diff --git "a/last-checkpoint/trainer_state.json" "b/last-checkpoint/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/last-checkpoint/trainer_state.json"
@@ -0,0 +1,4605 @@
+{
+  "best_metric": null,
+  "best_model_checkpoint": null,
+  "epoch": 0.7610993657505285,
+  "eval_steps": 20,
+  "global_step": 600,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0,
+      "eval_accuracy": 0.7339955849889624,
+      "eval_f1": 0.2445141065830721,
+      "eval_loss": 0.5994934439659119,
+      "eval_precision": 0.6,
+      "eval_recall": 0.15354330708661418,
+      "eval_runtime": 52.3271,
+      "eval_samples_per_second": 5.294,
+      "eval_steps_per_second": 0.172,
+      "step": 0
+    },
+    {
+      "epoch": 0.0012684989429175475,
+      "grad_norm": 2.466090202331543,
+      "learning_rate": 2.5316455696202533e-07,
+      "loss": 0.7205,
+      "step": 1
+    },
+    {
+      "epoch": 0.002536997885835095,
+      "grad_norm": 1.759368896484375,
+      "learning_rate": 5.063291139240507e-07,
+      "loss": 0.6242,
+      "step": 2
+    },
+    {
+      "epoch": 0.0038054968287526427,
+      "grad_norm": 1.8663183450698853,
+      "learning_rate": 7.59493670886076e-07,
+      "loss": 0.6844,
+      "step": 3
+    },
+    {
+      "epoch": 0.00507399577167019,
+      "grad_norm": 2.109837293624878,
+      "learning_rate": 1.0126582278481013e-06,
+      "loss": 0.6534,
+      "step": 4
+    },
+    {
+      "epoch": 0.006342494714587738,
+      "grad_norm": 2.1454806327819824,
+      "learning_rate": 1.2658227848101267e-06,
+      "loss": 0.6766,
+      "step": 5
+    },
+    {
+      "epoch": 0.007610993657505285,
+      "grad_norm": 2.128432512283325,
+      "learning_rate": 1.518987341772152e-06,
+      "loss": 0.6845,
+      "step": 6
+    },
+    {
+      "epoch": 0.008879492600422832,
+      "grad_norm": 2.348109722137451,
+      "learning_rate": 1.7721518987341774e-06,
+      "loss": 0.6689,
+      "step": 7
+    },
+    {
+      "epoch": 0.01014799154334038,
+      "grad_norm": 2.1332168579101562,
+      "learning_rate": 2.0253164556962026e-06,
+      "loss": 0.6957,
+      "step": 8
+    },
+    {
+      "epoch": 0.011416490486257928,
+      "grad_norm": 2.224982976913452,
+      "learning_rate": 2.278481012658228e-06,
+      "loss": 0.6278,
+      "step": 9
+    },
+    {
+      "epoch": 0.012684989429175475,
+      "grad_norm": 1.956710934638977,
+      "learning_rate": 2.5316455696202535e-06,
+      "loss": 0.6717,
+      "step": 10
+    },
+    {
+      "epoch": 0.013953488372093023,
+      "grad_norm": 2.072009325027466,
+      "learning_rate": 2.7848101265822785e-06,
+      "loss": 0.6349,
+      "step": 11
+    },
+    {
+      "epoch": 0.01522198731501057,
+      "grad_norm": 1.9736876487731934,
+      "learning_rate": 3.037974683544304e-06,
+      "loss": 0.6542,
+      "step": 12
+    },
+    {
+      "epoch": 0.01649048625792812,
+      "grad_norm": 2.037200689315796,
+      "learning_rate": 3.2911392405063294e-06,
+      "loss": 0.5939,
+      "step": 13
+    },
+    {
+      "epoch": 0.017758985200845664,
+      "grad_norm": 2.2937052249908447,
+      "learning_rate": 3.544303797468355e-06,
+      "loss": 0.6793,
+      "step": 14
+    },
+    {
+      "epoch": 0.019027484143763214,
+      "grad_norm": 2.500192403793335,
+      "learning_rate": 3.7974683544303802e-06,
+      "loss": 0.6384,
+      "step": 15
+    },
+    {
+      "epoch": 0.02029598308668076,
+      "grad_norm": 2.0733048915863037,
+      "learning_rate": 4.050632911392405e-06,
+      "loss": 0.6608,
+      "step": 16
+    },
+    {
+      "epoch": 0.02156448202959831,
+      "grad_norm": 2.258061170578003,
+      "learning_rate": 4.303797468354431e-06,
+      "loss": 0.7128,
+      "step": 17
+    },
+    {
+      "epoch": 0.022832980972515855,
+      "grad_norm": 1.9154382944107056,
+      "learning_rate": 4.556962025316456e-06,
+      "loss": 0.6418,
+      "step": 18
+    },
+    {
+      "epoch": 0.024101479915433405,
+      "grad_norm": 2.2492971420288086,
+      "learning_rate": 4.8101265822784815e-06,
+      "loss": 0.6382,
+      "step": 19
+    },
+    {
+      "epoch": 0.02536997885835095,
+      "grad_norm": 1.9588701725006104,
+      "learning_rate": 5.063291139240507e-06,
+      "loss": 0.6057,
+      "step": 20
+    },
+    {
+      "epoch": 0.02536997885835095,
+      "eval_accuracy": 0.7373068432671082,
+      "eval_f1": 0.25157232704402516,
+      "eval_loss": 0.591836154460907,
+      "eval_precision": 0.625,
+      "eval_recall": 0.15748031496062992,
+      "eval_runtime": 52.8906,
+      "eval_samples_per_second": 5.237,
+      "eval_steps_per_second": 0.17,
+      "step": 20
+    },
+    {
+      "epoch": 0.0266384778012685,
+      "grad_norm": 2.0472803115844727,
+      "learning_rate": 5.3164556962025316e-06,
+      "loss": 0.6482,
+      "step": 21
+    },
+    {
+      "epoch": 0.027906976744186046,
+      "grad_norm": 2.471733570098877,
+      "learning_rate": 5.569620253164557e-06,
+      "loss": 0.6122,
+      "step": 22
+    },
+    {
+      "epoch": 0.029175475687103596,
+      "grad_norm": 2.041800022125244,
+      "learning_rate": 5.8227848101265824e-06,
+      "loss": 0.7014,
+      "step": 23
+    },
+    {
+      "epoch": 0.03044397463002114,
+      "grad_norm": 1.9668631553649902,
+      "learning_rate": 6.075949367088608e-06,
+      "loss": 0.6282,
+      "step": 24
+    },
+    {
+      "epoch": 0.03171247357293869,
+      "grad_norm": 1.7734192609786987,
+      "learning_rate": 6.329113924050634e-06,
+      "loss": 0.6623,
+      "step": 25
+    },
+    {
+      "epoch": 0.03298097251585624,
+      "grad_norm": 2.1032049655914307,
+      "learning_rate": 6.582278481012659e-06,
+      "loss": 0.6185,
+      "step": 26
+    },
+    {
+      "epoch": 0.034249471458773786,
+      "grad_norm": 1.768869400024414,
+      "learning_rate": 6.835443037974684e-06,
+      "loss": 0.6106,
+      "step": 27
+    },
+    {
+      "epoch": 0.03551797040169133,
+      "grad_norm": 1.874267339706421,
+      "learning_rate": 7.08860759493671e-06,
+      "loss": 0.6634,
+      "step": 28
+    },
+    {
+      "epoch": 0.03678646934460888,
+      "grad_norm": 1.702706217765808,
+      "learning_rate": 7.341772151898735e-06,
+      "loss": 0.5611,
+      "step": 29
+    },
+    {
+      "epoch": 0.03805496828752643,
+      "grad_norm": 1.583400845527649,
+      "learning_rate": 7.5949367088607605e-06,
+      "loss": 0.5927,
+      "step": 30
+    },
+    {
+      "epoch": 0.03932346723044398,
+      "grad_norm": 1.9068820476531982,
+      "learning_rate": 7.848101265822786e-06,
+      "loss": 0.6602,
+      "step": 31
+    },
+    {
+      "epoch": 0.04059196617336152,
+      "grad_norm": 1.8527209758758545,
+      "learning_rate": 8.10126582278481e-06,
+      "loss": 0.5679,
+      "step": 32
+    },
+    {
+      "epoch": 0.04186046511627907,
+      "grad_norm": 2.0851504802703857,
+      "learning_rate": 8.354430379746837e-06,
+      "loss": 0.5437,
+      "step": 33
+    },
+    {
+      "epoch": 0.04312896405919662,
+      "grad_norm": 1.8767458200454712,
+      "learning_rate": 8.607594936708861e-06,
+      "loss": 0.5738,
+      "step": 34
+    },
+    {
+      "epoch": 0.04439746300211417,
+      "grad_norm": 2.203929901123047,
+      "learning_rate": 8.860759493670886e-06,
+      "loss": 0.6506,
+      "step": 35
+    },
+    {
+      "epoch": 0.04566596194503171,
+      "grad_norm": 1.9388232231140137,
+      "learning_rate": 9.113924050632912e-06,
+      "loss": 0.5631,
+      "step": 36
+    },
+    {
+      "epoch": 0.04693446088794926,
+      "grad_norm": 1.4374593496322632,
+      "learning_rate": 9.367088607594937e-06,
+      "loss": 0.529,
+      "step": 37
+    },
+    {
+      "epoch": 0.04820295983086681,
+      "grad_norm": 1.4115264415740967,
+      "learning_rate": 9.620253164556963e-06,
+      "loss": 0.501,
+      "step": 38
+    },
+    {
+      "epoch": 0.04947145877378435,
+      "grad_norm": 1.4638575315475464,
+      "learning_rate": 9.87341772151899e-06,
+      "loss": 0.5635,
+      "step": 39
+    },
+    {
+      "epoch": 0.0507399577167019,
+      "grad_norm": 1.6112034320831299,
+      "learning_rate": 1.0126582278481014e-05,
+      "loss": 0.5356,
+      "step": 40
+    },
+    {
+      "epoch": 0.0507399577167019,
+      "eval_accuracy": 0.7516556291390728,
+      "eval_f1": 0.3283582089552239,
+      "eval_loss": 0.5521447062492371,
+      "eval_precision": 0.6790123456790124,
+      "eval_recall": 0.21653543307086615,
+      "eval_runtime": 52.1586,
+      "eval_samples_per_second": 5.311,
+      "eval_steps_per_second": 0.173,
+      "step": 40
+    },
+    {
+      "epoch": 0.05200845665961945,
+      "grad_norm": 1.6395788192749023,
+      "learning_rate": 1.0379746835443039e-05,
+      "loss": 0.5908,
+      "step": 41
+    },
+    {
+      "epoch": 0.053276955602537,
+      "grad_norm": 1.3024195432662964,
+      "learning_rate": 1.0632911392405063e-05,
+      "loss": 0.5528,
+      "step": 42
+    },
+    {
+      "epoch": 0.05454545454545454,
+      "grad_norm": 1.3428813219070435,
+      "learning_rate": 1.088607594936709e-05,
+      "loss": 0.5177,
+      "step": 43
+    },
+    {
+      "epoch": 0.05581395348837209,
+      "grad_norm": 2.2030651569366455,
+      "learning_rate": 1.1139240506329114e-05,
+      "loss": 0.5615,
+      "step": 44
+    },
+    {
+      "epoch": 0.05708245243128964,
+      "grad_norm": 1.491940975189209,
+      "learning_rate": 1.139240506329114e-05,
+      "loss": 0.5186,
+      "step": 45
+    },
+    {
+      "epoch": 0.05835095137420719,
+      "grad_norm": 2.706481695175171,
+      "learning_rate": 1.1645569620253165e-05,
+      "loss": 0.5678,
+      "step": 46
+    },
+    {
+      "epoch": 0.059619450317124734,
+      "grad_norm": 1.7402979135513306,
+      "learning_rate": 1.189873417721519e-05,
+      "loss": 0.5478,
+      "step": 47
+    },
+    {
+      "epoch": 0.06088794926004228,
+      "grad_norm": 1.7591081857681274,
+      "learning_rate": 1.2151898734177216e-05,
+      "loss": 0.5296,
+      "step": 48
+    },
+    {
+      "epoch": 0.06215644820295983,
+      "grad_norm": 1.5506426095962524,
+      "learning_rate": 1.240506329113924e-05,
+      "loss": 0.5186,
+      "step": 49
+    },
+    {
+      "epoch": 0.06342494714587738,
+      "grad_norm": 1.4400081634521484,
+      "learning_rate": 1.2658227848101268e-05,
+      "loss": 0.5178,
+      "step": 50
+    },
+    {
+      "epoch": 0.06469344608879493,
+      "grad_norm": 1.6742987632751465,
+      "learning_rate": 1.2911392405063293e-05,
+      "loss": 0.5777,
+      "step": 51
+    },
+    {
+      "epoch": 0.06596194503171247,
+      "grad_norm": 1.9172340631484985,
+      "learning_rate": 1.3164556962025317e-05,
+      "loss": 0.5675,
+      "step": 52
+    },
+    {
+      "epoch": 0.06723044397463002,
+      "grad_norm": 2.08730149269104,
+      "learning_rate": 1.3417721518987344e-05,
+      "loss": 0.5832,
+      "step": 53
+    },
+    {
+      "epoch": 0.06849894291754757,
+      "grad_norm": 1.4685484170913696,
+      "learning_rate": 1.3670886075949368e-05,
+      "loss": 0.4869,
+      "step": 54
+    },
+    {
+      "epoch": 0.06976744186046512,
+      "grad_norm": 1.7338310480117798,
+      "learning_rate": 1.3924050632911395e-05,
+      "loss": 0.5132,
+      "step": 55
+    },
+    {
+      "epoch": 0.07103594080338266,
+      "grad_norm": 2.538879156112671,
+      "learning_rate": 1.417721518987342e-05,
+      "loss": 0.5726,
+      "step": 56
+    },
+    {
+      "epoch": 0.07230443974630021,
+      "grad_norm": 2.6859829425811768,
+      "learning_rate": 1.4430379746835444e-05,
+      "loss": 0.5126,
+      "step": 57
+    },
+    {
+      "epoch": 0.07357293868921776,
+      "grad_norm": 2.0791189670562744,
+      "learning_rate": 1.468354430379747e-05,
+      "loss": 0.5027,
+      "step": 58
+    },
+    {
+      "epoch": 0.07484143763213531,
+      "grad_norm": 2.305135726928711,
+      "learning_rate": 1.4936708860759495e-05,
+      "loss": 0.5724,
+      "step": 59
+    },
+    {
+      "epoch": 0.07610993657505286,
+      "grad_norm": 1.8028602600097656,
+      "learning_rate": 1.5189873417721521e-05,
+      "loss": 0.5141,
+      "step": 60
+    },
+    {
+      "epoch": 0.07610993657505286,
+      "eval_accuracy": 0.7626931567328918,
+      "eval_f1": 0.5454545454545454,
+      "eval_loss": 0.502124547958374,
+      "eval_precision": 0.589041095890411,
+      "eval_recall": 0.5078740157480315,
+      "eval_runtime": 52.5689,
+      "eval_samples_per_second": 5.269,
+      "eval_steps_per_second": 0.171,
+      "step": 60
+    },
+    {
+      "epoch": 0.0773784355179704,
+      "grad_norm": 2.500967025756836,
+      "learning_rate": 1.5443037974683546e-05,
+      "loss": 0.4764,
+      "step": 61
+    },
+    {
+      "epoch": 0.07864693446088795,
+      "grad_norm": 1.9441800117492676,
+      "learning_rate": 1.5696202531645572e-05,
+      "loss": 0.575,
+      "step": 62
+    },
+    {
+      "epoch": 0.0799154334038055,
+      "grad_norm": 2.2089622020721436,
+      "learning_rate": 1.5949367088607598e-05,
+      "loss": 0.4854,
+      "step": 63
+    },
+    {
+      "epoch": 0.08118393234672304,
+      "grad_norm": 1.727460503578186,
+      "learning_rate": 1.620253164556962e-05,
+      "loss": 0.4829,
+      "step": 64
+    },
+    {
+      "epoch": 0.0824524312896406,
+      "grad_norm": 2.3789401054382324,
+      "learning_rate": 1.6455696202531647e-05,
+      "loss": 0.5193,
+      "step": 65
+    },
+    {
+      "epoch": 0.08372093023255814,
+      "grad_norm": 1.817589282989502,
+      "learning_rate": 1.6708860759493674e-05,
+      "loss": 0.5115,
+      "step": 66
+    },
+    {
+      "epoch": 0.08498942917547568,
+      "grad_norm": 1.985190987586975,
+      "learning_rate": 1.6962025316455696e-05,
+      "loss": 0.5309,
+      "step": 67
+    },
+    {
+      "epoch": 0.08625792811839324,
+      "grad_norm": 2.5799453258514404,
+      "learning_rate": 1.7215189873417723e-05,
+      "loss": 0.455,
+      "step": 68
+    },
+    {
+      "epoch": 0.08752642706131078,
+      "grad_norm": 1.9298819303512573,
+      "learning_rate": 1.746835443037975e-05,
+      "loss": 0.4207,
+      "step": 69
+    },
+    {
+      "epoch": 0.08879492600422834,
+      "grad_norm": 1.6359901428222656,
+      "learning_rate": 1.7721518987341772e-05,
+      "loss": 0.4287,
+      "step": 70
+    },
+    {
+      "epoch": 0.09006342494714588,
+      "grad_norm": 3.6526267528533936,
+      "learning_rate": 1.7974683544303798e-05,
+      "loss": 0.5659,
+      "step": 71
+    },
+    {
+      "epoch": 0.09133192389006342,
+      "grad_norm": 2.615048408508301,
+      "learning_rate": 1.8227848101265824e-05,
+      "loss": 0.4652,
+      "step": 72
+    },
+    {
+      "epoch": 0.09260042283298098,
+      "grad_norm": 2.702597141265869,
+      "learning_rate": 1.848101265822785e-05,
+      "loss": 0.404,
+      "step": 73
+    },
+    {
+      "epoch": 0.09386892177589852,
+      "grad_norm": 2.3155453205108643,
