss": 0.6763,
+      "step": 530
+    },
+    {
+      "epoch": 0.07431770468859342,
+      "grad_norm": 0.5118350667595736,
+      "learning_rate": 4.9745536047023324e-05,
+      "loss": 0.693,
+      "step": 531
+    },
+    {
+      "epoch": 0.07445766270118964,
+      "grad_norm": 0.5342169754955468,
+      "learning_rate": 4.974392060662042e-05,
+      "loss": 0.6564,
+      "step": 532
+    },
+    {
+      "epoch": 0.07459762071378587,
+      "grad_norm": 0.494630398936876,
+      "learning_rate": 4.9742300081090774e-05,
+      "loss": 0.651,
+      "step": 533
+    },
+    {
+      "epoch": 0.07473757872638208,
+      "grad_norm": 0.5020465590373978,
+      "learning_rate": 4.974067447076742e-05,
+      "loss": 0.6924,
+      "step": 534
+    },
+    {
+      "epoch": 0.0748775367389783,
+      "grad_norm": 0.4949817767414362,
+      "learning_rate": 4.973904377598443e-05,
+      "loss": 0.6757,
+      "step": 535
+    },
+    {
+      "epoch": 0.07501749475157453,
+      "grad_norm": 0.503752121503341,
+      "learning_rate": 4.973740799707692e-05,
+      "loss": 0.6316,
+      "step": 536
+    },
+    {
+      "epoch": 0.07515745276417075,
+      "grad_norm": 0.5134153985614953,
+      "learning_rate": 4.973576713438108e-05,
+      "loss": 0.6768,
+      "step": 537
+    },
+    {
+      "epoch": 0.07529741077676697,
+      "grad_norm": 0.5293943650054048,
+      "learning_rate": 4.973412118823412e-05,
+      "loss": 0.6915,
+      "step": 538
+    },
+    {
+      "epoch": 0.07543736878936319,
+      "grad_norm": 0.5248241753853723,
+      "learning_rate": 4.973247015897428e-05,
+      "loss": 0.6878,
+      "step": 539
+    },
+    {
+      "epoch": 0.07557732680195942,
+      "grad_norm": 0.536619020426942,
+      "learning_rate": 4.973081404694088e-05,
+      "loss": 0.6913,
+      "step": 540
+    },
+    {
+      "epoch": 0.07571728481455563,
+      "grad_norm": 0.4893689470456731,
+      "learning_rate": 4.972915285247426e-05,
+      "loss": 0.6593,
+      "step": 541
+    },
+    {
+      "epoch": 0.07585724282715185,
+      "grad_norm": 0.5456811185536261,
+      "learning_rate": 4.9727486575915823e-05,
+      "loss": 0.6797,
+      "step": 542
+    },
+    {
+      "epoch": 0.07599720083974808,
+      "grad_norm": 0.547854073554072,
+      "learning_rate": 4.9725815217607994e-05,
+      "loss": 0.6729,
+      "step": 543
+    },
+    {
+      "epoch": 0.0761371588523443,
+      "grad_norm": 0.5192786810194786,
+      "learning_rate": 4.972413877789426e-05,
+      "loss": 0.6241,
+      "step": 544
+    },
+    {
+      "epoch": 0.07627711686494051,
+      "grad_norm": 0.47716140290292464,
+      "learning_rate": 4.972245725711914e-05,
+      "loss": 0.662,
+      "step": 545
+    },
+    {
+      "epoch": 0.07641707487753674,
+      "grad_norm": 0.5273241129689946,
+      "learning_rate": 4.972077065562821e-05,
+      "loss": 0.6603,
+      "step": 546
+    },
+    {
+      "epoch": 0.07655703289013296,
+      "grad_norm": 0.5220860206770274,
+      "learning_rate": 4.971907897376809e-05,
+      "loss": 0.6862,
+      "step": 547
+    },
+    {
+      "epoch": 0.07669699090272918,
+      "grad_norm": 0.4954377301463657,
+      "learning_rate": 4.971738221188643e-05,
+      "loss": 0.6357,
+      "step": 548
+    },
+    {
+      "epoch": 0.0768369489153254,
+      "grad_norm": 0.5031579615121048,
+      "learning_rate": 4.9715680370331926e-05,
+      "loss": 0.7242,
+      "step": 549
+    },
+    {
+      "epoch": 0.07697690692792163,
+      "grad_norm": 0.4883815277737791,
+      "learning_rate": 4.9713973449454335e-05,
+      "loss": 0.6764,
+      "step": 550
+    },
+    {
+      "epoch": 0.07711686494051785,
+      "grad_norm": 0.5324244397343785,
+      "learning_rate": 4.971226144960443e-05,
+      "loss": 0.7038,
+      "step": 551
+    },
+    {
+      "epoch": 0.07725682295311406,
+      "grad_norm": 0.5072266809492816,
+      "learning_rate": 4.971054437113406e-05,
+      "loss": 0.6778,
+      "step": 552
+    },
+    {
+      "epoch": 0.07739678096571029,
+      "grad_norm": 0.48810600146086575,
+      "learning_rate": 4.97088222143961e-05,
+      "loss": 0.6719,
+      "step": 553
+    },
+    {
+      "epoch": 0.07753673897830651,
+      "grad_norm": 0.5236886736407549,
+      "learning_rate": 4.970709497974447e-05,
+      "loss": 0.6472,
+      "step": 554
+    },
+    {
+      "epoch": 0.07767669699090272,
+      "grad_norm": 0.521109484700154,
+      "learning_rate": 4.9705362667534126e-05,
+      "loss": 0.637,
+      "step": 555
+    },
+    {
+      "epoch": 0.07781665500349895,
+      "grad_norm": 0.4825990181972682,
+      "learning_rate": 4.970362527812109e-05,
+      "loss": 0.6906,
+      "step": 556
+    },
+    {
+      "epoch": 0.07795661301609517,
+      "grad_norm": 0.5261802281784332,
+      "learning_rate": 4.970188281186241e-05,
+      "loss": 0.6585,
+      "step": 557
+    },
+    {
+      "epoch": 0.07809657102869139,
+      "grad_norm": 0.4913754461237493,
+      "learning_rate": 4.970013526911617e-05,
+      "loss": 0.6868,
+      "step": 558
+    },
+    {
+      "epoch": 0.07823652904128761,
+      "grad_norm": 0.5026456765855343,
+      "learning_rate": 4.969838265024151e-05,
+      "loss": 0.7251,
+      "step": 559
+    },
+    {
+      "epoch": 0.07837648705388384,
+      "grad_norm": 0.500311336929775,
+      "learning_rate": 4.969662495559862e-05,
+      "loss": 0.6589,
+      "step": 560
+    },
+    {
+      "epoch": 0.07851644506648006,
+      "grad_norm": 0.5085930979997164,
+      "learning_rate": 4.969486218554871e-05,
+      "loss": 0.6596,
+      "step": 561
+    },
+    {
+      "epoch": 0.07865640307907627,
+      "grad_norm": 0.5160633817299983,
+      "learning_rate": 4.9693094340454055e-05,
+      "loss": 0.6414,
+      "step": 562
+    },
+    {
+      "epoch": 0.0787963610916725,
+      "grad_norm": 0.5037304772878324,
+      "learning_rate": 4.969132142067797e-05,
+      "loss": 0.6766,
+      "step": 563
+    },
+    {
+      "epoch": 0.07893631910426872,
+      "grad_norm": 0.4848388430973453,
+      "learning_rate": 4.96895434265848e-05,
+      "loss": 0.6928,
+      "step": 564
+    },
+    {
+      "epoch": 0.07907627711686493,
+      "grad_norm": 0.5085119763538183,
+      "learning_rate": 4.968776035853996e-05,
+      "loss": 0.6998,
+      "step": 565
+    },
+    {
+      "epoch": 0.07921623512946116,
+      "grad_norm": 0.9757272802585409,
+      "learning_rate": 4.968597221690986e-05,
+      "loss": 0.6367,
+      "step": 566
+    },
+    {
+      "epoch": 0.07935619314205739,
+      "grad_norm": 0.49870651337329763,
+      "learning_rate": 4.9684179002062e-05,
+      "loss": 0.6814,
+      "step": 567
+    },
+    {
+      "epoch": 0.07949615115465361,
+      "grad_norm": 0.5044807130296783,
+      "learning_rate": 4.9682380714364897e-05,
+      "loss": 0.6108,
+      "step": 568
+    },
+    {
+      "epoch": 0.07963610916724982,
+      "grad_norm": 0.708447636643108,
+      "learning_rate": 4.968057735418812e-05,
+      "loss": 0.6621,
+      "step": 569
+    },
+    {
+      "epoch": 0.07977606717984605,
+      "grad_norm": 0.5176890020141557,
+      "learning_rate": 4.967876892190227e-05,
+      "loss": 0.6712,
+      "step": 570
+    },
+    {
+      "epoch": 0.07991602519244227,
+      "grad_norm": 0.5165402375548048,
+      "learning_rate": 4.967695541787901e-05,
+      "loss": 0.6516,
+      "step": 571
+    },
+    {
+      "epoch": 0.08005598320503848,
+      "grad_norm": 0.5023727855771402,
+      "learning_rate": 4.967513684249103e-05,
+      "loss": 0.6761,
+      "step": 572
+    },
+    {
+      "epoch": 0.08019594121763471,
+      "grad_norm": 0.5273373598923502,
+      "learning_rate": 4.967331319611206e-05,
+      "loss": 0.6513,
+      "step": 573
+    },
+    {
+      "epoch": 0.08033589923023093,
+      "grad_norm": 0.5038764795022782,
+      "learning_rate": 4.967148447911688e-05,
+      "loss": 0.6332,
+      "step": 574
+    },
+    {
+      "epoch": 0.08047585724282715,
+      "grad_norm": 0.5050569409944193,
+      "learning_rate": 4.966965069188132e-05,
+      "loss": 0.6617,
+      "step": 575
+    },
+    {
+      "epoch": 0.08061581525542337,
+      "grad_norm": 0.48919260228995454,
+      "learning_rate": 4.9667811834782224e-05,
+      "loss": 0.6654,
+      "step": 576
+    },
+    {
+      "epoch": 0.0807557732680196,
+      "grad_norm": 0.5353958441988039,
+      "learning_rate": 4.9665967908197506e-05,
+      "loss": 0.6592,
+      "step": 577
+    },
+    {
+      "epoch": 0.08089573128061582,
+      "grad_norm": 0.5361921590438847,
+      "learning_rate": 4.966411891250612e-05,
+      "loss": 0.674,
+      "step": 578
+    },
+    {
+      "epoch": 0.08103568929321203,
+      "grad_norm": 0.5342727473087102,
+      "learning_rate": 4.9662264848088034e-05,
+      "loss": 0.6885,
+      "step": 579
+    },
+    {
+      "epoch": 0.08117564730580826,
+      "grad_norm": 0.539903295130481,
+      "learning_rate": 4.96604057153243e-05,
+      "loss": 0.6662,
+      "step": 580
+    },
+    {
+      "epoch": 0.08131560531840448,
+      "grad_norm": 0.49180697521465716,
+      "learning_rate": 4.965854151459697e-05,
+      "loss": 0.6423,
+      "step": 581
+    },
+    {
+      "epoch": 0.0814555633310007,
+      "grad_norm": 0.5088216974014643,
+      "learning_rate": 4.965667224628916e-05,
+      "loss": 0.696,
+      "step": 582
+    },
+    {
+      "epoch": 0.08159552134359692,
+      "grad_norm": 0.49968640535207737,
+      "learning_rate": 4.965479791078502e-05,
+      "loss": 0.6543,
+      "step": 583
+    },
+    {
+      "epoch": 0.08173547935619314,
+      "grad_norm": 0.4859083259348156,
+      "learning_rate": 4.965291850846976e-05,
+      "loss": 0.6729,
+      "step": 584
+    },
+    {
+      "epoch": 0.08187543736878937,
+      "grad_norm": 0.4993128882253819,
+      "learning_rate": 4.96510340397296e-05,
+      "loss": 0.6521,
+      "step": 585
+    },
+    {
+      "epoch": 0.08201539538138558,
+      "grad_norm": 0.5193117598814236,
+      "learning_rate": 4.964914450495183e-05,
+      "loss": 0.6525,
+      "step": 586
+    },
+    {
+      "epoch": 0.0821553533939818,
+      "grad_norm": 0.5719370954745235,
+      "learning_rate": 4.964724990452476e-05,
+      "loss": 0.6622,
+      "step": 587
+    },
+    {
+      "epoch": 0.08229531140657803,
+      "grad_norm": 0.4700969122107685,
+      "learning_rate": 4.964535023883776e-05,
+      "loss": 0.6954,
+      "step": 588
+    },
+    {
+      "epoch": 0.08243526941917424,
+      "grad_norm": 0.484560721731152,
+      "learning_rate": 4.964344550828122e-05,
+      "loss": 0.6869,
