{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.96,
  "eval_steps": 50,
  "global_step": 3000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 0.2421875,
      "learning_rate": 5e-06,
      "loss": 0.7,
      "step": 10
    },
    {
      "epoch": 0.0064,
      "grad_norm": 0.30859375,
      "learning_rate": 1e-05,
      "loss": 0.6988,
      "step": 20
    },
    {
      "epoch": 0.0096,
      "grad_norm": 0.28515625,
      "learning_rate": 1.5e-05,
      "loss": 0.6977,
      "step": 30
    },
    {
      "epoch": 0.0128,
      "grad_norm": 0.244140625,
      "learning_rate": 2e-05,
      "loss": 0.6969,
      "step": 40
    },
    {
      "epoch": 0.016,
      "grad_norm": 0.259765625,
      "learning_rate": 2.5e-05,
      "loss": 0.6941,
      "step": 50
    },
    {
      "epoch": 0.016,
      "eval_loss": 0.6939687728881836,
      "eval_runtime": 192.7405,
      "eval_samples_per_second": 2.594,
      "eval_steps_per_second": 0.083,
      "step": 50
    },
    {
      "epoch": 0.0192,
      "grad_norm": 0.232421875,
      "learning_rate": 3e-05,
      "loss": 0.6871,
      "step": 60
    },
    {
      "epoch": 0.0224,
      "grad_norm": 0.302734375,
      "learning_rate": 3.5e-05,
      "loss": 0.6828,
      "step": 70
    },
    {
      "epoch": 0.0256,
      "grad_norm": 0.8203125,
      "learning_rate": 4e-05,
      "loss": 0.6687,
      "step": 80
    },
    {
      "epoch": 0.0288,
      "grad_norm": 1.2421875,
      "learning_rate": 4.5e-05,
      "loss": 0.6309,
      "step": 90
    },
    {
      "epoch": 0.032,
      "grad_norm": 2.03125,
      "learning_rate": 5e-05,
      "loss": 0.5781,
      "step": 100
    },
    {
      "epoch": 0.032,
      "eval_loss": 0.5389062762260437,
      "eval_runtime": 108.3608,
      "eval_samples_per_second": 4.614,
      "eval_steps_per_second": 0.148,
      "step": 100
    },
    {
      "epoch": 0.0352,
      "grad_norm": 2.328125,
      "learning_rate": 4.9834710743801654e-05,
      "loss": 0.4818,
      "step": 110
    },
    {
      "epoch": 0.0384,
      "grad_norm": 2.265625,
      "learning_rate": 4.9669421487603305e-05,
      "loss": 0.4291,
      "step": 120
    },
    {
      "epoch": 0.0416,
      "grad_norm": 2.203125,
      "learning_rate": 4.9504132231404956e-05,
      "loss": 0.3615,
      "step": 130
    },
    {
      "epoch": 0.0448,
      "grad_norm": 2.515625,
      "learning_rate": 4.9338842975206614e-05,
      "loss": 0.3408,
      "step": 140
    },
    {
      "epoch": 0.048,
      "grad_norm": 1.515625,
      "learning_rate": 4.917355371900827e-05,
      "loss": 0.306,
      "step": 150
    },
    {
      "epoch": 0.048,
      "eval_loss": 0.2862031161785126,
      "eval_runtime": 110.0641,
      "eval_samples_per_second": 4.543,
      "eval_steps_per_second": 0.145,
      "step": 150
    },
    {
      "epoch": 0.0512,
      "grad_norm": 1.7265625,
      "learning_rate": 4.900826446280992e-05,
      "loss": 0.2904,
      "step": 160
    },
    {
      "epoch": 0.0544,
      "grad_norm": 1.5703125,
      "learning_rate": 4.8842975206611575e-05,
      "loss": 0.2648,
      "step": 170
    },
    {
      "epoch": 0.0576,
      "grad_norm": 1.6875,
      "learning_rate": 4.8677685950413226e-05,
      "loss": 0.2373,
      "step": 180
    },
    {
      "epoch": 0.0608,
      "grad_norm": 1.0859375,
      "learning_rate": 4.851239669421488e-05,
      "loss": 0.1749,
      "step": 190
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.8671875,
      "learning_rate": 4.834710743801653e-05,
      "loss": 0.1556,
      "step": 200
    },
    {
      "epoch": 0.064,
      "eval_loss": 0.16703906655311584,
      "eval_runtime": 109.9961,
      "eval_samples_per_second": 4.546,
      "eval_steps_per_second": 0.145,
      "step": 200
    },
    {
      "epoch": 0.0672,
      "grad_norm": 2.5625,
      "learning_rate": 4.8181818181818186e-05,
      "loss": 0.1382,
      "step": 210
    },
    {
      "epoch": 0.0704,
      "grad_norm": 1.859375,
      "learning_rate": 4.801652892561984e-05,
      "loss": 0.125,
      "step": 220
    },
    {
      "epoch": 0.0736,
      "grad_norm": 1.3515625,
      "learning_rate": 4.785123966942149e-05,
      "loss": 0.1729,
      "step": 230
    },
    {
      "epoch": 0.0768,
      "grad_norm": 1.1953125,
      "learning_rate": 4.768595041322314e-05,
      "loss": 0.1252,
      "step": 240
    },
    {
      "epoch": 0.08,
      "grad_norm": 1.0078125,
      "learning_rate": 4.75206611570248e-05,
      "loss": 0.1108,
      "step": 250
    },
    {
      "epoch": 0.08,
      "eval_loss": 0.13416016101837158,
      "eval_runtime": 110.9679,
      "eval_samples_per_second": 4.506,
      "eval_steps_per_second": 0.144,
      "step": 250
    },
    {
      "epoch": 0.0832,
      "grad_norm": 1.7890625,
      "learning_rate": 4.735537190082645e-05,
      "loss": 0.1502,
      "step": 260
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.66796875,
      "learning_rate": 4.71900826446281e-05,
      "loss": 0.1291,
      "step": 270
    },
    {
      "epoch": 0.0896,
      "grad_norm": 2.546875,
      "learning_rate": 4.702479338842976e-05,
      "loss": 0.1326,
      "step": 280
    },
    {
      "epoch": 0.0928,
      "grad_norm": 1.7265625,
      "learning_rate": 4.685950413223141e-05,
      "loss": 0.1374,
      "step": 290
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.15625,
      "learning_rate": 4.669421487603306e-05,
      "loss": 0.1081,
      "step": 300
    },
    {
      "epoch": 0.096,
      "eval_loss": 0.12271875143051147,
      "eval_runtime": 110.6168,
      "eval_samples_per_second": 4.52,
      "eval_steps_per_second": 0.145,
      "step": 300
    },
    {
      "epoch": 0.0992,
      "grad_norm": 1.140625,
      "learning_rate": 4.652892561983471e-05,
      "loss": 0.1266,
      "step": 310
    },
    {
      "epoch": 0.1024,
      "grad_norm": 1.2265625,
      "learning_rate": 4.636363636363636e-05,
      "loss": 0.1368,
      "step": 320
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.7109375,
      "learning_rate": 4.6198347107438014e-05,
      "loss": 0.1064,
      "step": 330
    },
    {
      "epoch": 0.1088,
      "grad_norm": 1.4375,
      "learning_rate": 4.603305785123967e-05,
      "loss": 0.1233,
      "step": 340
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.6328125,
      "learning_rate": 4.586776859504133e-05,
      "loss": 0.0881,
      "step": 350
    },
    {
      "epoch": 0.112,
      "eval_loss": 0.1180742159485817,
      "eval_runtime": 109.5127,
      "eval_samples_per_second": 4.566,
      "eval_steps_per_second": 0.146,
      "step": 350
    },
    {
      "epoch": 0.1152,
      "grad_norm": 1.109375,
      "learning_rate": 4.570247933884298e-05,
      "loss": 0.1136,
      "step": 360
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.89453125,
      "learning_rate": 4.553719008264463e-05,
      "loss": 0.1338,
      "step": 370
    },
    {
      "epoch": 0.1216,
      "grad_norm": 2.65625,
      "learning_rate": 4.5371900826446284e-05,
      "loss": 0.1636,
      "step": 380
    },
    {
      "epoch": 0.1248,
      "grad_norm": 2.375,
      "learning_rate": 4.5206611570247935e-05,