+      "learning_rate": 1.8734177215189874e-05,
+      "loss": 0.4769,
+      "step": 74
+    },
+    {
+      "epoch": 0.09513742071881606,
+      "grad_norm": 2.9285926818847656,
+      "learning_rate": 1.89873417721519e-05,
+      "loss": 0.4264,
+      "step": 75
+    },
+    {
+      "epoch": 0.09640591966173362,
+      "grad_norm": 1.856951117515564,
+      "learning_rate": 1.9240506329113926e-05,
+      "loss": 0.4323,
+      "step": 76
+    },
+    {
+      "epoch": 0.09767441860465116,
+      "grad_norm": 1.961795449256897,
+      "learning_rate": 1.949367088607595e-05,
+      "loss": 0.3662,
+      "step": 77
+    },
+    {
+      "epoch": 0.0989429175475687,
+      "grad_norm": 2.0264902114868164,
+      "learning_rate": 1.974683544303798e-05,
+      "loss": 0.3677,
+      "step": 78
+    },
+    {
+      "epoch": 0.10021141649048626,
+      "grad_norm": 2.0433084964752197,
+      "learning_rate": 2e-05,
+      "loss": 0.4287,
+      "step": 79
+    },
+    {
+      "epoch": 0.1014799154334038,
+      "grad_norm": 2.14385724067688,
+      "learning_rate": 1.999990183050232e-05,
+      "loss": 0.3594,
+      "step": 80
+    },
+    {
+      "epoch": 0.1014799154334038,
+      "eval_accuracy": 0.7980132450331126,
+      "eval_f1": 0.5413533834586466,
+      "eval_loss": 0.4426968991756439,
+      "eval_precision": 0.7448275862068966,
+      "eval_recall": 0.4251968503937008,
+      "eval_runtime": 50.9956,
+      "eval_samples_per_second": 5.432,
+      "eval_steps_per_second": 0.176,
+      "step": 80
+    },
+    {
+      "epoch": 0.10274841437632136,
+      "grad_norm": 3.1786880493164062,
+      "learning_rate": 1.9999607323936722e-05,
+      "loss": 0.4351,
+      "step": 81
+    },
+    {
+      "epoch": 0.1040169133192389,
+      "grad_norm": 2.1636154651641846,
+      "learning_rate": 1.9999116486085527e-05,
+      "loss": 0.3564,
+      "step": 82
+    },
+    {
+      "epoch": 0.10528541226215644,
+      "grad_norm": 2.686142921447754,
+      "learning_rate": 1.999842932658579e-05,
+      "loss": 0.3642,
+      "step": 83
+    },
+    {
+      "epoch": 0.106553911205074,
+      "grad_norm": 3.7086920738220215,
+      "learning_rate": 1.9997545858929135e-05,
+      "loss": 0.3724,
+      "step": 84
+    },
+    {
+      "epoch": 0.10782241014799154,
+      "grad_norm": 3.1358773708343506,
+      "learning_rate": 1.9996466100461473e-05,
+      "loss": 0.302,
+      "step": 85
+    },
+    {
+      "epoch": 0.10909090909090909,
+      "grad_norm": 5.590456962585449,
+      "learning_rate": 1.9995190072382676e-05,
+      "loss": 0.3492,
+      "step": 86
+    },
+    {
+      "epoch": 0.11035940803382664,
+      "grad_norm": 3.6015288829803467,
+      "learning_rate": 1.9993717799746152e-05,
+      "loss": 0.3645,
+      "step": 87
+    },
+    {
+      "epoch": 0.11162790697674418,
+      "grad_norm": 4.678368091583252,
+      "learning_rate": 1.9992049311458353e-05,
+      "loss": 0.254,
+      "step": 88
+    },
+    {
+      "epoch": 0.11289640591966173,
+      "grad_norm": 3.184845447540283,
+      "learning_rate": 1.9990184640278212e-05,
+      "loss": 0.3374,
+      "step": 89
+    },
+    {
+      "epoch": 0.11416490486257928,
+      "grad_norm": 2.6212780475616455,
+      "learning_rate": 1.998812382281649e-05,
+      "loss": 0.3546,
+      "step": 90
+    },
+    {
+      "epoch": 0.11543340380549683,
+      "grad_norm": 4.838207721710205,
+      "learning_rate": 1.9985866899535076e-05,
+      "loss": 0.3734,
+      "step": 91
+    },
+    {
+      "epoch": 0.11670190274841438,
+      "grad_norm": 3.615828037261963,
+      "learning_rate": 1.9983413914746175e-05,
+      "loss": 0.3111,
+      "step": 92
+    },
+    {
+      "epoch": 0.11797040169133192,
+      "grad_norm": 3.7412872314453125,
+      "learning_rate": 1.998076491661144e-05,
+      "loss": 0.3374,
+      "step": 93
+    },
+    {
+      "epoch": 0.11923890063424947,
+      "grad_norm": 5.049579620361328,
+      "learning_rate": 1.997791995714104e-05,
+      "loss": 0.3831,
+      "step": 94
+    },
+    {
+      "epoch": 0.12050739957716702,
+      "grad_norm": 3.293745279312134,
+      "learning_rate": 1.9974879092192618e-05,
+      "loss": 0.2982,
+      "step": 95
+    },
+    {
+      "epoch": 0.12177589852008457,
+      "grad_norm": 4.374386310577393,
+      "learning_rate": 1.9971642381470207e-05,
+      "loss": 0.367,
+      "step": 96
+    },
+    {
+      "epoch": 0.12304439746300211,
+      "grad_norm": 4.193471431732178,
+      "learning_rate": 1.996820988852307e-05,
+      "loss": 0.3096,
+      "step": 97
+    },
+    {
+      "epoch": 0.12431289640591967,
+      "grad_norm": 4.823442459106445,
+      "learning_rate": 1.9964581680744425e-05,
+      "loss": 0.3464,
+      "step": 98
+    },
+    {
+      "epoch": 0.12558139534883722,
+      "grad_norm": 4.056851387023926,
+      "learning_rate": 1.9960757829370138e-05,
+      "loss": 0.3194,
+      "step": 99
+    },
+    {
+      "epoch": 0.12684989429175475,
+      "grad_norm": 5.191165924072266,
+      "learning_rate": 1.995673840947732e-05,
+      "loss": 0.3988,
+      "step": 100
+    },
+    {
+      "epoch": 0.12684989429175475,
+      "eval_accuracy": 0.8245033112582781,
+      "eval_f1": 0.6310904872389791,
+      "eval_loss": 0.40671199560165405,
+      "eval_precision": 0.768361581920904,
+      "eval_recall": 0.5354330708661418,
+      "eval_runtime": 50.9945,
+      "eval_samples_per_second": 5.432,
+      "eval_steps_per_second": 0.176,
+      "step": 100
+    },
+    {
+      "epoch": 0.1281183932346723,
+      "grad_norm": 3.31948184967041,
+      "learning_rate": 1.9952523499982866e-05,
+      "loss": 0.2718,
+      "step": 101
+    },
+    {
+      "epoch": 0.12938689217758986,
+      "grad_norm": 3.6124682426452637,
+      "learning_rate": 1.9948113183641875e-05,
+      "loss": 0.3701,
+      "step": 102
+    },
+    {
+      "epoch": 0.1306553911205074,
+      "grad_norm": 4.924780368804932,
+      "learning_rate": 1.9943507547046063e-05,
+      "loss": 0.3369,
+      "step": 103
+    },
+    {
+      "epoch": 0.13192389006342495,
+      "grad_norm": 2.997985363006592,
+      "learning_rate": 1.9938706680622035e-05,
+      "loss": 0.3062,
+      "step": 104
+    },
+    {
+      "epoch": 0.1331923890063425,
+      "grad_norm": 3.6868772506713867,
+      "learning_rate": 1.9933710678629512e-05,
+      "loss": 0.332,
+      "step": 105
+    },
+    {
+      "epoch": 0.13446088794926003,
+      "grad_norm": 3.6897387504577637,
+      "learning_rate": 1.9928519639159508e-05,
+      "loss": 0.3083,
+      "step": 106
+    },
+    {
+      "epoch": 0.1357293868921776,
+      "grad_norm": 3.886378049850464,
+      "learning_rate": 1.992313366413236e-05,
+      "loss": 0.3331,
+      "step": 107
+    },
+    {
+      "epoch": 0.13699788583509515,
+      "grad_norm": 4.580509662628174,
+      "learning_rate": 1.9917552859295764e-05,
+      "loss": 0.2677,
+      "step": 108
+    },
+    {
+      "epoch": 0.13826638477801267,
+      "grad_norm": 3.936422348022461,
+      "learning_rate": 1.9911777334222682e-05,
+      "loss": 0.3305,
+      "step": 109
+    },
+    {
+      "epoch": 0.13953488372093023,
+      "grad_norm": 3.6973016262054443,
+      "learning_rate": 1.9905807202309196e-05,
+      "loss": 0.319,
+      "step": 110
+    },
+    {
+      "epoch": 0.1408033826638478,
+      "grad_norm": 3.311084270477295,
+      "learning_rate": 1.9899642580772274e-05,
+      "loss": 0.3307,
+      "step": 111
+    },
+    {
+      "epoch": 0.14207188160676532,
+      "grad_norm": 6.772805690765381,
+      "learning_rate": 1.9893283590647473e-05,
+      "loss": 0.2844,
+      "step": 112
+    },
+    {
+      "epoch": 0.14334038054968287,
+      "grad_norm": 3.7769129276275635,
+      "learning_rate": 1.9886730356786567e-05,
+      "loss": 0.3657,
+      "step": 113
+    },
+    {
+      "epoch": 0.14460887949260043,
+      "grad_norm": 3.9703691005706787,
+      "learning_rate": 1.9879983007855095e-05,
+      "loss": 0.3125,
+      "step": 114
+    },
+    {
+      "epoch": 0.14587737843551796,
+      "grad_norm": 4.326747417449951,
+      "learning_rate": 1.9873041676329825e-05,
+      "loss": 0.3402,
+      "step": 115
+    },
+    {
+      "epoch": 0.1471458773784355,
+      "grad_norm": 3.051367998123169,
+      "learning_rate": 1.9865906498496164e-05,
+      "loss": 0.3178,
+      "step": 116
+    },
+    {
+      "epoch": 0.14841437632135307,
+      "grad_norm": 3.126115083694458,
+      "learning_rate": 1.9858577614445476e-05,
+      "loss": 0.2708,
+      "step": 117
+    },
+    {
+      "epoch": 0.14968287526427063,
+      "grad_norm": 3.873002052307129,
+      "learning_rate": 1.9851055168072334e-05,
+      "loss": 0.3772,
+      "step": 118
+    },
+    {
+      "epoch": 0.15095137420718815,
+      "grad_norm": 4.09826135635376,
+      "learning_rate": 1.9843339307071697e-05,
+      "loss": 0.3596,
+      "step": 119
+    },
+    {
+      "epoch": 0.1522198731501057,
+      "grad_norm": 3.2259912490844727,
+      "learning_rate": 1.9835430182935998e-05,
+      "loss": 0.3205,
+      "step": 120
+    },
+    {
+      "epoch": 0.1522198731501057,
+      "eval_accuracy": 0.8200883002207505,
+      "eval_f1": 0.5914786967418546,
+      "eval_loss": 0.3738110363483429,
+      "eval_precision": 0.8137931034482758,
+      "eval_recall": 0.4645669291338583,
+      "eval_runtime": 52.2662,
+      "eval_samples_per_second": 5.3,
+      "eval_steps_per_second": 0.172,
+      "step": 120
+    },
+    {
+      "epoch": 0.15348837209302327,
+      "grad_norm": 3.804227590560913,
+      "learning_rate": 1.9827327950952197e-05,
+      "loss": 0.339,
+      "step": 121
+    },
+    {
+      "epoch": 0.1547568710359408,
+      "grad_norm": 4.679238796234131,
+      "learning_rate": 1.9819032770198692e-05,
+      "loss": 0.3301,
+      "step": 122
+    },
+    {
+      "epoch": 0.15602536997885835,
+      "grad_norm": 3.627534866333008,
+      "learning_rate": 1.981054480354223e-05,
+      "loss": 0.3087,
+      "step": 123
+    },
+    {
+      "epoch": 0.1572938689217759,
+      "grad_norm": 3.4424080848693848,
+      "learning_rate": 1.9801864217634703e-05,
+      "loss": 0.3192,
+      "step": 124
+    },
+    {
+      "epoch": 0.15856236786469344,
+      "grad_norm": 4.500999450683594,
+      "learning_rate": 1.9792991182909857e-05,
+      "loss": 0.3274,
+      "step": 125
+    },
+    {
+      "epoch": 0.159830866807611,
+      "grad_norm": 3.3262839317321777,
+      "learning_rate": 1.9783925873579967e-05,
+      "loss": 0.306,
+      "step": 126
+    },
+    {
+      "epoch": 0.16109936575052855,
+      "grad_norm": 4.585197448730469,
+      "learning_rate": 1.9774668467632397e-05,
+      "loss": 0.3007,
+      "step": 127
+    },
+    {
+      "epoch": 0.16236786469344608,
+      "grad_norm": 4.023516654968262,
+      "learning_rate": 1.976521914682614e-05,
+      "loss": 0.2867,
+      "step": 128
+    },
+    {
+      "epoch": 0.16363636363636364,
+      "grad_norm": 2.462892770767212,
+      "learning_rate": 1.97555780966882e-05,
+      "loss": 0.2486,
+      "step": 129
+    },
+    {
+      "epoch": 0.1649048625792812,
+      "grad_norm": 3.1057772636413574,
+      "learning_rate": 1.974574550650999e-05,
+      "loss": 0.2761,
+      "step": 130
+    },
+    {
+      "epoch": 0.16617336152219872,
+      "grad_norm": 4.618252754211426,
+      "learning_rate": 1.97357215693436e-05,
+      "loss": 0.2376,
+      "step": 131
+    },
+    {
+      "epoch": 0.16744186046511628,
+      "grad_norm": 3.48673152923584,
+      "learning_rate": 1.9725506481998003e-05,
+      "loss": 0.2945,
+      "step": 132
+    },
+    {
+      "epoch": 0.16871035940803383,
+      "grad_norm": 2.9228079319000244,
+      "learning_rate": 1.9715100445035197e-05,
+      "loss": 0.307,
+      "step": 133
+    },
+    {
+      "epoch": 0.16997885835095136,
+      "grad_norm": 6.0457611083984375,
+      "learning_rate": 1.970450366276627e-05,
+      "loss": 0.2083,
+      "step": 134
+    },
+    {
+      "epoch": 0.17124735729386892,
+      "grad_norm": 2.6761562824249268,
+      "learning_rate": 1.9693716343247374e-05,
+      "loss": 0.2134,
+      "step": 135
+    },
+    {
+      "epoch": 0.17251585623678647,
+      "grad_norm": 4.628048419952393,
+      "learning_rate": 1.9682738698275663e-05,
+      "loss": 0.2501,
+      "step": 136
+    },
+    {
+      "epoch": 0.173784355179704,
+      "grad_norm": 4.1137590408325195,
+      "learning_rate": 1.9671570943385118e-05,
+      "loss": 0.2737,
+      "step": 137
+    },
+    {
+      "epoch": 0.17505285412262156,
+      "grad_norm": 4.245756149291992,
+      "learning_rate": 1.9660213297842304e-05,
+      "loss": 0.3059,
+      "step": 138
+    },
+    {
+      "epoch": 0.17632135306553912,
+      "grad_norm": 4.939450263977051,
+      "learning_rate": 1.9648665984642104e-05,
+      "loss": 0.2831,
+      "step": 139
+    },
+    {
+      "epoch": 0.17758985200845667,
+      "grad_norm": 4.256004333496094,
+      "learning_rate": 1.96369292305033e-05,
+      "loss": 0.3026,
+      "step": 140
+    },
+    {
+      "epoch": 0.17758985200845667,
+      "eval_accuracy": 0.8289183222958058,
+      "eval_f1": 0.6191646191646192,
+      "eval_loss": 0.36796802282333374,
+      "eval_precision": 0.8235294117647058,
+      "eval_recall": 0.49606299212598426,
+      "eval_runtime": 52.4513,
+      "eval_samples_per_second": 5.281,
+      "eval_steps_per_second": 0.172,
+      "step": 140
+    },
+    {
+      "epoch": 0.1788583509513742,
+      "grad_norm": 3.7234766483306885,
+      "learning_rate": 1.9625003265864148e-05,
+      "loss": 0.2006,
+      "step": 141
+    },
+    {
+      "epoch": 0.18012684989429176,
+      "grad_norm": 4.156630039215088,
+      "learning_rate": 1.9612888324877834e-05,
+      "loss": 0.2776,
+      "step": 142
+    },
+    {
+      "epoch": 0.1813953488372093,
+      "grad_norm": 3.76343035697937,
+      "learning_rate": 1.9600584645407897e-05,
+      "loss": 0.3226,
+      "step": 143
+    },
+    {
+      "epoch": 0.18266384778012684,
+      "grad_norm": 4.899490833282471,
+      "learning_rate": 1.958809246902354e-05,
+      "loss": 0.2916,
+      "step": 144
+    },
+    {
+      "epoch": 0.1839323467230444,
+      "grad_norm": 3.958932638168335,
+      "learning_rate": 1.9575412040994902e-05,
+      "loss": 0.2549,
+      "step": 145
+    },
+    {
+      "epoch": 0.18520084566596196,
+      "grad_norm": 2.987992525100708,
+      "learning_rate": 1.956254361028823e-05,
+      "loss": 0.2549,
+      "step": 146
+    },
+    {
+      "epoch": 0.18646934460887948,
+      "grad_norm": 4.582171440124512,
+      "learning_rate": 1.9549487429561003e-05,
+      "loss": 0.3249,
+      "step": 147
+    },
+    {
+      "epoch": 0.18773784355179704,
+      "grad_norm": 3.215688705444336,
+      "learning_rate": 1.953624375515696e-05,
+      "loss": 0.2555,
+      "step": 148
+    },
+    {
+      "epoch": 0.1890063424947146,
+      "grad_norm": 4.125192642211914,
