{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.6484383443207609,
  "eval_steps": 500,
  "global_step": 6000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00010807305738679347,
      "grad_norm": 1.1723264455795288,
      "learning_rate": 8e-05,
      "loss": 2.0628,
      "step": 1
    },
    {
      "epoch": 0.00021614611477358694,
      "grad_norm": 0.7034432291984558,
      "learning_rate": 8e-05,
      "loss": 2.1468,
      "step": 2
    },
    {
      "epoch": 0.0003242191721603804,
      "grad_norm": 1.2023898363113403,
      "learning_rate": 8e-05,
      "loss": 2.0087,
      "step": 3
    },
    {
      "epoch": 0.0004322922295471739,
      "grad_norm": 0.7881628274917603,
      "learning_rate": 8e-05,
      "loss": 2.1078,
      "step": 4
    },
    {
      "epoch": 0.0005403652869339674,
      "grad_norm": 0.4852694571018219,
      "learning_rate": 8e-05,
      "loss": 2.0177,
      "step": 5
    },
    {
      "epoch": 0.0006484383443207608,
      "grad_norm": 0.5214868187904358,
      "learning_rate": 8e-05,
      "loss": 2.1888,
      "step": 6
    },
    {
      "epoch": 0.0007565114017075543,
      "grad_norm": 0.4730895757675171,
      "learning_rate": 8e-05,
      "loss": 1.8638,
      "step": 7
    },
    {
      "epoch": 0.0008645844590943478,
      "grad_norm": 0.49415621161460876,
      "learning_rate": 8e-05,
      "loss": 2.0839,
      "step": 8
    },
    {
      "epoch": 0.0009726575164811412,
      "grad_norm": 0.48213955760002136,
      "learning_rate": 8e-05,
      "loss": 2.0302,
      "step": 9
    },
    {
      "epoch": 0.0010807305738679347,
      "grad_norm": 0.4757329821586609,
      "learning_rate": 8e-05,
      "loss": 1.8849,
      "step": 10
    },
    {
      "epoch": 0.0011888036312547282,
      "grad_norm": 0.4667082130908966,
      "learning_rate": 8e-05,
      "loss": 2.0472,
      "step": 11
    },
    {
      "epoch": 0.0012968766886415216,
      "grad_norm": 0.3837345540523529,
      "learning_rate": 8e-05,
      "loss": 1.8171,
      "step": 12
    },
    {
      "epoch": 0.001404949746028315,
      "grad_norm": 0.3936898410320282,
      "learning_rate": 8e-05,
      "loss": 1.9349,
      "step": 13
    },
    {
      "epoch": 0.0015130228034151087,
      "grad_norm": 0.3783639669418335,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 14
    },
    {
      "epoch": 0.0016210958608019021,
      "grad_norm": 0.4342162609100342,
      "learning_rate": 8e-05,
      "loss": 1.9042,
      "step": 15
    },
    {
      "epoch": 0.0017291689181886955,
      "grad_norm": 0.41809847950935364,
      "learning_rate": 8e-05,
      "loss": 1.8576,
      "step": 16
    },
    {
      "epoch": 0.001837241975575489,
      "grad_norm": 0.4019904136657715,
      "learning_rate": 8e-05,
      "loss": 1.8917,
      "step": 17
    },
    {
      "epoch": 0.0019453150329622824,
      "grad_norm": 0.3751857280731201,
      "learning_rate": 8e-05,
      "loss": 1.8635,
      "step": 18
    },
    {
      "epoch": 0.002053388090349076,
      "grad_norm": 0.4546978771686554,
      "learning_rate": 8e-05,
      "loss": 2.0914,
      "step": 19
    },
    {
      "epoch": 0.0021614611477358695,
      "grad_norm": 0.41806647181510925,
      "learning_rate": 8e-05,
      "loss": 1.9004,
      "step": 20
    },
    {
      "epoch": 0.002269534205122663,
      "grad_norm": 0.40970832109451294,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 21
    },
    {
      "epoch": 0.0023776072625094563,
      "grad_norm": 0.40093794465065,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 22
    },
    {
      "epoch": 0.0024856803198962498,
      "grad_norm": 0.43758541345596313,
      "learning_rate": 8e-05,
      "loss": 2.0376,
      "step": 23
    },
    {
      "epoch": 0.002593753377283043,
      "grad_norm": 0.3870728611946106,
      "learning_rate": 8e-05,
      "loss": 1.8323,
      "step": 24
    },
    {
      "epoch": 0.0027018264346698366,
      "grad_norm": 0.4140308201313019,
      "learning_rate": 8e-05,
      "loss": 1.9468,
      "step": 25
    },
    {
      "epoch": 0.00280989949205663,
      "grad_norm": 0.4336138963699341,
      "learning_rate": 8e-05,
      "loss": 2.0692,
      "step": 26
    },
    {
      "epoch": 0.002917972549443424,
      "grad_norm": 0.4042690396308899,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 27
    },
    {
      "epoch": 0.0030260456068302174,
      "grad_norm": 0.4046931564807892,
      "learning_rate": 8e-05,
      "loss": 2.1057,
      "step": 28
    },
    {
      "epoch": 0.003134118664217011,
      "grad_norm": 0.4208378493785858,
      "learning_rate": 8e-05,
      "loss": 1.9416,
      "step": 29
    },
    {
      "epoch": 0.0032421917216038042,
      "grad_norm": 0.44069036841392517,
      "learning_rate": 8e-05,
      "loss": 1.8726,
      "step": 30
    },
    {
      "epoch": 0.0033502647789905977,
      "grad_norm": 0.39752721786499023,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 31
    },
    {
      "epoch": 0.003458337836377391,
      "grad_norm": 0.4102960228919983,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 32
    },
    {
      "epoch": 0.0035664108937641845,
      "grad_norm": 0.3908040225505829,
      "learning_rate": 8e-05,
      "loss": 1.9257,
      "step": 33
    },
    {
      "epoch": 0.003674483951150978,
      "grad_norm": 0.41252148151397705,
      "learning_rate": 8e-05,
      "loss": 2.0182,
      "step": 34
    },
    {
      "epoch": 0.0037825570085377714,
      "grad_norm": 0.4197140634059906,
      "learning_rate": 8e-05,
      "loss": 1.917,
      "step": 35
    },
    {
      "epoch": 0.003890630065924565,
      "grad_norm": 0.4203090965747833,
      "learning_rate": 8e-05,
      "loss": 1.9532,
      "step": 36
    },
    {
      "epoch": 0.003998703123311359,
      "grad_norm": 0.39940884709358215,
      "learning_rate": 8e-05,
      "loss": 1.8866,
      "step": 37
    },
    {
      "epoch": 0.004106776180698152,
      "grad_norm": 0.3918643891811371,
      "learning_rate": 8e-05,
      "loss": 1.815,
      "step": 38
    },
    {
      "epoch": 0.0042148492380849455,
      "grad_norm": 0.3772515654563904,
      "learning_rate": 8e-05,
      "loss": 1.8445,
      "step": 39
    },
    {
      "epoch": 0.004322922295471739,
      "grad_norm": 0.40030282735824585,
      "learning_rate": 8e-05,
      "loss": 1.914,
      "step": 40
    },
    {
      "epoch": 0.004430995352858532,
      "grad_norm": 0.40037593245506287,
      "learning_rate": 8e-05,
      "loss": 1.9396,
      "step": 41
    },
    {
      "epoch": 0.004539068410245326,
      "grad_norm": 0.43869730830192566,
      "learning_rate": 8e-05,
      "loss": 1.9802,
      "step": 42
    },
    {
      "epoch": 0.004647141467632119,
      "grad_norm": 0.45838189125061035,
      "learning_rate": 8e-05,
      "loss": 1.9425,
      "step": 43
    },
    {
      "epoch": 0.004755214525018913,
      "grad_norm": 0.40124866366386414,
      "learning_rate": 8e-05,
      "loss": 1.9398,
      "step": 44
    },
    {
      "epoch": 0.004863287582405706,
      "grad_norm": 0.40170255303382874,
      "learning_rate": 8e-05,
      "loss": 2.006,
      "step": 45
    },
    {
      "epoch": 0.0049713606397924996,
      "grad_norm": 0.4152085483074188,
      "learning_rate": 8e-05,
      "loss": 1.8671,
      "step": 46
    },
    {
      "epoch": 0.005079433697179293,
      "grad_norm": 0.35721901059150696,
      "learning_rate": 8e-05,
      "loss": 1.8356,
      "step": 47
    },
    {
      "epoch": 0.005187506754566086,
      "grad_norm": 0.37504079937934875,
      "learning_rate": 8e-05,
      "loss": 1.878,
      "step": 48
    },
    {
      "epoch": 0.00529557981195288,
      "grad_norm": 0.36614570021629333,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 49
    },
    {
      "epoch": 0.005403652869339673,
      "grad_norm": 0.4576716125011444,
      "learning_rate": 8e-05,
      "loss": 2.044,
      "step": 50
    },
    {
      "epoch": 0.005511725926726467,
      "grad_norm": 0.3740018308162689,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 51
    },
    {
      "epoch": 0.00561979898411326,
      "grad_norm": 0.43559035658836365,
      "learning_rate": 8e-05,
      "loss": 2.1981,
      "step": 52
    },
    {
      "epoch": 0.005727872041500054,
      "grad_norm": 0.38865408301353455,
      "learning_rate": 8e-05,
      "loss": 1.8949,
      "step": 53
    },
    {
      "epoch": 0.005835945098886848,
      "grad_norm": 0.37158384919166565,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 54
    },
    {
      "epoch": 0.005944018156273641,
      "grad_norm": 0.37958189845085144,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 55
    },
    {
      "epoch": 0.006052091213660435,
      "grad_norm": 0.368605375289917,
      "learning_rate": 8e-05,
      "loss": 1.8265,
      "step": 56
    },
    {
      "epoch": 0.006160164271047228,
      "grad_norm": 0.39544960856437683,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 57
    },
    {
      "epoch": 0.006268237328434022,
      "grad_norm": 0.44783273339271545,
      "learning_rate": 8e-05,
      "loss": 1.9542,
      "step": 58
    },
    {
      "epoch": 0.006376310385820815,
      "grad_norm": 0.3987225294113159,
      "learning_rate": 8e-05,
      "loss": 1.8495,
      "step": 59
    },
    {
      "epoch": 0.0064843834432076084,
      "grad_norm": 0.45929187536239624,
      "learning_rate": 8e-05,
      "loss": 2.2551,
      "step": 60
    },
    {
      "epoch": 0.006592456500594402,
      "grad_norm": 0.37282443046569824,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 61
    },
    {
      "epoch": 0.006700529557981195,
      "grad_norm": 0.40441951155662537,
      "learning_rate": 8e-05,
      "loss": 1.8879,
      "step": 62
    },
    {
      "epoch": 0.006808602615367989,
      "grad_norm": 0.3969983458518982,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 63
    },
    {
      "epoch": 0.006916675672754782,
      "grad_norm": 0.3913993835449219,
      "learning_rate": 8e-05,
      "loss": 2.0212,
      "step": 64
    },
    {
      "epoch": 0.007024748730141576,
      "grad_norm": 0.4376554787158966,
      "learning_rate": 8e-05,
      "loss": 2.0607,
      "step": 65
    },
    {
      "epoch": 0.007132821787528369,
      "grad_norm": 0.37726086378097534,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 66
    },
    {
      "epoch": 0.0072408948449151625,
      "grad_norm": 0.41789546608924866,
      "learning_rate": 8e-05,
      "loss": 2.0039,
      "step": 67
    },
    {
      "epoch": 0.007348967902301956,
      "grad_norm": 0.39505037665367126,
      "learning_rate": 8e-05,
      "loss": 1.9182,
      "step": 68
    },
    {
      "epoch": 0.007457040959688749,
      "grad_norm": 0.4041613042354584,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 69
    },
    {
      "epoch": 0.007565114017075543,
      "grad_norm": 0.3840755224227905,
      "learning_rate": 8e-05,
      "loss": 1.8629,
      "step": 70
    },
    {
      "epoch": 0.007673187074462336,
      "grad_norm": 0.3758792579174042,
      "learning_rate": 8e-05,
      "loss": 1.9285,
      "step": 71
    },
    {
      "epoch": 0.00778126013184913,
      "grad_norm": 0.3771783411502838,
      "learning_rate": 8e-05,
      "loss": 1.8302,
      "step": 72
    },
    {
      "epoch": 0.007889333189235924,
      "grad_norm": 0.386371374130249,
      "learning_rate": 8e-05,
      "loss": 1.9723,
      "step": 73
    },
    {
      "epoch": 0.007997406246622717,
      "grad_norm": 0.3956718444824219,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 74
    },
    {
      "epoch": 0.00810547930400951,
      "grad_norm": 0.37973344326019287,
      "learning_rate": 8e-05,
      "loss": 1.8916,
      "step": 75
    },
    {
      "epoch": 0.008213552361396304,
      "grad_norm": 0.4194611608982086,
      "learning_rate": 8e-05,
      "loss": 1.9883,
      "step": 76
    },
    {
      "epoch": 0.008321625418783098,
      "grad_norm": 0.3735119700431824,
      "learning_rate": 8e-05,
      "loss": 1.8715,
      "step": 77
    },
    {
      "epoch": 0.008429698476169891,
      "grad_norm": 0.3977673053741455,
      "learning_rate": 8e-05,
      "loss": 1.8878,
      "step": 78
    },
    {
      "epoch": 0.008537771533556685,
      "grad_norm": 0.4270234704017639,
      "learning_rate": 8e-05,
      "loss": 1.9834,
      "step": 79
    },
    {
      "epoch": 0.008645844590943478,
      "grad_norm": 0.4047071933746338,
      "learning_rate": 8e-05,
      "loss": 1.9555,
      "step": 80
    },
    {
      "epoch": 0.008753917648330271,
      "grad_norm": 0.4528905749320984,
      "learning_rate": 8e-05,
      "loss": 2.0715,
      "step": 81
    },
    {
      "epoch": 0.008861990705717065,
      "grad_norm": 0.383327454328537,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 82
    },
    {
      "epoch": 0.008970063763103858,
      "grad_norm": 0.3786550760269165,
      "learning_rate": 8e-05,
      "loss": 1.8481,
      "step": 83
    },
    {
      "epoch": 0.009078136820490652,
      "grad_norm": 0.39775529503822327,
      "learning_rate": 8e-05,
      "loss": 1.9016,
      "step": 84
    },
    {
      "epoch": 0.009186209877877445,
      "grad_norm": 0.36153027415275574,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 85
    },
    {
      "epoch": 0.009294282935264239,
      "grad_norm": 0.3887314796447754,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 86
    },
    {
      "epoch": 0.009402355992651032,
      "grad_norm": 0.3864157199859619,
      "learning_rate": 8e-05,
      "loss": 1.9485,
      "step": 87
    },
    {
      "epoch": 0.009510429050037825,
      "grad_norm": 0.3642929792404175,
      "learning_rate": 8e-05,
      "loss": 1.5979,
      "step": 88
    },
    {
      "epoch": 0.009618502107424619,
      "grad_norm": 0.3677292466163635,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 89
    },
    {
      "epoch": 0.009726575164811412,
      "grad_norm": 0.3736303150653839,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 90
    },
    {
      "epoch": 0.009834648222198206,
      "grad_norm": 0.39751774072647095,
      "learning_rate": 8e-05,
      "loss": 1.9064,
      "step": 91
    },
    {
      "epoch": 0.009942721279584999,
      "grad_norm": 0.38238614797592163,
      "learning_rate": 8e-05,
      "loss": 1.9303,
      "step": 92
    },
    {
      "epoch": 0.010050794336971793,
      "grad_norm": 0.3611041307449341,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 93
    },
    {
      "epoch": 0.010158867394358586,
      "grad_norm": 0.3902607262134552,
      "learning_rate": 8e-05,
      "loss": 1.841,
      "step": 94
    },
    {
      "epoch": 0.01026694045174538,
      "grad_norm": 0.43703216314315796,
      "learning_rate": 8e-05,
      "loss": 2.0891,
      "step": 95
    },
    {
      "epoch": 0.010375013509132173,
      "grad_norm": 0.37732377648353577,
      "learning_rate": 8e-05,
      "loss": 1.8453,
      "step": 96
    },
    {
      "epoch": 0.010483086566518966,
      "grad_norm": 0.3939928114414215,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 97
    },
    {
      "epoch": 0.01059115962390576,
      "grad_norm": 0.42359834909439087,
      "learning_rate": 8e-05,
      "loss": 1.9811,
      "step": 98
    },
    {
      "epoch": 0.010699232681292553,
      "grad_norm": 0.35451656579971313,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 99
    },
    {
      "epoch": 0.010807305738679347,
      "grad_norm": 0.3824112117290497,
      "learning_rate": 8e-05,
      "loss": 2.0372,
      "step": 100
    },
    {
      "epoch": 0.01091537879606614,
      "grad_norm": 0.38967257738113403,
      "learning_rate": 8e-05,
      "loss": 1.6906,
      "step": 101
    },
    {
      "epoch": 0.011023451853452933,
      "grad_norm": 0.39197489619255066,
      "learning_rate": 8e-05,
      "loss": 1.7435,
      "step": 102
    },
    {
      "epoch": 0.011131524910839727,
      "grad_norm": 0.4271197021007538,
      "learning_rate": 8e-05,
      "loss": 1.9524,
      "step": 103
    },
    {
      "epoch": 0.01123959796822652,
      "grad_norm": 0.38411059975624084,
      "learning_rate": 8e-05,
      "loss": 1.9472,
      "step": 104
    },
    {
      "epoch": 0.011347671025613315,
      "grad_norm": 0.4045114815235138,
      "learning_rate": 8e-05,
      "loss": 1.9772,
      "step": 105
    },
    {
      "epoch": 0.011455744083000109,
      "grad_norm": 0.4735422432422638,
      "learning_rate": 8e-05,
      "loss": 1.7999,
      "step": 106
    },
    {
      "epoch": 0.011563817140386902,
      "grad_norm": 0.40350067615509033,
      "learning_rate": 8e-05,
      "loss": 1.9294,
      "step": 107
    },
    {
      "epoch": 0.011671890197773696,
      "grad_norm": 0.39403191208839417,
      "learning_rate": 8e-05,
      "loss": 1.9155,
      "step": 108
    },
    {
      "epoch": 0.01177996325516049,
      "grad_norm": 0.38301682472229004,
      "learning_rate": 8e-05,
      "loss": 1.8057,
      "step": 109
    },
    {
      "epoch": 0.011888036312547283,
      "grad_norm": 0.39158180356025696,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 110
    },
    {
      "epoch": 0.011996109369934076,
      "grad_norm": 0.383095920085907,
      "learning_rate": 8e-05,
      "loss": 1.8454,
      "step": 111
    },
    {
      "epoch": 0.01210418242732087,
      "grad_norm": 0.39831259846687317,
      "learning_rate": 8e-05,
      "loss": 1.9985,
      "step": 112
    },
    {
      "epoch": 0.012212255484707663,
      "grad_norm": 0.40717723965644836,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 113
    },
    {
      "epoch": 0.012320328542094456,
      "grad_norm": 0.37554696202278137,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 114
    },
    {
      "epoch": 0.01242840159948125,
      "grad_norm": 0.4222501218318939,
      "learning_rate": 8e-05,
      "loss": 1.9523,
      "step": 115
    },
    {
      "epoch": 0.012536474656868043,
      "grad_norm": 0.5975373387336731,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 116
    },
    {
      "epoch": 0.012644547714254837,
      "grad_norm": 0.38136371970176697,
      "learning_rate": 8e-05,
      "loss": 1.7762,
      "step": 117
    },
    {
      "epoch": 0.01275262077164163,
      "grad_norm": 0.41384100914001465,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 118
    },
    {
      "epoch": 0.012860693829028423,
      "grad_norm": 0.4296122193336487,
      "learning_rate": 8e-05,
      "loss": 1.9382,
      "step": 119
    },
    {
      "epoch": 0.012968766886415217,
      "grad_norm": 0.3701978921890259,
      "learning_rate": 8e-05,
      "loss": 1.8538,
      "step": 120
    },
    {
      "epoch": 0.01307683994380201,
      "grad_norm": 0.3837432563304901,
      "learning_rate": 8e-05,
      "loss": 1.9257,
      "step": 121
    },
    {
      "epoch": 0.013184913001188804,
      "grad_norm": 0.39746588468551636,
      "learning_rate": 8e-05,
      "loss": 1.8845,
      "step": 122
    },
    {
      "epoch": 0.013292986058575597,
      "grad_norm": 0.42879873514175415,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 123
    },
    {
      "epoch": 0.01340105911596239,
      "grad_norm": 0.39715635776519775,
      "learning_rate": 8e-05,
      "loss": 1.8587,
      "step": 124
    },
    {
      "epoch": 0.013509132173349184,
      "grad_norm": 0.38489505648612976,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 125
    },
    {
      "epoch": 0.013617205230735977,
      "grad_norm": 0.419987291097641,
      "learning_rate": 8e-05,
      "loss": 1.8885,
      "step": 126
    },
    {
      "epoch": 0.013725278288122771,
      "grad_norm": 0.4687266945838928,
      "learning_rate": 8e-05,
      "loss": 1.9872,
      "step": 127
    },
    {
      "epoch": 0.013833351345509564,
      "grad_norm": 0.3826073408126831,
      "learning_rate": 8e-05,
      "loss": 1.8377,
      "step": 128
    },
    {
      "epoch": 0.013941424402896358,
      "grad_norm": 0.37263014912605286,
      "learning_rate": 8e-05,
      "loss": 1.8564,
      "step": 129
    },
    {
      "epoch": 0.014049497460283151,
      "grad_norm": 0.4510330855846405,
      "learning_rate": 8e-05,
      "loss": 1.9611,
      "step": 130
    },
    {
      "epoch": 0.014157570517669945,
      "grad_norm": 0.40301069617271423,
      "learning_rate": 8e-05,
      "loss": 1.9504,
      "step": 131
    },
    {
      "epoch": 0.014265643575056738,
      "grad_norm": 0.3926645517349243,
      "learning_rate": 8e-05,
      "loss": 1.9395,
      "step": 132
    },
    {
      "epoch": 0.014373716632443531,
      "grad_norm": 0.41794946789741516,
      "learning_rate": 8e-05,
      "loss": 1.8927,
      "step": 133
    },
    {
      "epoch": 0.014481789689830325,
      "grad_norm": 0.40233609080314636,
      "learning_rate": 8e-05,
      "loss": 1.8041,
      "step": 134
    },
    {
      "epoch": 0.014589862747217118,
      "grad_norm": 0.4317133128643036,
      "learning_rate": 8e-05,
      "loss": 1.9633,
      "step": 135
    },
    {
      "epoch": 0.014697935804603912,
      "grad_norm": 0.3882908821105957,
      "learning_rate": 8e-05,
      "loss": 1.8945,
      "step": 136
    },
    {
      "epoch": 0.014806008861990705,
      "grad_norm": 0.4546521008014679,
      "learning_rate": 8e-05,
      "loss": 2.0693,
      "step": 137
    },
    {
      "epoch": 0.014914081919377499,
      "grad_norm": 0.3725680708885193,
      "learning_rate": 8e-05,
      "loss": 1.8717,
      "step": 138
    },
    {
      "epoch": 0.015022154976764292,
      "grad_norm": 0.38171207904815674,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 139
    },
    {
      "epoch": 0.015130228034151086,
      "grad_norm": 0.3819541931152344,
      "learning_rate": 8e-05,
      "loss": 1.8663,
      "step": 140
    },
    {
      "epoch": 0.015238301091537879,
      "grad_norm": 0.45130491256713867,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 141
    },
    {
      "epoch": 0.015346374148924672,
      "grad_norm": 0.4184626638889313,
      "learning_rate": 8e-05,
      "loss": 1.9189,
      "step": 142
    },
    {
      "epoch": 0.015454447206311466,
      "grad_norm": 0.38237568736076355,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 143
    },
    {
      "epoch": 0.01556252026369826,
      "grad_norm": 0.38273581862449646,
      "learning_rate": 8e-05,
      "loss": 1.7842,
      "step": 144
    },
    {
      "epoch": 0.015670593321085054,
      "grad_norm": 0.4038029611110687,
      "learning_rate": 8e-05,
      "loss": 1.825,
      "step": 145
    },
    {
      "epoch": 0.015778666378471848,
      "grad_norm": 0.4233596920967102,
      "learning_rate": 8e-05,
      "loss": 2.0048,
      "step": 146
    },
    {
      "epoch": 0.01588673943585864,
      "grad_norm": 0.3932502567768097,
      "learning_rate": 8e-05,
      "loss": 1.9008,
      "step": 147
    },
    {
      "epoch": 0.015994812493245435,
      "grad_norm": 0.3992982804775238,
      "learning_rate": 8e-05,
      "loss": 1.9438,
      "step": 148
    },
    {
      "epoch": 0.016102885550632228,
      "grad_norm": 0.3768817186355591,
      "learning_rate": 8e-05,
      "loss": 1.7915,
      "step": 149
    },
    {
      "epoch": 0.01621095860801902,
      "grad_norm": 0.39052727818489075,
      "learning_rate": 8e-05,
      "loss": 1.883,
      "step": 150
    },
    {
      "epoch": 0.016319031665405815,
      "grad_norm": 0.38049882650375366,
      "learning_rate": 8e-05,
      "loss": 1.9111,
      "step": 151
    },
    {
      "epoch": 0.01642710472279261,
      "grad_norm": 0.4055333733558655,
      "learning_rate": 8e-05,
      "loss": 2.0699,
      "step": 152
    },
    {
      "epoch": 0.016535177780179402,
      "grad_norm": 0.3887892961502075,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 153
    },
    {
      "epoch": 0.016643250837566195,
      "grad_norm": 0.3787854313850403,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 154
    },
    {
      "epoch": 0.01675132389495299,
      "grad_norm": 0.36630550026893616,
      "learning_rate": 8e-05,
      "loss": 1.8937,
      "step": 155
    },
    {
      "epoch": 0.016859396952339782,
      "grad_norm": 0.39160409569740295,
      "learning_rate": 8e-05,
      "loss": 1.7759,
      "step": 156
    },
    {
      "epoch": 0.016967470009726576,
      "grad_norm": 0.3865620493888855,
      "learning_rate": 8e-05,
      "loss": 1.8826,
      "step": 157
    },
    {
      "epoch": 0.01707554306711337,
      "grad_norm": 0.4001010060310364,
      "learning_rate": 8e-05,
      "loss": 1.8734,
      "step": 158
    },
    {
      "epoch": 0.017183616124500162,
      "grad_norm": 0.4675810933113098,
      "learning_rate": 8e-05,
      "loss": 2.0344,
      "step": 159
    },
    {
      "epoch": 0.017291689181886956,
      "grad_norm": 0.3979979455471039,
      "learning_rate": 8e-05,
      "loss": 1.884,
      "step": 160
    },
    {
      "epoch": 0.01739976223927375,
      "grad_norm": 0.3515948951244354,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 161
    },
    {
      "epoch": 0.017507835296660543,
      "grad_norm": 0.37273120880126953,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 162
    },
    {
      "epoch": 0.017615908354047336,
      "grad_norm": 0.3933907151222229,
      "learning_rate": 8e-05,
      "loss": 1.9517,
      "step": 163
    },
    {
      "epoch": 0.01772398141143413,
      "grad_norm": 0.3715977072715759,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 164
    },
    {
      "epoch": 0.017832054468820923,
      "grad_norm": 0.40160197019577026,
      "learning_rate": 8e-05,
      "loss": 2.0037,
      "step": 165
    },
    {
      "epoch": 0.017940127526207716,
      "grad_norm": 0.37912601232528687,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 166
    },
    {
      "epoch": 0.01804820058359451,
      "grad_norm": 0.38048484921455383,
      "learning_rate": 8e-05,
      "loss": 1.9331,
      "step": 167
    },
    {
      "epoch": 0.018156273640981303,
      "grad_norm": 0.4064115285873413,
      "learning_rate": 8e-05,
      "loss": 2.0729,
      "step": 168
    },
    {
      "epoch": 0.018264346698368097,
      "grad_norm": 0.37034860253334045,
      "learning_rate": 8e-05,
      "loss": 1.7719,
      "step": 169
    },
    {
      "epoch": 0.01837241975575489,
      "grad_norm": 0.4136922061443329,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 170
    },
    {
      "epoch": 0.018480492813141684,
      "grad_norm": 0.3788284659385681,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 171
    },
    {
      "epoch": 0.018588565870528477,
      "grad_norm": 0.6103711724281311,
      "learning_rate": 8e-05,
      "loss": 2.1371,
      "step": 172
    },
    {
      "epoch": 0.01869663892791527,
      "grad_norm": 0.42394280433654785,
      "learning_rate": 8e-05,
      "loss": 1.8954,
      "step": 173
    },
    {
      "epoch": 0.018804711985302064,
      "grad_norm": 0.38320788741111755,
      "learning_rate": 8e-05,
      "loss": 1.8101,
      "step": 174
    },
    {
      "epoch": 0.018912785042688857,
      "grad_norm": 0.3783103823661804,
      "learning_rate": 8e-05,
      "loss": 1.7826,
      "step": 175
    },
    {
      "epoch": 0.01902085810007565,
      "grad_norm": 0.37476250529289246,
      "learning_rate": 8e-05,
      "loss": 1.8191,
      "step": 176
    },
    {
      "epoch": 0.019128931157462444,
      "grad_norm": 0.3916049003601074,
      "learning_rate": 8e-05,
      "loss": 1.8888,
      "step": 177
    },
    {
      "epoch": 0.019237004214849238,
      "grad_norm": 0.38591471314430237,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 178
    },
    {
      "epoch": 0.01934507727223603,
      "grad_norm": 0.42106011509895325,
      "learning_rate": 8e-05,
      "loss": 1.8419,
      "step": 179
    },
    {
      "epoch": 0.019453150329622824,
      "grad_norm": 0.3971245586872101,
      "learning_rate": 8e-05,
      "loss": 1.9166,
      "step": 180
    },
    {
      "epoch": 0.019561223387009618,
      "grad_norm": 0.36734268069267273,
      "learning_rate": 8e-05,
      "loss": 1.8008,
      "step": 181
    },
    {
      "epoch": 0.01966929644439641,
      "grad_norm": 0.39223712682724,
      "learning_rate": 8e-05,
      "loss": 1.8541,
      "step": 182
    },
    {
      "epoch": 0.019777369501783205,
      "grad_norm": 0.37673863768577576,
      "learning_rate": 8e-05,
      "loss": 1.8661,
      "step": 183
    },
    {
      "epoch": 0.019885442559169998,
      "grad_norm": 0.4065092206001282,
      "learning_rate": 8e-05,
      "loss": 1.8001,
      "step": 184
    },
    {
      "epoch": 0.01999351561655679,
      "grad_norm": 0.42216357588768005,
      "learning_rate": 8e-05,
      "loss": 2.0559,
      "step": 185
    },
    {
      "epoch": 0.020101588673943585,
      "grad_norm": 0.39214301109313965,
      "learning_rate": 8e-05,
      "loss": 1.8752,
      "step": 186
    },
    {
      "epoch": 0.02020966173133038,
      "grad_norm": 0.3629594147205353,
      "learning_rate": 8e-05,
      "loss": 1.7512,
      "step": 187
    },
    {
      "epoch": 0.020317734788717172,
      "grad_norm": 0.402019739151001,
      "learning_rate": 8e-05,
      "loss": 1.9354,
      "step": 188
    },
    {
      "epoch": 0.020425807846103965,
      "grad_norm": 0.3779628574848175,
      "learning_rate": 8e-05,
      "loss": 1.9256,
      "step": 189
    },
    {
      "epoch": 0.02053388090349076,
      "grad_norm": 0.39732345938682556,
      "learning_rate": 8e-05,
      "loss": 1.9088,
      "step": 190
    },
    {
      "epoch": 0.020641953960877552,
      "grad_norm": 0.41548261046409607,
      "learning_rate": 8e-05,
      "loss": 1.9202,
      "step": 191
    },
    {
      "epoch": 0.020750027018264346,
      "grad_norm": 0.44871053099632263,
      "learning_rate": 8e-05,
      "loss": 1.9754,
      "step": 192
    },
    {
      "epoch": 0.02085810007565114,
      "grad_norm": 0.4330733120441437,
      "learning_rate": 8e-05,
      "loss": 1.9946,
      "step": 193
    },
    {
      "epoch": 0.020966173133037933,
      "grad_norm": 0.3993016481399536,
      "learning_rate": 8e-05,
      "loss": 1.968,
      "step": 194
    },
    {
      "epoch": 0.021074246190424726,
      "grad_norm": 0.3680805265903473,
      "learning_rate": 8e-05,
      "loss": 1.7862,
      "step": 195
    },
    {
      "epoch": 0.02118231924781152,
      "grad_norm": 0.3532881736755371,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 196
    },
    {
      "epoch": 0.021290392305198313,
      "grad_norm": 0.3720663785934448,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 197
    },
    {
      "epoch": 0.021398465362585106,
      "grad_norm": 0.3809856176376343,
      "learning_rate": 8e-05,
      "loss": 1.9466,
      "step": 198
    },
    {
      "epoch": 0.0215065384199719,
      "grad_norm": 0.3803189694881439,
      "learning_rate": 8e-05,
      "loss": 1.8712,
      "step": 199
    },
    {
      "epoch": 0.021614611477358693,
      "grad_norm": 0.3800376355648041,
      "learning_rate": 8e-05,
      "loss": 1.8415,
      "step": 200
    },
    {
      "epoch": 0.021722684534745487,
      "grad_norm": 0.373121976852417,
      "learning_rate": 8e-05,
      "loss": 1.8615,
      "step": 201
    },
    {
      "epoch": 0.02183075759213228,
      "grad_norm": 0.3480569124221802,
      "learning_rate": 8e-05,
      "loss": 1.7897,
      "step": 202
    },
    {
      "epoch": 0.021938830649519073,
      "grad_norm": 0.37026509642601013,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 203
    },
    {
      "epoch": 0.022046903706905867,
      "grad_norm": 0.40730607509613037,
      "learning_rate": 8e-05,
      "loss": 1.9032,
      "step": 204
    },
    {
      "epoch": 0.02215497676429266,
      "grad_norm": 0.36526039242744446,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 205
    },
    {
      "epoch": 0.022263049821679454,
      "grad_norm": 0.3851311504840851,
      "learning_rate": 8e-05,
      "loss": 1.8557,
      "step": 206
    },
    {
      "epoch": 0.022371122879066247,
      "grad_norm": 0.38706260919570923,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 207
    },
    {
      "epoch": 0.02247919593645304,
      "grad_norm": 0.42131349444389343,
      "learning_rate": 8e-05,
      "loss": 1.9376,
      "step": 208
    },
    {
      "epoch": 0.022587268993839837,
      "grad_norm": 0.39509931206703186,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 209
    },
    {
      "epoch": 0.02269534205122663,
      "grad_norm": 0.4486461877822876,
      "learning_rate": 8e-05,
      "loss": 2.046,
      "step": 210
    },
    {
      "epoch": 0.022803415108613424,
      "grad_norm": 0.37511327862739563,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 211
    },
    {
      "epoch": 0.022911488166000218,
      "grad_norm": 0.392880380153656,
      "learning_rate": 8e-05,
      "loss": 1.9183,
      "step": 212
    },
    {
      "epoch": 0.02301956122338701,
      "grad_norm": 0.3588014543056488,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 213
    },
    {
      "epoch": 0.023127634280773805,
      "grad_norm": 0.409408837556839,
      "learning_rate": 8e-05,
      "loss": 1.9709,
      "step": 214
    },
    {
      "epoch": 0.023235707338160598,
      "grad_norm": 0.38684749603271484,
      "learning_rate": 8e-05,
      "loss": 1.6974,
      "step": 215
    },
    {
      "epoch": 0.02334378039554739,
      "grad_norm": 0.40352535247802734,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 216
    },
    {
      "epoch": 0.023451853452934185,
      "grad_norm": 0.38765749335289,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 217
    },
    {
      "epoch": 0.02355992651032098,
      "grad_norm": 0.395373672246933,
      "learning_rate": 8e-05,
      "loss": 1.8828,
      "step": 218
    },
    {
      "epoch": 0.023667999567707772,
      "grad_norm": 0.40965431928634644,
      "learning_rate": 8e-05,
      "loss": 1.9358,
      "step": 219
    },
    {
      "epoch": 0.023776072625094565,
      "grad_norm": 0.38884827494621277,
      "learning_rate": 8e-05,
      "loss": 1.8027,
      "step": 220
    },
    {
      "epoch": 0.02388414568248136,
      "grad_norm": 0.4290681779384613,
      "learning_rate": 8e-05,
      "loss": 1.8723,
      "step": 221
    },
    {
      "epoch": 0.023992218739868152,
      "grad_norm": 0.4041561484336853,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 222
    },
    {
      "epoch": 0.024100291797254945,
      "grad_norm": 0.37803947925567627,
      "learning_rate": 8e-05,
      "loss": 1.7776,
      "step": 223
    },
    {
      "epoch": 0.02420836485464174,
      "grad_norm": 0.3696417808532715,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 224
    },
    {
      "epoch": 0.024316437912028532,
      "grad_norm": 0.3745357394218445,
      "learning_rate": 8e-05,
      "loss": 1.7107,
      "step": 225
    },
    {
      "epoch": 0.024424510969415326,
      "grad_norm": 0.3825242221355438,
      "learning_rate": 8e-05,
      "loss": 1.8554,
      "step": 226
    },
    {
      "epoch": 0.02453258402680212,
      "grad_norm": 0.37862396240234375,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 227
    },
    {
      "epoch": 0.024640657084188913,
      "grad_norm": 0.4100801944732666,
      "learning_rate": 8e-05,
      "loss": 1.9325,
      "step": 228
    },
    {
      "epoch": 0.024748730141575706,
      "grad_norm": 0.37161606550216675,
      "learning_rate": 8e-05,
      "loss": 1.7675,
      "step": 229
    },
    {
      "epoch": 0.0248568031989625,
      "grad_norm": 0.38659462332725525,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 230
    },
    {
      "epoch": 0.024964876256349293,
      "grad_norm": 0.35655543208122253,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 231
    },
    {
      "epoch": 0.025072949313736086,
      "grad_norm": 0.3984413146972656,
      "learning_rate": 8e-05,
      "loss": 1.8603,
      "step": 232
    },
    {
      "epoch": 0.02518102237112288,
      "grad_norm": 0.3585137128829956,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 233
    },
    {
      "epoch": 0.025289095428509673,
      "grad_norm": 0.39096739888191223,
      "learning_rate": 8e-05,
      "loss": 1.9319,
      "step": 234
    },
    {
      "epoch": 0.025397168485896467,
      "grad_norm": 0.37001100182533264,
      "learning_rate": 8e-05,
      "loss": 1.7834,
      "step": 235
    },
    {
      "epoch": 0.02550524154328326,
      "grad_norm": 0.39547184109687805,
      "learning_rate": 8e-05,
      "loss": 1.8208,
      "step": 236
    },
    {
      "epoch": 0.025613314600670054,
      "grad_norm": 0.4055747985839844,
      "learning_rate": 8e-05,
      "loss": 1.6514,
      "step": 237
    },
    {
      "epoch": 0.025721387658056847,
      "grad_norm": 0.33110472559928894,
      "learning_rate": 8e-05,
      "loss": 1.5289,
      "step": 238
    },
    {
      "epoch": 0.02582946071544364,
      "grad_norm": 0.4217626750469208,
      "learning_rate": 8e-05,
      "loss": 2.0152,
      "step": 239
    },
    {
      "epoch": 0.025937533772830434,
      "grad_norm": 0.41001901030540466,
      "learning_rate": 8e-05,
      "loss": 1.8881,
      "step": 240
    },
    {
      "epoch": 0.026045606830217227,
      "grad_norm": 0.3775026798248291,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 241
    },
    {
      "epoch": 0.02615367988760402,
      "grad_norm": 0.3907907009124756,
      "learning_rate": 8e-05,
      "loss": 1.9618,
      "step": 242
    },
    {
      "epoch": 0.026261752944990814,
      "grad_norm": 0.386920690536499,
      "learning_rate": 8e-05,
      "loss": 1.8331,
      "step": 243
    },
    {
      "epoch": 0.026369826002377608,
      "grad_norm": 0.3928714990615845,
      "learning_rate": 8e-05,
      "loss": 1.8987,
      "step": 244
    },
    {
      "epoch": 0.0264778990597644,
      "grad_norm": 0.3885044455528259,
      "learning_rate": 8e-05,
      "loss": 1.8571,
      "step": 245
    },
    {
      "epoch": 0.026585972117151194,
      "grad_norm": 0.37548908591270447,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 246
    },
    {
      "epoch": 0.026694045174537988,
      "grad_norm": 0.4012278616428375,
      "learning_rate": 8e-05,
      "loss": 1.8952,
      "step": 247
    },
    {
      "epoch": 0.02680211823192478,
      "grad_norm": 0.3742852210998535,
      "learning_rate": 8e-05,
      "loss": 1.8653,
      "step": 248
    },
    {
      "epoch": 0.026910191289311575,
      "grad_norm": 0.4425727427005768,
      "learning_rate": 8e-05,
      "loss": 1.9555,
      "step": 249
    },
    {
      "epoch": 0.027018264346698368,
      "grad_norm": 0.4133990705013275,
      "learning_rate": 8e-05,
      "loss": 2.0247,
      "step": 250
    },
    {
      "epoch": 0.02712633740408516,
      "grad_norm": 0.3764892816543579,
      "learning_rate": 8e-05,
      "loss": 1.8752,
      "step": 251
    },
    {
      "epoch": 0.027234410461471955,
      "grad_norm": 0.38350343704223633,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 252
    },
    {
      "epoch": 0.02734248351885875,
      "grad_norm": 0.38418692350387573,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 253
    },
    {
      "epoch": 0.027450556576245542,
      "grad_norm": 0.3836568295955658,
      "learning_rate": 8e-05,
      "loss": 1.9281,
      "step": 254
    },
    {
      "epoch": 0.027558629633632335,
      "grad_norm": 0.38728609681129456,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 255
    },
    {
      "epoch": 0.02766670269101913,
      "grad_norm": 0.3881751000881195,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 256
    },
    {
      "epoch": 0.027774775748405922,
      "grad_norm": 0.43750160932540894,
      "learning_rate": 8e-05,
      "loss": 2.0151,
      "step": 257
    },
    {
      "epoch": 0.027882848805792716,
      "grad_norm": 0.3566257357597351,
      "learning_rate": 8e-05,
      "loss": 1.7051,
      "step": 258
    },
    {
      "epoch": 0.02799092186317951,
      "grad_norm": 0.393274188041687,
      "learning_rate": 8e-05,
      "loss": 1.9421,
      "step": 259
    },
    {
      "epoch": 0.028098994920566302,
      "grad_norm": 0.3983575999736786,
      "learning_rate": 8e-05,
      "loss": 1.8355,
      "step": 260
    },
    {
      "epoch": 0.028207067977953096,
      "grad_norm": 0.3800559341907501,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 261
    },
    {
      "epoch": 0.02831514103533989,
      "grad_norm": 0.3782293200492859,
      "learning_rate": 8e-05,
      "loss": 1.8482,
      "step": 262
    },
    {
      "epoch": 0.028423214092726683,
      "grad_norm": 0.40121495723724365,
      "learning_rate": 8e-05,
      "loss": 1.8133,
      "step": 263
    },
    {
      "epoch": 0.028531287150113476,
      "grad_norm": 0.4046691060066223,
      "learning_rate": 8e-05,
      "loss": 1.8663,
      "step": 264
    },
    {
      "epoch": 0.02863936020750027,
      "grad_norm": 0.3844401240348816,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 265
    },
    {
      "epoch": 0.028747433264887063,
      "grad_norm": 0.37493830919265747,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 266
    },
    {
      "epoch": 0.028855506322273856,
      "grad_norm": 0.3654527962207794,
      "learning_rate": 8e-05,
      "loss": 1.8184,
      "step": 267
    },
    {
      "epoch": 0.02896357937966065,
      "grad_norm": 0.38567790389060974,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 268
    },
    {
      "epoch": 0.029071652437047443,
      "grad_norm": 0.37764883041381836,
      "learning_rate": 8e-05,
      "loss": 1.7943,
      "step": 269
    },
    {
      "epoch": 0.029179725494434237,
      "grad_norm": 0.37799185514450073,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 270
    },
    {
      "epoch": 0.02928779855182103,
      "grad_norm": 0.37777647376060486,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 271
    },
    {
      "epoch": 0.029395871609207824,
      "grad_norm": 0.38331955671310425,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 272
    },
    {
      "epoch": 0.029503944666594617,
      "grad_norm": 0.38704758882522583,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 273
    },
    {
      "epoch": 0.02961201772398141,
      "grad_norm": 0.3921975791454315,
      "learning_rate": 8e-05,
      "loss": 1.8602,
      "step": 274
    },
    {
      "epoch": 0.029720090781368204,
      "grad_norm": 0.3866276741027832,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 275
    },
    {
      "epoch": 0.029828163838754997,
      "grad_norm": 0.360526978969574,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 276
    },
    {
      "epoch": 0.02993623689614179,
      "grad_norm": 0.406816303730011,
      "learning_rate": 8e-05,
      "loss": 1.8358,
      "step": 277
    },
    {
      "epoch": 0.030044309953528584,
      "grad_norm": 0.41753488779067993,
      "learning_rate": 8e-05,
      "loss": 1.9256,
      "step": 278
    },
    {
      "epoch": 0.030152383010915378,
      "grad_norm": 0.43013501167297363,
      "learning_rate": 8e-05,
      "loss": 1.9959,
      "step": 279
    },
    {
      "epoch": 0.03026045606830217,
      "grad_norm": 0.38277286291122437,
      "learning_rate": 8e-05,
      "loss": 1.887,
      "step": 280
    },
    {
      "epoch": 0.030368529125688964,
      "grad_norm": 0.35029610991477966,
      "learning_rate": 8e-05,
      "loss": 1.6034,
      "step": 281
    },
    {
      "epoch": 0.030476602183075758,
      "grad_norm": 0.41762253642082214,
      "learning_rate": 8e-05,
      "loss": 1.8519,
      "step": 282
    },
    {
      "epoch": 0.03058467524046255,
      "grad_norm": 0.38342058658599854,
      "learning_rate": 8e-05,
      "loss": 1.9659,
      "step": 283
    },
    {
      "epoch": 0.030692748297849345,
      "grad_norm": 0.37613609433174133,
      "learning_rate": 8e-05,
      "loss": 1.8346,
      "step": 284
    },
    {
      "epoch": 0.030800821355236138,
      "grad_norm": 0.39876094460487366,
      "learning_rate": 8e-05,
      "loss": 1.8719,
      "step": 285
    },
    {
      "epoch": 0.03090889441262293,
      "grad_norm": 0.37250667810440063,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 286
    },
    {
      "epoch": 0.031016967470009725,
      "grad_norm": 0.3599224388599396,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 287
    },
    {
      "epoch": 0.03112504052739652,
      "grad_norm": 0.380977988243103,
      "learning_rate": 8e-05,
      "loss": 1.9555,
      "step": 288
    },
    {
      "epoch": 0.031233113584783312,
      "grad_norm": 0.3830679655075073,
      "learning_rate": 8e-05,
      "loss": 1.9513,
      "step": 289
    },
    {
      "epoch": 0.03134118664217011,
      "grad_norm": 0.4136650264263153,
      "learning_rate": 8e-05,
      "loss": 1.8796,
      "step": 290
    },
    {
      "epoch": 0.0314492596995569,
      "grad_norm": 0.35918280482292175,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 291
    },
    {
      "epoch": 0.031557332756943696,
      "grad_norm": 0.3885393738746643,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 292
    },
    {
      "epoch": 0.031665405814330486,
      "grad_norm": 0.39085906744003296,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 293
    },
    {
      "epoch": 0.03177347887171728,
      "grad_norm": 0.39614802598953247,
      "learning_rate": 8e-05,
      "loss": 1.9149,
      "step": 294
    },
    {
      "epoch": 0.03188155192910407,
      "grad_norm": 0.3644985556602478,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 295
    },
    {
      "epoch": 0.03198962498649087,
      "grad_norm": 0.42358651757240295,
      "learning_rate": 8e-05,
      "loss": 1.9755,
      "step": 296
    },
    {
      "epoch": 0.03209769804387766,
      "grad_norm": 0.3941521942615509,
      "learning_rate": 8e-05,
      "loss": 1.851,
      "step": 297
    },
    {
      "epoch": 0.032205771101264456,
      "grad_norm": 0.38100939989089966,
      "learning_rate": 8e-05,
      "loss": 1.7567,
      "step": 298
    },
    {
      "epoch": 0.032313844158651246,
      "grad_norm": 0.39728134870529175,
      "learning_rate": 8e-05,
      "loss": 1.9144,
      "step": 299
    },
    {
      "epoch": 0.03242191721603804,
      "grad_norm": 0.35281136631965637,
      "learning_rate": 8e-05,
      "loss": 1.6103,
      "step": 300
    },
    {
      "epoch": 0.03252999027342483,
      "grad_norm": 0.3991057574748993,
      "learning_rate": 8e-05,
      "loss": 1.9268,
      "step": 301
    },
    {
      "epoch": 0.03263806333081163,
      "grad_norm": 0.3573606312274933,
      "learning_rate": 8e-05,
      "loss": 1.8599,
      "step": 302
    },
    {
      "epoch": 0.03274613638819842,
      "grad_norm": 0.39089342951774597,
      "learning_rate": 8e-05,
      "loss": 1.7917,
      "step": 303
    },
    {
      "epoch": 0.03285420944558522,
      "grad_norm": 0.4095103442668915,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 304
    },
    {
      "epoch": 0.03296228250297201,
      "grad_norm": 0.4048735201358795,
      "learning_rate": 8e-05,
      "loss": 1.9357,
      "step": 305
    },
    {
      "epoch": 0.033070355560358804,
      "grad_norm": 0.4243958592414856,
      "learning_rate": 8e-05,
      "loss": 1.8902,
      "step": 306
    },
    {
      "epoch": 0.033178428617745594,
      "grad_norm": 0.38280507922172546,
      "learning_rate": 8e-05,
      "loss": 1.8338,
      "step": 307
    },
    {
      "epoch": 0.03328650167513239,
      "grad_norm": 0.4853745102882385,
      "learning_rate": 8e-05,
      "loss": 1.9228,
      "step": 308
    },
    {
      "epoch": 0.03339457473251918,
      "grad_norm": 0.40214666724205017,
      "learning_rate": 8e-05,
      "loss": 1.8908,
      "step": 309
    },
    {
      "epoch": 0.03350264778990598,
      "grad_norm": 0.3799232244491577,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 310
    },
    {
      "epoch": 0.03361072084729277,
      "grad_norm": 0.392907053232193,
      "learning_rate": 8e-05,
      "loss": 1.8322,
      "step": 311
    },
    {
      "epoch": 0.033718793904679564,
      "grad_norm": 0.3697127103805542,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 312
    },
    {
      "epoch": 0.033826866962066354,
      "grad_norm": 0.4245065152645111,
      "learning_rate": 8e-05,
      "loss": 1.9461,
      "step": 313
    },
    {
      "epoch": 0.03393494001945315,
      "grad_norm": 0.4249866306781769,
      "learning_rate": 8e-05,
      "loss": 2.0409,
      "step": 314
    },
    {
      "epoch": 0.03404301307683994,
      "grad_norm": 0.4139290452003479,
      "learning_rate": 8e-05,
      "loss": 1.9226,
      "step": 315
    },
    {
      "epoch": 0.03415108613422674,
      "grad_norm": 0.4352912902832031,
      "learning_rate": 8e-05,
      "loss": 1.9583,
      "step": 316
    },
    {
      "epoch": 0.03425915919161353,
      "grad_norm": 0.40120869874954224,
      "learning_rate": 8e-05,
      "loss": 1.926,
      "step": 317
    },
    {
      "epoch": 0.034367232249000325,
      "grad_norm": 0.3811039328575134,
      "learning_rate": 8e-05,
      "loss": 1.7982,
      "step": 318
    },
    {
      "epoch": 0.034475305306387115,
      "grad_norm": 0.41412317752838135,
      "learning_rate": 8e-05,
      "loss": 1.8325,
      "step": 319
    },
    {
      "epoch": 0.03458337836377391,
      "grad_norm": 0.3677211105823517,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 320
    },
    {
      "epoch": 0.0346914514211607,
      "grad_norm": 0.40204325318336487,
      "learning_rate": 8e-05,
      "loss": 1.9544,
      "step": 321
    },
    {
      "epoch": 0.0347995244785475,
      "grad_norm": 0.3774854242801666,
      "learning_rate": 8e-05,
      "loss": 1.8842,
      "step": 322
    },
    {
      "epoch": 0.03490759753593429,
      "grad_norm": 0.41708892583847046,
      "learning_rate": 8e-05,
      "loss": 1.9697,
      "step": 323
    },
    {
      "epoch": 0.035015670593321085,
      "grad_norm": 0.3694591522216797,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 324
    },
    {
      "epoch": 0.035123743650707875,
      "grad_norm": 0.3871074616909027,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 325
    },
    {
      "epoch": 0.03523181670809467,
      "grad_norm": 0.40115368366241455,
      "learning_rate": 8e-05,
      "loss": 1.9473,
      "step": 326
    },
    {
      "epoch": 0.03533988976548146,
      "grad_norm": 0.34573066234588623,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 327
    },
    {
      "epoch": 0.03544796282286826,
      "grad_norm": 0.36036252975463867,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 328
    },
    {
      "epoch": 0.03555603588025505,
      "grad_norm": 0.3882525563240051,
      "learning_rate": 8e-05,
      "loss": 1.8671,
      "step": 329
    },
    {
      "epoch": 0.035664108937641846,
      "grad_norm": 0.37284401059150696,
      "learning_rate": 8e-05,
      "loss": 1.8203,
      "step": 330
    },
    {
      "epoch": 0.035772181995028636,
      "grad_norm": 0.3881271183490753,
      "learning_rate": 8e-05,
      "loss": 1.775,
      "step": 331
    },
    {
      "epoch": 0.03588025505241543,
      "grad_norm": 0.4110361635684967,
      "learning_rate": 8e-05,
      "loss": 2.008,
      "step": 332
    },
    {
      "epoch": 0.03598832810980222,
      "grad_norm": 0.35942691564559937,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 333
    },
    {
      "epoch": 0.03609640116718902,
      "grad_norm": 0.3989504277706146,
      "learning_rate": 8e-05,
      "loss": 1.8574,
      "step": 334
    },
    {
      "epoch": 0.03620447422457582,
      "grad_norm": 0.40056127309799194,
      "learning_rate": 8e-05,
      "loss": 1.9016,
      "step": 335
    },
    {
      "epoch": 0.03631254728196261,
      "grad_norm": 0.3873950242996216,
      "learning_rate": 8e-05,
      "loss": 1.8326,
      "step": 336
    },
    {
      "epoch": 0.036420620339349404,
      "grad_norm": 0.3783419132232666,
      "learning_rate": 8e-05,
      "loss": 1.7917,
      "step": 337
    },
    {
      "epoch": 0.036528693396736193,
      "grad_norm": 0.4255695343017578,
      "learning_rate": 8e-05,
      "loss": 1.8075,
      "step": 338
    },
    {
      "epoch": 0.03663676645412299,
      "grad_norm": 0.42915570735931396,
      "learning_rate": 8e-05,
      "loss": 1.9533,
      "step": 339
    },
    {
      "epoch": 0.03674483951150978,
      "grad_norm": 0.4353823959827423,
      "learning_rate": 8e-05,
      "loss": 1.9186,
      "step": 340
    },
    {
      "epoch": 0.03685291256889658,
      "grad_norm": 0.40885356068611145,
      "learning_rate": 8e-05,
      "loss": 1.8309,
      "step": 341
    },
    {
      "epoch": 0.03696098562628337,
      "grad_norm": 0.3685755729675293,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 342
    },
    {
      "epoch": 0.037069058683670164,
      "grad_norm": 0.3689305782318115,
      "learning_rate": 8e-05,
      "loss": 1.7394,
      "step": 343
    },
    {
      "epoch": 0.037177131741056954,
      "grad_norm": 0.38720768690109253,
      "learning_rate": 8e-05,
      "loss": 1.8383,
      "step": 344
    },
    {
      "epoch": 0.03728520479844375,
      "grad_norm": 0.3552577495574951,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 345
    },
    {
      "epoch": 0.03739327785583054,
      "grad_norm": 0.4078991711139679,
      "learning_rate": 8e-05,
      "loss": 1.9238,
      "step": 346
    },
    {
      "epoch": 0.03750135091321734,
      "grad_norm": 0.38582518696784973,
      "learning_rate": 8e-05,
      "loss": 1.7514,
      "step": 347
    },
    {
      "epoch": 0.03760942397060413,
      "grad_norm": 0.3723450005054474,
      "learning_rate": 8e-05,
      "loss": 1.7024,
      "step": 348
    },
    {
      "epoch": 0.037717497027990925,
      "grad_norm": 0.40603965520858765,
      "learning_rate": 8e-05,
      "loss": 1.882,
      "step": 349
    },
    {
      "epoch": 0.037825570085377715,
      "grad_norm": 0.38014644384384155,
      "learning_rate": 8e-05,
      "loss": 1.8715,
      "step": 350
    },
    {
      "epoch": 0.03793364314276451,
      "grad_norm": 0.3696843087673187,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 351
    },
    {
      "epoch": 0.0380417162001513,
      "grad_norm": 0.37243345379829407,
      "learning_rate": 8e-05,
      "loss": 1.6032,
      "step": 352
    },
    {
      "epoch": 0.0381497892575381,
      "grad_norm": 0.4105185270309448,
      "learning_rate": 8e-05,
      "loss": 1.9086,
      "step": 353
    },
    {
      "epoch": 0.03825786231492489,
      "grad_norm": 0.37460798025131226,
      "learning_rate": 8e-05,
      "loss": 1.6184,
      "step": 354
    },
    {
      "epoch": 0.038365935372311685,
      "grad_norm": 0.40039703249931335,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 355
    },
    {
      "epoch": 0.038474008429698475,
      "grad_norm": 0.40266090631484985,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 356
    },
    {
      "epoch": 0.03858208148708527,
      "grad_norm": 0.3960215747356415,
      "learning_rate": 8e-05,
      "loss": 1.8895,
      "step": 357
    },
    {
      "epoch": 0.03869015454447206,
      "grad_norm": 0.3723769187927246,
      "learning_rate": 8e-05,
      "loss": 1.7954,
      "step": 358
    },
    {
      "epoch": 0.03879822760185886,
      "grad_norm": 0.3833503723144531,
      "learning_rate": 8e-05,
      "loss": 1.7724,
      "step": 359
    },
    {
      "epoch": 0.03890630065924565,
      "grad_norm": 0.37874698638916016,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 360
    },
    {
      "epoch": 0.039014373716632446,
      "grad_norm": 0.3774586319923401,
      "learning_rate": 8e-05,
      "loss": 1.8124,
      "step": 361
    },
    {
      "epoch": 0.039122446774019236,
      "grad_norm": 0.40446069836616516,
      "learning_rate": 8e-05,
      "loss": 1.3698,
      "step": 362
    },
    {
      "epoch": 0.03923051983140603,
      "grad_norm": 0.36500561237335205,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 363
    },
    {
      "epoch": 0.03933859288879282,
      "grad_norm": 0.3995642066001892,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 364
    },
    {
      "epoch": 0.03944666594617962,
      "grad_norm": 0.3842376470565796,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 365
    },
    {
      "epoch": 0.03955473900356641,
      "grad_norm": 0.36607101559638977,
      "learning_rate": 8e-05,
      "loss": 1.6731,
      "step": 366
    },
    {
      "epoch": 0.039662812060953206,
      "grad_norm": 0.3802371323108673,
      "learning_rate": 8e-05,
      "loss": 1.7763,
      "step": 367
    },
    {
      "epoch": 0.039770885118339996,
      "grad_norm": 0.3834855258464813,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 368
    },
    {
      "epoch": 0.03987895817572679,
      "grad_norm": 0.4152221381664276,
      "learning_rate": 8e-05,
      "loss": 1.7594,
      "step": 369
    },
    {
      "epoch": 0.03998703123311358,
      "grad_norm": 0.37589240074157715,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 370
    },
    {
      "epoch": 0.04009510429050038,
      "grad_norm": 0.43800219893455505,
      "learning_rate": 8e-05,
      "loss": 1.835,
      "step": 371
    },
    {
      "epoch": 0.04020317734788717,
      "grad_norm": 0.397070050239563,
      "learning_rate": 8e-05,
      "loss": 1.9621,
      "step": 372
    },
    {
      "epoch": 0.04031125040527397,
      "grad_norm": 0.43010687828063965,
      "learning_rate": 8e-05,
      "loss": 1.8171,
      "step": 373
    },
    {
      "epoch": 0.04041932346266076,
      "grad_norm": 0.4090173840522766,
      "learning_rate": 8e-05,
      "loss": 2.0476,
      "step": 374
    },
    {
      "epoch": 0.040527396520047554,
      "grad_norm": 0.37031230330467224,
      "learning_rate": 8e-05,
      "loss": 1.9466,
      "step": 375
    },
    {
      "epoch": 0.040635469577434344,
      "grad_norm": 0.41040468215942383,
      "learning_rate": 8e-05,
      "loss": 1.8215,
      "step": 376
    },
    {
      "epoch": 0.04074354263482114,
      "grad_norm": 0.37586167454719543,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 377
    },
    {
      "epoch": 0.04085161569220793,
      "grad_norm": 0.39795875549316406,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 378
    },
    {
      "epoch": 0.04095968874959473,
      "grad_norm": 0.4134315252304077,
      "learning_rate": 8e-05,
      "loss": 1.8278,
      "step": 379
    },
    {
      "epoch": 0.04106776180698152,
      "grad_norm": 0.3780001699924469,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 380
    },
    {
      "epoch": 0.041175834864368314,
      "grad_norm": 0.40022245049476624,
      "learning_rate": 8e-05,
      "loss": 1.8331,
      "step": 381
    },
    {
      "epoch": 0.041283907921755104,
      "grad_norm": 0.37059682607650757,
      "learning_rate": 8e-05,
      "loss": 1.7966,
      "step": 382
    },
    {
      "epoch": 0.0413919809791419,
      "grad_norm": 0.39825180172920227,
      "learning_rate": 8e-05,
      "loss": 1.8483,
      "step": 383
    },
    {
      "epoch": 0.04150005403652869,
      "grad_norm": 0.3850351870059967,
      "learning_rate": 8e-05,
      "loss": 1.8579,
      "step": 384
    },
    {
      "epoch": 0.04160812709391549,
      "grad_norm": 0.36618584394454956,
      "learning_rate": 8e-05,
      "loss": 1.6181,
      "step": 385
    },
    {
      "epoch": 0.04171620015130228,
      "grad_norm": 0.39024895429611206,
      "learning_rate": 8e-05,
      "loss": 1.8127,
      "step": 386
    },
    {
      "epoch": 0.041824273208689075,
      "grad_norm": 0.41545161604881287,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 387
    },
    {
      "epoch": 0.041932346266075865,
      "grad_norm": 0.38076847791671753,
      "learning_rate": 8e-05,
      "loss": 1.8583,
      "step": 388
    },
    {
      "epoch": 0.04204041932346266,
      "grad_norm": 0.37488022446632385,
      "learning_rate": 8e-05,
      "loss": 1.6256,
      "step": 389
    },
    {
      "epoch": 0.04214849238084945,
      "grad_norm": 0.3739245533943176,
      "learning_rate": 8e-05,
      "loss": 1.6497,
      "step": 390
    },
    {
      "epoch": 0.04225656543823625,
      "grad_norm": 0.4112280607223511,
      "learning_rate": 8e-05,
      "loss": 1.9784,
      "step": 391
    },
    {
      "epoch": 0.04236463849562304,
      "grad_norm": 0.3635033071041107,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 392
    },
    {
      "epoch": 0.042472711553009836,
      "grad_norm": 0.3553190529346466,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 393
    },
    {
      "epoch": 0.042580784610396626,
      "grad_norm": 0.39443516731262207,
      "learning_rate": 8e-05,
      "loss": 1.7807,
      "step": 394
    },
    {
      "epoch": 0.04268885766778342,
      "grad_norm": 0.4253872036933899,
      "learning_rate": 8e-05,
      "loss": 1.8484,
      "step": 395
    },
    {
      "epoch": 0.04279693072517021,
      "grad_norm": 0.3850438892841339,
      "learning_rate": 8e-05,
      "loss": 1.8233,
      "step": 396
    },
    {
      "epoch": 0.04290500378255701,
      "grad_norm": 0.3941638469696045,
      "learning_rate": 8e-05,
      "loss": 1.9134,
      "step": 397
    },
    {
      "epoch": 0.0430130768399438,
      "grad_norm": 0.38364243507385254,
      "learning_rate": 8e-05,
      "loss": 1.7757,
      "step": 398
    },
    {
      "epoch": 0.043121149897330596,
      "grad_norm": 0.3588034510612488,
      "learning_rate": 8e-05,
      "loss": 1.7323,
      "step": 399
    },
    {
      "epoch": 0.043229222954717386,
      "grad_norm": 0.40070948004722595,
      "learning_rate": 8e-05,
      "loss": 1.8595,
      "step": 400
    },
    {
      "epoch": 0.04333729601210418,
      "grad_norm": 0.3728334605693817,
      "learning_rate": 8e-05,
      "loss": 1.8523,
      "step": 401
    },
    {
      "epoch": 0.04344536906949097,
      "grad_norm": 0.3865524232387543,
      "learning_rate": 8e-05,
      "loss": 1.8993,
      "step": 402
    },
    {
      "epoch": 0.04355344212687777,
      "grad_norm": 0.3635477125644684,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 403
    },
    {
      "epoch": 0.04366151518426456,
      "grad_norm": 0.3658127188682556,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 404
    },
    {
      "epoch": 0.04376958824165136,
      "grad_norm": 0.38784506916999817,
      "learning_rate": 8e-05,
      "loss": 1.8504,
      "step": 405
    },
    {
      "epoch": 0.04387766129903815,
      "grad_norm": 0.39218786358833313,
      "learning_rate": 8e-05,
      "loss": 1.8261,
      "step": 406
    },
    {
      "epoch": 0.043985734356424944,
      "grad_norm": 0.3855704665184021,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 407
    },
    {
      "epoch": 0.044093807413811734,
      "grad_norm": 0.4157221019268036,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 408
    },
    {
      "epoch": 0.04420188047119853,
      "grad_norm": 0.43718111515045166,
      "learning_rate": 8e-05,
      "loss": 1.8945,
      "step": 409
    },
    {
      "epoch": 0.04430995352858532,
      "grad_norm": 0.3822411000728607,
      "learning_rate": 8e-05,
      "loss": 1.8144,
      "step": 410
    },
    {
      "epoch": 0.04441802658597212,
      "grad_norm": 0.4077363610267639,
      "learning_rate": 8e-05,
      "loss": 1.8817,
      "step": 411
    },
    {
      "epoch": 0.04452609964335891,
      "grad_norm": 0.35306328535079956,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 412
    },
    {
      "epoch": 0.044634172700745704,
      "grad_norm": 0.4037925899028778,
      "learning_rate": 8e-05,
      "loss": 1.8688,
      "step": 413
    },
    {
      "epoch": 0.044742245758132494,
      "grad_norm": 0.39644160866737366,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 414
    },
    {
      "epoch": 0.04485031881551929,
      "grad_norm": 0.38850751519203186,
      "learning_rate": 8e-05,
      "loss": 1.8966,
      "step": 415
    },
    {
      "epoch": 0.04495839187290608,
      "grad_norm": 0.3764535188674927,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 416
    },
    {
      "epoch": 0.04506646493029288,
      "grad_norm": 0.3724212050437927,
      "learning_rate": 8e-05,
      "loss": 1.8014,
      "step": 417
    },
    {
      "epoch": 0.045174537987679675,
      "grad_norm": 0.37624552845954895,
      "learning_rate": 8e-05,
      "loss": 1.6876,
      "step": 418
    },
    {
      "epoch": 0.045282611045066465,
      "grad_norm": 0.4294884204864502,
      "learning_rate": 8e-05,
      "loss": 1.9347,
      "step": 419
    },
    {
      "epoch": 0.04539068410245326,
      "grad_norm": 0.40191271901130676,
      "learning_rate": 8e-05,
      "loss": 1.6823,
      "step": 420
    },
    {
      "epoch": 0.04549875715984005,
      "grad_norm": 0.37208858132362366,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 421
    },
    {
      "epoch": 0.04560683021722685,
      "grad_norm": 0.3788667321205139,
      "learning_rate": 8e-05,
      "loss": 1.8079,
      "step": 422
    },
    {
      "epoch": 0.04571490327461364,
      "grad_norm": 0.3846014142036438,
      "learning_rate": 8e-05,
      "loss": 1.8295,
      "step": 423
    },
    {
      "epoch": 0.045822976332000435,
      "grad_norm": 0.36874285340309143,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 424
    },
    {
      "epoch": 0.045931049389387225,
      "grad_norm": 0.3773528039455414,
      "learning_rate": 8e-05,
      "loss": 1.8967,
      "step": 425
    },
    {
      "epoch": 0.04603912244677402,
      "grad_norm": 0.395060658454895,
      "learning_rate": 8e-05,
      "loss": 1.9069,
      "step": 426
    },
    {
      "epoch": 0.04614719550416081,
      "grad_norm": 0.397066593170166,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 427
    },
    {
      "epoch": 0.04625526856154761,
      "grad_norm": 0.40131664276123047,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 428
    },
    {
      "epoch": 0.0463633416189344,
      "grad_norm": 0.3970150649547577,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 429
    },
    {
      "epoch": 0.046471414676321196,
      "grad_norm": 0.3858741819858551,
      "learning_rate": 8e-05,
      "loss": 1.8595,
      "step": 430
    },
    {
      "epoch": 0.046579487733707986,
      "grad_norm": 0.35100990533828735,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 431
    },
    {
      "epoch": 0.04668756079109478,
      "grad_norm": 0.3665204644203186,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 432
    },
    {
      "epoch": 0.04679563384848157,
      "grad_norm": 0.389402836561203,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 433
    },
    {
      "epoch": 0.04690370690586837,
      "grad_norm": 0.3686143457889557,
      "learning_rate": 8e-05,
      "loss": 1.8754,
      "step": 434
    },
    {
      "epoch": 0.04701177996325516,
      "grad_norm": 0.3859221339225769,
      "learning_rate": 8e-05,
      "loss": 1.88,
      "step": 435
    },
    {
      "epoch": 0.04711985302064196,
      "grad_norm": 0.374782532453537,
      "learning_rate": 8e-05,
      "loss": 1.7727,
      "step": 436
    },
    {
      "epoch": 0.04722792607802875,
      "grad_norm": 0.39135491847991943,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 437
    },
    {
      "epoch": 0.047335999135415543,
      "grad_norm": 0.4159727394580841,
      "learning_rate": 8e-05,
      "loss": 1.96,
      "step": 438
    },
    {
      "epoch": 0.04744407219280233,
      "grad_norm": 0.38746562600135803,
      "learning_rate": 8e-05,
      "loss": 1.8876,
      "step": 439
    },
    {
      "epoch": 0.04755214525018913,
      "grad_norm": 0.4302104115486145,
      "learning_rate": 8e-05,
      "loss": 1.9793,
      "step": 440
    },
    {
      "epoch": 0.04766021830757592,
      "grad_norm": 0.42846155166625977,
      "learning_rate": 8e-05,
      "loss": 1.9228,
      "step": 441
    },
    {
      "epoch": 0.04776829136496272,
      "grad_norm": 0.41069167852401733,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 442
    },
    {
      "epoch": 0.04787636442234951,
      "grad_norm": 0.40765511989593506,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 443
    },
    {
      "epoch": 0.047984437479736304,
      "grad_norm": 0.3719295561313629,
      "learning_rate": 8e-05,
      "loss": 1.7205,
      "step": 444
    },
    {
      "epoch": 0.048092510537123094,
      "grad_norm": 0.41722410917282104,
      "learning_rate": 8e-05,
      "loss": 1.8728,
      "step": 445
    },
    {
      "epoch": 0.04820058359450989,
      "grad_norm": 0.41826048493385315,
      "learning_rate": 8e-05,
      "loss": 1.8338,
      "step": 446
    },
    {
      "epoch": 0.04830865665189668,
      "grad_norm": 0.3610713481903076,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 447
    },
    {
      "epoch": 0.04841672970928348,
      "grad_norm": 0.3895103335380554,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 448
    },
    {
      "epoch": 0.04852480276667027,
      "grad_norm": 0.35636061429977417,
      "learning_rate": 8e-05,
      "loss": 1.5445,
      "step": 449
    },
    {
      "epoch": 0.048632875824057065,
      "grad_norm": 0.4193304181098938,
      "learning_rate": 8e-05,
      "loss": 1.9749,
      "step": 450
    },
    {
      "epoch": 0.048740948881443855,
      "grad_norm": 0.39705803990364075,
      "learning_rate": 8e-05,
      "loss": 1.8425,
      "step": 451
    },
    {
      "epoch": 0.04884902193883065,
      "grad_norm": 0.3794572353363037,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 452
    },
    {
      "epoch": 0.04895709499621744,
      "grad_norm": 0.3695708215236664,
      "learning_rate": 8e-05,
      "loss": 1.8051,
      "step": 453
    },
    {
      "epoch": 0.04906516805360424,
      "grad_norm": 0.4181847870349884,
      "learning_rate": 8e-05,
      "loss": 1.9491,
      "step": 454
    },
    {
      "epoch": 0.04917324111099103,
      "grad_norm": 0.3856743574142456,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 455
    },
    {
      "epoch": 0.049281314168377825,
      "grad_norm": 0.4461413621902466,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 456
    },
    {
      "epoch": 0.049389387225764615,
      "grad_norm": 0.41560980677604675,
      "learning_rate": 8e-05,
      "loss": 2.0494,
      "step": 457
    },
    {
      "epoch": 0.04949746028315141,
      "grad_norm": 0.3937462568283081,
      "learning_rate": 8e-05,
      "loss": 1.9181,
      "step": 458
    },
    {
      "epoch": 0.0496055333405382,
      "grad_norm": 0.3602689206600189,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 459
    },
    {
      "epoch": 0.049713606397925,
      "grad_norm": 0.3674399256706238,
      "learning_rate": 8e-05,
      "loss": 1.8149,
      "step": 460
    },
    {
      "epoch": 0.04982167945531179,
      "grad_norm": 0.39698418974876404,
      "learning_rate": 8e-05,
      "loss": 1.7706,
      "step": 461
    },
    {
      "epoch": 0.049929752512698586,
      "grad_norm": 0.40804529190063477,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 462
    },
    {
      "epoch": 0.050037825570085376,
      "grad_norm": 0.35640743374824524,
      "learning_rate": 8e-05,
      "loss": 1.5543,
      "step": 463
    },
    {
      "epoch": 0.05014589862747217,
      "grad_norm": 0.3880807161331177,
      "learning_rate": 8e-05,
      "loss": 1.709,
      "step": 464
    },
    {
      "epoch": 0.05025397168485896,
      "grad_norm": 0.3852868676185608,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 465
    },
    {
      "epoch": 0.05036204474224576,
      "grad_norm": 0.4220997989177704,
      "learning_rate": 8e-05,
      "loss": 2.0106,
      "step": 466
    },
    {
      "epoch": 0.05047011779963255,
      "grad_norm": 0.36986929178237915,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 467
    },
    {
      "epoch": 0.050578190857019346,
      "grad_norm": 0.3972550630569458,
      "learning_rate": 8e-05,
      "loss": 1.9064,
      "step": 468
    },
    {
      "epoch": 0.050686263914406136,
      "grad_norm": 0.3819347023963928,
      "learning_rate": 8e-05,
      "loss": 1.7779,
      "step": 469
    },
    {
      "epoch": 0.05079433697179293,
      "grad_norm": 0.3585180938243866,
      "learning_rate": 8e-05,
      "loss": 1.7928,
      "step": 470
    },
    {
      "epoch": 0.05090241002917972,
      "grad_norm": 0.4086778163909912,
      "learning_rate": 8e-05,
      "loss": 1.973,
      "step": 471
    },
    {
      "epoch": 0.05101048308656652,
      "grad_norm": 0.4315330684185028,
      "learning_rate": 8e-05,
      "loss": 1.8072,
      "step": 472
    },
    {
      "epoch": 0.05111855614395331,
      "grad_norm": 0.4038466513156891,
      "learning_rate": 8e-05,
      "loss": 1.8291,
      "step": 473
    },
    {
      "epoch": 0.05122662920134011,
      "grad_norm": 0.3962635397911072,
      "learning_rate": 8e-05,
      "loss": 1.8123,
      "step": 474
    },
    {
      "epoch": 0.0513347022587269,
      "grad_norm": 0.40047988295555115,
      "learning_rate": 8e-05,
      "loss": 2.1662,
      "step": 475
    },
    {
      "epoch": 0.051442775316113694,
      "grad_norm": 0.4026619493961334,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 476
    },
    {
      "epoch": 0.051550848373500484,
      "grad_norm": 0.3865167200565338,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 477
    },
    {
      "epoch": 0.05165892143088728,
      "grad_norm": 0.3880349397659302,
      "learning_rate": 8e-05,
      "loss": 1.9196,
      "step": 478
    },
    {
      "epoch": 0.05176699448827407,
      "grad_norm": 0.3584767282009125,
      "learning_rate": 8e-05,
      "loss": 1.795,
      "step": 479
    },
    {
      "epoch": 0.05187506754566087,
      "grad_norm": 0.3752886652946472,
      "learning_rate": 8e-05,
      "loss": 1.8087,
      "step": 480
    },
    {
      "epoch": 0.05198314060304766,
      "grad_norm": 0.3656434118747711,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 481
    },
    {
      "epoch": 0.052091213660434454,
      "grad_norm": 0.4040050208568573,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 482
    },
    {
      "epoch": 0.052199286717821244,
      "grad_norm": 0.43846285343170166,
      "learning_rate": 8e-05,
      "loss": 1.8349,
      "step": 483
    },
    {
      "epoch": 0.05230735977520804,
      "grad_norm": 0.3863964378833771,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 484
    },
    {
      "epoch": 0.05241543283259483,
      "grad_norm": 0.35015442967414856,
      "learning_rate": 8e-05,
      "loss": 1.7225,
      "step": 485
    },
    {
      "epoch": 0.05252350588998163,
      "grad_norm": 0.42012444138526917,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 486
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 0.41372033953666687,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 487
    },
    {
      "epoch": 0.052739652004755215,
      "grad_norm": 0.39479345083236694,
      "learning_rate": 8e-05,
      "loss": 1.7406,
      "step": 488
    },
    {
      "epoch": 0.052847725062142005,
      "grad_norm": 0.37600812315940857,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 489
    },
    {
      "epoch": 0.0529557981195288,
      "grad_norm": 0.3889562487602234,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 490
    },
    {
      "epoch": 0.05306387117691559,
      "grad_norm": 0.4306604564189911,
      "learning_rate": 8e-05,
      "loss": 1.9344,
      "step": 491
    },
    {
      "epoch": 0.05317194423430239,
      "grad_norm": 0.3745219111442566,
      "learning_rate": 8e-05,
      "loss": 1.7519,
      "step": 492
    },
    {
      "epoch": 0.05328001729168918,
      "grad_norm": 0.4172973930835724,
      "learning_rate": 8e-05,
      "loss": 1.9364,
      "step": 493
    },
    {
      "epoch": 0.053388090349075976,
      "grad_norm": 0.38838228583335876,
      "learning_rate": 8e-05,
      "loss": 1.8311,
      "step": 494
    },
    {
      "epoch": 0.053496163406462766,
      "grad_norm": 0.3883071839809418,
      "learning_rate": 8e-05,
      "loss": 1.7687,
      "step": 495
    },
    {
      "epoch": 0.05360423646384956,
      "grad_norm": 0.4117523431777954,
      "learning_rate": 8e-05,
      "loss": 1.8404,
      "step": 496
    },
    {
      "epoch": 0.05371230952123635,
      "grad_norm": 0.4005739986896515,
      "learning_rate": 8e-05,
      "loss": 1.9648,
      "step": 497
    },
    {
      "epoch": 0.05382038257862315,
      "grad_norm": 0.38154447078704834,
      "learning_rate": 8e-05,
      "loss": 1.812,
      "step": 498
    },
    {
      "epoch": 0.05392845563600994,
      "grad_norm": 0.4300518333911896,
      "learning_rate": 8e-05,
      "loss": 1.8817,
      "step": 499
    },
    {
      "epoch": 0.054036528693396736,
      "grad_norm": 0.363741934299469,
      "learning_rate": 8e-05,
      "loss": 1.7924,
      "step": 500
    },
    {
      "epoch": 0.05414460175078353,
      "grad_norm": 0.3813784420490265,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 501
    },
    {
      "epoch": 0.05425267480817032,
      "grad_norm": 0.4035930037498474,
      "learning_rate": 8e-05,
      "loss": 1.958,
      "step": 502
    },
    {
      "epoch": 0.05436074786555712,
      "grad_norm": 0.37766215205192566,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 503
    },
    {
      "epoch": 0.05446882092294391,
      "grad_norm": 0.3796464800834656,
      "learning_rate": 8e-05,
      "loss": 1.9206,
      "step": 504
    },
    {
      "epoch": 0.05457689398033071,
      "grad_norm": 0.376295804977417,
      "learning_rate": 8e-05,
      "loss": 1.9684,
      "step": 505
    },
    {
      "epoch": 0.0546849670377175,
      "grad_norm": 0.37519073486328125,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 506
    },
    {
      "epoch": 0.054793040095104294,
      "grad_norm": 0.37861186265945435,
      "learning_rate": 8e-05,
      "loss": 1.7127,
      "step": 507
    },
    {
      "epoch": 0.054901113152491084,
      "grad_norm": 0.37276217341423035,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 508
    },
    {
      "epoch": 0.05500918620987788,
      "grad_norm": 0.368045449256897,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 509
    },
    {
      "epoch": 0.05511725926726467,
      "grad_norm": 0.4535576403141022,
      "learning_rate": 8e-05,
      "loss": 1.8451,
      "step": 510
    },
    {
      "epoch": 0.05522533232465147,
      "grad_norm": 0.40646564960479736,
      "learning_rate": 8e-05,
      "loss": 1.9395,
      "step": 511
    },
    {
      "epoch": 0.05533340538203826,
      "grad_norm": 0.3889763653278351,
      "learning_rate": 8e-05,
      "loss": 1.7983,
      "step": 512
    },
    {
      "epoch": 0.055441478439425054,
      "grad_norm": 0.3923338055610657,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 513
    },
    {
      "epoch": 0.055549551496811844,
      "grad_norm": 0.36657455563545227,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 514
    },
    {
      "epoch": 0.05565762455419864,
      "grad_norm": 0.37522923946380615,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 515
    },
    {
      "epoch": 0.05576569761158543,
      "grad_norm": 0.37318137288093567,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 516
    },
    {
      "epoch": 0.05587377066897223,
      "grad_norm": 0.4317561089992523,
      "learning_rate": 8e-05,
      "loss": 2.0737,
      "step": 517
    },
    {
      "epoch": 0.05598184372635902,
      "grad_norm": 0.38995981216430664,
      "learning_rate": 8e-05,
      "loss": 1.9067,
      "step": 518
    },
    {
      "epoch": 0.056089916783745815,
      "grad_norm": 0.362949013710022,
      "learning_rate": 8e-05,
      "loss": 1.7898,
      "step": 519
    },
    {
      "epoch": 0.056197989841132605,
      "grad_norm": 0.3555454909801483,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 520
    },
    {
      "epoch": 0.0563060628985194,
      "grad_norm": 0.3759154975414276,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 521
    },
    {
      "epoch": 0.05641413595590619,
      "grad_norm": 0.3732134997844696,
      "learning_rate": 8e-05,
      "loss": 1.846,
      "step": 522
    },
    {
      "epoch": 0.05652220901329299,
      "grad_norm": 0.41401079297065735,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 523
    },
    {
      "epoch": 0.05663028207067978,
      "grad_norm": 0.38953521847724915,
      "learning_rate": 8e-05,
      "loss": 1.8102,
      "step": 524
    },
    {
      "epoch": 0.056738355128066575,
      "grad_norm": 0.405595600605011,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 525
    },
    {
      "epoch": 0.056846428185453365,
      "grad_norm": 0.3897184133529663,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 526
    },
    {
      "epoch": 0.05695450124284016,
      "grad_norm": 0.3700886368751526,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 527
    },
    {
      "epoch": 0.05706257430022695,
      "grad_norm": 0.412516325712204,
      "learning_rate": 8e-05,
      "loss": 1.7578,
      "step": 528
    },
    {
      "epoch": 0.05717064735761375,
      "grad_norm": 0.3819820284843445,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 529
    },
    {
      "epoch": 0.05727872041500054,
      "grad_norm": 0.3711232542991638,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 530
    },
    {
      "epoch": 0.057386793472387336,
      "grad_norm": 0.3691806197166443,
      "learning_rate": 8e-05,
      "loss": 1.7707,
      "step": 531
    },
    {
      "epoch": 0.057494866529774126,
      "grad_norm": 0.38428232073783875,
      "learning_rate": 8e-05,
      "loss": 1.9595,
      "step": 532
    },
    {
      "epoch": 0.05760293958716092,
      "grad_norm": 0.38135918974876404,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 533
    },
    {
      "epoch": 0.05771101264454771,
      "grad_norm": 0.36597949266433716,
      "learning_rate": 8e-05,
      "loss": 1.8373,
      "step": 534
    },
    {
      "epoch": 0.05781908570193451,
      "grad_norm": 0.3876015841960907,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 535
    },
    {
      "epoch": 0.0579271587593213,
      "grad_norm": 0.39781370759010315,
      "learning_rate": 8e-05,
      "loss": 1.924,
      "step": 536
    },
    {
      "epoch": 0.0580352318167081,
      "grad_norm": 0.447880357503891,
      "learning_rate": 8e-05,
      "loss": 1.9407,
      "step": 537
    },
    {
      "epoch": 0.05814330487409489,
      "grad_norm": 0.41973045468330383,
      "learning_rate": 8e-05,
      "loss": 1.9179,
      "step": 538
    },
    {
      "epoch": 0.058251377931481683,
      "grad_norm": 0.4000081717967987,
      "learning_rate": 8e-05,
      "loss": 1.7945,
      "step": 539
    },
    {
      "epoch": 0.05835945098886847,
      "grad_norm": 0.4480554163455963,
      "learning_rate": 8e-05,
      "loss": 1.9586,
      "step": 540
    },
    {
      "epoch": 0.05846752404625527,
      "grad_norm": 0.38606417179107666,
      "learning_rate": 8e-05,
      "loss": 1.7578,
      "step": 541
    },
    {
      "epoch": 0.05857559710364206,
      "grad_norm": 0.3771899342536926,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 542
    },
    {
      "epoch": 0.05868367016102886,
      "grad_norm": 0.4201655089855194,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 543
    },
    {
      "epoch": 0.05879174321841565,
      "grad_norm": 0.4522247314453125,
      "learning_rate": 8e-05,
      "loss": 1.9192,
      "step": 544
    },
    {
      "epoch": 0.058899816275802444,
      "grad_norm": 0.4100610911846161,
      "learning_rate": 8e-05,
      "loss": 1.9789,
      "step": 545
    },
    {
      "epoch": 0.059007889333189234,
      "grad_norm": 0.4234224259853363,
      "learning_rate": 8e-05,
      "loss": 1.8208,
      "step": 546
    },
    {
      "epoch": 0.05911596239057603,
      "grad_norm": 0.4051947295665741,
      "learning_rate": 8e-05,
      "loss": 1.9268,
      "step": 547
    },
    {
      "epoch": 0.05922403544796282,
      "grad_norm": 0.41093116998672485,
      "learning_rate": 8e-05,
      "loss": 1.8171,
      "step": 548
    },
    {
      "epoch": 0.05933210850534962,
      "grad_norm": 0.3778577744960785,
      "learning_rate": 8e-05,
      "loss": 1.8015,
      "step": 549
    },
    {
      "epoch": 0.05944018156273641,
      "grad_norm": 0.4059986174106598,
      "learning_rate": 8e-05,
      "loss": 2.0483,
      "step": 550
    },
    {
      "epoch": 0.059548254620123205,
      "grad_norm": 0.4149869382381439,
      "learning_rate": 8e-05,
      "loss": 1.9023,
      "step": 551
    },
    {
      "epoch": 0.059656327677509995,
      "grad_norm": 0.40040323138237,
      "learning_rate": 8e-05,
      "loss": 1.8701,
      "step": 552
    },
    {
      "epoch": 0.05976440073489679,
      "grad_norm": 0.3790455162525177,
      "learning_rate": 8e-05,
      "loss": 1.8952,
      "step": 553
    },
    {
      "epoch": 0.05987247379228358,
      "grad_norm": 0.3911406099796295,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 554
    },
    {
      "epoch": 0.05998054684967038,
      "grad_norm": 0.37344563007354736,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 555
    },
    {
      "epoch": 0.06008861990705717,
      "grad_norm": 0.40307167172431946,
      "learning_rate": 8e-05,
      "loss": 1.7686,
      "step": 556
    },
    {
      "epoch": 0.060196692964443965,
      "grad_norm": 0.43773356080055237,
      "learning_rate": 8e-05,
      "loss": 2.0326,
      "step": 557
    },
    {
      "epoch": 0.060304766021830755,
      "grad_norm": 0.3883941173553467,
      "learning_rate": 8e-05,
      "loss": 1.9221,
      "step": 558
    },
    {
      "epoch": 0.06041283907921755,
      "grad_norm": 0.42291250824928284,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 559
    },
    {
      "epoch": 0.06052091213660434,
      "grad_norm": 0.383587121963501,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 560
    },
    {
      "epoch": 0.06062898519399114,
      "grad_norm": 0.3637489974498749,
      "learning_rate": 8e-05,
      "loss": 1.7317,
      "step": 561
    },
    {
      "epoch": 0.06073705825137793,
      "grad_norm": 0.3671761751174927,
      "learning_rate": 8e-05,
      "loss": 1.8846,
      "step": 562
    },
    {
      "epoch": 0.060845131308764726,
      "grad_norm": 0.35723164677619934,
      "learning_rate": 8e-05,
      "loss": 1.8169,
      "step": 563
    },
    {
      "epoch": 0.060953204366151516,
      "grad_norm": 0.3727544844150543,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 564
    },
    {
      "epoch": 0.06106127742353831,
      "grad_norm": 0.38152262568473816,
      "learning_rate": 8e-05,
      "loss": 1.9426,
      "step": 565
    },
    {
      "epoch": 0.0611693504809251,
      "grad_norm": 0.43421468138694763,
      "learning_rate": 8e-05,
      "loss": 2.0618,
      "step": 566
    },
    {
      "epoch": 0.0612774235383119,
      "grad_norm": 0.3680697977542877,
      "learning_rate": 8e-05,
      "loss": 1.8818,
      "step": 567
    },
    {
      "epoch": 0.06138549659569869,
      "grad_norm": 0.4068283438682556,
      "learning_rate": 8e-05,
      "loss": 1.8411,
      "step": 568
    },
    {
      "epoch": 0.061493569653085486,
      "grad_norm": 0.38021689653396606,
      "learning_rate": 8e-05,
      "loss": 1.87,
      "step": 569
    },
    {
      "epoch": 0.061601642710472276,
      "grad_norm": 0.45228666067123413,
      "learning_rate": 8e-05,
      "loss": 1.963,
      "step": 570
    },
    {
      "epoch": 0.06170971576785907,
      "grad_norm": 0.3976708948612213,
      "learning_rate": 8e-05,
      "loss": 1.983,
      "step": 571
    },
    {
      "epoch": 0.06181778882524586,
      "grad_norm": 0.3886009752750397,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 572
    },
    {
      "epoch": 0.06192586188263266,
      "grad_norm": 0.37844374775886536,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 573
    },
    {
      "epoch": 0.06203393494001945,
      "grad_norm": 0.36841094493865967,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 574
    },
    {
      "epoch": 0.06214200799740625,
      "grad_norm": 0.38314294815063477,
      "learning_rate": 8e-05,
      "loss": 1.8936,
      "step": 575
    },
    {
      "epoch": 0.06225008105479304,
      "grad_norm": 0.3733077049255371,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 576
    },
    {
      "epoch": 0.062358154112179834,
      "grad_norm": 0.4181191325187683,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 577
    },
    {
      "epoch": 0.062466227169566624,
      "grad_norm": 0.37213918566703796,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 578
    },
    {
      "epoch": 0.06257430022695341,
      "grad_norm": 0.4090023636817932,
      "learning_rate": 8e-05,
      "loss": 1.9122,
      "step": 579
    },
    {
      "epoch": 0.06268237328434022,
      "grad_norm": 0.3619495630264282,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 580
    },
    {
      "epoch": 0.06279044634172701,
      "grad_norm": 0.362935334444046,
      "learning_rate": 8e-05,
      "loss": 1.5268,
      "step": 581
    },
    {
      "epoch": 0.0628985193991138,
      "grad_norm": 0.4175827205181122,
      "learning_rate": 8e-05,
      "loss": 1.9571,
      "step": 582
    },
    {
      "epoch": 0.06300659245650059,
      "grad_norm": 0.40070852637290955,
      "learning_rate": 8e-05,
      "loss": 1.7267,
      "step": 583
    },
    {
      "epoch": 0.06311466551388739,
      "grad_norm": 0.37877357006073,
      "learning_rate": 8e-05,
      "loss": 1.8666,
      "step": 584
    },
    {
      "epoch": 0.06322273857127418,
      "grad_norm": 0.4106859564781189,
      "learning_rate": 8e-05,
      "loss": 1.7925,
      "step": 585
    },
    {
      "epoch": 0.06333081162866097,
      "grad_norm": 0.3773735761642456,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 586
    },
    {
      "epoch": 0.06343888468604778,
      "grad_norm": 0.4011212885379791,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 587
    },
    {
      "epoch": 0.06354695774343457,
      "grad_norm": 0.4007323980331421,
      "learning_rate": 8e-05,
      "loss": 1.7764,
      "step": 588
    },
    {
      "epoch": 0.06365503080082136,
      "grad_norm": 0.41673195362091064,
      "learning_rate": 8e-05,
      "loss": 1.9663,
      "step": 589
    },
    {
      "epoch": 0.06376310385820814,
      "grad_norm": 0.40779486298561096,
      "learning_rate": 8e-05,
      "loss": 1.9656,
      "step": 590
    },
    {
      "epoch": 0.06387117691559495,
      "grad_norm": 0.3688437342643738,
      "learning_rate": 8e-05,
      "loss": 1.8285,
      "step": 591
    },
    {
      "epoch": 0.06397924997298174,
      "grad_norm": 0.4052095115184784,
      "learning_rate": 8e-05,
      "loss": 1.7549,
      "step": 592
    },
    {
      "epoch": 0.06408732303036853,
      "grad_norm": 0.38654428720474243,
      "learning_rate": 8e-05,
      "loss": 1.9232,
      "step": 593
    },
    {
      "epoch": 0.06419539608775532,
      "grad_norm": 0.3872584104537964,
      "learning_rate": 8e-05,
      "loss": 1.8093,
      "step": 594
    },
    {
      "epoch": 0.06430346914514212,
      "grad_norm": 0.3800426423549652,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 595
    },
    {
      "epoch": 0.06441154220252891,
      "grad_norm": 0.39111658930778503,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 596
    },
    {
      "epoch": 0.0645196152599157,
      "grad_norm": 0.3776978552341461,
      "learning_rate": 8e-05,
      "loss": 1.7043,
      "step": 597
    },
    {
      "epoch": 0.06462768831730249,
      "grad_norm": 0.366039514541626,
      "learning_rate": 8e-05,
      "loss": 1.7973,
      "step": 598
    },
    {
      "epoch": 0.0647357613746893,
      "grad_norm": 0.4075569212436676,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 599
    },
    {
      "epoch": 0.06484383443207609,
      "grad_norm": 0.4031198024749756,
      "learning_rate": 8e-05,
      "loss": 1.8905,
      "step": 600
    },
    {
      "epoch": 0.06495190748946288,
      "grad_norm": 0.37056124210357666,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 601
    },
    {
      "epoch": 0.06505998054684967,
      "grad_norm": 0.4212067723274231,
      "learning_rate": 8e-05,
      "loss": 1.813,
      "step": 602
    },
    {
      "epoch": 0.06516805360423647,
      "grad_norm": 0.3861415982246399,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 603
    },
    {
      "epoch": 0.06527612666162326,
      "grad_norm": 0.346788614988327,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 604
    },
    {
      "epoch": 0.06538419971901005,
      "grad_norm": 0.42646998167037964,
      "learning_rate": 8e-05,
      "loss": 1.8703,
      "step": 605
    },
    {
      "epoch": 0.06549227277639684,
      "grad_norm": 0.39374110102653503,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 606
    },
    {
      "epoch": 0.06560034583378364,
      "grad_norm": 0.4954928755760193,
      "learning_rate": 8e-05,
      "loss": 2.0197,
      "step": 607
    },
    {
      "epoch": 0.06570841889117043,
      "grad_norm": 0.4148617386817932,
      "learning_rate": 8e-05,
      "loss": 1.7897,
      "step": 608
    },
    {
      "epoch": 0.06581649194855722,
      "grad_norm": 0.3995443880558014,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 609
    },
    {
      "epoch": 0.06592456500594401,
      "grad_norm": 0.3937922716140747,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 610
    },
    {
      "epoch": 0.06603263806333082,
      "grad_norm": 0.36282509565353394,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 611
    },
    {
      "epoch": 0.06614071112071761,
      "grad_norm": 0.384789377450943,
      "learning_rate": 8e-05,
      "loss": 1.764,
      "step": 612
    },
    {
      "epoch": 0.0662487841781044,
      "grad_norm": 0.39510613679885864,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 613
    },
    {
      "epoch": 0.06635685723549119,
      "grad_norm": 0.3740288019180298,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 614
    },
    {
      "epoch": 0.06646493029287799,
      "grad_norm": 0.38287287950515747,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 615
    },
    {
      "epoch": 0.06657300335026478,
      "grad_norm": 0.4186026453971863,
      "learning_rate": 8e-05,
      "loss": 1.9726,
      "step": 616
    },
    {
      "epoch": 0.06668107640765157,
      "grad_norm": 0.38106051087379456,
      "learning_rate": 8e-05,
      "loss": 1.8186,
      "step": 617
    },
    {
      "epoch": 0.06678914946503836,
      "grad_norm": 0.3739719092845917,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 618
    },
    {
      "epoch": 0.06689722252242516,
      "grad_norm": 0.38507819175720215,
      "learning_rate": 8e-05,
      "loss": 1.909,
      "step": 619
    },
    {
      "epoch": 0.06700529557981195,
      "grad_norm": 0.4046095609664917,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 620
    },
    {
      "epoch": 0.06711336863719874,
      "grad_norm": 0.3768567442893982,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 621
    },
    {
      "epoch": 0.06722144169458553,
      "grad_norm": 0.39734476804733276,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 622
    },
    {
      "epoch": 0.06732951475197234,
      "grad_norm": 0.40313541889190674,
      "learning_rate": 8e-05,
      "loss": 1.965,
      "step": 623
    },
    {
      "epoch": 0.06743758780935913,
      "grad_norm": 0.4095149636268616,
      "learning_rate": 8e-05,
      "loss": 1.831,
      "step": 624
    },
    {
      "epoch": 0.06754566086674592,
      "grad_norm": 0.38132134079933167,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 625
    },
    {
      "epoch": 0.06765373392413271,
      "grad_norm": 0.39820095896720886,
      "learning_rate": 8e-05,
      "loss": 1.9829,
      "step": 626
    },
    {
      "epoch": 0.06776180698151951,
      "grad_norm": 0.4287107586860657,
      "learning_rate": 8e-05,
      "loss": 1.9297,
      "step": 627
    },
    {
      "epoch": 0.0678698800389063,
      "grad_norm": 0.3656159043312073,
      "learning_rate": 8e-05,
      "loss": 1.7225,
      "step": 628
    },
    {
      "epoch": 0.06797795309629309,
      "grad_norm": 0.4076426029205322,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 629
    },
    {
      "epoch": 0.06808602615367988,
      "grad_norm": 0.3812955915927887,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 630
    },
    {
      "epoch": 0.06819409921106669,
      "grad_norm": 0.36205512285232544,
      "learning_rate": 8e-05,
      "loss": 1.7549,
      "step": 631
    },
    {
      "epoch": 0.06830217226845348,
      "grad_norm": 0.4323030710220337,
      "learning_rate": 8e-05,
      "loss": 1.9556,
      "step": 632
    },
    {
      "epoch": 0.06841024532584027,
      "grad_norm": 0.3783770799636841,
      "learning_rate": 8e-05,
      "loss": 1.8469,
      "step": 633
    },
    {
      "epoch": 0.06851831838322706,
      "grad_norm": 0.4081025719642639,
      "learning_rate": 8e-05,
      "loss": 1.9734,
      "step": 634
    },
    {
      "epoch": 0.06862639144061386,
      "grad_norm": 0.4132733941078186,
      "learning_rate": 8e-05,
      "loss": 1.9484,
      "step": 635
    },
    {
      "epoch": 0.06873446449800065,
      "grad_norm": 0.37275877594947815,
      "learning_rate": 8e-05,
      "loss": 1.8193,
      "step": 636
    },
    {
      "epoch": 0.06884253755538744,
      "grad_norm": 0.43893173336982727,
      "learning_rate": 8e-05,
      "loss": 2.0941,
      "step": 637
    },
    {
      "epoch": 0.06895061061277423,
      "grad_norm": 0.3599644601345062,
      "learning_rate": 8e-05,
      "loss": 1.7418,
      "step": 638
    },
    {
      "epoch": 0.06905868367016103,
      "grad_norm": 0.3741653561592102,
      "learning_rate": 8e-05,
      "loss": 1.9187,
      "step": 639
    },
    {
      "epoch": 0.06916675672754782,
      "grad_norm": 0.40093180537223816,
      "learning_rate": 8e-05,
      "loss": 1.8992,
      "step": 640
    },
    {
      "epoch": 0.06927482978493461,
      "grad_norm": 0.38753965497016907,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 641
    },
    {
      "epoch": 0.0693829028423214,
      "grad_norm": 0.362502783536911,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 642
    },
    {
      "epoch": 0.06949097589970821,
      "grad_norm": 0.367779016494751,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 643
    },
    {
      "epoch": 0.069599048957095,
      "grad_norm": 0.37093988060951233,
      "learning_rate": 8e-05,
      "loss": 1.88,
      "step": 644
    },
    {
      "epoch": 0.06970712201448179,
      "grad_norm": 0.37644118070602417,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 645
    },
    {
      "epoch": 0.06981519507186858,
      "grad_norm": 0.4614153206348419,
      "learning_rate": 8e-05,
      "loss": 2.0094,
      "step": 646
    },
    {
      "epoch": 0.06992326812925538,
      "grad_norm": 0.3555675745010376,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 647
    },
    {
      "epoch": 0.07003134118664217,
      "grad_norm": 0.3816656470298767,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 648
    },
    {
      "epoch": 0.07013941424402896,
      "grad_norm": 0.36184394359588623,
      "learning_rate": 8e-05,
      "loss": 1.8526,
      "step": 649
    },
    {
      "epoch": 0.07024748730141575,
      "grad_norm": 0.3917531371116638,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 650
    },
    {
      "epoch": 0.07035556035880255,
      "grad_norm": 0.42352578043937683,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 651
    },
    {
      "epoch": 0.07046363341618934,
      "grad_norm": 0.39063504338264465,
      "learning_rate": 8e-05,
      "loss": 1.9292,
      "step": 652
    },
    {
      "epoch": 0.07057170647357613,
      "grad_norm": 0.37439724802970886,
      "learning_rate": 8e-05,
      "loss": 1.7928,
      "step": 653
    },
    {
      "epoch": 0.07067977953096292,
      "grad_norm": 0.40234991908073425,
      "learning_rate": 8e-05,
      "loss": 1.884,
      "step": 654
    },
    {
      "epoch": 0.07078785258834973,
      "grad_norm": 0.35308724641799927,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 655
    },
    {
      "epoch": 0.07089592564573652,
      "grad_norm": 0.3786294162273407,
      "learning_rate": 8e-05,
      "loss": 1.8821,
      "step": 656
    },
    {
      "epoch": 0.07100399870312331,
      "grad_norm": 0.3727152645587921,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 657
    },
    {
      "epoch": 0.0711120717605101,
      "grad_norm": 0.41945216059684753,
      "learning_rate": 8e-05,
      "loss": 1.8857,
      "step": 658
    },
    {
      "epoch": 0.0712201448178969,
      "grad_norm": 0.3821808695793152,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 659
    },
    {
      "epoch": 0.07132821787528369,
      "grad_norm": 0.3622840940952301,
      "learning_rate": 8e-05,
      "loss": 1.8347,
      "step": 660
    },
    {
      "epoch": 0.07143629093267048,
      "grad_norm": 0.38230574131011963,
      "learning_rate": 8e-05,
      "loss": 1.8955,
      "step": 661
    },
    {
      "epoch": 0.07154436399005727,
      "grad_norm": 0.3756974935531616,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 662
    },
    {
      "epoch": 0.07165243704744408,
      "grad_norm": 0.40111440420150757,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 663
    },
    {
      "epoch": 0.07176051010483087,
      "grad_norm": 0.3803350627422333,
      "learning_rate": 8e-05,
      "loss": 1.8397,
      "step": 664
    },
    {
      "epoch": 0.07186858316221766,
      "grad_norm": 0.3880308270454407,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 665
    },
    {
      "epoch": 0.07197665621960445,
      "grad_norm": 0.3626590967178345,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 666
    },
    {
      "epoch": 0.07208472927699125,
      "grad_norm": 0.372896283864975,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 667
    },
    {
      "epoch": 0.07219280233437804,
      "grad_norm": 0.39188075065612793,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 668
    },
    {
      "epoch": 0.07230087539176483,
      "grad_norm": 0.39207613468170166,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 669
    },
    {
      "epoch": 0.07240894844915163,
      "grad_norm": 0.36483195424079895,
      "learning_rate": 8e-05,
      "loss": 1.7094,
      "step": 670
    },
    {
      "epoch": 0.07251702150653842,
      "grad_norm": 0.39603325724601746,
      "learning_rate": 8e-05,
      "loss": 1.86,
      "step": 671
    },
    {
      "epoch": 0.07262509456392521,
      "grad_norm": 0.39191511273384094,
      "learning_rate": 8e-05,
      "loss": 1.8873,
      "step": 672
    },
    {
      "epoch": 0.072733167621312,
      "grad_norm": 0.3677518963813782,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 673
    },
    {
      "epoch": 0.07284124067869881,
      "grad_norm": 0.3947051167488098,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 674
    },
    {
      "epoch": 0.0729493137360856,
      "grad_norm": 0.40861445665359497,
      "learning_rate": 8e-05,
      "loss": 1.8345,
      "step": 675
    },
    {
      "epoch": 0.07305738679347239,
      "grad_norm": 0.3512199819087982,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 676
    },
    {
      "epoch": 0.07316545985085918,
      "grad_norm": 0.41326385736465454,
      "learning_rate": 8e-05,
      "loss": 1.9878,
      "step": 677
    },
    {
      "epoch": 0.07327353290824598,
      "grad_norm": 0.3899990916252136,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 678
    },
    {
      "epoch": 0.07338160596563277,
      "grad_norm": 0.3928191065788269,
      "learning_rate": 8e-05,
      "loss": 1.8901,
      "step": 679
    },
    {
      "epoch": 0.07348967902301956,
      "grad_norm": 0.46000251173973083,
      "learning_rate": 8e-05,
      "loss": 2.1777,
      "step": 680
    },
    {
      "epoch": 0.07359775208040635,
      "grad_norm": 0.3591993451118469,
      "learning_rate": 8e-05,
      "loss": 1.6818,
      "step": 681
    },
    {
      "epoch": 0.07370582513779315,
      "grad_norm": 0.4386027753353119,
      "learning_rate": 8e-05,
      "loss": 1.9455,
      "step": 682
    },
    {
      "epoch": 0.07381389819517994,
      "grad_norm": 0.36685577034950256,
      "learning_rate": 8e-05,
      "loss": 1.795,
      "step": 683
    },
    {
      "epoch": 0.07392197125256673,
      "grad_norm": 0.3886784613132477,
      "learning_rate": 8e-05,
      "loss": 1.6913,
      "step": 684
    },
    {
      "epoch": 0.07403004430995352,
      "grad_norm": 0.41767385601997375,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 685
    },
    {
      "epoch": 0.07413811736734033,
      "grad_norm": 0.36953943967819214,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 686
    },
    {
      "epoch": 0.07424619042472712,
      "grad_norm": 0.37622350454330444,
      "learning_rate": 8e-05,
      "loss": 1.5588,
      "step": 687
    },
    {
      "epoch": 0.07435426348211391,
      "grad_norm": 0.38845396041870117,
      "learning_rate": 8e-05,
      "loss": 1.7859,
      "step": 688
    },
    {
      "epoch": 0.0744623365395007,
      "grad_norm": 0.37625575065612793,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 689
    },
    {
      "epoch": 0.0745704095968875,
      "grad_norm": 0.40193358063697815,
      "learning_rate": 8e-05,
      "loss": 1.8252,
      "step": 690
    },
    {
      "epoch": 0.07467848265427429,
      "grad_norm": 0.36179593205451965,
      "learning_rate": 8e-05,
      "loss": 1.7285,
      "step": 691
    },
    {
      "epoch": 0.07478655571166108,
      "grad_norm": 0.4251675307750702,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 692
    },
    {
      "epoch": 0.07489462876904787,
      "grad_norm": 0.4131140410900116,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 693
    },
    {
      "epoch": 0.07500270182643468,
      "grad_norm": 0.4186745882034302,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 694
    },
    {
      "epoch": 0.07511077488382147,
      "grad_norm": 0.38539329171180725,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 695
    },
    {
      "epoch": 0.07521884794120826,
      "grad_norm": 0.38900861144065857,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 696
    },
    {
      "epoch": 0.07532692099859505,
      "grad_norm": 0.43458014726638794,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 697
    },
    {
      "epoch": 0.07543499405598185,
      "grad_norm": 0.44995778799057007,
      "learning_rate": 8e-05,
      "loss": 2.0044,
      "step": 698
    },
    {
      "epoch": 0.07554306711336864,
      "grad_norm": 0.38817331194877625,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 699
    },
    {
      "epoch": 0.07565114017075543,
      "grad_norm": 0.3792326748371124,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 700
    },
    {
      "epoch": 0.07575921322814222,
      "grad_norm": 0.45768672227859497,
      "learning_rate": 8e-05,
      "loss": 1.9822,
      "step": 701
    },
    {
      "epoch": 0.07586728628552902,
      "grad_norm": 0.38582563400268555,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 702
    },
    {
      "epoch": 0.07597535934291581,
      "grad_norm": 0.46318045258522034,
      "learning_rate": 8e-05,
      "loss": 1.9612,
      "step": 703
    },
    {
      "epoch": 0.0760834324003026,
      "grad_norm": 0.41352084279060364,
      "learning_rate": 8e-05,
      "loss": 1.8463,
      "step": 704
    },
    {
      "epoch": 0.07619150545768939,
      "grad_norm": 0.3876722753047943,
      "learning_rate": 8e-05,
      "loss": 1.8781,
      "step": 705
    },
    {
      "epoch": 0.0762995785150762,
      "grad_norm": 0.3922761082649231,
      "learning_rate": 8e-05,
      "loss": 1.8374,
      "step": 706
    },
    {
      "epoch": 0.07640765157246299,
      "grad_norm": 0.376577228307724,
      "learning_rate": 8e-05,
      "loss": 1.8531,
      "step": 707
    },
    {
      "epoch": 0.07651572462984978,
      "grad_norm": 0.4063990116119385,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 708
    },
    {
      "epoch": 0.07662379768723657,
      "grad_norm": 0.35337427258491516,
      "learning_rate": 8e-05,
      "loss": 1.7098,
      "step": 709
    },
    {
      "epoch": 0.07673187074462337,
      "grad_norm": 0.3628653883934021,
      "learning_rate": 8e-05,
      "loss": 1.6349,
      "step": 710
    },
    {
      "epoch": 0.07683994380201016,
      "grad_norm": 0.36892956495285034,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 711
    },
    {
      "epoch": 0.07694801685939695,
      "grad_norm": 0.4008506238460541,
      "learning_rate": 8e-05,
      "loss": 1.8488,
      "step": 712
    },
    {
      "epoch": 0.07705608991678374,
      "grad_norm": 0.37482285499572754,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 713
    },
    {
      "epoch": 0.07716416297417054,
      "grad_norm": 0.37777504324913025,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 714
    },
    {
      "epoch": 0.07727223603155733,
      "grad_norm": 0.40664300322532654,
      "learning_rate": 8e-05,
      "loss": 1.9474,
      "step": 715
    },
    {
      "epoch": 0.07738030908894412,
      "grad_norm": 0.4404623210430145,
      "learning_rate": 8e-05,
      "loss": 1.8737,
      "step": 716
    },
    {
      "epoch": 0.07748838214633091,
      "grad_norm": 0.36503866314888,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 717
    },
    {
      "epoch": 0.07759645520371772,
      "grad_norm": 0.3754958510398865,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 718
    },
    {
      "epoch": 0.07770452826110451,
      "grad_norm": 0.3980618715286255,
      "learning_rate": 8e-05,
      "loss": 1.8587,
      "step": 719
    },
    {
      "epoch": 0.0778126013184913,
      "grad_norm": 0.3966750204563141,
      "learning_rate": 8e-05,
      "loss": 1.849,
      "step": 720
    },
    {
      "epoch": 0.07792067437587809,
      "grad_norm": 0.395704060792923,
      "learning_rate": 8e-05,
      "loss": 1.8363,
      "step": 721
    },
    {
      "epoch": 0.07802874743326489,
      "grad_norm": 0.39950525760650635,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 722
    },
    {
      "epoch": 0.07813682049065168,
      "grad_norm": 0.38623759150505066,
      "learning_rate": 8e-05,
      "loss": 1.643,
      "step": 723
    },
    {
      "epoch": 0.07824489354803847,
      "grad_norm": 0.4051452875137329,
      "learning_rate": 8e-05,
      "loss": 1.8128,
      "step": 724
    },
    {
      "epoch": 0.07835296660542526,
      "grad_norm": 0.3887885510921478,
      "learning_rate": 8e-05,
      "loss": 1.8749,
      "step": 725
    },
    {
      "epoch": 0.07846103966281207,
      "grad_norm": 0.4060894846916199,
      "learning_rate": 8e-05,
      "loss": 1.9156,
      "step": 726
    },
    {
      "epoch": 0.07856911272019886,
      "grad_norm": 0.4750705361366272,
      "learning_rate": 8e-05,
      "loss": 1.9148,
      "step": 727
    },
    {
      "epoch": 0.07867718577758565,
      "grad_norm": 0.4595370590686798,
      "learning_rate": 8e-05,
      "loss": 1.7999,
      "step": 728
    },
    {
      "epoch": 0.07878525883497244,
      "grad_norm": 0.41915518045425415,
      "learning_rate": 8e-05,
      "loss": 1.823,
      "step": 729
    },
    {
      "epoch": 0.07889333189235924,
      "grad_norm": 0.37480318546295166,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 730
    },
    {
      "epoch": 0.07900140494974603,
      "grad_norm": 0.41208967566490173,
      "learning_rate": 8e-05,
      "loss": 1.8462,
      "step": 731
    },
    {
      "epoch": 0.07910947800713282,
      "grad_norm": 0.423060804605484,
      "learning_rate": 8e-05,
      "loss": 1.8749,
      "step": 732
    },
    {
      "epoch": 0.07921755106451961,
      "grad_norm": 0.42225298285484314,
      "learning_rate": 8e-05,
      "loss": 1.9046,
      "step": 733
    },
    {
      "epoch": 0.07932562412190641,
      "grad_norm": 0.39896702766418457,
      "learning_rate": 8e-05,
      "loss": 1.922,
      "step": 734
    },
    {
      "epoch": 0.0794336971792932,
      "grad_norm": 0.41620367765426636,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 735
    },
    {
      "epoch": 0.07954177023667999,
      "grad_norm": 0.4114861786365509,
      "learning_rate": 8e-05,
      "loss": 1.9433,
      "step": 736
    },
    {
      "epoch": 0.07964984329406678,
      "grad_norm": 0.4142431318759918,
      "learning_rate": 8e-05,
      "loss": 1.8558,
      "step": 737
    },
    {
      "epoch": 0.07975791635145359,
      "grad_norm": 0.43747401237487793,
      "learning_rate": 8e-05,
      "loss": 1.9381,
      "step": 738
    },
    {
      "epoch": 0.07986598940884038,
      "grad_norm": 0.376428484916687,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 739
    },
    {
      "epoch": 0.07997406246622717,
      "grad_norm": 0.3887920379638672,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 740
    },
    {
      "epoch": 0.08008213552361396,
      "grad_norm": 0.37597936391830444,
      "learning_rate": 8e-05,
      "loss": 1.7208,
