{
  "best_metric": 0.5025785565376282,
  "best_model_checkpoint": "models/E-Coli-FFT/KCYHSM/checkpoint-21500",
  "epoch": 8.96,
  "eval_steps": 500,
  "global_step": 28000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.16,
      "grad_norm": 0.560947835445404,
      "learning_rate": 4.996016e-05,
      "loss": 2.0872,
      "step": 500
    },
    {
      "epoch": 0.16,
      "eval_accuracy_per_token": 0.30261626839637756,
      "eval_loss": 2.0467002391815186,
      "eval_runtime": 226.6763,
      "eval_samples_per_second": 110.289,
      "eval_steps_per_second": 6.895,
      "step": 500
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.5276467204093933,
      "learning_rate": 4.992016000000001e-05,
      "loss": 2.033,
      "step": 1000
    },
    {
      "epoch": 0.32,
      "eval_accuracy_per_token": 0.31899595260620117,
      "eval_loss": 2.0086700916290283,
      "eval_runtime": 215.1833,
      "eval_samples_per_second": 116.18,
      "eval_steps_per_second": 7.264,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.3343302011489868,
      "learning_rate": 4.988016e-05,
      "loss": 1.9439,
      "step": 1500
    },
    {
      "epoch": 0.48,
      "eval_accuracy_per_token": 0.3831162750720978,
      "eval_loss": 1.8512518405914307,
      "eval_runtime": 215.5299,
      "eval_samples_per_second": 115.993,
      "eval_steps_per_second": 7.252,
      "step": 1500
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.4562740325927734,
      "learning_rate": 4.984016e-05,
      "loss": 1.7198,
      "step": 2000
    },
    {
      "epoch": 0.64,
      "eval_accuracy_per_token": 0.47227564454078674,
      "eval_loss": 1.6024138927459717,
      "eval_runtime": 215.7037,
      "eval_samples_per_second": 115.9,
      "eval_steps_per_second": 7.246,
      "step": 2000
    },
    {
      "epoch": 0.8,
      "grad_norm": 2.136140823364258,
      "learning_rate": 4.980016e-05,
      "loss": 1.5124,
      "step": 2500
    },
    {
      "epoch": 0.8,
      "eval_accuracy_per_token": 0.5323951840400696,
      "eval_loss": 1.4281255006790161,
      "eval_runtime": 215.8041,
      "eval_samples_per_second": 115.846,
      "eval_steps_per_second": 7.243,
      "step": 2500
    },
    {
      "epoch": 0.96,
      "grad_norm": 1.6090725660324097,
      "learning_rate": 4.976016e-05,
      "loss": 1.3463,
      "step": 3000
    },
    {
      "epoch": 0.96,
      "eval_accuracy_per_token": 0.5815901160240173,
      "eval_loss": 1.2842084169387817,
      "eval_runtime": 215.7888,
      "eval_samples_per_second": 115.854,
      "eval_steps_per_second": 7.243,
      "step": 3000
    },
    {
      "epoch": 1.12,
      "grad_norm": 2.4967901706695557,
      "learning_rate": 4.972016e-05,
      "loss": 1.1657,
      "step": 3500
    },
    {
      "epoch": 1.12,
      "eval_accuracy_per_token": 0.6198335886001587,
      "eval_loss": 1.1727232933044434,
      "eval_runtime": 215.8392,
      "eval_samples_per_second": 115.827,
      "eval_steps_per_second": 7.242,
      "step": 3500
    },
    {
      "epoch": 1.28,
      "grad_norm": 2.750229597091675,
      "learning_rate": 4.968016e-05,
      "loss": 1.0624,
      "step": 4000
    },
    {
      "epoch": 1.28,
      "eval_accuracy_per_token": 0.6496000289916992,
      "eval_loss": 1.0832889080047607,
      "eval_runtime": 216.0152,
