{
  "best_metric": 1.1361786127090454,
  "best_model_checkpoint": "./vit-base-beans/checkpoint-480",
  "epoch": 15.0,
  "global_step": 1950,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "learning_rate": 0.00019897435897435898,
      "loss": 3.2519,
      "step": 10
    },
    {
      "epoch": 0.15,
      "learning_rate": 0.00019794871794871796,
      "loss": 3.1095,
      "step": 20
    },
    {
      "epoch": 0.23,
      "learning_rate": 0.00019692307692307696,
      "loss": 2.9002,
      "step": 30
    },
    {
      "epoch": 0.31,
      "learning_rate": 0.0001958974358974359,
      "loss": 2.7917,
      "step": 40
    },
    {
      "epoch": 0.31,
      "eval_accuracy": 0.35233160621761656,
      "eval_loss": 2.6724953651428223,
      "eval_runtime": 20.1054,
      "eval_samples_per_second": 57.596,
      "eval_steps_per_second": 7.212,
      "step": 40
    },
    {
      "epoch": 0.38,
      "learning_rate": 0.00019487179487179487,
      "loss": 2.6291,
      "step": 50
    },
    {
      "epoch": 0.46,
      "learning_rate": 0.00019384615384615385,
      "loss": 2.492,
      "step": 60
    },
    {
      "epoch": 0.54,
      "learning_rate": 0.00019282051282051282,
      "loss": 2.4302,
      "step": 70
    },
    {
      "epoch": 0.62,
      "learning_rate": 0.00019179487179487182,
      "loss": 2.3207,
      "step": 80
    },
    {
      "epoch": 0.62,
      "eval_accuracy": 0.4481865284974093,
      "eval_loss": 2.263643980026245,
      "eval_runtime": 20.6875,
      "eval_samples_per_second": 55.976,
      "eval_steps_per_second": 7.009,
      "step": 80
    },
    {
      "epoch": 0.69,
      "learning_rate": 0.0001907692307692308,
      "loss": 2.1778,
      "step": 90
    },
    {
      "epoch": 0.77,
      "learning_rate": 0.00018974358974358974,
      "loss": 2.1215,
      "step": 100
    },
    {
      "epoch": 0.85,
      "learning_rate": 0.0001887179487179487,
      "loss": 2.0333,
      "step": 110
    },
    {
      "epoch": 0.92,
      "learning_rate": 0.0001876923076923077,
      "loss": 1.9258,
      "step": 120
    },
    {
      "epoch": 0.92,
      "eval_accuracy": 0.5535405872193437,
      "eval_loss": 1.8725202083587646,
      "eval_runtime": 19.974,
      "eval_samples_per_second": 57.975,
      "eval_steps_per_second": 7.259,
      "step": 120
    },
    {
      "epoch": 1.0,
      "learning_rate": 0.0001866666666666667,
      "loss": 1.981,
      "step": 130
    },
    {
      "epoch": 1.08,
      "learning_rate": 0.00018564102564102566,
      "loss": 1.5576,
      "step": 140
    },
    {
      "epoch": 1.15,
      "learning_rate": 0.00018461538461538463,
      "loss": 1.4327,
      "step": 150
    },
    {
      "epoch": 1.23,
      "learning_rate": 0.00018358974358974358,
      "loss": 1.4339,
      "step": 160
    },
    {
      "epoch": 1.23,
      "eval_accuracy": 0.6260794473229706,
      "eval_loss": 1.5738734006881714,
      "eval_runtime": 20.2011,
      "eval_samples_per_second": 57.323,
      "eval_steps_per_second": 7.178,
      "step": 160
    },
    {
      "epoch": 1.31,
      "learning_rate": 0.00018256410256410258,
      "loss": 1.2945,
      "step": 170
    },
    {
      "epoch": 1.38,
      "learning_rate": 0.00018153846153846155,
      "loss": 1.3482,
      "step": 180
    },
    {
      "epoch": 1.46,
      "learning_rate": 0.00018051282051282052,
      "loss": 1.2537,
      "step": 190
    },
    {
      "epoch": 1.54,
      "learning_rate": 0.0001794871794871795,
      "loss": 1.3328,
      "step": 200
    },
    {
      "epoch": 1.54,
      "eval_accuracy": 0.6226252158894646,
      "eval_loss": 1.484718918800354,
      "eval_runtime": 20.1617,
      "eval_samples_per_second": 57.436,
      "eval_steps_per_second": 7.192,
      "step": 200
    },
    {
      "epoch": 1.62,
      "learning_rate": 0.00017846153846153847,
      "loss": 1.1274,
      "step": 210
    },
    {
      "epoch": 1.69,
      "learning_rate": 0.00017743589743589744,
      "loss": 1.169,
      "step": 220
    },
    {
      "epoch": 1.77,
      "learning_rate": 0.00017641025641025642,
      "loss": 1.1404,
      "step": 230
    },
    {
      "epoch": 1.85,
      "learning_rate": 0.0001753846153846154,
      "loss": 1.0914,
      "step": 240
    },
    {
      "epoch": 1.85,
      "eval_accuracy": 0.6442141623488774,
      "eval_loss": 1.3511847257614136,
      "eval_runtime": 19.8926,
      "eval_samples_per_second": 58.213,
      "eval_steps_per_second": 7.289,
