{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.730659025787966,
  "eval_steps": 10,
  "global_step": 2000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.03,
      "grad_norm": 31.5,
      "learning_rate": 8.163265306122449e-07,
      "loss": 2.1773,
      "step": 10
    },
    {
      "epoch": 0.03,
      "eval_loss": 2.17791485786438,
      "eval_runtime": 1.0288,
      "eval_samples_per_second": 151.63,
      "eval_steps_per_second": 7.776,
      "step": 10
    },
    {
      "epoch": 0.06,
      "grad_norm": 28.0,
      "learning_rate": 1.6326530612244897e-06,
      "loss": 2.1852,
      "step": 20
    },
    {
      "epoch": 0.06,
      "eval_loss": 2.0857961177825928,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.901,
      "eval_steps_per_second": 7.79,
      "step": 20
    },
    {
      "epoch": 0.09,
      "grad_norm": 34.75,
      "learning_rate": 2.4489795918367347e-06,
      "loss": 1.9915,
      "step": 30
    },
    {
      "epoch": 0.09,
      "eval_loss": 1.8705158233642578,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.892,
      "eval_steps_per_second": 7.789,
      "step": 30
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.9375,
      "learning_rate": 3.2653061224489794e-06,
      "loss": 1.7352,
      "step": 40
    },
    {
      "epoch": 0.11,
      "eval_loss": 1.5546011924743652,
      "eval_runtime": 1.0295,
      "eval_samples_per_second": 151.524,
      "eval_steps_per_second": 7.77,
      "step": 40
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.25,
      "learning_rate": 4.081632653061225e-06,
      "loss": 1.318,
      "step": 50
    },
    {
      "epoch": 0.14,
      "eval_loss": 1.0024280548095703,
      "eval_runtime": 1.0273,
      "eval_samples_per_second": 151.848,
      "eval_steps_per_second": 7.787,
      "step": 50
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.625,
      "learning_rate": 4.897959183673469e-06,
      "loss": 0.9294,
      "step": 60
    },
    {
      "epoch": 0.17,
      "eval_loss": 0.8546950221061707,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.702,
      "eval_steps_per_second": 7.78,
      "step": 60
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.75,
      "learning_rate": 5.7142857142857145e-06,
      "loss": 0.8652,
      "step": 70
    },
    {
      "epoch": 0.2,
      "eval_loss": 0.8076146245002747,
      "eval_runtime": 1.0288,
      "eval_samples_per_second": 151.638,
      "eval_steps_per_second": 7.776,
      "step": 70
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.625,
      "learning_rate": 6.530612244897959e-06,
      "loss": 0.7794,
      "step": 80
    },
    {
      "epoch": 0.23,
      "eval_loss": 0.770220160484314,
      "eval_runtime": 1.0308,
      "eval_samples_per_second": 151.338,
      "eval_steps_per_second": 7.761,
      "step": 80
    },
    {
      "epoch": 0.26,
      "grad_norm": 5.6875,
      "learning_rate": 7.346938775510205e-06,
      "loss": 0.8108,
      "step": 90
    },
    {
      "epoch": 0.26,
      "eval_loss": 0.7469730973243713,
      "eval_runtime": 1.0319,
      "eval_samples_per_second": 151.174,
      "eval_steps_per_second": 7.752,
      "step": 90
    },
    {
      "epoch": 0.29,
      "grad_norm": 8.125,
      "learning_rate": 8.16326530612245e-06,
      "loss": 0.7018,
      "step": 100
    },
    {
      "epoch": 0.29,
      "eval_loss": 0.7343297600746155,
      "eval_runtime": 1.0334,
      "eval_samples_per_second": 150.952,
      "eval_steps_per_second": 7.741,
      "step": 100
    },
    {
      "epoch": 0.32,
      "grad_norm": 6.46875,
      "learning_rate": 8.979591836734695e-06,
      "loss": 0.7466,
      "step": 110
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.7234026789665222,
      "eval_runtime": 1.03,
      "eval_samples_per_second": 151.46,
      "eval_steps_per_second": 7.767,
      "step": 110
    },
    {
      "epoch": 0.34,
      "grad_norm": 3.984375,
      "learning_rate": 9.795918367346939e-06,
      "loss": 0.7353,
      "step": 120
    },
    {
      "epoch": 0.34,
      "eval_loss": 0.7116101384162903,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.673,
      "eval_steps_per_second": 7.778,
      "step": 120
    },
    {
      "epoch": 0.37,
      "grad_norm": 3.90625,
      "learning_rate": 1.0612244897959186e-05,
      "loss": 0.768,
      "step": 130
    },
    {
      "epoch": 0.37,
      "eval_loss": 0.702937662601471,
      "eval_runtime": 1.0286,
      "eval_samples_per_second": 151.657,
      "eval_steps_per_second": 7.777,
      "step": 130
    },
    {
      "epoch": 0.4,
      "grad_norm": 4.125,
      "learning_rate": 1.1428571428571429e-05,
      "loss": 0.7104,
      "step": 140
    },
    {
      "epoch": 0.4,
      "eval_loss": 0.6971867680549622,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.763,
      "eval_steps_per_second": 7.783,
      "step": 140
    },
    {
      "epoch": 0.43,
      "grad_norm": 7.03125,
      "learning_rate": 1.2244897959183674e-05,
      "loss": 0.7251,
      "step": 150
    },
    {
      "epoch": 0.43,
      "eval_loss": 0.6914936900138855,
      "eval_runtime": 1.0288,
      "eval_samples_per_second": 151.633,
      "eval_steps_per_second": 7.776,
      "step": 150
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.3125,
      "learning_rate": 1.3061224489795918e-05,
      "loss": 0.673,
      "step": 160
    },
    {
      "epoch": 0.46,
      "eval_loss": 0.6913846135139465,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.729,
      "eval_steps_per_second": 7.781,
      "step": 160
    },
    {
      "epoch": 0.49,
      "grad_norm": 3.46875,
      "learning_rate": 1.3877551020408165e-05,
      "loss": 0.7261,
      "step": 170
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.6843172311782837,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.672,
      "eval_steps_per_second": 7.778,
      "step": 170
    },
    {
      "epoch": 0.52,
      "grad_norm": 4.0,
      "learning_rate": 1.469387755102041e-05,
      "loss": 0.7126,
      "step": 180
    },
    {
      "epoch": 0.52,
      "eval_loss": 0.6830589175224304,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.702,
      "eval_steps_per_second": 7.78,
      "step": 180
    },
    {
      "epoch": 0.54,
      "grad_norm": 4.6875,
      "learning_rate": 1.5510204081632655e-05,
      "loss": 0.6471,
      "step": 190
    },
    {
      "epoch": 0.54,
      "eval_loss": 0.6809999346733093,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.706,
      "eval_steps_per_second": 7.78,
      "step": 190
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.25,
      "learning_rate": 1.63265306122449e-05,
      "loss": 0.7056,
      "step": 200
    },
    {
      "epoch": 0.57,
      "eval_loss": 0.6814644932746887,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.767,
      "eval_steps_per_second": 7.783,
      "step": 200
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.1875,
      "learning_rate": 1.7142857142857142e-05,
      "loss": 0.7437,
      "step": 210
    },
    {
      "epoch": 0.6,
      "eval_loss": 0.6767860651016235,
      "eval_runtime": 1.0287,
      "eval_samples_per_second": 151.643,
      "eval_steps_per_second": 7.777,
      "step": 210
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.6875,
      "learning_rate": 1.795918367346939e-05,
      "loss": 0.698,
      "step": 220
    },
    {
      "epoch": 0.63,
      "eval_loss": 0.6760666966438293,
      "eval_runtime": 1.0302,
      "eval_samples_per_second": 151.422,
      "eval_steps_per_second": 7.765,
      "step": 220
    },
    {
      "epoch": 0.66,
      "grad_norm": 7.0625,
      "learning_rate": 1.8775510204081636e-05,
      "loss": 0.6675,
      "step": 230
    },
    {
      "epoch": 0.66,
      "eval_loss": 0.6715320944786072,
      "eval_runtime": 1.0272,
      "eval_samples_per_second": 151.869,
      "eval_steps_per_second": 7.788,
      "step": 230
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.125,
      "learning_rate": 1.9591836734693877e-05,
      "loss": 0.7056,
      "step": 240
    },
    {
      "epoch": 0.69,
      "eval_loss": 0.6754013299942017,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.704,
      "eval_steps_per_second": 7.78,
      "step": 240
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.125,
      "learning_rate": 1.9999744640197457e-05,
      "loss": 0.6884,
      "step": 250
    },
    {
      "epoch": 0.72,
      "eval_loss": 0.672713577747345,
      "eval_runtime": 1.0287,
      "eval_samples_per_second": 151.648,
      "eval_steps_per_second": 7.777,
      "step": 250
