{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.9812889812889813,
  "eval_steps": 500,
  "global_step": 360,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.008316008316008316,
      "grad_norm": 8.161552429199219,
      "learning_rate": 0.0,
      "loss": 4.8011,
      "step": 1
    },
    {
      "epoch": 0.016632016632016633,
      "grad_norm": 8.788352966308594,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 4.8042,
      "step": 2
    },
    {
      "epoch": 0.02494802494802495,
      "grad_norm": 7.977554798126221,
      "learning_rate": 4.000000000000001e-06,
      "loss": 4.8113,
      "step": 3
    },
    {
      "epoch": 0.033264033264033266,
      "grad_norm": 8.319157600402832,
      "learning_rate": 6e-06,
      "loss": 4.8424,
      "step": 4
    },
    {
      "epoch": 0.04158004158004158,
      "grad_norm": 6.344866752624512,
      "learning_rate": 8.000000000000001e-06,
      "loss": 4.7736,
      "step": 5
    },
    {
      "epoch": 0.0498960498960499,
      "grad_norm": 5.9034647941589355,
      "learning_rate": 1e-05,
      "loss": 4.7142,
      "step": 6
    },
    {
      "epoch": 0.058212058212058215,
      "grad_norm": 5.377430438995361,
      "learning_rate": 1.2e-05,
      "loss": 4.6148,
      "step": 7
    },
    {
      "epoch": 0.06652806652806653,
      "grad_norm": 6.9711222648620605,
      "learning_rate": 1.4000000000000001e-05,
      "loss": 4.5419,
      "step": 8
    },
    {
      "epoch": 0.07484407484407485,
      "grad_norm": 6.142094612121582,
      "learning_rate": 1.6000000000000003e-05,
      "loss": 4.4492,
      "step": 9
    },
    {
      "epoch": 0.08316008316008316,
      "grad_norm": 4.155233383178711,
      "learning_rate": 1.8e-05,
      "loss": 4.2954,
      "step": 10
    },
    {
      "epoch": 0.09147609147609148,
      "grad_norm": 3.6919031143188477,
      "learning_rate": 2e-05,
      "loss": 4.208,
      "step": 11
    },
    {
      "epoch": 0.0997920997920998,
      "grad_norm": 3.5758368968963623,
      "learning_rate": 2.2000000000000003e-05,
      "loss": 4.0939,
      "step": 12
    },
    {
      "epoch": 0.10810810810810811,
      "grad_norm": 2.7365787029266357,
      "learning_rate": 2.4e-05,
      "loss": 4.0046,
      "step": 13
    },
    {
      "epoch": 0.11642411642411643,
      "grad_norm": 2.0502889156341553,
      "learning_rate": 2.6000000000000002e-05,
      "loss": 3.9557,
      "step": 14
    },
    {
      "epoch": 0.12474012474012475,
      "grad_norm": 2.2007663249969482,
      "learning_rate": 2.8000000000000003e-05,
      "loss": 3.8579,
      "step": 15
    },
    {
      "epoch": 0.13305613305613306,
      "grad_norm": 2.462459087371826,
      "learning_rate": 3e-05,
      "loss": 3.8167,
      "step": 16
    },
    {
      "epoch": 0.14137214137214138,
      "grad_norm": 2.7364518642425537,
      "learning_rate": 3.2000000000000005e-05,
      "loss": 3.8004,
      "step": 17
    },
    {
      "epoch": 0.1496881496881497,
      "grad_norm": 2.85400128364563,
      "learning_rate": 3.4000000000000007e-05,
      "loss": 3.6991,
      "step": 18
    },
    {
      "epoch": 0.158004158004158,
      "grad_norm": 2.499318838119507,
      "learning_rate": 3.6e-05,
      "loss": 3.6456,
      "step": 19
    },
    {
      "epoch": 0.16632016632016633,
      "grad_norm": 2.214195728302002,
      "learning_rate": 3.8e-05,
      "loss": 3.6195,
      "step": 20
    },
    {
      "epoch": 0.17463617463617465,
      "grad_norm": 2.195711135864258,
      "learning_rate": 4e-05,
      "loss": 3.5519,
      "step": 21
    },
    {
      "epoch": 0.18295218295218296,
      "grad_norm": 1.992714762687683,
      "learning_rate": 4.2e-05,
      "loss": 3.5034,
      "step": 22
    },
    {
      "epoch": 0.19126819126819128,
      "grad_norm": 2.6501753330230713,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 3.4722,
      "step": 23
    },
    {
      "epoch": 0.1995841995841996,
      "grad_norm": 2.2270214557647705,
      "learning_rate": 4.600000000000001e-05,
      "loss": 3.4126,
      "step": 24
    },
    {
      "epoch": 0.2079002079002079,
      "grad_norm": 3.029968023300171,
      "learning_rate": 4.8e-05,
      "loss": 3.3517,
      "step": 25
    },
    {
      "epoch": 0.21621621621621623,
      "grad_norm": 1.7878334522247314,
      "learning_rate": 5e-05,
      "loss": 3.3321,
      "step": 26
    },
    {
      "epoch": 0.22453222453222454,
      "grad_norm": 1.6116269826889038,
      "learning_rate": 5.2000000000000004e-05,
      "loss": 3.2636,
      "step": 27
    },
    {
      "epoch": 0.23284823284823286,
      "grad_norm": 2.6376936435699463,
      "learning_rate": 5.4000000000000005e-05,
      "loss": 3.2541,
      "step": 28
    },
    {
      "epoch": 0.24116424116424118,
      "grad_norm": 10.313434600830078,
      "learning_rate": 5.6000000000000006e-05,
      "loss": 3.2762,
      "step": 29
    },
    {
      "epoch": 0.2494802494802495,
      "grad_norm": 5.4037251472473145,
      "learning_rate": 5.8e-05,
      "loss": 3.2097,
      "step": 30
    },
    {
      "epoch": 0.2577962577962578,
      "grad_norm": 15.105055809020996,
      "learning_rate": 6e-05,
      "loss": 3.2342,
      "step": 31
    },
    {
      "epoch": 0.2661122661122661,
      "grad_norm": 15.099345207214355,
      "learning_rate": 6.2e-05,
      "loss": 3.2316,
      "step": 32
    },
    {
      "epoch": 0.27442827442827444,
      "grad_norm": 1.7807673215866089,
      "learning_rate": 6.400000000000001e-05,
      "loss": 3.1318,
      "step": 33
    },
    {
      "epoch": 0.28274428274428276,
      "grad_norm": 7.983642101287842,
      "learning_rate": 6.6e-05,
      "loss": 3.1442,
      "step": 34
    },
    {
      "epoch": 0.2910602910602911,
      "grad_norm": 4.913079738616943,
      "learning_rate": 6.800000000000001e-05,
      "loss": 3.1082,
      "step": 35
    },
    {
      "epoch": 0.2993762993762994,
      "grad_norm": 12.353167533874512,
      "learning_rate": 7e-05,
      "loss": 3.1656,
      "step": 36
    },
    {
      "epoch": 0.3076923076923077,
      "grad_norm": 10.911690711975098,
      "learning_rate": 7.2e-05,
      "loss": 3.1361,
      "step": 37
    },
    {
      "epoch": 0.316008316008316,
      "grad_norm": 1.3475183248519897,
      "learning_rate": 7.4e-05,
      "loss": 3.057,
      "step": 38
    },
    {
      "epoch": 0.32432432432432434,
      "grad_norm": 7.743471145629883,
      "learning_rate": 7.6e-05,
      "loss": 3.0549,
      "step": 39
    },
    {
      "epoch": 0.33264033264033266,
      "grad_norm": 4.499805450439453,
      "learning_rate": 7.800000000000001e-05,
      "loss": 2.9976,
      "step": 40
    },
    {
      "epoch": 0.340956340956341,
      "grad_norm": 4.716672420501709,
      "learning_rate": 8e-05,
      "loss": 2.9922,
      "step": 41
    },
    {
      "epoch": 0.3492723492723493,
      "grad_norm": 5.240478038787842,
      "learning_rate": 8.2e-05,
      "loss": 3.0115,
      "step": 42
    },
    {
      "epoch": 0.3575883575883576,
      "grad_norm": 1.5897458791732788,
      "learning_rate": 8.4e-05,
      "loss": 2.9899,
      "step": 43
    },
    {
      "epoch": 0.3659043659043659,
      "grad_norm": 2.463665723800659,
      "learning_rate": 8.6e-05,
      "loss": 2.9799,
      "step": 44
    },
    {
      "epoch": 0.37422037422037424,
      "grad_norm": 1.827469825744629,
      "learning_rate": 8.800000000000001e-05,
      "loss": 2.928,
      "step": 45
    },
    {
      "epoch": 0.38253638253638256,
      "grad_norm": 1.878302812576294,
      "learning_rate": 9e-05,
      "loss": 2.9717,
      "step": 46
    },
    {
      "epoch": 0.3908523908523909,
      "grad_norm": 2.1960086822509766,
