{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.7621572363515865,
  "eval_steps": 500,
  "global_step": 2300,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0016568635572860575,
      "grad_norm": 1.7647515535354614,
      "learning_rate": 4.9999945820982895e-05,
      "loss": 3.3455,
      "num_input_tokens_seen": 128544,
      "step": 5,
      "train_runtime": 123.5268,
      "train_tokens_per_second": 1040.616
    },
    {
      "epoch": 0.003313727114572115,
      "grad_norm": 1.6262233257293701,
      "learning_rate": 4.999972571912837e-05,
      "loss": 2.9109,
      "num_input_tokens_seen": 273256,
      "step": 10,
      "train_runtime": 261.7686,
      "train_tokens_per_second": 1043.884
    },
    {
      "epoch": 0.004970590671858173,
      "grad_norm": 1.5207526683807373,
      "learning_rate": 4.99993363097373e-05,
      "loss": 2.2435,
      "num_input_tokens_seen": 401856,
      "step": 15,
      "train_runtime": 388.1031,
      "train_tokens_per_second": 1035.436
    },
    {
      "epoch": 0.00662745422914423,
      "grad_norm": 1.258175015449524,
      "learning_rate": 4.999877759544695e-05,
      "loss": 1.8225,
      "num_input_tokens_seen": 534584,
      "step": 20,
      "train_runtime": 524.4923,
      "train_tokens_per_second": 1019.241
    },
    {
      "epoch": 0.008284317786430288,
      "grad_norm": 1.4143387079238892,
      "learning_rate": 4.999804958004111e-05,
      "loss": 1.5736,
      "num_input_tokens_seen": 663592,
      "step": 25,
      "train_runtime": 646.2405,
      "train_tokens_per_second": 1026.85
    },
    {
      "epoch": 0.009941181343716346,
      "grad_norm": 0.9895913004875183,
      "learning_rate": 4.9997152268450184e-05,
      "loss": 1.36,
      "num_input_tokens_seen": 826480,
      "step": 30,
      "train_runtime": 801.1986,
      "train_tokens_per_second": 1031.554
    },
    {
      "epoch": 0.011598044901002403,
      "grad_norm": 4.952330589294434,
      "learning_rate": 4.999608566675111e-05,
      "loss": 1.1473,
      "num_input_tokens_seen": 955544,
      "step": 35,
      "train_runtime": 928.6777,
      "train_tokens_per_second": 1028.93
    },
    {
      "epoch": 0.01325490845828846,
      "grad_norm": 0.7959768772125244,
      "learning_rate": 4.999484978216731e-05,
      "loss": 1.1066,
      "num_input_tokens_seen": 1090352,
      "step": 40,
      "train_runtime": 1061.3778,
      "train_tokens_per_second": 1027.299
    },
    {
      "epoch": 0.014911772015574518,
      "grad_norm": 0.894897997379303,
      "learning_rate": 4.999344462306866e-05,
      "loss": 1.0412,
      "num_input_tokens_seen": 1201624,
      "step": 45,
      "train_runtime": 1163.7802,
      "train_tokens_per_second": 1032.518
    },
    {
      "epoch": 0.016568635572860577,
      "grad_norm": 0.6151811480522156,
      "learning_rate": 4.999187019897143e-05,
      "loss": 0.9477,
      "num_input_tokens_seen": 1347080,
      "step": 50,
      "train_runtime": 1305.0287,
      "train_tokens_per_second": 1032.222
    },
    {
      "epoch": 0.018225499130146632,
      "grad_norm": 0.6643481850624084,
      "learning_rate": 4.999012652053821e-05,
      "loss": 0.9077,
      "num_input_tokens_seen": 1514680,
      "step": 55,
      "train_runtime": 1481.8935,
      "train_tokens_per_second": 1022.125
    },
    {
      "epoch": 0.01988236268743269,
      "grad_norm": 0.799781858921051,
      "learning_rate": 4.998821359957783e-05,
      "loss": 0.8658,
      "num_input_tokens_seen": 1645168,
      "step": 60,
      "train_runtime": 1615.466,
      "train_tokens_per_second": 1018.386
    },
    {
      "epoch": 0.021539226244718747,
      "grad_norm": 0.6649002432823181,
      "learning_rate": 4.9986131449045325e-05,
      "loss": 0.8667,
      "num_input_tokens_seen": 1773656,
      "step": 65,
      "train_runtime": 1726.036,
      "train_tokens_per_second": 1027.589
    },
    {
      "epoch": 0.023196089802004806,
      "grad_norm": 0.6551240086555481,
      "learning_rate": 4.99838800830418e-05,
      "loss": 0.8498,
      "num_input_tokens_seen": 1896888,
      "step": 70,
      "train_runtime": 1850.7767,
      "train_tokens_per_second": 1024.915
    },
    {
      "epoch": 0.02485295335929086,
      "grad_norm": 4.035894870758057,
      "learning_rate": 4.9981459516814344e-05,
      "loss": 0.7846,
      "num_input_tokens_seen": 2026264,
      "step": 75,
      "train_runtime": 1982.4682,
      "train_tokens_per_second": 1022.092
    },
    {
      "epoch": 0.02650981691657692,
      "grad_norm": 0.6371873021125793,
      "learning_rate": 4.9978869766755954e-05,
      "loss": 0.8202,
      "num_input_tokens_seen": 2169480,
      "step": 80,
      "train_runtime": 2124.2418,
      "train_tokens_per_second": 1021.296
    },
    {
      "epoch": 0.028166680473862976,
      "grad_norm": 0.542093813419342,
      "learning_rate": 4.9976110850405377e-05,
      "loss": 0.7494,
      "num_input_tokens_seen": 2288704,
      "step": 85,
      "train_runtime": 2238.91,
      "train_tokens_per_second": 1022.24
    },
    {
      "epoch": 0.029823544031149035,
      "grad_norm": 0.5972122550010681,
      "learning_rate": 4.9973182786447034e-05,
      "loss": 0.7981,
      "num_input_tokens_seen": 2411008,
      "step": 90,
      "train_runtime": 2345.7791,
      "train_tokens_per_second": 1027.807
    },
    {
      "epoch": 0.031480407588435094,
      "grad_norm": 0.6272138953208923,
      "learning_rate": 4.997008559471089e-05,
      "loss": 0.7024,
      "num_input_tokens_seen": 2508152,
      "step": 95,
      "train_runtime": 2437.28,
      "train_tokens_per_second": 1029.078
    },
    {
      "epoch": 0.03313727114572115,
      "grad_norm": 0.6879661679267883,
      "learning_rate": 4.9966819296172276e-05,
      "loss": 0.6319,
      "num_input_tokens_seen": 2609472,
      "step": 100,
      "train_runtime": 2530.7928,
      "train_tokens_per_second": 1031.089
    },
    {
      "epoch": 0.034794134703007205,
      "grad_norm": 0.6695439219474792,
      "learning_rate": 4.9963383912951786e-05,
      "loss": 0.627,
      "num_input_tokens_seen": 2712568,
      "step": 105,
      "train_runtime": 2629.2895,
      "train_tokens_per_second": 1031.673
    },
    {
      "epoch": 0.036450998260293264,
      "grad_norm": 0.7428164482116699,
      "learning_rate": 4.9959779468315136e-05,
      "loss": 0.6415,
      "num_input_tokens_seen": 2830312,
      "step": 110,
      "train_runtime": 2740.7139,
      "train_tokens_per_second": 1032.692
    },
    {
      "epoch": 0.038107861817579324,
      "grad_norm": 0.7692319750785828,
      "learning_rate": 4.995600598667299e-05,
      "loss": 0.6833,
      "num_input_tokens_seen": 2986688,
      "step": 115,
      "train_runtime": 2894.4688,
      "train_tokens_per_second": 1031.86
    },
    {
      "epoch": 0.03976472537486538,
      "grad_norm": 0.6514962315559387,
      "learning_rate": 4.995206349358077e-05,
      "loss": 0.6488,
      "num_input_tokens_seen": 3093000,
      "step": 120,
      "train_runtime": 3001.3489,
      "train_tokens_per_second": 1030.537
    },
    {
      "epoch": 0.041421588932151435,
      "grad_norm": 0.82510906457901,
      "learning_rate": 4.994795201573852e-05,
      "loss": 0.6406,
      "num_input_tokens_seen": 3223184,
      "step": 125,
      "train_runtime": 3136.5623,
      "train_tokens_per_second": 1027.617
    },
    {
      "epoch": 0.043078452489437494,
      "grad_norm": 0.7509822249412537,
      "learning_rate": 4.9943671580990715e-05,
      "loss": 0.6041,
      "num_input_tokens_seen": 3328912,
      "step": 130,
      "train_runtime": 3229.0788,
      "train_tokens_per_second": 1030.917
    },
    {
      "epoch": 0.04473531604672355,
      "grad_norm": 0.620428740978241,
      "learning_rate": 4.993922221832607e-05,
      "loss": 0.6418,
      "num_input_tokens_seen": 3434680,
      "step": 135,
      "train_runtime": 3328.6227,
      "train_tokens_per_second": 1031.862
    },
    {
      "epoch": 0.04639217960400961,
      "grad_norm": 1.1372716426849365,
      "learning_rate": 4.9934603957877335e-05,
      "loss": 0.6634,
      "num_input_tokens_seen": 3603232,
      "step": 140,
      "train_runtime": 3491.5943,
      "train_tokens_per_second": 1031.973
    },
    {
      "epoch": 0.048049043161295664,
      "grad_norm": 0.5722112059593201,
      "learning_rate": 4.9929816830921115e-05,
      "loss": 0.5866,
      "num_input_tokens_seen": 3721720,
      "step": 145,
      "train_runtime": 3605.8586,
      "train_tokens_per_second": 1032.131
    },
    {
      "epoch": 0.04970590671858172,
      "grad_norm": 0.5459595918655396,
      "learning_rate": 4.9924860869877633e-05,
      "loss": 0.5363,
      "num_input_tokens_seen": 3842472,
      "step": 150,
      "train_runtime": 3724.3207,
      "train_tokens_per_second": 1031.724
    },
    {
      "epoch": 0.05136277027586778,
      "grad_norm": 0.7168979048728943,
      "learning_rate": 4.991973610831051e-05,
      "loss": 0.6169,
      "num_input_tokens_seen": 3956800,
      "step": 155,
      "train_runtime": 3839.349,
      "train_tokens_per_second": 1030.591
    },
    {
      "epoch": 0.05301963383315384,
      "grad_norm": 0.6107162237167358,
      "learning_rate": 4.9914442580926555e-05,
      "loss": 0.6585,
      "num_input_tokens_seen": 4089080,
      "step": 160,
      "train_runtime": 3971.3037,
      "train_tokens_per_second": 1029.657
    },
    {
      "epoch": 0.0546764973904399,
      "grad_norm": 0.5629305839538574,
      "learning_rate": 4.9908980323575535e-05,
      "loss": 0.6225,
      "num_input_tokens_seen": 4193568,
      "step": 165,
      "train_runtime": 4070.9448,
      "train_tokens_per_second": 1030.122
    },
    {
      "epoch": 0.05633336094772595,
      "grad_norm": 0.7005344033241272,
      "learning_rate": 4.990334937324991e-05,
      "loss": 0.6845,
      "num_input_tokens_seen": 4324064,
      "step": 170,
      "train_runtime": 4203.6347,
      "train_tokens_per_second": 1028.649
    },
    {
      "epoch": 0.05799022450501201,
      "grad_norm": 0.5772902369499207,
      "learning_rate": 4.989754976808458e-05,
      "loss": 0.5972,
      "num_input_tokens_seen": 4463584,
      "step": 175,
      "train_runtime": 4343.4188,
      "train_tokens_per_second": 1027.666
    },
    {
      "epoch": 0.05964708806229807,
      "grad_norm": 0.7041813731193542,
      "learning_rate": 4.989158154735666e-05,
      "loss": 0.5332,
      "num_input_tokens_seen": 4601480,
      "step": 180,
      "train_runtime": 4486.1595,
      "train_tokens_per_second": 1025.706
    },
    {
      "epoch": 0.06130395161958413,
      "grad_norm": 0.6148178577423096,
      "learning_rate": 4.988544475148518e-05,
      "loss": 0.6253,
      "num_input_tokens_seen": 4734368,
      "step": 185,
      "train_runtime": 4618.2499,
      "train_tokens_per_second": 1025.143
    },
    {
      "epoch": 0.06296081517687019,
      "grad_norm": 0.745169460773468,
      "learning_rate": 4.987913942203082e-05,
      "loss": 0.6261,
      "num_input_tokens_seen": 4863880,
      "step": 190,
      "train_runtime": 4745.5749,
      "train_tokens_per_second": 1024.93
    },
    {
      "epoch": 0.06461767873415625,
      "grad_norm": 0.5965126156806946,
      "learning_rate": 4.987266560169565e-05,
      "loss": 0.5186,
      "num_input_tokens_seen": 5011824,
      "step": 195,
      "train_runtime": 4900.5102,
      "train_tokens_per_second": 1022.715
    },
    {
      "epoch": 0.0662745422914423,
      "grad_norm": 0.568074107170105,
      "learning_rate": 4.986602333432282e-05,
      "loss": 0.5941,
      "num_input_tokens_seen": 5179888,
      "step": 200,
      "train_runtime": 5079.9699,
      "train_tokens_per_second": 1019.669
    },
    {
      "epoch": 0.06793140584872835,
      "grad_norm": 0.5502363443374634,
      "learning_rate": 4.985921266489625e-05,
      "loss": 0.5913,
      "num_input_tokens_seen": 5338664,
      "step": 205,
      "train_runtime": 5257.7076,
      "train_tokens_per_second": 1015.398
    },
    {
      "epoch": 0.06958826940601441,
      "grad_norm": 0.6765833497047424,
      "learning_rate": 4.9852233639540355e-05,
      "loss": 0.5628,
      "num_input_tokens_seen": 5473200,
      "step": 210,
      "train_runtime": 5386.1387,
      "train_tokens_per_second": 1016.164
    },
    {
      "epoch": 0.07124513296330047,
      "grad_norm": 0.6165874600410461,
      "learning_rate": 4.984508630551972e-05,
      "loss": 0.5859,
      "num_input_tokens_seen": 5610360,
      "step": 215,
      "train_runtime": 5521.1634,
      "train_tokens_per_second": 1016.155
    },
    {
      "epoch": 0.07290199652058653,
      "grad_norm": 0.6182301640510559,
      "learning_rate": 4.9837770711238763e-05,
      "loss": 0.5505,
      "num_input_tokens_seen": 5757168,
      "step": 220,
      "train_runtime": 5669.98,
      "train_tokens_per_second": 1015.377
    },
    {
      "epoch": 0.07455886007787259,
      "grad_norm": 0.6024623513221741,
      "learning_rate": 4.9830286906241456e-05,
      "loss": 0.5931,
      "num_input_tokens_seen": 5880584,
      "step": 225,
      "train_runtime": 5784.6269,
      "train_tokens_per_second": 1016.588
    },
    {
      "epoch": 0.07621572363515865,
      "grad_norm": 0.5623393058776855,
      "learning_rate": 4.9822634941210935e-05,
      "loss": 0.519,
      "num_input_tokens_seen": 6020704,
      "step": 230,
      "train_runtime": 5921.8783,
      "train_tokens_per_second": 1016.688
    },
    {
      "epoch": 0.0778725871924447,
      "grad_norm": 0.5631406903266907,
      "learning_rate": 4.981481486796917e-05,
      "loss": 0.5079,
      "num_input_tokens_seen": 6118584,
      "step": 235,
      "train_runtime": 6019.1943,
      "train_tokens_per_second": 1016.512
    },
    {
      "epoch": 0.07952945074973077,
      "grad_norm": 0.6267772316932678,
      "learning_rate": 4.980682673947665e-05,
      "loss": 0.544,
      "num_input_tokens_seen": 6263880,
      "step": 240,
      "train_runtime": 6173.7946,
      "train_tokens_per_second": 1014.592
    },
    {
      "epoch": 0.08118631430701681,
      "grad_norm": 0.6894953846931458,
      "learning_rate": 4.9798670609831975e-05,
      "loss": 0.5525,
      "num_input_tokens_seen": 6428192,
      "step": 245,
      "train_runtime": 6344.0719,
      "train_tokens_per_second": 1013.26
    },
    {
      "epoch": 0.08284317786430287,
      "grad_norm": 0.7429189085960388,
      "learning_rate": 4.979034653427152e-05,
      "loss": 0.5691,
      "num_input_tokens_seen": 6547824,
      "step": 250,
      "train_runtime": 6471.2371,
      "train_tokens_per_second": 1011.835
    },
    {
      "epoch": 0.08450004142158893,
      "grad_norm": 0.6777173280715942,
      "learning_rate": 4.978185456916906e-05,
      "loss": 0.5679,
      "num_input_tokens_seen": 6713080,
      "step": 255,
      "train_runtime": 6641.485,
      "train_tokens_per_second": 1010.78
    },
    {
      "epoch": 0.08615690497887499,
      "grad_norm": 0.6823623180389404,
      "learning_rate": 4.977319477203536e-05,
      "loss": 0.5725,
      "num_input_tokens_seen": 6860960,
      "step": 260,
      "train_runtime": 6785.018,
      "train_tokens_per_second": 1011.193
    },
    {
      "epoch": 0.08781376853616105,
      "grad_norm": 0.7490869164466858,
      "learning_rate": 4.9764367201517836e-05,
      "loss": 0.5773,
      "num_input_tokens_seen": 7004400,
      "step": 265,
      "train_runtime": 6928.5479,
      "train_tokens_per_second": 1010.948
    },
    {
      "epoch": 0.0894706320934471,
      "grad_norm": 0.7022630572319031,
      "learning_rate": 4.9755371917400096e-05,
      "loss": 0.581,
      "num_input_tokens_seen": 7114152,
      "step": 270,
      "train_runtime": 7028.1637,
      "train_tokens_per_second": 1012.235
    },
    {
      "epoch": 0.09112749565073316,
      "grad_norm": 0.7099390029907227,
      "learning_rate": 4.9746208980601596e-05,
      "loss": 0.5374,
      "num_input_tokens_seen": 7229656,
      "step": 275,
      "train_runtime": 7138.8234,
      "train_tokens_per_second": 1012.724
    },
    {
      "epoch": 0.09278435920801922,
      "grad_norm": 0.6199091076850891,
      "learning_rate": 4.973687845317719e-05,
      "loss": 0.4653,
      "num_input_tokens_seen": 7329744,
      "step": 280,
      "train_runtime": 7224.596,
      "train_tokens_per_second": 1014.554
    },
    {
      "epoch": 0.09444122276530528,
      "grad_norm": 0.6699258089065552,
      "learning_rate": 4.97273803983167e-05,
      "loss": 0.5485,
      "num_input_tokens_seen": 7485064,
      "step": 285,
      "train_runtime": 7387.6287,
      "train_tokens_per_second": 1013.189
    },
    {
      "epoch": 0.09609808632259133,
      "grad_norm": 0.6561999917030334,
      "learning_rate": 4.9717714880344536e-05,
