{
  "best_metric": 0.6666322350502014,
  "best_model_checkpoint": "./checkpoints/llava-v1.6-mistral/checkpoint-125",
  "epoch": 4.992,
  "eval_steps": 1.0,
  "global_step": 155,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.032,
      "grad_norm": 1.3562940564009667,
      "learning_rate": 0.0,
      "loss": 1.2069,
      "step": 1
    },
    {
      "epoch": 0.032,
      "eval_loss": 1.189530849456787,
      "eval_runtime": 100.6384,
      "eval_samples_per_second": 1.987,
      "eval_steps_per_second": 0.497,
      "step": 1
    },
    {
      "epoch": 0.064,
      "grad_norm": 1.3726335882966116,
      "learning_rate": 7.737056144690831e-06,
      "loss": 1.242,
      "step": 2
    },
    {
      "epoch": 0.064,
      "eval_loss": 1.189530849456787,
      "eval_runtime": 95.3713,
      "eval_samples_per_second": 2.097,
      "eval_steps_per_second": 0.524,
      "step": 2
    },
    {
      "epoch": 0.096,
      "grad_norm": 1.267071497789874,
      "learning_rate": 1.2262943855309169e-05,
      "loss": 1.1588,
      "step": 3
    },
    {
      "epoch": 0.096,
      "eval_loss": 1.1675384044647217,
      "eval_runtime": 99.4768,
      "eval_samples_per_second": 2.011,
      "eval_steps_per_second": 0.503,
      "step": 3
    },
    {
      "epoch": 0.128,
      "grad_norm": 1.2066781722006759,
      "learning_rate": 1.5474112289381662e-05,
      "loss": 1.1196,
      "step": 4
    },
    {
      "epoch": 0.128,
      "eval_loss": 1.1391011476516724,
      "eval_runtime": 95.4777,
      "eval_samples_per_second": 2.095,
      "eval_steps_per_second": 0.524,
      "step": 4
    },
    {
      "epoch": 0.16,
      "grad_norm": 1.2746210039866817,
      "learning_rate": 1.7964888034078543e-05,
      "loss": 1.1462,
      "step": 5
    },
    {
      "epoch": 0.16,
      "eval_loss": 1.1023523807525635,
      "eval_runtime": 95.4138,
      "eval_samples_per_second": 2.096,
      "eval_steps_per_second": 0.524,
      "step": 5
    },
    {
      "epoch": 0.192,
      "grad_norm": 1.0552650848120237,
      "learning_rate": 1.9999999999999998e-05,
      "loss": 1.1147,
      "step": 6
    },
    {
      "epoch": 0.192,
      "eval_loss": 1.063745379447937,
      "eval_runtime": 97.6633,
      "eval_samples_per_second": 2.048,
      "eval_steps_per_second": 0.512,
      "step": 6
    },
    {
      "epoch": 0.224,
      "grad_norm": 1.0779287740182022,
      "learning_rate": 2e-05,
      "loss": 1.1773,
      "step": 7
    },
    {
      "epoch": 0.224,
      "eval_loss": 1.024451732635498,
      "eval_runtime": 97.388,
      "eval_samples_per_second": 2.054,
      "eval_steps_per_second": 0.513,
      "step": 7
    },
    {
      "epoch": 0.256,
      "grad_norm": 0.9918200495785524,
      "learning_rate": 2e-05,
      "loss": 1.0434,
      "step": 8
    },
    {
      "epoch": 0.256,
      "eval_loss": 0.9906812906265259,
      "eval_runtime": 97.2423,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 8
    },
    {
      "epoch": 0.288,
      "grad_norm": 0.9119239308208719,
      "learning_rate": 2e-05,
      "loss": 1.0527,
      "step": 9
    },
    {
      "epoch": 0.288,
      "eval_loss": 0.9626247882843018,
      "eval_runtime": 97.2048,
      "eval_samples_per_second": 2.058,
      "eval_steps_per_second": 0.514,
      "step": 9
    },
    {
      "epoch": 0.32,
      "grad_norm": 0.7292073828818868,
      "learning_rate": 2e-05,
      "loss": 0.9801,
      "step": 10
    },
    {
      "epoch": 0.32,
      "eval_loss": 0.9422003030776978,
      "eval_runtime": 97.143,
      "eval_samples_per_second": 2.059,
      "eval_steps_per_second": 0.515,
      "step": 10
    },
    {
      "epoch": 0.352,
      "grad_norm": 0.8195137858461904,
      "learning_rate": 2e-05,
      "loss": 1.0104,
      "step": 11
    },
    {
      "epoch": 0.352,
      "eval_loss": 0.9264165759086609,
      "eval_runtime": 97.2279,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 11
    },
    {
      "epoch": 0.384,
      "grad_norm": 0.7060464565699565,
      "learning_rate": 2e-05,
      "loss": 0.9764,
      "step": 12
    },
    {
      "epoch": 0.384,
      "eval_loss": 0.9138455986976624,
      "eval_runtime": 97.0954,
      "eval_samples_per_second": 2.06,
      "eval_steps_per_second": 0.515,
      "step": 12
    },
    {
      "epoch": 0.416,
      "grad_norm": 0.8207130544099034,
      "learning_rate": 2e-05,
      "loss": 1.0371,
      "step": 13
    },
    {
      "epoch": 0.416,
      "eval_loss": 0.902658224105835,
      "eval_runtime": 100.8924,
      "eval_samples_per_second": 1.982,
      "eval_steps_per_second": 0.496,
      "step": 13
    },
    {
      "epoch": 0.448,
      "grad_norm": 0.7962544586919155,
      "learning_rate": 2e-05,
      "loss": 0.9455,
      "step": 14
    },
    {
      "epoch": 0.448,
      "eval_loss": 0.8926536440849304,
      "eval_runtime": 100.856,
      "eval_samples_per_second": 1.983,
      "eval_steps_per_second": 0.496,
      "step": 14
    },
    {
      "epoch": 0.48,
      "grad_norm": 0.8246421268957655,
      "learning_rate": 2e-05,
      "loss": 0.9456,
      "step": 15
    },
    {
      "epoch": 0.48,
      "eval_loss": 0.8823295831680298,
      "eval_runtime": 97.0568,
      "eval_samples_per_second": 2.061,
      "eval_steps_per_second": 0.515,
      "step": 15
    },
    {
      "epoch": 0.512,
      "grad_norm": 0.7040845257818263,
      "learning_rate": 2e-05,
      "loss": 0.9283,
      "step": 16
    },
    {
      "epoch": 0.512,
      "eval_loss": 0.8721897602081299,
      "eval_runtime": 98.318,
      "eval_samples_per_second": 2.034,
      "eval_steps_per_second": 0.509,
      "step": 16
    },
    {
      "epoch": 0.544,
      "grad_norm": 0.7904707927869743,
      "learning_rate": 2e-05,
      "loss": 0.9786,
      "step": 17
    },
    {
      "epoch": 0.544,
      "eval_loss": 0.861838698387146,
      "eval_runtime": 101.4487,
      "eval_samples_per_second": 1.971,
      "eval_steps_per_second": 0.493,
      "step": 17
    },
    {
      "epoch": 0.576,
      "grad_norm": 0.7605583546638608,
      "learning_rate": 2e-05,
      "loss": 0.9361,
      "step": 18
    },
    {
      "epoch": 0.576,
      "eval_loss": 0.8519415259361267,
      "eval_runtime": 98.2448,
      "eval_samples_per_second": 2.036,
      "eval_steps_per_second": 0.509,
      "step": 18
    },
    {
      "epoch": 0.608,
      "grad_norm": 0.7005232039249588,
      "learning_rate": 2e-05,
      "loss": 0.8707,
      "step": 19
    },
    {
      "epoch": 0.608,
      "eval_loss": 0.842995822429657,
      "eval_runtime": 98.3165,
      "eval_samples_per_second": 2.034,
      "eval_steps_per_second": 0.509,
      "step": 19
    },
    {
      "epoch": 0.64,
