{
  "best_global_step": 3000,
  "best_metric": 2.296126127243042,
  "best_model_checkpoint": "/kaggle/working/checkpoints/checkpoint-3000",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 3351,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0008954555630176853,
      "grad_norm": 331.45263671875,
      "learning_rate": 0.0,
      "loss": 105.3310546875,
      "num_input_tokens_seen": 8192,
      "step": 1,
      "train_runtime": 11.6116,
      "train_tokens_per_second": 705.503
    },
    {
      "epoch": 0.04477277815088426,
      "grad_norm": 183.94589233398438,
      "learning_rate": 0.0002916666666666667,
      "loss": 94.84019252232143,
      "num_input_tokens_seen": 409600,
      "step": 50,
      "train_runtime": 171.6655,
      "train_tokens_per_second": 2386.035
    },
    {
      "epoch": 0.08954555630176853,
      "grad_norm": 7.918169021606445,
      "learning_rate": 0.0005892857142857143,
      "loss": 44.2767529296875,
      "num_input_tokens_seen": 819200,
      "step": 100,
      "train_runtime": 344.252,
      "train_tokens_per_second": 2379.652
    },
    {
      "epoch": 0.1343183344526528,
      "grad_norm": 1.343171238899231,
      "learning_rate": 0.0008869047619047619,
      "loss": 20.50336669921875,
      "num_input_tokens_seen": 1228800,
      "step": 150,
      "train_runtime": 516.956,
      "train_tokens_per_second": 2376.991
    },
    {
      "epoch": 0.17909111260353705,
      "grad_norm": 0.5876757502555847,
      "learning_rate": 0.0009902607602890355,
      "loss": 16.116962890625,
      "num_input_tokens_seen": 1638400,
      "step": 200,
      "train_runtime": 689.6942,
      "train_tokens_per_second": 2375.546
    },
    {
      "epoch": 0.2238638907544213,
      "grad_norm": 0.5408444404602051,
      "learning_rate": 0.0009745523091423187,
      "loss": 14.958853759765624,
      "num_input_tokens_seen": 2048000,
      "step": 250,
      "train_runtime": 862.3198,
      "train_tokens_per_second": 2374.989
    },
    {
      "epoch": 0.2686366689053056,
      "grad_norm": 0.5192769169807434,
      "learning_rate": 0.0009588438579956017,
      "loss": 14.43544677734375,
      "num_input_tokens_seen": 2457600,
      "step": 300,
      "train_runtime": 1035.0406,
      "train_tokens_per_second": 2374.4
    },
    {
      "epoch": 0.31340944705618984,
      "grad_norm": 0.5099465847015381,
      "learning_rate": 0.0009431354068488848,
      "loss": 14.1262158203125,
      "num_input_tokens_seen": 2867200,
      "step": 350,
      "train_runtime": 1207.3524,
      "train_tokens_per_second": 2374.783
    },
    {
      "epoch": 0.3581822252070741,
      "grad_norm": 0.5959077477455139,
      "learning_rate": 0.0009274269557021679,
      "loss": 13.80364013671875,
      "num_input_tokens_seen": 3276800,
      "step": 400,
      "train_runtime": 1380.1593,
      "train_tokens_per_second": 2374.219
    },
    {
      "epoch": 0.40295500335795836,
      "grad_norm": 0.509476363658905,
      "learning_rate": 0.0009117185045554508,
      "loss": 13.52845458984375,
      "num_input_tokens_seen": 3686400,
      "step": 450,
      "train_runtime": 1552.8761,
      "train_tokens_per_second": 2373.918
    },
    {
      "epoch": 0.4477277815088426,
      "grad_norm": 0.51153963804245,
      "learning_rate": 0.0008960100534087339,
      "loss": 13.331270751953125,
      "num_input_tokens_seen": 4096000,
      "step": 500,
      "train_runtime": 1725.4211,
      "train_tokens_per_second": 2373.913
    },
    {
      "epoch": 0.4477277815088426,
      "eval_loss": 2.617971658706665,
      "eval_runtime": 13.0161,
      "eval_samples_per_second": 55.47,
      "eval_steps_per_second": 6.991,
      "num_input_tokens_seen": 4096000,
      "step": 500
    },
    {
      "epoch": 0.4925005596597269,