      "loss": 0.4892,
      "num_input_tokens_seen": 7625952,
      "step": 290,
      "train_runtime": 7532.8837,
      "train_tokens_per_second": 1012.355
    },
    {
      "epoch": 0.09775494987987739,
      "grad_norm": 0.5645545125007629,
      "learning_rate": 4.970788196471922e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 7761616,
      "step": 295,
      "train_runtime": 7671.3844,
      "train_tokens_per_second": 1011.762
    },
    {
      "epoch": 0.09941181343716345,
      "grad_norm": 0.6368781924247742,
      "learning_rate": 4.969788171803293e-05,
      "loss": 0.6039,
      "num_input_tokens_seen": 7894504,
      "step": 300,
      "train_runtime": 7814.5594,
      "train_tokens_per_second": 1010.23
    },
    {
      "epoch": 0.1010686769944495,
      "grad_norm": 0.6165225505828857,
      "learning_rate": 4.9687714208011126e-05,
      "loss": 0.503,
      "num_input_tokens_seen": 8028240,
      "step": 305,
      "train_runtime": 7949.434,
      "train_tokens_per_second": 1009.913
    },
    {
      "epoch": 0.10272554055173556,
      "grad_norm": 1.9813154935836792,
      "learning_rate": 4.9677379503511975e-05,
      "loss": 0.4886,
      "num_input_tokens_seen": 8175832,
      "step": 310,
      "train_runtime": 8096.4803,
      "train_tokens_per_second": 1009.801
    },
    {
      "epoch": 0.10438240410902162,
      "grad_norm": 0.6594364643096924,
      "learning_rate": 4.9666877674526e-05,
      "loss": 0.5094,
      "num_input_tokens_seen": 8314656,
      "step": 315,
      "train_runtime": 8230.6926,
      "train_tokens_per_second": 1010.201
    },
    {
      "epoch": 0.10603926766630768,
      "grad_norm": 0.698659360408783,
      "learning_rate": 4.965620879217553e-05,
      "loss": 0.507,
      "num_input_tokens_seen": 8437216,
      "step": 320,
      "train_runtime": 8349.2198,
      "train_tokens_per_second": 1010.539
    },
    {
      "epoch": 0.10769613122359374,
      "grad_norm": 0.5399906039237976,
      "learning_rate": 4.9645372928714246e-05,
      "loss": 0.5436,
      "num_input_tokens_seen": 8559784,
      "step": 325,
      "train_runtime": 8471.4585,
      "train_tokens_per_second": 1010.426
    },
    {
      "epoch": 0.1093529947808798,
      "grad_norm": 0.6274646520614624,
      "learning_rate": 4.963437015752668e-05,
      "loss": 0.4642,
      "num_input_tokens_seen": 8700488,
      "step": 330,
      "train_runtime": 8607.8341,
      "train_tokens_per_second": 1010.764
    },
    {
      "epoch": 0.11100985833816585,
      "grad_norm": 0.4852447509765625,
      "learning_rate": 4.962320055312774e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 8823808,
      "step": 335,
      "train_runtime": 8725.0664,
      "train_tokens_per_second": 1011.317
    },
    {
      "epoch": 0.1126667218954519,
      "grad_norm": 0.631702184677124,
      "learning_rate": 4.961186419116217e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 8944632,
      "step": 340,
      "train_runtime": 8843.1976,
      "train_tokens_per_second": 1011.47
    },
    {
      "epoch": 0.11432358545273796,
      "grad_norm": 0.6279223561286926,
      "learning_rate": 4.960036114840409e-05,
      "loss": 0.5721,
      "num_input_tokens_seen": 9064864,
      "step": 345,
      "train_runtime": 8963.9728,
      "train_tokens_per_second": 1011.255
    },
    {
      "epoch": 0.11598044901002402,
      "grad_norm": 0.61720210313797,
      "learning_rate": 4.958869150275642e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 9197904,
      "step": 350,
      "train_runtime": 9092.9935,
      "train_tokens_per_second": 1011.538
    },
    {
      "epoch": 0.11763731256731008,
      "grad_norm": 0.6143398284912109,
      "learning_rate": 4.957685533325039e-05,
      "loss": 0.5219,
      "num_input_tokens_seen": 9333088,
      "step": 355,
      "train_runtime": 9237.2769,
      "train_tokens_per_second": 1010.372
    },
    {
      "epoch": 0.11929417612459614,
      "grad_norm": 0.7272740006446838,
      "learning_rate": 4.956485272004499e-05,
      "loss": 0.5528,
      "num_input_tokens_seen": 9439152,
      "step": 360,
      "train_runtime": 9339.2326,
      "train_tokens_per_second": 1010.699
    },
    {
      "epoch": 0.1209510396818822,
      "grad_norm": 0.8610154986381531,
      "learning_rate": 4.955268374442641e-05,
      "loss": 0.4877,
      "num_input_tokens_seen": 9571840,
      "step": 365,
      "train_runtime": 9471.7093,
      "train_tokens_per_second": 1010.572
    },
    {
      "epoch": 0.12260790323916826,
      "grad_norm": 0.5370221138000488,
      "learning_rate": 4.954034848880754e-05,
      "loss": 0.5215,
      "num_input_tokens_seen": 9712648,
      "step": 370,
      "train_runtime": 9618.2428,
      "train_tokens_per_second": 1009.815
    },
    {
      "epoch": 0.12426476679645432,
      "grad_norm": 0.5596175193786621,
      "learning_rate": 4.952784703672736e-05,
      "loss": 0.5208,
      "num_input_tokens_seen": 9860400,
      "step": 375,
      "train_runtime": 9768.9327,
      "train_tokens_per_second": 1009.363
    },
    {
      "epoch": 0.12592163035374038,
      "grad_norm": 0.5850361585617065,
      "learning_rate": 4.95151794728504e-05,
      "loss": 0.5133,
      "num_input_tokens_seen": 9983640,
      "step": 380,
      "train_runtime": 9889.0076,
      "train_tokens_per_second": 1009.569
    },
    {
      "epoch": 0.12757849391102644,
      "grad_norm": 0.6175872087478638,
      "learning_rate": 4.950234588296618e-05,
      "loss": 0.4994,
      "num_input_tokens_seen": 10119112,
      "step": 385,
      "train_runtime": 10025.91,
      "train_tokens_per_second": 1009.296
    },
    {
      "epoch": 0.1292353574683125,
      "grad_norm": 0.5896696448326111,
      "learning_rate": 4.9489346353988574e-05,
      "loss": 0.4707,
      "num_input_tokens_seen": 10255320,
      "step": 390,
      "train_runtime": 10160.4137,
      "train_tokens_per_second": 1009.341
    },
    {
      "epoch": 0.13089222102559855,
      "grad_norm": 0.6871618628501892,
      "learning_rate": 4.9476180973955276e-05,
      "loss": 0.4935,
      "num_input_tokens_seen": 10379800,
      "step": 395,
      "train_runtime": 10279.4724,
      "train_tokens_per_second": 1009.76
    },
    {
      "epoch": 0.1325490845828846,
      "grad_norm": 0.592153787612915,
      "learning_rate": 4.94628498320272e-05,
      "loss": 0.5135,
      "num_input_tokens_seen": 10508312,
      "step": 400,
      "train_runtime": 10399.4018,
      "train_tokens_per_second": 1010.473
    },
    {
      "epoch": 0.13420594814017064,
      "grad_norm": 1.5047481060028076,
      "learning_rate": 4.944935301848784e-05,
      "loss": 0.5071,
      "num_input_tokens_seen": 10681712,
      "step": 405,
      "train_runtime": 10574.5353,
      "train_tokens_per_second": 1010.135
    },
    {
      "epoch": 0.1358628116974567,
      "grad_norm": 0.7214350700378418,
      "learning_rate": 4.9435690624742694e-05,
      "loss": 0.5498,
      "num_input_tokens_seen": 10830936,
      "step": 410,
      "train_runtime": 10719.878,
      "train_tokens_per_second": 1010.36
    },
    {
      "epoch": 0.13751967525474276,
      "grad_norm": 0.754260778427124,
      "learning_rate": 4.942186274331862e-05,
      "loss": 0.5425,
      "num_input_tokens_seen": 10938640,
      "step": 415,
      "train_runtime": 10824.8948,
      "train_tokens_per_second": 1010.508
    },
    {
      "epoch": 0.13917653881202882,
      "grad_norm": 0.5813269019126892,
      "learning_rate": 4.940786946786323e-05,
      "loss": 0.5857,
      "num_input_tokens_seen": 11063976,
      "step": 420,
      "train_runtime": 10946.6703,
      "train_tokens_per_second": 1010.716
    },
    {
      "epoch": 0.14083340236931488,
      "grad_norm": 0.679603099822998,
      "learning_rate": 4.9393710893144244e-05,
      "loss": 0.5529,
      "num_input_tokens_seen": 11201280,
      "step": 425,
      "train_runtime": 11080.4366,
      "train_tokens_per_second": 1010.906
    },
    {
      "epoch": 0.14249026592660094,
      "grad_norm": 0.629304826259613,
      "learning_rate": 4.937938711504885e-05,
      "loss": 0.4777,
      "num_input_tokens_seen": 11312184,
      "step": 430,
      "train_runtime": 11184.9978,
      "train_tokens_per_second": 1011.371
    },
    {
      "epoch": 0.144147129483887,
      "grad_norm": 0.9316897392272949,
      "learning_rate": 4.936489823058304e-05,
      "loss": 0.567,
      "num_input_tokens_seen": 11450552,
      "step": 435,
      "train_runtime": 11317.6743,
      "train_tokens_per_second": 1011.741
    },
    {
      "epoch": 0.14580399304117306,
      "grad_norm": 0.7755406498908997,
      "learning_rate": 4.935024433787102e-05,
      "loss": 0.5334,
      "num_input_tokens_seen": 11591568,
      "step": 440,
      "train_runtime": 11448.1534,
      "train_tokens_per_second": 1012.527
    },
    {
      "epoch": 0.14746085659845912,
      "grad_norm": 0.7222505211830139,
      "learning_rate": 4.933542553615443e-05,
      "loss": 0.5625,
      "num_input_tokens_seen": 11732416,
      "step": 445,
      "train_runtime": 11573.0583,
      "train_tokens_per_second": 1013.77
    },
    {
      "epoch": 0.14911772015574518,
      "grad_norm": 0.6651027798652649,
      "learning_rate": 4.932044192579177e-05,
      "loss": 0.5006,
      "num_input_tokens_seen": 11883936,
      "step": 450,
      "train_runtime": 11718.0766,
      "train_tokens_per_second": 1014.154
    },
    {
      "epoch": 0.15077458371303123,
      "grad_norm": 0.7629894018173218,
      "learning_rate": 4.930529360825768e-05,
      "loss": 0.6206,
      "num_input_tokens_seen": 12001800,
      "step": 455,
      "train_runtime": 11842.751,
      "train_tokens_per_second": 1013.43
    },
    {
      "epoch": 0.1524314472703173,
      "grad_norm": 0.8542412519454956,
      "learning_rate": 4.9289980686142254e-05,
      "loss": 0.473,
      "num_input_tokens_seen": 12092984,
      "step": 460,
      "train_runtime": 11930.9084,
      "train_tokens_per_second": 1013.585
    },
    {
      "epoch": 0.15408831082760335,
      "grad_norm": 0.7211665511131287,
      "learning_rate": 4.9274503263150374e-05,
      "loss": 0.4729,
      "num_input_tokens_seen": 12213448,
      "step": 465,
      "train_runtime": 12050.5869,
      "train_tokens_per_second": 1013.515
    },
    {
      "epoch": 0.1557451743848894,
      "grad_norm": 0.717004120349884,
      "learning_rate": 4.9258861444100936e-05,
      "loss": 0.5154,
      "num_input_tokens_seen": 12358784,
      "step": 470,
      "train_runtime": 12198.8825,
      "train_tokens_per_second": 1013.108
    },
    {
      "epoch": 0.15740203794217547,
      "grad_norm": 0.5700079202651978,
      "learning_rate": 4.924305533492623e-05,
      "loss": 0.5066,
      "num_input_tokens_seen": 12468080,
      "step": 475,
      "train_runtime": 12294.9334,
      "train_tokens_per_second": 1014.083
    },
    {
      "epoch": 0.15905890149946153,
      "grad_norm": 0.6611483693122864,
      "learning_rate": 4.9227085042671176e-05,
      "loss": 0.4968,
      "num_input_tokens_seen": 12614272,
      "step": 480,
      "train_runtime": 12449.3903,
      "train_tokens_per_second": 1013.244
    },
    {
      "epoch": 0.1607157650567476,
      "grad_norm": 0.5922725796699524,
      "learning_rate": 4.921095067549258e-05,
      "loss": 0.4779,
      "num_input_tokens_seen": 12753680,
      "step": 485,
      "train_runtime": 12573.0677,
      "train_tokens_per_second": 1014.365
    },
    {
      "epoch": 0.16237262861403362,
      "grad_norm": 0.7947571873664856,
      "learning_rate": 4.9194652342658445e-05,
      "loss": 0.5219,
      "num_input_tokens_seen": 12894896,
      "step": 490,
      "train_runtime": 12702.447,
      "train_tokens_per_second": 1015.151
    },
    {
      "epoch": 0.16402949217131968,
      "grad_norm": 0.6631523370742798,
      "learning_rate": 4.91781901545472e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 12997368,
      "step": 495,
      "train_runtime": 12799.9343,
      "train_tokens_per_second": 1015.425
    },
    {
      "epoch": 0.16568635572860574,
      "grad_norm": 0.6136907339096069,
      "learning_rate": 4.916156422264697e-05,
      "loss": 0.5263,
      "num_input_tokens_seen": 13162752,
      "step": 500,
      "train_runtime": 12966.1956,
      "train_tokens_per_second": 1015.159
    },
    {
      "epoch": 0.1673432192858918,
      "grad_norm": 0.5820109844207764,
      "learning_rate": 4.914477465955483e-05,
      "loss": 0.5058,
      "num_input_tokens_seen": 13288912,
      "step": 505,
      "train_runtime": 13088.7565,
      "train_tokens_per_second": 1015.292
    },
    {
      "epoch": 0.16900008284317786,
      "grad_norm": 0.6264848113059998,
      "learning_rate": 4.9127821578975986e-05,
      "loss": 0.5033,
      "num_input_tokens_seen": 13433264,
      "step": 510,
      "train_runtime": 13228.646,
      "train_tokens_per_second": 1015.468
    },
    {
      "epoch": 0.17065694640046392,
      "grad_norm": 0.7444716691970825,
      "learning_rate": 4.9110705095723076e-05,
      "loss": 0.4713,
      "num_input_tokens_seen": 13556072,
      "step": 515,
      "train_runtime": 13341.2541,
      "train_tokens_per_second": 1016.102
    },
    {
      "epoch": 0.17231380995774997,
      "grad_norm": 0.6482500433921814,
      "learning_rate": 4.909342532571537e-05,
      "loss": 0.5581,
      "num_input_tokens_seen": 13678008,
      "step": 520,
      "train_runtime": 13453.6429,
      "train_tokens_per_second": 1016.677
    },
    {
      "epoch": 0.17397067351503603,
      "grad_norm": 0.7409155964851379,
      "learning_rate": 4.9075982385977945e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 13785456,
      "step": 525,
      "train_runtime": 13549.909,
      "train_tokens_per_second": 1017.384
    },
    {
      "epoch": 0.1756275370723221,
      "grad_norm": 0.640856921672821,
      "learning_rate": 4.905837639464096e-05,
      "loss": 0.5081,
      "num_input_tokens_seen": 13950768,
      "step": 530,
      "train_runtime": 13725.5278,
      "train_tokens_per_second": 1016.41
    },
    {
      "epoch": 0.17728440062960815,
      "grad_norm": 0.5745977759361267,
      "learning_rate": 4.90406074709388e-05,
      "loss": 0.4516,
      "num_input_tokens_seen": 14096496,
      "step": 535,
      "train_runtime": 13873.1182,
      "train_tokens_per_second": 1016.101
    },
    {
      "epoch": 0.1789412641868942,
      "grad_norm": 0.6930714249610901,
      "learning_rate": 4.902267573520928e-05,
      "loss": 0.5106,
      "num_input_tokens_seen": 14200392,
      "step": 540,
      "train_runtime": 13976.5077,
      "train_tokens_per_second": 1016.019
    },
    {
      "epoch": 0.18059812774418027,
      "grad_norm": 0.5499430298805237,
      "learning_rate": 4.9004581308892875e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 14341928,
      "step": 545,
      "train_runtime": 14101.8349,
      "train_tokens_per_second": 1017.026
    },
    {
      "epoch": 0.18225499130146633,
      "grad_norm": 0.59327632188797,
      "learning_rate": 4.8986324314531815e-05,
      "loss": 0.4373,
      "num_input_tokens_seen": 14481648,
      "step": 550,
      "train_runtime": 14243.8972,
      "train_tokens_per_second": 1016.691
    },
    {
      "epoch": 0.1839118548587524,
      "grad_norm": 0.8996002078056335,
      "learning_rate": 4.8967904875769347e-05,
      "loss": 0.508,
      "num_input_tokens_seen": 14587488,
      "step": 555,
      "train_runtime": 14340.867,
      "train_tokens_per_second": 1017.197
    },
    {
      "epoch": 0.18556871841603845,
      "grad_norm": 0.7360817790031433,
      "learning_rate": 4.894932311734881e-05,
      "loss": 0.5258,
      "num_input_tokens_seen": 14729456,
      "step": 560,
      "train_runtime": 14480.1902,
      "train_tokens_per_second": 1017.214
    },
    {
      "epoch": 0.1872255819733245,
      "grad_norm": 0.638360321521759,
      "learning_rate": 4.893057916511287e-05,
      "loss": 0.5182,
      "num_input_tokens_seen": 14862528,
      "step": 565,
      "train_runtime": 14616.9642,
      "train_tokens_per_second": 1016.8
    },
    {
      "epoch": 0.18888244553061057,
      "grad_norm": 0.5993677973747253,
      "learning_rate": 4.891167314600261e-05,
      "loss": 0.5219,
      "num_input_tokens_seen": 14980776,
      "step": 570,
      "train_runtime": 14732.7936,
      "train_tokens_per_second": 1016.832
    },
    {
      "epoch": 0.19053930908789662,
      "grad_norm": 0.7454946637153625,
      "learning_rate": 4.8892605188056704e-05,
      "loss": 0.4943,
      "num_input_tokens_seen": 15098552,
      "step": 575,
      "train_runtime": 14852.4236,
      "train_tokens_per_second": 1016.572
    },
    {
      "epoch": 0.19219617264518266,
      "grad_norm": 0.7516965270042419,