      "step": 240
    },
    {
      "epoch": 1.92,
      "learning_rate": 0.00017435897435897436,
      "loss": 1.1632,
      "step": 250
    },
    {
      "epoch": 2.0,
      "learning_rate": 0.00017333333333333334,
      "loss": 1.029,
      "step": 260
    },
    {
      "epoch": 2.08,
      "learning_rate": 0.00017230769230769234,
      "loss": 0.7509,
      "step": 270
    },
    {
      "epoch": 2.15,
      "learning_rate": 0.00017128205128205128,
      "loss": 0.6594,
      "step": 280
    },
    {
      "epoch": 2.15,
      "eval_accuracy": 0.6450777202072538,
      "eval_loss": 1.2636480331420898,
      "eval_runtime": 20.434,
      "eval_samples_per_second": 56.67,
      "eval_steps_per_second": 7.096,
      "step": 280
    },
    {
      "epoch": 2.23,
      "learning_rate": 0.00017025641025641026,
      "loss": 0.557,
      "step": 290
    },
    {
      "epoch": 2.31,
      "learning_rate": 0.00016923076923076923,
      "loss": 0.5061,
      "step": 300
    },
    {
      "epoch": 2.38,
      "learning_rate": 0.00016820512820512823,
      "loss": 0.53,
      "step": 310
    },
    {
      "epoch": 2.46,
      "learning_rate": 0.0001671794871794872,
      "loss": 0.5651,
      "step": 320
    },
    {
      "epoch": 2.46,
      "eval_accuracy": 0.6649395509499136,
      "eval_loss": 1.2330708503723145,
      "eval_runtime": 20.6893,
      "eval_samples_per_second": 55.971,
      "eval_steps_per_second": 7.008,
      "step": 320
    },
    {
      "epoch": 2.54,
      "learning_rate": 0.00016615384615384617,
      "loss": 0.5417,
      "step": 330
    },
    {
      "epoch": 2.62,
      "learning_rate": 0.00016512820512820512,
      "loss": 0.5584,
      "step": 340
    },
    {
      "epoch": 2.69,
      "learning_rate": 0.0001641025641025641,
      "loss": 0.6261,
      "step": 350
    },
    {
      "epoch": 2.77,
      "learning_rate": 0.0001630769230769231,
      "loss": 0.6847,
      "step": 360
    },
    {
      "epoch": 2.77,
      "eval_accuracy": 0.6381692573402418,
      "eval_loss": 1.3150428533554077,
      "eval_runtime": 20.8049,
      "eval_samples_per_second": 55.66,
      "eval_steps_per_second": 6.97,
      "step": 360
    },
    {
      "epoch": 2.85,
      "learning_rate": 0.00016205128205128207,
      "loss": 0.6079,
      "step": 370
    },
    {
      "epoch": 2.92,
      "learning_rate": 0.00016102564102564104,
      "loss": 0.6087,
      "step": 380
    },
    {
      "epoch": 3.0,
      "learning_rate": 0.00016,
      "loss": 0.6049,
      "step": 390
    },
    {
      "epoch": 3.08,
      "learning_rate": 0.00015897435897435896,
      "loss": 0.2965,
      "step": 400
    },
    {
      "epoch": 3.08,
      "eval_accuracy": 0.6735751295336787,
      "eval_loss": 1.1951243877410889,
      "eval_runtime": 20.3022,
      "eval_samples_per_second": 57.038,
      "eval_steps_per_second": 7.142,
      "step": 400
    },
    {
      "epoch": 3.15,
      "learning_rate": 0.00015794871794871796,
      "loss": 0.3409,
      "step": 410
    },
    {
      "epoch": 3.23,
      "learning_rate": 0.00015692307692307693,
      "loss": 0.2707,
      "step": 420
    },
    {
      "epoch": 3.31,
      "learning_rate": 0.0001558974358974359,
      "loss": 0.2439,
      "step": 430
    },
    {
      "epoch": 3.38,
      "learning_rate": 0.00015487179487179488,
      "loss": 0.3052,
      "step": 440
    },
    {
      "epoch": 3.38,
      "eval_accuracy": 0.5915371329879102,
      "eval_loss": 1.4183249473571777,
      "eval_runtime": 19.9664,
      "eval_samples_per_second": 57.997,
      "eval_steps_per_second": 7.262,
      "step": 440
    },
    {
      "epoch": 3.46,
      "learning_rate": 0.00015384615384615385,
      "loss": 0.2819,
      "step": 450
    },
    {
      "epoch": 3.54,
      "learning_rate": 0.00015282051282051282,
      "loss": 0.2755,
      "step": 460
    },
    {
      "epoch": 3.62,
      "learning_rate": 0.0001517948717948718,
      "loss": 0.2574,
      "step": 470
    },
    {
      "epoch": 3.69,
      "learning_rate": 0.00015076923076923077,
      "loss": 0.2592,
      "step": 480
    },
    {
      "epoch": 3.69,
      "eval_accuracy": 0.697754749568221,
      "eval_loss": 1.1361786127090454,
      "eval_runtime": 19.9742,
      "eval_samples_per_second": 57.975,
      "eval_steps_per_second": 7.259,
      "step": 480
    },
    {
      "epoch": 3.77,
      "learning_rate": 0.00014974358974358974,
      "loss": 0.2762,