      "eval_samples_per_second": 115.733,
      "eval_steps_per_second": 7.236,
      "step": 4000
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.4956400394439697,
      "learning_rate": 4.9640160000000003e-05,
      "loss": 0.9957,
      "step": 4500
    },
    {
      "epoch": 1.44,
      "eval_accuracy_per_token": 0.6765108108520508,
      "eval_loss": 1.0036195516586304,
      "eval_runtime": 215.7199,
      "eval_samples_per_second": 115.891,
      "eval_steps_per_second": 7.246,
      "step": 4500
    },
    {
      "epoch": 1.6,
      "grad_norm": 2.816002368927002,
      "learning_rate": 4.9600160000000004e-05,
      "loss": 0.913,
      "step": 5000
    },
    {
      "epoch": 1.6,
      "eval_accuracy_per_token": 0.6984472274780273,
      "eval_loss": 0.9418182969093323,
      "eval_runtime": 215.8333,
      "eval_samples_per_second": 115.83,
      "eval_steps_per_second": 7.242,
      "step": 5000
    },
    {
      "epoch": 1.76,
      "grad_norm": 2.293124198913574,
      "learning_rate": 4.956016e-05,
      "loss": 0.8747,
      "step": 5500
    },
    {
      "epoch": 1.76,
      "eval_accuracy_per_token": 0.7186679244041443,
      "eval_loss": 0.8788002133369446,
      "eval_runtime": 215.9244,
      "eval_samples_per_second": 115.781,
      "eval_steps_per_second": 7.239,
      "step": 5500
    },
    {
      "epoch": 1.92,
      "grad_norm": 2.6259233951568604,
      "learning_rate": 4.9520160000000005e-05,
      "loss": 0.8064,
      "step": 6000
    },
    {
      "epoch": 1.92,
      "eval_accuracy_per_token": 0.7395341992378235,
      "eval_loss": 0.8194286823272705,
      "eval_runtime": 215.8802,
      "eval_samples_per_second": 115.805,
      "eval_steps_per_second": 7.24,
      "step": 6000
    },
    {
      "epoch": 2.08,
      "grad_norm": 2.7774460315704346,
      "learning_rate": 4.9480160000000005e-05,
      "loss": 0.7329,
      "step": 6500
    },
    {
      "epoch": 2.08,
      "eval_accuracy_per_token": 0.7511833310127258,
      "eval_loss": 0.7874646782875061,
      "eval_runtime": 215.9936,
      "eval_samples_per_second": 115.744,
      "eval_steps_per_second": 7.236,
      "step": 6500
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.8111910820007324,
      "learning_rate": 4.9440160000000005e-05,
      "loss": 0.6541,
      "step": 7000
    },
    {
      "epoch": 2.24,
      "eval_accuracy_per_token": 0.7654264569282532,
      "eval_loss": 0.7462261915206909,
      "eval_runtime": 215.7774,
      "eval_samples_per_second": 115.86,
      "eval_steps_per_second": 7.244,
      "step": 7000
    },
    {
      "epoch": 2.4,
      "grad_norm": 2.734023332595825,
      "learning_rate": 4.940016e-05,
      "loss": 0.6368,
      "step": 7500
    },
    {
      "epoch": 2.4,
      "eval_accuracy_per_token": 0.7762272953987122,
      "eval_loss": 0.7165001630783081,
      "eval_runtime": 215.5714,
      "eval_samples_per_second": 115.971,
      "eval_steps_per_second": 7.25,
      "step": 7500
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.085864782333374,
      "learning_rate": 4.936016e-05,
      "loss": 0.6019,
      "step": 8000
    },
    {
      "epoch": 2.56,
      "eval_accuracy_per_token": 0.7840808033943176,
      "eval_loss": 0.6941312551498413,
      "eval_runtime": 215.5763,
      "eval_samples_per_second": 115.968,
      "eval_steps_per_second": 7.25,