      "learning_rate": 4.8873375420410516e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 15218872,
      "step": 580,
      "train_runtime": 14961.0224,
      "train_tokens_per_second": 1017.235
    },
    {
      "epoch": 0.19385303620246871,
      "grad_norm": 0.5247603058815002,
      "learning_rate": 4.885398397329527e-05,
      "loss": 0.5031,
      "num_input_tokens_seen": 15374096,
      "step": 585,
      "train_runtime": 15115.8819,
      "train_tokens_per_second": 1017.082
    },
    {
      "epoch": 0.19550989975975477,
      "grad_norm": 0.6330128312110901,
      "learning_rate": 4.883443097803712e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 15522768,
      "step": 590,
      "train_runtime": 15272.0057,
      "train_tokens_per_second": 1016.42
    },
    {
      "epoch": 0.19716676331704083,
      "grad_norm": 0.6485143899917603,
      "learning_rate": 4.8814716567056306e-05,
      "loss": 0.4536,
      "num_input_tokens_seen": 15650968,
      "step": 595,
      "train_runtime": 15395.8012,
      "train_tokens_per_second": 1016.574
    },
    {
      "epoch": 0.1988236268743269,
      "grad_norm": 0.6386798620223999,
      "learning_rate": 4.879484087386622e-05,
      "loss": 0.4727,
      "num_input_tokens_seen": 15775608,
      "step": 600,
      "train_runtime": 15522.3137,
      "train_tokens_per_second": 1016.318
    },
    {
      "epoch": 0.20048049043161295,
      "grad_norm": 0.5361050367355347,
      "learning_rate": 4.877480403307253e-05,
      "loss": 0.4582,
      "num_input_tokens_seen": 15908760,
      "step": 605,
      "train_runtime": 15654.4043,
      "train_tokens_per_second": 1016.248
    },
    {
      "epoch": 0.202137353988899,
      "grad_norm": 0.5473469495773315,
      "learning_rate": 4.8754606180372256e-05,
      "loss": 0.4749,
      "num_input_tokens_seen": 16057264,
      "step": 610,
      "train_runtime": 15801.4891,
      "train_tokens_per_second": 1016.187
    },
    {
      "epoch": 0.20379421754618507,
      "grad_norm": 0.5276517271995544,
      "learning_rate": 4.873424745255283e-05,
      "loss": 0.4481,
      "num_input_tokens_seen": 16180472,
      "step": 615,
      "train_runtime": 15921.2051,
      "train_tokens_per_second": 1016.284
    },
    {
      "epoch": 0.20545108110347113,
      "grad_norm": 1.151261806488037,
      "learning_rate": 4.8713727987491225e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 16286928,
      "step": 620,
      "train_runtime": 16029.2925,
      "train_tokens_per_second": 1016.073
    },
    {
      "epoch": 0.2071079446607572,
      "grad_norm": 0.6508522629737854,
      "learning_rate": 4.869304792415296e-05,
      "loss": 0.4936,
      "num_input_tokens_seen": 16428448,
      "step": 625,
      "train_runtime": 16171.9755,
      "train_tokens_per_second": 1015.859
    },
    {
      "epoch": 0.20876480821804325,
      "grad_norm": 0.6947662830352783,
      "learning_rate": 4.86722074025912e-05,
      "loss": 0.4881,
      "num_input_tokens_seen": 16542896,
      "step": 630,
      "train_runtime": 16284.8466,
      "train_tokens_per_second": 1015.846
    },
    {
      "epoch": 0.2104216717753293,
      "grad_norm": 0.5938307046890259,
      "learning_rate": 4.865120656394577e-05,
      "loss": 0.5152,
      "num_input_tokens_seen": 16681864,
      "step": 635,
      "train_runtime": 16421.7251,
      "train_tokens_per_second": 1015.841
    },
    {
      "epoch": 0.21207853533261536,
      "grad_norm": 0.7247110605239868,
      "learning_rate": 4.8630045550442264e-05,
      "loss": 0.5496,
      "num_input_tokens_seen": 16823992,
      "step": 640,
      "train_runtime": 16567.1309,
      "train_tokens_per_second": 1015.504
    },
    {
      "epoch": 0.21373539888990142,
      "grad_norm": 0.7758472561836243,
      "learning_rate": 4.8608724505391004e-05,
      "loss": 0.5368,
      "num_input_tokens_seen": 16951432,
      "step": 645,
      "train_runtime": 16693.7974,
      "train_tokens_per_second": 1015.433
    },
    {
      "epoch": 0.21539226244718748,
      "grad_norm": 0.622420608997345,
      "learning_rate": 4.858724357318612e-05,
      "loss": 0.4739,
      "num_input_tokens_seen": 17076416,
      "step": 650,
      "train_runtime": 16805.9911,
      "train_tokens_per_second": 1016.091
    },
    {
      "epoch": 0.21704912600447354,
      "grad_norm": 0.6357735991477966,
      "learning_rate": 4.856560289930456e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 17181888,
      "step": 655,
      "train_runtime": 16900.5584,
      "train_tokens_per_second": 1016.646
    },
    {
      "epoch": 0.2187059895617596,
      "grad_norm": 0.7810402512550354,
      "learning_rate": 4.854380263030509e-05,
      "loss": 0.477,
      "num_input_tokens_seen": 17311240,
      "step": 660,
      "train_runtime": 17010.1629,
      "train_tokens_per_second": 1017.7
    },
    {
      "epoch": 0.22036285311904566,
      "grad_norm": 0.8222882151603699,
      "learning_rate": 4.8521842913827345e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 17456904,
      "step": 665,
      "train_runtime": 17157.1499,
      "train_tokens_per_second": 1017.471
    },
    {
      "epoch": 0.2220197166763317,
      "grad_norm": 1.101054072380066,
      "learning_rate": 4.8499723898590754e-05,
      "loss": 0.4673,
      "num_input_tokens_seen": 17561448,
      "step": 670,
      "train_runtime": 17257.9631,
      "train_tokens_per_second": 1017.585
    },
    {
      "epoch": 0.22367658023361775,
      "grad_norm": 0.8191339373588562,
      "learning_rate": 4.847744573439361e-05,
      "loss": 0.4345,
      "num_input_tokens_seen": 17695224,
      "step": 675,
      "train_runtime": 17399.2967,
      "train_tokens_per_second": 1017.008
    },
    {
      "epoch": 0.2253334437909038,
      "grad_norm": 0.6663082242012024,
      "learning_rate": 4.845500857211202e-05,
      "loss": 0.5216,
      "num_input_tokens_seen": 17817200,
      "step": 680,
      "train_runtime": 17514.0095,
      "train_tokens_per_second": 1017.311
    },
    {
      "epoch": 0.22699030734818987,
      "grad_norm": 0.645950436592102,
      "learning_rate": 4.843241256369886e-05,
      "loss": 0.5036,
      "num_input_tokens_seen": 17951952,
      "step": 685,
      "train_runtime": 17652.4248,
      "train_tokens_per_second": 1016.968
    },
    {
      "epoch": 0.22864717090547593,
      "grad_norm": 0.7596031427383423,
      "learning_rate": 4.84096578621828e-05,
      "loss": 0.4953,
      "num_input_tokens_seen": 18076992,
      "step": 690,
      "train_runtime": 17778.7907,
      "train_tokens_per_second": 1016.773
    },
    {
      "epoch": 0.230304034462762,
      "grad_norm": 0.7966089248657227,
      "learning_rate": 4.8386744621667223e-05,
      "loss": 0.4766,
      "num_input_tokens_seen": 18182976,
      "step": 695,
      "train_runtime": 17870.4232,
      "train_tokens_per_second": 1017.49
    },
    {
      "epoch": 0.23196089802004805,
      "grad_norm": 0.7758664488792419,
      "learning_rate": 4.8363672997329216e-05,
      "loss": 0.5377,
      "num_input_tokens_seen": 18346096,
      "step": 700,
      "train_runtime": 18029.399,
      "train_tokens_per_second": 1017.566
    },
    {
      "epoch": 0.2336177615773341,
      "grad_norm": 0.5894461870193481,
      "learning_rate": 4.834044314541847e-05,
      "loss": 0.4915,
      "num_input_tokens_seen": 18473144,
      "step": 705,
      "train_runtime": 18157.6438,
      "train_tokens_per_second": 1017.376
    },
    {
      "epoch": 0.23527462513462016,
      "grad_norm": 0.7481614351272583,
      "learning_rate": 4.831705522325629e-05,
      "loss": 0.4704,
      "num_input_tokens_seen": 18613272,
      "step": 710,
      "train_runtime": 18298.4103,
      "train_tokens_per_second": 1017.207
    },
    {
      "epoch": 0.23693148869190622,
      "grad_norm": 1.054065465927124,
      "learning_rate": 4.829350938923446e-05,
      "loss": 0.4751,
      "num_input_tokens_seen": 18779912,
      "step": 715,
      "train_runtime": 18475.478,
      "train_tokens_per_second": 1016.478
    },
    {
      "epoch": 0.23858835224919228,
      "grad_norm": 0.5754175782203674,
      "learning_rate": 4.8269805802814226e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 18920312,
      "step": 720,
      "train_runtime": 18616.5274,
      "train_tokens_per_second": 1016.318
    },
    {
      "epoch": 0.24024521580647834,
      "grad_norm": 0.6922730803489685,
      "learning_rate": 4.8245944624525177e-05,
      "loss": 0.4977,
      "num_input_tokens_seen": 19048384,
      "step": 725,
      "train_runtime": 18742.9111,
      "train_tokens_per_second": 1016.298
    },
    {
      "epoch": 0.2419020793637644,
      "grad_norm": 0.5471542477607727,
      "learning_rate": 4.822192601596417e-05,
      "loss": 0.5602,
      "num_input_tokens_seen": 19198856,
      "step": 730,
      "train_runtime": 18898.2627,
      "train_tokens_per_second": 1015.906
    },
    {
      "epoch": 0.24355894292105046,
      "grad_norm": 0.6527897119522095,
      "learning_rate": 4.819775013979426e-05,
      "loss": 0.499,
      "num_input_tokens_seen": 19328632,
      "step": 735,
      "train_runtime": 19025.2832,
      "train_tokens_per_second": 1015.945
    },
    {
      "epoch": 0.24521580647833652,
      "grad_norm": 0.7256128787994385,
      "learning_rate": 4.817341715974356e-05,
      "loss": 0.4986,
      "num_input_tokens_seen": 19441640,
      "step": 740,
      "train_runtime": 19134.4385,
      "train_tokens_per_second": 1016.055
    },
    {
      "epoch": 0.24687267003562258,
      "grad_norm": 0.8111450672149658,
      "learning_rate": 4.814892724060416e-05,
      "loss": 0.4362,
      "num_input_tokens_seen": 19593872,
      "step": 745,
      "train_runtime": 19279.7378,
      "train_tokens_per_second": 1016.293
    },
    {
      "epoch": 0.24852953359290864,
      "grad_norm": 0.6983108520507812,
      "learning_rate": 4.8124280548230984e-05,
      "loss": 0.5314,
      "num_input_tokens_seen": 19719584,
      "step": 750,
      "train_runtime": 19403.9683,
      "train_tokens_per_second": 1016.266
    },
    {
      "epoch": 0.2501863971501947,
      "grad_norm": 0.7815086841583252,
      "learning_rate": 4.80994772495407e-05,
      "loss": 0.496,
      "num_input_tokens_seen": 19839272,
      "step": 755,
      "train_runtime": 19517.6395,
      "train_tokens_per_second": 1016.479
    },
    {
      "epoch": 0.25184326070748075,
      "grad_norm": 0.7361581325531006,
      "learning_rate": 4.807451751251056e-05,
      "loss": 0.4449,
      "num_input_tokens_seen": 19986048,
      "step": 760,
      "train_runtime": 19669.2928,
      "train_tokens_per_second": 1016.104
    },
    {
      "epoch": 0.2535001242647668,
      "grad_norm": 0.7071053981781006,
      "learning_rate": 4.80494015061773e-05,
      "loss": 0.5582,
      "num_input_tokens_seen": 20081312,
      "step": 765,
      "train_runtime": 19766.2573,
      "train_tokens_per_second": 1015.939
    },
    {
      "epoch": 0.25515698782205287,
      "grad_norm": 0.8530505895614624,
      "learning_rate": 4.802412940063594e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 20209320,
      "step": 770,
      "train_runtime": 19893.9064,
      "train_tokens_per_second": 1015.855
    },
    {
      "epoch": 0.25681385137933893,
      "grad_norm": 0.7236202359199524,
      "learning_rate": 4.799870136703867e-05,
      "loss": 0.525,
      "num_input_tokens_seen": 20365272,
      "step": 775,
      "train_runtime": 20047.4398,
      "train_tokens_per_second": 1015.854
    },
    {
      "epoch": 0.258470714936625,
      "grad_norm": 0.6288323402404785,
      "learning_rate": 4.7973117577593694e-05,
      "loss": 0.4706,
      "num_input_tokens_seen": 20513368,
      "step": 780,
      "train_runtime": 20192.416,
      "train_tokens_per_second": 1015.895
    },
    {
      "epoch": 0.26012757849391105,
      "grad_norm": 0.6292834281921387,
      "learning_rate": 4.7947378205564055e-05,
      "loss": 0.4763,
      "num_input_tokens_seen": 20629096,
      "step": 785,
      "train_runtime": 20303.0947,
      "train_tokens_per_second": 1016.057
    },
    {
      "epoch": 0.2617844420511971,
      "grad_norm": 0.863003134727478,
      "learning_rate": 4.792148342526644e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 20726104,
      "step": 790,
      "train_runtime": 20390.2873,
      "train_tokens_per_second": 1016.469
    },
    {
      "epoch": 0.26344130560848317,
      "grad_norm": 0.6601041555404663,
      "learning_rate": 4.7895433412070035e-05,
      "loss": 0.4842,
      "num_input_tokens_seen": 20869504,
      "step": 795,
      "train_runtime": 20533.775,
      "train_tokens_per_second": 1016.35
    },
    {
      "epoch": 0.2650981691657692,
      "grad_norm": 0.6043739318847656,
      "learning_rate": 4.7869228342395324e-05,
      "loss": 0.3958,
      "num_input_tokens_seen": 20984776,
      "step": 800,
      "train_runtime": 20646.2729,
      "train_tokens_per_second": 1016.395
    },
    {
      "epoch": 0.26675503272305523,
      "grad_norm": 0.7576766610145569,
      "learning_rate": 4.784286839371288e-05,
      "loss": 0.532,
      "num_input_tokens_seen": 21132704,
      "step": 805,
      "train_runtime": 20797.5354,
      "train_tokens_per_second": 1016.116
    },
    {
      "epoch": 0.2684118962803413,
      "grad_norm": 0.4533199369907379,
      "learning_rate": 4.781635374454219e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 21273584,
      "step": 810,
      "train_runtime": 20944.7541,
      "train_tokens_per_second": 1015.7
    },
    {
      "epoch": 0.27006875983762735,
      "grad_norm": 0.8381832838058472,
      "learning_rate": 4.7789684574450396e-05,
      "loss": 0.451,
      "num_input_tokens_seen": 21395240,
      "step": 815,
      "train_runtime": 21058.9961,
      "train_tokens_per_second": 1015.967
    },
    {
      "epoch": 0.2717256233949134,
      "grad_norm": 0.6072661876678467,
      "learning_rate": 4.776286106405117e-05,
      "loss": 0.4731,
      "num_input_tokens_seen": 21527376,
      "step": 820,
      "train_runtime": 21184.9796,
      "train_tokens_per_second": 1016.162
    },
    {
      "epoch": 0.27338248695219947,
      "grad_norm": 0.7162862420082092,
      "learning_rate": 4.773588339500339e-05,
      "loss": 0.4841,
      "num_input_tokens_seen": 21654400,
      "step": 825,
      "train_runtime": 21299.744,
      "train_tokens_per_second": 1016.651
    },
    {
      "epoch": 0.2750393505094855,
      "grad_norm": 0.5779826045036316,
      "learning_rate": 4.7708751750009964e-05,
      "loss": 0.505,
      "num_input_tokens_seen": 21770072,
      "step": 830,
      "train_runtime": 21398.7078,
      "train_tokens_per_second": 1017.355
    },
    {
      "epoch": 0.2766962140667716,
      "grad_norm": 0.8403722047805786,
      "learning_rate": 4.7681466312816594e-05,
      "loss": 0.5268,
      "num_input_tokens_seen": 21891520,
      "step": 835,
      "train_runtime": 21525.1163,
      "train_tokens_per_second": 1017.022
    },
    {
      "epoch": 0.27835307762405764,
      "grad_norm": 0.975309431552887,
      "learning_rate": 4.765402726821051e-05,
      "loss": 0.4872,
      "num_input_tokens_seen": 22047216,
      "step": 840,
      "train_runtime": 21668.9049,
      "train_tokens_per_second": 1017.459
    },
    {
      "epoch": 0.2800099411813437,
      "grad_norm": 0.49896493554115295,
      "learning_rate": 4.762643480201923e-05,
      "loss": 0.4418,
      "num_input_tokens_seen": 22186472,
      "step": 845,
      "train_runtime": 21802.9384,
      "train_tokens_per_second": 1017.591
    },
    {
      "epoch": 0.28166680473862976,
      "grad_norm": 0.8691105246543884,
      "learning_rate": 4.759868910110931e-05,
      "loss": 0.4832,
      "num_input_tokens_seen": 22330160,
      "step": 850,
      "train_runtime": 21942.6906,
      "train_tokens_per_second": 1017.658
    },
    {
      "epoch": 0.2833236682959158,
      "grad_norm": 0.5267112255096436,
      "learning_rate": 4.757079035338506e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 22456120,
      "step": 855,
      "train_runtime": 22059.1573,
      "train_tokens_per_second": 1017.995
    },
    {
      "epoch": 0.2849805318532019,
      "grad_norm": 0.7115445733070374,
      "learning_rate": 4.7542738747787275e-05,
      "loss": 0.4443,
      "num_input_tokens_seen": 22564112,
      "step": 860,
      "train_runtime": 22157.264,
      "train_tokens_per_second": 1018.362
    },
    {
      "epoch": 0.28663739541048794,
      "grad_norm": 0.8554672598838806,
      "learning_rate": 4.7514534474291986e-05,
      "loss": 0.46,
      "num_input_tokens_seen": 22680328,
      "step": 865,
      "train_runtime": 22274.2279,
      "train_tokens_per_second": 1018.232
    },