      "step": 741
    },
    {
      "epoch": 0.08019020858100076,
      "grad_norm": 0.40306341648101807,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 742
    },
    {
      "epoch": 0.08029828163838755,
      "grad_norm": 0.39108285307884216,
      "learning_rate": 8e-05,
      "loss": 1.8865,
      "step": 743
    },
    {
      "epoch": 0.08040635469577434,
      "grad_norm": 0.3984650671482086,
      "learning_rate": 8e-05,
      "loss": 1.9034,
      "step": 744
    },
    {
      "epoch": 0.08051442775316113,
      "grad_norm": 0.38007602095603943,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 745
    },
    {
      "epoch": 0.08062250081054793,
      "grad_norm": 0.40538743138313293,
      "learning_rate": 8e-05,
      "loss": 1.8086,
      "step": 746
    },
    {
      "epoch": 0.08073057386793472,
      "grad_norm": 0.39091217517852783,
      "learning_rate": 8e-05,
      "loss": 1.8807,
      "step": 747
    },
    {
      "epoch": 0.08083864692532151,
      "grad_norm": 0.412941038608551,
      "learning_rate": 8e-05,
      "loss": 1.9038,
      "step": 748
    },
    {
      "epoch": 0.0809467199827083,
      "grad_norm": 0.43014633655548096,
      "learning_rate": 8e-05,
      "loss": 2.0818,
      "step": 749
    },
    {
      "epoch": 0.08105479304009511,
      "grad_norm": 0.43024763464927673,
      "learning_rate": 8e-05,
      "loss": 1.8385,
      "step": 750
    },
    {
      "epoch": 0.0811628660974819,
      "grad_norm": 0.4032343327999115,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 751
    },
    {
      "epoch": 0.08127093915486869,
      "grad_norm": 0.3984857499599457,
      "learning_rate": 8e-05,
      "loss": 1.7652,
      "step": 752
    },
    {
      "epoch": 0.08137901221225549,
      "grad_norm": 0.40351131558418274,
      "learning_rate": 8e-05,
      "loss": 1.875,
      "step": 753
    },
    {
      "epoch": 0.08148708526964228,
      "grad_norm": 0.4276335537433624,
      "learning_rate": 8e-05,
      "loss": 1.9624,
      "step": 754
    },
    {
      "epoch": 0.08159515832702907,
      "grad_norm": 0.4173552691936493,
      "learning_rate": 8e-05,
      "loss": 1.9534,
      "step": 755
    },
    {
      "epoch": 0.08170323138441586,
      "grad_norm": 0.4000588655471802,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 756
    },
    {
      "epoch": 0.08181130444180267,
      "grad_norm": 0.3658292889595032,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 757
    },
    {
      "epoch": 0.08191937749918946,
      "grad_norm": 0.3618623614311218,
      "learning_rate": 8e-05,
      "loss": 1.6193,
      "step": 758
    },
    {
      "epoch": 0.08202745055657625,
      "grad_norm": 0.3873171806335449,
      "learning_rate": 8e-05,
      "loss": 1.8274,
      "step": 759
    },
    {
      "epoch": 0.08213552361396304,
      "grad_norm": 0.39139628410339355,
      "learning_rate": 8e-05,
      "loss": 1.9344,
      "step": 760
    },
    {
      "epoch": 0.08224359667134984,
      "grad_norm": 0.369528591632843,
      "learning_rate": 8e-05,
      "loss": 1.9108,
      "step": 761
    },
    {
      "epoch": 0.08235166972873663,
      "grad_norm": 0.38032254576683044,
      "learning_rate": 8e-05,
      "loss": 1.7506,
      "step": 762
    },
    {
      "epoch": 0.08245974278612342,
      "grad_norm": 0.3675854802131653,
      "learning_rate": 8e-05,
      "loss": 1.5945,
      "step": 763
    },
    {
      "epoch": 0.08256781584351021,
      "grad_norm": 0.41050955653190613,
      "learning_rate": 8e-05,
      "loss": 2.0557,
      "step": 764
    },
    {
      "epoch": 0.08267588890089701,
      "grad_norm": 0.38829997181892395,
      "learning_rate": 8e-05,
      "loss": 1.83,
      "step": 765
    },
    {
      "epoch": 0.0827839619582838,
      "grad_norm": 0.39948633313179016,
      "learning_rate": 8e-05,
      "loss": 1.9633,
      "step": 766
    },
    {
      "epoch": 0.08289203501567059,
      "grad_norm": 0.37214839458465576,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 767
    },
    {
      "epoch": 0.08300010807305738,
      "grad_norm": 0.35905367136001587,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 768
    },
    {
      "epoch": 0.08310818113044419,
      "grad_norm": 0.4149291217327118,
      "learning_rate": 8e-05,
      "loss": 1.7792,
      "step": 769
    },
    {
      "epoch": 0.08321625418783098,
      "grad_norm": 0.4004831314086914,
      "learning_rate": 8e-05,
      "loss": 1.8439,
      "step": 770
    },
    {
      "epoch": 0.08332432724521777,
      "grad_norm": 0.3987247943878174,
      "learning_rate": 8e-05,
      "loss": 1.8321,
      "step": 771
    },
    {
      "epoch": 0.08343240030260456,
      "grad_norm": 0.37861669063568115,
      "learning_rate": 8e-05,
      "loss": 1.7538,
      "step": 772
    },
    {
      "epoch": 0.08354047335999136,
      "grad_norm": 0.3814734220504761,
      "learning_rate": 8e-05,
      "loss": 1.7378,
      "step": 773
    },
    {
      "epoch": 0.08364854641737815,
      "grad_norm": 0.3898855149745941,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 774
    },
    {
      "epoch": 0.08375661947476494,
      "grad_norm": 0.3908434510231018,
      "learning_rate": 8e-05,
      "loss": 1.6661,
      "step": 775
    },
    {
      "epoch": 0.08386469253215173,
      "grad_norm": 0.4093868136405945,
      "learning_rate": 8e-05,
      "loss": 1.9138,
      "step": 776
    },
    {
      "epoch": 0.08397276558953853,
      "grad_norm": 0.3769180178642273,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 777
    },
    {
      "epoch": 0.08408083864692532,
      "grad_norm": 0.433907687664032,
      "learning_rate": 8e-05,
      "loss": 1.8802,
      "step": 778
    },
    {
      "epoch": 0.08418891170431211,
      "grad_norm": 0.4165599048137665,
      "learning_rate": 8e-05,
      "loss": 1.9292,
      "step": 779
    },
    {
      "epoch": 0.0842969847616989,
      "grad_norm": 0.37741199135780334,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 780
    },
    {
      "epoch": 0.08440505781908571,
      "grad_norm": 0.36487072706222534,
      "learning_rate": 8e-05,
      "loss": 1.5165,
      "step": 781
    },
    {
      "epoch": 0.0845131308764725,
      "grad_norm": 0.36852407455444336,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 782
    },
    {
      "epoch": 0.08462120393385929,
      "grad_norm": 0.37867751717567444,
      "learning_rate": 8e-05,
      "loss": 1.6509,
      "step": 783
    },
    {
      "epoch": 0.08472927699124608,
      "grad_norm": 0.3854752779006958,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 784
    },
    {
      "epoch": 0.08483735004863288,
      "grad_norm": 0.3982992172241211,
      "learning_rate": 8e-05,
      "loss": 1.7303,
      "step": 785
    },
    {
      "epoch": 0.08494542310601967,
      "grad_norm": 0.404774010181427,
      "learning_rate": 8e-05,
      "loss": 2.0125,
      "step": 786
    },
    {
      "epoch": 0.08505349616340646,
      "grad_norm": 0.3976024091243744,
      "learning_rate": 8e-05,
      "loss": 1.8651,
      "step": 787
    },
    {
      "epoch": 0.08516156922079325,
      "grad_norm": 0.3817993998527527,
      "learning_rate": 8e-05,
      "loss": 1.8008,
      "step": 788
    },
    {
      "epoch": 0.08526964227818006,
      "grad_norm": 0.38240939378738403,
      "learning_rate": 8e-05,
      "loss": 1.8826,
      "step": 789
    },
    {
      "epoch": 0.08537771533556685,
      "grad_norm": 0.3701499104499817,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 790
    },
    {
      "epoch": 0.08548578839295363,
      "grad_norm": 0.35764646530151367,
      "learning_rate": 8e-05,
      "loss": 1.7125,
      "step": 791
    },
    {
      "epoch": 0.08559386145034042,
      "grad_norm": 0.398835152387619,
      "learning_rate": 8e-05,
      "loss": 1.8763,
      "step": 792
    },
    {
      "epoch": 0.08570193450772723,
      "grad_norm": 0.3656446635723114,
      "learning_rate": 8e-05,
      "loss": 1.7451,
      "step": 793
    },
    {
      "epoch": 0.08581000756511402,
      "grad_norm": 0.38311994075775146,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 794
    },
    {
      "epoch": 0.08591808062250081,
      "grad_norm": 0.3806648850440979,
      "learning_rate": 8e-05,
      "loss": 1.7417,
      "step": 795
    },
    {
      "epoch": 0.0860261536798876,
      "grad_norm": 0.3972926735877991,
      "learning_rate": 8e-05,
      "loss": 1.9324,
      "step": 796
    },
    {
      "epoch": 0.0861342267372744,
      "grad_norm": 0.3899213671684265,
      "learning_rate": 8e-05,
      "loss": 1.9501,
      "step": 797
    },
    {
      "epoch": 0.08624229979466119,
      "grad_norm": 0.39491838216781616,
      "learning_rate": 8e-05,
      "loss": 1.6727,
      "step": 798
    },
    {
      "epoch": 0.08635037285204798,
      "grad_norm": 0.42522281408309937,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 799
    },
    {
      "epoch": 0.08645844590943477,
      "grad_norm": 0.36641034483909607,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 800
    },
    {
      "epoch": 0.08656651896682158,
      "grad_norm": 0.43661749362945557,
      "learning_rate": 8e-05,
      "loss": 1.9789,
      "step": 801
    },
    {
      "epoch": 0.08667459202420837,
      "grad_norm": 0.4124746322631836,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 802
    },
    {
      "epoch": 0.08678266508159516,
      "grad_norm": 0.38758140802383423,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 803
    },
    {
      "epoch": 0.08689073813898195,
      "grad_norm": 0.41409361362457275,
      "learning_rate": 8e-05,
      "loss": 1.9696,
      "step": 804
    },
    {
      "epoch": 0.08699881119636875,
      "grad_norm": 0.3868120312690735,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 805
    },
    {
      "epoch": 0.08710688425375554,
      "grad_norm": 0.41382551193237305,
      "learning_rate": 8e-05,
      "loss": 2.0019,
      "step": 806
    },
    {
      "epoch": 0.08721495731114233,
      "grad_norm": 0.395884245634079,
      "learning_rate": 8e-05,
      "loss": 1.6038,
      "step": 807
    },
    {
      "epoch": 0.08732303036852912,
      "grad_norm": 0.3913949429988861,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 808
    },
    {
      "epoch": 0.08743110342591592,
      "grad_norm": 0.37184062600135803,
      "learning_rate": 8e-05,
      "loss": 1.8912,
      "step": 809
    },
    {
      "epoch": 0.08753917648330271,
      "grad_norm": 0.4517929255962372,
      "learning_rate": 8e-05,
      "loss": 1.9396,
      "step": 810
    },
    {
      "epoch": 0.0876472495406895,
      "grad_norm": 0.39372360706329346,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 811
    },
    {
      "epoch": 0.0877553225980763,
      "grad_norm": 0.35442546010017395,
      "learning_rate": 8e-05,
      "loss": 1.6597,
      "step": 812
    },
    {
      "epoch": 0.0878633956554631,
      "grad_norm": 0.3787603974342346,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 813
    },
    {
      "epoch": 0.08797146871284989,
      "grad_norm": 0.3756096065044403,
      "learning_rate": 8e-05,
      "loss": 1.9025,
      "step": 814
    },
    {
      "epoch": 0.08807954177023668,
      "grad_norm": 0.3945809602737427,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 815
    },
    {
      "epoch": 0.08818761482762347,
      "grad_norm": 0.3575534224510193,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 816
    },
    {
      "epoch": 0.08829568788501027,
      "grad_norm": 0.3838723599910736,
      "learning_rate": 8e-05,
      "loss": 1.7987,
      "step": 817
    },
    {
      "epoch": 0.08840376094239706,
      "grad_norm": 0.3883250653743744,
      "learning_rate": 8e-05,
      "loss": 1.8999,
      "step": 818
    },
    {
      "epoch": 0.08851183399978385,
      "grad_norm": 0.38830623030662537,
      "learning_rate": 8e-05,
      "loss": 1.8702,
      "step": 819
    },
    {
      "epoch": 0.08861990705717064,
      "grad_norm": 0.3867105543613434,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 820
    },
    {
      "epoch": 0.08872798011455744,
      "grad_norm": 0.4300100803375244,
      "learning_rate": 8e-05,
      "loss": 1.9542,
      "step": 821
    },
    {
      "epoch": 0.08883605317194423,
      "grad_norm": 0.3763100802898407,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 822
    },
    {
      "epoch": 0.08894412622933102,
      "grad_norm": 0.3884502649307251,
      "learning_rate": 8e-05,
      "loss": 1.9922,
      "step": 823
    },
    {
      "epoch": 0.08905219928671781,
      "grad_norm": 0.37160724401474,
      "learning_rate": 8e-05,
      "loss": 1.5586,
      "step": 824
    },
    {
      "epoch": 0.08916027234410462,
      "grad_norm": 0.3885810077190399,
      "learning_rate": 8e-05,
      "loss": 1.8479,
      "step": 825
    },
    {
      "epoch": 0.08926834540149141,
      "grad_norm": 0.4155563414096832,
      "learning_rate": 8e-05,
      "loss": 1.9966,
      "step": 826
    },
    {
      "epoch": 0.0893764184588782,
      "grad_norm": 0.38042008876800537,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 827
    },
    {
      "epoch": 0.08948449151626499,
      "grad_norm": 0.37227728962898254,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 828
    },
    {
      "epoch": 0.08959256457365179,
      "grad_norm": 0.38358941674232483,
      "learning_rate": 8e-05,
      "loss": 1.7891,
      "step": 829
    },
    {
      "epoch": 0.08970063763103858,
      "grad_norm": 0.38548195362091064,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 830
    },
    {
      "epoch": 0.08980871068842537,
      "grad_norm": 0.4204581081867218,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 831
    },
    {
      "epoch": 0.08991678374581216,
      "grad_norm": 0.3924368917942047,
      "learning_rate": 8e-05,
      "loss": 1.7948,
      "step": 832
    },
    {
      "epoch": 0.09002485680319897,
      "grad_norm": 0.3928157091140747,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 833
    },
    {
      "epoch": 0.09013292986058576,
      "grad_norm": 0.35608869791030884,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 834
    },
    {
      "epoch": 0.09024100291797255,
      "grad_norm": 0.37497037649154663,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 835
    },
    {
      "epoch": 0.09034907597535935,
      "grad_norm": 0.389941930770874,
      "learning_rate": 8e-05,
      "loss": 1.7929,
      "step": 836
    },
    {
      "epoch": 0.09045714903274614,
      "grad_norm": 0.4363395869731903,
      "learning_rate": 8e-05,
      "loss": 1.991,
      "step": 837
    },
    {
      "epoch": 0.09056522209013293,
      "grad_norm": 0.36937981843948364,
      "learning_rate": 8e-05,
      "loss": 1.4477,
      "step": 838
    },
    {
      "epoch": 0.09067329514751972,
      "grad_norm": 0.3966974914073944,
      "learning_rate": 8e-05,
      "loss": 1.7744,
      "step": 839
    },
    {
      "epoch": 0.09078136820490652,
      "grad_norm": 0.3682551383972168,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 840
    },
    {
      "epoch": 0.09088944126229331,
      "grad_norm": 0.44152846932411194,
      "learning_rate": 8e-05,
      "loss": 2.0222,
      "step": 841
    },
    {
      "epoch": 0.0909975143196801,
      "grad_norm": 0.393815815448761,
      "learning_rate": 8e-05,
      "loss": 1.9188,
      "step": 842
    },
    {
      "epoch": 0.0911055873770669,
      "grad_norm": 0.39585718512535095,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 843
    },
    {
      "epoch": 0.0912136604344537,
      "grad_norm": 0.3918764591217041,
      "learning_rate": 8e-05,
      "loss": 1.881,
      "step": 844
    },
    {
      "epoch": 0.09132173349184049,
      "grad_norm": 0.3769766390323639,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 845
    },
    {
      "epoch": 0.09142980654922728,
      "grad_norm": 0.38504549860954285,
      "learning_rate": 8e-05,
      "loss": 1.7987,
      "step": 846
    },
    {
      "epoch": 0.09153787960661407,
      "grad_norm": 0.41448888182640076,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 847
    },
    {
      "epoch": 0.09164595266400087,
      "grad_norm": 0.39184752106666565,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 848
    },
    {
      "epoch": 0.09175402572138766,
      "grad_norm": 0.3676890432834625,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 849
    },
    {
      "epoch": 0.09186209877877445,
      "grad_norm": 0.39263418316841125,
      "learning_rate": 8e-05,
      "loss": 1.874,
      "step": 850
    },
    {
      "epoch": 0.09197017183616124,
      "grad_norm": 0.37928488850593567,
      "learning_rate": 8e-05,
      "loss": 1.7637,
      "step": 851
    },
    {
      "epoch": 0.09207824489354804,
      "grad_norm": 0.4056071639060974,
      "learning_rate": 8e-05,
      "loss": 1.9423,
      "step": 852
    },
    {
      "epoch": 0.09218631795093483,
      "grad_norm": 0.36004483699798584,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 853
    },
    {
      "epoch": 0.09229439100832162,
      "grad_norm": 0.3662221133708954,
      "learning_rate": 8e-05,
      "loss": 1.6923,
      "step": 854
    },
    {
      "epoch": 0.09240246406570841,
      "grad_norm": 0.3881687521934509,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 855
    },
    {
      "epoch": 0.09251053712309522,
      "grad_norm": 0.3656984269618988,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 856
    },
    {
      "epoch": 0.09261861018048201,
      "grad_norm": 0.35692664980888367,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 857
    },
    {
      "epoch": 0.0927266832378688,
      "grad_norm": 0.3823276460170746,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 858
    },
    {
      "epoch": 0.09283475629525559,
      "grad_norm": 0.386101096868515,
      "learning_rate": 8e-05,
      "loss": 1.8226,
      "step": 859
    },
    {
      "epoch": 0.09294282935264239,
      "grad_norm": 0.3880169689655304,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 860
    },
    {
      "epoch": 0.09305090241002918,
      "grad_norm": 0.4168223440647125,
      "learning_rate": 8e-05,
      "loss": 1.8139,
      "step": 861
    },
    {
      "epoch": 0.09315897546741597,
      "grad_norm": 0.3633286952972412,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 862
    },
    {
      "epoch": 0.09326704852480276,
      "grad_norm": 0.41162189841270447,
      "learning_rate": 8e-05,
      "loss": 1.7998,
      "step": 863
    },
    {
      "epoch": 0.09337512158218957,
      "grad_norm": 0.3691965341567993,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 864
    },
    {
      "epoch": 0.09348319463957636,
      "grad_norm": 0.3983602225780487,
      "learning_rate": 8e-05,
      "loss": 1.8441,
      "step": 865
    },
    {
      "epoch": 0.09359126769696315,
      "grad_norm": 0.3840714991092682,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 866
    },
    {
      "epoch": 0.09369934075434994,
      "grad_norm": 0.42957112193107605,
      "learning_rate": 8e-05,
      "loss": 1.8453,
      "step": 867
    },
    {
      "epoch": 0.09380741381173674,
      "grad_norm": 0.4000215232372284,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 868
    },
    {
      "epoch": 0.09391548686912353,
      "grad_norm": 0.38217616081237793,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 869
    },
    {
      "epoch": 0.09402355992651032,
      "grad_norm": 0.4226932227611542,
      "learning_rate": 8e-05,
      "loss": 1.8034,
      "step": 870
    },
    {
      "epoch": 0.09413163298389711,
      "grad_norm": 0.3689088225364685,
      "learning_rate": 8e-05,
      "loss": 1.6566,
      "step": 871
    },
    {
      "epoch": 0.09423970604128391,
      "grad_norm": 0.4080781638622284,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 872
    },
    {
      "epoch": 0.0943477790986707,
      "grad_norm": 0.42132335901260376,
      "learning_rate": 8e-05,
      "loss": 1.771,
      "step": 873
    },
    {
      "epoch": 0.0944558521560575,
      "grad_norm": 0.3799995183944702,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 874
    },
    {
      "epoch": 0.09456392521344428,
      "grad_norm": 0.3889107406139374,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 875
    },
    {
      "epoch": 0.09467199827083109,
      "grad_norm": 0.38771361112594604,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 876
    },
    {
      "epoch": 0.09478007132821788,
      "grad_norm": 0.4059380292892456,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 877
    },
    {
      "epoch": 0.09488814438560467,
      "grad_norm": 0.4534737169742584,
      "learning_rate": 8e-05,
      "loss": 1.789,
      "step": 878
    },
    {
      "epoch": 0.09499621744299146,
      "grad_norm": 0.40548110008239746,
      "learning_rate": 8e-05,
      "loss": 1.8294,
      "step": 879
    },
    {
      "epoch": 0.09510429050037826,
      "grad_norm": 0.3889337182044983,
      "learning_rate": 8e-05,
      "loss": 1.9375,
      "step": 880
    },
    {
      "epoch": 0.09521236355776505,
      "grad_norm": 0.3814393877983093,
      "learning_rate": 8e-05,
      "loss": 1.8976,
      "step": 881
    },
    {
      "epoch": 0.09532043661515184,
      "grad_norm": 0.38665857911109924,
      "learning_rate": 8e-05,
      "loss": 1.9455,
      "step": 882
    },
    {
      "epoch": 0.09542850967253863,
      "grad_norm": 0.3986888527870178,
      "learning_rate": 8e-05,
      "loss": 1.9337,
      "step": 883
    },
    {
      "epoch": 0.09553658272992543,
      "grad_norm": 0.4116247296333313,
      "learning_rate": 8e-05,
      "loss": 1.9296,
      "step": 884
    },
    {
      "epoch": 0.09564465578731222,
      "grad_norm": 0.43145492672920227,
      "learning_rate": 8e-05,
      "loss": 2.0282,
      "step": 885
    },
    {
      "epoch": 0.09575272884469901,
      "grad_norm": 0.42363113164901733,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 886
    },
    {
      "epoch": 0.0958608019020858,
      "grad_norm": 0.4092237055301666,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 887
    },
    {
      "epoch": 0.09596887495947261,
      "grad_norm": 0.3851903975009918,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 888
    },
    {
      "epoch": 0.0960769480168594,
      "grad_norm": 0.38634058833122253,
      "learning_rate": 8e-05,
      "loss": 1.7738,
      "step": 889
    },
    {
      "epoch": 0.09618502107424619,
      "grad_norm": 0.39109164476394653,
      "learning_rate": 8e-05,
      "loss": 1.9104,
      "step": 890
    },
    {
      "epoch": 0.09629309413163298,
      "grad_norm": 0.4568932056427002,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 891
    },
    {
      "epoch": 0.09640116718901978,
      "grad_norm": 0.4418637454509735,
      "learning_rate": 8e-05,
      "loss": 1.9375,
      "step": 892
    },
    {
      "epoch": 0.09650924024640657,
      "grad_norm": 0.41826122999191284,
      "learning_rate": 8e-05,
      "loss": 1.7832,
      "step": 893
    },
    {
      "epoch": 0.09661731330379336,
      "grad_norm": 0.38933613896369934,
      "learning_rate": 8e-05,
      "loss": 1.8712,
      "step": 894
    },
    {
      "epoch": 0.09672538636118015,
      "grad_norm": 0.3884674906730652,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 895
    },
    {
      "epoch": 0.09683345941856696,
      "grad_norm": 0.4698336124420166,
      "learning_rate": 8e-05,
      "loss": 2.0276,
      "step": 896
    },
    {
      "epoch": 0.09694153247595375,
      "grad_norm": 0.38476020097732544,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 897
    },
    {
      "epoch": 0.09704960553334054,
      "grad_norm": 0.4123043417930603,
      "learning_rate": 8e-05,
      "loss": 1.9763,
      "step": 898
    },
    {
      "epoch": 0.09715767859072733,
      "grad_norm": 0.3964158296585083,
      "learning_rate": 8e-05,
      "loss": 1.7222,
      "step": 899
    },
    {
      "epoch": 0.09726575164811413,
      "grad_norm": 0.42433246970176697,
      "learning_rate": 8e-05,
      "loss": 1.8012,
      "step": 900
    },
    {
      "epoch": 0.09737382470550092,
      "grad_norm": 0.47174257040023804,
      "learning_rate": 8e-05,
      "loss": 1.8627,
      "step": 901
    },
    {
      "epoch": 0.09748189776288771,
      "grad_norm": 0.3662796914577484,
      "learning_rate": 8e-05,
      "loss": 1.6606,
      "step": 902
    },
    {
      "epoch": 0.0975899708202745,
      "grad_norm": 0.39581525325775146,
      "learning_rate": 8e-05,
      "loss": 1.9064,
      "step": 903
    },
    {
      "epoch": 0.0976980438776613,
      "grad_norm": 0.3845064342021942,
      "learning_rate": 8e-05,
      "loss": 1.7938,
      "step": 904
    },
    {
      "epoch": 0.09780611693504809,
      "grad_norm": 0.3746867775917053,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 905
    },
    {
      "epoch": 0.09791418999243488,
      "grad_norm": 0.3892257511615753,
      "learning_rate": 8e-05,
      "loss": 1.7624,
      "step": 906
    },
    {
      "epoch": 0.09802226304982167,
      "grad_norm": 0.4154015779495239,
      "learning_rate": 8e-05,
      "loss": 1.7878,
      "step": 907
    },
    {
      "epoch": 0.09813033610720848,
      "grad_norm": 0.410176545381546,
      "learning_rate": 8e-05,
      "loss": 1.9207,
      "step": 908
    },
    {
      "epoch": 0.09823840916459527,
      "grad_norm": 0.3932543396949768,
      "learning_rate": 8e-05,
      "loss": 1.8713,
      "step": 909
    },
    {
      "epoch": 0.09834648222198206,
      "grad_norm": 0.37113407254219055,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 910
    },
    {
      "epoch": 0.09845455527936885,
      "grad_norm": 0.3920171856880188,
      "learning_rate": 8e-05,
      "loss": 1.7281,
      "step": 911
    },
    {
      "epoch": 0.09856262833675565,
      "grad_norm": 0.3956649899482727,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 912
    },
    {
      "epoch": 0.09867070139414244,
      "grad_norm": 0.38215869665145874,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 913
    },
    {
      "epoch": 0.09877877445152923,
      "grad_norm": 0.40719112753868103,
      "learning_rate": 8e-05,
      "loss": 1.8719,
      "step": 914
    },
    {
      "epoch": 0.09888684750891602,
      "grad_norm": 0.4015348255634308,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 915
    },
    {
      "epoch": 0.09899492056630282,
      "grad_norm": 0.38722726702690125,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 916
    },
    {
      "epoch": 0.09910299362368961,
      "grad_norm": 0.4500625729560852,
      "learning_rate": 8e-05,
      "loss": 2.0419,
      "step": 917
    },
    {
      "epoch": 0.0992110666810764,
      "grad_norm": 0.4296802580356598,
      "learning_rate": 8e-05,
      "loss": 1.8555,
      "step": 918
    },
    {
      "epoch": 0.09931913973846321,
      "grad_norm": 0.39559996128082275,
      "learning_rate": 8e-05,
      "loss": 1.8795,
      "step": 919
    },
    {
      "epoch": 0.09942721279585,
      "grad_norm": 0.41484421491622925,
      "learning_rate": 8e-05,
      "loss": 1.9728,
      "step": 920
    },
    {
      "epoch": 0.09953528585323679,
      "grad_norm": 0.39266663789749146,
      "learning_rate": 8e-05,
      "loss": 1.8983,
      "step": 921
    },
    {
      "epoch": 0.09964335891062358,
      "grad_norm": 0.36321285367012024,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 922
    },
    {
      "epoch": 0.09975143196801038,
      "grad_norm": 0.39684033393859863,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 923
    },
    {
      "epoch": 0.09985950502539717,
      "grad_norm": 0.4456006586551666,
      "learning_rate": 8e-05,
      "loss": 1.8512,
      "step": 924
    },
    {
      "epoch": 0.09996757808278396,
      "grad_norm": 0.3925921618938446,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 925
    },
    {
      "epoch": 0.10007565114017075,
      "grad_norm": 0.37765389680862427,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 926
    },
    {
      "epoch": 0.10018372419755756,
      "grad_norm": 0.39557576179504395,
      "learning_rate": 8e-05,
      "loss": 1.7489,
      "step": 927
    },
    {
      "epoch": 0.10029179725494435,
      "grad_norm": 0.37633559107780457,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 928
    },
    {
      "epoch": 0.10039987031233114,
      "grad_norm": 0.3852573037147522,
      "learning_rate": 8e-05,
      "loss": 1.7551,
      "step": 929
    },
    {
      "epoch": 0.10050794336971793,
      "grad_norm": 0.3863203227519989,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 930
    },
    {
      "epoch": 0.10061601642710473,
      "grad_norm": 0.3638008236885071,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 931
    },
    {
      "epoch": 0.10072408948449152,
      "grad_norm": 0.3683527112007141,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 932
    },
    {
      "epoch": 0.10083216254187831,
      "grad_norm": 0.3892306089401245,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 933
    },
    {
      "epoch": 0.1009402355992651,
      "grad_norm": 0.502811849117279,
      "learning_rate": 8e-05,
      "loss": 1.8612,
      "step": 934
    },
    {
      "epoch": 0.1010483086566519,
      "grad_norm": 0.3959799110889435,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 935
    },
    {
      "epoch": 0.10115638171403869,
      "grad_norm": 0.4079091548919678,
      "learning_rate": 8e-05,
      "loss": 1.8726,
      "step": 936
    },
    {
      "epoch": 0.10126445477142548,
      "grad_norm": 0.3714248538017273,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 937
    },
    {
      "epoch": 0.10137252782881227,
      "grad_norm": 0.4109494090080261,
      "learning_rate": 8e-05,
      "loss": 1.8327,
      "step": 938
    },
    {
      "epoch": 0.10148060088619908,
      "grad_norm": 0.3767405152320862,
      "learning_rate": 8e-05,
      "loss": 1.7504,
      "step": 939
    },
    {
      "epoch": 0.10158867394358587,
      "grad_norm": 0.3870100677013397,
      "learning_rate": 8e-05,
      "loss": 1.8685,
      "step": 940
    },
    {
      "epoch": 0.10169674700097266,
      "grad_norm": 0.3935485780239105,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 941
    },
    {
      "epoch": 0.10180482005835945,
      "grad_norm": 0.4005415141582489,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 942
    },
    {
      "epoch": 0.10191289311574625,
      "grad_norm": 0.41724643111228943,
      "learning_rate": 8e-05,
      "loss": 2.0737,
      "step": 943
    },
    {
      "epoch": 0.10202096617313304,
      "grad_norm": 0.3955412209033966,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 944
    },
    {
      "epoch": 0.10212903923051983,
      "grad_norm": 0.38798463344573975,
      "learning_rate": 8e-05,
      "loss": 1.8015,
      "step": 945
    },
    {
      "epoch": 0.10223711228790662,
      "grad_norm": 0.3683830499649048,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 946
    },
    {
      "epoch": 0.10234518534529342,
      "grad_norm": 0.3578316867351532,
      "learning_rate": 8e-05,
      "loss": 1.606,
      "step": 947
    },
    {
      "epoch": 0.10245325840268021,
      "grad_norm": 0.38520461320877075,
      "learning_rate": 8e-05,
      "loss": 1.653,
      "step": 948
    },
    {
      "epoch": 0.102561331460067,
      "grad_norm": 0.3985229730606079,
      "learning_rate": 8e-05,
      "loss": 1.8523,
      "step": 949
    },
    {
      "epoch": 0.1026694045174538,
      "grad_norm": 0.39425259828567505,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 950
    },
    {
      "epoch": 0.1027774775748406,
      "grad_norm": 0.3792390525341034,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 951
    },
    {
      "epoch": 0.10288555063222739,
      "grad_norm": 0.3755030930042267,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 952
    },
    {
      "epoch": 0.10299362368961418,
      "grad_norm": 0.3919469714164734,
      "learning_rate": 8e-05,
      "loss": 1.8146,
      "step": 953
    },
    {
      "epoch": 0.10310169674700097,
      "grad_norm": 0.3702966570854187,
      "learning_rate": 8e-05,
      "loss": 1.742,
      "step": 954
    },
    {
      "epoch": 0.10320976980438777,
      "grad_norm": 0.4130384922027588,
      "learning_rate": 8e-05,
      "loss": 1.8827,
      "step": 955
    },
    {
      "epoch": 0.10331784286177456,
      "grad_norm": 0.42664065957069397,
      "learning_rate": 8e-05,
      "loss": 1.8536,
      "step": 956
    },
    {
      "epoch": 0.10342591591916135,
      "grad_norm": 0.36221182346343994,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 957
    },
    {
      "epoch": 0.10353398897654814,
      "grad_norm": 0.43861445784568787,
      "learning_rate": 8e-05,
      "loss": 1.9205,
      "step": 958
    },
    {
      "epoch": 0.10364206203393495,
      "grad_norm": 0.38644519448280334,
      "learning_rate": 8e-05,
      "loss": 1.5372,
      "step": 959
    },
    {
      "epoch": 0.10375013509132174,
      "grad_norm": 0.3858511447906494,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 960
    },
    {
      "epoch": 0.10385820814870853,
      "grad_norm": 0.3766767978668213,
      "learning_rate": 8e-05,
      "loss": 1.6319,
      "step": 961
    },
    {
      "epoch": 0.10396628120609532,
      "grad_norm": 0.383675217628479,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 962
    },
    {
      "epoch": 0.10407435426348212,
      "grad_norm": 0.354854017496109,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 963
    },
    {
      "epoch": 0.10418242732086891,
      "grad_norm": 0.41565173864364624,
      "learning_rate": 8e-05,
      "loss": 1.9405,
      "step": 964
    },
    {
      "epoch": 0.1042905003782557,
      "grad_norm": 0.3621530532836914,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 965
    },
    {
      "epoch": 0.10439857343564249,
      "grad_norm": 0.39552217721939087,
      "learning_rate": 8e-05,
      "loss": 1.9871,
      "step": 966
    },
    {
      "epoch": 0.10450664649302929,
      "grad_norm": 0.41175055503845215,
      "learning_rate": 8e-05,
      "loss": 1.8022,
      "step": 967
    },
    {
      "epoch": 0.10461471955041608,
      "grad_norm": 0.3924751877784729,
      "learning_rate": 8e-05,
      "loss": 1.7594,
      "step": 968
    },
    {
      "epoch": 0.10472279260780287,
      "grad_norm": 0.4085668623447418,
      "learning_rate": 8e-05,
      "loss": 1.8636,
      "step": 969
    },
    {
      "epoch": 0.10483086566518966,
      "grad_norm": 0.38495364785194397,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 970
    },
    {
      "epoch": 0.10493893872257647,
      "grad_norm": 0.37489065527915955,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 971
    },
    {
      "epoch": 0.10504701177996326,
      "grad_norm": 0.38128164410591125,
      "learning_rate": 8e-05,
      "loss": 1.8449,
      "step": 972
    },
    {
      "epoch": 0.10515508483735005,
      "grad_norm": 0.38440579175949097,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 973
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 0.4034498631954193,
      "learning_rate": 8e-05,
      "loss": 1.8798,
      "step": 974
    },
    {
      "epoch": 0.10537123095212364,
      "grad_norm": 0.393085241317749,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 975
    },
    {
      "epoch": 0.10547930400951043,
      "grad_norm": 0.39501169323921204,
      "learning_rate": 8e-05,
      "loss": 1.878,
      "step": 976
    },
    {
      "epoch": 0.10558737706689722,
      "grad_norm": 0.40644970536231995,
      "learning_rate": 8e-05,
      "loss": 1.8885,
      "step": 977
    },
    {
      "epoch": 0.10569545012428401,
      "grad_norm": 0.3920840919017792,
      "learning_rate": 8e-05,
      "loss": 1.7743,
      "step": 978
    },
    {
      "epoch": 0.10580352318167081,
      "grad_norm": 0.41720208525657654,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 979
    },
    {
      "epoch": 0.1059115962390576,
      "grad_norm": 0.3972610533237457,
      "learning_rate": 8e-05,
      "loss": 1.7891,
      "step": 980
    },
    {
      "epoch": 0.1060196692964444,
      "grad_norm": 0.3888673484325409,
      "learning_rate": 8e-05,
      "loss": 1.8439,
      "step": 981
    },
    {
      "epoch": 0.10612774235383118,
      "grad_norm": 0.3454180061817169,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 982
    },
    {
      "epoch": 0.10623581541121799,
      "grad_norm": 0.37489598989486694,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 983
    },
    {
      "epoch": 0.10634388846860478,
      "grad_norm": 0.4719237983226776,
      "learning_rate": 8e-05,
      "loss": 1.9419,
      "step": 984
    },
    {
      "epoch": 0.10645196152599157,
      "grad_norm": 0.3999921381473541,
      "learning_rate": 8e-05,
      "loss": 1.8531,
      "step": 985
    },
    {
      "epoch": 0.10656003458337836,
      "grad_norm": 0.38797518610954285,
      "learning_rate": 8e-05,
      "loss": 1.7473,
      "step": 986
    },
    {
      "epoch": 0.10666810764076516,
      "grad_norm": 0.3866576850414276,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 987
    },
    {
      "epoch": 0.10677618069815195,
      "grad_norm": 0.4167454242706299,
      "learning_rate": 8e-05,
      "loss": 1.7504,
      "step": 988
    },
    {
      "epoch": 0.10688425375553874,
      "grad_norm": 0.37719035148620605,
      "learning_rate": 8e-05,
      "loss": 1.6591,
      "step": 989
    },
    {
      "epoch": 0.10699232681292553,
      "grad_norm": 0.4179949462413788,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 990
    },
    {
      "epoch": 0.10710039987031233,
      "grad_norm": 0.40129154920578003,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 991
    },
    {
      "epoch": 0.10720847292769912,
      "grad_norm": 0.4385087490081787,
      "learning_rate": 8e-05,
      "loss": 2.018,
      "step": 992
    },
    {
      "epoch": 0.10731654598508591,
      "grad_norm": 0.4139760732650757,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 993
    },
    {
      "epoch": 0.1074246190424727,
      "grad_norm": 0.38537269830703735,
      "learning_rate": 8e-05,
      "loss": 1.7084,
      "step": 994
    },
    {
      "epoch": 0.10753269209985951,
      "grad_norm": 0.3762817978858948,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 995
    },
    {
      "epoch": 0.1076407651572463,
      "grad_norm": 0.3645487427711487,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 996
    },
    {
      "epoch": 0.10774883821463309,
      "grad_norm": 0.3987293243408203,
      "learning_rate": 8e-05,
      "loss": 1.9136,
      "step": 997
    },
    {
      "epoch": 0.10785691127201988,
      "grad_norm": 0.3873227834701538,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 998
    },
    {
      "epoch": 0.10796498432940668,
      "grad_norm": 0.37445971369743347,
      "learning_rate": 8e-05,
      "loss": 1.6392,
      "step": 999
    },
    {
      "epoch": 0.10807305738679347,
      "grad_norm": 0.3835374712944031,
      "learning_rate": 8e-05,
      "loss": 1.8741,
      "step": 1000
    },
    {
      "epoch": 0.10818113044418026,
      "grad_norm": 0.4019988477230072,
      "learning_rate": 8e-05,
      "loss": 1.9206,
      "step": 1001
    },
    {
      "epoch": 0.10828920350156707,
      "grad_norm": 0.38331493735313416,
      "learning_rate": 8e-05,
      "loss": 1.9473,
      "step": 1002
    },
    {
      "epoch": 0.10839727655895386,
      "grad_norm": 0.4246036410331726,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 1003
    },
    {
      "epoch": 0.10850534961634065,
      "grad_norm": 0.4011020362377167,
      "learning_rate": 8e-05,
      "loss": 1.884,
      "step": 1004
    },
    {
      "epoch": 0.10861342267372744,
      "grad_norm": 0.40727895498275757,
      "learning_rate": 8e-05,
      "loss": 1.8266,
      "step": 1005
    },
    {
      "epoch": 0.10872149573111424,
      "grad_norm": 0.41984468698501587,
      "learning_rate": 8e-05,
      "loss": 1.8404,
      "step": 1006
    },
    {
      "epoch": 0.10882956878850103,
      "grad_norm": 0.400930792093277,
      "learning_rate": 8e-05,
      "loss": 1.8573,
      "step": 1007
    },
    {
      "epoch": 0.10893764184588782,
      "grad_norm": 0.354104220867157,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 1008
    },
    {
      "epoch": 0.10904571490327461,
      "grad_norm": 0.39942649006843567,
      "learning_rate": 8e-05,
      "loss": 1.7588,
      "step": 1009
    },
    {
      "epoch": 0.10915378796066141,
      "grad_norm": 0.42665526270866394,
      "learning_rate": 8e-05,
      "loss": 1.9818,
      "step": 1010
    },
    {
      "epoch": 0.1092618610180482,
      "grad_norm": 0.4933208227157593,
      "learning_rate": 8e-05,
      "loss": 2.1123,
      "step": 1011
    },
    {
      "epoch": 0.109369934075435,
      "grad_norm": 0.3701803982257843,
      "learning_rate": 8e-05,
      "loss": 1.7619,
      "step": 1012
    },
    {
      "epoch": 0.10947800713282178,
      "grad_norm": 0.3931812047958374,
      "learning_rate": 8e-05,
      "loss": 1.9035,
      "step": 1013
    },
    {
      "epoch": 0.10958608019020859,
      "grad_norm": 0.376873254776001,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 1014
    },
    {
      "epoch": 0.10969415324759538,
      "grad_norm": 0.35913553833961487,
      "learning_rate": 8e-05,
      "loss": 1.7635,
      "step": 1015
    },
    {
      "epoch": 0.10980222630498217,
      "grad_norm": 0.3640599548816681,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 1016
    },
    {
      "epoch": 0.10991029936236896,
      "grad_norm": 0.3719395101070404,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 1017
    },
    {
      "epoch": 0.11001837241975576,
      "grad_norm": 0.37454622983932495,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 1018
    },
    {
      "epoch": 0.11012644547714255,
      "grad_norm": 0.35914841294288635,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1019
    },
    {
      "epoch": 0.11023451853452934,
      "grad_norm": 0.4011459946632385,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 1020
    },
    {
      "epoch": 0.11034259159191613,
      "grad_norm": 0.39066827297210693,
      "learning_rate": 8e-05,
      "loss": 1.7464,
      "step": 1021
    },
    {
      "epoch": 0.11045066464930293,
      "grad_norm": 0.37952175736427307,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 1022
    },
    {
      "epoch": 0.11055873770668972,
      "grad_norm": 0.3708139955997467,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 1023
    },
    {
      "epoch": 0.11066681076407651,
      "grad_norm": 0.39974668622016907,
      "learning_rate": 8e-05,
      "loss": 1.7923,
      "step": 1024
    },
    {
      "epoch": 0.1107748838214633,
      "grad_norm": 0.41978588700294495,
      "learning_rate": 8e-05,
      "loss": 1.8768,
      "step": 1025
    },
    {
      "epoch": 0.11088295687885011,
      "grad_norm": 0.4098535180091858,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1026
    },
    {
      "epoch": 0.1109910299362369,
      "grad_norm": 0.39209285378456116,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1027
    },
    {
      "epoch": 0.11109910299362369,
      "grad_norm": 0.3783665895462036,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 1028
    },
    {
      "epoch": 0.11120717605101048,
      "grad_norm": 0.42844903469085693,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 1029
    },
    {
      "epoch": 0.11131524910839728,
      "grad_norm": 0.3889824151992798,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 1030
    },
    {
      "epoch": 0.11142332216578407,
      "grad_norm": 0.4054432809352875,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 1031
    },
    {
      "epoch": 0.11153139522317086,
      "grad_norm": 0.4063291549682617,
      "learning_rate": 8e-05,
      "loss": 1.8885,
      "step": 1032
    },
    {
      "epoch": 0.11163946828055765,
      "grad_norm": 0.398786723613739,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 1033
    },
    {
      "epoch": 0.11174754133794446,
      "grad_norm": 0.3663095235824585,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 1034
    },
    {
      "epoch": 0.11185561439533125,
      "grad_norm": 0.37868762016296387,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 1035
    },
    {
      "epoch": 0.11196368745271804,
      "grad_norm": 0.3953869044780731,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 1036
    },
    {
      "epoch": 0.11207176051010483,
      "grad_norm": 0.3921975791454315,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 1037
    },
    {
      "epoch": 0.11217983356749163,
      "grad_norm": 0.43811070919036865,
      "learning_rate": 8e-05,
      "loss": 1.863,
      "step": 1038
    },
    {
      "epoch": 0.11228790662487842,
      "grad_norm": 0.3938697874546051,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 1039
    },
    {
      "epoch": 0.11239597968226521,
      "grad_norm": 0.37724462151527405,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 1040
    },
    {
      "epoch": 0.112504052739652,
      "grad_norm": 0.44374632835388184,
      "learning_rate": 8e-05,
      "loss": 2.021,
      "step": 1041
    },
    {
      "epoch": 0.1126121257970388,
      "grad_norm": 0.41529354453086853,
      "learning_rate": 8e-05,
      "loss": 1.8786,
      "step": 1042
    },
    {
      "epoch": 0.1127201988544256,
      "grad_norm": 0.38080450892448425,
      "learning_rate": 8e-05,
      "loss": 1.8324,
      "step": 1043
    },
    {
      "epoch": 0.11282827191181238,
      "grad_norm": 0.38980257511138916,
      "learning_rate": 8e-05,
      "loss": 1.8743,
      "step": 1044
    },
    {
      "epoch": 0.11293634496919917,
      "grad_norm": 0.4001608192920685,
      "learning_rate": 8e-05,
      "loss": 1.7822,
      "step": 1045
    },
    {
      "epoch": 0.11304441802658598,
      "grad_norm": 0.38472118973731995,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 1046
    },
    {
      "epoch": 0.11315249108397277,
      "grad_norm": 0.3805444836616516,
      "learning_rate": 8e-05,
      "loss": 1.7319,
      "step": 1047
    },
    {
      "epoch": 0.11326056414135956,
      "grad_norm": 0.4139706790447235,
      "learning_rate": 8e-05,
      "loss": 1.8513,
      "step": 1048
    },
    {
      "epoch": 0.11336863719874635,
      "grad_norm": 0.3881073594093323,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 1049
    },
    {
      "epoch": 0.11347671025613315,
      "grad_norm": 0.3923235535621643,
      "learning_rate": 8e-05,
      "loss": 1.57,
      "step": 1050
    },
    {
      "epoch": 0.11358478331351994,
      "grad_norm": 0.3660266399383545,
      "learning_rate": 8e-05,
      "loss": 1.6073,
      "step": 1051
    },
    {
      "epoch": 0.11369285637090673,
      "grad_norm": 0.3726440966129303,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 1052
    },
    {
      "epoch": 0.11380092942829352,
      "grad_norm": 0.3979420065879822,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 1053
    },
    {
      "epoch": 0.11390900248568032,
      "grad_norm": 0.37223440408706665,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 1054
    },
    {
      "epoch": 0.11401707554306711,
      "grad_norm": 0.37689220905303955,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 1055
    },
    {
      "epoch": 0.1141251486004539,
      "grad_norm": 0.39760494232177734,
      "learning_rate": 8e-05,
      "loss": 1.8539,
      "step": 1056
    },
    {
      "epoch": 0.1142332216578407,
      "grad_norm": 0.4190789461135864,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 1057
    },
    {
      "epoch": 0.1143412947152275,
      "grad_norm": 0.4178256690502167,
      "learning_rate": 8e-05,
      "loss": 1.8906,
      "step": 1058
    },
    {
      "epoch": 0.11444936777261429,
      "grad_norm": 0.3803950250148773,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 1059
    },
    {
      "epoch": 0.11455744083000108,
      "grad_norm": 0.37108492851257324,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 1060
    },
    {
      "epoch": 0.11466551388738787,
      "grad_norm": 0.3765726387500763,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 1061
    },
    {
      "epoch": 0.11477358694477467,
      "grad_norm": 0.42391684651374817,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 1062
    },
    {
      "epoch": 0.11488166000216146,
      "grad_norm": 0.3822234869003296,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 1063
    },
    {
      "epoch": 0.11498973305954825,
      "grad_norm": 0.4084148108959198,
      "learning_rate": 8e-05,
      "loss": 1.803,
      "step": 1064
    },
    {
      "epoch": 0.11509780611693504,
      "grad_norm": 0.388151615858078,
      "learning_rate": 8e-05,
      "loss": 1.8203,
      "step": 1065
    },
    {
      "epoch": 0.11520587917432185,
      "grad_norm": 0.435252845287323,
      "learning_rate": 8e-05,
      "loss": 1.9584,
      "step": 1066
    },
    {
      "epoch": 0.11531395223170864,
      "grad_norm": 0.3781786859035492,
      "learning_rate": 8e-05,
      "loss": 1.4814,
      "step": 1067
    },
    {
      "epoch": 0.11542202528909543,
      "grad_norm": 0.4196843206882477,
      "learning_rate": 8e-05,
      "loss": 1.9193,
      "step": 1068
    },
    {
      "epoch": 0.11553009834648222,
      "grad_norm": 0.40066418051719666,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 1069
    },
    {
      "epoch": 0.11563817140386902,
      "grad_norm": 0.38373470306396484,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 1070
    },
    {
      "epoch": 0.11574624446125581,
      "grad_norm": 0.4519558846950531,
      "learning_rate": 8e-05,
      "loss": 1.9599,
      "step": 1071
    },
    {
      "epoch": 0.1158543175186426,
      "grad_norm": 0.44331082701683044,
      "learning_rate": 8e-05,
      "loss": 1.9116,
      "step": 1072
    },
    {
      "epoch": 0.11596239057602939,
      "grad_norm": 0.3911808729171753,
      "learning_rate": 8e-05,
      "loss": 1.7904,
      "step": 1073
    },
    {
      "epoch": 0.1160704636334162,
      "grad_norm": 0.35664913058280945,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 1074
    },
    {
      "epoch": 0.11617853669080298,
      "grad_norm": 0.4297114312648773,
      "learning_rate": 8e-05,
      "loss": 1.8517,
      "step": 1075
    },
    {
      "epoch": 0.11628660974818977,
      "grad_norm": 0.45054349303245544,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1076
    },
    {
      "epoch": 0.11639468280557656,
      "grad_norm": 0.399105429649353,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 1077
    },
    {
      "epoch": 0.11650275586296337,
      "grad_norm": 0.4434087574481964,
      "learning_rate": 8e-05,
      "loss": 1.7475,
      "step": 1078
    },
    {
      "epoch": 0.11661082892035016,
      "grad_norm": 0.40028512477874756,
      "learning_rate": 8e-05,
      "loss": 1.9234,
      "step": 1079
    },
    {
      "epoch": 0.11671890197773695,
      "grad_norm": 0.39779356122016907,
      "learning_rate": 8e-05,
      "loss": 1.8101,
      "step": 1080
    },
    {
      "epoch": 0.11682697503512374,
      "grad_norm": 0.44353628158569336,
      "learning_rate": 8e-05,
      "loss": 1.9948,
      "step": 1081
    },
    {
      "epoch": 0.11693504809251054,
      "grad_norm": 0.42460140585899353,
      "learning_rate": 8e-05,
      "loss": 1.939,
      "step": 1082
    },
    {
      "epoch": 0.11704312114989733,
      "grad_norm": 0.38076382875442505,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 1083
    },
    {
      "epoch": 0.11715119420728412,
      "grad_norm": 0.38609376549720764,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 1084
    },
    {
      "epoch": 0.11725926726467092,
      "grad_norm": 0.41954711079597473,
      "learning_rate": 8e-05,
      "loss": 1.9258,
      "step": 1085
    },
    {
      "epoch": 0.11736734032205771,
      "grad_norm": 0.3955862820148468,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 1086
    },
    {
      "epoch": 0.1174754133794445,
      "grad_norm": 0.3895527422428131,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 1087
    },
    {
      "epoch": 0.1175834864368313,
      "grad_norm": 0.40267693996429443,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 1088
    },
    {
      "epoch": 0.1176915594942181,
      "grad_norm": 0.3973911702632904,
      "learning_rate": 8e-05,
      "loss": 1.7907,
      "step": 1089
    },
    {
      "epoch": 0.11779963255160489,
      "grad_norm": 0.43264514207839966,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 1090
    },
    {
      "epoch": 0.11790770560899168,
      "grad_norm": 0.3992522358894348,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 1091
    },
    {
      "epoch": 0.11801577866637847,
      "grad_norm": 0.40067043900489807,
      "learning_rate": 8e-05,
      "loss": 1.8082,
      "step": 1092
    },
    {
      "epoch": 0.11812385172376527,
      "grad_norm": 0.4115132689476013,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 1093
    },
    {
      "epoch": 0.11823192478115206,
      "grad_norm": 0.3945947289466858,
      "learning_rate": 8e-05,
      "loss": 1.869,
      "step": 1094
    },
    {
      "epoch": 0.11833999783853885,
      "grad_norm": 0.38537564873695374,
      "learning_rate": 8e-05,
      "loss": 1.8144,
      "step": 1095
    },
    {
      "epoch": 0.11844807089592564,
      "grad_norm": 0.39341843128204346,
      "learning_rate": 8e-05,
      "loss": 1.7204,
      "step": 1096
    },
    {
      "epoch": 0.11855614395331245,
      "grad_norm": 0.4371182322502136,
      "learning_rate": 8e-05,
      "loss": 1.7994,
      "step": 1097
    },
    {
      "epoch": 0.11866421701069924,
      "grad_norm": 0.3756963610649109,
      "learning_rate": 8e-05,
      "loss": 1.8059,
      "step": 1098
    },
    {
      "epoch": 0.11877229006808603,
      "grad_norm": 0.3809126913547516,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 1099
    },
    {
      "epoch": 0.11888036312547282,
      "grad_norm": 0.4054781198501587,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 1100
    },
    {
      "epoch": 0.11898843618285962,
      "grad_norm": 0.3998783826828003,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 1101
    },
    {
      "epoch": 0.11909650924024641,
      "grad_norm": 0.4328933358192444,
      "learning_rate": 8e-05,
      "loss": 1.9099,
      "step": 1102
    },
    {
      "epoch": 0.1192045822976332,
      "grad_norm": 0.42348945140838623,
      "learning_rate": 8e-05,
      "loss": 1.8363,
      "step": 1103
    },
    {
      "epoch": 0.11931265535501999,
      "grad_norm": 0.43993648886680603,
      "learning_rate": 8e-05,
      "loss": 1.9065,
      "step": 1104
    },
    {
      "epoch": 0.11942072841240679,
      "grad_norm": 0.3862455487251282,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 1105
    },
    {
      "epoch": 0.11952880146979358,
      "grad_norm": 0.3712832033634186,
      "learning_rate": 8e-05,
      "loss": 1.8108,
      "step": 1106
    },
    {
      "epoch": 0.11963687452718037,
      "grad_norm": 0.39608529210090637,
      "learning_rate": 8e-05,
      "loss": 1.8439,
      "step": 1107
    },
    {
      "epoch": 0.11974494758456716,
      "grad_norm": 0.4896920621395111,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 1108
    },
    {
      "epoch": 0.11985302064195397,
      "grad_norm": 0.37824660539627075,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 1109
    },
    {
      "epoch": 0.11996109369934076,
      "grad_norm": 0.37427079677581787,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 1110
    },
    {
      "epoch": 0.12006916675672755,
      "grad_norm": 0.42056891322135925,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 1111
    },
    {
      "epoch": 0.12017723981411434,
      "grad_norm": 0.4501860439777374,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 1112
    },
    {
      "epoch": 0.12028531287150114,
      "grad_norm": 0.5504316687583923,
      "learning_rate": 8e-05,
      "loss": 1.8175,
      "step": 1113
    },
    {
      "epoch": 0.12039338592888793,
      "grad_norm": 0.434026300907135,
      "learning_rate": 8e-05,
      "loss": 1.9403,
      "step": 1114
    },
    {
      "epoch": 0.12050145898627472,
      "grad_norm": 0.42353349924087524,
      "learning_rate": 8e-05,
      "loss": 1.7768,
      "step": 1115
    },
    {
      "epoch": 0.12060953204366151,
      "grad_norm": 0.4138243794441223,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 1116
    },
    {
      "epoch": 0.12071760510104831,
      "grad_norm": 0.45309898257255554,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 1117
    },
    {
      "epoch": 0.1208256781584351,
      "grad_norm": 0.4231041371822357,
      "learning_rate": 8e-05,
      "loss": 1.9682,
      "step": 1118
    },
    {
      "epoch": 0.1209337512158219,
      "grad_norm": 0.422085165977478,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 1119
    },
    {
      "epoch": 0.12104182427320868,
      "grad_norm": 0.40817442536354065,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 1120
    },
    {
      "epoch": 0.12114989733059549,
      "grad_norm": 0.392391562461853,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 1121
    },
    {
      "epoch": 0.12125797038798228,
      "grad_norm": 0.38499951362609863,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 1122
    },
    {
      "epoch": 0.12136604344536907,
      "grad_norm": 0.39962854981422424,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 1123
    },
    {
      "epoch": 0.12147411650275586,
      "grad_norm": 0.3966427445411682,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 1124
    },
    {
      "epoch": 0.12158218956014266,
      "grad_norm": 0.4463091492652893,
      "learning_rate": 8e-05,
      "loss": 1.9746,
      "step": 1125
    },
    {
      "epoch": 0.12169026261752945,
      "grad_norm": 0.43111205101013184,
      "learning_rate": 8e-05,
      "loss": 1.8713,
      "step": 1126
    },
    {
      "epoch": 0.12179833567491624,
      "grad_norm": 0.426467627286911,
      "learning_rate": 8e-05,
      "loss": 1.9456,
      "step": 1127
    },
    {
      "epoch": 0.12190640873230303,
      "grad_norm": 0.4001787006855011,
      "learning_rate": 8e-05,
      "loss": 1.7865,
      "step": 1128
    },
    {
      "epoch": 0.12201448178968984,
      "grad_norm": 0.3556313216686249,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 1129
    },
    {
      "epoch": 0.12212255484707663,
      "grad_norm": 0.3689104914665222,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 1130
    },
    {
      "epoch": 0.12223062790446342,
      "grad_norm": 0.396584153175354,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 1131
    },
    {
      "epoch": 0.1223387009618502,
      "grad_norm": 0.43432778120040894,
      "learning_rate": 8e-05,
      "loss": 1.9351,
      "step": 1132
    },
    {
      "epoch": 0.12244677401923701,
      "grad_norm": 0.4293030798435211,
      "learning_rate": 8e-05,
      "loss": 1.8471,
      "step": 1133
    },
    {
      "epoch": 0.1225548470766238,
      "grad_norm": 0.4375183582305908,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 1134
    },
    {
      "epoch": 0.12266292013401059,
      "grad_norm": 0.42041000723838806,
      "learning_rate": 8e-05,
      "loss": 1.913,
      "step": 1135
    },
    {
      "epoch": 0.12277099319139738,
      "grad_norm": 0.374152272939682,
      "learning_rate": 8e-05,
      "loss": 1.869,
      "step": 1136
    },
    {
      "epoch": 0.12287906624878418,
      "grad_norm": 0.3762070834636688,
      "learning_rate": 8e-05,
      "loss": 1.7341,
      "step": 1137
    },
    {
      "epoch": 0.12298713930617097,
      "grad_norm": 0.3948916494846344,
      "learning_rate": 8e-05,
      "loss": 1.6633,
      "step": 1138
    },
    {
      "epoch": 0.12309521236355776,
      "grad_norm": 0.4330386221408844,
      "learning_rate": 8e-05,
      "loss": 1.923,
      "step": 1139
    },
    {
      "epoch": 0.12320328542094455,
      "grad_norm": 0.4007166922092438,
      "learning_rate": 8e-05,
      "loss": 1.8437,
      "step": 1140
    },
    {
      "epoch": 0.12331135847833136,
      "grad_norm": 0.382861465215683,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 1141
    },
    {
      "epoch": 0.12341943153571815,
      "grad_norm": 0.39344456791877747,
      "learning_rate": 8e-05,
      "loss": 1.7205,
      "step": 1142
    },
    {
      "epoch": 0.12352750459310494,
      "grad_norm": 0.4214554727077484,
      "learning_rate": 8e-05,
      "loss": 1.7862,
      "step": 1143
    },
    {
      "epoch": 0.12363557765049173,
      "grad_norm": 0.3869437277317047,
      "learning_rate": 8e-05,
      "loss": 1.8092,
      "step": 1144
    },
    {
      "epoch": 0.12374365070787853,
      "grad_norm": 0.38726913928985596,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 1145
    },
    {
      "epoch": 0.12385172376526532,
      "grad_norm": 0.4158177077770233,
      "learning_rate": 8e-05,
      "loss": 1.8606,
      "step": 1146
    },
    {
      "epoch": 0.12395979682265211,
      "grad_norm": 0.39502185583114624,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 1147
    },
    {
      "epoch": 0.1240678698800389,
      "grad_norm": 0.40828996896743774,
      "learning_rate": 8e-05,
      "loss": 1.9724,
      "step": 1148
    },
    {
      "epoch": 0.1241759429374257,
      "grad_norm": 0.47126397490501404,
      "learning_rate": 8e-05,
      "loss": 2.0076,
      "step": 1149
    },
    {
      "epoch": 0.1242840159948125,
      "grad_norm": 0.404422789812088,
      "learning_rate": 8e-05,
      "loss": 1.8727,
      "step": 1150
    },
    {
      "epoch": 0.12439208905219928,
      "grad_norm": 0.3805296719074249,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 1151
    },
    {
      "epoch": 0.12450016210958607,
      "grad_norm": 0.4066793918609619,
      "learning_rate": 8e-05,
      "loss": 2.0601,
      "step": 1152
    },
    {
      "epoch": 0.12460823516697288,
      "grad_norm": 0.4022175967693329,
      "learning_rate": 8e-05,
      "loss": 1.7348,
      "step": 1153
    },
    {
      "epoch": 0.12471630822435967,
      "grad_norm": 0.37392136454582214,
      "learning_rate": 8e-05,
      "loss": 1.56,
      "step": 1154
    },
    {
      "epoch": 0.12482438128174646,
      "grad_norm": 0.3967081606388092,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 1155
    },
    {
      "epoch": 0.12493245433913325,
      "grad_norm": 0.37934428453445435,
      "learning_rate": 8e-05,
      "loss": 1.8092,
      "step": 1156
    },
    {
      "epoch": 0.12504052739652005,
      "grad_norm": 0.3877156972885132,
      "learning_rate": 8e-05,
      "loss": 1.9138,
      "step": 1157
    },
    {
      "epoch": 0.12514860045390683,
      "grad_norm": 0.39702361822128296,
      "learning_rate": 8e-05,
      "loss": 1.914,
      "step": 1158
    },
    {
      "epoch": 0.12525667351129363,
      "grad_norm": 0.40220436453819275,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 1159
    },
    {
      "epoch": 0.12536474656868044,
      "grad_norm": 0.36265698075294495,
      "learning_rate": 8e-05,
      "loss": 1.6799,
      "step": 1160
    },
    {
      "epoch": 0.1254728196260672,
      "grad_norm": 0.3873903453350067,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 1161
    },
    {
      "epoch": 0.12558089268345402,
      "grad_norm": 0.4082430303096771,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 1162
    },
    {
      "epoch": 0.12568896574084082,
      "grad_norm": 0.3927316963672638,
      "learning_rate": 8e-05,
      "loss": 1.746,
      "step": 1163
    },
    {
      "epoch": 0.1257970387982276,
      "grad_norm": 0.4038935601711273,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 1164
    },
    {
      "epoch": 0.1259051118556144,
      "grad_norm": 0.37971723079681396,
      "learning_rate": 8e-05,
      "loss": 1.5279,
      "step": 1165
    },
    {
      "epoch": 0.12601318491300117,
      "grad_norm": 0.406156450510025,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 1166
    },
    {
      "epoch": 0.12612125797038798,
      "grad_norm": 0.3836826682090759,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 1167
    },
    {
      "epoch": 0.12622933102777478,
      "grad_norm": 0.3874610960483551,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 1168
    },
    {
      "epoch": 0.12633740408516156,
      "grad_norm": 0.4173485040664673,
      "learning_rate": 8e-05,
      "loss": 1.9109,
      "step": 1169
    },
    {
      "epoch": 0.12644547714254836,
      "grad_norm": 0.39089012145996094,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 1170
    },
    {
      "epoch": 0.12655355019993517,
      "grad_norm": 0.42487382888793945,
      "learning_rate": 8e-05,
      "loss": 1.9391,
      "step": 1171
    },
    {
      "epoch": 0.12666162325732194,
      "grad_norm": 0.3710365295410156,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 1172
    },
    {
      "epoch": 0.12676969631470875,
      "grad_norm": 0.39773672819137573,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 1173
    },
    {
      "epoch": 0.12687776937209555,
      "grad_norm": 0.4346335828304291,
      "learning_rate": 8e-05,
      "loss": 1.7477,
      "step": 1174
    },
    {
      "epoch": 0.12698584242948233,
      "grad_norm": 0.405969500541687,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 1175
    },
    {
      "epoch": 0.12709391548686913,
      "grad_norm": 0.3760949373245239,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 1176
    },
    {
      "epoch": 0.1272019885442559,
      "grad_norm": 0.36301136016845703,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 1177
    },
    {
      "epoch": 0.1273100616016427,
      "grad_norm": 0.3719019889831543,
      "learning_rate": 8e-05,
      "loss": 1.5926,
      "step": 1178
    },
    {
      "epoch": 0.1274181346590295,
      "grad_norm": 0.43588805198669434,
      "learning_rate": 8e-05,
      "loss": 1.9668,
      "step": 1179
    },
    {
      "epoch": 0.1275262077164163,
      "grad_norm": 0.3988209068775177,
      "learning_rate": 8e-05,
      "loss": 1.9151,
      "step": 1180
    },
    {
      "epoch": 0.1276342807738031,
      "grad_norm": 0.3863646984100342,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 1181
    },
    {
      "epoch": 0.1277423538311899,
      "grad_norm": 0.37904486060142517,
      "learning_rate": 8e-05,
      "loss": 1.6526,
      "step": 1182
    },
    {
      "epoch": 0.12785042688857667,
      "grad_norm": 0.3899856209754944,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 1183
    },
    {
      "epoch": 0.12795849994596348,
      "grad_norm": 0.35133469104766846,
      "learning_rate": 8e-05,
      "loss": 1.5142,
      "step": 1184
    },
    {
      "epoch": 0.12806657300335025,
      "grad_norm": 0.37702104449272156,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 1185
    },
    {
      "epoch": 0.12817464606073706,
      "grad_norm": 0.3756886422634125,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 1186
    },
    {
      "epoch": 0.12828271911812386,
      "grad_norm": 0.385052889585495,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 1187
    },
    {
      "epoch": 0.12839079217551064,
      "grad_norm": 0.41698354482650757,
      "learning_rate": 8e-05,
      "loss": 1.746,
      "step": 1188
    },
    {
      "epoch": 0.12849886523289744,
      "grad_norm": 0.39301857352256775,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 1189
    },
    {
      "epoch": 0.12860693829028425,
      "grad_norm": 0.45109376311302185,
      "learning_rate": 8e-05,
      "loss": 1.7994,
      "step": 1190
    },
    {
      "epoch": 0.12871501134767102,
      "grad_norm": 0.42747899889945984,
      "learning_rate": 8e-05,
      "loss": 1.9418,
      "step": 1191
    },
    {
      "epoch": 0.12882308440505782,
      "grad_norm": 0.39833468198776245,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 1192
    },
    {
      "epoch": 0.1289311574624446,
      "grad_norm": 0.3804018497467041,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 1193
    },
    {
      "epoch": 0.1290392305198314,
      "grad_norm": 0.3843596279621124,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 1194
    },
    {
      "epoch": 0.1291473035772182,
      "grad_norm": 0.3819177746772766,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 1195
    },
    {
      "epoch": 0.12925537663460498,
      "grad_norm": 0.3951333165168762,
      "learning_rate": 8e-05,
      "loss": 1.8473,
      "step": 1196
    },
    {
      "epoch": 0.1293634496919918,
      "grad_norm": 0.4196566343307495,
      "learning_rate": 8e-05,
      "loss": 1.8762,
      "step": 1197
    },
    {
      "epoch": 0.1294715227493786,
      "grad_norm": 0.3610372543334961,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 1198
    },
    {
      "epoch": 0.12957959580676537,
      "grad_norm": 0.4164235293865204,
      "learning_rate": 8e-05,
      "loss": 1.9844,
      "step": 1199
    },
    {
      "epoch": 0.12968766886415217,
      "grad_norm": 0.4453701972961426,
      "learning_rate": 8e-05,
      "loss": 1.8425,
      "step": 1200
    },
    {
      "epoch": 0.12979574192153895,
      "grad_norm": 0.4126196503639221,
      "learning_rate": 8e-05,
      "loss": 1.9212,
      "step": 1201
    },
    {
      "epoch": 0.12990381497892575,
      "grad_norm": 0.4031369686126709,
      "learning_rate": 8e-05,
      "loss": 1.8251,
      "step": 1202
    },
    {
      "epoch": 0.13001188803631256,
      "grad_norm": 0.4103247821331024,
      "learning_rate": 8e-05,
      "loss": 1.8567,
      "step": 1203
    },
    {
      "epoch": 0.13011996109369933,
      "grad_norm": 0.435825377702713,
      "learning_rate": 8e-05,
      "loss": 1.8846,
      "step": 1204
    },
    {
      "epoch": 0.13022803415108614,
      "grad_norm": 0.39917829632759094,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 1205
    },
    {
      "epoch": 0.13033610720847294,
      "grad_norm": 0.4305744171142578,
      "learning_rate": 8e-05,
      "loss": 1.8639,
      "step": 1206
    },
    {
      "epoch": 0.13044418026585972,
      "grad_norm": 0.3868190050125122,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 1207
    },
    {
      "epoch": 0.13055225332324652,
      "grad_norm": 0.3874880075454712,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 1208
    },
    {
      "epoch": 0.1306603263806333,
      "grad_norm": 0.371711403131485,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 1209
    },
    {
      "epoch": 0.1307683994380201,
      "grad_norm": 0.4109328091144562,
      "learning_rate": 8e-05,
      "loss": 1.7617,
      "step": 1210
    },
    {
      "epoch": 0.1308764724954069,
      "grad_norm": 0.3655467927455902,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 1211
    },
    {
      "epoch": 0.13098454555279368,
      "grad_norm": 0.37783315777778625,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 1212
    },
    {
      "epoch": 0.13109261861018048,
      "grad_norm": 0.3934309184551239,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 1213
    },
    {
      "epoch": 0.1312006916675673,
      "grad_norm": 0.4209797978401184,
      "learning_rate": 8e-05,
      "loss": 1.8484,
      "step": 1214
    },
    {
      "epoch": 0.13130876472495406,
      "grad_norm": 0.43852007389068604,
      "learning_rate": 8e-05,
      "loss": 1.7234,
      "step": 1215
    },
    {
      "epoch": 0.13141683778234087,
      "grad_norm": 0.5381610989570618,
      "learning_rate": 8e-05,
      "loss": 2.104,
      "step": 1216
    },
    {
      "epoch": 0.13152491083972764,
      "grad_norm": 0.41665229201316833,
      "learning_rate": 8e-05,
      "loss": 1.8719,
      "step": 1217
    },
    {
      "epoch": 0.13163298389711445,
      "grad_norm": 0.42526260018348694,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 1218
    },
    {
      "epoch": 0.13174105695450125,
      "grad_norm": 0.39679017663002014,
      "learning_rate": 8e-05,
      "loss": 1.7775,
      "step": 1219
    },
    {
      "epoch": 0.13184913001188803,
      "grad_norm": 0.4550996422767639,
      "learning_rate": 8e-05,
      "loss": 2.046,
      "step": 1220
    },
    {
      "epoch": 0.13195720306927483,
      "grad_norm": 0.39569374918937683,
      "learning_rate": 8e-05,
      "loss": 1.8554,
      "step": 1221
    },
    {
      "epoch": 0.13206527612666163,
      "grad_norm": 0.4711671769618988,
      "learning_rate": 8e-05,
      "loss": 1.9245,
      "step": 1222
    },
    {
      "epoch": 0.1321733491840484,
      "grad_norm": 0.4085521399974823,
      "learning_rate": 8e-05,
      "loss": 1.8913,
      "step": 1223
    },
    {
      "epoch": 0.13228142224143521,
      "grad_norm": 0.41472187638282776,
      "learning_rate": 8e-05,
      "loss": 1.7752,
      "step": 1224
    },
    {
      "epoch": 0.132389495298822,
      "grad_norm": 0.3706962764263153,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 1225
    },
    {
      "epoch": 0.1324975683562088,
      "grad_norm": 0.3827354311943054,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 1226
    },
    {
      "epoch": 0.1326056414135956,
      "grad_norm": 0.4170599579811096,
      "learning_rate": 8e-05,
      "loss": 1.8833,
      "step": 1227
    },
    {
      "epoch": 0.13271371447098237,
      "grad_norm": 0.39253947138786316,
      "learning_rate": 8e-05,
      "loss": 1.8254,
      "step": 1228
    },
    {
      "epoch": 0.13282178752836918,
      "grad_norm": 0.3908419907093048,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 1229
    },
    {
      "epoch": 0.13292986058575598,
      "grad_norm": 0.38681429624557495,
      "learning_rate": 8e-05,
      "loss": 1.515,
      "step": 1230
    },
    {
      "epoch": 0.13303793364314276,
      "grad_norm": 0.4256155490875244,
      "learning_rate": 8e-05,
      "loss": 1.8359,
      "step": 1231
    },
    {
      "epoch": 0.13314600670052956,
      "grad_norm": 0.3645309507846832,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1232
    },
    {
      "epoch": 0.13325407975791634,
      "grad_norm": 0.404152512550354,
      "learning_rate": 8e-05,
      "loss": 1.8185,
      "step": 1233
    },
    {
      "epoch": 0.13336215281530314,
      "grad_norm": 0.4428636133670807,
      "learning_rate": 8e-05,
      "loss": 1.9203,
      "step": 1234
    },
    {
      "epoch": 0.13347022587268995,
      "grad_norm": 0.3958292603492737,
      "learning_rate": 8e-05,
      "loss": 1.8179,
      "step": 1235
    },
    {
      "epoch": 0.13357829893007672,
      "grad_norm": 0.4069244861602783,
      "learning_rate": 8e-05,
      "loss": 1.9301,
      "step": 1236
    },
    {
      "epoch": 0.13368637198746353,
      "grad_norm": 0.40535345673561096,
      "learning_rate": 8e-05,
      "loss": 1.864,
      "step": 1237
    },
    {
      "epoch": 0.13379444504485033,
      "grad_norm": 0.3730022609233856,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 1238
    },
    {
      "epoch": 0.1339025181022371,
      "grad_norm": 0.377554327249527,
      "learning_rate": 8e-05,
      "loss": 1.7631,
      "step": 1239
    },
    {
      "epoch": 0.1340105911596239,
      "grad_norm": 0.4147307276725769,
      "learning_rate": 8e-05,
      "loss": 1.9833,
      "step": 1240
    },
    {
      "epoch": 0.13411866421701069,
      "grad_norm": 0.3798055350780487,
      "learning_rate": 8e-05,
      "loss": 1.5689,
      "step": 1241
    },
    {
      "epoch": 0.1342267372743975,
      "grad_norm": 0.3920557200908661,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 1242
    },
    {
      "epoch": 0.1343348103317843,
      "grad_norm": 0.41031184792518616,
      "learning_rate": 8e-05,
      "loss": 1.8194,
      "step": 1243
    },
    {
      "epoch": 0.13444288338917107,
      "grad_norm": 0.3833477199077606,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 1244
    },
    {
      "epoch": 0.13455095644655787,
      "grad_norm": 0.3701001703739166,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 1245
    },
    {
      "epoch": 0.13465902950394468,
      "grad_norm": 0.4241451025009155,
      "learning_rate": 8e-05,
      "loss": 1.7582,
      "step": 1246
    },
    {
      "epoch": 0.13476710256133145,
      "grad_norm": 0.4205266237258911,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 1247
    },
    {
      "epoch": 0.13487517561871826,
      "grad_norm": 0.3984798192977905,
      "learning_rate": 8e-05,
      "loss": 1.8517,
      "step": 1248
    },
    {
      "epoch": 0.13498324867610503,
      "grad_norm": 0.4076390266418457,
      "learning_rate": 8e-05,
      "loss": 1.9,
      "step": 1249
    },
    {
      "epoch": 0.13509132173349184,
      "grad_norm": 0.4262385368347168,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 1250
    },
    {
      "epoch": 0.13519939479087864,
      "grad_norm": 0.445039302110672,
      "learning_rate": 8e-05,
      "loss": 1.8663,
      "step": 1251
    },
    {
      "epoch": 0.13530746784826542,
      "grad_norm": 0.4578566551208496,
      "learning_rate": 8e-05,
      "loss": 2.0516,
      "step": 1252
    },
    {
      "epoch": 0.13541554090565222,
      "grad_norm": 0.3906839191913605,
      "learning_rate": 8e-05,
      "loss": 1.6006,
      "step": 1253
    },
    {
      "epoch": 0.13552361396303902,
      "grad_norm": 0.40319880843162537,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 1254
    },
    {
      "epoch": 0.1356316870204258,
      "grad_norm": 0.41592854261398315,
      "learning_rate": 8e-05,
      "loss": 1.8351,
      "step": 1255
    },
    {
      "epoch": 0.1357397600778126,
      "grad_norm": 0.4496399462223053,
      "learning_rate": 8e-05,
      "loss": 1.9483,
      "step": 1256
    },
    {
      "epoch": 0.1358478331351994,
      "grad_norm": 0.37799370288848877,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 1257
    },
    {
      "epoch": 0.13595590619258618,
      "grad_norm": 0.43647775053977966,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 1258
    },
    {
      "epoch": 0.136063979249973,
      "grad_norm": 0.3720110356807709,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 1259
    },
    {
      "epoch": 0.13617205230735976,
      "grad_norm": 0.4136310815811157,
      "learning_rate": 8e-05,
      "loss": 1.7783,
      "step": 1260
    },
    {
      "epoch": 0.13628012536474657,
      "grad_norm": 0.42075321078300476,
      "learning_rate": 8e-05,
      "loss": 1.8577,
      "step": 1261
    },
    {
      "epoch": 0.13638819842213337,
      "grad_norm": 0.40346062183380127,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 1262
    },
    {
      "epoch": 0.13649627147952015,
      "grad_norm": 0.42158621549606323,
      "learning_rate": 8e-05,
      "loss": 2.0141,
      "step": 1263
    },
    {
      "epoch": 0.13660434453690695,
      "grad_norm": 0.37127885222435,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 1264
    },
    {
      "epoch": 0.13671241759429376,
      "grad_norm": 0.43421873450279236,
      "learning_rate": 8e-05,
      "loss": 1.9164,
      "step": 1265
    },
    {
      "epoch": 0.13682049065168053,
      "grad_norm": 0.3877571225166321,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 1266
    },
    {
      "epoch": 0.13692856370906734,
      "grad_norm": 0.4213945269584656,
      "learning_rate": 8e-05,
      "loss": 1.8817,
      "step": 1267
    },
    {
      "epoch": 0.1370366367664541,
      "grad_norm": 0.39213433861732483,
      "learning_rate": 8e-05,
      "loss": 1.9049,
      "step": 1268
    },
    {
      "epoch": 0.13714470982384092,
      "grad_norm": 0.3990940451622009,
      "learning_rate": 8e-05,
      "loss": 1.8441,
      "step": 1269
    },
    {
      "epoch": 0.13725278288122772,
      "grad_norm": 0.4409485459327698,
      "learning_rate": 8e-05,
      "loss": 1.8561,
      "step": 1270
    },
    {
      "epoch": 0.1373608559386145,
      "grad_norm": 0.3895263373851776,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 1271
    },
    {
      "epoch": 0.1374689289960013,
      "grad_norm": 0.4066556692123413,
      "learning_rate": 8e-05,
      "loss": 1.9368,
      "step": 1272
    },
    {
      "epoch": 0.1375770020533881,
      "grad_norm": 0.3820422887802124,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 1273
    },
    {
      "epoch": 0.13768507511077488,
      "grad_norm": 0.4340505301952362,
      "learning_rate": 8e-05,
      "loss": 1.9366,
      "step": 1274
    },
    {
      "epoch": 0.13779314816816168,
      "grad_norm": 0.38965660333633423,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 1275
    },
    {
      "epoch": 0.13790122122554846,
      "grad_norm": 0.39434656500816345,
      "learning_rate": 8e-05,
      "loss": 1.6939,
      "step": 1276
    },
    {
      "epoch": 0.13800929428293526,
      "grad_norm": 0.3872416019439697,
      "learning_rate": 8e-05,
      "loss": 1.8784,
      "step": 1277
    },
    {
      "epoch": 0.13811736734032207,
      "grad_norm": 0.3595607876777649,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 1278
    },
    {
      "epoch": 0.13822544039770884,
      "grad_norm": 0.4131299555301666,
      "learning_rate": 8e-05,
      "loss": 1.9515,
      "step": 1279
    },
    {
      "epoch": 0.13833351345509565,
      "grad_norm": 0.39136162400245667,
      "learning_rate": 8e-05,
      "loss": 1.7523,
      "step": 1280
    },
    {
      "epoch": 0.13844158651248245,
      "grad_norm": 0.37980449199676514,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 1281
    },
    {
      "epoch": 0.13854965956986923,
      "grad_norm": 0.38759034872055054,
      "learning_rate": 8e-05,
      "loss": 1.7095,
      "step": 1282
    },
    {
      "epoch": 0.13865773262725603,
      "grad_norm": 0.39524465799331665,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 1283
    },
    {
      "epoch": 0.1387658056846428,
      "grad_norm": 0.36527031660079956,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 1284
    },
    {
      "epoch": 0.1388738787420296,
      "grad_norm": 0.37598392367362976,
      "learning_rate": 8e-05,
      "loss": 1.8104,
      "step": 1285
    },
    {
      "epoch": 0.13898195179941641,
      "grad_norm": 0.36540475487709045,
      "learning_rate": 8e-05,
      "loss": 1.6183,
      "step": 1286
    },
    {
      "epoch": 0.1390900248568032,
      "grad_norm": 0.40018120408058167,
      "learning_rate": 8e-05,
      "loss": 1.7977,
      "step": 1287
    },
    {
      "epoch": 0.13919809791419,
      "grad_norm": 0.40901613235473633,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 1288
    },
    {
      "epoch": 0.1393061709715768,
      "grad_norm": 0.3922746777534485,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 1289
    },
    {
      "epoch": 0.13941424402896357,
      "grad_norm": 0.41608402132987976,
      "learning_rate": 8e-05,
      "loss": 1.9427,
      "step": 1290
    },
    {
      "epoch": 0.13952231708635038,
      "grad_norm": 0.3799397647380829,
      "learning_rate": 8e-05,
      "loss": 1.5807,
      "step": 1291
    },
    {
      "epoch": 0.13963039014373715,
      "grad_norm": 0.3964638411998749,
      "learning_rate": 8e-05,
      "loss": 1.8796,
      "step": 1292
    },
    {
      "epoch": 0.13973846320112396,
      "grad_norm": 0.4084264636039734,
      "learning_rate": 8e-05,
      "loss": 1.792,
      "step": 1293
    },
    {
      "epoch": 0.13984653625851076,
      "grad_norm": 0.39225301146507263,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 1294
    },
    {
      "epoch": 0.13995460931589754,
      "grad_norm": 0.4328615069389343,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 1295
    },
    {
      "epoch": 0.14006268237328434,
      "grad_norm": 0.39989063143730164,
      "learning_rate": 8e-05,
      "loss": 1.9212,
      "step": 1296
    },
    {
      "epoch": 0.14017075543067115,
      "grad_norm": 0.39057958126068115,
      "learning_rate": 8e-05,
      "loss": 1.8271,
      "step": 1297
    },
    {
      "epoch": 0.14027882848805792,
      "grad_norm": 0.42341816425323486,
      "learning_rate": 8e-05,
      "loss": 1.8955,
      "step": 1298
    },
    {
      "epoch": 0.14038690154544473,
      "grad_norm": 0.41987112164497375,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 1299
    },
    {
      "epoch": 0.1404949746028315,
      "grad_norm": 0.42120304703712463,
      "learning_rate": 8e-05,
      "loss": 1.8321,
      "step": 1300
    },
    {
      "epoch": 0.1406030476602183,
      "grad_norm": 0.40507838129997253,
      "learning_rate": 8e-05,
      "loss": 1.6779,
      "step": 1301
    },
    {
      "epoch": 0.1407111207176051,
      "grad_norm": 0.3748818039894104,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 1302
    },
    {
      "epoch": 0.14081919377499189,
      "grad_norm": 0.4027266204357147,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 1303
    },
    {
      "epoch": 0.1409272668323787,
      "grad_norm": 0.39383551478385925,
      "learning_rate": 8e-05,
      "loss": 1.9536,
      "step": 1304
    },
    {
      "epoch": 0.1410353398897655,
      "grad_norm": 0.4074350595474243,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 1305
    },
    {
      "epoch": 0.14114341294715227,
      "grad_norm": 0.40671730041503906,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 1306
    },
    {
      "epoch": 0.14125148600453907,
      "grad_norm": 0.4193304777145386,
      "learning_rate": 8e-05,
      "loss": 1.8397,
      "step": 1307
    },
    {
      "epoch": 0.14135955906192585,
      "grad_norm": 0.39455175399780273,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 1308
    },
    {
      "epoch": 0.14146763211931265,
      "grad_norm": 0.42480507493019104,
      "learning_rate": 8e-05,
      "loss": 1.8144,
      "step": 1309
    },
    {
      "epoch": 0.14157570517669946,
      "grad_norm": 0.39346614480018616,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 1310
    },
    {
      "epoch": 0.14168377823408623,
      "grad_norm": 0.3644057810306549,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 1311
    },
    {
      "epoch": 0.14179185129147304,
      "grad_norm": 0.43666326999664307,
      "learning_rate": 8e-05,
      "loss": 1.94,
      "step": 1312
    },
    {
      "epoch": 0.14189992434885984,
      "grad_norm": 0.3853846490383148,
      "learning_rate": 8e-05,
      "loss": 1.7882,
      "step": 1313
    },
    {
      "epoch": 0.14200799740624662,
      "grad_norm": 0.41207680106163025,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 1314
    },
    {
      "epoch": 0.14211607046363342,
      "grad_norm": 0.38116154074668884,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 1315
    },
    {
      "epoch": 0.1422241435210202,
      "grad_norm": 0.3970804810523987,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 1316
    },
    {
      "epoch": 0.142332216578407,
      "grad_norm": 0.4064047932624817,
      "learning_rate": 8e-05,
      "loss": 1.823,
      "step": 1317
    },
    {
      "epoch": 0.1424402896357938,
      "grad_norm": 0.42533430457115173,
      "learning_rate": 8e-05,
      "loss": 1.9631,
      "step": 1318
    },
    {
      "epoch": 0.14254836269318058,
      "grad_norm": 0.39447474479675293,
      "learning_rate": 8e-05,
      "loss": 1.5401,
      "step": 1319
    },
    {
      "epoch": 0.14265643575056738,
      "grad_norm": 0.39933955669403076,
      "learning_rate": 8e-05,
      "loss": 1.7127,
      "step": 1320
    },
    {
      "epoch": 0.1427645088079542,
      "grad_norm": 0.43245768547058105,
      "learning_rate": 8e-05,
      "loss": 1.8622,
      "step": 1321
    },
    {
      "epoch": 0.14287258186534096,
      "grad_norm": 0.4235605299472809,
      "learning_rate": 8e-05,
      "loss": 1.7858,
      "step": 1322
    },
    {
      "epoch": 0.14298065492272777,
      "grad_norm": 0.40082457661628723,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1323
    },
    {
      "epoch": 0.14308872798011454,
      "grad_norm": 0.39223310351371765,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 1324
    },
    {
      "epoch": 0.14319680103750135,
      "grad_norm": 0.4682566821575165,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 1325
    },
    {
      "epoch": 0.14330487409488815,
      "grad_norm": 0.37665924429893494,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1326
    },
    {
      "epoch": 0.14341294715227493,
      "grad_norm": 0.38328301906585693,
      "learning_rate": 8e-05,
      "loss": 1.6465,
      "step": 1327
    },
    {
      "epoch": 0.14352102020966173,
      "grad_norm": 0.4310777485370636,
      "learning_rate": 8e-05,
      "loss": 1.8698,
      "step": 1328
    },
    {
      "epoch": 0.14362909326704854,
      "grad_norm": 0.42203181982040405,
      "learning_rate": 8e-05,
      "loss": 1.8241,
      "step": 1329
    },
    {
      "epoch": 0.1437371663244353,
      "grad_norm": 0.4148368537425995,
      "learning_rate": 8e-05,
      "loss": 1.7585,
      "step": 1330
    },
    {
      "epoch": 0.14384523938182212,
      "grad_norm": 0.3902142643928528,
      "learning_rate": 8e-05,
      "loss": 1.8508,
      "step": 1331
    },
    {
      "epoch": 0.1439533124392089,
      "grad_norm": 0.3930110037326813,
      "learning_rate": 8e-05,
      "loss": 1.8194,
      "step": 1332
    },
    {
      "epoch": 0.1440613854965957,
      "grad_norm": 0.4176648259162903,
      "learning_rate": 8e-05,
      "loss": 1.8251,
      "step": 1333
    },
    {
      "epoch": 0.1441694585539825,
      "grad_norm": 0.42532312870025635,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 1334
    },
    {
      "epoch": 0.14427753161136928,
      "grad_norm": 0.40888214111328125,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 1335
    },
    {
      "epoch": 0.14438560466875608,
      "grad_norm": 0.413207471370697,
      "learning_rate": 8e-05,
      "loss": 1.7954,
      "step": 1336
    },
    {
      "epoch": 0.14449367772614288,
      "grad_norm": 0.3896910846233368,
      "learning_rate": 8e-05,
      "loss": 1.953,
      "step": 1337
    },
    {
      "epoch": 0.14460175078352966,
      "grad_norm": 0.4006461501121521,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 1338
    },
    {
      "epoch": 0.14470982384091646,
      "grad_norm": 0.41400015354156494,
      "learning_rate": 8e-05,
      "loss": 1.8913,
      "step": 1339
    },
    {
      "epoch": 0.14481789689830327,
      "grad_norm": 0.4077097773551941,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 1340
    },
    {
      "epoch": 0.14492596995569004,
      "grad_norm": 0.4090961813926697,
      "learning_rate": 8e-05,
      "loss": 1.7902,
      "step": 1341
    },
    {
      "epoch": 0.14503404301307685,
      "grad_norm": 0.43312036991119385,
      "learning_rate": 8e-05,
      "loss": 1.9318,
      "step": 1342
    },
    {
      "epoch": 0.14514211607046362,
      "grad_norm": 0.4037969410419464,
      "learning_rate": 8e-05,
      "loss": 1.9378,
      "step": 1343
    },
    {
      "epoch": 0.14525018912785043,
      "grad_norm": 0.36318567395210266,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 1344
    },
    {
      "epoch": 0.14535826218523723,
      "grad_norm": 0.40497443079948425,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 1345
    },
    {
      "epoch": 0.145466335242624,
      "grad_norm": 0.3913719654083252,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 1346
    },
    {
      "epoch": 0.1455744083000108,
      "grad_norm": 0.4039226472377777,
      "learning_rate": 8e-05,
      "loss": 1.7127,
      "step": 1347
    },
    {
      "epoch": 0.14568248135739761,
      "grad_norm": 0.4129995107650757,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 1348
    },
    {
      "epoch": 0.1457905544147844,
      "grad_norm": 0.3982151448726654,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 1349
    },
    {
      "epoch": 0.1458986274721712,
      "grad_norm": 0.4084954857826233,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 1350
    },
    {
      "epoch": 0.14600670052955797,
      "grad_norm": 0.42660748958587646,
      "learning_rate": 8e-05,
      "loss": 1.6466,
      "step": 1351
    },
    {
      "epoch": 0.14611477358694477,
      "grad_norm": 0.4494730234146118,
      "learning_rate": 8e-05,
      "loss": 1.5923,
      "step": 1352
    },
    {
      "epoch": 0.14622284664433158,
      "grad_norm": 0.439540833234787,
      "learning_rate": 8e-05,
      "loss": 1.8691,
      "step": 1353
    },
    {
      "epoch": 0.14633091970171835,
      "grad_norm": 0.3815706670284271,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 1354
    },
    {
      "epoch": 0.14643899275910516,
      "grad_norm": 0.38582998514175415,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 1355
    },
    {
      "epoch": 0.14654706581649196,
      "grad_norm": 0.36811143159866333,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 1356
    },
    {
      "epoch": 0.14665513887387874,
      "grad_norm": 0.3913086950778961,
      "learning_rate": 8e-05,
      "loss": 1.8581,
      "step": 1357
    },
    {
      "epoch": 0.14676321193126554,
      "grad_norm": 0.41038426756858826,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 1358
    },
    {
      "epoch": 0.14687128498865232,
      "grad_norm": 0.40928444266319275,
      "learning_rate": 8e-05,
      "loss": 1.9031,
      "step": 1359
    },
    {
      "epoch": 0.14697935804603912,
      "grad_norm": 0.40353333950042725,
      "learning_rate": 8e-05,
      "loss": 1.8741,
      "step": 1360
    },
    {
      "epoch": 0.14708743110342593,
      "grad_norm": 0.4253388047218323,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 1361
    },
    {
      "epoch": 0.1471955041608127,
      "grad_norm": 0.3913874924182892,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 1362
    },
    {
      "epoch": 0.1473035772181995,
      "grad_norm": 0.37585681676864624,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 1363
    },
    {
      "epoch": 0.1474116502755863,
      "grad_norm": 0.40192946791648865,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 1364
    },
    {
      "epoch": 0.14751972333297309,
      "grad_norm": 0.39640966057777405,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 1365
    },
    {
      "epoch": 0.1476277963903599,
      "grad_norm": 0.4274112284183502,
      "learning_rate": 8e-05,
      "loss": 1.8124,
      "step": 1366
    },
    {
      "epoch": 0.14773586944774666,
      "grad_norm": 0.3968859314918518,
      "learning_rate": 8e-05,
      "loss": 1.8123,
      "step": 1367
    },
    {
      "epoch": 0.14784394250513347,
      "grad_norm": 0.392220675945282,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 1368
    },
    {
      "epoch": 0.14795201556252027,
      "grad_norm": 0.3905208110809326,
      "learning_rate": 8e-05,
      "loss": 1.8876,
      "step": 1369
    },
    {
      "epoch": 0.14806008861990705,
      "grad_norm": 0.3752386271953583,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 1370
    },
    {
      "epoch": 0.14816816167729385,
      "grad_norm": 0.405080646276474,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 1371
    },
    {
      "epoch": 0.14827623473468066,
      "grad_norm": 0.4143310487270355,
      "learning_rate": 8e-05,
      "loss": 1.9211,
      "step": 1372
    },
    {
      "epoch": 0.14838430779206743,
      "grad_norm": 0.41946837306022644,
      "learning_rate": 8e-05,
      "loss": 2.0229,
      "step": 1373
    },
    {
      "epoch": 0.14849238084945424,
      "grad_norm": 0.3996814489364624,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 1374
    },
    {
      "epoch": 0.148600453906841,
      "grad_norm": 0.3907395005226135,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 1375
    },
    {
      "epoch": 0.14870852696422782,
      "grad_norm": 0.3808145523071289,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 1376
    },
    {
      "epoch": 0.14881660002161462,
      "grad_norm": 0.43856972455978394,
      "learning_rate": 8e-05,
      "loss": 1.953,
      "step": 1377
    },
    {
      "epoch": 0.1489246730790014,
      "grad_norm": 0.38694772124290466,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 1378
    },
    {
      "epoch": 0.1490327461363882,
      "grad_norm": 0.3729066550731659,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1379
    },
    {
      "epoch": 0.149140819193775,
      "grad_norm": 0.4187312126159668,
      "learning_rate": 8e-05,
      "loss": 1.9331,
      "step": 1380
    },
    {
      "epoch": 0.14924889225116178,
      "grad_norm": 0.43742454051971436,
      "learning_rate": 8e-05,
      "loss": 1.9266,
      "step": 1381
    },
    {
      "epoch": 0.14935696530854858,
      "grad_norm": 0.38988253474235535,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 1382
    },
    {
      "epoch": 0.14946503836593536,
      "grad_norm": 0.3989926278591156,
      "learning_rate": 8e-05,
      "loss": 1.8657,
      "step": 1383
    },
    {
      "epoch": 0.14957311142332216,
      "grad_norm": 0.3895411193370819,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 1384
    },
    {
      "epoch": 0.14968118448070897,
      "grad_norm": 0.42214536666870117,
      "learning_rate": 8e-05,
      "loss": 1.7703,
      "step": 1385
    },
    {
      "epoch": 0.14978925753809574,
      "grad_norm": 0.4188408851623535,
      "learning_rate": 8e-05,
      "loss": 1.8905,
      "step": 1386
    },
    {
      "epoch": 0.14989733059548255,
      "grad_norm": 0.3974918723106384,
      "learning_rate": 8e-05,
      "loss": 1.7996,
      "step": 1387
    },
    {
      "epoch": 0.15000540365286935,
      "grad_norm": 0.3807678818702698,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 1388
    },
    {
      "epoch": 0.15011347671025613,
      "grad_norm": 0.414058119058609,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 1389
    },
    {
      "epoch": 0.15022154976764293,
      "grad_norm": 0.40361881256103516,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 1390
    },
    {
      "epoch": 0.1503296228250297,
      "grad_norm": 0.3801555633544922,
      "learning_rate": 8e-05,
      "loss": 1.8158,
      "step": 1391
    },
    {
      "epoch": 0.1504376958824165,
      "grad_norm": 0.4011727571487427,
      "learning_rate": 8e-05,
      "loss": 1.8058,
      "step": 1392
    },
    {
      "epoch": 0.15054576893980331,
      "grad_norm": 0.4003799557685852,
      "learning_rate": 8e-05,
      "loss": 1.8063,
      "step": 1393
    },
    {
      "epoch": 0.1506538419971901,
      "grad_norm": 0.38795170187950134,
      "learning_rate": 8e-05,
      "loss": 1.8607,
      "step": 1394
    },
    {
      "epoch": 0.1507619150545769,
      "grad_norm": 0.37822023034095764,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 1395
    },
    {
      "epoch": 0.1508699881119637,
      "grad_norm": 0.37043896317481995,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 1396
    },
    {
      "epoch": 0.15097806116935047,
      "grad_norm": 0.42876702547073364,
      "learning_rate": 8e-05,
      "loss": 1.9964,
      "step": 1397
    },
    {
      "epoch": 0.15108613422673728,
      "grad_norm": 0.39591673016548157,
      "learning_rate": 8e-05,
      "loss": 1.8964,
      "step": 1398
    },
    {
      "epoch": 0.15119420728412405,
      "grad_norm": 0.4254635274410248,
      "learning_rate": 8e-05,
      "loss": 1.8273,
      "step": 1399
    },
    {
      "epoch": 0.15130228034151086,
      "grad_norm": 0.3852804899215698,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 1400
    },
    {
      "epoch": 0.15141035339889766,
      "grad_norm": 0.36417242884635925,
      "learning_rate": 8e-05,
      "loss": 1.7668,
      "step": 1401
    },
    {
      "epoch": 0.15151842645628444,
      "grad_norm": 0.40901950001716614,
      "learning_rate": 8e-05,
      "loss": 1.8645,
      "step": 1402
    },
    {
      "epoch": 0.15162649951367124,
      "grad_norm": 0.36646831035614014,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 1403
    },
    {
      "epoch": 0.15173457257105805,
      "grad_norm": 0.44161540269851685,
      "learning_rate": 8e-05,
      "loss": 1.8868,
      "step": 1404
    },
    {
      "epoch": 0.15184264562844482,
      "grad_norm": 0.4340607523918152,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1405
    },
    {
      "epoch": 0.15195071868583163,
      "grad_norm": 0.3681427240371704,
      "learning_rate": 8e-05,
      "loss": 1.6132,
      "step": 1406
    },
    {
      "epoch": 0.1520587917432184,
      "grad_norm": 0.38938456773757935,
      "learning_rate": 8e-05,
      "loss": 1.8009,
      "step": 1407
    },
    {
      "epoch": 0.1521668648006052,
      "grad_norm": 0.40488049387931824,
      "learning_rate": 8e-05,
      "loss": 1.8626,
      "step": 1408
    },
    {
      "epoch": 0.152274937857992,
      "grad_norm": 0.4419094920158386,
      "learning_rate": 8e-05,
      "loss": 1.9319,
      "step": 1409
    },
    {
      "epoch": 0.15238301091537879,
      "grad_norm": 0.4495726227760315,
      "learning_rate": 8e-05,
      "loss": 1.9899,
      "step": 1410
    },
    {
      "epoch": 0.1524910839727656,
      "grad_norm": 0.41366398334503174,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 1411
    },
    {
      "epoch": 0.1525991570301524,
      "grad_norm": 0.42947179079055786,
      "learning_rate": 8e-05,
      "loss": 1.8431,
      "step": 1412
    },
    {
      "epoch": 0.15270723008753917,
      "grad_norm": 0.418077290058136,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 1413
    },
    {
      "epoch": 0.15281530314492597,
      "grad_norm": 0.4075997769832611,
      "learning_rate": 8e-05,
      "loss": 1.9473,
      "step": 1414
    },
    {
      "epoch": 0.15292337620231275,
      "grad_norm": 0.41921573877334595,
      "learning_rate": 8e-05,
      "loss": 1.8206,
      "step": 1415
    },
    {
      "epoch": 0.15303144925969955,
      "grad_norm": 0.3738824129104614,
      "learning_rate": 8e-05,
      "loss": 1.796,
      "step": 1416
    },
    {
      "epoch": 0.15313952231708636,
      "grad_norm": 0.39668792486190796,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 1417
    },
    {
      "epoch": 0.15324759537447313,
      "grad_norm": 0.4195786118507385,
      "learning_rate": 8e-05,
      "loss": 1.9421,
      "step": 1418
    },
    {
      "epoch": 0.15335566843185994,
      "grad_norm": 0.4117182195186615,
      "learning_rate": 8e-05,
      "loss": 1.9057,
      "step": 1419
    },
    {
      "epoch": 0.15346374148924674,
      "grad_norm": 0.3847416937351227,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 1420
    },
    {
      "epoch": 0.15357181454663352,
      "grad_norm": 0.40914633870124817,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 1421
    },
    {
      "epoch": 0.15367988760402032,
      "grad_norm": 0.39840057492256165,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 1422
    },
    {
      "epoch": 0.15378796066140712,
      "grad_norm": 0.39120814204216003,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 1423
    },
    {
      "epoch": 0.1538960337187939,
      "grad_norm": 0.38623711466789246,
      "learning_rate": 8e-05,
      "loss": 1.8024,
      "step": 1424
    },
    {
      "epoch": 0.1540041067761807,
      "grad_norm": 0.39663103222846985,
      "learning_rate": 8e-05,
      "loss": 1.5894,
      "step": 1425
    },
    {
      "epoch": 0.15411217983356748,
      "grad_norm": 0.3620992600917816,
      "learning_rate": 8e-05,
      "loss": 1.6247,
      "step": 1426
    },
    {
      "epoch": 0.15422025289095428,
      "grad_norm": 0.40339988470077515,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 1427
    },
    {
      "epoch": 0.1543283259483411,
      "grad_norm": 0.45053473114967346,
      "learning_rate": 8e-05,
      "loss": 1.9153,
      "step": 1428
    },
    {
      "epoch": 0.15443639900572786,
      "grad_norm": 0.4230833053588867,
      "learning_rate": 8e-05,
      "loss": 1.8565,
      "step": 1429
    },
    {
      "epoch": 0.15454447206311467,
      "grad_norm": 0.4107688367366791,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 1430
    },
    {
      "epoch": 0.15465254512050147,
      "grad_norm": 0.3871934711933136,
      "learning_rate": 8e-05,
      "loss": 1.7464,
      "step": 1431
    },
    {
      "epoch": 0.15476061817788825,
      "grad_norm": 0.3824099600315094,
      "learning_rate": 8e-05,
      "loss": 1.5255,
      "step": 1432
    },
    {
      "epoch": 0.15486869123527505,
      "grad_norm": 0.4330695867538452,
      "learning_rate": 8e-05,
      "loss": 1.843,
      "step": 1433
    },
    {
      "epoch": 0.15497676429266183,
      "grad_norm": 0.42967167496681213,
      "learning_rate": 8e-05,
      "loss": 1.8849,
      "step": 1434
    },
    {
      "epoch": 0.15508483735004863,
      "grad_norm": 0.4230102598667145,
      "learning_rate": 8e-05,
      "loss": 1.937,
      "step": 1435
    },
    {
      "epoch": 0.15519291040743544,
      "grad_norm": 0.4063728451728821,
      "learning_rate": 8e-05,
      "loss": 1.9183,
      "step": 1436
    },
    {
      "epoch": 0.1553009834648222,
      "grad_norm": 0.37054744362831116,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 1437
    },
    {
      "epoch": 0.15540905652220902,
      "grad_norm": 0.4215314984321594,
      "learning_rate": 8e-05,
      "loss": 2.0292,
      "step": 1438
    },
    {
      "epoch": 0.15551712957959582,
      "grad_norm": 0.38475707173347473,
      "learning_rate": 8e-05,
      "loss": 1.6941,
      "step": 1439
    },
    {
      "epoch": 0.1556252026369826,
      "grad_norm": 0.38501936197280884,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 1440
    },
    {
      "epoch": 0.1557332756943694,
      "grad_norm": 0.3942856788635254,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 1441
    },
    {
      "epoch": 0.15584134875175618,
      "grad_norm": 0.3707219064235687,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 1442
    },
    {
      "epoch": 0.15594942180914298,
      "grad_norm": 0.3891001045703888,
      "learning_rate": 8e-05,
      "loss": 1.6159,
      "step": 1443
    },
    {
      "epoch": 0.15605749486652978,
      "grad_norm": 0.42341604828834534,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 1444
    },
    {
      "epoch": 0.15616556792391656,
      "grad_norm": 0.37214645743370056,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1445
    },
    {
      "epoch": 0.15627364098130336,
      "grad_norm": 0.3912259340286255,
      "learning_rate": 8e-05,
      "loss": 1.6161,
      "step": 1446
    },
    {
      "epoch": 0.15638171403869017,
      "grad_norm": 0.391072541475296,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 1447
    },
    {
      "epoch": 0.15648978709607694,
      "grad_norm": 0.41521722078323364,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 1448
    },
    {
      "epoch": 0.15659786015346375,
      "grad_norm": 0.39517128467559814,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 1449
    },
    {
      "epoch": 0.15670593321085052,
      "grad_norm": 0.39088064432144165,
      "learning_rate": 8e-05,
      "loss": 1.8907,
      "step": 1450
    },
    {
      "epoch": 0.15681400626823733,
      "grad_norm": 0.3810468018054962,
      "learning_rate": 8e-05,
      "loss": 1.783,
      "step": 1451
    },
    {
      "epoch": 0.15692207932562413,
      "grad_norm": 0.4340175688266754,
      "learning_rate": 8e-05,
      "loss": 1.9077,
      "step": 1452
    },
    {
      "epoch": 0.1570301523830109,
      "grad_norm": 0.37263646721839905,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 1453
    },
    {
      "epoch": 0.1571382254403977,
      "grad_norm": 0.37111836671829224,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 1454
    },
    {
      "epoch": 0.15724629849778451,
      "grad_norm": 0.44027191400527954,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 1455
    },
    {
      "epoch": 0.1573543715551713,
      "grad_norm": 0.42350974678993225,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 1456
    },
    {
      "epoch": 0.1574624446125581,
      "grad_norm": 0.406864732503891,
      "learning_rate": 8e-05,
      "loss": 1.9433,
      "step": 1457
    },
    {
      "epoch": 0.15757051766994487,
      "grad_norm": 0.42471247911453247,
      "learning_rate": 8e-05,
      "loss": 1.9248,
      "step": 1458
    },
    {
      "epoch": 0.15767859072733167,
      "grad_norm": 0.3934206962585449,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 1459
    },
    {
      "epoch": 0.15778666378471848,
      "grad_norm": 0.4054968059062958,
      "learning_rate": 8e-05,
      "loss": 1.8239,
      "step": 1460
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 0.47308802604675293,
      "learning_rate": 8e-05,
      "loss": 1.8693,
      "step": 1461
    },
    {
      "epoch": 0.15800280989949206,
      "grad_norm": 0.4168563783168793,
      "learning_rate": 8e-05,
      "loss": 1.81,
      "step": 1462
    },
    {
      "epoch": 0.15811088295687886,
      "grad_norm": 0.38998469710350037,
      "learning_rate": 8e-05,
      "loss": 1.8573,
      "step": 1463
    },
    {
      "epoch": 0.15821895601426564,
      "grad_norm": 0.40494465827941895,
      "learning_rate": 8e-05,
      "loss": 1.7594,
      "step": 1464
    },
    {
      "epoch": 0.15832702907165244,
      "grad_norm": 0.4300422966480255,
      "learning_rate": 8e-05,
      "loss": 1.7378,
      "step": 1465
    },
    {
      "epoch": 0.15843510212903922,
      "grad_norm": 0.4164920747280121,
      "learning_rate": 8e-05,
      "loss": 1.7996,
      "step": 1466
    },
    {
      "epoch": 0.15854317518642602,
      "grad_norm": 0.3775363564491272,
      "learning_rate": 8e-05,
      "loss": 1.6456,
      "step": 1467
    },
    {
      "epoch": 0.15865124824381283,
      "grad_norm": 0.42360371351242065,
      "learning_rate": 8e-05,
      "loss": 1.8423,
      "step": 1468
    },
    {
      "epoch": 0.1587593213011996,
      "grad_norm": 0.3892897367477417,
      "learning_rate": 8e-05,
      "loss": 1.7802,
      "step": 1469
    },
    {
      "epoch": 0.1588673943585864,
      "grad_norm": 0.4231686294078827,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 1470
    },
    {
      "epoch": 0.1589754674159732,
      "grad_norm": 0.41029295325279236,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 1471
    },
    {
      "epoch": 0.15908354047335999,
      "grad_norm": 0.3937508165836334,
      "learning_rate": 8e-05,
      "loss": 1.7834,
      "step": 1472
    },
    {
      "epoch": 0.1591916135307468,
      "grad_norm": 0.39608779549598694,
      "learning_rate": 8e-05,
      "loss": 1.6651,
      "step": 1473
    },
    {
      "epoch": 0.15929968658813357,
      "grad_norm": 0.3780694305896759,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 1474
    },
    {
      "epoch": 0.15940775964552037,
      "grad_norm": 0.3769053518772125,
      "learning_rate": 8e-05,
      "loss": 1.6875,
      "step": 1475
    },
    {
      "epoch": 0.15951583270290717,
      "grad_norm": 0.6450949311256409,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 1476
    },
    {
      "epoch": 0.15962390576029395,
      "grad_norm": 0.37835192680358887,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 1477
    },
    {
      "epoch": 0.15973197881768075,
      "grad_norm": 0.3712300658226013,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 1478
    },
    {
      "epoch": 0.15984005187506756,
      "grad_norm": 0.3998165428638458,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 1479
    },
    {
      "epoch": 0.15994812493245433,
      "grad_norm": 0.38400542736053467,
      "learning_rate": 8e-05,
      "loss": 1.8248,
      "step": 1480
    },
    {
      "epoch": 0.16005619798984114,
      "grad_norm": 0.39449164271354675,
      "learning_rate": 8e-05,
      "loss": 1.8189,
      "step": 1481
    },
    {
      "epoch": 0.1601642710472279,
      "grad_norm": 0.38770464062690735,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 1482
    },
    {
      "epoch": 0.16027234410461472,
      "grad_norm": 0.389159619808197,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 1483
    },
    {
      "epoch": 0.16038041716200152,
      "grad_norm": 0.36357641220092773,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 1484
    },
    {
      "epoch": 0.1604884902193883,
      "grad_norm": 0.41774117946624756,
      "learning_rate": 8e-05,
      "loss": 1.9602,
      "step": 1485
    },
    {
      "epoch": 0.1605965632767751,
      "grad_norm": 0.38568463921546936,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1486
    },
    {
      "epoch": 0.1607046363341619,
      "grad_norm": 0.4055615961551666,
      "learning_rate": 8e-05,
      "loss": 1.6312,
      "step": 1487
    },
    {
      "epoch": 0.16081270939154868,
      "grad_norm": 0.4082942008972168,
      "learning_rate": 8e-05,
      "loss": 1.8102,
      "step": 1488
    },
    {
      "epoch": 0.16092078244893548,
      "grad_norm": 0.39933234453201294,
      "learning_rate": 8e-05,