      "grad_norm": 0.653693259495668,
      "learning_rate": 2e-05,
      "loss": 0.9332,
      "step": 20
    },
    {
      "epoch": 0.64,
      "eval_loss": 0.8355565071105957,
      "eval_runtime": 98.237,
      "eval_samples_per_second": 2.036,
      "eval_steps_per_second": 0.509,
      "step": 20
    },
    {
      "epoch": 0.672,
      "grad_norm": 0.6600451404133434,
      "learning_rate": 2e-05,
      "loss": 0.8848,
      "step": 21
    },
    {
      "epoch": 0.672,
      "eval_loss": 0.8297473788261414,
      "eval_runtime": 98.1825,
      "eval_samples_per_second": 2.037,
      "eval_steps_per_second": 0.509,
      "step": 21
    },
    {
      "epoch": 0.704,
      "grad_norm": 0.6666525650519819,
      "learning_rate": 2e-05,
      "loss": 0.9261,
      "step": 22
    },
    {
      "epoch": 0.704,
      "eval_loss": 0.8248497247695923,
      "eval_runtime": 98.1541,
      "eval_samples_per_second": 2.038,
      "eval_steps_per_second": 0.509,
      "step": 22
    },
    {
      "epoch": 0.736,
      "grad_norm": 0.6759718697778233,
      "learning_rate": 2e-05,
      "loss": 1.0103,
      "step": 23
    },
    {
      "epoch": 0.736,
      "eval_loss": 0.8214980959892273,
      "eval_runtime": 98.2494,
      "eval_samples_per_second": 2.036,
      "eval_steps_per_second": 0.509,
      "step": 23
    },
    {
      "epoch": 0.768,
      "grad_norm": 0.651870418762904,
      "learning_rate": 2e-05,
      "loss": 0.9538,
      "step": 24
    },
    {
      "epoch": 0.768,
      "eval_loss": 0.818403959274292,
      "eval_runtime": 97.9476,
      "eval_samples_per_second": 2.042,
      "eval_steps_per_second": 0.51,
      "step": 24
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.6985872283132858,
      "learning_rate": 2e-05,
      "loss": 0.8216,
      "step": 25
    },
    {
      "epoch": 0.8,
      "eval_loss": 0.8149588108062744,
      "eval_runtime": 95.6175,
      "eval_samples_per_second": 2.092,
      "eval_steps_per_second": 0.523,
      "step": 25
    },
    {
      "epoch": 0.832,
      "grad_norm": 0.7801182369335715,
      "learning_rate": 2e-05,
      "loss": 0.8478,
      "step": 26
    },
    {
      "epoch": 0.832,
      "eval_loss": 0.8100517392158508,
      "eval_runtime": 99.7733,
      "eval_samples_per_second": 2.005,
      "eval_steps_per_second": 0.501,
      "step": 26
    },
    {
      "epoch": 0.864,
      "grad_norm": 0.6727766374526198,
      "learning_rate": 2e-05,
      "loss": 0.9015,
      "step": 27
    },
    {
      "epoch": 0.864,
      "eval_loss": 0.8051578998565674,
      "eval_runtime": 95.7387,
      "eval_samples_per_second": 2.089,
      "eval_steps_per_second": 0.522,
      "step": 27
    },
    {
      "epoch": 0.896,
      "grad_norm": 0.7398381040653764,
      "learning_rate": 2e-05,
      "loss": 0.8434,
      "step": 28
    },
    {
      "epoch": 0.896,
      "eval_loss": 0.7998965382575989,
      "eval_runtime": 95.7753,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.522,
      "step": 28
    },
    {
      "epoch": 0.928,
      "grad_norm": 0.6837341164482282,
      "learning_rate": 2e-05,
      "loss": 0.9124,
      "step": 29
    },
    {
      "epoch": 0.928,
      "eval_loss": 0.7946869134902954,
      "eval_runtime": 95.933,
      "eval_samples_per_second": 2.085,
      "eval_steps_per_second": 0.521,
      "step": 29
    },
    {
      "epoch": 0.96,
      "grad_norm": 0.7285508052728015,
      "learning_rate": 2e-05,
      "loss": 0.8252,
      "step": 30
    },
    {
      "epoch": 0.96,
      "eval_loss": 0.7891057729721069,
      "eval_runtime": 95.871,
      "eval_samples_per_second": 2.086,
      "eval_steps_per_second": 0.522,
      "step": 30
    },
    {
      "epoch": 0.992,
      "grad_norm": 0.706524788728722,
      "learning_rate": 2e-05,
      "loss": 0.8382,
      "step": 31
    },
    {
      "epoch": 0.992,
      "eval_loss": 0.7835636138916016,
      "eval_runtime": 95.7491,
      "eval_samples_per_second": 2.089,
      "eval_steps_per_second": 0.522,
      "step": 31
    },
    {
      "epoch": 1.024,
      "grad_norm": 0.7392763969473979,
      "learning_rate": 2e-05,
      "loss": 0.8586,
      "step": 32
    },
    {
      "epoch": 1.024,
      "eval_loss": 0.7780101299285889,
      "eval_runtime": 99.7493,
      "eval_samples_per_second": 2.005,
      "eval_steps_per_second": 0.501,
      "step": 32
    },
    {
      "epoch": 1.056,
      "grad_norm": 0.7407143277408553,
      "learning_rate": 2e-05,
      "loss": 0.8438,
      "step": 33
    },
    {
      "epoch": 1.056,
      "eval_loss": 0.7728690505027771,
      "eval_runtime": 95.8863,
      "eval_samples_per_second": 2.086,
      "eval_steps_per_second": 0.521,
      "step": 33
    },
    {
      "epoch": 1.088,
      "grad_norm": 0.7146296362894146,
      "learning_rate": 2e-05,
      "loss": 0.8357,
      "step": 34
    },
    {
      "epoch": 1.088,
      "eval_loss": 0.7683935761451721,
      "eval_runtime": 97.824,
      "eval_samples_per_second": 2.044,
      "eval_steps_per_second": 0.511,
      "step": 34
    },
    {
      "epoch": 1.12,
      "grad_norm": 0.7117142678212836,
      "learning_rate": 2e-05,
      "loss": 0.7884,
      "step": 35
    },
    {
      "epoch": 1.12,
      "eval_loss": 0.7646369338035583,
      "eval_runtime": 95.8185,
      "eval_samples_per_second": 2.087,
      "eval_steps_per_second": 0.522,
      "step": 35
    },
    {
      "epoch": 1.152,
      "grad_norm": 0.7735767817076163,
      "learning_rate": 2e-05,
      "loss": 0.8851,
      "step": 36
    },
    {
      "epoch": 1.152,
      "eval_loss": 0.7616337537765503,
      "eval_runtime": 96.1525,
      "eval_samples_per_second": 2.08,
      "eval_steps_per_second": 0.52,
      "step": 36
    },
    {
      "epoch": 1.184,
      "grad_norm": 0.7386952203362822,
      "learning_rate": 2e-05,
      "loss": 0.7943,
      "step": 37
    },
    {
      "epoch": 1.184,
      "eval_loss": 0.759408712387085,
      "eval_runtime": 96.2772,
      "eval_samples_per_second": 2.077,
      "eval_steps_per_second": 0.519,
      "step": 37
    },
    {
      "epoch": 1.216,
      "grad_norm": 0.6688305832985106,
      "learning_rate": 2e-05,
      "loss": 0.8433,
      "step": 38
    },
    {
      "epoch": 1.216,
      "eval_loss": 0.7577520608901978,
      "eval_runtime": 95.7726,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.522,
      "step": 38
    },
    {
      "epoch": 1.248,
      "grad_norm": 0.6861808214947546,
      "learning_rate": 2e-05,
      "loss": 0.8132,
      "step": 39
    },
    {
      "epoch": 1.248,
      "eval_loss": 0.7565059065818787,
      "eval_runtime": 96.1577,
      "eval_samples_per_second": 2.08,
      "eval_steps_per_second": 0.52,