      "grad_norm": 0.5071282982826233,
      "learning_rate": 0.000880301602262017,
      "loss": 13.067698974609375,
      "num_input_tokens_seen": 4505600,
      "step": 550,
      "train_runtime": 1950.7794,
      "train_tokens_per_second": 2309.641
    },
    {
      "epoch": 0.5372733378106112,
      "grad_norm": 0.49176469445228577,
      "learning_rate": 0.0008645931511153,
      "loss": 13.0263720703125,
      "num_input_tokens_seen": 4915200,
      "step": 600,
      "train_runtime": 2123.5094,
      "train_tokens_per_second": 2314.659
    },
    {
      "epoch": 0.5820461159614954,
      "grad_norm": 0.5232459306716919,
      "learning_rate": 0.000848884699968583,
      "loss": 13.060653076171874,
      "num_input_tokens_seen": 5324800,
      "step": 650,
      "train_runtime": 2296.0456,
      "train_tokens_per_second": 2319.118
    },
    {
      "epoch": 0.6268188941123797,
      "grad_norm": 0.4987340271472931,
      "learning_rate": 0.0008331762488218661,
      "loss": 12.73482177734375,
      "num_input_tokens_seen": 5734400,
      "step": 700,
      "train_runtime": 2468.9137,
      "train_tokens_per_second": 2322.641
    },
    {
      "epoch": 0.6715916722632639,
      "grad_norm": 0.4983995258808136,
      "learning_rate": 0.0008174677976751492,
      "loss": 12.742056884765624,
      "num_input_tokens_seen": 6144000,
      "step": 750,
      "train_runtime": 2641.6641,
      "train_tokens_per_second": 2325.807
    },
    {
      "epoch": 0.7163644504141482,
      "grad_norm": 0.5150266885757446,
      "learning_rate": 0.0008017593465284323,
      "loss": 12.57524169921875,
      "num_input_tokens_seen": 6553600,
      "step": 800,
      "train_runtime": 2813.8488,
      "train_tokens_per_second": 2329.052
    },
    {
      "epoch": 0.7611372285650324,
      "grad_norm": 0.5139651894569397,
      "learning_rate": 0.0007860508953817153,
      "loss": 12.592891845703125,
      "num_input_tokens_seen": 6963200,
      "step": 850,
      "train_runtime": 2986.0731,
      "train_tokens_per_second": 2331.892
    },
    {
      "epoch": 0.8059100067159167,
      "grad_norm": 0.6878384351730347,
      "learning_rate": 0.0007703424442349984,
      "loss": 12.511319580078125,
      "num_input_tokens_seen": 7372800,
      "step": 900,
      "train_runtime": 3158.5594,
      "train_tokens_per_second": 2334.229
    },
    {
      "epoch": 0.8506827848668009,
      "grad_norm": 0.4996120035648346,
      "learning_rate": 0.0007546339930882815,
      "loss": 12.498067626953125,
      "num_input_tokens_seen": 7782400,
      "step": 950,
      "train_runtime": 3331.5807,
      "train_tokens_per_second": 2335.948
    },
    {
      "epoch": 0.8954555630176853,
      "grad_norm": 0.5436695218086243,
      "learning_rate": 0.0007389255419415646,
      "loss": 12.31913330078125,
      "num_input_tokens_seen": 8192000,
      "step": 1000,
      "train_runtime": 3503.8172,
      "train_tokens_per_second": 2338.022
    },
    {
      "epoch": 0.8954555630176853,
      "eval_loss": 2.4708571434020996,
      "eval_runtime": 12.9321,
      "eval_samples_per_second": 55.83,
      "eval_steps_per_second": 7.037,
      "num_input_tokens_seen": 8192000,
      "step": 1000
    },
    {
      "epoch": 0.9402283411685695,
      "grad_norm": 0.5411441326141357,
      "learning_rate": 0.0007232170907948476,
      "loss": 12.345836181640625,
      "num_input_tokens_seen": 8601600,
      "step": 1050,
      "train_runtime": 3734.2712,
      "train_tokens_per_second": 2303.421
    },
    {
      "epoch": 0.9850011193194538,
      "grad_norm": 0.5038391947746277,
      "learning_rate": 0.0007075086396481307,
      "loss": 12.264193115234375,
      "num_input_tokens_seen": 9011200,
      "step": 1100,
      "train_runtime": 3906.8195,
      "train_tokens_per_second": 2306.531
