{
  "best_metric": 0.8704663212435233,
  "best_model_checkpoint": "./vit-base-skin/checkpoint-3000",
  "epoch": 6.0,
  "global_step": 3756,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 0.0001994675186368477,
      "loss": 1.2833,
      "step": 10
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019893503727369543,
      "loss": 1.0875,
      "step": 20
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019840255591054313,
      "loss": 0.8154,
      "step": 30
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.00019787007454739086,
      "loss": 1.037,
      "step": 40
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019733759318423856,
      "loss": 0.8253,
      "step": 50
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.00019680511182108628,
      "loss": 0.838,
      "step": 60
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.00019627263045793398,
      "loss": 0.9288,
      "step": 70
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.00019574014909478168,
      "loss": 0.8994,
      "step": 80
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.0001952076677316294,
      "loss": 0.9625,
      "step": 90
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.0001946751863684771,
      "loss": 0.8057,
      "step": 100
    },
    {
      "epoch": 0.16,
      "eval_accuracy": 0.7253886010362695,
      "eval_f1": 0.6902091795647346,
      "eval_loss": 0.7590692639350891,
      "eval_precision": 0.6778860287362657,
      "eval_recall": 0.7253886010362695,
      "eval_runtime": 1.2933,
      "eval_samples_per_second": 149.234,
      "eval_steps_per_second": 19.331,
      "step": 100
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019414270500532483,
      "loss": 0.8193,
      "step": 110
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019361022364217253,
      "loss": 0.7711,
      "step": 120
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00019307774227902025,
      "loss": 0.7055,
      "step": 130
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00019254526091586795,
      "loss": 0.7175,
      "step": 140
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00019201277955271565,
      "loss": 0.8349,
      "step": 150
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.00019148029818956338,
      "loss": 0.8284,
      "step": 160
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00019094781682641108,
      "loss": 0.7171,
      "step": 170
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.0001904153354632588,
      "loss": 0.8533,
      "step": 180
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.0001898828541001065,
      "loss": 0.7037,
      "step": 190
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018935037273695422,
      "loss": 0.7619,
      "step": 200
    },
    {
      "epoch": 0.32,
      "eval_accuracy": 0.7409326424870466,
      "eval_f1": 0.6952852008863443,
      "eval_loss": 0.7081080675125122,
      "eval_precision": 0.6919837080261991,
      "eval_recall": 0.7409326424870466,
      "eval_runtime": 1.3028,
      "eval_samples_per_second": 148.147,
      "eval_steps_per_second": 19.19,
      "step": 200
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018881789137380192,
      "loss": 0.7863,
      "step": 210
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018828541001064962,
      "loss": 0.6954,
      "step": 220
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.00018775292864749735,
      "loss": 0.7077,
      "step": 230
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018722044728434505,
      "loss": 0.685,
      "step": 240
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00018668796592119277,
      "loss": 0.5368,
      "step": 250
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00018615548455804047,
      "loss": 0.7447,
      "step": 260
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.0001856230031948882,
      "loss": 0.8099,
      "step": 270
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.0001850905218317359,
      "loss": 0.7634,
      "step": 280
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0001845580404685836,
      "loss": 0.7319,
      "step": 290
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.00018402555910543132,
      "loss": 0.6315,
      "step": 300
    },
    {
      "epoch": 0.48,
      "eval_accuracy": 0.8134715025906736,
      "eval_f1": 0.8038528184141907,
      "eval_loss": 0.5953513979911804,
      "eval_precision": 0.8688087603854711,
      "eval_recall": 0.8134715025906736,
      "eval_runtime": 1.3543,
      "eval_samples_per_second": 142.507,
      "eval_steps_per_second": 18.459,
      "step": 300
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.00018349307774227902,
      "loss": 0.5614,
      "step": 310
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00018296059637912674,
      "loss": 0.5914,
      "step": 320
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00018242811501597444,
      "loss": 0.6407,
      "step": 330
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00018189563365282217,
      "loss": 0.7443,
      "step": 340
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00018136315228966987,
      "loss": 0.6917,
      "step": 350
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00018083067092651756,
      "loss": 0.6157,
      "step": 360
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0001802981895633653,
      "loss": 0.6469,
      "step": 370
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00017976570820021302,
      "loss": 0.6911,
      "step": 380
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00017923322683706071,
      "loss": 0.6223,
      "step": 390
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.0001787007454739084,
      "loss": 0.8311,
      "step": 400
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.7927461139896373,
      "eval_f1": 0.7805736577757303,
      "eval_loss": 0.597364604473114,
      "eval_precision": 0.7984743458839895,
      "eval_recall": 0.7927461139896373,
      "eval_runtime": 1.4421,
      "eval_samples_per_second": 133.832,
      "eval_steps_per_second": 17.336,
      "step": 400
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.00017816826411075614,
      "loss": 0.6898,
      "step": 410
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00017763578274760384,
      "loss": 0.6062,
      "step": 420
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00017710330138445154,
      "loss": 0.7281,
      "step": 430
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00017657082002129926,
      "loss": 0.6312,
      "step": 440
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.000176038338658147,
      "loss": 0.6278,
      "step": 450
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00017550585729499469,
      "loss": 0.5219,
      "step": 460
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.0001749733759318424,
      "loss": 0.6421,
      "step": 470
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.0001744408945686901,
      "loss": 0.5929,
      "step": 480
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.0001739084132055378,
      "loss": 0.7804,
      "step": 490
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.0001733759318423855,
      "loss": 0.5666,
      "step": 500
    },