      "step": 39
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.7450886292190618,
      "learning_rate": 2e-05,
      "loss": 0.9067,
      "step": 40
    },
    {
      "epoch": 1.28,
      "eval_loss": 0.7562046647071838,
      "eval_runtime": 95.9983,
      "eval_samples_per_second": 2.083,
      "eval_steps_per_second": 0.521,
      "step": 40
    },
    {
      "epoch": 1.312,
      "grad_norm": 0.7508586498301053,
      "learning_rate": 2e-05,
      "loss": 0.8642,
      "step": 41
    },
    {
      "epoch": 1.312,
      "eval_loss": 0.755351722240448,
      "eval_runtime": 96.6244,
      "eval_samples_per_second": 2.07,
      "eval_steps_per_second": 0.517,
      "step": 41
    },
    {
      "epoch": 1.3439999999999999,
      "grad_norm": 0.7717273184776977,
      "learning_rate": 2e-05,
      "loss": 0.9406,
      "step": 42
    },
    {
      "epoch": 1.3439999999999999,
      "eval_loss": 0.753773033618927,
      "eval_runtime": 100.8127,
      "eval_samples_per_second": 1.984,
      "eval_steps_per_second": 0.496,
      "step": 42
    },
    {
      "epoch": 1.376,
      "grad_norm": 0.7980115302909253,
      "learning_rate": 2e-05,
      "loss": 0.7732,
      "step": 43
    },
    {
      "epoch": 1.376,
      "eval_loss": 0.7511720061302185,
      "eval_runtime": 97.4372,
      "eval_samples_per_second": 2.053,
      "eval_steps_per_second": 0.513,
      "step": 43
    },
    {
      "epoch": 1.408,
      "grad_norm": 0.794554904521861,
      "learning_rate": 2e-05,
      "loss": 0.8962,
      "step": 44
    },
    {
      "epoch": 1.408,
      "eval_loss": 0.7484354376792908,
      "eval_runtime": 96.7621,
      "eval_samples_per_second": 2.067,
      "eval_steps_per_second": 0.517,
      "step": 44
    },
    {
      "epoch": 1.44,
      "grad_norm": 0.7964935137573818,
      "learning_rate": 2e-05,
      "loss": 0.7725,
      "step": 45
    },
    {
      "epoch": 1.44,
      "eval_loss": 0.7454251050949097,
      "eval_runtime": 96.5292,
      "eval_samples_per_second": 2.072,
      "eval_steps_per_second": 0.518,
      "step": 45
    },
    {
      "epoch": 1.472,
      "grad_norm": 0.7477437094752549,
      "learning_rate": 2e-05,
      "loss": 0.8238,
      "step": 46
    },
    {
      "epoch": 1.472,
      "eval_loss": 0.7427694201469421,
      "eval_runtime": 96.4788,
      "eval_samples_per_second": 2.073,
      "eval_steps_per_second": 0.518,
      "step": 46
    },
    {
      "epoch": 1.504,
      "grad_norm": 0.7443339980176984,
      "learning_rate": 2e-05,
      "loss": 0.8273,
      "step": 47
    },
    {
      "epoch": 1.504,
      "eval_loss": 0.7407447099685669,
      "eval_runtime": 96.5686,
      "eval_samples_per_second": 2.071,
      "eval_steps_per_second": 0.518,
      "step": 47
    },
    {
      "epoch": 1.536,
      "grad_norm": 0.7901404920428264,
      "learning_rate": 2e-05,
      "loss": 0.7974,
      "step": 48
    },
    {
      "epoch": 1.536,
      "eval_loss": 0.7381842136383057,
      "eval_runtime": 96.6071,
      "eval_samples_per_second": 2.07,
      "eval_steps_per_second": 0.518,
      "step": 48
    },
    {
      "epoch": 1.568,
      "grad_norm": 0.8677109521747711,
      "learning_rate": 2e-05,
      "loss": 0.8011,
      "step": 49
    },
    {
      "epoch": 1.568,
      "eval_loss": 0.73555588722229,
      "eval_runtime": 96.4189,
      "eval_samples_per_second": 2.074,
      "eval_steps_per_second": 0.519,
      "step": 49
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.8170436579363425,
      "learning_rate": 2e-05,
      "loss": 0.7762,
      "step": 50
    },
    {
      "epoch": 1.6,
      "eval_loss": 0.7331439256668091,
      "eval_runtime": 100.5875,
      "eval_samples_per_second": 1.988,
      "eval_steps_per_second": 0.497,
      "step": 50
    },
    {
      "epoch": 1.6320000000000001,
      "grad_norm": 0.8190539924095326,
      "learning_rate": 2e-05,
      "loss": 0.7469,
      "step": 51
    },
    {
      "epoch": 1.6320000000000001,
      "eval_loss": 0.7306647300720215,
      "eval_runtime": 101.1429,
      "eval_samples_per_second": 1.977,
      "eval_steps_per_second": 0.494,
      "step": 51
    },
    {
      "epoch": 1.6640000000000001,
      "grad_norm": 0.8044932603192197,
      "learning_rate": 2e-05,
      "loss": 0.8052,
      "step": 52
    },
    {
      "epoch": 1.6640000000000001,
      "eval_loss": 0.7287429571151733,
      "eval_runtime": 97.25,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 52
    },
    {
      "epoch": 1.696,
      "grad_norm": 0.7811170846407103,
      "learning_rate": 2e-05,
      "loss": 0.755,
      "step": 53
    },
    {
      "epoch": 1.696,
      "eval_loss": 0.7264651656150818,
      "eval_runtime": 97.67,
      "eval_samples_per_second": 2.048,
      "eval_steps_per_second": 0.512,
      "step": 53
    },
    {
      "epoch": 1.728,
      "grad_norm": 0.808405093976468,
      "learning_rate": 2e-05,
      "loss": 0.7838,
      "step": 54
    },
    {
      "epoch": 1.728,
      "eval_loss": 0.7242828607559204,
      "eval_runtime": 101.2008,
      "eval_samples_per_second": 1.976,
      "eval_steps_per_second": 0.494,
      "step": 54
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.8172106270954758,
      "learning_rate": 2e-05,
      "loss": 0.7123,
      "step": 55
    },
    {
      "epoch": 1.76,
      "eval_loss": 0.7217574119567871,
      "eval_runtime": 99.6145,
      "eval_samples_per_second": 2.008,
      "eval_steps_per_second": 0.502,
      "step": 55
    },
    {
      "epoch": 1.792,
      "grad_norm": 0.83212530473105,
      "learning_rate": 2e-05,
      "loss": 0.8141,
      "step": 56
    },
    {
      "epoch": 1.792,
      "eval_loss": 0.7192932367324829,
      "eval_runtime": 97.1377,
      "eval_samples_per_second": 2.059,
      "eval_steps_per_second": 0.515,
      "step": 56
    },
    {
      "epoch": 1.8239999999999998,
      "grad_norm": 0.8632612959683078,
      "learning_rate": 2e-05,
      "loss": 0.8005,
      "step": 57
    },
    {
      "epoch": 1.8239999999999998,
      "eval_loss": 0.717003345489502,
      "eval_runtime": 97.2871,
      "eval_samples_per_second": 2.056,
      "eval_steps_per_second": 0.514,
      "step": 57
    },
    {
      "epoch": 1.8559999999999999,
      "grad_norm": 0.8362081570841255,
      "learning_rate": 2e-05,
      "loss": 0.8254,
      "step": 58
    },
    {
      "epoch": 1.8559999999999999,
      "eval_loss": 0.715446949005127,
      "eval_runtime": 97.5751,
      "eval_samples_per_second": 2.05,
      "eval_steps_per_second": 0.512,
      "step": 58
    },
    {
      "epoch": 1.888,
      "grad_norm": 0.8862779841020042,
      "learning_rate": 2e-05,
      "loss": 0.7949,
      "step": 59