      "loss": 1.9435,
      "step": 1489
    },
    {
      "epoch": 0.16102885550632226,
      "grad_norm": 0.38904184103012085,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 1490
    },
    {
      "epoch": 0.16113692856370906,
      "grad_norm": 0.3767629861831665,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 1491
    },
    {
      "epoch": 0.16124500162109587,
      "grad_norm": 0.4121721684932709,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 1492
    },
    {
      "epoch": 0.16135307467848264,
      "grad_norm": 0.4112269878387451,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 1493
    },
    {
      "epoch": 0.16146114773586945,
      "grad_norm": 0.42926132678985596,
      "learning_rate": 8e-05,
      "loss": 1.8433,
      "step": 1494
    },
    {
      "epoch": 0.16156922079325625,
      "grad_norm": 0.42812579870224,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 1495
    },
    {
      "epoch": 0.16167729385064303,
      "grad_norm": 0.39054766297340393,
      "learning_rate": 8e-05,
      "loss": 1.6143,
      "step": 1496
    },
    {
      "epoch": 0.16178536690802983,
      "grad_norm": 0.46550020575523376,
      "learning_rate": 8e-05,
      "loss": 2.0209,
      "step": 1497
    },
    {
      "epoch": 0.1618934399654166,
      "grad_norm": 0.4069766402244568,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 1498
    },
    {
      "epoch": 0.1620015130228034,
      "grad_norm": 0.4025830328464508,
      "learning_rate": 8e-05,
      "loss": 1.8052,
      "step": 1499
    },
    {
      "epoch": 0.16210958608019022,
      "grad_norm": 0.447213739156723,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 1500
    },
    {
      "epoch": 0.162217659137577,
      "grad_norm": 0.4063083529472351,
      "learning_rate": 8e-05,
      "loss": 1.8233,
      "step": 1501
    },
    {
      "epoch": 0.1623257321949638,
      "grad_norm": 0.39696040749549866,
      "learning_rate": 8e-05,
      "loss": 1.7626,
      "step": 1502
    },
    {
      "epoch": 0.1624338052523506,
      "grad_norm": 0.4074174761772156,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 1503
    },
    {
      "epoch": 0.16254187830973738,
      "grad_norm": 0.37727272510528564,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 1504
    },
    {
      "epoch": 0.16264995136712418,
      "grad_norm": 0.39036694169044495,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 1505
    },
    {
      "epoch": 0.16275802442451098,
      "grad_norm": 0.3832937180995941,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 1506
    },
    {
      "epoch": 0.16286609748189776,
      "grad_norm": 0.39932259917259216,
      "learning_rate": 8e-05,
      "loss": 1.5892,
      "step": 1507
    },
    {
      "epoch": 0.16297417053928456,
      "grad_norm": 0.3797241449356079,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 1508
    },
    {
      "epoch": 0.16308224359667134,
      "grad_norm": 0.400756299495697,
      "learning_rate": 8e-05,
      "loss": 1.8202,
      "step": 1509
    },
    {
      "epoch": 0.16319031665405814,
      "grad_norm": 0.43164893984794617,
      "learning_rate": 8e-05,
      "loss": 1.9634,
      "step": 1510
    },
    {
      "epoch": 0.16329838971144495,
      "grad_norm": 0.4244467616081238,
      "learning_rate": 8e-05,
      "loss": 1.9454,
      "step": 1511
    },
    {
      "epoch": 0.16340646276883172,
      "grad_norm": 0.4116677939891815,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 1512
    },
    {
      "epoch": 0.16351453582621853,
      "grad_norm": 0.366131067276001,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 1513
    },
    {
      "epoch": 0.16362260888360533,
      "grad_norm": 0.40200701355934143,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 1514
    },
    {
      "epoch": 0.1637306819409921,
      "grad_norm": 0.37406814098358154,
      "learning_rate": 8e-05,
      "loss": 1.6505,
      "step": 1515
    },
    {
      "epoch": 0.1638387549983789,
      "grad_norm": 0.38084056973457336,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 1516
    },
    {
      "epoch": 0.1639468280557657,
      "grad_norm": 0.37383729219436646,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 1517
    },
    {
      "epoch": 0.1640549011131525,
      "grad_norm": 0.37796831130981445,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 1518
    },
    {
      "epoch": 0.1641629741705393,
      "grad_norm": 0.4145317077636719,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 1519
    },
    {
      "epoch": 0.16427104722792607,
      "grad_norm": 0.43362754583358765,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 1520
    },
    {
      "epoch": 0.16437912028531287,
      "grad_norm": 0.43673744797706604,
      "learning_rate": 8e-05,
      "loss": 1.8532,
      "step": 1521
    },
    {
      "epoch": 0.16448719334269968,
      "grad_norm": 0.38099274039268494,
      "learning_rate": 8e-05,
      "loss": 1.7953,
      "step": 1522
    },
    {
      "epoch": 0.16459526640008645,
      "grad_norm": 0.41307979822158813,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 1523
    },
    {
      "epoch": 0.16470333945747326,
      "grad_norm": 0.42952287197113037,
      "learning_rate": 8e-05,
      "loss": 1.8443,
      "step": 1524
    },
    {
      "epoch": 0.16481141251486003,
      "grad_norm": 0.4296877980232239,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 1525
    },
    {
      "epoch": 0.16491948557224684,
      "grad_norm": 0.4383057653903961,
      "learning_rate": 8e-05,
      "loss": 1.9252,
      "step": 1526
    },
    {
      "epoch": 0.16502755862963364,
      "grad_norm": 0.41945961117744446,
      "learning_rate": 8e-05,
      "loss": 1.9316,
      "step": 1527
    },
    {
      "epoch": 0.16513563168702042,
      "grad_norm": 0.4046679139137268,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 1528
    },
    {
      "epoch": 0.16524370474440722,
      "grad_norm": 0.40208953619003296,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 1529
    },
    {
      "epoch": 0.16535177780179403,
      "grad_norm": 0.3992357552051544,
      "learning_rate": 8e-05,
      "loss": 1.9465,
      "step": 1530
    },
    {
      "epoch": 0.1654598508591808,
      "grad_norm": 0.4016571044921875,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 1531
    },
    {
      "epoch": 0.1655679239165676,
      "grad_norm": 0.4040756821632385,
      "learning_rate": 8e-05,
      "loss": 1.8131,
      "step": 1532
    },
    {
      "epoch": 0.16567599697395438,
      "grad_norm": 0.39432433247566223,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 1533
    },
    {
      "epoch": 0.16578407003134119,
      "grad_norm": 0.391748309135437,
      "learning_rate": 8e-05,
      "loss": 1.8039,
      "step": 1534
    },
    {
      "epoch": 0.165892143088728,
      "grad_norm": 0.39344322681427,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 1535
    },
    {
      "epoch": 0.16600021614611477,
      "grad_norm": 0.39443305134773254,
      "learning_rate": 8e-05,
      "loss": 1.7881,
      "step": 1536
    },
    {
      "epoch": 0.16610828920350157,
      "grad_norm": 0.4263917803764343,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 1537
    },
    {
      "epoch": 0.16621636226088837,
      "grad_norm": 0.3904288411140442,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 1538
    },
    {
      "epoch": 0.16632443531827515,
      "grad_norm": 0.40540793538093567,
      "learning_rate": 8e-05,
      "loss": 1.7255,
      "step": 1539
    },
    {
      "epoch": 0.16643250837566195,
      "grad_norm": 0.3984065651893616,
      "learning_rate": 8e-05,
      "loss": 1.7534,
      "step": 1540
    },
    {
      "epoch": 0.16654058143304873,
      "grad_norm": 0.39166060090065,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 1541
    },
    {
      "epoch": 0.16664865449043553,
      "grad_norm": 0.3923915922641754,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 1542
    },
    {
      "epoch": 0.16675672754782234,
      "grad_norm": 0.4101788401603699,
      "learning_rate": 8e-05,
      "loss": 1.83,
      "step": 1543
    },
    {
      "epoch": 0.1668648006052091,
      "grad_norm": 0.4310155510902405,
      "learning_rate": 8e-05,
      "loss": 1.5683,
      "step": 1544
    },
    {
      "epoch": 0.16697287366259592,
      "grad_norm": 0.3802625238895416,
      "learning_rate": 8e-05,
      "loss": 1.8909,
      "step": 1545
    },
    {
      "epoch": 0.16708094671998272,
      "grad_norm": 0.39147937297821045,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 1546
    },
    {
      "epoch": 0.1671890197773695,
      "grad_norm": 0.3863431513309479,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1547
    },
    {
      "epoch": 0.1672970928347563,
      "grad_norm": 0.38621121644973755,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 1548
    },
    {
      "epoch": 0.16740516589214308,
      "grad_norm": 0.41960301995277405,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 1549
    },
    {
      "epoch": 0.16751323894952988,
      "grad_norm": 0.4684583246707916,
      "learning_rate": 8e-05,
      "loss": 1.9733,
      "step": 1550
    },
    {
      "epoch": 0.16762131200691668,
      "grad_norm": 0.42110803723335266,
      "learning_rate": 8e-05,
      "loss": 1.8237,
      "step": 1551
    },
    {
      "epoch": 0.16772938506430346,
      "grad_norm": 0.3823283016681671,
      "learning_rate": 8e-05,
      "loss": 1.9261,
      "step": 1552
    },
    {
      "epoch": 0.16783745812169026,
      "grad_norm": 0.4116194248199463,
      "learning_rate": 8e-05,
      "loss": 1.8561,
      "step": 1553
    },
    {
      "epoch": 0.16794553117907707,
      "grad_norm": 0.4266180694103241,
      "learning_rate": 8e-05,
      "loss": 1.7411,
      "step": 1554
    },
    {
      "epoch": 0.16805360423646384,
      "grad_norm": 0.4048207104206085,
      "learning_rate": 8e-05,
      "loss": 1.9864,
      "step": 1555
    },
    {
      "epoch": 0.16816167729385065,
      "grad_norm": 0.42027974128723145,
      "learning_rate": 8e-05,
      "loss": 1.8378,
      "step": 1556
    },
    {
      "epoch": 0.16826975035123742,
      "grad_norm": 0.44342485070228577,
      "learning_rate": 8e-05,
      "loss": 1.8709,
      "step": 1557
    },
    {
      "epoch": 0.16837782340862423,
      "grad_norm": 0.37940144538879395,
      "learning_rate": 8e-05,
      "loss": 1.6951,
      "step": 1558
    },
    {
      "epoch": 0.16848589646601103,
      "grad_norm": 0.4132850468158722,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 1559
    },
    {
      "epoch": 0.1685939695233978,
      "grad_norm": 0.40772318840026855,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 1560
    },
    {
      "epoch": 0.1687020425807846,
      "grad_norm": 0.4053058624267578,
      "learning_rate": 8e-05,
      "loss": 1.7826,
      "step": 1561
    },
    {
      "epoch": 0.16881011563817142,
      "grad_norm": 0.39743009209632874,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 1562
    },
    {
      "epoch": 0.1689181886955582,
      "grad_norm": 0.3872872292995453,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 1563
    },
    {
      "epoch": 0.169026261752945,
      "grad_norm": 0.4274870753288269,
      "learning_rate": 8e-05,
      "loss": 1.9388,
      "step": 1564
    },
    {
      "epoch": 0.16913433481033177,
      "grad_norm": 0.41087085008621216,
      "learning_rate": 8e-05,
      "loss": 1.8447,
      "step": 1565
    },
    {
      "epoch": 0.16924240786771858,
      "grad_norm": 0.4330870807170868,
      "learning_rate": 8e-05,
      "loss": 1.9165,
      "step": 1566
    },
    {
      "epoch": 0.16935048092510538,
      "grad_norm": 0.401475191116333,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 1567
    },
    {
      "epoch": 0.16945855398249215,
      "grad_norm": 0.41262686252593994,
      "learning_rate": 8e-05,
      "loss": 1.8916,
      "step": 1568
    },
    {
      "epoch": 0.16956662703987896,
      "grad_norm": 0.38764384388923645,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 1569
    },
    {
      "epoch": 0.16967470009726576,
      "grad_norm": 0.4097070097923279,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 1570
    },
    {
      "epoch": 0.16978277315465254,
      "grad_norm": 0.3946172297000885,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 1571
    },
    {
      "epoch": 0.16989084621203934,
      "grad_norm": 0.3948056995868683,
      "learning_rate": 8e-05,
      "loss": 1.7234,
      "step": 1572
    },
    {
      "epoch": 0.16999891926942612,
      "grad_norm": 0.4473390281200409,
      "learning_rate": 8e-05,
      "loss": 1.9237,
      "step": 1573
    },
    {
      "epoch": 0.17010699232681292,
      "grad_norm": 0.44131919741630554,
      "learning_rate": 8e-05,
      "loss": 1.8131,
      "step": 1574
    },
    {
      "epoch": 0.17021506538419973,
      "grad_norm": 0.41968056559562683,
      "learning_rate": 8e-05,
      "loss": 1.8726,
      "step": 1575
    },
    {
      "epoch": 0.1703231384415865,
      "grad_norm": 0.37494173645973206,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1576
    },
    {
      "epoch": 0.1704312114989733,
      "grad_norm": 0.3846365213394165,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 1577
    },
    {
      "epoch": 0.1705392845563601,
      "grad_norm": 0.3821294605731964,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 1578
    },
    {
      "epoch": 0.1706473576137469,
      "grad_norm": 0.4591141641139984,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 1579
    },
    {
      "epoch": 0.1707554306711337,
      "grad_norm": 0.38429707288742065,
      "learning_rate": 8e-05,
      "loss": 1.742,
      "step": 1580
    },
    {
      "epoch": 0.17086350372852047,
      "grad_norm": 0.40170061588287354,
      "learning_rate": 8e-05,
      "loss": 1.7721,
      "step": 1581
    },
    {
      "epoch": 0.17097157678590727,
      "grad_norm": 0.4049871563911438,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 1582
    },
    {
      "epoch": 0.17107964984329407,
      "grad_norm": 0.43921029567718506,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 1583
    },
    {
      "epoch": 0.17118772290068085,
      "grad_norm": 0.41958072781562805,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 1584
    },
    {
      "epoch": 0.17129579595806765,
      "grad_norm": 0.3943561017513275,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 1585
    },
    {
      "epoch": 0.17140386901545446,
      "grad_norm": 0.39220091700553894,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 1586
    },
    {
      "epoch": 0.17151194207284123,
      "grad_norm": 0.4093630313873291,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 1587
    },
    {
      "epoch": 0.17162001513022804,
      "grad_norm": 0.3826999366283417,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 1588
    },
    {
      "epoch": 0.17172808818761484,
      "grad_norm": 0.42454931139945984,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 1589
    },
    {
      "epoch": 0.17183616124500162,
      "grad_norm": 0.39567896723747253,
      "learning_rate": 8e-05,
      "loss": 1.6761,
      "step": 1590
    },
    {
      "epoch": 0.17194423430238842,
      "grad_norm": 0.4217839539051056,
      "learning_rate": 8e-05,
      "loss": 1.7874,
      "step": 1591
    },
    {
      "epoch": 0.1720523073597752,
      "grad_norm": 0.41408225893974304,
      "learning_rate": 8e-05,
      "loss": 1.8345,
      "step": 1592
    },
    {
      "epoch": 0.172160380417162,
      "grad_norm": 0.39375677704811096,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 1593
    },
    {
      "epoch": 0.1722684534745488,
      "grad_norm": 0.42064109444618225,
      "learning_rate": 8e-05,
      "loss": 1.7923,
      "step": 1594
    },
    {
      "epoch": 0.17237652653193558,
      "grad_norm": 0.39786162972450256,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 1595
    },
    {
      "epoch": 0.17248459958932238,
      "grad_norm": 0.4089573919773102,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1596
    },
    {
      "epoch": 0.1725926726467092,
      "grad_norm": 0.3979000747203827,
      "learning_rate": 8e-05,
      "loss": 1.81,
      "step": 1597
    },
    {
      "epoch": 0.17270074570409596,
      "grad_norm": 0.38609254360198975,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 1598
    },
    {
      "epoch": 0.17280881876148277,
      "grad_norm": 0.41199037432670593,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 1599
    },
    {
      "epoch": 0.17291689181886954,
      "grad_norm": 0.39344334602355957,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 1600
    },
    {
      "epoch": 0.17302496487625635,
      "grad_norm": 0.40112656354904175,
      "learning_rate": 8e-05,
      "loss": 1.9248,
      "step": 1601
    },
    {
      "epoch": 0.17313303793364315,
      "grad_norm": 0.4231383800506592,
      "learning_rate": 8e-05,
      "loss": 1.9891,
      "step": 1602
    },
    {
      "epoch": 0.17324111099102993,
      "grad_norm": 0.39771127700805664,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 1603
    },
    {
      "epoch": 0.17334918404841673,
      "grad_norm": 0.4173750877380371,
      "learning_rate": 8e-05,
      "loss": 1.8647,
      "step": 1604
    },
    {
      "epoch": 0.17345725710580354,
      "grad_norm": 0.421533465385437,
      "learning_rate": 8e-05,
      "loss": 1.9422,
      "step": 1605
    },
    {
      "epoch": 0.1735653301631903,
      "grad_norm": 0.3959817588329315,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 1606
    },
    {
      "epoch": 0.17367340322057712,
      "grad_norm": 0.3841110169887543,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 1607
    },
    {
      "epoch": 0.1737814762779639,
      "grad_norm": 0.42447787523269653,
      "learning_rate": 8e-05,
      "loss": 1.8965,
      "step": 1608
    },
    {
      "epoch": 0.1738895493353507,
      "grad_norm": 0.4233480393886566,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 1609
    },
    {
      "epoch": 0.1739976223927375,
      "grad_norm": 0.38594192266464233,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 1610
    },
    {
      "epoch": 0.17410569545012428,
      "grad_norm": 0.4162929952144623,
      "learning_rate": 8e-05,
      "loss": 1.8578,
      "step": 1611
    },
    {
      "epoch": 0.17421376850751108,
      "grad_norm": 0.411483496427536,
      "learning_rate": 8e-05,
      "loss": 1.7596,
      "step": 1612
    },
    {
      "epoch": 0.17432184156489788,
      "grad_norm": 0.40065616369247437,
      "learning_rate": 8e-05,
      "loss": 1.8135,
      "step": 1613
    },
    {
      "epoch": 0.17442991462228466,
      "grad_norm": 0.41462215781211853,
      "learning_rate": 8e-05,
      "loss": 1.6421,
      "step": 1614
    },
    {
      "epoch": 0.17453798767967146,
      "grad_norm": 0.37826454639434814,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 1615
    },
    {
      "epoch": 0.17464606073705824,
      "grad_norm": 0.39087581634521484,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 1616
    },
    {
      "epoch": 0.17475413379444504,
      "grad_norm": 0.3825644552707672,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 1617
    },
    {
      "epoch": 0.17486220685183185,
      "grad_norm": 0.3963903784751892,
      "learning_rate": 8e-05,
      "loss": 1.9153,
      "step": 1618
    },
    {
      "epoch": 0.17497027990921862,
      "grad_norm": 0.45367926359176636,
      "learning_rate": 8e-05,
      "loss": 1.9203,
      "step": 1619
    },
    {
      "epoch": 0.17507835296660543,
      "grad_norm": 0.4260541796684265,
      "learning_rate": 8e-05,
      "loss": 1.8634,
      "step": 1620
    },
    {
      "epoch": 0.17518642602399223,
      "grad_norm": 0.3804470896720886,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 1621
    },
    {
      "epoch": 0.175294499081379,
      "grad_norm": 0.39194267988204956,
      "learning_rate": 8e-05,
      "loss": 1.8965,
      "step": 1622
    },
    {
      "epoch": 0.1754025721387658,
      "grad_norm": 0.4451465606689453,
      "learning_rate": 8e-05,
      "loss": 1.9663,
      "step": 1623
    },
    {
      "epoch": 0.1755106451961526,
      "grad_norm": 0.4089840054512024,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 1624
    },
    {
      "epoch": 0.1756187182535394,
      "grad_norm": 0.39795318245887756,
      "learning_rate": 8e-05,
      "loss": 1.7506,
      "step": 1625
    },
    {
      "epoch": 0.1757267913109262,
      "grad_norm": 0.4202139377593994,
      "learning_rate": 8e-05,
      "loss": 1.8825,
      "step": 1626
    },
    {
      "epoch": 0.17583486436831297,
      "grad_norm": 0.4123469293117523,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 1627
    },
    {
      "epoch": 0.17594293742569977,
      "grad_norm": 0.3948230445384979,
      "learning_rate": 8e-05,
      "loss": 1.874,
      "step": 1628
    },
    {
      "epoch": 0.17605101048308658,
      "grad_norm": 0.4066675007343292,
      "learning_rate": 8e-05,
      "loss": 1.6634,
      "step": 1629
    },
    {
      "epoch": 0.17615908354047335,
      "grad_norm": 0.40411874651908875,
      "learning_rate": 8e-05,
      "loss": 1.8422,
      "step": 1630
    },
    {
      "epoch": 0.17626715659786016,
      "grad_norm": 0.42432641983032227,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 1631
    },
    {
      "epoch": 0.17637522965524693,
      "grad_norm": 0.41799548268318176,
      "learning_rate": 8e-05,
      "loss": 1.7984,
      "step": 1632
    },
    {
      "epoch": 0.17648330271263374,
      "grad_norm": 0.49466055631637573,
      "learning_rate": 8e-05,
      "loss": 2.0501,
      "step": 1633
    },
    {
      "epoch": 0.17659137577002054,
      "grad_norm": 0.4492991864681244,
      "learning_rate": 8e-05,
      "loss": 1.9428,
      "step": 1634
    },
    {
      "epoch": 0.17669944882740732,
      "grad_norm": 0.4964216947555542,
      "learning_rate": 8e-05,
      "loss": 2.0221,
      "step": 1635
    },
    {
      "epoch": 0.17680752188479412,
      "grad_norm": 0.45653513073921204,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 1636
    },
    {
      "epoch": 0.17691559494218093,
      "grad_norm": 0.4358455538749695,
      "learning_rate": 8e-05,
      "loss": 1.7904,
      "step": 1637
    },
    {
      "epoch": 0.1770236679995677,
      "grad_norm": 0.42530080676078796,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 1638
    },
    {
      "epoch": 0.1771317410569545,
      "grad_norm": 0.41657865047454834,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 1639
    },
    {
      "epoch": 0.17723981411434128,
      "grad_norm": 0.3984321653842926,
      "learning_rate": 8e-05,
      "loss": 1.7898,
      "step": 1640
    },
    {
      "epoch": 0.17734788717172809,
      "grad_norm": 0.3740922212600708,
      "learning_rate": 8e-05,
      "loss": 1.7341,
      "step": 1641
    },
    {
      "epoch": 0.1774559602291149,
      "grad_norm": 0.37876397371292114,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 1642
    },
    {
      "epoch": 0.17756403328650167,
      "grad_norm": 0.4296639859676361,
      "learning_rate": 8e-05,
      "loss": 1.9479,
      "step": 1643
    },
    {
      "epoch": 0.17767210634388847,
      "grad_norm": 0.41031017899513245,
      "learning_rate": 8e-05,
      "loss": 1.8497,
      "step": 1644
    },
    {
      "epoch": 0.17778017940127527,
      "grad_norm": 0.3900047838687897,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 1645
    },
    {
      "epoch": 0.17788825245866205,
      "grad_norm": 0.4054920971393585,
      "learning_rate": 8e-05,
      "loss": 1.8047,
      "step": 1646
    },
    {
      "epoch": 0.17799632551604885,
      "grad_norm": 0.40660908818244934,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 1647
    },
    {
      "epoch": 0.17810439857343563,
      "grad_norm": 0.40105628967285156,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 1648
    },
    {
      "epoch": 0.17821247163082243,
      "grad_norm": 0.3951002061367035,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 1649
    },
    {
      "epoch": 0.17832054468820924,
      "grad_norm": 0.4031952917575836,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 1650
    },
    {
      "epoch": 0.178428617745596,
      "grad_norm": 0.39081084728240967,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 1651
    },
    {
      "epoch": 0.17853669080298282,
      "grad_norm": 0.3917630612850189,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 1652
    },
    {
      "epoch": 0.17864476386036962,
      "grad_norm": 0.3830966353416443,
      "learning_rate": 8e-05,
      "loss": 1.836,
      "step": 1653
    },
    {
      "epoch": 0.1787528369177564,
      "grad_norm": 0.4036678671836853,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 1654
    },
    {
      "epoch": 0.1788609099751432,
      "grad_norm": 0.4197758734226227,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 1655
    },
    {
      "epoch": 0.17896898303252998,
      "grad_norm": 0.39697912335395813,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 1656
    },
    {
      "epoch": 0.17907705608991678,
      "grad_norm": 0.3879580497741699,
      "learning_rate": 8e-05,
      "loss": 1.6357,
      "step": 1657
    },
    {
      "epoch": 0.17918512914730358,
      "grad_norm": 0.41479039192199707,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 1658
    },
    {
      "epoch": 0.17929320220469036,
      "grad_norm": 0.4015040397644043,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 1659
    },
    {
      "epoch": 0.17940127526207716,
      "grad_norm": 0.4239853024482727,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 1660
    },
    {
      "epoch": 0.17950934831946397,
      "grad_norm": 0.38968244194984436,
      "learning_rate": 8e-05,
      "loss": 1.8714,
      "step": 1661
    },
    {
      "epoch": 0.17961742137685074,
      "grad_norm": 0.4354327321052551,
      "learning_rate": 8e-05,
      "loss": 1.9562,
      "step": 1662
    },
    {
      "epoch": 0.17972549443423755,
      "grad_norm": 0.3744049370288849,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 1663
    },
    {
      "epoch": 0.17983356749162432,
      "grad_norm": 0.39872750639915466,
      "learning_rate": 8e-05,
      "loss": 1.6579,
      "step": 1664
    },
    {
      "epoch": 0.17994164054901113,
      "grad_norm": 0.42813172936439514,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 1665
    },
    {
      "epoch": 0.18004971360639793,
      "grad_norm": 0.38545700907707214,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 1666
    },
    {
      "epoch": 0.1801577866637847,
      "grad_norm": 0.4221579134464264,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 1667
    },
    {
      "epoch": 0.1802658597211715,
      "grad_norm": 0.40555936098098755,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 1668
    },
    {
      "epoch": 0.18037393277855832,
      "grad_norm": 0.40024468302726746,
      "learning_rate": 8e-05,
      "loss": 1.7813,
      "step": 1669
    },
    {
      "epoch": 0.1804820058359451,
      "grad_norm": 0.3726314306259155,
      "learning_rate": 8e-05,
      "loss": 1.5782,
      "step": 1670
    },
    {
      "epoch": 0.1805900788933319,
      "grad_norm": 0.40825381875038147,
      "learning_rate": 8e-05,
      "loss": 1.7515,
      "step": 1671
    },
    {
      "epoch": 0.1806981519507187,
      "grad_norm": 0.3919685482978821,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 1672
    },
    {
      "epoch": 0.18080622500810548,
      "grad_norm": 0.40825751423835754,
      "learning_rate": 8e-05,
      "loss": 1.7563,
      "step": 1673
    },
    {
      "epoch": 0.18091429806549228,
      "grad_norm": 0.40762433409690857,
      "learning_rate": 8e-05,
      "loss": 1.7962,
      "step": 1674
    },
    {
      "epoch": 0.18102237112287906,
      "grad_norm": 0.4419548809528351,
      "learning_rate": 8e-05,
      "loss": 1.923,
      "step": 1675
    },
    {
      "epoch": 0.18113044418026586,
      "grad_norm": 0.40403756499290466,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 1676
    },
    {
      "epoch": 0.18123851723765266,
      "grad_norm": 0.4162483513355255,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 1677
    },
    {
      "epoch": 0.18134659029503944,
      "grad_norm": 0.4178827106952667,
      "learning_rate": 8e-05,
      "loss": 1.8297,
      "step": 1678
    },
    {
      "epoch": 0.18145466335242624,
      "grad_norm": 0.42341357469558716,
      "learning_rate": 8e-05,
      "loss": 1.9399,
      "step": 1679
    },
    {
      "epoch": 0.18156273640981305,
      "grad_norm": 0.40514135360717773,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 1680
    },
    {
      "epoch": 0.18167080946719982,
      "grad_norm": 0.39004138112068176,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 1681
    },
    {
      "epoch": 0.18177888252458663,
      "grad_norm": 0.41486695408821106,
      "learning_rate": 8e-05,
      "loss": 1.8338,
      "step": 1682
    },
    {
      "epoch": 0.1818869555819734,
      "grad_norm": 0.40724533796310425,
      "learning_rate": 8e-05,
      "loss": 1.832,
      "step": 1683
    },
    {
      "epoch": 0.1819950286393602,
      "grad_norm": 0.4158163070678711,
      "learning_rate": 8e-05,
      "loss": 1.8503,
      "step": 1684
    },
    {
      "epoch": 0.182103101696747,
      "grad_norm": 0.4342936873435974,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 1685
    },
    {
      "epoch": 0.1822111747541338,
      "grad_norm": 0.4315701127052307,
      "learning_rate": 8e-05,
      "loss": 1.8692,
      "step": 1686
    },
    {
      "epoch": 0.1823192478115206,
      "grad_norm": 0.43985339999198914,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 1687
    },
    {
      "epoch": 0.1824273208689074,
      "grad_norm": 0.39966312050819397,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 1688
    },
    {
      "epoch": 0.18253539392629417,
      "grad_norm": 0.4266275465488434,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 1689
    },
    {
      "epoch": 0.18264346698368097,
      "grad_norm": 0.39274120330810547,
      "learning_rate": 8e-05,
      "loss": 1.7636,
      "step": 1690
    },
    {
      "epoch": 0.18275154004106775,
      "grad_norm": 0.40413227677345276,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 1691
    },
    {
      "epoch": 0.18285961309845455,
      "grad_norm": 0.43808189034461975,
      "learning_rate": 8e-05,
      "loss": 1.924,
      "step": 1692
    },
    {
      "epoch": 0.18296768615584136,
      "grad_norm": 0.3845285177230835,
      "learning_rate": 8e-05,
      "loss": 1.7429,
      "step": 1693
    },
    {
      "epoch": 0.18307575921322813,
      "grad_norm": 0.39391860365867615,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 1694
    },
    {
      "epoch": 0.18318383227061494,
      "grad_norm": 0.44354450702667236,
      "learning_rate": 8e-05,
      "loss": 1.9661,
      "step": 1695
    },
    {
      "epoch": 0.18329190532800174,
      "grad_norm": 0.45251691341400146,
      "learning_rate": 8e-05,
      "loss": 1.9636,
      "step": 1696
    },
    {
      "epoch": 0.18339997838538852,
      "grad_norm": 0.35056108236312866,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 1697
    },
    {
      "epoch": 0.18350805144277532,
      "grad_norm": 0.3910863697528839,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 1698
    },
    {
      "epoch": 0.1836161245001621,
      "grad_norm": 0.41822412610054016,
      "learning_rate": 8e-05,
      "loss": 1.9572,
      "step": 1699
    },
    {
      "epoch": 0.1837241975575489,
      "grad_norm": 0.4434035122394562,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 1700
    },
    {
      "epoch": 0.1838322706149357,
      "grad_norm": 0.4022752046585083,
      "learning_rate": 8e-05,
      "loss": 1.7455,
      "step": 1701
    },
    {
      "epoch": 0.18394034367232248,
      "grad_norm": 0.3968043327331543,
      "learning_rate": 8e-05,
      "loss": 1.8076,
      "step": 1702
    },
    {
      "epoch": 0.18404841672970929,
      "grad_norm": 0.46727845072746277,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 1703
    },
    {
      "epoch": 0.1841564897870961,
      "grad_norm": 0.4306187331676483,
      "learning_rate": 8e-05,
      "loss": 1.8459,
      "step": 1704
    },
    {
      "epoch": 0.18426456284448287,
      "grad_norm": 0.41112759709358215,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 1705
    },
    {
      "epoch": 0.18437263590186967,
      "grad_norm": 0.40800511837005615,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 1706
    },
    {
      "epoch": 0.18448070895925645,
      "grad_norm": 0.3675437867641449,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 1707
    },
    {
      "epoch": 0.18458878201664325,
      "grad_norm": 0.3988402187824249,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 1708
    },
    {
      "epoch": 0.18469685507403005,
      "grad_norm": 0.38849347829818726,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 1709
    },
    {
      "epoch": 0.18480492813141683,
      "grad_norm": 0.4062400162220001,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 1710
    },
    {
      "epoch": 0.18491300118880363,
      "grad_norm": 0.4006682336330414,
      "learning_rate": 8e-05,
      "loss": 1.7956,
      "step": 1711
    },
    {
      "epoch": 0.18502107424619044,
      "grad_norm": 0.4089118540287018,
      "learning_rate": 8e-05,
      "loss": 1.7267,
      "step": 1712
    },
    {
      "epoch": 0.1851291473035772,
      "grad_norm": 0.3867059051990509,
      "learning_rate": 8e-05,
      "loss": 1.7768,
      "step": 1713
    },
    {
      "epoch": 0.18523722036096402,
      "grad_norm": 0.4014013409614563,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 1714
    },
    {
      "epoch": 0.1853452934183508,
      "grad_norm": 0.4211176335811615,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 1715
    },
    {
      "epoch": 0.1854533664757376,
      "grad_norm": 0.4259680509567261,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 1716
    },
    {
      "epoch": 0.1855614395331244,
      "grad_norm": 0.3822132349014282,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 1717
    },
    {
      "epoch": 0.18566951259051118,
      "grad_norm": 0.390841007232666,
      "learning_rate": 8e-05,
      "loss": 1.8454,
      "step": 1718
    },
    {
      "epoch": 0.18577758564789798,
      "grad_norm": 0.39502573013305664,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 1719
    },
    {
      "epoch": 0.18588565870528478,
      "grad_norm": 0.39001142978668213,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 1720
    },
    {
      "epoch": 0.18599373176267156,
      "grad_norm": 0.39064866304397583,
      "learning_rate": 8e-05,
      "loss": 1.7801,
      "step": 1721
    },
    {
      "epoch": 0.18610180482005836,
      "grad_norm": 0.424930214881897,
      "learning_rate": 8e-05,
      "loss": 1.8509,
      "step": 1722
    },
    {
      "epoch": 0.18620987787744514,
      "grad_norm": 0.403729647397995,
      "learning_rate": 8e-05,
      "loss": 1.8065,
      "step": 1723
    },
    {
      "epoch": 0.18631795093483194,
      "grad_norm": 0.4183821678161621,
      "learning_rate": 8e-05,
      "loss": 1.8903,
      "step": 1724
    },
    {
      "epoch": 0.18642602399221875,
      "grad_norm": 0.42360934615135193,
      "learning_rate": 8e-05,
      "loss": 1.831,
      "step": 1725
    },
    {
      "epoch": 0.18653409704960552,
      "grad_norm": 0.45420917868614197,
      "learning_rate": 8e-05,
      "loss": 1.9454,
      "step": 1726
    },
    {
      "epoch": 0.18664217010699233,
      "grad_norm": 0.4610454738140106,
      "learning_rate": 8e-05,
      "loss": 1.7409,
      "step": 1727
    },
    {
      "epoch": 0.18675024316437913,
      "grad_norm": 0.39992135763168335,
      "learning_rate": 8e-05,
      "loss": 1.8105,
      "step": 1728
    },
    {
      "epoch": 0.1868583162217659,
      "grad_norm": 0.3831930458545685,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 1729
    },
    {
      "epoch": 0.1869663892791527,
      "grad_norm": 0.4260097146034241,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 1730
    },
    {
      "epoch": 0.1870744623365395,
      "grad_norm": 0.4500424861907959,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 1731
    },
    {
      "epoch": 0.1871825353939263,
      "grad_norm": 0.40182438492774963,
      "learning_rate": 8e-05,
      "loss": 1.8624,
      "step": 1732
    },
    {
      "epoch": 0.1872906084513131,
      "grad_norm": 0.434728741645813,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 1733
    },
    {
      "epoch": 0.18739868150869987,
      "grad_norm": 0.39172428846359253,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 1734
    },
    {
      "epoch": 0.18750675456608668,
      "grad_norm": 0.4302802085876465,
      "learning_rate": 8e-05,
      "loss": 1.8458,
      "step": 1735
    },
    {
      "epoch": 0.18761482762347348,
      "grad_norm": 0.43964317440986633,
      "learning_rate": 8e-05,
      "loss": 1.9203,
      "step": 1736
    },
    {
      "epoch": 0.18772290068086026,
      "grad_norm": 0.39464548230171204,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 1737
    },
    {
      "epoch": 0.18783097373824706,
      "grad_norm": 0.3988627791404724,
      "learning_rate": 8e-05,
      "loss": 1.7583,
      "step": 1738
    },
    {
      "epoch": 0.18793904679563384,
      "grad_norm": 0.39701953530311584,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 1739
    },
    {
      "epoch": 0.18804711985302064,
      "grad_norm": 0.4275576174259186,
      "learning_rate": 8e-05,
      "loss": 1.8525,
      "step": 1740
    },
    {
      "epoch": 0.18815519291040744,
      "grad_norm": 0.41196730732917786,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 1741
    },
    {
      "epoch": 0.18826326596779422,
      "grad_norm": 0.38394638895988464,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 1742
    },
    {
      "epoch": 0.18837133902518102,
      "grad_norm": 0.38760048151016235,
      "learning_rate": 8e-05,
      "loss": 1.7206,
      "step": 1743
    },
    {
      "epoch": 0.18847941208256783,
      "grad_norm": 0.45226436853408813,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 1744
    },
    {
      "epoch": 0.1885874851399546,
      "grad_norm": 0.4036504626274109,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 1745
    },
    {
      "epoch": 0.1886955581973414,
      "grad_norm": 0.38352468609809875,
      "learning_rate": 8e-05,
      "loss": 1.6651,
      "step": 1746
    },
    {
      "epoch": 0.18880363125472818,
      "grad_norm": 0.3903023600578308,
      "learning_rate": 8e-05,
      "loss": 1.8432,
      "step": 1747
    },
    {
      "epoch": 0.188911704312115,
      "grad_norm": 0.4147639870643616,
      "learning_rate": 8e-05,
      "loss": 1.7717,
      "step": 1748
    },
    {
      "epoch": 0.1890197773695018,
      "grad_norm": 0.4254249632358551,
      "learning_rate": 8e-05,
      "loss": 1.8182,
      "step": 1749
    },
    {
      "epoch": 0.18912785042688857,
      "grad_norm": 0.38011205196380615,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 1750
    },
    {
      "epoch": 0.18923592348427537,
      "grad_norm": 0.3863469064235687,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 1751
    },
    {
      "epoch": 0.18934399654166217,
      "grad_norm": 0.43946200609207153,
      "learning_rate": 8e-05,
      "loss": 1.8622,
      "step": 1752
    },
    {
      "epoch": 0.18945206959904895,
      "grad_norm": 0.41051772236824036,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 1753
    },
    {
      "epoch": 0.18956014265643575,
      "grad_norm": 0.3877480924129486,
      "learning_rate": 8e-05,
      "loss": 1.7273,
      "step": 1754
    },
    {
      "epoch": 0.18966821571382256,
      "grad_norm": 0.3889375627040863,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 1755
    },
    {
      "epoch": 0.18977628877120933,
      "grad_norm": 0.38260290026664734,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 1756
    },
    {
      "epoch": 0.18988436182859614,
      "grad_norm": 0.4558852016925812,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 1757
    },
    {
      "epoch": 0.1899924348859829,
      "grad_norm": 0.3826185464859009,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 1758
    },
    {
      "epoch": 0.19010050794336972,
      "grad_norm": 0.4348844289779663,
      "learning_rate": 8e-05,
      "loss": 1.8526,
      "step": 1759
    },
    {
      "epoch": 0.19020858100075652,
      "grad_norm": 0.37563440203666687,
      "learning_rate": 8e-05,
      "loss": 1.5528,
      "step": 1760
    },
    {
      "epoch": 0.1903166540581433,
      "grad_norm": 0.3701506555080414,
      "learning_rate": 8e-05,
      "loss": 1.4962,
      "step": 1761
    },
    {
      "epoch": 0.1904247271155301,
      "grad_norm": 0.3919129967689514,
      "learning_rate": 8e-05,
      "loss": 1.8852,
      "step": 1762
    },
    {
      "epoch": 0.1905328001729169,
      "grad_norm": 0.39255207777023315,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 1763
    },
    {
      "epoch": 0.19064087323030368,
      "grad_norm": 0.3984370231628418,
      "learning_rate": 8e-05,
      "loss": 1.6902,
      "step": 1764
    },
    {
      "epoch": 0.19074894628769049,
      "grad_norm": 0.3872072100639343,
      "learning_rate": 8e-05,
      "loss": 1.8414,
      "step": 1765
    },
    {
      "epoch": 0.19085701934507726,
      "grad_norm": 0.3949081599712372,
      "learning_rate": 8e-05,
      "loss": 1.8285,
      "step": 1766
    },
    {
      "epoch": 0.19096509240246407,
      "grad_norm": 0.43212148547172546,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 1767
    },
    {
      "epoch": 0.19107316545985087,
      "grad_norm": 0.4102305769920349,
      "learning_rate": 8e-05,
      "loss": 1.8686,
      "step": 1768
    },
    {
      "epoch": 0.19118123851723764,
      "grad_norm": 0.4029206931591034,
      "learning_rate": 8e-05,
      "loss": 1.8696,
      "step": 1769
    },
    {
      "epoch": 0.19128931157462445,
      "grad_norm": 0.4047209620475769,
      "learning_rate": 8e-05,
      "loss": 1.8488,
      "step": 1770
    },
    {
      "epoch": 0.19139738463201125,
      "grad_norm": 0.40032076835632324,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 1771
    },
    {
      "epoch": 0.19150545768939803,
      "grad_norm": 0.38848739862442017,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 1772
    },
    {
      "epoch": 0.19161353074678483,
      "grad_norm": 0.37916311621665955,
      "learning_rate": 8e-05,
      "loss": 1.7168,
      "step": 1773
    },
    {
      "epoch": 0.1917216038041716,
      "grad_norm": 0.3941713869571686,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 1774
    },
    {
      "epoch": 0.1918296768615584,
      "grad_norm": 0.38708704710006714,
      "learning_rate": 8e-05,
      "loss": 1.796,
      "step": 1775
    },
    {
      "epoch": 0.19193774991894522,
      "grad_norm": 0.4021908640861511,
      "learning_rate": 8e-05,
      "loss": 1.8375,
      "step": 1776
    },
    {
      "epoch": 0.192045822976332,
      "grad_norm": 0.42692363262176514,
      "learning_rate": 8e-05,
      "loss": 1.8269,
      "step": 1777
    },
    {
      "epoch": 0.1921538960337188,
      "grad_norm": 0.4367794096469879,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 1778
    },
    {
      "epoch": 0.1922619690911056,
      "grad_norm": 0.381561815738678,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 1779
    },
    {
      "epoch": 0.19237004214849238,
      "grad_norm": 0.4228097200393677,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 1780
    },
    {
      "epoch": 0.19247811520587918,
      "grad_norm": 0.39721131324768066,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 1781
    },
    {
      "epoch": 0.19258618826326596,
      "grad_norm": 0.41058483719825745,
      "learning_rate": 8e-05,
      "loss": 1.8903,
      "step": 1782
    },
    {
      "epoch": 0.19269426132065276,
      "grad_norm": 0.41253435611724854,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 1783
    },
    {
      "epoch": 0.19280233437803956,
      "grad_norm": 0.3929107189178467,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 1784
    },
    {
      "epoch": 0.19291040743542634,
      "grad_norm": 0.4096716642379761,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 1785
    },
    {
      "epoch": 0.19301848049281314,
      "grad_norm": 0.3902013897895813,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 1786
    },
    {
      "epoch": 0.19312655355019995,
      "grad_norm": 0.39768725633621216,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 1787
    },
    {
      "epoch": 0.19323462660758672,
      "grad_norm": 0.459332138299942,
      "learning_rate": 8e-05,
      "loss": 1.9549,
      "step": 1788
    },
    {
      "epoch": 0.19334269966497353,
      "grad_norm": 0.42200252413749695,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 1789
    },
    {
      "epoch": 0.1934507727223603,
      "grad_norm": 0.42415651679039,
      "learning_rate": 8e-05,
      "loss": 1.8386,
      "step": 1790
    },
    {
      "epoch": 0.1935588457797471,
      "grad_norm": 0.4034770429134369,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 1791
    },
    {
      "epoch": 0.1936669188371339,
      "grad_norm": 0.3873521089553833,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 1792
    },
    {
      "epoch": 0.1937749918945207,
      "grad_norm": 0.4399714469909668,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 1793
    },
    {
      "epoch": 0.1938830649519075,
      "grad_norm": 0.5273546576499939,
      "learning_rate": 8e-05,
      "loss": 1.8696,
      "step": 1794
    },
    {
      "epoch": 0.1939911380092943,
      "grad_norm": 0.4082815647125244,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 1795
    },
    {
      "epoch": 0.19409921106668107,
      "grad_norm": 0.37324103713035583,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 1796
    },
    {
      "epoch": 0.19420728412406787,
      "grad_norm": 0.37753596901893616,
      "learning_rate": 8e-05,
      "loss": 1.6706,
      "step": 1797
    },
    {
      "epoch": 0.19431535718145465,
      "grad_norm": 0.4144304692745209,
      "learning_rate": 8e-05,
      "loss": 1.8495,
      "step": 1798
    },
    {
      "epoch": 0.19442343023884145,
      "grad_norm": 0.4201388955116272,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 1799
    },
    {
      "epoch": 0.19453150329622826,
      "grad_norm": 0.42909055948257446,
      "learning_rate": 8e-05,
      "loss": 1.9186,
      "step": 1800
    },
    {
      "epoch": 0.19463957635361503,
      "grad_norm": 0.43580907583236694,
      "learning_rate": 8e-05,
      "loss": 1.8834,
      "step": 1801
    },
    {
      "epoch": 0.19474764941100184,
      "grad_norm": 0.4149489402770996,
      "learning_rate": 8e-05,
      "loss": 1.7259,
      "step": 1802
    },
    {
      "epoch": 0.19485572246838864,
      "grad_norm": 0.4395343065261841,
      "learning_rate": 8e-05,
      "loss": 1.9383,
      "step": 1803
    },
    {
      "epoch": 0.19496379552577542,
      "grad_norm": 0.37827184796333313,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 1804
    },
    {
      "epoch": 0.19507186858316222,
      "grad_norm": 0.38574326038360596,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 1805
    },
    {
      "epoch": 0.195179941640549,
      "grad_norm": 0.38325217366218567,
      "learning_rate": 8e-05,
      "loss": 1.6641,
      "step": 1806
    },
    {
      "epoch": 0.1952880146979358,
      "grad_norm": 0.39836716651916504,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 1807
    },
    {
      "epoch": 0.1953960877553226,
      "grad_norm": 0.4187996983528137,
      "learning_rate": 8e-05,
      "loss": 1.8833,
      "step": 1808
    },
    {
      "epoch": 0.19550416081270938,
      "grad_norm": 0.4088004529476166,
      "learning_rate": 8e-05,
      "loss": 1.8172,
      "step": 1809
    },
    {
      "epoch": 0.19561223387009619,
      "grad_norm": 0.4228382706642151,
      "learning_rate": 8e-05,
      "loss": 1.9816,
      "step": 1810
    },
    {
      "epoch": 0.195720306927483,
      "grad_norm": 0.41184115409851074,
      "learning_rate": 8e-05,
      "loss": 1.988,
      "step": 1811
    },
    {
      "epoch": 0.19582837998486977,
      "grad_norm": 0.38767316937446594,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 1812
    },
    {
      "epoch": 0.19593645304225657,
      "grad_norm": 0.41191694140434265,
      "learning_rate": 8e-05,
      "loss": 1.5037,
      "step": 1813
    },
    {
      "epoch": 0.19604452609964335,
      "grad_norm": 0.36803147196769714,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 1814
    },
    {
      "epoch": 0.19615259915703015,
      "grad_norm": 0.4548899829387665,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 1815
    },
    {
      "epoch": 0.19626067221441695,
      "grad_norm": 0.4324498772621155,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 1816
    },
    {
      "epoch": 0.19636874527180373,
      "grad_norm": 0.4745006561279297,
      "learning_rate": 8e-05,
      "loss": 1.7268,
      "step": 1817
    },
    {
      "epoch": 0.19647681832919053,
      "grad_norm": 0.4554423987865448,
      "learning_rate": 8e-05,
      "loss": 2.0013,
      "step": 1818
    },
    {
      "epoch": 0.19658489138657734,
      "grad_norm": 0.39314043521881104,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 1819
    },
    {
      "epoch": 0.1966929644439641,
      "grad_norm": 0.38300931453704834,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 1820
    },
    {
      "epoch": 0.19680103750135092,
      "grad_norm": 0.40099063515663147,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 1821
    },
    {
      "epoch": 0.1969091105587377,
      "grad_norm": 0.44942736625671387,
      "learning_rate": 8e-05,
      "loss": 2.015,
      "step": 1822
    },
    {
      "epoch": 0.1970171836161245,
      "grad_norm": 0.40747639536857605,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 1823
    },
    {
      "epoch": 0.1971252566735113,
      "grad_norm": 0.43060368299484253,
      "learning_rate": 8e-05,
      "loss": 1.9496,
      "step": 1824
    },
    {
      "epoch": 0.19723332973089808,
      "grad_norm": 0.38507574796676636,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 1825
    },
    {
      "epoch": 0.19734140278828488,
      "grad_norm": 0.3747859597206116,
      "learning_rate": 8e-05,
      "loss": 1.7247,
      "step": 1826
    },
    {
      "epoch": 0.19744947584567168,
      "grad_norm": 0.3978947699069977,
      "learning_rate": 8e-05,
      "loss": 1.8154,
      "step": 1827
    },
    {
      "epoch": 0.19755754890305846,
      "grad_norm": 0.42735981941223145,
      "learning_rate": 8e-05,
      "loss": 1.903,
      "step": 1828
    },
    {
      "epoch": 0.19766562196044526,
      "grad_norm": 0.41382256150245667,
      "learning_rate": 8e-05,
      "loss": 1.7953,
      "step": 1829
    },
    {
      "epoch": 0.19777369501783204,
      "grad_norm": 0.3746395409107208,
      "learning_rate": 8e-05,
      "loss": 1.7092,
      "step": 1830
    },
    {
      "epoch": 0.19788176807521884,
      "grad_norm": 0.3796350657939911,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 1831
    },
    {
      "epoch": 0.19798984113260565,
      "grad_norm": 0.39823511242866516,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1832
    },
    {
      "epoch": 0.19809791418999242,
      "grad_norm": 0.4178105294704437,
      "learning_rate": 8e-05,
      "loss": 1.851,
      "step": 1833
    },
    {
      "epoch": 0.19820598724737923,
      "grad_norm": 0.411088228225708,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 1834
    },
    {
      "epoch": 0.19831406030476603,
      "grad_norm": 0.429667592048645,
      "learning_rate": 8e-05,
      "loss": 1.9989,
      "step": 1835
    },
    {
      "epoch": 0.1984221333621528,
      "grad_norm": 0.3817291557788849,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 1836
    },
    {
      "epoch": 0.1985302064195396,
      "grad_norm": 0.43113991618156433,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 1837
    },
    {
      "epoch": 0.19863827947692642,
      "grad_norm": 0.4241942763328552,
      "learning_rate": 8e-05,
      "loss": 1.876,
      "step": 1838
    },
    {
      "epoch": 0.1987463525343132,
      "grad_norm": 0.39345821738243103,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 1839
    },
    {
      "epoch": 0.1988544255917,
      "grad_norm": 0.40366899967193604,
      "learning_rate": 8e-05,
      "loss": 1.9571,
      "step": 1840
    },
    {
      "epoch": 0.19896249864908677,
      "grad_norm": 0.4471379816532135,
      "learning_rate": 8e-05,
      "loss": 1.8502,
      "step": 1841
    },
    {
      "epoch": 0.19907057170647358,
      "grad_norm": 0.4049915671348572,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 1842
    },
    {
      "epoch": 0.19917864476386038,
      "grad_norm": 0.37762022018432617,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 1843
    },
    {
      "epoch": 0.19928671782124716,
      "grad_norm": 0.4370713233947754,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 1844
    },
    {
      "epoch": 0.19939479087863396,
      "grad_norm": 0.4234843850135803,
      "learning_rate": 8e-05,
      "loss": 1.8695,
      "step": 1845
    },
    {
      "epoch": 0.19950286393602076,
      "grad_norm": 0.4381996691226959,
      "learning_rate": 8e-05,
      "loss": 1.8797,
      "step": 1846
    },
    {
      "epoch": 0.19961093699340754,
      "grad_norm": 0.38635092973709106,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 1847
    },
    {
      "epoch": 0.19971901005079434,
      "grad_norm": 0.41247323155403137,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 1848
    },
    {
      "epoch": 0.19982708310818112,
      "grad_norm": 0.40663942694664,
      "learning_rate": 8e-05,
      "loss": 1.7778,
      "step": 1849
    },
    {
      "epoch": 0.19993515616556792,
      "grad_norm": 0.40080758929252625,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 1850
    },
    {
      "epoch": 0.20004322922295473,
      "grad_norm": 0.4519895017147064,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 1851
    },
    {
      "epoch": 0.2001513022803415,
      "grad_norm": 0.39314255118370056,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 1852
    },
    {
      "epoch": 0.2002593753377283,
      "grad_norm": 0.4010434150695801,
      "learning_rate": 8e-05,
      "loss": 1.8474,
      "step": 1853
    },
    {
      "epoch": 0.2003674483951151,
      "grad_norm": 0.394466370344162,
      "learning_rate": 8e-05,
      "loss": 1.7932,
      "step": 1854
    },
    {
      "epoch": 0.2004755214525019,
      "grad_norm": 0.3911910951137543,
      "learning_rate": 8e-05,
      "loss": 1.7853,
      "step": 1855
    },
    {
      "epoch": 0.2005835945098887,
      "grad_norm": 0.3744804263114929,
      "learning_rate": 8e-05,
      "loss": 1.6609,
      "step": 1856
    },
    {
      "epoch": 0.20069166756727547,
      "grad_norm": 0.4006586968898773,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 1857
    },
    {
      "epoch": 0.20079974062466227,
      "grad_norm": 0.3824820816516876,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 1858
    },
    {
      "epoch": 0.20090781368204907,
      "grad_norm": 0.4241723120212555,
      "learning_rate": 8e-05,
      "loss": 1.9644,
      "step": 1859
    },
    {
      "epoch": 0.20101588673943585,
      "grad_norm": 0.40553411841392517,
      "learning_rate": 8e-05,
      "loss": 1.7589,
      "step": 1860
    },
    {
      "epoch": 0.20112395979682265,
      "grad_norm": 0.3845817744731903,
      "learning_rate": 8e-05,
      "loss": 1.7608,
      "step": 1861
    },
    {
      "epoch": 0.20123203285420946,
      "grad_norm": 0.3958916664123535,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 1862
    },
    {
      "epoch": 0.20134010591159623,
      "grad_norm": 0.4184294044971466,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 1863
    },
    {
      "epoch": 0.20144817896898304,
      "grad_norm": 0.38523563742637634,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 1864
    },
    {
      "epoch": 0.20155625202636981,
      "grad_norm": 0.3942261040210724,
      "learning_rate": 8e-05,
      "loss": 1.7687,
      "step": 1865
    },
    {
      "epoch": 0.20166432508375662,
      "grad_norm": 0.3839040994644165,
      "learning_rate": 8e-05,
      "loss": 1.6311,
      "step": 1866
    },
    {
      "epoch": 0.20177239814114342,
      "grad_norm": 0.4111294150352478,
      "learning_rate": 8e-05,
      "loss": 1.8629,
      "step": 1867
    },
    {
      "epoch": 0.2018804711985302,
      "grad_norm": 0.40295520424842834,
      "learning_rate": 8e-05,
      "loss": 1.9253,
      "step": 1868
    },
    {
      "epoch": 0.201988544255917,
      "grad_norm": 0.43922826647758484,
      "learning_rate": 8e-05,
      "loss": 1.7522,
      "step": 1869
    },
    {
      "epoch": 0.2020966173133038,
      "grad_norm": 0.389254093170166,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 1870
    },
    {
      "epoch": 0.20220469037069058,
      "grad_norm": 0.3941856920719147,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 1871
    },
    {
      "epoch": 0.20231276342807739,
      "grad_norm": 0.4009362459182739,
      "learning_rate": 8e-05,
      "loss": 1.8853,
      "step": 1872
    },
    {
      "epoch": 0.20242083648546416,
      "grad_norm": 0.37995773553848267,
      "learning_rate": 8e-05,
      "loss": 1.7786,
      "step": 1873
    },
    {
      "epoch": 0.20252890954285097,
      "grad_norm": 0.43568530678749084,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 1874
    },
    {
      "epoch": 0.20263698260023777,
      "grad_norm": 0.4306122362613678,
      "learning_rate": 8e-05,
      "loss": 1.8391,
      "step": 1875
    },
    {
      "epoch": 0.20274505565762455,
      "grad_norm": 0.3934189975261688,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 1876
    },
    {
      "epoch": 0.20285312871501135,
      "grad_norm": 0.41781559586524963,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 1877
    },
    {
      "epoch": 0.20296120177239815,
      "grad_norm": 0.4345553517341614,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 1878
    },
    {
      "epoch": 0.20306927482978493,
      "grad_norm": 0.43291541934013367,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 1879
    },
    {
      "epoch": 0.20317734788717173,
      "grad_norm": 0.4012182354927063,
      "learning_rate": 8e-05,
      "loss": 1.8445,
      "step": 1880
    },
    {
      "epoch": 0.2032854209445585,
      "grad_norm": 0.3933817148208618,
      "learning_rate": 8e-05,
      "loss": 1.7455,
      "step": 1881
    },
    {
      "epoch": 0.2033934940019453,
      "grad_norm": 0.41162464022636414,
      "learning_rate": 8e-05,
      "loss": 1.8337,
      "step": 1882
    },
    {
      "epoch": 0.20350156705933212,
      "grad_norm": 0.4029616713523865,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 1883
    },
    {
      "epoch": 0.2036096401167189,
      "grad_norm": 0.43065595626831055,
      "learning_rate": 8e-05,
      "loss": 1.9639,
      "step": 1884
    },
    {
      "epoch": 0.2037177131741057,
      "grad_norm": 0.4122518002986908,
      "learning_rate": 8e-05,
      "loss": 1.8364,
      "step": 1885
    },
    {
      "epoch": 0.2038257862314925,
      "grad_norm": 0.4069730043411255,
      "learning_rate": 8e-05,
      "loss": 1.749,
      "step": 1886
    },
    {
      "epoch": 0.20393385928887928,
      "grad_norm": 0.40174949169158936,
      "learning_rate": 8e-05,
      "loss": 1.8452,
      "step": 1887
    },
    {
      "epoch": 0.20404193234626608,
      "grad_norm": 0.3764544129371643,
      "learning_rate": 8e-05,
      "loss": 1.5908,
      "step": 1888
    },
    {
      "epoch": 0.20415000540365286,
      "grad_norm": 0.4010041058063507,
      "learning_rate": 8e-05,
      "loss": 1.7874,
      "step": 1889
    },
    {
      "epoch": 0.20425807846103966,
      "grad_norm": 0.3901523947715759,
      "learning_rate": 8e-05,
      "loss": 1.7813,
      "step": 1890
    },
    {
      "epoch": 0.20436615151842646,
      "grad_norm": 0.4592389166355133,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 1891
    },
    {
      "epoch": 0.20447422457581324,
      "grad_norm": 0.4369499385356903,
      "learning_rate": 8e-05,
      "loss": 1.8737,
      "step": 1892
    },
    {
      "epoch": 0.20458229763320004,
      "grad_norm": 0.44792553782463074,
      "learning_rate": 8e-05,
      "loss": 1.9751,
      "step": 1893
    },
    {
      "epoch": 0.20469037069058685,
      "grad_norm": 0.44780823588371277,
      "learning_rate": 8e-05,
      "loss": 1.9071,
      "step": 1894
    },
    {
      "epoch": 0.20479844374797362,
      "grad_norm": 0.40773823857307434,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 1895
    },
    {
      "epoch": 0.20490651680536043,
      "grad_norm": 0.3884875774383545,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 1896
    },
    {
      "epoch": 0.2050145898627472,
      "grad_norm": 0.4168465733528137,
      "learning_rate": 8e-05,
      "loss": 1.8818,
      "step": 1897
    },
    {
      "epoch": 0.205122662920134,
      "grad_norm": 0.4159698486328125,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 1898
    },
    {
      "epoch": 0.2052307359775208,
      "grad_norm": 0.39883124828338623,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 1899
    },
    {
      "epoch": 0.2053388090349076,
      "grad_norm": 0.4033334255218506,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 1900
    },
    {
      "epoch": 0.2054468820922944,
      "grad_norm": 0.4412439465522766,
      "learning_rate": 8e-05,
      "loss": 1.7427,
      "step": 1901
    },
    {
      "epoch": 0.2055549551496812,
      "grad_norm": 0.39643287658691406,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 1902
    },
    {
      "epoch": 0.20566302820706797,
      "grad_norm": 0.3855389952659607,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 1903
    },
    {
      "epoch": 0.20577110126445478,
      "grad_norm": 0.4045558571815491,
      "learning_rate": 8e-05,
      "loss": 1.8559,
      "step": 1904
    },
    {
      "epoch": 0.20587917432184155,
      "grad_norm": 0.3895322382450104,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 1905
    },
    {
      "epoch": 0.20598724737922836,
      "grad_norm": 0.400478333234787,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 1906
    },
    {
      "epoch": 0.20609532043661516,
      "grad_norm": 0.3809603750705719,
      "learning_rate": 8e-05,
      "loss": 1.6608,
      "step": 1907
    },
    {
      "epoch": 0.20620339349400194,
      "grad_norm": 0.39927205443382263,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 1908
    },
    {
      "epoch": 0.20631146655138874,
      "grad_norm": 0.408145546913147,
      "learning_rate": 8e-05,
      "loss": 1.8589,
      "step": 1909
    },
    {
      "epoch": 0.20641953960877554,
      "grad_norm": 0.3801189959049225,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 1910
    },
    {
      "epoch": 0.20652761266616232,
      "grad_norm": 0.4106265604496002,
      "learning_rate": 8e-05,
      "loss": 1.9212,
      "step": 1911
    },
    {
      "epoch": 0.20663568572354912,
      "grad_norm": 0.3763452172279358,
      "learning_rate": 8e-05,
      "loss": 1.4875,
      "step": 1912
    },
    {
      "epoch": 0.2067437587809359,
      "grad_norm": 0.37636640667915344,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 1913
    },
    {
      "epoch": 0.2068518318383227,
      "grad_norm": 0.3921518325805664,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 1914
    },
    {
      "epoch": 0.2069599048957095,
      "grad_norm": 0.41918063163757324,
      "learning_rate": 8e-05,
      "loss": 1.8087,
      "step": 1915
    },
    {
      "epoch": 0.20706797795309628,
      "grad_norm": 0.39651623368263245,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 1916
    },
    {
      "epoch": 0.2071760510104831,
      "grad_norm": 0.4217573404312134,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 1917
    },
    {
      "epoch": 0.2072841240678699,
      "grad_norm": 0.4523226320743561,
      "learning_rate": 8e-05,
      "loss": 1.8346,
      "step": 1918
    },
    {
      "epoch": 0.20739219712525667,
      "grad_norm": 0.3777927756309509,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 1919
    },
    {
      "epoch": 0.20750027018264347,
      "grad_norm": 0.43349751830101013,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 1920
    },
    {
      "epoch": 0.20760834324003027,
      "grad_norm": 0.3863975405693054,
      "learning_rate": 8e-05,
      "loss": 1.6674,
      "step": 1921
    },
    {
      "epoch": 0.20771641629741705,
      "grad_norm": 0.41900762915611267,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 1922
    },
    {
      "epoch": 0.20782448935480385,
      "grad_norm": 0.40639373660087585,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 1923
    },
    {
      "epoch": 0.20793256241219063,
      "grad_norm": 0.43557679653167725,
      "learning_rate": 8e-05,
      "loss": 2.0082,
      "step": 1924
    },
    {
      "epoch": 0.20804063546957743,
      "grad_norm": 0.48247018456459045,
      "learning_rate": 8e-05,
      "loss": 1.8391,
      "step": 1925
    },
    {
      "epoch": 0.20814870852696424,
      "grad_norm": 0.3648173213005066,
      "learning_rate": 8e-05,
      "loss": 1.5431,
      "step": 1926
    },
    {
      "epoch": 0.20825678158435101,
      "grad_norm": 0.4107900559902191,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 1927
    },
    {
      "epoch": 0.20836485464173782,
      "grad_norm": 0.4090164601802826,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 1928
    },
    {
      "epoch": 0.20847292769912462,
      "grad_norm": 0.3935917615890503,
      "learning_rate": 8e-05,
      "loss": 1.7686,
      "step": 1929
    },
    {
      "epoch": 0.2085810007565114,
      "grad_norm": 0.441463828086853,
      "learning_rate": 8e-05,
      "loss": 1.8282,
      "step": 1930
    },
    {
      "epoch": 0.2086890738138982,
      "grad_norm": 0.37793996930122375,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 1931
    },
    {
      "epoch": 0.20879714687128498,
      "grad_norm": 0.40165776014328003,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 1932
    },
    {
      "epoch": 0.20890521992867178,
      "grad_norm": 0.4515363872051239,
      "learning_rate": 8e-05,
      "loss": 1.7441,
      "step": 1933
    },
    {
      "epoch": 0.20901329298605859,
      "grad_norm": 0.4473559558391571,
      "learning_rate": 8e-05,
      "loss": 1.8352,
      "step": 1934
    },
    {
      "epoch": 0.20912136604344536,
      "grad_norm": 0.4118614196777344,
      "learning_rate": 8e-05,
      "loss": 1.9185,
      "step": 1935
    },
    {
      "epoch": 0.20922943910083217,
      "grad_norm": 0.40480929613113403,
      "learning_rate": 8e-05,
      "loss": 1.8015,
      "step": 1936
    },
    {
      "epoch": 0.20933751215821897,
      "grad_norm": 0.4183764159679413,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 1937
    },
    {
      "epoch": 0.20944558521560575,
      "grad_norm": 0.4226767420768738,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 1938
    },
    {
      "epoch": 0.20955365827299255,
      "grad_norm": 0.3856644928455353,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 1939
    },
    {
      "epoch": 0.20966173133037933,
      "grad_norm": 0.42725104093551636,
      "learning_rate": 8e-05,
      "loss": 1.8035,
      "step": 1940
    },
    {
      "epoch": 0.20976980438776613,
      "grad_norm": 0.40143096446990967,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 1941
    },
    {
      "epoch": 0.20987787744515293,
      "grad_norm": 0.37382349371910095,
      "learning_rate": 8e-05,
      "loss": 1.5586,
      "step": 1942
    },
    {
      "epoch": 0.2099859505025397,
      "grad_norm": 0.36848628520965576,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 1943
    },
    {
      "epoch": 0.2100940235599265,
      "grad_norm": 0.39335399866104126,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 1944
    },
    {
      "epoch": 0.21020209661731332,
      "grad_norm": 0.44837284088134766,
      "learning_rate": 8e-05,
      "loss": 1.9666,
      "step": 1945
    },
    {
      "epoch": 0.2103101696747001,
      "grad_norm": 0.3967665135860443,
      "learning_rate": 8e-05,
      "loss": 1.6899,
      "step": 1946
    },
    {
      "epoch": 0.2104182427320869,
      "grad_norm": 0.4034571349620819,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 1947
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.45629194378852844,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 1948
    },
    {
      "epoch": 0.21063438884686048,
      "grad_norm": 0.4800003170967102,
      "learning_rate": 8e-05,
      "loss": 1.9834,
      "step": 1949
    },
    {
      "epoch": 0.21074246190424728,
      "grad_norm": 0.4878099262714386,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 1950
    },
    {
      "epoch": 0.21085053496163406,
      "grad_norm": 0.37781500816345215,
      "learning_rate": 8e-05,
      "loss": 1.7892,
      "step": 1951
    },
    {
      "epoch": 0.21095860801902086,
      "grad_norm": 0.450330913066864,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 1952
    },
    {
      "epoch": 0.21106668107640766,
      "grad_norm": 0.4006896913051605,
      "learning_rate": 8e-05,
      "loss": 1.5718,
      "step": 1953
    },
    {
      "epoch": 0.21117475413379444,
      "grad_norm": 0.3953251838684082,
      "learning_rate": 8e-05,
      "loss": 1.7143,
      "step": 1954
    },
    {
      "epoch": 0.21128282719118124,
      "grad_norm": 0.38859155774116516,
      "learning_rate": 8e-05,
      "loss": 1.4811,
      "step": 1955
    },
    {
      "epoch": 0.21139090024856802,
      "grad_norm": 0.38773059844970703,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 1956
    },
    {
      "epoch": 0.21149897330595482,
      "grad_norm": 0.43714427947998047,
      "learning_rate": 8e-05,
      "loss": 1.8427,
      "step": 1957
    },
    {
      "epoch": 0.21160704636334163,
      "grad_norm": 0.3862868845462799,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 1958
    },
    {
      "epoch": 0.2117151194207284,
      "grad_norm": 0.4144759476184845,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 1959
    },
    {
      "epoch": 0.2118231924781152,
      "grad_norm": 0.44727823138237,
      "learning_rate": 8e-05,
      "loss": 1.7786,
      "step": 1960
    },
    {
      "epoch": 0.211931265535502,
      "grad_norm": 0.4221794605255127,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 1961
    },
    {
      "epoch": 0.2120393385928888,
      "grad_norm": 0.4277529716491699,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 1962
    },
    {
      "epoch": 0.2121474116502756,
      "grad_norm": 0.44442683458328247,
      "learning_rate": 8e-05,
      "loss": 1.8777,
      "step": 1963
    },
    {
      "epoch": 0.21225548470766237,
      "grad_norm": 0.4745279550552368,
      "learning_rate": 8e-05,
      "loss": 1.8766,
      "step": 1964
    },
    {
      "epoch": 0.21236355776504917,
      "grad_norm": 0.39079684019088745,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1965
    },
    {
      "epoch": 0.21247163082243598,
      "grad_norm": 0.4045807123184204,
      "learning_rate": 8e-05,
      "loss": 1.7471,
      "step": 1966
    },
    {
      "epoch": 0.21257970387982275,
      "grad_norm": 0.4092879891395569,
      "learning_rate": 8e-05,
      "loss": 1.9372,
      "step": 1967
    },
    {
      "epoch": 0.21268777693720956,
      "grad_norm": 0.44031816720962524,
      "learning_rate": 8e-05,
      "loss": 1.8662,
      "step": 1968
    },
    {
      "epoch": 0.21279584999459636,
      "grad_norm": 0.4324405789375305,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 1969
    },
    {
      "epoch": 0.21290392305198313,
      "grad_norm": 0.4031631350517273,
      "learning_rate": 8e-05,
      "loss": 1.8373,
      "step": 1970
    },
    {
      "epoch": 0.21301199610936994,
      "grad_norm": 0.42081740498542786,
      "learning_rate": 8e-05,
      "loss": 1.7494,
      "step": 1971
    },
    {
      "epoch": 0.21312006916675671,
      "grad_norm": 0.3989444673061371,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 1972
    },
    {
      "epoch": 0.21322814222414352,
      "grad_norm": 0.4398438036441803,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 1973
    },
    {
      "epoch": 0.21333621528153032,
      "grad_norm": 0.3941102623939514,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1974
    },
    {
      "epoch": 0.2134442883389171,
      "grad_norm": 0.41559019684791565,
      "learning_rate": 8e-05,
      "loss": 1.766,
      "step": 1975
    },
    {
      "epoch": 0.2135523613963039,
      "grad_norm": 0.407122403383255,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 1976
    },
    {
      "epoch": 0.2136604344536907,
      "grad_norm": 0.43038615584373474,
      "learning_rate": 8e-05,
      "loss": 1.9142,
      "step": 1977
    },
    {
      "epoch": 0.21376850751107748,
      "grad_norm": 0.3964691758155823,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 1978
    },
    {
      "epoch": 0.2138765805684643,
      "grad_norm": 0.41280561685562134,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 1979
    },
    {
      "epoch": 0.21398465362585106,
      "grad_norm": 0.45882347226142883,
      "learning_rate": 8e-05,
      "loss": 1.8673,
      "step": 1980
    },
    {
      "epoch": 0.21409272668323787,
      "grad_norm": 0.4497271478176117,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 1981
    },
    {
      "epoch": 0.21420079974062467,
      "grad_norm": 0.40378624200820923,
      "learning_rate": 8e-05,
      "loss": 1.7925,
      "step": 1982
    },
    {
      "epoch": 0.21430887279801145,
      "grad_norm": 0.41565781831741333,
      "learning_rate": 8e-05,
      "loss": 1.8378,
      "step": 1983
    },
    {
      "epoch": 0.21441694585539825,
      "grad_norm": 0.3891652822494507,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 1984
    },
    {
      "epoch": 0.21452501891278505,
      "grad_norm": 0.40945783257484436,
      "learning_rate": 8e-05,
      "loss": 1.9026,
      "step": 1985
    },
    {
      "epoch": 0.21463309197017183,
      "grad_norm": 0.38832518458366394,
      "learning_rate": 8e-05,
      "loss": 1.7932,
      "step": 1986
    },
    {
      "epoch": 0.21474116502755863,
      "grad_norm": 0.3966149091720581,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 1987
    },
    {
      "epoch": 0.2148492380849454,
      "grad_norm": 0.4058310091495514,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 1988
    },
    {
      "epoch": 0.2149573111423322,
      "grad_norm": 0.39273446798324585,
      "learning_rate": 8e-05,
      "loss": 1.721,
      "step": 1989
    },
    {
      "epoch": 0.21506538419971902,
      "grad_norm": 0.4174811542034149,
      "learning_rate": 8e-05,
      "loss": 1.7548,
      "step": 1990
    },
    {
      "epoch": 0.2151734572571058,
      "grad_norm": 0.4127671420574188,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 1991
    },
    {
      "epoch": 0.2152815303144926,
      "grad_norm": 0.41961410641670227,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 1992
    },
    {
      "epoch": 0.2153896033718794,
      "grad_norm": 0.4488968849182129,
      "learning_rate": 8e-05,
      "loss": 1.8144,
      "step": 1993
    },
    {
      "epoch": 0.21549767642926618,
      "grad_norm": 0.47075217962265015,
      "learning_rate": 8e-05,
      "loss": 1.9579,
      "step": 1994
    },
    {
      "epoch": 0.21560574948665298,
      "grad_norm": 0.39560678601264954,
      "learning_rate": 8e-05,
      "loss": 1.8373,
      "step": 1995
    },
    {
      "epoch": 0.21571382254403976,
      "grad_norm": 0.4414810240268707,
      "learning_rate": 8e-05,
      "loss": 1.8861,
      "step": 1996
    },
    {
      "epoch": 0.21582189560142656,
      "grad_norm": 0.4048781991004944,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 1997
    },
    {
      "epoch": 0.21592996865881336,
      "grad_norm": 0.399124413728714,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 1998
    },
    {
      "epoch": 0.21603804171620014,
      "grad_norm": 0.41796571016311646,
      "learning_rate": 8e-05,
      "loss": 1.8494,
      "step": 1999
    },
    {
      "epoch": 0.21614611477358694,
      "grad_norm": 0.4011225700378418,
      "learning_rate": 8e-05,
      "loss": 1.7377,
      "step": 2000
    },
    {
      "epoch": 0.21625418783097375,
      "grad_norm": 0.4368759095668793,
      "learning_rate": 8e-05,
      "loss": 1.8163,
      "step": 2001
    },
    {
      "epoch": 0.21636226088836052,
      "grad_norm": 0.41631463170051575,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 2002
    },
    {
      "epoch": 0.21647033394574733,
      "grad_norm": 0.39685797691345215,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 2003
    },
    {
      "epoch": 0.21657840700313413,
      "grad_norm": 0.4026358425617218,
      "learning_rate": 8e-05,
      "loss": 1.5878,
      "step": 2004
    },
    {
      "epoch": 0.2166864800605209,
      "grad_norm": 0.3996177911758423,
      "learning_rate": 8e-05,
      "loss": 1.6381,
      "step": 2005
    },
    {
      "epoch": 0.2167945531179077,
      "grad_norm": 0.42654404044151306,
      "learning_rate": 8e-05,
      "loss": 1.8624,
      "step": 2006
    },
    {
      "epoch": 0.2169026261752945,
      "grad_norm": 0.39209839701652527,
      "learning_rate": 8e-05,
      "loss": 1.7602,
      "step": 2007
    },
    {
      "epoch": 0.2170106992326813,
      "grad_norm": 0.4005126655101776,
      "learning_rate": 8e-05,
      "loss": 1.6504,
      "step": 2008
    },
    {
      "epoch": 0.2171187722900681,
      "grad_norm": 0.4537847936153412,
      "learning_rate": 8e-05,
      "loss": 1.7836,
      "step": 2009
    },
    {
      "epoch": 0.21722684534745487,
      "grad_norm": 0.4858954846858978,
      "learning_rate": 8e-05,
      "loss": 1.8003,
      "step": 2010
    },
    {
      "epoch": 0.21733491840484168,
      "grad_norm": 0.4000581204891205,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 2011
    },
    {
      "epoch": 0.21744299146222848,
      "grad_norm": 0.4298788905143738,
      "learning_rate": 8e-05,
      "loss": 1.8021,
      "step": 2012
    },
    {
      "epoch": 0.21755106451961526,
      "grad_norm": 0.4221879243850708,
      "learning_rate": 8e-05,
      "loss": 1.6742,
      "step": 2013
    },
    {
      "epoch": 0.21765913757700206,
      "grad_norm": 0.38335832953453064,
      "learning_rate": 8e-05,
      "loss": 1.4721,
      "step": 2014
    },
    {
      "epoch": 0.21776721063438884,
      "grad_norm": 0.4195748567581177,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 2015
    },
    {
      "epoch": 0.21787528369177564,
      "grad_norm": 0.4003381133079529,
      "learning_rate": 8e-05,
      "loss": 1.6965,
      "step": 2016
    },
    {
      "epoch": 0.21798335674916244,
      "grad_norm": 0.4233470559120178,
      "learning_rate": 8e-05,
      "loss": 1.7136,
      "step": 2017
    },
    {
      "epoch": 0.21809142980654922,
      "grad_norm": 0.44031623005867004,
      "learning_rate": 8e-05,
      "loss": 1.8877,
      "step": 2018
    },
    {
      "epoch": 0.21819950286393602,
      "grad_norm": 0.41461220383644104,
      "learning_rate": 8e-05,
      "loss": 1.7547,
      "step": 2019
    },
    {
      "epoch": 0.21830757592132283,
      "grad_norm": 0.41144871711730957,
      "learning_rate": 8e-05,
      "loss": 1.7747,
      "step": 2020
    },
    {
      "epoch": 0.2184156489787096,
      "grad_norm": 0.4442959427833557,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 2021
    },
    {
      "epoch": 0.2185237220360964,
      "grad_norm": 0.4150540828704834,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 2022
    },
    {
      "epoch": 0.21863179509348318,
      "grad_norm": 0.4000658094882965,
      "learning_rate": 8e-05,
      "loss": 1.6812,
      "step": 2023
    },
    {
      "epoch": 0.21873986815087,
      "grad_norm": 0.41589972376823425,
      "learning_rate": 8e-05,
      "loss": 1.7601,
      "step": 2024
    },
    {
      "epoch": 0.2188479412082568,
      "grad_norm": 0.39739900827407837,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 2025
    },
    {
      "epoch": 0.21895601426564357,
      "grad_norm": 0.39198148250579834,
      "learning_rate": 8e-05,
      "loss": 1.8171,
      "step": 2026
    },
    {
      "epoch": 0.21906408732303037,
      "grad_norm": 0.4336865246295929,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 2027
    },
    {
      "epoch": 0.21917216038041717,
      "grad_norm": 0.43984130024909973,
      "learning_rate": 8e-05,
      "loss": 1.8055,
      "step": 2028
    },
    {
      "epoch": 0.21928023343780395,
      "grad_norm": 0.3988138437271118,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 2029
    },
    {
      "epoch": 0.21938830649519075,
      "grad_norm": 0.4244089722633362,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 2030
    },
    {
      "epoch": 0.21949637955257753,
      "grad_norm": 0.42099395394325256,
      "learning_rate": 8e-05,
      "loss": 1.8515,
      "step": 2031
    },
    {
      "epoch": 0.21960445260996433,
      "grad_norm": 0.3779638707637787,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 2032
    },
    {
      "epoch": 0.21971252566735114,
      "grad_norm": 0.40045905113220215,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 2033
    },
    {
      "epoch": 0.21982059872473791,
      "grad_norm": 0.4167577922344208,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 2034
    },
    {
      "epoch": 0.21992867178212472,
      "grad_norm": 0.40886619687080383,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 2035
    },
    {
      "epoch": 0.22003674483951152,
      "grad_norm": 0.3812326490879059,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 2036
    },
    {
      "epoch": 0.2201448178968983,
      "grad_norm": 0.41522809863090515,
      "learning_rate": 8e-05,
      "loss": 1.8755,
      "step": 2037
    },
    {
      "epoch": 0.2202528909542851,
      "grad_norm": 0.3953542411327362,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 2038
    },
    {
      "epoch": 0.22036096401167188,
      "grad_norm": 0.3817477822303772,
      "learning_rate": 8e-05,
      "loss": 1.589,
      "step": 2039
    },
    {
      "epoch": 0.22046903706905868,
      "grad_norm": 0.4113268554210663,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 2040
    },
    {
      "epoch": 0.22057711012644549,
      "grad_norm": 0.4051961898803711,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 2041
    },
    {
      "epoch": 0.22068518318383226,
      "grad_norm": 0.4062207341194153,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 2042
    },
    {
      "epoch": 0.22079325624121907,
      "grad_norm": 0.4326634407043457,
      "learning_rate": 8e-05,
      "loss": 1.7764,
      "step": 2043
    },
    {
      "epoch": 0.22090132929860587,
      "grad_norm": 0.3997977375984192,
      "learning_rate": 8e-05,
      "loss": 1.8325,
      "step": 2044
    },
    {
      "epoch": 0.22100940235599265,
      "grad_norm": 0.4219909608364105,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 2045
    },
    {
      "epoch": 0.22111747541337945,
      "grad_norm": 0.40244346857070923,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 2046
    },
    {
      "epoch": 0.22122554847076623,
      "grad_norm": 0.43411484360694885,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 2047
    },
    {
      "epoch": 0.22133362152815303,
      "grad_norm": 0.38347703218460083,
      "learning_rate": 8e-05,
      "loss": 1.6935,
      "step": 2048
    },
    {
      "epoch": 0.22144169458553983,
      "grad_norm": 0.3826943337917328,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 2049
    },
    {
      "epoch": 0.2215497676429266,
      "grad_norm": 0.4427852928638458,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 2050
    },
    {
      "epoch": 0.2216578407003134,
      "grad_norm": 0.43519556522369385,
      "learning_rate": 8e-05,
      "loss": 1.9327,
      "step": 2051
    },
    {
      "epoch": 0.22176591375770022,
      "grad_norm": 0.39829114079475403,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 2052
    },
    {
      "epoch": 0.221873986815087,
      "grad_norm": 0.435432493686676,
      "learning_rate": 8e-05,
      "loss": 1.846,
      "step": 2053
    },
    {
      "epoch": 0.2219820598724738,
      "grad_norm": 0.4441291093826294,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 2054
    },
    {
      "epoch": 0.22209013292986057,
      "grad_norm": 0.41369879245758057,
      "learning_rate": 8e-05,
      "loss": 1.7074,
      "step": 2055
    },
    {
      "epoch": 0.22219820598724738,
      "grad_norm": 0.41253504157066345,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 2056
    },
    {
      "epoch": 0.22230627904463418,
      "grad_norm": 0.42553967237472534,
      "learning_rate": 8e-05,
      "loss": 1.9426,
      "step": 2057
    },
    {
      "epoch": 0.22241435210202096,
      "grad_norm": 0.4109209477901459,
      "learning_rate": 8e-05,
      "loss": 1.8384,
      "step": 2058
    },
    {
      "epoch": 0.22252242515940776,
      "grad_norm": 0.4235013425350189,
      "learning_rate": 8e-05,
      "loss": 1.8582,
      "step": 2059
    },
    {
      "epoch": 0.22263049821679456,
      "grad_norm": 0.414936363697052,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 2060
    },
    {
      "epoch": 0.22273857127418134,
      "grad_norm": 0.40648406744003296,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 2061
    },
    {
      "epoch": 0.22284664433156814,
      "grad_norm": 0.4109076261520386,
      "learning_rate": 8e-05,
      "loss": 1.9404,
      "step": 2062
    },
    {
      "epoch": 0.22295471738895492,
      "grad_norm": 0.3946150839328766,
      "learning_rate": 8e-05,
      "loss": 1.7329,
      "step": 2063
    },
    {
      "epoch": 0.22306279044634172,
      "grad_norm": 0.40817445516586304,
      "learning_rate": 8e-05,
      "loss": 1.8245,
      "step": 2064
    },
    {
      "epoch": 0.22317086350372853,
      "grad_norm": 0.41996556520462036,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 2065
    },
    {
      "epoch": 0.2232789365611153,
      "grad_norm": 0.3988870680332184,
      "learning_rate": 8e-05,
      "loss": 1.7138,
      "step": 2066
    },
    {
      "epoch": 0.2233870096185021,
      "grad_norm": 0.4383637607097626,
      "learning_rate": 8e-05,
      "loss": 1.8725,
      "step": 2067
    },
    {
      "epoch": 0.2234950826758889,
      "grad_norm": 0.4138929545879364,
      "learning_rate": 8e-05,
      "loss": 1.9078,
      "step": 2068
    },
    {
      "epoch": 0.2236031557332757,
      "grad_norm": 0.4102814197540283,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 2069
    },
    {
      "epoch": 0.2237112287906625,
      "grad_norm": 0.38822874426841736,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 2070
    },
    {
      "epoch": 0.22381930184804927,
      "grad_norm": 0.4457188844680786,
      "learning_rate": 8e-05,
      "loss": 1.8449,
      "step": 2071
    },
    {
      "epoch": 0.22392737490543607,
      "grad_norm": 0.40503206849098206,
      "learning_rate": 8e-05,
      "loss": 1.82,
      "step": 2072
    },
    {
      "epoch": 0.22403544796282288,
      "grad_norm": 0.4034186601638794,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 2073
    },
    {
      "epoch": 0.22414352102020965,
      "grad_norm": 0.3628835380077362,
      "learning_rate": 8e-05,
      "loss": 1.6766,
      "step": 2074
    },
    {
      "epoch": 0.22425159407759646,
      "grad_norm": 0.42108389735221863,
      "learning_rate": 8e-05,
      "loss": 1.8606,
      "step": 2075
    },
    {
      "epoch": 0.22435966713498326,
      "grad_norm": 0.38381531834602356,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 2076
    },
    {
      "epoch": 0.22446774019237004,
      "grad_norm": 0.42808178067207336,
      "learning_rate": 8e-05,
      "loss": 1.818,
      "step": 2077
    },
    {
      "epoch": 0.22457581324975684,
      "grad_norm": 0.39599108695983887,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 2078
    },
    {
      "epoch": 0.22468388630714362,
      "grad_norm": 0.4346920847892761,
      "learning_rate": 8e-05,
      "loss": 1.8304,
      "step": 2079
    },
    {
      "epoch": 0.22479195936453042,
      "grad_norm": 0.3833211362361908,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 2080
    },
    {
      "epoch": 0.22490003242191722,
      "grad_norm": 0.40808841586112976,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 2081
    },
    {
      "epoch": 0.225008105479304,
      "grad_norm": 0.4266483187675476,
      "learning_rate": 8e-05,
      "loss": 1.7537,
      "step": 2082
    },
    {
      "epoch": 0.2251161785366908,
      "grad_norm": 0.44365939497947693,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 2083
    },
    {
      "epoch": 0.2252242515940776,
      "grad_norm": 0.41822969913482666,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 2084
    },
    {
      "epoch": 0.22533232465146438,
      "grad_norm": 0.43157148361206055,
      "learning_rate": 8e-05,
      "loss": 1.8657,
      "step": 2085
    },
    {
      "epoch": 0.2254403977088512,
      "grad_norm": 0.39073726534843445,
      "learning_rate": 8e-05,
      "loss": 1.8184,
      "step": 2086
    },
    {
      "epoch": 0.225548470766238,
      "grad_norm": 0.4464273452758789,
      "learning_rate": 8e-05,
      "loss": 1.8458,
      "step": 2087
    },
    {
      "epoch": 0.22565654382362477,
      "grad_norm": 0.4644491374492645,
      "learning_rate": 8e-05,
      "loss": 1.6314,
      "step": 2088
    },
    {
      "epoch": 0.22576461688101157,
      "grad_norm": 0.47318634390830994,
      "learning_rate": 8e-05,
      "loss": 2.0921,
      "step": 2089
    },
    {
      "epoch": 0.22587268993839835,
      "grad_norm": 0.411937415599823,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 2090
    },
    {
      "epoch": 0.22598076299578515,
      "grad_norm": 0.3743872046470642,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 2091
    },
    {
      "epoch": 0.22608883605317195,
      "grad_norm": 0.4300547242164612,
      "learning_rate": 8e-05,
      "loss": 1.919,
      "step": 2092
    },
    {
      "epoch": 0.22619690911055873,
      "grad_norm": 0.40563881397247314,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 2093
    },
    {
      "epoch": 0.22630498216794553,
      "grad_norm": 0.4179266691207886,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 2094
    },
    {
      "epoch": 0.22641305522533234,
      "grad_norm": 0.4229338467121124,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 2095
    },
    {
      "epoch": 0.22652112828271911,
      "grad_norm": 0.40138331055641174,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 2096
    },
    {
      "epoch": 0.22662920134010592,
      "grad_norm": 0.49631330370903015,
      "learning_rate": 8e-05,
      "loss": 2.0145,
      "step": 2097
    },
    {
      "epoch": 0.2267372743974927,
      "grad_norm": 0.4203665256500244,
      "learning_rate": 8e-05,
      "loss": 1.9017,
      "step": 2098
    },
    {
      "epoch": 0.2268453474548795,
      "grad_norm": 0.4118797183036804,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 2099
    },
    {
      "epoch": 0.2269534205122663,
      "grad_norm": 0.3954211175441742,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 2100
    },
    {
      "epoch": 0.22706149356965308,
      "grad_norm": 0.4308033585548401,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 2101
    },
    {
      "epoch": 0.22716956662703988,
      "grad_norm": 0.37827587127685547,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 2102
    },
    {
      "epoch": 0.22727763968442669,
      "grad_norm": 0.43319177627563477,
      "learning_rate": 8e-05,
      "loss": 1.8362,
      "step": 2103
    },
    {
      "epoch": 0.22738571274181346,
      "grad_norm": 0.40607067942619324,
      "learning_rate": 8e-05,
      "loss": 1.7249,
      "step": 2104
    },
    {
      "epoch": 0.22749378579920027,
      "grad_norm": 0.3948572278022766,
      "learning_rate": 8e-05,
      "loss": 1.8566,
      "step": 2105
    },
    {
      "epoch": 0.22760185885658704,
      "grad_norm": 0.4113921821117401,
      "learning_rate": 8e-05,
      "loss": 1.7355,
      "step": 2106
    },
    {
      "epoch": 0.22770993191397385,
      "grad_norm": 0.44707611203193665,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 2107
    },
    {
      "epoch": 0.22781800497136065,
      "grad_norm": 0.4125959277153015,
      "learning_rate": 8e-05,
      "loss": 1.8627,
      "step": 2108
    },
    {
      "epoch": 0.22792607802874743,
      "grad_norm": 0.4361126720905304,
      "learning_rate": 8e-05,
      "loss": 1.8158,
      "step": 2109
    },
    {
      "epoch": 0.22803415108613423,
      "grad_norm": 0.3867851793766022,
      "learning_rate": 8e-05,
      "loss": 1.7563,
      "step": 2110
    },
    {
      "epoch": 0.22814222414352103,
      "grad_norm": 0.4616968035697937,
      "learning_rate": 8e-05,
      "loss": 1.7571,
      "step": 2111
    },
    {
      "epoch": 0.2282502972009078,
      "grad_norm": 0.40942755341529846,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 2112
    },
    {
      "epoch": 0.2283583702582946,
      "grad_norm": 0.4035147428512573,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 2113
    },
    {
      "epoch": 0.2284664433156814,
      "grad_norm": 0.3799296021461487,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 2114
    },
    {
      "epoch": 0.2285745163730682,
      "grad_norm": 0.38475579023361206,
      "learning_rate": 8e-05,
      "loss": 1.7206,
      "step": 2115
    },
    {
      "epoch": 0.228682589430455,
      "grad_norm": 0.4855664372444153,
      "learning_rate": 8e-05,
      "loss": 1.8709,
      "step": 2116
    },
    {
      "epoch": 0.22879066248784177,
      "grad_norm": 0.3974844217300415,
      "learning_rate": 8e-05,
      "loss": 1.7507,
      "step": 2117
    },
    {
      "epoch": 0.22889873554522858,
      "grad_norm": 0.41040197014808655,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 2118
    },
    {
      "epoch": 0.22900680860261538,
      "grad_norm": 0.3944547176361084,
      "learning_rate": 8e-05,
      "loss": 1.7379,
      "step": 2119
    },
    {
      "epoch": 0.22911488166000216,
      "grad_norm": 0.39876946806907654,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 2120
    },
    {
      "epoch": 0.22922295471738896,
      "grad_norm": 0.4057289659976959,
      "learning_rate": 8e-05,
      "loss": 1.8239,
      "step": 2121
    },
    {
      "epoch": 0.22933102777477574,
      "grad_norm": 0.41026437282562256,
      "learning_rate": 8e-05,
      "loss": 1.8182,
      "step": 2122
    },
    {
      "epoch": 0.22943910083216254,
      "grad_norm": 0.41718176007270813,
      "learning_rate": 8e-05,
      "loss": 1.7591,
      "step": 2123
    },
    {
      "epoch": 0.22954717388954934,
      "grad_norm": 0.3898949921131134,
      "learning_rate": 8e-05,
      "loss": 1.8004,
      "step": 2124
    },
    {
      "epoch": 0.22965524694693612,
      "grad_norm": 0.40688979625701904,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 2125
    },
    {
      "epoch": 0.22976332000432292,
      "grad_norm": 0.404819518327713,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 2126
    },
    {
      "epoch": 0.22987139306170973,
      "grad_norm": 0.40449798107147217,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 2127
    },
    {
      "epoch": 0.2299794661190965,
      "grad_norm": 0.42569756507873535,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 2128
    },
    {
      "epoch": 0.2300875391764833,
      "grad_norm": 0.4155648648738861,
      "learning_rate": 8e-05,
      "loss": 1.9333,
      "step": 2129
    },
    {
      "epoch": 0.23019561223387008,
      "grad_norm": 0.3870641887187958,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 2130
    },
    {
      "epoch": 0.2303036852912569,
      "grad_norm": 0.43959441781044006,
      "learning_rate": 8e-05,
      "loss": 1.9309,
      "step": 2131
    },
    {
      "epoch": 0.2304117583486437,
      "grad_norm": 0.407684326171875,
      "learning_rate": 8e-05,
      "loss": 1.7463,
      "step": 2132
    },
    {
      "epoch": 0.23051983140603047,
      "grad_norm": 0.4135707914829254,
      "learning_rate": 8e-05,
      "loss": 1.9085,
      "step": 2133
    },
    {
      "epoch": 0.23062790446341727,
      "grad_norm": 0.4095306098461151,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 2134
    },
    {
      "epoch": 0.23073597752080408,
      "grad_norm": 0.43231603503227234,
      "learning_rate": 8e-05,
      "loss": 1.6056,
      "step": 2135
    },
    {
      "epoch": 0.23084405057819085,
      "grad_norm": 0.41548362374305725,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 2136
    },
    {
      "epoch": 0.23095212363557766,
      "grad_norm": 0.4058496952056885,
      "learning_rate": 8e-05,
      "loss": 1.789,
      "step": 2137
    },
    {
      "epoch": 0.23106019669296443,
      "grad_norm": 0.4908026158809662,
      "learning_rate": 8e-05,
      "loss": 1.9144,
      "step": 2138
    },
    {
      "epoch": 0.23116826975035124,
      "grad_norm": 0.42164385318756104,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 2139
    },
    {
      "epoch": 0.23127634280773804,
      "grad_norm": 0.3719654083251953,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 2140
    },
    {
      "epoch": 0.23138441586512482,
      "grad_norm": 0.39414840936660767,
      "learning_rate": 8e-05,
      "loss": 1.7985,
      "step": 2141
    },
    {
      "epoch": 0.23149248892251162,
      "grad_norm": 0.3884771168231964,
      "learning_rate": 8e-05,
      "loss": 1.6008,
      "step": 2142
    },
    {
      "epoch": 0.23160056197989842,
      "grad_norm": 0.40360984206199646,
      "learning_rate": 8e-05,
      "loss": 1.5755,
      "step": 2143
    },
    {
      "epoch": 0.2317086350372852,
      "grad_norm": 0.4341905415058136,
      "learning_rate": 8e-05,
      "loss": 1.9465,
      "step": 2144
    },
    {
      "epoch": 0.231816708094672,
      "grad_norm": 0.5481971502304077,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 2145
    },
    {
      "epoch": 0.23192478115205878,
      "grad_norm": 0.42595043778419495,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 2146
    },
    {
      "epoch": 0.23203285420944558,
      "grad_norm": 0.4164745807647705,
      "learning_rate": 8e-05,
      "loss": 1.81,
      "step": 2147
    },
    {
      "epoch": 0.2321409272668324,
      "grad_norm": 0.4031875729560852,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 2148
    },
    {
      "epoch": 0.23224900032421916,
      "grad_norm": 0.42535606026649475,
      "learning_rate": 8e-05,
      "loss": 1.8092,
      "step": 2149
    },
    {
      "epoch": 0.23235707338160597,
      "grad_norm": 0.4086892902851105,
      "learning_rate": 8e-05,
      "loss": 1.8256,
      "step": 2150
    },
    {
      "epoch": 0.23246514643899277,
      "grad_norm": 0.39617395401000977,
      "learning_rate": 8e-05,
      "loss": 1.6094,
      "step": 2151
    },
    {
      "epoch": 0.23257321949637955,
      "grad_norm": 0.4389045536518097,
      "learning_rate": 8e-05,
      "loss": 1.9004,
      "step": 2152
    },
    {
      "epoch": 0.23268129255376635,
      "grad_norm": 0.4232478141784668,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 2153
    },
    {
      "epoch": 0.23278936561115313,
      "grad_norm": 0.42816585302352905,
      "learning_rate": 8e-05,
      "loss": 1.7556,
      "step": 2154
    },
    {
      "epoch": 0.23289743866853993,
      "grad_norm": 0.5027056932449341,
      "learning_rate": 8e-05,
      "loss": 2.1277,
      "step": 2155
    },
    {
      "epoch": 0.23300551172592673,
      "grad_norm": 0.38895758986473083,
      "learning_rate": 8e-05,
      "loss": 1.8178,
      "step": 2156
    },
    {
      "epoch": 0.2331135847833135,
      "grad_norm": 0.4399850070476532,
      "learning_rate": 8e-05,
      "loss": 1.7008,
      "step": 2157
    },
    {
      "epoch": 0.2332216578407003,
      "grad_norm": 0.42571064829826355,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 2158
    },
    {
      "epoch": 0.23332973089808712,
      "grad_norm": 0.4157092869281769,
      "learning_rate": 8e-05,
      "loss": 1.9287,
      "step": 2159
    },
    {
      "epoch": 0.2334378039554739,
      "grad_norm": 0.4536384046077728,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 2160
    },
    {
      "epoch": 0.2335458770128607,
      "grad_norm": 0.421044260263443,
      "learning_rate": 8e-05,
      "loss": 1.8248,
      "step": 2161
    },
    {
      "epoch": 0.23365395007024747,
      "grad_norm": 0.5051353573799133,
      "learning_rate": 8e-05,
      "loss": 1.8853,
      "step": 2162
    },
    {
      "epoch": 0.23376202312763428,
      "grad_norm": 0.4663080871105194,
      "learning_rate": 8e-05,
      "loss": 1.8036,
      "step": 2163
    },
    {
      "epoch": 0.23387009618502108,
      "grad_norm": 0.3989972174167633,
      "learning_rate": 8e-05,
      "loss": 1.8023,
      "step": 2164
    },
    {
      "epoch": 0.23397816924240786,
      "grad_norm": 0.4314860999584198,
      "learning_rate": 8e-05,
      "loss": 1.8678,
      "step": 2165
    },
    {
      "epoch": 0.23408624229979466,
      "grad_norm": 0.4305175840854645,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 2166
    },
    {
      "epoch": 0.23419431535718147,
      "grad_norm": 0.41123655438423157,
      "learning_rate": 8e-05,
      "loss": 1.8163,
      "step": 2167
    },
    {
      "epoch": 0.23430238841456824,
      "grad_norm": 0.4057143032550812,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 2168
    },
    {
      "epoch": 0.23441046147195505,
      "grad_norm": 0.4469870924949646,
      "learning_rate": 8e-05,
      "loss": 1.7853,
      "step": 2169
    },
    {
      "epoch": 0.23451853452934185,
      "grad_norm": 0.40779000520706177,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 2170
    },
    {
      "epoch": 0.23462660758672862,
      "grad_norm": 0.4358156621456146,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 2171
    },
    {
      "epoch": 0.23473468064411543,
      "grad_norm": 0.3816567361354828,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 2172
    },
    {
      "epoch": 0.2348427537015022,
      "grad_norm": 0.3964163362979889,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 2173
    },
    {
      "epoch": 0.234950826758889,
      "grad_norm": 0.4072316288948059,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 2174
    },
    {
      "epoch": 0.2350588998162758,
      "grad_norm": 0.45334392786026,
      "learning_rate": 8e-05,
      "loss": 1.8981,
      "step": 2175
    },
    {
      "epoch": 0.2351669728736626,
      "grad_norm": 0.43720847368240356,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 2176
    },
    {
      "epoch": 0.2352750459310494,
      "grad_norm": 0.4206996262073517,
      "learning_rate": 8e-05,
      "loss": 1.864,
      "step": 2177
    },
    {
      "epoch": 0.2353831189884362,
      "grad_norm": 0.41845884919166565,
      "learning_rate": 8e-05,
      "loss": 1.7948,
      "step": 2178
    },
    {
      "epoch": 0.23549119204582297,
      "grad_norm": 0.39587926864624023,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 2179
    },
    {
      "epoch": 0.23559926510320978,
      "grad_norm": 0.4395233690738678,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 2180
    },
    {
      "epoch": 0.23570733816059655,
      "grad_norm": 0.41402581334114075,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 2181
    },
    {
      "epoch": 0.23581541121798336,
      "grad_norm": 0.41368502378463745,
      "learning_rate": 8e-05,
      "loss": 1.9309,
      "step": 2182
    },
    {
      "epoch": 0.23592348427537016,
      "grad_norm": 0.39879417419433594,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 2183
    },
    {
      "epoch": 0.23603155733275694,
      "grad_norm": 0.4331943392753601,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 2184
    },
    {
      "epoch": 0.23613963039014374,
      "grad_norm": 0.40085241198539734,
      "learning_rate": 8e-05,
      "loss": 1.6506,
      "step": 2185
    },
    {
      "epoch": 0.23624770344753054,
      "grad_norm": 0.42094576358795166,
      "learning_rate": 8e-05,
      "loss": 1.8189,
      "step": 2186
    },
    {
      "epoch": 0.23635577650491732,
      "grad_norm": 0.43225178122520447,
      "learning_rate": 8e-05,
      "loss": 1.8415,
      "step": 2187
    },
    {
      "epoch": 0.23646384956230412,
      "grad_norm": 0.4020864963531494,
      "learning_rate": 8e-05,
      "loss": 1.8562,
      "step": 2188
    },
    {
      "epoch": 0.2365719226196909,
      "grad_norm": 0.4108979105949402,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 2189
    },
    {
      "epoch": 0.2366799956770777,
      "grad_norm": 0.4383971691131592,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 2190
    },
    {
      "epoch": 0.2367880687344645,
      "grad_norm": 0.39930155873298645,
      "learning_rate": 8e-05,
      "loss": 1.8181,
      "step": 2191
    },
    {
      "epoch": 0.23689614179185128,
      "grad_norm": 0.40897202491760254,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 2192
    },
    {
      "epoch": 0.2370042148492381,
      "grad_norm": 0.4251718521118164,
      "learning_rate": 8e-05,
      "loss": 1.8277,
      "step": 2193
    },
    {
      "epoch": 0.2371122879066249,
      "grad_norm": 0.49700266122817993,
      "learning_rate": 8e-05,
      "loss": 1.8192,
      "step": 2194
    },
    {
      "epoch": 0.23722036096401167,
      "grad_norm": 0.37709859013557434,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 2195
    },
    {
      "epoch": 0.23732843402139847,
      "grad_norm": 0.4134148359298706,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 2196
    },
    {
      "epoch": 0.23743650707878525,
      "grad_norm": 0.42398807406425476,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 2197
    },
    {
      "epoch": 0.23754458013617205,
      "grad_norm": 0.3958914279937744,
      "learning_rate": 8e-05,
      "loss": 1.7093,
      "step": 2198
    },
    {
      "epoch": 0.23765265319355885,
      "grad_norm": 0.41832754015922546,
      "learning_rate": 8e-05,
      "loss": 1.6368,
      "step": 2199
    },
    {
      "epoch": 0.23776072625094563,
      "grad_norm": 0.44016245007514954,
      "learning_rate": 8e-05,
      "loss": 1.9092,
      "step": 2200
    },
    {
      "epoch": 0.23786879930833243,
      "grad_norm": 0.428435355424881,
      "learning_rate": 8e-05,
      "loss": 1.8773,
      "step": 2201
    },
    {
      "epoch": 0.23797687236571924,
      "grad_norm": 0.4015369713306427,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 2202
    },
    {
      "epoch": 0.23808494542310601,
      "grad_norm": 0.42845839262008667,
      "learning_rate": 8e-05,
      "loss": 1.7904,
      "step": 2203
    },
    {
      "epoch": 0.23819301848049282,
      "grad_norm": 0.38100507855415344,
      "learning_rate": 8e-05,
      "loss": 1.7049,
      "step": 2204
    },
    {
      "epoch": 0.2383010915378796,
      "grad_norm": 0.3930203914642334,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 2205
    },
    {
      "epoch": 0.2384091645952664,
      "grad_norm": 0.4229487180709839,
      "learning_rate": 8e-05,
      "loss": 1.8626,
      "step": 2206
    },
    {
      "epoch": 0.2385172376526532,
      "grad_norm": 0.4219011068344116,
      "learning_rate": 8e-05,
      "loss": 1.8177,
      "step": 2207
    },
    {
      "epoch": 0.23862531071003998,
      "grad_norm": 0.40031224489212036,
      "learning_rate": 8e-05,
      "loss": 1.7463,
      "step": 2208
    },
    {
      "epoch": 0.23873338376742678,
      "grad_norm": 0.4321628212928772,
      "learning_rate": 8e-05,
      "loss": 1.7669,
      "step": 2209
    },
    {
      "epoch": 0.23884145682481359,
      "grad_norm": 0.39966800808906555,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 2210
    },
    {
      "epoch": 0.23894952988220036,
      "grad_norm": 0.4115314781665802,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 2211
    },
    {
      "epoch": 0.23905760293958717,
      "grad_norm": 0.4219301640987396,
      "learning_rate": 8e-05,
      "loss": 1.6744,
      "step": 2212
    },
    {
      "epoch": 0.23916567599697394,
      "grad_norm": 0.4124618470668793,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 2213
    },
    {
      "epoch": 0.23927374905436075,
      "grad_norm": 0.3870982527732849,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 2214
    },
    {
      "epoch": 0.23938182211174755,
      "grad_norm": 0.4296342730522156,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 2215
    },
    {
      "epoch": 0.23948989516913433,
      "grad_norm": 0.4123176038265228,
      "learning_rate": 8e-05,
      "loss": 1.7902,
      "step": 2216
    },
    {
      "epoch": 0.23959796822652113,
      "grad_norm": 0.3864714205265045,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 2217
    },
    {
      "epoch": 0.23970604128390793,
      "grad_norm": 0.49615928530693054,
      "learning_rate": 8e-05,
      "loss": 1.9513,
      "step": 2218
    },
    {
      "epoch": 0.2398141143412947,
      "grad_norm": 0.40624842047691345,
      "learning_rate": 8e-05,
      "loss": 1.895,
      "step": 2219
    },
    {
      "epoch": 0.2399221873986815,
      "grad_norm": 0.40863385796546936,
      "learning_rate": 8e-05,
      "loss": 1.7559,
      "step": 2220
    },
    {
      "epoch": 0.2400302604560683,
      "grad_norm": 0.41547104716300964,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 2221
    },
    {
      "epoch": 0.2401383335134551,
      "grad_norm": 0.42994973063468933,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 2222
    },
    {
      "epoch": 0.2402464065708419,
      "grad_norm": 0.4213298559188843,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 2223
    },
    {
      "epoch": 0.24035447962822867,
      "grad_norm": 0.43688419461250305,
      "learning_rate": 8e-05,
      "loss": 1.7762,
      "step": 2224
    },
    {
      "epoch": 0.24046255268561548,
      "grad_norm": 0.425220787525177,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 2225
    },
    {
      "epoch": 0.24057062574300228,
      "grad_norm": 0.46399176120758057,
      "learning_rate": 8e-05,
      "loss": 1.9178,
      "step": 2226
    },
    {
      "epoch": 0.24067869880038906,
      "grad_norm": 0.4246191084384918,
      "learning_rate": 8e-05,
      "loss": 1.6505,
      "step": 2227
    },
    {
      "epoch": 0.24078677185777586,
      "grad_norm": 0.4078132212162018,
      "learning_rate": 8e-05,
      "loss": 1.7052,
      "step": 2228
    },
    {
      "epoch": 0.24089484491516264,
      "grad_norm": 0.40307313203811646,
      "learning_rate": 8e-05,
      "loss": 1.8242,
      "step": 2229
    },
    {
      "epoch": 0.24100291797254944,
      "grad_norm": 0.4141753613948822,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 2230
    },
    {
      "epoch": 0.24111099102993624,
      "grad_norm": 0.4402991533279419,
      "learning_rate": 8e-05,
      "loss": 1.8904,
      "step": 2231
    },
    {
      "epoch": 0.24121906408732302,
      "grad_norm": 0.4160575270652771,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 2232
    },
    {
      "epoch": 0.24132713714470982,
      "grad_norm": 0.42949071526527405,
      "learning_rate": 8e-05,
      "loss": 1.8475,
      "step": 2233
    },
    {
      "epoch": 0.24143521020209663,
      "grad_norm": 0.40415865182876587,
      "learning_rate": 8e-05,
      "loss": 1.872,
      "step": 2234
    },
    {
      "epoch": 0.2415432832594834,
      "grad_norm": 0.4034290313720703,
      "learning_rate": 8e-05,
      "loss": 1.8749,
      "step": 2235
    },
    {
      "epoch": 0.2416513563168702,
      "grad_norm": 0.43408945202827454,
      "learning_rate": 8e-05,
      "loss": 1.8503,
      "step": 2236
    },
    {
      "epoch": 0.24175942937425698,
      "grad_norm": 0.4203198552131653,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 2237
    },
    {
      "epoch": 0.2418675024316438,
      "grad_norm": 0.43557363748550415,
      "learning_rate": 8e-05,
      "loss": 1.8077,
