{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.05102040816326531,
  "eval_steps": 500,
  "global_step": 25,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0020408163265306124,
      "grad_norm": 3.8715779781341553,
      "learning_rate": 0.0,
      "loss": 1.313438892364502,
      "memory/device_reserved (GiB)": 75.76,
      "memory/max_active (GiB)": 74.61,
      "memory/max_allocated (GiB)": 74.61,
      "ppl": 3.71894,
      "step": 1,
      "tokens/total": 249728,
      "tokens/trainable": 1992
    },
    {
      "epoch": 0.004081632653061225,
      "grad_norm": 3.5599887371063232,
      "learning_rate": 3.448275862068966e-06,
      "loss": 1.2385082244873047,
      "memory/device_reserved (GiB)": 75.77,
      "memory/max_active (GiB)": 74.28,
      "memory/max_allocated (GiB)": 74.28,
      "ppl": 3.45046,
      "step": 2,
      "tokens/total": 483392,
      "tokens/train_per_sec_per_gpu": 11.85,
      "tokens/trainable": 3913
    },
    {
      "epoch": 0.006122448979591836,
      "grad_norm": 3.4575953483581543,
      "learning_rate": 6.896551724137932e-06,
      "loss": 1.2344105243682861,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 75.5,
      "memory/max_allocated (GiB)": 75.5,
      "ppl": 3.43635,
      "step": 3,
      "tokens/total": 726144,
      "tokens/train_per_sec_per_gpu": 12.0,
      "tokens/trainable": 5903
    },
    {
      "epoch": 0.00816326530612245,
      "grad_norm": 3.6946189403533936,
      "learning_rate": 1.0344827586206897e-05,
      "loss": 1.3252272605895996,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 72.18,
      "memory/max_allocated (GiB)": 72.18,
      "ppl": 3.76304,
      "step": 4,
      "tokens/total": 943168,
      "tokens/train_per_sec_per_gpu": 15.27,
      "tokens/trainable": 7867
    },
    {
      "epoch": 0.01020408163265306,
      "grad_norm": 3.77402663230896,
      "learning_rate": 1.3793103448275863e-05,
      "loss": 1.3327311277389526,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 75.42,
      "memory/max_allocated (GiB)": 75.42,
      "ppl": 3.79138,
      "step": 5,
      "tokens/total": 1169664,
      "tokens/train_per_sec_per_gpu": 12.41,
      "tokens/trainable": 9745
    },
    {
      "epoch": 0.012244897959183673,
      "grad_norm": 3.8298497200012207,
      "learning_rate": 1.7241379310344828e-05,
      "loss": 1.2867913246154785,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 73.05,
      "memory/max_allocated (GiB)": 73.05,
      "ppl": 3.62115,
      "step": 6,
      "tokens/total": 1406336,
      "tokens/train_per_sec_per_gpu": 13.89,
      "tokens/trainable": 11650
    },
    {
      "epoch": 0.014285714285714285,
      "grad_norm": 3.6937243938446045,
      "learning_rate": 2.0689655172413793e-05,
      "loss": 1.2978110313415527,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 72.54,
      "memory/max_allocated (GiB)": 72.54,
      "ppl": 3.66127,
      "step": 7,
      "tokens/total": 1647104,
      "tokens/train_per_sec_per_gpu": 13.04,
      "tokens/trainable": 13571
    },
    {
      "epoch": 0.0163265306122449,
      "grad_norm": 3.1459977626800537,
      "learning_rate": 2.413793103448276e-05,
      "loss": 1.116440773010254,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 74.44,
      "memory/max_allocated (GiB)": 74.44,
      "ppl": 3.05397,
      "step": 8,
      "tokens/total": 1913472,
      "tokens/train_per_sec_per_gpu": 11.21,
      "tokens/trainable": 15743
    },
    {
      "epoch": 0.018367346938775512,
      "grad_norm": 2.7553977966308594,
      "learning_rate": 2.7586206896551727e-05,
      "loss": 1.096015214920044,