      "step": 490
    },
    {
      "epoch": 3.85,
      "learning_rate": 0.00014871794871794872,
      "loss": 0.2489,
      "step": 500
    },
    {
      "epoch": 3.92,
      "learning_rate": 0.00014769230769230772,
      "loss": 0.2119,
      "step": 510
    },
    {
      "epoch": 4.0,
      "learning_rate": 0.00014666666666666666,
      "loss": 0.3058,
      "step": 520
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6303972366148531,
      "eval_loss": 1.4736298322677612,
      "eval_runtime": 20.1619,
      "eval_samples_per_second": 57.435,
      "eval_steps_per_second": 7.192,
      "step": 520
    },
    {
      "epoch": 4.08,
      "learning_rate": 0.00014564102564102564,
      "loss": 0.2272,
      "step": 530
    },
    {
      "epoch": 4.15,
      "learning_rate": 0.0001446153846153846,
      "loss": 0.1024,
      "step": 540
    },
    {
      "epoch": 4.23,
      "learning_rate": 0.0001435897435897436,
      "loss": 0.1202,
      "step": 550
    },
    {
      "epoch": 4.31,
      "learning_rate": 0.00014256410256410258,
      "loss": 0.1412,
      "step": 560
    },
    {
      "epoch": 4.31,
      "eval_accuracy": 0.6528497409326425,
      "eval_loss": 1.2578930854797363,
      "eval_runtime": 20.8549,
      "eval_samples_per_second": 55.527,
      "eval_steps_per_second": 6.953,
      "step": 560
    },
    {
      "epoch": 4.38,
      "learning_rate": 0.00014153846153846156,
      "loss": 0.1405,
      "step": 570
    },
    {
      "epoch": 4.46,
      "learning_rate": 0.0001405128205128205,
      "loss": 0.1154,
      "step": 580
    },
    {
      "epoch": 4.54,
      "learning_rate": 0.00013948717948717947,
      "loss": 0.1481,
      "step": 590
    },
    {
      "epoch": 4.62,
      "learning_rate": 0.00013846153846153847,
      "loss": 0.1141,
      "step": 600
    },
    {
      "epoch": 4.62,
      "eval_accuracy": 0.6675302245250432,
      "eval_loss": 1.2793298959732056,
      "eval_runtime": 20.7805,
      "eval_samples_per_second": 55.725,
      "eval_steps_per_second": 6.978,
      "step": 600
    },
    {
      "epoch": 4.69,
      "learning_rate": 0.00013743589743589745,
      "loss": 0.1616,
      "step": 610
    },
    {
      "epoch": 4.77,
      "learning_rate": 0.00013641025641025642,
      "loss": 0.1792,
      "step": 620
    },
    {
      "epoch": 4.85,
      "learning_rate": 0.0001353846153846154,
      "loss": 0.0885,
      "step": 630
    },
    {
      "epoch": 4.92,
      "learning_rate": 0.00013435897435897437,
      "loss": 0.1553,
      "step": 640
    },
    {
      "epoch": 4.92,
      "eval_accuracy": 0.6588946459412781,
      "eval_loss": 1.3182828426361084,
      "eval_runtime": 20.1017,
      "eval_samples_per_second": 57.607,
      "eval_steps_per_second": 7.213,
      "step": 640
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.00013333333333333334,
      "loss": 0.1181,
      "step": 650
    },
    {
      "epoch": 5.08,
      "learning_rate": 0.0001323076923076923,
      "loss": 0.1392,
      "step": 660
    },
    {
      "epoch": 5.15,
      "learning_rate": 0.00013128205128205129,
      "loss": 0.0673,
      "step": 670
    },
    {
      "epoch": 5.23,
      "learning_rate": 0.00013025641025641026,
      "loss": 0.059,
      "step": 680
    },
    {
      "epoch": 5.23,
      "eval_accuracy": 0.6882556131260794,
      "eval_loss": 1.2099286317825317,
      "eval_runtime": 21.3732,
      "eval_samples_per_second": 54.18,
      "eval_steps_per_second": 6.784,
      "step": 680
    },
    {
      "epoch": 5.31,
      "learning_rate": 0.00012923076923076923,
      "loss": 0.0786,
      "step": 690
    },
    {
      "epoch": 5.38,
      "learning_rate": 0.00012820512820512823,
      "loss": 0.0537,
      "step": 700
    },
    {
      "epoch": 5.46,
      "learning_rate": 0.00012717948717948718,
      "loss": 0.0639,
      "step": 710
    },
    {
      "epoch": 5.54,
      "learning_rate": 0.00012615384615384615,
      "loss": 0.0914,
      "step": 720
    },
    {
      "epoch": 5.54,
      "eval_accuracy": 0.6718480138169257,
      "eval_loss": 1.298492193222046,
      "eval_runtime": 20.1222,
      "eval_samples_per_second": 57.548,
      "eval_steps_per_second": 7.206,
      "step": 720
    },
    {
      "epoch": 5.62,
      "learning_rate": 0.00012512820512820512,
      "loss": 0.1004,
      "step": 730
    },
    {
      "epoch": 5.69,
      "learning_rate": 0.00012410256410256412,
      "loss": 0.0592,
