{
  "best_metric": 0.2815741002559662,
  "best_model_checkpoint": "finetuned-food/checkpoint-5000",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 5096,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 0.00019960753532182103,
      "loss": 0.4744,
      "step": 10
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001992150706436421,
      "loss": 0.4761,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 0.0001988226059654631,
      "loss": 0.5317,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 0.00019843014128728415,
      "loss": 0.477,
      "step": 40
    },
    {
      "epoch": 0.04,
      "learning_rate": 0.0001980376766091052,
      "loss": 0.3772,
      "step": 50
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019764521193092621,
      "loss": 0.3498,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 0.00019725274725274728,
      "loss": 0.7838,
      "step": 70
    },
    {
      "epoch": 0.06,
      "learning_rate": 0.0001968602825745683,
      "loss": 0.6395,
      "step": 80
    },
    {
      "epoch": 0.07,
      "learning_rate": 0.00019646781789638932,
      "loss": 0.5704,
      "step": 90
    },
    {
      "epoch": 0.08,
      "learning_rate": 0.00019607535321821039,
      "loss": 0.556,
      "step": 100
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.0001956828885400314,
      "loss": 0.6918,
      "step": 110
    },
    {
      "epoch": 0.09,
      "learning_rate": 0.00019529042386185245,
      "loss": 0.9462,
      "step": 120
    },
    {
      "epoch": 0.1,
      "learning_rate": 0.0001948979591836735,
      "loss": 0.9653,
      "step": 130
    },
    {
      "epoch": 0.11,
      "learning_rate": 0.0001945054945054945,
      "loss": 0.4661,
      "step": 140
    },
    {
      "epoch": 0.12,
      "learning_rate": 0.00019411302982731555,
      "loss": 0.7594,
      "step": 150
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0001937205651491366,
      "loss": 0.5468,
      "step": 160
    },
    {
      "epoch": 0.13,
      "learning_rate": 0.0001933281004709576,
      "loss": 0.5591,
      "step": 170
    },
    {
      "epoch": 0.14,
      "learning_rate": 0.00019293563579277868,
      "loss": 0.7155,
      "step": 180
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.0001925431711145997,
      "loss": 0.5256,
      "step": 190
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019215070643642074,
      "loss": 0.7468,
      "step": 200
    },
    {
      "epoch": 0.16,
      "learning_rate": 0.00019175824175824178,
      "loss": 0.915,
      "step": 210
    },
    {
      "epoch": 0.17,
      "learning_rate": 0.0001913657770800628,
      "loss": 0.9691,
      "step": 220
    },
    {
      "epoch": 0.18,
      "learning_rate": 0.00019097331240188384,
      "loss": 1.1112,
      "step": 230
    },
    {
      "epoch": 0.19,
      "learning_rate": 0.00019058084772370488,
      "loss": 1.0109,
      "step": 240
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.0001901883830455259,
      "loss": 0.923,
      "step": 250
    },
    {
      "epoch": 0.2,
      "learning_rate": 0.00018979591836734697,
      "loss": 0.8972,
      "step": 260
    },
    {
      "epoch": 0.21,
      "learning_rate": 0.00018940345368916798,
      "loss": 0.9926,
      "step": 270
    },
    {
      "epoch": 0.22,
      "learning_rate": 0.00018901098901098903,
      "loss": 1.0361,
      "step": 280
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00018861852433281007,
      "loss": 0.9765,
      "step": 290
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018822605965463109,
      "loss": 0.8471,
      "step": 300
    },
    {
      "epoch": 0.24,
      "learning_rate": 0.00018783359497645213,
      "loss": 0.9841,
      "step": 310
    },
    {
      "epoch": 0.25,
      "learning_rate": 0.00018744113029827317,
      "loss": 1.1251,
      "step": 320
    },
    {
      "epoch": 0.26,
      "learning_rate": 0.0001870486656200942,
      "loss": 0.9555,
      "step": 330
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018665620094191523,
      "loss": 1.0878,
      "step": 340
    },
    {
      "epoch": 0.27,
      "learning_rate": 0.00018626373626373627,
      "loss": 1.0991,
      "step": 350
    },
    {
      "epoch": 0.28,
      "learning_rate": 0.00018587127158555732,
      "loss": 0.8107,
      "step": 360
    },
    {
      "epoch": 0.29,
      "learning_rate": 0.00018547880690737836,
      "loss": 0.7627,
      "step": 370
    },
    {
      "epoch": 0.3,
      "learning_rate": 0.00018508634222919938,
      "loss": 0.8323,
      "step": 380
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018469387755102042,
      "loss": 0.858,
      "step": 390
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.00018430141287284146,
      "loss": 1.0723,
      "step": 400
    },
    {
      "epoch": 0.32,
      "learning_rate": 0.00018390894819466248,
      "loss": 0.9763,
      "step": 410
    },
    {
      "epoch": 0.33,
      "learning_rate": 0.00018351648351648352,
      "loss": 0.8474,
      "step": 420
    },
    {
      "epoch": 0.34,
      "learning_rate": 0.00018312401883830456,
      "loss": 1.0441,
      "step": 430
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.0001827315541601256,
      "loss": 0.8014,
      "step": 440
    },
    {
      "epoch": 0.35,
      "learning_rate": 0.00018233908948194662,
      "loss": 0.8023,
      "step": 450
    },
    {
      "epoch": 0.36,
      "learning_rate": 0.00018194662480376767,
      "loss": 0.8577,
      "step": 460
    },
    {
      "epoch": 0.37,
      "learning_rate": 0.0001815541601255887,
      "loss": 0.7513,
      "step": 470
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018116169544740975,
      "loss": 0.9424,
      "step": 480
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00018076923076923077,
      "loss": 0.7058,
      "step": 490
    },
    {
      "epoch": 0.39,
      "learning_rate": 0.0001803767660910518,
      "loss": 0.8456,
      "step": 500
    },
    {
      "epoch": 0.39,
      "eval_accuracy": 0.7633507853403141,
      "eval_loss": 0.8592807054519653,
      "eval_runtime": 104.955,
      "eval_samples_per_second": 45.496,
      "eval_steps_per_second": 5.688,
      "step": 500
    },
    {
      "epoch": 0.4,
      "learning_rate": 0.00017998430141287285,
      "loss": 0.9776,
      "step": 510
    },
    {
      "epoch": 0.41,
      "learning_rate": 0.0001795918367346939,
      "loss": 0.8952,
      "step": 520
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.0001791993720565149,
      "loss": 1.1562,
      "step": 530
    },
    {
      "epoch": 0.42,
      "learning_rate": 0.00017880690737833596,
      "loss": 0.8117,
      "step": 540
    },
    {
      "epoch": 0.43,
      "learning_rate": 0.000178414442700157,
      "loss": 0.7151,
      "step": 550
    },
    {
      "epoch": 0.44,
      "learning_rate": 0.00017802197802197802,
      "loss": 0.811,
      "step": 560
    },
    {
      "epoch": 0.45,
      "learning_rate": 0.00017762951334379906,
      "loss": 1.0324,
      "step": 570
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.0001772370486656201,
      "loss": 0.9562,
      "step": 580
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00017684458398744114,
      "loss": 0.8997,
      "step": 590
    },
    {
      "epoch": 0.47,
      "learning_rate": 0.0001764521193092622,
      "loss": 1.0177,
      "step": 600
    },
    {
      "epoch": 0.48,
      "learning_rate": 0.0001760596546310832,
      "loss": 0.7006,
      "step": 610
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.00017566718995290425,
      "loss": 0.777,
      "step": 620
    },
    {
      "epoch": 0.49,
      "learning_rate": 0.0001752747252747253,
      "loss": 0.7591,
      "step": 630
    },
    {
      "epoch": 0.5,
      "learning_rate": 0.0001748822605965463,
      "loss": 0.7587,
      "step": 640
    },
    {
      "epoch": 0.51,