+      "step": 589
+    },
+    {
+      "epoch": 0.08257522743177047,
+      "grad_norm": 0.4925497467039491,
+      "learning_rate": 4.964153571324658e-05,
+      "loss": 0.6675,
+      "step": 590
+    },
+    {
+      "epoch": 0.08271518544436669,
+      "grad_norm": 0.48466585548673363,
+      "learning_rate": 4.9639620854126326e-05,
+      "loss": 0.6593,
+      "step": 591
+    },
+    {
+      "epoch": 0.08285514345696292,
+      "grad_norm": 0.48477008272792094,
+      "learning_rate": 4.963770093131399e-05,
+      "loss": 0.7009,
+      "step": 592
+    },
+    {
+      "epoch": 0.08299510146955913,
+      "grad_norm": 0.5346681837659809,
+      "learning_rate": 4.963577594520412e-05,
+      "loss": 0.6611,
+      "step": 593
+    },
+    {
+      "epoch": 0.08313505948215535,
+      "grad_norm": 0.5094553222628161,
+      "learning_rate": 4.963384589619233e-05,
+      "loss": 0.6488,
+      "step": 594
+    },
+    {
+      "epoch": 0.08327501749475158,
+      "grad_norm": 0.5908935657202756,
+      "learning_rate": 4.9631910784675265e-05,
+      "loss": 0.6724,
+      "step": 595
+    },
+    {
+      "epoch": 0.08341497550734779,
+      "grad_norm": 0.515639184329749,
+      "learning_rate": 4.96299706110506e-05,
+      "loss": 0.6859,
+      "step": 596
+    },
+    {
+      "epoch": 0.08355493351994402,
+      "grad_norm": 0.55185217811717,
+      "learning_rate": 4.962802537571707e-05,
+      "loss": 0.6623,
+      "step": 597
+    },
+    {
+      "epoch": 0.08369489153254024,
+      "grad_norm": 0.49915022546087007,
+      "learning_rate": 4.962607507907444e-05,
+      "loss": 0.6648,
+      "step": 598
+    },
+    {
+      "epoch": 0.08383484954513645,
+      "grad_norm": 0.5284448258033697,
+      "learning_rate": 4.962411972152352e-05,
+      "loss": 0.6342,
+      "step": 599
+    },
+    {
+      "epoch": 0.08397480755773268,
+      "grad_norm": 0.4774184931571651,
+      "learning_rate": 4.9622159303466144e-05,
+      "loss": 0.686,
+      "step": 600
+    },
+    {
+      "epoch": 0.0841147655703289,
+      "grad_norm": 0.49606716653441585,
+      "learning_rate": 4.962019382530521e-05,
+      "loss": 0.6664,
+      "step": 601
+    },
+    {
+      "epoch": 0.08425472358292513,
+      "grad_norm": 0.5155895908215095,
+      "learning_rate": 4.9618223287444624e-05,
+      "loss": 0.693,
+      "step": 602
+    },
+    {
+      "epoch": 0.08439468159552134,
+      "grad_norm": 0.48698005510262393,
+      "learning_rate": 4.9616247690289375e-05,
+      "loss": 0.6824,
+      "step": 603
+    },
+    {
+      "epoch": 0.08453463960811756,
+      "grad_norm": 0.5096390231519464,
+      "learning_rate": 4.9614267034245454e-05,
+      "loss": 0.6709,
+      "step": 604
+    },
+    {
+      "epoch": 0.08467459762071379,
+      "grad_norm": 0.47879312610079056,
+      "learning_rate": 4.961228131971991e-05,
+      "loss": 0.6191,
+      "step": 605
+    },
+    {
+      "epoch": 0.08481455563331,
+      "grad_norm": 0.5036694948790285,
+      "learning_rate": 4.9610290547120835e-05,
+      "loss": 0.6677,
+      "step": 606
+    },
+    {
+      "epoch": 0.08495451364590623,
+      "grad_norm": 0.5387911945134125,
+      "learning_rate": 4.960829471685734e-05,
+      "loss": 0.7108,
+      "step": 607
+    },
+    {
+      "epoch": 0.08509447165850245,
+      "grad_norm": 0.48418619588152695,
+      "learning_rate": 4.9606293829339595e-05,
+      "loss": 0.6581,
+      "step": 608
+    },
+    {
+      "epoch": 0.08523442967109868,
+      "grad_norm": 0.46835555910688126,
+      "learning_rate": 4.9604287884978803e-05,
+      "loss": 0.6285,
+      "step": 609
+    },
+    {
+      "epoch": 0.08537438768369489,
+      "grad_norm": 0.49536583738424655,
+      "learning_rate": 4.9602276884187206e-05,
+      "loss": 0.6862,
+      "step": 610
+    },
+    {
+      "epoch": 0.08551434569629111,
+      "grad_norm": 0.5219805753845016,
+      "learning_rate": 4.9600260827378074e-05,
+      "loss": 0.7262,
+      "step": 611
+    },
+    {
+      "epoch": 0.08565430370888734,
+      "grad_norm": 0.5105322442141146,
+      "learning_rate": 4.959823971496574e-05,
+      "loss": 0.6818,
+      "step": 612
+    },
+    {
+      "epoch": 0.08579426172148355,
+      "grad_norm": 0.5306990676448388,
+      "learning_rate": 4.9596213547365566e-05,
+      "loss": 0.7141,
+      "step": 613
+    },
+    {
+      "epoch": 0.08593421973407978,
+      "grad_norm": 0.518971911880668,
+      "learning_rate": 4.959418232499394e-05,
+      "loss": 0.6447,
+      "step": 614
+    },
+    {
+      "epoch": 0.086074177746676,
+      "grad_norm": 0.48723426006398907,
+      "learning_rate": 4.959214604826831e-05,
+      "loss": 0.6691,
+      "step": 615
+    },
+    {
+      "epoch": 0.08621413575927221,
+      "grad_norm": 0.47908534192693764,
+      "learning_rate": 4.9590104717607135e-05,
+      "loss": 0.7039,
+      "step": 616
+    },
+    {
+      "epoch": 0.08635409377186844,
+      "grad_norm": 0.47477383891868286,
+      "learning_rate": 4.958805833342994e-05,
+      "loss": 0.6664,
+      "step": 617
+    },
+    {
+      "epoch": 0.08649405178446466,
+      "grad_norm": 0.4984833292893388,
+      "learning_rate": 4.958600689615728e-05,
+      "loss": 0.6829,
+      "step": 618
+    },
+    {
+      "epoch": 0.08663400979706089,
+      "grad_norm": 0.49546276658635413,
+      "learning_rate": 4.958395040621073e-05,
+      "loss": 0.6482,
+      "step": 619
+    },
+    {
+      "epoch": 0.0867739678096571,
+      "grad_norm": 0.51460494631631,
+      "learning_rate": 4.958188886401295e-05,
+      "loss": 0.684,
+      "step": 620
+    },
+    {
+      "epoch": 0.08691392582225332,
+      "grad_norm": 0.5030849606038066,
+      "learning_rate": 4.9579822269987574e-05,
+      "loss": 0.6708,
+      "step": 621
+    },
+    {
+      "epoch": 0.08705388383484955,
+      "grad_norm": 0.50101210673068,
+      "learning_rate": 4.957775062455933e-05,
+      "loss": 0.6639,
+      "step": 622
+    },
+    {
+      "epoch": 0.08719384184744576,
+      "grad_norm": 0.46852010673845496,
+      "learning_rate": 4.9575673928153957e-05,
+      "loss": 0.6603,
+      "step": 623
+    },
+    {
+      "epoch": 0.08733379986004199,
+      "grad_norm": 0.48823643441327075,
+      "learning_rate": 4.957359218119824e-05,
+      "loss": 0.6719,
+      "step": 624
+    },
+    {
+      "epoch": 0.08747375787263821,
+      "grad_norm": 0.4891031572821987,
+      "learning_rate": 4.957150538411999e-05,
+      "loss": 0.6656,
+      "step": 625
+    },
+    {
+      "epoch": 0.08761371588523444,
+      "grad_norm": 0.563701350648365,
+      "learning_rate": 4.956941353734807e-05,
+      "loss": 0.6663,
+      "step": 626
+    },
+    {
+      "epoch": 0.08775367389783065,
+      "grad_norm": 0.48687644197380436,
+      "learning_rate": 4.956731664131238e-05,
+      "loss": 0.6913,
+      "step": 627
+    },
+    {
+      "epoch": 0.08789363191042687,
+      "grad_norm": 0.4907175129397562,
+      "learning_rate": 4.956521469644384e-05,
+      "loss": 0.706,
+      "step": 628
+    },
+    {
+      "epoch": 0.0880335899230231,
+      "grad_norm": 0.5313788653762662,
+      "learning_rate": 4.9563107703174436e-05,
+      "loss": 0.6644,
+      "step": 629
+    },
+    {
+      "epoch": 0.08817354793561931,
+      "grad_norm": 0.5033134935760109,
+      "learning_rate": 4.956099566193717e-05,
+      "loss": 0.6654,
+      "step": 630
+    },
+    {
+      "epoch": 0.08831350594821553,
+      "grad_norm": 0.4589870741107357,
+      "learning_rate": 4.955887857316609e-05,
+      "loss": 0.6417,
+      "step": 631
+    },
+    {
+      "epoch": 0.08845346396081176,
+      "grad_norm": 0.48309495215754206,
+      "learning_rate": 4.955675643729628e-05,
+      "loss": 0.665,
+      "step": 632
+    },
+    {
+      "epoch": 0.08859342197340797,
+      "grad_norm": 0.5259807553857797,
+      "learning_rate": 4.955462925476385e-05,
+      "loss": 0.6945,
+      "step": 633
+    },
+    {
+      "epoch": 0.0887333799860042,
+      "grad_norm": 0.5163974444977445,
+      "learning_rate": 4.9552497026005974e-05,
+      "loss": 0.6972,
+      "step": 634
+    },
+    {
+      "epoch": 0.08887333799860042,
+      "grad_norm": 0.4973389672383331,
+      "learning_rate": 4.955035975146084e-05,
+      "loss": 0.6551,
+      "step": 635
+    },
+    {
+      "epoch": 0.08901329601119665,
+      "grad_norm": 0.47160543121321047,
+      "learning_rate": 4.9548217431567665e-05,
+      "loss": 0.6499,
+      "step": 636
+    },
+    {
+      "epoch": 0.08915325402379286,
+      "grad_norm": 0.4981939411665721,
+      "learning_rate": 4.954607006676675e-05,
+      "loss": 0.6541,
+      "step": 637
+    },
+    {
+      "epoch": 0.08929321203638908,
+      "grad_norm": 0.4495878467816103,
+      "learning_rate": 4.954391765749936e-05,
+      "loss": 0.6113,
+      "step": 638
+    },
+    {
+      "epoch": 0.08943317004898531,
+      "grad_norm": 0.47812065178812635,
+      "learning_rate": 4.954176020420788e-05,
+      "loss": 0.6269,
+      "step": 639
+    },
+    {
+      "epoch": 0.08957312806158152,
+      "grad_norm": 0.5080513968505054,
+      "learning_rate": 4.953959770733565e-05,
+      "loss": 0.7068,
+      "step": 640
+    },
+    {
+      "epoch": 0.08971308607417774,
+      "grad_norm": 0.5272266307181286,
+      "learning_rate": 4.95374301673271e-05,
+      "loss": 0.6986,
+      "step": 641
+    },
+    {
+      "epoch": 0.08985304408677397,
+      "grad_norm": 0.5013094523061675,
+      "learning_rate": 4.953525758462769e-05,
+      "loss": 0.6807,
+      "step": 642
+    },
+    {
+      "epoch": 0.0899930020993702,
+      "grad_norm": 0.6486797331723868,
+      "learning_rate": 4.95330799596839e-05,
+      "loss": 0.6414,
+      "step": 643
+    },
+    {
+      "epoch": 0.0901329601119664,
+      "grad_norm": 0.4872602306676202,
+      "learning_rate": 4.953089729294326e-05,
+      "loss": 0.6689,
+      "step": 644
+    },
+    {
+      "epoch": 0.09027291812456263,
+      "grad_norm": 0.5040432352418639,
+      "learning_rate": 4.952870958485432e-05,
+      "loss": 0.6865,
+      "step": 645
+    },
+    {
+      "epoch": 0.09041287613715886,
+      "grad_norm": 0.46173609966688006,
+      "learning_rate": 4.952651683586668e-05,
+      "loss": 0.6481,
+      "step": 646
+    },
+    {
+      "epoch": 0.09055283414975507,
+      "grad_norm": 0.45658386339702023,
+      "learning_rate": 4.952431904643097e-05,
+      "loss": 0.6539,
+      "step": 647
+    },
+    {
+      "epoch": 0.0906927921623513,
+      "grad_norm": 0.503432394082156,
+      "learning_rate": 4.952211621699887e-05,