      "learning_rate": 9.200000000000001e-05,
      "loss": 2.9037,
      "step": 47
    },
    {
      "epoch": 0.3991683991683992,
      "grad_norm": 3.8804755210876465,
      "learning_rate": 9.4e-05,
      "loss": 2.9174,
      "step": 48
    },
    {
      "epoch": 0.4074844074844075,
      "grad_norm": 1.1289819478988647,
      "learning_rate": 9.6e-05,
      "loss": 2.8997,
      "step": 49
    },
    {
      "epoch": 0.4158004158004158,
      "grad_norm": 4.365386009216309,
      "learning_rate": 9.8e-05,
      "loss": 2.8705,
      "step": 50
    },
    {
      "epoch": 0.42411642411642414,
      "grad_norm": 4.703887462615967,
      "learning_rate": 0.0001,
      "loss": 2.876,
      "step": 51
    },
    {
      "epoch": 0.43243243243243246,
      "grad_norm": 2.8481767177581787,
      "learning_rate": 0.00010200000000000001,
      "loss": 2.8617,
      "step": 52
    },
    {
      "epoch": 0.4407484407484408,
      "grad_norm": 10.58674144744873,
      "learning_rate": 0.00010400000000000001,
      "loss": 2.8998,
      "step": 53
    },
    {
      "epoch": 0.4490644490644491,
      "grad_norm": 6.807188510894775,
      "learning_rate": 0.00010600000000000002,
      "loss": 2.8795,
      "step": 54
    },
    {
      "epoch": 0.4573804573804574,
      "grad_norm": 2.453004837036133,
      "learning_rate": 0.00010800000000000001,
      "loss": 2.8707,
      "step": 55
    },
    {
      "epoch": 0.4656964656964657,
      "grad_norm": 1.8895411491394043,
      "learning_rate": 0.00011000000000000002,
      "loss": 2.8367,
      "step": 56
    },
    {
      "epoch": 0.47401247401247404,
      "grad_norm": 3.585893154144287,
      "learning_rate": 0.00011200000000000001,
      "loss": 2.8564,
      "step": 57
    },
    {
      "epoch": 0.48232848232848236,
      "grad_norm": 2.117868661880493,
      "learning_rate": 0.00011399999999999999,
      "loss": 2.8326,
      "step": 58
    },
    {
      "epoch": 0.49064449064449067,
      "grad_norm": 1.4010989665985107,
      "learning_rate": 0.000116,
      "loss": 2.784,
      "step": 59
    },
    {
      "epoch": 0.498960498960499,
      "grad_norm": 0.8910171985626221,
      "learning_rate": 0.000118,
      "loss": 2.7828,
      "step": 60
    },
    {
      "epoch": 0.5072765072765073,
      "grad_norm": 0.8965553641319275,
      "learning_rate": 0.00012,
      "loss": 2.7768,
      "step": 61
    },
    {
      "epoch": 0.5155925155925156,
      "grad_norm": 1.0502556562423706,
      "learning_rate": 0.000122,
      "loss": 2.7733,
      "step": 62
    },
    {
      "epoch": 0.5239085239085239,
      "grad_norm": 2.425708532333374,
      "learning_rate": 0.000124,
      "loss": 2.7897,
      "step": 63
    },
    {
      "epoch": 0.5322245322245323,
      "grad_norm": 0.8981500864028931,
      "learning_rate": 0.000126,
      "loss": 2.7452,
      "step": 64
    },
    {
      "epoch": 0.5405405405405406,
      "grad_norm": 0.7442967891693115,
      "learning_rate": 0.00012800000000000002,
      "loss": 2.7297,
      "step": 65
    },
    {
      "epoch": 0.5488565488565489,
      "grad_norm": 1.2096165418624878,
      "learning_rate": 0.00013000000000000002,
      "loss": 2.7521,
      "step": 66
    },
    {
      "epoch": 0.5571725571725572,
      "grad_norm": 5.993701457977295,
      "learning_rate": 0.000132,
      "loss": 2.7633,
      "step": 67
    },
    {
      "epoch": 0.5654885654885655,
      "grad_norm": 1.6413137912750244,
      "learning_rate": 0.000134,
      "loss": 2.7269,
      "step": 68
    },
    {
      "epoch": 0.5738045738045738,
      "grad_norm": 2.8430962562561035,
      "learning_rate": 0.00013600000000000003,
      "loss": 2.7717,
      "step": 69
    },
    {
      "epoch": 0.5821205821205822,
      "grad_norm": 2.3216440677642822,
      "learning_rate": 0.000138,
      "loss": 2.8099,
      "step": 70
    },
    {
      "epoch": 0.5904365904365905,
      "grad_norm": 1.4732354879379272,
      "learning_rate": 0.00014,
      "loss": 2.7525,
      "step": 71
    },
    {
      "epoch": 0.5987525987525988,
      "grad_norm": 1.524367332458496,
      "learning_rate": 0.000142,
      "loss": 2.7422,
      "step": 72
    },
    {
      "epoch": 0.6070686070686071,
      "grad_norm": 1.230338215827942,
      "learning_rate": 0.000144,
      "loss": 2.7555,
      "step": 73
    },
    {
      "epoch": 0.6153846153846154,
      "grad_norm": 0.9941631555557251,
      "learning_rate": 0.000146,
      "loss": 2.7042,
      "step": 74
    },
    {
      "epoch": 0.6237006237006237,
      "grad_norm": 1.3642252683639526,
      "learning_rate": 0.000148,
      "loss": 2.7088,
      "step": 75
    },
    {
      "epoch": 0.632016632016632,
      "grad_norm": 0.681107223033905,
      "learning_rate": 0.00015000000000000001,
      "loss": 2.7168,
      "step": 76
    },
    {
      "epoch": 0.6403326403326404,
      "grad_norm": 0.8406685590744019,
      "learning_rate": 0.000152,
      "loss": 2.6938,
      "step": 77
    },
    {
      "epoch": 0.6486486486486487,
      "grad_norm": 0.6661787033081055,
      "learning_rate": 0.000154,
      "loss": 2.7035,
      "step": 78
    },
    {
      "epoch": 0.656964656964657,
      "grad_norm": 0.5472131967544556,
      "learning_rate": 0.00015600000000000002,
      "loss": 2.6913,
      "step": 79
    },
    {
      "epoch": 0.6652806652806653,
      "grad_norm": 0.5465010404586792,
      "learning_rate": 0.00015800000000000002,
      "loss": 2.6913,
      "step": 80
    },
    {
      "epoch": 0.6735966735966736,
      "grad_norm": 0.6352857351303101,
      "learning_rate": 0.00016,
      "loss": 2.6834,
      "step": 81
    },
    {
      "epoch": 0.681912681912682,
      "grad_norm": 0.7992680668830872,
      "learning_rate": 0.000162,
      "loss": 2.6984,
      "step": 82
    },
    {
      "epoch": 0.6902286902286903,
      "grad_norm": 0.543773889541626,
      "learning_rate": 0.000164,
      "loss": 2.6631,
      "step": 83
    },
    {
      "epoch": 0.6985446985446986,
      "grad_norm": 0.4968494474887848,
      "learning_rate": 0.000166,
      "loss": 2.6707,
      "step": 84
    },
    {
      "epoch": 0.7068607068607069,
      "grad_norm": 0.5793298482894897,
      "learning_rate": 0.000168,
      "loss": 2.6776,
      "step": 85
    },
    {
      "epoch": 0.7151767151767152,
      "grad_norm": 1.137745976448059,
      "learning_rate": 0.00017,
      "loss": 2.6852,
      "step": 86
    },
    {
      "epoch": 0.7234927234927235,
      "grad_norm": 1.2862417697906494,
      "learning_rate": 0.000172,
      "loss": 2.6967,
      "step": 87
    },
    {
      "epoch": 0.7318087318087318,
      "grad_norm": 0.5603763461112976,
      "learning_rate": 0.000174,
      "loss": 2.6928,
      "step": 88
    },
    {
      "epoch": 0.7401247401247402,
      "grad_norm": 0.9993265867233276,
      "learning_rate": 0.00017600000000000002,
      "loss": 2.667,
      "step": 89
    },
    {
      "epoch": 0.7484407484407485,
      "grad_norm": 0.982528567314148,
      "learning_rate": 0.00017800000000000002,
      "loss": 2.6673,
      "step": 90
    },
    {
      "epoch": 0.7567567567567568,
      "grad_norm": 0.6244588494300842,
      "learning_rate": 0.00018,
      "loss": 2.6803,
      "step": 91
    },
    {
      "epoch": 0.7650727650727651,
      "grad_norm": 0.5460201501846313,
      "learning_rate": 0.000182,
      "loss": 2.6341,
      "step": 92
    },
    {
      "epoch": 0.7733887733887734,
      "grad_norm": 0.618466854095459,
      "learning_rate": 0.00018400000000000003,
      "loss": 2.6338,
      "step": 93
    },
    {
      "epoch": 0.7817047817047817,
      "grad_norm": 0.5517793893814087,
      "learning_rate": 0.00018600000000000002,
