{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 27.0,
  "eval_steps": 500,
  "global_step": 3375,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.08,
      "grad_norm": 10.4375,
      "learning_rate": 8e-05,
      "loss": 2.8794,
      "step": 10
    },
    {
      "epoch": 0.16,
      "grad_norm": 11.8125,
      "learning_rate": 8e-05,
      "loss": 1.3765,
      "step": 20
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.375,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 30
    },
    {
      "epoch": 0.32,
      "grad_norm": 7.5,
      "learning_rate": 8e-05,
      "loss": 1.0576,
      "step": 40
    },
    {
      "epoch": 0.4,
      "grad_norm": 18.125,
      "learning_rate": 8e-05,
      "loss": 1.2152,
      "step": 50
    },
    {
      "epoch": 0.48,
      "grad_norm": 8.9375,
      "learning_rate": 8e-05,
      "loss": 1.1611,
      "step": 60
    },
    {
      "epoch": 0.56,
      "grad_norm": 7.03125,
      "learning_rate": 8e-05,
      "loss": 1.3899,
      "step": 70
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.875,
      "learning_rate": 8e-05,
      "loss": 1.1901,
      "step": 80
    },
    {
      "epoch": 0.72,
      "grad_norm": 7.8125,
      "learning_rate": 8e-05,
      "loss": 1.0439,
      "step": 90
    },
    {
      "epoch": 0.8,
      "grad_norm": 7.4375,
      "learning_rate": 8e-05,
      "loss": 1.0462,
      "step": 100
    },
    {
      "epoch": 0.88,
      "grad_norm": 7.4375,
      "learning_rate": 8e-05,
      "loss": 1.1264,
      "step": 110
    },
    {
      "epoch": 0.96,
      "grad_norm": 6.78125,
      "learning_rate": 8e-05,
      "loss": 0.9896,
      "step": 120
    },
    {
      "epoch": 1.04,
      "grad_norm": 7.09375,
      "learning_rate": 8e-05,
      "loss": 0.9057,
      "step": 130
    },
    {
      "epoch": 1.12,
      "grad_norm": 7.125,
      "learning_rate": 8e-05,
      "loss": 0.8523,
      "step": 140
    },
    {
      "epoch": 1.2,
      "grad_norm": 6.6875,
      "learning_rate": 8e-05,
      "loss": 0.7495,
      "step": 150
    },
    {
      "epoch": 1.28,
      "grad_norm": 7.90625,
      "learning_rate": 8e-05,
      "loss": 0.7579,
      "step": 160
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 7.0,
      "learning_rate": 8e-05,
      "loss": 0.8312,
      "step": 170
    },
    {
      "epoch": 1.44,
      "grad_norm": 7.53125,
      "learning_rate": 8e-05,
      "loss": 0.8347,
      "step": 180
    },
    {
      "epoch": 1.52,
      "grad_norm": 7.03125,
      "learning_rate": 8e-05,
      "loss": 0.77,
      "step": 190
    },
    {
      "epoch": 1.6,
      "grad_norm": 18.375,
      "learning_rate": 8e-05,
      "loss": 0.7835,
      "step": 200
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 8.0,
      "learning_rate": 8e-05,
      "loss": 0.7207,
      "step": 210
    },
    {
      "epoch": 1.76,
      "grad_norm": 6.90625,
      "learning_rate": 8e-05,
      "loss": 0.7986,
      "step": 220
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 7.65625,
      "learning_rate": 8e-05,
      "loss": 0.8527,
      "step": 230
    },
    {
      "epoch": 1.92,
      "grad_norm": 7.28125,
      "learning_rate": 8e-05,
      "loss": 0.9771,
      "step": 240
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.84375,
      "learning_rate": 8e-05,
      "loss": 0.7728,
      "step": 250
    },
    {
      "epoch": 2.08,
      "grad_norm": 7.375,
      "learning_rate": 8e-05,
      "loss": 0.5319,
      "step": 260
    },
    {
      "epoch": 2.16,
      "grad_norm": 6.9375,
      "learning_rate": 8e-05,
      "loss": 0.624,
      "step": 270
    },
    {
      "epoch": 2.24,
      "grad_norm": 6.8125,
      "learning_rate": 8e-05,
      "loss": 0.8059,
      "step": 280
    },
    {
      "epoch": 2.32,
      "grad_norm": 7.375,
      "learning_rate": 8e-05,
      "loss": 0.5929,
      "step": 290
    },
    {
      "epoch": 2.4,
      "grad_norm": 6.65625,
      "learning_rate": 8e-05,
      "loss": 0.5317,
      "step": 300
    },
    {
      "epoch": 2.48,
      "grad_norm": 10.0,
      "learning_rate": 8e-05,
      "loss": 0.5592,
      "step": 310
    },
    {
      "epoch": 2.56,
      "grad_norm": 6.4375,
      "learning_rate": 8e-05,
      "loss": 0.6473,
      "step": 320
    },
    {
      "epoch": 2.64,
      "grad_norm": 8.5625,
      "learning_rate": 8e-05,
      "loss": 0.6117,
      "step": 330
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 8.0625,
      "learning_rate": 8e-05,
      "loss": 0.6364,
      "step": 340
    },
    {
      "epoch": 2.8,
      "grad_norm": 6.59375,
      "learning_rate": 8e-05,
      "loss": 0.5996,
      "step": 350
    },
    {
      "epoch": 2.88,
      "grad_norm": 7.0625,
      "learning_rate": 8e-05,
      "loss": 0.6544,
      "step": 360
    },
    {
      "epoch": 2.96,
      "grad_norm": 6.84375,
      "learning_rate": 8e-05,
      "loss": 0.5443,
      "step": 370
    },
    {
      "epoch": 3.04,
      "grad_norm": 5.6875,
      "learning_rate": 8e-05,
      "loss": 0.5767,
      "step": 380
    },
    {
      "epoch": 3.12,
      "grad_norm": 6.0625,
      "learning_rate": 8e-05,
      "loss": 0.4166,
      "step": 390
    },
    {
      "epoch": 3.2,
      "grad_norm": 7.125,
      "learning_rate": 8e-05,
      "loss": 0.5334,
      "step": 400
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 14.5,
      "learning_rate": 8e-05,
      "loss": 0.4994,
      "step": 410
    },
    {
      "epoch": 3.36,
      "grad_norm": 7.6875,
      "learning_rate": 8e-05,
      "loss": 0.4856,
      "step": 420
    },
    {
      "epoch": 3.44,
      "grad_norm": 6.875,
      "learning_rate": 8e-05,
      "loss": 0.5321,
      "step": 430
    },
    {
      "epoch": 3.52,
      "grad_norm": 7.375,
