{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9914794732765299,
  "eval_steps": 500,
  "global_step": 3200,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "grad_norm": 12.876287460327148,
      "learning_rate": 1.4099996659119809e-05,
      "loss": 4.3352,
      "step": 1
    },
    {
      "epoch": 0.0,
      "grad_norm": 14.893763542175293,
      "learning_rate": 1.4099986636482399e-05,
      "loss": 3.9062,
      "step": 2
    },
    {
      "epoch": 0.0,
      "grad_norm": 14.747922897338867,
      "learning_rate": 1.409996993209727e-05,
      "loss": 2.1353,
      "step": 3
    },
    {
      "epoch": 0.0,
      "grad_norm": 9.3056640625,
      "learning_rate": 1.4099946545980254e-05,
      "loss": 1.2115,
      "step": 4
    },
    {
      "epoch": 0.0,
      "grad_norm": 16.262678146362305,
      "learning_rate": 1.4099916478153516e-05,
      "loss": 2.1268,
      "step": 5
    },
    {
      "epoch": 0.0,
      "grad_norm": 24.459659576416016,
      "learning_rate": 1.4099879728645554e-05,
      "loss": 2.7806,
      "step": 6
    },
    {
      "epoch": 0.0,
      "grad_norm": 13.100327491760254,
      "learning_rate": 1.4099836297491197e-05,
      "loss": 2.453,
      "step": 7
    },
    {
      "epoch": 0.0,
      "grad_norm": 10.096444129943848,
      "learning_rate": 1.4099786184731606e-05,
      "loss": 2.2396,
      "step": 8
    },
    {
      "epoch": 0.0,
      "grad_norm": 12.478560447692871,
      "learning_rate": 1.409972939041428e-05,
      "loss": 1.4557,
      "step": 9
    },
    {
      "epoch": 0.0,
      "grad_norm": 11.49439525604248,
      "learning_rate": 1.4099665914593043e-05,
      "loss": 1.7847,
      "step": 10
    },
    {
      "epoch": 0.0,
      "grad_norm": 20.96303939819336,
      "learning_rate": 1.4099595757328058e-05,
      "loss": 2.6818,
      "step": 11
    },
    {
      "epoch": 0.0,
      "grad_norm": 14.429734230041504,
      "learning_rate": 1.4099518918685816e-05,
      "loss": 3.0543,
      "step": 12
    },
    {
      "epoch": 0.0,
      "grad_norm": 13.766142845153809,
      "learning_rate": 1.4099435398739144e-05,
      "loss": 2.4683,
      "step": 13
    },
    {
      "epoch": 0.0,
      "grad_norm": 19.67556381225586,
      "learning_rate": 1.4099345197567196e-05,
      "loss": 6.3905,
      "step": 14
    },
    {
      "epoch": 0.0,
      "grad_norm": 11.328873634338379,
      "learning_rate": 1.4099248315255468e-05,
      "loss": 1.104,
      "step": 15
    },
    {
      "epoch": 0.0,
      "grad_norm": 12.752955436706543,
      "learning_rate": 1.4099144751895776e-05,
      "loss": 3.7395,
      "step": 16
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.726831436157227,
      "learning_rate": 1.4099034507586277e-05,
      "loss": 5.7476,
      "step": 17
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.659893035888672,
      "learning_rate": 1.4098917582431455e-05,
      "loss": 2.6832,
      "step": 18
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.84981918334961,
      "learning_rate": 1.4098793976542129e-05,
      "loss": 1.827,
      "step": 19
    },
    {
      "epoch": 0.01,
      "grad_norm": 21.144001007080078,
      "learning_rate": 1.409866369003545e-05,
      "loss": 5.6839,
      "step": 20
    },
    {
      "epoch": 0.01,
      "grad_norm": 16.51398468017578,
      "learning_rate": 1.4098526723034896e-05,
      "loss": 7.7836,
      "step": 21
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.258636474609375,
      "learning_rate": 1.4098383075670282e-05,
      "loss": 2.5929,
      "step": 22
    },
    {
      "epoch": 0.01,
      "grad_norm": 15.997798919677734,
      "learning_rate": 1.4098232748077751e-05,
      "loss": 3.2264,
      "step": 23
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.296121597290039,
      "learning_rate": 1.4098075740399781e-05,
      "loss": 1.9712,
      "step": 24
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.37089729309082,
      "learning_rate": 1.409791205278518e-05,
      "loss": 1.8801,
      "step": 25
    },
    {
      "epoch": 0.01,
      "grad_norm": 19.558774948120117,
      "learning_rate": 1.4097741685389079e-05,
      "loss": 3.0265,
      "step": 26
    },
    {
      "epoch": 0.01,
      "grad_norm": 9.684329986572266,
      "learning_rate": 1.4097564638372954e-05,
      "loss": 0.8368,
      "step": 27
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.683094024658203,
      "learning_rate": 1.4097380911904599e-05,
      "loss": 1.2745,
      "step": 28
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.189740180969238,
      "learning_rate": 1.4097190506158147e-05,
      "loss": 1.1601,
      "step": 29
    },
    {
      "epoch": 0.01,
      "grad_norm": 18.130149841308594,
      "learning_rate": 1.4096993421314056e-05,
      "loss": 1.8264,
      "step": 30
    },
    {
      "epoch": 0.01,
      "grad_norm": 19.705230712890625,
      "learning_rate": 1.409678965755912e-05,
      "loss": 1.8733,
      "step": 31
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.09542465209961,
      "learning_rate": 1.4096579215086455e-05,
      "loss": 2.4501,
      "step": 32
    },
    {
      "epoch": 0.01,
      "grad_norm": 19.853260040283203,
      "learning_rate": 1.4096362094095518e-05,
      "loss": 2.7891,
      "step": 33
    },
    {
      "epoch": 0.01,
      "grad_norm": 14.29506778717041,
      "learning_rate": 1.4096138294792083e-05,
      "loss": 3.4722,
      "step": 34
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.516790390014648,
      "learning_rate": 1.4095907817388264e-05,
      "loss": 2.1471,
      "step": 35
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.4302978515625,
      "learning_rate": 1.4095670662102496e-05,
      "loss": 2.6135,
      "step": 36
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.625223159790039,
      "learning_rate": 1.4095426829159551e-05,
      "loss": 1.1707,
      "step": 37
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.80229949951172,
      "learning_rate": 1.4095176318790526e-05,
      "loss": 3.7891,
      "step": 38
    },
    {
      "epoch": 0.01,
      "grad_norm": 8.8538179397583,
      "learning_rate": 1.409491913123284e-05,
      "loss": 0.9273,
      "step": 39
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.519434928894043,
      "learning_rate": 1.4094655266730254e-05,
      "loss": 2.6715,
      "step": 40
    },
    {
      "epoch": 0.01,
      "grad_norm": 10.914387702941895,
      "learning_rate": 1.4094384725532849e-05,
      "loss": 1.5982,
      "step": 41
    },
    {
      "epoch": 0.01,
      "grad_norm": 17.040918350219727,
      "learning_rate": 1.4094107507897032e-05,
      "loss": 3.0822,
      "step": 42
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.09429931640625,
      "learning_rate": 1.4093823614085546e-05,
      "loss": 1.5496,
      "step": 43
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.157367706298828,
      "learning_rate": 1.409353304436745e-05,
      "loss": 1.4657,
      "step": 44
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.823859214782715,
      "learning_rate": 1.4093235799018141e-05,
      "loss": 1.637,
      "step": 45
    },
    {
      "epoch": 0.01,
      "grad_norm": 11.13479232788086,
      "learning_rate": 1.4092931878319336e-05,
      "loss": 2.6957,
      "step": 46
    },
    {
      "epoch": 0.01,
      "grad_norm": 12.315866470336914,
      "learning_rate": 1.4092621282559084e-05,
      "loss": 2.2981,
      "step": 47
    },
    {
      "epoch": 0.01,
      "grad_norm": 13.145035743713379,
      "learning_rate": 1.4092304012031754e-05,
      "loss": 2.8959,
      "step": 48
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.662394523620605,
      "learning_rate": 1.4091980067038048e-05,
      "loss": 4.7688,
      "step": 49
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.96580982208252,
      "learning_rate": 1.4091649447884988e-05,
      "loss": 2.0613,
      "step": 50
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.999765396118164,
      "learning_rate": 1.4091312154885925e-05,
      "loss": 2.4087,
      "step": 51
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.088006019592285,
      "learning_rate": 1.4090968188360538e-05,
      "loss": 1.6578,
      "step": 52
    },
    {
      "epoch": 0.02,
      "grad_norm": 17.08218002319336,
      "learning_rate": 1.4090617548634819e-05,
      "loss": 1.2898,
      "step": 53
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.499032974243164,
      "learning_rate": 1.40902602360411e-05,
      "loss": 1.1168,
      "step": 54
    },
    {
      "epoch": 0.02,
      "grad_norm": 9.868734359741211,
      "learning_rate": 1.4089896250918028e-05,
      "loss": 1.3226,
      "step": 55
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.432489395141602,
      "learning_rate": 1.4089525593610576e-05,
      "loss": 3.777,
      "step": 56
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.179858207702637,
      "learning_rate": 1.4089148264470039e-05,
      "loss": 1.6965,
      "step": 57
    },
    {
      "epoch": 0.02,
      "grad_norm": 16.56346893310547,
      "learning_rate": 1.408876426385404e-05,
      "loss": 2.4936,
      "step": 58
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.435903549194336,
      "learning_rate": 1.4088373592126525e-05,
      "loss": 3.2397,
      "step": 59
    },
    {
      "epoch": 0.02,
      "grad_norm": 17.36369514465332,
      "learning_rate": 1.4087976249657753e-05,
      "loss": 2.8435,
      "step": 60
    },
    {
      "epoch": 0.02,
      "grad_norm": 17.762537002563477,
      "learning_rate": 1.4087572236824316e-05,
      "loss": 5.0264,
      "step": 61
    },
    {
      "epoch": 0.02,
      "grad_norm": 16.8850040435791,
      "learning_rate": 1.4087161554009124e-05,
      "loss": 3.2372,
      "step": 62
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.00660514831543,
      "learning_rate": 1.4086744201601408e-05,
      "loss": 3.6194,
      "step": 63
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.710498809814453,
      "learning_rate": 1.4086320179996723e-05,
      "loss": 3.4525,
      "step": 64
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.125383377075195,
      "learning_rate": 1.4085889489596941e-05,
      "loss": 1.8647,
      "step": 65
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.840350151062012,
      "learning_rate": 1.4085452130810259e-05,
      "loss": 1.2356,
      "step": 66
    },
    {
      "epoch": 0.02,
      "grad_norm": 19.095186233520508,
      "learning_rate": 1.4085008104051187e-05,
      "loss": 5.6668,
      "step": 67
    },
    {
      "epoch": 0.02,
      "grad_norm": 28.46858024597168,
      "learning_rate": 1.4084557409740563e-05,
      "loss": 3.5707,
      "step": 68
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.306602478027344,
      "learning_rate": 1.408410004830554e-05,
      "loss": 2.8939,
      "step": 69
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.545845031738281,
      "learning_rate": 1.4083636020179586e-05,
      "loss": 1.1766,
      "step": 70
    },
    {
      "epoch": 0.02,
      "grad_norm": 16.976905822753906,
      "learning_rate": 1.4083165325802496e-05,
      "loss": 2.7053,
      "step": 71
    },
    {
      "epoch": 0.02,
      "grad_norm": 15.33132553100586,
      "learning_rate": 1.4082687965620378e-05,
      "loss": 1.5187,
      "step": 72
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.40718936920166,
      "learning_rate": 1.4082203940085658e-05,
      "loss": 4.6465,
      "step": 73
    },
    {
      "epoch": 0.02,
      "grad_norm": 17.957759857177734,
      "learning_rate": 1.408171324965708e-05,
      "loss": 5.891,
      "step": 74
    },
    {
      "epoch": 0.02,
      "grad_norm": 11.309581756591797,
      "learning_rate": 1.4081215894799704e-05,
      "loss": 1.3456,
      "step": 75
    },
    {
      "epoch": 0.02,
      "grad_norm": 16.916112899780273,
      "learning_rate": 1.4080711875984907e-05,
      "loss": 7.3883,
      "step": 76
    },
    {
      "epoch": 0.02,
      "grad_norm": 17.052242279052734,
      "learning_rate": 1.408020119369038e-05,
      "loss": 4.8764,
      "step": 77
    },
    {
      "epoch": 0.02,
      "grad_norm": 13.979994773864746,
      "learning_rate": 1.4079683848400135e-05,
      "loss": 2.1218,
      "step": 78
    },
    {
      "epoch": 0.02,
      "grad_norm": 12.151473999023438,
      "learning_rate": 1.407915984060449e-05,
      "loss": 1.468,
      "step": 79
    },
    {
      "epoch": 0.02,
      "grad_norm": 14.50087833404541,
      "learning_rate": 1.4078629170800088e-05,
      "loss": 2.8174,
      "step": 80
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.1065092086792,
      "learning_rate": 1.4078091839489877e-05,
      "loss": 2.0711,
      "step": 81
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.599384307861328,
      "learning_rate": 1.4077547847183122e-05,
      "loss": 1.5431,
      "step": 82
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.681108474731445,
      "learning_rate": 1.40769971943954e-05,
      "loss": 4.8315,
      "step": 83
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.349267959594727,
      "learning_rate": 1.4076439881648607e-05,
      "loss": 5.1925,
      "step": 84
    },
    {
      "epoch": 0.03,
      "grad_norm": 22.059520721435547,
      "learning_rate": 1.4075875909470941e-05,
      "loss": 3.7472,
      "step": 85
    },
    {
      "epoch": 0.03,
      "grad_norm": 13.839190483093262,
      "learning_rate": 1.407530527839692e-05,
      "loss": 5.2631,
      "step": 86
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.494612693786621,
      "learning_rate": 1.4074727988967365e-05,
      "loss": 2.383,
      "step": 87
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.313505172729492,
      "learning_rate": 1.4074144041729414e-05,
      "loss": 1.9434,
      "step": 88
    },
    {
      "epoch": 0.03,
      "grad_norm": 27.630678176879883,
      "learning_rate": 1.4073553437236517e-05,
      "loss": 3.9264,
      "step": 89
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.239396095275879,
      "learning_rate": 1.4072956176048426e-05,
      "loss": 2.1466,
      "step": 90
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.377116203308105,
      "learning_rate": 1.4072352258731206e-05,
      "loss": 1.8366,
      "step": 91
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.560379981994629,
      "learning_rate": 1.407174168585723e-05,
      "loss": 1.4054,
      "step": 92
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.187500953674316,
      "learning_rate": 1.4071124458005179e-05,
      "loss": 2.5571,
      "step": 93
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.817543029785156,
      "learning_rate": 1.4070500575760041e-05,
      "loss": 1.2708,
      "step": 94
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.608696937561035,
      "learning_rate": 1.4069870039713115e-05,
      "loss": 2.3783,
      "step": 95
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.780579566955566,
      "learning_rate": 1.4069232850461997e-05,
      "loss": 5.6665,
      "step": 96
    },
    {
      "epoch": 0.03,
      "grad_norm": 8.685426712036133,
      "learning_rate": 1.40685890086106e-05,
      "loss": 0.9811,
      "step": 97
    },
    {
      "epoch": 0.03,
      "grad_norm": 22.659534454345703,
      "learning_rate": 1.4067938514769132e-05,
      "loss": 8.1078,
      "step": 98
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.488410949707031,
      "learning_rate": 1.4067281369554112e-05,
      "loss": 2.6526,
      "step": 99
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.394866943359375,
      "learning_rate": 1.4066617573588359e-05,
      "loss": 1.8688,
      "step": 100
    },
    {
      "epoch": 0.03,
      "grad_norm": 18.25127410888672,
      "learning_rate": 1.4065947127500999e-05,
      "loss": 5.0283,
      "step": 101
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.696410179138184,
      "learning_rate": 1.4065270031927457e-05,
      "loss": 4.6801,
      "step": 102
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.974501609802246,
      "learning_rate": 1.4064586287509465e-05,
      "loss": 4.2069,
      "step": 103
    },
    {
      "epoch": 0.03,
      "grad_norm": 20.77216339111328,
      "learning_rate": 1.4063895894895051e-05,
      "loss": 6.9798,
      "step": 104
    },
    {
      "epoch": 0.03,
      "grad_norm": 16.946361541748047,
      "learning_rate": 1.4063198854738549e-05,
      "loss": 4.9877,
      "step": 105
    },
    {
      "epoch": 0.03,
      "grad_norm": 20.423288345336914,
      "learning_rate": 1.4062495167700588e-05,
      "loss": 5.478,
      "step": 106
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.431811332702637,
      "learning_rate": 1.4061784834448102e-05,
      "loss": 5.1891,
      "step": 107
    },
    {
      "epoch": 0.03,
      "grad_norm": 12.419416427612305,
      "learning_rate": 1.4061067855654323e-05,
      "loss": 1.5543,
      "step": 108
    },
    {
      "epoch": 0.03,
      "grad_norm": 15.559722900390625,
      "learning_rate": 1.4060344231998774e-05,
      "loss": 2.5583,
      "step": 109
    },
    {
      "epoch": 0.03,
      "grad_norm": 11.121515274047852,
      "learning_rate": 1.405961396416729e-05,
      "loss": 1.8617,
      "step": 110
    },
    {
      "epoch": 0.03,
      "grad_norm": 13.224935531616211,
      "learning_rate": 1.4058877052851988e-05,
      "loss": 2.214,
      "step": 111
    },
    {
      "epoch": 0.03,
      "grad_norm": 10.346721649169922,
      "learning_rate": 1.4058133498751294e-05,
      "loss": 1.0174,
      "step": 112
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.830920219421387,
      "learning_rate": 1.405738330256992e-05,
      "loss": 2.2093,
      "step": 113
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.912773132324219,
      "learning_rate": 1.4056626465018882e-05,
      "loss": 1.6604,
      "step": 114
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.618285179138184,
      "learning_rate": 1.4055862986815482e-05,
      "loss": 2.3275,
      "step": 115
    },
    {
      "epoch": 0.04,
      "grad_norm": 17.236412048339844,
      "learning_rate": 1.4055092868683323e-05,
      "loss": 5.0808,
      "step": 116
    },
    {
      "epoch": 0.04,
      "grad_norm": 19.668128967285156,
      "learning_rate": 1.4054316111352295e-05,
      "loss": 3.5998,
      "step": 117
    },
    {
      "epoch": 0.04,
      "grad_norm": 24.973814010620117,
      "learning_rate": 1.4053532715558585e-05,
      "loss": 2.1121,
      "step": 118
    },
    {
      "epoch": 0.04,
      "grad_norm": 12.31823444366455,
      "learning_rate": 1.405274268204467e-05,
      "loss": 1.8002,
      "step": 119
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.556957244873047,
      "learning_rate": 1.4051946011559316e-05,
      "loss": 1.3042,
      "step": 120
    },
    {
      "epoch": 0.04,
      "grad_norm": 17.19139862060547,
      "learning_rate": 1.4051142704857584e-05,
      "loss": 4.2733,
      "step": 121
    },
    {
      "epoch": 0.04,
      "grad_norm": 22.5864315032959,
      "learning_rate": 1.405033276270082e-05,
      "loss": 5.6605,
      "step": 122
    },
    {
      "epoch": 0.04,
      "grad_norm": 18.1478328704834,
      "learning_rate": 1.4049516185856663e-05,
      "loss": 2.2058,
      "step": 123
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.945579528808594,
      "learning_rate": 1.4048692975099036e-05,
      "loss": 1.4621,
      "step": 124
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.165715217590332,
      "learning_rate": 1.4047863131208152e-05,
      "loss": 2.9207,
      "step": 125
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.359014511108398,
      "learning_rate": 1.4047026654970509e-05,
      "loss": 1.7454,
      "step": 126
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.294341087341309,
      "learning_rate": 1.4046183547178893e-05,
      "loss": 1.0382,
      "step": 127
    },
    {
      "epoch": 0.04,
      "grad_norm": 19.4971866607666,
      "learning_rate": 1.4045333808632374e-05,
      "loss": 4.2891,
      "step": 128
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.049226760864258,
      "learning_rate": 1.4044477440136308e-05,
      "loss": 0.9562,
      "step": 129
    },
    {
      "epoch": 0.04,
      "grad_norm": 9.53789234161377,
      "learning_rate": 1.4043614442502331e-05,
      "loss": 1.303,
      "step": 130
    },
    {
      "epoch": 0.04,
      "grad_norm": 8.629888534545898,
      "learning_rate": 1.4042744816548363e-05,
      "loss": 1.0109,
      "step": 131
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.556740760803223,
      "learning_rate": 1.4041868563098612e-05,
      "loss": 1.2822,
      "step": 132
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.855392456054688,
      "learning_rate": 1.4040985682983557e-05,
      "loss": 2.5222,
      "step": 133
    },
    {
      "epoch": 0.04,
      "grad_norm": 20.333721160888672,
      "learning_rate": 1.4040096177039966e-05,
      "loss": 3.1756,
      "step": 134
    },
    {
      "epoch": 0.04,
      "grad_norm": 19.936588287353516,
      "learning_rate": 1.4039200046110882e-05,
      "loss": 2.2128,
      "step": 135
    },
    {
      "epoch": 0.04,
      "grad_norm": 11.913959503173828,
      "learning_rate": 1.4038297291045633e-05,
      "loss": 1.2175,
      "step": 136
    },
    {
      "epoch": 0.04,
      "grad_norm": 16.783187866210938,
      "learning_rate": 1.4037387912699815e-05,
      "loss": 2.6931,
      "step": 137
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.196114540100098,
      "learning_rate": 1.4036471911935311e-05,
      "loss": 0.658,
      "step": 138
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.82783031463623,
      "learning_rate": 1.4035549289620276e-05,
      "loss": 3.1209,
      "step": 139
    },
    {
      "epoch": 0.04,
      "grad_norm": 19.238548278808594,
      "learning_rate": 1.403462004662914e-05,
      "loss": 2.9366,
      "step": 140
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.233245849609375,
      "learning_rate": 1.4033684183842608e-05,
      "loss": 0.8557,
      "step": 141
    },
    {
      "epoch": 0.04,
      "grad_norm": 13.334661483764648,
      "learning_rate": 1.4032741702147666e-05,
      "loss": 1.3891,
      "step": 142
    },
    {
      "epoch": 0.04,
      "grad_norm": 15.853718757629395,
      "learning_rate": 1.4031792602437564e-05,
      "loss": 2.0343,
      "step": 143
    },
    {
      "epoch": 0.04,
      "grad_norm": 10.005762100219727,
      "learning_rate": 1.4030836885611825e-05,
      "loss": 1.4376,
      "step": 144
    },
    {
      "epoch": 0.04,
      "grad_norm": 14.192896842956543,
      "learning_rate": 1.4029874552576251e-05,
      "loss": 2.3502,
      "step": 145
    },
    {
      "epoch": 0.05,
      "grad_norm": 16.873315811157227,
      "learning_rate": 1.4028905604242907e-05,
      "loss": 5.7526,
      "step": 146
    },
    {
      "epoch": 0.05,
      "grad_norm": 15.167844772338867,
      "learning_rate": 1.402793004153013e-05,
      "loss": 0.8518,
      "step": 147
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.539708137512207,
      "learning_rate": 1.402694786536253e-05,
      "loss": 4.8231,
      "step": 148
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.99748420715332,
      "learning_rate": 1.4025959076670982e-05,
      "loss": 1.6392,
      "step": 149
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.567449569702148,
      "learning_rate": 1.4024963676392621e-05,
      "loss": 2.9904,
      "step": 150
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.029122352600098,
      "learning_rate": 1.4023961665470862e-05,
      "loss": 2.3267,
      "step": 151
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.277032852172852,
      "learning_rate": 1.4022953044855374e-05,
      "loss": 3.3196,
      "step": 152
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.015501976013184,
      "learning_rate": 1.40219378155021e-05,
      "loss": 2.5454,
      "step": 153
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.410118103027344,
      "learning_rate": 1.4020915978373233e-05,
      "loss": 1.3893,
      "step": 154
    },
    {
      "epoch": 0.05,
      "grad_norm": 17.930435180664062,
      "learning_rate": 1.4019887534437244e-05,
      "loss": 1.8514,
      "step": 155
    },
    {
      "epoch": 0.05,
      "grad_norm": 16.743331909179688,
      "learning_rate": 1.4018852484668857e-05,
      "loss": 2.6936,
      "step": 156
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.879019737243652,
      "learning_rate": 1.4017810830049058e-05,
      "loss": 1.5375,
      "step": 157
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.569742202758789,
      "learning_rate": 1.4016762571565094e-05,
      "loss": 3.0512,
      "step": 158
    },
    {
      "epoch": 0.05,
      "grad_norm": 9.77190113067627,
      "learning_rate": 1.4015707710210466e-05,
      "loss": 1.0528,
      "step": 159
    },
    {
      "epoch": 0.05,
      "grad_norm": 17.27251434326172,
      "learning_rate": 1.4014646246984943e-05,
      "loss": 2.5514,
      "step": 160
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.10989761352539,
      "learning_rate": 1.4013578182894543e-05,
      "loss": 1.5854,
      "step": 161
    },
    {
      "epoch": 0.05,
      "grad_norm": 20.85812759399414,
      "learning_rate": 1.4012503518951542e-05,
      "loss": 1.6049,
      "step": 162
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.70592975616455,
      "learning_rate": 1.4011422256174472e-05,
      "loss": 2.8355,
      "step": 163
    },
    {
      "epoch": 0.05,
      "grad_norm": 14.205408096313477,
      "learning_rate": 1.4010334395588117e-05,
      "loss": 2.7484,
      "step": 164
    },
    {
      "epoch": 0.05,
      "grad_norm": 19.103593826293945,
      "learning_rate": 1.400923993822352e-05,
      "loss": 1.7546,
      "step": 165
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.702698707580566,
      "learning_rate": 1.4008138885117966e-05,
      "loss": 1.2193,
      "step": 166
    },
    {
      "epoch": 0.05,
      "grad_norm": 18.2764835357666,
      "learning_rate": 1.4007031237315003e-05,
      "loss": 2.5745,
      "step": 167
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.94648551940918,
      "learning_rate": 1.400591699586442e-05,
      "loss": 1.5701,
      "step": 168
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.69957160949707,
      "learning_rate": 1.400479616182226e-05,
      "loss": 1.8477,
      "step": 169
    },
    {
      "epoch": 0.05,
      "grad_norm": 23.002796173095703,
      "learning_rate": 1.4003668736250814e-05,
      "loss": 2.777,
      "step": 170
    },
    {
      "epoch": 0.05,
      "grad_norm": 12.981588363647461,
      "learning_rate": 1.4002534720218615e-05,
      "loss": 5.5741,
      "step": 171
    },
    {
      "epoch": 0.05,
      "grad_norm": 16.71035385131836,
      "learning_rate": 1.4001394114800454e-05,
      "loss": 4.5123,
      "step": 172
    },
    {
      "epoch": 0.05,
      "grad_norm": 13.110565185546875,
      "learning_rate": 1.4000246921077352e-05,
      "loss": 2.8851,
      "step": 173
    },
    {
      "epoch": 0.05,
      "grad_norm": 11.49026107788086,
      "learning_rate": 1.3999093140136585e-05,
      "loss": 3.6336,
      "step": 174
    },
    {
      "epoch": 0.05,
      "grad_norm": 10.91289234161377,
      "learning_rate": 1.3997932773071671e-05,
      "loss": 1.3063,
      "step": 175
    },
    {
      "epoch": 0.05,
      "grad_norm": 17.452695846557617,
      "learning_rate": 1.3996765820982366e-05,
      "loss": 3.5251,
      "step": 176
    },
    {
      "epoch": 0.05,
      "grad_norm": 18.41887855529785,
      "learning_rate": 1.3995592284974667e-05,
      "loss": 1.9724,
      "step": 177
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.130496978759766,
      "learning_rate": 1.3994412166160817e-05,
      "loss": 2.8272,
      "step": 178
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.504462242126465,
      "learning_rate": 1.3993225465659294e-05,
      "loss": 2.5295,
      "step": 179
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.741791725158691,
      "learning_rate": 1.3992032184594812e-05,
      "loss": 0.7621,
      "step": 180
    },
    {
      "epoch": 0.06,
      "grad_norm": 13.722132682800293,
      "learning_rate": 1.3990832324098323e-05,
      "loss": 2.4034,
      "step": 181
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.14046573638916,
      "learning_rate": 1.3989625885307021e-05,
      "loss": 2.0134,
      "step": 182
    },
    {
      "epoch": 0.06,
      "grad_norm": 15.722448348999023,
      "learning_rate": 1.3988412869364323e-05,
      "loss": 5.0702,
      "step": 183
    },
    {
      "epoch": 0.06,
      "grad_norm": 7.00212287902832,
      "learning_rate": 1.398719327741989e-05,
      "loss": 0.8542,
      "step": 184
    },
    {
      "epoch": 0.06,
      "grad_norm": 13.306962013244629,
      "learning_rate": 1.398596711062961e-05,
      "loss": 2.0083,
      "step": 185
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.195137977600098,
      "learning_rate": 1.3984734370155603e-05,
      "loss": 1.4735,
      "step": 186
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.470940589904785,
      "learning_rate": 1.3983495057166222e-05,
      "loss": 5.086,
      "step": 187
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.928512573242188,
      "learning_rate": 1.398224917283605e-05,
      "loss": 2.2687,
      "step": 188
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.653017044067383,
      "learning_rate": 1.398099671834589e-05,
      "loss": 5.4149,
      "step": 189
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.542997360229492,
      "learning_rate": 1.3979737694882781e-05,
      "loss": 1.5893,
      "step": 190
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.671969413757324,
      "learning_rate": 1.3978472103639984e-05,
      "loss": 2.5069,
      "step": 191
    },
    {
      "epoch": 0.06,
      "grad_norm": 18.104692459106445,
      "learning_rate": 1.3977199945816983e-05,
      "loss": 2.2704,
      "step": 192
    },
    {
      "epoch": 0.06,
      "grad_norm": 16.34786605834961,
      "learning_rate": 1.397592122261949e-05,
      "loss": 4.7383,
      "step": 193
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.198836326599121,
      "learning_rate": 1.3974635935259439e-05,
      "loss": 2.0531,
      "step": 194
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.655235290527344,
      "learning_rate": 1.3973344084954981e-05,
      "loss": 1.5275,
      "step": 195
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.924242973327637,
      "learning_rate": 1.3972045672930486e-05,
      "loss": 4.6574,
      "step": 196
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.535531997680664,
      "learning_rate": 1.3970740700416554e-05,
      "loss": 5.0241,
      "step": 197
    },
    {
      "epoch": 0.06,
      "grad_norm": 13.114269256591797,
      "learning_rate": 1.3969429168649992e-05,
      "loss": 5.2485,
      "step": 198
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.427637100219727,
      "learning_rate": 1.3968111078873827e-05,
      "loss": 2.0893,
      "step": 199
    },
    {
      "epoch": 0.06,
      "grad_norm": 9.180805206298828,
      "learning_rate": 1.39667864323373e-05,
      "loss": 0.7392,
      "step": 200
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.39584732055664,
      "learning_rate": 1.3965455230295872e-05,
      "loss": 3.086,
      "step": 201
    },
    {
      "epoch": 0.06,
      "grad_norm": 14.689746856689453,
      "learning_rate": 1.3964117474011208e-05,
      "loss": 1.6546,
      "step": 202
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.625497817993164,
      "learning_rate": 1.3962773164751195e-05,
      "loss": 2.9516,
      "step": 203
    },
    {
      "epoch": 0.06,
      "grad_norm": 10.085124969482422,
      "learning_rate": 1.396142230378992e-05,
      "loss": 1.0635,
      "step": 204
    },
    {
      "epoch": 0.06,
      "grad_norm": 12.808405876159668,
      "learning_rate": 1.396006489240769e-05,
      "loss": 5.0727,
      "step": 205
    },
    {
      "epoch": 0.06,
      "grad_norm": 17.288103103637695,
      "learning_rate": 1.3958700931891013e-05,
      "loss": 3.326,
      "step": 206
    },
    {
      "epoch": 0.06,
      "grad_norm": 15.653545379638672,
      "learning_rate": 1.3957330423532604e-05,
      "loss": 1.8213,
      "step": 207
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.816011428833008,
      "learning_rate": 1.3955953368631391e-05,
      "loss": 1.4498,
      "step": 208
    },
    {
      "epoch": 0.06,
      "grad_norm": 11.988968849182129,
      "learning_rate": 1.3954569768492497e-05,
      "loss": 1.7693,
      "step": 209
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.014742851257324,
      "learning_rate": 1.3953179624427257e-05,
      "loss": 2.9494,
      "step": 210
    },
    {
      "epoch": 0.07,
      "grad_norm": 15.824225425720215,
      "learning_rate": 1.3951782937753205e-05,
      "loss": 4.9419,
      "step": 211
    },
    {
      "epoch": 0.07,
      "grad_norm": 18.75404167175293,
      "learning_rate": 1.395037970979407e-05,
      "loss": 4.0696,
      "step": 212
    },
    {
      "epoch": 0.07,
      "grad_norm": 22.080371856689453,
      "learning_rate": 1.394896994187979e-05,
      "loss": 4.0442,
      "step": 213
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.431167602539062,
      "learning_rate": 1.3947553635346495e-05,
      "loss": 1.5332,
      "step": 214
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.02080249786377,
      "learning_rate": 1.3946130791536518e-05,
      "loss": 2.3295,
      "step": 215
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.056671142578125,
      "learning_rate": 1.394470141179838e-05,
      "loss": 5.4534,
      "step": 216
    },
    {
      "epoch": 0.07,
      "grad_norm": 6.8713698387146,
      "learning_rate": 1.3943265497486799e-05,
      "loss": 1.199,
      "step": 217
    },
    {
      "epoch": 0.07,
      "grad_norm": 18.43778419494629,
      "learning_rate": 1.3941823049962691e-05,
      "loss": 5.0473,
      "step": 218
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.254149436950684,
      "learning_rate": 1.3940374070593162e-05,
      "loss": 2.1543,
      "step": 219
    },
    {
      "epoch": 0.07,
      "grad_norm": 19.002323150634766,
      "learning_rate": 1.3938918560751505e-05,
      "loss": 2.6575,
      "step": 220
    },
    {
      "epoch": 0.07,
      "grad_norm": 16.23662757873535,
      "learning_rate": 1.3937456521817205e-05,
      "loss": 6.1765,
      "step": 221
    },
    {
      "epoch": 0.07,
      "grad_norm": 17.02012825012207,
      "learning_rate": 1.3935987955175937e-05,
      "loss": 6.0835,
      "step": 222
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.135772705078125,
      "learning_rate": 1.3934512862219556e-05,
      "loss": 1.0666,
      "step": 223
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.325325965881348,
      "learning_rate": 1.3933031244346113e-05,
      "loss": 3.2572,
      "step": 224
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.025094985961914,
      "learning_rate": 1.3931543102959834e-05,
      "loss": 1.2652,
      "step": 225
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.37082290649414,
      "learning_rate": 1.393004843947113e-05,
      "loss": 2.2742,
      "step": 226
    },
    {
      "epoch": 0.07,
      "grad_norm": 10.832401275634766,
      "learning_rate": 1.3928547255296597e-05,
      "loss": 1.2151,
      "step": 227
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.949665069580078,
      "learning_rate": 1.3927039551859006e-05,
      "loss": 2.5363,
      "step": 228
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.304301261901855,
      "learning_rate": 1.3925525330587312e-05,
      "loss": 5.016,
      "step": 229
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.987974166870117,
      "learning_rate": 1.3924004592916641e-05,
      "loss": 5.0027,
      "step": 230
    },
    {
      "epoch": 0.07,
      "grad_norm": 18.76968002319336,
      "learning_rate": 1.39224773402883e-05,
      "loss": 2.382,
      "step": 231
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.265201568603516,
      "learning_rate": 1.392094357414977e-05,
      "loss": 1.8023,
      "step": 232
    },
    {
      "epoch": 0.07,
      "grad_norm": 9.563916206359863,
      "learning_rate": 1.3919403295954707e-05,
      "loss": 1.0785,
      "step": 233
    },
    {
      "epoch": 0.07,
      "grad_norm": 11.332926750183105,
      "learning_rate": 1.3917856507162931e-05,
      "loss": 1.4225,
      "step": 234
    },
    {
      "epoch": 0.07,
      "grad_norm": 15.125313758850098,
      "learning_rate": 1.3916303209240444e-05,
      "loss": 1.3242,
      "step": 235
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.948397636413574,
      "learning_rate": 1.3914743403659406e-05,
      "loss": 1.6152,
      "step": 236
    },
    {
      "epoch": 0.07,
      "grad_norm": 12.970022201538086,
      "learning_rate": 1.3913177091898151e-05,
      "loss": 1.9887,
      "step": 237
    },
    {
      "epoch": 0.07,
      "grad_norm": 15.360918045043945,
      "learning_rate": 1.391160427544118e-05,
      "loss": 1.711,
      "step": 238
    },
    {
      "epoch": 0.07,
      "grad_norm": 13.110067367553711,
      "learning_rate": 1.3910024955779158e-05,
      "loss": 2.768,
      "step": 239
    },
    {
      "epoch": 0.07,
      "grad_norm": 15.645023345947266,
      "learning_rate": 1.3908439134408911e-05,
      "loss": 2.8457,
      "step": 240
    },
    {
      "epoch": 0.07,
      "grad_norm": 14.423534393310547,
      "learning_rate": 1.3906846812833429e-05,
      "loss": 1.2738,
      "step": 241
    },
    {
      "epoch": 0.07,
      "grad_norm": 15.688253402709961,
      "learning_rate": 1.3905247992561863e-05,
      "loss": 3.8152,
      "step": 242
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.980319023132324,
      "learning_rate": 1.390364267510952e-05,
      "loss": 1.3614,
      "step": 243
    },
    {
      "epoch": 0.08,
      "grad_norm": 12.396924018859863,
      "learning_rate": 1.390203086199787e-05,
      "loss": 2.2875,
      "step": 244
    },
    {
      "epoch": 0.08,
      "grad_norm": 9.623881340026855,
      "learning_rate": 1.390041255475454e-05,
      "loss": 1.0153,
      "step": 245
    },
    {
      "epoch": 0.08,
      "grad_norm": 14.678177833557129,
      "learning_rate": 1.3898787754913304e-05,
      "loss": 2.5339,
      "step": 246
    },
    {
      "epoch": 0.08,
      "grad_norm": 18.33745765686035,
      "learning_rate": 1.3897156464014096e-05,
      "loss": 1.5622,
      "step": 247
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.040255546569824,
      "learning_rate": 1.3895518683603002e-05,
      "loss": 1.2037,
      "step": 248
    },
    {
      "epoch": 0.08,
      "grad_norm": 20.958837509155273,
      "learning_rate": 1.3893874415232256e-05,
      "loss": 2.3203,
      "step": 249
    },
    {
      "epoch": 0.08,
      "grad_norm": 11.519689559936523,
      "learning_rate": 1.389222366046024e-05,
      "loss": 0.971,
      "step": 250
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.441787719726562,
      "learning_rate": 1.3890566420851487e-05,
      "loss": 0.9101,
      "step": 251
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.943977355957031,
      "learning_rate": 1.3888902697976677e-05,
      "loss": 1.9499,
      "step": 252
    },
    {
      "epoch": 0.08,
      "grad_norm": 18.179977416992188,
      "learning_rate": 1.3887232493412629e-05,
      "loss": 1.6955,
      "step": 253
    },
    {
      "epoch": 0.08,
      "grad_norm": 12.42634391784668,
      "learning_rate": 1.388555580874231e-05,
      "loss": 2.1465,
      "step": 254
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.933977127075195,
      "learning_rate": 1.3883872645554827e-05,
      "loss": 0.9518,
      "step": 255
    },
    {
      "epoch": 0.08,
      "grad_norm": 20.83428382873535,
      "learning_rate": 1.3882183005445427e-05,
      "loss": 3.556,
      "step": 256
    },
    {
      "epoch": 0.08,
      "grad_norm": 13.598001480102539,
      "learning_rate": 1.3880486890015494e-05,
      "loss": 2.433,
      "step": 257
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.594669342041016,
      "learning_rate": 1.3878784300872555e-05,
      "loss": 1.9596,
      "step": 258
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.173829078674316,
      "learning_rate": 1.3877075239630266e-05,
      "loss": 2.4792,
      "step": 259
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.16698169708252,
      "learning_rate": 1.387535970790842e-05,
      "loss": 0.8115,
      "step": 260
    },
    {
      "epoch": 0.08,
      "grad_norm": 24.462717056274414,
      "learning_rate": 1.3873637707332943e-05,
      "loss": 5.0937,
      "step": 261
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.96473217010498,
      "learning_rate": 1.387190923953589e-05,
      "loss": 0.8035,
      "step": 262
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.601380348205566,
      "learning_rate": 1.3870174306155446e-05,
      "loss": 2.7259,
      "step": 263
    },
    {
      "epoch": 0.08,
      "grad_norm": 13.688828468322754,
      "learning_rate": 1.3868432908835925e-05,
      "loss": 2.0762,
      "step": 264
    },
    {
      "epoch": 0.08,
      "grad_norm": 12.210821151733398,
      "learning_rate": 1.3866685049227767e-05,
      "loss": 2.1519,
      "step": 265
    },
    {
      "epoch": 0.08,
      "grad_norm": 14.092357635498047,
      "learning_rate": 1.3864930728987538e-05,
      "loss": 1.3382,
      "step": 266
    },
    {
      "epoch": 0.08,
      "grad_norm": 26.011722564697266,
      "learning_rate": 1.3863169949777922e-05,
      "loss": 3.3422,
      "step": 267
    },
    {
      "epoch": 0.08,
      "grad_norm": 10.784296989440918,
      "learning_rate": 1.386140271326773e-05,
      "loss": 1.18,
      "step": 268
    },
    {
      "epoch": 0.08,
      "grad_norm": 14.321930885314941,
      "learning_rate": 1.3859629021131891e-05,
      "loss": 5.137,
      "step": 269
    },
    {
      "epoch": 0.08,
      "grad_norm": 17.20169448852539,
      "learning_rate": 1.3857848875051453e-05,
      "loss": 2.4465,
      "step": 270
    },
    {
      "epoch": 0.08,
      "grad_norm": 16.64161491394043,
      "learning_rate": 1.385606227671358e-05,
      "loss": 2.3278,
      "step": 271
    },
    {
      "epoch": 0.08,
      "grad_norm": 18.673120498657227,
      "learning_rate": 1.3854269227811553e-05,
      "loss": 2.082,
      "step": 272
    },
    {
      "epoch": 0.08,
      "grad_norm": 13.86021900177002,
      "learning_rate": 1.3852469730044761e-05,
      "loss": 2.2902,
      "step": 273
    },
    {
      "epoch": 0.08,
      "grad_norm": 15.379273414611816,
      "learning_rate": 1.3850663785118716e-05,
      "loss": 4.6714,
      "step": 274
    },
    {
      "epoch": 0.09,
      "grad_norm": 13.341683387756348,
      "learning_rate": 1.384885139474503e-05,
      "loss": 1.8605,
      "step": 275
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.289388656616211,
      "learning_rate": 1.3847032560641427e-05,
      "loss": 1.2596,
      "step": 276
    },
    {
      "epoch": 0.09,
      "grad_norm": 8.88835620880127,
      "learning_rate": 1.3845207284531742e-05,
      "loss": 0.8958,
      "step": 277
    },
    {
      "epoch": 0.09,
      "grad_norm": 18.075733184814453,
      "learning_rate": 1.3843375568145908e-05,
      "loss": 2.1797,
      "step": 278
    },
    {
      "epoch": 0.09,
      "grad_norm": 20.14925765991211,
      "learning_rate": 1.3841537413219967e-05,
      "loss": 2.904,
      "step": 279
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.297093391418457,
      "learning_rate": 1.3839692821496065e-05,
      "loss": 1.7218,
      "step": 280
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.621965408325195,
      "learning_rate": 1.3837841794722444e-05,
      "loss": 1.7314,
      "step": 281
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.384159088134766,
      "learning_rate": 1.383598433465345e-05,
      "loss": 1.74,
      "step": 282
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.62529754638672,
      "learning_rate": 1.3834120443049516e-05,
      "loss": 1.7181,
      "step": 283
    },
    {
      "epoch": 0.09,
      "grad_norm": 18.35346221923828,
      "learning_rate": 1.3832250121677187e-05,
      "loss": 1.9667,
      "step": 284
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.861061096191406,
      "learning_rate": 1.3830373372309083e-05,
      "loss": 2.3541,
      "step": 285
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.295331001281738,
      "learning_rate": 1.3828490196723935e-05,
      "loss": 2.0765,
      "step": 286
    },
    {
      "epoch": 0.09,
      "grad_norm": 23.350387573242188,
      "learning_rate": 1.3826600596706548e-05,
      "loss": 5.8657,
      "step": 287
    },
    {
      "epoch": 0.09,
      "grad_norm": 17.5737247467041,
      "learning_rate": 1.382470457404783e-05,
      "loss": 3.4766,
      "step": 288
    },
    {
      "epoch": 0.09,
      "grad_norm": 17.406465530395508,
      "learning_rate": 1.3822802130544763e-05,
      "loss": 1.1343,
      "step": 289
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.825047492980957,
      "learning_rate": 1.3820893268000426e-05,
      "loss": 2.3149,
      "step": 290
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.59432601928711,
      "learning_rate": 1.3818977988223976e-05,
      "loss": 5.5533,
      "step": 291
    },
    {
      "epoch": 0.09,
      "grad_norm": 17.41448974609375,
      "learning_rate": 1.381705629303065e-05,
      "loss": 4.7277,
      "step": 292
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.198013305664062,
      "learning_rate": 1.3815128184241771e-05,
      "loss": 2.3272,
      "step": 293
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.00592803955078,
      "learning_rate": 1.3813193663684735e-05,
      "loss": 2.0677,
      "step": 294
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.295944213867188,
      "learning_rate": 1.381125273319302e-05,
      "loss": 3.5203,
      "step": 295
    },
    {
      "epoch": 0.09,
      "grad_norm": 15.805230140686035,
      "learning_rate": 1.3809305394606176e-05,
      "loss": 1.6731,
      "step": 296
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.561698913574219,
      "learning_rate": 1.3807351649769828e-05,
      "loss": 1.9388,
      "step": 297
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.345500946044922,
      "learning_rate": 1.3805391500535669e-05,
      "loss": 0.9415,
      "step": 298
    },
    {
      "epoch": 0.09,
      "grad_norm": 15.150613784790039,
      "learning_rate": 1.3803424948761468e-05,
      "loss": 1.6428,
      "step": 299
    },
    {
      "epoch": 0.09,
      "grad_norm": 18.358455657958984,
      "learning_rate": 1.3801451996311053e-05,
      "loss": 3.3602,
      "step": 300
    },
    {
      "epoch": 0.09,
      "grad_norm": 13.325264930725098,
      "learning_rate": 1.3799472645054333e-05,
      "loss": 1.349,
      "step": 301
    },
    {
      "epoch": 0.09,
      "grad_norm": 11.161480903625488,
      "learning_rate": 1.3797486896867266e-05,
      "loss": 1.3497,
      "step": 302
    },
    {
      "epoch": 0.09,
      "grad_norm": 19.63141441345215,
      "learning_rate": 1.379549475363188e-05,
      "loss": 2.321,
      "step": 303
    },
    {
      "epoch": 0.09,
      "grad_norm": 10.052618980407715,
      "learning_rate": 1.3793496217236267e-05,
      "loss": 2.1645,
      "step": 304
    },
    {
      "epoch": 0.09,
      "grad_norm": 12.28446102142334,
      "learning_rate": 1.3791491289574571e-05,
      "loss": 1.6359,
      "step": 305
    },
    {
      "epoch": 0.09,
      "grad_norm": 16.950092315673828,
      "learning_rate": 1.3789479972546999e-05,
      "loss": 2.7423,
      "step": 306
    },
    {
      "epoch": 0.1,
      "grad_norm": 17.705276489257812,
      "learning_rate": 1.378746226805981e-05,
      "loss": 2.0395,
      "step": 307
    },
    {
      "epoch": 0.1,
      "grad_norm": 17.328041076660156,
      "learning_rate": 1.378543817802532e-05,
      "loss": 3.1794,
      "step": 308
    },
    {
      "epoch": 0.1,
      "grad_norm": 18.853071212768555,
      "learning_rate": 1.3783407704361897e-05,
      "loss": 2.7778,
      "step": 309
    },
    {
      "epoch": 0.1,
      "grad_norm": 14.925019264221191,
      "learning_rate": 1.3781370848993955e-05,
      "loss": 3.5123,
      "step": 310
    },
    {
      "epoch": 0.1,
      "grad_norm": 18.84942054748535,
      "learning_rate": 1.377932761385196e-05,
      "loss": 2.1698,
      "step": 311
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.201909065246582,
      "learning_rate": 1.3777278000872426e-05,
      "loss": 1.3381,
      "step": 312
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.145633697509766,
      "learning_rate": 1.3775222011997907e-05,
      "loss": 2.2904,
      "step": 313
    },
    {
      "epoch": 0.1,
      "grad_norm": 16.376514434814453,
      "learning_rate": 1.3773159649177003e-05,
      "loss": 5.1685,
      "step": 314
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.942585945129395,
      "learning_rate": 1.3771090914364354e-05,
      "loss": 2.2097,
      "step": 315
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.932531356811523,
      "learning_rate": 1.3769015809520639e-05,
      "loss": 5.0479,
      "step": 316
    },
    {
      "epoch": 0.1,
      "grad_norm": 21.648984909057617,
      "learning_rate": 1.3766934336612577e-05,
      "loss": 6.155,
      "step": 317
    },
    {
      "epoch": 0.1,
      "grad_norm": 19.75204849243164,
      "learning_rate": 1.3764846497612917e-05,
      "loss": 1.9287,
      "step": 318
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.067526817321777,
      "learning_rate": 1.3762752294500447e-05,
      "loss": 2.5628,
      "step": 319
    },
    {
      "epoch": 0.1,
      "grad_norm": 11.769791603088379,
      "learning_rate": 1.3760651729259985e-05,
      "loss": 1.8491,
      "step": 320
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.101560592651367,
      "learning_rate": 1.3758544803882376e-05,
      "loss": 1.8547,
      "step": 321
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.543452262878418,
      "learning_rate": 1.3756431520364499e-05,
      "loss": 4.2893,
      "step": 322
    },
    {
      "epoch": 0.1,
      "grad_norm": 22.693283081054688,
      "learning_rate": 1.375431188070925e-05,
      "loss": 2.9688,
      "step": 323
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.341410636901855,
      "learning_rate": 1.3752185886925557e-05,
      "loss": 2.0474,
      "step": 324
    },
    {
      "epoch": 0.1,
      "grad_norm": 13.230315208435059,
      "learning_rate": 1.375005354102837e-05,
      "loss": 1.388,
      "step": 325
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.137369155883789,
      "learning_rate": 1.3747914845038655e-05,
      "loss": 1.8743,
      "step": 326
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.931097984313965,
      "learning_rate": 1.3745769800983398e-05,
      "loss": 0.8339,
      "step": 327
    },
    {
      "epoch": 0.1,
      "grad_norm": 17.05511474609375,
      "learning_rate": 1.3743618410895602e-05,
      "loss": 3.0675,
      "step": 328
    },
    {
      "epoch": 0.1,
      "grad_norm": 15.675482749938965,
      "learning_rate": 1.3741460676814283e-05,
      "loss": 2.7984,
      "step": 329
    },
    {
      "epoch": 0.1,
      "grad_norm": 16.468854904174805,
      "learning_rate": 1.3739296600784477e-05,
      "loss": 4.7487,
      "step": 330
    },
    {
      "epoch": 0.1,
      "grad_norm": 14.824066162109375,
      "learning_rate": 1.3737126184857218e-05,
      "loss": 2.1714,
      "step": 331
    },
    {
      "epoch": 0.1,
      "grad_norm": 16.126014709472656,
      "learning_rate": 1.3734949431089558e-05,
      "loss": 4.4093,
      "step": 332
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.778393745422363,
      "learning_rate": 1.3732766341544554e-05,
      "loss": 2.6112,
      "step": 333
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.41983699798584,
      "learning_rate": 1.3730576918291266e-05,
      "loss": 2.6381,
      "step": 334
    },
    {
      "epoch": 0.1,
      "grad_norm": 9.529387474060059,
      "learning_rate": 1.3728381163404758e-05,
      "loss": 1.1488,
      "step": 335
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.54234790802002,
      "learning_rate": 1.3726179078966098e-05,
      "loss": 1.3704,
      "step": 336
    },
    {
      "epoch": 0.1,
      "grad_norm": 12.4400053024292,
      "learning_rate": 1.3723970667062345e-05,
      "loss": 1.4149,
      "step": 337
    },
    {
      "epoch": 0.1,
      "grad_norm": 10.626307487487793,
      "learning_rate": 1.3721755929786563e-05,
      "loss": 1.2803,
      "step": 338
    },
    {
      "epoch": 0.11,
      "grad_norm": 17.947769165039062,
      "learning_rate": 1.3719534869237809e-05,
      "loss": 6.1637,
      "step": 339
    },
    {
      "epoch": 0.11,
      "grad_norm": 14.783670425415039,
      "learning_rate": 1.371730748752113e-05,
      "loss": 3.3733,
      "step": 340
    },
    {
      "epoch": 0.11,
      "grad_norm": 15.04702091217041,
      "learning_rate": 1.3715073786747564e-05,
      "loss": 2.0311,
      "step": 341
    },
    {
      "epoch": 0.11,
      "grad_norm": 16.079864501953125,
      "learning_rate": 1.3712833769034142e-05,
      "loss": 2.1547,
      "step": 342
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.409101486206055,
      "learning_rate": 1.371058743650388e-05,
      "loss": 1.2826,
      "step": 343
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.11122989654541,
      "learning_rate": 1.3708334791285779e-05,
      "loss": 1.2537,
      "step": 344
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.088030815124512,
      "learning_rate": 1.3706075835514823e-05,
      "loss": 1.6252,
      "step": 345
    },
    {
      "epoch": 0.11,
      "grad_norm": 15.13451862335205,
      "learning_rate": 1.3703810571331973e-05,
      "loss": 6.6792,
      "step": 346
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.995061874389648,
      "learning_rate": 1.3701539000884178e-05,
      "loss": 1.1791,
      "step": 347
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.764225959777832,
      "learning_rate": 1.3699261126324355e-05,
      "loss": 3.4514,
      "step": 348
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.78808307647705,
      "learning_rate": 1.36969769498114e-05,
      "loss": 0.7701,
      "step": 349
    },
    {
      "epoch": 0.11,
      "grad_norm": 16.577268600463867,
      "learning_rate": 1.3694686473510184e-05,
      "loss": 4.2284,
      "step": 350
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.578121185302734,
      "learning_rate": 1.369238969959154e-05,
      "loss": 1.2305,
      "step": 351
    },
    {
      "epoch": 0.11,
      "grad_norm": 16.763050079345703,
      "learning_rate": 1.369008663023228e-05,
      "loss": 2.4429,
      "step": 352
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.74651050567627,
      "learning_rate": 1.3687777267615179e-05,
      "loss": 0.9373,
      "step": 353
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.947549819946289,
      "learning_rate": 1.3685461613928972e-05,
      "loss": 2.1318,
      "step": 354
    },
    {
      "epoch": 0.11,
      "grad_norm": 9.887306213378906,
      "learning_rate": 1.3683139671368361e-05,
      "loss": 0.9985,
      "step": 355
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.960177421569824,
      "learning_rate": 1.368081144213401e-05,
      "loss": 1.8042,
      "step": 356
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.04848861694336,
      "learning_rate": 1.3678476928432535e-05,
      "loss": 2.0705,
      "step": 357
    },
    {
      "epoch": 0.11,
      "grad_norm": 18.955495834350586,
      "learning_rate": 1.3676136132476515e-05,
      "loss": 5.26,
      "step": 358
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.15633487701416,
      "learning_rate": 1.367378905648448e-05,
      "loss": 0.9436,
      "step": 359
    },
    {
      "epoch": 0.11,
      "grad_norm": 17.71897315979004,
      "learning_rate": 1.367143570268091e-05,
      "loss": 1.9179,
      "step": 360
    },
    {
      "epoch": 0.11,
      "grad_norm": 16.045982360839844,
      "learning_rate": 1.366907607329624e-05,
      "loss": 1.6184,
      "step": 361
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.041046142578125,
      "learning_rate": 1.3666710170566847e-05,
      "loss": 1.4901,
      "step": 362
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.907164573669434,
      "learning_rate": 1.366433799673506e-05,
      "loss": 2.5307,
      "step": 363
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.025712966918945,
      "learning_rate": 1.3661959554049143e-05,
      "loss": 1.6155,
      "step": 364
    },
    {
      "epoch": 0.11,
      "grad_norm": 10.814041137695312,
      "learning_rate": 1.3659574844763313e-05,
      "loss": 1.7437,
      "step": 365
    },
    {
      "epoch": 0.11,
      "grad_norm": 16.084716796875,
      "learning_rate": 1.3657183871137716e-05,
      "loss": 2.8582,
      "step": 366
    },
    {
      "epoch": 0.11,
      "grad_norm": 13.304903030395508,
      "learning_rate": 1.3654786635438439e-05,
      "loss": 3.8795,
      "step": 367
    },
    {
      "epoch": 0.11,
      "grad_norm": 12.790328025817871,
      "learning_rate": 1.3652383139937504e-05,
      "loss": 3.2627,
      "step": 368
    },
    {
      "epoch": 0.11,
      "grad_norm": 11.926814079284668,
      "learning_rate": 1.3649973386912866e-05,
      "loss": 1.9194,
      "step": 369
    },
    {
      "epoch": 0.11,
      "grad_norm": 15.041619300842285,
      "learning_rate": 1.3647557378648409e-05,
      "loss": 6.1905,
      "step": 370
    },
    {
      "epoch": 0.11,
      "grad_norm": 18.288606643676758,
      "learning_rate": 1.364513511743395e-05,
      "loss": 2.6999,
      "step": 371
    },
    {
      "epoch": 0.12,
      "grad_norm": 13.575446128845215,
      "learning_rate": 1.3642706605565228e-05,
      "loss": 2.5753,
      "step": 372
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.08791446685791,
      "learning_rate": 1.3640271845343907e-05,
      "loss": 1.9656,
      "step": 373
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.84411907196045,
      "learning_rate": 1.3637830839077572e-05,
      "loss": 1.5011,
      "step": 374
    },
    {
      "epoch": 0.12,
      "grad_norm": 13.908230781555176,
      "learning_rate": 1.3635383589079732e-05,
      "loss": 1.8645,
      "step": 375
    },
    {
      "epoch": 0.12,
      "grad_norm": 22.58487319946289,
      "learning_rate": 1.363293009766981e-05,
      "loss": 2.4851,
      "step": 376
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.689922332763672,
      "learning_rate": 1.3630470367173147e-05,
      "loss": 2.8248,
      "step": 377
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.89048194885254,
      "learning_rate": 1.3628004399920988e-05,
      "loss": 1.8686,
      "step": 378
    },
    {
      "epoch": 0.12,
      "grad_norm": 13.554402351379395,
      "learning_rate": 1.3625532198250505e-05,
      "loss": 1.7365,
      "step": 379
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.678657531738281,
      "learning_rate": 1.3623053764504769e-05,
      "loss": 1.3599,
      "step": 380
    },
    {
      "epoch": 0.12,
      "grad_norm": 19.57718849182129,
      "learning_rate": 1.3620569101032753e-05,
      "loss": 1.7862,
      "step": 381
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.472726821899414,
      "learning_rate": 1.3618078210189348e-05,
      "loss": 1.0532,
      "step": 382
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.552109718322754,
      "learning_rate": 1.361558109433533e-05,
      "loss": 5.6745,
      "step": 383
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.24566650390625,
      "learning_rate": 1.3613077755837395e-05,
      "loss": 4.911,
      "step": 384
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.16314697265625,
      "learning_rate": 1.3610568197068114e-05,
      "loss": 1.9886,
      "step": 385
    },
    {
      "epoch": 0.12,
      "grad_norm": 14.017110824584961,
      "learning_rate": 1.3608052420405975e-05,
      "loss": 3.0355,
      "step": 386
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.076988220214844,
      "learning_rate": 1.3605530428235346e-05,
      "loss": 1.3674,
      "step": 387
    },
    {
      "epoch": 0.12,
      "grad_norm": 17.93915367126465,
      "learning_rate": 1.3603002222946485e-05,
      "loss": 6.2092,
      "step": 388
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.35929298400879,
      "learning_rate": 1.3600467806935547e-05,
      "loss": 2.6678,
      "step": 389
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.46489429473877,
      "learning_rate": 1.3597927182604568e-05,
      "loss": 1.1638,
      "step": 390
    },
    {
      "epoch": 0.12,
      "grad_norm": 13.621467590332031,
      "learning_rate": 1.3595380352361468e-05,
      "loss": 2.3239,
      "step": 391
    },
    {
      "epoch": 0.12,
      "grad_norm": 9.548460006713867,
      "learning_rate": 1.3592827318620053e-05,
      "loss": 1.1291,
      "step": 392
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.570243835449219,
      "learning_rate": 1.3590268083800002e-05,
      "loss": 1.7025,
      "step": 393
    },
    {
      "epoch": 0.12,
      "grad_norm": 15.140338897705078,
      "learning_rate": 1.3587702650326873e-05,
      "loss": 2.715,
      "step": 394
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.13969612121582,
      "learning_rate": 1.3585131020632103e-05,
      "loss": 2.1997,
      "step": 395
    },
    {
      "epoch": 0.12,
      "grad_norm": 17.66150665283203,
      "learning_rate": 1.3582553197153e-05,
      "loss": 2.4102,
      "step": 396
    },
    {
      "epoch": 0.12,
      "grad_norm": 14.307275772094727,
      "learning_rate": 1.3579969182332738e-05,
      "loss": 3.3309,
      "step": 397
    },
    {
      "epoch": 0.12,
      "grad_norm": 13.166078567504883,
      "learning_rate": 1.3577378978620364e-05,
      "loss": 1.7019,
      "step": 398
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.583813667297363,
      "learning_rate": 1.3574782588470789e-05,
      "loss": 1.9107,
      "step": 399
    },
    {
      "epoch": 0.12,
      "grad_norm": 10.487409591674805,
      "learning_rate": 1.3572180014344785e-05,
      "loss": 1.4231,
      "step": 400
    },
    {
      "epoch": 0.12,
      "grad_norm": 18.598587036132812,
      "learning_rate": 1.3569571258708989e-05,
      "loss": 2.8625,
      "step": 401
    },
    {
      "epoch": 0.12,
