{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.8361344537815127,
  "eval_steps": 50,
  "global_step": 2700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01050420168067227,
      "grad_norm": 2.5582146644592285,
      "learning_rate": 4.98249299719888e-05,
      "loss": 1.6787,
      "step": 10
    },
    {
      "epoch": 0.02100840336134454,
      "grad_norm": 0.9345070719718933,
      "learning_rate": 4.96498599439776e-05,
      "loss": 0.518,
      "step": 20
    },
    {
      "epoch": 0.031512605042016806,
      "grad_norm": 1.6358414888381958,
      "learning_rate": 4.947478991596639e-05,
      "loss": 0.4604,
      "step": 30
    },
    {
      "epoch": 0.04201680672268908,
      "grad_norm": 0.7778844237327576,
      "learning_rate": 4.9299719887955186e-05,
      "loss": 0.3771,
      "step": 40
    },
    {
      "epoch": 0.052521008403361345,
      "grad_norm": 0.7006077766418457,
      "learning_rate": 4.912464985994398e-05,
      "loss": 0.3842,
      "step": 50
    },
    {
      "epoch": 0.052521008403361345,
      "eval_loss": 0.42603224515914917,
      "eval_runtime": 13.5673,
      "eval_samples_per_second": 35.379,
      "eval_steps_per_second": 2.211,
      "step": 50
    },
    {
      "epoch": 0.06302521008403361,
      "grad_norm": 0.6415153741836548,
      "learning_rate": 4.8949579831932775e-05,
      "loss": 0.3399,
      "step": 60
    },
    {
      "epoch": 0.07352941176470588,
      "grad_norm": 0.6030780076980591,
      "learning_rate": 4.877450980392157e-05,
      "loss": 0.3447,
      "step": 70
    },
    {
      "epoch": 0.08403361344537816,
      "grad_norm": 0.688852071762085,
      "learning_rate": 4.859943977591036e-05,
      "loss": 0.3219,
      "step": 80
    },
    {
      "epoch": 0.09453781512605042,
      "grad_norm": 0.6371557712554932,
      "learning_rate": 4.8424369747899164e-05,
      "loss": 0.3379,
      "step": 90
    },
    {
      "epoch": 0.10504201680672269,
      "grad_norm": 0.7739270329475403,
      "learning_rate": 4.824929971988796e-05,
      "loss": 0.3177,
      "step": 100
    },
    {
      "epoch": 0.10504201680672269,
      "eval_loss": 0.3801896274089813,
      "eval_runtime": 13.6107,
      "eval_samples_per_second": 35.266,
      "eval_steps_per_second": 2.204,
      "step": 100
    },
    {
      "epoch": 0.11554621848739496,
      "grad_norm": 0.649507462978363,
      "learning_rate": 4.807422969187675e-05,
      "loss": 0.3415,
      "step": 110
    },
    {
      "epoch": 0.12605042016806722,
      "grad_norm": 0.594717264175415,
      "learning_rate": 4.7899159663865554e-05,
      "loss": 0.3325,
      "step": 120
    },
    {
      "epoch": 0.13655462184873948,
      "grad_norm": 0.627918541431427,
      "learning_rate": 4.772408963585435e-05,
      "loss": 0.3222,
      "step": 130
    },
    {
      "epoch": 0.14705882352941177,
      "grad_norm": 0.5384674668312073,
      "learning_rate": 4.7549019607843135e-05,
      "loss": 0.3426,
      "step": 140
    },
    {
      "epoch": 0.15756302521008403,
      "grad_norm": 0.5673420429229736,
      "learning_rate": 4.7373949579831936e-05,
      "loss": 0.3061,
      "step": 150
    },
    {
      "epoch": 0.15756302521008403,
      "eval_loss": 0.3653399348258972,
      "eval_runtime": 13.5947,
      "eval_samples_per_second": 35.308,
      "eval_steps_per_second": 2.207,
      "step": 150
    },
    {
      "epoch": 0.16806722689075632,
      "grad_norm": 0.6111018657684326,
      "learning_rate": 4.719887955182073e-05,
      "loss": 0.3271,
      "step": 160
    },
    {
      "epoch": 0.17857142857142858,
      "grad_norm": 0.7422594428062439,
      "learning_rate": 4.7023809523809525e-05,
      "loss": 0.315,
      "step": 170
    },
    {
      "epoch": 0.18907563025210083,
      "grad_norm": 0.7226534485816956,
      "learning_rate": 4.684873949579832e-05,
      "loss": 0.3031,
      "step": 180
    },
    {
      "epoch": 0.19957983193277312,
      "grad_norm": 0.6302976012229919,
      "learning_rate": 4.667366946778712e-05,
      "loss": 0.3161,
      "step": 190
    },
    {
      "epoch": 0.21008403361344538,
      "grad_norm": 0.6225076913833618,
      "learning_rate": 4.6498599439775914e-05,
      "loss": 0.3038,
      "step": 200
    },
    {
      "epoch": 0.21008403361344538,
      "eval_loss": 0.35061606764793396,
      "eval_runtime": 13.5616,
      "eval_samples_per_second": 35.394,
      "eval_steps_per_second": 2.212,
      "step": 200
    },
    {
      "epoch": 0.22058823529411764,
      "grad_norm": 0.6001319885253906,
      "learning_rate": 4.632352941176471e-05,
      "loss": 0.3129,
      "step": 210
    },
    {
      "epoch": 0.23109243697478993,
      "grad_norm": 0.5385990142822266,
      "learning_rate": 4.61484593837535e-05,
      "loss": 0.2991,
      "step": 220
    },
    {
      "epoch": 0.2415966386554622,
      "grad_norm": 0.4513624906539917,
      "learning_rate": 4.59733893557423e-05,
      "loss": 0.2896,
      "step": 230
    },
    {
      "epoch": 0.25210084033613445,
      "grad_norm": 0.6142160892486572,
      "learning_rate": 4.579831932773109e-05,
      "loss": 0.3059,
      "step": 240
    },
    {
      "epoch": 0.26260504201680673,
      "grad_norm": 0.6714802384376526,
      "learning_rate": 4.562324929971989e-05,
      "loss": 0.2897,
      "step": 250
    },
    {
      "epoch": 0.26260504201680673,
      "eval_loss": 0.3456435203552246,
      "eval_runtime": 13.5552,
      "eval_samples_per_second": 35.411,
      "eval_steps_per_second": 2.213,
      "step": 250
    },
    {
      "epoch": 0.27310924369747897,
      "grad_norm": 0.6518235206604004,
      "learning_rate": 4.5448179271708687e-05,
      "loss": 0.312,
      "step": 260
    },
    {
      "epoch": 0.28361344537815125,
      "grad_norm": 0.6250632405281067,
      "learning_rate": 4.527310924369748e-05,
      "loss": 0.2959,
      "step": 270
    },
    {
      "epoch": 0.29411764705882354,
      "grad_norm": 0.5683826804161072,
      "learning_rate": 4.5098039215686275e-05,
      "loss": 0.3027,
      "step": 280
    },
    {
      "epoch": 0.30462184873949577,
      "grad_norm": 0.560312807559967,
      "learning_rate": 4.4922969187675076e-05,
      "loss": 0.3002,
      "step": 290
    },
    {
      "epoch": 0.31512605042016806,
      "grad_norm": 0.66291743516922,
      "learning_rate": 4.474789915966387e-05,
      "loss": 0.2925,
      "step": 300
    },
    {
      "epoch": 0.31512605042016806,
      "eval_loss": 0.3431606888771057,
      "eval_runtime": 13.5629,
      "eval_samples_per_second": 35.391,
      "eval_steps_per_second": 2.212,
      "step": 300
    },
    {
      "epoch": 0.32563025210084034,
      "grad_norm": 0.6478439569473267,
      "learning_rate": 4.4572829131652665e-05,
      "loss": 0.2893,
      "step": 310
    },
    {
      "epoch": 0.33613445378151263,
      "grad_norm": 0.5832348465919495,
      "learning_rate": 4.439775910364146e-05,
      "loss": 0.2842,
      "step": 320
    },
    {
      "epoch": 0.34663865546218486,
      "grad_norm": 0.525932252407074,
      "learning_rate": 4.422268907563025e-05,
      "loss": 0.2837,
      "step": 330
    },
    {
      "epoch": 0.35714285714285715,
      "grad_norm": 0.5487508177757263,
      "learning_rate": 4.404761904761905e-05,
      "loss": 0.2706,
      "step": 340
    },
    {