      "step": 2238
    },
    {
      "epoch": 0.2419755754890306,
      "grad_norm": 0.4128033518791199,
      "learning_rate": 8e-05,
      "loss": 1.8522,
      "step": 2239
    },
    {
      "epoch": 0.24208364854641737,
      "grad_norm": 0.450779527425766,
      "learning_rate": 8e-05,
      "loss": 1.8406,
      "step": 2240
    },
    {
      "epoch": 0.24219172160380417,
      "grad_norm": 0.4271756410598755,
      "learning_rate": 8e-05,
      "loss": 1.7964,
      "step": 2241
    },
    {
      "epoch": 0.24229979466119098,
      "grad_norm": 0.42603281140327454,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 2242
    },
    {
      "epoch": 0.24240786771857775,
      "grad_norm": 0.49153056740760803,
      "learning_rate": 8e-05,
      "loss": 2.0031,
      "step": 2243
    },
    {
      "epoch": 0.24251594077596456,
      "grad_norm": 0.42202678322792053,
      "learning_rate": 8e-05,
      "loss": 1.7801,
      "step": 2244
    },
    {
      "epoch": 0.24262401383335133,
      "grad_norm": 0.369491845369339,
      "learning_rate": 8e-05,
      "loss": 1.4895,
      "step": 2245
    },
    {
      "epoch": 0.24273208689073814,
      "grad_norm": 0.45280393958091736,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 2246
    },
    {
      "epoch": 0.24284015994812494,
      "grad_norm": 0.4014429450035095,
      "learning_rate": 8e-05,
      "loss": 1.4784,
      "step": 2247
    },
    {
      "epoch": 0.24294823300551172,
      "grad_norm": 0.40739092230796814,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 2248
    },
    {
      "epoch": 0.24305630606289852,
      "grad_norm": 0.42792797088623047,
      "learning_rate": 8e-05,
      "loss": 1.9371,
      "step": 2249
    },
    {
      "epoch": 0.24316437912028532,
      "grad_norm": 0.4220174551010132,
      "learning_rate": 8e-05,
      "loss": 1.7889,
      "step": 2250
    },
    {
      "epoch": 0.2432724521776721,
      "grad_norm": 0.5363878607749939,
      "learning_rate": 8e-05,
      "loss": 1.4759,
      "step": 2251
    },
    {
      "epoch": 0.2433805252350589,
      "grad_norm": 0.42898300290107727,
      "learning_rate": 8e-05,
      "loss": 1.8691,
      "step": 2252
    },
    {
      "epoch": 0.2434885982924457,
      "grad_norm": 0.42363789677619934,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 2253
    },
    {
      "epoch": 0.24359667134983248,
      "grad_norm": 0.4130452573299408,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 2254
    },
    {
      "epoch": 0.2437047444072193,
      "grad_norm": 0.405157208442688,
      "learning_rate": 8e-05,
      "loss": 1.8158,
      "step": 2255
    },
    {
      "epoch": 0.24381281746460606,
      "grad_norm": 0.40729719400405884,
      "learning_rate": 8e-05,
      "loss": 1.7895,
      "step": 2256
    },
    {
      "epoch": 0.24392089052199287,
      "grad_norm": 0.40244317054748535,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 2257
    },
    {
      "epoch": 0.24402896357937967,
      "grad_norm": 0.38459038734436035,
      "learning_rate": 8e-05,
      "loss": 1.5896,
      "step": 2258
    },
    {
      "epoch": 0.24413703663676645,
      "grad_norm": 0.407066285610199,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 2259
    },
    {
      "epoch": 0.24424510969415325,
      "grad_norm": 0.4670327305793762,
      "learning_rate": 8e-05,
      "loss": 1.8216,
      "step": 2260
    },
    {
      "epoch": 0.24435318275154005,
      "grad_norm": 0.3835848271846771,
      "learning_rate": 8e-05,
      "loss": 1.5857,
      "step": 2261
    },
    {
      "epoch": 0.24446125580892683,
      "grad_norm": 0.3995159864425659,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 2262
    },
    {
      "epoch": 0.24456932886631363,
      "grad_norm": 0.4269305169582367,
      "learning_rate": 8e-05,
      "loss": 1.9002,
      "step": 2263
    },
    {
      "epoch": 0.2446774019237004,
      "grad_norm": 0.4051995575428009,
      "learning_rate": 8e-05,
      "loss": 1.8266,
      "step": 2264
    },
    {
      "epoch": 0.24478547498108721,
      "grad_norm": 0.3955152928829193,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 2265
    },
    {
      "epoch": 0.24489354803847402,
      "grad_norm": 0.4389755427837372,
      "learning_rate": 8e-05,
      "loss": 1.8102,
      "step": 2266
    },
    {
      "epoch": 0.2450016210958608,
      "grad_norm": 0.40770477056503296,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 2267
    },
    {
      "epoch": 0.2451096941532476,
      "grad_norm": 0.4001387655735016,
      "learning_rate": 8e-05,
      "loss": 1.5861,
      "step": 2268
    },
    {
      "epoch": 0.2452177672106344,
      "grad_norm": 0.45649421215057373,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 2269
    },
    {
      "epoch": 0.24532584026802118,
      "grad_norm": 0.4416021406650543,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 2270
    },
    {
      "epoch": 0.24543391332540798,
      "grad_norm": 0.40319493412971497,
      "learning_rate": 8e-05,
      "loss": 1.743,
      "step": 2271
    },
    {
      "epoch": 0.24554198638279476,
      "grad_norm": 0.4448947012424469,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 2272
    },
    {
      "epoch": 0.24565005944018156,
      "grad_norm": 0.3970223665237427,
      "learning_rate": 8e-05,
      "loss": 1.849,
      "step": 2273
    },
    {
      "epoch": 0.24575813249756837,
      "grad_norm": 0.4911063015460968,
      "learning_rate": 8e-05,
      "loss": 1.9903,
      "step": 2274
    },
    {
      "epoch": 0.24586620555495514,
      "grad_norm": 0.41442355513572693,
      "learning_rate": 8e-05,
      "loss": 1.7839,
      "step": 2275
    },
    {
      "epoch": 0.24597427861234195,
      "grad_norm": 0.40242794156074524,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 2276
    },
    {
      "epoch": 0.24608235166972875,
      "grad_norm": 0.36094552278518677,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 2277
    },
    {
      "epoch": 0.24619042472711553,
      "grad_norm": 0.4204481244087219,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 2278
    },
    {
      "epoch": 0.24629849778450233,
      "grad_norm": 0.44344159960746765,
      "learning_rate": 8e-05,
      "loss": 1.5746,
      "step": 2279
    },
    {
      "epoch": 0.2464065708418891,
      "grad_norm": 0.4481213688850403,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 2280
    },
    {
      "epoch": 0.2465146438992759,
      "grad_norm": 0.45191749930381775,
      "learning_rate": 8e-05,
      "loss": 1.5836,
      "step": 2281
    },
    {
      "epoch": 0.2466227169566627,
      "grad_norm": 0.40741199254989624,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 2282
    },
    {
      "epoch": 0.2467307900140495,
      "grad_norm": 0.43323972821235657,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 2283
    },
    {
      "epoch": 0.2468388630714363,
      "grad_norm": 0.40379416942596436,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 2284
    },
    {
      "epoch": 0.2469469361288231,
      "grad_norm": 0.43622714281082153,
      "learning_rate": 8e-05,
      "loss": 1.9507,
      "step": 2285
    },
    {
      "epoch": 0.24705500918620987,
      "grad_norm": 0.41648638248443604,
      "learning_rate": 8e-05,
      "loss": 1.8118,
      "step": 2286
    },
    {
      "epoch": 0.24716308224359668,
      "grad_norm": 0.4070819914340973,
      "learning_rate": 8e-05,
      "loss": 1.6609,
      "step": 2287
    },
    {
      "epoch": 0.24727115530098345,
      "grad_norm": 0.3831266760826111,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 2288
    },
    {
      "epoch": 0.24737922835837026,
      "grad_norm": 0.40246596932411194,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 2289
    },
    {
      "epoch": 0.24748730141575706,
      "grad_norm": 0.4052984118461609,
      "learning_rate": 8e-05,
      "loss": 1.4576,
      "step": 2290
    },
    {
      "epoch": 0.24759537447314384,
      "grad_norm": 0.41676801443099976,
      "learning_rate": 8e-05,
      "loss": 1.862,
      "step": 2291
    },
    {
      "epoch": 0.24770344753053064,
      "grad_norm": 0.38884052634239197,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 2292
    },
    {
      "epoch": 0.24781152058791744,
      "grad_norm": 0.383065402507782,
      "learning_rate": 8e-05,
      "loss": 1.8258,
      "step": 2293
    },
    {
      "epoch": 0.24791959364530422,
      "grad_norm": 0.3977891802787781,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 2294
    },
    {
      "epoch": 0.24802766670269102,
      "grad_norm": 0.448526531457901,
      "learning_rate": 8e-05,
      "loss": 1.7475,
      "step": 2295
    },
    {
      "epoch": 0.2481357397600778,
      "grad_norm": 0.38719141483306885,
      "learning_rate": 8e-05,
      "loss": 1.7108,
      "step": 2296
    },
    {
      "epoch": 0.2482438128174646,
      "grad_norm": 0.4766417145729065,
      "learning_rate": 8e-05,
      "loss": 1.7013,
      "step": 2297
    },
    {
      "epoch": 0.2483518858748514,
      "grad_norm": 0.4068244397640228,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 2298
    },
    {
      "epoch": 0.24845995893223818,
      "grad_norm": 0.40211567282676697,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 2299
    },
    {
      "epoch": 0.248568031989625,
      "grad_norm": 0.4327814280986786,
      "learning_rate": 8e-05,
      "loss": 1.8347,
      "step": 2300
    },
    {
      "epoch": 0.2486761050470118,
      "grad_norm": 0.3885396122932434,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 2301
    },
    {
      "epoch": 0.24878417810439857,
      "grad_norm": 0.3891986608505249,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 2302
    },
    {
      "epoch": 0.24889225116178537,
      "grad_norm": 0.38505733013153076,
      "learning_rate": 8e-05,
      "loss": 1.7217,
      "step": 2303
    },
    {
      "epoch": 0.24900032421917215,
      "grad_norm": 0.3769560754299164,
      "learning_rate": 8e-05,
      "loss": 1.6218,
      "step": 2304
    },
    {
      "epoch": 0.24910839727655895,
      "grad_norm": 0.41368505358695984,
      "learning_rate": 8e-05,
      "loss": 1.8187,
      "step": 2305
    },
    {
      "epoch": 0.24921647033394576,
      "grad_norm": 0.44439974427223206,
      "learning_rate": 8e-05,
      "loss": 1.8913,
      "step": 2306
    },
    {
      "epoch": 0.24932454339133253,
      "grad_norm": 0.41765937209129333,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 2307
    },
    {
      "epoch": 0.24943261644871934,
      "grad_norm": 0.40139949321746826,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 2308
    },
    {
      "epoch": 0.24954068950610614,
      "grad_norm": 0.48666390776634216,
      "learning_rate": 8e-05,
      "loss": 1.8818,
      "step": 2309
    },
    {
      "epoch": 0.24964876256349292,
      "grad_norm": 0.42327880859375,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 2310
    },
    {
      "epoch": 0.24975683562087972,
      "grad_norm": 0.41644787788391113,
      "learning_rate": 8e-05,
      "loss": 1.6234,
      "step": 2311
    },
    {
      "epoch": 0.2498649086782665,
      "grad_norm": 0.37401750683784485,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 2312
    },
    {
      "epoch": 0.2499729817356533,
      "grad_norm": 0.39776623249053955,
      "learning_rate": 8e-05,
      "loss": 1.5088,
      "step": 2313
    },
    {
      "epoch": 0.2500810547930401,
      "grad_norm": 0.4314245879650116,
      "learning_rate": 8e-05,
      "loss": 1.8678,
      "step": 2314
    },
    {
      "epoch": 0.2501891278504269,
      "grad_norm": 0.4315459430217743,
      "learning_rate": 8e-05,
      "loss": 1.873,
      "step": 2315
    },
    {
      "epoch": 0.25029720090781366,
      "grad_norm": 0.4031311273574829,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 2316
    },
    {
      "epoch": 0.25040527396520046,
      "grad_norm": 0.3810919523239136,
      "learning_rate": 8e-05,
      "loss": 1.6935,
      "step": 2317
    },
    {
      "epoch": 0.25051334702258726,
      "grad_norm": 0.4532819986343384,
      "learning_rate": 8e-05,
      "loss": 1.9868,
      "step": 2318
    },
    {
      "epoch": 0.25062142007997407,
      "grad_norm": 0.3963771462440491,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 2319
    },
    {
      "epoch": 0.25072949313736087,
      "grad_norm": 0.3919433653354645,
      "learning_rate": 8e-05,
      "loss": 1.6527,
      "step": 2320
    },
    {
      "epoch": 0.2508375661947477,
      "grad_norm": 0.44893065094947815,
      "learning_rate": 8e-05,
      "loss": 1.866,
      "step": 2321
    },
    {
      "epoch": 0.2509456392521344,
      "grad_norm": 0.40720558166503906,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 2322
    },
    {
      "epoch": 0.2510537123095212,
      "grad_norm": 0.40114784240722656,
      "learning_rate": 8e-05,
      "loss": 1.8209,
      "step": 2323
    },
    {
      "epoch": 0.25116178536690803,
      "grad_norm": 0.4403935372829437,
      "learning_rate": 8e-05,
      "loss": 1.8585,
      "step": 2324
    },
    {
      "epoch": 0.25126985842429483,
      "grad_norm": 0.4276638925075531,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 2325
    },
    {
      "epoch": 0.25137793148168164,
      "grad_norm": 0.4108809232711792,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 2326
    },
    {
      "epoch": 0.2514860045390684,
      "grad_norm": 0.3799722492694855,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 2327
    },
    {
      "epoch": 0.2515940775964552,
      "grad_norm": 0.45009803771972656,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 2328
    },
    {
      "epoch": 0.251702150653842,
      "grad_norm": 0.4261547029018402,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 2329
    },
    {
      "epoch": 0.2518102237112288,
      "grad_norm": 0.4598546326160431,
      "learning_rate": 8e-05,
      "loss": 1.9218,
      "step": 2330
    },
    {
      "epoch": 0.2519182967686156,
      "grad_norm": 0.41281354427337646,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 2331
    },
    {
      "epoch": 0.25202636982600235,
      "grad_norm": 0.39392974972724915,
      "learning_rate": 8e-05,
      "loss": 1.6933,
      "step": 2332
    },
    {
      "epoch": 0.25213444288338915,
      "grad_norm": 0.42078039050102234,
      "learning_rate": 8e-05,
      "loss": 1.6697,
      "step": 2333
    },
    {
      "epoch": 0.25224251594077596,
      "grad_norm": 0.45850658416748047,
      "learning_rate": 8e-05,
      "loss": 2.0073,
      "step": 2334
    },
    {
      "epoch": 0.25235058899816276,
      "grad_norm": 0.4402531683444977,
      "learning_rate": 8e-05,
      "loss": 1.9171,
      "step": 2335
    },
    {
      "epoch": 0.25245866205554957,
      "grad_norm": 0.49597933888435364,
      "learning_rate": 8e-05,
      "loss": 1.6742,
      "step": 2336
    },
    {
      "epoch": 0.25256673511293637,
      "grad_norm": 0.4358651638031006,
      "learning_rate": 8e-05,
      "loss": 1.8715,
      "step": 2337
    },
    {
      "epoch": 0.2526748081703231,
      "grad_norm": 0.4081408679485321,
      "learning_rate": 8e-05,
      "loss": 1.5572,
      "step": 2338
    },
    {
      "epoch": 0.2527828812277099,
      "grad_norm": 0.40265968441963196,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 2339
    },
    {
      "epoch": 0.2528909542850967,
      "grad_norm": 0.41433006525039673,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 2340
    },
    {
      "epoch": 0.25299902734248353,
      "grad_norm": 0.4183926582336426,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 2341
    },
    {
      "epoch": 0.25310710039987033,
      "grad_norm": 0.43320661783218384,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 2342
    },
    {
      "epoch": 0.2532151734572571,
      "grad_norm": 0.4419490098953247,
      "learning_rate": 8e-05,
      "loss": 1.8008,
      "step": 2343
    },
    {
      "epoch": 0.2533232465146439,
      "grad_norm": 0.39533042907714844,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 2344
    },
    {
      "epoch": 0.2534313195720307,
      "grad_norm": 0.38746196031570435,
      "learning_rate": 8e-05,
      "loss": 1.7079,
      "step": 2345
    },
    {
      "epoch": 0.2535393926294175,
      "grad_norm": 0.39679914712905884,
      "learning_rate": 8e-05,
      "loss": 1.5854,
      "step": 2346
    },
    {
      "epoch": 0.2536474656868043,
      "grad_norm": 0.46514543890953064,
      "learning_rate": 8e-05,
      "loss": 1.6957,
      "step": 2347
    },
    {
      "epoch": 0.2537555387441911,
      "grad_norm": 0.4088696539402008,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 2348
    },
    {
      "epoch": 0.25386361180157785,
      "grad_norm": 0.44849875569343567,
      "learning_rate": 8e-05,
      "loss": 1.8765,
      "step": 2349
    },
    {
      "epoch": 0.25397168485896465,
      "grad_norm": 0.40090954303741455,
      "learning_rate": 8e-05,
      "loss": 1.6413,
      "step": 2350
    },
    {
      "epoch": 0.25407975791635146,
      "grad_norm": 0.4090440273284912,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 2351
    },
    {
      "epoch": 0.25418783097373826,
      "grad_norm": 0.41185376048088074,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 2352
    },
    {
      "epoch": 0.25429590403112506,
      "grad_norm": 0.403613418340683,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 2353
    },
    {
      "epoch": 0.2544039770885118,
      "grad_norm": 0.4229564666748047,
      "learning_rate": 8e-05,
      "loss": 1.8863,
      "step": 2354
    },
    {
      "epoch": 0.2545120501458986,
      "grad_norm": 0.44485747814178467,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 2355
    },
    {
      "epoch": 0.2546201232032854,
      "grad_norm": 0.4516328275203705,
      "learning_rate": 8e-05,
      "loss": 1.8927,
      "step": 2356
    },
    {
      "epoch": 0.2547281962606722,
      "grad_norm": 0.392254114151001,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 2357
    },
    {
      "epoch": 0.254836269318059,
      "grad_norm": 0.42300093173980713,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 2358
    },
    {
      "epoch": 0.2549443423754458,
      "grad_norm": 0.40058445930480957,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 2359
    },
    {
      "epoch": 0.2550524154328326,
      "grad_norm": 0.3937632143497467,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 2360
    },
    {
      "epoch": 0.2551604884902194,
      "grad_norm": 0.42444199323654175,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 2361
    },
    {
      "epoch": 0.2552685615476062,
      "grad_norm": 0.41478320956230164,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 2362
    },
    {
      "epoch": 0.255376634604993,
      "grad_norm": 0.411935418844223,
      "learning_rate": 8e-05,
      "loss": 1.7537,
      "step": 2363
    },
    {
      "epoch": 0.2554847076623798,
      "grad_norm": 0.40553733706474304,
      "learning_rate": 8e-05,
      "loss": 1.6323,
      "step": 2364
    },
    {
      "epoch": 0.25559278071976654,
      "grad_norm": 0.43338435888290405,
      "learning_rate": 8e-05,
      "loss": 1.8136,
      "step": 2365
    },
    {
      "epoch": 0.25570085377715335,
      "grad_norm": 0.4088203012943268,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 2366
    },
    {
      "epoch": 0.25580892683454015,
      "grad_norm": 0.4171682298183441,
      "learning_rate": 8e-05,
      "loss": 1.661,
      "step": 2367
    },
    {
      "epoch": 0.25591699989192696,
      "grad_norm": 0.3987179100513458,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 2368
    },
    {
      "epoch": 0.25602507294931376,
      "grad_norm": 0.42314612865448,
      "learning_rate": 8e-05,
      "loss": 1.6302,
      "step": 2369
    },
    {
      "epoch": 0.2561331460067005,
      "grad_norm": 0.37354180216789246,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 2370
    },
    {
      "epoch": 0.2562412190640873,
      "grad_norm": 0.38359418511390686,
      "learning_rate": 8e-05,
      "loss": 1.5438,
      "step": 2371
    },
    {
      "epoch": 0.2563492921214741,
      "grad_norm": 0.38693997263908386,
      "learning_rate": 8e-05,
      "loss": 1.5209,
      "step": 2372
    },
    {
      "epoch": 0.2564573651788609,
      "grad_norm": 0.4193786382675171,
      "learning_rate": 8e-05,
      "loss": 1.8685,
      "step": 2373
    },
    {
      "epoch": 0.2565654382362477,
      "grad_norm": 0.441204309463501,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 2374
    },
    {
      "epoch": 0.25667351129363447,
      "grad_norm": 0.42206552624702454,
      "learning_rate": 8e-05,
      "loss": 1.9132,
      "step": 2375
    },
    {
      "epoch": 0.2567815843510213,
      "grad_norm": 0.4019717872142792,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 2376
    },
    {
      "epoch": 0.2568896574084081,
      "grad_norm": 0.42373108863830566,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 2377
    },
    {
      "epoch": 0.2569977304657949,
      "grad_norm": 0.3850076496601105,
      "learning_rate": 8e-05,
      "loss": 1.5346,
      "step": 2378
    },
    {
      "epoch": 0.2571058035231817,
      "grad_norm": 0.39058008790016174,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 2379
    },
    {
      "epoch": 0.2572138765805685,
      "grad_norm": 0.40224769711494446,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 2380
    },
    {
      "epoch": 0.25732194963795524,
      "grad_norm": 0.3849408030509949,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 2381
    },
    {
      "epoch": 0.25743002269534204,
      "grad_norm": 0.41313230991363525,
      "learning_rate": 8e-05,
      "loss": 1.6772,
      "step": 2382
    },
    {
      "epoch": 0.25753809575272885,
      "grad_norm": 0.43629127740859985,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 2383
    },
    {
      "epoch": 0.25764616881011565,
      "grad_norm": 0.4019980728626251,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 2384
    },
    {
      "epoch": 0.25775424186750245,
      "grad_norm": 0.4354315400123596,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 2385
    },
    {
      "epoch": 0.2578623149248892,
      "grad_norm": 0.40154799818992615,
      "learning_rate": 8e-05,
      "loss": 1.6773,
      "step": 2386
    },
    {
      "epoch": 0.257970387982276,
      "grad_norm": 0.3961428105831146,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 2387
    },
    {
      "epoch": 0.2580784610396628,
      "grad_norm": 0.39879611134529114,
      "learning_rate": 8e-05,
      "loss": 1.6939,
      "step": 2388
    },
    {
      "epoch": 0.2581865340970496,
      "grad_norm": 0.4064280390739441,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 2389
    },
    {
      "epoch": 0.2582946071544364,
      "grad_norm": 0.39445245265960693,
      "learning_rate": 8e-05,
      "loss": 1.6661,
      "step": 2390
    },
    {
      "epoch": 0.25840268021182317,
      "grad_norm": 0.4098775088787079,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 2391
    },
    {
      "epoch": 0.25851075326920997,
      "grad_norm": 0.4036963880062103,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 2392
    },
    {
      "epoch": 0.2586188263265968,
      "grad_norm": 0.4389079511165619,
      "learning_rate": 8e-05,
      "loss": 1.861,
      "step": 2393
    },
    {
      "epoch": 0.2587268993839836,
      "grad_norm": 0.42100366950035095,
      "learning_rate": 8e-05,
      "loss": 1.7743,
      "step": 2394
    },
    {
      "epoch": 0.2588349724413704,
      "grad_norm": 0.4010426998138428,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 2395
    },
    {
      "epoch": 0.2589430454987572,
      "grad_norm": 0.4156354069709778,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 2396
    },
    {
      "epoch": 0.25905111855614393,
      "grad_norm": 0.4719531238079071,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 2397
    },
    {
      "epoch": 0.25915919161353074,
      "grad_norm": 0.3908827304840088,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 2398
    },
    {
      "epoch": 0.25926726467091754,
      "grad_norm": 0.419207900762558,
      "learning_rate": 8e-05,
      "loss": 1.8141,
      "step": 2399
    },
    {
      "epoch": 0.25937533772830434,
      "grad_norm": 0.47012627124786377,
      "learning_rate": 8e-05,
      "loss": 1.4727,
      "step": 2400
    },
    {
      "epoch": 0.25948341078569115,
      "grad_norm": 0.3906846344470978,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 2401
    },
    {
      "epoch": 0.2595914838430779,
      "grad_norm": 0.3847828209400177,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 2402
    },
    {
      "epoch": 0.2596995569004647,
      "grad_norm": 0.3805268108844757,
      "learning_rate": 8e-05,
      "loss": 1.4542,
      "step": 2403
    },
    {
      "epoch": 0.2598076299578515,
      "grad_norm": 0.40234705805778503,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 2404
    },
    {
      "epoch": 0.2599157030152383,
      "grad_norm": 0.4033569395542145,
      "learning_rate": 8e-05,
      "loss": 1.8212,
      "step": 2405
    },
    {
      "epoch": 0.2600237760726251,
      "grad_norm": 0.38345420360565186,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 2406
    },
    {
      "epoch": 0.26013184913001186,
      "grad_norm": 0.3733268976211548,
      "learning_rate": 8e-05,
      "loss": 1.6001,
      "step": 2407
    },
    {
      "epoch": 0.26023992218739866,
      "grad_norm": 0.40710151195526123,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 2408
    },
    {
      "epoch": 0.26034799524478547,
      "grad_norm": 0.40899458527565,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 2409
    },
    {
      "epoch": 0.26045606830217227,
      "grad_norm": 0.39452487230300903,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 2410
    },
    {
      "epoch": 0.2605641413595591,
      "grad_norm": 0.4244062304496765,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 2411
    },
    {
      "epoch": 0.2606722144169459,
      "grad_norm": 0.45577219128608704,
      "learning_rate": 8e-05,
      "loss": 1.8801,
      "step": 2412
    },
    {
      "epoch": 0.26078028747433263,
      "grad_norm": 0.5084968209266663,
      "learning_rate": 8e-05,
      "loss": 1.6557,
      "step": 2413
    },
    {
      "epoch": 0.26088836053171943,
      "grad_norm": 0.45207586884498596,
      "learning_rate": 8e-05,
      "loss": 1.9198,
      "step": 2414
    },
    {
      "epoch": 0.26099643358910624,
      "grad_norm": 0.42489832639694214,
      "learning_rate": 8e-05,
      "loss": 1.7323,
      "step": 2415
    },
    {
      "epoch": 0.26110450664649304,
      "grad_norm": 0.38597649335861206,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 2416
    },
    {
      "epoch": 0.26121257970387984,
      "grad_norm": 0.47974640130996704,
      "learning_rate": 8e-05,
      "loss": 1.8777,
      "step": 2417
    },
    {
      "epoch": 0.2613206527612666,
      "grad_norm": 0.39351141452789307,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 2418
    },
    {
      "epoch": 0.2614287258186534,
      "grad_norm": 0.3972243070602417,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 2419
    },
    {
      "epoch": 0.2615367988760402,
      "grad_norm": 0.3908000588417053,
      "learning_rate": 8e-05,
      "loss": 1.696,
      "step": 2420
    },
    {
      "epoch": 0.261644871933427,
      "grad_norm": 0.4204675257205963,
      "learning_rate": 8e-05,
      "loss": 1.8643,
      "step": 2421
    },
    {
      "epoch": 0.2617529449908138,
      "grad_norm": 0.42249107360839844,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 2422
    },
    {
      "epoch": 0.2618610180482006,
      "grad_norm": 0.4392317831516266,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 2423
    },
    {
      "epoch": 0.26196909110558736,
      "grad_norm": 0.4058781564235687,
      "learning_rate": 8e-05,
      "loss": 1.5353,
      "step": 2424
    },
    {
      "epoch": 0.26207716416297416,
      "grad_norm": 0.432900995016098,
      "learning_rate": 8e-05,
      "loss": 1.8427,
      "step": 2425
    },
    {
      "epoch": 0.26218523722036097,
      "grad_norm": 0.4441083073616028,
      "learning_rate": 8e-05,
      "loss": 1.8593,
      "step": 2426
    },
    {
      "epoch": 0.26229331027774777,
      "grad_norm": 0.3977499008178711,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 2427
    },
    {
      "epoch": 0.2624013833351346,
      "grad_norm": 0.41435614228248596,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 2428
    },
    {
      "epoch": 0.2625094563925213,
      "grad_norm": 0.4837009310722351,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 2429
    },
    {
      "epoch": 0.2626175294499081,
      "grad_norm": 0.4283585548400879,
      "learning_rate": 8e-05,
      "loss": 1.9623,
      "step": 2430
    },
    {
      "epoch": 0.26272560250729493,
      "grad_norm": 0.46270594000816345,
      "learning_rate": 8e-05,
      "loss": 2.0128,
      "step": 2431
    },
    {
      "epoch": 0.26283367556468173,
      "grad_norm": 0.42940205335617065,
      "learning_rate": 8e-05,
      "loss": 1.8531,
      "step": 2432
    },
    {
      "epoch": 0.26294174862206854,
      "grad_norm": 0.37251758575439453,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 2433
    },
    {
      "epoch": 0.2630498216794553,
      "grad_norm": 0.4033619463443756,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 2434
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.4370531737804413,
      "learning_rate": 8e-05,
      "loss": 1.9177,
      "step": 2435
    },
    {
      "epoch": 0.2632659677942289,
      "grad_norm": 0.41248819231987,
      "learning_rate": 8e-05,
      "loss": 1.7334,
      "step": 2436
    },
    {
      "epoch": 0.2633740408516157,
      "grad_norm": 0.4334859251976013,
      "learning_rate": 8e-05,
      "loss": 1.688,
      "step": 2437
    },
    {
      "epoch": 0.2634821139090025,
      "grad_norm": 0.43158021569252014,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 2438
    },
    {
      "epoch": 0.2635901869663893,
      "grad_norm": 0.3843032419681549,
      "learning_rate": 8e-05,
      "loss": 1.564,
      "step": 2439
    },
    {
      "epoch": 0.26369826002377605,
      "grad_norm": 0.4330095648765564,
      "learning_rate": 8e-05,
      "loss": 1.8548,
      "step": 2440
    },
    {
      "epoch": 0.26380633308116286,
      "grad_norm": 0.4230830669403076,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 2441
    },
    {
      "epoch": 0.26391440613854966,
      "grad_norm": 0.3759700655937195,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 2442
    },
    {
      "epoch": 0.26402247919593647,
      "grad_norm": 0.4251384139060974,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 2443
    },
    {
      "epoch": 0.26413055225332327,
      "grad_norm": 0.41189393401145935,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 2444
    },
    {
      "epoch": 0.26423862531071,
      "grad_norm": 0.41445931792259216,
      "learning_rate": 8e-05,
      "loss": 1.6974,
      "step": 2445
    },
    {
      "epoch": 0.2643466983680968,
      "grad_norm": 0.42019417881965637,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 2446
    },
    {
      "epoch": 0.2644547714254836,
      "grad_norm": 0.37610286474227905,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 2447
    },
    {
      "epoch": 0.26456284448287043,
      "grad_norm": 0.47326356172561646,
      "learning_rate": 8e-05,
      "loss": 1.625,
      "step": 2448
    },
    {
      "epoch": 0.26467091754025723,
      "grad_norm": 0.4132389426231384,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 2449
    },
    {
      "epoch": 0.264778990597644,
      "grad_norm": 0.4711586833000183,
      "learning_rate": 8e-05,
      "loss": 1.8691,
      "step": 2450
    },
    {
      "epoch": 0.2648870636550308,
      "grad_norm": 0.41372689604759216,
      "learning_rate": 8e-05,
      "loss": 1.7076,
      "step": 2451
    },
    {
      "epoch": 0.2649951367124176,
      "grad_norm": 0.40434136986732483,
      "learning_rate": 8e-05,
      "loss": 1.7588,
      "step": 2452
    },
    {
      "epoch": 0.2651032097698044,
      "grad_norm": 0.44160497188568115,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 2453
    },
    {
      "epoch": 0.2652112828271912,
      "grad_norm": 0.42203038930892944,
      "learning_rate": 8e-05,
      "loss": 1.8835,
      "step": 2454
    },
    {
      "epoch": 0.265319355884578,
      "grad_norm": 0.38666704297065735,
      "learning_rate": 8e-05,
      "loss": 1.6272,
      "step": 2455
    },
    {
      "epoch": 0.26542742894196475,
      "grad_norm": 0.46934300661087036,
      "learning_rate": 8e-05,
      "loss": 1.9964,
      "step": 2456
    },
    {
      "epoch": 0.26553550199935155,
      "grad_norm": 0.41169750690460205,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 2457
    },
    {
      "epoch": 0.26564357505673836,
      "grad_norm": 0.41503971815109253,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 2458
    },
    {
      "epoch": 0.26575164811412516,
      "grad_norm": 0.4179888963699341,
      "learning_rate": 8e-05,
      "loss": 1.5367,
      "step": 2459
    },
    {
      "epoch": 0.26585972117151196,
      "grad_norm": 0.41574546694755554,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 2460
    },
    {
      "epoch": 0.2659677942288987,
      "grad_norm": 0.42333751916885376,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 2461
    },
    {
      "epoch": 0.2660758672862855,
      "grad_norm": 0.440268874168396,
      "learning_rate": 8e-05,
      "loss": 1.9519,
      "step": 2462
    },
    {
      "epoch": 0.2661839403436723,
      "grad_norm": 0.42688265442848206,
      "learning_rate": 8e-05,
      "loss": 1.7239,
      "step": 2463
    },
    {
      "epoch": 0.2662920134010591,
      "grad_norm": 0.4086346924304962,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 2464
    },
    {
      "epoch": 0.26640008645844593,
      "grad_norm": 0.42593589425086975,
      "learning_rate": 8e-05,
      "loss": 1.7189,
      "step": 2465
    },
    {
      "epoch": 0.2665081595158327,
      "grad_norm": 0.39568060636520386,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 2466
    },
    {
      "epoch": 0.2666162325732195,
      "grad_norm": 0.4496956765651703,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 2467
    },
    {
      "epoch": 0.2667243056306063,
      "grad_norm": 0.40430474281311035,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 2468
    },
    {
      "epoch": 0.2668323786879931,
      "grad_norm": 0.41903093457221985,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 2469
    },
    {
      "epoch": 0.2669404517453799,
      "grad_norm": 0.43242520093917847,
      "learning_rate": 8e-05,
      "loss": 1.7538,
      "step": 2470
    },
    {
      "epoch": 0.2670485248027667,
      "grad_norm": 0.3734981417655945,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 2471
    },
    {
      "epoch": 0.26715659786015344,
      "grad_norm": 0.4402594268321991,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 2472
    },
    {
      "epoch": 0.26726467091754025,
      "grad_norm": 0.3927295207977295,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 2473
    },
    {
      "epoch": 0.26737274397492705,
      "grad_norm": 0.43620219826698303,
      "learning_rate": 8e-05,
      "loss": 1.3694,
      "step": 2474
    },
    {
      "epoch": 0.26748081703231386,
      "grad_norm": 0.40087005496025085,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 2475
    },
    {
      "epoch": 0.26758889008970066,
      "grad_norm": 0.403878390789032,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 2476
    },
    {
      "epoch": 0.2676969631470874,
      "grad_norm": 0.43468159437179565,
      "learning_rate": 8e-05,
      "loss": 1.8833,
      "step": 2477
    },
    {
      "epoch": 0.2678050362044742,
      "grad_norm": 0.4263123869895935,
      "learning_rate": 8e-05,
      "loss": 1.5457,
      "step": 2478
    },
    {
      "epoch": 0.267913109261861,
      "grad_norm": 0.41384702920913696,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 2479
    },
    {
      "epoch": 0.2680211823192478,
      "grad_norm": 0.4466177821159363,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 2480
    },
    {
      "epoch": 0.2681292553766346,
      "grad_norm": 0.4065958261489868,
      "learning_rate": 8e-05,
      "loss": 1.7614,
      "step": 2481
    },
    {
      "epoch": 0.26823732843402137,
      "grad_norm": 0.42063117027282715,
      "learning_rate": 8e-05,
      "loss": 1.8342,
      "step": 2482
    },
    {
      "epoch": 0.2683454014914082,
      "grad_norm": 0.393279492855072,
      "learning_rate": 8e-05,
      "loss": 1.6586,
      "step": 2483
    },
    {
      "epoch": 0.268453474548795,
      "grad_norm": 0.4375686049461365,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 2484
    },
    {
      "epoch": 0.2685615476061818,
      "grad_norm": 0.41318395733833313,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 2485
    },
    {
      "epoch": 0.2686696206635686,
      "grad_norm": 0.393378347158432,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 2486
    },
    {
      "epoch": 0.2687776937209554,
      "grad_norm": 0.411169171333313,
      "learning_rate": 8e-05,
      "loss": 1.8216,
      "step": 2487
    },
    {
      "epoch": 0.26888576677834214,
      "grad_norm": 0.42677047848701477,
      "learning_rate": 8e-05,
      "loss": 1.4333,
      "step": 2488
    },
    {
      "epoch": 0.26899383983572894,
      "grad_norm": 0.3910801410675049,
      "learning_rate": 8e-05,
      "loss": 1.5064,
      "step": 2489
    },
    {
      "epoch": 0.26910191289311575,
      "grad_norm": 0.45566707849502563,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 2490
    },
    {
      "epoch": 0.26920998595050255,
      "grad_norm": 0.42279165983200073,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 2491
    },
    {
      "epoch": 0.26931805900788935,
      "grad_norm": 0.42037433385849,
      "learning_rate": 8e-05,
      "loss": 1.7939,
      "step": 2492
    },
    {
      "epoch": 0.2694261320652761,
      "grad_norm": 0.4562150239944458,
      "learning_rate": 8e-05,
      "loss": 1.8644,
      "step": 2493
    },
    {
      "epoch": 0.2695342051226629,
      "grad_norm": 0.3911263644695282,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 2494
    },
    {
      "epoch": 0.2696422781800497,
      "grad_norm": 0.4105684161186218,
      "learning_rate": 8e-05,
      "loss": 1.6649,
      "step": 2495
    },
    {
      "epoch": 0.2697503512374365,
      "grad_norm": 0.43488675355911255,
      "learning_rate": 8e-05,
      "loss": 1.8082,
      "step": 2496
    },
    {
      "epoch": 0.2698584242948233,
      "grad_norm": 0.38742998242378235,
      "learning_rate": 8e-05,
      "loss": 1.7498,
      "step": 2497
    },
    {
      "epoch": 0.26996649735221007,
      "grad_norm": 0.46524062752723694,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 2498
    },
    {
      "epoch": 0.27007457040959687,
      "grad_norm": 0.4174506962299347,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 2499
    },
    {
      "epoch": 0.2701826434669837,
      "grad_norm": 0.4334715008735657,
      "learning_rate": 8e-05,
      "loss": 1.7507,
      "step": 2500
    },
    {
      "epoch": 0.2702907165243705,
      "grad_norm": 0.41707009077072144,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 2501
    },
    {
      "epoch": 0.2703987895817573,
      "grad_norm": 0.4460618793964386,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 2502
    },
    {
      "epoch": 0.2705068626391441,
      "grad_norm": 0.40407586097717285,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 2503
    },
    {
      "epoch": 0.27061493569653083,
      "grad_norm": 0.41150081157684326,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 2504
    },
    {
      "epoch": 0.27072300875391764,
      "grad_norm": 0.4427175223827362,
      "learning_rate": 8e-05,
      "loss": 1.8482,
      "step": 2505
    },
    {
      "epoch": 0.27083108181130444,
      "grad_norm": 0.42899563908576965,
      "learning_rate": 8e-05,
      "loss": 1.7687,
      "step": 2506
    },
    {
      "epoch": 0.27093915486869125,
      "grad_norm": 0.4219948649406433,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 2507
    },
    {
      "epoch": 0.27104722792607805,
      "grad_norm": 0.3983110785484314,
      "learning_rate": 8e-05,
      "loss": 1.5448,
      "step": 2508
    },
    {
      "epoch": 0.2711553009834648,
      "grad_norm": 0.39553090929985046,
      "learning_rate": 8e-05,
      "loss": 1.6553,
      "step": 2509
    },
    {
      "epoch": 0.2712633740408516,
      "grad_norm": 0.4068582057952881,
      "learning_rate": 8e-05,
      "loss": 1.6113,
      "step": 2510
    },
    {
      "epoch": 0.2713714470982384,
      "grad_norm": 0.43375200033187866,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 2511
    },
    {
      "epoch": 0.2714795201556252,
      "grad_norm": 0.3985610604286194,
      "learning_rate": 8e-05,
      "loss": 1.7513,
      "step": 2512
    },
    {
      "epoch": 0.271587593213012,
      "grad_norm": 0.39924851059913635,
      "learning_rate": 8e-05,
      "loss": 1.5851,
      "step": 2513
    },
    {
      "epoch": 0.2716956662703988,
      "grad_norm": 0.4428696930408478,
      "learning_rate": 8e-05,
      "loss": 1.8266,
      "step": 2514
    },
    {
      "epoch": 0.27180373932778557,
      "grad_norm": 0.413267582654953,
      "learning_rate": 8e-05,
      "loss": 1.8538,
      "step": 2515
    },
    {
      "epoch": 0.27191181238517237,
      "grad_norm": 0.4045692980289459,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 2516
    },
    {
      "epoch": 0.2720198854425592,
      "grad_norm": 0.3988894820213318,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 2517
    },
    {
      "epoch": 0.272127958499946,
      "grad_norm": 0.4966161251068115,
      "learning_rate": 8e-05,
      "loss": 2.0964,
      "step": 2518
    },
    {
      "epoch": 0.2722360315573328,
      "grad_norm": 0.3914951682090759,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 2519
    },
    {
      "epoch": 0.27234410461471953,
      "grad_norm": 0.3956215977668762,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 2520
    },
    {
      "epoch": 0.27245217767210633,
      "grad_norm": 0.4265586733818054,
      "learning_rate": 8e-05,
      "loss": 1.8811,
      "step": 2521
    },
    {
      "epoch": 0.27256025072949314,
      "grad_norm": 0.42329099774360657,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 2522
    },
    {
      "epoch": 0.27266832378687994,
      "grad_norm": 0.37288519740104675,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 2523
    },
    {
      "epoch": 0.27277639684426674,
      "grad_norm": 0.4394083619117737,
      "learning_rate": 8e-05,
      "loss": 1.616,
      "step": 2524
    },
    {
      "epoch": 0.2728844699016535,
      "grad_norm": 0.43142348527908325,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 2525
    },
    {
      "epoch": 0.2729925429590403,
      "grad_norm": 0.43149372935295105,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 2526
    },
    {
      "epoch": 0.2731006160164271,
      "grad_norm": 0.408775269985199,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 2527
    },
    {
      "epoch": 0.2732086890738139,
      "grad_norm": 0.398120641708374,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 2528
    },
    {
      "epoch": 0.2733167621312007,
      "grad_norm": 0.4108511805534363,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 2529
    },
    {
      "epoch": 0.2734248351885875,
      "grad_norm": 0.4107043743133545,
      "learning_rate": 8e-05,
      "loss": 1.8044,
      "step": 2530
    },
    {
      "epoch": 0.27353290824597426,
      "grad_norm": 0.4478759169578552,
      "learning_rate": 8e-05,
      "loss": 1.8004,
      "step": 2531
    },
    {
      "epoch": 0.27364098130336106,
      "grad_norm": 0.393679678440094,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 2532
    },
    {
      "epoch": 0.27374905436074787,
      "grad_norm": 0.410746693611145,
      "learning_rate": 8e-05,
      "loss": 1.8286,
      "step": 2533
    },
    {
      "epoch": 0.27385712741813467,
      "grad_norm": 0.41027042269706726,
      "learning_rate": 8e-05,
      "loss": 1.7999,
      "step": 2534
    },
    {
      "epoch": 0.2739652004755215,
      "grad_norm": 0.39821162819862366,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 2535
    },
    {
      "epoch": 0.2740732735329082,
      "grad_norm": 0.39470362663269043,
      "learning_rate": 8e-05,
      "loss": 1.5181,
      "step": 2536
    },
    {
      "epoch": 0.274181346590295,
      "grad_norm": 0.42366161942481995,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 2537
    },
    {
      "epoch": 0.27428941964768183,
      "grad_norm": 0.4173755943775177,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 2538
    },
    {
      "epoch": 0.27439749270506864,
      "grad_norm": 0.4384559690952301,
      "learning_rate": 8e-05,
      "loss": 1.6933,
      "step": 2539
    },
    {
      "epoch": 0.27450556576245544,
      "grad_norm": 0.4002607464790344,
      "learning_rate": 8e-05,
      "loss": 1.5651,
      "step": 2540
    },
    {
      "epoch": 0.2746136388198422,
      "grad_norm": 0.4011465311050415,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 2541
    },
    {
      "epoch": 0.274721711877229,
      "grad_norm": 0.45403096079826355,
      "learning_rate": 8e-05,
      "loss": 1.9005,
      "step": 2542
    },
    {
      "epoch": 0.2748297849346158,
      "grad_norm": 0.44568508863449097,
      "learning_rate": 8e-05,
      "loss": 1.8943,
      "step": 2543
    },
    {
      "epoch": 0.2749378579920026,
      "grad_norm": 0.3821237087249756,
      "learning_rate": 8e-05,
      "loss": 1.7311,
      "step": 2544
    },
    {
      "epoch": 0.2750459310493894,
      "grad_norm": 0.4141777753829956,
      "learning_rate": 8e-05,
      "loss": 1.8507,
      "step": 2545
    },
    {
      "epoch": 0.2751540041067762,
      "grad_norm": 0.37729793787002563,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 2546
    },
    {
      "epoch": 0.27526207716416295,
      "grad_norm": 0.44752874970436096,
      "learning_rate": 8e-05,
      "loss": 1.7488,
      "step": 2547
    },
    {
      "epoch": 0.27537015022154976,
      "grad_norm": 0.3928857743740082,
      "learning_rate": 8e-05,
      "loss": 1.651,
      "step": 2548
    },
    {
      "epoch": 0.27547822327893656,
      "grad_norm": 0.4284208118915558,
      "learning_rate": 8e-05,
      "loss": 1.83,
      "step": 2549
    },
    {
      "epoch": 0.27558629633632337,
      "grad_norm": 0.3888256549835205,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 2550
    },
    {
      "epoch": 0.27569436939371017,
      "grad_norm": 0.432528555393219,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 2551
    },
    {
      "epoch": 0.2758024424510969,
      "grad_norm": 0.43047648668289185,
      "learning_rate": 8e-05,
      "loss": 1.8892,
      "step": 2552
    },
    {
      "epoch": 0.2759105155084837,
      "grad_norm": 0.4089474678039551,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 2553
    },
    {
      "epoch": 0.2760185885658705,
      "grad_norm": 0.4205721616744995,
      "learning_rate": 8e-05,
      "loss": 1.7483,
      "step": 2554
    },
    {
      "epoch": 0.27612666162325733,
      "grad_norm": 0.4499327540397644,
      "learning_rate": 8e-05,
      "loss": 1.8528,
      "step": 2555
    },
    {
      "epoch": 0.27623473468064413,
      "grad_norm": 0.46635958552360535,
      "learning_rate": 8e-05,
      "loss": 1.8322,
      "step": 2556
    },
    {
      "epoch": 0.2763428077380309,
      "grad_norm": 0.39498627185821533,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 2557
    },
    {
      "epoch": 0.2764508807954177,
      "grad_norm": 0.39723965525627136,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 2558
    },
    {
      "epoch": 0.2765589538528045,
      "grad_norm": 0.482094943523407,
      "learning_rate": 8e-05,
      "loss": 1.8672,
      "step": 2559
    },
    {
      "epoch": 0.2766670269101913,
      "grad_norm": 0.5061137676239014,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 2560
    },
    {
      "epoch": 0.2767750999675781,
      "grad_norm": 0.49184197187423706,
      "learning_rate": 8e-05,
      "loss": 1.9019,
      "step": 2561
    },
    {
      "epoch": 0.2768831730249649,
      "grad_norm": 0.3949839174747467,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 2562
    },
    {
      "epoch": 0.27699124608235165,
      "grad_norm": 0.45889267325401306,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 2563
    },
    {
      "epoch": 0.27709931913973845,
      "grad_norm": 0.48324844241142273,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 2564
    },
    {
      "epoch": 0.27720739219712526,
      "grad_norm": 0.4245380163192749,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 2565
    },
    {
      "epoch": 0.27731546525451206,
      "grad_norm": 0.43598639965057373,
      "learning_rate": 8e-05,
      "loss": 1.8718,
      "step": 2566
    },
    {
      "epoch": 0.27742353831189887,
      "grad_norm": 0.48248490691185,
      "learning_rate": 8e-05,
      "loss": 1.9361,
      "step": 2567
    },
    {
      "epoch": 0.2775316113692856,
      "grad_norm": 0.45552703738212585,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 2568
    },
    {
      "epoch": 0.2776396844266724,
      "grad_norm": 0.45249608159065247,
      "learning_rate": 8e-05,
      "loss": 1.8458,
      "step": 2569
    },
    {
      "epoch": 0.2777477574840592,
      "grad_norm": 0.4165706932544708,
      "learning_rate": 8e-05,
      "loss": 1.7417,
      "step": 2570
    },
    {
      "epoch": 0.277855830541446,
      "grad_norm": 0.45733922719955444,
      "learning_rate": 8e-05,
      "loss": 1.7662,
      "step": 2571
    },
    {
      "epoch": 0.27796390359883283,
      "grad_norm": 0.4043046236038208,
      "learning_rate": 8e-05,
      "loss": 1.4815,
      "step": 2572
    },
    {
      "epoch": 0.2780719766562196,
      "grad_norm": 0.40772545337677,
      "learning_rate": 8e-05,
      "loss": 1.7084,
      "step": 2573
    },
    {
      "epoch": 0.2781800497136064,
      "grad_norm": 0.4113202691078186,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 2574
    },
    {
      "epoch": 0.2782881227709932,
      "grad_norm": 0.45793914794921875,
      "learning_rate": 8e-05,
      "loss": 1.8868,
      "step": 2575
    },
    {
      "epoch": 0.27839619582838,
      "grad_norm": 0.4352041780948639,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 2576
    },
    {
      "epoch": 0.2785042688857668,
      "grad_norm": 0.41779395937919617,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 2577
    },
    {
      "epoch": 0.2786123419431536,
      "grad_norm": 0.4109345078468323,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 2578
    },
    {
      "epoch": 0.27872041500054034,
      "grad_norm": 0.44685643911361694,
      "learning_rate": 8e-05,
      "loss": 1.8827,
      "step": 2579
    },
    {
      "epoch": 0.27882848805792715,
      "grad_norm": 0.43174901604652405,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 2580
    },
    {
      "epoch": 0.27893656111531395,
      "grad_norm": 0.43596211075782776,
      "learning_rate": 8e-05,
      "loss": 1.7966,
      "step": 2581
    },
    {
      "epoch": 0.27904463417270076,
      "grad_norm": 0.4405350983142853,
      "learning_rate": 8e-05,
      "loss": 2.0004,
      "step": 2582
    },
    {
      "epoch": 0.27915270723008756,
      "grad_norm": 0.4111610949039459,
      "learning_rate": 8e-05,
      "loss": 1.7682,
      "step": 2583
    },
    {
      "epoch": 0.2792607802874743,
      "grad_norm": 0.41903555393218994,
      "learning_rate": 8e-05,
      "loss": 1.8689,
      "step": 2584
    },
    {
      "epoch": 0.2793688533448611,
      "grad_norm": 0.44346681237220764,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 2585
    },
    {
      "epoch": 0.2794769264022479,
      "grad_norm": 0.4203447997570038,
      "learning_rate": 8e-05,
      "loss": 1.6215,
      "step": 2586
    },
    {
      "epoch": 0.2795849994596347,
      "grad_norm": 0.45564162731170654,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 2587
    },
    {
      "epoch": 0.2796930725170215,
      "grad_norm": 0.4261377155780792,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 2588
    },
    {
      "epoch": 0.2798011455744083,
      "grad_norm": 0.37291792035102844,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 2589
    },
    {
      "epoch": 0.2799092186317951,
      "grad_norm": 0.41625186800956726,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 2590
    },
    {
      "epoch": 0.2800172916891819,
      "grad_norm": 0.4086818993091583,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 2591
    },
    {
      "epoch": 0.2801253647465687,
      "grad_norm": 0.4203695058822632,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 2592
    },
    {
      "epoch": 0.2802334378039555,
      "grad_norm": 0.4166378080844879,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 2593
    },
    {
      "epoch": 0.2803415108613423,
      "grad_norm": 0.4331427216529846,
      "learning_rate": 8e-05,
      "loss": 1.8432,
      "step": 2594
    },
    {
      "epoch": 0.28044958391872904,
      "grad_norm": 0.4097088873386383,
      "learning_rate": 8e-05,
      "loss": 1.6169,
      "step": 2595
    },
    {
      "epoch": 0.28055765697611584,
      "grad_norm": 0.4232092499732971,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 2596
    },
    {
      "epoch": 0.28066573003350265,
      "grad_norm": 0.43558764457702637,
      "learning_rate": 8e-05,
      "loss": 1.8586,
      "step": 2597
    },
    {
      "epoch": 0.28077380309088945,
      "grad_norm": 0.40416523814201355,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 2598
    },
    {
      "epoch": 0.28088187614827625,
      "grad_norm": 0.4078659117221832,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 2599
    },
    {
      "epoch": 0.280989949205663,
      "grad_norm": 0.40268275141716003,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 2600
    },
    {
      "epoch": 0.2810980222630498,
      "grad_norm": 0.4392794370651245,
      "learning_rate": 8e-05,
      "loss": 1.8354,
      "step": 2601
    },
    {
      "epoch": 0.2812060953204366,
      "grad_norm": 0.47986552119255066,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 2602
    },
    {
      "epoch": 0.2813141683778234,
      "grad_norm": 0.4133641719818115,
      "learning_rate": 8e-05,
      "loss": 1.8499,
      "step": 2603
    },
    {
      "epoch": 0.2814222414352102,
      "grad_norm": 0.4388870298862457,
      "learning_rate": 8e-05,
      "loss": 1.8069,
      "step": 2604
    },
    {
      "epoch": 0.281530314492597,
      "grad_norm": 0.43099692463874817,
      "learning_rate": 8e-05,
      "loss": 1.894,
      "step": 2605
    },
    {
      "epoch": 0.28163838754998377,
      "grad_norm": 0.4233704209327698,
      "learning_rate": 8e-05,
      "loss": 1.5506,
      "step": 2606
    },
    {
      "epoch": 0.2817464606073706,
      "grad_norm": 0.43325451016426086,
      "learning_rate": 8e-05,
      "loss": 1.8765,
      "step": 2607
    },
    {
      "epoch": 0.2818545336647574,
      "grad_norm": 0.40693485736846924,
      "learning_rate": 8e-05,
      "loss": 1.8751,
      "step": 2608
    },
    {
      "epoch": 0.2819626067221442,
      "grad_norm": 0.39769694209098816,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 2609
    },
    {
      "epoch": 0.282070679779531,
      "grad_norm": 0.4136699140071869,
      "learning_rate": 8e-05,
      "loss": 1.6205,
      "step": 2610
    },
    {
      "epoch": 0.28217875283691773,
      "grad_norm": 0.39564478397369385,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 2611
    },
    {
      "epoch": 0.28228682589430454,
      "grad_norm": 0.4290805160999298,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 2612
    },
    {
      "epoch": 0.28239489895169134,
      "grad_norm": 0.45722201466560364,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 2613
    },
    {
      "epoch": 0.28250297200907815,
      "grad_norm": 0.4488467276096344,
      "learning_rate": 8e-05,
      "loss": 1.9101,
      "step": 2614
    },
    {
      "epoch": 0.28261104506646495,
      "grad_norm": 0.4266910254955292,
      "learning_rate": 8e-05,
      "loss": 1.8692,
      "step": 2615
    },
    {
      "epoch": 0.2827191181238517,
      "grad_norm": 0.43244487047195435,
      "learning_rate": 8e-05,
      "loss": 1.7823,
      "step": 2616
    },
    {
      "epoch": 0.2828271911812385,
      "grad_norm": 0.4141051769256592,
      "learning_rate": 8e-05,
      "loss": 1.6034,
      "step": 2617
    },
    {
      "epoch": 0.2829352642386253,
      "grad_norm": 0.4533381462097168,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 2618
    },
    {
      "epoch": 0.2830433372960121,
      "grad_norm": 0.4231864809989929,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 2619
    },
    {
      "epoch": 0.2831514103533989,
      "grad_norm": 0.40418967604637146,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 2620
    },
    {
      "epoch": 0.2832594834107857,
      "grad_norm": 0.432481586933136,
      "learning_rate": 8e-05,
      "loss": 1.571,
      "step": 2621
    },
    {
      "epoch": 0.28336755646817247,
      "grad_norm": 0.44598978757858276,
      "learning_rate": 8e-05,
      "loss": 1.7871,
      "step": 2622
    },
    {
      "epoch": 0.28347562952555927,
      "grad_norm": 0.3994063138961792,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 2623
    },
    {
      "epoch": 0.2835837025829461,
      "grad_norm": 0.45349523425102234,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 2624
    },
    {
      "epoch": 0.2836917756403329,
      "grad_norm": 0.46148356795310974,
      "learning_rate": 8e-05,
      "loss": 1.8276,
      "step": 2625
    },
    {
      "epoch": 0.2837998486977197,
      "grad_norm": 0.4497140347957611,
      "learning_rate": 8e-05,
      "loss": 1.8631,
      "step": 2626
    },
    {
      "epoch": 0.28390792175510643,
      "grad_norm": 0.4230700731277466,
      "learning_rate": 8e-05,
      "loss": 1.7994,
      "step": 2627
    },
    {
      "epoch": 0.28401599481249323,
      "grad_norm": 0.43483421206474304,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 2628
    },
    {
      "epoch": 0.28412406786988004,
      "grad_norm": 0.3968863785266876,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 2629
    },
    {
      "epoch": 0.28423214092726684,
      "grad_norm": 0.40389323234558105,
      "learning_rate": 8e-05,
      "loss": 1.6492,
      "step": 2630
    },
    {
      "epoch": 0.28434021398465364,
      "grad_norm": 0.4372112452983856,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 2631
    },
    {
      "epoch": 0.2844482870420404,
      "grad_norm": 0.4157763421535492,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 2632
    },
    {
      "epoch": 0.2845563600994272,
      "grad_norm": 0.4247700572013855,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 2633
    },
    {
      "epoch": 0.284664433156814,
      "grad_norm": 0.42127174139022827,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 2634
    },
    {
      "epoch": 0.2847725062142008,
      "grad_norm": 0.3880135118961334,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 2635
    },
    {
      "epoch": 0.2848805792715876,
      "grad_norm": 0.44634732604026794,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 2636
    },
    {
      "epoch": 0.2849886523289744,
      "grad_norm": 0.45084649324417114,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 2637
    },
    {
      "epoch": 0.28509672538636116,
      "grad_norm": 0.3923322260379791,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 2638
    },
    {
      "epoch": 0.28520479844374796,
      "grad_norm": 0.44180211424827576,
      "learning_rate": 8e-05,
      "loss": 1.9387,
      "step": 2639
    },
    {
      "epoch": 0.28531287150113477,
      "grad_norm": 0.3870996832847595,
      "learning_rate": 8e-05,
      "loss": 1.5748,
      "step": 2640
    },
    {
      "epoch": 0.28542094455852157,
      "grad_norm": 0.3972896933555603,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 2641
    },
    {
      "epoch": 0.2855290176159084,
      "grad_norm": 0.4453761875629425,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 2642
    },
    {
      "epoch": 0.2856370906732951,
      "grad_norm": 0.413048654794693,
      "learning_rate": 8e-05,
      "loss": 1.8625,
      "step": 2643
    },
    {
      "epoch": 0.28574516373068193,
      "grad_norm": 0.4676482379436493,
      "learning_rate": 8e-05,
      "loss": 2.0415,
      "step": 2644
    },
    {
      "epoch": 0.28585323678806873,
      "grad_norm": 0.42422935366630554,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 2645
    },
    {
      "epoch": 0.28596130984545554,
      "grad_norm": 0.42110303044319153,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 2646
    },
    {
      "epoch": 0.28606938290284234,
      "grad_norm": 0.437878280878067,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 2647
    },
    {
      "epoch": 0.2861774559602291,
      "grad_norm": 0.42526158690452576,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 2648
    },
    {
      "epoch": 0.2862855290176159,
      "grad_norm": 0.4351373612880707,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 2649
    },
    {
      "epoch": 0.2863936020750027,
      "grad_norm": 0.4095854163169861,
      "learning_rate": 8e-05,
      "loss": 1.6571,
      "step": 2650
    },
    {
      "epoch": 0.2865016751323895,
      "grad_norm": 0.41261667013168335,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 2651
    },
    {
      "epoch": 0.2866097481897763,
      "grad_norm": 0.4083942770957947,
      "learning_rate": 8e-05,
      "loss": 1.8502,
      "step": 2652
    },
    {
      "epoch": 0.2867178212471631,
      "grad_norm": 0.40233129262924194,
      "learning_rate": 8e-05,
      "loss": 1.6876,
      "step": 2653
    },
    {
      "epoch": 0.28682589430454986,
      "grad_norm": 0.41242367029190063,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 2654
    },
    {
      "epoch": 0.28693396736193666,
      "grad_norm": 0.44550609588623047,
      "learning_rate": 8e-05,
      "loss": 1.5811,
      "step": 2655
    },
    {
      "epoch": 0.28704204041932346,
      "grad_norm": 0.3940154016017914,
      "learning_rate": 8e-05,
      "loss": 1.7092,
      "step": 2656
    },
    {
      "epoch": 0.28715011347671027,
      "grad_norm": 0.4076324999332428,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 2657
    },
    {
      "epoch": 0.28725818653409707,
      "grad_norm": 0.4387036859989166,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 2658
    },
    {
      "epoch": 0.2873662595914838,
      "grad_norm": 0.4345681667327881,
      "learning_rate": 8e-05,
      "loss": 1.9093,
      "step": 2659
    },
    {
      "epoch": 0.2874743326488706,
      "grad_norm": 0.46855491399765015,
      "learning_rate": 8e-05,
      "loss": 1.9176,
      "step": 2660
    },
    {
      "epoch": 0.2875824057062574,
      "grad_norm": 0.4813562333583832,
      "learning_rate": 8e-05,
      "loss": 2.047,
      "step": 2661
    },
    {
      "epoch": 0.28769047876364423,
      "grad_norm": 0.43084409832954407,
      "learning_rate": 8e-05,
      "loss": 1.7919,
      "step": 2662
    },
    {
      "epoch": 0.28779855182103103,
      "grad_norm": 0.45375123620033264,
      "learning_rate": 8e-05,
      "loss": 1.6621,
      "step": 2663
    },
    {
      "epoch": 0.2879066248784178,
      "grad_norm": 0.41913649439811707,
      "learning_rate": 8e-05,
      "loss": 1.8618,
      "step": 2664
    },
    {
      "epoch": 0.2880146979358046,
      "grad_norm": 0.44569680094718933,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 2665
    },
    {
      "epoch": 0.2881227709931914,
      "grad_norm": 0.4128209352493286,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 2666
    },
    {
      "epoch": 0.2882308440505782,
      "grad_norm": 0.4183307886123657,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 2667
    },
    {
      "epoch": 0.288338917107965,
      "grad_norm": 0.4090079963207245,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 2668
    },
    {
      "epoch": 0.2884469901653518,
      "grad_norm": 0.3991641104221344,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 2669
    },
    {
      "epoch": 0.28855506322273855,
      "grad_norm": 0.45228347182273865,
      "learning_rate": 8e-05,
      "loss": 1.9403,
      "step": 2670
    },
    {
      "epoch": 0.28866313628012535,
      "grad_norm": 0.49158990383148193,
      "learning_rate": 8e-05,
      "loss": 1.9571,
      "step": 2671
    },
    {
      "epoch": 0.28877120933751216,
      "grad_norm": 0.4102950096130371,
      "learning_rate": 8e-05,
      "loss": 1.7943,
      "step": 2672
    },
    {
      "epoch": 0.28887928239489896,
      "grad_norm": 0.4324445426464081,
      "learning_rate": 8e-05,
      "loss": 1.9612,
      "step": 2673
    },
    {
      "epoch": 0.28898735545228577,
      "grad_norm": 0.44105595350265503,
      "learning_rate": 8e-05,
      "loss": 1.7952,
      "step": 2674
    },
    {
      "epoch": 0.2890954285096725,
      "grad_norm": 0.3939342200756073,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 2675
    },
    {
      "epoch": 0.2892035015670593,
      "grad_norm": 0.41624149680137634,
      "learning_rate": 8e-05,
      "loss": 1.8024,
      "step": 2676
    },
    {
      "epoch": 0.2893115746244461,
      "grad_norm": 0.40756556391716003,
      "learning_rate": 8e-05,
      "loss": 1.8868,
      "step": 2677
    },
    {
      "epoch": 0.2894196476818329,
      "grad_norm": 0.40334221720695496,
      "learning_rate": 8e-05,
      "loss": 1.4755,
      "step": 2678
    },
    {
      "epoch": 0.28952772073921973,
      "grad_norm": 0.4137203097343445,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 2679
    },
    {
      "epoch": 0.28963579379660653,
      "grad_norm": 0.40662693977355957,
      "learning_rate": 8e-05,
      "loss": 1.7207,
      "step": 2680
    },
    {
      "epoch": 0.2897438668539933,
      "grad_norm": 0.4186086356639862,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 2681
    },
    {
      "epoch": 0.2898519399113801,
      "grad_norm": 0.3902864456176758,
      "learning_rate": 8e-05,
      "loss": 1.7285,
      "step": 2682
    },
    {
      "epoch": 0.2899600129687669,
      "grad_norm": 0.4270440638065338,
      "learning_rate": 8e-05,
      "loss": 1.8031,
      "step": 2683
    },
    {
      "epoch": 0.2900680860261537,
      "grad_norm": 0.43234196305274963,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 2684
    },
    {
      "epoch": 0.2901761590835405,
      "grad_norm": 0.42159348726272583,
      "learning_rate": 8e-05,
      "loss": 1.8793,
      "step": 2685
    },
    {
      "epoch": 0.29028423214092725,
      "grad_norm": 0.41953572630882263,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 2686
    },
    {
      "epoch": 0.29039230519831405,
      "grad_norm": 0.41235029697418213,
      "learning_rate": 8e-05,
      "loss": 1.582,
      "step": 2687
    },
    {
      "epoch": 0.29050037825570085,
      "grad_norm": 0.4163140058517456,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 2688
    },
    {
      "epoch": 0.29060845131308766,
      "grad_norm": 0.44079524278640747,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 2689
    },
    {
      "epoch": 0.29071652437047446,
      "grad_norm": 0.44385239481925964,
      "learning_rate": 8e-05,
      "loss": 1.7122,
      "step": 2690
    },
    {
      "epoch": 0.2908245974278612,
      "grad_norm": 0.43547478318214417,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 2691
    },
    {
      "epoch": 0.290932670485248,
      "grad_norm": 0.4902149736881256,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 2692
    },
    {
      "epoch": 0.2910407435426348,
      "grad_norm": 0.424389511346817,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 2693
    },
    {
      "epoch": 0.2911488166000216,
      "grad_norm": 0.4626428186893463,
      "learning_rate": 8e-05,
      "loss": 1.8848,
      "step": 2694
    },
    {
      "epoch": 0.2912568896574084,
      "grad_norm": 0.40812188386917114,
      "learning_rate": 8e-05,
      "loss": 1.848,
      "step": 2695
    },
    {
      "epoch": 0.29136496271479523,
      "grad_norm": 0.43489310145378113,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 2696
    },
    {
      "epoch": 0.291473035772182,
      "grad_norm": 0.476345032453537,
      "learning_rate": 8e-05,
      "loss": 1.8693,
      "step": 2697
    },
    {
      "epoch": 0.2915811088295688,
      "grad_norm": 0.38556936383247375,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 2698
    },
    {
      "epoch": 0.2916891818869556,
      "grad_norm": 0.4237300455570221,
      "learning_rate": 8e-05,
      "loss": 1.8904,
      "step": 2699
    },
    {
      "epoch": 0.2917972549443424,
      "grad_norm": 0.47219276428222656,
      "learning_rate": 8e-05,
      "loss": 1.775,
      "step": 2700
    },
    {
      "epoch": 0.2919053280017292,
      "grad_norm": 0.4487742781639099,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 2701
    },
    {
      "epoch": 0.29201340105911594,
      "grad_norm": 0.46204251050949097,
      "learning_rate": 8e-05,
      "loss": 1.8553,
      "step": 2702
    },
    {
      "epoch": 0.29212147411650274,
      "grad_norm": 0.41393131017684937,
      "learning_rate": 8e-05,
      "loss": 1.6608,
      "step": 2703
    },
    {
      "epoch": 0.29222954717388955,
      "grad_norm": 0.4223105013370514,
      "learning_rate": 8e-05,
      "loss": 1.8232,
      "step": 2704
    },
    {
      "epoch": 0.29233762023127635,
      "grad_norm": 0.382004052400589,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 2705
    },
    {
      "epoch": 0.29244569328866316,
      "grad_norm": 0.4383132755756378,
      "learning_rate": 8e-05,
      "loss": 1.8261,
      "step": 2706
    },
    {
      "epoch": 0.2925537663460499,
      "grad_norm": 0.43705326318740845,
      "learning_rate": 8e-05,
      "loss": 1.6717,
      "step": 2707
    },
    {
      "epoch": 0.2926618394034367,
      "grad_norm": 0.4731813073158264,
      "learning_rate": 8e-05,
      "loss": 1.8386,
      "step": 2708
    },
    {
      "epoch": 0.2927699124608235,
      "grad_norm": 0.4198560416698456,
      "learning_rate": 8e-05,
      "loss": 1.593,
      "step": 2709
    },
    {
      "epoch": 0.2928779855182103,
      "grad_norm": 0.3872883915901184,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 2710
    },
    {
      "epoch": 0.2929860585755971,
      "grad_norm": 0.4473097026348114,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 2711
    },
    {
      "epoch": 0.2930941316329839,
      "grad_norm": 0.427852600812912,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 2712
    },
    {
      "epoch": 0.29320220469037067,
      "grad_norm": 0.42789483070373535,
      "learning_rate": 8e-05,
      "loss": 1.9307,
      "step": 2713
    },
    {
      "epoch": 0.2933102777477575,
      "grad_norm": 0.44084006547927856,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 2714
    },
    {
      "epoch": 0.2934183508051443,
      "grad_norm": 0.3870580792427063,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 2715
    },
    {
      "epoch": 0.2935264238625311,
      "grad_norm": 0.403288871049881,
      "learning_rate": 8e-05,
      "loss": 1.7169,
      "step": 2716
    },
    {
      "epoch": 0.2936344969199179,
      "grad_norm": 0.3895527720451355,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 2717
    },
    {
      "epoch": 0.29374256997730464,
      "grad_norm": 0.3983418047428131,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 2718
    },
    {
      "epoch": 0.29385064303469144,
      "grad_norm": 0.41533082723617554,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 2719
    },
    {
      "epoch": 0.29395871609207824,
      "grad_norm": 0.44997069239616394,
      "learning_rate": 8e-05,
      "loss": 1.8967,
      "step": 2720
    },
    {
      "epoch": 0.29406678914946505,
      "grad_norm": 0.43294957280158997,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 2721
    },
    {
      "epoch": 0.29417486220685185,
      "grad_norm": 0.3927350640296936,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 2722
    },
    {
      "epoch": 0.2942829352642386,
      "grad_norm": 0.39901843667030334,
      "learning_rate": 8e-05,
      "loss": 1.7072,
      "step": 2723
    },
    {
      "epoch": 0.2943910083216254,
      "grad_norm": 0.40084460377693176,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 2724
    },
    {
      "epoch": 0.2944990813790122,
      "grad_norm": 0.42582255601882935,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 2725
    },
    {
      "epoch": 0.294607154436399,
      "grad_norm": 0.3913150429725647,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 2726
    },
    {
      "epoch": 0.2947152274937858,
      "grad_norm": 0.4502480924129486,
      "learning_rate": 8e-05,
      "loss": 1.8629,
      "step": 2727
    },
    {
      "epoch": 0.2948233005511726,
      "grad_norm": 0.40841928124427795,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 2728
    },
    {
      "epoch": 0.29493137360855937,
      "grad_norm": 0.3975891172885895,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 2729
    },
    {
      "epoch": 0.29503944666594617,
      "grad_norm": 0.39112409949302673,
      "learning_rate": 8e-05,
      "loss": 1.6278,
      "step": 2730
    },
    {
      "epoch": 0.295147519723333,
      "grad_norm": 0.4311474561691284,
      "learning_rate": 8e-05,
      "loss": 1.7403,
      "step": 2731
    },
    {
      "epoch": 0.2952555927807198,
      "grad_norm": 0.39621272683143616,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 2732
    },
    {
      "epoch": 0.2953636658381066,
      "grad_norm": 0.4121188223361969,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 2733
    },
    {
      "epoch": 0.29547173889549333,
      "grad_norm": 0.4411282539367676,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 2734
    },
    {
      "epoch": 0.29557981195288013,
      "grad_norm": 0.4187590479850769,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 2735
    },
    {
      "epoch": 0.29568788501026694,
      "grad_norm": 0.442374587059021,
      "learning_rate": 8e-05,
      "loss": 1.8104,
      "step": 2736
    },
    {
      "epoch": 0.29579595806765374,
      "grad_norm": 0.3992833197116852,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 2737
    },
    {
      "epoch": 0.29590403112504055,
      "grad_norm": 0.4301396608352661,
      "learning_rate": 8e-05,
      "loss": 1.994,
      "step": 2738
    },
    {
      "epoch": 0.2960121041824273,
      "grad_norm": 0.48271286487579346,
      "learning_rate": 8e-05,
      "loss": 2.0146,
      "step": 2739
    },
    {
      "epoch": 0.2961201772398141,
      "grad_norm": 0.442011296749115,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 2740
    },
    {
      "epoch": 0.2962282502972009,
      "grad_norm": 0.4715065062046051,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 2741
    },
    {
      "epoch": 0.2963363233545877,
      "grad_norm": 0.4178426265716553,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 2742
    },
    {
      "epoch": 0.2964443964119745,
      "grad_norm": 0.4351113438606262,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 2743
    },
    {
      "epoch": 0.2965524694693613,
      "grad_norm": 0.4375176429748535,
      "learning_rate": 8e-05,
      "loss": 1.9057,
      "step": 2744
    },
    {
      "epoch": 0.29666054252674806,
      "grad_norm": 0.40137892961502075,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 2745
    },
    {
      "epoch": 0.29676861558413486,
      "grad_norm": 0.47238534688949585,
      "learning_rate": 8e-05,
      "loss": 1.8875,
      "step": 2746
    },
    {
      "epoch": 0.29687668864152167,
      "grad_norm": 0.4001219868659973,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 2747
    },
    {
      "epoch": 0.2969847616989085,
      "grad_norm": 0.4148417115211487,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 2748
    },
    {
      "epoch": 0.2970928347562953,
      "grad_norm": 0.43985894322395325,
      "learning_rate": 8e-05,
      "loss": 1.8562,
      "step": 2749
    },
    {
      "epoch": 0.297200907813682,
      "grad_norm": 0.4560765326023102,
      "learning_rate": 8e-05,
      "loss": 1.9788,
      "step": 2750
    },
    {
      "epoch": 0.29730898087106883,
      "grad_norm": 0.40677377581596375,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 2751
    },
    {
      "epoch": 0.29741705392845563,
      "grad_norm": 0.4125247597694397,
      "learning_rate": 8e-05,
      "loss": 1.8109,
      "step": 2752
    },
    {
      "epoch": 0.29752512698584244,
      "grad_norm": 0.43198519945144653,
      "learning_rate": 8e-05,
      "loss": 1.9059,
      "step": 2753
    },
    {
      "epoch": 0.29763320004322924,
      "grad_norm": 0.4434976577758789,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 2754
    },
    {
      "epoch": 0.29774127310061604,
      "grad_norm": 0.41696038842201233,
      "learning_rate": 8e-05,
      "loss": 1.8667,
      "step": 2755
    },
    {
      "epoch": 0.2978493461580028,
      "grad_norm": 0.420543372631073,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 2756
    },
    {
      "epoch": 0.2979574192153896,
      "grad_norm": 0.44601401686668396,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 2757
    },
    {
      "epoch": 0.2980654922727764,
      "grad_norm": 0.40696612000465393,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 2758
    },
    {
      "epoch": 0.2981735653301632,
      "grad_norm": 0.44170981645584106,
      "learning_rate": 8e-05,
      "loss": 1.7285,
      "step": 2759
    },
    {
      "epoch": 0.29828163838755,
      "grad_norm": 0.41703689098358154,
      "learning_rate": 8e-05,
      "loss": 1.8023,
      "step": 2760
    },
    {
      "epoch": 0.29838971144493676,
      "grad_norm": 0.4410800039768219,
      "learning_rate": 8e-05,
      "loss": 1.8775,
      "step": 2761
    },
    {
      "epoch": 0.29849778450232356,
      "grad_norm": 0.4132538139820099,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 2762
    },
    {
      "epoch": 0.29860585755971036,
      "grad_norm": 0.3803849518299103,
      "learning_rate": 8e-05,
      "loss": 1.6628,
      "step": 2763
    },
    {
      "epoch": 0.29871393061709717,
      "grad_norm": 0.46084192395210266,
      "learning_rate": 8e-05,
      "loss": 1.6086,
      "step": 2764
    },
    {
      "epoch": 0.29882200367448397,
      "grad_norm": 0.38871505856513977,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 2765
    },
    {
      "epoch": 0.2989300767318707,
      "grad_norm": 0.4175260663032532,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 2766
    },
    {
      "epoch": 0.2990381497892575,
      "grad_norm": 0.4131663143634796,
      "learning_rate": 8e-05,
      "loss": 1.7036,
      "step": 2767
    },
    {
      "epoch": 0.2991462228466443,
      "grad_norm": 0.4121876060962677,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 2768
    },
    {
      "epoch": 0.29925429590403113,
      "grad_norm": 0.4035332202911377,
      "learning_rate": 8e-05,
      "loss": 1.7834,
      "step": 2769
    },
    {
      "epoch": 0.29936236896141794,
      "grad_norm": 0.4183337688446045,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 2770
    },
    {
      "epoch": 0.29947044201880474,
      "grad_norm": 0.5087496638298035,
      "learning_rate": 8e-05,
      "loss": 1.7726,
      "step": 2771
    },
    {
      "epoch": 0.2995785150761915,
      "grad_norm": 0.42605289816856384,
      "learning_rate": 8e-05,
      "loss": 1.7945,
      "step": 2772
    },
    {
      "epoch": 0.2996865881335783,
      "grad_norm": 0.4138619899749756,
      "learning_rate": 8e-05,
      "loss": 1.7379,
      "step": 2773
    },
    {
      "epoch": 0.2997946611909651,
      "grad_norm": 0.40151774883270264,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 2774
    },
    {
      "epoch": 0.2999027342483519,
      "grad_norm": 0.42041271924972534,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 2775
    },
    {
      "epoch": 0.3000108073057387,
      "grad_norm": 0.5054858922958374,
      "learning_rate": 8e-05,
      "loss": 2.0121,
      "step": 2776
    },
    {
      "epoch": 0.30011888036312545,
      "grad_norm": 0.46087968349456787,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 2777
    },
    {
      "epoch": 0.30022695342051225,
      "grad_norm": 0.5071647763252258,
      "learning_rate": 8e-05,
      "loss": 1.8784,
      "step": 2778
    },
    {
      "epoch": 0.30033502647789906,
      "grad_norm": 0.4668528437614441,
      "learning_rate": 8e-05,
      "loss": 1.8716,
      "step": 2779
    },
    {
      "epoch": 0.30044309953528586,
      "grad_norm": 0.41213715076446533,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 2780
    },
    {
      "epoch": 0.30055117259267267,
      "grad_norm": 0.38157880306243896,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 2781
    },
    {
      "epoch": 0.3006592456500594,
      "grad_norm": 0.4388939440250397,
      "learning_rate": 8e-05,
      "loss": 1.9329,
      "step": 2782
    },
    {
      "epoch": 0.3007673187074462,
      "grad_norm": 0.403331458568573,
      "learning_rate": 8e-05,
      "loss": 1.6124,
      "step": 2783
    },
    {
      "epoch": 0.300875391764833,
      "grad_norm": 0.4186929166316986,
      "learning_rate": 8e-05,
      "loss": 1.8364,
      "step": 2784
    },
    {
      "epoch": 0.3009834648222198,
      "grad_norm": 0.42931273579597473,
      "learning_rate": 8e-05,
      "loss": 1.8342,
      "step": 2785
    },
    {
      "epoch": 0.30109153787960663,
      "grad_norm": 0.40989255905151367,
      "learning_rate": 8e-05,
      "loss": 1.6131,
      "step": 2786
    },
    {
      "epoch": 0.30119961093699343,
      "grad_norm": 0.421018123626709,
      "learning_rate": 8e-05,
      "loss": 1.5218,
      "step": 2787
    },
    {
      "epoch": 0.3013076839943802,
      "grad_norm": 0.4028407335281372,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 2788
    },
    {
      "epoch": 0.301415757051767,
      "grad_norm": 0.39464160799980164,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 2789
    },
    {
      "epoch": 0.3015238301091538,
      "grad_norm": 0.41043978929519653,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 2790
    },
    {
      "epoch": 0.3016319031665406,
      "grad_norm": 0.43838074803352356,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 2791
    },
    {
      "epoch": 0.3017399762239274,
      "grad_norm": 0.4186965525150299,
      "learning_rate": 8e-05,
      "loss": 1.8426,
      "step": 2792
    },
    {
      "epoch": 0.30184804928131415,
      "grad_norm": 0.41501665115356445,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 2793
    },
    {
      "epoch": 0.30195612233870095,
      "grad_norm": 0.4248347282409668,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 2794
    },
    {
      "epoch": 0.30206419539608775,
      "grad_norm": 0.40848565101623535,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 2795
    },
    {
      "epoch": 0.30217226845347456,
      "grad_norm": 0.40484899282455444,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 2796
    },
    {
      "epoch": 0.30228034151086136,
      "grad_norm": 0.4032011330127716,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 2797
    },
    {
      "epoch": 0.3023884145682481,
      "grad_norm": 0.41208723187446594,
      "learning_rate": 8e-05,
      "loss": 1.796,
      "step": 2798
    },
    {
      "epoch": 0.3024964876256349,
      "grad_norm": 0.4223407506942749,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 2799
    },
    {
      "epoch": 0.3026045606830217,
      "grad_norm": 0.43031761050224304,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 2800
    },
    {
      "epoch": 0.3027126337404085,
      "grad_norm": 0.41539400815963745,
      "learning_rate": 8e-05,
      "loss": 1.7881,
      "step": 2801
    },
    {
      "epoch": 0.3028207067977953,
      "grad_norm": 0.45449548959732056,
      "learning_rate": 8e-05,
      "loss": 1.721,
      "step": 2802
    },
    {
      "epoch": 0.30292877985518213,
      "grad_norm": 0.4383353292942047,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 2803
    },
    {
      "epoch": 0.3030368529125689,
      "grad_norm": 0.4502035975456238,
      "learning_rate": 8e-05,
      "loss": 1.9415,
      "step": 2804
    },
    {
      "epoch": 0.3031449259699557,
      "grad_norm": 0.43714046478271484,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 2805
    },
    {
      "epoch": 0.3032529990273425,
      "grad_norm": 0.42778247594833374,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 2806
    },
    {
      "epoch": 0.3033610720847293,
      "grad_norm": 0.4090943932533264,
      "learning_rate": 8e-05,
      "loss": 1.6866,
      "step": 2807
    },
    {
      "epoch": 0.3034691451421161,
      "grad_norm": 0.4114381968975067,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 2808
    },
    {
      "epoch": 0.30357721819950284,
      "grad_norm": 0.4783458113670349,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 2809
    },
    {
      "epoch": 0.30368529125688964,
      "grad_norm": 0.42815500497817993,
      "learning_rate": 8e-05,
      "loss": 1.8127,
      "step": 2810
    },
    {
      "epoch": 0.30379336431427645,
      "grad_norm": 0.4532529413700104,
      "learning_rate": 8e-05,
      "loss": 1.9258,
      "step": 2811
    },
    {
      "epoch": 0.30390143737166325,
      "grad_norm": 0.4239599406719208,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 2812
    },
    {
      "epoch": 0.30400951042905006,
      "grad_norm": 0.41526493430137634,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 2813
    },
    {
      "epoch": 0.3041175834864368,
      "grad_norm": 0.4147811532020569,
      "learning_rate": 8e-05,
      "loss": 1.8296,
      "step": 2814
    },
    {
      "epoch": 0.3042256565438236,
      "grad_norm": 0.43341419100761414,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 2815
    },
    {
      "epoch": 0.3043337296012104,
      "grad_norm": 0.3984733521938324,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2816
    },
    {
      "epoch": 0.3044418026585972,
      "grad_norm": 0.4410656690597534,
      "learning_rate": 8e-05,
      "loss": 1.7737,
      "step": 2817
    },
    {
      "epoch": 0.304549875715984,
      "grad_norm": 0.4217013716697693,
      "learning_rate": 8e-05,
      "loss": 1.6643,
      "step": 2818
    },
    {
      "epoch": 0.3046579487733708,
      "grad_norm": 0.415360689163208,
      "learning_rate": 8e-05,
      "loss": 1.8239,
      "step": 2819
    },
    {
      "epoch": 0.30476602183075757,
      "grad_norm": 0.4012530744075775,
      "learning_rate": 8e-05,
      "loss": 1.7435,
      "step": 2820
    },
    {
      "epoch": 0.3048740948881444,
      "grad_norm": 0.4037805497646332,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 2821
    },
    {
      "epoch": 0.3049821679455312,
      "grad_norm": 0.39182204008102417,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 2822
    },
    {
      "epoch": 0.305090241002918,
      "grad_norm": 0.40487608313560486,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 2823
    },
    {
      "epoch": 0.3051983140603048,
      "grad_norm": 0.3899393081665039,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 2824
    },
    {
      "epoch": 0.30530638711769154,
      "grad_norm": 0.43534284830093384,
      "learning_rate": 8e-05,
      "loss": 1.8761,
      "step": 2825
    },
    {
      "epoch": 0.30541446017507834,
      "grad_norm": 0.40957513451576233,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 2826
    },
    {
      "epoch": 0.30552253323246514,
      "grad_norm": 0.41790148615837097,
      "learning_rate": 8e-05,
      "loss": 1.6395,
      "step": 2827
    },
    {
      "epoch": 0.30563060628985195,
      "grad_norm": 0.43441545963287354,
      "learning_rate": 8e-05,
      "loss": 1.8112,
      "step": 2828
    },
    {
      "epoch": 0.30573867934723875,
      "grad_norm": 0.40939152240753174,
      "learning_rate": 8e-05,
      "loss": 1.863,
      "step": 2829
    },
    {
      "epoch": 0.3058467524046255,
      "grad_norm": 0.41446584463119507,
      "learning_rate": 8e-05,
      "loss": 1.8308,
      "step": 2830
    },
    {
      "epoch": 0.3059548254620123,
      "grad_norm": 0.397671103477478,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 2831
    },
    {
      "epoch": 0.3060628985193991,
      "grad_norm": 0.4030945301055908,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 2832
    },
    {
      "epoch": 0.3061709715767859,
      "grad_norm": 0.4101508855819702,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 2833
    },
    {
      "epoch": 0.3062790446341727,
      "grad_norm": 0.41523465514183044,
      "learning_rate": 8e-05,
      "loss": 1.8232,
      "step": 2834
    },
    {
      "epoch": 0.3063871176915595,
      "grad_norm": 0.4247133433818817,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 2835
    },
    {
      "epoch": 0.30649519074894627,
      "grad_norm": 0.39822450280189514,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 2836
    },
    {
      "epoch": 0.30660326380633307,
      "grad_norm": 0.4244658350944519,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 2837
    },
    {
      "epoch": 0.3067113368637199,
      "grad_norm": 0.3994199335575104,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 2838
    },
    {
      "epoch": 0.3068194099211067,
      "grad_norm": 0.42141014337539673,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 2839
    },
    {
      "epoch": 0.3069274829784935,
      "grad_norm": 0.41372114419937134,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 2840
    },
    {
      "epoch": 0.30703555603588023,
      "grad_norm": 0.4196571409702301,
      "learning_rate": 8e-05,
      "loss": 1.8215,
      "step": 2841
    },
    {
      "epoch": 0.30714362909326703,
      "grad_norm": 0.42548659443855286,
      "learning_rate": 8e-05,
      "loss": 1.9188,
      "step": 2842
    },
    {
      "epoch": 0.30725170215065384,
      "grad_norm": 0.4108986556529999,
      "learning_rate": 8e-05,
      "loss": 1.7441,
      "step": 2843
    },
    {
      "epoch": 0.30735977520804064,
      "grad_norm": 0.41645917296409607,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 2844
    },
    {
      "epoch": 0.30746784826542745,
      "grad_norm": 0.40933653712272644,
      "learning_rate": 8e-05,
      "loss": 1.5394,
      "step": 2845
    },
    {
      "epoch": 0.30757592132281425,
      "grad_norm": 0.4148993492126465,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 2846
    },
    {
      "epoch": 0.307683994380201,
      "grad_norm": 0.45025232434272766,
      "learning_rate": 8e-05,
      "loss": 1.9006,
      "step": 2847
    },
    {
      "epoch": 0.3077920674375878,
      "grad_norm": 0.4246053695678711,
      "learning_rate": 8e-05,
      "loss": 1.7728,
      "step": 2848
    },
    {
      "epoch": 0.3079001404949746,
      "grad_norm": 0.42586448788642883,
      "learning_rate": 8e-05,
      "loss": 1.8847,
      "step": 2849
    },
    {
      "epoch": 0.3080082135523614,
      "grad_norm": 0.4433842599391937,
      "learning_rate": 8e-05,
      "loss": 1.812,
      "step": 2850
    },
    {
      "epoch": 0.3081162866097482,
      "grad_norm": 0.43536049127578735,
      "learning_rate": 8e-05,
      "loss": 1.8466,
      "step": 2851
    },
    {
      "epoch": 0.30822435966713496,
      "grad_norm": 0.4822477698326111,
      "learning_rate": 8e-05,
      "loss": 1.8922,
      "step": 2852
    },
    {
      "epoch": 0.30833243272452177,
      "grad_norm": 0.4344266355037689,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 2853
    },
    {
      "epoch": 0.30844050578190857,
      "grad_norm": 0.4310300350189209,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 2854
    },
    {
      "epoch": 0.3085485788392954,
      "grad_norm": 0.43923258781433105,
      "learning_rate": 8e-05,
      "loss": 1.7852,
      "step": 2855
    },
    {
      "epoch": 0.3086566518966822,
      "grad_norm": 0.3957637846469879,
      "learning_rate": 8e-05,
      "loss": 1.7473,
      "step": 2856
    },
    {
      "epoch": 0.3087647249540689,
      "grad_norm": 0.4079345762729645,
      "learning_rate": 8e-05,
      "loss": 1.829,
      "step": 2857
    },
    {
      "epoch": 0.30887279801145573,
      "grad_norm": 0.3995129466056824,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 2858
    },
    {
      "epoch": 0.30898087106884253,
      "grad_norm": 0.4515286982059479,
      "learning_rate": 8e-05,
      "loss": 1.9296,
      "step": 2859
    },
    {
      "epoch": 0.30908894412622934,
      "grad_norm": 0.45590901374816895,
      "learning_rate": 8e-05,
      "loss": 1.9119,
      "step": 2860
    },
    {
      "epoch": 0.30919701718361614,
      "grad_norm": 0.46267810463905334,
      "learning_rate": 8e-05,
      "loss": 1.8239,
      "step": 2861
    },
    {
      "epoch": 0.30930509024100294,
      "grad_norm": 0.42581766843795776,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 2862
    },
    {
      "epoch": 0.3094131632983897,
      "grad_norm": 0.4547431468963623,
      "learning_rate": 8e-05,
      "loss": 1.951,
      "step": 2863
    },
    {
      "epoch": 0.3095212363557765,
      "grad_norm": 0.40804821252822876,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 2864
    },
    {
      "epoch": 0.3096293094131633,
      "grad_norm": 0.41765260696411133,
      "learning_rate": 8e-05,
      "loss": 1.7534,
      "step": 2865
    },
    {
      "epoch": 0.3097373824705501,
      "grad_norm": 0.41498079895973206,
      "learning_rate": 8e-05,
      "loss": 1.5889,
      "step": 2866
    },
    {
      "epoch": 0.3098454555279369,
      "grad_norm": 0.419209748506546,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 2867
    },
    {
      "epoch": 0.30995352858532366,
      "grad_norm": 0.40154528617858887,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 2868
    },
    {
      "epoch": 0.31006160164271046,
      "grad_norm": 0.45431065559387207,
      "learning_rate": 8e-05,
      "loss": 1.9379,
      "step": 2869
    },
    {
      "epoch": 0.31016967470009726,
      "grad_norm": 0.3904293477535248,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 2870
    },
    {
      "epoch": 0.31027774775748407,
      "grad_norm": 0.4148043990135193,
      "learning_rate": 8e-05,
      "loss": 1.8978,
      "step": 2871
    },
    {
      "epoch": 0.31038582081487087,
      "grad_norm": 0.434434175491333,
      "learning_rate": 8e-05,
      "loss": 1.8167,
      "step": 2872
    },
    {
      "epoch": 0.3104938938722576,
      "grad_norm": 0.4073236584663391,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 2873
    },
    {
      "epoch": 0.3106019669296444,
      "grad_norm": 0.43892669677734375,
      "learning_rate": 8e-05,
      "loss": 1.9825,
      "step": 2874
    },
    {
      "epoch": 0.31071003998703123,
      "grad_norm": 0.45627570152282715,
      "learning_rate": 8e-05,
      "loss": 1.9675,
      "step": 2875
    },
    {
      "epoch": 0.31081811304441803,
      "grad_norm": 0.4521730840206146,
      "learning_rate": 8e-05,
      "loss": 1.8681,
      "step": 2876
    },
    {
      "epoch": 0.31092618610180484,
      "grad_norm": 0.420754075050354,
      "learning_rate": 8e-05,
      "loss": 1.8831,
      "step": 2877
    },
    {
      "epoch": 0.31103425915919164,
      "grad_norm": 0.4400045573711395,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 2878
    },
    {
      "epoch": 0.3111423322165784,
      "grad_norm": 0.3806029260158539,
      "learning_rate": 8e-05,
      "loss": 1.6193,
      "step": 2879
    },
    {
      "epoch": 0.3112504052739652,
      "grad_norm": 0.43675264716148376,
      "learning_rate": 8e-05,
      "loss": 1.9278,
      "step": 2880
    },
    {
      "epoch": 0.311358478331352,
      "grad_norm": 0.4348922371864319,
      "learning_rate": 8e-05,
      "loss": 1.9059,
      "step": 2881
    },
    {
      "epoch": 0.3114665513887388,
      "grad_norm": 0.40584638714790344,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 2882
    },
    {
      "epoch": 0.3115746244461256,
      "grad_norm": 0.411067396402359,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 2883
    },
    {
      "epoch": 0.31168269750351235,
      "grad_norm": 0.406663715839386,
      "learning_rate": 8e-05,
      "loss": 1.4833,
      "step": 2884
    },
    {
      "epoch": 0.31179077056089916,
      "grad_norm": 0.44400888681411743,
      "learning_rate": 8e-05,
      "loss": 1.8271,
      "step": 2885
    },
    {
      "epoch": 0.31189884361828596,
      "grad_norm": 0.42654529213905334,
      "learning_rate": 8e-05,
      "loss": 1.8319,
      "step": 2886
    },
    {
      "epoch": 0.31200691667567276,
      "grad_norm": 0.4209960103034973,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 2887
    },
    {
      "epoch": 0.31211498973305957,
      "grad_norm": 0.40348756313323975,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 2888
    },
    {
      "epoch": 0.3122230627904463,
      "grad_norm": 0.3908636271953583,
      "learning_rate": 8e-05,
      "loss": 1.6577,
      "step": 2889
    },
    {
      "epoch": 0.3123311358478331,
      "grad_norm": 0.3886580765247345,
      "learning_rate": 8e-05,
      "loss": 1.5938,
      "step": 2890
    },
    {
      "epoch": 0.3124392089052199,
      "grad_norm": 0.4375034272670746,
      "learning_rate": 8e-05,
      "loss": 1.7138,
      "step": 2891
    },
    {
      "epoch": 0.3125472819626067,
      "grad_norm": 0.43737682700157166,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 2892
    },
    {
      "epoch": 0.31265535501999353,
      "grad_norm": 0.45155423879623413,
      "learning_rate": 8e-05,
      "loss": 1.9524,
      "step": 2893
    },
    {
      "epoch": 0.31276342807738033,
      "grad_norm": 0.4622035622596741,
      "learning_rate": 8e-05,
      "loss": 1.8715,
      "step": 2894
    },
    {
      "epoch": 0.3128715011347671,
      "grad_norm": 0.43890511989593506,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 2895
    },
    {
      "epoch": 0.3129795741921539,
      "grad_norm": 0.4559171199798584,
      "learning_rate": 8e-05,
      "loss": 1.8733,
      "step": 2896
    },
    {
      "epoch": 0.3130876472495407,
      "grad_norm": 0.39365658164024353,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 2897
    },
    {
      "epoch": 0.3131957203069275,
      "grad_norm": 0.44903647899627686,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 2898
    },
    {
      "epoch": 0.3133037933643143,
      "grad_norm": 0.426870733499527,
      "learning_rate": 8e-05,
      "loss": 1.747,
      "step": 2899
    },
    {
      "epoch": 0.31341186642170105,
      "grad_norm": 0.41739729046821594,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 2900
    },
    {
      "epoch": 0.31351993947908785,
      "grad_norm": 0.4392854869365692,
      "learning_rate": 8e-05,
      "loss": 1.9773,
      "step": 2901
    },
    {
      "epoch": 0.31362801253647465,
      "grad_norm": 0.4247322082519531,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 2902
    },
    {
      "epoch": 0.31373608559386146,
      "grad_norm": 0.48353421688079834,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 2903
    },
    {
      "epoch": 0.31384415865124826,
      "grad_norm": 0.3734360635280609,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 2904
    },
    {
      "epoch": 0.313952231708635,
      "grad_norm": 0.3831199109554291,
      "learning_rate": 8e-05,
      "loss": 1.6449,
      "step": 2905
    },
    {
      "epoch": 0.3140603047660218,
      "grad_norm": 0.4375872313976288,
      "learning_rate": 8e-05,
      "loss": 1.8016,
      "step": 2906
    },
    {
      "epoch": 0.3141683778234086,
      "grad_norm": 0.4429185688495636,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 2907
    },
    {
      "epoch": 0.3142764508807954,
      "grad_norm": 0.4258683919906616,
      "learning_rate": 8e-05,
      "loss": 1.7891,
      "step": 2908
    },
    {
      "epoch": 0.3143845239381822,
      "grad_norm": 0.4079177677631378,
      "learning_rate": 8e-05,
      "loss": 1.8717,
      "step": 2909
    },
    {
      "epoch": 0.31449259699556903,
      "grad_norm": 0.4668319821357727,
      "learning_rate": 8e-05,
      "loss": 1.8094,
      "step": 2910
    },
    {
      "epoch": 0.3146006700529558,
      "grad_norm": 0.4111725986003876,
      "learning_rate": 8e-05,
      "loss": 1.7348,
      "step": 2911
    },
    {
      "epoch": 0.3147087431103426,
      "grad_norm": 0.43277454376220703,
      "learning_rate": 8e-05,
      "loss": 1.8044,
      "step": 2912
    },
    {
      "epoch": 0.3148168161677294,
      "grad_norm": 0.41419991850852966,
      "learning_rate": 8e-05,
      "loss": 1.5453,
      "step": 2913
    },
    {
      "epoch": 0.3149248892251162,
      "grad_norm": 0.3926936984062195,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 2914
    },
    {
      "epoch": 0.315032962282503,
      "grad_norm": 0.4333323836326599,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 2915
    },
    {
      "epoch": 0.31514103533988974,
      "grad_norm": 0.4374672472476959,
      "learning_rate": 8e-05,
      "loss": 1.7705,
      "step": 2916
    },
    {
      "epoch": 0.31524910839727655,
      "grad_norm": 0.46449971199035645,
      "learning_rate": 8e-05,
      "loss": 1.6685,
      "step": 2917
    },
    {
      "epoch": 0.31535718145466335,
      "grad_norm": 0.38487470149993896,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 2918
    },
    {
      "epoch": 0.31546525451205015,
      "grad_norm": 0.4252181351184845,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 2919
    },
    {
      "epoch": 0.31557332756943696,
      "grad_norm": 0.44563227891921997,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 2920
    },
    {
      "epoch": 0.31568140062682376,
      "grad_norm": 0.4090970754623413,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 2921
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.40127551555633545,
      "learning_rate": 8e-05,
      "loss": 1.5416,
      "step": 2922
    },
    {
      "epoch": 0.3158975467415973,
      "grad_norm": 0.39282330870628357,
      "learning_rate": 8e-05,
      "loss": 1.5344,
      "step": 2923
    },
    {
      "epoch": 0.3160056197989841,
      "grad_norm": 0.41666939854621887,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 2924
    },
    {
      "epoch": 0.3161136928563709,
      "grad_norm": 0.40843063592910767,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 2925
    },
    {
      "epoch": 0.3162217659137577,
      "grad_norm": 0.42825227975845337,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 2926
    },
    {
      "epoch": 0.3163298389711445,
      "grad_norm": 0.3995383381843567,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 2927
    },
    {
      "epoch": 0.3164379120285313,
      "grad_norm": 0.4784610867500305,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 2928
    },
    {
      "epoch": 0.3165459850859181,
      "grad_norm": 0.4881022870540619,
      "learning_rate": 8e-05,
      "loss": 1.762,
      "step": 2929
    },
    {
      "epoch": 0.3166540581433049,
      "grad_norm": 0.46046000719070435,
      "learning_rate": 8e-05,
      "loss": 1.8875,
      "step": 2930
    },
    {
      "epoch": 0.3167621312006917,
      "grad_norm": 0.4014468789100647,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 2931
    },
    {
      "epoch": 0.31687020425807844,
      "grad_norm": 0.4047388732433319,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 2932
    },
    {
      "epoch": 0.31697827731546524,
      "grad_norm": 0.4481785297393799,
      "learning_rate": 8e-05,
      "loss": 1.8756,
      "step": 2933
    },
    {
      "epoch": 0.31708635037285204,
      "grad_norm": 0.4250706434249878,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 2934
    },
    {
      "epoch": 0.31719442343023885,
      "grad_norm": 0.3816010057926178,
      "learning_rate": 8e-05,
      "loss": 1.4613,
      "step": 2935
    },
    {
      "epoch": 0.31730249648762565,
      "grad_norm": 0.45400470495224,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 2936
    },
    {
      "epoch": 0.31741056954501246,
      "grad_norm": 0.4166319668292999,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 2937
    },
    {
      "epoch": 0.3175186426023992,
      "grad_norm": 0.4325452744960785,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 2938
    },
    {
      "epoch": 0.317626715659786,
      "grad_norm": 0.4477466642856598,
      "learning_rate": 8e-05,
      "loss": 1.8174,
      "step": 2939
    },
    {
      "epoch": 0.3177347887171728,
      "grad_norm": 0.40406152606010437,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 2940
    },
    {
      "epoch": 0.3178428617745596,
      "grad_norm": 0.46426570415496826,
      "learning_rate": 8e-05,
      "loss": 1.8688,
      "step": 2941
    },
    {
      "epoch": 0.3179509348319464,
      "grad_norm": 0.3949672281742096,
      "learning_rate": 8e-05,
      "loss": 1.5921,
      "step": 2942
    },
    {
      "epoch": 0.31805900788933317,
      "grad_norm": 0.5116055011749268,
      "learning_rate": 8e-05,
      "loss": 1.9048,
      "step": 2943
    },
    {
      "epoch": 0.31816708094671997,
      "grad_norm": 0.4458138644695282,
      "learning_rate": 8e-05,
      "loss": 1.9113,
      "step": 2944
    },
    {
      "epoch": 0.3182751540041068,
      "grad_norm": 0.4163878560066223,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 2945
    },
    {
      "epoch": 0.3183832270614936,
      "grad_norm": 0.4034114181995392,
      "learning_rate": 8e-05,
      "loss": 1.8564,
      "step": 2946
    },
    {
      "epoch": 0.3184913001188804,
      "grad_norm": 0.41495972871780396,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 2947
    },
    {
      "epoch": 0.31859937317626713,
      "grad_norm": 0.4329721927642822,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 2948
    },
    {
      "epoch": 0.31870744623365393,
      "grad_norm": 0.42824700474739075,
      "learning_rate": 8e-05,
      "loss": 1.9206,
      "step": 2949
    },
    {
      "epoch": 0.31881551929104074,
      "grad_norm": 0.3824887275695801,
      "learning_rate": 8e-05,
      "loss": 1.702,
      "step": 2950
    },
    {
      "epoch": 0.31892359234842754,
      "grad_norm": 0.4088220000267029,
      "learning_rate": 8e-05,
      "loss": 1.4874,
      "step": 2951
    },
    {
      "epoch": 0.31903166540581435,
      "grad_norm": 0.4099370837211609,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 2952
    },
    {
      "epoch": 0.31913973846320115,
      "grad_norm": 0.47194668650627136,
      "learning_rate": 8e-05,
      "loss": 1.8939,
      "step": 2953
    },
    {
      "epoch": 0.3192478115205879,
      "grad_norm": 0.40296220779418945,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 2954
    },
    {
      "epoch": 0.3193558845779747,
      "grad_norm": 0.40350601077079773,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 2955
    },
    {
      "epoch": 0.3194639576353615,
      "grad_norm": 0.389812707901001,
      "learning_rate": 8e-05,
      "loss": 1.709,
      "step": 2956
    },
    {
      "epoch": 0.3195720306927483,
      "grad_norm": 0.4088842272758484,
      "learning_rate": 8e-05,
      "loss": 1.5742,
      "step": 2957
    },
    {
      "epoch": 0.3196801037501351,
      "grad_norm": 0.4472031891345978,
      "learning_rate": 8e-05,
      "loss": 1.8024,
      "step": 2958
    },
    {
      "epoch": 0.31978817680752186,
      "grad_norm": 0.45153364539146423,
      "learning_rate": 8e-05,
      "loss": 1.8952,
      "step": 2959
    },
    {
      "epoch": 0.31989624986490867,
      "grad_norm": 0.42959079146385193,
      "learning_rate": 8e-05,
      "loss": 1.6147,
      "step": 2960
    },
    {
      "epoch": 0.32000432292229547,
      "grad_norm": 0.4306739270687103,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 2961
    },
    {
      "epoch": 0.3201123959796823,
      "grad_norm": 0.4601111114025116,
      "learning_rate": 8e-05,
      "loss": 1.911,
      "step": 2962
    },
    {
      "epoch": 0.3202204690370691,
      "grad_norm": 0.39708569645881653,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 2963
    },
    {
      "epoch": 0.3203285420944558,
      "grad_norm": 0.3992665112018585,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 2964
    },
    {
      "epoch": 0.32043661515184263,
      "grad_norm": 0.4220523536205292,
      "learning_rate": 8e-05,
      "loss": 1.8455,
      "step": 2965
    },
    {
      "epoch": 0.32054468820922943,
      "grad_norm": 0.47129711508750916,
      "learning_rate": 8e-05,
      "loss": 1.909,
      "step": 2966
    },
    {
      "epoch": 0.32065276126661624,
      "grad_norm": 0.46551841497421265,
      "learning_rate": 8e-05,
      "loss": 1.8191,
      "step": 2967
    },
    {
      "epoch": 0.32076083432400304,
      "grad_norm": 0.39408302307128906,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 2968
    },
    {
      "epoch": 0.32086890738138985,
      "grad_norm": 0.43195945024490356,
      "learning_rate": 8e-05,
      "loss": 1.9179,
      "step": 2969
    },
    {
      "epoch": 0.3209769804387766,
      "grad_norm": 0.3993338346481323,
      "learning_rate": 8e-05,
      "loss": 1.8044,
      "step": 2970
    },
    {
      "epoch": 0.3210850534961634,
      "grad_norm": 0.4461846351623535,
      "learning_rate": 8e-05,
      "loss": 1.824,
      "step": 2971
    },
    {
      "epoch": 0.3211931265535502,
      "grad_norm": 0.4449719488620758,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 2972
    },
    {
      "epoch": 0.321301199610937,
      "grad_norm": 0.4419134557247162,
      "learning_rate": 8e-05,
      "loss": 1.8076,
      "step": 2973
    },
    {
      "epoch": 0.3214092726683238,
      "grad_norm": 0.44134676456451416,
      "learning_rate": 8e-05,
      "loss": 1.724,
      "step": 2974
    },
    {
      "epoch": 0.32151734572571056,
      "grad_norm": 0.41631853580474854,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 2975
    },
    {
      "epoch": 0.32162541878309736,
      "grad_norm": 0.40216460824012756,
      "learning_rate": 8e-05,
      "loss": 1.6748,
      "step": 2976
    },
    {
      "epoch": 0.32173349184048416,
      "grad_norm": 0.44889023900032043,
      "learning_rate": 8e-05,
      "loss": 1.9453,
      "step": 2977
    },
    {
      "epoch": 0.32184156489787097,
      "grad_norm": 0.417413592338562,
      "learning_rate": 8e-05,
      "loss": 1.7929,
      "step": 2978
    },
    {
      "epoch": 0.3219496379552578,
      "grad_norm": 0.4462122917175293,
      "learning_rate": 8e-05,
      "loss": 1.9467,
      "step": 2979
    },
    {
      "epoch": 0.3220577110126445,
      "grad_norm": 0.40477171540260315,
      "learning_rate": 8e-05,
      "loss": 1.7609,
      "step": 2980
    },
    {
      "epoch": 0.3221657840700313,
      "grad_norm": 0.41635435819625854,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 2981
    },
    {
      "epoch": 0.32227385712741813,
      "grad_norm": 0.4197012484073639,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 2982
    },
    {
      "epoch": 0.32238193018480493,
      "grad_norm": 0.41912776231765747,
      "learning_rate": 8e-05,
      "loss": 1.8963,
      "step": 2983
    },
    {
      "epoch": 0.32249000324219174,
      "grad_norm": 0.48309093713760376,
      "learning_rate": 8e-05,
      "loss": 1.9153,
      "step": 2984
    },
    {
      "epoch": 0.32259807629957854,
      "grad_norm": 0.4058014750480652,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 2985
    },
    {
      "epoch": 0.3227061493569653,
      "grad_norm": 0.42677488923072815,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 2986
    },
    {
      "epoch": 0.3228142224143521,
      "grad_norm": 0.45646682381629944,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 2987
    },
    {
      "epoch": 0.3229222954717389,
      "grad_norm": 0.43742141127586365,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 2988
    },
    {
      "epoch": 0.3230303685291257,
      "grad_norm": 0.4010315537452698,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 2989
    },