    {
      "epoch": 0.8,
      "eval_accuracy": 0.772020725388601,
      "eval_f1": 0.7727207125131491,
      "eval_loss": 0.6151149868965149,
      "eval_precision": 0.790252614397692,
      "eval_recall": 0.772020725388601,
      "eval_runtime": 1.4778,
      "eval_samples_per_second": 130.599,
      "eval_steps_per_second": 16.917,
      "step": 500
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00017284345047923323,
      "loss": 0.6114,
      "step": 510
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00017231096911608096,
      "loss": 0.5761,
      "step": 520
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00017177848775292866,
      "loss": 0.5115,
      "step": 530
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00017124600638977638,
      "loss": 0.6255,
      "step": 540
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00017071352502662408,
      "loss": 0.6064,
      "step": 550
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00017018104366347178,
      "loss": 0.6261,
      "step": 560
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.00016964856230031948,
      "loss": 0.56,
      "step": 570
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.0001691160809371672,
      "loss": 0.4957,
      "step": 580
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00016858359957401493,
      "loss": 0.5808,
      "step": 590
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00016805111821086263,
      "loss": 0.5816,
      "step": 600
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.8031088082901554,
      "eval_f1": 0.7440025135199124,
      "eval_loss": 0.4912366271018982,
      "eval_precision": 0.7008499532379302,
      "eval_recall": 0.8031088082901554,
      "eval_runtime": 1.3782,
      "eval_samples_per_second": 140.041,
      "eval_steps_per_second": 18.14,
      "step": 600
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00016751863684771035,
      "loss": 0.7771,
      "step": 610
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00016698615548455805,
      "loss": 0.5708,
      "step": 620
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00016645367412140575,
      "loss": 0.5156,
      "step": 630
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00016592119275825345,
      "loss": 0.3287,
      "step": 640
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00016538871139510118,
      "loss": 0.4675,
      "step": 650
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.0001648562300319489,
      "loss": 0.5591,
      "step": 660
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.0001643237486687966,
      "loss": 0.5379,
      "step": 670
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00016379126730564433,
      "loss": 0.5062,
      "step": 680
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00016325878594249202,
      "loss": 0.3909,
      "step": 690
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00016272630457933972,
      "loss": 0.3715,
      "step": 700
    },
    {
      "epoch": 1.12,
      "eval_accuracy": 0.7979274611398963,
      "eval_f1": 0.7764465194154111,
      "eval_loss": 0.5772222280502319,
      "eval_precision": 0.8023955988846563,
      "eval_recall": 0.7979274611398963,
      "eval_runtime": 1.4782,
      "eval_samples_per_second": 130.565,
      "eval_steps_per_second": 16.913,
      "step": 700
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.00016219382321618745,
      "loss": 0.597,
      "step": 710
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00016166134185303515,
      "loss": 0.4993,
      "step": 720
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.00016112886048988287,
      "loss": 0.4315,
      "step": 730
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00016059637912673057,
      "loss": 0.456,
      "step": 740
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.0001600638977635783,
      "loss": 0.4603,
      "step": 750
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.000159531416400426,
      "loss": 0.5013,
      "step": 760
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.0001589989350372737,
      "loss": 0.3764,
      "step": 770
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00015846645367412142,
      "loss": 0.4753,
      "step": 780
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00015793397231096912,
      "loss": 0.4748,
      "step": 790
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.00015740149094781684,
      "loss": 0.5411,
      "step": 800
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.8341968911917098,
      "eval_f1": 0.8300510265180505,
      "eval_loss": 0.5023627281188965,
      "eval_precision": 0.8447228277489082,
      "eval_recall": 0.8341968911917098,
      "eval_runtime": 1.423,
      "eval_samples_per_second": 135.63,
      "eval_steps_per_second": 17.569,
      "step": 800
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00015686900958466454,
      "loss": 0.468,
      "step": 810
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00015633652822151227,
      "loss": 0.5218,
      "step": 820
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00015580404685835997,
      "loss": 0.3141,
      "step": 830
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00015527156549520767,
      "loss": 0.5387,
      "step": 840
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.0001547390841320554,
      "loss": 0.3781,
      "step": 850
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.0001542066027689031,
      "loss": 0.5936,
      "step": 860
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00015367412140575082,
      "loss": 0.5951,
      "step": 870
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00015314164004259851,
      "loss": 0.3731,
      "step": 880
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00015260915867944624,
      "loss": 0.4994,
      "step": 890
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00015207667731629394,
      "loss": 0.474,
      "step": 900
    },
    {
      "epoch": 1.44,
      "eval_accuracy": 0.8341968911917098,
      "eval_f1": 0.8196455563647475,
      "eval_loss": 0.437396377325058,
      "eval_precision": 0.8259791733884998,
      "eval_recall": 0.8341968911917098,
      "eval_runtime": 1.495,
      "eval_samples_per_second": 129.099,
      "eval_steps_per_second": 16.723,
      "step": 900
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.00015154419595314164,
      "loss": 0.4212,
      "step": 910
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00015101171458998936,
      "loss": 0.4708,
      "step": 920
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00015047923322683706,
      "loss": 0.4641,
      "step": 930
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001499467518636848,
      "loss": 0.5121,
      "step": 940
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.00014941427050053249,
      "loss": 0.4798,
      "step": 950
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.0001488817891373802,
      "loss": 0.3917,
      "step": 960
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0001483493077742279,
      "loss": 0.4626,
      "step": 970
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.0001478168264110756,
      "loss": 0.4031,
      "step": 980