+      "loss": 0.7104,
+      "step": 648
+    },
+    {
+      "epoch": 0.09083275017494752,
+      "grad_norm": 0.46517000786275325,
+      "learning_rate": 4.951990834802307e-05,
+      "loss": 0.6525,
+      "step": 649
+    },
+    {
+      "epoch": 0.09097270818754374,
+      "grad_norm": 0.5062356868953571,
+      "learning_rate": 4.951769543995731e-05,
+      "loss": 0.6823,
+      "step": 650
+    },
+    {
+      "epoch": 0.09111266620013996,
+      "grad_norm": 0.5119419441222778,
+      "learning_rate": 4.951547749325638e-05,
+      "loss": 0.6635,
+      "step": 651
+    },
+    {
+      "epoch": 0.09125262421273618,
+      "grad_norm": 0.46695225104802485,
+      "learning_rate": 4.951325450837607e-05,
+      "loss": 0.6971,
+      "step": 652
+    },
+    {
+      "epoch": 0.0913925822253324,
+      "grad_norm": 0.49434009422947434,
+      "learning_rate": 4.951102648577324e-05,
+      "loss": 0.6914,
+      "step": 653
+    },
+    {
+      "epoch": 0.09153254023792862,
+      "grad_norm": 0.4604670405526644,
+      "learning_rate": 4.950879342590577e-05,
+      "loss": 0.6519,
+      "step": 654
+    },
+    {
+      "epoch": 0.09167249825052484,
+      "grad_norm": 0.5184837228643522,
+      "learning_rate": 4.9506555329232574e-05,
+      "loss": 0.6886,
+      "step": 655
+    },
+    {
+      "epoch": 0.09181245626312107,
+      "grad_norm": 0.48694633832428874,
+      "learning_rate": 4.9504312196213596e-05,
+      "loss": 0.6636,
+      "step": 656
+    },
+    {
+      "epoch": 0.09195241427571728,
+      "grad_norm": 0.45470193445575985,
+      "learning_rate": 4.9502064027309836e-05,
+      "loss": 0.6402,
+      "step": 657
+    },
+    {
+      "epoch": 0.0920923722883135,
+      "grad_norm": 0.49263294738528823,
+      "learning_rate": 4.9499810822983314e-05,
+      "loss": 0.6513,
+      "step": 658
+    },
+    {
+      "epoch": 0.09223233030090973,
+      "grad_norm": 0.48137872453923874,
+      "learning_rate": 4.949755258369707e-05,
+      "loss": 0.6387,
+      "step": 659
+    },
+    {
+      "epoch": 0.09237228831350595,
+      "grad_norm": 0.5244416908958357,
+      "learning_rate": 4.949528930991522e-05,
+      "loss": 0.6634,
+      "step": 660
+    },
+    {
+      "epoch": 0.09251224632610217,
+      "grad_norm": 0.4835025657565747,
+      "learning_rate": 4.949302100210287e-05,
+      "loss": 0.6582,
+      "step": 661
+    },
+    {
+      "epoch": 0.09265220433869839,
+      "grad_norm": 0.4601647504352115,
+      "learning_rate": 4.9490747660726186e-05,
+      "loss": 0.633,
+      "step": 662
+    },
+    {
+      "epoch": 0.09279216235129462,
+      "grad_norm": 0.4807820429695223,
+      "learning_rate": 4.948846928625236e-05,
+      "loss": 0.6484,
+      "step": 663
+    },
+    {
+      "epoch": 0.09293212036389083,
+      "grad_norm": 0.4621826268890677,
+      "learning_rate": 4.948618587914963e-05,
+      "loss": 0.6201,
+      "step": 664
+    },
+    {
+      "epoch": 0.09307207837648705,
+      "grad_norm": 0.5355555893122882,
+      "learning_rate": 4.9483897439887256e-05,
+      "loss": 0.6939,
+      "step": 665
+    },
+    {
+      "epoch": 0.09321203638908328,
+      "grad_norm": 0.48631181937028944,
+      "learning_rate": 4.948160396893553e-05,
+      "loss": 0.6463,
+      "step": 666
+    },
+    {
+      "epoch": 0.0933519944016795,
+      "grad_norm": 0.4699689810513121,
+      "learning_rate": 4.947930546676579e-05,
+      "loss": 0.6417,
+      "step": 667
+    },
+    {
+      "epoch": 0.09349195241427571,
+      "grad_norm": 0.5070622795052616,
+      "learning_rate": 4.94770019338504e-05,
+      "loss": 0.6536,
+      "step": 668
+    },
+    {
+      "epoch": 0.09363191042687194,
+      "grad_norm": 0.5074909040261282,
+      "learning_rate": 4.947469337066275e-05,
+      "loss": 0.6313,
+      "step": 669
+    },
+    {
+      "epoch": 0.09377186843946816,
+      "grad_norm": 0.5145354321514217,
+      "learning_rate": 4.9472379777677287e-05,
+      "loss": 0.6679,
+      "step": 670
+    },
+    {
+      "epoch": 0.09391182645206438,
+      "grad_norm": 0.45833527492215737,
+      "learning_rate": 4.947006115536947e-05,
+      "loss": 0.6604,
+      "step": 671
+    },
+    {
+      "epoch": 0.0940517844646606,
+      "grad_norm": 0.45998883685850256,
+      "learning_rate": 4.9467737504215805e-05,
+      "loss": 0.6484,
+      "step": 672
+    },
+    {
+      "epoch": 0.09419174247725683,
+      "grad_norm": 0.46701832176416125,
+      "learning_rate": 4.946540882469381e-05,
+      "loss": 0.6622,
+      "step": 673
+    },
+    {
+      "epoch": 0.09433170048985304,
+      "grad_norm": 0.4937012715997403,
+      "learning_rate": 4.946307511728208e-05,
+      "loss": 0.6423,
+      "step": 674
+    },
+    {
+      "epoch": 0.09447165850244926,
+      "grad_norm": 0.45944259724419273,
+      "learning_rate": 4.9460736382460195e-05,
+      "loss": 0.6263,
+      "step": 675
+    },
+    {
+      "epoch": 0.09461161651504549,
+      "grad_norm": 0.4953211118671407,
+      "learning_rate": 4.945839262070879e-05,
+      "loss": 0.6401,
+      "step": 676
+    },
+    {
+      "epoch": 0.09475157452764171,
+      "grad_norm": 0.4954445380856775,
+      "learning_rate": 4.9456043832509544e-05,
+      "loss": 0.6766,
+      "step": 677
+    },
+    {
+      "epoch": 0.09489153254023792,
+      "grad_norm": 0.4756639472671179,
+      "learning_rate": 4.9453690018345144e-05,
+      "loss": 0.6516,
+      "step": 678
+    },
+    {
+      "epoch": 0.09503149055283415,
+      "grad_norm": 0.5216854708766936,
+      "learning_rate": 4.9451331178699324e-05,
+      "loss": 0.6233,
+      "step": 679
+    },
+    {
+      "epoch": 0.09517144856543037,
+      "grad_norm": 0.5149096513895491,
+      "learning_rate": 4.944896731405686e-05,
+      "loss": 0.6561,
+      "step": 680
+    },
+    {
+      "epoch": 0.09531140657802659,
+      "grad_norm": 0.4884876631431996,
+      "learning_rate": 4.944659842490354e-05,
+      "loss": 0.6594,
+      "step": 681
+    },
+    {
+      "epoch": 0.09545136459062281,
+      "grad_norm": 0.4960871887261993,
+      "learning_rate": 4.944422451172619e-05,
+      "loss": 0.6391,
+      "step": 682
+    },
+    {
+      "epoch": 0.09559132260321904,
+      "grad_norm": 0.49440117471489403,
+      "learning_rate": 4.944184557501269e-05,
+      "loss": 0.6714,
+      "step": 683
+    },
+    {
+      "epoch": 0.09573128061581526,
+      "grad_norm": 0.47570363710472247,
+      "learning_rate": 4.943946161525192e-05,
+      "loss": 0.6786,
+      "step": 684
+    },
+    {
+      "epoch": 0.09587123862841147,
+      "grad_norm": 0.48841469758633105,
+      "learning_rate": 4.9437072632933814e-05,
+      "loss": 0.6932,
+      "step": 685
+    },
+    {
+      "epoch": 0.0960111966410077,
+      "grad_norm": 0.47367349893999716,
+      "learning_rate": 4.943467862854934e-05,
+      "loss": 0.6331,
+      "step": 686
+    },
+    {
+      "epoch": 0.09615115465360392,
+      "grad_norm": 0.4966192406884947,
+      "learning_rate": 4.943227960259048e-05,
+      "loss": 0.6632,
+      "step": 687
+    },
+    {
+      "epoch": 0.09629111266620013,
+      "grad_norm": 0.46145851197097154,
+      "learning_rate": 4.942987555555026e-05,
+      "loss": 0.6428,
+      "step": 688
+    },
+    {
+      "epoch": 0.09643107067879636,
+      "grad_norm": 0.47559863610819186,
+      "learning_rate": 4.942746648792274e-05,
+      "loss": 0.6499,
+      "step": 689
+    },
+    {
+      "epoch": 0.09657102869139259,
+      "grad_norm": 0.46991178970104935,
+      "learning_rate": 4.9425052400203e-05,
+      "loss": 0.6134,
+      "step": 690
+    },
+    {
+      "epoch": 0.0967109867039888,
+      "grad_norm": 0.486056912045251,
+      "learning_rate": 4.942263329288716e-05,
+      "loss": 0.6594,
+      "step": 691
+    },
+    {
+      "epoch": 0.09685094471658502,
+      "grad_norm": 0.5045620121297718,
+      "learning_rate": 4.942020916647238e-05,
+      "loss": 0.6606,
+      "step": 692
+    },
+    {
+      "epoch": 0.09699090272918125,
+      "grad_norm": 0.49357726954668424,
+      "learning_rate": 4.941778002145684e-05,
+      "loss": 0.6436,
+      "step": 693
+    },
+    {
+      "epoch": 0.09713086074177747,
+      "grad_norm": 0.44767321778450064,
+      "learning_rate": 4.941534585833975e-05,
+      "loss": 0.6662,
+      "step": 694
+    },
+    {
+      "epoch": 0.09727081875437368,
+      "grad_norm": 0.49515786663501604,
+      "learning_rate": 4.941290667762136e-05,
+      "loss": 0.6441,
+      "step": 695
+    },
+    {
+      "epoch": 0.09741077676696991,
+      "grad_norm": 0.5193186895726914,
+      "learning_rate": 4.9410462479802945e-05,
+      "loss": 0.6667,
+      "step": 696
+    },
+    {
+      "epoch": 0.09755073477956613,
+      "grad_norm": 0.48819228314517343,
+      "learning_rate": 4.94080132653868e-05,
+      "loss": 0.6525,
+      "step": 697
+    },
+    {
+      "epoch": 0.09769069279216235,
+      "grad_norm": 0.47504662151869315,
+      "learning_rate": 4.940555903487628e-05,
+      "loss": 0.6507,
+      "step": 698
+    },
+    {
+      "epoch": 0.09783065080475857,
+      "grad_norm": 0.5087429079412467,
+      "learning_rate": 4.9403099788775754e-05,
+      "loss": 0.6885,
+      "step": 699
+    },
+    {
+      "epoch": 0.0979706088173548,
+      "grad_norm": 0.4991373858117139,
+      "learning_rate": 4.940063552759061e-05,
+      "loss": 0.6614,
+      "step": 700
+    },
+    {
+      "epoch": 0.09811056682995102,
+      "grad_norm": 0.5030115759681447,
+      "learning_rate": 4.939816625182729e-05,
+      "loss": 0.6415,
+      "step": 701
+    },
+    {
+      "epoch": 0.09825052484254723,
+      "grad_norm": 0.47371776821243294,
+      "learning_rate": 4.939569196199325e-05,
+      "loss": 0.6391,
+      "step": 702
+    },
+    {
+      "epoch": 0.09839048285514346,
+      "grad_norm": 0.4881159343567594,
+      "learning_rate": 4.9393212658596976e-05,
+      "loss": 0.6244,
+      "step": 703
+    },
+    {
+      "epoch": 0.09853044086773968,
+      "grad_norm": 0.49870488221888004,
+      "learning_rate": 4.9390728342148006e-05,
+      "loss": 0.6616,
+      "step": 704
+    },
+    {
+      "epoch": 0.0986703988803359,
+      "grad_norm": 0.4768129759031237,
+      "learning_rate": 4.938823901315688e-05,
+      "loss": 0.6559,
+      "step": 705
+    },
+    {
+      "epoch": 0.09881035689293212,
+      "grad_norm": 0.4729780422275052,
+      "learning_rate": 4.938574467213518e-05,
+      "loss": 0.6256,
+      "step": 706
+    },
+    {
+      "epoch": 0.09895031490552834,
+      "grad_norm": 0.49227905209471745,