      "grad_norm": 12.454787254333496,
      "learning_rate": 1.3566956324035895e-05,
      "loss": 1.6535,
      "step": 402
    },
    {
      "epoch": 0.12,
      "grad_norm": 11.811790466308594,
      "learning_rate": 1.356433521280385e-05,
      "loss": 2.4083,
      "step": 403
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.476715087890625,
      "learning_rate": 1.356170792749706e-05,
      "loss": 2.9695,
      "step": 404
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.166971206665039,
      "learning_rate": 1.3559074470605582e-05,
      "loss": 1.5907,
      "step": 405
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.776312828063965,
      "learning_rate": 1.3556434844625316e-05,
      "loss": 1.1981,
      "step": 406
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.133995056152344,
      "learning_rate": 1.3553789052058017e-05,
      "loss": 1.6438,
      "step": 407
    },
    {
      "epoch": 0.13,
      "grad_norm": 19.32341766357422,
      "learning_rate": 1.3551137095411283e-05,
      "loss": 2.4166,
      "step": 408
    },
    {
      "epoch": 0.13,
      "grad_norm": 12.969614028930664,
      "learning_rate": 1.3548478977198546e-05,
      "loss": 1.886,
      "step": 409
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.233928680419922,
      "learning_rate": 1.354581469993909e-05,
      "loss": 1.7834,
      "step": 410
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.977611541748047,
      "learning_rate": 1.3543144266158026e-05,
      "loss": 1.0913,
      "step": 411
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.674370765686035,
      "learning_rate": 1.3540467678386306e-05,
      "loss": 5.5188,
      "step": 412
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.832097053527832,
      "learning_rate": 1.3537784939160712e-05,
      "loss": 3.5768,
      "step": 413
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.40325927734375,
      "learning_rate": 1.3535096051023858e-05,
      "loss": 1.9102,
      "step": 414
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.826641082763672,
      "learning_rate": 1.3532401016524185e-05,
      "loss": 2.6502,
      "step": 415
    },
    {
      "epoch": 0.13,
      "grad_norm": 20.26589584350586,
      "learning_rate": 1.3529699838215955e-05,
      "loss": 3.2667,
      "step": 416
    },
    {
      "epoch": 0.13,
      "grad_norm": 12.244672775268555,
      "learning_rate": 1.3526992518659262e-05,
      "loss": 1.8443,
      "step": 417
    },
    {
      "epoch": 0.13,
      "grad_norm": 17.309690475463867,
      "learning_rate": 1.3524279060420013e-05,
      "loss": 2.3227,
      "step": 418
    },
    {
      "epoch": 0.13,
      "grad_norm": 20.20816421508789,
      "learning_rate": 1.3521559466069934e-05,
      "loss": 2.2727,
      "step": 419
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.342791557312012,
      "learning_rate": 1.3518833738186571e-05,
      "loss": 1.4227,
      "step": 420
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.979636192321777,
      "learning_rate": 1.3516101879353277e-05,
      "loss": 2.3114,
      "step": 421
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.170080184936523,
      "learning_rate": 1.351336389215922e-05,
      "loss": 2.8633,
      "step": 422
    },
    {
      "epoch": 0.13,
      "grad_norm": 18.355520248413086,
      "learning_rate": 1.3510619779199375e-05,
      "loss": 2.6312,
      "step": 423
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.05256462097168,
      "learning_rate": 1.3507869543074524e-05,
      "loss": 1.1738,
      "step": 424
    },
    {
      "epoch": 0.13,
      "grad_norm": 11.840422630310059,
      "learning_rate": 1.350511318639125e-05,
      "loss": 2.0828,
      "step": 425
    },
    {
      "epoch": 0.13,
      "grad_norm": 15.11091423034668,
      "learning_rate": 1.3502350711761941e-05,
      "loss": 1.6218,
      "step": 426
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.374017715454102,
      "learning_rate": 1.3499582121804774e-05,
      "loss": 1.4773,
      "step": 427
    },
    {
      "epoch": 0.13,
      "grad_norm": 12.468167304992676,
      "learning_rate": 1.3496807419143736e-05,
      "loss": 2.0516,
      "step": 428
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.024086952209473,
      "learning_rate": 1.3494026606408593e-05,
      "loss": 0.9726,
      "step": 429
    },
    {
      "epoch": 0.13,
      "grad_norm": 10.078760147094727,
      "learning_rate": 1.3491239686234912e-05,
      "loss": 1.5166,
      "step": 430
    },
    {
      "epoch": 0.13,
      "grad_norm": 20.189712524414062,
      "learning_rate": 1.3488446661264046e-05,
      "loss": 4.3669,
      "step": 431
    },
    {
      "epoch": 0.13,
      "grad_norm": 14.545948028564453,
      "learning_rate": 1.3485647534143134e-05,
      "loss": 3.1417,
      "step": 432
    },
    {
      "epoch": 0.13,
      "grad_norm": 13.813590049743652,
      "learning_rate": 1.3482842307525094e-05,
      "loss": 1.9284,
      "step": 433
    },
    {
      "epoch": 0.13,
      "grad_norm": 6.276806354522705,
      "learning_rate": 1.348003098406863e-05,
      "loss": 0.9223,
      "step": 434
    },
    {
      "epoch": 0.13,
      "grad_norm": 23.6711368560791,
      "learning_rate": 1.3477213566438224e-05,
      "loss": 2.635,
      "step": 435
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.814998626708984,
      "learning_rate": 1.3474390057304129e-05,
      "loss": 2.1365,
      "step": 436
    },
    {
      "epoch": 0.14,
      "grad_norm": 19.597017288208008,
      "learning_rate": 1.347156045934238e-05,
      "loss": 6.3379,
      "step": 437
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.655862808227539,
      "learning_rate": 1.3468724775234775e-05,
      "loss": 1.3053,
      "step": 438
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.103118896484375,
      "learning_rate": 1.3465883007668883e-05,
      "loss": 2.5197,
      "step": 439
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.11824893951416,
      "learning_rate": 1.3463035159338038e-05,
      "loss": 4.5738,
      "step": 440
    },
    {
      "epoch": 0.14,
      "grad_norm": 22.6025447845459,
      "learning_rate": 1.3460181232941342e-05,
      "loss": 2.4246,
      "step": 441
    },
    {
      "epoch": 0.14,
      "grad_norm": 22.4508056640625,
      "learning_rate": 1.3457321231183648e-05,
      "loss": 3.8746,
      "step": 442
    },
    {
      "epoch": 0.14,
      "grad_norm": 17.415388107299805,
      "learning_rate": 1.3454455156775576e-05,
      "loss": 8.516,
      "step": 443
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.681063652038574,
      "learning_rate": 1.3451583012433499e-05,
      "loss": 1.5769,
      "step": 444
    },
    {
      "epoch": 0.14,
      "grad_norm": 15.727627754211426,
      "learning_rate": 1.3448704800879536e-05,
      "loss": 2.9668,
      "step": 445
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.169587135314941,
      "learning_rate": 1.3445820524841567e-05,
      "loss": 1.6273,
      "step": 446
    },
    {
      "epoch": 0.14,
      "grad_norm": 10.842358589172363,
      "learning_rate": 1.3442930187053217e-05,
      "loss": 1.948,
      "step": 447
    },
    {
      "epoch": 0.14,
      "grad_norm": 16.87100601196289,
      "learning_rate": 1.3440033790253847e-05,
      "loss": 2.1458,
      "step": 448
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.896727561950684,
      "learning_rate": 1.3437131337188573e-05,
      "loss": 2.6653,
      "step": 449
    },
    {
      "epoch": 0.14,
      "grad_norm": 11.050118446350098,
      "learning_rate": 1.3434222830608246e-05,
      "loss": 1.8485,
      "step": 450
    },
    {
      "epoch": 0.14,
      "grad_norm": 23.6415958404541,
      "learning_rate": 1.3431308273269449e-05,
      "loss": 2.607,
      "step": 451
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.02411937713623,
      "learning_rate": 1.3428387667934508e-05,
      "loss": 2.3825,
      "step": 452
    },
    {
      "epoch": 0.14,
      "grad_norm": 13.187039375305176,
      "learning_rate": 1.3425461017371479e-05,
      "loss": 1.4177,
      "step": 453
    },
    {
      "epoch": 0.14,
      "grad_norm": 11.099510192871094,
      "learning_rate": 1.3422528324354144e-05,
      "loss": 1.9126,
      "step": 454
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.535717964172363,
      "learning_rate": 1.341958959166201e-05,
      "loss": 2.7452,
      "step": 455
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.59811019897461,
      "learning_rate": 1.341664482208032e-05,
      "loss": 1.9522,
      "step": 456
    },
    {
      "epoch": 0.14,
      "grad_norm": 16.89012336730957,
      "learning_rate": 1.3413694018400026e-05,
      "loss": 6.0934,
      "step": 457
    },
    {
      "epoch": 0.14,
      "grad_norm": 17.078168869018555,
      "learning_rate": 1.3410737183417803e-05,
      "loss": 1.7698,
      "step": 458
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.786270141601562,
      "learning_rate": 1.3407774319936045e-05,
      "loss": 1.9643,
      "step": 459
    },
    {
      "epoch": 0.14,
      "grad_norm": 17.108491897583008,
      "learning_rate": 1.3404805430762857e-05,
      "loss": 2.8331,
      "step": 460
    },
    {
      "epoch": 0.14,
      "grad_norm": 19.87467384338379,
      "learning_rate": 1.3401830518712055e-05,
      "loss": 5.4639,
      "step": 461
    },
    {
      "epoch": 0.14,
      "grad_norm": 12.71600341796875,
      "learning_rate": 1.3398849586603164e-05,
      "loss": 2.2668,
      "step": 462
    },
    {
      "epoch": 0.14,
      "grad_norm": 6.157910346984863,
      "learning_rate": 1.3395862637261415e-05,
      "loss": 0.699,
      "step": 463
    },
    {
      "epoch": 0.14,
      "grad_norm": 20.636333465576172,
      "learning_rate": 1.3392869673517742e-05,
      "loss": 3.6988,
      "step": 464
    },
    {
      "epoch": 0.14,
      "grad_norm": 14.185775756835938,
      "learning_rate": 1.3389870698208776e-05,
      "loss": 3.227,
      "step": 465
    },
    {
      "epoch": 0.14,
      "grad_norm": 27.643817901611328,
      "learning_rate": 1.3386865714176849e-05,
      "loss": 2.6686,
      "step": 466
    },
    {
      "epoch": 0.14,
      "grad_norm": 18.93267822265625,
      "learning_rate": 1.3383854724269985e-05,
      "loss": 5.1713,
      "step": 467
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.180096626281738,
      "learning_rate": 1.3380837731341908e-05,
      "loss": 3.6741,
      "step": 468
    },
    {
      "epoch": 0.15,
      "grad_norm": 18.503711700439453,
      "learning_rate": 1.3377814738252022e-05,
      "loss": 5.9296,
      "step": 469
    },
    {
      "epoch": 0.15,
      "grad_norm": 17.148578643798828,
      "learning_rate": 1.337478574786542e-05,
      "loss": 2.8339,
      "step": 470
    },
    {
      "epoch": 0.15,
      "grad_norm": 12.4744873046875,
      "learning_rate": 1.3371750763052883e-05,
      "loss": 1.9957,
      "step": 471
    },
    {
      "epoch": 0.15,
      "grad_norm": 15.115225791931152,
      "learning_rate": 1.3368709786690868e-05,
      "loss": 2.205,
      "step": 472
    },
    {
      "epoch": 0.15,
      "grad_norm": 21.09969711303711,
      "learning_rate": 1.3365662821661518e-05,
      "loss": 2.9407,
      "step": 473
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.111128807067871,
      "learning_rate": 1.3362609870852641e-05,
      "loss": 5.3922,
      "step": 474
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.798126220703125,
      "learning_rate": 1.3359550937157732e-05,
      "loss": 0.9604,
      "step": 475
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.442574501037598,
      "learning_rate": 1.3356486023475942e-05,
      "loss": 1.9992,
      "step": 476
    },
    {
      "epoch": 0.15,
      "grad_norm": 15.014910697937012,
      "learning_rate": 1.3353415132712097e-05,
      "loss": 2.1064,
      "step": 477
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.74669075012207,
      "learning_rate": 1.3350338267776692e-05,
      "loss": 2.6105,
      "step": 478
    },
    {
      "epoch": 0.15,
      "grad_norm": 14.472315788269043,
      "learning_rate": 1.3347255431585876e-05,
      "loss": 3.1315,
      "step": 479
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.927700996398926,
      "learning_rate": 1.334416662706146e-05,
      "loss": 1.6392,
      "step": 480
    },
    {
      "epoch": 0.15,
      "grad_norm": 23.360815048217773,
      "learning_rate": 1.3341071857130912e-05,
      "loss": 4.32,
      "step": 481
    },
    {
      "epoch": 0.15,
      "grad_norm": 12.88177490234375,
      "learning_rate": 1.3337971124727359e-05,
      "loss": 3.0116,
      "step": 482
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.984797477722168,
      "learning_rate": 1.3334864432789568e-05,
      "loss": 2.155,
      "step": 483
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.673538208007812,
      "learning_rate": 1.3331751784261964e-05,
      "loss": 2.4724,
      "step": 484
    },
    {
      "epoch": 0.15,
      "grad_norm": 17.153629302978516,
      "learning_rate": 1.3328633182094614e-05,
      "loss": 3.5844,
      "step": 485
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.198391914367676,
      "learning_rate": 1.3325508629243226e-05,
      "loss": 1.4172,
      "step": 486
    },
    {
      "epoch": 0.15,
      "grad_norm": 15.56090259552002,
      "learning_rate": 1.332237812866915e-05,
      "loss": 4.2267,
      "step": 487
    },
    {
      "epoch": 0.15,
      "grad_norm": 21.435331344604492,
      "learning_rate": 1.3319241683339371e-05,
      "loss": 4.9318,
      "step": 488
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.9794921875,
      "learning_rate": 1.3316099296226514e-05,
      "loss": 3.6705,
      "step": 489
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.50817584991455,
      "learning_rate": 1.3312950970308827e-05,
      "loss": 1.3616,
      "step": 490
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.41978645324707,
      "learning_rate": 1.3309796708570194e-05,
      "loss": 5.9281,
      "step": 491
    },
    {
      "epoch": 0.15,
      "grad_norm": 17.76799964904785,
      "learning_rate": 1.3306636514000118e-05,
      "loss": 1.8393,
      "step": 492
    },
    {
      "epoch": 0.15,
      "grad_norm": 15.68265438079834,
      "learning_rate": 1.3303470389593731e-05,
      "loss": 2.5935,
      "step": 493
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.82168960571289,
      "learning_rate": 1.3300298338351783e-05,
      "loss": 5.2416,
      "step": 494
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.42543888092041,
      "learning_rate": 1.3297120363280635e-05,
      "loss": 1.201,
      "step": 495
    },
    {
      "epoch": 0.15,
      "grad_norm": 14.174787521362305,
      "learning_rate": 1.3293936467392277e-05,
      "loss": 1.582,
      "step": 496
    },
    {
      "epoch": 0.15,
      "grad_norm": 16.048009872436523,
      "learning_rate": 1.3290746653704298e-05,
      "loss": 1.8858,
      "step": 497
    },
    {
      "epoch": 0.15,
      "grad_norm": 11.09557819366455,
      "learning_rate": 1.3287550925239899e-05,
      "loss": 1.8468,
      "step": 498
    },
    {
      "epoch": 0.15,
      "grad_norm": 13.926736831665039,
      "learning_rate": 1.3284349285027888e-05,
      "loss": 6.1665,
      "step": 499
    },
    {
      "epoch": 0.15,
      "grad_norm": 10.572114944458008,
      "learning_rate": 1.3281141736102673e-05,
      "loss": 1.5575,
      "step": 500
    },
    {
      "epoch": 0.16,
      "grad_norm": 17.857831954956055,
      "learning_rate": 1.3277928281504268e-05,
      "loss": 6.2458,
      "step": 501
    },
    {
      "epoch": 0.16,
      "grad_norm": 15.5122709274292,
      "learning_rate": 1.3274708924278278e-05,
      "loss": 2.3332,
      "step": 502
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.162206649780273,
      "learning_rate": 1.3271483667475906e-05,
      "loss": 2.3407,
      "step": 503
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.517176628112793,
      "learning_rate": 1.3268252514153947e-05,
      "loss": 5.2842,
      "step": 504
    },
    {
      "epoch": 0.16,
      "grad_norm": 19.05678939819336,
      "learning_rate": 1.326501546737478e-05,
      "loss": 2.3107,
      "step": 505
    },
    {
      "epoch": 0.16,
      "grad_norm": 11.515403747558594,
      "learning_rate": 1.3261772530206374e-05,
      "loss": 2.2464,
      "step": 506
    },
    {
      "epoch": 0.16,
      "grad_norm": 23.113420486450195,
      "learning_rate": 1.3258523705722276e-05,
      "loss": 6.0012,
      "step": 507
    },
    {
      "epoch": 0.16,
      "grad_norm": 15.704401969909668,
      "learning_rate": 1.3255268997001622e-05,
      "loss": 2.0706,
      "step": 508
    },
    {
      "epoch": 0.16,
      "grad_norm": 15.194841384887695,
      "learning_rate": 1.3252008407129117e-05,
      "loss": 3.7562,
      "step": 509
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.791894912719727,
      "learning_rate": 1.3248741939195037e-05,
      "loss": 3.2929,
      "step": 510
    },
    {
      "epoch": 0.16,
      "grad_norm": 13.705718040466309,
      "learning_rate": 1.3245469596295241e-05,
      "loss": 1.7648,
      "step": 511
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.339812278747559,
      "learning_rate": 1.3242191381531145e-05,
      "loss": 1.8999,
      "step": 512
    },
    {
      "epoch": 0.16,
      "grad_norm": 13.354348182678223,
      "learning_rate": 1.3238907298009734e-05,
      "loss": 1.6501,
      "step": 513
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.52679443359375,
      "learning_rate": 1.3235617348843556e-05,
      "loss": 5.294,
      "step": 514
    },
    {
      "epoch": 0.16,
      "grad_norm": 23.997087478637695,
      "learning_rate": 1.3232321537150719e-05,
      "loss": 3.2144,
      "step": 515
    },
    {
      "epoch": 0.16,
      "grad_norm": 22.67416763305664,
      "learning_rate": 1.3229019866054885e-05,
      "loss": 5.2678,
      "step": 516
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.221373558044434,
      "learning_rate": 1.3225712338685269e-05,
      "loss": 2.6089,
      "step": 517
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.597294807434082,
      "learning_rate": 1.322239895817664e-05,
      "loss": 2.6505,
      "step": 518
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.288494110107422,
      "learning_rate": 1.321907972766931e-05,
      "loss": 1.4499,
      "step": 519
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.892966270446777,
      "learning_rate": 1.321575465030914e-05,
      "loss": 2.4837,
      "step": 520
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.720625877380371,
      "learning_rate": 1.3212423729247526e-05,
      "loss": 2.8317,
      "step": 521
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.73737907409668,
      "learning_rate": 1.3209086967641413e-05,
      "loss": 2.3667,
      "step": 522
    },
    {
      "epoch": 0.16,
      "grad_norm": 14.303455352783203,
      "learning_rate": 1.320574436865327e-05,
      "loss": 2.9824,
      "step": 523
    },
    {
      "epoch": 0.16,
      "grad_norm": 13.44922924041748,
      "learning_rate": 1.3202395935451102e-05,
      "loss": 1.7328,
      "step": 524
    },
    {
      "epoch": 0.16,
      "grad_norm": 19.608003616333008,
      "learning_rate": 1.319904167120845e-05,
      "loss": 5.1686,
      "step": 525
    },
    {
      "epoch": 0.16,
      "grad_norm": 15.829699516296387,
      "learning_rate": 1.3195681579104374e-05,
      "loss": 3.0279,
      "step": 526
    },
    {
      "epoch": 0.16,
      "grad_norm": 20.571901321411133,
      "learning_rate": 1.319231566232346e-05,
      "loss": 1.9538,
      "step": 527
    },
    {
      "epoch": 0.16,
      "grad_norm": 24.14697265625,
      "learning_rate": 1.3188943924055812e-05,
      "loss": 2.0403,
      "step": 528
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.41229248046875,
      "learning_rate": 1.3185566367497059e-05,
      "loss": 1.7593,
      "step": 529
    },
    {
      "epoch": 0.16,
      "grad_norm": 12.556974411010742,
      "learning_rate": 1.3182182995848333e-05,
      "loss": 1.9916,
      "step": 530
    },
    {
      "epoch": 0.16,
      "grad_norm": 16.047039031982422,
      "learning_rate": 1.3178793812316288e-05,
      "loss": 1.9223,
      "step": 531
    },
    {
      "epoch": 0.16,
      "grad_norm": 10.137266159057617,
      "learning_rate": 1.3175398820113079e-05,
      "loss": 1.8314,
      "step": 532
    },
    {
      "epoch": 0.17,
      "grad_norm": 16.968952178955078,
      "learning_rate": 1.3171998022456371e-05,
      "loss": 2.4476,
      "step": 533
    },
    {
      "epoch": 0.17,
      "grad_norm": 14.670219421386719,
      "learning_rate": 1.3168591422569328e-05,
      "loss": 5.2239,
      "step": 534
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.691137313842773,
      "learning_rate": 1.3165179023680615e-05,
      "loss": 2.215,
      "step": 535
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.702433586120605,
      "learning_rate": 1.3161760829024394e-05,
      "loss": 1.2759,
      "step": 536
    },
    {
      "epoch": 0.17,
      "grad_norm": 13.136799812316895,
      "learning_rate": 1.3158336841840314e-05,
      "loss": 3.76,
      "step": 537
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.581886291503906,
      "learning_rate": 1.3154907065373523e-05,
      "loss": 1.1901,
      "step": 538
    },
    {
      "epoch": 0.17,
      "grad_norm": 18.704313278198242,
      "learning_rate": 1.3151471502874648e-05,
      "loss": 2.523,
      "step": 539
    },
    {
      "epoch": 0.17,
      "grad_norm": 17.407930374145508,
      "learning_rate": 1.3148030157599808e-05,
      "loss": 2.3264,
      "step": 540
    },
    {
      "epoch": 0.17,
      "grad_norm": 18.318424224853516,
      "learning_rate": 1.3144583032810595e-05,
      "loss": 3.9198,
      "step": 541
    },
    {
      "epoch": 0.17,
      "grad_norm": 16.8125057220459,
      "learning_rate": 1.314113013177408e-05,
      "loss": 1.7953,
      "step": 542
    },
    {
      "epoch": 0.17,
      "grad_norm": 12.802943229675293,
      "learning_rate": 1.3137671457762812e-05,
      "loss": 2.2133,
      "step": 543
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.229947090148926,
      "learning_rate": 1.3134207014054812e-05,
      "loss": 5.0568,
      "step": 544
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.939119338989258,
      "learning_rate": 1.313073680393356e-05,
      "loss": 1.1138,
      "step": 545
    },
    {
      "epoch": 0.17,
      "grad_norm": 14.192569732666016,
      "learning_rate": 1.3127260830688016e-05,
      "loss": 5.0628,
      "step": 546
    },
    {
      "epoch": 0.17,
      "grad_norm": 20.657682418823242,
      "learning_rate": 1.3123779097612589e-05,
      "loss": 4.2917,
      "step": 547
    },
    {
      "epoch": 0.17,
      "grad_norm": 16.68743133544922,
      "learning_rate": 1.3120291608007155e-05,
      "loss": 2.6444,
      "step": 548
    },
    {
      "epoch": 0.17,
      "grad_norm": 18.14390754699707,
      "learning_rate": 1.3116798365177041e-05,
      "loss": 4.8532,
      "step": 549
    },
    {
      "epoch": 0.17,
      "grad_norm": 16.764875411987305,
      "learning_rate": 1.311329937243303e-05,
      "loss": 1.8359,
      "step": 550
    },
    {
      "epoch": 0.17,
      "grad_norm": 17.864337921142578,
      "learning_rate": 1.3109794633091354e-05,
      "loss": 3.1203,
      "step": 551
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.85683536529541,
      "learning_rate": 1.3106284150473688e-05,
      "loss": 3.0269,
      "step": 552
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.915852546691895,
      "learning_rate": 1.3102767927907157e-05,
      "loss": 1.3493,
      "step": 553
    },
    {
      "epoch": 0.17,
      "grad_norm": 12.88638973236084,
      "learning_rate": 1.3099245968724319e-05,
      "loss": 1.7223,
      "step": 554
    },
    {
      "epoch": 0.17,
      "grad_norm": 14.523603439331055,
      "learning_rate": 1.3095718276263174e-05,
      "loss": 2.2196,
      "step": 555
    },
    {
      "epoch": 0.17,
      "grad_norm": 15.751676559448242,
      "learning_rate": 1.3092184853867154e-05,
      "loss": 5.3548,
      "step": 556
    },
    {
      "epoch": 0.17,
      "grad_norm": 10.33692741394043,
      "learning_rate": 1.3088645704885122e-05,
      "loss": 1.6451,
      "step": 557
    },
    {
      "epoch": 0.17,
      "grad_norm": 17.207609176635742,
      "learning_rate": 1.3085100832671366e-05,
      "loss": 3.7781,
      "step": 558
    },
    {
      "epoch": 0.17,
      "grad_norm": 19.840381622314453,
      "learning_rate": 1.30815502405856e-05,
      "loss": 7.1694,
      "step": 559
    },
    {
      "epoch": 0.17,
      "grad_norm": 14.988025665283203,
      "learning_rate": 1.3077993931992962e-05,
      "loss": 4.8949,
      "step": 560
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.11162281036377,
      "learning_rate": 1.3074431910264007e-05,
      "loss": 5.5227,
      "step": 561
    },
    {
      "epoch": 0.17,
      "grad_norm": 14.494366645812988,
      "learning_rate": 1.30708641787747e-05,
      "loss": 3.0651,
      "step": 562
    },
    {
      "epoch": 0.17,
      "grad_norm": 11.259400367736816,
      "learning_rate": 1.306729074090642e-05,
      "loss": 1.4138,
      "step": 563
    },
    {
      "epoch": 0.17,
      "grad_norm": 8.006839752197266,
      "learning_rate": 1.3063711600045956e-05,
      "loss": 1.6043,
      "step": 564
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.398590087890625,
      "learning_rate": 1.3060126759585505e-05,
      "loss": 3.1046,
      "step": 565
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.973245620727539,
      "learning_rate": 1.3056536222922657e-05,
      "loss": 1.1749,
      "step": 566
    },
    {
      "epoch": 0.18,
      "grad_norm": 14.982606887817383,
      "learning_rate": 1.3052939993460407e-05,
      "loss": 4.2511,
      "step": 567
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.401564598083496,
      "learning_rate": 1.3049338074607144e-05,
      "loss": 2.8225,
      "step": 568
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.368091583251953,
      "learning_rate": 1.3045730469776653e-05,
      "loss": 2.1884,
      "step": 569
    },
    {
      "epoch": 0.18,
      "grad_norm": 7.658648490905762,
      "learning_rate": 1.30421171823881e-05,
      "loss": 0.8779,
      "step": 570
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.248800277709961,
      "learning_rate": 1.303849821586604e-05,
      "loss": 1.454,
      "step": 571
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.093027114868164,
      "learning_rate": 1.3034873573640416e-05,
      "loss": 5.9006,
      "step": 572
    },
    {
      "epoch": 0.18,
      "grad_norm": 15.373260498046875,
      "learning_rate": 1.3031243259146544e-05,
      "loss": 1.8576,
      "step": 573
    },
    {
      "epoch": 0.18,
      "grad_norm": 21.09121322631836,
      "learning_rate": 1.302760727582512e-05,
      "loss": 2.4911,
      "step": 574
    },
    {
      "epoch": 0.18,
      "grad_norm": 15.990294456481934,
      "learning_rate": 1.3023965627122207e-05,
      "loss": 1.9747,
      "step": 575
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.601783752441406,
      "learning_rate": 1.302031831648924e-05,
      "loss": 2.3455,
      "step": 576
    },
    {
      "epoch": 0.18,
      "grad_norm": 16.258012771606445,
      "learning_rate": 1.3016665347383024e-05,
      "loss": 2.3826,
      "step": 577
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.222940444946289,
      "learning_rate": 1.3013006723265727e-05,
      "loss": 1.3702,
      "step": 578
    },
    {
      "epoch": 0.18,
      "grad_norm": 11.493943214416504,
      "learning_rate": 1.3009342447604867e-05,
      "loss": 2.4295,
      "step": 579
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.920111656188965,
      "learning_rate": 1.300567252387333e-05,
      "loss": 3.0531,
      "step": 580
    },
    {
      "epoch": 0.18,
      "grad_norm": 18.79691505432129,
      "learning_rate": 1.3001996955549348e-05,
      "loss": 5.7553,
      "step": 581
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.230389595031738,
      "learning_rate": 1.2998315746116507e-05,
      "loss": 1.6072,
      "step": 582
    },
    {
      "epoch": 0.18,
      "grad_norm": 14.716042518615723,
      "learning_rate": 1.2994628899063736e-05,
      "loss": 2.5715,
      "step": 583
    },
    {
      "epoch": 0.18,
      "grad_norm": 17.834688186645508,
      "learning_rate": 1.2990936417885309e-05,
      "loss": 3.0267,
      "step": 584
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.204316139221191,
      "learning_rate": 1.2987238306080838e-05,
      "loss": 2.471,
      "step": 585
    },
    {
      "epoch": 0.18,
      "grad_norm": 18.392969131469727,
      "learning_rate": 1.2983534567155274e-05,
      "loss": 1.9733,
      "step": 586
    },
    {
      "epoch": 0.18,
      "grad_norm": 16.151996612548828,
      "learning_rate": 1.2979825204618902e-05,
      "loss": 4.7029,
      "step": 587
    },
    {
      "epoch": 0.18,
      "grad_norm": 21.312026977539062,
      "learning_rate": 1.2976110221987332e-05,
      "loss": 3.0766,
      "step": 588
    },
    {
      "epoch": 0.18,
      "grad_norm": 26.476966857910156,
      "learning_rate": 1.2972389622781504e-05,
      "loss": 1.6432,
      "step": 589
    },
    {
      "epoch": 0.18,
      "grad_norm": 19.200429916381836,
      "learning_rate": 1.2968663410527683e-05,
      "loss": 2.3434,
      "step": 590
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.3663330078125,
      "learning_rate": 1.296493158875745e-05,
      "loss": 1.3629,
      "step": 591
    },
    {
      "epoch": 0.18,
      "grad_norm": 10.774557113647461,
      "learning_rate": 1.2961194161007703e-05,
      "loss": 1.1199,
      "step": 592
    },
    {
      "epoch": 0.18,
      "grad_norm": 16.63807487487793,
      "learning_rate": 1.2957451130820656e-05,
      "loss": 3.0353,
      "step": 593
    },
    {
      "epoch": 0.18,
      "grad_norm": 9.413061141967773,
      "learning_rate": 1.295370250174383e-05,
      "loss": 1.0713,
      "step": 594
    },
    {
      "epoch": 0.18,
      "grad_norm": 15.048751831054688,
      "learning_rate": 1.2949948277330057e-05,
      "loss": 2.6175,
      "step": 595
    },
    {
      "epoch": 0.18,
      "grad_norm": 13.444994926452637,
      "learning_rate": 1.2946188461137461e-05,
      "loss": 1.7131,
      "step": 596
    },
    {
      "epoch": 0.18,
      "grad_norm": 14.747013092041016,
      "learning_rate": 1.2942423056729482e-05,
      "loss": 2.4692,
      "step": 597
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.79983139038086,
      "learning_rate": 1.2938652067674841e-05,
      "loss": 1.1351,
      "step": 598
    },
    {
      "epoch": 0.19,
      "grad_norm": 17.529685974121094,
      "learning_rate": 1.2934875497547564e-05,
      "loss": 2.5644,
      "step": 599
    },
    {
      "epoch": 0.19,
      "grad_norm": 14.526769638061523,
      "learning_rate": 1.2931093349926957e-05,
      "loss": 1.9986,
      "step": 600
    },
    {
      "epoch": 0.19,
      "grad_norm": 30.357128143310547,
      "learning_rate": 1.2927305628397617e-05,
      "loss": 2.2509,
      "step": 601
    },
    {
      "epoch": 0.19,
      "grad_norm": 18.066158294677734,
      "learning_rate": 1.2923512336549426e-05,
      "loss": 5.8853,
      "step": 602
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.021641731262207,
      "learning_rate": 1.2919713477977541e-05,
      "loss": 3.4449,
      "step": 603
    },
    {
      "epoch": 0.19,
      "grad_norm": 14.591251373291016,
      "learning_rate": 1.2915909056282392e-05,
      "loss": 2.6123,
      "step": 604
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.648482322692871,
      "learning_rate": 1.2912099075069694e-05,
      "loss": 1.0796,
      "step": 605
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.213525772094727,
      "learning_rate": 1.2908283537950416e-05,
      "loss": 0.7585,
      "step": 606
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.706143379211426,
      "learning_rate": 1.2904462448540806e-05,
      "loss": 1.93,
      "step": 607
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.988866806030273,
      "learning_rate": 1.2900635810462365e-05,
      "loss": 1.1424,
      "step": 608
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.18970775604248,
      "learning_rate": 1.2896803627341854e-05,
      "loss": 2.6315,
      "step": 609
    },
    {
      "epoch": 0.19,
      "grad_norm": 14.60741138458252,
      "learning_rate": 1.2892965902811291e-05,
      "loss": 3.0535,
      "step": 610
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.429036140441895,
      "learning_rate": 1.2889122640507951e-05,
      "loss": 1.5899,
      "step": 611
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.819948196411133,
      "learning_rate": 1.2885273844074347e-05,
      "loss": 1.2347,
      "step": 612
    },
    {
      "epoch": 0.19,
      "grad_norm": 7.144796848297119,
      "learning_rate": 1.288141951715825e-05,
      "loss": 0.8606,
      "step": 613
    },
    {
      "epoch": 0.19,
      "grad_norm": 21.950687408447266,
      "learning_rate": 1.287755966341266e-05,
      "loss": 2.5471,
      "step": 614
    },
    {
      "epoch": 0.19,
      "grad_norm": 20.954431533813477,
      "learning_rate": 1.2873694286495825e-05,
      "loss": 5.1857,
      "step": 615
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.520883560180664,
      "learning_rate": 1.2869823390071218e-05,
      "loss": 1.6997,
      "step": 616
    },
    {
      "epoch": 0.19,
      "grad_norm": 20.543258666992188,
      "learning_rate": 1.2865946977807552e-05,
      "loss": 5.732,
      "step": 617
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.407218933105469,
      "learning_rate": 1.2862065053378762e-05,
      "loss": 1.45,
      "step": 618
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.259573936462402,
      "learning_rate": 1.2858177620464013e-05,
      "loss": 2.0572,
      "step": 619
    },
    {
      "epoch": 0.19,
      "grad_norm": 16.575164794921875,
      "learning_rate": 1.2854284682747688e-05,
      "loss": 2.4111,
      "step": 620
    },
    {
      "epoch": 0.19,
      "grad_norm": 19.309242248535156,
      "learning_rate": 1.285038624391938e-05,
      "loss": 4.1687,
      "step": 621
    },
    {
      "epoch": 0.19,
      "grad_norm": 21.330078125,
      "learning_rate": 1.2846482307673907e-05,
      "loss": 3.9547,
      "step": 622
    },
    {
      "epoch": 0.19,
      "grad_norm": 10.402637481689453,
      "learning_rate": 1.2842572877711293e-05,
      "loss": 1.5042,
      "step": 623
    },
    {
      "epoch": 0.19,
      "grad_norm": 12.396795272827148,
      "learning_rate": 1.2838657957736767e-05,
      "loss": 2.5599,
      "step": 624
    },
    {
      "epoch": 0.19,
      "grad_norm": 16.713336944580078,
      "learning_rate": 1.2834737551460762e-05,
      "loss": 2.0968,
      "step": 625
    },
    {
      "epoch": 0.19,
      "grad_norm": 13.006550788879395,
      "learning_rate": 1.2830811662598913e-05,
      "loss": 2.6239,
      "step": 626
    },
    {
      "epoch": 0.19,
      "grad_norm": 16.9903621673584,
      "learning_rate": 1.282688029487205e-05,
      "loss": 4.884,
      "step": 627
    },
    {
      "epoch": 0.19,
      "grad_norm": 18.870378494262695,
      "learning_rate": 1.2822943452006195e-05,
      "loss": 5.6409,
      "step": 628
    },
    {
      "epoch": 0.19,
      "grad_norm": 11.85020637512207,
      "learning_rate": 1.2819001137732556e-05,
      "loss": 1.9056,
      "step": 629
    },
    {
      "epoch": 0.2,
      "grad_norm": 15.124698638916016,
      "learning_rate": 1.2815053355787535e-05,
      "loss": 1.8045,
      "step": 630
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.318387985229492,
      "learning_rate": 1.2811100109912706e-05,
      "loss": 3.9206,
      "step": 631
    },
    {
      "epoch": 0.2,
      "grad_norm": 15.948100090026855,
      "learning_rate": 1.2807141403854832e-05,
      "loss": 2.8284,
      "step": 632
    },
    {
      "epoch": 0.2,
      "grad_norm": 21.467557907104492,
      "learning_rate": 1.2803177241365842e-05,
      "loss": 2.9303,
      "step": 633
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.317900657653809,
      "learning_rate": 1.2799207626202844e-05,
      "loss": 1.7439,
      "step": 634
    },
    {
      "epoch": 0.2,
      "grad_norm": 22.261106491088867,
      "learning_rate": 1.2795232562128104e-05,
      "loss": 4.7986,
      "step": 635
    },
    {
      "epoch": 0.2,
      "grad_norm": 18.067625045776367,
      "learning_rate": 1.2791252052909065e-05,
      "loss": 3.3529,
      "step": 636
    },
    {
      "epoch": 0.2,
      "grad_norm": 20.194984436035156,
      "learning_rate": 1.278726610231832e-05,
      "loss": 6.0626,
      "step": 637
    },
    {
      "epoch": 0.2,
      "grad_norm": 17.769323348999023,
      "learning_rate": 1.278327471413362e-05,
      "loss": 3.8201,
      "step": 638
    },
    {
      "epoch": 0.2,
      "grad_norm": 14.183991432189941,
      "learning_rate": 1.2779277892137882e-05,
      "loss": 3.6506,
      "step": 639
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.68132495880127,
      "learning_rate": 1.2775275640119154e-05,
      "loss": 5.0845,
      "step": 640
    },
    {
      "epoch": 0.2,
      "grad_norm": 16.511796951293945,
      "learning_rate": 1.2771267961870646e-05,
      "loss": 2.6582,
      "step": 641
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.455010414123535,
      "learning_rate": 1.2767254861190701e-05,
      "loss": 1.3631,
      "step": 642
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.576130867004395,
      "learning_rate": 1.2763236341882811e-05,
      "loss": 2.3939,
      "step": 643
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.145406723022461,
      "learning_rate": 1.2759212407755592e-05,
      "loss": 2.4653,
      "step": 644
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.685088157653809,
      "learning_rate": 1.2755183062622799e-05,
      "loss": 4.6326,
      "step": 645
    },
    {
      "epoch": 0.2,
      "grad_norm": 14.559864044189453,
      "learning_rate": 1.2751148310303317e-05,
      "loss": 4.432,
      "step": 646
    },
    {
      "epoch": 0.2,
      "grad_norm": 15.920525550842285,
      "learning_rate": 1.274710815462115e-05,
      "loss": 2.2933,
      "step": 647
    },
    {
      "epoch": 0.2,
      "grad_norm": 7.618566036224365,
      "learning_rate": 1.2743062599405426e-05,
      "loss": 0.9646,
      "step": 648
    },
    {
      "epoch": 0.2,
      "grad_norm": 11.278463363647461,
      "learning_rate": 1.273901164849039e-05,
      "loss": 1.205,
      "step": 649
    },
    {
      "epoch": 0.2,
      "grad_norm": 14.634114265441895,
      "learning_rate": 1.2734955305715405e-05,
      "loss": 4.1018,
      "step": 650
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.614447593688965,
      "learning_rate": 1.2730893574924937e-05,
      "loss": 2.924,
      "step": 651
    },
    {
      "epoch": 0.2,
      "grad_norm": 12.723915100097656,
      "learning_rate": 1.2726826459968565e-05,
      "loss": 2.6999,
      "step": 652
    },
    {
      "epoch": 0.2,
      "grad_norm": 17.578689575195312,
      "learning_rate": 1.2722753964700963e-05,
      "loss": 1.9297,
      "step": 653
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.462774276733398,
      "learning_rate": 1.2718676092981914e-05,
      "loss": 1.9175,
      "step": 654
    },
    {
      "epoch": 0.2,
      "grad_norm": 14.999417304992676,
      "learning_rate": 1.2714592848676291e-05,
      "loss": 2.4162,
      "step": 655
    },
    {
      "epoch": 0.2,
      "grad_norm": 15.338152885437012,
      "learning_rate": 1.2710504235654058e-05,
      "loss": 2.0275,
      "step": 656
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.914753913879395,
      "learning_rate": 1.2706410257790273e-05,
      "loss": 2.2412,
      "step": 657
    },
    {
      "epoch": 0.2,
      "grad_norm": 10.793363571166992,
      "learning_rate": 1.2702310918965072e-05,
      "loss": 1.4223,
      "step": 658
    },
    {
      "epoch": 0.2,
      "grad_norm": 13.35788631439209,
      "learning_rate": 1.2698206223063677e-05,
      "loss": 1.6989,
      "step": 659
    },
    {
      "epoch": 0.2,
      "grad_norm": 9.177985191345215,
      "learning_rate": 1.2694096173976381e-05,
      "loss": 1.2592,
      "step": 660
    },
    {
      "epoch": 0.2,
      "grad_norm": 21.1748104095459,
      "learning_rate": 1.2689980775598558e-05,
      "loss": 6.016,
      "step": 661
    },
    {
      "epoch": 0.21,
      "grad_norm": 16.317060470581055,
      "learning_rate": 1.2685860031830648e-05,
      "loss": 1.6618,
      "step": 662
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.94815444946289,
      "learning_rate": 1.268173394657816e-05,
      "loss": 1.2826,
      "step": 663
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.04676055908203,
      "learning_rate": 1.267760252375166e-05,
      "loss": 4.4989,
      "step": 664
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.131474494934082,
      "learning_rate": 1.2673465767266773e-05,
      "loss": 1.3328,
      "step": 665
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.820852279663086,
      "learning_rate": 1.266932368104419e-05,
      "loss": 1.5155,
      "step": 666
    },
    {
      "epoch": 0.21,
      "grad_norm": 22.58245849609375,
      "learning_rate": 1.2665176269009639e-05,
      "loss": 1.598,
      "step": 667
    },
    {
      "epoch": 0.21,
      "grad_norm": 20.890731811523438,
      "learning_rate": 1.2661023535093905e-05,
      "loss": 3.1583,
      "step": 668
    },
    {
      "epoch": 0.21,
      "grad_norm": 16.39836311340332,
      "learning_rate": 1.2656865483232815e-05,
      "loss": 3.6485,
      "step": 669
    },
    {
      "epoch": 0.21,
      "grad_norm": 14.225503921508789,
      "learning_rate": 1.265270211736723e-05,
      "loss": 1.0935,
      "step": 670
    },
    {
      "epoch": 0.21,
      "grad_norm": 17.626684188842773,
      "learning_rate": 1.264853344144306e-05,
      "loss": 2.3145,
      "step": 671
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.931720733642578,
      "learning_rate": 1.2644359459411233e-05,
      "loss": 1.8545,
      "step": 672
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.659822463989258,
      "learning_rate": 1.2640180175227718e-05,
      "loss": 3.7829,
      "step": 673
    },
    {
      "epoch": 0.21,
      "grad_norm": 16.827177047729492,
      "learning_rate": 1.2635995592853505e-05,
      "loss": 2.8283,
      "step": 674
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.98625659942627,
      "learning_rate": 1.26318057162546e-05,
      "loss": 0.7866,
      "step": 675
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.108333587646484,
      "learning_rate": 1.2627610549402034e-05,
      "loss": 1.6244,
      "step": 676
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.255348205566406,
      "learning_rate": 1.2623410096271851e-05,
      "loss": 2.0722,
      "step": 677
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.850077629089355,
      "learning_rate": 1.2619204360845104e-05,
      "loss": 2.4558,
      "step": 678
    },
    {
      "epoch": 0.21,
      "grad_norm": 9.59620189666748,
      "learning_rate": 1.2614993347107847e-05,
      "loss": 1.6123,
      "step": 679
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.261030197143555,
      "learning_rate": 1.2610777059051148e-05,
      "loss": 2.0041,
      "step": 680
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.192977905273438,
      "learning_rate": 1.2606555500671064e-05,
      "loss": 1.6192,
      "step": 681
    },
    {
      "epoch": 0.21,
      "grad_norm": 16.09978675842285,
      "learning_rate": 1.2602328675968651e-05,
      "loss": 3.3949,
      "step": 682
    },
    {
      "epoch": 0.21,
      "grad_norm": 22.485755920410156,
      "learning_rate": 1.2598096588949958e-05,
      "loss": 3.2858,
      "step": 683
    },
    {
      "epoch": 0.21,
      "grad_norm": 19.057348251342773,
      "learning_rate": 1.2593859243626016e-05,
      "loss": 4.2207,
      "step": 684
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.834187507629395,
      "learning_rate": 1.2589616644012844e-05,
      "loss": 1.1678,
      "step": 685
    },
    {
      "epoch": 0.21,
      "grad_norm": 8.711618423461914,
      "learning_rate": 1.258536879413144e-05,
      "loss": 0.8833,
      "step": 686
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.831104278564453,
      "learning_rate": 1.258111569800778e-05,
      "loss": 1.3791,
      "step": 687
    },
    {
      "epoch": 0.21,
      "grad_norm": 11.20345687866211,
      "learning_rate": 1.2576857359672804e-05,
      "loss": 2.4059,
      "step": 688
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.189892768859863,
      "learning_rate": 1.2572593783162433e-05,
      "loss": 1.1859,
      "step": 689
    },
    {
      "epoch": 0.21,
      "grad_norm": 10.442441940307617,
      "learning_rate": 1.256832497251754e-05,
      "loss": 1.0376,
      "step": 690
    },
    {
      "epoch": 0.21,
      "grad_norm": 15.368931770324707,
      "learning_rate": 1.256405093178397e-05,
      "loss": 1.7561,
      "step": 691
    },
    {
      "epoch": 0.21,
      "grad_norm": 13.209639549255371,
      "learning_rate": 1.2559771665012516e-05,
      "loss": 2.3928,
      "step": 692
    },
    {
      "epoch": 0.21,
      "grad_norm": 12.749165534973145,
      "learning_rate": 1.2555487176258932e-05,
      "loss": 4.5768,
      "step": 693
    },
    {
      "epoch": 0.22,
      "grad_norm": 15.11466121673584,
      "learning_rate": 1.2551197469583912e-05,
      "loss": 3.2755,
      "step": 694
    },
    {
      "epoch": 0.22,
      "grad_norm": 14.278471946716309,
      "learning_rate": 1.25469025490531e-05,
      "loss": 1.6433,
      "step": 695
    },
    {
      "epoch": 0.22,
      "grad_norm": 17.323156356811523,
      "learning_rate": 1.2542602418737088e-05,
      "loss": 2.4764,
      "step": 696
    },
    {
      "epoch": 0.22,
      "grad_norm": 9.389161109924316,
      "learning_rate": 1.2538297082711393e-05,
      "loss": 1.1465,
      "step": 697
    },
    {
      "epoch": 0.22,
      "grad_norm": 19.667522430419922,
      "learning_rate": 1.2533986545056475e-05,
      "loss": 3.4705,
      "step": 698
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.19377326965332,
      "learning_rate": 1.2529670809857722e-05,
      "loss": 1.4053,
      "step": 699
    },
    {
      "epoch": 0.22,
      "grad_norm": 21.029172897338867,
      "learning_rate": 1.2525349881205445e-05,
      "loss": 1.9345,
      "step": 700
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.366401672363281,
      "learning_rate": 1.2521023763194882e-05,
      "loss": 2.8612,
      "step": 701
    },
    {
      "epoch": 0.22,
      "grad_norm": 13.524468421936035,
      "learning_rate": 1.2516692459926185e-05,
      "loss": 4.4014,
      "step": 702
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.896738052368164,
      "learning_rate": 1.2512355975504422e-05,
      "loss": 1.7042,
      "step": 703
    },
    {
      "epoch": 0.22,
      "grad_norm": 18.412193298339844,
      "learning_rate": 1.250801431403957e-05,
      "loss": 4.6858,
      "step": 704
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.425787925720215,
      "learning_rate": 1.250366747964652e-05,
      "loss": 1.6511,
      "step": 705
    },
    {
      "epoch": 0.22,
      "grad_norm": 16.743928909301758,
      "learning_rate": 1.2499315476445055e-05,
      "loss": 1.4475,
      "step": 706
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.760894775390625,
      "learning_rate": 1.2494958308559864e-05,
      "loss": 1.4689,
      "step": 707
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.353875160217285,
      "learning_rate": 1.2490595980120529e-05,
      "loss": 1.8056,
      "step": 708
    },
    {
      "epoch": 0.22,
      "grad_norm": 17.23015022277832,
      "learning_rate": 1.2486228495261522e-05,
      "loss": 2.3585,
      "step": 709
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.202420234680176,
      "learning_rate": 1.2481855858122201e-05,
      "loss": 2.073,
      "step": 710
    },
    {
      "epoch": 0.22,
      "grad_norm": 16.27634620666504,
      "learning_rate": 1.2477478072846814e-05,
      "loss": 2.0413,
      "step": 711
    },
    {
      "epoch": 0.22,
      "grad_norm": 7.528199672698975,
      "learning_rate": 1.2473095143584479e-05,
      "loss": 0.8323,
      "step": 712
    },
    {
      "epoch": 0.22,
      "grad_norm": 13.777722358703613,
      "learning_rate": 1.2468707074489194e-05,
      "loss": 1.9507,
      "step": 713
    },
    {
      "epoch": 0.22,
      "grad_norm": 10.845873832702637,
      "learning_rate": 1.2464313869719831e-05,
      "loss": 2.0242,
      "step": 714
    },
    {
      "epoch": 0.22,
      "grad_norm": 18.222116470336914,
      "learning_rate": 1.2459915533440124e-05,
      "loss": 1.6319,
      "step": 715
    },
    {
      "epoch": 0.22,
      "grad_norm": 11.19504451751709,
      "learning_rate": 1.2455512069818672e-05,
      "loss": 1.1383,
      "step": 716
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.988763809204102,
      "learning_rate": 1.2451103483028936e-05,
      "loss": 1.3111,
      "step": 717
    },
    {
      "epoch": 0.22,
      "grad_norm": 17.553518295288086,
      "learning_rate": 1.2446689777249232e-05,
      "loss": 2.2308,
      "step": 718
    },
    {
      "epoch": 0.22,
      "grad_norm": 20.2987003326416,
      "learning_rate": 1.2442270956662722e-05,
      "loss": 2.2082,
      "step": 719
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.516862869262695,
      "learning_rate": 1.2437847025457428e-05,
      "loss": 1.6221,
      "step": 720
    },
    {
      "epoch": 0.22,
      "grad_norm": 21.578792572021484,
      "learning_rate": 1.24334179878262e-05,
      "loss": 6.6848,
      "step": 721
    },
    {
      "epoch": 0.22,
      "grad_norm": 17.269962310791016,
      "learning_rate": 1.2428983847966741e-05,
      "loss": 2.9452,
      "step": 722
    },
    {
      "epoch": 0.22,
      "grad_norm": 24.642614364624023,
      "learning_rate": 1.242454461008158e-05,
      "loss": 4.4548,
      "step": 723
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.710604667663574,
      "learning_rate": 1.2420100278378088e-05,
      "loss": 1.6188,
      "step": 724
    },
    {
      "epoch": 0.22,
      "grad_norm": 12.620211601257324,
      "learning_rate": 1.241565085706845e-05,
      "loss": 1.69,
      "step": 725
    },
    {
      "epoch": 0.22,
      "grad_norm": 18.085607528686523,
      "learning_rate": 1.241119635036969e-05,
      "loss": 3.7645,
      "step": 726
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.195068359375,
      "learning_rate": 1.240673676250364e-05,
      "loss": 2.1039,
      "step": 727
    },
    {
      "epoch": 0.23,
      "grad_norm": 19.199190139770508,
      "learning_rate": 1.2402272097696953e-05,
      "loss": 3.1963,
      "step": 728
    },
    {
      "epoch": 0.23,
      "grad_norm": 16.378055572509766,
      "learning_rate": 1.2397802360181093e-05,
      "loss": 1.9512,
      "step": 729
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.149024963378906,
      "learning_rate": 1.2393327554192333e-05,
      "loss": 1.2123,
      "step": 730
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.235203742980957,
      "learning_rate": 1.2388847683971747e-05,
      "loss": 1.1037,
      "step": 731
    },
    {
      "epoch": 0.23,
      "grad_norm": 9.219218254089355,
      "learning_rate": 1.238436275376521e-05,
      "loss": 1.3665,
      "step": 732
    },
    {
      "epoch": 0.23,
      "grad_norm": 22.13263702392578,
      "learning_rate": 1.2379872767823393e-05,
      "loss": 3.0206,
      "step": 733
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.911103248596191,
      "learning_rate": 1.237537773040176e-05,
      "loss": 2.5949,
      "step": 734
    },
    {
      "epoch": 0.23,
      "grad_norm": 15.790095329284668,
      "learning_rate": 1.2370877645760557e-05,
      "loss": 3.5434,
      "step": 735
    },
    {
      "epoch": 0.23,
      "grad_norm": 18.366098403930664,
      "learning_rate": 1.2366372518164822e-05,
      "loss": 6.0189,
      "step": 736
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.551661491394043,
      "learning_rate": 1.2361862351884366e-05,
      "loss": 1.5336,
      "step": 737
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.93724250793457,
      "learning_rate": 1.2357347151193778e-05,
      "loss": 1.6109,
      "step": 738
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.013720512390137,
      "learning_rate": 1.2352826920372418e-05,
      "loss": 1.5895,
      "step": 739
    },
    {
      "epoch": 0.23,
      "grad_norm": 10.229809761047363,
      "learning_rate": 1.2348301663704419e-05,
      "loss": 1.3494,
      "step": 740
    },
    {
      "epoch": 0.23,
      "grad_norm": 16.16636085510254,
      "learning_rate": 1.2343771385478661e-05,
      "loss": 2.0953,
      "step": 741
    },
    {
      "epoch": 0.23,
      "grad_norm": 16.416259765625,
      "learning_rate": 1.2339236089988801e-05,
      "loss": 3.5324,
      "step": 742
    },
    {
      "epoch": 0.23,
      "grad_norm": 20.941978454589844,
      "learning_rate": 1.2334695781533243e-05,
      "loss": 4.7186,
      "step": 743
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.121553421020508,
      "learning_rate": 1.2330150464415144e-05,
      "loss": 1.2178,
      "step": 744
    },
    {
      "epoch": 0.23,
      "grad_norm": 12.87922477722168,
      "learning_rate": 1.2325600142942406e-05,
      "loss": 1.7759,
      "step": 745
    },
    {
      "epoch": 0.23,
      "grad_norm": 20.307462692260742,
      "learning_rate": 1.2321044821427681e-05,
      "loss": 3.7992,
      "step": 746
    },
    {
      "epoch": 0.23,
      "grad_norm": 15.959997177124023,
      "learning_rate": 1.2316484504188348e-05,
      "loss": 1.5344,
      "step": 747
    },
    {
      "epoch": 0.23,
      "grad_norm": 16.901805877685547,
      "learning_rate": 1.231191919554653e-05,
      "loss": 2.1397,
      "step": 748
    },
    {
      "epoch": 0.23,
      "grad_norm": 28.746898651123047,
      "learning_rate": 1.2307348899829073e-05,
      "loss": 2.3627,
      "step": 749
    },
    {
      "epoch": 0.23,
      "grad_norm": 18.00472640991211,
      "learning_rate": 1.2302773621367562e-05,
      "loss": 2.5594,
      "step": 750
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.613627433776855,
      "learning_rate": 1.2298193364498294e-05,
      "loss": 2.1991,
      "step": 751
    },
    {
      "epoch": 0.23,
      "grad_norm": 15.26014518737793,
      "learning_rate": 1.2293608133562285e-05,
      "loss": 4.995,
      "step": 752
    },
    {
      "epoch": 0.23,
      "grad_norm": 11.997806549072266,
      "learning_rate": 1.2289017932905271e-05,
      "loss": 2.2007,
      "step": 753
    },
    {
      "epoch": 0.23,
      "grad_norm": 13.420331954956055,
      "learning_rate": 1.2284422766877692e-05,
      "loss": 2.6394,
      "step": 754
    },
    {
      "epoch": 0.23,
      "grad_norm": 14.628358840942383,
      "learning_rate": 1.2279822639834698e-05,
      "loss": 4.2071,
      "step": 755
    },
    {
      "epoch": 0.23,
      "grad_norm": 14.1344575881958,
      "learning_rate": 1.2275217556136138e-05,
      "loss": 1.6072,
      "step": 756
    },
    {
      "epoch": 0.23,
      "grad_norm": 15.807295799255371,
      "learning_rate": 1.2270607520146563e-05,
      "loss": 6.007,
      "step": 757
    },
    {
      "epoch": 0.23,
      "grad_norm": 6.969516754150391,
      "learning_rate": 1.2265992536235213e-05,
      "loss": 0.7291,
      "step": 758
    },
    {
      "epoch": 0.24,
      "grad_norm": 16.10032081604004,
      "learning_rate": 1.2261372608776021e-05,
      "loss": 3.4565,
      "step": 759
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.56027889251709,
      "learning_rate": 1.2256747742147601e-05,
      "loss": 1.2174,
      "step": 760
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.300409317016602,
      "learning_rate": 1.2252117940733255e-05,
      "loss": 1.1443,
      "step": 761
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.430366516113281,
      "learning_rate": 1.2247483208920955e-05,
      "loss": 1.8809,
      "step": 762
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.15425968170166,
      "learning_rate": 1.2242843551103349e-05,
      "loss": 0.8325,
      "step": 763
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.97874641418457,
      "learning_rate": 1.2238198971677754e-05,
      "loss": 6.974,
      "step": 764
    },
    {
      "epoch": 0.24,
      "grad_norm": 8.714863777160645,
      "learning_rate": 1.2233549475046153e-05,
      "loss": 0.8588,
      "step": 765
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.327173233032227,
      "learning_rate": 1.2228895065615184e-05,
      "loss": 2.2164,
      "step": 766
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.299591064453125,
      "learning_rate": 1.2224235747796147e-05,
      "loss": 2.6141,
      "step": 767
    },
    {
      "epoch": 0.24,
      "grad_norm": 16.821212768554688,
      "learning_rate": 1.221957152600499e-05,
      "loss": 2.2579,
      "step": 768
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.770112991333008,
      "learning_rate": 1.2214902404662314e-05,
      "loss": 2.2605,
      "step": 769
    },
    {
      "epoch": 0.24,
      "grad_norm": 15.763861656188965,
      "learning_rate": 1.2210228388193355e-05,
      "loss": 2.2584,
      "step": 770
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.93879222869873,
      "learning_rate": 1.2205549481027998e-05,
      "loss": 1.5642,
      "step": 771
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.689042091369629,
      "learning_rate": 1.2200865687600757e-05,
      "loss": 1.1944,
      "step": 772
    },
    {
      "epoch": 0.24,
      "grad_norm": 19.434879302978516,
      "learning_rate": 1.219617701235078e-05,
      "loss": 4.4623,
      "step": 773
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.151241302490234,
      "learning_rate": 1.2191483459721837e-05,
      "loss": 1.1613,
      "step": 774
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.30809211730957,
      "learning_rate": 1.2186785034162328e-05,
      "loss": 1.4964,
      "step": 775
    },
    {
      "epoch": 0.24,
      "grad_norm": 20.536462783813477,
      "learning_rate": 1.2182081740125266e-05,
      "loss": 7.906,
      "step": 776
    },
    {
      "epoch": 0.24,
      "grad_norm": 17.299827575683594,
      "learning_rate": 1.2177373582068284e-05,
      "loss": 3.6637,
      "step": 777
    },
    {
      "epoch": 0.24,
      "grad_norm": 16.93096923828125,
      "learning_rate": 1.2172660564453615e-05,
      "loss": 2.4282,
      "step": 778
    },
    {
      "epoch": 0.24,
      "grad_norm": 11.090471267700195,
      "learning_rate": 1.2167942691748105e-05,
      "loss": 1.3211,
      "step": 779
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.850017547607422,
      "learning_rate": 1.2163219968423203e-05,
      "loss": 2.5532,
      "step": 780
    },
    {
      "epoch": 0.24,
      "grad_norm": 18.67405891418457,
      "learning_rate": 1.215849239895495e-05,
      "loss": 2.0105,
      "step": 781
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.775568008422852,
      "learning_rate": 1.2153759987823984e-05,
      "loss": 0.8751,
      "step": 782
    },
    {
      "epoch": 0.24,
      "grad_norm": 19.743520736694336,
      "learning_rate": 1.2149022739515528e-05,
      "loss": 1.1554,
      "step": 783
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.606451034545898,
      "learning_rate": 1.2144280658519392e-05,
      "loss": 1.904,
      "step": 784
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.49717903137207,
      "learning_rate": 1.2139533749329968e-05,
      "loss": 1.2586,
      "step": 785
    },
    {
      "epoch": 0.24,
      "grad_norm": 14.731224060058594,
      "learning_rate": 1.2134782016446219e-05,
      "loss": 2.3142,
      "step": 786
    },
    {
      "epoch": 0.24,
      "grad_norm": 10.49559211730957,
      "learning_rate": 1.2130025464371686e-05,
      "loss": 1.1727,
      "step": 787
    },
    {
      "epoch": 0.24,
      "grad_norm": 13.33672046661377,
      "learning_rate": 1.2125264097614471e-05,
      "loss": 1.6586,
      "step": 788
    },
    {
      "epoch": 0.24,
      "grad_norm": 12.418742179870605,
      "learning_rate": 1.2120497920687241e-05,
      "loss": 1.5077,
      "step": 789
    },
    {
      "epoch": 0.24,
      "grad_norm": 15.242766380310059,
      "learning_rate": 1.2115726938107232e-05,
      "loss": 2.5938,
      "step": 790
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.032227516174316,
      "learning_rate": 1.2110951154396216e-05,
      "loss": 1.9316,
      "step": 791
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.62277603149414,
      "learning_rate": 1.210617057408053e-05,
      "loss": 2.3232,
      "step": 792
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.518356323242188,
      "learning_rate": 1.210138520169105e-05,
      "loss": 2.179,
      "step": 793
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.969557762145996,
      "learning_rate": 1.20965950417632e-05,
      "loss": 1.8217,
      "step": 794
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.493850708007812,
      "learning_rate": 1.2091800098836934e-05,
      "loss": 4.8096,
      "step": 795
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.271809577941895,
      "learning_rate": 1.2087000377456743e-05,
      "loss": 1.2367,
      "step": 796
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.41793441772461,
      "learning_rate": 1.2082195882171647e-05,
      "loss": 5.0848,
      "step": 797
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.58703327178955,
      "learning_rate": 1.2077386617535193e-05,
      "loss": 1.7298,
      "step": 798
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.00687026977539,
      "learning_rate": 1.2072572588105442e-05,
      "loss": 2.0278,
      "step": 799
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.369733810424805,
      "learning_rate": 1.2067753798444974e-05,
      "loss": 1.6132,
      "step": 800
    },
    {
      "epoch": 0.25,
      "grad_norm": 9.346023559570312,
      "learning_rate": 1.2062930253120885e-05,
      "loss": 1.0617,
      "step": 801
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.150537490844727,
      "learning_rate": 1.2058101956704766e-05,
      "loss": 3.207,
      "step": 802
