{
  "best_metric": 0.3162495493888855,
  "best_model_checkpoint": "mikhail_panzo/zlm_b64_le4_s12000/checkpoint-9500",
  "epoch": 7.956448911222781,
  "eval_steps": 500,
  "global_step": 9500,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.04187604690117253,
      "grad_norm": 3.0184223651885986,
      "learning_rate": 2.35e-06,
      "loss": 1.1206,
      "step": 50
    },
    {
      "epoch": 0.08375209380234507,
      "grad_norm": 3.4442005157470703,
      "learning_rate": 4.85e-06,
      "loss": 0.8185,
      "step": 100
    },
    {
      "epoch": 0.12562814070351758,
      "grad_norm": 2.6812493801116943,
      "learning_rate": 7.35e-06,
      "loss": 0.7576,
      "step": 150
    },
    {
      "epoch": 0.16750418760469013,
      "grad_norm": 4.518482685089111,
      "learning_rate": 9.85e-06,
      "loss": 0.6938,
      "step": 200
    },
    {
      "epoch": 0.20938023450586266,
      "grad_norm": 4.683146953582764,
      "learning_rate": 1.235e-05,
      "loss": 0.6386,
      "step": 250
    },
    {
      "epoch": 0.25125628140703515,
      "grad_norm": 2.7392578125,
      "learning_rate": 1.485e-05,
      "loss": 0.6226,
      "step": 300
    },
    {
      "epoch": 0.2931323283082077,
      "grad_norm": 2.017056465148926,
      "learning_rate": 1.7349999999999998e-05,
      "loss": 0.6128,
      "step": 350
    },
    {
      "epoch": 0.33500837520938026,
      "grad_norm": 3.6669230461120605,
      "learning_rate": 1.985e-05,
      "loss": 0.5675,
      "step": 400
    },
    {
      "epoch": 0.3768844221105528,
      "grad_norm": 4.16896915435791,
      "learning_rate": 2.235e-05,
      "loss": 0.5655,
      "step": 450
    },
    {
      "epoch": 0.4187604690117253,
      "grad_norm": 3.691459894180298,
      "learning_rate": 2.485e-05,
      "loss": 0.5487,
      "step": 500
    },
    {
      "epoch": 0.4187604690117253,
      "eval_loss": 0.47458764910697937,
      "eval_runtime": 214.7196,
      "eval_samples_per_second": 39.535,
      "eval_steps_per_second": 4.946,
      "step": 500
    },
    {
      "epoch": 0.46063651591289784,
      "grad_norm": 3.3792996406555176,
      "learning_rate": 2.7350000000000004e-05,
      "loss": 0.5232,
      "step": 550
    },
    {
      "epoch": 0.5025125628140703,
      "grad_norm": 3.6263363361358643,
      "learning_rate": 2.985e-05,
      "loss": 0.5132,
      "step": 600
    },
    {
      "epoch": 0.5443886097152428,
      "grad_norm": 2.652538776397705,
      "learning_rate": 3.235e-05,
      "loss": 0.5067,
      "step": 650
    },
    {
      "epoch": 0.5862646566164154,
      "grad_norm": 3.5355172157287598,
      "learning_rate": 3.485e-05,
      "loss": 0.4981,
      "step": 700
    },
    {
      "epoch": 0.628140703517588,
      "grad_norm": 2.5035619735717773,
      "learning_rate": 3.735e-05,
      "loss": 0.5094,
      "step": 750
    },
    {
      "epoch": 0.6700167504187605,
      "grad_norm": 2.9585518836975098,
      "learning_rate": 3.9850000000000006e-05,
      "loss": 0.5053,
      "step": 800
    },
    {
      "epoch": 0.711892797319933,
      "grad_norm": 9.997116088867188,
      "learning_rate": 4.235e-05,
      "loss": 0.502,
      "step": 850
    },
    {
      "epoch": 0.7537688442211056,
      "grad_norm": 1.8630377054214478,
      "learning_rate": 4.4850000000000006e-05,
      "loss": 0.483,
      "step": 900
    },
    {
      "epoch": 0.7956448911222781,
      "grad_norm": 1.5534099340438843,
      "learning_rate": 4.735e-05,
      "loss": 0.4743,
      "step": 950
    },
    {
      "epoch": 0.8375209380234506,
      "grad_norm": 2.8155503273010254,
      "learning_rate": 4.9850000000000006e-05,
      "loss": 0.483,
      "step": 1000
    },
    {
      "epoch": 0.8375209380234506,
      "eval_loss": 0.4226582944393158,
      "eval_runtime": 213.2339,
      "eval_samples_per_second": 39.811,
      "eval_steps_per_second": 4.98,
      "step": 1000
    },
    {
      "epoch": 0.8793969849246231,
      "grad_norm": 4.2074055671691895,
      "learning_rate": 5.235e-05,
      "loss": 0.4625,
      "step": 1050
    },
    {
      "epoch": 0.9212730318257957,
      "grad_norm": 2.215416431427002,
      "learning_rate": 5.485e-05,
      "loss": 0.461,
      "step": 1100
    },
    {
      "epoch": 0.9631490787269682,
      "grad_norm": 3.3321831226348877,
      "learning_rate": 5.7350000000000005e-05,
      "loss": 0.4807,
      "step": 1150
    },
    {
      "epoch": 1.0050251256281406,
      "grad_norm": 3.5796420574188232,