    },
    {
      "epoch": 1.0295500335795835,
      "grad_norm": 0.5121947526931763,
      "learning_rate": 0.0006918001885014138,
      "loss": 11.926865234375,
      "num_input_tokens_seen": 9418368,
      "step": 1150,
      "train_runtime": 4078.9352,
      "train_tokens_per_second": 2309.026
    },
    {
      "epoch": 1.074322811730468,
      "grad_norm": 0.5182917714118958,
      "learning_rate": 0.0006760917373546969,
      "loss": 12.025079345703125,
      "num_input_tokens_seen": 9827968,
      "step": 1200,
      "train_runtime": 4251.8799,
      "train_tokens_per_second": 2311.441
    },
    {
      "epoch": 1.1190955898813522,
      "grad_norm": 0.518774688243866,
      "learning_rate": 0.0006603832862079799,
      "loss": 11.881273193359375,
      "num_input_tokens_seen": 10237568,
      "step": 1250,
      "train_runtime": 4424.9422,
      "train_tokens_per_second": 2313.605
    },
    {
      "epoch": 1.1638683680322364,
      "grad_norm": 0.49377894401550293,
      "learning_rate": 0.000644674835061263,
      "loss": 11.88373779296875,
      "num_input_tokens_seen": 10647168,
      "step": 1300,
      "train_runtime": 4597.7399,
      "train_tokens_per_second": 2315.74
    },
    {
      "epoch": 1.2086411461831206,
      "grad_norm": 0.5340304970741272,
      "learning_rate": 0.0006289663839145461,
      "loss": 11.859654541015624,
      "num_input_tokens_seen": 11056768,
      "step": 1350,
      "train_runtime": 4770.4911,
      "train_tokens_per_second": 2317.742
    },
    {
      "epoch": 1.253413924334005,
      "grad_norm": 0.5341930985450745,
      "learning_rate": 0.0006132579327678292,
      "loss": 11.792633056640625,
      "num_input_tokens_seen": 11466368,
      "step": 1400,
      "train_runtime": 4943.2665,
      "train_tokens_per_second": 2319.593
    },
    {
      "epoch": 1.2981867024848892,
      "grad_norm": 0.5263528823852539,
      "learning_rate": 0.0005975494816211122,
      "loss": 11.814337158203125,
      "num_input_tokens_seen": 11875968,
      "step": 1450,
      "train_runtime": 5116.4191,
      "train_tokens_per_second": 2321.148
    },
    {
      "epoch": 1.3429594806357734,
      "grad_norm": 0.49405381083488464,
      "learning_rate": 0.0005818410304743952,
      "loss": 11.8128759765625,
      "num_input_tokens_seen": 12285568,
      "step": 1500,
      "train_runtime": 5288.8716,
      "train_tokens_per_second": 2322.909
    },
    {
      "epoch": 1.3429594806357734,
      "eval_loss": 2.402881622314453,
      "eval_runtime": 12.9253,
      "eval_samples_per_second": 55.86,
      "eval_steps_per_second": 7.04,
      "num_input_tokens_seen": 12285568,
      "step": 1500
    },
    {
      "epoch": 1.3877322587866576,
      "grad_norm": 0.5319132208824158,
      "learning_rate": 0.0005661325793276783,
      "loss": 11.62824462890625,
      "num_input_tokens_seen": 12695168,
      "step": 1550,
      "train_runtime": 5520.1498,
      "train_tokens_per_second": 2299.787
    },
    {
      "epoch": 1.432505036937542,
      "grad_norm": 0.5469394326210022,
      "learning_rate": 0.0005504241281809613,
      "loss": 11.611361083984375,
      "num_input_tokens_seen": 13104768,
      "step": 1600,
      "train_runtime": 5692.7354,
      "train_tokens_per_second": 2302.016
    },
    {
      "epoch": 1.4772778150884263,
      "grad_norm": 0.5172468423843384,
      "learning_rate": 0.0005347156770342444,
      "loss": 11.5665576171875,
      "num_input_tokens_seen": 13514368,
      "step": 1650,
      "train_runtime": 5866.6814,
      "train_tokens_per_second": 2303.58
    },
    {
      "epoch": 1.5220505932393105,
      "grad_norm": 0.5269997119903564,
      "learning_rate": 0.0005190072258875275,
      "loss": 11.613292236328125,
      "num_input_tokens_seen": 13923968,
      "step": 1700,
