diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,4678 @@
+{
+  "best_global_step": 60000,
+  "best_metric": 0.9576732752290649,
+  "best_model_checkpoint": "./polyglot-tagger-v2.2/checkpoint-60000",
+  "epoch": 2.0,
+  "eval_steps": 2500,
+  "global_step": 62154,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.003217865589754316,
+      "grad_norm": 12.599648475646973,
+      "learning_rate": 4.9920359108022015e-05,
+      "loss": 8.340531616210937,
+      "step": 100
+    },
+    {
+      "epoch": 0.006435731179508632,
+      "grad_norm": 16.90213966369629,
+      "learning_rate": 4.9839913762589704e-05,
+      "loss": 2.2957839965820312,
+      "step": 200
+    },
+    {
+      "epoch": 0.009653596769262947,
+      "grad_norm": 4.964681625366211,
+      "learning_rate": 4.975946841715739e-05,
+      "loss": 1.2708330535888672,
+      "step": 300
+    },
+    {
+      "epoch": 0.012871462359017264,
+      "grad_norm": 9.05567455291748,
+      "learning_rate": 4.9679023071725076e-05,
+      "loss": 0.8362879180908203,
+      "step": 400
+    },
+    {
+      "epoch": 0.01608932794877158,
+      "grad_norm": 6.991761207580566,
+      "learning_rate": 4.959857772629276e-05,
+      "loss": 0.7179710388183593,
+      "step": 500
+    },
+    {
+      "epoch": 0.019307193538525894,
+      "grad_norm": 6.725792407989502,
+      "learning_rate": 4.951813238086044e-05,
+      "loss": 0.6352015686035156,
+      "step": 600
+    },
+    {
+      "epoch": 0.022525059128280213,
+      "grad_norm": 8.771743774414062,
+      "learning_rate": 4.943768703542813e-05,
+      "loss": 0.5740464782714844,
+      "step": 700
+    },
+    {
+      "epoch": 0.025742924718034528,
+      "grad_norm": 9.227128028869629,
+      "learning_rate": 4.935724168999582e-05,
+      "loss": 0.5493228149414062,
+      "step": 800
+    },
+    {
+      "epoch": 0.028960790307788843,
+      "grad_norm": 11.525238990783691,
+      "learning_rate": 4.927679634456351e-05,
+      "loss": 0.5246843719482421,
+      "step": 900
+    },
+    {
+      "epoch": 0.03217865589754316,
+      "grad_norm": 5.464803695678711,
+      "learning_rate": 4.919635099913119e-05,
+      "loss": 0.5209083938598633,
+      "step": 1000
+    },
+    {
+      "epoch": 0.03539652148729747,
+      "grad_norm": 16.38031005859375,
+      "learning_rate": 4.911590565369888e-05,
+      "loss": 0.4483478546142578,
+      "step": 1100
+    },
+    {
+      "epoch": 0.03861438707705179,
+      "grad_norm": 8.507850646972656,
+      "learning_rate": 4.903546030826657e-05,
+      "loss": 0.4467829513549805,
+      "step": 1200
+    },
+    {
+      "epoch": 0.04183225266680611,
+      "grad_norm": 10.285760879516602,
+      "learning_rate": 4.895501496283425e-05,
+      "loss": 0.44222518920898435,
+      "step": 1300
+    },
+    {
+      "epoch": 0.045050118256560426,
+      "grad_norm": 8.84359359741211,
+      "learning_rate": 4.887456961740194e-05,
+      "loss": 0.3752718734741211,
+      "step": 1400
+    },
+    {
+      "epoch": 0.04826798384631474,
+      "grad_norm": 5.227468490600586,
+      "learning_rate": 4.879412427196963e-05,
+      "loss": 0.4211923217773437,
+      "step": 1500
+    },
+    {
+      "epoch": 0.051485849436069056,
+      "grad_norm": 5.467648983001709,
+      "learning_rate": 4.871367892653732e-05,
+      "loss": 0.3770075988769531,
+      "step": 1600
+    },
+    {
+      "epoch": 0.05470371502582337,
+      "grad_norm": 7.041491508483887,
+      "learning_rate": 4.8633233581105e-05,
+      "loss": 0.39339431762695315,
+      "step": 1700
+    },
+    {
+      "epoch": 0.057921580615577686,
+      "grad_norm": 4.243207931518555,
+      "learning_rate": 4.855278823567269e-05,
+      "loss": 0.38666561126708987,
+      "step": 1800
+    },
+    {
+      "epoch": 0.061139446205332,
+      "grad_norm": 10.47419548034668,
+      "learning_rate": 4.847234289024038e-05,
+      "loss": 0.37535915374755857,
+      "step": 1900
+    },
+    {
+      "epoch": 0.06435731179508632,
+      "grad_norm": 7.903866291046143,
+      "learning_rate": 4.839189754480806e-05,
+      "loss": 0.3448226165771484,
+      "step": 2000
+    },
+    {
+      "epoch": 0.06757517738484063,
+      "grad_norm": 4.801281929016113,
+      "learning_rate": 4.831145219937574e-05,
+      "loss": 0.3735826873779297,
+      "step": 2100
+    },
+    {
+      "epoch": 0.07079304297459495,
+      "grad_norm": 6.289346218109131,
+      "learning_rate": 4.823100685394343e-05,
+      "loss": 0.3530204391479492,
+      "step": 2200
+    },
+    {
+      "epoch": 0.07401090856434926,
+      "grad_norm": 6.725964546203613,
+      "learning_rate": 4.815056150851112e-05,
+      "loss": 0.37652462005615234,
+      "step": 2300
+    },
+    {
+      "epoch": 0.07722877415410358,
+      "grad_norm": 14.810294151306152,
+      "learning_rate": 4.80701161630788e-05,
+      "loss": 0.3361575698852539,
+      "step": 2400
+    },
+    {
+      "epoch": 0.08044663974385789,
+      "grad_norm": 10.04340934753418,
+      "learning_rate": 4.798967081764649e-05,
+      "loss": 0.33932579040527344,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08044663974385789,
+      "eval_accuracy": 0.9722890395064231,
+      "eval_f1": 0.8388115225531707,
+      "eval_loss": 0.10779489576816559,
+      "eval_precision": 0.7998833070774257,
+      "eval_recall": 0.8817226431355398,
+      "eval_runtime": 47.993,
+      "eval_samples_per_second": 520.91,
+      "eval_steps_per_second": 28.942,
+      "step": 2500
+    },
+    {
+      "epoch": 0.08366450533361222,
+      "grad_norm": 5.665945529937744,
+      "learning_rate": 4.790922547221418e-05,
+      "loss": 0.3147930717468262,
+      "step": 2600
+    },
+    {
+      "epoch": 0.08688237092336654,
+      "grad_norm": 5.679959297180176,
+      "learning_rate": 4.782878012678187e-05,
+      "loss": 0.363752555847168,
+      "step": 2700
+    },
+    {
+      "epoch": 0.09010023651312085,
+      "grad_norm": 5.174692153930664,
+      "learning_rate": 4.774833478134955e-05,
+      "loss": 0.3410203552246094,
+      "step": 2800
+    },
+    {
+      "epoch": 0.09331810210287517,
+      "grad_norm": 5.585685729980469,
+      "learning_rate": 4.766788943591724e-05,
+      "loss": 0.3211274337768555,
+      "step": 2900
+    },
+    {
+      "epoch": 0.09653596769262948,
+      "grad_norm": 5.2783684730529785,
+      "learning_rate": 4.758744409048493e-05,
+      "loss": 0.3486956787109375,
+      "step": 3000
+    },
+    {
+      "epoch": 0.0997538332823838,
+      "grad_norm": 6.57411527633667,
+      "learning_rate": 4.750699874505261e-05,
+      "loss": 0.33664627075195314,
+      "step": 3100
+    },
+    {
+      "epoch": 0.10297169887213811,
+      "grad_norm": 8.044240951538086,
+      "learning_rate": 4.74265533996203e-05,
+      "loss": 0.31327726364135744,
+      "step": 3200
+    },
+    {
+      "epoch": 0.10618956446189243,
+      "grad_norm": 6.461747169494629,
+      "learning_rate": 4.734610805418799e-05,
+      "loss": 0.32588130950927735,
+      "step": 3300
+    },
+    {
+      "epoch": 0.10940743005164674,
+      "grad_norm": 2.9853570461273193,
+      "learning_rate": 4.726566270875568e-05,
+      "loss": 0.28595542907714844,
+      "step": 3400
+    },
+    {
+      "epoch": 0.11262529564140106,
+      "grad_norm": 3.4314420223236084,
+      "learning_rate": 4.718521736332336e-05,
+      "loss": 0.32015865325927734,
+      "step": 3500
+    },
+    {
+      "epoch": 0.11584316123115537,
+      "grad_norm": 5.1868205070495605,
+      "learning_rate": 4.7104772017891044e-05,
+      "loss": 0.29926179885864257,
+      "step": 3600
+    },
+    {
+      "epoch": 0.11906102682090969,
+      "grad_norm": 5.119034767150879,
+      "learning_rate": 4.702432667245873e-05,
+      "loss": 0.30344295501708984,
+      "step": 3700
+    },
+    {
+      "epoch": 0.122278892410664,
+      "grad_norm": 10.863602638244629,
+      "learning_rate": 4.694388132702642e-05,
+      "loss": 0.282437686920166,
+      "step": 3800
+    },
+    {
+      "epoch": 0.12549675800041832,
+      "grad_norm": 3.252387046813965,
+      "learning_rate": 4.6863435981594104e-05,
+      "loss": 0.2832884407043457,
+      "step": 3900
+    },
+    {
+      "epoch": 0.12871462359017263,
+      "grad_norm": 12.616849899291992,
+      "learning_rate": 4.678299063616179e-05,
+      "loss": 0.299547119140625,
+      "step": 4000
+    },
+    {
+      "epoch": 0.13193248917992695,
+      "grad_norm": 7.093480110168457,
+      "learning_rate": 4.670254529072948e-05,
+      "loss": 0.26917409896850586,
+      "step": 4100
+    },
+    {
+      "epoch": 0.13515035476968126,
+      "grad_norm": 8.318635940551758,
+      "learning_rate": 4.6622099945297164e-05,
+      "loss": 0.28246700286865234,
+      "step": 4200
+    },
+    {
+      "epoch": 0.13836822035943558,
+      "grad_norm": 6.450869560241699,
+      "learning_rate": 4.654165459986485e-05,
+      "loss": 0.2775975036621094,
+      "step": 4300
+    },
+    {
+      "epoch": 0.1415860859491899,
+      "grad_norm": 7.253055572509766,
+      "learning_rate": 4.646120925443254e-05,
+      "loss": 0.2734929466247559,
+      "step": 4400
+    },
+    {
+      "epoch": 0.1448039515389442,
+      "grad_norm": 2.5073742866516113,
+      "learning_rate": 4.638076390900023e-05,
+      "loss": 0.2605240631103516,
+      "step": 4500
+    },
+    {
+      "epoch": 0.14802181712869852,
+      "grad_norm": 4.528222560882568,
+      "learning_rate": 4.6300318563567914e-05,
+      "loss": 0.2719747543334961,
+      "step": 4600
+    },
+    {
+      "epoch": 0.15123968271845284,
+      "grad_norm": 6.474442481994629,
+      "learning_rate": 4.62198732181356e-05,
+      "loss": 0.27226806640625,
+      "step": 4700
+    },
+    {
+      "epoch": 0.15445754830820715,
+      "grad_norm": 4.291607856750488,
+      "learning_rate": 4.613942787270329e-05,
+      "loss": 0.29810441970825197,
+      "step": 4800
+    },
+    {
+      "epoch": 0.15767541389796147,
+      "grad_norm": 2.773998737335205,
+      "learning_rate": 4.6058982527270974e-05,
+      "loss": 0.2813092041015625,
+      "step": 4900
+    },
+    {
+      "epoch": 0.16089327948771578,
+      "grad_norm": 8.771388053894043,
+      "learning_rate": 4.597853718183866e-05,
+      "loss": 0.3041937828063965,
+      "step": 5000
+    },
+    {
+      "epoch": 0.16089327948771578,
+      "eval_accuracy": 0.9766050453886422,
+      "eval_f1": 0.8726156897994607,
+      "eval_loss": 0.09095935523509979,
+      "eval_precision": 0.8421638308564865,
+      "eval_recall": 0.9053523880577816,
+      "eval_runtime": 38.7855,
+      "eval_samples_per_second": 644.572,
+      "eval_steps_per_second": 35.812,
+      "step": 5000
+    },
+    {
+      "epoch": 0.16411114507747013,
+      "grad_norm": 6.708637237548828,
+      "learning_rate": 4.5898091836406345e-05,
+      "loss": 0.25013633728027346,
+      "step": 5100
+    },
+    {
+      "epoch": 0.16732901066722444,
+      "grad_norm": 4.836703777313232,
+      "learning_rate": 4.5817646490974034e-05,
+      "loss": 0.27049646377563474,
+      "step": 5200
+    },
+    {
+      "epoch": 0.17054687625697876,
+      "grad_norm": 2.853039026260376,
+      "learning_rate": 4.5737201145541717e-05,
+      "loss": 0.2682499885559082,
+      "step": 5300
+    },
+    {
+      "epoch": 0.17376474184673307,
+      "grad_norm": 2.6769561767578125,
+      "learning_rate": 4.5656755800109406e-05,
+      "loss": 0.25310728073120115,
+      "step": 5400
+    },
+    {
+      "epoch": 0.1769826074364874,
+      "grad_norm": 3.740751028060913,
+      "learning_rate": 4.5576310454677095e-05,
+      "loss": 0.2738076400756836,
+      "step": 5500
+    },
+    {
+      "epoch": 0.1802004730262417,
+      "grad_norm": 5.003463268280029,
+      "learning_rate": 4.549586510924478e-05,
+      "loss": 0.23465911865234376,
+      "step": 5600
+    },
+    {
+      "epoch": 0.18341833861599602,
+      "grad_norm": 4.04716682434082,
+      "learning_rate": 4.5415419763812466e-05,
+      "loss": 0.2690970230102539,
+      "step": 5700
+    },
+    {
+      "epoch": 0.18663620420575033,
+      "grad_norm": 4.878298282623291,
+      "learning_rate": 4.5334974418380155e-05,
+      "loss": 0.254027156829834,
+      "step": 5800
+    },
+    {
+      "epoch": 0.18985406979550465,
+      "grad_norm": 6.428683757781982,
+      "learning_rate": 4.5254529072947844e-05,
+      "loss": 0.2576713562011719,
+      "step": 5900
+    },
+    {
+      "epoch": 0.19307193538525896,
+      "grad_norm": 4.079443454742432,
+      "learning_rate": 4.5174083727515526e-05,
+      "loss": 0.27297264099121094,
+      "step": 6000
+    },
+    {
+      "epoch": 0.19628980097501328,
+      "grad_norm": 1.6320624351501465,
+      "learning_rate": 4.5093638382083215e-05,
+      "loss": 0.24345703125,
+      "step": 6100
+    },
+    {
+      "epoch": 0.1995076665647676,
+      "grad_norm": 4.895444869995117,
+      "learning_rate": 4.5013193036650904e-05,
+      "loss": 0.2668925857543945,
+      "step": 6200
+    },
+    {
+      "epoch": 0.2027255321545219,
+      "grad_norm": 15.62102222442627,
+      "learning_rate": 4.493274769121859e-05,
+      "loss": 0.25465171813964843,
+      "step": 6300
+    },
+    {
+      "epoch": 0.20594339774427622,
+      "grad_norm": 3.3204216957092285,
+      "learning_rate": 4.4852302345786276e-05,
+      "loss": 0.2394422721862793,
+      "step": 6400
+    },
+    {
+      "epoch": 0.20916126333403054,
+      "grad_norm": 2.538597822189331,
+      "learning_rate": 4.4771857000353965e-05,
+      "loss": 0.2686458206176758,
+      "step": 6500
+    },
+    {
+      "epoch": 0.21237912892378485,
+      "grad_norm": 3.806776523590088,
+      "learning_rate": 4.469141165492165e-05,
+      "loss": 0.2445933151245117,
+      "step": 6600
+    },
+    {
+      "epoch": 0.21559699451353917,
+      "grad_norm": 5.519633769989014,
+      "learning_rate": 4.4610966309489336e-05,
+      "loss": 0.2316486930847168,
+      "step": 6700
+    },
+    {
+      "epoch": 0.21881486010329348,
+      "grad_norm": 2.570289134979248,
+      "learning_rate": 4.453052096405702e-05,
+      "loss": 0.25618520736694333,
+      "step": 6800
+    },
+    {
+      "epoch": 0.2220327256930478,
+      "grad_norm": 1.9880187511444092,
+      "learning_rate": 4.445007561862471e-05,
+      "loss": 0.22628854751586913,
+      "step": 6900
+    },
+    {
+      "epoch": 0.22525059128280212,
+      "grad_norm": 5.40452766418457,
+      "learning_rate": 4.4369630273192396e-05,
+      "loss": 0.23788238525390626,
+      "step": 7000
+    },
+    {
+      "epoch": 0.22846845687255643,
+      "grad_norm": 3.7318766117095947,
+      "learning_rate": 4.428918492776008e-05,
+      "loss": 0.23460798263549804,
+      "step": 7100
+    },
+    {
+      "epoch": 0.23168632246231075,
+      "grad_norm": 3.2017786502838135,
+      "learning_rate": 4.420873958232777e-05,
+      "loss": 0.2477778434753418,
+      "step": 7200
+    },
+    {
+      "epoch": 0.23490418805206506,
+      "grad_norm": 5.915706157684326,
+      "learning_rate": 4.412829423689546e-05,
+      "loss": 0.2488957405090332,
+      "step": 7300
+    },
+    {
+      "epoch": 0.23812205364181938,
+      "grad_norm": 2.817690849304199,
+      "learning_rate": 4.404784889146314e-05,
+      "loss": 0.24542184829711913,
+      "step": 7400
+    },
+    {
+      "epoch": 0.2413399192315737,
+      "grad_norm": 4.39451789855957,
+      "learning_rate": 4.396740354603083e-05,
+      "loss": 0.2323160743713379,
+      "step": 7500
+    },
+    {
+      "epoch": 0.2413399192315737,
+      "eval_accuracy": 0.9765879819893235,
+      "eval_f1": 0.8910072212088652,
+      "eval_loss": 0.08586511015892029,
+      "eval_precision": 0.8661088374239462,
+      "eval_recall": 0.9173795037367669,
+      "eval_runtime": 38.8792,
+      "eval_samples_per_second": 643.017,
+      "eval_steps_per_second": 35.726,
+      "step": 7500
+    },
+    {
+      "epoch": 0.244557784821328,
+      "grad_norm": 6.589978218078613,
+      "learning_rate": 4.388695820059852e-05,
+      "loss": 0.25242706298828127,
+      "step": 7600
+    },
+    {
+      "epoch": 0.24777565041108232,
+      "grad_norm": 8.314870834350586,
+      "learning_rate": 4.3806512855166206e-05,
+      "loss": 0.23692258834838867,
+      "step": 7700
+    },
+    {