      "step": 740
    },
    {
      "epoch": 5.77,
      "learning_rate": 0.0001230769230769231,
      "loss": 0.0688,
      "step": 750
    },
    {
      "epoch": 5.85,
      "learning_rate": 0.00012205128205128207,
      "loss": 0.0546,
      "step": 760
    },
    {
      "epoch": 5.85,
      "eval_accuracy": 0.690846286701209,
      "eval_loss": 1.282209873199463,
      "eval_runtime": 20.42,
      "eval_samples_per_second": 56.709,
      "eval_steps_per_second": 7.101,
      "step": 760
    },
    {
      "epoch": 5.92,
      "learning_rate": 0.00012102564102564103,
      "loss": 0.0285,
      "step": 770
    },
    {
      "epoch": 6.0,
      "learning_rate": 0.00012,
      "loss": 0.0318,
      "step": 780
    },
    {
      "epoch": 6.08,
      "learning_rate": 0.00011897435897435898,
      "loss": 0.0249,
      "step": 790
    },
    {
      "epoch": 6.15,
      "learning_rate": 0.00011794871794871796,
      "loss": 0.0306,
      "step": 800
    },
    {
      "epoch": 6.15,
      "eval_accuracy": 0.6735751295336787,
      "eval_loss": 1.3418058156967163,
      "eval_runtime": 20.9104,
      "eval_samples_per_second": 55.379,
      "eval_steps_per_second": 6.934,
      "step": 800
    },
    {
      "epoch": 6.23,
      "learning_rate": 0.00011692307692307694,
      "loss": 0.0689,
      "step": 810
    },
    {
      "epoch": 6.31,
      "learning_rate": 0.00011589743589743591,
      "loss": 0.0617,
      "step": 820
    },
    {
      "epoch": 6.38,
      "learning_rate": 0.00011487179487179487,
      "loss": 0.0717,
      "step": 830
    },
    {
      "epoch": 6.46,
      "learning_rate": 0.00011384615384615384,
      "loss": 0.0207,
      "step": 840
    },
    {
      "epoch": 6.46,
      "eval_accuracy": 0.697754749568221,
      "eval_loss": 1.3155086040496826,
      "eval_runtime": 20.9506,
      "eval_samples_per_second": 55.273,
      "eval_steps_per_second": 6.921,
      "step": 840
    },
    {
      "epoch": 6.54,
      "learning_rate": 0.00011282051282051283,
      "loss": 0.0192,
      "step": 850
    },
    {
      "epoch": 6.62,
      "learning_rate": 0.0001117948717948718,
      "loss": 0.0448,
      "step": 860
    },
    {
      "epoch": 6.69,
      "learning_rate": 0.00011076923076923077,
      "loss": 0.0213,
      "step": 870
    },
    {
      "epoch": 6.77,
      "learning_rate": 0.00010974358974358976,
      "loss": 0.0505,
      "step": 880
    },
    {
      "epoch": 6.77,
      "eval_accuracy": 0.6614853195164075,
      "eval_loss": 1.4209389686584473,
      "eval_runtime": 20.0879,
      "eval_samples_per_second": 57.647,
      "eval_steps_per_second": 7.218,
      "step": 880
    },
    {
      "epoch": 6.85,
      "learning_rate": 0.00010871794871794872,
      "loss": 0.0356,
      "step": 890
    },
    {
      "epoch": 6.92,
      "learning_rate": 0.0001076923076923077,
      "loss": 0.022,
      "step": 900
    },
    {
      "epoch": 7.0,
      "learning_rate": 0.00010666666666666667,
      "loss": 0.102,
      "step": 910
    },
    {
      "epoch": 7.08,
      "learning_rate": 0.00010564102564102565,
      "loss": 0.016,
      "step": 920
    },
    {
      "epoch": 7.08,
      "eval_accuracy": 0.697754749568221,
      "eval_loss": 1.3212494850158691,
      "eval_runtime": 20.3129,
      "eval_samples_per_second": 57.008,
      "eval_steps_per_second": 7.138,
      "step": 920
    },
    {
      "epoch": 7.15,
      "learning_rate": 0.00010461538461538463,
      "loss": 0.038,
      "step": 930
    },
    {
      "epoch": 7.23,
      "learning_rate": 0.0001035897435897436,
      "loss": 0.0161,
      "step": 940
    },
    {
      "epoch": 7.31,
      "learning_rate": 0.00010256410256410256,
      "loss": 0.0266,
      "step": 950
    },
    {
      "epoch": 7.38,
      "learning_rate": 0.00010153846153846153,
      "loss": 0.0268,
      "step": 960
    },
    {
      "epoch": 7.38,
      "eval_accuracy": 0.7081174438687392,
      "eval_loss": 1.2981479167938232,
      "eval_runtime": 20.6143,
      "eval_samples_per_second": 56.175,
      "eval_steps_per_second": 7.034,
      "step": 960
    },
    {
      "epoch": 7.46,
      "learning_rate": 0.00010051282051282052,
      "loss": 0.0235,
      "step": 970
    },
    {
      "epoch": 7.54,
      "learning_rate": 9.948717948717949e-05,
      "loss": 0.0126,
      "step": 980
    },
    {
      "epoch": 7.62,
      "learning_rate": 9.846153846153848e-05,
      "loss": 0.0143,
      "step": 990
    },
