diff --git "a/trainer_state.json" "b/trainer_state.json"
new file mode 100644--- /dev/null
+++ "b/trainer_state.json"
@@ -0,0 +1,7327 @@
+{
+  "best_metric": 0.2870275791624106,
+  "best_model_checkpoint": "/media/cse/HDD/Shawon/shawon/MY DATA/ViViT_WLASL_200_epochs_p20/checkpoint-57160",
+  "epoch": 51.005000699888015,
+  "eval_steps": 500,
+  "global_step": 92885,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 0.0002799552071668533,
+      "grad_norm": 46.26704025268555,
+      "learning_rate": 1.3577827547592386e-07,
+      "loss": 30.8454,
+      "step": 100
+    },
+    {
+      "epoch": 0.0005599104143337066,
+      "grad_norm": 45.19619369506836,
+      "learning_rate": 2.757558790593505e-07,
+      "loss": 30.9112,
+      "step": 200
+    },
+    {
+      "epoch": 0.0008398656215005599,
+      "grad_norm": 44.790931701660156,
+      "learning_rate": 4.1573348264277716e-07,
+      "loss": 30.966,
+      "step": 300
+    },
+    {
+      "epoch": 0.0011198208286674132,
+      "grad_norm": 46.15544509887695,
+      "learning_rate": 5.557110862262039e-07,
+      "loss": 30.9534,
+      "step": 400
+    },
+    {
+      "epoch": 0.0013997760358342665,
+      "grad_norm": 45.963539123535156,
+      "learning_rate": 6.956886898096305e-07,
+      "loss": 30.9846,
+      "step": 500
+    },
+    {
+      "epoch": 0.0016797312430011197,
+      "grad_norm": 46.944801330566406,
+      "learning_rate": 8.356662933930571e-07,
+      "loss": 30.8929,
+      "step": 600
+    },
+    {
+      "epoch": 0.001959686450167973,
+      "grad_norm": 44.26920700073242,
+      "learning_rate": 9.756438969764837e-07,
+      "loss": 30.7879,
+      "step": 700
+    },
+    {
+      "epoch": 0.0022396416573348264,
+      "grad_norm": 44.74138259887695,
+      "learning_rate": 1.1156215005599104e-06,
+      "loss": 30.8412,
+      "step": 800
+    },
+    {
+      "epoch": 0.0025195968645016797,
+      "grad_norm": 44.76321792602539,
+      "learning_rate": 1.2555991041433373e-06,
+      "loss": 30.8549,
+      "step": 900
+    },
+    {
+      "epoch": 0.002799552071668533,
+      "grad_norm": 44.44749450683594,
+      "learning_rate": 1.395576707726764e-06,
+      "loss": 31.0151,
+      "step": 1000
+    },
+    {
+      "epoch": 0.003079507278835386,
+      "grad_norm": 44.81037139892578,
+      "learning_rate": 1.5355543113101904e-06,
+      "loss": 30.8179,
+      "step": 1100
+    },
+    {
+      "epoch": 0.0033594624860022394,
+      "grad_norm": 44.556419372558594,
+      "learning_rate": 1.675531914893617e-06,
+      "loss": 30.923,
+      "step": 1200
+    },
+    {
+      "epoch": 0.003639417693169093,
+      "grad_norm": 44.91850280761719,
+      "learning_rate": 1.815509518477044e-06,
+      "loss": 30.8516,
+      "step": 1300
+    },
+    {
+      "epoch": 0.003919372900335946,
+      "grad_norm": 42.26573181152344,
+      "learning_rate": 1.9554871220604704e-06,
+      "loss": 30.8223,
+      "step": 1400
+    },
+    {
+      "epoch": 0.0041993281075028,
+      "grad_norm": 42.67766189575195,
+      "learning_rate": 2.095464725643897e-06,
+      "loss": 30.8226,
+      "step": 1500
+    },
+    {
+      "epoch": 0.004479283314669653,
+      "grad_norm": 39.56558609008789,
+      "learning_rate": 2.2354423292273237e-06,
+      "loss": 30.7627,
+      "step": 1600
+    },
+    {
+      "epoch": 0.004759238521836506,
+      "grad_norm": 36.79159164428711,
+      "learning_rate": 2.3754199328107503e-06,
+      "loss": 30.7847,
+      "step": 1700
+    },
+    {
+      "epoch": 0.005,
+      "eval_accuracy": 0.0002553626149131767,
+      "eval_f1": 2.9184298847220195e-05,
+      "eval_loss": 7.671034336090088,
+      "eval_precision": 1.5476522115950103e-05,
+      "eval_recall": 0.0002553626149131767,
+      "eval_runtime": 261.8693,
+      "eval_samples_per_second": 14.954,
+      "eval_steps_per_second": 7.477,
+      "eval_top_10_accuracy": 0.004341164453524004,
+      "eval_top_1_accuracy": 0.0002553626149131767,
+      "eval_top_5_accuracy": 0.0012768130745658835,
+      "step": 1786
+    },
+    {
+      "epoch": 1.0000384938409854,
+      "grad_norm": 37.38892364501953,
+      "learning_rate": 2.515397536394177e-06,
+      "loss": 30.7364,
+      "step": 1800
+    },
+    {
+      "epoch": 1.0003184490481523,
+      "grad_norm": 38.72145080566406,
+      "learning_rate": 2.6553751399776037e-06,
+      "loss": 30.544,
+      "step": 1900
+    },
+    {
+      "epoch": 1.000598404255319,
+      "grad_norm": 38.69617462158203,
+      "learning_rate": 2.7953527435610303e-06,
+      "loss": 30.5909,
+      "step": 2000
+    },
+    {
+      "epoch": 1.000878359462486,
+      "grad_norm": 35.71845626831055,
+      "learning_rate": 2.933930571108623e-06,
+      "loss": 30.5322,
+      "step": 2100
+    },
+    {
+      "epoch": 1.0011583146696528,
+      "grad_norm": 37.04778289794922,
+      "learning_rate": 3.0739081746920493e-06,
+      "loss": 30.5843,
+      "step": 2200
+    },
+    {
+      "epoch": 1.0014382698768196,
+      "grad_norm": 34.91965866088867,
+      "learning_rate": 3.213885778275476e-06,
+      "loss": 30.4873,
+      "step": 2300
+    },
+    {
+      "epoch": 1.0017182250839867,
+      "grad_norm": 31.99736785888672,
+      "learning_rate": 3.353863381858903e-06,
+      "loss": 30.4577,
+      "step": 2400
+    },
+    {
+      "epoch": 1.0019981802911535,
+      "grad_norm": 32.1821403503418,
+      "learning_rate": 3.4938409854423293e-06,
+      "loss": 30.521,
+      "step": 2500
+    },
+    {
+      "epoch": 1.0022781354983203,
+      "grad_norm": 31.080778121948242,
+      "learning_rate": 3.633818589025756e-06,
+      "loss": 30.4738,
+      "step": 2600
+    },
+    {
+      "epoch": 1.0025580907054872,
+      "grad_norm": 33.43671417236328,
+      "learning_rate": 3.773796192609183e-06,
+      "loss": 30.518,
+      "step": 2700
+    },
+    {
+      "epoch": 1.002838045912654,
+      "grad_norm": 30.197689056396484,
+      "learning_rate": 3.913773796192609e-06,
+      "loss": 30.4191,
+      "step": 2800
+    },
+    {
+      "epoch": 1.0031180011198209,
+      "grad_norm": 33.55823516845703,
+      "learning_rate": 4.053751399776036e-06,
+      "loss": 30.4412,
+      "step": 2900
+    },
+    {
+      "epoch": 1.0033979563269877,
+      "grad_norm": 33.3959846496582,
+      "learning_rate": 4.193729003359463e-06,
+      "loss": 30.4303,
+      "step": 3000
+    },
+    {
+      "epoch": 1.0036779115341545,
+      "grad_norm": 33.3631706237793,
+      "learning_rate": 4.333706606942889e-06,
+      "loss": 30.3455,
+      "step": 3100
+    },
+    {
+      "epoch": 1.0039578667413214,
+      "grad_norm": 31.35999870300293,
+      "learning_rate": 4.473684210526316e-06,
+      "loss": 30.3703,
+      "step": 3200
+    },
+    {
+      "epoch": 1.0042378219484882,
+      "grad_norm": 34.44354248046875,
+      "learning_rate": 4.613661814109743e-06,
+      "loss": 30.5096,
+      "step": 3300
+    },
+    {
+      "epoch": 1.004517777155655,
+      "grad_norm": 34.77235794067383,
+      "learning_rate": 4.753639417693169e-06,
+      "loss": 30.4935,
+      "step": 3400
+    },
+    {
+      "epoch": 1.0047977323628219,
+      "grad_norm": 34.465576171875,
+      "learning_rate": 4.893617021276596e-06,
+      "loss": 30.4187,
+      "step": 3500
+    },
+    {
+      "epoch": 1.0049993001119821,
+      "eval_accuracy": 0.0020429009193054137,
+      "eval_f1": 0.0003999515192909813,
+      "eval_loss": 7.604901313781738,
+      "eval_precision": 0.0002671184173273292,
+      "eval_recall": 0.0020429009193054137,
+      "eval_runtime": 221.6582,
+      "eval_samples_per_second": 17.667,
+      "eval_steps_per_second": 8.833,
+      "eval_top_10_accuracy": 0.011491317671092951,
+      "eval_top_1_accuracy": 0.0020429009193054137,
+      "eval_top_5_accuracy": 0.0056179775280898875,
+      "step": 3572
+    },
+    {
+      "epoch": 2.000076987681971,
+      "grad_norm": 33.962181091308594,
+      "learning_rate": 5.0335946248600226e-06,
+      "loss": 30.4182,
+      "step": 3600
+    },
+    {
+      "epoch": 2.000356942889138,
+      "grad_norm": 41.75550079345703,
+      "learning_rate": 5.173572228443449e-06,
+      "loss": 30.1018,
+      "step": 3700
+    },
+    {
+      "epoch": 2.0006368980963045,
+      "grad_norm": 37.583744049072266,
+      "learning_rate": 5.313549832026876e-06,
+      "loss": 30.0925,
+      "step": 3800
+    },
+    {
+      "epoch": 2.0009168533034716,
+      "grad_norm": 38.75505447387695,
+      "learning_rate": 5.4535274356103025e-06,
+      "loss": 30.0771,
+      "step": 3900
+    },
+    {
+      "epoch": 2.001196808510638,
+      "grad_norm": 42.10888671875,
+      "learning_rate": 5.593505039193729e-06,
+      "loss": 30.072,
+      "step": 4000
+    },
+    {
+      "epoch": 2.0014767637178053,
+      "grad_norm": 39.6405029296875,
+      "learning_rate": 5.7320828667413216e-06,
+      "loss": 30.0103,
+      "step": 4100
+    },
+    {
+      "epoch": 2.001756718924972,
+      "grad_norm": 37.14402389526367,
+      "learning_rate": 5.872060470324748e-06,
+      "loss": 30.0757,
+      "step": 4200
+    },
+    {
+      "epoch": 2.002036674132139,
+      "grad_norm": 41.2797737121582,
+      "learning_rate": 6.012038073908175e-06,
+      "loss": 29.9727,
+      "step": 4300
+    },
+    {
+      "epoch": 2.0023166293393055,
+      "grad_norm": 42.102169036865234,
+      "learning_rate": 6.1520156774916016e-06,
+      "loss": 30.1018,
+      "step": 4400
+    },
+    {
+      "epoch": 2.0025965845464726,
+      "grad_norm": 44.31977462768555,
+      "learning_rate": 6.291993281075029e-06,
+      "loss": 29.9593,
+      "step": 4500
+    },
+    {
+      "epoch": 2.002876539753639,
+      "grad_norm": 41.01313781738281,
+      "learning_rate": 6.431970884658455e-06,
+      "loss": 30.0683,
+      "step": 4600
+    },
+    {
+      "epoch": 2.0031564949608063,
+      "grad_norm": 45.978580474853516,
+      "learning_rate": 6.5719484882418815e-06,
+      "loss": 30.031,
+      "step": 4700
+    },
+    {
+      "epoch": 2.0034364501679733,
+      "grad_norm": 43.09092712402344,
+      "learning_rate": 6.711926091825309e-06,
+      "loss": 29.9496,
+      "step": 4800
+    },
+    {
+      "epoch": 2.00371640537514,
+      "grad_norm": 48.287994384765625,
+      "learning_rate": 6.851903695408735e-06,
+      "loss": 29.9667,
+      "step": 4900
+    },
+    {
+      "epoch": 2.003996360582307,
+      "grad_norm": 40.08522033691406,
+      "learning_rate": 6.9918812989921615e-06,
+      "loss": 29.8607,
+      "step": 5000
+    },
+    {
+      "epoch": 2.0042763157894736,
+      "grad_norm": 43.70709991455078,
+      "learning_rate": 7.131858902575589e-06,
+      "loss": 29.7901,
+      "step": 5100
+    },
+    {
+      "epoch": 2.0045562709966407,
+      "grad_norm": 44.53886413574219,
+      "learning_rate": 7.271836506159015e-06,
+      "loss": 29.8422,
+      "step": 5200
+    },
+    {
+      "epoch": 2.0048362262038073,
+      "grad_norm": 46.5408821105957,
+      "learning_rate": 7.4118141097424415e-06,
+      "loss": 29.7427,
+      "step": 5300
+    },
+    {
+      "epoch": 2.0049986002239644,
+      "eval_accuracy": 0.003575076608784474,
+      "eval_f1": 0.0008090265726319296,
+      "eval_loss": 7.486084938049316,
+      "eval_precision": 0.0008795780348085499,
+      "eval_recall": 0.003575076608784474,
+      "eval_runtime": 222.3048,
+      "eval_samples_per_second": 17.615,
+      "eval_steps_per_second": 8.808,
+      "eval_top_10_accuracy": 0.027068437180796732,
+      "eval_top_1_accuracy": 0.003575076608784474,
+      "eval_top_5_accuracy": 0.016087844739530132,
+      "step": 5358
+    },
+    {
+      "epoch": 3.0001154815229563,
+      "grad_norm": 44.443878173828125,
+      "learning_rate": 7.551791713325867e-06,
+      "loss": 29.5441,
+      "step": 5400
+    },
+    {
+      "epoch": 3.0003954367301233,
+      "grad_norm": 52.79859924316406,
+      "learning_rate": 7.691769316909295e-06,
+      "loss": 28.9231,
+      "step": 5500
+    },
+    {
+      "epoch": 3.00067539193729,
+      "grad_norm": 47.09991455078125,
+      "learning_rate": 7.831746920492722e-06,
+      "loss": 29.1026,
+      "step": 5600
+    },
+    {
+      "epoch": 3.000955347144457,
+      "grad_norm": 49.0944709777832,
+      "learning_rate": 7.971724524076148e-06,
+      "loss": 29.1107,
+      "step": 5700
+    },
+    {
+      "epoch": 3.0012353023516236,
+      "grad_norm": 49.614070892333984,
+      "learning_rate": 8.111702127659574e-06,
+      "loss": 29.0765,
+      "step": 5800
+    },
+    {
+      "epoch": 3.0015152575587907,
+      "grad_norm": 48.07614517211914,
+      "learning_rate": 8.251679731243001e-06,
+      "loss": 29.0133,
+      "step": 5900
+    },
+    {
+      "epoch": 3.0017952127659573,
+      "grad_norm": 48.23404312133789,
+      "learning_rate": 8.391657334826427e-06,
+      "loss": 28.8593,
+      "step": 6000
+    },
+    {
+      "epoch": 3.0020751679731243,
+      "grad_norm": 49.85590744018555,
+      "learning_rate": 8.53023516237402e-06,
+      "loss": 28.9609,
+      "step": 6100
+    },
+    {
+      "epoch": 3.002355123180291,
+      "grad_norm": 50.691890716552734,
+      "learning_rate": 8.670212765957447e-06,
+      "loss": 28.9257,
+      "step": 6200
+    },
+    {
+      "epoch": 3.002635078387458,
+      "grad_norm": 49.11813735961914,
+      "learning_rate": 8.810190369540875e-06,
+      "loss": 28.7913,
+      "step": 6300
+    },
+    {
+      "epoch": 3.0029150335946246,
+      "grad_norm": 49.65032958984375,
+      "learning_rate": 8.9501679731243e-06,
+      "loss": 28.8195,
+      "step": 6400
+    },
+    {
+      "epoch": 3.0031949888017917,
+      "grad_norm": 49.955528259277344,
+      "learning_rate": 9.090145576707728e-06,
+      "loss": 28.7372,
+      "step": 6500
+    },
+    {
+      "epoch": 3.0034749440089588,
+      "grad_norm": 48.818843841552734,
+      "learning_rate": 9.230123180291154e-06,
+      "loss": 28.705,
+      "step": 6600
+    },
+    {
+      "epoch": 3.0037548992161254,
+      "grad_norm": 48.66638946533203,
+      "learning_rate": 9.37010078387458e-06,
+      "loss": 28.6832,
+      "step": 6700
+    },
+    {
+      "epoch": 3.0040348544232924,
+      "grad_norm": 50.39780044555664,
+      "learning_rate": 9.510078387458007e-06,
+      "loss": 28.6613,
+      "step": 6800
+    },
+    {
+      "epoch": 3.004314809630459,
+      "grad_norm": 51.30269241333008,
+      "learning_rate": 9.650055991041433e-06,
+      "loss": 28.4991,
+      "step": 6900
+    },
+    {
+      "epoch": 3.004594764837626,
+      "grad_norm": 51.30686569213867,
+      "learning_rate": 9.79003359462486e-06,
+      "loss": 28.5717,
+      "step": 7000
+    },
+    {
+      "epoch": 3.0048747200447927,
+      "grad_norm": 50.400569915771484,
+      "learning_rate": 9.930011198208288e-06,
+      "loss": 28.4534,
+      "step": 7100
+    },
+    {
+      "epoch": 3.005000699888018,
+      "eval_accuracy": 0.011491317671092951,
+      "eval_f1": 0.004413303303826739,
+      "eval_loss": 7.249152183532715,
+      "eval_precision": 0.0048748011754680045,
+      "eval_recall": 0.011491317671092951,
+      "eval_runtime": 220.6205,
+      "eval_samples_per_second": 17.75,
+      "eval_steps_per_second": 8.875,
+      "eval_top_10_accuracy": 0.06052093973442288,
+      "eval_top_1_accuracy": 0.011491317671092951,
+      "eval_top_5_accuracy": 0.037538304392236974,
+      "step": 7145
+    },
+    {
+      "epoch": 4.000153975363942,
+      "grad_norm": 50.86111068725586,
+      "learning_rate": 1.0069988801791714e-05,
+      "loss": 27.9795,
+      "step": 7200
+    },
+    {
+      "epoch": 4.000433930571108,
+      "grad_norm": 49.65926742553711,
+      "learning_rate": 1.020996640537514e-05,
+      "loss": 27.7397,
+      "step": 7300
+    },
+    {
+      "epoch": 4.000713885778276,
+      "grad_norm": 50.528770446777344,
+      "learning_rate": 1.0349944008958567e-05,
+      "loss": 27.5468,
+      "step": 7400
+    },
+    {
+      "epoch": 4.000993840985442,
+      "grad_norm": 50.71484375,
+      "learning_rate": 1.0489921612541993e-05,
+      "loss": 27.4879,
+      "step": 7500
+    },
+    {
+      "epoch": 4.001273796192609,
+      "grad_norm": 49.60582733154297,
+      "learning_rate": 1.062989921612542e-05,
+      "loss": 27.4364,
+      "step": 7600
+    },
+    {
+      "epoch": 4.001553751399776,
+      "grad_norm": 50.794559478759766,
+      "learning_rate": 1.0769876819708848e-05,
+      "loss": 27.3306,
+      "step": 7700
+    },
+    {
+      "epoch": 4.001833706606943,
+      "grad_norm": 52.04481887817383,
+      "learning_rate": 1.0909854423292274e-05,
+      "loss": 27.1802,
+      "step": 7800
+    },
+    {
+      "epoch": 4.00211366181411,
+      "grad_norm": 53.98943328857422,
+      "learning_rate": 1.1049832026875701e-05,
+      "loss": 27.4401,
+      "step": 7900
+    },
+    {
+      "epoch": 4.002393617021276,
+      "grad_norm": 50.8287467956543,
+      "learning_rate": 1.1189809630459127e-05,
+      "loss": 27.3483,
+      "step": 8000
+    },
+    {
+      "epoch": 4.002673572228444,
+      "grad_norm": 47.586692810058594,
+      "learning_rate": 1.132838745800672e-05,
+      "loss": 27.3851,
+      "step": 8100
+    },
+    {
+      "epoch": 4.0029535274356105,
+      "grad_norm": 50.97801971435547,
+      "learning_rate": 1.1468365061590145e-05,
+      "loss": 27.0601,
+      "step": 8200
+    },
+    {
+      "epoch": 4.003233482642777,
+      "grad_norm": 49.2835693359375,
+      "learning_rate": 1.1608342665173573e-05,
+      "loss": 27.1775,
+      "step": 8300
+    },
+    {
+      "epoch": 4.003513437849944,
+      "grad_norm": 49.225276947021484,
+      "learning_rate": 1.1748320268756998e-05,
+      "loss": 27.1897,
+      "step": 8400
+    },
+    {
+      "epoch": 4.003793393057111,
+      "grad_norm": 50.65304946899414,
+      "learning_rate": 1.1888297872340426e-05,
+      "loss": 27.2348,
+      "step": 8500
+    },
+    {
+      "epoch": 4.004073348264278,
+      "grad_norm": 51.88926315307617,
+      "learning_rate": 1.2028275475923854e-05,
+      "loss": 26.9358,
+      "step": 8600
+    },
+    {
+      "epoch": 4.0043533034714445,
+      "grad_norm": 53.61826705932617,
+      "learning_rate": 1.216825307950728e-05,
+      "loss": 26.8982,
+      "step": 8700
+    },
+    {
+      "epoch": 4.004633258678611,
+      "grad_norm": 48.89099884033203,
+      "learning_rate": 1.2308230683090707e-05,
+      "loss": 26.8073,
+      "step": 8800
+    },
+    {
+      "epoch": 4.004913213885779,
+      "grad_norm": 49.341552734375,
+      "learning_rate": 1.2448208286674133e-05,
+      "loss": 26.8561,
+      "step": 8900
+    },
+    {
+      "epoch": 4.005,
+      "eval_accuracy": 0.031920326864147086,
+      "eval_f1": 0.01254087277909593,
+      "eval_loss": 6.92970085144043,
+      "eval_precision": 0.011491031769049,
+      "eval_recall": 0.031920326864147086,
+      "eval_runtime": 224.0829,
+      "eval_samples_per_second": 17.476,
+      "eval_steps_per_second": 8.738,
+      "eval_top_10_accuracy": 0.12078651685393259,
+      "eval_top_1_accuracy": 0.031664964249233915,
+      "eval_top_5_accuracy": 0.07660878447395301,
+      "step": 8931
+    },
+    {
+      "epoch": 5.0001924692049275,
+      "grad_norm": 53.48250961303711,
+      "learning_rate": 1.2588185890257558e-05,
+      "loss": 26.0557,
+      "step": 9000
+    },
+    {
+      "epoch": 5.000472424412094,
+      "grad_norm": 49.2156982421875,
+      "learning_rate": 1.2728163493840986e-05,
+      "loss": 25.6844,
+      "step": 9100
+    },
+    {
+      "epoch": 5.000752379619261,
+      "grad_norm": 51.55862808227539,
+      "learning_rate": 1.2868141097424413e-05,
+      "loss": 25.5654,
+      "step": 9200
+    },
+    {
+      "epoch": 5.001032334826427,
+      "grad_norm": 54.36677551269531,
+      "learning_rate": 1.3008118701007841e-05,
+      "loss": 25.6638,
+      "step": 9300
+    },
+    {
+      "epoch": 5.001312290033595,
+      "grad_norm": 50.89421463012695,
+      "learning_rate": 1.3148096304591265e-05,
+      "loss": 25.8723,
+      "step": 9400
+    },
+    {
+      "epoch": 5.0015922452407615,
+      "grad_norm": 51.31764602661133,
+      "learning_rate": 1.3288073908174693e-05,
+      "loss": 25.5003,
+      "step": 9500
+    },
+    {
+      "epoch": 5.001872200447928,
+      "grad_norm": 50.556793212890625,
+      "learning_rate": 1.3428051511758118e-05,
+      "loss": 25.6165,
+      "step": 9600
+    },
+    {
+      "epoch": 5.002152155655095,
+      "grad_norm": 51.41532897949219,
+      "learning_rate": 1.3568029115341546e-05,
+      "loss": 25.5103,
+      "step": 9700
+    },
+    {
+      "epoch": 5.002432110862262,
+      "grad_norm": 50.99569320678711,
+      "learning_rate": 1.3708006718924973e-05,
+      "loss": 25.2506,
+      "step": 9800
+    },
+    {
+      "epoch": 5.002712066069429,
+      "grad_norm": 51.33198165893555,
+      "learning_rate": 1.3847984322508401e-05,
+      "loss": 25.4653,
+      "step": 9900
+    },
+    {
+      "epoch": 5.0029920212765955,
+      "grad_norm": 49.71004867553711,
+      "learning_rate": 1.3987961926091825e-05,
+      "loss": 25.4142,
+      "step": 10000
+    },
+    {
+      "epoch": 5.003271976483763,
+      "grad_norm": 49.55411911010742,
+      "learning_rate": 1.4126539753639417e-05,
+      "loss": 25.2711,
+      "step": 10100
+    },
+    {
+      "epoch": 5.00355193169093,
+      "grad_norm": 50.0079345703125,
+      "learning_rate": 1.4266517357222845e-05,
+      "loss": 25.3369,
+      "step": 10200
+    },
+    {
+      "epoch": 5.003831886898096,
+      "grad_norm": 49.790077209472656,
+      "learning_rate": 1.4406494960806272e-05,
+      "loss": 25.2757,
+      "step": 10300
+    },
+    {
+      "epoch": 5.004111842105263,
+      "grad_norm": 49.824241638183594,
+      "learning_rate": 1.4546472564389698e-05,
+      "loss": 25.0612,
+      "step": 10400
+    },
+    {
+      "epoch": 5.00439179731243,
+      "grad_norm": 50.10408020019531,
+      "learning_rate": 1.4686450167973126e-05,
+      "loss": 25.0615,
+      "step": 10500
+    },
+    {
+      "epoch": 5.004671752519597,
+      "grad_norm": 50.95100021362305,
+      "learning_rate": 1.4826427771556553e-05,
+      "loss": 25.2522,
+      "step": 10600
+    },
+    {
+      "epoch": 5.004951707726764,
+      "grad_norm": 50.80625915527344,
+      "learning_rate": 1.4966405375139977e-05,
+      "loss": 24.997,
+      "step": 10700
+    },
+    {
+      "epoch": 5.004999300111982,
+      "eval_accuracy": 0.045965270684371805,
+      "eval_f1": 0.020995371903437376,
+      "eval_loss": 6.5685930252075195,
+      "eval_precision": 0.021173968378675568,
+      "eval_recall": 0.045965270684371805,
+      "eval_runtime": 221.4663,
+      "eval_samples_per_second": 17.682,
+      "eval_steps_per_second": 8.841,
+      "eval_top_10_accuracy": 0.19458631256384065,
+      "eval_top_1_accuracy": 0.045965270684371805,
+      "eval_top_5_accuracy": 0.12921348314606743,
+      "step": 10717
+    },
+    {
+      "epoch": 6.0002309630459125,
+      "grad_norm": 50.683738708496094,
+      "learning_rate": 1.5106382978723405e-05,
+      "loss": 24.0698,
+      "step": 10800
+    },
+    {
+      "epoch": 6.000510918253079,
+      "grad_norm": 51.920692443847656,
+      "learning_rate": 1.5246360582306832e-05,
+      "loss": 23.6795,
+      "step": 10900
+    },
+    {
+      "epoch": 6.000790873460247,
+      "grad_norm": 51.51482391357422,
+      "learning_rate": 1.538633818589026e-05,
+      "loss": 23.793,
+      "step": 11000
+    },
+    {
+      "epoch": 6.001070828667413,
+      "grad_norm": 53.38042068481445,
+      "learning_rate": 1.5526315789473686e-05,
+      "loss": 23.7737,
+      "step": 11100
+    },
+    {
+      "epoch": 6.00135078387458,
+      "grad_norm": 51.36591720581055,
+      "learning_rate": 1.566629339305711e-05,
+      "loss": 23.4639,
+      "step": 11200
+    },
+    {
+      "epoch": 6.0016307390817465,
+      "grad_norm": 51.339813232421875,
+      "learning_rate": 1.5806270996640537e-05,
+      "loss": 23.6826,
+      "step": 11300
+    },
+    {
+      "epoch": 6.001910694288914,
+      "grad_norm": 51.70831298828125,
+      "learning_rate": 1.5946248600223963e-05,
+      "loss": 23.3406,
+      "step": 11400
+    },
+    {
+      "epoch": 6.002190649496081,
+      "grad_norm": 52.543033599853516,
+      "learning_rate": 1.6086226203807392e-05,
+      "loss": 23.6466,
+      "step": 11500
+    },
+    {
+      "epoch": 6.002470604703247,
+      "grad_norm": 50.490440368652344,
+      "learning_rate": 1.6226203807390818e-05,
+      "loss": 23.6253,
+      "step": 11600
