{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 4.940092165898617,
  "eval_steps": 500,
  "global_step": 360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.013824884792626729,
      "grad_norm": 34.963134765625,
      "learning_rate": 5.0000000000000004e-08,
      "loss": 2.5476,
      "step": 1
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 35.32600021362305,
      "learning_rate": 1.0000000000000001e-07,
      "loss": 2.6058,
      "step": 2
    },
    {
      "epoch": 0.041474654377880185,
      "grad_norm": 34.955448150634766,
      "learning_rate": 1.5000000000000002e-07,
      "loss": 2.5871,
      "step": 3
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 35.09806442260742,
      "learning_rate": 2.0000000000000002e-07,
      "loss": 2.5912,
      "step": 4
    },
    {
      "epoch": 0.06912442396313365,
      "grad_norm": 34.88739776611328,
      "learning_rate": 2.5000000000000004e-07,
      "loss": 2.592,
      "step": 5
    },
    {
      "epoch": 0.08294930875576037,
      "grad_norm": 34.84288024902344,
      "learning_rate": 3.0000000000000004e-07,
      "loss": 2.5609,
      "step": 6
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 35.0090217590332,
      "learning_rate": 3.5000000000000004e-07,
      "loss": 2.5651,
      "step": 7
    },
    {
      "epoch": 0.11059907834101383,
      "grad_norm": 35.03983688354492,
      "learning_rate": 4.0000000000000003e-07,
      "loss": 2.5437,
      "step": 8
    },
    {
      "epoch": 0.12442396313364056,
      "grad_norm": 34.802833557128906,
      "learning_rate": 4.5000000000000003e-07,
      "loss": 2.5448,
      "step": 9
    },
    {
      "epoch": 0.1382488479262673,
      "grad_norm": 34.5220947265625,
      "learning_rate": 5.000000000000001e-07,
      "loss": 2.504,
      "step": 10
    },
    {
      "epoch": 0.15207373271889402,
      "grad_norm": 34.401580810546875,
      "learning_rate": 5.5e-07,
      "loss": 2.4814,
      "step": 11
    },
    {
      "epoch": 0.16589861751152074,
      "grad_norm": 33.76997375488281,
      "learning_rate": 6.000000000000001e-07,
      "loss": 2.4282,
      "step": 12
    },
    {
      "epoch": 0.17972350230414746,
      "grad_norm": 33.53415298461914,
      "learning_rate": 6.5e-07,
      "loss": 2.4216,
      "step": 13
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 32.401580810546875,
      "learning_rate": 7.000000000000001e-07,
      "loss": 2.3362,
      "step": 14
    },
    {
      "epoch": 0.2073732718894009,
      "grad_norm": 33.636661529541016,
      "learning_rate": 7.5e-07,
      "loss": 2.2978,
      "step": 15
    },
    {
      "epoch": 0.22119815668202766,
      "grad_norm": 31.3782901763916,
      "learning_rate": 8.000000000000001e-07,
      "loss": 2.1358,
      "step": 16
    },
    {
      "epoch": 0.2350230414746544,
      "grad_norm": 30.72391700744629,
      "learning_rate": 8.500000000000001e-07,
      "loss": 2.0652,
      "step": 17
    },
    {
      "epoch": 0.2488479262672811,
      "grad_norm": 30.817584991455078,
      "learning_rate": 9.000000000000001e-07,
      "loss": 2.0115,
      "step": 18
    },
    {
      "epoch": 0.2626728110599078,
      "grad_norm": 29.683996200561523,
      "learning_rate": 9.500000000000001e-07,
      "loss": 1.8668,
      "step": 19
    },
    {
      "epoch": 0.2764976958525346,
      "grad_norm": 29.506683349609375,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.7796,
      "step": 20
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 27.55340003967285,
      "learning_rate": 1.0500000000000001e-06,
      "loss": 1.5656,
      "step": 21
    },
    {
      "epoch": 0.30414746543778803,
      "grad_norm": 27.78036880493164,
      "learning_rate": 1.1e-06,
      "loss": 1.5112,
      "step": 22
    },
    {
      "epoch": 0.31797235023041476,
      "grad_norm": 26.36115264892578,
      "learning_rate": 1.1500000000000002e-06,
      "loss": 1.3283,
      "step": 23
    },
    {
      "epoch": 0.3317972350230415,
      "grad_norm": 25.388761520385742,
      "learning_rate": 1.2000000000000002e-06,
      "loss": 1.137,
      "step": 24
    },
    {
      "epoch": 0.3456221198156682,
      "grad_norm": 25.21432876586914,
      "learning_rate": 1.25e-06,
      "loss": 0.9867,
      "step": 25
    },
    {
      "epoch": 0.35944700460829493,
      "grad_norm": 24.924489974975586,
      "learning_rate": 1.3e-06,
      "loss": 0.7122,
      "step": 26
    },
    {
      "epoch": 0.37327188940092165,
      "grad_norm": 21.881420135498047,
      "learning_rate": 1.3500000000000002e-06,
      "loss": 0.4952,
      "step": 27
    },
    {
      "epoch": 0.3870967741935484,
      "grad_norm": 17.67154884338379,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 0.3602,
      "step": 28
    },
    {
      "epoch": 0.4009216589861751,
      "grad_norm": 11.489490509033203,
      "learning_rate": 1.45e-06,
      "loss": 0.2432,
      "step": 29
    },
    {
      "epoch": 0.4147465437788018,
      "grad_norm": 7.622438907623291,
      "learning_rate": 1.5e-06,
      "loss": 0.189,
      "step": 30
    },
    {
      "epoch": 0.42857142857142855,
      "grad_norm": 4.340638637542725,
      "learning_rate": 1.5500000000000002e-06,
      "loss": 0.1302,
      "step": 31
    },
    {
      "epoch": 0.4423963133640553,
      "grad_norm": 3.079514980316162,
      "learning_rate": 1.6000000000000001e-06,
      "loss": 0.1075,
      "step": 32
    },
    {
      "epoch": 0.45622119815668205,
      "grad_norm": 2.355943441390991,
      "learning_rate": 1.6500000000000003e-06,
      "loss": 0.0998,
      "step": 33
    },
    {
      "epoch": 0.4700460829493088,
      "grad_norm": 1.9480725526809692,
      "learning_rate": 1.7000000000000002e-06,
      "loss": 0.0926,
      "step": 34
    },
    {
      "epoch": 0.4838709677419355,
      "grad_norm": 1.8598166704177856,
      "learning_rate": 1.75e-06,
      "loss": 0.0733,
      "step": 35
    },
    {
      "epoch": 0.4976958525345622,
      "grad_norm": 0.9892730712890625,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 0.0664,
      "step": 36
    },
    {
      "epoch": 0.511520737327189,
      "grad_norm": 0.8992418050765991,
      "learning_rate": 1.85e-06,
      "loss": 0.0709,
      "step": 37
    },
    {
      "epoch": 0.5253456221198156,
      "grad_norm": 0.7340101599693298,
      "learning_rate": 1.9000000000000002e-06,
      "loss": 0.0535,
      "step": 38
    },
    {
      "epoch": 0.5391705069124424,
      "grad_norm": 0.7032178044319153,
      "learning_rate": 1.9500000000000004e-06,
      "loss": 0.0573,
      "step": 39
    },
    {
      "epoch": 0.5529953917050692,
      "grad_norm": 0.6449429392814636,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 0.0576,
      "step": 40
    },
    {
      "epoch": 0.5668202764976958,
      "grad_norm": 0.6358592510223389,
      "learning_rate": 2.05e-06,
      "loss": 0.0502,
      "step": 41
    },
    {
      "epoch": 0.5806451612903226,
      "grad_norm": 0.572036623954773,
      "learning_rate": 2.1000000000000002e-06,
      "loss": 0.0556,
      "step": 42
    },
    {
      "epoch": 0.5944700460829493,
      "grad_norm": 0.6538863778114319,
      "learning_rate": 2.15e-06,
      "loss": 0.0556,
      "step": 43
    },
    {
      "epoch": 0.6082949308755761,
      "grad_norm": 0.3532159626483917,
      "learning_rate": 2.2e-06,
      "loss": 0.0452,
      "step": 44
    },
    {
      "epoch": 0.6221198156682027,
      "grad_norm": 0.4853012263774872,
      "learning_rate": 2.25e-06,
      "loss": 0.0471,
      "step": 45
    },
    {
      "epoch": 0.6359447004608295,
      "grad_norm": 0.4761648178100586,
      "learning_rate": 2.3000000000000004e-06,