      "learning_rate": 0.00017448979591836735,
      "loss": 0.7518,
      "step": 650
    },
    {
      "epoch": 0.52,
      "learning_rate": 0.0001740973312401884,
      "loss": 0.6905,
      "step": 660
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017370486656200943,
      "loss": 0.6435,
      "step": 670
    },
    {
      "epoch": 0.53,
      "learning_rate": 0.00017331240188383048,
      "loss": 0.8539,
      "step": 680
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.0001729199372056515,
      "loss": 0.9524,
      "step": 690
    },
    {
      "epoch": 0.55,
      "learning_rate": 0.00017252747252747254,
      "loss": 0.8727,
      "step": 700
    },
    {
      "epoch": 0.56,
      "learning_rate": 0.00017213500784929358,
      "loss": 0.7349,
      "step": 710
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.0001717425431711146,
      "loss": 0.6538,
      "step": 720
    },
    {
      "epoch": 0.57,
      "learning_rate": 0.00017135007849293564,
      "loss": 0.9937,
      "step": 730
    },
    {
      "epoch": 0.58,
      "learning_rate": 0.00017095761381475668,
      "loss": 0.7501,
      "step": 740
    },
    {
      "epoch": 0.59,
      "learning_rate": 0.0001705651491365777,
      "loss": 0.8535,
      "step": 750
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00017017268445839877,
      "loss": 0.9108,
      "step": 760
    },
    {
      "epoch": 0.6,
      "learning_rate": 0.00016978021978021978,
      "loss": 0.7519,
      "step": 770
    },
    {
      "epoch": 0.61,
      "learning_rate": 0.00016938775510204083,
      "loss": 0.661,
      "step": 780
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00016899529042386187,
      "loss": 0.7516,
      "step": 790
    },
    {
      "epoch": 0.63,
      "learning_rate": 0.0001686028257456829,
      "loss": 0.6468,
      "step": 800
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016821036106750393,
      "loss": 0.7219,
      "step": 810
    },
    {
      "epoch": 0.64,
      "learning_rate": 0.00016781789638932497,
      "loss": 0.9463,
      "step": 820
    },
    {
      "epoch": 0.65,
      "learning_rate": 0.000167425431711146,
      "loss": 0.9053,
      "step": 830
    },
    {
      "epoch": 0.66,
      "learning_rate": 0.00016703296703296706,
      "loss": 0.7573,
      "step": 840
    },
    {
      "epoch": 0.67,
      "learning_rate": 0.00016664050235478807,
      "loss": 0.7801,
      "step": 850
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.0001662480376766091,
      "loss": 0.8736,
      "step": 860
    },
    {
      "epoch": 0.68,
      "learning_rate": 0.00016585557299843016,
      "loss": 0.4921,
      "step": 870
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.00016546310832025118,
      "loss": 0.7076,
      "step": 880
    },
    {
      "epoch": 0.7,
      "learning_rate": 0.00016507064364207222,
      "loss": 0.7124,
      "step": 890
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016467817896389326,
      "loss": 0.8552,
      "step": 900
    },
    {
      "epoch": 0.71,
      "learning_rate": 0.00016428571428571428,
      "loss": 0.9654,
      "step": 910
    },
    {
      "epoch": 0.72,
      "learning_rate": 0.00016389324960753535,
      "loss": 0.6044,
      "step": 920
    },
    {
      "epoch": 0.73,
      "learning_rate": 0.00016350078492935637,
      "loss": 0.7184,
      "step": 930
    },
    {
      "epoch": 0.74,
      "learning_rate": 0.00016310832025117738,
      "loss": 0.649,
      "step": 940
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016271585557299845,
      "loss": 0.6896,
      "step": 950
    },
    {
      "epoch": 0.75,
      "learning_rate": 0.00016232339089481947,
      "loss": 0.4548,
      "step": 960
    },
    {
      "epoch": 0.76,
      "learning_rate": 0.0001619309262166405,
      "loss": 0.9116,
      "step": 970
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00016153846153846155,
      "loss": 0.6909,
      "step": 980
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016114599686028257,
      "loss": 0.7693,
      "step": 990
    },
    {
      "epoch": 0.78,
      "learning_rate": 0.00016075353218210364,
      "loss": 0.7824,
      "step": 1000
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.81717277486911,
      "eval_loss": 0.6625257730484009,
      "eval_runtime": 104.6909,
      "eval_samples_per_second": 45.61,
      "eval_steps_per_second": 5.703,
      "step": 1000
    },
    {
      "epoch": 0.79,
      "learning_rate": 0.00016036106750392466,
      "loss": 0.8368,
      "step": 1010
    },
    {
      "epoch": 0.8,
      "learning_rate": 0.00015996860282574567,
      "loss": 0.6853,
      "step": 1020
    },
    {
      "epoch": 0.81,
      "learning_rate": 0.00015957613814756674,
      "loss": 0.6957,
      "step": 1030
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.00015918367346938776,
      "loss": 0.7415,
      "step": 1040
    },
    {
      "epoch": 0.82,
      "learning_rate": 0.0001587912087912088,
      "loss": 0.719,
      "step": 1050
    },
    {
      "epoch": 0.83,
      "learning_rate": 0.00015839874411302984,
      "loss": 0.6999,
      "step": 1060
    },
    {
      "epoch": 0.84,
      "learning_rate": 0.00015800627943485086,
      "loss": 0.6508,
      "step": 1070
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.00015761381475667193,
      "loss": 0.7951,
      "step": 1080
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015722135007849295,
      "loss": 0.646,
      "step": 1090
    },
    {
      "epoch": 0.86,
      "learning_rate": 0.00015682888540031396,
      "loss": 0.568,
      "step": 1100
    },
    {
      "epoch": 0.87,
      "learning_rate": 0.00015643642072213503,
      "loss": 0.7428,
      "step": 1110
    },
    {
      "epoch": 0.88,
      "learning_rate": 0.00015604395604395605,
      "loss": 0.6068,
      "step": 1120
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.0001556514913657771,
      "loss": 0.5023,
      "step": 1130
    },
    {
      "epoch": 0.89,
      "learning_rate": 0.00015525902668759813,
      "loss": 0.7015,
      "step": 1140
    },
    {
      "epoch": 0.9,
      "learning_rate": 0.00015486656200941915,
      "loss": 0.4978,
      "step": 1150
    },
    {
      "epoch": 0.91,
      "learning_rate": 0.0001544740973312402,
      "loss": 0.572,
      "step": 1160
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.00015408163265306124,
      "loss": 0.6404,
      "step": 1170
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00015368916797488225,
      "loss": 0.7149,
      "step": 1180
    },
    {
      "epoch": 0.93,
      "learning_rate": 0.00015329670329670332,
      "loss": 0.6737,
      "step": 1190
    },
    {
      "epoch": 0.94,
      "learning_rate": 0.00015290423861852434,
      "loss": 0.5757,
      "step": 1200
    },
    {
      "epoch": 0.95,
      "learning_rate": 0.00015251177394034538,
      "loss": 0.5646,
      "step": 1210
    },
    {
      "epoch": 0.96,
      "learning_rate": 0.00015211930926216642,
      "loss": 0.8011,
      "step": 1220
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015172684458398744,
      "loss": 0.7641,
      "step": 1230
    },
    {
      "epoch": 0.97,
      "learning_rate": 0.00015133437990580848,
      "loss": 0.7895,
      "step": 1240
    },
    {
      "epoch": 0.98,
      "learning_rate": 0.00015094191522762953,
      "loss": 0.7088,
      "step": 1250
    },
    {
      "epoch": 0.99,
      "learning_rate": 0.00015054945054945054,
      "loss": 0.6559,
      "step": 1260
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00015015698587127159,
      "loss": 0.7573,
      "step": 1270
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.00014976452119309263,
      "loss": 0.8222,
      "step": 1280
    },
    {
      "epoch": 1.01,
      "learning_rate": 0.00014937205651491367,
      "loss": 0.5375,
      "step": 1290
    },
    {
      "epoch": 1.02,