+    },
+    {
+      "epoch": 6.002750559910415,
+      "grad_norm": 51.624053955078125,
+      "learning_rate": 1.6366181410974247e-05,
+      "loss": 23.296,
+      "step": 11700
+    },
+    {
+      "epoch": 6.003030515117581,
+      "grad_norm": 48.22161102294922,
+      "learning_rate": 1.650615901455767e-05,
+      "loss": 23.2675,
+      "step": 11800
+    },
+    {
+      "epoch": 6.003310470324748,
+      "grad_norm": 49.2001838684082,
+      "learning_rate": 1.66461366181411e-05,
+      "loss": 23.3034,
+      "step": 11900
+    },
+    {
+      "epoch": 6.003590425531915,
+      "grad_norm": 46.579803466796875,
+      "learning_rate": 1.6786114221724525e-05,
+      "loss": 23.2177,
+      "step": 12000
+    },
+    {
+      "epoch": 6.003870380739082,
+      "grad_norm": 51.24250411987305,
+      "learning_rate": 1.692469204927212e-05,
+      "loss": 23.0834,
+      "step": 12100
+    },
+    {
+      "epoch": 6.004150335946249,
+      "grad_norm": 53.71959686279297,
+      "learning_rate": 1.7064669652855545e-05,
+      "loss": 23.277,
+      "step": 12200
+    },
+    {
+      "epoch": 6.004430291153415,
+      "grad_norm": 51.754024505615234,
+      "learning_rate": 1.720464725643897e-05,
+      "loss": 23.0179,
+      "step": 12300
+    },
+    {
+      "epoch": 6.004710246360582,
+      "grad_norm": 52.60391616821289,
+      "learning_rate": 1.7344624860022396e-05,
+      "loss": 23.0741,
+      "step": 12400
+    },
+    {
+      "epoch": 6.004990201567749,
+      "grad_norm": 51.15227127075195,
+      "learning_rate": 1.7484602463605822e-05,
+      "loss": 22.8381,
+      "step": 12500
+    },
+    {
+      "epoch": 6.004998600223964,
+      "eval_accuracy": 0.07482124616956078,
+      "eval_f1": 0.03821121445416449,
+      "eval_loss": 6.204948425292969,
+      "eval_precision": 0.03864142268606078,
+      "eval_recall": 0.07482124616956078,
+      "eval_runtime": 224.4924,
+      "eval_samples_per_second": 17.444,
+      "eval_steps_per_second": 8.722,
+      "eval_top_10_accuracy": 0.2604698672114402,
+      "eval_top_1_accuracy": 0.0745658835546476,
+      "eval_top_5_accuracy": 0.18590398365679264,
+      "step": 12503
+    },
+    {
+      "epoch": 7.000269456886898,
+      "grad_norm": 51.505096435546875,
+      "learning_rate": 1.762458006718925e-05,
+      "loss": 21.753,
+      "step": 12600
+    },
+    {
+      "epoch": 7.000549412094065,
+      "grad_norm": 53.034732818603516,
+      "learning_rate": 1.7764557670772677e-05,
+      "loss": 21.4792,
+      "step": 12700
+    },
+    {
+      "epoch": 7.000829367301232,
+      "grad_norm": 53.07753372192383,
+      "learning_rate": 1.7904535274356103e-05,
+      "loss": 21.5145,
+      "step": 12800
+    },
+    {
+      "epoch": 7.001109322508398,
+      "grad_norm": 52.74058532714844,
+      "learning_rate": 1.8044512877939532e-05,
+      "loss": 21.4575,
+      "step": 12900
+    },
+    {
+      "epoch": 7.001389277715566,
+      "grad_norm": 53.695003509521484,
+      "learning_rate": 1.8184490481522955e-05,
+      "loss": 21.6702,
+      "step": 13000
+    },
+    {
+      "epoch": 7.001669232922732,
+      "grad_norm": 54.33465576171875,
+      "learning_rate": 1.8324468085106384e-05,
+      "loss": 21.3756,
+      "step": 13100
+    },
+    {
+      "epoch": 7.001949188129899,
+      "grad_norm": 55.21706008911133,
+      "learning_rate": 1.846444568868981e-05,
+      "loss": 21.2789,
+      "step": 13200
+    },
+    {
+      "epoch": 7.0022291433370665,
+      "grad_norm": 52.99767303466797,
+      "learning_rate": 1.860442329227324e-05,
+      "loss": 21.119,
+      "step": 13300
+    },
+    {
+      "epoch": 7.002509098544233,
+      "grad_norm": 52.56631851196289,
+      "learning_rate": 1.8744400895856665e-05,
+      "loss": 21.3768,
+      "step": 13400
+    },
+    {
+      "epoch": 7.0027890537514,
+      "grad_norm": 59.85367965698242,
+      "learning_rate": 1.888437849944009e-05,
+      "loss": 21.1635,
+      "step": 13500
+    },
+    {
+      "epoch": 7.003069008958566,
+      "grad_norm": 53.9185905456543,
+      "learning_rate": 1.9024356103023516e-05,
+      "loss": 21.239,
+      "step": 13600
+    },
+    {
+      "epoch": 7.003348964165734,
+      "grad_norm": 53.914737701416016,
+      "learning_rate": 1.9164333706606942e-05,
+      "loss": 21.2056,
+      "step": 13700
+    },
+    {
+      "epoch": 7.0036289193729,
+      "grad_norm": 53.44260025024414,
+      "learning_rate": 1.930431131019037e-05,
+      "loss": 20.8527,
+      "step": 13800
+    },
+    {
+      "epoch": 7.003908874580067,
+      "grad_norm": 50.94429016113281,
+      "learning_rate": 1.9444288913773797e-05,
+      "loss": 21.0278,
+      "step": 13900
+    },
+    {
+      "epoch": 7.004188829787234,
+      "grad_norm": 51.96830749511719,
+      "learning_rate": 1.9584266517357226e-05,
+      "loss": 20.979,
+      "step": 14000
+    },
+    {
+      "epoch": 7.004468784994401,
+      "grad_norm": 51.84593200683594,
+      "learning_rate": 1.9722844344904817e-05,
+      "loss": 21.0503,
+      "step": 14100
+    },
+    {
+      "epoch": 7.004748740201568,
+      "grad_norm": 52.195037841796875,
+      "learning_rate": 1.986142217245241e-05,
+      "loss": 20.56,
+      "step": 14200
+    },
+    {
+      "epoch": 7.005000699888018,
+      "eval_accuracy": 0.10214504596527069,
+      "eval_f1": 0.05936071255358349,
+      "eval_loss": 5.805036544799805,
+      "eval_precision": 0.05913221935096704,
+      "eval_recall": 0.10214504596527069,
+      "eval_runtime": 223.6709,
+      "eval_samples_per_second": 17.508,
+      "eval_steps_per_second": 8.754,
+      "eval_top_10_accuracy": 0.34627170582226763,
+      "eval_top_1_accuracy": 0.10214504596527069,
+      "eval_top_5_accuracy": 0.2553626149131767,
+      "step": 14290
+    },
+    {
+      "epoch": 8.000027995520716,
+      "grad_norm": 51.8380012512207,
+      "learning_rate": 2.0001399776035833e-05,
+      "loss": 20.5048,
+      "step": 14300
+    },
+    {
+      "epoch": 8.000307950727883,
+      "grad_norm": 56.783084869384766,
+      "learning_rate": 2.0141377379619263e-05,
+      "loss": 19.3519,
+      "step": 14400
+    },
+    {
+      "epoch": 8.00058790593505,
+      "grad_norm": 54.16028594970703,
+      "learning_rate": 2.028135498320269e-05,
+      "loss": 18.8881,
+      "step": 14500
+    },
+    {
+      "epoch": 8.000867861142217,
+      "grad_norm": 52.2464714050293,
+      "learning_rate": 2.0421332586786114e-05,
+      "loss": 19.0802,
+      "step": 14600
+    },
+    {
+      "epoch": 8.001147816349384,
+      "grad_norm": 54.0433464050293,
+      "learning_rate": 2.0561310190369543e-05,
+      "loss": 19.1179,
+      "step": 14700
+    },
+    {
+      "epoch": 8.001427771556552,
+      "grad_norm": 54.63380432128906,
+      "learning_rate": 2.070128779395297e-05,
+      "loss": 19.0229,
+      "step": 14800
+    },
+    {
+      "epoch": 8.001707726763717,
+      "grad_norm": 56.0116081237793,
+      "learning_rate": 2.0841265397536395e-05,
+      "loss": 19.1359,
+      "step": 14900
+    },
+    {
+      "epoch": 8.001987681970885,
+      "grad_norm": 52.73674392700195,
+      "learning_rate": 2.098124300111982e-05,
+      "loss": 18.7772,
+      "step": 15000
+    },
+    {
+      "epoch": 8.002267637178052,
+      "grad_norm": 48.63119888305664,
+      "learning_rate": 2.112122060470325e-05,
+      "loss": 18.9625,
+      "step": 15100
+    },
+    {
+      "epoch": 8.002547592385218,
+      "grad_norm": 52.9703369140625,
+      "learning_rate": 2.1261198208286676e-05,
+      "loss": 18.926,
+      "step": 15200
+    },
+    {
+      "epoch": 8.002827547592386,
+      "grad_norm": 54.72795104980469,
+      "learning_rate": 2.14011758118701e-05,
+      "loss": 18.4723,
+      "step": 15300
+    },
+    {
+      "epoch": 8.003107502799551,
+      "grad_norm": 52.77252197265625,
+      "learning_rate": 2.1541153415453527e-05,
+      "loss": 18.5272,
+      "step": 15400
+    },
+    {
+      "epoch": 8.003387458006719,
+      "grad_norm": 56.37332534790039,
+      "learning_rate": 2.1681131019036953e-05,
+      "loss": 18.6133,
+      "step": 15500
+    },
+    {
+      "epoch": 8.003667413213886,
+      "grad_norm": 55.63984680175781,
+      "learning_rate": 2.1821108622620383e-05,
+      "loss": 18.7277,
+      "step": 15600
+    },
+    {
+      "epoch": 8.003947368421052,
+      "grad_norm": 53.248497009277344,
+      "learning_rate": 2.196108622620381e-05,
+      "loss": 18.8904,
+      "step": 15700
+    },
+    {
+      "epoch": 8.00422732362822,
+      "grad_norm": 56.491886138916016,
+      "learning_rate": 2.2101063829787234e-05,
+      "loss": 18.7138,
+      "step": 15800
+    },
+    {
+      "epoch": 8.004507278835387,
+      "grad_norm": 53.40477752685547,
+      "learning_rate": 2.2241041433370663e-05,
+      "loss": 18.5271,
+      "step": 15900
+    },
+    {
+      "epoch": 8.004787234042553,
+      "grad_norm": 54.428314208984375,
+      "learning_rate": 2.2381019036954086e-05,
+      "loss": 18.323,
+      "step": 16000
+    },
+    {
+      "epoch": 8.005,
+      "eval_accuracy": 0.13253319713993872,
+      "eval_f1": 0.07727072976763863,
+      "eval_loss": 5.4230194091796875,
+      "eval_precision": 0.072025240634506,
+      "eval_recall": 0.13253319713993872,
+      "eval_runtime": 228.6261,
+      "eval_samples_per_second": 17.128,
+      "eval_steps_per_second": 8.564,
+      "eval_top_10_accuracy": 0.4101123595505618,
+      "eval_top_1_accuracy": 0.1317671092951992,
+      "eval_top_5_accuracy": 0.31128702757916243,
+      "step": 16076
+    },
+    {
+      "epoch": 9.000066489361702,
+      "grad_norm": 58.01615905761719,
+      "learning_rate": 2.2520996640537515e-05,
+      "loss": 18.0109,
+      "step": 16100
+    },
+    {
+      "epoch": 9.00034644456887,
+      "grad_norm": 55.21884536743164,
+      "learning_rate": 2.266097424412094e-05,
+      "loss": 16.6007,
+      "step": 16200
+    },
+    {
+      "epoch": 9.000626399776035,
+      "grad_norm": 55.47819137573242,
+      "learning_rate": 2.280095184770437e-05,
+      "loss": 16.6149,
+      "step": 16300
+    },
+    {
+      "epoch": 9.000906354983202,
+      "grad_norm": 53.711814880371094,
+      "learning_rate": 2.2940929451287796e-05,
+      "loss": 16.6412,
+      "step": 16400
+    },
+    {
+      "epoch": 9.00118631019037,
+      "grad_norm": 54.73662185668945,
+      "learning_rate": 2.308090705487122e-05,
+      "loss": 16.2739,
+      "step": 16500
+    },
+    {
+      "epoch": 9.001466265397536,
+      "grad_norm": 58.08010482788086,
+      "learning_rate": 2.3220884658454647e-05,
+      "loss": 16.38,
+      "step": 16600
+    },
+    {
+      "epoch": 9.001746220604703,
+      "grad_norm": 55.03322982788086,
+      "learning_rate": 2.3360862262038073e-05,
+      "loss": 16.4484,
+      "step": 16700
+    },
+    {
+      "epoch": 9.00202617581187,
+      "grad_norm": 57.478553771972656,
+      "learning_rate": 2.3500839865621502e-05,
+      "loss": 16.333,
+      "step": 16800
+    },
+    {
+      "epoch": 9.002306131019036,
+      "grad_norm": 56.241336822509766,
+      "learning_rate": 2.3640817469204928e-05,
+      "loss": 16.2893,
+      "step": 16900
+    },
+    {
+      "epoch": 9.002586086226204,
+      "grad_norm": 54.41828536987305,
+      "learning_rate": 2.3780795072788357e-05,
+      "loss": 16.3389,
+      "step": 17000
+    },
+    {
+      "epoch": 9.002866041433371,
+      "grad_norm": 55.62673568725586,
+      "learning_rate": 2.392077267637178e-05,
+      "loss": 16.1732,
+      "step": 17100
+    },
+    {
+      "epoch": 9.003145996640537,
+      "grad_norm": 57.23749542236328,
+      "learning_rate": 2.406075027995521e-05,
+      "loss": 16.371,
+      "step": 17200
+    },
+    {
+      "epoch": 9.003425951847705,
+      "grad_norm": 54.90141677856445,
+      "learning_rate": 2.4200727883538635e-05,
+      "loss": 16.3029,
+      "step": 17300
+    },
+    {
+      "epoch": 9.00370590705487,
+      "grad_norm": 53.32387924194336,
+      "learning_rate": 2.434070548712206e-05,
+      "loss": 15.9532,
+      "step": 17400
+    },
+    {
+      "epoch": 9.003985862262038,
+      "grad_norm": 57.22005844116211,
+      "learning_rate": 2.448068309070549e-05,
+      "loss": 16.1949,
+      "step": 17500
+    },
+    {
+      "epoch": 9.004265817469205,
+      "grad_norm": 53.49818420410156,
+      "learning_rate": 2.4620660694288916e-05,
+      "loss": 16.0044,
+      "step": 17600
+    },
+    {
+      "epoch": 9.004545772676371,
+      "grad_norm": 56.9859619140625,
+      "learning_rate": 2.476063829787234e-05,
+      "loss": 15.9484,
+      "step": 17700
+    },
+    {
+      "epoch": 9.004825727883539,
+      "grad_norm": 53.9587516784668,
+      "learning_rate": 2.4900615901455767e-05,
+      "loss": 15.752,
+      "step": 17800
+    },
+    {
+      "epoch": 9.004999300111981,
+      "eval_accuracy": 0.16189989785495404,
+      "eval_f1": 0.10126287636515371,
+      "eval_loss": 5.050577163696289,
+      "eval_precision": 0.0953715780728942,
+      "eval_recall": 0.16189989785495404,
+      "eval_runtime": 231.5088,
+      "eval_samples_per_second": 16.915,
+      "eval_steps_per_second": 8.458,
+      "eval_top_10_accuracy": 0.4846782431052094,
+      "eval_top_1_accuracy": 0.16215526046986722,
+      "eval_top_5_accuracy": 0.359805924412666,
+      "step": 17862
+    },
+    {
+      "epoch": 10.000104983202688,
+      "grad_norm": 56.3593635559082,
+      "learning_rate": 2.503919372900336e-05,
+      "loss": 15.1907,
+      "step": 17900
+    },
+    {
+      "epoch": 10.000384938409855,
+      "grad_norm": 53.651336669921875,
+      "learning_rate": 2.5179171332586787e-05,
+      "loss": 13.7771,
+      "step": 18000
+    },
+    {
+      "epoch": 10.00066489361702,
+      "grad_norm": 58.42094421386719,
+      "learning_rate": 2.5319148936170213e-05,
+      "loss": 13.9784,
+      "step": 18100
+    },
+    {
+      "epoch": 10.000944848824188,
+      "grad_norm": 55.46380615234375,
+      "learning_rate": 2.5459126539753642e-05,
+      "loss": 13.5654,
+      "step": 18200
+    },
+    {
+      "epoch": 10.001224804031356,
+      "grad_norm": 58.18601608276367,
+      "learning_rate": 2.5599104143337065e-05,
+      "loss": 13.6288,
+      "step": 18300
+    },
+    {
+      "epoch": 10.001504759238522,
+      "grad_norm": 58.32529067993164,
+      "learning_rate": 2.5739081746920497e-05,
+      "loss": 13.7221,
+      "step": 18400
+    },
+    {
+      "epoch": 10.001784714445689,
+      "grad_norm": 56.736839294433594,
+      "learning_rate": 2.587905935050392e-05,
+      "loss": 13.6705,
+      "step": 18500
+    },
+    {
+      "epoch": 10.002064669652855,
+      "grad_norm": 58.743064880371094,
+      "learning_rate": 2.6019036954087345e-05,
+      "loss": 13.9223,
+      "step": 18600
+    },
+    {
+      "epoch": 10.002344624860022,
+      "grad_norm": 55.71434783935547,
+      "learning_rate": 2.6159014557670775e-05,
+      "loss": 13.5166,
+      "step": 18700
+    },
+    {
+      "epoch": 10.00262458006719,
+      "grad_norm": 51.18729019165039,
+      "learning_rate": 2.62989921612542e-05,
+      "loss": 13.3252,
+      "step": 18800
+    },
+    {
+      "epoch": 10.002904535274356,
+      "grad_norm": 56.28409194946289,
+      "learning_rate": 2.643896976483763e-05,
+      "loss": 13.422,
+      "step": 18900
+    },
+    {
+      "epoch": 10.003184490481523,
+      "grad_norm": 56.53131866455078,
+      "learning_rate": 2.6578947368421052e-05,
+      "loss": 13.6352,
+      "step": 19000
+    },
+    {
+      "epoch": 10.00346444568869,
+      "grad_norm": 54.37139892578125,
+      "learning_rate": 2.6718924972004485e-05,
+      "loss": 13.7026,
+      "step": 19100
+    },
+    {
+      "epoch": 10.003744400895856,
+      "grad_norm": 54.172210693359375,
+      "learning_rate": 2.6858902575587907e-05,
+      "loss": 13.3519,
+      "step": 19200
+    },
+    {
+      "epoch": 10.004024356103024,
+      "grad_norm": 59.11737823486328,
+      "learning_rate": 2.6998880179171333e-05,
+      "loss": 13.4746,
+      "step": 19300
+    },
+    {
+      "epoch": 10.00430431131019,
+      "grad_norm": 60.201297760009766,
+      "learning_rate": 2.7138857782754762e-05,
+      "loss": 13.135,
+      "step": 19400
+    },
+    {
+      "epoch": 10.004584266517357,
+      "grad_norm": 62.821327209472656,
+      "learning_rate": 2.7278835386338188e-05,
+      "loss": 13.5812,
+      "step": 19500
+    },
+    {
+      "epoch": 10.004864221724524,
+      "grad_norm": 56.00190734863281,
+      "learning_rate": 2.7418812989921617e-05,
+      "loss": 13.1613,
+      "step": 19600
+    },
+    {
+      "epoch": 10.004998600223963,
+      "eval_accuracy": 0.1940755873340143,
+      "eval_f1": 0.13521619871280463,
+      "eval_loss": 4.658937454223633,
+      "eval_precision": 0.1293644795170035,
+      "eval_recall": 0.1940755873340143,
+      "eval_runtime": 218.1856,
+      "eval_samples_per_second": 17.948,
+      "eval_steps_per_second": 8.974,
+      "eval_top_10_accuracy": 0.556435137895812,
+      "eval_top_1_accuracy": 0.19330949948927478,
+      "eval_top_5_accuracy": 0.432073544433095,
+      "step": 19648
+    },
+    {
+      "epoch": 11.000143477043673,
+      "grad_norm": 50.50157928466797,
+      "learning_rate": 2.755879059350504e-05,
+      "loss": 12.0765,
+      "step": 19700
+    },
+    {
+      "epoch": 11.00042343225084,
+      "grad_norm": 60.919158935546875,
+      "learning_rate": 2.7698768197088465e-05,
+      "loss": 10.9856,
+      "step": 19800
+    },
+    {
+      "epoch": 11.000703387458007,
+      "grad_norm": 55.98445510864258,
+      "learning_rate": 2.7838745800671895e-05,
+      "loss": 11.1553,
+      "step": 19900
+    },
+    {
+      "epoch": 11.000983342665174,
+      "grad_norm": 53.11532211303711,
+      "learning_rate": 2.797872340425532e-05,
+      "loss": 10.8397,
+      "step": 20000
+    },
+    {
+      "epoch": 11.00126329787234,
+      "grad_norm": 58.633155822753906,
+      "learning_rate": 2.811870100783875e-05,
+      "loss": 11.1483,
+      "step": 20100
+    },
+    {
+      "epoch": 11.001543253079507,
+      "grad_norm": 55.615665435791016,
+      "learning_rate": 2.8258678611422172e-05,
+      "loss": 11.0123,
+      "step": 20200
+    },
+    {
+      "epoch": 11.001823208286675,
+      "grad_norm": 57.123165130615234,
+      "learning_rate": 2.8398656215005598e-05,
+      "loss": 10.5229,
+      "step": 20300
+    },
+    {
+      "epoch": 11.00210316349384,
+      "grad_norm": 51.48651885986328,
+      "learning_rate": 2.8538633818589027e-05,
+      "loss": 11.0409,
+      "step": 20400
+    },
+    {
+      "epoch": 11.002383118701008,
+      "grad_norm": 59.33769226074219,
+      "learning_rate": 2.8678611422172453e-05,
+      "loss": 11.1909,
+      "step": 20500
+    },
+    {
+      "epoch": 11.002663073908174,
+      "grad_norm": 56.830570220947266,
+      "learning_rate": 2.8818589025755882e-05,
+      "loss": 11.045,
+      "step": 20600
+    },
+    {
+      "epoch": 11.002943029115341,
+      "grad_norm": 59.105621337890625,
+      "learning_rate": 2.8958566629339308e-05,
+      "loss": 11.0254,
+      "step": 20700
+    },
+    {
+      "epoch": 11.003222984322509,
+      "grad_norm": 57.58058547973633,
+      "learning_rate": 2.9098544232922737e-05,
+      "loss": 11.1625,
+      "step": 20800
+    },
+    {
+      "epoch": 11.003502939529675,
+      "grad_norm": 60.10997009277344,
+      "learning_rate": 2.923852183650616e-05,
+      "loss": 10.6298,
+      "step": 20900
+    },
+    {
+      "epoch": 11.003782894736842,
+      "grad_norm": 54.05520248413086,
+      "learning_rate": 2.9378499440089585e-05,
+      "loss": 10.8325,
+      "step": 21000
+    },
+    {
+      "epoch": 11.00406284994401,
+      "grad_norm": 50.31863784790039,
+      "learning_rate": 2.9518477043673014e-05,
+      "loss": 10.8705,
+      "step": 21100
+    },
+    {
+      "epoch": 11.004342805151175,
+      "grad_norm": 58.63640594482422,
+      "learning_rate": 2.965845464725644e-05,
+      "loss": 10.7095,
+      "step": 21200
+    },
+    {
+      "epoch": 11.004622760358343,
+      "grad_norm": 56.01993179321289,
+      "learning_rate": 2.979703247480403e-05,
+      "loss": 10.4509,
+      "step": 21300
+    },
+    {
+      "epoch": 11.00490271556551,
+      "grad_norm": 61.992340087890625,
+      "learning_rate": 2.9937010078387464e-05,
+      "loss": 10.1883,
+      "step": 21400
+    },
+    {
+      "epoch": 11.005000699888019,
+      "eval_accuracy": 0.22625127681307455,
+      "eval_f1": 0.16842648729319404,
+      "eval_loss": 4.3067474365234375,
+      "eval_precision": 0.16332781901519114,
+      "eval_recall": 0.22625127681307455,
+      "eval_runtime": 226.1299,
+      "eval_samples_per_second": 17.317,
+      "eval_steps_per_second": 8.659,
+      "eval_top_10_accuracy": 0.6138917262512769,
+      "eval_top_1_accuracy": 0.22625127681307455,
+      "eval_top_5_accuracy": 0.4946373850868233,
+      "step": 21435
+    },
+    {
+      "epoch": 12.00018197088466,
+      "grad_norm": 44.793460845947266,
+      "learning_rate": 3.0076987681970886e-05,
+      "loss": 9.2143,
+      "step": 21500
+    },
+    {
+      "epoch": 12.000461926091825,
+      "grad_norm": 53.28424072265625,
+      "learning_rate": 3.0216965285554312e-05,
+      "loss": 8.0788,
+      "step": 21600
+    },
+    {
+      "epoch": 12.000741881298993,
+      "grad_norm": 56.62223434448242,
+      "learning_rate": 3.035694288913774e-05,
+      "loss": 8.3003,
+      "step": 21700
+    },
+    {
+      "epoch": 12.001021836506158,
+      "grad_norm": 50.635013580322266,
+      "learning_rate": 3.0496920492721163e-05,
+      "loss": 8.0193,
+      "step": 21800
+    },
+    {
+      "epoch": 12.001301791713326,
+      "grad_norm": 46.23341751098633,
+      "learning_rate": 3.0636898096304596e-05,
+      "loss": 8.2961,
+      "step": 21900
+    },
+    {
+      "epoch": 12.001581746920493,
+      "grad_norm": 61.92805862426758,
+      "learning_rate": 3.077687569988802e-05,
+      "loss": 8.3955,
+      "step": 22000
+    },
+    {
+      "epoch": 12.001861702127659,
+      "grad_norm": 59.257286071777344,
+      "learning_rate": 3.091685330347144e-05,
+      "loss": 8.398,
+      "step": 22100
+    },
+    {
+      "epoch": 12.002141657334827,
+      "grad_norm": 58.87281799316406,
+      "learning_rate": 3.1056830907054873e-05,
+      "loss": 8.2248,
+      "step": 22200
+    },
+    {
+      "epoch": 12.002421612541994,
+      "grad_norm": 60.63299560546875,
+      "learning_rate": 3.11968085106383e-05,
+      "loss": 8.4425,
+      "step": 22300
+    },
+    {
+      "epoch": 12.00270156774916,
+      "grad_norm": 60.67166519165039,
+      "learning_rate": 3.1336786114221725e-05,
+      "loss": 8.2765,
+      "step": 22400
+    },
+    {
+      "epoch": 12.002981522956327,
+      "grad_norm": 53.16400146484375,
+      "learning_rate": 3.147676371780515e-05,
+      "loss": 8.1763,
+      "step": 22500
+    },
+    {
+      "epoch": 12.003261478163493,
+      "grad_norm": 56.289981842041016,
+      "learning_rate": 3.161674132138858e-05,
+      "loss": 8.4768,
+      "step": 22600
+    },
+    {
+      "epoch": 12.00354143337066,
+      "grad_norm": 59.1424560546875,
+      "learning_rate": 3.175671892497201e-05,
+      "loss": 8.1219,
+      "step": 22700
+    },
+    {
+      "epoch": 12.003821388577828,
+      "grad_norm": 65.8774642944336,
+      "learning_rate": 3.189669652855543e-05,
+      "loss": 8.3595,
+      "step": 22800
+    },
+    {
+      "epoch": 12.004101343784994,
+      "grad_norm": 62.18092727661133,
+      "learning_rate": 3.203667413213886e-05,
+      "loss": 8.351,
+      "step": 22900
+    },
+    {
+      "epoch": 12.004381298992161,
+      "grad_norm": 58.203758239746094,
+      "learning_rate": 3.217665173572229e-05,
+      "loss": 8.0542,
+      "step": 23000
+    },
+    {
+      "epoch": 12.004661254199329,
+      "grad_norm": 57.017539978027344,
+      "learning_rate": 3.231662933930571e-05,
+      "loss": 8.0881,
+      "step": 23100
+    },
+    {
+      "epoch": 12.004941209406494,
+      "grad_norm": 60.21388244628906,
+      "learning_rate": 3.245660694288914e-05,
+      "loss": 8.1852,
+      "step": 23200
+    },
+    {
+      "epoch": 12.005,
+      "eval_accuracy": 0.24719101123595505,
+      "eval_f1": 0.19268227724832526,
+      "eval_loss": 4.012407302856445,
+      "eval_precision": 0.18863082228029204,
+      "eval_recall": 0.24719101123595505,
+      "eval_runtime": 226.6718,
+      "eval_samples_per_second": 17.276,
+      "eval_steps_per_second": 8.638,
+      "eval_top_10_accuracy": 0.6527068437180796,