    {
      "epoch": 0.288294258967774,
      "grad_norm": 0.5774676203727722,
      "learning_rate": 4.7486177723909115e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 22806000,
      "step": 870,
      "train_runtime": 22390.6093,
      "train_tokens_per_second": 1018.552
    },
    {
      "epoch": 0.28995112252506006,
      "grad_norm": 0.7077370882034302,
      "learning_rate": 4.745766868868123e-05,
      "loss": 0.525,
      "num_input_tokens_seen": 22929440,
      "step": 875,
      "train_runtime": 22517.1234,
      "train_tokens_per_second": 1018.311
    },
    {
      "epoch": 0.2916079860823461,
      "grad_norm": 0.7406365275382996,
      "learning_rate": 4.742900756168225e-05,
      "loss": 0.4979,
      "num_input_tokens_seen": 23053296,
      "step": 880,
      "train_runtime": 22630.4269,
      "train_tokens_per_second": 1018.686
    },
    {
      "epoch": 0.2932648496396322,
      "grad_norm": 0.9088826775550842,
      "learning_rate": 4.740019453701608e-05,
      "loss": 0.4272,
      "num_input_tokens_seen": 23188032,
      "step": 885,
      "train_runtime": 22762.5568,
      "train_tokens_per_second": 1018.692
    },
    {
      "epoch": 0.29492171319691823,
      "grad_norm": 0.6111741065979004,
      "learning_rate": 4.7371229809815354e-05,
      "loss": 0.4327,
      "num_input_tokens_seen": 23317104,
      "step": 890,
      "train_runtime": 22882.0829,
      "train_tokens_per_second": 1019.011
    },
    {
      "epoch": 0.2965785767542043,
      "grad_norm": 0.5999329686164856,
      "learning_rate": 4.734211357624009e-05,
      "loss": 0.4286,
      "num_input_tokens_seen": 23454328,
      "step": 895,
      "train_runtime": 23010.0356,
      "train_tokens_per_second": 1019.309
    },
    {
      "epoch": 0.29823544031149035,
      "grad_norm": 0.6912147402763367,
      "learning_rate": 4.731284603347635e-05,
      "loss": 0.4811,
      "num_input_tokens_seen": 23580128,
      "step": 900,
      "train_runtime": 23129.2048,
      "train_tokens_per_second": 1019.496
    },
    {
      "epoch": 0.2998923038687764,
      "grad_norm": 0.9737749099731445,
      "learning_rate": 4.7283427379734956e-05,
      "loss": 0.5664,
      "num_input_tokens_seen": 23711216,
      "step": 905,
      "train_runtime": 23261.6946,
      "train_tokens_per_second": 1019.325
    },
    {
      "epoch": 0.30154916742606247,
      "grad_norm": 0.8089159727096558,
      "learning_rate": 4.725385781425006e-05,
      "loss": 0.487,
      "num_input_tokens_seen": 23869968,
      "step": 910,
      "train_runtime": 23436.9702,
      "train_tokens_per_second": 1018.475
    },
    {
      "epoch": 0.30320603098334853,
      "grad_norm": 0.6630802750587463,
      "learning_rate": 4.722413753727789e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 24024800,
      "step": 915,
      "train_runtime": 23597.8535,
      "train_tokens_per_second": 1018.093
    },
    {
      "epoch": 0.3048628945406346,
      "grad_norm": 0.7434173822402954,
      "learning_rate": 4.7194266750095315e-05,
      "loss": 0.5344,
      "num_input_tokens_seen": 24159920,
      "step": 920,
      "train_runtime": 23722.4187,
      "train_tokens_per_second": 1018.443
    },
    {
      "epoch": 0.30651975809792065,
      "grad_norm": 0.6653624773025513,
      "learning_rate": 4.7164245654998545e-05,
      "loss": 0.4799,
      "num_input_tokens_seen": 24294144,
      "step": 925,
      "train_runtime": 23844.6795,
      "train_tokens_per_second": 1018.85
    },
    {
      "epoch": 0.3081766216552067,
      "grad_norm": 0.6320964097976685,
      "learning_rate": 4.7134074455301706e-05,
      "loss": 0.4712,
      "num_input_tokens_seen": 24435744,
      "step": 930,
      "train_runtime": 23998.4,
      "train_tokens_per_second": 1018.224
    },
    {
      "epoch": 0.30983348521249277,
      "grad_norm": 0.5859976410865784,
      "learning_rate": 4.71037533553355e-05,
      "loss": 0.5068,
      "num_input_tokens_seen": 24538392,
      "step": 935,
      "train_runtime": 24090.4359,
      "train_tokens_per_second": 1018.595
    },
    {
      "epoch": 0.3114903487697788,
      "grad_norm": 1.0807486772537231,
      "learning_rate": 4.7073282560445814e-05,
      "loss": 0.5675,
      "num_input_tokens_seen": 24693424,
      "step": 940,
      "train_runtime": 24232.1804,
      "train_tokens_per_second": 1019.034
    },
    {
      "epoch": 0.3131472123270649,
      "grad_norm": 0.6204106211662292,
      "learning_rate": 4.7042662276992315e-05,
      "loss": 0.4526,
      "num_input_tokens_seen": 24806920,
      "step": 945,
      "train_runtime": 24335.9202,
      "train_tokens_per_second": 1019.354
    },
    {
      "epoch": 0.31480407588435094,
      "grad_norm": 0.778038501739502,
      "learning_rate": 4.701189271234707e-05,
      "loss": 0.5055,
      "num_input_tokens_seen": 24942160,
      "step": 950,
      "train_runtime": 24469.8571,
      "train_tokens_per_second": 1019.301
    },
    {
      "epoch": 0.316460939441637,
      "grad_norm": 0.6957627534866333,
      "learning_rate": 4.698097407489315e-05,
      "loss": 0.4946,
      "num_input_tokens_seen": 25090208,
      "step": 955,
      "train_runtime": 24616.0094,
      "train_tokens_per_second": 1019.264
    },
    {
      "epoch": 0.31811780299892306,
      "grad_norm": 0.8292206525802612,
      "learning_rate": 4.694990657402315e-05,
      "loss": 0.4686,
      "num_input_tokens_seen": 25218944,
      "step": 960,
      "train_runtime": 24741.7225,
      "train_tokens_per_second": 1019.288
    },
    {
      "epoch": 0.3197746665562091,
      "grad_norm": 0.703926682472229,
      "learning_rate": 4.691869042013789e-05,
      "loss": 0.4905,
      "num_input_tokens_seen": 25346280,
      "step": 965,
      "train_runtime": 24878.1043,
      "train_tokens_per_second": 1018.819
    },
    {
      "epoch": 0.3214315301134952,
      "grad_norm": 0.7936163544654846,
      "learning_rate": 4.6887325824644886e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 25503528,
      "step": 970,
      "train_runtime": 25039.8774,
      "train_tokens_per_second": 1018.516
    },
    {
      "epoch": 0.32308839367078124,
      "grad_norm": 0.792598307132721,
      "learning_rate": 4.685581299995696e-05,
      "loss": 0.4104,
      "num_input_tokens_seen": 25618096,
      "step": 975,
      "train_runtime": 25148.1375,
      "train_tokens_per_second": 1018.688
    },
    {
      "epoch": 0.32474525722806724,
      "grad_norm": 0.5693343877792358,
      "learning_rate": 4.682415215949081e-05,
      "loss": 0.4274,
      "num_input_tokens_seen": 25754144,
      "step": 980,
      "train_runtime": 25280.5939,
      "train_tokens_per_second": 1018.732
    },
    {
      "epoch": 0.3264021207853533,
      "grad_norm": 0.5096243619918823,
      "learning_rate": 4.6792343517665534e-05,
      "loss": 0.5274,
      "num_input_tokens_seen": 25890040,
      "step": 985,
      "train_runtime": 25408.2832,
      "train_tokens_per_second": 1018.961
    },
    {
      "epoch": 0.32805898434263936,
      "grad_norm": 0.590783953666687,
      "learning_rate": 4.6760387289901205e-05,
      "loss": 0.493,
      "num_input_tokens_seen": 26028320,
      "step": 990,
      "train_runtime": 25539.628,
      "train_tokens_per_second": 1019.135
    },
    {
      "epoch": 0.3297158478999254,
      "grad_norm": 0.677506685256958,
      "learning_rate": 4.672828369261742e-05,
      "loss": 0.5047,
      "num_input_tokens_seen": 26156424,
      "step": 995,
      "train_runtime": 25664.1967,
      "train_tokens_per_second": 1019.18
    },
    {
      "epoch": 0.3313727114572115,
      "grad_norm": 0.5288072824478149,
      "learning_rate": 4.66960329432318e-05,
      "loss": 0.4638,
      "num_input_tokens_seen": 26274640,
      "step": 1000,
      "train_runtime": 25779.7003,
      "train_tokens_per_second": 1019.199
    },
    {
      "epoch": 0.33302957501449754,
      "grad_norm": 0.8450011610984802,
      "learning_rate": 4.666363526015854e-05,
      "loss": 0.436,
      "num_input_tokens_seen": 26401464,
      "step": 1005,
      "train_runtime": 25910.4284,
      "train_tokens_per_second": 1018.951
    },
    {
      "epoch": 0.3346864385717836,
      "grad_norm": 0.6777960658073425,
      "learning_rate": 4.663109086280691e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 26538432,
      "step": 1010,
      "train_runtime": 26055.2105,
      "train_tokens_per_second": 1018.546
    },
    {
      "epoch": 0.33634330212906965,
      "grad_norm": 0.6541361808776855,
      "learning_rate": 4.6598399971579806e-05,
      "loss": 0.4257,
      "num_input_tokens_seen": 26666400,
      "step": 1015,
      "train_runtime": 26182.4508,
      "train_tokens_per_second": 1018.484
    },
    {
      "epoch": 0.3380001656863557,
      "grad_norm": 0.7865561246871948,
      "learning_rate": 4.6565562807872225e-05,
      "loss": 0.4854,
      "num_input_tokens_seen": 26807848,
      "step": 1020,
      "train_runtime": 26333.8028,
      "train_tokens_per_second": 1018.001
    },
    {
      "epoch": 0.3396570292436418,
      "grad_norm": 0.8189525604248047,
      "learning_rate": 4.653257959406978e-05,
      "loss": 0.4913,
      "num_input_tokens_seen": 26964048,
      "step": 1025,
      "train_runtime": 26493.7758,
      "train_tokens_per_second": 1017.75
    },
    {
      "epoch": 0.34131389280092783,
      "grad_norm": 0.7746168375015259,
      "learning_rate": 4.64994505535472e-05,
      "loss": 0.545,
      "num_input_tokens_seen": 27090552,
      "step": 1030,
      "train_runtime": 26609.8765,
      "train_tokens_per_second": 1018.064
    },
    {
      "epoch": 0.3429707563582139,
      "grad_norm": 0.5827233791351318,
      "learning_rate": 4.6466175910666776e-05,
      "loss": 0.4826,
      "num_input_tokens_seen": 27217864,
      "step": 1035,
      "train_runtime": 26738.1578,
      "train_tokens_per_second": 1017.941
    },
    {
      "epoch": 0.34462761991549995,
      "grad_norm": 0.6815202832221985,
      "learning_rate": 4.6432755890776904e-05,
      "loss": 0.4878,
      "num_input_tokens_seen": 27364584,
      "step": 1040,
      "train_runtime": 26871.7981,
      "train_tokens_per_second": 1018.338
    },
    {
      "epoch": 0.346284483472786,
      "grad_norm": 0.6521704196929932,
      "learning_rate": 4.639919072021052e-05,
      "loss": 0.4545,
      "num_input_tokens_seen": 27496112,
      "step": 1045,
      "train_runtime": 27002.1734,
      "train_tokens_per_second": 1018.293
    },
    {
      "epoch": 0.34794134703007207,
      "grad_norm": 0.7422869205474854,
      "learning_rate": 4.636548062628355e-05,
      "loss": 0.4901,
      "num_input_tokens_seen": 27619776,
      "step": 1050,
      "train_runtime": 27125.6179,
      "train_tokens_per_second": 1018.217
    },
    {
      "epoch": 0.3495982105873581,
      "grad_norm": 0.7333402037620544,
      "learning_rate": 4.6331625837293455e-05,
      "loss": 0.5289,
      "num_input_tokens_seen": 27745288,
      "step": 1055,
      "train_runtime": 27247.781,
      "train_tokens_per_second": 1018.259
    },
    {
      "epoch": 0.3512550741446442,
      "grad_norm": 0.7111274600028992,
      "learning_rate": 4.6297626582517555e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 27881320,
      "step": 1060,
      "train_runtime": 27372.1731,
      "train_tokens_per_second": 1018.601
    },
    {
      "epoch": 0.35291193770193025,
      "grad_norm": 0.6821667551994324,
      "learning_rate": 4.626348309221159e-05,
      "loss": 0.4308,
      "num_input_tokens_seen": 28009016,
      "step": 1065,
      "train_runtime": 27501.3437,
      "train_tokens_per_second": 1018.46
    },
    {
      "epoch": 0.3545688012592163,
      "grad_norm": 0.7939398884773254,
      "learning_rate": 4.622919559760811e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 28146384,
      "step": 1070,
      "train_runtime": 27643.409,
      "train_tokens_per_second": 1018.195
    },
    {
      "epoch": 0.35622566481650236,
      "grad_norm": 0.774495542049408,
      "learning_rate": 4.619476433091492e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 28301104,
      "step": 1075,
      "train_runtime": 27796.8413,
      "train_tokens_per_second": 1018.141
    },
    {
      "epoch": 0.3578825283737884,
      "grad_norm": 0.6645853519439697,
      "learning_rate": 4.616018952531347e-05,
      "loss": 0.4566,
      "num_input_tokens_seen": 28418080,
      "step": 1080,
      "train_runtime": 27905.0563,
      "train_tokens_per_second": 1018.385
    },
    {
      "epoch": 0.3595393919310745,
      "grad_norm": 0.5797256231307983,
      "learning_rate": 4.612547141495735e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 28532104,
      "step": 1085,
      "train_runtime": 28007.626,
      "train_tokens_per_second": 1018.726
    },
    {
      "epoch": 0.36119625548836054,
      "grad_norm": 0.6727553606033325,
      "learning_rate": 4.6090610234970656e-05,
      "loss": 0.4407,
      "num_input_tokens_seen": 28618904,
      "step": 1090,
      "train_runtime": 28083.1218,
      "train_tokens_per_second": 1019.078
    },
    {
      "epoch": 0.3628531190456466,
      "grad_norm": 0.6134670972824097,
      "learning_rate": 4.6055606221446386e-05,
      "loss": 0.4861,
      "num_input_tokens_seen": 28741384,
      "step": 1095,
      "train_runtime": 28203.9934,
      "train_tokens_per_second": 1019.054
    },
    {
      "epoch": 0.36450998260293266,
      "grad_norm": 0.6938008069992065,
      "learning_rate": 4.602045961144488e-05,
      "loss": 0.4299,
      "num_input_tokens_seen": 28864992,
      "step": 1100,
      "train_runtime": 28330.6643,
      "train_tokens_per_second": 1018.86
    },
    {
      "epoch": 0.3661668461602187,
      "grad_norm": 0.7549251914024353,
      "learning_rate": 4.598517064299219e-05,
      "loss": 0.545,
      "num_input_tokens_seen": 28993344,
      "step": 1105,
      "train_runtime": 28467.3439,
      "train_tokens_per_second": 1018.477
    },
    {
      "epoch": 0.3678237097175048,
      "grad_norm": 0.8208213448524475,
      "learning_rate": 4.594973955507846e-05,
      "loss": 0.4689,
      "num_input_tokens_seen": 29131872,
      "step": 1110,
      "train_runtime": 28609.0166,
      "train_tokens_per_second": 1018.276
    },
    {
      "epoch": 0.36948057327479084,
      "grad_norm": 0.6004647016525269,
      "learning_rate": 4.591416658765635e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 29288216,
      "step": 1115,
      "train_runtime": 28773.4715,
      "train_tokens_per_second": 1017.89
    },
    {
      "epoch": 0.3711374368320769,
      "grad_norm": 0.8791599273681641,
      "learning_rate": 4.587845198163934e-05,
      "loss": 0.4478,
      "num_input_tokens_seen": 29429064,
      "step": 1120,
      "train_runtime": 28913.1137,
      "train_tokens_per_second": 1017.845
    },
    {
      "epoch": 0.37279430038936295,
      "grad_norm": 0.7069061398506165,
      "learning_rate": 4.5842595978900175e-05,
      "loss": 0.4657,
      "num_input_tokens_seen": 29570624,
      "step": 1125,
      "train_runtime": 29058.8889,
      "train_tokens_per_second": 1017.61
    },
    {
      "epoch": 0.374451163946649,
      "grad_norm": 0.8361027836799622,
      "learning_rate": 4.580659882226918e-05,
      "loss": 0.4119,
      "num_input_tokens_seen": 29719240,
      "step": 1130,
      "train_runtime": 29195.5922,
      "train_tokens_per_second": 1017.936
    },
    {
      "epoch": 0.37610802750393507,
      "grad_norm": 0.8364062309265137,
      "learning_rate": 4.5770460755532615e-05,
      "loss": 0.4492,
      "num_input_tokens_seen": 29828520,
      "step": 1135,
      "train_runtime": 29297.6079,
      "train_tokens_per_second": 1018.121
    },
    {
      "epoch": 0.37776489106122113,
      "grad_norm": 0.7573168873786926,
      "learning_rate": 4.573418202343106e-05,
      "loss": 0.4343,
      "num_input_tokens_seen": 29978488,
      "step": 1140,
      "train_runtime": 29449.4041,
      "train_tokens_per_second": 1017.966
    },
    {
      "epoch": 0.3794217546185072,
      "grad_norm": 1.2546659708023071,
      "learning_rate": 4.5697762871657724e-05,
      "loss": 0.4564,
      "num_input_tokens_seen": 30089840,
      "step": 1145,
      "train_runtime": 29547.987,
      "train_tokens_per_second": 1018.338
    },
    {
      "epoch": 0.38107861817579325,
      "grad_norm": 0.7081366181373596,
      "learning_rate": 4.566120354685677e-05,
      "loss": 0.4808,
      "num_input_tokens_seen": 30231264,
      "step": 1150,
      "train_runtime": 29683.8005,
      "train_tokens_per_second": 1018.443
    },
    {
      "epoch": 0.3827354817330793,
      "grad_norm": 0.6155252456665039,
      "learning_rate": 4.562450429662169e-05,
      "loss": 0.4753,
      "num_input_tokens_seen": 30370856,