      "epoch": 0.36764705882352944,
      "grad_norm": 0.5392388701438904,
      "learning_rate": 4.387254901960784e-05,
      "loss": 0.2835,
      "step": 350
    },
    {
      "epoch": 0.36764705882352944,
      "eval_loss": 0.33528536558151245,
      "eval_runtime": 13.5508,
      "eval_samples_per_second": 35.422,
      "eval_steps_per_second": 2.214,
      "step": 350
    },
    {
      "epoch": 0.37815126050420167,
      "grad_norm": 0.6706260442733765,
      "learning_rate": 4.369747899159664e-05,
      "loss": 0.2844,
      "step": 360
    },
    {
      "epoch": 0.38865546218487396,
      "grad_norm": 0.6042625904083252,
      "learning_rate": 4.352240896358544e-05,
      "loss": 0.2758,
      "step": 370
    },
    {
      "epoch": 0.39915966386554624,
      "grad_norm": 0.534008264541626,
      "learning_rate": 4.334733893557423e-05,
      "loss": 0.2918,
      "step": 380
    },
    {
      "epoch": 0.4096638655462185,
      "grad_norm": 0.48162588477134705,
      "learning_rate": 4.317226890756303e-05,
      "loss": 0.273,
      "step": 390
    },
    {
      "epoch": 0.42016806722689076,
      "grad_norm": 0.5669644474983215,
      "learning_rate": 4.2997198879551826e-05,
      "loss": 0.285,
      "step": 400
    },
    {
      "epoch": 0.42016806722689076,
      "eval_loss": 0.3348632752895355,
      "eval_runtime": 13.5507,
      "eval_samples_per_second": 35.423,
      "eval_steps_per_second": 2.214,
      "step": 400
    },
    {
      "epoch": 0.43067226890756305,
      "grad_norm": 0.6257824897766113,
      "learning_rate": 4.2822128851540614e-05,
      "loss": 0.299,
      "step": 410
    },
    {
      "epoch": 0.4411764705882353,
      "grad_norm": 0.5430576205253601,
      "learning_rate": 4.2647058823529415e-05,
      "loss": 0.2868,
      "step": 420
    },
    {
      "epoch": 0.45168067226890757,
      "grad_norm": 0.5633955597877502,
      "learning_rate": 4.247198879551821e-05,
      "loss": 0.2589,
      "step": 430
    },
    {
      "epoch": 0.46218487394957986,
      "grad_norm": 0.5294789671897888,
      "learning_rate": 4.2296918767507e-05,
      "loss": 0.2777,
      "step": 440
    },
    {
      "epoch": 0.4726890756302521,
      "grad_norm": 0.5480856895446777,
      "learning_rate": 4.21218487394958e-05,
      "loss": 0.2704,
      "step": 450
    },
    {
      "epoch": 0.4726890756302521,
      "eval_loss": 0.329515278339386,
      "eval_runtime": 13.5423,
      "eval_samples_per_second": 35.445,
      "eval_steps_per_second": 2.215,
      "step": 450
    },
    {
      "epoch": 0.4831932773109244,
      "grad_norm": 0.5051332116127014,
      "learning_rate": 4.19467787114846e-05,
      "loss": 0.2438,
      "step": 460
    },
    {
      "epoch": 0.49369747899159666,
      "grad_norm": 0.6251511573791504,
      "learning_rate": 4.177170868347339e-05,
      "loss": 0.2748,
      "step": 470
    },
    {
      "epoch": 0.5042016806722689,
      "grad_norm": 0.4729413092136383,
      "learning_rate": 4.159663865546219e-05,
      "loss": 0.2689,
      "step": 480
    },
    {
      "epoch": 0.5147058823529411,
      "grad_norm": 0.5220003724098206,
      "learning_rate": 4.142156862745099e-05,
      "loss": 0.2899,
      "step": 490
    },
    {
      "epoch": 0.5252100840336135,
      "grad_norm": 0.54283207654953,
      "learning_rate": 4.1246498599439776e-05,
      "loss": 0.272,
      "step": 500
    },
    {
      "epoch": 0.5252100840336135,
      "eval_loss": 0.32714489102363586,
      "eval_runtime": 13.5497,
      "eval_samples_per_second": 35.425,
      "eval_steps_per_second": 2.214,
      "step": 500
    },
    {
      "epoch": 0.5357142857142857,
      "grad_norm": 0.5851682424545288,
      "learning_rate": 4.107142857142857e-05,
      "loss": 0.2691,
      "step": 510
    },
    {
      "epoch": 0.5462184873949579,
      "grad_norm": 0.6026607751846313,
      "learning_rate": 4.089635854341737e-05,
      "loss": 0.2716,
      "step": 520
    },
    {
      "epoch": 0.5567226890756303,
      "grad_norm": 0.522422730922699,
      "learning_rate": 4.0721288515406165e-05,
      "loss": 0.2774,
      "step": 530
    },
    {
      "epoch": 0.5672268907563025,
      "grad_norm": 0.516901433467865,
      "learning_rate": 4.054621848739496e-05,
      "loss": 0.2726,
      "step": 540
    },
    {
      "epoch": 0.5777310924369747,
      "grad_norm": 0.667030394077301,
      "learning_rate": 4.0371148459383754e-05,
      "loss": 0.2622,
      "step": 550
    },
    {
      "epoch": 0.5777310924369747,
      "eval_loss": 0.3262839615345001,
      "eval_runtime": 13.5448,
      "eval_samples_per_second": 35.438,
      "eval_steps_per_second": 2.215,
      "step": 550
    },
    {
      "epoch": 0.5882352941176471,
      "grad_norm": 0.542658269405365,
      "learning_rate": 4.0196078431372555e-05,
      "loss": 0.2572,
      "step": 560
    },
    {
      "epoch": 0.5987394957983193,
      "grad_norm": 0.5408573746681213,
      "learning_rate": 4.002100840336135e-05,
      "loss": 0.2636,
      "step": 570
    },
    {
      "epoch": 0.6092436974789915,
      "grad_norm": 0.5691037774085999,
      "learning_rate": 3.984593837535014e-05,
      "loss": 0.268,
      "step": 580
    },
    {
      "epoch": 0.6197478991596639,
      "grad_norm": 0.5530794858932495,
      "learning_rate": 3.967086834733894e-05,
      "loss": 0.2583,
      "step": 590
    },
    {
      "epoch": 0.6302521008403361,
      "grad_norm": 0.546229362487793,
      "learning_rate": 3.949579831932773e-05,
      "loss": 0.2622,
      "step": 600
    },
    {
      "epoch": 0.6302521008403361,
      "eval_loss": 0.3219989836215973,
      "eval_runtime": 13.5524,
      "eval_samples_per_second": 35.418,
      "eval_steps_per_second": 2.214,
      "step": 600
    },
    {
      "epoch": 0.6407563025210085,
      "grad_norm": 0.5098925232887268,
      "learning_rate": 3.9320728291316526e-05,
      "loss": 0.2553,
      "step": 610
    },
    {
      "epoch": 0.6512605042016807,
      "grad_norm": 0.5201871991157532,
      "learning_rate": 3.914565826330533e-05,
      "loss": 0.2584,
      "step": 620
    },
    {
      "epoch": 0.6617647058823529,
      "grad_norm": 0.47408100962638855,
      "learning_rate": 3.897058823529412e-05,
      "loss": 0.2686,
      "step": 630
    },
    {
      "epoch": 0.6722689075630253,
      "grad_norm": 0.5591098666191101,
      "learning_rate": 3.8795518207282915e-05,
      "loss": 0.2772,
      "step": 640
    },
    {
      "epoch": 0.6827731092436975,
      "grad_norm": 0.5344163179397583,
      "learning_rate": 3.862044817927171e-05,
      "loss": 0.263,
      "step": 650
    },
    {
      "epoch": 0.6827731092436975,
      "eval_loss": 0.31990015506744385,
      "eval_runtime": 13.5501,
      "eval_samples_per_second": 35.424,
      "eval_steps_per_second": 2.214,
      "step": 650
    },
    {
      "epoch": 0.6932773109243697,
      "grad_norm": 0.6538853049278259,
      "learning_rate": 3.844537815126051e-05,
      "loss": 0.2605,
      "step": 660
    },
    {
      "epoch": 0.7037815126050421,
      "grad_norm": 0.43679994344711304,
      "learning_rate": 3.82703081232493e-05,
      "loss": 0.2486,
      "step": 670
    },
    {
      "epoch": 0.7142857142857143,
      "grad_norm": 0.5602915287017822,
      "learning_rate": 3.809523809523809e-05,