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00014728434504792333,
      "loss": 0.4627,
      "step": 990
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00014675186368477103,
      "loss": 0.4386,
      "step": 1000
    },
    {
      "epoch": 1.6,
      "eval_accuracy": 0.7564766839378239,
      "eval_f1": 0.7807925220524523,
      "eval_loss": 0.6611098051071167,
      "eval_precision": 0.8456463174136767,
      "eval_recall": 0.7564766839378239,
      "eval_runtime": 1.4882,
      "eval_samples_per_second": 129.689,
      "eval_steps_per_second": 16.799,
      "step": 1000
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.00014621938232161876,
      "loss": 0.4531,
      "step": 1010
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00014568690095846646,
      "loss": 0.4607,
      "step": 1020
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00014515441959531418,
      "loss": 0.4729,
      "step": 1030
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00014462193823216188,
      "loss": 0.5212,
      "step": 1040
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00014408945686900958,
      "loss": 0.4455,
      "step": 1050
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.0001435569755058573,
      "loss": 0.4562,
      "step": 1060
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.000143024494142705,
      "loss": 0.3637,
      "step": 1070
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00014249201277955273,
      "loss": 0.4477,
      "step": 1080
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00014195953141640043,
      "loss": 0.5181,
      "step": 1090
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.00014142705005324815,
      "loss": 0.4091,
      "step": 1100
    },
    {
      "epoch": 1.76,
      "eval_accuracy": 0.8031088082901554,
      "eval_f1": 0.7855269798617716,
      "eval_loss": 0.5261281728744507,
      "eval_precision": 0.8287814365795575,
      "eval_recall": 0.8031088082901554,
      "eval_runtime": 1.5518,
      "eval_samples_per_second": 124.372,
      "eval_steps_per_second": 16.11,
      "step": 1100
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00014089456869009585,
      "loss": 0.5717,
      "step": 1110
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00014036208732694355,
      "loss": 0.3932,
      "step": 1120
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00013982960596379128,
      "loss": 0.2944,
      "step": 1130
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.000139297124600639,
      "loss": 0.3272,
      "step": 1140
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.0001387646432374867,
      "loss": 0.5272,
      "step": 1150
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0001382321618743344,
      "loss": 0.4034,
      "step": 1160
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.00013769968051118212,
      "loss": 0.3946,
      "step": 1170
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00013716719914802982,
      "loss": 0.6408,
      "step": 1180
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00013663471778487752,
      "loss": 0.2895,
      "step": 1190
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00013610223642172525,
      "loss": 0.4023,
      "step": 1200
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.844559585492228,
      "eval_f1": 0.8462326114605659,
      "eval_loss": 0.42789995670318604,
      "eval_precision": 0.8686702660338749,
      "eval_recall": 0.844559585492228,
      "eval_runtime": 1.3766,
      "eval_samples_per_second": 140.2,
      "eval_steps_per_second": 18.161,
      "step": 1200
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00013556975505857297,
      "loss": 0.4638,
      "step": 1210
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00013503727369542067,
      "loss": 0.2626,
      "step": 1220
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00013450479233226837,
      "loss": 0.3227,
      "step": 1230
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0001339723109691161,
      "loss": 0.4861,
      "step": 1240
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.0001334398296059638,
      "loss": 0.4213,
      "step": 1250
    },
    {
      "epoch": 2.01,
      "learning_rate": 0.0001329073482428115,
      "loss": 0.3679,
      "step": 1260
    },
    {
      "epoch": 2.03,
      "learning_rate": 0.00013242811501597445,
      "loss": 0.2227,
      "step": 1270
    },
    {
      "epoch": 2.04,
      "learning_rate": 0.00013189563365282217,
      "loss": 0.2959,
      "step": 1280
    },
    {
      "epoch": 2.06,
      "learning_rate": 0.00013136315228966987,
      "loss": 0.2047,
      "step": 1290
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.00013083067092651757,
      "loss": 0.28,
      "step": 1300
    },
    {
      "epoch": 2.08,
      "eval_accuracy": 0.8238341968911918,
      "eval_f1": 0.8023299870466322,
      "eval_loss": 0.5927017331123352,
      "eval_precision": 0.8468071334174415,
      "eval_recall": 0.8238341968911918,
      "eval_runtime": 1.4255,
      "eval_samples_per_second": 135.394,
      "eval_steps_per_second": 17.538,
      "step": 1300
    },
    {
      "epoch": 2.09,
      "learning_rate": 0.00013029818956336527,
      "loss": 0.2691,
      "step": 1310
    },
    {
      "epoch": 2.11,
      "learning_rate": 0.00012976570820021302,
      "loss": 0.2565,
      "step": 1320
    },
    {
      "epoch": 2.12,
      "learning_rate": 0.00012923322683706072,
      "loss": 0.2351,
      "step": 1330
    },
    {
      "epoch": 2.14,
      "learning_rate": 0.00012870074547390842,
      "loss": 0.3709,
      "step": 1340
    },
    {
      "epoch": 2.16,
      "learning_rate": 0.00012816826411075614,
      "loss": 0.2818,
      "step": 1350
    },
    {
      "epoch": 2.17,
      "learning_rate": 0.00012763578274760384,
      "loss": 0.2715,
      "step": 1360
    },
    {
      "epoch": 2.19,
      "learning_rate": 0.00012710330138445154,
      "loss": 0.2061,
      "step": 1370
    },
    {
      "epoch": 2.2,
      "learning_rate": 0.00012657082002129924,
      "loss": 0.2897,
      "step": 1380
    },
    {
      "epoch": 2.22,
      "learning_rate": 0.000126038338658147,
      "loss": 0.3189,
      "step": 1390
    },
    {
      "epoch": 2.24,
      "learning_rate": 0.0001255058572949947,
      "loss": 0.2408,
      "step": 1400
    },
    {
      "epoch": 2.24,
      "eval_accuracy": 0.844559585492228,
      "eval_f1": 0.8399499972143294,
      "eval_loss": 0.46048107743263245,
      "eval_precision": 0.8502584577872601,
      "eval_recall": 0.844559585492228,
      "eval_runtime": 1.3777,
      "eval_samples_per_second": 140.091,
      "eval_steps_per_second": 18.147,
      "step": 1400
    },
    {
      "epoch": 2.25,
      "learning_rate": 0.0001249733759318424,
      "loss": 0.2581,
      "step": 1410
    },
    {
      "epoch": 2.27,
      "learning_rate": 0.00012444089456869011,
      "loss": 0.3245,
      "step": 1420
    },
    {
      "epoch": 2.28,
      "learning_rate": 0.0001239084132055378,
      "loss": 0.2711,
      "step": 1430
    },
    {
      "epoch": 2.3,
      "learning_rate": 0.0001233759318423855,
      "loss": 0.1715,
      "step": 1440
    },
    {
      "epoch": 2.32,
      "learning_rate": 0.00012284345047923324,