      "loss": 0.1067,
      "step": 390
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.078125,
      "learning_rate": 4.504132231404959e-05,
      "loss": 0.1287,
      "step": 400
    },
    {
      "epoch": 0.128,
      "eval_loss": 0.1135859340429306,
      "eval_runtime": 115.1853,
      "eval_samples_per_second": 4.341,
      "eval_steps_per_second": 0.139,
      "step": 400
    },
    {
      "epoch": 0.1312,
      "grad_norm": 1.5859375,
      "learning_rate": 4.487603305785124e-05,
      "loss": 0.122,
      "step": 410
    },
    {
      "epoch": 0.1344,
      "grad_norm": 1.6796875,
      "learning_rate": 4.4710743801652896e-05,
      "loss": 0.1232,
      "step": 420
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.99609375,
      "learning_rate": 4.454545454545455e-05,
      "loss": 0.11,
      "step": 430
    },
    {
      "epoch": 0.1408,
      "grad_norm": 1.6328125,
      "learning_rate": 4.43801652892562e-05,
      "loss": 0.0937,
      "step": 440
    },
    {
      "epoch": 0.144,
      "grad_norm": 1.8671875,
      "learning_rate": 4.4214876033057856e-05,
      "loss": 0.1162,
      "step": 450
    },
    {
      "epoch": 0.144,
      "eval_loss": 0.10819531232118607,
      "eval_runtime": 110.4172,
      "eval_samples_per_second": 4.528,
      "eval_steps_per_second": 0.145,
      "step": 450
    },
    {
      "epoch": 0.1472,
      "grad_norm": 3.78125,
      "learning_rate": 4.404958677685951e-05,
      "loss": 0.1171,
      "step": 460
    },
    {
      "epoch": 0.1504,
      "grad_norm": 1.71875,
      "learning_rate": 4.388429752066116e-05,
      "loss": 0.1032,
      "step": 470
    },
    {
      "epoch": 0.1536,
      "grad_norm": 2.796875,
      "learning_rate": 4.371900826446281e-05,
      "loss": 0.1217,
      "step": 480
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.703125,
      "learning_rate": 4.355371900826447e-05,
      "loss": 0.1197,
      "step": 490
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.734375,
      "learning_rate": 4.338842975206612e-05,
      "loss": 0.0878,
      "step": 500
    },
    {
      "epoch": 0.16,
      "eval_loss": 0.10691406577825546,
      "eval_runtime": 110.9181,
      "eval_samples_per_second": 4.508,
      "eval_steps_per_second": 0.144,
      "step": 500
    },
    {
      "epoch": 0.1632,
      "grad_norm": 1.3125,
      "learning_rate": 4.322314049586777e-05,
      "loss": 0.1102,
      "step": 510
    },
    {
      "epoch": 0.1664,
      "grad_norm": 1.171875,
      "learning_rate": 4.305785123966942e-05,
      "loss": 0.0982,
      "step": 520
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.90625,
      "learning_rate": 4.289256198347107e-05,
      "loss": 0.1096,
      "step": 530
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.93359375,
      "learning_rate": 4.2727272727272724e-05,
      "loss": 0.1269,
      "step": 540
    },
    {
      "epoch": 0.176,
      "grad_norm": 1.609375,
      "learning_rate": 4.256198347107438e-05,
      "loss": 0.0994,
      "step": 550
    },
    {
      "epoch": 0.176,
      "eval_loss": 0.10447265952825546,
      "eval_runtime": 110.3941,
      "eval_samples_per_second": 4.529,
      "eval_steps_per_second": 0.145,
      "step": 550
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.703125,
      "learning_rate": 4.239669421487604e-05,
      "loss": 0.1118,
      "step": 560
    },
    {
      "epoch": 0.1824,
      "grad_norm": 1.234375,
      "learning_rate": 4.223140495867769e-05,
      "loss": 0.098,
      "step": 570
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.9375,
      "learning_rate": 4.206611570247934e-05,
      "loss": 0.0811,
      "step": 580
    },
    {
      "epoch": 0.1888,
      "grad_norm": 1.3984375,
      "learning_rate": 4.1900826446280994e-05,
      "loss": 0.1208,
      "step": 590
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.34375,
      "learning_rate": 4.1735537190082645e-05,
      "loss": 0.1275,
      "step": 600
    },
    {
      "epoch": 0.192,
      "eval_loss": 0.10120312869548798,
      "eval_runtime": 109.2003,
      "eval_samples_per_second": 4.579,
      "eval_steps_per_second": 0.147,
      "step": 600
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.5859375,
      "learning_rate": 4.1570247933884296e-05,
      "loss": 0.0872,
      "step": 610
    },
    {
      "epoch": 0.1984,
      "grad_norm": 1.0390625,
      "learning_rate": 4.1404958677685954e-05,
      "loss": 0.1025,
      "step": 620
    },
    {
      "epoch": 0.2016,
      "grad_norm": 1.0546875,
      "learning_rate": 4.1239669421487606e-05,
      "loss": 0.0932,
      "step": 630
    },
    {
      "epoch": 0.2048,
      "grad_norm": 1.3046875,
      "learning_rate": 4.107438016528926e-05,
      "loss": 0.0838,
      "step": 640
    },
    {
      "epoch": 0.208,
      "grad_norm": 1.0703125,
      "learning_rate": 4.0909090909090915e-05,
      "loss": 0.0921,
      "step": 650
    },
    {
      "epoch": 0.208,
      "eval_loss": 0.10271484404802322,
      "eval_runtime": 110.3072,
      "eval_samples_per_second": 4.533,
      "eval_steps_per_second": 0.145,
      "step": 650
    },
    {
      "epoch": 0.2112,
      "grad_norm": 1.453125,
      "learning_rate": 4.0743801652892566e-05,
      "loss": 0.0842,
      "step": 660
    },
    {
      "epoch": 0.2144,
      "grad_norm": 1.9609375,
      "learning_rate": 4.057851239669422e-05,
      "loss": 0.0969,
      "step": 670
    },
    {
      "epoch": 0.2176,
      "grad_norm": 1.5859375,
      "learning_rate": 4.041322314049587e-05,
      "loss": 0.1037,
      "step": 680
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.96875,
      "learning_rate": 4.024793388429752e-05,
      "loss": 0.0872,
      "step": 690
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.703125,
      "learning_rate": 4.008264462809918e-05,
      "loss": 0.0722,
      "step": 700
    },
    {
      "epoch": 0.224,
      "eval_loss": 0.10465820133686066,
      "eval_runtime": 109.5448,
      "eval_samples_per_second": 4.564,
      "eval_steps_per_second": 0.146,
      "step": 700
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.66015625,
      "learning_rate": 3.991735537190083e-05,
      "loss": 0.0926,
      "step": 710
    },
    {
      "epoch": 0.2304,
      "grad_norm": 3.234375,
      "learning_rate": 3.975206611570248e-05,
      "loss": 0.0945,
      "step": 720
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.5390625,
      "learning_rate": 3.958677685950413e-05,
      "loss": 0.076,
      "step": 730
    },
    {
      "epoch": 0.2368,
      "grad_norm": 1.0234375,
      "learning_rate": 3.942148760330578e-05,
      "loss": 0.0959,
      "step": 740
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.455078125,
      "learning_rate": 3.925619834710744e-05,
      "loss": 0.0916,
      "step": 750
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.10287109017372131,
      "eval_runtime": 122.5858,
      "eval_samples_per_second": 4.079,
      "eval_steps_per_second": 0.131,
      "step": 750
    },
    {
      "epoch": 0.2432,
      "grad_norm": 1.609375,
      "learning_rate": 3.909090909090909e-05,