+      "epoch": 0.25099351600083664,
+      "grad_norm": 1.829958200454712,
+      "learning_rate": 4.372606750973389e-05,
+      "loss": 0.2302927017211914,
+      "step": 7800
+    },
+    {
+      "epoch": 0.254211381590591,
+      "grad_norm": 2.879436492919922,
+      "learning_rate": 4.364562216430158e-05,
+      "loss": 0.22530841827392578,
+      "step": 7900
+    },
+    {
+      "epoch": 0.25742924718034527,
+      "grad_norm": 4.005130767822266,
+      "learning_rate": 4.3565176818869266e-05,
+      "loss": 0.22251163482666014,
+      "step": 8000
+    },
+    {
+      "epoch": 0.2606471127700996,
+      "grad_norm": 6.088663578033447,
+      "learning_rate": 4.348473147343695e-05,
+      "loss": 0.2417246437072754,
+      "step": 8100
+    },
+    {
+      "epoch": 0.2638649783598539,
+      "grad_norm": 2.821535587310791,
+      "learning_rate": 4.340428612800463e-05,
+      "loss": 0.2411360549926758,
+      "step": 8200
+    },
+    {
+      "epoch": 0.26708284394960824,
+      "grad_norm": 3.160569190979004,
+      "learning_rate": 4.332384078257232e-05,
+      "loss": 0.23668193817138672,
+      "step": 8300
+    },
+    {
+      "epoch": 0.2703007095393625,
+      "grad_norm": 1.9559069871902466,
+      "learning_rate": 4.324339543714001e-05,
+      "loss": 0.20959726333618164,
+      "step": 8400
+    },
+    {
+      "epoch": 0.27351857512911687,
+      "grad_norm": 3.570617914199829,
+      "learning_rate": 4.316295009170769e-05,
+      "loss": 0.23034614562988281,
+      "step": 8500
+    },
+    {
+      "epoch": 0.27673644071887116,
+      "grad_norm": 15.547330856323242,
+      "learning_rate": 4.308250474627538e-05,
+      "loss": 0.22180200576782227,
+      "step": 8600
+    },
+    {
+      "epoch": 0.2799543063086255,
+      "grad_norm": 1.9708352088928223,
+      "learning_rate": 4.300205940084307e-05,
+      "loss": 0.22687753677368164,
+      "step": 8700
+    },
+    {
+      "epoch": 0.2831721718983798,
+      "grad_norm": 2.9480574131011963,
+      "learning_rate": 4.292161405541076e-05,
+      "loss": 0.22114805221557618,
+      "step": 8800
+    },
+    {
+      "epoch": 0.28639003748813413,
+      "grad_norm": 1.7439308166503906,
+      "learning_rate": 4.284116870997844e-05,
+      "loss": 0.20100093841552735,
+      "step": 8900
+    },
+    {
+      "epoch": 0.2896079030778884,
+      "grad_norm": 8.816022872924805,
+      "learning_rate": 4.276072336454613e-05,
+      "loss": 0.21243839263916015,
+      "step": 9000
+    },
+    {
+      "epoch": 0.29282576866764276,
+      "grad_norm": 2.90824556350708,
+      "learning_rate": 4.268027801911382e-05,
+      "loss": 0.22030885696411132,
+      "step": 9100
+    },
+    {
+      "epoch": 0.29604363425739705,
+      "grad_norm": 1.6333738565444946,
+      "learning_rate": 4.25998326736815e-05,
+      "loss": 0.20564973831176758,
+      "step": 9200
+    },
+    {
+      "epoch": 0.2992614998471514,
+      "grad_norm": 2.048818349838257,
+      "learning_rate": 4.251938732824919e-05,
+      "loss": 0.22784378051757812,
+      "step": 9300
+    },
+    {
+      "epoch": 0.3024793654369057,
+      "grad_norm": 2.5882511138916016,
+      "learning_rate": 4.243894198281688e-05,
+      "loss": 0.2064011001586914,
+      "step": 9400
+    },
+    {
+      "epoch": 0.30569723102666,
+      "grad_norm": 4.423720359802246,
+      "learning_rate": 4.235849663738457e-05,
+      "loss": 0.22651321411132813,
+      "step": 9500
+    },
+    {
+      "epoch": 0.3089150966164143,
+      "grad_norm": 6.1782732009887695,
+      "learning_rate": 4.227805129195225e-05,
+      "loss": 0.2322876739501953,
+      "step": 9600
+    },
+    {
+      "epoch": 0.31213296220616865,
+      "grad_norm": 6.551365852355957,
+      "learning_rate": 4.219760594651993e-05,
+      "loss": 0.19593393325805664,
+      "step": 9700
+    },
+    {
+      "epoch": 0.31535082779592294,
+      "grad_norm": 3.024858236312866,
+      "learning_rate": 4.211716060108762e-05,
+      "loss": 0.21738645553588867,
+      "step": 9800
+    },
+    {
+      "epoch": 0.3185686933856773,
+      "grad_norm": 3.413978338241577,
+      "learning_rate": 4.203671525565531e-05,
+      "loss": 0.23522287368774414,
+      "step": 9900
+    },
+    {
+      "epoch": 0.32178655897543157,
+      "grad_norm": 5.521674156188965,
+      "learning_rate": 4.195626991022299e-05,
+      "loss": 0.2253404998779297,
+      "step": 10000
+    },
+    {
+      "epoch": 0.32178655897543157,
+      "eval_accuracy": 0.9818083480354339,
+      "eval_f1": 0.9029431672892276,
+      "eval_loss": 0.07069914042949677,
+      "eval_precision": 0.8804620179673655,
+      "eval_recall": 0.9266024362948766,
+      "eval_runtime": 38.4445,
+      "eval_samples_per_second": 650.288,
+      "eval_steps_per_second": 36.13,
+      "step": 10000
+    },
+    {
+      "epoch": 0.3250044245651859,
+      "grad_norm": 3.399749755859375,
+      "learning_rate": 4.187582456479068e-05,
+      "loss": 0.20138219833374024,
+      "step": 10100
+    },
+    {
+      "epoch": 0.32822229015494025,
+      "grad_norm": 6.1640753746032715,
+      "learning_rate": 4.179537921935837e-05,
+      "loss": 0.2074588966369629,
+      "step": 10200
+    },
+    {
+      "epoch": 0.33144015574469454,
+      "grad_norm": 2.652939796447754,
+      "learning_rate": 4.171493387392605e-05,
+      "loss": 0.20934364318847656,
+      "step": 10300
+    },
+    {
+      "epoch": 0.3346580213344489,
+      "grad_norm": 2.3221635818481445,
+      "learning_rate": 4.163448852849374e-05,
+      "loss": 0.20557579040527343,
+      "step": 10400
+    },
+    {
+      "epoch": 0.3378758869242032,
+      "grad_norm": 4.995593547821045,
+      "learning_rate": 4.155404318306143e-05,
+      "loss": 0.21284431457519531,
+      "step": 10500
+    },
+    {
+      "epoch": 0.3410937525139575,
+      "grad_norm": 5.589059829711914,
+      "learning_rate": 4.147359783762912e-05,
+      "loss": 0.22939311981201171,
+      "step": 10600
+    },
+    {
+      "epoch": 0.3443116181037118,
+      "grad_norm": 3.8397393226623535,
+      "learning_rate": 4.13931524921968e-05,
+      "loss": 0.20622701644897462,
+      "step": 10700
+    },
+    {
+      "epoch": 0.34752948369346615,
+      "grad_norm": 6.400503158569336,
+      "learning_rate": 4.131270714676449e-05,
+      "loss": 0.2185076332092285,
+      "step": 10800
+    },
+    {
+      "epoch": 0.35074734928322043,
+      "grad_norm": 8.725358009338379,
+      "learning_rate": 4.123226180133218e-05,
+      "loss": 0.22046119689941407,
+      "step": 10900
+    },
+    {
+      "epoch": 0.3539652148729748,
+      "grad_norm": 2.3542211055755615,
+      "learning_rate": 4.115181645589986e-05,
+      "loss": 0.21081886291503907,
+      "step": 11000
+    },
+    {
+      "epoch": 0.35718308046272906,
+      "grad_norm": 2.4933526515960693,
+      "learning_rate": 4.107137111046755e-05,
+      "loss": 0.21268159866333008,
+      "step": 11100
+    },
+    {
+      "epoch": 0.3604009460524834,
+      "grad_norm": 2.3393359184265137,
+      "learning_rate": 4.0990925765035234e-05,
+      "loss": 0.20191749572753906,
+      "step": 11200
+    },
+    {
+      "epoch": 0.3636188116422377,
+      "grad_norm": 7.511013507843018,
+      "learning_rate": 4.091048041960292e-05,
+      "loss": 0.22198728561401368,
+      "step": 11300
+    },
+    {
+      "epoch": 0.36683667723199204,
+      "grad_norm": 2.7252819538116455,
+      "learning_rate": 4.0830035074170606e-05,
+      "loss": 0.18651321411132812,
+      "step": 11400
+    },
+    {
+      "epoch": 0.3700545428217463,
+      "grad_norm": 1.7606490850448608,
+      "learning_rate": 4.0749589728738295e-05,
+      "loss": 0.20632423400878908,
+      "step": 11500
+    },
+    {
+      "epoch": 0.37327240841150067,
+      "grad_norm": 1.7523146867752075,
+      "learning_rate": 4.0669144383305984e-05,
+      "loss": 0.19867513656616212,
+      "step": 11600
+    },
+    {
+      "epoch": 0.37649027400125495,
+      "grad_norm": 1.9296563863754272,
+      "learning_rate": 4.058869903787367e-05,
+      "loss": 0.21059492111206055,
+      "step": 11700
+    },
+    {
+      "epoch": 0.3797081395910093,
+      "grad_norm": 2.119272232055664,
+      "learning_rate": 4.0508253692441355e-05,
+      "loss": 0.20072717666625978,
+      "step": 11800
+    },
+    {
+      "epoch": 0.3829260051807636,
+      "grad_norm": 1.8676605224609375,
+      "learning_rate": 4.0427808347009044e-05,
+      "loss": 0.2199223518371582,
+      "step": 11900
+    },
+    {
+      "epoch": 0.3861438707705179,
+      "grad_norm": 7.326332092285156,
+      "learning_rate": 4.034736300157673e-05,
+      "loss": 0.22979413986206054,
+      "step": 12000
+    },
+    {
+      "epoch": 0.3893617363602722,
+      "grad_norm": 6.403603553771973,
+      "learning_rate": 4.0266917656144415e-05,
+      "loss": 0.20364051818847656,
+      "step": 12100
+    },
+    {
+      "epoch": 0.39257960195002656,
+      "grad_norm": 7.924391269683838,
+      "learning_rate": 4.0186472310712104e-05,
+      "loss": 0.21068944931030273,
+      "step": 12200
+    },
+    {
+      "epoch": 0.39579746753978085,
+      "grad_norm": 1.8581446409225464,
+      "learning_rate": 4.010602696527979e-05,
+      "loss": 0.1954026412963867,
+      "step": 12300
+    },
+    {
+      "epoch": 0.3990153331295352,
+      "grad_norm": 4.413888931274414,
+      "learning_rate": 4.002558161984748e-05,
+      "loss": 0.20379119873046875,
+      "step": 12400
+    },
+    {
+      "epoch": 0.4022331987192895,
+      "grad_norm": 7.339420795440674,
+      "learning_rate": 3.9945136274415165e-05,
+      "loss": 0.21170831680297852,
+      "step": 12500
+    },
+    {
+      "epoch": 0.4022331987192895,
+      "eval_accuracy": 0.9805167004142786,
+      "eval_f1": 0.9110038183596832,
+      "eval_loss": 0.07153428345918655,
+      "eval_precision": 0.8942803330689929,
+      "eval_recall": 0.9283646981644178,
+      "eval_runtime": 38.6581,
+      "eval_samples_per_second": 646.695,
+      "eval_steps_per_second": 35.93,
+      "step": 12500
+    },
+    {
+      "epoch": 0.4054510643090438,
+      "grad_norm": 6.818169116973877,
+      "learning_rate": 3.9864690928982854e-05,
+      "loss": 0.1984395217895508,
+      "step": 12600
+    },
+    {
+      "epoch": 0.4086689298987981,
+      "grad_norm": 7.034976959228516,
+      "learning_rate": 3.9784245583550536e-05,
+      "loss": 0.20739023208618165,
+      "step": 12700
+    },
+    {
+      "epoch": 0.41188679548855245,
+      "grad_norm": 1.944700002670288,
+      "learning_rate": 3.9703800238118225e-05,
+      "loss": 0.2069017219543457,
+      "step": 12800
+    },
+    {
+      "epoch": 0.41510466107830674,
+      "grad_norm": 4.255757808685303,
+      "learning_rate": 3.962335489268591e-05,
+      "loss": 0.20037817001342773,
+      "step": 12900
+    },
+    {
+      "epoch": 0.4183225266680611,
+      "grad_norm": 3.274329900741577,
+      "learning_rate": 3.9542909547253596e-05,
+      "loss": 0.20747562408447265,
+      "step": 13000
+    },
+    {
+      "epoch": 0.42154039225781537,
+      "grad_norm": 8.684778213500977,
+      "learning_rate": 3.9462464201821285e-05,
+      "loss": 0.2101104164123535,
+      "step": 13100
+    },
+    {
+      "epoch": 0.4247582578475697,
+      "grad_norm": 2.2195351123809814,
+      "learning_rate": 3.938201885638897e-05,
+      "loss": 0.2074994659423828,
+      "step": 13200
+    },
+    {
+      "epoch": 0.427976123437324,
+      "grad_norm": 5.522794723510742,
+      "learning_rate": 3.9301573510956657e-05,
+      "loss": 0.20569501876831053,
+      "step": 13300
+    },
+    {
+      "epoch": 0.43119398902707834,
+      "grad_norm": 8.248220443725586,
+      "learning_rate": 3.9221128165524346e-05,
+      "loss": 0.20211294174194336,
+      "step": 13400
+    },
+    {
+      "epoch": 0.4344118546168327,
+      "grad_norm": 1.3259437084197998,
+      "learning_rate": 3.9140682820092035e-05,
+      "loss": 0.19619279861450195,
+      "step": 13500
+    },
+    {
+      "epoch": 0.43762972020658697,
+      "grad_norm": 3.264453172683716,
+      "learning_rate": 3.906023747465972e-05,
+      "loss": 0.20112485885620118,
+      "step": 13600
+    },
+    {
+      "epoch": 0.4408475857963413,
+      "grad_norm": 6.523902893066406,
+      "learning_rate": 3.8979792129227406e-05,
+      "loss": 0.17402765274047852,
+      "step": 13700
+    },
+    {
+      "epoch": 0.4440654513860956,
+      "grad_norm": 6.627138137817383,
+      "learning_rate": 3.8899346783795095e-05,
+      "loss": 0.18946147918701173,
+      "step": 13800
+    },
+    {
+      "epoch": 0.44728331697584994,
+      "grad_norm": 1.7494491338729858,
+      "learning_rate": 3.881890143836278e-05,
+      "loss": 0.20346969604492188,
+      "step": 13900
+    },
+    {
+      "epoch": 0.45050118256560423,
+      "grad_norm": 6.50437068939209,
+      "learning_rate": 3.8738456092930466e-05,
+      "loss": 0.19907255172729493,
+      "step": 14000
+    },
+    {
+      "epoch": 0.4537190481553586,
+      "grad_norm": 1.948766827583313,
+      "learning_rate": 3.8658010747498155e-05,
+      "loss": 0.19304462432861327,
+      "step": 14100
+    },
+    {
+      "epoch": 0.45693691374511286,
+      "grad_norm": 2.1899731159210205,
+      "learning_rate": 3.857756540206584e-05,
+      "loss": 0.19804704666137696,
+      "step": 14200
+    },
+    {
+      "epoch": 0.4601547793348672,
+      "grad_norm": 2.7552740573883057,
+      "learning_rate": 3.8497120056633527e-05,
+      "loss": 0.17073585510253905,
+      "step": 14300
+    },
+    {
+      "epoch": 0.4633726449246215,
+      "grad_norm": 3.2083916664123535,
+      "learning_rate": 3.841667471120121e-05,
+      "loss": 0.19360818862915039,
+      "step": 14400
+    },
+    {
+      "epoch": 0.46659051051437583,
+      "grad_norm": 5.908588409423828,
+      "learning_rate": 3.83362293657689e-05,
+      "loss": 0.17787866592407225,
+      "step": 14500
+    },
+    {
+      "epoch": 0.4698083761041301,
+      "grad_norm": 2.6561052799224854,
+      "learning_rate": 3.825578402033658e-05,
+      "loss": 0.20204214096069337,
+      "step": 14600
+    },
+    {
+      "epoch": 0.47302624169388446,
+      "grad_norm": 2.6253390312194824,
+      "learning_rate": 3.817533867490427e-05,
+      "loss": 0.18690492630004882,
+      "step": 14700
+    },
+    {
+      "epoch": 0.47624410728363875,
+      "grad_norm": 1.7459347248077393,
+      "learning_rate": 3.809489332947196e-05,
+      "loss": 0.17405952453613283,
+      "step": 14800
+    },
+    {
+      "epoch": 0.4794619728733931,
+      "grad_norm": 4.657139301300049,
+      "learning_rate": 3.801444798403965e-05,
+      "loss": 0.1884486961364746,
+      "step": 14900
+    },
+    {
+      "epoch": 0.4826798384631474,
+      "grad_norm": 3.4389936923980713,
+      "learning_rate": 3.793400263860733e-05,
+      "loss": 0.1895262336730957,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4826798384631474,
+      "eval_accuracy": 0.9843394189343752,
+      "eval_f1": 0.9178045953844602,
+      "eval_loss": 0.05817751958966255,
+      "eval_precision": 0.8991940957952288,
+      "eval_recall": 0.9372017339627738,
+      "eval_runtime": 38.9423,
+      "eval_samples_per_second": 641.975,
+      "eval_steps_per_second": 35.668,
+      "step": 15000
+    },
+    {
+      "epoch": 0.4858977040529017,
+      "grad_norm": 2.704439640045166,
+      "learning_rate": 3.785355729317502e-05,
+      "loss": 0.1874333381652832,
+      "step": 15100
+    },
+    {
+      "epoch": 0.489115569642656,
+      "grad_norm": 3.225116729736328,
+      "learning_rate": 3.777311194774271e-05,
+      "loss": 0.20594432830810547,
+      "step": 15200
+    },
+    {
+      "epoch": 0.49233343523241035,
+      "grad_norm": 6.360146999359131,
+      "learning_rate": 3.76926666023104e-05,
+      "loss": 0.18481210708618165,
+      "step": 15300
+    },
+    {
+      "epoch": 0.49555130082216464,
+      "grad_norm": 6.072609901428223,
+      "learning_rate": 3.761222125687808e-05,
+      "loss": 0.19338289260864258,
+      "step": 15400
+    },
+    {
+      "epoch": 0.498769166411919,
+      "grad_norm": 6.828498363494873,
+      "learning_rate": 3.753177591144577e-05,
+      "loss": 0.17112209320068358,
+      "step": 15500
+    },
+    {
+      "epoch": 0.5019870320016733,