      "learning_rate": 0.00014897959183673472,
      "loss": 0.5014,
      "step": 1300
    },
    {
      "epoch": 1.03,
      "learning_rate": 0.00014858712715855573,
      "loss": 0.4494,
      "step": 1310
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014819466248037677,
      "loss": 0.6057,
      "step": 1320
    },
    {
      "epoch": 1.04,
      "learning_rate": 0.00014780219780219782,
      "loss": 0.5502,
      "step": 1330
    },
    {
      "epoch": 1.05,
      "learning_rate": 0.00014740973312401883,
      "loss": 0.5677,
      "step": 1340
    },
    {
      "epoch": 1.06,
      "learning_rate": 0.00014701726844583988,
      "loss": 0.5548,
      "step": 1350
    },
    {
      "epoch": 1.07,
      "learning_rate": 0.00014662480376766092,
      "loss": 0.5428,
      "step": 1360
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014623233908948196,
      "loss": 0.6924,
      "step": 1370
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00014583987441130298,
      "loss": 0.4367,
      "step": 1380
    },
    {
      "epoch": 1.09,
      "learning_rate": 0.00014544740973312402,
      "loss": 0.4269,
      "step": 1390
    },
    {
      "epoch": 1.1,
      "learning_rate": 0.00014505494505494506,
      "loss": 0.4049,
      "step": 1400
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.0001446624803767661,
      "loss": 0.4435,
      "step": 1410
    },
    {
      "epoch": 1.11,
      "learning_rate": 0.00014427001569858712,
      "loss": 0.5317,
      "step": 1420
    },
    {
      "epoch": 1.12,
      "learning_rate": 0.00014387755102040817,
      "loss": 0.7213,
      "step": 1430
    },
    {
      "epoch": 1.13,
      "learning_rate": 0.0001434850863422292,
      "loss": 0.8834,
      "step": 1440
    },
    {
      "epoch": 1.14,
      "learning_rate": 0.00014309262166405025,
      "loss": 0.5434,
      "step": 1450
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00014270015698587127,
      "loss": 0.5588,
      "step": 1460
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.0001423076923076923,
      "loss": 0.6011,
      "step": 1470
    },
    {
      "epoch": 1.16,
      "learning_rate": 0.00014191522762951335,
      "loss": 0.4124,
      "step": 1480
    },
    {
      "epoch": 1.17,
      "learning_rate": 0.0001415227629513344,
      "loss": 0.5477,
      "step": 1490
    },
    {
      "epoch": 1.18,
      "learning_rate": 0.00014113029827315541,
      "loss": 0.4806,
      "step": 1500
    },
    {
      "epoch": 1.18,
      "eval_accuracy": 0.8617801047120419,
      "eval_loss": 0.4951171875,
      "eval_runtime": 104.5532,
      "eval_samples_per_second": 45.671,
      "eval_steps_per_second": 5.71,
      "step": 1500
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.00014073783359497646,
      "loss": 0.5714,
      "step": 1510
    },
    {
      "epoch": 1.19,
      "learning_rate": 0.0001403453689167975,
      "loss": 0.4919,
      "step": 1520
    },
    {
      "epoch": 1.2,
      "learning_rate": 0.00013995290423861854,
      "loss": 0.6668,
      "step": 1530
    },
    {
      "epoch": 1.21,
      "learning_rate": 0.00013956043956043956,
      "loss": 0.6993,
      "step": 1540
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.0001391679748822606,
      "loss": 0.7433,
      "step": 1550
    },
    {
      "epoch": 1.22,
      "learning_rate": 0.00013877551020408165,
      "loss": 0.6475,
      "step": 1560
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00013838304552590266,
      "loss": 0.5059,
      "step": 1570
    },
    {
      "epoch": 1.24,
      "learning_rate": 0.0001379905808477237,
      "loss": 0.6121,
      "step": 1580
    },
    {
      "epoch": 1.25,
      "learning_rate": 0.00013759811616954475,
      "loss": 0.7062,
      "step": 1590
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.0001372056514913658,
      "loss": 0.4882,
      "step": 1600
    },
    {
      "epoch": 1.26,
      "learning_rate": 0.00013681318681318683,
      "loss": 0.5246,
      "step": 1610
    },
    {
      "epoch": 1.27,
      "learning_rate": 0.00013642072213500785,
      "loss": 0.5211,
      "step": 1620
    },
    {
      "epoch": 1.28,
      "learning_rate": 0.0001360282574568289,
      "loss": 0.5506,
      "step": 1630
    },
    {
      "epoch": 1.29,
      "learning_rate": 0.00013563579277864994,
      "loss": 0.3996,
      "step": 1640
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.00013524332810047095,
      "loss": 0.5601,
      "step": 1650
    },
    {
      "epoch": 1.3,
      "learning_rate": 0.000134850863422292,
      "loss": 0.6195,
      "step": 1660
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00013445839874411304,
      "loss": 0.5741,
      "step": 1670
    },
    {
      "epoch": 1.32,
      "learning_rate": 0.00013406593406593405,
      "loss": 0.5964,
      "step": 1680
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013367346938775512,
      "loss": 0.4962,
      "step": 1690
    },
    {
      "epoch": 1.33,
      "learning_rate": 0.00013328100470957614,
      "loss": 0.4467,
      "step": 1700
    },
    {
      "epoch": 1.34,
      "learning_rate": 0.00013288854003139718,
      "loss": 0.5517,
      "step": 1710
    },
    {
      "epoch": 1.35,
      "learning_rate": 0.00013249607535321823,
      "loss": 0.4578,
      "step": 1720
    },
    {
      "epoch": 1.36,
      "learning_rate": 0.00013210361067503924,
      "loss": 0.6583,
      "step": 1730
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013171114599686029,
      "loss": 0.5374,
      "step": 1740
    },
    {
      "epoch": 1.37,
      "learning_rate": 0.00013131868131868133,
      "loss": 0.7157,
      "step": 1750
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00013092621664050234,
      "loss": 0.3906,
      "step": 1760
    },
    {
      "epoch": 1.39,
      "learning_rate": 0.00013053375196232341,
      "loss": 0.6177,
      "step": 1770
    },
    {
      "epoch": 1.4,
      "learning_rate": 0.00013014128728414443,
      "loss": 0.4234,
      "step": 1780
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012974882260596545,
      "loss": 0.4714,
      "step": 1790
    },
    {
      "epoch": 1.41,
      "learning_rate": 0.00012935635792778652,
      "loss": 0.5318,
      "step": 1800
    },
    {
      "epoch": 1.42,
      "learning_rate": 0.00012896389324960753,
      "loss": 0.6005,
      "step": 1810
    },
    {
      "epoch": 1.43,
      "learning_rate": 0.00012857142857142858,
      "loss": 0.514,
      "step": 1820
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012817896389324962,
      "loss": 0.4319,
      "step": 1830
    },
    {
      "epoch": 1.44,
      "learning_rate": 0.00012778649921507063,
      "loss": 0.3819,
      "step": 1840
    },
    {
      "epoch": 1.45,
      "learning_rate": 0.0001273940345368917,
      "loss": 0.7177,
      "step": 1850
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00012700156985871272,
      "loss": 0.5326,
      "step": 1860
    },
    {
      "epoch": 1.47,
      "learning_rate": 0.00012660910518053374,
      "loss": 0.5763,
      "step": 1870
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.0001262166405023548,
      "loss": 0.6364,
      "step": 1880
    },
    {
      "epoch": 1.48,
      "learning_rate": 0.00012582417582417582,
      "loss": 0.6008,
      "step": 1890
    },
    {
      "epoch": 1.49,
      "learning_rate": 0.00012543171114599687,
      "loss": 0.6064,
      "step": 1900
    },
    {
      "epoch": 1.5,
      "learning_rate": 0.0001250392464678179,
      "loss": 0.4222,
      "step": 1910
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012464678178963893,
      "loss": 0.4328,
      "step": 1920
    },
    {
      "epoch": 1.51,
      "learning_rate": 0.00012425431711146,
      "loss": 0.4382,
      "step": 1930
    },
    {
      "epoch": 1.52,
      "learning_rate": 0.000123861852433281,
      "loss": 0.4715,
      "step": 1940