      "loss": 0.0891,
      "step": 760
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.8359375,
      "learning_rate": 3.892561983471075e-05,
      "loss": 0.0954,
      "step": 770
    },
    {
      "epoch": 0.2496,
      "grad_norm": 1.203125,
      "learning_rate": 3.87603305785124e-05,
      "loss": 0.11,
      "step": 780
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.365234375,
      "learning_rate": 3.859504132231405e-05,
      "loss": 0.067,
      "step": 790
    },
    {
      "epoch": 0.256,
      "grad_norm": 1.1171875,
      "learning_rate": 3.8429752066115703e-05,
      "loss": 0.0909,
      "step": 800
    },
    {
      "epoch": 0.256,
      "eval_loss": 0.09786719083786011,
      "eval_runtime": 112.8442,
      "eval_samples_per_second": 4.431,
      "eval_steps_per_second": 0.142,
      "step": 800
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.61328125,
      "learning_rate": 3.8264462809917355e-05,
      "loss": 0.0801,
      "step": 810
    },
    {
      "epoch": 0.2624,
      "grad_norm": 1.9296875,
      "learning_rate": 3.8099173553719006e-05,
      "loss": 0.1088,
      "step": 820
    },
    {
      "epoch": 0.2656,
      "grad_norm": 1.7421875,
      "learning_rate": 3.7933884297520664e-05,
      "loss": 0.08,
      "step": 830
    },
    {
      "epoch": 0.2688,
      "grad_norm": 1.1875,
      "learning_rate": 3.7768595041322315e-05,
      "loss": 0.1055,
      "step": 840
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.85546875,
      "learning_rate": 3.760330578512397e-05,
      "loss": 0.0951,
      "step": 850
    },
    {
      "epoch": 0.272,
      "eval_loss": 0.09604492038488388,
      "eval_runtime": 109.2932,
      "eval_samples_per_second": 4.575,
      "eval_steps_per_second": 0.146,
      "step": 850
    },
    {
      "epoch": 0.2752,
      "grad_norm": 1.7265625,
      "learning_rate": 3.7438016528925624e-05,
      "loss": 0.116,
      "step": 860
    },
    {
      "epoch": 0.2784,
      "grad_norm": 1.640625,
      "learning_rate": 3.7272727272727276e-05,
      "loss": 0.0771,
      "step": 870
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.92578125,
      "learning_rate": 3.710743801652893e-05,
      "loss": 0.08,
      "step": 880
    },
    {
      "epoch": 0.2848,
      "grad_norm": 1.7890625,
      "learning_rate": 3.694214876033058e-05,
      "loss": 0.105,
      "step": 890
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.60546875,
      "learning_rate": 3.6776859504132236e-05,
      "loss": 0.0686,
      "step": 900
    },
    {
      "epoch": 0.288,
      "eval_loss": 0.08956640958786011,
      "eval_runtime": 109.9587,
      "eval_samples_per_second": 4.547,
      "eval_steps_per_second": 0.146,
      "step": 900
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.484375,
      "learning_rate": 3.661157024793389e-05,
      "loss": 0.0867,
      "step": 910
    },
    {
      "epoch": 0.2944,
      "grad_norm": 1.25,
      "learning_rate": 3.644628099173554e-05,
      "loss": 0.0975,
      "step": 920
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.625,
      "learning_rate": 3.628099173553719e-05,
      "loss": 0.0905,
      "step": 930
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.44140625,
      "learning_rate": 3.611570247933884e-05,
      "loss": 0.0983,
      "step": 940
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.51953125,
      "learning_rate": 3.59504132231405e-05,
      "loss": 0.0772,
      "step": 950
    },
    {
      "epoch": 0.304,
      "eval_loss": 0.09039648622274399,
      "eval_runtime": 111.9644,
      "eval_samples_per_second": 4.466,
      "eval_steps_per_second": 0.143,
      "step": 950
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.828125,
      "learning_rate": 3.578512396694215e-05,
      "loss": 0.0907,
      "step": 960
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.84765625,
      "learning_rate": 3.56198347107438e-05,
      "loss": 0.0939,
      "step": 970
    },
    {
      "epoch": 0.3136,
      "grad_norm": 2.203125,
      "learning_rate": 3.545454545454546e-05,
      "loss": 0.0984,
      "step": 980
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.494140625,
      "learning_rate": 3.528925619834711e-05,
      "loss": 0.0924,
      "step": 990
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.91015625,
      "learning_rate": 3.512396694214876e-05,
      "loss": 0.0761,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.09162402153015137,
      "eval_runtime": 112.9796,
      "eval_samples_per_second": 4.426,
      "eval_steps_per_second": 0.142,
      "step": 1000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 0.7265625,
      "learning_rate": 3.495867768595041e-05,
      "loss": 0.0845,
      "step": 1010
    },
    {
      "epoch": 0.3264,
      "grad_norm": 1.0859375,
      "learning_rate": 3.4793388429752064e-05,
      "loss": 0.0852,
      "step": 1020
    },
    {
      "epoch": 0.3296,
      "grad_norm": 1.40625,
      "learning_rate": 3.4628099173553716e-05,
      "loss": 0.1208,
      "step": 1030
    },
    {
      "epoch": 0.3328,
      "grad_norm": 1.3203125,
      "learning_rate": 3.4462809917355374e-05,
      "loss": 0.0807,
      "step": 1040
    },
    {
      "epoch": 0.336,
      "grad_norm": 1.515625,
      "learning_rate": 3.429752066115703e-05,
      "loss": 0.0756,
      "step": 1050
    },
    {
      "epoch": 0.336,
      "eval_loss": 0.09120605140924454,
      "eval_runtime": 111.5564,
      "eval_samples_per_second": 4.482,
      "eval_steps_per_second": 0.143,
      "step": 1050
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.84765625,
      "learning_rate": 3.413223140495868e-05,
      "loss": 0.1092,
      "step": 1060
    },
    {
      "epoch": 0.3424,
      "grad_norm": 2.671875,
      "learning_rate": 3.3966942148760334e-05,
      "loss": 0.1217,
      "step": 1070
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.9921875,
      "learning_rate": 3.3801652892561985e-05,
      "loss": 0.0921,
      "step": 1080
    },
    {
      "epoch": 0.3488,
      "grad_norm": 1.0703125,
      "learning_rate": 3.3636363636363636e-05,
      "loss": 0.1023,
      "step": 1090
    },
    {
      "epoch": 0.352,
      "grad_norm": 2.34375,
      "learning_rate": 3.347107438016529e-05,
      "loss": 0.107,
      "step": 1100
    },
    {
      "epoch": 0.352,
      "eval_loss": 0.08991601318120956,
      "eval_runtime": 109.5083,
      "eval_samples_per_second": 4.566,
      "eval_steps_per_second": 0.146,
      "step": 1100
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.98828125,
      "learning_rate": 3.3305785123966946e-05,
      "loss": 0.0888,
      "step": 1110
    },
    {
      "epoch": 0.3584,
      "grad_norm": 2.640625,
      "learning_rate": 3.31404958677686e-05,
      "loss": 0.0896,
      "step": 1120
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.4375,
      "learning_rate": 3.297520661157025e-05,
      "loss": 0.072,
      "step": 1130
    },
    {
      "epoch": 0.3648,
      "grad_norm": 1.515625,
      "learning_rate": 3.28099173553719e-05,
      "loss": 0.1046,
      "step": 1140