+      "learning_rate": 4.9383245319595514e-05,
+      "loss": 0.6564,
+      "step": 707
+    },
+    {
+      "epoch": 0.09909027291812457,
+      "grad_norm": 0.49477666423952565,
+      "learning_rate": 4.9380740956051545e-05,
+      "loss": 0.6031,
+      "step": 708
+    },
+    {
+      "epoch": 0.09923023093072078,
+      "grad_norm": 0.4711224186000299,
+      "learning_rate": 4.9378231582017926e-05,
+      "loss": 0.6302,
+      "step": 709
+    },
+    {
+      "epoch": 0.099370188943317,
+      "grad_norm": 0.49742927700008693,
+      "learning_rate": 4.9375717198010366e-05,
+      "loss": 0.6732,
+      "step": 710
+    },
+    {
+      "epoch": 0.09951014695591323,
+      "grad_norm": 0.4750405851316029,
+      "learning_rate": 4.937319780454559e-05,
+      "loss": 0.6852,
+      "step": 711
+    },
+    {
+      "epoch": 0.09965010496850944,
+      "grad_norm": 0.5319507911786844,
+      "learning_rate": 4.937067340214137e-05,
+      "loss": 0.6426,
+      "step": 712
+    },
+    {
+      "epoch": 0.09979006298110567,
+      "grad_norm": 0.5222169180449785,
+      "learning_rate": 4.936814399131648e-05,
+      "loss": 0.6738,
+      "step": 713
+    },
+    {
+      "epoch": 0.09993002099370189,
+      "grad_norm": 0.5105935227729471,
+      "learning_rate": 4.936560957259077e-05,
+      "loss": 0.6668,
+      "step": 714
+    },
+    {
+      "epoch": 0.1000699790062981,
+      "grad_norm": 0.5163958109604428,
+      "learning_rate": 4.9363070146485044e-05,
+      "loss": 0.6986,
+      "step": 715
+    },
+    {
+      "epoch": 0.10020993701889433,
+      "grad_norm": 0.5021401060702906,
+      "learning_rate": 4.936052571352122e-05,
+      "loss": 0.6468,
+      "step": 716
+    },
+    {
+      "epoch": 0.10034989503149055,
+      "grad_norm": 0.4685192631602669,
+      "learning_rate": 4.9357976274222185e-05,
+      "loss": 0.6609,
+      "step": 717
+    },
+    {
+      "epoch": 0.10048985304408678,
+      "grad_norm": 0.48726170019620224,
+      "learning_rate": 4.935542182911188e-05,
+      "loss": 0.6761,
+      "step": 718
+    },
+    {
+      "epoch": 0.10062981105668299,
+      "grad_norm": 0.48175436628421253,
+      "learning_rate": 4.935286237871527e-05,
+      "loss": 0.645,
+      "step": 719
+    },
+    {
+      "epoch": 0.10076976906927922,
+      "grad_norm": 0.49385381693887187,
+      "learning_rate": 4.935029792355834e-05,
+      "loss": 0.6243,
+      "step": 720
+    },
+    {
+      "epoch": 0.10090972708187544,
+      "grad_norm": 0.4743721480397386,
+      "learning_rate": 4.934772846416812e-05,
+      "loss": 0.6999,
+      "step": 721
+    },
+    {
+      "epoch": 0.10104968509447165,
+      "grad_norm": 0.49107710855800335,
+      "learning_rate": 4.934515400107266e-05,
+      "loss": 0.6637,
+      "step": 722
+    },
+    {
+      "epoch": 0.10118964310706788,
+      "grad_norm": 0.49159205634657455,
+      "learning_rate": 4.934257453480103e-05,
+      "loss": 0.6218,
+      "step": 723
+    },
+    {
+      "epoch": 0.1013296011196641,
+      "grad_norm": 0.49141594694855456,
+      "learning_rate": 4.933999006588335e-05,
+      "loss": 0.674,
+      "step": 724
+    },
+    {
+      "epoch": 0.10146955913226033,
+      "grad_norm": 0.44737735842488807,
+      "learning_rate": 4.933740059485075e-05,
+      "loss": 0.6446,
+      "step": 725
+    },
+    {
+      "epoch": 0.10160951714485654,
+      "grad_norm": 0.4856033404998713,
+      "learning_rate": 4.9334806122235376e-05,
+      "loss": 0.6557,
+      "step": 726
+    },
+    {
+      "epoch": 0.10174947515745277,
+      "grad_norm": 0.48253432462131046,
+      "learning_rate": 4.933220664857044e-05,
+      "loss": 0.6452,
+      "step": 727
+    },
+    {
+      "epoch": 0.10188943317004899,
+      "grad_norm": 0.5065000807014411,
+      "learning_rate": 4.9329602174390153e-05,
+      "loss": 0.6936,
+      "step": 728
+    },
+    {
+      "epoch": 0.1020293911826452,
+      "grad_norm": 0.46004243943931644,
+      "learning_rate": 4.932699270022976e-05,
+      "loss": 0.6666,
+      "step": 729
+    },
+    {
+      "epoch": 0.10216934919524143,
+      "grad_norm": 0.4729412035182845,
+      "learning_rate": 4.932437822662553e-05,
+      "loss": 0.619,
+      "step": 730
+    },
+    {
+      "epoch": 0.10230930720783765,
+      "grad_norm": 0.4829984181448435,
+      "learning_rate": 4.932175875411478e-05,
+      "loss": 0.6387,
+      "step": 731
+    },
+    {
+      "epoch": 0.10244926522043386,
+      "grad_norm": 0.4826696857482811,
+      "learning_rate": 4.931913428323581e-05,
+      "loss": 0.6433,
+      "step": 732
+    },
+    {
+      "epoch": 0.10258922323303009,
+      "grad_norm": 0.45648188028698533,
+      "learning_rate": 4.931650481452801e-05,
+      "loss": 0.6035,
+      "step": 733
+    },
+    {
+      "epoch": 0.10272918124562631,
+      "grad_norm": 0.4691725557215748,
+      "learning_rate": 4.931387034853173e-05,
+      "loss": 0.6362,
+      "step": 734
+    },
+    {
+      "epoch": 0.10286913925822254,
+      "grad_norm": 0.49511735363849857,
+      "learning_rate": 4.93112308857884e-05,
+      "loss": 0.6128,
+      "step": 735
+    },
+    {
+      "epoch": 0.10300909727081875,
+      "grad_norm": 0.5067311529916844,
+      "learning_rate": 4.9308586426840454e-05,
+      "loss": 0.6581,
+      "step": 736
+    },
+    {
+      "epoch": 0.10314905528341498,
+      "grad_norm": 0.49461941413672095,
+      "learning_rate": 4.9305936972231346e-05,
+      "loss": 0.6545,
+      "step": 737
+    },
+    {
+      "epoch": 0.1032890132960112,
+      "grad_norm": 0.46083627671042743,
+      "learning_rate": 4.9303282522505565e-05,
+      "loss": 0.6384,
+      "step": 738
+    },
+    {
+      "epoch": 0.10342897130860741,
+      "grad_norm": 0.48817979810198836,
+      "learning_rate": 4.930062307820865e-05,
+      "loss": 0.6392,
+      "step": 739
+    },
+    {
+      "epoch": 0.10356892932120364,
+      "grad_norm": 0.4835520562427775,
+      "learning_rate": 4.9297958639887116e-05,
+      "loss": 0.6333,
+      "step": 740
+    },
+    {
+      "epoch": 0.10370888733379986,
+      "grad_norm": 0.5167572191592301,
+      "learning_rate": 4.929528920808854e-05,
+      "loss": 0.635,
+      "step": 741
+    },
+    {
+      "epoch": 0.10384884534639609,
+      "grad_norm": 0.4595745343183344,
+      "learning_rate": 4.9292614783361536e-05,
+      "loss": 0.6376,
+      "step": 742
+    },
+    {
+      "epoch": 0.1039888033589923,
+      "grad_norm": 0.43859471943688627,
+      "learning_rate": 4.92899353662557e-05,
+      "loss": 0.6585,
+      "step": 743
+    },
+    {
+      "epoch": 0.10412876137158852,
+      "grad_norm": 0.4953753280512198,
+      "learning_rate": 4.928725095732169e-05,
+      "loss": 0.6396,
+      "step": 744
+    },
+    {
+      "epoch": 0.10426871938418475,
+      "grad_norm": 0.4880410414717604,
+      "learning_rate": 4.928456155711117e-05,
+      "loss": 0.6403,
+      "step": 745
+    },
+    {
+      "epoch": 0.10440867739678096,
+      "grad_norm": 0.4737693275569935,
+      "learning_rate": 4.928186716617686e-05,
+      "loss": 0.6301,
+      "step": 746
+    },
+    {
+      "epoch": 0.10454863540937719,
+      "grad_norm": 0.5034731514446417,
+      "learning_rate": 4.927916778507248e-05,
+      "loss": 0.6729,
+      "step": 747
+    },
+    {
+      "epoch": 0.10468859342197341,
+      "grad_norm": 0.46500169965537136,
+      "learning_rate": 4.9276463414352757e-05,
+      "loss": 0.6296,
+      "step": 748
+    },
+    {
+      "epoch": 0.10482855143456962,
+      "grad_norm": 0.48081470356503164,
+      "learning_rate": 4.9273754054573496e-05,
+      "loss": 0.614,
+      "step": 749
+    },
+    {
+      "epoch": 0.10496850944716585,
+      "grad_norm": 0.4506016223155311,
+      "learning_rate": 4.927103970629148e-05,
+      "loss": 0.6313,
+      "step": 750
+    },
+    {
+      "epoch": 0.10510846745976207,
+      "grad_norm": 0.4806095102495363,
+      "learning_rate": 4.926832037006453e-05,
+      "loss": 0.6862,
+      "step": 751
+    },
+    {
+      "epoch": 0.1052484254723583,
+      "grad_norm": 0.5264224078411822,
+      "learning_rate": 4.926559604645152e-05,
+      "loss": 0.6387,
+      "step": 752
+    },
+    {
+      "epoch": 0.10538838348495451,
+      "grad_norm": 0.4993673425121446,
+      "learning_rate": 4.9262866736012304e-05,
+      "loss": 0.6761,
+      "step": 753
+    },
+    {
+      "epoch": 0.10552834149755073,
+      "grad_norm": 0.5130179673023382,
+      "learning_rate": 4.926013243930779e-05,
+      "loss": 0.6582,
+      "step": 754
+    },
+    {
+      "epoch": 0.10566829951014696,
+      "grad_norm": 0.5161547254751858,
+      "learning_rate": 4.925739315689991e-05,
+      "loss": 0.65,
+      "step": 755
+    },
+    {
+      "epoch": 0.10580825752274317,
+      "grad_norm": 3.5932005324517964,
+      "learning_rate": 4.925464888935162e-05,
+      "loss": 0.6722,
+      "step": 756
+    },
+    {
+      "epoch": 0.1059482155353394,
+      "grad_norm": 0.47238634638015287,
+      "learning_rate": 4.925189963722687e-05,
+      "loss": 0.6734,
+      "step": 757
+    },
+    {
+      "epoch": 0.10608817354793562,
+      "grad_norm": 0.4958610236105621,
+      "learning_rate": 4.924914540109068e-05,
+      "loss": 0.705,
+      "step": 758
+    },
+    {
+      "epoch": 0.10622813156053185,
+      "grad_norm": 0.47904033540016167,
+      "learning_rate": 4.924638618150906e-05,
+      "loss": 0.6611,
+      "step": 759
+    },
+    {
+      "epoch": 0.10636808957312806,
+      "grad_norm": 0.4595734849503931,
+      "learning_rate": 4.924362197904908e-05,
+      "loss": 0.637,
+      "step": 760
+    },
+    {
+      "epoch": 0.10650804758572428,
+      "grad_norm": 0.4706049650368603,
+      "learning_rate": 4.924085279427879e-05,
+      "loss": 0.607,
+      "step": 761
+    },
+    {
+      "epoch": 0.10664800559832051,
+      "grad_norm": 0.4829263701881818,
+      "learning_rate": 4.923807862776728e-05,
+      "loss": 0.6311,
+      "step": 762
+    },
+    {
+      "epoch": 0.10678796361091672,
+      "grad_norm": 0.4658603465917324,
+      "learning_rate": 4.92352994800847e-05,
+      "loss": 0.6575,
+      "step": 763
+    },
+    {
+      "epoch": 0.10692792162351294,
+      "grad_norm": 0.48975156023608535,
+      "learning_rate": 4.9232515351802166e-05,
+      "loss": 0.6754,
+      "step": 764
+    },
+    {
+      "epoch": 0.10706787963610917,
+      "grad_norm": 0.4574370958416068,
+      "learning_rate": 4.922972624349185e-05,
+      "loss": 0.6556,
+      "step": 765
+    },
+    {
+      "epoch": 0.1072078376487054,
+      "grad_norm": 0.47320816538424315,
