{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 39,
  "global_step": 195,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02564102564102564,
      "grad_norm": 0.22871212661266327,
      "learning_rate": 0.0,
      "loss": 1.7578125,
      "step": 1
    },
    {
      "epoch": 0.05128205128205128,
      "grad_norm": 0.18538884818553925,
      "learning_rate": 5.000000000000001e-07,
      "loss": 1.6953125,
      "step": 2
    },
    {
      "epoch": 0.07692307692307693,
      "grad_norm": 0.21811293065547943,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.818359375,
      "step": 3
    },
    {
      "epoch": 0.10256410256410256,
      "grad_norm": 0.21101585030555725,
      "learning_rate": 1.5e-06,
      "loss": 1.759765625,
      "step": 4
    },
    {
      "epoch": 0.1282051282051282,
      "grad_norm": 0.2201356142759323,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 1.796875,
      "step": 5
    },
    {
      "epoch": 0.15384615384615385,
      "grad_norm": 0.16987788677215576,
      "learning_rate": 2.5e-06,
      "loss": 1.716796875,
      "step": 6
    },
    {
      "epoch": 0.1794871794871795,
      "grad_norm": 0.20609858632087708,
      "learning_rate": 3e-06,
      "loss": 1.8017578125,
      "step": 7
    },
    {
      "epoch": 0.20512820512820512,
      "grad_norm": 0.18626707792282104,
      "learning_rate": 3.5e-06,
      "loss": 1.477294921875,
      "step": 8
    },
    {
      "epoch": 0.23076923076923078,
      "grad_norm": 0.242230623960495,
      "learning_rate": 4.000000000000001e-06,
      "loss": 1.7978515625,
      "step": 9
    },
    {
      "epoch": 0.2564102564102564,
      "grad_norm": 0.2741856276988983,
      "learning_rate": 4.5e-06,
      "loss": 1.8447265625,
      "step": 10
    },
    {
      "epoch": 0.28205128205128205,
      "grad_norm": 0.2998216450214386,
      "learning_rate": 5e-06,
      "loss": 1.87548828125,
      "step": 11
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 0.19347317516803741,
      "learning_rate": 5.500000000000001e-06,
      "loss": 1.71875,
      "step": 12
    },
    {
      "epoch": 0.3333333333333333,
      "grad_norm": 0.17854830622673035,
      "learning_rate": 6e-06,
      "loss": 1.7529296875,
      "step": 13
    },
    {
      "epoch": 0.358974358974359,
      "grad_norm": 0.1743069887161255,
      "learning_rate": 6.5000000000000004e-06,
      "loss": 1.7802734375,
      "step": 14
    },
    {
      "epoch": 0.38461538461538464,
      "grad_norm": 0.15998415648937225,
      "learning_rate": 7e-06,
      "loss": 1.69921875,
      "step": 15
    },
    {
      "epoch": 0.41025641025641024,
      "grad_norm": 0.1925395131111145,
      "learning_rate": 7.500000000000001e-06,
      "loss": 1.4715576171875,
      "step": 16
    },
    {
      "epoch": 0.4358974358974359,
      "grad_norm": 0.15806065499782562,
      "learning_rate": 8.000000000000001e-06,
      "loss": 1.7373046875,
      "step": 17
    },
    {
      "epoch": 0.46153846153846156,
      "grad_norm": 0.1433170586824417,
      "learning_rate": 8.5e-06,
      "loss": 1.7685546875,
      "step": 18
    },
    {
      "epoch": 0.48717948717948717,
      "grad_norm": 0.11917092651128769,
      "learning_rate": 9e-06,
      "loss": 1.5859375,
      "step": 19
    },
    {
      "epoch": 0.5128205128205128,
      "grad_norm": 0.11352289468050003,
      "learning_rate": 9.5e-06,
      "loss": 1.681640625,
      "step": 20
    },
    {
      "epoch": 0.5384615384615384,
      "grad_norm": 0.11791280657052994,
      "learning_rate": 1e-05,
      "loss": 1.673828125,
      "step": 21
    },
    {
      "epoch": 0.5641025641025641,
      "grad_norm": 0.7906798124313354,
      "learning_rate": 9.999274905680763e-06,
      "loss": 1.6484375,
      "step": 22
    },
    {
      "epoch": 0.5897435897435898,
      "grad_norm": 0.12193252891302109,
      "learning_rate": 9.997099856394947e-06,
      "loss": 1.6640625,
      "step": 23
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.11531448364257812,
      "learning_rate": 9.993475553082945e-06,
      "loss": 1.65234375,
      "step": 24
    },
    {
      "epoch": 0.6410256410256411,
      "grad_norm": 0.1119362935423851,
      "learning_rate": 9.988403163727754e-06,
      "loss": 1.724609375,
      "step": 25
    },
    {
      "epoch": 0.6666666666666666,
      "grad_norm": 0.10708431899547577,
      "learning_rate": 9.981884322978577e-06,
      "loss": 1.69921875,
      "step": 26