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.6640625,
      "learning_rate": 3.264462809917356e-05,
      "loss": 0.0791,
      "step": 1150
    },
    {
      "epoch": 0.368,
      "eval_loss": 0.08706250041723251,
      "eval_runtime": 108.4302,
      "eval_samples_per_second": 4.611,
      "eval_steps_per_second": 0.148,
      "step": 1150
    },
    {
      "epoch": 0.3712,
      "grad_norm": 1.4140625,
      "learning_rate": 3.247933884297521e-05,
      "loss": 0.0791,
      "step": 1160
    },
    {
      "epoch": 0.3744,
      "grad_norm": 1.734375,
      "learning_rate": 3.231404958677686e-05,
      "loss": 0.0854,
      "step": 1170
    },
    {
      "epoch": 0.3776,
      "grad_norm": 1.09375,
      "learning_rate": 3.214876033057852e-05,
      "loss": 0.0912,
      "step": 1180
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.48828125,
      "learning_rate": 3.198347107438017e-05,
      "loss": 0.1034,
      "step": 1190
    },
    {
      "epoch": 0.384,
      "grad_norm": 1.3515625,
      "learning_rate": 3.181818181818182e-05,
      "loss": 0.0863,
      "step": 1200
    },
    {
      "epoch": 0.384,
      "eval_loss": 0.08619336038827896,
      "eval_runtime": 108.0491,
      "eval_samples_per_second": 4.628,
      "eval_steps_per_second": 0.148,
      "step": 1200
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.80078125,
      "learning_rate": 3.165289256198347e-05,
      "loss": 0.0904,
      "step": 1210
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.7890625,
      "learning_rate": 3.148760330578512e-05,
      "loss": 0.1262,
      "step": 1220
    },
    {
      "epoch": 0.3936,
      "grad_norm": 1.21875,
      "learning_rate": 3.1322314049586774e-05,
      "loss": 0.08,
      "step": 1230
    },
    {
      "epoch": 0.3968,
      "grad_norm": 1.609375,
      "learning_rate": 3.1157024793388425e-05,
      "loss": 0.0861,
      "step": 1240
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.28125,
      "learning_rate": 3.099173553719008e-05,
      "loss": 0.0803,
      "step": 1250
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.08419531583786011,
      "eval_runtime": 110.4296,
      "eval_samples_per_second": 4.528,
      "eval_steps_per_second": 0.145,
      "step": 1250
    },
    {
      "epoch": 0.4032,
      "grad_norm": 1.140625,
      "learning_rate": 3.082644628099174e-05,
      "loss": 0.0889,
      "step": 1260
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.7890625,
      "learning_rate": 3.066115702479339e-05,
      "loss": 0.0879,
      "step": 1270
    },
    {
      "epoch": 0.4096,
      "grad_norm": 1.015625,
      "learning_rate": 3.0495867768595044e-05,
      "loss": 0.0776,
      "step": 1280
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.8984375,
      "learning_rate": 3.0330578512396695e-05,
      "loss": 0.0903,
      "step": 1290
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.7578125,
      "learning_rate": 3.016528925619835e-05,
      "loss": 0.0877,
      "step": 1300
    },
    {
      "epoch": 0.416,
      "eval_loss": 0.08515820652246475,
      "eval_runtime": 107.6996,
      "eval_samples_per_second": 4.643,
      "eval_steps_per_second": 0.149,
      "step": 1300
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.84765625,
      "learning_rate": 3e-05,
      "loss": 0.0912,
      "step": 1310
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.671875,
      "learning_rate": 2.9834710743801652e-05,
      "loss": 0.0634,
      "step": 1320
    },
    {
      "epoch": 0.4256,
      "grad_norm": 1.03125,
      "learning_rate": 2.9669421487603307e-05,
      "loss": 0.076,
      "step": 1330
    },
    {
      "epoch": 0.4288,
      "grad_norm": 1.7421875,
      "learning_rate": 2.950413223140496e-05,
      "loss": 0.0911,
      "step": 1340
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.859375,
      "learning_rate": 2.9338842975206616e-05,
      "loss": 0.0804,
      "step": 1350
    },
    {
      "epoch": 0.432,
      "eval_loss": 0.0840449184179306,
      "eval_runtime": 106.9192,
      "eval_samples_per_second": 4.676,
      "eval_steps_per_second": 0.15,
      "step": 1350
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.74609375,
      "learning_rate": 2.9173553719008267e-05,
      "loss": 0.0779,
      "step": 1360
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.93359375,
      "learning_rate": 2.9008264462809918e-05,
      "loss": 0.0749,
      "step": 1370
    },
    {
      "epoch": 0.4416,
      "grad_norm": 1.359375,
      "learning_rate": 2.8842975206611573e-05,
      "loss": 0.0772,
      "step": 1380
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.86328125,
      "learning_rate": 2.8677685950413224e-05,
      "loss": 0.0853,
      "step": 1390
    },
    {
      "epoch": 0.448,
      "grad_norm": 1.5390625,
      "learning_rate": 2.8512396694214875e-05,
      "loss": 0.1136,
      "step": 1400
    },
    {
      "epoch": 0.448,
      "eval_loss": 0.0832226574420929,
      "eval_runtime": 109.8641,
      "eval_samples_per_second": 4.551,
      "eval_steps_per_second": 0.146,
      "step": 1400
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.59765625,
      "learning_rate": 2.834710743801653e-05,
      "loss": 0.0655,
      "step": 1410
    },
    {
      "epoch": 0.4544,
      "grad_norm": 2.03125,
      "learning_rate": 2.818181818181818e-05,
      "loss": 0.0952,
      "step": 1420
    },
    {
      "epoch": 0.4576,
      "grad_norm": 1.1796875,
      "learning_rate": 2.8016528925619832e-05,
      "loss": 0.0812,
      "step": 1430
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.54296875,
      "learning_rate": 2.785123966942149e-05,
      "loss": 0.0649,
      "step": 1440
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.671875,
      "learning_rate": 2.7685950413223145e-05,
      "loss": 0.0799,
      "step": 1450
    },
    {
      "epoch": 0.464,
      "eval_loss": 0.08352148532867432,
      "eval_runtime": 109.4479,
      "eval_samples_per_second": 4.568,
      "eval_steps_per_second": 0.146,
      "step": 1450
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.62890625,
      "learning_rate": 2.7520661157024796e-05,
      "loss": 0.0699,
      "step": 1460
    },
    {
      "epoch": 0.4704,
      "grad_norm": 1.734375,
      "learning_rate": 2.7355371900826447e-05,
      "loss": 0.0688,
      "step": 1470
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.93359375,
      "learning_rate": 2.7190082644628102e-05,
      "loss": 0.085,
      "step": 1480
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.490234375,
      "learning_rate": 2.7024793388429753e-05,
      "loss": 0.0874,
      "step": 1490
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.703125,
      "learning_rate": 2.6859504132231405e-05,
      "loss": 0.0896,
      "step": 1500
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.08368749916553497,
      "eval_runtime": 110.4491,
      "eval_samples_per_second": 4.527,
      "eval_steps_per_second": 0.145,
      "step": 1500
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.828125,
      "learning_rate": 2.669421487603306e-05,
      "loss": 0.1036,
      "step": 1510