      "learning_rate": 8e-05,
      "loss": 0.5155,
      "step": 440
    },
    {
      "epoch": 3.6,
      "grad_norm": 6.25,
      "learning_rate": 8e-05,
      "loss": 0.512,
      "step": 450
    },
    {
      "epoch": 3.68,
      "grad_norm": 14.9375,
      "learning_rate": 8e-05,
      "loss": 0.5088,
      "step": 460
    },
    {
      "epoch": 3.76,
      "grad_norm": 6.65625,
      "learning_rate": 8e-05,
      "loss": 0.494,
      "step": 470
    },
    {
      "epoch": 3.84,
      "grad_norm": 7.34375,
      "learning_rate": 8e-05,
      "loss": 0.4257,
      "step": 480
    },
    {
      "epoch": 3.92,
      "grad_norm": 6.5,
      "learning_rate": 8e-05,
      "loss": 0.5417,
      "step": 490
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.53125,
      "learning_rate": 8e-05,
      "loss": 0.457,
      "step": 500
    },
    {
      "epoch": 4.0,
      "eval_loss": 5.955487251281738,
      "eval_runtime": 7.9651,
      "eval_samples_per_second": 62.774,
      "eval_steps_per_second": 62.774,
      "step": 500
    },
    {
      "epoch": 4.08,
      "grad_norm": 15.3125,
      "learning_rate": 8e-05,
      "loss": 0.4382,
      "step": 510
    },
    {
      "epoch": 4.16,
      "grad_norm": 5.5625,
      "learning_rate": 8e-05,
      "loss": 0.349,
      "step": 520
    },
    {
      "epoch": 4.24,
      "grad_norm": 6.25,
      "learning_rate": 8e-05,
      "loss": 0.3689,
      "step": 530
    },
    {
      "epoch": 4.32,
      "grad_norm": 7.0625,
      "learning_rate": 8e-05,
      "loss": 0.3873,
      "step": 540
    },
    {
      "epoch": 4.4,
      "grad_norm": 5.9375,
      "learning_rate": 8e-05,
      "loss": 0.4832,
      "step": 550
    },
    {
      "epoch": 4.48,
      "grad_norm": 6.0625,
      "learning_rate": 8e-05,
      "loss": 0.4306,
      "step": 560
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 6.5625,
      "learning_rate": 8e-05,
      "loss": 0.38,
      "step": 570
    },
    {
      "epoch": 4.64,
      "grad_norm": 13.4375,
      "learning_rate": 8e-05,
      "loss": 0.3777,
      "step": 580
    },
    {
      "epoch": 4.72,
      "grad_norm": 6.03125,
      "learning_rate": 8e-05,
      "loss": 0.3718,
      "step": 590
    },
    {
      "epoch": 4.8,
      "grad_norm": 16.125,
      "learning_rate": 8e-05,
      "loss": 0.3895,
      "step": 600
    },
    {
      "epoch": 4.88,
      "grad_norm": 6.3125,
      "learning_rate": 8e-05,
      "loss": 0.4697,
      "step": 610
    },
    {
      "epoch": 4.96,
      "grad_norm": 5.46875,
      "learning_rate": 8e-05,
      "loss": 0.4896,
      "step": 620
    },
    {
      "epoch": 5.04,
      "grad_norm": 3.671875,
      "learning_rate": 8e-05,
      "loss": 0.4276,
      "step": 630
    },
    {
      "epoch": 5.12,
      "grad_norm": 7.34375,
      "learning_rate": 8e-05,
      "loss": 0.2484,
      "step": 640
    },
    {
      "epoch": 5.2,
      "grad_norm": 5.8125,
      "learning_rate": 8e-05,
      "loss": 0.3089,
      "step": 650
    },
    {
      "epoch": 5.28,
      "grad_norm": 3.546875,
      "learning_rate": 8e-05,
      "loss": 0.2647,
      "step": 660
    },
    {
      "epoch": 5.36,
      "grad_norm": 5.46875,
      "learning_rate": 8e-05,
      "loss": 0.2967,
      "step": 670
    },
    {
      "epoch": 5.44,
      "grad_norm": 7.84375,
      "learning_rate": 8e-05,
      "loss": 0.306,
      "step": 680
    },
    {
      "epoch": 5.52,
      "grad_norm": 13.9375,
      "learning_rate": 8e-05,
      "loss": 0.3192,
      "step": 690
    },
    {
      "epoch": 5.6,
      "grad_norm": 6.34375,
      "learning_rate": 8e-05,
      "loss": 0.2927,
      "step": 700
    },
    {
      "epoch": 5.68,
      "grad_norm": 6.4375,
      "learning_rate": 8e-05,
      "loss": 0.3072,
      "step": 710
    },
    {
      "epoch": 5.76,
      "grad_norm": 5.78125,
      "learning_rate": 8e-05,
      "loss": 0.3086,
      "step": 720
    },
    {
      "epoch": 5.84,
      "grad_norm": 5.21875,
      "learning_rate": 8e-05,
      "loss": 0.3277,
      "step": 730
    },
    {
      "epoch": 5.92,
      "grad_norm": 5.65625,
      "learning_rate": 8e-05,
      "loss": 0.2554,
      "step": 740
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.59375,
      "learning_rate": 8e-05,
      "loss": 0.3697,
      "step": 750
    },
    {
      "epoch": 6.08,
      "grad_norm": 5.09375,
      "learning_rate": 8e-05,
      "loss": 0.2716,
      "step": 760
    },
    {
      "epoch": 6.16,
      "grad_norm": 2.734375,
      "learning_rate": 8e-05,
      "loss": 0.2084,
      "step": 770
    },
    {
      "epoch": 6.24,
      "grad_norm": 5.875,
      "learning_rate": 8e-05,
      "loss": 0.307,
      "step": 780
    },
    {
      "epoch": 6.32,
      "grad_norm": 2.96875,
      "learning_rate": 8e-05,
      "loss": 0.2431,
      "step": 790
    },
    {
      "epoch": 6.4,
      "grad_norm": 2.46875,
      "learning_rate": 8e-05,
      "loss": 0.2562,
      "step": 800
    },
    {
      "epoch": 6.48,
      "grad_norm": 4.96875,
      "learning_rate": 8e-05,
      "loss": 0.1857,
      "step": 810
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 3.796875,
      "learning_rate": 8e-05,
      "loss": 0.2965,
      "step": 820
    },
    {
      "epoch": 6.64,
      "grad_norm": 12.5625,
      "learning_rate": 8e-05,
      "loss": 0.2505,
      "step": 830
    },
    {
      "epoch": 6.72,
      "grad_norm": 15.3125,
      "learning_rate": 8e-05,
      "loss": 0.2911,
      "step": 840
    },
    {
      "epoch": 6.8,
      "grad_norm": 4.46875,
      "learning_rate": 8e-05,
      "loss": 0.2646,
      "step": 850
    },
    {
      "epoch": 6.88,
      "grad_norm": 13.5,
      "learning_rate": 8e-05,
      "loss": 0.2276,
      "step": 860
    },
    {