+      "eval_top_1_accuracy": 0.24744637385086823,
+      "eval_top_5_accuracy": 0.5314096016343207,
+      "step": 23221
+    },
+    {
+      "epoch": 13.000220464725643,
+      "grad_norm": 40.455081939697266,
+      "learning_rate": 3.2596584546472564e-05,
+      "loss": 6.7682,
+      "step": 23300
+    },
+    {
+      "epoch": 13.000500419932811,
+      "grad_norm": 41.39524841308594,
+      "learning_rate": 3.2736562150056e-05,
+      "loss": 5.8189,
+      "step": 23400
+    },
+    {
+      "epoch": 13.000780375139978,
+      "grad_norm": 58.26043701171875,
+      "learning_rate": 3.2876539753639416e-05,
+      "loss": 6.1814,
+      "step": 23500
+    },
+    {
+      "epoch": 13.001060330347144,
+      "grad_norm": 50.91895294189453,
+      "learning_rate": 3.301511758118701e-05,
+      "loss": 6.0026,
+      "step": 23600
+    },
+    {
+      "epoch": 13.001340285554312,
+      "grad_norm": 63.752769470214844,
+      "learning_rate": 3.315509518477044e-05,
+      "loss": 5.9956,
+      "step": 23700
+    },
+    {
+      "epoch": 13.001620240761477,
+      "grad_norm": 51.161293029785156,
+      "learning_rate": 3.329507278835387e-05,
+      "loss": 5.9912,
+      "step": 23800
+    },
+    {
+      "epoch": 13.001900195968645,
+      "grad_norm": 48.86064147949219,
+      "learning_rate": 3.343505039193729e-05,
+      "loss": 5.9778,
+      "step": 23900
+    },
+    {
+      "epoch": 13.002180151175812,
+      "grad_norm": 57.04310607910156,
+      "learning_rate": 3.357502799552072e-05,
+      "loss": 6.0001,
+      "step": 24000
+    },
+    {
+      "epoch": 13.002460106382978,
+      "grad_norm": 46.83991241455078,
+      "learning_rate": 3.3715005599104146e-05,
+      "loss": 5.8923,
+      "step": 24100
+    },
+    {
+      "epoch": 13.002740061590146,
+      "grad_norm": 59.89580154418945,
+      "learning_rate": 3.385498320268757e-05,
+      "loss": 6.0239,
+      "step": 24200
+    },
+    {
+      "epoch": 13.003020016797313,
+      "grad_norm": 58.387611389160156,
+      "learning_rate": 3.3994960806271e-05,
+      "loss": 6.1917,
+      "step": 24300
+    },
+    {
+      "epoch": 13.003299972004479,
+      "grad_norm": 59.23344802856445,
+      "learning_rate": 3.413493840985442e-05,
+      "loss": 5.6954,
+      "step": 24400
+    },
+    {
+      "epoch": 13.003579927211646,
+      "grad_norm": 63.508567810058594,
+      "learning_rate": 3.427491601343785e-05,
+      "loss": 5.912,
+      "step": 24500
+    },
+    {
+      "epoch": 13.003859882418814,
+      "grad_norm": 58.91965103149414,
+      "learning_rate": 3.4414893617021275e-05,
+      "loss": 6.0244,
+      "step": 24600
+    },
+    {
+      "epoch": 13.00413983762598,
+      "grad_norm": 57.9660530090332,
+      "learning_rate": 3.455487122060471e-05,
+      "loss": 5.9531,
+      "step": 24700
+    },
+    {
+      "epoch": 13.004419792833147,
+      "grad_norm": 51.39337921142578,
+      "learning_rate": 3.469484882418813e-05,
+      "loss": 5.8885,
+      "step": 24800
+    },
+    {
+      "epoch": 13.004699748040313,
+      "grad_norm": 43.0623779296875,
+      "learning_rate": 3.483482642777156e-05,
+      "loss": 5.954,
+      "step": 24900
+    },
+    {
+      "epoch": 13.00497970324748,
+      "grad_norm": 62.9230842590332,
+      "learning_rate": 3.4974804031354985e-05,
+      "loss": 5.88,
+      "step": 25000
+    },
+    {
+      "epoch": 13.004999300111981,
+      "eval_accuracy": 0.26506639427987744,
+      "eval_f1": 0.21861788640092267,
+      "eval_loss": 3.7566869258880615,
+      "eval_precision": 0.21875192535304894,
+      "eval_recall": 0.26506639427987744,
+      "eval_runtime": 227.0931,
+      "eval_samples_per_second": 17.244,
+      "eval_steps_per_second": 8.622,
+      "eval_top_10_accuracy": 0.6884576098059244,
+      "eval_top_1_accuracy": 0.26506639427987744,
+      "eval_top_5_accuracy": 0.5720122574055159,
+      "step": 25007
+    },
+    {
+      "epoch": 14.00025895856663,
+      "grad_norm": 33.66314697265625,
+      "learning_rate": 3.511478163493841e-05,
+      "loss": 4.0538,
+      "step": 25100
+    },
+    {
+      "epoch": 14.000538913773797,
+      "grad_norm": 51.570316314697266,
+      "learning_rate": 3.5254759238521836e-05,
+      "loss": 3.9683,
+      "step": 25200
+    },
+    {
+      "epoch": 14.000818868980963,
+      "grad_norm": 39.41959762573242,
+      "learning_rate": 3.539473684210526e-05,
+      "loss": 4.1115,
+      "step": 25300
+    },
+    {
+      "epoch": 14.00109882418813,
+      "grad_norm": 46.20742416381836,
+      "learning_rate": 3.5534714445688695e-05,
+      "loss": 4.2485,
+      "step": 25400
+    },
+    {
+      "epoch": 14.001378779395298,
+      "grad_norm": 32.3474235534668,
+      "learning_rate": 3.567469204927212e-05,
+      "loss": 4.0841,
+      "step": 25500
+    },
+    {
+      "epoch": 14.001658734602463,
+      "grad_norm": 50.43800735473633,
+      "learning_rate": 3.581326987681971e-05,
+      "loss": 3.9711,
+      "step": 25600
+    },
+    {
+      "epoch": 14.00193868980963,
+      "grad_norm": 42.714744567871094,
+      "learning_rate": 3.5953247480403134e-05,
+      "loss": 4.0905,
+      "step": 25700
+    },
+    {
+      "epoch": 14.002218645016796,
+      "grad_norm": 50.72501754760742,
+      "learning_rate": 3.6093225083986566e-05,
+      "loss": 4.2036,
+      "step": 25800
+    },
+    {
+      "epoch": 14.002498600223964,
+      "grad_norm": 44.00167465209961,
+      "learning_rate": 3.623320268756999e-05,
+      "loss": 3.9163,
+      "step": 25900
+    },
+    {
+      "epoch": 14.002778555431131,
+      "grad_norm": 36.54977798461914,
+      "learning_rate": 3.637318029115342e-05,
+      "loss": 4.1644,
+      "step": 26000
+    },
+    {
+      "epoch": 14.003058510638297,
+      "grad_norm": 31.148479461669922,
+      "learning_rate": 3.6513157894736844e-05,
+      "loss": 4.1178,
+      "step": 26100
+    },
+    {
+      "epoch": 14.003338465845465,
+      "grad_norm": 50.2994384765625,
+      "learning_rate": 3.665313549832027e-05,
+      "loss": 3.9082,
+      "step": 26200
+    },
+    {
+      "epoch": 14.003618421052632,
+      "grad_norm": 53.5435905456543,
+      "learning_rate": 3.6793113101903695e-05,
+      "loss": 4.2392,
+      "step": 26300
+    },
+    {
+      "epoch": 14.003898376259798,
+      "grad_norm": 32.26350402832031,
+      "learning_rate": 3.693309070548712e-05,
+      "loss": 4.026,
+      "step": 26400
+    },
+    {
+      "epoch": 14.004178331466965,
+      "grad_norm": 51.61219024658203,
+      "learning_rate": 3.7073068309070554e-05,
+      "loss": 4.1701,
+      "step": 26500
+    },
+    {
+      "epoch": 14.004458286674133,
+      "grad_norm": 60.40871810913086,
+      "learning_rate": 3.721304591265398e-05,
+      "loss": 4.1327,
+      "step": 26600
+    },
+    {
+      "epoch": 14.004738241881299,
+      "grad_norm": 55.58240509033203,
+      "learning_rate": 3.7353023516237405e-05,
+      "loss": 4.5173,
+      "step": 26700
+    },
+    {
+      "epoch": 14.004998600223963,
+      "eval_accuracy": 0.2763023493360572,
+      "eval_f1": 0.2372876851202756,
+      "eval_loss": 3.6064484119415283,
+      "eval_precision": 0.24398663297186698,
+      "eval_recall": 0.2763023493360572,
+      "eval_runtime": 223.8625,
+      "eval_samples_per_second": 17.493,
+      "eval_steps_per_second": 8.746,
+      "eval_top_10_accuracy": 0.7124616956077631,
+      "eval_top_1_accuracy": 0.2763023493360572,
+      "eval_top_5_accuracy": 0.5926966292134831,
+      "step": 26793
+    },
+    {
+      "epoch": 15.000017497200448,
+      "grad_norm": 46.20799255371094,
+      "learning_rate": 3.749300111982083e-05,
+      "loss": 4.0528,
+      "step": 26800
+    },
+    {
+      "epoch": 15.000297452407615,
+      "grad_norm": 52.91650390625,
+      "learning_rate": 3.763297872340426e-05,
+      "loss": 2.4598,
+      "step": 26900
+    },
+    {
+      "epoch": 15.00057740761478,
+      "grad_norm": 29.833009719848633,
+      "learning_rate": 3.777295632698768e-05,
+      "loss": 2.5363,
+      "step": 27000
+    },
+    {
+      "epoch": 15.000857362821948,
+      "grad_norm": 42.154937744140625,
+      "learning_rate": 3.791293393057111e-05,
+      "loss": 2.4495,
+      "step": 27100
+    },
+    {
+      "epoch": 15.001137318029116,
+      "grad_norm": 40.74216079711914,
+      "learning_rate": 3.805291153415454e-05,
+      "loss": 2.5793,
+      "step": 27200
+    },
+    {
+      "epoch": 15.001417273236282,
+      "grad_norm": 41.93460464477539,
+      "learning_rate": 3.819288913773797e-05,
+      "loss": 2.6526,
+      "step": 27300
+    },
+    {
+      "epoch": 15.001697228443449,
+      "grad_norm": 29.095109939575195,
+      "learning_rate": 3.8332866741321386e-05,
+      "loss": 2.7321,
+      "step": 27400
+    },
+    {
+      "epoch": 15.001977183650617,
+      "grad_norm": 46.61332702636719,
+      "learning_rate": 3.847284434490482e-05,
+      "loss": 2.8783,
+      "step": 27500
+    },
+    {
+      "epoch": 15.002257138857782,
+      "grad_norm": 47.55717086791992,
+      "learning_rate": 3.8612821948488244e-05,
+      "loss": 3.0291,
+      "step": 27600
+    },
+    {
+      "epoch": 15.00253709406495,
+      "grad_norm": 35.126800537109375,
+      "learning_rate": 3.875279955207167e-05,
+      "loss": 2.8221,
+      "step": 27700
+    },
+    {
+      "epoch": 15.002817049272117,
+      "grad_norm": 33.11594772338867,
+      "learning_rate": 3.8891377379619264e-05,
+      "loss": 2.8732,
+      "step": 27800
+    },
+    {
+      "epoch": 15.003097004479283,
+      "grad_norm": 39.2805290222168,
+      "learning_rate": 3.903135498320269e-05,
+      "loss": 2.6688,
+      "step": 27900
+    },
+    {
+      "epoch": 15.00337695968645,
+      "grad_norm": 47.020259857177734,
+      "learning_rate": 3.9171332586786116e-05,
+      "loss": 2.8454,
+      "step": 28000
+    },
+    {
+      "epoch": 15.003656914893616,
+      "grad_norm": 28.50092124938965,
+      "learning_rate": 3.931131019036954e-05,
+      "loss": 2.9415,
+      "step": 28100
+    },
+    {
+      "epoch": 15.003936870100784,
+      "grad_norm": 51.885154724121094,
+      "learning_rate": 3.945128779395297e-05,
+      "loss": 2.94,
+      "step": 28200
+    },
+    {
+      "epoch": 15.004216825307951,
+      "grad_norm": 37.81270217895508,
+      "learning_rate": 3.95912653975364e-05,
+      "loss": 2.8525,
+      "step": 28300
+    },
+    {
+      "epoch": 15.004496780515117,
+      "grad_norm": 38.692378997802734,
+      "learning_rate": 3.973124300111982e-05,
+      "loss": 2.9696,
+      "step": 28400
+    },
+    {
+      "epoch": 15.004776735722285,
+      "grad_norm": 52.6309928894043,
+      "learning_rate": 3.987122060470325e-05,
+      "loss": 2.851,
+      "step": 28500
+    },
+    {
+      "epoch": 15.005000699888019,
+      "eval_accuracy": 0.27783452502553624,
+      "eval_f1": 0.2434374171463338,
+      "eval_loss": 3.5264151096343994,
+      "eval_precision": 0.25338318332444987,
+      "eval_recall": 0.27783452502553624,
+      "eval_runtime": 226.3306,
+      "eval_samples_per_second": 17.302,
+      "eval_steps_per_second": 8.651,
+      "eval_top_10_accuracy": 0.719611848825332,
+      "eval_top_1_accuracy": 0.2775791624106231,
+      "eval_top_5_accuracy": 0.608273748723187,
+      "step": 28580
+    },
+    {
+      "epoch": 16.00005599104143,
+      "grad_norm": 35.61888885498047,
+      "learning_rate": 4.001119820828668e-05,
+      "loss": 2.6647,
+      "step": 28600
+    },
+    {
+      "epoch": 16.0003359462486,
+      "grad_norm": 33.327659606933594,
+      "learning_rate": 4.0151175811870103e-05,
+      "loss": 1.6989,
+      "step": 28700
+    },
+    {
+      "epoch": 16.000615901455767,
+      "grad_norm": 18.319982528686523,
+      "learning_rate": 4.029115341545353e-05,
+      "loss": 1.6956,
+      "step": 28800
+    },
+    {
+      "epoch": 16.000895856662932,
+      "grad_norm": 25.141307830810547,
+      "learning_rate": 4.0431131019036955e-05,
+      "loss": 1.715,
+      "step": 28900
+    },
+    {
+      "epoch": 16.0011758118701,
+      "grad_norm": 43.47077560424805,
+      "learning_rate": 4.057110862262039e-05,
+      "loss": 2.0691,
+      "step": 29000
+    },
+    {
+      "epoch": 16.001455767077267,
+      "grad_norm": 16.283327102661133,
+      "learning_rate": 4.071108622620381e-05,
+      "loss": 1.8248,
+      "step": 29100
+    },
+    {
+      "epoch": 16.001735722284433,
+      "grad_norm": 35.769107818603516,
+      "learning_rate": 4.085106382978723e-05,
+      "loss": 2.0076,
+      "step": 29200
+    },
+    {
+      "epoch": 16.002015677491602,
+      "grad_norm": 49.17667770385742,
+      "learning_rate": 4.0991041433370665e-05,
+      "loss": 2.0197,
+      "step": 29300
+    },
+    {
+      "epoch": 16.002295632698768,
+      "grad_norm": 57.587589263916016,
+      "learning_rate": 4.113101903695409e-05,
+      "loss": 1.9761,
+      "step": 29400
+    },
+    {
+      "epoch": 16.002575587905934,
+      "grad_norm": 49.32463836669922,
+      "learning_rate": 4.127099664053752e-05,
+      "loss": 1.9658,
+      "step": 29500
+    },
+    {
+      "epoch": 16.002855543113103,
+      "grad_norm": 38.368656158447266,
+      "learning_rate": 4.141097424412094e-05,
+      "loss": 2.1054,
+      "step": 29600
+    },
+    {
+      "epoch": 16.00313549832027,
+      "grad_norm": 43.137908935546875,
+      "learning_rate": 4.155095184770437e-05,
+      "loss": 2.1054,
+      "step": 29700
+    },
+    {
+      "epoch": 16.003415453527435,
+      "grad_norm": 34.75947952270508,
+      "learning_rate": 4.1690929451287794e-05,
+      "loss": 2.1547,
+      "step": 29800
+    },
+    {
+      "epoch": 16.003695408734604,
+      "grad_norm": 55.931907653808594,
+      "learning_rate": 4.183090705487122e-05,
+      "loss": 2.0834,
+      "step": 29900
+    },
+    {
+      "epoch": 16.00397536394177,
+      "grad_norm": 22.097393035888672,
+      "learning_rate": 4.197088465845465e-05,
+      "loss": 1.9993,
+      "step": 30000
+    },
+    {
+      "epoch": 16.004255319148935,
+      "grad_norm": 36.39588928222656,
+      "learning_rate": 4.211086226203808e-05,
+      "loss": 2.208,
+      "step": 30100
+    },
+    {
+      "epoch": 16.004535274356105,
+      "grad_norm": 25.969524383544922,
+      "learning_rate": 4.2249440089585666e-05,
+      "loss": 2.025,
+      "step": 30200
+    },
+    {
+      "epoch": 16.00481522956327,
+      "grad_norm": 30.56204605102539,
+      "learning_rate": 4.238941769316909e-05,
+      "loss": 2.2493,
+      "step": 30300
+    },
+    {
+      "epoch": 16.005,
+      "eval_accuracy": 0.27247191011235955,
+      "eval_f1": 0.23831057018449936,
+      "eval_loss": 3.500147581100464,
+      "eval_precision": 0.24710587923462707,
+      "eval_recall": 0.27247191011235955,
+      "eval_runtime": 226.3004,
+      "eval_samples_per_second": 17.304,
+      "eval_steps_per_second": 8.652,
+      "eval_top_10_accuracy": 0.7183350357507661,
+      "eval_top_1_accuracy": 0.27247191011235955,
+      "eval_top_5_accuracy": 0.6085291113381001,
+      "step": 30366
+    },
+    {
+      "epoch": 17.000094484882418,
+      "grad_norm": 9.153155326843262,
+      "learning_rate": 4.2529395296752524e-05,
+      "loss": 1.7014,
+      "step": 30400
+    },
+    {
+      "epoch": 17.000374440089587,
+      "grad_norm": 23.366758346557617,
+      "learning_rate": 4.266937290033595e-05,
+      "loss": 1.2216,
+      "step": 30500
+    },
+    {
+      "epoch": 17.000654395296753,
+      "grad_norm": 41.90321350097656,
+      "learning_rate": 4.2809350503919376e-05,
+      "loss": 1.3224,
+      "step": 30600
+    },
+    {
+      "epoch": 17.00093435050392,
+      "grad_norm": 16.407875061035156,
+      "learning_rate": 4.29493281075028e-05,
+      "loss": 1.3376,
+      "step": 30700
+    },
+    {
+      "epoch": 17.001214305711088,
+      "grad_norm": 16.32012176513672,
+      "learning_rate": 4.3089305711086234e-05,
+      "loss": 1.2108,
+      "step": 30800
+    },
+    {
+      "epoch": 17.001494260918253,
+      "grad_norm": 38.008888244628906,
+      "learning_rate": 4.322928331466965e-05,
+      "loss": 1.5091,
+      "step": 30900
+    },
+    {
+      "epoch": 17.00177421612542,
+      "grad_norm": 20.518709182739258,
+      "learning_rate": 4.336926091825308e-05,
+      "loss": 1.3844,
+      "step": 31000
+    },
+    {
+      "epoch": 17.00205417133259,
+      "grad_norm": 46.348655700683594,
+      "learning_rate": 4.350923852183651e-05,
+      "loss": 1.6624,
+      "step": 31100
+    },
+    {
+      "epoch": 17.002334126539754,
+      "grad_norm": 32.697288513183594,
+      "learning_rate": 4.364921612541993e-05,
+      "loss": 1.4262,
+      "step": 31200
+    },
+    {
+      "epoch": 17.00261408174692,
+      "grad_norm": 27.621122360229492,
+      "learning_rate": 4.378919372900336e-05,
+      "loss": 1.6205,
+      "step": 31300
+    },
+    {
+      "epoch": 17.00289403695409,
+      "grad_norm": 32.8564338684082,
+      "learning_rate": 4.392917133258679e-05,
+      "loss": 1.7276,
+      "step": 31400
+    },
+    {
+      "epoch": 17.003173992161255,
+      "grad_norm": 32.846004486083984,
+      "learning_rate": 4.4069148936170215e-05,
+      "loss": 1.5652,
+      "step": 31500
+    },
+    {
+      "epoch": 17.00345394736842,
+      "grad_norm": 43.118534088134766,
+      "learning_rate": 4.420912653975364e-05,
+      "loss": 1.7943,
+      "step": 31600
+    },
+    {
+      "epoch": 17.003733902575586,
+      "grad_norm": 37.35938262939453,
+      "learning_rate": 4.4349104143337066e-05,
+      "loss": 1.5109,
+      "step": 31700
+    },
+    {
+      "epoch": 17.004013857782756,
+      "grad_norm": 16.283411026000977,
+      "learning_rate": 4.44890817469205e-05,
+      "loss": 1.7202,
+      "step": 31800
+    },
+    {
+      "epoch": 17.00429381298992,
+      "grad_norm": 57.66909408569336,
+      "learning_rate": 4.462905935050392e-05,
+      "loss": 2.0454,
+      "step": 31900
+    },
+    {
+      "epoch": 17.004573768197087,
+      "grad_norm": 30.193222045898438,
+      "learning_rate": 4.4769036954087344e-05,
+      "loss": 1.8866,
+      "step": 32000
+    },
+    {
+      "epoch": 17.004853723404256,
+      "grad_norm": 44.54677200317383,
+      "learning_rate": 4.4909014557670776e-05,
+      "loss": 1.9096,
+      "step": 32100
+    },
+    {
+      "epoch": 17.004999300111983,
+      "eval_accuracy": 0.2814096016343207,
+      "eval_f1": 0.243171253299741,
+      "eval_loss": 3.4932773113250732,
+      "eval_precision": 0.24868229744378875,
+      "eval_recall": 0.2814096016343207,
+      "eval_runtime": 226.1393,
+      "eval_samples_per_second": 17.317,
+      "eval_steps_per_second": 8.658,
+      "eval_top_10_accuracy": 0.7127170582226762,
+      "eval_top_1_accuracy": 0.2808988764044944,
+      "eval_top_5_accuracy": 0.6013789581205311,
+      "step": 32152
+    },
+    {
+      "epoch": 18.000132978723403,
+      "grad_norm": 37.95355224609375,
+      "learning_rate": 4.50489921612542e-05,
+      "loss": 1.5346,
+      "step": 32200
+    },
+    {
+      "epoch": 18.000412933930573,
+      "grad_norm": 15.054861068725586,
+      "learning_rate": 4.518896976483763e-05,
+      "loss": 0.9482,
+      "step": 32300
+    },
+    {
+      "epoch": 18.00069288913774,
+      "grad_norm": 40.022178649902344,
+      "learning_rate": 4.5328947368421054e-05,
+      "loss": 0.9161,
+      "step": 32400
+    },
+    {
+      "epoch": 18.000972844344904,
+      "grad_norm": 22.443218231201172,
+      "learning_rate": 4.5468924972004486e-05,
+      "loss": 1.1182,
+      "step": 32500
+    },
+    {
+      "epoch": 18.00125279955207,
+      "grad_norm": 51.938507080078125,
+      "learning_rate": 4.5608902575587905e-05,
+      "loss": 1.188,
+      "step": 32600
+    },
+    {
+      "epoch": 18.00153275475924,
+      "grad_norm": 17.9680118560791,
+      "learning_rate": 4.574888017917133e-05,
+      "loss": 1.1892,
+      "step": 32700
+    },
+    {
+      "epoch": 18.001812709966405,
+      "grad_norm": 16.573867797851562,
+      "learning_rate": 4.5888857782754764e-05,
+      "loss": 1.2303,
+      "step": 32800
+    },
+    {
+      "epoch": 18.00209266517357,
+      "grad_norm": 8.892263412475586,
+      "learning_rate": 4.602883538633819e-05,
+      "loss": 1.1585,
+      "step": 32900
+    },
+    {
+      "epoch": 18.00237262038074,
+      "grad_norm": 8.436017036437988,
+      "learning_rate": 4.6168812989921615e-05,
+      "loss": 1.2501,
+      "step": 33000
+    },
+    {
+      "epoch": 18.002652575587906,
+      "grad_norm": 52.22962188720703,
+      "learning_rate": 4.630879059350504e-05,
+      "loss": 1.3674,
+      "step": 33100
+    },
+    {
+      "epoch": 18.00293253079507,
+      "grad_norm": 9.320150375366211,
+      "learning_rate": 4.644876819708847e-05,
+      "loss": 1.3614,
+      "step": 33200
+    },
+    {
+      "epoch": 18.00321248600224,
+      "grad_norm": 11.164777755737305,
+      "learning_rate": 4.658874580067189e-05,
+      "loss": 1.5928,
+      "step": 33300
+    },
+    {
+      "epoch": 18.003492441209406,
+      "grad_norm": 39.596923828125,
+      "learning_rate": 4.672872340425532e-05,
+      "loss": 1.5899,
+      "step": 33400
+    },
+    {
+      "epoch": 18.003772396416572,
+      "grad_norm": 14.535404205322266,
+      "learning_rate": 4.686870100783875e-05,
+      "loss": 1.467,
+      "step": 33500
+    },
+    {
+      "epoch": 18.00405235162374,
+      "grad_norm": 20.554059982299805,
+      "learning_rate": 4.700867861142218e-05,
+      "loss": 1.5696,
+      "step": 33600
+    },
+    {
+      "epoch": 18.004332306830907,
+      "grad_norm": 52.87431335449219,
+      "learning_rate": 4.7147256438969764e-05,
+      "loss": 1.5463,
+      "step": 33700
+    },
+    {
+      "epoch": 18.004612262038073,
+      "grad_norm": 31.22138786315918,
+      "learning_rate": 4.728723404255319e-05,
+      "loss": 1.595,
+      "step": 33800
+    },
+    {
+      "epoch": 18.004892217245242,
+      "grad_norm": 50.54197311401367,
+      "learning_rate": 4.742721164613662e-05,
+      "loss": 1.7109,
+      "step": 33900
+    },
+    {
+      "epoch": 18.004998600223963,
+      "eval_accuracy": 0.27783452502553624,
+      "eval_f1": 0.24180912707733293,
+      "eval_loss": 3.517486810684204,
+      "eval_precision": 0.25168317098858467,
+      "eval_recall": 0.27783452502553624,
+      "eval_runtime": 227.7357,
+      "eval_samples_per_second": 17.195,
+      "eval_steps_per_second": 8.598,
+      "eval_top_10_accuracy": 0.7068437180796732,
+      "eval_top_1_accuracy": 0.2775791624106231,
+      "eval_top_5_accuracy": 0.5919305413687436,
+      "step": 33938
+    },
+    {
+      "epoch": 19.00017147256439,
+      "grad_norm": 46.21772003173828,
+      "learning_rate": 4.756718924972005e-05,
+      "loss": 1.2682,
+      "step": 34000
+    },
+    {
+      "epoch": 19.000451427771555,
+      "grad_norm": 4.63399600982666,
+      "learning_rate": 4.7707166853303474e-05,
+      "loss": 0.8455,
+      "step": 34100
+    },
+    {
+      "epoch": 19.000731382978724,
+      "grad_norm": 13.730817794799805,
+      "learning_rate": 4.78471444568869e-05,
+      "loss": 0.8606,
+      "step": 34200
+    },
+    {
+      "epoch": 19.00101133818589,
+      "grad_norm": 10.35786247253418,
+      "learning_rate": 4.798712206047033e-05,
+      "loss": 1.0884,
+      "step": 34300
+    },
+    {
+      "epoch": 19.001291293393056,
+      "grad_norm": 43.86582565307617,
+      "learning_rate": 4.812709966405375e-05,
+      "loss": 1.3047,
+      "step": 34400
+    },
+    {
+      "epoch": 19.001571248600225,
+      "grad_norm": 5.779742240905762,
+      "learning_rate": 4.826707726763718e-05,
+      "loss": 1.0745,
+      "step": 34500
+    },
+    {
+      "epoch": 19.00185120380739,
+      "grad_norm": 9.534099578857422,
+      "learning_rate": 4.840705487122061e-05,
+      "loss": 1.3069,
+      "step": 34600
+    },
+    {
+      "epoch": 19.002131159014557,
+      "grad_norm": 40.05844497680664,
+      "learning_rate": 4.8547032474804036e-05,
+      "loss": 0.9951,
+      "step": 34700
+    },
+    {
+      "epoch": 19.002411114221726,
+      "grad_norm": 35.630767822265625,
+      "learning_rate": 4.868701007838746e-05,
+      "loss": 1.1791,
+      "step": 34800
+    },
+    {
+      "epoch": 19.00269106942889,
+      "grad_norm": 23.145824432373047,
+      "learning_rate": 4.882698768197089e-05,
+      "loss": 1.1415,
+      "step": 34900
+    },
+    {
+      "epoch": 19.002971024636057,
+      "grad_norm": 5.915820598602295,
+      "learning_rate": 4.8966965285554314e-05,