    },
    {
      "epoch": 0.4864,
      "grad_norm": 1.78125,
      "learning_rate": 2.652892561983471e-05,
      "loss": 0.0917,
      "step": 1520
    },
    {
      "epoch": 0.4896,
      "grad_norm": 1.1328125,
      "learning_rate": 2.636363636363636e-05,
      "loss": 0.0985,
      "step": 1530
    },
    {
      "epoch": 0.4928,
      "grad_norm": 1.171875,
      "learning_rate": 2.619834710743802e-05,
      "loss": 0.0881,
      "step": 1540
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.7734375,
      "learning_rate": 2.6033057851239674e-05,
      "loss": 0.078,
      "step": 1550
    },
    {
      "epoch": 0.496,
      "eval_loss": 0.08341991901397705,
      "eval_runtime": 109.2192,
      "eval_samples_per_second": 4.578,
      "eval_steps_per_second": 0.146,
      "step": 1550
    },
    {
      "epoch": 0.4992,
      "grad_norm": 1.2265625,
      "learning_rate": 2.5867768595041325e-05,
      "loss": 0.0784,
      "step": 1560
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.71484375,
      "learning_rate": 2.5702479338842977e-05,
      "loss": 0.0748,
      "step": 1570
    },
    {
      "epoch": 0.5056,
      "grad_norm": 1.203125,
      "learning_rate": 2.553719008264463e-05,
      "loss": 0.0776,
      "step": 1580
    },
    {
      "epoch": 0.5088,
      "grad_norm": 1.40625,
      "learning_rate": 2.5371900826446283e-05,
      "loss": 0.0927,
      "step": 1590
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.6640625,
      "learning_rate": 2.5206611570247934e-05,
      "loss": 0.0878,
      "step": 1600
    },
    {
      "epoch": 0.512,
      "eval_loss": 0.08232617378234863,
      "eval_runtime": 110.112,
      "eval_samples_per_second": 4.541,
      "eval_steps_per_second": 0.145,
      "step": 1600
    },
    {
      "epoch": 0.5152,
      "grad_norm": 1.8125,
      "learning_rate": 2.504132231404959e-05,
      "loss": 0.0929,
      "step": 1610
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.75,
      "learning_rate": 2.4876033057851243e-05,
      "loss": 0.0796,
      "step": 1620
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.9921875,
      "learning_rate": 2.4710743801652894e-05,
      "loss": 0.0932,
      "step": 1630
    },
    {
      "epoch": 0.5248,
      "grad_norm": 1.0078125,
      "learning_rate": 2.4545454545454545e-05,
      "loss": 0.1084,
      "step": 1640
    },
    {
      "epoch": 0.528,
      "grad_norm": 1.03125,
      "learning_rate": 2.43801652892562e-05,
      "loss": 0.0763,
      "step": 1650
    },
    {
      "epoch": 0.528,
      "eval_loss": 0.08151757717132568,
      "eval_runtime": 111.3314,
      "eval_samples_per_second": 4.491,
      "eval_steps_per_second": 0.144,
      "step": 1650
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.859375,
      "learning_rate": 2.421487603305785e-05,
      "loss": 0.062,
      "step": 1660
    },
    {
      "epoch": 0.5344,
      "grad_norm": 1.0859375,
      "learning_rate": 2.4049586776859506e-05,
      "loss": 0.0628,
      "step": 1670
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.59375,
      "learning_rate": 2.3884297520661157e-05,
      "loss": 0.0908,
      "step": 1680
    },
    {
      "epoch": 0.5408,
      "grad_norm": 1.3515625,
      "learning_rate": 2.3719008264462812e-05,
      "loss": 0.0927,
      "step": 1690
    },
    {
      "epoch": 0.544,
      "grad_norm": 1.109375,
      "learning_rate": 2.3553719008264463e-05,
      "loss": 0.0958,
      "step": 1700
    },
    {
      "epoch": 0.544,
      "eval_loss": 0.08200781047344208,
      "eval_runtime": 114.049,
      "eval_samples_per_second": 4.384,
      "eval_steps_per_second": 0.14,
      "step": 1700
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.53125,
      "learning_rate": 2.3388429752066114e-05,
      "loss": 0.0716,
      "step": 1710
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.421875,
      "learning_rate": 2.3223140495867772e-05,
      "loss": 0.0715,
      "step": 1720
    },
    {
      "epoch": 0.5536,
      "grad_norm": 2.703125,
      "learning_rate": 2.3057851239669423e-05,
      "loss": 0.0941,
      "step": 1730
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.89453125,
      "learning_rate": 2.2892561983471075e-05,
      "loss": 0.0785,
      "step": 1740
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.76953125,
      "learning_rate": 2.272727272727273e-05,
      "loss": 0.0857,
      "step": 1750
    },
    {
      "epoch": 0.56,
      "eval_loss": 0.08109960705041885,
      "eval_runtime": 112.2299,
      "eval_samples_per_second": 4.455,
      "eval_steps_per_second": 0.143,
      "step": 1750
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.953125,
      "learning_rate": 2.256198347107438e-05,
      "loss": 0.0601,
      "step": 1760
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.322265625,
      "learning_rate": 2.2396694214876035e-05,
      "loss": 0.0866,
      "step": 1770
    },
    {
      "epoch": 0.5696,
      "grad_norm": 1.015625,
      "learning_rate": 2.2231404958677686e-05,
      "loss": 0.0708,
      "step": 1780
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.72265625,
      "learning_rate": 2.206611570247934e-05,
      "loss": 0.0762,
      "step": 1790
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.5703125,
      "learning_rate": 2.1900826446280992e-05,
      "loss": 0.0831,
      "step": 1800
    },
    {
      "epoch": 0.576,
      "eval_loss": 0.07973046600818634,
      "eval_runtime": 108.6444,
      "eval_samples_per_second": 4.602,
      "eval_steps_per_second": 0.147,
      "step": 1800
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.87109375,
      "learning_rate": 2.1735537190082643e-05,
      "loss": 0.0677,
      "step": 1810
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.8203125,
      "learning_rate": 2.1570247933884298e-05,
      "loss": 0.0738,
      "step": 1820
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.58203125,
      "learning_rate": 2.1404958677685953e-05,
      "loss": 0.064,
      "step": 1830
    },
    {
      "epoch": 0.5888,
      "grad_norm": 1.3359375,
      "learning_rate": 2.1239669421487604e-05,
      "loss": 0.0777,
      "step": 1840
    },
    {
      "epoch": 0.592,
      "grad_norm": 1.0078125,
      "learning_rate": 2.1074380165289255e-05,
      "loss": 0.0792,
      "step": 1850
    },
    {
      "epoch": 0.592,
      "eval_loss": 0.08046093583106995,
      "eval_runtime": 110.6488,
      "eval_samples_per_second": 4.519,
      "eval_steps_per_second": 0.145,
      "step": 1850
    },
    {
      "epoch": 0.5952,
      "grad_norm": 1.28125,
      "learning_rate": 2.090909090909091e-05,
      "loss": 0.0778,
      "step": 1860
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.8125,
      "learning_rate": 2.0743801652892564e-05,
      "loss": 0.0831,
      "step": 1870
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.90234375,
      "learning_rate": 2.0578512396694216e-05,
      "loss": 0.0721,
      "step": 1880
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.69140625,
      "learning_rate": 2.0413223140495867e-05,
      "loss": 0.0661,
      "step": 1890
