{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.5403652869339673,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00010807305738679347,
      "grad_norm": 0.4351649880409241,
      "learning_rate": 8e-05,
      "loss": 1.8983,
      "step": 1
    },
    {
      "epoch": 0.00021614611477358694,
      "grad_norm": 0.5632684230804443,
      "learning_rate": 8e-05,
      "loss": 2.0631,
      "step": 2
    },
    {
      "epoch": 0.0003242191721603804,
      "grad_norm": 0.4193551540374756,
      "learning_rate": 8e-05,
      "loss": 1.9084,
      "step": 3
    },
    {
      "epoch": 0.0004322922295471739,
      "grad_norm": 0.43743789196014404,
      "learning_rate": 8e-05,
      "loss": 2.0345,
      "step": 4
    },
    {
      "epoch": 0.0005403652869339674,
      "grad_norm": 0.35983482003211975,
      "learning_rate": 8e-05,
      "loss": 1.9538,
      "step": 5
    },
    {
      "epoch": 0.0006484383443207608,
      "grad_norm": 0.37199866771698,
      "learning_rate": 8e-05,
      "loss": 2.1294,
      "step": 6
    },
    {
      "epoch": 0.0007565114017075543,
      "grad_norm": 0.35996878147125244,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 7
    },
    {
      "epoch": 0.0008645844590943478,
      "grad_norm": 0.40022915601730347,
      "learning_rate": 8e-05,
      "loss": 2.0355,
      "step": 8
    },
    {
      "epoch": 0.0009726575164811412,
      "grad_norm": 0.4122651517391205,
      "learning_rate": 8e-05,
      "loss": 1.9823,
      "step": 9
    },
    {
      "epoch": 0.0010807305738679347,
      "grad_norm": 0.43956854939460754,
      "learning_rate": 8e-05,
      "loss": 1.858,
      "step": 10
    },
    {
      "epoch": 0.0011888036312547282,
      "grad_norm": 0.398956835269928,
      "learning_rate": 8e-05,
      "loss": 1.9997,
      "step": 11
    },
    {
      "epoch": 0.0012968766886415216,
      "grad_norm": 0.34920763969421387,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 12
    },
    {
      "epoch": 0.001404949746028315,
      "grad_norm": 0.35869136452674866,
      "learning_rate": 8e-05,
      "loss": 1.8985,
      "step": 13
    },
    {
      "epoch": 0.0015130228034151087,
      "grad_norm": 0.32896265387535095,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 14
    },
    {
      "epoch": 0.0016210958608019021,
      "grad_norm": 0.36838844418525696,
      "learning_rate": 8e-05,
      "loss": 1.8666,
      "step": 15
    },
    {
      "epoch": 0.0017291689181886955,
      "grad_norm": 0.3589898943901062,
      "learning_rate": 8e-05,
      "loss": 1.8331,
      "step": 16
    },
    {
      "epoch": 0.001837241975575489,
      "grad_norm": 0.3534690737724304,
      "learning_rate": 8e-05,
      "loss": 1.8592,
      "step": 17
    },
    {
      "epoch": 0.0019453150329622824,
      "grad_norm": 0.33605045080184937,
      "learning_rate": 8e-05,
      "loss": 1.8403,
      "step": 18
    },
    {
      "epoch": 0.002053388090349076,
      "grad_norm": 0.38296014070510864,
      "learning_rate": 8e-05,
      "loss": 2.063,
      "step": 19
    },
    {
      "epoch": 0.0021614611477358695,
      "grad_norm": 0.356327623128891,
      "learning_rate": 8e-05,
      "loss": 1.8731,
      "step": 20
    },
    {
      "epoch": 0.002269534205122663,
      "grad_norm": 0.3825637102127075,
      "learning_rate": 8e-05,
      "loss": 1.8076,
      "step": 21
    },
    {
      "epoch": 0.0023776072625094563,
      "grad_norm": 0.3552597463130951,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 22
    },
    {
      "epoch": 0.0024856803198962498,
      "grad_norm": 0.3945937156677246,
      "learning_rate": 8e-05,
      "loss": 2.0094,
      "step": 23
    },
    {
      "epoch": 0.002593753377283043,
      "grad_norm": 0.3497529923915863,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 24
    },
    {
      "epoch": 0.0027018264346698366,
      "grad_norm": 0.35496705770492554,
      "learning_rate": 8e-05,
      "loss": 1.9212,
      "step": 25
    },
    {
      "epoch": 0.00280989949205663,
      "grad_norm": 0.40153196454048157,
      "learning_rate": 8e-05,
      "loss": 2.0488,
      "step": 26
    },
    {
      "epoch": 0.002917972549443424,
      "grad_norm": 0.3893981873989105,
      "learning_rate": 8e-05,
      "loss": 1.8708,
      "step": 27
    },
    {
      "epoch": 0.0030260456068302174,
      "grad_norm": 0.3695599138736725,
      "learning_rate": 8e-05,
      "loss": 2.0791,
      "step": 28
    },
    {
      "epoch": 0.003134118664217011,
      "grad_norm": 0.3641146421432495,
      "learning_rate": 8e-05,
      "loss": 1.9072,
      "step": 29
    },
    {
      "epoch": 0.0032421917216038042,
      "grad_norm": 0.3596847355365753,
      "learning_rate": 8e-05,
      "loss": 1.8498,
      "step": 30
    },
    {
      "epoch": 0.0033502647789905977,
      "grad_norm": 0.3638761341571808,
      "learning_rate": 8e-05,
      "loss": 1.8096,
      "step": 31
    },
    {
      "epoch": 0.003458337836377391,
      "grad_norm": 0.39277321100234985,
      "learning_rate": 8e-05,
      "loss": 1.8466,
      "step": 32
    },
    {
      "epoch": 0.0035664108937641845,
      "grad_norm": 0.3584340214729309,
      "learning_rate": 8e-05,
      "loss": 1.9072,
      "step": 33
    },
    {
      "epoch": 0.003674483951150978,
      "grad_norm": 0.38852962851524353,
      "learning_rate": 8e-05,
      "loss": 1.9932,
      "step": 34
    },
    {
      "epoch": 0.0037825570085377714,
      "grad_norm": 0.39563801884651184,
      "learning_rate": 8e-05,
      "loss": 1.8998,
      "step": 35
    },
    {
      "epoch": 0.003890630065924565,
      "grad_norm": 0.3966459035873413,
      "learning_rate": 8e-05,
      "loss": 1.941,
      "step": 36
    },
    {
      "epoch": 0.003998703123311359,
      "grad_norm": 0.3564274311065674,
      "learning_rate": 8e-05,
      "loss": 1.8725,
      "step": 37
    },
    {
      "epoch": 0.004106776180698152,
      "grad_norm": 0.34590858221054077,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 38
    },
    {
      "epoch": 0.0042148492380849455,
      "grad_norm": 0.3475266993045807,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 39
    },
    {
      "epoch": 0.004322922295471739,
      "grad_norm": 0.36428943276405334,
      "learning_rate": 8e-05,
      "loss": 1.8987,
      "step": 40
    },
    {
      "epoch": 0.004430995352858532,
      "grad_norm": 0.36737462878227234,
      "learning_rate": 8e-05,
      "loss": 1.9232,
      "step": 41
    },
    {
      "epoch": 0.004539068410245326,
      "grad_norm": 0.41655054688453674,
      "learning_rate": 8e-05,
      "loss": 1.9765,
      "step": 42
    },
    {
      "epoch": 0.004647141467632119,
      "grad_norm": 0.37395918369293213,
      "learning_rate": 8e-05,
      "loss": 1.9251,
      "step": 43
    },
    {
      "epoch": 0.004755214525018913,
      "grad_norm": 0.36078399419784546,
      "learning_rate": 8e-05,
      "loss": 1.9252,
      "step": 44
    },
    {
      "epoch": 0.004863287582405706,
      "grad_norm": 0.37375158071517944,
      "learning_rate": 8e-05,
      "loss": 1.9898,
      "step": 45
    },
    {
      "epoch": 0.0049713606397924996,
      "grad_norm": 0.35973864793777466,
      "learning_rate": 8e-05,
      "loss": 1.8468,
      "step": 46
    },
    {
      "epoch": 0.005079433697179293,
      "grad_norm": 0.33282211422920227,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 47
    },
    {
      "epoch": 0.005187506754566086,
      "grad_norm": 0.3648369312286377,
      "learning_rate": 8e-05,
      "loss": 1.8562,
      "step": 48
    },
    {
      "epoch": 0.00529557981195288,
      "grad_norm": 0.3680543005466461,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 49
    },
    {
      "epoch": 0.005403652869339673,
      "grad_norm": 0.3996305465698242,
      "learning_rate": 8e-05,
      "loss": 2.0178,
      "step": 50
    },
    {
      "epoch": 0.005511725926726467,
      "grad_norm": 0.34245601296424866,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 51
    },
    {
      "epoch": 0.00561979898411326,
      "grad_norm": 0.4385335147380829,
      "learning_rate": 8e-05,
      "loss": 2.1892,
      "step": 52
    },
    {
      "epoch": 0.005727872041500054,
      "grad_norm": 0.37781935930252075,
      "learning_rate": 8e-05,
      "loss": 1.8742,
      "step": 53
    },
    {
      "epoch": 0.005835945098886848,
      "grad_norm": 0.35104045271873474,
      "learning_rate": 8e-05,
      "loss": 1.8235,
      "step": 54
    },
    {
      "epoch": 0.005944018156273641,
      "grad_norm": 0.3527902066707611,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 55
    },
    {
      "epoch": 0.006052091213660435,
      "grad_norm": 0.3233564794063568,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 56
    },
    {
      "epoch": 0.006160164271047228,
      "grad_norm": 0.36856353282928467,