      "loss": 0.2359,
      "step": 680
    },
    {
      "epoch": 0.7247899159663865,
      "grad_norm": 0.5354353189468384,
      "learning_rate": 3.792016806722689e-05,
      "loss": 0.2541,
      "step": 690
    },
    {
      "epoch": 0.7352941176470589,
      "grad_norm": 0.5954485535621643,
      "learning_rate": 3.774509803921569e-05,
      "loss": 0.2649,
      "step": 700
    },
    {
      "epoch": 0.7352941176470589,
      "eval_loss": 0.3204093873500824,
      "eval_runtime": 13.5523,
      "eval_samples_per_second": 35.418,
      "eval_steps_per_second": 2.214,
      "step": 700
    },
    {
      "epoch": 0.7457983193277311,
      "grad_norm": 0.6093761324882507,
      "learning_rate": 3.757002801120448e-05,
      "loss": 0.2456,
      "step": 710
    },
    {
      "epoch": 0.7563025210084033,
      "grad_norm": 0.4796586334705353,
      "learning_rate": 3.739495798319328e-05,
      "loss": 0.253,
      "step": 720
    },
    {
      "epoch": 0.7668067226890757,
      "grad_norm": 0.5846813917160034,
      "learning_rate": 3.721988795518208e-05,
      "loss": 0.2442,
      "step": 730
    },
    {
      "epoch": 0.7773109243697479,
      "grad_norm": 0.4811939597129822,
      "learning_rate": 3.704481792717087e-05,
      "loss": 0.2522,
      "step": 740
    },
    {
      "epoch": 0.7878151260504201,
      "grad_norm": 0.5718042850494385,
      "learning_rate": 3.6869747899159665e-05,
      "loss": 0.2562,
      "step": 750
    },
    {
      "epoch": 0.7878151260504201,
      "eval_loss": 0.32099905610084534,
      "eval_runtime": 13.5408,
      "eval_samples_per_second": 35.448,
      "eval_steps_per_second": 2.216,
      "step": 750
    },
    {
      "epoch": 0.7983193277310925,
      "grad_norm": 0.5630698204040527,
      "learning_rate": 3.669467787114846e-05,
      "loss": 0.2476,
      "step": 760
    },
    {
      "epoch": 0.8088235294117647,
      "grad_norm": 0.6513442397117615,
      "learning_rate": 3.6519607843137254e-05,
      "loss": 0.2544,
      "step": 770
    },
    {
      "epoch": 0.819327731092437,
      "grad_norm": 0.6139647960662842,
      "learning_rate": 3.634453781512605e-05,
      "loss": 0.258,
      "step": 780
    },
    {
      "epoch": 0.8298319327731093,
      "grad_norm": 0.5916554927825928,
      "learning_rate": 3.616946778711485e-05,
      "loss": 0.2415,
      "step": 790
    },
    {
      "epoch": 0.8403361344537815,
      "grad_norm": 0.5163634419441223,
      "learning_rate": 3.5994397759103643e-05,
      "loss": 0.252,
      "step": 800
    },
    {
      "epoch": 0.8403361344537815,
      "eval_loss": 0.3215568959712982,
      "eval_runtime": 13.5293,
      "eval_samples_per_second": 35.479,
      "eval_steps_per_second": 2.217,
      "step": 800
    },
    {
      "epoch": 0.8508403361344538,
      "grad_norm": 0.5768859386444092,
      "learning_rate": 3.581932773109244e-05,
      "loss": 0.2421,
      "step": 810
    },
    {
      "epoch": 0.8613445378151261,
      "grad_norm": 0.6197952032089233,
      "learning_rate": 3.564425770308123e-05,
      "loss": 0.2672,
      "step": 820
    },
    {
      "epoch": 0.8718487394957983,
      "grad_norm": 0.5396980047225952,
      "learning_rate": 3.546918767507003e-05,
      "loss": 0.2393,
      "step": 830
    },
    {
      "epoch": 0.8823529411764706,
      "grad_norm": 0.5783377885818481,
      "learning_rate": 3.529411764705883e-05,
      "loss": 0.2374,
      "step": 840
    },
    {
      "epoch": 0.8928571428571429,
      "grad_norm": 0.5808666944503784,
      "learning_rate": 3.511904761904762e-05,
      "loss": 0.2405,
      "step": 850
    },
    {
      "epoch": 0.8928571428571429,
      "eval_loss": 0.3207303583621979,
      "eval_runtime": 13.5299,
      "eval_samples_per_second": 35.477,
      "eval_steps_per_second": 2.217,
      "step": 850
    },
    {
      "epoch": 0.9033613445378151,
      "grad_norm": 0.4931146204471588,
      "learning_rate": 3.4943977591036416e-05,
      "loss": 0.2382,
      "step": 860
    },
    {
      "epoch": 0.9138655462184874,
      "grad_norm": 0.6456460952758789,
      "learning_rate": 3.476890756302521e-05,
      "loss": 0.2416,
      "step": 870
    },
    {
      "epoch": 0.9243697478991597,
      "grad_norm": 0.5459381937980652,
      "learning_rate": 3.4593837535014004e-05,
      "loss": 0.2611,
      "step": 880
    },
    {
      "epoch": 0.9348739495798319,
      "grad_norm": 0.5317162275314331,
      "learning_rate": 3.4418767507002805e-05,
      "loss": 0.2383,
      "step": 890
    },
    {
      "epoch": 0.9453781512605042,
      "grad_norm": 0.5790566205978394,
      "learning_rate": 3.42436974789916e-05,
      "loss": 0.2455,
      "step": 900
    },
    {
      "epoch": 0.9453781512605042,
      "eval_loss": 0.31993839144706726,
      "eval_runtime": 13.5323,
      "eval_samples_per_second": 35.471,
      "eval_steps_per_second": 2.217,
      "step": 900
    },
    {
      "epoch": 0.9558823529411765,
      "grad_norm": 0.5805277228355408,
      "learning_rate": 3.4068627450980394e-05,
      "loss": 0.2393,
      "step": 910
    },
    {
      "epoch": 0.9663865546218487,
      "grad_norm": 0.6085871458053589,
      "learning_rate": 3.389355742296919e-05,
      "loss": 0.2505,
      "step": 920
    },
    {
      "epoch": 0.976890756302521,
      "grad_norm": 0.6251375079154968,
      "learning_rate": 3.371848739495799e-05,
      "loss": 0.2378,
      "step": 930
    },
    {
      "epoch": 0.9873949579831933,
      "grad_norm": 0.6169071197509766,
      "learning_rate": 3.3543417366946776e-05,
      "loss": 0.239,
      "step": 940
    },
    {
      "epoch": 0.9978991596638656,
      "grad_norm": 0.5361204147338867,
      "learning_rate": 3.336834733893557e-05,
      "loss": 0.2484,
      "step": 950
    },
    {
      "epoch": 0.9978991596638656,
      "eval_loss": 0.31845951080322266,
      "eval_runtime": 13.5483,
      "eval_samples_per_second": 35.429,
      "eval_steps_per_second": 2.214,
      "step": 950
    },
    {
      "epoch": 1.0084033613445378,
      "grad_norm": 0.47690704464912415,
      "learning_rate": 3.319327731092437e-05,
      "loss": 0.2006,
      "step": 960
    },
    {
      "epoch": 1.01890756302521,
      "grad_norm": 0.7460448741912842,
      "learning_rate": 3.3018207282913166e-05,
      "loss": 0.1983,
      "step": 970
    },
    {
      "epoch": 1.0294117647058822,
      "grad_norm": 0.5729458332061768,
      "learning_rate": 3.284313725490196e-05,
      "loss": 0.1989,
      "step": 980
    },
    {
      "epoch": 1.0399159663865547,
      "grad_norm": 0.5524929761886597,
      "learning_rate": 3.266806722689076e-05,
      "loss": 0.2039,
      "step": 990
    },
    {
      "epoch": 1.050420168067227,
      "grad_norm": 0.6426274180412292,
      "learning_rate": 3.2492997198879555e-05,
      "loss": 0.2184,
      "step": 1000
    },
    {
      "epoch": 1.050420168067227,
      "eval_loss": 0.3305407762527466,
      "eval_runtime": 13.5592,
      "eval_samples_per_second": 35.4,
      "eval_steps_per_second": 2.213,
      "step": 1000
    },
    {
      "epoch": 1.0609243697478992,
      "grad_norm": 0.4944634437561035,
      "learning_rate": 3.231792717086835e-05,
      "loss": 0.2042,
      "step": 1010
    },
    {
      "epoch": 1.0714285714285714,
      "grad_norm": 0.5576530694961548,