+      "learning_rate": 1.9522812847101077e-05,
+      "loss": 0.3004,
+      "step": 149
+    },
+    {
+      "epoch": 0.19027484143763213,
+      "grad_norm": 4.248358726501465,
+      "learning_rate": 1.9509194969094447e-05,
+      "loss": 0.2206,
+      "step": 150
+    },
+    {
+      "epoch": 0.19154334038054968,
+      "grad_norm": 5.806552886962891,
+      "learning_rate": 1.9495390388509123e-05,
+      "loss": 0.2926,
+      "step": 151
+    },
+    {
+      "epoch": 0.19281183932346724,
+      "grad_norm": 3.595050096511841,
+      "learning_rate": 1.948139937638285e-05,
+      "loss": 0.2746,
+      "step": 152
+    },
+    {
+      "epoch": 0.19408033826638477,
+      "grad_norm": 4.2712836265563965,
+      "learning_rate": 1.9467222207413763e-05,
+      "loss": 0.275,
+      "step": 153
+    },
+    {
+      "epoch": 0.19534883720930232,
+      "grad_norm": 3.809781074523926,
+      "learning_rate": 1.945285915995496e-05,
+      "loss": 0.2291,
+      "step": 154
+    },
+    {
+      "epoch": 0.19661733615221988,
+      "grad_norm": 3.5505263805389404,
+      "learning_rate": 1.9438310516009084e-05,
+      "loss": 0.2637,
+      "step": 155
+    },
+    {
+      "epoch": 0.1978858350951374,
+      "grad_norm": 2.8975532054901123,
+      "learning_rate": 1.9423576561222742e-05,
+      "loss": 0.2066,
+      "step": 156
+    },
+    {
+      "epoch": 0.19915433403805496,
+      "grad_norm": 4.945555686950684,
+      "learning_rate": 1.9408657584880924e-05,
+      "loss": 0.2785,
+      "step": 157
+    },
+    {
+      "epoch": 0.20042283298097252,
+      "grad_norm": 6.6838274002075195,
+      "learning_rate": 1.9393553879901314e-05,
+      "loss": 0.3395,
+      "step": 158
+    },
+    {
+      "epoch": 0.20169133192389008,
+      "grad_norm": 4.039582252502441,
+      "learning_rate": 1.9378265742828533e-05,
+      "loss": 0.3043,
+      "step": 159
+    },
+    {
+      "epoch": 0.2029598308668076,
+      "grad_norm": 3.475989580154419,
+      "learning_rate": 1.9362793473828338e-05,
+      "loss": 0.2886,
+      "step": 160
+    },
+    {
+      "epoch": 0.2029598308668076,
+      "eval_accuracy": 0.8432671081677704,
+      "eval_f1": 0.6536585365853659,
+      "eval_loss": 0.34669193625450134,
+      "eval_precision": 0.8589743589743589,
+      "eval_recall": 0.5275590551181102,
+      "eval_runtime": 52.3104,
+      "eval_samples_per_second": 5.295,
+      "eval_steps_per_second": 0.172,
+      "step": 160
+    },
+    {
+      "epoch": 0.20422832980972516,
+      "grad_norm": 5.376806735992432,
+      "learning_rate": 1.9347137376681692e-05,
+      "loss": 0.2629,
+      "step": 161
+    },
+    {
+      "epoch": 0.20549682875264272,
+      "grad_norm": 5.16301155090332,
+      "learning_rate": 1.933129775877884e-05,
+      "loss": 0.3097,
+      "step": 162
+    },
+    {
+      "epoch": 0.20676532769556025,
+      "grad_norm": 6.509466648101807,
+      "learning_rate": 1.9315274931113253e-05,
+      "loss": 0.2566,
+      "step": 163
+    },
+    {
+      "epoch": 0.2080338266384778,
+      "grad_norm": 4.453103065490723,
+      "learning_rate": 1.929906920827551e-05,
+      "loss": 0.2863,
+      "step": 164
+    },
+    {
+      "epoch": 0.20930232558139536,
+      "grad_norm": 3.955231189727783,
+      "learning_rate": 1.9282680908447152e-05,
+      "loss": 0.3102,
+      "step": 165
+    },
+    {
+      "epoch": 0.2105708245243129,
+      "grad_norm": 5.5910258293151855,
+      "learning_rate": 1.926611035339441e-05,
+      "loss": 0.2271,
+      "step": 166
+    },
+    {
+      "epoch": 0.21183932346723044,
+      "grad_norm": 4.199539661407471,
+      "learning_rate": 1.9249357868461896e-05,
+      "loss": 0.27,
+      "step": 167
+    },
+    {
+      "epoch": 0.213107822410148,
+      "grad_norm": 3.3116366863250732,
+      "learning_rate": 1.9232423782566217e-05,
+      "loss": 0.3221,
+      "step": 168
+    },
+    {
+      "epoch": 0.21437632135306553,
+      "grad_norm": 4.038525104522705,
+      "learning_rate": 1.9215308428189514e-05,
+      "loss": 0.2448,
+      "step": 169
+    },
+    {
+      "epoch": 0.2156448202959831,
+      "grad_norm": 4.98528528213501,
+      "learning_rate": 1.9198012141372934e-05,
+      "loss": 0.317,
+      "step": 170
+    },
+    {
+      "epoch": 0.21691331923890064,
+      "grad_norm": 3.0479323863983154,
+      "learning_rate": 1.918053526171004e-05,
+      "loss": 0.2867,
+      "step": 171
+    },
+    {
+      "epoch": 0.21818181818181817,
+      "grad_norm": 3.0250513553619385,
+      "learning_rate": 1.916287813234012e-05,
+      "loss": 0.1996,
+      "step": 172
+    },
+    {
+      "epoch": 0.21945031712473573,
+      "grad_norm": 2.8498682975769043,
+      "learning_rate": 1.914504109994149e-05,
+      "loss": 0.2383,
+      "step": 173
+    },
+    {
+      "epoch": 0.22071881606765328,
+      "grad_norm": 3.0769264698028564,
+      "learning_rate": 1.912702451472465e-05,
+      "loss": 0.2481,
+      "step": 174
+    },
+    {
+      "epoch": 0.2219873150105708,
+      "grad_norm": 3.5798423290252686,
+      "learning_rate": 1.910882873042542e-05,
+      "loss": 0.2883,
+      "step": 175
+    },
+    {
+      "epoch": 0.22325581395348837,
+      "grad_norm": 4.30120849609375,
+      "learning_rate": 1.9090454104298003e-05,
+      "loss": 0.2893,
+      "step": 176
+    },
+    {
+      "epoch": 0.22452431289640593,
+      "grad_norm": 4.033363342285156,
+      "learning_rate": 1.907190099710796e-05,
+      "loss": 0.2408,
+      "step": 177
+    },
+    {
+      "epoch": 0.22579281183932345,
+      "grad_norm": 6.6475348472595215,
+      "learning_rate": 1.9053169773125143e-05,
+      "loss": 0.2549,
+      "step": 178
+    },
+    {
+      "epoch": 0.227061310782241,
+      "grad_norm": 5.1474199295043945,
+      "learning_rate": 1.903426080011651e-05,
+      "loss": 0.2493,
+      "step": 179
+    },
+    {
+      "epoch": 0.22832980972515857,
+      "grad_norm": 4.56413459777832,
+      "learning_rate": 1.901517444933894e-05,
+      "loss": 0.2345,
+      "step": 180
+    },
+    {
+      "epoch": 0.22832980972515857,
+      "eval_accuracy": 0.8587196467991169,
+      "eval_f1": 0.7253218884120172,
+      "eval_loss": 0.3289457857608795,
+      "eval_precision": 0.7971698113207547,
+      "eval_recall": 0.6653543307086615,
+      "eval_runtime": 51.489,
+      "eval_samples_per_second": 5.38,
+      "eval_steps_per_second": 0.175,
+      "step": 180
+    },
+    {
+      "epoch": 0.22959830866807612,
+      "grad_norm": 6.834506988525391,
+      "learning_rate": 1.8995911095531932e-05,
+      "loss": 0.2355,
+      "step": 181
+    },
+    {
+      "epoch": 0.23086680761099365,
+      "grad_norm": 6.830471038818359,
+      "learning_rate": 1.8976471116910236e-05,
+      "loss": 0.2727,
+      "step": 182
+    },
+    {
+      "epoch": 0.2321353065539112,
+      "grad_norm": 4.903042316436768,
+      "learning_rate": 1.895685489515644e-05,
+      "loss": 0.25,
+      "step": 183
+    },
+    {
+      "epoch": 0.23340380549682876,
+      "grad_norm": 3.5585076808929443,
+      "learning_rate": 1.8937062815413466e-05,
+      "loss": 0.2276,
+      "step": 184
+    },
+    {
+      "epoch": 0.2346723044397463,
+      "grad_norm": 3.2517757415771484,
+      "learning_rate": 1.891709526627703e-05,
+      "loss": 0.2492,
+      "step": 185
+    },
+    {
+      "epoch": 0.23594080338266385,
+      "grad_norm": 5.2381181716918945,
+      "learning_rate": 1.8896952639787977e-05,
+      "loss": 0.1813,
+      "step": 186
+    },
+    {
+      "epoch": 0.2372093023255814,
+      "grad_norm": 6.287901401519775,
+      "learning_rate": 1.8876635331424616e-05,
+      "loss": 0.2281,
+      "step": 187
+    },
+    {
+      "epoch": 0.23847780126849893,
+      "grad_norm": 4.748634338378906,
+      "learning_rate": 1.8856143740094938e-05,
+      "loss": 0.2654,
+      "step": 188
+    },
+    {
+      "epoch": 0.2397463002114165,
+      "grad_norm": 7.864667892456055,
+      "learning_rate": 1.8835478268128784e-05,
+      "loss": 0.2791,
+      "step": 189
+    },
+    {
+      "epoch": 0.24101479915433405,
+      "grad_norm": 5.30699348449707,
+      "learning_rate": 1.881463932126996e-05,
+      "loss": 0.2772,
+      "step": 190
+    },
+    {
+      "epoch": 0.24228329809725158,
+      "grad_norm": 4.305821895599365,
+      "learning_rate": 1.879362730866825e-05,
+      "loss": 0.2552,
+      "step": 191
+    },
+    {
+      "epoch": 0.24355179704016913,
+      "grad_norm": 9.22236442565918,
+      "learning_rate": 1.8772442642871405e-05,
+      "loss": 0.3735,
+      "step": 192
+    },
+    {
+      "epoch": 0.2448202959830867,
+      "grad_norm": 3.3223769664764404,
+      "learning_rate": 1.8751085739817017e-05,
+      "loss": 0.2447,
+      "step": 193
+    },
+    {
+      "epoch": 0.24608879492600422,
+      "grad_norm": 4.844041347503662,
+      "learning_rate": 1.8729557018824382e-05,
+      "loss": 0.3255,
+      "step": 194
+    },
+    {
+      "epoch": 0.24735729386892177,
+      "grad_norm": 3.9604134559631348,
+      "learning_rate": 1.8707856902586244e-05,
+      "loss": 0.2706,
+      "step": 195
+    },
+    {
+      "epoch": 0.24862579281183933,
+      "grad_norm": 3.333854913711548,
+      "learning_rate": 1.8685985817160503e-05,
+      "loss": 0.2665,
+      "step": 196
+    },
+    {
+      "epoch": 0.24989429175475686,
+      "grad_norm": 2.807447910308838,
+      "learning_rate": 1.8663944191961853e-05,
+      "loss": 0.2699,
+      "step": 197
+    },
+    {
+      "epoch": 0.25116279069767444,
+      "grad_norm": 4.559386730194092,
+      "learning_rate": 1.864173245975335e-05,
+      "loss": 0.2557,
+      "step": 198
+    },
+    {
+      "epoch": 0.25243128964059197,
+      "grad_norm": 3.0040392875671387,
+      "learning_rate": 1.861935105663791e-05,
+      "loss": 0.2343,
+      "step": 199
+    },
+    {
+      "epoch": 0.2536997885835095,
+      "grad_norm": 3.234032154083252,
+      "learning_rate": 1.8596800422049756e-05,
+      "loss": 0.2964,
+      "step": 200
+    },
+    {
+      "epoch": 0.2536997885835095,
+      "eval_accuracy": 0.8377483443708609,
+      "eval_f1": 0.6388206388206388,
+      "eval_loss": 0.33216196298599243,
+      "eval_precision": 0.8496732026143791,
+      "eval_recall": 0.5118110236220472,
+      "eval_runtime": 51.0562,
+      "eval_samples_per_second": 5.425,
+      "eval_steps_per_second": 0.176,
+      "step": 200
+    },
+    {
+      "epoch": 0.2549682875264271,
+      "grad_norm": 2.9896929264068604,
+      "learning_rate": 1.8574080998745785e-05,
+      "loss": 0.2425,
+      "step": 201
+    },
+    {
+      "epoch": 0.2562367864693446,
+      "grad_norm": 4.189578056335449,
+      "learning_rate": 1.8551193232796862e-05,
+      "loss": 0.274,
+      "step": 202
+    },
+    {
+      "epoch": 0.25750528541226214,
+      "grad_norm": 3.2401742935180664,
+      "learning_rate": 1.852813757357909e-05,
+      "loss": 0.2548,
+      "step": 203
+    },
+    {
+      "epoch": 0.2587737843551797,
+      "grad_norm": 3.4946937561035156,
+      "learning_rate": 1.8504914473764973e-05,
+      "loss": 0.2924,
+      "step": 204
+    },
+    {
+      "epoch": 0.26004228329809725,
+      "grad_norm": 2.9049434661865234,
+      "learning_rate": 1.8481524389314508e-05,
+      "loss": 0.2132,
+      "step": 205
+    },
+    {
+      "epoch": 0.2613107822410148,
+      "grad_norm": 3.942204475402832,
+      "learning_rate": 1.845796777946627e-05,
+      "loss": 0.251,
+      "step": 206
+    },
+    {
+      "epoch": 0.26257928118393237,
+      "grad_norm": 3.098745822906494,
+      "learning_rate": 1.8434245106728367e-05,
+      "loss": 0.2698,
+      "step": 207
+    },
+    {
+      "epoch": 0.2638477801268499,
+      "grad_norm": 3.4184677600860596,
+      "learning_rate": 1.8410356836869374e-05,
+      "loss": 0.2239,
+      "step": 208
+    },
+    {
+      "epoch": 0.2651162790697674,
+      "grad_norm": 2.8000166416168213,
+      "learning_rate": 1.8386303438909182e-05,
+      "loss": 0.2144,
+      "step": 209
+    },
+    {
+      "epoch": 0.266384778012685,
+      "grad_norm": 4.3039937019348145,
+      "learning_rate": 1.8362085385109788e-05,
+      "loss": 0.2572,
+      "step": 210
+    },
+    {
+      "epoch": 0.26765327695560254,
+      "grad_norm": 3.855637311935425,
+      "learning_rate": 1.8337703150966025e-05,
+      "loss": 0.2392,
+      "step": 211
+    },
+    {
+      "epoch": 0.26892177589852007,
+      "grad_norm": 3.2118887901306152,
+      "learning_rate": 1.8313157215196237e-05,
+      "loss": 0.21,
+      "step": 212
+    },
+    {
+      "epoch": 0.27019027484143765,
+      "grad_norm": 4.038395881652832,
+      "learning_rate": 1.8288448059732852e-05,
+      "loss": 0.1917,
+      "step": 213
+    },
+    {
+      "epoch": 0.2714587737843552,
+      "grad_norm": 6.29147481918335,
+      "learning_rate": 1.8263576169712955e-05,
+      "loss": 0.2542,
+      "step": 214
+    },
+    {
+      "epoch": 0.2727272727272727,
+      "grad_norm": 6.392573356628418,
+      "learning_rate": 1.823854203346872e-05,
+      "loss": 0.2953,
+      "step": 215
+    },
+    {
+      "epoch": 0.2739957716701903,
+      "grad_norm": 5.477741718292236,
+      "learning_rate": 1.8213346142517884e-05,
+      "loss": 0.3066,
+      "step": 216
+    },
+    {
+      "epoch": 0.2752642706131078,
+      "grad_norm": 7.538582801818848,
+      "learning_rate": 1.8187988991554027e-05,
+      "loss": 0.303,
+      "step": 217
+    },
+    {
+      "epoch": 0.27653276955602535,
+      "grad_norm": 4.740074634552002,
+      "learning_rate": 1.8162471078436903e-05,
+      "loss": 0.2752,
+      "step": 218
+    },
+    {
+      "epoch": 0.27780126849894293,
+      "grad_norm": 3.831221580505371,
+      "learning_rate": 1.8136792904182657e-05,
+      "loss": 0.2749,
+      "step": 219
+    },
+    {
+      "epoch": 0.27906976744186046,
+      "grad_norm": 3.5587270259857178,
+      "learning_rate": 1.8110954972953984e-05,
+      "loss": 0.2655,
+      "step": 220
+    },
+    {
+      "epoch": 0.27906976744186046,
+      "eval_accuracy": 0.8278145695364238,
+      "eval_f1": 0.5979381443298969,
+      "eval_loss": 0.3495219349861145,
+      "eval_precision": 0.8656716417910447,
+      "eval_recall": 0.4566929133858268,
+      "eval_runtime": 51.5039,
+      "eval_samples_per_second": 5.378,
+      "eval_steps_per_second": 0.175,
+      "step": 220
+    },
+    {
+      "epoch": 0.280338266384778,
+      "grad_norm": 4.122521877288818,
+      "learning_rate": 1.8084957792050226e-05,
+      "loss": 0.2141,
+      "step": 221
+    },
+    {
+      "epoch": 0.2816067653276956,
+      "grad_norm": 3.4336111545562744,
+      "learning_rate": 1.8058801871897423e-05,
+      "loss": 0.1997,
+      "step": 222
+    },
+    {
+      "epoch": 0.2828752642706131,
+      "grad_norm": 4.062242031097412,
+      "learning_rate": 1.803248772603828e-05,
+      "loss": 0.2811,
+      "step": 223
+    },
+    {
+      "epoch": 0.28414376321353063,
+      "grad_norm": 4.814560413360596,
