{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 9.0625,
  "eval_steps": 500,
  "global_step": 1160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.078125,
      "grad_norm": 6.125,
      "learning_rate": 1e-05,
      "loss": 12.2019,
      "step": 10
    },
    {
      "epoch": 0.15625,
      "grad_norm": 6.5,
      "learning_rate": 2e-05,
      "loss": 11.8976,
      "step": 20
    },
    {
      "epoch": 0.234375,
      "grad_norm": 7.46875,
      "learning_rate": 3e-05,
      "loss": 11.1732,
      "step": 30
    },
    {
      "epoch": 0.3125,
      "grad_norm": 7.0625,
      "learning_rate": 4e-05,
      "loss": 9.8831,
      "step": 40
    },
    {
      "epoch": 0.390625,
      "grad_norm": 7.375,
      "learning_rate": 5e-05,
      "loss": 8.4578,
      "step": 50
    },
    {
      "epoch": 0.46875,
      "grad_norm": 8.9375,
      "learning_rate": 6e-05,
      "loss": 7.2717,
      "step": 60
    },
    {
      "epoch": 0.546875,
      "grad_norm": 12.25,
      "learning_rate": 7e-05,
      "loss": 6.1268,
      "step": 70
    },
    {
      "epoch": 0.625,
      "grad_norm": 13.4375,
      "learning_rate": 8e-05,
      "loss": 4.9462,
      "step": 80
    },
    {
      "epoch": 0.703125,
      "grad_norm": 15.375,
      "learning_rate": 9e-05,
      "loss": 3.5646,
      "step": 90
    },
    {
      "epoch": 0.78125,
      "grad_norm": 15.375,
      "learning_rate": 0.0001,
      "loss": 2.0469,
      "step": 100
    },
    {
      "epoch": 0.859375,
      "grad_norm": 3.71875,
      "learning_rate": 9.915254237288136e-05,
      "loss": 0.6762,
      "step": 110
    },
    {
      "epoch": 0.9375,
      "grad_norm": 3.734375,
      "learning_rate": 9.830508474576272e-05,
      "loss": 0.3449,
      "step": 120
    },
    {
      "epoch": 1.015625,
      "grad_norm": 2.609375,
      "learning_rate": 9.745762711864407e-05,
      "loss": 0.2995,
      "step": 130
    },
    {
      "epoch": 1.09375,
      "grad_norm": 2.375,
      "learning_rate": 9.661016949152543e-05,
      "loss": 0.2842,
      "step": 140
    },
    {
      "epoch": 1.171875,
      "grad_norm": 1.703125,
      "learning_rate": 9.576271186440679e-05,
      "loss": 0.2791,
      "step": 150
    },
    {
      "epoch": 1.25,
      "grad_norm": 1.0390625,
      "learning_rate": 9.491525423728815e-05,
      "loss": 0.2459,
      "step": 160
    },
    {
      "epoch": 1.328125,
      "grad_norm": 1.6484375,
      "learning_rate": 9.40677966101695e-05,
      "loss": 0.2479,
      "step": 170
    },
    {
      "epoch": 1.40625,
      "grad_norm": 1.0,
      "learning_rate": 9.322033898305085e-05,
      "loss": 0.2573,
      "step": 180
    },
    {
      "epoch": 1.484375,
      "grad_norm": 1.1171875,
      "learning_rate": 9.237288135593221e-05,
      "loss": 0.2424,
      "step": 190
    },
    {
      "epoch": 1.5625,
      "grad_norm": 0.8125,
      "learning_rate": 9.152542372881357e-05,
      "loss": 0.2339,
      "step": 200
    },
    {
      "epoch": 1.640625,
      "grad_norm": 0.95703125,
      "learning_rate": 9.067796610169493e-05,
      "loss": 0.2326,
      "step": 210
    },
    {
      "epoch": 1.71875,
      "grad_norm": 1.90625,
      "learning_rate": 8.983050847457629e-05,
      "loss": 0.2267,
      "step": 220
    },
    {
      "epoch": 1.796875,
      "grad_norm": 0.88671875,
      "learning_rate": 8.898305084745763e-05,
      "loss": 0.2233,
      "step": 230
    },
    {
      "epoch": 1.875,
      "grad_norm": 1.109375,
      "learning_rate": 8.813559322033899e-05,
      "loss": 0.2211,