    },
    {
      "epoch": 1.53,
      "learning_rate": 0.00012346938775510203,
      "loss": 0.5137,
      "step": 1950
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001230769230769231,
      "loss": 0.5555,
      "step": 1960
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.0001226844583987441,
      "loss": 0.6333,
      "step": 1970
    },
    {
      "epoch": 1.55,
      "learning_rate": 0.00012229199372056516,
      "loss": 0.5118,
      "step": 1980
    },
    {
      "epoch": 1.56,
      "learning_rate": 0.0001218995290423862,
      "loss": 0.3767,
      "step": 1990
    },
    {
      "epoch": 1.57,
      "learning_rate": 0.00012150706436420723,
      "loss": 0.6206,
      "step": 2000
    },
    {
      "epoch": 1.57,
      "eval_accuracy": 0.88,
      "eval_loss": 0.4433819651603699,
      "eval_runtime": 104.7073,
      "eval_samples_per_second": 45.603,
      "eval_steps_per_second": 5.702,
      "step": 2000
    },
    {
      "epoch": 1.58,
      "learning_rate": 0.00012111459968602827,
      "loss": 0.4409,
      "step": 2010
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.0001207221350078493,
      "loss": 0.5458,
      "step": 2020
    },
    {
      "epoch": 1.59,
      "learning_rate": 0.00012032967032967033,
      "loss": 0.369,
      "step": 2030
    },
    {
      "epoch": 1.6,
      "learning_rate": 0.00011993720565149137,
      "loss": 0.5078,
      "step": 2040
    },
    {
      "epoch": 1.61,
      "learning_rate": 0.0001195447409733124,
      "loss": 0.5815,
      "step": 2050
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011915227629513343,
      "loss": 0.4357,
      "step": 2060
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00011875981161695449,
      "loss": 0.5955,
      "step": 2070
    },
    {
      "epoch": 1.63,
      "learning_rate": 0.00011836734693877552,
      "loss": 0.4177,
      "step": 2080
    },
    {
      "epoch": 1.64,
      "learning_rate": 0.00011797488226059654,
      "loss": 0.5155,
      "step": 2090
    },
    {
      "epoch": 1.65,
      "learning_rate": 0.00011758241758241759,
      "loss": 0.5691,
      "step": 2100
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011718995290423862,
      "loss": 0.5891,
      "step": 2110
    },
    {
      "epoch": 1.66,
      "learning_rate": 0.00011679748822605966,
      "loss": 0.5013,
      "step": 2120
    },
    {
      "epoch": 1.67,
      "learning_rate": 0.0001164050235478807,
      "loss": 0.4859,
      "step": 2130
    },
    {
      "epoch": 1.68,
      "learning_rate": 0.00011601255886970172,
      "loss": 0.4267,
      "step": 2140
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00011562009419152278,
      "loss": 0.3648,
      "step": 2150
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011522762951334381,
      "loss": 0.5093,
      "step": 2160
    },
    {
      "epoch": 1.7,
      "learning_rate": 0.00011483516483516483,
      "loss": 0.6334,
      "step": 2170
    },
    {
      "epoch": 1.71,
      "learning_rate": 0.00011444270015698588,
      "loss": 0.4285,
      "step": 2180
    },
    {
      "epoch": 1.72,
      "learning_rate": 0.00011405023547880691,
      "loss": 0.4107,
      "step": 2190
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011365777080062794,
      "loss": 0.4215,
      "step": 2200
    },
    {
      "epoch": 1.73,
      "learning_rate": 0.00011326530612244898,
      "loss": 0.574,
      "step": 2210
    },
    {
      "epoch": 1.74,
      "learning_rate": 0.00011287284144427001,
      "loss": 0.5277,
      "step": 2220
    },
    {
      "epoch": 1.75,
      "learning_rate": 0.00011248037676609107,
      "loss": 0.4283,
      "step": 2230
    },
    {
      "epoch": 1.76,
      "learning_rate": 0.0001120879120879121,
      "loss": 0.5907,
      "step": 2240
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011169544740973312,
      "loss": 0.4093,
      "step": 2250
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00011130298273155417,
      "loss": 0.5197,
      "step": 2260
    },
    {
      "epoch": 1.78,
      "learning_rate": 0.0001109105180533752,
      "loss": 0.6211,
      "step": 2270
    },
    {
      "epoch": 1.79,
      "learning_rate": 0.00011051805337519623,
      "loss": 0.6174,
      "step": 2280
    },
    {
      "epoch": 1.8,
      "learning_rate": 0.00011012558869701728,
      "loss": 0.6469,
      "step": 2290
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.0001097331240188383,
      "loss": 0.5121,
      "step": 2300
    },
    {
      "epoch": 1.81,
      "learning_rate": 0.00010934065934065933,
      "loss": 0.5431,
      "step": 2310
    },
    {
      "epoch": 1.82,
      "learning_rate": 0.00010894819466248039,
      "loss": 0.4473,
      "step": 2320
    },
    {
      "epoch": 1.83,
      "learning_rate": 0.0001085557299843014,
      "loss": 0.4877,
      "step": 2330
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010816326530612246,
      "loss": 0.5579,
      "step": 2340
    },
    {
      "epoch": 1.84,
      "learning_rate": 0.00010777080062794349,
      "loss": 0.4271,
      "step": 2350
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.00010737833594976452,
      "loss": 0.5546,
      "step": 2360
    },
    {
      "epoch": 1.86,
      "learning_rate": 0.00010698587127158557,
      "loss": 0.3829,
      "step": 2370
    },
    {
      "epoch": 1.87,
      "learning_rate": 0.0001065934065934066,
      "loss": 0.5093,
      "step": 2380
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010620094191522762,
      "loss": 0.3551,
      "step": 2390
    },
    {
      "epoch": 1.88,
      "learning_rate": 0.00010580847723704868,
      "loss": 0.4126,
      "step": 2400
    },
    {
      "epoch": 1.89,
      "learning_rate": 0.0001054160125588697,
      "loss": 0.4307,
      "step": 2410
    },
    {
      "epoch": 1.9,
      "learning_rate": 0.00010502354788069075,
      "loss": 0.4657,
      "step": 2420
    },
    {
      "epoch": 1.91,
      "learning_rate": 0.00010463108320251178,
      "loss": 0.39,
      "step": 2430
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010423861852433281,
      "loss": 0.4949,
      "step": 2440
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00010384615384615386,
      "loss": 0.5374,
      "step": 2450
    },
    {
      "epoch": 1.93,
      "learning_rate": 0.00010345368916797489,
      "loss": 0.3725,
      "step": 2460
    },
    {
      "epoch": 1.94,
      "learning_rate": 0.00010306122448979591,
      "loss": 0.6204,
      "step": 2470
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010266875981161697,
      "loss": 0.6382,
      "step": 2480
    },
    {
      "epoch": 1.95,
      "learning_rate": 0.00010227629513343799,
      "loss": 0.4172,
      "step": 2490
    },
    {
      "epoch": 1.96,
      "learning_rate": 0.00010188383045525902,
      "loss": 0.5096,
      "step": 2500
    },
    {
      "epoch": 1.96,
      "eval_accuracy": 0.8682722513089005,
      "eval_loss": 0.4937422275543213,
      "eval_runtime": 104.7964,
      "eval_samples_per_second": 45.565,
      "eval_steps_per_second": 5.697,
      "step": 2500
    },
    {
      "epoch": 1.97,
      "learning_rate": 0.00010149136577708007,
      "loss": 0.7216,
      "step": 2510
    },
    {
      "epoch": 1.98,
      "learning_rate": 0.0001010989010989011,
      "loss": 0.5306,
      "step": 2520
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010070643642072215,
      "loss": 0.3693,
      "step": 2530
    },
    {
      "epoch": 1.99,
      "learning_rate": 0.00010031397174254318,
      "loss": 0.2895,
      "step": 2540
    },
    {
      "epoch": 2.0,
      "learning_rate": 9.992150706436422e-05,
      "loss": 0.3191,
      "step": 2550
    },
    {
      "epoch": 2.01,
      "learning_rate": 9.952904238618525e-05,
      "loss": 0.3354,
      "step": 2560
    },
    {
      "epoch": 2.02,
      "learning_rate": 9.913657770800628e-05,