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.21484375,
      "learning_rate": 2.024793388429752e-05,
      "loss": 0.0725,
      "step": 1900
    },
    {
      "epoch": 0.608,
      "eval_loss": 0.08086328208446503,
      "eval_runtime": 108.5455,
      "eval_samples_per_second": 4.606,
      "eval_steps_per_second": 0.147,
      "step": 1900
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.5,
      "learning_rate": 2.0082644628099173e-05,
      "loss": 0.0813,
      "step": 1910
    },
    {
      "epoch": 0.6144,
      "grad_norm": 1.2109375,
      "learning_rate": 1.9917355371900827e-05,
      "loss": 0.0724,
      "step": 1920
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.921875,
      "learning_rate": 1.9752066115702482e-05,
      "loss": 0.0585,
      "step": 1930
    },
    {
      "epoch": 0.6208,
      "grad_norm": 1.890625,
      "learning_rate": 1.9586776859504133e-05,
      "loss": 0.0928,
      "step": 1940
    },
    {
      "epoch": 0.624,
      "grad_norm": 1.0546875,
      "learning_rate": 1.9421487603305784e-05,
      "loss": 0.0795,
      "step": 1950
    },
    {
      "epoch": 0.624,
      "eval_loss": 0.08059374988079071,
      "eval_runtime": 107.6528,
      "eval_samples_per_second": 4.645,
      "eval_steps_per_second": 0.149,
      "step": 1950
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.765625,
      "learning_rate": 1.925619834710744e-05,
      "loss": 0.0675,
      "step": 1960
    },
    {
      "epoch": 0.6304,
      "grad_norm": 1.6328125,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 0.0896,
      "step": 1970
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.76953125,
      "learning_rate": 1.8925619834710745e-05,
      "loss": 0.075,
      "step": 1980
    },
    {
      "epoch": 0.6368,
      "grad_norm": 1.7890625,
      "learning_rate": 1.8760330578512396e-05,
      "loss": 0.0808,
      "step": 1990
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.21875,
      "learning_rate": 1.859504132231405e-05,
      "loss": 0.0774,
      "step": 2000
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.07988671958446503,
      "eval_runtime": 107.6118,
      "eval_samples_per_second": 4.646,
      "eval_steps_per_second": 0.149,
      "step": 2000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.63671875,
      "learning_rate": 1.8429752066115705e-05,
      "loss": 0.0795,
      "step": 2010
    },
    {
      "epoch": 0.6464,
      "grad_norm": 1.4140625,
      "learning_rate": 1.8264462809917356e-05,
      "loss": 0.1115,
      "step": 2020
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.8203125,
      "learning_rate": 1.8099173553719008e-05,
      "loss": 0.079,
      "step": 2030
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.28515625,
      "learning_rate": 1.7933884297520662e-05,
      "loss": 0.0646,
      "step": 2040
    },
    {
      "epoch": 0.656,
      "grad_norm": 1.25,
      "learning_rate": 1.7768595041322314e-05,
      "loss": 0.0846,
      "step": 2050
    },
    {
      "epoch": 0.656,
      "eval_loss": 0.07978320121765137,
      "eval_runtime": 107.8393,
      "eval_samples_per_second": 4.637,
      "eval_steps_per_second": 0.148,
      "step": 2050
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.55859375,
      "learning_rate": 1.7603305785123968e-05,
      "loss": 0.0949,
      "step": 2060
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.6171875,
      "learning_rate": 1.7438016528925623e-05,
      "loss": 0.0822,
      "step": 2070
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.75390625,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 0.0666,
      "step": 2080
    },
    {
      "epoch": 0.6688,
      "grad_norm": 1.1796875,
      "learning_rate": 1.7107438016528925e-05,
      "loss": 0.0745,
      "step": 2090
    },
    {
      "epoch": 0.672,
      "grad_norm": 1.2265625,
      "learning_rate": 1.694214876033058e-05,
      "loss": 0.1087,
      "step": 2100
    },
    {
      "epoch": 0.672,
      "eval_loss": 0.07949023693799973,
      "eval_runtime": 112.9895,
      "eval_samples_per_second": 4.425,
      "eval_steps_per_second": 0.142,
      "step": 2100
    },
    {
      "epoch": 0.6752,
      "grad_norm": 1.953125,
      "learning_rate": 1.6776859504132234e-05,
      "loss": 0.0917,
      "step": 2110
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.56640625,
      "learning_rate": 1.6611570247933886e-05,
      "loss": 0.075,
      "step": 2120
    },
    {
      "epoch": 0.6816,
      "grad_norm": 1.296875,
      "learning_rate": 1.6446280991735537e-05,
      "loss": 0.0788,
      "step": 2130
    },
    {
      "epoch": 0.6848,
      "grad_norm": 1.203125,
      "learning_rate": 1.628099173553719e-05,
      "loss": 0.0821,
      "step": 2140
    },
    {
      "epoch": 0.688,
      "grad_norm": 1.296875,
      "learning_rate": 1.6115702479338843e-05,
      "loss": 0.0836,
      "step": 2150
    },
    {
      "epoch": 0.688,
      "eval_loss": 0.0802304670214653,
      "eval_runtime": 109.4516,
      "eval_samples_per_second": 4.568,
      "eval_steps_per_second": 0.146,
      "step": 2150
    },
    {
      "epoch": 0.6912,
      "grad_norm": 1.21875,
      "learning_rate": 1.5950413223140497e-05,
      "loss": 0.0845,
      "step": 2160
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.69921875,
      "learning_rate": 1.578512396694215e-05,
      "loss": 0.0701,
      "step": 2170
    },
    {
      "epoch": 0.6976,
      "grad_norm": 1.5078125,
      "learning_rate": 1.5619834710743803e-05,
      "loss": 0.0789,
      "step": 2180
    },
    {
      "epoch": 0.7008,
      "grad_norm": 1.1015625,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.0611,
      "step": 2190
    },
    {
      "epoch": 0.704,
      "grad_norm": 1.578125,
      "learning_rate": 1.5289256198347106e-05,
      "loss": 0.0782,
      "step": 2200
    },
    {
      "epoch": 0.704,
      "eval_loss": 0.07987890392541885,
      "eval_runtime": 108.1627,
      "eval_samples_per_second": 4.623,
      "eval_steps_per_second": 0.148,
      "step": 2200
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.69921875,
      "learning_rate": 1.5123966942148762e-05,
      "loss": 0.0692,
      "step": 2210
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.9140625,
      "learning_rate": 1.4958677685950415e-05,
      "loss": 0.0656,
      "step": 2220
    },
    {
      "epoch": 0.7136,
      "grad_norm": 1.2890625,
      "learning_rate": 1.4793388429752068e-05,
      "loss": 0.0854,
      "step": 2230
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.5703125,
      "learning_rate": 1.4628099173553719e-05,
      "loss": 0.0798,
      "step": 2240
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.458984375,
      "learning_rate": 1.4462809917355372e-05,
      "loss": 0.0885,
      "step": 2250
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.07889453321695328,
      "eval_runtime": 110.4069,
      "eval_samples_per_second": 4.529,
      "eval_steps_per_second": 0.145,
      "step": 2250
    },
    {
      "epoch": 0.7232,
      "grad_norm": 1.234375,
      "learning_rate": 1.4297520661157027e-05,
      "loss": 0.0813,