      "loss": 2.6288,
      "step": 94
    },
    {
      "epoch": 0.7900207900207901,
      "grad_norm": 0.5810732841491699,
      "learning_rate": 0.000188,
      "loss": 2.6575,
      "step": 95
    },
    {
      "epoch": 0.7983367983367984,
      "grad_norm": 0.5090439915657043,
      "learning_rate": 0.00019,
      "loss": 2.6646,
      "step": 96
    },
    {
      "epoch": 0.8066528066528067,
      "grad_norm": 0.5052196979522705,
      "learning_rate": 0.000192,
      "loss": 2.6515,
      "step": 97
    },
    {
      "epoch": 0.814968814968815,
      "grad_norm": 0.4681943953037262,
      "learning_rate": 0.000194,
      "loss": 2.6507,
      "step": 98
    },
    {
      "epoch": 0.8232848232848233,
      "grad_norm": 0.47244778275489807,
      "learning_rate": 0.000196,
      "loss": 2.6259,
      "step": 99
    },
    {
      "epoch": 0.8316008316008316,
      "grad_norm": 0.6378790140151978,
      "learning_rate": 0.00019800000000000002,
      "loss": 2.6371,
      "step": 100
    },
    {
      "epoch": 0.83991683991684,
      "grad_norm": 0.5825613737106323,
      "learning_rate": 0.0002,
      "loss": 2.6397,
      "step": 101
    },
    {
      "epoch": 0.8482328482328483,
      "grad_norm": 0.4540935754776001,
      "learning_rate": 0.00019999270008556108,
      "loss": 2.6295,
      "step": 102
    },
    {
      "epoch": 0.8565488565488566,
      "grad_norm": 0.4203695058822632,
      "learning_rate": 0.00019997080140801932,
      "loss": 2.6144,
      "step": 103
    },
    {
      "epoch": 0.8648648648648649,
      "grad_norm": 0.4233225882053375,
      "learning_rate": 0.00019993430716454413,
      "loss": 2.6378,
      "step": 104
    },
    {
      "epoch": 0.8731808731808732,
      "grad_norm": 0.4105020761489868,
      "learning_rate": 0.00019988322268323268,
      "loss": 2.6502,
      "step": 105
    },
    {
      "epoch": 0.8814968814968815,
      "grad_norm": 0.38928887248039246,
      "learning_rate": 0.00019981755542233177,
      "loss": 2.6154,
      "step": 106
    },
    {
      "epoch": 0.8898128898128899,
      "grad_norm": 0.4148002862930298,
      "learning_rate": 0.00019973731496914914,
      "loss": 2.6405,
      "step": 107
    },
    {
      "epoch": 0.8981288981288982,
      "grad_norm": 0.39650681614875793,
      "learning_rate": 0.00019964251303865362,
      "loss": 2.6215,
      "step": 108
    },
    {
      "epoch": 0.9064449064449065,
      "grad_norm": 0.4093710780143738,
      "learning_rate": 0.00019953316347176488,
      "loss": 2.6382,
      "step": 109
    },
    {
      "epoch": 0.9147609147609148,
      "grad_norm": 0.43350887298583984,
      "learning_rate": 0.00019940928223333252,
      "loss": 2.6281,
      "step": 110
    },
    {
      "epoch": 0.9230769230769231,
      "grad_norm": 0.5919923782348633,
      "learning_rate": 0.0001992708874098054,
      "loss": 2.6348,
      "step": 111
    },
    {
      "epoch": 0.9313929313929314,
      "grad_norm": 2.143718957901001,
      "learning_rate": 0.00019911799920659093,
      "loss": 2.6295,
      "step": 112
    },
    {
      "epoch": 0.9397089397089398,
      "grad_norm": 0.8955861330032349,
      "learning_rate": 0.0001989506399451051,
      "loss": 2.5963,
      "step": 113
    },
    {
      "epoch": 0.9480249480249481,
      "grad_norm": 0.7444430589675903,
      "learning_rate": 0.00019876883405951377,
      "loss": 2.6131,
      "step": 114
    },
    {
      "epoch": 0.9563409563409564,
      "grad_norm": 0.6534228324890137,
      "learning_rate": 0.0001985726080931651,
      "loss": 2.621,
      "step": 115
    },
    {
      "epoch": 0.9646569646569647,
      "grad_norm": 0.6738382577896118,
      "learning_rate": 0.00019836199069471437,
      "loss": 2.6123,
      "step": 116
    },
    {
      "epoch": 0.972972972972973,
      "grad_norm": 0.46829137206077576,
      "learning_rate": 0.00019813701261394136,
      "loss": 2.6216,
      "step": 117
    },
    {
      "epoch": 0.9812889812889813,
      "grad_norm": 0.5436323881149292,
      "learning_rate": 0.00019789770669726087,
      "loss": 2.6364,
      "step": 118
    },
    {
      "epoch": 0.9896049896049897,
      "grad_norm": 0.3988193869590759,
      "learning_rate": 0.00019764410788292722,
      "loss": 2.6036,
      "step": 119
    },
    {
      "epoch": 0.997920997920998,
      "grad_norm": 0.43007829785346985,
      "learning_rate": 0.00019737625319593335,
      "loss": 2.6067,
      "step": 120
    },
    {
      "epoch": 1.0,
      "grad_norm": 0.6972622871398926,
      "learning_rate": 0.0001970941817426052,
      "loss": 2.5875,
      "step": 121
    },
    {
      "epoch": 1.0083160083160083,
      "grad_norm": 0.7436782121658325,
      "learning_rate": 0.00019679793470489228,
      "loss": 2.6127,
      "step": 122
    },
    {
      "epoch": 1.0166320166320166,
      "grad_norm": 2.4193379878997803,
      "learning_rate": 0.00019648755533435518,
      "loss": 2.575,
      "step": 123
    },
    {
      "epoch": 1.024948024948025,
      "grad_norm": 0.6707261204719543,
      "learning_rate": 0.00019616308894585078,
      "loss": 2.5704,
      "step": 124
    },
    {
      "epoch": 1.0332640332640333,
      "grad_norm": 0.46323344111442566,
      "learning_rate": 0.00019582458291091663,
      "loss": 2.5438,
      "step": 125
    },
    {
      "epoch": 1.0415800415800416,
      "grad_norm": 0.5376307964324951,
      "learning_rate": 0.00019547208665085457,
      "loss": 2.5629,
      "step": 126
    },
    {
      "epoch": 1.04989604989605,
      "grad_norm": 0.4404616951942444,
      "learning_rate": 0.00019510565162951537,
      "loss": 2.5539,
      "step": 127
    },
    {
      "epoch": 1.0582120582120582,
      "grad_norm": 0.45081833004951477,
      "learning_rate": 0.00019472533134578507,
      "loss": 2.558,
      "step": 128
    },
    {
      "epoch": 1.0665280665280665,
      "grad_norm": 0.6300471425056458,
      "learning_rate": 0.0001943311813257743,
      "loss": 2.5688,
      "step": 129
    },
    {
      "epoch": 1.0748440748440748,
      "grad_norm": 0.49277371168136597,
      "learning_rate": 0.00019392325911471155,
      "loss": 2.5545,
      "step": 130
    },
    {
      "epoch": 1.0831600831600832,
      "grad_norm": 0.4726185202598572,
      "learning_rate": 0.0001935016242685415,
      "loss": 2.569,
      "step": 131
    },
    {
      "epoch": 1.0914760914760915,
      "grad_norm": 0.5317509174346924,
      "learning_rate": 0.00019306633834523024,
      "loss": 2.5738,
      "step": 132
    },
    {
      "epoch": 1.0997920997920998,
      "grad_norm": 0.4991399645805359,
      "learning_rate": 0.00019261746489577765,
      "loss": 2.5491,
      "step": 133
    },
    {
      "epoch": 1.1081081081081081,
      "grad_norm": 0.6071000099182129,
      "learning_rate": 0.0001921550694549393,
      "loss": 2.5667,
      "step": 134
    },
    {
      "epoch": 1.1164241164241164,
      "grad_norm": 0.4629747271537781,
      "learning_rate": 0.00019167921953165825,
      "loss": 2.5514,
      "step": 135
    },
    {
      "epoch": 1.1247401247401247,
      "grad_norm": 0.41420885920524597,
      "learning_rate": 0.00019118998459920902,
      "loss": 2.5432,
      "step": 136
    },
    {
      "epoch": 1.133056133056133,
      "grad_norm": 0.41042062640190125,
      "learning_rate": 0.00019068743608505455,
      "loss": 2.5783,
      "step": 137
    },
    {
      "epoch": 1.1413721413721414,
      "grad_norm": 0.40009400248527527,
      "learning_rate": 0.00019017164736041795,
      "loss": 2.5598,
      "step": 138
    },
    {
      "epoch": 1.1496881496881497,
      "grad_norm": 0.44067704677581787,
      "learning_rate": 0.00018964269372957038,