      "loss": 0.5719,
      "step": 2570
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.874411302982732e-05,
      "loss": 0.3117,
      "step": 2580
    },
    {
      "epoch": 2.03,
      "learning_rate": 9.835164835164835e-05,
      "loss": 0.4133,
      "step": 2590
    },
    {
      "epoch": 2.04,
      "learning_rate": 9.79591836734694e-05,
      "loss": 0.3336,
      "step": 2600
    },
    {
      "epoch": 2.05,
      "learning_rate": 9.756671899529042e-05,
      "loss": 0.1835,
      "step": 2610
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.717425431711147e-05,
      "loss": 0.3862,
      "step": 2620
    },
    {
      "epoch": 2.06,
      "learning_rate": 9.67817896389325e-05,
      "loss": 0.4571,
      "step": 2630
    },
    {
      "epoch": 2.07,
      "learning_rate": 9.638932496075354e-05,
      "loss": 0.5059,
      "step": 2640
    },
    {
      "epoch": 2.08,
      "learning_rate": 9.599686028257457e-05,
      "loss": 0.3877,
      "step": 2650
    },
    {
      "epoch": 2.09,
      "learning_rate": 9.560439560439561e-05,
      "loss": 0.4755,
      "step": 2660
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.521193092621664e-05,
      "loss": 0.33,
      "step": 2670
    },
    {
      "epoch": 2.1,
      "learning_rate": 9.481946624803768e-05,
      "loss": 0.3211,
      "step": 2680
    },
    {
      "epoch": 2.11,
      "learning_rate": 9.442700156985871e-05,
      "loss": 0.5344,
      "step": 2690
    },
    {
      "epoch": 2.12,
      "learning_rate": 9.403453689167976e-05,
      "loss": 0.503,
      "step": 2700
    },
    {
      "epoch": 2.13,
      "learning_rate": 9.364207221350079e-05,
      "loss": 0.2102,
      "step": 2710
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.324960753532183e-05,
      "loss": 0.5119,
      "step": 2720
    },
    {
      "epoch": 2.14,
      "learning_rate": 9.285714285714286e-05,
      "loss": 0.2937,
      "step": 2730
    },
    {
      "epoch": 2.15,
      "learning_rate": 9.246467817896389e-05,
      "loss": 0.2749,
      "step": 2740
    },
    {
      "epoch": 2.16,
      "learning_rate": 9.207221350078493e-05,
      "loss": 0.5112,
      "step": 2750
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.167974882260597e-05,
      "loss": 0.4431,
      "step": 2760
    },
    {
      "epoch": 2.17,
      "learning_rate": 9.1287284144427e-05,
      "loss": 0.3892,
      "step": 2770
    },
    {
      "epoch": 2.18,
      "learning_rate": 9.089481946624803e-05,
      "loss": 0.4078,
      "step": 2780
    },
    {
      "epoch": 2.19,
      "learning_rate": 9.050235478806908e-05,
      "loss": 0.4673,
      "step": 2790
    },
    {
      "epoch": 2.2,
      "learning_rate": 9.010989010989012e-05,
      "loss": 0.3158,
      "step": 2800
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.971742543171115e-05,
      "loss": 0.2539,
      "step": 2810
    },
    {
      "epoch": 2.21,
      "learning_rate": 8.932496075353218e-05,
      "loss": 0.4073,
      "step": 2820
    },
    {
      "epoch": 2.22,
      "learning_rate": 8.893249607535322e-05,
      "loss": 0.4748,
      "step": 2830
    },
    {
      "epoch": 2.23,
      "learning_rate": 8.854003139717426e-05,
      "loss": 0.4274,
      "step": 2840
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.81475667189953e-05,
      "loss": 0.2248,
      "step": 2850
    },
    {
      "epoch": 2.24,
      "learning_rate": 8.775510204081632e-05,
      "loss": 0.2974,
      "step": 2860
    },
    {
      "epoch": 2.25,
      "learning_rate": 8.736263736263737e-05,
      "loss": 0.3354,
      "step": 2870
    },
    {
      "epoch": 2.26,
      "learning_rate": 8.697017268445841e-05,
      "loss": 0.2687,
      "step": 2880
    },
    {
      "epoch": 2.27,
      "learning_rate": 8.657770800627944e-05,
      "loss": 0.3089,
      "step": 2890
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.618524332810047e-05,
      "loss": 0.326,
      "step": 2900
    },
    {
      "epoch": 2.28,
      "learning_rate": 8.579277864992151e-05,
      "loss": 0.437,
      "step": 2910
    },
    {
      "epoch": 2.29,
      "learning_rate": 8.540031397174256e-05,
      "loss": 0.3589,
      "step": 2920
    },
    {
      "epoch": 2.3,
      "learning_rate": 8.500784929356358e-05,
      "loss": 0.3766,
      "step": 2930
    },
    {
      "epoch": 2.31,
      "learning_rate": 8.461538461538461e-05,
      "loss": 0.2418,
      "step": 2940
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.422291993720566e-05,
      "loss": 0.3981,
      "step": 2950
    },
    {
      "epoch": 2.32,
      "learning_rate": 8.38304552590267e-05,
      "loss": 0.3544,
      "step": 2960
    },
    {
      "epoch": 2.33,
      "learning_rate": 8.343799058084773e-05,
      "loss": 0.2367,
      "step": 2970
    },
    {
      "epoch": 2.34,
      "learning_rate": 8.304552590266876e-05,
      "loss": 0.3049,
      "step": 2980
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.26530612244898e-05,
      "loss": 0.4299,
      "step": 2990
    },
    {
      "epoch": 2.35,
      "learning_rate": 8.226059654631083e-05,
      "loss": 0.4576,
      "step": 3000
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.8906806282722514,
      "eval_loss": 0.4059741497039795,
      "eval_runtime": 104.8886,
      "eval_samples_per_second": 45.524,
      "eval_steps_per_second": 5.692,
      "step": 3000
    },
    {
      "epoch": 2.36,
      "learning_rate": 8.186813186813188e-05,
      "loss": 0.2715,
      "step": 3010
    },
    {
      "epoch": 2.37,
      "learning_rate": 8.14756671899529e-05,
      "loss": 0.4298,
      "step": 3020
    },
    {
      "epoch": 2.38,
      "learning_rate": 8.108320251177395e-05,
      "loss": 0.2927,
      "step": 3030
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.069073783359498e-05,
      "loss": 0.2767,
      "step": 3040
    },
    {
      "epoch": 2.39,
      "learning_rate": 8.029827315541602e-05,
      "loss": 0.2875,
      "step": 3050
    },
    {
      "epoch": 2.4,
      "learning_rate": 7.990580847723705e-05,
      "loss": 0.3713,
      "step": 3060
    },
    {
      "epoch": 2.41,
      "learning_rate": 7.951334379905809e-05,
      "loss": 0.3841,
      "step": 3070
    },
    {
      "epoch": 2.42,
      "learning_rate": 7.912087912087912e-05,
      "loss": 0.3523,
      "step": 3080
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.872841444270017e-05,
      "loss": 0.3103,
      "step": 3090
    },
    {
      "epoch": 2.43,
      "learning_rate": 7.83359497645212e-05,
      "loss": 0.4337,
      "step": 3100
    },
    {
      "epoch": 2.44,
      "learning_rate": 7.794348508634224e-05,
      "loss": 0.3,
      "step": 3110
    },
    {
      "epoch": 2.45,
      "learning_rate": 7.755102040816327e-05,
      "loss": 0.3451,
      "step": 3120
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.715855572998431e-05,
      "loss": 0.4831,
      "step": 3130
    },
    {
      "epoch": 2.46,
      "learning_rate": 7.676609105180534e-05,
      "loss": 0.4635,
      "step": 3140
    },
    {
      "epoch": 2.47,
      "learning_rate": 7.637362637362637e-05,
      "loss": 0.3224,
      "step": 3150
    },
    {
      "epoch": 2.48,
      "learning_rate": 7.598116169544741e-05,
      "loss": 0.4548,
      "step": 3160
    },
    {
      "epoch": 2.49,
      "learning_rate": 7.558869701726846e-05,
      "loss": 0.3212,
      "step": 3170
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.519623233908949e-05,
      "loss": 0.3922,
      "step": 3180
    },
    {
      "epoch": 2.5,
      "learning_rate": 7.480376766091051e-05,
      "loss": 0.4831,
      "step": 3190
    },
    {
      "epoch": 2.51,
      "learning_rate": 7.441130298273156e-05,
      "loss": 0.1519,
      "step": 3200
    },
    {
      "epoch": 2.52,
      "learning_rate": 7.40188383045526e-05,
      "loss": 0.301,
      "step": 3210
    },
    {
      "epoch": 2.53,