      "learning_rate": 5.9850000000000005e-05,
      "loss": 0.4456,
      "step": 1200
    },
    {
      "epoch": 1.0469011725293131,
      "grad_norm": 2.3565518856048584,
      "learning_rate": 6.235000000000001e-05,
      "loss": 0.4664,
      "step": 1250
    },
    {
      "epoch": 1.0887772194304857,
      "grad_norm": 3.147794485092163,
      "learning_rate": 6.485e-05,
      "loss": 0.4681,
      "step": 1300
    },
    {
      "epoch": 1.1306532663316582,
      "grad_norm": 3.028409957885742,
      "learning_rate": 6.735e-05,
      "loss": 0.4663,
      "step": 1350
    },
    {
      "epoch": 1.1725293132328307,
      "grad_norm": 4.820915222167969,
      "learning_rate": 6.985e-05,
      "loss": 0.4493,
      "step": 1400
    },
    {
      "epoch": 1.2144053601340032,
      "grad_norm": 1.93254816532135,
      "learning_rate": 7.235000000000001e-05,
      "loss": 0.4413,
      "step": 1450
    },
    {
      "epoch": 1.2562814070351758,
      "grad_norm": 2.237015724182129,
      "learning_rate": 7.485e-05,
      "loss": 0.432,
      "step": 1500
    },
    {
      "epoch": 1.2562814070351758,
      "eval_loss": 0.39829039573669434,
      "eval_runtime": 217.3782,
      "eval_samples_per_second": 39.052,
      "eval_steps_per_second": 4.885,
      "step": 1500
    },
    {
      "epoch": 1.2981574539363483,
      "grad_norm": 1.709120273590088,
      "learning_rate": 7.735e-05,
      "loss": 0.4361,
      "step": 1550
    },
    {
      "epoch": 1.3400335008375208,
      "grad_norm": 3.733328104019165,
      "learning_rate": 7.985e-05,
      "loss": 0.4357,
      "step": 1600
    },
    {
      "epoch": 1.3819095477386933,
      "grad_norm": 4.131324768066406,
      "learning_rate": 8.235000000000001e-05,
      "loss": 0.4445,
      "step": 1650
    },
    {
      "epoch": 1.4237855946398659,
      "grad_norm": 2.2893927097320557,
      "learning_rate": 8.485e-05,
      "loss": 0.4379,
      "step": 1700
    },
    {
      "epoch": 1.4656616415410384,
      "grad_norm": 2.1464734077453613,
      "learning_rate": 8.735000000000001e-05,
      "loss": 0.4393,
      "step": 1750
    },
    {
      "epoch": 1.507537688442211,
      "grad_norm": 2.01698637008667,
      "learning_rate": 8.985e-05,
      "loss": 0.4341,
      "step": 1800
    },
    {
      "epoch": 1.5494137353433834,
      "grad_norm": 2.259115219116211,
      "learning_rate": 9.235000000000001e-05,
      "loss": 0.4324,
      "step": 1850
    },
    {
      "epoch": 1.591289782244556,
      "grad_norm": 2.0072858333587646,
      "learning_rate": 9.485e-05,
      "loss": 0.4327,
      "step": 1900
    },
    {
      "epoch": 1.6331658291457285,
      "grad_norm": 2.060303211212158,
      "learning_rate": 9.735000000000001e-05,
      "loss": 0.4279,
      "step": 1950
    },
    {
      "epoch": 1.675041876046901,
      "grad_norm": 1.2846741676330566,
      "learning_rate": 9.985000000000001e-05,
      "loss": 0.429,
      "step": 2000
    },
    {
      "epoch": 1.675041876046901,
      "eval_loss": 0.3953319191932678,
      "eval_runtime": 217.8108,
      "eval_samples_per_second": 38.974,
      "eval_steps_per_second": 4.876,
      "step": 2000
    },
    {
      "epoch": 1.7169179229480735,
      "grad_norm": 2.253840923309326,
      "learning_rate": 9.953e-05,
      "loss": 0.4264,
      "step": 2050
    },
    {
      "epoch": 1.758793969849246,
      "grad_norm": 1.3114299774169922,
      "learning_rate": 9.903e-05,
      "loss": 0.4355,
      "step": 2100
    },
    {
      "epoch": 1.8006700167504186,
      "grad_norm": 1.8178610801696777,
      "learning_rate": 9.853e-05,
      "loss": 0.4256,
      "step": 2150
    },
    {
      "epoch": 1.8425460636515911,
      "grad_norm": 2.1193735599517822,
      "learning_rate": 9.803e-05,
      "loss": 0.4305,
      "step": 2200
    },
    {
      "epoch": 1.8844221105527639,
      "grad_norm": 1.6871263980865479,
      "learning_rate": 9.753e-05,
      "loss": 0.4106,
      "step": 2250
    },
    {
      "epoch": 1.9262981574539364,
      "grad_norm": 1.2699921131134033,
      "learning_rate": 9.703000000000001e-05,
      "loss": 0.4156,
      "step": 2300
    },
    {
      "epoch": 1.968174204355109,
      "grad_norm": 1.7966769933700562,
      "learning_rate": 9.653000000000001e-05,
      "loss": 0.4257,
      "step": 2350
    },
    {
      "epoch": 2.0100502512562812,
      "grad_norm": 1.7838683128356934,
      "learning_rate": 9.603000000000001e-05,