      "loss": 0.0469,
      "step": 46
    },
    {
      "epoch": 0.6497695852534562,
      "grad_norm": 0.6094638109207153,
      "learning_rate": 2.35e-06,
      "loss": 0.047,
      "step": 47
    },
    {
      "epoch": 0.663594470046083,
      "grad_norm": 0.5211306214332581,
      "learning_rate": 2.4000000000000003e-06,
      "loss": 0.0402,
      "step": 48
    },
    {
      "epoch": 0.6774193548387096,
      "grad_norm": 0.2997778356075287,
      "learning_rate": 2.4500000000000003e-06,
      "loss": 0.0425,
      "step": 49
    },
    {
      "epoch": 0.6912442396313364,
      "grad_norm": 0.37834689021110535,
      "learning_rate": 2.5e-06,
      "loss": 0.0446,
      "step": 50
    },
    {
      "epoch": 0.7050691244239631,
      "grad_norm": 0.31011995673179626,
      "learning_rate": 2.55e-06,
      "loss": 0.0406,
      "step": 51
    },
    {
      "epoch": 0.7188940092165899,
      "grad_norm": 0.3113131523132324,
      "learning_rate": 2.6e-06,
      "loss": 0.0368,
      "step": 52
    },
    {
      "epoch": 0.7327188940092166,
      "grad_norm": 0.5685846209526062,
      "learning_rate": 2.6500000000000005e-06,
      "loss": 0.0389,
      "step": 53
    },
    {
      "epoch": 0.7465437788018433,
      "grad_norm": 0.29334983229637146,
      "learning_rate": 2.7000000000000004e-06,
      "loss": 0.0423,
      "step": 54
    },
    {
      "epoch": 0.7603686635944701,
      "grad_norm": 0.5776861906051636,
      "learning_rate": 2.7500000000000004e-06,
      "loss": 0.0399,
      "step": 55
    },
    {
      "epoch": 0.7741935483870968,
      "grad_norm": 0.35423165559768677,
      "learning_rate": 2.8000000000000003e-06,
      "loss": 0.0357,
      "step": 56
    },
    {
      "epoch": 0.7880184331797235,
      "grad_norm": 0.37902742624282837,
      "learning_rate": 2.85e-06,
      "loss": 0.0407,
      "step": 57
    },
    {
      "epoch": 0.8018433179723502,
      "grad_norm": 0.26948878169059753,
      "learning_rate": 2.9e-06,
      "loss": 0.0351,
      "step": 58
    },
    {
      "epoch": 0.815668202764977,
      "grad_norm": 0.35688117146492004,
      "learning_rate": 2.95e-06,
      "loss": 0.0377,
      "step": 59
    },
    {
      "epoch": 0.8294930875576036,
      "grad_norm": 0.5287911891937256,
      "learning_rate": 3e-06,
      "loss": 0.0377,
      "step": 60
    },
    {
      "epoch": 0.8433179723502304,
      "grad_norm": 0.2950785756111145,
      "learning_rate": 3.05e-06,
      "loss": 0.0361,
      "step": 61
    },
    {
      "epoch": 0.8571428571428571,
      "grad_norm": 0.2789723575115204,
      "learning_rate": 3.1000000000000004e-06,
      "loss": 0.032,
      "step": 62
    },
    {
      "epoch": 0.8709677419354839,
      "grad_norm": 0.2802198529243469,
      "learning_rate": 3.1500000000000003e-06,
      "loss": 0.0394,
      "step": 63
    },
    {
      "epoch": 0.8847926267281107,
      "grad_norm": 0.286981463432312,
      "learning_rate": 3.2000000000000003e-06,
      "loss": 0.033,
      "step": 64
    },
    {
      "epoch": 0.8986175115207373,
      "grad_norm": 0.37392762303352356,
      "learning_rate": 3.2500000000000002e-06,
      "loss": 0.0335,
      "step": 65
    },
    {
      "epoch": 0.9124423963133641,
      "grad_norm": 0.25025588274002075,
      "learning_rate": 3.3000000000000006e-06,
      "loss": 0.0311,
      "step": 66
    },
    {
      "epoch": 0.9262672811059908,
      "grad_norm": 0.4292861521244049,
      "learning_rate": 3.3500000000000005e-06,
      "loss": 0.0362,
      "step": 67
    },
    {
      "epoch": 0.9400921658986175,
      "grad_norm": 0.4717651307582855,
      "learning_rate": 3.4000000000000005e-06,
      "loss": 0.0303,
      "step": 68
    },
    {
      "epoch": 0.9539170506912442,
      "grad_norm": 0.49291253089904785,
      "learning_rate": 3.45e-06,
      "loss": 0.0352,
      "step": 69
    },
    {
      "epoch": 0.967741935483871,
      "grad_norm": 0.3729935586452484,
      "learning_rate": 3.5e-06,
      "loss": 0.0297,
      "step": 70
    },
    {
      "epoch": 0.9815668202764977,
      "grad_norm": 0.27150583267211914,
      "learning_rate": 3.5500000000000003e-06,
      "loss": 0.0326,
      "step": 71
    },
    {
      "epoch": 0.9953917050691244,
      "grad_norm": 0.34516096115112305,
      "learning_rate": 3.6000000000000003e-06,
      "loss": 0.0336,
      "step": 72
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.34516096115112305,
      "learning_rate": 3.65e-06,
      "loss": 0.0274,
      "step": 73
    },
    {
      "epoch": 1.0138248847926268,
      "grad_norm": 0.6282734870910645,
      "learning_rate": 3.7e-06,
      "loss": 0.0289,
      "step": 74
    },
    {
      "epoch": 1.0276497695852536,
      "grad_norm": 0.2935558557510376,
      "learning_rate": 3.7500000000000005e-06,
      "loss": 0.0308,
      "step": 75
    },
    {
      "epoch": 1.0414746543778801,
      "grad_norm": 0.3166769742965698,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 0.0277,
      "step": 76
    },
    {
      "epoch": 1.055299539170507,
      "grad_norm": 0.38190239667892456,
      "learning_rate": 3.85e-06,
      "loss": 0.0338,
      "step": 77
    },
    {
      "epoch": 1.0691244239631337,
      "grad_norm": 0.2779421806335449,
      "learning_rate": 3.900000000000001e-06,
      "loss": 0.03,
      "step": 78
    },
    {
      "epoch": 1.0829493087557605,
      "grad_norm": 0.4055996537208557,
      "learning_rate": 3.95e-06,
      "loss": 0.0295,
      "step": 79
    },
    {
      "epoch": 1.096774193548387,
      "grad_norm": 0.2987312972545624,
      "learning_rate": 4.000000000000001e-06,
      "loss": 0.028,
      "step": 80
    },
    {
      "epoch": 1.1105990783410138,
      "grad_norm": 0.2674776017665863,
      "learning_rate": 4.05e-06,
      "loss": 0.0243,
      "step": 81
    },
    {
      "epoch": 1.1244239631336406,
      "grad_norm": 0.29042816162109375,
      "learning_rate": 4.1e-06,
      "loss": 0.0318,
      "step": 82
    },
    {
      "epoch": 1.1382488479262673,
      "grad_norm": 0.2904883027076721,
      "learning_rate": 4.15e-06,
      "loss": 0.0257,
      "step": 83
    },
    {
      "epoch": 1.1520737327188941,
      "grad_norm": 0.30603015422821045,
      "learning_rate": 4.2000000000000004e-06,
      "loss": 0.0284,
      "step": 84
    },
    {
      "epoch": 1.1658986175115207,
      "grad_norm": 0.23131045699119568,
      "learning_rate": 4.25e-06,
      "loss": 0.0285,
      "step": 85
    },
    {
      "epoch": 1.1797235023041475,
      "grad_norm": 0.26788002252578735,
      "learning_rate": 4.3e-06,
      "loss": 0.0269,
      "step": 86
    },
    {
      "epoch": 1.1935483870967742,
      "grad_norm": 0.2639651894569397,
      "learning_rate": 4.350000000000001e-06,
      "loss": 0.0289,
      "step": 87
    },
    {
      "epoch": 1.2073732718894008,
      "grad_norm": 0.25068584084510803,
      "learning_rate": 4.4e-06,
      "loss": 0.0275,
      "step": 88
    },
    {
      "epoch": 1.2211981566820276,
      "grad_norm": 0.25494542717933655,
      "learning_rate": 4.450000000000001e-06,
      "loss": 0.0275,
      "step": 89
    },
    {
      "epoch": 1.2350230414746544,
      "grad_norm": 0.31125035881996155,
      "learning_rate": 4.5e-06,
      "loss": 0.0251,
      "step": 90
    },
    {
      "epoch": 1.2488479262672811,
      "grad_norm": 0.2691773474216461,
      "learning_rate": 4.5500000000000005e-06,
      "loss": 0.0267,
      "step": 91
    },
    {
      "epoch": 1.262672811059908,
      "grad_norm": 0.20079147815704346,
      "learning_rate": 4.600000000000001e-06,
      "loss": 0.0263,
      "step": 92
    },
    {
      "epoch": 1.2764976958525347,
      "grad_norm": 0.28027331829071045,