+      "loss": 1.1801,
+      "step": 35000
+    },
+    {
+      "epoch": 19.003250979843227,
+      "grad_norm": 7.371408939361572,
+      "learning_rate": 4.910694288913774e-05,
+      "loss": 1.2856,
+      "step": 35100
+    },
+    {
+      "epoch": 19.003530935050392,
+      "grad_norm": 22.583425521850586,
+      "learning_rate": 4.9246920492721165e-05,
+      "loss": 1.3024,
+      "step": 35200
+    },
+    {
+      "epoch": 19.003810890257558,
+      "grad_norm": 33.190792083740234,
+      "learning_rate": 4.93868980963046e-05,
+      "loss": 1.2717,
+      "step": 35300
+    },
+    {
+      "epoch": 19.004090845464727,
+      "grad_norm": 37.30441665649414,
+      "learning_rate": 4.952687569988802e-05,
+      "loss": 1.4424,
+      "step": 35400
+    },
+    {
+      "epoch": 19.004370800671893,
+      "grad_norm": 25.534439086914062,
+      "learning_rate": 4.966685330347144e-05,
+      "loss": 1.5503,
+      "step": 35500
+    },
+    {
+      "epoch": 19.00465075587906,
+      "grad_norm": 17.24628257751465,
+      "learning_rate": 4.9806830907054875e-05,
+      "loss": 1.6463,
+      "step": 35600
+    },
+    {
+      "epoch": 19.004930711086224,
+      "grad_norm": 15.461166381835938,
+      "learning_rate": 4.99468085106383e-05,
+      "loss": 1.4479,
+      "step": 35700
+    },
+    {
+      "epoch": 19.00500069988802,
+      "eval_accuracy": 0.28345250255362614,
+      "eval_f1": 0.24778589478265647,
+      "eval_loss": 3.518174171447754,
+      "eval_precision": 0.25799831269060075,
+      "eval_recall": 0.28345250255362614,
+      "eval_runtime": 225.6187,
+      "eval_samples_per_second": 17.357,
+      "eval_steps_per_second": 8.678,
+      "eval_top_10_accuracy": 0.715270684371808,
+      "eval_top_1_accuracy": 0.28345250255362614,
+      "eval_top_5_accuracy": 0.5985699693564862,
+      "step": 35725
+    },
+    {
+      "epoch": 20.000209966405375,
+      "grad_norm": 14.617806434631348,
+      "learning_rate": 4.999035709841981e-05,
+      "loss": 1.0037,
+      "step": 35800
+    },
+    {
+      "epoch": 20.00048992161254,
+      "grad_norm": 31.072298049926758,
+      "learning_rate": 4.997480403135499e-05,
+      "loss": 0.9614,
+      "step": 35900
+    },
+    {
+      "epoch": 20.00076987681971,
+      "grad_norm": 17.07422637939453,
+      "learning_rate": 4.995925096429016e-05,
+      "loss": 0.9069,
+      "step": 36000
+    },
+    {
+      "epoch": 20.001049832026876,
+      "grad_norm": 7.916373252868652,
+      "learning_rate": 4.9943697897225334e-05,
+      "loss": 0.8297,
+      "step": 36100
+    },
+    {
+      "epoch": 20.00132978723404,
+      "grad_norm": 15.103555679321289,
+      "learning_rate": 4.992814483016051e-05,
+      "loss": 0.9506,
+      "step": 36200
+    },
+    {
+      "epoch": 20.00160974244121,
+      "grad_norm": 11.108366012573242,
+      "learning_rate": 4.9912591763095685e-05,
+      "loss": 1.2248,
+      "step": 36300
+    },
+    {
+      "epoch": 20.001889697648377,
+      "grad_norm": 39.42802047729492,
+      "learning_rate": 4.989703869603086e-05,
+      "loss": 1.3035,
+      "step": 36400
+    },
+    {
+      "epoch": 20.002169652855542,
+      "grad_norm": 11.509014129638672,
+      "learning_rate": 4.9881485628966035e-05,
+      "loss": 0.9473,
+      "step": 36500
+    },
+    {
+      "epoch": 20.00244960806271,
+      "grad_norm": 4.467772483825684,
+      "learning_rate": 4.986593256190121e-05,
+      "loss": 1.2713,
+      "step": 36600
+    },
+    {
+      "epoch": 20.002729563269877,
+      "grad_norm": 5.143234729766846,
+      "learning_rate": 4.9850379494836386e-05,
+      "loss": 1.1735,
+      "step": 36700
+    },
+    {
+      "epoch": 20.003009518477043,
+      "grad_norm": 3.546262502670288,
+      "learning_rate": 4.983482642777156e-05,
+      "loss": 1.2637,
+      "step": 36800
+    },
+    {
+      "epoch": 20.00328947368421,
+      "grad_norm": 42.26567840576172,
+      "learning_rate": 4.9819273360706736e-05,
+      "loss": 1.1088,
+      "step": 36900
+    },
+    {
+      "epoch": 20.003569428891378,
+      "grad_norm": 32.91080856323242,
+      "learning_rate": 4.980387582431256e-05,
+      "loss": 1.4778,
+      "step": 37000
+    },
+    {
+      "epoch": 20.003849384098544,
+      "grad_norm": 3.737553834915161,
+      "learning_rate": 4.978832275724773e-05,
+      "loss": 1.2936,
+      "step": 37100
+    },
+    {
+      "epoch": 20.00412933930571,
+      "grad_norm": 14.842428207397461,
+      "learning_rate": 4.977276969018291e-05,
+      "loss": 1.3193,
+      "step": 37200
+    },
+    {
+      "epoch": 20.00440929451288,
+      "grad_norm": 39.189353942871094,
+      "learning_rate": 4.975721662311808e-05,
+      "loss": 1.1632,
+      "step": 37300
+    },
+    {
+      "epoch": 20.004689249720045,
+      "grad_norm": 45.225460052490234,
+      "learning_rate": 4.974166355605326e-05,
+      "loss": 1.1817,
+      "step": 37400
+    },
+    {
+      "epoch": 20.00496920492721,
+      "grad_norm": 31.468297958374023,
+      "learning_rate": 4.972611048898843e-05,
+      "loss": 1.2076,
+      "step": 37500
+    },
+    {
+      "epoch": 20.005,
+      "eval_accuracy": 0.28421859039836567,
+      "eval_f1": 0.2493302460198215,
+      "eval_loss": 3.538327217102051,
+      "eval_precision": 0.2599856384394178,
+      "eval_recall": 0.28421859039836567,
+      "eval_runtime": 228.2995,
+      "eval_samples_per_second": 17.153,
+      "eval_steps_per_second": 8.576,
+      "eval_top_10_accuracy": 0.7070990806945863,
+      "eval_top_1_accuracy": 0.28421859039836567,
+      "eval_top_5_accuracy": 0.5980592441266599,
+      "step": 37511
+    },
+    {
+      "epoch": 21.00024846024636,
+      "grad_norm": 39.312591552734375,
+      "learning_rate": 4.9710557421923605e-05,
+      "loss": 0.8217,
+      "step": 37600
+    },
+    {
+      "epoch": 21.000528415453527,
+      "grad_norm": 40.887447357177734,
+      "learning_rate": 4.9695004354858784e-05,
+      "loss": 0.7459,
+      "step": 37700
+    },
+    {
+      "epoch": 21.000808370660696,
+      "grad_norm": 31.419702529907227,
+      "learning_rate": 4.9679451287793955e-05,
+      "loss": 0.9545,
+      "step": 37800
+    },
+    {
+      "epoch": 21.001088325867862,
+      "grad_norm": 32.3198127746582,
+      "learning_rate": 4.9663898220729134e-05,
+      "loss": 0.923,
+      "step": 37900
+    },
+    {
+      "epoch": 21.001368281075028,
+      "grad_norm": 5.489797592163086,
+      "learning_rate": 4.9648345153664306e-05,
+      "loss": 0.7275,
+      "step": 38000
+    },
+    {
+      "epoch": 21.001648236282193,
+      "grad_norm": 14.01230525970459,
+      "learning_rate": 4.963279208659948e-05,
+      "loss": 1.002,
+      "step": 38100
+    },
+    {
+      "epoch": 21.001928191489363,
+      "grad_norm": 32.620460510253906,
+      "learning_rate": 4.9617239019534657e-05,
+      "loss": 0.936,
+      "step": 38200
+    },
+    {
+      "epoch": 21.00220814669653,
+      "grad_norm": 39.93928909301758,
+      "learning_rate": 4.960168595246983e-05,
+      "loss": 0.9593,
+      "step": 38300
+    },
+    {
+      "epoch": 21.002488101903694,
+      "grad_norm": 40.96775817871094,
+      "learning_rate": 4.958613288540501e-05,
+      "loss": 0.8924,
+      "step": 38400
+    },
+    {
+      "epoch": 21.002768057110863,
+      "grad_norm": 24.79659652709961,
+      "learning_rate": 4.957057981834018e-05,
+      "loss": 1.0028,
+      "step": 38500
+    },
+    {
+      "epoch": 21.00304801231803,
+      "grad_norm": 11.152565956115723,
+      "learning_rate": 4.955502675127535e-05,
+      "loss": 1.1162,
+      "step": 38600
+    },
+    {
+      "epoch": 21.003327967525195,
+      "grad_norm": 37.34316635131836,
+      "learning_rate": 4.953947368421053e-05,
+      "loss": 1.1321,
+      "step": 38700
+    },
+    {
+      "epoch": 21.003607922732364,
+      "grad_norm": 5.6185526847839355,
+      "learning_rate": 4.95239206171457e-05,
+      "loss": 1.2126,
+      "step": 38800
+    },
+    {
+      "epoch": 21.00388787793953,
+      "grad_norm": 2.8256924152374268,
+      "learning_rate": 4.950836755008088e-05,
+      "loss": 0.9701,
+      "step": 38900
+    },
+    {
+      "epoch": 21.004167833146695,
+      "grad_norm": 23.547252655029297,
+      "learning_rate": 4.949281448301605e-05,
+      "loss": 0.9903,
+      "step": 39000
+    },
+    {
+      "epoch": 21.004447788353865,
+      "grad_norm": 46.85295486450195,
+      "learning_rate": 4.9477261415951224e-05,
+      "loss": 1.3198,
+      "step": 39100
+    },
+    {
+      "epoch": 21.00472774356103,
+      "grad_norm": 47.5512580871582,
+      "learning_rate": 4.9461863879557054e-05,
+      "loss": 1.4093,
+      "step": 39200
+    },
+    {
+      "epoch": 21.004999300111983,
+      "eval_accuracy": 0.27017364657814097,
+      "eval_f1": 0.23594364949777136,
+      "eval_loss": 3.6272244453430176,
+      "eval_precision": 0.2458368630291511,
+      "eval_recall": 0.27017364657814097,
+      "eval_runtime": 218.2444,
+      "eval_samples_per_second": 17.943,
+      "eval_steps_per_second": 8.972,
+      "eval_top_10_accuracy": 0.6917773237997957,
+      "eval_top_1_accuracy": 0.26991828396322776,
+      "eval_top_5_accuracy": 0.5898876404494382,
+      "step": 39297
+    },
+    {
+      "epoch": 22.000006998880178,
+      "grad_norm": 53.94511413574219,
+      "learning_rate": 4.9446310812492226e-05,
+      "loss": 1.3015,
+      "step": 39300
+    },
+    {
+      "epoch": 22.000286954087347,
+      "grad_norm": 6.432727336883545,
+      "learning_rate": 4.9430757745427405e-05,
+      "loss": 0.7464,
+      "step": 39400
+    },
+    {
+      "epoch": 22.000566909294513,
+      "grad_norm": 35.21794509887695,
+      "learning_rate": 4.941520467836258e-05,
+      "loss": 0.596,
+      "step": 39500
+    },
+    {
+      "epoch": 22.00084686450168,
+      "grad_norm": 41.56246566772461,
+      "learning_rate": 4.939965161129775e-05,
+      "loss": 0.7766,
+      "step": 39600
+    },
+    {
+      "epoch": 22.001126819708848,
+      "grad_norm": 2.1987500190734863,
+      "learning_rate": 4.938409854423293e-05,
+      "loss": 0.7652,
+      "step": 39700
+    },
+    {
+      "epoch": 22.001406774916013,
+      "grad_norm": 40.78813934326172,
+      "learning_rate": 4.93685454771681e-05,
+      "loss": 0.889,
+      "step": 39800
+    },
+    {
+      "epoch": 22.00168673012318,
+      "grad_norm": 57.54431915283203,
+      "learning_rate": 4.935299241010328e-05,
+      "loss": 0.9513,
+      "step": 39900
+    },
+    {
+      "epoch": 22.00196668533035,
+      "grad_norm": 22.238609313964844,
+      "learning_rate": 4.933743934303845e-05,
+      "loss": 0.9372,
+      "step": 40000
+    },
+    {
+      "epoch": 22.002246640537514,
+      "grad_norm": 10.823844909667969,
+      "learning_rate": 4.932188627597362e-05,
+      "loss": 0.7748,
+      "step": 40100
+    },
+    {
+      "epoch": 22.00252659574468,
+      "grad_norm": 70.64641571044922,
+      "learning_rate": 4.93063332089088e-05,
+      "loss": 0.9411,
+      "step": 40200
+    },
+    {
+      "epoch": 22.00280655095185,
+      "grad_norm": 31.34319305419922,
+      "learning_rate": 4.929078014184397e-05,
+      "loss": 1.337,
+      "step": 40300
+    },
+    {
+      "epoch": 22.003086506159015,
+      "grad_norm": 24.273786544799805,
+      "learning_rate": 4.927522707477915e-05,
+      "loss": 1.2788,
+      "step": 40400
+    },
+    {
+      "epoch": 22.00336646136618,
+      "grad_norm": 12.706284523010254,
+      "learning_rate": 4.925967400771432e-05,
+      "loss": 1.3021,
+      "step": 40500
+    },
+    {
+      "epoch": 22.00364641657335,
+      "grad_norm": 13.552841186523438,
+      "learning_rate": 4.9244120940649495e-05,
+      "loss": 0.8038,
+      "step": 40600
+    },
+    {
+      "epoch": 22.003926371780516,
+      "grad_norm": 55.14998245239258,
+      "learning_rate": 4.9228567873584674e-05,
+      "loss": 1.3207,
+      "step": 40700
+    },
+    {
+      "epoch": 22.00420632698768,
+      "grad_norm": 46.6702995300293,
+      "learning_rate": 4.9213014806519845e-05,
+      "loss": 1.3924,
+      "step": 40800
+    },
+    {
+      "epoch": 22.00448628219485,
+      "grad_norm": 4.238155841827393,
+      "learning_rate": 4.9197461739455024e-05,
+      "loss": 1.0947,
+      "step": 40900
+    },
+    {
+      "epoch": 22.004766237402016,
+      "grad_norm": 11.425141334533691,
+      "learning_rate": 4.9181908672390196e-05,
+      "loss": 1.3122,
+      "step": 41000
+    },
+    {
+      "epoch": 22.004998600223963,
+      "eval_accuracy": 0.2722165474974464,
+      "eval_f1": 0.23637877440977845,
+      "eval_loss": 3.679060459136963,
+      "eval_precision": 0.24200232503648353,
+      "eval_recall": 0.2722165474974464,
+      "eval_runtime": 226.0774,
+      "eval_samples_per_second": 17.322,
+      "eval_steps_per_second": 8.661,
+      "eval_top_10_accuracy": 0.695097037793667,
+      "eval_top_1_accuracy": 0.27247191011235955,
+      "eval_top_5_accuracy": 0.5822267620020429,
+      "step": 41083
+    },
+    {
+      "epoch": 23.000045492721163,
+      "grad_norm": 7.7932448387146,
+      "learning_rate": 4.916635560532537e-05,
+      "loss": 1.0572,
+      "step": 41100
+    },
+    {
+      "epoch": 23.000325447928333,
+      "grad_norm": 3.319699764251709,
+      "learning_rate": 4.9150802538260547e-05,
+      "loss": 0.669,
+      "step": 41200
+    },
+    {
+      "epoch": 23.0006054031355,
+      "grad_norm": 38.11969757080078,
+      "learning_rate": 4.913524947119572e-05,
+      "loss": 0.884,
+      "step": 41300
+    },
+    {
+      "epoch": 23.000885358342664,
+      "grad_norm": 37.383209228515625,
+      "learning_rate": 4.91196964041309e-05,
+      "loss": 0.6125,
+      "step": 41400
+    },
+    {
+      "epoch": 23.001165313549834,
+      "grad_norm": 21.38405418395996,
+      "learning_rate": 4.910429886773672e-05,
+      "loss": 0.987,
+      "step": 41500
+    },
+    {
+      "epoch": 23.001445268757,
+      "grad_norm": 44.750972747802734,
+      "learning_rate": 4.908874580067189e-05,
+      "loss": 0.8492,
+      "step": 41600
+    },
+    {
+      "epoch": 23.001725223964165,
+      "grad_norm": 27.360767364501953,
+      "learning_rate": 4.907319273360707e-05,
+      "loss": 0.8555,
+      "step": 41700
+    },
+    {
+      "epoch": 23.002005179171334,
+      "grad_norm": 34.80257797241211,
+      "learning_rate": 4.905763966654224e-05,
+      "loss": 0.9643,
+      "step": 41800
+    },
+    {
+      "epoch": 23.0022851343785,
+      "grad_norm": 34.58405685424805,
+      "learning_rate": 4.904208659947742e-05,
+      "loss": 0.748,
+      "step": 41900
+    },
+    {
+      "epoch": 23.002565089585666,
+      "grad_norm": 45.360355377197266,
+      "learning_rate": 4.9026533532412594e-05,
+      "loss": 1.115,
+      "step": 42000
+    },
+    {
+      "epoch": 23.00284504479283,
+      "grad_norm": 53.12576675415039,
+      "learning_rate": 4.9010980465347766e-05,
+      "loss": 1.0141,
+      "step": 42100
+    },
+    {
+      "epoch": 23.003125,
+      "grad_norm": 6.911433696746826,
+      "learning_rate": 4.8995427398282944e-05,
+      "loss": 1.0746,
+      "step": 42200
+    },
+    {
+      "epoch": 23.003404955207166,
+      "grad_norm": 2.8777873516082764,
+      "learning_rate": 4.8979874331218116e-05,
+      "loss": 1.088,
+      "step": 42300
+    },
+    {
+      "epoch": 23.003684910414332,
+      "grad_norm": 22.604394912719727,
+      "learning_rate": 4.8964321264153295e-05,
+      "loss": 1.1257,
+      "step": 42400
+    },
+    {
+      "epoch": 23.0039648656215,
+      "grad_norm": 15.631820678710938,
+      "learning_rate": 4.894876819708847e-05,
+      "loss": 1.1837,
+      "step": 42500
+    },
+    {
+      "epoch": 23.004244820828667,
+      "grad_norm": 55.372398376464844,
+      "learning_rate": 4.893321513002364e-05,
+      "loss": 1.2726,
+      "step": 42600
+    },
+    {
+      "epoch": 23.004524776035833,
+      "grad_norm": 33.215152740478516,
+      "learning_rate": 4.891766206295882e-05,
+      "loss": 1.2988,
+      "step": 42700
+    },
+    {
+      "epoch": 23.004804731243002,
+      "grad_norm": 3.4601495265960693,
+      "learning_rate": 4.890210899589399e-05,
+      "loss": 1.4774,
+      "step": 42800
+    },
+    {
+      "epoch": 23.00500069988802,
+      "eval_accuracy": 0.276046986721144,
+      "eval_f1": 0.24171610224981013,
+      "eval_loss": 3.7163209915161133,
+      "eval_precision": 0.25483022396965194,
+      "eval_recall": 0.276046986721144,
+      "eval_runtime": 223.2644,
+      "eval_samples_per_second": 17.54,
+      "eval_steps_per_second": 8.77,
+      "eval_top_10_accuracy": 0.6940755873340143,
+      "eval_top_1_accuracy": 0.27579162410623087,
+      "eval_top_5_accuracy": 0.5812053115423902,
+      "step": 42870
+    },
+    {
+      "epoch": 24.00008398656215,
+      "grad_norm": 6.2556962966918945,
+      "learning_rate": 4.888655592882917e-05,
+      "loss": 0.8567,
+      "step": 42900
+    },
+    {
+      "epoch": 24.00036394176932,
+      "grad_norm": 0.7903256416320801,
+      "learning_rate": 4.887100286176434e-05,
+      "loss": 0.7553,
+      "step": 43000
+    },
+    {
+      "epoch": 24.000643896976484,
+      "grad_norm": 2.442410945892334,
+      "learning_rate": 4.885544979469951e-05,
+      "loss": 0.7151,
+      "step": 43100
+    },
+    {
+      "epoch": 24.00092385218365,
+      "grad_norm": 0.9351925849914551,
+      "learning_rate": 4.883989672763469e-05,
+      "loss": 0.7546,
+      "step": 43200
+    },
+    {
+      "epoch": 24.001203807390816,
+      "grad_norm": 36.959136962890625,
+      "learning_rate": 4.882434366056986e-05,
+      "loss": 0.6823,
+      "step": 43300
+    },
+    {
+      "epoch": 24.001483762597985,
+      "grad_norm": 35.2205696105957,
+      "learning_rate": 4.880879059350504e-05,
+      "loss": 0.9348,
+      "step": 43400
+    },
+    {
+      "epoch": 24.00176371780515,
+      "grad_norm": 27.507863998413086,
+      "learning_rate": 4.8793393057110865e-05,
+      "loss": 0.6786,
+      "step": 43500
+    },
+    {
+      "epoch": 24.002043673012317,
+      "grad_norm": 4.061860084533691,
+      "learning_rate": 4.8777839990046037e-05,
+      "loss": 1.0083,
+      "step": 43600
+    },
+    {
+      "epoch": 24.002323628219486,
+      "grad_norm": 44.9168701171875,
+      "learning_rate": 4.8762286922981215e-05,
+      "loss": 0.9745,
+      "step": 43700
+    },
+    {
+      "epoch": 24.00260358342665,
+      "grad_norm": 1.4848639965057373,
+      "learning_rate": 4.874673385591639e-05,
+      "loss": 1.0088,
+      "step": 43800
+    },
+    {
+      "epoch": 24.002883538633817,
+      "grad_norm": 33.28787612915039,
+      "learning_rate": 4.8731180788851566e-05,
+      "loss": 0.7772,
+      "step": 43900
+    },
+    {
+      "epoch": 24.003163493840987,
+      "grad_norm": 46.953208923339844,
+      "learning_rate": 4.871562772178674e-05,
+      "loss": 1.189,
+      "step": 44000
+    },
+    {
+      "epoch": 24.003443449048152,
+      "grad_norm": 3.805981159210205,
+      "learning_rate": 4.870007465472191e-05,
+      "loss": 1.1122,
+      "step": 44100
+    },
+    {
+      "epoch": 24.003723404255318,
+      "grad_norm": 5.7398505210876465,
+      "learning_rate": 4.868452158765709e-05,
+      "loss": 0.9651,
+      "step": 44200
+    },
+    {
+      "epoch": 24.004003359462487,
+      "grad_norm": 42.19569778442383,
+      "learning_rate": 4.866896852059226e-05,
+      "loss": 1.1016,
+      "step": 44300
+    },
+    {
+      "epoch": 24.004283314669653,
+      "grad_norm": 32.88911437988281,
+      "learning_rate": 4.865341545352744e-05,
+      "loss": 0.821,
+      "step": 44400
+    },
+    {
+      "epoch": 24.00456326987682,
+      "grad_norm": 6.485898971557617,
+      "learning_rate": 4.863786238646261e-05,
+      "loss": 1.2985,
+      "step": 44500
+    },
+    {
+      "epoch": 24.004843225083988,
+      "grad_norm": 46.899261474609375,
+      "learning_rate": 4.862230931939778e-05,
+      "loss": 1.1338,
+      "step": 44600
+    },
+    {
+      "epoch": 24.005,
+      "eval_accuracy": 0.267364657814096,
+      "eval_f1": 0.23461128701357323,
+      "eval_loss": 3.7576823234558105,
+      "eval_precision": 0.2471144003255852,
+      "eval_recall": 0.267364657814096,
+      "eval_runtime": 228.6369,
+      "eval_samples_per_second": 17.128,
+      "eval_steps_per_second": 8.564,
+      "eval_top_10_accuracy": 0.6897344228804902,
+      "eval_top_1_accuracy": 0.2676200204290092,
+      "eval_top_5_accuracy": 0.5817160367722165,
+      "step": 44656
+    },
+    {
+      "epoch": 25.000122480403135,
+      "grad_norm": 29.33512306213379,
+      "learning_rate": 4.860675625233296e-05,
+      "loss": 0.884,
+      "step": 44700
+    },
+    {
+      "epoch": 25.0004024356103,
+      "grad_norm": 29.168378829956055,
+      "learning_rate": 4.859120318526813e-05,
+      "loss": 0.7048,
+      "step": 44800
+    },
+    {
+      "epoch": 25.00068239081747,
+      "grad_norm": 4.4006547927856445,
+      "learning_rate": 4.857565011820331e-05,
+      "loss": 0.5229,
+      "step": 44900
+    },
+    {
+      "epoch": 25.000962346024636,
+      "grad_norm": 0.7437628507614136,
+      "learning_rate": 4.8560097051138484e-05,
+      "loss": 0.6759,
+      "step": 45000
+    },
+    {
+      "epoch": 25.0012423012318,
+      "grad_norm": 1.8170311450958252,
+      "learning_rate": 4.854454398407366e-05,
+      "loss": 0.7961,
+      "step": 45100
+    },
+    {
+      "epoch": 25.00152225643897,
+      "grad_norm": 65.61835479736328,
+      "learning_rate": 4.8528990917008834e-05,
+      "loss": 0.7743,
+      "step": 45200
+    },
+    {
+      "epoch": 25.001802211646137,
+      "grad_norm": 38.23652267456055,
+      "learning_rate": 4.851343784994401e-05,
+      "loss": 1.0217,
+      "step": 45300
+    },
+    {
+      "epoch": 25.002082166853302,
+      "grad_norm": 6.596869468688965,
+      "learning_rate": 4.8497884782879185e-05,
+      "loss": 0.8333,
+      "step": 45400
+    },
+    {
+      "epoch": 25.00236212206047,
+      "grad_norm": 42.07466506958008,
+      "learning_rate": 4.8482331715814364e-05,
+      "loss": 0.7881,
+      "step": 45500
+    },
+    {
+      "epoch": 25.002642077267637,
+      "grad_norm": 2.7012946605682373,
+      "learning_rate": 4.8466778648749536e-05,
+      "loss": 0.9458,
+      "step": 45600
+    },
+    {
+      "epoch": 25.002922032474803,
+      "grad_norm": 30.443016052246094,
+      "learning_rate": 4.845138111235536e-05,
+      "loss": 0.9779,
+      "step": 45700
+    },
+    {
+      "epoch": 25.003201987681972,
+      "grad_norm": 10.28647518157959,
+      "learning_rate": 4.843582804529053e-05,
+      "loss": 1.0701,
+      "step": 45800
+    },
+    {
+      "epoch": 25.003481942889138,
+      "grad_norm": 41.59596252441406,
+      "learning_rate": 4.842027497822571e-05,
+      "loss": 1.08,
+      "step": 45900
+    },
+    {
+      "epoch": 25.003761898096304,
+      "grad_norm": 33.521087646484375,
+      "learning_rate": 4.840472191116088e-05,
+      "loss": 1.1218,
+      "step": 46000
+    },
+    {
+      "epoch": 25.004041853303473,
+      "grad_norm": 1.2339781522750854,
+      "learning_rate": 4.8389168844096054e-05,
+      "loss": 1.1475,
+      "step": 46100
+    },
+    {
+      "epoch": 25.00432180851064,
+      "grad_norm": 1.9241328239440918,
+      "learning_rate": 4.837361577703123e-05,
+      "loss": 1.1846,
+      "step": 46200
+    },
+    {
+      "epoch": 25.004601763717805,
+      "grad_norm": 37.55961608886719,
+      "learning_rate": 4.8358062709966404e-05,
+      "loss": 1.348,
+      "step": 46300
+    },
+    {
+      "epoch": 25.00488171892497,
+      "grad_norm": 3.4552130699157715,
+      "learning_rate": 4.834250964290158e-05,
+      "loss": 1.1446,
+      "step": 46400
+    },
+    {
+      "epoch": 25.004999300111983,
+      "eval_accuracy": 0.274514811031665,
+      "eval_f1": 0.24301366205430505,
+      "eval_loss": 3.76619553565979,
+      "eval_precision": 0.2533079046421743,
+      "eval_recall": 0.274514811031665,
+      "eval_runtime": 228.3549,
+      "eval_samples_per_second": 17.149,
+      "eval_steps_per_second": 8.574,
+      "eval_top_10_accuracy": 0.7073544433094995,
+      "eval_top_1_accuracy": 0.274514811031665,
+      "eval_top_5_accuracy": 0.5957609805924413,
+      "step": 46442
+    },
+    {
+      "epoch": 26.00016097424412,
+      "grad_norm": 0.36602893471717834,
+      "learning_rate": 4.8326956575836755e-05,
+      "loss": 0.652,
+      "step": 46500
+    },
+    {
+      "epoch": 26.000440929451287,
+      "grad_norm": 3.7436516284942627,