    },
    {
      "epoch": 0.6923076923076923,
      "grad_norm": 0.09046280384063721,
      "learning_rate": 9.973921131624033e-06,
      "loss": 1.6318359375,
      "step": 27
    },
    {
      "epoch": 0.717948717948718,
      "grad_norm": 0.0954422801733017,
      "learning_rate": 9.964516155915152e-06,
      "loss": 1.7197265625,
      "step": 28
    },
    {
      "epoch": 0.7435897435897436,
      "grad_norm": 0.08470303565263748,
      "learning_rate": 9.953672426738358e-06,
      "loss": 1.59765625,
      "step": 29
    },
    {
      "epoch": 0.7692307692307693,
      "grad_norm": 0.07736426591873169,
      "learning_rate": 9.941393438638731e-06,
      "loss": 1.591796875,
      "step": 30
    },
    {
      "epoch": 0.7948717948717948,
      "grad_norm": 0.07388478517532349,
      "learning_rate": 9.927683148693834e-06,
      "loss": 1.662109375,
      "step": 31
    },
    {
      "epoch": 0.8205128205128205,
      "grad_norm": 0.07508436590433121,
      "learning_rate": 9.91254597523849e-06,
      "loss": 1.6943359375,
      "step": 32
    },
    {
      "epoch": 0.8461538461538461,
      "grad_norm": 0.07030631601810455,
      "learning_rate": 9.895986796440914e-06,
      "loss": 1.6796875,
      "step": 33
    },
    {
      "epoch": 0.8717948717948718,
      "grad_norm": 0.06741949915885925,
      "learning_rate": 9.878010948730654e-06,
      "loss": 1.4033203125,
      "step": 34
    },
    {
      "epoch": 0.8974358974358975,
      "grad_norm": 0.07069171965122223,
      "learning_rate": 9.858624225078841e-06,
      "loss": 1.64599609375,
      "step": 35
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.06931110471487045,
      "learning_rate": 9.83783287313134e-06,
      "loss": 1.5986328125,
      "step": 36
    },
    {
      "epoch": 0.9487179487179487,
      "grad_norm": 0.07381527125835419,
      "learning_rate": 9.815643593195346e-06,
      "loss": 1.6845703125,
      "step": 37
    },
    {
      "epoch": 0.9743589743589743,
      "grad_norm": 0.07050381600856781,
      "learning_rate": 9.792063536080115e-06,
      "loss": 1.6376953125,
      "step": 38
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.0710814893245697,
      "learning_rate": 9.76710030079252e-06,
      "loss": 1.6044921875,
      "step": 39
    },
    {
      "epoch": 1.0,
      "eval_loss": 1.63037109375,
      "eval_runtime": 2.0359,
      "eval_samples_per_second": 1.474,
      "eval_steps_per_second": 0.491,
      "step": 39
    },
    {
      "epoch": 1.0256410256410255,
      "grad_norm": 0.06843537092208862,
      "learning_rate": 9.740761932088157e-06,
      "loss": 1.61328125,
      "step": 40
    },
    {
      "epoch": 1.0512820512820513,
      "grad_norm": 0.061435483396053314,
      "learning_rate": 9.713056917878818e-06,
      "loss": 1.556640625,
      "step": 41
    },
    {
      "epoch": 1.0769230769230769,
      "grad_norm": 0.06764902174472809,
      "learning_rate": 9.683994186497132e-06,
      "loss": 1.666015625,
      "step": 42
    },
    {
      "epoch": 1.1025641025641026,
      "grad_norm": 0.060427065938711166,
      "learning_rate": 9.653583103819289e-06,
      "loss": 1.6044921875,
      "step": 43
    },
    {
      "epoch": 1.1282051282051282,
      "grad_norm": 0.06229192763566971,
      "learning_rate": 9.62183347024675e-06,
      "loss": 1.63671875,
      "step": 44
    },
    {
      "epoch": 1.1538461538461537,
      "grad_norm": 0.0625888928771019,
      "learning_rate": 9.588755517547936e-06,
      "loss": 1.5673828125,
      "step": 45
    },
    {
      "epoch": 1.1794871794871795,
      "grad_norm": 0.056661974638700485,
      "learning_rate": 9.554359905560887e-06,
      "loss": 1.640625,
      "step": 46
    },
    {
      "epoch": 1.205128205128205,
      "grad_norm": 0.04868312180042267,
      "learning_rate": 9.518657718757985e-06,
      "loss": 1.33642578125,
      "step": 47
    },
    {
      "epoch": 1.2307692307692308,
      "grad_norm": 0.06196633726358414,
      "learning_rate": 9.481660462673824e-06,
      "loss": 1.6376953125,
      "step": 48
    },
    {
      "epoch": 1.2564102564102564,
      "grad_norm": 0.07045412808656693,
      "learning_rate": 9.443380060197387e-06,
      "loss": 1.669921875,
      "step": 49
    },
    {
      "epoch": 1.282051282051282,
      "grad_norm": 0.08120285719633102,
      "learning_rate": 9.40382884772973e-06,
      "loss": 1.693359375,
      "step": 50