      "learning_rate": 7.362637362637363e-05,
      "loss": 0.3742,
      "step": 3220
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.323390894819466e-05,
      "loss": 0.2249,
      "step": 3230
    },
    {
      "epoch": 2.54,
      "learning_rate": 7.28414442700157e-05,
      "loss": 0.5049,
      "step": 3240
    },
    {
      "epoch": 2.55,
      "learning_rate": 7.244897959183675e-05,
      "loss": 0.2744,
      "step": 3250
    },
    {
      "epoch": 2.56,
      "learning_rate": 7.205651491365776e-05,
      "loss": 0.4095,
      "step": 3260
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.16640502354788e-05,
      "loss": 0.5031,
      "step": 3270
    },
    {
      "epoch": 2.57,
      "learning_rate": 7.127158555729985e-05,
      "loss": 0.3179,
      "step": 3280
    },
    {
      "epoch": 2.58,
      "learning_rate": 7.087912087912089e-05,
      "loss": 0.2982,
      "step": 3290
    },
    {
      "epoch": 2.59,
      "learning_rate": 7.048665620094191e-05,
      "loss": 0.2934,
      "step": 3300
    },
    {
      "epoch": 2.6,
      "learning_rate": 7.009419152276295e-05,
      "loss": 0.395,
      "step": 3310
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.9701726844584e-05,
      "loss": 0.2843,
      "step": 3320
    },
    {
      "epoch": 2.61,
      "learning_rate": 6.930926216640504e-05,
      "loss": 0.2671,
      "step": 3330
    },
    {
      "epoch": 2.62,
      "learning_rate": 6.891679748822605e-05,
      "loss": 0.2192,
      "step": 3340
    },
    {
      "epoch": 2.63,
      "learning_rate": 6.85243328100471e-05,
      "loss": 0.298,
      "step": 3350
    },
    {
      "epoch": 2.64,
      "learning_rate": 6.813186813186814e-05,
      "loss": 0.3189,
      "step": 3360
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.773940345368918e-05,
      "loss": 0.3521,
      "step": 3370
    },
    {
      "epoch": 2.65,
      "learning_rate": 6.73469387755102e-05,
      "loss": 0.293,
      "step": 3380
    },
    {
      "epoch": 2.66,
      "learning_rate": 6.695447409733124e-05,
      "loss": 0.3281,
      "step": 3390
    },
    {
      "epoch": 2.67,
      "learning_rate": 6.656200941915228e-05,
      "loss": 0.2706,
      "step": 3400
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.616954474097331e-05,
      "loss": 0.2927,
      "step": 3410
    },
    {
      "epoch": 2.68,
      "learning_rate": 6.577708006279434e-05,
      "loss": 0.2787,
      "step": 3420
    },
    {
      "epoch": 2.69,
      "learning_rate": 6.538461538461539e-05,
      "loss": 0.3786,
      "step": 3430
    },
    {
      "epoch": 2.7,
      "learning_rate": 6.499215070643643e-05,
      "loss": 0.2341,
      "step": 3440
    },
    {
      "epoch": 2.71,
      "learning_rate": 6.459968602825746e-05,
      "loss": 0.4556,
      "step": 3450
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.420722135007849e-05,
      "loss": 0.1937,
      "step": 3460
    },
    {
      "epoch": 2.72,
      "learning_rate": 6.381475667189953e-05,
      "loss": 0.4824,
      "step": 3470
    },
    {
      "epoch": 2.73,
      "learning_rate": 6.342229199372057e-05,
      "loss": 0.4207,
      "step": 3480
    },
    {
      "epoch": 2.74,
      "learning_rate": 6.30298273155416e-05,
      "loss": 0.4383,
      "step": 3490
    },
    {
      "epoch": 2.75,
      "learning_rate": 6.263736263736263e-05,
      "loss": 0.3284,
      "step": 3500
    },
    {
      "epoch": 2.75,
      "eval_accuracy": 0.9080628272251309,
      "eval_loss": 0.34141793847084045,
      "eval_runtime": 105.2957,
      "eval_samples_per_second": 45.348,
      "eval_steps_per_second": 5.67,
      "step": 3500
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.224489795918368e-05,
      "loss": 0.2761,
      "step": 3510
    },
    {
      "epoch": 2.76,
      "learning_rate": 6.185243328100472e-05,
      "loss": 0.2809,
      "step": 3520
    },
    {
      "epoch": 2.77,
      "learning_rate": 6.145996860282575e-05,
      "loss": 0.369,
      "step": 3530
    },
    {
      "epoch": 2.78,
      "learning_rate": 6.106750392464678e-05,
      "loss": 0.4629,
      "step": 3540
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.067503924646782e-05,
      "loss": 0.4081,
      "step": 3550
    },
    {
      "epoch": 2.79,
      "learning_rate": 6.028257456828885e-05,
      "loss": 0.216,
      "step": 3560
    },
    {
      "epoch": 2.8,
      "learning_rate": 5.9890109890109894e-05,
      "loss": 0.4566,
      "step": 3570
    },
    {
      "epoch": 2.81,
      "learning_rate": 5.949764521193093e-05,
      "loss": 0.2259,
      "step": 3580
    },
    {
      "epoch": 2.82,
      "learning_rate": 5.910518053375197e-05,
      "loss": 0.2951,
      "step": 3590
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.8712715855572997e-05,
      "loss": 0.346,
      "step": 3600
    },
    {
      "epoch": 2.83,
      "learning_rate": 5.832025117739404e-05,
      "loss": 0.3088,
      "step": 3610
    },
    {
      "epoch": 2.84,
      "learning_rate": 5.7927786499215076e-05,
      "loss": 0.3857,
      "step": 3620
    },
    {
      "epoch": 2.85,
      "learning_rate": 5.753532182103611e-05,
      "loss": 0.32,
      "step": 3630
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.714285714285714e-05,
      "loss": 0.3178,
      "step": 3640
    },
    {
      "epoch": 2.86,
      "learning_rate": 5.6750392464678185e-05,
      "loss": 0.3984,
      "step": 3650
    },
    {
      "epoch": 2.87,
      "learning_rate": 5.635792778649922e-05,
      "loss": 0.4079,
      "step": 3660
    },
    {
      "epoch": 2.88,
      "learning_rate": 5.596546310832025e-05,
      "loss": 0.2981,
      "step": 3670
    },
    {
      "epoch": 2.89,
      "learning_rate": 5.557299843014129e-05,
      "loss": 0.179,
      "step": 3680
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.518053375196233e-05,
      "loss": 0.3578,
      "step": 3690
    },
    {
      "epoch": 2.9,
      "learning_rate": 5.4788069073783366e-05,
      "loss": 0.235,
      "step": 3700
    },
    {
      "epoch": 2.91,
      "learning_rate": 5.4395604395604396e-05,
      "loss": 0.2339,
      "step": 3710
    },
    {
      "epoch": 2.92,
      "learning_rate": 5.400313971742543e-05,
      "loss": 0.1819,
      "step": 3720
    },
    {
      "epoch": 2.93,
      "learning_rate": 5.3610675039246475e-05,
      "loss": 0.2206,
      "step": 3730
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.321821036106751e-05,
      "loss": 0.1959,
      "step": 3740
    },
    {
      "epoch": 2.94,
      "learning_rate": 5.282574568288854e-05,
      "loss": 0.3743,
      "step": 3750
    },
    {
      "epoch": 2.95,
      "learning_rate": 5.243328100470958e-05,
      "loss": 0.3278,
      "step": 3760
    },
    {
      "epoch": 2.96,
      "learning_rate": 5.2040816326530614e-05,
      "loss": 0.3743,
      "step": 3770
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.164835164835166e-05,
      "loss": 0.2863,
      "step": 3780
    },
    {
      "epoch": 2.97,
      "learning_rate": 5.1255886970172686e-05,
      "loss": 0.347,
      "step": 3790
    },
    {
      "epoch": 2.98,
      "learning_rate": 5.086342229199372e-05,
      "loss": 0.2294,
      "step": 3800
    },
    {
      "epoch": 2.99,
      "learning_rate": 5.047095761381476e-05,
      "loss": 0.4746,
      "step": 3810
    },
    {
      "epoch": 3.0,
      "learning_rate": 5.007849293563579e-05,
      "loss": 0.3479,
      "step": 3820
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.968602825745683e-05,
      "loss": 0.1066,
      "step": 3830
    },
    {
      "epoch": 3.01,
      "learning_rate": 4.929356357927787e-05,
      "loss": 0.2885,
      "step": 3840
    },
    {
      "epoch": 3.02,
      "learning_rate": 4.8901098901098904e-05,
      "loss": 0.1881,
      "step": 3850
    },
    {
      "epoch": 3.03,
      "learning_rate": 4.850863422291994e-05,
      "loss": 0.3133,