    {
      "epoch": 0.3231384415865125,
      "grad_norm": 0.42877694964408875,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 2990
    },
    {
      "epoch": 0.32324651464389925,
      "grad_norm": 0.4029404819011688,
      "learning_rate": 8e-05,
      "loss": 1.7535,
      "step": 2991
    },
    {
      "epoch": 0.32335458770128606,
      "grad_norm": 0.4304974675178528,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 2992
    },
    {
      "epoch": 0.32346266075867286,
      "grad_norm": 0.4735688269138336,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 2993
    },
    {
      "epoch": 0.32357073381605966,
      "grad_norm": 0.4035581648349762,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 2994
    },
    {
      "epoch": 0.32367880687344647,
      "grad_norm": 0.42899224162101746,
      "learning_rate": 8e-05,
      "loss": 1.7463,
      "step": 2995
    },
    {
      "epoch": 0.3237868799308332,
      "grad_norm": 0.40100765228271484,
      "learning_rate": 8e-05,
      "loss": 1.6415,
      "step": 2996
    },
    {
      "epoch": 0.32389495298822,
      "grad_norm": 0.40182361006736755,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 2997
    },
    {
      "epoch": 0.3240030260456068,
      "grad_norm": 0.43836691975593567,
      "learning_rate": 8e-05,
      "loss": 1.775,
      "step": 2998
    },
    {
      "epoch": 0.3241110991029936,
      "grad_norm": 0.4050801992416382,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 2999
    },
    {
      "epoch": 0.32421917216038043,
      "grad_norm": 0.43715113401412964,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 3000
    },
    {
      "epoch": 0.32432724521776723,
      "grad_norm": 0.4274667203426361,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 3001
    },
    {
      "epoch": 0.324435318275154,
      "grad_norm": 0.46873313188552856,
      "learning_rate": 8e-05,
      "loss": 1.9075,
      "step": 3002
    },
    {
      "epoch": 0.3245433913325408,
      "grad_norm": 0.39148345589637756,
      "learning_rate": 8e-05,
      "loss": 1.5117,
      "step": 3003
    },
    {
      "epoch": 0.3246514643899276,
      "grad_norm": 0.44787365198135376,
      "learning_rate": 8e-05,
      "loss": 1.9059,
      "step": 3004
    },
    {
      "epoch": 0.3247595374473144,
      "grad_norm": 0.43918052315711975,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 3005
    },
    {
      "epoch": 0.3248676105047012,
      "grad_norm": 0.4280099868774414,
      "learning_rate": 8e-05,
      "loss": 1.8753,
      "step": 3006
    },
    {
      "epoch": 0.32497568356208795,
      "grad_norm": 0.41691848635673523,
      "learning_rate": 8e-05,
      "loss": 1.4978,
      "step": 3007
    },
    {
      "epoch": 0.32508375661947475,
      "grad_norm": 0.49442532658576965,
      "learning_rate": 8e-05,
      "loss": 1.8377,
      "step": 3008
    },
    {
      "epoch": 0.32519182967686155,
      "grad_norm": 0.4707207977771759,
      "learning_rate": 8e-05,
      "loss": 1.6946,
      "step": 3009
    },
    {
      "epoch": 0.32529990273424836,
      "grad_norm": 0.4249975085258484,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 3010
    },
    {
      "epoch": 0.32540797579163516,
      "grad_norm": 0.39736315608024597,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 3011
    },
    {
      "epoch": 0.32551604884902197,
      "grad_norm": 0.4484423100948334,
      "learning_rate": 8e-05,
      "loss": 1.7611,
      "step": 3012
    },
    {
      "epoch": 0.3256241219064087,
      "grad_norm": 0.46918249130249023,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 3013
    },
    {
      "epoch": 0.3257321949637955,
      "grad_norm": 0.44564807415008545,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 3014
    },
    {
      "epoch": 0.3258402680211823,
      "grad_norm": 0.3841039836406708,
      "learning_rate": 8e-05,
      "loss": 1.578,
      "step": 3015
    },
    {
      "epoch": 0.3259483410785691,
      "grad_norm": 0.4091604948043823,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 3016
    },
    {
      "epoch": 0.32605641413595593,
      "grad_norm": 0.43443185091018677,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 3017
    },
    {
      "epoch": 0.3261644871933427,
      "grad_norm": 0.43932217359542847,
      "learning_rate": 8e-05,
      "loss": 1.8661,
      "step": 3018
    },
    {
      "epoch": 0.3262725602507295,
      "grad_norm": 0.3930162787437439,
      "learning_rate": 8e-05,
      "loss": 1.5633,
      "step": 3019
    },
    {
      "epoch": 0.3263806333081163,
      "grad_norm": 0.4259348213672638,
      "learning_rate": 8e-05,
      "loss": 1.6437,
      "step": 3020
    },
    {
      "epoch": 0.3264887063655031,
      "grad_norm": 0.42816415429115295,
      "learning_rate": 8e-05,
      "loss": 1.8355,
      "step": 3021
    },
    {
      "epoch": 0.3265967794228899,
      "grad_norm": 0.43021613359451294,
      "learning_rate": 8e-05,
      "loss": 1.8354,
      "step": 3022
    },
    {
      "epoch": 0.32670485248027664,
      "grad_norm": 0.43637916445732117,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 3023
    },
    {
      "epoch": 0.32681292553766345,
      "grad_norm": 0.4315795600414276,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 3024
    },
    {
      "epoch": 0.32692099859505025,
      "grad_norm": 0.4215269088745117,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 3025
    },
    {
      "epoch": 0.32702907165243705,
      "grad_norm": 0.40472617745399475,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 3026
    },
    {
      "epoch": 0.32713714470982386,
      "grad_norm": 0.46722322702407837,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 3027
    },
    {
      "epoch": 0.32724521776721066,
      "grad_norm": 0.4307945966720581,
      "learning_rate": 8e-05,
      "loss": 1.8617,
      "step": 3028
    },
    {
      "epoch": 0.3273532908245974,
      "grad_norm": 0.40561720728874207,
      "learning_rate": 8e-05,
      "loss": 1.6063,
      "step": 3029
    },
    {
      "epoch": 0.3274613638819842,
      "grad_norm": 0.398682177066803,
      "learning_rate": 8e-05,
      "loss": 1.6,
      "step": 3030
    },
    {
      "epoch": 0.327569436939371,
      "grad_norm": 0.42276349663734436,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 3031
    },
    {
      "epoch": 0.3276775099967578,
      "grad_norm": 0.41630324721336365,
      "learning_rate": 8e-05,
      "loss": 1.6214,
      "step": 3032
    },
    {
      "epoch": 0.3277855830541446,
      "grad_norm": 0.3930799663066864,
      "learning_rate": 8e-05,
      "loss": 1.7514,
      "step": 3033
    },
    {
      "epoch": 0.3278936561115314,
      "grad_norm": 0.4715534448623657,
      "learning_rate": 8e-05,
      "loss": 1.7982,
      "step": 3034
    },
    {
      "epoch": 0.3280017291689182,
      "grad_norm": 0.44665688276290894,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 3035
    },
    {
      "epoch": 0.328109802226305,
      "grad_norm": 0.4343736469745636,
      "learning_rate": 8e-05,
      "loss": 1.8544,
      "step": 3036
    },
    {
      "epoch": 0.3282178752836918,
      "grad_norm": 0.415327250957489,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 3037
    },
    {
      "epoch": 0.3283259483410786,
      "grad_norm": 0.4311169683933258,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 3038
    },
    {
      "epoch": 0.32843402139846534,
      "grad_norm": 0.446220338344574,
      "learning_rate": 8e-05,
      "loss": 1.8411,
      "step": 3039
    },
    {
      "epoch": 0.32854209445585214,
      "grad_norm": 0.40852731466293335,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 3040
    },
    {
      "epoch": 0.32865016751323894,
      "grad_norm": 0.4058588743209839,
      "learning_rate": 8e-05,
      "loss": 1.7878,
      "step": 3041
    },
    {
      "epoch": 0.32875824057062575,
      "grad_norm": 0.43577367067337036,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 3042
    },
    {
      "epoch": 0.32886631362801255,
      "grad_norm": 0.41561010479927063,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 3043
    },
    {
      "epoch": 0.32897438668539936,
      "grad_norm": 0.407105952501297,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 3044
    },
    {
      "epoch": 0.3290824597427861,
      "grad_norm": 0.4181126058101654,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 3045
    },
    {
      "epoch": 0.3291905328001729,
      "grad_norm": 0.4350561499595642,
      "learning_rate": 8e-05,
      "loss": 1.8571,
      "step": 3046
    },
    {
      "epoch": 0.3292986058575597,
      "grad_norm": 0.4740116596221924,
      "learning_rate": 8e-05,
      "loss": 1.9415,
      "step": 3047
    },
    {
      "epoch": 0.3294066789149465,
      "grad_norm": 0.4130246937274933,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 3048
    },
    {
      "epoch": 0.3295147519723333,
      "grad_norm": 0.42230039834976196,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 3049
    },
    {
      "epoch": 0.32962282502972007,
      "grad_norm": 0.3944128155708313,
      "learning_rate": 8e-05,
      "loss": 1.7506,
      "step": 3050
    },
    {
      "epoch": 0.32973089808710687,
      "grad_norm": 0.4058973789215088,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 3051
    },
    {
      "epoch": 0.3298389711444937,
      "grad_norm": 0.4101555347442627,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 3052
    },
    {
      "epoch": 0.3299470442018805,
      "grad_norm": 0.4520561397075653,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 3053
    },
    {
      "epoch": 0.3300551172592673,
      "grad_norm": 0.4959138035774231,
      "learning_rate": 8e-05,
      "loss": 1.9331,
      "step": 3054
    },
    {
      "epoch": 0.33016319031665403,
      "grad_norm": 0.4211522340774536,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 3055
    },
    {
      "epoch": 0.33027126337404084,
      "grad_norm": 0.40640613436698914,
      "learning_rate": 8e-05,
      "loss": 1.7206,
      "step": 3056
    },
    {
      "epoch": 0.33037933643142764,
      "grad_norm": 0.46440303325653076,
      "learning_rate": 8e-05,
      "loss": 1.9088,
      "step": 3057
    },
    {
      "epoch": 0.33048740948881444,
      "grad_norm": 0.4670504927635193,
      "learning_rate": 8e-05,
      "loss": 1.9221,
      "step": 3058
    },
    {
      "epoch": 0.33059548254620125,
      "grad_norm": 0.42659303545951843,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 3059
    },
    {
      "epoch": 0.33070355560358805,
      "grad_norm": 0.39858758449554443,
      "learning_rate": 8e-05,
      "loss": 1.6466,
      "step": 3060
    },
    {
      "epoch": 0.3308116286609748,
      "grad_norm": 0.402540385723114,
      "learning_rate": 8e-05,
      "loss": 1.6359,
      "step": 3061
    },
    {
      "epoch": 0.3309197017183616,
      "grad_norm": 0.46204978227615356,
      "learning_rate": 8e-05,
      "loss": 1.8235,
      "step": 3062
    },
    {
      "epoch": 0.3310277747757484,
      "grad_norm": 0.4290003776550293,
      "learning_rate": 8e-05,
      "loss": 1.9033,
      "step": 3063
    },
    {
      "epoch": 0.3311358478331352,
      "grad_norm": 0.45631512999534607,
      "learning_rate": 8e-05,
      "loss": 1.7768,
      "step": 3064
    },
    {
      "epoch": 0.331243920890522,
      "grad_norm": 0.4152870774269104,
      "learning_rate": 8e-05,
      "loss": 1.7582,
      "step": 3065
    },
    {
      "epoch": 0.33135199394790876,
      "grad_norm": 0.4072698652744293,
      "learning_rate": 8e-05,
      "loss": 1.6421,
      "step": 3066
    },
    {
      "epoch": 0.33146006700529557,
      "grad_norm": 0.41352584958076477,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 3067
    },
    {
      "epoch": 0.33156814006268237,
      "grad_norm": 0.4378058910369873,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 3068
    },
    {
      "epoch": 0.3316762131200692,
      "grad_norm": 0.4091898202896118,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 3069
    },
    {
      "epoch": 0.331784286177456,
      "grad_norm": 0.41846904158592224,
      "learning_rate": 8e-05,
      "loss": 1.7473,
      "step": 3070
    },
    {
      "epoch": 0.3318923592348427,
      "grad_norm": 0.41793161630630493,
      "learning_rate": 8e-05,
      "loss": 1.7682,
      "step": 3071
    },
    {
      "epoch": 0.33200043229222953,
      "grad_norm": 0.4743127226829529,
      "learning_rate": 8e-05,
      "loss": 1.9957,
      "step": 3072
    },
    {
      "epoch": 0.33210850534961633,
      "grad_norm": 0.4209557771682739,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 3073
    },
    {
      "epoch": 0.33221657840700314,
      "grad_norm": 0.41568198800086975,
      "learning_rate": 8e-05,
      "loss": 1.6482,
      "step": 3074
    },
    {
      "epoch": 0.33232465146438994,
      "grad_norm": 0.4137724041938782,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 3075
    },
    {
      "epoch": 0.33243272452177675,
      "grad_norm": 0.43064218759536743,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 3076
    },
    {
      "epoch": 0.3325407975791635,
      "grad_norm": 0.43294286727905273,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 3077
    },
    {
      "epoch": 0.3326488706365503,
      "grad_norm": 0.4668363332748413,
      "learning_rate": 8e-05,
      "loss": 1.9558,
      "step": 3078
    },
    {
      "epoch": 0.3327569436939371,
      "grad_norm": 0.4464457035064697,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 3079
    },
    {
      "epoch": 0.3328650167513239,
      "grad_norm": 0.4177311956882477,
      "learning_rate": 8e-05,
      "loss": 1.7329,
      "step": 3080
    },
    {
      "epoch": 0.3329730898087107,
      "grad_norm": 0.43946969509124756,
      "learning_rate": 8e-05,
      "loss": 1.9191,
      "step": 3081
    },
    {
      "epoch": 0.33308116286609746,
      "grad_norm": 0.39584052562713623,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 3082
    },
    {
      "epoch": 0.33318923592348426,
      "grad_norm": 0.4504165053367615,
      "learning_rate": 8e-05,
      "loss": 1.8455,
      "step": 3083
    },
    {
      "epoch": 0.33329730898087107,
      "grad_norm": 0.4235582947731018,
      "learning_rate": 8e-05,
      "loss": 1.7664,
      "step": 3084
    },
    {
      "epoch": 0.33340538203825787,
      "grad_norm": 0.37980496883392334,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 3085
    },
    {
      "epoch": 0.3335134550956447,
      "grad_norm": 0.4000647962093353,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 3086
    },
    {
      "epoch": 0.3336215281530315,
      "grad_norm": 0.4460447132587433,
      "learning_rate": 8e-05,
      "loss": 1.8504,
      "step": 3087
    },
    {
      "epoch": 0.3337296012104182,
      "grad_norm": 0.4252430498600006,
      "learning_rate": 8e-05,
      "loss": 1.7547,
      "step": 3088
    },
    {
      "epoch": 0.33383767426780503,
      "grad_norm": 0.40704989433288574,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 3089
    },
    {
      "epoch": 0.33394574732519183,
      "grad_norm": 0.45068660378456116,
      "learning_rate": 8e-05,
      "loss": 1.534,
      "step": 3090
    },
    {
      "epoch": 0.33405382038257864,
      "grad_norm": 0.4116709530353546,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 3091
    },
    {
      "epoch": 0.33416189343996544,
      "grad_norm": 0.41801294684410095,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 3092
    },
    {
      "epoch": 0.3342699664973522,
      "grad_norm": 0.4651156961917877,
      "learning_rate": 8e-05,
      "loss": 1.9222,
      "step": 3093
    },
    {
      "epoch": 0.334378039554739,
      "grad_norm": 0.40070900321006775,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 3094
    },
    {
      "epoch": 0.3344861126121258,
      "grad_norm": 0.40931007266044617,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 3095
    },
    {
      "epoch": 0.3345941856695126,
      "grad_norm": 0.40640679001808167,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 3096
    },
    {
      "epoch": 0.3347022587268994,
      "grad_norm": 0.4158477783203125,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 3097
    },
    {
      "epoch": 0.33481033178428615,
      "grad_norm": 0.49278485774993896,
      "learning_rate": 8e-05,
      "loss": 2.0003,
      "step": 3098
    },
    {
      "epoch": 0.33491840484167296,
      "grad_norm": 0.4292268753051758,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 3099
    },
    {
      "epoch": 0.33502647789905976,
      "grad_norm": 0.41057267785072327,
      "learning_rate": 8e-05,
      "loss": 1.8693,
      "step": 3100
    },
    {
      "epoch": 0.33513455095644656,
      "grad_norm": 0.40299174189567566,
      "learning_rate": 8e-05,
      "loss": 1.7457,
      "step": 3101
    },
    {
      "epoch": 0.33524262401383337,
      "grad_norm": 0.4210195243358612,
      "learning_rate": 8e-05,
      "loss": 1.7728,
      "step": 3102
    },
    {
      "epoch": 0.33535069707122017,
      "grad_norm": 0.4025794565677643,
      "learning_rate": 8e-05,
      "loss": 1.5578,
      "step": 3103
    },
    {
      "epoch": 0.3354587701286069,
      "grad_norm": 0.4015677571296692,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 3104
    },
    {
      "epoch": 0.3355668431859937,
      "grad_norm": 0.3864572048187256,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 3105
    },
    {
      "epoch": 0.33567491624338053,
      "grad_norm": 0.3936927914619446,
      "learning_rate": 8e-05,
      "loss": 1.7594,
      "step": 3106
    },
    {
      "epoch": 0.33578298930076733,
      "grad_norm": 0.4125633239746094,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 3107
    },
    {
      "epoch": 0.33589106235815414,
      "grad_norm": 0.3989204466342926,
      "learning_rate": 8e-05,
      "loss": 1.7663,
      "step": 3108
    },
    {
      "epoch": 0.3359991354155409,
      "grad_norm": 0.4242277443408966,
      "learning_rate": 8e-05,
      "loss": 1.7957,
      "step": 3109
    },
    {
      "epoch": 0.3361072084729277,
      "grad_norm": 0.4633646309375763,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 3110
    },
    {
      "epoch": 0.3362152815303145,
      "grad_norm": 0.3928966820240021,
      "learning_rate": 8e-05,
      "loss": 1.7486,
      "step": 3111
    },
    {
      "epoch": 0.3363233545877013,
      "grad_norm": 0.46732333302497864,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 3112
    },
    {
      "epoch": 0.3364314276450881,
      "grad_norm": 0.5171822309494019,
      "learning_rate": 8e-05,
      "loss": 1.9128,
      "step": 3113
    },
    {
      "epoch": 0.33653950070247485,
      "grad_norm": 0.41021767258644104,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 3114
    },
    {
      "epoch": 0.33664757375986165,
      "grad_norm": 0.42380577325820923,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 3115
    },
    {
      "epoch": 0.33675564681724846,
      "grad_norm": 0.4045974910259247,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 3116
    },
    {
      "epoch": 0.33686371987463526,
      "grad_norm": 0.4008439779281616,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 3117
    },
    {
      "epoch": 0.33697179293202206,
      "grad_norm": 0.4360436201095581,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 3118
    },
    {
      "epoch": 0.33707986598940887,
      "grad_norm": 0.4404592216014862,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 3119
    },
    {
      "epoch": 0.3371879390467956,
      "grad_norm": 0.3943673074245453,
      "learning_rate": 8e-05,
      "loss": 1.6298,
      "step": 3120
    },
    {
      "epoch": 0.3372960121041824,
      "grad_norm": 0.4213535189628601,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 3121
    },
    {
      "epoch": 0.3374040851615692,
      "grad_norm": 0.4204069972038269,
      "learning_rate": 8e-05,
      "loss": 1.5223,
      "step": 3122
    },
    {
      "epoch": 0.337512158218956,
      "grad_norm": 0.43627771735191345,
      "learning_rate": 8e-05,
      "loss": 1.7281,
      "step": 3123
    },
    {
      "epoch": 0.33762023127634283,
      "grad_norm": 0.47002917528152466,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 3124
    },
    {
      "epoch": 0.3377283043337296,
      "grad_norm": 0.42439714074134827,
      "learning_rate": 8e-05,
      "loss": 1.8664,
      "step": 3125
    },
    {
      "epoch": 0.3378363773911164,
      "grad_norm": 0.42767518758773804,
      "learning_rate": 8e-05,
      "loss": 1.7897,
      "step": 3126
    },
    {
      "epoch": 0.3379444504485032,
      "grad_norm": 0.4275858700275421,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 3127
    },
    {
      "epoch": 0.33805252350589,
      "grad_norm": 0.4136587977409363,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 3128
    },
    {
      "epoch": 0.3381605965632768,
      "grad_norm": 0.4272938668727875,
      "learning_rate": 8e-05,
      "loss": 1.6756,
      "step": 3129
    },
    {
      "epoch": 0.33826866962066354,
      "grad_norm": 0.4314397871494293,
      "learning_rate": 8e-05,
      "loss": 1.7496,
      "step": 3130
    },
    {
      "epoch": 0.33837674267805035,
      "grad_norm": 0.4179172217845917,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 3131
    },
    {
      "epoch": 0.33848481573543715,
      "grad_norm": 0.43751034140586853,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 3132
    },
    {
      "epoch": 0.33859288879282395,
      "grad_norm": 0.41688793897628784,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 3133
    },
    {
      "epoch": 0.33870096185021076,
      "grad_norm": 0.43119215965270996,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 3134
    },
    {
      "epoch": 0.33880903490759756,
      "grad_norm": 0.44571852684020996,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 3135
    },
    {
      "epoch": 0.3389171079649843,
      "grad_norm": 0.4326762557029724,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 3136
    },
    {
      "epoch": 0.3390251810223711,
      "grad_norm": 0.47711998224258423,
      "learning_rate": 8e-05,
      "loss": 1.9799,
      "step": 3137
    },
    {
      "epoch": 0.3391332540797579,
      "grad_norm": 0.42300349473953247,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 3138
    },
    {
      "epoch": 0.3392413271371447,
      "grad_norm": 0.4372170567512512,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 3139
    },
    {
      "epoch": 0.3393494001945315,
      "grad_norm": 0.41927826404571533,
      "learning_rate": 8e-05,
      "loss": 1.8407,
      "step": 3140
    },
    {
      "epoch": 0.3394574732519183,
      "grad_norm": 0.4428921937942505,
      "learning_rate": 8e-05,
      "loss": 1.886,
      "step": 3141
    },
    {
      "epoch": 0.3395655463093051,
      "grad_norm": 0.4116774797439575,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 3142
    },
    {
      "epoch": 0.3396736193666919,
      "grad_norm": 0.43969953060150146,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 3143
    },
    {
      "epoch": 0.3397816924240787,
      "grad_norm": 0.4841902554035187,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 3144
    },
    {
      "epoch": 0.3398897654814655,
      "grad_norm": 0.4282861351966858,
      "learning_rate": 8e-05,
      "loss": 1.7256,
      "step": 3145
    },
    {
      "epoch": 0.33999783853885224,
      "grad_norm": 0.4630432724952698,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 3146
    },
    {
      "epoch": 0.34010591159623904,
      "grad_norm": 0.4280772805213928,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 3147
    },
    {
      "epoch": 0.34021398465362584,
      "grad_norm": 0.4484170079231262,
      "learning_rate": 8e-05,
      "loss": 1.8894,
      "step": 3148
    },
    {
      "epoch": 0.34032205771101265,
      "grad_norm": 0.3957655429840088,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 3149
    },
    {
      "epoch": 0.34043013076839945,
      "grad_norm": 0.41571497917175293,
      "learning_rate": 8e-05,
      "loss": 1.7307,
      "step": 3150
    },
    {
      "epoch": 0.34053820382578626,
      "grad_norm": 0.3842501938343048,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 3151
    },
    {
      "epoch": 0.340646276883173,
      "grad_norm": 0.42243802547454834,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 3152
    },
    {
      "epoch": 0.3407543499405598,
      "grad_norm": 0.39638441801071167,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 3153
    },
    {
      "epoch": 0.3408624229979466,
      "grad_norm": 0.43051430583000183,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 3154
    },
    {
      "epoch": 0.3409704960553334,
      "grad_norm": 0.4188895523548126,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 3155
    },
    {
      "epoch": 0.3410785691127202,
      "grad_norm": 0.44626763463020325,
      "learning_rate": 8e-05,
      "loss": 1.7943,
      "step": 3156
    },
    {
      "epoch": 0.34118664217010697,
      "grad_norm": 0.43667277693748474,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 3157
    },
    {
      "epoch": 0.3412947152274938,
      "grad_norm": 0.4267333149909973,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 3158
    },
    {
      "epoch": 0.3414027882848806,
      "grad_norm": 0.41768816113471985,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 3159
    },
    {
      "epoch": 0.3415108613422674,
      "grad_norm": 0.407987505197525,
      "learning_rate": 8e-05,
      "loss": 1.7387,
      "step": 3160
    },
    {
      "epoch": 0.3416189343996542,
      "grad_norm": 0.43792250752449036,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 3161
    },
    {
      "epoch": 0.34172700745704093,
      "grad_norm": 0.4117439389228821,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 3162
    },
    {
      "epoch": 0.34183508051442774,
      "grad_norm": 0.4106815457344055,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 3163
    },
    {
      "epoch": 0.34194315357181454,
      "grad_norm": 0.4210023581981659,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 3164
    },
    {
      "epoch": 0.34205122662920134,
      "grad_norm": 0.41770291328430176,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 3165
    },
    {
      "epoch": 0.34215929968658815,
      "grad_norm": 0.4387108087539673,
      "learning_rate": 8e-05,
      "loss": 1.8686,
      "step": 3166
    },
    {
      "epoch": 0.34226737274397495,
      "grad_norm": 0.4490572512149811,
      "learning_rate": 8e-05,
      "loss": 1.8955,
      "step": 3167
    },
    {
      "epoch": 0.3423754458013617,
      "grad_norm": 0.45487114787101746,
      "learning_rate": 8e-05,
      "loss": 1.9459,
      "step": 3168
    },
    {
      "epoch": 0.3424835188587485,
      "grad_norm": 0.42484787106513977,
      "learning_rate": 8e-05,
      "loss": 1.7255,
      "step": 3169
    },
    {
      "epoch": 0.3425915919161353,
      "grad_norm": 0.4667893946170807,
      "learning_rate": 8e-05,
      "loss": 1.8813,
      "step": 3170
    },
    {
      "epoch": 0.3426996649735221,
      "grad_norm": 0.4319876432418823,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 3171
    },
    {
      "epoch": 0.3428077380309089,
      "grad_norm": 0.38514816761016846,
      "learning_rate": 8e-05,
      "loss": 1.5598,
      "step": 3172
    },
    {
      "epoch": 0.34291581108829566,
      "grad_norm": 0.4200936555862427,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 3173
    },
    {
      "epoch": 0.34302388414568247,
      "grad_norm": 0.4609629511833191,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 3174
    },
    {
      "epoch": 0.34313195720306927,
      "grad_norm": 0.4079894423484802,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 3175
    },
    {
      "epoch": 0.3432400302604561,
      "grad_norm": 0.490003377199173,
      "learning_rate": 8e-05,
      "loss": 2.0016,
      "step": 3176
    },
    {
      "epoch": 0.3433481033178429,
      "grad_norm": 0.4922468960285187,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 3177
    },
    {
      "epoch": 0.3434561763752297,
      "grad_norm": 0.45914337038993835,
      "learning_rate": 8e-05,
      "loss": 1.8994,
      "step": 3178
    },
    {
      "epoch": 0.34356424943261643,
      "grad_norm": 0.4114702343940735,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 3179
    },
    {
      "epoch": 0.34367232249000323,
      "grad_norm": 0.4870302081108093,
      "learning_rate": 8e-05,
      "loss": 1.9911,
      "step": 3180
    },
    {
      "epoch": 0.34378039554739004,
      "grad_norm": 0.43201684951782227,
      "learning_rate": 8e-05,
      "loss": 1.8083,
      "step": 3181
    },
    {
      "epoch": 0.34388846860477684,
      "grad_norm": 0.4474494159221649,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 3182
    },
    {
      "epoch": 0.34399654166216365,
      "grad_norm": 0.4362868368625641,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 3183
    },
    {
      "epoch": 0.3441046147195504,
      "grad_norm": 0.4266641139984131,
      "learning_rate": 8e-05,
      "loss": 1.8221,
      "step": 3184
    },
    {
      "epoch": 0.3442126877769372,
      "grad_norm": 0.4078013300895691,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 3185
    },
    {
      "epoch": 0.344320760834324,
      "grad_norm": 0.40682029724121094,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 3186
    },
    {
      "epoch": 0.3444288338917108,
      "grad_norm": 0.38171419501304626,
      "learning_rate": 8e-05,
      "loss": 1.5783,
      "step": 3187
    },
    {
      "epoch": 0.3445369069490976,
      "grad_norm": 0.4314795434474945,
      "learning_rate": 8e-05,
      "loss": 1.8457,
      "step": 3188
    },
    {
      "epoch": 0.34464498000648436,
      "grad_norm": 0.4193111062049866,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 3189
    },
    {
      "epoch": 0.34475305306387116,
      "grad_norm": 0.4842628538608551,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 3190
    },
    {
      "epoch": 0.34486112612125797,
      "grad_norm": 0.4759855568408966,
      "learning_rate": 8e-05,
      "loss": 1.9449,
      "step": 3191
    },
    {
      "epoch": 0.34496919917864477,
      "grad_norm": 0.3995366096496582,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 3192
    },
    {
      "epoch": 0.3450772722360316,
      "grad_norm": 0.42914876341819763,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 3193
    },
    {
      "epoch": 0.3451853452934184,
      "grad_norm": 0.46592947840690613,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 3194
    },
    {
      "epoch": 0.3452934183508051,
      "grad_norm": 0.4161090552806854,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 3195
    },
    {
      "epoch": 0.34540149140819193,
      "grad_norm": 0.41606152057647705,
      "learning_rate": 8e-05,
      "loss": 1.647,
      "step": 3196
    },
    {
      "epoch": 0.34550956446557873,
      "grad_norm": 0.405917227268219,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 3197
    },
    {
      "epoch": 0.34561763752296554,
      "grad_norm": 0.3937787115573883,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 3198
    },
    {
      "epoch": 0.34572571058035234,
      "grad_norm": 0.4889167845249176,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 3199
    },
    {
      "epoch": 0.3458337836377391,
      "grad_norm": 0.4293760657310486,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 3200
    },
    {
      "epoch": 0.3459418566951259,
      "grad_norm": 0.43282389640808105,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 3201
    },
    {
      "epoch": 0.3460499297525127,
      "grad_norm": 0.47534146904945374,
      "learning_rate": 8e-05,
      "loss": 1.8148,
      "step": 3202
    },
    {
      "epoch": 0.3461580028098995,
      "grad_norm": 0.39827343821525574,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 3203
    },
    {
      "epoch": 0.3462660758672863,
      "grad_norm": 0.4283888041973114,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 3204
    },
    {
      "epoch": 0.34637414892467305,
      "grad_norm": 0.42122387886047363,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 3205
    },
    {
      "epoch": 0.34648222198205986,
      "grad_norm": 0.40423500537872314,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 3206
    },
    {
      "epoch": 0.34659029503944666,
      "grad_norm": 0.4198627769947052,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 3207
    },
    {
      "epoch": 0.34669836809683346,
      "grad_norm": 0.43148568272590637,
      "learning_rate": 8e-05,
      "loss": 1.7561,
      "step": 3208
    },
    {
      "epoch": 0.34680644115422027,
      "grad_norm": 0.46417883038520813,
      "learning_rate": 8e-05,
      "loss": 1.8446,
      "step": 3209
    },
    {
      "epoch": 0.3469145142116071,
      "grad_norm": 0.40573036670684814,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 3210
    },
    {
      "epoch": 0.3470225872689938,
      "grad_norm": 0.39909347891807556,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 3211
    },
    {
      "epoch": 0.3471306603263806,
      "grad_norm": 0.4573175311088562,
      "learning_rate": 8e-05,
      "loss": 1.7014,
      "step": 3212
    },
    {
      "epoch": 0.34723873338376743,
      "grad_norm": 0.4475870430469513,
      "learning_rate": 8e-05,
      "loss": 1.7592,
      "step": 3213
    },
    {
      "epoch": 0.34734680644115423,
      "grad_norm": 0.47634467482566833,
      "learning_rate": 8e-05,
      "loss": 1.8494,
      "step": 3214
    },
    {
      "epoch": 0.34745487949854104,
      "grad_norm": 0.45016056299209595,
      "learning_rate": 8e-05,
      "loss": 1.8226,
      "step": 3215
    },
    {
      "epoch": 0.3475629525559278,
      "grad_norm": 0.42781057953834534,
      "learning_rate": 8e-05,
      "loss": 1.889,
      "step": 3216
    },
    {
      "epoch": 0.3476710256133146,
      "grad_norm": 0.39698705077171326,
      "learning_rate": 8e-05,
      "loss": 1.606,
      "step": 3217
    },
    {
      "epoch": 0.3477790986707014,
      "grad_norm": 0.42816075682640076,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 3218
    },
    {
      "epoch": 0.3478871717280882,
      "grad_norm": 0.4026806950569153,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 3219
    },
    {
      "epoch": 0.347995244785475,
      "grad_norm": 0.41708090901374817,
      "learning_rate": 8e-05,
      "loss": 1.6257,
      "step": 3220
    },
    {
      "epoch": 0.34810331784286175,
      "grad_norm": 0.4317580759525299,
      "learning_rate": 8e-05,
      "loss": 1.559,
      "step": 3221
    },
    {
      "epoch": 0.34821139090024855,
      "grad_norm": 0.5018089413642883,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 3222
    },
    {
      "epoch": 0.34831946395763536,
      "grad_norm": 0.4581655263900757,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 3223
    },
    {
      "epoch": 0.34842753701502216,
      "grad_norm": 0.42933425307273865,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 3224
    },
    {
      "epoch": 0.34853561007240896,
      "grad_norm": 0.41575950384140015,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 3225
    },
    {
      "epoch": 0.34864368312979577,
      "grad_norm": 0.40958285331726074,
      "learning_rate": 8e-05,
      "loss": 1.5311,
      "step": 3226
    },
    {
      "epoch": 0.3487517561871825,
      "grad_norm": 0.39247745275497437,
      "learning_rate": 8e-05,
      "loss": 1.4573,
      "step": 3227
    },
    {
      "epoch": 0.3488598292445693,
      "grad_norm": 0.4475451707839966,
      "learning_rate": 8e-05,
      "loss": 1.4847,
      "step": 3228
    },
    {
      "epoch": 0.3489679023019561,
      "grad_norm": 0.4168192744255066,
      "learning_rate": 8e-05,
      "loss": 1.5586,
      "step": 3229
    },
    {
      "epoch": 0.3490759753593429,
      "grad_norm": 0.4561598300933838,
      "learning_rate": 8e-05,
      "loss": 1.9003,
      "step": 3230
    },
    {
      "epoch": 0.34918404841672973,
      "grad_norm": 0.4506653845310211,
      "learning_rate": 8e-05,
      "loss": 1.8161,
      "step": 3231
    },
    {
      "epoch": 0.3492921214741165,
      "grad_norm": 0.4137413203716278,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 3232
    },
    {
      "epoch": 0.3494001945315033,
      "grad_norm": 0.41496944427490234,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 3233
    },
    {
      "epoch": 0.3495082675888901,
      "grad_norm": 0.4065919816493988,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 3234
    },
    {
      "epoch": 0.3496163406462769,
      "grad_norm": 0.4513411819934845,
      "learning_rate": 8e-05,
      "loss": 1.8436,
      "step": 3235
    },
    {
      "epoch": 0.3497244137036637,
      "grad_norm": 0.4702899158000946,
      "learning_rate": 8e-05,
      "loss": 1.9185,
      "step": 3236
    },
    {
      "epoch": 0.34983248676105044,
      "grad_norm": 0.4147980511188507,
      "learning_rate": 8e-05,
      "loss": 1.5375,
      "step": 3237
    },
    {
      "epoch": 0.34994055981843725,
      "grad_norm": 0.4604896903038025,
      "learning_rate": 8e-05,
      "loss": 1.4544,
      "step": 3238
    },
    {
      "epoch": 0.35004863287582405,
      "grad_norm": 0.4378202259540558,
      "learning_rate": 8e-05,
      "loss": 1.8506,
      "step": 3239
    },
    {
      "epoch": 0.35015670593321085,
      "grad_norm": 0.41581329703330994,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 3240
    },
    {
      "epoch": 0.35026477899059766,
      "grad_norm": 0.42399853467941284,
      "learning_rate": 8e-05,
      "loss": 1.7941,
      "step": 3241
    },
    {
      "epoch": 0.35037285204798446,
      "grad_norm": 0.41465359926223755,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 3242
    },
    {
      "epoch": 0.3504809251053712,
      "grad_norm": 0.39972224831581116,
      "learning_rate": 8e-05,
      "loss": 1.5469,
      "step": 3243
    },
    {
      "epoch": 0.350588998162758,
      "grad_norm": 0.40840935707092285,
      "learning_rate": 8e-05,
      "loss": 1.6906,
      "step": 3244
    },
    {
      "epoch": 0.3506970712201448,
      "grad_norm": 0.42177098989486694,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 3245
    },
    {
      "epoch": 0.3508051442775316,
      "grad_norm": 0.48361465334892273,
      "learning_rate": 8e-05,
      "loss": 1.9457,
      "step": 3246
    },
    {
      "epoch": 0.3509132173349184,
      "grad_norm": 0.43588781356811523,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 3247
    },
    {
      "epoch": 0.3510212903923052,
      "grad_norm": 0.4372222125530243,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 3248
    },
    {
      "epoch": 0.351129363449692,
      "grad_norm": 0.4324086010456085,
      "learning_rate": 8e-05,
      "loss": 1.7946,
      "step": 3249
    },
    {
      "epoch": 0.3512374365070788,
      "grad_norm": 0.4174424409866333,
      "learning_rate": 8e-05,
      "loss": 1.7768,
      "step": 3250
    },
    {
      "epoch": 0.3513455095644656,
      "grad_norm": 0.40522029995918274,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 3251
    },
    {
      "epoch": 0.3514535826218524,
      "grad_norm": 0.3941443860530853,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 3252
    },
    {
      "epoch": 0.3515616556792392,
      "grad_norm": 0.4551820456981659,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 3253
    },
    {
      "epoch": 0.35166972873662594,
      "grad_norm": 0.44897589087486267,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 3254
    },
    {
      "epoch": 0.35177780179401275,
      "grad_norm": 0.4171803295612335,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 3255
    },
    {
      "epoch": 0.35188587485139955,
      "grad_norm": 0.42700546979904175,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 3256
    },
    {
      "epoch": 0.35199394790878635,
      "grad_norm": 0.41935810446739197,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 3257
    },
    {
      "epoch": 0.35210202096617316,
      "grad_norm": 0.4345460832118988,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 3258
    },
    {
      "epoch": 0.3522100940235599,
      "grad_norm": 0.42895886301994324,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 3259
    },
    {
      "epoch": 0.3523181670809467,
      "grad_norm": 0.4170784652233124,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 3260
    },
    {
      "epoch": 0.3524262401383335,
      "grad_norm": 0.4589855968952179,
      "learning_rate": 8e-05,
      "loss": 1.7512,
      "step": 3261
    },
    {
      "epoch": 0.3525343131957203,
      "grad_norm": 0.4081254005432129,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 3262
    },
    {
      "epoch": 0.3526423862531071,
      "grad_norm": 0.4228208065032959,
      "learning_rate": 8e-05,
      "loss": 1.8453,
      "step": 3263
    },
    {
      "epoch": 0.35275045931049387,
      "grad_norm": 0.4077918529510498,
      "learning_rate": 8e-05,
      "loss": 1.5739,
      "step": 3264
    },
    {
      "epoch": 0.3528585323678807,
      "grad_norm": 0.40267708897590637,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 3265
    },
    {
      "epoch": 0.3529666054252675,
      "grad_norm": 0.44638392329216003,
      "learning_rate": 8e-05,
      "loss": 1.9884,
      "step": 3266
    },
    {
      "epoch": 0.3530746784826543,
      "grad_norm": 0.41052767634391785,
      "learning_rate": 8e-05,
      "loss": 1.7374,
      "step": 3267
    },
    {
      "epoch": 0.3531827515400411,
      "grad_norm": 0.4469383955001831,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 3268
    },
    {
      "epoch": 0.3532908245974279,
      "grad_norm": 0.4395454227924347,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 3269
    },
    {
      "epoch": 0.35339889765481464,
      "grad_norm": 0.4277908504009247,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 3270
    },
    {
      "epoch": 0.35350697071220144,
      "grad_norm": 0.4358438551425934,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 3271
    },
    {
      "epoch": 0.35361504376958824,
      "grad_norm": 0.42627719044685364,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 3272
    },
    {
      "epoch": 0.35372311682697505,
      "grad_norm": 0.40813136100769043,
      "learning_rate": 8e-05,
      "loss": 1.5264,
      "step": 3273
    },
    {
      "epoch": 0.35383118988436185,
      "grad_norm": 0.4586943984031677,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 3274
    },
    {
      "epoch": 0.3539392629417486,
      "grad_norm": 0.44319677352905273,
      "learning_rate": 8e-05,
      "loss": 1.8773,
      "step": 3275
    },
    {
      "epoch": 0.3540473359991354,
      "grad_norm": 0.43819814920425415,
      "learning_rate": 8e-05,
      "loss": 1.7398,
      "step": 3276
    },
    {
      "epoch": 0.3541554090565222,
      "grad_norm": 0.4026510417461395,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 3277
    },
    {
      "epoch": 0.354263482113909,
      "grad_norm": 0.44974517822265625,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 3278
    },
    {
      "epoch": 0.3543715551712958,
      "grad_norm": 0.4952731728553772,
      "learning_rate": 8e-05,
      "loss": 1.9736,
      "step": 3279
    },
    {
      "epoch": 0.35447962822868256,
      "grad_norm": 0.4657551646232605,
      "learning_rate": 8e-05,
      "loss": 1.7697,
      "step": 3280
    },
    {
      "epoch": 0.35458770128606937,
      "grad_norm": 0.4027966260910034,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 3281
    },
    {
      "epoch": 0.35469577434345617,
      "grad_norm": 0.41006502509117126,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 3282
    },
    {
      "epoch": 0.354803847400843,
      "grad_norm": 0.42665717005729675,
      "learning_rate": 8e-05,
      "loss": 1.854,
      "step": 3283
    },
    {
      "epoch": 0.3549119204582298,
      "grad_norm": 0.42648714780807495,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 3284
    },
    {
      "epoch": 0.3550199935156166,
      "grad_norm": 0.4522663354873657,
      "learning_rate": 8e-05,
      "loss": 1.8573,
      "step": 3285
    },
    {
      "epoch": 0.35512806657300333,
      "grad_norm": 0.41837015748023987,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 3286
    },
    {
      "epoch": 0.35523613963039014,
      "grad_norm": 0.4212339222431183,
      "learning_rate": 8e-05,
      "loss": 1.8053,
      "step": 3287
    },
    {
      "epoch": 0.35534421268777694,
      "grad_norm": 0.4300413131713867,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 3288
    },
    {
      "epoch": 0.35545228574516374,
      "grad_norm": 0.4104471206665039,
      "learning_rate": 8e-05,
      "loss": 1.8967,
      "step": 3289
    },
    {
      "epoch": 0.35556035880255055,
      "grad_norm": 0.4114314615726471,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 3290
    },
    {
      "epoch": 0.3556684318599373,
      "grad_norm": 0.41653650999069214,
      "learning_rate": 8e-05,
      "loss": 1.7433,
      "step": 3291
    },
    {
      "epoch": 0.3557765049173241,
      "grad_norm": 0.41024404764175415,
      "learning_rate": 8e-05,
      "loss": 1.8103,
      "step": 3292
    },
    {
      "epoch": 0.3558845779747109,
      "grad_norm": 0.40495941042900085,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 3293
    },
    {
      "epoch": 0.3559926510320977,
      "grad_norm": 0.4135931134223938,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 3294
    },
    {
      "epoch": 0.3561007240894845,
      "grad_norm": 0.4157228171825409,
      "learning_rate": 8e-05,
      "loss": 1.647,
      "step": 3295
    },
    {
      "epoch": 0.35620879714687126,
      "grad_norm": 0.4284229576587677,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 3296
    },
    {
      "epoch": 0.35631687020425806,
      "grad_norm": 0.4180998206138611,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 3297
    },
    {
      "epoch": 0.35642494326164487,
      "grad_norm": 0.40335705876350403,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 3298
    },
    {
      "epoch": 0.35653301631903167,
      "grad_norm": 0.42258456349372864,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 3299
    },
    {
      "epoch": 0.3566410893764185,
      "grad_norm": 0.4341300427913666,
      "learning_rate": 8e-05,
      "loss": 1.8053,
      "step": 3300
    },
    {
      "epoch": 0.3567491624338053,
      "grad_norm": 0.39528360962867737,
      "learning_rate": 8e-05,
      "loss": 1.6994,
      "step": 3301
    },
    {
      "epoch": 0.356857235491192,
      "grad_norm": 0.4275352954864502,
      "learning_rate": 8e-05,
      "loss": 1.7606,
      "step": 3302
    },
    {
      "epoch": 0.35696530854857883,
      "grad_norm": 0.4210830628871918,
      "learning_rate": 8e-05,
      "loss": 1.7013,
      "step": 3303
    },
    {
      "epoch": 0.35707338160596563,
      "grad_norm": 0.44945716857910156,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 3304
    },
    {
      "epoch": 0.35718145466335244,
      "grad_norm": 0.3992268443107605,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 3305
    },
    {
      "epoch": 0.35728952772073924,
      "grad_norm": 0.41074949502944946,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 3306
    },
    {
      "epoch": 0.357397600778126,
      "grad_norm": 0.3998318016529083,
      "learning_rate": 8e-05,
      "loss": 1.7721,
      "step": 3307
    },
    {
      "epoch": 0.3575056738355128,
      "grad_norm": 0.45301395654678345,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 3308
    },
    {
      "epoch": 0.3576137468928996,
      "grad_norm": 0.39202815294265747,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 3309
    },
    {
      "epoch": 0.3577218199502864,
      "grad_norm": 0.43178096413612366,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 3310
    },
    {
      "epoch": 0.3578298930076732,
      "grad_norm": 0.5284368991851807,
      "learning_rate": 8e-05,
      "loss": 1.8043,
      "step": 3311
    },
    {
      "epoch": 0.35793796606505995,
      "grad_norm": 0.40143412351608276,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 3312
    },
    {
      "epoch": 0.35804603912244676,
      "grad_norm": 0.4219238758087158,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 3313
    },
    {
      "epoch": 0.35815411217983356,
      "grad_norm": 0.4476851522922516,
      "learning_rate": 8e-05,
      "loss": 1.9229,
      "step": 3314
    },
    {
      "epoch": 0.35826218523722037,
      "grad_norm": 0.4168033003807068,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 3315
    },
    {
      "epoch": 0.35837025829460717,
      "grad_norm": 0.40560925006866455,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 3316
    },
    {
      "epoch": 0.358478331351994,
      "grad_norm": 0.4348239600658417,
      "learning_rate": 8e-05,
      "loss": 1.8231,
      "step": 3317
    },
    {
      "epoch": 0.3585864044093807,
      "grad_norm": 0.455912321805954,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 3318
    },
    {
      "epoch": 0.3586944774667675,
      "grad_norm": 0.4181908071041107,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 3319
    },
    {
      "epoch": 0.35880255052415433,
      "grad_norm": 0.38540831208229065,
      "learning_rate": 8e-05,
      "loss": 1.5013,
      "step": 3320
    },
    {
      "epoch": 0.35891062358154113,
      "grad_norm": 0.42669978737831116,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 3321
    },
    {
      "epoch": 0.35901869663892794,
      "grad_norm": 0.46832171082496643,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 3322
    },
    {
      "epoch": 0.3591267696963147,
      "grad_norm": 0.4478270709514618,
      "learning_rate": 8e-05,
      "loss": 1.9505,
      "step": 3323
    },
    {
      "epoch": 0.3592348427537015,
      "grad_norm": 0.41637474298477173,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 3324
    },
    {
      "epoch": 0.3593429158110883,
      "grad_norm": 0.40347734093666077,
      "learning_rate": 8e-05,
      "loss": 1.5702,
      "step": 3325
    },
    {
      "epoch": 0.3594509888684751,
      "grad_norm": 0.41043171286582947,
      "learning_rate": 8e-05,
      "loss": 1.8084,
      "step": 3326
    },
    {
      "epoch": 0.3595590619258619,
      "grad_norm": 0.4649650752544403,
      "learning_rate": 8e-05,
      "loss": 1.9093,
      "step": 3327
    },
    {
      "epoch": 0.35966713498324865,
      "grad_norm": 0.4051147997379303,
      "learning_rate": 8e-05,
      "loss": 1.788,
      "step": 3328
    },
    {
      "epoch": 0.35977520804063545,
      "grad_norm": 0.42034104466438293,
      "learning_rate": 8e-05,
      "loss": 1.6549,
      "step": 3329
    },
    {
      "epoch": 0.35988328109802226,
      "grad_norm": 0.4450497627258301,
      "learning_rate": 8e-05,
      "loss": 1.879,
      "step": 3330
    },
    {
      "epoch": 0.35999135415540906,
      "grad_norm": 0.4852311313152313,
      "learning_rate": 8e-05,
      "loss": 1.9935,
      "step": 3331
    },
    {
      "epoch": 0.36009942721279586,
      "grad_norm": 0.4377184808254242,
      "learning_rate": 8e-05,
      "loss": 1.7307,
      "step": 3332
    },
    {
      "epoch": 0.36020750027018267,
      "grad_norm": 0.4104357957839966,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 3333
    },
    {
      "epoch": 0.3603155733275694,
      "grad_norm": 0.4191974997520447,
      "learning_rate": 8e-05,
      "loss": 1.5785,
      "step": 3334
    },
    {
      "epoch": 0.3604236463849562,
      "grad_norm": 0.4283120632171631,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 3335
    },
    {
      "epoch": 0.360531719442343,
      "grad_norm": 0.42885103821754456,
      "learning_rate": 8e-05,
      "loss": 1.7895,
      "step": 3336
    },
    {
      "epoch": 0.3606397924997298,
      "grad_norm": 0.4270329177379608,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 3337
    },
    {
      "epoch": 0.36074786555711663,
      "grad_norm": 0.46543481945991516,
      "learning_rate": 8e-05,
      "loss": 1.8118,
      "step": 3338
    },
    {
      "epoch": 0.3608559386145034,
      "grad_norm": 0.42734381556510925,
      "learning_rate": 8e-05,
      "loss": 1.7965,
      "step": 3339
    },
    {
      "epoch": 0.3609640116718902,
      "grad_norm": 0.41826266050338745,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 3340
    },
    {
      "epoch": 0.361072084729277,
      "grad_norm": 0.43543606996536255,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 3341
    },
    {
      "epoch": 0.3611801577866638,
      "grad_norm": 0.42861783504486084,
      "learning_rate": 8e-05,
      "loss": 1.6951,
      "step": 3342
    },
    {
      "epoch": 0.3612882308440506,
      "grad_norm": 0.4121268391609192,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 3343
    },
    {
      "epoch": 0.3613963039014374,
      "grad_norm": 0.4242764711380005,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 3344
    },
    {
      "epoch": 0.36150437695882415,
      "grad_norm": 0.4200666844844818,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 3345
    },
    {
      "epoch": 0.36161245001621095,
      "grad_norm": 0.46164193749427795,
      "learning_rate": 8e-05,
      "loss": 1.8954,
      "step": 3346
    },
    {
      "epoch": 0.36172052307359776,
      "grad_norm": 0.4136163294315338,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 3347
    },
    {
      "epoch": 0.36182859613098456,
      "grad_norm": 0.43137362599372864,
      "learning_rate": 8e-05,
      "loss": 1.9206,
      "step": 3348
    },
    {
      "epoch": 0.36193666918837136,
      "grad_norm": 0.4365239143371582,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 3349
    },
    {
      "epoch": 0.3620447422457581,
      "grad_norm": 0.4212830364704132,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 3350
    },
    {
      "epoch": 0.3621528153031449,
      "grad_norm": 0.4354706406593323,
      "learning_rate": 8e-05,
      "loss": 1.7903,
      "step": 3351
    },
    {
      "epoch": 0.3622608883605317,
      "grad_norm": 0.44342532753944397,
      "learning_rate": 8e-05,
      "loss": 1.8508,
      "step": 3352
    },
    {
      "epoch": 0.3623689614179185,
      "grad_norm": 0.429250031709671,
      "learning_rate": 8e-05,
      "loss": 1.6933,
      "step": 3353
    },
    {
      "epoch": 0.3624770344753053,
      "grad_norm": 0.45404109358787537,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 3354
    },
    {
      "epoch": 0.3625851075326921,
      "grad_norm": 0.43464040756225586,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 3355
    },
    {
      "epoch": 0.3626931805900789,
      "grad_norm": 0.4617641270160675,
      "learning_rate": 8e-05,
      "loss": 1.9045,
      "step": 3356
    },
    {
      "epoch": 0.3628012536474657,
      "grad_norm": 0.3890913128852844,
      "learning_rate": 8e-05,
      "loss": 1.7581,
      "step": 3357
    },
    {
      "epoch": 0.3629093267048525,
      "grad_norm": 0.38741564750671387,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 3358
    },
    {
      "epoch": 0.3630173997622393,
      "grad_norm": 0.40005654096603394,
      "learning_rate": 8e-05,
      "loss": 1.6965,
      "step": 3359
    },
    {
      "epoch": 0.3631254728196261,
      "grad_norm": 0.4294668734073639,
      "learning_rate": 8e-05,
      "loss": 1.8984,
      "step": 3360
    },
    {
      "epoch": 0.36323354587701284,
      "grad_norm": 0.42550942301750183,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 3361
    },
    {
      "epoch": 0.36334161893439965,
      "grad_norm": 0.4105474650859833,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 3362
    },
    {
      "epoch": 0.36344969199178645,
      "grad_norm": 0.42685431241989136,
      "learning_rate": 8e-05,
      "loss": 1.7925,
      "step": 3363
    },
    {
      "epoch": 0.36355776504917325,
      "grad_norm": 0.4384251534938812,
      "learning_rate": 8e-05,
      "loss": 1.812,
      "step": 3364
    },
    {
      "epoch": 0.36366583810656006,
      "grad_norm": 0.44612717628479004,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 3365
    },
    {
      "epoch": 0.3637739111639468,
      "grad_norm": 0.4351370334625244,
      "learning_rate": 8e-05,
      "loss": 1.8546,
      "step": 3366
    },
    {
      "epoch": 0.3638819842213336,
      "grad_norm": 0.39765840768814087,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 3367
    },
    {
      "epoch": 0.3639900572787204,
      "grad_norm": 0.47329598665237427,
      "learning_rate": 8e-05,
      "loss": 1.8263,
      "step": 3368
    },
    {
      "epoch": 0.3640981303361072,
      "grad_norm": 0.41933950781822205,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 3369
    },
    {
      "epoch": 0.364206203393494,
      "grad_norm": 0.43658775091171265,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 3370
    },
    {
      "epoch": 0.36431427645088077,
      "grad_norm": 0.4096856713294983,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 3371
    },
    {
      "epoch": 0.3644223495082676,
      "grad_norm": 0.4300333559513092,
      "learning_rate": 8e-05,
      "loss": 1.8617,
      "step": 3372
    },
    {
      "epoch": 0.3645304225656544,
      "grad_norm": 0.42703747749328613,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 3373
    },
    {
      "epoch": 0.3646384956230412,
      "grad_norm": 0.39638009667396545,
      "learning_rate": 8e-05,
      "loss": 1.669,
      "step": 3374
    },
    {
      "epoch": 0.364746568680428,
      "grad_norm": 0.4091653823852539,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 3375
    },
    {
      "epoch": 0.3648546417378148,
      "grad_norm": 0.4178166091442108,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 3376
    },
    {
      "epoch": 0.36496271479520154,
      "grad_norm": 0.3981482684612274,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 3377
    },
    {
      "epoch": 0.36507078785258834,
      "grad_norm": 0.4674065113067627,
      "learning_rate": 8e-05,
      "loss": 1.8874,
      "step": 3378
    },
    {
      "epoch": 0.36517886090997514,
      "grad_norm": 0.4299635589122772,
      "learning_rate": 8e-05,
      "loss": 1.7187,
      "step": 3379
    },
    {
      "epoch": 0.36528693396736195,
      "grad_norm": 0.43582287430763245,
      "learning_rate": 8e-05,
      "loss": 1.824,
      "step": 3380
    },
    {
      "epoch": 0.36539500702474875,
      "grad_norm": 0.407050222158432,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 3381
    },
    {
      "epoch": 0.3655030800821355,
      "grad_norm": 0.48025399446487427,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 3382
    },
    {
      "epoch": 0.3656111531395223,
      "grad_norm": 0.4449569284915924,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 3383
    },
    {
      "epoch": 0.3657192261969091,
      "grad_norm": 0.4259520471096039,
      "learning_rate": 8e-05,
      "loss": 1.7783,
      "step": 3384
    },
    {
      "epoch": 0.3658272992542959,
      "grad_norm": 0.45011281967163086,
      "learning_rate": 8e-05,
      "loss": 1.8622,
      "step": 3385
    },
    {
      "epoch": 0.3659353723116827,
      "grad_norm": 0.4105873703956604,
      "learning_rate": 8e-05,
      "loss": 1.7107,
      "step": 3386
    },
    {
      "epoch": 0.36604344536906946,
      "grad_norm": 0.4146641790866852,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 3387
    },
    {
      "epoch": 0.36615151842645627,
      "grad_norm": 0.4537700414657593,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 3388
    },
    {
      "epoch": 0.36625959148384307,
      "grad_norm": 0.4525362253189087,
      "learning_rate": 8e-05,
      "loss": 1.8598,
      "step": 3389
    },
    {
      "epoch": 0.3663676645412299,
      "grad_norm": 0.4512573778629303,
      "learning_rate": 8e-05,
      "loss": 1.8833,
      "step": 3390
    },
    {
      "epoch": 0.3664757375986167,
      "grad_norm": 0.4179229438304901,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 3391
    },
    {
      "epoch": 0.3665838106560035,
      "grad_norm": 0.4680153727531433,
      "learning_rate": 8e-05,
      "loss": 1.851,
      "step": 3392
    },
    {
      "epoch": 0.36669188371339023,
      "grad_norm": 0.42009806632995605,
      "learning_rate": 8e-05,
      "loss": 1.7504,
      "step": 3393
    },
    {
      "epoch": 0.36679995677077704,
      "grad_norm": 0.45841076970100403,
      "learning_rate": 8e-05,
      "loss": 1.5156,
      "step": 3394
    },
    {
      "epoch": 0.36690802982816384,
      "grad_norm": 0.40569525957107544,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 3395
    },
    {
      "epoch": 0.36701610288555064,
      "grad_norm": 0.4220332205295563,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 3396
    },
    {
      "epoch": 0.36712417594293745,
      "grad_norm": 0.4452635645866394,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 3397
    },
    {
      "epoch": 0.3672322490003242,
      "grad_norm": 0.5187171697616577,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 3398
    },
    {
      "epoch": 0.367340322057711,
      "grad_norm": 0.4053468406200409,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 3399
    },
    {
      "epoch": 0.3674483951150978,
      "grad_norm": 0.4059796631336212,
      "learning_rate": 8e-05,
      "loss": 1.6094,
      "step": 3400
    },
    {
      "epoch": 0.3675564681724846,
      "grad_norm": 0.41496139764785767,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 3401
    },
    {
      "epoch": 0.3676645412298714,
      "grad_norm": 0.4336828291416168,
      "learning_rate": 8e-05,
      "loss": 1.7218,
      "step": 3402
    },
    {
      "epoch": 0.36777261428725816,
      "grad_norm": 0.4598367214202881,
      "learning_rate": 8e-05,
      "loss": 1.835,
      "step": 3403
    },
    {
      "epoch": 0.36788068734464496,
      "grad_norm": 0.3855069577693939,
      "learning_rate": 8e-05,
      "loss": 1.5542,
      "step": 3404
    },
    {
      "epoch": 0.36798876040203177,
      "grad_norm": 0.4354601502418518,
      "learning_rate": 8e-05,
      "loss": 1.8587,
      "step": 3405
    },
    {
      "epoch": 0.36809683345941857,
      "grad_norm": 0.42919349670410156,
      "learning_rate": 8e-05,
      "loss": 1.5253,
      "step": 3406
    },
    {
      "epoch": 0.3682049065168054,
      "grad_norm": 0.4500512182712555,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 3407
    },
    {
      "epoch": 0.3683129795741922,
      "grad_norm": 0.43949973583221436,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 3408
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.43005630373954773,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 3409
    },
    {
      "epoch": 0.36852912568896573,
      "grad_norm": 0.395633339881897,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 3410
    },
    {
      "epoch": 0.36863719874635253,
      "grad_norm": 0.42412295937538147,
      "learning_rate": 8e-05,
      "loss": 1.8047,
      "step": 3411
    },
    {
      "epoch": 0.36874527180373934,
      "grad_norm": 0.473602294921875,
      "learning_rate": 8e-05,
      "loss": 1.4586,
      "step": 3412
    },
    {
      "epoch": 0.36885334486112614,
      "grad_norm": 0.3959374725818634,
      "learning_rate": 8e-05,
      "loss": 1.6306,
      "step": 3413
    },
    {
      "epoch": 0.3689614179185129,
      "grad_norm": 0.41564103960990906,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 3414
    },
    {
      "epoch": 0.3690694909758997,
      "grad_norm": 0.4456472098827362,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 3415
    },
    {
      "epoch": 0.3691775640332865,
      "grad_norm": 0.43390417098999023,
      "learning_rate": 8e-05,
      "loss": 1.655,
      "step": 3416
    },
    {
      "epoch": 0.3692856370906733,
      "grad_norm": 0.4974266588687897,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 3417
    },
    {
      "epoch": 0.3693937101480601,
      "grad_norm": 0.44012171030044556,
      "learning_rate": 8e-05,
      "loss": 1.8881,
      "step": 3418
    },
    {
      "epoch": 0.3695017832054469,
      "grad_norm": 0.4110865890979767,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 3419
    },
    {
      "epoch": 0.36960985626283366,
      "grad_norm": 0.4619143009185791,
      "learning_rate": 8e-05,
      "loss": 2.0757,
      "step": 3420
    },
    {
      "epoch": 0.36971792932022046,
      "grad_norm": 0.3967478275299072,
      "learning_rate": 8e-05,
      "loss": 1.6631,
      "step": 3421
    },
    {
      "epoch": 0.36982600237760727,
      "grad_norm": 0.42890664935112,
      "learning_rate": 8e-05,
      "loss": 1.8469,
      "step": 3422
    },
    {
      "epoch": 0.36993407543499407,
      "grad_norm": 0.4297618865966797,
      "learning_rate": 8e-05,
      "loss": 1.7214,
      "step": 3423
    },
    {
      "epoch": 0.3700421484923809,
      "grad_norm": 0.4428226947784424,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 3424
    },
    {
      "epoch": 0.3701502215497676,
      "grad_norm": 0.44171640276908875,
      "learning_rate": 8e-05,
      "loss": 1.7913,
      "step": 3425
    },
    {
      "epoch": 0.3702582946071544,
      "grad_norm": 0.4209842383861542,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 3426
    },
    {
      "epoch": 0.37036636766454123,
      "grad_norm": 0.4423057734966278,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 3427
    },
    {
      "epoch": 0.37047444072192803,
      "grad_norm": 0.438478946685791,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 3428
    },
    {
      "epoch": 0.37058251377931484,
      "grad_norm": 0.42629921436309814,
      "learning_rate": 8e-05,
      "loss": 1.7247,
      "step": 3429
    },
    {
      "epoch": 0.3706905868367016,
      "grad_norm": 0.43275707960128784,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 3430
    },
    {
      "epoch": 0.3707986598940884,
      "grad_norm": 0.4102197289466858,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 3431
    },
    {
      "epoch": 0.3709067329514752,
      "grad_norm": 0.43349966406822205,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 3432
    },
    {
      "epoch": 0.371014806008862,
      "grad_norm": 0.41081467270851135,
      "learning_rate": 8e-05,
      "loss": 1.7903,
      "step": 3433
    },
    {
      "epoch": 0.3711228790662488,
      "grad_norm": 0.43383312225341797,
      "learning_rate": 8e-05,
      "loss": 1.7859,
      "step": 3434
    },
    {
      "epoch": 0.3712309521236356,
      "grad_norm": 0.40081873536109924,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 3435
    },
    {
      "epoch": 0.37133902518102235,
      "grad_norm": 0.4512002766132355,
      "learning_rate": 8e-05,
      "loss": 1.9434,
      "step": 3436
    },
    {
      "epoch": 0.37144709823840916,
      "grad_norm": 0.4437725841999054,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 3437
    },
    {
      "epoch": 0.37155517129579596,
      "grad_norm": 0.4484667479991913,
      "learning_rate": 8e-05,
      "loss": 1.8536,
      "step": 3438
    },
    {
      "epoch": 0.37166324435318276,
      "grad_norm": 0.444848895072937,
      "learning_rate": 8e-05,
      "loss": 1.7813,
      "step": 3439
    },
    {
      "epoch": 0.37177131741056957,
      "grad_norm": 0.43107283115386963,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 3440
    },
    {
      "epoch": 0.3718793904679563,
      "grad_norm": 0.4174618124961853,
      "learning_rate": 8e-05,
      "loss": 1.6437,
      "step": 3441
    },
    {
      "epoch": 0.3719874635253431,
      "grad_norm": 0.4822980761528015,
      "learning_rate": 8e-05,
      "loss": 1.9911,
      "step": 3442
    },
    {
      "epoch": 0.3720955365827299,
      "grad_norm": 0.43369293212890625,
      "learning_rate": 8e-05,
      "loss": 1.6776,
      "step": 3443
    },
    {
      "epoch": 0.37220360964011673,
      "grad_norm": 0.40353596210479736,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 3444
    },
    {
      "epoch": 0.37231168269750353,
      "grad_norm": 0.4588055908679962,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 3445
    },
    {
      "epoch": 0.3724197557548903,
      "grad_norm": 0.4066534638404846,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 3446
    },
    {
      "epoch": 0.3725278288122771,
      "grad_norm": 0.43900787830352783,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 3447
    },
    {
      "epoch": 0.3726359018696639,
      "grad_norm": 0.4334293007850647,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 3448
    },
    {
      "epoch": 0.3727439749270507,
      "grad_norm": 0.4250822961330414,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 3449
    },
    {
      "epoch": 0.3728520479844375,
      "grad_norm": 0.4160313010215759,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 3450
    },
    {
      "epoch": 0.3729601210418243,
      "grad_norm": 0.4243246018886566,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 3451
    },
    {
      "epoch": 0.37306819409921105,
      "grad_norm": 0.41720810532569885,
      "learning_rate": 8e-05,
      "loss": 1.7076,
      "step": 3452
    },
    {
      "epoch": 0.37317626715659785,
      "grad_norm": 0.45689088106155396,
      "learning_rate": 8e-05,
      "loss": 1.9267,
      "step": 3453
    },
    {
      "epoch": 0.37328434021398466,
      "grad_norm": 0.4210926592350006,
      "learning_rate": 8e-05,
      "loss": 1.6675,
      "step": 3454
    },
    {
      "epoch": 0.37339241327137146,
      "grad_norm": 0.4649599492549896,
      "learning_rate": 8e-05,
      "loss": 1.8704,
      "step": 3455
    },
    {
      "epoch": 0.37350048632875826,
      "grad_norm": 0.4419330060482025,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 3456
    },
    {
      "epoch": 0.373608559386145,
      "grad_norm": 0.4155466556549072,
      "learning_rate": 8e-05,
      "loss": 1.6557,
      "step": 3457
    },
    {
      "epoch": 0.3737166324435318,
      "grad_norm": 0.4767361581325531,
      "learning_rate": 8e-05,
      "loss": 1.8759,
      "step": 3458
    },
    {
      "epoch": 0.3738247055009186,
      "grad_norm": 0.4284636378288269,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 3459
    },
    {
      "epoch": 0.3739327785583054,
      "grad_norm": 0.40283095836639404,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 3460
    },
    {
      "epoch": 0.3740408516156922,
      "grad_norm": 0.448589950799942,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 3461
    },
    {
      "epoch": 0.374148924673079,
      "grad_norm": 0.40256980061531067,
      "learning_rate": 8e-05,
      "loss": 1.8124,
      "step": 3462
    },
    {
      "epoch": 0.3742569977304658,
      "grad_norm": 0.42498722672462463,
      "learning_rate": 8e-05,
      "loss": 1.7686,
      "step": 3463
    },
    {
      "epoch": 0.3743650707878526,
      "grad_norm": 0.3985738754272461,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 3464
    },
    {
      "epoch": 0.3744731438452394,
      "grad_norm": 0.4589521884918213,
      "learning_rate": 8e-05,
      "loss": 1.9622,
      "step": 3465
    },
    {
      "epoch": 0.3745812169026262,
      "grad_norm": 0.48362651467323303,
      "learning_rate": 8e-05,
      "loss": 1.7215,
      "step": 3466
    },
    {
      "epoch": 0.374689289960013,
      "grad_norm": 0.4078406095504761,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 3467
    },
    {
      "epoch": 0.37479736301739974,
      "grad_norm": 0.40207311511039734,
      "learning_rate": 8e-05,
      "loss": 1.7827,
      "step": 3468
    },
    {
      "epoch": 0.37490543607478655,
      "grad_norm": 0.4413811266422272,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 3469
    },
    {
      "epoch": 0.37501350913217335,
      "grad_norm": 0.44355201721191406,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 3470
    },
    {
      "epoch": 0.37512158218956015,
      "grad_norm": 0.45571786165237427,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 3471
    },
    {
      "epoch": 0.37522965524694696,
      "grad_norm": 0.44833147525787354,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 3472
    },
    {
      "epoch": 0.3753377283043337,
      "grad_norm": 0.42024585604667664,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 3473
    },
    {
      "epoch": 0.3754458013617205,
      "grad_norm": 0.44539734721183777,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 3474
    },
    {
      "epoch": 0.3755538744191073,
      "grad_norm": 0.48538270592689514,
      "learning_rate": 8e-05,
      "loss": 1.7726,
      "step": 3475
    },
    {
      "epoch": 0.3756619474764941,
      "grad_norm": 0.43293142318725586,
      "learning_rate": 8e-05,
      "loss": 1.7216,
      "step": 3476
    },
    {
      "epoch": 0.3757700205338809,
      "grad_norm": 0.4222375452518463,
      "learning_rate": 8e-05,
      "loss": 1.8013,
      "step": 3477
    },
    {
      "epoch": 0.37587809359126767,
      "grad_norm": 0.3983985185623169,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 3478
    },
    {
      "epoch": 0.3759861666486545,
      "grad_norm": 0.4303421974182129,
      "learning_rate": 8e-05,
      "loss": 1.6576,
      "step": 3479
    },
    {
      "epoch": 0.3760942397060413,
      "grad_norm": 0.4115367531776428,
      "learning_rate": 8e-05,
      "loss": 1.762,
      "step": 3480
    },
    {
      "epoch": 0.3762023127634281,
      "grad_norm": 0.4182387888431549,
      "learning_rate": 8e-05,
      "loss": 1.656,
      "step": 3481
    },
    {
      "epoch": 0.3763103858208149,
      "grad_norm": 0.4294019341468811,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 3482
    },
    {
      "epoch": 0.3764184588782017,
      "grad_norm": 0.45268142223358154,
      "learning_rate": 8e-05,
      "loss": 1.8522,
      "step": 3483
    },
    {
      "epoch": 0.37652653193558844,
      "grad_norm": 0.42809242010116577,
      "learning_rate": 8e-05,
      "loss": 1.694,
      "step": 3484
    },
    {
      "epoch": 0.37663460499297524,
      "grad_norm": 0.4289371967315674,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 3485
    },
    {
      "epoch": 0.37674267805036205,
      "grad_norm": 0.45978111028671265,
      "learning_rate": 8e-05,
      "loss": 1.8426,
      "step": 3486
    },
    {
      "epoch": 0.37685075110774885,
      "grad_norm": 0.42952287197113037,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 3487
    },
    {
      "epoch": 0.37695882416513565,
      "grad_norm": 0.41361501812934875,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 3488
    },
    {
      "epoch": 0.3770668972225224,
      "grad_norm": 0.4540669620037079,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 3489
    },
    {
      "epoch": 0.3771749702799092,
      "grad_norm": 0.40815770626068115,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 3490
    },
    {
      "epoch": 0.377283043337296,
      "grad_norm": 0.45647183060646057,
      "learning_rate": 8e-05,
      "loss": 1.6365,
      "step": 3491
    },
    {
      "epoch": 0.3773911163946828,
      "grad_norm": 0.43735361099243164,
      "learning_rate": 8e-05,
      "loss": 1.7094,
      "step": 3492
    },
    {
      "epoch": 0.3774991894520696,
      "grad_norm": 0.44445347785949707,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 3493
    },
    {
      "epoch": 0.37760726250945637,
      "grad_norm": 0.42541736364364624,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 3494
    },
    {
      "epoch": 0.37771533556684317,
      "grad_norm": 0.40757301449775696,
      "learning_rate": 8e-05,
      "loss": 1.734,
      "step": 3495
    },
    {
      "epoch": 0.37782340862423,
      "grad_norm": 0.41255390644073486,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 3496
    },
    {
      "epoch": 0.3779314816816168,
      "grad_norm": 0.40854141116142273,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 3497
    },
    {
      "epoch": 0.3780395547390036,
      "grad_norm": 0.42410773038864136,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 3498
    },
    {
      "epoch": 0.3781476277963904,
      "grad_norm": 0.4516099989414215,
      "learning_rate": 8e-05,
      "loss": 1.8524,
      "step": 3499
    },
    {
      "epoch": 0.37825570085377713,
      "grad_norm": 0.42901453375816345,
      "learning_rate": 8e-05,
      "loss": 1.7042,
      "step": 3500
    },
    {
      "epoch": 0.37836377391116394,
      "grad_norm": 0.4416302442550659,
      "learning_rate": 8e-05,
      "loss": 1.8485,
      "step": 3501
    },
    {
      "epoch": 0.37847184696855074,
      "grad_norm": 0.47387412190437317,
      "learning_rate": 8e-05,
      "loss": 1.8425,
      "step": 3502
    },
    {
      "epoch": 0.37857992002593754,
      "grad_norm": 0.43347400426864624,
      "learning_rate": 8e-05,
      "loss": 1.7411,
      "step": 3503
    },
    {
      "epoch": 0.37868799308332435,
      "grad_norm": 0.43483591079711914,
      "learning_rate": 8e-05,
      "loss": 1.7538,
      "step": 3504
    },
    {
      "epoch": 0.3787960661407111,
      "grad_norm": 0.46430715918540955,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 3505
    },
    {
      "epoch": 0.3789041391980979,
      "grad_norm": 0.4164539873600006,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 3506
    },
    {
      "epoch": 0.3790122122554847,
      "grad_norm": 0.4325180649757385,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 3507
    },
    {
      "epoch": 0.3791202853128715,
      "grad_norm": 0.4040311574935913,
      "learning_rate": 8e-05,
      "loss": 1.5847,
      "step": 3508
    },
    {
      "epoch": 0.3792283583702583,
      "grad_norm": 0.41692501306533813,
      "learning_rate": 8e-05,
      "loss": 1.7944,
      "step": 3509
    },
    {
      "epoch": 0.3793364314276451,
      "grad_norm": 0.4129464626312256,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 3510
    },
    {
      "epoch": 0.37944450448503186,
      "grad_norm": 0.40561172366142273,
      "learning_rate": 8e-05,
      "loss": 1.6533,
      "step": 3511
    },
    {
      "epoch": 0.37955257754241867,
      "grad_norm": 0.4534440338611603,
      "learning_rate": 8e-05,
      "loss": 1.8591,
      "step": 3512
    },
    {
      "epoch": 0.37966065059980547,
      "grad_norm": 0.402233362197876,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 3513
    },
    {
      "epoch": 0.3797687236571923,
      "grad_norm": 0.3915117383003235,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 3514
    },
    {
      "epoch": 0.3798767967145791,
      "grad_norm": 0.4126312732696533,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 3515
    },
    {
      "epoch": 0.3799848697719658,
      "grad_norm": 0.41190609335899353,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 3516
    },
    {
      "epoch": 0.38009294282935263,
      "grad_norm": 0.46126875281333923,
      "learning_rate": 8e-05,
      "loss": 1.8752,
      "step": 3517
    },
    {
      "epoch": 0.38020101588673944,
      "grad_norm": 0.38166698813438416,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 3518
    },
    {
      "epoch": 0.38030908894412624,
      "grad_norm": 0.43661558628082275,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 3519
    },
    {
      "epoch": 0.38041716200151304,
      "grad_norm": 0.45391857624053955,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 3520
    },
    {
      "epoch": 0.3805252350588998,
      "grad_norm": 0.42047610878944397,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 3521
    },
    {
      "epoch": 0.3806333081162866,
      "grad_norm": 0.4686938226222992,
      "learning_rate": 8e-05,
      "loss": 1.9078,
      "step": 3522
    },
    {
      "epoch": 0.3807413811736734,
      "grad_norm": 0.4189848005771637,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 3523
    },
    {
      "epoch": 0.3808494542310602,
      "grad_norm": 0.4441189169883728,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 3524
    },
    {
      "epoch": 0.380957527288447,
      "grad_norm": 0.39134642481803894,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 3525
    },
    {
      "epoch": 0.3810656003458338,
      "grad_norm": 0.449124813079834,
      "learning_rate": 8e-05,
      "loss": 1.8654,
      "step": 3526
    },
    {
      "epoch": 0.38117367340322056,
      "grad_norm": 0.4344721734523773,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 3527
    },
    {
      "epoch": 0.38128174646060736,
      "grad_norm": 0.41393086314201355,
      "learning_rate": 8e-05,
      "loss": 1.791,
      "step": 3528
    },
    {
      "epoch": 0.38138981951799417,
      "grad_norm": 0.4432050585746765,
      "learning_rate": 8e-05,
      "loss": 1.8722,
      "step": 3529
    },
    {
      "epoch": 0.38149789257538097,
      "grad_norm": 0.4568476378917694,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 3530
    },
    {
      "epoch": 0.3816059656327678,
      "grad_norm": 0.4061969816684723,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 3531
    },
    {
      "epoch": 0.3817140386901545,
      "grad_norm": 0.41773661971092224,
      "learning_rate": 8e-05,
      "loss": 1.7578,
      "step": 3532
    },
    {
      "epoch": 0.3818221117475413,
      "grad_norm": 0.41755226254463196,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 3533
    },
    {
      "epoch": 0.38193018480492813,
      "grad_norm": 0.4604392647743225,
      "learning_rate": 8e-05,
      "loss": 1.7716,
      "step": 3534
    },
    {
      "epoch": 0.38203825786231493,
      "grad_norm": 0.4502955675125122,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 3535
    },
    {
      "epoch": 0.38214633091970174,
      "grad_norm": 0.4575846791267395,
      "learning_rate": 8e-05,
      "loss": 1.7869,
      "step": 3536
    },
    {
      "epoch": 0.3822544039770885,
      "grad_norm": 0.40406206250190735,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 3537
    },
    {
      "epoch": 0.3823624770344753,
      "grad_norm": 0.4484672546386719,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 3538
    },
    {
      "epoch": 0.3824705500918621,
      "grad_norm": 0.40729930996894836,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 3539
    },
    {
      "epoch": 0.3825786231492489,
      "grad_norm": 0.4250151515007019,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 3540
    },
    {
      "epoch": 0.3826866962066357,
      "grad_norm": 0.4275742471218109,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 3541
    },
    {
      "epoch": 0.3827947692640225,
      "grad_norm": 0.4672510623931885,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 3542
    },
    {
      "epoch": 0.38290284232140925,
      "grad_norm": 0.42522722482681274,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 3543
    },
    {
      "epoch": 0.38301091537879606,
      "grad_norm": 0.44650694727897644,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 3544
    },
    {
      "epoch": 0.38311898843618286,
      "grad_norm": 0.4204098582267761,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 3545
    },
    {
      "epoch": 0.38322706149356967,
      "grad_norm": 0.47380244731903076,
      "learning_rate": 8e-05,
      "loss": 1.8292,
      "step": 3546
    },
    {
      "epoch": 0.38333513455095647,
      "grad_norm": 0.4091149866580963,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 3547
    },
    {
      "epoch": 0.3834432076083432,
      "grad_norm": 0.40014055371284485,
      "learning_rate": 8e-05,
      "loss": 1.6728,
      "step": 3548
    },
    {
      "epoch": 0.38355128066573,
      "grad_norm": 0.44214335083961487,
      "learning_rate": 8e-05,
      "loss": 1.8072,
      "step": 3549
    },
    {
      "epoch": 0.3836593537231168,
      "grad_norm": 0.4597921073436737,
      "learning_rate": 8e-05,
      "loss": 1.8648,
      "step": 3550
    },
    {
      "epoch": 0.38376742678050363,
      "grad_norm": 0.4153591990470886,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 3551
    },
    {
      "epoch": 0.38387549983789043,
      "grad_norm": 0.43814411759376526,
      "learning_rate": 8e-05,
      "loss": 1.8351,
      "step": 3552
    },
    {
      "epoch": 0.3839835728952772,
      "grad_norm": 0.41224783658981323,
      "learning_rate": 8e-05,
      "loss": 1.8161,
      "step": 3553
    },
    {
      "epoch": 0.384091645952664,
      "grad_norm": 0.44498416781425476,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 3554
    },
    {
      "epoch": 0.3841997190100508,
      "grad_norm": 0.45412561297416687,
      "learning_rate": 8e-05,
      "loss": 1.8389,
      "step": 3555
    },
    {
      "epoch": 0.3843077920674376,
      "grad_norm": 0.48190823197364807,
      "learning_rate": 8e-05,
      "loss": 1.9023,
      "step": 3556
    },
    {
      "epoch": 0.3844158651248244,
      "grad_norm": 0.4716581106185913,
      "learning_rate": 8e-05,
      "loss": 1.9739,
      "step": 3557
    },
    {
      "epoch": 0.3845239381822112,
      "grad_norm": 0.40748459100723267,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 3558
    },
    {
      "epoch": 0.38463201123959795,
      "grad_norm": 0.43396174907684326,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 3559
    },
    {
      "epoch": 0.38474008429698475,
      "grad_norm": 0.42178258299827576,
      "learning_rate": 8e-05,
      "loss": 1.7464,
      "step": 3560
    },
    {
      "epoch": 0.38484815735437156,
      "grad_norm": 0.43012893199920654,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 3561
    },
    {
      "epoch": 0.38495623041175836,
      "grad_norm": 0.44692155718803406,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 3562
    },
    {
      "epoch": 0.38506430346914516,
      "grad_norm": 0.4929838478565216,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 3563
    },
    {
      "epoch": 0.3851723765265319,
      "grad_norm": 0.4478960335254669,
      "learning_rate": 8e-05,
      "loss": 1.8294,
      "step": 3564
    },
    {
      "epoch": 0.3852804495839187,
      "grad_norm": 0.4289861023426056,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 3565
    },
    {
      "epoch": 0.3853885226413055,
      "grad_norm": 0.41241443157196045,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 3566
    },
    {
      "epoch": 0.3854965956986923,
      "grad_norm": 0.4765373170375824,
      "learning_rate": 8e-05,
      "loss": 1.783,
      "step": 3567
    },
    {
      "epoch": 0.3856046687560791,
      "grad_norm": 0.4731141924858093,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 3568
    },
    {
      "epoch": 0.3857127418134659,
      "grad_norm": 0.442623108625412,
      "learning_rate": 8e-05,
      "loss": 1.8262,
      "step": 3569
    },
    {
      "epoch": 0.3858208148708527,
      "grad_norm": 0.43833914399147034,
      "learning_rate": 8e-05,
      "loss": 1.764,
      "step": 3570
    },
    {
      "epoch": 0.3859288879282395,
      "grad_norm": 0.4178556203842163,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 3571
    },
    {
      "epoch": 0.3860369609856263,
      "grad_norm": 0.4163842499256134,
      "learning_rate": 8e-05,
      "loss": 1.7092,
      "step": 3572
    },
    {
      "epoch": 0.3861450340430131,
      "grad_norm": 0.4549366235733032,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 3573
    },
    {
      "epoch": 0.3862531071003999,
      "grad_norm": 0.4554690718650818,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 3574
    },
    {
      "epoch": 0.38636118015778664,
      "grad_norm": 0.44290637969970703,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 3575
    },
    {
      "epoch": 0.38646925321517345,
      "grad_norm": 0.4339907169342041,
      "learning_rate": 8e-05,
      "loss": 1.7072,
      "step": 3576
    },
    {
      "epoch": 0.38657732627256025,
      "grad_norm": 0.4544125497341156,
      "learning_rate": 8e-05,
      "loss": 1.9069,
      "step": 3577
    },
    {
      "epoch": 0.38668539932994705,
      "grad_norm": 0.41420575976371765,
      "learning_rate": 8e-05,
      "loss": 1.4907,
      "step": 3578
    },
    {
      "epoch": 0.38679347238733386,
      "grad_norm": 0.4325450658798218,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 3579
    },
    {
      "epoch": 0.3869015454447206,
      "grad_norm": 0.4321989417076111,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 3580
    },
    {
      "epoch": 0.3870096185021074,
      "grad_norm": 0.3936842381954193,
      "learning_rate": 8e-05,
      "loss": 1.5363,
      "step": 3581
    },
    {
      "epoch": 0.3871176915594942,
      "grad_norm": 0.4554964005947113,
      "learning_rate": 8e-05,
      "loss": 1.6674,
      "step": 3582
    },
    {
      "epoch": 0.387225764616881,
      "grad_norm": 0.43081119656562805,
      "learning_rate": 8e-05,
      "loss": 1.7611,
      "step": 3583
    },
    {
      "epoch": 0.3873338376742678,
      "grad_norm": 0.4263867139816284,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 3584
    },
    {
      "epoch": 0.3874419107316546,
      "grad_norm": 0.4466182589530945,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 3585
    },
    {
      "epoch": 0.3875499837890414,
      "grad_norm": 0.4231431186199188,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 3586
    },
    {
      "epoch": 0.3876580568464282,
      "grad_norm": 0.4337088167667389,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 3587
    },
    {
      "epoch": 0.387766129903815,
      "grad_norm": 0.4360045790672302,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 3588
    },
    {
      "epoch": 0.3878742029612018,
      "grad_norm": 0.48119670152664185,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 3589
    },
    {
      "epoch": 0.3879822760185886,
      "grad_norm": 0.4417213797569275,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 3590
    },
    {
      "epoch": 0.38809034907597534,
      "grad_norm": 0.42232853174209595,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 3591
    },
    {
      "epoch": 0.38819842213336214,
      "grad_norm": 0.4357171654701233,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 3592
    },
    {
      "epoch": 0.38830649519074895,
      "grad_norm": 0.43087711930274963,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 3593
    },
    {
      "epoch": 0.38841456824813575,
      "grad_norm": 0.4553661346435547,
      "learning_rate": 8e-05,
      "loss": 1.8452,
      "step": 3594
    },
    {
      "epoch": 0.38852264130552255,
      "grad_norm": 0.40859055519104004,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 3595
    },
    {
      "epoch": 0.3886307143629093,
      "grad_norm": 0.46088337898254395,
      "learning_rate": 8e-05,
      "loss": 1.7633,
      "step": 3596
    },
    {
      "epoch": 0.3887387874202961,
      "grad_norm": 0.4466537535190582,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 3597
    },
    {
      "epoch": 0.3888468604776829,
      "grad_norm": 0.49543100595474243,
      "learning_rate": 8e-05,
      "loss": 1.946,
      "step": 3598
    },
    {
      "epoch": 0.3889549335350697,
      "grad_norm": 0.4200584888458252,
      "learning_rate": 8e-05,
      "loss": 1.7606,
      "step": 3599
    },
    {
      "epoch": 0.3890630065924565,
      "grad_norm": 0.4552920162677765,
      "learning_rate": 8e-05,
      "loss": 1.6724,
      "step": 3600
    },
    {
      "epoch": 0.3891710796498433,
      "grad_norm": 0.5082268714904785,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 3601
    },
    {
      "epoch": 0.38927915270723007,
      "grad_norm": 0.41033533215522766,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 3602
    },
    {
      "epoch": 0.3893872257646169,
      "grad_norm": 0.4281781017780304,
      "learning_rate": 8e-05,
      "loss": 1.7878,
      "step": 3603
    },
    {
      "epoch": 0.3894952988220037,
      "grad_norm": 0.4289511442184448,
      "learning_rate": 8e-05,
      "loss": 1.6703,
      "step": 3604
    },
    {
      "epoch": 0.3896033718793905,
      "grad_norm": 0.46695029735565186,
      "learning_rate": 8e-05,
      "loss": 1.6638,
      "step": 3605
    },
    {
      "epoch": 0.3897114449367773,
      "grad_norm": 0.4272889792919159,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 3606
    },
    {
      "epoch": 0.38981951799416403,
      "grad_norm": 0.4509229063987732,
      "learning_rate": 8e-05,
      "loss": 1.8257,
      "step": 3607
    },
    {
      "epoch": 0.38992759105155084,
      "grad_norm": 0.4530782103538513,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 3608
    },
    {
      "epoch": 0.39003566410893764,
      "grad_norm": 0.4436373710632324,
      "learning_rate": 8e-05,
      "loss": 1.7559,
      "step": 3609
    },
    {
      "epoch": 0.39014373716632444,
      "grad_norm": 0.5030590891838074,
      "learning_rate": 8e-05,
      "loss": 1.8553,
      "step": 3610
    },
    {
      "epoch": 0.39025181022371125,
      "grad_norm": 0.436802476644516,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 3611
    },
    {
      "epoch": 0.390359883281098,
      "grad_norm": 0.46174871921539307,
      "learning_rate": 8e-05,
      "loss": 1.4905,
      "step": 3612
    },
    {
      "epoch": 0.3904679563384848,
      "grad_norm": 0.4613874554634094,
      "learning_rate": 8e-05,
      "loss": 1.8844,
      "step": 3613
    },
    {
      "epoch": 0.3905760293958716,
      "grad_norm": 0.4229370355606079,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 3614
    },
    {
      "epoch": 0.3906841024532584,
      "grad_norm": 0.38881102204322815,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 3615
    },
    {
      "epoch": 0.3907921755106452,
      "grad_norm": 0.4257299304008484,
      "learning_rate": 8e-05,
      "loss": 1.757,
      "step": 3616
    },
    {
      "epoch": 0.390900248568032,
      "grad_norm": 0.4028431475162506,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 3617
    },
    {
      "epoch": 0.39100832162541876,
      "grad_norm": 0.46209847927093506,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 3618
    },
    {
      "epoch": 0.39111639468280557,
      "grad_norm": 0.45247483253479004,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 3619
    },
    {
      "epoch": 0.39122446774019237,
      "grad_norm": 0.4258858859539032,
      "learning_rate": 8e-05,
      "loss": 1.8033,
      "step": 3620
    },
    {
      "epoch": 0.3913325407975792,
      "grad_norm": 0.4167170822620392,
      "learning_rate": 8e-05,
      "loss": 1.6176,
      "step": 3621
    },
    {
      "epoch": 0.391440613854966,
      "grad_norm": 0.4460950195789337,
      "learning_rate": 8e-05,
      "loss": 1.7562,
      "step": 3622
    },
    {
      "epoch": 0.39154868691235273,
      "grad_norm": 0.4062543511390686,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 3623
    },
    {
      "epoch": 0.39165675996973953,
      "grad_norm": 0.42487162351608276,
      "learning_rate": 8e-05,
      "loss": 1.6488,
      "step": 3624
    },
    {
      "epoch": 0.39176483302712634,
      "grad_norm": 0.4117380380630493,
      "learning_rate": 8e-05,
      "loss": 1.6161,
      "step": 3625
    },
    {
      "epoch": 0.39187290608451314,
      "grad_norm": 0.4491898715496063,
      "learning_rate": 8e-05,
      "loss": 1.7866,
      "step": 3626
    },
    {
      "epoch": 0.39198097914189994,
      "grad_norm": 0.4315648674964905,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 3627
    },
    {
      "epoch": 0.3920890521992867,
      "grad_norm": 0.43977344036102295,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 3628
    },
    {
      "epoch": 0.3921971252566735,
      "grad_norm": 0.45263731479644775,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 3629
    },
    {
      "epoch": 0.3923051983140603,
      "grad_norm": 0.42903581261634827,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 3630
    },
    {
      "epoch": 0.3924132713714471,
      "grad_norm": 0.4342464506626129,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 3631
    },
    {
      "epoch": 0.3925213444288339,
      "grad_norm": 0.45859020948410034,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 3632
    },
    {
      "epoch": 0.3926294174862207,
      "grad_norm": 0.4312132000923157,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 3633
    },
    {
      "epoch": 0.39273749054360746,
      "grad_norm": 0.464608371257782,
      "learning_rate": 8e-05,
      "loss": 1.7043,
      "step": 3634
    },
    {
      "epoch": 0.39284556360099426,
      "grad_norm": 0.4446944296360016,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 3635
    },
    {
      "epoch": 0.39295363665838107,
      "grad_norm": 0.4601576626300812,
      "learning_rate": 8e-05,
      "loss": 1.9314,
      "step": 3636
    },
    {
      "epoch": 0.39306170971576787,
      "grad_norm": 0.445232093334198,
      "learning_rate": 8e-05,
      "loss": 1.8512,
      "step": 3637
    },
    {
      "epoch": 0.3931697827731547,
      "grad_norm": 0.5111165046691895,
      "learning_rate": 8e-05,
      "loss": 1.9864,
      "step": 3638
    },
    {
      "epoch": 0.3932778558305414,
      "grad_norm": 0.4238782823085785,
      "learning_rate": 8e-05,
      "loss": 1.8263,
      "step": 3639
    },
    {
      "epoch": 0.3933859288879282,
      "grad_norm": 0.42111656069755554,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 3640
    },
    {
      "epoch": 0.39349400194531503,
      "grad_norm": 0.4486598074436188,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 3641
    },
    {
      "epoch": 0.39360207500270183,
      "grad_norm": 0.462300568819046,
      "learning_rate": 8e-05,
      "loss": 1.7249,
      "step": 3642
    },
    {
      "epoch": 0.39371014806008864,
      "grad_norm": 0.46267035603523254,
      "learning_rate": 8e-05,
      "loss": 1.6619,
      "step": 3643
    },
    {
      "epoch": 0.3938182211174754,
      "grad_norm": 0.422735333442688,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 3644
    },
    {
      "epoch": 0.3939262941748622,
      "grad_norm": 0.4415607750415802,
      "learning_rate": 8e-05,
      "loss": 1.6748,
      "step": 3645
    },
    {
      "epoch": 0.394034367232249,
      "grad_norm": 0.4602021872997284,
      "learning_rate": 8e-05,
      "loss": 1.8811,
      "step": 3646
    },
    {
      "epoch": 0.3941424402896358,
      "grad_norm": 0.4422406554222107,
      "learning_rate": 8e-05,
      "loss": 1.7879,
      "step": 3647
    },
    {
      "epoch": 0.3942505133470226,
      "grad_norm": 0.40833306312561035,
      "learning_rate": 8e-05,
      "loss": 1.5864,
      "step": 3648
    },
    {
      "epoch": 0.3943585864044094,
      "grad_norm": 0.4625471532344818,
      "learning_rate": 8e-05,
      "loss": 1.8937,
      "step": 3649
    },
    {
      "epoch": 0.39446665946179615,
      "grad_norm": 0.43749943375587463,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 3650
    },
    {
      "epoch": 0.39457473251918296,
      "grad_norm": 0.417495459318161,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 3651
    },
    {
      "epoch": 0.39468280557656976,
      "grad_norm": 0.41670066118240356,
      "learning_rate": 8e-05,
      "loss": 1.703,
      "step": 3652
    },
    {
      "epoch": 0.39479087863395657,
      "grad_norm": 0.45128265023231506,
      "learning_rate": 8e-05,
      "loss": 1.8139,
      "step": 3653
    },
    {
      "epoch": 0.39489895169134337,
      "grad_norm": 0.4394649863243103,
      "learning_rate": 8e-05,
      "loss": 1.744,
      "step": 3654
    },
    {
      "epoch": 0.3950070247487301,
      "grad_norm": 0.45002442598342896,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 3655
    },
    {
      "epoch": 0.3951150978061169,
      "grad_norm": 0.42135199904441833,
      "learning_rate": 8e-05,
      "loss": 1.8687,
      "step": 3656
    },
    {
      "epoch": 0.3952231708635037,
      "grad_norm": 0.4521159529685974,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 3657
    },
    {
      "epoch": 0.39533124392089053,
      "grad_norm": 0.4786424934864044,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 3658
    },
    {
      "epoch": 0.39543931697827733,
      "grad_norm": 0.4152093231678009,
      "learning_rate": 8e-05,
      "loss": 1.6627,
      "step": 3659
    },
    {
      "epoch": 0.3955473900356641,
      "grad_norm": 0.3987724184989929,
      "learning_rate": 8e-05,
      "loss": 1.6459,
      "step": 3660
    },
    {
      "epoch": 0.3956554630930509,
      "grad_norm": 0.41067513823509216,
      "learning_rate": 8e-05,
      "loss": 1.4181,
      "step": 3661
    },
    {
      "epoch": 0.3957635361504377,
      "grad_norm": 0.41946113109588623,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 3662
    },
    {
      "epoch": 0.3958716092078245,
      "grad_norm": 0.4285235106945038,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 3663
    },
    {
      "epoch": 0.3959796822652113,
      "grad_norm": 0.43019378185272217,
      "learning_rate": 8e-05,
      "loss": 1.6767,
      "step": 3664
    },
    {
      "epoch": 0.3960877553225981,
      "grad_norm": 0.45729175209999084,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 3665
    },
    {
      "epoch": 0.39619582837998485,
      "grad_norm": 0.4403662085533142,
      "learning_rate": 8e-05,
      "loss": 1.9084,
      "step": 3666
    },
    {
      "epoch": 0.39630390143737165,
      "grad_norm": 0.4186180830001831,
      "learning_rate": 8e-05,
      "loss": 1.7596,
      "step": 3667
    },
    {
      "epoch": 0.39641197449475846,
      "grad_norm": 0.4136730432510376,
      "learning_rate": 8e-05,
      "loss": 1.6993,
      "step": 3668
    },
    {
      "epoch": 0.39652004755214526,
      "grad_norm": 0.4288462698459625,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 3669
    },
    {
      "epoch": 0.39662812060953206,
      "grad_norm": 0.44506528973579407,
      "learning_rate": 8e-05,
      "loss": 1.827,
      "step": 3670
    },
    {
      "epoch": 0.3967361936669188,
      "grad_norm": 0.4478132426738739,
      "learning_rate": 8e-05,
      "loss": 1.7312,
      "step": 3671
    },
    {
      "epoch": 0.3968442667243056,
      "grad_norm": 0.4010065793991089,
      "learning_rate": 8e-05,
      "loss": 1.6775,
      "step": 3672
    },
    {
      "epoch": 0.3969523397816924,
      "grad_norm": 0.42949843406677246,
      "learning_rate": 8e-05,
      "loss": 1.7515,
      "step": 3673
    },
    {
      "epoch": 0.3970604128390792,
      "grad_norm": 0.48675793409347534,
      "learning_rate": 8e-05,
      "loss": 1.9118,
      "step": 3674
    },
    {
      "epoch": 0.39716848589646603,
      "grad_norm": 0.4166284203529358,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 3675
    },
    {
      "epoch": 0.39727655895385283,
      "grad_norm": 0.472807914018631,
      "learning_rate": 8e-05,
      "loss": 1.9306,
      "step": 3676
    },
    {
      "epoch": 0.3973846320112396,
      "grad_norm": 0.4107438623905182,
      "learning_rate": 8e-05,
      "loss": 1.4809,
      "step": 3677
    },
    {
      "epoch": 0.3974927050686264,
      "grad_norm": 0.4212261438369751,
      "learning_rate": 8e-05,
      "loss": 1.6445,
      "step": 3678
    },
    {
      "epoch": 0.3976007781260132,
      "grad_norm": 0.4266141355037689,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 3679
    },
    {
      "epoch": 0.3977088511834,
      "grad_norm": 0.4156089425086975,
      "learning_rate": 8e-05,
      "loss": 1.855,
      "step": 3680
    },
    {
      "epoch": 0.3978169242407868,
      "grad_norm": 0.44603171944618225,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 3681
    },
    {
      "epoch": 0.39792499729817354,
      "grad_norm": 0.4508771002292633,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 3682
    },
    {
      "epoch": 0.39803307035556035,
      "grad_norm": 0.4938910901546478,
      "learning_rate": 8e-05,
      "loss": 1.7327,
      "step": 3683
    },
    {
      "epoch": 0.39814114341294715,
      "grad_norm": 0.41987448930740356,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 3684
    },
    {
      "epoch": 0.39824921647033396,
      "grad_norm": 0.5075860023498535,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 3685
    },
    {
      "epoch": 0.39835728952772076,
      "grad_norm": 0.4174606502056122,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 3686
    },
    {
      "epoch": 0.3984653625851075,
      "grad_norm": 0.41603541374206543,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 3687
    },
    {
      "epoch": 0.3985734356424943,
      "grad_norm": 0.4386812448501587,
      "learning_rate": 8e-05,
      "loss": 1.894,
      "step": 3688
    },
    {
      "epoch": 0.3986815086998811,
      "grad_norm": 0.43827855587005615,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 3689
    },
    {
      "epoch": 0.3987895817572679,
      "grad_norm": 0.416027307510376,
      "learning_rate": 8e-05,
      "loss": 1.6433,
      "step": 3690
    },
    {
      "epoch": 0.3988976548146547,
      "grad_norm": 0.43987277150154114,
      "learning_rate": 8e-05,
      "loss": 1.8197,
      "step": 3691
    },
    {
      "epoch": 0.3990057278720415,
      "grad_norm": 0.39380329847335815,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 3692
    },
    {
      "epoch": 0.3991138009294283,
      "grad_norm": 0.4272485375404358,
      "learning_rate": 8e-05,
      "loss": 1.6737,
      "step": 3693
    },
    {
      "epoch": 0.3992218739868151,
      "grad_norm": 0.43056023120880127,
      "learning_rate": 8e-05,
      "loss": 1.8267,
      "step": 3694
    },
    {
      "epoch": 0.3993299470442019,
      "grad_norm": 0.45389553904533386,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 3695
    },
    {
      "epoch": 0.3994380201015887,
      "grad_norm": 0.43354108929634094,
      "learning_rate": 8e-05,
      "loss": 1.5589,
      "step": 3696
    },
    {
      "epoch": 0.3995460931589755,
      "grad_norm": 0.41877955198287964,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 3697
    },
    {
      "epoch": 0.39965416621636224,
      "grad_norm": 0.44419270753860474,
      "learning_rate": 8e-05,
      "loss": 1.6616,
      "step": 3698
    },
    {
      "epoch": 0.39976223927374904,
      "grad_norm": 0.4599677324295044,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 3699
    },
    {
      "epoch": 0.39987031233113585,
      "grad_norm": 0.4581482410430908,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 3700
    },
    {
      "epoch": 0.39997838538852265,
      "grad_norm": 0.4417431950569153,
      "learning_rate": 8e-05,
      "loss": 1.8308,
      "step": 3701
    },
    {
      "epoch": 0.40008645844590945,
      "grad_norm": 0.4106607735157013,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 3702
    },
    {
      "epoch": 0.4001945315032962,
      "grad_norm": 0.4542486369609833,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 3703
    },
    {
      "epoch": 0.400302604560683,
      "grad_norm": 0.4553987383842468,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 3704
    },
    {
      "epoch": 0.4004106776180698,
      "grad_norm": 0.411516010761261,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 3705
    },
    {
      "epoch": 0.4005187506754566,
      "grad_norm": 0.4551837146282196,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 3706
    },
    {
      "epoch": 0.4006268237328434,
      "grad_norm": 0.46361780166625977,
      "learning_rate": 8e-05,
      "loss": 1.7095,
      "step": 3707
    },
    {
      "epoch": 0.4007348967902302,
      "grad_norm": 0.4496322572231293,
      "learning_rate": 8e-05,
      "loss": 1.6008,
      "step": 3708
    },
    {
      "epoch": 0.40084296984761697,
      "grad_norm": 0.4417829215526581,
      "learning_rate": 8e-05,
      "loss": 1.6189,
      "step": 3709
    },
    {
      "epoch": 0.4009510429050038,
      "grad_norm": 0.45281141996383667,
      "learning_rate": 8e-05,
      "loss": 1.6365,
      "step": 3710
    },
    {
      "epoch": 0.4010591159623906,
      "grad_norm": 0.4428302049636841,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 3711
    },
    {
      "epoch": 0.4011671890197774,
      "grad_norm": 0.4292880594730377,
      "learning_rate": 8e-05,
      "loss": 1.6654,
      "step": 3712
    },
    {
      "epoch": 0.4012752620771642,
      "grad_norm": 0.3965036869049072,
      "learning_rate": 8e-05,
      "loss": 1.6369,
      "step": 3713
    },
    {
      "epoch": 0.40138333513455093,
      "grad_norm": 0.4361380636692047,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 3714
    },
    {
      "epoch": 0.40149140819193774,
      "grad_norm": 0.4497394263744354,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 3715
    },
    {
      "epoch": 0.40159948124932454,
      "grad_norm": 0.4493385851383209,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 3716
    },
    {
      "epoch": 0.40170755430671135,
      "grad_norm": 0.45345962047576904,
      "learning_rate": 8e-05,
      "loss": 1.7042,
      "step": 3717
    },
    {
      "epoch": 0.40181562736409815,
      "grad_norm": 0.5164238214492798,
      "learning_rate": 8e-05,
      "loss": 1.7897,
      "step": 3718
    },
    {
      "epoch": 0.4019237004214849,
      "grad_norm": 0.43206316232681274,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 3719
    },
    {
      "epoch": 0.4020317734788717,
      "grad_norm": 0.45956137776374817,
      "learning_rate": 8e-05,
      "loss": 1.7934,
      "step": 3720
    },
    {
      "epoch": 0.4021398465362585,
      "grad_norm": 0.415053129196167,
      "learning_rate": 8e-05,
      "loss": 1.5853,
      "step": 3721
    },
    {
      "epoch": 0.4022479195936453,
      "grad_norm": 0.3946075439453125,
      "learning_rate": 8e-05,
      "loss": 1.567,
      "step": 3722
    },
    {
      "epoch": 0.4023559926510321,
      "grad_norm": 0.41574814915657043,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 3723
    },
    {
      "epoch": 0.4024640657084189,
      "grad_norm": 0.3900175392627716,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 3724
    },
    {
      "epoch": 0.40257213876580566,
      "grad_norm": 0.4094412922859192,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 3725
    },
    {
      "epoch": 0.40268021182319247,
      "grad_norm": 0.437096506357193,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 3726
    },
    {
      "epoch": 0.4027882848805793,
      "grad_norm": 0.42509353160858154,
      "learning_rate": 8e-05,
      "loss": 1.8651,
      "step": 3727
    },
    {
      "epoch": 0.4028963579379661,
      "grad_norm": 0.42150071263313293,
      "learning_rate": 8e-05,
      "loss": 1.6523,
      "step": 3728
    },
    {
      "epoch": 0.4030044309953529,
      "grad_norm": 0.4493981599807739,
      "learning_rate": 8e-05,
      "loss": 1.8487,
      "step": 3729
    },
    {
      "epoch": 0.40311250405273963,
      "grad_norm": 0.4103095829486847,
      "learning_rate": 8e-05,
      "loss": 1.6702,
      "step": 3730
    },
    {
      "epoch": 0.40322057711012643,
      "grad_norm": 0.40782836079597473,
      "learning_rate": 8e-05,
      "loss": 1.6887,
      "step": 3731
    },
    {
      "epoch": 0.40332865016751324,
      "grad_norm": 0.4058775007724762,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 3732
    },
    {
      "epoch": 0.40343672322490004,
      "grad_norm": 0.48961141705513,
      "learning_rate": 8e-05,
      "loss": 1.9858,
      "step": 3733
    },
    {
      "epoch": 0.40354479628228684,
      "grad_norm": 0.49177929759025574,
      "learning_rate": 8e-05,
      "loss": 1.8574,
      "step": 3734
    },
    {
      "epoch": 0.4036528693396736,
      "grad_norm": 0.4561348259449005,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 3735
    },
    {
      "epoch": 0.4037609423970604,
      "grad_norm": 0.46109312772750854,
      "learning_rate": 8e-05,
      "loss": 1.9005,
      "step": 3736
    },
    {
      "epoch": 0.4038690154544472,
      "grad_norm": 0.4023257791996002,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 3737
    },
    {
      "epoch": 0.403977088511834,
      "grad_norm": 0.41945680975914,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 3738
    },
    {
      "epoch": 0.4040851615692208,
      "grad_norm": 0.41501882672309875,
      "learning_rate": 8e-05,
      "loss": 1.6801,
      "step": 3739
    },
    {
      "epoch": 0.4041932346266076,
      "grad_norm": 0.455325722694397,
      "learning_rate": 8e-05,
      "loss": 1.9328,
      "step": 3740
    },
    {
      "epoch": 0.40430130768399436,
      "grad_norm": 0.4204757511615753,
      "learning_rate": 8e-05,
      "loss": 1.5771,
      "step": 3741
    },
    {
      "epoch": 0.40440938074138116,
      "grad_norm": 0.4104888439178467,
      "learning_rate": 8e-05,
      "loss": 1.6497,
      "step": 3742
    },
    {
      "epoch": 0.40451745379876797,
      "grad_norm": 0.42231041193008423,
