{
  "best_metric": 0.9479,
  "best_model_checkpoint": "facebook/convnextv2-base-22k-224-finetuned-critique-100k/checkpoint-2000",
  "epoch": 3.0572342694632066,
  "global_step": 2150,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 1.4204545454545456e-06,
      "loss": 0.7049,
      "step": 10
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.840909090909091e-06,
      "loss": 0.692,
      "step": 20
    },
    {
      "epoch": 0.04,
      "learning_rate": 4.2613636363636365e-06,
      "loss": 0.6786,
      "step": 30
    },
    {
      "epoch": 0.06,
      "learning_rate": 5.681818181818182e-06,
      "loss": 0.6466,
      "step": 40
    },
    {
      "epoch": 0.07,
      "learning_rate": 7.102272727272728e-06,
      "loss": 0.6277,
      "step": 50
    },
    {
      "epoch": 0.07,
      "eval_accuracy": 0.6767,
      "eval_loss": 0.5986741185188293,
      "eval_runtime": 70.2999,
      "eval_samples_per_second": 142.248,
      "eval_steps_per_second": 4.452,
      "step": 50
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.522727272727273e-06,
      "loss": 0.6005,
      "step": 60
    },
    {
      "epoch": 0.1,
      "learning_rate": 9.943181818181819e-06,
      "loss": 0.578,
      "step": 70
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.1363636363636365e-05,
      "loss": 0.5766,
      "step": 80
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.2784090909090909e-05,
      "loss": 0.5208,
      "step": 90
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.4204545454545456e-05,
      "loss": 0.5459,
      "step": 100
    },
    {
      "epoch": 0.14,
      "eval_accuracy": 0.7401,
      "eval_loss": 0.5187093615531921,
      "eval_runtime": 70.0121,
      "eval_samples_per_second": 142.832,
      "eval_steps_per_second": 4.471,
      "step": 100
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.5625e-05,
      "loss": 0.5079,
      "step": 110
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.7045454545454546e-05,
      "loss": 0.5067,
      "step": 120
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.8465909090909092e-05,
      "loss": 0.4788,
      "step": 130
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.9886363636363638e-05,
      "loss": 0.4715,
      "step": 140
    },
    {
      "epoch": 0.21,
      "learning_rate": 2.1306818181818183e-05,
      "loss": 0.4397,
      "step": 150
    },
    {
      "epoch": 0.21,
      "eval_accuracy": 0.7768,
      "eval_loss": 0.444758802652359,
      "eval_runtime": 70.1963,
      "eval_samples_per_second": 142.458,
      "eval_steps_per_second": 4.459,
      "step": 150
    },
    {
      "epoch": 0.23,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.4371,
      "step": 160
    },
    {
      "epoch": 0.24,
      "learning_rate": 2.4147727272727275e-05,
      "loss": 0.4659,
      "step": 170
    },
    {
      "epoch": 0.26,
      "learning_rate": 2.5568181818181817e-05,
      "loss": 0.4076,
      "step": 180
    },
    {
      "epoch": 0.27,
      "learning_rate": 2.6988636363636367e-05,
      "loss": 0.3909,
      "step": 190
    },
    {
      "epoch": 0.28,
      "learning_rate": 2.8409090909090912e-05,
      "loss": 0.4197,
      "step": 200
    },
    {
      "epoch": 0.28,
      "eval_accuracy": 0.8401,
      "eval_loss": 0.36862310767173767,
      "eval_runtime": 70.2245,
      "eval_samples_per_second": 142.4,
      "eval_steps_per_second": 4.457,
      "step": 200
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.9829545454545455e-05,
      "loss": 0.3837,
      "step": 210
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.125e-05,
      "loss": 0.3702,
      "step": 220
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.267045454545455e-05,
      "loss": 0.389,
      "step": 230
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.409090909090909e-05,
      "loss": 0.3463,
      "step": 240
    },
    {
      "epoch": 0.36,
      "learning_rate": 3.5511363636363635e-05,
      "loss": 0.3397,
      "step": 250
    },
    {
      "epoch": 0.36,
      "eval_accuracy": 0.8664,
      "eval_loss": 0.31527072191238403,
      "eval_runtime": 70.1314,
      "eval_samples_per_second": 142.589,
      "eval_steps_per_second": 4.463,
      "step": 250
    },
    {
      "epoch": 0.37,
      "learning_rate": 3.6931818181818184e-05,
      "loss": 0.3365,
      "step": 260
    },
    {
      "epoch": 0.38,
      "learning_rate": 3.835227272727273e-05,
      "loss": 0.3594,
      "step": 270
    },
    {
      "epoch": 0.4,