      "loss": 2.56,
      "step": 139
    },
    {
      "epoch": 1.158004158004158,
      "grad_norm": 0.4079325795173645,
      "learning_rate": 0.0001891006524188368,
      "loss": 2.5748,
      "step": 140
    },
    {
      "epoch": 1.1663201663201663,
      "grad_norm": 0.4190751314163208,
      "learning_rate": 0.000188545602565321,
      "loss": 2.5569,
      "step": 141
    },
    {
      "epoch": 1.1746361746361746,
      "grad_norm": 0.39320653676986694,
      "learning_rate": 0.00018797762520535177,
      "loss": 2.5689,
      "step": 142
    },
    {
      "epoch": 1.182952182952183,
      "grad_norm": 0.3822016716003418,
      "learning_rate": 0.0001873968032626518,
      "loss": 2.543,
      "step": 143
    },
    {
      "epoch": 1.1912681912681913,
      "grad_norm": 0.37677645683288574,
      "learning_rate": 0.00018680322153623075,
      "loss": 2.5315,
      "step": 144
    },
    {
      "epoch": 1.1995841995841996,
      "grad_norm": 0.3912084996700287,
      "learning_rate": 0.00018619696668800492,
      "loss": 2.5344,
      "step": 145
    },
    {
      "epoch": 1.207900207900208,
      "grad_norm": 0.3857469856739044,
      "learning_rate": 0.00018557812723014476,
      "loss": 2.5631,
      "step": 146
    },
    {
      "epoch": 1.2162162162162162,
      "grad_norm": 0.38990500569343567,
      "learning_rate": 0.0001849467935121521,
      "loss": 2.5472,
      "step": 147
    },
    {
      "epoch": 1.2245322245322245,
      "grad_norm": 0.38198038935661316,
      "learning_rate": 0.00018430305770766948,
      "loss": 2.5287,
      "step": 148
    },
    {
      "epoch": 1.2328482328482329,
      "grad_norm": 0.5182828307151794,
      "learning_rate": 0.00018364701380102266,
      "loss": 2.5404,
      "step": 149
    },
    {
      "epoch": 1.2411642411642412,
      "grad_norm": 0.4881569445133209,
      "learning_rate": 0.00018297875757349952,
      "loss": 2.5547,
      "step": 150
    },
    {
      "epoch": 1.2494802494802495,
      "grad_norm": 0.4347280263900757,
      "learning_rate": 0.00018229838658936564,
      "loss": 2.5338,
      "step": 151
    },
    {
      "epoch": 1.2577962577962578,
      "grad_norm": 0.47588396072387695,
      "learning_rate": 0.0001816060001816205,
      "loss": 2.5507,
      "step": 152
    },
    {
      "epoch": 1.2661122661122661,
      "grad_norm": 0.3899973928928375,
      "learning_rate": 0.00018090169943749476,
      "loss": 2.5596,
      "step": 153
    },
    {
      "epoch": 1.2744282744282744,
      "grad_norm": 0.39740344882011414,
      "learning_rate": 0.00018018558718369186,
      "loss": 2.539,
      "step": 154
    },
    {
      "epoch": 1.2827442827442828,
      "grad_norm": 0.42894020676612854,
      "learning_rate": 0.00017945776797137543,
      "loss": 2.5211,
      "step": 155
    },
    {
      "epoch": 1.291060291060291,
      "grad_norm": 0.42803955078125,
      "learning_rate": 0.00017871834806090501,
      "loss": 2.5267,
      "step": 156
    },
    {
      "epoch": 1.2993762993762994,
      "grad_norm": 0.4332488477230072,
      "learning_rate": 0.00017796743540632223,
      "loss": 2.5422,
      "step": 157
    },
    {
      "epoch": 1.3076923076923077,
      "grad_norm": 0.5245593786239624,
      "learning_rate": 0.00017720513963958968,
      "loss": 2.5275,
      "step": 158
    },
    {
      "epoch": 1.316008316008316,
      "grad_norm": 0.4242333173751831,
      "learning_rate": 0.00017643157205458483,
      "loss": 2.5284,
      "step": 159
    },
    {
      "epoch": 1.3243243243243243,
      "grad_norm": 0.4454910457134247,
      "learning_rate": 0.00017564684559085136,
      "loss": 2.5204,
      "step": 160
    },
    {
      "epoch": 1.3326403326403327,
      "grad_norm": 0.5117389559745789,
      "learning_rate": 0.00017485107481711012,
      "loss": 2.5484,
      "step": 161
    },
    {
      "epoch": 1.340956340956341,
      "grad_norm": 0.5533547401428223,
      "learning_rate": 0.00017404437591453235,
      "loss": 2.5256,
      "step": 162
    },
    {
      "epoch": 1.3492723492723493,
      "grad_norm": 0.4423222541809082,
      "learning_rate": 0.00017322686665977737,
      "loss": 2.5516,
      "step": 163
    },
    {
      "epoch": 1.3575883575883576,
      "grad_norm": 0.6984363198280334,
      "learning_rate": 0.00017239866640779745,
      "loss": 2.5392,
      "step": 164
    },
    {
      "epoch": 1.365904365904366,
      "grad_norm": 0.46573567390441895,
      "learning_rate": 0.00017155989607441213,
      "loss": 2.5215,
      "step": 165
    },
    {
      "epoch": 1.3742203742203742,
      "grad_norm": 0.43190667033195496,
      "learning_rate": 0.00017071067811865476,
      "loss": 2.5109,
      "step": 166
    },
    {
      "epoch": 1.3825363825363826,
      "grad_norm": 0.6366004943847656,
      "learning_rate": 0.00016985113652489374,
      "loss": 2.5607,
      "step": 167
    },
    {
      "epoch": 1.3908523908523909,
      "grad_norm": 0.4087092876434326,
      "learning_rate": 0.00016898139678473076,
      "loss": 2.5446,
      "step": 168
    },
    {
      "epoch": 1.3991683991683992,
      "grad_norm": 0.40001702308654785,
      "learning_rate": 0.00016810158587867973,
      "loss": 2.5087,
      "step": 169
    },
    {
      "epoch": 1.4074844074844075,
      "grad_norm": 0.40865230560302734,
      "learning_rate": 0.00016721183225762727,
      "loss": 2.5256,
      "step": 170
    },
    {
      "epoch": 1.4158004158004158,
      "grad_norm": 0.41592201590538025,
      "learning_rate": 0.00016631226582407952,
      "loss": 2.5457,
      "step": 171
    },
    {
      "epoch": 1.4241164241164241,
      "grad_norm": 0.47380271553993225,
      "learning_rate": 0.00016540301791319645,
      "loss": 2.5239,
      "step": 172
    },
    {
      "epoch": 1.4324324324324325,
      "grad_norm": 0.3751722574234009,
      "learning_rate": 0.00016448422127361706,
      "loss": 2.5277,
      "step": 173
    },
    {
      "epoch": 1.4407484407484408,
      "grad_norm": 0.531104326248169,
      "learning_rate": 0.00016355601004807856,
      "loss": 2.5095,
      "step": 174
    },
    {
      "epoch": 1.449064449064449,
      "grad_norm": 0.3729030191898346,
      "learning_rate": 0.00016261851975383137,
      "loss": 2.508,
      "step": 175
    },
    {
      "epoch": 1.4573804573804574,
      "grad_norm": 0.4520551562309265,
      "learning_rate": 0.00016167188726285434,
      "loss": 2.5198,
      "step": 176
    },
    {
      "epoch": 1.4656964656964657,
      "grad_norm": 0.4487551152706146,
      "learning_rate": 0.00016071625078187114,
      "loss": 2.5153,
      "step": 177
    },
    {
      "epoch": 1.474012474012474,
      "grad_norm": 0.41815704107284546,
      "learning_rate": 0.00015975174983217275,
      "loss": 2.5218,
      "step": 178
    },
    {
      "epoch": 1.4823284823284824,
      "grad_norm": 0.4343583583831787,
      "learning_rate": 0.00015877852522924732,
      "loss": 2.518,
      "step": 179
    },
    {
      "epoch": 1.4906444906444907,
      "grad_norm": 0.4115164279937744,
      "learning_rate": 0.0001577967190622215,
      "loss": 2.5137,
      "step": 180
    },
    {
      "epoch": 1.498960498960499,
      "grad_norm": 0.366964191198349,
      "learning_rate": 0.00015680647467311557,
      "loss": 2.5379,
      "step": 181
    },
    {
      "epoch": 1.5072765072765073,
      "grad_norm": 0.42494460940361023,
      "learning_rate": 0.00015580793663591585,
      "loss": 2.5471,
      "step": 182
    },
    {
      "epoch": 1.5155925155925156,
      "grad_norm": 0.39186251163482666,
      "learning_rate": 0.00015480125073546704,
      "loss": 2.5131,
      "step": 183
    },
    {