    },
    {
      "epoch": 0.74,
      "grad_norm": 7.3125,
      "learning_rate": 1.999770184002678e-05,
      "loss": 0.6497,
      "step": 260
    },
    {
      "epoch": 0.74,
      "eval_loss": 0.6706756949424744,
      "eval_runtime": 1.035,
      "eval_samples_per_second": 150.73,
      "eval_steps_per_second": 7.73,
      "step": 260
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.4375,
      "learning_rate": 1.999361665699934e-05,
      "loss": 0.6699,
      "step": 270
    },
    {
      "epoch": 0.77,
      "eval_loss": 0.6661806106567383,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.622,
      "eval_steps_per_second": 7.775,
      "step": 270
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.6875,
      "learning_rate": 1.99874899256577e-05,
      "loss": 0.6864,
      "step": 280
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.6693017482757568,
      "eval_runtime": 1.0292,
      "eval_samples_per_second": 151.572,
      "eval_steps_per_second": 7.773,
      "step": 280
    },
    {
      "epoch": 0.83,
      "grad_norm": 7.1875,
      "learning_rate": 1.997932289760261e-05,
      "loss": 0.6304,
      "step": 290
    },
    {
      "epoch": 0.83,
      "eval_loss": 0.6646809577941895,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.69,
      "eval_steps_per_second": 7.779,
      "step": 290
    },
    {
      "epoch": 0.86,
      "grad_norm": 2.796875,
      "learning_rate": 1.9969117241237302e-05,
      "loss": 0.6425,
      "step": 300
    },
    {
      "epoch": 0.86,
      "eval_loss": 0.6653949022293091,
      "eval_runtime": 1.0295,
      "eval_samples_per_second": 151.53,
      "eval_steps_per_second": 7.771,
      "step": 300
    },
    {
      "epoch": 0.89,
      "grad_norm": 3.421875,
      "learning_rate": 1.995687504142667e-05,
      "loss": 0.7201,
      "step": 310
    },
    {
      "epoch": 0.89,
      "eval_loss": 0.664738118648529,
      "eval_runtime": 1.0308,
      "eval_samples_per_second": 151.342,
      "eval_steps_per_second": 7.761,
      "step": 310
    },
    {
      "epoch": 0.92,
      "grad_norm": 3.0625,
      "learning_rate": 1.994259879907137e-05,
      "loss": 0.6947,
      "step": 320
    },
    {
      "epoch": 0.92,
      "eval_loss": 0.6645806431770325,
      "eval_runtime": 1.0295,
      "eval_samples_per_second": 151.523,
      "eval_steps_per_second": 7.77,
      "step": 320
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.4375,
      "learning_rate": 1.99262914305969e-05,
      "loss": 0.6722,
      "step": 330
    },
    {
      "epoch": 0.95,
      "eval_loss": 0.6644319295883179,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.722,
      "eval_steps_per_second": 7.781,
      "step": 330
    },
    {
      "epoch": 0.97,
      "grad_norm": 3.59375,
      "learning_rate": 1.990795626735784e-05,
      "loss": 0.616,
      "step": 340
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.668482780456543,
      "eval_runtime": 1.029,
      "eval_samples_per_second": 151.601,
      "eval_steps_per_second": 7.774,
      "step": 340
    },
    {
      "epoch": 1.0,
      "grad_norm": 3.078125,
      "learning_rate": 1.9887597054957304e-05,
      "loss": 0.6448,
      "step": 350
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.6687355041503906,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.549,
      "eval_steps_per_second": 7.772,
      "step": 350
    },
    {
      "epoch": 1.03,
      "grad_norm": 4.25,
      "learning_rate": 1.986521795248175e-05,
      "loss": 0.5472,
      "step": 360
    },
    {
      "epoch": 1.03,
      "eval_loss": 0.674789547920227,
      "eval_runtime": 1.03,
      "eval_samples_per_second": 151.454,
      "eval_steps_per_second": 7.767,
      "step": 360
    },
    {
      "epoch": 1.06,
      "grad_norm": 5.375,
      "learning_rate": 1.9840823531651357e-05,
      "loss": 0.5425,
      "step": 370
    },
    {
      "epoch": 1.06,
      "eval_loss": 0.6699011325836182,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.729,
      "eval_steps_per_second": 7.781,
      "step": 370
    },
    {
      "epoch": 1.09,
      "grad_norm": 6.9375,
      "learning_rate": 1.9814418775886083e-05,
      "loss": 0.5447,
      "step": 380
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.6747943162918091,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.708,
      "eval_steps_per_second": 7.78,
      "step": 380
    },
    {
      "epoch": 1.12,
      "grad_norm": 4.28125,
      "learning_rate": 1.978600907928764e-05,
      "loss": 0.5412,
      "step": 390
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.6677992939949036,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.683,
      "eval_steps_per_second": 7.779,
      "step": 390
    },
    {
      "epoch": 1.15,
      "grad_norm": 4.5625,
      "learning_rate": 1.9755600245537522e-05,
      "loss": 0.5712,
      "step": 400
    },
    {
      "epoch": 1.15,
      "eval_loss": 0.6665674448013306,
      "eval_runtime": 1.029,
      "eval_samples_per_second": 151.602,
      "eval_steps_per_second": 7.774,
      "step": 400
    },
    {
      "epoch": 1.17,
      "grad_norm": 3.828125,
      "learning_rate": 1.972319848671145e-05,
      "loss": 0.5601,
      "step": 410
    },
    {
      "epoch": 1.17,
      "eval_loss": 0.6693459749221802,
      "eval_runtime": 1.0316,
      "eval_samples_per_second": 151.221,
      "eval_steps_per_second": 7.755,
      "step": 410
    },
    {
      "epoch": 1.2,
      "grad_norm": 2.515625,
      "learning_rate": 1.968881042201029e-05,
      "loss": 0.5176,
      "step": 420
    },
    {
      "epoch": 1.2,
      "eval_loss": 0.6797675490379333,
      "eval_runtime": 1.0313,
      "eval_samples_per_second": 151.268,
      "eval_steps_per_second": 7.757,
      "step": 420
    },
    {
      "epoch": 1.23,
      "grad_norm": 3.140625,
      "learning_rate": 1.9652443076407884e-05,
      "loss": 0.5072,
      "step": 430
    },
    {
      "epoch": 1.23,
      "eval_loss": 0.6783838272094727,
      "eval_runtime": 1.0297,
      "eval_samples_per_second": 151.493,
      "eval_steps_per_second": 7.769,
      "step": 430
    },
    {
      "epoch": 1.26,
      "grad_norm": 3.375,
      "learning_rate": 1.9614103879215917e-05,
      "loss": 0.5552,
      "step": 440
    },
    {
      "epoch": 1.26,
      "eval_loss": 0.6780270338058472,
      "eval_runtime": 1.0303,
      "eval_samples_per_second": 151.416,
      "eval_steps_per_second": 7.765,
      "step": 440
    },
    {
      "epoch": 1.29,
      "grad_norm": 6.75,
      "learning_rate": 1.9573800662566257e-05,
      "loss": 0.5193,
      "step": 450
    },
    {
      "epoch": 1.29,
      "eval_loss": 0.6783522367477417,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.739,
      "eval_steps_per_second": 7.781,
      "step": 450
    },
    {
      "epoch": 1.32,
      "grad_norm": 6.0,
      "learning_rate": 1.9531541659810927e-05,
      "loss": 0.5513,
      "step": 460
    },
    {
      "epoch": 1.32,
      "eval_loss": 0.6770094037055969,
      "eval_runtime": 1.0314,
      "eval_samples_per_second": 151.253,
      "eval_steps_per_second": 7.757,
      "step": 460
    },
    {
      "epoch": 1.35,
      "grad_norm": 9.625,
      "learning_rate": 1.9487335503840186e-05,
      "loss": 0.5697,
      "step": 470
    },
    {
      "epoch": 1.35,
      "eval_loss": 0.6728355288505554,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.728,
      "eval_steps_per_second": 7.781,
      "step": 470
    },
    {
      "epoch": 1.38,
      "grad_norm": 3.40625,
      "learning_rate": 1.9441191225318934e-05,
      "loss": 0.5153,
      "step": 480
    },
    {
      "epoch": 1.38,
      "eval_loss": 0.6756737232208252,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.706,
      "eval_steps_per_second": 7.78,
      "step": 480
    },
    {
      "epoch": 1.4,
      "grad_norm": 3.203125,
      "learning_rate": 1.9393118250841897e-05,
      "loss": 0.5808,
      "step": 490
    },
    {
      "epoch": 1.4,
      "eval_loss": 0.6708477139472961,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.79,
      "eval_steps_per_second": 7.784,
      "step": 490
    },
    {
      "epoch": 1.43,
      "grad_norm": 4.0625,
      "learning_rate": 1.9343126401007893e-05,
      "loss": 0.5525,
      "step": 500
    },
    {
      "epoch": 1.43,
      "eval_loss": 0.6698306202888489,
      "eval_runtime": 1.0275,
      "eval_samples_per_second": 151.828,
      "eval_steps_per_second": 7.786,
      "step": 500
    },
    {
      "epoch": 1.46,
      "grad_norm": 5.84375,
      "learning_rate": 1.9291225888413652e-05,