      "learning_rate": 3.9772727272727275e-05,
      "loss": 0.3503,
      "step": 280
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.119318181818182e-05,
      "loss": 0.3577,
      "step": 290
    },
    {
      "epoch": 0.43,
      "learning_rate": 4.261363636363637e-05,
      "loss": 0.3345,
      "step": 300
    },
    {
      "epoch": 0.43,
      "eval_accuracy": 0.8701,
      "eval_loss": 0.30711129307746887,
      "eval_runtime": 70.0389,
      "eval_samples_per_second": 142.778,
      "eval_steps_per_second": 4.469,
      "step": 300
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.4034090909090916e-05,
      "loss": 0.3235,
      "step": 310
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.545454545454546e-05,
      "loss": 0.3294,
      "step": 320
    },
    {
      "epoch": 0.47,
      "learning_rate": 4.6875e-05,
      "loss": 0.3156,
      "step": 330
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.829545454545455e-05,
      "loss": 0.3153,
      "step": 340
    },
    {
      "epoch": 0.5,
      "learning_rate": 4.971590909090909e-05,
      "loss": 0.3177,
      "step": 350
    },
    {
      "epoch": 0.5,
      "eval_accuracy": 0.8938,
      "eval_loss": 0.2576267719268799,
      "eval_runtime": 70.2029,
      "eval_samples_per_second": 142.444,
      "eval_steps_per_second": 4.459,
      "step": 350
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.9873537780588055e-05,
      "loss": 0.3315,
      "step": 360
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.9715460006323114e-05,
      "loss": 0.325,
      "step": 370
    },
    {
      "epoch": 0.54,
      "learning_rate": 4.955738223205817e-05,
      "loss": 0.3342,
      "step": 380
    },
    {
      "epoch": 0.55,
      "learning_rate": 4.939930445779324e-05,
      "loss": 0.3282,
      "step": 390
    },
    {
      "epoch": 0.57,
      "learning_rate": 4.92412266835283e-05,
      "loss": 0.3182,
      "step": 400
    },
    {
      "epoch": 0.57,
      "eval_accuracy": 0.8926,
      "eval_loss": 0.2545712888240814,
      "eval_runtime": 70.0504,
      "eval_samples_per_second": 142.754,
      "eval_steps_per_second": 4.468,
      "step": 400
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.908314890926336e-05,
      "loss": 0.2951,
      "step": 410
    },
    {
      "epoch": 0.6,
      "learning_rate": 4.892507113499842e-05,
      "loss": 0.2918,
      "step": 420
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.876699336073348e-05,
      "loss": 0.3223,
      "step": 430
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.8608915586468546e-05,
      "loss": 0.3029,
      "step": 440
    },
    {
      "epoch": 0.64,
      "learning_rate": 4.845083781220361e-05,
      "loss": 0.2596,
      "step": 450
    },
    {
      "epoch": 0.64,
      "eval_accuracy": 0.9004,
      "eval_loss": 0.23197658360004425,
      "eval_runtime": 70.1495,
      "eval_samples_per_second": 142.553,
      "eval_steps_per_second": 4.462,
      "step": 450
    },
    {
      "epoch": 0.65,
      "learning_rate": 4.829276003793867e-05,
      "loss": 0.2766,
      "step": 460
    },
    {
      "epoch": 0.67,
      "learning_rate": 4.813468226367373e-05,
      "loss": 0.2983,
      "step": 470
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.7976604489408795e-05,
      "loss": 0.2607,
      "step": 480
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.7818526715143853e-05,
      "loss": 0.2698,
      "step": 490
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.766044894087892e-05,
      "loss": 0.2563,
      "step": 500
    },
    {
      "epoch": 0.71,
      "eval_accuracy": 0.9082,
      "eval_loss": 0.22052288055419922,
      "eval_runtime": 70.0882,
      "eval_samples_per_second": 142.677,
      "eval_steps_per_second": 4.466,
      "step": 500
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.750237116661397e-05,
      "loss": 0.2989,
      "step": 510
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.734429339234904e-05,
      "loss": 0.2548,
      "step": 520
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.71862156180841e-05,
      "loss": 0.2855,
      "step": 530
    },
    {
      "epoch": 0.77,
      "learning_rate": 4.702813784381916e-05,
      "loss": 0.2886,
      "step": 540
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.687006006955422e-05,
      "loss": 0.2543,
      "step": 550
    },
    {
      "epoch": 0.78,
      "eval_accuracy": 0.9147,
      "eval_loss": 0.2142086923122406,