      "loss": 0.4156,
      "step": 2400
    },
    {
      "epoch": 2.0519262981574538,
      "grad_norm": 1.4161527156829834,
      "learning_rate": 9.553e-05,
      "loss": 0.4096,
      "step": 2450
    },
    {
      "epoch": 2.0938023450586263,
      "grad_norm": 1.5864896774291992,
      "learning_rate": 9.503e-05,
      "loss": 0.4168,
      "step": 2500
    },
    {
      "epoch": 2.0938023450586263,
      "eval_loss": 0.3701198697090149,
      "eval_runtime": 220.4399,
      "eval_samples_per_second": 38.509,
      "eval_steps_per_second": 4.818,
      "step": 2500
    },
    {
      "epoch": 2.135678391959799,
      "grad_norm": 1.2460856437683105,
      "learning_rate": 9.453e-05,
      "loss": 0.415,
      "step": 2550
    },
    {
      "epoch": 2.1775544388609713,
      "grad_norm": 1.994315266609192,
      "learning_rate": 9.403e-05,
      "loss": 0.4075,
      "step": 2600
    },
    {
      "epoch": 2.219430485762144,
      "grad_norm": 2.148324966430664,
      "learning_rate": 9.353000000000001e-05,
      "loss": 0.4077,
      "step": 2650
    },
    {
      "epoch": 2.2613065326633164,
      "grad_norm": 1.7242035865783691,
      "learning_rate": 9.303000000000001e-05,
      "loss": 0.4044,
      "step": 2700
    },
    {
      "epoch": 2.3031825795644894,
      "grad_norm": 1.912110447883606,
      "learning_rate": 9.253000000000001e-05,
      "loss": 0.4075,
      "step": 2750
    },
    {
      "epoch": 2.3450586264656614,
      "grad_norm": 1.3661189079284668,
      "learning_rate": 9.203000000000001e-05,
      "loss": 0.4041,
      "step": 2800
    },
    {
      "epoch": 2.3869346733668344,
      "grad_norm": 1.7083531618118286,
      "learning_rate": 9.153e-05,
      "loss": 0.396,
      "step": 2850
    },
    {
      "epoch": 2.4288107202680065,
      "grad_norm": 1.7829670906066895,
      "learning_rate": 9.103e-05,
      "loss": 0.4042,
      "step": 2900
    },
    {
      "epoch": 2.4706867671691795,
      "grad_norm": 2.188753366470337,
      "learning_rate": 9.053e-05,
      "loss": 0.3963,
      "step": 2950
    },
    {
      "epoch": 2.5125628140703515,
      "grad_norm": 1.2608036994934082,
      "learning_rate": 9.003e-05,
      "loss": 0.4021,
      "step": 3000
    },
    {
      "epoch": 2.5125628140703515,
      "eval_loss": 0.3613213300704956,
      "eval_runtime": 214.7226,
      "eval_samples_per_second": 39.535,
      "eval_steps_per_second": 4.946,
      "step": 3000
    },
    {
      "epoch": 2.5544388609715245,
      "grad_norm": 1.501607060432434,
      "learning_rate": 8.953e-05,
      "loss": 0.3897,
      "step": 3050
    },
    {
      "epoch": 2.5963149078726966,
      "grad_norm": 1.321588397026062,
      "learning_rate": 8.903e-05,
      "loss": 0.3948,
      "step": 3100
    },
    {
      "epoch": 2.6381909547738696,
      "grad_norm": 1.5581566095352173,
      "learning_rate": 8.853000000000001e-05,
      "loss": 0.4005,
      "step": 3150
    },
    {
      "epoch": 2.6800670016750416,
      "grad_norm": 1.3824856281280518,
      "learning_rate": 8.803e-05,
      "loss": 0.3981,
      "step": 3200
    },
    {
      "epoch": 2.7219430485762146,
      "grad_norm": 1.7730051279067993,
      "learning_rate": 8.753e-05,
      "loss": 0.3982,
      "step": 3250
    },
    {
      "epoch": 2.7638190954773867,
      "grad_norm": 1.516622543334961,
      "learning_rate": 8.703e-05,
      "loss": 0.3906,
      "step": 3300
    },
    {
      "epoch": 2.8056951423785597,
      "grad_norm": 1.2814849615097046,
      "learning_rate": 8.653e-05,
      "loss": 0.3859,
      "step": 3350
    },
    {
      "epoch": 2.8475711892797317,
      "grad_norm": 1.239619255065918,
      "learning_rate": 8.603e-05,
      "loss": 0.3942,
      "step": 3400
    },
    {
      "epoch": 2.8894472361809047,
      "grad_norm": 2.24064040184021,
      "learning_rate": 8.553e-05,
      "loss": 0.3938,
      "step": 3450
    },
    {
      "epoch": 2.931323283082077,
      "grad_norm": 1.1459012031555176,
      "learning_rate": 8.503e-05,
      "loss": 0.3925,
      "step": 3500
    },
    {
      "epoch": 2.931323283082077,
      "eval_loss": 0.35087427496910095,
      "eval_runtime": 217.1018,
      "eval_samples_per_second": 39.101,
      "eval_steps_per_second": 4.892,
      "step": 3500
    },
    {
      "epoch": 2.9731993299832498,
      "grad_norm": 1.8844960927963257,
      "learning_rate": 8.453e-05,
      "loss": 0.3803,
      "step": 3550
    },
    {