    },
    {
      "epoch": 1.888,
      "eval_loss": 0.7145370244979858,
      "eval_runtime": 97.2406,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 59
    },
    {
      "epoch": 1.92,
      "grad_norm": 0.875656256488838,
      "learning_rate": 2e-05,
      "loss": 0.8277,
      "step": 60
    },
    {
      "epoch": 1.92,
      "eval_loss": 0.7140547037124634,
      "eval_runtime": 97.1485,
      "eval_samples_per_second": 2.059,
      "eval_steps_per_second": 0.515,
      "step": 60
    },
    {
      "epoch": 1.952,
      "grad_norm": 0.8850690645543244,
      "learning_rate": 2e-05,
      "loss": 0.8117,
      "step": 61
    },
    {
      "epoch": 1.952,
      "eval_loss": 0.7128701210021973,
      "eval_runtime": 97.8906,
      "eval_samples_per_second": 2.043,
      "eval_steps_per_second": 0.511,
      "step": 61
    },
    {
      "epoch": 1.984,
      "grad_norm": 0.8961641824408927,
      "learning_rate": 2e-05,
      "loss": 0.704,
      "step": 62
    },
    {
      "epoch": 1.984,
      "eval_loss": 0.7107064127922058,
      "eval_runtime": 102.2163,
      "eval_samples_per_second": 1.957,
      "eval_steps_per_second": 0.489,
      "step": 62
    },
    {
      "epoch": 2.016,
      "grad_norm": 0.8602247372004996,
      "learning_rate": 2e-05,
      "loss": 0.8016,
      "step": 63
    },
    {
      "epoch": 2.016,
      "eval_loss": 0.7091581225395203,
      "eval_runtime": 98.2114,
      "eval_samples_per_second": 2.036,
      "eval_steps_per_second": 0.509,
      "step": 63
    },
    {
      "epoch": 2.048,
      "grad_norm": 0.7881703384753784,
      "learning_rate": 2e-05,
      "loss": 0.7912,
      "step": 64
    },
    {
      "epoch": 2.048,
      "eval_loss": 0.7081906199455261,
      "eval_runtime": 98.9403,
      "eval_samples_per_second": 2.021,
      "eval_steps_per_second": 0.505,
      "step": 64
    },
    {
      "epoch": 2.08,
      "grad_norm": 0.8436680506716614,
      "learning_rate": 2e-05,
      "loss": 0.6965,
      "step": 65
    },
    {
      "epoch": 2.08,
      "eval_loss": 0.7070262432098389,
      "eval_runtime": 97.7451,
      "eval_samples_per_second": 2.046,
      "eval_steps_per_second": 0.512,
      "step": 65
    },
    {
      "epoch": 2.112,
      "grad_norm": 0.8694446846234115,
      "learning_rate": 2e-05,
      "loss": 0.7015,
      "step": 66
    },
    {
      "epoch": 2.112,
      "eval_loss": 0.7055197358131409,
      "eval_runtime": 97.994,
      "eval_samples_per_second": 2.041,
      "eval_steps_per_second": 0.51,
      "step": 66
    },
    {
      "epoch": 2.144,
      "grad_norm": 0.9518475022730614,
      "learning_rate": 2e-05,
      "loss": 0.6952,
      "step": 67
    },
    {
      "epoch": 2.144,
      "eval_loss": 0.7035844326019287,
      "eval_runtime": 98.0296,
      "eval_samples_per_second": 2.04,
      "eval_steps_per_second": 0.51,
      "step": 67
    },
    {
      "epoch": 2.176,
      "grad_norm": 0.8662066024104106,
      "learning_rate": 2e-05,
      "loss": 0.7731,
      "step": 68
    },
    {
      "epoch": 2.176,
      "eval_loss": 0.7019283771514893,
      "eval_runtime": 100.0982,
      "eval_samples_per_second": 1.998,
      "eval_steps_per_second": 0.5,
      "step": 68
    },
    {
      "epoch": 2.208,
      "grad_norm": 0.9574631942368209,
      "learning_rate": 2e-05,
      "loss": 0.7572,
      "step": 69
    },
    {
      "epoch": 2.208,
      "eval_loss": 0.7011401653289795,
      "eval_runtime": 101.9267,
      "eval_samples_per_second": 1.962,
      "eval_steps_per_second": 0.491,
      "step": 69
    },
    {
      "epoch": 2.24,
      "grad_norm": 0.9477542521472675,
      "learning_rate": 2e-05,
      "loss": 0.7393,
      "step": 70
    },
    {
      "epoch": 2.24,
      "eval_loss": 0.7006180286407471,
      "eval_runtime": 101.9672,
      "eval_samples_per_second": 1.961,
      "eval_steps_per_second": 0.49,
      "step": 70
    },
    {
      "epoch": 2.2720000000000002,
      "grad_norm": 0.9612956116350272,
      "learning_rate": 2e-05,
      "loss": 0.7175,
      "step": 71
    },
    {
      "epoch": 2.2720000000000002,
      "eval_loss": 0.7002539038658142,
      "eval_runtime": 101.8755,
      "eval_samples_per_second": 1.963,
      "eval_steps_per_second": 0.491,
      "step": 71
    },
    {
      "epoch": 2.304,
      "grad_norm": 0.9934844303955727,
      "learning_rate": 2e-05,
      "loss": 0.791,
      "step": 72
    },
    {
      "epoch": 2.304,
      "eval_loss": 0.6997203230857849,
      "eval_runtime": 97.9926,
      "eval_samples_per_second": 2.041,
      "eval_steps_per_second": 0.51,
      "step": 72
    },
    {
      "epoch": 2.336,
      "grad_norm": 1.0540430431227044,
      "learning_rate": 2e-05,
      "loss": 0.7542,
      "step": 73
    },
    {
      "epoch": 2.336,
      "eval_loss": 0.6988361477851868,
      "eval_runtime": 100.3704,
      "eval_samples_per_second": 1.993,
      "eval_steps_per_second": 0.498,
      "step": 73
    },
    {
      "epoch": 2.368,
      "grad_norm": 1.0249397957961794,
      "learning_rate": 2e-05,
      "loss": 0.8084,
      "step": 74
    },
    {
      "epoch": 2.368,
      "eval_loss": 0.6980065703392029,
      "eval_runtime": 101.8958,
      "eval_samples_per_second": 1.963,
      "eval_steps_per_second": 0.491,
      "step": 74
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.0445498365690145,
      "learning_rate": 2e-05,
      "loss": 0.7964,
      "step": 75
    },
    {
      "epoch": 2.4,
      "eval_loss": 0.6971798539161682,
      "eval_runtime": 98.1624,
      "eval_samples_per_second": 2.037,
      "eval_steps_per_second": 0.509,
      "step": 75
    },
    {
      "epoch": 2.432,
      "grad_norm": 0.9685893079320761,
      "learning_rate": 2e-05,
      "loss": 0.8047,
      "step": 76
    },
    {
      "epoch": 2.432,
      "eval_loss": 0.696861743927002,
      "eval_runtime": 98.6404,
      "eval_samples_per_second": 2.028,
      "eval_steps_per_second": 0.507,
      "step": 76
    },
    {
      "epoch": 2.464,
      "grad_norm": 0.9753091933204456,
      "learning_rate": 2e-05,
      "loss": 0.684,
      "step": 77
    },
    {
      "epoch": 2.464,
      "eval_loss": 0.6957904100418091,
      "eval_runtime": 98.8766,
      "eval_samples_per_second": 2.023,
      "eval_steps_per_second": 0.506,
      "step": 77
    },
    {
      "epoch": 2.496,
      "grad_norm": 0.9389149478503764,
      "learning_rate": 2e-05,
      "loss": 0.7434,
      "step": 78
    },
    {
      "epoch": 2.496,
      "eval_loss": 0.6943306922912598,
      "eval_runtime": 98.5751,
      "eval_samples_per_second": 2.029,
      "eval_steps_per_second": 0.507,
      "step": 78
    },
    {
      "epoch": 2.528,
