{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1872,
  "eval_steps": 500,
  "global_step": 185500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0032,
      "grad_norm": 1.7230606079101562,
      "learning_rate": 4.99208e-05,
      "loss": 1.2281,
      "step": 500
    },
    {
      "epoch": 0.0064,
      "grad_norm": 3.655383348464966,
      "learning_rate": 4.9840800000000006e-05,
      "loss": 0.7566,
      "step": 1000
    },
    {
      "epoch": 0.0096,
      "grad_norm": 1.2925927639007568,
      "learning_rate": 4.97608e-05,
      "loss": 0.6764,
      "step": 1500
    },
    {
      "epoch": 0.0128,
      "grad_norm": 1.286004900932312,
      "learning_rate": 4.968080000000001e-05,
      "loss": 0.6304,
      "step": 2000
    },
    {
      "epoch": 0.016,
      "grad_norm": 1.2140214443206787,
      "learning_rate": 4.96008e-05,
      "loss": 0.5981,
      "step": 2500
    },
    {
      "epoch": 0.0192,
      "grad_norm": 1.2525482177734375,
      "learning_rate": 4.95208e-05,
      "loss": 0.5767,
      "step": 3000
    },
    {
      "epoch": 0.0224,
      "grad_norm": 1.2310410737991333,
      "learning_rate": 4.94408e-05,
      "loss": 0.5597,
      "step": 3500
    },
    {
      "epoch": 0.0256,
      "grad_norm": 1.1735206842422485,
      "learning_rate": 4.9360800000000004e-05,
      "loss": 0.5418,
      "step": 4000
    },
    {
      "epoch": 0.0288,
      "grad_norm": 1.114688754081726,
      "learning_rate": 4.9280800000000004e-05,
      "loss": 0.5335,
      "step": 4500
    },
    {
      "epoch": 0.032,
      "grad_norm": 0.8874593377113342,
      "learning_rate": 4.9200800000000005e-05,
      "loss": 0.5237,
      "step": 5000
    },
    {
      "epoch": 0.0352,
      "grad_norm": 1.1261299848556519,
      "learning_rate": 4.91208e-05,
      "loss": 0.5135,
      "step": 5500
    },
    {
      "epoch": 0.0384,
      "grad_norm": 0.9994556307792664,
      "learning_rate": 4.9040800000000007e-05,
      "loss": 0.5059,
      "step": 6000
    },
    {
      "epoch": 0.0416,
      "grad_norm": 1.2349673509597778,
      "learning_rate": 4.89608e-05,
      "loss": 0.4939,
      "step": 6500
    },
    {
      "epoch": 0.0448,
      "grad_norm": 0.9770995378494263,
      "learning_rate": 4.88808e-05,
      "loss": 0.4824,
      "step": 7000
    },
    {
      "epoch": 0.048,
      "grad_norm": 0.981966495513916,
      "learning_rate": 4.88008e-05,
      "loss": 0.4875,
      "step": 7500
    },
    {
      "epoch": 0.0512,
      "grad_norm": 1.0177415609359741,
      "learning_rate": 4.87208e-05,
      "loss": 0.4785,
      "step": 8000
    },
    {
      "epoch": 0.0544,
      "grad_norm": 1.0521667003631592,
      "learning_rate": 4.8640800000000004e-05,
      "loss": 0.4731,
      "step": 8500
    },
    {
      "epoch": 0.0576,
      "grad_norm": 0.8560615181922913,
      "learning_rate": 4.85608e-05,
      "loss": 0.4633,
      "step": 9000
    },
    {
      "epoch": 0.0608,
      "grad_norm": 1.0170217752456665,
      "learning_rate": 4.8480800000000005e-05,
      "loss": 0.4576,
      "step": 9500
    },
    {
      "epoch": 0.064,
      "grad_norm": 0.9891325831413269,
      "learning_rate": 4.84008e-05,
      "loss": 0.4556,
      "step": 10000
    },
    {
      "epoch": 0.0672,
      "grad_norm": 1.0609711408615112,
      "learning_rate": 4.832080000000001e-05,
      "loss": 0.4493,
      "step": 10500
    },
    {
      "epoch": 0.0704,
      "grad_norm": 0.8623799681663513,
      "learning_rate": 4.82408e-05,
      "loss": 0.4459,
      "step": 11000
    },
    {
      "epoch": 0.0736,
      "grad_norm": 0.9587870240211487,
      "learning_rate": 4.81608e-05,
      "loss": 0.4418,
      "step": 11500
    },
    {
      "epoch": 0.0768,
      "grad_norm": 0.8939447999000549,
      "learning_rate": 4.80808e-05,
      "loss": 0.4327,
      "step": 12000
    },
    {
      "epoch": 0.08,
      "grad_norm": 0.9886033535003662,
      "learning_rate": 4.80008e-05,
      "loss": 0.438,
      "step": 12500
    },
    {
      "epoch": 0.0832,
      "grad_norm": 0.9157513976097107,
      "learning_rate": 4.7920800000000004e-05,
      "loss": 0.4323,
      "step": 13000
    },
    {
      "epoch": 0.0864,
      "grad_norm": 0.9085854887962341,
      "learning_rate": 4.7840800000000005e-05,
      "loss": 0.4303,
      "step": 13500
    },
    {
      "epoch": 0.0896,
      "grad_norm": 0.9123984575271606,
      "learning_rate": 4.77608e-05,
      "loss": 0.4247,
      "step": 14000
    },
    {
      "epoch": 0.0928,
      "grad_norm": 0.839026689529419,
      "learning_rate": 4.7680960000000004e-05,
      "loss": 0.4233,
      "step": 14500
    },
    {
      "epoch": 0.096,
      "grad_norm": 0.8110847473144531,
      "learning_rate": 4.760096e-05,
      "loss": 0.4207,
      "step": 15000
    },
    {
      "epoch": 0.0992,
      "grad_norm": 0.8462579250335693,
      "learning_rate": 4.7520960000000005e-05,
      "loss": 0.421,
      "step": 15500
    },
    {
      "epoch": 0.1024,
      "grad_norm": 0.8980106711387634,
      "learning_rate": 4.744096e-05,
      "loss": 0.417,
      "step": 16000
    },
    {
      "epoch": 0.1056,
      "grad_norm": 0.8297702074050903,
      "learning_rate": 4.736096000000001e-05,
      "loss": 0.4139,
      "step": 16500
    },
    {
      "epoch": 0.1088,
      "grad_norm": 0.9856173992156982,
      "learning_rate": 4.728096e-05,
      "loss": 0.419,
      "step": 17000
    },
    {
      "epoch": 0.112,
      "grad_norm": 0.934256911277771,
      "learning_rate": 4.720096e-05,
      "loss": 0.4098,
      "step": 17500
    },
    {
      "epoch": 0.1152,
      "grad_norm": 0.9190649390220642,
      "learning_rate": 4.712096e-05,
      "loss": 0.412,
      "step": 18000
    },
    {
      "epoch": 0.1184,
      "grad_norm": 0.9078772664070129,
      "learning_rate": 4.704096e-05,
      "loss": 0.4043,
      "step": 18500
    },
    {
      "epoch": 0.1216,
      "grad_norm": 1.082939624786377,
      "learning_rate": 4.696112e-05,
      "loss": 0.4045,
      "step": 19000
    },
    {
      "epoch": 0.1248,
      "grad_norm": 0.9159390926361084,
      "learning_rate": 4.688112e-05,
      "loss": 0.4098,
      "step": 19500
    },
    {
      "epoch": 0.128,
      "grad_norm": 0.8420547842979431,
      "learning_rate": 4.680128e-05,
      "loss": 0.4033,
      "step": 20000
    },
    {
      "epoch": 0.1312,
      "grad_norm": 0.7658286094665527,
      "learning_rate": 4.672128e-05,
      "loss": 0.4002,
      "step": 20500
    },
    {
      "epoch": 0.1344,
      "grad_norm": 0.9074057340621948,
      "learning_rate": 4.664128e-05,
      "loss": 0.3964,
      "step": 21000
    },
    {
      "epoch": 0.1376,
      "grad_norm": 0.6065025329589844,
      "learning_rate": 4.656128e-05,
      "loss": 0.3984,
      "step": 21500
    },
    {
      "epoch": 0.1408,
      "grad_norm": 0.7523757219314575,
      "learning_rate": 4.6481280000000004e-05,
      "loss": 0.3959,
      "step": 22000
    },
    {
      "epoch": 0.144,
      "grad_norm": 0.807826042175293,
      "learning_rate": 4.6401280000000004e-05,
      "loss": 0.3921,
      "step": 22500
    },
    {
      "epoch": 0.1472,
      "grad_norm": 0.8530682325363159,
      "learning_rate": 4.632128e-05,
      "loss": 0.4002,
      "step": 23000
    },