      "eval_runtime": 70.0483,
      "eval_samples_per_second": 142.759,
      "eval_steps_per_second": 4.468,
      "step": 550
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.6711982295289285e-05,
      "loss": 0.2673,
      "step": 560
    },
    {
      "epoch": 0.81,
      "learning_rate": 4.6553904521024344e-05,
      "loss": 0.2962,
      "step": 570
    },
    {
      "epoch": 0.82,
      "learning_rate": 4.639582674675941e-05,
      "loss": 0.2509,
      "step": 580
    },
    {
      "epoch": 0.84,
      "learning_rate": 4.623774897249447e-05,
      "loss": 0.2532,
      "step": 590
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.607967119822953e-05,
      "loss": 0.2768,
      "step": 600
    },
    {
      "epoch": 0.85,
      "eval_accuracy": 0.9132,
      "eval_loss": 0.21360518038272858,
      "eval_runtime": 70.2799,
      "eval_samples_per_second": 142.288,
      "eval_steps_per_second": 4.454,
      "step": 600
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.592159342396459e-05,
      "loss": 0.3061,
      "step": 610
    },
    {
      "epoch": 0.88,
      "learning_rate": 4.576351564969966e-05,
      "loss": 0.2522,
      "step": 620
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.560543787543472e-05,
      "loss": 0.2834,
      "step": 630
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.5447360101169776e-05,
      "loss": 0.2256,
      "step": 640
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.528928232690484e-05,
      "loss": 0.2486,
      "step": 650
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.9175,
      "eval_loss": 0.20521819591522217,
      "eval_runtime": 70.0587,
      "eval_samples_per_second": 142.738,
      "eval_steps_per_second": 4.468,
      "step": 650
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.51312045526399e-05,
      "loss": 0.2392,
      "step": 660
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.4973126778374966e-05,
      "loss": 0.2454,
      "step": 670
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.4815049004110025e-05,
      "loss": 0.2272,
      "step": 680
    },
    {
      "epoch": 0.98,
      "learning_rate": 4.4656971229845084e-05,
      "loss": 0.2668,
      "step": 690
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.449889345558015e-05,
      "loss": 0.2504,
      "step": 700
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9058,
      "eval_loss": 0.23142580687999725,
      "eval_runtime": 70.0882,
      "eval_samples_per_second": 142.677,
      "eval_steps_per_second": 4.466,
      "step": 700
    },
    {
      "epoch": 1.01,
      "learning_rate": 4.434081568131521e-05,
      "loss": 0.222,
      "step": 710
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.418273790705027e-05,
      "loss": 0.2342,
      "step": 720
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.402466013278533e-05,
      "loss": 0.228,
      "step": 730
    },
    {
      "epoch": 1.05,
      "learning_rate": 4.386658235852039e-05,
      "loss": 0.256,
      "step": 740
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.370850458425546e-05,
      "loss": 0.2437,
      "step": 750
    },
    {
      "epoch": 1.07,
      "eval_accuracy": 0.9235,
      "eval_loss": 0.1942846179008484,
      "eval_runtime": 70.1793,
      "eval_samples_per_second": 142.492,
      "eval_steps_per_second": 4.46,
      "step": 750
    },
    {
      "epoch": 1.08,
      "learning_rate": 4.3550426809990516e-05,
      "loss": 0.208,
      "step": 760
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.3392349035725575e-05,
      "loss": 0.2361,
      "step": 770
    },
    {
      "epoch": 1.11,
      "learning_rate": 4.323427126146064e-05,
      "loss": 0.204,
      "step": 780
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.3076193487195706e-05,
      "loss": 0.1855,
      "step": 790
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.2918115712930765e-05,
      "loss": 0.212,
      "step": 800
    },
    {
      "epoch": 1.14,
      "eval_accuracy": 0.9183,
      "eval_loss": 0.20191900432109833,
      "eval_runtime": 70.1451,
      "eval_samples_per_second": 142.562,
      "eval_steps_per_second": 4.462,
      "step": 800
    },
    {
      "epoch": 1.15,
      "learning_rate": 4.2760037938665824e-05,
      "loss": 0.1913,
      "step": 810
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.260196016440089e-05,
      "loss": 0.2169,
      "step": 820
    },
    {
      "epoch": 1.18,
      "learning_rate": 4.244388239013595e-05,