+      "grad_norm": 1.9960297346115112,
+      "learning_rate": 3.745133056601346e-05,
+      "loss": 0.177659912109375,
+      "step": 15600
+    },
+    {
+      "epoch": 0.5052048975914276,
+      "grad_norm": 2.489583969116211,
+      "learning_rate": 3.737088522058114e-05,
+      "loss": 0.1999117088317871,
+      "step": 15700
+    },
+    {
+      "epoch": 0.508422763181182,
+      "grad_norm": 1.4370602369308472,
+      "learning_rate": 3.729043987514882e-05,
+      "loss": 0.18860567092895508,
+      "step": 15800
+    },
+    {
+      "epoch": 0.5116406287709362,
+      "grad_norm": 1.5181185007095337,
+      "learning_rate": 3.720999452971651e-05,
+      "loss": 0.17629049301147462,
+      "step": 15900
+    },
+    {
+      "epoch": 0.5148584943606905,
+      "grad_norm": 6.286506652832031,
+      "learning_rate": 3.71295491842842e-05,
+      "loss": 0.18211414337158202,
+      "step": 16000
+    },
+    {
+      "epoch": 0.5180763599504449,
+      "grad_norm": 3.1108760833740234,
+      "learning_rate": 3.704910383885188e-05,
+      "loss": 0.18931018829345703,
+      "step": 16100
+    },
+    {
+      "epoch": 0.5212942255401992,
+      "grad_norm": 0.5399872064590454,
+      "learning_rate": 3.696865849341957e-05,
+      "loss": 0.18857126235961913,
+      "step": 16200
+    },
+    {
+      "epoch": 0.5245120911299535,
+      "grad_norm": 1.3765555620193481,
+      "learning_rate": 3.688821314798726e-05,
+      "loss": 0.18414220809936524,
+      "step": 16300
+    },
+    {
+      "epoch": 0.5277299567197078,
+      "grad_norm": 3.104323148727417,
+      "learning_rate": 3.680776780255494e-05,
+      "loss": 0.18788623809814453,
+      "step": 16400
+    },
+    {
+      "epoch": 0.5309478223094621,
+      "grad_norm": 1.609277367591858,
+      "learning_rate": 3.672732245712263e-05,
+      "loss": 0.17455484390258788,
+      "step": 16500
+    },
+    {
+      "epoch": 0.5341656878992165,
+      "grad_norm": 2.4873690605163574,
+      "learning_rate": 3.664687711169032e-05,
+      "loss": 0.17273141860961913,
+      "step": 16600
+    },
+    {
+      "epoch": 0.5373835534889707,
+      "grad_norm": 3.004922866821289,
+      "learning_rate": 3.656643176625801e-05,
+      "loss": 0.17702035903930663,
+      "step": 16700
+    },
+    {
+      "epoch": 0.540601419078725,
+      "grad_norm": 4.731108665466309,
+      "learning_rate": 3.648598642082569e-05,
+      "loss": 0.1878594970703125,
+      "step": 16800
+    },
+    {
+      "epoch": 0.5438192846684794,
+      "grad_norm": 3.8425683975219727,
+      "learning_rate": 3.640554107539338e-05,
+      "loss": 0.19668659210205078,
+      "step": 16900
+    },
+    {
+      "epoch": 0.5470371502582337,
+      "grad_norm": 1.5426945686340332,
+      "learning_rate": 3.632509572996107e-05,
+      "loss": 0.16921035766601564,
+      "step": 17000
+    },
+    {
+      "epoch": 0.5502550158479881,
+      "grad_norm": 1.886399507522583,
+      "learning_rate": 3.624465038452876e-05,
+      "loss": 0.15211493492126466,
+      "step": 17100
+    },
+    {
+      "epoch": 0.5534728814377423,
+      "grad_norm": 5.827595233917236,
+      "learning_rate": 3.616420503909644e-05,
+      "loss": 0.1691839599609375,
+      "step": 17200
+    },
+    {
+      "epoch": 0.5566907470274967,
+      "grad_norm": 5.728984832763672,
+      "learning_rate": 3.608375969366412e-05,
+      "loss": 0.1721731948852539,
+      "step": 17300
+    },
+    {
+      "epoch": 0.559908612617251,
+      "grad_norm": 2.2809104919433594,
+      "learning_rate": 3.600331434823181e-05,
+      "loss": 0.1917552185058594,
+      "step": 17400
+    },
+    {
+      "epoch": 0.5631264782070053,
+      "grad_norm": 2.936466693878174,
+      "learning_rate": 3.5922869002799494e-05,
+      "loss": 0.18653520584106445,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5631264782070053,
+      "eval_accuracy": 0.9851181304305561,
+      "eval_f1": 0.9213907594344987,
+      "eval_loss": 0.055659033358097076,
+      "eval_precision": 0.9052643338588151,
+      "eval_recall": 0.9381021597355321,
+      "eval_runtime": 38.4689,
+      "eval_samples_per_second": 649.876,
+      "eval_steps_per_second": 36.107,
+      "step": 17500
+    },
+    {
+      "epoch": 0.5663443437967596,
+      "grad_norm": 5.648510456085205,
+      "learning_rate": 3.5842423657367184e-05,
+      "loss": 0.17743080139160156,
+      "step": 17600
+    },
+    {
+      "epoch": 0.5695622093865139,
+      "grad_norm": 2.0115761756896973,
+      "learning_rate": 3.576197831193487e-05,
+      "loss": 0.18602855682373046,
+      "step": 17700
+    },
+    {
+      "epoch": 0.5727800749762683,
+      "grad_norm": 5.880187511444092,
+      "learning_rate": 3.568153296650256e-05,
+      "loss": 0.18167713165283203,
+      "step": 17800
+    },
+    {
+      "epoch": 0.5759979405660226,
+      "grad_norm": 4.690539360046387,
+      "learning_rate": 3.5601087621070244e-05,
+      "loss": 0.17695968627929687,
+      "step": 17900
+    },
+    {
+      "epoch": 0.5792158061557768,
+      "grad_norm": 0.8806695342063904,
+      "learning_rate": 3.552064227563793e-05,
+      "loss": 0.19892288208007813,
+      "step": 18000
+    },
+    {
+      "epoch": 0.5824336717455312,
+      "grad_norm": 1.943228006362915,
+      "learning_rate": 3.544019693020562e-05,
+      "loss": 0.17242523193359374,
+      "step": 18100
+    },
+    {
+      "epoch": 0.5856515373352855,
+      "grad_norm": 3.5915210247039795,
+      "learning_rate": 3.5359751584773304e-05,
+      "loss": 0.181878662109375,
+      "step": 18200
+    },
+    {
+      "epoch": 0.5888694029250399,
+      "grad_norm": 2.8968303203582764,
+      "learning_rate": 3.527930623934099e-05,
+      "loss": 0.18603469848632812,
+      "step": 18300
+    },
+    {
+      "epoch": 0.5920872685147941,
+      "grad_norm": 5.2887678146362305,
+      "learning_rate": 3.519886089390868e-05,
+      "loss": 0.16267045974731445,
+      "step": 18400
+    },
+    {
+      "epoch": 0.5953051341045484,
+      "grad_norm": 1.6592695713043213,
+      "learning_rate": 3.511841554847637e-05,
+      "loss": 0.17949144363403322,
+      "step": 18500
+    },
+    {
+      "epoch": 0.5985229996943028,
+      "grad_norm": 0.979318380355835,
+      "learning_rate": 3.5037970203044054e-05,
+      "loss": 0.1779278564453125,
+      "step": 18600
+    },
+    {
+      "epoch": 0.6017408652840571,
+      "grad_norm": 1.6524354219436646,
+      "learning_rate": 3.495752485761174e-05,
+      "loss": 0.1590627098083496,
+      "step": 18700
+    },
+    {
+      "epoch": 0.6049587308738114,
+      "grad_norm": 3.5658767223358154,
+      "learning_rate": 3.4877079512179425e-05,
+      "loss": 0.16901378631591796,
+      "step": 18800
+    },
+    {
+      "epoch": 0.6081765964635657,
+      "grad_norm": 3.9737935066223145,
+      "learning_rate": 3.4796634166747114e-05,
+      "loss": 0.17097728729248046,
+      "step": 18900
+    },
+    {
+      "epoch": 0.61139446205332,
+      "grad_norm": 3.3529911041259766,
+      "learning_rate": 3.4716188821314796e-05,
+      "loss": 0.17232301712036133,
+      "step": 19000
+    },
+    {
+      "epoch": 0.6146123276430744,
+      "grad_norm": 3.2977654933929443,
+      "learning_rate": 3.4635743475882485e-05,
+      "loss": 0.16828372955322266,
+      "step": 19100
+    },
+    {
+      "epoch": 0.6178301932328286,
+      "grad_norm": 3.297731876373291,
+      "learning_rate": 3.4555298130450174e-05,
+      "loss": 0.17515941619873046,
+      "step": 19200
+    },
+    {
+      "epoch": 0.621048058822583,
+      "grad_norm": 7.258795261383057,
+      "learning_rate": 3.4474852785017856e-05,
+      "loss": 0.15848381996154784,
+      "step": 19300
+    },
+    {
+      "epoch": 0.6242659244123373,
+      "grad_norm": 3.2193796634674072,
+      "learning_rate": 3.4394407439585546e-05,
+      "loss": 0.1693815231323242,
+      "step": 19400
+    },
+    {
+      "epoch": 0.6274837900020916,
+      "grad_norm": 1.8956769704818726,
+      "learning_rate": 3.4313962094153235e-05,
+      "loss": 0.1632681083679199,
+      "step": 19500
+    },
+    {
+      "epoch": 0.6307016555918459,
+      "grad_norm": 1.5155613422393799,
+      "learning_rate": 3.4233516748720924e-05,
+      "loss": 0.1833029556274414,
+      "step": 19600
+    },
+    {
+      "epoch": 0.6339195211816002,
+      "grad_norm": 4.2836079597473145,
+      "learning_rate": 3.4153071403288606e-05,
+      "loss": 0.17777244567871095,
+      "step": 19700
+    },
+    {
+      "epoch": 0.6371373867713546,
+      "grad_norm": 5.861156463623047,
+      "learning_rate": 3.4072626057856295e-05,
+      "loss": 0.17992183685302734,
+      "step": 19800
+    },
+    {
+      "epoch": 0.6403552523611089,
+      "grad_norm": 3.4068503379821777,
+      "learning_rate": 3.3992180712423984e-05,
+      "loss": 0.17448163986206056,
+      "step": 19900
+    },
+    {
+      "epoch": 0.6435731179508631,
+      "grad_norm": 2.3339030742645264,
+      "learning_rate": 3.3911735366991666e-05,
+      "loss": 0.16662296295166015,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6435731179508631,
+      "eval_accuracy": 0.985229301062481,
+      "eval_f1": 0.92316609857673,
+      "eval_loss": 0.056044481694698334,
+      "eval_precision": 0.9047148608263973,
+      "eval_recall": 0.9423856137687964,
+      "eval_runtime": 38.8504,
+      "eval_samples_per_second": 643.494,
+      "eval_steps_per_second": 35.753,
+      "step": 20000
+    },
+    {
+      "epoch": 0.6467909835406175,
+      "grad_norm": 3.3607404232025146,
+      "learning_rate": 3.3831290021559355e-05,
+      "loss": 0.17825605392456054,
+      "step": 20100
+    },
+    {
+      "epoch": 0.6500088491303718,
+      "grad_norm": 2.690410852432251,
+      "learning_rate": 3.3750844676127044e-05,
+      "loss": 0.1796116065979004,
+      "step": 20200
+    },
+    {
+      "epoch": 0.6532267147201262,
+      "grad_norm": 5.667492866516113,
+      "learning_rate": 3.3670399330694727e-05,
+      "loss": 0.15698086738586425,
+      "step": 20300
+    },
+    {
+      "epoch": 0.6564445803098805,
+      "grad_norm": 2.9585955142974854,
+      "learning_rate": 3.3589953985262416e-05,
+      "loss": 0.14914603233337403,
+      "step": 20400
+    },
+    {
+      "epoch": 0.6596624458996347,
+      "grad_norm": 3.539393901824951,
+      "learning_rate": 3.35095086398301e-05,
+      "loss": 0.15797216415405274,
+      "step": 20500
+    },
+    {
+      "epoch": 0.6628803114893891,
+      "grad_norm": 3.268399238586426,
+      "learning_rate": 3.342906329439779e-05,
+      "loss": 0.16447280883789062,
+      "step": 20600
+    },
+    {
+      "epoch": 0.6660981770791434,
+      "grad_norm": 1.261047601699829,
+      "learning_rate": 3.3348617948965476e-05,
+      "loss": 0.18879241943359376,
+      "step": 20700
+    },
+    {
+      "epoch": 0.6693160426688978,
+      "grad_norm": 2.1420369148254395,
+      "learning_rate": 3.326817260353316e-05,
+      "loss": 0.15382347106933594,
+      "step": 20800
+    },
+    {
+      "epoch": 0.672533908258652,
+      "grad_norm": 2.8946328163146973,
+      "learning_rate": 3.318772725810085e-05,
+      "loss": 0.15514832496643066,
+      "step": 20900
+    },
+    {
+      "epoch": 0.6757517738484063,
+      "grad_norm": 12.205283164978027,
+      "learning_rate": 3.3107281912668536e-05,
+      "loss": 0.18069747924804688,
+      "step": 21000
+    },
+    {
+      "epoch": 0.6789696394381607,
+      "grad_norm": 1.8224153518676758,
+      "learning_rate": 3.302683656723622e-05,
+      "loss": 0.17442039489746095,
+      "step": 21100
+    },
+    {
+      "epoch": 0.682187505027915,
+      "grad_norm": 3.392157554626465,
+      "learning_rate": 3.294639122180391e-05,
+      "loss": 0.16294483184814454,
+      "step": 21200
+    },
+    {
+      "epoch": 0.6854053706176693,
+      "grad_norm": 1.9432575702667236,
+      "learning_rate": 3.2865945876371597e-05,
+      "loss": 0.17553016662597656,
+      "step": 21300
+    },
+    {
+      "epoch": 0.6886232362074236,
+      "grad_norm": 3.2393903732299805,
+      "learning_rate": 3.2785500530939286e-05,
+      "loss": 0.17193687438964844,
+      "step": 21400
+    },
+    {
+      "epoch": 0.691841101797178,
+      "grad_norm": 3.1432416439056396,
+      "learning_rate": 3.270505518550697e-05,
+      "loss": 0.17834131240844728,
+      "step": 21500
+    },
+    {
+      "epoch": 0.6950589673869323,
+      "grad_norm": 1.5839548110961914,
+      "learning_rate": 3.262460984007466e-05,
+      "loss": 0.16497987747192383,
+      "step": 21600
+    },
+    {
+      "epoch": 0.6982768329766865,
+      "grad_norm": 3.1208765506744385,
+      "learning_rate": 3.2544164494642346e-05,
+      "loss": 0.18163761138916015,
+      "step": 21700
+    },
+    {
+      "epoch": 0.7014946985664409,
+      "grad_norm": 8.323920249938965,
+      "learning_rate": 3.246371914921003e-05,
+      "loss": 0.1704619598388672,
+      "step": 21800
+    },
+    {
+      "epoch": 0.7047125641561952,
+      "grad_norm": 0.9279898405075073,
+      "learning_rate": 3.238327380377772e-05,
+      "loss": 0.1771290397644043,
+      "step": 21900
+    },
+    {
+      "epoch": 0.7079304297459496,
+      "grad_norm": 3.2795073986053467,
+      "learning_rate": 3.23028284583454e-05,
+      "loss": 0.1524122428894043,
+      "step": 22000
+    },
+    {
+      "epoch": 0.7111482953357038,
+      "grad_norm": 5.583573818206787,
+      "learning_rate": 3.222238311291309e-05,
+      "loss": 0.16980670928955077,
+      "step": 22100
+    },
+    {
+      "epoch": 0.7143661609254581,
+      "grad_norm": 4.969387531280518,
+      "learning_rate": 3.214193776748077e-05,
+      "loss": 0.1560416603088379,
+      "step": 22200
+    },
+    {
+      "epoch": 0.7175840265152125,
+      "grad_norm": 3.3920233249664307,
+      "learning_rate": 3.206149242204846e-05,
+      "loss": 0.160349063873291,
+      "step": 22300
+    },
+    {
+      "epoch": 0.7208018921049668,
+      "grad_norm": 2.8945395946502686,
+      "learning_rate": 3.198104707661615e-05,
+      "loss": 0.18149927139282226,
+      "step": 22400
+    },
+    {
+      "epoch": 0.724019757694721,
+      "grad_norm": 1.1273301839828491,
+      "learning_rate": 3.190060173118384e-05,
+      "loss": 0.16230302810668945,
+      "step": 22500
+    },
+    {
+      "epoch": 0.724019757694721,
+      "eval_accuracy": 0.9838156242825616,
+      "eval_f1": 0.9249598360552048,
+      "eval_loss": 0.05883582681417465,
+      "eval_precision": 0.9098816590137007,
+      "eval_recall": 0.9405461725473045,
+      "eval_runtime": 38.7455,
+      "eval_samples_per_second": 645.236,
+      "eval_steps_per_second": 35.849,
+      "step": 22500
+    },
+    {
+      "epoch": 0.7272376232844754,
+      "grad_norm": 2.699639081954956,
+      "learning_rate": 3.182015638575152e-05,
+      "loss": 0.17641155242919923,
+      "step": 22600
+    },
+    {
+      "epoch": 0.7304554888742297,
+      "grad_norm": 4.614871978759766,
+      "learning_rate": 3.173971104031921e-05,
+      "loss": 0.15983717918395995,
+      "step": 22700
+    },
+    {
+      "epoch": 0.7336733544639841,
+      "grad_norm": 2.631415843963623,
+      "learning_rate": 3.16592656948869e-05,
+      "loss": 0.15378172874450682,
+      "step": 22800
+    },
+    {
+      "epoch": 0.7368912200537383,
+      "grad_norm": 1.4634898900985718,
+      "learning_rate": 3.157882034945458e-05,
+      "loss": 0.16743200302124023,
+      "step": 22900
+    },
+    {
+      "epoch": 0.7401090856434926,
+      "grad_norm": 3.4641151428222656,
+      "learning_rate": 3.149837500402227e-05,
+      "loss": 0.15191415786743165,
+      "step": 23000
+    },
+    {
+      "epoch": 0.743326951233247,
+      "grad_norm": 1.500380039215088,
+      "learning_rate": 3.141792965858996e-05,
+      "loss": 0.15310038566589357,
+      "step": 23100
+    },
+    {
+      "epoch": 0.7465448168230013,
+      "grad_norm": 2.184008836746216,
+      "learning_rate": 3.133748431315765e-05,
+      "loss": 0.16115007400512696,
+      "step": 23200
+    },
+    {
+      "epoch": 0.7497626824127556,
+      "grad_norm": 1.1047471761703491,
+      "learning_rate": 3.125703896772533e-05,
+      "loss": 0.153313045501709,
+      "step": 23300
+    },
+    {
+      "epoch": 0.7529805480025099,
+      "grad_norm": 3.3007121086120605,