      "epoch": 6.96,
      "grad_norm": 3.421875,
      "learning_rate": 8e-05,
      "loss": 0.1909,
      "step": 870
    },
    {
      "epoch": 7.04,
      "grad_norm": 5.5,
      "learning_rate": 8e-05,
      "loss": 0.2293,
      "step": 880
    },
    {
      "epoch": 7.12,
      "grad_norm": 2.21875,
      "learning_rate": 8e-05,
      "loss": 0.1238,
      "step": 890
    },
    {
      "epoch": 7.2,
      "grad_norm": 6.3125,
      "learning_rate": 8e-05,
      "loss": 0.3137,
      "step": 900
    },
    {
      "epoch": 7.28,
      "grad_norm": 12.0,
      "learning_rate": 8e-05,
      "loss": 0.214,
      "step": 910
    },
    {
      "epoch": 7.36,
      "grad_norm": 4.34375,
      "learning_rate": 8e-05,
      "loss": 0.2143,
      "step": 920
    },
    {
      "epoch": 7.44,
      "grad_norm": 2.671875,
      "learning_rate": 8e-05,
      "loss": 0.1859,
      "step": 930
    },
    {
      "epoch": 7.52,
      "grad_norm": 4.625,
      "learning_rate": 8e-05,
      "loss": 0.2029,
      "step": 940
    },
    {
      "epoch": 7.6,
      "grad_norm": 2.078125,
      "learning_rate": 8e-05,
      "loss": 0.206,
      "step": 950
    },
    {
      "epoch": 7.68,
      "grad_norm": 3.203125,
      "learning_rate": 8e-05,
      "loss": 0.1454,
      "step": 960
    },
    {
      "epoch": 7.76,
      "grad_norm": 13.6875,
      "learning_rate": 8e-05,
      "loss": 0.2195,
      "step": 970
    },
    {
      "epoch": 7.84,
      "grad_norm": 7.5625,
      "learning_rate": 8e-05,
      "loss": 0.2111,
      "step": 980
    },
    {
      "epoch": 7.92,
      "grad_norm": 4.34375,
      "learning_rate": 8e-05,
      "loss": 0.2387,
      "step": 990
    },
    {
      "epoch": 8.0,
      "grad_norm": 14.625,
      "learning_rate": 8e-05,
      "loss": 0.2577,
      "step": 1000
    },
    {
      "epoch": 8.0,
      "eval_loss": 6.3644208908081055,
      "eval_runtime": 7.8125,
      "eval_samples_per_second": 64.0,
      "eval_steps_per_second": 64.0,
      "step": 1000
    },
    {
      "epoch": 8.08,
      "grad_norm": 4.46875,
      "learning_rate": 8e-05,
      "loss": 0.1625,
      "step": 1010
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.4375,
      "learning_rate": 8e-05,
      "loss": 0.1557,
      "step": 1020
    },
    {
      "epoch": 8.24,
      "grad_norm": 2.078125,
      "learning_rate": 8e-05,
      "loss": 0.2838,
      "step": 1030
    },
    {
      "epoch": 8.32,
      "grad_norm": 2.3125,
      "learning_rate": 8e-05,
      "loss": 0.1527,
      "step": 1040
    },
    {
      "epoch": 8.4,
      "grad_norm": 11.5,
      "learning_rate": 8e-05,
      "loss": 0.2098,
      "step": 1050
    },
    {
      "epoch": 8.48,
      "grad_norm": 2.21875,
      "learning_rate": 8e-05,
      "loss": 0.2563,
      "step": 1060
    },
    {
      "epoch": 8.56,
      "grad_norm": 2.46875,
      "learning_rate": 8e-05,
      "loss": 0.1923,
      "step": 1070
    },
    {
      "epoch": 8.64,
      "grad_norm": 1.875,
      "learning_rate": 8e-05,
      "loss": 0.179,
      "step": 1080
    },
    {
      "epoch": 8.72,
      "grad_norm": 2.875,
      "learning_rate": 8e-05,
      "loss": 0.1626,
      "step": 1090
    },
    {
      "epoch": 8.8,
      "grad_norm": 8.875,
      "learning_rate": 8e-05,
      "loss": 0.2404,
      "step": 1100
    },
    {
      "epoch": 8.88,
      "grad_norm": 2.015625,
      "learning_rate": 8e-05,
      "loss": 0.1682,
      "step": 1110
    },
    {
      "epoch": 8.96,
      "grad_norm": 2.25,
      "learning_rate": 8e-05,
      "loss": 0.2182,
      "step": 1120
    },
    {
      "epoch": 9.04,
      "grad_norm": 1.9609375,
      "learning_rate": 8e-05,
      "loss": 0.1742,
      "step": 1130
    },
    {
      "epoch": 9.12,
      "grad_norm": 2.859375,
      "learning_rate": 8e-05,
      "loss": 0.1374,
      "step": 1140
    },
    {
      "epoch": 9.2,
      "grad_norm": 2.78125,
      "learning_rate": 8e-05,
      "loss": 0.1037,
      "step": 1150
    },
    {
      "epoch": 9.28,
      "grad_norm": 13.75,
      "learning_rate": 8e-05,
      "loss": 0.1504,
      "step": 1160
    },
    {
      "epoch": 9.36,
      "grad_norm": 15.0625,
      "learning_rate": 8e-05,
      "loss": 0.2686,
      "step": 1170
    },
    {
      "epoch": 9.44,
      "grad_norm": 1.46875,
      "learning_rate": 8e-05,
      "loss": 0.1604,
      "step": 1180
    },
    {
      "epoch": 9.52,
      "grad_norm": 1.6640625,
      "learning_rate": 8e-05,
      "loss": 0.1988,
      "step": 1190
    },
    {
      "epoch": 9.6,
      "grad_norm": 2.25,
      "learning_rate": 8e-05,
      "loss": 0.1162,
      "step": 1200
    },
    {
      "epoch": 9.68,
      "grad_norm": 2.28125,
      "learning_rate": 8e-05,
      "loss": 0.2101,
      "step": 1210
    },
    {
      "epoch": 9.76,
      "grad_norm": 1.546875,
      "learning_rate": 8e-05,
      "loss": 0.1573,
      "step": 1220
    },
    {
      "epoch": 9.84,
      "grad_norm": 1.25,
      "learning_rate": 8e-05,
      "loss": 0.1335,
      "step": 1230
    },
    {
      "epoch": 9.92,
      "grad_norm": 3.28125,
      "learning_rate": 8e-05,
      "loss": 0.2523,
      "step": 1240
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.9296875,
      "learning_rate": 8e-05,
      "loss": 0.1335,
      "step": 1250
    },
    {
      "epoch": 10.08,
      "grad_norm": 1.1171875,
      "learning_rate": 8e-05,
      "loss": 0.125,
      "step": 1260
    },
    {
      "epoch": 10.16,
      "grad_norm": 1.5078125,
      "learning_rate": 8e-05,
      "loss": 0.1788,
      "step": 1270
    },
    {
      "epoch": 10.24,
      "grad_norm": 5.59375,
      "learning_rate": 8e-05,
      "loss": 0.1387,
      "step": 1280
    },
    {
      "epoch": 10.32,
      "grad_norm": 1.7890625,