      "step": 1155,
      "train_runtime": 29831.2648,
      "train_tokens_per_second": 1018.088
    },
    {
      "epoch": 0.3843923452903653,
      "grad_norm": 0.7011456489562988,
      "learning_rate": 4.55876653694936e-05,
      "loss": 0.4928,
      "num_input_tokens_seen": 30489096,
      "step": 1160,
      "train_runtime": 29946.3253,
      "train_tokens_per_second": 1018.125
    },
    {
      "epoch": 0.38604920884765137,
      "grad_norm": 0.9465348720550537,
      "learning_rate": 4.5550687014959566e-05,
      "loss": 0.5198,
      "num_input_tokens_seen": 30633632,
      "step": 1165,
      "train_runtime": 30074.2507,
      "train_tokens_per_second": 1018.6
    },
    {
      "epoch": 0.38770607240493743,
      "grad_norm": 0.687272310256958,
      "learning_rate": 4.551356948345088e-05,
      "loss": 0.4733,
      "num_input_tokens_seen": 30774456,
      "step": 1170,
      "train_runtime": 30219.2808,
      "train_tokens_per_second": 1018.372
    },
    {
      "epoch": 0.3893629359622235,
      "grad_norm": 0.8126049637794495,
      "learning_rate": 4.547631302634143e-05,
      "loss": 0.4321,
      "num_input_tokens_seen": 30897720,
      "step": 1175,
      "train_runtime": 30339.2442,
      "train_tokens_per_second": 1018.408
    },
    {
      "epoch": 0.39101979951950955,
      "grad_norm": 0.8145015239715576,
      "learning_rate": 4.543891789594593e-05,
      "loss": 0.4828,
      "num_input_tokens_seen": 31000536,
      "step": 1180,
      "train_runtime": 30433.9256,
      "train_tokens_per_second": 1018.618
    },
    {
      "epoch": 0.3926766630767956,
      "grad_norm": 0.8269211649894714,
      "learning_rate": 4.540138434551827e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 31148312,
      "step": 1185,
      "train_runtime": 30579.0581,
      "train_tokens_per_second": 1018.616
    },
    {
      "epoch": 0.39433352663408167,
      "grad_norm": 0.8473024964332581,
      "learning_rate": 4.5363712629249745e-05,
      "loss": 0.4973,
      "num_input_tokens_seen": 31279984,
      "step": 1190,
      "train_runtime": 30716.2786,
      "train_tokens_per_second": 1018.352
    },
    {
      "epoch": 0.3959903901913677,
      "grad_norm": 0.6580644249916077,
      "learning_rate": 4.532590300226737e-05,
      "loss": 0.4574,
      "num_input_tokens_seen": 31395512,
      "step": 1195,
      "train_runtime": 30824.2907,
      "train_tokens_per_second": 1018.532
    },
    {
      "epoch": 0.3976472537486538,
      "grad_norm": 0.6749723553657532,
      "learning_rate": 4.528795572063216e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 31518936,
      "step": 1200,
      "train_runtime": 30946.5583,
      "train_tokens_per_second": 1018.496
    },
    {
      "epoch": 0.39930411730593984,
      "grad_norm": 0.6987890005111694,
      "learning_rate": 4.524987104133735e-05,
      "loss": 0.5001,
      "num_input_tokens_seen": 31643304,
      "step": 1205,
      "train_runtime": 31066.0372,
      "train_tokens_per_second": 1018.582
    },
    {
      "epoch": 0.4009609808632259,
      "grad_norm": 0.7823147773742676,
      "learning_rate": 4.521164922230672e-05,
      "loss": 0.5183,
      "num_input_tokens_seen": 31792320,
      "step": 1210,
      "train_runtime": 31212.7566,
      "train_tokens_per_second": 1018.568
    },
    {
      "epoch": 0.40261784442051196,
      "grad_norm": 0.655124306678772,
      "learning_rate": 4.5173290522392774e-05,
      "loss": 0.4145,
      "num_input_tokens_seen": 31926968,
      "step": 1215,
      "train_runtime": 31336.0845,
      "train_tokens_per_second": 1018.856
    },
    {
      "epoch": 0.404274707977798,
      "grad_norm": 0.9212071299552917,
      "learning_rate": 4.513479520137505e-05,
      "loss": 0.4674,
      "num_input_tokens_seen": 32073192,
      "step": 1220,
      "train_runtime": 31480.032,
      "train_tokens_per_second": 1018.842
    },
    {
      "epoch": 0.4059315715350841,
      "grad_norm": 0.9372535943984985,
      "learning_rate": 4.509616351995831e-05,
      "loss": 0.4579,
      "num_input_tokens_seen": 32194280,
      "step": 1225,
      "train_runtime": 31605.7585,
      "train_tokens_per_second": 1018.621
    },
    {
      "epoch": 0.40758843509237014,
      "grad_norm": 0.6691573262214661,
      "learning_rate": 4.5057395739770834e-05,
      "loss": 0.3967,
      "num_input_tokens_seen": 32353200,
      "step": 1230,
      "train_runtime": 31762.9057,
      "train_tokens_per_second": 1018.584
    },
    {
      "epoch": 0.4092452986496562,
      "grad_norm": 0.6633516550064087,
      "learning_rate": 4.50184921233626e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 32461776,
      "step": 1235,
      "train_runtime": 31869.6086,
      "train_tokens_per_second": 1018.581
    },
    {
      "epoch": 0.41090216220694226,
      "grad_norm": 0.6780272722244263,
      "learning_rate": 4.49794529342035e-05,
      "loss": 0.4618,
      "num_input_tokens_seen": 32595400,
      "step": 1240,
      "train_runtime": 32001.112,
      "train_tokens_per_second": 1018.571
    },
    {
      "epoch": 0.4125590257642283,
      "grad_norm": 0.7955459952354431,
      "learning_rate": 4.49402784366816e-05,
      "loss": 0.5142,
      "num_input_tokens_seen": 32744048,
      "step": 1245,
      "train_runtime": 32156.2774,
      "train_tokens_per_second": 1018.279
    },
    {
      "epoch": 0.4142158893215144,
      "grad_norm": 0.6670683026313782,
      "learning_rate": 4.490096889610133e-05,
      "loss": 0.4176,
      "num_input_tokens_seen": 32866160,
      "step": 1250,
      "train_runtime": 32271.1941,
      "train_tokens_per_second": 1018.436
    },
    {
      "epoch": 0.41587275287880043,
      "grad_norm": 0.7879221439361572,
      "learning_rate": 4.486152457868166e-05,
      "loss": 0.4682,
      "num_input_tokens_seen": 33006096,
      "step": 1255,
      "train_runtime": 32407.7967,
      "train_tokens_per_second": 1018.462
    },
    {
      "epoch": 0.4175296164360865,
      "grad_norm": 0.6026702523231506,
      "learning_rate": 4.482194575155432e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 33143768,
      "step": 1260,
      "train_runtime": 32535.334,
      "train_tokens_per_second": 1018.701
    },
    {
      "epoch": 0.41918647999337255,
      "grad_norm": 0.6875720620155334,
      "learning_rate": 4.478223268276201e-05,
      "loss": 0.4556,
      "num_input_tokens_seen": 33262816,
      "step": 1265,
      "train_runtime": 32644.1223,
      "train_tokens_per_second": 1018.953
    },
    {
      "epoch": 0.4208433435506586,
      "grad_norm": 1.0256553888320923,
      "learning_rate": 4.474238564125656e-05,
      "loss": 0.4787,
      "num_input_tokens_seen": 33415904,
      "step": 1270,
      "train_runtime": 32796.7193,
      "train_tokens_per_second": 1018.879
    },
    {
      "epoch": 0.42250020710794467,
      "grad_norm": 0.613653838634491,
      "learning_rate": 4.470240489689709e-05,
      "loss": 0.4377,
      "num_input_tokens_seen": 33535392,
      "step": 1275,
      "train_runtime": 32898.8551,
      "train_tokens_per_second": 1019.348
    },
    {
      "epoch": 0.42415707066523073,
      "grad_norm": 0.7099058032035828,
      "learning_rate": 4.466229072044824e-05,
      "loss": 0.4064,
      "num_input_tokens_seen": 33640704,
      "step": 1280,
      "train_runtime": 33002.6339,
      "train_tokens_per_second": 1019.334
    },
    {
      "epoch": 0.4258139342225168,
      "grad_norm": 0.8235899806022644,
      "learning_rate": 4.462204338357828e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 33761800,
      "step": 1285,
      "train_runtime": 33123.2523,
      "train_tokens_per_second": 1019.278
    },
    {
      "epoch": 0.42747079777980285,
      "grad_norm": 0.7067025899887085,
      "learning_rate": 4.4581663158857295e-05,
      "loss": 0.4651,
      "num_input_tokens_seen": 33871080,
      "step": 1290,
      "train_runtime": 33228.9382,
      "train_tokens_per_second": 1019.325
    },
    {
      "epoch": 0.4291276613370889,
      "grad_norm": 0.7987658381462097,
      "learning_rate": 4.4541150319755354e-05,
      "loss": 0.4865,
      "num_input_tokens_seen": 33988640,
      "step": 1295,
      "train_runtime": 33349.9713,
      "train_tokens_per_second": 1019.151
    },
    {
      "epoch": 0.43078452489437496,
      "grad_norm": 0.7586035132408142,
      "learning_rate": 4.450050514064061e-05,
      "loss": 0.4279,
      "num_input_tokens_seen": 34109296,
      "step": 1300,
      "train_runtime": 33467.1104,
      "train_tokens_per_second": 1019.189
    },
    {
      "epoch": 0.432441388451661,
      "grad_norm": 0.9816948771476746,
      "learning_rate": 4.4459727896777496e-05,
      "loss": 0.4663,
      "num_input_tokens_seen": 34223840,
      "step": 1305,
      "train_runtime": 33579.3929,
      "train_tokens_per_second": 1019.192
    },
    {
      "epoch": 0.4340982520089471,
      "grad_norm": 0.7071145176887512,
      "learning_rate": 4.441881886432484e-05,
      "loss": 0.4482,
      "num_input_tokens_seen": 34361848,
      "step": 1310,
      "train_runtime": 33719.3459,
      "train_tokens_per_second": 1019.054
    },
    {
      "epoch": 0.43575511556623314,
      "grad_norm": 0.6951707601547241,
      "learning_rate": 4.437777832033396e-05,
      "loss": 0.3762,
      "num_input_tokens_seen": 34469784,
      "step": 1315,
      "train_runtime": 33820.053,
      "train_tokens_per_second": 1019.211
    },
    {
      "epoch": 0.4374119791235192,
      "grad_norm": 1.4176478385925293,
      "learning_rate": 4.433660654274686e-05,
      "loss": 0.5211,
      "num_input_tokens_seen": 34602144,
      "step": 1320,
      "train_runtime": 33954.6676,
      "train_tokens_per_second": 1019.069
    },
    {
      "epoch": 0.43906884268080526,
      "grad_norm": 0.8272822499275208,
      "learning_rate": 4.4295303810394284e-05,
      "loss": 0.4417,
      "num_input_tokens_seen": 34707544,
      "step": 1325,
      "train_runtime": 34055.7902,
      "train_tokens_per_second": 1019.138
    },
    {
      "epoch": 0.4407257062380913,
      "grad_norm": 0.7135868668556213,
      "learning_rate": 4.425387040299385e-05,
      "loss": 0.422,
      "num_input_tokens_seen": 34817584,
      "step": 1330,
      "train_runtime": 34164.0751,
      "train_tokens_per_second": 1019.129
    },
    {
      "epoch": 0.4423825697953774,
      "grad_norm": 0.9297186136245728,
      "learning_rate": 4.4212306601148166e-05,
      "loss": 0.4765,
      "num_input_tokens_seen": 34941672,
      "step": 1335,
      "train_runtime": 34280.9453,
      "train_tokens_per_second": 1019.274
    },
    {
      "epoch": 0.4440394333526634,
      "grad_norm": 0.8199912905693054,
      "learning_rate": 4.417061268634292e-05,
      "loss": 0.4334,
      "num_input_tokens_seen": 35083888,
      "step": 1340,
      "train_runtime": 34415.5254,
      "train_tokens_per_second": 1019.42
    },
    {
      "epoch": 0.44569629690994944,
      "grad_norm": 0.7122479677200317,
      "learning_rate": 4.4128788940944965e-05,
      "loss": 0.447,
      "num_input_tokens_seen": 35234408,
      "step": 1345,
      "train_runtime": 34571.68,
      "train_tokens_per_second": 1019.17
    },
    {
      "epoch": 0.4473531604672355,
      "grad_norm": 0.5812869071960449,
      "learning_rate": 4.408683564820042e-05,
      "loss": 0.3651,
      "num_input_tokens_seen": 35355080,
      "step": 1350,
      "train_runtime": 34686.0924,
      "train_tokens_per_second": 1019.287
    },
    {
      "epoch": 0.44901002402452156,
      "grad_norm": 0.7574847340583801,
      "learning_rate": 4.404475309223275e-05,
      "loss": 0.5477,
      "num_input_tokens_seen": 35500448,
      "step": 1355,
      "train_runtime": 34841.3778,
      "train_tokens_per_second": 1018.916
    },
    {
      "epoch": 0.4506668875818076,
      "grad_norm": 0.6787741780281067,
      "learning_rate": 4.400254155804084e-05,
      "loss": 0.4504,
      "num_input_tokens_seen": 35606888,
      "step": 1360,
      "train_runtime": 34925.0018,
      "train_tokens_per_second": 1019.524
    },
    {
      "epoch": 0.4523237511390937,
      "grad_norm": 0.7749010324478149,
      "learning_rate": 4.396020133149705e-05,
      "loss": 0.4821,
      "num_input_tokens_seen": 35731304,
      "step": 1365,
      "train_runtime": 35036.8635,
      "train_tokens_per_second": 1019.82
    },
    {
      "epoch": 0.45398061469637974,
      "grad_norm": 0.7534992098808289,
      "learning_rate": 4.3917732699345315e-05,
      "loss": 0.4488,
      "num_input_tokens_seen": 35864544,
      "step": 1370,
      "train_runtime": 35172.2573,
      "train_tokens_per_second": 1019.683
    },
    {
      "epoch": 0.4556374782536658,
      "grad_norm": 0.7417518496513367,
      "learning_rate": 4.387513594919917e-05,
      "loss": 0.3998,
      "num_input_tokens_seen": 35970208,
      "step": 1375,
      "train_runtime": 35275.3181,
      "train_tokens_per_second": 1019.699
    },
    {
      "epoch": 0.45729434181095185,
      "grad_norm": 0.8546368479728699,
      "learning_rate": 4.383241136953981e-05,
      "loss": 0.3939,
      "num_input_tokens_seen": 36110112,
      "step": 1380,
      "train_runtime": 35409.9069,
      "train_tokens_per_second": 1019.774
    },
    {
      "epoch": 0.4589512053682379,
      "grad_norm": 0.6396704912185669,
      "learning_rate": 4.378955924971414e-05,
      "loss": 0.3707,
      "num_input_tokens_seen": 36228592,
      "step": 1385,
      "train_runtime": 35510.3951,
      "train_tokens_per_second": 1020.225
    },
    {
      "epoch": 0.460608068925524,
      "grad_norm": 0.6036563515663147,
      "learning_rate": 4.3746579879932826e-05,
      "loss": 0.4245,
      "num_input_tokens_seen": 36330832,
      "step": 1390,
      "train_runtime": 35611.3312,
      "train_tokens_per_second": 1020.204
    },
    {
      "epoch": 0.46226493248281003,
      "grad_norm": 0.8486759662628174,
      "learning_rate": 4.37034735512683e-05,
      "loss": 0.482,
      "num_input_tokens_seen": 36466744,
      "step": 1395,
      "train_runtime": 35742.6765,
      "train_tokens_per_second": 1020.258
    },
    {
      "epoch": 0.4639217960400961,
      "grad_norm": 0.7479493618011475,
      "learning_rate": 4.366024055565282e-05,
      "loss": 0.4883,
      "num_input_tokens_seen": 36605776,
      "step": 1400,
      "train_runtime": 35886.9712,
      "train_tokens_per_second": 1020.03
    },
    {
      "epoch": 0.46557865959738215,
      "grad_norm": 0.7107446193695068,
      "learning_rate": 4.361688118587648e-05,
      "loss": 0.4108,
      "num_input_tokens_seen": 36750096,
      "step": 1405,
      "train_runtime": 36028.985,
      "train_tokens_per_second": 1020.015
    },
    {
      "epoch": 0.4672355231546682,
      "grad_norm": 0.757632851600647,
      "learning_rate": 4.357339573558523e-05,
      "loss": 0.5153,
      "num_input_tokens_seen": 36865328,
      "step": 1410,
      "train_runtime": 36146.1224,
      "train_tokens_per_second": 1019.897
    },
    {
      "epoch": 0.46889238671195427,
      "grad_norm": 0.8993147611618042,
      "learning_rate": 4.352978449927887e-05,
      "loss": 0.3911,
      "num_input_tokens_seen": 36991168,
      "step": 1415,
      "train_runtime": 36277.3452,
      "train_tokens_per_second": 1019.677
    },
    {
      "epoch": 0.4705492502692403,
      "grad_norm": 0.6440050005912781,
      "learning_rate": 4.348604777230909e-05,
      "loss": 0.4324,
      "num_input_tokens_seen": 37106032,
      "step": 1420,
      "train_runtime": 36385.4649,
      "train_tokens_per_second": 1019.804
    },
    {
      "epoch": 0.4722061138265264,
      "grad_norm": 0.6604807376861572,
      "learning_rate": 4.344218585087743e-05,
      "loss": 0.4824,
      "num_input_tokens_seen": 37242568,
      "step": 1425,
      "train_runtime": 36515.9691,
      "train_tokens_per_second": 1019.898
    },
    {
      "epoch": 0.47386297738381244,
      "grad_norm": 0.802575409412384,
      "learning_rate": 4.339819903203331e-05,
      "loss": 0.4383,
      "num_input_tokens_seen": 37381192,
      "step": 1430,
      "train_runtime": 36664.9918,
      "train_tokens_per_second": 1019.534
    },
    {
      "epoch": 0.4755198409410985,
      "grad_norm": 0.6782968044281006,
      "learning_rate": 4.335408761367198e-05,
      "loss": 0.4231,
      "num_input_tokens_seen": 37481032,
      "step": 1435,
      "train_runtime": 36762.552,
      "train_tokens_per_second": 1019.544
    },
    {
      "epoch": 0.47717670449838456,
      "grad_norm": 0.8068889379501343,
      "learning_rate": 4.3309851894532566e-05,
      "loss": 0.416,
      "num_input_tokens_seen": 37601408,
      "step": 1440,
      "train_runtime": 36883.5702,
      "train_tokens_per_second": 1019.462
    },
    {
      "epoch": 0.4788335680556706,