      "train_runtime": 6040.4262,
      "train_tokens_per_second": 2305.13
    },
    {
      "epoch": 1.566823371390195,
      "grad_norm": 0.5272865891456604,
      "learning_rate": 0.0005032987747408106,
      "loss": 11.535517578125,
      "num_input_tokens_seen": 14333568,
      "step": 1750,
      "train_runtime": 6214.6971,
      "train_tokens_per_second": 2306.398
    },
    {
      "epoch": 1.611596149541079,
      "grad_norm": 0.48300108313560486,
      "learning_rate": 0.0004875903235940936,
      "loss": 11.529622802734375,
      "num_input_tokens_seen": 14743168,
      "step": 1800,
      "train_runtime": 6388.2344,
      "train_tokens_per_second": 2307.863
    },
    {
      "epoch": 1.6563689276919633,
      "grad_norm": 0.5525538921356201,
      "learning_rate": 0.00047188187244737666,
      "loss": 11.44923828125,
      "num_input_tokens_seen": 15152768,
      "step": 1850,
      "train_runtime": 6561.8162,
      "train_tokens_per_second": 2309.234
    },
    {
      "epoch": 1.7011417058428475,
      "grad_norm": 0.5223372578620911,
      "learning_rate": 0.00045617342130065976,
      "loss": 11.444879150390625,
      "num_input_tokens_seen": 15562368,
      "step": 1900,
      "train_runtime": 6735.1477,
      "train_tokens_per_second": 2310.62
    },
    {
      "epoch": 1.7459144839937317,
      "grad_norm": 0.5123465657234192,
      "learning_rate": 0.0004404649701539428,
      "loss": 11.449864501953124,
      "num_input_tokens_seen": 15971968,
      "step": 1950,
      "train_runtime": 6908.8412,
      "train_tokens_per_second": 2311.816
    },
    {
      "epoch": 1.790687262144616,
      "grad_norm": 0.5051376819610596,
      "learning_rate": 0.0004247565190072259,
      "loss": 11.4521826171875,
      "num_input_tokens_seen": 16381568,
      "step": 2000,
      "train_runtime": 7082.8117,
      "train_tokens_per_second": 2312.862
    },
    {
      "epoch": 1.790687262144616,
      "eval_loss": 2.3497679233551025,
      "eval_runtime": 13.1104,
      "eval_samples_per_second": 55.071,
      "eval_steps_per_second": 6.941,
      "num_input_tokens_seen": 16381568,
      "step": 2000
    },
    {
      "epoch": 1.8354600402955004,
      "grad_norm": 0.5100616216659546,
      "learning_rate": 0.00040904806786050895,
      "loss": 11.323531494140624,
      "num_input_tokens_seen": 16791168,
      "step": 2050,
      "train_runtime": 7316.4951,
      "train_tokens_per_second": 2294.974
    },
    {
      "epoch": 1.8802328184463846,
      "grad_norm": 0.5197663307189941,
      "learning_rate": 0.00039333961671379205,
      "loss": 11.3488134765625,
      "num_input_tokens_seen": 17200768,
      "step": 2100,
      "train_runtime": 7490.4692,
      "train_tokens_per_second": 2296.354
    },
    {
      "epoch": 1.925005596597269,
      "grad_norm": 0.5394102334976196,
      "learning_rate": 0.0003776311655670751,
      "loss": 11.24622802734375,
      "num_input_tokens_seen": 17610368,
      "step": 2150,
      "train_runtime": 7664.0786,
      "train_tokens_per_second": 2297.78
    },
    {
      "epoch": 1.9697783747481532,
      "grad_norm": 0.5379300713539124,
      "learning_rate": 0.0003619227144203582,
      "loss": 11.337928466796875,
      "num_input_tokens_seen": 18019968,
      "step": 2200,
      "train_runtime": 7837.6015,
      "train_tokens_per_second": 2299.169
    },
    {
      "epoch": 2.014327289008283,
      "grad_norm": 0.5014926791191101,
      "learning_rate": 0.00034621426327364124,
      "loss": 11.289034423828125,
      "num_input_tokens_seen": 18427136,
      "step": 2250,
      "train_runtime": 8010.7858,
      "train_tokens_per_second": 2300.291
    },
    {
      "epoch": 2.059100067159167,
      "grad_norm": 0.5547139048576355,
      "learning_rate": 0.0003305058121269243,