      "epoch": 1.523908523908524,
      "grad_norm": 0.3860762119293213,
      "learning_rate": 0.00015378656394618787,
      "loss": 2.5355,
      "step": 184
    },
    {
      "epoch": 1.5322245322245323,
      "grad_norm": 0.4578627943992615,
      "learning_rate": 0.0001527640244106133,
      "loss": 2.5119,
      "step": 185
    },
    {
      "epoch": 1.5405405405405406,
      "grad_norm": 0.3482917547225952,
      "learning_rate": 0.00015173378141776568,
      "loss": 2.512,
      "step": 186
    },
    {
      "epoch": 1.5488565488565489,
      "grad_norm": 0.3526865541934967,
      "learning_rate": 0.00015069598538135906,
      "loss": 2.5142,
      "step": 187
    },
    {
      "epoch": 1.5571725571725572,
      "grad_norm": 0.3888881206512451,
      "learning_rate": 0.0001496507878178388,
      "loss": 2.5202,
      "step": 188
    },
    {
      "epoch": 1.5654885654885655,
      "grad_norm": 0.39839354157447815,
      "learning_rate": 0.0001485983413242606,
      "loss": 2.5178,
      "step": 189
    },
    {
      "epoch": 1.5738045738045738,
      "grad_norm": 0.42406460642814636,
      "learning_rate": 0.00014753879955601163,
      "loss": 2.4996,
      "step": 190
    },
    {
      "epoch": 1.5821205821205822,
      "grad_norm": 0.4432888627052307,
      "learning_rate": 0.00014647231720437686,
      "loss": 2.5232,
      "step": 191
    },
    {
      "epoch": 1.5904365904365905,
      "grad_norm": 0.43413785099983215,
      "learning_rate": 0.00014539904997395468,
      "loss": 2.5243,
      "step": 192
    },
    {
      "epoch": 1.5987525987525988,
      "grad_norm": 0.3551112413406372,
      "learning_rate": 0.00014431915455992414,
      "loss": 2.5134,
      "step": 193
    },
    {
      "epoch": 1.607068607068607,
      "grad_norm": 0.36720749735832214,
      "learning_rate": 0.00014323278862516775,
      "loss": 2.502,
      "step": 194
    },
    {
      "epoch": 1.6153846153846154,
      "grad_norm": 0.35447460412979126,
      "learning_rate": 0.00014214011077725292,
      "loss": 2.5036,
      "step": 195
    },
    {
      "epoch": 1.6237006237006237,
      "grad_norm": 0.3743102550506592,
      "learning_rate": 0.0001410412805452757,
      "loss": 2.4936,
      "step": 196
    },
    {
      "epoch": 1.632016632016632,
      "grad_norm": 0.36030158400535583,
      "learning_rate": 0.00013993645835656953,
      "loss": 2.4851,
      "step": 197
    },
    {
      "epoch": 1.6403326403326404,
      "grad_norm": 0.3682660162448883,
      "learning_rate": 0.0001388258055132835,
      "loss": 2.5107,
      "step": 198
    },
    {
      "epoch": 1.6486486486486487,
      "grad_norm": 0.37054452300071716,
      "learning_rate": 0.00013770948416883205,
      "loss": 2.4875,
      "step": 199
    },
    {
      "epoch": 1.656964656964657,
      "grad_norm": 0.4111086428165436,
      "learning_rate": 0.00013658765730422125,
      "loss": 2.5055,
      "step": 200
    },
    {
      "epoch": 1.6652806652806653,
      "grad_norm": 0.36359384655952454,
      "learning_rate": 0.00013546048870425356,
      "loss": 2.5099,
      "step": 201
    },
    {
      "epoch": 1.6735966735966736,
      "grad_norm": 0.40381795167922974,
      "learning_rate": 0.00013432814293361584,
      "loss": 2.5162,
      "step": 202
    },
    {
      "epoch": 1.681912681912682,
      "grad_norm": 0.3458056151866913,
      "learning_rate": 0.00013319078531285285,
      "loss": 2.4827,
      "step": 203
    },
    {
      "epoch": 1.6902286902286903,
      "grad_norm": 0.35546690225601196,
      "learning_rate": 0.00013204858189423097,
      "loss": 2.5162,
      "step": 204
    },
    {
      "epoch": 1.6985446985446986,
      "grad_norm": 0.37076064944267273,
      "learning_rate": 0.00013090169943749476,
      "loss": 2.5201,
      "step": 205
    },
    {
      "epoch": 1.706860706860707,
      "grad_norm": 0.3626950681209564,
      "learning_rate": 0.00012975030538552032,
      "loss": 2.5196,
      "step": 206
    },
    {
      "epoch": 1.7151767151767152,
      "grad_norm": 0.38436341285705566,
      "learning_rate": 0.00012859456783986893,
      "loss": 2.5157,
      "step": 207
    },
    {
      "epoch": 1.7234927234927235,
      "grad_norm": 0.38007652759552,
      "learning_rate": 0.0001274346555362446,
      "loss": 2.4935,
      "step": 208
    },
    {
      "epoch": 1.7318087318087318,
      "grad_norm": 0.38916251063346863,
      "learning_rate": 0.0001262707378198587,
      "loss": 2.5083,
      "step": 209
    },
    {
      "epoch": 1.7401247401247402,
      "grad_norm": 0.37072858214378357,
      "learning_rate": 0.00012510298462070619,
      "loss": 2.4915,
      "step": 210
    },
    {
      "epoch": 1.7484407484407485,
      "grad_norm": 0.4798509478569031,
      "learning_rate": 0.0001239315664287558,
      "loss": 2.5189,
      "step": 211
    },
    {
      "epoch": 1.7567567567567568,
      "grad_norm": 0.4451827108860016,
      "learning_rate": 0.000122756654269059,
      "loss": 2.5231,
      "step": 212
    },
    {
      "epoch": 1.7650727650727651,
      "grad_norm": 0.35244473814964294,
      "learning_rate": 0.00012157841967678063,
      "loss": 2.5061,
      "step": 213
    },
    {
      "epoch": 1.7733887733887734,
      "grad_norm": 0.34157419204711914,
      "learning_rate": 0.00012039703467215488,
      "loss": 2.5044,
      "step": 214
    },
    {
      "epoch": 1.7817047817047817,
      "grad_norm": 0.40339285135269165,
      "learning_rate": 0.00011921267173537086,
      "loss": 2.506,
      "step": 215
    },
    {
      "epoch": 1.79002079002079,
      "grad_norm": 0.3662133514881134,
      "learning_rate": 0.0001180255037813906,
      "loss": 2.5012,
      "step": 216
    },
    {
      "epoch": 1.7983367983367984,
      "grad_norm": 0.3495447039604187,
      "learning_rate": 0.00011683570413470383,
      "loss": 2.5,
      "step": 217
    },
    {
      "epoch": 1.8066528066528067,
      "grad_norm": 0.36315813660621643,
      "learning_rate": 0.0001156434465040231,
      "loss": 2.4923,
      "step": 218
    },
    {
      "epoch": 1.814968814968815,
      "grad_norm": 0.3807941973209381,
      "learning_rate": 0.00011444890495692213,
      "loss": 2.5123,
      "step": 219
    },
    {
      "epoch": 1.8232848232848233,
      "grad_norm": 0.36616984009742737,
      "learning_rate": 0.00011325225389442277,
      "loss": 2.4954,
      "step": 220
    },
    {
      "epoch": 1.8316008316008316,
      "grad_norm": 0.372138112783432,
      "learning_rate": 0.0001120536680255323,
      "loss": 2.488,
      "step": 221
    },
    {
      "epoch": 1.83991683991684,
      "grad_norm": 0.37164467573165894,
      "learning_rate": 0.00011085332234173664,
      "loss": 2.4883,
      "step": 222
    },
    {
      "epoch": 1.8482328482328483,
      "grad_norm": 0.34948042035102844,
      "learning_rate": 0.00010965139209145152,
      "loss": 2.4864,
      "step": 223
    },
    {
      "epoch": 1.8565488565488566,
      "grad_norm": 0.35938721895217896,
      "learning_rate": 0.00010844805275443673,
      "loss": 2.4928,
      "step": 224
    },
    {
      "epoch": 1.864864864864865,
      "grad_norm": 0.3879775404930115,
      "learning_rate": 0.00010724348001617625,
      "loss": 2.4938,
      "step": 225
    },
    {
      "epoch": 1.8731808731808732,
      "grad_norm": 0.36195841431617737,
      "learning_rate": 0.00010603784974222861,
      "loss": 2.4925,
      "step": 226
    },
    {
      "epoch": 1.8814968814968815,
      "grad_norm": 0.35239464044570923,
      "learning_rate": 0.00010483133795255071,
      "loss": 2.4973,
      "step": 227
    },
    {
      "epoch": 1.8898128898128899,
      "grad_norm": 0.3456074297428131,