      "grad_norm": 0.6284390687942505,
      "learning_rate": 4.3265492174195964e-05,
      "loss": 0.411,
      "num_input_tokens_seen": 37726312,
      "step": 1445,
      "train_runtime": 37007.504,
      "train_tokens_per_second": 1019.423
    },
    {
      "epoch": 0.4804904316129567,
      "grad_norm": 1.264664888381958,
      "learning_rate": 4.322100875308287e-05,
      "loss": 0.4411,
      "num_input_tokens_seen": 37838272,
      "step": 1450,
      "train_runtime": 37110.9876,
      "train_tokens_per_second": 1019.598
    },
    {
      "epoch": 0.48214729517024274,
      "grad_norm": 0.7960128784179688,
      "learning_rate": 4.317640193245174e-05,
      "loss": 0.3874,
      "num_input_tokens_seen": 38016128,
      "step": 1455,
      "train_runtime": 37303.8716,
      "train_tokens_per_second": 1019.093
    },
    {
      "epoch": 0.4838041587275288,
      "grad_norm": 0.7297863960266113,
      "learning_rate": 4.313167201439671e-05,
      "loss": 0.4717,
      "num_input_tokens_seen": 38119904,
      "step": 1460,
      "train_runtime": 37395.2998,
      "train_tokens_per_second": 1019.377
    },
    {
      "epoch": 0.48546102228481486,
      "grad_norm": 0.8983017206192017,
      "learning_rate": 4.308681930184559e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 38260784,
      "step": 1465,
      "train_runtime": 37546.8187,
      "train_tokens_per_second": 1019.015
    },
    {
      "epoch": 0.4871178858421009,
      "grad_norm": 0.6402216553688049,
      "learning_rate": 4.304184409855782e-05,
      "loss": 0.4315,
      "num_input_tokens_seen": 38388952,
      "step": 1470,
      "train_runtime": 37679.3491,
      "train_tokens_per_second": 1018.833
    },
    {
      "epoch": 0.488774749399387,
      "grad_norm": 0.8213521838188171,
      "learning_rate": 4.299674670912236e-05,
      "loss": 0.4451,
      "num_input_tokens_seen": 38529688,
      "step": 1475,
      "train_runtime": 37812.937,
      "train_tokens_per_second": 1018.955
    },
    {
      "epoch": 0.49043161295667304,
      "grad_norm": 0.7938104867935181,
      "learning_rate": 4.295152743895569e-05,
      "loss": 0.4217,
      "num_input_tokens_seen": 38649104,
      "step": 1480,
      "train_runtime": 37937.1221,
      "train_tokens_per_second": 1018.767
    },
    {
      "epoch": 0.4920884765139591,
      "grad_norm": 0.8412910103797913,
      "learning_rate": 4.290618659429969e-05,
      "loss": 0.5077,
      "num_input_tokens_seen": 38778080,
      "step": 1485,
      "train_runtime": 38053.4266,
      "train_tokens_per_second": 1019.043
    },
    {
      "epoch": 0.49374534007124515,
      "grad_norm": 0.6579925417900085,
      "learning_rate": 4.2860724482219604e-05,
      "loss": 0.5089,
      "num_input_tokens_seen": 38905816,
      "step": 1490,
      "train_runtime": 38187.1748,
      "train_tokens_per_second": 1018.819
    },
    {
      "epoch": 0.4954022036285312,
      "grad_norm": 0.7261325716972351,
      "learning_rate": 4.281514141060193e-05,
      "loss": 0.5008,
      "num_input_tokens_seen": 39006448,
      "step": 1495,
      "train_runtime": 38279.439,
      "train_tokens_per_second": 1018.992
    },
    {
      "epoch": 0.49705906718581727,
      "grad_norm": 0.6003370881080627,
      "learning_rate": 4.276943768815237e-05,
      "loss": 0.4654,
      "num_input_tokens_seen": 39121072,
      "step": 1500,
      "train_runtime": 38392.2699,
      "train_tokens_per_second": 1018.983
    },
    {
      "epoch": 0.49871593074310333,
      "grad_norm": 0.8921909332275391,
      "learning_rate": 4.27236136243937e-05,
      "loss": 0.4527,
      "num_input_tokens_seen": 39251768,
      "step": 1505,
      "train_runtime": 38523.7324,
      "train_tokens_per_second": 1018.898
    },
    {
      "epoch": 0.5003727943003894,
      "grad_norm": 0.6896916627883911,
      "learning_rate": 4.267766952966369e-05,
      "loss": 0.4445,
      "num_input_tokens_seen": 39367384,
      "step": 1510,
      "train_runtime": 38632.1276,
      "train_tokens_per_second": 1019.032
    },
    {
      "epoch": 0.5020296578576754,
      "grad_norm": 0.7951368689537048,
      "learning_rate": 4.263160571511303e-05,
      "loss": 0.4391,
      "num_input_tokens_seen": 39470208,
      "step": 1515,
      "train_runtime": 38731.4281,
      "train_tokens_per_second": 1019.074
    },
    {
      "epoch": 0.5036865214149615,
      "grad_norm": 0.6856791973114014,
      "learning_rate": 4.258542249270316e-05,
      "loss": 0.4521,
      "num_input_tokens_seen": 39597784,
      "step": 1520,
      "train_runtime": 38853.3949,
      "train_tokens_per_second": 1019.159
    },
    {
      "epoch": 0.5053433849722475,
      "grad_norm": 0.8171475529670715,
      "learning_rate": 4.253912017520424e-05,
      "loss": 0.3907,
      "num_input_tokens_seen": 39684384,
      "step": 1525,
      "train_runtime": 38934.7067,
      "train_tokens_per_second": 1019.255
    },
    {
      "epoch": 0.5070002485295336,
      "grad_norm": 0.7384614944458008,
      "learning_rate": 4.249269907619294e-05,
      "loss": 0.4497,
      "num_input_tokens_seen": 39803384,
      "step": 1530,
      "train_runtime": 39052.8561,
      "train_tokens_per_second": 1019.218
    },
    {
      "epoch": 0.5086571120868196,
      "grad_norm": 0.5552036762237549,
      "learning_rate": 4.2446159510050405e-05,
      "loss": 0.427,
      "num_input_tokens_seen": 39973640,
      "step": 1535,
      "train_runtime": 39223.129,
      "train_tokens_per_second": 1019.134
    },
    {
      "epoch": 0.5103139756441057,
      "grad_norm": 0.7796664237976074,
      "learning_rate": 4.239950179196005e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 40099160,
      "step": 1540,
      "train_runtime": 39355.8753,
      "train_tokens_per_second": 1018.886
    },
    {
      "epoch": 0.5119708392013917,
      "grad_norm": 0.7219895124435425,
      "learning_rate": 4.2352726237905496e-05,
      "loss": 0.4833,
      "num_input_tokens_seen": 40236784,
      "step": 1545,
      "train_runtime": 39492.4363,
      "train_tokens_per_second": 1018.848
    },
    {
      "epoch": 0.5136277027586779,
      "grad_norm": 1.109179139137268,
      "learning_rate": 4.230583316466834e-05,
      "loss": 0.4489,
      "num_input_tokens_seen": 40354776,
      "step": 1550,
      "train_runtime": 39607.2655,
      "train_tokens_per_second": 1018.873
    },
    {
      "epoch": 0.5152845663159639,
      "grad_norm": 0.7547450661659241,
      "learning_rate": 4.225882288982612e-05,
      "loss": 0.4776,
      "num_input_tokens_seen": 40473600,
      "step": 1555,
      "train_runtime": 39717.5528,
      "train_tokens_per_second": 1019.036
    },
    {
      "epoch": 0.51694142987325,
      "grad_norm": 0.7080791592597961,
      "learning_rate": 4.221169573175007e-05,
      "loss": 0.4128,
      "num_input_tokens_seen": 40598160,
      "step": 1560,
      "train_runtime": 39837.9224,
      "train_tokens_per_second": 1019.083
    },
    {
      "epoch": 0.518598293430536,
      "grad_norm": 0.6733773946762085,
      "learning_rate": 4.216445200960302e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 40715336,
      "step": 1565,
      "train_runtime": 39952.7077,
      "train_tokens_per_second": 1019.088
    },
    {
      "epoch": 0.5202551569878221,
      "grad_norm": 0.9929623603820801,
      "learning_rate": 4.211709204333719e-05,
      "loss": 0.4351,
      "num_input_tokens_seen": 40848512,
      "step": 1570,
      "train_runtime": 40084.0175,
      "train_tokens_per_second": 1019.072
    },
    {
      "epoch": 0.5219120205451081,
      "grad_norm": 0.6040559411048889,
      "learning_rate": 4.2069616153692094e-05,
      "loss": 0.4458,
      "num_input_tokens_seen": 40973512,
      "step": 1575,
      "train_runtime": 40197.2692,
      "train_tokens_per_second": 1019.311
    },
    {
      "epoch": 0.5235688841023942,
      "grad_norm": 0.7203269600868225,
      "learning_rate": 4.202202466219228e-05,
      "loss": 0.4194,
      "num_input_tokens_seen": 41106864,
      "step": 1580,
      "train_runtime": 40329.1266,
      "train_tokens_per_second": 1019.285
    },
    {
      "epoch": 0.5252257476596802,
      "grad_norm": 0.7061666250228882,
      "learning_rate": 4.197431789114522e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 41249192,
      "step": 1585,
      "train_runtime": 40477.5359,
      "train_tokens_per_second": 1019.064
    },
    {
      "epoch": 0.5268826112169663,
      "grad_norm": 0.6258950233459473,
      "learning_rate": 4.1926496163639096e-05,
      "loss": 0.3707,
      "num_input_tokens_seen": 41369112,
      "step": 1590,
      "train_runtime": 40593.6803,
      "train_tokens_per_second": 1019.102
    },
    {
      "epoch": 0.5285394747742523,
      "grad_norm": 0.7151850461959839,
      "learning_rate": 4.187855980354061e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 41502480,
      "step": 1595,
      "train_runtime": 40733.7606,
      "train_tokens_per_second": 1018.872
    },
    {
      "epoch": 0.5301963383315385,
      "grad_norm": 0.5882954001426697,
      "learning_rate": 4.183050913549279e-05,
      "loss": 0.4098,
      "num_input_tokens_seen": 41634216,
      "step": 1600,
      "train_runtime": 40861.978,
      "train_tokens_per_second": 1018.899
    },
    {
      "epoch": 0.5318532018888245,
      "grad_norm": 0.8259994387626648,
      "learning_rate": 4.1782344484912845e-05,
      "loss": 0.4184,
      "num_input_tokens_seen": 41768440,
      "step": 1605,
      "train_runtime": 41003.2089,
      "train_tokens_per_second": 1018.663
    },
    {
      "epoch": 0.5335100654461105,
      "grad_norm": 0.5952545404434204,
      "learning_rate": 4.173406617798987e-05,
      "loss": 0.3997,
      "num_input_tokens_seen": 41917512,
      "step": 1610,
      "train_runtime": 41150.0286,
      "train_tokens_per_second": 1018.651
    },
    {
      "epoch": 0.5351669290033966,
      "grad_norm": 0.7103211879730225,
      "learning_rate": 4.168567454168271e-05,
      "loss": 0.4077,
      "num_input_tokens_seen": 42060440,
      "step": 1615,
      "train_runtime": 41284.3349,
      "train_tokens_per_second": 1018.799
    },
    {
      "epoch": 0.5368237925606826,
      "grad_norm": 0.8635889887809753,
      "learning_rate": 4.16371699037177e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 42190520,
      "step": 1620,
      "train_runtime": 41402.4174,
      "train_tokens_per_second": 1019.035
    },
    {
      "epoch": 0.5384806561179687,
      "grad_norm": 0.7823556065559387,
      "learning_rate": 4.158855259258647e-05,
      "loss": 0.4581,
      "num_input_tokens_seen": 42320608,
      "step": 1625,
      "train_runtime": 41531.8193,
      "train_tokens_per_second": 1018.992
    },
    {
      "epoch": 0.5401375196752547,
      "grad_norm": 0.5964196920394897,
      "learning_rate": 4.1539822937543735e-05,
      "loss": 0.4557,
      "num_input_tokens_seen": 42466344,
      "step": 1630,
      "train_runtime": 41681.0934,
      "train_tokens_per_second": 1018.839
    },
    {
      "epoch": 0.5417943832325408,
      "grad_norm": 0.7347302436828613,
      "learning_rate": 4.149098126860503e-05,
      "loss": 0.4168,
      "num_input_tokens_seen": 42594184,
      "step": 1635,
      "train_runtime": 41806.2638,
      "train_tokens_per_second": 1018.847
    },
    {
      "epoch": 0.5434512467898268,
      "grad_norm": 0.8562350869178772,
      "learning_rate": 4.144202791654448e-05,
      "loss": 0.4283,
      "num_input_tokens_seen": 42725008,
      "step": 1640,
      "train_runtime": 41928.3679,
      "train_tokens_per_second": 1019.0
    },
    {
      "epoch": 0.5451081103471129,
      "grad_norm": 0.8074847459793091,
      "learning_rate": 4.139296321289258e-05,
      "loss": 0.4771,
      "num_input_tokens_seen": 42878216,
      "step": 1645,
      "train_runtime": 42083.5289,
      "train_tokens_per_second": 1018.884
    },
    {
      "epoch": 0.5467649739043989,
      "grad_norm": 0.8461412787437439,
      "learning_rate": 4.134378748993394e-05,
      "loss": 0.472,
      "num_input_tokens_seen": 43002696,
      "step": 1650,
      "train_runtime": 42210.6894,
      "train_tokens_per_second": 1018.763
    },
    {
      "epoch": 0.548421837461685,
      "grad_norm": 1.114956259727478,
      "learning_rate": 4.1294501080705033e-05,
      "loss": 0.4846,
      "num_input_tokens_seen": 43141120,
      "step": 1655,
      "train_runtime": 42347.8848,
      "train_tokens_per_second": 1018.731
    },
    {
      "epoch": 0.550078701018971,
      "grad_norm": 0.926898717880249,
      "learning_rate": 4.124510431899196e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 43307576,
      "step": 1660,
      "train_runtime": 42522.2093,
      "train_tokens_per_second": 1018.47
    },
    {
      "epoch": 0.5517355645762572,
      "grad_norm": 1.0194493532180786,
      "learning_rate": 4.1195597539328125e-05,
      "loss": 0.485,
      "num_input_tokens_seen": 43456208,
      "step": 1665,
      "train_runtime": 42682.4802,
      "train_tokens_per_second": 1018.128
    },
    {
      "epoch": 0.5533924281335432,
      "grad_norm": 0.8047080039978027,
      "learning_rate": 4.1145981076992066e-05,
      "loss": 0.4758,
      "num_input_tokens_seen": 43592072,
      "step": 1670,
      "train_runtime": 42819.2812,
      "train_tokens_per_second": 1018.048
    },
    {
      "epoch": 0.5550492916908293,
      "grad_norm": 0.7355964183807373,
      "learning_rate": 4.10962552680051e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 43718944,
      "step": 1675,
      "train_runtime": 42942.3074,
      "train_tokens_per_second": 1018.086
    },
    {
      "epoch": 0.5567061552481153,
      "grad_norm": 0.7899766564369202,
      "learning_rate": 4.1046420449129085e-05,
      "loss": 0.4604,
      "num_input_tokens_seen": 43855464,
      "step": 1680,
      "train_runtime": 43077.5777,
      "train_tokens_per_second": 1018.058
    },
    {
      "epoch": 0.5583630188054014,
      "grad_norm": 0.7572499513626099,
      "learning_rate": 4.0996476957864146e-05,
      "loss": 0.4797,
      "num_input_tokens_seen": 44000512,
      "step": 1685,
      "train_runtime": 43222.6579,
      "train_tokens_per_second": 1017.996
    },
    {
      "epoch": 0.5600198823626874,
      "grad_norm": 0.68511962890625,
      "learning_rate": 4.094642513244638e-05,
      "loss": 0.4001,
      "num_input_tokens_seen": 44106472,
      "step": 1690,
      "train_runtime": 43323.2997,
      "train_tokens_per_second": 1018.077
    },
    {
      "epoch": 0.5616767459199735,
      "grad_norm": 0.682883083820343,
      "learning_rate": 4.089626531184555e-05,
      "loss": 0.4312,
      "num_input_tokens_seen": 44220368,
      "step": 1695,
      "train_runtime": 43430.9598,
      "train_tokens_per_second": 1018.176
    },
    {
      "epoch": 0.5633336094772595,
      "grad_norm": 0.6726131439208984,
      "learning_rate": 4.084599783576281e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 44332752,
      "step": 1700,
      "train_runtime": 43536.2719,
      "train_tokens_per_second": 1018.295
    },
    {
      "epoch": 0.5649904730345456,
      "grad_norm": 0.7754923701286316,
      "learning_rate": 4.0795623044628386e-05,
      "loss": 0.3575,
      "num_input_tokens_seen": 44450776,
      "step": 1705,
      "train_runtime": 43654.8581,
      "train_tokens_per_second": 1018.232
    },
    {
      "epoch": 0.5666473365918316,
      "grad_norm": 0.8166598677635193,
      "learning_rate": 4.074514127959931e-05,
      "loss": 0.5342,
      "num_input_tokens_seen": 44591832,
      "step": 1710,
      "train_runtime": 43795.3755,
      "train_tokens_per_second": 1018.186
    },
    {
      "epoch": 0.5683042001491178,
      "grad_norm": 0.6547060012817383,
      "learning_rate": 4.0694552882557045e-05,
      "loss": 0.4388,
      "num_input_tokens_seen": 44754072,
      "step": 1715,
      "train_runtime": 43966.9859,
      "train_tokens_per_second": 1017.902
    },
    {
      "epoch": 0.5699610637064038,
      "grad_norm": 0.8993564248085022,
      "learning_rate": 4.064385819610524e-05,
      "loss": 0.4248,
      "num_input_tokens_seen": 44886344,
      "step": 1720,
      "train_runtime": 44081.1525,
      "train_tokens_per_second": 1018.266
    },
    {
      "epoch": 0.5716179272636899,
      "grad_norm": 0.7738493084907532,
      "learning_rate": 4.059305756356737e-05,
      "loss": 0.4329,
      "num_input_tokens_seen": 45011600,
      "step": 1725,
      "train_runtime": 44200.3293,
      "train_tokens_per_second": 1018.354
    },
    {
      "epoch": 0.5732747908209759,
      "grad_norm": 0.8054126501083374,
      "learning_rate": 4.0542151328984384e-05,
      "loss": 0.4425,
      "num_input_tokens_seen": 45108344,
      "step": 1730,