+      "learning_rate": 1.8006015871122105e-05,
+      "loss": 0.1923,
+      "step": 224
+    },
+    {
+      "epoch": 0.2854122621564482,
+      "grad_norm": 3.1431543827056885,
+      "learning_rate": 1.797938682689462e-05,
+      "loss": 0.223,
+      "step": 225
+    },
+    {
+      "epoch": 0.28668076109936574,
+      "grad_norm": 4.914591312408447,
+      "learning_rate": 1.7952601116187822e-05,
+      "loss": 0.2928,
+      "step": 226
+    },
+    {
+      "epoch": 0.2879492600422833,
+      "grad_norm": 2.977471113204956,
+      "learning_rate": 1.7925659264909657e-05,
+      "loss": 0.2562,
+      "step": 227
+    },
+    {
+      "epoch": 0.28921775898520086,
+      "grad_norm": 5.882057189941406,
+      "learning_rate": 1.7898561802033717e-05,
+      "loss": 0.2599,
+      "step": 228
+    },
+    {
+      "epoch": 0.2904862579281184,
+      "grad_norm": 7.490647792816162,
+      "learning_rate": 1.7871309259588884e-05,
+      "loss": 0.2616,
+      "step": 229
+    },
+    {
+      "epoch": 0.2917547568710359,
+      "grad_norm": 8.199599266052246,
+      "learning_rate": 1.784390217264882e-05,
+      "loss": 0.2665,
+      "step": 230
+    },
+    {
+      "epoch": 0.2930232558139535,
+      "grad_norm": 2.947110414505005,
+      "learning_rate": 1.781634107932153e-05,
+      "loss": 0.183,
+      "step": 231
+    },
+    {
+      "epoch": 0.294291754756871,
+      "grad_norm": 5.069705486297607,
+      "learning_rate": 1.7788626520738745e-05,
+      "loss": 0.2423,
+      "step": 232
+    },
+    {
+      "epoch": 0.2955602536997886,
+      "grad_norm": 3.6471478939056396,
+      "learning_rate": 1.776075904104533e-05,
+      "loss": 0.24,
+      "step": 233
+    },
+    {
+      "epoch": 0.29682875264270614,
+      "grad_norm": 3.3612494468688965,
+      "learning_rate": 1.773273918738857e-05,
+      "loss": 0.2007,
+      "step": 234
+    },
+    {
+      "epoch": 0.29809725158562367,
+      "grad_norm": 3.5518970489501953,
+      "learning_rate": 1.7704567509907468e-05,
+      "loss": 0.2526,
+      "step": 235
+    },
+    {
+      "epoch": 0.29936575052854125,
+      "grad_norm": 5.473038673400879,
+      "learning_rate": 1.7676244561721905e-05,
+      "loss": 0.2802,
+      "step": 236
+    },
+    {
+      "epoch": 0.3006342494714588,
+      "grad_norm": 4.366296768188477,
+      "learning_rate": 1.7647770898921798e-05,
+      "loss": 0.2698,
+      "step": 237
+    },
+    {
+      "epoch": 0.3019027484143763,
+      "grad_norm": 6.12213659286499,
+      "learning_rate": 1.7619147080556188e-05,
+      "loss": 0.2266,
+      "step": 238
+    },
+    {
+      "epoch": 0.3031712473572939,
+      "grad_norm": 4.8955864906311035,
+      "learning_rate": 1.759037366862224e-05,
+      "loss": 0.2574,
+      "step": 239
+    },
+    {
+      "epoch": 0.3044397463002114,
+      "grad_norm": 4.285499572753906,
+      "learning_rate": 1.7561451228054247e-05,
+      "loss": 0.3252,
+      "step": 240
+    },
+    {
+      "epoch": 0.3044397463002114,
+      "eval_accuracy": 0.8454746136865342,
+      "eval_f1": 0.6713615023474179,
+      "eval_loss": 0.31889277696609497,
+      "eval_precision": 0.8313953488372093,
+      "eval_recall": 0.562992125984252,
+      "eval_runtime": 51.2215,
+      "eval_samples_per_second": 5.408,
+      "eval_steps_per_second": 0.176,
+      "step": 240
+    },
+    {
+      "epoch": 0.30570824524312895,
+      "grad_norm": 3.903001308441162,
+      "learning_rate": 1.7532380326712487e-05,
+      "loss": 0.2511,
+      "step": 241
+    },
+    {
+      "epoch": 0.30697674418604654,
+      "grad_norm": 5.383487701416016,
+      "learning_rate": 1.7503161535372126e-05,
+      "loss": 0.2861,
+      "step": 242
+    },
+    {
+      "epoch": 0.30824524312896406,
+      "grad_norm": 5.267589569091797,
+      "learning_rate": 1.7473795427711975e-05,
+      "loss": 0.1947,
+      "step": 243
+    },
+    {
+      "epoch": 0.3095137420718816,
+      "grad_norm": 6.931105613708496,
+      "learning_rate": 1.7444282580303237e-05,
+      "loss": 0.2462,
+      "step": 244
+    },
+    {
+      "epoch": 0.3107822410147992,
+      "grad_norm": 7.340319633483887,
+      "learning_rate": 1.74146235725982e-05,
+      "loss": 0.2709,
+      "step": 245
+    },
+    {
+      "epoch": 0.3120507399577167,
+      "grad_norm": 4.371819972991943,
+      "learning_rate": 1.738481898691884e-05,
+      "loss": 0.2763,
+      "step": 246
+    },
+    {
+      "epoch": 0.31331923890063423,
+      "grad_norm": 4.855105400085449,
+      "learning_rate": 1.7354869408445393e-05,
+      "loss": 0.2306,
+      "step": 247
+    },
+    {
+      "epoch": 0.3145877378435518,
+      "grad_norm": 2.958298683166504,
+      "learning_rate": 1.7324775425204877e-05,
+      "loss": 0.2072,
+      "step": 248
+    },
+    {
+      "epoch": 0.31585623678646935,
+      "grad_norm": 3.649681806564331,
+      "learning_rate": 1.729453762805954e-05,
+      "loss": 0.2535,
+      "step": 249
+    },
+    {
+      "epoch": 0.3171247357293869,
+      "grad_norm": 3.00809383392334,
+      "learning_rate": 1.7264156610695247e-05,
+      "loss": 0.2002,
+      "step": 250
+    },
+    {
+      "epoch": 0.31839323467230446,
+      "grad_norm": 2.804321050643921,
+      "learning_rate": 1.7233632969609844e-05,
+      "loss": 0.2373,
+      "step": 251
+    },
+    {
+      "epoch": 0.319661733615222,
+      "grad_norm": 4.483391761779785,
+      "learning_rate": 1.7202967304101434e-05,
+      "loss": 0.226,
+      "step": 252
+    },
+    {
+      "epoch": 0.3209302325581395,
+      "grad_norm": 4.620050430297852,
+      "learning_rate": 1.7172160216256607e-05,
+      "loss": 0.2272,
+      "step": 253
+    },
+    {
+      "epoch": 0.3221987315010571,
+      "grad_norm": 4.082276821136475,
+      "learning_rate": 1.714121231093864e-05,
+      "loss": 0.2021,
+      "step": 254
+    },
+    {
+      "epoch": 0.32346723044397463,
+      "grad_norm": 4.720620632171631,
+      "learning_rate": 1.7110124195775592e-05,
+      "loss": 0.2471,
+      "step": 255
+    },
+    {
+      "epoch": 0.32473572938689216,
+      "grad_norm": 3.695242166519165,
+      "learning_rate": 1.707889648114839e-05,
+      "loss": 0.2629,
+      "step": 256
+    },
+    {
+      "epoch": 0.32600422832980974,
+      "grad_norm": 5.539944648742676,
+      "learning_rate": 1.704752978017885e-05,
+      "loss": 0.219,
+      "step": 257
+    },
+    {
+      "epoch": 0.32727272727272727,
+      "grad_norm": 4.941290378570557,
+      "learning_rate": 1.701602470871763e-05,
+      "loss": 0.2,
+      "step": 258
+    },
+    {
+      "epoch": 0.3285412262156448,
+      "grad_norm": 5.238450527191162,
+      "learning_rate": 1.698438188533213e-05,
+      "loss": 0.1941,
+      "step": 259
+    },
+    {
+      "epoch": 0.3298097251585624,
+      "grad_norm": 3.7015936374664307,
+      "learning_rate": 1.6952601931294372e-05,
+      "loss": 0.2561,
+      "step": 260
+    },
+    {
+      "epoch": 0.3298097251585624,
+      "eval_accuracy": 0.8532008830022075,
+      "eval_f1": 0.6997742663656885,
+      "eval_loss": 0.3227859139442444,
+      "eval_precision": 0.8201058201058201,
+      "eval_recall": 0.610236220472441,
+      "eval_runtime": 51.1034,
+      "eval_samples_per_second": 5.42,
+      "eval_steps_per_second": 0.176,
+      "step": 260
+    },
+    {
+      "epoch": 0.3310782241014799,
+      "grad_norm": 3.9863884449005127,
+      "learning_rate": 1.6920685470568775e-05,
+      "loss": 0.2591,
+      "step": 261
+    },
+    {
+      "epoch": 0.33234672304439744,
+      "grad_norm": 3.281353712081909,
+      "learning_rate": 1.6888633129799932e-05,
+      "loss": 0.1717,
+      "step": 262
+    },
+    {
+      "epoch": 0.333615221987315,
+      "grad_norm": 3.945847749710083,
+      "learning_rate": 1.6856445538300273e-05,
+      "loss": 0.2916,
+      "step": 263
+    },
+    {
+      "epoch": 0.33488372093023255,
+      "grad_norm": 4.858996391296387,
+      "learning_rate": 1.6824123328037737e-05,
+      "loss": 0.2521,
+      "step": 264
+    },
+    {
+      "epoch": 0.3361522198731501,
+      "grad_norm": 7.984918594360352,
+      "learning_rate": 1.6791667133623354e-05,
+      "loss": 0.2878,
+      "step": 265
+    },
+    {
+      "epoch": 0.33742071881606767,
+      "grad_norm": 2.9552502632141113,
+      "learning_rate": 1.675907759229879e-05,
+      "loss": 0.2246,
+      "step": 266
+    },
+    {
+      "epoch": 0.3386892177589852,
+      "grad_norm": 5.446991443634033,
+      "learning_rate": 1.6726355343923816e-05,
+      "loss": 0.2362,
+      "step": 267
+    },
+    {
+      "epoch": 0.3399577167019027,
+      "grad_norm": 4.872383117675781,
+      "learning_rate": 1.6693501030963775e-05,
+      "loss": 0.2282,
+      "step": 268
+    },
+    {
+      "epoch": 0.3412262156448203,
+      "grad_norm": 4.628652095794678,
+      "learning_rate": 1.6660515298476943e-05,
+      "loss": 0.2489,
+      "step": 269
+    },
+    {
+      "epoch": 0.34249471458773784,
+      "grad_norm": 3.3619189262390137,
+      "learning_rate": 1.6627398794101883e-05,
+      "loss": 0.2158,
+      "step": 270
+    },
+    {
+      "epoch": 0.34376321353065536,
+      "grad_norm": 4.453815937042236,
+      "learning_rate": 1.659415216804471e-05,
+      "loss": 0.267,
+      "step": 271
+    },
+    {
+      "epoch": 0.34503171247357295,
+      "grad_norm": 4.466940879821777,
+      "learning_rate": 1.6560776073066343e-05,
+      "loss": 0.2187,
+      "step": 272
+    },
+    {
+      "epoch": 0.3463002114164905,
+      "grad_norm": 4.6378397941589355,
+      "learning_rate": 1.652727116446968e-05,
+      "loss": 0.2104,
+      "step": 273
+    },
+    {
+      "epoch": 0.347568710359408,
+      "grad_norm": 3.28175950050354,
+      "learning_rate": 1.6493638100086726e-05,
+      "loss": 0.2269,
+      "step": 274
+    },
+    {
+      "epoch": 0.3488372093023256,
+      "grad_norm": 5.272284507751465,
+      "learning_rate": 1.6459877540265686e-05,
+      "loss": 0.257,
+      "step": 275
+    },
+    {
+      "epoch": 0.3501057082452431,
+      "grad_norm": 2.739626407623291,
+      "learning_rate": 1.6425990147858004e-05,
+      "loss": 0.1847,
+      "step": 276
+    },
+    {
+      "epoch": 0.3513742071881607,
+      "grad_norm": 4.007655620574951,
+      "learning_rate": 1.639197658820534e-05,
+      "loss": 0.2498,
+      "step": 277
+    },
+    {
+      "epoch": 0.35264270613107823,
+      "grad_norm": 2.493644952774048,
+      "learning_rate": 1.6357837529126502e-05,
+      "loss": 0.1925,
+      "step": 278
+    },
+    {
+      "epoch": 0.35391120507399576,
+      "grad_norm": 5.886753559112549,
+      "learning_rate": 1.6323573640904352e-05,
+      "loss": 0.2331,
+      "step": 279
+    },
+    {
+      "epoch": 0.35517970401691334,
+      "grad_norm": 3.9009897708892822,
+      "learning_rate": 1.628918559627262e-05,
+      "loss": 0.1661,
+      "step": 280
+    },
+    {
+      "epoch": 0.35517970401691334,
+      "eval_accuracy": 0.8498896247240618,
+      "eval_f1": 0.6746411483253588,
+      "eval_loss": 0.31411004066467285,
+      "eval_precision": 0.8597560975609756,
+      "eval_recall": 0.5551181102362205,
+      "eval_runtime": 51.0464,
+      "eval_samples_per_second": 5.426,
+      "eval_steps_per_second": 0.176,
+      "step": 280
+    },
+    {
+      "epoch": 0.3564482029598309,
+      "grad_norm": 5.003305912017822,
+      "learning_rate": 1.625467407040273e-05,
+      "loss": 0.2532,
+      "step": 281
+    },
+    {
+      "epoch": 0.3577167019027484,
+      "grad_norm": 3.5860018730163574,
+      "learning_rate": 1.6220039740890507e-05,
+      "loss": 0.2194,
+      "step": 282
+    },
+    {
+      "epoch": 0.358985200845666,
+      "grad_norm": 3.8231117725372314,
+      "learning_rate": 1.6185283287742902e-05,
+      "loss": 0.2183,
+      "step": 283
+    },
+    {
+      "epoch": 0.3602536997885835,
+      "grad_norm": 5.111902236938477,
+      "learning_rate": 1.615040539336462e-05,
+      "loss": 0.2831,
+      "step": 284
+    },
+    {
+      "epoch": 0.36152219873150104,
+      "grad_norm": 7.390100002288818,
+      "learning_rate": 1.6115406742544736e-05,
+      "loss": 0.2675,
+      "step": 285
+    },
+    {
+      "epoch": 0.3627906976744186,
+      "grad_norm": 4.8447489738464355,
+      "learning_rate": 1.6080288022443242e-05,
+      "loss": 0.184,
+      "step": 286
+    },
+    {
+      "epoch": 0.36405919661733616,
+      "grad_norm": 4.323414325714111,
+      "learning_rate": 1.6045049922577567e-05,
+      "loss": 0.2267,
+      "step": 287
+    },
+    {
+      "epoch": 0.3653276955602537,
+      "grad_norm": 4.628475189208984,
+      "learning_rate": 1.6009693134809018e-05,
+      "loss": 0.2527,
+      "step": 288
+    },
+    {
+      "epoch": 0.36659619450317127,
+      "grad_norm": 5.454392910003662,
+      "learning_rate": 1.597421835332922e-05,
+      "loss": 0.2165,
+      "step": 289
+    },
+    {
+      "epoch": 0.3678646934460888,
+      "grad_norm": 3.9720702171325684,
+      "learning_rate": 1.5938626274646457e-05,
+      "loss": 0.2671,
+      "step": 290
+    },
+    {
+      "epoch": 0.3691331923890063,
+      "grad_norm": 4.625940322875977,
+      "learning_rate": 1.590291759757204e-05,
+      "loss": 0.242,
+      "step": 291
+    },
+    {
+      "epoch": 0.3704016913319239,
+      "grad_norm": 3.9591894149780273,
+      "learning_rate": 1.5867093023206538e-05,
+      "loss": 0.185,
+      "step": 292
+    },
+    {
+      "epoch": 0.37167019027484144,
+      "grad_norm": 9.09524154663086,
+      "learning_rate": 1.5831153254926048e-05,
+      "loss": 0.2799,
+      "step": 293
+    },
+    {
+      "epoch": 0.37293868921775897,
+      "grad_norm": 3.936962127685547,
+      "learning_rate": 1.579509899836837e-05,
+      "loss": 0.2206,
+      "step": 294
+    },
+    {
+      "epoch": 0.37420718816067655,
+      "grad_norm": 3.293510913848877,
+      "learning_rate": 1.5758930961419154e-05,
+      "loss": 0.202,
+      "step": 295
+    },
+    {
+      "epoch": 0.3754756871035941,
+      "grad_norm": 4.638730525970459,
+      "learning_rate": 1.5722649854198007e-05,
+      "loss": 0.2745,
+      "step": 296
+    },
+    {
+      "epoch": 0.3767441860465116,
+      "grad_norm": 11.325010299682617,
+      "learning_rate": 1.568625638904454e-05,
+      "loss": 0.258,
+      "step": 297
+    },
+    {
+      "epoch": 0.3780126849894292,
+      "grad_norm": 3.792722702026367,
+      "learning_rate": 1.564975128050439e-05,
+      "loss": 0.2376,
+      "step": 298
+    },
+    {
+      "epoch": 0.3792811839323467,
+      "grad_norm": 4.620094299316406,
+      "learning_rate": 1.5613135245315196e-05,
+      "loss": 0.2361,
+      "step": 299
+    },
+    {
+      "epoch": 0.38054968287526425,
+      "grad_norm": 4.535460948944092,
+      "learning_rate": 1.557640900239251e-05,
+      "loss": 0.1812,
+      "step": 300
+    },
+    {
+      "epoch": 0.38054968287526425,
+      "eval_accuracy": 0.8300220750551877,
+      "eval_f1": 0.6169154228855721,