      "loss": 0.2997,
      "step": 1450
    },
    {
      "epoch": 2.33,
      "learning_rate": 0.00012231096911608096,
      "loss": 0.3211,
      "step": 1460
    },
    {
      "epoch": 2.35,
      "learning_rate": 0.00012177848775292866,
      "loss": 0.2979,
      "step": 1470
    },
    {
      "epoch": 2.36,
      "learning_rate": 0.00012124600638977637,
      "loss": 0.3101,
      "step": 1480
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.00012071352502662407,
      "loss": 0.2338,
      "step": 1490
    },
    {
      "epoch": 2.4,
      "learning_rate": 0.00012018104366347178,
      "loss": 0.2145,
      "step": 1500
    },
    {
      "epoch": 2.4,
      "eval_accuracy": 0.8341968911917098,
      "eval_f1": 0.8399037708608313,
      "eval_loss": 0.48647597432136536,
      "eval_precision": 0.857462518686231,
      "eval_recall": 0.8341968911917098,
      "eval_runtime": 1.4386,
      "eval_samples_per_second": 134.156,
      "eval_steps_per_second": 17.378,
      "step": 1500
    },
    {
      "epoch": 2.41,
      "learning_rate": 0.00011964856230031948,
      "loss": 0.1552,
      "step": 1510
    },
    {
      "epoch": 2.43,
      "learning_rate": 0.0001191160809371672,
      "loss": 0.2341,
      "step": 1520
    },
    {
      "epoch": 2.44,
      "learning_rate": 0.00011858359957401492,
      "loss": 0.2611,
      "step": 1530
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.00011805111821086263,
      "loss": 0.3338,
      "step": 1540
    },
    {
      "epoch": 2.48,
      "learning_rate": 0.00011751863684771035,
      "loss": 0.2027,
      "step": 1550
    },
    {
      "epoch": 2.49,
      "learning_rate": 0.00011698615548455804,
      "loss": 0.246,
      "step": 1560
    },
    {
      "epoch": 2.51,
      "learning_rate": 0.00011645367412140576,
      "loss": 0.2501,
      "step": 1570
    },
    {
      "epoch": 2.52,
      "learning_rate": 0.00011592119275825345,
      "loss": 0.2001,
      "step": 1580
    },
    {
      "epoch": 2.54,
      "learning_rate": 0.00011538871139510117,
      "loss": 0.2775,
      "step": 1590
    },
    {
      "epoch": 2.56,
      "learning_rate": 0.00011485623003194889,
      "loss": 0.3194,
      "step": 1600
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.8497409326424871,
      "eval_f1": 0.8435052257391918,
      "eval_loss": 0.47271010279655457,
      "eval_precision": 0.8476361340579566,
      "eval_recall": 0.8497409326424871,
      "eval_runtime": 1.4929,
      "eval_samples_per_second": 129.279,
      "eval_steps_per_second": 16.746,
      "step": 1600
    },
    {
      "epoch": 2.57,
      "learning_rate": 0.0001143237486687966,
      "loss": 0.1092,
      "step": 1610
    },
    {
      "epoch": 2.59,
      "learning_rate": 0.00011379126730564432,
      "loss": 0.2677,
      "step": 1620
    },
    {
      "epoch": 2.6,
      "learning_rate": 0.00011325878594249201,
      "loss": 0.3687,
      "step": 1630
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.00011272630457933973,
      "loss": 0.2458,
      "step": 1640
    },
    {
      "epoch": 2.64,
      "learning_rate": 0.00011219382321618744,
      "loss": 0.3641,
      "step": 1650
    },
    {
      "epoch": 2.65,
      "learning_rate": 0.00011166134185303514,
      "loss": 0.1551,
      "step": 1660
    },
    {
      "epoch": 2.67,
      "learning_rate": 0.00011112886048988286,
      "loss": 0.4259,
      "step": 1670
    },
    {
      "epoch": 2.68,
      "learning_rate": 0.00011059637912673058,
      "loss": 0.2504,
      "step": 1680
    },
    {
      "epoch": 2.7,
      "learning_rate": 0.00011006389776357829,
      "loss": 0.4175,
      "step": 1690
    },
    {
      "epoch": 2.72,
      "learning_rate": 0.00010953141640042599,
      "loss": 0.2391,
      "step": 1700
    },
    {
      "epoch": 2.72,
      "eval_accuracy": 0.844559585492228,
      "eval_f1": 0.8402048896568737,
      "eval_loss": 0.4676208198070526,
      "eval_precision": 0.8422952210750975,
      "eval_recall": 0.844559585492228,
      "eval_runtime": 1.4255,
      "eval_samples_per_second": 135.389,
      "eval_steps_per_second": 17.537,
      "step": 1700
    },
    {
      "epoch": 2.73,
      "learning_rate": 0.0001089989350372737,
      "loss": 0.3173,
      "step": 1710
    },
    {
      "epoch": 2.75,
      "learning_rate": 0.00010846645367412141,
      "loss": 0.1635,
      "step": 1720
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.00010793397231096911,
      "loss": 0.1839,
      "step": 1730
    },
    {
      "epoch": 2.78,
      "learning_rate": 0.00010740149094781683,
      "loss": 0.2295,
      "step": 1740
    },
    {
      "epoch": 2.8,
      "learning_rate": 0.00010686900958466455,
      "loss": 0.2353,
      "step": 1750
    },
    {
      "epoch": 2.81,
      "learning_rate": 0.00010633652822151226,
      "loss": 0.2398,
      "step": 1760
    },
    {
      "epoch": 2.83,
      "learning_rate": 0.00010580404685835996,
      "loss": 0.2059,
      "step": 1770
    },
    {
      "epoch": 2.84,
      "learning_rate": 0.00010527156549520767,
      "loss": 0.2171,
      "step": 1780
    },
    {
      "epoch": 2.86,
      "learning_rate": 0.00010473908413205538,
      "loss": 0.245,
      "step": 1790
    },
    {
      "epoch": 2.88,
      "learning_rate": 0.00010420660276890308,
      "loss": 0.1828,
      "step": 1800
    },
    {
      "epoch": 2.88,
      "eval_accuracy": 0.8601036269430051,
      "eval_f1": 0.8624874419033628,
      "eval_loss": 0.433723121881485,
      "eval_precision": 0.8708761065207271,
      "eval_recall": 0.8601036269430051,
      "eval_runtime": 1.4311,
      "eval_samples_per_second": 134.86,
      "eval_steps_per_second": 17.469,
      "step": 1800
    },
    {
      "epoch": 2.89,
      "learning_rate": 0.0001036741214057508,
      "loss": 0.2974,
      "step": 1810
    },
    {
      "epoch": 2.91,
      "learning_rate": 0.00010314164004259852,
      "loss": 0.3614,
      "step": 1820
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.00010260915867944623,
      "loss": 0.1654,
      "step": 1830
    },
    {
      "epoch": 2.94,
      "learning_rate": 0.00010207667731629393,
      "loss": 0.2466,
      "step": 1840
    },
    {
      "epoch": 2.96,
      "learning_rate": 0.00010154419595314164,
      "loss": 0.1736,
      "step": 1850
    },
    {
      "epoch": 2.97,
      "learning_rate": 0.00010101171458998935,
      "loss": 0.2877,
      "step": 1860
    },
    {
      "epoch": 2.99,
      "learning_rate": 0.00010047923322683705,
      "loss": 0.3447,
      "step": 1870
    },
    {
      "epoch": 3.0,
      "learning_rate": 9.994675186368478e-05,
      "loss": 0.2438,
      "step": 1880
    },
    {
      "epoch": 3.02,
      "learning_rate": 9.941427050053249e-05,
      "loss": 0.1226,
      "step": 1890
    },
    {
      "epoch": 3.04,
      "learning_rate": 9.88817891373802e-05,
      "loss": 0.1232,
      "step": 1900
    },
    {
      "epoch": 3.04,
      "eval_accuracy": 0.8601036269430051,
      "eval_f1": 0.864604353489067,
      "eval_loss": 0.45493900775909424,
      "eval_precision": 0.8726125083980296,
      "eval_recall": 0.8601036269430051,
      "eval_runtime": 1.495,
      "eval_samples_per_second": 129.094,
      "eval_steps_per_second": 16.722,
      "step": 1900
    },
    {