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.04695580527186394,
      "learning_rate": 9.363019571208398e-06,
      "loss": 1.5693359375,
      "step": 51
    },
    {
      "epoch": 1.3333333333333333,
      "grad_norm": 0.04755212739109993,
      "learning_rate": 9.320965381999872e-06,
      "loss": 1.5966796875,
      "step": 52
    },
    {
      "epoch": 1.358974358974359,
      "grad_norm": 0.049953095614910126,
      "learning_rate": 9.277679832661348e-06,
      "loss": 1.6279296875,
      "step": 53
    },
    {
      "epoch": 1.3846153846153846,
      "grad_norm": 0.04847012087702751,
      "learning_rate": 9.233176872573241e-06,
      "loss": 1.55859375,
      "step": 54
    },
    {
      "epoch": 1.4102564102564101,
      "grad_norm": 0.052262645214796066,
      "learning_rate": 9.187470843443794e-06,
      "loss": 1.343017578125,
      "step": 55
    },
    {
      "epoch": 1.435897435897436,
      "grad_norm": 0.04185536503791809,
      "learning_rate": 9.140576474687263e-06,
      "loss": 1.607421875,
      "step": 56
    },
    {
      "epoch": 1.4615384615384617,
      "grad_norm": 0.04539763554930687,
      "learning_rate": 9.092508878677158e-06,
      "loss": 1.634765625,
      "step": 57
    },
    {
      "epoch": 1.4871794871794872,
      "grad_norm": 0.04300449788570404,
      "learning_rate": 9.043283545876055e-06,
      "loss": 1.4736328125,
      "step": 58
    },
    {
      "epoch": 1.5128205128205128,
      "grad_norm": 0.04862791672348976,
      "learning_rate": 8.99291633984359e-06,
      "loss": 1.56640625,
      "step": 59
    },
    {
      "epoch": 1.5384615384615383,
      "grad_norm": 0.04195116460323334,
      "learning_rate": 8.941423492124194e-06,
      "loss": 1.564453125,
      "step": 60
    },
    {
      "epoch": 1.564102564102564,
      "grad_norm": 0.048699039965867996,
      "learning_rate": 8.888821597016251e-06,
      "loss": 1.5224609375,
      "step": 61
    },
    {
      "epoch": 1.5897435897435899,
      "grad_norm": 0.046053119003772736,
      "learning_rate": 8.835127606224347e-06,
      "loss": 1.5634765625,
      "step": 62
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.043240975588560104,
      "learning_rate": 8.780358823396352e-06,
      "loss": 1.5419921875,
      "step": 63
    },
    {
      "epoch": 1.641025641025641,
      "grad_norm": 0.04077160358428955,
      "learning_rate": 8.724532898547065e-06,
      "loss": 1.6240234375,
      "step": 64
    },
    {
      "epoch": 1.6666666666666665,
      "grad_norm": 0.04109228029847145,
      "learning_rate": 8.667667822370253e-06,
      "loss": 1.599609375,
      "step": 65
    },
    {
      "epoch": 1.6923076923076923,
      "grad_norm": 0.041688188910484314,
      "learning_rate": 8.609781920440891e-06,
      "loss": 1.5517578125,
      "step": 66
    },
    {
      "epoch": 1.717948717948718,
      "grad_norm": 0.038294363766908646,
      "learning_rate": 8.550893847309496e-06,
      "loss": 1.6328125,
      "step": 67
    },
    {
      "epoch": 1.7435897435897436,
      "grad_norm": 0.0398561917245388,
      "learning_rate": 8.491022580490418e-06,
      "loss": 1.51953125,
      "step": 68
    },
    {
      "epoch": 1.7692307692307692,
      "grad_norm": 0.038095083087682724,
      "learning_rate": 8.430187414346082e-06,
      "loss": 1.517578125,
      "step": 69
    },
    {
      "epoch": 1.7948717948717947,
      "grad_norm": 0.03960049897432327,
      "learning_rate": 8.368407953869105e-06,
      "loss": 1.5830078125,
      "step": 70
    },
    {
      "epoch": 1.8205128205128205,
      "grad_norm": 0.03845981881022453,
      "learning_rate": 8.305704108364301e-06,
      "loss": 1.6142578125,
      "step": 71
    },
    {
      "epoch": 1.8461538461538463,
      "grad_norm": 0.03888195753097534,
      "learning_rate": 8.242096085032637e-06,
      "loss": 1.603515625,
      "step": 72
    },
    {
      "epoch": 1.8717948717948718,
      "grad_norm": 0.03578997030854225,
      "learning_rate": 8.177604382459168e-06,
      "loss": 1.3448486328125,
      "step": 73
    },
    {
      "epoch": 1.8974358974358974,
      "grad_norm": 0.038249146193265915,
      "learning_rate": 8.112249784007081e-06,
      "loss": 1.57080078125,
      "step": 74
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.04393129050731659,
      "learning_rate": 8.046053351119965e-06,
      "loss": 1.5263671875,
      "step": 75
    },
    {
      "epoch": 1.9487179487179487,
      "grad_norm": 0.03695112466812134,