      "loss": 0.5412,
      "step": 510
    },
    {
      "epoch": 1.46,
      "eval_loss": 0.6722325086593628,
      "eval_runtime": 1.0528,
      "eval_samples_per_second": 148.179,
      "eval_steps_per_second": 7.599,
      "step": 510
    },
    {
      "epoch": 1.49,
      "grad_norm": 3.65625,
      "learning_rate": 1.923742731556752e-05,
      "loss": 0.5236,
      "step": 520
    },
    {
      "epoch": 1.49,
      "eval_loss": 0.6703674793243408,
      "eval_runtime": 1.0436,
      "eval_samples_per_second": 149.477,
      "eval_steps_per_second": 7.666,
      "step": 520
    },
    {
      "epoch": 1.52,
      "grad_norm": 3.59375,
      "learning_rate": 1.9181741672723516e-05,
      "loss": 0.5794,
      "step": 530
    },
    {
      "epoch": 1.52,
      "eval_loss": 0.6666515469551086,
      "eval_runtime": 1.0472,
      "eval_samples_per_second": 148.969,
      "eval_steps_per_second": 7.639,
      "step": 530
    },
    {
      "epoch": 1.55,
      "grad_norm": 4.21875,
      "learning_rate": 1.9124180335636213e-05,
      "loss": 0.4933,
      "step": 540
    },
    {
      "epoch": 1.55,
      "eval_loss": 0.6719895601272583,
      "eval_runtime": 1.0512,
      "eval_samples_per_second": 148.395,
      "eval_steps_per_second": 7.61,
      "step": 540
    },
    {
      "epoch": 1.58,
      "grad_norm": 6.09375,
      "learning_rate": 1.90647550632368e-05,
      "loss": 0.5346,
      "step": 550
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.6691425442695618,
      "eval_runtime": 1.0539,
      "eval_samples_per_second": 148.025,
      "eval_steps_per_second": 7.591,
      "step": 550
    },
    {
      "epoch": 1.6,
      "grad_norm": 4.6875,
      "learning_rate": 1.9003477995230942e-05,
      "loss": 0.5743,
      "step": 560
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.6666125059127808,
      "eval_runtime": 1.0572,
      "eval_samples_per_second": 147.555,
      "eval_steps_per_second": 7.567,
      "step": 560
    },
    {
      "epoch": 1.63,
      "grad_norm": 3.25,
      "learning_rate": 1.894036164961879e-05,
      "loss": 0.5662,
      "step": 570
    },
    {
      "epoch": 1.63,
      "eval_loss": 0.66337651014328,
      "eval_runtime": 1.0492,
      "eval_samples_per_second": 148.684,
      "eval_steps_per_second": 7.625,
      "step": 570
    },
    {
      "epoch": 1.66,
      "grad_norm": 3.359375,
      "learning_rate": 1.8875418920137764e-05,
      "loss": 0.5924,
      "step": 580
    },
    {
      "epoch": 1.66,
      "eval_loss": 0.6664850115776062,
      "eval_runtime": 1.0522,
      "eval_samples_per_second": 148.261,
      "eval_steps_per_second": 7.603,
      "step": 580
    },
    {
      "epoch": 1.69,
      "grad_norm": 3.640625,
      "learning_rate": 1.880866307362853e-05,
      "loss": 0.5186,
      "step": 590
    },
    {
      "epoch": 1.69,
      "eval_loss": 0.6705337762832642,
      "eval_runtime": 1.0529,
      "eval_samples_per_second": 148.163,
      "eval_steps_per_second": 7.598,
      "step": 590
    },
    {
      "epoch": 1.72,
      "grad_norm": 4.125,
      "learning_rate": 1.8740107747324785e-05,
      "loss": 0.5155,
      "step": 600
    },
    {
      "epoch": 1.72,
      "eval_loss": 0.6648845076560974,
      "eval_runtime": 1.0494,
      "eval_samples_per_second": 148.657,
      "eval_steps_per_second": 7.623,
      "step": 600
    },
    {
      "epoch": 1.75,
      "grad_norm": 3.171875,
      "learning_rate": 1.8669766946067398e-05,
      "loss": 0.5317,
      "step": 610
    },
    {
      "epoch": 1.75,
      "eval_loss": 0.6645151972770691,
      "eval_runtime": 1.029,
      "eval_samples_per_second": 151.608,
      "eval_steps_per_second": 7.775,
      "step": 610
    },
    {
      "epoch": 1.78,
      "grad_norm": 3.390625,
      "learning_rate": 1.8597655039443384e-05,
      "loss": 0.5151,
      "step": 620
    },
    {
      "epoch": 1.78,
      "eval_loss": 0.6744809150695801,
      "eval_runtime": 1.0296,
      "eval_samples_per_second": 151.522,
      "eval_steps_per_second": 7.77,
      "step": 620
    },
    {
      "epoch": 1.81,
      "grad_norm": 2.78125,
      "learning_rate": 1.8523786758850436e-05,
      "loss": 0.5269,
      "step": 630
    },
    {
      "epoch": 1.81,
      "eval_loss": 0.6621160507202148,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.733,
      "eval_steps_per_second": 7.781,
      "step": 630
    },
    {
      "epoch": 1.83,
      "grad_norm": 3.15625,
      "learning_rate": 1.8448177194487524e-05,
      "loss": 0.5522,
      "step": 640
    },
    {
      "epoch": 1.83,
      "eval_loss": 0.6631208658218384,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.707,
      "eval_steps_per_second": 7.78,
      "step": 640
    },
    {
      "epoch": 1.86,
      "grad_norm": 3.03125,
      "learning_rate": 1.837084179227217e-05,
      "loss": 0.573,
      "step": 650
    },
    {
      "epoch": 1.86,
      "eval_loss": 0.6646651029586792,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.773,
      "eval_steps_per_second": 7.783,
      "step": 650
    },
    {
      "epoch": 1.89,
      "grad_norm": 2.71875,
      "learning_rate": 1.829179635068509e-05,
      "loss": 0.5536,
      "step": 660
    },
    {
      "epoch": 1.89,
      "eval_loss": 0.6614721417427063,
      "eval_runtime": 1.0275,
      "eval_samples_per_second": 151.818,
      "eval_steps_per_second": 7.786,
      "step": 660
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.640625,
      "learning_rate": 1.821105701754279e-05,
      "loss": 0.5298,
      "step": 670
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.6630841493606567,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.541,
      "eval_steps_per_second": 7.771,
      "step": 670
    },
    {
      "epoch": 1.95,
      "grad_norm": 2.984375,
      "learning_rate": 1.812864028669881e-05,
      "loss": 0.5331,
      "step": 680
    },
    {
      "epoch": 1.95,
      "eval_loss": 0.6609542369842529,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.722,
      "eval_steps_per_second": 7.781,
      "step": 680
    },
    {
      "epoch": 1.98,
      "grad_norm": 3.6875,
      "learning_rate": 1.8044562994674266e-05,
      "loss": 0.505,
      "step": 690
    },
    {
      "epoch": 1.98,
      "eval_loss": 0.66065913438797,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.78,
      "eval_steps_per_second": 7.784,
      "step": 690
    },
    {
      "epoch": 2.01,
      "grad_norm": 2.90625,
      "learning_rate": 1.7958842317218413e-05,
      "loss": 0.4993,
      "step": 700
    },
    {
      "epoch": 2.01,
      "eval_loss": 0.658078134059906,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.803,
      "eval_steps_per_second": 7.785,
      "step": 700
    },
    {
      "epoch": 2.03,
      "grad_norm": 3.84375,
      "learning_rate": 1.7871495765799875e-05,
      "loss": 0.3185,
      "step": 710
    },
    {
      "epoch": 2.03,
      "eval_loss": 0.7684138417243958,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.733,
      "eval_steps_per_second": 7.781,
      "step": 710
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.0,
      "learning_rate": 1.7782541184029316e-05,
      "loss": 0.3207,
      "step": 720
    },
    {
      "epoch": 2.06,
      "eval_loss": 0.6996563076972961,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.729,
      "eval_steps_per_second": 7.781,
      "step": 720
    },
    {
      "epoch": 2.09,
      "grad_norm": 4.21875,
      "learning_rate": 1.769199674401427e-05,
      "loss": 0.3208,
      "step": 730
    },
    {
      "epoch": 2.09,
      "eval_loss": 0.7326269745826721,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.714,
      "eval_steps_per_second": 7.78,
      "step": 730
    },
    {
      "epoch": 2.12,
      "grad_norm": 3.015625,
      "learning_rate": 1.759988094264682e-05,
      "loss": 0.2998,
      "step": 740
    },
    {
      "epoch": 2.12,
      "eval_loss": 0.7228878736495972,
      "eval_runtime": 1.0298,
      "eval_samples_per_second": 151.481,
      "eval_steps_per_second": 7.768,
      "step": 740
    },
    {
      "epoch": 2.15,
      "grad_norm": 2.65625,
      "learning_rate": 1.7506212597824976e-05,
      "loss": 0.3,
      "step": 750
    },
    {
      "epoch": 2.15,
      "eval_loss": 0.716642439365387,
      "eval_runtime": 1.0288,
      "eval_samples_per_second": 151.63,
      "eval_steps_per_second": 7.776,
      "step": 750
    },
    {
      "epoch": 2.18,
      "grad_norm": 3.296875,
      "learning_rate": 1.7411010844608448e-05,
      "loss": 0.3212,
      "step": 760