      "step": 240
    },
    {
      "epoch": 1.953125,
      "grad_norm": 0.80859375,
      "learning_rate": 8.728813559322035e-05,
      "loss": 0.2287,
      "step": 250
    },
    {
      "epoch": 2.03125,
      "grad_norm": 0.87109375,
      "learning_rate": 8.644067796610171e-05,
      "loss": 0.2169,
      "step": 260
    },
    {
      "epoch": 2.109375,
      "grad_norm": 1.2890625,
      "learning_rate": 8.559322033898305e-05,
      "loss": 0.2161,
      "step": 270
    },
    {
      "epoch": 2.1875,
      "grad_norm": 0.75390625,
      "learning_rate": 8.474576271186441e-05,
      "loss": 0.2192,
      "step": 280
    },
    {
      "epoch": 2.265625,
      "grad_norm": 0.56640625,
      "learning_rate": 8.389830508474577e-05,
      "loss": 0.2136,
      "step": 290
    },
    {
      "epoch": 2.34375,
      "grad_norm": 0.7265625,
      "learning_rate": 8.305084745762712e-05,
      "loss": 0.1974,
      "step": 300
    },
    {
      "epoch": 2.421875,
      "grad_norm": 0.78515625,
      "learning_rate": 8.220338983050848e-05,
      "loss": 0.2046,
      "step": 310
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.8984375,
      "learning_rate": 8.135593220338983e-05,
      "loss": 0.1961,
      "step": 320
    },
    {
      "epoch": 2.578125,
      "grad_norm": 0.51171875,
      "learning_rate": 8.050847457627118e-05,
      "loss": 0.196,
      "step": 330
    },
    {
      "epoch": 2.65625,
      "grad_norm": 0.5234375,
      "learning_rate": 7.966101694915254e-05,
      "loss": 0.1955,
      "step": 340
    },
    {
      "epoch": 2.734375,
      "grad_norm": 0.8125,
      "learning_rate": 7.88135593220339e-05,
      "loss": 0.1867,
      "step": 350
    },
    {
      "epoch": 2.8125,
      "grad_norm": 0.57421875,
      "learning_rate": 7.796610169491526e-05,
      "loss": 0.184,
      "step": 360
    },
    {
      "epoch": 2.890625,
      "grad_norm": 0.5,
      "learning_rate": 7.711864406779662e-05,
      "loss": 0.1911,
      "step": 370
    },
    {
      "epoch": 2.96875,
      "grad_norm": 0.72265625,
      "learning_rate": 7.627118644067796e-05,
      "loss": 0.185,
      "step": 380
    },
    {
      "epoch": 3.046875,
      "grad_norm": 0.87109375,
      "learning_rate": 7.542372881355932e-05,
      "loss": 0.1682,
      "step": 390
    },
    {
      "epoch": 3.125,
      "grad_norm": 0.94140625,
      "learning_rate": 7.457627118644068e-05,
      "loss": 0.166,
      "step": 400
    },
    {
      "epoch": 3.203125,
      "grad_norm": 0.73828125,
      "learning_rate": 7.372881355932204e-05,
      "loss": 0.1683,
      "step": 410
    },
    {
      "epoch": 3.28125,
      "grad_norm": 0.640625,
      "learning_rate": 7.288135593220338e-05,
      "loss": 0.1672,
      "step": 420
    },
    {
      "epoch": 3.359375,
      "grad_norm": 0.478515625,
      "learning_rate": 7.203389830508474e-05,
      "loss": 0.1518,
      "step": 430
    },
    {
      "epoch": 3.4375,
      "grad_norm": 0.6796875,
      "learning_rate": 7.11864406779661e-05,
      "loss": 0.1555,
      "step": 440
    },
    {
      "epoch": 3.515625,
      "grad_norm": 0.5,
      "learning_rate": 7.033898305084746e-05,
      "loss": 0.1632,
      "step": 450
    },
    {
      "epoch": 3.59375,
      "grad_norm": 15.8125,
      "learning_rate": 6.949152542372882e-05,
      "loss": 0.1647,
      "step": 460
    },
    {
      "epoch": 3.671875,
      "grad_norm": 0.71484375,
      "learning_rate": 6.864406779661017e-05,
      "loss": 0.156,
      "step": 470
    },
    {
      "epoch": 3.75,
