{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.0,
  "eval_steps": 500,
  "global_step": 1400,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.007142857142857143,
      "grad_norm": 2.519733190536499,
      "learning_rate": 9.976190476190477e-05,
      "loss": 1.2999,
      "step": 10
    },
    {
      "epoch": 0.014285714285714285,
      "grad_norm": 2.4232161045074463,
      "learning_rate": 9.952380952380953e-05,
      "loss": 1.1228,
      "step": 20
    },
    {
      "epoch": 0.02142857142857143,
      "grad_norm": 2.7126998901367188,
      "learning_rate": 9.92857142857143e-05,
      "loss": 0.9913,
      "step": 30
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 7.3811235427856445,
      "learning_rate": 9.904761904761905e-05,
      "loss": 0.9306,
      "step": 40
    },
    {
      "epoch": 0.03571428571428571,
      "grad_norm": 6.5697197914123535,
      "learning_rate": 9.880952380952381e-05,
      "loss": 0.7457,
      "step": 50
    },
    {
      "epoch": 0.04285714285714286,
      "grad_norm": 1.2972687482833862,
      "learning_rate": 9.857142857142858e-05,
      "loss": 0.6408,
      "step": 60
    },
    {
      "epoch": 0.05,
      "grad_norm": 3.600625991821289,
      "learning_rate": 9.833333333333333e-05,
      "loss": 0.6541,
      "step": 70
    },
    {
      "epoch": 0.05714285714285714,
      "grad_norm": 1.5594621896743774,
      "learning_rate": 9.80952380952381e-05,
      "loss": 0.538,
      "step": 80
    },
    {
      "epoch": 0.06428571428571428,
      "grad_norm": 4.595429420471191,
      "learning_rate": 9.785714285714286e-05,
      "loss": 0.556,
      "step": 90
    },
    {
      "epoch": 0.07142857142857142,
      "grad_norm": 2.620643138885498,
      "learning_rate": 9.761904761904762e-05,
      "loss": 0.5521,
      "step": 100
    },
    {
      "epoch": 0.07857142857142857,
      "grad_norm": 3.503641366958618,
      "learning_rate": 9.738095238095239e-05,
      "loss": 0.5912,
      "step": 110
    },
    {
      "epoch": 0.08571428571428572,
      "grad_norm": 1.5058848857879639,
      "learning_rate": 9.714285714285715e-05,
      "loss": 0.5421,
      "step": 120
    },
    {
      "epoch": 0.09285714285714286,
      "grad_norm": 3.7034990787506104,
      "learning_rate": 9.69047619047619e-05,
      "loss": 0.5568,
      "step": 130
    },
    {
      "epoch": 0.1,
      "grad_norm": 3.714170455932617,
      "learning_rate": 9.666666666666667e-05,
      "loss": 0.5286,
      "step": 140
    },
    {
      "epoch": 0.10714285714285714,
      "grad_norm": 2.069328546524048,
      "learning_rate": 9.642857142857143e-05,
      "loss": 0.4755,
      "step": 150
    },
    {
      "epoch": 0.11428571428571428,
      "grad_norm": 1.2878583669662476,
      "learning_rate": 9.61904761904762e-05,
      "loss": 0.4817,
      "step": 160
    },
    {
      "epoch": 0.12142857142857143,
      "grad_norm": 3.1945483684539795,
      "learning_rate": 9.595238095238096e-05,
      "loss": 0.5035,
      "step": 170
    },
    {
      "epoch": 0.12857142857142856,
      "grad_norm": 1.9643166065216064,
      "learning_rate": 9.571428571428573e-05,
      "loss": 0.4932,
      "step": 180
    },
    {
      "epoch": 0.1357142857142857,
      "grad_norm": 3.5002968311309814,
      "learning_rate": 9.547619047619049e-05,
      "loss": 0.5531,
      "step": 190
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 4.7095794677734375,
      "learning_rate": 9.523809523809524e-05,