      "grad_norm": 1.0675154845211299,
      "learning_rate": 2e-05,
      "loss": 0.7208,
      "step": 79
    },
    {
      "epoch": 2.528,
      "eval_loss": 0.6920651197433472,
      "eval_runtime": 99.0851,
      "eval_samples_per_second": 2.018,
      "eval_steps_per_second": 0.505,
      "step": 79
    },
    {
      "epoch": 2.56,
      "grad_norm": 0.9937936593307737,
      "learning_rate": 2e-05,
      "loss": 0.6948,
      "step": 80
    },
    {
      "epoch": 2.56,
      "eval_loss": 0.6899142265319824,
      "eval_runtime": 98.7974,
      "eval_samples_per_second": 2.024,
      "eval_steps_per_second": 0.506,
      "step": 80
    },
    {
      "epoch": 2.592,
      "grad_norm": 0.9650832276698476,
      "learning_rate": 2e-05,
      "loss": 0.7666,
      "step": 81
    },
    {
      "epoch": 2.592,
      "eval_loss": 0.6886695623397827,
      "eval_runtime": 98.6813,
      "eval_samples_per_second": 2.027,
      "eval_steps_per_second": 0.507,
      "step": 81
    },
    {
      "epoch": 2.624,
      "grad_norm": 0.9961610958296112,
      "learning_rate": 2e-05,
      "loss": 0.7741,
      "step": 82
    },
    {
      "epoch": 2.624,
      "eval_loss": 0.687745213508606,
      "eval_runtime": 98.624,
      "eval_samples_per_second": 2.028,
      "eval_steps_per_second": 0.507,
      "step": 82
    },
    {
      "epoch": 2.656,
      "grad_norm": 1.0261499699526089,
      "learning_rate": 2e-05,
      "loss": 0.7747,
      "step": 83
    },
    {
      "epoch": 2.656,
      "eval_loss": 0.6869972944259644,
      "eval_runtime": 98.9604,
      "eval_samples_per_second": 2.021,
      "eval_steps_per_second": 0.505,
      "step": 83
    },
    {
      "epoch": 2.6879999999999997,
      "grad_norm": 1.025059237745532,
      "learning_rate": 2e-05,
      "loss": 0.7526,
      "step": 84
    },
    {
      "epoch": 2.6879999999999997,
      "eval_loss": 0.6862147450447083,
      "eval_runtime": 98.4387,
      "eval_samples_per_second": 2.032,
      "eval_steps_per_second": 0.508,
      "step": 84
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 1.1383626266013125,
      "learning_rate": 2e-05,
      "loss": 0.683,
      "step": 85
    },
    {
      "epoch": 2.7199999999999998,
      "eval_loss": 0.6845572590827942,
      "eval_runtime": 98.4979,
      "eval_samples_per_second": 2.03,
      "eval_steps_per_second": 0.508,
      "step": 85
    },
    {
      "epoch": 2.752,
      "grad_norm": 1.0427739679421295,
      "learning_rate": 2e-05,
      "loss": 0.7362,
      "step": 86
    },
    {
      "epoch": 2.752,
      "eval_loss": 0.683280348777771,
      "eval_runtime": 103.6707,
      "eval_samples_per_second": 1.929,
      "eval_steps_per_second": 0.482,
      "step": 86
    },
    {
      "epoch": 2.784,
      "grad_norm": 1.1280129141879938,
      "learning_rate": 2e-05,
      "loss": 0.7743,
      "step": 87
    },
    {
      "epoch": 2.784,
      "eval_loss": 0.6824235320091248,
      "eval_runtime": 99.2995,
      "eval_samples_per_second": 2.014,
      "eval_steps_per_second": 0.504,
      "step": 87
    },
    {
      "epoch": 2.816,
      "grad_norm": 1.0017715497784696,
      "learning_rate": 2e-05,
      "loss": 0.7164,
      "step": 88
    },
    {
      "epoch": 2.816,
      "eval_loss": 0.6819013357162476,
      "eval_runtime": 99.5206,
      "eval_samples_per_second": 2.01,
      "eval_steps_per_second": 0.502,
      "step": 88
    },
    {
      "epoch": 2.848,
      "grad_norm": 1.0769673642284994,
      "learning_rate": 2e-05,
      "loss": 0.7013,
      "step": 89
    },
    {
      "epoch": 2.848,
      "eval_loss": 0.6820746064186096,
      "eval_runtime": 96.5197,
      "eval_samples_per_second": 2.072,
      "eval_steps_per_second": 0.518,
      "step": 89
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.0389167281844591,
      "learning_rate": 2e-05,
      "loss": 0.6805,
      "step": 90
    },
    {
      "epoch": 2.88,
      "eval_loss": 0.6826525926589966,
      "eval_runtime": 96.0942,
      "eval_samples_per_second": 2.081,
      "eval_steps_per_second": 0.52,
      "step": 90
    },
    {
      "epoch": 2.912,
      "grad_norm": 0.9705743838620626,
      "learning_rate": 2e-05,
      "loss": 0.7681,
      "step": 91
    },
    {
      "epoch": 2.912,
      "eval_loss": 0.6836435198783875,
      "eval_runtime": 96.2441,
      "eval_samples_per_second": 2.078,
      "eval_steps_per_second": 0.52,
      "step": 91
    },
    {
      "epoch": 2.944,
      "grad_norm": 1.037823791993831,
      "learning_rate": 2e-05,
      "loss": 0.7532,
      "step": 92
    },
    {
      "epoch": 2.944,
      "eval_loss": 0.6845746040344238,
      "eval_runtime": 96.2394,
      "eval_samples_per_second": 2.078,
      "eval_steps_per_second": 0.52,
      "step": 92
    },
    {
      "epoch": 2.976,
      "grad_norm": 1.1323835942146157,
      "learning_rate": 2e-05,
      "loss": 0.7171,
      "step": 93
    },
    {
      "epoch": 2.976,
      "eval_loss": 0.684663712978363,
      "eval_runtime": 96.23,
      "eval_samples_per_second": 2.078,
      "eval_steps_per_second": 0.52,
      "step": 93
    },
    {
      "epoch": 3.008,
      "grad_norm": 1.1957864756602699,
      "learning_rate": 2e-05,
      "loss": 0.7166,
      "step": 94
    },
    {
      "epoch": 3.008,
      "eval_loss": 0.6830846667289734,
      "eval_runtime": 96.6549,
      "eval_samples_per_second": 2.069,
      "eval_steps_per_second": 0.517,
      "step": 94
    },
    {
      "epoch": 3.04,
      "grad_norm": 1.1077357794232636,
      "learning_rate": 2e-05,
      "loss": 0.6667,
      "step": 95
    },
    {
      "epoch": 3.04,
      "eval_loss": 0.6810076832771301,
      "eval_runtime": 96.3239,
      "eval_samples_per_second": 2.076,
      "eval_steps_per_second": 0.519,
      "step": 95
    },
    {
      "epoch": 3.072,
      "grad_norm": 1.1851219157184936,
      "learning_rate": 2e-05,
      "loss": 0.7115,
      "step": 96
    },
    {
      "epoch": 3.072,
      "eval_loss": 0.6796395778656006,
      "eval_runtime": 96.9109,
      "eval_samples_per_second": 2.064,
      "eval_steps_per_second": 0.516,
      "step": 96
    },
    {
      "epoch": 3.104,
      "grad_norm": 1.0812671042616444,
      "learning_rate": 2e-05,
      "loss": 0.7333,
      "step": 97
    },
    {
      "epoch": 3.104,
      "eval_loss": 0.6794567108154297,
      "eval_runtime": 96.7403,
      "eval_samples_per_second": 2.067,
      "eval_steps_per_second": 0.517,
      "step": 97
    },
    {
      "epoch": 3.136,
      "grad_norm": 1.130095597839828,
      "learning_rate": 2e-05,
      "loss": 0.6328,
      "step": 98
    },
    {
      "epoch": 3.136,
      "eval_loss": 0.6792007684707642,
      "eval_runtime": 96.8136,