+      "learning_rate": 3.117659362229301e-05,
+      "loss": 0.16872770309448243,
+      "step": 23400
+    },
+    {
+      "epoch": 0.7561984135922643,
+      "grad_norm": 6.547897815704346,
+      "learning_rate": 3.10961482768607e-05,
+      "loss": 0.18551017761230468,
+      "step": 23500
+    },
+    {
+      "epoch": 0.7594162791820186,
+      "grad_norm": 2.700106382369995,
+      "learning_rate": 3.1015702931428383e-05,
+      "loss": 0.15708372116088867,
+      "step": 23600
+    },
+    {
+      "epoch": 0.7626341447717729,
+      "grad_norm": 1.265189290046692,
+      "learning_rate": 3.093525758599607e-05,
+      "loss": 0.17149799346923827,
+      "step": 23700
+    },
+    {
+      "epoch": 0.7658520103615272,
+      "grad_norm": 6.2682204246521,
+      "learning_rate": 3.085481224056376e-05,
+      "loss": 0.15998153686523436,
+      "step": 23800
+    },
+    {
+      "epoch": 0.7690698759512815,
+      "grad_norm": 3.071823835372925,
+      "learning_rate": 3.077436689513145e-05,
+      "loss": 0.18916090011596678,
+      "step": 23900
+    },
+    {
+      "epoch": 0.7722877415410359,
+      "grad_norm": 2.476935625076294,
+      "learning_rate": 3.069392154969913e-05,
+      "loss": 0.14498735427856446,
+      "step": 24000
+    },
+    {
+      "epoch": 0.7755056071307902,
+      "grad_norm": 4.1790032386779785,
+      "learning_rate": 3.061347620426682e-05,
+      "loss": 0.1612476921081543,
+      "step": 24100
+    },
+    {
+      "epoch": 0.7787234727205444,
+      "grad_norm": 1.0075314044952393,
+      "learning_rate": 3.053303085883451e-05,
+      "loss": 0.14984708786010742,
+      "step": 24200
+    },
+    {
+      "epoch": 0.7819413383102988,
+      "grad_norm": 9.008003234863281,
+      "learning_rate": 3.04525855134022e-05,
+      "loss": 0.17614728927612305,
+      "step": 24300
+    },
+    {
+      "epoch": 0.7851592039000531,
+      "grad_norm": 0.790449321269989,
+      "learning_rate": 3.0372140167969882e-05,
+      "loss": 0.1434452724456787,
+      "step": 24400
+    },
+    {
+      "epoch": 0.7883770694898075,
+      "grad_norm": 2.6443631649017334,
+      "learning_rate": 3.029169482253757e-05,
+      "loss": 0.1625998878479004,
+      "step": 24500
+    },
+    {
+      "epoch": 0.7915949350795617,
+      "grad_norm": 1.1716026067733765,
+      "learning_rate": 3.0211249477105257e-05,
+      "loss": 0.1629466438293457,
+      "step": 24600
+    },
+    {
+      "epoch": 0.794812800669316,
+      "grad_norm": 1.9636585712432861,
+      "learning_rate": 3.013080413167294e-05,
+      "loss": 0.1526513195037842,
+      "step": 24700
+    },
+    {
+      "epoch": 0.7980306662590704,
+      "grad_norm": 6.442037582397461,
+      "learning_rate": 3.0050358786240628e-05,
+      "loss": 0.16380990982055665,
+      "step": 24800
+    },
+    {
+      "epoch": 0.8012485318488247,
+      "grad_norm": 2.0020787715911865,
+      "learning_rate": 2.9969913440808317e-05,
+      "loss": 0.16642404556274415,
+      "step": 24900
+    },
+    {
+      "epoch": 0.804466397438579,
+      "grad_norm": 3.425807476043701,
+      "learning_rate": 2.9889468095376006e-05,
+      "loss": 0.15638751029968262,
+      "step": 25000
+    },
+    {
+      "epoch": 0.804466397438579,
+      "eval_accuracy": 0.9871683237123339,
+      "eval_f1": 0.9358503491978916,
+      "eval_loss": 0.047558050602674484,
+      "eval_precision": 0.9231722273671564,
+      "eval_recall": 0.948881542557981,
+      "eval_runtime": 39.0981,
+      "eval_samples_per_second": 639.417,
+      "eval_steps_per_second": 35.526,
+      "step": 25000
+    },
+    {
+      "epoch": 0.8076842630283333,
+      "grad_norm": 1.7773306369781494,
+      "learning_rate": 2.980902274994369e-05,
+      "loss": 0.16818260192871093,
+      "step": 25100
+    },
+    {
+      "epoch": 0.8109021286180876,
+      "grad_norm": 2.457798957824707,
+      "learning_rate": 2.9728577404511378e-05,
+      "loss": 0.1736893844604492,
+      "step": 25200
+    },
+    {
+      "epoch": 0.814119994207842,
+      "grad_norm": 1.6925740242004395,
+      "learning_rate": 2.9648132059079063e-05,
+      "loss": 0.17085908889770507,
+      "step": 25300
+    },
+    {
+      "epoch": 0.8173378597975962,
+      "grad_norm": 2.4007983207702637,
+      "learning_rate": 2.956768671364675e-05,
+      "loss": 0.16143144607543947,
+      "step": 25400
+    },
+    {
+      "epoch": 0.8205557253873506,
+      "grad_norm": 2.328049898147583,
+      "learning_rate": 2.9487241368214435e-05,
+      "loss": 0.16922952651977538,
+      "step": 25500
+    },
+    {
+      "epoch": 0.8237735909771049,
+      "grad_norm": 2.7495510578155518,
+      "learning_rate": 2.9406796022782124e-05,
+      "loss": 0.16985824584960937,
+      "step": 25600
+    },
+    {
+      "epoch": 0.8269914565668592,
+      "grad_norm": 18.3980712890625,
+      "learning_rate": 2.9326350677349813e-05,
+      "loss": 0.15559816360473633,
+      "step": 25700
+    },
+    {
+      "epoch": 0.8302093221566135,
+      "grad_norm": 2.3473148345947266,
+      "learning_rate": 2.9245905331917495e-05,
+      "loss": 0.17103708267211915,
+      "step": 25800
+    },
+    {
+      "epoch": 0.8334271877463678,
+      "grad_norm": 2.7980849742889404,
+      "learning_rate": 2.9165459986485184e-05,
+      "loss": 0.1549097156524658,
+      "step": 25900
+    },
+    {
+      "epoch": 0.8366450533361222,
+      "grad_norm": 0.4983047842979431,
+      "learning_rate": 2.908501464105287e-05,
+      "loss": 0.15225051879882812,
+      "step": 26000
+    },
+    {
+      "epoch": 0.8398629189258765,
+      "grad_norm": 1.1898616552352905,
+      "learning_rate": 2.900456929562056e-05,
+      "loss": 0.14232300758361816,
+      "step": 26100
+    },
+    {
+      "epoch": 0.8430807845156307,
+      "grad_norm": 1.2203755378723145,
+      "learning_rate": 2.892412395018824e-05,
+      "loss": 0.16007034301757814,
+      "step": 26200
+    },
+    {
+      "epoch": 0.8462986501053851,
+      "grad_norm": 3.160578489303589,
+      "learning_rate": 2.884367860475593e-05,
+      "loss": 0.16384252548217773,
+      "step": 26300
+    },
+    {
+      "epoch": 0.8495165156951394,
+      "grad_norm": 5.011970043182373,
+      "learning_rate": 2.876323325932362e-05,
+      "loss": 0.15432297706604003,
+      "step": 26400
+    },
+    {
+      "epoch": 0.8527343812848938,
+      "grad_norm": 0.8514712452888489,
+      "learning_rate": 2.86827879138913e-05,
+      "loss": 0.15506745338439942,
+      "step": 26500
+    },
+    {
+      "epoch": 0.855952246874648,
+      "grad_norm": 9.957659721374512,
+      "learning_rate": 2.860234256845899e-05,
+      "loss": 0.15732930183410646,
+      "step": 26600
+    },
+    {
+      "epoch": 0.8591701124644023,
+      "grad_norm": 8.271772384643555,
+      "learning_rate": 2.852189722302668e-05,
+      "loss": 0.1552633285522461,
+      "step": 26700
+    },
+    {
+      "epoch": 0.8623879780541567,
+      "grad_norm": 2.02590274810791,
+      "learning_rate": 2.8441451877594365e-05,
+      "loss": 0.15745894432067872,
+      "step": 26800
+    },
+    {
+      "epoch": 0.865605843643911,
+      "grad_norm": 4.529170989990234,
+      "learning_rate": 2.836100653216205e-05,
+      "loss": 0.1467604446411133,
+      "step": 26900
+    },
+    {
+      "epoch": 0.8688237092336654,
+      "grad_norm": 5.628500938415527,
+      "learning_rate": 2.8280561186729736e-05,
+      "loss": 0.1447477436065674,
+      "step": 27000
+    },
+    {
+      "epoch": 0.8720415748234196,
+      "grad_norm": 12.87569522857666,
+      "learning_rate": 2.8200115841297425e-05,
+      "loss": 0.14656261444091798,
+      "step": 27100
+    },
+    {
+      "epoch": 0.8752594404131739,
+      "grad_norm": 3.9200944900512695,
+      "learning_rate": 2.8119670495865107e-05,
+      "loss": 0.16447486877441406,
+      "step": 27200
+    },
+    {
+      "epoch": 0.8784773060029283,
+      "grad_norm": 1.2977664470672607,
+      "learning_rate": 2.8039225150432797e-05,
+      "loss": 0.15925183296203613,
+      "step": 27300
+    },
+    {
+      "epoch": 0.8816951715926826,
+      "grad_norm": 2.786237955093384,
+      "learning_rate": 2.7958779805000486e-05,
+      "loss": 0.15786813735961913,
+      "step": 27400
+    },
+    {
+      "epoch": 0.8849130371824369,
+      "grad_norm": 4.701409339904785,
+      "learning_rate": 2.787833445956817e-05,
+      "loss": 0.14735154151916505,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8849130371824369,
+      "eval_accuracy": 0.985545749558937,
+      "eval_f1": 0.9339101573386709,
+      "eval_loss": 0.052501607686281204,
+      "eval_precision": 0.9199600673862857,
+      "eval_recall": 0.9482898341930256,
+      "eval_runtime": 39.0393,
+      "eval_samples_per_second": 640.38,
+      "eval_steps_per_second": 35.58,
+      "step": 27500
+    },
+    {
+      "epoch": 0.8881309027721912,
+      "grad_norm": 1.7882884740829468,
+      "learning_rate": 2.7797889114135857e-05,
+      "loss": 0.14846997261047362,
+      "step": 27600
+    },
+    {
+      "epoch": 0.8913487683619455,
+      "grad_norm": 1.8413563966751099,
+      "learning_rate": 2.7717443768703542e-05,
+      "loss": 0.15191203117370605,
+      "step": 27700
+    },
+    {
+      "epoch": 0.8945666339516999,
+      "grad_norm": 15.72754192352295,
+      "learning_rate": 2.763699842327123e-05,
+      "loss": 0.1671205520629883,
+      "step": 27800
+    },
+    {
+      "epoch": 0.8977844995414541,
+      "grad_norm": 7.281972885131836,
+      "learning_rate": 2.755655307783892e-05,
+      "loss": 0.15349125862121582,
+      "step": 27900
+    },
+    {
+      "epoch": 0.9010023651312085,
+      "grad_norm": 2.539133071899414,
+      "learning_rate": 2.7476107732406603e-05,
+      "loss": 0.1401151466369629,
+      "step": 28000
+    },
+    {
+      "epoch": 0.9042202307209628,
+      "grad_norm": 1.8258554935455322,
+      "learning_rate": 2.7395662386974292e-05,
+      "loss": 0.14998969078063965,
+      "step": 28100
+    },
+    {
+      "epoch": 0.9074380963107171,
+      "grad_norm": 6.016611099243164,
+      "learning_rate": 2.731521704154198e-05,
+      "loss": 0.14337385177612305,
+      "step": 28200
+    },
+    {
+      "epoch": 0.9106559619004714,
+      "grad_norm": 3.221945285797119,
+      "learning_rate": 2.7234771696109663e-05,
+      "loss": 0.15551527023315428,
+      "step": 28300
+    },
+    {
+      "epoch": 0.9138738274902257,
+      "grad_norm": 1.4967395067214966,
+      "learning_rate": 2.715432635067735e-05,
+      "loss": 0.13511601448059082,
+      "step": 28400
+    },
+    {
+      "epoch": 0.9170916930799801,
+      "grad_norm": 2.9666309356689453,
+      "learning_rate": 2.7073881005245038e-05,
+      "loss": 0.1585793113708496,
+      "step": 28500
+    },
+    {
+      "epoch": 0.9203095586697344,
+      "grad_norm": 0.9828789234161377,
+      "learning_rate": 2.6993435659812727e-05,
+      "loss": 0.15394386291503906,
+      "step": 28600
+    },
+    {
+      "epoch": 0.9235274242594886,
+      "grad_norm": 0.7447288036346436,
+      "learning_rate": 2.691299031438041e-05,
+      "loss": 0.1461707878112793,
+      "step": 28700
+    },
+    {
+      "epoch": 0.926745289849243,
+      "grad_norm": 1.8215960264205933,
+      "learning_rate": 2.6832544968948098e-05,
+      "loss": 0.14361499786376952,
+      "step": 28800
+    },
+    {
+      "epoch": 0.9299631554389973,
+      "grad_norm": 5.320079326629639,
+      "learning_rate": 2.6752099623515787e-05,
+      "loss": 0.15090553283691407,
+      "step": 28900
+    },
+    {
+      "epoch": 0.9331810210287517,
+      "grad_norm": 1.9044791460037231,
+      "learning_rate": 2.667165427808347e-05,
+      "loss": 0.16826240539550782,
+      "step": 29000
+    },
+    {
+      "epoch": 0.9363988866185059,
+      "grad_norm": 2.378955602645874,
+      "learning_rate": 2.659120893265116e-05,
+      "loss": 0.14869707107543945,
+      "step": 29100
+    },
+    {
+      "epoch": 0.9396167522082602,
+      "grad_norm": 3.2531590461730957,
+      "learning_rate": 2.6510763587218844e-05,
+      "loss": 0.15361614227294923,
+      "step": 29200
+    },
+    {
+      "epoch": 0.9428346177980146,
+      "grad_norm": 3.1207923889160156,
+      "learning_rate": 2.6430318241786533e-05,
+      "loss": 0.14827921867370605,
+      "step": 29300
+    },
+    {
+      "epoch": 0.9460524833877689,
+      "grad_norm": 8.111202239990234,
+      "learning_rate": 2.6349872896354215e-05,
+      "loss": 0.14765314102172852,
+      "step": 29400
+    },
+    {
+      "epoch": 0.9492703489775232,
+      "grad_norm": 2.2104053497314453,
+      "learning_rate": 2.6269427550921905e-05,
+      "loss": 0.17542333602905275,
+      "step": 29500
+    },
+    {
+      "epoch": 0.9524882145672775,
+      "grad_norm": 0.9287620782852173,
+      "learning_rate": 2.6188982205489594e-05,
+      "loss": 0.14969840049743652,
+      "step": 29600
+    },
+    {
+      "epoch": 0.9557060801570318,
+      "grad_norm": 2.604642868041992,
+      "learning_rate": 2.6108536860057283e-05,
+      "loss": 0.16406068801879883,
+      "step": 29700
+    },
+    {
+      "epoch": 0.9589239457467862,
+      "grad_norm": 7.544447422027588,
+      "learning_rate": 2.6028091514624965e-05,
+      "loss": 0.1531221103668213,
+      "step": 29800
+    },
+    {
+      "epoch": 0.9621418113365404,
+      "grad_norm": 3.618058204650879,
+      "learning_rate": 2.594764616919265e-05,
+      "loss": 0.16164751052856446,
+      "step": 29900
+    },
+    {
+      "epoch": 0.9653596769262948,
+      "grad_norm": 2.8580358028411865,
+      "learning_rate": 2.586720082376034e-05,
+      "loss": 0.15803168296813966,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9653596769262948,
+      "eval_accuracy": 0.9877764012153277,
+      "eval_f1": 0.9371598218703612,
+      "eval_loss": 0.04550632834434509,
+      "eval_precision": 0.9245153134660876,
+      "eval_recall": 0.9501550018651677,
+      "eval_runtime": 39.0526,
+      "eval_samples_per_second": 640.163,
+      "eval_steps_per_second": 35.567,
+      "step": 30000
+    },
+    {
+      "epoch": 0.9685775425160491,
+      "grad_norm": 1.4846822023391724,
+      "learning_rate": 2.5786755478328022e-05,
+      "loss": 0.138107852935791,
+      "step": 30100
+    },
+    {
+      "epoch": 0.9717954081058034,
+      "grad_norm": 2.2517590522766113,
+      "learning_rate": 2.570631013289571e-05,
+      "loss": 0.13565559387207032,
+      "step": 30200
+    },
+    {
+      "epoch": 0.9750132736955578,
+      "grad_norm": 3.6406161785125732,
+      "learning_rate": 2.56258647874634e-05,
+      "loss": 0.1619549560546875,
+      "step": 30300
+    },
+    {
+      "epoch": 0.978231139285312,
+      "grad_norm": 2.2304012775421143,
+      "learning_rate": 2.554541944203109e-05,
+      "loss": 0.13655080795288085,
+      "step": 30400
+    },
+    {
+      "epoch": 0.9814490048750664,
+      "grad_norm": 4.99883508682251,
+      "learning_rate": 2.546497409659877e-05,
+      "loss": 0.16668659210205078,
+      "step": 30500
+    },
+    {
+      "epoch": 0.9846668704648207,
+      "grad_norm": 2.1127796173095703,
+      "learning_rate": 2.538452875116646e-05,
+      "loss": 0.14745793342590333,
+      "step": 30600
+    },
+    {
+      "epoch": 0.987884736054575,
+      "grad_norm": 0.5811536908149719,
+      "learning_rate": 2.5304083405734146e-05,
+      "loss": 0.1281309223175049,
+      "step": 30700
+    },
+    {
+      "epoch": 0.9911026016443293,
+      "grad_norm": 3.0965957641601562,
+      "learning_rate": 2.5223638060301828e-05,
+      "loss": 0.14012389183044432,
+      "step": 30800
+    },
+    {
+      "epoch": 0.9943204672340836,
+      "grad_norm": 3.952867031097412,
+      "learning_rate": 2.5143192714869517e-05,
+      "loss": 0.18041374206542968,
+      "step": 30900
+    },
+    {
+      "epoch": 0.997538332823838,
+      "grad_norm": 2.5378260612487793,
+      "learning_rate": 2.5062747369437206e-05,
+      "loss": 0.14559303283691405,
+      "step": 31000
+    },
+    {
+      "epoch": 1.0007401090856436,
+      "grad_norm": 2.2179856300354004,
+      "learning_rate": 2.4982302024004892e-05,
+      "loss": 0.13755855560302735,
+      "step": 31100
+    },
+    {
+      "epoch": 1.0039579746753977,
+      "grad_norm": 1.619863510131836,