      "step": 8000
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 2.5468504428863525,
      "learning_rate": 4.9320320000000004e-05,
      "loss": 0.5855,
      "step": 8500
    },
    {
      "epoch": 2.7199999999999998,
      "eval_accuracy_per_token": 0.7905924916267395,
      "eval_loss": 0.6726610064506531,
      "eval_runtime": 215.5941,
      "eval_samples_per_second": 115.959,
      "eval_steps_per_second": 7.25,
      "step": 8500
    },
    {
      "epoch": 2.88,
      "grad_norm": 2.7991392612457275,
      "learning_rate": 4.9280320000000005e-05,
      "loss": 0.565,
      "step": 9000
    },
    {
      "epoch": 2.88,
      "eval_accuracy_per_token": 0.7970213890075684,
      "eval_loss": 0.6574403047561646,
      "eval_runtime": 215.6086,
      "eval_samples_per_second": 115.951,
      "eval_steps_per_second": 7.249,
      "step": 9000
    },
    {
      "epoch": 3.04,
      "grad_norm": 2.0875983238220215,
      "learning_rate": 4.924032e-05,
      "loss": 0.5437,
      "step": 9500
    },
    {
      "epoch": 3.04,
      "eval_accuracy_per_token": 0.8036227226257324,
      "eval_loss": 0.6450381278991699,
      "eval_runtime": 215.7698,
      "eval_samples_per_second": 115.864,
      "eval_steps_per_second": 7.244,
      "step": 9500
    },
    {
      "epoch": 3.2,
      "grad_norm": 2.820004940032959,
      "learning_rate": 4.920032e-05,
      "loss": 0.4618,
      "step": 10000
    },
    {
      "epoch": 3.2,
      "eval_accuracy_per_token": 0.8065482378005981,
      "eval_loss": 0.6290712952613831,
      "eval_runtime": 215.5858,
      "eval_samples_per_second": 115.963,
      "eval_steps_per_second": 7.25,
      "step": 10000
    },
    {
      "epoch": 3.36,
      "grad_norm": 2.534975051879883,
      "learning_rate": 4.916048e-05,
      "loss": 0.4836,
      "step": 10500
    },
    {
      "epoch": 3.36,
      "eval_accuracy_per_token": 0.8109478950500488,
      "eval_loss": 0.621654748916626,
      "eval_runtime": 215.6832,
      "eval_samples_per_second": 115.911,
      "eval_steps_per_second": 7.247,
      "step": 10500
    },
    {
      "epoch": 3.52,
      "grad_norm": 2.2555694580078125,
      "learning_rate": 4.9120480000000004e-05,
      "loss": 0.4793,
      "step": 11000
    },
    {
      "epoch": 3.52,
      "eval_accuracy_per_token": 0.8153020739555359,
      "eval_loss": 0.6032074093818665,
      "eval_runtime": 215.7864,
      "eval_samples_per_second": 115.855,
      "eval_steps_per_second": 7.243,
      "step": 11000
    },
    {
      "epoch": 3.68,
      "grad_norm": 2.6011011600494385,
      "learning_rate": 4.9080480000000004e-05,
      "loss": 0.4639,
      "step": 11500
    },
    {
      "epoch": 3.68,
      "eval_accuracy_per_token": 0.8189014792442322,
      "eval_loss": 0.5884661078453064,
      "eval_runtime": 215.9181,
      "eval_samples_per_second": 115.785,
      "eval_steps_per_second": 7.239,
      "step": 11500
    },
    {
      "epoch": 3.84,
      "grad_norm": 2.201638698577881,
      "learning_rate": 4.904048e-05,
      "loss": 0.4734,
      "step": 12000
    },
    {
      "epoch": 3.84,
      "eval_accuracy_per_token": 0.8225154280662537,
      "eval_loss": 0.5901506543159485,
      "eval_runtime": 215.7481,
      "eval_samples_per_second": 115.876,
      "eval_steps_per_second": 7.245,
      "step": 12000