      "memory/device_reserved (GiB)": 76.4,
      "memory/max_active (GiB)": 75.26,
      "memory/max_allocated (GiB)": 75.26,
      "ppl": 2.99222,
      "step": 9,
      "tokens/total": 2159104,
      "tokens/train_per_sec_per_gpu": 13.31,
      "tokens/trainable": 17993
    },
    {
      "epoch": 0.02040816326530612,
      "grad_norm": 3.3356995582580566,
      "learning_rate": 3.103448275862069e-05,
      "loss": 1.1540675163269043,
      "memory/device_reserved (GiB)": 76.59,
      "memory/max_active (GiB)": 75.69,
      "memory/max_allocated (GiB)": 75.69,
      "ppl": 3.17107,
      "step": 10,
      "tokens/total": 2406336,
      "tokens/train_per_sec_per_gpu": 9.84,
      "tokens/trainable": 19603
    },
    {
      "epoch": 0.022448979591836733,
      "grad_norm": 2.382963180541992,
      "learning_rate": 3.4482758620689657e-05,
      "loss": 1.0211127996444702,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 76.81,
      "memory/max_allocated (GiB)": 76.81,
      "ppl": 2.77628,
      "step": 11,
      "tokens/total": 2637440,
      "tokens/train_per_sec_per_gpu": 13.03,
      "tokens/trainable": 21674
    },
    {
      "epoch": 0.024489795918367346,
      "grad_norm": 1.8799129724502563,
      "learning_rate": 3.793103448275862e-05,
      "loss": 0.7622053623199463,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 73.65,
      "memory/max_allocated (GiB)": 73.65,
      "ppl": 2.143,
      "step": 12,
      "tokens/total": 2872896,
      "tokens/train_per_sec_per_gpu": 10.32,
      "tokens/trainable": 23481
    },
    {
      "epoch": 0.026530612244897958,
      "grad_norm": 1.7561086416244507,
      "learning_rate": 4.1379310344827587e-05,
      "loss": 0.7509362697601318,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 74.77,
      "memory/max_allocated (GiB)": 74.77,
      "ppl": 2.11898,
      "step": 13,
      "tokens/total": 3128768,
      "tokens/train_per_sec_per_gpu": 11.06,
      "tokens/trainable": 25355
    },
    {
      "epoch": 0.02857142857142857,
      "grad_norm": 1.4080203771591187,
      "learning_rate": 4.482758620689655e-05,
      "loss": 0.687152624130249,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 75.83,
      "memory/max_allocated (GiB)": 75.83,
      "ppl": 1.98805,
      "step": 14,
      "tokens/total": 3354624,
      "tokens/train_per_sec_per_gpu": 13.35,
      "tokens/trainable": 27379
    },
    {
      "epoch": 0.030612244897959183,
      "grad_norm": 1.6408438682556152,
      "learning_rate": 4.827586206896552e-05,
      "loss": 0.6175031065940857,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 76.32,
      "memory/max_allocated (GiB)": 76.32,
      "ppl": 1.85429,
      "step": 15,
      "tokens/total": 3632576,
      "tokens/train_per_sec_per_gpu": 10.37,
      "tokens/trainable": 29348
    },
    {
      "epoch": 0.0326530612244898,
      "grad_norm": 0.8767540454864502,
      "learning_rate": 5.172413793103449e-05,
      "loss": 0.4765658676624298,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 76.86,
      "memory/max_allocated (GiB)": 76.86,
      "ppl": 1.61053,
      "step": 16,
      "tokens/total": 3893632,
      "tokens/train_per_sec_per_gpu": 11.17,
      "tokens/trainable": 31409
    },
    {
      "epoch": 0.03469387755102041,
      "grad_norm": 0.6025304198265076,
      "learning_rate": 5.517241379310345e-05,
      "loss": 0.5316739678382874,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 73.6,
      "memory/max_allocated (GiB)": 73.6,
      "ppl": 1.70178,
      "step": 17,
      "tokens/total": 4120000,
      "tokens/train_per_sec_per_gpu": 16.29,
      "tokens/trainable": 33895