+      "eval_loss": 0.33304014801979065,
+      "eval_precision": 0.8378378378378378,
+      "eval_recall": 0.4881889763779528,
+      "eval_runtime": 50.8332,
+      "eval_samples_per_second": 5.449,
+      "eval_steps_per_second": 0.177,
+      "step": 300
+    },
+    {
+      "epoch": 0.38181818181818183,
+      "grad_norm": 3.9105491638183594,
+      "learning_rate": 1.5539573272815698e-05,
+      "loss": 0.2252,
+      "step": 301
+    },
+    {
+      "epoch": 0.38308668076109936,
+      "grad_norm": 5.97371244430542,
+      "learning_rate": 1.5502628779813774e-05,
+      "loss": 0.3113,
+      "step": 302
+    },
+    {
+      "epoch": 0.3843551797040169,
+      "grad_norm": 2.844496965408325,
+      "learning_rate": 1.54655762487512e-05,
+      "loss": 0.1796,
+      "step": 303
+    },
+    {
+      "epoch": 0.3856236786469345,
+      "grad_norm": 5.323549270629883,
+      "learning_rate": 1.542841640711365e-05,
+      "loss": 0.2381,
+      "step": 304
+    },
+    {
+      "epoch": 0.386892177589852,
+      "grad_norm": 3.897357940673828,
+      "learning_rate": 1.539114998449372e-05,
+      "loss": 0.2174,
+      "step": 305
+    },
+    {
+      "epoch": 0.38816067653276953,
+      "grad_norm": 3.345202922821045,
+      "learning_rate": 1.5353777712576613e-05,
+      "loss": 0.2236,
+      "step": 306
+    },
+    {
+      "epoch": 0.3894291754756871,
+      "grad_norm": 4.982361316680908,
+      "learning_rate": 1.531630032512575e-05,
+      "loss": 0.2484,
+      "step": 307
+    },
+    {
+      "epoch": 0.39069767441860465,
+      "grad_norm": 3.730459213256836,
+      "learning_rate": 1.5278718557968405e-05,
+      "loss": 0.2516,
+      "step": 308
+    },
+    {
+      "epoch": 0.3919661733615222,
+      "grad_norm": 5.202997207641602,
+      "learning_rate": 1.5241033148981209e-05,
+      "loss": 0.2951,
+      "step": 309
+    },
+    {
+      "epoch": 0.39323467230443976,
+      "grad_norm": 3.5829482078552246,
+      "learning_rate": 1.5203244838075702e-05,
+      "loss": 0.2491,
+      "step": 310
+    },
+    {
+      "epoch": 0.3945031712473573,
+      "grad_norm": 6.405603408813477,
+      "learning_rate": 1.5165354367183778e-05,
+      "loss": 0.2377,
+      "step": 311
+    },
+    {
+      "epoch": 0.3957716701902748,
+      "grad_norm": 4.980127334594727,
+      "learning_rate": 1.5127362480243139e-05,
+      "loss": 0.1729,
+      "step": 312
+    },
+    {
+      "epoch": 0.3970401691331924,
+      "grad_norm": 3.1744191646575928,
+      "learning_rate": 1.508926992318268e-05,
+      "loss": 0.2174,
+      "step": 313
+    },
+    {
+      "epoch": 0.39830866807610993,
+      "grad_norm": 5.434924125671387,
+      "learning_rate": 1.505107744390783e-05,
+      "loss": 0.2544,
+      "step": 314
+    },
+    {
+      "epoch": 0.39957716701902746,
+      "grad_norm": 4.3313212394714355,
+      "learning_rate": 1.5012785792285901e-05,
+      "loss": 0.229,
+      "step": 315
+    },
+    {
+      "epoch": 0.40084566596194504,
+      "grad_norm": 3.850680351257324,
+      "learning_rate": 1.4974395720131328e-05,
+      "loss": 0.2952,
+      "step": 316
+    },
+    {
+      "epoch": 0.40211416490486257,
+      "grad_norm": 3.698664665222168,
+      "learning_rate": 1.4935907981190931e-05,
+      "loss": 0.2758,
+      "step": 317
+    },
+    {
+      "epoch": 0.40338266384778015,
+      "grad_norm": 4.980349540710449,
+      "learning_rate": 1.4897323331129113e-05,
+      "loss": 0.2672,
+      "step": 318
+    },
+    {
+      "epoch": 0.4046511627906977,
+      "grad_norm": 4.738137245178223,
+      "learning_rate": 1.4858642527513009e-05,
+      "loss": 0.2359,
+      "step": 319
+    },
+    {
+      "epoch": 0.4059196617336152,
+      "grad_norm": 5.489634990692139,
+      "learning_rate": 1.4819866329797639e-05,
+      "loss": 0.3265,
+      "step": 320
+    },
+    {
+      "epoch": 0.4059196617336152,
+      "eval_accuracy": 0.8543046357615894,
+      "eval_f1": 0.7,
+      "eval_loss": 0.2961479127407074,
+      "eval_precision": 0.8279569892473119,
+      "eval_recall": 0.6062992125984252,
+      "eval_runtime": 51.0166,
+      "eval_samples_per_second": 5.43,
+      "eval_steps_per_second": 0.176,
+      "step": 320
+    },
+    {
+      "epoch": 0.4071881606765328,
+      "grad_norm": 3.285163164138794,
+      "learning_rate": 1.4780995499310973e-05,
+      "loss": 0.1567,
+      "step": 321
+    },
+    {
+      "epoch": 0.4084566596194503,
+      "grad_norm": 3.621706485748291,
+      "learning_rate": 1.4742030799238989e-05,
+      "loss": 0.2204,
+      "step": 322
+    },
+    {
+      "epoch": 0.40972515856236785,
+      "grad_norm": 4.652901649475098,
+      "learning_rate": 1.4702972994610694e-05,
+      "loss": 0.1918,
+      "step": 323
+    },
+    {
+      "epoch": 0.41099365750528544,
+      "grad_norm": 3.95822811126709,
+      "learning_rate": 1.46638228522831e-05,
+      "loss": 0.2537,
+      "step": 324
+    },
+    {
+      "epoch": 0.41226215644820297,
+      "grad_norm": 6.265679836273193,
+      "learning_rate": 1.4624581140926169e-05,
+      "loss": 0.2054,
+      "step": 325
+    },
+    {
+      "epoch": 0.4135306553911205,
+      "grad_norm": 3.0921337604522705,
+      "learning_rate": 1.458524863100772e-05,
+      "loss": 0.2118,
+      "step": 326
+    },
+    {
+      "epoch": 0.4147991543340381,
+      "grad_norm": 3.311448812484741,
+      "learning_rate": 1.45458260947783e-05,
+      "loss": 0.1814,
+      "step": 327
+    },
+    {
+      "epoch": 0.4160676532769556,
+      "grad_norm": 4.024649620056152,
+      "learning_rate": 1.4506314306256025e-05,
+      "loss": 0.2323,
+      "step": 328
+    },
+    {
+      "epoch": 0.41733615221987314,
+      "grad_norm": 4.3980536460876465,
+      "learning_rate": 1.4466714041211383e-05,
+      "loss": 0.2106,
+      "step": 329
+    },
+    {
+      "epoch": 0.4186046511627907,
+      "grad_norm": 5.472320079803467,
+      "learning_rate": 1.4427026077151998e-05,
+      "loss": 0.2769,
+      "step": 330
+    },
+    {
+      "epoch": 0.41987315010570825,
+      "grad_norm": 5.5313401222229,
+      "learning_rate": 1.4387251193307369e-05,
+      "loss": 0.2501,
+      "step": 331
+    },
+    {
+      "epoch": 0.4211416490486258,
+      "grad_norm": 5.064823627471924,
+      "learning_rate": 1.434739017061357e-05,
+      "loss": 0.3063,
+      "step": 332
+    },
+    {
+      "epoch": 0.42241014799154336,
+      "grad_norm": 7.245492458343506,
+      "learning_rate": 1.4307443791697915e-05,
+      "loss": 0.251,
+      "step": 333
+    },
+    {
+      "epoch": 0.4236786469344609,
+      "grad_norm": 4.918938636779785,
+      "learning_rate": 1.4267412840863597e-05,
+      "loss": 0.2897,
+      "step": 334
+    },
+    {
+      "epoch": 0.4249471458773784,
+      "grad_norm": 7.725310325622559,
+      "learning_rate": 1.4227298104074279e-05,
+      "loss": 0.2339,
+      "step": 335
+    },
+    {
+      "epoch": 0.426215644820296,
+      "grad_norm": 3.9810047149658203,
+      "learning_rate": 1.4187100368938678e-05,
+      "loss": 0.2219,
+      "step": 336
+    },
+    {
+      "epoch": 0.42748414376321353,
+      "grad_norm": 4.304353713989258,
+      "learning_rate": 1.4146820424695082e-05,
+      "loss": 0.2547,
+      "step": 337
+    },
+    {
+      "epoch": 0.42875264270613106,
+      "grad_norm": 6.903933525085449,
+      "learning_rate": 1.4106459062195873e-05,
+      "loss": 0.2123,
+      "step": 338
+    },
+    {
+      "epoch": 0.43002114164904864,
+      "grad_norm": 4.345834732055664,
+      "learning_rate": 1.4066017073891988e-05,
+      "loss": 0.2094,
+      "step": 339
+    },
+    {
+      "epoch": 0.4312896405919662,
+      "grad_norm": 4.44146203994751,
+      "learning_rate": 1.4025495253817359e-05,
+      "loss": 0.2217,
+      "step": 340
+    },
+    {
+      "epoch": 0.4312896405919662,
+      "eval_accuracy": 0.8664459161147903,
+      "eval_f1": 0.7430997876857749,
+      "eval_loss": 0.29700928926467896,
+      "eval_precision": 0.8064516129032258,
+      "eval_recall": 0.6889763779527559,
+      "eval_runtime": 51.4883,
+      "eval_samples_per_second": 5.38,
+      "eval_steps_per_second": 0.175,
+      "step": 340
+    },
+    {
+      "epoch": 0.4325581395348837,
+      "grad_norm": 4.186248779296875,
+      "learning_rate": 1.3984894397573332e-05,
+      "loss": 0.2687,
+      "step": 341
+    },
+    {
+      "epoch": 0.4338266384778013,
+      "grad_norm": 4.99129056930542,
+      "learning_rate": 1.3944215302313042e-05,
+      "loss": 0.202,
+      "step": 342
+    },
+    {
+      "epoch": 0.4350951374207188,
+      "grad_norm": 4.0888848304748535,
+      "learning_rate": 1.3903458766725757e-05,
+      "loss": 0.2868,
+      "step": 343
+    },
+    {
+      "epoch": 0.43636363636363634,
+      "grad_norm": 6.3171515464782715,
+      "learning_rate": 1.3862625591021201e-05,
+      "loss": 0.2175,
+      "step": 344
+    },
+    {
+      "epoch": 0.4376321353065539,
+      "grad_norm": 5.681170463562012,
+      "learning_rate": 1.3821716576913841e-05,
+      "loss": 0.2224,
+      "step": 345
+    },
+    {
+      "epoch": 0.43890063424947146,
+      "grad_norm": 3.9973504543304443,
+      "learning_rate": 1.3780732527607157e-05,
+      "loss": 0.2135,
+      "step": 346
+    },
+    {
+      "epoch": 0.440169133192389,
+      "grad_norm": 3.7838385105133057,
+      "learning_rate": 1.3739674247777853e-05,
+      "loss": 0.2475,
+      "step": 347
+    },
+    {
+      "epoch": 0.44143763213530657,
+      "grad_norm": 3.6724958419799805,
+      "learning_rate": 1.3698542543560068e-05,
+      "loss": 0.2295,
+      "step": 348
+    },
+    {
+      "epoch": 0.4427061310782241,
+      "grad_norm": 3.7347805500030518,
+      "learning_rate": 1.3657338222529553e-05,
+      "loss": 0.2553,
+      "step": 349
+    },
+    {
+      "epoch": 0.4439746300211416,
+      "grad_norm": 4.058570384979248,
+      "learning_rate": 1.3616062093687803e-05,
+      "loss": 0.2698,
+      "step": 350
+    },
+    {
+      "epoch": 0.4452431289640592,
+      "grad_norm": 2.770801067352295,
+      "learning_rate": 1.3574714967446193e-05,
+      "loss": 0.1499,
+      "step": 351
+    },
+    {
+      "epoch": 0.44651162790697674,
+      "grad_norm": 5.817844867706299,
+      "learning_rate": 1.3533297655610038e-05,
+      "loss": 0.2548,
+      "step": 352
+    },
+    {
+      "epoch": 0.44778012684989427,
+      "grad_norm": 3.270364999771118,
+      "learning_rate": 1.3491810971362682e-05,
+      "loss": 0.2621,
+      "step": 353
+    },
+    {
+      "epoch": 0.44904862579281185,
+      "grad_norm": 4.292186260223389,
+      "learning_rate": 1.3450255729249513e-05,
+      "loss": 0.2542,
+      "step": 354
+    },
+    {
+      "epoch": 0.4503171247357294,
+      "grad_norm": 3.6082029342651367,
+      "learning_rate": 1.340863274516198e-05,
+      "loss": 0.2861,
+      "step": 355
+    },
+    {
+      "epoch": 0.4515856236786469,
+      "grad_norm": 5.060634136199951,
+      "learning_rate": 1.3366942836321574e-05,
+      "loss": 0.2482,
+      "step": 356
+    },
+    {
+      "epoch": 0.4528541226215645,
+      "grad_norm": 3.972484827041626,
+      "learning_rate": 1.3325186821263777e-05,
+      "loss": 0.2121,
+      "step": 357
+    },
+    {
+      "epoch": 0.454122621564482,
+      "grad_norm": 3.285651445388794,
+      "learning_rate": 1.3283365519821987e-05,
+      "loss": 0.1904,
+      "step": 358
+    },
+    {
+      "epoch": 0.4553911205073996,
+      "grad_norm": 3.950873374938965,
+      "learning_rate": 1.3241479753111441e-05,
+      "loss": 0.2098,
+      "step": 359
+    },
+    {
+      "epoch": 0.45665961945031713,
+      "grad_norm": 2.7246978282928467,
+      "learning_rate": 1.3199530343513077e-05,
+      "loss": 0.2058,
+      "step": 360
+    },
+    {
+      "epoch": 0.45665961945031713,
+      "eval_accuracy": 0.8520971302428256,
+      "eval_f1": 0.6912442396313364,
+      "eval_loss": 0.3054248094558716,
+      "eval_precision": 0.8333333333333334,
+      "eval_recall": 0.5905511811023622,
+      "eval_runtime": 50.5884,
+      "eval_samples_per_second": 5.476,
+      "eval_steps_per_second": 0.178,
+      "step": 360
+    },
+    {
+      "epoch": 0.45792811839323466,
+      "grad_norm": 4.469756603240967,
+      "learning_rate": 1.3157518114657382e-05,
+      "loss": 0.2589,
+      "step": 361
+    },
+    {
+      "epoch": 0.45919661733615225,
+      "grad_norm": 4.308989524841309,
+      "learning_rate": 1.311544389140824e-05,
+      "loss": 0.2859,
+      "step": 362
+    },
+    {
+      "epoch": 0.4604651162790698,
+      "grad_norm": 3.6170761585235596,
+      "learning_rate": 1.3073308499846722e-05,
+      "loss": 0.2399,
+      "step": 363
+    },
+    {
+      "epoch": 0.4617336152219873,
+      "grad_norm": 4.052624225616455,
+      "learning_rate": 1.3031112767254874e-05,
+      "loss": 0.272,
+      "step": 364
+    },
+    {
+      "epoch": 0.4630021141649049,
+      "grad_norm": 3.574326992034912,
+      "learning_rate": 1.2988857522099471e-05,
+      "loss": 0.2331,
+      "step": 365
+    },
+    {
+      "epoch": 0.4642706131078224,
+      "grad_norm": 3.4199771881103516,
+      "learning_rate": 1.2946543594015752e-05,
+      "loss": 0.2147,
+      "step": 366
+    },
+    {
+      "epoch": 0.46553911205073994,
+      "grad_norm": 4.001230716705322,
+      "learning_rate": 1.2904171813791127e-05,
+      "loss": 0.2274,
+      "step": 367
+    },
+    {
+      "epoch": 0.46680761099365753,
+      "grad_norm": 3.2300639152526855,
+      "learning_rate": 1.2861743013348873e-05,
+      "loss": 0.257,
+      "step": 368
+    },
+    {
+      "epoch": 0.46807610993657506,
+      "grad_norm": 3.6694083213806152,
+      "learning_rate": 1.2819258025731798e-05,
+      "loss": 0.175,
+      "step": 369
+    },
+    {
+      "epoch": 0.4693446088794926,
+      "grad_norm": 4.965548992156982,
+      "learning_rate": 1.2776717685085878e-05,
+      "loss": 0.2061,
+      "step": 370
+    },
+    {
+      "epoch": 0.47061310782241017,
+      "grad_norm": 2.8742663860321045,
+      "learning_rate": 1.2734122826643884e-05,
+      "loss": 0.1796,
+      "step": 371
+    },
+    {
+      "epoch": 0.4718816067653277,
+      "grad_norm": 5.175198078155518,
+      "learning_rate": 1.2691474286708995e-05,
+      "loss": 0.2637,
+      "step": 372
+    },
+    {
+      "epoch": 0.4731501057082452,
+      "grad_norm": 3.802945613861084,
+      "learning_rate": 1.2648772902638358e-05,
+      "loss": 0.2745,
+      "step": 373
+    },
+    {
+      "epoch": 0.4744186046511628,
+      "grad_norm": 3.0676567554473877,
+      "learning_rate": 1.2606019512826655e-05,
+      "loss": 0.1904,
+      "step": 374
+    },
+    {
+      "epoch": 0.47568710359408034,
+      "grad_norm": 4.040560722351074,
+      "learning_rate": 1.2563214956689646e-05,
+      "loss": 0.2882,
+      "step": 375
+    },
+    {
+      "epoch": 0.47695560253699787,
+      "grad_norm": 5.133072376251221,
+      "learning_rate": 1.2520360074647686e-05,