      "grad_norm": 1.28125,
      "learning_rate": 6.779661016949152e-05,
      "loss": 0.1445,
      "step": 480
    },
    {
      "epoch": 3.828125,
      "grad_norm": 0.6171875,
      "learning_rate": 6.694915254237288e-05,
      "loss": 0.125,
      "step": 490
    },
    {
      "epoch": 3.90625,
      "grad_norm": 2.015625,
      "learning_rate": 6.610169491525424e-05,
      "loss": 0.1408,
      "step": 500
    },
    {
      "epoch": 3.984375,
      "grad_norm": 8.9375,
      "learning_rate": 6.52542372881356e-05,
      "loss": 0.1295,
      "step": 510
    },
    {
      "epoch": 4.0625,
      "grad_norm": 0.96484375,
      "learning_rate": 6.440677966101695e-05,
      "loss": 0.127,
      "step": 520
    },
    {
      "epoch": 4.140625,
      "grad_norm": 0.71875,
      "learning_rate": 6.35593220338983e-05,
      "loss": 0.1191,
      "step": 530
    },
    {
      "epoch": 4.21875,
      "grad_norm": 0.447265625,
      "learning_rate": 6.271186440677966e-05,
      "loss": 0.1288,
      "step": 540
    },
    {
      "epoch": 4.296875,
      "grad_norm": 0.8046875,
      "learning_rate": 6.186440677966102e-05,
      "loss": 0.105,
      "step": 550
    },
    {
      "epoch": 4.375,
      "grad_norm": 0.83203125,
      "learning_rate": 6.101694915254238e-05,
      "loss": 0.1171,
      "step": 560
    },
    {
      "epoch": 4.453125,
      "grad_norm": 1.3515625,
      "learning_rate": 6.016949152542373e-05,
      "loss": 0.1279,
      "step": 570
    },
    {
      "epoch": 4.53125,
      "grad_norm": 0.8203125,
      "learning_rate": 5.932203389830509e-05,
      "loss": 0.1049,
      "step": 580
    },
    {
      "epoch": 4.609375,
      "grad_norm": 31.0,
      "learning_rate": 5.8474576271186446e-05,
      "loss": 0.1511,
      "step": 590
    },
    {
      "epoch": 4.6875,
      "grad_norm": 0.54296875,
      "learning_rate": 5.76271186440678e-05,
      "loss": 0.1048,
      "step": 600
    },
    {
      "epoch": 4.765625,
      "grad_norm": 1.328125,
      "learning_rate": 5.677966101694916e-05,
      "loss": 0.1087,
      "step": 610
    },
    {
      "epoch": 4.84375,
      "grad_norm": 0.81640625,
      "learning_rate": 5.593220338983051e-05,
      "loss": 0.1018,
      "step": 620
    },
    {
      "epoch": 4.921875,
      "grad_norm": 1.25,
      "learning_rate": 5.508474576271186e-05,
      "loss": 0.1116,
      "step": 630
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7734375,
      "learning_rate": 5.423728813559322e-05,
      "loss": 0.1077,
      "step": 640
    },
    {
      "epoch": 5.078125,
      "grad_norm": 0.52734375,
      "learning_rate": 5.338983050847458e-05,
      "loss": 0.1047,
      "step": 650
    },
    {
      "epoch": 5.15625,
      "grad_norm": 0.71875,
      "learning_rate": 5.254237288135594e-05,
      "loss": 0.101,
      "step": 660
    },
    {
      "epoch": 5.234375,
      "grad_norm": 0.66796875,
      "learning_rate": 5.1694915254237284e-05,
      "loss": 0.0999,
      "step": 670
    },
    {
      "epoch": 5.3125,
      "grad_norm": 0.498046875,
      "learning_rate": 5.0847457627118643e-05,
      "loss": 0.0964,
      "step": 680
    },
    {
      "epoch": 5.390625,
      "grad_norm": 0.52734375,
      "learning_rate": 5e-05,
      "loss": 0.1011,
      "step": 690
    },
    {
      "epoch": 5.46875,
      "grad_norm": 0.78515625,
      "learning_rate": 4.915254237288136e-05,
      "loss": 0.1019,
      "step": 700
    },
    {
      "epoch": 5.546875,
      "grad_norm": 0.7734375,
      "learning_rate": 4.8305084745762714e-05,