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.722019374370575,
      "eval_runtime": 1.0268,
      "eval_samples_per_second": 151.933,
      "eval_steps_per_second": 7.791,
      "step": 760
    },
    {
      "epoch": 2.21,
      "grad_norm": 3.5625,
      "learning_rate": 1.731429513130964e-05,
      "loss": 0.3192,
      "step": 770
    },
    {
      "epoch": 2.21,
      "eval_loss": 0.7228615880012512,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.694,
      "eval_steps_per_second": 7.779,
      "step": 770
    },
    {
      "epoch": 2.23,
      "grad_norm": 3.359375,
      "learning_rate": 1.7216085215520644e-05,
      "loss": 0.298,
      "step": 780
    },
    {
      "epoch": 2.23,
      "eval_loss": 0.7244638204574585,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.546,
      "eval_steps_per_second": 7.772,
      "step": 780
    },
    {
      "epoch": 2.26,
      "grad_norm": 4.0,
      "learning_rate": 1.711640116007706e-05,
      "loss": 0.3152,
      "step": 790
    },
    {
      "epoch": 2.26,
      "eval_loss": 0.7284898161888123,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.719,
      "eval_steps_per_second": 7.78,
      "step": 790
    },
    {
      "epoch": 2.29,
      "grad_norm": 4.28125,
      "learning_rate": 1.701526332895945e-05,
      "loss": 0.2995,
      "step": 800
    },
    {
      "epoch": 2.29,
      "eval_loss": 0.7223809957504272,
      "eval_runtime": 1.0287,
      "eval_samples_per_second": 151.648,
      "eval_steps_per_second": 7.777,
      "step": 800
    },
    {
      "epoch": 2.32,
      "grad_norm": 3.46875,
      "learning_rate": 1.6912692383133276e-05,
      "loss": 0.313,
      "step": 810
    },
    {
      "epoch": 2.32,
      "eval_loss": 0.7280778288841248,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.759,
      "eval_steps_per_second": 7.783,
      "step": 810
    },
    {
      "epoch": 2.35,
      "grad_norm": 3.1875,
      "learning_rate": 1.680870927632818e-05,
      "loss": 0.2983,
      "step": 820
    },
    {
      "epoch": 2.35,
      "eval_loss": 0.7330511212348938,
      "eval_runtime": 1.0274,
      "eval_samples_per_second": 151.84,
      "eval_steps_per_second": 7.787,
      "step": 820
    },
    {
      "epoch": 2.38,
      "grad_norm": 2.515625,
      "learning_rate": 1.6703335250757428e-05,
      "loss": 0.3204,
      "step": 830
    },
    {
      "epoch": 2.38,
      "eval_loss": 0.7196641564369202,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.708,
      "eval_steps_per_second": 7.78,
      "step": 830
    },
    {
      "epoch": 2.41,
      "grad_norm": 3.40625,
      "learning_rate": 1.659659183277847e-05,
      "loss": 0.2942,
      "step": 840
    },
    {
      "epoch": 2.41,
      "eval_loss": 0.7265558242797852,
      "eval_runtime": 1.0302,
      "eval_samples_per_second": 151.425,
      "eval_steps_per_second": 7.765,
      "step": 840
    },
    {
      "epoch": 2.44,
      "grad_norm": 4.0,
      "learning_rate": 1.6488500828495384e-05,
      "loss": 0.3184,
      "step": 850
    },
    {
      "epoch": 2.44,
      "eval_loss": 0.7329144477844238,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.798,
      "eval_steps_per_second": 7.784,
      "step": 850
    },
    {
      "epoch": 2.46,
      "grad_norm": 3.15625,
      "learning_rate": 1.6379084319304245e-05,
      "loss": 0.3255,
      "step": 860
    },
    {
      "epoch": 2.46,
      "eval_loss": 0.729471743106842,
      "eval_runtime": 1.0288,
      "eval_samples_per_second": 151.628,
      "eval_steps_per_second": 7.776,
      "step": 860
    },
    {
      "epoch": 2.49,
      "grad_norm": 3.375,
      "learning_rate": 1.6268364657382196e-05,
      "loss": 0.3239,
      "step": 870
    },
    {
      "epoch": 2.49,
      "eval_loss": 0.7192624807357788,
      "eval_runtime": 1.0271,
      "eval_samples_per_second": 151.878,
      "eval_steps_per_second": 7.789,
      "step": 870
    },
    {
      "epoch": 2.52,
      "grad_norm": 3.578125,
      "learning_rate": 1.6156364461121255e-05,
      "loss": 0.3255,
      "step": 880
    },
    {
      "epoch": 2.52,
      "eval_loss": 0.7301867604255676,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.724,
      "eval_steps_per_second": 7.781,
      "step": 880
    },
    {
      "epoch": 2.55,
      "grad_norm": 4.0,
      "learning_rate": 1.6043106610507683e-05,
      "loss": 0.3167,
      "step": 890
    },
    {
      "epoch": 2.55,
      "eval_loss": 0.7298671007156372,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.783,
      "eval_steps_per_second": 7.784,
      "step": 890
    },
    {
      "epoch": 2.58,
      "grad_norm": 3.140625,
      "learning_rate": 1.5928614242447965e-05,
      "loss": 0.3157,
      "step": 900
    },
    {
      "epoch": 2.58,
      "eval_loss": 0.726569652557373,
      "eval_runtime": 1.0303,
      "eval_samples_per_second": 151.412,
      "eval_steps_per_second": 7.765,
      "step": 900
    },
    {
      "epoch": 2.61,
      "grad_norm": 3.28125,
      "learning_rate": 1.581291074604226e-05,
      "loss": 0.3128,
      "step": 910
    },
    {
      "epoch": 2.61,
      "eval_loss": 0.7234473824501038,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.547,
      "eval_steps_per_second": 7.772,
      "step": 910
    },
    {
      "epoch": 2.64,
      "grad_norm": 3.1875,
      "learning_rate": 1.5696019757806373e-05,
      "loss": 0.2989,
      "step": 920
    },
    {
      "epoch": 2.64,
      "eval_loss": 0.7260815501213074,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.791,
      "eval_steps_per_second": 7.784,
      "step": 920
    },
    {
      "epoch": 2.66,
      "grad_norm": 3.28125,
      "learning_rate": 1.5577965156843153e-05,
      "loss": 0.3263,
      "step": 930
    },
    {
      "epoch": 2.66,
      "eval_loss": 0.7242666482925415,
      "eval_runtime": 1.0291,
      "eval_samples_per_second": 151.596,
      "eval_steps_per_second": 7.774,
      "step": 930
    },
    {
      "epoch": 2.69,
      "grad_norm": 3.484375,
      "learning_rate": 1.5458771059964348e-05,
      "loss": 0.3189,
      "step": 940
    },
    {
      "epoch": 2.69,
      "eval_loss": 0.7277385592460632,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.759,
      "eval_steps_per_second": 7.783,
      "step": 940
    },
    {
      "epoch": 2.72,
      "grad_norm": 3.5625,
      "learning_rate": 1.533846181676389e-05,
      "loss": 0.3129,
      "step": 950
    },
    {
      "epoch": 2.72,
      "eval_loss": 0.729342520236969,
      "eval_runtime": 1.0269,
      "eval_samples_per_second": 151.913,
      "eval_steps_per_second": 7.79,
      "step": 950
    },
    {
      "epoch": 2.75,
      "grad_norm": 3.328125,
      "learning_rate": 1.521706200464364e-05,
      "loss": 0.305,
      "step": 960
    },
    {
      "epoch": 2.75,
      "eval_loss": 0.7262902855873108,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.621,
      "eval_steps_per_second": 7.775,
      "step": 960
    },
    {
      "epoch": 2.78,
      "grad_norm": 2.296875,
      "learning_rate": 1.509459642379259e-05,
      "loss": 0.3115,
      "step": 970
    },
    {
      "epoch": 2.78,
      "eval_loss": 0.7204219698905945,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.688,
      "eval_steps_per_second": 7.779,
      "step": 970
    },
    {
      "epoch": 2.81,
      "grad_norm": 2.34375,
      "learning_rate": 1.4971090092120544e-05,
      "loss": 0.302,
      "step": 980
    },
    {
      "epoch": 2.81,
      "eval_loss": 0.7241439819335938,
      "eval_runtime": 1.0293,
      "eval_samples_per_second": 151.563,
      "eval_steps_per_second": 7.772,
      "step": 980
    },
    {
      "epoch": 2.84,
      "grad_norm": 2.953125,
      "learning_rate": 1.4846568240147327e-05,
      "loss": 0.3218,
      "step": 990
    },
    {
      "epoch": 2.84,
      "eval_loss": 0.7206692695617676,
      "eval_runtime": 1.0291,
      "eval_samples_per_second": 151.585,
      "eval_steps_per_second": 7.774,
      "step": 990
    },
    {
      "epoch": 2.87,
      "grad_norm": 3.328125,
      "learning_rate": 1.4721056305848571e-05,
      "loss": 0.3202,
      "step": 1000
    },
    {
      "epoch": 2.87,
      "eval_loss": 0.728424608707428,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.678,
      "eval_steps_per_second": 7.778,
      "step": 1000
    },
    {
      "epoch": 2.89,
      "grad_norm": 2.640625,
      "learning_rate": 1.4594579929459107e-05,
      "loss": 0.328,
      "step": 1010
    },
    {
      "epoch": 2.89,
      "eval_loss": 0.7225694060325623,
      "eval_runtime": 1.024,