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.115771770477295,
      "learning_rate": 4.9000480000000005e-05,
      "loss": 0.4589,
      "step": 12500
    },
    {
      "epoch": 4.0,
      "eval_accuracy_per_token": 0.824912428855896,
      "eval_loss": 0.57036292552948,
      "eval_runtime": 215.9409,
      "eval_samples_per_second": 115.772,
      "eval_steps_per_second": 7.238,
      "step": 12500
    },
    {
      "epoch": 4.16,
      "grad_norm": 2.13425350189209,
      "learning_rate": 4.8960480000000005e-05,
      "loss": 0.3846,
      "step": 13000
    },
    {
      "epoch": 4.16,
      "eval_accuracy_per_token": 0.8280689716339111,
      "eval_loss": 0.5732296705245972,
      "eval_runtime": 215.8629,
      "eval_samples_per_second": 115.814,
      "eval_steps_per_second": 7.241,
      "step": 13000
    },
    {
      "epoch": 4.32,
      "grad_norm": 2.227823257446289,
      "learning_rate": 4.892048e-05,
      "loss": 0.3907,
      "step": 13500
    },
    {
      "epoch": 4.32,
      "eval_accuracy_per_token": 0.8309552073478699,
      "eval_loss": 0.5606500506401062,
      "eval_runtime": 215.8455,
      "eval_samples_per_second": 115.824,
      "eval_steps_per_second": 7.241,
      "step": 13500
    },
    {
      "epoch": 4.48,
      "grad_norm": 1.9877572059631348,
      "learning_rate": 4.888048e-05,
      "loss": 0.3781,
      "step": 14000
    },
    {
      "epoch": 4.48,
      "eval_accuracy_per_token": 0.833513617515564,
      "eval_loss": 0.5528165102005005,
      "eval_runtime": 216.0002,
      "eval_samples_per_second": 115.741,
      "eval_steps_per_second": 7.236,
      "step": 14000
    },
    {
      "epoch": 4.64,
      "grad_norm": 2.0553159713745117,
      "learning_rate": 4.884048e-05,
      "loss": 0.3914,
      "step": 14500
    },
    {
      "epoch": 4.64,
      "eval_accuracy_per_token": 0.8335487246513367,
      "eval_loss": 0.549156904220581,
      "eval_runtime": 215.7201,
      "eval_samples_per_second": 115.891,
      "eval_steps_per_second": 7.246,
      "step": 14500
    },
    {
      "epoch": 4.8,
      "grad_norm": 2.2741925716400146,
      "learning_rate": 4.880048000000001e-05,
      "loss": 0.3894,
      "step": 15000
    },
    {
      "epoch": 4.8,
      "eval_accuracy_per_token": 0.8362560272216797,
      "eval_loss": 0.5504783391952515,
      "eval_runtime": 215.7979,
      "eval_samples_per_second": 115.849,
      "eval_steps_per_second": 7.243,
      "step": 15000
    },
    {
      "epoch": 4.96,
      "grad_norm": 2.08858585357666,
      "learning_rate": 4.876048e-05,
      "loss": 0.3944,
      "step": 15500
    },
    {
      "epoch": 4.96,
      "eval_accuracy_per_token": 0.8395103812217712,
      "eval_loss": 0.5324631929397583,
      "eval_runtime": 215.9188,
      "eval_samples_per_second": 115.784,
      "eval_steps_per_second": 7.239,
      "step": 15500
    },
    {
      "epoch": 5.12,
      "grad_norm": 2.1368110179901123,
      "learning_rate": 4.872048e-05,
      "loss": 0.3307,
      "step": 16000
    },
    {
      "epoch": 5.12,
      "eval_accuracy_per_token": 0.839469850063324,
      "eval_loss": 0.5465147495269775,
      "eval_runtime": 216.142,
      "eval_samples_per_second": 115.665,
      "eval_steps_per_second": 7.231,
      "step": 16000
    },
    {
      "epoch": 5.28,
      "grad_norm": 2.2641913890838623,