      "learning_rate": 8e-05,
      "loss": 1.7915,
      "step": 3743
    },
    {
      "epoch": 0.40462552685615477,
      "grad_norm": 0.427582710981369,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 3744
    },
    {
      "epoch": 0.4047335999135416,
      "grad_norm": 0.4144611656665802,
      "learning_rate": 8e-05,
      "loss": 1.8426,
      "step": 3745
    },
    {
      "epoch": 0.4048416729709283,
      "grad_norm": 0.42702651023864746,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 3746
    },
    {
      "epoch": 0.4049497460283151,
      "grad_norm": 0.41470932960510254,
      "learning_rate": 8e-05,
      "loss": 1.5833,
      "step": 3747
    },
    {
      "epoch": 0.40505781908570193,
      "grad_norm": 0.38062769174575806,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 3748
    },
    {
      "epoch": 0.40516589214308874,
      "grad_norm": 0.41118744015693665,
      "learning_rate": 8e-05,
      "loss": 1.638,
      "step": 3749
    },
    {
      "epoch": 0.40527396520047554,
      "grad_norm": 0.40478622913360596,
      "learning_rate": 8e-05,
      "loss": 1.7724,
      "step": 3750
    },
    {
      "epoch": 0.4053820382578623,
      "grad_norm": 0.44077643752098083,
      "learning_rate": 8e-05,
      "loss": 1.4922,
      "step": 3751
    },
    {
      "epoch": 0.4054901113152491,
      "grad_norm": 0.4371686577796936,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 3752
    },
    {
      "epoch": 0.4055981843726359,
      "grad_norm": 0.4593043029308319,
      "learning_rate": 8e-05,
      "loss": 1.7918,
      "step": 3753
    },
    {
      "epoch": 0.4057062574300227,
      "grad_norm": 0.4050823450088501,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 3754
    },
    {
      "epoch": 0.4058143304874095,
      "grad_norm": 0.4197343587875366,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 3755
    },
    {
      "epoch": 0.4059224035447963,
      "grad_norm": 0.41836997866630554,
      "learning_rate": 8e-05,
      "loss": 1.6693,
      "step": 3756
    },
    {
      "epoch": 0.40603047660218305,
      "grad_norm": 0.4243151843547821,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 3757
    },
    {
      "epoch": 0.40613854965956986,
      "grad_norm": 0.4050559997558594,
      "learning_rate": 8e-05,
      "loss": 1.5782,
      "step": 3758
    },
    {
      "epoch": 0.40624662271695666,
      "grad_norm": 0.4481789171695709,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 3759
    },
    {
      "epoch": 0.40635469577434347,
      "grad_norm": 0.4258626103401184,
      "learning_rate": 8e-05,
      "loss": 1.7628,
      "step": 3760
    },
    {
      "epoch": 0.40646276883173027,
      "grad_norm": 0.46520403027534485,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 3761
    },
    {
      "epoch": 0.406570841889117,
      "grad_norm": 0.47400322556495667,
      "learning_rate": 8e-05,
      "loss": 1.7677,
      "step": 3762
    },
    {
      "epoch": 0.4066789149465038,
      "grad_norm": 0.4666513502597809,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 3763
    },
    {
      "epoch": 0.4067869880038906,
      "grad_norm": 0.4544326663017273,
      "learning_rate": 8e-05,
      "loss": 1.7911,
      "step": 3764
    },
    {
      "epoch": 0.40689506106127743,
      "grad_norm": 0.43219226598739624,
      "learning_rate": 8e-05,
      "loss": 1.8022,
      "step": 3765
    },
    {
      "epoch": 0.40700313411866423,
      "grad_norm": 0.4155332148075104,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 3766
    },
    {
      "epoch": 0.40711120717605104,
      "grad_norm": 0.4450971186161041,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 3767
    },
    {
      "epoch": 0.4072192802334378,
      "grad_norm": 0.4451694190502167,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 3768
    },
    {
      "epoch": 0.4073273532908246,
      "grad_norm": 0.4120102524757385,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 3769
    },
    {
      "epoch": 0.4074354263482114,
      "grad_norm": 0.423669695854187,
      "learning_rate": 8e-05,
      "loss": 1.7084,
      "step": 3770
    },
    {
      "epoch": 0.4075434994055982,
      "grad_norm": 0.4332157373428345,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 3771
    },
    {
      "epoch": 0.407651572462985,
      "grad_norm": 0.40930598974227905,
      "learning_rate": 8e-05,
      "loss": 1.5603,
      "step": 3772
    },
    {
      "epoch": 0.40775964552037175,
      "grad_norm": 0.4044739007949829,
      "learning_rate": 8e-05,
      "loss": 1.6115,
      "step": 3773
    },
    {
      "epoch": 0.40786771857775855,
      "grad_norm": 0.4433550536632538,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 3774
    },
    {
      "epoch": 0.40797579163514536,
      "grad_norm": 0.4044461250305176,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 3775
    },
    {
      "epoch": 0.40808386469253216,
      "grad_norm": 0.43961507081985474,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 3776
    },
    {
      "epoch": 0.40819193774991896,
      "grad_norm": 0.41893184185028076,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 3777
    },
    {
      "epoch": 0.4083000108073057,
      "grad_norm": 0.4280010759830475,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 3778
    },
    {
      "epoch": 0.4084080838646925,
      "grad_norm": 0.44841232895851135,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 3779
    },
    {
      "epoch": 0.4085161569220793,
      "grad_norm": 0.44160908460617065,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 3780
    },
    {
      "epoch": 0.4086242299794661,
      "grad_norm": 0.44584354758262634,
      "learning_rate": 8e-05,
      "loss": 1.8265,
      "step": 3781
    },
    {
      "epoch": 0.40873230303685293,
      "grad_norm": 0.4068160057067871,
      "learning_rate": 8e-05,
      "loss": 1.5188,
      "step": 3782
    },
    {
      "epoch": 0.40884037609423973,
      "grad_norm": 0.42283895611763,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 3783
    },
    {
      "epoch": 0.4089484491516265,
      "grad_norm": 0.47673606872558594,
      "learning_rate": 8e-05,
      "loss": 1.9443,
      "step": 3784
    },
    {
      "epoch": 0.4090565222090133,
      "grad_norm": 0.46055489778518677,
      "learning_rate": 8e-05,
      "loss": 1.734,
      "step": 3785
    },
    {
      "epoch": 0.4091645952664001,
      "grad_norm": 0.4494338929653168,
      "learning_rate": 8e-05,
      "loss": 1.7578,
      "step": 3786
    },
    {
      "epoch": 0.4092726683237869,
      "grad_norm": 0.4553321897983551,
      "learning_rate": 8e-05,
      "loss": 1.8312,
      "step": 3787
    },
    {
      "epoch": 0.4093807413811737,
      "grad_norm": 0.44312456250190735,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 3788
    },
    {
      "epoch": 0.40948881443856044,
      "grad_norm": 0.41232019662857056,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 3789
    },
    {
      "epoch": 0.40959688749594725,
      "grad_norm": 0.43518000841140747,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 3790
    },
    {
      "epoch": 0.40970496055333405,
      "grad_norm": 0.41473284363746643,
      "learning_rate": 8e-05,
      "loss": 1.7222,
      "step": 3791
    },
    {
      "epoch": 0.40981303361072086,
      "grad_norm": 0.4263463020324707,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 3792
    },
    {
      "epoch": 0.40992110666810766,
      "grad_norm": 0.4522159695625305,
      "learning_rate": 8e-05,
      "loss": 1.7403,
      "step": 3793
    },
    {
      "epoch": 0.4100291797254944,
      "grad_norm": 0.45743653178215027,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 3794
    },
    {
      "epoch": 0.4101372527828812,
      "grad_norm": 0.47667717933654785,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 3795
    },
    {
      "epoch": 0.410245325840268,
      "grad_norm": 0.4799433648586273,
      "learning_rate": 8e-05,
      "loss": 1.7754,
      "step": 3796
    },
    {
      "epoch": 0.4103533988976548,
      "grad_norm": 0.5021691918373108,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 3797
    },
    {
      "epoch": 0.4104614719550416,
      "grad_norm": 0.4571269750595093,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 3798
    },
    {
      "epoch": 0.4105695450124284,
      "grad_norm": 0.4468599855899811,
      "learning_rate": 8e-05,
      "loss": 1.9898,
      "step": 3799
    },
    {
      "epoch": 0.4106776180698152,
      "grad_norm": 0.4466179311275482,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 3800
    },
    {
      "epoch": 0.410785691127202,
      "grad_norm": 0.4502711296081543,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 3801
    },
    {
      "epoch": 0.4108937641845888,
      "grad_norm": 0.4419698119163513,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 3802
    },
    {
      "epoch": 0.4110018372419756,
      "grad_norm": 0.41177424788475037,
      "learning_rate": 8e-05,
      "loss": 1.5969,
      "step": 3803
    },
    {
      "epoch": 0.4111099102993624,
      "grad_norm": 0.4002532362937927,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 3804
    },
    {
      "epoch": 0.41121798335674914,
      "grad_norm": 0.45052140951156616,
      "learning_rate": 8e-05,
      "loss": 1.7535,
      "step": 3805
    },
    {
      "epoch": 0.41132605641413594,
      "grad_norm": 0.4219045341014862,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 3806
    },
    {
      "epoch": 0.41143412947152275,
      "grad_norm": 0.44657668471336365,
      "learning_rate": 8e-05,
      "loss": 1.8121,
      "step": 3807
    },
    {
      "epoch": 0.41154220252890955,
      "grad_norm": 0.40233510732650757,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 3808
    },
    {
      "epoch": 0.41165027558629635,
      "grad_norm": 0.4653548002243042,
      "learning_rate": 8e-05,
      "loss": 1.9041,
      "step": 3809
    },
    {
      "epoch": 0.4117583486436831,
      "grad_norm": 0.41046056151390076,
      "learning_rate": 8e-05,
      "loss": 1.6583,
      "step": 3810
    },
    {
      "epoch": 0.4118664217010699,
      "grad_norm": 0.4101243019104004,
      "learning_rate": 8e-05,
      "loss": 1.6987,
      "step": 3811
    },
    {
      "epoch": 0.4119744947584567,
      "grad_norm": 0.452740877866745,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 3812
    },
    {
      "epoch": 0.4120825678158435,
      "grad_norm": 0.45041924715042114,
      "learning_rate": 8e-05,
      "loss": 1.6616,
      "step": 3813
    },
    {
      "epoch": 0.4121906408732303,
      "grad_norm": 0.4307399094104767,
      "learning_rate": 8e-05,
      "loss": 1.8094,
      "step": 3814
    },
    {
      "epoch": 0.4122987139306171,
      "grad_norm": 0.4191710650920868,
      "learning_rate": 8e-05,
      "loss": 1.6966,
      "step": 3815
    },
    {
      "epoch": 0.41240678698800387,
      "grad_norm": 0.424173504114151,
      "learning_rate": 8e-05,
      "loss": 1.6531,
      "step": 3816
    },
    {
      "epoch": 0.4125148600453907,
      "grad_norm": 0.41179588437080383,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 3817
    },
    {
      "epoch": 0.4126229331027775,
      "grad_norm": 0.43925076723098755,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 3818
    },
    {
      "epoch": 0.4127310061601643,
      "grad_norm": 0.4063188433647156,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 3819
    },
    {
      "epoch": 0.4128390792175511,
      "grad_norm": 0.4454279839992523,
      "learning_rate": 8e-05,
      "loss": 1.823,
      "step": 3820
    },
    {
      "epoch": 0.41294715227493783,
      "grad_norm": 0.4137673079967499,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 3821
    },
    {
      "epoch": 0.41305522533232464,
      "grad_norm": 0.45062750577926636,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 3822
    },
    {
      "epoch": 0.41316329838971144,
      "grad_norm": 0.4643123745918274,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 3823
    },
    {
      "epoch": 0.41327137144709825,
      "grad_norm": 0.446298748254776,
      "learning_rate": 8e-05,
      "loss": 1.8019,
      "step": 3824
    },
    {
      "epoch": 0.41337944450448505,
      "grad_norm": 0.4567425549030304,
      "learning_rate": 8e-05,
      "loss": 1.8072,
      "step": 3825
    },
    {
      "epoch": 0.4134875175618718,
      "grad_norm": 0.4653870463371277,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 3826
    },
    {
      "epoch": 0.4135955906192586,
      "grad_norm": 0.41211429238319397,
      "learning_rate": 8e-05,
      "loss": 1.4613,
      "step": 3827
    },
    {
      "epoch": 0.4137036636766454,
      "grad_norm": 0.48436239361763,
      "learning_rate": 8e-05,
      "loss": 1.5992,
      "step": 3828
    },
    {
      "epoch": 0.4138117367340322,
      "grad_norm": 0.4384808838367462,
      "learning_rate": 8e-05,
      "loss": 1.3596,
      "step": 3829
    },
    {
      "epoch": 0.413919809791419,
      "grad_norm": 0.4746789038181305,
      "learning_rate": 8e-05,
      "loss": 1.9285,
      "step": 3830
    },
    {
      "epoch": 0.4140278828488058,
      "grad_norm": 0.41284096240997314,
      "learning_rate": 8e-05,
      "loss": 1.6679,
      "step": 3831
    },
    {
      "epoch": 0.41413595590619257,
      "grad_norm": 0.46529728174209595,
      "learning_rate": 8e-05,
      "loss": 1.7672,
      "step": 3832
    },
    {
      "epoch": 0.41424402896357937,
      "grad_norm": 0.4134504497051239,
      "learning_rate": 8e-05,
      "loss": 1.6005,
      "step": 3833
    },
    {
      "epoch": 0.4143521020209662,
      "grad_norm": 0.4389565587043762,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 3834
    },
    {
      "epoch": 0.414460175078353,
      "grad_norm": 0.4391131103038788,
      "learning_rate": 8e-05,
      "loss": 1.738,
      "step": 3835
    },
    {
      "epoch": 0.4145682481357398,
      "grad_norm": 0.48091068863868713,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 3836
    },
    {
      "epoch": 0.41467632119312653,
      "grad_norm": 0.4540415108203888,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 3837
    },
    {
      "epoch": 0.41478439425051333,
      "grad_norm": 0.5060266852378845,
      "learning_rate": 8e-05,
      "loss": 1.8854,
      "step": 3838
    },
    {
      "epoch": 0.41489246730790014,
      "grad_norm": 0.4474870562553406,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 3839
    },
    {
      "epoch": 0.41500054036528694,
      "grad_norm": 0.4300170838832855,
      "learning_rate": 8e-05,
      "loss": 1.8997,
      "step": 3840
    },
    {
      "epoch": 0.41510861342267374,
      "grad_norm": 0.4334309697151184,
      "learning_rate": 8e-05,
      "loss": 1.6298,
      "step": 3841
    },
    {
      "epoch": 0.41521668648006055,
      "grad_norm": 0.5116660594940186,
      "learning_rate": 8e-05,
      "loss": 2.1035,
      "step": 3842
    },
    {
      "epoch": 0.4153247595374473,
      "grad_norm": 0.4035719633102417,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 3843
    },
    {
      "epoch": 0.4154328325948341,
      "grad_norm": 0.47815337777137756,
      "learning_rate": 8e-05,
      "loss": 1.8453,
      "step": 3844
    },
    {
      "epoch": 0.4155409056522209,
      "grad_norm": 0.466966450214386,
      "learning_rate": 8e-05,
      "loss": 1.7601,
      "step": 3845
    },
    {
      "epoch": 0.4156489787096077,
      "grad_norm": 0.4599195420742035,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 3846
    },
    {
      "epoch": 0.4157570517669945,
      "grad_norm": 0.4383247196674347,
      "learning_rate": 8e-05,
      "loss": 1.7776,
      "step": 3847
    },
    {
      "epoch": 0.41586512482438126,
      "grad_norm": 0.4247230291366577,
      "learning_rate": 8e-05,
      "loss": 1.7827,
      "step": 3848
    },
    {
      "epoch": 0.41597319788176806,
      "grad_norm": 0.48084914684295654,
      "learning_rate": 8e-05,
      "loss": 2.0259,
      "step": 3849
    },
    {
      "epoch": 0.41608127093915487,
      "grad_norm": 0.44634896516799927,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 3850
    },
    {
      "epoch": 0.41618934399654167,
      "grad_norm": 0.40854907035827637,
      "learning_rate": 8e-05,
      "loss": 1.6184,
      "step": 3851
    },
    {
      "epoch": 0.4162974170539285,
      "grad_norm": 0.4604286551475525,
      "learning_rate": 8e-05,
      "loss": 1.8213,
      "step": 3852
    },
    {
      "epoch": 0.4164054901113152,
      "grad_norm": 0.3957708179950714,
      "learning_rate": 8e-05,
      "loss": 1.5675,
      "step": 3853
    },
    {
      "epoch": 0.41651356316870203,
      "grad_norm": 0.42494916915893555,
      "learning_rate": 8e-05,
      "loss": 1.8354,
      "step": 3854
    },
    {
      "epoch": 0.41662163622608883,
      "grad_norm": 0.4209285378456116,
      "learning_rate": 8e-05,
      "loss": 1.6778,
      "step": 3855
    },
    {
      "epoch": 0.41672970928347564,
      "grad_norm": 0.5240936875343323,
      "learning_rate": 8e-05,
      "loss": 1.946,
      "step": 3856
    },
    {
      "epoch": 0.41683778234086244,
      "grad_norm": 0.42781636118888855,
      "learning_rate": 8e-05,
      "loss": 1.5432,
      "step": 3857
    },
    {
      "epoch": 0.41694585539824924,
      "grad_norm": 0.45328959822654724,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 3858
    },
    {
      "epoch": 0.417053928455636,
      "grad_norm": 0.4705790579319,
      "learning_rate": 8e-05,
      "loss": 1.8861,
      "step": 3859
    },
    {
      "epoch": 0.4171620015130228,
      "grad_norm": 0.46849310398101807,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 3860
    },
    {
      "epoch": 0.4172700745704096,
      "grad_norm": 0.45972418785095215,
      "learning_rate": 8e-05,
      "loss": 1.9071,
      "step": 3861
    },
    {
      "epoch": 0.4173781476277964,
      "grad_norm": 0.4606461524963379,
      "learning_rate": 8e-05,
      "loss": 1.9175,
      "step": 3862
    },
    {
      "epoch": 0.4174862206851832,
      "grad_norm": 0.44186118245124817,
      "learning_rate": 8e-05,
      "loss": 1.702,
      "step": 3863
    },
    {
      "epoch": 0.41759429374256996,
      "grad_norm": 0.4235588610172272,
      "learning_rate": 8e-05,
      "loss": 1.7697,
      "step": 3864
    },
    {
      "epoch": 0.41770236679995676,
      "grad_norm": 0.4608592092990875,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 3865
    },
    {
      "epoch": 0.41781043985734356,
      "grad_norm": 0.4199691712856293,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 3866
    },
    {
      "epoch": 0.41791851291473037,
      "grad_norm": 0.4544784724712372,
      "learning_rate": 8e-05,
      "loss": 1.7357,
      "step": 3867
    },
    {
      "epoch": 0.41802658597211717,
      "grad_norm": 0.4341640770435333,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 3868
    },
    {
      "epoch": 0.4181346590295039,
      "grad_norm": 0.44106733798980713,
      "learning_rate": 8e-05,
      "loss": 1.7723,
      "step": 3869
    },
    {
      "epoch": 0.4182427320868907,
      "grad_norm": 0.41517964005470276,
      "learning_rate": 8e-05,
      "loss": 1.6951,
      "step": 3870
    },
    {
      "epoch": 0.4183508051442775,
      "grad_norm": 0.4602586030960083,
      "learning_rate": 8e-05,
      "loss": 1.8451,
      "step": 3871
    },
    {
      "epoch": 0.41845887820166433,
      "grad_norm": 0.4380829334259033,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 3872
    },
    {
      "epoch": 0.41856695125905113,
      "grad_norm": 0.43103355169296265,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 3873
    },
    {
      "epoch": 0.41867502431643794,
      "grad_norm": 0.40566712617874146,
      "learning_rate": 8e-05,
      "loss": 1.5463,
      "step": 3874
    },
    {
      "epoch": 0.4187830973738247,
      "grad_norm": 0.4314163029193878,
      "learning_rate": 8e-05,
      "loss": 1.709,
      "step": 3875
    },
    {
      "epoch": 0.4188911704312115,
      "grad_norm": 0.45007848739624023,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 3876
    },
    {
      "epoch": 0.4189992434885983,
      "grad_norm": 0.4296343922615051,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 3877
    },
    {
      "epoch": 0.4191073165459851,
      "grad_norm": 0.4539518356323242,
      "learning_rate": 8e-05,
      "loss": 1.4683,
      "step": 3878
    },
    {
      "epoch": 0.4192153896033719,
      "grad_norm": 0.4159621000289917,
      "learning_rate": 8e-05,
      "loss": 1.8385,
      "step": 3879
    },
    {
      "epoch": 0.41932346266075865,
      "grad_norm": 0.4153398275375366,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 3880
    },
    {
      "epoch": 0.41943153571814545,
      "grad_norm": 0.44329744577407837,
      "learning_rate": 8e-05,
      "loss": 1.7719,
      "step": 3881
    },
    {
      "epoch": 0.41953960877553226,
      "grad_norm": 0.4715214967727661,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 3882
    },
    {
      "epoch": 0.41964768183291906,
      "grad_norm": 0.4703550338745117,
      "learning_rate": 8e-05,
      "loss": 1.5975,
      "step": 3883
    },
    {
      "epoch": 0.41975575489030587,
      "grad_norm": 0.4259623885154724,
      "learning_rate": 8e-05,
      "loss": 1.7976,
      "step": 3884
    },
    {
      "epoch": 0.4198638279476926,
      "grad_norm": 0.4545591473579407,
      "learning_rate": 8e-05,
      "loss": 1.8323,
      "step": 3885
    },
    {
      "epoch": 0.4199719010050794,
      "grad_norm": 0.4051617383956909,
      "learning_rate": 8e-05,
      "loss": 1.4994,
      "step": 3886
    },
    {
      "epoch": 0.4200799740624662,
      "grad_norm": 0.40334779024124146,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 3887
    },
    {
      "epoch": 0.420188047119853,
      "grad_norm": 0.5452941060066223,
      "learning_rate": 8e-05,
      "loss": 1.9778,
      "step": 3888
    },
    {
      "epoch": 0.42029612017723983,
      "grad_norm": 0.4229658842086792,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 3889
    },
    {
      "epoch": 0.42040419323462663,
      "grad_norm": 0.4005073308944702,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 3890
    },
    {
      "epoch": 0.4205122662920134,
      "grad_norm": 0.435597687959671,
      "learning_rate": 8e-05,
      "loss": 1.7433,
      "step": 3891
    },
    {
      "epoch": 0.4206203393494002,
      "grad_norm": 0.4457031786441803,
      "learning_rate": 8e-05,
      "loss": 1.8362,
      "step": 3892
    },
    {
      "epoch": 0.420728412406787,
      "grad_norm": 0.4417300820350647,
      "learning_rate": 8e-05,
      "loss": 1.9025,
      "step": 3893
    },
    {
      "epoch": 0.4208364854641738,
      "grad_norm": 0.40059104561805725,
      "learning_rate": 8e-05,
      "loss": 1.6118,
      "step": 3894
    },
    {
      "epoch": 0.4209445585215606,
      "grad_norm": 0.421177476644516,
      "learning_rate": 8e-05,
      "loss": 1.7087,
      "step": 3895
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.41959044337272644,
      "learning_rate": 8e-05,
      "loss": 1.5762,
      "step": 3896
    },
    {
      "epoch": 0.42116070463633415,
      "grad_norm": 0.42461884021759033,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 3897
    },
    {
      "epoch": 0.42126877769372095,
      "grad_norm": 0.47070473432540894,
      "learning_rate": 8e-05,
      "loss": 1.7635,
      "step": 3898
    },
    {
      "epoch": 0.42137685075110776,
      "grad_norm": 0.5346662998199463,
      "learning_rate": 8e-05,
      "loss": 1.8489,
      "step": 3899
    },
    {
      "epoch": 0.42148492380849456,
      "grad_norm": 0.44554057717323303,
      "learning_rate": 8e-05,
      "loss": 1.759,
      "step": 3900
    },
    {
      "epoch": 0.4215929968658813,
      "grad_norm": 0.4116251468658447,
      "learning_rate": 8e-05,
      "loss": 1.5537,
      "step": 3901
    },
    {
      "epoch": 0.4217010699232681,
      "grad_norm": 0.4758415222167969,
      "learning_rate": 8e-05,
      "loss": 1.7982,
      "step": 3902
    },
    {
      "epoch": 0.4218091429806549,
      "grad_norm": 0.43455684185028076,
      "learning_rate": 8e-05,
      "loss": 1.6458,
      "step": 3903
    },
    {
      "epoch": 0.4219172160380417,
      "grad_norm": 0.43807342648506165,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 3904
    },
    {
      "epoch": 0.4220252890954285,
      "grad_norm": 0.4677625596523285,
      "learning_rate": 8e-05,
      "loss": 1.9072,
      "step": 3905
    },
    {
      "epoch": 0.42213336215281533,
      "grad_norm": 0.4622395634651184,
      "learning_rate": 8e-05,
      "loss": 1.7458,
      "step": 3906
    },
    {
      "epoch": 0.4222414352102021,
      "grad_norm": 0.42269957065582275,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 3907
    },
    {
      "epoch": 0.4223495082675889,
      "grad_norm": 0.4763660132884979,
      "learning_rate": 8e-05,
      "loss": 1.9308,
      "step": 3908
    },
    {
      "epoch": 0.4224575813249757,
      "grad_norm": 0.4237555265426636,
      "learning_rate": 8e-05,
      "loss": 1.4913,
      "step": 3909
    },
    {
      "epoch": 0.4225656543823625,
      "grad_norm": 0.4171382188796997,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 3910
    },
    {
      "epoch": 0.4226737274397493,
      "grad_norm": 0.4028118848800659,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 3911
    },
    {
      "epoch": 0.42278180049713604,
      "grad_norm": 0.43647482991218567,
      "learning_rate": 8e-05,
      "loss": 1.8401,
      "step": 3912
    },
    {
      "epoch": 0.42288987355452284,
      "grad_norm": 0.4465855062007904,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 3913
    },
    {
      "epoch": 0.42299794661190965,
      "grad_norm": 0.47485458850860596,
      "learning_rate": 8e-05,
      "loss": 1.8615,
      "step": 3914
    },
    {
      "epoch": 0.42310601966929645,
      "grad_norm": 0.4156360626220703,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 3915
    },
    {
      "epoch": 0.42321409272668326,
      "grad_norm": 0.4197590947151184,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 3916
    },
    {
      "epoch": 0.42332216578407,
      "grad_norm": 0.4879176914691925,
      "learning_rate": 8e-05,
      "loss": 1.9281,
      "step": 3917
    },
    {
      "epoch": 0.4234302388414568,
      "grad_norm": 0.4127993881702423,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 3918
    },
    {
      "epoch": 0.4235383118988436,
      "grad_norm": 0.42108452320098877,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 3919
    },
    {
      "epoch": 0.4236463849562304,
      "grad_norm": 0.48254895210266113,
      "learning_rate": 8e-05,
      "loss": 1.8387,
      "step": 3920
    },
    {
      "epoch": 0.4237544580136172,
      "grad_norm": 0.4550391435623169,
      "learning_rate": 8e-05,
      "loss": 1.6528,
      "step": 3921
    },
    {
      "epoch": 0.423862531071004,
      "grad_norm": 0.4663132131099701,
      "learning_rate": 8e-05,
      "loss": 1.7958,
      "step": 3922
    },
    {
      "epoch": 0.42397060412839077,
      "grad_norm": 0.4111482501029968,
      "learning_rate": 8e-05,
      "loss": 1.6418,
      "step": 3923
    },
    {
      "epoch": 0.4240786771857776,
      "grad_norm": 0.4419543743133545,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 3924
    },
    {
      "epoch": 0.4241867502431644,
      "grad_norm": 0.4386090040206909,
      "learning_rate": 8e-05,
      "loss": 1.8581,
      "step": 3925
    },
    {
      "epoch": 0.4242948233005512,
      "grad_norm": 0.4159349501132965,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 3926
    },
    {
      "epoch": 0.424402896357938,
      "grad_norm": 0.44141486287117004,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 3927
    },
    {
      "epoch": 0.42451096941532473,
      "grad_norm": 0.4438622295856476,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 3928
    },
    {
      "epoch": 0.42461904247271154,
      "grad_norm": 0.42246687412261963,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 3929
    },
    {
      "epoch": 0.42472711553009834,
      "grad_norm": 0.4424813985824585,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 3930
    },
    {
      "epoch": 0.42483518858748515,
      "grad_norm": 0.4495694637298584,
      "learning_rate": 8e-05,
      "loss": 1.863,
      "step": 3931
    },
    {
      "epoch": 0.42494326164487195,
      "grad_norm": 0.46549341082572937,
      "learning_rate": 8e-05,
      "loss": 1.8488,
      "step": 3932
    },
    {
      "epoch": 0.42505133470225875,
      "grad_norm": 0.42901504039764404,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 3933
    },
    {
      "epoch": 0.4251594077596455,
      "grad_norm": 0.4948921501636505,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 3934
    },
    {
      "epoch": 0.4252674808170323,
      "grad_norm": 0.4233351945877075,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 3935
    },
    {
      "epoch": 0.4253755538744191,
      "grad_norm": 0.4696337580680847,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 3936
    },
    {
      "epoch": 0.4254836269318059,
      "grad_norm": 0.44807466864585876,
      "learning_rate": 8e-05,
      "loss": 1.6882,
      "step": 3937
    },
    {
      "epoch": 0.4255916999891927,
      "grad_norm": 0.4644996225833893,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 3938
    },
    {
      "epoch": 0.42569977304657947,
      "grad_norm": 0.48724403977394104,
      "learning_rate": 8e-05,
      "loss": 1.5863,
      "step": 3939
    },
    {
      "epoch": 0.42580784610396627,
      "grad_norm": 0.4845251739025116,
      "learning_rate": 8e-05,
      "loss": 1.8595,
      "step": 3940
    },
    {
      "epoch": 0.4259159191613531,
      "grad_norm": 0.4272240698337555,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 3941
    },
    {
      "epoch": 0.4260239922187399,
      "grad_norm": 0.39636537432670593,
      "learning_rate": 8e-05,
      "loss": 1.5794,
      "step": 3942
    },
    {
      "epoch": 0.4261320652761267,
      "grad_norm": 0.4332127571105957,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 3943
    },
    {
      "epoch": 0.42624013833351343,
      "grad_norm": 0.4377022683620453,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 3944
    },
    {
      "epoch": 0.42634821139090023,
      "grad_norm": 0.4560088515281677,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 3945
    },
    {
      "epoch": 0.42645628444828704,
      "grad_norm": 0.4121147394180298,
      "learning_rate": 8e-05,
      "loss": 1.6286,
      "step": 3946
    },
    {
      "epoch": 0.42656435750567384,
      "grad_norm": 0.5279379487037659,
      "learning_rate": 8e-05,
      "loss": 2.0069,
      "step": 3947
    },
    {
      "epoch": 0.42667243056306065,
      "grad_norm": 0.5125395059585571,
      "learning_rate": 8e-05,
      "loss": 1.9261,
      "step": 3948
    },
    {
      "epoch": 0.42678050362044745,
      "grad_norm": 0.44626861810684204,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 3949
    },
    {
      "epoch": 0.4268885766778342,
      "grad_norm": 0.41344085335731506,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 3950
    },
    {
      "epoch": 0.426996649735221,
      "grad_norm": 0.4169583022594452,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 3951
    },
    {
      "epoch": 0.4271047227926078,
      "grad_norm": 0.4648887813091278,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 3952
    },
    {
      "epoch": 0.4272127958499946,
      "grad_norm": 0.4631357192993164,
      "learning_rate": 8e-05,
      "loss": 1.9267,
      "step": 3953
    },
    {
      "epoch": 0.4273208689073814,
      "grad_norm": 0.4239581227302551,
      "learning_rate": 8e-05,
      "loss": 1.5714,
      "step": 3954
    },
    {
      "epoch": 0.42742894196476816,
      "grad_norm": 0.44858697056770325,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 3955
    },
    {
      "epoch": 0.42753701502215496,
      "grad_norm": 0.4333905875682831,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 3956
    },
    {
      "epoch": 0.42764508807954177,
      "grad_norm": 0.4059256911277771,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 3957
    },
    {
      "epoch": 0.4277531611369286,
      "grad_norm": 0.4172080457210541,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 3958
    },
    {
      "epoch": 0.4278612341943154,
      "grad_norm": 0.49334973096847534,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 3959
    },
    {
      "epoch": 0.4279693072517021,
      "grad_norm": 0.5323154926300049,
      "learning_rate": 8e-05,
      "loss": 1.9391,
      "step": 3960
    },
    {
      "epoch": 0.42807738030908893,
      "grad_norm": 0.5308656096458435,
      "learning_rate": 8e-05,
      "loss": 1.9644,
      "step": 3961
    },
    {
      "epoch": 0.42818545336647573,
      "grad_norm": 0.42655327916145325,
      "learning_rate": 8e-05,
      "loss": 1.8783,
      "step": 3962
    },
    {
      "epoch": 0.42829352642386254,
      "grad_norm": 0.4288577139377594,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 3963
    },
    {
      "epoch": 0.42840159948124934,
      "grad_norm": 0.43760186433792114,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 3964
    },
    {
      "epoch": 0.42850967253863614,
      "grad_norm": 0.47798895835876465,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 3965
    },
    {
      "epoch": 0.4286177455960229,
      "grad_norm": 0.44441962242126465,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 3966
    },
    {
      "epoch": 0.4287258186534097,
      "grad_norm": 0.44476059079170227,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 3967
    },
    {
      "epoch": 0.4288338917107965,
      "grad_norm": 0.4548526704311371,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 3968
    },
    {
      "epoch": 0.4289419647681833,
      "grad_norm": 0.45173656940460205,
      "learning_rate": 8e-05,
      "loss": 1.8132,
      "step": 3969
    },
    {
      "epoch": 0.4290500378255701,
      "grad_norm": 0.4032886326313019,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 3970
    },
    {
      "epoch": 0.42915811088295686,
      "grad_norm": 0.48568806052207947,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 3971
    },
    {
      "epoch": 0.42926618394034366,
      "grad_norm": 0.4596042335033417,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 3972
    },
    {
      "epoch": 0.42937425699773046,
      "grad_norm": 0.43071603775024414,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 3973
    },
    {
      "epoch": 0.42948233005511727,
      "grad_norm": 0.4300806522369385,
      "learning_rate": 8e-05,
      "loss": 1.7862,
      "step": 3974
    },
    {
      "epoch": 0.42959040311250407,
      "grad_norm": 0.4340526759624481,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 3975
    },
    {
      "epoch": 0.4296984761698908,
      "grad_norm": 0.5000108480453491,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 3976
    },
    {
      "epoch": 0.4298065492272776,
      "grad_norm": 0.43857866525650024,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 3977
    },
    {
      "epoch": 0.4299146222846644,
      "grad_norm": 0.5507070422172546,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 3978
    },
    {
      "epoch": 0.43002269534205123,
      "grad_norm": 0.4173576235771179,
      "learning_rate": 8e-05,
      "loss": 1.7054,
      "step": 3979
    },
    {
      "epoch": 0.43013076839943803,
      "grad_norm": 0.4519742727279663,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 3980
    },
    {
      "epoch": 0.43023884145682484,
      "grad_norm": 0.4171552062034607,
      "learning_rate": 8e-05,
      "loss": 1.5649,
      "step": 3981
    },
    {
      "epoch": 0.4303469145142116,
      "grad_norm": 0.4072751998901367,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 3982
    },
    {
      "epoch": 0.4304549875715984,
      "grad_norm": 0.4172516167163849,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 3983
    },
    {
      "epoch": 0.4305630606289852,
      "grad_norm": 0.5008341073989868,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 3984
    },
    {
      "epoch": 0.430671133686372,
      "grad_norm": 0.42623138427734375,
      "learning_rate": 8e-05,
      "loss": 1.5248,
      "step": 3985
    },
    {
      "epoch": 0.4307792067437588,
      "grad_norm": 0.4561443030834198,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 3986
    },
    {
      "epoch": 0.43088727980114555,
      "grad_norm": 0.4338952898979187,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 3987
    },
    {
      "epoch": 0.43099535285853235,
      "grad_norm": 0.4297468066215515,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 3988
    },
    {
      "epoch": 0.43110342591591916,
      "grad_norm": 0.47269296646118164,
      "learning_rate": 8e-05,
      "loss": 1.9283,
      "step": 3989
    },
    {
      "epoch": 0.43121149897330596,
      "grad_norm": 0.44260260462760925,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 3990
    },
    {
      "epoch": 0.43131957203069277,
      "grad_norm": 0.4273146092891693,
      "learning_rate": 8e-05,
      "loss": 1.6034,
      "step": 3991
    },
    {
      "epoch": 0.4314276450880795,
      "grad_norm": 0.4350992441177368,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 3992
    },
    {
      "epoch": 0.4315357181454663,
      "grad_norm": 0.5168647766113281,
      "learning_rate": 8e-05,
      "loss": 2.0562,
      "step": 3993
    },
    {
      "epoch": 0.4316437912028531,
      "grad_norm": 0.46659040451049805,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 3994
    },
    {
      "epoch": 0.4317518642602399,
      "grad_norm": 0.43766263127326965,
      "learning_rate": 8e-05,
      "loss": 1.7704,
      "step": 3995
    },
    {
      "epoch": 0.43185993731762673,
      "grad_norm": 0.4720185697078705,
      "learning_rate": 8e-05,
      "loss": 1.7435,
      "step": 3996
    },
    {
      "epoch": 0.43196801037501353,
      "grad_norm": 0.43264809250831604,
      "learning_rate": 8e-05,
      "loss": 1.702,
      "step": 3997
    },
    {
      "epoch": 0.4320760834324003,
      "grad_norm": 0.42666003108024597,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 3998
    },
    {
      "epoch": 0.4321841564897871,
      "grad_norm": 0.44149795174598694,
      "learning_rate": 8e-05,
      "loss": 1.8009,
      "step": 3999
    },
    {
      "epoch": 0.4322922295471739,
      "grad_norm": 0.4723605513572693,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 4000
    },
    {
      "epoch": 0.4324003026045607,
      "grad_norm": 0.43791231513023376,
      "learning_rate": 8e-05,
      "loss": 1.6285,
      "step": 4001
    },
    {
      "epoch": 0.4325083756619475,
      "grad_norm": 0.43205708265304565,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 4002
    },
    {
      "epoch": 0.43261644871933425,
      "grad_norm": 0.42249298095703125,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 4003
    },
    {
      "epoch": 0.43272452177672105,
      "grad_norm": 0.4272950291633606,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 4004
    },
    {
      "epoch": 0.43283259483410785,
      "grad_norm": 0.45676127076148987,
      "learning_rate": 8e-05,
      "loss": 1.9477,
      "step": 4005
    },
    {
      "epoch": 0.43294066789149466,
      "grad_norm": 0.41241440176963806,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 4006
    },
    {
      "epoch": 0.43304874094888146,
      "grad_norm": 0.43888169527053833,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 4007
    },
    {
      "epoch": 0.43315681400626826,
      "grad_norm": 0.4230831265449524,
      "learning_rate": 8e-05,
      "loss": 1.409,
      "step": 4008
    },
    {
      "epoch": 0.433264887063655,
      "grad_norm": 0.4168720841407776,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 4009
    },
    {
      "epoch": 0.4333729601210418,
      "grad_norm": 0.43469861149787903,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 4010
    },
    {
      "epoch": 0.4334810331784286,
      "grad_norm": 0.48600974678993225,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 4011
    },
    {
      "epoch": 0.4335891062358154,
      "grad_norm": 0.424825519323349,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 4012
    },
    {
      "epoch": 0.43369717929320223,
      "grad_norm": 0.43900173902511597,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 4013
    },
    {
      "epoch": 0.433805252350589,
      "grad_norm": 0.4591785669326782,
      "learning_rate": 8e-05,
      "loss": 1.7528,
      "step": 4014
    },
    {
      "epoch": 0.4339133254079758,
      "grad_norm": 0.47181758284568787,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 4015
    },
    {
      "epoch": 0.4340213984653626,
      "grad_norm": 0.51241135597229,
      "learning_rate": 8e-05,
      "loss": 1.9418,
      "step": 4016
    },
    {
      "epoch": 0.4341294715227494,
      "grad_norm": 0.46633362770080566,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 4017
    },
    {
      "epoch": 0.4342375445801362,
      "grad_norm": 0.44052186608314514,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 4018
    },
    {
      "epoch": 0.43434561763752294,
      "grad_norm": 0.41213348507881165,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 4019
    },
    {
      "epoch": 0.43445369069490974,
      "grad_norm": 0.4440227150917053,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 4020
    },
    {
      "epoch": 0.43456176375229655,
      "grad_norm": 0.5059786438941956,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 4021
    },
    {
      "epoch": 0.43466983680968335,
      "grad_norm": 0.44048696756362915,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 4022
    },
    {
      "epoch": 0.43477790986707016,
      "grad_norm": 0.4356512725353241,
      "learning_rate": 8e-05,
      "loss": 1.6008,
      "step": 4023
    },
    {
      "epoch": 0.43488598292445696,
      "grad_norm": 0.4280588626861572,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 4024
    },
    {
      "epoch": 0.4349940559818437,
      "grad_norm": 0.5010555386543274,
      "learning_rate": 8e-05,
      "loss": 1.8927,
      "step": 4025
    },
    {
      "epoch": 0.4351021290392305,
      "grad_norm": 0.4446280896663666,
      "learning_rate": 8e-05,
      "loss": 1.8525,
      "step": 4026
    },
    {
      "epoch": 0.4352102020966173,
      "grad_norm": 0.4276418089866638,
      "learning_rate": 8e-05,
      "loss": 1.6341,
      "step": 4027
    },
    {
      "epoch": 0.4353182751540041,
      "grad_norm": 0.4852413237094879,
      "learning_rate": 8e-05,
      "loss": 1.8269,
      "step": 4028
    },
    {
      "epoch": 0.4354263482113909,
      "grad_norm": 0.4630221724510193,
      "learning_rate": 8e-05,
      "loss": 1.846,
      "step": 4029
    },
    {
      "epoch": 0.43553442126877767,
      "grad_norm": 0.42097532749176025,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 4030
    },
    {
      "epoch": 0.4356424943261645,
      "grad_norm": 0.4665873348712921,
      "learning_rate": 8e-05,
      "loss": 1.8624,
      "step": 4031
    },
    {
      "epoch": 0.4357505673835513,
      "grad_norm": 0.3981325328350067,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 4032
    },
    {
      "epoch": 0.4358586404409381,
      "grad_norm": 0.4287027418613434,
      "learning_rate": 8e-05,
      "loss": 1.813,
      "step": 4033
    },
    {
      "epoch": 0.4359667134983249,
      "grad_norm": 0.43227389454841614,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 4034
    },
    {
      "epoch": 0.43607478655571164,
      "grad_norm": 0.4619078040122986,
      "learning_rate": 8e-05,
      "loss": 1.9209,
      "step": 4035
    },
    {
      "epoch": 0.43618285961309844,
      "grad_norm": 0.41055402159690857,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 4036
    },
    {
      "epoch": 0.43629093267048524,
      "grad_norm": 0.45075321197509766,
      "learning_rate": 8e-05,
      "loss": 1.7635,
      "step": 4037
    },
    {
      "epoch": 0.43639900572787205,
      "grad_norm": 0.4336980879306793,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 4038
    },
    {
      "epoch": 0.43650707878525885,
      "grad_norm": 0.3988364636898041,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 4039
    },
    {
      "epoch": 0.43661515184264565,
      "grad_norm": 0.4167577922344208,
      "learning_rate": 8e-05,
      "loss": 1.6461,
      "step": 4040
    },
    {
      "epoch": 0.4367232249000324,
      "grad_norm": 0.4930783212184906,
      "learning_rate": 8e-05,
      "loss": 1.7706,
      "step": 4041
    },
    {
      "epoch": 0.4368312979574192,
      "grad_norm": 0.4373396933078766,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 4042
    },
    {
      "epoch": 0.436939371014806,
      "grad_norm": 0.42410796880722046,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 4043
    },
    {
      "epoch": 0.4370474440721928,
      "grad_norm": 0.46465402841567993,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 4044
    },
    {
      "epoch": 0.4371555171295796,
      "grad_norm": 0.43713822960853577,
      "learning_rate": 8e-05,
      "loss": 1.7125,
      "step": 4045
    },
    {
      "epoch": 0.43726359018696637,
      "grad_norm": 0.43840843439102173,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 4046
    },
    {
      "epoch": 0.43737166324435317,
      "grad_norm": 0.4863823652267456,
      "learning_rate": 8e-05,
      "loss": 1.95,
      "step": 4047
    },
    {
      "epoch": 0.43747973630174,
      "grad_norm": 0.4560789167881012,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 4048
    },
    {
      "epoch": 0.4375878093591268,
      "grad_norm": 0.412541002035141,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 4049
    },
    {
      "epoch": 0.4376958824165136,
      "grad_norm": 0.42028868198394775,
      "learning_rate": 8e-05,
      "loss": 1.6211,
      "step": 4050
    },
    {
      "epoch": 0.43780395547390033,
      "grad_norm": 0.41196298599243164,
      "learning_rate": 8e-05,
      "loss": 1.6549,
      "step": 4051
    },
    {
      "epoch": 0.43791202853128713,
      "grad_norm": 0.4353453814983368,
      "learning_rate": 8e-05,
      "loss": 1.6985,
      "step": 4052
    },
    {
      "epoch": 0.43802010158867394,
      "grad_norm": 0.4784531891345978,
      "learning_rate": 8e-05,
      "loss": 1.89,
      "step": 4053
    },
    {
      "epoch": 0.43812817464606074,
      "grad_norm": 0.5705502033233643,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 4054
    },
    {
      "epoch": 0.43823624770344755,
      "grad_norm": 0.4793296158313751,
      "learning_rate": 8e-05,
      "loss": 1.933,
      "step": 4055
    },
    {
      "epoch": 0.43834432076083435,
      "grad_norm": 0.4391688406467438,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 4056
    },
    {
      "epoch": 0.4384523938182211,
      "grad_norm": 0.4441629648208618,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 4057
    },
    {
      "epoch": 0.4385604668756079,
      "grad_norm": 0.4286699891090393,
      "learning_rate": 8e-05,
      "loss": 1.4511,
      "step": 4058
    },
    {
      "epoch": 0.4386685399329947,
      "grad_norm": 0.4293336868286133,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 4059
    },
    {
      "epoch": 0.4387766129903815,
      "grad_norm": 0.48418128490448,
      "learning_rate": 8e-05,
      "loss": 1.9767,
      "step": 4060
    },
    {
      "epoch": 0.4388846860477683,
      "grad_norm": 0.424579381942749,
      "learning_rate": 8e-05,
      "loss": 1.6505,
      "step": 4061
    },
    {
      "epoch": 0.43899275910515506,
      "grad_norm": 0.45915284752845764,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 4062
    },
    {
      "epoch": 0.43910083216254187,
      "grad_norm": 0.4583372175693512,
      "learning_rate": 8e-05,
      "loss": 1.8522,
      "step": 4063
    },
    {
      "epoch": 0.43920890521992867,
      "grad_norm": 0.4252242147922516,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 4064
    },
    {
      "epoch": 0.4393169782773155,
      "grad_norm": 0.4561458230018616,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 4065
    },
    {
      "epoch": 0.4394250513347023,
      "grad_norm": 0.4322243928909302,
      "learning_rate": 8e-05,
      "loss": 1.7852,
      "step": 4066
    },
    {
      "epoch": 0.439533124392089,
      "grad_norm": 0.4290982484817505,
      "learning_rate": 8e-05,
      "loss": 1.8609,
      "step": 4067
    },
    {
      "epoch": 0.43964119744947583,
      "grad_norm": 0.43716999888420105,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 4068
    },
    {
      "epoch": 0.43974927050686263,
      "grad_norm": 0.5319225192070007,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 4069
    },
    {
      "epoch": 0.43985734356424944,
      "grad_norm": 0.4733983278274536,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 4070
    },
    {
      "epoch": 0.43996541662163624,
      "grad_norm": 0.4398679733276367,
      "learning_rate": 8e-05,
      "loss": 1.7504,
      "step": 4071
    },
    {
      "epoch": 0.44007348967902304,
      "grad_norm": 0.4421903192996979,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 4072
    },
    {
      "epoch": 0.4401815627364098,
      "grad_norm": 0.505876898765564,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 4073
    },
    {
      "epoch": 0.4402896357937966,
      "grad_norm": 0.45107635855674744,
      "learning_rate": 8e-05,
      "loss": 1.8334,
      "step": 4074
    },
    {
      "epoch": 0.4403977088511834,
      "grad_norm": 0.4307292401790619,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 4075
    },
    {
      "epoch": 0.4405057819085702,
      "grad_norm": 0.4368915855884552,
      "learning_rate": 8e-05,
      "loss": 1.7924,
      "step": 4076
    },
    {
      "epoch": 0.440613854965957,
      "grad_norm": 0.4004994034767151,
      "learning_rate": 8e-05,
      "loss": 1.703,
      "step": 4077
    },
    {
      "epoch": 0.44072192802334376,
      "grad_norm": 0.44151172041893005,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 4078
    },
    {
      "epoch": 0.44083000108073056,
      "grad_norm": 0.3944567143917084,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 4079
    },
    {
      "epoch": 0.44093807413811736,
      "grad_norm": 0.39417627453804016,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 4080
    },
    {
      "epoch": 0.44104614719550417,
      "grad_norm": 0.43266168236732483,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 4081
    },
    {
      "epoch": 0.44115422025289097,
      "grad_norm": 0.415542334318161,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 4082
    },
    {
      "epoch": 0.4412622933102777,
      "grad_norm": 0.46871110796928406,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 4083
    },
    {
      "epoch": 0.4413703663676645,
      "grad_norm": 0.4024544954299927,
      "learning_rate": 8e-05,
      "loss": 1.5368,
      "step": 4084
    },
    {
      "epoch": 0.4414784394250513,
      "grad_norm": 0.44527488946914673,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 4085
    },
    {
      "epoch": 0.44158651248243813,
      "grad_norm": 0.4082452058792114,
      "learning_rate": 8e-05,
      "loss": 1.5787,
      "step": 4086
    },
    {
      "epoch": 0.44169458553982494,
      "grad_norm": 0.46607914566993713,
      "learning_rate": 8e-05,
      "loss": 1.8282,
      "step": 4087
    },
    {
      "epoch": 0.44180265859721174,
      "grad_norm": 0.46515774726867676,
      "learning_rate": 8e-05,
      "loss": 1.9321,
      "step": 4088
    },
    {
      "epoch": 0.4419107316545985,
      "grad_norm": 0.4445117712020874,
      "learning_rate": 8e-05,
      "loss": 1.7307,
      "step": 4089
    },
    {
      "epoch": 0.4420188047119853,
      "grad_norm": 0.41419950127601624,
      "learning_rate": 8e-05,
      "loss": 1.6216,
      "step": 4090
    },
    {
      "epoch": 0.4421268777693721,
      "grad_norm": 0.42533260583877563,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 4091
    },
    {
      "epoch": 0.4422349508267589,
      "grad_norm": 0.4718379080295563,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 4092
    },
    {
      "epoch": 0.4423430238841457,
      "grad_norm": 0.44216299057006836,
      "learning_rate": 8e-05,
      "loss": 1.6407,
      "step": 4093
    },
    {
      "epoch": 0.44245109694153245,
      "grad_norm": 0.4264194667339325,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 4094
    },
    {
      "epoch": 0.44255916999891926,
      "grad_norm": 0.4055792987346649,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 4095
    },
    {
      "epoch": 0.44266724305630606,
      "grad_norm": 0.44283828139305115,
      "learning_rate": 8e-05,
      "loss": 1.7664,
      "step": 4096
    },
    {
      "epoch": 0.44277531611369286,
      "grad_norm": 0.4330628514289856,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 4097
    },
    {
      "epoch": 0.44288338917107967,
      "grad_norm": 0.42181286215782166,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 4098
    },
    {
      "epoch": 0.44299146222846647,
      "grad_norm": 0.42756059765815735,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 4099
    },
    {
      "epoch": 0.4430995352858532,
      "grad_norm": 0.48497799038887024,
      "learning_rate": 8e-05,
      "loss": 1.9628,
      "step": 4100
    },
    {
      "epoch": 0.44320760834324,
      "grad_norm": 0.43913137912750244,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 4101
    },
    {
      "epoch": 0.4433156814006268,
      "grad_norm": 0.42771115899086,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 4102
    },
    {
      "epoch": 0.44342375445801363,
      "grad_norm": 0.46608003973960876,
      "learning_rate": 8e-05,
      "loss": 1.8412,
      "step": 4103
    },
    {
      "epoch": 0.44353182751540043,
      "grad_norm": 0.5131593346595764,
      "learning_rate": 8e-05,
      "loss": 1.78,
      "step": 4104
    },
    {
      "epoch": 0.4436399005727872,
      "grad_norm": 0.45873725414276123,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 4105
    },
    {
      "epoch": 0.443747973630174,
      "grad_norm": 0.4305790364742279,
      "learning_rate": 8e-05,
      "loss": 1.8203,
      "step": 4106
    },
    {
      "epoch": 0.4438560466875608,
      "grad_norm": 0.38517969846725464,
      "learning_rate": 8e-05,
      "loss": 1.5196,
      "step": 4107
    },
    {
      "epoch": 0.4439641197449476,
      "grad_norm": 0.40546494722366333,
      "learning_rate": 8e-05,
      "loss": 1.5384,
      "step": 4108
    },
    {
      "epoch": 0.4440721928023344,
      "grad_norm": 0.42575764656066895,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 4109
    },
    {
      "epoch": 0.44418026585972115,
      "grad_norm": 0.43408989906311035,
      "learning_rate": 8e-05,
      "loss": 1.8245,
      "step": 4110
    },
    {
      "epoch": 0.44428833891710795,
      "grad_norm": 0.41283321380615234,
      "learning_rate": 8e-05,
      "loss": 1.5025,
      "step": 4111
    },
    {
      "epoch": 0.44439641197449475,
      "grad_norm": 0.49838370084762573,
      "learning_rate": 8e-05,
      "loss": 1.7832,
      "step": 4112
    },
    {
      "epoch": 0.44450448503188156,
      "grad_norm": 0.49152541160583496,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 4113
    },
    {
      "epoch": 0.44461255808926836,
      "grad_norm": 0.4938238263130188,
      "learning_rate": 8e-05,
      "loss": 2.0161,
      "step": 4114
    },
    {
      "epoch": 0.44472063114665517,
      "grad_norm": 0.46763646602630615,
      "learning_rate": 8e-05,
      "loss": 1.8689,
      "step": 4115
    },
    {
      "epoch": 0.4448287042040419,
      "grad_norm": 0.46870937943458557,
      "learning_rate": 8e-05,
      "loss": 1.8618,
      "step": 4116
    },
    {
      "epoch": 0.4449367772614287,
      "grad_norm": 0.42840611934661865,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 4117
    },
    {
      "epoch": 0.4450448503188155,
      "grad_norm": 0.42922306060791016,
      "learning_rate": 8e-05,
      "loss": 1.8023,
      "step": 4118
    },
    {
      "epoch": 0.4451529233762023,
      "grad_norm": 0.44846782088279724,
      "learning_rate": 8e-05,
      "loss": 1.8447,
      "step": 4119
    },
    {
      "epoch": 0.44526099643358913,
      "grad_norm": 0.42636898159980774,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 4120
    },
    {
      "epoch": 0.4453690694909759,
      "grad_norm": 0.44884470105171204,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 4121
    },
    {
      "epoch": 0.4454771425483627,
      "grad_norm": 0.4449501931667328,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 4122
    },
    {
      "epoch": 0.4455852156057495,
      "grad_norm": 0.4221552312374115,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 4123
    },
    {
      "epoch": 0.4456932886631363,
      "grad_norm": 0.4202118515968323,
      "learning_rate": 8e-05,
      "loss": 1.6403,
      "step": 4124
    },
    {
      "epoch": 0.4458013617205231,
      "grad_norm": 0.45338883996009827,
      "learning_rate": 8e-05,
      "loss": 1.8511,
      "step": 4125
    },
    {
      "epoch": 0.44590943477790984,
      "grad_norm": 0.4452354609966278,
      "learning_rate": 8e-05,
      "loss": 1.7379,
      "step": 4126
    },
    {
      "epoch": 0.44601750783529664,
      "grad_norm": 0.4416992962360382,
      "learning_rate": 8e-05,
      "loss": 1.8041,
      "step": 4127
    },
    {
      "epoch": 0.44612558089268345,
      "grad_norm": 0.45178642868995667,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 4128
    },
    {
      "epoch": 0.44623365395007025,
      "grad_norm": 0.4444006085395813,
      "learning_rate": 8e-05,
      "loss": 1.7641,
      "step": 4129
    },
    {
      "epoch": 0.44634172700745706,
      "grad_norm": 0.4688357710838318,
      "learning_rate": 8e-05,
      "loss": 1.8039,
      "step": 4130
    },
    {
      "epoch": 0.44644980006484386,
      "grad_norm": 0.40843430161476135,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 4131
    },
    {
      "epoch": 0.4465578731222306,
      "grad_norm": 0.4462106227874756,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 4132
    },
    {
      "epoch": 0.4466659461796174,
      "grad_norm": 0.45174431800842285,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 4133
    },
    {
      "epoch": 0.4467740192370042,
      "grad_norm": 0.4264303743839264,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 4134
    },
    {
      "epoch": 0.446882092294391,
      "grad_norm": 0.4621368944644928,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 4135
    },
    {
      "epoch": 0.4469901653517778,
      "grad_norm": 0.41699913144111633,
      "learning_rate": 8e-05,
      "loss": 1.602,
      "step": 4136
    },
    {
      "epoch": 0.4470982384091646,
      "grad_norm": 0.42032426595687866,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 4137
    },
    {
      "epoch": 0.4472063114665514,
      "grad_norm": 0.4329681396484375,
      "learning_rate": 8e-05,
      "loss": 1.7701,
      "step": 4138
    },
    {
      "epoch": 0.4473143845239382,
      "grad_norm": 0.4930082857608795,
      "learning_rate": 8e-05,
      "loss": 1.6717,
      "step": 4139
    },
    {
      "epoch": 0.447422457581325,
      "grad_norm": 0.4174818992614746,
      "learning_rate": 8e-05,
      "loss": 1.7029,
      "step": 4140
    },
    {
      "epoch": 0.4475305306387118,
      "grad_norm": 0.47103065252304077,
      "learning_rate": 8e-05,
      "loss": 1.8053,
      "step": 4141
    },
    {
      "epoch": 0.44763860369609854,
      "grad_norm": 0.4103153944015503,
      "learning_rate": 8e-05,
      "loss": 1.5641,
      "step": 4142
    },
    {
      "epoch": 0.44774667675348534,
      "grad_norm": 0.42930904030799866,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 4143
    },
    {
      "epoch": 0.44785474981087214,
      "grad_norm": 0.4400421977043152,
      "learning_rate": 8e-05,
      "loss": 1.6126,
      "step": 4144
    },
    {
      "epoch": 0.44796282286825895,
      "grad_norm": 0.4135350286960602,
      "learning_rate": 8e-05,
      "loss": 1.6365,
      "step": 4145
    },
    {
      "epoch": 0.44807089592564575,
      "grad_norm": 0.41760432720184326,
      "learning_rate": 8e-05,
      "loss": 1.6238,
      "step": 4146
    },
    {
      "epoch": 0.44817896898303256,
      "grad_norm": 0.4600202143192291,
      "learning_rate": 8e-05,
      "loss": 1.6383,
      "step": 4147
    },
    {
      "epoch": 0.4482870420404193,
      "grad_norm": 0.4699585735797882,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 4148
    },
    {
      "epoch": 0.4483951150978061,
      "grad_norm": 0.5149980187416077,
      "learning_rate": 8e-05,
      "loss": 1.9467,
      "step": 4149
    },
    {
      "epoch": 0.4485031881551929,
      "grad_norm": 0.43702277541160583,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 4150
    },
    {
      "epoch": 0.4486112612125797,
      "grad_norm": 0.41213831305503845,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 4151
    },
    {
      "epoch": 0.4487193342699665,
      "grad_norm": 0.4379148483276367,
      "learning_rate": 8e-05,
      "loss": 1.8179,
      "step": 4152
    },
    {
      "epoch": 0.44882740732735327,
      "grad_norm": 0.42742499709129333,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 4153
    },
    {
      "epoch": 0.44893548038474007,
      "grad_norm": 0.4379568099975586,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 4154
    },
    {
      "epoch": 0.4490435534421269,
      "grad_norm": 0.4359889328479767,
      "learning_rate": 8e-05,
      "loss": 1.7598,
      "step": 4155
    },
    {
      "epoch": 0.4491516264995137,
      "grad_norm": 0.42240601778030396,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 4156
    },
    {
      "epoch": 0.4492596995569005,
      "grad_norm": 0.4301482141017914,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 4157
    },
    {
      "epoch": 0.44936777261428723,
      "grad_norm": 0.42765045166015625,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 4158
    },
    {
      "epoch": 0.44947584567167403,
      "grad_norm": 0.41019725799560547,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 4159
    },
    {
      "epoch": 0.44958391872906084,
      "grad_norm": 0.44172990322113037,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 4160
    },
    {
      "epoch": 0.44969199178644764,
      "grad_norm": 0.41132068634033203,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 4161
    },
    {
      "epoch": 0.44980006484383445,
      "grad_norm": 0.4046846330165863,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 4162
    },
    {
      "epoch": 0.44990813790122125,
      "grad_norm": 0.42835360765457153,
      "learning_rate": 8e-05,
      "loss": 1.4765,
      "step": 4163
    },
    {
      "epoch": 0.450016210958608,
      "grad_norm": 0.449691504240036,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 4164
    },
    {
      "epoch": 0.4501242840159948,
      "grad_norm": 0.43543747067451477,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 4165
    },
    {
      "epoch": 0.4502323570733816,
      "grad_norm": 0.4547455608844757,
      "learning_rate": 8e-05,
      "loss": 1.8777,
      "step": 4166
    },
    {
      "epoch": 0.4503404301307684,
      "grad_norm": 0.4088083803653717,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 4167
    },
    {
      "epoch": 0.4504485031881552,
      "grad_norm": 0.4379052519798279,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 4168
    },
    {
      "epoch": 0.45055657624554196,
      "grad_norm": 0.47236260771751404,
      "learning_rate": 8e-05,
      "loss": 1.8196,
      "step": 4169
    },
    {
      "epoch": 0.45066464930292877,
      "grad_norm": 0.4441351890563965,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 4170
    },
    {
      "epoch": 0.45077272236031557,
      "grad_norm": 0.42088690400123596,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 4171
    },
    {
      "epoch": 0.4508807954177024,
      "grad_norm": 0.454499751329422,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 4172
    },
    {
      "epoch": 0.4509888684750892,
      "grad_norm": 0.44542646408081055,
      "learning_rate": 8e-05,
      "loss": 1.8701,
      "step": 4173
    },
    {
      "epoch": 0.451096941532476,
      "grad_norm": 0.42925041913986206,
      "learning_rate": 8e-05,
      "loss": 1.7707,
      "step": 4174
    },
    {
      "epoch": 0.45120501458986273,
      "grad_norm": 0.44744524359703064,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 4175
    },
    {
      "epoch": 0.45131308764724953,
      "grad_norm": 0.44414469599723816,
      "learning_rate": 8e-05,
      "loss": 1.6554,
      "step": 4176
    },
    {
      "epoch": 0.45142116070463634,
      "grad_norm": 0.4502558410167694,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 4177
    },
    {
      "epoch": 0.45152923376202314,
      "grad_norm": 0.451296329498291,
      "learning_rate": 8e-05,
      "loss": 1.8746,
      "step": 4178
    },
    {
      "epoch": 0.45163730681940994,
      "grad_norm": 0.4535653293132782,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 4179
    },
    {
      "epoch": 0.4517453798767967,
      "grad_norm": 0.4156627357006073,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 4180
    },
    {
      "epoch": 0.4518534529341835,
      "grad_norm": 0.4419694244861603,
      "learning_rate": 8e-05,
      "loss": 1.7317,
      "step": 4181
    },
    {
      "epoch": 0.4519615259915703,
      "grad_norm": 0.48256924748420715,
      "learning_rate": 8e-05,
      "loss": 1.5618,
      "step": 4182
    },
    {
      "epoch": 0.4520695990489571,
      "grad_norm": 0.4032396972179413,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 4183
    },
    {
      "epoch": 0.4521776721063439,
      "grad_norm": 0.46799328923225403,
      "learning_rate": 8e-05,
      "loss": 1.8787,
      "step": 4184
    },
    {
      "epoch": 0.45228574516373066,
      "grad_norm": 0.4412846565246582,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 4185
    },
    {
      "epoch": 0.45239381822111746,
      "grad_norm": 0.4270966053009033,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 4186
    },
    {
      "epoch": 0.45250189127850426,
      "grad_norm": 0.457460880279541,
      "learning_rate": 8e-05,
      "loss": 1.9034,
      "step": 4187
    },
    {
      "epoch": 0.45260996433589107,
      "grad_norm": 0.4366394281387329,
      "learning_rate": 8e-05,
      "loss": 1.6223,
      "step": 4188
    },
    {
      "epoch": 0.4527180373932779,
      "grad_norm": 0.4557481110095978,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 4189
    },
    {
      "epoch": 0.4528261104506647,
      "grad_norm": 0.4028531610965729,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 4190
    },
    {
      "epoch": 0.4529341835080514,
      "grad_norm": 0.4565776586532593,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 4191
    },
    {
      "epoch": 0.45304225656543823,
      "grad_norm": 0.5025374889373779,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 4192
    },
    {
      "epoch": 0.45315032962282503,
      "grad_norm": 0.44019415974617004,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 4193
    },
    {
      "epoch": 0.45325840268021184,
      "grad_norm": 0.4534400701522827,
      "learning_rate": 8e-05,
      "loss": 1.8236,
      "step": 4194
    },
    {
      "epoch": 0.45336647573759864,
      "grad_norm": 0.4240535497665405,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 4195
    },
    {
      "epoch": 0.4534745487949854,
      "grad_norm": 0.4639144539833069,
      "learning_rate": 8e-05,
      "loss": 1.8841,
      "step": 4196
    },
    {
      "epoch": 0.4535826218523722,
      "grad_norm": 0.44167816638946533,
      "learning_rate": 8e-05,
      "loss": 1.6295,
      "step": 4197
    },
    {
      "epoch": 0.453690694909759,
      "grad_norm": 0.46004346013069153,
      "learning_rate": 8e-05,
      "loss": 1.7839,
      "step": 4198
    },
    {
      "epoch": 0.4537987679671458,
      "grad_norm": 0.4166390895843506,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 4199
    },
    {
      "epoch": 0.4539068410245326,
      "grad_norm": 0.46320784091949463,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 4200
    },
    {
      "epoch": 0.45401491408191935,
      "grad_norm": 0.4946112334728241,
      "learning_rate": 8e-05,
      "loss": 1.841,
      "step": 4201
    },
    {
      "epoch": 0.45412298713930616,
      "grad_norm": 0.45610031485557556,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 4202
    },
    {
      "epoch": 0.45423106019669296,
      "grad_norm": 0.427095890045166,
      "learning_rate": 8e-05,
      "loss": 1.7518,
      "step": 4203
    },
    {
      "epoch": 0.45433913325407976,
      "grad_norm": 0.4529431164264679,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 4204
    },
    {
      "epoch": 0.45444720631146657,
      "grad_norm": 0.4154171943664551,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 4205
    },
    {
      "epoch": 0.45455527936885337,
      "grad_norm": 0.4628804624080658,
      "learning_rate": 8e-05,
      "loss": 1.8499,
      "step": 4206
    },
    {
      "epoch": 0.4546633524262401,
      "grad_norm": 0.4254166781902313,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 4207
    },
    {
      "epoch": 0.4547714254836269,
      "grad_norm": 0.40442150831222534,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 4208
    },
    {
      "epoch": 0.4548794985410137,
      "grad_norm": 0.44500336050987244,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 4209
    },
    {
      "epoch": 0.45498757159840053,
      "grad_norm": 0.42877069115638733,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 4210
    },
    {
      "epoch": 0.45509564465578733,
      "grad_norm": 0.4354883134365082,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 4211
    },
    {
      "epoch": 0.4552037177131741,
      "grad_norm": 0.41898852586746216,
      "learning_rate": 8e-05,
      "loss": 1.7775,
      "step": 4212
    },
    {
      "epoch": 0.4553117907705609,
      "grad_norm": 0.4746129512786865,
      "learning_rate": 8e-05,
      "loss": 1.614,
      "step": 4213
    },
    {
      "epoch": 0.4554198638279477,
      "grad_norm": 0.4032132029533386,
      "learning_rate": 8e-05,
      "loss": 1.6705,
      "step": 4214
    },
    {
      "epoch": 0.4555279368853345,
      "grad_norm": 0.3982903063297272,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 4215
    },
    {
      "epoch": 0.4556360099427213,
      "grad_norm": 0.45867207646369934,
      "learning_rate": 8e-05,
      "loss": 1.8432,
      "step": 4216
    },
    {
      "epoch": 0.45574408300010805,
      "grad_norm": 0.4441671371459961,
      "learning_rate": 8e-05,
      "loss": 1.6226,
      "step": 4217
    },
    {
      "epoch": 0.45585215605749485,
      "grad_norm": 0.44110044836997986,
      "learning_rate": 8e-05,
      "loss": 1.8905,
      "step": 4218
    },
    {
      "epoch": 0.45596022911488165,
      "grad_norm": 0.41997888684272766,
      "learning_rate": 8e-05,
      "loss": 1.7924,
      "step": 4219
    },
    {
      "epoch": 0.45606830217226846,
      "grad_norm": 0.4172268509864807,
      "learning_rate": 8e-05,
      "loss": 1.7003,
      "step": 4220
    },
    {
      "epoch": 0.45617637522965526,
      "grad_norm": 0.49546557664871216,
      "learning_rate": 8e-05,
      "loss": 1.85,
      "step": 4221
    },
    {
      "epoch": 0.45628444828704207,
      "grad_norm": 0.44726142287254333,
      "learning_rate": 8e-05,
      "loss": 1.8328,
      "step": 4222
    },
    {
      "epoch": 0.4563925213444288,
      "grad_norm": 0.4518565535545349,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 4223
    },
    {
      "epoch": 0.4565005944018156,
      "grad_norm": 0.4169279932975769,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 4224
    },
    {
      "epoch": 0.4566086674592024,
      "grad_norm": 0.46425676345825195,
      "learning_rate": 8e-05,
      "loss": 1.8969,
      "step": 4225
    },
    {
      "epoch": 0.4567167405165892,
      "grad_norm": 0.4517851769924164,
      "learning_rate": 8e-05,
      "loss": 1.8366,
      "step": 4226
    },
    {
      "epoch": 0.45682481357397603,
      "grad_norm": 0.4327676594257355,
      "learning_rate": 8e-05,
      "loss": 1.6306,
      "step": 4227
    },
    {
      "epoch": 0.4569328866313628,
      "grad_norm": 0.4507025182247162,
      "learning_rate": 8e-05,
      "loss": 1.5439,
      "step": 4228
    },
    {
      "epoch": 0.4570409596887496,
      "grad_norm": 0.4557878077030182,
      "learning_rate": 8e-05,
      "loss": 1.622,
      "step": 4229
    },
    {
      "epoch": 0.4571490327461364,
      "grad_norm": 0.46204566955566406,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 4230
    },
    {
      "epoch": 0.4572571058035232,
      "grad_norm": 0.43274375796318054,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 4231
    },
    {
      "epoch": 0.45736517886091,
      "grad_norm": 0.4206010401248932,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 4232
    },
    {
      "epoch": 0.45747325191829674,
      "grad_norm": 0.4545277953147888,
      "learning_rate": 8e-05,
      "loss": 1.7139,
      "step": 4233
    },
    {
      "epoch": 0.45758132497568355,
      "grad_norm": 0.46365052461624146,
      "learning_rate": 8e-05,
      "loss": 1.7398,
      "step": 4234
    },
    {
      "epoch": 0.45768939803307035,
      "grad_norm": 0.4478015899658203,
      "learning_rate": 8e-05,
      "loss": 1.4424,
      "step": 4235
    },
    {
      "epoch": 0.45779747109045715,
      "grad_norm": 0.44459959864616394,
      "learning_rate": 8e-05,
      "loss": 1.7265,
      "step": 4236
    },
    {
      "epoch": 0.45790554414784396,
      "grad_norm": 0.4216412603855133,
      "learning_rate": 8e-05,
      "loss": 1.4975,
      "step": 4237
    },
    {
      "epoch": 0.45801361720523076,
      "grad_norm": 0.44938787817955017,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 4238
    },
    {
      "epoch": 0.4581216902626175,
      "grad_norm": 0.4305035173892975,
      "learning_rate": 8e-05,
      "loss": 1.5852,
      "step": 4239
    },
    {
      "epoch": 0.4582297633200043,
      "grad_norm": 0.42821428179740906,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 4240
    },
    {
      "epoch": 0.4583378363773911,
      "grad_norm": 0.4074019491672516,
      "learning_rate": 8e-05,
      "loss": 1.6464,
      "step": 4241
    },
    {
      "epoch": 0.4584459094347779,
      "grad_norm": 0.432125449180603,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 4242
    },
    {
      "epoch": 0.4585539824921647,
      "grad_norm": 0.5118075013160706,
      "learning_rate": 8e-05,
      "loss": 1.9402,
      "step": 4243
    },
    {
      "epoch": 0.4586620555495515,
      "grad_norm": 0.447480171918869,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 4244
    },
    {
      "epoch": 0.4587701286069383,
      "grad_norm": 0.46424180269241333,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 4245
    },
    {
      "epoch": 0.4588782016643251,
      "grad_norm": 0.4634558856487274,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 4246
    },
    {
      "epoch": 0.4589862747217119,
      "grad_norm": 0.4220263659954071,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 4247
    },
    {
      "epoch": 0.4590943477790987,
      "grad_norm": 0.4639836847782135,
      "learning_rate": 8e-05,
      "loss": 1.4606,
      "step": 4248
    },
    {
      "epoch": 0.45920242083648544,
      "grad_norm": 0.4912407398223877,
      "learning_rate": 8e-05,
      "loss": 1.8933,
      "step": 4249
    },
    {
      "epoch": 0.45931049389387224,
      "grad_norm": 0.43755170702934265,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 4250
    },
    {
      "epoch": 0.45941856695125904,
      "grad_norm": 0.43365931510925293,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 4251
    },
    {
      "epoch": 0.45952664000864585,
      "grad_norm": 0.41320517659187317,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 4252
    },
    {
      "epoch": 0.45963471306603265,
      "grad_norm": 0.41983312368392944,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 4253
    },
    {
      "epoch": 0.45974278612341946,
      "grad_norm": 0.42392438650131226,
      "learning_rate": 8e-05,
      "loss": 1.6714,
      "step": 4254
    },
    {
      "epoch": 0.4598508591808062,
      "grad_norm": 0.44541335105895996,
      "learning_rate": 8e-05,
      "loss": 1.8523,
      "step": 4255
    },
    {
      "epoch": 0.459958932238193,
      "grad_norm": 0.44276705384254456,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 4256
    },
    {
      "epoch": 0.4600670052955798,
      "grad_norm": 0.4369412958621979,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 4257
    },
    {
      "epoch": 0.4601750783529666,
      "grad_norm": 0.44248175621032715,
      "learning_rate": 8e-05,
      "loss": 1.7145,
      "step": 4258
    },
    {
      "epoch": 0.4602831514103534,
      "grad_norm": 0.4203895628452301,
      "learning_rate": 8e-05,
      "loss": 1.666,
      "step": 4259
    },
    {
      "epoch": 0.46039122446774017,
      "grad_norm": 0.447632759809494,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 4260
    },
    {
      "epoch": 0.46049929752512697,
      "grad_norm": 0.49456626176834106,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 4261
    },
    {
      "epoch": 0.4606073705825138,
      "grad_norm": 0.4450577199459076,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 4262
    },
    {
      "epoch": 0.4607154436399006,
      "grad_norm": 0.4206874668598175,
      "learning_rate": 8e-05,
      "loss": 1.6913,
      "step": 4263
    },
    {
      "epoch": 0.4608235166972874,
      "grad_norm": 0.42525988817214966,
      "learning_rate": 8e-05,
      "loss": 1.5595,
      "step": 4264
    },
    {
      "epoch": 0.4609315897546742,
      "grad_norm": 0.38732588291168213,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 4265
    },
    {
      "epoch": 0.46103966281206094,
      "grad_norm": 0.5574738383293152,
      "learning_rate": 8e-05,
      "loss": 2.1112,
      "step": 4266
    },
    {
      "epoch": 0.46114773586944774,
      "grad_norm": 0.42982858419418335,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 4267
    },
    {
      "epoch": 0.46125580892683454,
      "grad_norm": 0.42815694212913513,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 4268
    },
    {
      "epoch": 0.46136388198422135,
      "grad_norm": 0.40028703212738037,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 4269
    },
    {
      "epoch": 0.46147195504160815,
      "grad_norm": 0.477752685546875,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 4270
    },
    {
      "epoch": 0.4615800280989949,
      "grad_norm": 0.44198429584503174,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 4271
    },
    {
      "epoch": 0.4616881011563817,
      "grad_norm": 0.47476726770401,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 4272
    },
    {
      "epoch": 0.4617961742137685,
      "grad_norm": 0.5240785479545593,
      "learning_rate": 8e-05,
      "loss": 1.8255,
      "step": 4273
    },
    {
      "epoch": 0.4619042472711553,
      "grad_norm": 0.429733008146286,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 4274
    },
    {
      "epoch": 0.4620123203285421,
      "grad_norm": 0.45686212182044983,
      "learning_rate": 8e-05,
      "loss": 1.7874,
      "step": 4275
    },
    {
      "epoch": 0.46212039338592886,
      "grad_norm": 0.421690970659256,
      "learning_rate": 8e-05,
      "loss": 1.6838,
      "step": 4276
    },
    {
      "epoch": 0.46222846644331567,
      "grad_norm": 0.4692758023738861,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 4277
    },
    {
      "epoch": 0.46233653950070247,
      "grad_norm": 0.44985368847846985,
      "learning_rate": 8e-05,
      "loss": 1.8883,
      "step": 4278
    },
    {
      "epoch": 0.4624446125580893,
      "grad_norm": 0.43135830760002136,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 4279
    },
    {
      "epoch": 0.4625526856154761,
      "grad_norm": 0.45663487911224365,
      "learning_rate": 8e-05,
      "loss": 1.5877,
      "step": 4280
    },
    {
      "epoch": 0.4626607586728629,
      "grad_norm": 0.4858367443084717,
      "learning_rate": 8e-05,
      "loss": 1.8739,
      "step": 4281
    },
    {
      "epoch": 0.46276883173024963,
      "grad_norm": 0.44199952483177185,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 4282
    },
    {
      "epoch": 0.46287690478763643,
      "grad_norm": 0.4228383004665375,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 4283
    },
    {
      "epoch": 0.46298497784502324,
      "grad_norm": 0.4293738305568695,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 4284
    },
    {
      "epoch": 0.46309305090241004,
      "grad_norm": 0.44804099202156067,
      "learning_rate": 8e-05,
      "loss": 1.7566,
      "step": 4285
    },
    {
      "epoch": 0.46320112395979685,
      "grad_norm": 0.45113250613212585,
      "learning_rate": 8e-05,
      "loss": 1.8708,
      "step": 4286
    },
    {
      "epoch": 0.4633091970171836,
      "grad_norm": 0.44327476620674133,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 4287
    },
    {
      "epoch": 0.4634172700745704,
      "grad_norm": 0.4531184136867523,
      "learning_rate": 8e-05,
      "loss": 1.6846,
      "step": 4288
    },
    {
      "epoch": 0.4635253431319572,
      "grad_norm": 0.46570998430252075,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 4289
    },
    {
      "epoch": 0.463633416189344,
      "grad_norm": 0.537976086139679,
      "learning_rate": 8e-05,
      "loss": 1.6917,
      "step": 4290
    },
    {
      "epoch": 0.4637414892467308,
      "grad_norm": 0.44095200300216675,
      "learning_rate": 8e-05,
      "loss": 1.7609,
      "step": 4291
    },
    {
      "epoch": 0.46384956230411756,
      "grad_norm": 0.443903386592865,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 4292
    },
    {
      "epoch": 0.46395763536150436,
      "grad_norm": 0.40788063406944275,
      "learning_rate": 8e-05,
      "loss": 1.6518,
      "step": 4293
    },
    {
      "epoch": 0.46406570841889117,
      "grad_norm": 0.47895747423171997,
      "learning_rate": 8e-05,
      "loss": 1.8656,
      "step": 4294
    },
    {
      "epoch": 0.46417378147627797,
      "grad_norm": 0.43863290548324585,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 4295
    },
    {
      "epoch": 0.4642818545336648,
      "grad_norm": 0.446875661611557,
      "learning_rate": 8e-05,
      "loss": 1.6685,
      "step": 4296
    },
    {
      "epoch": 0.4643899275910516,
      "grad_norm": 0.4459078013896942,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 4297
    },
    {
      "epoch": 0.4644980006484383,
      "grad_norm": 0.4376997947692871,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 4298
    },
    {
      "epoch": 0.46460607370582513,
      "grad_norm": 0.40418586134910583,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 4299
    },
    {
      "epoch": 0.46471414676321193,
      "grad_norm": 0.44401949644088745,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 4300
    },
    {
      "epoch": 0.46482221982059874,
      "grad_norm": 0.4577077031135559,
      "learning_rate": 8e-05,
      "loss": 1.7989,
      "step": 4301
    },
    {
      "epoch": 0.46493029287798554,
      "grad_norm": 0.42888376116752625,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 4302
    },
    {
      "epoch": 0.4650383659353723,
      "grad_norm": 0.4292757213115692,
      "learning_rate": 8e-05,
      "loss": 1.7338,
      "step": 4303
    },
    {
      "epoch": 0.4651464389927591,
      "grad_norm": 0.4548374116420746,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 4304
    },
    {
      "epoch": 0.4652545120501459,
      "grad_norm": 0.44983065128326416,
      "learning_rate": 8e-05,
      "loss": 1.8761,
      "step": 4305
    },
    {
      "epoch": 0.4653625851075327,
      "grad_norm": 0.46727675199508667,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 4306
    },
    {
      "epoch": 0.4654706581649195,
      "grad_norm": 0.3978903293609619,
      "learning_rate": 8e-05,
      "loss": 1.3426,
      "step": 4307
    },
    {
      "epoch": 0.46557873122230625,
      "grad_norm": 0.453128457069397,
      "learning_rate": 8e-05,
      "loss": 1.7911,
      "step": 4308
    },
    {
      "epoch": 0.46568680427969306,
      "grad_norm": 0.4320334196090698,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 4309
    },
    {
      "epoch": 0.46579487733707986,
      "grad_norm": 0.4326897859573364,
      "learning_rate": 8e-05,
      "loss": 1.6903,
      "step": 4310
    },
    {
      "epoch": 0.46590295039446666,
      "grad_norm": 0.43983137607574463,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 4311
    },
    {
      "epoch": 0.46601102345185347,
      "grad_norm": 0.4165794551372528,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 4312
    },
    {
      "epoch": 0.46611909650924027,
      "grad_norm": 0.46860310435295105,
      "learning_rate": 8e-05,
      "loss": 1.7415,
      "step": 4313
    },
    {
      "epoch": 0.466227169566627,
      "grad_norm": 0.42817097902297974,
      "learning_rate": 8e-05,
      "loss": 1.6431,
      "step": 4314
    },
    {
      "epoch": 0.4663352426240138,
      "grad_norm": 0.41663217544555664,
      "learning_rate": 8e-05,
      "loss": 1.593,
      "step": 4315
    },
    {
      "epoch": 0.4664433156814006,
      "grad_norm": 0.4294176995754242,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 4316
    },
    {
      "epoch": 0.46655138873878743,
      "grad_norm": 0.4219122529029846,
      "learning_rate": 8e-05,
      "loss": 1.791,
      "step": 4317
    },
    {
      "epoch": 0.46665946179617424,
      "grad_norm": 0.4432951807975769,
      "learning_rate": 8e-05,
      "loss": 1.6982,
      "step": 4318
    },
    {
      "epoch": 0.466767534853561,
      "grad_norm": 0.41083037853240967,
      "learning_rate": 8e-05,
      "loss": 1.5087,
      "step": 4319
    },
    {
      "epoch": 0.4668756079109478,
      "grad_norm": 0.4323696196079254,
      "learning_rate": 8e-05,
      "loss": 1.5386,
      "step": 4320
    },
    {
      "epoch": 0.4669836809683346,
      "grad_norm": 0.405820369720459,
      "learning_rate": 8e-05,
      "loss": 1.7521,
      "step": 4321
    },
    {
      "epoch": 0.4670917540257214,
      "grad_norm": 0.4635251760482788,
      "learning_rate": 8e-05,
      "loss": 1.8244,
      "step": 4322
    },
    {
      "epoch": 0.4671998270831082,
      "grad_norm": 0.5216931104660034,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 4323
    },
    {
      "epoch": 0.46730790014049495,
      "grad_norm": 0.43576428294181824,
      "learning_rate": 8e-05,
      "loss": 1.599,
      "step": 4324
    },
    {
      "epoch": 0.46741597319788175,
      "grad_norm": 0.4944605529308319,
      "learning_rate": 8e-05,
      "loss": 1.8444,
      "step": 4325
    },
    {
      "epoch": 0.46752404625526855,
      "grad_norm": 0.418350487947464,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 4326
    },
    {
      "epoch": 0.46763211931265536,
      "grad_norm": 0.4392555356025696,
      "learning_rate": 8e-05,
      "loss": 1.7589,
      "step": 4327
    },
    {
      "epoch": 0.46774019237004216,
      "grad_norm": 0.47517502307891846,
      "learning_rate": 8e-05,
      "loss": 1.886,
      "step": 4328
    },
    {
      "epoch": 0.46784826542742897,
      "grad_norm": 0.46860334277153015,
      "learning_rate": 8e-05,
      "loss": 1.7728,
      "step": 4329
    },
    {
      "epoch": 0.4679563384848157,
      "grad_norm": 0.44872376322746277,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 4330
    },
    {
      "epoch": 0.4680644115422025,
      "grad_norm": 0.47071564197540283,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 4331
    },
    {
      "epoch": 0.4681724845995893,
      "grad_norm": 0.46122756600379944,
      "learning_rate": 8e-05,
      "loss": 1.8181,
      "step": 4332
    },
    {
      "epoch": 0.4682805576569761,
      "grad_norm": 0.44584915041923523,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 4333
    },
    {
      "epoch": 0.46838863071436293,
      "grad_norm": 0.41938602924346924,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 4334
    },
    {
      "epoch": 0.4684967037717497,
      "grad_norm": 0.433266282081604,
      "learning_rate": 8e-05,
      "loss": 1.7036,
      "step": 4335
    },
    {
      "epoch": 0.4686047768291365,
      "grad_norm": 0.4607677161693573,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 4336
    },
    {
      "epoch": 0.4687128498865233,
      "grad_norm": 0.40440666675567627,
      "learning_rate": 8e-05,
      "loss": 1.59,
      "step": 4337
    },
    {
      "epoch": 0.4688209229439101,
      "grad_norm": 0.42527639865875244,
      "learning_rate": 8e-05,
      "loss": 1.8059,
      "step": 4338
    },
    {
      "epoch": 0.4689289960012969,
      "grad_norm": 0.4593566954135895,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 4339
    },
    {
      "epoch": 0.4690370690586837,
      "grad_norm": 0.4832436442375183,
      "learning_rate": 8e-05,
      "loss": 1.924,
      "step": 4340
    },
    {
      "epoch": 0.46914514211607045,
      "grad_norm": 0.4470100700855255,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 4341
    },
    {
      "epoch": 0.46925321517345725,
      "grad_norm": 0.6064956188201904,
      "learning_rate": 8e-05,
      "loss": 1.9722,
      "step": 4342
    },
    {
      "epoch": 0.46936128823084405,
      "grad_norm": 0.41135895252227783,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 4343
    },
    {
      "epoch": 0.46946936128823086,
      "grad_norm": 0.45011064410209656,
      "learning_rate": 8e-05,
      "loss": 1.7102,
      "step": 4344
    },
    {
      "epoch": 0.46957743434561766,
      "grad_norm": 0.4148107171058655,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 4345
    },
    {
      "epoch": 0.4696855074030044,
      "grad_norm": 0.4377236068248749,
      "learning_rate": 8e-05,
      "loss": 1.619,
      "step": 4346
    },
    {
      "epoch": 0.4697935804603912,
      "grad_norm": 0.47588443756103516,
      "learning_rate": 8e-05,
      "loss": 1.8251,
      "step": 4347
    },
    {
      "epoch": 0.469901653517778,
      "grad_norm": 0.4610872268676758,
      "learning_rate": 8e-05,
      "loss": 1.9095,
      "step": 4348
    },
    {
      "epoch": 0.4700097265751648,
      "grad_norm": 0.4461185038089752,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 4349
    },
    {
      "epoch": 0.4701177996325516,
      "grad_norm": 0.44669854640960693,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 4350
    },
    {
      "epoch": 0.4702258726899384,
      "grad_norm": 0.4431760907173157,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 4351
    },
    {
      "epoch": 0.4703339457473252,
      "grad_norm": 0.5018522143363953,
      "learning_rate": 8e-05,
      "loss": 1.9256,
      "step": 4352
    },
    {
      "epoch": 0.470442018804712,
      "grad_norm": 0.5249881744384766,
      "learning_rate": 8e-05,
      "loss": 1.8146,
      "step": 4353
    },
    {
      "epoch": 0.4705500918620988,
      "grad_norm": 0.43779247999191284,
      "learning_rate": 8e-05,
      "loss": 1.5307,
      "step": 4354
    },
    {
      "epoch": 0.4706581649194856,
      "grad_norm": 0.42912545800209045,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 4355
    },
    {
      "epoch": 0.4707662379768724,
      "grad_norm": 0.4825234115123749,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 4356
    },
    {
      "epoch": 0.47087431103425914,
      "grad_norm": 0.42840513586997986,
      "learning_rate": 8e-05,
      "loss": 1.8128,
      "step": 4357
    },
    {
      "epoch": 0.47098238409164594,
      "grad_norm": 0.4618459641933441,
      "learning_rate": 8e-05,
      "loss": 1.5712,
      "step": 4358
    },
    {
      "epoch": 0.47109045714903275,
      "grad_norm": 0.4747450649738312,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 4359
    },
    {
      "epoch": 0.47119853020641955,
      "grad_norm": 0.4267098307609558,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 4360
    },
    {
      "epoch": 0.47130660326380636,
      "grad_norm": 0.4711913466453552,
      "learning_rate": 8e-05,
      "loss": 1.9219,
      "step": 4361
    },
    {
      "epoch": 0.4714146763211931,
      "grad_norm": 0.43364372849464417,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 4362
    },
    {
      "epoch": 0.4715227493785799,
      "grad_norm": 0.5234556794166565,
      "learning_rate": 8e-05,
      "loss": 2.0267,
      "step": 4363
    },
    {
      "epoch": 0.4716308224359667,
      "grad_norm": 0.4821789264678955,
      "learning_rate": 8e-05,
      "loss": 1.8555,
      "step": 4364
    },
    {
      "epoch": 0.4717388954933535,
      "grad_norm": 0.4472949802875519,
      "learning_rate": 8e-05,
      "loss": 1.4265,
      "step": 4365
    },
    {
      "epoch": 0.4718469685507403,
      "grad_norm": 0.5096022486686707,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 4366
    },
    {
      "epoch": 0.47195504160812707,
      "grad_norm": 0.5463072061538696,
      "learning_rate": 8e-05,
      "loss": 2.1129,
      "step": 4367
    },
    {
      "epoch": 0.47206311466551387,
      "grad_norm": 0.44732269644737244,
      "learning_rate": 8e-05,
      "loss": 1.6373,
      "step": 4368
    },
    {
      "epoch": 0.4721711877229007,
      "grad_norm": 0.44337233901023865,
      "learning_rate": 8e-05,
      "loss": 1.7932,
      "step": 4369
    },
    {
      "epoch": 0.4722792607802875,
      "grad_norm": 0.4418798089027405,
      "learning_rate": 8e-05,
      "loss": 1.6553,
      "step": 4370
    },
    {
      "epoch": 0.4723873338376743,
      "grad_norm": 0.44732338190078735,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 4371
    },
    {
      "epoch": 0.4724954068950611,
      "grad_norm": 0.4226166605949402,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 4372
    },
    {
      "epoch": 0.47260347995244784,
      "grad_norm": 0.5153347849845886,
      "learning_rate": 8e-05,
      "loss": 1.9174,
      "step": 4373
    },
    {
      "epoch": 0.47271155300983464,
      "grad_norm": 0.4198814928531647,
      "learning_rate": 8e-05,
      "loss": 1.6567,
      "step": 4374
    },
    {
      "epoch": 0.47281962606722144,
      "grad_norm": 0.4688476026058197,
      "learning_rate": 8e-05,
      "loss": 1.6643,
      "step": 4375
    },
    {
      "epoch": 0.47292769912460825,
      "grad_norm": 0.42378145456314087,
      "learning_rate": 8e-05,
      "loss": 1.6372,
      "step": 4376
    },
    {
      "epoch": 0.47303577218199505,
      "grad_norm": 0.4594070613384247,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 4377
    },
    {
      "epoch": 0.4731438452393818,
      "grad_norm": 0.4955171048641205,
      "learning_rate": 8e-05,
      "loss": 1.8419,
      "step": 4378
    },
    {
      "epoch": 0.4732519182967686,
      "grad_norm": 0.4390750825405121,
      "learning_rate": 8e-05,
      "loss": 1.7737,
      "step": 4379
    },
    {
      "epoch": 0.4733599913541554,
      "grad_norm": 0.4249928593635559,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 4380
    },
    {
      "epoch": 0.4734680644115422,
      "grad_norm": 0.447437047958374,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 4381
    },
    {
      "epoch": 0.473576137468929,
      "grad_norm": 0.4561867415904999,
      "learning_rate": 8e-05,
      "loss": 1.5922,
      "step": 4382
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 0.4181942939758301,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 4383
    },
    {
      "epoch": 0.47379228358370257,
      "grad_norm": 0.4354732632637024,
      "learning_rate": 8e-05,
      "loss": 1.8329,
      "step": 4384
    },
    {
      "epoch": 0.47390035664108937,
      "grad_norm": 0.4717293977737427,
      "learning_rate": 8e-05,
      "loss": 1.9425,
      "step": 4385
    },
    {
      "epoch": 0.4740084296984762,
      "grad_norm": 0.4507363736629486,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 4386
    },
    {
      "epoch": 0.474116502755863,
      "grad_norm": 0.46960678696632385,
      "learning_rate": 8e-05,
      "loss": 1.746,
      "step": 4387
    },
    {
      "epoch": 0.4742245758132498,
      "grad_norm": 0.4162602424621582,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 4388
    },
    {
      "epoch": 0.47433264887063653,
      "grad_norm": 0.5015707612037659,
      "learning_rate": 8e-05,
      "loss": 1.9575,
      "step": 4389
    },
    {
      "epoch": 0.47444072192802333,
      "grad_norm": 0.49720466136932373,
      "learning_rate": 8e-05,
      "loss": 1.7673,
      "step": 4390
    },
    {
      "epoch": 0.47454879498541014,
      "grad_norm": 0.4223114550113678,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 4391
    },
    {
      "epoch": 0.47465686804279694,
      "grad_norm": 0.4195987284183502,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 4392
    },
    {
      "epoch": 0.47476494110018375,
      "grad_norm": 0.47932130098342896,
      "learning_rate": 8e-05,
      "loss": 1.8237,
      "step": 4393
    },
    {
      "epoch": 0.4748730141575705,
      "grad_norm": 0.43170279264450073,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 4394
    },
    {
      "epoch": 0.4749810872149573,
      "grad_norm": 0.45804262161254883,
      "learning_rate": 8e-05,
      "loss": 1.663,
      "step": 4395
    },
    {
      "epoch": 0.4750891602723441,
      "grad_norm": 0.42211320996284485,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 4396
    },
    {
      "epoch": 0.4751972333297309,
      "grad_norm": 0.42635220289230347,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 4397
    },
    {
      "epoch": 0.4753053063871177,
      "grad_norm": 0.4456947445869446,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 4398
    },
    {
      "epoch": 0.47541337944450446,
      "grad_norm": 0.4579963684082031,
      "learning_rate": 8e-05,
      "loss": 1.5785,
      "step": 4399
    },
    {
      "epoch": 0.47552145250189126,
      "grad_norm": 0.4507114589214325,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 4400
    },
    {
      "epoch": 0.47562952555927807,
      "grad_norm": 0.4416195750236511,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 4401
    },
    {
      "epoch": 0.47573759861666487,
      "grad_norm": 0.4365001320838928,
      "learning_rate": 8e-05,
      "loss": 1.7664,
      "step": 4402
    },
    {
      "epoch": 0.4758456716740517,
      "grad_norm": 0.4335675835609436,
      "learning_rate": 8e-05,
      "loss": 1.7544,
      "step": 4403
    },
    {
      "epoch": 0.4759537447314385,
      "grad_norm": 0.4506111145019531,
      "learning_rate": 8e-05,
      "loss": 1.8302,
      "step": 4404
    },
    {
      "epoch": 0.4760618177888252,
      "grad_norm": 0.40573450922966003,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 4405
    },
    {
      "epoch": 0.47616989084621203,
      "grad_norm": 0.45827099680900574,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 4406
    },
    {
      "epoch": 0.47627796390359883,
      "grad_norm": 0.4925473928451538,
      "learning_rate": 8e-05,
      "loss": 2.0702,
      "step": 4407
    },
    {
      "epoch": 0.47638603696098564,
      "grad_norm": 0.44004854559898376,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 4408
    },
    {
      "epoch": 0.47649411001837244,
      "grad_norm": 0.4239892363548279,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 4409
    },
    {
      "epoch": 0.4766021830757592,
      "grad_norm": 0.4138792157173157,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 4410
    },
    {
      "epoch": 0.476710256133146,
      "grad_norm": 0.5054914355278015,
      "learning_rate": 8e-05,
      "loss": 1.8722,
      "step": 4411
    },
    {
      "epoch": 0.4768183291905328,
      "grad_norm": 0.45110830664634705,
      "learning_rate": 8e-05,
      "loss": 1.5035,
      "step": 4412
    },
    {
      "epoch": 0.4769264022479196,
      "grad_norm": 0.4241587221622467,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 4413
    },
    {
      "epoch": 0.4770344753053064,
      "grad_norm": 0.49000638723373413,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 4414
    },
    {
      "epoch": 0.47714254836269315,
      "grad_norm": 0.45393577218055725,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 4415
    },
    {
      "epoch": 0.47725062142007996,
      "grad_norm": 0.4646742045879364,
      "learning_rate": 8e-05,
      "loss": 1.8183,
      "step": 4416
    },
    {
      "epoch": 0.47735869447746676,
      "grad_norm": 0.42487719655036926,
      "learning_rate": 8e-05,
      "loss": 1.5823,
      "step": 4417
    },
    {
      "epoch": 0.47746676753485356,
      "grad_norm": 0.43750110268592834,
      "learning_rate": 8e-05,
      "loss": 1.7336,
      "step": 4418
    },
    {
      "epoch": 0.47757484059224037,
      "grad_norm": 0.4200732111930847,
      "learning_rate": 8e-05,
      "loss": 1.5415,
      "step": 4419
    },
    {
      "epoch": 0.47768291364962717,
      "grad_norm": 0.4106476902961731,
      "learning_rate": 8e-05,
      "loss": 1.6586,
      "step": 4420
    },
    {
      "epoch": 0.4777909867070139,
      "grad_norm": 0.4483850598335266,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 4421
    },
    {
      "epoch": 0.4778990597644007,
      "grad_norm": 0.44243308901786804,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 4422
    },
    {
      "epoch": 0.47800713282178753,
      "grad_norm": 0.4649585485458374,
      "learning_rate": 8e-05,
      "loss": 1.7535,
      "step": 4423
    },
    {
      "epoch": 0.47811520587917433,
      "grad_norm": 0.4760155975818634,
      "learning_rate": 8e-05,
      "loss": 1.813,
      "step": 4424
    },
    {
      "epoch": 0.47822327893656114,
      "grad_norm": 0.46781569719314575,
      "learning_rate": 8e-05,
      "loss": 1.9266,
      "step": 4425
    },
    {
      "epoch": 0.4783313519939479,
      "grad_norm": 0.45242437720298767,
      "learning_rate": 8e-05,
      "loss": 1.7369,
      "step": 4426
    },
    {
      "epoch": 0.4784394250513347,
      "grad_norm": 0.4705638587474823,
      "learning_rate": 8e-05,
      "loss": 1.81,
      "step": 4427
    },
    {
      "epoch": 0.4785474981087215,
      "grad_norm": 0.4353788495063782,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 4428
    },
    {
      "epoch": 0.4786555711661083,
      "grad_norm": 0.4197918772697449,
      "learning_rate": 8e-05,
      "loss": 1.5577,
      "step": 4429
    },
    {
      "epoch": 0.4787636442234951,
      "grad_norm": 0.4873145818710327,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 4430
    },
    {
      "epoch": 0.4788717172808819,
      "grad_norm": 0.45300522446632385,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 4431
    },
    {
      "epoch": 0.47897979033826865,
      "grad_norm": 0.4565155804157257,
      "learning_rate": 8e-05,
      "loss": 1.749,
      "step": 4432
    },
    {
      "epoch": 0.47908786339565546,
      "grad_norm": 0.4308806359767914,
      "learning_rate": 8e-05,
      "loss": 1.5408,
      "step": 4433
    },
    {
      "epoch": 0.47919593645304226,
      "grad_norm": 0.43796491622924805,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 4434
    },
    {
      "epoch": 0.47930400951042906,
      "grad_norm": 0.5056694149971008,
      "learning_rate": 8e-05,
      "loss": 1.8521,
      "step": 4435
    },
    {
      "epoch": 0.47941208256781587,
      "grad_norm": 0.4281952679157257,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 4436
    },
    {
      "epoch": 0.4795201556252026,
      "grad_norm": 0.4764220118522644,
      "learning_rate": 8e-05,
      "loss": 1.7011,
      "step": 4437
    },
    {
      "epoch": 0.4796282286825894,
      "grad_norm": 0.43882760405540466,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 4438
    },
    {
      "epoch": 0.4797363017399762,
      "grad_norm": 0.4294537901878357,
      "learning_rate": 8e-05,
      "loss": 1.6579,
      "step": 4439
    },
    {
      "epoch": 0.479844374797363,
      "grad_norm": 0.4423680603504181,
      "learning_rate": 8e-05,
      "loss": 1.6761,
      "step": 4440
    },
    {
      "epoch": 0.47995244785474983,
      "grad_norm": 0.40551096200942993,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 4441
    },
    {
      "epoch": 0.4800605209121366,
      "grad_norm": 0.4423576295375824,
      "learning_rate": 8e-05,
      "loss": 1.8854,
      "step": 4442
    },
    {
      "epoch": 0.4801685939695234,
      "grad_norm": 0.4424543082714081,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 4443
    },
    {
      "epoch": 0.4802766670269102,
      "grad_norm": 0.40267324447631836,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 4444
    },
    {
      "epoch": 0.480384740084297,
      "grad_norm": 0.46327799558639526,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 4445
    },
    {
      "epoch": 0.4804928131416838,
      "grad_norm": 0.47183164954185486,
      "learning_rate": 8e-05,
      "loss": 1.8294,
      "step": 4446
    },
    {
      "epoch": 0.4806008861990706,
      "grad_norm": 0.42047256231307983,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 4447
    },
    {
      "epoch": 0.48070895925645735,
      "grad_norm": 0.4473212659358978,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 4448
    },
    {
      "epoch": 0.48081703231384415,
      "grad_norm": 0.4599415063858032,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 4449
    },
    {
      "epoch": 0.48092510537123095,
      "grad_norm": 0.42555320262908936,
      "learning_rate": 8e-05,
      "loss": 1.5737,
      "step": 4450
    },
    {
      "epoch": 0.48103317842861776,
      "grad_norm": 0.4502354562282562,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 4451
    },
    {
      "epoch": 0.48114125148600456,
      "grad_norm": 0.48069891333580017,
      "learning_rate": 8e-05,
      "loss": 1.8428,
      "step": 4452
    },
    {
      "epoch": 0.4812493245433913,
      "grad_norm": 0.4971717894077301,
      "learning_rate": 8e-05,
      "loss": 1.8602,
      "step": 4453
    },
    {
      "epoch": 0.4813573976007781,
      "grad_norm": 0.4574388265609741,
      "learning_rate": 8e-05,
      "loss": 1.7209,
      "step": 4454
    },
    {
      "epoch": 0.4814654706581649,
      "grad_norm": 0.4342834949493408,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 4455
    },
    {
      "epoch": 0.4815735437155517,
      "grad_norm": 0.4914734959602356,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 4456
    },
    {
      "epoch": 0.4816816167729385,
      "grad_norm": 0.4828100800514221,
      "learning_rate": 8e-05,
      "loss": 1.9207,
      "step": 4457
    },
    {
      "epoch": 0.4817896898303253,
      "grad_norm": 0.5023161768913269,
      "learning_rate": 8e-05,
      "loss": 1.706,
      "step": 4458
    },
    {
      "epoch": 0.4818977628877121,
      "grad_norm": 0.44304654002189636,
      "learning_rate": 8e-05,
      "loss": 1.6266,
      "step": 4459
    },
    {
      "epoch": 0.4820058359450989,
      "grad_norm": 0.4469166100025177,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 4460
    },
    {
      "epoch": 0.4821139090024857,
      "grad_norm": 0.46235737204551697,
      "learning_rate": 8e-05,
      "loss": 1.7958,
      "step": 4461
    },
    {
      "epoch": 0.4822219820598725,
      "grad_norm": 0.47271183133125305,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 4462
    },
    {
      "epoch": 0.4823300551172593,
      "grad_norm": 0.4280488193035126,
      "learning_rate": 8e-05,
      "loss": 1.8014,
      "step": 4463
    },
    {
      "epoch": 0.48243812817464604,
      "grad_norm": 0.43174272775650024,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 4464
    },
    {
      "epoch": 0.48254620123203285,
      "grad_norm": 0.4262846112251282,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 4465
    },
    {
      "epoch": 0.48265427428941965,
      "grad_norm": 0.410413920879364,
      "learning_rate": 8e-05,
      "loss": 1.6539,
      "step": 4466
    },
    {
      "epoch": 0.48276234734680645,
      "grad_norm": 0.488225519657135,
      "learning_rate": 8e-05,
      "loss": 1.8058,
      "step": 4467
    },
    {
      "epoch": 0.48287042040419326,
      "grad_norm": 0.4624020755290985,
      "learning_rate": 8e-05,
      "loss": 1.8024,
      "step": 4468
    },
    {
      "epoch": 0.48297849346158,
      "grad_norm": 0.4150160849094391,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 4469
    },
    {
      "epoch": 0.4830865665189668,
      "grad_norm": 0.45198357105255127,
      "learning_rate": 8e-05,
      "loss": 1.6445,
      "step": 4470
    },
    {
      "epoch": 0.4831946395763536,
      "grad_norm": 0.43618258833885193,
      "learning_rate": 8e-05,
      "loss": 1.5575,
      "step": 4471
    },
    {
      "epoch": 0.4833027126337404,
      "grad_norm": 0.5713172554969788,
      "learning_rate": 8e-05,
      "loss": 1.8966,
      "step": 4472
    },
    {
      "epoch": 0.4834107856911272,
      "grad_norm": 0.504959762096405,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 4473
    },
    {
      "epoch": 0.48351885874851397,
      "grad_norm": 0.4237705171108246,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 4474
    },
    {
      "epoch": 0.4836269318059008,
      "grad_norm": 0.3990875482559204,
      "learning_rate": 8e-05,
      "loss": 1.5175,
      "step": 4475
    },
    {
      "epoch": 0.4837350048632876,
      "grad_norm": 0.42895573377609253,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 4476
    },
    {
      "epoch": 0.4838430779206744,
      "grad_norm": 0.44917482137680054,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 4477
    },
    {
      "epoch": 0.4839511509780612,
      "grad_norm": 0.4524247944355011,
      "learning_rate": 8e-05,
      "loss": 1.8297,
      "step": 4478
    },
    {
      "epoch": 0.484059224035448,
      "grad_norm": 0.4127720296382904,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 4479
    },
    {
      "epoch": 0.48416729709283474,
      "grad_norm": 0.4942583441734314,
      "learning_rate": 8e-05,
      "loss": 1.8264,
      "step": 4480
    },
    {
      "epoch": 0.48427537015022154,
      "grad_norm": 0.4435935914516449,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 4481
    },
    {
      "epoch": 0.48438344320760834,
      "grad_norm": 0.4167121946811676,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 4482
    },
    {
      "epoch": 0.48449151626499515,
      "grad_norm": 0.4711264967918396,
      "learning_rate": 8e-05,
      "loss": 1.8344,
      "step": 4483
    },
    {
      "epoch": 0.48459958932238195,
      "grad_norm": 0.40489253401756287,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 4484
    },
    {
      "epoch": 0.4847076623797687,
      "grad_norm": 0.45474231243133545,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 4485
    },
    {
      "epoch": 0.4848157354371555,
      "grad_norm": 0.4472688138484955,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 4486
    },
    {
      "epoch": 0.4849238084945423,
      "grad_norm": 0.449089914560318,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 4487
    },
    {
      "epoch": 0.4850318815519291,
      "grad_norm": 0.42619356513023376,
      "learning_rate": 8e-05,
      "loss": 1.5952,
      "step": 4488
    },
    {
      "epoch": 0.4851399546093159,
      "grad_norm": 0.4337788224220276,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 4489
    },
    {
      "epoch": 0.48524802766670266,
      "grad_norm": 0.5049014091491699,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 4490
    },
    {
      "epoch": 0.48535610072408947,
      "grad_norm": 0.40378502011299133,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 4491
    },
    {
      "epoch": 0.48546417378147627,
      "grad_norm": 0.42512020468711853,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 4492
    },
    {
      "epoch": 0.4855722468388631,
      "grad_norm": 0.45626673102378845,
      "learning_rate": 8e-05,
      "loss": 1.7993,
      "step": 4493
    },
    {
      "epoch": 0.4856803198962499,
      "grad_norm": 0.43610769510269165,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 4494
    },
    {
      "epoch": 0.4857883929536367,
      "grad_norm": 0.4629031717777252,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 4495