      "step": 3860
    },
    {
      "epoch": 3.04,
      "learning_rate": 4.811616954474098e-05,
      "loss": 0.2348,
      "step": 3870
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.772370486656201e-05,
      "loss": 0.2296,
      "step": 3880
    },
    {
      "epoch": 3.05,
      "learning_rate": 4.733124018838305e-05,
      "loss": 0.2402,
      "step": 3890
    },
    {
      "epoch": 3.06,
      "learning_rate": 4.6938775510204086e-05,
      "loss": 0.1956,
      "step": 3900
    },
    {
      "epoch": 3.07,
      "learning_rate": 4.654631083202512e-05,
      "loss": 0.1585,
      "step": 3910
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.2759,
      "step": 3920
    },
    {
      "epoch": 3.08,
      "learning_rate": 4.5761381475667194e-05,
      "loss": 0.2603,
      "step": 3930
    },
    {
      "epoch": 3.09,
      "learning_rate": 4.5368916797488224e-05,
      "loss": 0.2199,
      "step": 3940
    },
    {
      "epoch": 3.1,
      "learning_rate": 4.497645211930927e-05,
      "loss": 0.2321,
      "step": 3950
    },
    {
      "epoch": 3.11,
      "learning_rate": 4.4583987441130297e-05,
      "loss": 0.2188,
      "step": 3960
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.419152276295134e-05,
      "loss": 0.2422,
      "step": 3970
    },
    {
      "epoch": 3.12,
      "learning_rate": 4.379905808477237e-05,
      "loss": 0.1887,
      "step": 3980
    },
    {
      "epoch": 3.13,
      "learning_rate": 4.340659340659341e-05,
      "loss": 0.2668,
      "step": 3990
    },
    {
      "epoch": 3.14,
      "learning_rate": 4.301412872841444e-05,
      "loss": 0.2022,
      "step": 4000
    },
    {
      "epoch": 3.14,
      "eval_accuracy": 0.9118324607329843,
      "eval_loss": 0.3330060839653015,
      "eval_runtime": 105.6575,
      "eval_samples_per_second": 45.193,
      "eval_steps_per_second": 5.65,
      "step": 4000
    },
    {
      "epoch": 3.15,
      "learning_rate": 4.2621664050235485e-05,
      "loss": 0.4154,
      "step": 4010
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.2229199372056514e-05,
      "loss": 0.2389,
      "step": 4020
    },
    {
      "epoch": 3.16,
      "learning_rate": 4.183673469387756e-05,
      "loss": 0.2303,
      "step": 4030
    },
    {
      "epoch": 3.17,
      "learning_rate": 4.144427001569859e-05,
      "loss": 0.136,
      "step": 4040
    },
    {
      "epoch": 3.18,
      "learning_rate": 4.105180533751963e-05,
      "loss": 0.2214,
      "step": 4050
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.065934065934066e-05,
      "loss": 0.1767,
      "step": 4060
    },
    {
      "epoch": 3.19,
      "learning_rate": 4.0266875981161696e-05,
      "loss": 0.183,
      "step": 4070
    },
    {
      "epoch": 3.2,
      "learning_rate": 3.987441130298273e-05,
      "loss": 0.2733,
      "step": 4080
    },
    {
      "epoch": 3.21,
      "learning_rate": 3.948194662480377e-05,
      "loss": 0.1869,
      "step": 4090
    },
    {
      "epoch": 3.22,
      "learning_rate": 3.9089481946624805e-05,
      "loss": 0.2429,
      "step": 4100
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.869701726844584e-05,
      "loss": 0.166,
      "step": 4110
    },
    {
      "epoch": 3.23,
      "learning_rate": 3.830455259026688e-05,
      "loss": 0.2535,
      "step": 4120
    },
    {
      "epoch": 3.24,
      "learning_rate": 3.7912087912087914e-05,
      "loss": 0.1876,
      "step": 4130
    },
    {
      "epoch": 3.25,
      "learning_rate": 3.751962323390895e-05,
      "loss": 0.2684,
      "step": 4140
    },
    {
      "epoch": 3.26,
      "learning_rate": 3.7127158555729986e-05,
      "loss": 0.3673,
      "step": 4150
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.673469387755102e-05,
      "loss": 0.2569,
      "step": 4160
    },
    {
      "epoch": 3.27,
      "learning_rate": 3.634222919937206e-05,
      "loss": 0.1942,
      "step": 4170
    },
    {
      "epoch": 3.28,
      "learning_rate": 3.5949764521193095e-05,
      "loss": 0.2461,
      "step": 4180
    },
    {
      "epoch": 3.29,
      "learning_rate": 3.555729984301413e-05,
      "loss": 0.2283,
      "step": 4190
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.516483516483517e-05,
      "loss": 0.3326,
      "step": 4200
    },
    {
      "epoch": 3.3,
      "learning_rate": 3.4772370486656204e-05,
      "loss": 0.1666,
      "step": 4210
    },
    {
      "epoch": 3.31,
      "learning_rate": 3.4379905808477234e-05,
      "loss": 0.1527,
      "step": 4220
    },
    {
      "epoch": 3.32,
      "learning_rate": 3.398744113029828e-05,
      "loss": 0.1704,
      "step": 4230
    },
    {
      "epoch": 3.33,
      "learning_rate": 3.3594976452119306e-05,
      "loss": 0.1885,
      "step": 4240
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.320251177394035e-05,
      "loss": 0.3355,
      "step": 4250
    },
    {
      "epoch": 3.34,
      "learning_rate": 3.281004709576138e-05,
      "loss": 0.2072,
      "step": 4260
    },
    {
      "epoch": 3.35,
      "learning_rate": 3.241758241758242e-05,
      "loss": 0.1902,
      "step": 4270
    },
    {
      "epoch": 3.36,
      "learning_rate": 3.202511773940345e-05,
      "loss": 0.2432,
      "step": 4280
    },
    {
      "epoch": 3.37,
      "learning_rate": 3.1632653061224494e-05,
      "loss": 0.151,
      "step": 4290
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.1240188383045524e-05,
      "loss": 0.3722,
      "step": 4300
    },
    {
      "epoch": 3.38,
      "learning_rate": 3.084772370486657e-05,
      "loss": 0.1488,
      "step": 4310
    },
    {
      "epoch": 3.39,
      "learning_rate": 3.04552590266876e-05,
      "loss": 0.1683,
      "step": 4320
    },
    {
      "epoch": 3.4,
      "learning_rate": 3.006279434850864e-05,
      "loss": 0.1746,
      "step": 4330
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9670329670329673e-05,
      "loss": 0.2985,
      "step": 4340
    },
    {
      "epoch": 3.41,
      "learning_rate": 2.9277864992150706e-05,
      "loss": 0.1778,
      "step": 4350
    },
    {
      "epoch": 3.42,
      "learning_rate": 2.8885400313971745e-05,
      "loss": 0.0781,
      "step": 4360
    },
    {
      "epoch": 3.43,
      "learning_rate": 2.8492935635792778e-05,
      "loss": 0.2062,
      "step": 4370
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.8100470957613818e-05,
      "loss": 0.1906,
      "step": 4380
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.770800627943485e-05,
      "loss": 0.2594,
      "step": 4390
    },
    {
      "epoch": 3.45,
      "learning_rate": 2.731554160125589e-05,
      "loss": 0.2087,
      "step": 4400
    },
    {
      "epoch": 3.46,
      "learning_rate": 2.6923076923076923e-05,
      "loss": 0.3252,
      "step": 4410
    },
    {
      "epoch": 3.47,
      "learning_rate": 2.6530612244897963e-05,
      "loss": 0.1681,
      "step": 4420
    },
    {
      "epoch": 3.48,
      "learning_rate": 2.6138147566718996e-05,
      "loss": 0.267,
      "step": 4430
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.5745682888540036e-05,
      "loss": 0.2589,
      "step": 4440
    },
    {
      "epoch": 3.49,
      "learning_rate": 2.535321821036107e-05,
      "loss": 0.2525,
      "step": 4450
    },
    {
      "epoch": 3.5,
      "learning_rate": 2.4960753532182105e-05,
      "loss": 0.3417,
      "step": 4460
    },
    {
      "epoch": 3.51,
      "learning_rate": 2.456828885400314e-05,
      "loss": 0.2417,
      "step": 4470
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.4175824175824177e-05,
      "loss": 0.1734,
      "step": 4480
    },
    {
      "epoch": 3.52,
      "learning_rate": 2.3783359497645214e-05,
      "loss": 0.212,
      "step": 4490
    },
    {
      "epoch": 3.53,
      "learning_rate": 2.339089481946625e-05,
      "loss": 0.1332,
      "step": 4500
    },
    {
      "epoch": 3.53,