      "learning_rate": 4.868048e-05,
      "loss": 0.3125,
      "step": 16500
    },
    {
      "epoch": 5.28,
      "eval_accuracy_per_token": 0.8418903946876526,
      "eval_loss": 0.5389049053192139,
      "eval_runtime": 216.0271,
      "eval_samples_per_second": 115.726,
      "eval_steps_per_second": 7.235,
      "step": 16500
    },
    {
      "epoch": 5.44,
      "grad_norm": 1.6883608102798462,
      "learning_rate": 4.864048e-05,
      "loss": 0.3213,
      "step": 17000
    },
    {
      "epoch": 5.44,
      "eval_accuracy_per_token": 0.8448163270950317,
      "eval_loss": 0.5376200675964355,
      "eval_runtime": 216.0239,
      "eval_samples_per_second": 115.728,
      "eval_steps_per_second": 7.235,
      "step": 17000
    },
    {
      "epoch": 5.6,
      "grad_norm": 2.2113685607910156,
      "learning_rate": 4.860048e-05,
      "loss": 0.3269,
      "step": 17500
    },
    {
      "epoch": 5.6,
      "eval_accuracy_per_token": 0.8447655439376831,
      "eval_loss": 0.5422470569610596,
      "eval_runtime": 215.062,
      "eval_samples_per_second": 116.246,
      "eval_steps_per_second": 7.268,
      "step": 17500
    },
    {
      "epoch": 5.76,
      "grad_norm": 1.9853556156158447,
      "learning_rate": 4.856048e-05,
      "loss": 0.3374,
      "step": 18000
    },
    {
      "epoch": 5.76,
      "eval_accuracy_per_token": 0.847917914390564,
      "eval_loss": 0.5255292057991028,
      "eval_runtime": 213.7043,
      "eval_samples_per_second": 116.984,
      "eval_steps_per_second": 7.314,
      "step": 18000
    },
    {
      "epoch": 5.92,
      "grad_norm": 2.584245204925537,
      "learning_rate": 4.852048e-05,
      "loss": 0.3229,
      "step": 18500
    },
    {
      "epoch": 5.92,
      "eval_accuracy_per_token": 0.8491668701171875,
      "eval_loss": 0.5113908052444458,
      "eval_runtime": 214.1955,
      "eval_samples_per_second": 116.716,
      "eval_steps_per_second": 7.297,
      "step": 18500
    },
    {
      "epoch": 6.08,
      "grad_norm": 2.0866963863372803,
      "learning_rate": 4.848056e-05,
      "loss": 0.3061,
      "step": 19000
    },
    {
      "epoch": 6.08,
      "eval_accuracy_per_token": 0.8502768874168396,
      "eval_loss": 0.5418649911880493,
      "eval_runtime": 214.0933,
      "eval_samples_per_second": 116.771,
      "eval_steps_per_second": 7.301,
      "step": 19000
    },
    {
      "epoch": 6.24,
      "grad_norm": 2.0000994205474854,
      "learning_rate": 4.844056e-05,
      "loss": 0.2556,
      "step": 19500
    },
    {
      "epoch": 6.24,
      "eval_accuracy_per_token": 0.85043865442276,
      "eval_loss": 0.542536199092865,
      "eval_runtime": 213.1241,
      "eval_samples_per_second": 117.303,
      "eval_steps_per_second": 7.334,
      "step": 19500
    },
    {
      "epoch": 6.4,
      "grad_norm": 2.849112033843994,
      "learning_rate": 4.840064e-05,
      "loss": 0.2672,
      "step": 20000
    },
    {
      "epoch": 6.4,
      "eval_accuracy_per_token": 0.8512768149375916,
      "eval_loss": 0.5348747372627258,
      "eval_runtime": 215.5865,
      "eval_samples_per_second": 115.963,
      "eval_steps_per_second": 7.25,
      "step": 20000
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 2.216937303543091,
      "learning_rate": 4.836064e-05,
      "loss": 0.2791,
      "step": 20500
    },
    {