      "loss": 11.22334716796875,
      "num_input_tokens_seen": 18836736,
      "step": 2300,
      "train_runtime": 8184.6927,
      "train_tokens_per_second": 2301.459
    },
    {
      "epoch": 2.1038728453100517,
      "grad_norm": 0.5273997783660889,
      "learning_rate": 0.00031479736098020733,
      "loss": 11.20933349609375,
      "num_input_tokens_seen": 19246336,
      "step": 2350,
      "train_runtime": 8358.1995,
      "train_tokens_per_second": 2302.689
    },
    {
      "epoch": 2.148645623460936,
      "grad_norm": 0.5178916454315186,
      "learning_rate": 0.00029908890983349043,
      "loss": 11.13134033203125,
      "num_input_tokens_seen": 19655936,
      "step": 2400,
      "train_runtime": 8531.9507,
      "train_tokens_per_second": 2303.803
    },
    {
      "epoch": 2.19341840161182,
      "grad_norm": 0.5172515511512756,
      "learning_rate": 0.0002833804586867735,
      "loss": 11.15967041015625,
      "num_input_tokens_seen": 20065536,
      "step": 2450,
      "train_runtime": 8706.1654,
      "train_tokens_per_second": 2304.75
    },
    {
      "epoch": 2.2381911797627043,
      "grad_norm": 0.5247936248779297,
      "learning_rate": 0.00026767200754005657,
      "loss": 11.007467041015625,
      "num_input_tokens_seen": 20475136,
      "step": 2500,
      "train_runtime": 8879.7506,
      "train_tokens_per_second": 2305.823
    },
    {
      "epoch": 2.2381911797627043,
      "eval_loss": 2.3171231746673584,
      "eval_runtime": 13.2491,
      "eval_samples_per_second": 54.494,
      "eval_steps_per_second": 6.868,
      "num_input_tokens_seen": 20475136,
      "step": 2500
    },
    {
      "epoch": 2.2829639579135885,
      "grad_norm": 0.5802310705184937,
      "learning_rate": 0.0002519635563933396,
      "loss": 11.119674072265624,
      "num_input_tokens_seen": 20884736,
      "step": 2550,
      "train_runtime": 9116.3008,
      "train_tokens_per_second": 2290.922
    },
    {
      "epoch": 2.3277367360644727,
      "grad_norm": 0.4896214008331299,
      "learning_rate": 0.0002362551052466227,
      "loss": 11.020155029296875,
      "num_input_tokens_seen": 21294336,
      "step": 2600,
      "train_runtime": 9289.9843,
      "train_tokens_per_second": 2292.182
    },
    {
      "epoch": 2.372509514215357,
      "grad_norm": 0.5538379549980164,
      "learning_rate": 0.00022054665409990576,
      "loss": 11.0175,
      "num_input_tokens_seen": 21703936,
      "step": 2650,
      "train_runtime": 9463.3153,
      "train_tokens_per_second": 2293.481
    },
    {
      "epoch": 2.417282292366241,
      "grad_norm": 0.5305850505828857,
      "learning_rate": 0.0002048382029531888,
      "loss": 11.08681884765625,
      "num_input_tokens_seen": 22113536,
      "step": 2700,
      "train_runtime": 9636.7051,
      "train_tokens_per_second": 2294.72
    },
    {
      "epoch": 2.462055070517126,
      "grad_norm": 0.5520696640014648,
      "learning_rate": 0.00018912975180647188,
      "loss": 11.049239501953124,
      "num_input_tokens_seen": 22523136,
      "step": 2750,
      "train_runtime": 9810.3229,
      "train_tokens_per_second": 2295.861
    },
    {
      "epoch": 2.50682784866801,
      "grad_norm": 0.5960472226142883,
      "learning_rate": 0.00017342130065975495,
      "loss": 11.0529296875,
      "num_input_tokens_seen": 22932736,
      "step": 2800,
      "train_runtime": 9984.0418,
      "train_tokens_per_second": 2296.939
    },
    {
      "epoch": 2.551600626818894,
      "grad_norm": 0.5422330498695374,
      "learning_rate": 0.00015771284951303802,
      "loss": 11.079190673828125,
      "num_input_tokens_seen": 23342336,
      "step": 2850,
      "train_runtime": 10157.5512,
      "train_tokens_per_second": 2298.028
    },
    {
      "epoch": 2.5963734049697784,