      "epoch": 3.05,
      "learning_rate": 9.83493077742279e-05,
      "loss": 0.1233,
      "step": 1910
    },
    {
      "epoch": 3.07,
      "learning_rate": 9.781682641107561e-05,
      "loss": 0.1722,
      "step": 1920
    },
    {
      "epoch": 3.08,
      "learning_rate": 9.728434504792332e-05,
      "loss": 0.114,
      "step": 1930
    },
    {
      "epoch": 3.1,
      "learning_rate": 9.675186368477104e-05,
      "loss": 0.073,
      "step": 1940
    },
    {
      "epoch": 3.12,
      "learning_rate": 9.621938232161875e-05,
      "loss": 0.115,
      "step": 1950
    },
    {
      "epoch": 3.13,
      "learning_rate": 9.568690095846646e-05,
      "loss": 0.086,
      "step": 1960
    },
    {
      "epoch": 3.15,
      "learning_rate": 9.515441959531417e-05,
      "loss": 0.0939,
      "step": 1970
    },
    {
      "epoch": 3.16,
      "learning_rate": 9.462193823216187e-05,
      "loss": 0.082,
      "step": 1980
    },
    {
      "epoch": 3.18,
      "learning_rate": 9.408945686900958e-05,
      "loss": 0.1224,
      "step": 1990
    },
    {
      "epoch": 3.19,
      "learning_rate": 9.355697550585731e-05,
      "loss": 0.0929,
      "step": 2000
    },
    {
      "epoch": 3.19,
      "eval_accuracy": 0.8497409326424871,
      "eval_f1": 0.8520891403446254,
      "eval_loss": 0.5939182043075562,
      "eval_precision": 0.86060710532408,
      "eval_recall": 0.8497409326424871,
      "eval_runtime": 1.4945,
      "eval_samples_per_second": 129.138,
      "eval_steps_per_second": 16.728,
      "step": 2000
    },
    {
      "epoch": 3.21,
      "learning_rate": 9.302449414270501e-05,
      "loss": 0.0937,
      "step": 2010
    },
    {
      "epoch": 3.23,
      "learning_rate": 9.249201277955272e-05,
      "loss": 0.0459,
      "step": 2020
    },
    {
      "epoch": 3.24,
      "learning_rate": 9.195953141640043e-05,
      "loss": 0.2179,
      "step": 2030
    },
    {
      "epoch": 3.26,
      "learning_rate": 9.142705005324814e-05,
      "loss": 0.1042,
      "step": 2040
    },
    {
      "epoch": 3.27,
      "learning_rate": 9.089456869009584e-05,
      "loss": 0.07,
      "step": 2050
    },
    {
      "epoch": 3.29,
      "learning_rate": 9.036208732694355e-05,
      "loss": 0.1633,
      "step": 2060
    },
    {
      "epoch": 3.31,
      "learning_rate": 8.982960596379128e-05,
      "loss": 0.1169,
      "step": 2070
    },
    {
      "epoch": 3.32,
      "learning_rate": 8.929712460063898e-05,
      "loss": 0.0455,
      "step": 2080
    },
    {
      "epoch": 3.34,
      "learning_rate": 8.876464323748669e-05,
      "loss": 0.0547,
      "step": 2090
    },
    {
      "epoch": 3.35,
      "learning_rate": 8.82321618743344e-05,
      "loss": 0.0559,
      "step": 2100
    },
    {
      "epoch": 3.35,
      "eval_accuracy": 0.8290155440414507,
      "eval_f1": 0.8236693358455016,
      "eval_loss": 0.580689013004303,
      "eval_precision": 0.8242500909768313,
      "eval_recall": 0.8290155440414507,
      "eval_runtime": 1.4214,
      "eval_samples_per_second": 135.781,
      "eval_steps_per_second": 17.588,
      "step": 2100
    },
    {
      "epoch": 3.37,
      "learning_rate": 8.769968051118212e-05,
      "loss": 0.175,
      "step": 2110
    },
    {
      "epoch": 3.39,
      "learning_rate": 8.716719914802983e-05,
      "loss": 0.136,
      "step": 2120
    },
    {
      "epoch": 3.4,
      "learning_rate": 8.663471778487753e-05,
      "loss": 0.0599,
      "step": 2130
    },
    {
      "epoch": 3.42,
      "learning_rate": 8.610223642172525e-05,
      "loss": 0.0872,
      "step": 2140
    },
    {
      "epoch": 3.43,
      "learning_rate": 8.556975505857295e-05,
      "loss": 0.1796,
      "step": 2150
    },
    {
      "epoch": 3.45,
      "learning_rate": 8.503727369542066e-05,
      "loss": 0.0727,
      "step": 2160
    },
    {
      "epoch": 3.47,
      "learning_rate": 8.450479233226837e-05,
      "loss": 0.0664,
      "step": 2170
    },
    {
      "epoch": 3.48,
      "learning_rate": 8.397231096911609e-05,
      "loss": 0.0176,
      "step": 2180
    },
    {
      "epoch": 3.5,
      "learning_rate": 8.34398296059638e-05,
      "loss": 0.0838,
      "step": 2190
    },
    {
      "epoch": 3.51,
      "learning_rate": 8.29073482428115e-05,
      "loss": 0.1833,
      "step": 2200
    },
    {
      "epoch": 3.51,
      "eval_accuracy": 0.8601036269430051,
      "eval_f1": 0.8609735593630476,
      "eval_loss": 0.5234777331352234,
      "eval_precision": 0.8636060098945884,
      "eval_recall": 0.8601036269430051,
      "eval_runtime": 1.429,
      "eval_samples_per_second": 135.056,
      "eval_steps_per_second": 17.494,
      "step": 2200
    },
    {
      "epoch": 3.53,
      "learning_rate": 8.237486687965922e-05,
      "loss": 0.0984,
      "step": 2210
    },
    {
      "epoch": 3.55,
      "learning_rate": 8.184238551650692e-05,
      "loss": 0.1189,
      "step": 2220
    },
    {
      "epoch": 3.56,
      "learning_rate": 8.130990415335463e-05,
      "loss": 0.0526,
      "step": 2230
    },
    {
      "epoch": 3.58,
      "learning_rate": 8.077742279020235e-05,
      "loss": 0.1402,
      "step": 2240
    },
    {
      "epoch": 3.59,
      "learning_rate": 8.024494142705006e-05,
      "loss": 0.1102,
      "step": 2250
    },
    {
      "epoch": 3.61,
      "learning_rate": 7.971246006389777e-05,
      "loss": 0.1409,
      "step": 2260
    },
    {
      "epoch": 3.63,
      "learning_rate": 7.917997870074548e-05,
      "loss": 0.0704,
      "step": 2270
    },
    {
      "epoch": 3.64,
      "learning_rate": 7.86474973375932e-05,
      "loss": 0.1156,
      "step": 2280
    },
    {
      "epoch": 3.66,
      "learning_rate": 7.811501597444089e-05,
      "loss": 0.0989,
      "step": 2290
    },
    {
      "epoch": 3.67,
      "learning_rate": 7.75825346112886e-05,
      "loss": 0.1395,
      "step": 2300
    },
    {
      "epoch": 3.67,
      "eval_accuracy": 0.8134715025906736,
      "eval_f1": 0.820806815758547,
      "eval_loss": 0.674992561340332,
      "eval_precision": 0.8465714575985837,
      "eval_recall": 0.8134715025906736,
      "eval_runtime": 1.6253,
      "eval_samples_per_second": 118.748,
      "eval_steps_per_second": 15.382,
      "step": 2300
    },
    {
      "epoch": 3.69,
      "learning_rate": 7.705005324813632e-05,
      "loss": 0.1346,
      "step": 2310
    },
    {
      "epoch": 3.71,
      "learning_rate": 7.651757188498403e-05,
      "loss": 0.1195,
      "step": 2320
    },
    {
      "epoch": 3.72,
      "learning_rate": 7.598509052183174e-05,
      "loss": 0.0902,
      "step": 2330
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.545260915867945e-05,
      "loss": 0.0753,
      "step": 2340
    },
    {
      "epoch": 3.75,
      "learning_rate": 7.492012779552717e-05,
      "loss": 0.0883,
      "step": 2350
    },
    {
      "epoch": 3.77,
      "learning_rate": 7.438764643237486e-05,
      "loss": 0.0992,
      "step": 2360
    },
    {
      "epoch": 3.79,
      "learning_rate": 7.385516506922258e-05,
      "loss": 0.1442,
      "step": 2370
    },
    {
      "epoch": 3.8,
      "learning_rate": 7.332268370607029e-05,
      "loss": 0.1138,
      "step": 2380
    },
    {
      "epoch": 3.82,
      "learning_rate": 7.2790202342918e-05,
      "loss": 0.1553,