      "step": 2260
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.76171875,
      "learning_rate": 1.413223140495868e-05,
      "loss": 0.0615,
      "step": 2270
    },
    {
      "epoch": 0.7296,
      "grad_norm": 1.0390625,
      "learning_rate": 1.396694214876033e-05,
      "loss": 0.0773,
      "step": 2280
    },
    {
      "epoch": 0.7328,
      "grad_norm": 1.1484375,
      "learning_rate": 1.3801652892561984e-05,
      "loss": 0.0854,
      "step": 2290
    },
    {
      "epoch": 0.736,
      "grad_norm": 1.390625,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.0926,
      "step": 2300
    },
    {
      "epoch": 0.736,
      "eval_loss": 0.07876367121934891,
      "eval_runtime": 120.6799,
      "eval_samples_per_second": 4.143,
      "eval_steps_per_second": 0.133,
      "step": 2300
    },
    {
      "epoch": 0.7392,
      "grad_norm": 1.046875,
      "learning_rate": 1.3471074380165291e-05,
      "loss": 0.0992,
      "step": 2310
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.44140625,
      "learning_rate": 1.3305785123966944e-05,
      "loss": 0.0629,
      "step": 2320
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.7421875,
      "learning_rate": 1.3140495867768595e-05,
      "loss": 0.0799,
      "step": 2330
    },
    {
      "epoch": 0.7488,
      "grad_norm": 1.203125,
      "learning_rate": 1.2975206611570248e-05,
      "loss": 0.092,
      "step": 2340
    },
    {
      "epoch": 0.752,
      "grad_norm": 1.34375,
      "learning_rate": 1.2809917355371901e-05,
      "loss": 0.064,
      "step": 2350
    },
    {
      "epoch": 0.752,
      "eval_loss": 0.07899804413318634,
      "eval_runtime": 110.2904,
      "eval_samples_per_second": 4.533,
      "eval_steps_per_second": 0.145,
      "step": 2350
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.6875,
      "learning_rate": 1.2644628099173556e-05,
      "loss": 0.0833,
      "step": 2360
    },
    {
      "epoch": 0.7584,
      "grad_norm": 2.015625,
      "learning_rate": 1.2479338842975207e-05,
      "loss": 0.0841,
      "step": 2370
    },
    {
      "epoch": 0.7616,
      "grad_norm": 1.1875,
      "learning_rate": 1.231404958677686e-05,
      "loss": 0.0798,
      "step": 2380
    },
    {
      "epoch": 0.7648,
      "grad_norm": 1.7421875,
      "learning_rate": 1.2148760330578513e-05,
      "loss": 0.0956,
      "step": 2390
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.99609375,
      "learning_rate": 1.1983471074380166e-05,
      "loss": 0.0798,
      "step": 2400
    },
    {
      "epoch": 0.768,
      "eval_loss": 0.07969531416893005,
      "eval_runtime": 110.6153,
      "eval_samples_per_second": 4.52,
      "eval_steps_per_second": 0.145,
      "step": 2400
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.5078125,
      "learning_rate": 1.1818181818181819e-05,
      "loss": 0.0711,
      "step": 2410
    },
    {
      "epoch": 0.7744,
      "grad_norm": 1.4765625,
      "learning_rate": 1.1652892561983472e-05,
      "loss": 0.0708,
      "step": 2420
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.9609375,
      "learning_rate": 1.1487603305785125e-05,
      "loss": 0.0656,
      "step": 2430
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.71484375,
      "learning_rate": 1.1322314049586777e-05,
      "loss": 0.0723,
      "step": 2440
    },
    {
      "epoch": 0.784,
      "grad_norm": 1.09375,
      "learning_rate": 1.115702479338843e-05,
      "loss": 0.0789,
      "step": 2450
    },
    {
      "epoch": 0.784,
      "eval_loss": 0.0796191394329071,
      "eval_runtime": 110.6578,
      "eval_samples_per_second": 4.518,
      "eval_steps_per_second": 0.145,
      "step": 2450
    },
    {
      "epoch": 0.7872,
      "grad_norm": 1.046875,
      "learning_rate": 1.0991735537190083e-05,
      "loss": 0.0755,
      "step": 2460
    },
    {
      "epoch": 0.7904,
      "grad_norm": 1.140625,
      "learning_rate": 1.0826446280991736e-05,
      "loss": 0.0833,
      "step": 2470
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.5625,
      "learning_rate": 1.0661157024793389e-05,
      "loss": 0.1063,
      "step": 2480
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.48046875,
      "learning_rate": 1.0495867768595042e-05,
      "loss": 0.0747,
      "step": 2490
    },
    {
      "epoch": 0.8,
      "grad_norm": 1.390625,
      "learning_rate": 1.0330578512396695e-05,
      "loss": 0.0997,
      "step": 2500
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.0799550786614418,
      "eval_runtime": 106.8244,
      "eval_samples_per_second": 4.681,
      "eval_steps_per_second": 0.15,
      "step": 2500
    },
    {
      "epoch": 0.8032,
      "grad_norm": 1.046875,
      "learning_rate": 1.0165289256198348e-05,
      "loss": 0.0737,
      "step": 2510
    },
    {
      "epoch": 0.8064,
      "grad_norm": 1.1328125,
      "learning_rate": 1e-05,
      "loss": 0.099,
      "step": 2520
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.58203125,
      "learning_rate": 9.834710743801654e-06,
      "loss": 0.0668,
      "step": 2530
    },
    {
      "epoch": 0.8128,
      "grad_norm": 1.140625,
      "learning_rate": 9.669421487603305e-06,
      "loss": 0.0772,
      "step": 2540
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.7734375,
      "learning_rate": 9.50413223140496e-06,
      "loss": 0.0755,
      "step": 2550
    },
    {
      "epoch": 0.816,
      "eval_loss": 0.07922851294279099,
      "eval_runtime": 106.4516,
      "eval_samples_per_second": 4.697,
      "eval_steps_per_second": 0.15,
      "step": 2550
    },
    {
      "epoch": 0.8192,
      "grad_norm": 2.125,
      "learning_rate": 9.338842975206613e-06,
      "loss": 0.0735,
      "step": 2560
    },
    {
      "epoch": 0.8224,
      "grad_norm": 1.7265625,
      "learning_rate": 9.173553719008265e-06,
      "loss": 0.0879,
      "step": 2570
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.2578125,
      "learning_rate": 9.008264462809918e-06,
      "loss": 0.075,
      "step": 2580
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.9375,
      "learning_rate": 8.84297520661157e-06,
      "loss": 0.0718,
      "step": 2590
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.828125,
      "learning_rate": 8.677685950413224e-06,
      "loss": 0.069,
      "step": 2600
    },
    {
      "epoch": 0.832,
      "eval_loss": 0.07956640422344208,
      "eval_runtime": 116.0737,
      "eval_samples_per_second": 4.308,
      "eval_steps_per_second": 0.138,
      "step": 2600
    },
    {
      "epoch": 0.8352,
      "grad_norm": 1.7109375,
      "learning_rate": 8.512396694214875e-06,
      "loss": 0.1006,
      "step": 2610
    },
    {
      "epoch": 0.8384,
      "grad_norm": 1.34375,
      "learning_rate": 8.34710743801653e-06,
      "loss": 0.0771,
      "step": 2620
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.9375,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.0668,
      "step": 2630
    },
    {
      "epoch": 0.8448,
      "grad_norm": 1.1875,
      "learning_rate": 8.016528925619834e-06,
      "loss": 0.0663,
      "step": 2640
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.466796875,