+      "learning_rate": 2.490185667857258e-05,
+      "loss": 0.11792629241943359,
+      "step": 31200
+    },
+    {
+      "epoch": 1.007175840265152,
+      "grad_norm": 1.596127986907959,
+      "learning_rate": 2.4821411333140267e-05,
+      "loss": 0.1253152847290039,
+      "step": 31300
+    },
+    {
+      "epoch": 1.0103937058549064,
+      "grad_norm": 5.185761451721191,
+      "learning_rate": 2.4740965987707952e-05,
+      "loss": 0.11788433074951171,
+      "step": 31400
+    },
+    {
+      "epoch": 1.0136115714446607,
+      "grad_norm": 4.540038585662842,
+      "learning_rate": 2.4660520642275638e-05,
+      "loss": 0.10257969856262207,
+      "step": 31500
+    },
+    {
+      "epoch": 1.016829437034415,
+      "grad_norm": 0.530720055103302,
+      "learning_rate": 2.4580075296843323e-05,
+      "loss": 0.11232701301574707,
+      "step": 31600
+    },
+    {
+      "epoch": 1.0200473026241694,
+      "grad_norm": 1.8916836977005005,
+      "learning_rate": 2.4499629951411013e-05,
+      "loss": 0.11638439178466797,
+      "step": 31700
+    },
+    {
+      "epoch": 1.0232651682139238,
+      "grad_norm": 1.7947441339492798,
+      "learning_rate": 2.4419184605978698e-05,
+      "loss": 0.1160141658782959,
+      "step": 31800
+    },
+    {
+      "epoch": 1.0264830338036781,
+      "grad_norm": 0.8208370804786682,
+      "learning_rate": 2.4338739260546387e-05,
+      "loss": 0.09769786834716797,
+      "step": 31900
+    },
+    {
+      "epoch": 1.0297008993934322,
+      "grad_norm": 2.4557855129241943,
+      "learning_rate": 2.4258293915114073e-05,
+      "loss": 0.1321820545196533,
+      "step": 32000
+    },
+    {
+      "epoch": 1.0329187649831866,
+      "grad_norm": 1.885315179824829,
+      "learning_rate": 2.4177848569681762e-05,
+      "loss": 0.10417740821838378,
+      "step": 32100
+    },
+    {
+      "epoch": 1.036136630572941,
+      "grad_norm": 0.7314161658287048,
+      "learning_rate": 2.4097403224249444e-05,
+      "loss": 0.11431515693664551,
+      "step": 32200
+    },
+    {
+      "epoch": 1.0393544961626953,
+      "grad_norm": 1.2714358568191528,
+      "learning_rate": 2.4016957878817133e-05,
+      "loss": 0.121004638671875,
+      "step": 32300
+    },
+    {
+      "epoch": 1.0425723617524496,
+      "grad_norm": 3.3896045684814453,
+      "learning_rate": 2.393651253338482e-05,
+      "loss": 0.11515800476074219,
+      "step": 32400
+    },
+    {
+      "epoch": 1.045790227342204,
+      "grad_norm": 1.288989782333374,
+      "learning_rate": 2.3856067187952504e-05,
+      "loss": 0.12445229530334473,
+      "step": 32500
+    },
+    {
+      "epoch": 1.045790227342204,
+      "eval_accuracy": 0.9881399033280868,
+      "eval_f1": 0.9421632544130543,
+      "eval_loss": 0.04471902176737785,
+      "eval_precision": 0.9309708286765537,
+      "eval_recall": 0.9536280727029496,
+      "eval_runtime": 38.6457,
+      "eval_samples_per_second": 646.903,
+      "eval_steps_per_second": 35.942,
+      "step": 32500
+    },
+    {
+      "epoch": 1.0490080929319583,
+      "grad_norm": 1.3189036846160889,
+      "learning_rate": 2.3775621842520194e-05,
+      "loss": 0.11217801094055176,
+      "step": 32600
+    },
+    {
+      "epoch": 1.0522259585217126,
+      "grad_norm": 3.2094037532806396,
+      "learning_rate": 2.369517649708788e-05,
+      "loss": 0.12271183013916015,
+      "step": 32700
+    },
+    {
+      "epoch": 1.055443824111467,
+      "grad_norm": 9.113394737243652,
+      "learning_rate": 2.3614731151655568e-05,
+      "loss": 0.11827351570129395,
+      "step": 32800
+    },
+    {
+      "epoch": 1.058661689701221,
+      "grad_norm": 1.4682255983352661,
+      "learning_rate": 2.3534285806223254e-05,
+      "loss": 0.12056323051452637,
+      "step": 32900
+    },
+    {
+      "epoch": 1.0618795552909754,
+      "grad_norm": 1.7021067142486572,
+      "learning_rate": 2.345384046079094e-05,
+      "loss": 0.10995281219482422,
+      "step": 33000
+    },
+    {
+      "epoch": 1.0650974208807298,
+      "grad_norm": 0.7944279313087463,
+      "learning_rate": 2.3373395115358625e-05,
+      "loss": 0.10448121070861817,
+      "step": 33100
+    },
+    {
+      "epoch": 1.0683152864704841,
+      "grad_norm": 1.9725573062896729,
+      "learning_rate": 2.3292949769926314e-05,
+      "loss": 0.12772043228149413,
+      "step": 33200
+    },
+    {
+      "epoch": 1.0715331520602385,
+      "grad_norm": 4.521146297454834,
+      "learning_rate": 2.3212504424494e-05,
+      "loss": 0.11528082847595215,
+      "step": 33300
+    },
+    {
+      "epoch": 1.0747510176499928,
+      "grad_norm": 3.5572879314422607,
+      "learning_rate": 2.3132059079061685e-05,
+      "loss": 0.1119957447052002,
+      "step": 33400
+    },
+    {
+      "epoch": 1.0779688832397472,
+      "grad_norm": 4.957686424255371,
+      "learning_rate": 2.3051613733629375e-05,
+      "loss": 0.1071746826171875,
+      "step": 33500
+    },
+    {
+      "epoch": 1.0811867488295013,
+      "grad_norm": 1.005800724029541,
+      "learning_rate": 2.297116838819706e-05,
+      "loss": 0.11798532485961914,
+      "step": 33600
+    },
+    {
+      "epoch": 1.0844046144192556,
+      "grad_norm": 1.6847596168518066,
+      "learning_rate": 2.2890723042764746e-05,
+      "loss": 0.12464913368225097,
+      "step": 33700
+    },
+    {
+      "epoch": 1.08762248000901,
+      "grad_norm": 2.2746946811676025,
+      "learning_rate": 2.281027769733243e-05,
+      "loss": 0.11306841850280762,
+      "step": 33800
+    },
+    {
+      "epoch": 1.0908403455987643,
+      "grad_norm": 7.0843706130981445,
+      "learning_rate": 2.272983235190012e-05,
+      "loss": 0.10698895454406739,
+      "step": 33900
+    },
+    {
+      "epoch": 1.0940582111885186,
+      "grad_norm": 1.731086015701294,
+      "learning_rate": 2.2649387006467806e-05,
+      "loss": 0.11211064338684082,
+      "step": 34000
+    },
+    {
+      "epoch": 1.097276076778273,
+      "grad_norm": 3.32621431350708,
+      "learning_rate": 2.2568941661035495e-05,
+      "loss": 0.10381333351135254,
+      "step": 34100
+    },
+    {
+      "epoch": 1.1004939423680273,
+      "grad_norm": 1.3450391292572021,
+      "learning_rate": 2.248849631560318e-05,
+      "loss": 0.12325675010681153,
+      "step": 34200
+    },
+    {
+      "epoch": 1.1037118079577817,
+      "grad_norm": 6.684245586395264,
+      "learning_rate": 2.2408050970170866e-05,
+      "loss": 0.12590248107910157,
+      "step": 34300
+    },
+    {
+      "epoch": 1.106929673547536,
+      "grad_norm": 1.1298221349716187,
+      "learning_rate": 2.2327605624738556e-05,
+      "loss": 0.11126781463623046,
+      "step": 34400
+    },
+    {
+      "epoch": 1.1101475391372901,
+      "grad_norm": 1.185738205909729,
+      "learning_rate": 2.2247160279306238e-05,
+      "loss": 0.11205435752868652,
+      "step": 34500
+    },
+    {
+      "epoch": 1.1133654047270445,
+      "grad_norm": 5.897251605987549,
+      "learning_rate": 2.2166714933873927e-05,
+      "loss": 0.10614056587219238,
+      "step": 34600
+    },
+    {
+      "epoch": 1.1165832703167988,
+      "grad_norm": 0.8509872555732727,
+      "learning_rate": 2.2086269588441612e-05,
+      "loss": 0.11909424781799316,
+      "step": 34700
+    },
+    {
+      "epoch": 1.1198011359065532,
+      "grad_norm": 1.7664456367492676,
+      "learning_rate": 2.20058242430093e-05,
+      "loss": 0.10704885482788086,
+      "step": 34800
+    },
+    {
+      "epoch": 1.1230190014963075,
+      "grad_norm": 4.05808687210083,
+      "learning_rate": 2.1925378897576987e-05,
+      "loss": 0.1148582649230957,
+      "step": 34900
+    },
+    {
+      "epoch": 1.1262368670860619,
+      "grad_norm": 0.689573347568512,
+      "learning_rate": 2.1844933552144676e-05,
+      "loss": 0.11986784934997559,
+      "step": 35000
+    },
+    {
+      "epoch": 1.1262368670860619,
+      "eval_accuracy": 0.988106293602156,
+      "eval_f1": 0.9430045230472126,
+      "eval_loss": 0.04454037919640541,
+      "eval_precision": 0.9315594603074993,
+      "eval_recall": 0.9547343100809097,
+      "eval_runtime": 38.4818,
+      "eval_samples_per_second": 649.658,
+      "eval_steps_per_second": 36.095,
+      "step": 35000
+    },
+    {
+      "epoch": 1.1294547326758162,
+      "grad_norm": 1.2704740762710571,
+      "learning_rate": 2.1764488206712362e-05,
+      "loss": 0.11099970817565918,
+      "step": 35100
+    },
+    {
+      "epoch": 1.1326725982655705,
+      "grad_norm": 5.144682884216309,
+      "learning_rate": 2.1684042861280047e-05,
+      "loss": 0.13487217903137208,
+      "step": 35200
+    },
+    {
+      "epoch": 1.1358904638553247,
+      "grad_norm": 0.7087175846099854,
+      "learning_rate": 2.1603597515847733e-05,
+      "loss": 0.12489970207214356,
+      "step": 35300
+    },
+    {
+      "epoch": 1.139108329445079,
+      "grad_norm": 1.559787631034851,
+      "learning_rate": 2.152315217041542e-05,
+      "loss": 0.10530885696411132,
+      "step": 35400
+    },
+    {
+      "epoch": 1.1423261950348333,
+      "grad_norm": 1.2589160203933716,
+      "learning_rate": 2.1442706824983108e-05,
+      "loss": 0.12130918502807617,
+      "step": 35500
+    },
+    {
+      "epoch": 1.1455440606245877,
+      "grad_norm": 1.1291875839233398,
+      "learning_rate": 2.1362261479550793e-05,
+      "loss": 0.11352409362792969,
+      "step": 35600
+    },
+    {
+      "epoch": 1.148761926214342,
+      "grad_norm": 5.766851425170898,
+      "learning_rate": 2.1281816134118483e-05,
+      "loss": 0.12505414009094237,
+      "step": 35700
+    },
+    {
+      "epoch": 1.1519797918040964,
+      "grad_norm": 4.1735334396362305,
+      "learning_rate": 2.1201370788686168e-05,
+      "loss": 0.10489568710327149,
+      "step": 35800
+    },
+    {
+      "epoch": 1.1551976573938507,
+      "grad_norm": 0.7965012192726135,
+      "learning_rate": 2.1120925443253857e-05,
+      "loss": 0.10613764762878418,
+      "step": 35900
+    },
+    {
+      "epoch": 1.158415522983605,
+      "grad_norm": 11.224800109863281,
+      "learning_rate": 2.104048009782154e-05,
+      "loss": 0.10370283126831055,
+      "step": 36000
+    },
+    {
+      "epoch": 1.1616333885733594,
+      "grad_norm": 2.968989372253418,
+      "learning_rate": 2.0960034752389225e-05,
+      "loss": 0.1149828052520752,
+      "step": 36100
+    },
+    {
+      "epoch": 1.1648512541631135,
+      "grad_norm": 2.068894863128662,
+      "learning_rate": 2.0879589406956914e-05,
+      "loss": 0.10543420791625976,
+      "step": 36200
+    },
+    {
+      "epoch": 1.1680691197528679,
+      "grad_norm": 1.2261930704116821,
+      "learning_rate": 2.07991440615246e-05,
+      "loss": 0.11141412734985351,
+      "step": 36300
+    },
+    {
+      "epoch": 1.1712869853426222,
+      "grad_norm": 3.737278699874878,
+      "learning_rate": 2.071869871609229e-05,
+      "loss": 0.11346253395080566,
+      "step": 36400
+    },
+    {
+      "epoch": 1.1745048509323766,
+      "grad_norm": 2.6356663703918457,
+      "learning_rate": 2.0638253370659974e-05,
+      "loss": 0.12470024108886718,
+      "step": 36500
+    },
+    {
+      "epoch": 1.177722716522131,
+      "grad_norm": 1.931261658668518,
+      "learning_rate": 2.0557808025227664e-05,
+      "loss": 0.12472087860107423,
+      "step": 36600
+    },
+    {
+      "epoch": 1.1809405821118852,
+      "grad_norm": 3.1169793605804443,
+      "learning_rate": 2.047736267979535e-05,
+      "loss": 0.10721843719482421,
+      "step": 36700
+    },
+    {
+      "epoch": 1.1841584477016396,
+      "grad_norm": 3.3439762592315674,
+      "learning_rate": 2.0396917334363035e-05,
+      "loss": 0.115928316116333,
+      "step": 36800
+    },
+    {
+      "epoch": 1.1873763132913937,
+      "grad_norm": 5.898672103881836,
+      "learning_rate": 2.031647198893072e-05,
+      "loss": 0.1268135643005371,
+      "step": 36900
+    },
+    {
+      "epoch": 1.190594178881148,
+      "grad_norm": 1.7046561241149902,
+      "learning_rate": 2.0236026643498406e-05,
+      "loss": 0.11421506881713867,
+      "step": 37000
+    },
+    {
+      "epoch": 1.1938120444709024,
+      "grad_norm": 1.223960041999817,
+      "learning_rate": 2.0155581298066095e-05,
+      "loss": 0.11065722465515136,
+      "step": 37100
+    },
+    {
+      "epoch": 1.1970299100606567,
+      "grad_norm": 5.873408794403076,
+      "learning_rate": 2.007513595263378e-05,
+      "loss": 0.12281797409057617,
+      "step": 37200
+    },
+    {
+      "epoch": 1.200247775650411,
+      "grad_norm": 1.8486683368682861,
+      "learning_rate": 1.999469060720147e-05,
+      "loss": 0.1131003189086914,
+      "step": 37300
+    },
+    {
+      "epoch": 1.2034656412401654,
+      "grad_norm": 3.4886019229888916,
+      "learning_rate": 1.9914245261769155e-05,
+      "loss": 0.11905000686645507,
+      "step": 37400
+    },
+    {
+      "epoch": 1.2066835068299198,
+      "grad_norm": 0.7239132523536682,
+      "learning_rate": 1.983379991633684e-05,
+      "loss": 0.10931044578552246,
+      "step": 37500
+    },
+    {
+      "epoch": 1.2066835068299198,
+      "eval_accuracy": 0.9886145760727707,
+      "eval_f1": 0.9460371209613769,
+      "eval_loss": 0.043151482939720154,
+      "eval_precision": 0.9363463404185619,
+      "eval_recall": 0.9559305900361457,
+      "eval_runtime": 38.5909,
+      "eval_samples_per_second": 647.822,
+      "eval_steps_per_second": 35.993,
+      "step": 37500
+    },
+    {
+      "epoch": 1.209901372419674,
+      "grad_norm": 2.1548306941986084,
+      "learning_rate": 1.9753354570904527e-05,
+      "loss": 0.11063767433166503,
+      "step": 37600
+    },
+    {
+      "epoch": 1.2131192380094284,
+      "grad_norm": 1.2177164554595947,
+      "learning_rate": 1.9672909225472216e-05,
+      "loss": 0.11377819061279297,
+      "step": 37700
+    },
+    {
+      "epoch": 1.2163371035991826,
+      "grad_norm": 0.8512117266654968,
+      "learning_rate": 1.95924638800399e-05,
+      "loss": 0.1323404121398926,
+      "step": 37800
+    },
+    {
+      "epoch": 1.219554969188937,
+      "grad_norm": 1.077217936515808,
+      "learning_rate": 1.9512018534607587e-05,
+      "loss": 0.0988365364074707,
+      "step": 37900
+    },
+    {
+      "epoch": 1.2227728347786913,
+      "grad_norm": 2.1567108631134033,
+      "learning_rate": 1.9431573189175276e-05,
+      "loss": 0.12120766639709472,
+      "step": 38000
+    },
+    {
+      "epoch": 1.2259907003684456,
+      "grad_norm": 3.18619966506958,
+      "learning_rate": 1.9351127843742962e-05,
+      "loss": 0.11067086219787597,
+      "step": 38100
+    },
+    {
+      "epoch": 1.2292085659582,
+      "grad_norm": 1.0777678489685059,
+      "learning_rate": 1.927068249831065e-05,
+      "loss": 0.11158087730407715,
+      "step": 38200
+    },
+    {
+      "epoch": 1.2324264315479543,
+      "grad_norm": 2.5809645652770996,
+      "learning_rate": 1.9190237152878333e-05,
+      "loss": 0.11114447593688964,
+      "step": 38300
+    },
+    {
+      "epoch": 1.2356442971377086,
+      "grad_norm": 1.8847070932388306,
+      "learning_rate": 1.9109791807446022e-05,
+      "loss": 0.11262718200683594,
+      "step": 38400
+    },
+    {
+      "epoch": 1.238862162727463,
+      "grad_norm": 7.437971591949463,
+      "learning_rate": 1.9029346462013708e-05,
+      "loss": 0.11268477439880371,
+      "step": 38500
+    },
+    {
+      "epoch": 1.242080028317217,
+      "grad_norm": 3.669605255126953,
+      "learning_rate": 1.8948901116581397e-05,
+      "loss": 0.11750943183898926,
+      "step": 38600
+    },
+    {
+      "epoch": 1.2452978939069714,
+      "grad_norm": 2.447467803955078,
+      "learning_rate": 1.8868455771149082e-05,
+      "loss": 0.10881879806518555,
+      "step": 38700
+    },
+    {
+      "epoch": 1.2485157594967258,
+      "grad_norm": 4.701931953430176,
+      "learning_rate": 1.8788010425716768e-05,
+      "loss": 0.12167869567871094,
+      "step": 38800
+    },
+    {
+      "epoch": 1.2517336250864801,
+      "grad_norm": 1.6426016092300415,
+      "learning_rate": 1.8707565080284457e-05,
+      "loss": 0.12139597892761231,
+      "step": 38900
+    },
+    {
+      "epoch": 1.2549514906762345,
+      "grad_norm": 4.622367858886719,
+      "learning_rate": 1.8627119734852143e-05,