      "train_runtime": 44287.6866,
      "train_tokens_per_second": 1018.53
    },
    {
      "epoch": 0.574931654378262,
      "grad_norm": 0.6723814010620117,
      "learning_rate": 4.0491139837112446e-05,
      "loss": 0.4277,
      "num_input_tokens_seen": 45230712,
      "step": 1735,
      "train_runtime": 44413.8698,
      "train_tokens_per_second": 1018.392
    },
    {
      "epoch": 0.576588517935548,
      "grad_norm": 0.7484299540519714,
      "learning_rate": 4.044002343342054e-05,
      "loss": 0.4095,
      "num_input_tokens_seen": 45378280,
      "step": 1740,
      "train_runtime": 44565.7765,
      "train_tokens_per_second": 1018.232
    },
    {
      "epoch": 0.5782453814928341,
      "grad_norm": 0.7660119533538818,
      "learning_rate": 4.038880246408816e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 45492728,
      "step": 1745,
      "train_runtime": 44679.3109,
      "train_tokens_per_second": 1018.206
    },
    {
      "epoch": 0.5799022450501201,
      "grad_norm": 0.8029608726501465,
      "learning_rate": 4.0337477276002964e-05,
      "loss": 0.4297,
      "num_input_tokens_seen": 45625600,
      "step": 1750,
      "train_runtime": 44814.867,
      "train_tokens_per_second": 1018.091
    },
    {
      "epoch": 0.5815591086074062,
      "grad_norm": 0.6781209111213684,
      "learning_rate": 4.0286048216758406e-05,
      "loss": 0.4313,
      "num_input_tokens_seen": 45728776,
      "step": 1755,
      "train_runtime": 44912.4208,
      "train_tokens_per_second": 1018.177
    },
    {
      "epoch": 0.5832159721646922,
      "grad_norm": 0.9115334153175354,
      "learning_rate": 4.02345156346514e-05,
      "loss": 0.5091,
      "num_input_tokens_seen": 45877888,
      "step": 1760,
      "train_runtime": 45056.1562,
      "train_tokens_per_second": 1018.238
    },
    {
      "epoch": 0.5848728357219783,
      "grad_norm": 0.8252612352371216,
      "learning_rate": 4.0182879878679955e-05,
      "loss": 0.3957,
      "num_input_tokens_seen": 45984440,
      "step": 1765,
      "train_runtime": 45156.5866,
      "train_tokens_per_second": 1018.333
    },
    {
      "epoch": 0.5865296992792643,
      "grad_norm": 0.8437846302986145,
      "learning_rate": 4.013114129854082e-05,
      "loss": 0.5587,
      "num_input_tokens_seen": 46139328,
      "step": 1770,
      "train_runtime": 45296.4299,
      "train_tokens_per_second": 1018.608
    },
    {
      "epoch": 0.5881865628365505,
      "grad_norm": 0.7869289517402649,
      "learning_rate": 4.0079300244627094e-05,
      "loss": 0.5021,
      "num_input_tokens_seen": 46265296,
      "step": 1775,
      "train_runtime": 45413.3679,
      "train_tokens_per_second": 1018.759
    },
    {
      "epoch": 0.5898434263938365,
      "grad_norm": 0.8693633675575256,
      "learning_rate": 4.002735706802587e-05,
      "loss": 0.3806,
      "num_input_tokens_seen": 46396184,
      "step": 1780,
      "train_runtime": 45545.1332,
      "train_tokens_per_second": 1018.686
    },
    {
      "epoch": 0.5915002899511225,
      "grad_norm": 0.9264876246452332,
      "learning_rate": 3.9975312120515884e-05,
      "loss": 0.4291,
      "num_input_tokens_seen": 46506352,
      "step": 1785,
      "train_runtime": 45648.8139,
      "train_tokens_per_second": 1018.786
    },
    {
      "epoch": 0.5931571535084086,
      "grad_norm": 0.8768818974494934,
      "learning_rate": 3.9923165754565046e-05,
      "loss": 0.3802,
      "num_input_tokens_seen": 46647728,
      "step": 1790,
      "train_runtime": 45792.9732,
      "train_tokens_per_second": 1018.666
    },
    {
      "epoch": 0.5948140170656946,
      "grad_norm": 0.731304407119751,
      "learning_rate": 3.987091832332816e-05,
      "loss": 0.4057,
      "num_input_tokens_seen": 46761904,
      "step": 1795,
      "train_runtime": 45893.9641,
      "train_tokens_per_second": 1018.912
    },
    {
      "epoch": 0.5964708806229807,
      "grad_norm": 0.9606395959854126,
      "learning_rate": 3.981857018064444e-05,
      "loss": 0.4643,
      "num_input_tokens_seen": 46884408,
      "step": 1800,
      "train_runtime": 46018.6808,
      "train_tokens_per_second": 1018.813
    },
    {
      "epoch": 0.5981277441802667,
      "grad_norm": 0.7286819815635681,
      "learning_rate": 3.9766121681035214e-05,
      "loss": 0.48,
      "num_input_tokens_seen": 47046952,
      "step": 1805,
      "train_runtime": 46188.7293,
      "train_tokens_per_second": 1018.581
    },
    {
      "epoch": 0.5997846077375528,
      "grad_norm": 0.9200510382652283,
      "learning_rate": 3.97135731797014e-05,
      "loss": 0.4531,
      "num_input_tokens_seen": 47197592,
      "step": 1810,
      "train_runtime": 46342.5647,
      "train_tokens_per_second": 1018.45
    },
    {
      "epoch": 0.6014414712948388,
      "grad_norm": 0.7427248358726501,
      "learning_rate": 3.96609250325212e-05,
      "loss": 0.443,
      "num_input_tokens_seen": 47340456,
      "step": 1815,
      "train_runtime": 46493.5149,
      "train_tokens_per_second": 1018.216
    },
    {
      "epoch": 0.6030983348521249,
      "grad_norm": 0.8811022043228149,
      "learning_rate": 3.960817759604767e-05,
      "loss": 0.4246,
      "num_input_tokens_seen": 47453888,
      "step": 1820,
      "train_runtime": 46609.1471,
      "train_tokens_per_second": 1018.124
    },
    {
      "epoch": 0.6047551984094109,
      "grad_norm": 0.6852868795394897,
      "learning_rate": 3.955533122750624e-05,
      "loss": 0.5002,
      "num_input_tokens_seen": 47622176,
      "step": 1825,
      "train_runtime": 46787.8652,
      "train_tokens_per_second": 1017.832
    },
    {
      "epoch": 0.6064120619666971,
      "grad_norm": 0.8648651838302612,
      "learning_rate": 3.9502386284792404e-05,
      "loss": 0.4415,
      "num_input_tokens_seen": 47758736,
      "step": 1830,
      "train_runtime": 46919.5082,
      "train_tokens_per_second": 1017.887
    },
    {
      "epoch": 0.6080689255239831,
      "grad_norm": 0.8359082937240601,
      "learning_rate": 3.944934312646919e-05,
      "loss": 0.486,
      "num_input_tokens_seen": 47894376,
      "step": 1835,
      "train_runtime": 47059.4228,
      "train_tokens_per_second": 1017.743
    },
    {
      "epoch": 0.6097257890812692,
      "grad_norm": 0.980783224105835,
      "learning_rate": 3.9396202111764804e-05,
      "loss": 0.4857,
      "num_input_tokens_seen": 48011720,
      "step": 1840,
      "train_runtime": 47159.6969,
      "train_tokens_per_second": 1018.067
    },
    {
      "epoch": 0.6113826526385552,
      "grad_norm": 0.604836642742157,
      "learning_rate": 3.934296360057016e-05,
      "loss": 0.4142,
      "num_input_tokens_seen": 48119680,
      "step": 1845,
      "train_runtime": 47257.0906,
      "train_tokens_per_second": 1018.253
    },
    {
      "epoch": 0.6130395161958413,
      "grad_norm": 0.6503739953041077,
      "learning_rate": 3.928962795343647e-05,
      "loss": 0.4394,
      "num_input_tokens_seen": 48270672,
      "step": 1850,
      "train_runtime": 47415.5937,
      "train_tokens_per_second": 1018.034
    },
    {
      "epoch": 0.6146963797531273,
      "grad_norm": 0.7693909406661987,
      "learning_rate": 3.9236195531572774e-05,
      "loss": 0.4908,
      "num_input_tokens_seen": 48415480,
      "step": 1855,
      "train_runtime": 47548.4808,
      "train_tokens_per_second": 1018.234
    },
    {
      "epoch": 0.6163532433104134,
      "grad_norm": 1.0091125965118408,
      "learning_rate": 3.918266669684351e-05,
      "loss": 0.3586,
      "num_input_tokens_seen": 48511936,
      "step": 1860,
      "train_runtime": 47640.2401,
      "train_tokens_per_second": 1018.297
    },
    {
      "epoch": 0.6180101068676994,
      "grad_norm": 0.7991073727607727,
      "learning_rate": 3.912904181176606e-05,
      "loss": 0.4223,
      "num_input_tokens_seen": 48658504,
      "step": 1865,
      "train_runtime": 47788.8779,
      "train_tokens_per_second": 1018.197
    },
    {
      "epoch": 0.6196669704249855,
      "grad_norm": 0.759983479976654,
      "learning_rate": 3.90753212395083e-05,
      "loss": 0.5293,
      "num_input_tokens_seen": 48787640,
      "step": 1870,
      "train_runtime": 47914.5408,
      "train_tokens_per_second": 1018.222
    },
    {
      "epoch": 0.6213238339822715,
      "grad_norm": 0.7874640226364136,
      "learning_rate": 3.902150534388613e-05,
      "loss": 0.3879,
      "num_input_tokens_seen": 48908000,
      "step": 1875,
      "train_runtime": 48029.0387,
      "train_tokens_per_second": 1018.301
    },
    {
      "epoch": 0.6229806975395576,
      "grad_norm": 0.7168433666229248,
      "learning_rate": 3.896759448936102e-05,
      "loss": 0.4337,
      "num_input_tokens_seen": 49035760,
      "step": 1880,
      "train_runtime": 48151.327,
      "train_tokens_per_second": 1018.368
    },
    {
      "epoch": 0.6246375610968437,
      "grad_norm": 0.7323624491691589,
      "learning_rate": 3.891358904103752e-05,
      "loss": 0.4107,
      "num_input_tokens_seen": 49191504,
      "step": 1885,
      "train_runtime": 48315.1266,
      "train_tokens_per_second": 1018.139
    },
    {
      "epoch": 0.6262944246541298,
      "grad_norm": 0.7390926480293274,
      "learning_rate": 3.8859489364660836e-05,
      "loss": 0.4759,
      "num_input_tokens_seen": 49324768,
      "step": 1890,
      "train_runtime": 48448.1642,
      "train_tokens_per_second": 1018.094
    },
    {
      "epoch": 0.6279512882114158,
      "grad_norm": 0.6986560225486755,
      "learning_rate": 3.880529582661431e-05,
      "loss": 0.4465,
      "num_input_tokens_seen": 49495864,
      "step": 1895,
      "train_runtime": 48620.8358,
      "train_tokens_per_second": 1017.997
    },
    {
      "epoch": 0.6296081517687019,
      "grad_norm": 0.6945990920066833,
      "learning_rate": 3.875100879391692e-05,
      "loss": 0.4496,
      "num_input_tokens_seen": 49622640,
      "step": 1900,
      "train_runtime": 48734.1179,
      "train_tokens_per_second": 1018.232
    },
    {
      "epoch": 0.6312650153259879,
      "grad_norm": 1.0046573877334595,
      "learning_rate": 3.869662863422087e-05,
      "loss": 0.3925,
      "num_input_tokens_seen": 49724800,
      "step": 1905,
      "train_runtime": 48830.4018,
      "train_tokens_per_second": 1018.316
    },
    {
      "epoch": 0.632921878883274,
      "grad_norm": 0.6133820414543152,
      "learning_rate": 3.864215571580903e-05,
      "loss": 0.4538,
      "num_input_tokens_seen": 49877408,
      "step": 1910,
      "train_runtime": 48987.392,
      "train_tokens_per_second": 1018.168
    },
    {
      "epoch": 0.63457874244056,
      "grad_norm": 0.75047367811203,
      "learning_rate": 3.8587590407592473e-05,
      "loss": 0.4187,
      "num_input_tokens_seen": 49979280,
      "step": 1915,
      "train_runtime": 49077.9524,
      "train_tokens_per_second": 1018.365
    },
    {
      "epoch": 0.6362356059978461,
      "grad_norm": 0.6862044930458069,
      "learning_rate": 3.8532933079107966e-05,
      "loss": 0.4087,
      "num_input_tokens_seen": 50093736,
      "step": 1920,
      "train_runtime": 49190.3328,
      "train_tokens_per_second": 1018.365
    },
    {
      "epoch": 0.6378924695551321,
      "grad_norm": 0.7010332345962524,
      "learning_rate": 3.847818410051548e-05,
      "loss": 0.4708,
      "num_input_tokens_seen": 50234160,
      "step": 1925,
      "train_runtime": 49327.32,
      "train_tokens_per_second": 1018.384
    },
    {
      "epoch": 0.6395493331124182,
      "grad_norm": 0.8258742094039917,
      "learning_rate": 3.842334384259568e-05,
      "loss": 0.4118,
      "num_input_tokens_seen": 50362656,
      "step": 1930,
      "train_runtime": 49453.5936,
      "train_tokens_per_second": 1018.382
    },
    {
      "epoch": 0.6412061966697042,
      "grad_norm": 0.745025098323822,
      "learning_rate": 3.836841267674738e-05,
      "loss": 0.4357,
      "num_input_tokens_seen": 50490104,
      "step": 1935,
      "train_runtime": 49579.7869,
      "train_tokens_per_second": 1018.361
    },
    {
      "epoch": 0.6428630602269904,
      "grad_norm": 0.7710812091827393,
      "learning_rate": 3.831339097498509e-05,
      "loss": 0.3858,
      "num_input_tokens_seen": 50635768,
      "step": 1940,
      "train_runtime": 49714.0995,
      "train_tokens_per_second": 1018.539
    },
    {
      "epoch": 0.6445199237842764,
      "grad_norm": 0.7151515483856201,
      "learning_rate": 3.825827910993645e-05,
      "loss": 0.3994,
      "num_input_tokens_seen": 50774144,
      "step": 1945,
      "train_runtime": 49862.1474,
      "train_tokens_per_second": 1018.29
    },
    {
      "epoch": 0.6461767873415625,
      "grad_norm": 0.6541534066200256,
      "learning_rate": 3.820307745483971e-05,
      "loss": 0.4533,
      "num_input_tokens_seen": 50918568,
      "step": 1950,
      "train_runtime": 50007.1978,
      "train_tokens_per_second": 1018.225
    },
    {
      "epoch": 0.6478336508988485,
      "grad_norm": 0.7350675463676453,
      "learning_rate": 3.814778638354122e-05,
      "loss": 0.4282,
      "num_input_tokens_seen": 51025816,
      "step": 1955,
      "train_runtime": 50109.5594,
      "train_tokens_per_second": 1018.285
    },
    {
      "epoch": 0.6494905144561345,
      "grad_norm": 0.9542429447174072,
      "learning_rate": 3.8092406270492895e-05,
      "loss": 0.4535,
      "num_input_tokens_seen": 51125448,
      "step": 1960,
      "train_runtime": 50201.7004,
      "train_tokens_per_second": 1018.401
    },
    {
      "epoch": 0.6511473780134206,
      "grad_norm": 0.6744038462638855,
      "learning_rate": 3.803693749074968e-05,
      "loss": 0.509,
      "num_input_tokens_seen": 51239520,
      "step": 1965,
      "train_runtime": 50305.6229,
      "train_tokens_per_second": 1018.564
    },
    {
      "epoch": 0.6528042415707066,
      "grad_norm": 0.8454577326774597,
      "learning_rate": 3.7981380419966986e-05,
      "loss": 0.5137,
      "num_input_tokens_seen": 51373376,
      "step": 1970,
      "train_runtime": 50442.9352,
      "train_tokens_per_second": 1018.445
    },
    {
      "epoch": 0.6544611051279927,
      "grad_norm": 0.8060897588729858,
      "learning_rate": 3.792573543439817e-05,
      "loss": 0.4856,
      "num_input_tokens_seen": 51515824,
      "step": 1975,
      "train_runtime": 50587.5006,
      "train_tokens_per_second": 1018.351
    },
    {
      "epoch": 0.6561179686852787,
      "grad_norm": 0.7242020964622498,
      "learning_rate": 3.787000291089199e-05,
      "loss": 0.4058,
      "num_input_tokens_seen": 51672872,
      "step": 1980,
      "train_runtime": 50752.7237,
      "train_tokens_per_second": 1018.13
    },
    {
      "epoch": 0.6577748322425648,
      "grad_norm": 0.7966690063476562,
      "learning_rate": 3.781418322689003e-05,
      "loss": 0.4018,
      "num_input_tokens_seen": 51791296,
      "step": 1985,
      "train_runtime": 50859.0343,
      "train_tokens_per_second": 1018.33
    },
    {
      "epoch": 0.6594316957998508,
      "grad_norm": 0.874972939491272,
      "learning_rate": 3.775827676042417e-05,
      "loss": 0.4363,
      "num_input_tokens_seen": 51913024,
      "step": 1990,
      "train_runtime": 50972.3791,
      "train_tokens_per_second": 1018.454
    },
    {
      "epoch": 0.661088559357137,
      "grad_norm": 0.8509311079978943,
      "learning_rate": 3.770228389011401e-05,
      "loss": 0.4847,
      "num_input_tokens_seen": 52039184,
      "step": 1995,
      "train_runtime": 51100.0137,
      "train_tokens_per_second": 1018.379
    },
    {
      "epoch": 0.662745422914423,
      "grad_norm": 0.7952942848205566,
      "learning_rate": 3.76462049951643e-05,
      "loss": 0.4273,
      "num_input_tokens_seen": 52169784,
      "step": 2000,
      "train_runtime": 51230.5794,
      "train_tokens_per_second": 1018.333
    },
    {
      "epoch": 0.6644022864717091,
      "grad_norm": 0.9261770248413086,
      "learning_rate": 3.759004045536239e-05,
      "loss": 0.4631,
      "num_input_tokens_seen": 52312536,
      "step": 2005,
      "train_runtime": 51364.7856,
      "train_tokens_per_second": 1018.451
    },
    {
      "epoch": 0.6660591500289951,
      "grad_norm": 0.6793546080589294,
      "learning_rate": 3.7533790651075665e-05,
      "loss": 0.4306,
      "num_input_tokens_seen": 52421120,
      "step": 2010,
      "train_runtime": 51469.9647,
      "train_tokens_per_second": 1018.48
    },
    {
      "epoch": 0.6677160135862812,
      "grad_norm": 0.6238763332366943,
      "learning_rate": 3.7477455963248915e-05,
      "loss": 0.4079,
      "num_input_tokens_seen": 52534000,
      "step": 2015,
      "train_runtime": 51575.0872,
      "train_tokens_per_second": 1018.593
    },
    {
      "epoch": 0.6693728771435672,
      "grad_norm": 0.766726553440094,