+      "learning_rate": 4.8311403508771927e-05,
+      "loss": 0.5429,
+      "step": 46600
+    },
+    {
+      "epoch": 26.000720884658456,
+      "grad_norm": 34.448280334472656,
+      "learning_rate": 4.8295850441707105e-05,
+      "loss": 0.6339,
+      "step": 46700
+    },
+    {
+      "epoch": 26.001000839865622,
+      "grad_norm": 1.2372946739196777,
+      "learning_rate": 4.828029737464228e-05,
+      "loss": 0.7041,
+      "step": 46800
+    },
+    {
+      "epoch": 26.001280795072788,
+      "grad_norm": 7.009436130523682,
+      "learning_rate": 4.8264744307577456e-05,
+      "loss": 0.6515,
+      "step": 46900
+    },
+    {
+      "epoch": 26.001560750279957,
+      "grad_norm": 26.225847244262695,
+      "learning_rate": 4.8249191240512635e-05,
+      "loss": 0.7411,
+      "step": 47000
+    },
+    {
+      "epoch": 26.001840705487123,
+      "grad_norm": 50.131893157958984,
+      "learning_rate": 4.8233638173447806e-05,
+      "loss": 0.9812,
+      "step": 47100
+    },
+    {
+      "epoch": 26.00212066069429,
+      "grad_norm": 31.176666259765625,
+      "learning_rate": 4.8218085106382985e-05,
+      "loss": 0.7793,
+      "step": 47200
+    },
+    {
+      "epoch": 26.002400615901454,
+      "grad_norm": 36.463321685791016,
+      "learning_rate": 4.820253203931816e-05,
+      "loss": 0.9363,
+      "step": 47300
+    },
+    {
+      "epoch": 26.002680571108623,
+      "grad_norm": 0.5882710218429565,
+      "learning_rate": 4.8186978972253336e-05,
+      "loss": 0.8734,
+      "step": 47400
+    },
+    {
+      "epoch": 26.00296052631579,
+      "grad_norm": 56.16053771972656,
+      "learning_rate": 4.817142590518851e-05,
+      "loss": 0.8313,
+      "step": 47500
+    },
+    {
+      "epoch": 26.003240481522955,
+      "grad_norm": 27.409542083740234,
+      "learning_rate": 4.815587283812368e-05,
+      "loss": 1.0755,
+      "step": 47600
+    },
+    {
+      "epoch": 26.003520436730124,
+      "grad_norm": 37.072975158691406,
+      "learning_rate": 4.814031977105886e-05,
+      "loss": 1.1273,
+      "step": 47700
+    },
+    {
+      "epoch": 26.00380039193729,
+      "grad_norm": 12.40834903717041,
+      "learning_rate": 4.8124922234664675e-05,
+      "loss": 1.0032,
+      "step": 47800
+    },
+    {
+      "epoch": 26.004080347144455,
+      "grad_norm": 33.848052978515625,
+      "learning_rate": 4.8109369167599854e-05,
+      "loss": 1.1193,
+      "step": 47900
+    },
+    {
+      "epoch": 26.004360302351625,
+      "grad_norm": 35.6231803894043,
+      "learning_rate": 4.8093816100535026e-05,
+      "loss": 1.1809,
+      "step": 48000
+    },
+    {
+      "epoch": 26.00464025755879,
+      "grad_norm": 53.72087860107422,
+      "learning_rate": 4.80782630334702e-05,
+      "loss": 1.2457,
+      "step": 48100
+    },
+    {
+      "epoch": 26.004920212765956,
+      "grad_norm": 27.564422607421875,
+      "learning_rate": 4.8062709966405376e-05,
+      "loss": 1.0876,
+      "step": 48200
+    },
+    {
+      "epoch": 26.004998600223963,
+      "eval_accuracy": 0.27860061287027577,
+      "eval_f1": 0.24350325505841336,
+      "eval_loss": 3.8689017295837402,
+      "eval_precision": 0.25253788318139697,
+      "eval_recall": 0.27860061287027577,
+      "eval_runtime": 228.9607,
+      "eval_samples_per_second": 17.103,
+      "eval_steps_per_second": 8.552,
+      "eval_top_10_accuracy": 0.6851378958120531,
+      "eval_top_1_accuracy": 0.278855975485189,
+      "eval_top_5_accuracy": 0.5781409601634321,
+      "step": 48228
+    },
+    {
+      "epoch": 27.000199468085107,
+      "grad_norm": 8.388084411621094,
+      "learning_rate": 4.804715689934055e-05,
+      "loss": 0.7608,
+      "step": 48300
+    },
+    {
+      "epoch": 27.000479423292273,
+      "grad_norm": 1.5461616516113281,
+      "learning_rate": 4.803160383227573e-05,
+      "loss": 0.7676,
+      "step": 48400
+    },
+    {
+      "epoch": 27.00075937849944,
+      "grad_norm": 2.8455636501312256,
+      "learning_rate": 4.80160507652109e-05,
+      "loss": 0.5519,
+      "step": 48500
+    },
+    {
+      "epoch": 27.001039333706608,
+      "grad_norm": 37.09521484375,
+      "learning_rate": 4.800049769814608e-05,
+      "loss": 0.738,
+      "step": 48600
+    },
+    {
+      "epoch": 27.001319288913773,
+      "grad_norm": 24.946674346923828,
+      "learning_rate": 4.798494463108125e-05,
+      "loss": 0.8532,
+      "step": 48700
+    },
+    {
+      "epoch": 27.00159924412094,
+      "grad_norm": 1.9327168464660645,
+      "learning_rate": 4.796939156401643e-05,
+      "loss": 1.0297,
+      "step": 48800
+    },
+    {
+      "epoch": 27.00187919932811,
+      "grad_norm": 37.39794921875,
+      "learning_rate": 4.79538384969516e-05,
+      "loss": 0.7483,
+      "step": 48900
+    },
+    {
+      "epoch": 27.002159154535274,
+      "grad_norm": 1.6338342428207397,
+      "learning_rate": 4.793828542988678e-05,
+      "loss": 0.7587,
+      "step": 49000
+    },
+    {
+      "epoch": 27.00243910974244,
+      "grad_norm": 13.684679985046387,
+      "learning_rate": 4.792273236282195e-05,
+      "loss": 0.7818,
+      "step": 49100
+    },
+    {
+      "epoch": 27.00271906494961,
+      "grad_norm": 31.797103881835938,
+      "learning_rate": 4.790717929575713e-05,
+      "loss": 0.8512,
+      "step": 49200
+    },
+    {
+      "epoch": 27.002999020156775,
+      "grad_norm": 41.46651840209961,
+      "learning_rate": 4.78916262286923e-05,
+      "loss": 1.3144,
+      "step": 49300
+    },
+    {
+      "epoch": 27.00327897536394,
+      "grad_norm": 6.321496963500977,
+      "learning_rate": 4.787607316162748e-05,
+      "loss": 0.9538,
+      "step": 49400
+    },
+    {
+      "epoch": 27.00355893057111,
+      "grad_norm": 0.416471391916275,
+      "learning_rate": 4.786052009456265e-05,
+      "loss": 1.2089,
+      "step": 49500
+    },
+    {
+      "epoch": 27.003838885778276,
+      "grad_norm": 31.824552536010742,
+      "learning_rate": 4.784496702749782e-05,
+      "loss": 1.0261,
+      "step": 49600
+    },
+    {
+      "epoch": 27.00411884098544,
+      "grad_norm": 51.06576156616211,
+      "learning_rate": 4.7829413960433e-05,
+      "loss": 1.1528,
+      "step": 49700
+    },
+    {
+      "epoch": 27.00439879619261,
+      "grad_norm": 20.590635299682617,
+      "learning_rate": 4.7813860893368174e-05,
+      "loss": 0.9639,
+      "step": 49800
+    },
+    {
+      "epoch": 27.004678751399776,
+      "grad_norm": 6.22679328918457,
+      "learning_rate": 4.779830782630335e-05,
+      "loss": 1.1261,
+      "step": 49900
+    },
+    {
+      "epoch": 27.004958706606942,
+      "grad_norm": 49.03010559082031,
+      "learning_rate": 4.7782754759238525e-05,
+      "loss": 1.0755,
+      "step": 50000
+    },
+    {
+      "epoch": 27.00500069988802,
+      "eval_accuracy": 0.2773237997957099,
+      "eval_f1": 0.24651992442655182,
+      "eval_loss": 3.8659050464630127,
+      "eval_precision": 0.2626311954133711,
+      "eval_recall": 0.2773237997957099,
+      "eval_runtime": 217.8846,
+      "eval_samples_per_second": 17.973,
+      "eval_steps_per_second": 8.986,
+      "eval_top_10_accuracy": 0.70097037793667,
+      "eval_top_1_accuracy": 0.2773237997957099,
+      "eval_top_5_accuracy": 0.5855464759959143,
+      "step": 50015
+    },
+    {
+      "epoch": 28.000237961926093,
+      "grad_norm": 0.523884117603302,
+      "learning_rate": 4.776735722284434e-05,
+      "loss": 0.6872,
+      "step": 50100
+    },
+    {
+      "epoch": 28.00051791713326,
+      "grad_norm": 23.129125595092773,
+      "learning_rate": 4.775180415577952e-05,
+      "loss": 0.5709,
+      "step": 50200
+    },
+    {
+      "epoch": 28.000797872340424,
+      "grad_norm": 33.83433532714844,
+      "learning_rate": 4.773625108871469e-05,
+      "loss": 0.7503,
+      "step": 50300
+    },
+    {
+      "epoch": 28.001077827547594,
+      "grad_norm": 0.6037617325782776,
+      "learning_rate": 4.772069802164987e-05,
+      "loss": 0.9381,
+      "step": 50400
+    },
+    {
+      "epoch": 28.00135778275476,
+      "grad_norm": 0.9508163332939148,
+      "learning_rate": 4.770514495458505e-05,
+      "loss": 0.7982,
+      "step": 50500
+    },
+    {
+      "epoch": 28.001637737961925,
+      "grad_norm": 5.3412251472473145,
+      "learning_rate": 4.768959188752022e-05,
+      "loss": 0.8581,
+      "step": 50600
+    },
+    {
+      "epoch": 28.001917693169094,
+      "grad_norm": 44.1540641784668,
+      "learning_rate": 4.76740388204554e-05,
+      "loss": 0.9748,
+      "step": 50700
+    },
+    {
+      "epoch": 28.00219764837626,
+      "grad_norm": 122.2665786743164,
+      "learning_rate": 4.765848575339057e-05,
+      "loss": 0.8151,
+      "step": 50800
+    },
+    {
+      "epoch": 28.002477603583426,
+      "grad_norm": 1.555612564086914,
+      "learning_rate": 4.764293268632575e-05,
+      "loss": 0.827,
+      "step": 50900
+    },
+    {
+      "epoch": 28.002757558790595,
+      "grad_norm": 27.170162200927734,
+      "learning_rate": 4.762737961926092e-05,
+      "loss": 0.7203,
+      "step": 51000
+    },
+    {
+      "epoch": 28.00303751399776,
+      "grad_norm": 32.98384094238281,
+      "learning_rate": 4.7611826552196094e-05,
+      "loss": 0.9387,
+      "step": 51100
+    },
+    {
+      "epoch": 28.003317469204926,
+      "grad_norm": 4.3118157386779785,
+      "learning_rate": 4.759627348513127e-05,
+      "loss": 0.9075,
+      "step": 51200
+    },
+    {
+      "epoch": 28.003597424412096,
+      "grad_norm": 18.410526275634766,
+      "learning_rate": 4.7580720418066445e-05,
+      "loss": 1.0489,
+      "step": 51300
+    },
+    {
+      "epoch": 28.00387737961926,
+      "grad_norm": 0.6224632859230042,
+      "learning_rate": 4.7565167351001623e-05,
+      "loss": 0.777,
+      "step": 51400
+    },
+    {
+      "epoch": 28.004157334826427,
+      "grad_norm": 21.2976016998291,
+      "learning_rate": 4.7549614283936795e-05,
+      "loss": 0.8111,
+      "step": 51500
+    },
+    {
+      "epoch": 28.004437290033593,
+      "grad_norm": 39.73042678833008,
+      "learning_rate": 4.753406121687197e-05,
+      "loss": 1.1039,
+      "step": 51600
+    },
+    {
+      "epoch": 28.004717245240762,
+      "grad_norm": 38.503684997558594,
+      "learning_rate": 4.7518508149807146e-05,
+      "loss": 1.1345,
+      "step": 51700
+    },
+    {
+      "epoch": 28.004997200447928,
+      "grad_norm": 1.6712877750396729,
+      "learning_rate": 4.750295508274232e-05,
+      "loss": 1.1092,
+      "step": 51800
+    },
+    {
+      "epoch": 28.005,
+      "eval_accuracy": 0.278855975485189,
+      "eval_f1": 0.2448271325501993,
+      "eval_loss": 3.91512393951416,
+      "eval_precision": 0.2539590289973334,
+      "eval_recall": 0.278855975485189,
+      "eval_runtime": 227.3347,
+      "eval_samples_per_second": 17.226,
+      "eval_steps_per_second": 8.613,
+      "eval_top_10_accuracy": 0.6966292134831461,
+      "eval_top_1_accuracy": 0.27860061287027577,
+      "eval_top_5_accuracy": 0.5891215526046987,
+      "step": 51801
+    },
+    {
+      "epoch": 29.00027645576708,
+      "grad_norm": 41.43709945678711,
+      "learning_rate": 4.7487402015677497e-05,
+      "loss": 0.6971,
+      "step": 51900
+    },
+    {
+      "epoch": 29.000556410974244,
+      "grad_norm": 33.58243942260742,
+      "learning_rate": 4.747184894861267e-05,
+      "loss": 0.5892,
+      "step": 52000
+    },
+    {
+      "epoch": 29.00083636618141,
+      "grad_norm": 0.9046162962913513,
+      "learning_rate": 4.745629588154784e-05,
+      "loss": 0.7421,
+      "step": 52100
+    },
+    {
+      "epoch": 29.00111632138858,
+      "grad_norm": 1.261534333229065,
+      "learning_rate": 4.744074281448302e-05,
+      "loss": 0.7439,
+      "step": 52200
+    },
+    {
+      "epoch": 29.001396276595745,
+      "grad_norm": 25.14644432067871,
+      "learning_rate": 4.742518974741819e-05,
+      "loss": 0.6257,
+      "step": 52300
+    },
+    {
+      "epoch": 29.00167623180291,
+      "grad_norm": 30.892152786254883,
+      "learning_rate": 4.740963668035337e-05,
+      "loss": 0.6543,
+      "step": 52400
+    },
+    {
+      "epoch": 29.00195618701008,
+      "grad_norm": 38.89350509643555,
+      "learning_rate": 4.739423914395919e-05,
+      "loss": 0.8098,
+      "step": 52500
+    },
+    {
+      "epoch": 29.002236142217246,
+      "grad_norm": 1.8801395893096924,
+      "learning_rate": 4.737868607689437e-05,
+      "loss": 0.7963,
+      "step": 52600
+    },
+    {
+      "epoch": 29.00251609742441,
+      "grad_norm": 4.187483310699463,
+      "learning_rate": 4.7363133009829544e-05,
+      "loss": 0.9501,
+      "step": 52700
+    },
+    {
+      "epoch": 29.002796052631577,
+      "grad_norm": 0.38438934087753296,
+      "learning_rate": 4.7347579942764716e-05,
+      "loss": 1.0176,
+      "step": 52800
+    },
+    {
+      "epoch": 29.003076007838747,
+      "grad_norm": 24.411516189575195,
+      "learning_rate": 4.7332026875699894e-05,
+      "loss": 0.8995,
+      "step": 52900
+    },
+    {
+      "epoch": 29.003355963045912,
+      "grad_norm": 37.18138885498047,
+      "learning_rate": 4.7316473808635066e-05,
+      "loss": 0.9454,
+      "step": 53000
+    },
+    {
+      "epoch": 29.003635918253078,
+      "grad_norm": 34.37055206298828,
+      "learning_rate": 4.7300920741570245e-05,
+      "loss": 1.382,
+      "step": 53100
+    },
+    {
+      "epoch": 29.003915873460247,
+      "grad_norm": 0.7984645962715149,
+      "learning_rate": 4.728536767450542e-05,
+      "loss": 1.2185,
+      "step": 53200
+    },
+    {
+      "epoch": 29.004195828667413,
+      "grad_norm": 0.9875502586364746,
+      "learning_rate": 4.726981460744059e-05,
+      "loss": 0.9532,
+      "step": 53300
+    },
+    {
+      "epoch": 29.00447578387458,
+      "grad_norm": 38.7867317199707,
+      "learning_rate": 4.725426154037577e-05,
+      "loss": 0.9921,
+      "step": 53400
+    },
+    {
+      "epoch": 29.004755739081748,
+      "grad_norm": 2.064009666442871,
+      "learning_rate": 4.723870847331094e-05,
+      "loss": 0.9877,
+      "step": 53500
+    },
+    {
+      "epoch": 29.004999300111983,
+      "eval_accuracy": 0.283197139938713,
+      "eval_f1": 0.24799378699098912,
+      "eval_loss": 4.014481544494629,
+      "eval_precision": 0.25962104387530993,
+      "eval_recall": 0.283197139938713,
+      "eval_runtime": 225.1257,
+      "eval_samples_per_second": 17.395,
+      "eval_steps_per_second": 8.697,
+      "eval_top_10_accuracy": 0.6902451481103167,
+      "eval_top_1_accuracy": 0.283197139938713,
+      "eval_top_5_accuracy": 0.585291113381001,
+      "step": 53587
+    },
+    {
+      "epoch": 30.000034994400895,
+      "grad_norm": 5.7924675941467285,
+      "learning_rate": 4.722315540624611e-05,
+      "loss": 0.7719,
+      "step": 53600
+    },
+    {
+      "epoch": 30.00031494960806,
+      "grad_norm": 0.8171272873878479,
+      "learning_rate": 4.720760233918129e-05,
+      "loss": 0.568,
+      "step": 53700
+    },
+    {
+      "epoch": 30.00059490481523,
+      "grad_norm": 31.9758358001709,
+      "learning_rate": 4.719204927211646e-05,
+      "loss": 0.7045,
+      "step": 53800
+    },
+    {
+      "epoch": 30.000874860022396,
+      "grad_norm": 7.561930179595947,
+      "learning_rate": 4.717649620505164e-05,
+      "loss": 0.589,
+      "step": 53900
+    },
+    {
+      "epoch": 30.00115481522956,
+      "grad_norm": 2.9528915882110596,
+      "learning_rate": 4.716094313798681e-05,
+      "loss": 0.8783,
+      "step": 54000
+    },
+    {
+      "epoch": 30.00143477043673,
+      "grad_norm": 38.09792709350586,
+      "learning_rate": 4.7145390070921984e-05,
+      "loss": 0.6687,
+      "step": 54100
+    },
+    {
+      "epoch": 30.001714725643897,
+      "grad_norm": 17.62958526611328,
+      "learning_rate": 4.712983700385716e-05,
+      "loss": 0.7255,
+      "step": 54200
+    },
+    {
+      "epoch": 30.001994680851062,
+      "grad_norm": 47.918392181396484,
+      "learning_rate": 4.7114283936792335e-05,
+      "loss": 0.8257,
+      "step": 54300
+    },
+    {
+      "epoch": 30.00227463605823,
+      "grad_norm": 4.136358737945557,
+      "learning_rate": 4.7098730869727513e-05,
+      "loss": 0.9338,
+      "step": 54400
+    },
+    {
+      "epoch": 30.002554591265397,
+      "grad_norm": 40.67402267456055,
+      "learning_rate": 4.7083177802662685e-05,
+      "loss": 0.8837,
+      "step": 54500
+    },
+    {
+      "epoch": 30.002834546472563,
+      "grad_norm": 28.413721084594727,
+      "learning_rate": 4.7067780266268516e-05,
+      "loss": 1.0067,
+      "step": 54600
+    },
+    {
+      "epoch": 30.003114501679732,
+      "grad_norm": 3.6661288738250732,
+      "learning_rate": 4.705222719920369e-05,
+      "loss": 0.94,
+      "step": 54700
+    },
+    {
+      "epoch": 30.003394456886898,
+      "grad_norm": 0.3771117329597473,
+      "learning_rate": 4.7036829662809505e-05,
+      "loss": 1.0855,
+      "step": 54800
+    },
+    {
+      "epoch": 30.003674412094064,
+      "grad_norm": 68.63765716552734,
+      "learning_rate": 4.702127659574468e-05,
+      "loss": 1.2102,
+      "step": 54900
+    },
+    {
+      "epoch": 30.003954367301233,
+      "grad_norm": 1.7344911098480225,
+      "learning_rate": 4.7005723528679855e-05,
+      "loss": 0.8625,
+      "step": 55000
+    },
+    {
+      "epoch": 30.0042343225084,
+      "grad_norm": 1.0122687816619873,
+      "learning_rate": 4.6990170461615034e-05,
+      "loss": 1.2124,
+      "step": 55100
+    },
+    {
+      "epoch": 30.004514277715565,
+      "grad_norm": 0.7474389672279358,
+      "learning_rate": 4.6974617394550206e-05,
+      "loss": 1.1678,
+      "step": 55200
+    },
+    {
+      "epoch": 30.004794232922734,
+      "grad_norm": 6.55318546295166,
+      "learning_rate": 4.695906432748538e-05,
+      "loss": 0.946,
+      "step": 55300
+    },
+    {
+      "epoch": 30.004998600223963,
+      "eval_accuracy": 0.2770684371807967,
+      "eval_f1": 0.24776271203710667,
+      "eval_loss": 3.9974467754364014,
+      "eval_precision": 0.2626098273936753,
+      "eval_recall": 0.2770684371807967,
+      "eval_runtime": 224.4315,
+      "eval_samples_per_second": 17.449,
+      "eval_steps_per_second": 8.724,
+      "eval_top_10_accuracy": 0.6915219611848825,
+      "eval_top_1_accuracy": 0.2763023493360572,
+      "eval_top_5_accuracy": 0.5883554647599591,
+      "step": 55373
+    },
+    {
+      "epoch": 31.00007348824188,
+      "grad_norm": 1.8461582660675049,
+      "learning_rate": 4.6943511260420556e-05,
+      "loss": 0.7635,
+      "step": 55400
+    },
+    {
+      "epoch": 31.000353443449047,
+      "grad_norm": 19.579442977905273,
+      "learning_rate": 4.692795819335573e-05,
+      "loss": 0.4863,
+      "step": 55500
+    },
+    {
+      "epoch": 31.000633398656216,
+      "grad_norm": 0.31262636184692383,
+      "learning_rate": 4.691240512629091e-05,
+      "loss": 0.7288,
+      "step": 55600
+    },
+    {
+      "epoch": 31.000913353863382,
+      "grad_norm": 0.9600892066955566,
+      "learning_rate": 4.689685205922608e-05,
+      "loss": 0.7114,
+      "step": 55700
+    },
+    {
+      "epoch": 31.001193309070548,
+      "grad_norm": 33.14706039428711,
+      "learning_rate": 4.688129899216126e-05,
+      "loss": 0.5404,
+      "step": 55800
+    },
+    {
+      "epoch": 31.001473264277717,
+      "grad_norm": 71.96259307861328,
+      "learning_rate": 4.686574592509643e-05,
+      "loss": 1.0186,
+      "step": 55900
+    },
+    {
+      "epoch": 31.001753219484883,
+      "grad_norm": 69.26034545898438,
+      "learning_rate": 4.685019285803161e-05,
+      "loss": 0.5707,
+      "step": 56000
+    },
+    {
+      "epoch": 31.00203317469205,
+      "grad_norm": 19.351402282714844,
+      "learning_rate": 4.683463979096679e-05,
+      "loss": 0.9263,
+      "step": 56100
+    },
+    {
+      "epoch": 31.002313129899218,
+      "grad_norm": 34.91017532348633,
+      "learning_rate": 4.681908672390196e-05,
+      "loss": 1.0285,
+      "step": 56200
+    },
+    {
+      "epoch": 31.002593085106383,
+      "grad_norm": 4.248389720916748,
+      "learning_rate": 4.680353365683713e-05,
+      "loss": 0.9372,
+      "step": 56300
+    },
+    {
+      "epoch": 31.00287304031355,
+      "grad_norm": 11.464951515197754,
+      "learning_rate": 4.678798058977231e-05,
+      "loss": 1.2444,
+      "step": 56400
+    },
+    {
+      "epoch": 31.00315299552072,
+      "grad_norm": 3.1078383922576904,
+      "learning_rate": 4.677242752270748e-05,
+      "loss": 0.9295,
+      "step": 56500
+    },
+    {
+      "epoch": 31.003432950727884,
+      "grad_norm": 36.12800979614258,
+      "learning_rate": 4.675687445564266e-05,
+      "loss": 0.9508,
+      "step": 56600
+    },
+    {
+      "epoch": 31.00371290593505,
+      "grad_norm": 31.067119598388672,
+      "learning_rate": 4.674132138857783e-05,
+      "loss": 0.9861,
+      "step": 56700
+    },
+    {
+      "epoch": 31.003992861142216,
+      "grad_norm": 2.1897337436676025,
+      "learning_rate": 4.6725768321513003e-05,
+      "loss": 1.031,
+      "step": 56800
+    },
+    {
+      "epoch": 31.004272816349385,
+      "grad_norm": 32.6429443359375,
+      "learning_rate": 4.671021525444818e-05,
+      "loss": 0.9505,
+      "step": 56900
+    },
+    {
+      "epoch": 31.00455277155655,
+      "grad_norm": 24.413639068603516,
+      "learning_rate": 4.6694662187383354e-05,
+      "loss": 0.9977,
+      "step": 57000
+    },
+    {
+      "epoch": 31.004832726763716,
+      "grad_norm": 3.9543466567993164,
+      "learning_rate": 4.667910912031853e-05,
+      "loss": 1.1572,
+      "step": 57100
+    },
+    {
+      "epoch": 31.00500069988802,
+      "eval_accuracy": 0.2870275791624106,
+      "eval_f1": 0.2544566698227792,
+      "eval_loss": 4.0120158195495605,
+      "eval_precision": 0.26635475344936527,
+      "eval_recall": 0.2870275791624106,
+      "eval_runtime": 226.0085,
+      "eval_samples_per_second": 17.327,
+      "eval_steps_per_second": 8.663,
+      "eval_top_10_accuracy": 0.68488253319714,
+      "eval_top_1_accuracy": 0.28677221654749746,
+      "eval_top_5_accuracy": 0.5786516853932584,
+      "step": 57160
+    },
+    {
+      "epoch": 32.00011198208286,
+      "grad_norm": 30.936607360839844,
+      "learning_rate": 4.6663556053253705e-05,
+      "loss": 0.8606,
+      "step": 57200
+    },
+    {
+      "epoch": 32.00039193729003,
+      "grad_norm": 0.24527552723884583,
+      "learning_rate": 4.6648002986188877e-05,
+      "loss": 0.4935,
+      "step": 57300
+    },
+    {
+      "epoch": 32.0006718924972,
+      "grad_norm": 0.4248438775539398,
+      "learning_rate": 4.6632449919124055e-05,
+      "loss": 0.7442,
+      "step": 57400
+    },
+    {
+      "epoch": 32.000951847704364,
+      "grad_norm": 16.248767852783203,
+      "learning_rate": 4.661689685205923e-05,
+      "loss": 0.4462,
+      "step": 57500
+    },
+    {
+      "epoch": 32.00123180291153,
+      "grad_norm": 2.6000514030456543,
+      "learning_rate": 4.6601343784994406e-05,
+      "loss": 0.6116,
+      "step": 57600
+    },
+    {
+      "epoch": 32.0015117581187,
+      "grad_norm": 0.5838096141815186,
+      "learning_rate": 4.658579071792958e-05,
+      "loss": 0.7441,
+      "step": 57700
+    },
+    {
+      "epoch": 32.001791713325865,
+      "grad_norm": 44.025821685791016,
+      "learning_rate": 4.657023765086475e-05,
+      "loss": 0.8305,
+      "step": 57800
+    },
+    {
+      "epoch": 32.002071668533034,
+      "grad_norm": 2.842480421066284,
+      "learning_rate": 4.655468458379993e-05,
+      "loss": 0.804,
+      "step": 57900
+    },
+    {
+      "epoch": 32.0023516237402,
+      "grad_norm": 38.31273651123047,
+      "learning_rate": 4.65391315167351e-05,
+      "loss": 1.0669,
+      "step": 58000
+    },
+    {
+      "epoch": 32.002631578947366,
+      "grad_norm": 2.9648852348327637,
+      "learning_rate": 4.652357844967028e-05,
+      "loss": 0.9346,
+      "step": 58100
+    },
+    {
+      "epoch": 32.002911534154535,
+      "grad_norm": 29.513328552246094,
+      "learning_rate": 4.650802538260545e-05,
+      "loss": 0.7173,
+      "step": 58200
+    },
+    {
+      "epoch": 32.003191489361704,
+      "grad_norm": 43.86397933959961,
+      "learning_rate": 4.649247231554062e-05,
+      "loss": 1.083,
+      "step": 58300
+    },
+    {
+      "epoch": 32.003471444568866,
+      "grad_norm": 24.38750648498535,
+      "learning_rate": 4.64769192484758e-05,
+      "loss": 1.1133,
+      "step": 58400
+    },