      "learning_rate": 8e-05,
      "loss": 1.7645,
      "step": 57
    },
    {
      "epoch": 0.006268237328434022,
      "grad_norm": 0.4024446904659271,
      "learning_rate": 8e-05,
      "loss": 1.939,
      "step": 58
    },
    {
      "epoch": 0.006376310385820815,
      "grad_norm": 0.3551691174507141,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 59
    },
    {
      "epoch": 0.0064843834432076084,
      "grad_norm": 0.44394415616989136,
      "learning_rate": 8e-05,
      "loss": 2.2443,
      "step": 60
    },
    {
      "epoch": 0.006592456500594402,
      "grad_norm": 0.3416070342063904,
      "learning_rate": 8e-05,
      "loss": 1.5767,
      "step": 61
    },
    {
      "epoch": 0.006700529557981195,
      "grad_norm": 0.3482869565486908,
      "learning_rate": 8e-05,
      "loss": 1.8669,
      "step": 62
    },
    {
      "epoch": 0.006808602615367989,
      "grad_norm": 0.3702720105648041,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 63
    },
    {
      "epoch": 0.006916675672754782,
      "grad_norm": 0.36599811911582947,
      "learning_rate": 8e-05,
      "loss": 2.0098,
      "step": 64
    },
    {
      "epoch": 0.007024748730141576,
      "grad_norm": 0.3971049189567566,
      "learning_rate": 8e-05,
      "loss": 2.0508,
      "step": 65
    },
    {
      "epoch": 0.007132821787528369,
      "grad_norm": 0.35264477133750916,
      "learning_rate": 8e-05,
      "loss": 1.7954,
      "step": 66
    },
    {
      "epoch": 0.0072408948449151625,
      "grad_norm": 0.3791993260383606,
      "learning_rate": 8e-05,
      "loss": 1.9896,
      "step": 67
    },
    {
      "epoch": 0.007348967902301956,
      "grad_norm": 0.37191087007522583,
      "learning_rate": 8e-05,
      "loss": 1.9054,
      "step": 68
    },
    {
      "epoch": 0.007457040959688749,
      "grad_norm": 0.4020589590072632,
      "learning_rate": 8e-05,
      "loss": 1.7792,
      "step": 69
    },
    {
      "epoch": 0.007565114017075543,
      "grad_norm": 0.34911972284317017,
      "learning_rate": 8e-05,
      "loss": 1.8529,
      "step": 70
    },
    {
      "epoch": 0.007673187074462336,
      "grad_norm": 0.37512168288230896,
      "learning_rate": 8e-05,
      "loss": 1.9126,
      "step": 71
    },
    {
      "epoch": 0.00778126013184913,
      "grad_norm": 0.3415440618991852,
      "learning_rate": 8e-05,
      "loss": 1.8138,
      "step": 72
    },
    {
      "epoch": 0.007889333189235924,
      "grad_norm": 0.3596620559692383,
      "learning_rate": 8e-05,
      "loss": 1.959,
      "step": 73
    },
    {
      "epoch": 0.007997406246622717,
      "grad_norm": 0.3739068806171417,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 74
    },
    {
      "epoch": 0.00810547930400951,
      "grad_norm": 0.33997228741645813,
      "learning_rate": 8e-05,
      "loss": 1.8809,
      "step": 75
    },
    {
      "epoch": 0.008213552361396304,
      "grad_norm": 0.3936220407485962,
      "learning_rate": 8e-05,
      "loss": 1.9793,
      "step": 76
    },
    {
      "epoch": 0.008321625418783098,
      "grad_norm": 0.3442775011062622,
      "learning_rate": 8e-05,
      "loss": 1.8598,
      "step": 77
    },
    {
      "epoch": 0.008429698476169891,
      "grad_norm": 0.34543630480766296,
      "learning_rate": 8e-05,
      "loss": 1.8743,
      "step": 78
    },
    {
      "epoch": 0.008537771533556685,
      "grad_norm": 0.3811757266521454,
      "learning_rate": 8e-05,
      "loss": 1.9644,
      "step": 79
    },
    {
      "epoch": 0.008645844590943478,
      "grad_norm": 0.3636032044887543,
      "learning_rate": 8e-05,
      "loss": 1.9381,
      "step": 80
    },
    {
      "epoch": 0.008753917648330271,
      "grad_norm": 0.3840232193470001,
      "learning_rate": 8e-05,
      "loss": 2.0628,
      "step": 81
    },
    {
      "epoch": 0.008861990705717065,
      "grad_norm": 0.3367981016635895,
      "learning_rate": 8e-05,
      "loss": 1.782,
      "step": 82
    },
    {
      "epoch": 0.008970063763103858,
      "grad_norm": 0.33157339692115784,
      "learning_rate": 8e-05,
      "loss": 1.8371,
      "step": 83
    },
    {
      "epoch": 0.009078136820490652,
      "grad_norm": 0.3375696837902069,
      "learning_rate": 8e-05,
      "loss": 1.8922,
      "step": 84
    },
    {
      "epoch": 0.009186209877877445,
      "grad_norm": 0.3356796205043793,
      "learning_rate": 8e-05,
      "loss": 1.6773,
      "step": 85
    },
    {
      "epoch": 0.009294282935264239,
      "grad_norm": 0.3588196635246277,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 86
    },
    {
      "epoch": 0.009402355992651032,
      "grad_norm": 0.3561270534992218,
      "learning_rate": 8e-05,
      "loss": 1.9395,
      "step": 87
    },
    {
      "epoch": 0.009510429050037825,
      "grad_norm": 0.3281303942203522,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 88
    },
    {
      "epoch": 0.009618502107424619,
      "grad_norm": 0.34098535776138306,
      "learning_rate": 8e-05,
      "loss": 1.7778,
      "step": 89
    },
    {
      "epoch": 0.009726575164811412,
      "grad_norm": 0.3468838334083557,
      "learning_rate": 8e-05,
      "loss": 1.7544,
      "step": 90
    },
    {
      "epoch": 0.009834648222198206,
      "grad_norm": 0.3873237073421478,
      "learning_rate": 8e-05,
      "loss": 1.8939,
      "step": 91
    },
    {
      "epoch": 0.009942721279584999,
      "grad_norm": 0.34803327918052673,
      "learning_rate": 8e-05,
      "loss": 1.923,
      "step": 92
    },
    {
      "epoch": 0.010050794336971793,
      "grad_norm": 0.3591421842575073,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 93
    },
    {
      "epoch": 0.010158867394358586,
      "grad_norm": 0.3685607314109802,
      "learning_rate": 8e-05,
      "loss": 1.8363,
      "step": 94
    },
    {
      "epoch": 0.01026694045174538,
      "grad_norm": 0.4085378050804138,
      "learning_rate": 8e-05,
      "loss": 2.0675,
      "step": 95
    },
    {
      "epoch": 0.010375013509132173,
      "grad_norm": 0.33424437046051025,
      "learning_rate": 8e-05,
      "loss": 1.8288,
      "step": 96
    },
    {
      "epoch": 0.010483086566518966,
      "grad_norm": 0.3550828993320465,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 97
    },
    {
      "epoch": 0.01059115962390576,
      "grad_norm": 0.3696897327899933,
      "learning_rate": 8e-05,
      "loss": 1.9669,
      "step": 98
    },
    {
      "epoch": 0.010699232681292553,
      "grad_norm": 0.33994463086128235,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 99
    },
    {
      "epoch": 0.010807305738679347,
      "grad_norm": 0.36720365285873413,
      "learning_rate": 8e-05,
      "loss": 2.028,
      "step": 100
    },
    {
      "epoch": 0.01091537879606614,
      "grad_norm": 0.3576091527938843,
      "learning_rate": 8e-05,
      "loss": 1.6761,
      "step": 101
    },
    {
      "epoch": 0.011023451853452933,
      "grad_norm": 0.3334646224975586,
      "learning_rate": 8e-05,
      "loss": 1.7336,
      "step": 102
    },
    {
      "epoch": 0.011131524910839727,
      "grad_norm": 0.39445286989212036,
      "learning_rate": 8e-05,
      "loss": 1.9391,
      "step": 103
    },
    {
      "epoch": 0.01123959796822652,
      "grad_norm": 0.34280478954315186,
      "learning_rate": 8e-05,
      "loss": 1.933,
      "step": 104
    },
    {
      "epoch": 0.011347671025613315,
      "grad_norm": 0.36647966504096985,
      "learning_rate": 8e-05,
      "loss": 1.9672,
      "step": 105
    },
    {
      "epoch": 0.011455744083000109,
      "grad_norm": 0.4370229244232178,
      "learning_rate": 8e-05,
      "loss": 1.7903,
      "step": 106
    },
    {
      "epoch": 0.011563817140386902,
      "grad_norm": 0.38026702404022217,
      "learning_rate": 8e-05,
      "loss": 1.9277,
      "step": 107
    },
    {
      "epoch": 0.011671890197773696,
      "grad_norm": 0.3587406277656555,
      "learning_rate": 8e-05,
      "loss": 1.9038,
      "step": 108
    },
    {
      "epoch": 0.01177996325516049,
      "grad_norm": 0.3514784872531891,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 109
    },
    {
      "epoch": 0.011888036312547283,
      "grad_norm": 0.3730924427509308,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 110
    },
    {
      "epoch": 0.011996109369934076,
      "grad_norm": 0.3412942886352539,
      "learning_rate": 8e-05,
      "loss": 1.8384,
      "step": 111
    },
    {
      "epoch": 0.01210418242732087,
      "grad_norm": 0.36606597900390625,
      "learning_rate": 8e-05,
      "loss": 1.9919,
      "step": 112
    },
    {
      "epoch": 0.012212255484707663,
      "grad_norm": 0.3667832911014557,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 113
    },
    {
      "epoch": 0.012320328542094456,