      "step": 2390
    },
    {
      "epoch": 3.83,
      "learning_rate": 7.225772097976571e-05,
      "loss": 0.0485,
      "step": 2400
    },
    {
      "epoch": 3.83,
      "eval_accuracy": 0.8860103626943006,
      "eval_f1": 0.885639852691223,
      "eval_loss": 0.44308164715766907,
      "eval_precision": 0.8887689502974477,
      "eval_recall": 0.8860103626943006,
      "eval_runtime": 1.5974,
      "eval_samples_per_second": 120.822,
      "eval_steps_per_second": 15.651,
      "step": 2400
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.172523961661343e-05,
      "loss": 0.1022,
      "step": 2410
    },
    {
      "epoch": 3.87,
      "learning_rate": 7.119275825346114e-05,
      "loss": 0.1077,
      "step": 2420
    },
    {
      "epoch": 3.88,
      "learning_rate": 7.066027689030884e-05,
      "loss": 0.1688,
      "step": 2430
    },
    {
      "epoch": 3.9,
      "learning_rate": 7.012779552715655e-05,
      "loss": 0.0722,
      "step": 2440
    },
    {
      "epoch": 3.91,
      "learning_rate": 6.959531416400426e-05,
      "loss": 0.0801,
      "step": 2450
    },
    {
      "epoch": 3.93,
      "learning_rate": 6.906283280085197e-05,
      "loss": 0.021,
      "step": 2460
    },
    {
      "epoch": 3.95,
      "learning_rate": 6.853035143769968e-05,
      "loss": 0.1538,
      "step": 2470
    },
    {
      "epoch": 3.96,
      "learning_rate": 6.79978700745474e-05,
      "loss": 0.0944,
      "step": 2480
    },
    {
      "epoch": 3.98,
      "learning_rate": 6.746538871139511e-05,
      "loss": 0.0663,
      "step": 2490
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.693290734824281e-05,
      "loss": 0.1206,
      "step": 2500
    },
    {
      "epoch": 3.99,
      "eval_accuracy": 0.8393782383419689,
      "eval_f1": 0.8375225030319982,
      "eval_loss": 0.5491106510162354,
      "eval_precision": 0.8477221585625083,
      "eval_recall": 0.8393782383419689,
      "eval_runtime": 1.4337,
      "eval_samples_per_second": 134.621,
      "eval_steps_per_second": 17.438,
      "step": 2500
    },
    {
      "epoch": 4.01,
      "learning_rate": 6.640042598509052e-05,
      "loss": 0.0565,
      "step": 2510
    },
    {
      "epoch": 4.03,
      "learning_rate": 6.586794462193823e-05,
      "loss": 0.0323,
      "step": 2520
    },
    {
      "epoch": 4.04,
      "learning_rate": 6.533546325878594e-05,
      "loss": 0.0655,
      "step": 2530
    },
    {
      "epoch": 4.06,
      "learning_rate": 6.480298189563366e-05,
      "loss": 0.0289,
      "step": 2540
    },
    {
      "epoch": 4.07,
      "learning_rate": 6.427050053248137e-05,
      "loss": 0.0975,
      "step": 2550
    },
    {
      "epoch": 4.09,
      "learning_rate": 6.373801916932908e-05,
      "loss": 0.0777,
      "step": 2560
    },
    {
      "epoch": 4.11,
      "learning_rate": 6.320553780617678e-05,
      "loss": 0.0391,
      "step": 2570
    },
    {
      "epoch": 4.12,
      "learning_rate": 6.26730564430245e-05,
      "loss": 0.0266,
      "step": 2580
    },
    {
      "epoch": 4.14,
      "learning_rate": 6.214057507987222e-05,
      "loss": 0.0325,
      "step": 2590
    },
    {
      "epoch": 4.15,
      "learning_rate": 6.160809371671991e-05,
      "loss": 0.0485,
      "step": 2600
    },
    {
      "epoch": 4.15,
      "eval_accuracy": 0.8652849740932642,
      "eval_f1": 0.8676890818288896,
      "eval_loss": 0.5288814306259155,
      "eval_precision": 0.8744135727666489,
      "eval_recall": 0.8652849740932642,
      "eval_runtime": 1.6432,
      "eval_samples_per_second": 117.451,
      "eval_steps_per_second": 15.214,
      "step": 2600
    },
    {
      "epoch": 4.17,
      "learning_rate": 6.107561235356763e-05,
      "loss": 0.0644,
      "step": 2610
    },
    {
      "epoch": 4.19,
      "learning_rate": 6.054313099041534e-05,
      "loss": 0.0089,
      "step": 2620
    },
    {
      "epoch": 4.2,
      "learning_rate": 6.001064962726305e-05,
      "loss": 0.0103,
      "step": 2630
    },
    {
      "epoch": 4.22,
      "learning_rate": 5.9478168264110756e-05,
      "loss": 0.0293,
      "step": 2640
    },
    {
      "epoch": 4.23,
      "learning_rate": 5.8945686900958475e-05,
      "loss": 0.0282,
      "step": 2650
    },
    {
      "epoch": 4.25,
      "learning_rate": 5.841320553780618e-05,
      "loss": 0.0214,
      "step": 2660
    },
    {
      "epoch": 4.27,
      "learning_rate": 5.7880724174653886e-05,
      "loss": 0.0148,
      "step": 2670
    },
    {
      "epoch": 4.28,
      "learning_rate": 5.73482428115016e-05,
      "loss": 0.0185,
      "step": 2680
    },
    {
      "epoch": 4.3,
      "learning_rate": 5.681576144834932e-05,
      "loss": 0.0058,
      "step": 2690
    },
    {
      "epoch": 4.31,
      "learning_rate": 5.628328008519702e-05,
      "loss": 0.0494,
      "step": 2700
    },
    {
      "epoch": 4.31,
      "eval_accuracy": 0.8601036269430051,
      "eval_f1": 0.8602912574839866,
      "eval_loss": 0.5664997696876526,
      "eval_precision": 0.8633398594607762,
      "eval_recall": 0.8601036269430051,
      "eval_runtime": 1.4897,
      "eval_samples_per_second": 129.556,
      "eval_steps_per_second": 16.782,
      "step": 2700
    },
    {
      "epoch": 4.33,
      "learning_rate": 5.575079872204473e-05,
      "loss": 0.0061,
      "step": 2710
    },
    {
      "epoch": 4.35,
      "learning_rate": 5.521831735889245e-05,
      "loss": 0.0181,
      "step": 2720
    },
    {
      "epoch": 4.36,
      "learning_rate": 5.468583599574015e-05,
      "loss": 0.0119,
      "step": 2730
    },
    {
      "epoch": 4.38,
      "learning_rate": 5.415335463258786e-05,
      "loss": 0.0052,
      "step": 2740
    },
    {
      "epoch": 4.39,
      "learning_rate": 5.362087326943557e-05,
      "loss": 0.0087,
      "step": 2750
    },
    {
      "epoch": 4.41,
      "learning_rate": 5.308839190628329e-05,
      "loss": 0.0473,
      "step": 2760
    },
    {
      "epoch": 4.42,
      "learning_rate": 5.2555910543130994e-05,
      "loss": 0.0859,
      "step": 2770
    },
    {
      "epoch": 4.44,
      "learning_rate": 5.20234291799787e-05,
      "loss": 0.0787,
      "step": 2780
    },
    {
      "epoch": 4.46,
      "learning_rate": 5.149094781682642e-05,
      "loss": 0.0431,
      "step": 2790
    },
    {
      "epoch": 4.47,
      "learning_rate": 5.095846645367412e-05,
      "loss": 0.0062,
      "step": 2800
    },
    {
      "epoch": 4.47,
      "eval_accuracy": 0.8497409326424871,
      "eval_f1": 0.8478513270593213,
      "eval_loss": 0.6186147928237915,
      "eval_precision": 0.8546696007885869,
      "eval_recall": 0.8497409326424871,
      "eval_runtime": 1.4697,
      "eval_samples_per_second": 131.324,
      "eval_steps_per_second": 17.011,
      "step": 2800
    },
    {
      "epoch": 4.49,
      "learning_rate": 5.0425985090521835e-05,
      "loss": 0.0604,
      "step": 2810
    },
    {
      "epoch": 4.5,
      "learning_rate": 4.989350372736955e-05,
      "loss": 0.0073,
      "step": 2820
    },
    {
      "epoch": 4.52,
      "learning_rate": 4.936102236421725e-05,
      "loss": 0.0554,
      "step": 2830
    },
    {
      "epoch": 4.54,
      "learning_rate": 4.8828541001064965e-05,
      "loss": 0.0458,
      "step": 2840
    },
    {
      "epoch": 4.55,