      "epoch": 3.0150753768844223,
      "grad_norm": 1.5738004446029663,
      "learning_rate": 8.403000000000001e-05,
      "loss": 0.3895,
      "step": 3600
    },
    {
      "epoch": 3.056951423785595,
      "grad_norm": 1.19874906539917,
      "learning_rate": 8.353000000000001e-05,
      "loss": 0.3902,
      "step": 3650
    },
    {
      "epoch": 3.0988274706867673,
      "grad_norm": 2.4810609817504883,
      "learning_rate": 8.303000000000001e-05,
      "loss": 0.3863,
      "step": 3700
    },
    {
      "epoch": 3.14070351758794,
      "grad_norm": 1.4734585285186768,
      "learning_rate": 8.253000000000001e-05,
      "loss": 0.391,
      "step": 3750
    },
    {
      "epoch": 3.1825795644891124,
      "grad_norm": 1.6449118852615356,
      "learning_rate": 8.203000000000001e-05,
      "loss": 0.3888,
      "step": 3800
    },
    {
      "epoch": 3.224455611390285,
      "grad_norm": 2.5607736110687256,
      "learning_rate": 8.153e-05,
      "loss": 0.3893,
      "step": 3850
    },
    {
      "epoch": 3.2663316582914574,
      "grad_norm": 1.0670469999313354,
      "learning_rate": 8.103e-05,
      "loss": 0.3853,
      "step": 3900
    },
    {
      "epoch": 3.30820770519263,
      "grad_norm": 2.191486120223999,
      "learning_rate": 8.053e-05,
      "loss": 0.3832,
      "step": 3950
    },
    {
      "epoch": 3.3500837520938025,
      "grad_norm": 1.0259532928466797,
      "learning_rate": 8.003e-05,
      "loss": 0.3839,
      "step": 4000
    },
    {
      "epoch": 3.3500837520938025,
      "eval_loss": 0.35058465600013733,
      "eval_runtime": 214.7925,
      "eval_samples_per_second": 39.522,
      "eval_steps_per_second": 4.944,
      "step": 4000
    },
    {
      "epoch": 3.391959798994975,
      "grad_norm": 2.1106534004211426,
      "learning_rate": 7.953e-05,
      "loss": 0.4088,
      "step": 4050
    },
    {
      "epoch": 3.4338358458961475,
      "grad_norm": 1.7350833415985107,
      "learning_rate": 7.903000000000001e-05,
      "loss": 0.3923,
      "step": 4100
    },
    {
      "epoch": 3.47571189279732,
      "grad_norm": 1.1461760997772217,
      "learning_rate": 7.853000000000001e-05,
      "loss": 0.3924,
      "step": 4150
    },
    {
      "epoch": 3.5175879396984926,
      "grad_norm": 1.3420217037200928,
      "learning_rate": 7.803e-05,
      "loss": 0.3833,
      "step": 4200
    },
    {
      "epoch": 3.559463986599665,
      "grad_norm": 1.1494131088256836,
      "learning_rate": 7.753e-05,
      "loss": 0.3821,
      "step": 4250
    },
    {
      "epoch": 3.6013400335008376,
      "grad_norm": 2.058499813079834,
      "learning_rate": 7.703e-05,
      "loss": 0.3784,
      "step": 4300
    },
    {
      "epoch": 3.64321608040201,
      "grad_norm": 1.8158782720565796,
      "learning_rate": 7.653e-05,
      "loss": 0.3833,
      "step": 4350
    },
    {
      "epoch": 3.6850921273031827,
      "grad_norm": 1.8433794975280762,
      "learning_rate": 7.603e-05,
      "loss": 0.3768,
      "step": 4400
    },
    {
      "epoch": 3.726968174204355,
      "grad_norm": 1.6182619333267212,
      "learning_rate": 7.553e-05,
      "loss": 0.3772,
      "step": 4450
    },
    {
      "epoch": 3.7688442211055277,
      "grad_norm": 1.5222620964050293,
      "learning_rate": 7.503e-05,
      "loss": 0.3798,
      "step": 4500
    },
    {
      "epoch": 3.7688442211055277,
      "eval_loss": 0.3422691822052002,
      "eval_runtime": 214.805,
      "eval_samples_per_second": 39.52,
      "eval_steps_per_second": 4.944,
      "step": 4500
    },
    {
      "epoch": 3.8107202680067003,
      "grad_norm": 1.981819748878479,
      "learning_rate": 7.453e-05,
      "loss": 0.3739,
      "step": 4550
    },
    {
      "epoch": 3.852596314907873,
      "grad_norm": 1.1871320009231567,
      "learning_rate": 7.403e-05,
      "loss": 0.3769,
      "step": 4600
    },
    {
      "epoch": 3.8944723618090453,
      "grad_norm": 1.5213547945022583,
      "learning_rate": 7.353e-05,
      "loss": 0.3734,
      "step": 4650
    },
    {
      "epoch": 3.936348408710218,
      "grad_norm": 1.6221413612365723,
      "learning_rate": 7.303e-05,
      "loss": 0.3704,
      "step": 4700
    },
    {
      "epoch": 3.9782244556113904,
      "grad_norm": 2.075160264968872,
      "learning_rate": 7.253e-05,
      "loss": 0.3701,
      "step": 4750
    },
    {
      "epoch": 4.0201005025125625,
      "grad_norm": 1.217353105545044,
      "learning_rate": 7.203000000000001e-05,