+      "learning_rate": 4.922693215572695e-05,
+      "loss": 0.6808,
+      "step": 766
+    },
+    {
+      "epoch": 0.1073477956613016,
+      "grad_norm": 0.4938979904546163,
+      "learning_rate": 4.9224133089081675e-05,
+      "loss": 0.6396,
+      "step": 767
+    },
+    {
+      "epoch": 0.10748775367389783,
+      "grad_norm": 0.4815436050811902,
+      "learning_rate": 4.922132904413126e-05,
+      "loss": 0.6885,
+      "step": 768
+    },
+    {
+      "epoch": 0.10762771168649406,
+      "grad_norm": 0.4526917995255154,
+      "learning_rate": 4.921852002145196e-05,
+      "loss": 0.6613,
+      "step": 769
+    },
+    {
+      "epoch": 0.10776766969909027,
+      "grad_norm": 0.48451528868058824,
+      "learning_rate": 4.921570602162108e-05,
+      "loss": 0.6273,
+      "step": 770
+    },
+    {
+      "epoch": 0.1079076277116865,
+      "grad_norm": 0.44910249964176424,
+      "learning_rate": 4.921288704521689e-05,
+      "loss": 0.6445,
+      "step": 771
+    },
+    {
+      "epoch": 0.10804758572428272,
+      "grad_norm": 0.4803448598027863,
+      "learning_rate": 4.9210063092818755e-05,
+      "loss": 0.6369,
+      "step": 772
+    },
+    {
+      "epoch": 0.10818754373687893,
+      "grad_norm": 0.500963530964286,
+      "learning_rate": 4.9207234165007e-05,
+      "loss": 0.6572,
+      "step": 773
+    },
+    {
+      "epoch": 0.10832750174947516,
+      "grad_norm": 0.4929298522075444,
+      "learning_rate": 4.920440026236301e-05,
+      "loss": 0.6346,
+      "step": 774
+    },
+    {
+      "epoch": 0.10846745976207138,
+      "grad_norm": 0.45662320216713365,
+      "learning_rate": 4.920156138546917e-05,
+      "loss": 0.6593,
+      "step": 775
+    },
+    {
+      "epoch": 0.1086074177746676,
+      "grad_norm": 0.46973861985195237,
+      "learning_rate": 4.919871753490891e-05,
+      "loss": 0.6906,
+      "step": 776
+    },
+    {
+      "epoch": 0.10874737578726382,
+      "grad_norm": 0.47814740177046505,
+      "learning_rate": 4.919586871126667e-05,
+      "loss": 0.646,
+      "step": 777
+    },
+    {
+      "epoch": 0.10888733379986004,
+      "grad_norm": 0.5173386479220408,
+      "learning_rate": 4.91930149151279e-05,
+      "loss": 0.6629,
+      "step": 778
+    },
+    {
+      "epoch": 0.10902729181245627,
+      "grad_norm": 0.47384818148671254,
+      "learning_rate": 4.9190156147079094e-05,
+      "loss": 0.6458,
+      "step": 779
+    },
+    {
+      "epoch": 0.10916724982505248,
+      "grad_norm": 0.4998121952757577,
+      "learning_rate": 4.918729240770775e-05,
+      "loss": 0.6627,
+      "step": 780
+    },
+    {
+      "epoch": 0.1093072078376487,
+      "grad_norm": 0.49480732704046404,
+      "learning_rate": 4.918442369760241e-05,
+      "loss": 0.6503,
+      "step": 781
+    },
+    {
+      "epoch": 0.10944716585024493,
+      "grad_norm": 0.48089249137852297,
+      "learning_rate": 4.9181550017352615e-05,
+      "loss": 0.6334,
+      "step": 782
+    },
+    {
+      "epoch": 0.10958712386284115,
+      "grad_norm": 0.46605303630427286,
+      "learning_rate": 4.917867136754893e-05,
+      "loss": 0.598,
+      "step": 783
+    },
+    {
+      "epoch": 0.10972708187543737,
+      "grad_norm": 0.4766867229123813,
+      "learning_rate": 4.9175787748782955e-05,
+      "loss": 0.6478,
+      "step": 784
+    },
+    {
+      "epoch": 0.10986703988803359,
+      "grad_norm": 0.47706432841684887,
+      "learning_rate": 4.9172899161647295e-05,
+      "loss": 0.6579,
+      "step": 785
+    },
+    {
+      "epoch": 0.11000699790062982,
+      "grad_norm": 0.47350025007351243,
+      "learning_rate": 4.9170005606735594e-05,
+      "loss": 0.622,
+      "step": 786
+    },
+    {
+      "epoch": 0.11014695591322603,
+      "grad_norm": 0.4737951373989515,
+      "learning_rate": 4.9167107084642496e-05,
+      "loss": 0.6186,
+      "step": 787
+    },
+    {
+      "epoch": 0.11028691392582225,
+      "grad_norm": 0.4973982793226407,
+      "learning_rate": 4.916420359596368e-05,
+      "loss": 0.6041,
+      "step": 788
+    },
+    {
+      "epoch": 0.11042687193841848,
+      "grad_norm": 0.45566949261190215,
+      "learning_rate": 4.916129514129585e-05,
+      "loss": 0.6196,
+      "step": 789
+    },
+    {
+      "epoch": 0.11056682995101469,
+      "grad_norm": 0.47407807015226433,
+      "learning_rate": 4.915838172123671e-05,
+      "loss": 0.5968,
+      "step": 790
+    },
+    {
+      "epoch": 0.11070678796361091,
+      "grad_norm": 0.4508773406494618,
+      "learning_rate": 4.915546333638501e-05,
+      "loss": 0.6558,
+      "step": 791
+    },
+    {
+      "epoch": 0.11084674597620714,
+      "grad_norm": 0.47903719726456023,
+      "learning_rate": 4.915253998734051e-05,
+      "loss": 0.6281,
+      "step": 792
+    },
+    {
+      "epoch": 0.11098670398880336,
+      "grad_norm": 0.4864004849243157,
+      "learning_rate": 4.914961167470396e-05,
+      "loss": 0.6632,
+      "step": 793
+    },
+    {
+      "epoch": 0.11112666200139958,
+      "grad_norm": 0.4828968005003782,
+      "learning_rate": 4.9146678399077196e-05,
+      "loss": 0.6641,
+      "step": 794
+    },
+    {
+      "epoch": 0.1112666200139958,
+      "grad_norm": 0.5082783883297906,
+      "learning_rate": 4.9143740161063015e-05,
+      "loss": 0.6881,
+      "step": 795
+    },
+    {
+      "epoch": 0.11140657802659203,
+      "grad_norm": 0.4811155277843086,
+      "learning_rate": 4.914079696126526e-05,
+      "loss": 0.6542,
+      "step": 796
+    },
+    {
+      "epoch": 0.11154653603918824,
+      "grad_norm": 0.4697422705172665,
+      "learning_rate": 4.913784880028878e-05,
+      "loss": 0.6398,
+      "step": 797
+    },
+    {
+      "epoch": 0.11168649405178446,
+      "grad_norm": 0.48609960976538413,
+      "learning_rate": 4.9134895678739456e-05,
+      "loss": 0.6361,
+      "step": 798
+    },
+    {
+      "epoch": 0.11182645206438069,
+      "grad_norm": 0.4505917244706054,
+      "learning_rate": 4.9131937597224185e-05,
+      "loss": 0.6531,
+      "step": 799
+    },
+    {
+      "epoch": 0.11196641007697691,
+      "grad_norm": 0.4609463571323594,
+      "learning_rate": 4.912897455635089e-05,
+      "loss": 0.6091,
+      "step": 800
+    },
+    {
+      "epoch": 0.11210636808957312,
+      "grad_norm": 0.4633046052805601,
+      "learning_rate": 4.91260065567285e-05,
+      "loss": 0.6326,
+      "step": 801
+    },
+    {
+      "epoch": 0.11224632610216935,
+      "grad_norm": 0.48176393499918024,
+      "learning_rate": 4.912303359896697e-05,
+      "loss": 0.6256,
+      "step": 802
+    },
+    {
+      "epoch": 0.11238628411476558,
+      "grad_norm": 0.44238158654462206,
+      "learning_rate": 4.912005568367727e-05,
+      "loss": 0.617,
+      "step": 803
+    },
+    {
+      "epoch": 0.11252624212736179,
+      "grad_norm": 0.46784130274686425,
+      "learning_rate": 4.91170728114714e-05,
+      "loss": 0.6194,
+      "step": 804
+    },
+    {
+      "epoch": 0.11266620013995801,
+      "grad_norm": 0.4813594149903505,
+      "learning_rate": 4.9114084982962356e-05,
+      "loss": 0.6809,
+      "step": 805
+    },
+    {
+      "epoch": 0.11280615815255424,
+      "grad_norm": 0.4746068016576439,
+      "learning_rate": 4.911109219876417e-05,
+      "loss": 0.6482,
+      "step": 806
+    },
+    {
+      "epoch": 0.11294611616515045,
+      "grad_norm": 0.4837726952509842,
+      "learning_rate": 4.9108094459491916e-05,
+      "loss": 0.593,
+      "step": 807
+    },
+    {
+      "epoch": 0.11308607417774667,
+      "grad_norm": 0.4939387206527619,
+      "learning_rate": 4.910509176576162e-05,
+      "loss": 0.6372,
+      "step": 808
+    },
+    {
+      "epoch": 0.1132260321903429,
+      "grad_norm": 0.49309033390988954,
+      "learning_rate": 4.910208411819039e-05,
+      "loss": 0.6212,
+      "step": 809
+    },
+    {
+      "epoch": 0.11336599020293912,
+      "grad_norm": 0.48931195891148005,
+      "learning_rate": 4.909907151739633e-05,
+      "loss": 0.6668,
+      "step": 810
+    },
+    {
+      "epoch": 0.11350594821553533,
+      "grad_norm": 0.47504035399989575,
+      "learning_rate": 4.909605396399856e-05,
+      "loss": 0.66,
+      "step": 811
+    },
+    {
+      "epoch": 0.11364590622813156,
+      "grad_norm": 0.475509126808091,
+      "learning_rate": 4.90930314586172e-05,
+      "loss": 0.6693,
+      "step": 812
+    },
+    {
+      "epoch": 0.11378586424072779,
+      "grad_norm": 0.46395502226780827,
+      "learning_rate": 4.909000400187341e-05,
+      "loss": 0.6028,
+      "step": 813
+    },
+    {
+      "epoch": 0.113925822253324,
+      "grad_norm": 0.48820830409318683,
+      "learning_rate": 4.908697159438937e-05,
+      "loss": 0.62,
+      "step": 814
+    },
+    {
+      "epoch": 0.11406578026592022,
+      "grad_norm": 0.4909040802882437,
+      "learning_rate": 4.908393423678829e-05,
+      "loss": 0.6261,
+      "step": 815
+    },
+    {
+      "epoch": 0.11420573827851645,
+      "grad_norm": 0.4853877606631391,
+      "learning_rate": 4.908089192969434e-05,
+      "loss": 0.6574,
+      "step": 816
+    },
+    {
+      "epoch": 0.11434569629111267,
+      "grad_norm": 0.4604083047219368,
+      "learning_rate": 4.907784467373277e-05,
+      "loss": 0.6091,
+      "step": 817
+    },
+    {
+      "epoch": 0.11448565430370888,
+      "grad_norm": 0.4533069892751884,
+      "learning_rate": 4.9074792469529815e-05,
+      "loss": 0.6021,
+      "step": 818
+    },
+    {
+      "epoch": 0.11462561231630511,
+      "grad_norm": 0.4520978783038299,
+      "learning_rate": 4.907173531771273e-05,
+      "loss": 0.6427,
+      "step": 819
+    },
+    {
+      "epoch": 0.11476557032890133,
+      "grad_norm": 0.45052658755884484,
+      "learning_rate": 4.9068673218909796e-05,
+      "loss": 0.6294,
+      "step": 820
+    },
+    {
+      "epoch": 0.11490552834149755,
+      "grad_norm": 0.4793207729603878,
+      "learning_rate": 4.90656061737503e-05,
+      "loss": 0.6575,
+      "step": 821
+    },
+    {
+      "epoch": 0.11504548635409377,
+      "grad_norm": 0.4713321633366388,
+      "learning_rate": 4.906253418286456e-05,
+      "loss": 0.6507,
+      "step": 822
+    },
+    {
+      "epoch": 0.11518544436669,
+      "grad_norm": 0.4933569292067938,
+      "learning_rate": 4.90594572468839e-05,
+      "loss": 0.6084,
+      "step": 823
+    },
+    {
+      "epoch": 0.11532540237928622,
+      "grad_norm": 0.4727488359524651,
+      "learning_rate": 4.9056375366440654e-05,
+      "loss": 0.6047,
+      "step": 824
+    },
+    {
+      "epoch": 0.11546536039188243,
+      "grad_norm": 0.4790836053521045,