    },
    {
      "epoch": 0.48589646601102343,
      "grad_norm": 0.5478720664978027,
      "learning_rate": 8e-05,
      "loss": 2.0373,
      "step": 4496
    },
    {
      "epoch": 0.48600453906841024,
      "grad_norm": 0.45136258006095886,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 4497
    },
    {
      "epoch": 0.48611261212579704,
      "grad_norm": 0.42797863483428955,
      "learning_rate": 8e-05,
      "loss": 1.5433,
      "step": 4498
    },
    {
      "epoch": 0.48622068518318384,
      "grad_norm": 0.4159573018550873,
      "learning_rate": 8e-05,
      "loss": 1.594,
      "step": 4499
    },
    {
      "epoch": 0.48632875824057065,
      "grad_norm": 0.4392431080341339,
      "learning_rate": 8e-05,
      "loss": 1.6396,
      "step": 4500
    },
    {
      "epoch": 0.4864368312979574,
      "grad_norm": 0.4375099539756775,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 4501
    },
    {
      "epoch": 0.4865449043553442,
      "grad_norm": 0.4600321650505066,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 4502
    },
    {
      "epoch": 0.486652977412731,
      "grad_norm": 0.4499640166759491,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 4503
    },
    {
      "epoch": 0.4867610504701178,
      "grad_norm": 0.4487142860889435,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 4504
    },
    {
      "epoch": 0.4868691235275046,
      "grad_norm": 0.45476698875427246,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 4505
    },
    {
      "epoch": 0.4869771965848914,
      "grad_norm": 0.44930553436279297,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 4506
    },
    {
      "epoch": 0.48708526964227816,
      "grad_norm": 0.4324619174003601,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 4507
    },
    {
      "epoch": 0.48719334269966497,
      "grad_norm": 0.4158931374549866,
      "learning_rate": 8e-05,
      "loss": 1.6316,
      "step": 4508
    },
    {
      "epoch": 0.48730141575705177,
      "grad_norm": 0.4308388829231262,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 4509
    },
    {
      "epoch": 0.4874094888144386,
      "grad_norm": 0.41997256875038147,
      "learning_rate": 8e-05,
      "loss": 1.5803,
      "step": 4510
    },
    {
      "epoch": 0.4875175618718254,
      "grad_norm": 0.44430962204933167,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 4511
    },
    {
      "epoch": 0.4876256349292121,
      "grad_norm": 0.41465994715690613,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 4512
    },
    {
      "epoch": 0.48773370798659893,
      "grad_norm": 0.4189044237136841,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 4513
    },
    {
      "epoch": 0.48784178104398573,
      "grad_norm": 0.4861762821674347,
      "learning_rate": 8e-05,
      "loss": 1.6301,
      "step": 4514
    },
    {
      "epoch": 0.48794985410137254,
      "grad_norm": 0.5393797159194946,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 4515
    },
    {
      "epoch": 0.48805792715875934,
      "grad_norm": 0.4770389497280121,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 4516
    },
    {
      "epoch": 0.4881660002161461,
      "grad_norm": 0.43225303292274475,
      "learning_rate": 8e-05,
      "loss": 1.6224,
      "step": 4517
    },
    {
      "epoch": 0.4882740732735329,
      "grad_norm": 0.4160992205142975,
      "learning_rate": 8e-05,
      "loss": 1.5356,
      "step": 4518
    },
    {
      "epoch": 0.4883821463309197,
      "grad_norm": 0.47697389125823975,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 4519
    },
    {
      "epoch": 0.4884902193883065,
      "grad_norm": 0.48435404896736145,
      "learning_rate": 8e-05,
      "loss": 1.8806,
      "step": 4520
    },
    {
      "epoch": 0.4885982924456933,
      "grad_norm": 0.42291489243507385,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 4521
    },
    {
      "epoch": 0.4887063655030801,
      "grad_norm": 0.4529842436313629,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 4522
    },
    {
      "epoch": 0.48881443856046686,
      "grad_norm": 0.4725779891014099,
      "learning_rate": 8e-05,
      "loss": 1.8292,
      "step": 4523
    },
    {
      "epoch": 0.48892251161785366,
      "grad_norm": 0.5167995095252991,
      "learning_rate": 8e-05,
      "loss": 1.8592,
      "step": 4524
    },
    {
      "epoch": 0.48903058467524047,
      "grad_norm": 0.44742295145988464,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 4525
    },
    {
      "epoch": 0.48913865773262727,
      "grad_norm": 0.4509267807006836,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 4526
    },
    {
      "epoch": 0.4892467307900141,
      "grad_norm": 0.49812835454940796,
      "learning_rate": 8e-05,
      "loss": 1.6471,
      "step": 4527
    },
    {
      "epoch": 0.4893548038474008,
      "grad_norm": 0.4709514379501343,
      "learning_rate": 8e-05,
      "loss": 1.4631,
      "step": 4528
    },
    {
      "epoch": 0.4894628769047876,
      "grad_norm": 0.44753924012184143,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 4529
    },
    {
      "epoch": 0.48957094996217443,
      "grad_norm": 0.45856133103370667,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 4530
    },
    {
      "epoch": 0.48967902301956123,
      "grad_norm": 0.4306870102882385,
      "learning_rate": 8e-05,
      "loss": 1.6288,
      "step": 4531
    },
    {
      "epoch": 0.48978709607694804,
      "grad_norm": 0.49145621061325073,
      "learning_rate": 8e-05,
      "loss": 1.8303,
      "step": 4532
    },
    {
      "epoch": 0.4898951691343348,
      "grad_norm": 0.4750341773033142,
      "learning_rate": 8e-05,
      "loss": 1.9609,
      "step": 4533
    },
    {
      "epoch": 0.4900032421917216,
      "grad_norm": 0.42234480381011963,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 4534
    },
    {
      "epoch": 0.4901113152491084,
      "grad_norm": 0.4818591773509979,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 4535
    },
    {
      "epoch": 0.4902193883064952,
      "grad_norm": 0.4404877722263336,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 4536
    },
    {
      "epoch": 0.490327461363882,
      "grad_norm": 0.4319358170032501,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 4537
    },
    {
      "epoch": 0.4904355344212688,
      "grad_norm": 0.4615265727043152,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 4538
    },
    {
      "epoch": 0.49054360747865555,
      "grad_norm": 0.4597000181674957,
      "learning_rate": 8e-05,
      "loss": 1.5547,
      "step": 4539
    },
    {
      "epoch": 0.49065168053604236,
      "grad_norm": 0.4872746765613556,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 4540
    },
    {
      "epoch": 0.49075975359342916,
      "grad_norm": 0.5038478970527649,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 4541
    },
    {
      "epoch": 0.49086782665081596,
      "grad_norm": 0.43366026878356934,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 4542
    },
    {
      "epoch": 0.49097589970820277,
      "grad_norm": 0.4377911388874054,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 4543
    },
    {
      "epoch": 0.4910839727655895,
      "grad_norm": 0.4432450830936432,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 4544
    },
    {
      "epoch": 0.4911920458229763,
      "grad_norm": 0.4857546389102936,
      "learning_rate": 8e-05,
      "loss": 1.8814,
      "step": 4545
    },
    {
      "epoch": 0.4913001188803631,
      "grad_norm": 0.4277060627937317,
      "learning_rate": 8e-05,
      "loss": 1.7089,
      "step": 4546
    },
    {
      "epoch": 0.4914081919377499,
      "grad_norm": 0.4704058766365051,
      "learning_rate": 8e-05,
      "loss": 1.695,
      "step": 4547
    },
    {
      "epoch": 0.49151626499513673,
      "grad_norm": 0.4580753445625305,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 4548
    },
    {
      "epoch": 0.4916243380525235,
      "grad_norm": 0.41151848435401917,
      "learning_rate": 8e-05,
      "loss": 1.6051,
      "step": 4549
    },
    {
      "epoch": 0.4917324111099103,
      "grad_norm": 0.4697320759296417,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 4550
    },
    {
      "epoch": 0.4918404841672971,
      "grad_norm": 0.4736131429672241,
      "learning_rate": 8e-05,
      "loss": 1.8517,
      "step": 4551
    },
    {
      "epoch": 0.4919485572246839,
      "grad_norm": 0.4511648118495941,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 4552
    },
    {
      "epoch": 0.4920566302820707,
      "grad_norm": 0.42551884055137634,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 4553
    },
    {
      "epoch": 0.4921647033394575,
      "grad_norm": 0.4448295533657074,
      "learning_rate": 8e-05,
      "loss": 1.5363,
      "step": 4554
    },
    {
      "epoch": 0.49227277639684425,
      "grad_norm": 0.49205800890922546,
      "learning_rate": 8e-05,
      "loss": 1.8382,
      "step": 4555
    },
    {
      "epoch": 0.49238084945423105,
      "grad_norm": 0.4500883221626282,
      "learning_rate": 8e-05,
      "loss": 1.6057,
      "step": 4556
    },
    {
      "epoch": 0.49248892251161785,
      "grad_norm": 0.417587012052536,
      "learning_rate": 8e-05,
      "loss": 1.6514,
      "step": 4557
    },
    {
      "epoch": 0.49259699556900466,
      "grad_norm": 0.45526498556137085,
      "learning_rate": 8e-05,
      "loss": 1.6882,
      "step": 4558
    },
    {
      "epoch": 0.49270506862639146,
      "grad_norm": 0.43814635276794434,
      "learning_rate": 8e-05,
      "loss": 1.4582,
      "step": 4559
    },
    {
      "epoch": 0.4928131416837782,
      "grad_norm": 0.43992918729782104,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 4560
    },
    {
      "epoch": 0.492921214741165,
      "grad_norm": 0.4380009174346924,
      "learning_rate": 8e-05,
      "loss": 1.768,
      "step": 4561
    },
    {
      "epoch": 0.4930292877985518,
      "grad_norm": 0.4475932717323303,
      "learning_rate": 8e-05,
      "loss": 1.7451,
      "step": 4562
    },
    {
      "epoch": 0.4931373608559386,
      "grad_norm": 0.44631823897361755,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 4563
    },
    {
      "epoch": 0.4932454339133254,
      "grad_norm": 0.45139577984809875,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 4564
    },
    {
      "epoch": 0.4933535069707122,
      "grad_norm": 0.45419371128082275,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 4565
    },
    {
      "epoch": 0.493461580028099,
      "grad_norm": 0.454950213432312,
      "learning_rate": 8e-05,
      "loss": 1.8395,
      "step": 4566
    },
    {
      "epoch": 0.4935696530854858,
      "grad_norm": 0.4649523198604584,
      "learning_rate": 8e-05,
      "loss": 1.8511,
      "step": 4567
    },
    {
      "epoch": 0.4936777261428726,
      "grad_norm": 0.4300883710384369,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 4568
    },
    {
      "epoch": 0.4937857992002594,
      "grad_norm": 0.46401679515838623,
      "learning_rate": 8e-05,
      "loss": 1.865,
      "step": 4569
    },
    {
      "epoch": 0.4938938722576462,
      "grad_norm": 0.44116657972335815,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 4570
    },
    {
      "epoch": 0.49400194531503294,
      "grad_norm": 0.4432836174964905,
      "learning_rate": 8e-05,
      "loss": 1.6967,
      "step": 4571
    },
    {
      "epoch": 0.49411001837241975,
      "grad_norm": 0.4341505169868469,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 4572
    },
    {
      "epoch": 0.49421809142980655,
      "grad_norm": 0.44092002511024475,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 4573
    },
    {
      "epoch": 0.49432616448719335,
      "grad_norm": 0.5332693457603455,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 4574
    },
    {
      "epoch": 0.49443423754458016,
      "grad_norm": 0.4682213068008423,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 4575
    },
    {
      "epoch": 0.4945423106019669,
      "grad_norm": 0.42639344930648804,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 4576
    },
    {
      "epoch": 0.4946503836593537,
      "grad_norm": 0.43404456973075867,
      "learning_rate": 8e-05,
      "loss": 1.6485,
      "step": 4577
    },
    {
      "epoch": 0.4947584567167405,
      "grad_norm": 0.44460850954055786,
      "learning_rate": 8e-05,
      "loss": 1.7709,
      "step": 4578
    },
    {
      "epoch": 0.4948665297741273,
      "grad_norm": 0.4197717010974884,
      "learning_rate": 8e-05,
      "loss": 1.681,
      "step": 4579
    },
    {
      "epoch": 0.4949746028315141,
      "grad_norm": 0.47279971837997437,
      "learning_rate": 8e-05,
      "loss": 1.9966,
      "step": 4580
    },
    {
      "epoch": 0.49508267588890087,
      "grad_norm": 0.48262476921081543,
      "learning_rate": 8e-05,
      "loss": 1.843,
      "step": 4581
    },
    {
      "epoch": 0.4951907489462877,
      "grad_norm": 0.4513663649559021,
      "learning_rate": 8e-05,
      "loss": 1.7431,
      "step": 4582
    },
    {
      "epoch": 0.4952988220036745,
      "grad_norm": 0.4799826145172119,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 4583
    },
    {
      "epoch": 0.4954068950610613,
      "grad_norm": 0.44708481431007385,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 4584
    },
    {
      "epoch": 0.4955149681184481,
      "grad_norm": 0.4802453815937042,
      "learning_rate": 8e-05,
      "loss": 1.8467,
      "step": 4585
    },
    {
      "epoch": 0.4956230411758349,
      "grad_norm": 0.4554511606693268,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 4586
    },
    {
      "epoch": 0.49573111423322164,
      "grad_norm": 0.4931447505950928,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 4587
    },
    {
      "epoch": 0.49583918729060844,
      "grad_norm": 0.46515023708343506,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 4588
    },
    {
      "epoch": 0.49594726034799524,
      "grad_norm": 0.48708656430244446,
      "learning_rate": 8e-05,
      "loss": 1.8517,
      "step": 4589
    },
    {
      "epoch": 0.49605533340538205,
      "grad_norm": 0.4388846755027771,
      "learning_rate": 8e-05,
      "loss": 1.6072,
      "step": 4590
    },
    {
      "epoch": 0.49616340646276885,
      "grad_norm": 0.502586841583252,
      "learning_rate": 8e-05,
      "loss": 1.8727,
      "step": 4591
    },
    {
      "epoch": 0.4962714795201556,
      "grad_norm": 0.4653155207633972,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 4592
    },
    {
      "epoch": 0.4963795525775424,
      "grad_norm": 0.4580000638961792,
      "learning_rate": 8e-05,
      "loss": 1.5687,
      "step": 4593
    },
    {
      "epoch": 0.4964876256349292,
      "grad_norm": 0.4914253354072571,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 4594
    },
    {
      "epoch": 0.496595698692316,
      "grad_norm": 0.49912506341934204,
      "learning_rate": 8e-05,
      "loss": 1.7449,
      "step": 4595
    },
    {
      "epoch": 0.4967037717497028,
      "grad_norm": 0.4680221974849701,
      "learning_rate": 8e-05,
      "loss": 1.712,
      "step": 4596
    },
    {
      "epoch": 0.4968118448070896,
      "grad_norm": 0.5013942718505859,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 4597
    },
    {
      "epoch": 0.49691991786447637,
      "grad_norm": 0.4714498519897461,
      "learning_rate": 8e-05,
      "loss": 1.7543,
      "step": 4598
    },
    {
      "epoch": 0.49702799092186317,
      "grad_norm": 0.4427943527698517,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 4599
    },
    {
      "epoch": 0.49713606397925,
      "grad_norm": 0.4628116488456726,
      "learning_rate": 8e-05,
      "loss": 1.7452,
      "step": 4600
    },
    {
      "epoch": 0.4972441370366368,
      "grad_norm": 0.45211493968963623,
      "learning_rate": 8e-05,
      "loss": 1.6349,
      "step": 4601
    },
    {
      "epoch": 0.4973522100940236,
      "grad_norm": 0.44300568103790283,
      "learning_rate": 8e-05,
      "loss": 1.5249,
      "step": 4602
    },
    {
      "epoch": 0.49746028315141033,
      "grad_norm": 0.509817361831665,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 4603
    },
    {
      "epoch": 0.49756835620879714,
      "grad_norm": 0.45611703395843506,
      "learning_rate": 8e-05,
      "loss": 1.7348,
      "step": 4604
    },
    {
      "epoch": 0.49767642926618394,
      "grad_norm": 0.4715808629989624,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 4605
    },
    {
      "epoch": 0.49778450232357074,
      "grad_norm": 0.524392306804657,
      "learning_rate": 8e-05,
      "loss": 1.8938,
      "step": 4606
    },
    {
      "epoch": 0.49789257538095755,
      "grad_norm": 0.44749632477760315,
      "learning_rate": 8e-05,
      "loss": 1.8126,
      "step": 4607
    },
    {
      "epoch": 0.4980006484383443,
      "grad_norm": 0.4259788393974304,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 4608
    },
    {
      "epoch": 0.4981087214957311,
      "grad_norm": 0.5179271101951599,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 4609
    },
    {
      "epoch": 0.4982167945531179,
      "grad_norm": 0.4774622619152069,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 4610
    },
    {
      "epoch": 0.4983248676105047,
      "grad_norm": 0.44147804379463196,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 4611
    },
    {
      "epoch": 0.4984329406678915,
      "grad_norm": 0.45315268635749817,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 4612
    },
    {
      "epoch": 0.4985410137252783,
      "grad_norm": 0.5256526470184326,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 4613
    },
    {
      "epoch": 0.49864908678266506,
      "grad_norm": 0.4579021632671356,
      "learning_rate": 8e-05,
      "loss": 1.6354,
      "step": 4614
    },
    {
      "epoch": 0.49875715984005187,
      "grad_norm": 0.48104017972946167,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 4615
    },
    {
      "epoch": 0.49886523289743867,
      "grad_norm": 0.45232218503952026,
      "learning_rate": 8e-05,
      "loss": 1.7725,
      "step": 4616
    },
    {
      "epoch": 0.4989733059548255,
      "grad_norm": 0.44204428791999817,
      "learning_rate": 8e-05,
      "loss": 1.9113,
      "step": 4617
    },
    {
      "epoch": 0.4990813790122123,
      "grad_norm": 0.42797884345054626,
      "learning_rate": 8e-05,
      "loss": 1.6112,
      "step": 4618
    },
    {
      "epoch": 0.499189452069599,
      "grad_norm": 0.4407442510128021,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 4619
    },
    {
      "epoch": 0.49929752512698583,
      "grad_norm": 0.46989181637763977,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 4620
    },
    {
      "epoch": 0.49940559818437263,
      "grad_norm": 0.4359860122203827,
      "learning_rate": 8e-05,
      "loss": 1.719,
      "step": 4621
    },
    {
      "epoch": 0.49951367124175944,
      "grad_norm": 0.5156635642051697,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 4622
    },
    {
      "epoch": 0.49962174429914624,
      "grad_norm": 0.46487459540367126,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 4623
    },
    {
      "epoch": 0.499729817356533,
      "grad_norm": 0.5006021857261658,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 4624
    },
    {
      "epoch": 0.4998378904139198,
      "grad_norm": 0.49178966879844666,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 4625
    },
    {
      "epoch": 0.4999459634713066,
      "grad_norm": 0.4357178211212158,
      "learning_rate": 8e-05,
      "loss": 1.8588,
      "step": 4626
    },
    {
      "epoch": 0.5000540365286934,
      "grad_norm": 0.43304675817489624,
      "learning_rate": 8e-05,
      "loss": 1.5837,
      "step": 4627
    },
    {
      "epoch": 0.5001621095860802,
      "grad_norm": 0.45611679553985596,
      "learning_rate": 8e-05,
      "loss": 1.6285,
      "step": 4628
    },
    {
      "epoch": 0.500270182643467,
      "grad_norm": 0.4722634553909302,
      "learning_rate": 8e-05,
      "loss": 1.8343,
      "step": 4629
    },
    {
      "epoch": 0.5003782557008538,
      "grad_norm": 0.4585272967815399,
      "learning_rate": 8e-05,
      "loss": 1.7185,
      "step": 4630
    },
    {
      "epoch": 0.5004863287582406,
      "grad_norm": 0.4825398325920105,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 4631
    },
    {
      "epoch": 0.5005944018156273,
      "grad_norm": 0.49977102875709534,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 4632
    },
    {
      "epoch": 0.5007024748730141,
      "grad_norm": 0.5336390137672424,
      "learning_rate": 8e-05,
      "loss": 1.9985,
      "step": 4633
    },
    {
      "epoch": 0.5008105479304009,
      "grad_norm": 0.49954524636268616,
      "learning_rate": 8e-05,
      "loss": 1.8599,
      "step": 4634
    },
    {
      "epoch": 0.5009186209877877,
      "grad_norm": 0.4020881652832031,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 4635
    },
    {
      "epoch": 0.5010266940451745,
      "grad_norm": 0.43205851316452026,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 4636
    },
    {
      "epoch": 0.5011347671025613,
      "grad_norm": 0.5870148539543152,
      "learning_rate": 8e-05,
      "loss": 1.8955,
      "step": 4637
    },
    {
      "epoch": 0.5012428401599481,
      "grad_norm": 0.44627925753593445,
      "learning_rate": 8e-05,
      "loss": 1.6924,
      "step": 4638
    },
    {
      "epoch": 0.5013509132173349,
      "grad_norm": 0.4929617643356323,
      "learning_rate": 8e-05,
      "loss": 1.8718,
      "step": 4639
    },
    {
      "epoch": 0.5014589862747217,
      "grad_norm": 0.4403620958328247,
      "learning_rate": 8e-05,
      "loss": 1.8481,
      "step": 4640
    },
    {
      "epoch": 0.5015670593321085,
      "grad_norm": 0.506755530834198,
      "learning_rate": 8e-05,
      "loss": 1.9544,
      "step": 4641
    },
    {
      "epoch": 0.5016751323894953,
      "grad_norm": 0.43373963236808777,
      "learning_rate": 8e-05,
      "loss": 1.6221,
      "step": 4642
    },
    {
      "epoch": 0.501783205446882,
      "grad_norm": 0.4588828384876251,
      "learning_rate": 8e-05,
      "loss": 1.6699,
      "step": 4643
    },
    {
      "epoch": 0.5018912785042688,
      "grad_norm": 0.43538978695869446,
      "learning_rate": 8e-05,
      "loss": 1.5902,
      "step": 4644
    },
    {
      "epoch": 0.5019993515616556,
      "grad_norm": 0.41972601413726807,
      "learning_rate": 8e-05,
      "loss": 1.6668,
      "step": 4645
    },
    {
      "epoch": 0.5021074246190425,
      "grad_norm": 0.42788490653038025,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 4646
    },
    {
      "epoch": 0.5022154976764293,
      "grad_norm": 0.42410722374916077,
      "learning_rate": 8e-05,
      "loss": 1.5939,
      "step": 4647
    },
    {
      "epoch": 0.5023235707338161,
      "grad_norm": 0.48062261939048767,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 4648
    },
    {
      "epoch": 0.5024316437912029,
      "grad_norm": 0.455432265996933,
      "learning_rate": 8e-05,
      "loss": 1.8367,
      "step": 4649
    },
    {
      "epoch": 0.5025397168485897,
      "grad_norm": 0.3904498815536499,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 4650
    },
    {
      "epoch": 0.5026477899059765,
      "grad_norm": 0.4600466787815094,
      "learning_rate": 8e-05,
      "loss": 1.893,
      "step": 4651
    },
    {
      "epoch": 0.5027558629633633,
      "grad_norm": 0.4255983531475067,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 4652
    },
    {
      "epoch": 0.5028639360207501,
      "grad_norm": 0.47301390767097473,
      "learning_rate": 8e-05,
      "loss": 1.8465,
      "step": 4653
    },
    {
      "epoch": 0.5029720090781368,
      "grad_norm": 0.42934802174568176,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 4654
    },
    {
      "epoch": 0.5030800821355236,
      "grad_norm": 0.4951838254928589,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 4655
    },
    {
      "epoch": 0.5031881551929104,
      "grad_norm": 0.4629475474357605,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 4656
    },
    {
      "epoch": 0.5032962282502972,
      "grad_norm": 0.41975584626197815,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 4657
    },
    {
      "epoch": 0.503404301307684,
      "grad_norm": 0.43556612730026245,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 4658
    },
    {
      "epoch": 0.5035123743650708,
      "grad_norm": 0.40583810210227966,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 4659
    },
    {
      "epoch": 0.5036204474224576,
      "grad_norm": 0.43540892004966736,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 4660
    },
    {
      "epoch": 0.5037285204798444,
      "grad_norm": 0.45372042059898376,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 4661
    },
    {
      "epoch": 0.5038365935372312,
      "grad_norm": 0.44067713618278503,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 4662
    },
    {
      "epoch": 0.503944666594618,
      "grad_norm": 0.43811023235321045,
      "learning_rate": 8e-05,
      "loss": 1.6987,
      "step": 4663
    },
    {
      "epoch": 0.5040527396520047,
      "grad_norm": 0.44988855719566345,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 4664
    },
    {
      "epoch": 0.5041608127093915,
      "grad_norm": 0.42364558577537537,
      "learning_rate": 8e-05,
      "loss": 1.635,
      "step": 4665
    },
    {
      "epoch": 0.5042688857667783,
      "grad_norm": 0.4460824728012085,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 4666
    },
    {
      "epoch": 0.5043769588241651,
      "grad_norm": 0.4717217683792114,
      "learning_rate": 8e-05,
      "loss": 1.6054,
      "step": 4667
    },
    {
      "epoch": 0.5044850318815519,
      "grad_norm": 0.45729678869247437,
      "learning_rate": 8e-05,
      "loss": 1.8189,
      "step": 4668
    },
    {
      "epoch": 0.5045931049389387,
      "grad_norm": 0.46349775791168213,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 4669
    },
    {
      "epoch": 0.5047011779963255,
      "grad_norm": 0.42460963129997253,
      "learning_rate": 8e-05,
      "loss": 1.6582,
      "step": 4670
    },
    {
      "epoch": 0.5048092510537123,
      "grad_norm": 0.432793527841568,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 4671
    },
    {
      "epoch": 0.5049173241110991,
      "grad_norm": 0.42932596802711487,
      "learning_rate": 8e-05,
      "loss": 1.6325,
      "step": 4672
    },
    {
      "epoch": 0.5050253971684859,
      "grad_norm": 0.44093799591064453,
      "learning_rate": 8e-05,
      "loss": 1.6235,
      "step": 4673
    },
    {
      "epoch": 0.5051334702258727,
      "grad_norm": 0.42190471291542053,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 4674
    },
    {
      "epoch": 0.5052415432832594,
      "grad_norm": 0.4190448820590973,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 4675
    },
    {
      "epoch": 0.5053496163406462,
      "grad_norm": 0.4684920907020569,
      "learning_rate": 8e-05,
      "loss": 1.7783,
      "step": 4676
    },
    {
      "epoch": 0.505457689398033,
      "grad_norm": 0.44839295744895935,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 4677
    },
    {
      "epoch": 0.5055657624554198,
      "grad_norm": 0.46807414293289185,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 4678
    },
    {
      "epoch": 0.5056738355128066,
      "grad_norm": 0.43201470375061035,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 4679
    },
    {
      "epoch": 0.5057819085701935,
      "grad_norm": 0.4609532356262207,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 4680
    },
    {
      "epoch": 0.5058899816275803,
      "grad_norm": 0.41021278500556946,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 4681
    },
    {
      "epoch": 0.5059980546849671,
      "grad_norm": 0.4348876178264618,
      "learning_rate": 8e-05,
      "loss": 1.711,
      "step": 4682
    },
    {
      "epoch": 0.5061061277423539,
      "grad_norm": 0.4698355793952942,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 4683
    },
    {
      "epoch": 0.5062142007997407,
      "grad_norm": 0.4202800691127777,
      "learning_rate": 8e-05,
      "loss": 1.7985,
      "step": 4684
    },
    {
      "epoch": 0.5063222738571275,
      "grad_norm": 0.4434875547885895,
      "learning_rate": 8e-05,
      "loss": 1.791,
      "step": 4685
    },
    {
      "epoch": 0.5064303469145142,
      "grad_norm": 0.4359697997570038,
      "learning_rate": 8e-05,
      "loss": 1.7567,
      "step": 4686
    },
    {
      "epoch": 0.506538419971901,
      "grad_norm": 0.5532411336898804,
      "learning_rate": 8e-05,
      "loss": 1.8189,
      "step": 4687
    },
    {
      "epoch": 0.5066464930292878,
      "grad_norm": 0.4119906723499298,
      "learning_rate": 8e-05,
      "loss": 1.5573,
      "step": 4688
    },
    {
      "epoch": 0.5067545660866746,
      "grad_norm": 0.5013248920440674,
      "learning_rate": 8e-05,
      "loss": 1.661,
      "step": 4689
    },
    {
      "epoch": 0.5068626391440614,
      "grad_norm": 0.461455374956131,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 4690
    },
    {
      "epoch": 0.5069707122014482,
      "grad_norm": 0.4014069736003876,
      "learning_rate": 8e-05,
      "loss": 1.5187,
      "step": 4691
    },
    {
      "epoch": 0.507078785258835,
      "grad_norm": 0.47217753529548645,
      "learning_rate": 8e-05,
      "loss": 1.8277,
      "step": 4692
    },
    {
      "epoch": 0.5071868583162218,
      "grad_norm": 0.43741557002067566,
      "learning_rate": 8e-05,
      "loss": 1.7754,
      "step": 4693
    },
    {
      "epoch": 0.5072949313736086,
      "grad_norm": 0.43695297837257385,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 4694
    },
    {
      "epoch": 0.5074030044309954,
      "grad_norm": 0.40093517303466797,
      "learning_rate": 8e-05,
      "loss": 1.4183,
      "step": 4695
    },
    {
      "epoch": 0.5075110774883822,
      "grad_norm": 0.4221629798412323,
      "learning_rate": 8e-05,
      "loss": 1.7015,
      "step": 4696
    },
    {
      "epoch": 0.5076191505457689,
      "grad_norm": 0.46503254771232605,
      "learning_rate": 8e-05,
      "loss": 1.7626,
      "step": 4697
    },
    {
      "epoch": 0.5077272236031557,
      "grad_norm": 0.4630764126777649,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 4698
    },
    {
      "epoch": 0.5078352966605425,
      "grad_norm": 0.47782421112060547,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 4699
    },
    {
      "epoch": 0.5079433697179293,
      "grad_norm": 0.44084304571151733,
      "learning_rate": 8e-05,
      "loss": 1.672,
      "step": 4700
    },
    {
      "epoch": 0.5080514427753161,
      "grad_norm": 0.410439133644104,
      "learning_rate": 8e-05,
      "loss": 1.4789,
      "step": 4701
    },
    {
      "epoch": 0.5081595158327029,
      "grad_norm": 0.49456390738487244,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 4702
    },
    {
      "epoch": 0.5082675888900897,
      "grad_norm": 0.4171033501625061,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 4703
    },
    {
      "epoch": 0.5083756619474765,
      "grad_norm": 0.46798089146614075,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 4704
    },
    {
      "epoch": 0.5084837350048633,
      "grad_norm": 0.44995465874671936,
      "learning_rate": 8e-05,
      "loss": 1.9601,
      "step": 4705
    },
    {
      "epoch": 0.5085918080622501,
      "grad_norm": 0.4399876594543457,
      "learning_rate": 8e-05,
      "loss": 1.7834,
      "step": 4706
    },
    {
      "epoch": 0.5086998811196368,
      "grad_norm": 0.46327292919158936,
      "learning_rate": 8e-05,
      "loss": 1.9034,
      "step": 4707
    },
    {
      "epoch": 0.5088079541770236,
      "grad_norm": 0.41943588852882385,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 4708
    },
    {
      "epoch": 0.5089160272344104,
      "grad_norm": 0.42165428400039673,
      "learning_rate": 8e-05,
      "loss": 1.6405,
      "step": 4709
    },
    {
      "epoch": 0.5090241002917972,
      "grad_norm": 0.4739735722541809,
      "learning_rate": 8e-05,
      "loss": 1.9183,
      "step": 4710
    },
    {
      "epoch": 0.509132173349184,
      "grad_norm": 0.41496315598487854,
      "learning_rate": 8e-05,
      "loss": 1.6242,
      "step": 4711
    },
    {
      "epoch": 0.5092402464065708,
      "grad_norm": 0.4529750645160675,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 4712
    },
    {
      "epoch": 0.5093483194639576,
      "grad_norm": 0.428808331489563,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 4713
    },
    {
      "epoch": 0.5094563925213444,
      "grad_norm": 0.47122007608413696,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 4714
    },
    {
      "epoch": 0.5095644655787313,
      "grad_norm": 0.4569855332374573,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 4715
    },
    {
      "epoch": 0.509672538636118,
      "grad_norm": 0.42573824524879456,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 4716
    },
    {
      "epoch": 0.5097806116935049,
      "grad_norm": 0.428300678730011,
      "learning_rate": 8e-05,
      "loss": 1.6571,
      "step": 4717
    },
    {
      "epoch": 0.5098886847508916,
      "grad_norm": 0.45448943972587585,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 4718
    },
    {
      "epoch": 0.5099967578082784,
      "grad_norm": 0.5374927520751953,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 4719
    },
    {
      "epoch": 0.5101048308656652,
      "grad_norm": 0.4421241879463196,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 4720
    },
    {
      "epoch": 0.510212903923052,
      "grad_norm": 0.4842216670513153,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 4721
    },
    {
      "epoch": 0.5103209769804388,
      "grad_norm": 0.4844920039176941,
      "learning_rate": 8e-05,
      "loss": 1.8493,
      "step": 4722
    },
    {
      "epoch": 0.5104290500378256,
      "grad_norm": 0.4363443851470947,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 4723
    },
    {
      "epoch": 0.5105371230952124,
      "grad_norm": 0.43398985266685486,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 4724
    },
    {
      "epoch": 0.5106451961525992,
      "grad_norm": 0.4409353733062744,
      "learning_rate": 8e-05,
      "loss": 1.6244,
      "step": 4725
    },
    {
      "epoch": 0.510753269209986,
      "grad_norm": 0.4134727120399475,
      "learning_rate": 8e-05,
      "loss": 1.4711,
      "step": 4726
    },
    {
      "epoch": 0.5108613422673728,
      "grad_norm": 0.4271588623523712,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 4727
    },
    {
      "epoch": 0.5109694153247596,
      "grad_norm": 0.4987278878688812,
      "learning_rate": 8e-05,
      "loss": 1.9603,
      "step": 4728
    },
    {
      "epoch": 0.5110774883821463,
      "grad_norm": 0.500177264213562,
      "learning_rate": 8e-05,
      "loss": 1.8834,
      "step": 4729
    },
    {
      "epoch": 0.5111855614395331,
      "grad_norm": 0.4465075135231018,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 4730
    },
    {
      "epoch": 0.5112936344969199,
      "grad_norm": 0.41860055923461914,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 4731
    },
    {
      "epoch": 0.5114017075543067,
      "grad_norm": 0.4418352246284485,
      "learning_rate": 8e-05,
      "loss": 1.7029,
      "step": 4732
    },
    {
      "epoch": 0.5115097806116935,
      "grad_norm": 0.42539575695991516,
      "learning_rate": 8e-05,
      "loss": 1.7555,
      "step": 4733
    },
    {
      "epoch": 0.5116178536690803,
      "grad_norm": 0.49134010076522827,
      "learning_rate": 8e-05,
      "loss": 1.8836,
      "step": 4734
    },
    {
      "epoch": 0.5117259267264671,
      "grad_norm": 0.4557126462459564,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 4735
    },
    {
      "epoch": 0.5118339997838539,
      "grad_norm": 0.4767579138278961,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 4736
    },
    {
      "epoch": 0.5119420728412407,
      "grad_norm": 0.4333094358444214,
      "learning_rate": 8e-05,
      "loss": 1.8174,
      "step": 4737
    },
    {
      "epoch": 0.5120501458986275,
      "grad_norm": 0.440217524766922,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 4738
    },
    {
      "epoch": 0.5121582189560142,
      "grad_norm": 0.44460123777389526,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 4739
    },
    {
      "epoch": 0.512266292013401,
      "grad_norm": 0.42648983001708984,
      "learning_rate": 8e-05,
      "loss": 1.6148,
      "step": 4740
    },
    {
      "epoch": 0.5123743650707878,
      "grad_norm": 0.4787532091140747,
      "learning_rate": 8e-05,
      "loss": 1.9095,
      "step": 4741
    },
    {
      "epoch": 0.5124824381281746,
      "grad_norm": 0.4235503375530243,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 4742
    },
    {
      "epoch": 0.5125905111855614,
      "grad_norm": 0.4492688775062561,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 4743
    },
    {
      "epoch": 0.5126985842429482,
      "grad_norm": 0.4298434257507324,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 4744
    },
    {
      "epoch": 0.512806657300335,
      "grad_norm": 0.4284496605396271,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 4745
    },
    {
      "epoch": 0.5129147303577218,
      "grad_norm": 0.49575749039649963,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 4746
    },
    {
      "epoch": 0.5130228034151086,
      "grad_norm": 0.44586774706840515,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 4747
    },
    {
      "epoch": 0.5131308764724954,
      "grad_norm": 0.4768214523792267,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 4748
    },
    {
      "epoch": 0.5132389495298822,
      "grad_norm": 0.45820415019989014,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 4749
    },
    {
      "epoch": 0.5133470225872689,
      "grad_norm": 0.45349815487861633,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 4750
    },
    {
      "epoch": 0.5134550956446557,
      "grad_norm": 0.4418184161186218,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 4751
    },
    {
      "epoch": 0.5135631687020425,
      "grad_norm": 0.46581295132637024,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 4752
    },
    {
      "epoch": 0.5136712417594294,
      "grad_norm": 0.46772778034210205,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 4753
    },
    {
      "epoch": 0.5137793148168162,
      "grad_norm": 0.4433102607727051,
      "learning_rate": 8e-05,
      "loss": 1.8202,
      "step": 4754
    },
    {
      "epoch": 0.513887387874203,
      "grad_norm": 0.5231044888496399,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 4755
    },
    {
      "epoch": 0.5139954609315898,
      "grad_norm": 0.4367322027683258,
      "learning_rate": 8e-05,
      "loss": 1.8639,
      "step": 4756
    },
    {
      "epoch": 0.5141035339889766,
      "grad_norm": 0.44936978816986084,
      "learning_rate": 8e-05,
      "loss": 1.7649,
      "step": 4757
    },
    {
      "epoch": 0.5142116070463634,
      "grad_norm": 0.6672779321670532,
      "learning_rate": 8e-05,
      "loss": 1.9907,
      "step": 4758
    },
    {
      "epoch": 0.5143196801037502,
      "grad_norm": 0.4358031153678894,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 4759
    },
    {
      "epoch": 0.514427753161137,
      "grad_norm": 0.4918586015701294,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 4760
    },
    {
      "epoch": 0.5145358262185237,
      "grad_norm": 0.42208048701286316,
      "learning_rate": 8e-05,
      "loss": 1.6151,
      "step": 4761
    },
    {
      "epoch": 0.5146438992759105,
      "grad_norm": 0.5672999620437622,
      "learning_rate": 8e-05,
      "loss": 2.0211,
      "step": 4762
    },
    {
      "epoch": 0.5147519723332973,
      "grad_norm": 0.4733365774154663,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 4763
    },
    {
      "epoch": 0.5148600453906841,
      "grad_norm": 0.4224005937576294,
      "learning_rate": 8e-05,
      "loss": 1.5298,
      "step": 4764
    },
    {
      "epoch": 0.5149681184480709,
      "grad_norm": 0.4628714621067047,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 4765
    },
    {
      "epoch": 0.5150761915054577,
      "grad_norm": 0.43693897128105164,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 4766
    },
    {
      "epoch": 0.5151842645628445,
      "grad_norm": 0.4698878526687622,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 4767
    },
    {
      "epoch": 0.5152923376202313,
      "grad_norm": 0.5049228668212891,
      "learning_rate": 8e-05,
      "loss": 1.7669,
      "step": 4768
    },
    {
      "epoch": 0.5154004106776181,
      "grad_norm": 0.4312134385108948,
      "learning_rate": 8e-05,
      "loss": 1.5906,
      "step": 4769
    },
    {
      "epoch": 0.5155084837350049,
      "grad_norm": 0.48732686042785645,
      "learning_rate": 8e-05,
      "loss": 1.8278,
      "step": 4770
    },
    {
      "epoch": 0.5156165567923917,
      "grad_norm": 0.4104337692260742,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 4771
    },
    {
      "epoch": 0.5157246298497784,
      "grad_norm": 0.47884124517440796,
      "learning_rate": 8e-05,
      "loss": 1.9375,
      "step": 4772
    },
    {
      "epoch": 0.5158327029071652,
      "grad_norm": 0.48802244663238525,
      "learning_rate": 8e-05,
      "loss": 1.878,
      "step": 4773
    },
    {
      "epoch": 0.515940775964552,
      "grad_norm": 0.4558298885822296,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 4774
    },
    {
      "epoch": 0.5160488490219388,
      "grad_norm": 0.45967429876327515,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 4775
    },
    {
      "epoch": 0.5161569220793256,
      "grad_norm": 0.4327090084552765,
      "learning_rate": 8e-05,
      "loss": 1.4693,
      "step": 4776
    },
    {
      "epoch": 0.5162649951367124,
      "grad_norm": 0.456434428691864,
      "learning_rate": 8e-05,
      "loss": 1.5342,
      "step": 4777
    },
    {
      "epoch": 0.5163730681940992,
      "grad_norm": 0.5191709995269775,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 4778
    },
    {
      "epoch": 0.516481141251486,
      "grad_norm": 0.45064646005630493,
      "learning_rate": 8e-05,
      "loss": 1.6413,
      "step": 4779
    },
    {
      "epoch": 0.5165892143088728,
      "grad_norm": 0.4676867127418518,
      "learning_rate": 8e-05,
      "loss": 1.8903,
      "step": 4780
    },
    {
      "epoch": 0.5166972873662596,
      "grad_norm": 0.5040070414543152,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 4781
    },
    {
      "epoch": 0.5168053604236463,
      "grad_norm": 0.44950851798057556,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 4782
    },
    {
      "epoch": 0.5169134334810331,
      "grad_norm": 0.4593349099159241,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 4783
    },
    {
      "epoch": 0.5170215065384199,
      "grad_norm": 0.4725874364376068,
      "learning_rate": 8e-05,
      "loss": 1.7217,
      "step": 4784
    },
    {
      "epoch": 0.5171295795958067,
      "grad_norm": 0.4770016670227051,
      "learning_rate": 8e-05,
      "loss": 1.8625,
      "step": 4785
    },
    {
      "epoch": 0.5172376526531935,
      "grad_norm": 0.4317431151866913,
      "learning_rate": 8e-05,
      "loss": 1.6737,
      "step": 4786
    },
    {
      "epoch": 0.5173457257105804,
      "grad_norm": 0.44504350423812866,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 4787
    },
    {
      "epoch": 0.5174537987679672,
      "grad_norm": 0.4686642289161682,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 4788
    },
    {
      "epoch": 0.517561871825354,
      "grad_norm": 0.4189186096191406,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 4789
    },
    {
      "epoch": 0.5176699448827408,
      "grad_norm": 0.43180379271507263,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 4790
    },
    {
      "epoch": 0.5177780179401276,
      "grad_norm": 0.6494812965393066,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 4791
    },
    {
      "epoch": 0.5178860909975144,
      "grad_norm": 0.5174448490142822,
      "learning_rate": 8e-05,
      "loss": 1.8506,
      "step": 4792
    },
    {
      "epoch": 0.5179941640549011,
      "grad_norm": 0.44236263632774353,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 4793
    },
    {
      "epoch": 0.5181022371122879,
      "grad_norm": 0.4522027373313904,
      "learning_rate": 8e-05,
      "loss": 1.8683,
      "step": 4794
    },
    {
      "epoch": 0.5182103101696747,
      "grad_norm": 0.4161589741706848,
      "learning_rate": 8e-05,
      "loss": 1.5133,
      "step": 4795
    },
    {
      "epoch": 0.5183183832270615,
      "grad_norm": 0.4465634226799011,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 4796
    },
    {
      "epoch": 0.5184264562844483,
      "grad_norm": 0.43160906434059143,
      "learning_rate": 8e-05,
      "loss": 1.6718,
      "step": 4797
    },
    {
      "epoch": 0.5185345293418351,
      "grad_norm": 0.4345472753047943,
      "learning_rate": 8e-05,
      "loss": 1.5101,
      "step": 4798
    },
    {
      "epoch": 0.5186426023992219,
      "grad_norm": 0.40643438696861267,
      "learning_rate": 8e-05,
      "loss": 1.4451,
      "step": 4799
    },
    {
      "epoch": 0.5187506754566087,
      "grad_norm": 0.47910672426223755,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 4800
    },
    {
      "epoch": 0.5188587485139955,
      "grad_norm": 0.5270683169364929,
      "learning_rate": 8e-05,
      "loss": 1.8702,
      "step": 4801
    },
    {
      "epoch": 0.5189668215713823,
      "grad_norm": 0.457313597202301,
      "learning_rate": 8e-05,
      "loss": 1.8211,
      "step": 4802
    },
    {
      "epoch": 0.5190748946287691,
      "grad_norm": 0.4754776060581207,
      "learning_rate": 8e-05,
      "loss": 1.9417,
      "step": 4803
    },
    {
      "epoch": 0.5191829676861558,
      "grad_norm": 0.5159242153167725,
      "learning_rate": 8e-05,
      "loss": 1.9411,
      "step": 4804
    },
    {
      "epoch": 0.5192910407435426,
      "grad_norm": 0.4583616256713867,
      "learning_rate": 8e-05,
      "loss": 1.685,
      "step": 4805
    },
    {
      "epoch": 0.5193991138009294,
      "grad_norm": 0.45002350211143494,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 4806
    },
    {
      "epoch": 0.5195071868583162,
      "grad_norm": 0.4588592052459717,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 4807
    },
    {
      "epoch": 0.519615259915703,
      "grad_norm": 0.44024157524108887,
      "learning_rate": 8e-05,
      "loss": 1.6613,
      "step": 4808
    },
    {
      "epoch": 0.5197233329730898,
      "grad_norm": 0.43906262516975403,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 4809
    },
    {
      "epoch": 0.5198314060304766,
      "grad_norm": 0.42438656091690063,
      "learning_rate": 8e-05,
      "loss": 1.8394,
      "step": 4810
    },
    {
      "epoch": 0.5199394790878634,
      "grad_norm": 0.4453434944152832,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 4811
    },
    {
      "epoch": 0.5200475521452502,
      "grad_norm": 0.4359395205974579,
      "learning_rate": 8e-05,
      "loss": 1.7418,
      "step": 4812
    },
    {
      "epoch": 0.520155625202637,
      "grad_norm": 0.41998815536499023,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 4813
    },
    {
      "epoch": 0.5202636982600237,
      "grad_norm": 0.45825058221817017,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 4814
    },
    {
      "epoch": 0.5203717713174105,
      "grad_norm": 0.4161607325077057,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 4815
    },
    {
      "epoch": 0.5204798443747973,
      "grad_norm": 0.4485475718975067,
      "learning_rate": 8e-05,
      "loss": 1.6458,
      "step": 4816
    },
    {
      "epoch": 0.5205879174321841,
      "grad_norm": 0.48913416266441345,
      "learning_rate": 8e-05,
      "loss": 1.8097,
      "step": 4817
    },
    {
      "epoch": 0.5206959904895709,
      "grad_norm": 0.4763086140155792,
      "learning_rate": 8e-05,
      "loss": 1.8689,
      "step": 4818
    },
    {
      "epoch": 0.5208040635469577,
      "grad_norm": 0.43153125047683716,
      "learning_rate": 8e-05,
      "loss": 1.859,
      "step": 4819
    },
    {
      "epoch": 0.5209121366043445,
      "grad_norm": 0.47921857237815857,
      "learning_rate": 8e-05,
      "loss": 1.6273,
      "step": 4820
    },
    {
      "epoch": 0.5210202096617313,
      "grad_norm": 0.4601854979991913,
      "learning_rate": 8e-05,
      "loss": 1.9109,
      "step": 4821
    },
    {
      "epoch": 0.5211282827191182,
      "grad_norm": 0.4273405969142914,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 4822
    },
    {
      "epoch": 0.521236355776505,
      "grad_norm": 0.4028484523296356,
      "learning_rate": 8e-05,
      "loss": 1.6851,
      "step": 4823
    },
    {
      "epoch": 0.5213444288338918,
      "grad_norm": 0.4264885187149048,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 4824
    },
    {
      "epoch": 0.5214525018912785,
      "grad_norm": 0.4144957959651947,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 4825
    },
    {
      "epoch": 0.5215605749486653,
      "grad_norm": 0.45830488204956055,
      "learning_rate": 8e-05,
      "loss": 1.5693,
      "step": 4826
    },
    {
      "epoch": 0.5216686480060521,
      "grad_norm": 0.47415199875831604,
      "learning_rate": 8e-05,
      "loss": 1.7723,
      "step": 4827
    },
    {
      "epoch": 0.5217767210634389,
      "grad_norm": 0.41898590326309204,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 4828
    },
    {
      "epoch": 0.5218847941208257,
      "grad_norm": 0.4730352461338043,
      "learning_rate": 8e-05,
      "loss": 1.8159,
      "step": 4829
    },
    {
      "epoch": 0.5219928671782125,
      "grad_norm": 0.5015882253646851,
      "learning_rate": 8e-05,
      "loss": 1.7577,
      "step": 4830
    },
    {
      "epoch": 0.5221009402355993,
      "grad_norm": 0.5227234363555908,
      "learning_rate": 8e-05,
      "loss": 1.9411,
      "step": 4831
    },
    {
      "epoch": 0.5222090132929861,
      "grad_norm": 0.46370232105255127,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 4832
    },
    {
      "epoch": 0.5223170863503729,
      "grad_norm": 0.46318167448043823,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 4833
    },
    {
      "epoch": 0.5224251594077597,
      "grad_norm": 0.4610770046710968,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 4834
    },
    {
      "epoch": 0.5225332324651465,
      "grad_norm": 0.45913204550743103,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 4835
    },
    {
      "epoch": 0.5226413055225332,
      "grad_norm": 0.4615631699562073,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 4836
    },
    {
      "epoch": 0.52274937857992,
      "grad_norm": 0.47996392846107483,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 4837
    },
    {
      "epoch": 0.5228574516373068,
      "grad_norm": 0.48143890500068665,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 4838
    },
    {
      "epoch": 0.5229655246946936,
      "grad_norm": 0.45626822113990784,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 4839
    },
    {
      "epoch": 0.5230735977520804,
      "grad_norm": 0.49381619691848755,
      "learning_rate": 8e-05,
      "loss": 1.8261,
      "step": 4840
    },
    {
      "epoch": 0.5231816708094672,
      "grad_norm": 0.47788092494010925,
      "learning_rate": 8e-05,
      "loss": 1.9525,
      "step": 4841
    },
    {
      "epoch": 0.523289743866854,
      "grad_norm": 0.4548017978668213,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 4842
    },
    {
      "epoch": 0.5233978169242408,
      "grad_norm": 0.4408176839351654,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 4843
    },
    {
      "epoch": 0.5235058899816276,
      "grad_norm": 0.4262688457965851,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 4844
    },
    {
      "epoch": 0.5236139630390144,
      "grad_norm": 0.42433732748031616,
      "learning_rate": 8e-05,
      "loss": 1.7113,
      "step": 4845
    },
    {
      "epoch": 0.5237220360964012,
      "grad_norm": 0.4683915376663208,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 4846
    },
    {
      "epoch": 0.5238301091537879,
      "grad_norm": 0.4398551881313324,
      "learning_rate": 8e-05,
      "loss": 1.594,
      "step": 4847
    },
    {
      "epoch": 0.5239381822111747,
      "grad_norm": 0.4380076825618744,
      "learning_rate": 8e-05,
      "loss": 1.5257,
      "step": 4848
    },
    {
      "epoch": 0.5240462552685615,
      "grad_norm": 0.49280938506126404,
      "learning_rate": 8e-05,
      "loss": 1.8252,
      "step": 4849
    },
    {
      "epoch": 0.5241543283259483,
      "grad_norm": 0.47574540972709656,
      "learning_rate": 8e-05,
      "loss": 1.7717,
      "step": 4850
    },
    {
      "epoch": 0.5242624013833351,
      "grad_norm": 0.537904679775238,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 4851
    },
    {
      "epoch": 0.5243704744407219,
      "grad_norm": 0.4343433976173401,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 4852
    },
    {
      "epoch": 0.5244785474981087,
      "grad_norm": 0.4533993899822235,
      "learning_rate": 8e-05,
      "loss": 1.7079,
      "step": 4853
    },
    {
      "epoch": 0.5245866205554955,
      "grad_norm": 0.4799986779689789,
      "learning_rate": 8e-05,
      "loss": 1.8635,
      "step": 4854
    },
    {
      "epoch": 0.5246946936128823,
      "grad_norm": 0.43388497829437256,
      "learning_rate": 8e-05,
      "loss": 1.623,
      "step": 4855
    },
    {
      "epoch": 0.5248027666702691,
      "grad_norm": 0.4563114047050476,
      "learning_rate": 8e-05,
      "loss": 1.6568,
      "step": 4856
    },
    {
      "epoch": 0.5249108397276558,
      "grad_norm": 0.4301362931728363,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 4857
    },
    {
      "epoch": 0.5250189127850426,
      "grad_norm": 0.44407838582992554,
      "learning_rate": 8e-05,
      "loss": 1.6582,
      "step": 4858
    },
    {
      "epoch": 0.5251269858424295,
      "grad_norm": 0.4078293442726135,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 4859
    },
    {
      "epoch": 0.5252350588998163,
      "grad_norm": 0.4413585364818573,
      "learning_rate": 8e-05,
      "loss": 1.7193,
      "step": 4860
    },
    {
      "epoch": 0.5253431319572031,
      "grad_norm": 0.44573846459388733,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 4861
    },
    {
      "epoch": 0.5254512050145899,
      "grad_norm": 0.44897791743278503,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 4862
    },
    {
      "epoch": 0.5255592780719767,
      "grad_norm": 0.5216456651687622,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 4863
    },
    {
      "epoch": 0.5256673511293635,
      "grad_norm": 0.45774805545806885,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 4864
    },
    {
      "epoch": 0.5257754241867503,
      "grad_norm": 0.45692551136016846,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 4865
    },
    {
      "epoch": 0.5258834972441371,
      "grad_norm": 0.4466839134693146,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 4866
    },
    {
      "epoch": 0.5259915703015239,
      "grad_norm": 0.5071262121200562,
      "learning_rate": 8e-05,
      "loss": 1.8438,
      "step": 4867
    },
    {
      "epoch": 0.5260996433589106,
      "grad_norm": 0.48343726992607117,
      "learning_rate": 8e-05,
      "loss": 1.8333,
      "step": 4868
    },
    {
      "epoch": 0.5262077164162974,
      "grad_norm": 0.4438244700431824,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 4869
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.4258307218551636,
      "learning_rate": 8e-05,
      "loss": 1.6347,
      "step": 4870
    },
    {
      "epoch": 0.526423862531071,
      "grad_norm": 0.47714754939079285,
      "learning_rate": 8e-05,
      "loss": 1.6765,
      "step": 4871
    },
    {
      "epoch": 0.5265319355884578,
      "grad_norm": 0.444656640291214,
      "learning_rate": 8e-05,
      "loss": 1.8214,
      "step": 4872
    },
    {
      "epoch": 0.5266400086458446,
      "grad_norm": 0.45282256603240967,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 4873
    },
    {
      "epoch": 0.5267480817032314,
      "grad_norm": 0.4798215329647064,
      "learning_rate": 8e-05,
      "loss": 1.8265,
      "step": 4874
    },
    {
      "epoch": 0.5268561547606182,
      "grad_norm": 0.48568686842918396,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 4875
    },
    {
      "epoch": 0.526964227818005,
      "grad_norm": 0.4150574505329132,
      "learning_rate": 8e-05,
      "loss": 1.6017,
      "step": 4876
    },
    {
      "epoch": 0.5270723008753918,
      "grad_norm": 0.4443899691104889,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 4877
    },
    {
      "epoch": 0.5271803739327786,
      "grad_norm": 0.5089095234870911,
      "learning_rate": 8e-05,
      "loss": 1.9227,
      "step": 4878
    },
    {
      "epoch": 0.5272884469901653,
      "grad_norm": 0.4880915582180023,
      "learning_rate": 8e-05,
      "loss": 1.9601,
      "step": 4879
    },
    {
      "epoch": 0.5273965200475521,
      "grad_norm": 0.446435809135437,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 4880
    },
    {
      "epoch": 0.5275045931049389,
      "grad_norm": 0.4820060133934021,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 4881
    },
    {
      "epoch": 0.5276126661623257,
      "grad_norm": 0.46026739478111267,
      "learning_rate": 8e-05,
      "loss": 1.8138,
      "step": 4882
    },
    {
      "epoch": 0.5277207392197125,
      "grad_norm": 0.4313663840293884,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 4883
    },
    {
      "epoch": 0.5278288122770993,
      "grad_norm": 0.4414142370223999,
      "learning_rate": 8e-05,
      "loss": 1.6455,
      "step": 4884
    },
    {
      "epoch": 0.5279368853344861,
      "grad_norm": 0.4558389484882355,
      "learning_rate": 8e-05,
      "loss": 1.4971,
      "step": 4885
    },
    {
      "epoch": 0.5280449583918729,
      "grad_norm": 0.45691511034965515,
      "learning_rate": 8e-05,
      "loss": 1.7946,
      "step": 4886
    },
    {
      "epoch": 0.5281530314492597,
      "grad_norm": 0.44813430309295654,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 4887
    },
    {
      "epoch": 0.5282611045066465,
      "grad_norm": 0.44934988021850586,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 4888
    },
    {
      "epoch": 0.5283691775640332,
      "grad_norm": 0.46946069598197937,
      "learning_rate": 8e-05,
      "loss": 1.8329,
      "step": 4889
    },
    {
      "epoch": 0.52847725062142,
      "grad_norm": 0.5024906992912292,
      "learning_rate": 8e-05,
      "loss": 1.7662,
      "step": 4890
    },
    {
      "epoch": 0.5285853236788068,
      "grad_norm": 0.4468110501766205,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 4891
    },
    {
      "epoch": 0.5286933967361936,
      "grad_norm": 0.4688645303249359,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 4892
    },
    {
      "epoch": 0.5288014697935804,
      "grad_norm": 0.4622570276260376,
      "learning_rate": 8e-05,
      "loss": 1.5504,
      "step": 4893
    },
    {
      "epoch": 0.5289095428509673,
      "grad_norm": 0.43761733174324036,
      "learning_rate": 8e-05,
      "loss": 1.6423,
      "step": 4894
    },
    {
      "epoch": 0.529017615908354,
      "grad_norm": 0.4924977719783783,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 4895
    },
    {
      "epoch": 0.5291256889657409,
      "grad_norm": 0.4652452766895294,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 4896
    },
    {
      "epoch": 0.5292337620231277,
      "grad_norm": 0.4209824204444885,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 4897
    },
    {
      "epoch": 0.5293418350805145,
      "grad_norm": 0.425800085067749,
      "learning_rate": 8e-05,
      "loss": 1.6725,
      "step": 4898
    },
    {
      "epoch": 0.5294499081379013,
      "grad_norm": 0.43406739830970764,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 4899
    },
    {
      "epoch": 0.529557981195288,
      "grad_norm": 0.44544360041618347,
      "learning_rate": 8e-05,
      "loss": 1.7211,
      "step": 4900
    },
    {
      "epoch": 0.5296660542526748,
      "grad_norm": 0.5272177457809448,
      "learning_rate": 8e-05,
      "loss": 1.8831,
      "step": 4901
    },
    {
      "epoch": 0.5297741273100616,
      "grad_norm": 0.4438045024871826,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 4902
    },
    {
      "epoch": 0.5298822003674484,
      "grad_norm": 0.43426713347435,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 4903
    },
    {
      "epoch": 0.5299902734248352,
      "grad_norm": 0.4566912353038788,
      "learning_rate": 8e-05,
      "loss": 1.7543,
      "step": 4904
    },
    {
      "epoch": 0.530098346482222,
      "grad_norm": 0.5420807003974915,
      "learning_rate": 8e-05,
      "loss": 1.8302,
      "step": 4905
    },
    {
      "epoch": 0.5302064195396088,
      "grad_norm": 0.4450613558292389,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 4906
    },
    {
      "epoch": 0.5303144925969956,
      "grad_norm": 0.4306768774986267,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 4907
    },
    {
      "epoch": 0.5304225656543824,
      "grad_norm": 0.45799019932746887,
      "learning_rate": 8e-05,
      "loss": 1.7834,
      "step": 4908
    },
    {
      "epoch": 0.5305306387117692,
      "grad_norm": 0.4409727156162262,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 4909
    },
    {
      "epoch": 0.530638711769156,
      "grad_norm": 0.4476011097431183,
      "learning_rate": 8e-05,
      "loss": 1.6532,
      "step": 4910
    },
    {
      "epoch": 0.5307467848265427,
      "grad_norm": 0.45628440380096436,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 4911
    },
    {
      "epoch": 0.5308548578839295,
      "grad_norm": 0.43801817297935486,
      "learning_rate": 8e-05,
      "loss": 1.5478,
      "step": 4912
    },
    {
      "epoch": 0.5309629309413163,
      "grad_norm": 0.44181472063064575,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 4913
    },
    {
      "epoch": 0.5310710039987031,
      "grad_norm": 0.5240151286125183,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 4914
    },
    {
      "epoch": 0.5311790770560899,
      "grad_norm": 0.4425276219844818,
      "learning_rate": 8e-05,
      "loss": 1.4831,
      "step": 4915
    },
    {
      "epoch": 0.5312871501134767,
      "grad_norm": 0.5097845792770386,
      "learning_rate": 8e-05,
      "loss": 1.9585,
      "step": 4916
    },
    {
      "epoch": 0.5313952231708635,
      "grad_norm": 0.487388551235199,
      "learning_rate": 8e-05,
      "loss": 1.783,
      "step": 4917
    },
    {
      "epoch": 0.5315032962282503,
      "grad_norm": 0.4468676447868347,
      "learning_rate": 8e-05,
      "loss": 1.8652,
      "step": 4918
    },
    {
      "epoch": 0.5316113692856371,
      "grad_norm": 0.4469374120235443,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 4919
    },
    {
      "epoch": 0.5317194423430239,
      "grad_norm": 0.4670659303665161,
      "learning_rate": 8e-05,
      "loss": 1.6481,
      "step": 4920
    },
    {
      "epoch": 0.5318275154004107,
      "grad_norm": 0.4773979187011719,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 4921
    },
    {
      "epoch": 0.5319355884577974,
      "grad_norm": 0.4227922558784485,
      "learning_rate": 8e-05,
      "loss": 1.5525,
      "step": 4922
    },
    {
      "epoch": 0.5320436615151842,
      "grad_norm": 0.39445480704307556,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 4923
    },
    {
      "epoch": 0.532151734572571,
      "grad_norm": 0.4215022921562195,
      "learning_rate": 8e-05,
      "loss": 1.6348,
      "step": 4924
    },
    {
      "epoch": 0.5322598076299578,
      "grad_norm": 0.5110618472099304,
      "learning_rate": 8e-05,
      "loss": 1.8852,
      "step": 4925
    },
    {
      "epoch": 0.5323678806873446,
      "grad_norm": 0.4417065680027008,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 4926
    },
    {
      "epoch": 0.5324759537447314,
      "grad_norm": 0.464196115732193,
      "learning_rate": 8e-05,
      "loss": 1.7143,
      "step": 4927
    },
    {
      "epoch": 0.5325840268021182,
      "grad_norm": 0.44176584482192993,
      "learning_rate": 8e-05,
      "loss": 1.7683,
      "step": 4928
    },
    {
      "epoch": 0.532692099859505,
      "grad_norm": 0.5265250205993652,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 4929
    },
    {
      "epoch": 0.5328001729168919,
      "grad_norm": 0.4447842836380005,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 4930
    },
    {
      "epoch": 0.5329082459742787,
      "grad_norm": 0.45426374673843384,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 4931
    },
    {
      "epoch": 0.5330163190316654,
      "grad_norm": 0.4775089621543884,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 4932
    },
    {
      "epoch": 0.5331243920890522,
      "grad_norm": 0.4525960087776184,
      "learning_rate": 8e-05,
      "loss": 1.735,
      "step": 4933
    },
    {
      "epoch": 0.533232465146439,
      "grad_norm": 0.42379099130630493,
      "learning_rate": 8e-05,
      "loss": 1.5959,
      "step": 4934
    },
    {
      "epoch": 0.5333405382038258,
      "grad_norm": 0.46634969115257263,
      "learning_rate": 8e-05,
      "loss": 1.7307,
      "step": 4935
    },
    {
      "epoch": 0.5334486112612126,
      "grad_norm": 0.463201642036438,
      "learning_rate": 8e-05,
      "loss": 1.8605,
      "step": 4936
    },
    {
      "epoch": 0.5335566843185994,
      "grad_norm": 0.4313330054283142,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 4937
    },
    {
      "epoch": 0.5336647573759862,
      "grad_norm": 0.4679643213748932,
      "learning_rate": 8e-05,
      "loss": 1.615,
      "step": 4938
    },
    {
      "epoch": 0.533772830433373,
      "grad_norm": 0.4722180664539337,
      "learning_rate": 8e-05,
      "loss": 1.6809,
      "step": 4939
    },
    {
      "epoch": 0.5338809034907598,
      "grad_norm": 0.4636044204235077,
      "learning_rate": 8e-05,
      "loss": 1.792,
      "step": 4940
    },
    {
      "epoch": 0.5339889765481466,
      "grad_norm": 0.47307100892066956,
      "learning_rate": 8e-05,
      "loss": 1.8111,
      "step": 4941
    },
    {
      "epoch": 0.5340970496055334,
      "grad_norm": 0.5155717730522156,
      "learning_rate": 8e-05,
      "loss": 1.8423,
      "step": 4942
    },
    {
      "epoch": 0.5342051226629201,
      "grad_norm": 0.4781176745891571,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 4943
    },
    {
      "epoch": 0.5343131957203069,
      "grad_norm": 0.4633389413356781,
      "learning_rate": 8e-05,
      "loss": 1.518,
      "step": 4944
    },
    {
      "epoch": 0.5344212687776937,
      "grad_norm": 0.4205099046230316,
      "learning_rate": 8e-05,
      "loss": 1.6651,
      "step": 4945
    },
    {
      "epoch": 0.5345293418350805,
      "grad_norm": 0.41915759444236755,
      "learning_rate": 8e-05,
      "loss": 1.5586,
      "step": 4946
    },
    {
      "epoch": 0.5346374148924673,
      "grad_norm": 0.47371968626976013,
      "learning_rate": 8e-05,
      "loss": 1.8039,
      "step": 4947
    },
    {
      "epoch": 0.5347454879498541,
      "grad_norm": 0.42298412322998047,
      "learning_rate": 8e-05,
      "loss": 1.6245,
      "step": 4948
    },
    {
      "epoch": 0.5348535610072409,
      "grad_norm": 0.4497411251068115,
      "learning_rate": 8e-05,
      "loss": 1.8829,
      "step": 4949
    },
    {
      "epoch": 0.5349616340646277,
      "grad_norm": 0.4378730356693268,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 4950
    },
    {
      "epoch": 0.5350697071220145,
      "grad_norm": 0.47933605313301086,
      "learning_rate": 8e-05,
      "loss": 1.6206,
      "step": 4951
    },
    {
      "epoch": 0.5351777801794013,
      "grad_norm": 0.4580546021461487,
      "learning_rate": 8e-05,
      "loss": 1.6121,
      "step": 4952
    },
    {
      "epoch": 0.5352858532367881,
      "grad_norm": 0.4575256407260895,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 4953
    },
    {
      "epoch": 0.5353939262941748,
      "grad_norm": 0.4162660539150238,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 4954
    },
    {
      "epoch": 0.5355019993515616,
      "grad_norm": 0.4566630423069,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 4955
    },
    {
      "epoch": 0.5356100724089484,
      "grad_norm": 0.4746178388595581,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 4956
    },
    {
      "epoch": 0.5357181454663352,
      "grad_norm": 0.44345900416374207,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 4957
    },
    {
      "epoch": 0.535826218523722,
      "grad_norm": 0.4483538866043091,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 4958
    },
    {
      "epoch": 0.5359342915811088,
      "grad_norm": 0.4410283863544464,
      "learning_rate": 8e-05,
      "loss": 1.7431,
      "step": 4959
    },
    {
      "epoch": 0.5360423646384956,
      "grad_norm": 0.4770946204662323,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 4960
    },
    {
      "epoch": 0.5361504376958824,
      "grad_norm": 0.5090518593788147,
      "learning_rate": 8e-05,
      "loss": 1.8597,
      "step": 4961
    },
    {
      "epoch": 0.5362585107532692,
      "grad_norm": 0.4397258460521698,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 4962
    },
    {
      "epoch": 0.536366583810656,
      "grad_norm": 0.5352802872657776,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 4963
    },
    {
      "epoch": 0.5364746568680427,
      "grad_norm": 0.5198618173599243,
      "learning_rate": 8e-05,
      "loss": 1.7727,
      "step": 4964
    },
    {
      "epoch": 0.5365827299254295,
      "grad_norm": 0.4422585666179657,
      "learning_rate": 8e-05,
      "loss": 1.7106,
      "step": 4965
    },
    {
      "epoch": 0.5366908029828164,
      "grad_norm": 0.4607478976249695,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 4966
    },
    {
      "epoch": 0.5367988760402032,
      "grad_norm": 0.4347396790981293,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 4967
    },
    {
      "epoch": 0.53690694909759,
      "grad_norm": 0.45329323410987854,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 4968
    },
    {
      "epoch": 0.5370150221549768,
      "grad_norm": 0.4701964259147644,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 4969
    },
    {
      "epoch": 0.5371230952123636,
      "grad_norm": 0.4837842881679535,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 4970
    },
    {
      "epoch": 0.5372311682697504,
      "grad_norm": 0.4332659840583801,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 4971
    },
    {
      "epoch": 0.5373392413271372,
      "grad_norm": 0.4506637156009674,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 4972
    },
    {
      "epoch": 0.537447314384524,
      "grad_norm": 0.5201629400253296,
      "learning_rate": 8e-05,
      "loss": 1.6127,
      "step": 4973
    },
    {
      "epoch": 0.5375553874419108,
      "grad_norm": 0.4797949194908142,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 4974
    },
    {
      "epoch": 0.5376634604992975,
      "grad_norm": 0.4289214611053467,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 4975
    },
    {
      "epoch": 0.5377715335566843,
      "grad_norm": 0.45744815468788147,
      "learning_rate": 8e-05,
      "loss": 1.7403,
      "step": 4976
    },
    {
      "epoch": 0.5378796066140711,
      "grad_norm": 0.4619590938091278,
      "learning_rate": 8e-05,
      "loss": 1.5918,
      "step": 4977
    },
    {
      "epoch": 0.5379876796714579,
      "grad_norm": 0.4494166672229767,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 4978
    },
    {
      "epoch": 0.5380957527288447,
      "grad_norm": 0.5188494324684143,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 4979
    },
    {
      "epoch": 0.5382038257862315,
      "grad_norm": 0.45929479598999023,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 4980
    },
    {
      "epoch": 0.5383118988436183,
      "grad_norm": 0.42217525839805603,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 4981
    },
    {
      "epoch": 0.5384199719010051,
      "grad_norm": 0.43096718192100525,
      "learning_rate": 8e-05,
      "loss": 1.5291,
      "step": 4982
    },
    {
      "epoch": 0.5385280449583919,
      "grad_norm": 0.47813504934310913,
      "learning_rate": 8e-05,
      "loss": 1.8793,
      "step": 4983
    },
    {
      "epoch": 0.5386361180157787,
      "grad_norm": 0.42990100383758545,
      "learning_rate": 8e-05,
      "loss": 1.7252,
      "step": 4984
    },
    {
      "epoch": 0.5387441910731655,
      "grad_norm": 0.4469013512134552,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 4985
    },
    {
      "epoch": 0.5388522641305522,
      "grad_norm": 0.4693702161312103,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 4986
    },
    {
      "epoch": 0.538960337187939,
      "grad_norm": 0.43346068263053894,
      "learning_rate": 8e-05,
      "loss": 1.8226,
      "step": 4987
    },
    {
      "epoch": 0.5390684102453258,
      "grad_norm": 0.46888434886932373,
      "learning_rate": 8e-05,
      "loss": 1.6538,
      "step": 4988
    },
    {
      "epoch": 0.5391764833027126,
      "grad_norm": 0.4954688549041748,
      "learning_rate": 8e-05,
      "loss": 1.9294,
      "step": 4989
    },
    {
      "epoch": 0.5392845563600994,
      "grad_norm": 0.4240321218967438,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 4990
    },
    {
      "epoch": 0.5393926294174862,
      "grad_norm": 0.46745309233665466,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 4991
    },
    {
      "epoch": 0.539500702474873,
      "grad_norm": 0.44480040669441223,
      "learning_rate": 8e-05,
      "loss": 1.7475,
      "step": 4992
    },
    {
      "epoch": 0.5396087755322598,
      "grad_norm": 0.4772859215736389,
      "learning_rate": 8e-05,
      "loss": 1.8297,
      "step": 4993
    },
    {
      "epoch": 0.5397168485896466,
      "grad_norm": 0.47060924768447876,
      "learning_rate": 8e-05,
      "loss": 1.8543,
      "step": 4994
    },
    {
      "epoch": 0.5398249216470334,
      "grad_norm": 0.5036466121673584,
      "learning_rate": 8e-05,
      "loss": 1.9622,
      "step": 4995
    },
    {
      "epoch": 0.5399329947044201,
      "grad_norm": 0.4292738139629364,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 4996
    },
    {
      "epoch": 0.5400410677618069,
      "grad_norm": 0.4759220480918884,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 4997
    },
    {
      "epoch": 0.5401491408191937,
      "grad_norm": 0.4892449676990509,
      "learning_rate": 8e-05,
      "loss": 1.8147,
      "step": 4998
    },
    {
      "epoch": 0.5402572138765805,
      "grad_norm": 0.48332148790359497,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 4999
    },
    {
      "epoch": 0.5403652869339673,
      "grad_norm": 0.49124816060066223,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 5000
    },
    {
      "epoch": 0.5404733599913542,
      "grad_norm": 0.4608619213104248,
      "learning_rate": 8e-05,
      "loss": 1.9248,
      "step": 5001
    },
    {
      "epoch": 0.540581433048741,
      "grad_norm": 0.4298100769519806,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 5002
    },
    {
      "epoch": 0.5406895061061278,
      "grad_norm": 0.4268058240413666,
      "learning_rate": 8e-05,
      "loss": 1.6474,
      "step": 5003
    },
    {
      "epoch": 0.5407975791635146,
      "grad_norm": 0.42413002252578735,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 5004
    },
    {
      "epoch": 0.5409056522209014,
      "grad_norm": 0.4683758616447449,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 5005
    },
    {
      "epoch": 0.5410137252782882,
      "grad_norm": 0.465787410736084,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 5006
    },
    {
      "epoch": 0.5411217983356749,
      "grad_norm": 0.4559248983860016,
      "learning_rate": 8e-05,
      "loss": 1.6238,
      "step": 5007
    },
    {
      "epoch": 0.5412298713930617,
      "grad_norm": 0.4304090738296509,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 5008
    },
    {
      "epoch": 0.5413379444504485,
      "grad_norm": 0.4878780245780945,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 5009
    },
    {
      "epoch": 0.5414460175078353,
      "grad_norm": 0.4272918999195099,
      "learning_rate": 8e-05,
      "loss": 1.7432,
      "step": 5010
    },
    {
      "epoch": 0.5415540905652221,
      "grad_norm": 0.43340176343917847,
      "learning_rate": 8e-05,
      "loss": 1.7078,
      "step": 5011
    },
    {
      "epoch": 0.5416621636226089,
      "grad_norm": 0.47149229049682617,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 5012
    },
    {
      "epoch": 0.5417702366799957,
      "grad_norm": 0.46065273880958557,
      "learning_rate": 8e-05,
      "loss": 1.6184,
      "step": 5013
    },
    {
      "epoch": 0.5418783097373825,
      "grad_norm": 0.4539535343647003,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 5014
    },
    {
      "epoch": 0.5419863827947693,
      "grad_norm": 0.5085490345954895,
      "learning_rate": 8e-05,
      "loss": 1.7662,
      "step": 5015
    },
    {
      "epoch": 0.5420944558521561,
      "grad_norm": 0.44535478949546814,
      "learning_rate": 8e-05,
      "loss": 1.6924,
      "step": 5016
    },
    {
      "epoch": 0.5422025289095429,
      "grad_norm": 0.4576006829738617,
      "learning_rate": 8e-05,
      "loss": 1.8539,
      "step": 5017
    },
    {
      "epoch": 0.5423106019669296,
      "grad_norm": 0.42282360792160034,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 5018
    },
    {
      "epoch": 0.5424186750243164,
      "grad_norm": 0.4533088207244873,
      "learning_rate": 8e-05,
      "loss": 1.9194,
      "step": 5019
    },
    {
      "epoch": 0.5425267480817032,
      "grad_norm": 0.4925113916397095,
      "learning_rate": 8e-05,
      "loss": 1.9089,
      "step": 5020
    },
    {
      "epoch": 0.54263482113909,
      "grad_norm": 0.4320696294307709,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 5021
    },
    {
      "epoch": 0.5427428941964768,
      "grad_norm": 0.487800270318985,
      "learning_rate": 8e-05,
      "loss": 1.9691,
      "step": 5022
    },
    {
      "epoch": 0.5428509672538636,
      "grad_norm": 0.442401260137558,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 5023
    },
    {
      "epoch": 0.5429590403112504,
      "grad_norm": 0.4439208507537842,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 5024
    },
    {
      "epoch": 0.5430671133686372,
      "grad_norm": 0.4210017919540405,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 5025
    },
    {
      "epoch": 0.543175186426024,
      "grad_norm": 0.4557841122150421,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 5026
    },
    {
      "epoch": 0.5432832594834108,
      "grad_norm": 0.42770662903785706,
      "learning_rate": 8e-05,
      "loss": 1.6191,
      "step": 5027
    },
    {
      "epoch": 0.5433913325407976,
      "grad_norm": 0.48616379499435425,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 5028
    },
    {
      "epoch": 0.5434994055981843,
      "grad_norm": 0.4155402183532715,
      "learning_rate": 8e-05,
      "loss": 1.6538,
      "step": 5029
    },
    {
      "epoch": 0.5436074786555711,
      "grad_norm": 0.44352200627326965,
      "learning_rate": 8e-05,
      "loss": 1.5661,
      "step": 5030
    },
    {
      "epoch": 0.5437155517129579,
      "grad_norm": 0.45088088512420654,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 5031
    },
    {
      "epoch": 0.5438236247703447,
      "grad_norm": 0.46003979444503784,
      "learning_rate": 8e-05,
      "loss": 1.8135,
      "step": 5032
    },
    {
      "epoch": 0.5439316978277315,
      "grad_norm": 0.43093323707580566,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 5033
    },
    {
      "epoch": 0.5440397708851183,
      "grad_norm": 0.45626404881477356,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 5034
    },
    {
      "epoch": 0.5441478439425051,
      "grad_norm": 0.43688467144966125,
      "learning_rate": 8e-05,
      "loss": 1.7507,
      "step": 5035
    },
    {
      "epoch": 0.544255916999892,
      "grad_norm": 0.48153892159461975,
      "learning_rate": 8e-05,
      "loss": 1.8184,
      "step": 5036
    },
    {
      "epoch": 0.5443639900572788,
      "grad_norm": 0.4999698996543884,
      "learning_rate": 8e-05,
      "loss": 1.7745,
      "step": 5037
    },
    {
      "epoch": 0.5444720631146656,
      "grad_norm": 0.44290104508399963,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 5038
    },
    {
      "epoch": 0.5445801361720523,
      "grad_norm": 0.4447813630104065,
      "learning_rate": 8e-05,
      "loss": 1.6046,
      "step": 5039
    },
    {
      "epoch": 0.5446882092294391,
      "grad_norm": 0.483098566532135,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 5040
    },
    {
      "epoch": 0.5447962822868259,
      "grad_norm": 0.45254209637641907,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 5041
    },
    {
      "epoch": 0.5449043553442127,
      "grad_norm": 0.4390370845794678,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 5042
    },
    {
      "epoch": 0.5450124284015995,
      "grad_norm": 0.48912790417671204,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 5043
    },
    {
      "epoch": 0.5451205014589863,
      "grad_norm": 0.47345584630966187,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 5044
    },
    {
      "epoch": 0.5452285745163731,
      "grad_norm": 0.43179264664649963,
      "learning_rate": 8e-05,
      "loss": 1.5701,
      "step": 5045
    },
    {
      "epoch": 0.5453366475737599,
      "grad_norm": 0.419025182723999,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 5046
    },
    {
      "epoch": 0.5454447206311467,
      "grad_norm": 0.4953727126121521,
      "learning_rate": 8e-05,
      "loss": 1.8734,
      "step": 5047
    },
    {
      "epoch": 0.5455527936885335,
      "grad_norm": 0.5148927569389343,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 5048
    },
    {
      "epoch": 0.5456608667459203,
      "grad_norm": 0.4547547399997711,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 5049
    },
    {
      "epoch": 0.545768939803307,
      "grad_norm": 0.431106835603714,
      "learning_rate": 8e-05,
      "loss": 1.6261,
      "step": 5050
    },
    {
      "epoch": 0.5458770128606938,
      "grad_norm": 0.43646594882011414,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 5051
    },
    {
      "epoch": 0.5459850859180806,
      "grad_norm": 0.45300766825675964,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 5052
    },
    {
      "epoch": 0.5460931589754674,
      "grad_norm": 0.4770621955394745,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 5053
    },
    {
      "epoch": 0.5462012320328542,
      "grad_norm": 0.4313872754573822,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 5054
    },
    {
      "epoch": 0.546309305090241,
      "grad_norm": 0.43316957354545593,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 5055
    },
    {
      "epoch": 0.5464173781476278,
      "grad_norm": 0.453786700963974,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 5056
    },
    {
      "epoch": 0.5465254512050146,
      "grad_norm": 0.4847680330276489,
      "learning_rate": 8e-05,
      "loss": 1.9646,
      "step": 5057
    },
    {
      "epoch": 0.5466335242624014,
      "grad_norm": 0.4404902458190918,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 5058
    },
    {
      "epoch": 0.5467415973197882,
      "grad_norm": 0.4532972276210785,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 5059
    },
    {
      "epoch": 0.546849670377175,
      "grad_norm": 0.4783410429954529,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 5060
    },
    {
      "epoch": 0.5469577434345617,
      "grad_norm": 0.47126471996307373,
      "learning_rate": 8e-05,
      "loss": 1.7518,
      "step": 5061
    },
    {
      "epoch": 0.5470658164919485,
      "grad_norm": 0.455576092004776,
      "learning_rate": 8e-05,
      "loss": 1.5435,
      "step": 5062
    },
    {
      "epoch": 0.5471738895493353,
      "grad_norm": 0.4345244765281677,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 5063
    },
    {
      "epoch": 0.5472819626067221,
      "grad_norm": 0.44765639305114746,
      "learning_rate": 8e-05,
      "loss": 1.5831,
      "step": 5064
    },
    {
      "epoch": 0.5473900356641089,
      "grad_norm": 0.4559739828109741,
      "learning_rate": 8e-05,
      "loss": 1.7252,
      "step": 5065
    },
    {
      "epoch": 0.5474981087214957,
      "grad_norm": 0.4227026402950287,
      "learning_rate": 8e-05,
      "loss": 1.6006,
      "step": 5066
    },
    {
      "epoch": 0.5476061817788825,
      "grad_norm": 0.4268631935119629,
      "learning_rate": 8e-05,
      "loss": 1.5026,
      "step": 5067
    },
    {
      "epoch": 0.5477142548362693,
      "grad_norm": 0.4592440128326416,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 5068
    },
    {
      "epoch": 0.5478223278936561,
      "grad_norm": 0.4364212453365326,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 5069
    },
    {
      "epoch": 0.547930400951043,
      "grad_norm": 0.4465663433074951,
      "learning_rate": 8e-05,
      "loss": 1.576,
      "step": 5070
    },
    {
      "epoch": 0.5480384740084296,
      "grad_norm": 0.4760459363460541,
      "learning_rate": 8e-05,
      "loss": 1.6982,
      "step": 5071
    },
    {
      "epoch": 0.5481465470658164,
      "grad_norm": 0.4570913314819336,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 5072
    },
    {
      "epoch": 0.5482546201232033,
      "grad_norm": 0.4355856776237488,
      "learning_rate": 8e-05,
      "loss": 1.7706,
      "step": 5073
    },
    {
      "epoch": 0.54836269318059,
      "grad_norm": 0.4121935963630676,
      "learning_rate": 8e-05,
      "loss": 1.433,
      "step": 5074
    },
    {
      "epoch": 0.5484707662379769,
      "grad_norm": 0.49367713928222656,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 5075
    },
    {
      "epoch": 0.5485788392953637,
      "grad_norm": 0.4488212466239929,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 5076
    },
    {
      "epoch": 0.5486869123527505,
      "grad_norm": 0.4449191093444824,
      "learning_rate": 8e-05,
      "loss": 1.5227,
      "step": 5077
    },
    {
      "epoch": 0.5487949854101373,
      "grad_norm": 0.44847336411476135,
      "learning_rate": 8e-05,
      "loss": 1.5143,
      "step": 5078
    },
    {
      "epoch": 0.5489030584675241,
      "grad_norm": 0.4335845410823822,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 5079
    },
    {
      "epoch": 0.5490111315249109,
      "grad_norm": 0.432660311460495,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 5080
    },
    {
      "epoch": 0.5491192045822977,
      "grad_norm": 0.46550822257995605,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 5081
    },
    {
      "epoch": 0.5492272776396844,
      "grad_norm": 0.4759705364704132,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 5082
    },
    {
      "epoch": 0.5493353506970712,
      "grad_norm": 0.4294305443763733,
      "learning_rate": 8e-05,
      "loss": 1.5805,
      "step": 5083
    },
    {
      "epoch": 0.549443423754458,
      "grad_norm": 0.4563569724559784,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 5084
    },
    {
      "epoch": 0.5495514968118448,
      "grad_norm": 0.47725149989128113,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 5085
    },
    {
      "epoch": 0.5496595698692316,
      "grad_norm": 0.46542760729789734,
      "learning_rate": 8e-05,
      "loss": 1.8418,
      "step": 5086
    },
    {
      "epoch": 0.5497676429266184,
      "grad_norm": 0.492119699716568,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 5087
    },
    {
      "epoch": 0.5498757159840052,
      "grad_norm": 0.42810022830963135,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 5088
    },
    {
      "epoch": 0.549983789041392,
      "grad_norm": 0.4774756133556366,
      "learning_rate": 8e-05,
      "loss": 1.7921,
      "step": 5089
    },
    {
      "epoch": 0.5500918620987788,
      "grad_norm": 0.4160921275615692,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 5090
    },
    {
      "epoch": 0.5501999351561656,
      "grad_norm": 0.4276072084903717,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 5091
    },
    {
      "epoch": 0.5503080082135524,
      "grad_norm": 0.49171584844589233,
      "learning_rate": 8e-05,
      "loss": 1.8609,
      "step": 5092
    },
    {
      "epoch": 0.5504160812709391,
      "grad_norm": 0.5261533856391907,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 5093
    },
    {
      "epoch": 0.5505241543283259,
      "grad_norm": 0.5092872977256775,
      "learning_rate": 8e-05,
      "loss": 1.9191,
      "step": 5094
    },
    {
      "epoch": 0.5506322273857127,
      "grad_norm": 0.47201305627822876,
      "learning_rate": 8e-05,
      "loss": 1.7232,
      "step": 5095
    },
    {
      "epoch": 0.5507403004430995,
      "grad_norm": 0.43081122636795044,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 5096
    },
    {
      "epoch": 0.5508483735004863,
      "grad_norm": 0.48573341965675354,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 5097
    },
    {
      "epoch": 0.5509564465578731,
      "grad_norm": 0.4304220974445343,
      "learning_rate": 8e-05,
      "loss": 1.6567,
      "step": 5098
    },
    {
      "epoch": 0.5510645196152599,
      "grad_norm": 0.43866583704948425,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 5099
    },
    {
      "epoch": 0.5511725926726467,
      "grad_norm": 0.47060632705688477,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 5100
    },
    {
      "epoch": 0.5512806657300335,
      "grad_norm": 0.5046303868293762,
      "learning_rate": 8e-05,
      "loss": 1.6626,
      "step": 5101
    },
    {
      "epoch": 0.5513887387874203,
      "grad_norm": 0.46397021412849426,
      "learning_rate": 8e-05,
      "loss": 1.6697,
      "step": 5102
    },
    {
      "epoch": 0.5514968118448071,
      "grad_norm": 0.4196062982082367,
      "learning_rate": 8e-05,
      "loss": 1.6103,
      "step": 5103
    },
    {
      "epoch": 0.5516048849021938,
      "grad_norm": 0.4503146708011627,
      "learning_rate": 8e-05,
      "loss": 1.5903,
      "step": 5104
    },
    {
      "epoch": 0.5517129579595806,
      "grad_norm": 0.4398947060108185,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 5105
    },
    {
      "epoch": 0.5518210310169674,
      "grad_norm": 0.45349499583244324,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 5106
    },
    {
      "epoch": 0.5519291040743542,
      "grad_norm": 0.47890377044677734,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 5107
    },
    {
      "epoch": 0.552037177131741,
      "grad_norm": 0.4844301640987396,
      "learning_rate": 8e-05,
      "loss": 1.803,
      "step": 5108
    },
    {
      "epoch": 0.5521452501891279,
      "grad_norm": 0.4603897035121918,
      "learning_rate": 8e-05,
      "loss": 1.6652,
      "step": 5109
    },
    {
      "epoch": 0.5522533232465147,
      "grad_norm": 0.4396999776363373,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 5110
    },
    {
      "epoch": 0.5523613963039015,
      "grad_norm": 0.45798346400260925,
      "learning_rate": 8e-05,
      "loss": 1.6311,
      "step": 5111
    },
    {
      "epoch": 0.5524694693612883,
      "grad_norm": 0.45687416195869446,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 5112
    },
    {
      "epoch": 0.5525775424186751,
      "grad_norm": 0.5315713882446289,
      "learning_rate": 8e-05,
      "loss": 1.886,
      "step": 5113
    },
    {
      "epoch": 0.5526856154760618,
      "grad_norm": 0.482028990983963,
      "learning_rate": 8e-05,
      "loss": 1.822,
      "step": 5114
    },
    {
      "epoch": 0.5527936885334486,
      "grad_norm": 0.45141491293907166,
      "learning_rate": 8e-05,
      "loss": 1.7331,
      "step": 5115
    },
    {
      "epoch": 0.5529017615908354,
      "grad_norm": 0.4277758300304413,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 5116
    },
    {
      "epoch": 0.5530098346482222,
      "grad_norm": 0.44303393363952637,
      "learning_rate": 8e-05,
      "loss": 1.3372,
      "step": 5117
    },
    {
      "epoch": 0.553117907705609,
      "grad_norm": 0.5126786231994629,
      "learning_rate": 8e-05,
      "loss": 1.8359,
      "step": 5118
    },
    {
      "epoch": 0.5532259807629958,
      "grad_norm": 0.4971819519996643,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 5119
    },
    {
      "epoch": 0.5533340538203826,
      "grad_norm": 0.4635187089443207,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 5120
    },
    {
      "epoch": 0.5534421268777694,
      "grad_norm": 0.4487621486186981,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 5121
    },
    {
      "epoch": 0.5535501999351562,
      "grad_norm": 0.549837052822113,
      "learning_rate": 8e-05,
      "loss": 1.9174,
      "step": 5122
    },
    {
      "epoch": 0.553658272992543,
      "grad_norm": 0.530081570148468,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 5123
    },
    {
      "epoch": 0.5537663460499298,
      "grad_norm": 0.4937395453453064,
      "learning_rate": 8e-05,
      "loss": 1.6612,
      "step": 5124
    },
    {
      "epoch": 0.5538744191073165,
      "grad_norm": 0.47606930136680603,
      "learning_rate": 8e-05,
      "loss": 1.892,
      "step": 5125
    },
    {
      "epoch": 0.5539824921647033,
      "grad_norm": 0.45198360085487366,
      "learning_rate": 8e-05,
      "loss": 1.5399,
      "step": 5126
    },
    {
      "epoch": 0.5540905652220901,
      "grad_norm": 0.479468435049057,
      "learning_rate": 8e-05,
      "loss": 1.4473,
      "step": 5127
    },
    {
      "epoch": 0.5541986382794769,
      "grad_norm": 0.48571234941482544,
      "learning_rate": 8e-05,
      "loss": 1.9392,
      "step": 5128
    },
    {
      "epoch": 0.5543067113368637,
      "grad_norm": 0.4660547077655792,
      "learning_rate": 8e-05,
      "loss": 1.7989,
      "step": 5129
    },
    {
      "epoch": 0.5544147843942505,
      "grad_norm": 0.45751211047172546,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 5130
    },
    {
      "epoch": 0.5545228574516373,
      "grad_norm": 0.4584125578403473,
      "learning_rate": 8e-05,
      "loss": 1.8585,
      "step": 5131
    },
    {
      "epoch": 0.5546309305090241,
      "grad_norm": 0.4545653760433197,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 5132
    },
    {
      "epoch": 0.5547390035664109,
      "grad_norm": 0.4518764615058899,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 5133
    },
    {
      "epoch": 0.5548470766237977,
      "grad_norm": 0.437589168548584,
      "learning_rate": 8e-05,
      "loss": 1.6685,
      "step": 5134
    },
    {
      "epoch": 0.5549551496811845,
      "grad_norm": 0.45255497097969055,
      "learning_rate": 8e-05,
      "loss": 1.7325,
      "step": 5135
    },
    {
      "epoch": 0.5550632227385712,
      "grad_norm": 0.4931819438934326,
      "learning_rate": 8e-05,
      "loss": 1.7674,
      "step": 5136
    },
    {
      "epoch": 0.555171295795958,
      "grad_norm": 0.46312427520751953,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 5137
    },
    {
      "epoch": 0.5552793688533448,
      "grad_norm": 0.42596176266670227,
      "learning_rate": 8e-05,
      "loss": 1.5937,
      "step": 5138
    },
    {
      "epoch": 0.5553874419107316,
      "grad_norm": 0.5149576663970947,
      "learning_rate": 8e-05,
      "loss": 1.8478,
      "step": 5139
    },
    {
      "epoch": 0.5554955149681184,
      "grad_norm": 0.43380165100097656,
      "learning_rate": 8e-05,
      "loss": 1.8206,
      "step": 5140
    },
    {
      "epoch": 0.5556035880255052,
      "grad_norm": 0.49397820234298706,
      "learning_rate": 8e-05,
      "loss": 1.8752,
      "step": 5141
    },
    {
      "epoch": 0.555711661082892,
      "grad_norm": 0.4422333538532257,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 5142
    },
    {
      "epoch": 0.5558197341402789,
      "grad_norm": 0.4713895618915558,
      "learning_rate": 8e-05,
      "loss": 1.6442,
      "step": 5143
    },
    {
      "epoch": 0.5559278071976657,
      "grad_norm": 0.4527498185634613,
      "learning_rate": 8e-05,
      "loss": 1.5708,
      "step": 5144
    },
    {
      "epoch": 0.5560358802550525,
      "grad_norm": 0.46678975224494934,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 5145
    },
    {
      "epoch": 0.5561439533124392,
      "grad_norm": 0.445804238319397,
      "learning_rate": 8e-05,
      "loss": 1.489,
      "step": 5146
    },
    {
      "epoch": 0.556252026369826,
      "grad_norm": 0.5070182085037231,
      "learning_rate": 8e-05,
      "loss": 1.591,
      "step": 5147
    },
    {
      "epoch": 0.5563600994272128,
      "grad_norm": 0.45364680886268616,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 5148
    },
    {
      "epoch": 0.5564681724845996,
      "grad_norm": 0.4734874367713928,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 5149
    },
    {
      "epoch": 0.5565762455419864,
      "grad_norm": 0.5121163725852966,
      "learning_rate": 8e-05,
      "loss": 1.8567,
      "step": 5150
    },
    {
      "epoch": 0.5566843185993732,
      "grad_norm": 0.4427231550216675,
      "learning_rate": 8e-05,
      "loss": 1.7879,
      "step": 5151
    },
    {
      "epoch": 0.55679239165676,
      "grad_norm": 0.4942622482776642,
      "learning_rate": 8e-05,
      "loss": 1.8663,
      "step": 5152
    },
    {
      "epoch": 0.5569004647141468,
      "grad_norm": 0.46653610467910767,
      "learning_rate": 8e-05,
      "loss": 1.6584,
      "step": 5153
    },
    {
      "epoch": 0.5570085377715336,
      "grad_norm": 0.4515688419342041,
      "learning_rate": 8e-05,
      "loss": 1.6334,
      "step": 5154
    },
    {
      "epoch": 0.5571166108289204,
      "grad_norm": 0.505327045917511,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 5155
    },
    {
      "epoch": 0.5572246838863072,
      "grad_norm": 0.4266228675842285,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 5156
    },
    {
      "epoch": 0.5573327569436939,
      "grad_norm": 0.4337596893310547,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 5157
    },
    {
      "epoch": 0.5574408300010807,
      "grad_norm": 0.4420917332172394,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 5158
    },
    {
      "epoch": 0.5575489030584675,
      "grad_norm": 0.4374214708805084,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 5159
    },
    {
      "epoch": 0.5576569761158543,
      "grad_norm": 0.4746066927909851,
      "learning_rate": 8e-05,
      "loss": 1.7281,
      "step": 5160
    },
    {
      "epoch": 0.5577650491732411,
      "grad_norm": 0.46112895011901855,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 5161
    },
    {
      "epoch": 0.5578731222306279,
      "grad_norm": 0.4244885742664337,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 5162
    },
    {
      "epoch": 0.5579811952880147,
      "grad_norm": 0.4253966212272644,
      "learning_rate": 8e-05,
      "loss": 1.5603,
      "step": 5163
    },
    {
      "epoch": 0.5580892683454015,
      "grad_norm": 0.4248102307319641,
      "learning_rate": 8e-05,
      "loss": 1.6413,
      "step": 5164
    },
    {
      "epoch": 0.5581973414027883,
      "grad_norm": 0.4568192958831787,
      "learning_rate": 8e-05,
      "loss": 1.6305,
      "step": 5165
    },
    {
      "epoch": 0.5583054144601751,
      "grad_norm": 0.4551163613796234,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 5166
    },
    {
      "epoch": 0.5584134875175619,
      "grad_norm": 0.47130024433135986,
      "learning_rate": 8e-05,
      "loss": 1.853,
      "step": 5167
    },
    {
      "epoch": 0.5585215605749486,
      "grad_norm": 0.5182979106903076,
      "learning_rate": 8e-05,
      "loss": 1.9163,
      "step": 5168
    },
    {
      "epoch": 0.5586296336323354,
      "grad_norm": 0.4995230436325073,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 5169
    },
    {
      "epoch": 0.5587377066897222,
      "grad_norm": 0.46121713519096375,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 5170
    },
    {
      "epoch": 0.558845779747109,
      "grad_norm": 0.4357644021511078,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 5171
    },
    {
      "epoch": 0.5589538528044958,
      "grad_norm": 0.41372138261795044,
      "learning_rate": 8e-05,
      "loss": 1.65,
      "step": 5172
    },
    {
      "epoch": 0.5590619258618826,
      "grad_norm": 0.44551050662994385,
      "learning_rate": 8e-05,
      "loss": 1.5736,
      "step": 5173
    },
    {
      "epoch": 0.5591699989192694,
      "grad_norm": 0.4462915062904358,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 5174
    },
    {
      "epoch": 0.5592780719766562,
      "grad_norm": 0.4524897038936615,
      "learning_rate": 8e-05,
      "loss": 1.6665,
      "step": 5175
    },
    {
      "epoch": 0.559386145034043,
      "grad_norm": 0.4689357876777649,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 5176
    },
    {
      "epoch": 0.5594942180914299,
      "grad_norm": 0.43674010038375854,
      "learning_rate": 8e-05,
      "loss": 1.6045,
      "step": 5177
    },
    {
      "epoch": 0.5596022911488167,
      "grad_norm": 0.5229376554489136,
      "learning_rate": 8e-05,
      "loss": 1.8543,
      "step": 5178
    },
    {
      "epoch": 0.5597103642062033,
      "grad_norm": 0.45374736189842224,
      "learning_rate": 8e-05,
      "loss": 1.627,
      "step": 5179
    },
    {
      "epoch": 0.5598184372635902,
      "grad_norm": 0.4578039050102234,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 5180
    },
    {
      "epoch": 0.559926510320977,
      "grad_norm": 0.4622832238674164,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 5181
    },
    {
      "epoch": 0.5600345833783638,
      "grad_norm": 0.48715877532958984,
      "learning_rate": 8e-05,
      "loss": 1.6424,
      "step": 5182
    },
    {
      "epoch": 0.5601426564357506,
      "grad_norm": 0.4390258193016052,
      "learning_rate": 8e-05,
      "loss": 1.6969,
      "step": 5183
    },
    {
      "epoch": 0.5602507294931374,
      "grad_norm": 0.49545007944107056,
      "learning_rate": 8e-05,
      "loss": 1.9469,
      "step": 5184
    },
    {
      "epoch": 0.5603588025505242,
      "grad_norm": 0.476457804441452,
      "learning_rate": 8e-05,
      "loss": 1.8559,
      "step": 5185
    },
    {
      "epoch": 0.560466875607911,
      "grad_norm": 0.4435414671897888,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 5186
    },
    {
      "epoch": 0.5605749486652978,
      "grad_norm": 0.43020933866500854,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 5187
    },
    {
      "epoch": 0.5606830217226846,
      "grad_norm": 0.4812830984592438,
      "learning_rate": 8e-05,
      "loss": 1.8858,
      "step": 5188
    },
    {
      "epoch": 0.5607910947800713,
      "grad_norm": 0.4395637810230255,
      "learning_rate": 8e-05,
      "loss": 1.6478,
      "step": 5189
    },
    {
      "epoch": 0.5608991678374581,
      "grad_norm": 0.46113893389701843,
      "learning_rate": 8e-05,
      "loss": 1.7677,
      "step": 5190
    },
    {
      "epoch": 0.5610072408948449,
      "grad_norm": 0.4880397319793701,
      "learning_rate": 8e-05,
      "loss": 1.8705,
      "step": 5191
    },
    {
      "epoch": 0.5611153139522317,
      "grad_norm": 0.5418736338615417,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 5192
    },
    {
      "epoch": 0.5612233870096185,
      "grad_norm": 0.4741430878639221,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 5193
    },
    {
      "epoch": 0.5613314600670053,
      "grad_norm": 0.5103474855422974,
      "learning_rate": 8e-05,
      "loss": 1.8418,
      "step": 5194
    },
    {
      "epoch": 0.5614395331243921,
      "grad_norm": 0.4208717942237854,
      "learning_rate": 8e-05,
      "loss": 1.5362,
      "step": 5195
    },
    {
      "epoch": 0.5615476061817789,
      "grad_norm": 0.419827401638031,
      "learning_rate": 8e-05,
      "loss": 1.5074,
      "step": 5196
    },
    {
      "epoch": 0.5616556792391657,
      "grad_norm": 0.4584641754627228,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 5197
    },
    {
      "epoch": 0.5617637522965525,
      "grad_norm": 0.4482277035713196,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 5198
    },
    {
      "epoch": 0.5618718253539393,
      "grad_norm": 0.45975780487060547,
      "learning_rate": 8e-05,
      "loss": 1.7921,
      "step": 5199
    },
    {
      "epoch": 0.561979898411326,
      "grad_norm": 0.47934210300445557,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 5200
    },
    {
      "epoch": 0.5620879714687128,
      "grad_norm": 0.45348939299583435,
      "learning_rate": 8e-05,
      "loss": 1.8574,
      "step": 5201
    },
    {
      "epoch": 0.5621960445260996,
      "grad_norm": 0.44837844371795654,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 5202
    },
    {
      "epoch": 0.5623041175834864,
      "grad_norm": 0.4296506345272064,
      "learning_rate": 8e-05,
      "loss": 1.5865,
      "step": 5203
    },
    {
      "epoch": 0.5624121906408732,
      "grad_norm": 0.45431363582611084,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 5204
    },
    {
      "epoch": 0.56252026369826,
      "grad_norm": 0.45959824323654175,
      "learning_rate": 8e-05,
      "loss": 1.9084,
      "step": 5205
    },
    {
      "epoch": 0.5626283367556468,
      "grad_norm": 0.4609025716781616,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 5206
    },
    {
      "epoch": 0.5627364098130336,
      "grad_norm": 0.4508279860019684,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 5207
    },
    {
      "epoch": 0.5628444828704204,
      "grad_norm": 0.43421608209609985,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 5208
    },
    {
      "epoch": 0.5629525559278072,
      "grad_norm": 0.45012593269348145,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 5209
    },
    {
      "epoch": 0.563060628985194,
      "grad_norm": 0.4817896783351898,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 5210
    },
    {
      "epoch": 0.5631687020425807,
      "grad_norm": 0.48874738812446594,
      "learning_rate": 8e-05,
      "loss": 1.8532,
      "step": 5211
    },
    {
      "epoch": 0.5632767750999675,
      "grad_norm": 0.4596792459487915,
      "learning_rate": 8e-05,
      "loss": 1.9286,
      "step": 5212
    },
    {
      "epoch": 0.5633848481573543,
      "grad_norm": 0.43587726354599,
      "learning_rate": 8e-05,
      "loss": 1.7182,
      "step": 5213
    },
    {
      "epoch": 0.5634929212147411,
      "grad_norm": 0.43641337752342224,
      "learning_rate": 8e-05,
      "loss": 1.7126,
      "step": 5214
    },
    {
      "epoch": 0.563600994272128,
      "grad_norm": 0.4742232859134674,
      "learning_rate": 8e-05,
      "loss": 1.8498,
      "step": 5215
    },
    {
      "epoch": 0.5637090673295148,
      "grad_norm": 0.4941873550415039,
      "learning_rate": 8e-05,
      "loss": 1.6665,
      "step": 5216
    },
    {
      "epoch": 0.5638171403869016,
      "grad_norm": 0.476133793592453,
      "learning_rate": 8e-05,
      "loss": 1.686,
      "step": 5217
    },
    {
      "epoch": 0.5639252134442884,
      "grad_norm": 0.42435476183891296,
      "learning_rate": 8e-05,
      "loss": 1.6852,
      "step": 5218
    },
    {
      "epoch": 0.5640332865016752,
      "grad_norm": 0.46463945508003235,
      "learning_rate": 8e-05,
      "loss": 1.804,
      "step": 5219
    },
    {
      "epoch": 0.564141359559062,
      "grad_norm": 0.4531163275241852,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 5220
    },
    {
      "epoch": 0.5642494326164487,
      "grad_norm": 0.48327764868736267,
      "learning_rate": 8e-05,
      "loss": 1.965,
      "step": 5221
    },
    {
      "epoch": 0.5643575056738355,
      "grad_norm": 0.4405801296234131,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 5222
    },
    {
      "epoch": 0.5644655787312223,
      "grad_norm": 0.473025918006897,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 5223
    },
    {
      "epoch": 0.5645736517886091,
      "grad_norm": 0.44665858149528503,
      "learning_rate": 8e-05,
      "loss": 1.7349,
      "step": 5224
    },
    {
      "epoch": 0.5646817248459959,
      "grad_norm": 0.4517905116081238,
      "learning_rate": 8e-05,
      "loss": 1.7902,
      "step": 5225
    },
    {
      "epoch": 0.5647897979033827,
      "grad_norm": 0.4762458801269531,
      "learning_rate": 8e-05,
      "loss": 1.8566,
      "step": 5226
    },
    {
      "epoch": 0.5648978709607695,
      "grad_norm": 0.41608330607414246,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 5227
    },
    {
      "epoch": 0.5650059440181563,
      "grad_norm": 0.4411550760269165,
      "learning_rate": 8e-05,
      "loss": 1.6755,
      "step": 5228
    },
    {
      "epoch": 0.5651140170755431,
      "grad_norm": 0.4607795178890228,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 5229
    },
    {
      "epoch": 0.5652220901329299,
      "grad_norm": 0.4429433047771454,
      "learning_rate": 8e-05,
      "loss": 1.7763,
      "step": 5230
    },
    {
      "epoch": 0.5653301631903167,
      "grad_norm": 0.467380553483963,
      "learning_rate": 8e-05,
      "loss": 1.742,
      "step": 5231
    },
    {
      "epoch": 0.5654382362477034,
      "grad_norm": 0.45474934577941895,
      "learning_rate": 8e-05,
      "loss": 1.5793,
      "step": 5232
    },
    {
      "epoch": 0.5655463093050902,
      "grad_norm": 0.4485014081001282,
      "learning_rate": 8e-05,
      "loss": 1.603,
      "step": 5233
    },
    {
      "epoch": 0.565654382362477,
      "grad_norm": 0.5176403522491455,
      "learning_rate": 8e-05,
      "loss": 1.9008,
      "step": 5234
    },
    {
      "epoch": 0.5657624554198638,
      "grad_norm": 0.48003119230270386,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 5235
    },
    {
      "epoch": 0.5658705284772506,
      "grad_norm": 0.4269717335700989,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 5236
    },
    {
      "epoch": 0.5659786015346374,
      "grad_norm": 0.48167166113853455,
      "learning_rate": 8e-05,
      "loss": 1.9197,
      "step": 5237
    },
    {
      "epoch": 0.5660866745920242,
      "grad_norm": 0.45992833375930786,
      "learning_rate": 8e-05,
      "loss": 1.5102,
      "step": 5238
    },
    {
      "epoch": 0.566194747649411,
      "grad_norm": 0.44162362813949585,
      "learning_rate": 8e-05,
      "loss": 1.6084,
      "step": 5239
    },
    {
      "epoch": 0.5663028207067978,
      "grad_norm": 0.48860490322113037,
      "learning_rate": 8e-05,
      "loss": 1.8281,
      "step": 5240
    },
    {
      "epoch": 0.5664108937641846,
      "grad_norm": 0.44954925775527954,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 5241
    },
    {
      "epoch": 0.5665189668215714,
      "grad_norm": 0.4339462220668793,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 5242
    },
    {
      "epoch": 0.5666270398789581,
      "grad_norm": 0.44080260396003723,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 5243
    },
    {
      "epoch": 0.5667351129363449,
      "grad_norm": 0.42393597960472107,
      "learning_rate": 8e-05,
      "loss": 1.671,
      "step": 5244
    },
    {
      "epoch": 0.5668431859937317,
      "grad_norm": 0.6082028746604919,
      "learning_rate": 8e-05,
      "loss": 1.954,
      "step": 5245
    },
    {
      "epoch": 0.5669512590511185,
      "grad_norm": 0.4891705811023712,
      "learning_rate": 8e-05,
      "loss": 1.8463,
      "step": 5246
    },
    {
      "epoch": 0.5670593321085053,
      "grad_norm": 0.43596941232681274,
      "learning_rate": 8e-05,
      "loss": 1.7624,
      "step": 5247
    },
    {
      "epoch": 0.5671674051658921,
      "grad_norm": 0.48031365871429443,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 5248
    },
    {
      "epoch": 0.567275478223279,
      "grad_norm": 0.4315033257007599,
      "learning_rate": 8e-05,