    {
      "epoch": 0.1504,
      "grad_norm": 0.8661518692970276,
      "learning_rate": 4.6241280000000006e-05,
      "loss": 0.3856,
      "step": 23500
    },
    {
      "epoch": 0.1536,
      "grad_norm": 0.7473235130310059,
      "learning_rate": 4.616144e-05,
      "loss": 0.3854,
      "step": 24000
    },
    {
      "epoch": 0.1568,
      "grad_norm": 0.7954819202423096,
      "learning_rate": 4.6081440000000005e-05,
      "loss": 0.3871,
      "step": 24500
    },
    {
      "epoch": 0.16,
      "grad_norm": 0.8758727312088013,
      "learning_rate": 4.600144e-05,
      "loss": 0.3842,
      "step": 25000
    },
    {
      "epoch": 0.1632,
      "grad_norm": 0.8430293798446655,
      "learning_rate": 4.592144000000001e-05,
      "loss": 0.3886,
      "step": 25500
    },
    {
      "epoch": 0.1664,
      "grad_norm": 0.6557173728942871,
      "learning_rate": 4.584144e-05,
      "loss": 0.3854,
      "step": 26000
    },
    {
      "epoch": 0.1696,
      "grad_norm": 0.7791888117790222,
      "learning_rate": 4.576144e-05,
      "loss": 0.3796,
      "step": 26500
    },
    {
      "epoch": 0.1728,
      "grad_norm": 0.736084520816803,
      "learning_rate": 4.56816e-05,
      "loss": 0.3806,
      "step": 27000
    },
    {
      "epoch": 0.176,
      "grad_norm": 0.7714269161224365,
      "learning_rate": 4.56016e-05,
      "loss": 0.3781,
      "step": 27500
    },
    {
      "epoch": 0.1792,
      "grad_norm": 0.766144335269928,
      "learning_rate": 4.552176e-05,
      "loss": 0.3766,
      "step": 28000
    },
    {
      "epoch": 0.1824,
      "grad_norm": 0.7035301923751831,
      "learning_rate": 4.544176e-05,
      "loss": 0.3737,
      "step": 28500
    },
    {
      "epoch": 0.1856,
      "grad_norm": 0.7573793530464172,
      "learning_rate": 4.536176e-05,
      "loss": 0.3753,
      "step": 29000
    },
    {
      "epoch": 0.1888,
      "grad_norm": 0.8799508213996887,
      "learning_rate": 4.528176e-05,
      "loss": 0.373,
      "step": 29500
    },
    {
      "epoch": 0.192,
      "grad_norm": 0.8543264269828796,
      "learning_rate": 4.520176e-05,
      "loss": 0.3735,
      "step": 30000
    },
    {
      "epoch": 0.1952,
      "grad_norm": 0.6768947243690491,
      "learning_rate": 4.512176e-05,
      "loss": 0.3697,
      "step": 30500
    },
    {
      "epoch": 0.1984,
      "grad_norm": 0.8239702582359314,
      "learning_rate": 4.504176e-05,
      "loss": 0.3675,
      "step": 31000
    },
    {
      "epoch": 0.2016,
      "grad_norm": 0.8310449123382568,
      "learning_rate": 4.4961760000000004e-05,
      "loss": 0.3695,
      "step": 31500
    },
    {
      "epoch": 0.2048,
      "grad_norm": 0.8459475040435791,
      "learning_rate": 4.488176e-05,
      "loss": 0.3694,
      "step": 32000
    },
    {
      "epoch": 0.208,
      "grad_norm": 0.7346063852310181,
      "learning_rate": 4.4801760000000006e-05,
      "loss": 0.3646,
      "step": 32500
    },
    {
      "epoch": 0.2112,
      "grad_norm": 0.6958354115486145,
      "learning_rate": 4.472176e-05,
      "loss": 0.3704,
      "step": 33000
    },
    {
      "epoch": 0.2144,
      "grad_norm": 0.8244686722755432,
      "learning_rate": 4.464176000000001e-05,
      "loss": 0.3647,
      "step": 33500
    },
    {
      "epoch": 0.2176,
      "grad_norm": 0.7559502124786377,
      "learning_rate": 4.456192e-05,
      "loss": 0.3665,
      "step": 34000
    },
    {
      "epoch": 0.2208,
      "grad_norm": 0.9046504497528076,
      "learning_rate": 4.4481920000000007e-05,
      "loss": 0.3637,
      "step": 34500
    },
    {
      "epoch": 0.224,
      "grad_norm": 0.7771899700164795,
      "learning_rate": 4.440192e-05,
      "loss": 0.3648,
      "step": 35000
    },
    {
      "epoch": 0.2272,
      "grad_norm": 0.6887528300285339,
      "learning_rate": 4.432192e-05,
      "loss": 0.3562,
      "step": 35500
    },
    {
      "epoch": 0.2304,
      "grad_norm": 0.7471407055854797,
      "learning_rate": 4.424192e-05,
      "loss": 0.3639,
      "step": 36000
    },
    {
      "epoch": 0.2336,
      "grad_norm": 0.7198163270950317,
      "learning_rate": 4.416192e-05,
      "loss": 0.3604,
      "step": 36500
    },
    {
      "epoch": 0.2368,
      "grad_norm": 0.7383478879928589,
      "learning_rate": 4.4081920000000004e-05,
      "loss": 0.3592,
      "step": 37000
    },
    {
      "epoch": 0.24,
      "grad_norm": 0.8052579760551453,
      "learning_rate": 4.4001920000000004e-05,
      "loss": 0.3563,
      "step": 37500
    },
    {
      "epoch": 0.2432,
      "grad_norm": 0.7765107154846191,
      "learning_rate": 4.392224e-05,
      "loss": 0.3548,
      "step": 38000
    },
    {
      "epoch": 0.2464,
      "grad_norm": 0.7250288724899292,
      "learning_rate": 4.384224e-05,
      "loss": 0.3605,
      "step": 38500
    },
    {
      "epoch": 0.2496,
      "grad_norm": 0.6914694309234619,
      "learning_rate": 4.376224e-05,
      "loss": 0.3551,
      "step": 39000
    },
    {
      "epoch": 0.2528,
      "grad_norm": 0.6636275053024292,
      "learning_rate": 4.368224e-05,
      "loss": 0.3587,
      "step": 39500
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.710564911365509,
      "learning_rate": 4.360224e-05,
      "loss": 0.3537,
      "step": 40000
    },
    {
      "epoch": 0.2592,
      "grad_norm": 0.6195800304412842,
      "learning_rate": 4.3522240000000004e-05,
      "loss": 0.3537,
      "step": 40500
    },
    {
      "epoch": 0.2624,
      "grad_norm": 0.7131514549255371,
      "learning_rate": 4.34424e-05,
      "loss": 0.3531,
      "step": 41000
    },
    {
      "epoch": 0.2656,
      "grad_norm": 0.6594410538673401,
      "learning_rate": 4.336256e-05,
      "loss": 0.3518,
      "step": 41500
    },
    {
      "epoch": 0.2688,
      "grad_norm": 0.7651230096817017,
      "learning_rate": 4.328256e-05,
      "loss": 0.3516,
      "step": 42000
    },
    {
      "epoch": 0.272,
      "grad_norm": 0.756515622138977,
      "learning_rate": 4.320256e-05,
      "loss": 0.3461,
      "step": 42500
    },
    {
      "epoch": 0.2752,
      "grad_norm": 0.7201528549194336,
      "learning_rate": 4.3122560000000003e-05,
      "loss": 0.3497,
      "step": 43000
    },
    {
      "epoch": 0.2784,
      "grad_norm": 0.7436856031417847,
      "learning_rate": 4.3042560000000004e-05,
      "loss": 0.3505,
      "step": 43500
    },
    {
      "epoch": 0.2816,
      "grad_norm": 0.7914199829101562,
      "learning_rate": 4.2962560000000005e-05,
      "loss": 0.3439,
      "step": 44000
    },
    {
      "epoch": 0.2848,
      "grad_norm": 0.7488194704055786,
      "learning_rate": 4.288256e-05,
      "loss": 0.349,
      "step": 44500
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.8654124736785889,
      "learning_rate": 4.280256e-05,
      "loss": 0.3491,
      "step": 45000
    },
    {
      "epoch": 0.2912,
      "grad_norm": 0.6817401647567749,
      "learning_rate": 4.272272e-05,
      "loss": 0.3447,
      "step": 45500
    },
    {
      "epoch": 0.2944,
      "grad_norm": 0.6439715623855591,
      "learning_rate": 4.2642720000000006e-05,
      "loss": 0.3453,
      "step": 46000
    },
    {
      "epoch": 0.2976,
      "grad_norm": 1.3840138912200928,
      "learning_rate": 4.256272e-05,
      "loss": 0.3445,