      "eval_samples_per_second": 152.344,
      "eval_steps_per_second": 7.813,
      "step": 1010
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.921875,
      "learning_rate": 1.446716494823504e-05,
      "loss": 0.3139,
      "step": 1020
    },
    {
      "epoch": 2.92,
      "eval_loss": 0.7302761673927307,
      "eval_runtime": 1.0249,
      "eval_samples_per_second": 152.213,
      "eval_steps_per_second": 7.806,
      "step": 1020
    },
    {
      "epoch": 2.95,
      "grad_norm": 3.546875,
      "learning_rate": 1.4338837391175582e-05,
      "loss": 0.3337,
      "step": 1030
    },
    {
      "epoch": 2.95,
      "eval_loss": 0.72865229845047,
      "eval_runtime": 1.0269,
      "eval_samples_per_second": 151.911,
      "eval_steps_per_second": 7.79,
      "step": 1030
    },
    {
      "epoch": 2.98,
      "grad_norm": 3.296875,
      "learning_rate": 1.4209623473705722e-05,
      "loss": 0.3162,
      "step": 1040
    },
    {
      "epoch": 2.98,
      "eval_loss": 0.7287803292274475,
      "eval_runtime": 1.0261,
      "eval_samples_per_second": 152.029,
      "eval_steps_per_second": 7.796,
      "step": 1040
    },
    {
      "epoch": 3.01,
      "grad_norm": 3.203125,
      "learning_rate": 1.4079549592320782e-05,
      "loss": 0.2787,
      "step": 1050
    },
    {
      "epoch": 3.01,
      "eval_loss": 0.7424219250679016,
      "eval_runtime": 1.0318,
      "eval_samples_per_second": 151.185,
      "eval_steps_per_second": 7.753,
      "step": 1050
    },
    {
      "epoch": 3.04,
      "grad_norm": 3.3125,
      "learning_rate": 1.3948642319194021e-05,
      "loss": 0.178,
      "step": 1060
    },
    {
      "epoch": 3.04,
      "eval_loss": 0.8832307457923889,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.625,
      "eval_steps_per_second": 7.776,
      "step": 1060
    },
    {
      "epoch": 3.07,
      "grad_norm": 3.265625,
      "learning_rate": 1.3816928396748321e-05,
      "loss": 0.1773,
      "step": 1070
    },
    {
      "epoch": 3.07,
      "eval_loss": 0.8206979632377625,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.673,
      "eval_steps_per_second": 7.778,
      "step": 1070
    },
    {
      "epoch": 3.09,
      "grad_norm": 2.703125,
      "learning_rate": 1.3684434732193107e-05,
      "loss": 0.1742,
      "step": 1080
    },
    {
      "epoch": 3.09,
      "eval_loss": 0.8235818147659302,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.551,
      "eval_steps_per_second": 7.772,
      "step": 1080
    },
    {
      "epoch": 3.12,
      "grad_norm": 3.0,
      "learning_rate": 1.3551188392027606e-05,
      "loss": 0.1864,
      "step": 1090
    },
    {
      "epoch": 3.12,
      "eval_loss": 0.8448302149772644,
      "eval_runtime": 1.0291,
      "eval_samples_per_second": 151.583,
      "eval_steps_per_second": 7.773,
      "step": 1090
    },
    {
      "epoch": 3.15,
      "grad_norm": 2.71875,
      "learning_rate": 1.3417216596511557e-05,
      "loss": 0.1687,
      "step": 1100
    },
    {
      "epoch": 3.15,
      "eval_loss": 0.8366988897323608,
      "eval_runtime": 1.0292,
      "eval_samples_per_second": 151.568,
      "eval_steps_per_second": 7.773,
      "step": 1100
    },
    {
      "epoch": 3.18,
      "grad_norm": 2.828125,
      "learning_rate": 1.328254671410452e-05,
      "loss": 0.1856,
      "step": 1110
    },
    {
      "epoch": 3.18,
      "eval_loss": 0.8296992182731628,
      "eval_runtime": 1.03,
      "eval_samples_per_second": 151.449,
      "eval_steps_per_second": 7.767,
      "step": 1110
    },
    {
      "epoch": 3.21,
      "grad_norm": 3.28125,
      "learning_rate": 1.3147206255874886e-05,
      "loss": 0.1802,
      "step": 1120
    },
    {
      "epoch": 3.21,
      "eval_loss": 0.8381376266479492,
      "eval_runtime": 1.0295,
      "eval_samples_per_second": 151.53,
      "eval_steps_per_second": 7.771,
      "step": 1120
    },
    {
      "epoch": 3.24,
      "grad_norm": 2.34375,
      "learning_rate": 1.3011222869879796e-05,
      "loss": 0.1795,
      "step": 1130
    },
    {
      "epoch": 3.24,
      "eval_loss": 0.8237268328666687,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.688,
      "eval_steps_per_second": 7.779,
      "step": 1130
    },
    {
      "epoch": 3.27,
      "grad_norm": 2.71875,
      "learning_rate": 1.287462433551704e-05,
      "loss": 0.1837,
      "step": 1140
    },
    {
      "epoch": 3.27,
      "eval_loss": 0.8503643870353699,
      "eval_runtime": 1.0292,
      "eval_samples_per_second": 151.576,
      "eval_steps_per_second": 7.773,
      "step": 1140
    },
    {
      "epoch": 3.3,
      "grad_norm": 3.375,
      "learning_rate": 1.273743855785012e-05,
      "loss": 0.1706,
      "step": 1150
    },
    {
      "epoch": 3.3,
      "eval_loss": 0.8583101034164429,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.76,
      "eval_steps_per_second": 7.783,
      "step": 1150
    },
    {
      "epoch": 3.32,
      "grad_norm": 2.421875,
      "learning_rate": 1.25996935619077e-05,
      "loss": 0.1832,
      "step": 1160
    },
    {
      "epoch": 3.32,
      "eval_loss": 0.8319224715232849,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.714,
      "eval_steps_per_second": 7.78,
      "step": 1160
    },
    {
      "epoch": 3.35,
      "grad_norm": 3.34375,
      "learning_rate": 1.2461417486958463e-05,
      "loss": 0.196,
      "step": 1170
    },
    {
      "epoch": 3.35,
      "eval_loss": 0.8305267095565796,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.768,
      "eval_steps_per_second": 7.783,
      "step": 1170
    },
    {
      "epoch": 3.38,
      "grad_norm": 3.0,
      "learning_rate": 1.2322638580762684e-05,
      "loss": 0.1769,
      "step": 1180
    },
    {
      "epoch": 3.38,
      "eval_loss": 0.8342500329017639,
      "eval_runtime": 1.028,
      "eval_samples_per_second": 151.75,
      "eval_steps_per_second": 7.782,
      "step": 1180
    },
    {
      "epoch": 3.41,
      "grad_norm": 2.28125,
      "learning_rate": 1.2183385193801655e-05,
      "loss": 0.175,
      "step": 1190
    },
    {
      "epoch": 3.41,
      "eval_loss": 0.8393198847770691,
      "eval_runtime": 1.0287,
      "eval_samples_per_second": 151.644,
      "eval_steps_per_second": 7.777,
      "step": 1190
    },
    {
      "epoch": 3.44,
      "grad_norm": 3.078125,
      "learning_rate": 1.2043685773486073e-05,
      "loss": 0.1795,
      "step": 1200
    },
    {
      "epoch": 3.44,
      "eval_loss": 0.8404285907745361,
      "eval_runtime": 1.029,
      "eval_samples_per_second": 151.601,
      "eval_steps_per_second": 7.774,
      "step": 1200
    },
    {
      "epoch": 3.47,
      "grad_norm": 2.734375,
      "learning_rate": 1.1903568858344667e-05,
      "loss": 0.1908,
      "step": 1210
    },
    {
      "epoch": 3.47,
      "eval_loss": 0.8395776152610779,
      "eval_runtime": 1.03,
      "eval_samples_per_second": 151.454,
      "eval_steps_per_second": 7.767,
      "step": 1210
    },
    {
      "epoch": 3.5,
      "grad_norm": 2.65625,
      "learning_rate": 1.1763063072194181e-05,
      "loss": 0.1937,
      "step": 1220
    },
    {
      "epoch": 3.5,
      "eval_loss": 0.8442292809486389,
      "eval_runtime": 1.029,
      "eval_samples_per_second": 151.598,
      "eval_steps_per_second": 7.774,
      "step": 1220
    },
    {
      "epoch": 3.52,
      "grad_norm": 3.78125,
      "learning_rate": 1.1622197118291982e-05,
      "loss": 0.1923,
      "step": 1230
    },
    {
      "epoch": 3.52,
      "eval_loss": 0.8377634882926941,
      "eval_runtime": 1.0302,
      "eval_samples_per_second": 151.425,
      "eval_steps_per_second": 7.765,
      "step": 1230
    },
    {
      "epoch": 3.55,
      "grad_norm": 3.46875,
      "learning_rate": 1.148099977347238e-05,
      "loss": 0.1879,
      "step": 1240
    },
    {
      "epoch": 3.55,
      "eval_loss": 0.8426746726036072,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.807,
      "eval_steps_per_second": 7.785,
      "step": 1240
    },
    {
      "epoch": 3.58,
      "grad_norm": 2.28125,
      "learning_rate": 1.1339499882267955e-05,
      "loss": 0.177,
      "step": 1250
    },
    {
      "epoch": 3.58,
      "eval_loss": 0.8470408916473389,
      "eval_runtime": 1.0317,
      "eval_samples_per_second": 151.209,
      "eval_steps_per_second": 7.754,
      "step": 1250
    },
    {
      "epoch": 3.61,
      "grad_norm": 3.234375,
      "learning_rate": 1.1197726351017052e-05,
      "loss": 0.1779,
      "step": 1260
    },
    {
      "epoch": 3.61,
      "eval_loss": 0.8529857993125916,
