{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.1060723371308483,
  "eval_steps": 500,
  "global_step": 10000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.011060723371308484,
      "grad_norm": 375.9931945800781,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 166.3671,
      "step": 100
    },
    {
      "epoch": 0.022121446742616967,
      "grad_norm": 400.7745361328125,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 81.792,
      "step": 200
    },
    {
      "epoch": 0.03318217011392545,
      "grad_norm": 905.6781616210938,
      "learning_rate": 3e-06,
      "loss": 122.6817,
      "step": 300
    },
    {
      "epoch": 0.044242893485233935,
      "grad_norm": 1132.2952880859375,
      "learning_rate": 4.000000000000001e-06,
      "loss": 77.7319,
      "step": 400
    },
    {
      "epoch": 0.05530361685654242,
      "grad_norm": 688.6167602539062,
      "learning_rate": 5e-06,
      "loss": 78.9042,
      "step": 500
    },
    {
      "epoch": 0.0663643402278509,
      "grad_norm": 703.4044799804688,
      "learning_rate": 6e-06,
      "loss": 83.0117,
      "step": 600
    },
    {
      "epoch": 0.07742506359915939,
      "grad_norm": 1175.0980224609375,
      "learning_rate": 7e-06,
      "loss": 157.7761,
      "step": 700
    },
    {
      "epoch": 0.08848578697046787,
      "grad_norm": 1873.5474853515625,
      "learning_rate": 8.000000000000001e-06,
      "loss": 126.2309,
      "step": 800
    },
    {
      "epoch": 0.09954651034177635,
      "grad_norm": 547.4945068359375,
      "learning_rate": 9e-06,
      "loss": 75.2334,
      "step": 900
    },
    {
      "epoch": 0.11060723371308484,
      "grad_norm": 573.0324096679688,
      "learning_rate": 1e-05,
      "loss": 81.3112,
      "step": 1000
    },
    {
      "epoch": 0.12166795708439332,
      "grad_norm": 680.4608764648438,
      "learning_rate": 9.88888888888889e-06,
      "loss": 60.3122,
      "step": 1100
    },
    {
      "epoch": 0.1327286804557018,
      "grad_norm": 414.4683837890625,
      "learning_rate": 9.777777777777779e-06,
      "loss": 109.5748,
      "step": 1200
    },
    {
      "epoch": 0.14378940382701028,
      "grad_norm": 168.61688232421875,
      "learning_rate": 9.666666666666667e-06,
      "loss": 103.7934,
      "step": 1300
    },
    {
      "epoch": 0.15485012719831878,
      "grad_norm": 1966.074462890625,
      "learning_rate": 9.555555555555556e-06,
      "loss": 67.2399,
      "step": 1400
    },
    {
      "epoch": 0.16591085056962726,
      "grad_norm": 514.1409301757812,
      "learning_rate": 9.444444444444445e-06,
      "loss": 75.3922,
      "step": 1500
    },
    {
      "epoch": 0.17697157394093574,
      "grad_norm": 732.1213989257812,
      "learning_rate": 9.333333333333334e-06,
      "loss": 51.2073,
      "step": 1600
    },
    {
      "epoch": 0.18803229731224422,
      "grad_norm": 225.4730987548828,
      "learning_rate": 9.222222222222224e-06,
      "loss": 87.8553,
      "step": 1700
    },
    {
      "epoch": 0.1990930206835527,
      "grad_norm": 1147.8402099609375,
      "learning_rate": 9.111111111111112e-06,
      "loss": 72.9169,
      "step": 1800
    },
    {
      "epoch": 0.21015374405486117,
      "grad_norm": 393.7732238769531,
      "learning_rate": 9e-06,
      "loss": 102.8502,
      "step": 1900
    },
    {
      "epoch": 0.22121446742616968,
      "grad_norm": 1204.0518798828125,
      "learning_rate": 8.888888888888888e-06,
      "loss": 93.0795,
      "step": 2000
    },
    {