+      "loss": 0.1883,
+      "step": 376
+    },
+    {
+      "epoch": 0.47822410147991545,
+      "grad_norm": 3.334303379058838,
+      "learning_rate": 1.2477455708109228e-05,
+      "loss": 0.2287,
+      "step": 377
+    },
+    {
+      "epoch": 0.479492600422833,
+      "grad_norm": 3.445974349975586,
+      "learning_rate": 1.243450269945429e-05,
+      "loss": 0.2616,
+      "step": 378
+    },
+    {
+      "epoch": 0.4807610993657505,
+      "grad_norm": 4.057299613952637,
+      "learning_rate": 1.239150189201793e-05,
+      "loss": 0.2113,
+      "step": 379
+    },
+    {
+      "epoch": 0.4820295983086681,
+      "grad_norm": 3.4984474182128906,
+      "learning_rate": 1.2348454130073679e-05,
+      "loss": 0.225,
+      "step": 380
+    },
+    {
+      "epoch": 0.4820295983086681,
+      "eval_accuracy": 0.8576158940397351,
+      "eval_f1": 0.7006960556844548,
+      "eval_loss": 0.30184322595596313,
+      "eval_precision": 0.8531073446327684,
+      "eval_recall": 0.594488188976378,
+      "eval_runtime": 51.4684,
+      "eval_samples_per_second": 5.382,
+      "eval_steps_per_second": 0.175,
+      "step": 380
+    },
+    {
+      "epoch": 0.4832980972515856,
+      "grad_norm": 5.044597625732422,
+      "learning_rate": 1.2305360258816976e-05,
+      "loss": 0.1806,
+      "step": 381
+    },
+    {
+      "epoch": 0.48456659619450315,
+      "grad_norm": 5.378483295440674,
+      "learning_rate": 1.2262221124348555e-05,
+      "loss": 0.2141,
+      "step": 382
+    },
+    {
+      "epoch": 0.48583509513742074,
+      "grad_norm": 2.9310171604156494,
+      "learning_rate": 1.2219037573657848e-05,
+      "loss": 0.2164,
+      "step": 383
+    },
+    {
+      "epoch": 0.48710359408033826,
+      "grad_norm": 3.2981674671173096,
+      "learning_rate": 1.2175810454606354e-05,
+      "loss": 0.2024,
+      "step": 384
+    },
+    {
+      "epoch": 0.4883720930232558,
+      "grad_norm": 2.9548192024230957,
+      "learning_rate": 1.2132540615910984e-05,
+      "loss": 0.1725,
+      "step": 385
+    },
+    {
+      "epoch": 0.4896405919661734,
+      "grad_norm": 3.742029905319214,
+      "learning_rate": 1.2089228907127403e-05,
+      "loss": 0.2754,
+      "step": 386
+    },
+    {
+      "epoch": 0.4909090909090909,
+      "grad_norm": 3.1863484382629395,
+      "learning_rate": 1.2045876178633354e-05,
+      "loss": 0.1871,
+      "step": 387
+    },
+    {
+      "epoch": 0.49217758985200843,
+      "grad_norm": 3.8084633350372314,
+      "learning_rate": 1.2002483281611945e-05,
+      "loss": 0.2319,
+      "step": 388
+    },
+    {
+      "epoch": 0.493446088794926,
+      "grad_norm": 5.892749786376953,
+      "learning_rate": 1.1959051068034965e-05,
+      "loss": 0.2589,
+      "step": 389
+    },
+    {
+      "epoch": 0.49471458773784355,
+      "grad_norm": 3.523524045944214,
+      "learning_rate": 1.1915580390646129e-05,
+      "loss": 0.1949,
+      "step": 390
+    },
+    {
+      "epoch": 0.4959830866807611,
+      "grad_norm": 3.9826090335845947,
+      "learning_rate": 1.1872072102944345e-05,
+      "loss": 0.202,
+      "step": 391
+    },
+    {
+      "epoch": 0.49725158562367866,
+      "grad_norm": 4.0874810218811035,
+      "learning_rate": 1.182852705916697e-05,
+      "loss": 0.1757,
+      "step": 392
+    },
+    {
+      "epoch": 0.4985200845665962,
+      "grad_norm": 3.664313316345215,
+      "learning_rate": 1.1784946114273015e-05,
+      "loss": 0.2164,
+      "step": 393
+    },
+    {
+      "epoch": 0.4997885835095137,
+      "grad_norm": 3.309058427810669,
+      "learning_rate": 1.1741330123926376e-05,
+      "loss": 0.1956,
+      "step": 394
+    },
+    {
+      "epoch": 0.5010570824524313,
+      "grad_norm": 4.173009872436523,
+      "learning_rate": 1.1697679944479023e-05,
+      "loss": 0.2445,
+      "step": 395
+    },
+    {
+      "epoch": 0.5023255813953489,
+      "grad_norm": 3.826289653778076,
+      "learning_rate": 1.1653996432954194e-05,
+      "loss": 0.2082,
+      "step": 396
+    },
+    {
+      "epoch": 0.5035940803382664,
+      "grad_norm": 4.404917240142822,
+      "learning_rate": 1.161028044702957e-05,
+      "loss": 0.2502,
+      "step": 397
+    },
+    {
+      "epoch": 0.5048625792811839,
+      "grad_norm": 4.247323036193848,
+      "learning_rate": 1.1566532845020422e-05,
+      "loss": 0.1902,
+      "step": 398
+    },
+    {
+      "epoch": 0.5061310782241015,
+      "grad_norm": 3.4806225299835205,
+      "learning_rate": 1.1522754485862773e-05,
+      "loss": 0.2221,
+      "step": 399
+    },
+    {
+      "epoch": 0.507399577167019,
+      "grad_norm": 5.911139965057373,
+      "learning_rate": 1.1478946229096532e-05,
+      "loss": 0.2045,
+      "step": 400
+    },
+    {
+      "epoch": 0.507399577167019,
+      "eval_accuracy": 0.8509933774834437,
+      "eval_f1": 0.6731234866828087,
+      "eval_loss": 0.3174249827861786,
+      "eval_precision": 0.8742138364779874,
+      "eval_recall": 0.547244094488189,
+      "eval_runtime": 51.0242,
+      "eval_samples_per_second": 5.429,
+      "eval_steps_per_second": 0.176,
+      "step": 400
+    },
+    {
+      "epoch": 0.5086680761099366,
+      "grad_norm": 4.422440528869629,
+      "learning_rate": 1.1435108934848609e-05,
+      "loss": 0.2941,
+      "step": 401
+    },
+    {
+      "epoch": 0.5099365750528542,
+      "grad_norm": 3.987483024597168,
+      "learning_rate": 1.1391243463816036e-05,
+      "loss": 0.2109,
+      "step": 402
+    },
+    {
+      "epoch": 0.5112050739957716,
+      "grad_norm": 4.6511640548706055,
+      "learning_rate": 1.1347350677249063e-05,
+      "loss": 0.2082,
+      "step": 403
+    },
+    {
+      "epoch": 0.5124735729386892,
+      "grad_norm": 5.202207565307617,
+      "learning_rate": 1.1303431436934254e-05,
+      "loss": 0.1939,
+      "step": 404
+    },
+    {
+      "epoch": 0.5137420718816068,
+      "grad_norm": 3.7034809589385986,
+      "learning_rate": 1.125948660517756e-05,
+      "loss": 0.1794,
+      "step": 405
+    },
+    {
+      "epoch": 0.5150105708245243,
+      "grad_norm": 4.559153079986572,
+      "learning_rate": 1.1215517044787387e-05,
+      "loss": 0.2288,
+      "step": 406
+    },
+    {
+      "epoch": 0.5162790697674419,
+      "grad_norm": 6.425766944885254,
+      "learning_rate": 1.1171523619057678e-05,
+      "loss": 0.3279,
+      "step": 407
+    },
+    {
+      "epoch": 0.5175475687103595,
+      "grad_norm": 3.744201898574829,
+      "learning_rate": 1.1127507191750926e-05,
+      "loss": 0.1917,
+      "step": 408
+    },
+    {
+      "epoch": 0.5188160676532769,
+      "grad_norm": 4.557126522064209,
+      "learning_rate": 1.1083468627081241e-05,
+      "loss": 0.2885,
+      "step": 409
+    },
+    {
+      "epoch": 0.5200845665961945,
+      "grad_norm": 3.1752729415893555,
+      "learning_rate": 1.1039408789697385e-05,
+      "loss": 0.1734,
+      "step": 410
+    },
+    {
+      "epoch": 0.5213530655391121,
+      "grad_norm": 3.41485595703125,
+      "learning_rate": 1.0995328544665776e-05,
+      "loss": 0.2242,
+      "step": 411
+    },
+    {
+      "epoch": 0.5226215644820296,
+      "grad_norm": 3.8267111778259277,
+      "learning_rate": 1.0951228757453514e-05,
+      "loss": 0.1678,
+      "step": 412
+    },
+    {
+      "epoch": 0.5238900634249471,
+      "grad_norm": 3.5933799743652344,
+      "learning_rate": 1.0907110293911391e-05,
+      "loss": 0.1935,
+      "step": 413
+    },
+    {
+      "epoch": 0.5251585623678647,
+      "grad_norm": 4.974501609802246,
+      "learning_rate": 1.086297402025689e-05,
+      "loss": 0.2342,
+      "step": 414
+    },
+    {
+      "epoch": 0.5264270613107822,
+      "grad_norm": 5.614207744598389,
+      "learning_rate": 1.0818820803057172e-05,
+      "loss": 0.1686,
+      "step": 415
+    },
+    {
+      "epoch": 0.5276955602536998,
+      "grad_norm": 4.421863555908203,
+      "learning_rate": 1.077465150921207e-05,
+      "loss": 0.2413,
+      "step": 416
+    },
+    {
+      "epoch": 0.5289640591966174,
+      "grad_norm": 3.7212207317352295,
+      "learning_rate": 1.073046700593706e-05,
+      "loss": 0.2188,
+      "step": 417
+    },
+    {
+      "epoch": 0.5302325581395348,
+      "grad_norm": 3.744774580001831,
+      "learning_rate": 1.068626816074624e-05,
+      "loss": 0.2309,
+      "step": 418
+    },
+    {
+      "epoch": 0.5315010570824524,
+      "grad_norm": 3.859792709350586,
+      "learning_rate": 1.0642055841435297e-05,
+      "loss": 0.1562,
+      "step": 419
+    },
+    {
+      "epoch": 0.53276955602537,
+      "grad_norm": 7.018131256103516,
+      "learning_rate": 1.0597830916064466e-05,
+      "loss": 0.2368,
+      "step": 420
+    },
+    {
+      "epoch": 0.53276955602537,
+      "eval_accuracy": 0.847682119205298,
+      "eval_f1": 0.6698564593301436,
+      "eval_loss": 0.315592885017395,
+      "eval_precision": 0.8536585365853658,
+      "eval_recall": 0.5511811023622047,
+      "eval_runtime": 51.9783,
+      "eval_samples_per_second": 5.329,
+      "eval_steps_per_second": 0.173,
+      "step": 420
+    },
+    {
+      "epoch": 0.5340380549682875,
+      "grad_norm": 5.954317569732666,
+      "learning_rate": 1.0553594252941489e-05,
+      "loss": 0.2667,
+      "step": 421
+    },
+    {
+      "epoch": 0.5353065539112051,
+      "grad_norm": 2.7260313034057617,
+      "learning_rate": 1.0509346720604568e-05,
+      "loss": 0.1736,
+      "step": 422
+    },
+    {
+      "epoch": 0.5365750528541227,
+      "grad_norm": 6.12712287902832,
+      "learning_rate": 1.0465089187805305e-05,
+      "loss": 0.2367,
+      "step": 423
+    },
+    {
+      "epoch": 0.5378435517970401,
+      "grad_norm": 4.741700172424316,
+      "learning_rate": 1.0420822523491651e-05,
+      "loss": 0.2163,
+      "step": 424
+    },
+    {
+      "epoch": 0.5391120507399577,
+      "grad_norm": 8.61501407623291,
+      "learning_rate": 1.037654759679085e-05,
+      "loss": 0.2217,
+      "step": 425
+    },
+    {
+      "epoch": 0.5403805496828753,
+      "grad_norm": 4.60408353805542,
+      "learning_rate": 1.0332265276992361e-05,
+      "loss": 0.2595,
+      "step": 426
+    },
+    {
+      "epoch": 0.5416490486257928,
+      "grad_norm": 5.619015693664551,
+      "learning_rate": 1.0287976433530803e-05,
+      "loss": 0.2045,
+      "step": 427
+    },
+    {
+      "epoch": 0.5429175475687104,
+      "grad_norm": 6.060495376586914,
+      "learning_rate": 1.0243681935968885e-05,
+      "loss": 0.2218,
+      "step": 428
+    },
+    {
+      "epoch": 0.5441860465116279,
+      "grad_norm": 4.551842212677002,
+      "learning_rate": 1.0199382653980317e-05,
+      "loss": 0.2108,
+      "step": 429
+    },
+    {
+      "epoch": 0.5454545454545454,
+      "grad_norm": 6.0335564613342285,
+      "learning_rate": 1.0155079457332747e-05,
+      "loss": 0.2044,
+      "step": 430
+    },
+    {
+      "epoch": 0.546723044397463,
+      "grad_norm": 3.9287242889404297,
+      "learning_rate": 1.0110773215870695e-05,
+      "loss": 0.1748,
+      "step": 431
+    },
+    {
+      "epoch": 0.5479915433403806,
+      "grad_norm": 5.402927398681641,
+      "learning_rate": 1.0066464799498452e-05,
+      "loss": 0.2594,
+      "step": 432
+    },
+    {
+      "epoch": 0.5492600422832981,
+      "grad_norm": 4.90135383605957,
+      "learning_rate": 1.0022155078163012e-05,
+      "loss": 0.1855,
+      "step": 433
+    },
+    {
+      "epoch": 0.5505285412262156,
+      "grad_norm": 2.856055736541748,
+      "learning_rate": 9.977844921836992e-06,
+      "loss": 0.1365,
+      "step": 434
+    },
+    {
+      "epoch": 0.5517970401691332,
+      "grad_norm": 5.056857109069824,
+      "learning_rate": 9.933535200501552e-06,
+      "loss": 0.1966,
+      "step": 435
+    },
+    {
+      "epoch": 0.5530655391120507,
+      "grad_norm": 5.74174690246582,
+      "learning_rate": 9.889226784129307e-06,
+      "loss": 0.2867,
+      "step": 436
+    },
+    {
+      "epoch": 0.5543340380549683,
+      "grad_norm": 4.972617149353027,
+      "learning_rate": 9.844920542667254e-06,
+      "loss": 0.2438,
+      "step": 437
+    },
+    {
+      "epoch": 0.5556025369978859,
+      "grad_norm": 5.448729038238525,
+      "learning_rate": 9.800617346019687e-06,
+      "loss": 0.2221,
+      "step": 438
+    },
+    {
+      "epoch": 0.5568710359408033,
+      "grad_norm": 4.500537395477295,
+      "learning_rate": 9.756318064031118e-06,
+      "loss": 0.2215,
+      "step": 439
+    },
+    {
+      "epoch": 0.5581395348837209,
+      "grad_norm": 4.469764232635498,
+      "learning_rate": 9.712023566469198e-06,
+      "loss": 0.2162,
+      "step": 440
+    },
+    {
+      "epoch": 0.5581395348837209,
+      "eval_accuracy": 0.8609271523178808,
+      "eval_f1": 0.7136363636363636,
+      "eval_loss": 0.2927682101726532,
+      "eval_precision": 0.8440860215053764,
+      "eval_recall": 0.6181102362204725,
+      "eval_runtime": 52.7005,
+      "eval_samples_per_second": 5.256,
+      "eval_steps_per_second": 0.171,
+      "step": 440
+    },
+    {
+      "epoch": 0.5594080338266385,
+      "grad_norm": 4.238883018493652,
+      "learning_rate": 9.66773472300764e-06,
+      "loss": 0.1607,
+      "step": 441
+    },
+    {
+      "epoch": 0.560676532769556,
+      "grad_norm": 3.689420223236084,
+      "learning_rate": 9.623452403209152e-06,
+      "loss": 0.2105,
+      "step": 442
+    },
+    {
+      "epoch": 0.5619450317124736,
+      "grad_norm": 3.286783456802368,
+      "learning_rate": 9.579177476508352e-06,
+      "loss": 0.2194,
+      "step": 443
+    },
+    {
+      "epoch": 0.5632135306553911,
+      "grad_norm": 4.903306484222412,
+      "learning_rate": 9.5349108121947e-06,
+      "loss": 0.2095,
+      "step": 444
+    },
+    {
+      "epoch": 0.5644820295983086,
+      "grad_norm": 3.390174627304077,
+      "learning_rate": 9.490653279395436e-06,
+      "loss": 0.1972,
+      "step": 445
+    },
+    {
+      "epoch": 0.5657505285412262,
+      "grad_norm": 4.267228603363037,
+      "learning_rate": 9.446405747058513e-06,
+      "loss": 0.2665,
+      "step": 446
+    },
+    {
+      "epoch": 0.5670190274841438,
+      "grad_norm": 4.4697675704956055,
+      "learning_rate": 9.402169083935539e-06,
+      "loss": 0.2574,
+      "step": 447
+    },
+    {
+      "epoch": 0.5682875264270613,
+      "grad_norm": 4.458617210388184,
+      "learning_rate": 9.357944158564708e-06,
+      "loss": 0.1765,
+      "step": 448
+    },
+    {
+      "epoch": 0.5695560253699788,
+      "grad_norm": 4.2114715576171875,
+      "learning_rate": 9.313731839253766e-06,
+      "loss": 0.2469,
+      "step": 449
+    },
+    {
+      "epoch": 0.5708245243128964,
+      "grad_norm": 3.167578935623169,
+      "learning_rate": 9.269532994062945e-06,
+      "loss": 0.2217,
+      "step": 450
+    },
+    {
+      "epoch": 0.5720930232558139,
+      "grad_norm": 3.5999159812927246,
+      "learning_rate": 9.225348490787935e-06,
+      "loss": 0.2433,
+      "step": 451