      "learning_rate": 8e-05,
      "loss": 0.2533,
      "step": 1290
    },
    {
      "epoch": 10.4,
      "grad_norm": 1.1171875,
      "learning_rate": 8e-05,
      "loss": 0.0953,
      "step": 1300
    },
    {
      "epoch": 10.48,
      "grad_norm": 2.734375,
      "learning_rate": 8e-05,
      "loss": 0.3212,
      "step": 1310
    },
    {
      "epoch": 10.56,
      "grad_norm": 3.625,
      "learning_rate": 8e-05,
      "loss": 0.1222,
      "step": 1320
    },
    {
      "epoch": 10.64,
      "grad_norm": 15.4375,
      "learning_rate": 8e-05,
      "loss": 0.2796,
      "step": 1330
    },
    {
      "epoch": 10.72,
      "grad_norm": 14.5,
      "learning_rate": 8e-05,
      "loss": 0.2214,
      "step": 1340
    },
    {
      "epoch": 10.8,
      "grad_norm": 3.15625,
      "learning_rate": 8e-05,
      "loss": 0.0851,
      "step": 1350
    },
    {
      "epoch": 10.88,
      "grad_norm": 3.125,
      "learning_rate": 8e-05,
      "loss": 0.1612,
      "step": 1360
    },
    {
      "epoch": 10.96,
      "grad_norm": 9.0625,
      "learning_rate": 8e-05,
      "loss": 0.1493,
      "step": 1370
    },
    {
      "epoch": 11.04,
      "grad_norm": 2.75,
      "learning_rate": 8e-05,
      "loss": 0.2357,
      "step": 1380
    },
    {
      "epoch": 11.12,
      "grad_norm": 1.8125,
      "learning_rate": 8e-05,
      "loss": 0.1377,
      "step": 1390
    },
    {
      "epoch": 11.2,
      "grad_norm": 3.546875,
      "learning_rate": 8e-05,
      "loss": 0.1757,
      "step": 1400
    },
    {
      "epoch": 11.28,
      "grad_norm": 1.375,
      "learning_rate": 8e-05,
      "loss": 0.1276,
      "step": 1410
    },
    {
      "epoch": 11.36,
      "grad_norm": 12.4375,
      "learning_rate": 8e-05,
      "loss": 0.1677,
      "step": 1420
    },
    {
      "epoch": 11.44,
      "grad_norm": 1.6171875,
      "learning_rate": 8e-05,
      "loss": 0.2043,
      "step": 1430
    },
    {
      "epoch": 11.52,
      "grad_norm": 12.4375,
      "learning_rate": 8e-05,
      "loss": 0.2517,
      "step": 1440
    },
    {
      "epoch": 11.6,
      "grad_norm": 11.875,
      "learning_rate": 8e-05,
      "loss": 0.1798,
      "step": 1450
    },
    {
      "epoch": 11.68,
      "grad_norm": 1.3828125,
      "learning_rate": 8e-05,
      "loss": 0.1363,
      "step": 1460
    },
    {
      "epoch": 11.76,
      "grad_norm": 2.046875,
      "learning_rate": 8e-05,
      "loss": 0.1562,
      "step": 1470
    },
    {
      "epoch": 11.84,
      "grad_norm": 1.984375,
      "learning_rate": 8e-05,
      "loss": 0.1433,
      "step": 1480
    },
    {
      "epoch": 11.92,
      "grad_norm": 2.859375,
      "learning_rate": 8e-05,
      "loss": 0.1044,
      "step": 1490
    },
    {
      "epoch": 12.0,
      "grad_norm": 2.78125,
      "learning_rate": 8e-05,
      "loss": 0.1167,
      "step": 1500
    },
    {
      "epoch": 12.0,
      "eval_loss": 6.880049705505371,
      "eval_runtime": 7.7521,
      "eval_samples_per_second": 64.499,
      "eval_steps_per_second": 64.499,
      "step": 1500
    },
    {
      "epoch": 12.08,
      "grad_norm": 1.78125,
      "learning_rate": 8e-05,
      "loss": 0.086,
      "step": 1510
    },
    {
      "epoch": 12.16,
      "grad_norm": 13.0625,
      "learning_rate": 8e-05,
      "loss": 0.1261,
      "step": 1520
    },
    {
      "epoch": 12.24,
      "grad_norm": 2.625,
      "learning_rate": 8e-05,
      "loss": 0.1597,
      "step": 1530
    },
    {
      "epoch": 12.32,
      "grad_norm": 10.375,
      "learning_rate": 8e-05,
      "loss": 0.2094,
      "step": 1540
    },
    {
      "epoch": 12.4,
      "grad_norm": 11.5625,
      "learning_rate": 8e-05,
      "loss": 0.1402,
      "step": 1550
    },
    {
      "epoch": 12.48,
      "grad_norm": 0.69921875,
      "learning_rate": 8e-05,
      "loss": 0.1614,
      "step": 1560
    },
    {
      "epoch": 12.56,
      "grad_norm": 1.71875,
      "learning_rate": 8e-05,
      "loss": 0.099,
      "step": 1570
    },
    {
      "epoch": 12.64,
      "grad_norm": 11.6875,
      "learning_rate": 8e-05,
      "loss": 0.1913,
      "step": 1580
    },
    {
      "epoch": 12.72,
      "grad_norm": 12.8125,
      "learning_rate": 8e-05,
      "loss": 0.1537,
      "step": 1590
    },
    {
      "epoch": 12.8,
      "grad_norm": 1.5859375,
      "learning_rate": 8e-05,
      "loss": 0.1909,
      "step": 1600
    },
    {
      "epoch": 12.88,
      "grad_norm": 12.9375,
      "learning_rate": 8e-05,
      "loss": 0.1719,
      "step": 1610
    },
    {
      "epoch": 12.96,
      "grad_norm": 2.390625,
      "learning_rate": 8e-05,
      "loss": 0.1278,
      "step": 1620
    },
    {
      "epoch": 13.04,
      "grad_norm": 1.5078125,
      "learning_rate": 8e-05,
      "loss": 0.1048,
      "step": 1630
    },
    {
      "epoch": 13.12,
      "grad_norm": 3.09375,
      "learning_rate": 8e-05,
      "loss": 0.1214,
      "step": 1640
    },
    {
      "epoch": 13.2,
      "grad_norm": 2.40625,
      "learning_rate": 8e-05,
      "loss": 0.1648,
      "step": 1650
    },
    {
      "epoch": 13.28,
      "grad_norm": 3.78125,
      "learning_rate": 8e-05,
      "loss": 0.1426,
      "step": 1660
    },
    {
      "epoch": 13.36,
      "grad_norm": 1.3984375,
      "learning_rate": 8e-05,
      "loss": 0.0681,
      "step": 1670
    },
    {
      "epoch": 13.44,
      "grad_norm": 12.9375,
      "learning_rate": 8e-05,
      "loss": 0.207,
      "step": 1680
    },
    {
      "epoch": 13.52,
      "grad_norm": 1.4140625,
      "learning_rate": 8e-05,
      "loss": 0.1404,
      "step": 1690
    },
    {
      "epoch": 13.6,
      "grad_norm": 1.71875,
      "learning_rate": 8e-05,
      "loss": 0.2456,
      "step": 1700
    },
    {
      "epoch": 13.68,