      "epoch": 0.23227519079747816,
      "grad_norm": 663.1642456054688,
      "learning_rate": 8.777777777777778e-06,
      "loss": 88.786,
      "step": 2100
    },
    {
      "epoch": 0.24333591416878664,
      "grad_norm": 1716.8380126953125,
      "learning_rate": 8.666666666666668e-06,
      "loss": 57.2407,
      "step": 2200
    },
    {
      "epoch": 0.2543966375400951,
      "grad_norm": 1476.3135986328125,
      "learning_rate": 8.555555555555556e-06,
      "loss": 73.0435,
      "step": 2300
    },
    {
      "epoch": 0.2654573609114036,
      "grad_norm": 800.91650390625,
      "learning_rate": 8.444444444444446e-06,
      "loss": 84.3321,
      "step": 2400
    },
    {
      "epoch": 0.2765180842827121,
      "grad_norm": 1679.648681640625,
      "learning_rate": 8.333333333333334e-06,
      "loss": 69.3359,
      "step": 2500
    },
    {
      "epoch": 0.28757880765402055,
      "grad_norm": 832.7955322265625,
      "learning_rate": 8.222222222222222e-06,
      "loss": 98.2768,
      "step": 2600
    },
    {
      "epoch": 0.29863953102532903,
      "grad_norm": 2618.217529296875,
      "learning_rate": 8.111111111111112e-06,
      "loss": 77.8276,
      "step": 2700
    },
    {
      "epoch": 0.30970025439663756,
      "grad_norm": 1804.9989013671875,
      "learning_rate": 8.000000000000001e-06,
      "loss": 56.5917,
      "step": 2800
    },
    {
      "epoch": 0.32076097776794604,
      "grad_norm": 174.3721160888672,
      "learning_rate": 7.88888888888889e-06,
      "loss": 51.9428,
      "step": 2900
    },
    {
      "epoch": 0.3318217011392545,
      "grad_norm": 169.36148071289062,
      "learning_rate": 7.77777777777778e-06,
      "loss": 82.8076,
      "step": 3000
    },
    {
      "epoch": 0.342882424510563,
      "grad_norm": 627.05419921875,
      "learning_rate": 7.666666666666667e-06,
      "loss": 52.4689,
      "step": 3100
    },
    {
      "epoch": 0.3539431478818715,
      "grad_norm": 641.0360717773438,
      "learning_rate": 7.555555555555556e-06,
      "loss": 100.1459,
      "step": 3200
    },
    {
      "epoch": 0.36500387125317996,
      "grad_norm": 268.31854248046875,
      "learning_rate": 7.444444444444445e-06,
      "loss": 115.3202,
      "step": 3300
    },
    {
      "epoch": 0.37606459462448844,
      "grad_norm": 400.4610900878906,
      "learning_rate": 7.333333333333333e-06,
      "loss": 70.1518,
      "step": 3400
    },
    {
      "epoch": 0.3871253179957969,
      "grad_norm": 786.57421875,
      "learning_rate": 7.222222222222223e-06,
      "loss": 61.8627,
      "step": 3500
    },
    {
      "epoch": 0.3981860413671054,
      "grad_norm": 2080.04541015625,
      "learning_rate": 7.111111111111112e-06,
      "loss": 55.4311,
      "step": 3600
    },
    {
      "epoch": 0.40924676473841387,
      "grad_norm": 850.0482177734375,
      "learning_rate": 7e-06,
      "loss": 48.0213,
      "step": 3700
    },
    {
      "epoch": 0.42030748810972235,
      "grad_norm": 399.0281677246094,
      "learning_rate": 6.88888888888889e-06,
      "loss": 105.0593,
      "step": 3800
    },
    {
      "epoch": 0.4313682114810309,
      "grad_norm": 627.0266723632812,
      "learning_rate": 6.777777777777779e-06,
      "loss": 78.3507,
      "step": 3900
    },
    {
      "epoch": 0.44242893485233936,
      "grad_norm": 227.25367736816406,
      "learning_rate": 6.666666666666667e-06,
      "loss": 76.2969,
      "step": 4000
    },
    {
      "epoch": 0.45348965822364784,
      "grad_norm": 522.318603515625,