+      "loss": 0.10588560104370118,
+      "step": 39000
+    },
+    {
+      "epoch": 1.2581693562659888,
+      "grad_norm": 1.7007027864456177,
+      "learning_rate": 1.854667438941983e-05,
+      "loss": 0.11511992454528809,
+      "step": 39100
+    },
+    {
+      "epoch": 1.2613872218557431,
+      "grad_norm": 2.8813023567199707,
+      "learning_rate": 1.8466229043987514e-05,
+      "loss": 0.09704429626464844,
+      "step": 39200
+    },
+    {
+      "epoch": 1.2646050874454975,
+      "grad_norm": 8.06043815612793,
+      "learning_rate": 1.8385783698555203e-05,
+      "loss": 0.11457429885864258,
+      "step": 39300
+    },
+    {
+      "epoch": 1.2678229530352518,
+      "grad_norm": 8.746806144714355,
+      "learning_rate": 1.830533835312289e-05,
+      "loss": 0.12408617973327636,
+      "step": 39400
+    },
+    {
+      "epoch": 1.271040818625006,
+      "grad_norm": 5.757315158843994,
+      "learning_rate": 1.8224893007690578e-05,
+      "loss": 0.10381578445434571,
+      "step": 39500
+    },
+    {
+      "epoch": 1.2742586842147603,
+      "grad_norm": 0.5180655121803284,
+      "learning_rate": 1.8144447662258263e-05,
+      "loss": 0.10882855415344238,
+      "step": 39600
+    },
+    {
+      "epoch": 1.2774765498045146,
+      "grad_norm": 2.3184099197387695,
+      "learning_rate": 1.806400231682595e-05,
+      "loss": 0.10384510040283203,
+      "step": 39700
+    },
+    {
+      "epoch": 1.280694415394269,
+      "grad_norm": 3.118455648422241,
+      "learning_rate": 1.7983556971393635e-05,
+      "loss": 0.1031002426147461,
+      "step": 39800
+    },
+    {
+      "epoch": 1.2839122809840233,
+      "grad_norm": 10.587071418762207,
+      "learning_rate": 1.790311162596132e-05,
+      "loss": 0.11156161308288574,
+      "step": 39900
+    },
+    {
+      "epoch": 1.2871301465737777,
+      "grad_norm": 2.5432746410369873,
+      "learning_rate": 1.782266628052901e-05,
+      "loss": 0.11637428283691406,
+      "step": 40000
+    },
+    {
+      "epoch": 1.2871301465737777,
+      "eval_accuracy": 0.989512214291476,
+      "eval_f1": 0.9486766456772123,
+      "eval_loss": 0.040956247597932816,
+      "eval_precision": 0.9392224099241068,
+      "eval_recall": 0.9583231499466176,
+      "eval_runtime": 38.6339,
+      "eval_samples_per_second": 647.1,
+      "eval_steps_per_second": 35.953,
+      "step": 40000
+    },
+    {
+      "epoch": 1.290348012163532,
+      "grad_norm": 1.766068458557129,
+      "learning_rate": 1.7742220935096695e-05,
+      "loss": 0.11921710968017578,
+      "step": 40100
+    },
+    {
+      "epoch": 1.2935658777532861,
+      "grad_norm": 1.7064458131790161,
+      "learning_rate": 1.7661775589664384e-05,
+      "loss": 0.12581809997558593,
+      "step": 40200
+    },
+    {
+      "epoch": 1.2967837433430405,
+      "grad_norm": 1.2497196197509766,
+      "learning_rate": 1.758133024423207e-05,
+      "loss": 0.10997524261474609,
+      "step": 40300
+    },
+    {
+      "epoch": 1.3000016089327948,
+      "grad_norm": 1.997997760772705,
+      "learning_rate": 1.750088489879976e-05,
+      "loss": 0.11178022384643554,
+      "step": 40400
+    },
+    {
+      "epoch": 1.3032194745225492,
+      "grad_norm": 1.8823689222335815,
+      "learning_rate": 1.7420439553367444e-05,
+      "loss": 0.09935328483581543,
+      "step": 40500
+    },
+    {
+      "epoch": 1.3064373401123035,
+      "grad_norm": 4.478331089019775,
+      "learning_rate": 1.733999420793513e-05,
+      "loss": 0.11027010917663574,
+      "step": 40600
+    },
+    {
+      "epoch": 1.3096552057020578,
+      "grad_norm": 3.932783365249634,
+      "learning_rate": 1.7259548862502816e-05,
+      "loss": 0.11079183578491211,
+      "step": 40700
+    },
+    {
+      "epoch": 1.3128730712918122,
+      "grad_norm": 2.1589443683624268,
+      "learning_rate": 1.71791035170705e-05,
+      "loss": 0.10328941345214844,
+      "step": 40800
+    },
+    {
+      "epoch": 1.3160909368815665,
+      "grad_norm": 2.9713399410247803,
+      "learning_rate": 1.709865817163819e-05,
+      "loss": 0.1030450439453125,
+      "step": 40900
+    },
+    {
+      "epoch": 1.3193088024713209,
+      "grad_norm": 4.732212066650391,
+      "learning_rate": 1.7018212826205876e-05,
+      "loss": 0.114616117477417,
+      "step": 41000
+    },
+    {
+      "epoch": 1.3225266680610752,
+      "grad_norm": 1.4628472328186035,
+      "learning_rate": 1.6937767480773565e-05,
+      "loss": 0.12563533782958985,
+      "step": 41100
+    },
+    {
+      "epoch": 1.3257445336508293,
+      "grad_norm": 3.7251741886138916,
+      "learning_rate": 1.685732213534125e-05,
+      "loss": 0.11473335266113281,
+      "step": 41200
+    },
+    {
+      "epoch": 1.3289623992405837,
+      "grad_norm": 4.524189472198486,
+      "learning_rate": 1.6776876789908936e-05,
+      "loss": 0.12875600814819335,
+      "step": 41300
+    },
+    {
+      "epoch": 1.332180264830338,
+      "grad_norm": 28.197296142578125,
+      "learning_rate": 1.6696431444476622e-05,
+      "loss": 0.11420975685119629,
+      "step": 41400
+    },
+    {
+      "epoch": 1.3353981304200924,
+      "grad_norm": 1.819801926612854,
+      "learning_rate": 1.6615986099044308e-05,
+      "loss": 0.09828336715698242,
+      "step": 41500
+    },
+    {
+      "epoch": 1.3386159960098467,
+      "grad_norm": 1.7372171878814697,
+      "learning_rate": 1.6535540753611997e-05,
+      "loss": 0.1114089298248291,
+      "step": 41600
+    },
+    {
+      "epoch": 1.341833861599601,
+      "grad_norm": 4.730764865875244,
+      "learning_rate": 1.6455095408179682e-05,
+      "loss": 0.11046221733093262,
+      "step": 41700
+    },
+    {
+      "epoch": 1.3450517271893552,
+      "grad_norm": 0.6920406222343445,
+      "learning_rate": 1.637465006274737e-05,
+      "loss": 0.1076749610900879,
+      "step": 41800
+    },
+    {
+      "epoch": 1.3482695927791095,
+      "grad_norm": 1.7678096294403076,
+      "learning_rate": 1.6294204717315057e-05,
+      "loss": 0.10196740150451661,
+      "step": 41900
+    },
+    {
+      "epoch": 1.3514874583688639,
+      "grad_norm": 7.219323635101318,
+      "learning_rate": 1.6213759371882746e-05,
+      "loss": 0.10450594902038574,
+      "step": 42000
+    },
+    {
+      "epoch": 1.3547053239586182,
+      "grad_norm": 3.9280190467834473,
+      "learning_rate": 1.613331402645043e-05,
+      "loss": 0.11356738090515137,
+      "step": 42100
+    },
+    {
+      "epoch": 1.3579231895483725,
+      "grad_norm": 0.7362053394317627,
+      "learning_rate": 1.6052868681018117e-05,
+      "loss": 0.11030459403991699,
+      "step": 42200
+    },
+    {
+      "epoch": 1.3611410551381269,
+      "grad_norm": 2.8355274200439453,
+      "learning_rate": 1.5972423335585803e-05,
+      "loss": 0.10215305328369141,
+      "step": 42300
+    },
+    {
+      "epoch": 1.3643589207278812,
+      "grad_norm": 2.266855001449585,
+      "learning_rate": 1.589197799015349e-05,
+      "loss": 0.10216259002685547,
+      "step": 42400
+    },
+    {
+      "epoch": 1.3675767863176356,
+      "grad_norm": 1.7423440217971802,
+      "learning_rate": 1.5811532644721178e-05,
+      "loss": 0.11463526725769042,
+      "step": 42500
+    },
+    {
+      "epoch": 1.3675767863176356,
+      "eval_accuracy": 0.9885556297842152,
+      "eval_f1": 0.9436256775454764,
+      "eval_loss": 0.043291687965393066,
+      "eval_precision": 0.9313770939571248,
+      "eval_recall": 0.9562007177679731,
+      "eval_runtime": 39.0115,
+      "eval_samples_per_second": 640.836,
+      "eval_steps_per_second": 35.605,
+      "step": 42500
+    },
+    {
+      "epoch": 1.37079465190739,
+      "grad_norm": 2.6859209537506104,
+      "learning_rate": 1.5731087299288863e-05,
+      "loss": 0.10509669303894043,
+      "step": 42600
+    },
+    {
+      "epoch": 1.3740125174971443,
+      "grad_norm": 1.952075481414795,
+      "learning_rate": 1.5650641953856552e-05,
+      "loss": 0.10759903907775879,
+      "step": 42700
+    },
+    {
+      "epoch": 1.3772303830868984,
+      "grad_norm": 2.712860107421875,
+      "learning_rate": 1.5570196608424238e-05,
+      "loss": 0.1039089298248291,
+      "step": 42800
+    },
+    {
+      "epoch": 1.3804482486766527,
+      "grad_norm": 0.966740608215332,
+      "learning_rate": 1.5489751262991924e-05,
+      "loss": 0.11702409744262696,
+      "step": 42900
+    },
+    {
+      "epoch": 1.383666114266407,
+      "grad_norm": 0.4163697063922882,
+      "learning_rate": 1.540930591755961e-05,
+      "loss": 0.10730401039123535,
+      "step": 43000
+    },
+    {
+      "epoch": 1.3868839798561614,
+      "grad_norm": 1.878928780555725,
+      "learning_rate": 1.53288605721273e-05,
+      "loss": 0.10289834022521972,
+      "step": 43100
+    },
+    {
+      "epoch": 1.3901018454459158,
+      "grad_norm": 0.8312145471572876,
+      "learning_rate": 1.5248415226694984e-05,
+      "loss": 0.0965440845489502,
+      "step": 43200
+    },
+    {
+      "epoch": 1.39331971103567,
+      "grad_norm": 2.221240282058716,
+      "learning_rate": 1.516796988126267e-05,
+      "loss": 0.11579631805419922,
+      "step": 43300
+    },
+    {
+      "epoch": 1.3965375766254244,
+      "grad_norm": 1.7558112144470215,
+      "learning_rate": 1.5087524535830357e-05,
+      "loss": 0.09858304977416993,
+      "step": 43400
+    },
+    {
+      "epoch": 1.3997554422151786,
+      "grad_norm": 2.686286449432373,
+      "learning_rate": 1.5007079190398043e-05,
+      "loss": 0.1057161808013916,
+      "step": 43500
+    },
+    {
+      "epoch": 1.402973307804933,
+      "grad_norm": 3.1591944694519043,
+      "learning_rate": 1.4926633844965732e-05,
+      "loss": 0.08528660774230958,
+      "step": 43600
+    },
+    {
+      "epoch": 1.4061911733946872,
+      "grad_norm": 5.5276994705200195,
+      "learning_rate": 1.4846188499533417e-05,
+      "loss": 0.11572813987731934,
+      "step": 43700
+    },
+    {
+      "epoch": 1.4094090389844416,
+      "grad_norm": 1.2573060989379883,
+      "learning_rate": 1.4765743154101105e-05,
+      "loss": 0.11284799575805664,
+      "step": 43800
+    },
+    {
+      "epoch": 1.412626904574196,
+      "grad_norm": 1.1216791868209839,
+      "learning_rate": 1.468529780866879e-05,
+      "loss": 0.11373154640197754,
+      "step": 43900
+    },
+    {
+      "epoch": 1.4158447701639503,
+      "grad_norm": 1.1175236701965332,
+      "learning_rate": 1.460485246323648e-05,
+      "loss": 0.11358193397521972,
+      "step": 44000
+    },
+    {
+      "epoch": 1.4190626357537046,
+      "grad_norm": 1.5515943765640259,
+      "learning_rate": 1.4524407117804165e-05,
+      "loss": 0.0964575481414795,
+      "step": 44100
+    },
+    {
+      "epoch": 1.422280501343459,
+      "grad_norm": 1.3432780504226685,
+      "learning_rate": 1.444396177237185e-05,
+      "loss": 0.12204754829406739,
+      "step": 44200
+    },
+    {
+      "epoch": 1.4254983669332133,
+      "grad_norm": 5.370590686798096,
+      "learning_rate": 1.4363516426939538e-05,
+      "loss": 0.10917285919189453,
+      "step": 44300
+    },
+    {
+      "epoch": 1.4287162325229676,
+      "grad_norm": 3.957749366760254,
+      "learning_rate": 1.4283071081507224e-05,
+      "loss": 0.1064769458770752,
+      "step": 44400
+    },
+    {
+      "epoch": 1.4319340981127218,
+      "grad_norm": 9.636327743530273,
+      "learning_rate": 1.4202625736074913e-05,
+      "loss": 0.11022495269775391,
+      "step": 44500
+    },
+    {
+      "epoch": 1.435151963702476,
+      "grad_norm": 2.991060495376587,
+      "learning_rate": 1.4122180390642597e-05,
+      "loss": 0.10742597579956055,
+      "step": 44600
+    },
+    {
+      "epoch": 1.4383698292922305,
+      "grad_norm": 2.648939847946167,
+      "learning_rate": 1.4041735045210286e-05,
+      "loss": 0.10721823692321777,
+      "step": 44700
+    },
+    {
+      "epoch": 1.4415876948819848,
+      "grad_norm": 0.6414406895637512,
+      "learning_rate": 1.3961289699777971e-05,
+      "loss": 0.09946588516235351,
+      "step": 44800
+    },
+    {
+      "epoch": 1.4448055604717391,
+      "grad_norm": 0.6889192461967468,
+      "learning_rate": 1.3880844354345659e-05,
+      "loss": 0.10531152725219727,
+      "step": 44900
+    },
+    {
+      "epoch": 1.4480234260614935,
+      "grad_norm": 9.125837326049805,
+      "learning_rate": 1.3800399008913344e-05,
+      "loss": 0.09133255004882812,
+      "step": 45000
+    },
+    {
+      "epoch": 1.4480234260614935,
+      "eval_accuracy": 0.9893700192971534,
+      "eval_f1": 0.9493190302072849,
+      "eval_loss": 0.043362099677324295,
+      "eval_precision": 0.9403339180474755,
+      "eval_recall": 0.9584775086505191,
+      "eval_runtime": 39.1132,
+      "eval_samples_per_second": 639.171,
+      "eval_steps_per_second": 35.512,
+      "step": 45000
+    },
+    {
+      "epoch": 1.4512412916512476,
+      "grad_norm": 2.54496431350708,
+      "learning_rate": 1.371995366348103e-05,
+      "loss": 0.11927374839782715,
+      "step": 45100
+    },
+    {
+      "epoch": 1.454459157241002,
+      "grad_norm": 2.2553350925445557,
+      "learning_rate": 1.3639508318048719e-05,
+      "loss": 0.10068492889404297,
+      "step": 45200
+    },
+    {
+      "epoch": 1.4576770228307563,
+      "grad_norm": 0.6272743940353394,
+      "learning_rate": 1.3559062972616405e-05,
+      "loss": 0.09143063545227051,
+      "step": 45300
+    },
+    {
+      "epoch": 1.4608948884205106,
+      "grad_norm": 2.9756674766540527,
+      "learning_rate": 1.3478617627184092e-05,
+      "loss": 0.11779884338378906,
+      "step": 45400
+    },
+    {
+      "epoch": 1.464112754010265,
+      "grad_norm": 1.4076873064041138,
+      "learning_rate": 1.3398172281751778e-05,
+      "loss": 0.09415918350219726,
+      "step": 45500
+    },
+    {
+      "epoch": 1.4673306196000193,
+      "grad_norm": 2.004279613494873,
+      "learning_rate": 1.3317726936319467e-05,
+      "loss": 0.10082883834838867,
+      "step": 45600
+    },
+    {
+      "epoch": 1.4705484851897737,
+      "grad_norm": 12.558138847351074,
+      "learning_rate": 1.3237281590887152e-05,
+      "loss": 0.10111323356628418,
+      "step": 45700
+    },
+    {
+      "epoch": 1.473766350779528,
+      "grad_norm": 3.433985948562622,
+      "learning_rate": 1.315683624545484e-05,
+      "loss": 0.10751867294311523,
+      "step": 45800
+    },
+    {
+      "epoch": 1.4769842163692823,
+      "grad_norm": 1.3402138948440552,
+      "learning_rate": 1.3076390900022525e-05,
+      "loss": 0.09543782234191894,
+      "step": 45900
+    },
+    {
+      "epoch": 1.4802020819590367,
+      "grad_norm": 2.1528687477111816,
+      "learning_rate": 1.2995945554590211e-05,
+      "loss": 0.10040821075439453,
+      "step": 46000
+    },
+    {
+      "epoch": 1.4834199475487908,
+      "grad_norm": 2.317426919937134,
+      "learning_rate": 1.2915500209157898e-05,
+      "loss": 0.10733115196228027,
+      "step": 46100
+    },
+    {
+      "epoch": 1.4866378131385451,
+      "grad_norm": 2.752028703689575,
+      "learning_rate": 1.2835054863725584e-05,
+      "loss": 0.09164957046508788,
+      "step": 46200
+    },
+    {
+      "epoch": 1.4898556787282995,
+      "grad_norm": 1.7984188795089722,
+      "learning_rate": 1.2754609518293273e-05,
+      "loss": 0.10080486297607422,
+      "step": 46300
+    },
+    {
+      "epoch": 1.4930735443180538,
+      "grad_norm": 2.4552342891693115,
+      "learning_rate": 1.2674164172860959e-05,
+      "loss": 0.10242198944091797,
+      "step": 46400
+    },
+    {
+      "epoch": 1.4962914099078082,
+      "grad_norm": 2.6899032592773438,
+      "learning_rate": 1.2593718827428646e-05,
+      "loss": 0.10129733085632324,
+      "step": 46500
+    },
+    {
+      "epoch": 1.4995092754975625,
+      "grad_norm": 1.909561038017273,
+      "learning_rate": 1.2513273481996332e-05,
+      "loss": 0.1006275749206543,
+      "step": 46600
+    },
+    {
+      "epoch": 1.5027271410873166,
+      "grad_norm": 19.42228889465332,
+      "learning_rate": 1.2432828136564019e-05,
+      "loss": 0.10171239852905273,
+      "step": 46700
+    },
+    {
+      "epoch": 1.505945006677071,
+      "grad_norm": 1.729035496711731,