      "grad_norm": 1.4921875,
      "learning_rate": 8e-05,
      "loss": 0.0857,
      "step": 1710
    },
    {
      "epoch": 13.76,
      "grad_norm": 13.25,
      "learning_rate": 8e-05,
      "loss": 0.2331,
      "step": 1720
    },
    {
      "epoch": 13.84,
      "grad_norm": 11.625,
      "learning_rate": 8e-05,
      "loss": 0.2518,
      "step": 1730
    },
    {
      "epoch": 13.92,
      "grad_norm": 9.4375,
      "learning_rate": 8e-05,
      "loss": 0.139,
      "step": 1740
    },
    {
      "epoch": 14.0,
      "grad_norm": 4.09375,
      "learning_rate": 8e-05,
      "loss": 0.1043,
      "step": 1750
    },
    {
      "epoch": 14.08,
      "grad_norm": 1.6171875,
      "learning_rate": 8e-05,
      "loss": 0.1867,
      "step": 1760
    },
    {
      "epoch": 14.16,
      "grad_norm": 2.75,
      "learning_rate": 8e-05,
      "loss": 0.1933,
      "step": 1770
    },
    {
      "epoch": 14.24,
      "grad_norm": 3.21875,
      "learning_rate": 8e-05,
      "loss": 0.1029,
      "step": 1780
    },
    {
      "epoch": 14.32,
      "grad_norm": 11.0,
      "learning_rate": 8e-05,
      "loss": 0.1679,
      "step": 1790
    },
    {
      "epoch": 14.4,
      "grad_norm": 2.703125,
      "learning_rate": 8e-05,
      "loss": 0.0926,
      "step": 1800
    },
    {
      "epoch": 14.48,
      "grad_norm": 1.984375,
      "learning_rate": 8e-05,
      "loss": 0.1764,
      "step": 1810
    },
    {
      "epoch": 14.56,
      "grad_norm": 1.7265625,
      "learning_rate": 8e-05,
      "loss": 0.2071,
      "step": 1820
    },
    {
      "epoch": 14.64,
      "grad_norm": 1.6953125,
      "learning_rate": 8e-05,
      "loss": 0.1027,
      "step": 1830
    },
    {
      "epoch": 14.72,
      "grad_norm": 2.34375,
      "learning_rate": 8e-05,
      "loss": 0.0807,
      "step": 1840
    },
    {
      "epoch": 14.8,
      "grad_norm": 2.796875,
      "learning_rate": 8e-05,
      "loss": 0.2069,
      "step": 1850
    },
    {
      "epoch": 14.88,
      "grad_norm": 14.0625,
      "learning_rate": 8e-05,
      "loss": 0.1714,
      "step": 1860
    },
    {
      "epoch": 14.96,
      "grad_norm": 12.25,
      "learning_rate": 8e-05,
      "loss": 0.0976,
      "step": 1870
    },
    {
      "epoch": 15.04,
      "grad_norm": 1.3046875,
      "learning_rate": 8e-05,
      "loss": 0.0776,
      "step": 1880
    },
    {
      "epoch": 15.12,
      "grad_norm": 0.9609375,
      "learning_rate": 8e-05,
      "loss": 0.1022,
      "step": 1890
    },
    {
      "epoch": 15.2,
      "grad_norm": 12.8125,
      "learning_rate": 8e-05,
      "loss": 0.1935,
      "step": 1900
    },
    {
      "epoch": 15.28,
      "grad_norm": 10.0625,
      "learning_rate": 8e-05,
      "loss": 0.1253,
      "step": 1910
    },
    {
      "epoch": 15.36,
      "grad_norm": 0.8046875,
      "learning_rate": 8e-05,
      "loss": 0.0874,
      "step": 1920
    },
    {
      "epoch": 15.44,
      "grad_norm": 0.89453125,
      "learning_rate": 8e-05,
      "loss": 0.1495,
      "step": 1930
    },
    {
      "epoch": 15.52,
      "grad_norm": 1.3515625,
      "learning_rate": 8e-05,
      "loss": 0.1414,
      "step": 1940
    },
    {
      "epoch": 15.6,
      "grad_norm": 3.171875,
      "learning_rate": 8e-05,
      "loss": 0.171,
      "step": 1950
    },
    {
      "epoch": 15.68,
      "grad_norm": 1.4453125,
      "learning_rate": 8e-05,
      "loss": 0.0812,
      "step": 1960
    },
    {
      "epoch": 15.76,
      "grad_norm": 14.625,
      "learning_rate": 8e-05,
      "loss": 0.1929,
      "step": 1970
    },
    {
      "epoch": 15.84,
      "grad_norm": 11.9375,
      "learning_rate": 8e-05,
      "loss": 0.1511,
      "step": 1980
    },
    {
      "epoch": 15.92,
      "grad_norm": 2.484375,
      "learning_rate": 8e-05,
      "loss": 0.1548,
      "step": 1990
    },
    {
      "epoch": 16.0,
      "grad_norm": 13.1875,
      "learning_rate": 8e-05,
      "loss": 0.1725,
      "step": 2000
    },
    {
      "epoch": 16.0,
      "eval_loss": 7.8348259925842285,
      "eval_runtime": 7.9081,
      "eval_samples_per_second": 63.227,
      "eval_steps_per_second": 63.227,
      "step": 2000
    },
    {
      "epoch": 16.08,
      "grad_norm": 0.890625,
      "learning_rate": 8e-05,
      "loss": 0.2574,
      "step": 2010
    },
    {
      "epoch": 16.16,
      "grad_norm": 0.87109375,
      "learning_rate": 8e-05,
      "loss": 0.1048,
      "step": 2020
    },
    {
      "epoch": 16.24,
      "grad_norm": 3.59375,
      "learning_rate": 8e-05,
      "loss": 0.1627,
      "step": 2030
    },
    {
      "epoch": 16.32,
      "grad_norm": 10.0625,
      "learning_rate": 8e-05,
      "loss": 0.2,
      "step": 2040
    },
    {
      "epoch": 16.4,
      "grad_norm": 2.171875,
      "learning_rate": 8e-05,
      "loss": 0.1846,
      "step": 2050
    },
    {
      "epoch": 16.48,
      "grad_norm": 2.5625,
      "learning_rate": 8e-05,
      "loss": 0.1296,
      "step": 2060
    },
    {
      "epoch": 16.56,
      "grad_norm": 2.203125,
      "learning_rate": 8e-05,
      "loss": 0.1036,
      "step": 2070
    },
    {
      "epoch": 16.64,
      "grad_norm": 1.8203125,
      "learning_rate": 8e-05,
      "loss": 0.1629,
      "step": 2080
    },
    {
      "epoch": 16.72,
      "grad_norm": 2.0625,
      "learning_rate": 8e-05,
      "loss": 0.1398,
      "step": 2090
    },
    {
      "epoch": 16.8,
      "grad_norm": 11.125,
      "learning_rate": 8e-05,
      "loss": 0.1421,
      "step": 2100
    },
    {
      "epoch": 16.88,
      "grad_norm": 1.484375,
      "learning_rate": 8e-05,
      "loss": 0.0845,
      "step": 2110
    },
    {
      "epoch": 16.96,
      "grad_norm": 1.53125,
      "learning_rate": 8e-05,
      "loss": 0.1685,
      "step": 2120