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.184139251708984,
      "learning_rate": 1.205326891377273e-05,
      "loss": 2.0503,
      "step": 803
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.664175987243652,
      "learning_rate": 1.204843112890537e-05,
      "loss": 1.9163,
      "step": 804
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.745572090148926,
      "learning_rate": 1.204358860668778e-05,
      "loss": 1.4228,
      "step": 805
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.390905380249023,
      "learning_rate": 1.2038741351709551e-05,
      "loss": 1.3138,
      "step": 806
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.42654800415039,
      "learning_rate": 1.203388936856475e-05,
      "loss": 1.5961,
      "step": 807
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.617491722106934,
      "learning_rate": 1.2029032661851928e-05,
      "loss": 1.5742,
      "step": 808
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.204427719116211,
      "learning_rate": 1.2024171236174117e-05,
      "loss": 4.9142,
      "step": 809
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.922470092773438,
      "learning_rate": 1.2019305096138813e-05,
      "loss": 3.9775,
      "step": 810
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.659815788269043,
      "learning_rate": 1.2014434246357985e-05,
      "loss": 2.4097,
      "step": 811
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.354887008666992,
      "learning_rate": 1.2009558691448071e-05,
      "loss": 1.5156,
      "step": 812
    },
    {
      "epoch": 0.25,
      "grad_norm": 12.20242691040039,
      "learning_rate": 1.200467843602996e-05,
      "loss": 4.2651,
      "step": 813
    },
    {
      "epoch": 0.25,
      "grad_norm": 13.313082695007324,
      "learning_rate": 1.1999793484729e-05,
      "loss": 2.9641,
      "step": 814
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.646648406982422,
      "learning_rate": 1.1994903842174985e-05,
      "loss": 4.5349,
      "step": 815
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.043517112731934,
      "learning_rate": 1.1990009513002166e-05,
      "loss": 3.5675,
      "step": 816
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.295183181762695,
      "learning_rate": 1.198511050184922e-05,
      "loss": 1.8846,
      "step": 817
    },
    {
      "epoch": 0.25,
      "grad_norm": 10.43750286102295,
      "learning_rate": 1.198020681335928e-05,
      "loss": 1.2817,
      "step": 818
    },
    {
      "epoch": 0.25,
      "grad_norm": 15.210440635681152,
      "learning_rate": 1.1975298452179894e-05,
      "loss": 3.1719,
      "step": 819
    },
    {
      "epoch": 0.25,
      "grad_norm": 19.98702621459961,
      "learning_rate": 1.1970385422963052e-05,
      "loss": 1.5404,
      "step": 820
    },
    {
      "epoch": 0.25,
      "grad_norm": 14.712196350097656,
      "learning_rate": 1.1965467730365157e-05,
      "loss": 6.1263,
      "step": 821
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.687849044799805,
      "learning_rate": 1.1960545379047047e-05,
      "loss": 2.2848,
      "step": 822
    },
    {
      "epoch": 0.25,
      "grad_norm": 16.070829391479492,
      "learning_rate": 1.1955618373673957e-05,
      "loss": 1.9679,
      "step": 823
    },
    {
      "epoch": 0.26,
      "grad_norm": 14.484185218811035,
      "learning_rate": 1.1950686718915552e-05,
      "loss": 2.8716,
      "step": 824
    },
    {
      "epoch": 0.26,
      "grad_norm": 11.286521911621094,
      "learning_rate": 1.1945750419445884e-05,
      "loss": 2.4105,
      "step": 825
    },
    {
      "epoch": 0.26,
      "grad_norm": 11.694199562072754,
      "learning_rate": 1.1940809479943422e-05,
      "loss": 0.9597,
      "step": 826
    },
    {
      "epoch": 0.26,
      "grad_norm": 21.48593521118164,
      "learning_rate": 1.1935863905091026e-05,
      "loss": 5.8054,
      "step": 827
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.373125076293945,
      "learning_rate": 1.1930913699575955e-05,
      "loss": 3.0395,
      "step": 828
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.502968788146973,
      "learning_rate": 1.1925958868089847e-05,
      "loss": 1.4822,
      "step": 829
    },
    {
      "epoch": 0.26,
      "grad_norm": 14.944729804992676,
      "learning_rate": 1.1920999415328728e-05,
      "loss": 1.9248,
      "step": 830
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.607522964477539,
      "learning_rate": 1.1916035345993015e-05,
      "loss": 3.673,
      "step": 831
    },
    {
      "epoch": 0.26,
      "grad_norm": 11.033047676086426,
      "learning_rate": 1.1911066664787484e-05,
      "loss": 1.2412,
      "step": 832
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.628982543945312,
      "learning_rate": 1.1906093376421291e-05,
      "loss": 1.7389,
      "step": 833
    },
    {
      "epoch": 0.26,
      "grad_norm": 20.70766258239746,
      "learning_rate": 1.1901115485607957e-05,
      "loss": 3.0788,
      "step": 834
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.678114891052246,
      "learning_rate": 1.1896132997065364e-05,
      "loss": 1.2587,
      "step": 835
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.934436798095703,
      "learning_rate": 1.1891145915515754e-05,
      "loss": 2.8138,
      "step": 836
    },
    {
      "epoch": 0.26,
      "grad_norm": 15.426923751831055,
      "learning_rate": 1.188615424568572e-05,
      "loss": 1.9037,
      "step": 837
    },
    {
      "epoch": 0.26,
      "grad_norm": 11.262444496154785,
      "learning_rate": 1.1881157992306204e-05,
      "loss": 1.4626,
      "step": 838
    },
    {
      "epoch": 0.26,
      "grad_norm": 19.111431121826172,
      "learning_rate": 1.187615716011249e-05,
      "loss": 4.544,
      "step": 839
    },
    {
      "epoch": 0.26,
      "grad_norm": 15.529975891113281,
      "learning_rate": 1.1871151753844207e-05,
      "loss": 3.2797,
      "step": 840
    },
    {
      "epoch": 0.26,
      "grad_norm": 18.435951232910156,
      "learning_rate": 1.1866141778245314e-05,
      "loss": 2.6301,
      "step": 841
    },
    {
      "epoch": 0.26,
      "grad_norm": 16.66749382019043,
      "learning_rate": 1.1861127238064103e-05,
      "loss": 2.8497,
      "step": 842
    },
    {
      "epoch": 0.26,
      "grad_norm": 18.804611206054688,
      "learning_rate": 1.1856108138053194e-05,
      "loss": 5.6834,
      "step": 843
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.824563026428223,
      "learning_rate": 1.1851084482969523e-05,
      "loss": 5.7809,
      "step": 844
    },
    {
      "epoch": 0.26,
      "grad_norm": 14.626198768615723,
      "learning_rate": 1.1846056277574352e-05,
      "loss": 3.5987,
      "step": 845
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.569183349609375,
      "learning_rate": 1.1841023526633243e-05,
      "loss": 1.5252,
      "step": 846
    },
    {
      "epoch": 0.26,
      "grad_norm": 17.080930709838867,
      "learning_rate": 1.1835986234916079e-05,
      "loss": 4.205,
      "step": 847
    },
    {
      "epoch": 0.26,
      "grad_norm": 18.488704681396484,
      "learning_rate": 1.183094440719704e-05,
      "loss": 2.2896,
      "step": 848
    },
    {
      "epoch": 0.26,
      "grad_norm": 10.405138969421387,
      "learning_rate": 1.1825898048254604e-05,
      "loss": 2.4845,
      "step": 849
    },
    {
      "epoch": 0.26,
      "grad_norm": 13.465655326843262,
      "learning_rate": 1.1820847162871548e-05,
      "loss": 2.3561,
      "step": 850
    },
    {
      "epoch": 0.26,
      "grad_norm": 12.562773704528809,
      "learning_rate": 1.181579175583493e-05,
      "loss": 2.2778,
      "step": 851
    },
    {
      "epoch": 0.26,
      "grad_norm": 16.159759521484375,
      "learning_rate": 1.1810731831936115e-05,
      "loss": 4.7586,
      "step": 852
    },
    {
      "epoch": 0.26,
      "grad_norm": 14.529434204101562,
      "learning_rate": 1.1805667395970719e-05,
      "loss": 1.5514,
      "step": 853
    },
    {
      "epoch": 0.26,
      "grad_norm": 16.40343475341797,
      "learning_rate": 1.1800598452738656e-05,
      "loss": 2.4718,
      "step": 854
    },
    {
      "epoch": 0.26,
      "grad_norm": 9.03700065612793,
      "learning_rate": 1.1795525007044106e-05,
      "loss": 0.8431,
      "step": 855
    },
    {
      "epoch": 0.27,
      "grad_norm": 20.445934295654297,
      "learning_rate": 1.1790447063695516e-05,
      "loss": 1.8509,
      "step": 856
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.171225547790527,
      "learning_rate": 1.178536462750559e-05,
      "loss": 1.9565,
      "step": 857
    },
    {
      "epoch": 0.27,
      "grad_norm": 16.317176818847656,
      "learning_rate": 1.1780277703291302e-05,
      "loss": 2.3285,
      "step": 858
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.370471954345703,
      "learning_rate": 1.1775186295873869e-05,
      "loss": 1.3541,
      "step": 859
    },
    {
      "epoch": 0.27,
      "grad_norm": 16.811317443847656,
      "learning_rate": 1.177009041007876e-05,
      "loss": 2.7627,
      "step": 860
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.549324035644531,
      "learning_rate": 1.1764990050735694e-05,
      "loss": 2.925,
      "step": 861
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.653358459472656,
      "learning_rate": 1.1759885222678624e-05,
      "loss": 1.5885,
      "step": 862
    },
    {
      "epoch": 0.27,
      "grad_norm": 12.484577178955078,
      "learning_rate": 1.1754775930745738e-05,
      "loss": 4.57,
      "step": 863
    },
    {
      "epoch": 0.27,
      "grad_norm": 10.131400108337402,
      "learning_rate": 1.1749662179779458e-05,
      "loss": 2.021,
      "step": 864
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.61589241027832,
      "learning_rate": 1.1744543974626428e-05,
      "loss": 4.4905,
      "step": 865
    },
    {
      "epoch": 0.27,
      "grad_norm": 24.758455276489258,
      "learning_rate": 1.1739421320137521e-05,
      "loss": 2.7249,
      "step": 866
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.530740737915039,
      "learning_rate": 1.173429422116782e-05,
      "loss": 1.3571,
      "step": 867
    },
    {
      "epoch": 0.27,
      "grad_norm": 22.10980796813965,
      "learning_rate": 1.172916268257662e-05,
      "loss": 6.113,
      "step": 868
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.13565444946289,
      "learning_rate": 1.1724026709227428e-05,
      "loss": 2.1237,
      "step": 869
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.099780082702637,
      "learning_rate": 1.1718886305987952e-05,
      "loss": 1.5074,
      "step": 870
    },
    {
      "epoch": 0.27,
      "grad_norm": 12.233208656311035,
      "learning_rate": 1.17137414777301e-05,
      "loss": 1.646,
      "step": 871
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.301345825195312,
      "learning_rate": 1.170859222932997e-05,
      "loss": 0.8234,
      "step": 872
    },
    {
      "epoch": 0.27,
      "grad_norm": 15.48371696472168,
      "learning_rate": 1.1703438565667858e-05,
      "loss": 2.4947,
      "step": 873
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.66010856628418,
      "learning_rate": 1.1698280491628229e-05,
      "loss": 4.3986,
      "step": 874
    },
    {
      "epoch": 0.27,
      "grad_norm": 16.98670768737793,
      "learning_rate": 1.1693118012099745e-05,
      "loss": 3.4212,
      "step": 875
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.71246337890625,
      "learning_rate": 1.1687951131975234e-05,
      "loss": 0.7751,
      "step": 876
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.071887016296387,
      "learning_rate": 1.1682779856151694e-05,
      "loss": 1.2658,
      "step": 877
    },
    {
      "epoch": 0.27,
      "grad_norm": 9.613807678222656,
      "learning_rate": 1.1677604189530297e-05,
      "loss": 2.1093,
      "step": 878
    },
    {
      "epoch": 0.27,
      "grad_norm": 18.976409912109375,
      "learning_rate": 1.1672424137016369e-05,
      "loss": 2.9181,
      "step": 879
    },
    {
      "epoch": 0.27,
      "grad_norm": 17.42270851135254,
      "learning_rate": 1.1667239703519393e-05,
      "loss": 3.4419,
      "step": 880
    },
    {
      "epoch": 0.27,
      "grad_norm": 18.03681755065918,
      "learning_rate": 1.1662050893953007e-05,
      "loss": 3.1427,
      "step": 881
    },
    {
      "epoch": 0.27,
      "grad_norm": 8.601777076721191,
      "learning_rate": 1.1656857713234999e-05,
      "loss": 0.9373,
      "step": 882
    },
    {
      "epoch": 0.27,
      "grad_norm": 14.285597801208496,
      "learning_rate": 1.165166016628729e-05,
      "loss": 1.3633,
      "step": 883
    },
    {
      "epoch": 0.27,
      "grad_norm": 12.230426788330078,
      "learning_rate": 1.1646458258035953e-05,
      "loss": 1.0753,
      "step": 884
    },
    {
      "epoch": 0.27,
      "grad_norm": 18.180387496948242,
      "learning_rate": 1.1641251993411182e-05,
      "loss": 2.7476,
      "step": 885
    },
    {
      "epoch": 0.27,
      "grad_norm": 20.50904655456543,
      "learning_rate": 1.1636041377347309e-05,
      "loss": 3.8631,
      "step": 886
    },
    {
      "epoch": 0.27,
      "grad_norm": 13.58751106262207,
      "learning_rate": 1.1630826414782781e-05,
      "loss": 2.4704,
      "step": 887
    },
    {
      "epoch": 0.28,
      "grad_norm": 13.96949291229248,
      "learning_rate": 1.1625607110660176e-05,
      "loss": 2.2461,
      "step": 888
    },
    {
      "epoch": 0.28,
      "grad_norm": 15.626188278198242,
      "learning_rate": 1.1620383469926176e-05,
      "loss": 2.2847,
      "step": 889
    },
    {
      "epoch": 0.28,
      "grad_norm": 17.77653694152832,
      "learning_rate": 1.161515549753158e-05,
      "loss": 4.4198,
      "step": 890
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.263466835021973,
      "learning_rate": 1.1609923198431286e-05,
      "loss": 0.9664,
      "step": 891
    },
    {
      "epoch": 0.28,
      "grad_norm": 20.309823989868164,
      "learning_rate": 1.1604686577584302e-05,
      "loss": 2.1832,
      "step": 892
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.75548267364502,
      "learning_rate": 1.1599445639953723e-05,
      "loss": 2.443,
      "step": 893
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.767431259155273,
      "learning_rate": 1.159420039050674e-05,
      "loss": 1.0003,
      "step": 894
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.956649780273438,
      "learning_rate": 1.158895083421463e-05,
      "loss": 3.4096,
      "step": 895
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.170507431030273,
      "learning_rate": 1.1583696976052751e-05,
      "loss": 1.2359,
      "step": 896
    },
    {
      "epoch": 0.28,
      "grad_norm": 21.288522720336914,
      "learning_rate": 1.157843882100054e-05,
      "loss": 6.4301,
      "step": 897
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.635541915893555,
      "learning_rate": 1.1573176374041503e-05,
      "loss": 3.6166,
      "step": 898
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.411420822143555,
      "learning_rate": 1.1567909640163217e-05,
      "loss": 2.014,
      "step": 899
    },
    {
      "epoch": 0.28,
      "grad_norm": 9.194472312927246,
      "learning_rate": 1.1562638624357319e-05,
      "loss": 1.1898,
      "step": 900
    },
    {
      "epoch": 0.28,
      "grad_norm": 17.868114471435547,
      "learning_rate": 1.1557363331619512e-05,
      "loss": 2.3214,
      "step": 901
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.548218727111816,
      "learning_rate": 1.1552083766949537e-05,
      "loss": 1.8704,
      "step": 902
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.413217544555664,
      "learning_rate": 1.1546799935351202e-05,
      "loss": 1.5204,
      "step": 903
    },
    {
      "epoch": 0.28,
      "grad_norm": 17.646013259887695,
      "learning_rate": 1.1541511841832348e-05,
      "loss": 2.3315,
      "step": 904
    },
    {
      "epoch": 0.28,
      "grad_norm": 15.216915130615234,
      "learning_rate": 1.1536219491404855e-05,
      "loss": 2.9786,
      "step": 905
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.426615715026855,
      "learning_rate": 1.1530922889084643e-05,
      "loss": 2.4699,
      "step": 906
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.518457412719727,
      "learning_rate": 1.1525622039891659e-05,
      "loss": 5.3688,
      "step": 907
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.908069610595703,
      "learning_rate": 1.1520316948849874e-05,
      "loss": 6.7208,
      "step": 908
    },
    {
      "epoch": 0.28,
      "grad_norm": 14.2049560546875,
      "learning_rate": 1.1515007620987282e-05,
      "loss": 1.9017,
      "step": 909
    },
    {
      "epoch": 0.28,
      "grad_norm": 16.437877655029297,
      "learning_rate": 1.1509694061335893e-05,
      "loss": 1.4594,
      "step": 910
    },
    {
      "epoch": 0.28,
      "grad_norm": 11.180994987487793,
      "learning_rate": 1.1504376274931718e-05,
      "loss": 2.1785,
      "step": 911
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.00964641571045,
      "learning_rate": 1.149905426681479e-05,
      "loss": 1.1223,
      "step": 912
    },
    {
      "epoch": 0.28,
      "grad_norm": 20.051992416381836,
      "learning_rate": 1.1493728042029132e-05,
      "loss": 2.0832,
      "step": 913
    },
    {
      "epoch": 0.28,
      "grad_norm": 7.1900129318237305,
      "learning_rate": 1.1488397605622767e-05,
      "loss": 0.883,
      "step": 914
    },
    {
      "epoch": 0.28,
      "grad_norm": 12.989784240722656,
      "learning_rate": 1.1483062962647708e-05,
      "loss": 5.7096,
      "step": 915
    },
    {
      "epoch": 0.28,
      "grad_norm": 15.41681957244873,
      "learning_rate": 1.1477724118159955e-05,
      "loss": 1.1806,
      "step": 916
    },
    {
      "epoch": 0.28,
      "grad_norm": 13.027620315551758,
      "learning_rate": 1.1472381077219493e-05,
      "loss": 2.3941,
      "step": 917
    },
    {
      "epoch": 0.28,
      "grad_norm": 10.86893367767334,
      "learning_rate": 1.1467033844890278e-05,
      "loss": 1.9049,
      "step": 918
    },
    {
      "epoch": 0.28,
      "grad_norm": 17.04452896118164,
      "learning_rate": 1.146168242624025e-05,
      "loss": 3.0468,
      "step": 919
    },
    {
      "epoch": 0.29,
      "grad_norm": 15.610978126525879,
      "learning_rate": 1.1456326826341303e-05,
      "loss": 2.4455,
      "step": 920
    },
    {
      "epoch": 0.29,
      "grad_norm": 17.980670928955078,
      "learning_rate": 1.1450967050269304e-05,
      "loss": 2.5141,
      "step": 921
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.66197681427002,
      "learning_rate": 1.1445603103104072e-05,
      "loss": 1.2295,
      "step": 922
    },
    {
      "epoch": 0.29,
      "grad_norm": 9.646084785461426,
      "learning_rate": 1.1440234989929381e-05,
      "loss": 1.0995,
      "step": 923
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.297891616821289,
      "learning_rate": 1.1434862715832955e-05,
      "loss": 1.1596,
      "step": 924
    },
    {
      "epoch": 0.29,
      "grad_norm": 16.568077087402344,
      "learning_rate": 1.1429486285906462e-05,
      "loss": 5.4638,
      "step": 925
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.074769973754883,
      "learning_rate": 1.1424105705245503e-05,
      "loss": 3.3299,
      "step": 926
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.440634727478027,
      "learning_rate": 1.1418720978949622e-05,
      "loss": 2.032,
      "step": 927
    },
    {
      "epoch": 0.29,
      "grad_norm": 15.466567993164062,
      "learning_rate": 1.1413332112122284e-05,
      "loss": 1.7181,
      "step": 928
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.589716911315918,
      "learning_rate": 1.140793910987088e-05,
      "loss": 1.3699,
      "step": 929
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.145682334899902,
      "learning_rate": 1.1402541977306725e-05,
      "loss": 0.8361,
      "step": 930
    },
    {
      "epoch": 0.29,
      "grad_norm": 17.095247268676758,
      "learning_rate": 1.1397140719545047e-05,
      "loss": 3.4324,
      "step": 931
    },
    {
      "epoch": 0.29,
      "grad_norm": 16.94806480407715,
      "learning_rate": 1.1391735341704977e-05,
      "loss": 3.3401,
      "step": 932
    },
    {
      "epoch": 0.29,
      "grad_norm": 11.145195007324219,
      "learning_rate": 1.1386325848909559e-05,
      "loss": 2.567,
      "step": 933
    },
    {
      "epoch": 0.29,
      "grad_norm": 16.297367095947266,
      "learning_rate": 1.1380912246285735e-05,
      "loss": 2.6329,
      "step": 934
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.540289878845215,
      "learning_rate": 1.1375494538964339e-05,
      "loss": 1.7408,
      "step": 935
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.331464767456055,
      "learning_rate": 1.13700727320801e-05,
      "loss": 4.355,
      "step": 936
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.369190216064453,
      "learning_rate": 1.1364646830771628e-05,
      "loss": 3.484,
      "step": 937
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.361828804016113,
      "learning_rate": 1.1359216840181416e-05,
      "loss": 1.7254,
      "step": 938
    },
    {
      "epoch": 0.29,
      "grad_norm": 13.768353462219238,
      "learning_rate": 1.1353782765455831e-05,
      "loss": 4.9616,
      "step": 939
    },
    {
      "epoch": 0.29,
      "grad_norm": 15.60986614227295,
      "learning_rate": 1.1348344611745116e-05,
      "loss": 2.9609,
      "step": 940
    },
    {
      "epoch": 0.29,
      "grad_norm": 12.908799171447754,
      "learning_rate": 1.1342902384203372e-05,
      "loss": 1.27,
      "step": 941
    },
    {
      "epoch": 0.29,
      "grad_norm": 12.849720001220703,
      "learning_rate": 1.1337456087988565e-05,
      "loss": 2.6348,
      "step": 942
    },
    {
      "epoch": 0.29,
      "grad_norm": 17.09627342224121,
      "learning_rate": 1.1332005728262522e-05,
      "loss": 5.6902,
      "step": 943
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.49213981628418,
      "learning_rate": 1.1326551310190911e-05,
      "loss": 1.6267,
      "step": 944
    },
    {
      "epoch": 0.29,
      "grad_norm": 14.392061233520508,
      "learning_rate": 1.1321092838943253e-05,
      "loss": 2.2689,
      "step": 945
    },
    {
      "epoch": 0.29,
      "grad_norm": 15.092098236083984,
      "learning_rate": 1.1315630319692912e-05,
      "loss": 1.7458,
      "step": 946
    },
    {
      "epoch": 0.29,
      "grad_norm": 12.850863456726074,
      "learning_rate": 1.1310163757617084e-05,
      "loss": 1.6304,
      "step": 947
    },
    {
      "epoch": 0.29,
      "grad_norm": 16.881410598754883,
      "learning_rate": 1.1304693157896797e-05,
      "loss": 2.6307,
      "step": 948
    },
    {
      "epoch": 0.29,
      "grad_norm": 10.893935203552246,
      "learning_rate": 1.1299218525716909e-05,
      "loss": 2.6573,
      "step": 949
    },
    {
      "epoch": 0.29,
      "grad_norm": 21.684673309326172,
      "learning_rate": 1.1293739866266097e-05,
      "loss": 2.6042,
      "step": 950
    },
    {
      "epoch": 0.29,
      "grad_norm": 24.443126678466797,
      "learning_rate": 1.128825718473686e-05,
      "loss": 8.0742,
      "step": 951
    },
    {
      "epoch": 0.29,
      "grad_norm": 12.001096725463867,
      "learning_rate": 1.12827704863255e-05,
      "loss": 1.7396,
      "step": 952
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.469615936279297,
      "learning_rate": 1.127727977623213e-05,
      "loss": 2.2511,
      "step": 953
    },
    {
      "epoch": 0.3,
      "grad_norm": 15.373016357421875,
      "learning_rate": 1.127178505966067e-05,
      "loss": 2.7536,
      "step": 954
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.056225776672363,
      "learning_rate": 1.1266286341818836e-05,
      "loss": 1.0603,
      "step": 955
    },
    {
      "epoch": 0.3,
      "grad_norm": 12.561933517456055,
      "learning_rate": 1.1260783627918125e-05,
      "loss": 1.4368,
      "step": 956
    },
    {
      "epoch": 0.3,
      "grad_norm": 19.184635162353516,
      "learning_rate": 1.125527692317384e-05,
      "loss": 1.4412,
      "step": 957
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.610422134399414,
      "learning_rate": 1.1249766232805048e-05,
      "loss": 1.268,
      "step": 958
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.994367599487305,
      "learning_rate": 1.1244251562034607e-05,
      "loss": 1.3654,
      "step": 959
    },
    {
      "epoch": 0.3,
      "grad_norm": 12.45901870727539,
      "learning_rate": 1.1238732916089138e-05,
      "loss": 3.1829,
      "step": 960
    },
    {
      "epoch": 0.3,
      "grad_norm": 16.608116149902344,
      "learning_rate": 1.123321030019904e-05,
      "loss": 3.3567,
      "step": 961
    },
    {
      "epoch": 0.3,
      "grad_norm": 16.66891098022461,
      "learning_rate": 1.1227683719598463e-05,
      "loss": 2.76,
      "step": 962
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.403820037841797,
      "learning_rate": 1.1222153179525323e-05,
      "loss": 1.8815,
      "step": 963
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.18431568145752,
      "learning_rate": 1.1216618685221285e-05,
      "loss": 1.7656,
      "step": 964
    },
    {
      "epoch": 0.3,
      "grad_norm": 15.747517585754395,
      "learning_rate": 1.1211080241931762e-05,
      "loss": 4.782,
      "step": 965
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.531620979309082,
      "learning_rate": 1.120553785490591e-05,
      "loss": 2.3307,
      "step": 966
    },
    {
      "epoch": 0.3,
      "grad_norm": 12.81969165802002,
      "learning_rate": 1.1199991529396624e-05,
      "loss": 1.9334,
      "step": 967
    },
    {
      "epoch": 0.3,
      "grad_norm": 16.989093780517578,
      "learning_rate": 1.1194441270660528e-05,
      "loss": 3.0846,
      "step": 968
    },
    {
      "epoch": 0.3,
      "grad_norm": 24.283540725708008,
      "learning_rate": 1.118888708395798e-05,
      "loss": 2.8705,
      "step": 969
    },
    {
      "epoch": 0.3,
      "grad_norm": 11.448086738586426,
      "learning_rate": 1.1183328974553058e-05,
      "loss": 1.2765,
      "step": 970
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.364688873291016,
      "learning_rate": 1.1177766947713552e-05,
      "loss": 1.5004,
      "step": 971
    },
    {
      "epoch": 0.3,
      "grad_norm": 18.686744689941406,
      "learning_rate": 1.1172201008710974e-05,
      "loss": 3.0106,
      "step": 972
    },
    {
      "epoch": 0.3,
      "grad_norm": 9.328819274902344,
      "learning_rate": 1.1166631162820538e-05,
      "loss": 1.6537,
      "step": 973
    },
    {
      "epoch": 0.3,
      "grad_norm": 15.435249328613281,
      "learning_rate": 1.1161057415321164e-05,
      "loss": 5.4632,
      "step": 974
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.374683380126953,
      "learning_rate": 1.1155479771495465e-05,
      "loss": 5.2145,
      "step": 975
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.833335876464844,
      "learning_rate": 1.1149898236629754e-05,
      "loss": 2.1236,
      "step": 976
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.178450584411621,
      "learning_rate": 1.1144312816014025e-05,
      "loss": 2.7498,
      "step": 977
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.957172393798828,
      "learning_rate": 1.1138723514941963e-05,
      "loss": 0.9344,
      "step": 978
    },
    {
      "epoch": 0.3,
      "grad_norm": 14.987807273864746,
      "learning_rate": 1.1133130338710917e-05,
      "loss": 2.0331,
      "step": 979
    },
    {
      "epoch": 0.3,
      "grad_norm": 17.33484649658203,
      "learning_rate": 1.1127533292621924e-05,
      "loss": 4.294,
      "step": 980
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.04723834991455,
      "learning_rate": 1.1121932381979679e-05,
      "loss": 2.3455,
      "step": 981
    },
    {
      "epoch": 0.3,
      "grad_norm": 13.156903266906738,
      "learning_rate": 1.1116327612092546e-05,
      "loss": 1.359,
      "step": 982
    },
    {
      "epoch": 0.3,
      "grad_norm": 15.709280967712402,
      "learning_rate": 1.111071898827254e-05,
      "loss": 2.018,
      "step": 983
    },
    {
      "epoch": 0.3,
      "grad_norm": 23.15648651123047,
      "learning_rate": 1.1105106515835333e-05,
      "loss": 3.0358,
      "step": 984
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.00515079498291,
      "learning_rate": 1.1099490200100241e-05,
      "loss": 1.1987,
      "step": 985
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.764809608459473,
      "learning_rate": 1.1093870046390231e-05,
      "loss": 1.7425,
      "step": 986
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.764495849609375,
      "learning_rate": 1.1088246060031897e-05,
      "loss": 2.7288,
      "step": 987
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.281590461730957,
      "learning_rate": 1.108261824635547e-05,
      "loss": 3.4756,
      "step": 988
    },
    {
      "epoch": 0.31,
      "grad_norm": 14.334712028503418,
      "learning_rate": 1.1076986610694813e-05,
      "loss": 2.1193,
      "step": 989
    },
    {
      "epoch": 0.31,
      "grad_norm": 18.38178253173828,
      "learning_rate": 1.1071351158387402e-05,
      "loss": 1.5241,
      "step": 990
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.551787376403809,
      "learning_rate": 1.1065711894774338e-05,
      "loss": 5.2846,
      "step": 991
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.72806167602539,
      "learning_rate": 1.106006882520033e-05,
      "loss": 2.9111,
      "step": 992
    },
    {
      "epoch": 0.31,
      "grad_norm": 16.107839584350586,
      "learning_rate": 1.1054421955013695e-05,
      "loss": 2.7331,
      "step": 993
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.12115478515625,
      "learning_rate": 1.1048771289566354e-05,
      "loss": 1.9713,
      "step": 994
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.031820297241211,
      "learning_rate": 1.1043116834213822e-05,
      "loss": 1.9452,
      "step": 995
    },
    {
      "epoch": 0.31,
      "grad_norm": 15.122635841369629,
      "learning_rate": 1.103745859431521e-05,
      "loss": 2.1204,
      "step": 996
    },
    {
      "epoch": 0.31,
      "grad_norm": 6.159744739532471,
      "learning_rate": 1.1031796575233211e-05,
      "loss": 0.8231,
      "step": 997
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.978343963623047,
      "learning_rate": 1.1026130782334106e-05,
      "loss": 1.0703,
      "step": 998
    },
    {
      "epoch": 0.31,
      "grad_norm": 11.258939743041992,
      "learning_rate": 1.1020461220987744e-05,
      "loss": 1.32,
      "step": 999
    },
    {
      "epoch": 0.31,
      "grad_norm": 15.83403205871582,
      "learning_rate": 1.1014787896567554e-05,
      "loss": 1.9714,
      "step": 1000
    },
    {
      "epoch": 0.31,
      "grad_norm": 14.88000774383545,
      "learning_rate": 1.1009110814450527e-05,
      "loss": 2.2317,
      "step": 1001
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.894899368286133,
      "learning_rate": 1.100342998001722e-05,
      "loss": 1.7284,
      "step": 1002
    },
    {
      "epoch": 0.31,
      "grad_norm": 17.018301010131836,
      "learning_rate": 1.0997745398651735e-05,
      "loss": 2.0597,
      "step": 1003
    },
    {
      "epoch": 0.31,
      "grad_norm": 18.89592933654785,
      "learning_rate": 1.0992057075741742e-05,
      "loss": 4.1914,
      "step": 1004
    },
    {
      "epoch": 0.31,
      "grad_norm": 17.937870025634766,
      "learning_rate": 1.098636501667844e-05,
      "loss": 9.2263,
      "step": 1005
    },
    {
      "epoch": 0.31,
      "grad_norm": 22.56218147277832,
      "learning_rate": 1.0980669226856584e-05,
      "loss": 2.8435,
      "step": 1006
    },
    {
      "epoch": 0.31,
      "grad_norm": 18.284509658813477,
      "learning_rate": 1.0974969711674452e-05,
      "loss": 4.8165,
      "step": 1007
    },
    {
      "epoch": 0.31,
      "grad_norm": 15.159753799438477,
      "learning_rate": 1.0969266476533864e-05,
      "loss": 2.9532,
      "step": 1008
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.554959297180176,
      "learning_rate": 1.0963559526840157e-05,
      "loss": 2.5036,
      "step": 1009
    },
    {
      "epoch": 0.31,
      "grad_norm": 13.92638111114502,
      "learning_rate": 1.0957848868002198e-05,
      "loss": 3.2055,
      "step": 1010
    },
    {
      "epoch": 0.31,
      "grad_norm": 15.291211128234863,
      "learning_rate": 1.0952134505432357e-05,
      "loss": 3.4983,
      "step": 1011
    },
    {
      "epoch": 0.31,
      "grad_norm": 10.960637092590332,
      "learning_rate": 1.0946416444546519e-05,
      "loss": 1.178,
      "step": 1012
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.908455848693848,
      "learning_rate": 1.0940694690764083e-05,
      "loss": 2.153,
      "step": 1013
    },
    {
      "epoch": 0.31,
      "grad_norm": 15.675530433654785,
      "learning_rate": 1.0934969249507937e-05,
      "loss": 4.4807,
      "step": 1014
    },
    {
      "epoch": 0.31,
      "grad_norm": 12.855313301086426,
      "learning_rate": 1.0929240126204468e-05,
      "loss": 3.9184,
      "step": 1015
    },
    {
      "epoch": 0.31,
      "grad_norm": 8.087355613708496,
      "learning_rate": 1.0923507326283553e-05,
      "loss": 1.181,
      "step": 1016
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.986209869384766,
      "learning_rate": 1.0917770855178553e-05,
      "loss": 1.4583,
      "step": 1017
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.826547622680664,
      "learning_rate": 1.0912030718326308e-05,
      "loss": 4.0032,
      "step": 1018
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.497063636779785,
      "learning_rate": 1.0906286921167134e-05,
      "loss": 1.0684,
      "step": 1019
    },
    {
      "epoch": 0.32,
      "grad_norm": 10.521944999694824,
      "learning_rate": 1.0900539469144812e-05,
      "loss": 1.7304,
      "step": 1020
    },
    {
      "epoch": 0.32,
      "grad_norm": 13.924857139587402,
      "learning_rate": 1.089478836770659e-05,
      "loss": 2.4864,
      "step": 1021
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.12009334564209,
      "learning_rate": 1.0889033622303182e-05,
      "loss": 1.3925,
      "step": 1022
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.136836051940918,
      "learning_rate": 1.0883275238388739e-05,
      "loss": 1.4287,
      "step": 1023
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.186498641967773,
      "learning_rate": 1.0877513221420873e-05,
      "loss": 1.7405,
      "step": 1024
    },
    {
      "epoch": 0.32,
      "grad_norm": 17.94057273864746,
      "learning_rate": 1.087174757686064e-05,
      "loss": 3.8773,
      "step": 1025
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.553104400634766,
      "learning_rate": 1.0865978310172522e-05,
      "loss": 2.3274,
      "step": 1026
    },
    {
      "epoch": 0.32,
      "grad_norm": 15.304353713989258,
      "learning_rate": 1.0860205426824449e-05,
      "loss": 5.7125,
      "step": 1027
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.027484893798828,
      "learning_rate": 1.0854428932287768e-05,
      "loss": 1.0223,
      "step": 1028
    },
    {
      "epoch": 0.32,
      "grad_norm": 14.752253532409668,
      "learning_rate": 1.0848648832037257e-05,
      "loss": 2.5004,
      "step": 1029
    },
    {
      "epoch": 0.32,
      "grad_norm": 19.452241897583008,
      "learning_rate": 1.0842865131551098e-05,
      "loss": 2.8527,
      "step": 1030
    },
    {
      "epoch": 0.32,
      "grad_norm": 17.181283950805664,
      "learning_rate": 1.0837077836310903e-05,
      "loss": 3.4927,
      "step": 1031
    },
    {
      "epoch": 0.32,
      "grad_norm": 9.379412651062012,
      "learning_rate": 1.083128695180168e-05,
      "loss": 0.9177,
      "step": 1032
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.588668823242188,
      "learning_rate": 1.0825492483511836e-05,
      "loss": 1.4227,
      "step": 1033
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.405271530151367,
      "learning_rate": 1.0819694436933183e-05,
      "loss": 1.6208,
      "step": 1034
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.79037094116211,
      "learning_rate": 1.081389281756092e-05,
      "loss": 2.357,
      "step": 1035
    },
    {
      "epoch": 0.32,
      "grad_norm": 13.481416702270508,
      "learning_rate": 1.0808087630893634e-05,
      "loss": 1.7349,
      "step": 1036
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.72651481628418,
      "learning_rate": 1.0802278882433291e-05,
      "loss": 1.3169,
      "step": 1037
    },
    {
      "epoch": 0.32,
      "grad_norm": 15.26197624206543,
      "learning_rate": 1.0796466577685234e-05,
      "loss": 2.0093,
      "step": 1038
    },
    {
      "epoch": 0.32,
      "grad_norm": 13.19687271118164,
      "learning_rate": 1.0790650722158175e-05,
      "loss": 2.2271,
      "step": 1039
    },
    {
      "epoch": 0.32,
      "grad_norm": 11.644251823425293,
      "learning_rate": 1.0784831321364192e-05,
      "loss": 1.1088,
      "step": 1040
    },
    {
      "epoch": 0.32,
      "grad_norm": 18.361989974975586,
      "learning_rate": 1.0779008380818725e-05,
      "loss": 6.0244,
      "step": 1041
    },
    {
      "epoch": 0.32,
      "grad_norm": 20.438385009765625,
      "learning_rate": 1.0773181906040568e-05,
      "loss": 2.9884,
      "step": 1042
    },
    {
      "epoch": 0.32,
      "grad_norm": 15.359336853027344,
      "learning_rate": 1.0767351902551863e-05,
      "loss": 1.2271,
      "step": 1043
    },
    {
      "epoch": 0.32,
      "grad_norm": 16.786720275878906,
      "learning_rate": 1.0761518375878097e-05,
      "loss": 5.1943,
      "step": 1044
    },
    {
      "epoch": 0.32,
      "grad_norm": 15.47581672668457,
      "learning_rate": 1.0755681331548095e-05,
      "loss": 2.0553,
      "step": 1045
    },
    {
      "epoch": 0.32,
      "grad_norm": 13.731276512145996,
      "learning_rate": 1.0749840775094023e-05,
      "loss": 2.6001,
      "step": 1046
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.26632022857666,
      "learning_rate": 1.0743996712051366e-05,
      "loss": 1.6236,
      "step": 1047
    },
    {
      "epoch": 0.32,
      "grad_norm": 12.464577674865723,
      "learning_rate": 1.0738149147958936e-05,
      "loss": 1.7641,
      "step": 1048
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.070165634155273,
      "learning_rate": 1.0732298088358867e-05,
      "loss": 2.8956,
      "step": 1049
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.91103172302246,
      "learning_rate": 1.07264435387966e-05,
      "loss": 3.1296,
      "step": 1050
    },
    {
      "epoch": 0.33,
      "grad_norm": 12.081746101379395,
      "learning_rate": 1.0720585504820885e-05,
      "loss": 2.3239,
      "step": 1051
    },
    {
      "epoch": 0.33,
      "grad_norm": 11.550629615783691,
      "learning_rate": 1.0714723991983782e-05,
      "loss": 2.0387,
      "step": 1052
    },
    {
      "epoch": 0.33,
      "grad_norm": 10.877471923828125,
      "learning_rate": 1.0708859005840637e-05,
      "loss": 0.9464,
      "step": 1053
    },
    {
      "epoch": 0.33,
      "grad_norm": 18.174089431762695,
      "learning_rate": 1.0702990551950092e-05,
      "loss": 2.5236,
      "step": 1054
    },
    {
      "epoch": 0.33,
      "grad_norm": 15.213211059570312,
      "learning_rate": 1.069711863587408e-05,
      "loss": 2.5513,
      "step": 1055
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.575151443481445,
      "learning_rate": 1.0691243263177808e-05,
      "loss": 4.2446,
      "step": 1056
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.19057273864746,
      "learning_rate": 1.0685364439429767e-05,
      "loss": 1.7243,
      "step": 1057
    },
    {
      "epoch": 0.33,
      "grad_norm": 12.374418258666992,
      "learning_rate": 1.0679482170201714e-05,
      "loss": 1.9149,
      "step": 1058
    },
    {
      "epoch": 0.33,
      "grad_norm": 19.93390464782715,
      "learning_rate": 1.0673596461068668e-05,
      "loss": 4.4496,
      "step": 1059
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.360801696777344,
      "learning_rate": 1.0667707317608917e-05,
      "loss": 1.2565,
      "step": 1060
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.6316556930542,
      "learning_rate": 1.0661814745403998e-05,
      "loss": 5.3259,
      "step": 1061
    },
    {
      "epoch": 0.33,
      "grad_norm": 16.264671325683594,
      "learning_rate": 1.0655918750038698e-05,
      "loss": 1.9203,
      "step": 1062
    },
    {
      "epoch": 0.33,
      "grad_norm": 25.273666381835938,
      "learning_rate": 1.065001933710105e-05,
      "loss": 7.867,
      "step": 1063
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.659454345703125,
      "learning_rate": 1.0644116512182325e-05,
      "loss": 1.921,
      "step": 1064
    },
    {
      "epoch": 0.33,
      "grad_norm": 19.446399688720703,
      "learning_rate": 1.063821028087703e-05,
      "loss": 6.1517,
      "step": 1065
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.461407661437988,
      "learning_rate": 1.0632300648782897e-05,
      "loss": 3.0714,
      "step": 1066
    },
    {
      "epoch": 0.33,
      "grad_norm": 19.374279022216797,
      "learning_rate": 1.0626387621500884e-05,
      "loss": 5.1723,
      "step": 1067
    },
    {
      "epoch": 0.33,
      "grad_norm": 19.21060562133789,
      "learning_rate": 1.0620471204635164e-05,
      "loss": 2.4489,
      "step": 1068
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.720632553100586,
      "learning_rate": 1.0614551403793124e-05,
      "loss": 2.9885,
      "step": 1069
    },
    {
      "epoch": 0.33,
      "grad_norm": 16.278717041015625,
      "learning_rate": 1.060862822458536e-05,
      "loss": 4.4297,
      "step": 1070
    },
    {
      "epoch": 0.33,
      "grad_norm": 16.912240982055664,
      "learning_rate": 1.060270167262567e-05,
      "loss": 2.9691,
      "step": 1071
    },
    {
      "epoch": 0.33,
      "grad_norm": 20.76288414001465,
      "learning_rate": 1.0596771753531042e-05,
      "loss": 6.1958,
      "step": 1072
    },
    {
      "epoch": 0.33,
      "grad_norm": 15.202351570129395,
      "learning_rate": 1.0590838472921663e-05,
      "loss": 2.0175,
      "step": 1073
    },
    {
      "epoch": 0.33,
      "grad_norm": 14.269200325012207,
      "learning_rate": 1.0584901836420905e-05,
      "loss": 2.3095,
      "step": 1074
    },
    {
      "epoch": 0.33,
      "grad_norm": 16.454748153686523,
      "learning_rate": 1.0578961849655319e-05,
      "loss": 5.6113,
      "step": 1075
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.027233123779297,
      "learning_rate": 1.0573018518254625e-05,
      "loss": 5.2381,
      "step": 1076
    },
    {
      "epoch": 0.33,
      "grad_norm": 13.088054656982422,
      "learning_rate": 1.0567071847851723e-05,
      "loss": 2.7067,
      "step": 1077
    },
    {
      "epoch": 0.33,
      "grad_norm": 12.304478645324707,
      "learning_rate": 1.0561121844082676e-05,
      "loss": 2.5341,
      "step": 1078
    },
    {
      "epoch": 0.33,
      "grad_norm": 17.764257431030273,
      "learning_rate": 1.0555168512586697e-05,
      "loss": 1.7237,
      "step": 1079
    },
    {
      "epoch": 0.33,
      "grad_norm": 23.56541633605957,
      "learning_rate": 1.0549211859006169e-05,
      "loss": 3.334,
      "step": 1080
    },
    {
      "epoch": 0.33,
      "grad_norm": 16.430952072143555,
      "learning_rate": 1.0543251888986605e-05,
      "loss": 2.1071,
      "step": 1081
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.087030410766602,
      "learning_rate": 1.0537288608176673e-05,
      "loss": 2.5736,
      "step": 1082
    },
    {
      "epoch": 0.34,
      "grad_norm": 17.93031120300293,
      "learning_rate": 1.0531322022228181e-05,
      "loss": 5.9253,
      "step": 1083
    },
    {
      "epoch": 0.34,
      "grad_norm": 18.3150577545166,
      "learning_rate": 1.0525352136796061e-05,
      "loss": 2.781,
      "step": 1084
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.447970390319824,
      "learning_rate": 1.0519378957538378e-05,
      "loss": 1.8882,
      "step": 1085
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.831107139587402,
      "learning_rate": 1.0513402490116313e-05,
      "loss": 1.3537,
      "step": 1086
    },
    {
      "epoch": 0.34,
      "grad_norm": 18.853090286254883,
      "learning_rate": 1.050742274019417e-05,
      "loss": 7.4965,
      "step": 1087
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.2085018157959,
      "learning_rate": 1.050143971343936e-05,
      "loss": 2.4386,
      "step": 1088
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.624937057495117,
      "learning_rate": 1.0495453415522405e-05,
      "loss": 2.6718,
      "step": 1089
    },
    {
      "epoch": 0.34,
      "grad_norm": 28.173011779785156,
      "learning_rate": 1.0489463852116918e-05,
      "loss": 2.2093,
      "step": 1090
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.845703125,
      "learning_rate": 1.0483471028899613e-05,
      "loss": 1.1327,
      "step": 1091
    },
    {
      "epoch": 0.34,
      "grad_norm": 13.236507415771484,
      "learning_rate": 1.0477474951550294e-05,
      "loss": 3.1908,
      "step": 1092
    },
    {
      "epoch": 0.34,
      "grad_norm": 7.714241981506348,
      "learning_rate": 1.0471475625751847e-05,
      "loss": 0.7375,
      "step": 1093
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.44623374938965,
      "learning_rate": 1.046547305719024e-05,
      "loss": 1.7071,
      "step": 1094
    },
    {
      "epoch": 0.34,
      "grad_norm": 6.655475616455078,
      "learning_rate": 1.0459467251554508e-05,
      "loss": 0.668,
      "step": 1095
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.086280822753906,
      "learning_rate": 1.0453458214536762e-05,
      "loss": 1.7233,
      "step": 1096
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.326202392578125,
      "learning_rate": 1.0447445951832167e-05,
      "loss": 1.9609,
      "step": 1097
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.662792205810547,
      "learning_rate": 1.0441430469138954e-05,
      "loss": 0.9869,
      "step": 1098
    },
    {
      "epoch": 0.34,
      "grad_norm": 12.256331443786621,
      "learning_rate": 1.0435411772158401e-05,
      "loss": 1.7208,
      "step": 1099
    },
    {
      "epoch": 0.34,
      "grad_norm": 23.35416030883789,
      "learning_rate": 1.042938986659483e-05,
      "loss": 2.9477,
      "step": 1100
    },
    {
      "epoch": 0.34,
      "grad_norm": 15.350637435913086,
      "learning_rate": 1.0423364758155609e-05,
      "loss": 1.8908,
      "step": 1101
    },
    {
      "epoch": 0.34,
      "grad_norm": 17.07219696044922,
      "learning_rate": 1.041733645255114e-05,
      "loss": 2.6845,
      "step": 1102
    },
    {
      "epoch": 0.34,
      "grad_norm": 19.381675720214844,
      "learning_rate": 1.0411304955494854e-05,
      "loss": 3.0835,
      "step": 1103
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.797784805297852,
      "learning_rate": 1.040527027270321e-05,
      "loss": 1.382,
      "step": 1104
    },
    {
      "epoch": 0.34,
      "grad_norm": 18.23086166381836,
      "learning_rate": 1.0399232409895677e-05,
      "loss": 1.6303,
      "step": 1105
    },
    {
      "epoch": 0.34,
      "grad_norm": 9.812505722045898,
      "learning_rate": 1.0393191372794753e-05,
      "loss": 0.8937,
      "step": 1106
    },
    {
      "epoch": 0.34,
      "grad_norm": 19.530162811279297,
      "learning_rate": 1.0387147167125929e-05,
      "loss": 3.8252,
      "step": 1107
    },
    {
      "epoch": 0.34,
      "grad_norm": 21.945730209350586,
      "learning_rate": 1.0381099798617711e-05,
      "loss": 4.0083,
      "step": 1108
    },
    {
      "epoch": 0.34,
      "grad_norm": 21.621074676513672,
      "learning_rate": 1.0375049273001596e-05,
      "loss": 5.3178,
      "step": 1109
    },
    {
      "epoch": 0.34,
      "grad_norm": 13.84105110168457,
      "learning_rate": 1.0368995596012079e-05,
      "loss": 1.3034,
      "step": 1110
    },
    {
      "epoch": 0.34,
      "grad_norm": 16.137035369873047,
      "learning_rate": 1.036293877338663e-05,
      "loss": 1.0091,
      "step": 1111
    },
    {
      "epoch": 0.34,
      "grad_norm": 14.98267650604248,
      "learning_rate": 1.0356878810865713e-05,
      "loss": 4.8692,
      "step": 1112
    },
    {
      "epoch": 0.34,
      "grad_norm": 10.179339408874512,
      "learning_rate": 1.0350815714192765e-05,
      "loss": 2.0231,
      "step": 1113
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.825215339660645,
      "learning_rate": 1.034474948911419e-05,
      "loss": 1.9089,
      "step": 1114
    },
    {
      "epoch": 0.35,
      "grad_norm": 15.7478609085083,
      "learning_rate": 1.0338680141379355e-05,
      "loss": 5.4506,
      "step": 1115
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.455516815185547,
      "learning_rate": 1.0332607676740596e-05,
      "loss": 4.8066,
      "step": 1116
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.75464916229248,
      "learning_rate": 1.0326532100953193e-05,
      "loss": 2.9567,
      "step": 1117
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.472881317138672,
      "learning_rate": 1.0320453419775378e-05,
      "loss": 1.5355,
      "step": 1118
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.322370529174805,
      "learning_rate": 1.0314371638968331e-05,
      "loss": 1.8469,
      "step": 1119
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.017380714416504,
      "learning_rate": 1.0308286764296164e-05,
      "loss": 2.8139,
      "step": 1120
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.160161972045898,
      "learning_rate": 1.0302198801525923e-05,
      "loss": 1.2088,
      "step": 1121
    },
    {
      "epoch": 0.35,
      "grad_norm": 20.627988815307617,
      "learning_rate": 1.029610775642758e-05,
      "loss": 4.5566,
      "step": 1122
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.355149269104004,
      "learning_rate": 1.0290013634774031e-05,
      "loss": 1.3943,
      "step": 1123
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.896719932556152,
      "learning_rate": 1.0283916442341085e-05,
      "loss": 2.469,
      "step": 1124
    },
    {
      "epoch": 0.35,
      "grad_norm": 25.20741081237793,
      "learning_rate": 1.0277816184907462e-05,
      "loss": 5.2271,
      "step": 1125
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.427926063537598,
      "learning_rate": 1.0271712868254787e-05,
      "loss": 2.7404,
      "step": 1126
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.849968910217285,
      "learning_rate": 1.0265606498167586e-05,
      "loss": 5.4348,
      "step": 1127
    },
    {
      "epoch": 0.35,
      "grad_norm": 15.994071960449219,
      "learning_rate": 1.0259497080433279e-05,
      "loss": 2.3489,
      "step": 1128
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.78647804260254,
      "learning_rate": 1.0253384620842169e-05,
      "loss": 5.6616,
      "step": 1129
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.617353439331055,
      "learning_rate": 1.0247269125187447e-05,
      "loss": 2.2434,
      "step": 1130
    },
    {
      "epoch": 0.35,
      "grad_norm": 11.257705688476562,
      "learning_rate": 1.0241150599265186e-05,
      "loss": 1.5982,
      "step": 1131
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.668039321899414,
      "learning_rate": 1.0235029048874318e-05,
      "loss": 1.4743,
      "step": 1132
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.072298049926758,
      "learning_rate": 1.0228904479816654e-05,
      "loss": 1.8054,
      "step": 1133
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.091022491455078,
      "learning_rate": 1.022277689789686e-05,
      "loss": 1.8684,
      "step": 1134
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.35072135925293,
      "learning_rate": 1.0216646308922457e-05,
      "loss": 2.476,
      "step": 1135
    },
    {
      "epoch": 0.35,
      "grad_norm": 8.345824241638184,
      "learning_rate": 1.021051271870382e-05,
      "loss": 1.0648,
      "step": 1136
    },
    {
      "epoch": 0.35,
      "grad_norm": 15.507397651672363,
      "learning_rate": 1.0204376133054166e-05,
      "loss": 6.1307,
      "step": 1137
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.02056312561035,
      "learning_rate": 1.0198236557789548e-05,
      "loss": 2.7045,
      "step": 1138
    },
    {
      "epoch": 0.35,
      "grad_norm": 14.535933494567871,
      "learning_rate": 1.0192093998728858e-05,
      "loss": 2.0451,
      "step": 1139
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.683833122253418,
      "learning_rate": 1.0185948461693812e-05,
      "loss": 1.9239,
      "step": 1140
    },
    {
      "epoch": 0.35,
      "grad_norm": 10.788311958312988,
      "learning_rate": 1.0179799952508952e-05,
      "loss": 1.2934,
      "step": 1141
    },
    {
      "epoch": 0.35,
      "grad_norm": 12.489497184753418,
      "learning_rate": 1.0173648477001632e-05,
      "loss": 1.8452,
      "step": 1142
    },
    {
      "epoch": 0.35,
      "grad_norm": 17.01746940612793,
      "learning_rate": 1.0167494041002023e-05,
      "loss": 2.1199,
      "step": 1143
    },
    {
      "epoch": 0.35,
      "grad_norm": 13.93575382232666,
      "learning_rate": 1.0161336650343097e-05,
      "loss": 2.2911,
      "step": 1144
    },
    {
      "epoch": 0.35,
      "grad_norm": 16.233535766601562,
      "learning_rate": 1.0155176310860629e-05,
      "loss": 4.7763,
      "step": 1145
    },
    {
      "epoch": 0.36,
      "grad_norm": 23.979536056518555,
      "learning_rate": 1.0149013028393188e-05,
      "loss": 3.326,
      "step": 1146
    },
    {
      "epoch": 0.36,
      "grad_norm": 17.001752853393555,
      "learning_rate": 1.0142846808782132e-05,
      "loss": 3.1564,
      "step": 1147
    },
    {
      "epoch": 0.36,
      "grad_norm": 23.573925018310547,
      "learning_rate": 1.0136677657871602e-05,
      "loss": 4.3396,
      "step": 1148
    },
    {
      "epoch": 0.36,
      "grad_norm": 13.067983627319336,
      "learning_rate": 1.0130505581508521e-05,
      "loss": 1.7841,
      "step": 1149
    },
    {
      "epoch": 0.36,
      "grad_norm": 19.41147804260254,
      "learning_rate": 1.0124330585542581e-05,
      "loss": 3.8713,
      "step": 1150
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.91067886352539,
      "learning_rate": 1.0118152675826241e-05,
      "loss": 4.796,
      "step": 1151
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.093422889709473,
      "learning_rate": 1.0111971858214724e-05,
      "loss": 1.9465,
      "step": 1152
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.869210243225098,
      "learning_rate": 1.010578813856601e-05,
      "loss": 4.7352,
      "step": 1153
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.492764472961426,
      "learning_rate": 1.0099601522740818e-05,
      "loss": 1.0493,
      "step": 1154
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.02536678314209,
      "learning_rate": 1.0093412016602632e-05,
      "loss": 1.4691,
      "step": 1155
    },
    {
      "epoch": 0.36,
      "grad_norm": 19.572359085083008,
      "learning_rate": 1.0087219626017658e-05,
      "loss": 2.0801,
      "step": 1156
    },
    {
      "epoch": 0.36,
      "grad_norm": 11.340232849121094,
      "learning_rate": 1.0081024356854843e-05,
      "loss": 1.668,
      "step": 1157
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.6257266998291,
      "learning_rate": 1.0074826214985863e-05,
      "loss": 1.0484,
      "step": 1158
    },
    {
      "epoch": 0.36,
      "grad_norm": 17.96036720275879,
      "learning_rate": 1.0068625206285112e-05,
      "loss": 2.1781,
      "step": 1159
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.40553092956543,
      "learning_rate": 1.0062421336629707e-05,
      "loss": 1.4709,
      "step": 1160
    },
    {
      "epoch": 0.36,
      "grad_norm": 13.915143013000488,
      "learning_rate": 1.0056214611899471e-05,
      "loss": 1.6584,
      "step": 1161
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.291485786437988,
      "learning_rate": 1.0050005037976941e-05,
      "loss": 4.2902,
      "step": 1162
    },
    {
      "epoch": 0.36,
      "grad_norm": 15.307587623596191,
      "learning_rate": 1.0043792620747342e-05,
      "loss": 5.2116,
      "step": 1163
    },
    {
      "epoch": 0.36,
      "grad_norm": 19.080533981323242,
      "learning_rate": 1.0037577366098603e-05,
      "loss": 2.7628,
      "step": 1164
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.373787879943848,
      "learning_rate": 1.0031359279921343e-05,
      "loss": 2.3492,
      "step": 1165
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.30225944519043,
      "learning_rate": 1.0025138368108858e-05,
      "loss": 5.119,
      "step": 1166
    },
    {
      "epoch": 0.36,
      "grad_norm": 17.795886993408203,
      "learning_rate": 1.0018914636557127e-05,
      "loss": 1.8234,
      "step": 1167
    },
    {
      "epoch": 0.36,
      "grad_norm": 13.938688278198242,
      "learning_rate": 1.00126880911648e-05,
      "loss": 3.5577,
      "step": 1168
    },
    {
      "epoch": 0.36,
      "grad_norm": 12.905827522277832,
      "learning_rate": 1.0006458737833194e-05,
      "loss": 0.7515,
      "step": 1169
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.679336547851562,
      "learning_rate": 1.0000226582466286e-05,
      "loss": 4.5681,
      "step": 1170
    },
    {
      "epoch": 0.36,
      "grad_norm": 15.47931957244873,
      "learning_rate": 9.99399163097071e-06,
      "loss": 1.555,
      "step": 1171
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.255754470825195,
      "learning_rate": 9.987753889255753e-06,
      "loss": 1.9615,
      "step": 1172
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.039870262145996,
      "learning_rate": 9.981513363233342e-06,
      "loss": 2.3602,
      "step": 1173
    },
    {
      "epoch": 0.36,
      "grad_norm": 15.20569133758545,
      "learning_rate": 9.975270058818041e-06,
      "loss": 4.2076,
      "step": 1174
    },
    {
      "epoch": 0.36,
      "grad_norm": 9.11036205291748,
      "learning_rate": 9.969023981927054e-06,
      "loss": 1.4167,
      "step": 1175
    },
    {
      "epoch": 0.36,
      "grad_norm": 16.150285720825195,
      "learning_rate": 9.962775138480211e-06,
      "loss": 3.0512,
      "step": 1176
    },
    {
      "epoch": 0.36,
      "grad_norm": 17.30279541015625,
      "learning_rate": 9.956523534399956e-06,
      "loss": 3.5126,
      "step": 1177
    },
    {
      "epoch": 0.36,
      "grad_norm": 14.633407592773438,
      "learning_rate": 9.95026917561136e-06,
      "loss": 5.3493,
      "step": 1178
    },
    {
      "epoch": 0.37,
      "grad_norm": 21.873212814331055,
      "learning_rate": 9.944012068042102e-06,
      "loss": 5.6664,
      "step": 1179
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.300834655761719,
      "learning_rate": 9.937752217622463e-06,
      "loss": 2.9006,
      "step": 1180
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.483316421508789,
      "learning_rate": 9.931489630285323e-06,
      "loss": 2.9011,
      "step": 1181
    },
    {
      "epoch": 0.37,
      "grad_norm": 16.026229858398438,
      "learning_rate": 9.92522431196616e-06,
      "loss": 1.76,
      "step": 1182
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.829645156860352,
      "learning_rate": 9.91895626860304e-06,
      "loss": 1.9771,
      "step": 1183
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.540634155273438,
      "learning_rate": 9.912685506136613e-06,
      "loss": 1.0964,
      "step": 1184
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.960216522216797,
      "learning_rate": 9.906412030510093e-06,
      "loss": 1.8555,
      "step": 1185
    },
    {
      "epoch": 0.37,
      "grad_norm": 17.319026947021484,
      "learning_rate": 9.900135847669287e-06,
      "loss": 2.5102,
      "step": 1186
    },
    {
      "epoch": 0.37,
      "grad_norm": 19.70650291442871,
      "learning_rate": 9.893856963562553e-06,
      "loss": 6.4272,
      "step": 1187
    },
    {
      "epoch": 0.37,
      "grad_norm": 10.795018196105957,
      "learning_rate": 9.88757538414081e-06,
      "loss": 1.5096,
      "step": 1188
    },
    {
      "epoch": 0.37,
      "grad_norm": 14.696696281433105,
      "learning_rate": 9.88129111535754e-06,
      "loss": 5.2524,
      "step": 1189
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.470584869384766,
      "learning_rate": 9.875004163168761e-06,
      "loss": 1.545,
      "step": 1190
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.876420021057129,
      "learning_rate": 9.868714533533047e-06,
      "loss": 1.419,
      "step": 1191
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.037657737731934,
      "learning_rate": 9.862422232411503e-06,
      "loss": 1.228,
      "step": 1192
    },
    {
      "epoch": 0.37,
      "grad_norm": 16.153223037719727,
      "learning_rate": 9.856127265767766e-06,
      "loss": 2.9238,
      "step": 1193
    },
    {
      "epoch": 0.37,
      "grad_norm": 8.563721656799316,
      "learning_rate": 9.849829639568e-06,
      "loss": 0.9769,
      "step": 1194
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.931380271911621,
      "learning_rate": 9.843529359780893e-06,
      "loss": 1.8056,
      "step": 1195
    },
    {
      "epoch": 0.37,
      "grad_norm": 14.038524627685547,
      "learning_rate": 9.83722643237764e-06,
      "loss": 2.5044,
      "step": 1196
    },
    {
      "epoch": 0.37,
      "grad_norm": 21.455846786499023,
      "learning_rate": 9.830920863331956e-06,
      "loss": 6.5308,
      "step": 1197
    },
    {
      "epoch": 0.37,
      "grad_norm": 9.880926132202148,
      "learning_rate": 9.824612658620052e-06,
      "loss": 1.3094,
      "step": 1198
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.45067310333252,
      "learning_rate": 9.818301824220637e-06,
      "loss": 1.7067,
      "step": 1199
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.626029968261719,
      "learning_rate": 9.811988366114913e-06,
      "loss": 2.5457,
      "step": 1200
    },
    {
      "epoch": 0.37,
      "grad_norm": 22.094850540161133,
      "learning_rate": 9.805672290286577e-06,
      "loss": 2.4551,
      "step": 1201
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.181697845458984,
      "learning_rate": 9.799353602721792e-06,
      "loss": 1.7186,