      "loss": 0.196,
      "step": 830
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.2285804615871014e-05,
      "loss": 0.2144,
      "step": 840
    },
    {
      "epoch": 1.21,
      "learning_rate": 4.212772684160607e-05,
      "loss": 0.1891,
      "step": 850
    },
    {
      "epoch": 1.21,
      "eval_accuracy": 0.9254,
      "eval_loss": 0.1845281422138214,
      "eval_runtime": 70.1164,
      "eval_samples_per_second": 142.62,
      "eval_steps_per_second": 4.464,
      "step": 850
    },
    {
      "epoch": 1.22,
      "learning_rate": 4.196964906734113e-05,
      "loss": 0.2297,
      "step": 860
    },
    {
      "epoch": 1.24,
      "learning_rate": 4.18115712930762e-05,
      "loss": 0.2199,
      "step": 870
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.165349351881126e-05,
      "loss": 0.2395,
      "step": 880
    },
    {
      "epoch": 1.27,
      "learning_rate": 4.1495415744546315e-05,
      "loss": 0.2128,
      "step": 890
    },
    {
      "epoch": 1.28,
      "learning_rate": 4.133733797028138e-05,
      "loss": 0.2105,
      "step": 900
    },
    {
      "epoch": 1.28,
      "eval_accuracy": 0.9288,
      "eval_loss": 0.18339484930038452,
      "eval_runtime": 70.2884,
      "eval_samples_per_second": 142.271,
      "eval_steps_per_second": 4.453,
      "step": 900
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.1179260196016446e-05,
      "loss": 0.2037,
      "step": 910
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.1021182421751505e-05,
      "loss": 0.2124,
      "step": 920
    },
    {
      "epoch": 1.32,
      "learning_rate": 4.0863104647486563e-05,
      "loss": 0.228,
      "step": 930
    },
    {
      "epoch": 1.34,
      "learning_rate": 4.070502687322162e-05,
      "loss": 0.1866,
      "step": 940
    },
    {
      "epoch": 1.35,
      "learning_rate": 4.054694909895669e-05,
      "loss": 0.2285,
      "step": 950
    },
    {
      "epoch": 1.35,
      "eval_accuracy": 0.9206,
      "eval_loss": 0.19937752187252045,
      "eval_runtime": 70.1468,
      "eval_samples_per_second": 142.558,
      "eval_steps_per_second": 4.462,
      "step": 950
    },
    {
      "epoch": 1.37,
      "learning_rate": 4.0388871324691753e-05,
      "loss": 0.2086,
      "step": 960
    },
    {
      "epoch": 1.38,
      "learning_rate": 4.023079355042681e-05,
      "loss": 0.216,
      "step": 970
    },
    {
      "epoch": 1.39,
      "learning_rate": 4.007271577616187e-05,
      "loss": 0.1899,
      "step": 980
    },
    {
      "epoch": 1.41,
      "learning_rate": 3.991463800189694e-05,
      "loss": 0.1884,
      "step": 990
    },
    {
      "epoch": 1.42,
      "learning_rate": 3.9756560227631995e-05,
      "loss": 0.2214,
      "step": 1000
    },
    {
      "epoch": 1.42,
      "eval_accuracy": 0.9251,
      "eval_loss": 0.18035535514354706,
      "eval_runtime": 70.1953,
      "eval_samples_per_second": 142.46,
      "eval_steps_per_second": 4.459,
      "step": 1000
    },
    {
      "epoch": 1.44,
      "learning_rate": 3.959848245336706e-05,
      "loss": 0.2057,
      "step": 1010
    },
    {
      "epoch": 1.45,
      "learning_rate": 3.944040467910212e-05,
      "loss": 0.2064,
      "step": 1020
    },
    {
      "epoch": 1.46,
      "learning_rate": 3.928232690483718e-05,
      "loss": 0.181,
      "step": 1030
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.9124249130572244e-05,
      "loss": 0.2079,
      "step": 1040
    },
    {
      "epoch": 1.49,
      "learning_rate": 3.896617135630731e-05,
      "loss": 0.1848,
      "step": 1050
    },
    {
      "epoch": 1.49,
      "eval_accuracy": 0.9196,
      "eval_loss": 0.19748303294181824,
      "eval_runtime": 70.0931,
      "eval_samples_per_second": 142.667,
      "eval_steps_per_second": 4.465,
      "step": 1050
    },
    {
      "epoch": 1.51,
      "learning_rate": 3.880809358204236e-05,
      "loss": 0.2065,
      "step": 1060
    },
    {
      "epoch": 1.52,
      "learning_rate": 3.865001580777743e-05,
      "loss": 0.1861,
      "step": 1070
    },
    {
      "epoch": 1.54,
      "learning_rate": 3.849193803351249e-05,
      "loss": 0.2154,
      "step": 1080
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.833386025924755e-05,
      "loss": 0.1989,
      "step": 1090
    },
    {
      "epoch": 1.56,
      "learning_rate": 3.817578248498261e-05,
      "loss": 0.191,
      "step": 1100
    },
    {
      "epoch": 1.56,
      "eval_accuracy": 0.9269,