      "step": 46500
    },
    {
      "epoch": 0.3008,
      "grad_norm": 0.7245766520500183,
      "learning_rate": 4.248272e-05,
      "loss": 0.3462,
      "step": 47000
    },
    {
      "epoch": 0.304,
      "grad_norm": 0.6877666711807251,
      "learning_rate": 4.240288e-05,
      "loss": 0.3465,
      "step": 47500
    },
    {
      "epoch": 0.3072,
      "grad_norm": 0.8494886159896851,
      "learning_rate": 4.2322880000000006e-05,
      "loss": 0.348,
      "step": 48000
    },
    {
      "epoch": 0.3104,
      "grad_norm": 0.6704971790313721,
      "learning_rate": 4.224288e-05,
      "loss": 0.3403,
      "step": 48500
    },
    {
      "epoch": 0.3136,
      "grad_norm": 0.6239964962005615,
      "learning_rate": 4.216288000000001e-05,
      "loss": 0.3382,
      "step": 49000
    },
    {
      "epoch": 0.3168,
      "grad_norm": 0.7317768335342407,
      "learning_rate": 4.208288e-05,
      "loss": 0.3385,
      "step": 49500
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7397735118865967,
      "learning_rate": 4.200288e-05,
      "loss": 0.3405,
      "step": 50000
    },
    {
      "epoch": 0.3232,
      "grad_norm": 1.1299536228179932,
      "learning_rate": 4.1922880000000003e-05,
      "loss": 0.3431,
      "step": 50500
    },
    {
      "epoch": 0.3264,
      "grad_norm": 0.6406556963920593,
      "learning_rate": 4.184304e-05,
      "loss": 0.3384,
      "step": 51000
    },
    {
      "epoch": 0.3296,
      "grad_norm": 0.8084424734115601,
      "learning_rate": 4.17632e-05,
      "loss": 0.3365,
      "step": 51500
    },
    {
      "epoch": 0.3328,
      "grad_norm": 0.7525010704994202,
      "learning_rate": 4.16832e-05,
      "loss": 0.3399,
      "step": 52000
    },
    {
      "epoch": 0.336,
      "grad_norm": 0.7382110953330994,
      "learning_rate": 4.16032e-05,
      "loss": 0.335,
      "step": 52500
    },
    {
      "epoch": 0.3392,
      "grad_norm": 0.6454793810844421,
      "learning_rate": 4.15232e-05,
      "loss": 0.3354,
      "step": 53000
    },
    {
      "epoch": 0.3424,
      "grad_norm": 0.639664351940155,
      "learning_rate": 4.14432e-05,
      "loss": 0.3371,
      "step": 53500
    },
    {
      "epoch": 0.3456,
      "grad_norm": 0.5574499368667603,
      "learning_rate": 4.1363200000000004e-05,
      "loss": 0.3341,
      "step": 54000
    },
    {
      "epoch": 0.3488,
      "grad_norm": 0.6772671341896057,
      "learning_rate": 4.12832e-05,
      "loss": 0.3331,
      "step": 54500
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.6943195462226868,
      "learning_rate": 4.120336e-05,
      "loss": 0.3365,
      "step": 55000
    },
    {
      "epoch": 0.3552,
      "grad_norm": 0.7460485100746155,
      "learning_rate": 4.112336e-05,
      "loss": 0.3308,
      "step": 55500
    },
    {
      "epoch": 0.3584,
      "grad_norm": 0.7071924805641174,
      "learning_rate": 4.1043360000000005e-05,
      "loss": 0.3312,
      "step": 56000
    },
    {
      "epoch": 0.3616,
      "grad_norm": 0.6678891181945801,
      "learning_rate": 4.0963519999999996e-05,
      "loss": 0.3314,
      "step": 56500
    },
    {
      "epoch": 0.3648,
      "grad_norm": 0.7100914120674133,
      "learning_rate": 4.0883520000000004e-05,
      "loss": 0.3307,
      "step": 57000
    },
    {
      "epoch": 0.368,
      "grad_norm": 0.6085671782493591,
      "learning_rate": 4.080352e-05,
      "loss": 0.3282,
      "step": 57500
    },
    {
      "epoch": 0.3712,
      "grad_norm": 0.6634243130683899,
      "learning_rate": 4.0723520000000005e-05,
      "loss": 0.3321,
      "step": 58000
    },
    {
      "epoch": 0.3744,
      "grad_norm": 0.7203409075737,
      "learning_rate": 4.064352e-05,
      "loss": 0.3318,
      "step": 58500
    },
    {
      "epoch": 0.3776,
      "grad_norm": 0.7934884428977966,
      "learning_rate": 4.056352e-05,
      "loss": 0.3239,
      "step": 59000
    },
    {
      "epoch": 0.3808,
      "grad_norm": 0.8591666221618652,
      "learning_rate": 4.048352e-05,
      "loss": 0.3275,
      "step": 59500
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.6306772232055664,
      "learning_rate": 4.040352e-05,
      "loss": 0.3308,
      "step": 60000
    },
    {
      "epoch": 0.3872,
      "grad_norm": 0.6059302687644958,
      "learning_rate": 4.032352e-05,
      "loss": 0.3266,
      "step": 60500
    },
    {
      "epoch": 0.3904,
      "grad_norm": 0.6875105500221252,
      "learning_rate": 4.024352e-05,
      "loss": 0.3265,
      "step": 61000
    },
    {
      "epoch": 0.3936,
      "grad_norm": 0.6397412419319153,
      "learning_rate": 4.0163520000000004e-05,
      "loss": 0.3268,
      "step": 61500
    },
    {
      "epoch": 0.3968,
      "grad_norm": 0.7801005840301514,
      "learning_rate": 4.0083520000000005e-05,
      "loss": 0.3314,
      "step": 62000
    },
    {
      "epoch": 0.4,
      "grad_norm": 0.6966884136199951,
      "learning_rate": 4.000352e-05,
      "loss": 0.3263,
      "step": 62500
    },
    {
      "epoch": 0.4032,
      "grad_norm": 0.7413304448127747,
      "learning_rate": 3.9923520000000006e-05,
      "loss": 0.3284,
      "step": 63000
    },
    {
      "epoch": 0.4064,
      "grad_norm": 0.7089780569076538,
      "learning_rate": 3.984352e-05,
      "loss": 0.3252,
      "step": 63500
    },
    {
      "epoch": 0.4096,
      "grad_norm": 0.6669878959655762,
      "learning_rate": 3.976352e-05,
      "loss": 0.3239,
      "step": 64000
    },
    {
      "epoch": 0.4128,
      "grad_norm": 0.7352403998374939,
      "learning_rate": 3.968368e-05,
      "loss": 0.3226,
      "step": 64500
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.6916635036468506,
      "learning_rate": 3.9603840000000005e-05,
      "loss": 0.3234,
      "step": 65000
    },
    {
      "epoch": 0.4192,
      "grad_norm": 0.6800302863121033,
      "learning_rate": 3.952384e-05,
      "loss": 0.3224,
      "step": 65500
    },
    {
      "epoch": 0.4224,
      "grad_norm": 0.6685224771499634,
      "learning_rate": 3.9443840000000006e-05,
      "loss": 0.3197,
      "step": 66000
    },
    {
      "epoch": 0.4256,
      "grad_norm": 0.7219159603118896,
      "learning_rate": 3.936384e-05,
      "loss": 0.3185,
      "step": 66500
    },
    {
      "epoch": 0.4288,
      "grad_norm": 0.5928858518600464,
      "learning_rate": 3.928384e-05,
      "loss": 0.3291,
      "step": 67000
    },
    {
      "epoch": 0.432,
      "grad_norm": 0.6616542339324951,
      "learning_rate": 3.920384e-05,
      "loss": 0.3266,
      "step": 67500
    },
    {
      "epoch": 0.4352,
      "grad_norm": 0.5957266092300415,
      "learning_rate": 3.912384e-05,
      "loss": 0.32,
      "step": 68000
    },
    {
      "epoch": 0.4384,
      "grad_norm": 0.6576407551765442,
      "learning_rate": 3.904384e-05,
      "loss": 0.3246,
      "step": 68500
    },
    {
      "epoch": 0.4416,
      "grad_norm": 0.6852056384086609,
      "learning_rate": 3.896416e-05,
      "loss": 0.3268,
      "step": 69000
    },
    {
      "epoch": 0.4448,
      "grad_norm": 0.780893087387085,
      "learning_rate": 3.888416e-05,
      "loss": 0.3229,
      "step": 69500
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.6741476655006409,
      "learning_rate": 3.880416e-05,