      "step": 1202
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.625029563903809,
      "learning_rate": 9.79303230940921e-06,
      "loss": 4.4049,
      "step": 1203
    },
    {
      "epoch": 0.37,
      "grad_norm": 14.522659301757812,
      "learning_rate": 9.786708416339944e-06,
      "loss": 1.8953,
      "step": 1204
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.440786361694336,
      "learning_rate": 9.780381929507575e-06,
      "loss": 2.1252,
      "step": 1205
    },
    {
      "epoch": 0.37,
      "grad_norm": 15.497150421142578,
      "learning_rate": 9.774052854908138e-06,
      "loss": 3.1987,
      "step": 1206
    },
    {
      "epoch": 0.37,
      "grad_norm": 17.17487907409668,
      "learning_rate": 9.767721198540129e-06,
      "loss": 1.7985,
      "step": 1207
    },
    {
      "epoch": 0.37,
      "grad_norm": 11.13756275177002,
      "learning_rate": 9.761386966404483e-06,
      "loss": 1.5107,
      "step": 1208
    },
    {
      "epoch": 0.37,
      "grad_norm": 13.055180549621582,
      "learning_rate": 9.755050164504577e-06,
      "loss": 3.8065,
      "step": 1209
    },
    {
      "epoch": 0.37,
      "grad_norm": 12.56474494934082,
      "learning_rate": 9.748710798846232e-06,
      "loss": 2.9978,
      "step": 1210
    },
    {
      "epoch": 0.38,
      "grad_norm": 14.121736526489258,
      "learning_rate": 9.742368875437683e-06,
      "loss": 2.078,
      "step": 1211
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.13382339477539,
      "learning_rate": 9.736024400289606e-06,
      "loss": 1.387,
      "step": 1212
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.474137306213379,
      "learning_rate": 9.729677379415085e-06,
      "loss": 1.4034,
      "step": 1213
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.429963111877441,
      "learning_rate": 9.723327818829618e-06,
      "loss": 1.1999,
      "step": 1214
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.799750328063965,
      "learning_rate": 9.716975724551116e-06,
      "loss": 1.0906,
      "step": 1215
    },
    {
      "epoch": 0.38,
      "grad_norm": 12.66462230682373,
      "learning_rate": 9.710621102599883e-06,
      "loss": 2.6997,
      "step": 1216
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.209826469421387,
      "learning_rate": 9.704263958998624e-06,
      "loss": 2.7838,
      "step": 1217
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.077875137329102,
      "learning_rate": 9.697904299772434e-06,
      "loss": 1.6925,
      "step": 1218
    },
    {
      "epoch": 0.38,
      "grad_norm": 9.607660293579102,
      "learning_rate": 9.691542130948792e-06,
      "loss": 1.4831,
      "step": 1219
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.674386978149414,
      "learning_rate": 9.685177458557548e-06,
      "loss": 2.0321,
      "step": 1220
    },
    {
      "epoch": 0.38,
      "grad_norm": 22.753320693969727,
      "learning_rate": 9.678810288630936e-06,
      "loss": 3.6639,
      "step": 1221
    },
    {
      "epoch": 0.38,
      "grad_norm": 16.516311645507812,
      "learning_rate": 9.672440627203556e-06,
      "loss": 7.3332,
      "step": 1222
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.29770278930664,
      "learning_rate": 9.666068480312358e-06,
      "loss": 0.9376,
      "step": 1223
    },
    {
      "epoch": 0.38,
      "grad_norm": 18.074369430541992,
      "learning_rate": 9.659693853996658e-06,
      "loss": 2.0509,
      "step": 1224
    },
    {
      "epoch": 0.38,
      "grad_norm": 14.756071090698242,
      "learning_rate": 9.653316754298122e-06,
      "loss": 1.5116,
      "step": 1225
    },
    {
      "epoch": 0.38,
      "grad_norm": 18.59358024597168,
      "learning_rate": 9.646937187260755e-06,
      "loss": 4.5909,
      "step": 1226
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.734687805175781,
      "learning_rate": 9.640555158930902e-06,
      "loss": 2.512,
      "step": 1227
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.989273071289062,
      "learning_rate": 9.634170675357244e-06,
      "loss": 1.2963,
      "step": 1228
    },
    {
      "epoch": 0.38,
      "grad_norm": 16.843233108520508,
      "learning_rate": 9.627783742590786e-06,
      "loss": 3.031,
      "step": 1229
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.894126892089844,
      "learning_rate": 9.621394366684852e-06,
      "loss": 2.2722,
      "step": 1230
    },
    {
      "epoch": 0.38,
      "grad_norm": 22.67670440673828,
      "learning_rate": 9.61500255369509e-06,
      "loss": 2.397,
      "step": 1231
    },
    {
      "epoch": 0.38,
      "grad_norm": 23.82426643371582,
      "learning_rate": 9.608608309679443e-06,
      "loss": 2.2971,
      "step": 1232
    },
    {
      "epoch": 0.38,
      "grad_norm": 13.017992973327637,
      "learning_rate": 9.602211640698177e-06,
      "loss": 1.5027,
      "step": 1233
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.917731285095215,
      "learning_rate": 9.595812552813842e-06,
      "loss": 4.8058,
      "step": 1234
    },
    {
      "epoch": 0.38,
      "grad_norm": 13.124893188476562,
      "learning_rate": 9.589411052091286e-06,
      "loss": 2.5112,
      "step": 1235
    },
    {
      "epoch": 0.38,
      "grad_norm": 11.611188888549805,
      "learning_rate": 9.583007144597642e-06,
      "loss": 1.5805,
      "step": 1236
    },
    {
      "epoch": 0.38,
      "grad_norm": 10.515594482421875,
      "learning_rate": 9.576600836402324e-06,
      "loss": 1.3343,
      "step": 1237
    },
    {
      "epoch": 0.38,
      "grad_norm": 18.802034378051758,
      "learning_rate": 9.570192133577026e-06,
      "loss": 2.038,
      "step": 1238
    },
    {
      "epoch": 0.38,
      "grad_norm": 16.2142391204834,
      "learning_rate": 9.563781042195703e-06,
      "loss": 3.2628,
      "step": 1239
    },
    {
      "epoch": 0.38,
      "grad_norm": 15.4701566696167,
      "learning_rate": 9.557367568334583e-06,
      "loss": 2.5495,
      "step": 1240
    },
    {
      "epoch": 0.38,
      "grad_norm": 14.809575080871582,
      "learning_rate": 9.550951718072145e-06,
      "loss": 2.47,
      "step": 1241
    },
    {
      "epoch": 0.38,
      "grad_norm": 17.822465896606445,
      "learning_rate": 9.544533497489123e-06,
      "loss": 1.7422,
      "step": 1242
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.080434799194336,
      "learning_rate": 9.538112912668501e-06,
      "loss": 1.3608,
      "step": 1243
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.433822631835938,
      "learning_rate": 9.531689969695495e-06,
      "loss": 2.3869,
      "step": 1244
    },
    {
      "epoch": 0.39,
      "grad_norm": 19.75357437133789,
      "learning_rate": 9.525264674657563e-06,
      "loss": 5.278,
      "step": 1245
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.760089874267578,
      "learning_rate": 9.518837033644393e-06,
      "loss": 1.2702,
      "step": 1246
    },
    {
      "epoch": 0.39,
      "grad_norm": 15.111987113952637,
      "learning_rate": 9.512407052747889e-06,
      "loss": 1.9018,
      "step": 1247
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.314002990722656,
      "learning_rate": 9.50597473806218e-06,
      "loss": 3.2879,
      "step": 1248
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.623244285583496,
      "learning_rate": 9.499540095683605e-06,
      "loss": 1.0273,
      "step": 1249
    },
    {
      "epoch": 0.39,
      "grad_norm": 8.672993659973145,
      "learning_rate": 9.49310313171071e-06,
      "loss": 1.1626,
      "step": 1250
    },
    {
      "epoch": 0.39,
      "grad_norm": 13.466089248657227,
      "learning_rate": 9.486663852244235e-06,
      "loss": 1.5149,
      "step": 1251
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.448925018310547,
      "learning_rate": 9.480222263387122e-06,
      "loss": 4.1081,
      "step": 1252
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.953042984008789,
      "learning_rate": 9.4737783712445e-06,
      "loss": 1.8349,
      "step": 1253
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.215497970581055,
      "learning_rate": 9.467332181923682e-06,
      "loss": 1.0758,
      "step": 1254
    },
    {
      "epoch": 0.39,
      "grad_norm": 15.898959159851074,
      "learning_rate": 9.46088370153415e-06,
      "loss": 2.8912,
      "step": 1255
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.336346626281738,
      "learning_rate": 9.45443293618757e-06,
      "loss": 3.0235,
      "step": 1256
    },
    {
      "epoch": 0.39,
      "grad_norm": 12.652429580688477,
      "learning_rate": 9.447979891997762e-06,
      "loss": 1.4084,
      "step": 1257
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.48835563659668,
      "learning_rate": 9.441524575080714e-06,
      "loss": 1.9351,
      "step": 1258
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.853721618652344,
      "learning_rate": 9.435066991554565e-06,
      "loss": 2.0826,
      "step": 1259
    },
    {
      "epoch": 0.39,
      "grad_norm": 11.199191093444824,
      "learning_rate": 9.428607147539601e-06,
      "loss": 1.7517,
      "step": 1260
    },
    {
      "epoch": 0.39,
      "grad_norm": 10.818768501281738,
      "learning_rate": 9.422145049158256e-06,
      "loss": 1.6581,
      "step": 1261
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.718265533447266,
      "learning_rate": 9.415680702535092e-06,
      "loss": 2.4333,
      "step": 1262
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.43951416015625,
      "learning_rate": 9.409214113796806e-06,
      "loss": 1.7366,
      "step": 1263
    },
    {
      "epoch": 0.39,
      "grad_norm": 13.1856050491333,
      "learning_rate": 9.402745289072221e-06,
      "loss": 5.1645,
      "step": 1264
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.550506591796875,
      "learning_rate": 9.396274234492283e-06,
      "loss": 8.1488,
      "step": 1265
    },
    {
      "epoch": 0.39,
      "grad_norm": 15.178678512573242,
      "learning_rate": 9.389800956190039e-06,
      "loss": 1.7309,
      "step": 1266
    },
    {
      "epoch": 0.39,
      "grad_norm": 16.878833770751953,
      "learning_rate": 9.383325460300657e-06,
      "loss": 1.7562,
      "step": 1267
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.218253135681152,
      "learning_rate": 9.376847752961399e-06,
      "loss": 3.9006,
      "step": 1268
    },
    {
      "epoch": 0.39,
      "grad_norm": 17.071884155273438,
      "learning_rate": 9.370367840311625e-06,
      "loss": 5.4157,
      "step": 1269
    },
    {
      "epoch": 0.39,
      "grad_norm": 18.59787940979004,
      "learning_rate": 9.363885728492784e-06,
      "loss": 3.4909,
      "step": 1270
    },
    {
      "epoch": 0.39,
      "grad_norm": 13.29552936553955,
      "learning_rate": 9.357401423648413e-06,
      "loss": 1.5064,
      "step": 1271
    },
    {
      "epoch": 0.39,
      "grad_norm": 15.856002807617188,
      "learning_rate": 9.350914931924126e-06,
      "loss": 4.8737,
      "step": 1272
    },
    {
      "epoch": 0.39,
      "grad_norm": 9.875200271606445,
      "learning_rate": 9.344426259467604e-06,
      "loss": 1.3462,
      "step": 1273
    },
    {
      "epoch": 0.39,
      "grad_norm": 14.881725311279297,
      "learning_rate": 9.337935412428602e-06,
      "loss": 2.0532,
      "step": 1274
    },
    {
      "epoch": 0.4,
      "grad_norm": 16.499113082885742,
      "learning_rate": 9.331442396958935e-06,
      "loss": 1.6289,
      "step": 1275
    },
    {
      "epoch": 0.4,
      "grad_norm": 22.04961395263672,
      "learning_rate": 9.324947219212467e-06,
      "loss": 4.3416,
      "step": 1276
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.246705055236816,
      "learning_rate": 9.318449885345123e-06,
      "loss": 0.8192,
      "step": 1277
    },
    {
      "epoch": 0.4,
      "grad_norm": 25.260831832885742,
      "learning_rate": 9.311950401514857e-06,
      "loss": 3.559,
      "step": 1278
    },
    {
      "epoch": 0.4,
      "grad_norm": 16.55936622619629,
      "learning_rate": 9.305448773881671e-06,
      "loss": 1.8204,
      "step": 1279
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.88330364227295,
      "learning_rate": 9.298945008607598e-06,
      "loss": 2.2244,
      "step": 1280
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.431914329528809,
      "learning_rate": 9.292439111856692e-06,
      "loss": 2.5618,
      "step": 1281
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.331016540527344,
      "learning_rate": 9.285931089795032e-06,
      "loss": 1.8892,
      "step": 1282
    },
    {
      "epoch": 0.4,
      "grad_norm": 7.850904941558838,
      "learning_rate": 9.279420948590708e-06,
      "loss": 0.8145,
      "step": 1283
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.035999298095703,
      "learning_rate": 9.272908694413818e-06,
      "loss": 5.0167,
      "step": 1284
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.117691040039062,
      "learning_rate": 9.26639433343647e-06,
      "loss": 1.2939,
      "step": 1285
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.571340560913086,
      "learning_rate": 9.259877871832759e-06,
      "loss": 1.8576,
      "step": 1286
    },
    {
      "epoch": 0.4,
      "grad_norm": 15.457513809204102,
      "learning_rate": 9.253359315778775e-06,
      "loss": 3.5042,
      "step": 1287
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.027514457702637,
      "learning_rate": 9.246838671452594e-06,
      "loss": 1.1628,
      "step": 1288
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.462739944458008,
      "learning_rate": 9.240315945034274e-06,
      "loss": 1.2288,
      "step": 1289
    },
    {
      "epoch": 0.4,
      "grad_norm": 16.443859100341797,
      "learning_rate": 9.233791142705834e-06,
      "loss": 3.6546,
      "step": 1290
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.439101219177246,
      "learning_rate": 9.227264270651276e-06,
      "loss": 2.3072,
      "step": 1291
    },
    {
      "epoch": 0.4,
      "grad_norm": 20.032997131347656,
      "learning_rate": 9.220735335056554e-06,
      "loss": 1.4739,
      "step": 1292
    },
    {
      "epoch": 0.4,
      "grad_norm": 17.97390365600586,
      "learning_rate": 9.214204342109585e-06,
      "loss": 2.3818,
      "step": 1293
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.600703239440918,
      "learning_rate": 9.207671298000225e-06,
      "loss": 2.8134,
      "step": 1294
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.583927154541016,
      "learning_rate": 9.201136208920286e-06,
      "loss": 2.2068,
      "step": 1295
    },
    {
      "epoch": 0.4,
      "grad_norm": 13.71022891998291,
      "learning_rate": 9.194599081063507e-06,
      "loss": 1.2469,
      "step": 1296
    },
    {
      "epoch": 0.4,
      "grad_norm": 17.376754760742188,
      "learning_rate": 9.188059920625567e-06,
      "loss": 2.8039,
      "step": 1297
    },
    {
      "epoch": 0.4,
      "grad_norm": 24.509260177612305,
      "learning_rate": 9.181518733804072e-06,
      "loss": 6.0718,
      "step": 1298
    },
    {
      "epoch": 0.4,
      "grad_norm": 10.079828262329102,
      "learning_rate": 9.174975526798544e-06,
      "loss": 1.4315,
      "step": 1299
    },
    {
      "epoch": 0.4,
      "grad_norm": 18.6185302734375,
      "learning_rate": 9.16843030581042e-06,
      "loss": 2.0979,
      "step": 1300
    },
    {
      "epoch": 0.4,
      "grad_norm": 22.407678604125977,
      "learning_rate": 9.161883077043049e-06,
      "loss": 3.9683,
      "step": 1301
    },
    {
      "epoch": 0.4,
      "grad_norm": 11.654013633728027,
      "learning_rate": 9.155333846701682e-06,
      "loss": 2.2643,
      "step": 1302
    },
    {
      "epoch": 0.4,
      "grad_norm": 9.130629539489746,
      "learning_rate": 9.148782620993464e-06,
      "loss": 1.0969,
      "step": 1303
    },
    {
      "epoch": 0.4,
      "grad_norm": 15.498295783996582,
      "learning_rate": 9.142229406127434e-06,
      "loss": 3.0316,
      "step": 1304
    },
    {
      "epoch": 0.4,
      "grad_norm": 12.560232162475586,
      "learning_rate": 9.135674208314516e-06,
      "loss": 3.3885,
      "step": 1305
    },
    {
      "epoch": 0.4,
      "grad_norm": 15.263082504272461,
      "learning_rate": 9.129117033767514e-06,
      "loss": 1.6005,
      "step": 1306
    },
    {
      "epoch": 0.4,
      "grad_norm": 14.036645889282227,
      "learning_rate": 9.122557888701104e-06,
      "loss": 3.1104,
      "step": 1307
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.84864616394043,
      "learning_rate": 9.115996779331826e-06,
      "loss": 1.8887,
      "step": 1308
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.828146934509277,
      "learning_rate": 9.109433711878092e-06,
      "loss": 2.8117,
      "step": 1309
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.600400924682617,
      "learning_rate": 9.102868692560162e-06,
      "loss": 4.8946,
      "step": 1310
    },
    {
      "epoch": 0.41,
      "grad_norm": 15.0155611038208,
      "learning_rate": 9.096301727600144e-06,
      "loss": 3.2513,
      "step": 1311
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.661171913146973,
      "learning_rate": 9.089732823221997e-06,
      "loss": 1.1866,
      "step": 1312
    },
    {
      "epoch": 0.41,
      "grad_norm": 22.12264060974121,
      "learning_rate": 9.083161985651512e-06,
      "loss": 1.7744,
      "step": 1313
    },
    {
      "epoch": 0.41,
      "grad_norm": 14.373431205749512,
      "learning_rate": 9.076589221116318e-06,
      "loss": 1.8977,
      "step": 1314
    },
    {
      "epoch": 0.41,
      "grad_norm": 17.13340187072754,
      "learning_rate": 9.070014535845864e-06,
      "loss": 4.0427,
      "step": 1315
    },
    {
      "epoch": 0.41,
      "grad_norm": 19.54230499267578,
      "learning_rate": 9.063437936071424e-06,
      "loss": 3.5795,
      "step": 1316
    },
    {
      "epoch": 0.41,
      "grad_norm": 23.94930076599121,
      "learning_rate": 9.056859428026088e-06,
      "loss": 3.257,
      "step": 1317
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.285595893859863,
      "learning_rate": 9.050279017944747e-06,
      "loss": 1.9428,
      "step": 1318
    },
    {
      "epoch": 0.41,
      "grad_norm": 10.203804969787598,
      "learning_rate": 9.043696712064099e-06,
      "loss": 1.1842,
      "step": 1319
    },
    {
      "epoch": 0.41,
      "grad_norm": 19.172683715820312,
      "learning_rate": 9.037112516622643e-06,
      "loss": 4.4253,
      "step": 1320
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.565506935119629,
      "learning_rate": 9.030526437860664e-06,
      "loss": 2.852,
      "step": 1321
    },
    {
      "epoch": 0.41,
      "grad_norm": 8.095751762390137,
      "learning_rate": 9.02393848202023e-06,
      "loss": 1.434,
      "step": 1322
    },
    {
      "epoch": 0.41,
      "grad_norm": 16.259660720825195,
      "learning_rate": 9.017348655345196e-06,
      "loss": 1.5106,
      "step": 1323
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.453413963317871,
      "learning_rate": 9.01075696408118e-06,
      "loss": 1.8904,
      "step": 1324
    },
    {
      "epoch": 0.41,
      "grad_norm": 17.38831901550293,
      "learning_rate": 9.004163414475576e-06,
      "loss": 6.3193,
      "step": 1325
    },
    {
      "epoch": 0.41,
      "grad_norm": 14.774956703186035,
      "learning_rate": 8.997568012777534e-06,
      "loss": 1.5099,
      "step": 1326
    },
    {
      "epoch": 0.41,
      "grad_norm": 17.295808792114258,
      "learning_rate": 8.990970765237961e-06,
      "loss": 5.6794,
      "step": 1327
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.591071128845215,
      "learning_rate": 8.984371678109513e-06,
      "loss": 1.3968,
      "step": 1328
    },
    {
      "epoch": 0.41,
      "grad_norm": 14.689584732055664,
      "learning_rate": 8.97777075764659e-06,
      "loss": 2.0836,
      "step": 1329
    },
    {
      "epoch": 0.41,
      "grad_norm": 18.209320068359375,
      "learning_rate": 8.971168010105329e-06,
      "loss": 1.9112,
      "step": 1330
    },
    {
      "epoch": 0.41,
      "grad_norm": 18.5660343170166,
      "learning_rate": 8.9645634417436e-06,
      "loss": 2.5502,
      "step": 1331
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.953259468078613,
      "learning_rate": 8.957957058820999e-06,
      "loss": 0.8663,
      "step": 1332
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.316336631774902,
      "learning_rate": 8.951348867598838e-06,
      "loss": 2.494,
      "step": 1333
    },
    {
      "epoch": 0.41,
      "grad_norm": 9.528247833251953,
      "learning_rate": 8.944738874340146e-06,
      "loss": 1.0284,
      "step": 1334
    },
    {
      "epoch": 0.41,
      "grad_norm": 12.781307220458984,
      "learning_rate": 8.93812708530966e-06,
      "loss": 2.5631,
      "step": 1335
    },
    {
      "epoch": 0.41,
      "grad_norm": 21.69401741027832,
      "learning_rate": 8.931513506773818e-06,
      "loss": 3.1673,
      "step": 1336
    },
    {
      "epoch": 0.41,
      "grad_norm": 11.969874382019043,
      "learning_rate": 8.924898145000754e-06,
      "loss": 2.8022,
      "step": 1337
    },
    {
      "epoch": 0.41,
      "grad_norm": 14.627640724182129,
      "learning_rate": 8.918281006260297e-06,
      "loss": 2.124,
      "step": 1338
    },
    {
      "epoch": 0.41,
      "grad_norm": 13.691817283630371,
      "learning_rate": 8.911662096823946e-06,
      "loss": 1.3643,
      "step": 1339
    },
    {
      "epoch": 0.42,
      "grad_norm": 16.14912223815918,
      "learning_rate": 8.905041422964897e-06,
      "loss": 2.7159,
      "step": 1340
    },
    {
      "epoch": 0.42,
      "grad_norm": 21.14331817626953,
      "learning_rate": 8.898418990958003e-06,
      "loss": 2.4386,
      "step": 1341
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.493985176086426,
      "learning_rate": 8.891794807079791e-06,
      "loss": 1.0417,
      "step": 1342
    },
    {
      "epoch": 0.42,
      "grad_norm": 7.941287994384766,
      "learning_rate": 8.88516887760845e-06,
      "loss": 0.9287,
      "step": 1343
    },
    {
      "epoch": 0.42,
      "grad_norm": 15.730162620544434,
      "learning_rate": 8.878541208823817e-06,
      "loss": 1.849,
      "step": 1344
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.380057334899902,
      "learning_rate": 8.871911807007382e-06,
      "loss": 2.0177,
      "step": 1345
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.666074752807617,
      "learning_rate": 8.86528067844227e-06,
      "loss": 1.9661,
      "step": 1346
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.915221214294434,
      "learning_rate": 8.858647829413261e-06,
      "loss": 1.5253,
      "step": 1347
    },
    {
      "epoch": 0.42,
      "grad_norm": 19.663179397583008,
      "learning_rate": 8.852013266206741e-06,
      "loss": 2.7269,
      "step": 1348
    },
    {
      "epoch": 0.42,
      "grad_norm": 14.238304138183594,
      "learning_rate": 8.845376995110742e-06,
      "loss": 1.6626,
      "step": 1349
    },
    {
      "epoch": 0.42,
      "grad_norm": 16.315189361572266,
      "learning_rate": 8.8387390224149e-06,
      "loss": 4.0402,
      "step": 1350
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.602688789367676,
      "learning_rate": 8.832099354410471e-06,
      "loss": 0.7924,
      "step": 1351
    },
    {
      "epoch": 0.42,
      "grad_norm": 28.27524185180664,
      "learning_rate": 8.825457997390316e-06,
      "loss": 7.6666,
      "step": 1352
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.535933494567871,
      "learning_rate": 8.818814957648897e-06,
      "loss": 1.5314,
      "step": 1353
    },
    {
      "epoch": 0.42,
      "grad_norm": 12.956826210021973,
      "learning_rate": 8.812170241482271e-06,
      "loss": 1.9593,
      "step": 1354
    },
    {
      "epoch": 0.42,
      "grad_norm": 15.91325855255127,
      "learning_rate": 8.805523855188082e-06,
      "loss": 6.0726,
      "step": 1355
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.713898658752441,
      "learning_rate": 8.798875805065565e-06,
      "loss": 1.4066,
      "step": 1356
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.808228492736816,
      "learning_rate": 8.792226097415512e-06,
      "loss": 1.284,
      "step": 1357
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.787436485290527,
      "learning_rate": 8.785574738540314e-06,
      "loss": 1.9648,
      "step": 1358
    },
    {
      "epoch": 0.42,
      "grad_norm": 10.245705604553223,
      "learning_rate": 8.778921734743904e-06,
      "loss": 1.0906,
      "step": 1359
    },
    {
      "epoch": 0.42,
      "grad_norm": 9.439340591430664,
      "learning_rate": 8.772267092331784e-06,
      "loss": 1.159,
      "step": 1360
    },
    {
      "epoch": 0.42,
      "grad_norm": 24.344266891479492,
      "learning_rate": 8.765610817611006e-06,
      "loss": 3.629,
      "step": 1361
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.541095733642578,
      "learning_rate": 8.758952916890174e-06,
      "loss": 1.2108,
      "step": 1362
    },
    {
      "epoch": 0.42,
      "grad_norm": 12.004585266113281,
      "learning_rate": 8.752293396479428e-06,
      "loss": 4.9698,
      "step": 1363
    },
    {
      "epoch": 0.42,
      "grad_norm": 14.796895980834961,
      "learning_rate": 8.74563226269044e-06,
      "loss": 5.4403,
      "step": 1364
    },
    {
      "epoch": 0.42,
      "grad_norm": 21.8315486907959,
      "learning_rate": 8.738969521836422e-06,
      "loss": 3.5713,
      "step": 1365
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.761943817138672,
      "learning_rate": 8.732305180232102e-06,
      "loss": 1.5461,
      "step": 1366
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.959177017211914,
      "learning_rate": 8.72563924419372e-06,
      "loss": 3.3256,
      "step": 1367
    },
    {
      "epoch": 0.42,
      "grad_norm": 13.9632568359375,
      "learning_rate": 8.71897172003904e-06,
      "loss": 2.1056,
      "step": 1368
    },
    {
      "epoch": 0.42,
      "grad_norm": 21.102760314941406,
      "learning_rate": 8.71230261408732e-06,
      "loss": 3.0056,
      "step": 1369
    },
    {
      "epoch": 0.42,
      "grad_norm": 8.353947639465332,
      "learning_rate": 8.705631932659325e-06,
      "loss": 1.3257,
      "step": 1370
    },
    {
      "epoch": 0.42,
      "grad_norm": 11.950407028198242,
      "learning_rate": 8.698959682077306e-06,
      "loss": 0.9416,
      "step": 1371
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.873854637145996,
      "learning_rate": 8.692285868665007e-06,
      "loss": 1.8988,
      "step": 1372
    },
    {
      "epoch": 0.43,
      "grad_norm": 16.22323226928711,
      "learning_rate": 8.68561049874765e-06,
      "loss": 3.3277,
      "step": 1373
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.70744514465332,
      "learning_rate": 8.678933578651935e-06,
      "loss": 2.6929,
      "step": 1374
    },
    {
      "epoch": 0.43,
      "grad_norm": 14.4379301071167,
      "learning_rate": 8.672255114706025e-06,
      "loss": 1.7289,
      "step": 1375
    },
    {
      "epoch": 0.43,
      "grad_norm": 14.871354103088379,
      "learning_rate": 8.665575113239554e-06,
      "loss": 1.4159,
      "step": 1376
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.862020492553711,
      "learning_rate": 8.658893580583613e-06,
      "loss": 1.3267,
      "step": 1377
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.802543640136719,
      "learning_rate": 8.652210523070735e-06,
      "loss": 2.1546,
      "step": 1378
    },
    {
      "epoch": 0.43,
      "grad_norm": 15.65637493133545,
      "learning_rate": 8.645525947034904e-06,
      "loss": 2.8595,
      "step": 1379
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.872232437133789,
      "learning_rate": 8.63883985881155e-06,
      "loss": 1.7195,
      "step": 1380
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.933382987976074,
      "learning_rate": 8.632152264737524e-06,
      "loss": 1.3872,
      "step": 1381
    },
    {
      "epoch": 0.43,
      "grad_norm": 10.490205764770508,
      "learning_rate": 8.62546317115111e-06,
      "loss": 1.3498,
      "step": 1382
    },
    {
      "epoch": 0.43,
      "grad_norm": 16.36244773864746,
      "learning_rate": 8.618772584392016e-06,
      "loss": 3.7617,
      "step": 1383
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.92510986328125,
      "learning_rate": 8.612080510801363e-06,
      "loss": 1.2954,
      "step": 1384
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.136679649353027,
      "learning_rate": 8.605386956721675e-06,
      "loss": 1.1087,
      "step": 1385
    },
    {
      "epoch": 0.43,
      "grad_norm": 15.491793632507324,
      "learning_rate": 8.59869192849689e-06,
      "loss": 2.1803,
      "step": 1386
    },
    {
      "epoch": 0.43,
      "grad_norm": 16.49875831604004,
      "learning_rate": 8.591995432472336e-06,
      "loss": 5.0587,
      "step": 1387
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.55068302154541,
      "learning_rate": 8.58529747499473e-06,
      "loss": 2.4693,
      "step": 1388
    },
    {
      "epoch": 0.43,
      "grad_norm": 9.14673137664795,
      "learning_rate": 8.578598062412183e-06,
      "loss": 1.4954,
      "step": 1389
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.553098678588867,
      "learning_rate": 8.571897201074177e-06,
      "loss": 1.6114,
      "step": 1390
    },
    {
      "epoch": 0.43,
      "grad_norm": 17.280818939208984,
      "learning_rate": 8.565194897331574e-06,
      "loss": 3.3327,
      "step": 1391
    },
    {
      "epoch": 0.43,
      "grad_norm": 13.013246536254883,
      "learning_rate": 8.558491157536597e-06,
      "loss": 0.8913,
      "step": 1392
    },
    {
      "epoch": 0.43,
      "grad_norm": 14.715285301208496,
      "learning_rate": 8.551785988042829e-06,
      "loss": 1.5306,
      "step": 1393
    },
    {
      "epoch": 0.43,
      "grad_norm": 14.097271919250488,
      "learning_rate": 8.545079395205217e-06,
      "loss": 6.3287,
      "step": 1394
    },
    {
      "epoch": 0.43,
      "grad_norm": 13.477266311645508,
      "learning_rate": 8.538371385380045e-06,
      "loss": 2.2853,
      "step": 1395
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.329754829406738,
      "learning_rate": 8.531661964924951e-06,
      "loss": 1.3214,
      "step": 1396
    },
    {
      "epoch": 0.43,
      "grad_norm": 15.463603019714355,
      "learning_rate": 8.524951140198904e-06,
      "loss": 5.906,
      "step": 1397
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.154081344604492,
      "learning_rate": 8.518238917562208e-06,
      "loss": 1.3551,
      "step": 1398
    },
    {
      "epoch": 0.43,
      "grad_norm": 17.804229736328125,
      "learning_rate": 8.511525303376481e-06,
      "loss": 2.2535,
      "step": 1399
    },
    {
      "epoch": 0.43,
      "grad_norm": 12.435379028320312,
      "learning_rate": 8.504810304004674e-06,
      "loss": 1.1441,
      "step": 1400
    },
    {
      "epoch": 0.43,
      "grad_norm": 13.349984169006348,
      "learning_rate": 8.498093925811045e-06,
      "loss": 2.4738,
      "step": 1401
    },
    {
      "epoch": 0.43,
      "grad_norm": 11.454378128051758,
      "learning_rate": 8.491376175161158e-06,
      "loss": 1.077,
      "step": 1402
    },
    {
      "epoch": 0.43,
      "grad_norm": 19.132919311523438,
      "learning_rate": 8.484657058421874e-06,
      "loss": 4.4179,
      "step": 1403
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.526947021484375,
      "learning_rate": 8.477936581961358e-06,
      "loss": 1.3308,
      "step": 1404
    },
    {
      "epoch": 0.44,
      "grad_norm": 18.873443603515625,
      "learning_rate": 8.471214752149058e-06,
      "loss": 4.2097,
      "step": 1405
    },
    {
      "epoch": 0.44,
      "grad_norm": 12.703725814819336,
      "learning_rate": 8.464491575355702e-06,
      "loss": 2.342,
      "step": 1406
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.23741626739502,
      "learning_rate": 8.4577670579533e-06,
      "loss": 1.5917,
      "step": 1407
    },
    {
      "epoch": 0.44,
      "grad_norm": 14.981216430664062,
      "learning_rate": 8.451041206315132e-06,
      "loss": 2.0091,
      "step": 1408
    },
    {
      "epoch": 0.44,
      "grad_norm": 18.34746742248535,
      "learning_rate": 8.444314026815741e-06,
      "loss": 7.7797,
      "step": 1409
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.101547241210938,
      "learning_rate": 8.437585525830926e-06,
      "loss": 5.0375,
      "step": 1410
    },
    {
      "epoch": 0.44,
      "grad_norm": 22.929941177368164,
      "learning_rate": 8.430855709737744e-06,
      "loss": 3.8084,
      "step": 1411
    },
    {
      "epoch": 0.44,
      "grad_norm": 17.568201065063477,
      "learning_rate": 8.424124584914495e-06,
      "loss": 1.9364,
      "step": 1412
    },
    {
      "epoch": 0.44,
      "grad_norm": 10.255859375,
      "learning_rate": 8.417392157740718e-06,
      "loss": 1.0762,
      "step": 1413
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.532630920410156,
      "learning_rate": 8.41065843459719e-06,
      "loss": 5.2018,
      "step": 1414
    },
    {
      "epoch": 0.44,
      "grad_norm": 14.951210975646973,
      "learning_rate": 8.403923421865912e-06,
      "loss": 1.2097,
      "step": 1415
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.789084434509277,
      "learning_rate": 8.397187125930114e-06,
      "loss": 1.6282,
      "step": 1416
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.053868293762207,
      "learning_rate": 8.390449553174235e-06,
      "loss": 2.3437,
      "step": 1417
    },
    {
      "epoch": 0.44,
      "grad_norm": 17.386356353759766,
      "learning_rate": 8.383710709983924e-06,
      "loss": 2.5349,
      "step": 1418
    },
    {
      "epoch": 0.44,
      "grad_norm": 14.942131042480469,
      "learning_rate": 8.376970602746046e-06,
      "loss": 1.484,
      "step": 1419
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.587835311889648,
      "learning_rate": 8.370229237848645e-06,
      "loss": 5.3537,
      "step": 1420
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.863311767578125,
      "learning_rate": 8.363486621680974e-06,
      "loss": 1.8422,
      "step": 1421
    },
    {
      "epoch": 0.44,
      "grad_norm": 15.610124588012695,
      "learning_rate": 8.356742760633463e-06,
      "loss": 1.3224,
      "step": 1422
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.711703300476074,
      "learning_rate": 8.349997661097723e-06,
      "loss": 3.4481,
      "step": 1423
    },
    {
      "epoch": 0.44,
      "grad_norm": 16.342193603515625,
      "learning_rate": 8.343251329466537e-06,
      "loss": 2.0732,
      "step": 1424
    },
    {
      "epoch": 0.44,
      "grad_norm": 25.153188705444336,
      "learning_rate": 8.336503772133865e-06,
      "loss": 5.9056,
      "step": 1425
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.40982723236084,
      "learning_rate": 8.329754995494821e-06,
      "loss": 2.0932,
      "step": 1426
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.98735523223877,
      "learning_rate": 8.32300500594567e-06,
      "loss": 2.5621,
      "step": 1427
    },
    {
      "epoch": 0.44,
      "grad_norm": 12.737622261047363,
      "learning_rate": 8.316253809883834e-06,
      "loss": 1.433,
      "step": 1428
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.427014350891113,
      "learning_rate": 8.309501413707881e-06,
      "loss": 1.0897,
      "step": 1429
    },
    {
      "epoch": 0.44,
      "grad_norm": 19.95100975036621,
      "learning_rate": 8.302747823817512e-06,
      "loss": 5.7549,
      "step": 1430
    },
    {
      "epoch": 0.44,
      "grad_norm": 11.455700874328613,
      "learning_rate": 8.295993046613552e-06,
      "loss": 1.9752,
      "step": 1431
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.410773277282715,
      "learning_rate": 8.289237088497966e-06,
      "loss": 2.232,
      "step": 1432
    },
    {
      "epoch": 0.44,
      "grad_norm": 9.498603820800781,
      "learning_rate": 8.282479955873832e-06,
      "loss": 1.2018,
      "step": 1433
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.638213157653809,
      "learning_rate": 8.275721655145333e-06,
      "loss": 2.6624,
      "step": 1434
    },
    {
      "epoch": 0.44,
      "grad_norm": 13.837607383728027,
      "learning_rate": 8.268962192717772e-06,
      "loss": 2.8477,
      "step": 1435
    },
    {
      "epoch": 0.44,
      "grad_norm": 8.231826782226562,
      "learning_rate": 8.262201574997547e-06,
      "loss": 1.0987,
      "step": 1436
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.587052345275879,
      "learning_rate": 8.255439808392148e-06,
      "loss": 6.8916,
      "step": 1437
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.058676719665527,
      "learning_rate": 8.248676899310161e-06,
      "loss": 1.1781,
      "step": 1438
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.14787483215332,
      "learning_rate": 8.241912854161249e-06,
      "loss": 1.4953,
      "step": 1439
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.948318481445312,
      "learning_rate": 8.235147679356154e-06,
      "loss": 1.6075,
      "step": 1440
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.693050384521484,
      "learning_rate": 8.228381381306688e-06,
      "loss": 1.8597,
      "step": 1441
    },
    {
      "epoch": 0.45,
      "grad_norm": 15.089034080505371,
      "learning_rate": 8.221613966425727e-06,
      "loss": 1.687,
      "step": 1442
    },
    {
      "epoch": 0.45,
      "grad_norm": 9.826478004455566,
      "learning_rate": 8.21484544112721e-06,
      "loss": 1.4379,
      "step": 1443
    },
    {
      "epoch": 0.45,
      "grad_norm": 12.025218963623047,
      "learning_rate": 8.208075811826121e-06,
      "loss": 1.1459,
      "step": 1444
    },
    {
      "epoch": 0.45,
      "grad_norm": 15.948945045471191,
      "learning_rate": 8.201305084938495e-06,
      "loss": 1.8357,
      "step": 1445
    },
    {
      "epoch": 0.45,
      "grad_norm": 20.572792053222656,
      "learning_rate": 8.194533266881407e-06,
      "loss": 3.8687,
      "step": 1446
    },
    {
      "epoch": 0.45,
      "grad_norm": 19.228185653686523,
      "learning_rate": 8.187760364072969e-06,
      "loss": 1.8467,
      "step": 1447
    },
    {
      "epoch": 0.45,
      "grad_norm": 12.649337768554688,
      "learning_rate": 8.18098638293231e-06,
      "loss": 2.9462,
      "step": 1448
    },
    {
      "epoch": 0.45,
      "grad_norm": 23.110645294189453,
      "learning_rate": 8.174211329879596e-06,
      "loss": 2.1425,
      "step": 1449
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.246688842773438,
      "learning_rate": 8.167435211335998e-06,
      "loss": 3.2245,
      "step": 1450
    },
    {
      "epoch": 0.45,
      "grad_norm": 25.37168312072754,
      "learning_rate": 8.160658033723702e-06,
      "loss": 3.2721,
      "step": 1451
    },
    {
      "epoch": 0.45,
      "grad_norm": 8.314859390258789,
      "learning_rate": 8.153879803465893e-06,
      "loss": 1.0462,
      "step": 1452
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.370716094970703,
      "learning_rate": 8.14710052698676e-06,
      "loss": 1.8953,
      "step": 1453
    },
    {
      "epoch": 0.45,
      "grad_norm": 17.507314682006836,
      "learning_rate": 8.14032021071148e-06,
      "loss": 1.5388,
      "step": 1454
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.845351219177246,
      "learning_rate": 8.133538861066215e-06,
      "loss": 6.5737,
      "step": 1455
    },
    {
      "epoch": 0.45,
      "grad_norm": 17.046667098999023,
      "learning_rate": 8.12675648447811e-06,
      "loss": 3.2323,
      "step": 1456
    },
    {
      "epoch": 0.45,
      "grad_norm": 10.533256530761719,
      "learning_rate": 8.119973087375277e-06,
      "loss": 1.7338,
      "step": 1457
    },
    {
      "epoch": 0.45,
      "grad_norm": 25.88167381286621,
      "learning_rate": 8.1131886761868e-06,
      "loss": 5.856,
      "step": 1458
    },
    {
      "epoch": 0.45,
      "grad_norm": 16.5262451171875,
      "learning_rate": 8.106403257342725e-06,
      "loss": 3.2691,
      "step": 1459
    },
    {
      "epoch": 0.45,
      "grad_norm": 19.846923828125,
      "learning_rate": 8.099616837274047e-06,
      "loss": 3.5881,
      "step": 1460
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.813898086547852,
      "learning_rate": 8.092829422412718e-06,
      "loss": 3.8796,
      "step": 1461
    },
    {
      "epoch": 0.45,
      "grad_norm": 15.77165412902832,
      "learning_rate": 8.086041019191626e-06,
      "loss": 2.4237,
      "step": 1462
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.56583023071289,
      "learning_rate": 8.0792516340446e-06,
      "loss": 2.3957,
      "step": 1463
    },
    {
      "epoch": 0.45,
      "grad_norm": 11.302949905395508,
      "learning_rate": 8.072461273406399e-06,
      "loss": 5.0651,
      "step": 1464
    },
    {
      "epoch": 0.45,
      "grad_norm": 13.356122016906738,
      "learning_rate": 8.065669943712703e-06,
      "loss": 1.3629,
      "step": 1465
    },
    {
      "epoch": 0.45,
      "grad_norm": 17.472110748291016,
      "learning_rate": 8.058877651400112e-06,
      "loss": 4.4384,
      "step": 1466
    },
    {
      "epoch": 0.45,
      "grad_norm": 18.280729293823242,
      "learning_rate": 8.052084402906144e-06,
      "loss": 2.0282,
      "step": 1467
    },
    {
      "epoch": 0.45,
      "grad_norm": 14.480973243713379,
      "learning_rate": 8.045290204669218e-06,
      "loss": 2.558,
      "step": 1468
    },
    {
      "epoch": 0.46,
      "grad_norm": 17.44344711303711,
      "learning_rate": 8.03849506312865e-06,
      "loss": 2.1654,
      "step": 1469
    },
    {
      "epoch": 0.46,
      "grad_norm": 16.810806274414062,
      "learning_rate": 8.031698984724656e-06,
      "loss": 2.3641,
      "step": 1470
    },
    {
      "epoch": 0.46,
      "grad_norm": 15.981531143188477,
      "learning_rate": 8.024901975898337e-06,
      "loss": 3.1273,
      "step": 1471
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.475582122802734,
      "learning_rate": 8.018104043091681e-06,
      "loss": 1.7297,
      "step": 1472
    },
    {
      "epoch": 0.46,
      "grad_norm": 15.845691680908203,
      "learning_rate": 8.011305192747539e-06,
      "loss": 5.2557,
      "step": 1473
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.488033294677734,
      "learning_rate": 8.004505431309649e-06,
      "loss": 4.55,
      "step": 1474
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.891829490661621,
      "learning_rate": 7.997704765222598e-06,
      "loss": 4.3927,
      "step": 1475
    },
    {
      "epoch": 0.46,
      "grad_norm": 17.33945083618164,
      "learning_rate": 7.990903200931836e-06,
      "loss": 5.5007,
      "step": 1476
    },
    {
      "epoch": 0.46,
      "grad_norm": 22.563276290893555,
      "learning_rate": 7.984100744883664e-06,
      "loss": 3.6111,
      "step": 1477
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.822282791137695,
      "learning_rate": 7.97729740352523e-06,
      "loss": 1.7473,
      "step": 1478
    },
    {
      "epoch": 0.46,
      "grad_norm": 22.096668243408203,
      "learning_rate": 7.97049318330452e-06,
      "loss": 3.3873,
      "step": 1479
    },
    {
      "epoch": 0.46,
      "grad_norm": 11.21049976348877,
      "learning_rate": 7.96368809067035e-06,
      "loss": 2.4169,
      "step": 1480
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.811819076538086,
      "learning_rate": 7.956882132072366e-06,
      "loss": 4.7143,
      "step": 1481
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.399057388305664,
      "learning_rate": 7.950075313961034e-06,
      "loss": 1.4499,
      "step": 1482
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.199051856994629,
      "learning_rate": 7.943267642787636e-06,
      "loss": 1.3204,
      "step": 1483
    },
    {
      "epoch": 0.46,
      "grad_norm": 10.940443992614746,
      "learning_rate": 7.936459125004257e-06,
      "loss": 2.3052,
      "step": 1484
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.73052406311035,
      "learning_rate": 7.929649767063791e-06,
      "loss": 1.48,
      "step": 1485
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.472209930419922,
      "learning_rate": 7.92283957541993e-06,
      "loss": 1.8377,
      "step": 1486
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.428339958190918,
      "learning_rate": 7.916028556527143e-06,
      "loss": 0.9211,
      "step": 1487
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.35672378540039,
      "learning_rate": 7.909216716840696e-06,
      "loss": 6.0892,
      "step": 1488
    },
    {
      "epoch": 0.46,
      "grad_norm": 16.624366760253906,
      "learning_rate": 7.902404062816633e-06,
      "loss": 5.7379,
      "step": 1489
    },
    {
      "epoch": 0.46,
      "grad_norm": 8.005545616149902,
      "learning_rate": 7.89559060091176e-06,
      "loss": 0.5775,
      "step": 1490
    },
    {
      "epoch": 0.46,
      "grad_norm": 14.2269287109375,
      "learning_rate": 7.888776337583654e-06,
      "loss": 4.0219,
      "step": 1491
    },
    {
      "epoch": 0.46,
      "grad_norm": 16.19810676574707,
      "learning_rate": 7.881961279290658e-06,
      "loss": 2.1824,
      "step": 1492
    },
    {
      "epoch": 0.46,
      "grad_norm": 17.658531188964844,
      "learning_rate": 7.875145432491855e-06,
      "loss": 4.4864,
      "step": 1493
    },
    {
      "epoch": 0.46,
      "grad_norm": 17.673933029174805,
      "learning_rate": 7.868328803647085e-06,
      "loss": 6.9672,
      "step": 1494
    },
    {
      "epoch": 0.46,
      "grad_norm": 9.747373580932617,
      "learning_rate": 7.861511399216929e-06,
      "loss": 1.101,
      "step": 1495
    },
    {
      "epoch": 0.46,
      "grad_norm": 12.639095306396484,
      "learning_rate": 7.854693225662701e-06,
      "loss": 5.1661,
      "step": 1496
    },
    {
      "epoch": 0.46,
      "grad_norm": 13.06181812286377,
      "learning_rate": 7.847874289446439e-06,
      "loss": 1.5592,
      "step": 1497
    },
    {
      "epoch": 0.46,
      "grad_norm": 18.21415138244629,
      "learning_rate": 7.841054597030913e-06,
      "loss": 2.8689,
      "step": 1498
    },
    {
      "epoch": 0.46,
      "grad_norm": 16.860862731933594,
      "learning_rate": 7.834234154879605e-06,
      "loss": 2.0175,
      "step": 1499
    },
    {
      "epoch": 0.46,
      "grad_norm": 15.08491039276123,
      "learning_rate": 7.827412969456708e-06,
      "loss": 1.1815,
      "step": 1500
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.941483497619629,
      "learning_rate": 7.820591047227116e-06,
      "loss": 1.7555,
      "step": 1501
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.32042694091797,
      "learning_rate": 7.81376839465643e-06,
      "loss": 4.6121,
      "step": 1502
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.87294864654541,
      "learning_rate": 7.806945018210935e-06,
      "loss": 1.4974,
      "step": 1503
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.907079696655273,
      "learning_rate": 7.800120924357604e-06,
      "loss": 2.2645,
      "step": 1504
    },
    {
      "epoch": 0.47,
      "grad_norm": 7.117086887359619,
      "learning_rate": 7.793296119564092e-06,
      "loss": 0.8516,
      "step": 1505
    },
    {
      "epoch": 0.47,
      "grad_norm": 15.370654106140137,
      "learning_rate": 7.786470610298726e-06,
      "loss": 1.7695,
      "step": 1506
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.78589630126953,
      "learning_rate": 7.779644403030505e-06,
      "loss": 3.1191,
      "step": 1507
    },
    {
      "epoch": 0.47,
      "grad_norm": 10.148062705993652,
      "learning_rate": 7.772817504229082e-06,
      "loss": 1.8681,
      "step": 1508
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.689626693725586,
      "learning_rate": 7.765989920364767e-06,
      "loss": 2.2163,
      "step": 1509
    },
    {
      "epoch": 0.47,
      "grad_norm": 14.091196060180664,
      "learning_rate": 7.759161657908529e-06,
      "loss": 2.1025,
      "step": 1510
    },
    {
      "epoch": 0.47,
      "grad_norm": 9.173744201660156,
      "learning_rate": 7.752332723331967e-06,
      "loss": 1.3896,
      "step": 1511
    },
    {
      "epoch": 0.47,
      "grad_norm": 17.604738235473633,
      "learning_rate": 7.745503123107325e-06,
      "loss": 2.8429,
      "step": 1512
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.202844619750977,
      "learning_rate": 7.738672863707473e-06,
      "loss": 0.9668,
      "step": 1513
    },
    {
      "epoch": 0.47,
      "grad_norm": 17.14142417907715,
      "learning_rate": 7.731841951605915e-06,
      "loss": 4.8675,
      "step": 1514
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.663723945617676,
      "learning_rate": 7.72501039327676e-06,
      "loss": 2.7225,
      "step": 1515
    },
    {
      "epoch": 0.47,
      "grad_norm": 15.818323135375977,
      "learning_rate": 7.71817819519474e-06,
      "loss": 5.5613,
      "step": 1516
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.62367057800293,
      "learning_rate": 7.711345363835186e-06,
      "loss": 3.1708,
      "step": 1517
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.760972023010254,
      "learning_rate": 7.70451190567404e-06,
      "loss": 2.2154,
      "step": 1518
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.931819915771484,
      "learning_rate": 7.697677827187824e-06,
      "loss": 1.8667,
      "step": 1519
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.671175003051758,
      "learning_rate": 7.69084313485366e-06,
      "loss": 2.2861,
      "step": 1520
    },
    {
      "epoch": 0.47,
      "grad_norm": 12.393848419189453,
      "learning_rate": 7.684007835149245e-06,
      "loss": 1.9357,
      "step": 1521
    },
    {
      "epoch": 0.47,
      "grad_norm": 11.369022369384766,
      "learning_rate": 7.67717193455285e-06,
      "loss": 1.3349,
      "step": 1522
    },
    {
      "epoch": 0.47,
      "grad_norm": 15.161210060119629,
      "learning_rate": 7.670335439543325e-06,
      "loss": 2.1201,
      "step": 1523
    },
    {
      "epoch": 0.47,
      "grad_norm": 15.16026782989502,
      "learning_rate": 7.663498356600075e-06,
      "loss": 2.0202,
      "step": 1524
    },
    {
      "epoch": 0.47,
      "grad_norm": 6.391641616821289,
      "learning_rate": 7.656660692203063e-06,
      "loss": 0.7958,
      "step": 1525
    },
    {
      "epoch": 0.47,
      "grad_norm": 14.75525951385498,
      "learning_rate": 7.649822452832805e-06,
      "loss": 2.917,
      "step": 1526
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.168161392211914,
      "learning_rate": 7.642983644970363e-06,
      "loss": 4.8536,
      "step": 1527
    },
    {
      "epoch": 0.47,
      "grad_norm": 17.493988037109375,
      "learning_rate": 7.636144275097334e-06,
      "loss": 5.3522,
      "step": 1528
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.88285255432129,
      "learning_rate": 7.629304349695853e-06,
      "loss": 6.0407,
      "step": 1529
    },
    {
      "epoch": 0.47,
      "grad_norm": 16.397403717041016,
      "learning_rate": 7.622463875248577e-06,
      "loss": 2.8976,
      "step": 1530
    },
    {
      "epoch": 0.47,
      "grad_norm": 20.834239959716797,
      "learning_rate": 7.615622858238685e-06,
      "loss": 2.0012,
      "step": 1531
    },
    {
      "epoch": 0.47,
      "grad_norm": 13.2479887008667,
      "learning_rate": 7.608781305149871e-06,
      "loss": 1.7579,
      "step": 1532
    },
    {
      "epoch": 0.47,
      "grad_norm": 8.735552787780762,
      "learning_rate": 7.6019392224663325e-06,
      "loss": 0.7553,
      "step": 1533
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.764557838439941,
      "learning_rate": 7.595096616672776e-06,
      "loss": 3.706,
      "step": 1534
    },
    {
      "epoch": 0.48,
      "grad_norm": 10.473408699035645,
      "learning_rate": 7.588253494254403e-06,
      "loss": 1.5389,
      "step": 1535
    },
    {
      "epoch": 0.48,
      "grad_norm": 15.223381996154785,
      "learning_rate": 7.581409861696897e-06,
      "loss": 4.5035,
      "step": 1536
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.932395935058594,
      "learning_rate": 7.574565725486429e-06,
      "loss": 3.2436,
      "step": 1537
    },
    {
      "epoch": 0.48,
      "grad_norm": 15.13408088684082,
      "learning_rate": 7.567721092109656e-06,
      "loss": 1.5579,
      "step": 1538
    },
    {
      "epoch": 0.48,
      "grad_norm": 13.730198860168457,
      "learning_rate": 7.560875968053692e-06,
      "loss": 2.021,
      "step": 1539
    },
    {
      "epoch": 0.48,
      "grad_norm": 27.59801483154297,
      "learning_rate": 7.5540303598061235e-06,
      "loss": 3.208,
      "step": 1540
    },
    {
      "epoch": 0.48,
      "grad_norm": 16.457393646240234,
      "learning_rate": 7.547184273854998e-06,
      "loss": 2.2529,
      "step": 1541
    },
    {
      "epoch": 0.48,
      "grad_norm": 13.959450721740723,
      "learning_rate": 7.540337716688809e-06,
      "loss": 2.5864,
      "step": 1542
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.607627868652344,
      "learning_rate": 7.533490694796502e-06,
      "loss": 4.8759,
      "step": 1543
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.483607292175293,
      "learning_rate": 7.526643214667463e-06,
      "loss": 1.1964,
      "step": 1544
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.84099578857422,
      "learning_rate": 7.519795282791509e-06,
      "loss": 1.5788,
      "step": 1545
    },
    {
      "epoch": 0.48,
      "grad_norm": 11.038904190063477,
      "learning_rate": 7.5129469056588865e-06,
      "loss": 1.5326,
      "step": 1546
    },
    {
      "epoch": 0.48,
      "grad_norm": 18.416704177856445,
      "learning_rate": 7.506098089760266e-06,
      "loss": 5.7714,
      "step": 1547
    },
    {
      "epoch": 0.48,
      "grad_norm": 20.013147354125977,
      "learning_rate": 7.4992488415867285e-06,
      "loss": 2.925,
      "step": 1548
    },
    {
      "epoch": 0.48,
      "grad_norm": 15.000293731689453,
      "learning_rate": 7.492399167629774e-06,
      "loss": 5.5733,
      "step": 1549
    },
    {
      "epoch": 0.48,
      "grad_norm": 15.219602584838867,
      "learning_rate": 7.485549074381296e-06,
      "loss": 4.8389,
      "step": 1550
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.503700256347656,
      "learning_rate": 7.4786985683335895e-06,
      "loss": 1.1435,
      "step": 1551
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.639745712280273,
      "learning_rate": 7.471847655979343e-06,
      "loss": 2.0322,
      "step": 1552
    },
    {
      "epoch": 0.48,
      "grad_norm": 13.07108211517334,
      "learning_rate": 7.46499634381163e-06,
      "loss": 2.7517,
      "step": 1553
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.419761657714844,
      "learning_rate": 7.4581446383238945e-06,
      "loss": 1.1141,
      "step": 1554
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.002758026123047,
      "learning_rate": 7.451292546009965e-06,
      "loss": 1.991,
      "step": 1555
    },
    {
      "epoch": 0.48,
      "grad_norm": 14.649391174316406,
      "learning_rate": 7.444440073364034e-06,
      "loss": 1.7529,
      "step": 1556
    },
    {
      "epoch": 0.48,
      "grad_norm": 17.31776237487793,
      "learning_rate": 7.437587226880645e-06,
      "loss": 2.3223,
      "step": 1557
    },
    {
      "epoch": 0.48,
      "grad_norm": 18.440576553344727,
      "learning_rate": 7.430734013054705e-06,
      "loss": 2.425,
      "step": 1558
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.171570777893066,
      "learning_rate": 7.42388043838147e-06,
      "loss": 2.0255,
      "step": 1559
    },
    {
      "epoch": 0.48,
      "grad_norm": 16.10645866394043,
      "learning_rate": 7.417026509356532e-06,
      "loss": 1.5669,
      "step": 1560
    },
    {
      "epoch": 0.48,
      "grad_norm": 16.45768165588379,
      "learning_rate": 7.410172232475823e-06,
      "loss": 2.2515,
      "step": 1561
    },
    {
      "epoch": 0.48,
      "grad_norm": 13.46390438079834,
      "learning_rate": 7.403317614235602e-06,
      "loss": 1.03,
      "step": 1562
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.802998542785645,
      "learning_rate": 7.396462661132454e-06,
      "loss": 0.9395,
      "step": 1563
    },
    {
      "epoch": 0.48,
      "grad_norm": 16.94371223449707,
      "learning_rate": 7.389607379663279e-06,
      "loss": 5.0141,
      "step": 1564
    },
    {
      "epoch": 0.48,
      "grad_norm": 12.43297004699707,
      "learning_rate": 7.3827517763252906e-06,
      "loss": 4.5835,
      "step": 1565
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.801984786987305,
      "learning_rate": 7.375895857616006e-06,
      "loss": 0.9518,
      "step": 1566
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.065658569335938,
      "learning_rate": 7.369039630033244e-06,
      "loss": 2.1409,
      "step": 1567
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.188167572021484,
      "learning_rate": 7.362183100075111e-06,
      "loss": 1.1734,
      "step": 1568
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.572041511535645,
      "learning_rate": 7.355326274240001e-06,
      "loss": 2.2057,
      "step": 1569
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.975041389465332,
      "learning_rate": 7.348469159026596e-06,
      "loss": 3.8706,
      "step": 1570
    },
    {
      "epoch": 0.49,
      "grad_norm": 10.509519577026367,
      "learning_rate": 7.341611760933841e-06,
      "loss": 1.0018,
      "step": 1571
    },
    {
      "epoch": 0.49,
      "grad_norm": 15.144615173339844,
      "learning_rate": 7.3347540864609555e-06,
      "loss": 3.0489,
      "step": 1572
    },
    {
      "epoch": 0.49,
      "grad_norm": 20.42429542541504,
      "learning_rate": 7.327896142107424e-06,
      "loss": 5.1546,
      "step": 1573
    },
    {
      "epoch": 0.49,
      "grad_norm": 16.69227409362793,
      "learning_rate": 7.321037934372978e-06,
      "loss": 2.9772,
      "step": 1574
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.439099311828613,
      "learning_rate": 7.314179469757607e-06,
      "loss": 1.4989,
      "step": 1575
    },
    {
      "epoch": 0.49,
      "grad_norm": 9.062239646911621,
      "learning_rate": 7.307320754761536e-06,
      "loss": 1.4532,
      "step": 1576
    },
    {
      "epoch": 0.49,
      "grad_norm": 21.60405158996582,
      "learning_rate": 7.300461795885237e-06,
      "loss": 2.0011,
      "step": 1577
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.495698928833008,
      "learning_rate": 7.293602599629405e-06,
      "loss": 1.81,
      "step": 1578
    },
    {
      "epoch": 0.49,
      "grad_norm": 11.463644981384277,
      "learning_rate": 7.286743172494962e-06,
      "loss": 1.8246,
      "step": 1579
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.608999252319336,
      "learning_rate": 7.279883520983053e-06,
      "loss": 1.7617,
      "step": 1580
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.266855239868164,
      "learning_rate": 7.273023651595029e-06,
      "loss": 1.3808,
      "step": 1581
    },
    {
      "epoch": 0.49,
      "grad_norm": 16.99720573425293,
      "learning_rate": 7.26616357083245e-06,
      "loss": 3.412,
      "step": 1582
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.8319673538208,
      "learning_rate": 7.25930328519708e-06,
      "loss": 2.5051,
      "step": 1583
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.773306846618652,
      "learning_rate": 7.252442801190875e-06,
      "loss": 1.4383,
      "step": 1584
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.512967109680176,
      "learning_rate": 7.245582125315974e-06,
      "loss": 1.763,
      "step": 1585
    },
    {
      "epoch": 0.49,
      "grad_norm": 16.133769989013672,
      "learning_rate": 7.238721264074704e-06,
      "loss": 2.7118,
      "step": 1586
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.554754257202148,
      "learning_rate": 7.2318602239695685e-06,
      "loss": 1.2321,
      "step": 1587
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.490348815917969,
      "learning_rate": 7.224999011503238e-06,
      "loss": 1.9627,
      "step": 1588
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.49413013458252,
      "learning_rate": 7.21813763317854e-06,
      "loss": 2.0481,
      "step": 1589
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.60959243774414,
      "learning_rate": 7.211276095498471e-06,
      "loss": 1.4864,
      "step": 1590
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.422796249389648,
      "learning_rate": 7.204414404966173e-06,
      "loss": 1.6967,
      "step": 1591
    },
    {
      "epoch": 0.49,
      "grad_norm": 14.81104564666748,
      "learning_rate": 7.197552568084928e-06,
      "loss": 1.9473,
      "step": 1592
    },
    {
      "epoch": 0.49,
      "grad_norm": 15.132305145263672,
      "learning_rate": 7.190690591358166e-06,
      "loss": 5.9458,
      "step": 1593
    },
    {
      "epoch": 0.49,
      "grad_norm": 16.62466049194336,
      "learning_rate": 7.183828481289444e-06,
      "loss": 4.3815,
      "step": 1594
    },
    {
      "epoch": 0.49,
      "grad_norm": 13.80553150177002,
      "learning_rate": 7.176966244382446e-06,
      "loss": 6.7008,
      "step": 1595
    },
    {
      "epoch": 0.49,
      "grad_norm": 12.077177047729492,
      "learning_rate": 7.170103887140977e-06,
      "loss": 1.7794,
      "step": 1596
    },
    {
      "epoch": 0.49,
      "grad_norm": 17.30846405029297,
      "learning_rate": 7.163241416068956e-06,
      "loss": 1.4915,
      "step": 1597
    },
    {
      "epoch": 0.5,
      "grad_norm": 13.239306449890137,
      "learning_rate": 7.156378837670409e-06,
      "loss": 1.6999,
      "step": 1598
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.074838638305664,
      "learning_rate": 7.149516158449465e-06,
      "loss": 3.0839,
      "step": 1599
    },
    {
      "epoch": 0.5,
      "grad_norm": 18.072620391845703,
      "learning_rate": 7.142653384910344e-06,
      "loss": 3.7544,
      "step": 1600
    },
    {
      "epoch": 0.5,
      "grad_norm": 19.225902557373047,
      "learning_rate": 7.1357905235573654e-06,
      "loss": 6.3921,
      "step": 1601
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.992647171020508,
      "learning_rate": 7.1289275808949235e-06,
      "loss": 5.8246,
      "step": 1602
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.987789154052734,
      "learning_rate": 7.122064563427488e-06,