      "loss": 0.3747,
      "step": 4800
    },
    {
      "epoch": 4.061976549413735,
      "grad_norm": 1.9273918867111206,
      "learning_rate": 7.153000000000001e-05,
      "loss": 0.3718,
      "step": 4850
    },
    {
      "epoch": 4.1038525963149075,
      "grad_norm": 1.2843226194381714,
      "learning_rate": 7.103e-05,
      "loss": 0.3689,
      "step": 4900
    },
    {
      "epoch": 4.1457286432160805,
      "grad_norm": 2.8989481925964355,
      "learning_rate": 7.053e-05,
      "loss": 0.3785,
      "step": 4950
    },
    {
      "epoch": 4.187604690117253,
      "grad_norm": 1.6432912349700928,
      "learning_rate": 7.003e-05,
      "loss": 0.3693,
      "step": 5000
    },
    {
      "epoch": 4.187604690117253,
      "eval_loss": 0.3375495374202728,
      "eval_runtime": 215.1803,
      "eval_samples_per_second": 39.451,
      "eval_steps_per_second": 4.935,
      "step": 5000
    },
    {
      "epoch": 4.2294807370184255,
      "grad_norm": 1.5097157955169678,
      "learning_rate": 6.953000000000001e-05,
      "loss": 0.3731,
      "step": 5050
    },
    {
      "epoch": 4.271356783919598,
      "grad_norm": 1.1934735774993896,
      "learning_rate": 6.903000000000001e-05,
      "loss": 0.3683,
      "step": 5100
    },
    {
      "epoch": 4.313232830820771,
      "grad_norm": 1.1494152545928955,
      "learning_rate": 6.853000000000001e-05,
      "loss": 0.3752,
      "step": 5150
    },
    {
      "epoch": 4.355108877721943,
      "grad_norm": 1.404390811920166,
      "learning_rate": 6.803000000000001e-05,
      "loss": 0.3697,
      "step": 5200
    },
    {
      "epoch": 4.396984924623116,
      "grad_norm": 1.0370264053344727,
      "learning_rate": 6.753e-05,
      "loss": 0.3688,
      "step": 5250
    },
    {
      "epoch": 4.438860971524288,
      "grad_norm": 1.6091570854187012,
      "learning_rate": 6.703e-05,
      "loss": 0.3692,
      "step": 5300
    },
    {
      "epoch": 4.480737018425461,
      "grad_norm": 2.325129270553589,
      "learning_rate": 6.653e-05,
      "loss": 0.3711,
      "step": 5350
    },
    {
      "epoch": 4.522613065326633,
      "grad_norm": 1.2980619668960571,
      "learning_rate": 6.603e-05,
      "loss": 0.379,
      "step": 5400
    },
    {
      "epoch": 4.564489112227806,
      "grad_norm": 1.407894253730774,
      "learning_rate": 6.553e-05,
      "loss": 0.3678,
      "step": 5450
    },
    {
      "epoch": 4.606365159128979,
      "grad_norm": 1.9543007612228394,
      "learning_rate": 6.503e-05,
      "loss": 0.3712,
      "step": 5500
    },
    {
      "epoch": 4.606365159128979,
      "eval_loss": 0.33667415380477905,
      "eval_runtime": 214.7408,
      "eval_samples_per_second": 39.531,
      "eval_steps_per_second": 4.945,
      "step": 5500
    },
    {
      "epoch": 4.648241206030151,
      "grad_norm": 1.7211672067642212,
      "learning_rate": 6.453000000000001e-05,
      "loss": 0.369,
      "step": 5550
    },
    {
      "epoch": 4.690117252931323,
      "grad_norm": 1.2072224617004395,
      "learning_rate": 6.403e-05,
      "loss": 0.3661,
      "step": 5600
    },
    {
      "epoch": 4.731993299832496,
      "grad_norm": 1.994110345840454,
      "learning_rate": 6.353e-05,
      "loss": 0.3685,
      "step": 5650
    },
    {
      "epoch": 4.773869346733669,
      "grad_norm": 1.9967550039291382,
      "learning_rate": 6.303e-05,
      "loss": 0.3647,
      "step": 5700
    },
    {
      "epoch": 4.815745393634841,
      "grad_norm": 1.83970046043396,
      "learning_rate": 6.253e-05,
      "loss": 0.369,
      "step": 5750
    },
    {
      "epoch": 4.857621440536013,
      "grad_norm": 1.1974071264266968,
      "learning_rate": 6.203e-05,
      "loss": 0.3662,
      "step": 5800
    },
    {
      "epoch": 4.899497487437186,
      "grad_norm": 1.50179123878479,
      "learning_rate": 6.153e-05,
      "loss": 0.3663,
      "step": 5850
    },
    {
      "epoch": 4.941373534338359,
      "grad_norm": 1.105136513710022,
      "learning_rate": 6.103e-05,
      "loss": 0.3654,
      "step": 5900
    },
    {
      "epoch": 4.983249581239531,
      "grad_norm": 1.079147219657898,
      "learning_rate": 6.053e-05,
      "loss": 0.3613,
      "step": 5950
    },
    {
      "epoch": 5.025125628140704,
      "grad_norm": 1.25212824344635,
      "learning_rate": 6.003e-05,
      "loss": 0.3668,
      "step": 6000
    },
    {
      "epoch": 5.025125628140704,
      "eval_loss": 0.3315671682357788,