+    {
+      "epoch": 32.003751399776036,
+      "grad_norm": 2.4515841007232666,
+      "learning_rate": 4.646136618141097e-05,
+      "loss": 0.8093,
+      "step": 58500
+    },
+    {
+      "epoch": 32.004031354983205,
+      "grad_norm": 1.9614719152450562,
+      "learning_rate": 4.644581311434615e-05,
+      "loss": 1.1983,
+      "step": 58600
+    },
+    {
+      "epoch": 32.00431131019037,
+      "grad_norm": 1.845036506652832,
+      "learning_rate": 4.6430260047281324e-05,
+      "loss": 1.0139,
+      "step": 58700
+    },
+    {
+      "epoch": 32.004591265397536,
+      "grad_norm": 37.803550720214844,
+      "learning_rate": 4.64147069802165e-05,
+      "loss": 0.9255,
+      "step": 58800
+    },
+    {
+      "epoch": 32.004871220604706,
+      "grad_norm": 3.2317845821380615,
+      "learning_rate": 4.6399309443822326e-05,
+      "loss": 1.0663,
+      "step": 58900
+    },
+    {
+      "epoch": 32.005,
+      "eval_accuracy": 0.2763023493360572,
+      "eval_f1": 0.24460618831487968,
+      "eval_loss": 4.123536109924316,
+      "eval_precision": 0.25515447978420397,
+      "eval_recall": 0.2763023493360572,
+      "eval_runtime": 225.915,
+      "eval_samples_per_second": 17.334,
+      "eval_steps_per_second": 8.667,
+      "eval_top_10_accuracy": 0.673391215526047,
+      "eval_top_1_accuracy": 0.2763023493360572,
+      "eval_top_5_accuracy": 0.5702247191011236,
+      "step": 58946
+    },
+    {
+      "epoch": 33.00015047592385,
+      "grad_norm": 34.167762756347656,
+      "learning_rate": 4.63837563767575e-05,
+      "loss": 0.8821,
+      "step": 59000
+    },
+    {
+      "epoch": 33.00043043113102,
+      "grad_norm": 37.967613220214844,
+      "learning_rate": 4.636820330969268e-05,
+      "loss": 0.5966,
+      "step": 59100
+    },
+    {
+      "epoch": 33.000710386338184,
+      "grad_norm": 31.84786605834961,
+      "learning_rate": 4.635265024262785e-05,
+      "loss": 0.8587,
+      "step": 59200
+    },
+    {
+      "epoch": 33.000990341545354,
+      "grad_norm": 1.0245041847229004,
+      "learning_rate": 4.633709717556302e-05,
+      "loss": 0.641,
+      "step": 59300
+    },
+    {
+      "epoch": 33.00127029675252,
+      "grad_norm": 28.23471450805664,
+      "learning_rate": 4.63215441084982e-05,
+      "loss": 0.8437,
+      "step": 59400
+    },
+    {
+      "epoch": 33.001550251959685,
+      "grad_norm": 8.509440422058105,
+      "learning_rate": 4.630599104143337e-05,
+      "loss": 0.5655,
+      "step": 59500
+    },
+    {
+      "epoch": 33.001830207166854,
+      "grad_norm": 1.8368055820465088,
+      "learning_rate": 4.629043797436855e-05,
+      "loss": 0.6445,
+      "step": 59600
+    },
+    {
+      "epoch": 33.002110162374024,
+      "grad_norm": 1.10395085811615,
+      "learning_rate": 4.627488490730372e-05,
+      "loss": 0.8858,
+      "step": 59700
+    },
+    {
+      "epoch": 33.002390117581186,
+      "grad_norm": 31.648338317871094,
+      "learning_rate": 4.6259331840238893e-05,
+      "loss": 0.89,
+      "step": 59800
+    },
+    {
+      "epoch": 33.002670072788355,
+      "grad_norm": 37.29691696166992,
+      "learning_rate": 4.624377877317407e-05,
+      "loss": 0.9659,
+      "step": 59900
+    },
+    {
+      "epoch": 33.00295002799552,
+      "grad_norm": 0.43430274724960327,
+      "learning_rate": 4.6228225706109244e-05,
+      "loss": 0.8978,
+      "step": 60000
+    },
+    {
+      "epoch": 33.00322998320269,
+      "grad_norm": 35.974586486816406,
+      "learning_rate": 4.621267263904442e-05,
+      "loss": 0.8231,
+      "step": 60100
+    },
+    {
+      "epoch": 33.003509938409856,
+      "grad_norm": 8.3503999710083,
+      "learning_rate": 4.6197119571979595e-05,
+      "loss": 1.1412,
+      "step": 60200
+    },
+    {
+      "epoch": 33.00378989361702,
+      "grad_norm": 34.63167190551758,
+      "learning_rate": 4.6181566504914767e-05,
+      "loss": 1.1708,
+      "step": 60300
+    },
+    {
+      "epoch": 33.00406984882419,
+      "grad_norm": 25.57069969177246,
+      "learning_rate": 4.6166013437849945e-05,
+      "loss": 1.0096,
+      "step": 60400
+    },
+    {
+      "epoch": 33.00434980403136,
+      "grad_norm": 42.470584869384766,
+      "learning_rate": 4.615046037078512e-05,
+      "loss": 0.9288,
+      "step": 60500
+    },
+    {
+      "epoch": 33.00462975923852,
+      "grad_norm": 39.73023986816406,
+      "learning_rate": 4.6134907303720296e-05,
+      "loss": 0.8984,
+      "step": 60600
+    },
+    {
+      "epoch": 33.00490971444569,
+      "grad_norm": 32.1846923828125,
+      "learning_rate": 4.611935423665547e-05,
+      "loss": 1.085,
+      "step": 60700
+    },
+    {
+      "epoch": 33.00499930011198,
+      "eval_accuracy": 0.27145045965270687,
+      "eval_f1": 0.2402946650128174,
+      "eval_loss": 4.142614364624023,
+      "eval_precision": 0.2528004594086391,
+      "eval_recall": 0.27145045965270687,
+      "eval_runtime": 226.7122,
+      "eval_samples_per_second": 17.273,
+      "eval_steps_per_second": 8.637,
+      "eval_top_10_accuracy": 0.6782431052093973,
+      "eval_top_1_accuracy": 0.27145045965270687,
+      "eval_top_5_accuracy": 0.5717568947906027,
+      "step": 60732
+    },
+    {
+      "epoch": 34.000188969764835,
+      "grad_norm": 1.583235502243042,
+      "learning_rate": 4.6103801169590646e-05,
+      "loss": 0.7598,
+      "step": 60800
+    },
+    {
+      "epoch": 34.000468924972004,
+      "grad_norm": 41.911720275878906,
+      "learning_rate": 4.608840363319647e-05,
+      "loss": 0.6443,
+      "step": 60900
+    },
+    {
+      "epoch": 34.000748880179174,
+      "grad_norm": 0.4762161076068878,
+      "learning_rate": 4.607285056613164e-05,
+      "loss": 0.6395,
+      "step": 61000
+    },
+    {
+      "epoch": 34.001028835386336,
+      "grad_norm": 44.91167068481445,
+      "learning_rate": 4.605729749906682e-05,
+      "loss": 0.6762,
+      "step": 61100
+    },
+    {
+      "epoch": 34.001308790593505,
+      "grad_norm": 2.2632224559783936,
+      "learning_rate": 4.604174443200199e-05,
+      "loss": 0.6015,
+      "step": 61200
+    },
+    {
+      "epoch": 34.001588745800674,
+      "grad_norm": 0.8883789777755737,
+      "learning_rate": 4.6026191364937164e-05,
+      "loss": 0.8364,
+      "step": 61300
+    },
+    {
+      "epoch": 34.00186870100784,
+      "grad_norm": 0.21396863460540771,
+      "learning_rate": 4.601063829787234e-05,
+      "loss": 0.8178,
+      "step": 61400
+    },
+    {
+      "epoch": 34.002148656215006,
+      "grad_norm": 19.145397186279297,
+      "learning_rate": 4.5995085230807515e-05,
+      "loss": 0.7655,
+      "step": 61500
+    },
+    {
+      "epoch": 34.002428611422175,
+      "grad_norm": 34.72240447998047,
+      "learning_rate": 4.5979532163742694e-05,
+      "loss": 0.724,
+      "step": 61600
+    },
+    {
+      "epoch": 34.00270856662934,
+      "grad_norm": 0.7193199396133423,
+      "learning_rate": 4.5963979096677865e-05,
+      "loss": 0.8898,
+      "step": 61700
+    },
+    {
+      "epoch": 34.00298852183651,
+      "grad_norm": 41.06938934326172,
+      "learning_rate": 4.594842602961304e-05,
+      "loss": 0.9544,
+      "step": 61800
+    },
+    {
+      "epoch": 34.003268477043676,
+      "grad_norm": 31.43454933166504,
+      "learning_rate": 4.5932872962548216e-05,
+      "loss": 0.8788,
+      "step": 61900
+    },
+    {
+      "epoch": 34.00354843225084,
+      "grad_norm": 51.27229309082031,
+      "learning_rate": 4.591731989548339e-05,
+      "loss": 0.7306,
+      "step": 62000
+    },
+    {
+      "epoch": 34.00382838745801,
+      "grad_norm": 1.796972632408142,
+      "learning_rate": 4.590176682841857e-05,
+      "loss": 0.9033,
+      "step": 62100
+    },
+    {
+      "epoch": 34.00410834266518,
+      "grad_norm": 39.01441955566406,
+      "learning_rate": 4.588621376135374e-05,
+      "loss": 0.9423,
+      "step": 62200
+    },
+    {
+      "epoch": 34.00438829787234,
+      "grad_norm": 0.1421416848897934,
+      "learning_rate": 4.587066069428892e-05,
+      "loss": 1.0109,
+      "step": 62300
+    },
+    {
+      "epoch": 34.00466825307951,
+      "grad_norm": 32.059959411621094,
+      "learning_rate": 4.585510762722409e-05,
+      "loss": 1.0664,
+      "step": 62400
+    },
+    {
+      "epoch": 34.00494820828668,
+      "grad_norm": 47.05455780029297,
+      "learning_rate": 4.583955456015927e-05,
+      "loss": 1.3799,
+      "step": 62500
+    },
+    {
+      "epoch": 34.00499860022396,
+      "eval_accuracy": 0.27553626149131766,
+      "eval_f1": 0.24451187615318176,
+      "eval_loss": 4.101657390594482,
+      "eval_precision": 0.25940964549613826,
+      "eval_recall": 0.27553626149131766,
+      "eval_runtime": 226.0648,
+      "eval_samples_per_second": 17.322,
+      "eval_steps_per_second": 8.661,
+      "eval_top_10_accuracy": 0.6838610827374872,
+      "eval_top_1_accuracy": 0.27553626149131766,
+      "eval_top_5_accuracy": 0.5743105209397344,
+      "step": 62518
+    },
+    {
+      "epoch": 35.000227463605825,
+      "grad_norm": 39.157588958740234,
+      "learning_rate": 4.582400149309444e-05,
+      "loss": 0.5925,
+      "step": 62600
+    },
+    {
+      "epoch": 35.00050741881299,
+      "grad_norm": 13.6599760055542,
+      "learning_rate": 4.580844842602962e-05,
+      "loss": 0.6346,
+      "step": 62700
+    },
+    {
+      "epoch": 35.000787374020156,
+      "grad_norm": 0.17690366506576538,
+      "learning_rate": 4.579289535896479e-05,
+      "loss": 0.7364,
+      "step": 62800
+    },
+    {
+      "epoch": 35.001067329227325,
+      "grad_norm": 5.204614162445068,
+      "learning_rate": 4.577734229189997e-05,
+      "loss": 0.6949,
+      "step": 62900
+    },
+    {
+      "epoch": 35.00134728443449,
+      "grad_norm": 0.2207849770784378,
+      "learning_rate": 4.5761944755505786e-05,
+      "loss": 0.7559,
+      "step": 63000
+    },
+    {
+      "epoch": 35.00162723964166,
+      "grad_norm": 41.076332092285156,
+      "learning_rate": 4.5746391688440964e-05,
+      "loss": 0.7937,
+      "step": 63100
+    },
+    {
+      "epoch": 35.001907194848826,
+      "grad_norm": 47.679779052734375,
+      "learning_rate": 4.5730838621376136e-05,
+      "loss": 0.9563,
+      "step": 63200
+    },
+    {
+      "epoch": 35.00218715005599,
+      "grad_norm": 40.40637969970703,
+      "learning_rate": 4.571528555431131e-05,
+      "loss": 0.9855,
+      "step": 63300
+    },
+    {
+      "epoch": 35.00246710526316,
+      "grad_norm": 41.624244689941406,
+      "learning_rate": 4.569973248724649e-05,
+      "loss": 0.7511,
+      "step": 63400
+    },
+    {
+      "epoch": 35.00274706047033,
+      "grad_norm": 32.867408752441406,
+      "learning_rate": 4.568417942018166e-05,
+      "loss": 0.7931,
+      "step": 63500
+    },
+    {
+      "epoch": 35.00302701567749,
+      "grad_norm": 28.88155174255371,
+      "learning_rate": 4.566862635311684e-05,
+      "loss": 0.7677,
+      "step": 63600
+    },
+    {
+      "epoch": 35.00330697088466,
+      "grad_norm": 3.3230557441711426,
+      "learning_rate": 4.565307328605201e-05,
+      "loss": 0.9521,
+      "step": 63700
+    },
+    {
+      "epoch": 35.00358692609183,
+      "grad_norm": 56.24114227294922,
+      "learning_rate": 4.563752021898718e-05,
+      "loss": 0.9281,
+      "step": 63800
+    },
+    {
+      "epoch": 35.00386688129899,
+      "grad_norm": 0.7060623168945312,
+      "learning_rate": 4.562196715192236e-05,
+      "loss": 0.8661,
+      "step": 63900
+    },
+    {
+      "epoch": 35.00414683650616,
+      "grad_norm": 0.19642525911331177,
+      "learning_rate": 4.560641408485753e-05,
+      "loss": 1.1495,
+      "step": 64000
+    },
+    {
+      "epoch": 35.00442679171333,
+      "grad_norm": 0.7218245267868042,
+      "learning_rate": 4.559086101779271e-05,
+      "loss": 1.2618,
+      "step": 64100
+    },
+    {
+      "epoch": 35.00470674692049,
+      "grad_norm": 0.09785791486501694,
+      "learning_rate": 4.557530795072788e-05,
+      "loss": 0.8072,
+      "step": 64200
+    },
+    {
+      "epoch": 35.00498670212766,
+      "grad_norm": 2.7851946353912354,
+      "learning_rate": 4.555975488366306e-05,
+      "loss": 0.8419,
+      "step": 64300
+    },
+    {
+      "epoch": 35.005000699888015,
+      "eval_accuracy": 0.27936670071501535,
+      "eval_f1": 0.24866188033961273,
+      "eval_loss": 4.176891803741455,
+      "eval_precision": 0.261978146413795,
+      "eval_recall": 0.27936670071501535,
+      "eval_runtime": 226.5401,
+      "eval_samples_per_second": 17.286,
+      "eval_steps_per_second": 8.643,
+      "eval_top_10_accuracy": 0.683605720122574,
+      "eval_top_1_accuracy": 0.27936670071501535,
+      "eval_top_5_accuracy": 0.5727783452502554,
+      "step": 64305
+    },
+    {
+      "epoch": 36.00026595744681,
+      "grad_norm": 0.516681969165802,
+      "learning_rate": 4.554420181659824e-05,
+      "loss": 0.506,
+      "step": 64400
+    },
+    {
+      "epoch": 36.000545912653976,
+      "grad_norm": 0.44977912306785583,
+      "learning_rate": 4.552864874953341e-05,
+      "loss": 0.5155,
+      "step": 64500
+    },
+    {
+      "epoch": 36.000825867861145,
+      "grad_norm": 36.72551727294922,
+      "learning_rate": 4.551309568246859e-05,
+      "loss": 0.7122,
+      "step": 64600
+    },
+    {
+      "epoch": 36.00110582306831,
+      "grad_norm": 4.889037609100342,
+      "learning_rate": 4.549754261540376e-05,
+      "loss": 0.5901,
+      "step": 64700
+    },
+    {
+      "epoch": 36.00138577827548,
+      "grad_norm": 1.347294569015503,
+      "learning_rate": 4.5481989548338934e-05,
+      "loss": 0.7308,
+      "step": 64800
+    },
+    {
+      "epoch": 36.001665733482646,
+      "grad_norm": 27.506364822387695,
+      "learning_rate": 4.546643648127411e-05,
+      "loss": 0.8497,
+      "step": 64900
+    },
+    {
+      "epoch": 36.00194568868981,
+      "grad_norm": 10.64461898803711,
+      "learning_rate": 4.5450883414209285e-05,
+      "loss": 0.613,
+      "step": 65000
+    },
+    {
+      "epoch": 36.00222564389698,
+      "grad_norm": 46.96599197387695,
+      "learning_rate": 4.543548587781511e-05,
+      "loss": 1.1151,
+      "step": 65100
+    },
+    {
+      "epoch": 36.00250559910414,
+      "grad_norm": 2.2545366287231445,
+      "learning_rate": 4.541993281075028e-05,
+      "loss": 0.7595,
+      "step": 65200
+    },
+    {
+      "epoch": 36.00278555431131,
+      "grad_norm": 0.24742832779884338,
+      "learning_rate": 4.540437974368545e-05,
+      "loss": 1.0187,
+      "step": 65300
+    },
+    {
+      "epoch": 36.00306550951848,
+      "grad_norm": 0.34370434284210205,
+      "learning_rate": 4.538882667662063e-05,
+      "loss": 0.748,
+      "step": 65400
+    },
+    {
+      "epoch": 36.00334546472564,
+      "grad_norm": 0.0902349129319191,
+      "learning_rate": 4.53732736095558e-05,
+      "loss": 0.8016,
+      "step": 65500
+    },
+    {
+      "epoch": 36.00362541993281,
+      "grad_norm": 5.595888137817383,
+      "learning_rate": 4.535772054249098e-05,
+      "loss": 1.2848,
+      "step": 65600
+    },
+    {
+      "epoch": 36.00390537513998,
+      "grad_norm": 0.23979386687278748,
+      "learning_rate": 4.534216747542615e-05,
+      "loss": 0.8729,
+      "step": 65700
+    },
+    {
+      "epoch": 36.00418533034714,
+      "grad_norm": 2.2355868816375732,
+      "learning_rate": 4.532661440836133e-05,
+      "loss": 0.8902,
+      "step": 65800
+    },
+    {
+      "epoch": 36.00446528555431,
+      "grad_norm": 1.7017910480499268,
+      "learning_rate": 4.5311061341296504e-05,
+      "loss": 1.1558,
+      "step": 65900
+    },
+    {
+      "epoch": 36.00474524076148,
+      "grad_norm": 0.6865089535713196,
+      "learning_rate": 4.529550827423168e-05,
+      "loss": 1.1308,
+      "step": 66000
+    },
+    {
+      "epoch": 36.005,
+      "eval_accuracy": 0.27936670071501535,
+      "eval_f1": 0.24678017904603605,
+      "eval_loss": 4.133280277252197,
+      "eval_precision": 0.2568876401518128,
+      "eval_recall": 0.27936670071501535,
+      "eval_runtime": 227.4478,
+      "eval_samples_per_second": 17.217,
+      "eval_steps_per_second": 8.609,
+      "eval_top_10_accuracy": 0.6889683350357507,
+      "eval_top_1_accuracy": 0.2796220633299285,
+      "eval_top_5_accuracy": 0.5804392236976507,
+      "step": 66091
+    },
+    {
+      "epoch": 37.00002449608063,
+      "grad_norm": 4.969370365142822,
+      "learning_rate": 4.5279955207166854e-05,
+      "loss": 0.9023,
+      "step": 66100
+    },
+    {
+      "epoch": 37.000304451287796,
+      "grad_norm": 3.5516724586486816,
+      "learning_rate": 4.526440214010203e-05,
+      "loss": 0.4191,
+      "step": 66200
+    },
+    {
+      "epoch": 37.00058440649496,
+      "grad_norm": 1.1282627582550049,
+      "learning_rate": 4.5248849073037205e-05,
+      "loss": 0.5857,
+      "step": 66300
+    },
+    {
+      "epoch": 37.00086436170213,
+      "grad_norm": 0.38395920395851135,
+      "learning_rate": 4.5233296005972384e-05,
+      "loss": 0.5167,
+      "step": 66400
+    },
+    {
+      "epoch": 37.0011443169093,
+      "grad_norm": 0.11064700782299042,
+      "learning_rate": 4.5217742938907556e-05,
+      "loss": 0.4823,
+      "step": 66500
+    },
+    {
+      "epoch": 37.00142427211646,
+      "grad_norm": 1.7022432088851929,
+      "learning_rate": 4.5202189871842734e-05,
+      "loss": 0.8452,
+      "step": 66600
+    },
+    {
+      "epoch": 37.00170422732363,
+      "grad_norm": 34.14882278442383,
+      "learning_rate": 4.5186636804777906e-05,
+      "loss": 0.6892,
+      "step": 66700
+    },
+    {
+      "epoch": 37.0019841825308,
+      "grad_norm": 29.29656219482422,
+      "learning_rate": 4.517108373771308e-05,
+      "loss": 0.7937,
+      "step": 66800
+    },
+    {
+      "epoch": 37.00226413773796,
+      "grad_norm": 23.31835174560547,
+      "learning_rate": 4.515553067064826e-05,
+      "loss": 0.8147,
+      "step": 66900
+    },
+    {
+      "epoch": 37.00254409294513,
+      "grad_norm": 29.693347930908203,
+      "learning_rate": 4.513997760358343e-05,
+      "loss": 0.9029,
+      "step": 67000
+    },
+    {
+      "epoch": 37.0028240481523,
+      "grad_norm": 32.401817321777344,
+      "learning_rate": 4.512458006718925e-05,
+      "loss": 0.8017,
+      "step": 67100
+    },
+    {
+      "epoch": 37.00310400335946,
+      "grad_norm": 66.0666732788086,
+      "learning_rate": 4.510933806146572e-05,
+      "loss": 0.9821,
+      "step": 67200
+    },
+    {
+      "epoch": 37.00338395856663,
+      "grad_norm": 2.4841063022613525,
+      "learning_rate": 4.50937849944009e-05,
+      "loss": 0.755,
+      "step": 67300
+    },
+    {
+      "epoch": 37.0036639137738,
+      "grad_norm": 28.219749450683594,
+      "learning_rate": 4.507823192733607e-05,
+      "loss": 1.0117,
+      "step": 67400
+    },
+    {
+      "epoch": 37.00394386898096,
+      "grad_norm": 0.13683520257472992,
+      "learning_rate": 4.506267886027124e-05,
+      "loss": 0.9562,
+      "step": 67500
+    },
+    {
+      "epoch": 37.00422382418813,
+      "grad_norm": 0.6364304423332214,
+      "learning_rate": 4.504712579320642e-05,
+      "loss": 1.1528,
+      "step": 67600
+    },
+    {
+      "epoch": 37.0045037793953,
+      "grad_norm": 2.150608539581299,
+      "learning_rate": 4.5031572726141594e-05,
+      "loss": 0.8858,
+      "step": 67700
+    },
+    {
+      "epoch": 37.00478373460246,
+      "grad_norm": 5.197592258453369,
+      "learning_rate": 4.501601965907677e-05,
+      "loss": 0.9785,
+      "step": 67800
+    },
+    {
+      "epoch": 37.00499930011198,
+      "eval_accuracy": 0.2770684371807967,
+      "eval_f1": 0.2470607814773372,
+      "eval_loss": 4.239144325256348,
+      "eval_precision": 0.26169534456877486,
+      "eval_recall": 0.2770684371807967,
+      "eval_runtime": 222.8835,
+      "eval_samples_per_second": 17.57,
+      "eval_steps_per_second": 8.785,
+      "eval_top_10_accuracy": 0.6654749744637385,
+      "eval_top_1_accuracy": 0.2770684371807967,
+      "eval_top_5_accuracy": 0.559244126659857,
+      "step": 67877
+    },
+    {
+      "epoch": 38.00006298992161,
+      "grad_norm": 19.93975067138672,
+      "learning_rate": 4.5000466592011944e-05,
+      "loss": 1.0973,
+      "step": 67900
+    },
+    {
+      "epoch": 38.00034294512878,
+      "grad_norm": 3.1067473888397217,
+      "learning_rate": 4.498491352494712e-05,
+      "loss": 0.5088,
+      "step": 68000
+    },
+    {
+      "epoch": 38.00062290033595,
+      "grad_norm": 11.543988227844238,
+      "learning_rate": 4.4969360457882295e-05,
+      "loss": 0.6467,
+      "step": 68100
+    },
+    {
+      "epoch": 38.00090285554311,
+      "grad_norm": 0.6623213291168213,
+      "learning_rate": 4.4953807390817474e-05,
+      "loss": 0.5682,
+      "step": 68200
+    },
+    {
+      "epoch": 38.00118281075028,
+      "grad_norm": 30.84617805480957,
+      "learning_rate": 4.4938254323752646e-05,
+      "loss": 0.6804,
+      "step": 68300
+    },
+    {
+      "epoch": 38.00146276595745,
+      "grad_norm": 44.52362823486328,
+      "learning_rate": 4.4922701256687824e-05,
+      "loss": 0.6082,
+      "step": 68400
+    },
+    {
+      "epoch": 38.00174272116461,
+      "grad_norm": 0.663332998752594,
+      "learning_rate": 4.4907148189622996e-05,
+      "loss": 0.6514,
+      "step": 68500
+    },
+    {
+      "epoch": 38.00202267637178,
+      "grad_norm": 33.60304641723633,
+      "learning_rate": 4.4891595122558175e-05,
+      "loss": 0.7611,
+      "step": 68600
+    },
+    {
+      "epoch": 38.00230263157895,
+      "grad_norm": 0.861744225025177,
+      "learning_rate": 4.487604205549335e-05,
+      "loss": 0.64,
+      "step": 68700
+    },
+    {
+      "epoch": 38.00258258678611,
+      "grad_norm": 37.43878936767578,
+      "learning_rate": 4.4860488988428525e-05,
+      "loss": 0.7982,
+      "step": 68800
+    },
+    {
+      "epoch": 38.00286254199328,
+      "grad_norm": 18.1073055267334,
+      "learning_rate": 4.48449359213637e-05,
+      "loss": 1.1362,
+      "step": 68900
+    },
+    {
+      "epoch": 38.00314249720045,
+      "grad_norm": 32.217166900634766,
+      "learning_rate": 4.482938285429887e-05,
+      "loss": 0.6346,
+      "step": 69000
+    },
+    {
+      "epoch": 38.00342245240761,
+      "grad_norm": 19.9949951171875,
+      "learning_rate": 4.481382978723405e-05,
+      "loss": 1.0563,
+      "step": 69100
+    },
+    {
+      "epoch": 38.00370240761478,
+      "grad_norm": 6.759098052978516,
+      "learning_rate": 4.479827672016922e-05,
+      "loss": 1.0793,
+      "step": 69200
+    },
+    {
+      "epoch": 38.00398236282195,
+      "grad_norm": 31.205677032470703,
+      "learning_rate": 4.47827236531044e-05,
+      "loss": 0.9697,
+      "step": 69300
+    },
+    {
+      "epoch": 38.00426231802911,
+      "grad_norm": 16.609729766845703,
+      "learning_rate": 4.476717058603957e-05,
+      "loss": 0.9377,
+      "step": 69400
+    },
+    {
+      "epoch": 38.00454227323628,
+      "grad_norm": 0.45362696051597595,
+      "learning_rate": 4.475161751897474e-05,
+      "loss": 1.1853,
+      "step": 69500
+    },
+    {
+      "epoch": 38.00482222844345,
+      "grad_norm": 47.770931243896484,
+      "learning_rate": 4.473606445190992e-05,
+      "loss": 0.923,
+      "step": 69600
+    },
+    {
+      "epoch": 38.00499860022396,
+      "eval_accuracy": 0.28038815117466803,
+      "eval_f1": 0.24841193213143883,
+      "eval_loss": 4.227442264556885,
+      "eval_precision": 0.25828937193211965,
+      "eval_recall": 0.28038815117466803,
+      "eval_runtime": 223.4537,
+      "eval_samples_per_second": 17.525,
+      "eval_steps_per_second": 8.762,
+      "eval_top_10_accuracy": 0.6682839632277835,
+      "eval_top_1_accuracy": 0.28038815117466803,
+      "eval_top_5_accuracy": 0.5768641470888662,
+      "step": 69663
+    },
+    {
+      "epoch": 39.0001014837626,
+      "grad_norm": 14.46917724609375,
+      "learning_rate": 4.472051138484509e-05,
+      "loss": 0.7307,
+      "step": 69700
+    },
+    {
+      "epoch": 39.00038143896977,
+      "grad_norm": 1.327642798423767,
+      "learning_rate": 4.470495831778027e-05,
+      "loss": 0.5555,
+      "step": 69800
+    },
+    {
+      "epoch": 39.00066139417693,
+      "grad_norm": 20.620176315307617,
+      "learning_rate": 4.4689405250715443e-05,
+      "loss": 0.4745,
+      "step": 69900
+    },
+    {
+      "epoch": 39.0009413493841,
+      "grad_norm": 0.19782160222530365,
+      "learning_rate": 4.4673852183650615e-05,