      "loss": 2.5646,
      "step": 1603
    },
    {
      "epoch": 0.5,
      "grad_norm": 13.11773681640625,
      "learning_rate": 7.115201477659608e-06,
      "loss": 4.8927,
      "step": 1604
    },
    {
      "epoch": 0.5,
      "grad_norm": 16.583314895629883,
      "learning_rate": 7.108338330095892e-06,
      "loss": 1.8204,
      "step": 1605
    },
    {
      "epoch": 0.5,
      "grad_norm": 9.47932243347168,
      "learning_rate": 7.101475127241006e-06,
      "loss": 1.191,
      "step": 1606
    },
    {
      "epoch": 0.5,
      "grad_norm": 11.21604061126709,
      "learning_rate": 7.0946118755996695e-06,
      "loss": 1.4498,
      "step": 1607
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.17425537109375,
      "learning_rate": 7.087748581676651e-06,
      "loss": 1.8093,
      "step": 1608
    },
    {
      "epoch": 0.5,
      "grad_norm": 12.51001262664795,
      "learning_rate": 7.080885251976759e-06,
      "loss": 1.7423,
      "step": 1609
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.638405799865723,
      "learning_rate": 7.074021893004827e-06,
      "loss": 1.3783,
      "step": 1610
    },
    {
      "epoch": 0.5,
      "grad_norm": 13.604153633117676,
      "learning_rate": 7.06715851126573e-06,
      "loss": 5.2793,
      "step": 1611
    },
    {
      "epoch": 0.5,
      "grad_norm": 15.841768264770508,
      "learning_rate": 7.060295113264354e-06,
      "loss": 5.742,
      "step": 1612
    },
    {
      "epoch": 0.5,
      "grad_norm": 16.581165313720703,
      "learning_rate": 7.053431705505605e-06,
      "loss": 2.986,
      "step": 1613
    },
    {
      "epoch": 0.5,
      "grad_norm": 23.555326461791992,
      "learning_rate": 7.046568294494397e-06,
      "loss": 2.2116,
      "step": 1614
    },
    {
      "epoch": 0.5,
      "grad_norm": 16.073575973510742,
      "learning_rate": 7.039704886735647e-06,
      "loss": 2.3679,
      "step": 1615
    },
    {
      "epoch": 0.5,
      "grad_norm": 21.275436401367188,
      "learning_rate": 7.032841488734272e-06,
      "loss": 2.3565,
      "step": 1616
    },
    {
      "epoch": 0.5,
      "grad_norm": 21.215978622436523,
      "learning_rate": 7.025978106995174e-06,
      "loss": 2.5173,
      "step": 1617
    },
    {
      "epoch": 0.5,
      "grad_norm": 18.45775032043457,
      "learning_rate": 7.019114748023242e-06,
      "loss": 2.4042,
      "step": 1618
    },
    {
      "epoch": 0.5,
      "grad_norm": 13.314290046691895,
      "learning_rate": 7.01225141832335e-06,
      "loss": 2.5132,
      "step": 1619
    },
    {
      "epoch": 0.5,
      "grad_norm": 24.080631256103516,
      "learning_rate": 7.005388124400333e-06,
      "loss": 2.0978,
      "step": 1620
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.010071754455566,
      "learning_rate": 6.9985248727589945e-06,
      "loss": 5.0885,
      "step": 1621
    },
    {
      "epoch": 0.5,
      "grad_norm": 8.733227729797363,
      "learning_rate": 6.991661669904109e-06,
      "loss": 0.8556,
      "step": 1622
    },
    {
      "epoch": 0.5,
      "grad_norm": 15.029264450073242,
      "learning_rate": 6.984798522340394e-06,
      "loss": 1.3525,
      "step": 1623
    },
    {
      "epoch": 0.5,
      "grad_norm": 16.164785385131836,
      "learning_rate": 6.977935436572511e-06,
      "loss": 1.9743,
      "step": 1624
    },
    {
      "epoch": 0.5,
      "grad_norm": 10.85263729095459,
      "learning_rate": 6.971072419105078e-06,
      "loss": 2.056,
      "step": 1625
    },
    {
      "epoch": 0.5,
      "grad_norm": 14.404804229736328,
      "learning_rate": 6.964209476442636e-06,
      "loss": 3.1708,
      "step": 1626
    },
    {
      "epoch": 0.5,
      "grad_norm": 13.646852493286133,
      "learning_rate": 6.957346615089655e-06,
      "loss": 2.62,
      "step": 1627
    },
    {
      "epoch": 0.5,
      "grad_norm": 23.522275924682617,
      "learning_rate": 6.9504838415505375e-06,
      "loss": 1.4201,
      "step": 1628
    },
    {
      "epoch": 0.5,
      "grad_norm": 16.97157096862793,
      "learning_rate": 6.943621162329593e-06,
      "loss": 2.1315,
      "step": 1629
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.685445785522461,
      "learning_rate": 6.936758583931044e-06,
      "loss": 1.3641,
      "step": 1630
    },
    {
      "epoch": 0.51,
      "grad_norm": 15.746297836303711,
      "learning_rate": 6.929896112859023e-06,
      "loss": 2.0654,
      "step": 1631
    },
    {
      "epoch": 0.51,
      "grad_norm": 17.248628616333008,
      "learning_rate": 6.923033755617555e-06,
      "loss": 5.3468,
      "step": 1632
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.610515594482422,
      "learning_rate": 6.916171518710556e-06,
      "loss": 2.0751,
      "step": 1633
    },
    {
      "epoch": 0.51,
      "grad_norm": 18.109254837036133,
      "learning_rate": 6.9093094086418344e-06,
      "loss": 4.59,
      "step": 1634
    },
    {
      "epoch": 0.51,
      "grad_norm": 19.82781410217285,
      "learning_rate": 6.902447431915073e-06,
      "loss": 3.4236,
      "step": 1635
    },
    {
      "epoch": 0.51,
      "grad_norm": 16.590665817260742,
      "learning_rate": 6.895585595033829e-06,
      "loss": 1.6634,
      "step": 1636
    },
    {
      "epoch": 0.51,
      "grad_norm": 16.019447326660156,
      "learning_rate": 6.88872390450153e-06,
      "loss": 2.4793,
      "step": 1637
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.343634605407715,
      "learning_rate": 6.881862366821462e-06,
      "loss": 1.5978,
      "step": 1638
    },
    {
      "epoch": 0.51,
      "grad_norm": 19.105314254760742,
      "learning_rate": 6.875000988496763e-06,
      "loss": 3.1883,
      "step": 1639
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.569679260253906,
      "learning_rate": 6.868139776030432e-06,
      "loss": 2.315,
      "step": 1640
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.843108177185059,
      "learning_rate": 6.8612787359252966e-06,
      "loss": 2.538,
      "step": 1641
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.337444305419922,
      "learning_rate": 6.854417874684026e-06,
      "loss": 1.852,
      "step": 1642
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.132347106933594,
      "learning_rate": 6.847557198809127e-06,
      "loss": 0.8586,
      "step": 1643
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.987866401672363,
      "learning_rate": 6.840696714802922e-06,
      "loss": 2.9912,
      "step": 1644
    },
    {
      "epoch": 0.51,
      "grad_norm": 17.009370803833008,
      "learning_rate": 6.8338364291675495e-06,
      "loss": 4.2418,
      "step": 1645
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.241983413696289,
      "learning_rate": 6.826976348404973e-06,
      "loss": 2.1306,
      "step": 1646
    },
    {
      "epoch": 0.51,
      "grad_norm": 14.771096229553223,
      "learning_rate": 6.8201164790169494e-06,
      "loss": 4.413,
      "step": 1647
    },
    {
      "epoch": 0.51,
      "grad_norm": 8.216975212097168,
      "learning_rate": 6.813256827505037e-06,
      "loss": 1.2711,
      "step": 1648
    },
    {
      "epoch": 0.51,
      "grad_norm": 17.802562713623047,
      "learning_rate": 6.806397400370596e-06,
      "loss": 1.8284,
      "step": 1649
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.894693374633789,
      "learning_rate": 6.7995382041147645e-06,
      "loss": 1.1183,
      "step": 1650
    },
    {
      "epoch": 0.51,
      "grad_norm": 15.150440216064453,
      "learning_rate": 6.792679245238464e-06,
      "loss": 4.8429,
      "step": 1651
    },
    {
      "epoch": 0.51,
      "grad_norm": 13.411111831665039,
      "learning_rate": 6.785820530242395e-06,
      "loss": 2.5226,
      "step": 1652
    },
    {
      "epoch": 0.51,
      "grad_norm": 7.378268241882324,
      "learning_rate": 6.7789620656270235e-06,
      "loss": 0.6978,
      "step": 1653
    },
    {
      "epoch": 0.51,
      "grad_norm": 15.269695281982422,
      "learning_rate": 6.772103857892577e-06,
      "loss": 3.3351,
      "step": 1654
    },
    {
      "epoch": 0.51,
      "grad_norm": 15.948509216308594,
      "learning_rate": 6.765245913539046e-06,
      "loss": 8.719,
      "step": 1655
    },
    {
      "epoch": 0.51,
      "grad_norm": 11.423715591430664,
      "learning_rate": 6.7583882390661615e-06,
      "loss": 1.6719,
      "step": 1656
    },
    {
      "epoch": 0.51,
      "grad_norm": 10.186993598937988,
      "learning_rate": 6.751530840973407e-06,
      "loss": 1.3835,
      "step": 1657
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.548033714294434,
      "learning_rate": 6.74467372576e-06,
      "loss": 1.9285,
      "step": 1658
    },
    {
      "epoch": 0.51,
      "grad_norm": 16.2474308013916,
      "learning_rate": 6.737816899924892e-06,
      "loss": 1.4393,
      "step": 1659
    },
    {
      "epoch": 0.51,
      "grad_norm": 12.502729415893555,
      "learning_rate": 6.730960369966756e-06,
      "loss": 0.9805,
      "step": 1660
    },
    {
      "epoch": 0.51,
      "grad_norm": 14.059015274047852,
      "learning_rate": 6.724104142383995e-06,
      "loss": 2.7396,
      "step": 1661
    },
    {
      "epoch": 0.51,
      "grad_norm": 14.44599723815918,
      "learning_rate": 6.717248223674711e-06,
      "loss": 1.7818,
      "step": 1662
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.191947937011719,
      "learning_rate": 6.710392620336721e-06,
      "loss": 1.2289,
      "step": 1663
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.764534950256348,
      "learning_rate": 6.703537338867547e-06,
      "loss": 1.9445,
      "step": 1664
    },
    {
      "epoch": 0.52,
      "grad_norm": 13.833651542663574,
      "learning_rate": 6.696682385764401e-06,
      "loss": 1.4044,
      "step": 1665
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.326014518737793,
      "learning_rate": 6.689827767524177e-06,
      "loss": 1.3071,
      "step": 1666
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.583909034729004,
      "learning_rate": 6.6829734906434685e-06,
      "loss": 4.8377,
      "step": 1667
    },
    {
      "epoch": 0.52,
      "grad_norm": 12.085514068603516,
      "learning_rate": 6.676119561618531e-06,
      "loss": 1.1961,
      "step": 1668
    },
    {
      "epoch": 0.52,
      "grad_norm": 19.26174545288086,
      "learning_rate": 6.669265986945295e-06,
      "loss": 6.4908,
      "step": 1669
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.472599983215332,
      "learning_rate": 6.662412773119357e-06,
      "loss": 1.5512,
      "step": 1670
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.320133209228516,
      "learning_rate": 6.655559926635969e-06,
      "loss": 1.3109,
      "step": 1671
    },
    {
      "epoch": 0.52,
      "grad_norm": 17.44525909423828,
      "learning_rate": 6.648707453990034e-06,
      "loss": 2.0053,
      "step": 1672
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.975166320800781,
      "learning_rate": 6.641855361676106e-06,
      "loss": 2.8781,
      "step": 1673
    },
    {
      "epoch": 0.52,
      "grad_norm": 16.109355926513672,
      "learning_rate": 6.635003656188374e-06,
      "loss": 5.3664,
      "step": 1674
    },
    {
      "epoch": 0.52,
      "grad_norm": 16.004348754882812,
      "learning_rate": 6.628152344020658e-06,
      "loss": 3.1915,
      "step": 1675
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.833353996276855,
      "learning_rate": 6.621301431666412e-06,
      "loss": 2.4988,
      "step": 1676
    },
    {
      "epoch": 0.52,
      "grad_norm": 14.966736793518066,
      "learning_rate": 6.614450925618708e-06,
      "loss": 1.7351,
      "step": 1677
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.273491859436035,
      "learning_rate": 6.607600832370228e-06,
      "loss": 1.276,
      "step": 1678
    },
    {
      "epoch": 0.52,
      "grad_norm": 9.694404602050781,
      "learning_rate": 6.600751158413272e-06,
      "loss": 2.3457,
      "step": 1679
    },
    {
      "epoch": 0.52,
      "grad_norm": 17.113826751708984,
      "learning_rate": 6.593901910239737e-06,
      "loss": 4.0584,
      "step": 1680
    },
    {
      "epoch": 0.52,
      "grad_norm": 18.73774528503418,
      "learning_rate": 6.587053094341113e-06,
      "loss": 4.5252,
      "step": 1681
    },
    {
      "epoch": 0.52,
      "grad_norm": 17.822935104370117,
      "learning_rate": 6.580204717208493e-06,
      "loss": 2.3868,
      "step": 1682
    },
    {
      "epoch": 0.52,
      "grad_norm": 16.173681259155273,
      "learning_rate": 6.573356785332539e-06,
      "loss": 2.5625,
      "step": 1683
    },
    {
      "epoch": 0.52,
      "grad_norm": 10.989896774291992,
      "learning_rate": 6.566509305203497e-06,
      "loss": 1.3217,
      "step": 1684
    },
    {
      "epoch": 0.52,
      "grad_norm": 11.736306190490723,
      "learning_rate": 6.559662283311192e-06,
      "loss": 1.5926,
      "step": 1685
    },
    {
      "epoch": 0.52,
      "grad_norm": 19.356401443481445,
      "learning_rate": 6.552815726145004e-06,
      "loss": 4.2014,
      "step": 1686
    },
    {
      "epoch": 0.52,
      "grad_norm": 15.877721786499023,
      "learning_rate": 6.545969640193877e-06,
      "loss": 3.4804,
      "step": 1687
    },
    {
      "epoch": 0.52,
      "grad_norm": 17.891929626464844,
      "learning_rate": 6.53912403194631e-06,
      "loss": 4.1462,
      "step": 1688
    },
    {
      "epoch": 0.52,
      "grad_norm": 13.31762409210205,
      "learning_rate": 6.532278907890346e-06,
      "loss": 2.0393,
      "step": 1689
    },
    {
      "epoch": 0.52,
      "grad_norm": 8.635902404785156,
      "learning_rate": 6.5254342745135695e-06,
      "loss": 0.7918,
      "step": 1690
    },
    {
      "epoch": 0.52,
      "grad_norm": 13.983869552612305,
      "learning_rate": 6.518590138303105e-06,
      "loss": 2.2377,
      "step": 1691
    },
    {
      "epoch": 0.52,
      "grad_norm": 19.41148567199707,
      "learning_rate": 6.511746505745599e-06,
      "loss": 4.7251,
      "step": 1692
    },
    {
      "epoch": 0.52,
      "grad_norm": 16.147748947143555,
      "learning_rate": 6.5049033833272235e-06,
      "loss": 1.7898,
      "step": 1693
    },
    {
      "epoch": 0.52,
      "grad_norm": 17.849939346313477,
      "learning_rate": 6.498060777533669e-06,
      "loss": 2.1782,
      "step": 1694
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.791414260864258,
      "learning_rate": 6.491218694850132e-06,
      "loss": 2.4036,
      "step": 1695
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.15569496154785,
      "learning_rate": 6.484377141761316e-06,
      "loss": 1.7765,
      "step": 1696
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.212052345275879,
      "learning_rate": 6.477536124751425e-06,
      "loss": 1.9845,
      "step": 1697
    },
    {
      "epoch": 0.53,
      "grad_norm": 18.090255737304688,
      "learning_rate": 6.47069565030415e-06,
      "loss": 2.8683,
      "step": 1698
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.535499572753906,
      "learning_rate": 6.463855724902666e-06,
      "loss": 3.7178,
      "step": 1699
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.417211532592773,
      "learning_rate": 6.45701635502964e-06,
      "loss": 2.572,
      "step": 1700
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.360429763793945,
      "learning_rate": 6.450177547167198e-06,
      "loss": 2.4635,
      "step": 1701
    },
    {
      "epoch": 0.53,
      "grad_norm": 13.608604431152344,
      "learning_rate": 6.443339307796939e-06,
      "loss": 1.9137,
      "step": 1702
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.365615844726562,
      "learning_rate": 6.436501643399927e-06,
      "loss": 3.9508,
      "step": 1703
    },
    {
      "epoch": 0.53,
      "grad_norm": 32.33718490600586,
      "learning_rate": 6.429664560456678e-06,
      "loss": 2.7259,
      "step": 1704
    },
    {
      "epoch": 0.53,
      "grad_norm": 23.710081100463867,
      "learning_rate": 6.42282806544715e-06,
      "loss": 2.7618,
      "step": 1705
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.952068328857422,
      "learning_rate": 6.415992164850757e-06,
      "loss": 2.031,
      "step": 1706
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.720380783081055,
      "learning_rate": 6.409156865146342e-06,
      "loss": 1.9651,
      "step": 1707
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.775768280029297,
      "learning_rate": 6.402322172812176e-06,
      "loss": 0.8801,
      "step": 1708
    },
    {
      "epoch": 0.53,
      "grad_norm": 7.3918023109436035,
      "learning_rate": 6.395488094325961e-06,
      "loss": 0.9,
      "step": 1709
    },
    {
      "epoch": 0.53,
      "grad_norm": 12.462408065795898,
      "learning_rate": 6.3886546361648145e-06,
      "loss": 2.0359,
      "step": 1710
    },
    {
      "epoch": 0.53,
      "grad_norm": 9.982486724853516,
      "learning_rate": 6.381821804805262e-06,
      "loss": 2.5078,
      "step": 1711
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.839106559753418,
      "learning_rate": 6.374989606723242e-06,
      "loss": 4.6393,
      "step": 1712
    },
    {
      "epoch": 0.53,
      "grad_norm": 20.70285987854004,
      "learning_rate": 6.368158048394088e-06,
      "loss": 6.814,
      "step": 1713
    },
    {
      "epoch": 0.53,
      "grad_norm": 10.38366985321045,
      "learning_rate": 6.361327136292526e-06,
      "loss": 1.0995,
      "step": 1714
    },
    {
      "epoch": 0.53,
      "grad_norm": 12.559454917907715,
      "learning_rate": 6.354496876892677e-06,
      "loss": 2.4292,
      "step": 1715
    },
    {
      "epoch": 0.53,
      "grad_norm": 12.452824592590332,
      "learning_rate": 6.347667276668035e-06,
      "loss": 2.3951,
      "step": 1716
    },
    {
      "epoch": 0.53,
      "grad_norm": 13.784807205200195,
      "learning_rate": 6.340838342091473e-06,
      "loss": 1.9206,
      "step": 1717
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.844255447387695,
      "learning_rate": 6.334010079635234e-06,
      "loss": 2.962,
      "step": 1718
    },
    {
      "epoch": 0.53,
      "grad_norm": 14.486095428466797,
      "learning_rate": 6.327182495770922e-06,
      "loss": 3.1889,
      "step": 1719
    },
    {
      "epoch": 0.53,
      "grad_norm": 13.291302680969238,
      "learning_rate": 6.320355596969496e-06,
      "loss": 2.257,
      "step": 1720
    },
    {
      "epoch": 0.53,
      "grad_norm": 13.94046688079834,
      "learning_rate": 6.313529389701275e-06,
      "loss": 3.026,
      "step": 1721
    },
    {
      "epoch": 0.53,
      "grad_norm": 15.41063404083252,
      "learning_rate": 6.306703880435911e-06,
      "loss": 1.7961,
      "step": 1722
    },
    {
      "epoch": 0.53,
      "grad_norm": 17.081377029418945,
      "learning_rate": 6.299879075642397e-06,
      "loss": 2.9324,
      "step": 1723
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.9708309173584,
      "learning_rate": 6.293054981789066e-06,
      "loss": 2.3154,
      "step": 1724
    },
    {
      "epoch": 0.53,
      "grad_norm": 11.27573013305664,
      "learning_rate": 6.286231605343573e-06,
      "loss": 1.9319,
      "step": 1725
    },
    {
      "epoch": 0.53,
      "grad_norm": 16.793363571166992,
      "learning_rate": 6.279408952772883e-06,
      "loss": 1.8106,
      "step": 1726
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.946924209594727,
      "learning_rate": 6.272587030543293e-06,
      "loss": 1.7559,
      "step": 1727
    },
    {
      "epoch": 0.54,
      "grad_norm": 17.80219078063965,
      "learning_rate": 6.2657658451203954e-06,
      "loss": 2.3931,
      "step": 1728
    },
    {
      "epoch": 0.54,
      "grad_norm": 16.310911178588867,
      "learning_rate": 6.258945402969087e-06,
      "loss": 3.2346,
      "step": 1729
    },
    {
      "epoch": 0.54,
      "grad_norm": 16.45994758605957,
      "learning_rate": 6.2521257105535616e-06,
      "loss": 2.502,
      "step": 1730
    },
    {
      "epoch": 0.54,
      "grad_norm": 17.91693878173828,
      "learning_rate": 6.2453067743373025e-06,
      "loss": 4.5985,
      "step": 1731
    },
    {
      "epoch": 0.54,
      "grad_norm": 16.53362464904785,
      "learning_rate": 6.238488600783072e-06,
      "loss": 2.1074,
      "step": 1732
    },
    {
      "epoch": 0.54,
      "grad_norm": 17.455644607543945,
      "learning_rate": 6.2316711963529155e-06,
      "loss": 5.2828,
      "step": 1733
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.842496871948242,
      "learning_rate": 6.224854567508148e-06,
      "loss": 1.6192,
      "step": 1734
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.071833610534668,
      "learning_rate": 6.218038720709345e-06,
      "loss": 1.0014,
      "step": 1735
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.850285530090332,
      "learning_rate": 6.211223662416347e-06,
      "loss": 1.2009,
      "step": 1736
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.262269020080566,
      "learning_rate": 6.204409399088243e-06,
      "loss": 1.4618,
      "step": 1737
    },
    {
      "epoch": 0.54,
      "grad_norm": 18.779258728027344,
      "learning_rate": 6.197595937183369e-06,
      "loss": 4.6567,
      "step": 1738
    },
    {
      "epoch": 0.54,
      "grad_norm": 15.584208488464355,
      "learning_rate": 6.190783283159304e-06,
      "loss": 2.299,
      "step": 1739
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.968876838684082,
      "learning_rate": 6.183971443472859e-06,
      "loss": 2.518,
      "step": 1740
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.949677467346191,
      "learning_rate": 6.177160424580071e-06,
      "loss": 1.4009,
      "step": 1741
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.316316604614258,
      "learning_rate": 6.1703502329362096e-06,
      "loss": 2.442,
      "step": 1742
    },
    {
      "epoch": 0.54,
      "grad_norm": 17.885141372680664,
      "learning_rate": 6.163540874995746e-06,
      "loss": 2.819,
      "step": 1743
    },
    {
      "epoch": 0.54,
      "grad_norm": 12.24764633178711,
      "learning_rate": 6.156732357212366e-06,
      "loss": 2.331,
      "step": 1744
    },
    {
      "epoch": 0.54,
      "grad_norm": 10.27761173248291,
      "learning_rate": 6.149924686038967e-06,
      "loss": 1.0084,
      "step": 1745
    },
    {
      "epoch": 0.54,
      "grad_norm": 11.830199241638184,
      "learning_rate": 6.143117867927637e-06,
      "loss": 1.8183,
      "step": 1746
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.59326171875,
      "learning_rate": 6.136311909329651e-06,
      "loss": 2.6811,
      "step": 1747
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.359524726867676,
      "learning_rate": 6.1295068166954815e-06,
      "loss": 1.511,
      "step": 1748
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.620701789855957,
      "learning_rate": 6.122702596474772e-06,
      "loss": 1.9805,
      "step": 1749
    },
    {
      "epoch": 0.54,
      "grad_norm": 12.86448860168457,
      "learning_rate": 6.115899255116336e-06,
      "loss": 1.7906,
      "step": 1750
    },
    {
      "epoch": 0.54,
      "grad_norm": 12.143721580505371,
      "learning_rate": 6.109096799068166e-06,
      "loss": 2.337,
      "step": 1751
    },
    {
      "epoch": 0.54,
      "grad_norm": 14.398964881896973,
      "learning_rate": 6.1022952347774054e-06,
      "loss": 4.1856,
      "step": 1752
    },
    {
      "epoch": 0.54,
      "grad_norm": 12.750125885009766,
      "learning_rate": 6.095494568690353e-06,
      "loss": 1.9975,
      "step": 1753
    },
    {
      "epoch": 0.54,
      "grad_norm": 15.612987518310547,
      "learning_rate": 6.088694807252462e-06,
      "loss": 1.4707,
      "step": 1754
    },
    {
      "epoch": 0.54,
      "grad_norm": 9.288985252380371,
      "learning_rate": 6.081895956908323e-06,
      "loss": 1.0508,
      "step": 1755
    },
    {
      "epoch": 0.54,
      "grad_norm": 13.260685920715332,
      "learning_rate": 6.0750980241016635e-06,
      "loss": 1.7748,
      "step": 1756
    },
    {
      "epoch": 0.54,
      "grad_norm": 16.665273666381836,
      "learning_rate": 6.068301015275347e-06,
      "loss": 5.0814,
      "step": 1757
    },
    {
      "epoch": 0.54,
      "grad_norm": 12.461769104003906,
      "learning_rate": 6.061504936871353e-06,
      "loss": 1.2751,
      "step": 1758
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.850014686584473,
      "learning_rate": 6.054709795330784e-06,
      "loss": 2.1483,
      "step": 1759
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.211248397827148,
      "learning_rate": 6.047915597093859e-06,
      "loss": 2.166,
      "step": 1760
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.773746490478516,
      "learning_rate": 6.04112234859989e-06,
      "loss": 3.4716,
      "step": 1761
    },
    {
      "epoch": 0.55,
      "grad_norm": 19.844953536987305,
      "learning_rate": 6.034330056287299e-06,
      "loss": 4.2191,
      "step": 1762
    },
    {
      "epoch": 0.55,
      "grad_norm": 18.28035545349121,
      "learning_rate": 6.027538726593603e-06,
      "loss": 6.1553,
      "step": 1763
    },
    {
      "epoch": 0.55,
      "grad_norm": 14.571393013000488,
      "learning_rate": 6.020748365955403e-06,
      "loss": 2.0841,
      "step": 1764
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.416486740112305,
      "learning_rate": 6.0139589808083746e-06,
      "loss": 2.7342,
      "step": 1765
    },
    {
      "epoch": 0.55,
      "grad_norm": 11.846794128417969,
      "learning_rate": 6.007170577587284e-06,
      "loss": 2.8779,
      "step": 1766
    },
    {
      "epoch": 0.55,
      "grad_norm": 14.38280200958252,
      "learning_rate": 6.000383162725954e-06,
      "loss": 2.1532,
      "step": 1767
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.458091735839844,
      "learning_rate": 5.993596742657276e-06,
      "loss": 2.2165,
      "step": 1768
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.689918518066406,
      "learning_rate": 5.986811323813201e-06,
      "loss": 4.93,
      "step": 1769
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.026179313659668,
      "learning_rate": 5.9800269126247255e-06,
      "loss": 2.9913,
      "step": 1770
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.321549415588379,
      "learning_rate": 5.973243515521892e-06,
      "loss": 4.5672,
      "step": 1771
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.819236755371094,
      "learning_rate": 5.966461138933786e-06,
      "loss": 3.0129,
      "step": 1772
    },
    {
      "epoch": 0.55,
      "grad_norm": 14.934184074401855,
      "learning_rate": 5.9596797892885224e-06,
      "loss": 1.6563,
      "step": 1773
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.410440444946289,
      "learning_rate": 5.952899473013241e-06,
      "loss": 1.7962,
      "step": 1774
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.91740894317627,
      "learning_rate": 5.946120196534109e-06,
      "loss": 2.1482,
      "step": 1775
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.390769958496094,
      "learning_rate": 5.939341966276302e-06,
      "loss": 1.7179,
      "step": 1776
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.820837020874023,
      "learning_rate": 5.932564788664004e-06,
      "loss": 3.3391,
      "step": 1777
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.530945777893066,
      "learning_rate": 5.9257886701204064e-06,
      "loss": 5.1542,
      "step": 1778
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.225460052490234,
      "learning_rate": 5.919013617067692e-06,
      "loss": 5.2362,
      "step": 1779
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.23628044128418,
      "learning_rate": 5.9122396359270325e-06,
      "loss": 3.1039,
      "step": 1780
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.73491668701172,
      "learning_rate": 5.905466733118594e-06,
      "loss": 8.1694,
      "step": 1781
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.948612213134766,
      "learning_rate": 5.8986949150615075e-06,
      "loss": 2.5072,
      "step": 1782
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.731588363647461,
      "learning_rate": 5.89192418817388e-06,
      "loss": 2.9993,
      "step": 1783
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.084639549255371,
      "learning_rate": 5.885154558872792e-06,
      "loss": 2.9031,
      "step": 1784
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.718095779418945,
      "learning_rate": 5.878386033574273e-06,
      "loss": 2.0702,
      "step": 1785
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.161492347717285,
      "learning_rate": 5.871618618693314e-06,
      "loss": 3.5457,
      "step": 1786
    },
    {
      "epoch": 0.55,
      "grad_norm": 16.949766159057617,
      "learning_rate": 5.864852320643849e-06,
      "loss": 3.3263,
      "step": 1787
    },
    {
      "epoch": 0.55,
      "grad_norm": 13.630675315856934,
      "learning_rate": 5.8580871458387525e-06,
      "loss": 2.5225,
      "step": 1788
    },
    {
      "epoch": 0.55,
      "grad_norm": 15.821239471435547,
      "learning_rate": 5.85132310068984e-06,
      "loss": 1.9543,
      "step": 1789
    },
    {
      "epoch": 0.55,
      "grad_norm": 12.509420394897461,
      "learning_rate": 5.844560191607854e-06,
      "loss": 2.6313,
      "step": 1790
    },
    {
      "epoch": 0.55,
      "grad_norm": 9.080439567565918,
      "learning_rate": 5.837798425002454e-06,
      "loss": 0.8224,
      "step": 1791
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.666498184204102,
      "learning_rate": 5.831037807282229e-06,
      "loss": 2.6361,
      "step": 1792
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.119965553283691,
      "learning_rate": 5.824278344854669e-06,
      "loss": 1.4584,
      "step": 1793
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.555170059204102,
      "learning_rate": 5.81752004412617e-06,
      "loss": 1.4938,
      "step": 1794
    },
    {
      "epoch": 0.56,
      "grad_norm": 20.039859771728516,
      "learning_rate": 5.810762911502034e-06,
      "loss": 1.9742,
      "step": 1795
    },
    {
      "epoch": 0.56,
      "grad_norm": 19.904695510864258,
      "learning_rate": 5.8040069533864486e-06,
      "loss": 5.4046,
      "step": 1796
    },
    {
      "epoch": 0.56,
      "grad_norm": 15.574432373046875,
      "learning_rate": 5.7972521761824896e-06,
      "loss": 2.7755,
      "step": 1797
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.398809432983398,
      "learning_rate": 5.790498586292119e-06,
      "loss": 1.321,
      "step": 1798
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.65013313293457,
      "learning_rate": 5.783746190116167e-06,
      "loss": 3.4494,
      "step": 1799
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.127032279968262,
      "learning_rate": 5.7769949940543306e-06,
      "loss": 2.4075,
      "step": 1800
    },
    {
      "epoch": 0.56,
      "grad_norm": 17.058229446411133,
      "learning_rate": 5.770245004505181e-06,
      "loss": 4.8438,
      "step": 1801
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.680752754211426,
      "learning_rate": 5.763496227866136e-06,
      "loss": 2.608,
      "step": 1802
    },
    {
      "epoch": 0.56,
      "grad_norm": 11.628544807434082,
      "learning_rate": 5.756748670533461e-06,
      "loss": 2.0497,
      "step": 1803
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.606917381286621,
      "learning_rate": 5.75000233890228e-06,
      "loss": 1.8469,
      "step": 1804
    },
    {
      "epoch": 0.56,
      "grad_norm": 12.258424758911133,
      "learning_rate": 5.743257239366539e-06,
      "loss": 1.6223,
      "step": 1805
    },
    {
      "epoch": 0.56,
      "grad_norm": 13.137167930603027,
      "learning_rate": 5.7365133783190266e-06,
      "loss": 2.2344,
      "step": 1806
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.135931015014648,
      "learning_rate": 5.729770762151356e-06,
      "loss": 4.8953,
      "step": 1807
    },
    {
      "epoch": 0.56,
      "grad_norm": 12.336984634399414,
      "learning_rate": 5.723029397253957e-06,
      "loss": 2.1529,
      "step": 1808
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.23254108428955,
      "learning_rate": 5.7162892900160755e-06,
      "loss": 1.3341,
      "step": 1809
    },
    {
      "epoch": 0.56,
      "grad_norm": 15.352858543395996,
      "learning_rate": 5.709550446825767e-06,
      "loss": 2.2873,
      "step": 1810
    },
    {
      "epoch": 0.56,
      "grad_norm": 14.300989151000977,
      "learning_rate": 5.7028128740698884e-06,
      "loss": 4.2608,
      "step": 1811
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.150107383728027,
      "learning_rate": 5.696076578134088e-06,
      "loss": 1.1745,
      "step": 1812
    },
    {
      "epoch": 0.56,
      "grad_norm": 12.48327922821045,
      "learning_rate": 5.689341565402811e-06,
      "loss": 1.9706,
      "step": 1813
    },
    {
      "epoch": 0.56,
      "grad_norm": 15.741283416748047,
      "learning_rate": 5.6826078422592834e-06,
      "loss": 3.9089,
      "step": 1814
    },
    {
      "epoch": 0.56,
      "grad_norm": 18.629844665527344,
      "learning_rate": 5.675875415085506e-06,
      "loss": 4.4346,
      "step": 1815
    },
    {
      "epoch": 0.56,
      "grad_norm": 9.940484046936035,
      "learning_rate": 5.669144290262256e-06,
      "loss": 2.4403,
      "step": 1816
    },
    {
      "epoch": 0.56,
      "grad_norm": 19.276111602783203,
      "learning_rate": 5.6624144741690745e-06,
      "loss": 2.4949,
      "step": 1817
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.42538833618164,
      "learning_rate": 5.655685973184258e-06,
      "loss": 2.155,
      "step": 1818
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.172608375549316,
      "learning_rate": 5.648958793684869e-06,
      "loss": 2.2949,
      "step": 1819
    },
    {
      "epoch": 0.56,
      "grad_norm": 8.87829303741455,
      "learning_rate": 5.642232942046701e-06,
      "loss": 1.1954,
      "step": 1820
    },
    {
      "epoch": 0.56,
      "grad_norm": 10.703292846679688,
      "learning_rate": 5.635508424644298e-06,
      "loss": 1.1126,
      "step": 1821
    },
    {
      "epoch": 0.56,
      "grad_norm": 17.50180435180664,
      "learning_rate": 5.628785247850944e-06,
      "loss": 5.5155,
      "step": 1822
    },
    {
      "epoch": 0.56,
      "grad_norm": 15.151026725769043,
      "learning_rate": 5.622063418038643e-06,
      "loss": 2.6611,
      "step": 1823
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.896838188171387,
      "learning_rate": 5.615342941578126e-06,
      "loss": 1.8171,
      "step": 1824
    },
    {
      "epoch": 0.57,
      "grad_norm": 21.477943420410156,
      "learning_rate": 5.608623824838845e-06,
      "loss": 4.216,
      "step": 1825
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.832585334777832,
      "learning_rate": 5.6019060741889564e-06,
      "loss": 1.3468,
      "step": 1826
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.833240509033203,
      "learning_rate": 5.595189695995325e-06,
      "loss": 3.0509,
      "step": 1827
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.890311241149902,
      "learning_rate": 5.588474696623519e-06,
      "loss": 5.038,
      "step": 1828
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.075677871704102,
      "learning_rate": 5.5817610824377955e-06,
      "loss": 2.7405,
      "step": 1829
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.089751243591309,
      "learning_rate": 5.575048859801096e-06,
      "loss": 1.1481,
      "step": 1830
    },
    {
      "epoch": 0.57,
      "grad_norm": 17.9642391204834,
      "learning_rate": 5.568338035075049e-06,
      "loss": 1.942,
      "step": 1831
    },
    {
      "epoch": 0.57,
      "grad_norm": 20.85904312133789,
      "learning_rate": 5.561628614619956e-06,
      "loss": 3.8661,
      "step": 1832
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.438445091247559,
      "learning_rate": 5.554920604794785e-06,
      "loss": 1.3036,
      "step": 1833
    },
    {
      "epoch": 0.57,
      "grad_norm": 11.098956108093262,
      "learning_rate": 5.548214011957172e-06,
      "loss": 1.763,
      "step": 1834
    },
    {
      "epoch": 0.57,
      "grad_norm": 16.185504913330078,
      "learning_rate": 5.5415088424634065e-06,
      "loss": 2.5151,
      "step": 1835
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.786161422729492,
      "learning_rate": 5.534805102668425e-06,
      "loss": 1.4129,
      "step": 1836
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.491531372070312,
      "learning_rate": 5.528102798925823e-06,
      "loss": 2.4639,
      "step": 1837
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.029722213745117,
      "learning_rate": 5.521401937587819e-06,
      "loss": 2.2583,
      "step": 1838
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.83185863494873,
      "learning_rate": 5.51470252500527e-06,
      "loss": 1.6569,
      "step": 1839
    },
    {
      "epoch": 0.57,
      "grad_norm": 17.041156768798828,
      "learning_rate": 5.5080045675276666e-06,
      "loss": 1.8535,
      "step": 1840
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.895615577697754,
      "learning_rate": 5.501308071503113e-06,
      "loss": 1.8716,
      "step": 1841
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.872110366821289,
      "learning_rate": 5.494613043278324e-06,
      "loss": 1.9894,
      "step": 1842
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.688057899475098,
      "learning_rate": 5.487919489198638e-06,
      "loss": 2.6675,
      "step": 1843
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.86407470703125,
      "learning_rate": 5.481227415607984e-06,
      "loss": 1.2891,
      "step": 1844
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.483810424804688,
      "learning_rate": 5.47453682884889e-06,
      "loss": 2.4939,
      "step": 1845
    },
    {
      "epoch": 0.57,
      "grad_norm": 12.794442176818848,
      "learning_rate": 5.467847735262478e-06,
      "loss": 5.7496,
      "step": 1846
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.446375846862793,
      "learning_rate": 5.461160141188452e-06,
      "loss": 3.5324,
      "step": 1847
    },
    {
      "epoch": 0.57,
      "grad_norm": 10.231986045837402,
      "learning_rate": 5.454474052965096e-06,
      "loss": 1.116,
      "step": 1848
    },
    {
      "epoch": 0.57,
      "grad_norm": 14.563923835754395,
      "learning_rate": 5.447789476929268e-06,
      "loss": 2.5878,
      "step": 1849
    },
    {
      "epoch": 0.57,
      "grad_norm": 13.371225357055664,
      "learning_rate": 5.4411064194163895e-06,
      "loss": 1.1454,
      "step": 1850
    },
    {
      "epoch": 0.57,
      "grad_norm": 15.974309921264648,
      "learning_rate": 5.434424886760445e-06,
      "loss": 2.1937,
      "step": 1851
    },
    {
      "epoch": 0.57,
      "grad_norm": 17.46932601928711,
      "learning_rate": 5.427744885293975e-06,
      "loss": 2.9106,
      "step": 1852
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.957886695861816,
      "learning_rate": 5.4210664213480676e-06,
      "loss": 2.1941,
      "step": 1853
    },
    {
      "epoch": 0.57,
      "grad_norm": 9.03040885925293,
      "learning_rate": 5.41438950125235e-06,
      "loss": 1.5123,
      "step": 1854
    },
    {
      "epoch": 0.57,
      "grad_norm": 15.304936408996582,
      "learning_rate": 5.407714131334994e-06,
      "loss": 2.1541,
      "step": 1855
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.13270092010498,
      "learning_rate": 5.401040317922695e-06,
      "loss": 2.0546,
      "step": 1856
    },
    {
      "epoch": 0.58,
      "grad_norm": 18.850942611694336,
      "learning_rate": 5.394368067340675e-06,
      "loss": 4.6624,
      "step": 1857
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.505763053894043,
      "learning_rate": 5.387697385912681e-06,
      "loss": 2.6678,
      "step": 1858
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.811540603637695,
      "learning_rate": 5.381028279960963e-06,
      "loss": 3.3744,
      "step": 1859
    },
    {
      "epoch": 0.58,
      "grad_norm": 11.627664566040039,
      "learning_rate": 5.3743607558062794e-06,
      "loss": 2.7316,
      "step": 1860
    },
    {
      "epoch": 0.58,
      "grad_norm": 19.460006713867188,
      "learning_rate": 5.367694819767901e-06,
      "loss": 5.1478,
      "step": 1861
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.877521514892578,
      "learning_rate": 5.3610304781635805e-06,
      "loss": 1.6275,
      "step": 1862
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.138850212097168,
      "learning_rate": 5.35436773730956e-06,
      "loss": 2.4314,
      "step": 1863
    },
    {
      "epoch": 0.58,
      "grad_norm": 19.668495178222656,
      "learning_rate": 5.347706603520575e-06,
      "loss": 2.4506,
      "step": 1864
    },
    {
      "epoch": 0.58,
      "grad_norm": 12.425284385681152,
      "learning_rate": 5.341047083109828e-06,
      "loss": 1.6937,
      "step": 1865
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.025413513183594,
      "learning_rate": 5.334389182388994e-06,
      "loss": 2.6373,
      "step": 1866
    },
    {
      "epoch": 0.58,
      "grad_norm": 12.790462493896484,
      "learning_rate": 5.327732907668218e-06,
      "loss": 1.6166,
      "step": 1867
    },
    {
      "epoch": 0.58,
      "grad_norm": 19.806283950805664,
      "learning_rate": 5.321078265256098e-06,
      "loss": 3.5507,
      "step": 1868
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.250811576843262,
      "learning_rate": 5.314425261459687e-06,
      "loss": 1.5591,
      "step": 1869
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.386801719665527,
      "learning_rate": 5.307773902584487e-06,
      "loss": 1.7944,
      "step": 1870
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.107830047607422,
      "learning_rate": 5.30112419493444e-06,
      "loss": 2.4954,
      "step": 1871
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.570332527160645,
      "learning_rate": 5.294476144811917e-06,
      "loss": 4.9374,
      "step": 1872
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.250971794128418,
      "learning_rate": 5.2878297585177304e-06,
      "loss": 2.2374,
      "step": 1873
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.113117218017578,
      "learning_rate": 5.281185042351105e-06,
      "loss": 2.4667,
      "step": 1874
    },
    {
      "epoch": 0.58,
      "grad_norm": 10.013445854187012,
      "learning_rate": 5.274542002609686e-06,
      "loss": 1.7648,
      "step": 1875
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.309882164001465,
      "learning_rate": 5.267900645589531e-06,
      "loss": 4.6993,
      "step": 1876
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.022863388061523,
      "learning_rate": 5.261260977585103e-06,
      "loss": 2.9688,
      "step": 1877
    },
    {
      "epoch": 0.58,
      "grad_norm": 18.01882553100586,
      "learning_rate": 5.254623004889259e-06,
      "loss": 3.1605,
      "step": 1878
    },
    {
      "epoch": 0.58,
      "grad_norm": 9.066216468811035,
      "learning_rate": 5.2479867337932595e-06,
      "loss": 0.7356,
      "step": 1879
    },
    {
      "epoch": 0.58,
      "grad_norm": 13.521183967590332,
      "learning_rate": 5.241352170586743e-06,
      "loss": 2.7033,
      "step": 1880
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.285184860229492,
      "learning_rate": 5.234719321557728e-06,
      "loss": 1.6921,
      "step": 1881
    },
    {
      "epoch": 0.58,
      "grad_norm": 17.454221725463867,
      "learning_rate": 5.2280881929926205e-06,
      "loss": 6.739,
      "step": 1882
    },
    {
      "epoch": 0.58,
      "grad_norm": 12.460721015930176,
      "learning_rate": 5.221458791176186e-06,
      "loss": 2.4783,
      "step": 1883
    },
    {
      "epoch": 0.58,
      "grad_norm": 16.056852340698242,
      "learning_rate": 5.21483112239155e-06,
      "loss": 2.9951,
      "step": 1884
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.775067329406738,
      "learning_rate": 5.208205192920209e-06,
      "loss": 2.9683,
      "step": 1885
    },
    {
      "epoch": 0.58,
      "grad_norm": 15.0515775680542,
      "learning_rate": 5.201581009041998e-06,
      "loss": 2.7254,
      "step": 1886
    },
    {
      "epoch": 0.58,
      "grad_norm": 14.527585983276367,
      "learning_rate": 5.194958577035105e-06,
      "loss": 3.066,
      "step": 1887
    },
    {
      "epoch": 0.58,
      "grad_norm": 18.543670654296875,
      "learning_rate": 5.188337903176055e-06,
      "loss": 2.9837,
      "step": 1888
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.011702537536621,
      "learning_rate": 5.181718993739707e-06,
      "loss": 2.7422,
      "step": 1889
    },
    {
      "epoch": 0.59,
      "grad_norm": 16.387422561645508,
      "learning_rate": 5.175101854999245e-06,
      "loss": 3.4034,
      "step": 1890
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.453161239624023,
      "learning_rate": 5.168486493226182e-06,
      "loss": 1.6797,
      "step": 1891
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.211752891540527,
      "learning_rate": 5.161872914690342e-06,
      "loss": 1.725,
      "step": 1892
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.550117492675781,
      "learning_rate": 5.1552611256598555e-06,
      "loss": 2.9156,
      "step": 1893
    },
    {
      "epoch": 0.59,
      "grad_norm": 15.480829238891602,
      "learning_rate": 5.148651132401165e-06,
      "loss": 1.5313,
      "step": 1894
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.496414184570312,
      "learning_rate": 5.142042941179003e-06,
      "loss": 2.5291,
      "step": 1895
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.85752010345459,
      "learning_rate": 5.135436558256399e-06,
      "loss": 1.4814,
      "step": 1896
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.321053504943848,
      "learning_rate": 5.128831989894673e-06,
      "loss": 2.3097,
      "step": 1897
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.116474151611328,
      "learning_rate": 5.122229242353413e-06,
      "loss": 4.8876,
      "step": 1898
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.035852432250977,
      "learning_rate": 5.115628321890488e-06,
      "loss": 2.701,
      "step": 1899
    },
    {
      "epoch": 0.59,
      "grad_norm": 15.801783561706543,
      "learning_rate": 5.1090292347620405e-06,
      "loss": 1.6895,
      "step": 1900
    },
    {
      "epoch": 0.59,
      "grad_norm": 16.488258361816406,
      "learning_rate": 5.1024319872224695e-06,
      "loss": 4.6689,
      "step": 1901
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.51988983154297,
      "learning_rate": 5.095836585524424e-06,
      "loss": 2.0975,
      "step": 1902
    },
    {
      "epoch": 0.59,
      "grad_norm": 13.482269287109375,
      "learning_rate": 5.0892430359188205e-06,
      "loss": 1.374,
      "step": 1903
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.935653686523438,
      "learning_rate": 5.082651344654806e-06,
      "loss": 2.5465,
      "step": 1904
    },
    {
      "epoch": 0.59,
      "grad_norm": 16.7248592376709,
      "learning_rate": 5.076061517979769e-06,
      "loss": 2.3183,
      "step": 1905
    },
    {
      "epoch": 0.59,
      "grad_norm": 19.343509674072266,
      "learning_rate": 5.069473562139337e-06,
      "loss": 4.0582,
      "step": 1906
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.782135009765625,
      "learning_rate": 5.062887483377358e-06,
      "loss": 6.0777,
      "step": 1907
    },
    {
      "epoch": 0.59,
      "grad_norm": 16.98220443725586,
      "learning_rate": 5.056303287935901e-06,
      "loss": 1.7017,
      "step": 1908
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.168288230895996,
      "learning_rate": 5.0497209820552556e-06,
      "loss": 0.841,
      "step": 1909
    },
    {
      "epoch": 0.59,
      "grad_norm": 11.430391311645508,
      "learning_rate": 5.043140571973915e-06,
      "loss": 4.2754,
      "step": 1910
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.67452335357666,
      "learning_rate": 5.0365620639285754e-06,
      "loss": 1.0665,
      "step": 1911
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.424131393432617,
      "learning_rate": 5.029985464154138e-06,
      "loss": 2.9558,
      "step": 1912
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.27757453918457,
      "learning_rate": 5.023410778883685e-06,
      "loss": 4.7597,
      "step": 1913
    },
    {
      "epoch": 0.59,
      "grad_norm": 17.786062240600586,
      "learning_rate": 5.016838014348489e-06,
      "loss": 3.6501,
      "step": 1914
    },
    {
      "epoch": 0.59,
      "grad_norm": 9.792977333068848,
      "learning_rate": 5.010267176778005e-06,
      "loss": 4.768,
      "step": 1915
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.924040794372559,
      "learning_rate": 5.003698272399859e-06,
      "loss": 4.1235,
      "step": 1916
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.380438804626465,
      "learning_rate": 4.997131307439839e-06,
      "loss": 2.0261,
      "step": 1917
    },
    {
      "epoch": 0.59,
      "grad_norm": 10.831319808959961,
      "learning_rate": 4.99056628812191e-06,
      "loss": 2.286,
      "step": 1918
    },
    {
      "epoch": 0.59,
      "grad_norm": 14.513365745544434,
      "learning_rate": 4.984003220668176e-06,
      "loss": 3.7361,
      "step": 1919
    },
    {
      "epoch": 0.59,
      "grad_norm": 12.797149658203125,
      "learning_rate": 4.9774421112988976e-06,
      "loss": 5.2536,
      "step": 1920
    },
    {
      "epoch": 0.6,
      "grad_norm": 28.588335037231445,
      "learning_rate": 4.970882966232487e-06,
      "loss": 3.5549,
      "step": 1921
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.213167190551758,
      "learning_rate": 4.964325791685487e-06,
      "loss": 4.5068,
      "step": 1922
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.606828689575195,
      "learning_rate": 4.9577705938725665e-06,
      "loss": 1.4982,
      "step": 1923
    },
    {
      "epoch": 0.6,
      "grad_norm": 10.224977493286133,
      "learning_rate": 4.951217379006538e-06,
      "loss": 0.9959,
      "step": 1924
    },
    {
      "epoch": 0.6,
      "grad_norm": 16.293413162231445,
      "learning_rate": 4.9446661532983206e-06,
      "loss": 3.7225,
      "step": 1925
    },
    {
      "epoch": 0.6,
      "grad_norm": 17.75890350341797,
      "learning_rate": 4.938116922956952e-06,
      "loss": 3.3565,
      "step": 1926
    },
    {
      "epoch": 0.6,
      "grad_norm": 18.865028381347656,
      "learning_rate": 4.931569694189582e-06,
      "loss": 2.1867,
      "step": 1927
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.345128059387207,
      "learning_rate": 4.925024473201459e-06,
      "loss": 4.05,
      "step": 1928
    },
    {
      "epoch": 0.6,
      "grad_norm": 14.990592956542969,
      "learning_rate": 4.91848126619593e-06,
      "loss": 3.244,
      "step": 1929
    },
    {
      "epoch": 0.6,
      "grad_norm": 9.98999309539795,
      "learning_rate": 4.911940079374434e-06,
      "loss": 1.2111,
      "step": 1930
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.376320838928223,
      "learning_rate": 4.905400918936496e-06,
      "loss": 3.0751,
      "step": 1931
    },
    {
      "epoch": 0.6,
      "grad_norm": 15.502862930297852,
      "learning_rate": 4.898863791079717e-06,
      "loss": 1.6701,
      "step": 1932
    },
    {
      "epoch": 0.6,
      "grad_norm": 16.26798439025879,
      "learning_rate": 4.892328701999776e-06,
      "loss": 6.2586,
      "step": 1933
    },
    {
      "epoch": 0.6,
      "grad_norm": 17.54841423034668,
      "learning_rate": 4.885795657890418e-06,
      "loss": 2.0316,
      "step": 1934
    },
    {
      "epoch": 0.6,
      "grad_norm": 23.011951446533203,
      "learning_rate": 4.879264664943445e-06,
      "loss": 3.3383,
      "step": 1935
    },
    {
      "epoch": 0.6,
      "grad_norm": 14.206123352050781,
      "learning_rate": 4.8727357293487254e-06,
      "loss": 1.4116,
      "step": 1936
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.767412185668945,
      "learning_rate": 4.866208857294168e-06,
      "loss": 1.9756,
      "step": 1937
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.327007293701172,
      "learning_rate": 4.859684054965728e-06,
      "loss": 1.6458,
      "step": 1938
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.891633987426758,
      "learning_rate": 4.853161328547408e-06,
      "loss": 1.5201,
      "step": 1939
    },
    {
      "epoch": 0.6,
      "grad_norm": 25.075712203979492,
      "learning_rate": 4.846640684221227e-06,
      "loss": 4.8558,
      "step": 1940
    },
    {
      "epoch": 0.6,
      "grad_norm": 18.45266342163086,
      "learning_rate": 4.840122128167242e-06,
      "loss": 2.6664,
      "step": 1941
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.337541580200195,
      "learning_rate": 4.833605666563532e-06,
      "loss": 1.2926,
      "step": 1942
    },
    {
      "epoch": 0.6,
      "grad_norm": 11.597476959228516,
      "learning_rate": 4.827091305586184e-06,
      "loss": 1.9385,
      "step": 1943
    },
    {
      "epoch": 0.6,
      "grad_norm": 19.565479278564453,
      "learning_rate": 4.820579051409294e-06,
      "loss": 2.9348,
      "step": 1944
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.648226737976074,
      "learning_rate": 4.81406891020497e-06,
      "loss": 1.6047,
      "step": 1945
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.827557563781738,
      "learning_rate": 4.807560888143311e-06,
      "loss": 2.0912,
      "step": 1946
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.061546325683594,
      "learning_rate": 4.8010549913924045e-06,
      "loss": 1.3529,
      "step": 1947
    },
    {
      "epoch": 0.6,
      "grad_norm": 16.40416717529297,
      "learning_rate": 4.794551226118331e-06,
      "loss": 2.5774,
      "step": 1948
    },
    {
      "epoch": 0.6,
      "grad_norm": 16.391469955444336,
      "learning_rate": 4.788049598485146e-06,
      "loss": 2.5718,
      "step": 1949
    },
    {
      "epoch": 0.6,
      "grad_norm": 8.078376770019531,
      "learning_rate": 4.7815501146548795e-06,
      "loss": 1.0108,
      "step": 1950
    },
    {
      "epoch": 0.6,
      "grad_norm": 13.772455215454102,
      "learning_rate": 4.775052780787534e-06,
      "loss": 1.3883,
      "step": 1951
    },
    {
      "epoch": 0.6,
      "grad_norm": 12.186179161071777,
      "learning_rate": 4.7685576030410676e-06,
      "loss": 2.1278,
      "step": 1952
    },
    {
      "epoch": 0.61,
      "grad_norm": 18.81841278076172,
      "learning_rate": 4.762064587571399e-06,
      "loss": 2.2821,
      "step": 1953
    },
    {
      "epoch": 0.61,
      "grad_norm": 20.560026168823242,
      "learning_rate": 4.755573740532398e-06,
      "loss": 3.2618,
      "step": 1954
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.923328399658203,
      "learning_rate": 4.7490850680758734e-06,
      "loss": 5.6892,
      "step": 1955
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.167757987976074,
      "learning_rate": 4.742598576351587e-06,
      "loss": 0.7712,
      "step": 1956
    },
    {
      "epoch": 0.61,
      "grad_norm": 20.166093826293945,
      "learning_rate": 4.736114271507217e-06,
      "loss": 3.5002,
      "step": 1957
    },
    {
      "epoch": 0.61,
      "grad_norm": 16.612627029418945,
      "learning_rate": 4.729632159688375e-06,
      "loss": 8.2562,
      "step": 1958
    },
    {
      "epoch": 0.61,
      "grad_norm": 12.197168350219727,
      "learning_rate": 4.723152247038602e-06,
      "loss": 1.71,
      "step": 1959
    },
    {
      "epoch": 0.61,
      "grad_norm": 12.618730545043945,
      "learning_rate": 4.7166745396993444e-06,
      "loss": 0.7106,
      "step": 1960
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.799409866333008,
      "learning_rate": 4.71019904380996e-06,
      "loss": 1.5867,
      "step": 1961
    },
    {
      "epoch": 0.61,
      "grad_norm": 27.86197853088379,
      "learning_rate": 4.703725765507719e-06,
      "loss": 2.0034,
      "step": 1962
    },
    {
      "epoch": 0.61,
      "grad_norm": 18.016983032226562,
      "learning_rate": 4.697254710927778e-06,
      "loss": 3.8767,
      "step": 1963
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.817669868469238,
      "learning_rate": 4.690785886203195e-06,
      "loss": 1.6525,
      "step": 1964
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.399222373962402,
      "learning_rate": 4.68431929746491e-06,
      "loss": 2.8169,
      "step": 1965
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.781970024108887,
      "learning_rate": 4.677854950841746e-06,
      "loss": 2.5669,
      "step": 1966
    },
    {
      "epoch": 0.61,
      "grad_norm": 23.760482788085938,
      "learning_rate": 4.671392852460399e-06,
      "loss": 2.4262,
      "step": 1967
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.917473793029785,
      "learning_rate": 4.664933008445436e-06,
      "loss": 1.3121,
      "step": 1968
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.976125717163086,
      "learning_rate": 4.658475424919288e-06,
      "loss": 2.0666,
      "step": 1969
    },
    {
      "epoch": 0.61,
      "grad_norm": 20.271209716796875,
      "learning_rate": 4.65202010800224e-06,
      "loss": 4.8662,
      "step": 1970
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.959456443786621,
      "learning_rate": 4.645567063812433e-06,
      "loss": 1.4667,
      "step": 1971
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.600061416625977,
      "learning_rate": 4.639116298465852e-06,
      "loss": 1.6943,
      "step": 1972
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.672968864440918,
      "learning_rate": 4.632667818076319e-06,
      "loss": 2.446,
      "step": 1973
    },
    {
      "epoch": 0.61,
      "grad_norm": 10.582765579223633,
      "learning_rate": 4.6262216287555e-06,
      "loss": 0.8156,
      "step": 1974
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.40039348602295,
      "learning_rate": 4.619777736612878e-06,
      "loss": 1.8299,
      "step": 1975
    },
    {
      "epoch": 0.61,
      "grad_norm": 11.918106079101562,
      "learning_rate": 4.613336147755764e-06,
      "loss": 1.7873,
      "step": 1976
    },
    {
      "epoch": 0.61,
      "grad_norm": 9.903485298156738,
      "learning_rate": 4.6068968682892905e-06,
      "loss": 0.9619,
      "step": 1977
    },
    {
      "epoch": 0.61,
      "grad_norm": 15.035801887512207,
      "learning_rate": 4.600459904316396e-06,
      "loss": 1.7628,
      "step": 1978
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.6061429977417,
      "learning_rate": 4.594025261937818e-06,
      "loss": 2.7498,
      "step": 1979
    },
    {
      "epoch": 0.61,
      "grad_norm": 17.603635787963867,
      "learning_rate": 4.587592947252112e-06,
      "loss": 2.1893,
      "step": 1980
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.130147933959961,
      "learning_rate": 4.581162966355609e-06,
      "loss": 2.1399,
      "step": 1981
    },
    {
      "epoch": 0.61,
      "grad_norm": 13.292744636535645,
      "learning_rate": 4.5747353253424365e-06,
      "loss": 3.2051,
      "step": 1982
    },
    {
      "epoch": 0.61,
      "grad_norm": 14.927051544189453,
      "learning_rate": 4.568310030304507e-06,
      "loss": 5.1818,
      "step": 1983
    },
    {
      "epoch": 0.61,
      "grad_norm": 8.793292999267578,
      "learning_rate": 4.561887087331501e-06,
      "loss": 1.0002,
      "step": 1984
    },
    {
      "epoch": 0.62,
      "grad_norm": 16.968029022216797,
      "learning_rate": 4.555466502510876e-06,
      "loss": 4.2384,
      "step": 1985
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.477005004882812,
      "learning_rate": 4.549048281927855e-06,
      "loss": 1.1315,
      "step": 1986
    },
    {
      "epoch": 0.62,
      "grad_norm": 18.990814208984375,
      "learning_rate": 4.542632431665419e-06,
      "loss": 3.8732,
      "step": 1987
    },
    {
      "epoch": 0.62,
      "grad_norm": 12.12499713897705,
      "learning_rate": 4.536218957804297e-06,
      "loss": 2.9292,
      "step": 1988
    },
    {
      "epoch": 0.62,
      "grad_norm": 9.619987487792969,
      "learning_rate": 4.529807866422976e-06,
      "loss": 1.3448,
      "step": 1989
    },
    {
      "epoch": 0.62,
      "grad_norm": 12.029317855834961,
      "learning_rate": 4.523399163597677e-06,
      "loss": 1.5169,
      "step": 1990
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.02697467803955,
      "learning_rate": 4.5169928554023594e-06,
      "loss": 1.5848,
      "step": 1991
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.536266326904297,
      "learning_rate": 4.510588947908717e-06,
      "loss": 1.9503,
      "step": 1992
    },
    {
      "epoch": 0.62,
      "grad_norm": 8.811053276062012,
      "learning_rate": 4.504187447186161e-06,
      "loss": 1.652,
      "step": 1993
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.713563919067383,
      "learning_rate": 4.497788359301823e-06,
      "loss": 3.1022,
      "step": 1994
    },
    {
      "epoch": 0.62,
      "grad_norm": 18.214853286743164,
      "learning_rate": 4.491391690320558e-06,
      "loss": 2.3007,
      "step": 1995
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.674553871154785,
      "learning_rate": 4.484997446304915e-06,
      "loss": 1.1045,
      "step": 1996
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.581334114074707,
      "learning_rate": 4.478605633315148e-06,
      "loss": 1.129,
      "step": 1997
    },
    {
      "epoch": 0.62,
      "grad_norm": 23.164079666137695,
      "learning_rate": 4.472216257409216e-06,
      "loss": 5.2292,
      "step": 1998
    },
    {
      "epoch": 0.62,
      "grad_norm": 16.813941955566406,
      "learning_rate": 4.465829324642759e-06,
      "loss": 2.1862,
      "step": 1999
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.300929069519043,
      "learning_rate": 4.459444841069098e-06,
      "loss": 1.0576,
      "step": 2000
    },
    {
      "epoch": 0.62,
      "grad_norm": 17.166240692138672,
      "learning_rate": 4.4530628127392454e-06,
      "loss": 1.7541,
      "step": 2001
    },
    {
      "epoch": 0.62,
      "grad_norm": 15.232935905456543,
      "learning_rate": 4.44668324570188e-06,
      "loss": 1.8813,
      "step": 2002
    },
    {
      "epoch": 0.62,
      "grad_norm": 17.39579200744629,
      "learning_rate": 4.440306146003342e-06,
      "loss": 3.1383,
      "step": 2003