      "grad_norm": 0.3357333540916443,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 114
    },
    {
      "epoch": 0.01242840159948125,
      "grad_norm": 0.3732669949531555,
      "learning_rate": 8e-05,
      "loss": 1.941,
      "step": 115
    },
    {
      "epoch": 0.012536474656868043,
      "grad_norm": 0.48029825091362,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 116
    },
    {
      "epoch": 0.012644547714254837,
      "grad_norm": 0.3361777663230896,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 117
    },
    {
      "epoch": 0.01275262077164163,
      "grad_norm": 0.3681136667728424,
      "learning_rate": 8e-05,
      "loss": 1.8102,
      "step": 118
    },
    {
      "epoch": 0.012860693829028423,
      "grad_norm": 0.4084520637989044,
      "learning_rate": 8e-05,
      "loss": 1.9245,
      "step": 119
    },
    {
      "epoch": 0.012968766886415217,
      "grad_norm": 0.34002935886383057,
      "learning_rate": 8e-05,
      "loss": 1.8387,
      "step": 120
    },
    {
      "epoch": 0.01307683994380201,
      "grad_norm": 0.35462895035743713,
      "learning_rate": 8e-05,
      "loss": 1.912,
      "step": 121
    },
    {
      "epoch": 0.013184913001188804,
      "grad_norm": 0.3921608328819275,
      "learning_rate": 8e-05,
      "loss": 1.8677,
      "step": 122
    },
    {
      "epoch": 0.013292986058575597,
      "grad_norm": 0.3691435754299164,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 123
    },
    {
      "epoch": 0.01340105911596239,
      "grad_norm": 0.3534267544746399,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 124
    },
    {
      "epoch": 0.013509132173349184,
      "grad_norm": 0.35075145959854126,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 125
    },
    {
      "epoch": 0.013617205230735977,
      "grad_norm": 0.37121960520744324,
      "learning_rate": 8e-05,
      "loss": 1.8728,
      "step": 126
    },
    {
      "epoch": 0.013725278288122771,
      "grad_norm": 0.4479074478149414,
      "learning_rate": 8e-05,
      "loss": 1.972,
      "step": 127
    },
    {
      "epoch": 0.013833351345509564,
      "grad_norm": 0.3476560413837433,
      "learning_rate": 8e-05,
      "loss": 1.8225,
      "step": 128
    },
    {
      "epoch": 0.013941424402896358,
      "grad_norm": 0.3355270028114319,
      "learning_rate": 8e-05,
      "loss": 1.8436,
      "step": 129
    },
    {
      "epoch": 0.014049497460283151,
      "grad_norm": 0.42383939027786255,
      "learning_rate": 8e-05,
      "loss": 1.9533,
      "step": 130
    },
    {
      "epoch": 0.014157570517669945,
      "grad_norm": 0.3658997416496277,
      "learning_rate": 8e-05,
      "loss": 1.943,
      "step": 131
    },
    {
      "epoch": 0.014265643575056738,
      "grad_norm": 0.38565921783447266,
      "learning_rate": 8e-05,
      "loss": 1.9349,
      "step": 132
    },
    {
      "epoch": 0.014373716632443531,
      "grad_norm": 0.390728622674942,
      "learning_rate": 8e-05,
      "loss": 1.882,
      "step": 133
    },
    {
      "epoch": 0.014481789689830325,
      "grad_norm": 0.37281593680381775,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 134
    },
    {
      "epoch": 0.014589862747217118,
      "grad_norm": 0.42595094442367554,
      "learning_rate": 8e-05,
      "loss": 1.9489,
      "step": 135
    },
    {
      "epoch": 0.014697935804603912,
      "grad_norm": 0.3504122495651245,
      "learning_rate": 8e-05,
      "loss": 1.8774,
      "step": 136
    },
    {
      "epoch": 0.014806008861990705,
      "grad_norm": 0.4228856861591339,
      "learning_rate": 8e-05,
      "loss": 2.0514,
      "step": 137
    },
    {
      "epoch": 0.014914081919377499,
      "grad_norm": 0.3680881857872009,
      "learning_rate": 8e-05,
      "loss": 1.86,
      "step": 138
    },
    {
      "epoch": 0.015022154976764292,
      "grad_norm": 0.344536155462265,
      "learning_rate": 8e-05,
      "loss": 1.8319,
      "step": 139
    },
    {
      "epoch": 0.015130228034151086,
      "grad_norm": 0.3578692376613617,
      "learning_rate": 8e-05,
      "loss": 1.8546,
      "step": 140
    },
    {
      "epoch": 0.015238301091537879,
      "grad_norm": 0.42693620920181274,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 141
    },
    {
      "epoch": 0.015346374148924672,
      "grad_norm": 0.4166564345359802,
      "learning_rate": 8e-05,
      "loss": 1.9141,
      "step": 142
    },
    {
      "epoch": 0.015454447206311466,
      "grad_norm": 0.38682886958122253,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 143
    },
    {
      "epoch": 0.01556252026369826,
      "grad_norm": 0.33764222264289856,
      "learning_rate": 8e-05,
      "loss": 1.7668,
      "step": 144
    },
    {
      "epoch": 0.015670593321085054,
      "grad_norm": 0.3671913146972656,
      "learning_rate": 8e-05,
      "loss": 1.8183,
      "step": 145
    },
    {
      "epoch": 0.015778666378471848,
      "grad_norm": 0.39653363823890686,
      "learning_rate": 8e-05,
      "loss": 1.9931,
      "step": 146
    },
    {
      "epoch": 0.01588673943585864,
      "grad_norm": 0.39502009749412537,
      "learning_rate": 8e-05,
      "loss": 1.8918,
      "step": 147
    },
    {
      "epoch": 0.015994812493245435,
      "grad_norm": 0.35358163714408875,
      "learning_rate": 8e-05,
      "loss": 1.9298,
      "step": 148
    },
    {
      "epoch": 0.016102885550632228,
      "grad_norm": 0.35190215706825256,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 149
    },
    {
      "epoch": 0.01621095860801902,
      "grad_norm": 0.3853912055492401,
      "learning_rate": 8e-05,
      "loss": 1.8762,
      "step": 150
    },
    {
      "epoch": 0.016319031665405815,
      "grad_norm": 0.3836837410926819,
      "learning_rate": 8e-05,
      "loss": 1.8991,
      "step": 151
    },
    {
      "epoch": 0.01642710472279261,
      "grad_norm": 0.3867194950580597,
      "learning_rate": 8e-05,
      "loss": 2.0555,
      "step": 152
    },
    {
      "epoch": 0.016535177780179402,
      "grad_norm": 0.3466987907886505,
      "learning_rate": 8e-05,
      "loss": 1.7332,
      "step": 153
    },
    {
      "epoch": 0.016643250837566195,
      "grad_norm": 0.4112001359462738,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 154
    },
    {
      "epoch": 0.01675132389495299,
      "grad_norm": 0.36368218064308167,
      "learning_rate": 8e-05,
      "loss": 1.8836,
      "step": 155
    },
    {
      "epoch": 0.016859396952339782,
      "grad_norm": 0.343652606010437,
      "learning_rate": 8e-05,
      "loss": 1.763,
      "step": 156
    },
    {
      "epoch": 0.016967470009726576,
      "grad_norm": 0.3557482957839966,
      "learning_rate": 8e-05,
      "loss": 1.8708,
      "step": 157
    },
    {
      "epoch": 0.01707554306711337,
      "grad_norm": 0.3625258207321167,
      "learning_rate": 8e-05,
      "loss": 1.86,
      "step": 158
    },
    {
      "epoch": 0.017183616124500162,
      "grad_norm": 0.4034252166748047,
      "learning_rate": 8e-05,
      "loss": 2.0207,
      "step": 159
    },
    {
      "epoch": 0.017291689181886956,
      "grad_norm": 0.3633774220943451,
      "learning_rate": 8e-05,
      "loss": 1.8668,
      "step": 160
    },
    {
      "epoch": 0.01739976223927375,
      "grad_norm": 0.326499342918396,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 161
    },
    {
      "epoch": 0.017507835296660543,
      "grad_norm": 0.33330976963043213,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 162
    },
    {
      "epoch": 0.017615908354047336,
      "grad_norm": 0.37760624289512634,
      "learning_rate": 8e-05,
      "loss": 1.9396,
      "step": 163
    },
    {
      "epoch": 0.01772398141143413,
      "grad_norm": 0.32992613315582275,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 164
    },
    {
      "epoch": 0.017832054468820923,
      "grad_norm": 0.35324472188949585,
      "learning_rate": 8e-05,
      "loss": 1.9957,
      "step": 165
    },
    {
      "epoch": 0.017940127526207716,
      "grad_norm": 0.33778947591781616,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 166
    },
    {
      "epoch": 0.01804820058359451,
      "grad_norm": 0.35893818736076355,
      "learning_rate": 8e-05,
      "loss": 1.9277,
      "step": 167
    },
    {
      "epoch": 0.018156273640981303,
      "grad_norm": 0.37994202971458435,
      "learning_rate": 8e-05,
      "loss": 2.0575,
      "step": 168
    },
    {
      "epoch": 0.018264346698368097,
      "grad_norm": 0.35307133197784424,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 169
    },
    {
      "epoch": 0.01837241975575489,
      "grad_norm": 0.3657943904399872,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 170
    },
    {
      "epoch": 0.018480492813141684,
      "grad_norm": 0.3480663597583771,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 171
    },
    {