      "learning_rate": 7.979036416534462e-06,
      "loss": 1.61328125,
      "step": 76
    },
    {
      "epoch": 1.9743589743589745,
      "grad_norm": 0.041780345141887665,
      "learning_rate": 7.911220577405485e-06,
      "loss": 1.5703125,
      "step": 77
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.035476040095090866,
      "learning_rate": 7.84262768834624e-06,
      "loss": 1.548828125,
      "step": 78
    },
    {
      "epoch": 2.0,
      "eval_loss": 1.56396484375,
      "eval_runtime": 2.0404,
      "eval_samples_per_second": 1.47,
      "eval_steps_per_second": 0.49,
      "step": 78
    },
    {
      "epoch": 2.0256410256410255,
      "grad_norm": 0.034847695380449295,
      "learning_rate": 7.773279854385256e-06,
      "loss": 1.5556640625,
      "step": 79
    },
    {
      "epoch": 2.051282051282051,
      "grad_norm": 0.03561405465006828,
      "learning_rate": 7.703199423842737e-06,
      "loss": 1.4990234375,
      "step": 80
    },
    {
      "epoch": 2.076923076923077,
      "grad_norm": 0.034786123782396317,
      "learning_rate": 7.632408981128493e-06,
      "loss": 1.603515625,
      "step": 81
    },
    {
      "epoch": 2.1025641025641026,
      "grad_norm": 0.033820539712905884,
      "learning_rate": 7.560931339463805e-06,
      "loss": 1.55078125,
      "step": 82
    },
    {
      "epoch": 2.128205128205128,
      "grad_norm": 0.033961810171604156,
      "learning_rate": 7.488789533529521e-06,
      "loss": 1.58203125,
      "step": 83
    },
    {
      "epoch": 2.1538461538461537,
      "grad_norm": 0.04089190438389778,
      "learning_rate": 7.416006812042827e-06,
      "loss": 1.5068359375,
      "step": 84
    },
    {
      "epoch": 2.1794871794871793,
      "grad_norm": 0.033873166888952255,
      "learning_rate": 7.342606630265009e-06,
      "loss": 1.5849609375,
      "step": 85
    },
    {
      "epoch": 2.2051282051282053,
      "grad_norm": 0.03078600950539112,
      "learning_rate": 7.268612642442657e-06,
      "loss": 1.2940673828125,
      "step": 86
    },
    {
      "epoch": 2.230769230769231,
      "grad_norm": 0.03978991508483887,
      "learning_rate": 7.194048694184764e-06,
      "loss": 1.5830078125,
      "step": 87
    },
    {
      "epoch": 2.2564102564102564,
      "grad_norm": 0.04575211927294731,
      "learning_rate": 7.118938814778132e-06,
      "loss": 1.62109375,
      "step": 88
    },
    {
      "epoch": 2.282051282051282,
      "grad_norm": 0.05950942635536194,
      "learning_rate": 7.043307209443591e-06,
      "loss": 1.63671875,
      "step": 89
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.03359539434313774,
      "learning_rate": 6.967178251535539e-06,
      "loss": 1.525390625,
      "step": 90
    },
    {
      "epoch": 2.3333333333333335,
      "grad_norm": 0.03355783224105835,
      "learning_rate": 6.890576474687264e-06,
      "loss": 1.5498046875,
      "step": 91
    },
    {
      "epoch": 2.358974358974359,
      "grad_norm": 0.035399310290813446,
      "learning_rate": 6.813526564904651e-06,
      "loss": 1.5751953125,
      "step": 92
    },
    {
      "epoch": 2.3846153846153846,
      "grad_norm": 0.037045758217573166,
      "learning_rate": 6.7360533526107564e-06,
      "loss": 1.5146484375,
      "step": 93
    },
    {
      "epoch": 2.41025641025641,
      "grad_norm": 0.040528059005737305,
      "learning_rate": 6.658181804643862e-06,
      "loss": 1.30712890625,
      "step": 94
    },
    {
      "epoch": 2.435897435897436,
      "grad_norm": 0.033201709389686584,
      "learning_rate": 6.579937016211547e-06,
      "loss": 1.5654296875,
      "step": 95
    },
    {
      "epoch": 2.4615384615384617,
      "grad_norm": 0.0346633717417717,
      "learning_rate": 6.501344202803415e-06,
      "loss": 1.5869140625,
      "step": 96
    },
    {
      "epoch": 2.4871794871794872,
      "grad_norm": 0.03557073324918747,
      "learning_rate": 6.422428692065038e-06,
      "loss": 1.4384765625,
      "step": 97
    },
    {
      "epoch": 2.5128205128205128,
      "grad_norm": 0.0386660173535347,
      "learning_rate": 6.343215915635762e-06,
      "loss": 1.525390625,
      "step": 98
    },
    {
      "epoch": 2.5384615384615383,
      "grad_norm": 0.032886434346437454,
      "learning_rate": 6.2637314009530106e-06,
      "loss": 1.5244140625,
      "step": 99
    },
    {
      "epoch": 2.564102564102564,
      "grad_norm": 0.03942364826798439,