      "loss": 0.3188,
      "step": 70000
    },
    {
      "epoch": 0.4512,
      "grad_norm": 0.5919800400733948,
      "learning_rate": 3.872416e-05,
      "loss": 0.3208,
      "step": 70500
    },
    {
      "epoch": 0.4544,
      "grad_norm": 0.6476633548736572,
      "learning_rate": 3.864416e-05,
      "loss": 0.322,
      "step": 71000
    },
    {
      "epoch": 0.4576,
      "grad_norm": 0.5667979717254639,
      "learning_rate": 3.8564159999999996e-05,
      "loss": 0.3151,
      "step": 71500
    },
    {
      "epoch": 0.4608,
      "grad_norm": 0.6126554608345032,
      "learning_rate": 3.8484160000000004e-05,
      "loss": 0.3185,
      "step": 72000
    },
    {
      "epoch": 0.464,
      "grad_norm": 0.7995546460151672,
      "learning_rate": 3.840416e-05,
      "loss": 0.3174,
      "step": 72500
    },
    {
      "epoch": 0.4672,
      "grad_norm": 0.5964981317520142,
      "learning_rate": 3.8324160000000005e-05,
      "loss": 0.3187,
      "step": 73000
    },
    {
      "epoch": 0.4704,
      "grad_norm": 0.7718212008476257,
      "learning_rate": 3.824416e-05,
      "loss": 0.3156,
      "step": 73500
    },
    {
      "epoch": 0.4736,
      "grad_norm": 0.7086686491966248,
      "learning_rate": 3.8164320000000005e-05,
      "loss": 0.3189,
      "step": 74000
    },
    {
      "epoch": 0.4768,
      "grad_norm": 0.7988029718399048,
      "learning_rate": 3.808432e-05,
      "loss": 0.3151,
      "step": 74500
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.6092699766159058,
      "learning_rate": 3.8004320000000006e-05,
      "loss": 0.3153,
      "step": 75000
    },
    {
      "epoch": 0.4832,
      "grad_norm": 0.6181166768074036,
      "learning_rate": 3.792432e-05,
      "loss": 0.3113,
      "step": 75500
    },
    {
      "epoch": 0.4864,
      "grad_norm": 0.5952243208885193,
      "learning_rate": 3.784432e-05,
      "loss": 0.3091,
      "step": 76000
    },
    {
      "epoch": 0.4896,
      "grad_norm": 0.5732501745223999,
      "learning_rate": 3.776432e-05,
      "loss": 0.3169,
      "step": 76500
    },
    {
      "epoch": 0.4928,
      "grad_norm": 0.5866090059280396,
      "learning_rate": 3.768432e-05,
      "loss": 0.3135,
      "step": 77000
    },
    {
      "epoch": 0.496,
      "grad_norm": 0.6748520135879517,
      "learning_rate": 3.760432e-05,
      "loss": 0.3134,
      "step": 77500
    },
    {
      "epoch": 0.4992,
      "grad_norm": 0.5922159552574158,
      "learning_rate": 3.752448e-05,
      "loss": 0.3156,
      "step": 78000
    },
    {
      "epoch": 0.5024,
      "grad_norm": 0.6446545124053955,
      "learning_rate": 3.744448e-05,
      "loss": 0.3171,
      "step": 78500
    },
    {
      "epoch": 0.5056,
      "grad_norm": 0.6506426334381104,
      "learning_rate": 3.736448e-05,
      "loss": 0.3138,
      "step": 79000
    },
    {
      "epoch": 0.5088,
      "grad_norm": 0.6826354265213013,
      "learning_rate": 3.728448e-05,
      "loss": 0.3164,
      "step": 79500
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.6866195797920227,
      "learning_rate": 3.72048e-05,
      "loss": 0.315,
      "step": 80000
    },
    {
      "epoch": 0.5152,
      "grad_norm": 0.5590147376060486,
      "learning_rate": 3.7124960000000005e-05,
      "loss": 0.3094,
      "step": 80500
    },
    {
      "epoch": 0.5184,
      "grad_norm": 0.6728788614273071,
      "learning_rate": 3.704496e-05,
      "loss": 0.3194,
      "step": 81000
    },
    {
      "epoch": 0.5216,
      "grad_norm": 0.6108749508857727,
      "learning_rate": 3.696496000000001e-05,
      "loss": 0.3128,
      "step": 81500
    },
    {
      "epoch": 0.5248,
      "grad_norm": 0.5888856649398804,
      "learning_rate": 3.688496e-05,
      "loss": 0.3121,
      "step": 82000
    },
    {
      "epoch": 0.528,
      "grad_norm": 0.727268397808075,
      "learning_rate": 3.680496e-05,
      "loss": 0.3193,
      "step": 82500
    },
    {
      "epoch": 0.5312,
      "grad_norm": 0.6358634233474731,
      "learning_rate": 3.672496e-05,
      "loss": 0.3092,
      "step": 83000
    },
    {
      "epoch": 0.5344,
      "grad_norm": 0.6482620239257812,
      "learning_rate": 3.664496e-05,
      "loss": 0.3098,
      "step": 83500
    },
    {
      "epoch": 0.5376,
      "grad_norm": 0.5968552827835083,
      "learning_rate": 3.6564960000000004e-05,
      "loss": 0.3108,
      "step": 84000
    },
    {
      "epoch": 0.5408,
      "grad_norm": 0.6621351838111877,
      "learning_rate": 3.6484960000000004e-05,
      "loss": 0.3065,
      "step": 84500
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.5520649552345276,
      "learning_rate": 3.640496e-05,
      "loss": 0.3088,
      "step": 85000
    },
    {
      "epoch": 0.5472,
      "grad_norm": 0.6885005831718445,
      "learning_rate": 3.632496e-05,
      "loss": 0.3075,
      "step": 85500
    },
    {
      "epoch": 0.5504,
      "grad_norm": 0.666653573513031,
      "learning_rate": 3.624512e-05,
      "loss": 0.3113,
      "step": 86000
    },
    {
      "epoch": 0.5536,
      "grad_norm": 0.6344409584999084,
      "learning_rate": 3.6165120000000005e-05,
      "loss": 0.3085,
      "step": 86500
    },
    {
      "epoch": 0.5568,
      "grad_norm": 0.5792534947395325,
      "learning_rate": 3.608512e-05,
      "loss": 0.3132,
      "step": 87000
    },
    {
      "epoch": 0.56,
      "grad_norm": 0.6864989995956421,
      "learning_rate": 3.600512e-05,
      "loss": 0.3079,
      "step": 87500
    },
    {
      "epoch": 0.5632,
      "grad_norm": 0.6077435612678528,
      "learning_rate": 3.592512e-05,
      "loss": 0.3095,
      "step": 88000
    },
    {
      "epoch": 0.5664,
      "grad_norm": 0.7073134779930115,
      "learning_rate": 3.584512e-05,
      "loss": 0.3116,
      "step": 88500
    },
    {
      "epoch": 0.5696,
      "grad_norm": 0.6477733850479126,
      "learning_rate": 3.576512e-05,
      "loss": 0.3062,
      "step": 89000
    },
    {
      "epoch": 0.5728,
      "grad_norm": 0.7786093354225159,
      "learning_rate": 3.568512e-05,
      "loss": 0.3017,
      "step": 89500
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.6447868943214417,
      "learning_rate": 3.560528e-05,
      "loss": 0.3077,
      "step": 90000
    },
    {
      "epoch": 0.5792,
      "grad_norm": 0.6663397550582886,
      "learning_rate": 3.552528e-05,
      "loss": 0.3089,
      "step": 90500
    },
    {
      "epoch": 0.5824,
      "grad_norm": 0.533214807510376,
      "learning_rate": 3.544528e-05,
      "loss": 0.3064,
      "step": 91000
    },
    {
      "epoch": 0.5856,
      "grad_norm": 0.6517444849014282,
      "learning_rate": 3.5365280000000004e-05,
      "loss": 0.3108,
      "step": 91500
    },
    {
      "epoch": 0.5888,
      "grad_norm": 0.7635303735733032,
      "learning_rate": 3.528544e-05,
      "loss": 0.3028,
      "step": 92000
    },
    {
      "epoch": 0.592,
      "grad_norm": 0.6636632680892944,
      "learning_rate": 3.520544e-05,
      "loss": 0.3015,
      "step": 92500
    },
    {
      "epoch": 0.5952,
      "grad_norm": 0.7296783924102783,
      "learning_rate": 3.5125440000000004e-05,
      "loss": 0.305,
      "step": 93000
    },
    {
      "epoch": 0.5984,
      "grad_norm": 0.5089054703712463,