      "eval_runtime": 216.0413,
      "eval_samples_per_second": 39.293,
      "eval_steps_per_second": 4.916,
      "step": 6000
    },
    {
      "epoch": 5.067001675041876,
      "grad_norm": 1.2671592235565186,
      "learning_rate": 5.953000000000001e-05,
      "loss": 0.3589,
      "step": 6050
    },
    {
      "epoch": 5.108877721943049,
      "grad_norm": 2.1920173168182373,
      "learning_rate": 5.903000000000001e-05,
      "loss": 0.3647,
      "step": 6100
    },
    {
      "epoch": 5.150753768844221,
      "grad_norm": 0.9611026048660278,
      "learning_rate": 5.8530000000000004e-05,
      "loss": 0.3639,
      "step": 6150
    },
    {
      "epoch": 5.192629815745394,
      "grad_norm": 1.5600687265396118,
      "learning_rate": 5.8030000000000005e-05,
      "loss": 0.3608,
      "step": 6200
    },
    {
      "epoch": 5.234505862646566,
      "grad_norm": 1.410090446472168,
      "learning_rate": 5.7530000000000007e-05,
      "loss": 0.3602,
      "step": 6250
    },
    {
      "epoch": 5.276381909547739,
      "grad_norm": 1.2535935640335083,
      "learning_rate": 5.703000000000001e-05,
      "loss": 0.3611,
      "step": 6300
    },
    {
      "epoch": 5.318257956448911,
      "grad_norm": 1.2604366540908813,
      "learning_rate": 5.653e-05,
      "loss": 0.3609,
      "step": 6350
    },
    {
      "epoch": 5.360134003350084,
      "grad_norm": 1.4509762525558472,
      "learning_rate": 5.6030000000000004e-05,
      "loss": 0.366,
      "step": 6400
    },
    {
      "epoch": 5.402010050251256,
      "grad_norm": 2.065939426422119,
      "learning_rate": 5.5530000000000005e-05,
      "loss": 0.3622,
      "step": 6450
    },
    {
      "epoch": 5.443886097152429,
      "grad_norm": 1.6851164102554321,
      "learning_rate": 5.503000000000001e-05,
      "loss": 0.3635,
      "step": 6500
    },
    {
      "epoch": 5.443886097152429,
      "eval_loss": 0.3290642201900482,
      "eval_runtime": 216.182,
      "eval_samples_per_second": 39.268,
      "eval_steps_per_second": 4.913,
      "step": 6500
    },
    {
      "epoch": 5.485762144053601,
      "grad_norm": 1.2512905597686768,
      "learning_rate": 5.453e-05,
      "loss": 0.3595,
      "step": 6550
    },
    {
      "epoch": 5.527638190954773,
      "grad_norm": 1.0366383790969849,
      "learning_rate": 5.403e-05,
      "loss": 0.3654,
      "step": 6600
    },
    {
      "epoch": 5.569514237855946,
      "grad_norm": 1.6947450637817383,
      "learning_rate": 5.3530000000000004e-05,
      "loss": 0.3573,
      "step": 6650
    },
    {
      "epoch": 5.611390284757119,
      "grad_norm": 1.5837823152542114,
      "learning_rate": 5.303e-05,
      "loss": 0.3575,
      "step": 6700
    },
    {
      "epoch": 5.653266331658291,
      "grad_norm": 1.6993337869644165,
      "learning_rate": 5.253e-05,
      "loss": 0.3606,
      "step": 6750
    },
    {
      "epoch": 5.695142378559464,
      "grad_norm": 1.208640456199646,
      "learning_rate": 5.203e-05,
      "loss": 0.3608,
      "step": 6800
    },
    {
      "epoch": 5.7370184254606365,
      "grad_norm": 1.5352587699890137,
      "learning_rate": 5.153e-05,
      "loss": 0.3613,
      "step": 6850
    },
    {
      "epoch": 5.778894472361809,
      "grad_norm": 1.9643837213516235,
      "learning_rate": 5.103e-05,
      "loss": 0.3542,
      "step": 6900
    },
    {
      "epoch": 5.8207705192629815,
      "grad_norm": 0.9766552448272705,
      "learning_rate": 5.053e-05,
      "loss": 0.3579,
      "step": 6950
    },
    {
      "epoch": 5.8626465661641545,
      "grad_norm": 1.5910046100616455,
      "learning_rate": 5.003e-05,
      "loss": 0.3543,
      "step": 7000
    },
    {
      "epoch": 5.8626465661641545,
      "eval_loss": 0.3250139653682709,
      "eval_runtime": 213.3758,
      "eval_samples_per_second": 39.784,
      "eval_steps_per_second": 4.977,
      "step": 7000
    },
    {
      "epoch": 5.9045226130653266,
      "grad_norm": 2.036315679550171,
      "learning_rate": 4.953e-05,
      "loss": 0.3615,
      "step": 7050
    },
    {
      "epoch": 5.9463986599664995,
      "grad_norm": 1.7725523710250854,
      "learning_rate": 4.903e-05,
      "loss": 0.3569,
      "step": 7100
    },
    {
      "epoch": 5.988274706867672,
      "grad_norm": 0.9128499031066895,
      "learning_rate": 4.8530000000000005e-05,
      "loss": 0.3607,
      "step": 7150
    },
    {
      "epoch": 6.030150753768845,
      "grad_norm": 1.2767729759216309,