      "learning_rate": 3.2142857142857144e-05,
      "loss": 0.2029,
      "step": 1020
    },
    {
      "epoch": 1.0819327731092436,
      "grad_norm": 0.674849271774292,
      "learning_rate": 3.196778711484594e-05,
      "loss": 0.2026,
      "step": 1030
    },
    {
      "epoch": 1.092436974789916,
      "grad_norm": 0.5424471497535706,
      "learning_rate": 3.179271708683473e-05,
      "loss": 0.2027,
      "step": 1040
    },
    {
      "epoch": 1.1029411764705883,
      "grad_norm": 0.6491550207138062,
      "learning_rate": 3.161764705882353e-05,
      "loss": 0.2029,
      "step": 1050
    },
    {
      "epoch": 1.1029411764705883,
      "eval_loss": 0.33292290568351746,
      "eval_runtime": 13.5611,
      "eval_samples_per_second": 35.395,
      "eval_steps_per_second": 2.212,
      "step": 1050
    },
    {
      "epoch": 1.1134453781512605,
      "grad_norm": 0.6113711595535278,
      "learning_rate": 3.144257703081233e-05,
      "loss": 0.2,
      "step": 1060
    },
    {
      "epoch": 1.1239495798319328,
      "grad_norm": 0.5068053603172302,
      "learning_rate": 3.126750700280112e-05,
      "loss": 0.1903,
      "step": 1070
    },
    {
      "epoch": 1.134453781512605,
      "grad_norm": 0.6518192291259766,
      "learning_rate": 3.1092436974789916e-05,
      "loss": 0.2006,
      "step": 1080
    },
    {
      "epoch": 1.1449579831932772,
      "grad_norm": 0.6932762861251831,
      "learning_rate": 3.091736694677872e-05,
      "loss": 0.2051,
      "step": 1090
    },
    {
      "epoch": 1.1554621848739495,
      "grad_norm": 0.5372537970542908,
      "learning_rate": 3.074229691876751e-05,
      "loss": 0.2037,
      "step": 1100
    },
    {
      "epoch": 1.1554621848739495,
      "eval_loss": 0.33518460392951965,
      "eval_runtime": 13.5628,
      "eval_samples_per_second": 35.391,
      "eval_steps_per_second": 2.212,
      "step": 1100
    },
    {
      "epoch": 1.165966386554622,
      "grad_norm": 0.6258675456047058,
      "learning_rate": 3.0567226890756306e-05,
      "loss": 0.1953,
      "step": 1110
    },
    {
      "epoch": 1.1764705882352942,
      "grad_norm": 0.5910756587982178,
      "learning_rate": 3.0392156862745097e-05,
      "loss": 0.2098,
      "step": 1120
    },
    {
      "epoch": 1.1869747899159664,
      "grad_norm": 0.6931313276290894,
      "learning_rate": 3.0217086834733894e-05,
      "loss": 0.2022,
      "step": 1130
    },
    {
      "epoch": 1.1974789915966386,
      "grad_norm": 0.6102430820465088,
      "learning_rate": 3.004201680672269e-05,
      "loss": 0.1946,
      "step": 1140
    },
    {
      "epoch": 1.2079831932773109,
      "grad_norm": 0.6068236827850342,
      "learning_rate": 2.9866946778711486e-05,
      "loss": 0.2089,
      "step": 1150
    },
    {
      "epoch": 1.2079831932773109,
      "eval_loss": 0.3338312804698944,
      "eval_runtime": 13.5402,
      "eval_samples_per_second": 35.45,
      "eval_steps_per_second": 2.216,
      "step": 1150
    },
    {
      "epoch": 1.2184873949579833,
      "grad_norm": 0.6477882862091064,
      "learning_rate": 2.969187675070028e-05,
      "loss": 0.2065,
      "step": 1160
    },
    {
      "epoch": 1.2289915966386555,
      "grad_norm": 0.6269820928573608,
      "learning_rate": 2.9516806722689078e-05,
      "loss": 0.1926,
      "step": 1170
    },
    {
      "epoch": 1.2394957983193278,
      "grad_norm": 0.6708040833473206,
      "learning_rate": 2.9341736694677872e-05,
      "loss": 0.1983,
      "step": 1180
    },
    {
      "epoch": 1.25,
      "grad_norm": 0.6625474095344543,
      "learning_rate": 2.916666666666667e-05,
      "loss": 0.1967,
      "step": 1190
    },
    {
      "epoch": 1.2605042016806722,
      "grad_norm": 0.5640371441841125,
      "learning_rate": 2.8991596638655467e-05,
      "loss": 0.205,
      "step": 1200
    },
    {
      "epoch": 1.2605042016806722,
      "eval_loss": 0.33764752745628357,
      "eval_runtime": 13.5241,
      "eval_samples_per_second": 35.492,
      "eval_steps_per_second": 2.218,
      "step": 1200
    },
    {
      "epoch": 1.2710084033613445,
      "grad_norm": 0.590160608291626,
      "learning_rate": 2.8816526610644258e-05,
      "loss": 0.2008,
      "step": 1210
    },
    {
      "epoch": 1.2815126050420167,
      "grad_norm": 0.6975618004798889,
      "learning_rate": 2.8641456582633052e-05,
      "loss": 0.1956,
      "step": 1220
    },
    {
      "epoch": 1.2920168067226891,
      "grad_norm": 0.6742042899131775,
      "learning_rate": 2.846638655462185e-05,
      "loss": 0.2089,
      "step": 1230
    },
    {
      "epoch": 1.3025210084033614,
      "grad_norm": 0.5806481838226318,
      "learning_rate": 2.8291316526610644e-05,
      "loss": 0.1937,
      "step": 1240
    },
    {
      "epoch": 1.3130252100840336,
      "grad_norm": 0.9894171953201294,
      "learning_rate": 2.8116246498599442e-05,
      "loss": 0.2098,
      "step": 1250
    },
    {
      "epoch": 1.3130252100840336,
      "eval_loss": 0.3376815617084503,
      "eval_runtime": 13.503,
      "eval_samples_per_second": 35.548,
      "eval_steps_per_second": 2.222,
      "step": 1250
    },
    {
      "epoch": 1.3235294117647058,
      "grad_norm": 0.8022045493125916,
      "learning_rate": 2.7941176470588236e-05,
      "loss": 0.2043,
      "step": 1260
    },
    {
      "epoch": 1.334033613445378,
      "grad_norm": 0.687003493309021,
      "learning_rate": 2.7766106442577034e-05,
      "loss": 0.2023,
      "step": 1270
    },
    {
      "epoch": 1.3445378151260505,
      "grad_norm": 0.7270589470863342,
      "learning_rate": 2.7591036414565828e-05,
      "loss": 0.1994,
      "step": 1280
    },
    {
      "epoch": 1.3550420168067228,
      "grad_norm": 0.5873326063156128,
      "learning_rate": 2.7415966386554626e-05,
      "loss": 0.1904,
      "step": 1290
    },
    {
      "epoch": 1.365546218487395,
      "grad_norm": 0.6190339922904968,
      "learning_rate": 2.7240896358543417e-05,
      "loss": 0.198,
      "step": 1300
    },
    {
      "epoch": 1.365546218487395,
      "eval_loss": 0.3378330171108246,
      "eval_runtime": 13.4926,
      "eval_samples_per_second": 35.575,
      "eval_steps_per_second": 2.223,
      "step": 1300
    },
    {
      "epoch": 1.3760504201680672,
      "grad_norm": 0.6693372130393982,
      "learning_rate": 2.706582633053221e-05,
      "loss": 0.1953,
      "step": 1310
    },
    {
      "epoch": 1.3865546218487395,
      "grad_norm": 0.5310758352279663,
      "learning_rate": 2.689075630252101e-05,
      "loss": 0.2013,
      "step": 1320
    },
    {
      "epoch": 1.3970588235294117,
      "grad_norm": 0.7299119830131531,
      "learning_rate": 2.6715686274509806e-05,
      "loss": 0.1967,
      "step": 1330
    },
    {
      "epoch": 1.407563025210084,
      "grad_norm": 0.7364310026168823,
      "learning_rate": 2.65406162464986e-05,
      "loss": 0.1996,
      "step": 1340
    },
    {
      "epoch": 1.4180672268907564,
      "grad_norm": 0.6624464988708496,
      "learning_rate": 2.6365546218487398e-05,
      "loss": 0.1889,
      "step": 1350
    },
    {
      "epoch": 1.4180672268907564,
      "eval_loss": 0.33763712644577026,
      "eval_runtime": 13.5125,
      "eval_samples_per_second": 35.523,
      "eval_steps_per_second": 2.22,
      "step": 1350