    },
    {
      "epoch": 0.62,
      "grad_norm": 16.826940536499023,
      "learning_rate": 4.433931519687644e-06,
      "loss": 2.4366,
      "step": 2004
    },
    {
      "epoch": 0.62,
      "grad_norm": 13.890028953552246,
      "learning_rate": 4.427559372796447e-06,
      "loss": 2.2855,
      "step": 2005
    },
    {
      "epoch": 0.62,
      "grad_norm": 22.110166549682617,
      "learning_rate": 4.421189711369063e-06,
      "loss": 2.8979,
      "step": 2006
    },
    {
      "epoch": 0.62,
      "grad_norm": 15.890730857849121,
      "learning_rate": 4.414822541442454e-06,
      "loss": 2.046,
      "step": 2007
    },
    {
      "epoch": 0.62,
      "grad_norm": 15.0035400390625,
      "learning_rate": 4.408457869051212e-06,
      "loss": 5.1188,
      "step": 2008
    },
    {
      "epoch": 0.62,
      "grad_norm": 24.372314453125,
      "learning_rate": 4.402095700227567e-06,
      "loss": 2.7475,
      "step": 2009
    },
    {
      "epoch": 0.62,
      "grad_norm": 19.236316680908203,
      "learning_rate": 4.395736041001377e-06,
      "loss": 4.2975,
      "step": 2010
    },
    {
      "epoch": 0.62,
      "grad_norm": 14.21186637878418,
      "learning_rate": 4.3893788974001204e-06,
      "loss": 1.0334,
      "step": 2011
    },
    {
      "epoch": 0.62,
      "grad_norm": 14.599188804626465,
      "learning_rate": 4.383024275448886e-06,
      "loss": 1.8703,
      "step": 2012
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.629423141479492,
      "learning_rate": 4.376672181170383e-06,
      "loss": 2.0404,
      "step": 2013
    },
    {
      "epoch": 0.62,
      "grad_norm": 14.760478019714355,
      "learning_rate": 4.370322620584918e-06,
      "loss": 1.4733,
      "step": 2014
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.846567153930664,
      "learning_rate": 4.363975599710395e-06,
      "loss": 1.4684,
      "step": 2015
    },
    {
      "epoch": 0.62,
      "grad_norm": 10.186484336853027,
      "learning_rate": 4.357631124562318e-06,
      "loss": 1.3573,
      "step": 2016
    },
    {
      "epoch": 0.62,
      "grad_norm": 11.071416854858398,
      "learning_rate": 4.351289201153772e-06,
      "loss": 0.8062,
      "step": 2017
    },
    {
      "epoch": 0.63,
      "grad_norm": 14.283429145812988,
      "learning_rate": 4.344949835495421e-06,
      "loss": 1.7092,
      "step": 2018
    },
    {
      "epoch": 0.63,
      "grad_norm": 27.083303451538086,
      "learning_rate": 4.338613033595518e-06,
      "loss": 2.3482,
      "step": 2019
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.069631576538086,
      "learning_rate": 4.332278801459873e-06,
      "loss": 3.7772,
      "step": 2020
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.28091049194336,
      "learning_rate": 4.325947145091861e-06,
      "loss": 4.7075,
      "step": 2021
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.421253204345703,
      "learning_rate": 4.3196180704924275e-06,
      "loss": 2.554,
      "step": 2022
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.664969444274902,
      "learning_rate": 4.313291583660058e-06,
      "loss": 1.8915,
      "step": 2023
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.2357177734375,
      "learning_rate": 4.306967690590791e-06,
      "loss": 3.0887,
      "step": 2024
    },
    {
      "epoch": 0.63,
      "grad_norm": 9.014602661132812,
      "learning_rate": 4.300646397278208e-06,
      "loss": 1.2436,
      "step": 2025
    },
    {
      "epoch": 0.63,
      "grad_norm": 15.402137756347656,
      "learning_rate": 4.294327709713425e-06,
      "loss": 4.4077,
      "step": 2026
    },
    {
      "epoch": 0.63,
      "grad_norm": 19.77958869934082,
      "learning_rate": 4.2880116338850855e-06,
      "loss": 3.137,
      "step": 2027
    },
    {
      "epoch": 0.63,
      "grad_norm": 19.124786376953125,
      "learning_rate": 4.281698175779366e-06,
      "loss": 3.9406,
      "step": 2028
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.08989143371582,
      "learning_rate": 4.2753873413799505e-06,
      "loss": 1.8425,
      "step": 2029
    },
    {
      "epoch": 0.63,
      "grad_norm": 15.249897003173828,
      "learning_rate": 4.269079136668045e-06,
      "loss": 2.791,
      "step": 2030
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.198994636535645,
      "learning_rate": 4.2627735676223604e-06,
      "loss": 1.4869,
      "step": 2031
    },
    {
      "epoch": 0.63,
      "grad_norm": 14.075993537902832,
      "learning_rate": 4.25647064021911e-06,
      "loss": 1.2025,
      "step": 2032
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.997515678405762,
      "learning_rate": 4.250170360432e-06,
      "loss": 1.106,
      "step": 2033
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.242206573486328,
      "learning_rate": 4.2438727342322375e-06,
      "loss": 1.3455,
      "step": 2034
    },
    {
      "epoch": 0.63,
      "grad_norm": 15.24692153930664,
      "learning_rate": 4.237577767588501e-06,
      "loss": 1.9019,
      "step": 2035
    },
    {
      "epoch": 0.63,
      "grad_norm": 14.23598861694336,
      "learning_rate": 4.231285466466954e-06,
      "loss": 3.0527,
      "step": 2036
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.469017028808594,
      "learning_rate": 4.224995836831241e-06,
      "loss": 1.6078,
      "step": 2037
    },
    {
      "epoch": 0.63,
      "grad_norm": 19.05677032470703,
      "learning_rate": 4.218708884642465e-06,
      "loss": 3.422,
      "step": 2038
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.673805236816406,
      "learning_rate": 4.21242461585919e-06,
      "loss": 3.2013,
      "step": 2039
    },
    {
      "epoch": 0.63,
      "grad_norm": 17.894207000732422,
      "learning_rate": 4.206143036437449e-06,
      "loss": 6.1287,
      "step": 2040
    },
    {
      "epoch": 0.63,
      "grad_norm": 13.790003776550293,
      "learning_rate": 4.199864152330714e-06,
      "loss": 2.0449,
      "step": 2041
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.329513549804688,
      "learning_rate": 4.193587969489906e-06,
      "loss": 1.8032,
      "step": 2042
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.738219261169434,
      "learning_rate": 4.1873144938633905e-06,
      "loss": 0.8659,
      "step": 2043
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.105698585510254,
      "learning_rate": 4.18104373139696e-06,
      "loss": 2.2742,
      "step": 2044
    },
    {
      "epoch": 0.63,
      "grad_norm": 10.643797874450684,
      "learning_rate": 4.17477568803384e-06,
      "loss": 1.5166,
      "step": 2045
    },
    {
      "epoch": 0.63,
      "grad_norm": 12.171561241149902,
      "learning_rate": 4.16851036971468e-06,
      "loss": 2.284,
      "step": 2046
    },
    {
      "epoch": 0.63,
      "grad_norm": 11.268860816955566,
      "learning_rate": 4.16224778237754e-06,
      "loss": 4.9508,
      "step": 2047
    },
    {
      "epoch": 0.63,
      "grad_norm": 14.62899398803711,
      "learning_rate": 4.1559879319578994e-06,
      "loss": 5.0698,
      "step": 2048
    },
    {
      "epoch": 0.63,
      "grad_norm": 15.234814643859863,
      "learning_rate": 4.14973082438864e-06,
      "loss": 3.7426,
      "step": 2049
    },
    {
      "epoch": 0.64,
      "grad_norm": 23.05950355529785,
      "learning_rate": 4.1434764656000465e-06,
      "loss": 3.1389,
      "step": 2050
    },
    {
      "epoch": 0.64,
      "grad_norm": 14.896961212158203,
      "learning_rate": 4.137224861519792e-06,
      "loss": 1.9389,
      "step": 2051
    },
    {
      "epoch": 0.64,
      "grad_norm": 16.13572120666504,
      "learning_rate": 4.1309760180729464e-06,
      "loss": 1.6235,
      "step": 2052
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.377883434295654,
      "learning_rate": 4.124729941181961e-06,
      "loss": 0.8666,
      "step": 2053
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.743086814880371,
      "learning_rate": 4.118486636766659e-06,
      "loss": 1.2833,
      "step": 2054
    },
    {
      "epoch": 0.64,
      "grad_norm": 8.719217300415039,
      "learning_rate": 4.112246110744247e-06,
      "loss": 1.0122,
      "step": 2055
    },
    {
      "epoch": 0.64,
      "grad_norm": 11.55544376373291,
      "learning_rate": 4.10600836902929e-06,
      "loss": 2.1538,
      "step": 2056
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.313085556030273,
      "learning_rate": 4.099773417533717e-06,
      "loss": 4.89,
      "step": 2057
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.50877571105957,
      "learning_rate": 4.093541262166809e-06,
      "loss": 2.432,
      "step": 2058
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.130254745483398,
      "learning_rate": 4.0873119088352035e-06,
      "loss": 5.1446,
      "step": 2059
    },
    {
      "epoch": 0.64,
      "grad_norm": 19.7545108795166,
      "learning_rate": 4.081085363442874e-06,
      "loss": 4.3046,
      "step": 2060
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.939685821533203,
      "learning_rate": 4.074861631891144e-06,
      "loss": 3.2587,
      "step": 2061
    },
    {
      "epoch": 0.64,
      "grad_norm": 7.611458778381348,
      "learning_rate": 4.0686407200786595e-06,
      "loss": 0.8931,
      "step": 2062
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.956474304199219,
      "learning_rate": 4.062422633901396e-06,
      "loss": 2.0707,
      "step": 2063
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.552716255187988,
      "learning_rate": 4.0562073792526615e-06,
      "loss": 1.3171,
      "step": 2064
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.638819694519043,
      "learning_rate": 4.049994962023063e-06,
      "loss": 1.3762,
      "step": 2065
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.033987998962402,
      "learning_rate": 4.0437853881005285e-06,
      "loss": 3.3857,
      "step": 2066
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.575906753540039,
      "learning_rate": 4.037578663370295e-06,
      "loss": 1.4654,
      "step": 2067
    },
    {
      "epoch": 0.64,
      "grad_norm": 15.086124420166016,
      "learning_rate": 4.031374793714891e-06,
      "loss": 2.941,
      "step": 2068
    },
    {
      "epoch": 0.64,
      "grad_norm": 14.043783187866211,
      "learning_rate": 4.025173785014139e-06,
      "loss": 2.1311,
      "step": 2069
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.548436164855957,
      "learning_rate": 4.0189756431451575e-06,
      "loss": 2.8225,
      "step": 2070
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.42125701904297,
      "learning_rate": 4.012780373982344e-06,
      "loss": 0.8218,
      "step": 2071
    },
    {
      "epoch": 0.64,
      "grad_norm": 14.614411354064941,
      "learning_rate": 4.0065879833973694e-06,
      "loss": 2.4332,
      "step": 2072
    },
    {
      "epoch": 0.64,
      "grad_norm": 17.507421493530273,
      "learning_rate": 4.000398477259182e-06,
      "loss": 3.5555,
      "step": 2073
    },
    {
      "epoch": 0.64,
      "grad_norm": 10.591798782348633,
      "learning_rate": 3.994211861433995e-06,
      "loss": 1.6596,
      "step": 2074
    },
    {
      "epoch": 0.64,
      "grad_norm": 18.85674285888672,
      "learning_rate": 3.988028141785275e-06,
      "loss": 3.2954,
      "step": 2075
    },
    {
      "epoch": 0.64,
      "grad_norm": 12.868513107299805,
      "learning_rate": 3.9818473241737585e-06,
      "loss": 1.5311,
      "step": 2076
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.88658332824707,
      "learning_rate": 3.97566941445742e-06,
      "loss": 2.2858,
      "step": 2077
    },
    {
      "epoch": 0.64,
      "grad_norm": 9.721027374267578,
      "learning_rate": 3.9694944184914804e-06,
      "loss": 1.3336,
      "step": 2078
    },
    {
      "epoch": 0.64,
      "grad_norm": 13.591902732849121,
      "learning_rate": 3.9633223421284e-06,
      "loss": 1.4743,
      "step": 2079
    },
    {
      "epoch": 0.64,
      "grad_norm": 21.40350341796875,
      "learning_rate": 3.957153191217872e-06,
      "loss": 2.8149,
      "step": 2080
    },
    {
      "epoch": 0.64,
      "grad_norm": 20.24881935119629,
      "learning_rate": 3.950986971606814e-06,
      "loss": 5.8637,
      "step": 2081
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.11694049835205,
      "learning_rate": 3.944823689139373e-06,
      "loss": 1.6955,
      "step": 2082
    },
    {
      "epoch": 0.65,
      "grad_norm": 19.558984756469727,
      "learning_rate": 3.938663349656905e-06,
      "loss": 2.7874,
      "step": 2083
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.004746437072754,
      "learning_rate": 3.932505958997977e-06,
      "loss": 1.889,
      "step": 2084
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.026920318603516,
      "learning_rate": 3.926351522998371e-06,
      "loss": 0.8632,
      "step": 2085
    },
    {
      "epoch": 0.65,
      "grad_norm": 14.440336227416992,
      "learning_rate": 3.920200047491051e-06,
      "loss": 2.5162,
      "step": 2086
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.653973579406738,
      "learning_rate": 3.9140515383061885e-06,
      "loss": 1.6231,
      "step": 2087
    },
    {
      "epoch": 0.65,
      "grad_norm": 16.52773094177246,
      "learning_rate": 3.907906001271145e-06,
      "loss": 2.3722,
      "step": 2088
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.918020248413086,
      "learning_rate": 3.9017634422104545e-06,
      "loss": 3.0244,
      "step": 2089
    },
    {
      "epoch": 0.65,
      "grad_norm": 15.039247512817383,
      "learning_rate": 3.895623866945835e-06,
      "loss": 2.6583,
      "step": 2090
    },
    {
      "epoch": 0.65,
      "grad_norm": 14.738115310668945,
      "learning_rate": 3.88948728129618e-06,
      "loss": 2.2832,
      "step": 2091
    },
    {
      "epoch": 0.65,
      "grad_norm": 15.432753562927246,
      "learning_rate": 3.883353691077543e-06,
      "loss": 4.0483,
      "step": 2092
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.476043701171875,
      "learning_rate": 3.8772231021031416e-06,
      "loss": 1.7617,
      "step": 2093
    },
    {
      "epoch": 0.65,
      "grad_norm": 15.05960750579834,
      "learning_rate": 3.871095520183348e-06,
      "loss": 1.767,
      "step": 2094
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.401081085205078,
      "learning_rate": 3.864970951125685e-06,
      "loss": 1.6661,
      "step": 2095
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.15285587310791,
      "learning_rate": 3.858849400734815e-06,
      "loss": 0.8678,
      "step": 2096
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.537773132324219,
      "learning_rate": 3.852730874812552e-06,
      "loss": 3.9522,
      "step": 2097
    },
    {
      "epoch": 0.65,
      "grad_norm": 8.932913780212402,
      "learning_rate": 3.846615379157833e-06,
      "loss": 0.8396,
      "step": 2098
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.99303913116455,
      "learning_rate": 3.8405029195667235e-06,
      "loss": 1.286,
      "step": 2099
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.972646713256836,
      "learning_rate": 3.834393501832415e-06,
      "loss": 1.9852,
      "step": 2100
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.994357109069824,
      "learning_rate": 3.828287131745215e-06,
      "loss": 1.8358,
      "step": 2101
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.910358428955078,
      "learning_rate": 3.822183815092538e-06,
      "loss": 1.6349,
      "step": 2102
    },
    {
      "epoch": 0.65,
      "grad_norm": 17.40057373046875,
      "learning_rate": 3.8160835576589156e-06,
      "loss": 5.6253,
      "step": 2103
    },
    {
      "epoch": 0.65,
      "grad_norm": 17.70642852783203,
      "learning_rate": 3.8099863652259697e-06,
      "loss": 2.2104,
      "step": 2104
    },
    {
      "epoch": 0.65,
      "grad_norm": 14.252694129943848,
      "learning_rate": 3.8038922435724186e-06,
      "loss": 2.2713,
      "step": 2105
    },
    {
      "epoch": 0.65,
      "grad_norm": 18.08612632751465,
      "learning_rate": 3.7978011984740795e-06,
      "loss": 4.3866,
      "step": 2106
    },
    {
      "epoch": 0.65,
      "grad_norm": 13.621631622314453,
      "learning_rate": 3.791713235703839e-06,
      "loss": 2.4693,
      "step": 2107
    },
    {
      "epoch": 0.65,
      "grad_norm": 11.246854782104492,
      "learning_rate": 3.78562836103167e-06,
      "loss": 1.8537,
      "step": 2108
    },
    {
      "epoch": 0.65,
      "grad_norm": 10.283203125,
      "learning_rate": 3.7795465802246238e-06,
      "loss": 0.9031,
      "step": 2109
    },
    {
      "epoch": 0.65,
      "grad_norm": 18.088642120361328,
      "learning_rate": 3.77346789904681e-06,
      "loss": 1.7515,
      "step": 2110
    },
    {
      "epoch": 0.65,
      "grad_norm": 9.636170387268066,
      "learning_rate": 3.7673923232594055e-06,
      "loss": 0.918,
      "step": 2111
    },
    {
      "epoch": 0.65,
      "grad_norm": 18.71010971069336,
      "learning_rate": 3.7613198586206453e-06,
      "loss": 4.5926,
      "step": 2112
    },
    {
      "epoch": 0.65,
      "grad_norm": 14.63417911529541,
      "learning_rate": 3.7552505108858128e-06,
      "loss": 2.1638,
      "step": 2113
    },
    {
      "epoch": 0.65,
      "grad_norm": 12.265777587890625,
      "learning_rate": 3.7491842858072365e-06,
      "loss": 1.8688,
      "step": 2114
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.674768447875977,
      "learning_rate": 3.7431211891342885e-06,
      "loss": 2.0126,
      "step": 2115
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.914712905883789,
      "learning_rate": 3.7370612266133733e-06,
      "loss": 2.3795,
      "step": 2116
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.552215576171875,
      "learning_rate": 3.731004403987924e-06,
      "loss": 1.2995,
      "step": 2117
    },
    {
      "epoch": 0.66,
      "grad_norm": 15.453949928283691,
      "learning_rate": 3.724950726998404e-06,
      "loss": 3.2372,
      "step": 2118
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.629971504211426,
      "learning_rate": 3.7189002013822904e-06,
      "loss": 2.3727,
      "step": 2119
    },
    {
      "epoch": 0.66,
      "grad_norm": 15.596941947937012,
      "learning_rate": 3.7128528328740733e-06,
      "loss": 3.0927,
      "step": 2120
    },
    {
      "epoch": 0.66,
      "grad_norm": 17.61478042602539,
      "learning_rate": 3.7068086272052516e-06,
      "loss": 3.1227,
      "step": 2121
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.685725212097168,
      "learning_rate": 3.700767590104325e-06,
      "loss": 0.7904,
      "step": 2122
    },
    {
      "epoch": 0.66,
      "grad_norm": 13.112590789794922,
      "learning_rate": 3.6947297272967926e-06,
      "loss": 2.4491,
      "step": 2123
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.789117813110352,
      "learning_rate": 3.6886950445051465e-06,
      "loss": 1.0209,
      "step": 2124
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.015466690063477,
      "learning_rate": 3.6826635474488588e-06,
      "loss": 2.0642,
      "step": 2125
    },
    {
      "epoch": 0.66,
      "grad_norm": 18.760639190673828,
      "learning_rate": 3.6766352418443893e-06,
      "loss": 1.7846,
      "step": 2126
    },
    {
      "epoch": 0.66,
      "grad_norm": 18.239532470703125,
      "learning_rate": 3.670610133405172e-06,
      "loss": 5.6831,
      "step": 2127
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.637475967407227,
      "learning_rate": 3.6645882278416003e-06,
      "loss": 1.2087,
      "step": 2128
    },
    {
      "epoch": 0.66,
      "grad_norm": 13.088603973388672,
      "learning_rate": 3.6585695308610465e-06,
      "loss": 1.7061,
      "step": 2129
    },
    {
      "epoch": 0.66,
      "grad_norm": 17.58694076538086,
      "learning_rate": 3.652554048167834e-06,
      "loss": 5.1298,
      "step": 2130
    },
    {
      "epoch": 0.66,
      "grad_norm": 13.216436386108398,
      "learning_rate": 3.6465417854632377e-06,
      "loss": 1.4959,
      "step": 2131
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.837897300720215,
      "learning_rate": 3.640532748445491e-06,
      "loss": 1.5755,
      "step": 2132
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.177626609802246,
      "learning_rate": 3.6345269428097607e-06,
      "loss": 2.7501,
      "step": 2133
    },
    {
      "epoch": 0.66,
      "grad_norm": 16.52323341369629,
      "learning_rate": 3.6285243742481533e-06,
      "loss": 3.8536,
      "step": 2134
    },
    {
      "epoch": 0.66,
      "grad_norm": 13.3431396484375,
      "learning_rate": 3.622525048449708e-06,
      "loss": 2.1923,
      "step": 2135
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.451409339904785,
      "learning_rate": 3.6165289711003897e-06,
      "loss": 1.0444,
      "step": 2136
    },
    {
      "epoch": 0.66,
      "grad_norm": 9.489715576171875,
      "learning_rate": 3.610536147883083e-06,
      "loss": 1.2464,
      "step": 2137
    },
    {
      "epoch": 0.66,
      "grad_norm": 10.092480659484863,
      "learning_rate": 3.604546584477596e-06,
      "loss": 2.2864,
      "step": 2138
    },
    {
      "epoch": 0.66,
      "grad_norm": 21.5133056640625,
      "learning_rate": 3.598560286560639e-06,
      "loss": 2.9101,
      "step": 2139
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.151141166687012,
      "learning_rate": 3.592577259805829e-06,
      "loss": 1.4456,
      "step": 2140
    },
    {
      "epoch": 0.66,
      "grad_norm": 12.522343635559082,
      "learning_rate": 3.58659750988369e-06,
      "loss": 3.3415,
      "step": 2141
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.61464786529541,
      "learning_rate": 3.5806210424616257e-06,
      "loss": 2.2732,
      "step": 2142
    },
    {
      "epoch": 0.66,
      "grad_norm": 14.204218864440918,
      "learning_rate": 3.574647863203939e-06,
      "loss": 4.3799,
      "step": 2143
    },
    {
      "epoch": 0.66,
      "grad_norm": 11.515316009521484,
      "learning_rate": 3.5686779777718193e-06,
      "loss": 2.4421,
      "step": 2144
    },
    {
      "epoch": 0.66,
      "grad_norm": 15.56559944152832,
      "learning_rate": 3.5627113918233276e-06,
      "loss": 3.1241,
      "step": 2145
    },
    {
      "epoch": 0.66,
      "grad_norm": 18.81997299194336,
      "learning_rate": 3.5567481110133953e-06,
      "loss": 4.0341,
      "step": 2146
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.367809295654297,
      "learning_rate": 3.5507881409938328e-06,
      "loss": 1.2594,
      "step": 2147
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.987909317016602,
      "learning_rate": 3.5448314874133033e-06,
      "loss": 1.0973,
      "step": 2148
    },
    {
      "epoch": 0.67,
      "grad_norm": 18.935338973999023,
      "learning_rate": 3.5388781559173267e-06,
      "loss": 5.2275,
      "step": 2149
    },
    {
      "epoch": 0.67,
      "grad_norm": 13.536104202270508,
      "learning_rate": 3.5329281521482783e-06,
      "loss": 1.608,
      "step": 2150
    },
    {
      "epoch": 0.67,
      "grad_norm": 20.221797943115234,
      "learning_rate": 3.526981481745377e-06,
      "loss": 2.6814,
      "step": 2151
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.287632942199707,
      "learning_rate": 3.5210381503446822e-06,
      "loss": 1.271,
      "step": 2152
    },
    {
      "epoch": 0.67,
      "grad_norm": 11.780200004577637,
      "learning_rate": 3.5150981635790942e-06,
      "loss": 1.9295,
      "step": 2153
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.957992553710938,
      "learning_rate": 3.509161527078336e-06,
      "loss": 1.063,
      "step": 2154
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.497386932373047,
      "learning_rate": 3.503228246468959e-06,
      "loss": 5.2318,
      "step": 2155
    },
    {
      "epoch": 0.67,
      "grad_norm": 15.921914100646973,
      "learning_rate": 3.4972983273743326e-06,
      "loss": 5.0767,
      "step": 2156
    },
    {
      "epoch": 0.67,
      "grad_norm": 17.922636032104492,
      "learning_rate": 3.4913717754146416e-06,
      "loss": 1.8815,
      "step": 2157
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.503357887268066,
      "learning_rate": 3.4854485962068766e-06,
      "loss": 1.9732,
      "step": 2158
    },
    {
      "epoch": 0.67,
      "grad_norm": 19.81252098083496,
      "learning_rate": 3.4795287953648377e-06,
      "loss": 3.1647,
      "step": 2159
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.471256256103516,
      "learning_rate": 3.4736123784991183e-06,
      "loss": 1.1755,
      "step": 2160
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.54334545135498,
      "learning_rate": 3.4676993512171023e-06,
      "loss": 0.9531,
      "step": 2161
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.592487335205078,
      "learning_rate": 3.4617897191229724e-06,
      "loss": 2.5702,
      "step": 2162
    },
    {
      "epoch": 0.67,
      "grad_norm": 16.094675064086914,
      "learning_rate": 3.455883487817677e-06,
      "loss": 3.1361,
      "step": 2163
    },
    {
      "epoch": 0.67,
      "grad_norm": 13.244039535522461,
      "learning_rate": 3.449980662898951e-06,
      "loss": 1.4257,
      "step": 2164
    },
    {
      "epoch": 0.67,
      "grad_norm": 10.110610008239746,
      "learning_rate": 3.4440812499613035e-06,
      "loss": 1.3565,
      "step": 2165
    },
    {
      "epoch": 0.67,
      "grad_norm": 15.21812629699707,
      "learning_rate": 3.4381852545960046e-06,
      "loss": 3.301,
      "step": 2166
    },
    {
      "epoch": 0.67,
      "grad_norm": 13.889304161071777,
      "learning_rate": 3.4322926823910824e-06,
      "loss": 2.4784,
      "step": 2167
    },
    {
      "epoch": 0.67,
      "grad_norm": 19.99413299560547,
      "learning_rate": 3.4264035389313316e-06,
      "loss": 2.0284,
      "step": 2168
    },
    {
      "epoch": 0.67,
      "grad_norm": 14.55456256866455,
      "learning_rate": 3.4205178297982878e-06,
      "loss": 2.0917,
      "step": 2169
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.45078182220459,
      "learning_rate": 3.4146355605702333e-06,
      "loss": 1.7994,
      "step": 2170
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.798308372497559,
      "learning_rate": 3.4087567368221925e-06,
      "loss": 1.5656,
      "step": 2171
    },
    {
      "epoch": 0.67,
      "grad_norm": 20.649717330932617,
      "learning_rate": 3.402881364125923e-06,
      "loss": 3.8393,
      "step": 2172
    },
    {
      "epoch": 0.67,
      "grad_norm": 9.314234733581543,
      "learning_rate": 3.3970094480499083e-06,
      "loss": 2.6209,
      "step": 2173
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.697712898254395,
      "learning_rate": 3.3911409941593645e-06,
      "loss": 4.9514,
      "step": 2174
    },
    {
      "epoch": 0.67,
      "grad_norm": 18.51091194152832,
      "learning_rate": 3.3852760080162196e-06,
      "loss": 3.1904,
      "step": 2175
    },
    {
      "epoch": 0.67,
      "grad_norm": 21.123741149902344,
      "learning_rate": 3.379414495179115e-06,
      "loss": 8.2013,
      "step": 2176
    },
    {
      "epoch": 0.67,
      "grad_norm": 12.869108200073242,
      "learning_rate": 3.3735564612034024e-06,
      "loss": 1.6296,
      "step": 2177
    },
    {
      "epoch": 0.67,
      "grad_norm": 8.646051406860352,
      "learning_rate": 3.3677019116411356e-06,
      "loss": 0.834,
      "step": 2178
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.026458740234375,
      "learning_rate": 3.361850852041064e-06,
      "loss": 2.2227,
      "step": 2179
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.233617782592773,
      "learning_rate": 3.3560032879486353e-06,
      "loss": 2.5529,
      "step": 2180
    },
    {
      "epoch": 0.68,
      "grad_norm": 18.442110061645508,
      "learning_rate": 3.3501592249059784e-06,
      "loss": 3.5626,
      "step": 2181
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.84872817993164,
      "learning_rate": 3.3443186684519025e-06,
      "loss": 4.7617,
      "step": 2182
    },
    {
      "epoch": 0.68,
      "grad_norm": 15.51069450378418,
      "learning_rate": 3.338481624121906e-06,
      "loss": 6.0042,
      "step": 2183
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.388606071472168,
      "learning_rate": 3.33264809744814e-06,
      "loss": 2.0355,
      "step": 2184
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.3096342086792,
      "learning_rate": 3.326818093959433e-06,
      "loss": 1.449,
      "step": 2185
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.437499046325684,
      "learning_rate": 3.3209916191812763e-06,
      "loss": 4.6765,
      "step": 2186
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.381603240966797,
      "learning_rate": 3.31516867863581e-06,
      "loss": 2.78,
      "step": 2187
    },
    {
      "epoch": 0.68,
      "grad_norm": 10.969701766967773,
      "learning_rate": 3.309349277841825e-06,
      "loss": 1.3352,
      "step": 2188
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.634147644042969,
      "learning_rate": 3.3035334223147673e-06,
      "loss": 1.3593,
      "step": 2189
    },
    {
      "epoch": 0.68,
      "grad_norm": 13.996097564697266,
      "learning_rate": 3.297721117566709e-06,
      "loss": 2.0998,
      "step": 2190
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.386398315429688,
      "learning_rate": 3.2919123691063667e-06,
      "loss": 1.8084,
      "step": 2191
    },
    {
      "epoch": 0.68,
      "grad_norm": 15.382070541381836,
      "learning_rate": 3.286107182439081e-06,
      "loss": 5.2602,
      "step": 2192
    },
    {
      "epoch": 0.68,
      "grad_norm": 21.93962860107422,
      "learning_rate": 3.280305563066819e-06,
      "loss": 2.948,
      "step": 2193
    },
    {
      "epoch": 0.68,
      "grad_norm": 18.650524139404297,
      "learning_rate": 3.2745075164881645e-06,
      "loss": 4.9346,
      "step": 2194
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.64752960205078,
      "learning_rate": 3.2687130481983222e-06,
      "loss": 2.3879,
      "step": 2195
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.175460815429688,
      "learning_rate": 3.2629221636890975e-06,
      "loss": 2.6153,
      "step": 2196
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.955414772033691,
      "learning_rate": 3.257134868448903e-06,
      "loss": 1.0509,
      "step": 2197
    },
    {
      "epoch": 0.68,
      "grad_norm": 16.26597785949707,
      "learning_rate": 3.2513511679627473e-06,
      "loss": 2.6854,
      "step": 2198
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.234977722167969,
      "learning_rate": 3.245571067712234e-06,
      "loss": 1.5645,
      "step": 2199
    },
    {
      "epoch": 0.68,
      "grad_norm": 8.558510780334473,
      "learning_rate": 3.239794573175552e-06,
      "loss": 0.9032,
      "step": 2200
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.81620979309082,
      "learning_rate": 3.234021689827479e-06,
      "loss": 3.2927,
      "step": 2201
    },
    {
      "epoch": 0.68,
      "grad_norm": 13.267828941345215,
      "learning_rate": 3.2282524231393627e-06,
      "loss": 2.4144,
      "step": 2202
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.47176456451416,
      "learning_rate": 3.2224867785791255e-06,
      "loss": 6.066,
      "step": 2203
    },
    {
      "epoch": 0.68,
      "grad_norm": 13.159913063049316,
      "learning_rate": 3.2167247616112605e-06,
      "loss": 2.0575,
      "step": 2204
    },
    {
      "epoch": 0.68,
      "grad_norm": 17.272306442260742,
      "learning_rate": 3.2109663776968216e-06,
      "loss": 5.412,
      "step": 2205
    },
    {
      "epoch": 0.68,
      "grad_norm": 20.280254364013672,
      "learning_rate": 3.2052116322934085e-06,
      "loss": 2.2086,
      "step": 2206
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.724082946777344,
      "learning_rate": 3.1994605308551892e-06,
      "loss": 3.3149,
      "step": 2207
    },
    {
      "epoch": 0.68,
      "grad_norm": 14.12836742401123,
      "learning_rate": 3.193713078832869e-06,
      "loss": 1.3179,
      "step": 2208
    },
    {
      "epoch": 0.68,
      "grad_norm": 11.377018928527832,
      "learning_rate": 3.187969281673692e-06,
      "loss": 2.8663,
      "step": 2209
    },
    {
      "epoch": 0.68,
      "grad_norm": 12.249744415283203,
      "learning_rate": 3.1822291448214475e-06,
      "loss": 1.7564,
      "step": 2210
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.492171287536621,
      "learning_rate": 3.1764926737164476e-06,
      "loss": 1.8586,
      "step": 2211
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.95061206817627,
      "learning_rate": 3.1707598737955327e-06,
      "loss": 0.9182,
      "step": 2212
    },
    {
      "epoch": 0.69,
      "grad_norm": 16.02012062072754,
      "learning_rate": 3.1650307504920644e-06,
      "loss": 1.7154,
      "step": 2213
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.300397872924805,
      "learning_rate": 3.1593053092359195e-06,
      "loss": 2.4413,
      "step": 2214
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.821404457092285,
      "learning_rate": 3.1535835554534806e-06,
      "loss": 1.5034,
      "step": 2215
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.430465698242188,
      "learning_rate": 3.1478654945676463e-06,
      "loss": 2.1052,
      "step": 2216
    },
    {
      "epoch": 0.69,
      "grad_norm": 16.182870864868164,
      "learning_rate": 3.142151131997804e-06,
      "loss": 2.2128,
      "step": 2217
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.833673477172852,
      "learning_rate": 3.136440473159843e-06,
      "loss": 1.4605,
      "step": 2218
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.153757095336914,
      "learning_rate": 3.1307335234661376e-06,
      "loss": 0.9416,
      "step": 2219
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.109410285949707,
      "learning_rate": 3.1250302883255502e-06,
      "loss": 1.2267,
      "step": 2220
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.069206237792969,
      "learning_rate": 3.119330773143418e-06,
      "loss": 1.1063,
      "step": 2221
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.913293838500977,
      "learning_rate": 3.113634983321561e-06,
      "loss": 1.8163,
      "step": 2222
    },
    {
      "epoch": 0.69,
      "grad_norm": 17.14433479309082,
      "learning_rate": 3.1079429242582614e-06,
      "loss": 3.1109,
      "step": 2223
    },
    {
      "epoch": 0.69,
      "grad_norm": 15.804655075073242,
      "learning_rate": 3.102254601348264e-06,
      "loss": 2.6792,
      "step": 2224
    },
    {
      "epoch": 0.69,
      "grad_norm": 11.95259952545166,
      "learning_rate": 3.0965700199827813e-06,
      "loss": 1.0476,
      "step": 2225
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.387849807739258,
      "learning_rate": 3.090889185549475e-06,
      "loss": 1.4064,
      "step": 2226
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.103630065917969,
      "learning_rate": 3.0852121034324464e-06,
      "loss": 1.5432,
      "step": 2227
    },
    {
      "epoch": 0.69,
      "grad_norm": 8.44951343536377,
      "learning_rate": 3.0795387790122575e-06,
      "loss": 0.7952,
      "step": 2228
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.66195297241211,
      "learning_rate": 3.0738692176658966e-06,
      "loss": 3.0626,
      "step": 2229
    },
    {
      "epoch": 0.69,
      "grad_norm": 20.553163528442383,
      "learning_rate": 3.0682034247667883e-06,
      "loss": 5.6836,
      "step": 2230
    },
    {
      "epoch": 0.69,
      "grad_norm": 13.766483306884766,
      "learning_rate": 3.062541405684791e-06,
      "loss": 1.8825,
      "step": 2231
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.576692581176758,
      "learning_rate": 3.0568831657861797e-06,
      "loss": 1.3714,
      "step": 2232
    },
    {
      "epoch": 0.69,
      "grad_norm": 12.139288902282715,
      "learning_rate": 3.0512287104336494e-06,
      "loss": 1.2121,
      "step": 2233
    },
    {
      "epoch": 0.69,
      "grad_norm": 15.085342407226562,
      "learning_rate": 3.045578044986308e-06,
      "loss": 2.1256,
      "step": 2234
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.91439151763916,
      "learning_rate": 3.0399311747996738e-06,
      "loss": 2.0942,
      "step": 2235
    },
    {
      "epoch": 0.69,
      "grad_norm": 20.10934829711914,
      "learning_rate": 3.0342881052256634e-06,
      "loss": 3.1579,
      "step": 2236
    },
    {
      "epoch": 0.69,
      "grad_norm": 10.833390235900879,
      "learning_rate": 3.0286488416125994e-06,
      "loss": 1.3159,
      "step": 2237
    },
    {
      "epoch": 0.69,
      "grad_norm": 18.666641235351562,
      "learning_rate": 3.0230133893051884e-06,
      "loss": 2.6354,
      "step": 2238
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.653985023498535,
      "learning_rate": 3.0173817536445305e-06,
      "loss": 5.5188,
      "step": 2239
    },
    {
      "epoch": 0.69,
      "grad_norm": 6.806528568267822,
      "learning_rate": 3.0117539399681053e-06,
      "loss": 0.6842,
      "step": 2240
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.189783096313477,
      "learning_rate": 3.0061299536097723e-06,
      "loss": 1.1556,
      "step": 2241
    },
    {
      "epoch": 0.69,
      "grad_norm": 9.61166763305664,
      "learning_rate": 3.0005097998997586e-06,
      "loss": 1.2695,
      "step": 2242
    },
    {
      "epoch": 0.69,
      "grad_norm": 14.16010856628418,
      "learning_rate": 2.9948934841646695e-06,
      "loss": 4.5669,
      "step": 2243
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.887718200683594,
      "learning_rate": 2.9892810117274626e-06,
      "loss": 2.2022,
      "step": 2244
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.175048828125,
      "learning_rate": 2.983672387907454e-06,
      "loss": 1.3168,
      "step": 2245
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.184016227722168,
      "learning_rate": 2.97806761802032e-06,
      "loss": 3.7449,
      "step": 2246
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.650161743164062,
      "learning_rate": 2.972466707378079e-06,
      "loss": 3.1845,
      "step": 2247
    },
    {
      "epoch": 0.7,
      "grad_norm": 20.005102157592773,
      "learning_rate": 2.966869661289084e-06,
      "loss": 4.9742,
      "step": 2248
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.802370071411133,
      "learning_rate": 2.9612764850580405e-06,
      "loss": 2.6339,
      "step": 2249
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.79102611541748,
      "learning_rate": 2.955687183985976e-06,
      "loss": 3.4665,
      "step": 2250
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.28844165802002,
      "learning_rate": 2.9501017633702462e-06,
      "loss": 1.7935,
      "step": 2251
    },
    {
      "epoch": 0.7,
      "grad_norm": 17.05403709411621,
      "learning_rate": 2.9445202285045356e-06,
      "loss": 4.087,
      "step": 2252
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.470844268798828,
      "learning_rate": 2.9389425846788387e-06,
      "loss": 2.7735,
      "step": 2253
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.51559066772461,
      "learning_rate": 2.9333688371794643e-06,
      "loss": 2.176,
      "step": 2254
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.657980918884277,
      "learning_rate": 2.9277989912890284e-06,
      "loss": 1.8321,
      "step": 2255
    },
    {
      "epoch": 0.7,
      "grad_norm": 10.267308235168457,
      "learning_rate": 2.922233052286451e-06,
      "loss": 1.1327,
      "step": 2256
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.085373878479004,
      "learning_rate": 2.916671025446944e-06,
      "loss": 1.5981,
      "step": 2257
    },
    {
      "epoch": 0.7,
      "grad_norm": 18.13713836669922,
      "learning_rate": 2.91111291604202e-06,
      "loss": 6.3258,
      "step": 2258
    },
    {
      "epoch": 0.7,
      "grad_norm": 15.099580764770508,
      "learning_rate": 2.9055587293394733e-06,
      "loss": 1.563,
      "step": 2259
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.139883995056152,
      "learning_rate": 2.9000084706033795e-06,
      "loss": 3.1056,
      "step": 2260
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.964570045471191,
      "learning_rate": 2.894462145094093e-06,
      "loss": 5.665,
      "step": 2261
    },
    {
      "epoch": 0.7,
      "grad_norm": 19.927221298217773,
      "learning_rate": 2.8889197580682424e-06,
      "loss": 2.7615,
      "step": 2262
    },
    {
      "epoch": 0.7,
      "grad_norm": 19.620908737182617,
      "learning_rate": 2.883381314778717e-06,
      "loss": 3.5992,
      "step": 2263
    },
    {
      "epoch": 0.7,
      "grad_norm": 15.602147102355957,
      "learning_rate": 2.877846820474679e-06,
      "loss": 3.0023,
      "step": 2264
    },
    {
      "epoch": 0.7,
      "grad_norm": 15.21510124206543,
      "learning_rate": 2.872316280401539e-06,
      "loss": 3.4408,
      "step": 2265
    },
    {
      "epoch": 0.7,
      "grad_norm": 13.735747337341309,
      "learning_rate": 2.86678969980096e-06,
      "loss": 1.3311,
      "step": 2266
    },
    {
      "epoch": 0.7,
      "grad_norm": 15.913676261901855,
      "learning_rate": 2.8612670839108617e-06,
      "loss": 1.8771,
      "step": 2267
    },
    {
      "epoch": 0.7,
      "grad_norm": 24.34699249267578,
      "learning_rate": 2.855748437965395e-06,
      "loss": 5.8405,
      "step": 2268
    },
    {
      "epoch": 0.7,
      "grad_norm": 9.142714500427246,
      "learning_rate": 2.850233767194954e-06,
      "loss": 1.0924,
      "step": 2269
    },
    {
      "epoch": 0.7,
      "grad_norm": 12.947314262390137,
      "learning_rate": 2.8447230768261632e-06,
      "loss": 4.7454,
      "step": 2270
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.288124084472656,
      "learning_rate": 2.8392163720818763e-06,
      "loss": 1.3408,
      "step": 2271
    },
    {
      "epoch": 0.7,
      "grad_norm": 11.081491470336914,
      "learning_rate": 2.8337136581811654e-06,
      "loss": 1.6393,
      "step": 2272
    },
    {
      "epoch": 0.7,
      "grad_norm": 16.007856369018555,
      "learning_rate": 2.8282149403393293e-06,
      "loss": 1.7916,
      "step": 2273
    },
    {
      "epoch": 0.7,
      "grad_norm": 14.11226749420166,
      "learning_rate": 2.8227202237678712e-06,
      "loss": 2.4778,
      "step": 2274
    },
    {
      "epoch": 0.7,
      "grad_norm": 8.911184310913086,
      "learning_rate": 2.8172295136745035e-06,
      "loss": 1.3889,
      "step": 2275
    },
    {
      "epoch": 0.71,
      "grad_norm": 14.678910255432129,
      "learning_rate": 2.811742815263144e-06,
      "loss": 1.8128,
      "step": 2276
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.682074546813965,
      "learning_rate": 2.8062601337339043e-06,
      "loss": 1.5763,
      "step": 2277
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.303553581237793,
      "learning_rate": 2.8007814742830922e-06,
      "loss": 2.4316,
      "step": 2278
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.91916275024414,
      "learning_rate": 2.7953068421032044e-06,
      "loss": 3.1608,
      "step": 2279
    },
    {
      "epoch": 0.71,
      "grad_norm": 16.613800048828125,
      "learning_rate": 2.789836242382919e-06,
      "loss": 2.4289,
      "step": 2280
    },
    {
      "epoch": 0.71,
      "grad_norm": 15.66029167175293,
      "learning_rate": 2.7843696803070877e-06,
      "loss": 2.2365,
      "step": 2281
    },
    {
      "epoch": 0.71,
      "grad_norm": 19.280305862426758,
      "learning_rate": 2.778907161056749e-06,
      "loss": 1.7019,
      "step": 2282
    },
    {
      "epoch": 0.71,
      "grad_norm": 14.321356773376465,
      "learning_rate": 2.7734486898090925e-06,
      "loss": 1.0882,
      "step": 2283
    },
    {
      "epoch": 0.71,
      "grad_norm": 9.382092475891113,
      "learning_rate": 2.76799427173748e-06,
      "loss": 1.1687,
      "step": 2284
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.250184059143066,
      "learning_rate": 2.7625439120114352e-06,
      "loss": 1.5645,
      "step": 2285
    },
    {
      "epoch": 0.71,
      "grad_norm": 15.593734741210938,
      "learning_rate": 2.7570976157966304e-06,
      "loss": 2.7318,
      "step": 2286
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.554723739624023,
      "learning_rate": 2.7516553882548847e-06,
      "loss": 0.8545,
      "step": 2287
    },
    {
      "epoch": 0.71,
      "grad_norm": 16.75957489013672,
      "learning_rate": 2.746217234544169e-06,
      "loss": 2.6925,
      "step": 2288
    },
    {
      "epoch": 0.71,
      "grad_norm": 16.92119026184082,
      "learning_rate": 2.740783159818586e-06,
      "loss": 2.4305,
      "step": 2289
    },
    {
      "epoch": 0.71,
      "grad_norm": 14.165618896484375,
      "learning_rate": 2.7353531692283744e-06,
      "loss": 1.7753,
      "step": 2290
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.619165420532227,
      "learning_rate": 2.7299272679199026e-06,
      "loss": 6.1848,
      "step": 2291
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.482929229736328,
      "learning_rate": 2.724505461035663e-06,
      "loss": 1.3093,
      "step": 2292
    },
    {
      "epoch": 0.71,
      "grad_norm": 18.101463317871094,
      "learning_rate": 2.7190877537142655e-06,
      "loss": 3.2845,
      "step": 2293
    },
    {
      "epoch": 0.71,
      "grad_norm": 22.22005844116211,
      "learning_rate": 2.713674151090442e-06,
      "loss": 2.3354,
      "step": 2294
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.901947021484375,
      "learning_rate": 2.7082646582950244e-06,
      "loss": 3.2988,
      "step": 2295
    },
    {
      "epoch": 0.71,
      "grad_norm": 11.771439552307129,
      "learning_rate": 2.702859280454956e-06,
      "loss": 2.8109,
      "step": 2296
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.15504264831543,
      "learning_rate": 2.6974580226932765e-06,
      "loss": 2.5444,
      "step": 2297
    },
    {
      "epoch": 0.71,
      "grad_norm": 24.20153045654297,
      "learning_rate": 2.6920608901291203e-06,
      "loss": 1.683,
      "step": 2298
    },
    {
      "epoch": 0.71,
      "grad_norm": 20.239795684814453,
      "learning_rate": 2.6866678878777185e-06,
      "loss": 2.9259,
      "step": 2299
    },
    {
      "epoch": 0.71,
      "grad_norm": 19.217615127563477,
      "learning_rate": 2.6812790210503795e-06,
      "loss": 1.3868,
      "step": 2300
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.615150451660156,
      "learning_rate": 2.6758942947544954e-06,
      "loss": 2.0394,
      "step": 2301
    },
    {
      "epoch": 0.71,
      "grad_norm": 13.607015609741211,
      "learning_rate": 2.6705137140935384e-06,
      "loss": 2.4767,
      "step": 2302
    },
    {
      "epoch": 0.71,
      "grad_norm": 27.116342544555664,
      "learning_rate": 2.665137284167047e-06,
      "loss": 3.3526,
      "step": 2303
    },
    {
      "epoch": 0.71,
      "grad_norm": 15.009269714355469,
      "learning_rate": 2.659765010070619e-06,
      "loss": 2.6932,
      "step": 2304
    },
    {
      "epoch": 0.71,
      "grad_norm": 12.386222839355469,
      "learning_rate": 2.654396896895929e-06,
      "loss": 1.9814,
      "step": 2305
    },
    {
      "epoch": 0.71,
      "grad_norm": 10.858598709106445,
      "learning_rate": 2.649032949730697e-06,
      "loss": 1.7919,
      "step": 2306
    },
    {
      "epoch": 0.71,
      "grad_norm": 21.328723907470703,
      "learning_rate": 2.6436731736586954e-06,
      "loss": 3.7241,
      "step": 2307
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.507683753967285,
      "learning_rate": 2.63831757375975e-06,
      "loss": 1.5231,
      "step": 2308
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.425612449645996,
      "learning_rate": 2.6329661551097206e-06,
      "loss": 1.3472,
      "step": 2309
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.5155611038208,
      "learning_rate": 2.627618922780509e-06,
      "loss": 2.5453,
      "step": 2310
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.072335243225098,
      "learning_rate": 2.622275881840047e-06,
      "loss": 2.5709,
      "step": 2311
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.811149597167969,
      "learning_rate": 2.616937037352295e-06,
      "loss": 1.7393,
      "step": 2312
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.45902156829834,
      "learning_rate": 2.611602394377233e-06,
      "loss": 5.5844,
      "step": 2313
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.028008460998535,
      "learning_rate": 2.6062719579708676e-06,
      "loss": 1.6212,
      "step": 2314
    },
    {
      "epoch": 0.72,
      "grad_norm": 15.309538841247559,
      "learning_rate": 2.6009457331852094e-06,
      "loss": 2.0222,
      "step": 2315
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.404561996459961,
      "learning_rate": 2.5956237250682816e-06,
      "loss": 2.1896,
      "step": 2316
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.74881362915039,
      "learning_rate": 2.5903059386641104e-06,
      "loss": 1.7404,
      "step": 2317
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.10749340057373,
      "learning_rate": 2.584992379012719e-06,
      "loss": 1.5973,
      "step": 2318
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.463422775268555,
      "learning_rate": 2.5796830511501252e-06,
      "loss": 1.3745,
      "step": 2319
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.964384078979492,
      "learning_rate": 2.574377960108342e-06,
      "loss": 1.6189,
      "step": 2320
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.361332893371582,
      "learning_rate": 2.5690771109153587e-06,
      "loss": 1.3829,
      "step": 2321
    },
    {
      "epoch": 0.72,
      "grad_norm": 17.673812866210938,
      "learning_rate": 2.5637805085951448e-06,
      "loss": 2.4582,
      "step": 2322
    },
    {
      "epoch": 0.72,
      "grad_norm": 18.889476776123047,
      "learning_rate": 2.558488158167653e-06,
      "loss": 4.2102,
      "step": 2323
    },
    {
      "epoch": 0.72,
      "grad_norm": 23.341211318969727,
      "learning_rate": 2.5532000646488008e-06,
      "loss": 2.0677,
      "step": 2324
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.450510025024414,
      "learning_rate": 2.547916233050463e-06,
      "loss": 1.6574,
      "step": 2325
    },
    {
      "epoch": 0.72,
      "grad_norm": 12.01343822479248,
      "learning_rate": 2.542636668380491e-06,
      "loss": 1.1428,
      "step": 2326
    },
    {
      "epoch": 0.72,
      "grad_norm": 23.03483009338379,
      "learning_rate": 2.537361375642682e-06,
      "loss": 2.9922,
      "step": 2327
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.131219863891602,
      "learning_rate": 2.5320903598367836e-06,
      "loss": 1.4113,
      "step": 2328
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.456703186035156,
      "learning_rate": 2.526823625958498e-06,
      "loss": 2.5972,
      "step": 2329
    },
    {
      "epoch": 0.72,
      "grad_norm": 11.621821403503418,
      "learning_rate": 2.5215611789994617e-06,
      "loss": 2.6253,
      "step": 2330
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.902911186218262,
      "learning_rate": 2.51630302394725e-06,
      "loss": 3.9074,
      "step": 2331
    },
    {
      "epoch": 0.72,
      "grad_norm": 16.977209091186523,
      "learning_rate": 2.5110491657853716e-06,
      "loss": 4.3072,
      "step": 2332
    },
    {
      "epoch": 0.72,
      "grad_norm": 16.70897674560547,
      "learning_rate": 2.5057996094932624e-06,
      "loss": 2.3071,
      "step": 2333
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.035141944885254,
      "learning_rate": 2.500554360046278e-06,
      "loss": 3.3543,
      "step": 2334
    },
    {
      "epoch": 0.72,
      "grad_norm": 10.244784355163574,
      "learning_rate": 2.4953134224156995e-06,
      "loss": 1.8466,
      "step": 2335
    },
    {
      "epoch": 0.72,
      "grad_norm": 14.622149467468262,
      "learning_rate": 2.4900768015687146e-06,
      "loss": 4.5747,
      "step": 2336
    },
    {
      "epoch": 0.72,
      "grad_norm": 17.696693420410156,
      "learning_rate": 2.4848445024684204e-06,
      "loss": 5.5025,
      "step": 2337
    },
    {
      "epoch": 0.72,
      "grad_norm": 9.789216995239258,
      "learning_rate": 2.4796165300738265e-06,
      "loss": 0.898,
      "step": 2338
    },
    {
      "epoch": 0.72,
      "grad_norm": 18.043411254882812,
      "learning_rate": 2.4743928893398266e-06,
      "loss": 1.8609,
      "step": 2339
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.23833179473877,
      "learning_rate": 2.469173585217219e-06,
      "loss": 2.3504,
      "step": 2340
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.880558013916016,
      "learning_rate": 2.4639586226526928e-06,
      "loss": 5.6096,
      "step": 2341
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.475369453430176,
      "learning_rate": 2.458748006588819e-06,
      "loss": 4.6972,
      "step": 2342
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.600180625915527,
      "learning_rate": 2.4535417419640467e-06,
      "loss": 2.1673,
      "step": 2343
    },
    {
      "epoch": 0.73,
      "grad_norm": 18.367475509643555,
      "learning_rate": 2.448339833712709e-06,
      "loss": 1.7335,
      "step": 2344
    },
    {
      "epoch": 0.73,
      "grad_norm": 18.15220832824707,
      "learning_rate": 2.4431422867650023e-06,
      "loss": 3.1931,
      "step": 2345
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.09410572052002,
      "learning_rate": 2.4379491060469934e-06,
      "loss": 2.4334,
      "step": 2346
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.841636657714844,
      "learning_rate": 2.432760296480609e-06,
      "loss": 2.0151,
      "step": 2347
    },
    {
      "epoch": 0.73,
      "grad_norm": 14.133190155029297,
      "learning_rate": 2.427575862983633e-06,
      "loss": 1.5586,
      "step": 2348
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.86910343170166,
      "learning_rate": 2.422395810469702e-06,
      "loss": 3.1122,
      "step": 2349
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.083471298217773,
      "learning_rate": 2.4172201438483046e-06,
      "loss": 2.3513,
      "step": 2350
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.80009651184082,
      "learning_rate": 2.412048868024767e-06,
      "loss": 7.2501,
      "step": 2351
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.160103797912598,
      "learning_rate": 2.406881987900256e-06,
      "loss": 3.9078,
      "step": 2352
    },
    {
      "epoch": 0.73,
      "grad_norm": 25.479278564453125,
      "learning_rate": 2.401719508371772e-06,
      "loss": 2.1761,
      "step": 2353
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.644292831420898,
      "learning_rate": 2.396561434332146e-06,
      "loss": 2.5117,
      "step": 2354
    },
    {
      "epoch": 0.73,
      "grad_norm": 9.581793785095215,
      "learning_rate": 2.3914077706700286e-06,
      "loss": 1.1493,
      "step": 2355
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.066975593566895,
      "learning_rate": 2.3862585222699007e-06,
      "loss": 1.4776,
      "step": 2356
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.616497993469238,
      "learning_rate": 2.381113694012049e-06,
      "loss": 1.3081,
      "step": 2357
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.0732421875,
      "learning_rate": 2.3759732907725716e-06,
      "loss": 1.9863,
      "step": 2358
    },
    {
      "epoch": 0.73,
      "grad_norm": 10.824091911315918,
      "learning_rate": 2.370837317423383e-06,
      "loss": 1.1054,
      "step": 2359
    },
    {
      "epoch": 0.73,
      "grad_norm": 23.744216918945312,
      "learning_rate": 2.365705778832184e-06,
      "loss": 2.0633,
      "step": 2360
    },
    {
      "epoch": 0.73,
      "grad_norm": 11.767815589904785,
      "learning_rate": 2.3605786798624796e-06,
      "loss": 1.481,
      "step": 2361
    },
    {
      "epoch": 0.73,
      "grad_norm": 23.761098861694336,
      "learning_rate": 2.3554560253735722e-06,
      "loss": 3.4828,
      "step": 2362
    },
    {
      "epoch": 0.73,
      "grad_norm": 24.555688858032227,
      "learning_rate": 2.350337820220545e-06,
      "loss": 1.5381,
      "step": 2363
    },
    {
      "epoch": 0.73,
      "grad_norm": 22.43702507019043,
      "learning_rate": 2.3452240692542623e-06,
      "loss": 6.4765,
      "step": 2364
    },
    {
      "epoch": 0.73,
      "grad_norm": 8.012927055358887,
      "learning_rate": 2.340114777321377e-06,
      "loss": 1.3097,
      "step": 2365
    },
    {
      "epoch": 0.73,
      "grad_norm": 17.169174194335938,
      "learning_rate": 2.335009949264306e-06,
      "loss": 5.5977,
      "step": 2366
    },
    {
      "epoch": 0.73,
      "grad_norm": 16.629274368286133,
      "learning_rate": 2.3299095899212405e-06,
      "loss": 2.5058,
      "step": 2367
    },
    {
      "epoch": 0.73,
      "grad_norm": 12.617037773132324,
      "learning_rate": 2.324813704126134e-06,
      "loss": 3.6558,
      "step": 2368
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.532926559448242,
      "learning_rate": 2.3197222967087014e-06,
      "loss": 1.8785,
      "step": 2369
    },
    {
      "epoch": 0.73,
      "grad_norm": 15.670842170715332,
      "learning_rate": 2.3146353724944096e-06,
      "loss": 2.7407,
      "step": 2370
    },
    {
      "epoch": 0.73,
      "grad_norm": 13.534579277038574,
      "learning_rate": 2.3095529363044857e-06,
      "loss": 1.2841,
      "step": 2371
    },
    {
      "epoch": 0.73,
      "grad_norm": 19.370758056640625,
      "learning_rate": 2.304474992955894e-06,
      "loss": 2.2037,
      "step": 2372
    },
    {
      "epoch": 0.74,
      "grad_norm": 12.927643775939941,
      "learning_rate": 2.299401547261344e-06,
      "loss": 1.8976,
      "step": 2373
    },
    {
      "epoch": 0.74,
      "grad_norm": 19.296655654907227,
      "learning_rate": 2.2943326040292827e-06,
      "loss": 1.4712,
      "step": 2374
    },
    {
      "epoch": 0.74,
      "grad_norm": 17.31259536743164,
      "learning_rate": 2.2892681680638883e-06,
      "loss": 4.794,
      "step": 2375
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.219792366027832,
      "learning_rate": 2.284208244165067e-06,
      "loss": 2.6991,
      "step": 2376
    },
    {
      "epoch": 0.74,
      "grad_norm": 11.96985149383545,
      "learning_rate": 2.2791528371284547e-06,
      "loss": 1.4655,
      "step": 2377
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.814947128295898,
      "learning_rate": 2.2741019517453987e-06,
      "loss": 1.4603,
      "step": 2378
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.676878929138184,
      "learning_rate": 2.269055592802961e-06,
      "loss": 3.1338,
      "step": 2379
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.400547981262207,
      "learning_rate": 2.2640137650839237e-06,
      "loss": 1.724,
      "step": 2380
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.442301750183105,
      "learning_rate": 2.2589764733667593e-06,
      "loss": 1.5981,
      "step": 2381
    },
    {
      "epoch": 0.74,
      "grad_norm": 15.594927787780762,
      "learning_rate": 2.2539437224256494e-06,
      "loss": 4.5581,
      "step": 2382
    },
    {
      "epoch": 0.74,
      "grad_norm": 17.237245559692383,
      "learning_rate": 2.248915517030477e-06,
      "loss": 4.9534,
      "step": 2383
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.386374473571777,
      "learning_rate": 2.2438918619468073e-06,
      "loss": 2.1693,
      "step": 2384
    },
    {
      "epoch": 0.74,
      "grad_norm": 18.36848258972168,
      "learning_rate": 2.2388727619358956e-06,
      "loss": 2.2726,
      "step": 2385
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.761434555053711,
      "learning_rate": 2.2338582217546863e-06,
      "loss": 1.4383,
      "step": 2386
    },
    {
      "epoch": 0.74,
      "grad_norm": 13.74380111694336,
      "learning_rate": 2.228848246155794e-06,
      "loss": 1.5859,
      "step": 2387
    },
    {
      "epoch": 0.74,
      "grad_norm": 9.30272102355957,
      "learning_rate": 2.223842839887511e-06,
      "loss": 0.9477,
      "step": 2388
    },
    {
      "epoch": 0.74,
      "grad_norm": 15.387125015258789,
      "learning_rate": 2.218842007693798e-06,
      "loss": 4.0006,
      "step": 2389
    },
    {
      "epoch": 0.74,
      "grad_norm": 10.58398723602295,
      "learning_rate": 2.2138457543142815e-06,
      "loss": 1.5678,
      "step": 2390
    },
    {
      "epoch": 0.74,
      "grad_norm": 16.065271377563477,
      "learning_rate": 2.2088540844842454e-06,
      "loss": 4.272,
      "step": 2391
    },
    {
      "epoch": 0.74,
      "grad_norm": 16.36202621459961,
      "learning_rate": 2.2038670029346358e-06,
      "loss": 3.4879,
      "step": 2392
    },
    {
      "epoch": 0.74,
      "grad_norm": 15.290525436401367,
      "learning_rate": 2.1988845143920443e-06,
      "loss": 1.4566,
      "step": 2393
    },
    {
      "epoch": 0.74,
      "grad_norm": 13.585494995117188,
      "learning_rate": 2.1939066235787106e-06,
      "loss": 2.9677,
      "step": 2394
    },
    {
      "epoch": 0.74,
      "grad_norm": 21.19524574279785,
      "learning_rate": 2.1889333352125187e-06,
      "loss": 7.8869,
      "step": 2395
    },
    {
      "epoch": 0.74,
      "grad_norm": 12.337952613830566,
      "learning_rate": 2.183964654006988e-06,
      "loss": 1.1113,
      "step": 2396
    },
    {
      "epoch": 0.74,
      "grad_norm": 19.695003509521484,
      "learning_rate": 2.1790005846712713e-06,
      "loss": 2.2414,
      "step": 2397
    },
    {
      "epoch": 0.74,
      "grad_norm": 15.903563499450684,
      "learning_rate": 2.174041131910156e-06,
      "loss": 4.9721,
      "step": 2398
    },
    {
      "epoch": 0.74,
      "grad_norm": 13.483502388000488,
      "learning_rate": 2.169086300424048e-06,
      "loss": 2.0157,
      "step": 2399
    },
    {
      "epoch": 0.74,
      "grad_norm": 18.105079650878906,
      "learning_rate": 2.1641360949089727e-06,
      "loss": 5.2384,
      "step": 2400
    },
    {
      "epoch": 0.74,
      "grad_norm": 14.571303367614746,
      "learning_rate": 2.15919052005658e-06,
      "loss": 1.2627,
      "step": 2401
    },
    {
      "epoch": 0.74,
      "grad_norm": 29.374582290649414,
      "learning_rate": 2.1542495805541187e-06,
      "loss": 3.0239,
      "step": 2402
    },
    {
      "epoch": 0.74,
      "grad_norm": 18.31540298461914,