      "learning_rate": 0.00010362412079579924,
      "loss": 2.4966,
      "step": 228
    },
    {
      "epoch": 1.8981288981288982,
      "grad_norm": 0.3808579444885254,
      "learning_rate": 0.00010241637452361323,
      "loss": 2.5087,
      "step": 229
    },
    {
      "epoch": 1.9064449064449065,
      "grad_norm": 0.39099177718162537,
      "learning_rate": 0.00010120827546488174,
      "loss": 2.4894,
      "step": 230
    },
    {
      "epoch": 1.9147609147609148,
      "grad_norm": 0.3821989595890045,
      "learning_rate": 0.0001,
      "loss": 2.4971,
      "step": 231
    },
    {
      "epoch": 1.9230769230769231,
      "grad_norm": 0.3640426695346832,
      "learning_rate": 9.879172453511827e-05,
      "loss": 2.4893,
      "step": 232
    },
    {
      "epoch": 1.9313929313929314,
      "grad_norm": 0.3754567801952362,
      "learning_rate": 9.75836254763868e-05,
      "loss": 2.4871,
      "step": 233
    },
    {
      "epoch": 1.9397089397089398,
      "grad_norm": 0.3793177902698517,
      "learning_rate": 9.63758792042008e-05,
      "loss": 2.4929,
      "step": 234
    },
    {
      "epoch": 1.948024948024948,
      "grad_norm": 0.35912153124809265,
      "learning_rate": 9.516866204744931e-05,
      "loss": 2.503,
      "step": 235
    },
    {
      "epoch": 1.9563409563409564,
      "grad_norm": 0.3523949384689331,
      "learning_rate": 9.396215025777139e-05,
      "loss": 2.4869,
      "step": 236
    },
    {
      "epoch": 1.9646569646569647,
      "grad_norm": 0.38750869035720825,
      "learning_rate": 9.275651998382377e-05,
      "loss": 2.4957,
      "step": 237
    },
    {
      "epoch": 1.972972972972973,
      "grad_norm": 0.3791241943836212,
      "learning_rate": 9.155194724556331e-05,
      "loss": 2.506,
      "step": 238
    },
    {
      "epoch": 1.9812889812889813,
      "grad_norm": 0.34388408064842224,
      "learning_rate": 9.034860790854849e-05,
      "loss": 2.5003,
      "step": 239
    },
    {
      "epoch": 1.9896049896049897,
      "grad_norm": 0.3853475749492645,
      "learning_rate": 8.914667765826338e-05,
      "loss": 2.4585,
      "step": 240
    },
    {
      "epoch": 1.997920997920998,
      "grad_norm": 0.3552019000053406,
      "learning_rate": 8.79463319744677e-05,
      "loss": 2.4799,
      "step": 241
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6879647374153137,
      "learning_rate": 8.674774610557728e-05,
      "loss": 2.485,
      "step": 242
    },
    {
      "epoch": 2.008316008316008,
      "grad_norm": 0.4841165542602539,
      "learning_rate": 8.55510950430779e-05,
      "loss": 2.3719,
      "step": 243
    },
    {
      "epoch": 2.0166320166320166,
      "grad_norm": 0.40582990646362305,
      "learning_rate": 8.435655349597689e-05,
      "loss": 2.3914,
      "step": 244
    },
    {
      "epoch": 2.024948024948025,
      "grad_norm": 0.44056642055511475,
      "learning_rate": 8.316429586529615e-05,
      "loss": 2.3862,
      "step": 245
    },
    {
      "epoch": 2.0332640332640333,
      "grad_norm": 0.5893064141273499,
      "learning_rate": 8.197449621860943e-05,
      "loss": 2.3759,
      "step": 246
    },
    {
      "epoch": 2.0415800415800414,
      "grad_norm": 0.5335010290145874,
      "learning_rate": 8.078732826462915e-05,
      "loss": 2.3966,
      "step": 247
    },
    {
      "epoch": 2.04989604989605,
      "grad_norm": 0.40315303206443787,
      "learning_rate": 7.960296532784515e-05,
      "loss": 2.3886,
      "step": 248
    },
    {
      "epoch": 2.0582120582120584,
      "grad_norm": 0.45403429865837097,
      "learning_rate": 7.84215803232194e-05,
      "loss": 2.3664,
      "step": 249
    },
    {
      "epoch": 2.0665280665280665,
      "grad_norm": 0.44575178623199463,
      "learning_rate": 7.7243345730941e-05,
      "loss": 2.3705,
      "step": 250
    },
    {
      "epoch": 2.0748440748440746,
      "grad_norm": 0.41231396794319153,
      "learning_rate": 7.606843357124426e-05,
      "loss": 2.3417,
      "step": 251
    },
    {
      "epoch": 2.083160083160083,
      "grad_norm": 0.45150724053382874,
      "learning_rate": 7.489701537929384e-05,
      "loss": 2.3467,
      "step": 252
    },
    {
      "epoch": 2.0914760914760917,
      "grad_norm": 0.45270782709121704,
      "learning_rate": 7.372926218014131e-05,
      "loss": 2.3589,
      "step": 253
    },
    {
      "epoch": 2.0997920997921,
      "grad_norm": 0.49504727125167847,
      "learning_rate": 7.256534446375542e-05,
      "loss": 2.3821,
      "step": 254
    },
    {
      "epoch": 2.108108108108108,
      "grad_norm": 0.40953919291496277,
      "learning_rate": 7.14054321601311e-05,
      "loss": 2.3754,
      "step": 255
    },
    {
      "epoch": 2.1164241164241164,
      "grad_norm": 0.429299533367157,
      "learning_rate": 7.024969461447972e-05,
      "loss": 2.3582,
      "step": 256
    },
    {
      "epoch": 2.124740124740125,
      "grad_norm": 0.4189452826976776,
      "learning_rate": 6.909830056250527e-05,
      "loss": 2.3636,
      "step": 257
    },
    {
      "epoch": 2.133056133056133,
      "grad_norm": 0.4161284267902374,
      "learning_rate": 6.795141810576906e-05,
      "loss": 2.3543,
      "step": 258
    },
    {
      "epoch": 2.141372141372141,
      "grad_norm": 0.4045678675174713,
      "learning_rate": 6.680921468714719e-05,
      "loss": 2.362,
      "step": 259
    },
    {
      "epoch": 2.1496881496881497,
      "grad_norm": 0.47194382548332214,
      "learning_rate": 6.567185706638417e-05,
      "loss": 2.377,
      "step": 260
    },
    {
      "epoch": 2.1580041580041582,
      "grad_norm": 0.47539058327674866,
      "learning_rate": 6.453951129574644e-05,
      "loss": 2.3645,
      "step": 261
    },
    {
      "epoch": 2.1663201663201663,
      "grad_norm": 0.43471434712409973,
      "learning_rate": 6.341234269577879e-05,
      "loss": 2.3441,
      "step": 262
    },
    {
      "epoch": 2.1746361746361744,
      "grad_norm": 0.45001259446144104,
      "learning_rate": 6.229051583116796e-05,
      "loss": 2.3124,
      "step": 263
    },
    {
      "epoch": 2.182952182952183,
      "grad_norm": 0.40879690647125244,
      "learning_rate": 6.117419448671651e-05,
      "loss": 2.3396,
      "step": 264
    },
    {
      "epoch": 2.1912681912681915,
      "grad_norm": 0.42216163873672485,
      "learning_rate": 6.006354164343046e-05,
      "loss": 2.3724,
      "step": 265
    },
    {
      "epoch": 2.1995841995841996,
      "grad_norm": 0.41845160722732544,
      "learning_rate": 5.8958719454724346e-05,
      "loss": 2.3541,
      "step": 266
    },
    {
      "epoch": 2.2079002079002077,
      "grad_norm": 0.423265665769577,
      "learning_rate": 5.785988922274711e-05,
      "loss": 2.3594,
      "step": 267
    },
    {
      "epoch": 2.2162162162162162,
      "grad_norm": 0.4284776747226715,
      "learning_rate": 5.676721137483225e-05,
      "loss": 2.3676,
      "step": 268
    },
    {
      "epoch": 2.2245322245322248,
      "grad_norm": 0.4213842749595642,
      "learning_rate": 5.568084544007588e-05,
      "loss": 2.3392,
      "step": 269
    },
    {
      "epoch": 2.232848232848233,
      "grad_norm": 0.4187883138656616,
      "learning_rate": 5.4600950026045326e-05,
      "loss": 2.3501,
      "step": 270
    },
    {
      "epoch": 2.241164241164241,
      "grad_norm": 0.41858911514282227,
      "learning_rate": 5.3527682795623146e-05,
      "loss": 2.3563,
      "step": 271
    },
    {
      "epoch": 2.2494802494802495,
      "grad_norm": 0.4647063910961151,
      "learning_rate": 5.246120044398839e-05,
      "loss": 2.3575,
      "step": 272
    },
    {
      "epoch": 2.257796257796258,