    },
    {
      "epoch": 17.04,
      "grad_norm": 1.1875,
      "learning_rate": 8e-05,
      "loss": 0.1122,
      "step": 2130
    },
    {
      "epoch": 17.12,
      "grad_norm": 9.8125,
      "learning_rate": 8e-05,
      "loss": 0.1509,
      "step": 2140
    },
    {
      "epoch": 17.2,
      "grad_norm": 6.40625,
      "learning_rate": 8e-05,
      "loss": 0.1171,
      "step": 2150
    },
    {
      "epoch": 17.28,
      "grad_norm": 0.83984375,
      "learning_rate": 8e-05,
      "loss": 0.1652,
      "step": 2160
    },
    {
      "epoch": 17.36,
      "grad_norm": 1.359375,
      "learning_rate": 8e-05,
      "loss": 0.0871,
      "step": 2170
    },
    {
      "epoch": 17.44,
      "grad_norm": 1.96875,
      "learning_rate": 8e-05,
      "loss": 0.2603,
      "step": 2180
    },
    {
      "epoch": 17.52,
      "grad_norm": 13.8125,
      "learning_rate": 8e-05,
      "loss": 0.1249,
      "step": 2190
    },
    {
      "epoch": 17.6,
      "grad_norm": 1.109375,
      "learning_rate": 8e-05,
      "loss": 0.1161,
      "step": 2200
    },
    {
      "epoch": 17.68,
      "grad_norm": 12.8125,
      "learning_rate": 8e-05,
      "loss": 0.1502,
      "step": 2210
    },
    {
      "epoch": 17.76,
      "grad_norm": 15.125,
      "learning_rate": 8e-05,
      "loss": 0.1375,
      "step": 2220
    },
    {
      "epoch": 17.84,
      "grad_norm": 0.87109375,
      "learning_rate": 8e-05,
      "loss": 0.1636,
      "step": 2230
    },
    {
      "epoch": 17.92,
      "grad_norm": 1.890625,
      "learning_rate": 8e-05,
      "loss": 0.1541,
      "step": 2240
    },
    {
      "epoch": 18.0,
      "grad_norm": 3.8125,
      "learning_rate": 8e-05,
      "loss": 0.106,
      "step": 2250
    },
    {
      "epoch": 18.08,
      "grad_norm": 1.2578125,
      "learning_rate": 8e-05,
      "loss": 0.1042,
      "step": 2260
    },
    {
      "epoch": 18.16,
      "grad_norm": 1.265625,
      "learning_rate": 8e-05,
      "loss": 0.2585,
      "step": 2270
    },
    {
      "epoch": 18.24,
      "grad_norm": 1.84375,
      "learning_rate": 8e-05,
      "loss": 0.1033,
      "step": 2280
    },
    {
      "epoch": 18.32,
      "grad_norm": 3.09375,
      "learning_rate": 8e-05,
      "loss": 0.137,
      "step": 2290
    },
    {
      "epoch": 18.4,
      "grad_norm": 1.6875,
      "learning_rate": 8e-05,
      "loss": 0.0679,
      "step": 2300
    },
    {
      "epoch": 18.48,
      "grad_norm": 3.3125,
      "learning_rate": 8e-05,
      "loss": 0.135,
      "step": 2310
    },
    {
      "epoch": 18.56,
      "grad_norm": 1.8046875,
      "learning_rate": 8e-05,
      "loss": 0.1582,
      "step": 2320
    },
    {
      "epoch": 18.64,
      "grad_norm": 14.0625,
      "learning_rate": 8e-05,
      "loss": 0.1295,
      "step": 2330
    },
    {
      "epoch": 18.72,
      "grad_norm": 3.265625,
      "learning_rate": 8e-05,
      "loss": 0.1386,
      "step": 2340
    },
    {
      "epoch": 18.8,
      "grad_norm": 1.265625,
      "learning_rate": 8e-05,
      "loss": 0.1629,
      "step": 2350
    },
    {
      "epoch": 18.88,
      "grad_norm": 1.6796875,
      "learning_rate": 8e-05,
      "loss": 0.0668,
      "step": 2360
    },
    {
      "epoch": 18.96,
      "grad_norm": 1.546875,
      "learning_rate": 8e-05,
      "loss": 0.133,
      "step": 2370
    },
    {
      "epoch": 19.04,
      "grad_norm": 11.5,
      "learning_rate": 8e-05,
      "loss": 0.1276,
      "step": 2380
    },
    {
      "epoch": 19.12,
      "grad_norm": 11.8125,
      "learning_rate": 8e-05,
      "loss": 0.1711,
      "step": 2390
    },
    {
      "epoch": 19.2,
      "grad_norm": 1.90625,
      "learning_rate": 8e-05,
      "loss": 0.1272,
      "step": 2400
    },
    {
      "epoch": 19.28,
      "grad_norm": 2.265625,
      "learning_rate": 8e-05,
      "loss": 0.1748,
      "step": 2410
    },
    {
      "epoch": 19.36,
      "grad_norm": 1.0703125,
      "learning_rate": 8e-05,
      "loss": 0.1185,
      "step": 2420
    },
    {
      "epoch": 19.44,
      "grad_norm": 2.65625,
      "learning_rate": 8e-05,
      "loss": 0.2216,
      "step": 2430
    },
    {
      "epoch": 19.52,
      "grad_norm": 0.91796875,
      "learning_rate": 8e-05,
      "loss": 0.0562,
      "step": 2440
    },
    {
      "epoch": 19.6,
      "grad_norm": 0.984375,
      "learning_rate": 8e-05,
      "loss": 0.1306,
      "step": 2450
    },
    {
      "epoch": 19.68,
      "grad_norm": 2.203125,
      "learning_rate": 8e-05,
      "loss": 0.097,
      "step": 2460
    },
    {
      "epoch": 19.76,
      "grad_norm": 11.5,
      "learning_rate": 8e-05,
      "loss": 0.1722,
      "step": 2470
    },
    {
      "epoch": 19.84,
      "grad_norm": 4.34375,
      "learning_rate": 8e-05,
      "loss": 0.1796,
      "step": 2480
    },
    {
      "epoch": 19.92,
      "grad_norm": 1.109375,
      "learning_rate": 8e-05,
      "loss": 0.1134,
      "step": 2490
    },
    {
      "epoch": 20.0,
      "grad_norm": 12.0,
      "learning_rate": 8e-05,
      "loss": 0.1808,
      "step": 2500
    },
    {
      "epoch": 20.0,
      "eval_loss": 8.226277351379395,
      "eval_runtime": 7.8324,
      "eval_samples_per_second": 63.838,
      "eval_steps_per_second": 63.838,
      "step": 2500
    },
    {
      "epoch": 20.08,
      "grad_norm": 1.171875,
      "learning_rate": 8e-05,
      "loss": 0.086,
      "step": 2510
    },
    {
      "epoch": 20.16,
      "grad_norm": 1.8515625,
      "learning_rate": 8e-05,
      "loss": 0.0983,
      "step": 2520
    },
    {
      "epoch": 20.24,
      "grad_norm": 8.1875,
      "learning_rate": 8e-05,
      "loss": 0.1646,
      "step": 2530
    },
    {
      "epoch": 20.32,
      "grad_norm": 1.328125,
      "learning_rate": 8e-05,