      "learning_rate": 3.50456e-05,
      "loss": 0.3092,
      "step": 93500
    },
    {
      "epoch": 0.6016,
      "grad_norm": 0.6761330366134644,
      "learning_rate": 3.49656e-05,
      "loss": 0.3055,
      "step": 94000
    },
    {
      "epoch": 0.6048,
      "grad_norm": 0.6327843070030212,
      "learning_rate": 3.4885600000000004e-05,
      "loss": 0.3055,
      "step": 94500
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.5940554141998291,
      "learning_rate": 3.48056e-05,
      "loss": 0.3017,
      "step": 95000
    },
    {
      "epoch": 0.6112,
      "grad_norm": 0.516828179359436,
      "learning_rate": 3.4725600000000005e-05,
      "loss": 0.3035,
      "step": 95500
    },
    {
      "epoch": 0.6144,
      "grad_norm": 0.5835782289505005,
      "learning_rate": 3.46456e-05,
      "loss": 0.2978,
      "step": 96000
    },
    {
      "epoch": 0.6176,
      "grad_norm": 0.5978230237960815,
      "learning_rate": 3.456560000000001e-05,
      "loss": 0.301,
      "step": 96500
    },
    {
      "epoch": 0.6208,
      "grad_norm": 0.5460017323493958,
      "learning_rate": 3.44856e-05,
      "loss": 0.3052,
      "step": 97000
    },
    {
      "epoch": 0.624,
      "grad_norm": 0.6875701546669006,
      "learning_rate": 3.44056e-05,
      "loss": 0.3028,
      "step": 97500
    },
    {
      "epoch": 0.6272,
      "grad_norm": 0.5780492424964905,
      "learning_rate": 3.43256e-05,
      "loss": 0.2988,
      "step": 98000
    },
    {
      "epoch": 0.6304,
      "grad_norm": 0.5191554427146912,
      "learning_rate": 3.42456e-05,
      "loss": 0.3052,
      "step": 98500
    },
    {
      "epoch": 0.6336,
      "grad_norm": 0.6811420917510986,
      "learning_rate": 3.416576e-05,
      "loss": 0.3032,
      "step": 99000
    },
    {
      "epoch": 0.6368,
      "grad_norm": 0.6301366686820984,
      "learning_rate": 3.408576e-05,
      "loss": 0.2979,
      "step": 99500
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.5777577757835388,
      "learning_rate": 3.400576e-05,
      "loss": 0.2991,
      "step": 100000
    },
    {
      "epoch": 0.6432,
      "grad_norm": 0.6444558501243591,
      "learning_rate": 3.392592e-05,
      "loss": 0.298,
      "step": 100500
    },
    {
      "epoch": 0.6464,
      "grad_norm": 0.4793080985546112,
      "learning_rate": 3.384592e-05,
      "loss": 0.3014,
      "step": 101000
    },
    {
      "epoch": 0.6496,
      "grad_norm": 0.6691552400588989,
      "learning_rate": 3.376608e-05,
      "loss": 0.3006,
      "step": 101500
    },
    {
      "epoch": 0.6528,
      "grad_norm": 0.6318476796150208,
      "learning_rate": 3.368608e-05,
      "loss": 0.3032,
      "step": 102000
    },
    {
      "epoch": 0.656,
      "grad_norm": 0.5805894136428833,
      "learning_rate": 3.360608e-05,
      "loss": 0.3014,
      "step": 102500
    },
    {
      "epoch": 0.6592,
      "grad_norm": 0.5658220648765564,
      "learning_rate": 3.352608e-05,
      "loss": 0.3,
      "step": 103000
    },
    {
      "epoch": 0.6624,
      "grad_norm": 0.6117516160011292,
      "learning_rate": 3.3446080000000004e-05,
      "loss": 0.3014,
      "step": 103500
    },
    {
      "epoch": 0.6656,
      "grad_norm": 0.6763502359390259,
      "learning_rate": 3.336608e-05,
      "loss": 0.3043,
      "step": 104000
    },
    {
      "epoch": 0.6688,
      "grad_norm": 0.6046746969223022,
      "learning_rate": 3.3286080000000005e-05,
      "loss": 0.2965,
      "step": 104500
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.7453213930130005,
      "learning_rate": 3.320608e-05,
      "loss": 0.2964,
      "step": 105000
    },
    {
      "epoch": 0.6752,
      "grad_norm": 0.6010546088218689,
      "learning_rate": 3.3126080000000007e-05,
      "loss": 0.2975,
      "step": 105500
    },
    {
      "epoch": 0.6784,
      "grad_norm": 0.7377296686172485,
      "learning_rate": 3.304608e-05,
      "loss": 0.2993,
      "step": 106000
    },
    {
      "epoch": 0.6816,
      "grad_norm": 0.6612259745597839,
      "learning_rate": 3.2966240000000006e-05,
      "loss": 0.298,
      "step": 106500
    },
    {
      "epoch": 0.6848,
      "grad_norm": 0.6570013165473938,
      "learning_rate": 3.288624e-05,
      "loss": 0.296,
      "step": 107000
    },
    {
      "epoch": 0.688,
      "grad_norm": 0.633602499961853,
      "learning_rate": 3.280624e-05,
      "loss": 0.2989,
      "step": 107500
    },
    {
      "epoch": 0.6912,
      "grad_norm": 0.5594373345375061,
      "learning_rate": 3.272624e-05,
      "loss": 0.2977,
      "step": 108000
    },
    {
      "epoch": 0.6944,
      "grad_norm": 0.5643302202224731,
      "learning_rate": 3.264624e-05,
      "loss": 0.2941,
      "step": 108500
    },
    {
      "epoch": 0.6976,
      "grad_norm": 0.5127794146537781,
      "learning_rate": 3.256624e-05,
      "loss": 0.2953,
      "step": 109000
    },
    {
      "epoch": 0.7008,
      "grad_norm": 0.6273791790008545,
      "learning_rate": 3.24864e-05,
      "loss": 0.2944,
      "step": 109500
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.5089157223701477,
      "learning_rate": 3.24064e-05,
      "loss": 0.3,
      "step": 110000
    },
    {
      "epoch": 0.7072,
      "grad_norm": 0.5816791653633118,
      "learning_rate": 3.232656e-05,
      "loss": 0.2957,
      "step": 110500
    },
    {
      "epoch": 0.7104,
      "grad_norm": 0.6407476663589478,
      "learning_rate": 3.224656e-05,
      "loss": 0.2974,
      "step": 111000
    },
    {
      "epoch": 0.7136,
      "grad_norm": 0.46444937586784363,
      "learning_rate": 3.216656e-05,
      "loss": 0.2969,
      "step": 111500
    },
    {
      "epoch": 0.7168,
      "grad_norm": 0.4997446835041046,
      "learning_rate": 3.2086559999999996e-05,
      "loss": 0.2966,
      "step": 112000
    },
    {
      "epoch": 0.72,
      "grad_norm": 0.6996490359306335,
      "learning_rate": 3.2006560000000003e-05,
      "loss": 0.2965,
      "step": 112500
    },
    {
      "epoch": 0.7232,
      "grad_norm": 0.5806016325950623,
      "learning_rate": 3.192672e-05,
      "loss": 0.2952,
      "step": 113000
    },
    {
      "epoch": 0.7264,
      "grad_norm": 0.6140916347503662,
      "learning_rate": 3.184672e-05,
      "loss": 0.2995,
      "step": 113500
    },
    {
      "epoch": 0.7296,
      "grad_norm": 0.45879319310188293,
      "learning_rate": 3.1766719999999997e-05,
      "loss": 0.292,
      "step": 114000
    },
    {
      "epoch": 0.7328,
      "grad_norm": 0.6141937971115112,
      "learning_rate": 3.1686720000000004e-05,
      "loss": 0.2945,
      "step": 114500
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.6565462946891785,
      "learning_rate": 3.160672e-05,
      "loss": 0.2982,
      "step": 115000
    },
    {
      "epoch": 0.7392,
      "grad_norm": 0.5997145175933838,
      "learning_rate": 3.1526720000000006e-05,
      "loss": 0.2957,
      "step": 115500
    },
    {
      "epoch": 0.7424,
      "grad_norm": 0.736965537071228,
      "learning_rate": 3.144672e-05,
      "loss": 0.2953,
      "step": 116000
    },
    {
      "epoch": 0.7456,
      "grad_norm": 0.6587550640106201,
      "learning_rate": 3.136672e-05,
      "loss": 0.2917,
      "step": 116500
    },
    {
      "epoch": 0.7488,
      "grad_norm": 0.7265971302986145,