      "grad_norm": 0.41441211104393005,
      "learning_rate": 5.14016586757394e-05,
      "loss": 2.3534,
      "step": 273
    },
    {
      "epoch": 2.266112266112266,
      "grad_norm": 0.4543509781360626,
      "learning_rate": 5.0349212182161254e-05,
      "loss": 2.3562,
      "step": 274
    },
    {
      "epoch": 2.274428274428274,
      "grad_norm": 0.4140624403953552,
      "learning_rate": 4.9304014618640995e-05,
      "loss": 2.3488,
      "step": 275
    },
    {
      "epoch": 2.2827442827442828,
      "grad_norm": 0.4368607699871063,
      "learning_rate": 4.826621858223431e-05,
      "loss": 2.3532,
      "step": 276
    },
    {
      "epoch": 2.2910602910602913,
      "grad_norm": 0.41528797149658203,
      "learning_rate": 4.723597558938672e-05,
      "loss": 2.3591,
      "step": 277
    },
    {
      "epoch": 2.2993762993762994,
      "grad_norm": 0.43311697244644165,
      "learning_rate": 4.6213436053812144e-05,
      "loss": 2.3347,
      "step": 278
    },
    {
      "epoch": 2.3076923076923075,
      "grad_norm": 0.42939522862434387,
      "learning_rate": 4.519874926453302e-05,
      "loss": 2.3505,
      "step": 279
    },
    {
      "epoch": 2.316008316008316,
      "grad_norm": 0.42948731780052185,
      "learning_rate": 4.419206336408418e-05,
      "loss": 2.3377,
      "step": 280
    },
    {
      "epoch": 2.3243243243243246,
      "grad_norm": 0.44148650765419006,
      "learning_rate": 4.3193525326884435e-05,
      "loss": 2.3413,
      "step": 281
    },
    {
      "epoch": 2.3326403326403327,
      "grad_norm": 0.424143522977829,
      "learning_rate": 4.220328093777851e-05,
      "loss": 2.3377,
      "step": 282
    },
    {
      "epoch": 2.3409563409563408,
      "grad_norm": 0.4698318541049957,
      "learning_rate": 4.12214747707527e-05,
      "loss": 2.3692,
      "step": 283
    },
    {
      "epoch": 2.3492723492723493,
      "grad_norm": 0.42582592368125916,
      "learning_rate": 4.0248250167827275e-05,
      "loss": 2.3572,
      "step": 284
    },
    {
      "epoch": 2.357588357588358,
      "grad_norm": 0.4405396282672882,
      "learning_rate": 3.9283749218128885e-05,
      "loss": 2.352,
      "step": 285
    },
    {
      "epoch": 2.365904365904366,
      "grad_norm": 0.4368214011192322,
      "learning_rate": 3.832811273714569e-05,
      "loss": 2.3425,
      "step": 286
    },
    {
      "epoch": 2.374220374220374,
      "grad_norm": 0.42838993668556213,
      "learning_rate": 3.738148024616863e-05,
      "loss": 2.3472,
      "step": 287
    },
    {
      "epoch": 2.3825363825363826,
      "grad_norm": 0.4351613223552704,
      "learning_rate": 3.644398995192147e-05,
      "loss": 2.3441,
      "step": 288
    },
    {
      "epoch": 2.390852390852391,
      "grad_norm": 0.4328824281692505,
      "learning_rate": 3.5515778726382966e-05,
      "loss": 2.3573,
      "step": 289
    },
    {
      "epoch": 2.399168399168399,
      "grad_norm": 0.42902839183807373,
      "learning_rate": 3.459698208680359e-05,
      "loss": 2.3672,
      "step": 290
    },
    {
      "epoch": 2.4074844074844073,
      "grad_norm": 0.44229575991630554,
      "learning_rate": 3.36877341759205e-05,
      "loss": 2.3371,
      "step": 291
    },
    {
      "epoch": 2.415800415800416,
      "grad_norm": 0.4500284492969513,
      "learning_rate": 3.2788167742372725e-05,
      "loss": 2.329,
      "step": 292
    },
    {
      "epoch": 2.4241164241164244,
      "grad_norm": 0.6344565749168396,
      "learning_rate": 3.1898414121320276e-05,
      "loss": 2.3618,
      "step": 293
    },
    {
      "epoch": 2.4324324324324325,
      "grad_norm": 0.42112040519714355,
      "learning_rate": 3.101860321526924e-05,
      "loss": 2.3349,
      "step": 294
    },
    {
      "epoch": 2.4407484407484406,
      "grad_norm": 0.49494612216949463,
      "learning_rate": 3.0148863475106314e-05,
      "loss": 2.3389,
      "step": 295
    },
    {
      "epoch": 2.449064449064449,
      "grad_norm": 0.4607335925102234,
      "learning_rate": 2.9289321881345254e-05,
      "loss": 2.3245,
      "step": 296
    },
    {
      "epoch": 2.4573804573804576,
      "grad_norm": 0.4578121602535248,
      "learning_rate": 2.84401039255879e-05,
      "loss": 2.3573,
      "step": 297
    },
    {
      "epoch": 2.4656964656964657,
      "grad_norm": 0.4579460620880127,
      "learning_rate": 2.7601333592202583e-05,
      "loss": 2.3372,
      "step": 298
    },
    {
      "epoch": 2.474012474012474,
      "grad_norm": 0.4420205354690552,
      "learning_rate": 2.677313334022268e-05,
      "loss": 2.3221,
      "step": 299
    },
    {
      "epoch": 2.4823284823284824,
      "grad_norm": 0.4563603103160858,
      "learning_rate": 2.59556240854677e-05,
      "loss": 2.3446,
      "step": 300
    },
    {
      "epoch": 2.490644490644491,
      "grad_norm": 0.44042691588401794,
      "learning_rate": 2.514892518288988e-05,
      "loss": 2.3279,
      "step": 301
    },
    {
      "epoch": 2.498960498960499,
      "grad_norm": 0.4372948408126831,
      "learning_rate": 2.4353154409148637e-05,
      "loss": 2.3629,
      "step": 302
    },
    {
      "epoch": 2.507276507276507,
      "grad_norm": 0.4709435701370239,
      "learning_rate": 2.356842794541516e-05,
      "loss": 2.353,
      "step": 303
    },
    {
      "epoch": 2.5155925155925156,
      "grad_norm": 0.42820531129837036,
      "learning_rate": 2.2794860360410342e-05,
      "loss": 2.3419,
      "step": 304
    },
    {
      "epoch": 2.523908523908524,
      "grad_norm": 0.4281867742538452,
      "learning_rate": 2.2032564593677774e-05,
      "loss": 2.3369,
      "step": 305
    },
    {
      "epoch": 2.5322245322245323,
      "grad_norm": 0.4326595664024353,
      "learning_rate": 2.1281651939094992e-05,
      "loss": 2.3215,
      "step": 306
    },
    {
      "epoch": 2.5405405405405403,
      "grad_norm": 0.4328608810901642,
      "learning_rate": 2.0542232028624586e-05,
      "loss": 2.3094,
      "step": 307
    },
    {
      "epoch": 2.548856548856549,
      "grad_norm": 0.42794665694236755,
      "learning_rate": 1.981441281630816e-05,
      "loss": 2.3303,
      "step": 308
    },
    {
      "epoch": 2.5571725571725574,
      "grad_norm": 0.43002575635910034,
      "learning_rate": 1.9098300562505266e-05,
      "loss": 2.3516,
      "step": 309
    },
    {
      "epoch": 2.5654885654885655,
      "grad_norm": 0.43052583932876587,
      "learning_rate": 1.8393999818379525e-05,
      "loss": 2.3592,
      "step": 310
    },
    {
      "epoch": 2.5738045738045736,
      "grad_norm": 0.41541633009910583,
      "learning_rate": 1.7701613410634365e-05,
      "loss": 2.333,
      "step": 311
    },
    {
      "epoch": 2.582120582120582,
      "grad_norm": 0.44588619470596313,
      "learning_rate": 1.7021242426500493e-05,
      "loss": 2.3445,
      "step": 312
    },
    {
      "epoch": 2.5904365904365907,
      "grad_norm": 0.4385538399219513,
      "learning_rate": 1.6352986198977325e-05,
      "loss": 2.3406,
      "step": 313
    },
    {
      "epoch": 2.598752598752599,
      "grad_norm": 0.43057897686958313,
      "learning_rate": 1.5696942292330576e-05,
      "loss": 2.3426,
      "step": 314
    },
    {
      "epoch": 2.607068607068607,
      "grad_norm": 0.4492436945438385,
      "learning_rate": 1.5053206487847914e-05,
      "loss": 2.33,
      "step": 315
    },
    {
      "epoch": 2.6153846153846154,
      "grad_norm": 0.43555155396461487,
      "learning_rate": 1.442187276985526e-05,
      "loss": 2.342,
      "step": 316
    },
    {
      "epoch": 2.623700623700624,
      "grad_norm": 0.4379512071609497,
      "learning_rate": 1.3803033311995072e-05,