+      "learning_rate": 1.2352382791131706e-05,
+      "loss": 0.11470888137817382,
+      "step": 46800
+    },
+    {
+      "epoch": 1.5091628722668253,
+      "grad_norm": 10.81437873840332,
+      "learning_rate": 1.2271937445699392e-05,
+      "loss": 0.09871227264404298,
+      "step": 46900
+    },
+    {
+      "epoch": 1.5123807378565797,
+      "grad_norm": 3.496492624282837,
+      "learning_rate": 1.219149210026708e-05,
+      "loss": 0.10564603805541992,
+      "step": 47000
+    },
+    {
+      "epoch": 1.515598603446334,
+      "grad_norm": 0.7296909093856812,
+      "learning_rate": 1.2111046754834767e-05,
+      "loss": 0.09953035354614258,
+      "step": 47100
+    },
+    {
+      "epoch": 1.5188164690360884,
+      "grad_norm": 0.6311995983123779,
+      "learning_rate": 1.2030601409402452e-05,
+      "loss": 0.10630006790161133,
+      "step": 47200
+    },
+    {
+      "epoch": 1.5220343346258427,
+      "grad_norm": 2.436532497406006,
+      "learning_rate": 1.1950156063970138e-05,
+      "loss": 0.09258381843566894,
+      "step": 47300
+    },
+    {
+      "epoch": 1.525252200215597,
+      "grad_norm": 0.557080864906311,
+      "learning_rate": 1.1869710718537825e-05,
+      "loss": 0.1228843879699707,
+      "step": 47400
+    },
+    {
+      "epoch": 1.5284700658053514,
+      "grad_norm": 7.565468788146973,
+      "learning_rate": 1.1789265373105513e-05,
+      "loss": 0.1068018627166748,
+      "step": 47500
+    },
+    {
+      "epoch": 1.5284700658053514,
+      "eval_accuracy": 0.989717492156007,
+      "eval_f1": 0.9515031847133757,
+      "eval_loss": 0.03974379226565361,
+      "eval_precision": 0.942391400345702,
+      "eval_recall": 0.9607928892090403,
+      "eval_runtime": 38.7298,
+      "eval_samples_per_second": 645.498,
+      "eval_steps_per_second": 35.864,
+      "step": 47500
+    },
+    {
+      "epoch": 1.5316879313951057,
+      "grad_norm": 2.948582410812378,
+      "learning_rate": 1.17088200276732e-05,
+      "loss": 0.10712558746337891,
+      "step": 47600
+    },
+    {
+      "epoch": 1.53490579698486,
+      "grad_norm": 6.355142593383789,
+      "learning_rate": 1.1628374682240886e-05,
+      "loss": 0.10939033508300781,
+      "step": 47700
+    },
+    {
+      "epoch": 1.5381236625746144,
+      "grad_norm": 0.6847637295722961,
+      "learning_rate": 1.1547929336808573e-05,
+      "loss": 0.09593602180480958,
+      "step": 47800
+    },
+    {
+      "epoch": 1.5413415281643685,
+      "grad_norm": 2.1787171363830566,
+      "learning_rate": 1.146748399137626e-05,
+      "loss": 0.09451790809631348,
+      "step": 47900
+    },
+    {
+      "epoch": 1.5445593937541229,
+      "grad_norm": 4.488222122192383,
+      "learning_rate": 1.1387038645943946e-05,
+      "loss": 0.1111738109588623,
+      "step": 48000
+    },
+    {
+      "epoch": 1.5477772593438772,
+      "grad_norm": 5.803431034088135,
+      "learning_rate": 1.1306593300511632e-05,
+      "loss": 0.09694368362426758,
+      "step": 48100
+    },
+    {
+      "epoch": 1.5509951249336316,
+      "grad_norm": 1.8297266960144043,
+      "learning_rate": 1.1226147955079319e-05,
+      "loss": 0.105281982421875,
+      "step": 48200
+    },
+    {
+      "epoch": 1.5542129905233857,
+      "grad_norm": 3.511244058609009,
+      "learning_rate": 1.1145702609647006e-05,
+      "loss": 0.10829096794128418,
+      "step": 48300
+    },
+    {
+      "epoch": 1.55743085611314,
+      "grad_norm": 7.947854995727539,
+      "learning_rate": 1.1065257264214692e-05,
+      "loss": 0.11103793144226075,
+      "step": 48400
+    },
+    {
+      "epoch": 1.5606487217028944,
+      "grad_norm": 3.7324254512786865,
+      "learning_rate": 1.098481191878238e-05,
+      "loss": 0.09467618942260742,
+      "step": 48500
+    },
+    {
+      "epoch": 1.5638665872926487,
+      "grad_norm": 8.261491775512695,
+      "learning_rate": 1.0904366573350067e-05,
+      "loss": 0.10849467277526856,
+      "step": 48600
+    },
+    {
+      "epoch": 1.567084452882403,
+      "grad_norm": 1.8083908557891846,
+      "learning_rate": 1.0823921227917754e-05,
+      "loss": 0.09797250747680664,
+      "step": 48700
+    },
+    {
+      "epoch": 1.5703023184721574,
+      "grad_norm": 0.7597082257270813,
+      "learning_rate": 1.074347588248544e-05,
+      "loss": 0.08994776725769044,
+      "step": 48800
+    },
+    {
+      "epoch": 1.5735201840619117,
+      "grad_norm": 1.0194493532180786,
+      "learning_rate": 1.0663030537053127e-05,
+      "loss": 0.1160097885131836,
+      "step": 48900
+    },
+    {
+      "epoch": 1.576738049651666,
+      "grad_norm": 3.0627150535583496,
+      "learning_rate": 1.0582585191620813e-05,
+      "loss": 0.08039675712585449,
+      "step": 49000
+    },
+    {
+      "epoch": 1.5799559152414204,
+      "grad_norm": 11.51146411895752,
+      "learning_rate": 1.05021398461885e-05,
+      "loss": 0.10077597618103028,
+      "step": 49100
+    },
+    {
+      "epoch": 1.5831737808311748,
+      "grad_norm": 3.410426616668701,
+      "learning_rate": 1.0421694500756186e-05,
+      "loss": 0.12183525085449219,
+      "step": 49200
+    },
+    {
+      "epoch": 1.5863916464209291,
+      "grad_norm": 4.240739822387695,
+      "learning_rate": 1.0341249155323873e-05,
+      "loss": 0.10310389518737793,
+      "step": 49300
+    },
+    {
+      "epoch": 1.5896095120106835,
+      "grad_norm": 4.760958194732666,
+      "learning_rate": 1.026080380989156e-05,
+      "loss": 0.10938605308532715,
+      "step": 49400
+    },
+    {
+      "epoch": 1.5928273776004376,
+      "grad_norm": 1.9922808408737183,
+      "learning_rate": 1.0180358464459248e-05,
+      "loss": 0.08123787879943847,
+      "step": 49500
+    },
+    {
+      "epoch": 1.596045243190192,
+      "grad_norm": 1.350643277168274,
+      "learning_rate": 1.0099913119026933e-05,
+      "loss": 0.12260754585266113,
+      "step": 49600
+    },
+    {
+      "epoch": 1.5992631087799463,
+      "grad_norm": 1.4271605014801025,
+      "learning_rate": 1.001946777359462e-05,
+      "loss": 0.09629259109497071,
+      "step": 49700
+    },
+    {
+      "epoch": 1.6024809743697006,
+      "grad_norm": 0.5193639397621155,
+      "learning_rate": 9.939022428162308e-06,
+      "loss": 0.10984975814819337,
+      "step": 49800
+    },
+    {
+      "epoch": 1.605698839959455,
+      "grad_norm": 9.31329345703125,
+      "learning_rate": 9.858577082729994e-06,
+      "loss": 0.09676812171936035,
+      "step": 49900
+    },
+    {
+      "epoch": 1.608916705549209,
+      "grad_norm": 1.29440176486969,
+      "learning_rate": 9.77813173729768e-06,
+      "loss": 0.09296704292297363,
+      "step": 50000
+    },
+    {
+      "epoch": 1.608916705549209,
+      "eval_accuracy": 0.9898457261872506,
+      "eval_f1": 0.9511017704750988,
+      "eval_loss": 0.038785140961408615,
+      "eval_precision": 0.9418761588819233,
+      "eval_recall": 0.9605098982518877,
+      "eval_runtime": 39.1145,
+      "eval_samples_per_second": 639.149,
+      "eval_steps_per_second": 35.511,
+      "step": 50000
+    },
+    {
+      "epoch": 1.6121345711389634,
+      "grad_norm": 2.5150434970855713,
+      "learning_rate": 9.697686391865367e-06,
+      "loss": 0.0893746280670166,
+      "step": 50100
+    },
+    {
+      "epoch": 1.6153524367287178,
+      "grad_norm": 3.5938878059387207,
+      "learning_rate": 9.617241046433054e-06,
+      "loss": 0.10516523361206055,
+      "step": 50200
+    },
+    {
+      "epoch": 1.618570302318472,
+      "grad_norm": 3.323432445526123,
+      "learning_rate": 9.53679570100074e-06,
+      "loss": 0.10093666076660156,
+      "step": 50300
+    },
+    {
+      "epoch": 1.6217881679082264,
+      "grad_norm": 3.201542854309082,
+      "learning_rate": 9.456350355568427e-06,
+      "loss": 0.09994193077087403,
+      "step": 50400
+    },
+    {
+      "epoch": 1.6250060334979808,
+      "grad_norm": 1.2016533613204956,
+      "learning_rate": 9.375905010136114e-06,
+      "loss": 0.09140503883361817,
+      "step": 50500
+    },
+    {
+      "epoch": 1.6282238990877351,
+      "grad_norm": 2.176725149154663,
+      "learning_rate": 9.295459664703802e-06,
+      "loss": 0.10795197486877442,
+      "step": 50600
+    },
+    {
+      "epoch": 1.6314417646774895,
+      "grad_norm": 2.084689140319824,
+      "learning_rate": 9.215014319271487e-06,
+      "loss": 0.10725319862365723,
+      "step": 50700
+    },
+    {
+      "epoch": 1.6346596302672438,
+      "grad_norm": 3.9410250186920166,
+      "learning_rate": 9.134568973839173e-06,
+      "loss": 0.1040007209777832,
+      "step": 50800
+    },
+    {
+      "epoch": 1.6378774958569982,
+      "grad_norm": 7.110863208770752,
+      "learning_rate": 9.05412362840686e-06,
+      "loss": 0.09510645866394044,
+      "step": 50900
+    },
+    {
+      "epoch": 1.6410953614467525,
+      "grad_norm": 2.1342461109161377,
+      "learning_rate": 8.973678282974548e-06,
+      "loss": 0.10462862014770508,
+      "step": 51000
+    },
+    {
+      "epoch": 1.6443132270365068,
+      "grad_norm": 2.3921563625335693,
+      "learning_rate": 8.893232937542233e-06,
+      "loss": 0.09732514381408691,
+      "step": 51100
+    },
+    {
+      "epoch": 1.647531092626261,
+      "grad_norm": 2.4982810020446777,
+      "learning_rate": 8.81278759210992e-06,
+      "loss": 0.08822668075561524,
+      "step": 51200
+    },
+    {
+      "epoch": 1.6507489582160153,
+      "grad_norm": 1.7426912784576416,
+      "learning_rate": 8.732342246677608e-06,
+      "loss": 0.08898674011230469,
+      "step": 51300
+    },
+    {
+      "epoch": 1.6539668238057696,
+      "grad_norm": 9.542119026184082,
+      "learning_rate": 8.651896901245295e-06,
+      "loss": 0.12028385162353515,
+      "step": 51400
+    },
+    {
+      "epoch": 1.657184689395524,
+      "grad_norm": 0.7342370748519897,
+      "learning_rate": 8.571451555812981e-06,
+      "loss": 0.10786977767944336,
+      "step": 51500
+    },
+    {
+      "epoch": 1.6604025549852781,
+      "grad_norm": 0.8785775303840637,
+      "learning_rate": 8.491006210380668e-06,
+      "loss": 0.09443745613098145,
+      "step": 51600
+    },
+    {
+      "epoch": 1.6636204205750325,
+      "grad_norm": 9.54134464263916,
+      "learning_rate": 8.410560864948354e-06,
+      "loss": 0.11094076156616212,
+      "step": 51700
+    },
+    {
+      "epoch": 1.6668382861647868,
+      "grad_norm": 23.422744750976562,
+      "learning_rate": 8.330115519516041e-06,
+      "loss": 0.10699220657348633,
+      "step": 51800
+    },
+    {
+      "epoch": 1.6700561517545411,
+      "grad_norm": 1.0714454650878906,
+      "learning_rate": 8.249670174083727e-06,
+      "loss": 0.10455569267272949,
+      "step": 51900
+    },
+    {
+      "epoch": 1.6732740173442955,
+      "grad_norm": 2.6620521545410156,
+      "learning_rate": 8.169224828651414e-06,
+      "loss": 0.09597659111022949,
+      "step": 52000
+    },
+    {
+      "epoch": 1.6764918829340498,
+      "grad_norm": 1.480083703994751,
+      "learning_rate": 8.088779483219102e-06,
+      "loss": 0.10422684669494629,
+      "step": 52100
+    },
+    {
+      "epoch": 1.6797097485238042,
+      "grad_norm": 2.0191445350646973,
+      "learning_rate": 8.008334137786787e-06,
+      "loss": 0.10290263175964355,
+      "step": 52200
+    },
+    {
+      "epoch": 1.6829276141135585,
+      "grad_norm": 1.1020231246948242,
+      "learning_rate": 7.927888792354475e-06,
+      "loss": 0.10773059844970703,
+      "step": 52300
+    },
+    {
+      "epoch": 1.6861454797033129,
+      "grad_norm": 2.2319722175598145,
+      "learning_rate": 7.847443446922162e-06,
+      "loss": 0.10715092658996582,
+      "step": 52400
+    },
+    {
+      "epoch": 1.6893633452930672,
+      "grad_norm": 3.737014055252075,
+      "learning_rate": 7.76699810148985e-06,
+      "loss": 0.08955263137817383,
+      "step": 52500
+    },
+    {
+      "epoch": 1.6893633452930672,
+      "eval_accuracy": 0.9901202917944698,
+      "eval_f1": 0.9537408288021826,
+      "eval_loss": 0.03707578778266907,
+      "eval_precision": 0.945334613829357,
+      "eval_recall": 0.962297886572079,
+      "eval_runtime": 38.6863,
+      "eval_samples_per_second": 646.224,
+      "eval_steps_per_second": 35.904,
+      "step": 52500
+    },
+    {
+      "epoch": 1.6925812108828215,
+      "grad_norm": 1.981602668762207,
+      "learning_rate": 7.686552756057535e-06,
+      "loss": 0.09512082099914551,
+      "step": 52600
+    },
+    {
+      "epoch": 1.6957990764725759,
+      "grad_norm": 0.9314746856689453,
+      "learning_rate": 7.606107410625221e-06,
+      "loss": 0.09555187225341796,
+      "step": 52700
+    },
+    {
+      "epoch": 1.69901694206233,
+      "grad_norm": 1.3338748216629028,
+      "learning_rate": 7.525662065192908e-06,
+      "loss": 0.08804367065429687,
+      "step": 52800
+    },
+    {
+      "epoch": 1.7022348076520843,
+      "grad_norm": 2.233142375946045,
+      "learning_rate": 7.445216719760595e-06,
+      "loss": 0.099904146194458,
+      "step": 52900
+    },
+    {
+      "epoch": 1.7054526732418387,
+      "grad_norm": 2.9169979095458984,
+      "learning_rate": 7.364771374328282e-06,
+      "loss": 0.09450019836425781,
+      "step": 53000
+    },
+    {
+      "epoch": 1.708670538831593,
+      "grad_norm": 1.4995619058609009,
+      "learning_rate": 7.284326028895968e-06,
+      "loss": 0.10291241645812989,
+      "step": 53100
+    },
+    {
+      "epoch": 1.7118884044213472,
+      "grad_norm": 0.7775556445121765,
+      "learning_rate": 7.203880683463656e-06,
+      "loss": 0.1126316738128662,
+      "step": 53200
+    },
+    {
+      "epoch": 1.7151062700111015,
+      "grad_norm": 4.055330753326416,
+      "learning_rate": 7.123435338031342e-06,
+      "loss": 0.10327481269836426,
+      "step": 53300
+    },
+    {
+      "epoch": 1.7183241356008558,
+      "grad_norm": 1.895665168762207,
+      "learning_rate": 7.04298999259903e-06,
+      "loss": 0.09612360954284668,
+      "step": 53400
+    },
+    {
+      "epoch": 1.7215420011906102,
+      "grad_norm": 2.202033758163452,
+      "learning_rate": 6.962544647166714e-06,
+      "loss": 0.09727460861206055,
+      "step": 53500
+    },
+    {
+      "epoch": 1.7247598667803645,
+      "grad_norm": 1.6725600957870483,
+      "learning_rate": 6.882099301734402e-06,
+      "loss": 0.1051039981842041,
+      "step": 53600
+    },
+    {
+      "epoch": 1.7279777323701189,
+      "grad_norm": 1.6951491832733154,
+      "learning_rate": 6.801653956302088e-06,
+      "loss": 0.09583574295043945,
+      "step": 53700
+    },
+    {
+      "epoch": 1.7311955979598732,
+      "grad_norm": 2.0571811199188232,
+      "learning_rate": 6.721208610869776e-06,
+      "loss": 0.09057200431823731,
+      "step": 53800
+    },
+    {
+      "epoch": 1.7344134635496276,
+      "grad_norm": 1.9436057806015015,
+      "learning_rate": 6.640763265437462e-06,
+      "loss": 0.10000760078430176,
+      "step": 53900
+    },
+    {
+      "epoch": 1.737631329139382,
+      "grad_norm": 0.3372837007045746,
+      "learning_rate": 6.560317920005149e-06,
+      "loss": 0.10574793815612793,
+      "step": 54000
+    },
+    {
+      "epoch": 1.7408491947291362,
+      "grad_norm": 2.41835880279541,
+      "learning_rate": 6.479872574572836e-06,
+      "loss": 0.10087454795837403,
+      "step": 54100
+    },
+    {
+      "epoch": 1.7440670603188906,
+      "grad_norm": 0.6734520792961121,
+      "learning_rate": 6.399427229140522e-06,
+      "loss": 0.08962540626525879,
+      "step": 54200
+    },
+    {
+      "epoch": 1.747284925908645,
+      "grad_norm": 2.5488386154174805,
+      "learning_rate": 6.31898188370821e-06,
+      "loss": 0.1008315372467041,
+      "step": 54300
+    },
+    {
+      "epoch": 1.7505027914983993,
+      "grad_norm": 3.4320902824401855,
+      "learning_rate": 6.238536538275895e-06,
+      "loss": 0.08948537826538086,
+      "step": 54400
+    },
+    {
+      "epoch": 1.7537206570881534,
+      "grad_norm": 3.819451093673706,
+      "learning_rate": 6.158091192843583e-06,
+      "loss": 0.08715433120727539,
+      "step": 54500
+    },
+    {
+      "epoch": 1.7569385226779077,
+      "grad_norm": 1.9689542055130005,
+      "learning_rate": 6.077645847411268e-06,
+      "loss": 0.0976317596435547,