      "loss": 1.7703,
      "step": 5249
    },
    {
      "epoch": 0.5673835512806658,
      "grad_norm": 0.47637084126472473,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 5250
    },
    {
      "epoch": 0.5674916243380526,
      "grad_norm": 0.42846372723579407,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 5251
    },
    {
      "epoch": 0.5675996973954394,
      "grad_norm": 0.48459893465042114,
      "learning_rate": 8e-05,
      "loss": 1.7641,
      "step": 5252
    },
    {
      "epoch": 0.5677077704528262,
      "grad_norm": 0.4880440831184387,
      "learning_rate": 8e-05,
      "loss": 1.9141,
      "step": 5253
    },
    {
      "epoch": 0.5678158435102129,
      "grad_norm": 0.43372347950935364,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 5254
    },
    {
      "epoch": 0.5679239165675997,
      "grad_norm": 0.4700317680835724,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 5255
    },
    {
      "epoch": 0.5680319896249865,
      "grad_norm": 0.46579593420028687,
      "learning_rate": 8e-05,
      "loss": 1.8306,
      "step": 5256
    },
    {
      "epoch": 0.5681400626823733,
      "grad_norm": 0.722809374332428,
      "learning_rate": 8e-05,
      "loss": 2.046,
      "step": 5257
    },
    {
      "epoch": 0.5682481357397601,
      "grad_norm": 0.46818938851356506,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 5258
    },
    {
      "epoch": 0.5683562087971469,
      "grad_norm": 0.4825763404369354,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 5259
    },
    {
      "epoch": 0.5684642818545337,
      "grad_norm": 0.4721335172653198,
      "learning_rate": 8e-05,
      "loss": 1.8015,
      "step": 5260
    },
    {
      "epoch": 0.5685723549119205,
      "grad_norm": 0.4445522129535675,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 5261
    },
    {
      "epoch": 0.5686804279693073,
      "grad_norm": 0.4699435830116272,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 5262
    },
    {
      "epoch": 0.5687885010266941,
      "grad_norm": 0.4484909176826477,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 5263
    },
    {
      "epoch": 0.5688965740840808,
      "grad_norm": 0.4674362242221832,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 5264
    },
    {
      "epoch": 0.5690046471414676,
      "grad_norm": 0.48993754386901855,
      "learning_rate": 8e-05,
      "loss": 1.5319,
      "step": 5265
    },
    {
      "epoch": 0.5691127201988544,
      "grad_norm": 0.43317797780036926,
      "learning_rate": 8e-05,
      "loss": 1.5556,
      "step": 5266
    },
    {
      "epoch": 0.5692207932562412,
      "grad_norm": 0.4868384003639221,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 5267
    },
    {
      "epoch": 0.569328866313628,
      "grad_norm": 0.45298099517822266,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 5268
    },
    {
      "epoch": 0.5694369393710148,
      "grad_norm": 0.4639950692653656,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 5269
    },
    {
      "epoch": 0.5695450124284016,
      "grad_norm": 0.4172171354293823,
      "learning_rate": 8e-05,
      "loss": 1.5093,
      "step": 5270
    },
    {
      "epoch": 0.5696530854857884,
      "grad_norm": 0.45202991366386414,
      "learning_rate": 8e-05,
      "loss": 1.6785,
      "step": 5271
    },
    {
      "epoch": 0.5697611585431752,
      "grad_norm": 0.4786117374897003,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 5272
    },
    {
      "epoch": 0.569869231600562,
      "grad_norm": 0.4332369565963745,
      "learning_rate": 8e-05,
      "loss": 1.7317,
      "step": 5273
    },
    {
      "epoch": 0.5699773046579488,
      "grad_norm": 0.43024006485939026,
      "learning_rate": 8e-05,
      "loss": 1.5201,
      "step": 5274
    },
    {
      "epoch": 0.5700853777153355,
      "grad_norm": 0.4180617332458496,
      "learning_rate": 8e-05,
      "loss": 1.488,
      "step": 5275
    },
    {
      "epoch": 0.5701934507727223,
      "grad_norm": 0.4562242329120636,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 5276
    },
    {
      "epoch": 0.5703015238301091,
      "grad_norm": 0.46327367424964905,
      "learning_rate": 8e-05,
      "loss": 1.848,
      "step": 5277
    },
    {
      "epoch": 0.5704095968874959,
      "grad_norm": 0.501304030418396,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 5278
    },
    {
      "epoch": 0.5705176699448827,
      "grad_norm": 0.48164886236190796,
      "learning_rate": 8e-05,
      "loss": 1.7488,
      "step": 5279
    },
    {
      "epoch": 0.5706257430022695,
      "grad_norm": 0.4413878321647644,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 5280
    },
    {
      "epoch": 0.5707338160596563,
      "grad_norm": 0.4641432464122772,
      "learning_rate": 8e-05,
      "loss": 1.7223,
      "step": 5281
    },
    {
      "epoch": 0.5708418891170431,
      "grad_norm": 0.45106151700019836,
      "learning_rate": 8e-05,
      "loss": 1.6812,
      "step": 5282
    },
    {
      "epoch": 0.57094996217443,
      "grad_norm": 0.429231733083725,
      "learning_rate": 8e-05,
      "loss": 1.6177,
      "step": 5283
    },
    {
      "epoch": 0.5710580352318168,
      "grad_norm": 0.4489170014858246,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 5284
    },
    {
      "epoch": 0.5711661082892036,
      "grad_norm": 0.4684273600578308,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 5285
    },
    {
      "epoch": 0.5712741813465902,
      "grad_norm": 0.42327946424484253,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 5286
    },
    {
      "epoch": 0.571382254403977,
      "grad_norm": 0.4480898678302765,
      "learning_rate": 8e-05,
      "loss": 1.5599,
      "step": 5287
    },
    {
      "epoch": 0.5714903274613639,
      "grad_norm": 0.41945308446884155,
      "learning_rate": 8e-05,
      "loss": 1.593,
      "step": 5288
    },
    {
      "epoch": 0.5715984005187507,
      "grad_norm": 0.4353809356689453,
      "learning_rate": 8e-05,
      "loss": 1.6401,
      "step": 5289
    },
    {
      "epoch": 0.5717064735761375,
      "grad_norm": 0.4442664682865143,
      "learning_rate": 8e-05,
      "loss": 1.7296,
      "step": 5290
    },
    {
      "epoch": 0.5718145466335243,
      "grad_norm": 0.4704211950302124,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 5291
    },
    {
      "epoch": 0.5719226196909111,
      "grad_norm": 0.503862738609314,
      "learning_rate": 8e-05,
      "loss": 1.8344,
      "step": 5292
    },
    {
      "epoch": 0.5720306927482979,
      "grad_norm": 0.4672357439994812,
      "learning_rate": 8e-05,
      "loss": 1.7491,
      "step": 5293
    },
    {
      "epoch": 0.5721387658056847,
      "grad_norm": 0.4646608829498291,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 5294
    },
    {
      "epoch": 0.5722468388630715,
      "grad_norm": 0.4516545534133911,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 5295
    },
    {
      "epoch": 0.5723549119204582,
      "grad_norm": 0.4524863660335541,
      "learning_rate": 8e-05,
      "loss": 1.7958,
      "step": 5296
    },
    {
      "epoch": 0.572462984977845,
      "grad_norm": 0.4419468641281128,
      "learning_rate": 8e-05,
      "loss": 1.7353,
      "step": 5297
    },
    {
      "epoch": 0.5725710580352318,
      "grad_norm": 0.4354066252708435,
      "learning_rate": 8e-05,
      "loss": 1.6492,
      "step": 5298
    },
    {
      "epoch": 0.5726791310926186,
      "grad_norm": 0.4203045666217804,
      "learning_rate": 8e-05,
      "loss": 1.7028,
      "step": 5299
    },
    {
      "epoch": 0.5727872041500054,
      "grad_norm": 0.440807044506073,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 5300
    },
    {
      "epoch": 0.5728952772073922,
      "grad_norm": 0.4620375633239746,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 5301
    },
    {
      "epoch": 0.573003350264779,
      "grad_norm": 0.4463610351085663,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 5302
    },
    {
      "epoch": 0.5731114233221658,
      "grad_norm": 0.43299809098243713,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 5303
    },
    {
      "epoch": 0.5732194963795526,
      "grad_norm": 0.46380361914634705,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 5304
    },
    {
      "epoch": 0.5733275694369394,
      "grad_norm": 0.47358763217926025,
      "learning_rate": 8e-05,
      "loss": 1.6038,
      "step": 5305
    },
    {
      "epoch": 0.5734356424943262,
      "grad_norm": 0.47635430097579956,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 5306
    },
    {
      "epoch": 0.5735437155517129,
      "grad_norm": 0.47478294372558594,
      "learning_rate": 8e-05,
      "loss": 1.8647,
      "step": 5307
    },
    {
      "epoch": 0.5736517886090997,
      "grad_norm": 0.45854952931404114,
      "learning_rate": 8e-05,
      "loss": 1.7194,
      "step": 5308
    },
    {
      "epoch": 0.5737598616664865,
      "grad_norm": 0.4625128209590912,
      "learning_rate": 8e-05,
      "loss": 1.8947,
      "step": 5309
    },
    {
      "epoch": 0.5738679347238733,
      "grad_norm": 0.49244216084480286,
      "learning_rate": 8e-05,
      "loss": 1.8196,
      "step": 5310
    },
    {
      "epoch": 0.5739760077812601,
      "grad_norm": 0.4399133324623108,
      "learning_rate": 8e-05,
      "loss": 1.6829,
      "step": 5311
    },
    {
      "epoch": 0.5740840808386469,
      "grad_norm": 0.44024139642715454,
      "learning_rate": 8e-05,
      "loss": 1.6617,
      "step": 5312
    },
    {
      "epoch": 0.5741921538960337,
      "grad_norm": 0.4127318859100342,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 5313
    },
    {
      "epoch": 0.5743002269534205,
      "grad_norm": 0.47349345684051514,
      "learning_rate": 8e-05,
      "loss": 1.7636,
      "step": 5314
    },
    {
      "epoch": 0.5744083000108073,
      "grad_norm": 0.4784218966960907,
      "learning_rate": 8e-05,
      "loss": 1.6234,
      "step": 5315
    },
    {
      "epoch": 0.5745163730681941,
      "grad_norm": 0.41884952783584595,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 5316
    },
    {
      "epoch": 0.574624446125581,
      "grad_norm": 0.4733394384384155,
      "learning_rate": 8e-05,
      "loss": 1.7878,
      "step": 5317
    },
    {
      "epoch": 0.5747325191829676,
      "grad_norm": 0.441173791885376,
      "learning_rate": 8e-05,
      "loss": 1.5215,
      "step": 5318
    },
    {
      "epoch": 0.5748405922403544,
      "grad_norm": 0.4283747673034668,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 5319
    },
    {
      "epoch": 0.5749486652977412,
      "grad_norm": 0.4305785298347473,
      "learning_rate": 8e-05,
      "loss": 1.5821,
      "step": 5320
    },
    {
      "epoch": 0.575056738355128,
      "grad_norm": 0.4923546314239502,
      "learning_rate": 8e-05,
      "loss": 1.7626,
      "step": 5321
    },
    {
      "epoch": 0.5751648114125149,
      "grad_norm": 0.4799255430698395,
      "learning_rate": 8e-05,
      "loss": 1.8763,
      "step": 5322
    },
    {
      "epoch": 0.5752728844699017,
      "grad_norm": 0.44048771262168884,
      "learning_rate": 8e-05,
      "loss": 1.6992,
      "step": 5323
    },
    {
      "epoch": 0.5753809575272885,
      "grad_norm": 0.43792060017585754,
      "learning_rate": 8e-05,
      "loss": 1.5944,
      "step": 5324
    },
    {
      "epoch": 0.5754890305846753,
      "grad_norm": 0.4940316081047058,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 5325
    },
    {
      "epoch": 0.5755971036420621,
      "grad_norm": 0.4586353600025177,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 5326
    },
    {
      "epoch": 0.5757051766994489,
      "grad_norm": 0.450091689825058,
      "learning_rate": 8e-05,
      "loss": 1.6334,
      "step": 5327
    },
    {
      "epoch": 0.5758132497568356,
      "grad_norm": 0.4620334804058075,
      "learning_rate": 8e-05,
      "loss": 1.7192,
      "step": 5328
    },
    {
      "epoch": 0.5759213228142224,
      "grad_norm": 0.5022056698799133,
      "learning_rate": 8e-05,
      "loss": 2.004,
      "step": 5329
    },
    {
      "epoch": 0.5760293958716092,
      "grad_norm": 0.4766848385334015,
      "learning_rate": 8e-05,
      "loss": 1.8281,
      "step": 5330
    },
    {
      "epoch": 0.576137468928996,
      "grad_norm": 0.5320010781288147,
      "learning_rate": 8e-05,
      "loss": 1.6338,
      "step": 5331
    },
    {
      "epoch": 0.5762455419863828,
      "grad_norm": 0.46637195348739624,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 5332
    },
    {
      "epoch": 0.5763536150437696,
      "grad_norm": 0.4210820198059082,
      "learning_rate": 8e-05,
      "loss": 1.5194,
      "step": 5333
    },
    {
      "epoch": 0.5764616881011564,
      "grad_norm": 0.44604042172431946,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 5334
    },
    {
      "epoch": 0.5765697611585432,
      "grad_norm": 0.470411479473114,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 5335
    },
    {
      "epoch": 0.57667783421593,
      "grad_norm": 0.45458075404167175,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 5336
    },
    {
      "epoch": 0.5767859072733168,
      "grad_norm": 0.46313443779945374,
      "learning_rate": 8e-05,
      "loss": 1.6517,
      "step": 5337
    },
    {
      "epoch": 0.5768939803307036,
      "grad_norm": 0.44694337248802185,
      "learning_rate": 8e-05,
      "loss": 1.5789,
      "step": 5338
    },
    {
      "epoch": 0.5770020533880903,
      "grad_norm": 0.4227500557899475,
      "learning_rate": 8e-05,
      "loss": 1.6488,
      "step": 5339
    },
    {
      "epoch": 0.5771101264454771,
      "grad_norm": 0.439653605222702,
      "learning_rate": 8e-05,
      "loss": 1.5608,
      "step": 5340
    },
    {
      "epoch": 0.5772181995028639,
      "grad_norm": 0.45096445083618164,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 5341
    },
    {
      "epoch": 0.5773262725602507,
      "grad_norm": 0.47008517384529114,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 5342
    },
    {
      "epoch": 0.5774343456176375,
      "grad_norm": 0.4318113625049591,
      "learning_rate": 8e-05,
      "loss": 1.6553,
      "step": 5343
    },
    {
      "epoch": 0.5775424186750243,
      "grad_norm": 0.447519451379776,
      "learning_rate": 8e-05,
      "loss": 1.6138,
      "step": 5344
    },
    {
      "epoch": 0.5776504917324111,
      "grad_norm": 0.486003041267395,
      "learning_rate": 8e-05,
      "loss": 1.7838,
      "step": 5345
    },
    {
      "epoch": 0.5777585647897979,
      "grad_norm": 0.45907121896743774,
      "learning_rate": 8e-05,
      "loss": 1.7369,
      "step": 5346
    },
    {
      "epoch": 0.5778666378471847,
      "grad_norm": 0.48016831278800964,
      "learning_rate": 8e-05,
      "loss": 1.7263,
      "step": 5347
    },
    {
      "epoch": 0.5779747109045715,
      "grad_norm": 0.45631250739097595,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 5348
    },
    {
      "epoch": 0.5780827839619583,
      "grad_norm": 0.4412658214569092,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 5349
    },
    {
      "epoch": 0.578190857019345,
      "grad_norm": 0.44465696811676025,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 5350
    },
    {
      "epoch": 0.5782989300767318,
      "grad_norm": 0.4294421672821045,
      "learning_rate": 8e-05,
      "loss": 1.553,
      "step": 5351
    },
    {
      "epoch": 0.5784070031341186,
      "grad_norm": 0.41040274500846863,
      "learning_rate": 8e-05,
      "loss": 1.4247,
      "step": 5352
    },
    {
      "epoch": 0.5785150761915054,
      "grad_norm": 0.4736253023147583,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 5353
    },
    {
      "epoch": 0.5786231492488922,
      "grad_norm": 0.45290717482566833,
      "learning_rate": 8e-05,
      "loss": 1.5658,
      "step": 5354
    },
    {
      "epoch": 0.578731222306279,
      "grad_norm": 0.46385419368743896,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 5355
    },
    {
      "epoch": 0.5788392953636659,
      "grad_norm": 0.4136492908000946,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 5356
    },
    {
      "epoch": 0.5789473684210527,
      "grad_norm": 0.47213953733444214,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 5357
    },
    {
      "epoch": 0.5790554414784395,
      "grad_norm": 0.41935959458351135,
      "learning_rate": 8e-05,
      "loss": 1.5964,
      "step": 5358
    },
    {
      "epoch": 0.5791635145358263,
      "grad_norm": 0.44835758209228516,
      "learning_rate": 8e-05,
      "loss": 1.6461,
      "step": 5359
    },
    {
      "epoch": 0.5792715875932131,
      "grad_norm": 0.4415118396282196,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 5360
    },
    {
      "epoch": 0.5793796606505998,
      "grad_norm": 0.4567587971687317,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 5361
    },
    {
      "epoch": 0.5794877337079866,
      "grad_norm": 0.4314841330051422,
      "learning_rate": 8e-05,
      "loss": 1.71,
      "step": 5362
    },
    {
      "epoch": 0.5795958067653734,
      "grad_norm": 0.44142475724220276,
      "learning_rate": 8e-05,
      "loss": 1.5506,
      "step": 5363
    },
    {
      "epoch": 0.5797038798227602,
      "grad_norm": 0.4625411629676819,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 5364
    },
    {
      "epoch": 0.579811952880147,
      "grad_norm": 0.4402400553226471,
      "learning_rate": 8e-05,
      "loss": 1.7694,
      "step": 5365
    },
    {
      "epoch": 0.5799200259375338,
      "grad_norm": 0.43658265471458435,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 5366
    },
    {
      "epoch": 0.5800280989949206,
      "grad_norm": 0.5148639678955078,
      "learning_rate": 8e-05,
      "loss": 2.065,
      "step": 5367
    },
    {
      "epoch": 0.5801361720523074,
      "grad_norm": 0.4412710964679718,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 5368
    },
    {
      "epoch": 0.5802442451096942,
      "grad_norm": 0.47894611954689026,
      "learning_rate": 8e-05,
      "loss": 1.8489,
      "step": 5369
    },
    {
      "epoch": 0.580352318167081,
      "grad_norm": 0.46814611554145813,
      "learning_rate": 8e-05,
      "loss": 1.4833,
      "step": 5370
    },
    {
      "epoch": 0.5804603912244677,
      "grad_norm": 0.47676411271095276,
      "learning_rate": 8e-05,
      "loss": 1.7172,
      "step": 5371
    },
    {
      "epoch": 0.5805684642818545,
      "grad_norm": 0.4330938160419464,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 5372
    },
    {
      "epoch": 0.5806765373392413,
      "grad_norm": 0.3913879096508026,
      "learning_rate": 8e-05,
      "loss": 1.4953,
      "step": 5373
    },
    {
      "epoch": 0.5807846103966281,
      "grad_norm": 0.4616716802120209,
      "learning_rate": 8e-05,
      "loss": 1.7296,
      "step": 5374
    },
    {
      "epoch": 0.5808926834540149,
      "grad_norm": 0.48646944761276245,
      "learning_rate": 8e-05,
      "loss": 1.8404,
      "step": 5375
    },
    {
      "epoch": 0.5810007565114017,
      "grad_norm": 0.4614700675010681,
      "learning_rate": 8e-05,
      "loss": 1.5998,
      "step": 5376
    },
    {
      "epoch": 0.5811088295687885,
      "grad_norm": 0.4333561360836029,
      "learning_rate": 8e-05,
      "loss": 1.6853,
      "step": 5377
    },
    {
      "epoch": 0.5812169026261753,
      "grad_norm": 0.4430297911167145,
      "learning_rate": 8e-05,
      "loss": 1.6778,
      "step": 5378
    },
    {
      "epoch": 0.5813249756835621,
      "grad_norm": 0.4299125671386719,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 5379
    },
    {
      "epoch": 0.5814330487409489,
      "grad_norm": 0.4476819932460785,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 5380
    },
    {
      "epoch": 0.5815411217983357,
      "grad_norm": 0.44013336300849915,
      "learning_rate": 8e-05,
      "loss": 1.6158,
      "step": 5381
    },
    {
      "epoch": 0.5816491948557224,
      "grad_norm": 0.43739473819732666,
      "learning_rate": 8e-05,
      "loss": 1.6116,
      "step": 5382
    },
    {
      "epoch": 0.5817572679131092,
      "grad_norm": 0.4678393006324768,
      "learning_rate": 8e-05,
      "loss": 1.738,
      "step": 5383
    },
    {
      "epoch": 0.581865340970496,
      "grad_norm": 0.4316672086715698,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 5384
    },
    {
      "epoch": 0.5819734140278828,
      "grad_norm": 0.435238242149353,
      "learning_rate": 8e-05,
      "loss": 1.5838,
      "step": 5385
    },
    {
      "epoch": 0.5820814870852696,
      "grad_norm": 0.45582816004753113,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 5386
    },
    {
      "epoch": 0.5821895601426564,
      "grad_norm": 0.418997198343277,
      "learning_rate": 8e-05,
      "loss": 1.5475,
      "step": 5387
    },
    {
      "epoch": 0.5822976332000432,
      "grad_norm": 0.42782124876976013,
      "learning_rate": 8e-05,
      "loss": 1.6267,
      "step": 5388
    },
    {
      "epoch": 0.58240570625743,
      "grad_norm": 0.5231521725654602,
      "learning_rate": 8e-05,
      "loss": 1.88,
      "step": 5389
    },
    {
      "epoch": 0.5825137793148168,
      "grad_norm": 0.43132421374320984,
      "learning_rate": 8e-05,
      "loss": 1.4919,
      "step": 5390
    },
    {
      "epoch": 0.5826218523722037,
      "grad_norm": 0.4598361849784851,
      "learning_rate": 8e-05,
      "loss": 1.5672,
      "step": 5391
    },
    {
      "epoch": 0.5827299254295905,
      "grad_norm": 0.4163326025009155,
      "learning_rate": 8e-05,
      "loss": 1.5282,
      "step": 5392
    },
    {
      "epoch": 0.5828379984869771,
      "grad_norm": 0.42676079273223877,
      "learning_rate": 8e-05,
      "loss": 1.5105,
      "step": 5393
    },
    {
      "epoch": 0.582946071544364,
      "grad_norm": 0.4402390420436859,
      "learning_rate": 8e-05,
      "loss": 1.6213,
      "step": 5394
    },
    {
      "epoch": 0.5830541446017508,
      "grad_norm": 0.4507570266723633,
      "learning_rate": 8e-05,
      "loss": 1.5688,
      "step": 5395
    },
    {
      "epoch": 0.5831622176591376,
      "grad_norm": 0.4449840188026428,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 5396
    },
    {
      "epoch": 0.5832702907165244,
      "grad_norm": 0.4511798322200775,
      "learning_rate": 8e-05,
      "loss": 1.6168,
      "step": 5397
    },
    {
      "epoch": 0.5833783637739112,
      "grad_norm": 0.4679395258426666,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 5398
    },
    {
      "epoch": 0.583486436831298,
      "grad_norm": 0.4895848035812378,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 5399
    },
    {
      "epoch": 0.5835945098886848,
      "grad_norm": 0.44089803099632263,
      "learning_rate": 8e-05,
      "loss": 1.6219,
      "step": 5400
    },
    {
      "epoch": 0.5837025829460716,
      "grad_norm": 0.45725780725479126,
      "learning_rate": 8e-05,
      "loss": 1.6541,
      "step": 5401
    },
    {
      "epoch": 0.5838106560034584,
      "grad_norm": 0.4796881079673767,
      "learning_rate": 8e-05,
      "loss": 1.6892,
      "step": 5402
    },
    {
      "epoch": 0.5839187290608451,
      "grad_norm": 0.4771806299686432,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 5403
    },
    {
      "epoch": 0.5840268021182319,
      "grad_norm": 0.4477314054965973,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 5404
    },
    {
      "epoch": 0.5841348751756187,
      "grad_norm": 0.5631340742111206,
      "learning_rate": 8e-05,
      "loss": 1.9507,
      "step": 5405
    },
    {
      "epoch": 0.5842429482330055,
      "grad_norm": 0.448894739151001,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 5406
    },
    {
      "epoch": 0.5843510212903923,
      "grad_norm": 0.4527323842048645,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 5407
    },
    {
      "epoch": 0.5844590943477791,
      "grad_norm": 0.46861597895622253,
      "learning_rate": 8e-05,
      "loss": 1.6572,
      "step": 5408
    },
    {
      "epoch": 0.5845671674051659,
      "grad_norm": 0.46755337715148926,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 5409
    },
    {
      "epoch": 0.5846752404625527,
      "grad_norm": 0.4251028895378113,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 5410
    },
    {
      "epoch": 0.5847833135199395,
      "grad_norm": 0.4343346953392029,
      "learning_rate": 8e-05,
      "loss": 1.6146,
      "step": 5411
    },
    {
      "epoch": 0.5848913865773263,
      "grad_norm": 0.44779694080352783,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 5412
    },
    {
      "epoch": 0.5849994596347131,
      "grad_norm": 0.432643324136734,
      "learning_rate": 8e-05,
      "loss": 1.82,
      "step": 5413
    },
    {
      "epoch": 0.5851075326920998,
      "grad_norm": 0.4284329116344452,
      "learning_rate": 8e-05,
      "loss": 1.5874,
      "step": 5414
    },
    {
      "epoch": 0.5852156057494866,
      "grad_norm": 0.509408175945282,
      "learning_rate": 8e-05,
      "loss": 1.6373,
      "step": 5415
    },
    {
      "epoch": 0.5853236788068734,
      "grad_norm": 0.557856559753418,
      "learning_rate": 8e-05,
      "loss": 1.9994,
      "step": 5416
    },
    {
      "epoch": 0.5854317518642602,
      "grad_norm": 0.5096309781074524,
      "learning_rate": 8e-05,
      "loss": 1.9702,
      "step": 5417
    },
    {
      "epoch": 0.585539824921647,
      "grad_norm": 0.4392206072807312,
      "learning_rate": 8e-05,
      "loss": 1.7043,
      "step": 5418
    },
    {
      "epoch": 0.5856478979790338,
      "grad_norm": 0.4567457139492035,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 5419
    },
    {
      "epoch": 0.5857559710364206,
      "grad_norm": 0.4543382227420807,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 5420
    },
    {
      "epoch": 0.5858640440938074,
      "grad_norm": 0.456405371427536,
      "learning_rate": 8e-05,
      "loss": 1.7829,
      "step": 5421
    },
    {
      "epoch": 0.5859721171511942,
      "grad_norm": 0.4807143211364746,
      "learning_rate": 8e-05,
      "loss": 1.8781,
      "step": 5422
    },
    {
      "epoch": 0.586080190208581,
      "grad_norm": 0.46283021569252014,
      "learning_rate": 8e-05,
      "loss": 1.7915,
      "step": 5423
    },
    {
      "epoch": 0.5861882632659678,
      "grad_norm": 0.4957137107849121,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 5424
    },
    {
      "epoch": 0.5862963363233545,
      "grad_norm": 0.4584107995033264,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 5425
    },
    {
      "epoch": 0.5864044093807413,
      "grad_norm": 0.4376767575740814,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 5426
    },
    {
      "epoch": 0.5865124824381281,
      "grad_norm": 0.4491543471813202,
      "learning_rate": 8e-05,
      "loss": 1.6926,
      "step": 5427
    },
    {
      "epoch": 0.586620555495515,
      "grad_norm": 0.44645845890045166,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 5428
    },
    {
      "epoch": 0.5867286285529018,
      "grad_norm": 0.4767771065235138,
      "learning_rate": 8e-05,
      "loss": 1.7653,
      "step": 5429
    },
    {
      "epoch": 0.5868367016102886,
      "grad_norm": 0.4638815224170685,
      "learning_rate": 8e-05,
      "loss": 1.756,
      "step": 5430
    },
    {
      "epoch": 0.5869447746676754,
      "grad_norm": 0.5069172978401184,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 5431
    },
    {
      "epoch": 0.5870528477250622,
      "grad_norm": 0.4385349154472351,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 5432
    },
    {
      "epoch": 0.587160920782449,
      "grad_norm": 0.4453005790710449,
      "learning_rate": 8e-05,
      "loss": 1.766,
      "step": 5433
    },
    {
      "epoch": 0.5872689938398358,
      "grad_norm": 0.428234726190567,
      "learning_rate": 8e-05,
      "loss": 1.5636,
      "step": 5434
    },
    {
      "epoch": 0.5873770668972226,
      "grad_norm": 0.4636060297489166,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 5435
    },
    {
      "epoch": 0.5874851399546093,
      "grad_norm": 0.43749555945396423,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 5436
    },
    {
      "epoch": 0.5875932130119961,
      "grad_norm": 0.45799723267555237,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 5437
    },
    {
      "epoch": 0.5877012860693829,
      "grad_norm": 0.4748808741569519,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 5438
    },
    {
      "epoch": 0.5878093591267697,
      "grad_norm": 0.4963321387767792,
      "learning_rate": 8e-05,
      "loss": 1.8308,
      "step": 5439
    },
    {
      "epoch": 0.5879174321841565,
      "grad_norm": 0.449841171503067,
      "learning_rate": 8e-05,
      "loss": 1.7091,
      "step": 5440
    },
    {
      "epoch": 0.5880255052415433,
      "grad_norm": 0.47024503350257874,
      "learning_rate": 8e-05,
      "loss": 1.8102,
      "step": 5441
    },
    {
      "epoch": 0.5881335782989301,
      "grad_norm": 0.5590707063674927,
      "learning_rate": 8e-05,
      "loss": 1.8367,
      "step": 5442
    },
    {
      "epoch": 0.5882416513563169,
      "grad_norm": 0.43544524908065796,
      "learning_rate": 8e-05,
      "loss": 1.6689,
      "step": 5443
    },
    {
      "epoch": 0.5883497244137037,
      "grad_norm": 0.4395105540752411,
      "learning_rate": 8e-05,
      "loss": 1.6073,
      "step": 5444
    },
    {
      "epoch": 0.5884577974710905,
      "grad_norm": 0.46062207221984863,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 5445
    },
    {
      "epoch": 0.5885658705284772,
      "grad_norm": 0.4263331890106201,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 5446
    },
    {
      "epoch": 0.588673943585864,
      "grad_norm": 0.47889626026153564,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 5447
    },
    {
      "epoch": 0.5887820166432508,
      "grad_norm": 0.4153454005718231,
      "learning_rate": 8e-05,
      "loss": 1.4535,
      "step": 5448
    },
    {
      "epoch": 0.5888900897006376,
      "grad_norm": 0.43225550651550293,
      "learning_rate": 8e-05,
      "loss": 1.4733,
      "step": 5449
    },
    {
      "epoch": 0.5889981627580244,
      "grad_norm": 0.4397936463356018,
      "learning_rate": 8e-05,
      "loss": 1.6925,
      "step": 5450
    },
    {
      "epoch": 0.5891062358154112,
      "grad_norm": 0.47318634390830994,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 5451
    },
    {
      "epoch": 0.589214308872798,
      "grad_norm": 0.4619483947753906,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 5452
    },
    {
      "epoch": 0.5893223819301848,
      "grad_norm": 0.45720791816711426,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 5453
    },
    {
      "epoch": 0.5894304549875716,
      "grad_norm": 0.4066876173019409,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 5454
    },
    {
      "epoch": 0.5895385280449584,
      "grad_norm": 0.4802696406841278,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 5455
    },
    {
      "epoch": 0.5896466011023452,
      "grad_norm": 0.45702221989631653,
      "learning_rate": 8e-05,
      "loss": 1.797,
      "step": 5456
    },
    {
      "epoch": 0.5897546741597319,
      "grad_norm": 0.4460969567298889,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 5457
    },
    {
      "epoch": 0.5898627472171187,
      "grad_norm": 0.5050446391105652,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 5458
    },
    {
      "epoch": 0.5899708202745055,
      "grad_norm": 0.5020354390144348,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 5459
    },
    {
      "epoch": 0.5900788933318923,
      "grad_norm": 0.47807377576828003,
      "learning_rate": 8e-05,
      "loss": 1.6561,
      "step": 5460
    },
    {
      "epoch": 0.5901869663892791,
      "grad_norm": 0.4683237373828888,
      "learning_rate": 8e-05,
      "loss": 1.7398,
      "step": 5461
    },
    {
      "epoch": 0.590295039446666,
      "grad_norm": 0.460338294506073,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 5462
    },
    {
      "epoch": 0.5904031125040528,
      "grad_norm": 0.45391228795051575,
      "learning_rate": 8e-05,
      "loss": 1.6858,
      "step": 5463
    },
    {
      "epoch": 0.5905111855614396,
      "grad_norm": 0.4856325387954712,
      "learning_rate": 8e-05,
      "loss": 1.8265,
      "step": 5464
    },
    {
      "epoch": 0.5906192586188264,
      "grad_norm": 0.46313026547431946,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 5465
    },
    {
      "epoch": 0.5907273316762132,
      "grad_norm": 0.4197196960449219,
      "learning_rate": 8e-05,
      "loss": 1.5502,
      "step": 5466
    },
    {
      "epoch": 0.5908354047336,
      "grad_norm": 0.45119696855545044,
      "learning_rate": 8e-05,
      "loss": 1.6569,
      "step": 5467
    },
    {
      "epoch": 0.5909434777909867,
      "grad_norm": 0.4836268424987793,
      "learning_rate": 8e-05,
      "loss": 1.832,
      "step": 5468
    },
    {
      "epoch": 0.5910515508483735,
      "grad_norm": 0.49243855476379395,
      "learning_rate": 8e-05,
      "loss": 1.7888,
      "step": 5469
    },
    {
      "epoch": 0.5911596239057603,
      "grad_norm": 0.4855518937110901,
      "learning_rate": 8e-05,
      "loss": 1.7371,
      "step": 5470
    },
    {
      "epoch": 0.5912676969631471,
      "grad_norm": 0.48916488885879517,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 5471
    },
    {
      "epoch": 0.5913757700205339,
      "grad_norm": 0.5016804933547974,
      "learning_rate": 8e-05,
      "loss": 1.8572,
      "step": 5472
    },
    {
      "epoch": 0.5914838430779207,
      "grad_norm": 0.4373093545436859,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 5473
    },
    {
      "epoch": 0.5915919161353075,
      "grad_norm": 0.5208323001861572,
      "learning_rate": 8e-05,
      "loss": 1.7705,
      "step": 5474
    },
    {
      "epoch": 0.5916999891926943,
      "grad_norm": 0.4394699037075043,
      "learning_rate": 8e-05,
      "loss": 1.5741,
      "step": 5475
    },
    {
      "epoch": 0.5918080622500811,
      "grad_norm": 0.4683206379413605,
      "learning_rate": 8e-05,
      "loss": 1.5597,
      "step": 5476
    },
    {
      "epoch": 0.5919161353074679,
      "grad_norm": 0.44892555475234985,
      "learning_rate": 8e-05,
      "loss": 1.5813,
      "step": 5477
    },
    {
      "epoch": 0.5920242083648546,
      "grad_norm": 0.4269934296607971,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 5478
    },
    {
      "epoch": 0.5921322814222414,
      "grad_norm": 0.5375174283981323,
      "learning_rate": 8e-05,
      "loss": 1.8593,
      "step": 5479
    },
    {
      "epoch": 0.5922403544796282,
      "grad_norm": 0.47564980387687683,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 5480
    },
    {
      "epoch": 0.592348427537015,
      "grad_norm": 0.45079874992370605,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 5481
    },
    {
      "epoch": 0.5924565005944018,
      "grad_norm": 0.41649505496025085,
      "learning_rate": 8e-05,
      "loss": 1.6008,
      "step": 5482
    },
    {
      "epoch": 0.5925645736517886,
      "grad_norm": 0.43143442273139954,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 5483
    },
    {
      "epoch": 0.5926726467091754,
      "grad_norm": 0.46150705218315125,
      "learning_rate": 8e-05,
      "loss": 1.7374,
      "step": 5484
    },
    {
      "epoch": 0.5927807197665622,
      "grad_norm": 0.4498525857925415,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 5485
    },
    {
      "epoch": 0.592888792823949,
      "grad_norm": 0.46347808837890625,
      "learning_rate": 8e-05,
      "loss": 1.8065,
      "step": 5486
    },
    {
      "epoch": 0.5929968658813358,
      "grad_norm": 0.4893917739391327,
      "learning_rate": 8e-05,
      "loss": 1.7874,
      "step": 5487
    },
    {
      "epoch": 0.5931049389387226,
      "grad_norm": 0.47290468215942383,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 5488
    },
    {
      "epoch": 0.5932130119961093,
      "grad_norm": 0.48606622219085693,
      "learning_rate": 8e-05,
      "loss": 1.8614,
      "step": 5489
    },
    {
      "epoch": 0.5933210850534961,
      "grad_norm": 0.49145811796188354,
      "learning_rate": 8e-05,
      "loss": 1.9145,
      "step": 5490
    },
    {
      "epoch": 0.5934291581108829,
      "grad_norm": 0.42695239186286926,
      "learning_rate": 8e-05,
      "loss": 1.6189,
      "step": 5491
    },
    {
      "epoch": 0.5935372311682697,
      "grad_norm": 0.4368320107460022,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 5492
    },
    {
      "epoch": 0.5936453042256565,
      "grad_norm": 0.4536571204662323,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 5493
    },
    {
      "epoch": 0.5937533772830433,
      "grad_norm": 0.4236263334751129,
      "learning_rate": 8e-05,
      "loss": 1.5117,
      "step": 5494
    },
    {
      "epoch": 0.5938614503404301,
      "grad_norm": 0.43621453642845154,
      "learning_rate": 8e-05,
      "loss": 1.6648,
      "step": 5495
    },
    {
      "epoch": 0.593969523397817,
      "grad_norm": 0.43696990609169006,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 5496
    },
    {
      "epoch": 0.5940775964552037,
      "grad_norm": 0.47802916169166565,
      "learning_rate": 8e-05,
      "loss": 1.7773,
      "step": 5497
    },
    {
      "epoch": 0.5941856695125906,
      "grad_norm": 0.4521600902080536,
      "learning_rate": 8e-05,
      "loss": 1.6308,
      "step": 5498
    },
    {
      "epoch": 0.5942937425699774,
      "grad_norm": 0.43578726053237915,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 5499
    },
    {
      "epoch": 0.594401815627364,
      "grad_norm": 0.45497050881385803,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 5500
    },
    {
      "epoch": 0.5945098886847509,
      "grad_norm": 0.46041977405548096,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 5501
    },
    {
      "epoch": 0.5946179617421377,
      "grad_norm": 0.4884103536605835,
      "learning_rate": 8e-05,
      "loss": 1.7705,
      "step": 5502
    },
    {
      "epoch": 0.5947260347995245,
      "grad_norm": 0.4735165238380432,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 5503
    },
    {
      "epoch": 0.5948341078569113,
      "grad_norm": 0.41975903511047363,
      "learning_rate": 8e-05,
      "loss": 1.545,
      "step": 5504
    },
    {
      "epoch": 0.5949421809142981,
      "grad_norm": 0.4886011481285095,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 5505
    },
    {
      "epoch": 0.5950502539716849,
      "grad_norm": 0.5399619936943054,
      "learning_rate": 8e-05,
      "loss": 1.8611,
      "step": 5506
    },
    {
      "epoch": 0.5951583270290717,
      "grad_norm": 0.4546045660972595,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 5507
    },
    {
      "epoch": 0.5952664000864585,
      "grad_norm": 0.49118444323539734,
      "learning_rate": 8e-05,
      "loss": 1.909,
      "step": 5508
    },
    {
      "epoch": 0.5953744731438453,
      "grad_norm": 0.4795960783958435,
      "learning_rate": 8e-05,
      "loss": 1.6022,
      "step": 5509
    },
    {
      "epoch": 0.5954825462012321,
      "grad_norm": 0.4376899003982544,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 5510
    },
    {
      "epoch": 0.5955906192586188,
      "grad_norm": 0.4274897873401642,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 5511
    },
    {
      "epoch": 0.5956986923160056,
      "grad_norm": 0.5143346190452576,
      "learning_rate": 8e-05,
      "loss": 1.9109,
      "step": 5512
    },
    {
      "epoch": 0.5958067653733924,
      "grad_norm": 0.5070353746414185,
      "learning_rate": 8e-05,
      "loss": 1.8688,
      "step": 5513
    },
    {
      "epoch": 0.5959148384307792,
      "grad_norm": 0.4658992290496826,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 5514
    },
    {
      "epoch": 0.596022911488166,
      "grad_norm": 0.5206338763237,
      "learning_rate": 8e-05,
      "loss": 1.6687,
      "step": 5515
    },
    {
      "epoch": 0.5961309845455528,
      "grad_norm": 0.47798627614974976,
      "learning_rate": 8e-05,
      "loss": 1.8364,
      "step": 5516
    },
    {
      "epoch": 0.5962390576029396,
      "grad_norm": 0.48321273922920227,
      "learning_rate": 8e-05,
      "loss": 1.8055,
      "step": 5517
    },
    {
      "epoch": 0.5963471306603264,
      "grad_norm": 0.43605631589889526,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 5518
    },
    {
      "epoch": 0.5964552037177132,
      "grad_norm": 0.46028855443000793,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 5519
    },
    {
      "epoch": 0.5965632767751,
      "grad_norm": 0.510981559753418,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 5520
    },
    {
      "epoch": 0.5966713498324867,
      "grad_norm": 0.4708881974220276,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 5521
    },
    {
      "epoch": 0.5967794228898735,
      "grad_norm": 0.4599258601665497,
      "learning_rate": 8e-05,
      "loss": 1.7366,
      "step": 5522
    },
    {
      "epoch": 0.5968874959472603,
      "grad_norm": 0.4705546498298645,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 5523
    },
    {
      "epoch": 0.5969955690046471,
      "grad_norm": 0.4217449724674225,
      "learning_rate": 8e-05,
      "loss": 1.5106,
      "step": 5524
    },
    {
      "epoch": 0.5971036420620339,
      "grad_norm": 0.4518130123615265,
      "learning_rate": 8e-05,
      "loss": 1.6616,
      "step": 5525
    },
    {
      "epoch": 0.5972117151194207,
      "grad_norm": 0.45554104447364807,
      "learning_rate": 8e-05,
      "loss": 1.6302,
      "step": 5526
    },
    {
      "epoch": 0.5973197881768075,
      "grad_norm": 0.47657129168510437,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 5527
    },
    {
      "epoch": 0.5974278612341943,
      "grad_norm": 0.46889933943748474,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 5528
    },
    {
      "epoch": 0.5975359342915811,
      "grad_norm": 0.45156508684158325,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 5529
    },
    {
      "epoch": 0.5976440073489679,
      "grad_norm": 0.4643718898296356,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 5530
    },
    {
      "epoch": 0.5977520804063547,
      "grad_norm": 0.4910826086997986,
      "learning_rate": 8e-05,
      "loss": 1.8517,
      "step": 5531
    },
    {
      "epoch": 0.5978601534637414,
      "grad_norm": 0.4594413936138153,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 5532
    },
    {
      "epoch": 0.5979682265211282,
      "grad_norm": 0.442728728055954,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 5533
    },
    {
      "epoch": 0.598076299578515,
      "grad_norm": 0.4397749602794647,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 5534
    },
    {
      "epoch": 0.5981843726359019,
      "grad_norm": 0.4432555139064789,
      "learning_rate": 8e-05,
      "loss": 1.7965,
      "step": 5535
    },
    {
      "epoch": 0.5982924456932887,
      "grad_norm": 0.4619430899620056,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 5536
    },
    {
      "epoch": 0.5984005187506755,
      "grad_norm": 0.41755831241607666,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 5537
    },
    {
      "epoch": 0.5985085918080623,
      "grad_norm": 0.5237672328948975,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 5538
    },
    {
      "epoch": 0.5986166648654491,
      "grad_norm": 0.46269693970680237,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 5539
    },
    {
      "epoch": 0.5987247379228359,
      "grad_norm": 0.5550700426101685,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 5540
    },
    {
      "epoch": 0.5988328109802227,
      "grad_norm": 0.47571927309036255,
      "learning_rate": 8e-05,
      "loss": 1.8148,
      "step": 5541
    },
    {
      "epoch": 0.5989408840376095,
      "grad_norm": 0.44056618213653564,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 5542
    },
    {
      "epoch": 0.5990489570949962,
      "grad_norm": 0.47720867395401,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 5543
    },
    {
      "epoch": 0.599157030152383,
      "grad_norm": 0.4912940263748169,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 5544
    },
    {
      "epoch": 0.5992651032097698,
      "grad_norm": 0.46885576844215393,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 5545
    },
    {
      "epoch": 0.5993731762671566,
      "grad_norm": 0.4718937575817108,
      "learning_rate": 8e-05,
      "loss": 1.7965,
      "step": 5546
    },
    {
      "epoch": 0.5994812493245434,
      "grad_norm": 0.5152108073234558,
      "learning_rate": 8e-05,
      "loss": 1.9146,
      "step": 5547
    },
    {
      "epoch": 0.5995893223819302,
      "grad_norm": 0.4348636865615845,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 5548
    },
    {
      "epoch": 0.599697395439317,
      "grad_norm": 0.47381797432899475,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 5549
    },
    {
      "epoch": 0.5998054684967038,
      "grad_norm": 0.4455678164958954,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 5550
    },
    {
      "epoch": 0.5999135415540906,
      "grad_norm": 0.48109400272369385,
      "learning_rate": 8e-05,
      "loss": 1.7204,
      "step": 5551
    },
    {
      "epoch": 0.6000216146114774,
      "grad_norm": 0.46083611249923706,
      "learning_rate": 8e-05,
      "loss": 1.805,
      "step": 5552
    },
    {
      "epoch": 0.6001296876688641,
      "grad_norm": 0.4300784468650818,
      "learning_rate": 8e-05,
      "loss": 1.5176,
      "step": 5553
    },
    {
      "epoch": 0.6002377607262509,
      "grad_norm": 0.4481351375579834,
      "learning_rate": 8e-05,
      "loss": 1.7435,
      "step": 5554
    },
    {
      "epoch": 0.6003458337836377,
      "grad_norm": 0.4938361644744873,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 5555
    },
    {
      "epoch": 0.6004539068410245,
      "grad_norm": 0.4823426604270935,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 5556
    },
    {
      "epoch": 0.6005619798984113,
      "grad_norm": 0.4543304145336151,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 5557
    },
    {
      "epoch": 0.6006700529557981,
      "grad_norm": 0.43063241243362427,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 5558
    },
    {
      "epoch": 0.6007781260131849,
      "grad_norm": 0.4758738577365875,
      "learning_rate": 8e-05,
      "loss": 1.5624,
      "step": 5559
    },
    {
      "epoch": 0.6008861990705717,
      "grad_norm": 0.47377416491508484,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 5560
    },
    {
      "epoch": 0.6009942721279585,
      "grad_norm": 0.462771475315094,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 5561
    },
    {
      "epoch": 0.6011023451853453,
      "grad_norm": 0.4375857412815094,
      "learning_rate": 8e-05,
      "loss": 1.5513,
      "step": 5562
    },
    {
      "epoch": 0.6012104182427321,
      "grad_norm": 0.47146275639533997,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 5563
    },
    {
      "epoch": 0.6013184913001188,
      "grad_norm": 0.4637908339500427,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 5564
    },
    {
      "epoch": 0.6014265643575056,
      "grad_norm": 0.44560885429382324,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 5565
    },
    {
      "epoch": 0.6015346374148924,
      "grad_norm": 0.47084662318229675,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 5566
    },
    {
      "epoch": 0.6016427104722792,
      "grad_norm": 0.4454510807991028,
      "learning_rate": 8e-05,
      "loss": 1.7157,
      "step": 5567
    },
    {
      "epoch": 0.601750783529666,
      "grad_norm": 0.4750213623046875,
      "learning_rate": 8e-05,
      "loss": 1.5186,
      "step": 5568
    },
    {
      "epoch": 0.6018588565870528,
      "grad_norm": 0.4305649995803833,
      "learning_rate": 8e-05,
      "loss": 1.5653,
      "step": 5569
    },
    {
      "epoch": 0.6019669296444397,
      "grad_norm": 0.4577915668487549,
      "learning_rate": 8e-05,
      "loss": 1.618,
      "step": 5570
    },
    {
      "epoch": 0.6020750027018265,
      "grad_norm": 0.4173718988895416,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 5571
    },
    {
      "epoch": 0.6021830757592133,
      "grad_norm": 0.4500075876712799,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 5572
    },
    {
      "epoch": 0.6022911488166001,
      "grad_norm": 0.47105881571769714,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 5573
    },
    {
      "epoch": 0.6023992218739869,
      "grad_norm": 0.47047746181488037,
      "learning_rate": 8e-05,
      "loss": 1.8327,
      "step": 5574
    },
    {
      "epoch": 0.6025072949313736,
      "grad_norm": 0.4290904402732849,
      "learning_rate": 8e-05,
      "loss": 1.6698,
      "step": 5575
    },
    {
      "epoch": 0.6026153679887604,
      "grad_norm": 0.43327367305755615,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 5576
    },
    {
      "epoch": 0.6027234410461472,
      "grad_norm": 0.43265700340270996,
      "learning_rate": 8e-05,
      "loss": 1.7456,
      "step": 5577
    },
    {
      "epoch": 0.602831514103534,
      "grad_norm": 0.441648930311203,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 5578
    },
    {
      "epoch": 0.6029395871609208,
      "grad_norm": 0.47210314869880676,
      "learning_rate": 8e-05,
      "loss": 1.7664,
      "step": 5579
    },
    {
      "epoch": 0.6030476602183076,
      "grad_norm": 0.41860827803611755,
      "learning_rate": 8e-05,
      "loss": 1.5122,
      "step": 5580
    },
    {
      "epoch": 0.6031557332756944,
      "grad_norm": 0.44679826498031616,
      "learning_rate": 8e-05,
      "loss": 1.6648,
      "step": 5581
    },
    {
      "epoch": 0.6032638063330812,
      "grad_norm": 0.45806702971458435,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 5582
    },
    {
      "epoch": 0.603371879390468,
      "grad_norm": 0.48387739062309265,
      "learning_rate": 8e-05,
      "loss": 1.8019,
      "step": 5583
    },
    {
      "epoch": 0.6034799524478548,
      "grad_norm": 0.4534894526004791,
      "learning_rate": 8e-05,
      "loss": 1.6574,
      "step": 5584
    },
    {
      "epoch": 0.6035880255052415,
      "grad_norm": 0.4613317847251892,
      "learning_rate": 8e-05,
      "loss": 1.5832,
      "step": 5585
    },
    {
      "epoch": 0.6036960985626283,
      "grad_norm": 0.47045212984085083,
      "learning_rate": 8e-05,
      "loss": 1.8001,
      "step": 5586
    },
    {
      "epoch": 0.6038041716200151,
      "grad_norm": 0.48350730538368225,
      "learning_rate": 8e-05,
      "loss": 1.8211,
      "step": 5587
    },
    {
      "epoch": 0.6039122446774019,
      "grad_norm": 0.48400986194610596,
      "learning_rate": 8e-05,
      "loss": 1.7763,
      "step": 5588
    },
    {
      "epoch": 0.6040203177347887,
      "grad_norm": 0.48703670501708984,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 5589
    },
    {
      "epoch": 0.6041283907921755,
      "grad_norm": 0.4657638370990753,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 5590
    },
    {
      "epoch": 0.6042364638495623,
      "grad_norm": 0.4672345221042633,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 5591
    },
    {
      "epoch": 0.6043445369069491,
      "grad_norm": 0.4663335084915161,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 5592
    },
    {
      "epoch": 0.6044526099643359,
      "grad_norm": 0.5408346056938171,
      "learning_rate": 8e-05,
      "loss": 1.8595,
      "step": 5593
    },
    {
      "epoch": 0.6045606830217227,
      "grad_norm": 0.4662058353424072,
      "learning_rate": 8e-05,
      "loss": 1.8564,
      "step": 5594
    },
    {
      "epoch": 0.6046687560791095,
      "grad_norm": 0.5240734815597534,
      "learning_rate": 8e-05,
      "loss": 2.0592,
      "step": 5595
    },
    {
      "epoch": 0.6047768291364962,
      "grad_norm": 0.47205379605293274,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 5596
    },
    {
      "epoch": 0.604884902193883,
      "grad_norm": 0.4636024534702301,
      "learning_rate": 8e-05,
      "loss": 1.6606,
      "step": 5597
    },
    {
      "epoch": 0.6049929752512698,
      "grad_norm": 0.42031776905059814,
      "learning_rate": 8e-05,
      "loss": 1.5028,
      "step": 5598
    },
    {
      "epoch": 0.6051010483086566,
      "grad_norm": 0.45109105110168457,
      "learning_rate": 8e-05,
      "loss": 1.4294,
      "step": 5599
    },
    {
      "epoch": 0.6052091213660434,
      "grad_norm": 0.4911045432090759,
      "learning_rate": 8e-05,
      "loss": 1.8946,
      "step": 5600
    },
    {
      "epoch": 0.6053171944234302,
      "grad_norm": 0.4353826642036438,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 5601
    },
    {
      "epoch": 0.605425267480817,
      "grad_norm": 0.4261130392551422,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 5602
    },
    {
      "epoch": 0.6055333405382038,
      "grad_norm": 0.5044853687286377,
      "learning_rate": 8e-05,
      "loss": 1.8383,
      "step": 5603
    },
    {
      "epoch": 0.6056414135955906,
      "grad_norm": 0.4478701651096344,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 5604
    },
    {
      "epoch": 0.6057494866529775,
      "grad_norm": 0.43872424960136414,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 5605
    },
    {
      "epoch": 0.6058575597103643,
      "grad_norm": 0.45928430557250977,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 5606
    },
    {
      "epoch": 0.605965632767751,
      "grad_norm": 0.4374661147594452,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 5607
    },
    {
      "epoch": 0.6060737058251378,
      "grad_norm": 0.4480030834674835,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 5608
    },
    {
      "epoch": 0.6061817788825246,
      "grad_norm": 0.49852874875068665,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 5609
    },
    {
      "epoch": 0.6062898519399114,
      "grad_norm": 0.4471527338027954,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 5610
    },
    {
      "epoch": 0.6063979249972982,
      "grad_norm": 0.44876629114151,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 5611
    },
    {
      "epoch": 0.606505998054685,
      "grad_norm": 0.45334723591804504,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 5612
    },
    {
      "epoch": 0.6066140711120718,
      "grad_norm": 0.4400753080844879,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 5613
    },
    {
      "epoch": 0.6067221441694586,
      "grad_norm": 0.4482342004776001,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 5614
    },
    {
      "epoch": 0.6068302172268454,
      "grad_norm": 0.44942206144332886,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 5615
    },
    {
      "epoch": 0.6069382902842322,
      "grad_norm": 0.44073784351348877,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 5616
    },
    {
      "epoch": 0.607046363341619,
      "grad_norm": 0.48111796379089355,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 5617
    },
    {
      "epoch": 0.6071544363990057,
      "grad_norm": 0.4740293323993683,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 5618
    },
    {
      "epoch": 0.6072625094563925,
      "grad_norm": 0.4592937231063843,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 5619
    },
    {
      "epoch": 0.6073705825137793,
      "grad_norm": 0.4932771623134613,
      "learning_rate": 8e-05,
      "loss": 1.9273,
      "step": 5620
    },
    {
      "epoch": 0.6074786555711661,
      "grad_norm": 0.4331694543361664,
      "learning_rate": 8e-05,
      "loss": 1.5885,
      "step": 5621
    },
    {
      "epoch": 0.6075867286285529,
      "grad_norm": 0.43562158942222595,
      "learning_rate": 8e-05,
      "loss": 1.7391,
      "step": 5622
    },
    {
      "epoch": 0.6076948016859397,
      "grad_norm": 0.4714029133319855,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 5623
    },
    {
      "epoch": 0.6078028747433265,
      "grad_norm": 0.4526682496070862,
      "learning_rate": 8e-05,
      "loss": 1.6197,
      "step": 5624
    },
    {
      "epoch": 0.6079109478007133,
      "grad_norm": 0.4675043225288391,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 5625
    },
    {
      "epoch": 0.6080190208581001,
      "grad_norm": 0.47051605582237244,
      "learning_rate": 8e-05,
      "loss": 1.8133,
      "step": 5626
    },
    {
      "epoch": 0.6081270939154869,
      "grad_norm": 0.45482879877090454,
      "learning_rate": 8e-05,
      "loss": 1.6455,
      "step": 5627
    },
    {
      "epoch": 0.6082351669728736,
      "grad_norm": 0.5210318565368652,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 5628
    },
    {
      "epoch": 0.6083432400302604,
      "grad_norm": 0.4596984386444092,
      "learning_rate": 8e-05,
      "loss": 1.7752,
      "step": 5629
    },
    {
      "epoch": 0.6084513130876472,
      "grad_norm": 0.4441293179988861,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 5630
    },
    {
      "epoch": 0.608559386145034,
      "grad_norm": 0.5124384164810181,
      "learning_rate": 8e-05,
      "loss": 1.9068,
      "step": 5631
    },
    {
      "epoch": 0.6086674592024208,
      "grad_norm": 0.43929874897003174,
      "learning_rate": 8e-05,
      "loss": 1.5372,
      "step": 5632
    },
    {
      "epoch": 0.6087755322598076,
      "grad_norm": 0.48780208826065063,
      "learning_rate": 8e-05,
      "loss": 1.7107,
      "step": 5633
    },
    {
      "epoch": 0.6088836053171944,
      "grad_norm": 0.499931275844574,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 5634
    },
    {
      "epoch": 0.6089916783745812,
      "grad_norm": 0.4616127610206604,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 5635
    },
    {
      "epoch": 0.609099751431968,
      "grad_norm": 0.46425071358680725,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 5636
    },
    {
      "epoch": 0.6092078244893548,
      "grad_norm": 0.49222898483276367,
      "learning_rate": 8e-05,
      "loss": 1.8924,
      "step": 5637
    },
    {
      "epoch": 0.6093158975467416,
      "grad_norm": 0.4356289803981781,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 5638
    },
    {
      "epoch": 0.6094239706041283,
      "grad_norm": 0.5093769431114197,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 5639
    },
    {
      "epoch": 0.6095320436615151,
      "grad_norm": 0.5056626200675964,
      "learning_rate": 8e-05,
      "loss": 1.9181,
      "step": 5640
    },
    {
      "epoch": 0.609640116718902,
      "grad_norm": 0.48479798436164856,
      "learning_rate": 8e-05,
      "loss": 1.8341,
      "step": 5641
    },
    {
      "epoch": 0.6097481897762888,
      "grad_norm": 0.4404919743537903,
      "learning_rate": 8e-05,
      "loss": 1.6996,
      "step": 5642
    },
    {
      "epoch": 0.6098562628336756,
      "grad_norm": 0.4920656681060791,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 5643
    },
    {
      "epoch": 0.6099643358910624,
      "grad_norm": 0.46269622445106506,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 5644
    },
    {
      "epoch": 0.6100724089484492,
      "grad_norm": 0.4360038638114929,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 5645
    },
    {
      "epoch": 0.610180482005836,
      "grad_norm": 0.459777295589447,
      "learning_rate": 8e-05,
      "loss": 1.7809,
      "step": 5646
    },
    {
      "epoch": 0.6102885550632228,
      "grad_norm": 0.460322767496109,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 5647
    },
    {
      "epoch": 0.6103966281206096,
      "grad_norm": 0.5161346793174744,
      "learning_rate": 8e-05,
      "loss": 1.7148,
      "step": 5648
    },
    {
      "epoch": 0.6105047011779964,
      "grad_norm": 0.4731173813343048,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 5649
    },
    {
      "epoch": 0.6106127742353831,
      "grad_norm": 0.4915755093097687,
      "learning_rate": 8e-05,
      "loss": 1.8744,
      "step": 5650
    },
    {
      "epoch": 0.6107208472927699,
      "grad_norm": 0.4605688750743866,
      "learning_rate": 8e-05,
      "loss": 1.499,
      "step": 5651
    },
    {
      "epoch": 0.6108289203501567,
      "grad_norm": 0.43510356545448303,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 5652
    },
    {
      "epoch": 0.6109369934075435,
      "grad_norm": 0.477292001247406,
      "learning_rate": 8e-05,
      "loss": 1.8863,
      "step": 5653
    },
    {
      "epoch": 0.6110450664649303,
      "grad_norm": 0.4315967559814453,
      "learning_rate": 8e-05,
      "loss": 1.4403,
      "step": 5654
    },
    {
      "epoch": 0.6111531395223171,
      "grad_norm": 0.465789258480072,
      "learning_rate": 8e-05,
      "loss": 1.721,
      "step": 5655
    },
    {
      "epoch": 0.6112612125797039,
      "grad_norm": 0.4950612485408783,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 5656
    },
    {
      "epoch": 0.6113692856370907,
      "grad_norm": 0.4368162453174591,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 5657
    },
    {
      "epoch": 0.6114773586944775,
      "grad_norm": 0.4500575363636017,
      "learning_rate": 8e-05,
      "loss": 1.6818,
      "step": 5658
    },
    {
      "epoch": 0.6115854317518643,
      "grad_norm": 0.4476412832736969,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 5659
    },
    {
      "epoch": 0.611693504809251,
      "grad_norm": 0.4372149705886841,
      "learning_rate": 8e-05,
      "loss": 1.5683,
      "step": 5660
    },
    {
      "epoch": 0.6118015778666378,
      "grad_norm": 0.47047483921051025,
      "learning_rate": 8e-05,
      "loss": 1.8897,
      "step": 5661
    },
    {
      "epoch": 0.6119096509240246,
      "grad_norm": 0.49580007791519165,
      "learning_rate": 8e-05,
      "loss": 1.8522,
      "step": 5662
    },
    {
      "epoch": 0.6120177239814114,
      "grad_norm": 0.4656510651111603,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 5663
    },
    {
      "epoch": 0.6121257970387982,
      "grad_norm": 0.4361419975757599,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 5664
    },
    {
      "epoch": 0.612233870096185,
      "grad_norm": 0.4603894352912903,
      "learning_rate": 8e-05,
      "loss": 1.5472,
      "step": 5665
    },
    {
      "epoch": 0.6123419431535718,
      "grad_norm": 0.4363880157470703,
      "learning_rate": 8e-05,
      "loss": 1.8388,
      "step": 5666
    },
    {
      "epoch": 0.6124500162109586,
      "grad_norm": 0.4491766095161438,
      "learning_rate": 8e-05,
      "loss": 1.7454,
      "step": 5667
    },
    {
      "epoch": 0.6125580892683454,
      "grad_norm": 0.429231196641922,
      "learning_rate": 8e-05,
      "loss": 1.5077,
      "step": 5668
    },
    {
      "epoch": 0.6126661623257322,
      "grad_norm": 0.4491557776927948,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 5669
    },
    {
      "epoch": 0.612774235383119,
      "grad_norm": 0.43479305505752563,
      "learning_rate": 8e-05,
      "loss": 1.6802,
      "step": 5670
    },
    {
      "epoch": 0.6128823084405057,
      "grad_norm": 0.4732877314090729,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 5671
    },
    {
      "epoch": 0.6129903814978925,
      "grad_norm": 0.4247016906738281,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 5672
    },
    {
      "epoch": 0.6130984545552793,
      "grad_norm": 0.46953335404396057,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 5673
    },
    {
      "epoch": 0.6132065276126661,
      "grad_norm": 0.4804176986217499,
      "learning_rate": 8e-05,
      "loss": 1.8075,
      "step": 5674
    },
    {
      "epoch": 0.613314600670053,
      "grad_norm": 0.4732964038848877,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 5675
    },
    {
      "epoch": 0.6134226737274397,
      "grad_norm": 0.4651418924331665,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 5676
    },
    {
      "epoch": 0.6135307467848266,
      "grad_norm": 0.41696280241012573,
      "learning_rate": 8e-05,
      "loss": 1.6075,
      "step": 5677
    },
    {
      "epoch": 0.6136388198422134,
      "grad_norm": 0.4377788007259369,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 5678
    },
    {
      "epoch": 0.6137468928996002,
      "grad_norm": 0.4535657465457916,
      "learning_rate": 8e-05,
      "loss": 1.6194,
      "step": 5679
    },
    {
      "epoch": 0.613854965956987,
      "grad_norm": 0.42283836007118225,
      "learning_rate": 8e-05,
      "loss": 1.5842,
      "step": 5680
    },
    {
      "epoch": 0.6139630390143738,
      "grad_norm": 0.46147292852401733,
      "learning_rate": 8e-05,
      "loss": 1.6132,
      "step": 5681
    },
    {
      "epoch": 0.6140711120717605,
      "grad_norm": 0.49645549058914185,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 5682
    },
    {
      "epoch": 0.6141791851291473,
      "grad_norm": 0.48584115505218506,
      "learning_rate": 8e-05,
      "loss": 1.8104,
      "step": 5683
    },
    {
      "epoch": 0.6142872581865341,
      "grad_norm": 0.4603438973426819,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 5684
    },
    {
      "epoch": 0.6143953312439209,
      "grad_norm": 0.4575071930885315,
      "learning_rate": 8e-05,
      "loss": 1.8857,
      "step": 5685
    },
    {
      "epoch": 0.6145034043013077,
      "grad_norm": 0.4673934578895569,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 5686
    },
    {
      "epoch": 0.6146114773586945,
      "grad_norm": 0.520446240901947,
      "learning_rate": 8e-05,
      "loss": 1.8608,
      "step": 5687
    },
    {
      "epoch": 0.6147195504160813,
      "grad_norm": 0.47799015045166016,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 5688
    },
    {
      "epoch": 0.6148276234734681,
      "grad_norm": 0.4579654932022095,
      "learning_rate": 8e-05,
      "loss": 1.6124,
      "step": 5689
    },
    {
      "epoch": 0.6149356965308549,
      "grad_norm": 0.4563894271850586,
      "learning_rate": 8e-05,
      "loss": 1.7122,
      "step": 5690
    },
    {
      "epoch": 0.6150437695882417,
      "grad_norm": 0.4307648539543152,
      "learning_rate": 8e-05,
      "loss": 1.5208,
      "step": 5691
    },
    {
      "epoch": 0.6151518426456285,
      "grad_norm": 0.4436216652393341,
      "learning_rate": 8e-05,
      "loss": 1.7211,
      "step": 5692
    },
    {
      "epoch": 0.6152599157030152,
      "grad_norm": 0.4701896607875824,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 5693
    },
    {
      "epoch": 0.615367988760402,
      "grad_norm": 0.4382934868335724,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 5694
    },
    {
      "epoch": 0.6154760618177888,
      "grad_norm": 0.4490520656108856,
      "learning_rate": 8e-05,
      "loss": 1.582,
      "step": 5695
    },
    {
      "epoch": 0.6155841348751756,
      "grad_norm": 0.45886728167533875,
      "learning_rate": 8e-05,
      "loss": 1.6994,
      "step": 5696
    },
    {
      "epoch": 0.6156922079325624,
      "grad_norm": 0.45676755905151367,
      "learning_rate": 8e-05,
      "loss": 1.6925,
      "step": 5697
    },
    {
      "epoch": 0.6158002809899492,
      "grad_norm": 0.46693795919418335,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 5698
    },
    {
      "epoch": 0.615908354047336,
      "grad_norm": 0.411001592874527,
      "learning_rate": 8e-05,
      "loss": 1.5685,
      "step": 5699
    },
    {
      "epoch": 0.6160164271047228,
      "grad_norm": 0.4623746871948242,
      "learning_rate": 8e-05,
      "loss": 1.5874,
      "step": 5700
    },
    {
      "epoch": 0.6161245001621096,
      "grad_norm": 0.47669318318367004,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 5701
    },
    {
      "epoch": 0.6162325732194964,
      "grad_norm": 0.4281429350376129,
      "learning_rate": 8e-05,
      "loss": 1.5948,
      "step": 5702
    },
    {
      "epoch": 0.6163406462768831,
      "grad_norm": 0.5077112913131714,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 5703
    },
    {
      "epoch": 0.6164487193342699,
      "grad_norm": 0.4525469243526459,
      "learning_rate": 8e-05,
      "loss": 1.7139,
      "step": 5704
    },
    {
      "epoch": 0.6165567923916567,
      "grad_norm": 0.45006120204925537,
      "learning_rate": 8e-05,
      "loss": 1.5009,
      "step": 5705
    },
    {
      "epoch": 0.6166648654490435,
      "grad_norm": 0.4362548291683197,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 5706
    },
    {
      "epoch": 0.6167729385064303,
      "grad_norm": 0.429889053106308,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 5707
    },
    {
      "epoch": 0.6168810115638171,
      "grad_norm": 0.4965493679046631,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 5708
    },
    {
      "epoch": 0.6169890846212039,
      "grad_norm": 0.4343239963054657,
      "learning_rate": 8e-05,
      "loss": 1.6534,
      "step": 5709
    },
    {
      "epoch": 0.6170971576785907,
      "grad_norm": 0.45086291432380676,
      "learning_rate": 8e-05,
      "loss": 1.6432,
      "step": 5710
    },
    {
      "epoch": 0.6172052307359776,
      "grad_norm": 0.4735439419746399,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 5711
    },
    {
      "epoch": 0.6173133037933644,
      "grad_norm": 0.4599758982658386,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 5712
    },
    {
      "epoch": 0.6174213768507512,
      "grad_norm": 0.4597901999950409,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 5713
    },
    {
      "epoch": 0.6175294499081379,
      "grad_norm": 0.44011133909225464,
      "learning_rate": 8e-05,
      "loss": 1.6125,
      "step": 5714
    },
    {
      "epoch": 0.6176375229655247,
      "grad_norm": 0.44547387957572937,
      "learning_rate": 8e-05,
      "loss": 1.5495,
      "step": 5715
    },
    {
      "epoch": 0.6177455960229115,
      "grad_norm": 0.4830479025840759,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 5716
    },
    {
      "epoch": 0.6178536690802983,
      "grad_norm": 0.4625515043735504,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 5717
    },
    {
      "epoch": 0.6179617421376851,
      "grad_norm": 0.5090802311897278,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 5718
    },
    {
      "epoch": 0.6180698151950719,
      "grad_norm": 0.514909565448761,
      "learning_rate": 8e-05,
      "loss": 1.6898,
      "step": 5719
    },
    {
      "epoch": 0.6181778882524587,
      "grad_norm": 0.44272735714912415,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 5720
    },
    {
      "epoch": 0.6182859613098455,
      "grad_norm": 0.43055665493011475,
      "learning_rate": 8e-05,
      "loss": 1.6417,
      "step": 5721
    },
    {
      "epoch": 0.6183940343672323,
      "grad_norm": 0.5243388414382935,
      "learning_rate": 8e-05,
      "loss": 1.9239,
      "step": 5722
    },
    {
      "epoch": 0.6185021074246191,
      "grad_norm": 0.45779767632484436,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 5723
    },
    {
      "epoch": 0.6186101804820059,
      "grad_norm": 0.5226853489875793,
      "learning_rate": 8e-05,
      "loss": 1.9079,
      "step": 5724
    },
    {
      "epoch": 0.6187182535393926,
      "grad_norm": 0.45591089129447937,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 5725
    },
    {
      "epoch": 0.6188263265967794,
      "grad_norm": 0.4691537320613861,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 5726
    },
    {
      "epoch": 0.6189343996541662,
      "grad_norm": 0.42020440101623535,
      "learning_rate": 8e-05,
      "loss": 1.5327,
      "step": 5727
    },
    {
      "epoch": 0.619042472711553,
      "grad_norm": 0.5019327998161316,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 5728
    },
    {
      "epoch": 0.6191505457689398,
      "grad_norm": 0.46830180287361145,
      "learning_rate": 8e-05,
      "loss": 1.8991,
      "step": 5729
    },
    {
      "epoch": 0.6192586188263266,
      "grad_norm": 0.46691957116127014,
      "learning_rate": 8e-05,
      "loss": 1.8576,
      "step": 5730
    },
    {
      "epoch": 0.6193666918837134,
      "grad_norm": 0.447125107049942,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 5731
    },
    {
      "epoch": 0.6194747649411002,
      "grad_norm": 0.4188820421695709,
      "learning_rate": 8e-05,
      "loss": 1.5902,
      "step": 5732
    },
    {
      "epoch": 0.619582837998487,
      "grad_norm": 0.42870327830314636,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 5733
    },
    {
      "epoch": 0.6196909110558738,
      "grad_norm": 0.4528655409812927,
      "learning_rate": 8e-05,
      "loss": 1.8358,
      "step": 5734
    },
    {
      "epoch": 0.6197989841132605,
      "grad_norm": 0.49897122383117676,
      "learning_rate": 8e-05,
      "loss": 1.8613,
      "step": 5735
    },
    {
      "epoch": 0.6199070571706473,
      "grad_norm": 0.4554516673088074,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 5736
    },
    {
      "epoch": 0.6200151302280341,
      "grad_norm": 0.4372716248035431,
      "learning_rate": 8e-05,
      "loss": 1.8193,
      "step": 5737
    },
    {
      "epoch": 0.6201232032854209,
      "grad_norm": 0.4544358253479004,
      "learning_rate": 8e-05,
      "loss": 1.6733,
      "step": 5738
    },
    {
      "epoch": 0.6202312763428077,
      "grad_norm": 0.4569014012813568,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 5739
    },
    {
      "epoch": 0.6203393494001945,
      "grad_norm": 0.44572606682777405,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 5740
    },
    {
      "epoch": 0.6204474224575813,
      "grad_norm": 0.43456554412841797,
      "learning_rate": 8e-05,
      "loss": 1.6429,
      "step": 5741
    },
    {
      "epoch": 0.6205554955149681,
      "grad_norm": 0.4171997606754303,
      "learning_rate": 8e-05,
      "loss": 1.3687,
      "step": 5742
    },
    {
      "epoch": 0.6206635685723549,
      "grad_norm": 0.4764534533023834,
      "learning_rate": 8e-05,
      "loss": 1.7218,
      "step": 5743
    },
    {
      "epoch": 0.6207716416297417,
      "grad_norm": 0.4435490369796753,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 5744
    },
    {
      "epoch": 0.6208797146871285,
      "grad_norm": 0.4177079498767853,
      "learning_rate": 8e-05,
      "loss": 1.6374,
      "step": 5745
    },
    {
      "epoch": 0.6209877877445152,
      "grad_norm": 0.4497847855091095,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 5746
    },
    {
      "epoch": 0.621095860801902,
      "grad_norm": 0.510424017906189,
      "learning_rate": 8e-05,
      "loss": 1.745,
      "step": 5747
    },
    {
      "epoch": 0.6212039338592888,
      "grad_norm": 0.4944458305835724,
      "learning_rate": 8e-05,
      "loss": 1.8689,
      "step": 5748
    },
    {
      "epoch": 0.6213120069166757,
      "grad_norm": 0.42158809304237366,
      "learning_rate": 8e-05,
      "loss": 1.6672,
      "step": 5749
    },
    {
      "epoch": 0.6214200799740625,
      "grad_norm": 0.44082900881767273,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 5750
    },
    {
      "epoch": 0.6215281530314493,
      "grad_norm": 0.41281527280807495,
      "learning_rate": 8e-05,
      "loss": 1.6584,
      "step": 5751
    },
    {
      "epoch": 0.6216362260888361,
      "grad_norm": 0.4595940411090851,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 5752
    },
    {
      "epoch": 0.6217442991462229,
      "grad_norm": 0.5178603529930115,
      "learning_rate": 8e-05,
      "loss": 1.9817,
      "step": 5753
    },
    {
      "epoch": 0.6218523722036097,
      "grad_norm": 0.47200581431388855,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 5754
    },
    {
      "epoch": 0.6219604452609965,
      "grad_norm": 0.465548574924469,
      "learning_rate": 8e-05,
      "loss": 1.6886,
      "step": 5755
    },
    {
      "epoch": 0.6220685183183833,
      "grad_norm": 0.47664961218833923,
      "learning_rate": 8e-05,
      "loss": 1.6647,
      "step": 5756
    },
    {
      "epoch": 0.62217659137577,
      "grad_norm": 0.47717034816741943,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 5757
    },
    {
      "epoch": 0.6222846644331568,
      "grad_norm": 0.44756773114204407,
      "learning_rate": 8e-05,
      "loss": 1.7164,
      "step": 5758
    },
    {
      "epoch": 0.6223927374905436,
      "grad_norm": 0.4292087256908417,
      "learning_rate": 8e-05,
      "loss": 1.5477,
      "step": 5759
    },
    {
      "epoch": 0.6225008105479304,
      "grad_norm": 0.4683057367801666,
      "learning_rate": 8e-05,
      "loss": 1.8381,
      "step": 5760
    },
    {
      "epoch": 0.6226088836053172,
      "grad_norm": 0.45287930965423584,
      "learning_rate": 8e-05,
      "loss": 1.7058,
      "step": 5761
    },
    {
      "epoch": 0.622716956662704,
      "grad_norm": 0.4704689085483551,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 5762
    },
    {
      "epoch": 0.6228250297200908,
      "grad_norm": 0.4658387303352356,
      "learning_rate": 8e-05,
      "loss": 1.7833,
      "step": 5763
    },
    {
      "epoch": 0.6229331027774776,
      "grad_norm": 0.5236711502075195,
      "learning_rate": 8e-05,
      "loss": 1.5662,
      "step": 5764
    },
    {
      "epoch": 0.6230411758348644,
      "grad_norm": 0.42707574367523193,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 5765
    },
    {
      "epoch": 0.6231492488922512,
      "grad_norm": 0.42540737986564636,
      "learning_rate": 8e-05,
      "loss": 1.6354,
      "step": 5766
    },
    {
      "epoch": 0.623257321949638,
      "grad_norm": 0.4672861099243164,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 5767
    },
    {
      "epoch": 0.6233653950070247,
      "grad_norm": 0.46566343307495117,
      "learning_rate": 8e-05,
      "loss": 1.5059,
      "step": 5768
    },
    {
      "epoch": 0.6234734680644115,
      "grad_norm": 0.43126392364501953,
      "learning_rate": 8e-05,
      "loss": 1.6216,
      "step": 5769
    },
    {
      "epoch": 0.6235815411217983,
      "grad_norm": 0.4801551401615143,
      "learning_rate": 8e-05,
      "loss": 1.5269,
      "step": 5770
    },
    {
      "epoch": 0.6236896141791851,
      "grad_norm": 0.4314405024051666,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 5771
    },
    {
      "epoch": 0.6237976872365719,
      "grad_norm": 0.4687712788581848,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 5772
    },
    {
      "epoch": 0.6239057602939587,
      "grad_norm": 0.43072938919067383,
      "learning_rate": 8e-05,
      "loss": 1.5837,
      "step": 5773
    },
    {
      "epoch": 0.6240138333513455,
      "grad_norm": 0.4832228720188141,
      "learning_rate": 8e-05,
      "loss": 1.7936,
      "step": 5774
    },
    {
      "epoch": 0.6241219064087323,
      "grad_norm": 0.5146892070770264,
      "learning_rate": 8e-05,
      "loss": 1.7736,
      "step": 5775
    },
    {
      "epoch": 0.6242299794661191,
      "grad_norm": 0.4391840100288391,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 5776
    },
    {
      "epoch": 0.6243380525235059,
      "grad_norm": 0.48520421981811523,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 5777
    },
    {
      "epoch": 0.6244461255808926,
      "grad_norm": 0.47812116146087646,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 5778
    },
    {
      "epoch": 0.6245541986382794,
      "grad_norm": 0.4548112750053406,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 5779
    },
    {
      "epoch": 0.6246622716956662,
      "grad_norm": 0.45246341824531555,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 5780
    },
    {
      "epoch": 0.624770344753053,
      "grad_norm": 0.424312561750412,
      "learning_rate": 8e-05,
      "loss": 1.6247,
      "step": 5781
    },
    {
      "epoch": 0.6248784178104398,
      "grad_norm": 0.4125204384326935,
      "learning_rate": 8e-05,
      "loss": 1.5171,
      "step": 5782
    },
    {
      "epoch": 0.6249864908678266,
      "grad_norm": 0.450283020734787,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 5783
    },
    {
      "epoch": 0.6250945639252135,
      "grad_norm": 0.4598381817340851,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 5784
    },
    {
      "epoch": 0.6252026369826003,
      "grad_norm": 0.4287654757499695,
      "learning_rate": 8e-05,
      "loss": 1.4106,
      "step": 5785
    },
    {
      "epoch": 0.6253107100399871,
      "grad_norm": 0.45009666681289673,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 5786
    },
    {
      "epoch": 0.6254187830973739,
      "grad_norm": 0.46841156482696533,
      "learning_rate": 8e-05,
      "loss": 1.743,
      "step": 5787
    },
    {
      "epoch": 0.6255268561547607,
      "grad_norm": 0.5230856537818909,
      "learning_rate": 8e-05,
      "loss": 1.5775,
      "step": 5788
    },
    {
      "epoch": 0.6256349292121474,
      "grad_norm": 0.4259295165538788,
      "learning_rate": 8e-05,
      "loss": 1.7112,
      "step": 5789
    },
    {
      "epoch": 0.6257430022695342,
      "grad_norm": 0.43393945693969727,
      "learning_rate": 8e-05,
      "loss": 1.7411,
      "step": 5790
    },
    {
      "epoch": 0.625851075326921,
      "grad_norm": 0.501250147819519,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 5791
    },
    {
      "epoch": 0.6259591483843078,
      "grad_norm": 0.4336906969547272,
      "learning_rate": 8e-05,
      "loss": 1.629,
      "step": 5792
    },
    {
      "epoch": 0.6260672214416946,
      "grad_norm": 0.5130109786987305,
      "learning_rate": 8e-05,
      "loss": 1.6233,
      "step": 5793
    },
    {
      "epoch": 0.6261752944990814,
      "grad_norm": 0.4786396920681,
      "learning_rate": 8e-05,
      "loss": 1.7053,
      "step": 5794
    },
    {
      "epoch": 0.6262833675564682,
      "grad_norm": 0.46680203080177307,
      "learning_rate": 8e-05,
      "loss": 1.5725,
      "step": 5795
    },
    {
      "epoch": 0.626391440613855,
      "grad_norm": 0.4778881371021271,
      "learning_rate": 8e-05,
      "loss": 1.6598,
      "step": 5796
    },
    {
      "epoch": 0.6264995136712418,
      "grad_norm": 0.45743969082832336,
      "learning_rate": 8e-05,
      "loss": 1.6179,
      "step": 5797
    },
    {
      "epoch": 0.6266075867286286,
      "grad_norm": 0.5267737507820129,
      "learning_rate": 8e-05,
      "loss": 1.9002,
      "step": 5798
    },
    {
      "epoch": 0.6267156597860154,
      "grad_norm": 0.4658590257167816,
      "learning_rate": 8e-05,
      "loss": 1.6606,
      "step": 5799
    },
    {
      "epoch": 0.6268237328434021,
      "grad_norm": 0.43808770179748535,
      "learning_rate": 8e-05,
      "loss": 1.6875,
      "step": 5800
    },
    {
      "epoch": 0.6269318059007889,
      "grad_norm": 0.44403275847435,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 5801
    },
    {
      "epoch": 0.6270398789581757,
      "grad_norm": 0.4962800443172455,
      "learning_rate": 8e-05,
      "loss": 1.7695,
      "step": 5802
    },
    {
      "epoch": 0.6271479520155625,
      "grad_norm": 0.46098706126213074,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 5803
    },
    {
      "epoch": 0.6272560250729493,
      "grad_norm": 0.45993494987487793,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 5804
    },
    {
      "epoch": 0.6273640981303361,
      "grad_norm": 0.46268653869628906,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 5805
    },
    {
      "epoch": 0.6274721711877229,
      "grad_norm": 0.5217958688735962,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 5806
    },
    {
      "epoch": 0.6275802442451097,
      "grad_norm": 0.47385331988334656,
      "learning_rate": 8e-05,
      "loss": 1.496,
      "step": 5807
    },
    {
      "epoch": 0.6276883173024965,
      "grad_norm": 0.4645078182220459,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 5808
    },
    {
      "epoch": 0.6277963903598833,
      "grad_norm": 0.43602481484413147,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 5809
    },
    {
      "epoch": 0.62790446341727,
      "grad_norm": 0.46625471115112305,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 5810
    },
    {
      "epoch": 0.6280125364746568,
      "grad_norm": 0.4461331069469452,
      "learning_rate": 8e-05,
      "loss": 1.6253,
      "step": 5811
    },
    {
      "epoch": 0.6281206095320436,
      "grad_norm": 0.4687572121620178,
      "learning_rate": 8e-05,
      "loss": 1.8483,
      "step": 5812
    },
    {
      "epoch": 0.6282286825894304,
      "grad_norm": 0.44872134923934937,
      "learning_rate": 8e-05,
      "loss": 1.7024,
      "step": 5813
    },
    {
      "epoch": 0.6283367556468172,
      "grad_norm": 0.4775184392929077,
      "learning_rate": 8e-05,
      "loss": 1.8608,
      "step": 5814
    },
    {
      "epoch": 0.628444828704204,
      "grad_norm": 0.4612930715084076,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 5815
    },
    {
      "epoch": 0.6285529017615908,
      "grad_norm": 0.46075528860092163,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 5816
    },
    {
      "epoch": 0.6286609748189776,
      "grad_norm": 0.5088425278663635,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 5817
    },
    {
      "epoch": 0.6287690478763645,
      "grad_norm": 0.47776079177856445,
      "learning_rate": 8e-05,
      "loss": 1.8133,
      "step": 5818
    },
    {
      "epoch": 0.6288771209337513,
      "grad_norm": 0.5069181323051453,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 5819
    },
    {
      "epoch": 0.6289851939911381,
      "grad_norm": 0.45624807476997375,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 5820
    },
    {
      "epoch": 0.6290932670485248,
      "grad_norm": 0.4311521053314209,
      "learning_rate": 8e-05,
      "loss": 1.5785,
      "step": 5821
    },
    {
      "epoch": 0.6292013401059116,
      "grad_norm": 0.49467381834983826,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 5822
    },
    {
      "epoch": 0.6293094131632984,
      "grad_norm": 0.48584577441215515,
      "learning_rate": 8e-05,
      "loss": 1.4911,
      "step": 5823
    },
    {
      "epoch": 0.6294174862206852,
      "grad_norm": 0.4531196653842926,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 5824
    },
    {
      "epoch": 0.629525559278072,
      "grad_norm": 0.44337621331214905,
      "learning_rate": 8e-05,
      "loss": 1.6557,
      "step": 5825
    },
    {
      "epoch": 0.6296336323354588,
      "grad_norm": 0.5295804142951965,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 5826
    },
    {
      "epoch": 0.6297417053928456,
      "grad_norm": 0.45692169666290283,
      "learning_rate": 8e-05,
      "loss": 1.8106,
      "step": 5827
    },
    {
      "epoch": 0.6298497784502324,
      "grad_norm": 0.5159560441970825,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 5828
    },
    {
      "epoch": 0.6299578515076192,
      "grad_norm": 0.43604031205177307,
      "learning_rate": 8e-05,
      "loss": 1.5902,
      "step": 5829
    },
    {
      "epoch": 0.630065924565006,
      "grad_norm": 0.48572859168052673,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 5830
    },
    {
      "epoch": 0.6301739976223928,
      "grad_norm": 0.45837894082069397,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 5831
    },
    {
      "epoch": 0.6302820706797795,
      "grad_norm": 0.45187926292419434,
      "learning_rate": 8e-05,
      "loss": 1.6769,
      "step": 5832
    },
    {
      "epoch": 0.6303901437371663,
      "grad_norm": 0.47859957814216614,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 5833
    },
    {
      "epoch": 0.6304982167945531,
      "grad_norm": 0.48727160692214966,
      "learning_rate": 8e-05,
      "loss": 1.88,
      "step": 5834
    },
    {
      "epoch": 0.6306062898519399,
      "grad_norm": 0.46846237778663635,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 5835
    },
    {
      "epoch": 0.6307143629093267,
      "grad_norm": 0.49860095977783203,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 5836
    },
    {
      "epoch": 0.6308224359667135,
      "grad_norm": 0.46450579166412354,
      "learning_rate": 8e-05,
      "loss": 1.4565,
      "step": 5837
    },
    {
      "epoch": 0.6309305090241003,
      "grad_norm": 0.4572513997554779,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 5838
    },
    {
      "epoch": 0.6310385820814871,
      "grad_norm": 0.5078173875808716,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 5839
    },
    {
      "epoch": 0.6311466551388739,
      "grad_norm": 0.4972337782382965,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 5840
    },
    {
      "epoch": 0.6312547281962607,
      "grad_norm": 0.4321762025356293,
      "learning_rate": 8e-05,
      "loss": 1.7093,
      "step": 5841
    },
    {
      "epoch": 0.6313628012536475,
      "grad_norm": 0.4722598195075989,
      "learning_rate": 8e-05,
      "loss": 1.5747,
      "step": 5842
    },
    {
      "epoch": 0.6314708743110342,
      "grad_norm": 0.44367337226867676,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 5843
    },
    {
      "epoch": 0.631578947368421,
      "grad_norm": 0.4662986397743225,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 5844
    },
    {
      "epoch": 0.6316870204258078,
      "grad_norm": 0.469190776348114,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 5845
    },
    {
      "epoch": 0.6317950934831946,
      "grad_norm": 0.451201468706131,
      "learning_rate": 8e-05,
      "loss": 1.3589,
      "step": 5846
    },
    {
      "epoch": 0.6319031665405814,
      "grad_norm": 0.4908949136734009,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 5847
    },
    {
      "epoch": 0.6320112395979682,
      "grad_norm": 0.46603408455848694,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 5848
    },
    {
      "epoch": 0.632119312655355,
      "grad_norm": 0.4675006866455078,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 5849
    },
    {
      "epoch": 0.6322273857127418,
      "grad_norm": 0.44438135623931885,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 5850
    },
    {
      "epoch": 0.6323354587701286,
      "grad_norm": 0.4587791860103607,
      "learning_rate": 8e-05,
      "loss": 1.588,
      "step": 5851
    },
    {
      "epoch": 0.6324435318275154,
      "grad_norm": 0.4653497338294983,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 5852
    },
    {
      "epoch": 0.6325516048849021,
      "grad_norm": 0.4402623474597931,
      "learning_rate": 8e-05,
      "loss": 1.5893,
      "step": 5853
    },
    {
      "epoch": 0.632659677942289,
      "grad_norm": 0.45831578969955444,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 5854
    },
    {
      "epoch": 0.6327677509996757,
      "grad_norm": 0.4931390881538391,
      "learning_rate": 8e-05,
      "loss": 1.7813,
      "step": 5855
    },
    {
      "epoch": 0.6328758240570626,
      "grad_norm": 0.46938833594322205,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 5856
    },
    {
      "epoch": 0.6329838971144494,
      "grad_norm": 0.43494829535484314,
      "learning_rate": 8e-05,
      "loss": 1.5954,
      "step": 5857
    },
    {
      "epoch": 0.6330919701718362,
      "grad_norm": 0.47373664379119873,
      "learning_rate": 8e-05,
      "loss": 1.7744,
      "step": 5858
    },
    {
      "epoch": 0.633200043229223,
      "grad_norm": 0.44712769985198975,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 5859
    },
    {
      "epoch": 0.6333081162866098,
      "grad_norm": 0.437565416097641,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 5860
    },
    {
      "epoch": 0.6334161893439966,
      "grad_norm": 0.459397554397583,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 5861
    },
    {
      "epoch": 0.6335242624013834,
      "grad_norm": 0.4738190174102783,
      "learning_rate": 8e-05,
      "loss": 1.5729,
      "step": 5862
    },
    {
      "epoch": 0.6336323354587702,
      "grad_norm": 0.5572516918182373,
      "learning_rate": 8e-05,
      "loss": 2.0648,
      "step": 5863
    },
    {
      "epoch": 0.6337404085161569,
      "grad_norm": 0.4255099892616272,
      "learning_rate": 8e-05,
      "loss": 1.6103,
      "step": 5864
    },
    {
      "epoch": 0.6338484815735437,
      "grad_norm": 0.46732527017593384,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 5865
    },
    {
      "epoch": 0.6339565546309305,
      "grad_norm": 0.461786687374115,
      "learning_rate": 8e-05,
      "loss": 1.6568,
      "step": 5866
    },
    {
      "epoch": 0.6340646276883173,
      "grad_norm": 0.48468831181526184,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 5867
    },
    {
      "epoch": 0.6341727007457041,
      "grad_norm": 0.5050501823425293,
      "learning_rate": 8e-05,
      "loss": 1.9284,
      "step": 5868
    },
    {
      "epoch": 0.6342807738030909,
      "grad_norm": 0.45135846734046936,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 5869
    },
    {
      "epoch": 0.6343888468604777,
      "grad_norm": 0.43670931458473206,
      "learning_rate": 8e-05,
      "loss": 1.6321,
      "step": 5870
    },
    {
      "epoch": 0.6344969199178645,
      "grad_norm": 0.4494459927082062,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 5871
    },
    {
      "epoch": 0.6346049929752513,
      "grad_norm": 0.4380071759223938,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 5872
    },
    {
      "epoch": 0.6347130660326381,
      "grad_norm": 0.47637027502059937,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 5873
    },
    {
      "epoch": 0.6348211390900249,
      "grad_norm": 0.4148869514465332,
      "learning_rate": 8e-05,
      "loss": 1.4941,
      "step": 5874
    },
    {
      "epoch": 0.6349292121474116,
      "grad_norm": 0.5438070893287659,
      "learning_rate": 8e-05,
      "loss": 1.7469,
      "step": 5875
    },
    {
      "epoch": 0.6350372852047984,
      "grad_norm": 0.42122015357017517,
      "learning_rate": 8e-05,
      "loss": 1.4498,
      "step": 5876
    },
    {
      "epoch": 0.6351453582621852,
      "grad_norm": 0.467573881149292,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 5877
    },
    {
      "epoch": 0.635253431319572,
      "grad_norm": 0.46532878279685974,
      "learning_rate": 8e-05,
      "loss": 1.7937,
      "step": 5878
    },
    {
      "epoch": 0.6353615043769588,
      "grad_norm": 0.4412324130535126,
      "learning_rate": 8e-05,
      "loss": 1.6157,
      "step": 5879
    },
    {
      "epoch": 0.6354695774343456,
      "grad_norm": 0.47655388712882996,
      "learning_rate": 8e-05,
      "loss": 1.7445,
      "step": 5880
    },
    {
      "epoch": 0.6355776504917324,
      "grad_norm": 0.5112617611885071,
      "learning_rate": 8e-05,
      "loss": 1.9348,
      "step": 5881
    },
    {
      "epoch": 0.6356857235491192,
      "grad_norm": 0.46213895082473755,
      "learning_rate": 8e-05,
      "loss": 1.3423,
      "step": 5882
    },
    {
      "epoch": 0.635793796606506,
      "grad_norm": 0.43682414293289185,
      "learning_rate": 8e-05,
      "loss": 1.6946,
      "step": 5883
    },
    {
      "epoch": 0.6359018696638928,
      "grad_norm": 0.4414867162704468,
      "learning_rate": 8e-05,
      "loss": 1.6097,
      "step": 5884
    },
    {
      "epoch": 0.6360099427212795,
      "grad_norm": 0.47384002804756165,
      "learning_rate": 8e-05,
      "loss": 1.818,
      "step": 5885
    },
    {
      "epoch": 0.6361180157786663,
      "grad_norm": 0.4790787696838379,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 5886
    },
    {
      "epoch": 0.6362260888360531,
      "grad_norm": 0.4511025846004486,
      "learning_rate": 8e-05,
      "loss": 1.449,
      "step": 5887
    },
    {
      "epoch": 0.6363341618934399,
      "grad_norm": 0.48787668347358704,
      "learning_rate": 8e-05,
      "loss": 1.8304,
      "step": 5888
    },
    {
      "epoch": 0.6364422349508267,
      "grad_norm": 0.4610356390476227,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 5889
    },
    {
      "epoch": 0.6365503080082136,
      "grad_norm": 0.47889548540115356,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 5890
    },
    {
      "epoch": 0.6366583810656004,
      "grad_norm": 0.46246275305747986,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 5891
    },
    {
      "epoch": 0.6367664541229872,
      "grad_norm": 0.48781657218933105,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 5892
    },
    {
      "epoch": 0.636874527180374,
      "grad_norm": 0.4480491578578949,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 5893
    },
    {
      "epoch": 0.6369826002377608,
      "grad_norm": 0.4417595863342285,
      "learning_rate": 8e-05,
      "loss": 1.55,
      "step": 5894
    },
    {
      "epoch": 0.6370906732951476,
      "grad_norm": 0.467379093170166,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 5895
    },
    {
      "epoch": 0.6371987463525343,
      "grad_norm": 0.45467954874038696,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 5896
    },
    {
      "epoch": 0.6373068194099211,
      "grad_norm": 0.4466654360294342,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 5897
    },
    {
      "epoch": 0.6374148924673079,
      "grad_norm": 0.4570397734642029,
      "learning_rate": 8e-05,
      "loss": 1.6452,
      "step": 5898
    },
    {
      "epoch": 0.6375229655246947,
      "grad_norm": 0.4655531644821167,
      "learning_rate": 8e-05,
      "loss": 1.749,
      "step": 5899
    },
    {
      "epoch": 0.6376310385820815,
      "grad_norm": 0.46655017137527466,
      "learning_rate": 8e-05,
      "loss": 1.6664,
      "step": 5900
    },
    {
      "epoch": 0.6377391116394683,
      "grad_norm": 0.47527891397476196,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 5901
    },
    {
      "epoch": 0.6378471846968551,
      "grad_norm": 0.44720250368118286,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 5902
    },
    {
      "epoch": 0.6379552577542419,
      "grad_norm": 0.4873383343219757,
      "learning_rate": 8e-05,
      "loss": 1.7941,
      "step": 5903
    },
    {
      "epoch": 0.6380633308116287,
      "grad_norm": 0.4687193036079407,
      "learning_rate": 8e-05,
      "loss": 1.8241,
      "step": 5904
    },
    {
      "epoch": 0.6381714038690155,
      "grad_norm": 0.445834219455719,
      "learning_rate": 8e-05,
      "loss": 1.7182,
      "step": 5905
    },
    {
      "epoch": 0.6382794769264023,
      "grad_norm": 0.45877718925476074,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 5906
    },
    {
      "epoch": 0.638387549983789,
      "grad_norm": 0.5117327570915222,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 5907
    },
    {
      "epoch": 0.6384956230411758,
      "grad_norm": 0.4483349919319153,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 5908
    },
    {
      "epoch": 0.6386036960985626,
      "grad_norm": 0.47284021973609924,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 5909
    },
    {
      "epoch": 0.6387117691559494,
      "grad_norm": 0.4378233253955841,
      "learning_rate": 8e-05,
      "loss": 1.7097,
      "step": 5910
    },
    {
      "epoch": 0.6388198422133362,
      "grad_norm": 0.43169206380844116,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 5911
    },
    {
      "epoch": 0.638927915270723,
      "grad_norm": 0.46599283814430237,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 5912
    },
    {
      "epoch": 0.6390359883281098,
      "grad_norm": 0.4385697841644287,
      "learning_rate": 8e-05,
      "loss": 1.5995,
      "step": 5913
    },
    {
      "epoch": 0.6391440613854966,
      "grad_norm": 0.4320792853832245,
      "learning_rate": 8e-05,
      "loss": 1.6947,
      "step": 5914
    },
    {
      "epoch": 0.6392521344428834,
      "grad_norm": 0.45624417066574097,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 5915
    },
    {
      "epoch": 0.6393602075002702,
      "grad_norm": 0.4883643388748169,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 5916
    },
    {
      "epoch": 0.6394682805576569,
      "grad_norm": 0.45986685156822205,
      "learning_rate": 8e-05,
      "loss": 1.8108,
      "step": 5917
    },
    {
      "epoch": 0.6395763536150437,
      "grad_norm": 0.45105069875717163,
      "learning_rate": 8e-05,
      "loss": 1.5934,
      "step": 5918
    },
    {
      "epoch": 0.6396844266724305,
      "grad_norm": 0.46297338604927063,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 5919
    },
    {
      "epoch": 0.6397924997298173,
      "grad_norm": 0.4720000624656677,
      "learning_rate": 8e-05,
      "loss": 1.8171,
      "step": 5920
    },
    {
      "epoch": 0.6399005727872041,
      "grad_norm": 0.43475326895713806,
      "learning_rate": 8e-05,
      "loss": 1.6701,
      "step": 5921
    },
    {
      "epoch": 0.6400086458445909,
      "grad_norm": 0.49164581298828125,
      "learning_rate": 8e-05,
      "loss": 1.868,
      "step": 5922
    },
    {
      "epoch": 0.6401167189019777,
      "grad_norm": 0.4984549582004547,
      "learning_rate": 8e-05,
      "loss": 1.886,
      "step": 5923
    },
    {
      "epoch": 0.6402247919593645,
      "grad_norm": 0.4466260075569153,
      "learning_rate": 8e-05,
      "loss": 1.7341,
      "step": 5924
    },
    {
      "epoch": 0.6403328650167514,
      "grad_norm": 0.45236361026763916,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 5925
    },
    {
      "epoch": 0.6404409380741382,
      "grad_norm": 0.5447743535041809,
      "learning_rate": 8e-05,
      "loss": 1.9852,
      "step": 5926
    },
    {
      "epoch": 0.640549011131525,
      "grad_norm": 0.45476287603378296,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 5927
    },
    {
      "epoch": 0.6406570841889117,
      "grad_norm": 0.496670126914978,
      "learning_rate": 8e-05,
      "loss": 1.8243,
      "step": 5928
    },
    {
      "epoch": 0.6407651572462985,
      "grad_norm": 0.48587891459465027,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 5929
    },
    {
      "epoch": 0.6408732303036853,
      "grad_norm": 0.4802110493183136,
      "learning_rate": 8e-05,
      "loss": 1.7537,
      "step": 5930
    },
    {
      "epoch": 0.6409813033610721,
      "grad_norm": 0.46664249897003174,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 5931
    },
    {
      "epoch": 0.6410893764184589,
      "grad_norm": 0.4421055018901825,
      "learning_rate": 8e-05,
      "loss": 1.5226,
      "step": 5932
    },
    {
      "epoch": 0.6411974494758457,
      "grad_norm": 0.4656575918197632,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 5933
    },
    {
      "epoch": 0.6413055225332325,
      "grad_norm": 0.4513384997844696,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 5934
    },
    {
      "epoch": 0.6414135955906193,
      "grad_norm": 0.49333086609840393,
      "learning_rate": 8e-05,
      "loss": 1.9898,
      "step": 5935
    },
    {
      "epoch": 0.6415216686480061,
      "grad_norm": 0.46963754296302795,
      "learning_rate": 8e-05,
      "loss": 1.5507,
      "step": 5936
    },
    {
      "epoch": 0.6416297417053929,
      "grad_norm": 0.47902289032936096,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 5937
    },
    {
      "epoch": 0.6417378147627797,
      "grad_norm": 0.4897211194038391,
      "learning_rate": 8e-05,
      "loss": 1.8406,
      "step": 5938
    },
    {
      "epoch": 0.6418458878201664,
      "grad_norm": 0.4949687123298645,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 5939
    },
    {
      "epoch": 0.6419539608775532,
      "grad_norm": 0.4815426170825958,
      "learning_rate": 8e-05,
      "loss": 1.7121,
      "step": 5940
    },
    {
      "epoch": 0.64206203393494,
      "grad_norm": 0.434255450963974,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 5941
    },
    {
      "epoch": 0.6421701069923268,
      "grad_norm": 0.4768272638320923,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 5942
    },
    {
      "epoch": 0.6422781800497136,
      "grad_norm": 0.43197935819625854,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 5943
    },
    {
      "epoch": 0.6423862531071004,
      "grad_norm": 0.4227067232131958,
      "learning_rate": 8e-05,
      "loss": 1.5713,
      "step": 5944
    },
    {
      "epoch": 0.6424943261644872,
      "grad_norm": 0.4498826265335083,
      "learning_rate": 8e-05,
      "loss": 1.6921,
      "step": 5945
    },
    {
      "epoch": 0.642602399221874,
      "grad_norm": 0.4489765763282776,
      "learning_rate": 8e-05,
      "loss": 1.7311,
      "step": 5946
    },
    {
      "epoch": 0.6427104722792608,
      "grad_norm": 0.4814077615737915,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 5947
    },
    {
      "epoch": 0.6428185453366476,
      "grad_norm": 0.4516551196575165,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 5948
    },
    {
      "epoch": 0.6429266183940344,
      "grad_norm": 0.4329946041107178,
      "learning_rate": 8e-05,
      "loss": 1.584,
      "step": 5949
    },
    {
      "epoch": 0.6430346914514211,
      "grad_norm": 0.5322399735450745,
      "learning_rate": 8e-05,
      "loss": 1.9674,
      "step": 5950
    },
    {
      "epoch": 0.6431427645088079,
      "grad_norm": 0.43984171748161316,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 5951
    },
    {
      "epoch": 0.6432508375661947,
      "grad_norm": 0.4450085461139679,
      "learning_rate": 8e-05,
      "loss": 1.6514,
      "step": 5952
    },
    {
      "epoch": 0.6433589106235815,
      "grad_norm": 0.4723295271396637,
      "learning_rate": 8e-05,
      "loss": 1.6268,
      "step": 5953
    },
    {
      "epoch": 0.6434669836809683,
      "grad_norm": 0.42486312985420227,
      "learning_rate": 8e-05,
      "loss": 1.468,
      "step": 5954
    },
    {
      "epoch": 0.6435750567383551,
      "grad_norm": 0.430436909198761,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 5955
    },
    {
      "epoch": 0.6436831297957419,
      "grad_norm": 0.44935479760169983,
      "learning_rate": 8e-05,
      "loss": 1.5943,
      "step": 5956
    },
    {
      "epoch": 0.6437912028531287,
      "grad_norm": 0.4501962959766388,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 5957
    },
    {
      "epoch": 0.6438992759105155,
      "grad_norm": 0.4817235767841339,
      "learning_rate": 8e-05,
      "loss": 1.5903,
      "step": 5958
    },
    {
      "epoch": 0.6440073489679023,
      "grad_norm": 0.4311402142047882,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 5959
    },
    {
      "epoch": 0.644115422025289,
      "grad_norm": 0.49984273314476013,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 5960
    },
    {
      "epoch": 0.6442234950826758,
      "grad_norm": 0.5932108759880066,
      "learning_rate": 8e-05,
      "loss": 1.983,
      "step": 5961
    },
    {
      "epoch": 0.6443315681400626,
      "grad_norm": 0.4644288420677185,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 5962
    },
    {
      "epoch": 0.6444396411974495,
      "grad_norm": 0.443280965089798,
      "learning_rate": 8e-05,
      "loss": 1.6403,
      "step": 5963
    },
    {
      "epoch": 0.6445477142548363,
      "grad_norm": 0.4332227110862732,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 5964
    },
    {
      "epoch": 0.6446557873122231,
      "grad_norm": 0.44066137075424194,
      "learning_rate": 8e-05,
      "loss": 1.6621,
      "step": 5965
    },
    {
      "epoch": 0.6447638603696099,
      "grad_norm": 0.4439776539802551,
      "learning_rate": 8e-05,
      "loss": 1.6416,
      "step": 5966
    },
    {
      "epoch": 0.6448719334269967,
      "grad_norm": 0.4665268063545227,
      "learning_rate": 8e-05,
      "loss": 1.7,
      "step": 5967
    },
    {
      "epoch": 0.6449800064843835,
      "grad_norm": 0.47857311367988586,
      "learning_rate": 8e-05,
      "loss": 1.6068,
      "step": 5968
    },
    {
      "epoch": 0.6450880795417703,
      "grad_norm": 0.43485772609710693,
      "learning_rate": 8e-05,
      "loss": 1.3372,
      "step": 5969
    },
    {
      "epoch": 0.6451961525991571,
      "grad_norm": 0.45961129665374756,
      "learning_rate": 8e-05,
      "loss": 1.7506,
      "step": 5970
    },
    {
      "epoch": 0.6453042256565438,
      "grad_norm": 0.44660940766334534,
      "learning_rate": 8e-05,
      "loss": 1.4899,
      "step": 5971
    },
    {
      "epoch": 0.6454122987139306,
      "grad_norm": 0.4821637272834778,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 5972
    },
    {
      "epoch": 0.6455203717713174,
      "grad_norm": 0.4739539921283722,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 5973
    },
    {
      "epoch": 0.6456284448287042,
      "grad_norm": 0.452079713344574,
      "learning_rate": 8e-05,
      "loss": 1.6477,
      "step": 5974
    },
    {
      "epoch": 0.645736517886091,
      "grad_norm": 0.4875281751155853,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 5975
    },
    {
      "epoch": 0.6458445909434778,
      "grad_norm": 0.46676871180534363,
      "learning_rate": 8e-05,
      "loss": 1.5501,
      "step": 5976
    },
    {
      "epoch": 0.6459526640008646,
      "grad_norm": 0.49397730827331543,
      "learning_rate": 8e-05,
      "loss": 1.9117,
      "step": 5977
    },
    {
      "epoch": 0.6460607370582514,
      "grad_norm": 0.49007442593574524,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 5978
    },
    {
      "epoch": 0.6461688101156382,
      "grad_norm": 0.4650939702987671,
      "learning_rate": 8e-05,
      "loss": 1.5179,
      "step": 5979
    },
    {
      "epoch": 0.646276883173025,
      "grad_norm": 0.49104219675064087,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 5980
    },
    {
      "epoch": 0.6463849562304118,
      "grad_norm": 0.5339421033859253,
      "learning_rate": 8e-05,
      "loss": 1.8928,
      "step": 5981
    },
    {
      "epoch": 0.6464930292877985,
      "grad_norm": 0.5299485921859741,
      "learning_rate": 8e-05,
      "loss": 1.8438,
      "step": 5982
    },
    {
      "epoch": 0.6466011023451853,
      "grad_norm": 0.49085965752601624,
      "learning_rate": 8e-05,
      "loss": 1.9302,
      "step": 5983
    },
    {
      "epoch": 0.6467091754025721,
      "grad_norm": 0.4650600850582123,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 5984
    },
    {
      "epoch": 0.6468172484599589,
      "grad_norm": 0.48735931515693665,
      "learning_rate": 8e-05,
      "loss": 1.7303,
      "step": 5985
    },
    {
      "epoch": 0.6469253215173457,
      "grad_norm": 0.5607811808586121,
      "learning_rate": 8e-05,
      "loss": 1.8891,
      "step": 5986
    },
    {
      "epoch": 0.6470333945747325,
      "grad_norm": 0.4745977818965912,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 5987
    },
    {
      "epoch": 0.6471414676321193,
      "grad_norm": 0.48732292652130127,
      "learning_rate": 8e-05,
      "loss": 1.734,
      "step": 5988
    },
    {
      "epoch": 0.6472495406895061,
      "grad_norm": 0.49691468477249146,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 5989
    },
    {
      "epoch": 0.6473576137468929,
      "grad_norm": 0.499629408121109,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 5990
    },
    {
      "epoch": 0.6474656868042797,
      "grad_norm": 0.43888160586357117,
      "learning_rate": 8e-05,
      "loss": 1.643,
      "step": 5991
    },
    {
      "epoch": 0.6475737598616664,
      "grad_norm": 0.4917463958263397,
      "learning_rate": 8e-05,
      "loss": 1.6043,
      "step": 5992
    },
    {
      "epoch": 0.6476818329190532,
      "grad_norm": 0.4818088710308075,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 5993
    },
    {
      "epoch": 0.64778990597644,
      "grad_norm": 0.4924224615097046,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 5994
    },
    {
      "epoch": 0.6478979790338268,
      "grad_norm": 0.47290971875190735,
      "learning_rate": 8e-05,
      "loss": 1.716,
      "step": 5995
    },
    {
      "epoch": 0.6480060520912136,
      "grad_norm": 0.4421630799770355,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 5996
    },
    {
      "epoch": 0.6481141251486005,
      "grad_norm": 0.44072914123535156,
      "learning_rate": 8e-05,
      "loss": 1.4917,
      "step": 5997
    },
    {
      "epoch": 0.6482221982059873,
      "grad_norm": 0.5024055242538452,
      "learning_rate": 8e-05,
      "loss": 1.4718,
      "step": 5998
    },
    {
      "epoch": 0.6483302712633741,
      "grad_norm": 0.4409164488315582,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 5999
    },
    {
      "epoch": 0.6484383443207609,
      "grad_norm": 0.4403681457042694,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 6000
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 9253,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 8.952125425975296e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}