      "learning_rate": 3.742103677340182e-05,
      "loss": 0.5103,
      "num_input_tokens_seen": 52668640,
      "step": 2020,
      "train_runtime": 51700.6465,
      "train_tokens_per_second": 1018.723
    },
    {
      "epoch": 0.6710297407008533,
      "grad_norm": 0.9717079997062683,
      "learning_rate": 3.736453346362634e-05,
      "loss": 0.4507,
      "num_input_tokens_seen": 52782088,
      "step": 2025,
      "train_runtime": 51799.1952,
      "train_tokens_per_second": 1018.975
    },
    {
      "epoch": 0.6726866042581393,
      "grad_norm": 0.7837896943092346,
      "learning_rate": 3.730794641658413e-05,
      "loss": 0.5127,
      "num_input_tokens_seen": 52911856,
      "step": 2030,
      "train_runtime": 51921.9208,
      "train_tokens_per_second": 1019.066
    },
    {
      "epoch": 0.6743434678154254,
      "grad_norm": 0.6400278210639954,
      "learning_rate": 3.725127601550391e-05,
      "loss": 0.4413,
      "num_input_tokens_seen": 53006056,
      "step": 2035,
      "train_runtime": 52009.022,
      "train_tokens_per_second": 1019.17
    },
    {
      "epoch": 0.6760003313727114,
      "grad_norm": 0.5347840189933777,
      "learning_rate": 3.719452264417894e-05,
      "loss": 0.4542,
      "num_input_tokens_seen": 53158248,
      "step": 2040,
      "train_runtime": 52163.2785,
      "train_tokens_per_second": 1019.074
    },
    {
      "epoch": 0.6776571949299975,
      "grad_norm": 0.8710319995880127,
      "learning_rate": 3.7137686686964385e-05,
      "loss": 0.4648,
      "num_input_tokens_seen": 53273032,
      "step": 2045,
      "train_runtime": 52274.2957,
      "train_tokens_per_second": 1019.106
    },
    {
      "epoch": 0.6793140584872835,
      "grad_norm": 0.8236331939697266,
      "learning_rate": 3.70807685287747e-05,
      "loss": 0.479,
      "num_input_tokens_seen": 53401240,
      "step": 2050,
      "train_runtime": 52398.7554,
      "train_tokens_per_second": 1019.132
    },
    {
      "epoch": 0.6809709220445697,
      "grad_norm": 0.6824526786804199,
      "learning_rate": 3.7023768555081035e-05,
      "loss": 0.4111,
      "num_input_tokens_seen": 53539816,
      "step": 2055,
      "train_runtime": 52525.4864,
      "train_tokens_per_second": 1019.311
    },
    {
      "epoch": 0.6826277856018557,
      "grad_norm": 0.7892778515815735,
      "learning_rate": 3.696668715190864e-05,
      "loss": 0.4382,
      "num_input_tokens_seen": 53689720,
      "step": 2060,
      "train_runtime": 52674.0959,
      "train_tokens_per_second": 1019.281
    },
    {
      "epoch": 0.6842846491591418,
      "grad_norm": 0.648322582244873,
      "learning_rate": 3.690952470583422e-05,
      "loss": 0.4174,
      "num_input_tokens_seen": 53831184,
      "step": 2065,
      "train_runtime": 52818.6933,
      "train_tokens_per_second": 1019.169
    },
    {
      "epoch": 0.6859415127164278,
      "grad_norm": 0.6113147139549255,
      "learning_rate": 3.685228160398334e-05,
      "loss": 0.4586,
      "num_input_tokens_seen": 53970696,
      "step": 2070,
      "train_runtime": 52954.1971,
      "train_tokens_per_second": 1019.196
    },
    {
      "epoch": 0.6875983762737139,
      "grad_norm": 0.7623670697212219,
      "learning_rate": 3.67949582340278e-05,
      "loss": 0.4888,
      "num_input_tokens_seen": 54098944,
      "step": 2075,
      "train_runtime": 53079.2252,
      "train_tokens_per_second": 1019.211
    },
    {
      "epoch": 0.6892552398309999,
      "grad_norm": 0.6129375100135803,
      "learning_rate": 3.673755498418299e-05,
      "loss": 0.3736,
      "num_input_tokens_seen": 54248112,
      "step": 2080,
      "train_runtime": 53233.235,
      "train_tokens_per_second": 1019.065
    },
    {
      "epoch": 0.690912103388286,
      "grad_norm": 0.9889698624610901,
      "learning_rate": 3.6680072243205295e-05,
      "loss": 0.4927,
      "num_input_tokens_seen": 54386008,
      "step": 2085,
      "train_runtime": 53380.5899,
      "train_tokens_per_second": 1018.835
    },
    {
      "epoch": 0.692568966945572,
      "grad_norm": 0.6624147295951843,
      "learning_rate": 3.6622510400389436e-05,
      "loss": 0.4957,
      "num_input_tokens_seen": 54511136,
      "step": 2090,
      "train_runtime": 53492.3676,
      "train_tokens_per_second": 1019.045
    },
    {
      "epoch": 0.6942258305028581,
      "grad_norm": 0.7150281667709351,
      "learning_rate": 3.656486984556585e-05,
      "loss": 0.4096,
      "num_input_tokens_seen": 54666912,
      "step": 2095,
      "train_runtime": 53657.8371,
      "train_tokens_per_second": 1018.806
    },
    {
      "epoch": 0.6958826940601441,
      "grad_norm": 0.6723034977912903,
      "learning_rate": 3.6507150969098017e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 54820392,
      "step": 2100,
      "train_runtime": 53810.4589,
      "train_tokens_per_second": 1018.768
    },
    {
      "epoch": 0.6975395576174303,
      "grad_norm": 0.710696280002594,
      "learning_rate": 3.6449354161879866e-05,
      "loss": 0.471,
      "num_input_tokens_seen": 54931024,
      "step": 2105,
      "train_runtime": 53916.0364,
      "train_tokens_per_second": 1018.825
    },
    {
      "epoch": 0.6991964211747163,
      "grad_norm": 0.6247450709342957,
      "learning_rate": 3.6391479815333106e-05,
      "loss": 0.4446,
      "num_input_tokens_seen": 55085920,
      "step": 2110,
      "train_runtime": 54064.1856,
      "train_tokens_per_second": 1018.899
    },
    {
      "epoch": 0.7008532847320024,
      "grad_norm": 0.8954891562461853,
      "learning_rate": 3.6333528321404535e-05,
      "loss": 0.4152,
      "num_input_tokens_seen": 55223584,
      "step": 2115,
      "train_runtime": 54204.7187,
      "train_tokens_per_second": 1018.797
    },
    {
      "epoch": 0.7025101482892884,
      "grad_norm": 0.7879125475883484,
      "learning_rate": 3.6275500072563465e-05,
      "loss": 0.4577,
      "num_input_tokens_seen": 55369776,
      "step": 2120,
      "train_runtime": 54356.7768,
      "train_tokens_per_second": 1018.636
    },
    {
      "epoch": 0.7041670118465745,
      "grad_norm": 0.8038482666015625,
      "learning_rate": 3.621739546179899e-05,
      "loss": 0.4667,
      "num_input_tokens_seen": 55480368,
      "step": 2125,
      "train_runtime": 54466.1548,
      "train_tokens_per_second": 1018.621
    },
    {
      "epoch": 0.7058238754038605,
      "grad_norm": 0.7599883079528809,
      "learning_rate": 3.615921488261737e-05,
      "loss": 0.4422,
      "num_input_tokens_seen": 55626568,
      "step": 2130,
      "train_runtime": 54603.9107,
      "train_tokens_per_second": 1018.729
    },
    {
      "epoch": 0.7074807389611466,
      "grad_norm": 0.7137490510940552,
      "learning_rate": 3.610095872903936e-05,
      "loss": 0.4683,
      "num_input_tokens_seen": 55769664,
      "step": 2135,
      "train_runtime": 54751.6664,
      "train_tokens_per_second": 1018.593
    },
    {
      "epoch": 0.7091376025184326,
      "grad_norm": 0.6420841813087463,
      "learning_rate": 3.604262739559751e-05,
      "loss": 0.4076,
      "num_input_tokens_seen": 55912016,
      "step": 2140,
      "train_runtime": 54894.8956,
      "train_tokens_per_second": 1018.529
    },
    {
      "epoch": 0.7107944660757186,
      "grad_norm": 0.7634726166725159,
      "learning_rate": 3.5984221277333526e-05,
      "loss": 0.4692,
      "num_input_tokens_seen": 56042632,
      "step": 2145,
      "train_runtime": 55023.3257,
      "train_tokens_per_second": 1018.525
    },
    {
      "epoch": 0.7124513296330047,
      "grad_norm": 0.6721476316452026,
      "learning_rate": 3.592574076979559e-05,
      "loss": 0.4037,
      "num_input_tokens_seen": 56168136,
      "step": 2150,
      "train_runtime": 55147.0028,
      "train_tokens_per_second": 1018.517
    },
    {
      "epoch": 0.7141081931902907,
      "grad_norm": 0.6763362884521484,
      "learning_rate": 3.586718626903569e-05,
      "loss": 0.4318,
      "num_input_tokens_seen": 56299744,
      "step": 2155,
      "train_runtime": 55278.2559,
      "train_tokens_per_second": 1018.479
    },
    {
      "epoch": 0.7157650567475768,
      "grad_norm": 0.7523173689842224,
      "learning_rate": 3.5808558171606896e-05,
      "loss": 0.4699,
      "num_input_tokens_seen": 56466432,
      "step": 2160,
      "train_runtime": 55455.4163,
      "train_tokens_per_second": 1018.231
    },
    {
      "epoch": 0.7174219203048628,
      "grad_norm": 0.7404528260231018,
      "learning_rate": 3.574985687456071e-05,
      "loss": 0.4804,
      "num_input_tokens_seen": 56561232,
      "step": 2165,
      "train_runtime": 55542.2988,
      "train_tokens_per_second": 1018.345
    },
    {
      "epoch": 0.719078783862149,
      "grad_norm": 0.7279338836669922,
      "learning_rate": 3.569108277544438e-05,
      "loss": 0.5392,
      "num_input_tokens_seen": 56694184,
      "step": 2170,
      "train_runtime": 55671.277,
      "train_tokens_per_second": 1018.374
    },
    {
      "epoch": 0.720735647419435,
      "grad_norm": 0.5898196697235107,
      "learning_rate": 3.563223627229818e-05,
      "loss": 0.4559,
      "num_input_tokens_seen": 56836392,
      "step": 2175,
      "train_runtime": 55827.1243,
      "train_tokens_per_second": 1018.078
    },
    {
      "epoch": 0.7223925109767211,
      "grad_norm": 0.7183842658996582,
      "learning_rate": 3.557331776365276e-05,
      "loss": 0.4743,
      "num_input_tokens_seen": 56979552,
      "step": 2180,
      "train_runtime": 55979.095,
      "train_tokens_per_second": 1017.872
    },
    {
      "epoch": 0.7240493745340071,
      "grad_norm": 0.5992622375488281,
      "learning_rate": 3.551432764852637e-05,
      "loss": 0.4755,
      "num_input_tokens_seen": 57115888,
      "step": 2185,
      "train_runtime": 56120.7483,
      "train_tokens_per_second": 1017.732
    },
    {
      "epoch": 0.7257062380912932,
      "grad_norm": 0.6807531118392944,
      "learning_rate": 3.545526632642228e-05,
      "loss": 0.4109,
      "num_input_tokens_seen": 57240384,
      "step": 2190,
      "train_runtime": 56241.4077,
      "train_tokens_per_second": 1017.762
    },
    {
      "epoch": 0.7273631016485792,
      "grad_norm": 0.5910036563873291,
      "learning_rate": 3.539613419732591e-05,
      "loss": 0.3413,
      "num_input_tokens_seen": 57358168,
      "step": 2195,
      "train_runtime": 56355.7649,
      "train_tokens_per_second": 1017.787
    },
    {
      "epoch": 0.7290199652058653,
      "grad_norm": 0.8397616744041443,
      "learning_rate": 3.533693166170229e-05,
      "loss": 0.4762,
      "num_input_tokens_seen": 57481824,
      "step": 2200,
      "train_runtime": 56477.9695,
      "train_tokens_per_second": 1017.774
    },
    {
      "epoch": 0.7306768287631513,
      "grad_norm": 0.6472870707511902,
      "learning_rate": 3.5277659120493226e-05,
      "loss": 0.4294,
      "num_input_tokens_seen": 57602600,
      "step": 2205,
      "train_runtime": 56596.0343,
      "train_tokens_per_second": 1017.785
    },
    {
      "epoch": 0.7323336923204374,
      "grad_norm": 0.7097488641738892,
      "learning_rate": 3.521831697511464e-05,
      "loss": 0.4795,
      "num_input_tokens_seen": 57737936,
      "step": 2210,
      "train_runtime": 56733.3653,
      "train_tokens_per_second": 1017.707
    },
    {
      "epoch": 0.7339905558777234,
      "grad_norm": 0.7325347661972046,
      "learning_rate": 3.515890562745384e-05,
      "loss": 0.5018,
      "num_input_tokens_seen": 57870232,
      "step": 2215,
      "train_runtime": 56864.4271,
      "train_tokens_per_second": 1017.688
    },
    {
      "epoch": 0.7356474194350096,
      "grad_norm": 0.6279223561286926,
      "learning_rate": 3.509942547986679e-05,
      "loss": 0.4387,
      "num_input_tokens_seen": 58018256,
      "step": 2220,
      "train_runtime": 57020.8062,
      "train_tokens_per_second": 1017.493
    },
    {
      "epoch": 0.7373042829922956,
      "grad_norm": 0.7424075603485107,
      "learning_rate": 3.503987693517541e-05,
      "loss": 0.4484,
      "num_input_tokens_seen": 58183856,
      "step": 2225,
      "train_runtime": 57198.9195,
      "train_tokens_per_second": 1017.219
    },
    {
      "epoch": 0.7389611465495817,
      "grad_norm": 0.6782631874084473,
      "learning_rate": 3.498026039666481e-05,
      "loss": 0.4534,
      "num_input_tokens_seen": 58317896,
      "step": 2230,
      "train_runtime": 57335.8307,
      "train_tokens_per_second": 1017.128
    },
    {
      "epoch": 0.7406180101068677,
      "grad_norm": 0.8886894583702087,
      "learning_rate": 3.49205762680806e-05,
      "loss": 0.4365,
      "num_input_tokens_seen": 58432600,
      "step": 2235,
      "train_runtime": 57437.4107,
      "train_tokens_per_second": 1017.327
    },
    {
      "epoch": 0.7422748736641538,
      "grad_norm": 0.6596097350120544,
      "learning_rate": 3.4860824953626124e-05,
      "loss": 0.4027,
      "num_input_tokens_seen": 58571992,
      "step": 2240,
      "train_runtime": 57575.1501,
      "train_tokens_per_second": 1017.314
    },
    {
      "epoch": 0.7439317372214398,
      "grad_norm": 0.7005575895309448,
      "learning_rate": 3.480100685795974e-05,
      "loss": 0.429,
      "num_input_tokens_seen": 58696752,
      "step": 2245,
      "train_runtime": 57698.2768,
      "train_tokens_per_second": 1017.305
    },
    {
      "epoch": 0.7455886007787259,
      "grad_norm": 0.7052701115608215,
      "learning_rate": 3.474112238619206e-05,
      "loss": 0.4486,
      "num_input_tokens_seen": 58838496,
      "step": 2250,
      "train_runtime": 57836.5811,
      "train_tokens_per_second": 1017.323
    },
    {
      "epoch": 0.7472454643360119,
      "grad_norm": 0.6943373680114746,
      "learning_rate": 3.468117194388323e-05,
      "loss": 0.47,
      "num_input_tokens_seen": 58940944,
      "step": 2255,
      "train_runtime": 57930.1699,
      "train_tokens_per_second": 1017.448
    },
    {
      "epoch": 0.748902327893298,
      "grad_norm": 0.8867368102073669,
      "learning_rate": 3.462115593704018e-05,
      "loss": 0.348,
      "num_input_tokens_seen": 59084640,
      "step": 2260,
      "train_runtime": 58081.1523,
      "train_tokens_per_second": 1017.277
    },
    {
      "epoch": 0.750559191450584,
      "grad_norm": 0.6269903779029846,
      "learning_rate": 3.456107477211385e-05,
      "loss": 0.4049,
      "num_input_tokens_seen": 59195096,
      "step": 2265,
      "train_runtime": 58184.0152,
      "train_tokens_per_second": 1017.377
    },
    {
      "epoch": 0.7522160550078701,
      "grad_norm": 0.6841360926628113,
      "learning_rate": 3.450092885599647e-05,
      "loss": 0.391,
      "num_input_tokens_seen": 59337576,
      "step": 2270,
      "train_runtime": 58311.3662,
      "train_tokens_per_second": 1017.599
    },
    {
      "epoch": 0.7538729185651561,
      "grad_norm": 0.7364853620529175,
      "learning_rate": 3.444071859601879e-05,
      "loss": 0.4097,
      "num_input_tokens_seen": 59443216,
      "step": 2275,
      "train_runtime": 58404.411,
      "train_tokens_per_second": 1017.786
    },
    {
      "epoch": 0.7555297821224423,
      "grad_norm": 0.7121084332466125,
      "learning_rate": 3.438044439994731e-05,
      "loss": 0.4503,
      "num_input_tokens_seen": 59549512,
      "step": 2280,
      "train_runtime": 58500.7453,
      "train_tokens_per_second": 1017.927
    },
    {
      "epoch": 0.7571866456797283,
      "grad_norm": 0.865563690662384,
      "learning_rate": 3.4320106675981535e-05,
      "loss": 0.4192,
      "num_input_tokens_seen": 59692952,
      "step": 2285,
      "train_runtime": 58639.2875,
      "train_tokens_per_second": 1017.969
    },
    {
      "epoch": 0.7588435092370144,
      "grad_norm": 0.7475072145462036,
      "learning_rate": 3.42597058327512e-05,
      "loss": 0.3951,
      "num_input_tokens_seen": 59839600,
      "step": 2290,
      "train_runtime": 58786.7982,
      "train_tokens_per_second": 1017.909
    },
    {
      "epoch": 0.7605003727943004,
      "grad_norm": 0.724461555480957,
      "learning_rate": 3.419924227931352e-05,
      "loss": 0.4681,
      "num_input_tokens_seen": 59969792,
      "step": 2295,
      "train_runtime": 58915.7515,
      "train_tokens_per_second": 1017.891
    },
    {
      "epoch": 0.7621572363515865,
      "grad_norm": 0.6593635082244873,
      "learning_rate": 3.4138716425150396e-05,
      "loss": 0.4476,
      "num_input_tokens_seen": 60080056,
      "step": 2300,
      "train_runtime": 59028.9849,
      "train_tokens_per_second": 1017.806
    }
  ],
  "logging_steps": 5,
  "max_steps": 6036,
  "num_input_tokens_seen": 60080056,
  "num_train_epochs": 2,
  "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": 4.770484873527706e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}