      "loss": 0.4601,
      "step": 200
    },
    {
      "epoch": 0.15,
      "grad_norm": 1.4850050210952759,
      "learning_rate": 9.5e-05,
      "loss": 0.432,
      "step": 210
    },
    {
      "epoch": 0.15714285714285714,
      "grad_norm": 2.860565423965454,
      "learning_rate": 9.476190476190476e-05,
      "loss": 0.4699,
      "step": 220
    },
    {
      "epoch": 0.16428571428571428,
      "grad_norm": 1.8467603921890259,
      "learning_rate": 9.452380952380952e-05,
      "loss": 0.42,
      "step": 230
    },
    {
      "epoch": 0.17142857142857143,
      "grad_norm": 1.6448352336883545,
      "learning_rate": 9.428571428571429e-05,
      "loss": 0.431,
      "step": 240
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 1.8279732465744019,
      "learning_rate": 9.404761904761905e-05,
      "loss": 0.4413,
      "step": 250
    },
    {
      "epoch": 0.18571428571428572,
      "grad_norm": 3.0112857818603516,
      "learning_rate": 9.380952380952381e-05,
      "loss": 0.4263,
      "step": 260
    },
    {
      "epoch": 0.19285714285714287,
      "grad_norm": 1.7717293500900269,
      "learning_rate": 9.357142857142858e-05,
      "loss": 0.4829,
      "step": 270
    },
    {
      "epoch": 0.2,
      "grad_norm": 4.515851974487305,
      "learning_rate": 9.333333333333334e-05,
      "loss": 0.4194,
      "step": 280
    },
    {
      "epoch": 0.20714285714285716,
      "grad_norm": 3.998619556427002,
      "learning_rate": 9.309523809523811e-05,
      "loss": 0.4422,
      "step": 290
    },
    {
      "epoch": 0.21428571428571427,
      "grad_norm": 3.5681912899017334,
      "learning_rate": 9.285714285714286e-05,
      "loss": 0.4677,
      "step": 300
    },
    {
      "epoch": 0.22142857142857142,
      "grad_norm": 3.286348342895508,
      "learning_rate": 9.261904761904762e-05,
      "loss": 0.4152,
      "step": 310
    },
    {
      "epoch": 0.22857142857142856,
      "grad_norm": 2.018721580505371,
      "learning_rate": 9.238095238095239e-05,
      "loss": 0.5065,
      "step": 320
    },
    {
      "epoch": 0.2357142857142857,
      "grad_norm": 1.8489034175872803,
      "learning_rate": 9.214285714285714e-05,
      "loss": 0.392,
      "step": 330
    },
    {
      "epoch": 0.24285714285714285,
      "grad_norm": 1.3958369493484497,
      "learning_rate": 9.19047619047619e-05,
      "loss": 0.3988,
      "step": 340
    },
    {
      "epoch": 0.25,
      "grad_norm": 2.9772961139678955,
      "learning_rate": 9.166666666666667e-05,
      "loss": 0.3955,
      "step": 350
    },
    {
      "epoch": 0.2571428571428571,
      "grad_norm": 3.2752914428710938,
      "learning_rate": 9.142857142857143e-05,
      "loss": 0.3998,
      "step": 360
    },
    {
      "epoch": 0.2642857142857143,
      "grad_norm": 1.8811057806015015,
      "learning_rate": 9.11904761904762e-05,
      "loss": 0.4775,
      "step": 370
    },
    {
      "epoch": 0.2714285714285714,
      "grad_norm": 2.3310513496398926,
      "learning_rate": 9.095238095238096e-05,
      "loss": 0.3703,
      "step": 380
    },
    {
      "epoch": 0.2785714285714286,
      "grad_norm": 2.967597484588623,
      "learning_rate": 9.071428571428571e-05,
      "loss": 0.4177,
      "step": 390
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 2.671673536300659,
      "learning_rate": 9.047619047619048e-05,
      "loss": 0.3823,
      "step": 400
    },
    {