+      "learning_rate": 4.9053288542168185e-05,
+      "loss": 0.6496,
+      "step": 825
+    },
+    {
+      "epoch": 0.11560531840447866,
+      "grad_norm": 0.4832677203184707,
+      "learning_rate": 4.905019677470086e-05,
+      "loss": 0.6247,
+      "step": 826
+    },
+    {
+      "epoch": 0.11574527641707488,
+      "grad_norm": 0.48906934804664143,
+      "learning_rate": 4.9047100064674076e-05,
+      "loss": 0.6419,
+      "step": 827
+    },
+    {
+      "epoch": 0.1158852344296711,
+      "grad_norm": 0.44450056169083824,
+      "learning_rate": 4.904399841272423e-05,
+      "loss": 0.6284,
+      "step": 828
+    },
+    {
+      "epoch": 0.11602519244226732,
+      "grad_norm": 0.4736815158989831,
+      "learning_rate": 4.9040891819488766e-05,
+      "loss": 0.6254,
+      "step": 829
+    },
+    {
+      "epoch": 0.11616515045486354,
+      "grad_norm": 0.44945213217864644,
+      "learning_rate": 4.903778028560609e-05,
+      "loss": 0.6441,
+      "step": 830
+    },
+    {
+      "epoch": 0.11630510846745976,
+      "grad_norm": 0.5007544417322715,
+      "learning_rate": 4.903466381171568e-05,
+      "loss": 0.6898,
+      "step": 831
+    },
+    {
+      "epoch": 0.11644506648005598,
+      "grad_norm": 0.5281054716405869,
+      "learning_rate": 4.9031542398457974e-05,
+      "loss": 0.6338,
+      "step": 832
+    },
+    {
+      "epoch": 0.1165850244926522,
+      "grad_norm": 0.484478263124759,
+      "learning_rate": 4.902841604647448e-05,
+      "loss": 0.6164,
+      "step": 833
+    },
+    {
+      "epoch": 0.11672498250524843,
+      "grad_norm": 0.44994143276964366,
+      "learning_rate": 4.902528475640768e-05,
+      "loss": 0.622,
+      "step": 834
+    },
+    {
+      "epoch": 0.11686494051784464,
+      "grad_norm": 0.4865019036883684,
+      "learning_rate": 4.902214852890109e-05,
+      "loss": 0.6106,
+      "step": 835
+    },
+    {
+      "epoch": 0.11700489853044087,
+      "grad_norm": 0.4846126418090665,
+      "learning_rate": 4.9019007364599246e-05,
+      "loss": 0.6271,
+      "step": 836
+    },
+    {
+      "epoch": 0.11714485654303709,
+      "grad_norm": 0.48440494490884967,
+      "learning_rate": 4.901586126414768e-05,
+      "loss": 0.6221,
+      "step": 837
+    },
+    {
+      "epoch": 0.1172848145556333,
+      "grad_norm": 0.5014410717293758,
+      "learning_rate": 4.9012710228192946e-05,
+      "loss": 0.6326,
+      "step": 838
+    },
+    {
+      "epoch": 0.11742477256822953,
+      "grad_norm": 0.463246332033912,
+      "learning_rate": 4.9009554257382616e-05,
+      "loss": 0.5973,
+      "step": 839
+    },
+    {
+      "epoch": 0.11756473058082575,
+      "grad_norm": 0.5021414470675243,
+      "learning_rate": 4.900639335236527e-05,
+      "loss": 0.6163,
+      "step": 840
+    },
+    {
+      "epoch": 0.11770468859342198,
+      "grad_norm": 0.4751809205646141,
+      "learning_rate": 4.900322751379052e-05,
+      "loss": 0.5995,
+      "step": 841
+    },
+    {
+      "epoch": 0.11784464660601819,
+      "grad_norm": 0.495594879328081,
+      "learning_rate": 4.900005674230896e-05,
+      "loss": 0.6666,
+      "step": 842
+    },
+    {
+      "epoch": 0.11798460461861442,
+      "grad_norm": 0.46838396983160246,
+      "learning_rate": 4.899688103857223e-05,
+      "loss": 0.6391,
+      "step": 843
+    },
+    {
+      "epoch": 0.11812456263121064,
+      "grad_norm": 0.48479222227164487,
+      "learning_rate": 4.899370040323295e-05,
+      "loss": 0.6425,
+      "step": 844
+    },
+    {
+      "epoch": 0.11826452064380685,
+      "grad_norm": 0.4758054425265797,
+      "learning_rate": 4.899051483694481e-05,
+      "loss": 0.6722,
+      "step": 845
+    },
+    {
+      "epoch": 0.11840447865640308,
+      "grad_norm": 0.4517669680918226,
+      "learning_rate": 4.898732434036244e-05,
+      "loss": 0.6417,
+      "step": 846
+    },
+    {
+      "epoch": 0.1185444366689993,
+      "grad_norm": 0.4637034408072003,
+      "learning_rate": 4.898412891414153e-05,
+      "loss": 0.6472,
+      "step": 847
+    },
+    {
+      "epoch": 0.11868439468159551,
+      "grad_norm": 0.4483579373594681,
+      "learning_rate": 4.8980928558938774e-05,
+      "loss": 0.6631,
+      "step": 848
+    },
+    {
+      "epoch": 0.11882435269419174,
+      "grad_norm": 0.4607581614793959,
+      "learning_rate": 4.897772327541188e-05,
+      "loss": 0.6564,
+      "step": 849
+    },
+    {
+      "epoch": 0.11896431070678797,
+      "grad_norm": 0.4627218264459414,
+      "learning_rate": 4.8974513064219564e-05,
+      "loss": 0.6421,
+      "step": 850
+    },
+    {
+      "epoch": 0.11910426871938419,
+      "grad_norm": 0.4381385195746151,
+      "learning_rate": 4.897129792602156e-05,
+      "loss": 0.6443,
+      "step": 851
+    },
+    {
+      "epoch": 0.1192442267319804,
+      "grad_norm": 0.44805468338464166,
+      "learning_rate": 4.8968077861478606e-05,
+      "loss": 0.6289,
+      "step": 852
+    },
+    {
+      "epoch": 0.11938418474457663,
+      "grad_norm": 0.4419209751144886,
+      "learning_rate": 4.896485287125246e-05,
+      "loss": 0.6148,
+      "step": 853
+    },
+    {
+      "epoch": 0.11952414275717285,
+      "grad_norm": 0.46951206043414845,
+      "learning_rate": 4.896162295600589e-05,
+      "loss": 0.6552,
+      "step": 854
+    },
+    {
+      "epoch": 0.11966410076976906,
+      "grad_norm": 0.6796790512869556,
+      "learning_rate": 4.8958388116402685e-05,
+      "loss": 0.6547,
+      "step": 855
+    },
+    {
+      "epoch": 0.11980405878236529,
+      "grad_norm": 0.4900439088338446,
+      "learning_rate": 4.8955148353107625e-05,
+      "loss": 0.6618,
+      "step": 856
+    },
+    {
+      "epoch": 0.11994401679496151,
+      "grad_norm": 0.5670961555372047,
+      "learning_rate": 4.8951903666786514e-05,
+      "loss": 0.6234,
+      "step": 857
+    },
+    {
+      "epoch": 0.12008397480755774,
+      "grad_norm": 0.4709993547132971,
+      "learning_rate": 4.894865405810618e-05,
+      "loss": 0.6528,
+      "step": 858
+    },
+    {
+      "epoch": 0.12022393282015395,
+      "grad_norm": 0.5029700711110836,
+      "learning_rate": 4.8945399527734436e-05,
+      "loss": 0.6508,
+      "step": 859
+    },
+    {
+      "epoch": 0.12036389083275018,
+      "grad_norm": 0.471348795948391,
+      "learning_rate": 4.8942140076340135e-05,
+      "loss": 0.6303,
+      "step": 860
+    },
+    {
+      "epoch": 0.1205038488453464,
+      "grad_norm": 0.47522173933299294,
+      "learning_rate": 4.893887570459312e-05,
+      "loss": 0.6503,
+      "step": 861
+    },
+    {
+      "epoch": 0.12064380685794261,
+      "grad_norm": 0.44742680670952745,
+      "learning_rate": 4.893560641316425e-05,
+      "loss": 0.6306,
+      "step": 862
+    },
+    {
+      "epoch": 0.12078376487053884,
+      "grad_norm": 0.46472651593001396,
+      "learning_rate": 4.89323322027254e-05,
+      "loss": 0.6322,
+      "step": 863
+    },
+    {
+      "epoch": 0.12092372288313506,
+      "grad_norm": 0.4573234609819833,
+      "learning_rate": 4.8929053073949456e-05,
+      "loss": 0.6014,
+      "step": 864
+    },
+    {
+      "epoch": 0.12106368089573127,
+      "grad_norm": 0.5281892171792254,
+      "learning_rate": 4.892576902751031e-05,
+      "loss": 0.666,
+      "step": 865
+    },
+    {
+      "epoch": 0.1212036389083275,
+      "grad_norm": 0.46580727241797515,
+      "learning_rate": 4.8922480064082864e-05,
+      "loss": 0.634,
+      "step": 866
+    },
+    {
+      "epoch": 0.12134359692092372,
+      "grad_norm": 0.46889330922237543,
+      "learning_rate": 4.8919186184343046e-05,
+      "loss": 0.6171,
+      "step": 867
+    },
+    {
+      "epoch": 0.12148355493351995,
+      "grad_norm": 0.4593728936842765,
+      "learning_rate": 4.891588738896776e-05,
+      "loss": 0.569,
+      "step": 868
+    },
+    {
+      "epoch": 0.12162351294611616,
+      "grad_norm": 0.4554124468260883,
+      "learning_rate": 4.891258367863497e-05,
+      "loss": 0.6426,
+      "step": 869
+    },
+    {
+      "epoch": 0.12176347095871239,
+      "grad_norm": 0.4613451131787414,
+      "learning_rate": 4.890927505402359e-05,
+      "loss": 0.6738,
+      "step": 870
+    },
+    {
+      "epoch": 0.12190342897130861,
+      "grad_norm": 0.4775869179312642,
+      "learning_rate": 4.8905961515813604e-05,
+      "loss": 0.6403,
+      "step": 871
+    },
+    {
+      "epoch": 0.12204338698390482,
+      "grad_norm": 0.4836931023229791,
+      "learning_rate": 4.890264306468596e-05,
+      "loss": 0.6472,
+      "step": 872
+    },
+    {
+      "epoch": 0.12218334499650105,
+      "grad_norm": 0.4661703058075851,
+      "learning_rate": 4.8899319701322646e-05,
+      "loss": 0.6656,
+      "step": 873
+    },
+    {
+      "epoch": 0.12232330300909727,
+      "grad_norm": 0.46602771617049493,
+      "learning_rate": 4.889599142640663e-05,
+      "loss": 0.6471,
+      "step": 874
+    },
+    {
+      "epoch": 0.1224632610216935,
+      "grad_norm": 0.46986425878775284,
+      "learning_rate": 4.889265824062193e-05,
+      "loss": 0.6538,
+      "step": 875
+    },
+    {
+      "epoch": 0.12260321903428971,
+      "grad_norm": 0.4350487912771349,
+      "learning_rate": 4.888932014465352e-05,
+      "loss": 0.5965,
+      "step": 876
+    },
+    {
+      "epoch": 0.12274317704688593,
+      "grad_norm": 0.45434867025136694,
+      "learning_rate": 4.888597713918743e-05,
+      "loss": 0.6091,
+      "step": 877
+    },
+    {
+      "epoch": 0.12288313505948216,
+      "grad_norm": 0.45288960332540396,
+      "learning_rate": 4.888262922491069e-05,
+      "loss": 0.6633,
+      "step": 878
+    },
+    {
+      "epoch": 0.12302309307207837,
+      "grad_norm": 0.46432887825126745,
+      "learning_rate": 4.887927640251132e-05,
+      "loss": 0.6597,
+      "step": 879
+    },
+    {
+      "epoch": 0.1231630510846746,
+      "grad_norm": 0.45520528507797114,
+      "learning_rate": 4.887591867267836e-05,
+      "loss": 0.6403,
+      "step": 880
+    },
+    {
+      "epoch": 0.12330300909727082,
+      "grad_norm": 0.4575189583703599,
+      "learning_rate": 4.887255603610185e-05,
+      "loss": 0.6531,
+      "step": 881
+    },
+    {
+      "epoch": 0.12344296710986705,
+      "grad_norm": 0.46480538006139166,
+      "learning_rate": 4.8869188493472854e-05,
+      "loss": 0.5969,
+      "step": 882
+    },
+    {
+      "epoch": 0.12358292512246326,
+      "grad_norm": 0.46565682699578237,
+      "learning_rate": 4.886581604548344e-05,
+      "loss": 0.6334,
+      "step": 883
+    },
+    {
+      "epoch": 0.12372288313505948,