      "loss": 0.1089,
      "step": 710
    },
    {
      "epoch": 5.625,
      "grad_norm": 0.6953125,
      "learning_rate": 4.745762711864407e-05,
      "loss": 0.0998,
      "step": 720
    },
    {
      "epoch": 5.703125,
      "grad_norm": 10.375,
      "learning_rate": 4.6610169491525425e-05,
      "loss": 0.1138,
      "step": 730
    },
    {
      "epoch": 5.78125,
      "grad_norm": 0.67578125,
      "learning_rate": 4.5762711864406784e-05,
      "loss": 0.1189,
      "step": 740
    },
    {
      "epoch": 5.859375,
      "grad_norm": 0.6328125,
      "learning_rate": 4.491525423728814e-05,
      "loss": 0.1006,
      "step": 750
    },
    {
      "epoch": 5.9375,
      "grad_norm": 0.90625,
      "learning_rate": 4.4067796610169495e-05,
      "loss": 0.1049,
      "step": 760
    },
    {
      "epoch": 6.015625,
      "grad_norm": 0.46875,
      "learning_rate": 4.3220338983050854e-05,
      "loss": 0.0988,
      "step": 770
    },
    {
      "epoch": 6.09375,
      "grad_norm": 0.953125,
      "learning_rate": 4.2372881355932206e-05,
      "loss": 0.1061,
      "step": 780
    },
    {
      "epoch": 6.171875,
      "grad_norm": 0.51171875,
      "learning_rate": 4.152542372881356e-05,
      "loss": 0.0999,
      "step": 790
    },
    {
      "epoch": 6.25,
      "grad_norm": 0.52734375,
      "learning_rate": 4.067796610169492e-05,
      "loss": 0.102,
      "step": 800
    },
    {
      "epoch": 6.328125,
      "grad_norm": 0.98828125,
      "learning_rate": 3.983050847457627e-05,
      "loss": 0.0933,
      "step": 810
    },
    {
      "epoch": 6.40625,
      "grad_norm": 0.45703125,
      "learning_rate": 3.898305084745763e-05,
      "loss": 0.0963,
      "step": 820
    },
    {
      "epoch": 6.484375,
      "grad_norm": 0.93359375,
      "learning_rate": 3.813559322033898e-05,
      "loss": 0.1018,
      "step": 830
    },
    {
      "epoch": 6.5625,
      "grad_norm": 0.6484375,
      "learning_rate": 3.728813559322034e-05,
      "loss": 0.1038,
      "step": 840
    },
    {
      "epoch": 6.640625,
      "grad_norm": 0.51953125,
      "learning_rate": 3.644067796610169e-05,
      "loss": 0.0978,
      "step": 850
    },
    {
      "epoch": 6.71875,
      "grad_norm": 0.515625,
      "learning_rate": 3.559322033898305e-05,
      "loss": 0.1033,
      "step": 860
    },
    {
      "epoch": 6.796875,
      "grad_norm": 0.58203125,
      "learning_rate": 3.474576271186441e-05,
      "loss": 0.0998,
      "step": 870
    },
    {
      "epoch": 6.875,
      "grad_norm": 0.546875,
      "learning_rate": 3.389830508474576e-05,
      "loss": 0.0943,
      "step": 880
    },
    {
      "epoch": 6.953125,
      "grad_norm": 0.796875,
      "learning_rate": 3.305084745762712e-05,
      "loss": 0.1026,
      "step": 890
    },
    {
      "epoch": 7.03125,
      "grad_norm": 0.60546875,
      "learning_rate": 3.2203389830508473e-05,
      "loss": 0.1009,
      "step": 900
    },
    {
      "epoch": 7.109375,
      "grad_norm": 0.82421875,
      "learning_rate": 3.135593220338983e-05,
      "loss": 0.0949,
      "step": 910
    },
    {
      "epoch": 7.1875,
      "grad_norm": 0.51953125,
      "learning_rate": 3.050847457627119e-05,
      "loss": 0.0969,
      "step": 920
    },
    {
      "epoch": 7.265625,
      "grad_norm": 0.60546875,
      "learning_rate": 2.9661016949152544e-05,
      "loss": 0.0965,
      "step": 930
    },
    {
      "epoch": 7.34375,
      "grad_norm": 0.546875,
      "learning_rate": 2.88135593220339e-05,
      "loss": 0.0929,
      "step": 940
    },
    {