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.6339524984359741,
      "learning_rate": 2.6190476190476192e-05,
      "loss": 0.1898,
      "step": 1360
    },
    {
      "epoch": 1.4390756302521008,
      "grad_norm": 0.7600531578063965,
      "learning_rate": 2.601540616246499e-05,
      "loss": 0.1976,
      "step": 1370
    },
    {
      "epoch": 1.449579831932773,
      "grad_norm": 0.7129259705543518,
      "learning_rate": 2.5840336134453784e-05,
      "loss": 0.1978,
      "step": 1380
    },
    {
      "epoch": 1.4600840336134453,
      "grad_norm": 0.6824067234992981,
      "learning_rate": 2.5665266106442575e-05,
      "loss": 0.1917,
      "step": 1390
    },
    {
      "epoch": 1.4705882352941178,
      "grad_norm": 0.8596220016479492,
      "learning_rate": 2.5490196078431373e-05,
      "loss": 0.1821,
      "step": 1400
    },
    {
      "epoch": 1.4705882352941178,
      "eval_loss": 0.3433762192726135,
      "eval_runtime": 13.5224,
      "eval_samples_per_second": 35.497,
      "eval_steps_per_second": 2.219,
      "step": 1400
    },
    {
      "epoch": 1.48109243697479,
      "grad_norm": 0.5555725693702698,
      "learning_rate": 2.5315126050420167e-05,
      "loss": 0.1952,
      "step": 1410
    },
    {
      "epoch": 1.4915966386554622,
      "grad_norm": 0.7239671349525452,
      "learning_rate": 2.5140056022408964e-05,
      "loss": 0.1989,
      "step": 1420
    },
    {
      "epoch": 1.5021008403361344,
      "grad_norm": 0.70365971326828,
      "learning_rate": 2.4964985994397762e-05,
      "loss": 0.1944,
      "step": 1430
    },
    {
      "epoch": 1.5126050420168067,
      "grad_norm": 0.8063983917236328,
      "learning_rate": 2.4789915966386556e-05,
      "loss": 0.1964,
      "step": 1440
    },
    {
      "epoch": 1.523109243697479,
      "grad_norm": 0.7536810636520386,
      "learning_rate": 2.4614845938375354e-05,
      "loss": 0.2004,
      "step": 1450
    },
    {
      "epoch": 1.523109243697479,
      "eval_loss": 0.3417983055114746,
      "eval_runtime": 13.5372,
      "eval_samples_per_second": 35.458,
      "eval_steps_per_second": 2.216,
      "step": 1450
    },
    {
      "epoch": 1.5336134453781511,
      "grad_norm": 0.608958899974823,
      "learning_rate": 2.4439775910364145e-05,
      "loss": 0.1907,
      "step": 1460
    },
    {
      "epoch": 1.5441176470588234,
      "grad_norm": 0.6138647794723511,
      "learning_rate": 2.4264705882352942e-05,
      "loss": 0.193,
      "step": 1470
    },
    {
      "epoch": 1.5546218487394958,
      "grad_norm": 0.8067657947540283,
      "learning_rate": 2.4089635854341737e-05,
      "loss": 0.188,
      "step": 1480
    },
    {
      "epoch": 1.565126050420168,
      "grad_norm": 0.7985292077064514,
      "learning_rate": 2.3914565826330534e-05,
      "loss": 0.197,
      "step": 1490
    },
    {
      "epoch": 1.5756302521008403,
      "grad_norm": 0.697371244430542,
      "learning_rate": 2.373949579831933e-05,
      "loss": 0.203,
      "step": 1500
    },
    {
      "epoch": 1.5756302521008403,
      "eval_loss": 0.34146031737327576,
      "eval_runtime": 13.5585,
      "eval_samples_per_second": 35.402,
      "eval_steps_per_second": 2.213,
      "step": 1500
    },
    {
      "epoch": 1.5861344537815127,
      "grad_norm": 0.6977027058601379,
      "learning_rate": 2.3564425770308123e-05,
      "loss": 0.197,
      "step": 1510
    },
    {
      "epoch": 1.596638655462185,
      "grad_norm": 0.80245441198349,
      "learning_rate": 2.338935574229692e-05,
      "loss": 0.1876,
      "step": 1520
    },
    {
      "epoch": 1.6071428571428572,
      "grad_norm": 0.7472719550132751,
      "learning_rate": 2.3214285714285715e-05,
      "loss": 0.1936,
      "step": 1530
    },
    {
      "epoch": 1.6176470588235294,
      "grad_norm": 0.7296733260154724,
      "learning_rate": 2.303921568627451e-05,
      "loss": 0.1762,
      "step": 1540
    },
    {
      "epoch": 1.6281512605042017,
      "grad_norm": 0.7784711718559265,
      "learning_rate": 2.2864145658263307e-05,
      "loss": 0.1828,
      "step": 1550
    },
    {
      "epoch": 1.6281512605042017,
      "eval_loss": 0.34527555108070374,
      "eval_runtime": 13.5577,
      "eval_samples_per_second": 35.404,
      "eval_steps_per_second": 2.213,
      "step": 1550
    },
    {
      "epoch": 1.638655462184874,
      "grad_norm": 0.8349794149398804,
      "learning_rate": 2.26890756302521e-05,
      "loss": 0.1904,
      "step": 1560
    },
    {
      "epoch": 1.6491596638655461,
      "grad_norm": 0.7088022828102112,
      "learning_rate": 2.25140056022409e-05,
      "loss": 0.1863,
      "step": 1570
    },
    {
      "epoch": 1.6596638655462184,
      "grad_norm": 0.6738799810409546,
      "learning_rate": 2.2338935574229693e-05,
      "loss": 0.1847,
      "step": 1580
    },
    {
      "epoch": 1.6701680672268906,
      "grad_norm": 0.7299010157585144,
      "learning_rate": 2.2163865546218487e-05,
      "loss": 0.1892,
      "step": 1590
    },
    {
      "epoch": 1.680672268907563,
      "grad_norm": 0.9265356063842773,
      "learning_rate": 2.1988795518207285e-05,
      "loss": 0.1916,
      "step": 1600
    },
    {
      "epoch": 1.680672268907563,
      "eval_loss": 0.34289926290512085,
      "eval_runtime": 13.5614,
      "eval_samples_per_second": 35.394,
      "eval_steps_per_second": 2.212,
      "step": 1600
    },
    {
      "epoch": 1.6911764705882353,
      "grad_norm": 0.6903817653656006,
      "learning_rate": 2.181372549019608e-05,
      "loss": 0.1849,
      "step": 1610
    },
    {
      "epoch": 1.7016806722689075,
      "grad_norm": 0.7027392983436584,
      "learning_rate": 2.1638655462184876e-05,
      "loss": 0.1782,
      "step": 1620
    },
    {
      "epoch": 1.71218487394958,
      "grad_norm": 0.6882468461990356,
      "learning_rate": 2.146358543417367e-05,
      "loss": 0.1849,
      "step": 1630
    },
    {
      "epoch": 1.7226890756302522,
      "grad_norm": 0.7120645642280579,
      "learning_rate": 2.1288515406162465e-05,
      "loss": 0.1916,
      "step": 1640
    },
    {
      "epoch": 1.7331932773109244,
      "grad_norm": 0.7592134475708008,
      "learning_rate": 2.1113445378151263e-05,
      "loss": 0.1748,
      "step": 1650
    },
    {
      "epoch": 1.7331932773109244,
      "eval_loss": 0.34829968214035034,
      "eval_runtime": 13.5535,
      "eval_samples_per_second": 35.415,
      "eval_steps_per_second": 2.213,
      "step": 1650
    },
    {
      "epoch": 1.7436974789915967,
      "grad_norm": 0.6609721183776855,
      "learning_rate": 2.0938375350140057e-05,
      "loss": 0.1877,
      "step": 1660
    },
    {
      "epoch": 1.754201680672269,
      "grad_norm": 0.8183338046073914,
      "learning_rate": 2.0763305322128854e-05,
      "loss": 0.1795,
      "step": 1670
    },
    {
      "epoch": 1.7647058823529411,
      "grad_norm": 0.6892154812812805,
      "learning_rate": 2.058823529411765e-05,
      "loss": 0.1812,
      "step": 1680
    },
    {
      "epoch": 1.7752100840336134,
      "grad_norm": 0.618888258934021,
      "learning_rate": 2.0413165266106443e-05,
      "loss": 0.1852,
      "step": 1690
    },
    {
      "epoch": 1.7857142857142856,
      "grad_norm": 0.6212893128395081,
      "learning_rate": 2.023809523809524e-05,
      "loss": 0.1935,
      "step": 1700
    },