+      "grad_norm": 0.4615157170447177,
+      "learning_rate": 4.8862438692826675e-05,
+      "loss": 0.6317,
+      "step": 884
+    },
+    {
+      "epoch": 0.12386284114765571,
+      "grad_norm": 0.45719602979674634,
+      "learning_rate": 4.885905643619664e-05,
+      "loss": 0.6615,
+      "step": 885
+    },
+    {
+      "epoch": 0.12400279916025192,
+      "grad_norm": 0.4666726722691908,
+      "learning_rate": 4.885566927628842e-05,
+      "loss": 0.6192,
+      "step": 886
+    },
+    {
+      "epoch": 0.12414275717284814,
+      "grad_norm": 0.49622217670592067,
+      "learning_rate": 4.8852277213798106e-05,
+      "loss": 0.6263,
+      "step": 887
+    },
+    {
+      "epoch": 0.12428271518544437,
+      "grad_norm": 0.47891535770456606,
+      "learning_rate": 4.8848880249422815e-05,
+      "loss": 0.6435,
+      "step": 888
+    },
+    {
+      "epoch": 0.12442267319804058,
+      "grad_norm": 0.478301172680431,
+      "learning_rate": 4.884547838386065e-05,
+      "loss": 0.671,
+      "step": 889
+    },
+    {
+      "epoch": 0.1245626312106368,
+      "grad_norm": 0.44791777443248404,
+      "learning_rate": 4.884207161781074e-05,
+      "loss": 0.5783,
+      "step": 890
+    },
+    {
+      "epoch": 0.12470258922323303,
+      "grad_norm": 0.4476426364893327,
+      "learning_rate": 4.883865995197319e-05,
+      "loss": 0.6164,
+      "step": 891
+    },
+    {
+      "epoch": 0.12484254723582926,
+      "grad_norm": 0.47381113823924553,
+      "learning_rate": 4.8835243387049144e-05,
+      "loss": 0.6427,
+      "step": 892
+    },
+    {
+      "epoch": 0.12498250524842547,
+      "grad_norm": 0.47937356676709797,
+      "learning_rate": 4.8831821923740745e-05,
+      "loss": 0.6783,
+      "step": 893
+    },
+    {
+      "epoch": 0.1251224632610217,
+      "grad_norm": 0.4634097903429328,
+      "learning_rate": 4.882839556275113e-05,
+      "loss": 0.6131,
+      "step": 894
+    },
+    {
+      "epoch": 0.12526242127361792,
+      "grad_norm": 0.4867484756619001,
+      "learning_rate": 4.8824964304784446e-05,
+      "loss": 0.649,
+      "step": 895
+    },
+    {
+      "epoch": 0.12540237928621414,
+      "grad_norm": 0.450881363852382,
+      "learning_rate": 4.882152815054587e-05,
+      "loss": 0.6561,
+      "step": 896
+    },
+    {
+      "epoch": 0.12554233729881037,
+      "grad_norm": 0.46114414970423057,
+      "learning_rate": 4.881808710074155e-05,
+      "loss": 0.6104,
+      "step": 897
+    },
+    {
+      "epoch": 0.12568229531140657,
+      "grad_norm": 0.45063006067640693,
+      "learning_rate": 4.881464115607865e-05,
+      "loss": 0.6445,
+      "step": 898
+    },
+    {
+      "epoch": 0.1258222533240028,
+      "grad_norm": 0.47221223029390763,
+      "learning_rate": 4.8811190317265376e-05,
+      "loss": 0.6246,
+      "step": 899
+    },
+    {
+      "epoch": 0.12596221133659902,
+      "grad_norm": 0.4781011312710143,
+      "learning_rate": 4.880773458501089e-05,
+      "loss": 0.6222,
+      "step": 900
+    },
+    {
+      "epoch": 0.12610216934919524,
+      "grad_norm": 0.45326509799123327,
+      "learning_rate": 4.8804273960025376e-05,
+      "loss": 0.651,
+      "step": 901
+    },
+    {
+      "epoch": 0.12624212736179147,
+      "grad_norm": 0.482160143883641,
+      "learning_rate": 4.880080844302004e-05,
+      "loss": 0.6286,
+      "step": 902
+    },
+    {
+      "epoch": 0.1263820853743877,
+      "grad_norm": 0.4630226354527779,
+      "learning_rate": 4.879733803470707e-05,
+      "loss": 0.636,
+      "step": 903
+    },
+    {
+      "epoch": 0.12652204338698392,
+      "grad_norm": 0.4731355871687041,
+      "learning_rate": 4.8793862735799676e-05,
+      "loss": 0.6418,
+      "step": 904
+    },
+    {
+      "epoch": 0.12666200139958012,
+      "grad_norm": 0.467195164605955,
+      "learning_rate": 4.879038254701207e-05,
+      "loss": 0.6306,
+      "step": 905
+    },
+    {
+      "epoch": 0.12680195941217634,
+      "grad_norm": 0.441351930375645,
+      "learning_rate": 4.878689746905946e-05,
+      "loss": 0.5974,
+      "step": 906
+    },
+    {
+      "epoch": 0.12694191742477257,
+      "grad_norm": 0.4714053662846153,
+      "learning_rate": 4.878340750265807e-05,
+      "loss": 0.6506,
+      "step": 907
+    },
+    {
+      "epoch": 0.1270818754373688,
+      "grad_norm": 0.4459074638142588,
+      "learning_rate": 4.877991264852512e-05,
+      "loss": 0.6474,
+      "step": 908
+    },
+    {
+      "epoch": 0.12722183344996502,
+      "grad_norm": 0.469617227274652,
+      "learning_rate": 4.877641290737884e-05,
+      "loss": 0.6656,
+      "step": 909
+    },
+    {
+      "epoch": 0.12736179146256124,
+      "grad_norm": 0.4557573307622375,
+      "learning_rate": 4.8772908279938464e-05,
+      "loss": 0.6908,
+      "step": 910
+    },
+    {
+      "epoch": 0.12750174947515747,
+      "grad_norm": 0.46053886117091863,
+      "learning_rate": 4.8769398766924226e-05,
+      "loss": 0.6619,
+      "step": 911
+    },
+    {
+      "epoch": 0.12764170748775366,
+      "grad_norm": 0.5409321614433314,
+      "learning_rate": 4.876588436905736e-05,
+      "loss": 0.6306,
+      "step": 912
+    },
+    {
+      "epoch": 0.1277816655003499,
+      "grad_norm": 0.4525734524473738,
+      "learning_rate": 4.8762365087060117e-05,
+      "loss": 0.6163,
+      "step": 913
+    },
+    {
+      "epoch": 0.12792162351294611,
+      "grad_norm": 0.44910287840912927,
+      "learning_rate": 4.8758840921655744e-05,
+      "loss": 0.5945,
+      "step": 914
+    },
+    {
+      "epoch": 0.12806158152554234,
+      "grad_norm": 0.4451813661618601,
+      "learning_rate": 4.8755311873568505e-05,
+      "loss": 0.613,
+      "step": 915
+    },
+    {
+      "epoch": 0.12820153953813856,
+      "grad_norm": 0.46559156418594866,
+      "learning_rate": 4.8751777943523634e-05,
+      "loss": 0.6414,
+      "step": 916
+    },
+    {
+      "epoch": 0.1283414975507348,
+      "grad_norm": 0.4947913830806307,
+      "learning_rate": 4.874823913224741e-05,
+      "loss": 0.6461,
+      "step": 917
+    },
+    {
+      "epoch": 0.128481455563331,
+      "grad_norm": 0.4786066687436742,
+      "learning_rate": 4.874469544046707e-05,
+      "loss": 0.6474,
+      "step": 918
+    },
+    {
+      "epoch": 0.1286214135759272,
+      "grad_norm": 0.4603973225093796,
+      "learning_rate": 4.8741146868910906e-05,
+      "loss": 0.66,
+      "step": 919
+    },
+    {
+      "epoch": 0.12876137158852344,
+      "grad_norm": 0.48095107605482856,
+      "learning_rate": 4.8737593418308156e-05,
+      "loss": 0.6534,
+      "step": 920
+    },
+    {
+      "epoch": 0.12890132960111966,
+      "grad_norm": 0.4623235801737886,
+      "learning_rate": 4.8734035089389115e-05,
+      "loss": 0.628,
+      "step": 921
+    },
+    {
+      "epoch": 0.1290412876137159,
+      "grad_norm": 0.4930237562486678,
+      "learning_rate": 4.873047188288505e-05,
+      "loss": 0.6409,
+      "step": 922
+    },
+    {
+      "epoch": 0.1291812456263121,
+      "grad_norm": 0.46495991811324616,
+      "learning_rate": 4.8726903799528234e-05,
+      "loss": 0.6293,
+      "step": 923
+    },
+    {
+      "epoch": 0.12932120363890834,
+      "grad_norm": 0.46304983740233435,
+      "learning_rate": 4.872333084005194e-05,
+      "loss": 0.6472,
+      "step": 924
+    },
+    {
+      "epoch": 0.12946116165150454,
+      "grad_norm": 0.44884446203631584,
+      "learning_rate": 4.871975300519045e-05,
+      "loss": 0.6371,
+      "step": 925
+    },
+    {
+      "epoch": 0.12960111966410076,
+      "grad_norm": 0.4675721819496966,
+      "learning_rate": 4.8716170295679053e-05,
+      "loss": 0.6653,
+      "step": 926
+    },
+    {
+      "epoch": 0.129741077676697,
+      "grad_norm": 0.47573888293206407,
+      "learning_rate": 4.8712582712254016e-05,
+      "loss": 0.6267,
+      "step": 927
+    },
+    {
+      "epoch": 0.1298810356892932,
+      "grad_norm": 0.46375806207869347,
+      "learning_rate": 4.870899025565264e-05,
+      "loss": 0.6467,
+      "step": 928
+    },
+    {
+      "epoch": 0.13002099370188944,
+      "grad_norm": 0.457474154948321,
+      "learning_rate": 4.8705392926613205e-05,
+      "loss": 0.6156,
+      "step": 929
+    },
+    {
+      "epoch": 0.13016095171448566,
+      "grad_norm": 0.44866595177001034,
+      "learning_rate": 4.870179072587499e-05,
+      "loss": 0.5962,
+      "step": 930
+    },
+    {
+      "epoch": 0.1303009097270819,
+      "grad_norm": 0.4703083744568931,
+      "learning_rate": 4.86981836541783e-05,
+      "loss": 0.6309,
+      "step": 931
+    },
+    {
+      "epoch": 0.13044086773967808,
+      "grad_norm": 0.6018616664400933,
+      "learning_rate": 4.869457171226441e-05,
+      "loss": 0.642,
+      "step": 932
+    },
+    {
+      "epoch": 0.1305808257522743,
+      "grad_norm": 0.46944918935149577,
+      "learning_rate": 4.869095490087562e-05,
+      "loss": 0.6273,
+      "step": 933
+    },
+    {
+      "epoch": 0.13072078376487054,
+      "grad_norm": 0.44148958617555917,
+      "learning_rate": 4.868733322075522e-05,
+      "loss": 0.6378,
+      "step": 934
+    },
+    {
+      "epoch": 0.13086074177746676,
+      "grad_norm": 0.4395583432649483,
+      "learning_rate": 4.86837066726475e-05,
+      "loss": 0.6202,
+      "step": 935
+    },
+    {
+      "epoch": 0.13100069979006299,
+      "grad_norm": 0.44852952705339616,
+      "learning_rate": 4.868007525729775e-05,
+      "loss": 0.6416,
+      "step": 936
+    },
+    {
+      "epoch": 0.1311406578026592,
+      "grad_norm": 0.4487713642870549,
+      "learning_rate": 4.8676438975452274e-05,
+      "loss": 0.6263,
+      "step": 937
+    },
+    {
+      "epoch": 0.13128061581525544,
+      "grad_norm": 0.4648656677758751,
+      "learning_rate": 4.8672797827858355e-05,
+      "loss": 0.6164,
+      "step": 938
+    },
+    {
+      "epoch": 0.13142057382785163,
+      "grad_norm": 0.4608329526156638,
+      "learning_rate": 4.866915181526428e-05,
+      "loss": 0.6692,
+      "step": 939
+    },
+    {
+      "epoch": 0.13156053184044786,
+      "grad_norm": 0.4654710571009042,
+      "learning_rate": 4.866550093841936e-05,
+      "loss": 0.6222,
+      "step": 940
+    },
+    {
+      "epoch": 0.13170048985304408,
+      "grad_norm": 0.4603807269880903,
+      "learning_rate": 4.866184519807387e-05,
+      "loss": 0.6165,
+      "step": 941
+    },
+    {
+      "epoch": 0.1318404478656403,
+      "grad_norm": 0.47496860741108315,
+      "learning_rate": 4.865818459497911e-05,
+      "loss": 0.6267,
+      "step": 942
+    },
+    {
+      "epoch": 0.13198040587823653,