+      "loss": 0.6574,
+      "step": 70000
+    },
+    {
+      "epoch": 39.00122130459127,
+      "grad_norm": 1.6866756677627563,
+      "learning_rate": 4.4658299116585794e-05,
+      "loss": 0.6071,
+      "step": 70100
+    },
+    {
+      "epoch": 39.00150125979843,
+      "grad_norm": 0.5778133273124695,
+      "learning_rate": 4.4642746049520966e-05,
+      "loss": 0.6044,
+      "step": 70200
+    },
+    {
+      "epoch": 39.0017812150056,
+      "grad_norm": 0.14593109488487244,
+      "learning_rate": 4.4627192982456145e-05,
+      "loss": 0.8319,
+      "step": 70300
+    },
+    {
+      "epoch": 39.00206117021276,
+      "grad_norm": 29.488271713256836,
+      "learning_rate": 4.4611639915391316e-05,
+      "loss": 1.0373,
+      "step": 70400
+    },
+    {
+      "epoch": 39.00234112541993,
+      "grad_norm": 36.36933135986328,
+      "learning_rate": 4.459608684832649e-05,
+      "loss": 0.7747,
+      "step": 70500
+    },
+    {
+      "epoch": 39.0026210806271,
+      "grad_norm": 0.10845132917165756,
+      "learning_rate": 4.458053378126167e-05,
+      "loss": 0.7393,
+      "step": 70600
+    },
+    {
+      "epoch": 39.00290103583426,
+      "grad_norm": 10.269367218017578,
+      "learning_rate": 4.456498071419684e-05,
+      "loss": 0.8106,
+      "step": 70700
+    },
+    {
+      "epoch": 39.00318099104143,
+      "grad_norm": 27.30816078186035,
+      "learning_rate": 4.454942764713202e-05,
+      "loss": 1.0046,
+      "step": 70800
+    },
+    {
+      "epoch": 39.0034609462486,
+      "grad_norm": 0.13129769265651703,
+      "learning_rate": 4.453387458006719e-05,
+      "loss": 1.0303,
+      "step": 70900
+    },
+    {
+      "epoch": 39.003740901455764,
+      "grad_norm": 1.489894986152649,
+      "learning_rate": 4.451832151300236e-05,
+      "loss": 1.0482,
+      "step": 71000
+    },
+    {
+      "epoch": 39.00402085666293,
+      "grad_norm": 38.621490478515625,
+      "learning_rate": 4.450276844593754e-05,
+      "loss": 0.7289,
+      "step": 71100
+    },
+    {
+      "epoch": 39.0043008118701,
+      "grad_norm": 1.9483338594436646,
+      "learning_rate": 4.448721537887271e-05,
+      "loss": 0.8243,
+      "step": 71200
+    },
+    {
+      "epoch": 39.004580767077265,
+      "grad_norm": 41.24127197265625,
+      "learning_rate": 4.447166231180789e-05,
+      "loss": 1.0618,
+      "step": 71300
+    },
+    {
+      "epoch": 39.004860722284434,
+      "grad_norm": 0.1476653516292572,
+      "learning_rate": 4.445610924474307e-05,
+      "loss": 0.9857,
+      "step": 71400
+    },
+    {
+      "epoch": 39.005000699888015,
+      "eval_accuracy": 0.2819203268641471,
+      "eval_f1": 0.24980865661407028,
+      "eval_loss": 4.2032647132873535,
+      "eval_precision": 0.2609707892266626,
+      "eval_recall": 0.2819203268641471,
+      "eval_runtime": 222.7035,
+      "eval_samples_per_second": 17.584,
+      "eval_steps_per_second": 8.792,
+      "eval_top_10_accuracy": 0.6861593462717058,
+      "eval_top_1_accuracy": 0.2814096016343207,
+      "eval_top_5_accuracy": 0.5829928498467825,
+      "step": 71450
+    },
+    {
+      "epoch": 40.00013997760358,
+      "grad_norm": 0.46101224422454834,
+      "learning_rate": 4.444055617767824e-05,
+      "loss": 0.8026,
+      "step": 71500
+    },
+    {
+      "epoch": 40.00041993281075,
+      "grad_norm": 1.734413504600525,
+      "learning_rate": 4.442500311061342e-05,
+      "loss": 0.4295,
+      "step": 71600
+    },
+    {
+      "epoch": 40.00069988801792,
+      "grad_norm": 0.5759927034378052,
+      "learning_rate": 4.440945004354859e-05,
+      "loss": 0.5855,
+      "step": 71700
+    },
+    {
+      "epoch": 40.00097984322508,
+      "grad_norm": 0.35544517636299133,
+      "learning_rate": 4.4394052507154415e-05,
+      "loss": 0.6253,
+      "step": 71800
+    },
+    {
+      "epoch": 40.00125979843225,
+      "grad_norm": 1.3745167255401611,
+      "learning_rate": 4.437849944008959e-05,
+      "loss": 0.7667,
+      "step": 71900
+    },
+    {
+      "epoch": 40.00153975363942,
+      "grad_norm": 38.11959457397461,
+      "learning_rate": 4.436294637302476e-05,
+      "loss": 0.7042,
+      "step": 72000
+    },
+    {
+      "epoch": 40.00181970884658,
+      "grad_norm": 0.17603681981563568,
+      "learning_rate": 4.434739330595994e-05,
+      "loss": 0.6599,
+      "step": 72100
+    },
+    {
+      "epoch": 40.00209966405375,
+      "grad_norm": 15.766741752624512,
+      "learning_rate": 4.433184023889511e-05,
+      "loss": 1.0079,
+      "step": 72200
+    },
+    {
+      "epoch": 40.00237961926092,
+      "grad_norm": 5.181120872497559,
+      "learning_rate": 4.431628717183029e-05,
+      "loss": 0.7113,
+      "step": 72300
+    },
+    {
+      "epoch": 40.00265957446808,
+      "grad_norm": 0.3363553285598755,
+      "learning_rate": 4.430073410476546e-05,
+      "loss": 0.8276,
+      "step": 72400
+    },
+    {
+      "epoch": 40.00293952967525,
+      "grad_norm": 33.76218795776367,
+      "learning_rate": 4.428518103770063e-05,
+      "loss": 0.8144,
+      "step": 72500
+    },
+    {
+      "epoch": 40.00321948488242,
+      "grad_norm": 0.08487240225076675,
+      "learning_rate": 4.426962797063581e-05,
+      "loss": 0.8064,
+      "step": 72600
+    },
+    {
+      "epoch": 40.003499440089584,
+      "grad_norm": 28.670927047729492,
+      "learning_rate": 4.425407490357098e-05,
+      "loss": 0.8729,
+      "step": 72700
+    },
+    {
+      "epoch": 40.00377939529675,
+      "grad_norm": 25.651275634765625,
+      "learning_rate": 4.423852183650616e-05,
+      "loss": 0.8185,
+      "step": 72800
+    },
+    {
+      "epoch": 40.00405935050392,
+      "grad_norm": 0.2041385918855667,
+      "learning_rate": 4.4222968769441333e-05,
+      "loss": 1.0474,
+      "step": 72900
+    },
+    {
+      "epoch": 40.004339305711085,
+      "grad_norm": 27.669574737548828,
+      "learning_rate": 4.420741570237651e-05,
+      "loss": 0.8934,
+      "step": 73000
+    },
+    {
+      "epoch": 40.004619260918254,
+      "grad_norm": 0.12979120016098022,
+      "learning_rate": 4.4191862635311684e-05,
+      "loss": 1.1923,
+      "step": 73100
+    },
+    {
+      "epoch": 40.00489921612542,
+      "grad_norm": 56.33195877075195,
+      "learning_rate": 4.417630956824686e-05,
+      "loss": 0.7679,
+      "step": 73200
+    },
+    {
+      "epoch": 40.005,
+      "eval_accuracy": 0.2844739530132789,
+      "eval_f1": 0.25178392253063286,
+      "eval_loss": 4.198288440704346,
+      "eval_precision": 0.2640017201048866,
+      "eval_recall": 0.2844739530132789,
+      "eval_runtime": 225.0507,
+      "eval_samples_per_second": 17.401,
+      "eval_steps_per_second": 8.7,
+      "eval_top_10_accuracy": 0.6833503575076609,
+      "eval_top_1_accuracy": 0.2844739530132789,
+      "eval_top_5_accuracy": 0.5860572012257406,
+      "step": 73236
+    },
+    {
+      "epoch": 41.00017847144457,
+      "grad_norm": 40.168697357177734,
+      "learning_rate": 4.4160756501182035e-05,
+      "loss": 0.8715,
+      "step": 73300
+    },
+    {
+      "epoch": 41.00045842665173,
+      "grad_norm": 20.46622085571289,
+      "learning_rate": 4.414520343411721e-05,
+      "loss": 0.5017,
+      "step": 73400
+    },
+    {
+      "epoch": 41.0007383818589,
+      "grad_norm": 0.17491298913955688,
+      "learning_rate": 4.4129650367052385e-05,
+      "loss": 0.6206,
+      "step": 73500
+    },
+    {
+      "epoch": 41.00101833706607,
+      "grad_norm": 19.896448135375977,
+      "learning_rate": 4.4114097299987564e-05,
+      "loss": 0.6051,
+      "step": 73600
+    },
+    {
+      "epoch": 41.00129829227323,
+      "grad_norm": 0.23201890289783478,
+      "learning_rate": 4.4098544232922736e-05,
+      "loss": 0.5626,
+      "step": 73700
+    },
+    {
+      "epoch": 41.0015782474804,
+      "grad_norm": 26.803802490234375,
+      "learning_rate": 4.4082991165857914e-05,
+      "loss": 0.7498,
+      "step": 73800
+    },
+    {
+      "epoch": 41.00185820268757,
+      "grad_norm": 0.059675004333257675,
+      "learning_rate": 4.4067438098793086e-05,
+      "loss": 0.5219,
+      "step": 73900
+    },
+    {
+      "epoch": 41.002138157894734,
+      "grad_norm": 1.5195122957229614,
+      "learning_rate": 4.405188503172826e-05,
+      "loss": 0.5764,
+      "step": 74000
+    },
+    {
+      "epoch": 41.0024181131019,
+      "grad_norm": 26.826339721679688,
+      "learning_rate": 4.403633196466344e-05,
+      "loss": 0.8804,
+      "step": 74100
+    },
+    {
+      "epoch": 41.00269806830907,
+      "grad_norm": 62.245506286621094,
+      "learning_rate": 4.402077889759861e-05,
+      "loss": 1.0212,
+      "step": 74200
+    },
+    {
+      "epoch": 41.002978023516235,
+      "grad_norm": 14.845590591430664,
+      "learning_rate": 4.400522583053379e-05,
+      "loss": 0.7602,
+      "step": 74300
+    },
+    {
+      "epoch": 41.003257978723404,
+      "grad_norm": 0.10195853561162949,
+      "learning_rate": 4.398967276346896e-05,
+      "loss": 0.9441,
+      "step": 74400
+    },
+    {
+      "epoch": 41.00353793393057,
+      "grad_norm": 20.029823303222656,
+      "learning_rate": 4.397411969640413e-05,
+      "loss": 0.9074,
+      "step": 74500
+    },
+    {
+      "epoch": 41.003817889137736,
+      "grad_norm": 0.1141628846526146,
+      "learning_rate": 4.395856662933931e-05,
+      "loss": 0.9763,
+      "step": 74600
+    },
+    {
+      "epoch": 41.004097844344905,
+      "grad_norm": 10.608264923095703,
+      "learning_rate": 4.394301356227448e-05,
+      "loss": 0.8389,
+      "step": 74700
+    },
+    {
+      "epoch": 41.004377799552074,
+      "grad_norm": 29.61349868774414,
+      "learning_rate": 4.3927460495209654e-05,
+      "loss": 1.0023,
+      "step": 74800
+    },
+    {
+      "epoch": 41.004657754759236,
+      "grad_norm": 40.82460403442383,
+      "learning_rate": 4.391190742814483e-05,
+      "loss": 0.783,
+      "step": 74900
+    },
+    {
+      "epoch": 41.004937709966406,
+      "grad_norm": 0.5806564092636108,
+      "learning_rate": 4.3896354361080004e-05,
+      "loss": 0.8991,
+      "step": 75000
+    },
+    {
+      "epoch": 41.00499930011198,
+      "eval_accuracy": 0.28115423901940756,
+      "eval_f1": 0.25330357027391814,
+      "eval_loss": 4.209905624389648,
+      "eval_precision": 0.2713199055512641,
+      "eval_recall": 0.28115423901940756,
+      "eval_runtime": 222.5071,
+      "eval_samples_per_second": 17.599,
+      "eval_steps_per_second": 8.8,
+      "eval_top_10_accuracy": 0.6882022471910112,
+      "eval_top_1_accuracy": 0.28115423901940756,
+      "eval_top_5_accuracy": 0.5873340143003064,
+      "step": 75022
+    },
+    {
+      "epoch": 42.00021696528555,
+      "grad_norm": 1.9568456411361694,
+      "learning_rate": 4.388080129401518e-05,
+      "loss": 0.5599,
+      "step": 75100
+    },
+    {
+      "epoch": 42.00049692049272,
+      "grad_norm": 0.06384823471307755,
+      "learning_rate": 4.3865248226950355e-05,
+      "loss": 0.4513,
+      "step": 75200
+    },
+    {
+      "epoch": 42.00077687569989,
+      "grad_norm": 37.387027740478516,
+      "learning_rate": 4.384969515988553e-05,
+      "loss": 0.6817,
+      "step": 75300
+    },
+    {
+      "epoch": 42.00105683090705,
+      "grad_norm": 3.1402885913848877,
+      "learning_rate": 4.3834142092820705e-05,
+      "loss": 0.579,
+      "step": 75400
+    },
+    {
+      "epoch": 42.00133678611422,
+      "grad_norm": 24.56217384338379,
+      "learning_rate": 4.381858902575588e-05,
+      "loss": 0.5672,
+      "step": 75500
+    },
+    {
+      "epoch": 42.00161674132139,
+      "grad_norm": 16.695140838623047,
+      "learning_rate": 4.3803035958691056e-05,
+      "loss": 0.7093,
+      "step": 75600
+    },
+    {
+      "epoch": 42.001896696528554,
+      "grad_norm": 0.4463357925415039,
+      "learning_rate": 4.378748289162623e-05,
+      "loss": 0.568,
+      "step": 75700
+    },
+    {
+      "epoch": 42.002176651735724,
+      "grad_norm": 10.364665985107422,
+      "learning_rate": 4.377208535523206e-05,
+      "loss": 0.7371,
+      "step": 75800
+    },
+    {
+      "epoch": 42.002456606942886,
+      "grad_norm": 0.2388359159231186,
+      "learning_rate": 4.375653228816723e-05,
+      "loss": 0.6993,
+      "step": 75900
+    },
+    {
+      "epoch": 42.002736562150055,
+      "grad_norm": 1.1912405490875244,
+      "learning_rate": 4.37409792211024e-05,
+      "loss": 0.58,
+      "step": 76000
+    },
+    {
+      "epoch": 42.003016517357224,
+      "grad_norm": 0.1253499835729599,
+      "learning_rate": 4.372542615403758e-05,
+      "loss": 0.8355,
+      "step": 76100
+    },
+    {
+      "epoch": 42.00329647256439,
+      "grad_norm": 0.13281774520874023,
+      "learning_rate": 4.370987308697275e-05,
+      "loss": 0.9043,
+      "step": 76200
+    },
+    {
+      "epoch": 42.003576427771556,
+      "grad_norm": 0.41041648387908936,
+      "learning_rate": 4.369432001990793e-05,
+      "loss": 0.7185,
+      "step": 76300
+    },
+    {
+      "epoch": 42.003856382978725,
+      "grad_norm": 50.560123443603516,
+      "learning_rate": 4.36787669528431e-05,
+      "loss": 1.075,
+      "step": 76400
+    },
+    {
+      "epoch": 42.00413633818589,
+      "grad_norm": 0.2037896066904068,
+      "learning_rate": 4.3663213885778275e-05,
+      "loss": 0.9476,
+      "step": 76500
+    },
+    {
+      "epoch": 42.00441629339306,
+      "grad_norm": 0.8026627898216248,
+      "learning_rate": 4.3647660818713454e-05,
+      "loss": 0.9955,
+      "step": 76600
+    },
+    {
+      "epoch": 42.004696248600226,
+      "grad_norm": 21.437389373779297,
+      "learning_rate": 4.3632107751648626e-05,
+      "loss": 1.0739,
+      "step": 76700
+    },
+    {
+      "epoch": 42.00497620380739,
+      "grad_norm": 13.451566696166992,
+      "learning_rate": 4.3616554684583804e-05,
+      "loss": 1.1176,
+      "step": 76800
+    },
+    {
+      "epoch": 42.00499860022396,
+      "eval_accuracy": 0.2765577119509704,
+      "eval_f1": 0.24552088339676698,
+      "eval_loss": 4.34188985824585,
+      "eval_precision": 0.2595869870982841,
+      "eval_recall": 0.2765577119509704,
+      "eval_runtime": 226.8112,
+      "eval_samples_per_second": 17.265,
+      "eval_steps_per_second": 8.633,
+      "eval_top_10_accuracy": 0.673391215526047,
+      "eval_top_1_accuracy": 0.27681307456588355,
+      "eval_top_5_accuracy": 0.5686925434116445,
+      "step": 76808
+    },
+    {
+      "epoch": 43.00025545912654,
+      "grad_norm": 0.03739053010940552,
+      "learning_rate": 4.3601001617518976e-05,
+      "loss": 0.4309,
+      "step": 76900
+    },
+    {
+      "epoch": 43.000535414333704,
+      "grad_norm": 0.10674834996461868,
+      "learning_rate": 4.358544855045415e-05,
+      "loss": 0.4421,
+      "step": 77000
+    },
+    {
+      "epoch": 43.000815369540874,
+      "grad_norm": 0.10748764872550964,
+      "learning_rate": 4.356989548338933e-05,
+      "loss": 0.6075,
+      "step": 77100
+    },
+    {
+      "epoch": 43.00109532474804,
+      "grad_norm": 9.194770812988281,
+      "learning_rate": 4.35543424163245e-05,
+      "loss": 0.6129,
+      "step": 77200
+    },
+    {
+      "epoch": 43.001375279955205,
+      "grad_norm": 0.1717982441186905,
+      "learning_rate": 4.353878934925968e-05,
+      "loss": 0.5179,
+      "step": 77300
+    },
+    {
+      "epoch": 43.001655235162374,
+      "grad_norm": 0.15550194680690765,
+      "learning_rate": 4.352323628219485e-05,
+      "loss": 0.7215,
+      "step": 77400
+    },
+    {
+      "epoch": 43.001935190369544,
+      "grad_norm": 0.15924519300460815,
+      "learning_rate": 4.350768321513002e-05,
+      "loss": 0.5473,
+      "step": 77500
+    },
+    {
+      "epoch": 43.002215145576706,
+      "grad_norm": 32.287445068359375,
+      "learning_rate": 4.34921301480652e-05,
+      "loss": 0.801,
+      "step": 77600
+    },
+    {
+      "epoch": 43.002495100783875,
+      "grad_norm": 0.5695823431015015,
+      "learning_rate": 4.347657708100037e-05,
+      "loss": 0.8792,
+      "step": 77700
+    },
+    {
+      "epoch": 43.002775055991044,
+      "grad_norm": 21.394012451171875,
+      "learning_rate": 4.346102401393555e-05,
+      "loss": 0.8703,
+      "step": 77800
+    },
+    {
+      "epoch": 43.00305501119821,
+      "grad_norm": 0.2652227580547333,
+      "learning_rate": 4.3445626477541374e-05,
+      "loss": 0.5404,
+      "step": 77900
+    },
+    {
+      "epoch": 43.003334966405376,
+      "grad_norm": 0.10218947380781174,
+      "learning_rate": 4.3430073410476546e-05,
+      "loss": 0.9025,
+      "step": 78000
+    },
+    {
+      "epoch": 43.003614921612545,
+      "grad_norm": 28.094913482666016,
+      "learning_rate": 4.3414520343411725e-05,
+      "loss": 0.989,
+      "step": 78100
+    },
+    {
+      "epoch": 43.00389487681971,
+      "grad_norm": 0.22375081479549408,
+      "learning_rate": 4.3398967276346897e-05,
+      "loss": 0.8974,
+      "step": 78200
+    },
+    {
+      "epoch": 43.00417483202688,
+      "grad_norm": 1.988282322883606,
+      "learning_rate": 4.3383414209282075e-05,
+      "loss": 0.7192,
+      "step": 78300
+    },
+    {
+      "epoch": 43.004454787234046,
+      "grad_norm": 32.35314178466797,
+      "learning_rate": 4.336786114221725e-05,
+      "loss": 0.8441,
+      "step": 78400
+    },
+    {
+      "epoch": 43.00473474244121,
+      "grad_norm": 27.084787368774414,
+      "learning_rate": 4.335230807515242e-05,
+      "loss": 1.2777,
+      "step": 78500
+    },
+    {
+      "epoch": 43.005000699888015,
+      "eval_accuracy": 0.27808988764044945,
+      "eval_f1": 0.2469082169774653,
+      "eval_loss": 4.310375213623047,
+      "eval_precision": 0.26096653920645235,
+      "eval_recall": 0.27808988764044945,
+      "eval_runtime": 223.359,
+      "eval_samples_per_second": 17.532,
+      "eval_steps_per_second": 8.766,
+      "eval_top_10_accuracy": 0.6741573033707865,
+      "eval_top_1_accuracy": 0.2783452502553626,
+      "eval_top_5_accuracy": 0.5773748723186926,
+      "step": 78595
+    },
+    {
+      "epoch": 44.000013997760355,
+      "grad_norm": 30.237857818603516,
+      "learning_rate": 4.33367550080876e-05,
+      "loss": 1.0311,
+      "step": 78600
+    },
+    {
+      "epoch": 44.000293952967525,
+      "grad_norm": 57.867767333984375,
+      "learning_rate": 4.332120194102277e-05,
+      "loss": 0.3452,
+      "step": 78700
+    },
+    {
+      "epoch": 44.000573908174694,
+      "grad_norm": 0.10751175135374069,
+      "learning_rate": 4.330564887395795e-05,
+      "loss": 0.4273,
+      "step": 78800
+    },
+    {
+      "epoch": 44.000853863381856,
+      "grad_norm": 3.9307987689971924,
+      "learning_rate": 4.329009580689312e-05,
+      "loss": 0.5516,
+      "step": 78900
+    },
+    {
+      "epoch": 44.001133818589025,
+      "grad_norm": 31.563825607299805,
+      "learning_rate": 4.327454273982829e-05,
+      "loss": 0.5658,
+      "step": 79000
+    },
+    {
+      "epoch": 44.001413773796195,
+      "grad_norm": 0.5761617422103882,
+      "learning_rate": 4.325898967276347e-05,
+      "loss": 0.8374,
+      "step": 79100
+    },
+    {
+      "epoch": 44.00169372900336,
+      "grad_norm": 5.779951572418213,
+      "learning_rate": 4.324343660569864e-05,
+      "loss": 0.6332,
+      "step": 79200
+    },
+    {
+      "epoch": 44.001973684210526,
+      "grad_norm": 25.423784255981445,
+      "learning_rate": 4.322788353863382e-05,
+      "loss": 0.6223,
+      "step": 79300
+    },
+    {
+      "epoch": 44.002253639417695,
+      "grad_norm": 15.803873062133789,
+      "learning_rate": 4.321233047156899e-05,
+      "loss": 0.5489,
+      "step": 79400
+    },
+    {
+      "epoch": 44.00253359462486,
+      "grad_norm": 0.1971481591463089,
+      "learning_rate": 4.3196777404504165e-05,
+      "loss": 0.7783,
+      "step": 79500
+    },
+    {
+      "epoch": 44.00281354983203,
+      "grad_norm": 2.9763810634613037,
+      "learning_rate": 4.3181224337439344e-05,
+      "loss": 0.8164,
+      "step": 79600
+    },
+    {
+      "epoch": 44.003093505039196,
+      "grad_norm": 26.30331039428711,
+      "learning_rate": 4.316567127037452e-05,
+      "loss": 1.0041,
+      "step": 79700
+    },
+    {
+      "epoch": 44.00337346024636,
+      "grad_norm": 44.34019470214844,
+      "learning_rate": 4.3150118203309694e-05,
+      "loss": 0.8961,
+      "step": 79800
+    },
+    {
+      "epoch": 44.00365341545353,
+      "grad_norm": 26.827043533325195,
+      "learning_rate": 4.313472066691552e-05,
+      "loss": 0.9644,
+      "step": 79900
+    },
+    {
+      "epoch": 44.0039333706607,
+      "grad_norm": 7.778226375579834,
+      "learning_rate": 4.311916759985069e-05,
+      "loss": 1.0374,
+      "step": 80000
+    },
+    {
+      "epoch": 44.00421332586786,
+      "grad_norm": 41.65769958496094,
+      "learning_rate": 4.310361453278587e-05,
+      "loss": 0.9236,
+      "step": 80100
+    },
+    {
+      "epoch": 44.00449328107503,
+      "grad_norm": 37.12181091308594,
+      "learning_rate": 4.308806146572104e-05,
+      "loss": 0.7795,
+      "step": 80200
+    },
+    {
+      "epoch": 44.0047732362822,
+      "grad_norm": 4.020895481109619,
+      "learning_rate": 4.307250839865622e-05,
+      "loss": 0.8072,
+      "step": 80300
+    },
+    {
+      "epoch": 44.005,
+      "eval_accuracy": 0.27808988764044945,
+      "eval_f1": 0.2500487491688747,
+      "eval_loss": 4.37076997756958,
+      "eval_precision": 0.2685824383028162,
+      "eval_recall": 0.27808988764044945,
+      "eval_runtime": 221.8318,
+      "eval_samples_per_second": 17.653,
+      "eval_steps_per_second": 8.827,
+      "eval_top_10_accuracy": 0.6616445352400409,
+      "eval_top_1_accuracy": 0.27783452502553624,
+      "eval_top_5_accuracy": 0.5638406537282942,
+      "step": 80381
+    },
+    {
+      "epoch": 45.000052491601345,
+      "grad_norm": 5.015646457672119,
+      "learning_rate": 4.305695533159139e-05,
+      "loss": 0.7486,
+      "step": 80400
+    },
+    {
+      "epoch": 45.000332446808514,
+      "grad_norm": 10.068531036376953,
+      "learning_rate": 4.304140226452656e-05,
+      "loss": 0.4467,
+      "step": 80500
+    },
+    {
+      "epoch": 45.000612402015676,
+      "grad_norm": 26.70047950744629,
+      "learning_rate": 4.302584919746174e-05,
+      "loss": 0.48,
+      "step": 80600
+    },
+    {
+      "epoch": 45.000892357222845,
+      "grad_norm": 0.11867395788431168,
+      "learning_rate": 4.3010296130396914e-05,
+      "loss": 0.6253,
+      "step": 80700
+    },
+    {
+      "epoch": 45.001172312430015,
+      "grad_norm": 39.205528259277344,
+      "learning_rate": 4.299474306333209e-05,
+      "loss": 0.6083,
+      "step": 80800
+    },
+    {
+      "epoch": 45.00145226763718,
+      "grad_norm": 0.874306857585907,
+      "learning_rate": 4.2979189996267264e-05,
+      "loss": 0.5311,
+      "step": 80900
+    },
+    {
+      "epoch": 45.001732222844346,
+      "grad_norm": 0.12406778335571289,
+      "learning_rate": 4.2963636929202436e-05,
+      "loss": 0.5437,
+      "step": 81000
+    },
+    {
+      "epoch": 45.00201217805151,
+      "grad_norm": 1.986485481262207,
+      "learning_rate": 4.2948083862137615e-05,
+      "loss": 0.6653,
+      "step": 81100
+    },
+    {
+      "epoch": 45.00229213325868,
+      "grad_norm": 0.10207056999206543,
+      "learning_rate": 4.2932530795072787e-05,
+      "loss": 0.9323,
+      "step": 81200
+    },
+    {
+      "epoch": 45.00257208846585,
+      "grad_norm": 0.07560595870018005,
+      "learning_rate": 4.2916977728007965e-05,
+      "loss": 0.6401,
+      "step": 81300
+    },
+    {
+      "epoch": 45.00285204367301,
+      "grad_norm": 29.77736473083496,
+      "learning_rate": 4.290142466094314e-05,
+      "loss": 0.7685,
+      "step": 81400
+    },
+    {
+      "epoch": 45.00313199888018,
+      "grad_norm": 0.9332723617553711,
+      "learning_rate": 4.2885871593878316e-05,
+      "loss": 0.8516,
+      "step": 81500
+    },
+    {
+      "epoch": 45.00341195408735,
+      "grad_norm": 0.36157646775245667,
+      "learning_rate": 4.287031852681349e-05,
+      "loss": 0.6851,
+      "step": 81600
+    },
+    {
+      "epoch": 45.00369190929451,
+      "grad_norm": 1.132853388786316,
+      "learning_rate": 4.2854765459748666e-05,
+      "loss": 0.9181,
+      "step": 81700
+    },
+    {
+      "epoch": 45.00397186450168,
+      "grad_norm": 33.407169342041016,
+      "learning_rate": 4.2839212392683845e-05,
+      "loss": 0.8689,
+      "step": 81800
+    },
+    {
+      "epoch": 45.00425181970885,