      "learning_rate": 3.128672e-05,
      "loss": 0.2908,
      "step": 117000
    },
    {
      "epoch": 0.752,
      "grad_norm": 0.6158114075660706,
      "learning_rate": 3.120672e-05,
      "loss": 0.2916,
      "step": 117500
    },
    {
      "epoch": 0.7552,
      "grad_norm": 0.6521216034889221,
      "learning_rate": 3.112672e-05,
      "loss": 0.2947,
      "step": 118000
    },
    {
      "epoch": 0.7584,
      "grad_norm": 0.5868868231773376,
      "learning_rate": 3.1046720000000004e-05,
      "loss": 0.2919,
      "step": 118500
    },
    {
      "epoch": 0.7616,
      "grad_norm": 0.6495432257652283,
      "learning_rate": 3.096672e-05,
      "loss": 0.2974,
      "step": 119000
    },
    {
      "epoch": 0.7648,
      "grad_norm": 0.6204816102981567,
      "learning_rate": 3.0886720000000005e-05,
      "loss": 0.2945,
      "step": 119500
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.6333968639373779,
      "learning_rate": 3.080672e-05,
      "loss": 0.292,
      "step": 120000
    },
    {
      "epoch": 0.7712,
      "grad_norm": 0.5613961815834045,
      "learning_rate": 3.0726880000000004e-05,
      "loss": 0.2938,
      "step": 120500
    },
    {
      "epoch": 0.7744,
      "grad_norm": 0.6623988151550293,
      "learning_rate": 3.064688e-05,
      "loss": 0.2954,
      "step": 121000
    },
    {
      "epoch": 0.7776,
      "grad_norm": 0.6134264469146729,
      "learning_rate": 3.0566880000000006e-05,
      "loss": 0.2915,
      "step": 121500
    },
    {
      "epoch": 0.7808,
      "grad_norm": 0.6159347891807556,
      "learning_rate": 3.048688e-05,
      "loss": 0.2887,
      "step": 122000
    },
    {
      "epoch": 0.784,
      "grad_norm": 0.6079424023628235,
      "learning_rate": 3.0407040000000005e-05,
      "loss": 0.2915,
      "step": 122500
    },
    {
      "epoch": 0.7872,
      "grad_norm": 0.7703385353088379,
      "learning_rate": 3.0327040000000002e-05,
      "loss": 0.2901,
      "step": 123000
    },
    {
      "epoch": 0.7904,
      "grad_norm": 0.5626256465911865,
      "learning_rate": 3.024704e-05,
      "loss": 0.2938,
      "step": 123500
    },
    {
      "epoch": 0.7936,
      "grad_norm": 0.554914653301239,
      "learning_rate": 3.016704e-05,
      "loss": 0.2913,
      "step": 124000
    },
    {
      "epoch": 0.7968,
      "grad_norm": 0.6610060930252075,
      "learning_rate": 3.008704e-05,
      "loss": 0.2912,
      "step": 124500
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.6194009780883789,
      "learning_rate": 3.0007040000000002e-05,
      "loss": 0.2901,
      "step": 125000
    },
    {
      "epoch": 0.8032,
      "grad_norm": 0.7150211930274963,
      "learning_rate": 2.992704e-05,
      "loss": 0.2895,
      "step": 125500
    },
    {
      "epoch": 0.8064,
      "grad_norm": 0.6945148706436157,
      "learning_rate": 2.9847040000000003e-05,
      "loss": 0.2878,
      "step": 126000
    },
    {
      "epoch": 0.8096,
      "grad_norm": 0.6546908617019653,
      "learning_rate": 2.9767200000000002e-05,
      "loss": 0.287,
      "step": 126500
    },
    {
      "epoch": 0.8128,
      "grad_norm": 0.535040020942688,
      "learning_rate": 2.9687360000000004e-05,
      "loss": 0.2901,
      "step": 127000
    },
    {
      "epoch": 0.816,
      "grad_norm": 0.6062806844711304,
      "learning_rate": 2.960736e-05,
      "loss": 0.2862,
      "step": 127500
    },
    {
      "epoch": 0.8192,
      "grad_norm": 0.6202298998832703,
      "learning_rate": 2.9527360000000005e-05,
      "loss": 0.2884,
      "step": 128000
    },
    {
      "epoch": 0.8224,
      "grad_norm": 0.5966545343399048,
      "learning_rate": 2.9447360000000003e-05,
      "loss": 0.2877,
      "step": 128500
    },
    {
      "epoch": 0.8256,
      "grad_norm": 0.5024796724319458,
      "learning_rate": 2.936736e-05,
      "loss": 0.2882,
      "step": 129000
    },
    {
      "epoch": 0.8288,
      "grad_norm": 0.5895559191703796,
      "learning_rate": 2.9287520000000002e-05,
      "loss": 0.288,
      "step": 129500
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.9302066564559937,
      "learning_rate": 2.920752e-05,
      "loss": 0.286,
      "step": 130000
    },
    {
      "epoch": 0.8352,
      "grad_norm": 0.573466956615448,
      "learning_rate": 2.9127520000000003e-05,
      "loss": 0.2848,
      "step": 130500
    },
    {
      "epoch": 0.8384,
      "grad_norm": 0.5901783108711243,
      "learning_rate": 2.904768e-05,
      "loss": 0.2883,
      "step": 131000
    },
    {
      "epoch": 0.8416,
      "grad_norm": 0.7780030369758606,
      "learning_rate": 2.8967680000000002e-05,
      "loss": 0.2914,
      "step": 131500
    },
    {
      "epoch": 0.8448,
      "grad_norm": 0.6630533933639526,
      "learning_rate": 2.888768e-05,
      "loss": 0.2878,
      "step": 132000
    },
    {
      "epoch": 0.848,
      "grad_norm": 0.6001667976379395,
      "learning_rate": 2.8807680000000004e-05,
      "loss": 0.2818,
      "step": 132500
    },
    {
      "epoch": 0.8512,
      "grad_norm": 0.6324682831764221,
      "learning_rate": 2.872768e-05,
      "loss": 0.2849,
      "step": 133000
    },
    {
      "epoch": 0.8544,
      "grad_norm": 0.6814092993736267,
      "learning_rate": 2.864768e-05,
      "loss": 0.288,
      "step": 133500
    },
    {
      "epoch": 0.8576,
      "grad_norm": 0.651709794998169,
      "learning_rate": 2.8567680000000003e-05,
      "loss": 0.2872,
      "step": 134000
    },
    {
      "epoch": 0.8608,
      "grad_norm": 0.5912330746650696,
      "learning_rate": 2.848768e-05,
      "loss": 0.2824,
      "step": 134500
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.5821974277496338,
      "learning_rate": 2.8407680000000004e-05,
      "loss": 0.2853,
      "step": 135000
    },
    {
      "epoch": 0.8672,
      "grad_norm": 0.6262611150741577,
      "learning_rate": 2.832784e-05,
      "loss": 0.2848,
      "step": 135500
    },
    {
      "epoch": 0.8704,
      "grad_norm": 0.5360976457595825,
      "learning_rate": 2.8247840000000004e-05,
      "loss": 0.2869,
      "step": 136000
    },
    {
      "epoch": 0.8736,
      "grad_norm": 0.6523284912109375,
      "learning_rate": 2.816784e-05,
      "loss": 0.2792,
      "step": 136500
    },
    {
      "epoch": 0.8768,
      "grad_norm": 0.6329330205917358,
      "learning_rate": 2.808784e-05,
      "loss": 0.2865,
      "step": 137000
    },
    {
      "epoch": 0.88,
      "grad_norm": 0.6053124666213989,
      "learning_rate": 2.8007840000000003e-05,
      "loss": 0.2844,
      "step": 137500
    },
    {
      "epoch": 0.8832,
      "grad_norm": 0.6887571811676025,
      "learning_rate": 2.7927999999999998e-05,
      "loss": 0.288,
      "step": 138000
    },
    {
      "epoch": 0.8864,
      "grad_norm": 0.7047476172447205,
      "learning_rate": 2.7848000000000002e-05,
      "loss": 0.2877,
      "step": 138500
    },
    {
      "epoch": 0.8896,
      "grad_norm": 0.598227858543396,
      "learning_rate": 2.7768e-05,
      "loss": 0.2867,
      "step": 139000
    },
    {
      "epoch": 0.8928,
      "grad_norm": 0.5094701051712036,
      "learning_rate": 2.7688000000000003e-05,
      "loss": 0.2832,
      "step": 139500
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.5749739408493042,