+      "grad_norm": 0.46532589078199804,
+      "learning_rate": 4.8654519129887364e-05,
+      "loss": 0.6293,
+      "step": 943
+    },
+    {
+      "epoch": 0.13212036389083276,
+      "grad_norm": 0.4590484934275629,
+      "learning_rate": 4.865084880355193e-05,
+      "loss": 0.6231,
+      "step": 944
+    },
+    {
+      "epoch": 0.13226032190342898,
+      "grad_norm": 0.47530046553555577,
+      "learning_rate": 4.86471736167271e-05,
+      "loss": 0.6133,
+      "step": 945
+    },
+    {
+      "epoch": 0.13240027991602518,
+      "grad_norm": 0.4527393981074044,
+      "learning_rate": 4.864349357016815e-05,
+      "loss": 0.6124,
+      "step": 946
+    },
+    {
+      "epoch": 0.1325402379286214,
+      "grad_norm": 0.46113362427703974,
+      "learning_rate": 4.863980866463138e-05,
+      "loss": 0.6456,
+      "step": 947
+    },
+    {
+      "epoch": 0.13268019594121763,
+      "grad_norm": 0.509871579156769,
+      "learning_rate": 4.8636118900874064e-05,
+      "loss": 0.6229,
+      "step": 948
+    },
+    {
+      "epoch": 0.13282015395381386,
+      "grad_norm": 0.4784635752003754,
+      "learning_rate": 4.86324242796545e-05,
+      "loss": 0.6596,
+      "step": 949
+    },
+    {
+      "epoch": 0.13296011196641008,
+      "grad_norm": 0.4414714405275805,
+      "learning_rate": 4.862872480173195e-05,
+      "loss": 0.6459,
+      "step": 950
+    },
+    {
+      "epoch": 0.1331000699790063,
+      "grad_norm": 0.4587139434548765,
+      "learning_rate": 4.862502046786671e-05,
+      "loss": 0.6126,
+      "step": 951
+    },
+    {
+      "epoch": 0.1332400279916025,
+      "grad_norm": 0.42461073662577176,
+      "learning_rate": 4.8621311278820056e-05,
+      "loss": 0.6353,
+      "step": 952
+    },
+    {
+      "epoch": 0.13337998600419873,
+      "grad_norm": 0.4324208829491985,
+      "learning_rate": 4.861759723535426e-05,
+      "loss": 0.6499,
+      "step": 953
+    },
+    {
+      "epoch": 0.13351994401679496,
+      "grad_norm": 0.4640550519654056,
+      "learning_rate": 4.86138783382326e-05,
+      "loss": 0.5896,
+      "step": 954
+    },
+    {
+      "epoch": 0.13365990202939118,
+      "grad_norm": 0.4638965920279924,
+      "learning_rate": 4.8610154588219345e-05,
+      "loss": 0.6201,
+      "step": 955
+    },
+    {
+      "epoch": 0.1337998600419874,
+      "grad_norm": 0.4695822317863163,
+      "learning_rate": 4.860642598607976e-05,
+      "loss": 0.6751,
+      "step": 956
+    },
+    {
+      "epoch": 0.13393981805458363,
+      "grad_norm": 0.4600752660391693,
+      "learning_rate": 4.860269253258012e-05,
+      "loss": 0.6654,
+      "step": 957
+    },
+    {
+      "epoch": 0.13407977606717986,
+      "grad_norm": 0.47652178074145596,
+      "learning_rate": 4.859895422848767e-05,
+      "loss": 0.659,
+      "step": 958
+    },
+    {
+      "epoch": 0.13421973407977605,
+      "grad_norm": 0.445979113412429,
+      "learning_rate": 4.859521107457069e-05,
+      "loss": 0.632,
+      "step": 959
+    },
+    {
+      "epoch": 0.13435969209237228,
+      "grad_norm": 0.45688889669250304,
+      "learning_rate": 4.859146307159842e-05,
+      "loss": 0.5705,
+      "step": 960
+    },
+    {
+      "epoch": 0.1344996501049685,
+      "grad_norm": 0.4567888614063869,
+      "learning_rate": 4.858771022034112e-05,
+      "loss": 0.6607,
+      "step": 961
+    },
+    {
+      "epoch": 0.13463960811756473,
+      "grad_norm": 0.4495345209098684,
+      "learning_rate": 4.858395252157004e-05,
+      "loss": 0.6415,
+      "step": 962
+    },
+    {
+      "epoch": 0.13477956613016095,
+      "grad_norm": 0.48532112670125027,
+      "learning_rate": 4.858018997605742e-05,
+      "loss": 0.6088,
+      "step": 963
+    },
+    {
+      "epoch": 0.13491952414275718,
+      "grad_norm": 0.440987511045589,
+      "learning_rate": 4.8576422584576514e-05,
+      "loss": 0.6129,
+      "step": 964
+    },
+    {
+      "epoch": 0.1350594821553534,
+      "grad_norm": 0.4669491385905085,
+      "learning_rate": 4.8572650347901544e-05,
+      "loss": 0.6191,
+      "step": 965
+    },
+    {
+      "epoch": 0.1351994401679496,
+      "grad_norm": 0.4665408837691419,
+      "learning_rate": 4.856887326680774e-05,
+      "loss": 0.6317,
+      "step": 966
+    },
+    {
+      "epoch": 0.13533939818054583,
+      "grad_norm": 0.48893208349101197,
+      "learning_rate": 4.856509134207136e-05,
+      "loss": 0.595,
+      "step": 967
+    },
+    {
+      "epoch": 0.13547935619314205,
+      "grad_norm": 0.4438150829055967,
+      "learning_rate": 4.856130457446959e-05,
+      "loss": 0.6065,
+      "step": 968
+    },
+    {
+      "epoch": 0.13561931420573828,
+      "grad_norm": 0.5115425422953542,
+      "learning_rate": 4.8557512964780674e-05,
+      "loss": 0.6448,
+      "step": 969
+    },
+    {
+      "epoch": 0.1357592722183345,
+      "grad_norm": 0.44504395647634876,
+      "learning_rate": 4.855371651378382e-05,
+      "loss": 0.6164,
+      "step": 970
+    },
+    {
+      "epoch": 0.13589923023093073,
+      "grad_norm": 0.5076347594918873,
+      "learning_rate": 4.854991522225923e-05,
+      "loss": 0.6207,
+      "step": 971
+    },
+    {
+      "epoch": 0.13603918824352695,
+      "grad_norm": 0.4602778279540183,
+      "learning_rate": 4.854610909098812e-05,
+      "loss": 0.5923,
+      "step": 972
+    },
+    {
+      "epoch": 0.13617914625612315,
+      "grad_norm": 0.49106399964431596,
+      "learning_rate": 4.8542298120752684e-05,
+      "loss": 0.5914,
+      "step": 973
+    },
+    {
+      "epoch": 0.13631910426871938,
+      "grad_norm": 0.44686877889055854,
+      "learning_rate": 4.85384823123361e-05,
+      "loss": 0.6163,
+      "step": 974
+    },
+    {
+      "epoch": 0.1364590622813156,
+      "grad_norm": 0.46546574820842096,
+      "learning_rate": 4.8534661666522584e-05,
+      "loss": 0.6625,
+      "step": 975
+    },
+    {
+      "epoch": 0.13659902029391183,
+      "grad_norm": 0.6072066700290619,
+      "learning_rate": 4.8530836184097297e-05,
+      "loss": 0.6048,
+      "step": 976
+    },
+    {
+      "epoch": 0.13673897830650805,
+      "grad_norm": 0.49583731751388566,
+      "learning_rate": 4.852700586584642e-05,
+      "loss": 0.6445,
+      "step": 977
+    },
+    {
+      "epoch": 0.13687893631910428,
+      "grad_norm": 0.4600581543242378,
+      "learning_rate": 4.852317071255712e-05,
+      "loss": 0.6276,
+      "step": 978
+    },
+    {
+      "epoch": 0.1370188943317005,
+      "grad_norm": 0.47143173605603533,
+      "learning_rate": 4.851933072501756e-05,
+      "loss": 0.6707,
+      "step": 979
+    },
+    {
+      "epoch": 0.1371588523442967,
+      "grad_norm": 0.43226562343994834,
+      "learning_rate": 4.85154859040169e-05,
+      "loss": 0.5993,
+      "step": 980
+    },
+    {
+      "epoch": 0.13729881035689293,
+      "grad_norm": 0.5032598563086278,
+      "learning_rate": 4.8511636250345294e-05,
+      "loss": 0.6388,
+      "step": 981
+    },
+    {
+      "epoch": 0.13743876836948915,
+      "grad_norm": 0.6566397034726049,
+      "learning_rate": 4.850778176479387e-05,
+      "loss": 0.626,
+      "step": 982
+    },
+    {
+      "epoch": 0.13757872638208538,
+      "grad_norm": 0.45320474044463904,
+      "learning_rate": 4.850392244815478e-05,
+      "loss": 0.6222,
+      "step": 983
+    },
+    {
+      "epoch": 0.1377186843946816,
+      "grad_norm": 0.4637289824351019,
+      "learning_rate": 4.8500058301221144e-05,
+      "loss": 0.6357,
+      "step": 984
+    },
+    {
+      "epoch": 0.13785864240727783,
+      "grad_norm": 0.4502718394375495,
+      "learning_rate": 4.849618932478708e-05,
+      "loss": 0.6161,
+      "step": 985
+    },
+    {
+      "epoch": 0.13799860041987405,
+      "grad_norm": 0.45481795684352794,
+      "learning_rate": 4.849231551964771e-05,
+      "loss": 0.6251,
+      "step": 986
+    },
+    {
+      "epoch": 0.13813855843247025,
+      "grad_norm": 0.505134985018315,
+      "learning_rate": 4.8488436886599144e-05,
+      "loss": 0.6358,
+      "step": 987
+    },
+    {
+      "epoch": 0.13827851644506647,
+      "grad_norm": 0.48619324349330834,
+      "learning_rate": 4.8484553426438464e-05,
+      "loss": 0.6115,
+      "step": 988
+    },
+    {
+      "epoch": 0.1384184744576627,
+      "grad_norm": 0.4650935755290328,
+      "learning_rate": 4.8480665139963774e-05,
+      "loss": 0.6004,
+      "step": 989
+    },
+    {
+      "epoch": 0.13855843247025892,
+      "grad_norm": 0.44207287204019446,
+      "learning_rate": 4.847677202797415e-05,
+      "loss": 0.597,
+      "step": 990
+    },
+    {
+      "epoch": 0.13869839048285515,
+      "grad_norm": 0.4629268299609161,
+      "learning_rate": 4.8472874091269674e-05,
+      "loss": 0.6414,
+      "step": 991
+    },
+    {
+      "epoch": 0.13883834849545137,
+      "grad_norm": 0.46784298534877244,
+      "learning_rate": 4.84689713306514e-05,
+      "loss": 0.6263,
+      "step": 992
+    },
+    {
+      "epoch": 0.13897830650804757,
+      "grad_norm": 0.448599541259752,
+      "learning_rate": 4.8465063746921395e-05,
+      "loss": 0.6395,
+      "step": 993
+    },
+    {
+      "epoch": 0.1391182645206438,
+      "grad_norm": 0.4582611197304845,
+      "learning_rate": 4.8461151340882706e-05,
+      "loss": 0.6533,
+      "step": 994
+    },
+    {
+      "epoch": 0.13925822253324002,
+      "grad_norm": 0.49137392446019373,
+      "learning_rate": 4.845723411333936e-05,
+      "loss": 0.6327,
+      "step": 995
+    },
+    {
+      "epoch": 0.13939818054583625,
+      "grad_norm": 0.4651834345736373,
+      "learning_rate": 4.84533120650964e-05,
+      "loss": 0.6444,
+      "step": 996
+    },
+    {
+      "epoch": 0.13953813855843247,
+      "grad_norm": 0.43613025151896734,
+      "learning_rate": 4.844938519695984e-05,
+      "loss": 0.6036,
+      "step": 997
+    },
+    {
+      "epoch": 0.1396780965710287,
+      "grad_norm": 0.4547252123073849,
+      "learning_rate": 4.84454535097367e-05,
+      "loss": 0.6715,
+      "step": 998
+    },
+    {
+      "epoch": 0.13981805458362492,
+      "grad_norm": 0.4411221787105127,
+      "learning_rate": 4.8441517004234975e-05,
+      "loss": 0.6207,
+      "step": 999
+    },
+    {
+      "epoch": 0.13995801259622112,
+      "grad_norm": 0.4652981257021806,
+      "learning_rate": 4.8437575681263656e-05,
+      "loss": 0.628,
+      "step": 1000
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 7145,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 175094048915456.0,
+  "train_batch_size": 1,
+  "trial_name": null,
+  "trial_params": null
+}