      "epoch": 0.018588565870528477,
      "grad_norm": 0.5380162596702576,
      "learning_rate": 8e-05,
      "loss": 2.1205,
      "step": 172
    },
    {
      "epoch": 0.01869663892791527,
      "grad_norm": 0.3727838695049286,
      "learning_rate": 8e-05,
      "loss": 1.8845,
      "step": 173
    },
    {
      "epoch": 0.018804711985302064,
      "grad_norm": 0.3410198390483856,
      "learning_rate": 8e-05,
      "loss": 1.7974,
      "step": 174
    },
    {
      "epoch": 0.018912785042688857,
      "grad_norm": 0.33915817737579346,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 175
    },
    {
      "epoch": 0.01902085810007565,
      "grad_norm": 0.34816646575927734,
      "learning_rate": 8e-05,
      "loss": 1.8082,
      "step": 176
    },
    {
      "epoch": 0.019128931157462444,
      "grad_norm": 0.3584868311882019,
      "learning_rate": 8e-05,
      "loss": 1.8774,
      "step": 177
    },
    {
      "epoch": 0.019237004214849238,
      "grad_norm": 0.3616390526294708,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 178
    },
    {
      "epoch": 0.01934507727223603,
      "grad_norm": 0.3825621008872986,
      "learning_rate": 8e-05,
      "loss": 1.8342,
      "step": 179
    },
    {
      "epoch": 0.019453150329622824,
      "grad_norm": 0.37360844016075134,
      "learning_rate": 8e-05,
      "loss": 1.9066,
      "step": 180
    },
    {
      "epoch": 0.019561223387009618,
      "grad_norm": 0.3455824553966522,
      "learning_rate": 8e-05,
      "loss": 1.7934,
      "step": 181
    },
    {
      "epoch": 0.01966929644439641,
      "grad_norm": 0.35564401745796204,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 182
    },
    {
      "epoch": 0.019777369501783205,
      "grad_norm": 0.3567356467247009,
      "learning_rate": 8e-05,
      "loss": 1.854,
      "step": 183
    },
    {
      "epoch": 0.019885442559169998,
      "grad_norm": 0.3364497125148773,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 184
    },
    {
      "epoch": 0.01999351561655679,
      "grad_norm": 0.3673102855682373,
      "learning_rate": 8e-05,
      "loss": 2.044,
      "step": 185
    },
    {
      "epoch": 0.020101588673943585,
      "grad_norm": 0.3575238287448883,
      "learning_rate": 8e-05,
      "loss": 1.862,
      "step": 186
    },
    {
      "epoch": 0.02020966173133038,
      "grad_norm": 0.3294106423854828,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 187
    },
    {
      "epoch": 0.020317734788717172,
      "grad_norm": 0.36280909180641174,
      "learning_rate": 8e-05,
      "loss": 1.9266,
      "step": 188
    },
    {
      "epoch": 0.020425807846103965,
      "grad_norm": 0.3469875454902649,
      "learning_rate": 8e-05,
      "loss": 1.9141,
      "step": 189
    },
    {
      "epoch": 0.02053388090349076,
      "grad_norm": 0.35601651668548584,
      "learning_rate": 8e-05,
      "loss": 1.9005,
      "step": 190
    },
    {
      "epoch": 0.020641953960877552,
      "grad_norm": 0.3853970766067505,
      "learning_rate": 8e-05,
      "loss": 1.9129,
      "step": 191
    },
    {
      "epoch": 0.020750027018264346,
      "grad_norm": 0.37526872754096985,
      "learning_rate": 8e-05,
      "loss": 1.9607,
      "step": 192
    },
    {
      "epoch": 0.02085810007565114,
      "grad_norm": 0.36681321263313293,
      "learning_rate": 8e-05,
      "loss": 1.9772,
      "step": 193
    },
    {
      "epoch": 0.020966173133037933,
      "grad_norm": 0.3470933735370636,
      "learning_rate": 8e-05,
      "loss": 1.9596,
      "step": 194
    },
    {
      "epoch": 0.021074246190424726,
      "grad_norm": 0.3369242250919342,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 195
    },
    {
      "epoch": 0.02118231924781152,
      "grad_norm": 0.3203667402267456,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 196
    },
    {
      "epoch": 0.021290392305198313,
      "grad_norm": 0.3353847563266754,
      "learning_rate": 8e-05,
      "loss": 1.8597,
      "step": 197
    },
    {
      "epoch": 0.021398465362585106,
      "grad_norm": 0.3403714597225189,
      "learning_rate": 8e-05,
      "loss": 1.9415,
      "step": 198
    },
    {
      "epoch": 0.0215065384199719,
      "grad_norm": 0.33768534660339355,
      "learning_rate": 8e-05,
      "loss": 1.8627,
      "step": 199
    },
    {
      "epoch": 0.021614611477358693,
      "grad_norm": 0.3423212766647339,
      "learning_rate": 8e-05,
      "loss": 1.8255,
      "step": 200
    },
    {
      "epoch": 0.021722684534745487,
      "grad_norm": 0.33419105410575867,
      "learning_rate": 8e-05,
      "loss": 1.8485,
      "step": 201
    },
    {
      "epoch": 0.02183075759213228,
      "grad_norm": 0.31438854336738586,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 202
    },
    {
      "epoch": 0.021938830649519073,
      "grad_norm": 0.3373056650161743,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 203
    },
    {
      "epoch": 0.022046903706905867,
      "grad_norm": 0.3650558292865753,
      "learning_rate": 8e-05,
      "loss": 1.8846,
      "step": 204
    },
    {
      "epoch": 0.02215497676429266,
      "grad_norm": 0.3386206030845642,
      "learning_rate": 8e-05,
      "loss": 1.7638,
      "step": 205
    },
    {
      "epoch": 0.022263049821679454,
      "grad_norm": 0.35893094539642334,
      "learning_rate": 8e-05,
      "loss": 1.8466,
      "step": 206
    },
    {
      "epoch": 0.022371122879066247,
      "grad_norm": 0.33917978405952454,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 207
    },
    {
      "epoch": 0.02247919593645304,
      "grad_norm": 0.378278911113739,
      "learning_rate": 8e-05,
      "loss": 1.9303,
      "step": 208
    },
    {
      "epoch": 0.022587268993839837,
      "grad_norm": 0.3625754415988922,
      "learning_rate": 8e-05,
      "loss": 1.814,
      "step": 209
    },
    {
      "epoch": 0.02269534205122663,
      "grad_norm": 0.4043956696987152,
      "learning_rate": 8e-05,
      "loss": 2.0295,
      "step": 210
    },
    {
      "epoch": 0.022803415108613424,
      "grad_norm": 0.34033194184303284,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 211
    },
    {
      "epoch": 0.022911488166000218,
      "grad_norm": 0.35087522864341736,
      "learning_rate": 8e-05,
      "loss": 1.9093,
      "step": 212
    },
    {
      "epoch": 0.02301956122338701,
      "grad_norm": 0.3271329998970032,
      "learning_rate": 8e-05,
      "loss": 1.6803,
      "step": 213
    },
    {
      "epoch": 0.023127634280773805,
      "grad_norm": 0.3723451793193817,
      "learning_rate": 8e-05,
      "loss": 1.9597,
      "step": 214
    },
    {
      "epoch": 0.023235707338160598,
      "grad_norm": 0.36175742745399475,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 215
    },
    {
      "epoch": 0.02334378039554739,
      "grad_norm": 0.3611111044883728,
      "learning_rate": 8e-05,
      "loss": 1.7868,
      "step": 216
    },
    {
      "epoch": 0.023451853452934185,
      "grad_norm": 0.3544881343841553,
      "learning_rate": 8e-05,
      "loss": 1.802,
      "step": 217
    },
    {
      "epoch": 0.02355992651032098,
      "grad_norm": 0.3522435426712036,
      "learning_rate": 8e-05,
      "loss": 1.8659,
      "step": 218
    },
    {
      "epoch": 0.023667999567707772,
      "grad_norm": 0.3727189898490906,
      "learning_rate": 8e-05,
      "loss": 1.927,
      "step": 219
    },
    {
      "epoch": 0.023776072625094565,
      "grad_norm": 0.3585643768310547,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 220
    },
    {
      "epoch": 0.02388414568248136,
      "grad_norm": 0.3653928339481354,
      "learning_rate": 8e-05,
      "loss": 1.8583,
      "step": 221
    },
    {
      "epoch": 0.023992218739868152,
      "grad_norm": 0.3753211200237274,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 222
    },
    {
      "epoch": 0.024100291797254945,
      "grad_norm": 0.3349682688713074,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 223
    },
    {
      "epoch": 0.02420836485464174,
      "grad_norm": 0.3413155674934387,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 224
    },
    {
      "epoch": 0.024316437912028532,
      "grad_norm": 0.33893683552742004,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 225
    },
    {
      "epoch": 0.024424510969415326,
      "grad_norm": 0.3363328278064728,
      "learning_rate": 8e-05,
      "loss": 1.8412,
      "step": 226
    },
    {
      "epoch": 0.02453258402680212,
      "grad_norm": 0.3394852876663208,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 227
    },
    {
      "epoch": 0.024640657084188913,
      "grad_norm": 0.3761345446109772,
      "learning_rate": 8e-05,
      "loss": 1.9233,
      "step": 228
    },
    {
      "epoch": 0.024748730141575706,
      "grad_norm": 0.3359405994415283,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 229
