{
  "best_metric": 0.09464961290359497,
  "best_model_checkpoint": "outputs/checkpoint-540",
  "epoch": 9.840546697038725,
  "eval_steps": 500,
  "global_step": 540,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.36446469248291574,
      "grad_norm": 3.396773099899292,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.3727,
      "step": 20
    },
    {
      "epoch": 0.7289293849658315,
      "grad_norm": 1.4876775741577148,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.2861,
      "step": 40
    },
    {
      "epoch": 0.9840546697038725,
      "eval_loss": 0.18749132752418518,
      "eval_runtime": 49.826,
      "eval_samples_per_second": 3.372,
      "eval_steps_per_second": 0.421,
      "step": 54
    },
    {
      "epoch": 1.0933940774487472,
      "grad_norm": 0.6102920770645142,
      "learning_rate": 5.8e-06,
      "loss": 0.1811,
      "step": 60
    },
    {
      "epoch": 1.4578587699316627,
      "grad_norm": 0.4133767783641815,
      "learning_rate": 7.800000000000002e-06,
      "loss": 0.1346,
      "step": 80
    },
    {
      "epoch": 1.8223234624145785,
      "grad_norm": 1.7365530729293823,
      "learning_rate": 9.800000000000001e-06,
      "loss": 0.1243,
      "step": 100
    },
    {
      "epoch": 1.9863325740318907,
      "eval_loss": 0.13598798215389252,
      "eval_runtime": 49.8253,
      "eval_samples_per_second": 3.372,
      "eval_steps_per_second": 0.421,
      "step": 109
    },
    {
      "epoch": 2.1867881548974943,
      "grad_norm": 0.4655854403972626,
      "learning_rate": 9.958763523679515e-06,
      "loss": 0.1038,
      "step": 120
    },
    {
      "epoch": 2.55125284738041,
      "grad_norm": 0.6029446125030518,
      "learning_rate": 9.817090706862895e-06,
      "loss": 0.0873,
      "step": 140
    },
    {
      "epoch": 2.9157175398633255,
      "grad_norm": 0.4066179394721985,
      "learning_rate": 9.577355814597031e-06,
      "loss": 0.0862,
      "step": 160
    },
    {
      "epoch": 2.988610478359909,
      "eval_loss": 0.11723620444536209,
      "eval_runtime": 49.7707,
      "eval_samples_per_second": 3.375,
      "eval_steps_per_second": 0.422,
      "step": 164
    },
    {
      "epoch": 3.2801822323462413,
      "grad_norm": 0.6009082198143005,
      "learning_rate": 9.244439157950114e-06,
      "loss": 0.0834,
      "step": 180
    },
    {
      "epoch": 3.644646924829157,
      "grad_norm": 0.6393954157829285,
      "learning_rate": 8.825117959999117e-06,
      "loss": 0.0756,
      "step": 200
    },
    {
      "epoch": 3.990888382687927,
      "eval_loss": 0.10594599694013596,
      "eval_runtime": 49.8676,
      "eval_samples_per_second": 3.369,
      "eval_steps_per_second": 0.421,
      "step": 219
    },
    {
      "epoch": 4.009111617312073,
      "grad_norm": 0.5410734415054321,
      "learning_rate": 8.327928391111841e-06,
      "loss": 0.0733,
      "step": 220
    },
    {
      "epoch": 4.373576309794989,
      "grad_norm": 0.544377326965332,
      "learning_rate": 7.762991797134513e-06,
      "loss": 0.0684,
      "step": 240
    },
    {
      "epoch": 4.738041002277904,
      "grad_norm": 0.5764002799987793,
      "learning_rate": 7.1418086579779075e-06,
      "loss": 0.0628,
      "step": 260
    },
    {
      "epoch": 4.993166287015946,
      "eval_loss": 0.0995541512966156,
      "eval_runtime": 49.7114,
      "eval_samples_per_second": 3.38,
      "eval_steps_per_second": 0.422,
      "step": 274
    },
    {
      "epoch": 5.10250569476082,
      "grad_norm": 0.48410555720329285,