      "learning_rate": 6.184000763025704e-06,
      "loss": 1.48388671875,
      "step": 100
    },
    {
      "epoch": 2.58974358974359,
      "grad_norm": 0.03862686827778816,
      "learning_rate": 6.10404969617945e-06,
      "loss": 1.52880859375,
      "step": 101
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 0.03463450446724892,
      "learning_rate": 6.023903965776203e-06,
      "loss": 1.5029296875,
      "step": 102
    },
    {
      "epoch": 2.641025641025641,
      "grad_norm": 0.034159280359745026,
      "learning_rate": 5.943589399910996e-06,
      "loss": 1.5859375,
      "step": 103
    },
    {
      "epoch": 2.6666666666666665,
      "grad_norm": 0.0343034490942955,
      "learning_rate": 5.863131881088478e-06,
      "loss": 1.5576171875,
      "step": 104
    },
    {
      "epoch": 2.6923076923076925,
      "grad_norm": 0.03469570353627205,
      "learning_rate": 5.782557337881911e-06,
      "loss": 1.517578125,
      "step": 105
    },
    {
      "epoch": 2.717948717948718,
      "grad_norm": 0.03211572766304016,
      "learning_rate": 5.701891736577318e-06,
      "loss": 1.599609375,
      "step": 106
    },
    {
      "epoch": 2.7435897435897436,
      "grad_norm": 0.03453078120946884,
      "learning_rate": 5.621161072805483e-06,
      "loss": 1.48828125,
      "step": 107
    },
    {
      "epoch": 2.769230769230769,
      "grad_norm": 0.033561594784259796,
      "learning_rate": 5.540391363164491e-06,
      "loss": 1.486328125,
      "step": 108
    },
    {
      "epoch": 2.7948717948717947,
      "grad_norm": 0.03299626708030701,
      "learning_rate": 5.45960863683551e-06,
      "loss": 1.5478515625,
      "step": 109
    },
    {
      "epoch": 2.8205128205128203,
      "grad_norm": 0.033326972275972366,
      "learning_rate": 5.378838927194518e-06,
      "loss": 1.5830078125,
      "step": 110
    },
    {
      "epoch": 2.8461538461538463,
      "grad_norm": 0.033495061099529266,
      "learning_rate": 5.298108263422686e-06,
      "loss": 1.5712890625,
      "step": 111
    },
    {
      "epoch": 2.871794871794872,
      "grad_norm": 0.031870946288108826,
      "learning_rate": 5.217442662118091e-06,
      "loss": 1.3175048828125,
      "step": 112
    },
    {
      "epoch": 2.8974358974358974,
      "grad_norm": 0.034332484006881714,
      "learning_rate": 5.136868118911523e-06,
      "loss": 1.5380859375,
      "step": 113
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 0.039312176406383514,
      "learning_rate": 5.0564106000890044e-06,
      "loss": 1.4951171875,
      "step": 114
    },
    {
      "epoch": 2.948717948717949,
      "grad_norm": 0.03217581659555435,
      "learning_rate": 4.976096034223798e-06,
      "loss": 1.5830078125,
      "step": 115
    },
    {
      "epoch": 2.9743589743589745,
      "grad_norm": 0.0373978465795517,
      "learning_rate": 4.895950303820552e-06,
      "loss": 1.5341796875,
      "step": 116
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.0322604663670063,
      "learning_rate": 4.815999236974298e-06,
      "loss": 1.525390625,
      "step": 117
    },
    {
      "epoch": 3.0,
      "eval_loss": 1.542724609375,
      "eval_runtime": 2.0242,
      "eval_samples_per_second": 1.482,
      "eval_steps_per_second": 0.494,
      "step": 117
    },
    {
      "epoch": 3.0256410256410255,
      "grad_norm": 0.03109799697995186,
      "learning_rate": 4.73626859904699e-06,
      "loss": 1.5341796875,
      "step": 118
    },
    {
      "epoch": 3.051282051282051,
      "grad_norm": 0.03308187052607536,
      "learning_rate": 4.656784084364239e-06,
      "loss": 1.474609375,
      "step": 119
    },
    {
      "epoch": 3.076923076923077,
      "grad_norm": 0.03196786716580391,
      "learning_rate": 4.577571307934964e-06,
      "loss": 1.5771484375,
      "step": 120
    },
    {
      "epoch": 3.1025641025641026,
      "grad_norm": 0.03189630061388016,
      "learning_rate": 4.4986557971965865e-06,
      "loss": 1.5263671875,
      "step": 121
    },
    {
      "epoch": 3.128205128205128,
      "grad_norm": 0.031320199370384216,
      "learning_rate": 4.420062983788453e-06,
      "loss": 1.5546875,
      "step": 122
    },
    {
      "epoch": 3.1538461538461537,
      "grad_norm": 0.03848275542259216,
      "learning_rate": 4.3418181953561415e-06,
      "loss": 1.4794921875,
      "step": 123
    },
    {
      "epoch": 3.1794871794871793,
      "grad_norm": 0.03228919208049774,
      "learning_rate": 4.263946647389246e-06,