      "learning_rate": 6.555555555555556e-06,
      "loss": 65.638,
      "step": 4100
    },
    {
      "epoch": 0.4645503815949563,
      "grad_norm": 934.457763671875,
      "learning_rate": 6.444444444444445e-06,
      "loss": 64.3987,
      "step": 4200
    },
    {
      "epoch": 0.4756111049662648,
      "grad_norm": 1235.8519287109375,
      "learning_rate": 6.333333333333333e-06,
      "loss": 90.3662,
      "step": 4300
    },
    {
      "epoch": 0.4866718283375733,
      "grad_norm": 70.5522232055664,
      "learning_rate": 6.222222222222223e-06,
      "loss": 116.0307,
      "step": 4400
    },
    {
      "epoch": 0.49773255170888175,
      "grad_norm": 849.7728271484375,
      "learning_rate": 6.111111111111112e-06,
      "loss": 63.1055,
      "step": 4500
    },
    {
      "epoch": 0.5087932750801902,
      "grad_norm": 200.14434814453125,
      "learning_rate": 6e-06,
      "loss": 55.2083,
      "step": 4600
    },
    {
      "epoch": 0.5198539984514987,
      "grad_norm": 622.9990234375,
      "learning_rate": 5.88888888888889e-06,
      "loss": 52.0711,
      "step": 4700
    },
    {
      "epoch": 0.5309147218228072,
      "grad_norm": 647.577392578125,
      "learning_rate": 5.777777777777778e-06,
      "loss": 85.6683,
      "step": 4800
    },
    {
      "epoch": 0.5419754451941157,
      "grad_norm": 680.5855712890625,
      "learning_rate": 5.666666666666667e-06,
      "loss": 100.0637,
      "step": 4900
    },
    {
      "epoch": 0.5530361685654241,
      "grad_norm": 364.47796630859375,
      "learning_rate": 5.555555555555557e-06,
      "loss": 53.0132,
      "step": 5000
    },
    {
      "epoch": 0.5640968919367326,
      "grad_norm": 28.153059005737305,
      "learning_rate": 5.444444444444445e-06,
      "loss": 81.5576,
      "step": 5100
    },
    {
      "epoch": 0.5751576153080411,
      "grad_norm": 334.4709777832031,
      "learning_rate": 5.333333333333334e-06,
      "loss": 67.3312,
      "step": 5200
    },
    {
      "epoch": 0.5862183386793496,
      "grad_norm": 1089.625,
      "learning_rate": 5.2222222222222226e-06,
      "loss": 50.9502,
      "step": 5300
    },
    {
      "epoch": 0.5972790620506581,
      "grad_norm": 444.6000061035156,
      "learning_rate": 5.1111111111111115e-06,
      "loss": 43.0887,
      "step": 5400
    },
    {
      "epoch": 0.6083397854219667,
      "grad_norm": 428.47821044921875,
      "learning_rate": 5e-06,
      "loss": 76.2733,
      "step": 5500
    },
    {
      "epoch": 0.6194005087932751,
      "grad_norm": 12874.7021484375,
      "learning_rate": 4.888888888888889e-06,
      "loss": 117.3838,
      "step": 5600
    },
    {
      "epoch": 0.6304612321645836,
      "grad_norm": 366.2122497558594,
      "learning_rate": 4.777777777777778e-06,
      "loss": 68.4892,
      "step": 5700
    },
    {
      "epoch": 0.6415219555358921,
      "grad_norm": 707.5550537109375,
      "learning_rate": 4.666666666666667e-06,
      "loss": 74.8324,
      "step": 5800
    },
    {
      "epoch": 0.6525826789072006,
      "grad_norm": 368.8702697753906,
      "learning_rate": 4.555555555555556e-06,
      "loss": 49.0018,
      "step": 5900
    },
    {
      "epoch": 0.663643402278509,
      "grad_norm": 961.3125,
      "learning_rate": 4.444444444444444e-06,
      "loss": 38.9093,
      "step": 6000
    },
    {
      "epoch": 0.6747041256498175,
      "grad_norm": 384.34490966796875,
      "learning_rate": 4.333333333333334e-06,
      "loss": 48.6872,
      "step": 6100
    },
    {
      "epoch": 0.685764849021126,