      "epoch": 0.29285714285714287,
      "grad_norm": 4.647867679595947,
      "learning_rate": 9.023809523809524e-05,
      "loss": 0.3888,
      "step": 410
    },
    {
      "epoch": 0.3,
      "grad_norm": 3.4788432121276855,
      "learning_rate": 9e-05,
      "loss": 0.3856,
      "step": 420
    },
    {
      "epoch": 0.30714285714285716,
      "grad_norm": 3.308420181274414,
      "learning_rate": 8.976190476190477e-05,
      "loss": 0.4143,
      "step": 430
    },
    {
      "epoch": 0.3142857142857143,
      "grad_norm": 3.499880075454712,
      "learning_rate": 8.952380952380953e-05,
      "loss": 0.4857,
      "step": 440
    },
    {
      "epoch": 0.32142857142857145,
      "grad_norm": 1.7907614707946777,
      "learning_rate": 8.92857142857143e-05,
      "loss": 0.356,
      "step": 450
    },
    {
      "epoch": 0.32857142857142857,
      "grad_norm": 3.0111501216888428,
      "learning_rate": 8.904761904761905e-05,
      "loss": 0.358,
      "step": 460
    },
    {
      "epoch": 0.3357142857142857,
      "grad_norm": 2.1389689445495605,
      "learning_rate": 8.880952380952381e-05,
      "loss": 0.3331,
      "step": 470
    },
    {
      "epoch": 0.34285714285714286,
      "grad_norm": 1.8007326126098633,
      "learning_rate": 8.857142857142857e-05,
      "loss": 0.3148,
      "step": 480
    },
    {
      "epoch": 0.35,
      "grad_norm": 3.603548288345337,
      "learning_rate": 8.833333333333333e-05,
      "loss": 0.4513,
      "step": 490
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 2.270984411239624,
      "learning_rate": 8.80952380952381e-05,
      "loss": 0.3392,
      "step": 500
    },
    {
      "epoch": 0.36428571428571427,
      "grad_norm": 1.5367858409881592,
      "learning_rate": 8.785714285714286e-05,
      "loss": 0.3665,
      "step": 510
    },
    {
      "epoch": 0.37142857142857144,
      "grad_norm": 1.5397262573242188,
      "learning_rate": 8.761904761904762e-05,
      "loss": 0.3875,
      "step": 520
    },
    {
      "epoch": 0.37857142857142856,
      "grad_norm": 3.6661577224731445,
      "learning_rate": 8.738095238095239e-05,
      "loss": 0.3617,
      "step": 530
    },
    {
      "epoch": 0.38571428571428573,
      "grad_norm": 2.5946245193481445,
      "learning_rate": 8.714285714285715e-05,
      "loss": 0.3318,
      "step": 540
    },
    {
      "epoch": 0.39285714285714285,
      "grad_norm": 2.508056640625,
      "learning_rate": 8.690476190476192e-05,
      "loss": 0.3681,
      "step": 550
    },
    {
      "epoch": 0.4,
      "grad_norm": 5.406072616577148,
      "learning_rate": 8.666666666666667e-05,
      "loss": 0.3589,
      "step": 560
    },
    {
      "epoch": 0.40714285714285714,
      "grad_norm": 2.0363543033599854,
      "learning_rate": 8.642857142857143e-05,
      "loss": 0.4328,
      "step": 570
    },
    {
      "epoch": 0.4142857142857143,
      "grad_norm": 1.796236515045166,
      "learning_rate": 8.61904761904762e-05,
      "loss": 0.4005,
      "step": 580
    },
    {
      "epoch": 0.42142857142857143,
      "grad_norm": 5.6172943115234375,
      "learning_rate": 8.595238095238096e-05,
      "loss": 0.3414,
      "step": 590
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 2.953678846359253,
      "learning_rate": 8.571428571428571e-05,
      "loss": 0.3917,
      "step": 600
    },
    {
      "epoch": 0.4357142857142857,
      "grad_norm": 1.8276687860488892,
      "learning_rate": 8.547619047619048e-05,