+    },
+    {
+      "epoch": 0.5733615221987315,
+      "grad_norm": 4.339929580688477,
+      "learning_rate": 9.181179196942831e-06,
+      "loss": 0.2543,
+      "step": 452
+    },
+    {
+      "epoch": 0.5746300211416491,
+      "grad_norm": 3.657891035079956,
+      "learning_rate": 9.137025979743114e-06,
+      "loss": 0.1899,
+      "step": 453
+    },
+    {
+      "epoch": 0.5758985200845665,
+      "grad_norm": 3.0815606117248535,
+      "learning_rate": 9.092889706088615e-06,
+      "loss": 0.1921,
+      "step": 454
+    },
+    {
+      "epoch": 0.5771670190274841,
+      "grad_norm": 3.4993326663970947,
+      "learning_rate": 9.048771242546493e-06,
+      "loss": 0.1811,
+      "step": 455
+    },
+    {
+      "epoch": 0.5784355179704017,
+      "grad_norm": 3.1785824298858643,
+      "learning_rate": 9.004671455334227e-06,
+      "loss": 0.1451,
+      "step": 456
+    },
+    {
+      "epoch": 0.5797040169133192,
+      "grad_norm": 4.485751152038574,
+      "learning_rate": 8.960591210302616e-06,
+      "loss": 0.2327,
+      "step": 457
+    },
+    {
+      "epoch": 0.5809725158562368,
+      "grad_norm": 3.2318646907806396,
+      "learning_rate": 8.91653137291876e-06,
+      "loss": 0.2231,
+      "step": 458
+    },
+    {
+      "epoch": 0.5822410147991544,
+      "grad_norm": 3.670243740081787,
+      "learning_rate": 8.872492808249077e-06,
+      "loss": 0.192,
+      "step": 459
+    },
+    {
+      "epoch": 0.5835095137420718,
+      "grad_norm": 3.487762928009033,
+      "learning_rate": 8.828476380942326e-06,
+      "loss": 0.1664,
+      "step": 460
+    },
+    {
+      "epoch": 0.5835095137420718,
+      "eval_accuracy": 0.8598233995584988,
+      "eval_f1": 0.7146067415730337,
+      "eval_loss": 0.29775571823120117,
+      "eval_precision": 0.8324607329842932,
+      "eval_recall": 0.6259842519685039,
+      "eval_runtime": 52.4539,
+      "eval_samples_per_second": 5.281,
+      "eval_steps_per_second": 0.172,
+      "step": 460
+    },
+    {
+      "epoch": 0.5847780126849894,
+      "grad_norm": 3.1561360359191895,
+      "learning_rate": 8.784482955212615e-06,
+      "loss": 0.1727,
+      "step": 461
+    },
+    {
+      "epoch": 0.586046511627907,
+      "grad_norm": 4.375553131103516,
+      "learning_rate": 8.740513394822442e-06,
+      "loss": 0.2016,
+      "step": 462
+    },
+    {
+      "epoch": 0.5873150105708245,
+      "grad_norm": 3.2783806324005127,
+      "learning_rate": 8.696568563065748e-06,
+      "loss": 0.1808,
+      "step": 463
+    },
+    {
+      "epoch": 0.588583509513742,
+      "grad_norm": 5.312812328338623,
+      "learning_rate": 8.652649322750937e-06,
+      "loss": 0.2486,
+      "step": 464
+    },
+    {
+      "epoch": 0.5898520084566596,
+      "grad_norm": 3.248567819595337,
+      "learning_rate": 8.608756536183964e-06,
+      "loss": 0.2222,
+      "step": 465
+    },
+    {
+      "epoch": 0.5911205073995772,
+      "grad_norm": 4.083889007568359,
+      "learning_rate": 8.564891065151391e-06,
+      "loss": 0.2293,
+      "step": 466
+    },
+    {
+      "epoch": 0.5923890063424947,
+      "grad_norm": 4.238209247589111,
+      "learning_rate": 8.521053770903468e-06,
+      "loss": 0.225,
+      "step": 467
+    },
+    {
+      "epoch": 0.5936575052854123,
+      "grad_norm": 5.558248043060303,
+      "learning_rate": 8.477245514137227e-06,
+      "loss": 0.2204,
+      "step": 468
+    },
+    {
+      "epoch": 0.5949260042283299,
+      "grad_norm": 4.5489420890808105,
+      "learning_rate": 8.43346715497958e-06,
+      "loss": 0.2116,
+      "step": 469
+    },
+    {
+      "epoch": 0.5961945031712473,
+      "grad_norm": 3.887915849685669,
+      "learning_rate": 8.389719552970432e-06,
+      "loss": 0.1674,
+      "step": 470
+    },
+    {
+      "epoch": 0.5974630021141649,
+      "grad_norm": 5.154160976409912,
+      "learning_rate": 8.346003567045806e-06,
+      "loss": 0.2238,
+      "step": 471
+    },
+    {
+      "epoch": 0.5987315010570825,
+      "grad_norm": 4.397689342498779,
+      "learning_rate": 8.30232005552098e-06,
+      "loss": 0.2215,
+      "step": 472
+    },
+    {
+      "epoch": 0.6,
+      "grad_norm": 5.090795040130615,
+      "learning_rate": 8.258669876073626e-06,
+      "loss": 0.1966,
+      "step": 473
+    },
+    {
+      "epoch": 0.6012684989429176,
+      "grad_norm": 2.8848166465759277,
+      "learning_rate": 8.215053885726986e-06,
+      "loss": 0.1757,
+      "step": 474
+    },
+    {
+      "epoch": 0.6025369978858351,
+      "grad_norm": 5.922531604766846,
+      "learning_rate": 8.171472940833033e-06,
+      "loss": 0.2721,
+      "step": 475
+    },
+    {
+      "epoch": 0.6038054968287526,
+      "grad_norm": 3.6184239387512207,
+      "learning_rate": 8.127927897055658e-06,
+      "loss": 0.1644,
+      "step": 476
+    },
+    {
+      "epoch": 0.6050739957716702,
+      "grad_norm": 3.496631145477295,
+      "learning_rate": 8.084419609353875e-06,
+      "loss": 0.1849,
+      "step": 477
+    },
+    {
+      "epoch": 0.6063424947145878,
+      "grad_norm": 4.242512226104736,
+      "learning_rate": 8.040948931965038e-06,
+      "loss": 0.2105,
+      "step": 478
+    },
+    {
+      "epoch": 0.6076109936575053,
+      "grad_norm": 3.4095211029052734,
+      "learning_rate": 7.997516718388056e-06,
+      "loss": 0.1628,
+      "step": 479
+    },
+    {
+      "epoch": 0.6088794926004228,
+      "grad_norm": 5.149474143981934,
+      "learning_rate": 7.95412382136665e-06,
+      "loss": 0.2282,
+      "step": 480
+    },
+    {
+      "epoch": 0.6088794926004228,
+      "eval_accuracy": 0.8587196467991169,
+      "eval_f1": 0.7037037037037037,
+      "eval_loss": 0.3031023144721985,
+      "eval_precision": 0.8539325842696629,
+      "eval_recall": 0.5984251968503937,
+      "eval_runtime": 51.9749,
+      "eval_samples_per_second": 5.329,
+      "eval_steps_per_second": 0.173,
+      "step": 480
+    },
+    {
+      "epoch": 0.6101479915433404,
+      "grad_norm": 4.154394149780273,
+      "learning_rate": 7.910771092872598e-06,
+      "loss": 0.2157,
+      "step": 481
+    },
+    {
+      "epoch": 0.6114164904862579,
+      "grad_norm": 5.618874549865723,
+      "learning_rate": 7.86745938408902e-06,
+      "loss": 0.2325,
+      "step": 482
+    },
+    {
+      "epoch": 0.6126849894291755,
+      "grad_norm": 4.56056547164917,
+      "learning_rate": 7.82418954539365e-06,
+      "loss": 0.2452,
+      "step": 483
+    },
+    {
+      "epoch": 0.6139534883720931,
+      "grad_norm": 5.774714946746826,
+      "learning_rate": 7.780962426342155e-06,
+      "loss": 0.2435,
+      "step": 484
+    },
+    {
+      "epoch": 0.6152219873150105,
+      "grad_norm": 5.51974630355835,
+      "learning_rate": 7.737778875651448e-06,
+      "loss": 0.1966,
+      "step": 485
+    },
+    {
+      "epoch": 0.6164904862579281,
+      "grad_norm": 3.314197063446045,
+      "learning_rate": 7.694639741183027e-06,
+      "loss": 0.1392,
+      "step": 486
+    },
+    {
+      "epoch": 0.6177589852008457,
+      "grad_norm": 4.631697654724121,
+      "learning_rate": 7.651545869926323e-06,
+      "loss": 0.2346,
+      "step": 487
+    },
+    {
+      "epoch": 0.6190274841437632,
+      "grad_norm": 4.110776424407959,
+      "learning_rate": 7.608498107982074e-06,
+      "loss": 0.221,
+      "step": 488
+    },
+    {
+      "epoch": 0.6202959830866808,
+      "grad_norm": 5.305356502532959,
+      "learning_rate": 7.565497300545714e-06,
+      "loss": 0.2169,
+      "step": 489
+    },
+    {
+      "epoch": 0.6215644820295984,
+      "grad_norm": 4.338992595672607,
+      "learning_rate": 7.5225442918907765e-06,
+      "loss": 0.1901,
+      "step": 490
+    },
+    {
+      "epoch": 0.6228329809725158,
+      "grad_norm": 6.824976921081543,
+      "learning_rate": 7.479639925352318e-06,
+      "loss": 0.2836,
+      "step": 491
+    },
+    {
+      "epoch": 0.6241014799154334,
+      "grad_norm": 3.5278079509735107,
+      "learning_rate": 7.436785043310357e-06,
+      "loss": 0.1524,
+      "step": 492
+    },
+    {
+      "epoch": 0.625369978858351,
+      "grad_norm": 4.220887184143066,
+      "learning_rate": 7.393980487173349e-06,
+      "loss": 0.253,
+      "step": 493
+    },
+    {
+      "epoch": 0.6266384778012685,
+      "grad_norm": 4.364911079406738,
+      "learning_rate": 7.351227097361645e-06,
+      "loss": 0.215,
+      "step": 494
+    },
+    {
+      "epoch": 0.627906976744186,
+      "grad_norm": 3.2984468936920166,
+      "learning_rate": 7.308525713291006e-06,
+      "loss": 0.2252,
+      "step": 495
+    },
+    {
+      "epoch": 0.6291754756871036,
+      "grad_norm": 4.866385459899902,
+      "learning_rate": 7.2658771733561175e-06,
+      "loss": 0.2572,
+      "step": 496
+    },
+    {
+      "epoch": 0.6304439746300211,
+      "grad_norm": 4.756896495819092,
+      "learning_rate": 7.223282314914128e-06,
+      "loss": 0.1821,
+      "step": 497
+    },
+    {
+      "epoch": 0.6317124735729387,
+      "grad_norm": 3.7380802631378174,
+      "learning_rate": 7.180741974268207e-06,
+      "loss": 0.1703,
+      "step": 498
+    },
+    {
+      "epoch": 0.6329809725158563,
+      "grad_norm": 3.99318265914917,
+      "learning_rate": 7.138256986651131e-06,
+      "loss": 0.2035,
+      "step": 499
+    },
+    {
+      "epoch": 0.6342494714587738,
+      "grad_norm": 4.226932525634766,
+      "learning_rate": 7.095828186208879e-06,
+      "loss": 0.1983,
+      "step": 500
+    },
+    {
+      "epoch": 0.6342494714587738,
+      "eval_accuracy": 0.8543046357615894,
+      "eval_f1": 0.7040358744394619,
+      "eval_loss": 0.2958061695098877,
+      "eval_precision": 0.8177083333333334,
+      "eval_recall": 0.6181102362204725,
+      "eval_runtime": 51.9773,
+      "eval_samples_per_second": 5.329,
+      "eval_steps_per_second": 0.173,
+      "step": 500
+    },
+    {
+      "epoch": 0.6355179704016913,
+      "grad_norm": 2.9807305335998535,
+      "learning_rate": 7.053456405984254e-06,
+      "loss": 0.2006,
+      "step": 501
+    },
+    {
+      "epoch": 0.6367864693446089,
+      "grad_norm": 3.490462303161621,
+      "learning_rate": 7.011142477900534e-06,
+      "loss": 0.209,
+      "step": 502
+    },
+    {
+      "epoch": 0.6380549682875264,
+      "grad_norm": 4.7009782791137695,
+      "learning_rate": 6.9688872327451295e-06,
+      "loss": 0.1958,
+      "step": 503
+    },
+    {
+      "epoch": 0.639323467230444,
+      "grad_norm": 4.062927722930908,
+      "learning_rate": 6.926691500153283e-06,
+      "loss": 0.2643,
+      "step": 504
+    },
+    {
+      "epoch": 0.6405919661733616,
+      "grad_norm": 5.526817321777344,
+      "learning_rate": 6.884556108591766e-06,
+      "loss": 0.2787,
+      "step": 505
+    },
+    {
+      "epoch": 0.641860465116279,
+      "grad_norm": 5.635672569274902,
+      "learning_rate": 6.842481885342624e-06,
+      "loss": 0.2394,
+      "step": 506
+    },
+    {
+      "epoch": 0.6431289640591966,
+      "grad_norm": 3.4105381965637207,
+      "learning_rate": 6.8004696564869275e-06,
+      "loss": 0.1787,
+      "step": 507
+    },
+    {
+      "epoch": 0.6443974630021142,
+      "grad_norm": 4.923336982727051,
+      "learning_rate": 6.75852024688856e-06,
+      "loss": 0.1556,
+      "step": 508
+    },
+    {
+      "epoch": 0.6456659619450317,
+      "grad_norm": 4.841919422149658,
+      "learning_rate": 6.716634480178016e-06,
+      "loss": 0.2467,
+      "step": 509
+    },
+    {
+      "epoch": 0.6469344608879493,
+      "grad_norm": 3.464397668838501,
+      "learning_rate": 6.674813178736226e-06,
+      "loss": 0.1951,
+      "step": 510
+    },
+    {
+      "epoch": 0.6482029598308668,
+      "grad_norm": 4.985055923461914,
+      "learning_rate": 6.633057163678427e-06,
+      "loss": 0.2217,
+      "step": 511
+    },
+    {
+      "epoch": 0.6494714587737843,
+      "grad_norm": 4.595572471618652,
+      "learning_rate": 6.5913672548380215e-06,
+      "loss": 0.2207,
+      "step": 512
+    },
+    {
+      "epoch": 0.6507399577167019,
+      "grad_norm": 3.702177047729492,
+      "learning_rate": 6.549744270750489e-06,
+      "loss": 0.1988,
+      "step": 513
+    },
+    {
+      "epoch": 0.6520084566596195,
+      "grad_norm": 6.06284236907959,
+      "learning_rate": 6.50818902863732e-06,
+      "loss": 0.2363,
+      "step": 514
+    },
+    {
+      "epoch": 0.653276955602537,
+      "grad_norm": 3.6575632095336914,
+      "learning_rate": 6.466702344389963e-06,
+      "loss": 0.1872,
+      "step": 515
+    },
+    {
+      "epoch": 0.6545454545454545,
+      "grad_norm": 3.904543161392212,
+      "learning_rate": 6.42528503255381e-06,
+      "loss": 0.2535,
+      "step": 516
+    },
+    {
+      "epoch": 0.6558139534883721,
+      "grad_norm": 2.966963291168213,
+      "learning_rate": 6.383937906312196e-06,
+      "loss": 0.1866,
+      "step": 517
+    },
+    {
+      "epoch": 0.6570824524312896,
+      "grad_norm": 3.8866353034973145,
+      "learning_rate": 6.342661777470449e-06,
+      "loss": 0.2196,
+      "step": 518
+    },
+    {
+      "epoch": 0.6583509513742072,
+      "grad_norm": 4.250112533569336,
+      "learning_rate": 6.301457456439932e-06,
+      "loss": 0.1956,
+      "step": 519
+    },
+    {
+      "epoch": 0.6596194503171248,
+      "grad_norm": 4.357515335083008,
+      "learning_rate": 6.260325752222148e-06,
+      "loss": 0.1843,
+      "step": 520
+    },
+    {
+      "epoch": 0.6596194503171248,
+      "eval_accuracy": 0.8609271523178808,
+      "eval_f1": 0.7096774193548387,
+      "eval_loss": 0.30549532175064087,
+      "eval_precision": 0.8555555555555555,
+      "eval_recall": 0.6062992125984252,
+      "eval_runtime": 50.4215,
+      "eval_samples_per_second": 5.494,
+      "eval_steps_per_second": 0.178,
+      "step": 520
+    },
+    {
+      "epoch": 0.6608879492600422,
+      "grad_norm": 3.0417070388793945,
+      "learning_rate": 6.2192674723928425e-06,
+      "loss": 0.2046,
+      "step": 521
+    },
+    {
+      "epoch": 0.6621564482029598,
+      "grad_norm": 3.6613922119140625,
+      "learning_rate": 6.178283423086159e-06,
+      "loss": 0.1782,
+      "step": 522
+    },
+    {
+      "epoch": 0.6634249471458774,
+      "grad_norm": 3.165771722793579,
+      "learning_rate": 6.137374408978803e-06,
+      "loss": 0.2041,
+      "step": 523
+    },
+    {
+      "epoch": 0.6646934460887949,
+      "grad_norm": 4.003574371337891,
+      "learning_rate": 6.096541233274247e-06,
+      "loss": 0.2098,
+      "step": 524
+    },
+    {
+      "epoch": 0.6659619450317125,
+      "grad_norm": 3.318516492843628,
+      "learning_rate": 6.05578469768696e-06,
+      "loss": 0.1887,
+      "step": 525
+    },
+    {
+      "epoch": 0.66723044397463,
+      "grad_norm": 4.255181312561035,
+      "learning_rate": 6.01510560242667e-06,
+      "loss": 0.2046,
+      "step": 526
+    },
+    {
+      "epoch": 0.6684989429175475,
+      "grad_norm": 3.862027883529663,