    },
    {
      "epoch": 0.0248568031989625,
      "grad_norm": 0.35076820850372314,
      "learning_rate": 8e-05,
      "loss": 1.7561,
      "step": 230
    },
    {
      "epoch": 0.024964876256349293,
      "grad_norm": 0.3408905267715454,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 231
    },
    {
      "epoch": 0.025072949313736086,
      "grad_norm": 0.3573218584060669,
      "learning_rate": 8e-05,
      "loss": 1.8407,
      "step": 232
    },
    {
      "epoch": 0.02518102237112288,
      "grad_norm": 0.31585562229156494,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 233
    },
    {
      "epoch": 0.025289095428509673,
      "grad_norm": 0.3589065372943878,
      "learning_rate": 8e-05,
      "loss": 1.9227,
      "step": 234
    },
    {
      "epoch": 0.025397168485896467,
      "grad_norm": 0.3374887704849243,
      "learning_rate": 8e-05,
      "loss": 1.7683,
      "step": 235
    },
    {
      "epoch": 0.02550524154328326,
      "grad_norm": 0.35544005036354065,
      "learning_rate": 8e-05,
      "loss": 1.8151,
      "step": 236
    },
    {
      "epoch": 0.025613314600670054,
      "grad_norm": 0.34715214371681213,
      "learning_rate": 8e-05,
      "loss": 1.6359,
      "step": 237
    },
    {
      "epoch": 0.025721387658056847,
      "grad_norm": 0.319793164730072,
      "learning_rate": 8e-05,
      "loss": 1.524,
      "step": 238
    },
    {
      "epoch": 0.02582946071544364,
      "grad_norm": 0.37795740365982056,
      "learning_rate": 8e-05,
      "loss": 2.0085,
      "step": 239
    },
    {
      "epoch": 0.025937533772830434,
      "grad_norm": 0.35202550888061523,
      "learning_rate": 8e-05,
      "loss": 1.8803,
      "step": 240
    },
    {
      "epoch": 0.026045606830217227,
      "grad_norm": 0.33381348848342896,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 241
    },
    {
      "epoch": 0.02615367988760402,
      "grad_norm": 0.36450648307800293,
      "learning_rate": 8e-05,
      "loss": 1.9532,
      "step": 242
    },
    {
      "epoch": 0.026261752944990814,
      "grad_norm": 0.3383338451385498,
      "learning_rate": 8e-05,
      "loss": 1.8238,
      "step": 243
    },
    {
      "epoch": 0.026369826002377608,
      "grad_norm": 0.3532298505306244,
      "learning_rate": 8e-05,
      "loss": 1.8886,
      "step": 244
    },
    {
      "epoch": 0.0264778990597644,
      "grad_norm": 0.3489563763141632,
      "learning_rate": 8e-05,
      "loss": 1.8382,
      "step": 245
    },
    {
      "epoch": 0.026585972117151194,
      "grad_norm": 0.3260352611541748,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 246
    },
    {
      "epoch": 0.026694045174537988,
      "grad_norm": 0.35486161708831787,
      "learning_rate": 8e-05,
      "loss": 1.8844,
      "step": 247
    },
    {
      "epoch": 0.02680211823192478,
      "grad_norm": 0.33174437284469604,
      "learning_rate": 8e-05,
      "loss": 1.855,
      "step": 248
    },
    {
      "epoch": 0.026910191289311575,
      "grad_norm": 0.4141625761985779,
      "learning_rate": 8e-05,
      "loss": 1.9501,
      "step": 249
    },
    {
      "epoch": 0.027018264346698368,
      "grad_norm": 0.37890875339508057,
      "learning_rate": 8e-05,
      "loss": 2.0182,
      "step": 250
    },
    {
      "epoch": 0.02712633740408516,
      "grad_norm": 0.33093369007110596,
      "learning_rate": 8e-05,
      "loss": 1.8655,
      "step": 251
    },
    {
      "epoch": 0.027234410461471955,
      "grad_norm": 0.34627413749694824,
      "learning_rate": 8e-05,
      "loss": 1.834,
      "step": 252
    },
    {
      "epoch": 0.02734248351885875,
      "grad_norm": 0.3526088297367096,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 253
    },
    {
      "epoch": 0.027450556576245542,
      "grad_norm": 0.3771674931049347,
      "learning_rate": 8e-05,
      "loss": 1.9183,
      "step": 254
    },
    {
      "epoch": 0.027558629633632335,
      "grad_norm": 0.3523033559322357,
      "learning_rate": 8e-05,
      "loss": 1.8241,
      "step": 255
    },
    {
      "epoch": 0.02766670269101913,
      "grad_norm": 0.352531760931015,
      "learning_rate": 8e-05,
      "loss": 1.7656,
      "step": 256
    },
    {
      "epoch": 0.027774775748405922,
      "grad_norm": 0.40698111057281494,
      "learning_rate": 8e-05,
      "loss": 2.003,
      "step": 257
    },
    {
      "epoch": 0.027882848805792716,
      "grad_norm": 0.32587379217147827,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 258
    },
    {
      "epoch": 0.02799092186317951,
      "grad_norm": 0.37126708030700684,
      "learning_rate": 8e-05,
      "loss": 1.9308,
      "step": 259
    },
    {
      "epoch": 0.028098994920566302,
      "grad_norm": 0.36422014236450195,
      "learning_rate": 8e-05,
      "loss": 1.8255,
      "step": 260
    },
    {
      "epoch": 0.028207067977953096,
      "grad_norm": 0.3457685708999634,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 261
    },
    {
      "epoch": 0.02831514103533989,
      "grad_norm": 0.35602548718452454,
      "learning_rate": 8e-05,
      "loss": 1.8368,
      "step": 262
    },
    {
      "epoch": 0.028423214092726683,
      "grad_norm": 0.3606073558330536,
      "learning_rate": 8e-05,
      "loss": 1.8022,
      "step": 263
    },
    {
      "epoch": 0.028531287150113476,
      "grad_norm": 0.3583906590938568,
      "learning_rate": 8e-05,
      "loss": 1.8543,
      "step": 264
    },
    {
      "epoch": 0.02863936020750027,
      "grad_norm": 0.3634568750858307,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 265
    },
    {
      "epoch": 0.028747433264887063,
      "grad_norm": 0.3363524377346039,
      "learning_rate": 8e-05,
      "loss": 1.7092,
      "step": 266
    },
    {
      "epoch": 0.028855506322273856,
      "grad_norm": 0.3427327573299408,
      "learning_rate": 8e-05,
      "loss": 1.8065,
      "step": 267
    },
    {
      "epoch": 0.02896357937966065,
      "grad_norm": 0.36411482095718384,
      "learning_rate": 8e-05,
      "loss": 1.6541,
      "step": 268
    },
    {
      "epoch": 0.029071652437047443,
      "grad_norm": 0.351688027381897,
      "learning_rate": 8e-05,
      "loss": 1.7847,
      "step": 269
    },
    {
      "epoch": 0.029179725494434237,
      "grad_norm": 0.33315518498420715,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 270
    },
    {
      "epoch": 0.02928779855182103,
      "grad_norm": 0.33381903171539307,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 271
    },
    {
      "epoch": 0.029395871609207824,
      "grad_norm": 0.3562837541103363,
      "learning_rate": 8e-05,
      "loss": 1.8022,
      "step": 272
    },
    {
      "epoch": 0.029503944666594617,
      "grad_norm": 0.3630593419075012,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 273
    },
    {
      "epoch": 0.02961201772398141,
      "grad_norm": 0.35505959391593933,
      "learning_rate": 8e-05,
      "loss": 1.8505,
      "step": 274
    },
    {
      "epoch": 0.029720090781368204,
      "grad_norm": 0.3518848121166229,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 275
    },
    {
      "epoch": 0.029828163838754997,
      "grad_norm": 0.3195616900920868,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 276
    },
    {
      "epoch": 0.02993623689614179,
      "grad_norm": 0.3616880178451538,
      "learning_rate": 8e-05,
      "loss": 1.8278,
      "step": 277
    },
    {
      "epoch": 0.030044309953528584,
      "grad_norm": 0.3618241548538208,
      "learning_rate": 8e-05,
      "loss": 1.9174,
      "step": 278
    },
    {
      "epoch": 0.030152383010915378,
      "grad_norm": 0.3775617182254791,
      "learning_rate": 8e-05,
      "loss": 1.979,
      "step": 279
    },
    {
      "epoch": 0.03026045606830217,
      "grad_norm": 0.33412155508995056,
      "learning_rate": 8e-05,
      "loss": 1.8776,
      "step": 280
    },
    {
      "epoch": 0.030368529125688964,
      "grad_norm": 0.31312301754951477,
      "learning_rate": 8e-05,
      "loss": 1.5912,
      "step": 281
    },
    {
      "epoch": 0.030476602183075758,
      "grad_norm": 0.3662278354167938,
      "learning_rate": 8e-05,
      "loss": 1.8465,
      "step": 282
    },
    {
      "epoch": 0.03058467524046255,
      "grad_norm": 0.36790311336517334,
      "learning_rate": 8e-05,
      "loss": 1.9502,
      "step": 283
    },
    {
      "epoch": 0.030692748297849345,
      "grad_norm": 0.3379440903663635,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 284
    },
    {
      "epoch": 0.030800821355236138,
      "grad_norm": 0.36950597167015076,
      "learning_rate": 8e-05,
      "loss": 1.8648,
      "step": 285
    },
    {
      "epoch": 0.03090889441262293,
      "grad_norm": 0.3285030424594879,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 286