      "eval_loss": 0.17946849763393402,
      "eval_runtime": 70.2956,
      "eval_samples_per_second": 142.256,
      "eval_steps_per_second": 4.453,
      "step": 1100
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.8017704710717676e-05,
      "loss": 0.1992,
      "step": 1110
    },
    {
      "epoch": 1.59,
      "learning_rate": 3.7859626936452735e-05,
      "loss": 0.2095,
      "step": 1120
    },
    {
      "epoch": 1.61,
      "learning_rate": 3.77015491621878e-05,
      "loss": 0.1967,
      "step": 1130
    },
    {
      "epoch": 1.62,
      "learning_rate": 3.754347138792286e-05,
      "loss": 0.1895,
      "step": 1140
    },
    {
      "epoch": 1.64,
      "learning_rate": 3.738539361365792e-05,
      "loss": 0.1794,
      "step": 1150
    },
    {
      "epoch": 1.64,
      "eval_accuracy": 0.9358,
      "eval_loss": 0.1606118232011795,
      "eval_runtime": 70.1775,
      "eval_samples_per_second": 142.496,
      "eval_steps_per_second": 4.46,
      "step": 1150
    },
    {
      "epoch": 1.65,
      "learning_rate": 3.7227315839392984e-05,
      "loss": 0.1727,
      "step": 1160
    },
    {
      "epoch": 1.66,
      "learning_rate": 3.706923806512804e-05,
      "loss": 0.2358,
      "step": 1170
    },
    {
      "epoch": 1.68,
      "learning_rate": 3.691116029086311e-05,
      "loss": 0.1766,
      "step": 1180
    },
    {
      "epoch": 1.69,
      "learning_rate": 3.675308251659817e-05,
      "loss": 0.2174,
      "step": 1190
    },
    {
      "epoch": 1.71,
      "learning_rate": 3.6595004742333226e-05,
      "loss": 0.2084,
      "step": 1200
    },
    {
      "epoch": 1.71,
      "eval_accuracy": 0.9293,
      "eval_loss": 0.18068891763687134,
      "eval_runtime": 70.0834,
      "eval_samples_per_second": 142.687,
      "eval_steps_per_second": 4.466,
      "step": 1200
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.643692696806829e-05,
      "loss": 0.1727,
      "step": 1210
    },
    {
      "epoch": 1.73,
      "learning_rate": 3.627884919380336e-05,
      "loss": 0.1725,
      "step": 1220
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.6120771419538416e-05,
      "loss": 0.1761,
      "step": 1230
    },
    {
      "epoch": 1.76,
      "learning_rate": 3.5962693645273475e-05,
      "loss": 0.1951,
      "step": 1240
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.580461587100854e-05,
      "loss": 0.199,
      "step": 1250
    },
    {
      "epoch": 1.78,
      "eval_accuracy": 0.9307,
      "eval_loss": 0.1697014719247818,
      "eval_runtime": 70.0908,
      "eval_samples_per_second": 142.672,
      "eval_steps_per_second": 4.466,
      "step": 1250
    },
    {
      "epoch": 1.79,
      "learning_rate": 3.56465380967436e-05,
      "loss": 0.1975,
      "step": 1260
    },
    {
      "epoch": 1.81,
      "learning_rate": 3.5488460322478665e-05,
      "loss": 0.1558,
      "step": 1270
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.5330382548213724e-05,
      "loss": 0.1607,
      "step": 1280
    },
    {
      "epoch": 1.83,
      "learning_rate": 3.517230477394878e-05,
      "loss": 0.1713,
      "step": 1290
    },
    {
      "epoch": 1.85,
      "learning_rate": 3.501422699968385e-05,
      "loss": 0.1874,
      "step": 1300
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.9372,
      "eval_loss": 0.16498488187789917,
      "eval_runtime": 70.1699,
      "eval_samples_per_second": 142.511,
      "eval_steps_per_second": 4.461,
      "step": 1300
    },
    {
      "epoch": 1.86,
      "learning_rate": 3.4856149225418914e-05,
      "loss": 0.1919,
      "step": 1310
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.4698071451153966e-05,
      "loss": 0.1874,
      "step": 1320
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.453999367688903e-05,
      "loss": 0.2021,
      "step": 1330
    },
    {
      "epoch": 1.91,
      "learning_rate": 3.438191590262409e-05,
      "loss": 0.1806,
      "step": 1340
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.4223838128359156e-05,
      "loss": 0.1681,
      "step": 1350
    },
    {
      "epoch": 1.92,
      "eval_accuracy": 0.939,
      "eval_loss": 0.15153267979621887,
      "eval_runtime": 70.086,
      "eval_samples_per_second": 142.682,
      "eval_steps_per_second": 4.466,
      "step": 1350
    },
    {
      "epoch": 1.93,
      "learning_rate": 3.4065760354094215e-05,
      "loss": 0.193,
      "step": 1360
    },
    {
      "epoch": 1.95,
      "learning_rate": 3.3907682579829273e-05,