      "eval_accuracy": 0.9208376963350785,
      "eval_loss": 0.3042859733104706,
      "eval_runtime": 105.8953,
      "eval_samples_per_second": 45.092,
      "eval_steps_per_second": 5.638,
      "step": 4500
    },
    {
      "epoch": 3.54,
      "learning_rate": 2.2998430141287286e-05,
      "loss": 0.2482,
      "step": 4510
    },
    {
      "epoch": 3.55,
      "learning_rate": 2.2605965463108323e-05,
      "loss": 0.151,
      "step": 4520
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.221350078492936e-05,
      "loss": 0.1627,
      "step": 4530
    },
    {
      "epoch": 3.56,
      "learning_rate": 2.1821036106750395e-05,
      "loss": 0.3471,
      "step": 4540
    },
    {
      "epoch": 3.57,
      "learning_rate": 2.1428571428571428e-05,
      "loss": 0.162,
      "step": 4550
    },
    {
      "epoch": 3.58,
      "learning_rate": 2.1036106750392464e-05,
      "loss": 0.2045,
      "step": 4560
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.06436420722135e-05,
      "loss": 0.0903,
      "step": 4570
    },
    {
      "epoch": 3.59,
      "learning_rate": 2.0251177394034537e-05,
      "loss": 0.1779,
      "step": 4580
    },
    {
      "epoch": 3.6,
      "learning_rate": 1.9858712715855573e-05,
      "loss": 0.1739,
      "step": 4590
    },
    {
      "epoch": 3.61,
      "learning_rate": 1.946624803767661e-05,
      "loss": 0.1945,
      "step": 4600
    },
    {
      "epoch": 3.62,
      "learning_rate": 1.9073783359497646e-05,
      "loss": 0.37,
      "step": 4610
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.8681318681318682e-05,
      "loss": 0.2932,
      "step": 4620
    },
    {
      "epoch": 3.63,
      "learning_rate": 1.828885400313972e-05,
      "loss": 0.1827,
      "step": 4630
    },
    {
      "epoch": 3.64,
      "learning_rate": 1.7896389324960755e-05,
      "loss": 0.1647,
      "step": 4640
    },
    {
      "epoch": 3.65,
      "learning_rate": 1.750392464678179e-05,
      "loss": 0.1071,
      "step": 4650
    },
    {
      "epoch": 3.66,
      "learning_rate": 1.7111459968602827e-05,
      "loss": 0.1995,
      "step": 4660
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6718995290423864e-05,
      "loss": 0.2714,
      "step": 4670
    },
    {
      "epoch": 3.67,
      "learning_rate": 1.6326530612244897e-05,
      "loss": 0.1401,
      "step": 4680
    },
    {
      "epoch": 3.68,
      "learning_rate": 1.5934065934065933e-05,
      "loss": 0.1516,
      "step": 4690
    },
    {
      "epoch": 3.69,
      "learning_rate": 1.554160125588697e-05,
      "loss": 0.262,
      "step": 4700
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.5149136577708006e-05,
      "loss": 0.1809,
      "step": 4710
    },
    {
      "epoch": 3.7,
      "learning_rate": 1.4756671899529042e-05,
      "loss": 0.2473,
      "step": 4720
    },
    {
      "epoch": 3.71,
      "learning_rate": 1.4364207221350078e-05,
      "loss": 0.074,
      "step": 4730
    },
    {
      "epoch": 3.72,
      "learning_rate": 1.3971742543171114e-05,
      "loss": 0.1139,
      "step": 4740
    },
    {
      "epoch": 3.73,
      "learning_rate": 1.357927786499215e-05,
      "loss": 0.2363,
      "step": 4750
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.3186813186813187e-05,
      "loss": 0.1619,
      "step": 4760
    },
    {
      "epoch": 3.74,
      "learning_rate": 1.2794348508634223e-05,
      "loss": 0.0841,
      "step": 4770
    },
    {
      "epoch": 3.75,
      "learning_rate": 1.240188383045526e-05,
      "loss": 0.2056,
      "step": 4780
    },
    {
      "epoch": 3.76,
      "learning_rate": 1.2009419152276296e-05,
      "loss": 0.2504,
      "step": 4790
    },
    {
      "epoch": 3.77,
      "learning_rate": 1.1616954474097332e-05,
      "loss": 0.2733,
      "step": 4800
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.1224489795918369e-05,
      "loss": 0.2208,
      "step": 4810
    },
    {
      "epoch": 3.78,
      "learning_rate": 1.0832025117739405e-05,
      "loss": 0.1542,
      "step": 4820
    },
    {
      "epoch": 3.79,
      "learning_rate": 1.0439560439560441e-05,
      "loss": 0.3311,
      "step": 4830
    },
    {
      "epoch": 3.8,
      "learning_rate": 1.0047095761381477e-05,
      "loss": 0.1597,
      "step": 4840
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.654631083202512e-06,
      "loss": 0.1415,
      "step": 4850
    },
    {
      "epoch": 3.81,
      "learning_rate": 9.262166405023548e-06,
      "loss": 0.229,
      "step": 4860
    },
    {
      "epoch": 3.82,
      "learning_rate": 8.869701726844585e-06,
      "loss": 0.1686,
      "step": 4870
    },
    {
      "epoch": 3.83,
      "learning_rate": 8.477237048665621e-06,
      "loss": 0.2163,
      "step": 4880
    },
    {
      "epoch": 3.84,
      "learning_rate": 8.084772370486657e-06,
      "loss": 0.1421,
      "step": 4890
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.692307692307694e-06,
      "loss": 0.2704,
      "step": 4900
    },
    {
      "epoch": 3.85,
      "learning_rate": 7.299843014128729e-06,
      "loss": 0.0956,
      "step": 4910
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.9073783359497645e-06,
      "loss": 0.1536,
      "step": 4920
    },
    {
      "epoch": 3.87,
      "learning_rate": 6.514913657770801e-06,
      "loss": 0.2418,
      "step": 4930
    },
    {
      "epoch": 3.88,
      "learning_rate": 6.122448979591837e-06,
      "loss": 0.3416,
      "step": 4940
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.729984301412873e-06,
      "loss": 0.2246,
      "step": 4950
    },
    {
      "epoch": 3.89,
      "learning_rate": 5.33751962323391e-06,
      "loss": 0.2371,
      "step": 4960
    },
    {
      "epoch": 3.9,
      "learning_rate": 4.945054945054945e-06,
      "loss": 0.1669,
      "step": 4970
    },
    {
      "epoch": 3.91,
      "learning_rate": 4.5525902668759815e-06,
      "loss": 0.2007,
      "step": 4980
    },
    {
      "epoch": 3.92,
      "learning_rate": 4.160125588697018e-06,
      "loss": 0.1772,
      "step": 4990
    },
    {
      "epoch": 3.92,
      "learning_rate": 3.767660910518053e-06,
      "loss": 0.1821,
      "step": 5000
    },
    {
      "epoch": 3.92,
      "eval_accuracy": 0.9281675392670157,
      "eval_loss": 0.2815741002559662,
      "eval_runtime": 105.8792,
      "eval_samples_per_second": 45.099,
      "eval_steps_per_second": 5.638,
      "step": 5000
    },
    {
      "epoch": 3.93,
      "learning_rate": 3.3751962323390895e-06,
      "loss": 0.2243,
      "step": 5010
    },
    {
      "epoch": 3.94,
      "learning_rate": 2.982731554160126e-06,
      "loss": 0.1931,
      "step": 5020
    },
    {
      "epoch": 3.95,
      "learning_rate": 2.5902668759811617e-06,
      "loss": 0.136,
      "step": 5030
    },
    {
      "epoch": 3.96,
      "learning_rate": 2.197802197802198e-06,
      "loss": 0.3308,
      "step": 5040
    },
    {
      "epoch": 3.96,
      "learning_rate": 1.805337519623234e-06,
      "loss": 0.1563,
      "step": 5050
    },
    {
      "epoch": 3.97,
      "learning_rate": 1.4128728414442702e-06,
      "loss": 0.1906,
      "step": 5060
    },
    {
      "epoch": 3.98,
      "learning_rate": 1.020408163265306e-06,
      "loss": 0.2072,
      "step": 5070
    },
    {
      "epoch": 3.99,
      "learning_rate": 6.279434850863423e-07,
      "loss": 0.2218,
      "step": 5080
    },
    {
      "epoch": 4.0,
      "learning_rate": 2.3547880690737833e-07,
      "loss": 0.2175,
      "step": 5090
    },
    {
      "epoch": 4.0,
      "step": 5096,
      "total_flos": 5.921465894918849e+18,
      "train_loss": 0.46419070048178757,
      "train_runtime": 5529.79,
      "train_samples_per_second": 13.815,
      "train_steps_per_second": 0.922
    }
  ],
  "logging_steps": 10,
  "max_steps": 5096,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 5.921465894918849e+18,
  "trial_name": null,
  "trial_params": null
}