      "loss": 0.4057,
      "step": 610
    },
    {
      "epoch": 0.44285714285714284,
      "grad_norm": 1.8668811321258545,
      "learning_rate": 8.523809523809524e-05,
      "loss": 0.4153,
      "step": 620
    },
    {
      "epoch": 0.45,
      "grad_norm": 2.87809419631958,
      "learning_rate": 8.5e-05,
      "loss": 0.3911,
      "step": 630
    },
    {
      "epoch": 0.45714285714285713,
      "grad_norm": 4.863674640655518,
      "learning_rate": 8.476190476190477e-05,
      "loss": 0.4458,
      "step": 640
    },
    {
      "epoch": 0.4642857142857143,
      "grad_norm": 1.7150845527648926,
      "learning_rate": 8.452380952380952e-05,
      "loss": 0.3951,
      "step": 650
    },
    {
      "epoch": 0.4714285714285714,
      "grad_norm": 2.0149145126342773,
      "learning_rate": 8.428571428571429e-05,
      "loss": 0.3238,
      "step": 660
    },
    {
      "epoch": 0.4785714285714286,
      "grad_norm": 2.3146402835845947,
      "learning_rate": 8.404761904761905e-05,
      "loss": 0.3129,
      "step": 670
    },
    {
      "epoch": 0.4857142857142857,
      "grad_norm": 1.8723537921905518,
      "learning_rate": 8.380952380952382e-05,
      "loss": 0.2969,
      "step": 680
    },
    {
      "epoch": 0.4928571428571429,
      "grad_norm": 1.6206679344177246,
      "learning_rate": 8.357142857142858e-05,
      "loss": 0.3673,
      "step": 690
    },
    {
      "epoch": 0.5,
      "grad_norm": 2.490722179412842,
      "learning_rate": 8.333333333333334e-05,
      "loss": 0.3493,
      "step": 700
    },
    {
      "epoch": 0.5071428571428571,
      "grad_norm": 1.7868791818618774,
      "learning_rate": 8.309523809523811e-05,
      "loss": 0.423,
      "step": 710
    },
    {
      "epoch": 0.5142857142857142,
      "grad_norm": 2.3152015209198,
      "learning_rate": 8.285714285714287e-05,
      "loss": 0.3448,
      "step": 720
    },
    {
      "epoch": 0.5214285714285715,
      "grad_norm": 1.934523105621338,
      "learning_rate": 8.261904761904762e-05,
      "loss": 0.372,
      "step": 730
    },
    {
      "epoch": 0.5285714285714286,
      "grad_norm": 2.421624183654785,
      "learning_rate": 8.238095238095238e-05,
      "loss": 0.3602,
      "step": 740
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 2.0156285762786865,
      "learning_rate": 8.214285714285714e-05,
      "loss": 0.334,
      "step": 750
    },
    {
      "epoch": 0.5428571428571428,
      "grad_norm": 1.6513855457305908,
      "learning_rate": 8.19047619047619e-05,
      "loss": 0.3553,
      "step": 760
    },
    {
      "epoch": 0.55,
      "grad_norm": 1.950451374053955,
      "learning_rate": 8.166666666666667e-05,
      "loss": 0.3696,
      "step": 770
    },
    {
      "epoch": 0.5571428571428572,
      "grad_norm": 2.4662423133850098,
      "learning_rate": 8.142857142857143e-05,
      "loss": 0.3232,
      "step": 780
    },
    {
      "epoch": 0.5642857142857143,
      "grad_norm": 2.060654401779175,
      "learning_rate": 8.11904761904762e-05,
      "loss": 0.3294,
      "step": 790
    },
    {
      "epoch": 0.5714285714285714,
      "grad_norm": 3.3581550121307373,
      "learning_rate": 8.095238095238096e-05,
      "loss": 0.3022,
      "step": 800
    },
    {
      "epoch": 0.5785714285714286,
      "grad_norm": 1.6778148412704468,
      "learning_rate": 8.071428571428573e-05,
      "loss": 0.2808,
      "step": 810
    },
    {
      "epoch": 0.5857142857142857,