+      "learning_rate": 5.974504746182644e-06,
+      "loss": 0.1841,
+      "step": 527
+    },
+    {
+      "epoch": 0.6697674418604651,
+      "grad_norm": 3.570117950439453,
+      "learning_rate": 5.933982926108015e-06,
+      "loss": 0.1597,
+      "step": 528
+    },
+    {
+      "epoch": 0.6710359408033827,
+      "grad_norm": 6.440456867218018,
+      "learning_rate": 5.893540937804128e-06,
+      "loss": 0.2544,
+      "step": 529
+    },
+    {
+      "epoch": 0.6723044397463002,
+      "grad_norm": 4.021892547607422,
+      "learning_rate": 5.85317957530492e-06,
+      "loss": 0.2522,
+      "step": 530
+    },
+    {
+      "epoch": 0.6735729386892177,
+      "grad_norm": 4.242708206176758,
+      "learning_rate": 5.812899631061328e-06,
+      "loss": 0.1897,
+      "step": 531
+    },
+    {
+      "epoch": 0.6748414376321353,
+      "grad_norm": 4.2478814125061035,
+      "learning_rate": 5.772701895925725e-06,
+      "loss": 0.2376,
+      "step": 532
+    },
+    {
+      "epoch": 0.6761099365750528,
+      "grad_norm": 3.1216530799865723,
+      "learning_rate": 5.732587159136409e-06,
+      "loss": 0.1965,
+      "step": 533
+    },
+    {
+      "epoch": 0.6773784355179704,
+      "grad_norm": 4.511528968811035,
+      "learning_rate": 5.692556208302089e-06,
+      "loss": 0.2142,
+      "step": 534
+    },
+    {
+      "epoch": 0.678646934460888,
+      "grad_norm": 3.4524171352386475,
+      "learning_rate": 5.652609829386431e-06,
+      "loss": 0.2075,
+      "step": 535
+    },
+    {
+      "epoch": 0.6799154334038054,
+      "grad_norm": 5.186423301696777,
+      "learning_rate": 5.612748806692631e-06,
+      "loss": 0.2504,
+      "step": 536
+    },
+    {
+      "epoch": 0.681183932346723,
+      "grad_norm": 7.290851593017578,
+      "learning_rate": 5.572973922848003e-06,
+      "loss": 0.2263,
+      "step": 537
+    },
+    {
+      "epoch": 0.6824524312896406,
+      "grad_norm": 7.404147624969482,
+      "learning_rate": 5.5332859587886174e-06,
+      "loss": 0.2877,
+      "step": 538
+    },
+    {
+      "epoch": 0.6837209302325581,
+      "grad_norm": 3.5259296894073486,
+      "learning_rate": 5.493685693743975e-06,
+      "loss": 0.2011,
+      "step": 539
+    },
+    {
+      "epoch": 0.6849894291754757,
+      "grad_norm": 5.679382801055908,
+      "learning_rate": 5.4541739052217015e-06,
+      "loss": 0.1915,
+      "step": 540
+    },
+    {
+      "epoch": 0.6849894291754757,
+      "eval_accuracy": 0.8675496688741722,
+      "eval_f1": 0.7424892703862661,
+      "eval_loss": 0.28183531761169434,
+      "eval_precision": 0.8160377358490566,
+      "eval_recall": 0.6811023622047244,
+      "eval_runtime": 50.4206,
+      "eval_samples_per_second": 5.494,
+      "eval_steps_per_second": 0.178,
+      "step": 540
+    },
+    {
+      "epoch": 0.6862579281183933,
+      "grad_norm": 4.955303192138672,
+      "learning_rate": 5.414751368992281e-06,
+      "loss": 0.1991,
+      "step": 541
+    },
+    {
+      "epoch": 0.6875264270613107,
+      "grad_norm": 5.0703349113464355,
+      "learning_rate": 5.3754188590738335e-06,
+      "loss": 0.1959,
+      "step": 542
+    },
+    {
+      "epoch": 0.6887949260042283,
+      "grad_norm": 3.5751047134399414,
+      "learning_rate": 5.3361771477169025e-06,
+      "loss": 0.2537,
+      "step": 543
+    },
+    {
+      "epoch": 0.6900634249471459,
+      "grad_norm": 4.129851341247559,
+      "learning_rate": 5.297027005389309e-06,
+      "loss": 0.2031,
+      "step": 544
+    },
+    {
+      "epoch": 0.6913319238900634,
+      "grad_norm": 3.951871633529663,
+      "learning_rate": 5.257969200761015e-06,
+      "loss": 0.1888,
+      "step": 545
+    },
+    {
+      "epoch": 0.692600422832981,
+      "grad_norm": 4.308104515075684,
+      "learning_rate": 5.219004500689031e-06,
+      "loss": 0.2216,
+      "step": 546
+    },
+    {
+      "epoch": 0.6938689217758985,
+      "grad_norm": 3.6957151889801025,
+      "learning_rate": 5.180133670202363e-06,
+      "loss": 0.2113,
+      "step": 547
+    },
+    {
+      "epoch": 0.695137420718816,
+      "grad_norm": 4.605830669403076,
+      "learning_rate": 5.141357472486994e-06,
+      "loss": 0.1663,
+      "step": 548
+    },
+    {
+      "epoch": 0.6964059196617336,
+      "grad_norm": 4.787519931793213,
+      "learning_rate": 5.102676668870894e-06,
+      "loss": 0.2201,
+      "step": 549
+    },
+    {
+      "epoch": 0.6976744186046512,
+      "grad_norm": 4.204073905944824,
+      "learning_rate": 5.064092018809074e-06,
+      "loss": 0.2575,
+      "step": 550
+    },
+    {
+      "epoch": 0.6989429175475688,
+      "grad_norm": 5.7078962326049805,
+      "learning_rate": 5.025604279868677e-06,
+      "loss": 0.3108,
+      "step": 551
+    },
+    {
+      "epoch": 0.7002114164904862,
+      "grad_norm": 4.210063457489014,
+      "learning_rate": 4.987214207714103e-06,
+      "loss": 0.1687,
+      "step": 552
+    },
+    {
+      "epoch": 0.7014799154334038,
+      "grad_norm": 3.3465261459350586,
+      "learning_rate": 4.948922556092173e-06,
+      "loss": 0.1805,
+      "step": 553
+    },
+    {
+      "epoch": 0.7027484143763214,
+      "grad_norm": 3.795522928237915,
+      "learning_rate": 4.910730076817326e-06,
+      "loss": 0.229,
+      "step": 554
+    },
+    {
+      "epoch": 0.7040169133192389,
+      "grad_norm": 4.867710113525391,
+      "learning_rate": 4.872637519756865e-06,
+      "loss": 0.2243,
+      "step": 555
+    },
+    {
+      "epoch": 0.7052854122621565,
+      "grad_norm": 4.932121753692627,
+      "learning_rate": 4.834645632816227e-06,
+      "loss": 0.2373,
+      "step": 556
+    },
+    {
+      "epoch": 0.706553911205074,
+      "grad_norm": 4.272938251495361,
+      "learning_rate": 4.796755161924305e-06,
+      "loss": 0.2484,
+      "step": 557
+    },
+    {
+      "epoch": 0.7078224101479915,
+      "grad_norm": 6.413752555847168,
+      "learning_rate": 4.7589668510187955e-06,
+      "loss": 0.2344,
+      "step": 558
+    },
+    {
+      "epoch": 0.7090909090909091,
+      "grad_norm": 3.5056440830230713,
+      "learning_rate": 4.721281442031601e-06,
+      "loss": 0.1616,
+      "step": 559
+    },
+    {
+      "epoch": 0.7103594080338267,
+      "grad_norm": 3.480790615081787,
+      "learning_rate": 4.683699674874253e-06,
+      "loss": 0.1582,
+      "step": 560
+    },
+    {
+      "epoch": 0.7103594080338267,
+      "eval_accuracy": 0.8675496688741722,
+      "eval_f1": 0.726027397260274,
+      "eval_loss": 0.288669228553772,
+      "eval_precision": 0.8641304347826086,
+      "eval_recall": 0.6259842519685039,
+      "eval_runtime": 52.0991,
+      "eval_samples_per_second": 5.317,
+      "eval_steps_per_second": 0.173,
+      "step": 560
+    },
+    {
+      "epoch": 0.7116279069767442,
+      "grad_norm": 3.7166647911071777,
+      "learning_rate": 4.646222287423391e-06,
+      "loss": 0.2111,
+      "step": 561
+    },
+    {
+      "epoch": 0.7128964059196617,
+      "grad_norm": 5.089128017425537,
+      "learning_rate": 4.608850015506282e-06,
+      "loss": 0.2095,
+      "step": 562
+    },
+    {
+      "epoch": 0.7141649048625793,
+      "grad_norm": 4.740091800689697,
+      "learning_rate": 4.571583592886352e-06,
+      "loss": 0.1926,
+      "step": 563
+    },
+    {
+      "epoch": 0.7154334038054968,
+      "grad_norm": 3.8879525661468506,
+      "learning_rate": 4.534423751248803e-06,
+      "loss": 0.2297,
+      "step": 564
+    },
+    {
+      "epoch": 0.7167019027484144,
+      "grad_norm": 4.693061351776123,
+      "learning_rate": 4.497371220186229e-06,
+      "loss": 0.2001,
+      "step": 565
+    },
+    {
+      "epoch": 0.717970401691332,
+      "grad_norm": 4.66520881652832,
+      "learning_rate": 4.460426727184305e-06,
+      "loss": 0.2073,
+      "step": 566
+    },
+    {
+      "epoch": 0.7192389006342494,
+      "grad_norm": 3.9349865913391113,
+      "learning_rate": 4.4235909976074935e-06,
+      "loss": 0.2098,
+      "step": 567
+    },
+    {
+      "epoch": 0.720507399577167,
+      "grad_norm": 5.9813618659973145,
+      "learning_rate": 4.386864754684808e-06,
+      "loss": 0.2579,
+      "step": 568
+    },
+    {
+      "epoch": 0.7217758985200846,
+      "grad_norm": 4.006646633148193,
+      "learning_rate": 4.350248719495612e-06,
+      "loss": 0.1935,
+      "step": 569
+    },
+    {
+      "epoch": 0.7230443974630021,
+      "grad_norm": 4.767092227935791,
+      "learning_rate": 4.3137436109554645e-06,
+      "loss": 0.2556,
+      "step": 570
+    },
+    {
+      "epoch": 0.7243128964059197,
+      "grad_norm": 4.517501354217529,
+      "learning_rate": 4.277350145801994e-06,
+      "loss": 0.185,
+      "step": 571
+    },
+    {
+      "epoch": 0.7255813953488373,
+      "grad_norm": 3.7447240352630615,
+      "learning_rate": 4.241069038580845e-06,
+      "loss": 0.2017,
+      "step": 572
+    },
+    {
+      "epoch": 0.7268498942917547,
+      "grad_norm": 5.2144904136657715,
+      "learning_rate": 4.204901001631631e-06,
+      "loss": 0.2049,
+      "step": 573
+    },
+    {
+      "epoch": 0.7281183932346723,
+      "grad_norm": 4.404168128967285,
+      "learning_rate": 4.168846745073952e-06,
+      "loss": 0.266,
+      "step": 574
+    },
+    {
+      "epoch": 0.7293868921775899,
+      "grad_norm": 4.323419570922852,
+      "learning_rate": 4.132906976793463e-06,
+      "loss": 0.1812,
+      "step": 575
+    },
+    {
+      "epoch": 0.7306553911205074,
+      "grad_norm": 3.2697091102600098,
+      "learning_rate": 4.097082402427962e-06,
+      "loss": 0.1758,
+      "step": 576
+    },
+    {
+      "epoch": 0.731923890063425,
+      "grad_norm": 4.7810959815979,
+      "learning_rate": 4.061373725353542e-06,
+      "loss": 0.2368,
+      "step": 577
+    },
+    {
+      "epoch": 0.7331923890063425,
+      "grad_norm": 4.363092422485352,
+      "learning_rate": 4.025781646670784e-06,
+      "loss": 0.1726,
+      "step": 578
+    },
+    {
+      "epoch": 0.73446088794926,
+      "grad_norm": 4.322168827056885,
+      "learning_rate": 3.990306865190983e-06,
+      "loss": 0.2365,
+      "step": 579
+    },
+    {
+      "epoch": 0.7357293868921776,
+      "grad_norm": 3.4661788940429688,
+      "learning_rate": 3.954950077422435e-06,
+      "loss": 0.2003,
+      "step": 580
+    },
+    {
+      "epoch": 0.7357293868921776,
+      "eval_accuracy": 0.8653421633554084,
+      "eval_f1": 0.7239819004524887,
+      "eval_loss": 0.2871930003166199,
+      "eval_precision": 0.851063829787234,
+      "eval_recall": 0.6299212598425197,
+      "eval_runtime": 51.4881,
+      "eval_samples_per_second": 5.38,
+      "eval_steps_per_second": 0.175,
+      "step": 580
+    },
+    {
+      "epoch": 0.7369978858350952,
+      "grad_norm": 4.783055305480957,
+      "learning_rate": 3.9197119775567595e-06,
+      "loss": 0.2522,
+      "step": 581
+    },
+    {
+      "epoch": 0.7382663847780127,
+      "grad_norm": 4.873471260070801,
+      "learning_rate": 3.884593257455268e-06,
+      "loss": 0.2188,
+      "step": 582
+    },
+    {
+      "epoch": 0.7395348837209302,
+      "grad_norm": 4.4479289054870605,
+      "learning_rate": 3.849594606635384e-06,
+      "loss": 0.1991,
+      "step": 583
+    },
+    {
+      "epoch": 0.7408033826638478,
+      "grad_norm": 2.947563409805298,
+      "learning_rate": 3.814716712257101e-06,
+      "loss": 0.1545,
+      "step": 584
+    },
+    {
+      "epoch": 0.7420718816067653,
+      "grad_norm": 3.4630439281463623,
+      "learning_rate": 3.7799602591094953e-06,
+      "loss": 0.1736,
+      "step": 585
+    },
+    {
+      "epoch": 0.7433403805496829,
+      "grad_norm": 3.2017877101898193,
+      "learning_rate": 3.745325929597272e-06,
+      "loss": 0.1384,
+      "step": 586
+    },
+    {
+      "epoch": 0.7446088794926005,
+      "grad_norm": 3.4015791416168213,
+      "learning_rate": 3.7108144037273806e-06,
+      "loss": 0.2033,
+      "step": 587
+    },
+    {
+      "epoch": 0.7458773784355179,
+      "grad_norm": 3.5852890014648438,
+      "learning_rate": 3.676426359095653e-06,
+      "loss": 0.1888,
+      "step": 588
+    },
+    {
+      "epoch": 0.7471458773784355,
+      "grad_norm": 3.599536657333374,
+      "learning_rate": 3.6421624708735002e-06,
+      "loss": 0.1998,
+      "step": 589
+    },
+    {
+      "epoch": 0.7484143763213531,
+      "grad_norm": 5.890554904937744,
+      "learning_rate": 3.6080234117946634e-06,
+      "loss": 0.2303,
+      "step": 590
+    },
+    {
+      "epoch": 0.7496828752642706,
+      "grad_norm": 4.978075981140137,
+      "learning_rate": 3.5740098521419985e-06,
+      "loss": 0.2299,
+      "step": 591
+    },
+    {
+      "epoch": 0.7509513742071882,
+      "grad_norm": 4.91259241104126,
+      "learning_rate": 3.5401224597343175e-06,
+      "loss": 0.222,
+      "step": 592
+    },
+    {
+      "epoch": 0.7522198731501057,
+      "grad_norm": 3.010634422302246,
+      "learning_rate": 3.5063618999132798e-06,
+      "loss": 0.2077,
+      "step": 593
+    },
+    {
+      "epoch": 0.7534883720930232,
+      "grad_norm": 4.121917724609375,
+      "learning_rate": 3.4727288355303247e-06,
+      "loss": 0.2408,
+      "step": 594
+    },
+    {
+      "epoch": 0.7547568710359408,
+      "grad_norm": 6.061969757080078,
+      "learning_rate": 3.4392239269336592e-06,
+      "loss": 0.227,
+      "step": 595
+    },
+    {
+      "epoch": 0.7560253699788584,
+      "grad_norm": 6.199504375457764,
+      "learning_rate": 3.4058478319552935e-06,
+      "loss": 0.2553,
+      "step": 596
+    },
+    {
+      "epoch": 0.7572938689217759,
+      "grad_norm": 3.981107473373413,
+      "learning_rate": 3.372601205898122e-06,
+      "loss": 0.2189,
+      "step": 597
+    },
+    {
+      "epoch": 0.7585623678646934,
+      "grad_norm": 5.17703104019165,
+      "learning_rate": 3.3394847015230613e-06,
+      "loss": 0.2252,
+      "step": 598
+    },
+    {
+      "epoch": 0.759830866807611,
+      "grad_norm": 3.7564640045166016,
+      "learning_rate": 3.3064989690362314e-06,
+      "loss": 0.2137,
+      "step": 599
+    },
+    {
+      "epoch": 0.7610993657505285,
+      "grad_norm": 5.007497310638428,
+      "learning_rate": 3.2736446560761903e-06,
+      "loss": 0.2345,
+      "step": 600
+    },
+    {
+      "epoch": 0.7610993657505285,
+      "eval_accuracy": 0.8686534216335541,
+      "eval_f1": 0.7407407407407407,
+      "eval_loss": 0.282692551612854,
+      "eval_precision": 0.8292682926829268,
+      "eval_recall": 0.6692913385826772,
+      "eval_runtime": 52.302,
+      "eval_samples_per_second": 5.296,
+      "eval_steps_per_second": 0.172,
+      "step": 600
+    }
+  ],
+  "logging_steps": 1,
+  "max_steps": 788,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 1,
+  "save_steps": 100,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": false
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 2.2467637165634355e+17,
+  "train_batch_size": 6,
+  "trial_name": null,
+  "trial_params": null
+}