      "learning_rate": 2.1493132810844506e-06,
      "loss": 6.1591,
      "step": 2403
    },
    {
      "epoch": 0.74,
      "grad_norm": 12.622870445251465,
      "learning_rate": 2.1443816263260427e-06,
      "loss": 2.1064,
      "step": 2404
    },
    {
      "epoch": 0.75,
      "grad_norm": 21.38667106628418,
      "learning_rate": 2.139454620952956e-06,
      "loss": 3.1859,
      "step": 2405
    },
    {
      "epoch": 0.75,
      "grad_norm": 12.097481727600098,
      "learning_rate": 2.1345322696348416e-06,
      "loss": 1.2453,
      "step": 2406
    },
    {
      "epoch": 0.75,
      "grad_norm": 16.879291534423828,
      "learning_rate": 2.1296145770369503e-06,
      "loss": 1.3218,
      "step": 2407
    },
    {
      "epoch": 0.75,
      "grad_norm": 16.653841018676758,
      "learning_rate": 2.1247015478201074e-06,
      "loss": 1.2542,
      "step": 2408
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.436256408691406,
      "learning_rate": 2.1197931866407225e-06,
      "loss": 1.5462,
      "step": 2409
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.92463207244873,
      "learning_rate": 2.1148894981507804e-06,
      "loss": 1.1283,
      "step": 2410
    },
    {
      "epoch": 0.75,
      "grad_norm": 18.83355712890625,
      "learning_rate": 2.109990486997837e-06,
      "loss": 2.3514,
      "step": 2411
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.721569061279297,
      "learning_rate": 2.1050961578250142e-06,
      "loss": 1.099,
      "step": 2412
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.510601043701172,
      "learning_rate": 2.100206515271002e-06,
      "loss": 1.9727,
      "step": 2413
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.138338088989258,
      "learning_rate": 2.0953215639700404e-06,
      "loss": 2.0488,
      "step": 2414
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.770391464233398,
      "learning_rate": 2.0904413085519295e-06,
      "loss": 3.2804,
      "step": 2415
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.164030075073242,
      "learning_rate": 2.0855657536420156e-06,
      "loss": 1.9102,
      "step": 2416
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.141904830932617,
      "learning_rate": 2.0806949038611903e-06,
      "loss": 2.0403,
      "step": 2417
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.000855445861816,
      "learning_rate": 2.0758287638258846e-06,
      "loss": 2.5367,
      "step": 2418
    },
    {
      "epoch": 0.75,
      "grad_norm": 23.206893920898438,
      "learning_rate": 2.070967338148072e-06,
      "loss": 2.5601,
      "step": 2419
    },
    {
      "epoch": 0.75,
      "grad_norm": 16.114166259765625,
      "learning_rate": 2.0661106314352506e-06,
      "loss": 2.7923,
      "step": 2420
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.76025676727295,
      "learning_rate": 2.061258648290447e-06,
      "loss": 1.8805,
      "step": 2421
    },
    {
      "epoch": 0.75,
      "grad_norm": 13.431325912475586,
      "learning_rate": 2.0564113933122185e-06,
      "loss": 2.7144,
      "step": 2422
    },
    {
      "epoch": 0.75,
      "grad_norm": 20.068647384643555,
      "learning_rate": 2.0515688710946337e-06,
      "loss": 4.914,
      "step": 2423
    },
    {
      "epoch": 0.75,
      "grad_norm": 24.792346954345703,
      "learning_rate": 2.0467310862272714e-06,
      "loss": 5.9553,
      "step": 2424
    },
    {
      "epoch": 0.75,
      "grad_norm": 12.623862266540527,
      "learning_rate": 2.0418980432952343e-06,
      "loss": 2.0634,
      "step": 2425
    },
    {
      "epoch": 0.75,
      "grad_norm": 12.7319974899292,
      "learning_rate": 2.0370697468791187e-06,
      "loss": 3.1398,
      "step": 2426
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.475244522094727,
      "learning_rate": 2.032246201555026e-06,
      "loss": 2.2193,
      "step": 2427
    },
    {
      "epoch": 0.75,
      "grad_norm": 14.422100067138672,
      "learning_rate": 2.0274274118945586e-06,
      "loss": 3.1081,
      "step": 2428
    },
    {
      "epoch": 0.75,
      "grad_norm": 13.511394500732422,
      "learning_rate": 2.022613382464808e-06,
      "loss": 2.309,
      "step": 2429
    },
    {
      "epoch": 0.75,
      "grad_norm": 7.807738780975342,
      "learning_rate": 2.017804117828353e-06,
      "loss": 1.4154,
      "step": 2430
    },
    {
      "epoch": 0.75,
      "grad_norm": 11.46981430053711,
      "learning_rate": 2.0129996225432587e-06,
      "loss": 2.2389,
      "step": 2431
    },
    {
      "epoch": 0.75,
      "grad_norm": 10.154441833496094,
      "learning_rate": 2.0081999011630687e-06,
      "loss": 1.2579,
      "step": 2432
    },
    {
      "epoch": 0.75,
      "grad_norm": 9.019061088562012,
      "learning_rate": 2.003404958236801e-06,
      "loss": 1.5883,
      "step": 2433
    },
    {
      "epoch": 0.75,
      "grad_norm": 13.277936935424805,
      "learning_rate": 1.99861479830895e-06,
      "loss": 4.5607,
      "step": 2434
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.37705135345459,
      "learning_rate": 1.9938294259194714e-06,
      "loss": 4.9968,
      "step": 2435
    },
    {
      "epoch": 0.75,
      "grad_norm": 15.416961669921875,
      "learning_rate": 1.989048845603786e-06,
      "loss": 2.2994,
      "step": 2436
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.432685852050781,
      "learning_rate": 1.984273061892771e-06,
      "loss": 1.6858,
      "step": 2437
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.396636009216309,
      "learning_rate": 1.979502079312759e-06,
      "loss": 1.3837,
      "step": 2438
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.962885856628418,
      "learning_rate": 1.9747359023855308e-06,
      "loss": 2.4708,
      "step": 2439
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.278225898742676,
      "learning_rate": 1.969974535628316e-06,
      "loss": 3.5953,
      "step": 2440
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.124987602233887,
      "learning_rate": 1.965217983553783e-06,
      "loss": 0.9435,
      "step": 2441
    },
    {
      "epoch": 0.76,
      "grad_norm": 14.870964050292969,
      "learning_rate": 1.9604662506700326e-06,
      "loss": 2.4343,
      "step": 2442
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.670650482177734,
      "learning_rate": 1.955719341480609e-06,
      "loss": 3.1242,
      "step": 2443
    },
    {
      "epoch": 0.76,
      "grad_norm": 16.715402603149414,
      "learning_rate": 1.950977260484476e-06,
      "loss": 1.8638,
      "step": 2444
    },
    {
      "epoch": 0.76,
      "grad_norm": 15.901594161987305,
      "learning_rate": 1.9462400121760185e-06,
      "loss": 2.4353,
      "step": 2445
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.394169807434082,
      "learning_rate": 1.9415076010450515e-06,
      "loss": 1.7736,
      "step": 2446
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.738875389099121,
      "learning_rate": 1.936780031576799e-06,
      "loss": 1.1602,
      "step": 2447
    },
    {
      "epoch": 0.76,
      "grad_norm": 17.010435104370117,
      "learning_rate": 1.9320573082518947e-06,
      "loss": 5.2137,
      "step": 2448
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.7792329788208,
      "learning_rate": 1.9273394355463864e-06,
      "loss": 1.8585,
      "step": 2449
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.063002586364746,
      "learning_rate": 1.9226264179317185e-06,
      "loss": 2.4567,
      "step": 2450
    },
    {
      "epoch": 0.76,
      "grad_norm": 12.918591499328613,
      "learning_rate": 1.917918259874735e-06,
      "loss": 0.7596,
      "step": 2451
    },
    {
      "epoch": 0.76,
      "grad_norm": 15.485755920410156,
      "learning_rate": 1.9132149658376742e-06,
      "loss": 2.6869,
      "step": 2452
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.972342491149902,
      "learning_rate": 1.908516540278166e-06,
      "loss": 1.1617,
      "step": 2453
    },
    {
      "epoch": 0.76,
      "grad_norm": 22.043930053710938,
      "learning_rate": 1.903822987649223e-06,
      "loss": 2.5623,
      "step": 2454
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.817049980163574,
      "learning_rate": 1.899134312399245e-06,
      "loss": 1.2553,
      "step": 2455
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.989843368530273,
      "learning_rate": 1.8944505189720034e-06,
      "loss": 1.54,
      "step": 2456
    },
    {
      "epoch": 0.76,
      "grad_norm": 15.252662658691406,
      "learning_rate": 1.889771611806647e-06,
      "loss": 3.6877,
      "step": 2457
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.54154109954834,
      "learning_rate": 1.88509759533769e-06,
      "loss": 1.6751,
      "step": 2458
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.486143112182617,
      "learning_rate": 1.8804284739950129e-06,
      "loss": 3.4475,
      "step": 2459
    },
    {
      "epoch": 0.76,
      "grad_norm": 9.09095287322998,
      "learning_rate": 1.8757642522038544e-06,
      "loss": 1.2124,
      "step": 2460
    },
    {
      "epoch": 0.76,
      "grad_norm": 10.430471420288086,
      "learning_rate": 1.8711049343848184e-06,
      "loss": 1.4979,
      "step": 2461
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.80158805847168,
      "learning_rate": 1.8664505249538497e-06,
      "loss": 1.4916,
      "step": 2462
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.338501930236816,
      "learning_rate": 1.8618010283222455e-06,
      "loss": 1.3531,
      "step": 2463
    },
    {
      "epoch": 0.76,
      "grad_norm": 17.377933502197266,
      "learning_rate": 1.8571564488966517e-06,
      "loss": 2.4318,
      "step": 2464
    },
    {
      "epoch": 0.76,
      "grad_norm": 14.895021438598633,
      "learning_rate": 1.8525167910790465e-06,
      "loss": 5.0413,
      "step": 2465
    },
    {
      "epoch": 0.76,
      "grad_norm": 15.143299102783203,
      "learning_rate": 1.8478820592667463e-06,
      "loss": 5.4828,
      "step": 2466
    },
    {
      "epoch": 0.76,
      "grad_norm": 11.624774932861328,
      "learning_rate": 1.8432522578523997e-06,
      "loss": 2.1534,
      "step": 2467
    },
    {
      "epoch": 0.76,
      "grad_norm": 13.456793785095215,
      "learning_rate": 1.8386273912239786e-06,
      "loss": 1.6974,
      "step": 2468
    },
    {
      "epoch": 0.76,
      "grad_norm": 15.70180606842041,
      "learning_rate": 1.8340074637647865e-06,
      "loss": 6.4944,
      "step": 2469
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.85049057006836,
      "learning_rate": 1.8293924798534375e-06,
      "loss": 1.6007,
      "step": 2470
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.042861938476562,
      "learning_rate": 1.824782443863862e-06,
      "loss": 1.3233,
      "step": 2471
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.435417175292969,
      "learning_rate": 1.8201773601653036e-06,
      "loss": 1.0364,
      "step": 2472
    },
    {
      "epoch": 0.77,
      "grad_norm": 6.530979156494141,
      "learning_rate": 1.8155772331223098e-06,
      "loss": 0.9042,
      "step": 2473
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.234294891357422,
      "learning_rate": 1.8109820670947296e-06,
      "loss": 1.4322,
      "step": 2474
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.820496559143066,
      "learning_rate": 1.8063918664377154e-06,
      "loss": 1.0702,
      "step": 2475
    },
    {
      "epoch": 0.77,
      "grad_norm": 11.390840530395508,
      "learning_rate": 1.8018066355017072e-06,
      "loss": 1.2658,
      "step": 2476
    },
    {
      "epoch": 0.77,
      "grad_norm": 18.800928115844727,
      "learning_rate": 1.7972263786324366e-06,
      "loss": 6.5576,
      "step": 2477
    },
    {
      "epoch": 0.77,
      "grad_norm": 17.33761978149414,
      "learning_rate": 1.792651100170926e-06,
      "loss": 4.825,
      "step": 2478
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.361677169799805,
      "learning_rate": 1.788080804453474e-06,
      "loss": 1.3772,
      "step": 2479
    },
    {
      "epoch": 0.77,
      "grad_norm": 22.153621673583984,
      "learning_rate": 1.7835154958116536e-06,
      "loss": 5.4864,
      "step": 2480
    },
    {
      "epoch": 0.77,
      "grad_norm": 14.149209976196289,
      "learning_rate": 1.7789551785723201e-06,
      "loss": 2.4508,
      "step": 2481
    },
    {
      "epoch": 0.77,
      "grad_norm": 14.432504653930664,
      "learning_rate": 1.7743998570575935e-06,
      "loss": 1.1354,
      "step": 2482
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.748077392578125,
      "learning_rate": 1.7698495355848553e-06,
      "loss": 2.4884,
      "step": 2483
    },
    {
      "epoch": 0.77,
      "grad_norm": 18.71636199951172,
      "learning_rate": 1.7653042184667574e-06,
      "loss": 5.2609,
      "step": 2484
    },
    {
      "epoch": 0.77,
      "grad_norm": 14.906200408935547,
      "learning_rate": 1.7607639100112006e-06,
      "loss": 2.6443,
      "step": 2485
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.056180953979492,
      "learning_rate": 1.7562286145213414e-06,
      "loss": 1.2123,
      "step": 2486
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.615121841430664,
      "learning_rate": 1.751698336295585e-06,
      "loss": 1.2911,
      "step": 2487
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.8436861038208,
      "learning_rate": 1.747173079627582e-06,
      "loss": 1.458,
      "step": 2488
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.317079544067383,
      "learning_rate": 1.7426528488062211e-06,
      "loss": 1.7317,
      "step": 2489
    },
    {
      "epoch": 0.77,
      "grad_norm": 10.851561546325684,
      "learning_rate": 1.7381376481156337e-06,
      "loss": 1.1819,
      "step": 2490
    },
    {
      "epoch": 0.77,
      "grad_norm": 8.39650821685791,
      "learning_rate": 1.7336274818351788e-06,
      "loss": 1.8603,
      "step": 2491
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.600789070129395,
      "learning_rate": 1.7291223542394436e-06,
      "loss": 3.2663,
      "step": 2492
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.00397777557373,
      "learning_rate": 1.7246222695982422e-06,
      "loss": 1.7545,
      "step": 2493
    },
    {
      "epoch": 0.77,
      "grad_norm": 18.165191650390625,
      "learning_rate": 1.7201272321766084e-06,
      "loss": 4.9447,
      "step": 2494
    },
    {
      "epoch": 0.77,
      "grad_norm": 9.986980438232422,
      "learning_rate": 1.7156372462347903e-06,
      "loss": 1.6906,
      "step": 2495
    },
    {
      "epoch": 0.77,
      "grad_norm": 19.583627700805664,
      "learning_rate": 1.711152316028254e-06,
      "loss": 4.1174,
      "step": 2496
    },
    {
      "epoch": 0.77,
      "grad_norm": 15.571231842041016,
      "learning_rate": 1.7066724458076684e-06,
      "loss": 2.7608,
      "step": 2497
    },
    {
      "epoch": 0.77,
      "grad_norm": 24.01877784729004,
      "learning_rate": 1.7021976398189062e-06,
      "loss": 3.5909,
      "step": 2498
    },
    {
      "epoch": 0.77,
      "grad_norm": 12.039501190185547,
      "learning_rate": 1.697727902303047e-06,
      "loss": 2.7924,
      "step": 2499
    },
    {
      "epoch": 0.77,
      "grad_norm": 16.051227569580078,
      "learning_rate": 1.6932632374963631e-06,
      "loss": 2.0519,
      "step": 2500
    },
    {
      "epoch": 0.77,
      "grad_norm": 18.50823211669922,
      "learning_rate": 1.6888036496303116e-06,
      "loss": 4.4319,
      "step": 2501
    },
    {
      "epoch": 0.78,
      "grad_norm": 21.07451057434082,
      "learning_rate": 1.6843491429315508e-06,
      "loss": 2.6907,
      "step": 2502
    },
    {
      "epoch": 0.78,
      "grad_norm": 15.278072357177734,
      "learning_rate": 1.679899721621915e-06,
      "loss": 1.422,
      "step": 2503
    },
    {
      "epoch": 0.78,
      "grad_norm": 12.383358001708984,
      "learning_rate": 1.6754553899184191e-06,
      "loss": 2.0169,
      "step": 2504
    },
    {
      "epoch": 0.78,
      "grad_norm": 16.6717529296875,
      "learning_rate": 1.6710161520332601e-06,
      "loss": 3.1959,
      "step": 2505
    },
    {
      "epoch": 0.78,
      "grad_norm": 8.391064643859863,
      "learning_rate": 1.666582012173801e-06,
      "loss": 0.8575,
      "step": 2506
    },
    {
      "epoch": 0.78,
      "grad_norm": 15.009490966796875,
      "learning_rate": 1.6621529745425744e-06,
      "loss": 2.3967,
      "step": 2507
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.876742362976074,
      "learning_rate": 1.6577290433372782e-06,
      "loss": 2.0324,
      "step": 2508
    },
    {
      "epoch": 0.78,
      "grad_norm": 18.65607261657715,
      "learning_rate": 1.6533102227507705e-06,
      "loss": 3.8017,
      "step": 2509
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.521139144897461,
      "learning_rate": 1.6488965169710638e-06,
      "loss": 1.4466,
      "step": 2510
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.26103687286377,
      "learning_rate": 1.6444879301813294e-06,
      "loss": 2.4333,
      "step": 2511
    },
    {
      "epoch": 0.78,
      "grad_norm": 9.338981628417969,
      "learning_rate": 1.6400844665598781e-06,
      "loss": 1.7815,
      "step": 2512
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.998225212097168,
      "learning_rate": 1.635686130280171e-06,
      "loss": 1.176,
      "step": 2513
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.324702262878418,
      "learning_rate": 1.631292925510807e-06,
      "loss": 0.9974,
      "step": 2514
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.276659965515137,
      "learning_rate": 1.6269048564155235e-06,
      "loss": 0.9944,
      "step": 2515
    },
    {
      "epoch": 0.78,
      "grad_norm": 14.636279106140137,
      "learning_rate": 1.6225219271531878e-06,
      "loss": 1.7558,
      "step": 2516
    },
    {
      "epoch": 0.78,
      "grad_norm": 20.32813262939453,
      "learning_rate": 1.6181441418777997e-06,
      "loss": 1.9024,
      "step": 2517
    },
    {
      "epoch": 0.78,
      "grad_norm": 7.731251239776611,
      "learning_rate": 1.6137715047384803e-06,
      "loss": 1.06,
      "step": 2518
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.945781707763672,
      "learning_rate": 1.609404019879471e-06,
      "loss": 2.0694,
      "step": 2519
    },
    {
      "epoch": 0.78,
      "grad_norm": 12.413304328918457,
      "learning_rate": 1.6050416914401355e-06,
      "loss": 3.1534,
      "step": 2520
    },
    {
      "epoch": 0.78,
      "grad_norm": 14.440204620361328,
      "learning_rate": 1.6006845235549472e-06,
      "loss": 1.7909,
      "step": 2521
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.835144996643066,
      "learning_rate": 1.5963325203534811e-06,
      "loss": 1.6483,
      "step": 2522
    },
    {
      "epoch": 0.78,
      "grad_norm": 12.865160942077637,
      "learning_rate": 1.5919856859604304e-06,
      "loss": 1.72,
      "step": 2523
    },
    {
      "epoch": 0.78,
      "grad_norm": 10.06251049041748,
      "learning_rate": 1.587644024495581e-06,
      "loss": 1.3708,
      "step": 2524
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.970108032226562,
      "learning_rate": 1.583307540073816e-06,
      "loss": 1.6351,
      "step": 2525
    },
    {
      "epoch": 0.78,
      "grad_norm": 14.133593559265137,
      "learning_rate": 1.578976236805119e-06,
      "loss": 1.863,
      "step": 2526
    },
    {
      "epoch": 0.78,
      "grad_norm": 11.517053604125977,
      "learning_rate": 1.5746501187945555e-06,
      "loss": 1.1233,
      "step": 2527
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.115801811218262,
      "learning_rate": 1.5703291901422795e-06,
      "loss": 1.1439,
      "step": 2528
    },
    {
      "epoch": 0.78,
      "grad_norm": 15.460615158081055,
      "learning_rate": 1.5660134549435257e-06,
      "loss": 1.745,
      "step": 2529
    },
    {
      "epoch": 0.78,
      "grad_norm": 13.269757270812988,
      "learning_rate": 1.561702917288609e-06,
      "loss": 1.1088,
      "step": 2530
    },
    {
      "epoch": 0.78,
      "grad_norm": 15.581395149230957,
      "learning_rate": 1.5573975812629126e-06,
      "loss": 2.7008,
      "step": 2531
    },
    {
      "epoch": 0.78,
      "grad_norm": 15.502326965332031,
      "learning_rate": 1.5530974509468995e-06,
      "loss": 5.0785,
      "step": 2532
    },
    {
      "epoch": 0.78,
      "grad_norm": 19.524160385131836,
      "learning_rate": 1.5488025304160903e-06,
      "loss": 4.5823,
      "step": 2533
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.589733123779297,
      "learning_rate": 1.5445128237410704e-06,
      "loss": 2.689,
      "step": 2534
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.67539119720459,
      "learning_rate": 1.5402283349874845e-06,
      "loss": 2.7477,
      "step": 2535
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.214372634887695,
      "learning_rate": 1.5359490682160318e-06,
      "loss": 3.1485,
      "step": 2536
    },
    {
      "epoch": 0.79,
      "grad_norm": 18.086790084838867,
      "learning_rate": 1.5316750274824598e-06,
      "loss": 2.002,
      "step": 2537
    },
    {
      "epoch": 0.79,
      "grad_norm": 15.74965763092041,
      "learning_rate": 1.5274062168375691e-06,
      "loss": 1.3451,
      "step": 2538
    },
    {
      "epoch": 0.79,
      "grad_norm": 16.10420036315918,
      "learning_rate": 1.5231426403271976e-06,
      "loss": 3.2361,
      "step": 2539
    },
    {
      "epoch": 0.79,
      "grad_norm": 15.154731750488281,
      "learning_rate": 1.5188843019922215e-06,
      "loss": 1.6669,
      "step": 2540
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.527267456054688,
      "learning_rate": 1.5146312058685597e-06,
      "loss": 1.3462,
      "step": 2541
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.436562538146973,
      "learning_rate": 1.5103833559871566e-06,
      "loss": 2.5334,
      "step": 2542
    },
    {
      "epoch": 0.79,
      "grad_norm": 18.57342529296875,
      "learning_rate": 1.5061407563739853e-06,
      "loss": 2.7942,
      "step": 2543
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.763381958007812,
      "learning_rate": 1.5019034110500433e-06,
      "loss": 2.4587,
      "step": 2544
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.633790969848633,
      "learning_rate": 1.4976713240313496e-06,
      "loss": 2.1891,
      "step": 2545
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.143424987792969,
      "learning_rate": 1.4934444993289352e-06,
      "loss": 2.2721,
      "step": 2546
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.036821365356445,
      "learning_rate": 1.489222940948852e-06,
      "loss": 1.8695,
      "step": 2547
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.992802619934082,
      "learning_rate": 1.485006652892153e-06,
      "loss": 1.069,
      "step": 2548
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.13967514038086,
      "learning_rate": 1.4807956391548982e-06,
      "loss": 1.1058,
      "step": 2549
    },
    {
      "epoch": 0.79,
      "grad_norm": 9.486446380615234,
      "learning_rate": 1.4765899037281503e-06,
      "loss": 1.4127,
      "step": 2550
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.577086448669434,
      "learning_rate": 1.4723894505979675e-06,
      "loss": 1.7006,
      "step": 2551
    },
    {
      "epoch": 0.79,
      "grad_norm": 11.6814603805542,
      "learning_rate": 1.4681942837454012e-06,
      "loss": 2.0487,
      "step": 2552
    },
    {
      "epoch": 0.79,
      "grad_norm": 16.907302856445312,
      "learning_rate": 1.4640044071464966e-06,
      "loss": 4.4646,
      "step": 2553
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.21091365814209,
      "learning_rate": 1.4598198247722822e-06,
      "loss": 1.5015,
      "step": 2554
    },
    {
      "epoch": 0.79,
      "grad_norm": 10.397065162658691,
      "learning_rate": 1.4556405405887654e-06,
      "loss": 1.2597,
      "step": 2555
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.379748344421387,
      "learning_rate": 1.4514665585569415e-06,
      "loss": 2.5205,
      "step": 2556
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.382526397705078,
      "learning_rate": 1.4472978826327705e-06,
      "loss": 3.0737,
      "step": 2557
    },
    {
      "epoch": 0.79,
      "grad_norm": 12.831366539001465,
      "learning_rate": 1.4431345167671862e-06,
      "loss": 2.0951,
      "step": 2558
    },
    {
      "epoch": 0.79,
      "grad_norm": 15.526214599609375,
      "learning_rate": 1.4389764649060951e-06,
      "loss": 2.0521,
      "step": 2559
    },
    {
      "epoch": 0.79,
      "grad_norm": 7.109898567199707,
      "learning_rate": 1.4348237309903618e-06,
      "loss": 0.6138,
      "step": 2560
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.867196083068848,
      "learning_rate": 1.43067631895581e-06,
      "loss": 1.8706,
      "step": 2561
    },
    {
      "epoch": 0.79,
      "grad_norm": 19.599706649780273,
      "learning_rate": 1.4265342327332263e-06,
      "loss": 2.4417,
      "step": 2562
    },
    {
      "epoch": 0.79,
      "grad_norm": 18.613494873046875,
      "learning_rate": 1.4223974762483427e-06,
      "loss": 2.3428,
      "step": 2563
    },
    {
      "epoch": 0.79,
      "grad_norm": 8.591452598571777,
      "learning_rate": 1.4182660534218415e-06,
      "loss": 0.9621,
      "step": 2564
    },
    {
      "epoch": 0.79,
      "grad_norm": 13.627431869506836,
      "learning_rate": 1.4141399681693522e-06,
      "loss": 3.8242,
      "step": 2565
    },
    {
      "epoch": 0.8,
      "grad_norm": 20.302942276000977,
      "learning_rate": 1.4100192244014433e-06,
      "loss": 3.7236,
      "step": 2566
    },
    {
      "epoch": 0.8,
      "grad_norm": 16.992944717407227,
      "learning_rate": 1.4059038260236196e-06,
      "loss": 2.4187,
      "step": 2567
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.403325080871582,
      "learning_rate": 1.4017937769363246e-06,
      "loss": 2.3914,
      "step": 2568
    },
    {
      "epoch": 0.8,
      "grad_norm": 18.10266876220703,
      "learning_rate": 1.397689081034929e-06,
      "loss": 2.1453,
      "step": 2569
    },
    {
      "epoch": 0.8,
      "grad_norm": 9.529816627502441,
      "learning_rate": 1.3935897422097285e-06,
      "loss": 1.4158,
      "step": 2570
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.614334106445312,
      "learning_rate": 1.3894957643459425e-06,
      "loss": 1.75,
      "step": 2571
    },
    {
      "epoch": 0.8,
      "grad_norm": 24.91597557067871,
      "learning_rate": 1.3854071513237115e-06,
      "loss": 3.66,
      "step": 2572
    },
    {
      "epoch": 0.8,
      "grad_norm": 15.958589553833008,
      "learning_rate": 1.381323907018087e-06,
      "loss": 1.594,
      "step": 2573
    },
    {
      "epoch": 0.8,
      "grad_norm": 20.908681869506836,
      "learning_rate": 1.3772460352990386e-06,
      "loss": 2.5328,
      "step": 2574
    },
    {
      "epoch": 0.8,
      "grad_norm": 15.671060562133789,
      "learning_rate": 1.3731735400314382e-06,
      "loss": 1.8023,
      "step": 2575
    },
    {
      "epoch": 0.8,
      "grad_norm": 21.42799949645996,
      "learning_rate": 1.3691064250750629e-06,
      "loss": 2.8076,
      "step": 2576
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.156183242797852,
      "learning_rate": 1.365044694284597e-06,
      "loss": 0.9839,
      "step": 2577
    },
    {
      "epoch": 0.8,
      "grad_norm": 15.905706405639648,
      "learning_rate": 1.3609883515096112e-06,
      "loss": 3.1947,
      "step": 2578
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.29395866394043,
      "learning_rate": 1.3569374005945752e-06,
      "loss": 1.5666,
      "step": 2579
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.16615104675293,
      "learning_rate": 1.3528918453788517e-06,
      "loss": 1.1203,
      "step": 2580
    },
    {
      "epoch": 0.8,
      "grad_norm": 17.40990447998047,
      "learning_rate": 1.3488516896966847e-06,
      "loss": 1.161,
      "step": 2581
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.006482124328613,
      "learning_rate": 1.3448169373772002e-06,
      "loss": 3.4506,
      "step": 2582
    },
    {
      "epoch": 0.8,
      "grad_norm": 17.400257110595703,
      "learning_rate": 1.3407875922444088e-06,
      "loss": 5.522,
      "step": 2583
    },
    {
      "epoch": 0.8,
      "grad_norm": 15.19652271270752,
      "learning_rate": 1.3367636581171906e-06,
      "loss": 2.3077,
      "step": 2584
    },
    {
      "epoch": 0.8,
      "grad_norm": 18.77398681640625,
      "learning_rate": 1.3327451388092991e-06,
      "loss": 2.8604,
      "step": 2585
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.453078269958496,
      "learning_rate": 1.3287320381293562e-06,
      "loss": 4.7492,
      "step": 2586
    },
    {
      "epoch": 0.8,
      "grad_norm": 15.172926902770996,
      "learning_rate": 1.3247243598808481e-06,
      "loss": 2.9449,
      "step": 2587
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.221792221069336,
      "learning_rate": 1.3207221078621198e-06,
      "loss": 2.237,
      "step": 2588
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.888528823852539,
      "learning_rate": 1.3167252858663798e-06,
      "loss": 2.8871,
      "step": 2589
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.226152420043945,
      "learning_rate": 1.3127338976816826e-06,
      "loss": 1.6073,
      "step": 2590
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.393001556396484,
      "learning_rate": 1.308747947090937e-06,
      "loss": 2.7073,
      "step": 2591
    },
    {
      "epoch": 0.8,
      "grad_norm": 11.04410457611084,
      "learning_rate": 1.3047674378718965e-06,
      "loss": 1.8482,
      "step": 2592
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.1553316116333,
      "learning_rate": 1.3007923737971589e-06,
      "loss": 2.3972,
      "step": 2593
    },
    {
      "epoch": 0.8,
      "grad_norm": 13.165019035339355,
      "learning_rate": 1.2968227586341577e-06,
      "loss": 1.7444,
      "step": 2594
    },
    {
      "epoch": 0.8,
      "grad_norm": 12.589109420776367,
      "learning_rate": 1.2928585961451686e-06,
      "loss": 1.3718,
      "step": 2595
    },
    {
      "epoch": 0.8,
      "grad_norm": 16.85529136657715,
      "learning_rate": 1.2888998900872945e-06,
      "loss": 3.3007,
      "step": 2596
    },
    {
      "epoch": 0.8,
      "grad_norm": 16.978525161743164,
      "learning_rate": 1.2849466442124665e-06,
      "loss": 3.3992,
      "step": 2597
    },
    {
      "epoch": 0.8,
      "grad_norm": 14.961907386779785,
      "learning_rate": 1.2809988622674466e-06,
      "loss": 2.7419,
      "step": 2598
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.649197578430176,
      "learning_rate": 1.277056547993809e-06,
      "loss": 1.668,
      "step": 2599
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.195618629455566,
      "learning_rate": 1.2731197051279508e-06,
      "loss": 1.4284,
      "step": 2600
    },
    {
      "epoch": 0.81,
      "grad_norm": 15.158956527709961,
      "learning_rate": 1.2691883374010877e-06,
      "loss": 1.3633,
      "step": 2601
    },
    {
      "epoch": 0.81,
      "grad_norm": 20.267412185668945,
      "learning_rate": 1.2652624485392395e-06,
      "loss": 5.4692,
      "step": 2602
    },
    {
      "epoch": 0.81,
      "grad_norm": 25.416479110717773,
      "learning_rate": 1.261342042263234e-06,
      "loss": 3.8892,
      "step": 2603
    },
    {
      "epoch": 0.81,
      "grad_norm": 15.120532989501953,
      "learning_rate": 1.257427122288708e-06,
      "loss": 1.7587,
      "step": 2604
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.286882400512695,
      "learning_rate": 1.2535176923260937e-06,
      "loss": 1.3639,
      "step": 2605
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.099992752075195,
      "learning_rate": 1.2496137560806214e-06,
      "loss": 1.2341,
      "step": 2606
    },
    {
      "epoch": 0.81,
      "grad_norm": 16.102886199951172,
      "learning_rate": 1.245715317252315e-06,
      "loss": 1.8498,
      "step": 2607
    },
    {
      "epoch": 0.81,
      "grad_norm": 16.775665283203125,
      "learning_rate": 1.2418223795359877e-06,
      "loss": 1.6815,
      "step": 2608
    },
    {
      "epoch": 0.81,
      "grad_norm": 7.794978618621826,
      "learning_rate": 1.237934946621237e-06,
      "loss": 0.7872,
      "step": 2609
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.038506507873535,
      "learning_rate": 1.2340530221924487e-06,
      "loss": 1.549,
      "step": 2610
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.117688179016113,
      "learning_rate": 1.2301766099287831e-06,
      "loss": 1.0437,
      "step": 2611
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.886580467224121,
      "learning_rate": 1.2263057135041772e-06,
      "loss": 1.0861,
      "step": 2612
    },
    {
      "epoch": 0.81,
      "grad_norm": 15.002225875854492,
      "learning_rate": 1.2224403365873406e-06,
      "loss": 2.5014,
      "step": 2613
    },
    {
      "epoch": 0.81,
      "grad_norm": 11.995274543762207,
      "learning_rate": 1.2185804828417515e-06,
      "loss": 1.7224,
      "step": 2614
    },
    {
      "epoch": 0.81,
      "grad_norm": 15.890597343444824,
      "learning_rate": 1.2147261559256523e-06,
      "loss": 2.7263,
      "step": 2615
    },
    {
      "epoch": 0.81,
      "grad_norm": 15.805421829223633,
      "learning_rate": 1.2108773594920514e-06,
      "loss": 1.3961,
      "step": 2616
    },
    {
      "epoch": 0.81,
      "grad_norm": 14.018182754516602,
      "learning_rate": 1.207034097188711e-06,
      "loss": 2.0382,
      "step": 2617
    },
    {
      "epoch": 0.81,
      "grad_norm": 14.638632774353027,
      "learning_rate": 1.2031963726581484e-06,
      "loss": 2.6442,
      "step": 2618
    },
    {
      "epoch": 0.81,
      "grad_norm": 17.171119689941406,
      "learning_rate": 1.1993641895376371e-06,
      "loss": 2.7081,
      "step": 2619
    },
    {
      "epoch": 0.81,
      "grad_norm": 16.15960121154785,
      "learning_rate": 1.1955375514591967e-06,
      "loss": 5.6566,
      "step": 2620
    },
    {
      "epoch": 0.81,
      "grad_norm": 9.930737495422363,
      "learning_rate": 1.1917164620495836e-06,
      "loss": 1.0565,
      "step": 2621
    },
    {
      "epoch": 0.81,
      "grad_norm": 18.560791015625,
      "learning_rate": 1.1879009249303071e-06,
      "loss": 1.7435,
      "step": 2622
    },
    {
      "epoch": 0.81,
      "grad_norm": 16.398456573486328,
      "learning_rate": 1.1840909437176075e-06,
      "loss": 2.5659,
      "step": 2623
    },
    {
      "epoch": 0.81,
      "grad_norm": 24.315637588500977,
      "learning_rate": 1.1802865220224601e-06,
      "loss": 2.9749,
      "step": 2624
    },
    {
      "epoch": 0.81,
      "grad_norm": 10.489819526672363,
      "learning_rate": 1.176487663450574e-06,
      "loss": 1.0255,
      "step": 2625
    },
    {
      "epoch": 0.81,
      "grad_norm": 14.841395378112793,
      "learning_rate": 1.1726943716023828e-06,
      "loss": 2.2453,
      "step": 2626
    },
    {
      "epoch": 0.81,
      "grad_norm": 20.017536163330078,
      "learning_rate": 1.168906650073044e-06,
      "loss": 2.8686,
      "step": 2627
    },
    {
      "epoch": 0.81,
      "grad_norm": 20.4256591796875,
      "learning_rate": 1.1651245024524376e-06,
      "loss": 2.2901,
      "step": 2628
    },
    {
      "epoch": 0.81,
      "grad_norm": 13.860508918762207,
      "learning_rate": 1.1613479323251603e-06,
      "loss": 3.3471,
      "step": 2629
    },
    {
      "epoch": 0.81,
      "grad_norm": 12.17288589477539,
      "learning_rate": 1.1575769432705193e-06,
      "loss": 2.8286,
      "step": 2630
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.398018836975098,
      "learning_rate": 1.1538115388625396e-06,
      "loss": 1.0916,
      "step": 2631
    },
    {
      "epoch": 0.82,
      "grad_norm": 15.136306762695312,
      "learning_rate": 1.1500517226699459e-06,
      "loss": 2.655,
      "step": 2632
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.895465850830078,
      "learning_rate": 1.146297498256171e-06,
      "loss": 1.1909,
      "step": 2633
    },
    {
      "epoch": 0.82,
      "grad_norm": 16.685121536254883,
      "learning_rate": 1.1425488691793465e-06,
      "loss": 2.6348,
      "step": 2634
    },
    {
      "epoch": 0.82,
      "grad_norm": 18.08514404296875,
      "learning_rate": 1.1388058389922978e-06,
      "loss": 3.771,
      "step": 2635
    },
    {
      "epoch": 0.82,
      "grad_norm": 14.868639945983887,
      "learning_rate": 1.1350684112425515e-06,
      "loss": 5.4822,
      "step": 2636
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.509803771972656,
      "learning_rate": 1.131336589472318e-06,
      "loss": 1.8023,
      "step": 2637
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.3806791305542,
      "learning_rate": 1.127610377218495e-06,
      "loss": 1.6137,
      "step": 2638
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.26029109954834,
      "learning_rate": 1.1238897780126684e-06,
      "loss": 2.9263,
      "step": 2639
    },
    {
      "epoch": 0.82,
      "grad_norm": 9.973566055297852,
      "learning_rate": 1.1201747953810984e-06,
      "loss": 1.2468,
      "step": 2640
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.38296890258789,
      "learning_rate": 1.1164654328447256e-06,
      "loss": 1.5823,
      "step": 2641
    },
    {
      "epoch": 0.82,
      "grad_norm": 8.711637496948242,
      "learning_rate": 1.1127616939191625e-06,
      "loss": 1.1944,
      "step": 2642
    },
    {
      "epoch": 0.82,
      "grad_norm": 15.474139213562012,
      "learning_rate": 1.109063582114693e-06,
      "loss": 3.4932,
      "step": 2643
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.657920837402344,
      "learning_rate": 1.1053711009362643e-06,
      "loss": 4.9638,
      "step": 2644
    },
    {
      "epoch": 0.82,
      "grad_norm": 18.262088775634766,
      "learning_rate": 1.1016842538834934e-06,
      "loss": 2.6687,
      "step": 2645
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.732860565185547,
      "learning_rate": 1.0980030444506517e-06,
      "loss": 3.2846,
      "step": 2646
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.444441795349121,
      "learning_rate": 1.0943274761266708e-06,
      "loss": 2.0819,
      "step": 2647
    },
    {
      "epoch": 0.82,
      "grad_norm": 15.731082916259766,
      "learning_rate": 1.0906575523951347e-06,
      "loss": 2.9561,
      "step": 2648
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.473064422607422,
      "learning_rate": 1.0869932767342762e-06,
      "loss": 1.7959,
      "step": 2649
    },
    {
      "epoch": 0.82,
      "grad_norm": 18.323083877563477,
      "learning_rate": 1.0833346526169758e-06,
      "loss": 2.6552,
      "step": 2650
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.62148666381836,
      "learning_rate": 1.0796816835107606e-06,
      "loss": 1.621,
      "step": 2651
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.119149208068848,
      "learning_rate": 1.0760343728777956e-06,
      "loss": 1.3556,
      "step": 2652
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.881817817687988,
      "learning_rate": 1.0723927241748803e-06,
      "loss": 1.7563,
      "step": 2653
    },
    {
      "epoch": 0.82,
      "grad_norm": 11.731474876403809,
      "learning_rate": 1.068756740853456e-06,
      "loss": 1.1546,
      "step": 2654
    },
    {
      "epoch": 0.82,
      "grad_norm": 18.9011173248291,
      "learning_rate": 1.0651264263595845e-06,
      "loss": 4.6636,
      "step": 2655
    },
    {
      "epoch": 0.82,
      "grad_norm": 13.250627517700195,
      "learning_rate": 1.0615017841339591e-06,
      "loss": 1.9599,
      "step": 2656
    },
    {
      "epoch": 0.82,
      "grad_norm": 10.588768005371094,
      "learning_rate": 1.0578828176119018e-06,
      "loss": 4.6439,
      "step": 2657
    },
    {
      "epoch": 0.82,
      "grad_norm": 14.118270874023438,
      "learning_rate": 1.0542695302233486e-06,
      "loss": 1.6572,
      "step": 2658
    },
    {
      "epoch": 0.82,
      "grad_norm": 16.81142234802246,
      "learning_rate": 1.0506619253928547e-06,
      "loss": 3.0529,
      "step": 2659
    },
    {
      "epoch": 0.82,
      "grad_norm": 15.711450576782227,
      "learning_rate": 1.0470600065395927e-06,
      "loss": 3.9799,
      "step": 2660
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.071063041687012,
      "learning_rate": 1.0434637770773433e-06,
      "loss": 2.9833,
      "step": 2661
    },
    {
      "epoch": 0.82,
      "grad_norm": 12.90408706665039,
      "learning_rate": 1.0398732404144961e-06,
      "loss": 1.855,
      "step": 2662
    },
    {
      "epoch": 0.83,
      "grad_norm": 14.79298210144043,
      "learning_rate": 1.0362883999540434e-06,
      "loss": 3.5146,
      "step": 2663
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.10693359375,
      "learning_rate": 1.0327092590935814e-06,
      "loss": 2.9858,
      "step": 2664
    },
    {
      "epoch": 0.83,
      "grad_norm": 17.625398635864258,
      "learning_rate": 1.0291358212253015e-06,
      "loss": 5.5017,
      "step": 2665
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.117279052734375,
      "learning_rate": 1.025568089735994e-06,
      "loss": 5.5637,
      "step": 2666
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.486377716064453,
      "learning_rate": 1.0220060680070375e-06,
      "loss": 2.4685,
      "step": 2667
    },
    {
      "epoch": 0.83,
      "grad_norm": 14.307598114013672,
      "learning_rate": 1.0184497594144005e-06,
      "loss": 1.7946,
      "step": 2668
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.106240272521973,
      "learning_rate": 1.0148991673286366e-06,
      "loss": 1.846,
      "step": 2669
    },
    {
      "epoch": 0.83,
      "grad_norm": 8.210283279418945,
      "learning_rate": 1.0113542951148806e-06,
      "loss": 0.7165,
      "step": 2670
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.936012268066406,
      "learning_rate": 1.007815146132846e-06,
      "loss": 2.8782,
      "step": 2671
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.700289726257324,
      "learning_rate": 1.004281723736826e-06,
      "loss": 5.2996,
      "step": 2672
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.953660011291504,
      "learning_rate": 1.0007540312756809e-06,
      "loss": 1.2916,
      "step": 2673
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.936169624328613,
      "learning_rate": 9.972320720928423e-07,
      "loss": 1.2098,
      "step": 2674
    },
    {
      "epoch": 0.83,
      "grad_norm": 13.651212692260742,
      "learning_rate": 9.937158495263123e-07,
      "loss": 1.5848,
      "step": 2675
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.4189510345459,
      "learning_rate": 9.902053669086485e-07,
      "loss": 1.9543,
      "step": 2676
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.403413772583008,
      "learning_rate": 9.867006275669705e-07,
      "loss": 4.6017,
      "step": 2677
    },
    {
      "epoch": 0.83,
      "grad_norm": 16.291194915771484,
      "learning_rate": 9.832016348229598e-07,
      "loss": 1.5132,
      "step": 2678
    },
    {
      "epoch": 0.83,
      "grad_norm": 13.160598754882812,
      "learning_rate": 9.797083919928462e-07,
      "loss": 2.8031,
      "step": 2679
    },
    {
      "epoch": 0.83,
      "grad_norm": 12.4559907913208,
      "learning_rate": 9.76220902387411e-07,
      "loss": 2.9015,
      "step": 2680
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.56799602508545,
      "learning_rate": 9.727391693119844e-07,
      "loss": 3.8978,
      "step": 2681
    },
    {
      "epoch": 0.83,
      "grad_norm": 10.057963371276855,
      "learning_rate": 9.692631960664399e-07,
      "loss": 1.8247,
      "step": 2682
    },
    {
      "epoch": 0.83,
      "grad_norm": 22.983314514160156,
      "learning_rate": 9.657929859451903e-07,
      "loss": 6.4451,
      "step": 2683
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.508147239685059,
      "learning_rate": 9.62328542237189e-07,
      "loss": 3.1221,
      "step": 2684
    },
    {
      "epoch": 0.83,
      "grad_norm": 11.232625961303711,
      "learning_rate": 9.588698682259217e-07,
      "loss": 1.0586,
      "step": 2685
    },
    {
      "epoch": 0.83,
      "grad_norm": 22.873655319213867,
      "learning_rate": 9.554169671894061e-07,
      "loss": 3.8315,
      "step": 2686
    },
    {
      "epoch": 0.83,
      "grad_norm": 13.896849632263184,
      "learning_rate": 9.519698424001922e-07,
      "loss": 5.5813,
      "step": 2687
    },
    {
      "epoch": 0.83,
      "grad_norm": 15.250011444091797,
      "learning_rate": 9.485284971253514e-07,
      "loss": 2.6697,
      "step": 2688
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.72279167175293,
      "learning_rate": 9.450929346264781e-07,
      "loss": 1.4129,
      "step": 2689
    },
    {
      "epoch": 0.83,
      "grad_norm": 21.19035530090332,
      "learning_rate": 9.416631581596877e-07,
      "loss": 2.8044,
      "step": 2690
    },
    {
      "epoch": 0.83,
      "grad_norm": 17.99174690246582,
      "learning_rate": 9.382391709756093e-07,
      "loss": 2.5248,
      "step": 2691
    },
    {
      "epoch": 0.83,
      "grad_norm": 18.81049156188965,
      "learning_rate": 9.348209763193857e-07,
      "loss": 1.5409,
      "step": 2692
    },
    {
      "epoch": 0.83,
      "grad_norm": 9.588401794433594,
      "learning_rate": 9.314085774306725e-07,
      "loss": 2.1723,
      "step": 2693
    },
    {
      "epoch": 0.83,
      "grad_norm": 14.631446838378906,
      "learning_rate": 9.2800197754363e-07,
      "loss": 1.3401,
      "step": 2694
    },
    {
      "epoch": 0.84,
      "grad_norm": 29.79292106628418,
      "learning_rate": 9.246011798869209e-07,
      "loss": 4.9964,
      "step": 2695
    },
    {
      "epoch": 0.84,
      "grad_norm": 16.508506774902344,
      "learning_rate": 9.212061876837127e-07,
      "loss": 1.0711,
      "step": 2696
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.153032302856445,
      "learning_rate": 9.17817004151669e-07,
      "loss": 2.5525,
      "step": 2697
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.247258186340332,
      "learning_rate": 9.144336325029431e-07,
      "loss": 1.8873,
      "step": 2698
    },
    {
      "epoch": 0.84,
      "grad_norm": 12.099528312683105,
      "learning_rate": 9.110560759441885e-07,
      "loss": 1.1924,
      "step": 2699
    },
    {
      "epoch": 0.84,
      "grad_norm": 7.610776424407959,
      "learning_rate": 9.076843376765418e-07,
      "loss": 1.0963,
      "step": 2700
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.039730072021484,
      "learning_rate": 9.043184208956258e-07,
      "loss": 4.1729,
      "step": 2701
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.46630859375,
      "learning_rate": 9.009583287915495e-07,
      "loss": 4.3558,
      "step": 2702
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.001320838928223,
      "learning_rate": 8.976040645488973e-07,
      "loss": 1.4631,
      "step": 2703
    },
    {
      "epoch": 0.84,
      "grad_norm": 15.730828285217285,
      "learning_rate": 8.942556313467316e-07,
      "loss": 3.5943,
      "step": 2704
    },
    {
      "epoch": 0.84,
      "grad_norm": 16.972795486450195,
      "learning_rate": 8.909130323585884e-07,
      "loss": 4.9959,
      "step": 2705
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.503647804260254,
      "learning_rate": 8.875762707524742e-07,
      "loss": 3.0095,
      "step": 2706
    },
    {
      "epoch": 0.84,
      "grad_norm": 13.763020515441895,
      "learning_rate": 8.842453496908605e-07,
      "loss": 1.6531,
      "step": 2707
    },
    {
      "epoch": 0.84,
      "grad_norm": 19.466567993164062,
      "learning_rate": 8.809202723306905e-07,
      "loss": 2.8807,
      "step": 2708
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.884944915771484,
      "learning_rate": 8.776010418233612e-07,
      "loss": 1.0522,
      "step": 2709
    },
    {
      "epoch": 0.84,
      "grad_norm": 16.030908584594727,
      "learning_rate": 8.742876613147325e-07,
      "loss": 1.9042,
      "step": 2710
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.34588623046875,
      "learning_rate": 8.709801339451175e-07,
      "loss": 1.8041,
      "step": 2711
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.672372817993164,
      "learning_rate": 8.676784628492829e-07,
      "loss": 3.5281,
      "step": 2712
    },
    {
      "epoch": 0.84,
      "grad_norm": 16.09285545349121,
      "learning_rate": 8.643826511564443e-07,
      "loss": 5.1449,
      "step": 2713
    },
    {
      "epoch": 0.84,
      "grad_norm": 23.066343307495117,
      "learning_rate": 8.610927019902668e-07,
      "loss": 5.009,
      "step": 2714
    },
    {
      "epoch": 0.84,
      "grad_norm": 13.979827880859375,
      "learning_rate": 8.578086184688562e-07,
      "loss": 1.3423,
      "step": 2715
    },
    {
      "epoch": 0.84,
      "grad_norm": 18.696386337280273,
      "learning_rate": 8.545304037047584e-07,
      "loss": 2.1157,
      "step": 2716
    },
    {
      "epoch": 0.84,
      "grad_norm": 15.990952491760254,
      "learning_rate": 8.512580608049616e-07,
      "loss": 2.9161,
      "step": 2717
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.257905006408691,
      "learning_rate": 8.479915928708854e-07,
      "loss": 1.6421,
      "step": 2718
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.700502395629883,
      "learning_rate": 8.44731002998378e-07,
      "loss": 2.5394,
      "step": 2719
    },
    {
      "epoch": 0.84,
      "grad_norm": 15.877397537231445,
      "learning_rate": 8.414762942777231e-07,
      "loss": 2.4409,
      "step": 2720
    },
    {
      "epoch": 0.84,
      "grad_norm": 10.039705276489258,
      "learning_rate": 8.382274697936282e-07,
      "loss": 1.0026,
      "step": 2721
    },
    {
      "epoch": 0.84,
      "grad_norm": 16.870891571044922,
      "learning_rate": 8.349845326252203e-07,
      "loss": 4.0844,
      "step": 2722
    },
    {
      "epoch": 0.84,
      "grad_norm": 11.927734375,
      "learning_rate": 8.317474858460537e-07,
      "loss": 2.5039,
      "step": 2723
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.01502513885498,
      "learning_rate": 8.285163325240943e-07,
      "loss": 2.0861,
      "step": 2724
    },
    {
      "epoch": 0.84,
      "grad_norm": 14.202402114868164,
      "learning_rate": 8.252910757217231e-07,
      "loss": 3.7536,
      "step": 2725
    },
    {
      "epoch": 0.84,
      "grad_norm": 17.157291412353516,
      "learning_rate": 8.220717184957337e-07,
      "loss": 3.1874,
      "step": 2726
    },
    {
      "epoch": 0.84,
      "grad_norm": 18.783302307128906,
      "learning_rate": 8.188582638973286e-07,
      "loss": 3.0514,
      "step": 2727
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.83509063720703,
      "learning_rate": 8.156507149721133e-07,
      "loss": 5.7338,
      "step": 2728
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.39986515045166,
      "learning_rate": 8.124490747601021e-07,
      "loss": 3.11,
      "step": 2729
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.768985748291016,
      "learning_rate": 8.092533462957026e-07,
      "loss": 5.6912,
      "step": 2730
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.61513328552246,
      "learning_rate": 8.06063532607723e-07,
      "loss": 2.2014,
      "step": 2731
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.662651062011719,
      "learning_rate": 8.028796367193647e-07,
      "loss": 1.2689,
      "step": 2732
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.922401428222656,
      "learning_rate": 7.997016616482206e-07,
      "loss": 1.6796,
      "step": 2733
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.682415962219238,
      "learning_rate": 7.965296104062696e-07,
      "loss": 1.2255,
      "step": 2734
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.56028938293457,
      "learning_rate": 7.933634859998832e-07,
      "loss": 0.9541,
      "step": 2735
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.88955020904541,
      "learning_rate": 7.902032914298073e-07,
      "loss": 2.2162,
      "step": 2736
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.706653594970703,
      "learning_rate": 7.870490296911723e-07,
      "loss": 1.963,
      "step": 2737
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.079216957092285,
      "learning_rate": 7.839007037734862e-07,
      "loss": 3.8992,
      "step": 2738
    },
    {
      "epoch": 0.85,
      "grad_norm": 12.343116760253906,
      "learning_rate": 7.8075831666063e-07,
      "loss": 2.2811,
      "step": 2739
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.426636695861816,
      "learning_rate": 7.77621871330851e-07,
      "loss": 1.6619,
      "step": 2740
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.37838077545166,
      "learning_rate": 7.744913707567756e-07,
      "loss": 3.2722,
      "step": 2741
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.811568260192871,
      "learning_rate": 7.713668179053881e-07,
      "loss": 1.3035,
      "step": 2742
    },
    {
      "epoch": 0.85,
      "grad_norm": 9.749939918518066,
      "learning_rate": 7.682482157380361e-07,
      "loss": 1.1341,
      "step": 2743
    },
    {
      "epoch": 0.85,
      "grad_norm": 16.110702514648438,
      "learning_rate": 7.651355672104327e-07,
      "loss": 2.4466,
      "step": 2744
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.312182426452637,
      "learning_rate": 7.62028875272643e-07,
      "loss": 2.4678,
      "step": 2745
    },
    {
      "epoch": 0.85,
      "grad_norm": 8.107397079467773,
      "learning_rate": 7.589281428690886e-07,
      "loss": 1.0105,
      "step": 2746
    },
    {
      "epoch": 0.85,
      "grad_norm": 18.883792877197266,
      "learning_rate": 7.558333729385425e-07,
      "loss": 5.5305,
      "step": 2747
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.187335014343262,
      "learning_rate": 7.527445684141272e-07,
      "loss": 2.1495,
      "step": 2748
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.755721092224121,
      "learning_rate": 7.496617322233088e-07,
      "loss": 4.6224,
      "step": 2749
    },
    {
      "epoch": 0.85,
      "grad_norm": 18.182004928588867,
      "learning_rate": 7.465848672879036e-07,
      "loss": 3.0456,
      "step": 2750
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.412226676940918,
      "learning_rate": 7.435139765240602e-07,
      "loss": 2.3355,
      "step": 2751
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.734976768493652,
      "learning_rate": 7.404490628422693e-07,
      "loss": 2.1919,
      "step": 2752
    },
    {
      "epoch": 0.85,
      "grad_norm": 18.193538665771484,
      "learning_rate": 7.373901291473594e-07,
      "loss": 5.7131,
      "step": 2753
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.647350311279297,
      "learning_rate": 7.343371783384843e-07,
      "loss": 2.2852,
      "step": 2754
    },
    {
      "epoch": 0.85,
      "grad_norm": 15.154494285583496,
      "learning_rate": 7.312902133091318e-07,
      "loss": 3.8039,
      "step": 2755
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.07168960571289,
      "learning_rate": 7.282492369471184e-07,
      "loss": 4.9915,
      "step": 2756
    },
    {
      "epoch": 0.85,
      "grad_norm": 13.386819839477539,
      "learning_rate": 7.252142521345818e-07,
      "loss": 1.7995,
      "step": 2757
    },
    {
      "epoch": 0.85,
      "grad_norm": 22.815553665161133,
      "learning_rate": 7.221852617479788e-07,
      "loss": 4.8514,
      "step": 2758
    },
    {
      "epoch": 0.85,
      "grad_norm": 11.641252517700195,
      "learning_rate": 7.191622686580927e-07,
      "loss": 0.8264,
      "step": 2759
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.47198486328125,
      "learning_rate": 7.161452757300145e-07,
      "loss": 0.7502,
      "step": 2760
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.305670738220215,
      "learning_rate": 7.131342858231538e-07,
      "loss": 1.82,
      "step": 2761
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.490869522094727,
      "learning_rate": 7.101293017912268e-07,
      "loss": 2.5695,
      "step": 2762
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.534713745117188,
      "learning_rate": 7.071303264822608e-07,
      "loss": 1.915,
      "step": 2763
    },
    {
      "epoch": 0.86,
      "grad_norm": 15.01233959197998,
      "learning_rate": 7.041373627385852e-07,
      "loss": 2.5617,
      "step": 2764
    },
    {
      "epoch": 0.86,
      "grad_norm": 17.969585418701172,
      "learning_rate": 7.011504133968359e-07,
      "loss": 2.0232,
      "step": 2765
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.038057327270508,
      "learning_rate": 6.981694812879459e-07,
      "loss": 1.5787,
      "step": 2766
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.305198669433594,
      "learning_rate": 6.951945692371438e-07,
      "loss": 1.3824,
      "step": 2767
    },
    {
      "epoch": 0.86,
      "grad_norm": 9.185097694396973,
      "learning_rate": 6.92225680063956e-07,
      "loss": 0.9478,
      "step": 2768
    },
    {
      "epoch": 0.86,
      "grad_norm": 15.012513160705566,
      "learning_rate": 6.892628165821988e-07,
      "loss": 3.1556,
      "step": 2769
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.005261421203613,
      "learning_rate": 6.863059815999756e-07,
      "loss": 1.8806,
      "step": 2770
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.431744575500488,
      "learning_rate": 6.83355177919681e-07,
      "loss": 1.5041,
      "step": 2771
    },
    {
      "epoch": 0.86,
      "grad_norm": 13.604820251464844,
      "learning_rate": 6.804104083379899e-07,
      "loss": 1.8549,
      "step": 2772
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.08254051208496,
      "learning_rate": 6.774716756458579e-07,
      "loss": 3.1237,
      "step": 2773
    },
    {
      "epoch": 0.86,
      "grad_norm": 24.12819480895996,
      "learning_rate": 6.745389826285227e-07,
      "loss": 2.3856,
      "step": 2774
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.128646850585938,
      "learning_rate": 6.716123320654928e-07,
      "loss": 3.163,
      "step": 2775
    },
    {
      "epoch": 0.86,
      "grad_norm": 8.595212936401367,
      "learning_rate": 6.686917267305512e-07,
      "loss": 0.8763,
      "step": 2776
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.716238975524902,
      "learning_rate": 6.657771693917553e-07,
      "loss": 1.8973,
      "step": 2777
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.591482162475586,
      "learning_rate": 6.628686628114276e-07,
      "loss": 1.7251,
      "step": 2778
    },
    {
      "epoch": 0.86,
      "grad_norm": 19.033681869506836,
      "learning_rate": 6.599662097461526e-07,
      "loss": 3.174,
      "step": 2779
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.152606010437012,
      "learning_rate": 6.570698129467846e-07,
      "loss": 2.4005,
      "step": 2780
    },
    {
      "epoch": 0.86,
      "grad_norm": 10.492902755737305,
      "learning_rate": 6.54179475158433e-07,
      "loss": 0.7674,
      "step": 2781
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.48991870880127,
      "learning_rate": 6.512951991204649e-07,
      "loss": 1.5215,
      "step": 2782
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.252331733703613,
      "learning_rate": 6.484169875665041e-07,
      "loss": 1.5332,
      "step": 2783
    },
    {
      "epoch": 0.86,
      "grad_norm": 12.655040740966797,
      "learning_rate": 6.455448432244255e-07,
      "loss": 2.0131,
      "step": 2784
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.361101150512695,
      "learning_rate": 6.426787688163527e-07,
      "loss": 3.2265,
      "step": 2785
    },
    {
      "epoch": 0.86,
      "grad_norm": 18.582107543945312,
      "learning_rate": 6.3981876705866e-07,
      "loss": 3.6963,
      "step": 2786
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.305501937866211,
      "learning_rate": 6.36964840661962e-07,
      "loss": 1.7268,
      "step": 2787
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.70504093170166,
      "learning_rate": 6.341169923311187e-07,
      "loss": 1.5501,
      "step": 2788
    },
    {
      "epoch": 0.86,
      "grad_norm": 11.306252479553223,
      "learning_rate": 6.312752247652273e-07,
      "loss": 3.1718,
      "step": 2789
    },
    {
      "epoch": 0.86,
      "grad_norm": 16.00751495361328,
      "learning_rate": 6.284395406576224e-07,
      "loss": 2.9333,
      "step": 2790
    },
    {
      "epoch": 0.86,
      "grad_norm": 14.812768936157227,
      "learning_rate": 6.25609942695871e-07,
      "loss": 4.0762,
      "step": 2791
    },
    {
      "epoch": 0.87,
      "grad_norm": 18.853811264038086,
      "learning_rate": 6.227864335617784e-07,
      "loss": 3.1675,
      "step": 2792
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.854783058166504,
      "learning_rate": 6.199690159313715e-07,
      "loss": 1.1855,
      "step": 2793
    },
    {
      "epoch": 0.87,
      "grad_norm": 14.672539710998535,
      "learning_rate": 6.171576924749069e-07,
      "loss": 1.5224,
      "step": 2794
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.445911407470703,
      "learning_rate": 6.143524658568685e-07,
      "loss": 1.1172,
      "step": 2795
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.489899635314941,
      "learning_rate": 6.11553338735955e-07,
      "loss": 2.6757,
      "step": 2796
    },
    {
      "epoch": 0.87,
      "grad_norm": 21.521224975585938,
      "learning_rate": 6.087603137650882e-07,
      "loss": 3.4554,
      "step": 2797
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.998003005981445,
      "learning_rate": 6.059733935914089e-07,
      "loss": 4.7792,
      "step": 2798
    },
    {
      "epoch": 0.87,
      "grad_norm": 17.415874481201172,
      "learning_rate": 6.03192580856267e-07,
      "loss": 4.8494,
      "step": 2799
    },
    {
      "epoch": 0.87,
      "grad_norm": 16.41072654724121,
      "learning_rate": 6.004178781952261e-07,
      "loss": 4.1986,
      "step": 2800
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.820795059204102,
      "learning_rate": 5.976492882380614e-07,
      "loss": 1.0387,
      "step": 2801
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.374963760375977,
      "learning_rate": 5.948868136087504e-07,