      "loss": 1.5625,
      "step": 124
    },
    {
      "epoch": 3.2051282051282053,
      "grad_norm": 0.028752503916621208,
      "learning_rate": 4.1864734350953504e-06,
      "loss": 1.2752685546875,
      "step": 125
    },
    {
      "epoch": 3.230769230769231,
      "grad_norm": 0.03756840154528618,
      "learning_rate": 4.109423525312738e-06,
      "loss": 1.5625,
      "step": 126
    },
    {
      "epoch": 3.2564102564102564,
      "grad_norm": 0.04252268746495247,
      "learning_rate": 4.032821748464462e-06,
      "loss": 1.5966796875,
      "step": 127
    },
    {
      "epoch": 3.282051282051282,
      "grad_norm": 0.05631447583436966,
      "learning_rate": 3.956692790556409e-06,
      "loss": 1.61279296875,
      "step": 128
    },
    {
      "epoch": 3.3076923076923075,
      "grad_norm": 0.032152410596609116,
      "learning_rate": 3.88106118522187e-06,
      "loss": 1.5029296875,
      "step": 129
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.031788863241672516,
      "learning_rate": 3.8059513058152365e-06,
      "loss": 1.5263671875,
      "step": 130
    },
    {
      "epoch": 3.358974358974359,
      "grad_norm": 0.033455390483140945,
      "learning_rate": 3.7313873575573445e-06,
      "loss": 1.5498046875,
      "step": 131
    },
    {
      "epoch": 3.3846153846153846,
      "grad_norm": 0.03509214520454407,
      "learning_rate": 3.657393369734994e-06,
      "loss": 1.490234375,
      "step": 132
    },
    {
      "epoch": 3.41025641025641,
      "grad_norm": 0.03831712529063225,
      "learning_rate": 3.5839931879571733e-06,
      "loss": 1.287353515625,
      "step": 133
    },
    {
      "epoch": 3.435897435897436,
      "grad_norm": 0.031902801245450974,
      "learning_rate": 3.5112104664704815e-06,
      "loss": 1.54296875,
      "step": 134
    },
    {
      "epoch": 3.4615384615384617,
      "grad_norm": 0.03318481519818306,
      "learning_rate": 3.4390686605361973e-06,
      "loss": 1.5673828125,
      "step": 135
    },
    {
      "epoch": 3.4871794871794872,
      "grad_norm": 0.0345056876540184,
      "learning_rate": 3.3675910188715067e-06,
      "loss": 1.4208984375,
      "step": 136
    },
    {
      "epoch": 3.5128205128205128,
      "grad_norm": 0.036722343415021896,
      "learning_rate": 3.2968005761572664e-06,
      "loss": 1.5048828125,
      "step": 137
    },
    {
      "epoch": 3.5384615384615383,
      "grad_norm": 0.03170086815953255,
      "learning_rate": 3.2267201456147457e-06,
      "loss": 1.5068359375,
      "step": 138
    },
    {
      "epoch": 3.564102564102564,
      "grad_norm": 0.03775108978152275,
      "learning_rate": 3.157372311653763e-06,
      "loss": 1.4658203125,
      "step": 139
    },
    {
      "epoch": 3.58974358974359,
      "grad_norm": 0.03674975037574768,
      "learning_rate": 3.0887794225945143e-06,
      "loss": 1.51025390625,
      "step": 140
    },
    {
      "epoch": 3.6153846153846154,
      "grad_norm": 0.03359014168381691,
      "learning_rate": 3.020963583465539e-06,
      "loss": 1.4833984375,
      "step": 141
    },
    {
      "epoch": 3.641025641025641,
      "grad_norm": 0.033196352422237396,
      "learning_rate": 2.953946648880035e-06,
      "loss": 1.568359375,
      "step": 142
    },
    {
      "epoch": 3.6666666666666665,
      "grad_norm": 0.03336072713136673,
      "learning_rate": 2.887750215992919e-06,
      "loss": 1.54150390625,
      "step": 143
    },
    {
      "epoch": 3.6923076923076925,
      "grad_norm": 0.033981941640377045,
      "learning_rate": 2.822395617540834e-06,
      "loss": 1.501953125,
      "step": 144
    },
    {
      "epoch": 3.717948717948718,
      "grad_norm": 0.03145177289843559,
      "learning_rate": 2.757903914967364e-06,
      "loss": 1.58203125,
      "step": 145
    },
    {
      "epoch": 3.7435897435897436,
      "grad_norm": 0.033963125199079514,
      "learning_rate": 2.6942958916356997e-06,
      "loss": 1.4736328125,
      "step": 146
    },
    {
      "epoch": 3.769230769230769,
      "grad_norm": 0.03317960724234581,
      "learning_rate": 2.6315920461308963e-06,
      "loss": 1.4736328125,
      "step": 147
    },
    {
      "epoch": 3.7948717948717947,
      "grad_norm": 0.03243858739733696,
      "learning_rate": 2.569812585653919e-06,
      "loss": 1.5322265625,
      "step": 148
    },
    {
      "epoch": 3.8205128205128203,
      "grad_norm": 0.03254690393805504,