    {
      "epoch": 7.69,
      "learning_rate": 9.743589743589744e-05,
      "loss": 0.0299,
      "step": 1000
    },
    {
      "epoch": 7.69,
      "eval_accuracy": 0.6925734024179621,
      "eval_loss": 1.3568987846374512,
      "eval_runtime": 20.4554,
      "eval_samples_per_second": 56.611,
      "eval_steps_per_second": 7.089,
      "step": 1000
    },
    {
      "epoch": 7.77,
      "learning_rate": 9.641025641025641e-05,
      "loss": 0.0122,
      "step": 1010
    },
    {
      "epoch": 7.85,
      "learning_rate": 9.53846153846154e-05,
      "loss": 0.0203,
      "step": 1020
    },
    {
      "epoch": 7.92,
      "learning_rate": 9.435897435897436e-05,
      "loss": 0.0105,
      "step": 1030
    },
    {
      "epoch": 8.0,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.0395,
      "step": 1040
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.697754749568221,
      "eval_loss": 1.315879225730896,
      "eval_runtime": 20.1743,
      "eval_samples_per_second": 57.4,
      "eval_steps_per_second": 7.187,
      "step": 1040
    },
    {
      "epoch": 8.08,
      "learning_rate": 9.230769230769232e-05,
      "loss": 0.0102,
      "step": 1050
    },
    {
      "epoch": 8.15,
      "learning_rate": 9.128205128205129e-05,
      "loss": 0.0103,
      "step": 1060
    },
    {
      "epoch": 8.23,
      "learning_rate": 9.025641025641026e-05,
      "loss": 0.0106,
      "step": 1070
    },
    {
      "epoch": 8.31,
      "learning_rate": 8.923076923076924e-05,
      "loss": 0.0106,
      "step": 1080
    },
    {
      "epoch": 8.31,
      "eval_accuracy": 0.7046632124352331,
      "eval_loss": 1.3094946146011353,
      "eval_runtime": 20.1865,
      "eval_samples_per_second": 57.365,
      "eval_steps_per_second": 7.183,
      "step": 1080
    },
    {
      "epoch": 8.38,
      "learning_rate": 8.820512820512821e-05,
      "loss": 0.0364,
      "step": 1090
    },
    {
      "epoch": 8.46,
      "learning_rate": 8.717948717948718e-05,
      "loss": 0.0098,
      "step": 1100
    },
    {
      "epoch": 8.54,
      "learning_rate": 8.615384615384617e-05,
      "loss": 0.0462,
      "step": 1110
    },
    {
      "epoch": 8.62,
      "learning_rate": 8.512820512820513e-05,
      "loss": 0.0093,
      "step": 1120
    },
    {
      "epoch": 8.62,
      "eval_accuracy": 0.7124352331606217,
      "eval_loss": 1.3362736701965332,
      "eval_runtime": 20.3723,
      "eval_samples_per_second": 56.842,
      "eval_steps_per_second": 7.117,
      "step": 1120
    },
    {
      "epoch": 8.69,
      "learning_rate": 8.410256410256411e-05,
      "loss": 0.0097,
      "step": 1130
    },
    {
      "epoch": 8.77,
      "learning_rate": 8.307692307692309e-05,
      "loss": 0.0087,
      "step": 1140
    },
    {
      "epoch": 8.85,
      "learning_rate": 8.205128205128205e-05,
      "loss": 0.0087,
      "step": 1150
    },
    {
      "epoch": 8.92,
      "learning_rate": 8.102564102564103e-05,
      "loss": 0.0364,
      "step": 1160
    },
    {
      "epoch": 8.92,
      "eval_accuracy": 0.7089810017271158,
      "eval_loss": 1.340015172958374,
      "eval_runtime": 21.051,
      "eval_samples_per_second": 55.009,
      "eval_steps_per_second": 6.888,
      "step": 1160
    },
    {
      "epoch": 9.0,
      "learning_rate": 8e-05,
      "loss": 0.0087,
      "step": 1170
    },
    {
      "epoch": 9.08,
      "learning_rate": 7.897435897435898e-05,
      "loss": 0.0101,
      "step": 1180
    },
    {
      "epoch": 9.15,
      "learning_rate": 7.794871794871795e-05,
      "loss": 0.0174,
      "step": 1190
    },
    {
      "epoch": 9.23,
      "learning_rate": 7.692307692307693e-05,
      "loss": 0.0324,
      "step": 1200
    },
    {
      "epoch": 9.23,
      "eval_accuracy": 0.7098445595854922,
      "eval_loss": 1.3313075304031372,
      "eval_runtime": 21.981,
      "eval_samples_per_second": 52.682,
      "eval_steps_per_second": 6.597,
      "step": 1200
    },
    {
      "epoch": 9.31,
      "learning_rate": 7.58974358974359e-05,
      "loss": 0.0076,
      "step": 1210
    },
    {
      "epoch": 9.38,
      "learning_rate": 7.487179487179487e-05,
      "loss": 0.0287,
      "step": 1220
    },
    {
      "epoch": 9.46,
      "learning_rate": 7.384615384615386e-05,
      "loss": 0.0083,
      "step": 1230
    },
    {
      "epoch": 9.54,
      "learning_rate": 7.282051282051282e-05,
      "loss": 0.0076,
      "step": 1240
    },
    {