      "loss": 0.1656,
      "step": 1370
    },
    {
      "epoch": 1.96,
      "learning_rate": 3.374960480556434e-05,
      "loss": 0.196,
      "step": 1380
    },
    {
      "epoch": 1.98,
      "learning_rate": 3.3591527031299405e-05,
      "loss": 0.1684,
      "step": 1390
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.3433449257034463e-05,
      "loss": 0.1696,
      "step": 1400
    },
    {
      "epoch": 1.99,
      "eval_accuracy": 0.9416,
      "eval_loss": 0.14727267622947693,
      "eval_runtime": 70.1305,
      "eval_samples_per_second": 142.591,
      "eval_steps_per_second": 4.463,
      "step": 1400
    },
    {
      "epoch": 2.0,
      "learning_rate": 3.327537148276952e-05,
      "loss": 0.1845,
      "step": 1410
    },
    {
      "epoch": 2.02,
      "learning_rate": 3.311729370850459e-05,
      "loss": 0.164,
      "step": 1420
    },
    {
      "epoch": 2.03,
      "learning_rate": 3.295921593423965e-05,
      "loss": 0.1623,
      "step": 1430
    },
    {
      "epoch": 2.05,
      "learning_rate": 3.280113815997471e-05,
      "loss": 0.1583,
      "step": 1440
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.264306038570977e-05,
      "loss": 0.1651,
      "step": 1450
    },
    {
      "epoch": 2.06,
      "eval_accuracy": 0.9428,
      "eval_loss": 0.14893431961536407,
      "eval_runtime": 70.1871,
      "eval_samples_per_second": 142.476,
      "eval_steps_per_second": 4.46,
      "step": 1450
    },
    {
      "epoch": 2.08,
      "learning_rate": 3.248498261144483e-05,
      "loss": 0.1367,
      "step": 1460
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.2326904837179896e-05,
      "loss": 0.1459,
      "step": 1470
    },
    {
      "epoch": 2.1,
      "learning_rate": 3.216882706291496e-05,
      "loss": 0.1502,
      "step": 1480
    },
    {
      "epoch": 2.12,
      "learning_rate": 3.201074928865001e-05,
      "loss": 0.1622,
      "step": 1490
    },
    {
      "epoch": 2.13,
      "learning_rate": 3.185267151438508e-05,
      "loss": 0.1627,
      "step": 1500
    },
    {
      "epoch": 2.13,
      "eval_accuracy": 0.9395,
      "eval_loss": 0.15285363793373108,
      "eval_runtime": 70.0671,
      "eval_samples_per_second": 142.72,
      "eval_steps_per_second": 4.467,
      "step": 1500
    },
    {
      "epoch": 2.15,
      "learning_rate": 3.1694593740120144e-05,
      "loss": 0.1668,
      "step": 1510
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.15365159658552e-05,
      "loss": 0.1609,
      "step": 1520
    },
    {
      "epoch": 2.18,
      "learning_rate": 3.137843819159026e-05,
      "loss": 0.1592,
      "step": 1530
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.122036041732532e-05,
      "loss": 0.1669,
      "step": 1540
    },
    {
      "epoch": 2.2,
      "learning_rate": 3.1062282643060386e-05,
      "loss": 0.1754,
      "step": 1550
    },
    {
      "epoch": 2.2,
      "eval_accuracy": 0.9379,
      "eval_loss": 0.15395458042621613,
      "eval_runtime": 70.1364,
      "eval_samples_per_second": 142.579,
      "eval_steps_per_second": 4.463,
      "step": 1550
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.090420486879545e-05,
      "loss": 0.1536,
      "step": 1560
    },
    {
      "epoch": 2.23,
      "learning_rate": 3.074612709453051e-05,
      "loss": 0.1497,
      "step": 1570
    },
    {
      "epoch": 2.25,
      "learning_rate": 3.058804932026557e-05,
      "loss": 0.1735,
      "step": 1580
    },
    {
      "epoch": 2.26,
      "learning_rate": 3.0429971546000635e-05,
      "loss": 0.1536,
      "step": 1590
    },
    {
      "epoch": 2.28,
      "learning_rate": 3.0271893771735694e-05,
      "loss": 0.1302,
      "step": 1600
    },
    {
      "epoch": 2.28,
      "eval_accuracy": 0.939,
      "eval_loss": 0.1578802913427353,
      "eval_runtime": 70.1747,
      "eval_samples_per_second": 142.501,
      "eval_steps_per_second": 4.46,
      "step": 1600
    },
    {
      "epoch": 2.29,
      "learning_rate": 3.0113815997470756e-05,
      "loss": 0.1627,
      "step": 1610
    },
    {
      "epoch": 2.3,
      "learning_rate": 2.9955738223205822e-05,
      "loss": 0.1252,
      "step": 1620
    },
    {
      "epoch": 2.32,
      "learning_rate": 2.9797660448940877e-05,
      "loss": 0.1447,
      "step": 1630
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.9639582674675943e-05,
      "loss": 0.1531,
      "step": 1640
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.9481504900411005e-05,