      "grad_norm": 0.5332451462745667,
      "learning_rate": 0.0001420043983663211,
      "loss": 11.045711669921875,
      "num_input_tokens_seen": 23751936,
      "step": 2900,
      "train_runtime": 10330.8794,
      "train_tokens_per_second": 2299.12
    },
    {
      "epoch": 2.6411461831206626,
      "grad_norm": 0.5149208903312683,
      "learning_rate": 0.00012629594721960417,
      "loss": 10.973729248046874,
      "num_input_tokens_seen": 24161536,
      "step": 2950,
      "train_runtime": 10505.0191,
      "train_tokens_per_second": 2299.999
    },
    {
      "epoch": 2.685918961271547,
      "grad_norm": 0.5152694582939148,
      "learning_rate": 0.00011058749607288723,
      "loss": 10.958800048828126,
      "num_input_tokens_seen": 24571136,
      "step": 3000,
      "train_runtime": 10678.159,
      "train_tokens_per_second": 2301.065
    },
    {
      "epoch": 2.685918961271547,
      "eval_loss": 2.296126127243042,
      "eval_runtime": 13.2029,
      "eval_samples_per_second": 54.685,
      "eval_steps_per_second": 6.892,
      "num_input_tokens_seen": 24571136,
      "step": 3000
    },
    {
      "epoch": 2.730691739422431,
      "grad_norm": 0.5589572191238403,
      "learning_rate": 9.487904492617027e-05,
      "loss": 11.127513427734375,
      "num_input_tokens_seen": 24980736,
      "step": 3050,
      "train_runtime": 10916.5795,
      "train_tokens_per_second": 2288.33
    },
    {
      "epoch": 2.7754645175733152,
      "grad_norm": 0.5186505317687988,
      "learning_rate": 7.917059377945334e-05,
      "loss": 11.1047265625,
      "num_input_tokens_seen": 25390336,
      "step": 3100,
      "train_runtime": 11090.2623,
      "train_tokens_per_second": 2289.426
    },
    {
      "epoch": 2.8202372957241995,
      "grad_norm": 0.510116457939148,
      "learning_rate": 6.346214263273642e-05,
      "loss": 11.018306884765625,
      "num_input_tokens_seen": 25799936,
      "step": 3150,
      "train_runtime": 11263.7763,
      "train_tokens_per_second": 2290.523
    },
    {
      "epoch": 2.865010073875084,
      "grad_norm": 0.5256752967834473,
      "learning_rate": 4.7753691486019474e-05,
      "loss": 10.991341552734376,
      "num_input_tokens_seen": 26209536,
      "step": 3200,
      "train_runtime": 11436.8692,
      "train_tokens_per_second": 2291.671
    },
    {
      "epoch": 2.9097828520259683,
      "grad_norm": 0.553755521774292,
      "learning_rate": 3.2045240339302546e-05,
      "loss": 10.97140625,
      "num_input_tokens_seen": 26619136,
      "step": 3250,
      "train_runtime": 11610.6191,
      "train_tokens_per_second": 2292.654
    },
    {
      "epoch": 2.9545556301768525,
      "grad_norm": 0.5288929343223572,
      "learning_rate": 1.633678919258561e-05,
      "loss": 11.016265869140625,
      "num_input_tokens_seen": 27028736,
      "step": 3300,
      "train_runtime": 11784.0858,
      "train_tokens_per_second": 2293.664
    },
    {
      "epoch": 2.9993284083277367,
      "grad_norm": 0.5394493937492371,
      "learning_rate": 6.283380458686773e-07,
      "loss": 10.937723388671875,
      "num_input_tokens_seen": 27438336,
      "step": 3350,
      "train_runtime": 11957.2304,
      "train_tokens_per_second": 2294.707
    },
    {
      "epoch": 3.0,
      "num_input_tokens_seen": 27444096,
      "step": 3351,
      "total_flos": 4.973428737441792e+16,
      "train_loss": 13.759503722582174,
      "train_runtime": 12000.3856,
      "train_samples_per_second": 17.867,
      "train_steps_per_second": 0.279
    }
  ],
  "logging_steps": 50,
  "max_steps": 3351,
  "num_input_tokens_seen": 27444096,
  "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": true
      },
      "attributes": {}
    }
  },
  "total_flos": 4.973428737441792e+16,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}