      "loss": 2.3202,
      "step": 317
    },
    {
      "epoch": 2.632016632016632,
      "grad_norm": 0.42732083797454834,
      "learning_rate": 1.3196778463769255e-05,
      "loss": 2.3104,
      "step": 318
    },
    {
      "epoch": 2.64033264033264,
      "grad_norm": 0.43830588459968567,
      "learning_rate": 1.260319673734821e-05,
      "loss": 2.3252,
      "step": 319
    },
    {
      "epoch": 2.6486486486486487,
      "grad_norm": 0.43504393100738525,
      "learning_rate": 1.2022374794648228e-05,
      "loss": 2.3416,
      "step": 320
    },
    {
      "epoch": 2.6569646569646572,
      "grad_norm": 0.43797558546066284,
      "learning_rate": 1.1454397434679021e-05,
      "loss": 2.3361,
      "step": 321
    },
    {
      "epoch": 2.6652806652806653,
      "grad_norm": 0.4464263916015625,
      "learning_rate": 1.0899347581163221e-05,
      "loss": 2.3515,
      "step": 322
    },
    {
      "epoch": 2.6735966735966734,
      "grad_norm": 0.43725448846817017,
      "learning_rate": 1.0357306270429624e-05,
      "loss": 2.3236,
      "step": 323
    },
    {
      "epoch": 2.681912681912682,
      "grad_norm": 0.4281315803527832,
      "learning_rate": 9.828352639582072e-06,
      "loss": 2.3296,
      "step": 324
    },
    {
      "epoch": 2.6902286902286905,
      "grad_norm": 0.4342389702796936,
      "learning_rate": 9.31256391494546e-06,
      "loss": 2.3097,
      "step": 325
    },
    {
      "epoch": 2.6985446985446986,
      "grad_norm": 0.44206634163856506,
      "learning_rate": 8.810015400790994e-06,
      "loss": 2.3303,
      "step": 326
    },
    {
      "epoch": 2.7068607068607067,
      "grad_norm": 0.4294271767139435,
      "learning_rate": 8.32078046834176e-06,
      "loss": 2.3183,
      "step": 327
    },
    {
      "epoch": 2.715176715176715,
      "grad_norm": 0.43030035495758057,
      "learning_rate": 7.844930545060703e-06,
      "loss": 2.3368,
      "step": 328
    },
    {
      "epoch": 2.7234927234927238,
      "grad_norm": 0.435197651386261,
      "learning_rate": 7.382535104222366e-06,
      "loss": 2.3312,
      "step": 329
    },
    {
      "epoch": 2.731808731808732,
      "grad_norm": 0.4398477375507355,
      "learning_rate": 6.9336616547697965e-06,
      "loss": 2.3146,
      "step": 330
    },
    {
      "epoch": 2.74012474012474,
      "grad_norm": 0.43261393904685974,
      "learning_rate": 6.498375731458528e-06,
      "loss": 2.3277,
      "step": 331
    },
    {
      "epoch": 2.7484407484407485,
      "grad_norm": 0.46077948808670044,
      "learning_rate": 6.076740885288479e-06,
      "loss": 2.3493,
      "step": 332
    },
    {
      "epoch": 2.756756756756757,
      "grad_norm": 0.4285019636154175,
      "learning_rate": 5.668818674225685e-06,
      "loss": 2.3537,
      "step": 333
    },
    {
      "epoch": 2.765072765072765,
      "grad_norm": 0.43278998136520386,
      "learning_rate": 5.274668654214932e-06,
      "loss": 2.3504,
      "step": 334
    },
    {
      "epoch": 2.773388773388773,
      "grad_norm": 0.43840229511260986,
      "learning_rate": 4.8943483704846475e-06,
      "loss": 2.333,
      "step": 335
    },
    {
      "epoch": 2.7817047817047817,
      "grad_norm": 0.4274958670139313,
      "learning_rate": 4.527913349145441e-06,
      "loss": 2.3467,
      "step": 336
    },
    {
      "epoch": 2.7900207900207903,
      "grad_norm": 0.4174635410308838,
      "learning_rate": 4.175417089083378e-06,
      "loss": 2.3248,
      "step": 337
    },
    {
      "epoch": 2.7983367983367984,
      "grad_norm": 0.4231819808483124,
      "learning_rate": 3.836911054149239e-06,
      "loss": 2.3169,
      "step": 338
    },
    {
      "epoch": 2.8066528066528065,
      "grad_norm": 0.4251345098018646,
      "learning_rate": 3.512444665644865e-06,
      "loss": 2.333,
      "step": 339
    },
    {
      "epoch": 2.814968814968815,
      "grad_norm": 0.42021313309669495,
      "learning_rate": 3.202065295107726e-06,
      "loss": 2.3475,
      "step": 340
    },
    {
      "epoch": 2.8232848232848236,
      "grad_norm": 0.4197041094303131,
      "learning_rate": 2.905818257394799e-06,
      "loss": 2.3128,
      "step": 341
    },
    {
      "epoch": 2.8316008316008316,
      "grad_norm": 0.4266311228275299,
      "learning_rate": 2.6237468040666512e-06,
      "loss": 2.3249,
      "step": 342
    },
    {
      "epoch": 2.8399168399168397,
      "grad_norm": 0.442434698343277,
      "learning_rate": 2.3558921170727888e-06,
      "loss": 2.3422,
      "step": 343
    },
    {
      "epoch": 2.8482328482328483,
      "grad_norm": 0.4329405725002289,
      "learning_rate": 2.1022933027391555e-06,
      "loss": 2.3188,
      "step": 344
    },
    {
      "epoch": 2.856548856548857,
      "grad_norm": 0.43024739623069763,
      "learning_rate": 1.8629873860586566e-06,
      "loss": 2.3365,
      "step": 345
    },
    {
      "epoch": 2.864864864864865,
      "grad_norm": 0.43431299924850464,
      "learning_rate": 1.6380093052856483e-06,
      "loss": 2.3476,
      "step": 346
    },
    {
      "epoch": 2.873180873180873,
      "grad_norm": 0.43452492356300354,
      "learning_rate": 1.4273919068349184e-06,
      "loss": 2.3327,
      "step": 347
    },
    {
      "epoch": 2.8814968814968815,
      "grad_norm": 0.4286072552204132,
      "learning_rate": 1.231165940486234e-06,
      "loss": 2.334,
      "step": 348
    },
    {
      "epoch": 2.88981288981289,
      "grad_norm": 0.4313005805015564,
      "learning_rate": 1.0493600548948878e-06,
      "loss": 2.3216,
      "step": 349
    },
    {
      "epoch": 2.898128898128898,
      "grad_norm": 0.4270586371421814,
      "learning_rate": 8.820007934090879e-07,
      "loss": 2.3438,
      "step": 350
    },
    {
      "epoch": 2.9064449064449063,
      "grad_norm": 0.4391648769378662,
      "learning_rate": 7.291125901946027e-07,
      "loss": 2.3528,
      "step": 351
    },
    {
      "epoch": 2.914760914760915,
      "grad_norm": 0.4252782166004181,
      "learning_rate": 5.907177666674812e-07,
      "loss": 2.3422,
      "step": 352
    },
    {
      "epoch": 2.9230769230769234,
      "grad_norm": 0.4223313629627228,
      "learning_rate": 4.668365282351372e-07,
      "loss": 2.3233,
      "step": 353
    },
    {
      "epoch": 2.9313929313929314,
      "grad_norm": 0.43984875082969666,
      "learning_rate": 3.5748696134639825e-07,
      "loss": 2.3437,
      "step": 354
    },
    {
      "epoch": 2.9397089397089395,
      "grad_norm": 0.4258709251880646,
      "learning_rate": 2.6268503085089547e-07,
      "loss": 2.3406,
      "step": 355
    },
    {
      "epoch": 2.948024948024948,
      "grad_norm": 0.42968347668647766,
      "learning_rate": 1.824445776682504e-07,
      "loss": 2.3277,
      "step": 356
    },
    {
      "epoch": 2.9563409563409566,
      "grad_norm": 0.4366348683834076,
      "learning_rate": 1.1677731676733584e-07,
      "loss": 2.3263,
      "step": 357
    },
    {
      "epoch": 2.9646569646569647,
      "grad_norm": 0.42044591903686523,
      "learning_rate": 6.569283545587724e-08,
      "loss": 2.3439,
      "step": 358
    },
    {
      "epoch": 2.972972972972973,
      "grad_norm": 0.43199941515922546,
      "learning_rate": 2.9198591980705848e-08,
      "loss": 2.3411,
      "step": 359
    },
    {
      "epoch": 2.9812889812889813,
      "grad_norm": 0.4222280979156494,
      "learning_rate": 7.2999144389296335e-09,
      "loss": 2.3358,
      "step": 360
    }
  ],
  "logging_steps": 1,
  "max_steps": 360,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 200,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 3.2413772944716595e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}