      "epoch": 6.5600000000000005,
      "eval_accuracy_per_token": 0.8536346554756165,
      "eval_loss": 0.5185777544975281,
      "eval_runtime": 215.5365,
      "eval_samples_per_second": 115.99,
      "eval_steps_per_second": 7.252,
      "step": 20500
    },
    {
      "epoch": 6.72,
      "grad_norm": 2.2237415313720703,
      "learning_rate": 4.832064e-05,
      "loss": 0.2792,
      "step": 21000
    },
    {
      "epoch": 6.72,
      "eval_accuracy_per_token": 0.8542323112487793,
      "eval_loss": 0.5137789249420166,
      "eval_runtime": 215.6897,
      "eval_samples_per_second": 115.907,
      "eval_steps_per_second": 7.247,
      "step": 21000
    },
    {
      "epoch": 6.88,
      "grad_norm": 2.6220571994781494,
      "learning_rate": 4.828064e-05,
      "loss": 0.2867,
      "step": 21500
    },
    {
      "epoch": 6.88,
      "eval_accuracy_per_token": 0.8552057147026062,
      "eval_loss": 0.5025785565376282,
      "eval_runtime": 215.6135,
      "eval_samples_per_second": 115.948,
      "eval_steps_per_second": 7.249,
      "step": 21500
    },
    {
      "epoch": 7.04,
      "grad_norm": 2.4813005924224854,
      "learning_rate": 4.82408e-05,
      "loss": 0.2628,
      "step": 22000
    },
    {
      "epoch": 7.04,
      "eval_accuracy_per_token": 0.8553439974784851,
      "eval_loss": 0.5604137778282166,
      "eval_runtime": 215.6781,
      "eval_samples_per_second": 115.913,
      "eval_steps_per_second": 7.247,
      "step": 22000
    },
    {
      "epoch": 7.2,
      "grad_norm": 2.316831588745117,
      "learning_rate": 4.82008e-05,
      "loss": 0.2124,
      "step": 22500
    },
    {
      "epoch": 7.2,
      "eval_accuracy_per_token": 0.8566614389419556,
      "eval_loss": 0.5436919927597046,
      "eval_runtime": 215.5335,
      "eval_samples_per_second": 115.991,
      "eval_steps_per_second": 7.252,
      "step": 22500
    },
    {
      "epoch": 7.36,
      "grad_norm": 2.0445761680603027,
      "learning_rate": 4.81608e-05,
      "loss": 0.2246,
      "step": 23000
    },
    {
      "epoch": 7.36,
      "eval_accuracy_per_token": 0.8579392433166504,
      "eval_loss": 0.5327755212783813,
      "eval_runtime": 215.6442,
      "eval_samples_per_second": 115.932,
      "eval_steps_per_second": 7.248,
      "step": 23000
    },
    {
      "epoch": 7.52,
      "grad_norm": 1.8908203840255737,
      "learning_rate": 4.81208e-05,
      "loss": 0.2266,
      "step": 23500
    },
    {
      "epoch": 7.52,
      "eval_accuracy_per_token": 0.8593510389328003,
      "eval_loss": 0.5242588520050049,
      "eval_runtime": 215.5467,
      "eval_samples_per_second": 115.984,
      "eval_steps_per_second": 7.251,
      "step": 23500
    },
    {
      "epoch": 7.68,
      "grad_norm": 1.698264241218567,
      "learning_rate": 4.80808e-05,
      "loss": 0.2335,
      "step": 24000
    },
    {
      "epoch": 7.68,
      "eval_accuracy_per_token": 0.8600181341171265,
      "eval_loss": 0.5188168883323669,
      "eval_runtime": 215.5491,
      "eval_samples_per_second": 115.983,
      "eval_steps_per_second": 7.251,
      "step": 24000
    },
    {
      "epoch": 7.84,
      "grad_norm": 2.196866035461426,
      "learning_rate": 4.80408e-05,
      "loss": 0.2375,
      "step": 24500
    },
    {
      "epoch": 7.84,
      "eval_accuracy_per_token": 0.8607679605484009,