      "learning_rate": 4.65e-06,
      "loss": 0.0227,
      "step": 93
    },
    {
      "epoch": 1.2903225806451613,
      "grad_norm": 0.40053099393844604,
      "learning_rate": 4.7e-06,
      "loss": 0.0246,
      "step": 94
    },
    {
      "epoch": 1.304147465437788,
      "grad_norm": 0.33066362142562866,
      "learning_rate": 4.75e-06,
      "loss": 0.0221,
      "step": 95
    },
    {
      "epoch": 1.3179723502304148,
      "grad_norm": 0.2531339228153229,
      "learning_rate": 4.800000000000001e-06,
      "loss": 0.0216,
      "step": 96
    },
    {
      "epoch": 1.3317972350230414,
      "grad_norm": 0.37544378638267517,
      "learning_rate": 4.85e-06,
      "loss": 0.0247,
      "step": 97
    },
    {
      "epoch": 1.3456221198156681,
      "grad_norm": 0.34273672103881836,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 0.0217,
      "step": 98
    },
    {
      "epoch": 1.359447004608295,
      "grad_norm": 0.2338661253452301,
      "learning_rate": 4.95e-06,
      "loss": 0.0237,
      "step": 99
    },
    {
      "epoch": 1.3732718894009217,
      "grad_norm": 0.30151981115341187,
      "learning_rate": 5e-06,
      "loss": 0.0248,
      "step": 100
    },
    {
      "epoch": 1.3870967741935485,
      "grad_norm": 0.3205336630344391,
      "learning_rate": 4.999888074163108e-06,
      "loss": 0.0232,
      "step": 101
    },
    {
      "epoch": 1.400921658986175,
      "grad_norm": 0.2705315351486206,
      "learning_rate": 4.999552306674345e-06,
      "loss": 0.0245,
      "step": 102
    },
    {
      "epoch": 1.4147465437788018,
      "grad_norm": 0.2564137578010559,
      "learning_rate": 4.998992727598557e-06,
      "loss": 0.0274,
      "step": 103
    },
    {
      "epoch": 1.4285714285714286,
      "grad_norm": 0.1967611312866211,
      "learning_rate": 4.998209387040829e-06,
      "loss": 0.0173,
      "step": 104
    },
    {
      "epoch": 1.4423963133640554,
      "grad_norm": 0.2568240761756897,
      "learning_rate": 4.9972023551419995e-06,
      "loss": 0.0223,
      "step": 105
    },
    {
      "epoch": 1.456221198156682,
      "grad_norm": 0.2236352413892746,
      "learning_rate": 4.995971722072379e-06,
      "loss": 0.0202,
      "step": 106
    },
    {
      "epoch": 1.4700460829493087,
      "grad_norm": 0.3389627933502197,
      "learning_rate": 4.9945175980236745e-06,
      "loss": 0.0214,
      "step": 107
    },
    {
      "epoch": 1.4838709677419355,
      "grad_norm": 0.31428012251853943,
      "learning_rate": 4.992840113199131e-06,
      "loss": 0.0188,
      "step": 108
    },
    {
      "epoch": 1.4976958525345623,
      "grad_norm": 0.41508516669273376,
      "learning_rate": 4.990939417801859e-06,
      "loss": 0.0213,
      "step": 109
    },
    {
      "epoch": 1.511520737327189,
      "grad_norm": 0.19615545868873596,
      "learning_rate": 4.988815682021398e-06,
      "loss": 0.0191,
      "step": 110
    },
    {
      "epoch": 1.5253456221198156,
      "grad_norm": 0.2059931755065918,
      "learning_rate": 4.986469096018472e-06,
      "loss": 0.0208,
      "step": 111
    },
    {
      "epoch": 1.5391705069124424,
      "grad_norm": 0.26946336030960083,
      "learning_rate": 4.983899869907963e-06,
      "loss": 0.0192,
      "step": 112
    },
    {
      "epoch": 1.5529953917050692,
      "grad_norm": 0.3227538466453552,
      "learning_rate": 4.981108233740096e-06,
      "loss": 0.0169,
      "step": 113
    },
    {
      "epoch": 1.5668202764976957,
      "grad_norm": 0.2811918258666992,
      "learning_rate": 4.978094437479843e-06,
      "loss": 0.0151,
      "step": 114
    },
    {
      "epoch": 1.5806451612903225,
      "grad_norm": 0.32980477809906006,
      "learning_rate": 4.97485875098454e-06,
      "loss": 0.0182,
      "step": 115
    },
    {
      "epoch": 1.5944700460829493,
      "grad_norm": 0.2759259045124054,
      "learning_rate": 4.971401463979722e-06,
      "loss": 0.0192,
      "step": 116
    },
    {
      "epoch": 1.608294930875576,
      "grad_norm": 0.2572178840637207,
      "learning_rate": 4.967722886033181e-06,
      "loss": 0.0198,
      "step": 117
    },
    {
      "epoch": 1.6221198156682028,
      "grad_norm": 0.3238658905029297,
      "learning_rate": 4.963823346527249e-06,
      "loss": 0.0186,
      "step": 118
    },
    {
      "epoch": 1.6359447004608296,
      "grad_norm": 0.3834918737411499,
      "learning_rate": 4.959703194629304e-06,
      "loss": 0.0188,
      "step": 119
    },
    {
      "epoch": 1.6497695852534562,
      "grad_norm": 0.23881244659423828,
      "learning_rate": 4.955362799260507e-06,
      "loss": 0.0182,
      "step": 120
    },
    {
      "epoch": 1.663594470046083,
      "grad_norm": 0.1885918825864792,
      "learning_rate": 4.950802549062764e-06,
      "loss": 0.0183,
      "step": 121
    },
    {
      "epoch": 1.6774193548387095,
      "grad_norm": 0.34959614276885986,
      "learning_rate": 4.946022852363932e-06,
      "loss": 0.0173,
      "step": 122
    },
    {
      "epoch": 1.6912442396313363,
      "grad_norm": 0.22990310192108154,
      "learning_rate": 4.9410241371412525e-06,
      "loss": 0.0135,
      "step": 123
    },
    {
      "epoch": 1.705069124423963,
      "grad_norm": 0.2790350615978241,
      "learning_rate": 4.935806850983034e-06,
      "loss": 0.0159,
      "step": 124
    },
    {
      "epoch": 1.7188940092165899,
      "grad_norm": 0.3218020796775818,
      "learning_rate": 4.9303714610485705e-06,
      "loss": 0.0176,
      "step": 125
    },
    {
      "epoch": 1.7327188940092166,
      "grad_norm": 0.2294609695672989,
      "learning_rate": 4.924718454026318e-06,
      "loss": 0.0149,
      "step": 126
    },
    {
      "epoch": 1.7465437788018434,
      "grad_norm": 0.3427927494049072,
      "learning_rate": 4.918848336090309e-06,
      "loss": 0.0165,
      "step": 127
    },
    {
      "epoch": 1.7603686635944702,
      "grad_norm": 0.22731825709342957,
      "learning_rate": 4.912761632854834e-06,
      "loss": 0.0145,
      "step": 128
    },
    {
      "epoch": 1.7741935483870968,
      "grad_norm": 0.35364386439323425,
      "learning_rate": 4.906458889327375e-06,
      "loss": 0.0161,
      "step": 129
    },
    {
      "epoch": 1.7880184331797235,
      "grad_norm": 0.29476454854011536,
      "learning_rate": 4.899940669859807e-06,
      "loss": 0.0154,
      "step": 130
    },
    {
      "epoch": 1.80184331797235,
      "grad_norm": 0.28667864203453064,
      "learning_rate": 4.893207558097867e-06,
      "loss": 0.0143,
      "step": 131
    },
    {
      "epoch": 1.8156682027649769,
      "grad_norm": 0.2731999158859253,
      "learning_rate": 4.8862601569288885e-06,
      "loss": 0.0141,
      "step": 132
    },
    {
      "epoch": 1.8294930875576036,
      "grad_norm": 0.2670470178127289,
      "learning_rate": 4.879099088427824e-06,
      "loss": 0.0131,
      "step": 133
    },
    {
      "epoch": 1.8433179723502304,
      "grad_norm": 0.23313525319099426,
      "learning_rate": 4.871724993801541e-06,
      "loss": 0.012,
      "step": 134
    },
    {
      "epoch": 1.8571428571428572,
      "grad_norm": 0.2192607820034027,
      "learning_rate": 4.864138533331411e-06,
      "loss": 0.0125,
      "step": 135
    },
    {
      "epoch": 1.870967741935484,
      "grad_norm": 0.26603585481643677,
      "learning_rate": 4.8563403863141825e-06,
      "loss": 0.0121,
      "step": 136
    },
    {
      "epoch": 1.8847926267281108,
      "grad_norm": 0.32500001788139343,
      "learning_rate": 4.84833125100116e-06,
      "loss": 0.0116,
      "step": 137
    },
    {
      "epoch": 1.8986175115207373,
      "grad_norm": 0.24893291294574738,
      "learning_rate": 4.840111844535682e-06,