      "loss": 0.1643,
      "step": 1650
    },
    {
      "epoch": 2.35,
      "eval_accuracy": 0.9401,
      "eval_loss": 0.15176580846309662,
      "eval_runtime": 70.1732,
      "eval_samples_per_second": 142.505,
      "eval_steps_per_second": 4.46,
      "step": 1650
    },
    {
      "epoch": 2.36,
      "learning_rate": 2.9323427126146064e-05,
      "loss": 0.1638,
      "step": 1660
    },
    {
      "epoch": 2.37,
      "learning_rate": 2.9165349351881126e-05,
      "loss": 0.1409,
      "step": 1670
    },
    {
      "epoch": 2.39,
      "learning_rate": 2.900727157761619e-05,
      "loss": 0.1415,
      "step": 1680
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.8849193803351247e-05,
      "loss": 0.1651,
      "step": 1690
    },
    {
      "epoch": 2.42,
      "learning_rate": 2.8691116029086313e-05,
      "loss": 0.1938,
      "step": 1700
    },
    {
      "epoch": 2.42,
      "eval_accuracy": 0.941,
      "eval_loss": 0.14790111780166626,
      "eval_runtime": 70.2486,
      "eval_samples_per_second": 142.352,
      "eval_steps_per_second": 4.456,
      "step": 1700
    },
    {
      "epoch": 2.43,
      "learning_rate": 2.8533038254821375e-05,
      "loss": 0.1597,
      "step": 1710
    },
    {
      "epoch": 2.45,
      "learning_rate": 2.8374960480556434e-05,
      "loss": 0.1395,
      "step": 1720
    },
    {
      "epoch": 2.46,
      "learning_rate": 2.8216882706291496e-05,
      "loss": 0.1628,
      "step": 1730
    },
    {
      "epoch": 2.47,
      "learning_rate": 2.805880493202656e-05,
      "loss": 0.1438,
      "step": 1740
    },
    {
      "epoch": 2.49,
      "learning_rate": 2.790072715776162e-05,
      "loss": 0.1441,
      "step": 1750
    },
    {
      "epoch": 2.49,
      "eval_accuracy": 0.9436,
      "eval_loss": 0.1451416313648224,
      "eval_runtime": 70.1105,
      "eval_samples_per_second": 142.632,
      "eval_steps_per_second": 4.464,
      "step": 1750
    },
    {
      "epoch": 2.5,
      "learning_rate": 2.7742649383496683e-05,
      "loss": 0.1703,
      "step": 1760
    },
    {
      "epoch": 2.52,
      "learning_rate": 2.758457160923174e-05,
      "loss": 0.1403,
      "step": 1770
    },
    {
      "epoch": 2.53,
      "learning_rate": 2.7426493834966804e-05,
      "loss": 0.1687,
      "step": 1780
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.726841606070187e-05,
      "loss": 0.1422,
      "step": 1790
    },
    {
      "epoch": 2.56,
      "learning_rate": 2.7110338286436925e-05,
      "loss": 0.1478,
      "step": 1800
    },
    {
      "epoch": 2.56,
      "eval_accuracy": 0.9472,
      "eval_loss": 0.1323779672384262,
      "eval_runtime": 70.0872,
      "eval_samples_per_second": 142.679,
      "eval_steps_per_second": 4.466,
      "step": 1800
    },
    {
      "epoch": 2.57,
      "learning_rate": 2.695226051217199e-05,
      "loss": 0.1602,
      "step": 1810
    },
    {
      "epoch": 2.59,
      "learning_rate": 2.6794182737907052e-05,
      "loss": 0.1401,
      "step": 1820
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.663610496364211e-05,
      "loss": 0.135,
      "step": 1830
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.6478027189377173e-05,
      "loss": 0.1465,
      "step": 1840
    },
    {
      "epoch": 2.63,
      "learning_rate": 2.631994941511224e-05,
      "loss": 0.1275,
      "step": 1850
    },
    {
      "epoch": 2.63,
      "eval_accuracy": 0.9466,
      "eval_loss": 0.1340366005897522,
      "eval_runtime": 70.3004,
      "eval_samples_per_second": 142.247,
      "eval_steps_per_second": 4.452,
      "step": 1850
    },
    {
      "epoch": 2.64,
      "learning_rate": 2.6161871640847298e-05,
      "loss": 0.1235,
      "step": 1860
    },
    {
      "epoch": 2.66,
      "learning_rate": 2.600379386658236e-05,
      "loss": 0.1347,
      "step": 1870
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.5845716092317422e-05,
      "loss": 0.1371,
      "step": 1880
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.568763831805248e-05,
      "loss": 0.1623,
      "step": 1890
    },
    {
      "epoch": 2.7,
      "learning_rate": 2.5529560543787547e-05,
      "loss": 0.1582,
      "step": 1900
    },
    {
      "epoch": 2.7,
      "eval_accuracy": 0.9391,
      "eval_loss": 0.15011067688465118,
      "eval_runtime": 70.1659,
      "eval_samples_per_second": 142.519,
      "eval_steps_per_second": 4.461,
      "step": 1900
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.537148276952261e-05,