+      "grad_norm": 1.4110562801361084,
+      "learning_rate": 4.282381485628966e-05,
+      "loss": 0.8921,
+      "step": 81900
+    },
+    {
+      "epoch": 45.00453177491601,
+      "grad_norm": 0.2189277559518814,
+      "learning_rate": 4.2808261789224834e-05,
+      "loss": 0.8344,
+      "step": 82000
+    },
+    {
+      "epoch": 45.00481173012318,
+      "grad_norm": 16.592561721801758,
+      "learning_rate": 4.279270872216001e-05,
+      "loss": 1.1258,
+      "step": 82100
+    },
+    {
+      "epoch": 45.00499930011198,
+      "eval_accuracy": 0.2801327885597549,
+      "eval_f1": 0.2509663358999416,
+      "eval_loss": 4.351675987243652,
+      "eval_precision": 0.26884382499806403,
+      "eval_recall": 0.2801327885597549,
+      "eval_runtime": 223.1566,
+      "eval_samples_per_second": 17.548,
+      "eval_steps_per_second": 8.774,
+      "eval_top_10_accuracy": 0.6731358529111338,
+      "eval_top_1_accuracy": 0.27987742594484166,
+      "eval_top_5_accuracy": 0.5707354443309499,
+      "step": 82167
+    },
+    {
+      "epoch": 46.00009098544233,
+      "grad_norm": 8.000289916992188,
+      "learning_rate": 4.2777155655095184e-05,
+      "loss": 0.8636,
+      "step": 82200
+    },
+    {
+      "epoch": 46.000370940649496,
+      "grad_norm": 0.263664573431015,
+      "learning_rate": 4.276160258803036e-05,
+      "loss": 0.3495,
+      "step": 82300
+    },
+    {
+      "epoch": 46.000650895856666,
+      "grad_norm": 20.22283935546875,
+      "learning_rate": 4.2746049520965535e-05,
+      "loss": 0.6638,
+      "step": 82400
+    },
+    {
+      "epoch": 46.00093085106383,
+      "grad_norm": 20.6286678314209,
+      "learning_rate": 4.273049645390071e-05,
+      "loss": 0.4812,
+      "step": 82500
+    },
+    {
+      "epoch": 46.001210806271,
+      "grad_norm": 1.6630127429962158,
+      "learning_rate": 4.2714943386835886e-05,
+      "loss": 0.654,
+      "step": 82600
+    },
+    {
+      "epoch": 46.001490761478166,
+      "grad_norm": 28.354536056518555,
+      "learning_rate": 4.269939031977106e-05,
+      "loss": 0.6758,
+      "step": 82700
+    },
+    {
+      "epoch": 46.00177071668533,
+      "grad_norm": 0.040363699197769165,
+      "learning_rate": 4.2683837252706236e-05,
+      "loss": 0.5949,
+      "step": 82800
+    },
+    {
+      "epoch": 46.0020506718925,
+      "grad_norm": 0.522068440914154,
+      "learning_rate": 4.266828418564141e-05,
+      "loss": 0.6394,
+      "step": 82900
+    },
+    {
+      "epoch": 46.00233062709967,
+      "grad_norm": 0.14378629624843597,
+      "learning_rate": 4.265273111857658e-05,
+      "loss": 0.7237,
+      "step": 83000
+    },
+    {
+      "epoch": 46.00261058230683,
+      "grad_norm": 25.07221031188965,
+      "learning_rate": 4.263717805151176e-05,
+      "loss": 0.6259,
+      "step": 83100
+    },
+    {
+      "epoch": 46.002890537514,
+      "grad_norm": 3.614312171936035,
+      "learning_rate": 4.262162498444694e-05,
+      "loss": 0.6505,
+      "step": 83200
+    },
+    {
+      "epoch": 46.00317049272117,
+      "grad_norm": 23.773239135742188,
+      "learning_rate": 4.260607191738211e-05,
+      "loss": 0.6757,
+      "step": 83300
+    },
+    {
+      "epoch": 46.00345044792833,
+      "grad_norm": 19.163311004638672,
+      "learning_rate": 4.259051885031729e-05,
+      "loss": 0.736,
+      "step": 83400
+    },
+    {
+      "epoch": 46.0037304031355,
+      "grad_norm": 26.841463088989258,
+      "learning_rate": 4.257496578325246e-05,
+      "loss": 0.7399,
+      "step": 83500
+    },
+    {
+      "epoch": 46.00401035834267,
+      "grad_norm": 0.11447547376155853,
+      "learning_rate": 4.255941271618764e-05,
+      "loss": 0.9328,
+      "step": 83600
+    },
+    {
+      "epoch": 46.00429031354983,
+      "grad_norm": 0.1793317198753357,
+      "learning_rate": 4.254385964912281e-05,
+      "loss": 0.6467,
+      "step": 83700
+    },
+    {
+      "epoch": 46.004570268757,
+      "grad_norm": 37.40285873413086,
+      "learning_rate": 4.252830658205799e-05,
+      "loss": 1.0173,
+      "step": 83800
+    },
+    {
+      "epoch": 46.00485022396417,
+      "grad_norm": 0.11013773828744888,
+      "learning_rate": 4.251275351499316e-05,
+      "loss": 0.9476,
+      "step": 83900
+    },
+    {
+      "epoch": 46.00499860022396,
+      "eval_accuracy": 0.28115423901940756,
+      "eval_f1": 0.24729637104677424,
+      "eval_loss": 4.371170520782471,
+      "eval_precision": 0.2582944003845434,
+      "eval_recall": 0.28115423901940756,
+      "eval_runtime": 223.6689,
+      "eval_samples_per_second": 17.508,
+      "eval_steps_per_second": 8.754,
+      "eval_top_10_accuracy": 0.6756894790602656,
+      "eval_top_1_accuracy": 0.2814096016343207,
+      "eval_top_5_accuracy": 0.5684371807967313,
+      "step": 83953
+    },
+    {
+      "epoch": 47.000129479283316,
+      "grad_norm": 0.257235050201416,
+      "learning_rate": 4.249735597859898e-05,
+      "loss": 0.7448,
+      "step": 84000
+    },
+    {
+      "epoch": 47.00040943449048,
+      "grad_norm": 18.50094985961914,
+      "learning_rate": 4.2481802911534156e-05,
+      "loss": 0.3183,
+      "step": 84100
+    },
+    {
+      "epoch": 47.00068938969765,
+      "grad_norm": 0.018771247938275337,
+      "learning_rate": 4.246624984446933e-05,
+      "loss": 0.6396,
+      "step": 84200
+    },
+    {
+      "epoch": 47.00096934490482,
+      "grad_norm": 0.19734026491641998,
+      "learning_rate": 4.245069677740451e-05,
+      "loss": 0.4782,
+      "step": 84300
+    },
+    {
+      "epoch": 47.00124930011198,
+      "grad_norm": 4.257106304168701,
+      "learning_rate": 4.243514371033968e-05,
+      "loss": 0.6645,
+      "step": 84400
+    },
+    {
+      "epoch": 47.00152925531915,
+      "grad_norm": 14.478767395019531,
+      "learning_rate": 4.241959064327485e-05,
+      "loss": 0.5289,
+      "step": 84500
+    },
+    {
+      "epoch": 47.00180921052632,
+      "grad_norm": 0.04039764776825905,
+      "learning_rate": 4.240403757621003e-05,
+      "loss": 0.6705,
+      "step": 84600
+    },
+    {
+      "epoch": 47.00208916573348,
+      "grad_norm": 31.222604751586914,
+      "learning_rate": 4.23884845091452e-05,
+      "loss": 0.5118,
+      "step": 84700
+    },
+    {
+      "epoch": 47.00236912094065,
+      "grad_norm": 4.133471488952637,
+      "learning_rate": 4.237293144208038e-05,
+      "loss": 0.716,
+      "step": 84800
+    },
+    {
+      "epoch": 47.00264907614782,
+      "grad_norm": 0.13676881790161133,
+      "learning_rate": 4.235737837501555e-05,
+      "loss": 0.6143,
+      "step": 84900
+    },
+    {
+      "epoch": 47.00292903135498,
+      "grad_norm": 0.7570433616638184,
+      "learning_rate": 4.234182530795073e-05,
+      "loss": 0.6359,
+      "step": 85000
+    },
+    {
+      "epoch": 47.00320898656215,
+      "grad_norm": 30.351444244384766,
+      "learning_rate": 4.23262722408859e-05,
+      "loss": 0.5945,
+      "step": 85100
+    },
+    {
+      "epoch": 47.00348894176932,
+      "grad_norm": 0.016572458669543266,
+      "learning_rate": 4.231071917382108e-05,
+      "loss": 0.9843,
+      "step": 85200
+    },
+    {
+      "epoch": 47.00376889697648,
+      "grad_norm": 4.526513576507568,
+      "learning_rate": 4.229516610675626e-05,
+      "loss": 0.8844,
+      "step": 85300
+    },
+    {
+      "epoch": 47.00404885218365,
+      "grad_norm": 19.184106826782227,
+      "learning_rate": 4.227961303969143e-05,
+      "loss": 0.845,
+      "step": 85400
+    },
+    {
+      "epoch": 47.00432880739082,
+      "grad_norm": 0.7215138077735901,
+      "learning_rate": 4.2264059972626604e-05,
+      "loss": 0.7167,
+      "step": 85500
+    },
+    {
+      "epoch": 47.00460876259798,
+      "grad_norm": 0.3852091133594513,
+      "learning_rate": 4.224850690556178e-05,
+      "loss": 0.9227,
+      "step": 85600
+    },
+    {
+      "epoch": 47.00488871780515,
+      "grad_norm": 20.54586410522461,
+      "learning_rate": 4.2232953838496954e-05,
+      "loss": 1.0757,
+      "step": 85700
+    },
+    {
+      "epoch": 47.005000699888015,
+      "eval_accuracy": 0.2676200204290092,
+      "eval_f1": 0.23724926687771428,
+      "eval_loss": 4.428029537200928,
+      "eval_precision": 0.251943858264083,
+      "eval_recall": 0.2676200204290092,
+      "eval_runtime": 219.2253,
+      "eval_samples_per_second": 17.863,
+      "eval_steps_per_second": 8.931,
+      "eval_top_10_accuracy": 0.6667517875383044,
+      "eval_top_1_accuracy": 0.2678753830439224,
+      "eval_top_5_accuracy": 0.5640960163432074,
+      "step": 85740
+    },
+    {
+      "epoch": 48.0001679731243,
+      "grad_norm": 1.403875708580017,
+      "learning_rate": 4.221740077143213e-05,
+      "loss": 0.6773,
+      "step": 85800
+    },
+    {
+      "epoch": 48.00044792833147,
+      "grad_norm": 1.0571541786193848,
+      "learning_rate": 4.2201847704367305e-05,
+      "loss": 0.4237,
+      "step": 85900
+    },
+    {
+      "epoch": 48.00072788353864,
+      "grad_norm": 0.06935632228851318,
+      "learning_rate": 4.218645016797312e-05,
+      "loss": 0.6378,
+      "step": 86000
+    },
+    {
+      "epoch": 48.0010078387458,
+      "grad_norm": 0.08270786702632904,
+      "learning_rate": 4.21708971009083e-05,
+      "loss": 0.56,
+      "step": 86100
+    },
+    {
+      "epoch": 48.00128779395297,
+      "grad_norm": 19.250934600830078,
+      "learning_rate": 4.215534403384347e-05,
+      "loss": 0.5513,
+      "step": 86200
+    },
+    {
+      "epoch": 48.00156774916013,
+      "grad_norm": 0.13220156729221344,
+      "learning_rate": 4.213979096677865e-05,
+      "loss": 0.5792,
+      "step": 86300
+    },
+    {
+      "epoch": 48.0018477043673,
+      "grad_norm": 19.1898250579834,
+      "learning_rate": 4.212423789971382e-05,
+      "loss": 0.7476,
+      "step": 86400
+    },
+    {
+      "epoch": 48.00212765957447,
+      "grad_norm": 35.42184829711914,
+      "learning_rate": 4.2108684832648995e-05,
+      "loss": 0.9539,
+      "step": 86500
+    },
+    {
+      "epoch": 48.00240761478163,
+      "grad_norm": 14.540600776672363,
+      "learning_rate": 4.209313176558417e-05,
+      "loss": 0.5537,
+      "step": 86600
+    },
+    {
+      "epoch": 48.0026875699888,
+      "grad_norm": 0.27380481362342834,
+      "learning_rate": 4.207757869851935e-05,
+      "loss": 0.6941,
+      "step": 86700
+    },
+    {
+      "epoch": 48.00296752519597,
+      "grad_norm": 0.0291057787835598,
+      "learning_rate": 4.2062025631454524e-05,
+      "loss": 0.989,
+      "step": 86800
+    },
+    {
+      "epoch": 48.00324748040313,
+      "grad_norm": 23.343826293945312,
+      "learning_rate": 4.20464725643897e-05,
+      "loss": 0.9426,
+      "step": 86900
+    },
+    {
+      "epoch": 48.0035274356103,
+      "grad_norm": 18.877023696899414,
+      "learning_rate": 4.2030919497324875e-05,
+      "loss": 0.6116,
+      "step": 87000
+    },
+    {
+      "epoch": 48.00380739081747,
+      "grad_norm": 52.804725646972656,
+      "learning_rate": 4.201536643026005e-05,
+      "loss": 0.8468,
+      "step": 87100
+    },
+    {
+      "epoch": 48.00408734602463,
+      "grad_norm": 0.39581963419914246,
+      "learning_rate": 4.1999813363195225e-05,
+      "loss": 0.6277,
+      "step": 87200
+    },
+    {
+      "epoch": 48.0043673012318,
+      "grad_norm": 0.08678244054317474,
+      "learning_rate": 4.1984260296130404e-05,
+      "loss": 0.734,
+      "step": 87300
+    },
+    {
+      "epoch": 48.00464725643897,
+      "grad_norm": 37.76765060424805,
+      "learning_rate": 4.1968707229065576e-05,
+      "loss": 0.8369,
+      "step": 87400
+    },
+    {
+      "epoch": 48.004927211646134,
+      "grad_norm": 25.679229736328125,
+      "learning_rate": 4.195315416200075e-05,
+      "loss": 0.7244,
+      "step": 87500
+    },
+    {
+      "epoch": 48.005,
+      "eval_accuracy": 0.267364657814096,
+      "eval_f1": 0.23883580701912732,
+      "eval_loss": 4.437381267547607,
+      "eval_precision": 0.2528897338332987,
+      "eval_recall": 0.267364657814096,
+      "eval_runtime": 227.9408,
+      "eval_samples_per_second": 17.18,
+      "eval_steps_per_second": 8.59,
+      "eval_top_10_accuracy": 0.6685393258426966,
+      "eval_top_1_accuracy": 0.2676200204290092,
+      "eval_top_5_accuracy": 0.5676710929519918,
+      "step": 87526
+    },
+    {
+      "epoch": 49.00020646696529,
+      "grad_norm": 0.036201562732458115,
+      "learning_rate": 4.1937601094935926e-05,
+      "loss": 0.4116,
+      "step": 87600
+    },
+    {
+      "epoch": 49.00048642217245,
+      "grad_norm": 0.03052523173391819,
+      "learning_rate": 4.19220480278711e-05,
+      "loss": 0.4815,
+      "step": 87700
+    },
+    {
+      "epoch": 49.00076637737962,
+      "grad_norm": 77.27123260498047,
+      "learning_rate": 4.190649496080628e-05,
+      "loss": 0.5239,
+      "step": 87800
+    },
+    {
+      "epoch": 49.00104633258679,
+      "grad_norm": 36.39849090576172,
+      "learning_rate": 4.189094189374145e-05,
+      "loss": 0.493,
+      "step": 87900
+    },
+    {
+      "epoch": 49.00132628779395,
+      "grad_norm": 34.98275375366211,
+      "learning_rate": 4.187538882667662e-05,
+      "loss": 0.7586,
+      "step": 88000
+    },
+    {
+      "epoch": 49.00160624300112,
+      "grad_norm": 2.225522041320801,
+      "learning_rate": 4.18598357596118e-05,
+      "loss": 0.4518,
+      "step": 88100
+    },
+    {
+      "epoch": 49.00188619820829,
+      "grad_norm": 20.050642013549805,
+      "learning_rate": 4.1844438223217616e-05,
+      "loss": 0.6425,
+      "step": 88200
+    },
+    {
+      "epoch": 49.00216615341545,
+      "grad_norm": 26.255083084106445,
+      "learning_rate": 4.1828885156152795e-05,
+      "loss": 0.5254,
+      "step": 88300
+    },
+    {
+      "epoch": 49.00244610862262,
+      "grad_norm": 68.98938751220703,
+      "learning_rate": 4.181333208908797e-05,
+      "loss": 0.5639,
+      "step": 88400
+    },
+    {
+      "epoch": 49.00272606382979,
+      "grad_norm": 0.42682796716690063,
+      "learning_rate": 4.1797779022023145e-05,
+      "loss": 0.5954,
+      "step": 88500
+    },
+    {
+      "epoch": 49.00300601903695,
+      "grad_norm": 0.07504092901945114,
+      "learning_rate": 4.178222595495832e-05,
+      "loss": 0.794,
+      "step": 88600
+    },
+    {
+      "epoch": 49.00328597424412,
+      "grad_norm": 2.559800148010254,
+      "learning_rate": 4.1766672887893496e-05,
+      "loss": 0.7073,
+      "step": 88700
+    },
+    {
+      "epoch": 49.00356592945129,
+      "grad_norm": 0.014354785904288292,
+      "learning_rate": 4.1751119820828675e-05,
+      "loss": 0.851,
+      "step": 88800
+    },
+    {
+      "epoch": 49.00384588465845,
+      "grad_norm": 0.1795697808265686,
+      "learning_rate": 4.1735566753763847e-05,
+      "loss": 0.9604,
+      "step": 88900
+    },
+    {
+      "epoch": 49.00412583986562,
+      "grad_norm": 27.99577522277832,
+      "learning_rate": 4.172001368669902e-05,
+      "loss": 0.5985,
+      "step": 89000
+    },
+    {
+      "epoch": 49.00440579507279,
+      "grad_norm": 24.083967208862305,
+      "learning_rate": 4.17044606196342e-05,
+      "loss": 0.9174,
+      "step": 89100
+    },
+    {
+      "epoch": 49.004685750279954,
+      "grad_norm": 0.6158522963523865,
+      "learning_rate": 4.168890755256937e-05,
+      "loss": 0.6783,
+      "step": 89200
+    },
+    {
+      "epoch": 49.00496570548712,
+      "grad_norm": 0.4537813067436218,
+      "learning_rate": 4.167335448550455e-05,
+      "loss": 0.9193,
+      "step": 89300
+    },
+    {
+      "epoch": 49.00499930011198,
+      "eval_accuracy": 0.27477017364657813,
+      "eval_f1": 0.2466602161316155,
+      "eval_loss": 4.401386260986328,
+      "eval_precision": 0.2617121945995394,
+      "eval_recall": 0.27477017364657813,
+      "eval_runtime": 219.9162,
+      "eval_samples_per_second": 17.807,
+      "eval_steps_per_second": 8.903,
+      "eval_top_10_accuracy": 0.666241062308478,
+      "eval_top_1_accuracy": 0.27477017364657813,
+      "eval_top_5_accuracy": 0.563585291113381,
+      "step": 89312
+    },
+    {
+      "epoch": 50.00024496080627,
+      "grad_norm": 0.04477471113204956,
+      "learning_rate": 4.165780141843972e-05,
+      "loss": 0.3977,
+      "step": 89400
+    },
+    {
+      "epoch": 50.00052491601344,
+      "grad_norm": 0.24003936350345612,
+      "learning_rate": 4.164224835137489e-05,
+      "loss": 0.4872,
+      "step": 89500
+    },
+    {
+      "epoch": 50.0008048712206,
+      "grad_norm": 27.157386779785156,
+      "learning_rate": 4.162669528431007e-05,
+      "loss": 0.5601,
+      "step": 89600
+    },
+    {
+      "epoch": 50.00108482642777,
+      "grad_norm": 0.051918625831604004,
+      "learning_rate": 4.161114221724524e-05,
+      "loss": 0.4453,
+      "step": 89700
+    },
+    {
+      "epoch": 50.00136478163494,
+      "grad_norm": 5.718064785003662,
+      "learning_rate": 4.159558915018042e-05,
+      "loss": 0.4906,
+      "step": 89800
+    },
+    {
+      "epoch": 50.0016447368421,
+      "grad_norm": 0.8962364792823792,
+      "learning_rate": 4.158003608311559e-05,
+      "loss": 0.5037,
+      "step": 89900
+    },
+    {
+      "epoch": 50.00192469204927,
+      "grad_norm": 0.05446602776646614,
+      "learning_rate": 4.1564483016050765e-05,
+      "loss": 0.61,
+      "step": 90000
+    },
+    {
+      "epoch": 50.00220464725644,
+      "grad_norm": 19.462753295898438,
+      "learning_rate": 4.154892994898594e-05,
+      "loss": 0.6704,
+      "step": 90100
+    },
+    {
+      "epoch": 50.0024846024636,
+      "grad_norm": 0.09495168179273605,
+      "learning_rate": 4.1533376881921115e-05,
+      "loss": 0.6498,
+      "step": 90200
+    },
+    {
+      "epoch": 50.00276455767077,
+      "grad_norm": 1.4602558612823486,
+      "learning_rate": 4.151797934552694e-05,
+      "loss": 0.7506,
+      "step": 90300
+    },
+    {
+      "epoch": 50.00304451287794,
+      "grad_norm": 21.6412296295166,
+      "learning_rate": 4.150242627846212e-05,
+      "loss": 0.75,
+      "step": 90400
+    },
+    {
+      "epoch": 50.003324468085104,
+      "grad_norm": 2.719374418258667,
+      "learning_rate": 4.148687321139729e-05,
+      "loss": 0.708,
+      "step": 90500
+    },
+    {
+      "epoch": 50.00360442329227,
+      "grad_norm": 5.177334308624268,
+      "learning_rate": 4.147132014433247e-05,
+      "loss": 0.8531,
+      "step": 90600
+    },
+    {
+      "epoch": 50.00388437849944,
+      "grad_norm": 38.69509506225586,
+      "learning_rate": 4.145576707726764e-05,
+      "loss": 0.886,
+      "step": 90700
+    },
+    {
+      "epoch": 50.004164333706605,
+      "grad_norm": 0.25815537571907043,
+      "learning_rate": 4.144021401020282e-05,
+      "loss": 0.7835,
+      "step": 90800
+    },
+    {
+      "epoch": 50.004444288913774,
+      "grad_norm": 27.770305633544922,
+      "learning_rate": 4.142466094313799e-05,
+      "loss": 0.8142,
+      "step": 90900
+    },
+    {
+      "epoch": 50.00472424412094,
+      "grad_norm": 18.81846046447754,
+      "learning_rate": 4.140910787607316e-05,
+      "loss": 0.8372,
+      "step": 91000
+    },
+    {
+      "epoch": 50.00499860022396,
+      "eval_accuracy": 0.27349336057201223,
+      "eval_f1": 0.24285441862576393,
+      "eval_loss": 4.476330280303955,
+      "eval_precision": 0.2554687395112048,
+      "eval_recall": 0.27349336057201223,
+      "eval_runtime": 227.1325,
+      "eval_samples_per_second": 17.241,
+      "eval_steps_per_second": 8.621,
+      "eval_top_10_accuracy": 0.6544943820224719,
+      "eval_top_1_accuracy": 0.27349336057201223,
+      "eval_top_5_accuracy": 0.5584780388151175,
+      "step": 91098
+    },
+    {
+      "epoch": 51.00000349944009,
+      "grad_norm": 0.6531506180763245,
+      "learning_rate": 4.139355480900834e-05,
+      "loss": 0.8241,
+      "step": 91100
+    },
+    {
+      "epoch": 51.00028345464726,
+      "grad_norm": 0.014770712703466415,
+      "learning_rate": 4.137800174194351e-05,
+      "loss": 0.3236,
+      "step": 91200
+    },
+    {
+      "epoch": 51.00056340985442,
+      "grad_norm": 27.77800178527832,
+      "learning_rate": 4.136244867487869e-05,
+      "loss": 0.3143,
+      "step": 91300
+    },
+    {
+      "epoch": 51.00084336506159,
+      "grad_norm": 0.24872513115406036,
+      "learning_rate": 4.1346895607813863e-05,
+      "loss": 0.3622,
+      "step": 91400
+    },
+    {
+      "epoch": 51.00112332026875,
+      "grad_norm": 17.729570388793945,
+      "learning_rate": 4.1331342540749035e-05,
+      "loss": 0.5148,
+      "step": 91500
+    },
+    {
+      "epoch": 51.00140327547592,
+      "grad_norm": 0.36014020442962646,
+      "learning_rate": 4.1315789473684214e-05,
+      "loss": 0.4913,
+      "step": 91600
+    },
+    {
+      "epoch": 51.00168323068309,
+      "grad_norm": 0.2546406090259552,
+      "learning_rate": 4.1300236406619386e-05,
+      "loss": 0.7763,
+      "step": 91700
+    },
+    {
+      "epoch": 51.001963185890254,
+      "grad_norm": 0.051592834293842316,
+      "learning_rate": 4.1284683339554565e-05,
+      "loss": 0.5542,
+      "step": 91800
+    },
+    {
+      "epoch": 51.00224314109742,
+      "grad_norm": 0.3881039321422577,
+      "learning_rate": 4.1269130272489737e-05,
+      "loss": 0.6371,
+      "step": 91900
+    },
+    {
+      "epoch": 51.00252309630459,
+      "grad_norm": 6.290889739990234,
+      "learning_rate": 4.125357720542491e-05,
+      "loss": 0.6698,
+      "step": 92000
+    },
+    {
+      "epoch": 51.002803051511755,
+      "grad_norm": 13.433815002441406,
+      "learning_rate": 4.123802413836009e-05,
+      "loss": 0.8478,
+      "step": 92100
+    },
+    {
+      "epoch": 51.003083006718924,
+      "grad_norm": 21.983304977416992,
+      "learning_rate": 4.122247107129526e-05,
+      "loss": 0.7023,
+      "step": 92200
+    },
+    {
+      "epoch": 51.00336296192609,
+      "grad_norm": 24.535905838012695,
+      "learning_rate": 4.120691800423044e-05,
+      "loss": 0.7172,
+      "step": 92300
+    },
+    {
+      "epoch": 51.003642917133256,
+      "grad_norm": 0.10872181504964828,
+      "learning_rate": 4.119152046783626e-05,
+      "loss": 0.767,
+      "step": 92400
+    },
+    {
+      "epoch": 51.003922872340425,
+      "grad_norm": 1.7466179132461548,
+      "learning_rate": 4.117596740077143e-05,
+      "loss": 0.8614,
+      "step": 92500
+    },
+    {
+      "epoch": 51.004202827547594,
+      "grad_norm": 24.38710594177246,
+      "learning_rate": 4.116041433370661e-05,
+      "loss": 0.9432,
+      "step": 92600
+    },
+    {
+      "epoch": 51.004482782754756,
+      "grad_norm": 0.1896403431892395,
+      "learning_rate": 4.1144861266641784e-05,
+      "loss": 0.611,
+      "step": 92700
+    },
+    {
+      "epoch": 51.004762737961926,
+      "grad_norm": 0.058731839060783386,
+      "learning_rate": 4.112930819957696e-05,
+      "loss": 0.7598,
+      "step": 92800
+    },
+    {
+      "epoch": 51.005000699888015,
+      "eval_accuracy": 0.2727272727272727,
+      "eval_f1": 0.24191750595869443,
+      "eval_loss": 4.436596393585205,
+      "eval_precision": 0.2553644020829925,
+      "eval_recall": 0.2727272727272727,
+      "eval_runtime": 222.7636,
+      "eval_samples_per_second": 17.579,
+      "eval_steps_per_second": 8.79,
+      "eval_top_10_accuracy": 0.6598569969356486,
+      "eval_top_1_accuracy": 0.27247191011235955,
+      "eval_top_5_accuracy": 0.56511746680286,
+      "step": 92885
+    },
+    {
+      "epoch": 51.005000699888015,
+      "step": 92885,
+      "total_flos": 1.9366677482157638e+21,
+      "train_loss": 6.364319215203574,
+      "train_runtime": 102825.8245,
+      "train_samples_per_second": 27.791,
+      "train_steps_per_second": 3.474
+    }
+  ],
+  "logging_steps": 100,
+  "max_steps": 357200,
+  "num_input_tokens_seen": 0,
+  "num_train_epochs": 9223372036854775807,
+  "save_steps": 500,
+  "stateful_callbacks": {
+    "EarlyStoppingCallback": {
+      "args": {
+        "early_stopping_patience": 20,
+        "early_stopping_threshold": 0.0
+      },
+      "attributes": {
+        "early_stopping_patience_counter": 20
+      }
+    },
+    "TrainerControl": {
+      "args": {
+        "should_epoch_stop": false,
+        "should_evaluate": false,
+        "should_log": false,
+        "should_save": true,
+        "should_training_stop": true
+      },
+      "attributes": {}
+    }
+  },
+  "total_flos": 1.9366677482157638e+21,
+  "train_batch_size": 2,
+  "trial_name": null,
+  "trial_params": null
+}