      "eval_runtime": 1.0272,
      "eval_samples_per_second": 151.863,
      "eval_steps_per_second": 7.788,
      "step": 1260
    },
    {
      "epoch": 3.64,
      "grad_norm": 2.5,
      "learning_rate": 1.1055708141958634e-05,
      "loss": 0.1921,
      "step": 1270
    },
    {
      "epoch": 3.64,
      "eval_loss": 0.8385663032531738,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.612,
      "eval_steps_per_second": 7.775,
      "step": 1270
    },
    {
      "epoch": 3.67,
      "grad_norm": 2.84375,
      "learning_rate": 1.091347426731573e-05,
      "loss": 0.181,
      "step": 1280
    },
    {
      "epoch": 3.67,
      "eval_loss": 0.8298706412315369,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.696,
      "eval_steps_per_second": 7.779,
      "step": 1280
    },
    {
      "epoch": 3.7,
      "grad_norm": 2.859375,
      "learning_rate": 1.0771053783368647e-05,
      "loss": 0.1741,
      "step": 1290
    },
    {
      "epoch": 3.7,
      "eval_loss": 0.8360311985015869,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.811,
      "eval_steps_per_second": 7.785,
      "step": 1290
    },
    {
      "epoch": 3.72,
      "grad_norm": 2.46875,
      "learning_rate": 1.0628475784519248e-05,
      "loss": 0.1729,
      "step": 1300
    },
    {
      "epoch": 3.72,
      "eval_loss": 0.8473387360572815,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.736,
      "eval_steps_per_second": 7.781,
      "step": 1300
    },
    {
      "epoch": 3.75,
      "grad_norm": 3.015625,
      "learning_rate": 1.0485769397347335e-05,
      "loss": 0.1778,
      "step": 1310
    },
    {
      "epoch": 3.75,
      "eval_loss": 0.8452015519142151,
      "eval_runtime": 1.028,
      "eval_samples_per_second": 151.746,
      "eval_steps_per_second": 7.782,
      "step": 1310
    },
    {
      "epoch": 3.78,
      "grad_norm": 3.515625,
      "learning_rate": 1.0342963774660566e-05,
      "loss": 0.1754,
      "step": 1320
    },
    {
      "epoch": 3.78,
      "eval_loss": 0.842901349067688,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.702,
      "eval_steps_per_second": 7.78,
      "step": 1320
    },
    {
      "epoch": 3.81,
      "grad_norm": 3.15625,
      "learning_rate": 1.0200088089538944e-05,
      "loss": 0.179,
      "step": 1330
    },
    {
      "epoch": 3.81,
      "eval_loss": 0.8430735468864441,
      "eval_runtime": 1.0292,
      "eval_samples_per_second": 151.567,
      "eval_steps_per_second": 7.773,
      "step": 1330
    },
    {
      "epoch": 3.84,
      "grad_norm": 2.203125,
      "learning_rate": 1.0057171529375192e-05,
      "loss": 0.1788,
      "step": 1340
    },
    {
      "epoch": 3.84,
      "eval_loss": 0.8360325694084167,
      "eval_runtime": 1.0296,
      "eval_samples_per_second": 151.513,
      "eval_steps_per_second": 7.77,
      "step": 1340
    },
    {
      "epoch": 3.87,
      "grad_norm": 2.453125,
      "learning_rate": 9.9142432899122e-06,
      "loss": 0.1776,
      "step": 1350
    },
    {
      "epoch": 3.87,
      "eval_loss": 0.8384283185005188,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.775,
      "eval_steps_per_second": 7.783,
      "step": 1350
    },
    {
      "epoch": 3.9,
      "grad_norm": 2.703125,
      "learning_rate": 9.77133256927877e-06,
      "loss": 0.1716,
      "step": 1360
    },
    {
      "epoch": 3.9,
      "eval_loss": 0.8421617150306702,
      "eval_runtime": 1.0297,
      "eval_samples_per_second": 151.496,
      "eval_steps_per_second": 7.769,
      "step": 1360
    },
    {
      "epoch": 3.93,
      "grad_norm": 2.625,
      "learning_rate": 9.628468562024858e-06,
      "loss": 0.1885,
      "step": 1370
    },
    {
      "epoch": 3.93,
      "eval_loss": 0.8416627645492554,
      "eval_runtime": 1.0264,
      "eval_samples_per_second": 151.987,
      "eval_steps_per_second": 7.794,
      "step": 1370
    },
    {
      "epoch": 3.95,
      "grad_norm": 3.140625,
      "learning_rate": 9.48568045315758e-06,
      "loss": 0.1858,
      "step": 1380
    },
    {
      "epoch": 3.95,
      "eval_loss": 0.8395534157752991,
      "eval_runtime": 1.0274,
      "eval_samples_per_second": 151.842,
      "eval_steps_per_second": 7.787,
      "step": 1380
    },
    {
      "epoch": 3.98,
      "grad_norm": 3.046875,
      "learning_rate": 9.342997412179123e-06,
      "loss": 0.1765,
      "step": 1390
    },
    {
      "epoch": 3.98,
      "eval_loss": 0.8416285514831543,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.814,
      "eval_steps_per_second": 7.785,
      "step": 1390
    },
    {
      "epoch": 4.01,
      "grad_norm": 1.84375,
      "learning_rate": 9.200448587127852e-06,
      "loss": 0.16,
      "step": 1400
    },
    {
      "epoch": 4.01,
      "eval_loss": 0.8607451319694519,
      "eval_runtime": 1.0298,
      "eval_samples_per_second": 151.48,
      "eval_steps_per_second": 7.768,
      "step": 1400
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.96875,
      "learning_rate": 9.058063098623808e-06,
      "loss": 0.1175,
      "step": 1410
    },
    {
      "epoch": 4.04,
      "eval_loss": 0.9675983786582947,
      "eval_runtime": 1.0302,
      "eval_samples_per_second": 151.433,
      "eval_steps_per_second": 7.766,
      "step": 1410
    },
    {
      "epoch": 4.07,
      "grad_norm": 2.9375,
      "learning_rate": 8.915870033919785e-06,
      "loss": 0.1266,
      "step": 1420
    },
    {
      "epoch": 4.07,
      "eval_loss": 0.9722825884819031,
      "eval_runtime": 1.0285,
      "eval_samples_per_second": 151.675,
      "eval_steps_per_second": 7.778,
      "step": 1420
    },
    {
      "epoch": 4.1,
      "grad_norm": 2.140625,
      "learning_rate": 8.77389844095923e-06,
      "loss": 0.1232,
      "step": 1430
    },
    {
      "epoch": 4.1,
      "eval_loss": 0.9478564262390137,
      "eval_runtime": 1.0286,
      "eval_samples_per_second": 151.668,
      "eval_steps_per_second": 7.778,
      "step": 1430
    },
    {
      "epoch": 4.13,
      "grad_norm": 1.9296875,
      "learning_rate": 8.632177322442198e-06,
      "loss": 0.1202,
      "step": 1440
    },
    {
      "epoch": 4.13,
      "eval_loss": 0.958242654800415,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.538,
      "eval_steps_per_second": 7.771,
      "step": 1440
    },
    {
      "epoch": 4.15,
      "grad_norm": 2.234375,
      "learning_rate": 8.490735629900497e-06,
      "loss": 0.1163,
      "step": 1450
    },
    {
      "epoch": 4.15,
      "eval_loss": 0.9646920561790466,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.781,
      "eval_steps_per_second": 7.784,
      "step": 1450
    },
    {
      "epoch": 4.18,
      "grad_norm": 2.1875,
      "learning_rate": 8.349602257783347e-06,
      "loss": 0.1179,
      "step": 1460
    },
    {
      "epoch": 4.18,
      "eval_loss": 0.968169093132019,
      "eval_runtime": 1.028,
      "eval_samples_per_second": 151.752,
      "eval_steps_per_second": 7.782,
      "step": 1460
    },
    {
      "epoch": 4.21,
      "grad_norm": 2.828125,
      "learning_rate": 8.208806037554645e-06,
      "loss": 0.1281,
      "step": 1470
    },
    {
      "epoch": 4.21,
      "eval_loss": 0.9741702079772949,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.895,
      "eval_steps_per_second": 7.789,
      "step": 1470
    },
    {
      "epoch": 4.24,
      "grad_norm": 2.9375,
      "learning_rate": 8.068375731803151e-06,
      "loss": 0.1259,
      "step": 1480
    },
    {
      "epoch": 4.24,
      "eval_loss": 0.9701966643333435,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.902,
      "eval_steps_per_second": 7.79,
      "step": 1480
    },
    {
      "epoch": 4.27,
      "grad_norm": 2.578125,
      "learning_rate": 7.928340028366687e-06,
      "loss": 0.1242,
      "step": 1490
    },
    {
      "epoch": 4.27,
      "eval_loss": 0.9706182479858398,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.719,
      "eval_steps_per_second": 7.78,
      "step": 1490
    },
    {
      "epoch": 4.3,
      "grad_norm": 2.953125,
      "learning_rate": 7.788727534471655e-06,
      "loss": 0.1237,
      "step": 1500
    },
    {
      "epoch": 4.3,
      "eval_loss": 0.9681195020675659,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.775,
      "eval_steps_per_second": 7.783,
      "step": 1500
    },
    {
      "epoch": 4.33,
      "grad_norm": 2.375,
      "learning_rate": 7.649566770889003e-06,
      "loss": 0.1213,
      "step": 1510
    },
    {
      "epoch": 4.33,
      "eval_loss": 0.9657440185546875,
      "eval_runtime": 1.0262,
      "eval_samples_per_second": 152.022,