      "epoch": 7.421875,
      "grad_norm": 0.796875,
      "learning_rate": 2.7966101694915255e-05,
      "loss": 0.0982,
      "step": 950
    },
    {
      "epoch": 7.5,
      "grad_norm": 0.63671875,
      "learning_rate": 2.711864406779661e-05,
      "loss": 0.1018,
      "step": 960
    },
    {
      "epoch": 7.578125,
      "grad_norm": 0.7265625,
      "learning_rate": 2.627118644067797e-05,
      "loss": 0.0926,
      "step": 970
    },
    {
      "epoch": 7.65625,
      "grad_norm": 0.8203125,
      "learning_rate": 2.5423728813559322e-05,
      "loss": 0.0944,
      "step": 980
    },
    {
      "epoch": 7.734375,
      "grad_norm": 0.59375,
      "learning_rate": 2.457627118644068e-05,
      "loss": 0.0957,
      "step": 990
    },
    {
      "epoch": 7.8125,
      "grad_norm": 2.703125,
      "learning_rate": 2.3728813559322036e-05,
      "loss": 0.0974,
      "step": 1000
    },
    {
      "epoch": 7.890625,
      "grad_norm": 0.77734375,
      "learning_rate": 2.2881355932203392e-05,
      "loss": 0.1005,
      "step": 1010
    },
    {
      "epoch": 7.96875,
      "grad_norm": 0.70703125,
      "learning_rate": 2.2033898305084748e-05,
      "loss": 0.0988,
      "step": 1020
    },
    {
      "epoch": 8.046875,
      "grad_norm": 0.73046875,
      "learning_rate": 2.1186440677966103e-05,
      "loss": 0.0968,
      "step": 1030
    },
    {
      "epoch": 8.125,
      "grad_norm": 0.7265625,
      "learning_rate": 2.033898305084746e-05,
      "loss": 0.0951,
      "step": 1040
    },
    {
      "epoch": 8.203125,
      "grad_norm": 0.6953125,
      "learning_rate": 1.9491525423728814e-05,
      "loss": 0.0925,
      "step": 1050
    },
    {
      "epoch": 8.28125,
      "grad_norm": 0.498046875,
      "learning_rate": 1.864406779661017e-05,
      "loss": 0.0953,
      "step": 1060
    },
    {
      "epoch": 8.359375,
      "grad_norm": 0.546875,
      "learning_rate": 1.7796610169491526e-05,
      "loss": 0.0966,
      "step": 1070
    },
    {
      "epoch": 8.4375,
      "grad_norm": 0.6796875,
      "learning_rate": 1.694915254237288e-05,
      "loss": 0.1083,
      "step": 1080
    },
    {
      "epoch": 8.515625,
      "grad_norm": 0.70703125,
      "learning_rate": 1.6101694915254237e-05,
      "loss": 0.0992,
      "step": 1090
    },
    {
      "epoch": 8.59375,
      "grad_norm": 0.52734375,
      "learning_rate": 1.5254237288135596e-05,
      "loss": 0.11,
      "step": 1100
    },
    {
      "epoch": 8.671875,
      "grad_norm": 0.58984375,
      "learning_rate": 1.440677966101695e-05,
      "loss": 0.1163,
      "step": 1110
    },
    {
      "epoch": 8.75,
      "grad_norm": 0.51953125,
      "learning_rate": 1.3559322033898305e-05,
      "loss": 0.1086,
      "step": 1120
    },
    {
      "epoch": 8.828125,
      "grad_norm": 0.93359375,
      "learning_rate": 1.2711864406779661e-05,
      "loss": 0.0908,
      "step": 1130
    },
    {
      "epoch": 8.90625,
      "grad_norm": 0.765625,
      "learning_rate": 1.1864406779661018e-05,
      "loss": 0.0967,
      "step": 1140
    },
    {
      "epoch": 8.984375,
      "grad_norm": 0.64453125,
      "learning_rate": 1.1016949152542374e-05,
      "loss": 0.0959,
      "step": 1150
    },
    {
      "epoch": 9.0625,
      "grad_norm": 0.5546875,
      "learning_rate": 1.016949152542373e-05,
      "loss": 0.0855,
      "step": 1160
    }
  ],
  "logging_steps": 10,
  "max_steps": 1280,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 4466660675659776.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}