      "loss": 0.1301,
      "step": 2540
    },
    {
      "epoch": 20.4,
      "grad_norm": 0.88671875,
      "learning_rate": 8e-05,
      "loss": 0.1124,
      "step": 2550
    },
    {
      "epoch": 20.48,
      "grad_norm": 0.91015625,
      "learning_rate": 8e-05,
      "loss": 0.1121,
      "step": 2560
    },
    {
      "epoch": 20.56,
      "grad_norm": 1.4453125,
      "learning_rate": 8e-05,
      "loss": 0.0874,
      "step": 2570
    },
    {
      "epoch": 20.64,
      "grad_norm": 12.75,
      "learning_rate": 8e-05,
      "loss": 0.1808,
      "step": 2580
    },
    {
      "epoch": 20.72,
      "grad_norm": 13.75,
      "learning_rate": 8e-05,
      "loss": 0.1845,
      "step": 2590
    },
    {
      "epoch": 20.8,
      "grad_norm": 2.078125,
      "learning_rate": 8e-05,
      "loss": 0.1648,
      "step": 2600
    },
    {
      "epoch": 20.88,
      "grad_norm": 1.328125,
      "learning_rate": 8e-05,
      "loss": 0.1531,
      "step": 2610
    },
    {
      "epoch": 20.96,
      "grad_norm": 0.92578125,
      "learning_rate": 8e-05,
      "loss": 0.0932,
      "step": 2620
    },
    {
      "epoch": 21.04,
      "grad_norm": 0.99609375,
      "learning_rate": 8e-05,
      "loss": 0.1656,
      "step": 2630
    },
    {
      "epoch": 21.12,
      "grad_norm": 1.9921875,
      "learning_rate": 8e-05,
      "loss": 0.115,
      "step": 2640
    },
    {
      "epoch": 21.2,
      "grad_norm": 2.21875,
      "learning_rate": 8e-05,
      "loss": 0.0872,
      "step": 2650
    },
    {
      "epoch": 21.28,
      "grad_norm": 12.0,
      "learning_rate": 8e-05,
      "loss": 0.1692,
      "step": 2660
    },
    {
      "epoch": 21.36,
      "grad_norm": 0.93359375,
      "learning_rate": 8e-05,
      "loss": 0.1691,
      "step": 2670
    },
    {
      "epoch": 21.44,
      "grad_norm": 1.6328125,
      "learning_rate": 8e-05,
      "loss": 0.101,
      "step": 2680
    },
    {
      "epoch": 21.52,
      "grad_norm": 0.8046875,
      "learning_rate": 8e-05,
      "loss": 0.1124,
      "step": 2690
    },
    {
      "epoch": 21.6,
      "grad_norm": 1.234375,
      "learning_rate": 8e-05,
      "loss": 0.1218,
      "step": 2700
    },
    {
      "epoch": 21.68,
      "grad_norm": 2.421875,
      "learning_rate": 8e-05,
      "loss": 0.1999,
      "step": 2710
    },
    {
      "epoch": 21.76,
      "grad_norm": 2.21875,
      "learning_rate": 8e-05,
      "loss": 0.1771,
      "step": 2720
    },
    {
      "epoch": 21.84,
      "grad_norm": 1.5,
      "learning_rate": 8e-05,
      "loss": 0.1491,
      "step": 2730
    },
    {
      "epoch": 21.92,
      "grad_norm": 13.625,
      "learning_rate": 8e-05,
      "loss": 0.1392,
      "step": 2740
    },
    {
      "epoch": 22.0,
      "grad_norm": 1.1953125,
      "learning_rate": 8e-05,
      "loss": 0.1795,
      "step": 2750
    },
    {
      "epoch": 22.08,
      "grad_norm": 7.96875,
      "learning_rate": 8e-05,
      "loss": 0.0752,
      "step": 2760
    },
    {
      "epoch": 22.16,
      "grad_norm": 2.453125,
      "learning_rate": 8e-05,
      "loss": 0.2007,
      "step": 2770
    },
    {
      "epoch": 22.24,
      "grad_norm": 1.21875,
      "learning_rate": 8e-05,
      "loss": 0.1152,
      "step": 2780
    },
    {
      "epoch": 22.32,
      "grad_norm": 1.140625,
      "learning_rate": 8e-05,
      "loss": 0.1126,
      "step": 2790
    },
    {
      "epoch": 22.4,
      "grad_norm": 5.1875,
      "learning_rate": 8e-05,
      "loss": 0.1385,
      "step": 2800
    },
    {
      "epoch": 22.48,
      "grad_norm": 1.7578125,
      "learning_rate": 8e-05,
      "loss": 0.1497,
      "step": 2810
    },
    {
      "epoch": 22.56,
      "grad_norm": 10.4375,
      "learning_rate": 8e-05,
      "loss": 0.2022,
      "step": 2820
    },
    {
      "epoch": 22.64,
      "grad_norm": 1.1171875,
      "learning_rate": 8e-05,
      "loss": 0.137,
      "step": 2830
    },
    {
      "epoch": 22.72,
      "grad_norm": 2.28125,
      "learning_rate": 8e-05,
      "loss": 0.164,
      "step": 2840
    },
    {
      "epoch": 22.8,
      "grad_norm": 13.625,
      "learning_rate": 8e-05,
      "loss": 0.1336,
      "step": 2850
    },
    {
      "epoch": 22.88,
      "grad_norm": 1.140625,
      "learning_rate": 8e-05,
      "loss": 0.1674,
      "step": 2860
    },
    {
      "epoch": 22.96,
      "grad_norm": 2.484375,
      "learning_rate": 8e-05,
      "loss": 0.1656,
      "step": 2870
    },
    {
      "epoch": 23.04,
      "grad_norm": 1.53125,
      "learning_rate": 8e-05,
      "loss": 0.0927,
      "step": 2880
    },
    {
      "epoch": 23.12,
      "grad_norm": 1.296875,
      "learning_rate": 8e-05,
      "loss": 0.0651,
      "step": 2890
    },
    {
      "epoch": 23.2,
      "grad_norm": 1.2890625,
      "learning_rate": 8e-05,
      "loss": 0.1061,
      "step": 2900
    },
    {
      "epoch": 23.28,
      "grad_norm": 0.9140625,
      "learning_rate": 8e-05,
      "loss": 0.1063,
      "step": 2910
    },
    {
      "epoch": 23.36,
      "grad_norm": 1.46875,
      "learning_rate": 8e-05,
      "loss": 0.1106,
      "step": 2920
    },
    {
      "epoch": 23.44,
      "grad_norm": 14.5625,
      "learning_rate": 8e-05,
      "loss": 0.1363,
      "step": 2930
    },
    {
      "epoch": 23.52,
      "grad_norm": 12.875,
      "learning_rate": 8e-05,
      "loss": 0.1571,
      "step": 2940
    },
    {
      "epoch": 23.6,
      "grad_norm": 13.75,
      "learning_rate": 8e-05,
      "loss": 0.1521,
      "step": 2950
    },
    {
      "epoch": 23.68,
      "grad_norm": 1.2265625,
      "learning_rate": 8e-05,
      "loss": 0.135,
      "step": 2960
    },
    {
      "epoch": 23.76,
      "grad_norm": 1.6953125,
      "learning_rate": 8e-05,