      "learning_rate": 4.829605963791268e-05,
      "loss": 0.0089,
      "step": 2850
    },
    {
      "epoch": 4.57,
      "learning_rate": 4.776357827476038e-05,
      "loss": 0.0069,
      "step": 2860
    },
    {
      "epoch": 4.58,
      "learning_rate": 4.7231096911608095e-05,
      "loss": 0.0574,
      "step": 2870
    },
    {
      "epoch": 4.6,
      "learning_rate": 4.669861554845581e-05,
      "loss": 0.0057,
      "step": 2880
    },
    {
      "epoch": 4.62,
      "learning_rate": 4.616613418530352e-05,
      "loss": 0.0819,
      "step": 2890
    },
    {
      "epoch": 4.63,
      "learning_rate": 4.563365282215123e-05,
      "loss": 0.0065,
      "step": 2900
    },
    {
      "epoch": 4.63,
      "eval_accuracy": 0.8756476683937824,
      "eval_f1": 0.8728464968594524,
      "eval_loss": 0.5823409557342529,
      "eval_precision": 0.8737292973366059,
      "eval_recall": 0.8756476683937824,
      "eval_runtime": 1.5056,
      "eval_samples_per_second": 128.19,
      "eval_steps_per_second": 16.605,
      "step": 2900
    },
    {
      "epoch": 4.65,
      "learning_rate": 4.5101171458998936e-05,
      "loss": 0.0171,
      "step": 2910
    },
    {
      "epoch": 4.66,
      "learning_rate": 4.456869009584665e-05,
      "loss": 0.0063,
      "step": 2920
    },
    {
      "epoch": 4.68,
      "learning_rate": 4.4036208732694354e-05,
      "loss": 0.0705,
      "step": 2930
    },
    {
      "epoch": 4.7,
      "learning_rate": 4.3503727369542066e-05,
      "loss": 0.0335,
      "step": 2940
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.297124600638978e-05,
      "loss": 0.0526,
      "step": 2950
    },
    {
      "epoch": 4.73,
      "learning_rate": 4.243876464323749e-05,
      "loss": 0.0058,
      "step": 2960
    },
    {
      "epoch": 4.74,
      "learning_rate": 4.19062832800852e-05,
      "loss": 0.0173,
      "step": 2970
    },
    {
      "epoch": 4.76,
      "learning_rate": 4.137380191693291e-05,
      "loss": 0.0483,
      "step": 2980
    },
    {
      "epoch": 4.78,
      "learning_rate": 4.084132055378062e-05,
      "loss": 0.0076,
      "step": 2990
    },
    {
      "epoch": 4.79,
      "learning_rate": 4.0308839190628325e-05,
      "loss": 0.0045,
      "step": 3000
    },
    {
      "epoch": 4.79,
      "eval_accuracy": 0.8704663212435233,
      "eval_f1": 0.8698843579901536,
      "eval_loss": 0.5800853967666626,
      "eval_precision": 0.8724272893958966,
      "eval_recall": 0.8704663212435233,
      "eval_runtime": 1.5036,
      "eval_samples_per_second": 128.362,
      "eval_steps_per_second": 16.627,
      "step": 3000
    },
    {
      "epoch": 4.81,
      "learning_rate": 3.9776357827476044e-05,
      "loss": 0.0324,
      "step": 3010
    },
    {
      "epoch": 4.82,
      "learning_rate": 3.9243876464323756e-05,
      "loss": 0.034,
      "step": 3020
    },
    {
      "epoch": 4.84,
      "learning_rate": 3.871139510117146e-05,
      "loss": 0.0043,
      "step": 3030
    },
    {
      "epoch": 4.86,
      "learning_rate": 3.8178913738019174e-05,
      "loss": 0.063,
      "step": 3040
    },
    {
      "epoch": 4.87,
      "learning_rate": 3.764643237486688e-05,
      "loss": 0.0054,
      "step": 3050
    },
    {
      "epoch": 4.89,
      "learning_rate": 3.711395101171459e-05,
      "loss": 0.0056,
      "step": 3060
    },
    {
      "epoch": 4.9,
      "learning_rate": 3.65814696485623e-05,
      "loss": 0.0076,
      "step": 3070
    },
    {
      "epoch": 4.92,
      "learning_rate": 3.6048988285410015e-05,
      "loss": 0.0377,
      "step": 3080
    },
    {
      "epoch": 4.94,
      "learning_rate": 3.551650692225773e-05,
      "loss": 0.0118,
      "step": 3090
    },
    {
      "epoch": 4.95,
      "learning_rate": 3.498402555910543e-05,
      "loss": 0.038,
      "step": 3100
    },
    {
      "epoch": 4.95,
      "eval_accuracy": 0.8393782383419689,
      "eval_f1": 0.8404874328190347,
      "eval_loss": 0.654182493686676,
      "eval_precision": 0.8471590163920663,
      "eval_recall": 0.8393782383419689,
      "eval_runtime": 1.4009,
      "eval_samples_per_second": 137.764,
      "eval_steps_per_second": 17.845,
      "step": 3100
    },
    {
      "epoch": 4.97,
      "learning_rate": 3.4451544195953145e-05,
      "loss": 0.0058,
      "step": 3110
    },
    {
      "epoch": 4.98,
      "learning_rate": 3.391906283280085e-05,
      "loss": 0.0324,
      "step": 3120
    },
    {
      "epoch": 5.0,
      "learning_rate": 3.338658146964856e-05,
      "loss": 0.0643,
      "step": 3130
    },
    {
      "epoch": 5.02,
      "learning_rate": 3.2854100106496274e-05,
      "loss": 0.0185,
      "step": 3140
    },
    {
      "epoch": 5.03,
      "learning_rate": 3.2321618743343987e-05,
      "loss": 0.0062,
      "step": 3150
    },
    {
      "epoch": 5.05,
      "learning_rate": 3.17891373801917e-05,
      "loss": 0.0041,
      "step": 3160
    },
    {
      "epoch": 5.06,
      "learning_rate": 3.1256656017039404e-05,
      "loss": 0.0032,
      "step": 3170
    },
    {
      "epoch": 5.08,
      "learning_rate": 3.0724174653887116e-05,
      "loss": 0.0252,
      "step": 3180
    },
    {
      "epoch": 5.1,
      "learning_rate": 3.0191693290734825e-05,
      "loss": 0.0029,
      "step": 3190
    },
    {
      "epoch": 5.11,
      "learning_rate": 2.9659211927582537e-05,
      "loss": 0.0035,
      "step": 3200
    },
    {
      "epoch": 5.11,
      "eval_accuracy": 0.8652849740932642,
      "eval_f1": 0.8653389841057969,
      "eval_loss": 0.6028582453727722,
      "eval_precision": 0.8713586643638457,
      "eval_recall": 0.8652849740932642,
      "eval_runtime": 1.4572,
      "eval_samples_per_second": 132.443,
      "eval_steps_per_second": 17.156,
      "step": 3200
    },
    {
      "epoch": 5.13,
      "learning_rate": 2.912673056443025e-05,
      "loss": 0.0033,
      "step": 3210
    },
    {
      "epoch": 5.14,
      "learning_rate": 2.8594249201277955e-05,
      "loss": 0.0029,
      "step": 3220
    },
    {
      "epoch": 5.16,
      "learning_rate": 2.806176783812567e-05,
      "loss": 0.0033,
      "step": 3230
    },
    {
      "epoch": 5.18,
      "learning_rate": 2.7529286474973375e-05,
      "loss": 0.0038,
      "step": 3240
    },
    {
      "epoch": 5.19,
      "learning_rate": 2.6996805111821088e-05,
      "loss": 0.0031,
      "step": 3250
    },
    {
      "epoch": 5.21,
      "learning_rate": 2.6464323748668796e-05,
      "loss": 0.003,
      "step": 3260
    },
    {
      "epoch": 5.22,
      "learning_rate": 2.593184238551651e-05,
      "loss": 0.0115,
      "step": 3270
    },
    {
      "epoch": 5.24,
      "learning_rate": 2.539936102236422e-05,
      "loss": 0.0219,
      "step": 3280
    },
    {
      "epoch": 5.26,
      "learning_rate": 2.486687965921193e-05,
      "loss": 0.003,
      "step": 3290
    },
    {
      "epoch": 5.27,
      "learning_rate": 2.4334398296059638e-05,
      "loss": 0.0031,
      "step": 3300
    },
    {
      "epoch": 5.27,
      "eval_accuracy": 0.8601036269430051,
      "eval_f1": 0.8581969638224265,
      "eval_loss": 0.6385093927383423,
      "eval_precision": 0.8653198213208406,
      "eval_recall": 0.8601036269430051,
      "eval_runtime": 1.5005,
      "eval_samples_per_second": 128.623,
      "eval_steps_per_second": 16.661,