      "loss": 0.0119,
      "step": 138
    },
    {
      "epoch": 1.912442396313364,
      "grad_norm": 0.17670764029026031,
      "learning_rate": 4.8316829028889076e-06,
      "loss": 0.0096,
      "step": 139
    },
    {
      "epoch": 1.9262672811059907,
      "grad_norm": 0.16747575998306274,
      "learning_rate": 4.823045180793914e-06,
      "loss": 0.0113,
      "step": 140
    },
    {
      "epoch": 1.9400921658986174,
      "grad_norm": 0.19587458670139313,
      "learning_rate": 4.8141994516781196e-06,
      "loss": 0.0111,
      "step": 141
    },
    {
      "epoch": 1.9539170506912442,
      "grad_norm": 0.237543985247612,
      "learning_rate": 4.805146507594034e-06,
      "loss": 0.0088,
      "step": 142
    },
    {
      "epoch": 1.967741935483871,
      "grad_norm": 0.22710399329662323,
      "learning_rate": 4.7958871591483305e-06,
      "loss": 0.0085,
      "step": 143
    },
    {
      "epoch": 1.9815668202764978,
      "grad_norm": 0.2946629822254181,
      "learning_rate": 4.786422235429269e-06,
      "loss": 0.0122,
      "step": 144
    },
    {
      "epoch": 1.9953917050691246,
      "grad_norm": 0.2763853371143341,
      "learning_rate": 4.776752583932455e-06,
      "loss": 0.0118,
      "step": 145
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.2763853371143341,
      "learning_rate": 4.766879070484957e-06,
      "loss": 0.0078,
      "step": 146
    },
    {
      "epoch": 2.013824884792627,
      "grad_norm": 0.2722196877002716,
      "learning_rate": 4.756802579167781e-06,
      "loss": 0.0076,
      "step": 147
    },
    {
      "epoch": 2.0276497695852536,
      "grad_norm": 0.18556565046310425,
      "learning_rate": 4.746524012236706e-06,
      "loss": 0.0091,
      "step": 148
    },
    {
      "epoch": 2.0414746543778803,
      "grad_norm": 0.24442361295223236,
      "learning_rate": 4.736044290041496e-06,
      "loss": 0.009,
      "step": 149
    },
    {
      "epoch": 2.055299539170507,
      "grad_norm": 0.24207571148872375,
      "learning_rate": 4.725364350943492e-06,
      "loss": 0.0085,
      "step": 150
    },
    {
      "epoch": 2.0691244239631335,
      "grad_norm": 0.18502290546894073,
      "learning_rate": 4.714485151231593e-06,
      "loss": 0.0059,
      "step": 151
    },
    {
      "epoch": 2.0829493087557602,
      "grad_norm": 0.3010450303554535,
      "learning_rate": 4.703407665036622e-06,
      "loss": 0.0071,
      "step": 152
    },
    {
      "epoch": 2.096774193548387,
      "grad_norm": 0.23272967338562012,
      "learning_rate": 4.692132884244113e-06,
      "loss": 0.0074,
      "step": 153
    },
    {
      "epoch": 2.110599078341014,
      "grad_norm": 0.25476181507110596,
      "learning_rate": 4.680661818405485e-06,
      "loss": 0.0082,
      "step": 154
    },
    {
      "epoch": 2.1244239631336406,
      "grad_norm": 0.24534538388252258,
      "learning_rate": 4.668995494647653e-06,
      "loss": 0.0065,
      "step": 155
    },
    {
      "epoch": 2.1382488479262673,
      "grad_norm": 0.1642732173204422,
      "learning_rate": 4.657134957581057e-06,
      "loss": 0.0054,
      "step": 156
    },
    {
      "epoch": 2.152073732718894,
      "grad_norm": 0.21100501716136932,
      "learning_rate": 4.645081269206128e-06,
      "loss": 0.0091,
      "step": 157
    },
    {
      "epoch": 2.165898617511521,
      "grad_norm": 0.19043587148189545,
      "learning_rate": 4.632835508818192e-06,
      "loss": 0.0047,
      "step": 158
    },
    {
      "epoch": 2.1797235023041477,
      "grad_norm": 0.1804375797510147,
      "learning_rate": 4.620398772910833e-06,
      "loss": 0.0068,
      "step": 159
    },
    {
      "epoch": 2.193548387096774,
      "grad_norm": 0.6586657762527466,
      "learning_rate": 4.607772175077712e-06,
      "loss": 0.0049,
      "step": 160
    },
    {
      "epoch": 2.207373271889401,
      "grad_norm": 0.18181656301021576,
      "learning_rate": 4.59495684591285e-06,
      "loss": 0.0071,
      "step": 161
    },
    {
      "epoch": 2.2211981566820276,
      "grad_norm": 0.760053813457489,
      "learning_rate": 4.581953932909403e-06,
      "loss": 0.0065,
      "step": 162
    },
    {
      "epoch": 2.2350230414746544,
      "grad_norm": 0.1935238242149353,
      "learning_rate": 4.5687646003569055e-06,
      "loss": 0.0066,
      "step": 163
    },
    {
      "epoch": 2.248847926267281,
      "grad_norm": 0.3035024404525757,
      "learning_rate": 4.555390029237026e-06,
      "loss": 0.0046,
      "step": 164
    },
    {
      "epoch": 2.262672811059908,
      "grad_norm": 0.16596420109272003,
      "learning_rate": 4.541831417117815e-06,
      "loss": 0.007,
      "step": 165
    },
    {
      "epoch": 2.2764976958525347,
      "grad_norm": 0.2578873336315155,
      "learning_rate": 4.528089978046481e-06,
      "loss": 0.0048,
      "step": 166
    },
    {
      "epoch": 2.2903225806451615,
      "grad_norm": 1.7751781940460205,
      "learning_rate": 4.514166942440679e-06,
      "loss": 0.0041,
      "step": 167
    },
    {
      "epoch": 2.3041474654377883,
      "grad_norm": 0.37872445583343506,
      "learning_rate": 4.5000635569783365e-06,
      "loss": 0.0045,
      "step": 168
    },
    {
      "epoch": 2.3179723502304146,
      "grad_norm": 0.22949594259262085,
      "learning_rate": 4.4857810844860325e-06,
      "loss": 0.0071,
      "step": 169
    },
    {
      "epoch": 2.3317972350230414,
      "grad_norm": 0.34662699699401855,
      "learning_rate": 4.471320803825915e-06,
      "loss": 0.006,
      "step": 170
    },
    {
      "epoch": 2.345622119815668,
      "grad_norm": 0.5892661213874817,
      "learning_rate": 4.4566840097811956e-06,
      "loss": 0.0055,
      "step": 171
    },
    {
      "epoch": 2.359447004608295,
      "grad_norm": 0.18866907060146332,
      "learning_rate": 4.4418720129402145e-06,
      "loss": 0.0036,
      "step": 172
    },
    {
      "epoch": 2.3732718894009217,
      "grad_norm": 0.1510942429304123,
      "learning_rate": 4.426886139579083e-06,
      "loss": 0.0065,
      "step": 173
    },
    {
      "epoch": 2.3870967741935485,
      "grad_norm": 0.21291828155517578,
      "learning_rate": 4.411727731542937e-06,
      "loss": 0.004,
      "step": 174
    },
    {
      "epoch": 2.4009216589861753,
      "grad_norm": 0.18649035692214966,
      "learning_rate": 4.39639814612578e-06,
      "loss": 0.0047,
      "step": 175
    },
    {
      "epoch": 2.4147465437788016,
      "grad_norm": 0.19008278846740723,
      "learning_rate": 4.3808987559489536e-06,
      "loss": 0.0071,
      "step": 176
    },
    {
      "epoch": 2.4285714285714284,
      "grad_norm": 0.26282456517219543,
      "learning_rate": 4.365230948838232e-06,
      "loss": 0.0044,
      "step": 177
    },
    {
      "epoch": 2.442396313364055,
      "grad_norm": 0.2351403385400772,
      "learning_rate": 4.349396127699552e-06,
      "loss": 0.0057,
      "step": 178
    },
    {
      "epoch": 2.456221198156682,
      "grad_norm": 0.20451441407203674,
      "learning_rate": 4.3333957103934025e-06,
      "loss": 0.003,
      "step": 179
    },
    {
      "epoch": 2.4700460829493087,
      "grad_norm": 0.22120380401611328,
      "learning_rate": 4.317231129607859e-06,
      "loss": 0.0045,
      "step": 180
    },
    {
      "epoch": 2.4838709677419355,
      "grad_norm": 0.18543967604637146,
      "learning_rate": 4.30090383273031e-06,
      "loss": 0.0062,
      "step": 181
    },
    {
      "epoch": 2.4976958525345623,
      "grad_norm": 0.18473730981349945,
      "learning_rate": 4.2844152817178476e-06,
      "loss": 0.0052,
      "step": 182
    },
    {
      "epoch": 2.511520737327189,