      "learning_rate": 2.7608e-05,
      "loss": 0.2821,
      "step": 140000
    },
    {
      "epoch": 0.8992,
      "grad_norm": 0.4442578852176666,
      "learning_rate": 2.7528320000000003e-05,
      "loss": 0.282,
      "step": 140500
    },
    {
      "epoch": 0.9024,
      "grad_norm": 0.5418574213981628,
      "learning_rate": 2.744832e-05,
      "loss": 0.2816,
      "step": 141000
    },
    {
      "epoch": 0.9056,
      "grad_norm": 0.5984327793121338,
      "learning_rate": 2.736832e-05,
      "loss": 0.285,
      "step": 141500
    },
    {
      "epoch": 0.9088,
      "grad_norm": 0.6572843194007874,
      "learning_rate": 2.728832e-05,
      "loss": 0.2817,
      "step": 142000
    },
    {
      "epoch": 0.912,
      "grad_norm": 0.590993344783783,
      "learning_rate": 2.7208320000000003e-05,
      "loss": 0.288,
      "step": 142500
    },
    {
      "epoch": 0.9152,
      "grad_norm": 0.6096624135971069,
      "learning_rate": 2.712832e-05,
      "loss": 0.2861,
      "step": 143000
    },
    {
      "epoch": 0.9184,
      "grad_norm": 0.5189167261123657,
      "learning_rate": 2.7048319999999998e-05,
      "loss": 0.2857,
      "step": 143500
    },
    {
      "epoch": 0.9216,
      "grad_norm": 0.5812899470329285,
      "learning_rate": 2.6968320000000002e-05,
      "loss": 0.2888,
      "step": 144000
    },
    {
      "epoch": 0.9248,
      "grad_norm": 0.515201210975647,
      "learning_rate": 2.688832e-05,
      "loss": 0.2791,
      "step": 144500
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.6398504972457886,
      "learning_rate": 2.6808320000000004e-05,
      "loss": 0.282,
      "step": 145000
    },
    {
      "epoch": 0.9312,
      "grad_norm": 0.5990891456604004,
      "learning_rate": 2.672832e-05,
      "loss": 0.28,
      "step": 145500
    },
    {
      "epoch": 0.9344,
      "grad_norm": 0.5883029699325562,
      "learning_rate": 2.664832e-05,
      "loss": 0.2777,
      "step": 146000
    },
    {
      "epoch": 0.9376,
      "grad_norm": 0.6432376503944397,
      "learning_rate": 2.656848e-05,
      "loss": 0.2804,
      "step": 146500
    },
    {
      "epoch": 0.9408,
      "grad_norm": 0.5375948548316956,
      "learning_rate": 2.6488479999999997e-05,
      "loss": 0.2807,
      "step": 147000
    },
    {
      "epoch": 0.944,
      "grad_norm": 0.6207411885261536,
      "learning_rate": 2.6408640000000003e-05,
      "loss": 0.283,
      "step": 147500
    },
    {
      "epoch": 0.9472,
      "grad_norm": 0.5854378342628479,
      "learning_rate": 2.632864e-05,
      "loss": 0.2854,
      "step": 148000
    },
    {
      "epoch": 0.9504,
      "grad_norm": 0.5260078310966492,
      "learning_rate": 2.6248800000000002e-05,
      "loss": 0.2836,
      "step": 148500
    },
    {
      "epoch": 0.9536,
      "grad_norm": 0.6284717917442322,
      "learning_rate": 2.61688e-05,
      "loss": 0.2824,
      "step": 149000
    },
    {
      "epoch": 0.9568,
      "grad_norm": 0.6092182397842407,
      "learning_rate": 2.608896e-05,
      "loss": 0.2804,
      "step": 149500
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.6028911471366882,
      "learning_rate": 2.600896e-05,
      "loss": 0.281,
      "step": 150000
    },
    {
      "epoch": 0.9632,
      "grad_norm": 0.5008478164672852,
      "learning_rate": 2.5928960000000003e-05,
      "loss": 0.277,
      "step": 150500
    },
    {
      "epoch": 0.9664,
      "grad_norm": 0.5233867168426514,
      "learning_rate": 2.584896e-05,
      "loss": 0.2807,
      "step": 151000
    },
    {
      "epoch": 0.9696,
      "grad_norm": 0.5762408375740051,
      "learning_rate": 2.5768960000000004e-05,
      "loss": 0.2831,
      "step": 151500
    },
    {
      "epoch": 0.9728,
      "grad_norm": 0.6097844243049622,
      "learning_rate": 2.568896e-05,
      "loss": 0.2803,
      "step": 152000
    },
    {
      "epoch": 0.976,
      "grad_norm": 0.6696804761886597,
      "learning_rate": 2.560896e-05,
      "loss": 0.2742,
      "step": 152500
    },
    {
      "epoch": 0.9792,
      "grad_norm": 0.6028556823730469,
      "learning_rate": 2.5528960000000003e-05,
      "loss": 0.282,
      "step": 153000
    },
    {
      "epoch": 0.9824,
      "grad_norm": 0.6651898622512817,
      "learning_rate": 2.544896e-05,
      "loss": 0.2849,
      "step": 153500
    },
    {
      "epoch": 0.9856,
      "grad_norm": 0.5219380855560303,
      "learning_rate": 2.536896e-05,
      "loss": 0.2785,
      "step": 154000
    },
    {
      "epoch": 0.9888,
      "grad_norm": 0.6161176562309265,
      "learning_rate": 2.5288960000000002e-05,
      "loss": 0.2808,
      "step": 154500
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.7915316224098206,
      "learning_rate": 2.5208960000000003e-05,
      "loss": 0.2777,
      "step": 155000
    },
    {
      "epoch": 0.9952,
      "grad_norm": 0.7261882424354553,
      "learning_rate": 2.512896e-05,
      "loss": 0.2767,
      "step": 155500
    },
    {
      "epoch": 0.9984,
      "grad_norm": 0.5452406406402588,
      "learning_rate": 2.5048959999999997e-05,
      "loss": 0.2764,
      "step": 156000
    },
    {
      "epoch": 1.0016,
      "grad_norm": 0.642181396484375,
      "learning_rate": 2.4969120000000003e-05,
      "loss": 0.2746,
      "step": 156500
    },
    {
      "epoch": 1.0048,
      "grad_norm": 0.5900291204452515,
      "learning_rate": 2.4889120000000003e-05,
      "loss": 0.2721,
      "step": 157000
    },
    {
      "epoch": 1.008,
      "grad_norm": 0.5960043668746948,
      "learning_rate": 2.480912e-05,
      "loss": 0.265,
      "step": 157500
    },
    {
      "epoch": 1.0112,
      "grad_norm": 0.582115650177002,
      "learning_rate": 2.472912e-05,
      "loss": 0.2673,
      "step": 158000
    },
    {
      "epoch": 1.0144,
      "grad_norm": 0.552392303943634,
      "learning_rate": 2.464912e-05,
      "loss": 0.2663,
      "step": 158500
    },
    {
      "epoch": 1.0176,
      "grad_norm": 0.5585765242576599,
      "learning_rate": 2.456912e-05,
      "loss": 0.2688,
      "step": 159000
    },
    {
      "epoch": 1.0208,
      "grad_norm": 0.6049332022666931,
      "learning_rate": 2.448912e-05,
      "loss": 0.266,
      "step": 159500
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.5749877095222473,
      "learning_rate": 2.440912e-05,
      "loss": 0.2689,
      "step": 160000
    },
    {
      "epoch": 1.0272,
      "grad_norm": 0.5832675695419312,
      "learning_rate": 2.4329120000000002e-05,
      "loss": 0.2703,
      "step": 160500
    },
    {
      "epoch": 1.0304,
      "grad_norm": 0.8549031019210815,
      "learning_rate": 2.424928e-05,
      "loss": 0.2623,
      "step": 161000
    },
    {
      "epoch": 1.0336,
      "grad_norm": 0.5572855472564697,
      "learning_rate": 2.416928e-05,
      "loss": 0.2711,
      "step": 161500
    },
    {
      "epoch": 1.0368,
      "grad_norm": 0.6818140745162964,
      "learning_rate": 2.408928e-05,
      "loss": 0.2652,
      "step": 162000
    },
    {
      "epoch": 1.04,
      "grad_norm": 0.6900683045387268,
      "learning_rate": 2.400928e-05,
      "loss": 0.2669,
      "step": 162500
    },
    {
      "epoch": 1.0432,
      "grad_norm": 0.6015618443489075,
      "learning_rate": 2.392944e-05,