    },
    {
      "epoch": 0.036734693877551024,
      "grad_norm": 0.6741383671760559,
      "learning_rate": 5.862068965517241e-05,
      "loss": 0.49132800102233887,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 73.19,
      "memory/max_allocated (GiB)": 73.19,
      "ppl": 1.63449,
      "step": 18,
      "tokens/total": 4334208,
      "tokens/train_per_sec_per_gpu": 12.95,
      "tokens/trainable": 35668
    },
    {
      "epoch": 0.03877551020408163,
      "grad_norm": 0.45101675391197205,
      "learning_rate": 6.206896551724138e-05,
      "loss": 0.45823076367378235,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 75.47,
      "memory/max_allocated (GiB)": 75.47,
      "ppl": 1.58127,
      "step": 19,
      "tokens/total": 4595520,
      "tokens/train_per_sec_per_gpu": 14.98,
      "tokens/trainable": 38006
    },
    {
      "epoch": 0.04081632653061224,
      "grad_norm": 0.4415302574634552,
      "learning_rate": 6.551724137931034e-05,
      "loss": 0.35276925563812256,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 73.57,
      "memory/max_allocated (GiB)": 73.57,
      "ppl": 1.423,
      "step": 20,
      "tokens/total": 4823424,
      "tokens/train_per_sec_per_gpu": 15.23,
      "tokens/trainable": 40245
    },
    {
      "epoch": 0.04285714285714286,
      "grad_norm": 0.3700725734233856,
      "learning_rate": 6.896551724137931e-05,
      "loss": 0.48657187819480896,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 74.82,
      "memory/max_allocated (GiB)": 74.82,
      "ppl": 1.62673,
      "step": 21,
      "tokens/total": 5095296,
      "tokens/train_per_sec_per_gpu": 12.47,
      "tokens/trainable": 42288
    },
    {
      "epoch": 0.044897959183673466,
      "grad_norm": 0.3767106533050537,
      "learning_rate": 7.241379310344828e-05,
      "loss": 0.3799547255039215,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 76.37,
      "memory/max_allocated (GiB)": 76.37,
      "ppl": 1.46222,
      "step": 22,
      "tokens/total": 5361536,
      "tokens/train_per_sec_per_gpu": 11.92,
      "tokens/trainable": 44219
    },
    {
      "epoch": 0.04693877551020408,
      "grad_norm": 0.38255029916763306,
      "learning_rate": 7.586206896551724e-05,
      "loss": 0.4809321463108063,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 76.18,
      "memory/max_allocated (GiB)": 76.18,
      "ppl": 1.61758,
      "step": 23,
      "tokens/total": 5611776,
      "tokens/train_per_sec_per_gpu": 11.9,
      "tokens/trainable": 46265
    },
    {
      "epoch": 0.04897959183673469,
      "grad_norm": 2.0013272762298584,
      "learning_rate": 7.931034482758621e-05,
      "loss": 0.44761645793914795,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 75.09,
      "memory/max_allocated (GiB)": 75.09,
      "ppl": 1.56458,
      "step": 24,
      "tokens/total": 5842944,
      "tokens/train_per_sec_per_gpu": 13.07,
      "tokens/trainable": 48369
    },
    {
      "epoch": 0.05102040816326531,
      "grad_norm": 0.3144684433937073,
      "learning_rate": 8.275862068965517e-05,
      "loss": 0.31974953413009644,
      "memory/device_reserved (GiB)": 77.69,
      "memory/max_active (GiB)": 75.72,
      "memory/max_allocated (GiB)": 75.72,
      "ppl": 1.37678,
      "step": 25,
      "tokens/total": 6081152,
      "tokens/train_per_sec_per_gpu": 13.43,
      "tokens/trainable": 50389
    }
  ],
  "logging_steps": 1,
  "max_steps": 980,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 25,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1261128083714867e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}