      "grad_norm": 0.21087361872196198,
      "learning_rate": 4.267766952966369e-06,
      "loss": 0.0041,
      "step": 183
    },
    {
      "epoch": 2.525345622119816,
      "grad_norm": 0.24977360665798187,
      "learning_rate": 4.2509603371783776e-06,
      "loss": 0.0053,
      "step": 184
    },
    {
      "epoch": 2.539170506912442,
      "grad_norm": 0.19377018511295319,
      "learning_rate": 4.233996939229502e-06,
      "loss": 0.0037,
      "step": 185
    },
    {
      "epoch": 2.5529953917050694,
      "grad_norm": 0.21130548417568207,
      "learning_rate": 4.216878278033753e-06,
      "loss": 0.0031,
      "step": 186
    },
    {
      "epoch": 2.5668202764976957,
      "grad_norm": 0.13288047909736633,
      "learning_rate": 4.199605886407515e-06,
      "loss": 0.0029,
      "step": 187
    },
    {
      "epoch": 2.5806451612903225,
      "grad_norm": 0.15998876094818115,
      "learning_rate": 4.1821813109322975e-06,
      "loss": 0.0031,
      "step": 188
    },
    {
      "epoch": 2.5944700460829493,
      "grad_norm": 0.19475506246089935,
      "learning_rate": 4.164606111816256e-06,
      "loss": 0.0022,
      "step": 189
    },
    {
      "epoch": 2.608294930875576,
      "grad_norm": 0.1446300446987152,
      "learning_rate": 4.146881862754485e-06,
      "loss": 0.0025,
      "step": 190
    },
    {
      "epoch": 2.622119815668203,
      "grad_norm": 0.13051164150238037,
      "learning_rate": 4.129010150788112e-06,
      "loss": 0.0019,
      "step": 191
    },
    {
      "epoch": 2.6359447004608296,
      "grad_norm": 0.1953984946012497,
      "learning_rate": 4.110992576162193e-06,
      "loss": 0.0021,
      "step": 192
    },
    {
      "epoch": 2.6497695852534564,
      "grad_norm": 0.23630598187446594,
      "learning_rate": 4.092830752182423e-06,
      "loss": 0.002,
      "step": 193
    },
    {
      "epoch": 2.6635944700460827,
      "grad_norm": 0.2919062376022339,
      "learning_rate": 4.074526305070679e-06,
      "loss": 0.0017,
      "step": 194
    },
    {
      "epoch": 2.6774193548387095,
      "grad_norm": 0.22015534341335297,
      "learning_rate": 4.056080873819412e-06,
      "loss": 0.0025,
      "step": 195
    },
    {
      "epoch": 2.6912442396313363,
      "grad_norm": 0.9449160099029541,
      "learning_rate": 4.037496110044885e-06,
      "loss": 0.0024,
      "step": 196
    },
    {
      "epoch": 2.705069124423963,
      "grad_norm": 0.25235581398010254,
      "learning_rate": 4.018773677839289e-06,
      "loss": 0.0031,
      "step": 197
    },
    {
      "epoch": 2.71889400921659,
      "grad_norm": 0.3098089098930359,
      "learning_rate": 3.999915253621739e-06,
      "loss": 0.0019,
      "step": 198
    },
    {
      "epoch": 2.7327188940092166,
      "grad_norm": 0.19896291196346283,
      "learning_rate": 3.980922525988167e-06,
      "loss": 0.0019,
      "step": 199
    },
    {
      "epoch": 2.7465437788018434,
      "grad_norm": 0.21136268973350525,
      "learning_rate": 3.961797195560118e-06,
      "loss": 0.0031,
      "step": 200
    },
    {
      "epoch": 2.76036866359447,
      "grad_norm": 0.2549005150794983,
      "learning_rate": 3.942540974832486e-06,
      "loss": 0.0017,
      "step": 201
    },
    {
      "epoch": 2.774193548387097,
      "grad_norm": 0.14762410521507263,
      "learning_rate": 3.9231555880201655e-06,
      "loss": 0.0022,
      "step": 202
    },
    {
      "epoch": 2.7880184331797233,
      "grad_norm": 0.16235944628715515,
      "learning_rate": 3.903642770903671e-06,
      "loss": 0.0012,
      "step": 203
    },
    {
      "epoch": 2.80184331797235,
      "grad_norm": 0.1506718099117279,
      "learning_rate": 3.884004270673711e-06,
      "loss": 0.0015,
      "step": 204
    },
    {
      "epoch": 2.815668202764977,
      "grad_norm": 0.10484135895967484,
      "learning_rate": 3.864241845774746e-06,
      "loss": 0.0016,
      "step": 205
    },
    {
      "epoch": 2.8294930875576036,
      "grad_norm": 0.7636306285858154,
      "learning_rate": 3.844357265747531e-06,
      "loss": 0.0018,
      "step": 206
    },
    {
      "epoch": 2.8433179723502304,
      "grad_norm": 0.2242082804441452,
      "learning_rate": 3.8243523110706736e-06,
      "loss": 0.0021,
      "step": 207
    },
    {
      "epoch": 2.857142857142857,
      "grad_norm": 0.3264133334159851,
      "learning_rate": 3.8042287730012117e-06,
      "loss": 0.0021,
      "step": 208
    },
    {
      "epoch": 2.870967741935484,
      "grad_norm": 0.12472204118967056,
      "learning_rate": 3.7839884534142157e-06,
      "loss": 0.0011,
      "step": 209
    },
    {
      "epoch": 2.8847926267281108,
      "grad_norm": 0.07526414096355438,
      "learning_rate": 3.7636331646414524e-06,
      "loss": 0.0017,
      "step": 210
    },
    {
      "epoch": 2.8986175115207375,
      "grad_norm": 0.16134843230247498,
      "learning_rate": 3.7431647293091076e-06,
      "loss": 0.0019,
      "step": 211
    },
    {
      "epoch": 2.912442396313364,
      "grad_norm": 0.14789307117462158,
      "learning_rate": 3.7225849801745835e-06,
      "loss": 0.0012,
      "step": 212
    },
    {
      "epoch": 2.9262672811059907,
      "grad_norm": 0.13681238889694214,
      "learning_rate": 3.701895759962397e-06,
      "loss": 0.0011,
      "step": 213
    },
    {
      "epoch": 2.9400921658986174,
      "grad_norm": 0.10747735947370529,
      "learning_rate": 3.6810989211991777e-06,
      "loss": 0.0007,
      "step": 214
    },
    {
      "epoch": 2.953917050691244,
      "grad_norm": 0.08121375739574432,
      "learning_rate": 3.6601963260477923e-06,
      "loss": 0.0005,
      "step": 215
    },
    {
      "epoch": 2.967741935483871,
      "grad_norm": 0.0884300246834755,
      "learning_rate": 3.6391898461406045e-06,
      "loss": 0.0014,
      "step": 216
    },
    {
      "epoch": 2.9815668202764978,
      "grad_norm": 0.18539245426654816,
      "learning_rate": 3.6180813624118898e-06,
      "loss": 0.002,
      "step": 217
    },
    {
      "epoch": 2.9953917050691246,
      "grad_norm": 0.1257522702217102,
      "learning_rate": 3.5968727649294134e-06,
      "loss": 0.0015,
      "step": 218
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.2422555685043335,
      "learning_rate": 3.575565952725193e-06,
      "loss": 0.0002,
      "step": 219
    },
    {
      "epoch": 3.013824884792627,
      "grad_norm": 0.06009506434202194,
      "learning_rate": 3.55416283362546e-06,
      "loss": 0.0003,
      "step": 220
    },
    {
      "epoch": 3.0276497695852536,
      "grad_norm": 0.0876953974366188,
      "learning_rate": 3.5326653240798283e-06,
      "loss": 0.0005,
      "step": 221
    },
    {
      "epoch": 3.0414746543778803,
      "grad_norm": 0.7512914538383484,
      "learning_rate": 3.5110753489896924e-06,
      "loss": 0.0007,
      "step": 222
    },
    {
      "epoch": 3.055299539170507,
      "grad_norm": 0.08451899141073227,
      "learning_rate": 3.4893948415358803e-06,
      "loss": 0.0009,
      "step": 223
    },
    {
      "epoch": 3.0691244239631335,
      "grad_norm": 0.15445305407047272,
      "learning_rate": 3.4676257430055438e-06,
      "loss": 0.0006,
      "step": 224
    },
    {
      "epoch": 3.0829493087557602,
      "grad_norm": 0.07909094542264938,
      "learning_rate": 3.4457700026183378e-06,
      "loss": 0.0004,
      "step": 225
    },
    {
      "epoch": 3.096774193548387,
      "grad_norm": 0.03637247905135155,
      "learning_rate": 3.4238295773518924e-06,
      "loss": 0.0003,
      "step": 226
    },
    {
      "epoch": 3.110599078341014,
      "grad_norm": 0.203308567404747,
      "learning_rate": 3.4018064317665745e-06,
      "loss": 0.0003,
      "step": 227