      "eval_samples_per_second": 2.066,
      "eval_steps_per_second": 0.516,
      "step": 98
    },
    {
      "epoch": 3.168,
      "grad_norm": 1.16102100344116,
      "learning_rate": 2e-05,
      "loss": 0.6625,
      "step": 99
    },
    {
      "epoch": 3.168,
      "eval_loss": 0.6789132952690125,
      "eval_runtime": 96.6982,
      "eval_samples_per_second": 2.068,
      "eval_steps_per_second": 0.517,
      "step": 99
    },
    {
      "epoch": 3.2,
      "grad_norm": 1.226689811951201,
      "learning_rate": 2e-05,
      "loss": 0.7522,
      "step": 100
    },
    {
      "epoch": 3.2,
      "eval_loss": 0.6786602735519409,
      "eval_runtime": 96.6253,
      "eval_samples_per_second": 2.07,
      "eval_steps_per_second": 0.517,
      "step": 100
    },
    {
      "epoch": 3.232,
      "grad_norm": 1.1623462595850367,
      "learning_rate": 2e-05,
      "loss": 0.6755,
      "step": 101
    },
    {
      "epoch": 3.232,
      "eval_loss": 0.6791322827339172,
      "eval_runtime": 96.5222,
      "eval_samples_per_second": 2.072,
      "eval_steps_per_second": 0.518,
      "step": 101
    },
    {
      "epoch": 3.2640000000000002,
      "grad_norm": 1.16303930181089,
      "learning_rate": 2e-05,
      "loss": 0.6752,
      "step": 102
    },
    {
      "epoch": 3.2640000000000002,
      "eval_loss": 0.680637538433075,
      "eval_runtime": 96.8789,
      "eval_samples_per_second": 2.064,
      "eval_steps_per_second": 0.516,
      "step": 102
    },
    {
      "epoch": 3.296,
      "grad_norm": 1.166028046661615,
      "learning_rate": 2e-05,
      "loss": 0.6732,
      "step": 103
    },
    {
      "epoch": 3.296,
      "eval_loss": 0.6818951964378357,
      "eval_runtime": 96.7579,
      "eval_samples_per_second": 2.067,
      "eval_steps_per_second": 0.517,
      "step": 103
    },
    {
      "epoch": 3.328,
      "grad_norm": 1.2872370423601793,
      "learning_rate": 2e-05,
      "loss": 0.7713,
      "step": 104
    },
    {
      "epoch": 3.328,
      "eval_loss": 0.682328462600708,
      "eval_runtime": 98.3142,
      "eval_samples_per_second": 2.034,
      "eval_steps_per_second": 0.509,
      "step": 104
    },
    {
      "epoch": 3.36,
      "grad_norm": 1.1363822202896854,
      "learning_rate": 2e-05,
      "loss": 0.7429,
      "step": 105
    },
    {
      "epoch": 3.36,
      "eval_loss": 0.6817943453788757,
      "eval_runtime": 100.7507,
      "eval_samples_per_second": 1.985,
      "eval_steps_per_second": 0.496,
      "step": 105
    },
    {
      "epoch": 3.416,
      "grad_norm": 1.2711044658075554,
      "learning_rate": 2e-05,
      "loss": 0.7057,
      "step": 106
    },
    {
      "epoch": 3.416,
      "eval_loss": 0.6794378161430359,
      "eval_runtime": 100.4096,
      "eval_samples_per_second": 1.992,
      "eval_steps_per_second": 0.498,
      "step": 106
    },
    {
      "epoch": 3.448,
      "grad_norm": 1.3212295597772596,
      "learning_rate": 2e-05,
      "loss": 0.6982,
      "step": 107
    },
    {
      "epoch": 3.448,
      "eval_loss": 0.6767404079437256,
      "eval_runtime": 96.6386,
      "eval_samples_per_second": 2.07,
      "eval_steps_per_second": 0.517,
      "step": 107
    },
    {
      "epoch": 3.48,
      "grad_norm": 1.2420948881737728,
      "learning_rate": 2e-05,
      "loss": 0.7092,
      "step": 108
    },
    {
      "epoch": 3.48,
      "eval_loss": 0.6751566529273987,
      "eval_runtime": 95.1851,
      "eval_samples_per_second": 2.101,
      "eval_steps_per_second": 0.525,
      "step": 108
    },
    {
      "epoch": 3.512,
      "grad_norm": 1.2965352636029341,
      "learning_rate": 2e-05,
      "loss": 0.6715,
      "step": 109
    },
    {
      "epoch": 3.512,
      "eval_loss": 0.6750080585479736,
      "eval_runtime": 95.3479,
      "eval_samples_per_second": 2.098,
      "eval_steps_per_second": 0.524,
      "step": 109
    },
    {
      "epoch": 3.544,
      "grad_norm": 1.2789534479099607,
      "learning_rate": 2e-05,
      "loss": 0.6732,
      "step": 110
    },
    {
      "epoch": 3.544,
      "eval_loss": 0.6744334697723389,
      "eval_runtime": 95.3541,
      "eval_samples_per_second": 2.097,
      "eval_steps_per_second": 0.524,
      "step": 110
    },
    {
      "epoch": 3.576,
      "grad_norm": 1.3881035995379567,
      "learning_rate": 2e-05,
      "loss": 0.6777,
      "step": 111
    },
    {
      "epoch": 3.576,
      "eval_loss": 0.6730498671531677,
      "eval_runtime": 96.972,
      "eval_samples_per_second": 2.062,
      "eval_steps_per_second": 0.516,
      "step": 111
    },
    {
      "epoch": 3.608,
      "grad_norm": 1.2398078245019133,
      "learning_rate": 2e-05,
      "loss": 0.6314,
      "step": 112
    },
    {
      "epoch": 3.608,
      "eval_loss": 0.6725335717201233,
      "eval_runtime": 97.1794,
      "eval_samples_per_second": 2.058,
      "eval_steps_per_second": 0.515,
      "step": 112
    },
    {
      "epoch": 3.64,
      "grad_norm": 1.3383993031041075,
      "learning_rate": 2e-05,
      "loss": 0.6776,
      "step": 113
    },
    {
      "epoch": 3.64,
      "eval_loss": 0.6719880104064941,
      "eval_runtime": 97.2319,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 113
    },
    {
      "epoch": 3.672,
      "grad_norm": 1.289557205987285,
      "learning_rate": 2e-05,
      "loss": 0.7475,
      "step": 114
    },
    {
      "epoch": 3.672,
      "eval_loss": 0.6723533868789673,
      "eval_runtime": 97.3108,
      "eval_samples_per_second": 2.055,
      "eval_steps_per_second": 0.514,
      "step": 114
    },
    {
      "epoch": 3.7039999999999997,
      "grad_norm": 1.121644844950096,
      "learning_rate": 2e-05,
      "loss": 0.6813,
      "step": 115
    },
    {
      "epoch": 3.7039999999999997,
      "eval_loss": 0.673932671546936,
      "eval_runtime": 97.0919,
      "eval_samples_per_second": 2.06,
      "eval_steps_per_second": 0.515,
      "step": 115
    },
    {
      "epoch": 3.7359999999999998,
      "grad_norm": 1.5370551636574723,
      "learning_rate": 2e-05,
      "loss": 0.7036,
      "step": 116
    },
    {
      "epoch": 3.7359999999999998,
      "eval_loss": 0.6737978458404541,
      "eval_runtime": 97.0064,
      "eval_samples_per_second": 2.062,
      "eval_steps_per_second": 0.515,
      "step": 116
    },
    {
      "epoch": 3.768,
      "grad_norm": 1.3017720066449985,
      "learning_rate": 2e-05,
      "loss": 0.7048,
      "step": 117
    },
    {
      "epoch": 3.768,
      "eval_loss": 0.6731936931610107,
      "eval_runtime": 96.9134,
      "eval_samples_per_second": 2.064,
      "eval_steps_per_second": 0.516,
      "step": 117
    },
    {
      "epoch": 3.8,
      "grad_norm": 1.3976157986974596,