      "epoch": 9.54,
      "eval_accuracy": 0.7072538860103627,
      "eval_loss": 1.3641475439071655,
      "eval_runtime": 20.2849,
      "eval_samples_per_second": 57.087,
      "eval_steps_per_second": 7.148,
      "step": 1240
    },
    {
      "epoch": 9.62,
      "learning_rate": 7.17948717948718e-05,
      "loss": 0.0076,
      "step": 1250
    },
    {
      "epoch": 9.69,
      "learning_rate": 7.076923076923078e-05,
      "loss": 0.0075,
      "step": 1260
    },
    {
      "epoch": 9.77,
      "learning_rate": 6.974358974358974e-05,
      "loss": 0.0078,
      "step": 1270
    },
    {
      "epoch": 9.85,
      "learning_rate": 6.871794871794872e-05,
      "loss": 0.0076,
      "step": 1280
    },
    {
      "epoch": 9.85,
      "eval_accuracy": 0.7098445595854922,
      "eval_loss": 1.357639193534851,
      "eval_runtime": 20.3563,
      "eval_samples_per_second": 56.887,
      "eval_steps_per_second": 7.123,
      "step": 1280
    },
    {
      "epoch": 9.92,
      "learning_rate": 6.76923076923077e-05,
      "loss": 0.007,
      "step": 1290
    },
    {
      "epoch": 10.0,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.0075,
      "step": 1300
    },
    {
      "epoch": 10.08,
      "learning_rate": 6.564102564102564e-05,
      "loss": 0.0126,
      "step": 1310
    },
    {
      "epoch": 10.15,
      "learning_rate": 6.461538461538462e-05,
      "loss": 0.021,
      "step": 1320
    },
    {
      "epoch": 10.15,
      "eval_accuracy": 0.7124352331606217,
      "eval_loss": 1.3460208177566528,
      "eval_runtime": 20.5844,
      "eval_samples_per_second": 56.256,
      "eval_steps_per_second": 7.044,
      "step": 1320
    },
    {
      "epoch": 10.23,
      "learning_rate": 6.358974358974359e-05,
      "loss": 0.0068,
      "step": 1330
    },
    {
      "epoch": 10.31,
      "learning_rate": 6.256410256410256e-05,
      "loss": 0.0068,
      "step": 1340
    },
    {
      "epoch": 10.38,
      "learning_rate": 6.153846153846155e-05,
      "loss": 0.0066,
      "step": 1350
    },
    {
      "epoch": 10.46,
      "learning_rate": 6.0512820512820515e-05,
      "loss": 0.0064,
      "step": 1360
    },
    {
      "epoch": 10.46,
      "eval_accuracy": 0.7184801381692574,
      "eval_loss": 1.3521218299865723,
      "eval_runtime": 21.5706,
      "eval_samples_per_second": 53.684,
      "eval_steps_per_second": 6.722,
      "step": 1360
    },
    {
      "epoch": 10.54,
      "learning_rate": 5.948717948717949e-05,
      "loss": 0.0061,
      "step": 1370
    },
    {
      "epoch": 10.62,
      "learning_rate": 5.846153846153847e-05,
      "loss": 0.0062,
      "step": 1380
    },
    {
      "epoch": 10.69,
      "learning_rate": 5.7435897435897434e-05,
      "loss": 0.0364,
      "step": 1390
    },
    {
      "epoch": 10.77,
      "learning_rate": 5.6410256410256414e-05,
      "loss": 0.0069,
      "step": 1400
    },
    {
      "epoch": 10.77,
      "eval_accuracy": 0.7141623488773747,
      "eval_loss": 1.3537815809249878,
      "eval_runtime": 20.0789,
      "eval_samples_per_second": 57.672,
      "eval_steps_per_second": 7.221,
      "step": 1400
    },
    {
      "epoch": 10.85,
      "learning_rate": 5.538461538461539e-05,
      "loss": 0.0069,
      "step": 1410
    },
    {
      "epoch": 10.92,
      "learning_rate": 5.435897435897436e-05,
      "loss": 0.0068,
      "step": 1420
    },
    {
      "epoch": 11.0,
      "learning_rate": 5.333333333333333e-05,
      "loss": 0.0063,
      "step": 1430
    },
    {
      "epoch": 11.08,
      "learning_rate": 5.230769230769231e-05,
      "loss": 0.0101,
      "step": 1440
    },
    {
      "epoch": 11.08,
      "eval_accuracy": 0.7150259067357513,
      "eval_loss": 1.3637244701385498,
      "eval_runtime": 20.418,
      "eval_samples_per_second": 56.715,
      "eval_steps_per_second": 7.102,
      "step": 1440
    },
    {
      "epoch": 11.15,
      "learning_rate": 5.128205128205128e-05,
      "loss": 0.0094,
      "step": 1450
    },
    {
      "epoch": 11.23,
      "learning_rate": 5.025641025641026e-05,
      "loss": 0.0152,
      "step": 1460
    },
    {
      "epoch": 11.31,
      "learning_rate": 4.923076923076924e-05,
      "loss": 0.006,
      "step": 1470
    },
    {
      "epoch": 11.38,
      "learning_rate": 4.8205128205128205e-05,
      "loss": 0.0115,
      "step": 1480
    },
    {
      "epoch": 11.38,
      "eval_accuracy": 0.7081174438687392,
      "eval_loss": 1.3893637657165527,