      "loss": 1.26,
      "step": 2802
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.477418899536133,
      "learning_rate": 5.921304569254768e-07,
      "loss": 5.2774,
      "step": 2803
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.892913818359375,
      "learning_rate": 5.893802208006254e-07,
      "loss": 1.2195,
      "step": 2804
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.48777961730957,
      "learning_rate": 5.866361078407806e-07,
      "loss": 2.0262,
      "step": 2805
    },
    {
      "epoch": 0.87,
      "grad_norm": 14.145560264587402,
      "learning_rate": 5.83898120646724e-07,
      "loss": 2.3832,
      "step": 2806
    },
    {
      "epoch": 0.87,
      "grad_norm": 16.891164779663086,
      "learning_rate": 5.811662618134309e-07,
      "loss": 1.7395,
      "step": 2807
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.735100746154785,
      "learning_rate": 5.784405339300669e-07,
      "loss": 2.7466,
      "step": 2808
    },
    {
      "epoch": 0.87,
      "grad_norm": 9.554878234863281,
      "learning_rate": 5.757209395799886e-07,
      "loss": 1.7883,
      "step": 2809
    },
    {
      "epoch": 0.87,
      "grad_norm": 15.633171081542969,
      "learning_rate": 5.730074813407394e-07,
      "loss": 1.6478,
      "step": 2810
    },
    {
      "epoch": 0.87,
      "grad_norm": 17.242467880249023,
      "learning_rate": 5.703001617840447e-07,
      "loss": 1.5473,
      "step": 2811
    },
    {
      "epoch": 0.87,
      "grad_norm": 11.249731063842773,
      "learning_rate": 5.675989834758162e-07,
      "loss": 0.9579,
      "step": 2812
    },
    {
      "epoch": 0.87,
      "grad_norm": 19.327224731445312,
      "learning_rate": 5.649039489761428e-07,
      "loss": 3.9334,
      "step": 2813
    },
    {
      "epoch": 0.87,
      "grad_norm": 18.37837028503418,
      "learning_rate": 5.622150608392878e-07,
      "loss": 5.0965,
      "step": 2814
    },
    {
      "epoch": 0.87,
      "grad_norm": 17.791837692260742,
      "learning_rate": 5.595323216136943e-07,
      "loss": 3.4107,
      "step": 2815
    },
    {
      "epoch": 0.87,
      "grad_norm": 10.968727111816406,
      "learning_rate": 5.568557338419748e-07,
      "loss": 0.996,
      "step": 2816
    },
    {
      "epoch": 0.87,
      "grad_norm": 18.726015090942383,
      "learning_rate": 5.54185300060912e-07,
      "loss": 4.8876,
      "step": 2817
    },
    {
      "epoch": 0.87,
      "grad_norm": 8.49133586883545,
      "learning_rate": 5.515210228014545e-07,
      "loss": 1.0867,
      "step": 2818
    },
    {
      "epoch": 0.87,
      "grad_norm": 14.631171226501465,
      "learning_rate": 5.488629045887192e-07,
      "loss": 1.7388,
      "step": 2819
    },
    {
      "epoch": 0.87,
      "grad_norm": 16.90894317626953,
      "learning_rate": 5.462109479419823e-07,
      "loss": 1.7231,
      "step": 2820
    },
    {
      "epoch": 0.87,
      "grad_norm": 18.553396224975586,
      "learning_rate": 5.435651553746843e-07,
      "loss": 1.7818,
      "step": 2821
    },
    {
      "epoch": 0.87,
      "grad_norm": 13.470282554626465,
      "learning_rate": 5.409255293944203e-07,
      "loss": 2.7616,
      "step": 2822
    },
    {
      "epoch": 0.87,
      "grad_norm": 6.711654186248779,
      "learning_rate": 5.382920725029411e-07,
      "loss": 0.6761,
      "step": 2823
    },
    {
      "epoch": 0.87,
      "grad_norm": 15.698296546936035,
      "learning_rate": 5.356647871961512e-07,
      "loss": 3.7117,
      "step": 2824
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.690383911132812,
      "learning_rate": 5.330436759641078e-07,
      "loss": 5.3394,
      "step": 2825
    },
    {
      "epoch": 0.88,
      "grad_norm": 16.919147491455078,
      "learning_rate": 5.304287412910115e-07,
      "loss": 2.627,
      "step": 2826
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.312337875366211,
      "learning_rate": 5.278199856552158e-07,
      "loss": 1.9178,
      "step": 2827
    },
    {
      "epoch": 0.88,
      "grad_norm": 14.843708992004395,
      "learning_rate": 5.252174115292121e-07,
      "loss": 2.1143,
      "step": 2828
    },
    {
      "epoch": 0.88,
      "grad_norm": 19.56746482849121,
      "learning_rate": 5.226210213796357e-07,
      "loss": 5.143,
      "step": 2829
    },
    {
      "epoch": 0.88,
      "grad_norm": 17.91864013671875,
      "learning_rate": 5.200308176672634e-07,
      "loss": 5.6695,
      "step": 2830
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.32984447479248,
      "learning_rate": 5.174468028470023e-07,
      "loss": 1.4222,
      "step": 2831
    },
    {
      "epoch": 0.88,
      "grad_norm": 18.887392044067383,
      "learning_rate": 5.148689793678974e-07,
      "loss": 1.6787,
      "step": 2832
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.73823356628418,
      "learning_rate": 5.122973496731283e-07,
      "loss": 2.3282,
      "step": 2833
    },
    {
      "epoch": 0.88,
      "grad_norm": 19.994966506958008,
      "learning_rate": 5.097319162000009e-07,
      "loss": 2.7318,
      "step": 2834
    },
    {
      "epoch": 0.88,
      "grad_norm": 13.204672813415527,
      "learning_rate": 5.071726813799483e-07,
      "loss": 1.5666,
      "step": 2835
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.245928764343262,
      "learning_rate": 5.046196476385316e-07,
      "loss": 1.7705,
      "step": 2836
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.357982635498047,
      "learning_rate": 5.020728173954322e-07,
      "loss": 5.0124,
      "step": 2837
    },
    {
      "epoch": 0.88,
      "grad_norm": 26.01468276977539,
      "learning_rate": 4.995321930644537e-07,
      "loss": 4.5425,
      "step": 2838
    },
    {
      "epoch": 0.88,
      "grad_norm": 14.262585639953613,
      "learning_rate": 4.969977770535164e-07,
      "loss": 2.3728,
      "step": 2839
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.951447486877441,
      "learning_rate": 4.944695717646566e-07,
      "loss": 2.1417,
      "step": 2840
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.154837608337402,
      "learning_rate": 4.919475795940252e-07,
      "loss": 1.6735,
      "step": 2841
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.00161361694336,
      "learning_rate": 4.894318029318848e-07,
      "loss": 2.6789,
      "step": 2842
    },
    {
      "epoch": 0.88,
      "grad_norm": 13.821235656738281,
      "learning_rate": 4.869222441626074e-07,
      "loss": 5.2744,
      "step": 2843
    },
    {
      "epoch": 0.88,
      "grad_norm": 19.3443660736084,
      "learning_rate": 4.844189056646698e-07,
      "loss": 2.4512,
      "step": 2844
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.478118896484375,
      "learning_rate": 4.819217898106546e-07,
      "loss": 1.2044,
      "step": 2845
    },
    {
      "epoch": 0.88,
      "grad_norm": 11.809135437011719,
      "learning_rate": 4.794308989672472e-07,
      "loss": 2.4362,
      "step": 2846
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.744396209716797,
      "learning_rate": 4.769462354952327e-07,
      "loss": 0.7586,
      "step": 2847
    },
    {
      "epoch": 0.88,
      "grad_norm": 17.750654220581055,
      "learning_rate": 4.744678017494944e-07,
      "loss": 1.5162,
      "step": 2848
    },
    {
      "epoch": 0.88,
      "grad_norm": 13.151311874389648,
      "learning_rate": 4.719956000790113e-07,
      "loss": 1.59,
      "step": 2849
    },
    {
      "epoch": 0.88,
      "grad_norm": 12.47098445892334,
      "learning_rate": 4.6952963282685497e-07,
      "loss": 1.8405,
      "step": 2850
    },
    {
      "epoch": 0.88,
      "grad_norm": 16.631267547607422,
      "learning_rate": 4.6706990233019073e-07,
      "loss": 2.7837,
      "step": 2851
    },
    {
      "epoch": 0.88,
      "grad_norm": 9.20441722869873,
      "learning_rate": 4.646164109202683e-07,
      "loss": 1.9357,
      "step": 2852
    },
    {
      "epoch": 0.88,
      "grad_norm": 14.25259017944336,
      "learning_rate": 4.621691609224279e-07,
      "loss": 3.417,
      "step": 2853
    },
    {
      "epoch": 0.88,
      "grad_norm": 15.585476875305176,
      "learning_rate": 4.5972815465609407e-07,
      "loss": 1.8087,
      "step": 2854
    },
    {
      "epoch": 0.88,
      "grad_norm": 21.953487396240234,
      "learning_rate": 4.572933944347732e-07,
      "loss": 3.7204,
      "step": 2855
    },
    {
      "epoch": 0.88,
      "grad_norm": 10.49705696105957,
      "learning_rate": 4.548648825660498e-07,
      "loss": 1.5159,
      "step": 2856
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.401366233825684,
      "learning_rate": 4.524426213515912e-07,
      "loss": 1.3898,
      "step": 2857
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.927726745605469,
      "learning_rate": 4.5002661308713556e-07,
      "loss": 5.4374,
      "step": 2858
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.92420482635498,
      "learning_rate": 4.4761686006249766e-07,
      "loss": 2.0195,
      "step": 2859
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.106597900390625,
      "learning_rate": 4.452133645615623e-07,
      "loss": 4.3874,
      "step": 2860
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.929544448852539,
      "learning_rate": 4.428161288622855e-07,
      "loss": 2.0026,
      "step": 2861
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.906898498535156,
      "learning_rate": 4.4042515523668703e-07,
      "loss": 5.4916,
      "step": 2862
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.238280296325684,
      "learning_rate": 4.380404459508562e-07,
      "loss": 4.5458,
      "step": 2863
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.196687698364258,
      "learning_rate": 4.3566200326494157e-07,
      "loss": 1.3555,
      "step": 2864
    },
    {
      "epoch": 0.89,
      "grad_norm": 19.44993782043457,
      "learning_rate": 4.332898294331532e-07,
      "loss": 2.3461,
      "step": 2865
    },
    {
      "epoch": 0.89,
      "grad_norm": 22.673126220703125,
      "learning_rate": 4.309239267037614e-07,
      "loss": 5.4574,
      "step": 2866
    },
    {
      "epoch": 0.89,
      "grad_norm": 17.445295333862305,
      "learning_rate": 4.285642973190916e-07,
      "loss": 5.0254,
      "step": 2867
    },
    {
      "epoch": 0.89,
      "grad_norm": 15.001561164855957,
      "learning_rate": 4.262109435155215e-07,
      "loss": 2.0039,
      "step": 2868
    },
    {
      "epoch": 0.89,
      "grad_norm": 13.708137512207031,
      "learning_rate": 4.2386386752348646e-07,
      "loss": 2.3565,
      "step": 2869
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.443163871765137,
      "learning_rate": 4.2152307156746634e-07,
      "loss": 1.2412,
      "step": 2870
    },
    {
      "epoch": 0.89,
      "grad_norm": 22.866369247436523,
      "learning_rate": 4.191885578659916e-07,
      "loss": 5.0355,
      "step": 2871
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.499114990234375,
      "learning_rate": 4.1686032863164085e-07,
      "loss": 3.6915,
      "step": 2872
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.884692192077637,
      "learning_rate": 4.145383860710302e-07,
      "loss": 1.6315,
      "step": 2873
    },
    {
      "epoch": 0.89,
      "grad_norm": 12.003686904907227,
      "learning_rate": 4.122227323848225e-07,
      "loss": 2.196,
      "step": 2874
    },
    {
      "epoch": 0.89,
      "grad_norm": 15.553420066833496,
      "learning_rate": 4.099133697677193e-07,
      "loss": 5.855,
      "step": 2875
    },
    {
      "epoch": 0.89,
      "grad_norm": 13.435556411743164,
      "learning_rate": 4.0761030040845973e-07,
      "loss": 1.3082,
      "step": 2876
    },
    {
      "epoch": 0.89,
      "grad_norm": 26.332712173461914,
      "learning_rate": 4.05313526489817e-07,
      "loss": 2.9261,
      "step": 2877
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.73967933654785,
      "learning_rate": 4.030230501885993e-07,
      "loss": 5.3095,
      "step": 2878
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.076337814331055,
      "learning_rate": 4.0073887367564516e-07,
      "loss": 7.312,
      "step": 2879
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.803072929382324,
      "learning_rate": 3.984609991158226e-07,
      "loss": 3.0275,
      "step": 2880
    },
    {
      "epoch": 0.89,
      "grad_norm": 10.968694686889648,
      "learning_rate": 3.9618942866802756e-07,
      "loss": 1.1406,
      "step": 2881
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.786733627319336,
      "learning_rate": 3.9392416448517924e-07,
      "loss": 1.3895,
      "step": 2882
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.944122314453125,
      "learning_rate": 3.916652087142217e-07,
      "loss": 2.8045,
      "step": 2883
    },
    {
      "epoch": 0.89,
      "grad_norm": 17.218578338623047,
      "learning_rate": 3.894125634961199e-07,
      "loss": 1.7795,
      "step": 2884
    },
    {
      "epoch": 0.89,
      "grad_norm": 11.743535041809082,
      "learning_rate": 3.8716623096585873e-07,
      "loss": 1.3452,
      "step": 2885
    },
    {
      "epoch": 0.89,
      "grad_norm": 9.584161758422852,
      "learning_rate": 3.849262132524373e-07,
      "loss": 1.0609,
      "step": 2886
    },
    {
      "epoch": 0.89,
      "grad_norm": 14.182889938354492,
      "learning_rate": 3.8269251247887283e-07,
      "loss": 2.6492,
      "step": 2887
    },
    {
      "epoch": 0.89,
      "grad_norm": 16.427518844604492,
      "learning_rate": 3.8046513076219284e-07,
      "loss": 4.0594,
      "step": 2888
    },
    {
      "epoch": 0.9,
      "grad_norm": 19.689199447631836,
      "learning_rate": 3.78244070213437e-07,
      "loss": 4.664,
      "step": 2889
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.083978652954102,
      "learning_rate": 3.7602933293765565e-07,
      "loss": 3.2096,
      "step": 2890
    },
    {
      "epoch": 0.9,
      "grad_norm": 14.33427906036377,
      "learning_rate": 3.738209210339037e-07,
      "loss": 3.754,
      "step": 2891
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.477331161499023,
      "learning_rate": 3.716188365952413e-07,
      "loss": 1.8141,
      "step": 2892
    },
    {
      "epoch": 0.9,
      "grad_norm": 17.20378875732422,
      "learning_rate": 3.694230817087345e-07,
      "loss": 4.5588,
      "step": 2893
    },
    {
      "epoch": 0.9,
      "grad_norm": 15.989679336547852,
      "learning_rate": 3.6723365845544847e-07,
      "loss": 5.9715,
      "step": 2894
    },
    {
      "epoch": 0.9,
      "grad_norm": 14.15831470489502,
      "learning_rate": 3.6505056891044337e-07,
      "loss": 1.9312,
      "step": 2895
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.890592575073242,
      "learning_rate": 3.6287381514278386e-07,
      "loss": 1.6506,
      "step": 2896
    },
    {
      "epoch": 0.9,
      "grad_norm": 17.143720626831055,
      "learning_rate": 3.6070339921552503e-07,
      "loss": 3.1037,
      "step": 2897
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.737471580505371,
      "learning_rate": 3.585393231857153e-07,
      "loss": 1.0532,
      "step": 2898
    },
    {
      "epoch": 0.9,
      "grad_norm": 14.59965991973877,
      "learning_rate": 3.5638158910439836e-07,
      "loss": 2.3546,
      "step": 2899
    },
    {
      "epoch": 0.9,
      "grad_norm": 13.561890602111816,
      "learning_rate": 3.542301990166027e-07,
      "loss": 2.0995,
      "step": 2900
    },
    {
      "epoch": 0.9,
      "grad_norm": 16.774505615234375,
      "learning_rate": 3.5208515496134557e-07,
      "loss": 1.2884,
      "step": 2901
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.472853660583496,
      "learning_rate": 3.4994645897163075e-07,
      "loss": 1.6445,
      "step": 2902
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.874034881591797,
      "learning_rate": 3.4781411307444306e-07,
      "loss": 1.0967,
      "step": 2903
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.599327087402344,
      "learning_rate": 3.456881192907505e-07,
      "loss": 1.184,
      "step": 2904
    },
    {
      "epoch": 0.9,
      "grad_norm": 13.708553314208984,
      "learning_rate": 3.435684796355029e-07,
      "loss": 1.7495,
      "step": 2905
    },
    {
      "epoch": 0.9,
      "grad_norm": 14.182079315185547,
      "learning_rate": 3.4145519611762485e-07,
      "loss": 2.5004,
      "step": 2906
    },
    {
      "epoch": 0.9,
      "grad_norm": 16.279220581054688,
      "learning_rate": 3.393482707400165e-07,
      "loss": 2.1366,
      "step": 2907
    },
    {
      "epoch": 0.9,
      "grad_norm": 15.938751220703125,
      "learning_rate": 3.37247705499554e-07,
      "loss": 3.5088,
      "step": 2908
    },
    {
      "epoch": 0.9,
      "grad_norm": 13.326894760131836,
      "learning_rate": 3.3515350238708466e-07,
      "loss": 2.4735,
      "step": 2909
    },
    {
      "epoch": 0.9,
      "grad_norm": 17.642976760864258,
      "learning_rate": 3.330656633874248e-07,
      "loss": 3.1495,
      "step": 2910
    },
    {
      "epoch": 0.9,
      "grad_norm": 18.092041015625,
      "learning_rate": 3.309841904793617e-07,
      "loss": 3.6098,
      "step": 2911
    },
    {
      "epoch": 0.9,
      "grad_norm": 10.679511070251465,
      "learning_rate": 3.2890908563564784e-07,
      "loss": 2.8008,
      "step": 2912
    },
    {
      "epoch": 0.9,
      "grad_norm": 9.876250267028809,
      "learning_rate": 3.268403508229979e-07,
      "loss": 0.7715,
      "step": 2913
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.585461616516113,
      "learning_rate": 3.247779880020936e-07,
      "loss": 1.6819,
      "step": 2914
    },
    {
      "epoch": 0.9,
      "grad_norm": 23.991846084594727,
      "learning_rate": 3.2272199912757546e-07,
      "loss": 3.7919,
      "step": 2915
    },
    {
      "epoch": 0.9,
      "grad_norm": 15.386876106262207,
      "learning_rate": 3.206723861480395e-07,
      "loss": 1.9322,
      "step": 2916
    },
    {
      "epoch": 0.9,
      "grad_norm": 12.495972633361816,
      "learning_rate": 3.18629151006046e-07,
      "loss": 1.1509,
      "step": 2917
    },
    {
      "epoch": 0.9,
      "grad_norm": 20.994876861572266,
      "learning_rate": 3.1659229563810424e-07,
      "loss": 4.6183,
      "step": 2918
    },
    {
      "epoch": 0.9,
      "grad_norm": 18.07956886291504,
      "learning_rate": 3.145618219746802e-07,
      "loss": 2.7403,
      "step": 2919
    },
    {
      "epoch": 0.9,
      "grad_norm": 11.444549560546875,
      "learning_rate": 3.1253773194019116e-07,
      "loss": 3.1144,
      "step": 2920
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.558086395263672,
      "learning_rate": 3.105200274530024e-07,
      "loss": 6.6997,
      "step": 2921
    },
    {
      "epoch": 0.91,
      "grad_norm": 17.11377716064453,
      "learning_rate": 3.085087104254305e-07,
      "loss": 5.6643,
      "step": 2922
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.213759422302246,
      "learning_rate": 3.065037827637346e-07,
      "loss": 1.05,
      "step": 2923
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.40109634399414,
      "learning_rate": 3.0450524636812055e-07,
      "loss": 1.237,
      "step": 2924
    },
    {
      "epoch": 0.91,
      "grad_norm": 19.069467544555664,
      "learning_rate": 3.0251310313273436e-07,
      "loss": 2.2991,
      "step": 2925
    },
    {
      "epoch": 0.91,
      "grad_norm": 15.392562866210938,
      "learning_rate": 3.0052735494566775e-07,
      "loss": 2.6311,
      "step": 2926
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.271350860595703,
      "learning_rate": 2.985480036889458e-07,
      "loss": 2.2844,
      "step": 2927
    },
    {
      "epoch": 0.91,
      "grad_norm": 18.259136199951172,
      "learning_rate": 2.9657505123853376e-07,
      "loss": 2.5049,
      "step": 2928
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.51673698425293,
      "learning_rate": 2.9460849946433253e-07,
      "loss": 1.6506,
      "step": 2929
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.291834831237793,
      "learning_rate": 2.926483502301739e-07,
      "loss": 1.4081,
      "step": 2930
    },
    {
      "epoch": 0.91,
      "grad_norm": 17.517236709594727,
      "learning_rate": 2.906946053938244e-07,
      "loss": 2.5833,
      "step": 2931
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.520174980163574,
      "learning_rate": 2.887472668069808e-07,
      "loss": 2.3661,
      "step": 2932
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.30176830291748,
      "learning_rate": 2.868063363152659e-07,
      "loss": 2.445,
      "step": 2933
    },
    {
      "epoch": 0.91,
      "grad_norm": 18.104732513427734,
      "learning_rate": 2.8487181575823033e-07,
      "loss": 2.9647,
      "step": 2934
    },
    {
      "epoch": 0.91,
      "grad_norm": 17.55752182006836,
      "learning_rate": 2.829437069693509e-07,
      "loss": 5.7919,
      "step": 2935
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.471856117248535,
      "learning_rate": 2.8102201177602596e-07,
      "loss": 2.0401,
      "step": 2936
    },
    {
      "epoch": 0.91,
      "grad_norm": 8.837617874145508,
      "learning_rate": 2.7910673199957454e-07,
      "loss": 1.6894,
      "step": 2937
    },
    {
      "epoch": 0.91,
      "grad_norm": 15.523258209228516,
      "learning_rate": 2.7719786945523714e-07,
      "loss": 3.0205,
      "step": 2938
    },
    {
      "epoch": 0.91,
      "grad_norm": 15.296859741210938,
      "learning_rate": 2.752954259521719e-07,
      "loss": 3.5861,
      "step": 2939
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.045928955078125,
      "learning_rate": 2.733994032934521e-07,
      "loss": 3.6416,
      "step": 2940
    },
    {
      "epoch": 0.91,
      "grad_norm": 17.02593994140625,
      "learning_rate": 2.7150980327606636e-07,
      "loss": 3.4657,
      "step": 2941
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.85177993774414,
      "learning_rate": 2.6962662769091695e-07,
      "loss": 1.2278,
      "step": 2942
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.752093315124512,
      "learning_rate": 2.677498783228158e-07,
      "loss": 1.7542,
      "step": 2943
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.72518539428711,
      "learning_rate": 2.658795569504847e-07,
      "loss": 2.2336,
      "step": 2944
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.09422492980957,
      "learning_rate": 2.6401566534655284e-07,
      "loss": 1.113,
      "step": 2945
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.671250343322754,
      "learning_rate": 2.6215820527755593e-07,
      "loss": 2.0924,
      "step": 2946
    },
    {
      "epoch": 0.91,
      "grad_norm": 14.427495956420898,
      "learning_rate": 2.6030717850393565e-07,
      "loss": 4.8395,
      "step": 2947
    },
    {
      "epoch": 0.91,
      "grad_norm": 9.298993110656738,
      "learning_rate": 2.5846258678003323e-07,
      "loss": 0.7284,
      "step": 2948
    },
    {
      "epoch": 0.91,
      "grad_norm": 12.819926261901855,
      "learning_rate": 2.566244318540941e-07,
      "loss": 2.285,
      "step": 2949
    },
    {
      "epoch": 0.91,
      "grad_norm": 10.828763961791992,
      "learning_rate": 2.5479271546826016e-07,
      "loss": 1.5147,
      "step": 2950
    },
    {
      "epoch": 0.91,
      "grad_norm": 13.688743591308594,
      "learning_rate": 2.5296743935857376e-07,
      "loss": 2.0614,
      "step": 2951
    },
    {
      "epoch": 0.91,
      "grad_norm": 18.027328491210938,
      "learning_rate": 2.511486052549705e-07,
      "loss": 5.5247,
      "step": 2952
    },
    {
      "epoch": 0.91,
      "grad_norm": 11.425618171691895,
      "learning_rate": 2.4933621488128485e-07,
      "loss": 2.7768,
      "step": 2953
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.18440055847168,
      "learning_rate": 2.4753026995523893e-07,
      "loss": 1.2978,
      "step": 2954
    },
    {
      "epoch": 0.92,
      "grad_norm": 15.074543952941895,
      "learning_rate": 2.4573077218844843e-07,
      "loss": 2.2961,
      "step": 2955
    },
    {
      "epoch": 0.92,
      "grad_norm": 22.014270782470703,
      "learning_rate": 2.439377232864206e-07,
      "loss": 2.1697,
      "step": 2956
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.891254425048828,
      "learning_rate": 2.4215112494854824e-07,
      "loss": 2.7444,
      "step": 2957
    },
    {
      "epoch": 0.92,
      "grad_norm": 14.182608604431152,
      "learning_rate": 2.4037097886810964e-07,
      "loss": 2.7952,
      "step": 2958
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.573148727416992,
      "learning_rate": 2.3859728673227103e-07,
      "loss": 2.1803,
      "step": 2959
    },
    {
      "epoch": 0.92,
      "grad_norm": 9.943724632263184,
      "learning_rate": 2.3683005022207926e-07,
      "loss": 1.5011,
      "step": 2960
    },
    {
      "epoch": 0.92,
      "grad_norm": 20.083030700683594,
      "learning_rate": 2.3506927101246287e-07,
      "loss": 2.7564,
      "step": 2961
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.7942533493042,
      "learning_rate": 2.3331495077223274e-07,
      "loss": 4.9782,
      "step": 2962
    },
    {
      "epoch": 0.92,
      "grad_norm": 20.043697357177734,
      "learning_rate": 2.3156709116407503e-07,
      "loss": 5.3203,
      "step": 2963
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.594661712646484,
      "learning_rate": 2.2982569384455443e-07,
      "loss": 6.6035,
      "step": 2964
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.41570281982422,
      "learning_rate": 2.2809076046411082e-07,
      "loss": 2.6747,
      "step": 2965
    },
    {
      "epoch": 0.92,
      "grad_norm": 10.65675163269043,
      "learning_rate": 2.2636229266705798e-07,
      "loss": 1.5477,
      "step": 2966
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.77856731414795,
      "learning_rate": 2.2464029209158018e-07,
      "loss": 1.2037,
      "step": 2967
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.271510124206543,
      "learning_rate": 2.2292476036973393e-07,
      "loss": 1.5294,
      "step": 2968
    },
    {
      "epoch": 0.92,
      "grad_norm": 14.198921203613281,
      "learning_rate": 2.212156991274456e-07,
      "loss": 3.7179,
      "step": 2969
    },
    {
      "epoch": 0.92,
      "grad_norm": 17.764951705932617,
      "learning_rate": 2.1951310998450583e-07,
      "loss": 2.4399,
      "step": 2970
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.994083404541016,
      "learning_rate": 2.1781699455457524e-07,
      "loss": 4.2348,
      "step": 2971
    },
    {
      "epoch": 0.92,
      "grad_norm": 18.061304092407227,
      "learning_rate": 2.1612735444517553e-07,
      "loss": 3.7625,
      "step": 2972
    },
    {
      "epoch": 0.92,
      "grad_norm": 12.471445083618164,
      "learning_rate": 2.1444419125769131e-07,
      "loss": 2.0589,
      "step": 2973
    },
    {
      "epoch": 0.92,
      "grad_norm": 14.454485893249512,
      "learning_rate": 2.1276750658737224e-07,
      "loss": 1.3984,
      "step": 2974
    },
    {
      "epoch": 0.92,
      "grad_norm": 22.4586181640625,
      "learning_rate": 2.110973020233238e-07,
      "loss": 1.4389,
      "step": 2975
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.39283561706543,
      "learning_rate": 2.0943357914851265e-07,
      "loss": 5.4744,
      "step": 2976
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.580299377441406,
      "learning_rate": 2.077763395397604e-07,
      "loss": 1.9475,
      "step": 2977
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.463586807250977,
      "learning_rate": 2.0612558476774527e-07,
      "loss": 2.557,
      "step": 2978
    },
    {
      "epoch": 0.92,
      "grad_norm": 22.39612579345703,
      "learning_rate": 2.0448131639699801e-07,
      "loss": 4.7626,
      "step": 2979
    },
    {
      "epoch": 0.92,
      "grad_norm": 17.558107376098633,
      "learning_rate": 2.0284353598590362e-07,
      "loss": 3.7814,
      "step": 2980
    },
    {
      "epoch": 0.92,
      "grad_norm": 20.783864974975586,
      "learning_rate": 2.0121224508669576e-07,
      "loss": 3.2047,
      "step": 2981
    },
    {
      "epoch": 0.92,
      "grad_norm": 16.20199203491211,
      "learning_rate": 1.995874452454607e-07,
      "loss": 2.1022,
      "step": 2982
    },
    {
      "epoch": 0.92,
      "grad_norm": 22.333803176879883,
      "learning_rate": 1.9796913800212956e-07,
      "loss": 2.7157,
      "step": 2983
    },
    {
      "epoch": 0.92,
      "grad_norm": 17.68119239807129,
      "learning_rate": 1.9635732489048134e-07,
      "loss": 1.9325,
      "step": 2984
    },
    {
      "epoch": 0.92,
      "grad_norm": 13.95864200592041,
      "learning_rate": 1.9475200743813983e-07,
      "loss": 1.6473,
      "step": 2985
    },
    {
      "epoch": 0.93,
      "grad_norm": 16.2840633392334,
      "learning_rate": 1.9315318716657281e-07,
      "loss": 3.1788,
      "step": 2986
    },
    {
      "epoch": 0.93,
      "grad_norm": 15.319509506225586,
      "learning_rate": 1.9156086559108976e-07,
      "loss": 9.2511,
      "step": 2987
    },
    {
      "epoch": 0.93,
      "grad_norm": 18.31650161743164,
      "learning_rate": 1.899750442208426e-07,
      "loss": 3.4509,
      "step": 2988
    },
    {
      "epoch": 0.93,
      "grad_norm": 18.275609970092773,
      "learning_rate": 1.8839572455881937e-07,
      "loss": 3.013,
      "step": 2989
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.587852478027344,
      "learning_rate": 1.868229081018483e-07,
      "loss": 1.4377,
      "step": 2990
    },
    {
      "epoch": 0.93,
      "grad_norm": 14.27921199798584,
      "learning_rate": 1.8525659634059525e-07,
      "loss": 2.0313,
      "step": 2991
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.033051490783691,
      "learning_rate": 1.8369679075955767e-07,
      "loss": 1.1735,
      "step": 2992
    },
    {
      "epoch": 0.93,
      "grad_norm": 18.491239547729492,
      "learning_rate": 1.821434928370684e-07,
      "loss": 5.3385,
      "step": 2993
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.413246154785156,
      "learning_rate": 1.80596704045294e-07,
      "loss": 1.5649,
      "step": 2994
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.972743034362793,
      "learning_rate": 1.7905642585022874e-07,
      "loss": 1.9612,
      "step": 2995
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.768912315368652,
      "learning_rate": 1.7752265971169908e-07,
      "loss": 1.735,
      "step": 2996
    },
    {
      "epoch": 0.93,
      "grad_norm": 13.970714569091797,
      "learning_rate": 1.7599540708335906e-07,
      "loss": 2.5504,
      "step": 2997
    },
    {
      "epoch": 0.93,
      "grad_norm": 14.848921775817871,
      "learning_rate": 1.7447466941268872e-07,
      "loss": 3.7117,
      "step": 2998
    },
    {
      "epoch": 0.93,
      "grad_norm": 7.468012809753418,
      "learning_rate": 1.7296044814099336e-07,
      "loss": 0.9068,
      "step": 2999
    },
    {
      "epoch": 0.93,
      "grad_norm": 10.256609916687012,
      "learning_rate": 1.7145274470340347e-07,
      "loss": 1.6465,
      "step": 3000
    },
    {
      "epoch": 0.93,
      "grad_norm": 15.5140380859375,
      "learning_rate": 1.6995156052886927e-07,
      "loss": 4.7481,
      "step": 3001
    },
    {
      "epoch": 0.93,
      "grad_norm": 8.324983596801758,
      "learning_rate": 1.684568970401655e-07,
      "loss": 0.8189,
      "step": 3002
    },
    {
      "epoch": 0.93,
      "grad_norm": 17.819440841674805,
      "learning_rate": 1.6696875565388651e-07,
      "loss": 2.9557,
      "step": 3003
    },
    {
      "epoch": 0.93,
      "grad_norm": 13.389687538146973,
      "learning_rate": 1.6548713778044263e-07,
      "loss": 2.1643,
      "step": 3004
    },
    {
      "epoch": 0.93,
      "grad_norm": 15.54904556274414,
      "learning_rate": 1.640120448240638e-07,
      "loss": 3.0798,
      "step": 3005
    },
    {
      "epoch": 0.93,
      "grad_norm": 14.200133323669434,
      "learning_rate": 1.6254347818279507e-07,
      "loss": 2.8567,
      "step": 3006
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.55955696105957,
      "learning_rate": 1.610814392484958e-07,
      "loss": 1.0734,
      "step": 3007
    },
    {
      "epoch": 0.93,
      "grad_norm": 15.413692474365234,
      "learning_rate": 1.5962592940683793e-07,
      "loss": 3.0384,
      "step": 3008
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.215126991271973,
      "learning_rate": 1.5817695003730858e-07,
      "loss": 0.9001,
      "step": 3009
    },
    {
      "epoch": 0.93,
      "grad_norm": 12.100921630859375,
      "learning_rate": 1.5673450251320202e-07,
      "loss": 1.7762,
      "step": 3010
    },
    {
      "epoch": 0.93,
      "grad_norm": 13.793547630310059,
      "learning_rate": 1.5529858820162204e-07,
      "loss": 2.3921,
      "step": 3011
    },
    {
      "epoch": 0.93,
      "grad_norm": 18.29058265686035,
      "learning_rate": 1.538692084634829e-07,
      "loss": 4.5271,
      "step": 3012
    },
    {
      "epoch": 0.93,
      "grad_norm": 9.583614349365234,
      "learning_rate": 1.5244636465350446e-07,
      "loss": 1.0375,
      "step": 3013
    },
    {
      "epoch": 0.93,
      "grad_norm": 16.553199768066406,
      "learning_rate": 1.5103005812020983e-07,
      "loss": 2.4258,
      "step": 3014
    },
    {
      "epoch": 0.93,
      "grad_norm": 19.734880447387695,
      "learning_rate": 1.4962029020593018e-07,
      "loss": 7.1062,
      "step": 3015
    },
    {
      "epoch": 0.93,
      "grad_norm": 11.817924499511719,
      "learning_rate": 1.482170622467961e-07,
      "loss": 1.8422,
      "step": 3016
    },
    {
      "epoch": 0.93,
      "grad_norm": 14.2987699508667,
      "learning_rate": 1.468203755727422e-07,
      "loss": 1.4806,
      "step": 3017
    },
    {
      "epoch": 0.94,
      "grad_norm": 14.055093765258789,
      "learning_rate": 1.4543023150750256e-07,
      "loss": 1.55,
      "step": 3018
    },
    {
      "epoch": 0.94,
      "grad_norm": 19.72188949584961,
      "learning_rate": 1.440466313686098e-07,
      "loss": 2.4603,
      "step": 3019
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.469618797302246,
      "learning_rate": 1.4266957646739674e-07,
      "loss": 1.2462,
      "step": 3020
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.99949836730957,
      "learning_rate": 1.4129906810898935e-07,
      "loss": 2.0896,
      "step": 3021
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.28267765045166,
      "learning_rate": 1.3993510759231143e-07,
      "loss": 4.4144,
      "step": 3022
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.911250114440918,
      "learning_rate": 1.3857769621007986e-07,
      "loss": 4.2494,
      "step": 3023
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.130582809448242,
      "learning_rate": 1.372268352488063e-07,
      "loss": 3.7505,
      "step": 3024
    },
    {
      "epoch": 0.94,
      "grad_norm": 20.72568130493164,
      "learning_rate": 1.3588252598879153e-07,
      "loss": 2.0615,
      "step": 3025
    },
    {
      "epoch": 0.94,
      "grad_norm": 18.65958023071289,
      "learning_rate": 1.3454476970412798e-07,
      "loss": 4.6672,
      "step": 3026
    },
    {
      "epoch": 0.94,
      "grad_norm": 17.683170318603516,
      "learning_rate": 1.3321356766269963e-07,
      "loss": 2.7539,
      "step": 3027
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.837944030761719,
      "learning_rate": 1.3188892112617414e-07,
      "loss": 3.0784,
      "step": 3028
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.589051246643066,
      "learning_rate": 1.3057083135000768e-07,
      "loss": 1.944,
      "step": 3029
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.951064109802246,
      "learning_rate": 1.292592995834456e-07,
      "loss": 2.2353,
      "step": 3030
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.508352279663086,
      "learning_rate": 1.2795432706951309e-07,
      "loss": 2.7971,
      "step": 3031
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.353891372680664,
      "learning_rate": 1.2665591504502062e-07,
      "loss": 1.9832,
      "step": 3032
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.772056579589844,
      "learning_rate": 1.2536406474056165e-07,
      "loss": 2.069,
      "step": 3033
    },
    {
      "epoch": 0.94,
      "grad_norm": 19.559324264526367,
      "learning_rate": 1.2407877738050943e-07,
      "loss": 2.9631,
      "step": 3034
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.815930366516113,
      "learning_rate": 1.2280005418301708e-07,
      "loss": 5.6007,
      "step": 3035
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.518260955810547,
      "learning_rate": 1.215278963600175e-07,
      "loss": 1.9506,
      "step": 3036
    },
    {
      "epoch": 0.94,
      "grad_norm": 20.898035049438477,
      "learning_rate": 1.2026230511722032e-07,
      "loss": 2.5529,
      "step": 3037
    },
    {
      "epoch": 0.94,
      "grad_norm": 8.328916549682617,
      "learning_rate": 1.1900328165411105e-07,
      "loss": 0.758,
      "step": 3038
    },
    {
      "epoch": 0.94,
      "grad_norm": 16.16352081298828,
      "learning_rate": 1.1775082716395116e-07,
      "loss": 1.857,
      "step": 3039
    },
    {
      "epoch": 0.94,
      "grad_norm": 15.048077583312988,
      "learning_rate": 1.165049428337772e-07,
      "loss": 4.866,
      "step": 3040
    },
    {
      "epoch": 0.94,
      "grad_norm": 17.457170486450195,
      "learning_rate": 1.1526562984439694e-07,
      "loss": 3.0742,
      "step": 3041
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.811216354370117,
      "learning_rate": 1.1403288937039175e-07,
      "loss": 1.2268,
      "step": 3042
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.016590118408203,
      "learning_rate": 1.1280672258011184e-07,
      "loss": 2.0475,
      "step": 3043
    },
    {
      "epoch": 0.94,
      "grad_norm": 9.95260238647461,
      "learning_rate": 1.1158713063567786e-07,
      "loss": 1.129,
      "step": 3044
    },
    {
      "epoch": 0.94,
      "grad_norm": 11.272063255310059,
      "learning_rate": 1.1037411469298089e-07,
      "loss": 1.6434,
      "step": 3045
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.511055946350098,
      "learning_rate": 1.0916767590167698e-07,
      "loss": 1.6796,
      "step": 3046
    },
    {
      "epoch": 0.94,
      "grad_norm": 10.552799224853516,
      "learning_rate": 1.0796781540518867e-07,
      "loss": 1.3846,
      "step": 3047
    },
    {
      "epoch": 0.94,
      "grad_norm": 13.904885292053223,
      "learning_rate": 1.0677453434070661e-07,
      "loss": 0.9898,
      "step": 3048
    },
    {
      "epoch": 0.94,
      "grad_norm": 12.032923698425293,
      "learning_rate": 1.0558783383918326e-07,
      "loss": 1.9281,
      "step": 3049
    },
    {
      "epoch": 0.95,
      "grad_norm": 14.337904930114746,
      "learning_rate": 1.0440771502533292e-07,
      "loss": 1.507,
      "step": 3050
    },
    {
      "epoch": 0.95,
      "grad_norm": 17.677717208862305,
      "learning_rate": 1.0323417901763561e-07,
      "loss": 4.5251,
      "step": 3051
    },
    {
      "epoch": 0.95,
      "grad_norm": 17.238025665283203,
      "learning_rate": 1.0206722692832925e-07,
      "loss": 4.1081,
      "step": 3052
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.109561920166016,
      "learning_rate": 1.0090685986341439e-07,
      "loss": 1.8459,
      "step": 3053
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.919336318969727,
      "learning_rate": 9.975307892264793e-08,
      "loss": 3.0997,
      "step": 3054
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.939140319824219,
      "learning_rate": 9.8605885199547e-08,
      "loss": 2.0562,
      "step": 3055
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.790946960449219,
      "learning_rate": 9.74652797813843e-08,
      "loss": 1.135,
      "step": 3056
    },
    {
      "epoch": 0.95,
      "grad_norm": 7.619593620300293,
      "learning_rate": 9.633126374918733e-08,
      "loss": 1.3455,
      "step": 3057
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.63952922821045,
      "learning_rate": 9.52038381777407e-08,
      "loss": 3.9738,
      "step": 3058
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.779718399047852,
      "learning_rate": 9.408300413558064e-08,
      "loss": 1.712,
      "step": 3059
    },
    {
      "epoch": 0.95,
      "grad_norm": 11.850090026855469,
      "learning_rate": 9.296876268499823e-08,
      "loss": 1.4411,
      "step": 3060
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.475018501281738,
      "learning_rate": 9.186111488203457e-08,
      "loss": 4.717,
      "step": 3061
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.185152053833008,
      "learning_rate": 9.076006177648243e-08,
      "loss": 1.4226,
      "step": 3062
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.603704452514648,
      "learning_rate": 8.966560441188388e-08,
      "loss": 2.6218,
      "step": 3063
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.038780212402344,
      "learning_rate": 8.85777438255295e-08,
      "loss": 2.4836,
      "step": 3064
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.55335807800293,
      "learning_rate": 8.74964810484592e-08,
      "loss": 1.6314,
      "step": 3065
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.07515525817871,
      "learning_rate": 8.642181710545822e-08,
      "loss": 2.2336,
      "step": 3066
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.747037887573242,
      "learning_rate": 8.535375301505804e-08,
      "loss": 0.8815,
      "step": 3067
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.257098197937012,
      "learning_rate": 8.429228978953389e-08,
      "loss": 8.3597,
      "step": 3068
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.84339714050293,
      "learning_rate": 8.323742843490879e-08,
      "loss": 2.8884,
      "step": 3069
    },
    {
      "epoch": 0.95,
      "grad_norm": 18.105701446533203,
      "learning_rate": 8.218916995094328e-08,
      "loss": 2.0507,
      "step": 3070
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.747038841247559,
      "learning_rate": 8.114751533114333e-08,
      "loss": 5.9551,
      "step": 3071
    },
    {
      "epoch": 0.95,
      "grad_norm": 10.658041000366211,
      "learning_rate": 8.011246556275555e-08,
      "loss": 1.7899,
      "step": 3072
    },
    {
      "epoch": 0.95,
      "grad_norm": 27.648420333862305,
      "learning_rate": 7.908402162676728e-08,
      "loss": 2.5052,
      "step": 3073
    },
    {
      "epoch": 0.95,
      "grad_norm": 16.756755828857422,
      "learning_rate": 7.806218449790182e-08,
      "loss": 1.9382,
      "step": 3074
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.293566703796387,
      "learning_rate": 7.704695514462554e-08,
      "loss": 1.938,
      "step": 3075
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.618069648742676,
      "learning_rate": 7.603833452913845e-08,
      "loss": 6.5341,
      "step": 3076
    },
    {
      "epoch": 0.95,
      "grad_norm": 12.356037139892578,
      "learning_rate": 7.503632360737888e-08,
      "loss": 1.3101,
      "step": 3077
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.837542533874512,
      "learning_rate": 7.40409233290204e-08,
      "loss": 1.2993,
      "step": 3078
    },
    {
      "epoch": 0.95,
      "grad_norm": 9.97978401184082,
      "learning_rate": 7.30521346374702e-08,
      "loss": 1.4106,
      "step": 3079
    },
    {
      "epoch": 0.95,
      "grad_norm": 13.803317070007324,
      "learning_rate": 7.206995846986914e-08,
      "loss": 4.2463,
      "step": 3080
    },
    {
      "epoch": 0.95,
      "grad_norm": 15.993304252624512,
      "learning_rate": 7.109439575709405e-08,
      "loss": 1.9746,
      "step": 3081
    },
    {
      "epoch": 0.95,
      "grad_norm": 14.088014602661133,
      "learning_rate": 7.012544742375074e-08,
      "loss": 1.7334,
      "step": 3082
    },
    {
      "epoch": 0.96,
      "grad_norm": 10.91285514831543,
      "learning_rate": 6.916311438817549e-08,
      "loss": 0.7295,
      "step": 3083
    },
    {
      "epoch": 0.96,
      "grad_norm": 8.432781219482422,
      "learning_rate": 6.820739756243829e-08,
      "loss": 1.5952,
      "step": 3084
    },
    {
      "epoch": 0.96,
      "grad_norm": 14.103729248046875,
      "learning_rate": 6.725829785233487e-08,
      "loss": 1.9413,
      "step": 3085
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.959000587463379,
      "learning_rate": 6.631581615739155e-08,
      "loss": 2.0251,
      "step": 3086
    },
    {
      "epoch": 0.96,
      "grad_norm": 14.486614227294922,
      "learning_rate": 6.537995337086198e-08,
      "loss": 2.2165,
      "step": 3087
    },
    {
      "epoch": 0.96,
      "grad_norm": 20.323394775390625,
      "learning_rate": 6.445071037972566e-08,
      "loss": 2.8939,
      "step": 3088
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.170675277709961,
      "learning_rate": 6.352808806468945e-08,
      "loss": 5.9418,
      "step": 3089
    },
    {
      "epoch": 0.96,
      "grad_norm": 25.27256965637207,
      "learning_rate": 6.261208730018529e-08,
      "loss": 1.3611,
      "step": 3090
    },
    {
      "epoch": 0.96,
      "grad_norm": 13.025799751281738,
      "learning_rate": 6.170270895436855e-08,
      "loss": 1.794,
      "step": 3091
    },
    {
      "epoch": 0.96,
      "grad_norm": 26.44091033935547,
      "learning_rate": 6.079995388911727e-08,
      "loss": 3.4504,
      "step": 3092
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.841899871826172,
      "learning_rate": 5.990382296003459e-08,
      "loss": 2.8998,
      "step": 3093
    },
    {
      "epoch": 0.96,
      "grad_norm": 14.168822288513184,
      "learning_rate": 5.901431701644315e-08,
      "loss": 2.3676,
      "step": 3094
    },
    {
      "epoch": 0.96,
      "grad_norm": 18.31540870666504,
      "learning_rate": 5.81314369013891e-08,
      "loss": 4.4452,
      "step": 3095
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.244515419006348,
      "learning_rate": 5.725518345163656e-08,
      "loss": 1.7953,
      "step": 3096
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.17121696472168,
      "learning_rate": 5.638555749766998e-08,
      "loss": 2.4318,
      "step": 3097
    },
    {
      "epoch": 0.96,
      "grad_norm": 20.633235931396484,
      "learning_rate": 5.552255986369262e-08,
      "loss": 1.9895,
      "step": 3098
    },
    {
      "epoch": 0.96,
      "grad_norm": 12.283513069152832,
      "learning_rate": 5.46661913676257e-08,
      "loss": 1.7769,
      "step": 3099
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.421818733215332,
      "learning_rate": 5.381645282110687e-08,
      "loss": 2.0828,
      "step": 3100
    },
    {
      "epoch": 0.96,
      "grad_norm": 18.69306755065918,
      "learning_rate": 5.2973345029491796e-08,
      "loss": 3.2615,
      "step": 3101
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.494669914245605,
      "learning_rate": 5.213686879184941e-08,
      "loss": 1.1403,
      "step": 3102
    },
    {
      "epoch": 0.96,
      "grad_norm": 19.049734115600586,
      "learning_rate": 5.130702490096508e-08,
      "loss": 1.6026,
      "step": 3103
    },
    {
      "epoch": 0.96,
      "grad_norm": 11.46169662475586,
      "learning_rate": 5.048381414333827e-08,
      "loss": 1.7153,
      "step": 3104
    },
    {
      "epoch": 0.96,
      "grad_norm": 9.212632179260254,
      "learning_rate": 4.966723729918015e-08,
      "loss": 0.8264,
      "step": 3105
    },
    {
      "epoch": 0.96,
      "grad_norm": 17.575876235961914,
      "learning_rate": 4.8857295142416755e-08,
      "loss": 3.2923,
      "step": 3106
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.128787994384766,
      "learning_rate": 4.8053988440684294e-08,
      "loss": 1.3332,
      "step": 3107
    },
    {
      "epoch": 0.96,
      "grad_norm": 17.256290435791016,
      "learning_rate": 4.7257317955331486e-08,
      "loss": 4.6119,
      "step": 3108
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.599822998046875,
      "learning_rate": 4.646728444141565e-08,
      "loss": 4.0102,
      "step": 3109
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.660905838012695,
      "learning_rate": 4.568388864770504e-08,
      "loss": 1.9701,
      "step": 3110
    },
    {
      "epoch": 0.96,
      "grad_norm": 15.143543243408203,
      "learning_rate": 4.4907131316677315e-08,
      "loss": 1.8301,
      "step": 3111
    },
    {
      "epoch": 0.96,
      "grad_norm": 13.858985900878906,
      "learning_rate": 4.413701318451793e-08,
      "loss": 1.576,
      "step": 3112
    },
    {
      "epoch": 0.96,
      "grad_norm": 16.94219970703125,
      "learning_rate": 4.337353498111781e-08,
      "loss": 2.0578,
      "step": 3113
    },
    {
      "epoch": 0.96,
      "grad_norm": 14.752725601196289,
      "learning_rate": 4.261669743007885e-08,
      "loss": 3.3421,
      "step": 3114
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.993329048156738,
      "learning_rate": 4.1866501248706036e-08,
      "loss": 6.4533,
      "step": 3115
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.602298736572266,
      "learning_rate": 4.11229471480114e-08,
      "loss": 1.2179,
      "step": 3116
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.814712524414062,
      "learning_rate": 4.038603583271089e-08,
      "loss": 1.283,
      "step": 3117
    },
    {
      "epoch": 0.97,
      "grad_norm": 14.151748657226562,
      "learning_rate": 3.9655768001225895e-08,
      "loss": 2.2273,
      "step": 3118
    },
    {
      "epoch": 0.97,
      "grad_norm": 14.59679126739502,
      "learning_rate": 3.89321443456786e-08,
      "loss": 2.1082,
      "step": 3119
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.78852367401123,
      "learning_rate": 3.821516555189821e-08,
      "loss": 1.6505,
      "step": 3120
    },
    {
      "epoch": 0.97,
      "grad_norm": 14.897425651550293,
      "learning_rate": 3.7504832299412366e-08,
      "loss": 1.7861,
      "step": 3121
    },
    {
      "epoch": 0.97,
      "grad_norm": 16.76914405822754,
      "learning_rate": 3.6801145261451817e-08,
      "loss": 4.539,
      "step": 3122
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.46123218536377,
      "learning_rate": 3.610410510494888e-08,
      "loss": 1.2907,
      "step": 3123
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.13853645324707,
      "learning_rate": 3.5413712490535053e-08,
      "loss": 2.8722,
      "step": 3124
    },
    {
      "epoch": 0.97,
      "grad_norm": 16.077030181884766,
      "learning_rate": 3.4729968072542645e-08,
      "loss": 1.5336,
      "step": 3125
    },
    {
      "epoch": 0.97,
      "grad_norm": 17.59398651123047,
      "learning_rate": 3.4052872499001574e-08,
      "loss": 3.9501,
      "step": 3126
    },
    {
      "epoch": 0.97,
      "grad_norm": 18.4521541595459,
      "learning_rate": 3.3382426411641764e-08,
      "loss": 1.9943,
      "step": 3127
    },
    {
      "epoch": 0.97,
      "grad_norm": 16.40480613708496,
      "learning_rate": 3.2718630445889206e-08,
      "loss": 1.7157,
      "step": 3128
    },
    {
      "epoch": 0.97,
      "grad_norm": 14.485817909240723,
      "learning_rate": 3.206148523086832e-08,
      "loss": 1.8532,
      "step": 3129
    },
    {
      "epoch": 0.97,
      "grad_norm": 18.246952056884766,
      "learning_rate": 3.14109913894004e-08,
      "loss": 4.0219,
      "step": 3130
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.424639701843262,
      "learning_rate": 3.076714953800199e-08,
      "loss": 1.6797,
      "step": 3131
    },
    {
      "epoch": 0.97,
      "grad_norm": 14.660819053649902,
      "learning_rate": 3.012996028688575e-08,
      "loss": 2.8193,
      "step": 3132
    },
    {
      "epoch": 0.97,
      "grad_norm": 13.350130081176758,
      "learning_rate": 2.949942423995884e-08,
      "loss": 3.0054,
      "step": 3133
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.52187728881836,
      "learning_rate": 2.887554199482214e-08,
      "loss": 2.7958,
      "step": 3134
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.523505210876465,
      "learning_rate": 2.8258314142771835e-08,
      "loss": 1.1539,
      "step": 3135
    },
    {
      "epoch": 0.97,
      "grad_norm": 15.402833938598633,
      "learning_rate": 2.764774126879626e-08,
      "loss": 2.2724,
      "step": 3136
    },
    {
      "epoch": 0.97,
      "grad_norm": 19.74759864807129,
      "learning_rate": 2.7043823951575144e-08,
      "loss": 1.6459,
      "step": 3137
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.21536636352539,
      "learning_rate": 2.644656276348349e-08,
      "loss": 1.2842,
      "step": 3138
    },
    {
      "epoch": 0.97,
      "grad_norm": 18.24991226196289,
      "learning_rate": 2.585595827058535e-08,
      "loss": 2.9906,
      "step": 3139
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.516609191894531,
      "learning_rate": 2.527201103263615e-08,
      "loss": 1.2808,
      "step": 3140
    },
    {
      "epoch": 0.97,
      "grad_norm": 13.044878005981445,
      "learning_rate": 2.4694721603082695e-08,
      "loss": 1.7913,
      "step": 3141
    },
    {
      "epoch": 0.97,
      "grad_norm": 19.25539779663086,
      "learning_rate": 2.4124090529060035e-08,
      "loss": 2.0492,
      "step": 3142
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.908823013305664,
      "learning_rate": 2.3560118351393824e-08,
      "loss": 1.1524,
      "step": 3143
    },
    {
      "epoch": 0.97,
      "grad_norm": 11.41439151763916,
      "learning_rate": 2.3002805604599526e-08,
      "loss": 0.7916,
      "step": 3144
    },
    {
      "epoch": 0.97,
      "grad_norm": 10.581273078918457,
      "learning_rate": 2.245215281687929e-08,
      "loss": 1.5594,
      "step": 3145
    },
    {
      "epoch": 0.97,
      "grad_norm": 12.786311149597168,
      "learning_rate": 2.1908160510124297e-08,
      "loss": 1.3038,
      "step": 3146
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.598114967346191,
      "learning_rate": 2.1370829199913194e-08,
      "loss": 1.4897,
      "step": 3147
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.448434829711914,
      "learning_rate": 2.0840159395508963e-08,
      "loss": 1.2894,
      "step": 3148
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.518309593200684,
      "learning_rate": 2.0316151599865964e-08,
      "loss": 1.4369,
      "step": 3149
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.300780296325684,
      "learning_rate": 1.979880630961976e-08,
      "loss": 1.6881,
      "step": 3150
    },
    {
      "epoch": 0.98,
      "grad_norm": 18.67576789855957,
      "learning_rate": 1.9288124015094174e-08,
      "loss": 2.6745,
      "step": 3151
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.411372184753418,
      "learning_rate": 1.8784105200296566e-08,
      "loss": 0.9355,
      "step": 3152
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.240246772766113,
      "learning_rate": 1.8286750342920988e-08,
      "loss": 1.6935,
      "step": 3153
    },
    {
      "epoch": 0.98,
      "grad_norm": 16.537799835205078,
      "learning_rate": 1.77960599143427e-08,
      "loss": 1.5408,
      "step": 3154
    },
    {
      "epoch": 0.98,
      "grad_norm": 14.367791175842285,
      "learning_rate": 1.7312034379622853e-08,
      "loss": 5.0619,
      "step": 3155
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.820769309997559,
      "learning_rate": 1.6834674197504595e-08,
      "loss": 1.0659,
      "step": 3156
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.2417573928833,
      "learning_rate": 1.636397982041462e-08,
      "loss": 4.2424,
      "step": 3157
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.1566801071167,
      "learning_rate": 1.5899951694462396e-08,
      "loss": 1.7313,
      "step": 3158
    },
    {
      "epoch": 0.98,
      "grad_norm": 19.42092514038086,
      "learning_rate": 1.5442590259437807e-08,
      "loss": 2.3418,
      "step": 3159
    },
    {
      "epoch": 0.98,
      "grad_norm": 16.40899658203125,
      "learning_rate": 1.499189594881273e-08,
      "loss": 3.196,
      "step": 3160
    },
    {
      "epoch": 0.98,
      "grad_norm": 17.25589942932129,
      "learning_rate": 1.4547869189741808e-08,
      "loss": 2.7817,
      "step": 3161
    },
    {
      "epoch": 0.98,
      "grad_norm": 7.910188674926758,
      "learning_rate": 1.4110510403058546e-08,
      "loss": 0.5826,
      "step": 3162
    },
    {
      "epoch": 0.98,
      "grad_norm": 10.37157154083252,
      "learning_rate": 1.3679820003276866e-08,
      "loss": 0.8978,
      "step": 3163
    },
    {
      "epoch": 0.98,
      "grad_norm": 18.129549026489258,
      "learning_rate": 1.3255798398591895e-08,
      "loss": 3.2344,
      "step": 3164
    },
    {
      "epoch": 0.98,
      "grad_norm": 17.443370819091797,
      "learning_rate": 1.2838445990876835e-08,
      "loss": 2.0711,
      "step": 3165
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.30059814453125,
      "learning_rate": 1.2427763175685308e-08,
      "loss": 3.252,
      "step": 3166
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.549692153930664,
      "learning_rate": 1.202375034224823e-08,
      "loss": 1.5618,
      "step": 3167
    },
    {
      "epoch": 0.98,
      "grad_norm": 19.086090087890625,
      "learning_rate": 1.1626407873477718e-08,
      "loss": 1.8955,
      "step": 3168
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.417947769165039,
      "learning_rate": 1.1235736145959269e-08,
      "loss": 1.935,
      "step": 3169
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.785294532775879,
      "learning_rate": 1.0851735529961149e-08,
      "loss": 5.1338,
      "step": 3170
    },
    {
      "epoch": 0.98,
      "grad_norm": 11.394896507263184,
      "learning_rate": 1.0474406389425783e-08,
      "loss": 1.197,
      "step": 3171
    },
    {
      "epoch": 0.98,
      "grad_norm": 13.535491943359375,
      "learning_rate": 1.010374908197289e-08,
      "loss": 2.2509,
      "step": 3172
    },
    {
      "epoch": 0.98,
      "grad_norm": 15.352787017822266,
      "learning_rate": 9.739763958900261e-09,
      "loss": 1.9027,
      "step": 3173
    },
    {
      "epoch": 0.98,
      "grad_norm": 16.808731079101562,
      "learning_rate": 9.382451365180627e-09,
      "loss": 3.5206,
      "step": 3174
    },
    {
      "epoch": 0.98,
      "grad_norm": 17.150636672973633,
      "learning_rate": 9.031811639463234e-09,
      "loss": 1.1038,
      "step": 3175
    },
    {
      "epoch": 0.98,
      "grad_norm": 9.438704490661621,
      "learning_rate": 8.687845114073828e-09,
      "loss": 1.4847,
      "step": 3176
    },
    {
      "epoch": 0.98,
      "grad_norm": 14.827310562133789,
      "learning_rate": 8.350552115011539e-09,
      "loss": 2.1309,
      "step": 3177
    },
    {
      "epoch": 0.98,
      "grad_norm": 12.88956356048584,
      "learning_rate": 8.019932961952003e-09,
      "loss": 1.8953,
      "step": 3178
    },
    {
      "epoch": 0.98,
      "grad_norm": 21.6605224609375,
      "learning_rate": 7.6959879682458e-09,
      "loss": 6.0578,
      "step": 3179
    },
    {
      "epoch": 0.99,
      "grad_norm": 13.935561180114746,
      "learning_rate": 7.378717440916882e-09,
      "loss": 1.2751,
      "step": 3180
    },
    {
      "epoch": 0.99,
      "grad_norm": 16.300317764282227,
      "learning_rate": 7.068121680664153e-09,
      "loss": 2.9739,
      "step": 3181
    },
    {
      "epoch": 0.99,
      "grad_norm": 15.14452838897705,
      "learning_rate": 6.764200981859886e-09,
      "loss": 1.763,
      "step": 3182
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.049345016479492,
      "learning_rate": 6.4669556325513025e-09,
      "loss": 1.719,
      "step": 3183
    },
    {
      "epoch": 0.99,
      "grad_norm": 21.240371704101562,
      "learning_rate": 6.176385914455868e-09,
      "loss": 3.6554,
      "step": 3184
    },
    {
      "epoch": 0.99,
      "grad_norm": 15.055821418762207,
      "learning_rate": 5.892492102967556e-09,
      "loss": 3.6622,
      "step": 3185
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.99221134185791,
      "learning_rate": 5.615274467152154e-09,
      "loss": 2.4352,
      "step": 3186
    },
    {
      "epoch": 0.99,
      "grad_norm": 14.632514953613281,
      "learning_rate": 5.344733269745694e-09,
      "loss": 1.2119,
      "step": 3187
    },
    {
      "epoch": 0.99,
      "grad_norm": 15.027057647705078,
      "learning_rate": 5.080868767159935e-09,
      "loss": 4.8931,
      "step": 3188
    },
    {
      "epoch": 0.99,
      "grad_norm": 13.23940658569336,
      "learning_rate": 4.8236812094768826e-09,
      "loss": 2.2475,
      "step": 3189
    },
    {
      "epoch": 0.99,
      "grad_norm": 8.672513008117676,
      "learning_rate": 4.5731708404487875e-09,
      "loss": 0.9679,
      "step": 3190
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.244919776916504,
      "learning_rate": 4.329337897503627e-09,
      "loss": 1.4577,
      "step": 3191
    },
    {
      "epoch": 0.99,
      "grad_norm": 15.41777229309082,
      "learning_rate": 4.0921826117372775e-09,
      "loss": 1.9401,
      "step": 3192
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.77115249633789,
      "learning_rate": 3.861705207916644e-09,
      "loss": 1.5778,
      "step": 3193
    },
    {
      "epoch": 0.99,
      "grad_norm": 14.487737655639648,
      "learning_rate": 3.637905904482791e-09,
      "loss": 1.9504,
      "step": 3194
    },
    {
      "epoch": 0.99,
      "grad_norm": 11.114299774169922,
      "learning_rate": 3.4207849135446836e-09,
      "loss": 1.9605,
      "step": 3195
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.995927810668945,
      "learning_rate": 3.210342440881531e-09,
      "loss": 1.1768,
      "step": 3196
    },
    {
      "epoch": 0.99,
      "grad_norm": 18.027769088745117,
      "learning_rate": 3.0065786859451382e-09,
      "loss": 5.1756,
      "step": 3197
    },
    {
      "epoch": 0.99,
      "grad_norm": 16.118772506713867,
      "learning_rate": 2.8094938418552095e-09,
      "loss": 2.346,
      "step": 3198
    },
    {
      "epoch": 0.99,
      "grad_norm": 12.7620849609375,
      "learning_rate": 2.6190880954024777e-09,
      "loss": 1.8738,
      "step": 3199
    },
    {
      "epoch": 0.99,
      "grad_norm": 10.473453521728516,
      "learning_rate": 2.4353616270479216e-09,
      "loss": 1.2152,
      "step": 3200
    }
  ],
  "logging_steps": 1,
  "max_steps": 3227,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 100,
  "total_flos": 0.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}