      "learning_rate": 6.477024471011001e-06,
      "loss": 0.0628,
      "step": 280
    },
    {
      "epoch": 5.466970387243736,
      "grad_norm": 0.6120467185974121,
      "learning_rate": 5.782172325201155e-06,
      "loss": 0.0594,
      "step": 300
    },
    {
      "epoch": 5.831435079726651,
      "grad_norm": 0.7353035807609558,
      "learning_rate": 5.071397406448937e-06,
      "loss": 0.0593,
      "step": 320
    },
    {
      "epoch": 5.995444191343964,
      "eval_loss": 0.09746743738651276,
      "eval_runtime": 49.7379,
      "eval_samples_per_second": 3.378,
      "eval_steps_per_second": 0.422,
      "step": 329
    },
    {
      "epoch": 6.195899772209567,
      "grad_norm": 0.5597842931747437,
      "learning_rate": 4.359169042394537e-06,
      "loss": 0.0606,
      "step": 340
    },
    {
      "epoch": 6.560364464692483,
      "grad_norm": 0.5548788905143738,
      "learning_rate": 3.6599861486331074e-06,
      "loss": 0.0548,
      "step": 360
    },
    {
      "epoch": 6.924829157175399,
      "grad_norm": 0.6110637784004211,
      "learning_rate": 2.9880820726046613e-06,
      "loss": 0.0498,
      "step": 380
    },
    {
      "epoch": 6.997722095671982,
      "eval_loss": 0.09508081525564194,
      "eval_runtime": 49.5881,
      "eval_samples_per_second": 3.388,
      "eval_steps_per_second": 0.423,
      "step": 384
    },
    {
      "epoch": 7.289293849658314,
      "grad_norm": 0.6793264150619507,
      "learning_rate": 2.3571348436857906e-06,
      "loss": 0.0485,
      "step": 400
    },
    {
      "epoch": 7.65375854214123,
      "grad_norm": 0.499203085899353,
      "learning_rate": 1.7799887279557238e-06,
      "loss": 0.0474,
      "step": 420
    },
    {
      "epoch": 8.0,
      "eval_loss": 0.09498950093984604,
      "eval_runtime": 49.5797,
      "eval_samples_per_second": 3.388,
      "eval_steps_per_second": 0.424,
      "step": 439
    },
    {
      "epoch": 8.018223234624147,
      "grad_norm": 0.86496502161026,
      "learning_rate": 1.2683927559787657e-06,
      "loss": 0.0523,
      "step": 440
    },
    {
      "epoch": 8.382687927107062,
      "grad_norm": 0.7341931462287903,
      "learning_rate": 8.327615464234129e-07,
      "loss": 0.0423,
      "step": 460
    },
    {
      "epoch": 8.747152619589977,
      "grad_norm": 0.5355525016784668,
      "learning_rate": 4.819632944595415e-07,
      "loss": 0.047,
      "step": 480
    },
    {
      "epoch": 8.984054669703873,
      "eval_loss": 0.09474514424800873,
      "eval_runtime": 49.5868,
      "eval_samples_per_second": 3.388,
      "eval_steps_per_second": 0.424,
      "step": 493
    },
    {
      "epoch": 9.111617312072893,
      "grad_norm": 0.5788146257400513,
      "learning_rate": 2.2313924087851657e-07,
      "loss": 0.0548,
      "step": 500
    },
    {
      "epoch": 9.476082004555808,
      "grad_norm": 0.4824052155017853,
      "learning_rate": 6.15582970243117e-08,
      "loss": 0.0451,
      "step": 520
    },
    {
      "epoch": 9.840546697038725,
      "grad_norm": 0.9220362901687622,
      "learning_rate": 5.09784952833492e-10,
      "loss": 0.0453,
      "step": 540
    },
    {
      "epoch": 9.840546697038725,
      "eval_loss": 0.09464961290359497,
      "eval_runtime": 49.5747,
      "eval_samples_per_second": 3.389,
      "eval_steps_per_second": 0.424,
      "step": 540
    }
  ],
  "logging_steps": 20,
  "max_steps": 540,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 2.0191775689433088e+17,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}