      "learning_rate": 4.8030000000000006e-05,
      "loss": 0.3548,
      "step": 7200
    },
    {
      "epoch": 6.072026800670017,
      "grad_norm": 1.359195351600647,
      "learning_rate": 4.753e-05,
      "loss": 0.3595,
      "step": 7250
    },
    {
      "epoch": 6.11390284757119,
      "grad_norm": 1.290307641029358,
      "learning_rate": 4.703e-05,
      "loss": 0.3515,
      "step": 7300
    },
    {
      "epoch": 6.155778894472362,
      "grad_norm": 1.5797748565673828,
      "learning_rate": 4.6530000000000003e-05,
      "loss": 0.3594,
      "step": 7350
    },
    {
      "epoch": 6.197654941373535,
      "grad_norm": 3.1323928833007812,
      "learning_rate": 4.603e-05,
      "loss": 0.3531,
      "step": 7400
    },
    {
      "epoch": 6.239530988274707,
      "grad_norm": 1.1986955404281616,
      "learning_rate": 4.553e-05,
      "loss": 0.3503,
      "step": 7450
    },
    {
      "epoch": 6.28140703517588,
      "grad_norm": 2.56941294670105,
      "learning_rate": 4.503e-05,
      "loss": 0.3526,
      "step": 7500
    },
    {
      "epoch": 6.28140703517588,
      "eval_loss": 0.32207000255584717,
      "eval_runtime": 213.5905,
      "eval_samples_per_second": 39.744,
      "eval_steps_per_second": 4.972,
      "step": 7500
    },
    {
      "epoch": 6.323283082077052,
      "grad_norm": 1.4174665212631226,
      "learning_rate": 4.453e-05,
      "loss": 0.3597,
      "step": 7550
    },
    {
      "epoch": 6.365159128978225,
      "grad_norm": 1.4059972763061523,
      "learning_rate": 4.4030000000000004e-05,
      "loss": 0.3495,
      "step": 7600
    },
    {
      "epoch": 6.407035175879397,
      "grad_norm": 1.2972431182861328,
      "learning_rate": 4.3530000000000005e-05,
      "loss": 0.3534,
      "step": 7650
    },
    {
      "epoch": 6.44891122278057,
      "grad_norm": 1.3449029922485352,
      "learning_rate": 4.3030000000000006e-05,
      "loss": 0.3505,
      "step": 7700
    },
    {
      "epoch": 6.490787269681742,
      "grad_norm": 1.2508536577224731,
      "learning_rate": 4.253e-05,
      "loss": 0.3583,
      "step": 7750
    },
    {
      "epoch": 6.532663316582915,
      "grad_norm": 1.881706714630127,
      "learning_rate": 4.203e-05,
      "loss": 0.3519,
      "step": 7800
    },
    {
      "epoch": 6.574539363484087,
      "grad_norm": 1.0013474225997925,
      "learning_rate": 4.1530000000000004e-05,
      "loss": 0.3512,
      "step": 7850
    },
    {
      "epoch": 6.61641541038526,
      "grad_norm": 1.5754601955413818,
      "learning_rate": 4.103e-05,
      "loss": 0.3517,
      "step": 7900
    },
    {
      "epoch": 6.658291457286432,
      "grad_norm": 1.0650445222854614,
      "learning_rate": 4.053e-05,
      "loss": 0.352,
      "step": 7950
    },
    {
      "epoch": 6.700167504187605,
      "grad_norm": 0.8827164769172668,
      "learning_rate": 4.003e-05,
      "loss": 0.3525,
      "step": 8000
    },
    {
      "epoch": 6.700167504187605,
      "eval_loss": 0.32177916169166565,
      "eval_runtime": 212.2379,
      "eval_samples_per_second": 39.998,
      "eval_steps_per_second": 5.004,
      "step": 8000
    },
    {
      "epoch": 6.742043551088777,
      "grad_norm": 1.0491342544555664,
      "learning_rate": 3.953e-05,
      "loss": 0.3508,
      "step": 8050
    },
    {
      "epoch": 6.78391959798995,
      "grad_norm": 1.8081963062286377,
      "learning_rate": 3.903e-05,
      "loss": 0.3598,
      "step": 8100
    },
    {
      "epoch": 6.825795644891122,
      "grad_norm": 1.4452039003372192,
      "learning_rate": 3.853e-05,
      "loss": 0.3475,
      "step": 8150
    },
    {
      "epoch": 6.867671691792295,
      "grad_norm": 1.9516081809997559,
      "learning_rate": 3.803000000000001e-05,
      "loss": 0.3576,
      "step": 8200
    },
    {
      "epoch": 6.909547738693467,
      "grad_norm": 2.1795125007629395,
      "learning_rate": 3.753e-05,
      "loss": 0.3523,
      "step": 8250
    },
    {
      "epoch": 6.95142378559464,
      "grad_norm": 1.8348519802093506,
      "learning_rate": 3.703e-05,
      "loss": 0.3554,
      "step": 8300
    },
    {
      "epoch": 6.993299832495812,
      "grad_norm": 1.1464494466781616,
      "learning_rate": 3.6530000000000004e-05,
      "loss": 0.3558,
      "step": 8350
    },
    {
      "epoch": 7.035175879396985,
      "grad_norm": 1.2294760942459106,
      "learning_rate": 3.6030000000000006e-05,
      "loss": 0.3502,
      "step": 8400