    },
    {
      "epoch": 3.1244239631336406,
      "grad_norm": 0.03239201754331589,
      "learning_rate": 3.3797025378295826e-06,
      "loss": 0.0002,
      "step": 228
    },
    {
      "epoch": 3.1382488479262673,
      "grad_norm": 0.07106538861989975,
      "learning_rate": 3.357519874738382e-06,
      "loss": 0.0004,
      "step": 229
    },
    {
      "epoch": 3.152073732718894,
      "grad_norm": 0.048268985003232956,
      "learning_rate": 3.3352604287434752e-06,
      "loss": 0.0003,
      "step": 230
    },
    {
      "epoch": 3.165898617511521,
      "grad_norm": 0.0841558575630188,
      "learning_rate": 3.31292619297056e-06,
      "loss": 0.0003,
      "step": 231
    },
    {
      "epoch": 3.1797235023041477,
      "grad_norm": 0.07029678672552109,
      "learning_rate": 3.29051916724206e-06,
      "loss": 0.0003,
      "step": 232
    },
    {
      "epoch": 3.193548387096774,
      "grad_norm": 0.11369964480400085,
      "learning_rate": 3.2680413578980623e-06,
      "loss": 0.0014,
      "step": 233
    },
    {
      "epoch": 3.207373271889401,
      "grad_norm": 0.0367964468896389,
      "learning_rate": 3.245494777616664e-06,
      "loss": 0.0001,
      "step": 234
    },
    {
      "epoch": 3.2211981566820276,
      "grad_norm": 0.13746097683906555,
      "learning_rate": 3.2228814452337587e-06,
      "loss": 0.0003,
      "step": 235
    },
    {
      "epoch": 3.2350230414746544,
      "grad_norm": 0.09046189486980438,
      "learning_rate": 3.2002033855622683e-06,
      "loss": 0.0004,
      "step": 236
    },
    {
      "epoch": 3.248847926267281,
      "grad_norm": 0.04587667062878609,
      "learning_rate": 3.177462629210838e-06,
      "loss": 0.0002,
      "step": 237
    },
    {
      "epoch": 3.262672811059908,
      "grad_norm": 0.11323168128728867,
      "learning_rate": 3.154661212402017e-06,
      "loss": 0.0003,
      "step": 238
    },
    {
      "epoch": 3.2764976958525347,
      "grad_norm": 0.04728177189826965,
      "learning_rate": 3.131801176789934e-06,
      "loss": 0.0002,
      "step": 239
    },
    {
      "epoch": 3.2903225806451615,
      "grad_norm": 0.527999997138977,
      "learning_rate": 3.1088845692774798e-06,
      "loss": 0.0008,
      "step": 240
    },
    {
      "epoch": 3.3041474654377883,
      "grad_norm": 0.026646027341485023,
      "learning_rate": 3.0859134418330373e-06,
      "loss": 0.0001,
      "step": 241
    },
    {
      "epoch": 3.3179723502304146,
      "grad_norm": 0.057450197637081146,
      "learning_rate": 3.0628898513067357e-06,
      "loss": 0.0004,
      "step": 242
    },
    {
      "epoch": 3.3317972350230414,
      "grad_norm": 0.08258494734764099,
      "learning_rate": 3.0398158592462847e-06,
      "loss": 0.0005,
      "step": 243
    },
    {
      "epoch": 3.345622119815668,
      "grad_norm": 0.01878846250474453,
      "learning_rate": 3.0166935317123824e-06,
      "loss": 0.0001,
      "step": 244
    },
    {
      "epoch": 3.359447004608295,
      "grad_norm": 0.041918545961380005,
      "learning_rate": 2.9935249390937184e-06,
      "loss": 0.0002,
      "step": 245
    },
    {
      "epoch": 3.3732718894009217,
      "grad_norm": 0.04018491134047508,
      "learning_rate": 2.970312155921585e-06,
      "loss": 0.0002,
      "step": 246
    },
    {
      "epoch": 3.3870967741935485,
      "grad_norm": 0.040825020521879196,
      "learning_rate": 2.9470572606841295e-06,
      "loss": 0.0002,
      "step": 247
    },
    {
      "epoch": 3.4009216589861753,
      "grad_norm": 0.050590481609106064,
      "learning_rate": 2.9237623356402423e-06,
      "loss": 0.0002,
      "step": 248
    },
    {
      "epoch": 3.4147465437788016,
      "grad_norm": 0.07999978959560394,
      "learning_rate": 2.900429466633107e-06,
      "loss": 0.0002,
      "step": 249
    },
    {
      "epoch": 3.4285714285714284,
      "grad_norm": 0.02137935161590576,
      "learning_rate": 2.8770607429034352e-06,
      "loss": 0.0001,
      "step": 250
    },
    {
      "epoch": 3.442396313364055,
      "grad_norm": 0.18967340886592865,
      "learning_rate": 2.8536582569023964e-06,
      "loss": 0.0007,
      "step": 251
    },
    {
      "epoch": 3.456221198156682,
      "grad_norm": 0.03681226819753647,
      "learning_rate": 2.8302241041042564e-06,
      "loss": 0.0001,
      "step": 252
    },
    {
      "epoch": 3.4700460829493087,
      "grad_norm": 0.03142761439085007,
      "learning_rate": 2.8067603828187446e-06,
      "loss": 0.0001,
      "step": 253
    },
    {
      "epoch": 3.4838709677419355,
      "grad_norm": 0.11318890005350113,
      "learning_rate": 2.7832691940031755e-06,
      "loss": 0.0005,
      "step": 254
    },
    {
      "epoch": 3.4976958525345623,
      "grad_norm": 0.047176819294691086,
      "learning_rate": 2.759752641074322e-06,
      "loss": 0.0002,
      "step": 255
    },
    {
      "epoch": 3.511520737327189,
      "grad_norm": 0.0642286092042923,
      "learning_rate": 2.7362128297200784e-06,
      "loss": 0.0002,
      "step": 256
    },
    {
      "epoch": 3.525345622119816,
      "grad_norm": 0.09328105300664902,
      "learning_rate": 2.712651867710914e-06,
      "loss": 0.0004,
      "step": 257
    },
    {
      "epoch": 3.539170506912442,
      "grad_norm": 0.08150269836187363,
      "learning_rate": 2.6890718647111424e-06,
      "loss": 0.0007,
      "step": 258
    },
    {
      "epoch": 3.5529953917050694,
      "grad_norm": 0.03366294875741005,
      "learning_rate": 2.665474932090017e-06,
      "loss": 0.0001,
      "step": 259
    },
    {
      "epoch": 3.5668202764976957,
      "grad_norm": 0.032316725701093674,
      "learning_rate": 2.6418631827326857e-06,
      "loss": 0.0001,
      "step": 260
    },
    {
      "epoch": 3.5806451612903225,
      "grad_norm": 0.02776617370545864,
      "learning_rate": 2.6182387308509927e-06,
      "loss": 0.0001,
      "step": 261
    },
    {
      "epoch": 3.5944700460829493,
      "grad_norm": 0.1258484572172165,
      "learning_rate": 2.5946036917941765e-06,
      "loss": 0.0003,
      "step": 262
    },
    {
      "epoch": 3.608294930875576,
      "grad_norm": 0.04412033408880234,
      "learning_rate": 2.570960181859458e-06,
      "loss": 0.0003,
      "step": 263
    },
    {
      "epoch": 3.622119815668203,
      "grad_norm": 0.016816483810544014,
      "learning_rate": 2.547310318102548e-06,
      "loss": 0.0001,
      "step": 264
    },
    {
      "epoch": 3.6359447004608296,
      "grad_norm": 0.028503524139523506,
      "learning_rate": 2.5236562181480794e-06,
      "loss": 0.0001,
      "step": 265
    },
    {
      "epoch": 3.6497695852534564,
      "grad_norm": 0.03991785645484924,
      "learning_rate": 2.5e-06,
      "loss": 0.0002,
      "step": 266
    },
    {
      "epoch": 3.6635944700460827,
      "grad_norm": 0.07638856768608093,
      "learning_rate": 2.4763437818519205e-06,
      "loss": 0.0002,
      "step": 267
    },
    {
      "epoch": 3.6774193548387095,
      "grad_norm": 0.032387226819992065,
      "learning_rate": 2.4526896818974534e-06,
      "loss": 0.0002,
      "step": 268
    },
    {
      "epoch": 3.6912442396313363,
      "grad_norm": 0.035975128412246704,
      "learning_rate": 2.429039818140543e-06,
      "loss": 0.0002,
      "step": 269
    },
    {
      "epoch": 3.705069124423963,
      "grad_norm": 0.021173926070332527,
      "learning_rate": 2.405396308205825e-06,
      "loss": 0.0001,
      "step": 270
    },
    {
      "epoch": 3.71889400921659,
      "grad_norm": 0.005446314811706543,
      "learning_rate": 2.381761269149009e-06,
      "loss": 0.0,
      "step": 271
    },
    {
      "epoch": 3.7327188940092166,
      "grad_norm": 0.04019308090209961,