    },
    {
      "epoch": 0.031016967470009725,
      "grad_norm": 0.33293578028678894,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 287
    },
    {
      "epoch": 0.03112504052739652,
      "grad_norm": 0.3433438539505005,
      "learning_rate": 8e-05,
      "loss": 1.9394,
      "step": 288
    },
    {
      "epoch": 0.031233113584783312,
      "grad_norm": 0.3556041717529297,
      "learning_rate": 8e-05,
      "loss": 1.9432,
      "step": 289
    },
    {
      "epoch": 0.03134118664217011,
      "grad_norm": 0.34017595648765564,
      "learning_rate": 8e-05,
      "loss": 1.8713,
      "step": 290
    },
    {
      "epoch": 0.0314492596995569,
      "grad_norm": 0.3288329541683197,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 291
    },
    {
      "epoch": 0.031557332756943696,
      "grad_norm": 0.35025882720947266,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 292
    },
    {
      "epoch": 0.031665405814330486,
      "grad_norm": 0.3555113673210144,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 293
    },
    {
      "epoch": 0.03177347887171728,
      "grad_norm": 0.36543551087379456,
      "learning_rate": 8e-05,
      "loss": 1.9049,
      "step": 294
    },
    {
      "epoch": 0.03188155192910407,
      "grad_norm": 0.3291500210762024,
      "learning_rate": 8e-05,
      "loss": 1.6394,
      "step": 295
    },
    {
      "epoch": 0.03198962498649087,
      "grad_norm": 0.35103124380111694,
      "learning_rate": 8e-05,
      "loss": 1.9655,
      "step": 296
    },
    {
      "epoch": 0.03209769804387766,
      "grad_norm": 0.3578075170516968,
      "learning_rate": 8e-05,
      "loss": 1.8393,
      "step": 297
    },
    {
      "epoch": 0.032205771101264456,
      "grad_norm": 0.3581794798374176,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 298
    },
    {
      "epoch": 0.032313844158651246,
      "grad_norm": 0.3574570119380951,
      "learning_rate": 8e-05,
      "loss": 1.9077,
      "step": 299
    },
    {
      "epoch": 0.03242191721603804,
      "grad_norm": 0.3462468385696411,
      "learning_rate": 8e-05,
      "loss": 1.6018,
      "step": 300
    },
    {
      "epoch": 0.03252999027342483,
      "grad_norm": 0.37435850501060486,
      "learning_rate": 8e-05,
      "loss": 1.9119,
      "step": 301
    },
    {
      "epoch": 0.03263806333081163,
      "grad_norm": 0.3438180983066559,
      "learning_rate": 8e-05,
      "loss": 1.8564,
      "step": 302
    },
    {
      "epoch": 0.03274613638819842,
      "grad_norm": 0.35696008801460266,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 303
    },
    {
      "epoch": 0.03285420944558522,
      "grad_norm": 0.3689241409301758,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 304
    },
    {
      "epoch": 0.03296228250297201,
      "grad_norm": 0.3736790120601654,
      "learning_rate": 8e-05,
      "loss": 1.93,
      "step": 305
    },
    {
      "epoch": 0.033070355560358804,
      "grad_norm": 0.4091819226741791,
      "learning_rate": 8e-05,
      "loss": 1.8834,
      "step": 306
    },
    {
      "epoch": 0.033178428617745594,
      "grad_norm": 0.3624974191188812,
      "learning_rate": 8e-05,
      "loss": 1.8233,
      "step": 307
    },
    {
      "epoch": 0.03328650167513239,
      "grad_norm": 0.40337708592414856,
      "learning_rate": 8e-05,
      "loss": 1.91,
      "step": 308
    },
    {
      "epoch": 0.03339457473251918,
      "grad_norm": 0.3695801794528961,
      "learning_rate": 8e-05,
      "loss": 1.8775,
      "step": 309
    },
    {
      "epoch": 0.03350264778990598,
      "grad_norm": 0.34969937801361084,
      "learning_rate": 8e-05,
      "loss": 1.7489,
      "step": 310
    },
    {
      "epoch": 0.03361072084729277,
      "grad_norm": 0.3449096083641052,
      "learning_rate": 8e-05,
      "loss": 1.8207,
      "step": 311
    },
    {
      "epoch": 0.033718793904679564,
      "grad_norm": 0.32441210746765137,
      "learning_rate": 8e-05,
      "loss": 1.5538,
      "step": 312
    },
    {
      "epoch": 0.033826866962066354,
      "grad_norm": 0.39607810974121094,
      "learning_rate": 8e-05,
      "loss": 1.9337,
      "step": 313
    },
    {
      "epoch": 0.03393494001945315,
      "grad_norm": 0.3832971751689911,
      "learning_rate": 8e-05,
      "loss": 2.0217,
      "step": 314
    },
    {
      "epoch": 0.03404301307683994,
      "grad_norm": 0.36978593468666077,
      "learning_rate": 8e-05,
      "loss": 1.9112,
      "step": 315
    },
    {
      "epoch": 0.03415108613422674,
      "grad_norm": 0.3912685513496399,
      "learning_rate": 8e-05,
      "loss": 1.9535,
      "step": 316
    },
    {
      "epoch": 0.03425915919161353,
      "grad_norm": 0.36194825172424316,
      "learning_rate": 8e-05,
      "loss": 1.9126,
      "step": 317
    },
    {
      "epoch": 0.034367232249000325,
      "grad_norm": 0.3311236798763275,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 318
    },
    {
      "epoch": 0.034475305306387115,
      "grad_norm": 0.35460561513900757,
      "learning_rate": 8e-05,
      "loss": 1.8197,
      "step": 319
    },
    {
      "epoch": 0.03458337836377391,
      "grad_norm": 0.34542927145957947,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 320
    },
    {
      "epoch": 0.0346914514211607,
      "grad_norm": 0.36881187558174133,
      "learning_rate": 8e-05,
      "loss": 1.9422,
      "step": 321
    },
    {
      "epoch": 0.0347995244785475,
      "grad_norm": 0.32884302735328674,
      "learning_rate": 8e-05,
      "loss": 1.8724,
      "step": 322
    },
    {
      "epoch": 0.03490759753593429,
      "grad_norm": 0.3654920160770416,
      "learning_rate": 8e-05,
      "loss": 1.9599,
      "step": 323
    },
    {
      "epoch": 0.035015670593321085,
      "grad_norm": 0.34346309304237366,
      "learning_rate": 8e-05,
      "loss": 1.7999,
      "step": 324
    },
    {
      "epoch": 0.035123743650707875,
      "grad_norm": 0.346317857503891,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 325
    },
    {
      "epoch": 0.03523181670809467,
      "grad_norm": 0.3559117317199707,
      "learning_rate": 8e-05,
      "loss": 1.9348,
      "step": 326
    },
    {
      "epoch": 0.03533988976548146,
      "grad_norm": 0.3103870451450348,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 327
    },
    {
      "epoch": 0.03544796282286826,
      "grad_norm": 0.3160318434238434,
      "learning_rate": 8e-05,
      "loss": 1.667,
      "step": 328
    },
    {
      "epoch": 0.03555603588025505,
      "grad_norm": 0.34836089611053467,
      "learning_rate": 8e-05,
      "loss": 1.8578,
      "step": 329
    },
    {
      "epoch": 0.035664108937641846,
      "grad_norm": 0.34697508811950684,
      "learning_rate": 8e-05,
      "loss": 1.8063,
      "step": 330
    },
    {
      "epoch": 0.035772181995028636,
      "grad_norm": 0.34704625606536865,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 331
    },
    {
      "epoch": 0.03588025505241543,
      "grad_norm": 0.3697121739387512,
      "learning_rate": 8e-05,
      "loss": 1.9914,
      "step": 332
    },
    {
      "epoch": 0.03598832810980222,
      "grad_norm": 0.3467378616333008,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 333
    },
    {
      "epoch": 0.03609640116718902,
      "grad_norm": 0.37638208270072937,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 334
    },
    {
      "epoch": 0.03620447422457582,
      "grad_norm": 0.3592894375324249,
      "learning_rate": 8e-05,
      "loss": 1.8893,
      "step": 335
    },
    {
      "epoch": 0.03631254728196261,
      "grad_norm": 0.3446558713912964,
      "learning_rate": 8e-05,
      "loss": 1.8213,
      "step": 336
    },
    {
      "epoch": 0.036420620339349404,
      "grad_norm": 0.3634584844112396,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 337
    },
    {
      "epoch": 0.036528693396736193,
      "grad_norm": 0.39033040404319763,
      "learning_rate": 8e-05,
      "loss": 1.7989,
      "step": 338
    },
    {
      "epoch": 0.03663676645412299,
      "grad_norm": 0.4093152582645416,
      "learning_rate": 8e-05,
      "loss": 1.9441,
      "step": 339
    },
    {
      "epoch": 0.03674483951150978,
      "grad_norm": 0.39712831377983093,
      "learning_rate": 8e-05,
      "loss": 1.9132,
      "step": 340
    },
    {
      "epoch": 0.03685291256889658,
      "grad_norm": 0.37408292293548584,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 341
    },
    {
      "epoch": 0.03696098562628337,
      "grad_norm": 0.35121020674705505,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 342
    },
    {
      "epoch": 0.037069058683670164,
      "grad_norm": 0.3425823152065277,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 343
    },
    {
      "epoch": 0.037177131741056954,
      "grad_norm": 0.3547641634941101,