+      "step": 54600
+    },
+    {
+      "epoch": 1.760156388267662,
+      "grad_norm": 2.7664942741394043,
+      "learning_rate": 5.997200501978956e-06,
+      "loss": 0.11424756050109863,
+      "step": 54700
+    },
+    {
+      "epoch": 1.7633742538574164,
+      "grad_norm": 1.188758134841919,
+      "learning_rate": 5.916755156546642e-06,
+      "loss": 0.09281901359558105,
+      "step": 54800
+    },
+    {
+      "epoch": 1.7665921194471705,
+      "grad_norm": 1.442600965499878,
+      "learning_rate": 5.83630981111433e-06,
+      "loss": 0.09055022239685058,
+      "step": 54900
+    },
+    {
+      "epoch": 1.7698099850369249,
+      "grad_norm": 2.042584180831909,
+      "learning_rate": 5.755864465682016e-06,
+      "loss": 0.08615251541137696,
+      "step": 55000
+    },
+    {
+      "epoch": 1.7698099850369249,
+      "eval_accuracy": 0.9903457354945594,
+      "eval_f1": 0.9534046525260119,
+      "eval_loss": 0.036240287125110626,
+      "eval_precision": 0.9445875312476327,
+      "eval_recall": 0.9623879291493549,
+      "eval_runtime": 38.9533,
+      "eval_samples_per_second": 641.795,
+      "eval_steps_per_second": 35.658,
+      "step": 55000
+    },
+    {
+      "epoch": 1.7730278506266792,
+      "grad_norm": 1.0843491554260254,
+      "learning_rate": 5.675419120249703e-06,
+      "loss": 0.08899230003356934,
+      "step": 55100
+    },
+    {
+      "epoch": 1.7762457162164336,
+      "grad_norm": 12.144911766052246,
+      "learning_rate": 5.594973774817389e-06,
+      "loss": 0.08881766319274903,
+      "step": 55200
+    },
+    {
+      "epoch": 1.779463581806188,
+      "grad_norm": 2.8250832557678223,
+      "learning_rate": 5.514528429385076e-06,
+      "loss": 0.0975367259979248,
+      "step": 55300
+    },
+    {
+      "epoch": 1.7826814473959423,
+      "grad_norm": 1.2422140836715698,
+      "learning_rate": 5.434083083952763e-06,
+      "loss": 0.10660443305969239,
+      "step": 55400
+    },
+    {
+      "epoch": 1.7858993129856966,
+      "grad_norm": 4.103642463684082,
+      "learning_rate": 5.353637738520449e-06,
+      "loss": 0.10672403335571289,
+      "step": 55500
+    },
+    {
+      "epoch": 1.789117178575451,
+      "grad_norm": 2.0695173740386963,
+      "learning_rate": 5.273192393088136e-06,
+      "loss": 0.08749826431274414,
+      "step": 55600
+    },
+    {
+      "epoch": 1.7923350441652053,
+      "grad_norm": 2.9659390449523926,
+      "learning_rate": 5.192747047655823e-06,
+      "loss": 0.10261861801147461,
+      "step": 55700
+    },
+    {
+      "epoch": 1.7955529097549596,
+      "grad_norm": 3.0467967987060547,
+      "learning_rate": 5.11230170222351e-06,
+      "loss": 0.1026226806640625,
+      "step": 55800
+    },
+    {
+      "epoch": 1.798770775344714,
+      "grad_norm": 2.826698064804077,
+      "learning_rate": 5.031856356791196e-06,
+      "loss": 0.08784229278564454,
+      "step": 55900
+    },
+    {
+      "epoch": 1.8019886409344683,
+      "grad_norm": 4.862614631652832,
+      "learning_rate": 4.951411011358883e-06,
+      "loss": 0.09360240936279297,
+      "step": 56000
+    },
+    {
+      "epoch": 1.8052065065242224,
+      "grad_norm": 1.5207270383834839,
+      "learning_rate": 4.87096566592657e-06,
+      "loss": 0.10859518051147461,
+      "step": 56100
+    },
+    {
+      "epoch": 1.8084243721139768,
+      "grad_norm": 2.692488670349121,
+      "learning_rate": 4.790520320494257e-06,
+      "loss": 0.0937327766418457,
+      "step": 56200
+    },
+    {
+      "epoch": 1.8116422377037311,
+      "grad_norm": 2.3452610969543457,
+      "learning_rate": 4.710074975061943e-06,
+      "loss": 0.0986541748046875,
+      "step": 56300
+    },
+    {
+      "epoch": 1.8148601032934855,
+      "grad_norm": 3.4485702514648438,
+      "learning_rate": 4.6296296296296296e-06,
+      "loss": 0.11264933586120605,
+      "step": 56400
+    },
+    {
+      "epoch": 1.8180779688832396,
+      "grad_norm": 0.7109314203262329,
+      "learning_rate": 4.549184284197316e-06,
+      "loss": 0.09030750274658203,
+      "step": 56500
+    },
+    {
+      "epoch": 1.821295834472994,
+      "grad_norm": 11.383523941040039,
+      "learning_rate": 4.468738938765003e-06,
+      "loss": 0.09089506149291993,
+      "step": 56600
+    },
+    {
+      "epoch": 1.8245137000627483,
+      "grad_norm": 2.5693790912628174,
+      "learning_rate": 4.38829359333269e-06,
+      "loss": 0.09092802047729492,
+      "step": 56700
+    },
+    {
+      "epoch": 1.8277315656525026,
+      "grad_norm": 4.628587245941162,
+      "learning_rate": 4.307848247900376e-06,
+      "loss": 0.10240273475646973,
+      "step": 56800
+    },
+    {
+      "epoch": 1.830949431242257,
+      "grad_norm": 2.5633656978607178,
+      "learning_rate": 4.227402902468063e-06,
+      "loss": 0.10268410682678222,
+      "step": 56900
+    },
+    {
+      "epoch": 1.8341672968320113,
+      "grad_norm": 1.9654608964920044,
+      "learning_rate": 4.14695755703575e-06,
+      "loss": 0.08537924766540528,
+      "step": 57000
+    },
+    {
+      "epoch": 1.8373851624217656,
+      "grad_norm": 5.80186653137207,
+      "learning_rate": 4.066512211603437e-06,
+      "loss": 0.09202526092529296,
+      "step": 57100
+    },
+    {
+      "epoch": 1.84060302801152,
+      "grad_norm": 1.1093907356262207,
+      "learning_rate": 3.986066866171124e-06,
+      "loss": 0.08391679763793945,
+      "step": 57200
+    },
+    {
+      "epoch": 1.8438208936012743,
+      "grad_norm": 2.4156336784362793,
+      "learning_rate": 3.90562152073881e-06,
+      "loss": 0.08964170455932617,
+      "step": 57300
+    },
+    {
+      "epoch": 1.8470387591910287,
+      "grad_norm": 0.8483083844184875,
+      "learning_rate": 3.825176175306497e-06,
+      "loss": 0.08196611404418945,
+      "step": 57400
+    },
+    {
+      "epoch": 1.850256624780783,
+      "grad_norm": 2.47489070892334,
+      "learning_rate": 3.7447308298741836e-06,
+      "loss": 0.09712170600891114,
+      "step": 57500
+    },
+    {
+      "epoch": 1.850256624780783,
+      "eval_accuracy": 0.990784730222486,
+      "eval_f1": 0.9556170744775851,
+      "eval_loss": 0.03545914590358734,
+      "eval_precision": 0.9478179448570814,
+      "eval_recall": 0.9635456194286155,
+      "eval_runtime": 39.1151,
+      "eval_samples_per_second": 639.14,
+      "eval_steps_per_second": 35.511,
+      "step": 57500
+    },
+    {
+      "epoch": 1.8534744903705374,
+      "grad_norm": 2.640423536300659,
+      "learning_rate": 3.6642854844418705e-06,
+      "loss": 0.09879606246948242,
+      "step": 57600
+    },
+    {
+      "epoch": 1.8566923559602917,
+      "grad_norm": 1.9036402702331543,
+      "learning_rate": 3.5838401390095566e-06,
+      "loss": 0.08787569999694825,
+      "step": 57700
+    },
+    {
+      "epoch": 1.8599102215500458,
+      "grad_norm": 2.5272982120513916,
+      "learning_rate": 3.5033947935772435e-06,
+      "loss": 0.0921150016784668,
+      "step": 57800
+    },
+    {
+      "epoch": 1.8631280871398002,
+      "grad_norm": 2.6478683948516846,
+      "learning_rate": 3.4229494481449304e-06,
+      "loss": 0.10192151069641113,
+      "step": 57900
+    },
+    {
+      "epoch": 1.8663459527295545,
+      "grad_norm": 1.6722396612167358,
+      "learning_rate": 3.3425041027126173e-06,
+      "loss": 0.08221748352050781,
+      "step": 58000
+    },
+    {
+      "epoch": 1.8695638183193088,
+      "grad_norm": 0.9039199948310852,
+      "learning_rate": 3.2620587572803042e-06,
+      "loss": 0.09699295043945312,
+      "step": 58100
+    },
+    {
+      "epoch": 1.872781683909063,
+      "grad_norm": 4.108741283416748,
+      "learning_rate": 3.1816134118479903e-06,
+      "loss": 0.10586893081665039,
+      "step": 58200
+    },
+    {
+      "epoch": 1.8759995494988173,
+      "grad_norm": 0.8022117018699646,
+      "learning_rate": 3.1011680664156772e-06,
+      "loss": 0.10414183616638184,
+      "step": 58300
+    },
+    {
+      "epoch": 1.8792174150885717,
+      "grad_norm": 0.757673978805542,
+      "learning_rate": 3.020722720983364e-06,
+      "loss": 0.09568047523498535,
+      "step": 58400
+    },
+    {
+      "epoch": 1.882435280678326,
+      "grad_norm": 2.535696506500244,
+      "learning_rate": 2.9402773755510506e-06,
+      "loss": 0.10529354095458984,
+      "step": 58500
+    },
+    {
+      "epoch": 1.8856531462680803,
+      "grad_norm": 3.3938286304473877,
+      "learning_rate": 2.8598320301187376e-06,
+      "loss": 0.09156073570251465,
+      "step": 58600
+    },
+    {
+      "epoch": 1.8888710118578347,
+      "grad_norm": 1.427871823310852,
+      "learning_rate": 2.779386684686424e-06,
+      "loss": 0.08934795379638671,
+      "step": 58700
+    },
+    {
+      "epoch": 1.892088877447589,
+      "grad_norm": 1.7206834554672241,
+      "learning_rate": 2.698941339254111e-06,
+      "loss": 0.10297789573669433,
+      "step": 58800
+    },
+    {
+      "epoch": 1.8953067430373434,
+      "grad_norm": 2.4081387519836426,
+      "learning_rate": 2.618495993821798e-06,
+      "loss": 0.10169391632080078,
+      "step": 58900
+    },
+    {
+      "epoch": 1.8985246086270977,
+      "grad_norm": 4.349306583404541,
+      "learning_rate": 2.5380506483894844e-06,
+      "loss": 0.09130579948425294,
+      "step": 59000
+    },
+    {
+      "epoch": 1.901742474216852,
+      "grad_norm": 1.038880467414856,
+      "learning_rate": 2.4576053029571713e-06,
+      "loss": 0.09172413825988769,
+      "step": 59100
+    },
+    {
+      "epoch": 1.9049603398066064,
+      "grad_norm": 1.437605619430542,
+      "learning_rate": 2.377159957524858e-06,
+      "loss": 0.10029353141784668,
+      "step": 59200
+    },
+    {
+      "epoch": 1.9081782053963607,
+      "grad_norm": 1.8541232347488403,
+      "learning_rate": 2.2967146120925443e-06,
+      "loss": 0.09666367530822755,
+      "step": 59300
+    },
+    {
+      "epoch": 1.9113960709861149,
+      "grad_norm": 4.376369953155518,
+      "learning_rate": 2.2162692666602312e-06,
+      "loss": 0.08466120719909669,
+      "step": 59400
+    },
+    {
+      "epoch": 1.9146139365758692,
+      "grad_norm": 1.9385297298431396,
+      "learning_rate": 2.1358239212279177e-06,
+      "loss": 0.08599072456359863,
+      "step": 59500
+    },
+    {
+      "epoch": 1.9178318021656235,
+      "grad_norm": 1.689815640449524,
+      "learning_rate": 2.0553785757956042e-06,
+      "loss": 0.09492490768432617,
+      "step": 59600
+    },
+    {
+      "epoch": 1.9210496677553779,
+      "grad_norm": 1.3797245025634766,
+      "learning_rate": 1.974933230363291e-06,
+      "loss": 0.09027796745300293,
+      "step": 59700
+    },
+    {
+      "epoch": 1.924267533345132,
+      "grad_norm": 0.4921301007270813,
+      "learning_rate": 1.894487884930978e-06,
+      "loss": 0.0897593879699707,
+      "step": 59800
+    },
+    {
+      "epoch": 1.9274853989348864,
+      "grad_norm": 1.489128828048706,
+      "learning_rate": 1.8140425394986648e-06,
+      "loss": 0.09694389343261718,
+      "step": 59900
+    },
+    {
+      "epoch": 1.9307032645246407,
+      "grad_norm": 0.7353283166885376,
+      "learning_rate": 1.7335971940663515e-06,
+      "loss": 0.0947257137298584,
+      "step": 60000
+    },
+    {
+      "epoch": 1.9307032645246407,
+      "eval_accuracy": 0.9908240277481897,
+      "eval_f1": 0.9576732752290649,
+      "eval_loss": 0.03446565568447113,
+      "eval_precision": 0.9507949387630905,
+      "eval_recall": 0.9646518568065757,
+      "eval_runtime": 38.8968,
+      "eval_samples_per_second": 642.726,
+      "eval_steps_per_second": 35.71,
+      "step": 60000
+    },
+    {
+      "epoch": 1.933921130114395,
+      "grad_norm": 1.3698302507400513,
+      "learning_rate": 1.653151848634038e-06,
+      "loss": 0.0832412052154541,
+      "step": 60100
+    },
+    {
+      "epoch": 1.9371389957041494,
+      "grad_norm": 4.799550533294678,
+      "learning_rate": 1.5727065032017249e-06,
+      "loss": 0.11331718444824218,
+      "step": 60200
+    },
+    {
+      "epoch": 1.9403568612939037,
+      "grad_norm": 1.4409183263778687,
+      "learning_rate": 1.4922611577694116e-06,
+      "loss": 0.09095165252685547,
+      "step": 60300
+    },
+    {
+      "epoch": 1.943574726883658,
+      "grad_norm": 0.7844113111495972,
+      "learning_rate": 1.4118158123370983e-06,
+      "loss": 0.07738580703735351,
+      "step": 60400
+    },
+    {
+      "epoch": 1.9467925924734124,
+      "grad_norm": 1.4404165744781494,
+      "learning_rate": 1.331370466904785e-06,
+      "loss": 0.08979734420776367,
+      "step": 60500
+    },
+    {
+      "epoch": 1.9500104580631668,
+      "grad_norm": 12.599478721618652,
+      "learning_rate": 1.2509251214724717e-06,
+      "loss": 0.10170120239257813,
+      "step": 60600
+    },
+    {
+      "epoch": 1.953228323652921,
+      "grad_norm": 16.47142791748047,
+      "learning_rate": 1.1704797760401584e-06,
+      "loss": 0.08839907646179199,
+      "step": 60700
+    },
+    {
+      "epoch": 1.9564461892426754,
+      "grad_norm": 1.4317628145217896,
+      "learning_rate": 1.0900344306078451e-06,
+      "loss": 0.10484655380249024,
+      "step": 60800
+    },
+    {
+      "epoch": 1.9596640548324298,
+      "grad_norm": 1.4019297361373901,
+      "learning_rate": 1.0095890851755318e-06,
+      "loss": 0.0975730323791504,
+      "step": 60900
+    },
+    {
+      "epoch": 1.9628819204221841,
+      "grad_norm": 1.428163766860962,
+      "learning_rate": 9.291437397432184e-07,
+      "loss": 0.09934920310974121,
+      "step": 61000
+    },
+    {
+      "epoch": 1.9660997860119382,
+      "grad_norm": 0.9141352772712708,
+      "learning_rate": 8.486983943109052e-07,
+      "loss": 0.09621753692626953,
+      "step": 61100
+    },
+    {
+      "epoch": 1.9693176516016926,
+      "grad_norm": 4.859970569610596,
+      "learning_rate": 7.682530488785919e-07,
+      "loss": 0.0880228328704834,
+      "step": 61200
+    },
+    {
+      "epoch": 1.972535517191447,
+      "grad_norm": 14.404598236083984,
+      "learning_rate": 6.878077034462786e-07,
+      "loss": 0.10370702743530273,
+      "step": 61300
+    },
+    {
+      "epoch": 1.9757533827812013,
+      "grad_norm": 1.890513300895691,
+      "learning_rate": 6.073623580139654e-07,
+      "loss": 0.08836669921875,
+      "step": 61400
+    },
+    {
+      "epoch": 1.9789712483709554,
+      "grad_norm": 3.365849494934082,
+      "learning_rate": 5.269170125816521e-07,
+      "loss": 0.0887114429473877,
+      "step": 61500
+    },
+    {
+      "epoch": 1.9821891139607097,
+      "grad_norm": 2.7111990451812744,
+      "learning_rate": 4.4647166714933875e-07,
+      "loss": 0.10604698181152344,
+      "step": 61600
+    },
+    {
+      "epoch": 1.985406979550464,
+      "grad_norm": 1.373115062713623,
+      "learning_rate": 3.6602632171702546e-07,
+      "loss": 0.08321118354797363,
+      "step": 61700
+    },
+    {
+      "epoch": 1.9886248451402184,
+      "grad_norm": 1.2105165719985962,
+      "learning_rate": 2.855809762847122e-07,
+      "loss": 0.09142358779907227,
+      "step": 61800
+    },
+    {
+      "epoch": 1.9918427107299728,
+      "grad_norm": 1.9538366794586182,
+      "learning_rate": 2.0513563085239887e-07,
+      "loss": 0.07895866394042969,
+      "step": 61900
+    },
+    {
+      "epoch": 1.995060576319727,
+      "grad_norm": 0.8726534843444824,
+      "learning_rate": 1.246902854200856e-07,
+      "loss": 0.10273164749145508,
+      "step": 62000
+    },
+    {
+      "epoch": 1.9982784419094815,
+      "grad_norm": 3.1368095874786377,
+      "learning_rate": 4.424493998777231e-08,
+      "loss": 0.08670869827270508,
+      "step": 62100
+    },
+    {
+      "epoch": 2.0,
+      "step": 62154,
+      "total_flos": 1.5480486185003853e+18,
+      "train_loss": 0.1780550116395906,
+      "train_runtime": 17614.9903,
+      "train_samples_per_second": 508.09,
+      "train_steps_per_second": 3.528
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 62154,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 2,
+  "save_steps": 2500,
+  "stateful_callbacks": {
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.5480486185003853e+18,
+  "train_batch_size": 18,
+  "trial_name": null,
+  "trial_params": null
+}