      "grad_norm": 5.509958267211914,
      "learning_rate": 8.047619047619048e-05,
      "loss": 0.3584,
      "step": 820
    },
    {
      "epoch": 0.5928571428571429,
      "grad_norm": 2.3136510848999023,
      "learning_rate": 8.023809523809524e-05,
      "loss": 0.2967,
      "step": 830
    },
    {
      "epoch": 0.6,
      "grad_norm": 2.757749557495117,
      "learning_rate": 8e-05,
      "loss": 0.3091,
      "step": 840
    },
    {
      "epoch": 0.6071428571428571,
      "grad_norm": 1.6453288793563843,
      "learning_rate": 7.976190476190477e-05,
      "loss": 0.2882,
      "step": 850
    },
    {
      "epoch": 0.6142857142857143,
      "grad_norm": 3.599712371826172,
      "learning_rate": 7.952380952380952e-05,
      "loss": 0.35,
      "step": 860
    },
    {
      "epoch": 0.6214285714285714,
      "grad_norm": 1.203050136566162,
      "learning_rate": 7.928571428571429e-05,
      "loss": 0.3147,
      "step": 870
    },
    {
      "epoch": 0.6285714285714286,
      "grad_norm": 3.0582826137542725,
      "learning_rate": 7.904761904761905e-05,
      "loss": 0.369,
      "step": 880
    },
    {
      "epoch": 0.6357142857142857,
      "grad_norm": 1.7133578062057495,
      "learning_rate": 7.880952380952382e-05,
      "loss": 0.4404,
      "step": 890
    },
    {
      "epoch": 0.6428571428571429,
      "grad_norm": 2.1532819271087646,
      "learning_rate": 7.857142857142858e-05,
      "loss": 0.3249,
      "step": 900
    },
    {
      "epoch": 0.65,
      "grad_norm": 3.529282808303833,
      "learning_rate": 7.833333333333333e-05,
      "loss": 0.3449,
      "step": 910
    },
    {
      "epoch": 0.6571428571428571,
      "grad_norm": 3.412729501724243,
      "learning_rate": 7.80952380952381e-05,
      "loss": 0.3678,
      "step": 920
    },
    {
      "epoch": 0.6642857142857143,
      "grad_norm": 1.8676691055297852,
      "learning_rate": 7.785714285714286e-05,
      "loss": 0.3522,
      "step": 930
    },
    {
      "epoch": 0.6714285714285714,
      "grad_norm": 1.8748834133148193,
      "learning_rate": 7.761904761904762e-05,
      "loss": 0.2561,
      "step": 940
    },
    {
      "epoch": 0.6785714285714286,
      "grad_norm": 3.1622543334960938,
      "learning_rate": 7.738095238095239e-05,
      "loss": 0.3245,
      "step": 950
    },
    {
      "epoch": 0.6857142857142857,
      "grad_norm": 2.17108416557312,
      "learning_rate": 7.714285714285715e-05,
      "loss": 0.335,
      "step": 960
    },
    {
      "epoch": 0.6928571428571428,
      "grad_norm": 3.686673402786255,
      "learning_rate": 7.690476190476192e-05,
      "loss": 0.319,
      "step": 970
    },
    {
      "epoch": 0.7,
      "grad_norm": 7.108774662017822,
      "learning_rate": 7.666666666666667e-05,
      "loss": 0.3337,
      "step": 980
    },
    {
      "epoch": 0.7071428571428572,
      "grad_norm": 2.532336711883545,
      "learning_rate": 7.642857142857143e-05,
      "loss": 0.377,
      "step": 990
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 1.8768038749694824,
      "learning_rate": 7.619047619047618e-05,
      "loss": 0.3745,
      "step": 1000
    },
    {
      "epoch": 0.7214285714285714,
      "grad_norm": 1.8926101922988892,
      "learning_rate": 7.595238095238095e-05,
      "loss": 0.3503,
      "step": 1010
    },
    {
      "epoch": 0.7285714285714285,
      "grad_norm": 2.2763147354125977,
      "learning_rate": 7.571428571428571e-05,