      "learning_rate": 2e-05,
      "loss": 0.6512,
      "step": 118
    },
    {
      "epoch": 3.8,
      "eval_loss": 0.6716210842132568,
      "eval_runtime": 97.4028,
      "eval_samples_per_second": 2.053,
      "eval_steps_per_second": 0.513,
      "step": 118
    },
    {
      "epoch": 3.832,
      "grad_norm": 1.3982632064891347,
      "learning_rate": 2e-05,
      "loss": 0.6481,
      "step": 119
    },
    {
      "epoch": 3.832,
      "eval_loss": 0.6690347194671631,
      "eval_runtime": 97.2096,
      "eval_samples_per_second": 2.057,
      "eval_steps_per_second": 0.514,
      "step": 119
    },
    {
      "epoch": 3.864,
      "grad_norm": 1.346077433712507,
      "learning_rate": 2e-05,
      "loss": 0.6005,
      "step": 120
    },
    {
      "epoch": 3.864,
      "eval_loss": 0.667186975479126,
      "eval_runtime": 97.326,
      "eval_samples_per_second": 2.055,
      "eval_steps_per_second": 0.514,
      "step": 120
    },
    {
      "epoch": 3.896,
      "grad_norm": 1.3142652448176726,
      "learning_rate": 2e-05,
      "loss": 0.6967,
      "step": 121
    },
    {
      "epoch": 3.896,
      "eval_loss": 0.6662881970405579,
      "eval_runtime": 101.7648,
      "eval_samples_per_second": 1.965,
      "eval_steps_per_second": 0.491,
      "step": 121
    },
    {
      "epoch": 3.928,
      "grad_norm": 1.3368669812290557,
      "learning_rate": 2e-05,
      "loss": 0.6954,
      "step": 122
    },
    {
      "epoch": 3.928,
      "eval_loss": 0.6659301519393921,
      "eval_runtime": 98.3123,
      "eval_samples_per_second": 2.034,
      "eval_steps_per_second": 0.509,
      "step": 122
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.3992228513792793,
      "learning_rate": 2e-05,
      "loss": 0.6168,
      "step": 123
    },
    {
      "epoch": 3.96,
      "eval_loss": 0.6662861704826355,
      "eval_runtime": 98.5527,
      "eval_samples_per_second": 2.029,
      "eval_steps_per_second": 0.507,
      "step": 123
    },
    {
      "epoch": 3.992,
      "grad_norm": 1.403303492690172,
      "learning_rate": 2e-05,
      "loss": 0.6818,
      "step": 124
    },
    {
      "epoch": 3.992,
      "eval_loss": 0.666582465171814,
      "eval_runtime": 95.7576,
      "eval_samples_per_second": 2.089,
      "eval_steps_per_second": 0.522,
      "step": 124
    },
    {
      "epoch": 4.024,
      "grad_norm": 1.37383602466541,
      "learning_rate": 2e-05,
      "loss": 0.6539,
      "step": 125
    },
    {
      "epoch": 4.024,
      "eval_loss": 0.6666322350502014,
      "eval_runtime": 95.5774,
      "eval_samples_per_second": 2.093,
      "eval_steps_per_second": 0.523,
      "step": 125
    },
    {
      "epoch": 4.064,
      "grad_norm": 1.3710348437226973,
      "learning_rate": 2e-05,
      "loss": 0.7181,
      "step": 126
    },
    {
      "epoch": 4.064,
      "eval_loss": 0.6673327684402466,
      "eval_runtime": 102.0166,
      "eval_samples_per_second": 1.96,
      "eval_steps_per_second": 0.49,
      "step": 126
    },
    {
      "epoch": 4.096,
      "grad_norm": 1.297767286241014,
      "learning_rate": 2e-05,
      "loss": 0.6452,
      "step": 127
    },
    {
      "epoch": 4.096,
      "eval_loss": 0.6687185764312744,
      "eval_runtime": 95.6988,
      "eval_samples_per_second": 2.09,
      "eval_steps_per_second": 0.522,
      "step": 127
    },
    {
      "epoch": 4.128,
      "grad_norm": 1.292569885504323,
      "learning_rate": 2e-05,
      "loss": 0.6145,
      "step": 128
    },
    {
      "epoch": 4.128,
      "eval_loss": 0.6715303063392639,
      "eval_runtime": 95.465,
      "eval_samples_per_second": 2.095,
      "eval_steps_per_second": 0.524,
      "step": 128
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.419162300463038,
      "learning_rate": 2e-05,
      "loss": 0.6088,
      "step": 129
    },
    {
      "epoch": 4.16,
      "eval_loss": 0.675749659538269,
      "eval_runtime": 97.8829,
      "eval_samples_per_second": 2.043,
      "eval_steps_per_second": 0.511,
      "step": 129
    },
    {
      "epoch": 4.192,
      "grad_norm": 1.5215794897159234,
      "learning_rate": 2e-05,
      "loss": 0.6391,
      "step": 130
    },
    {
      "epoch": 4.192,
      "eval_loss": 0.677629828453064,
      "eval_runtime": 95.4624,
      "eval_samples_per_second": 2.095,
      "eval_steps_per_second": 0.524,
      "step": 130
    },
    {
      "epoch": 4.224,
      "grad_norm": 1.4684908212096284,
      "learning_rate": 2e-05,
      "loss": 0.6102,
      "step": 131
    },
    {
      "epoch": 4.224,
      "eval_loss": 0.6779782176017761,
      "eval_runtime": 97.1518,
      "eval_samples_per_second": 2.059,
      "eval_steps_per_second": 0.515,
      "step": 131
    },
    {
      "epoch": 4.256,
      "grad_norm": 1.7036283256760842,
      "learning_rate": 2e-05,
      "loss": 0.6396,
      "step": 132
    },
    {
      "epoch": 4.256,
      "eval_loss": 0.6759344339370728,
      "eval_runtime": 97.8632,
      "eval_samples_per_second": 2.044,
      "eval_steps_per_second": 0.511,
      "step": 132
    },
    {
      "epoch": 4.288,
      "grad_norm": 1.6051118551476093,
      "learning_rate": 2e-05,
      "loss": 0.6566,
      "step": 133
    },
    {
      "epoch": 4.288,
      "eval_loss": 0.673258364200592,
      "eval_runtime": 97.4448,
      "eval_samples_per_second": 2.052,
      "eval_steps_per_second": 0.513,
      "step": 133
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.5572448859529642,
      "learning_rate": 2e-05,
      "loss": 0.6223,
      "step": 134
    },
    {
      "epoch": 4.32,
      "eval_loss": 0.6723945736885071,
      "eval_runtime": 96.9931,
      "eval_samples_per_second": 2.062,
      "eval_steps_per_second": 0.516,
      "step": 134
    },
    {
      "epoch": 4.352,
      "grad_norm": 1.607226224281089,
      "learning_rate": 2e-05,
      "loss": 0.6168,
      "step": 135
    },
    {
      "epoch": 4.352,
      "eval_loss": 0.6719383001327515,
      "eval_runtime": 97.0819,
      "eval_samples_per_second": 2.06,
      "eval_steps_per_second": 0.515,
      "step": 135
    },
    {
      "epoch": 4.384,
      "grad_norm": 1.3581330775842726,
      "learning_rate": 2e-05,
      "loss": 0.6197,
      "step": 136
    },
    {
      "epoch": 4.384,
      "eval_loss": 0.6716668605804443,
      "eval_runtime": 99.6743,
      "eval_samples_per_second": 2.007,
      "eval_steps_per_second": 0.502,
      "step": 136
    },
    {
      "epoch": 4.416,
      "grad_norm": 1.6769316310469216,
      "learning_rate": 2e-05,
      "loss": 0.638,
      "step": 137
    },
    {
      "epoch": 4.416,
      "eval_loss": 0.6708360314369202,
      "eval_runtime": 95.7987,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.522,