      "eval_loss": 0.5174950957298279,
      "eval_runtime": 215.872,
      "eval_samples_per_second": 115.809,
      "eval_steps_per_second": 7.24,
      "step": 24500
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.599553346633911,
      "learning_rate": 4.800088e-05,
      "loss": 0.2426,
      "step": 25000
    },
    {
      "epoch": 8.0,
      "eval_accuracy_per_token": 0.8615555763244629,
      "eval_loss": 0.5121429562568665,
      "eval_runtime": 215.6323,
      "eval_samples_per_second": 115.938,
      "eval_steps_per_second": 7.248,
      "step": 25000
    },
    {
      "epoch": 8.16,
      "grad_norm": 2.2151308059692383,
      "learning_rate": 4.796088e-05,
      "loss": 0.1708,
      "step": 25500
    },
    {
      "epoch": 8.16,
      "eval_accuracy_per_token": 0.8606916666030884,
      "eval_loss": 0.5612675547599792,
      "eval_runtime": 215.6202,
      "eval_samples_per_second": 115.945,
      "eval_steps_per_second": 7.249,
      "step": 25500
    },
    {
      "epoch": 8.32,
      "grad_norm": 2.0478286743164062,
      "learning_rate": 4.792088e-05,
      "loss": 0.1739,
      "step": 26000
    },
    {
      "epoch": 8.32,
      "eval_accuracy_per_token": 0.8610015511512756,
      "eval_loss": 0.5620591640472412,
      "eval_runtime": 215.5503,
      "eval_samples_per_second": 115.982,
      "eval_steps_per_second": 7.251,
      "step": 26000
    },
    {
      "epoch": 8.48,
      "grad_norm": 2.2109131813049316,
      "learning_rate": 4.788088e-05,
      "loss": 0.1893,
      "step": 26500
    },
    {
      "epoch": 8.48,
      "eval_accuracy_per_token": 0.861874520778656,
      "eval_loss": 0.5581731200218201,
      "eval_runtime": 215.6555,
      "eval_samples_per_second": 115.926,
      "eval_steps_per_second": 7.248,
      "step": 26500
    },
    {
      "epoch": 8.64,
      "grad_norm": 2.688985586166382,
      "learning_rate": 4.784096e-05,
      "loss": 0.1948,
      "step": 27000
    },
    {
      "epoch": 8.64,
      "eval_accuracy_per_token": 0.864142119884491,
      "eval_loss": 0.5396531820297241,
      "eval_runtime": 215.5944,
      "eval_samples_per_second": 115.959,
      "eval_steps_per_second": 7.25,
      "step": 27000
    },
    {
      "epoch": 8.8,
      "grad_norm": 2.5816333293914795,
      "learning_rate": 4.780096e-05,
      "loss": 0.2022,
      "step": 27500
    },
    {
      "epoch": 8.8,
      "eval_accuracy_per_token": 0.8647276759147644,
      "eval_loss": 0.537101149559021,
      "eval_runtime": 215.522,
      "eval_samples_per_second": 115.997,
      "eval_steps_per_second": 7.252,
      "step": 27500
    },
    {
      "epoch": 8.96,
      "grad_norm": 2.262164354324341,
      "learning_rate": 4.776096e-05,
      "loss": 0.2027,
      "step": 28000
    },
    {
      "epoch": 8.96,
      "eval_accuracy_per_token": 0.8652188777923584,
      "eval_loss": 0.5241075158119202,
      "eval_runtime": 215.8588,
      "eval_samples_per_second": 115.816,
      "eval_steps_per_second": 7.241,
      "step": 28000
    }
  ],
  "logging_steps": 500,
  "max_steps": 625000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 200,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 19,
        "early_stopping_threshold": 0.01
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.043038124900352e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}