      "loss": 0.307,
      "step": 1020
    },
    {
      "epoch": 0.7357142857142858,
      "grad_norm": 3.663257360458374,
      "learning_rate": 7.547619047619048e-05,
      "loss": 0.2876,
      "step": 1030
    },
    {
      "epoch": 0.7428571428571429,
      "grad_norm": 1.844648003578186,
      "learning_rate": 7.523809523809524e-05,
      "loss": 0.3512,
      "step": 1040
    },
    {
      "epoch": 0.75,
      "grad_norm": 2.5800540447235107,
      "learning_rate": 7.500000000000001e-05,
      "loss": 0.301,
      "step": 1050
    },
    {
      "epoch": 0.7571428571428571,
      "grad_norm": 2.4674232006073,
      "learning_rate": 7.476190476190477e-05,
      "loss": 0.3635,
      "step": 1060
    },
    {
      "epoch": 0.7642857142857142,
      "grad_norm": 3.5119080543518066,
      "learning_rate": 7.452380952380952e-05,
      "loss": 0.3688,
      "step": 1070
    },
    {
      "epoch": 0.7714285714285715,
      "grad_norm": 2.5902345180511475,
      "learning_rate": 7.428571428571429e-05,
      "loss": 0.3136,
      "step": 1080
    },
    {
      "epoch": 0.7785714285714286,
      "grad_norm": 1.734082818031311,
      "learning_rate": 7.404761904761905e-05,
      "loss": 0.3366,
      "step": 1090
    },
    {
      "epoch": 0.7857142857142857,
      "grad_norm": 2.0863406658172607,
      "learning_rate": 7.380952380952382e-05,
      "loss": 0.331,
      "step": 1100
    },
    {
      "epoch": 0.7928571428571428,
      "grad_norm": 2.9227652549743652,
      "learning_rate": 7.357142857142858e-05,
      "loss": 0.3131,
      "step": 1110
    },
    {
      "epoch": 0.8,
      "grad_norm": 5.268605709075928,
      "learning_rate": 7.333333333333333e-05,
      "loss": 0.3212,
      "step": 1120
    },
    {
      "epoch": 0.8071428571428572,
      "grad_norm": 2.633431911468506,
      "learning_rate": 7.30952380952381e-05,
      "loss": 0.2707,
      "step": 1130
    },
    {
      "epoch": 0.8142857142857143,
      "grad_norm": 1.652354121208191,
      "learning_rate": 7.285714285714286e-05,
      "loss": 0.3037,
      "step": 1140
    },
    {
      "epoch": 0.8214285714285714,
      "grad_norm": 3.1987786293029785,
      "learning_rate": 7.261904761904762e-05,
      "loss": 0.3603,
      "step": 1150
    },
    {
      "epoch": 0.8285714285714286,
      "grad_norm": 1.7624555826187134,
      "learning_rate": 7.238095238095238e-05,
      "loss": 0.308,
      "step": 1160
    },
    {
      "epoch": 0.8357142857142857,
      "grad_norm": 2.6379318237304688,
      "learning_rate": 7.214285714285714e-05,
      "loss": 0.3068,
      "step": 1170
    },
    {
      "epoch": 0.8428571428571429,
      "grad_norm": 2.6688101291656494,
      "learning_rate": 7.19047619047619e-05,
      "loss": 0.3134,
      "step": 1180
    },
    {
      "epoch": 0.85,
      "grad_norm": 3.698643207550049,
      "learning_rate": 7.166666666666667e-05,
      "loss": 0.2973,
      "step": 1190
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 2.9004178047180176,
      "learning_rate": 7.142857142857143e-05,
      "loss": 0.3116,
      "step": 1200
    },
    {
      "epoch": 0.8642857142857143,
      "grad_norm": 2.3414883613586426,
      "learning_rate": 7.11904761904762e-05,
      "loss": 0.334,
      "step": 1210
    },
    {
      "epoch": 0.8714285714285714,
      "grad_norm": 3.6203651428222656,
      "learning_rate": 7.095238095238096e-05,
      "loss": 0.3241,
      "step": 1220
    },
    {