      "loss": 0.0755,
      "step": 2970
    },
    {
      "epoch": 23.84,
      "grad_norm": 1.8828125,
      "learning_rate": 8e-05,
      "loss": 0.1685,
      "step": 2980
    },
    {
      "epoch": 23.92,
      "grad_norm": 11.3125,
      "learning_rate": 8e-05,
      "loss": 0.1397,
      "step": 2990
    },
    {
      "epoch": 24.0,
      "grad_norm": 11.1875,
      "learning_rate": 8e-05,
      "loss": 0.2037,
      "step": 3000
    },
    {
      "epoch": 24.0,
      "eval_loss": 8.587909698486328,
      "eval_runtime": 7.7605,
      "eval_samples_per_second": 64.429,
      "eval_steps_per_second": 64.429,
      "step": 3000
    },
    {
      "epoch": 24.08,
      "grad_norm": 9.8125,
      "learning_rate": 8e-05,
      "loss": 0.1701,
      "step": 3010
    },
    {
      "epoch": 24.16,
      "grad_norm": 0.875,
      "learning_rate": 8e-05,
      "loss": 0.1105,
      "step": 3020
    },
    {
      "epoch": 24.24,
      "grad_norm": 1.3203125,
      "learning_rate": 8e-05,
      "loss": 0.0702,
      "step": 3030
    },
    {
      "epoch": 24.32,
      "grad_norm": 1.1796875,
      "learning_rate": 8e-05,
      "loss": 0.0835,
      "step": 3040
    },
    {
      "epoch": 24.4,
      "grad_norm": 13.125,
      "learning_rate": 8e-05,
      "loss": 0.2214,
      "step": 3050
    },
    {
      "epoch": 24.48,
      "grad_norm": 1.203125,
      "learning_rate": 8e-05,
      "loss": 0.1072,
      "step": 3060
    },
    {
      "epoch": 24.56,
      "grad_norm": 0.921875,
      "learning_rate": 8e-05,
      "loss": 0.0679,
      "step": 3070
    },
    {
      "epoch": 24.64,
      "grad_norm": 14.25,
      "learning_rate": 8e-05,
      "loss": 0.1306,
      "step": 3080
    },
    {
      "epoch": 24.72,
      "grad_norm": 1.8984375,
      "learning_rate": 8e-05,
      "loss": 0.1081,
      "step": 3090
    },
    {
      "epoch": 24.8,
      "grad_norm": 0.95703125,
      "learning_rate": 8e-05,
      "loss": 0.1175,
      "step": 3100
    },
    {
      "epoch": 24.88,
      "grad_norm": 0.890625,
      "learning_rate": 8e-05,
      "loss": 0.0938,
      "step": 3110
    },
    {
      "epoch": 24.96,
      "grad_norm": 1.0,
      "learning_rate": 8e-05,
      "loss": 0.1949,
      "step": 3120
    },
    {
      "epoch": 25.04,
      "grad_norm": 11.5,
      "learning_rate": 8e-05,
      "loss": 0.182,
      "step": 3130
    },
    {
      "epoch": 25.12,
      "grad_norm": 11.6875,
      "learning_rate": 8e-05,
      "loss": 0.2094,
      "step": 3140
    },
    {
      "epoch": 25.2,
      "grad_norm": 1.5390625,
      "learning_rate": 8e-05,
      "loss": 0.0803,
      "step": 3150
    },
    {
      "epoch": 25.28,
      "grad_norm": 0.88671875,
      "learning_rate": 8e-05,
      "loss": 0.0533,
      "step": 3160
    },
    {
      "epoch": 25.36,
      "grad_norm": 0.96484375,
      "learning_rate": 8e-05,
      "loss": 0.0916,
      "step": 3170
    },
    {
      "epoch": 25.44,
      "grad_norm": 12.5625,
      "learning_rate": 8e-05,
      "loss": 0.0968,
      "step": 3180
    },
    {
      "epoch": 25.52,
      "grad_norm": 1.3671875,
      "learning_rate": 8e-05,
      "loss": 0.0717,
      "step": 3190
    },
    {
      "epoch": 25.6,
      "grad_norm": 1.1015625,
      "learning_rate": 8e-05,
      "loss": 0.1266,
      "step": 3200
    },
    {
      "epoch": 25.68,
      "grad_norm": 1.546875,
      "learning_rate": 8e-05,
      "loss": 0.2108,
      "step": 3210
    },
    {
      "epoch": 25.76,
      "grad_norm": 14.5,
      "learning_rate": 8e-05,
      "loss": 0.1072,
      "step": 3220
    },
    {
      "epoch": 25.84,
      "grad_norm": 1.0234375,
      "learning_rate": 8e-05,
      "loss": 0.1371,
      "step": 3230
    },
    {
      "epoch": 25.92,
      "grad_norm": 1.8828125,
      "learning_rate": 8e-05,
      "loss": 0.2164,
      "step": 3240
    },
    {
      "epoch": 26.0,
      "grad_norm": 1.0390625,
      "learning_rate": 8e-05,
      "loss": 0.1007,
      "step": 3250
    },
    {
      "epoch": 26.08,
      "grad_norm": 0.87109375,
      "learning_rate": 8e-05,
      "loss": 0.0686,
      "step": 3260
    },
    {
      "epoch": 26.16,
      "grad_norm": 0.69921875,
      "learning_rate": 8e-05,
      "loss": 0.1272,
      "step": 3270
    },
    {
      "epoch": 26.24,
      "grad_norm": 2.25,
      "learning_rate": 8e-05,
      "loss": 0.1243,
      "step": 3280
    },
    {
      "epoch": 26.32,
      "grad_norm": 1.234375,
      "learning_rate": 8e-05,
      "loss": 0.0689,
      "step": 3290
    },
    {
      "epoch": 26.4,
      "grad_norm": 1.1328125,
      "learning_rate": 8e-05,
      "loss": 0.1712,
      "step": 3300
    },
    {
      "epoch": 26.48,
      "grad_norm": 1.578125,
      "learning_rate": 8e-05,
      "loss": 0.1643,
      "step": 3310
    },
    {
      "epoch": 26.56,
      "grad_norm": 1.3515625,
      "learning_rate": 8e-05,
      "loss": 0.172,
      "step": 3320
    },
    {
      "epoch": 26.64,
      "grad_norm": 1.234375,
      "learning_rate": 8e-05,
      "loss": 0.1448,
      "step": 3330
    },
    {
      "epoch": 26.72,
      "grad_norm": 0.890625,
      "learning_rate": 8e-05,
      "loss": 0.154,
      "step": 3340
    },
    {
      "epoch": 26.8,
      "grad_norm": 0.93359375,
      "learning_rate": 8e-05,
      "loss": 0.1253,
      "step": 3350
    },
    {
      "epoch": 26.88,
      "grad_norm": 12.0,
      "learning_rate": 8e-05,
      "loss": 0.1776,
      "step": 3360
    },
    {
      "epoch": 26.96,
      "grad_norm": 2.09375,
      "learning_rate": 8e-05,
      "loss": 0.0906,
      "step": 3370
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 125,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2857020049096704.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}