      "learning_rate": 8e-05,
      "loss": 1.8282,
      "step": 344
    },
    {
      "epoch": 0.03728520479844375,
      "grad_norm": 0.3267326354980469,
      "learning_rate": 8e-05,
      "loss": 1.6574,
      "step": 345
    },
    {
      "epoch": 0.03739327785583054,
      "grad_norm": 0.37417852878570557,
      "learning_rate": 8e-05,
      "loss": 1.9083,
      "step": 346
    },
    {
      "epoch": 0.03750135091321734,
      "grad_norm": 0.38231778144836426,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 347
    },
    {
      "epoch": 0.03760942397060413,
      "grad_norm": 0.34262776374816895,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 348
    },
    {
      "epoch": 0.037717497027990925,
      "grad_norm": 0.3890533149242401,
      "learning_rate": 8e-05,
      "loss": 1.8726,
      "step": 349
    },
    {
      "epoch": 0.037825570085377715,
      "grad_norm": 0.34022581577301025,
      "learning_rate": 8e-05,
      "loss": 1.8635,
      "step": 350
    },
    {
      "epoch": 0.03793364314276451,
      "grad_norm": 0.33824262022972107,
      "learning_rate": 8e-05,
      "loss": 1.7204,
      "step": 351
    },
    {
      "epoch": 0.0380417162001513,
      "grad_norm": 0.3271998167037964,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 352
    },
    {
      "epoch": 0.0381497892575381,
      "grad_norm": 0.3670596480369568,
      "learning_rate": 8e-05,
      "loss": 1.8983,
      "step": 353
    },
    {
      "epoch": 0.03825786231492489,
      "grad_norm": 0.3383699357509613,
      "learning_rate": 8e-05,
      "loss": 1.6058,
      "step": 354
    },
    {
      "epoch": 0.038365935372311685,
      "grad_norm": 0.34746605157852173,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 355
    },
    {
      "epoch": 0.038474008429698475,
      "grad_norm": 0.3594323694705963,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 356
    },
    {
      "epoch": 0.03858208148708527,
      "grad_norm": 0.3453253209590912,
      "learning_rate": 8e-05,
      "loss": 1.8821,
      "step": 357
    },
    {
      "epoch": 0.03869015454447206,
      "grad_norm": 0.32800087332725525,
      "learning_rate": 8e-05,
      "loss": 1.7831,
      "step": 358
    },
    {
      "epoch": 0.03879822760185886,
      "grad_norm": 0.35044166445732117,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 359
    },
    {
      "epoch": 0.03890630065924565,
      "grad_norm": 0.3695552349090576,
      "learning_rate": 8e-05,
      "loss": 1.5305,
      "step": 360
    },
    {
      "epoch": 0.039014373716632446,
      "grad_norm": 0.3436754047870636,
      "learning_rate": 8e-05,
      "loss": 1.8017,
      "step": 361
    },
    {
      "epoch": 0.039122446774019236,
      "grad_norm": 0.41472095251083374,
      "learning_rate": 8e-05,
      "loss": 1.3636,
      "step": 362
    },
    {
      "epoch": 0.03923051983140603,
      "grad_norm": 0.3261834681034088,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 363
    },
    {
      "epoch": 0.03933859288879282,
      "grad_norm": 0.3855779767036438,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 364
    },
    {
      "epoch": 0.03944666594617962,
      "grad_norm": 0.361030638217926,
      "learning_rate": 8e-05,
      "loss": 1.8141,
      "step": 365
    },
    {
      "epoch": 0.03955473900356641,
      "grad_norm": 0.3460286855697632,
      "learning_rate": 8e-05,
      "loss": 1.6625,
      "step": 366
    },
    {
      "epoch": 0.039662812060953206,
      "grad_norm": 0.3418448865413666,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 367
    },
    {
      "epoch": 0.039770885118339996,
      "grad_norm": 0.3445992171764374,
      "learning_rate": 8e-05,
      "loss": 1.7121,
      "step": 368
    },
    {
      "epoch": 0.03987895817572679,
      "grad_norm": 0.3819223940372467,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 369
    },
    {
      "epoch": 0.03998703123311358,
      "grad_norm": 0.3545958995819092,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 370
    },
    {
      "epoch": 0.04009510429050038,
      "grad_norm": 0.39416635036468506,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 371
    },
    {
      "epoch": 0.04020317734788717,
      "grad_norm": 0.36554983258247375,
      "learning_rate": 8e-05,
      "loss": 1.9476,
      "step": 372
    },
    {
      "epoch": 0.04031125040527397,
      "grad_norm": 0.3332500159740448,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 373
    },
    {
      "epoch": 0.04041932346266076,
      "grad_norm": 0.3605744540691376,
      "learning_rate": 8e-05,
      "loss": 2.0361,
      "step": 374
    },
    {
      "epoch": 0.040527396520047554,
      "grad_norm": 0.3399654030799866,
      "learning_rate": 8e-05,
      "loss": 1.9372,
      "step": 375
    },
    {
      "epoch": 0.040635469577434344,
      "grad_norm": 0.3782304525375366,
      "learning_rate": 8e-05,
      "loss": 1.8155,
      "step": 376
    },
    {
      "epoch": 0.04074354263482114,
      "grad_norm": 0.33868125081062317,
      "learning_rate": 8e-05,
      "loss": 1.5434,
      "step": 377
    },
    {
      "epoch": 0.04085161569220793,
      "grad_norm": 0.3886982202529907,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 378
    },
    {
      "epoch": 0.04095968874959473,
      "grad_norm": 0.4014958441257477,
      "learning_rate": 8e-05,
      "loss": 1.8195,
      "step": 379
    },
    {
      "epoch": 0.04106776180698152,
      "grad_norm": 0.34100309014320374,
      "learning_rate": 8e-05,
      "loss": 1.6916,
      "step": 380
    },
    {
      "epoch": 0.041175834864368314,
      "grad_norm": 0.3505455255508423,
      "learning_rate": 8e-05,
      "loss": 1.8154,
      "step": 381
    },
    {
      "epoch": 0.041283907921755104,
      "grad_norm": 0.3408295810222626,
      "learning_rate": 8e-05,
      "loss": 1.79,
      "step": 382
    },
    {
      "epoch": 0.0413919809791419,
      "grad_norm": 0.3800085484981537,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 383
    },
    {
      "epoch": 0.04150005403652869,
      "grad_norm": 0.35123538970947266,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 384
    },
    {
      "epoch": 0.04160812709391549,
      "grad_norm": 0.3258672058582306,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 385
    },
    {
      "epoch": 0.04171620015130228,
      "grad_norm": 0.3586743175983429,
      "learning_rate": 8e-05,
      "loss": 1.8037,
      "step": 386
    },
    {
      "epoch": 0.041824273208689075,
      "grad_norm": 0.3811790645122528,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 387
    },
    {
      "epoch": 0.041932346266075865,
      "grad_norm": 0.36081990599632263,
      "learning_rate": 8e-05,
      "loss": 1.8521,
      "step": 388
    },
    {
      "epoch": 0.04204041932346266,
      "grad_norm": 0.35482174158096313,
      "learning_rate": 8e-05,
      "loss": 1.6141,
      "step": 389
    },
    {
      "epoch": 0.04214849238084945,
      "grad_norm": 0.3619999885559082,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 390
    },
    {
      "epoch": 0.04225656543823625,
      "grad_norm": 0.371709942817688,
      "learning_rate": 8e-05,
      "loss": 1.9687,
      "step": 391
    },
    {
      "epoch": 0.04236463849562304,
      "grad_norm": 0.32231834530830383,
      "learning_rate": 8e-05,
      "loss": 1.5901,
      "step": 392
    },
    {
      "epoch": 0.042472711553009836,
      "grad_norm": 0.33641085028648376,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 393
    },
    {
      "epoch": 0.042580784610396626,
      "grad_norm": 0.35316988825798035,
      "learning_rate": 8e-05,
      "loss": 1.7697,
      "step": 394
    },
    {
      "epoch": 0.04268885766778342,
      "grad_norm": 0.3627004623413086,
      "learning_rate": 8e-05,
      "loss": 1.8328,
      "step": 395
    },
    {
      "epoch": 0.04279693072517021,
      "grad_norm": 0.3517940044403076,
      "learning_rate": 8e-05,
      "loss": 1.8157,
      "step": 396
    },
    {
      "epoch": 0.04290500378255701,
      "grad_norm": 0.34521055221557617,
      "learning_rate": 8e-05,
      "loss": 1.9056,
      "step": 397
    },
    {
      "epoch": 0.0430130768399438,
      "grad_norm": 0.3528774380683899,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 398
    },
    {
      "epoch": 0.043121149897330596,
      "grad_norm": 0.32668769359588623,
      "learning_rate": 8e-05,
      "loss": 1.7195,
      "step": 399
    },
    {
      "epoch": 0.043229222954717386,
      "grad_norm": 0.36554402112960815,
      "learning_rate": 8e-05,
      "loss": 1.8537,
      "step": 400
    },
    {
      "epoch": 0.04333729601210418,
      "grad_norm": 0.34430208802223206,
      "learning_rate": 8e-05,
      "loss": 1.8433,
      "step": 401
    },
    {
      "epoch": 0.04344536906949097,
      "grad_norm": 0.35406693816185,