      "learning_rate": 2.508977419509586e-06,
      "loss": 1.564453125,
      "step": 149
    },
    {
      "epoch": 3.8461538461538463,
      "grad_norm": 0.03309546038508415,
      "learning_rate": 2.4491061526905075e-06,
      "loss": 1.5556640625,
      "step": 150
    },
    {
      "epoch": 3.871794871794872,
      "grad_norm": 0.031699035316705704,
      "learning_rate": 2.390218079559109e-06,
      "loss": 1.3040771484375,
      "step": 151
    },
    {
      "epoch": 3.8974358974358974,
      "grad_norm": 0.03398076817393303,
      "learning_rate": 2.332332177629748e-06,
      "loss": 1.52197265625,
      "step": 152
    },
    {
      "epoch": 3.9230769230769234,
      "grad_norm": 0.038682978600263596,
      "learning_rate": 2.275467101452935e-06,
      "loss": 1.478515625,
      "step": 153
    },
    {
      "epoch": 3.948717948717949,
      "grad_norm": 0.03174389898777008,
      "learning_rate": 2.2196411766036492e-06,
      "loss": 1.5703125,
      "step": 154
    },
    {
      "epoch": 3.9743589743589745,
      "grad_norm": 0.03697316348552704,
      "learning_rate": 2.1648723937756543e-06,
      "loss": 1.5205078125,
      "step": 155
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.03174947574734688,
      "learning_rate": 2.1111784029837513e-06,
      "loss": 1.5146484375,
      "step": 156
    },
    {
      "epoch": 4.0,
      "eval_loss": 1.52197265625,
      "eval_runtime": 2.0433,
      "eval_samples_per_second": 1.468,
      "eval_steps_per_second": 0.489,
      "step": 156
    },
    {
      "epoch": 4.0256410256410255,
      "grad_norm": 0.030488736927509308,
      "learning_rate": 2.0585765078758076e-06,
      "loss": 1.5224609375,
      "step": 157
    },
    {
      "epoch": 4.051282051282051,
      "grad_norm": 0.03281666338443756,
      "learning_rate": 2.0070836601564125e-06,
      "loss": 1.4599609375,
      "step": 158
    },
    {
      "epoch": 4.076923076923077,
      "grad_norm": 0.03166268765926361,
      "learning_rate": 1.956716454123946e-06,
      "loss": 1.560546875,
      "step": 159
    },
    {
      "epoch": 4.102564102564102,
      "grad_norm": 0.03180032968521118,
      "learning_rate": 1.9074911213228443e-06,
      "loss": 1.5146484375,
      "step": 160
    },
    {
      "epoch": 4.128205128205128,
      "grad_norm": 0.030374251306056976,
      "learning_rate": 1.8594235253127373e-06,
      "loss": 1.5458984375,
      "step": 161
    },
    {
      "epoch": 4.153846153846154,
      "grad_norm": 0.03813847526907921,
      "learning_rate": 1.8125291565562066e-06,
      "loss": 1.466796875,
      "step": 162
    },
    {
      "epoch": 4.17948717948718,
      "grad_norm": 0.0319676548242569,
      "learning_rate": 1.7668231274267611e-06,
      "loss": 1.5498046875,
      "step": 163
    },
    {
      "epoch": 4.205128205128205,
      "grad_norm": 0.027972424402832985,
      "learning_rate": 1.7223201673386525e-06,
      "loss": 1.267333984375,
      "step": 164
    },
    {
      "epoch": 4.230769230769231,
      "grad_norm": 0.03631917014718056,
      "learning_rate": 1.6790346180001293e-06,
      "loss": 1.5517578125,
      "step": 165
    },
    {
      "epoch": 4.256410256410256,
      "grad_norm": 0.04067970812320709,
      "learning_rate": 1.6369804287916028e-06,
      "loss": 1.5849609375,
      "step": 166
    },
    {
      "epoch": 4.282051282051282,
      "grad_norm": 0.054668668657541275,
      "learning_rate": 1.596171152270271e-06,
      "loss": 1.60107421875,
      "step": 167
    },
    {
      "epoch": 4.3076923076923075,
      "grad_norm": 0.03209273889660835,
      "learning_rate": 1.556619939802615e-06,
      "loss": 1.49609375,
      "step": 168
    },
    {
      "epoch": 4.333333333333333,
      "grad_norm": 0.031059790402650833,
      "learning_rate": 1.518339537326177e-06,
      "loss": 1.5166015625,
      "step": 169
    },
    {
      "epoch": 4.358974358974359,
      "grad_norm": 0.03289590775966644,
      "learning_rate": 1.4813422812420164e-06,
      "loss": 1.5400390625,
      "step": 170
    },
    {
      "epoch": 4.384615384615385,
      "grad_norm": 0.033618438988924026,
      "learning_rate": 1.4456400944391147e-06,
      "loss": 1.482421875,
      "step": 171
    },
    {
      "epoch": 4.410256410256411,
      "grad_norm": 0.03935592621564865,
      "learning_rate": 1.411244482452065e-06,
      "loss": 1.2802734375,
      "step": 172
    },
    {
      "epoch": 4.435897435897436,
      "grad_norm": 0.03122033178806305,