      "loss": 0.1726,
      "step": 1910
    },
    {
      "epoch": 2.73,
      "learning_rate": 2.5213404995257668e-05,
      "loss": 0.1549,
      "step": 1920
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.505532722099273e-05,
      "loss": 0.1374,
      "step": 1930
    },
    {
      "epoch": 2.76,
      "learning_rate": 2.4897249446727792e-05,
      "loss": 0.1382,
      "step": 1940
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.473917167246285e-05,
      "loss": 0.1472,
      "step": 1950
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 0.9451,
      "eval_loss": 0.135423481464386,
      "eval_runtime": 70.1579,
      "eval_samples_per_second": 142.536,
      "eval_steps_per_second": 4.461,
      "step": 1950
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.4581093898197917e-05,
      "loss": 0.1455,
      "step": 1960
    },
    {
      "epoch": 2.8,
      "learning_rate": 2.4423016123932975e-05,
      "loss": 0.1395,
      "step": 1970
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.4264938349668038e-05,
      "loss": 0.1409,
      "step": 1980
    },
    {
      "epoch": 2.83,
      "learning_rate": 2.41068605754031e-05,
      "loss": 0.1559,
      "step": 1990
    },
    {
      "epoch": 2.84,
      "learning_rate": 2.3948782801138162e-05,
      "loss": 0.1522,
      "step": 2000
    },
    {
      "epoch": 2.84,
      "eval_accuracy": 0.9479,
      "eval_loss": 0.13092650473117828,
      "eval_runtime": 70.111,
      "eval_samples_per_second": 142.631,
      "eval_steps_per_second": 4.464,
      "step": 2000
    },
    {
      "epoch": 2.86,
      "learning_rate": 2.379070502687322e-05,
      "loss": 0.13,
      "step": 2010
    },
    {
      "epoch": 2.87,
      "learning_rate": 2.3632627252608283e-05,
      "loss": 0.1348,
      "step": 2020
    },
    {
      "epoch": 2.89,
      "learning_rate": 2.3474549478343345e-05,
      "loss": 0.1295,
      "step": 2030
    },
    {
      "epoch": 2.9,
      "learning_rate": 2.3316471704078407e-05,
      "loss": 0.1401,
      "step": 2040
    },
    {
      "epoch": 2.92,
      "learning_rate": 2.315839392981347e-05,
      "loss": 0.1593,
      "step": 2050
    },
    {
      "epoch": 2.92,
      "eval_accuracy": 0.9452,
      "eval_loss": 0.14333376288414001,
      "eval_runtime": 70.1364,
      "eval_samples_per_second": 142.579,
      "eval_steps_per_second": 4.463,
      "step": 2050
    },
    {
      "epoch": 2.93,
      "learning_rate": 2.3000316155548532e-05,
      "loss": 0.1168,
      "step": 2060
    },
    {
      "epoch": 2.94,
      "learning_rate": 2.2842238381283594e-05,
      "loss": 0.1377,
      "step": 2070
    },
    {
      "epoch": 2.96,
      "learning_rate": 2.2684160607018653e-05,
      "loss": 0.1405,
      "step": 2080
    },
    {
      "epoch": 2.97,
      "learning_rate": 2.252608283275372e-05,
      "loss": 0.1709,
      "step": 2090
    },
    {
      "epoch": 2.99,
      "learning_rate": 2.2368005058488777e-05,
      "loss": 0.1541,
      "step": 2100
    },
    {
      "epoch": 2.99,
      "eval_accuracy": 0.9466,
      "eval_loss": 0.13812462985515594,
      "eval_runtime": 70.125,
      "eval_samples_per_second": 142.602,
      "eval_steps_per_second": 4.463,
      "step": 2100
    },
    {
      "epoch": 3.0,
      "learning_rate": 2.220992728422384e-05,
      "loss": 0.1462,
      "step": 2110
    },
    {
      "epoch": 3.01,
      "learning_rate": 2.20518495099589e-05,
      "loss": 0.1057,
      "step": 2120
    },
    {
      "epoch": 3.03,
      "learning_rate": 2.1893771735693964e-05,
      "loss": 0.1337,
      "step": 2130
    },
    {
      "epoch": 3.04,
      "learning_rate": 2.1735693961429023e-05,
      "loss": 0.1013,
      "step": 2140
    },
    {
      "epoch": 3.06,
      "learning_rate": 2.1577616187164085e-05,
      "loss": 0.1297,
      "step": 2150
    },
    {
      "epoch": 3.06,
      "eval_accuracy": 0.9479,
      "eval_loss": 0.13202042877674103,
      "eval_runtime": 70.2017,
      "eval_samples_per_second": 142.447,
      "eval_steps_per_second": 4.459,
      "step": 2150
    },
    {
      "epoch": 3.06,
      "step": 2150,
      "total_flos": 2.179291401202729e+19,
      "train_loss": 0.23635822601096573,
      "train_runtime": 8476.8194,
      "train_samples_per_second": 53.086,
      "train_steps_per_second": 0.415
    }
  ],
  "max_steps": 3515,
  "num_train_epochs": 5,
  "total_flos": 2.179291401202729e+19,
  "trial_name": null,
  "trial_params": null
}