      "learning_rate": 2.358136817267315e-06,
      "loss": 0.0001,
      "step": 272
    },
    {
      "epoch": 3.7465437788018434,
      "grad_norm": 0.0222685057669878,
      "learning_rate": 2.334525067909983e-06,
      "loss": 0.0001,
      "step": 273
    },
    {
      "epoch": 3.76036866359447,
      "grad_norm": 0.02486710622906685,
      "learning_rate": 2.3109281352888593e-06,
      "loss": 0.0001,
      "step": 274
    },
    {
      "epoch": 3.774193548387097,
      "grad_norm": 0.01929207146167755,
      "learning_rate": 2.2873481322890866e-06,
      "loss": 0.0001,
      "step": 275
    },
    {
      "epoch": 3.7880184331797233,
      "grad_norm": 0.010686581023037434,
      "learning_rate": 2.263787170279922e-06,
      "loss": 0.0,
      "step": 276
    },
    {
      "epoch": 3.80184331797235,
      "grad_norm": 0.04710806906223297,
      "learning_rate": 2.2402473589256793e-06,
      "loss": 0.0001,
      "step": 277
    },
    {
      "epoch": 3.815668202764977,
      "grad_norm": 0.00774085009470582,
      "learning_rate": 2.2167308059968258e-06,
      "loss": 0.0,
      "step": 278
    },
    {
      "epoch": 3.8294930875576036,
      "grad_norm": 0.00735470512881875,
      "learning_rate": 2.193239617181256e-06,
      "loss": 0.0,
      "step": 279
    },
    {
      "epoch": 3.8433179723502304,
      "grad_norm": 0.005572167690843344,
      "learning_rate": 2.169775895895745e-06,
      "loss": 0.0,
      "step": 280
    },
    {
      "epoch": 3.857142857142857,
      "grad_norm": 0.07026448100805283,
      "learning_rate": 2.146341743097604e-06,
      "loss": 0.0004,
      "step": 281
    },
    {
      "epoch": 3.870967741935484,
      "grad_norm": 0.03968067839741707,
      "learning_rate": 2.1229392570965656e-06,
      "loss": 0.0001,
      "step": 282
    },
    {
      "epoch": 3.8847926267281108,
      "grad_norm": 0.002730958629399538,
      "learning_rate": 2.0995705333668948e-06,
      "loss": 0.0,
      "step": 283
    },
    {
      "epoch": 3.8986175115207375,
      "grad_norm": 0.010703709907829762,
      "learning_rate": 2.0762376643597586e-06,
      "loss": 0.0,
      "step": 284
    },
    {
      "epoch": 3.912442396313364,
      "grad_norm": 0.03527766093611717,
      "learning_rate": 2.0529427393158704e-06,
      "loss": 0.0001,
      "step": 285
    },
    {
      "epoch": 3.9262672811059907,
      "grad_norm": 0.03926033526659012,
      "learning_rate": 2.0296878440784164e-06,
      "loss": 0.0001,
      "step": 286
    },
    {
      "epoch": 3.9400921658986174,
      "grad_norm": 0.007335775997489691,
      "learning_rate": 2.006475060906283e-06,
      "loss": 0.0,
      "step": 287
    },
    {
      "epoch": 3.953917050691244,
      "grad_norm": 0.005718631204217672,
      "learning_rate": 1.9833064682876175e-06,
      "loss": 0.0,
      "step": 288
    },
    {
      "epoch": 3.967741935483871,
      "grad_norm": 0.005941327195614576,
      "learning_rate": 1.9601841407537157e-06,
      "loss": 0.0,
      "step": 289
    },
    {
      "epoch": 3.9815668202764978,
      "grad_norm": 0.039281055331230164,
      "learning_rate": 1.937110148693265e-06,
      "loss": 0.0001,
      "step": 290
    },
    {
      "epoch": 3.9953917050691246,
      "grad_norm": 0.06976872682571411,
      "learning_rate": 1.9140865581669627e-06,
      "loss": 0.0001,
      "step": 291
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.06976872682571411,
      "learning_rate": 1.8911154307225204e-06,
      "loss": 0.0,
      "step": 292
    },
    {
      "epoch": 4.013824884792626,
      "grad_norm": 0.005908307619392872,
      "learning_rate": 1.8681988232100674e-06,
      "loss": 0.0,
      "step": 293
    },
    {
      "epoch": 4.027649769585254,
      "grad_norm": 0.022486088797450066,
      "learning_rate": 1.8453387875979834e-06,
      "loss": 0.0001,
      "step": 294
    },
    {
      "epoch": 4.04147465437788,
      "grad_norm": 0.0074249873869121075,
      "learning_rate": 1.822537370789163e-06,
      "loss": 0.0,
      "step": 295
    },
    {
      "epoch": 4.055299539170507,
      "grad_norm": 0.004768090322613716,
      "learning_rate": 1.7997966144377328e-06,
      "loss": 0.0,
      "step": 296
    },
    {
      "epoch": 4.0691244239631335,
      "grad_norm": 0.013053408823907375,
      "learning_rate": 1.7771185547662417e-06,
      "loss": 0.0,
      "step": 297
    },
    {
      "epoch": 4.082949308755761,
      "grad_norm": 0.00568437110632658,
      "learning_rate": 1.754505222383337e-06,
      "loss": 0.0,
      "step": 298
    },
    {
      "epoch": 4.096774193548387,
      "grad_norm": 0.006704168394207954,
      "learning_rate": 1.7319586421019383e-06,
      "loss": 0.0,
      "step": 299
    },
    {
      "epoch": 4.110599078341014,
      "grad_norm": 0.0039120810106396675,
      "learning_rate": 1.7094808327579401e-06,
      "loss": 0.0,
      "step": 300
    },
    {
      "epoch": 4.124423963133641,
      "grad_norm": 0.009206798858940601,
      "learning_rate": 1.6870738070294412e-06,
      "loss": 0.0,
      "step": 301
    },
    {
      "epoch": 4.138248847926267,
      "grad_norm": 0.005304583813995123,
      "learning_rate": 1.6647395712565256e-06,
      "loss": 0.0,
      "step": 302
    },
    {
      "epoch": 4.152073732718894,
      "grad_norm": 0.008103611879050732,
      "learning_rate": 1.6424801252616186e-06,
      "loss": 0.0001,
      "step": 303
    },
    {
      "epoch": 4.1658986175115205,
      "grad_norm": 0.028891608119010925,
      "learning_rate": 1.6202974621704176e-06,
      "loss": 0.0,
      "step": 304
    },
    {
      "epoch": 4.179723502304148,
      "grad_norm": 0.0035763406194746494,
      "learning_rate": 1.5981935682334266e-06,
      "loss": 0.0,
      "step": 305
    },
    {
      "epoch": 4.193548387096774,
      "grad_norm": 0.009718772955238819,
      "learning_rate": 1.5761704226481078e-06,
      "loss": 0.0,
      "step": 306
    },
    {
      "epoch": 4.207373271889401,
      "grad_norm": 0.01045698020607233,
      "learning_rate": 1.5542299973816626e-06,
      "loss": 0.0,
      "step": 307
    },
    {
      "epoch": 4.221198156682028,
      "grad_norm": 0.004575685132294893,
      "learning_rate": 1.5323742569944573e-06,
      "loss": 0.0,
      "step": 308
    },
    {
      "epoch": 4.235023041474655,
      "grad_norm": 0.003245371161028743,
      "learning_rate": 1.5106051584641208e-06,
      "loss": 0.0,
      "step": 309
    },
    {
      "epoch": 4.248847926267281,
      "grad_norm": 0.005619620904326439,
      "learning_rate": 1.4889246510103078e-06,
      "loss": 0.0,
      "step": 310
    },
    {
      "epoch": 4.2626728110599075,
      "grad_norm": 0.004715710878372192,
      "learning_rate": 1.4673346759201728e-06,
      "loss": 0.0,
      "step": 311
    },
    {
      "epoch": 4.276497695852535,
      "grad_norm": 0.007476332131773233,
      "learning_rate": 1.44583716637454e-06,
      "loss": 0.0,
      "step": 312
    },
    {
      "epoch": 4.290322580645161,
      "grad_norm": 0.01739400625228882,
      "learning_rate": 1.4244340472748076e-06,
      "loss": 0.0001,
      "step": 313
    },
    {
      "epoch": 4.304147465437788,
      "grad_norm": 0.00816753227263689,
      "learning_rate": 1.403127235070587e-06,
      "loss": 0.0,
      "step": 314
    },
    {
      "epoch": 4.317972350230415,
      "grad_norm": 0.010216044262051582,
      "learning_rate": 1.381918637588112e-06,
      "loss": 0.0,
      "step": 315
    },
    {
      "epoch": 4.331797235023042,
      "grad_norm": 0.004990486893802881,
      "learning_rate": 1.3608101538593965e-06,
      "loss": 0.0,
      "step": 316
    },
    {
      "epoch": 4.345622119815668,