      "step": 137
    },
    {
      "epoch": 4.448,
      "grad_norm": 1.5498003031158447,
      "learning_rate": 2e-05,
      "loss": 0.6481,
      "step": 138
    },
    {
      "epoch": 4.448,
      "eval_loss": 0.6702771782875061,
      "eval_runtime": 96.0198,
      "eval_samples_per_second": 2.083,
      "eval_steps_per_second": 0.521,
      "step": 138
    },
    {
      "epoch": 4.48,
      "grad_norm": 1.551391635180101,
      "learning_rate": 2e-05,
      "loss": 0.6396,
      "step": 139
    },
    {
      "epoch": 4.48,
      "eval_loss": 0.6709393262863159,
      "eval_runtime": 95.7365,
      "eval_samples_per_second": 2.089,
      "eval_steps_per_second": 0.522,
      "step": 139
    },
    {
      "epoch": 4.5120000000000005,
      "grad_norm": 1.534460089474332,
      "learning_rate": 2e-05,
      "loss": 0.5738,
      "step": 140
    },
    {
      "epoch": 4.5120000000000005,
      "eval_loss": 0.6725634932518005,
      "eval_runtime": 98.1018,
      "eval_samples_per_second": 2.039,
      "eval_steps_per_second": 0.51,
      "step": 140
    },
    {
      "epoch": 4.5440000000000005,
      "grad_norm": 1.4820795078494542,
      "learning_rate": 2e-05,
      "loss": 0.662,
      "step": 141
    },
    {
      "epoch": 4.5440000000000005,
      "eval_loss": 0.6750396490097046,
      "eval_runtime": 95.8735,
      "eval_samples_per_second": 2.086,
      "eval_steps_per_second": 0.522,
      "step": 141
    },
    {
      "epoch": 4.576,
      "grad_norm": 1.5013493191075218,
      "learning_rate": 2e-05,
      "loss": 0.6571,
      "step": 142
    },
    {
      "epoch": 4.576,
      "eval_loss": 0.6773877739906311,
      "eval_runtime": 95.7522,
      "eval_samples_per_second": 2.089,
      "eval_steps_per_second": 0.522,
      "step": 142
    },
    {
      "epoch": 4.608,
      "grad_norm": 1.5713746383626521,
      "learning_rate": 2e-05,
      "loss": 0.6652,
      "step": 143
    },
    {
      "epoch": 4.608,
      "eval_loss": 0.677890419960022,
      "eval_runtime": 95.6097,
      "eval_samples_per_second": 2.092,
      "eval_steps_per_second": 0.523,
      "step": 143
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.7992319985289944,
      "learning_rate": 2e-05,
      "loss": 0.6846,
      "step": 144
    },
    {
      "epoch": 4.64,
      "eval_loss": 0.6761859059333801,
      "eval_runtime": 95.5647,
      "eval_samples_per_second": 2.093,
      "eval_steps_per_second": 0.523,
      "step": 144
    },
    {
      "epoch": 4.672,
      "grad_norm": 1.8153564044346968,
      "learning_rate": 2e-05,
      "loss": 0.6343,
      "step": 145
    },
    {
      "epoch": 4.672,
      "eval_loss": 0.6728916764259338,
      "eval_runtime": 95.7986,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.522,
      "step": 145
    },
    {
      "epoch": 4.704,
      "grad_norm": 1.7257180198459412,
      "learning_rate": 2e-05,
      "loss": 0.6635,
      "step": 146
    },
    {
      "epoch": 4.704,
      "eval_loss": 0.670305073261261,
      "eval_runtime": 96.0378,
      "eval_samples_per_second": 2.083,
      "eval_steps_per_second": 0.521,
      "step": 146
    },
    {
      "epoch": 4.736,
      "grad_norm": 1.700890150951268,
      "learning_rate": 2e-05,
      "loss": 0.6492,
      "step": 147
    },
    {
      "epoch": 4.736,
      "eval_loss": 0.6693490743637085,
      "eval_runtime": 95.7653,
      "eval_samples_per_second": 2.088,
      "eval_steps_per_second": 0.522,
      "step": 147
    },
    {
      "epoch": 4.768,
      "grad_norm": 1.524592426429662,
      "learning_rate": 2e-05,
      "loss": 0.6196,
      "step": 148
    },
    {
      "epoch": 4.768,
      "eval_loss": 0.6692916750907898,
      "eval_runtime": 95.8987,
      "eval_samples_per_second": 2.086,
      "eval_steps_per_second": 0.521,
      "step": 148
    },
    {
      "epoch": 4.8,
      "grad_norm": 1.765585164658825,
      "learning_rate": 2e-05,
      "loss": 0.6634,
      "step": 149
    },
    {
      "epoch": 4.8,
      "eval_loss": 0.6684932112693787,
      "eval_runtime": 96.0068,
      "eval_samples_per_second": 2.083,
      "eval_steps_per_second": 0.521,
      "step": 149
    },
    {
      "epoch": 4.832,
      "grad_norm": 1.7560172174474262,
      "learning_rate": 2e-05,
      "loss": 0.6011,
      "step": 150
    },
    {
      "epoch": 4.832,
      "eval_loss": 0.6678715348243713,
      "eval_runtime": 95.8356,
      "eval_samples_per_second": 2.087,
      "eval_steps_per_second": 0.522,
      "step": 150
    },
    {
      "epoch": 4.864,
      "grad_norm": 1.5106799279904972,
      "learning_rate": 2e-05,
      "loss": 0.6502,
      "step": 151
    },
    {
      "epoch": 4.864,
      "eval_loss": 0.6691852807998657,
      "eval_runtime": 97.5603,
      "eval_samples_per_second": 2.05,
      "eval_steps_per_second": 0.513,
      "step": 151
    },
    {
      "epoch": 4.896,
      "grad_norm": 1.6515668480995418,
      "learning_rate": 2e-05,
      "loss": 0.6207,
      "step": 152
    },
    {
      "epoch": 4.896,
      "eval_loss": 0.6714971661567688,
      "eval_runtime": 96.9472,
      "eval_samples_per_second": 2.063,
      "eval_steps_per_second": 0.516,
      "step": 152
    },
    {
      "epoch": 4.928,
      "grad_norm": 1.648933444515062,
      "learning_rate": 2e-05,
      "loss": 0.6366,
      "step": 153
    },
    {
      "epoch": 4.928,
      "eval_loss": 0.6733483672142029,
      "eval_runtime": 100.8633,
      "eval_samples_per_second": 1.983,
      "eval_steps_per_second": 0.496,
      "step": 153
    },
    {
      "epoch": 4.96,
      "grad_norm": 1.6749689583250607,
      "learning_rate": 2e-05,
      "loss": 0.5815,
      "step": 154
    },
    {
      "epoch": 4.96,
      "eval_loss": 0.6746806502342224,
      "eval_runtime": 100.4948,
      "eval_samples_per_second": 1.99,
      "eval_steps_per_second": 0.498,
      "step": 154
    },
    {
      "epoch": 4.992,
      "grad_norm": 1.7662303190761401,
      "learning_rate": 2e-05,
      "loss": 0.6366,
      "step": 155
    },
    {
      "epoch": 4.992,
      "eval_loss": 0.6755189299583435,
      "eval_runtime": 100.872,
      "eval_samples_per_second": 1.983,
      "eval_steps_per_second": 0.496,
      "step": 155
    },
    {
      "epoch": 4.992,
      "step": 155,
      "total_flos": 135390371053568.0,
      "train_loss": 0.1233276440251258,
      "train_runtime": 3633.0506,
      "train_samples_per_second": 1.376,
      "train_steps_per_second": 0.043
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 155,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 5,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 135390371053568.0,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}