      "eval_steps_per_second": 7.796,
      "step": 1510
    },
    {
      "epoch": 4.36,
      "grad_norm": 2.34375,
      "learning_rate": 7.510886166107833e-06,
      "loss": 0.1263,
      "step": 1520
    },
    {
      "epoch": 4.36,
      "eval_loss": 0.9629148840904236,
      "eval_runtime": 1.0251,
      "eval_samples_per_second": 152.175,
      "eval_steps_per_second": 7.804,
      "step": 1520
    },
    {
      "epoch": 4.38,
      "grad_norm": 2.5625,
      "learning_rate": 7.3727140505279045e-06,
      "loss": 0.1371,
      "step": 1530
    },
    {
      "epoch": 4.38,
      "eval_loss": 0.9612699747085571,
      "eval_runtime": 1.0271,
      "eval_samples_per_second": 151.883,
      "eval_steps_per_second": 7.789,
      "step": 1530
    },
    {
      "epoch": 4.41,
      "grad_norm": 2.328125,
      "learning_rate": 7.235078650672141e-06,
      "loss": 0.1247,
      "step": 1540
    },
    {
      "epoch": 4.41,
      "eval_loss": 0.9614297151565552,
      "eval_runtime": 1.0264,
      "eval_samples_per_second": 151.994,
      "eval_steps_per_second": 7.795,
      "step": 1540
    },
    {
      "epoch": 4.44,
      "grad_norm": 2.46875,
      "learning_rate": 7.098008083420359e-06,
      "loss": 0.1283,
      "step": 1550
    },
    {
      "epoch": 4.44,
      "eval_loss": 0.9646217226982117,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.9,
      "eval_steps_per_second": 7.79,
      "step": 1550
    },
    {
      "epoch": 4.47,
      "grad_norm": 2.671875,
      "learning_rate": 6.961530350265427e-06,
      "loss": 0.1299,
      "step": 1560
    },
    {
      "epoch": 4.47,
      "eval_loss": 0.9594274759292603,
      "eval_runtime": 1.028,
      "eval_samples_per_second": 151.748,
      "eval_steps_per_second": 7.782,
      "step": 1560
    },
    {
      "epoch": 4.5,
      "grad_norm": 2.390625,
      "learning_rate": 6.825673331592952e-06,
      "loss": 0.128,
      "step": 1570
    },
    {
      "epoch": 4.5,
      "eval_loss": 0.9570373892784119,
      "eval_runtime": 1.0284,
      "eval_samples_per_second": 151.689,
      "eval_steps_per_second": 7.779,
      "step": 1570
    },
    {
      "epoch": 4.53,
      "grad_norm": 1.8203125,
      "learning_rate": 6.6904647809857484e-06,
      "loss": 0.1214,
      "step": 1580
    },
    {
      "epoch": 4.53,
      "eval_loss": 0.9542241096496582,
      "eval_runtime": 1.0301,
      "eval_samples_per_second": 151.448,
      "eval_steps_per_second": 7.767,
      "step": 1580
    },
    {
      "epoch": 4.56,
      "grad_norm": 3.125,
      "learning_rate": 6.55593231955417e-06,
      "loss": 0.1231,
      "step": 1590
    },
    {
      "epoch": 4.56,
      "eval_loss": 0.9523473978042603,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.54,
      "eval_steps_per_second": 7.771,
      "step": 1590
    },
    {
      "epoch": 4.58,
      "grad_norm": 2.3125,
      "learning_rate": 6.42210343029354e-06,
      "loss": 0.1224,
      "step": 1600
    },
    {
      "epoch": 4.58,
      "eval_loss": 0.9537341594696045,
      "eval_runtime": 1.0282,
      "eval_samples_per_second": 151.725,
      "eval_steps_per_second": 7.781,
      "step": 1600
    },
    {
      "epoch": 4.61,
      "grad_norm": 2.171875,
      "learning_rate": 6.289005452469778e-06,
      "loss": 0.1273,
      "step": 1610
    },
    {
      "epoch": 4.61,
      "eval_loss": 0.9624239802360535,
      "eval_runtime": 1.0299,
      "eval_samples_per_second": 151.472,
      "eval_steps_per_second": 7.768,
      "step": 1610
    },
    {
      "epoch": 4.64,
      "grad_norm": 3.28125,
      "learning_rate": 6.156665576034383e-06,
      "loss": 0.1304,
      "step": 1620
    },
    {
      "epoch": 4.64,
      "eval_loss": 0.9658766388893127,
      "eval_runtime": 1.0299,
      "eval_samples_per_second": 151.466,
      "eval_steps_per_second": 7.768,
      "step": 1620
    },
    {
      "epoch": 4.67,
      "grad_norm": 2.234375,
      "learning_rate": 6.025110836069939e-06,
      "loss": 0.1258,
      "step": 1630
    },
    {
      "epoch": 4.67,
      "eval_loss": 0.9638926386833191,
      "eval_runtime": 1.0292,
      "eval_samples_per_second": 151.567,
      "eval_steps_per_second": 7.773,
      "step": 1630
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.7109375,
      "learning_rate": 5.89436810726725e-06,
      "loss": 0.1184,
      "step": 1640
    },
    {
      "epoch": 4.7,
      "eval_loss": 0.965703547000885,
      "eval_runtime": 1.0286,
      "eval_samples_per_second": 151.668,
      "eval_steps_per_second": 7.778,
      "step": 1640
    },
    {
      "epoch": 4.73,
      "grad_norm": 2.265625,
      "learning_rate": 5.764464098435216e-06,
      "loss": 0.1231,
      "step": 1650
    },
    {
      "epoch": 4.73,
      "eval_loss": 0.9644750952720642,
      "eval_runtime": 1.0266,
      "eval_samples_per_second": 151.951,
      "eval_steps_per_second": 7.792,
      "step": 1650
    },
    {
      "epoch": 4.76,
      "grad_norm": 2.71875,
      "learning_rate": 5.635425347044639e-06,
      "loss": 0.1278,
      "step": 1660
    },
    {
      "epoch": 4.76,
      "eval_loss": 0.9634093046188354,
      "eval_runtime": 1.0271,
      "eval_samples_per_second": 151.887,
      "eval_steps_per_second": 7.789,
      "step": 1660
    },
    {
      "epoch": 4.79,
      "grad_norm": 2.296875,
      "learning_rate": 5.507278213807009e-06,
      "loss": 0.1294,
      "step": 1670
    },
    {
      "epoch": 4.79,
      "eval_loss": 0.9675451517105103,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.623,
      "eval_steps_per_second": 7.776,
      "step": 1670
    },
    {
      "epoch": 4.81,
      "grad_norm": 2.609375,
      "learning_rate": 5.380048877289381e-06,
      "loss": 0.1236,
      "step": 1680
    },
    {
      "epoch": 4.81,
      "eval_loss": 0.9684876799583435,
      "eval_runtime": 1.027,
      "eval_samples_per_second": 151.904,
      "eval_steps_per_second": 7.79,
      "step": 1680
    },
    {
      "epoch": 4.84,
      "grad_norm": 2.546875,
      "learning_rate": 5.253763328566494e-06,
      "loss": 0.1237,
      "step": 1690
    },
    {
      "epoch": 4.84,
      "eval_loss": 0.968744158744812,
      "eval_runtime": 1.0279,
      "eval_samples_per_second": 151.763,
      "eval_steps_per_second": 7.783,
      "step": 1690
    },
    {
      "epoch": 4.87,
      "grad_norm": 2.515625,
      "learning_rate": 5.128447365911185e-06,
      "loss": 0.1274,
      "step": 1700
    },
    {
      "epoch": 4.87,
      "eval_loss": 0.9674803614616394,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.797,
      "eval_steps_per_second": 7.784,
      "step": 1700
    },
    {
      "epoch": 4.9,
      "grad_norm": 3.546875,
      "learning_rate": 5.004126589524174e-06,
      "loss": 0.1307,
      "step": 1710
    },
    {
      "epoch": 4.9,
      "eval_loss": 0.9692357182502747,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.793,
      "eval_steps_per_second": 7.784,
      "step": 1710
    },
    {
      "epoch": 4.93,
      "grad_norm": 2.0625,
      "learning_rate": 4.880826396304312e-06,
      "loss": 0.1161,
      "step": 1720
    },
    {
      "epoch": 4.93,
      "eval_loss": 0.9701640605926514,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.617,
      "eval_steps_per_second": 7.775,
      "step": 1720
    },
    {
      "epoch": 4.96,
      "grad_norm": 2.171875,
      "learning_rate": 4.75857197466039e-06,
      "loss": 0.1221,
      "step": 1730
    },
    {
      "epoch": 4.96,
      "eval_loss": 0.9716487526893616,
      "eval_runtime": 1.0275,
      "eval_samples_per_second": 151.819,
      "eval_steps_per_second": 7.786,
      "step": 1730
    },
    {
      "epoch": 4.99,
      "grad_norm": 2.03125,
      "learning_rate": 4.6373882993655e-06,
      "loss": 0.1238,
      "step": 1740
    },
    {
      "epoch": 4.99,
      "eval_loss": 0.9733632206916809,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.736,
      "eval_steps_per_second": 7.781,
      "step": 1740
    },
    {
      "epoch": 5.01,
      "grad_norm": 1.6953125,
      "learning_rate": 4.5173001264550665e-06,
      "loss": 0.1129,
      "step": 1750
    },
    {
      "epoch": 5.01,
      "eval_loss": 0.9793078899383545,
      "eval_runtime": 1.0286,
      "eval_samples_per_second": 151.66,
      "eval_steps_per_second": 7.777,
      "step": 1750
    },
    {
      "epoch": 5.04,
      "grad_norm": 2.140625,
      "learning_rate": 4.398331988169559e-06,
      "loss": 0.1042,
      "step": 1760
    },
    {
      "epoch": 5.04,
      "eval_loss": 0.9991005063056946,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.816,