      "epoch": 0.8785714285714286,
      "grad_norm": 2.7472829818725586,
      "learning_rate": 7.071428571428573e-05,
      "loss": 0.2727,
      "step": 1230
    },
    {
      "epoch": 0.8857142857142857,
      "grad_norm": 3.2332115173339844,
      "learning_rate": 7.047619047619048e-05,
      "loss": 0.3329,
      "step": 1240
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 1.968595027923584,
      "learning_rate": 7.023809523809524e-05,
      "loss": 0.3478,
      "step": 1250
    },
    {
      "epoch": 0.9,
      "grad_norm": 2.3712966442108154,
      "learning_rate": 7e-05,
      "loss": 0.3964,
      "step": 1260
    },
    {
      "epoch": 0.9071428571428571,
      "grad_norm": 1.7704048156738281,
      "learning_rate": 6.976190476190476e-05,
      "loss": 0.3983,
      "step": 1270
    },
    {
      "epoch": 0.9142857142857143,
      "grad_norm": 1.8879201412200928,
      "learning_rate": 6.952380952380952e-05,
      "loss": 0.3296,
      "step": 1280
    },
    {
      "epoch": 0.9214285714285714,
      "grad_norm": 4.0522780418396,
      "learning_rate": 6.928571428571429e-05,
      "loss": 0.3722,
      "step": 1290
    },
    {
      "epoch": 0.9285714285714286,
      "grad_norm": 1.7849159240722656,
      "learning_rate": 6.904761904761905e-05,
      "loss": 0.3022,
      "step": 1300
    },
    {
      "epoch": 0.9357142857142857,
      "grad_norm": 3.376678705215454,
      "learning_rate": 6.880952380952382e-05,
      "loss": 0.3429,
      "step": 1310
    },
    {
      "epoch": 0.9428571428571428,
      "grad_norm": 4.100617408752441,
      "learning_rate": 6.857142857142858e-05,
      "loss": 0.2793,
      "step": 1320
    },
    {
      "epoch": 0.95,
      "grad_norm": 3.8000288009643555,
      "learning_rate": 6.833333333333333e-05,
      "loss": 0.3114,
      "step": 1330
    },
    {
      "epoch": 0.9571428571428572,
      "grad_norm": 4.846517562866211,
      "learning_rate": 6.80952380952381e-05,
      "loss": 0.3377,
      "step": 1340
    },
    {
      "epoch": 0.9642857142857143,
      "grad_norm": 2.0567409992218018,
      "learning_rate": 6.785714285714286e-05,
      "loss": 0.3198,
      "step": 1350
    },
    {
      "epoch": 0.9714285714285714,
      "grad_norm": 2.3197948932647705,
      "learning_rate": 6.761904761904763e-05,
      "loss": 0.2817,
      "step": 1360
    },
    {
      "epoch": 0.9785714285714285,
      "grad_norm": 2.7312827110290527,
      "learning_rate": 6.738095238095239e-05,
      "loss": 0.3378,
      "step": 1370
    },
    {
      "epoch": 0.9857142857142858,
      "grad_norm": 3.115516185760498,
      "learning_rate": 6.714285714285714e-05,
      "loss": 0.3638,
      "step": 1380
    },
    {
      "epoch": 0.9928571428571429,
      "grad_norm": 1.0742294788360596,
      "learning_rate": 6.69047619047619e-05,
      "loss": 0.3176,
      "step": 1390
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.3872761726379395,
      "learning_rate": 6.666666666666667e-05,
      "loss": 0.2921,
      "step": 1400
    },
    {
      "epoch": 1.0,
      "eval_loss": 0.2929375469684601,
      "eval_runtime": 68.5646,
      "eval_samples_per_second": 2.917,
      "eval_steps_per_second": 0.365,
      "step": 1400
    }
  ],
  "logging_steps": 10,
  "max_steps": 4200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "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": 8377941884928000.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}