      "loss": 0.2654,
      "step": 163000
    },
    {
      "epoch": 1.0464,
      "grad_norm": 0.5343177318572998,
      "learning_rate": 2.3849440000000002e-05,
      "loss": 0.2656,
      "step": 163500
    },
    {
      "epoch": 1.0496,
      "grad_norm": 0.6130079627037048,
      "learning_rate": 2.3769440000000003e-05,
      "loss": 0.2592,
      "step": 164000
    },
    {
      "epoch": 1.0528,
      "grad_norm": 0.7150599956512451,
      "learning_rate": 2.368944e-05,
      "loss": 0.2634,
      "step": 164500
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.6321354508399963,
      "learning_rate": 2.360944e-05,
      "loss": 0.2683,
      "step": 165000
    },
    {
      "epoch": 1.0592,
      "grad_norm": 0.6234462857246399,
      "learning_rate": 2.352976e-05,
      "loss": 0.2628,
      "step": 165500
    },
    {
      "epoch": 1.0624,
      "grad_norm": 0.6542537808418274,
      "learning_rate": 2.344976e-05,
      "loss": 0.2618,
      "step": 166000
    },
    {
      "epoch": 1.0656,
      "grad_norm": 0.6302633881568909,
      "learning_rate": 2.3369760000000002e-05,
      "loss": 0.2661,
      "step": 166500
    },
    {
      "epoch": 1.0688,
      "grad_norm": 0.5890353322029114,
      "learning_rate": 2.3289760000000002e-05,
      "loss": 0.2646,
      "step": 167000
    },
    {
      "epoch": 1.072,
      "grad_norm": 0.6490179300308228,
      "learning_rate": 2.320976e-05,
      "loss": 0.2635,
      "step": 167500
    },
    {
      "epoch": 1.0752,
      "grad_norm": 0.648162305355072,
      "learning_rate": 2.312976e-05,
      "loss": 0.2646,
      "step": 168000
    },
    {
      "epoch": 1.0784,
      "grad_norm": 0.675680935382843,
      "learning_rate": 2.304976e-05,
      "loss": 0.2626,
      "step": 168500
    },
    {
      "epoch": 1.0816,
      "grad_norm": 0.6192341446876526,
      "learning_rate": 2.2969760000000002e-05,
      "loss": 0.2641,
      "step": 169000
    },
    {
      "epoch": 1.0848,
      "grad_norm": 0.7046379446983337,
      "learning_rate": 2.288992e-05,
      "loss": 0.2643,
      "step": 169500
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.5477197170257568,
      "learning_rate": 2.280992e-05,
      "loss": 0.265,
      "step": 170000
    },
    {
      "epoch": 1.0912,
      "grad_norm": 0.5775583982467651,
      "learning_rate": 2.2729920000000002e-05,
      "loss": 0.2645,
      "step": 170500
    },
    {
      "epoch": 1.0944,
      "grad_norm": 0.6389047503471375,
      "learning_rate": 2.2649920000000003e-05,
      "loss": 0.2634,
      "step": 171000
    },
    {
      "epoch": 1.0976,
      "grad_norm": 0.6169374585151672,
      "learning_rate": 2.256992e-05,
      "loss": 0.2642,
      "step": 171500
    },
    {
      "epoch": 1.1008,
      "grad_norm": 0.5913782715797424,
      "learning_rate": 2.2490080000000002e-05,
      "loss": 0.2658,
      "step": 172000
    },
    {
      "epoch": 1.104,
      "grad_norm": 0.7547928690910339,
      "learning_rate": 2.241008e-05,
      "loss": 0.2674,
      "step": 172500
    },
    {
      "epoch": 1.1072,
      "grad_norm": 0.6277585625648499,
      "learning_rate": 2.233024e-05,
      "loss": 0.2686,
      "step": 173000
    },
    {
      "epoch": 1.1104,
      "grad_norm": 0.6357282996177673,
      "learning_rate": 2.225024e-05,
      "loss": 0.2639,
      "step": 173500
    },
    {
      "epoch": 1.1136,
      "grad_norm": 0.5262208580970764,
      "learning_rate": 2.2170400000000004e-05,
      "loss": 0.2641,
      "step": 174000
    },
    {
      "epoch": 1.1168,
      "grad_norm": 0.6878075003623962,
      "learning_rate": 2.20904e-05,
      "loss": 0.2654,
      "step": 174500
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.5332186222076416,
      "learning_rate": 2.2010400000000002e-05,
      "loss": 0.2638,
      "step": 175000
    },
    {
      "epoch": 1.1232,
      "grad_norm": 0.5562476515769958,
      "learning_rate": 2.19304e-05,
      "loss": 0.2648,
      "step": 175500
    },
    {
      "epoch": 1.1264,
      "grad_norm": 0.5924221277236938,
      "learning_rate": 2.18504e-05,
      "loss": 0.2627,
      "step": 176000
    },
    {
      "epoch": 1.1296,
      "grad_norm": 0.5250386595726013,
      "learning_rate": 2.17704e-05,
      "loss": 0.2619,
      "step": 176500
    },
    {
      "epoch": 1.1328,
      "grad_norm": 0.7426069378852844,
      "learning_rate": 2.16904e-05,
      "loss": 0.2628,
      "step": 177000
    },
    {
      "epoch": 1.1360000000000001,
      "grad_norm": 0.4925951063632965,
      "learning_rate": 2.16104e-05,
      "loss": 0.2661,
      "step": 177500
    },
    {
      "epoch": 1.1392,
      "grad_norm": 0.5707270503044128,
      "learning_rate": 2.15304e-05,
      "loss": 0.2622,
      "step": 178000
    },
    {
      "epoch": 1.1424,
      "grad_norm": 0.5793021321296692,
      "learning_rate": 2.14504e-05,
      "loss": 0.2671,
      "step": 178500
    },
    {
      "epoch": 1.1456,
      "grad_norm": 0.5736916661262512,
      "learning_rate": 2.13704e-05,
      "loss": 0.2648,
      "step": 179000
    },
    {
      "epoch": 1.1488,
      "grad_norm": 0.588550329208374,
      "learning_rate": 2.129056e-05,
      "loss": 0.2641,
      "step": 179500
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.5504462122917175,
      "learning_rate": 2.121056e-05,
      "loss": 0.2643,
      "step": 180000
    },
    {
      "epoch": 1.1552,
      "grad_norm": 0.5439949035644531,
      "learning_rate": 2.113056e-05,
      "loss": 0.2639,
      "step": 180500
    },
    {
      "epoch": 1.1584,
      "grad_norm": 0.6882042288780212,
      "learning_rate": 2.105056e-05,
      "loss": 0.2595,
      "step": 181000
    },
    {
      "epoch": 1.1616,
      "grad_norm": 0.6735561490058899,
      "learning_rate": 2.097056e-05,
      "loss": 0.2624,
      "step": 181500
    },
    {
      "epoch": 1.1648,
      "grad_norm": 0.5545785427093506,
      "learning_rate": 2.089056e-05,
      "loss": 0.2625,
      "step": 182000
    },
    {
      "epoch": 1.168,
      "grad_norm": 0.6497994065284729,
      "learning_rate": 2.081056e-05,
      "loss": 0.2611,
      "step": 182500
    },
    {
      "epoch": 1.1712,
      "grad_norm": 0.5887815356254578,
      "learning_rate": 2.073056e-05,
      "loss": 0.2632,
      "step": 183000
    },
    {
      "epoch": 1.1743999999999999,
      "grad_norm": 0.6037270426750183,
      "learning_rate": 2.0650560000000002e-05,
      "loss": 0.2645,
      "step": 183500
    },
    {
      "epoch": 1.1776,
      "grad_norm": 0.636946439743042,
      "learning_rate": 2.057072e-05,
      "loss": 0.2628,
      "step": 184000
    },
    {
      "epoch": 1.1808,
      "grad_norm": 0.5285276770591736,
      "learning_rate": 2.049072e-05,
      "loss": 0.2629,
      "step": 184500
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.4634397625923157,
      "learning_rate": 2.041072e-05,
      "loss": 0.2615,
      "step": 185000
    },
    {
      "epoch": 1.1872,
      "grad_norm": 0.5693604946136475,
      "learning_rate": 2.033072e-05,
      "loss": 0.2619,
      "step": 185500
    }
  ],
  "logging_steps": 500,
  "max_steps": 312500,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.0369350959104e+17,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}