      "step": 3300
    },
    {
      "epoch": 5.29,
      "learning_rate": 2.380191693290735e-05,
      "loss": 0.0031,
      "step": 3310
    },
    {
      "epoch": 5.3,
      "learning_rate": 2.326943556975506e-05,
      "loss": 0.0034,
      "step": 3320
    },
    {
      "epoch": 5.32,
      "learning_rate": 2.273695420660277e-05,
      "loss": 0.0031,
      "step": 3330
    },
    {
      "epoch": 5.34,
      "learning_rate": 2.220447284345048e-05,
      "loss": 0.0065,
      "step": 3340
    },
    {
      "epoch": 5.35,
      "learning_rate": 2.167199148029819e-05,
      "loss": 0.0031,
      "step": 3350
    },
    {
      "epoch": 5.37,
      "learning_rate": 2.11395101171459e-05,
      "loss": 0.003,
      "step": 3360
    },
    {
      "epoch": 5.38,
      "learning_rate": 2.0607028753993613e-05,
      "loss": 0.0028,
      "step": 3370
    },
    {
      "epoch": 5.4,
      "learning_rate": 2.007454739084132e-05,
      "loss": 0.0029,
      "step": 3380
    },
    {
      "epoch": 5.42,
      "learning_rate": 1.9542066027689034e-05,
      "loss": 0.0031,
      "step": 3390
    },
    {
      "epoch": 5.43,
      "learning_rate": 1.9009584664536742e-05,
      "loss": 0.0029,
      "step": 3400
    },
    {
      "epoch": 5.43,
      "eval_accuracy": 0.8704663212435233,
      "eval_f1": 0.867615361571679,
      "eval_loss": 0.6132028698921204,
      "eval_precision": 0.883001891602928,
      "eval_recall": 0.8704663212435233,
      "eval_runtime": 1.5211,
      "eval_samples_per_second": 126.882,
      "eval_steps_per_second": 16.436,
      "step": 3400
    },
    {
      "epoch": 5.45,
      "learning_rate": 1.847710330138445e-05,
      "loss": 0.0388,
      "step": 3410
    },
    {
      "epoch": 5.46,
      "learning_rate": 1.7944621938232163e-05,
      "loss": 0.003,
      "step": 3420
    },
    {
      "epoch": 5.48,
      "learning_rate": 1.7412140575079875e-05,
      "loss": 0.0231,
      "step": 3430
    },
    {
      "epoch": 5.5,
      "learning_rate": 1.6879659211927584e-05,
      "loss": 0.0032,
      "step": 3440
    },
    {
      "epoch": 5.51,
      "learning_rate": 1.6347177848775293e-05,
      "loss": 0.0044,
      "step": 3450
    },
    {
      "epoch": 5.53,
      "learning_rate": 1.5814696485623005e-05,
      "loss": 0.0031,
      "step": 3460
    },
    {
      "epoch": 5.54,
      "learning_rate": 1.5282215122470714e-05,
      "loss": 0.0028,
      "step": 3470
    },
    {
      "epoch": 5.56,
      "learning_rate": 1.4749733759318424e-05,
      "loss": 0.0033,
      "step": 3480
    },
    {
      "epoch": 5.58,
      "learning_rate": 1.4217252396166134e-05,
      "loss": 0.003,
      "step": 3490
    },
    {
      "epoch": 5.59,
      "learning_rate": 1.3684771033013847e-05,
      "loss": 0.0039,
      "step": 3500
    },
    {
      "epoch": 5.59,
      "eval_accuracy": 0.8652849740932642,
      "eval_f1": 0.8638885525134503,
      "eval_loss": 0.6398009657859802,
      "eval_precision": 0.8814954336431856,
      "eval_recall": 0.8652849740932642,
      "eval_runtime": 1.4856,
      "eval_samples_per_second": 129.91,
      "eval_steps_per_second": 16.828,
      "step": 3500
    },
    {
      "epoch": 5.61,
      "learning_rate": 1.3152289669861555e-05,
      "loss": 0.0029,
      "step": 3510
    },
    {
      "epoch": 5.62,
      "learning_rate": 1.2619808306709266e-05,
      "loss": 0.0031,
      "step": 3520
    },
    {
      "epoch": 5.64,
      "learning_rate": 1.2087326943556976e-05,
      "loss": 0.0024,
      "step": 3530
    },
    {
      "epoch": 5.65,
      "learning_rate": 1.1554845580404687e-05,
      "loss": 0.0025,
      "step": 3540
    },
    {
      "epoch": 5.67,
      "learning_rate": 1.1022364217252397e-05,
      "loss": 0.003,
      "step": 3550
    },
    {
      "epoch": 5.69,
      "learning_rate": 1.0489882854100107e-05,
      "loss": 0.0026,
      "step": 3560
    },
    {
      "epoch": 5.7,
      "learning_rate": 9.957401490947816e-06,
      "loss": 0.0024,
      "step": 3570
    },
    {
      "epoch": 5.72,
      "learning_rate": 9.424920127795528e-06,
      "loss": 0.0034,
      "step": 3580
    },
    {
      "epoch": 5.73,
      "learning_rate": 8.892438764643237e-06,
      "loss": 0.0055,
      "step": 3590
    },
    {
      "epoch": 5.75,
      "learning_rate": 8.359957401490947e-06,
      "loss": 0.0034,
      "step": 3600
    },
    {
      "epoch": 5.75,
      "eval_accuracy": 0.8652849740932642,
      "eval_f1": 0.8648885794188792,
      "eval_loss": 0.6220521926879883,
      "eval_precision": 0.8726252158894647,
      "eval_recall": 0.8652849740932642,
      "eval_runtime": 1.4634,
      "eval_samples_per_second": 131.885,
      "eval_steps_per_second": 17.083,
      "step": 3600
    },
    {
      "epoch": 5.77,
      "learning_rate": 7.82747603833866e-06,
      "loss": 0.0061,
      "step": 3610
    },
    {
      "epoch": 5.78,
      "learning_rate": 7.294994675186369e-06,
      "loss": 0.0023,
      "step": 3620
    },
    {
      "epoch": 5.8,
      "learning_rate": 6.762513312034079e-06,
      "loss": 0.0025,
      "step": 3630
    },
    {
      "epoch": 5.81,
      "learning_rate": 6.230031948881789e-06,
      "loss": 0.0025,
      "step": 3640
    },
    {
      "epoch": 5.83,
      "learning_rate": 5.6975505857295e-06,
      "loss": 0.0023,
      "step": 3650
    },
    {
      "epoch": 5.85,
      "learning_rate": 5.16506922257721e-06,
      "loss": 0.003,
      "step": 3660
    },
    {
      "epoch": 5.86,
      "learning_rate": 4.6325878594249205e-06,
      "loss": 0.0023,
      "step": 3670
    },
    {
      "epoch": 5.88,
      "learning_rate": 4.100106496272631e-06,
      "loss": 0.003,
      "step": 3680
    },
    {
      "epoch": 5.89,
      "learning_rate": 3.5676251331203413e-06,
      "loss": 0.0027,
      "step": 3690
    },
    {
      "epoch": 5.91,
      "learning_rate": 3.0351437699680513e-06,
      "loss": 0.003,
      "step": 3700
    },
    {
      "epoch": 5.91,
      "eval_accuracy": 0.8704663212435233,
      "eval_f1": 0.8684333767114856,
      "eval_loss": 0.6205737590789795,
      "eval_precision": 0.8849810017271157,
      "eval_recall": 0.8704663212435233,
      "eval_runtime": 1.4708,
      "eval_samples_per_second": 131.218,
      "eval_steps_per_second": 16.997,
      "step": 3700
    },
    {
      "epoch": 5.93,
      "learning_rate": 2.5026624068157617e-06,
      "loss": 0.0044,
      "step": 3710
    },
    {
      "epoch": 5.94,
      "learning_rate": 1.9701810436634718e-06,
      "loss": 0.0025,
      "step": 3720
    },
    {
      "epoch": 5.96,
      "learning_rate": 1.4376996805111822e-06,
      "loss": 0.053,
      "step": 3730
    },
    {
      "epoch": 5.97,
      "learning_rate": 9.052183173588925e-07,
      "loss": 0.0027,
      "step": 3740
    },
    {
      "epoch": 5.99,
      "learning_rate": 3.727369542066028e-07,
      "loss": 0.0023,
      "step": 3750
    },
    {
      "epoch": 6.0,
      "step": 3756,
      "total_flos": 4.65670232933972e+18,
      "train_loss": 0.2616841745980905,
      "train_runtime": 1322.2784,
      "train_samples_per_second": 45.444,
      "train_steps_per_second": 2.841
    }
  ],
  "max_steps": 3756,
  "num_train_epochs": 6,
  "total_flos": 4.65670232933972e+18,
  "trial_name": null,
  "trial_params": null
}