      "learning_rate": 1.3781665297532512e-06,
      "loss": 1.537109375,
      "step": 173
    },
    {
      "epoch": 4.461538461538462,
      "grad_norm": 0.031517501920461655,
      "learning_rate": 1.3464168961807122e-06,
      "loss": 1.556640625,
      "step": 174
    },
    {
      "epoch": 4.487179487179487,
      "grad_norm": 0.03323807194828987,
      "learning_rate": 1.316005813502869e-06,
      "loss": 1.412109375,
      "step": 175
    },
    {
      "epoch": 4.512820512820513,
      "grad_norm": 0.035576991736888885,
      "learning_rate": 1.2869430821211828e-06,
      "loss": 1.4970703125,
      "step": 176
    },
    {
      "epoch": 4.538461538461538,
      "grad_norm": 0.031376902014017105,
      "learning_rate": 1.2592380679118437e-06,
      "loss": 1.5,
      "step": 177
    },
    {
      "epoch": 4.564102564102564,
      "grad_norm": 0.0379689559340477,
      "learning_rate": 1.2328996992074813e-06,
      "loss": 1.45849609375,
      "step": 178
    },
    {
      "epoch": 4.589743589743589,
      "grad_norm": 0.0368792749941349,
      "learning_rate": 1.2079364639198862e-06,
      "loss": 1.50634765625,
      "step": 179
    },
    {
      "epoch": 4.615384615384615,
      "grad_norm": 0.03351832553744316,
      "learning_rate": 1.1843564068046551e-06,
      "loss": 1.4755859375,
      "step": 180
    },
    {
      "epoch": 4.641025641025641,
      "grad_norm": 0.03347695618867874,
      "learning_rate": 1.1621671268686605e-06,
      "loss": 1.5615234375,
      "step": 181
    },
    {
      "epoch": 4.666666666666667,
      "grad_norm": 0.033549714833498,
      "learning_rate": 1.1413757749211602e-06,
      "loss": 1.53369140625,
      "step": 182
    },
    {
      "epoch": 4.6923076923076925,
      "grad_norm": 0.0333133228123188,
      "learning_rate": 1.121989051269348e-06,
      "loss": 1.49609375,
      "step": 183
    },
    {
      "epoch": 4.717948717948718,
      "grad_norm": 0.03141152113676071,
      "learning_rate": 1.104013203559086e-06,
      "loss": 1.576171875,
      "step": 184
    },
    {
      "epoch": 4.743589743589744,
      "grad_norm": 0.033574286848306656,
      "learning_rate": 1.08745402476151e-06,
      "loss": 1.46484375,
      "step": 185
    },
    {
      "epoch": 4.769230769230769,
      "grad_norm": 0.03336711600422859,
      "learning_rate": 1.0723168513061665e-06,
      "loss": 1.4677734375,
      "step": 186
    },
    {
      "epoch": 4.794871794871795,
      "grad_norm": 0.03177247568964958,
      "learning_rate": 1.0586065613612696e-06,
      "loss": 1.5263671875,
      "step": 187
    },
    {
      "epoch": 4.82051282051282,
      "grad_norm": 0.032153815031051636,
      "learning_rate": 1.0463275732616432e-06,
      "loss": 1.5576171875,
      "step": 188
    },
    {
      "epoch": 4.846153846153846,
      "grad_norm": 0.03333226218819618,
      "learning_rate": 1.0354838440848502e-06,
      "loss": 1.5478515625,
      "step": 189
    },
    {
      "epoch": 4.871794871794872,
      "grad_norm": 0.03197743743658066,
      "learning_rate": 1.0260788683759687e-06,
      "loss": 1.2972412109375,
      "step": 190
    },
    {
      "epoch": 4.897435897435898,
      "grad_norm": 0.03383539244532585,
      "learning_rate": 1.0181156770214243e-06,
      "loss": 1.51611328125,
      "step": 191
    },
    {
      "epoch": 4.923076923076923,
      "grad_norm": 0.038513340055942535,
      "learning_rate": 1.011596836272247e-06,
      "loss": 1.4755859375,
      "step": 192
    },
    {
      "epoch": 4.948717948717949,
      "grad_norm": 0.031770892441272736,
      "learning_rate": 1.006524446917056e-06,
      "loss": 1.5625,
      "step": 193
    },
    {
      "epoch": 4.9743589743589745,
      "grad_norm": 0.037130434066057205,
      "learning_rate": 1.002900143605054e-06,
      "loss": 1.513671875,
      "step": 194
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.03213154524564743,
      "learning_rate": 1.0007250943192382e-06,
      "loss": 1.51171875,
      "step": 195
    },
    {
      "epoch": 5.0,
      "eval_loss": 1.521728515625,
      "eval_runtime": 2.0346,
      "eval_samples_per_second": 1.474,
      "eval_steps_per_second": 0.491,
      "step": 195
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 195,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 0,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.5737673183474483e+18,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}