      "learning_rate": 7.851239669421489e-06,
      "loss": 0.0687,
      "step": 2650
    },
    {
      "epoch": 0.848,
      "eval_loss": 0.07871288806200027,
      "eval_runtime": 108.1878,
      "eval_samples_per_second": 4.622,
      "eval_steps_per_second": 0.148,
      "step": 2650
    },
    {
      "epoch": 0.8512,
      "grad_norm": 1.0546875,
      "learning_rate": 7.68595041322314e-06,
      "loss": 0.0899,
      "step": 2660
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.9921875,
      "learning_rate": 7.520661157024795e-06,
      "loss": 0.0735,
      "step": 2670
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.416015625,
      "learning_rate": 7.355371900826447e-06,
      "loss": 0.0905,
      "step": 2680
    },
    {
      "epoch": 0.8608,
      "grad_norm": 1.2109375,
      "learning_rate": 7.190082644628099e-06,
      "loss": 0.0664,
      "step": 2690
    },
    {
      "epoch": 0.864,
      "grad_norm": 1.09375,
      "learning_rate": 7.0247933884297525e-06,
      "loss": 0.0791,
      "step": 2700
    },
    {
      "epoch": 0.864,
      "eval_loss": 0.07861328125,
      "eval_runtime": 109.1401,
      "eval_samples_per_second": 4.581,
      "eval_steps_per_second": 0.147,
      "step": 2700
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.470703125,
      "learning_rate": 6.859504132231405e-06,
      "loss": 0.0588,
      "step": 2710
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.4453125,
      "learning_rate": 6.694214876033058e-06,
      "loss": 0.1022,
      "step": 2720
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.77734375,
      "learning_rate": 6.528925619834711e-06,
      "loss": 0.0645,
      "step": 2730
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.79296875,
      "learning_rate": 6.363636363636363e-06,
      "loss": 0.0688,
      "step": 2740
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.0,
      "learning_rate": 6.198347107438017e-06,
      "loss": 0.0763,
      "step": 2750
    },
    {
      "epoch": 0.88,
      "eval_loss": 0.07813085615634918,
      "eval_runtime": 107.2687,
      "eval_samples_per_second": 4.661,
      "eval_steps_per_second": 0.149,
      "step": 2750
    },
    {
      "epoch": 0.8832,
      "grad_norm": 1.0625,
      "learning_rate": 6.03305785123967e-06,
      "loss": 0.0619,
      "step": 2760
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.81640625,
      "learning_rate": 5.867768595041322e-06,
      "loss": 0.0861,
      "step": 2770
    },
    {
      "epoch": 0.8896,
      "grad_norm": 1.140625,
      "learning_rate": 5.702479338842975e-06,
      "loss": 0.0573,
      "step": 2780
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.6015625,
      "learning_rate": 5.537190082644628e-06,
      "loss": 0.0546,
      "step": 2790
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.408203125,
      "learning_rate": 5.371900826446282e-06,
      "loss": 0.0656,
      "step": 2800
    },
    {
      "epoch": 0.896,
      "eval_loss": 0.0781230479478836,
      "eval_runtime": 120.1101,
      "eval_samples_per_second": 4.163,
      "eval_steps_per_second": 0.133,
      "step": 2800
    },
    {
      "epoch": 0.8992,
      "grad_norm": 1.4609375,
      "learning_rate": 5.206611570247935e-06,
      "loss": 0.0634,
      "step": 2810
    },
    {
      "epoch": 0.9024,
      "grad_norm": 1.515625,
      "learning_rate": 5.041322314049587e-06,
      "loss": 0.093,
      "step": 2820
    },
    {
      "epoch": 0.9056,
      "grad_norm": 1.8359375,
      "learning_rate": 4.87603305785124e-06,
      "loss": 0.0624,
      "step": 2830
    },
    {
      "epoch": 0.9088,
      "grad_norm": 1.0859375,
      "learning_rate": 4.710743801652893e-06,
      "loss": 0.0911,
      "step": 2840
    },
    {
      "epoch": 0.912,
      "grad_norm": 1.171875,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.074,
      "step": 2850
    },
    {
      "epoch": 0.912,
      "eval_loss": 0.07803711295127869,
      "eval_runtime": 109.0375,
      "eval_samples_per_second": 4.586,
      "eval_steps_per_second": 0.147,
      "step": 2850
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.62890625,
      "learning_rate": 4.3801652892561984e-06,
      "loss": 0.0734,
      "step": 2860
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.5703125,
      "learning_rate": 4.214876033057851e-06,
      "loss": 0.0521,
      "step": 2870
    },
    {
      "epoch": 0.9216,
      "grad_norm": 1.453125,
      "learning_rate": 4.049586776859504e-06,
      "loss": 0.0843,
      "step": 2880
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.302734375,
      "learning_rate": 3.884297520661157e-06,
      "loss": 0.0775,
      "step": 2890
    },
    {
      "epoch": 0.928,
      "grad_norm": 1.1328125,
      "learning_rate": 3.71900826446281e-06,
      "loss": 0.0974,
      "step": 2900
    },
    {
      "epoch": 0.928,
      "eval_loss": 0.07862695306539536,
      "eval_runtime": 109.8662,
      "eval_samples_per_second": 4.551,
      "eval_steps_per_second": 0.146,
      "step": 2900
    },
    {
      "epoch": 0.9312,
      "grad_norm": 1.3515625,
      "learning_rate": 3.553719008264463e-06,
      "loss": 0.0611,
      "step": 2910
    },
    {
      "epoch": 0.9344,
      "grad_norm": 1.0078125,
      "learning_rate": 3.3884297520661155e-06,
      "loss": 0.0697,
      "step": 2920
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.56640625,
      "learning_rate": 3.2231404958677685e-06,
      "loss": 0.076,
      "step": 2930
    },
    {
      "epoch": 0.9408,
      "grad_norm": 1.5,
      "learning_rate": 3.0578512396694214e-06,
      "loss": 0.084,
      "step": 2940
    },
    {
      "epoch": 0.944,
      "grad_norm": 1.015625,
      "learning_rate": 2.8925619834710747e-06,
      "loss": 0.0769,
      "step": 2950
    },
    {
      "epoch": 0.944,
      "eval_loss": 0.07859765738248825,
      "eval_runtime": 109.8176,
      "eval_samples_per_second": 4.553,
      "eval_steps_per_second": 0.146,
      "step": 2950
    },
    {
      "epoch": 0.9472,
      "grad_norm": 1.1640625,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.0722,
      "step": 2960
    },
    {
      "epoch": 0.9504,
      "grad_norm": 1.1015625,
      "learning_rate": 2.56198347107438e-06,
      "loss": 0.0839,
      "step": 2970
    },
    {
      "epoch": 0.9536,
      "grad_norm": 1.171875,
      "learning_rate": 2.3966942148760335e-06,
      "loss": 0.0818,
      "step": 2980
    },
    {
      "epoch": 0.9568,
      "grad_norm": 1.5,
      "learning_rate": 2.231404958677686e-06,
      "loss": 0.0757,
      "step": 2990
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.015625,
      "learning_rate": 2.066115702479339e-06,
      "loss": 0.0816,
      "step": 3000
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.07787304371595383,
      "eval_runtime": 108.0913,
      "eval_samples_per_second": 4.626,
      "eval_steps_per_second": 0.148,
      "step": 3000
    }
  ],
  "logging_steps": 10,
  "max_steps": 3125,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.75772653691904e+18,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}