{
  "best_metric": 0.3953319191932678,
  "best_model_checkpoint": "mikhail_panzo/zlm_b64_le4_s12000/checkpoint-2000",
  "epoch": 1.675041876046901,
  "eval_steps": 500,
  "global_step": 2000,
  "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
    }
  ],
  "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": 1.944641497187981e+16,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}