      "eval_runtime": 21.0579,
      "eval_samples_per_second": 54.991,
      "eval_steps_per_second": 6.886,
      "step": 1480
    },
    {
      "epoch": 11.46,
      "learning_rate": 4.717948717948718e-05,
      "loss": 0.0062,
      "step": 1490
    },
    {
      "epoch": 11.54,
      "learning_rate": 4.615384615384616e-05,
      "loss": 0.0058,
      "step": 1500
    },
    {
      "epoch": 11.62,
      "learning_rate": 4.512820512820513e-05,
      "loss": 0.0058,
      "step": 1510
    },
    {
      "epoch": 11.69,
      "learning_rate": 4.4102564102564104e-05,
      "loss": 0.0062,
      "step": 1520
    },
    {
      "epoch": 11.69,
      "eval_accuracy": 0.7072538860103627,
      "eval_loss": 1.3909190893173218,
      "eval_runtime": 20.5699,
      "eval_samples_per_second": 56.296,
      "eval_steps_per_second": 7.049,
      "step": 1520
    },
    {
      "epoch": 11.77,
      "learning_rate": 4.3076923076923084e-05,
      "loss": 0.0171,
      "step": 1530
    },
    {
      "epoch": 11.85,
      "learning_rate": 4.205128205128206e-05,
      "loss": 0.0159,
      "step": 1540
    },
    {
      "epoch": 11.92,
      "learning_rate": 4.1025641025641023e-05,
      "loss": 0.0054,
      "step": 1550
    },
    {
      "epoch": 12.0,
      "learning_rate": 4e-05,
      "loss": 0.0055,
      "step": 1560
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.7124352331606217,
      "eval_loss": 1.3742793798446655,
      "eval_runtime": 20.0602,
      "eval_samples_per_second": 57.726,
      "eval_steps_per_second": 7.228,
      "step": 1560
    },
    {
      "epoch": 12.08,
      "learning_rate": 3.8974358974358976e-05,
      "loss": 0.0056,
      "step": 1570
    },
    {
      "epoch": 12.15,
      "learning_rate": 3.794871794871795e-05,
      "loss": 0.0234,
      "step": 1580
    },
    {
      "epoch": 12.23,
      "learning_rate": 3.692307692307693e-05,
      "loss": 0.0053,
      "step": 1590
    },
    {
      "epoch": 12.31,
      "learning_rate": 3.58974358974359e-05,
      "loss": 0.0054,
      "step": 1600
    },
    {
      "epoch": 12.31,
      "eval_accuracy": 0.7158894645941278,
      "eval_loss": 1.3830854892730713,
      "eval_runtime": 20.2514,
      "eval_samples_per_second": 57.181,
      "eval_steps_per_second": 7.16,
      "step": 1600
    },
    {
      "epoch": 12.38,
      "learning_rate": 3.487179487179487e-05,
      "loss": 0.0052,
      "step": 1610
    },
    {
      "epoch": 12.46,
      "learning_rate": 3.384615384615385e-05,
      "loss": 0.0052,
      "step": 1620
    },
    {
      "epoch": 12.54,
      "learning_rate": 3.282051282051282e-05,
      "loss": 0.0056,
      "step": 1630
    },
    {
      "epoch": 12.62,
      "learning_rate": 3.1794871794871795e-05,
      "loss": 0.0054,
      "step": 1640
    },
    {
      "epoch": 12.62,
      "eval_accuracy": 0.7167530224525043,
      "eval_loss": 1.3866362571716309,
      "eval_runtime": 21.0166,
      "eval_samples_per_second": 55.099,
      "eval_steps_per_second": 6.899,
      "step": 1640
    },
    {
      "epoch": 12.69,
      "learning_rate": 3.0769230769230774e-05,
      "loss": 0.0052,
      "step": 1650
    },
    {
      "epoch": 12.77,
      "learning_rate": 2.9743589743589744e-05,
      "loss": 0.0052,
      "step": 1660
    },
    {
      "epoch": 12.85,
      "learning_rate": 2.8717948717948717e-05,
      "loss": 0.0058,
      "step": 1670
    },
    {
      "epoch": 12.92,
      "learning_rate": 2.7692307692307694e-05,
      "loss": 0.0147,
      "step": 1680
    },
    {
      "epoch": 12.92,
      "eval_accuracy": 0.7150259067357513,
      "eval_loss": 1.391139030456543,
      "eval_runtime": 20.4472,
      "eval_samples_per_second": 56.634,
      "eval_steps_per_second": 7.091,
      "step": 1680
    },
    {
      "epoch": 13.0,
      "learning_rate": 2.6666666666666667e-05,
      "loss": 0.0051,
      "step": 1690
    },
    {
      "epoch": 13.08,
      "learning_rate": 2.564102564102564e-05,
      "loss": 0.0144,
      "step": 1700
    },
    {
      "epoch": 13.15,
      "learning_rate": 2.461538461538462e-05,
      "loss": 0.0052,
      "step": 1710
    },
    {
      "epoch": 13.23,
      "learning_rate": 2.358974358974359e-05,
      "loss": 0.0058,
      "step": 1720
    },
    {
      "epoch": 13.23,
      "eval_accuracy": 0.7141623488773747,
      "eval_loss": 1.3986831903457642,
      "eval_runtime": 20.1538,
      "eval_samples_per_second": 57.458,