      "grad_norm": 0.004758649505674839,
      "learning_rate": 1.3398036739522088e-06,
      "loss": 0.0001,
      "step": 317
    },
    {
      "epoch": 4.359447004608295,
      "grad_norm": 0.041808340698480606,
      "learning_rate": 1.3189010788008234e-06,
      "loss": 0.0,
      "step": 318
    },
    {
      "epoch": 4.373271889400922,
      "grad_norm": 0.012711254879832268,
      "learning_rate": 1.2981042400376032e-06,
      "loss": 0.0,
      "step": 319
    },
    {
      "epoch": 4.387096774193548,
      "grad_norm": 0.0035697701387107372,
      "learning_rate": 1.277415019825417e-06,
      "loss": 0.0,
      "step": 320
    },
    {
      "epoch": 4.400921658986175,
      "grad_norm": 0.005487007088959217,
      "learning_rate": 1.2568352706908937e-06,
      "loss": 0.0,
      "step": 321
    },
    {
      "epoch": 4.414746543778802,
      "grad_norm": 0.01304635126143694,
      "learning_rate": 1.2363668353585486e-06,
      "loss": 0.0,
      "step": 322
    },
    {
      "epoch": 4.428571428571429,
      "grad_norm": 0.0019787494093179703,
      "learning_rate": 1.216011546585785e-06,
      "loss": 0.0,
      "step": 323
    },
    {
      "epoch": 4.442396313364055,
      "grad_norm": 0.00808583851903677,
      "learning_rate": 1.195771226998789e-06,
      "loss": 0.0,
      "step": 324
    },
    {
      "epoch": 4.456221198156682,
      "grad_norm": 0.0022094689775258303,
      "learning_rate": 1.1756476889293269e-06,
      "loss": 0.0,
      "step": 325
    },
    {
      "epoch": 4.470046082949309,
      "grad_norm": 0.012792594730854034,
      "learning_rate": 1.1556427342524698e-06,
      "loss": 0.0,
      "step": 326
    },
    {
      "epoch": 4.483870967741936,
      "grad_norm": 0.006805351935327053,
      "learning_rate": 1.1357581542252555e-06,
      "loss": 0.0,
      "step": 327
    },
    {
      "epoch": 4.497695852534562,
      "grad_norm": 0.003740285988897085,
      "learning_rate": 1.1159957293262888e-06,
      "loss": 0.0,
      "step": 328
    },
    {
      "epoch": 4.511520737327189,
      "grad_norm": 0.009705561213195324,
      "learning_rate": 1.0963572290963298e-06,
      "loss": 0.0,
      "step": 329
    },
    {
      "epoch": 4.525345622119816,
      "grad_norm": 0.040002401918172836,
      "learning_rate": 1.0768444119798357e-06,
      "loss": 0.0002,
      "step": 330
    },
    {
      "epoch": 4.539170506912442,
      "grad_norm": 0.0036789420992136,
      "learning_rate": 1.0574590251675145e-06,
      "loss": 0.0,
      "step": 331
    },
    {
      "epoch": 4.552995391705069,
      "grad_norm": 0.004043960478156805,
      "learning_rate": 1.0382028044398823e-06,
      "loss": 0.0002,
      "step": 332
    },
    {
      "epoch": 4.566820276497696,
      "grad_norm": 0.0512581467628479,
      "learning_rate": 1.0190774740118343e-06,
      "loss": 0.0,
      "step": 333
    },
    {
      "epoch": 4.580645161290323,
      "grad_norm": 0.004926969762891531,
      "learning_rate": 1.0000847463782615e-06,
      "loss": 0.0,
      "step": 334
    },
    {
      "epoch": 4.594470046082949,
      "grad_norm": 0.0043294900096952915,
      "learning_rate": 9.812263221607114e-07,
      "loss": 0.0,
      "step": 335
    },
    {
      "epoch": 4.6082949308755765,
      "grad_norm": 0.0023195091634988785,
      "learning_rate": 9.625038899551162e-07,
      "loss": 0.0,
      "step": 336
    },
    {
      "epoch": 4.622119815668203,
      "grad_norm": 0.0015059575671330094,
      "learning_rate": 9.439191261805894e-07,
      "loss": 0.0,
      "step": 337
    },
    {
      "epoch": 4.635944700460829,
      "grad_norm": 0.001368862227536738,
      "learning_rate": 9.254736949293216e-07,
      "loss": 0.0,
      "step": 338
    },
    {
      "epoch": 4.649769585253456,
      "grad_norm": 0.008128674700856209,
      "learning_rate": 9.07169247817579e-07,
      "loss": 0.0,
      "step": 339
    },
    {
      "epoch": 4.663594470046083,
      "grad_norm": 0.0029226879123598337,
      "learning_rate": 8.890074238378074e-07,
      "loss": 0.0,
      "step": 340
    },
    {
      "epoch": 4.67741935483871,
      "grad_norm": 0.0012331035686656833,
      "learning_rate": 8.709898492118885e-07,
      "loss": 0.0,
      "step": 341
    },
    {
      "epoch": 4.691244239631336,
      "grad_norm": 0.005286338273435831,
      "learning_rate": 8.531181372455161e-07,
      "loss": 0.0,
      "step": 342
    },
    {
      "epoch": 4.705069124423963,
      "grad_norm": 0.0026836844626814127,
      "learning_rate": 8.353938881837445e-07,
      "loss": 0.0,
      "step": 343
    },
    {
      "epoch": 4.71889400921659,
      "grad_norm": 0.013100259937345982,
      "learning_rate": 8.178186890677029e-07,
      "loss": 0.0,
      "step": 344
    },
    {
      "epoch": 4.732718894009217,
      "grad_norm": 0.005650435108691454,
      "learning_rate": 8.003941135924859e-07,
      "loss": 0.0,
      "step": 345
    },
    {
      "epoch": 4.746543778801843,
      "grad_norm": 0.007480297237634659,
      "learning_rate": 7.83121721966248e-07,
      "loss": 0.0,
      "step": 346
    },
    {
      "epoch": 4.76036866359447,
      "grad_norm": 0.014115474186837673,
      "learning_rate": 7.66003060770498e-07,
      "loss": 0.0,
      "step": 347
    },
    {
      "epoch": 4.774193548387097,
      "grad_norm": 0.0011564996093511581,
      "learning_rate": 7.490396628216237e-07,
      "loss": 0.0,
      "step": 348
    },
    {
      "epoch": 4.788018433179723,
      "grad_norm": 0.01101834885776043,
      "learning_rate": 7.322330470336314e-07,
      "loss": 0.0,
      "step": 349
    },
    {
      "epoch": 4.8018433179723505,
      "grad_norm": 0.003535608062520623,
      "learning_rate": 7.155847182821524e-07,
      "loss": 0.0,
      "step": 350
    },
    {
      "epoch": 4.815668202764977,
      "grad_norm": 0.03185940906405449,
      "learning_rate": 6.990961672696908e-07,
      "loss": 0.0001,
      "step": 351
    },
    {
      "epoch": 4.829493087557603,
      "grad_norm": 0.006516721565276384,
      "learning_rate": 6.827688703921407e-07,
      "loss": 0.0,
      "step": 352
    },
    {
      "epoch": 4.84331797235023,
      "grad_norm": 0.008277276530861855,
      "learning_rate": 6.666042896065983e-07,
      "loss": 0.0,
      "step": 353
    },
    {
      "epoch": 4.857142857142857,
      "grad_norm": 0.00266360049135983,
      "learning_rate": 6.506038723004484e-07,
      "loss": 0.0,
      "step": 354
    },
    {
      "epoch": 4.870967741935484,
      "grad_norm": 0.01671386882662773,
      "learning_rate": 6.347690511617693e-07,
      "loss": 0.0,
      "step": 355
    },
    {
      "epoch": 4.88479262672811,
      "grad_norm": 0.013981528580188751,
      "learning_rate": 6.191012440510469e-07,
      "loss": 0.0001,
      "step": 356
    },
    {
      "epoch": 4.8986175115207375,
      "grad_norm": 0.02350999414920807,
      "learning_rate": 6.036018538742208e-07,
      "loss": 0.0,
      "step": 357
    },
    {
      "epoch": 4.912442396313364,
      "grad_norm": 0.004093356896191835,
      "learning_rate": 5.882722684570638e-07,
      "loss": 0.0,
      "step": 358
    },
    {
      "epoch": 4.926267281105991,
      "grad_norm": 0.00656296918168664,
      "learning_rate": 5.731138604209169e-07,
      "loss": 0.0,
      "step": 359
    },
    {
      "epoch": 4.940092165898617,
      "grad_norm": 0.002148544415831566,
      "learning_rate": 5.581279870597866e-07,
      "loss": 0.0,
      "step": 360
    }
  ],
  "logging_steps": 1,
  "max_steps": 432,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 6,
  "save_steps": 72,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.298790820660537e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}