      "eval_steps_per_second": 7.785,
      "step": 1760
    },
    {
      "epoch": 5.07,
      "grad_norm": 2.234375,
      "learning_rate": 4.280508187942913e-06,
      "loss": 0.1076,
      "step": 1770
    },
    {
      "epoch": 5.07,
      "eval_loss": 1.0183870792388916,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.815,
      "eval_steps_per_second": 7.785,
      "step": 1770
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.71875,
      "learning_rate": 4.1638527954376865e-06,
      "loss": 0.1061,
      "step": 1780
    },
    {
      "epoch": 5.1,
      "eval_loss": 1.0275365114212036,
      "eval_runtime": 1.0276,
      "eval_samples_per_second": 151.803,
      "eval_steps_per_second": 7.785,
      "step": 1780
    },
    {
      "epoch": 5.13,
      "grad_norm": 1.8046875,
      "learning_rate": 4.048389641628002e-06,
      "loss": 0.1016,
      "step": 1790
    },
    {
      "epoch": 5.13,
      "eval_loss": 1.0309237241744995,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.702,
      "eval_steps_per_second": 7.78,
      "step": 1790
    },
    {
      "epoch": 5.16,
      "grad_norm": 2.015625,
      "learning_rate": 3.934142313931215e-06,
      "loss": 0.1067,
      "step": 1800
    },
    {
      "epoch": 5.16,
      "eval_loss": 1.0333725214004517,
      "eval_runtime": 1.0272,
      "eval_samples_per_second": 151.866,
      "eval_steps_per_second": 7.788,
      "step": 1800
    },
    {
      "epoch": 5.19,
      "grad_norm": 2.3125,
      "learning_rate": 3.821134151389367e-06,
      "loss": 0.104,
      "step": 1810
    },
    {
      "epoch": 5.19,
      "eval_loss": 1.0376129150390625,
      "eval_runtime": 1.0294,
      "eval_samples_per_second": 151.541,
      "eval_steps_per_second": 7.771,
      "step": 1810
    },
    {
      "epoch": 5.21,
      "grad_norm": 2.328125,
      "learning_rate": 3.7093882399013504e-06,
      "loss": 0.1086,
      "step": 1820
    },
    {
      "epoch": 5.21,
      "eval_loss": 1.0411092042922974,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.8,
      "eval_steps_per_second": 7.785,
      "step": 1820
    },
    {
      "epoch": 5.24,
      "grad_norm": 1.5625,
      "learning_rate": 3.5989274075068005e-06,
      "loss": 0.1025,
      "step": 1830
    },
    {
      "epoch": 5.24,
      "eval_loss": 1.0420621633529663,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.71,
      "eval_steps_per_second": 7.78,
      "step": 1830
    },
    {
      "epoch": 5.27,
      "grad_norm": 2.421875,
      "learning_rate": 3.489774219722689e-06,
      "loss": 0.107,
      "step": 1840
    },
    {
      "epoch": 5.27,
      "eval_loss": 1.0420392751693726,
      "eval_runtime": 1.0306,
      "eval_samples_per_second": 151.368,
      "eval_steps_per_second": 7.762,
      "step": 1840
    },
    {
      "epoch": 5.3,
      "grad_norm": 2.0,
      "learning_rate": 3.3819509749334843e-06,
      "loss": 0.0982,
      "step": 1850
    },
    {
      "epoch": 5.3,
      "eval_loss": 1.0423657894134521,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.71,
      "eval_steps_per_second": 7.78,
      "step": 1850
    },
    {
      "epoch": 5.33,
      "grad_norm": 1.6796875,
      "learning_rate": 3.2754796998359427e-06,
      "loss": 0.1067,
      "step": 1860
    },
    {
      "epoch": 5.33,
      "eval_loss": 1.0411757230758667,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.776,
      "eval_steps_per_second": 7.783,
      "step": 1860
    },
    {
      "epoch": 5.36,
      "grad_norm": 1.96875,
      "learning_rate": 3.1703821449393856e-06,
      "loss": 0.1079,
      "step": 1870
    },
    {
      "epoch": 5.36,
      "eval_loss": 1.04030179977417,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.741,
      "eval_steps_per_second": 7.782,
      "step": 1870
    },
    {
      "epoch": 5.39,
      "grad_norm": 1.2109375,
      "learning_rate": 3.066679780122368e-06,
      "loss": 0.1027,
      "step": 1880
    },
    {
      "epoch": 5.39,
      "eval_loss": 1.0408295392990112,
      "eval_runtime": 1.0296,
      "eval_samples_per_second": 151.52,
      "eval_steps_per_second": 7.77,
      "step": 1880
    },
    {
      "epoch": 5.42,
      "grad_norm": 1.75,
      "learning_rate": 2.964393790246728e-06,
      "loss": 0.1035,
      "step": 1890
    },
    {
      "epoch": 5.42,
      "eval_loss": 1.0431201457977295,
      "eval_runtime": 1.0283,
      "eval_samples_per_second": 151.702,
      "eval_steps_per_second": 7.78,
      "step": 1890
    },
    {
      "epoch": 5.44,
      "grad_norm": 1.5390625,
      "learning_rate": 2.863545070829804e-06,
      "loss": 0.1021,
      "step": 1900
    },
    {
      "epoch": 5.44,
      "eval_loss": 1.044161081314087,
      "eval_runtime": 1.028,
      "eval_samples_per_second": 151.75,
      "eval_steps_per_second": 7.782,
      "step": 1900
    },
    {
      "epoch": 5.47,
      "grad_norm": 2.15625,
      "learning_rate": 2.7641542237758134e-06,
      "loss": 0.1031,
      "step": 1910
    },
    {
      "epoch": 5.47,
      "eval_loss": 1.044289231300354,
      "eval_runtime": 1.031,
      "eval_samples_per_second": 151.303,
      "eval_steps_per_second": 7.759,
      "step": 1910
    },
    {
      "epoch": 5.5,
      "grad_norm": 2.59375,
      "learning_rate": 2.666241553167175e-06,
      "loss": 0.1064,
      "step": 1920
    },
    {
      "epoch": 5.5,
      "eval_loss": 1.0444979667663574,
      "eval_runtime": 1.0289,
      "eval_samples_per_second": 151.613,
      "eval_steps_per_second": 7.775,
      "step": 1920
    },
    {
      "epoch": 5.53,
      "grad_norm": 1.796875,
      "learning_rate": 2.569827061116684e-06,
      "loss": 0.1083,
      "step": 1930
    },
    {
      "epoch": 5.53,
      "eval_loss": 1.044089436531067,
      "eval_runtime": 1.0275,
      "eval_samples_per_second": 151.832,
      "eval_steps_per_second": 7.786,
      "step": 1930
    },
    {
      "epoch": 5.56,
      "grad_norm": 1.984375,
      "learning_rate": 2.474930443681377e-06,
      "loss": 0.1111,
      "step": 1940
    },
    {
      "epoch": 5.56,
      "eval_loss": 1.0448068380355835,
      "eval_runtime": 1.0281,
      "eval_samples_per_second": 151.742,
      "eval_steps_per_second": 7.782,
      "step": 1940
    },
    {
      "epoch": 5.59,
      "grad_norm": 2.109375,
      "learning_rate": 2.3815710868389274e-06,
      "loss": 0.1093,
      "step": 1950
    },
    {
      "epoch": 5.59,
      "eval_loss": 1.045329213142395,
      "eval_runtime": 1.0265,
      "eval_samples_per_second": 151.979,
      "eval_steps_per_second": 7.794,
      "step": 1950
    },
    {
      "epoch": 5.62,
      "grad_norm": 2.203125,
      "learning_rate": 2.2897680625273623e-06,
      "loss": 0.1062,
      "step": 1960
    },
    {
      "epoch": 5.62,
      "eval_loss": 1.0454120635986328,
      "eval_runtime": 1.0278,
      "eval_samples_per_second": 151.787,
      "eval_steps_per_second": 7.784,
      "step": 1960
    },
    {
      "epoch": 5.64,
      "grad_norm": 2.015625,
      "learning_rate": 2.199540124748957e-06,
      "loss": 0.1072,
      "step": 1970
    },
    {
      "epoch": 5.64,
      "eval_loss": 1.0454870462417603,
      "eval_runtime": 1.0277,
      "eval_samples_per_second": 151.79,
      "eval_steps_per_second": 7.784,
      "step": 1970
    },
    {
      "epoch": 5.67,
      "grad_norm": 1.90625,
      "learning_rate": 2.110905705739069e-06,
      "loss": 0.1046,
      "step": 1980
    },
    {
      "epoch": 5.67,
      "eval_loss": 1.0458413362503052,
      "eval_runtime": 1.0268,
      "eval_samples_per_second": 151.927,
      "eval_steps_per_second": 7.791,
      "step": 1980
    },
    {
      "epoch": 5.7,
      "grad_norm": 1.984375,
      "learning_rate": 2.0238829122006944e-06,
      "loss": 0.1078,
      "step": 1990
    },
    {
      "epoch": 5.7,
      "eval_loss": 1.0452886819839478,
      "eval_runtime": 1.0274,
      "eval_samples_per_second": 151.835,
      "eval_steps_per_second": 7.786,
      "step": 1990
    },
    {
      "epoch": 5.73,
      "grad_norm": 1.75,
      "learning_rate": 1.9384895216055533e-06,
      "loss": 0.1076,
      "step": 2000
    },
    {
      "epoch": 5.73,
      "eval_loss": 1.0452741384506226,
      "eval_runtime": 1.0267,
      "eval_samples_per_second": 151.95,
      "eval_steps_per_second": 7.792,
      "step": 2000
    }
  ],
  "logging_steps": 10,
  "max_steps": 2443,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "save_steps": 500,
  "total_flos": 9.135415373280051e+16,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}