      "eval_steps_per_second": 7.195,
      "step": 1720
    },
    {
      "epoch": 13.31,
      "learning_rate": 2.2564102564102566e-05,
      "loss": 0.0047,
      "step": 1730
    },
    {
      "epoch": 13.38,
      "learning_rate": 2.1538461538461542e-05,
      "loss": 0.0053,
      "step": 1740
    },
    {
      "epoch": 13.46,
      "learning_rate": 2.0512820512820512e-05,
      "loss": 0.005,
      "step": 1750
    },
    {
      "epoch": 13.54,
      "learning_rate": 1.9487179487179488e-05,
      "loss": 0.0096,
      "step": 1760
    },
    {
      "epoch": 13.54,
      "eval_accuracy": 0.7132987910189983,
      "eval_loss": 1.3994832038879395,
      "eval_runtime": 20.378,
      "eval_samples_per_second": 56.826,
      "eval_steps_per_second": 7.116,
      "step": 1760
    },
    {
      "epoch": 13.62,
      "learning_rate": 1.8461538461538465e-05,
      "loss": 0.0055,
      "step": 1770
    },
    {
      "epoch": 13.69,
      "learning_rate": 1.7435897435897434e-05,
      "loss": 0.0112,
      "step": 1780
    },
    {
      "epoch": 13.77,
      "learning_rate": 1.641025641025641e-05,
      "loss": 0.0047,
      "step": 1790
    },
    {
      "epoch": 13.85,
      "learning_rate": 1.5384615384615387e-05,
      "loss": 0.0048,
      "step": 1800
    },
    {
      "epoch": 13.85,
      "eval_accuracy": 0.7115716753022453,
      "eval_loss": 1.401125192642212,
      "eval_runtime": 20.4927,
      "eval_samples_per_second": 56.508,
      "eval_steps_per_second": 7.076,
      "step": 1800
    },
    {
      "epoch": 13.92,
      "learning_rate": 1.4358974358974359e-05,
      "loss": 0.0053,
      "step": 1810
    },
    {
      "epoch": 14.0,
      "learning_rate": 1.3333333333333333e-05,
      "loss": 0.0047,
      "step": 1820
    },
    {
      "epoch": 14.08,
      "learning_rate": 1.230769230769231e-05,
      "loss": 0.0146,
      "step": 1830
    },
    {
      "epoch": 14.15,
      "learning_rate": 1.1282051282051283e-05,
      "loss": 0.0054,
      "step": 1840
    },
    {
      "epoch": 14.15,
      "eval_accuracy": 0.7115716753022453,
      "eval_loss": 1.4026780128479004,
      "eval_runtime": 20.8793,
      "eval_samples_per_second": 55.462,
      "eval_steps_per_second": 6.945,
      "step": 1840
    },
    {
      "epoch": 14.23,
      "learning_rate": 1.0256410256410256e-05,
      "loss": 0.0046,
      "step": 1850
    },
    {
      "epoch": 14.31,
      "learning_rate": 9.230769230769232e-06,
      "loss": 0.0046,
      "step": 1860
    },
    {
      "epoch": 14.38,
      "learning_rate": 8.205128205128205e-06,
      "loss": 0.005,
      "step": 1870
    },
    {
      "epoch": 14.46,
      "learning_rate": 7.179487179487179e-06,
      "loss": 0.0048,
      "step": 1880
    },
    {
      "epoch": 14.46,
      "eval_accuracy": 0.7115716753022453,
      "eval_loss": 1.403483510017395,
      "eval_runtime": 21.9216,
      "eval_samples_per_second": 52.825,
      "eval_steps_per_second": 6.614,
      "step": 1880
    },
    {
      "epoch": 14.54,
      "learning_rate": 6.153846153846155e-06,
      "loss": 0.0089,
      "step": 1890
    },
    {
      "epoch": 14.62,
      "learning_rate": 5.128205128205128e-06,
      "loss": 0.0047,
      "step": 1900
    },
    {
      "epoch": 14.69,
      "learning_rate": 4.102564102564103e-06,
      "loss": 0.0048,
      "step": 1910
    },
    {
      "epoch": 14.77,
      "learning_rate": 3.0769230769230774e-06,
      "loss": 0.0047,
      "step": 1920
    },
    {
      "epoch": 14.77,
      "eval_accuracy": 0.7124352331606217,
      "eval_loss": 1.4049346446990967,
      "eval_runtime": 20.2662,
      "eval_samples_per_second": 57.139,
      "eval_steps_per_second": 7.155,
      "step": 1920
    },
    {
      "epoch": 14.85,
      "learning_rate": 2.0512820512820513e-06,
      "loss": 0.005,
      "step": 1930
    },
    {
      "epoch": 14.92,
      "learning_rate": 1.0256410256410257e-06,
      "loss": 0.0049,
      "step": 1940
    },
    {
      "epoch": 15.0,
      "learning_rate": 0.0,
      "loss": 0.0109,
      "step": 1950
    },
    {
      "epoch": 15.0,
      "step": 1950,
      "total_flos": 2.4159665336144486e+18,
      "train_loss": 0.33013623780929124,
      "train_runtime": 2443.4946,
      "train_samples_per_second": 12.756,
      "train_steps_per_second": 0.798
    }
  ],
  "max_steps": 1950,
  "num_train_epochs": 15,
  "total_flos": 2.4159665336144486e+18,
  "trial_name": null,
  "trial_params": null
}