    {
      "epoch": 1.7857142857142856,
      "eval_loss": 0.34380197525024414,
      "eval_runtime": 13.526,
      "eval_samples_per_second": 35.487,
      "eval_steps_per_second": 2.218,
      "step": 1700
    },
    {
      "epoch": 1.7962184873949578,
      "grad_norm": 0.6889869570732117,
      "learning_rate": 2.0063025210084035e-05,
      "loss": 0.182,
      "step": 1710
    },
    {
      "epoch": 1.8067226890756303,
      "grad_norm": 0.6283136606216431,
      "learning_rate": 1.988795518207283e-05,
      "loss": 0.1837,
      "step": 1720
    },
    {
      "epoch": 1.8172268907563025,
      "grad_norm": 0.6092699766159058,
      "learning_rate": 1.9712885154061627e-05,
      "loss": 0.178,
      "step": 1730
    },
    {
      "epoch": 1.8277310924369747,
      "grad_norm": 0.665622889995575,
      "learning_rate": 1.953781512605042e-05,
      "loss": 0.1785,
      "step": 1740
    },
    {
      "epoch": 1.8382352941176472,
      "grad_norm": 0.8595823049545288,
      "learning_rate": 1.936274509803922e-05,
      "loss": 0.184,
      "step": 1750
    },
    {
      "epoch": 1.8382352941176472,
      "eval_loss": 0.35087138414382935,
      "eval_runtime": 13.5028,
      "eval_samples_per_second": 35.548,
      "eval_steps_per_second": 2.222,
      "step": 1750
    },
    {
      "epoch": 1.8487394957983194,
      "grad_norm": 0.8481118679046631,
      "learning_rate": 1.9187675070028013e-05,
      "loss": 0.1841,
      "step": 1760
    },
    {
      "epoch": 1.8592436974789917,
      "grad_norm": 0.6994073987007141,
      "learning_rate": 1.9012605042016807e-05,
      "loss": 0.1765,
      "step": 1770
    },
    {
      "epoch": 1.8697478991596639,
      "grad_norm": 0.6001220941543579,
      "learning_rate": 1.88375350140056e-05,
      "loss": 0.1807,
      "step": 1780
    },
    {
      "epoch": 1.8802521008403361,
      "grad_norm": 0.7995849847793579,
      "learning_rate": 1.86624649859944e-05,
      "loss": 0.1731,
      "step": 1790
    },
    {
      "epoch": 1.8907563025210083,
      "grad_norm": 0.6178213953971863,
      "learning_rate": 1.8487394957983196e-05,
      "loss": 0.1723,
      "step": 1800
    },
    {
      "epoch": 1.8907563025210083,
      "eval_loss": 0.35629966855049133,
      "eval_runtime": 13.4945,
      "eval_samples_per_second": 35.57,
      "eval_steps_per_second": 2.223,
      "step": 1800
    },
    {
      "epoch": 1.9012605042016806,
      "grad_norm": 0.6287909746170044,
      "learning_rate": 1.8312324929971987e-05,
      "loss": 0.1756,
      "step": 1810
    },
    {
      "epoch": 1.9117647058823528,
      "grad_norm": 0.7970178723335266,
      "learning_rate": 1.8137254901960785e-05,
      "loss": 0.1815,
      "step": 1820
    },
    {
      "epoch": 1.9222689075630253,
      "grad_norm": 0.7358176708221436,
      "learning_rate": 1.796218487394958e-05,
      "loss": 0.1816,
      "step": 1830
    },
    {
      "epoch": 1.9327731092436975,
      "grad_norm": 0.7053601741790771,
      "learning_rate": 1.7787114845938377e-05,
      "loss": 0.1711,
      "step": 1840
    },
    {
      "epoch": 1.9432773109243697,
      "grad_norm": 0.7120839953422546,
      "learning_rate": 1.7612044817927174e-05,
      "loss": 0.1737,
      "step": 1850
    },
    {
      "epoch": 1.9432773109243697,
      "eval_loss": 0.35368964076042175,
      "eval_runtime": 13.499,
      "eval_samples_per_second": 35.558,
      "eval_steps_per_second": 2.222,
      "step": 1850
    },
    {
      "epoch": 1.9537815126050422,
      "grad_norm": 0.6667950749397278,
      "learning_rate": 1.7436974789915965e-05,
      "loss": 0.1729,
      "step": 1860
    },
    {
      "epoch": 1.9642857142857144,
      "grad_norm": 0.6258378028869629,
      "learning_rate": 1.7261904761904763e-05,
      "loss": 0.1811,
      "step": 1870
    },
    {
      "epoch": 1.9747899159663866,
      "grad_norm": 0.7531097531318665,
      "learning_rate": 1.7086834733893557e-05,
      "loss": 0.1742,
      "step": 1880
    },
    {
      "epoch": 1.9852941176470589,
      "grad_norm": 0.7142496109008789,
      "learning_rate": 1.6911764705882355e-05,
      "loss": 0.1738,
      "step": 1890
    },
    {
      "epoch": 1.995798319327731,
      "grad_norm": 0.7428011298179626,
      "learning_rate": 1.673669467787115e-05,
      "loss": 0.1694,
      "step": 1900
    },
    {
      "epoch": 1.995798319327731,
      "eval_loss": 0.3620971739292145,
      "eval_runtime": 13.5186,
      "eval_samples_per_second": 35.507,
      "eval_steps_per_second": 2.219,
      "step": 1900
    },
    {
      "epoch": 2.0063025210084033,
      "grad_norm": 0.5604604482650757,
      "learning_rate": 1.6561624649859943e-05,
      "loss": 0.1524,
      "step": 1910
    },
    {
      "epoch": 2.0168067226890756,
      "grad_norm": 0.8657930493354797,
      "learning_rate": 1.638655462184874e-05,
      "loss": 0.1496,
      "step": 1920
    },
    {
      "epoch": 2.027310924369748,
      "grad_norm": 0.642977237701416,
      "learning_rate": 1.6211484593837535e-05,
      "loss": 0.1452,
      "step": 1930
    },
    {
      "epoch": 2.03781512605042,
      "grad_norm": 0.7766443490982056,
      "learning_rate": 1.6036414565826333e-05,
      "loss": 0.1474,
      "step": 1940
    },
    {
      "epoch": 2.0483193277310923,
      "grad_norm": 0.6917587518692017,
      "learning_rate": 1.5861344537815127e-05,
      "loss": 0.149,
      "step": 1950
    },
    {
      "epoch": 2.0483193277310923,
      "eval_loss": 0.3917512893676758,
      "eval_runtime": 13.5475,
      "eval_samples_per_second": 35.431,
      "eval_steps_per_second": 2.214,
      "step": 1950
    },
    {
      "epoch": 2.0588235294117645,
      "grad_norm": 0.7516221404075623,
      "learning_rate": 1.568627450980392e-05,
      "loss": 0.1507,
      "step": 1960
    },
    {
      "epoch": 2.069327731092437,
      "grad_norm": 0.8124748468399048,
      "learning_rate": 1.551120448179272e-05,
      "loss": 0.1478,
      "step": 1970
    },
    {
      "epoch": 2.0798319327731094,
      "grad_norm": 0.664069652557373,
      "learning_rate": 1.5336134453781513e-05,
      "loss": 0.1505,
      "step": 1980
    },
    {
      "epoch": 2.0903361344537816,
      "grad_norm": 0.6899878978729248,
      "learning_rate": 1.5161064425770307e-05,
      "loss": 0.1472,
      "step": 1990
    },
    {
      "epoch": 2.100840336134454,
      "grad_norm": 0.680801510810852,
      "learning_rate": 1.4985994397759103e-05,
      "loss": 0.1527,
      "step": 2000
    },
    {
      "epoch": 2.100840336134454,
      "eval_loss": 0.3939915895462036,
      "eval_runtime": 13.5573,
      "eval_samples_per_second": 35.405,
      "eval_steps_per_second": 2.213,
      "step": 2000
    },
    {
      "epoch": 2.111344537815126,
      "grad_norm": 0.5556862950325012,
      "learning_rate": 1.48109243697479e-05,
      "loss": 0.1476,
      "step": 2010
    },
    {
      "epoch": 2.1218487394957983,
      "grad_norm": 0.6164736151695251,
      "learning_rate": 1.4635854341736697e-05,
      "loss": 0.1471,
      "step": 2020
    },
    {
      "epoch": 2.1323529411764706,
      "grad_norm": 0.7886316180229187,
      "learning_rate": 1.4460784313725493e-05,
      "loss": 0.1447,
      "step": 2030
    },
    {
      "epoch": 2.142857142857143,
      "grad_norm": 0.780383825302124,