      "grad_norm": 756.4180297851562,
      "learning_rate": 4.222222222222223e-06,
      "loss": 51.1286,
      "step": 6200
    },
    {
      "epoch": 0.6968255723924345,
      "grad_norm": 452.16973876953125,
      "learning_rate": 4.111111111111111e-06,
      "loss": 64.0528,
      "step": 6300
    },
    {
      "epoch": 0.707886295763743,
      "grad_norm": 951.4662475585938,
      "learning_rate": 4.000000000000001e-06,
      "loss": 38.149,
      "step": 6400
    },
    {
      "epoch": 0.7189470191350514,
      "grad_norm": 110.86688232421875,
      "learning_rate": 3.88888888888889e-06,
      "loss": 39.444,
      "step": 6500
    },
    {
      "epoch": 0.7300077425063599,
      "grad_norm": 241.35595703125,
      "learning_rate": 3.777777777777778e-06,
      "loss": 42.4553,
      "step": 6600
    },
    {
      "epoch": 0.7410684658776684,
      "grad_norm": 1219.59765625,
      "learning_rate": 3.6666666666666666e-06,
      "loss": 75.4871,
      "step": 6700
    },
    {
      "epoch": 0.7521291892489769,
      "grad_norm": 257.36553955078125,
      "learning_rate": 3.555555555555556e-06,
      "loss": 64.0951,
      "step": 6800
    },
    {
      "epoch": 0.7631899126202853,
      "grad_norm": 444.5009460449219,
      "learning_rate": 3.444444444444445e-06,
      "loss": 57.2763,
      "step": 6900
    },
    {
      "epoch": 0.7742506359915938,
      "grad_norm": 508.6488342285156,
      "learning_rate": 3.3333333333333333e-06,
      "loss": 49.7445,
      "step": 7000
    },
    {
      "epoch": 0.7853113593629023,
      "grad_norm": 371.861083984375,
      "learning_rate": 3.2222222222222227e-06,
      "loss": 49.065,
      "step": 7100
    },
    {
      "epoch": 0.7963720827342108,
      "grad_norm": 178.38575744628906,
      "learning_rate": 3.1111111111111116e-06,
      "loss": 35.8833,
      "step": 7200
    },
    {
      "epoch": 0.8074328061055193,
      "grad_norm": 379.4342346191406,
      "learning_rate": 3e-06,
      "loss": 74.4786,
      "step": 7300
    },
    {
      "epoch": 0.8184935294768277,
      "grad_norm": 1825.0079345703125,
      "learning_rate": 2.888888888888889e-06,
      "loss": 83.434,
      "step": 7400
    },
    {
      "epoch": 0.8295542528481362,
      "grad_norm": 225.136474609375,
      "learning_rate": 2.7777777777777783e-06,
      "loss": 49.8191,
      "step": 7500
    },
    {
      "epoch": 0.8406149762194447,
      "grad_norm": 1184.1737060546875,
      "learning_rate": 2.666666666666667e-06,
      "loss": 61.6105,
      "step": 7600
    },
    {
      "epoch": 0.8516756995907533,
      "grad_norm": 252.15902709960938,
      "learning_rate": 2.5555555555555557e-06,
      "loss": 73.2447,
      "step": 7700
    },
    {
      "epoch": 0.8627364229620618,
      "grad_norm": 973.193359375,
      "learning_rate": 2.4444444444444447e-06,
      "loss": 48.954,
      "step": 7800
    },
    {
      "epoch": 0.8737971463333702,
      "grad_norm": 677.3170166015625,
      "learning_rate": 2.3333333333333336e-06,
      "loss": 60.8879,
      "step": 7900
    },
    {
      "epoch": 0.8848578697046787,
      "grad_norm": 220.8177490234375,
      "learning_rate": 2.222222222222222e-06,
      "loss": 71.5866,
      "step": 8000
    },
    {
      "epoch": 0.8959185930759872,
      "grad_norm": 1242.9266357421875,
      "learning_rate": 2.1111111111111114e-06,
      "loss": 41.986,
      "step": 8100
    },
    {
      "epoch": 0.9069793164472957,
      "grad_norm": 537.8414306640625,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 34.6977,