    },
    {
      "epoch": 7.077051926298157,
      "grad_norm": 4.840690612792969,
      "learning_rate": 3.553e-05,
      "loss": 0.3534,
      "step": 8450
    },
    {
      "epoch": 7.11892797319933,
      "grad_norm": 1.1307451725006104,
      "learning_rate": 3.503e-05,
      "loss": 0.3513,
      "step": 8500
    },
    {
      "epoch": 7.11892797319933,
      "eval_loss": 0.31820428371429443,
      "eval_runtime": 211.821,
      "eval_samples_per_second": 40.076,
      "eval_steps_per_second": 5.014,
      "step": 8500
    },
    {
      "epoch": 7.160804020100502,
      "grad_norm": 1.680982232093811,
      "learning_rate": 3.453e-05,
      "loss": 0.3464,
      "step": 8550
    },
    {
      "epoch": 7.202680067001675,
      "grad_norm": 1.16056489944458,
      "learning_rate": 3.403e-05,
      "loss": 0.353,
      "step": 8600
    },
    {
      "epoch": 7.244556113902847,
      "grad_norm": 1.0772472620010376,
      "learning_rate": 3.353e-05,
      "loss": 0.3515,
      "step": 8650
    },
    {
      "epoch": 7.28643216080402,
      "grad_norm": 1.8330378532409668,
      "learning_rate": 3.303e-05,
      "loss": 0.3487,
      "step": 8700
    },
    {
      "epoch": 7.328308207705192,
      "grad_norm": 3.8991479873657227,
      "learning_rate": 3.253e-05,
      "loss": 0.3504,
      "step": 8750
    },
    {
      "epoch": 7.370184254606365,
      "grad_norm": 1.5016411542892456,
      "learning_rate": 3.2029999999999997e-05,
      "loss": 0.3451,
      "step": 8800
    },
    {
      "epoch": 7.4120603015075375,
      "grad_norm": 1.6319200992584229,
      "learning_rate": 3.154e-05,
      "loss": 0.3487,
      "step": 8850
    },
    {
      "epoch": 7.45393634840871,
      "grad_norm": 1.5115652084350586,
      "learning_rate": 3.104e-05,
      "loss": 0.3451,
      "step": 8900
    },
    {
      "epoch": 7.4958123953098825,
      "grad_norm": 1.1861990690231323,
      "learning_rate": 3.054e-05,
      "loss": 0.3467,
      "step": 8950
    },
    {
      "epoch": 7.5376884422110555,
      "grad_norm": 1.0466668605804443,
      "learning_rate": 3.004e-05,
      "loss": 0.346,
      "step": 9000
    },
    {
      "epoch": 7.5376884422110555,
      "eval_loss": 0.316290020942688,
      "eval_runtime": 212.3656,
      "eval_samples_per_second": 39.974,
      "eval_steps_per_second": 5.001,
      "step": 9000
    },
    {
      "epoch": 7.579564489112228,
      "grad_norm": 1.9901808500289917,
      "learning_rate": 2.9540000000000002e-05,
      "loss": 0.3469,
      "step": 9050
    },
    {
      "epoch": 7.6214405360134005,
      "grad_norm": 1.5441765785217285,
      "learning_rate": 2.904e-05,
      "loss": 0.3452,
      "step": 9100
    },
    {
      "epoch": 7.663316582914573,
      "grad_norm": 1.1541210412979126,
      "learning_rate": 2.854e-05,
      "loss": 0.3453,
      "step": 9150
    },
    {
      "epoch": 7.705192629815746,
      "grad_norm": 1.121564269065857,
      "learning_rate": 2.804e-05,
      "loss": 0.3464,
      "step": 9200
    },
    {
      "epoch": 7.747068676716918,
      "grad_norm": 1.4487674236297607,
      "learning_rate": 2.754e-05,
      "loss": 0.3535,
      "step": 9250
    },
    {
      "epoch": 7.788944723618091,
      "grad_norm": 1.9906679391860962,
      "learning_rate": 2.704e-05,
      "loss": 0.35,
      "step": 9300
    },
    {
      "epoch": 7.830820770519263,
      "grad_norm": 1.2831981182098389,
      "learning_rate": 2.6540000000000003e-05,
      "loss": 0.3494,
      "step": 9350
    },
    {
      "epoch": 7.872696817420436,
      "grad_norm": 1.4484026432037354,
      "learning_rate": 2.6050000000000003e-05,
      "loss": 0.3494,
      "step": 9400
    },
    {
      "epoch": 7.914572864321608,
      "grad_norm": 0.8587632179260254,
      "learning_rate": 2.555e-05,
      "loss": 0.3439,
      "step": 9450
    },
    {
      "epoch": 7.956448911222781,
      "grad_norm": 1.0778433084487915,
      "learning_rate": 2.5050000000000002e-05,
      "loss": 0.3448,
      "step": 9500
    },
    {
      "epoch": 7.956448911222781,
      "eval_loss": 0.3162495493888855,
      "eval_runtime": 212.2033,
      "eval_samples_per_second": 40.004,
      "eval_steps_per_second": 5.005,
      "step": 9500
    }
  ],
  "logging_steps": 50,
  "max_steps": 12000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 11,
  "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": 9.2392487225299e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}