      "learning_rate": 1.4285714285714285e-05,
      "loss": 0.1456,
      "step": 2040
    },
    {
      "epoch": 2.153361344537815,
      "grad_norm": 0.8698550462722778,
      "learning_rate": 1.4110644257703081e-05,
      "loss": 0.1529,
      "step": 2050
    },
    {
      "epoch": 2.153361344537815,
      "eval_loss": 0.4003276228904724,
      "eval_runtime": 13.5575,
      "eval_samples_per_second": 35.405,
      "eval_steps_per_second": 2.213,
      "step": 2050
    },
    {
      "epoch": 2.1638655462184873,
      "grad_norm": 0.709811806678772,
      "learning_rate": 1.3935574229691877e-05,
      "loss": 0.1434,
      "step": 2060
    },
    {
      "epoch": 2.1743697478991595,
      "grad_norm": 0.6641806364059448,
      "learning_rate": 1.3760504201680673e-05,
      "loss": 0.1479,
      "step": 2070
    },
    {
      "epoch": 2.184873949579832,
      "grad_norm": 0.6883408427238464,
      "learning_rate": 1.3585434173669467e-05,
      "loss": 0.152,
      "step": 2080
    },
    {
      "epoch": 2.1953781512605044,
      "grad_norm": 0.754300057888031,
      "learning_rate": 1.3410364145658263e-05,
      "loss": 0.1451,
      "step": 2090
    },
    {
      "epoch": 2.2058823529411766,
      "grad_norm": 0.7174961566925049,
      "learning_rate": 1.323529411764706e-05,
      "loss": 0.1469,
      "step": 2100
    },
    {
      "epoch": 2.2058823529411766,
      "eval_loss": 0.40417909622192383,
      "eval_runtime": 13.5536,
      "eval_samples_per_second": 35.415,
      "eval_steps_per_second": 2.213,
      "step": 2100
    },
    {
      "epoch": 2.216386554621849,
      "grad_norm": 0.8382938504219055,
      "learning_rate": 1.3060224089635855e-05,
      "loss": 0.145,
      "step": 2110
    },
    {
      "epoch": 2.226890756302521,
      "grad_norm": 0.6055848002433777,
      "learning_rate": 1.288515406162465e-05,
      "loss": 0.1415,
      "step": 2120
    },
    {
      "epoch": 2.2373949579831933,
      "grad_norm": 0.6858454942703247,
      "learning_rate": 1.2710084033613445e-05,
      "loss": 0.1477,
      "step": 2130
    },
    {
      "epoch": 2.2478991596638656,
      "grad_norm": 0.825764000415802,
      "learning_rate": 1.2535014005602241e-05,
      "loss": 0.1454,
      "step": 2140
    },
    {
      "epoch": 2.258403361344538,
      "grad_norm": 0.7165321707725525,
      "learning_rate": 1.2359943977591037e-05,
      "loss": 0.1583,
      "step": 2150
    },
    {
      "epoch": 2.258403361344538,
      "eval_loss": 0.4027842581272125,
      "eval_runtime": 13.5575,
      "eval_samples_per_second": 35.405,
      "eval_steps_per_second": 2.213,
      "step": 2150
    },
    {
      "epoch": 2.26890756302521,
      "grad_norm": 0.8113967776298523,
      "learning_rate": 1.2184873949579832e-05,
      "loss": 0.146,
      "step": 2160
    },
    {
      "epoch": 2.2794117647058822,
      "grad_norm": 0.5904905796051025,
      "learning_rate": 1.200980392156863e-05,
      "loss": 0.1445,
      "step": 2170
    },
    {
      "epoch": 2.2899159663865545,
      "grad_norm": 0.6510922908782959,
      "learning_rate": 1.1834733893557423e-05,
      "loss": 0.1417,
      "step": 2180
    },
    {
      "epoch": 2.3004201680672267,
      "grad_norm": 1.0300558805465698,
      "learning_rate": 1.165966386554622e-05,
      "loss": 0.1445,
      "step": 2190
    },
    {
      "epoch": 2.310924369747899,
      "grad_norm": 0.8051159977912903,
      "learning_rate": 1.1484593837535014e-05,
      "loss": 0.1441,
      "step": 2200
    },
    {
      "epoch": 2.310924369747899,
      "eval_loss": 0.4135133922100067,
      "eval_runtime": 13.5402,
      "eval_samples_per_second": 35.45,
      "eval_steps_per_second": 2.216,
      "step": 2200
    },
    {
      "epoch": 2.3214285714285716,
      "grad_norm": 0.7365185022354126,
      "learning_rate": 1.130952380952381e-05,
      "loss": 0.1445,
      "step": 2210
    },
    {
      "epoch": 2.331932773109244,
      "grad_norm": 0.6721594929695129,
      "learning_rate": 1.1134453781512606e-05,
      "loss": 0.1417,
      "step": 2220
    },
    {
      "epoch": 2.342436974789916,
      "grad_norm": 0.9220572113990784,
      "learning_rate": 1.0959383753501401e-05,
      "loss": 0.1455,
      "step": 2230
    },
    {
      "epoch": 2.3529411764705883,
      "grad_norm": 0.7778609395027161,
      "learning_rate": 1.0784313725490197e-05,
      "loss": 0.1451,
      "step": 2240
    },
    {
      "epoch": 2.3634453781512605,
      "grad_norm": 1.1159327030181885,
      "learning_rate": 1.0609243697478992e-05,
      "loss": 0.1458,
      "step": 2250
    },
    {
      "epoch": 2.3634453781512605,
      "eval_loss": 0.40705686807632446,
      "eval_runtime": 13.5021,
      "eval_samples_per_second": 35.55,
      "eval_steps_per_second": 2.222,
      "step": 2250
    },
    {
      "epoch": 2.3739495798319328,
      "grad_norm": 0.6801664233207703,
      "learning_rate": 1.0434173669467788e-05,
      "loss": 0.1428,
      "step": 2260
    },
    {
      "epoch": 2.384453781512605,
      "grad_norm": 0.8475140333175659,
      "learning_rate": 1.0259103641456584e-05,
      "loss": 0.1401,
      "step": 2270
    },
    {
      "epoch": 2.3949579831932772,
      "grad_norm": 1.0753763914108276,
      "learning_rate": 1.008403361344538e-05,
      "loss": 0.1485,
      "step": 2280
    },
    {
      "epoch": 2.4054621848739495,
      "grad_norm": 0.7163957953453064,
      "learning_rate": 9.908963585434174e-06,
      "loss": 0.148,
      "step": 2290
    },
    {
      "epoch": 2.4159663865546217,
      "grad_norm": 0.73530113697052,
      "learning_rate": 9.73389355742297e-06,
      "loss": 0.1462,
      "step": 2300
    },
    {
      "epoch": 2.4159663865546217,
      "eval_loss": 0.4083961546421051,
      "eval_runtime": 13.5005,
      "eval_samples_per_second": 35.554,
      "eval_steps_per_second": 2.222,
      "step": 2300
    },
    {
      "epoch": 2.426470588235294,
      "grad_norm": 0.8182764649391174,
      "learning_rate": 9.558823529411764e-06,
      "loss": 0.145,
      "step": 2310
    },
    {
      "epoch": 2.4369747899159666,
      "grad_norm": 0.8590428829193115,
      "learning_rate": 9.38375350140056e-06,
      "loss": 0.146,
      "step": 2320
    },
    {
      "epoch": 2.447478991596639,
      "grad_norm": 0.83616042137146,
      "learning_rate": 9.208683473389357e-06,
      "loss": 0.1455,
      "step": 2330
    },
    {
      "epoch": 2.457983193277311,
      "grad_norm": 0.7634994387626648,
      "learning_rate": 9.033613445378152e-06,
      "loss": 0.1363,
      "step": 2340
    },
    {
      "epoch": 2.4684873949579833,
      "grad_norm": 0.7826119661331177,
      "learning_rate": 8.858543417366948e-06,
      "loss": 0.1447,
      "step": 2350
    },
    {
      "epoch": 2.4684873949579833,
      "eval_loss": 0.4166198670864105,
      "eval_runtime": 13.5099,
      "eval_samples_per_second": 35.529,
      "eval_steps_per_second": 2.221,
      "step": 2350
    },
    {
      "epoch": 2.4789915966386555,
      "grad_norm": 0.7558380961418152,
      "learning_rate": 8.683473389355742e-06,
      "loss": 0.1449,
      "step": 2360
    },
    {
      "epoch": 2.4894957983193278,
      "grad_norm": 0.6801384091377258,
      "learning_rate": 8.508403361344538e-06,
      "loss": 0.1382,
      "step": 2370
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.9655725359916687,