      "step": 8200
    },
    {
      "epoch": 0.9180400398186042,
      "grad_norm": 485.3799743652344,
      "learning_rate": 1.888888888888889e-06,
      "loss": 71.6902,
      "step": 8300
    },
    {
      "epoch": 0.9291007631899126,
      "grad_norm": 336.6470947265625,
      "learning_rate": 1.777777777777778e-06,
      "loss": 53.9529,
      "step": 8400
    },
    {
      "epoch": 0.9401614865612211,
      "grad_norm": 1168.831787109375,
      "learning_rate": 1.6666666666666667e-06,
      "loss": 50.8682,
      "step": 8500
    },
    {
      "epoch": 0.9512222099325296,
      "grad_norm": 194.76914978027344,
      "learning_rate": 1.5555555555555558e-06,
      "loss": 74.3729,
      "step": 8600
    },
    {
      "epoch": 0.9622829333038381,
      "grad_norm": 1270.5018310546875,
      "learning_rate": 1.4444444444444445e-06,
      "loss": 88.4943,
      "step": 8700
    },
    {
      "epoch": 0.9733436566751466,
      "grad_norm": 351.2843017578125,
      "learning_rate": 1.3333333333333334e-06,
      "loss": 79.6729,
      "step": 8800
    },
    {
      "epoch": 0.984404380046455,
      "grad_norm": 35.92924880981445,
      "learning_rate": 1.2222222222222223e-06,
      "loss": 71.0744,
      "step": 8900
    },
    {
      "epoch": 0.9954651034177635,
      "grad_norm": 447.2767333984375,
      "learning_rate": 1.111111111111111e-06,
      "loss": 52.7668,
      "step": 9000
    },
    {
      "epoch": 1.0,
      "eval_loss": 60.977603912353516,
      "eval_runtime": 87.775,
      "eval_samples_per_second": 22.899,
      "eval_steps_per_second": 11.45,
      "step": 9041
    },
    {
      "epoch": 1.006525826789072,
      "grad_norm": 488.3475036621094,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 47.1465,
      "step": 9100
    },
    {
      "epoch": 1.0175865501603805,
      "grad_norm": 523.9360961914062,
      "learning_rate": 8.88888888888889e-07,
      "loss": 55.9864,
      "step": 9200
    },
    {
      "epoch": 1.028647273531689,
      "grad_norm": 473.5372619628906,
      "learning_rate": 7.777777777777779e-07,
      "loss": 60.8429,
      "step": 9300
    },
    {
      "epoch": 1.0397079969029974,
      "grad_norm": 2480.4365234375,
      "learning_rate": 6.666666666666667e-07,
      "loss": 53.3988,
      "step": 9400
    },
    {
      "epoch": 1.050768720274306,
      "grad_norm": 247.97398376464844,
      "learning_rate": 5.555555555555555e-07,
      "loss": 87.3904,
      "step": 9500
    },
    {
      "epoch": 1.0618294436456144,
      "grad_norm": 329.9159240722656,
      "learning_rate": 4.444444444444445e-07,
      "loss": 47.602,
      "step": 9600
    },
    {
      "epoch": 1.0728901670169229,
      "grad_norm": 210.892822265625,
      "learning_rate": 3.3333333333333335e-07,
      "loss": 61.5918,
      "step": 9700
    },
    {
      "epoch": 1.0839508903882313,
      "grad_norm": 255.29690551757812,
      "learning_rate": 2.2222222222222224e-07,
      "loss": 65.8683,
      "step": 9800
    },
    {
      "epoch": 1.0950116137595398,
      "grad_norm": 1664.9510498046875,
      "learning_rate": 1.1111111111111112e-07,
      "loss": 30.0188,
      "step": 9900
    },
    {
      "epoch": 1.1060723371308483,
      "grad_norm": 23.272554397583008,
      "learning_rate": 0.0,
      "loss": 46.2958,
      "step": 10000
    }
  ],
  "logging_steps": 100,
  "max_steps": 10000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 2,
  "save_steps": 5000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}