      "learning_rate": 8.333333333333334e-06,
      "loss": 0.1425,
      "step": 2380
    },
    {
      "epoch": 2.5105042016806722,
      "grad_norm": 0.9613827466964722,
      "learning_rate": 8.15826330532213e-06,
      "loss": 0.1374,
      "step": 2390
    },
    {
      "epoch": 2.5210084033613445,
      "grad_norm": 0.8105395436286926,
      "learning_rate": 7.983193277310924e-06,
      "loss": 0.1431,
      "step": 2400
    },
    {
      "epoch": 2.5210084033613445,
      "eval_loss": 0.4192233979701996,
      "eval_runtime": 13.5325,
      "eval_samples_per_second": 35.47,
      "eval_steps_per_second": 2.217,
      "step": 2400
    },
    {
      "epoch": 2.5315126050420167,
      "grad_norm": 0.7973752617835999,
      "learning_rate": 7.80812324929972e-06,
      "loss": 0.1436,
      "step": 2410
    },
    {
      "epoch": 2.542016806722689,
      "grad_norm": 0.6894790530204773,
      "learning_rate": 7.633053221288516e-06,
      "loss": 0.1425,
      "step": 2420
    },
    {
      "epoch": 2.552521008403361,
      "grad_norm": 0.6263079047203064,
      "learning_rate": 7.457983193277311e-06,
      "loss": 0.1473,
      "step": 2430
    },
    {
      "epoch": 2.5630252100840334,
      "grad_norm": 0.7887470722198486,
      "learning_rate": 7.282913165266108e-06,
      "loss": 0.1405,
      "step": 2440
    },
    {
      "epoch": 2.5735294117647056,
      "grad_norm": 0.6866456270217896,
      "learning_rate": 7.107843137254902e-06,
      "loss": 0.1409,
      "step": 2450
    },
    {
      "epoch": 2.5735294117647056,
      "eval_loss": 0.4140944480895996,
      "eval_runtime": 13.5509,
      "eval_samples_per_second": 35.422,
      "eval_steps_per_second": 2.214,
      "step": 2450
    },
    {
      "epoch": 2.5840336134453783,
      "grad_norm": 0.7917041778564453,
      "learning_rate": 6.932773109243698e-06,
      "loss": 0.1406,
      "step": 2460
    },
    {
      "epoch": 2.5945378151260505,
      "grad_norm": 0.6841098666191101,
      "learning_rate": 6.757703081232493e-06,
      "loss": 0.1415,
      "step": 2470
    },
    {
      "epoch": 2.6050420168067228,
      "grad_norm": 0.7307237982749939,
      "learning_rate": 6.582633053221289e-06,
      "loss": 0.1386,
      "step": 2480
    },
    {
      "epoch": 2.615546218487395,
      "grad_norm": 0.8012540340423584,
      "learning_rate": 6.407563025210084e-06,
      "loss": 0.1404,
      "step": 2490
    },
    {
      "epoch": 2.6260504201680672,
      "grad_norm": 0.7140881419181824,
      "learning_rate": 6.23249299719888e-06,
      "loss": 0.14,
      "step": 2500
    },
    {
      "epoch": 2.6260504201680672,
      "eval_loss": 0.42485204339027405,
      "eval_runtime": 13.5543,
      "eval_samples_per_second": 35.413,
      "eval_steps_per_second": 2.213,
      "step": 2500
    },
    {
      "epoch": 2.6365546218487395,
      "grad_norm": 0.8732954263687134,
      "learning_rate": 6.057422969187675e-06,
      "loss": 0.1459,
      "step": 2510
    },
    {
      "epoch": 2.6470588235294117,
      "grad_norm": 0.7600938677787781,
      "learning_rate": 5.882352941176471e-06,
      "loss": 0.1429,
      "step": 2520
    },
    {
      "epoch": 2.657563025210084,
      "grad_norm": 0.922292947769165,
      "learning_rate": 5.707282913165266e-06,
      "loss": 0.144,
      "step": 2530
    },
    {
      "epoch": 2.668067226890756,
      "grad_norm": 0.5363701581954956,
      "learning_rate": 5.532212885154062e-06,
      "loss": 0.1448,
      "step": 2540
    },
    {
      "epoch": 2.678571428571429,
      "grad_norm": 0.8106054663658142,
      "learning_rate": 5.357142857142857e-06,
      "loss": 0.1385,
      "step": 2550
    },
    {
      "epoch": 2.678571428571429,
      "eval_loss": 0.4227524399757385,
      "eval_runtime": 13.566,
      "eval_samples_per_second": 35.383,
      "eval_steps_per_second": 2.211,
      "step": 2550
    },
    {
      "epoch": 2.689075630252101,
      "grad_norm": 0.8006922006607056,
      "learning_rate": 5.182072829131653e-06,
      "loss": 0.1433,
      "step": 2560
    },
    {
      "epoch": 2.6995798319327733,
      "grad_norm": 0.6566533446311951,
      "learning_rate": 5.007002801120449e-06,
      "loss": 0.1374,
      "step": 2570
    },
    {
      "epoch": 2.7100840336134455,
      "grad_norm": 0.6680784225463867,
      "learning_rate": 4.831932773109244e-06,
      "loss": 0.1426,
      "step": 2580
    },
    {
      "epoch": 2.7205882352941178,
      "grad_norm": 0.7459174394607544,
      "learning_rate": 4.65686274509804e-06,
      "loss": 0.1412,
      "step": 2590
    },
    {
      "epoch": 2.73109243697479,
      "grad_norm": 0.6131850481033325,
      "learning_rate": 4.481792717086835e-06,
      "loss": 0.1417,
      "step": 2600
    },
    {
      "epoch": 2.73109243697479,
      "eval_loss": 0.42155376076698303,
      "eval_runtime": 13.5469,
      "eval_samples_per_second": 35.432,
      "eval_steps_per_second": 2.215,
      "step": 2600
    },
    {
      "epoch": 2.741596638655462,
      "grad_norm": 0.8135959506034851,
      "learning_rate": 4.30672268907563e-06,
      "loss": 0.137,
      "step": 2610
    },
    {
      "epoch": 2.7521008403361344,
      "grad_norm": 0.7350234985351562,
      "learning_rate": 4.131652661064426e-06,
      "loss": 0.1444,
      "step": 2620
    },
    {
      "epoch": 2.7626050420168067,
      "grad_norm": 2.5002171993255615,
      "learning_rate": 3.956582633053221e-06,
      "loss": 0.1448,
      "step": 2630
    },
    {
      "epoch": 2.773109243697479,
      "grad_norm": 0.9199714660644531,
      "learning_rate": 3.7815126050420167e-06,
      "loss": 0.1385,
      "step": 2640
    },
    {
      "epoch": 2.783613445378151,
      "grad_norm": 0.6327308416366577,
      "learning_rate": 3.606442577030812e-06,
      "loss": 0.1388,
      "step": 2650
    },
    {
      "epoch": 2.783613445378151,
      "eval_loss": 0.42358073592185974,
      "eval_runtime": 13.5314,
      "eval_samples_per_second": 35.473,
      "eval_steps_per_second": 2.217,
      "step": 2650
    },
    {
      "epoch": 2.7941176470588234,
      "grad_norm": 0.7996506690979004,
      "learning_rate": 3.431372549019608e-06,
      "loss": 0.1402,
      "step": 2660
    },
    {
      "epoch": 2.8046218487394956,
      "grad_norm": 0.7575409412384033,
      "learning_rate": 3.2563025210084036e-06,
      "loss": 0.1523,
      "step": 2670
    },
    {
      "epoch": 2.815126050420168,
      "grad_norm": 0.7814265489578247,
      "learning_rate": 3.081232492997199e-06,
      "loss": 0.1351,
      "step": 2680
    },
    {
      "epoch": 2.82563025210084,
      "grad_norm": 0.8243302702903748,
      "learning_rate": 2.9061624649859946e-06,
      "loss": 0.1362,
      "step": 2690
    },
    {
      "epoch": 2.8361344537815127,
      "grad_norm": 0.628792941570282,
      "learning_rate": 2.73109243697479e-06,
      "loss": 0.1423,
      "step": 2700
    },
    {
      "epoch": 2.8361344537815127,
      "eval_loss": 0.4288436770439148,
      "eval_runtime": 13.5168,
      "eval_samples_per_second": 35.511,
      "eval_steps_per_second": 2.219,
      "step": 2700
    }
  ],
  "logging_steps": 10,
  "max_steps": 2856,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2323898534854656e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}