      "learning_rate": 8e-05,
      "loss": 1.8906,
      "step": 402
    },
    {
      "epoch": 0.04355344212687777,
      "grad_norm": 0.3372669816017151,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 403
    },
    {
      "epoch": 0.04366151518426456,
      "grad_norm": 0.3269980251789093,
      "learning_rate": 8e-05,
      "loss": 1.7987,
      "step": 404
    },
    {
      "epoch": 0.04376958824165136,
      "grad_norm": 0.35388675332069397,
      "learning_rate": 8e-05,
      "loss": 1.8383,
      "step": 405
    },
    {
      "epoch": 0.04387766129903815,
      "grad_norm": 0.3406014144420624,
      "learning_rate": 8e-05,
      "loss": 1.8139,
      "step": 406
    },
    {
      "epoch": 0.043985734356424944,
      "grad_norm": 0.3585246801376343,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 407
    },
    {
      "epoch": 0.044093807413811734,
      "grad_norm": 0.38654810190200806,
      "learning_rate": 8e-05,
      "loss": 1.8619,
      "step": 408
    },
    {
      "epoch": 0.04420188047119853,
      "grad_norm": 0.38541606068611145,
      "learning_rate": 8e-05,
      "loss": 1.8839,
      "step": 409
    },
    {
      "epoch": 0.04430995352858532,
      "grad_norm": 0.34015166759490967,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 410
    },
    {
      "epoch": 0.04441802658597212,
      "grad_norm": 0.37512245774269104,
      "learning_rate": 8e-05,
      "loss": 1.8673,
      "step": 411
    },
    {
      "epoch": 0.04452609964335891,
      "grad_norm": 0.3188580870628357,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 412
    },
    {
      "epoch": 0.044634172700745704,
      "grad_norm": 0.3648746907711029,
      "learning_rate": 8e-05,
      "loss": 1.8527,
      "step": 413
    },
    {
      "epoch": 0.044742245758132494,
      "grad_norm": 0.45930737257003784,
      "learning_rate": 8e-05,
      "loss": 1.8655,
      "step": 414
    },
    {
      "epoch": 0.04485031881551929,
      "grad_norm": 0.3602280914783478,
      "learning_rate": 8e-05,
      "loss": 1.8813,
      "step": 415
    },
    {
      "epoch": 0.04495839187290608,
      "grad_norm": 0.34455743432044983,
      "learning_rate": 8e-05,
      "loss": 1.8455,
      "step": 416
    },
    {
      "epoch": 0.04506646493029288,
      "grad_norm": 0.331175833940506,
      "learning_rate": 8e-05,
      "loss": 1.7908,
      "step": 417
    },
    {
      "epoch": 0.045174537987679675,
      "grad_norm": 0.33429694175720215,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 418
    },
    {
      "epoch": 0.045282611045066465,
      "grad_norm": 0.374340683221817,
      "learning_rate": 8e-05,
      "loss": 1.923,
      "step": 419
    },
    {
      "epoch": 0.04539068410245326,
      "grad_norm": 0.37301379442214966,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 420
    },
    {
      "epoch": 0.04549875715984005,
      "grad_norm": 0.3516581952571869,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 421
    },
    {
      "epoch": 0.04560683021722685,
      "grad_norm": 0.34179389476776123,
      "learning_rate": 8e-05,
      "loss": 1.799,
      "step": 422
    },
    {
      "epoch": 0.04571490327461364,
      "grad_norm": 0.34728938341140747,
      "learning_rate": 8e-05,
      "loss": 1.8193,
      "step": 423
    },
    {
      "epoch": 0.045822976332000435,
      "grad_norm": 0.3493211567401886,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 424
    },
    {
      "epoch": 0.045931049389387225,
      "grad_norm": 0.34878382086753845,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 425
    },
    {
      "epoch": 0.04603912244677402,
      "grad_norm": 0.3423656225204468,
      "learning_rate": 8e-05,
      "loss": 1.8972,
      "step": 426
    },
    {
      "epoch": 0.04614719550416081,
      "grad_norm": 0.36060187220573425,
      "learning_rate": 8e-05,
      "loss": 1.7659,
      "step": 427
    },
    {
      "epoch": 0.04625526856154761,
      "grad_norm": 0.36862578988075256,
      "learning_rate": 8e-05,
      "loss": 1.848,
      "step": 428
    },
    {
      "epoch": 0.0463633416189344,
      "grad_norm": 0.3442145586013794,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 429
    },
    {
      "epoch": 0.046471414676321196,
      "grad_norm": 0.3400770425796509,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 430
    },
    {
      "epoch": 0.046579487733707986,
      "grad_norm": 0.3153742849826813,
      "learning_rate": 8e-05,
      "loss": 1.8187,
      "step": 431
    },
    {
      "epoch": 0.04668756079109478,
      "grad_norm": 0.36666610836982727,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 432
    },
    {
      "epoch": 0.04679563384848157,
      "grad_norm": 0.35173097252845764,
      "learning_rate": 8e-05,
      "loss": 1.794,
      "step": 433
    },
    {
      "epoch": 0.04690370690586837,
      "grad_norm": 0.3420298397541046,
      "learning_rate": 8e-05,
      "loss": 1.8638,
      "step": 434
    },
    {
      "epoch": 0.04701177996325516,
      "grad_norm": 0.34960559010505676,
      "learning_rate": 8e-05,
      "loss": 1.8717,
      "step": 435
    },
    {
      "epoch": 0.04711985302064196,
      "grad_norm": 0.3430729806423187,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 436
    },
    {
      "epoch": 0.04722792607802875,
      "grad_norm": 0.3420880436897278,
      "learning_rate": 8e-05,
      "loss": 1.7324,
      "step": 437
    },
    {
      "epoch": 0.047335999135415543,
      "grad_norm": 0.38300836086273193,
      "learning_rate": 8e-05,
      "loss": 1.9481,
      "step": 438
    },
    {
      "epoch": 0.04744407219280233,
      "grad_norm": 0.3362177014350891,
      "learning_rate": 8e-05,
      "loss": 1.8758,
      "step": 439
    },
    {
      "epoch": 0.04755214525018913,
      "grad_norm": 0.3980714678764343,
      "learning_rate": 8e-05,
      "loss": 1.9653,
      "step": 440
    },
    {
      "epoch": 0.04766021830757592,
      "grad_norm": 0.38389119505882263,
      "learning_rate": 8e-05,
      "loss": 1.912,
      "step": 441
    },
    {
      "epoch": 0.04776829136496272,
      "grad_norm": 0.36654260754585266,
      "learning_rate": 8e-05,
      "loss": 1.8661,
      "step": 442
    },
    {
      "epoch": 0.04787636442234951,
      "grad_norm": 0.3643897473812103,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 443
    },
    {
      "epoch": 0.047984437479736304,
      "grad_norm": 0.33685579895973206,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 444
    },
    {
      "epoch": 0.048092510537123094,
      "grad_norm": 0.3855327069759369,
      "learning_rate": 8e-05,
      "loss": 1.8651,
      "step": 445
    },
    {
      "epoch": 0.04820058359450989,
      "grad_norm": 0.37553495168685913,
      "learning_rate": 8e-05,
      "loss": 1.8213,
      "step": 446
    },
    {
      "epoch": 0.04830865665189668,
      "grad_norm": 0.3785769045352936,
      "learning_rate": 8e-05,
      "loss": 1.5616,
      "step": 447
    },
    {
      "epoch": 0.04841672970928348,
      "grad_norm": 0.36590754985809326,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 448
    },
    {
      "epoch": 0.04852480276667027,
      "grad_norm": 0.3443998396396637,
      "learning_rate": 8e-05,
      "loss": 1.5348,
      "step": 449
    },
    {
      "epoch": 0.048632875824057065,
      "grad_norm": 0.37892866134643555,
      "learning_rate": 8e-05,
      "loss": 1.959,
      "step": 450
    },
    {
      "epoch": 0.048740948881443855,
      "grad_norm": 0.35763201117515564,
      "learning_rate": 8e-05,
      "loss": 1.836,
      "step": 451
    },
    {
      "epoch": 0.04884902193883065,
      "grad_norm": 0.33352231979370117,
      "learning_rate": 8e-05,
      "loss": 1.7613,
      "step": 452
    },
    {
      "epoch": 0.04895709499621744,
      "grad_norm": 0.3498244285583496,
      "learning_rate": 8e-05,
      "loss": 1.7944,
      "step": 453
    },
    {
      "epoch": 0.04906516805360424,
      "grad_norm": 0.37402021884918213,
      "learning_rate": 8e-05,
      "loss": 1.935,
      "step": 454
    },
    {
      "epoch": 0.04917324111099103,
      "grad_norm": 0.3497718274593353,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 455
    },
    {
      "epoch": 0.049281314168377825,
      "grad_norm": 0.4347609281539917,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 456
    },
    {
      "epoch": 0.049389387225764615,
      "grad_norm": 0.374033123254776,
      "learning_rate": 8e-05,
      "loss": 2.0395,
      "step": 457
    },
    {
      "epoch": 0.04949746028315141,
      "grad_norm": 0.3523750305175781,
      "learning_rate": 8e-05,
      "loss": 1.9116,
      "step": 458
    },
    {
      "epoch": 0.0496055333405382,
      "grad_norm": 0.3229548931121826,
      "learning_rate": 8e-05,
      "loss": 1.6511,
      "step": 459
    },
    {
      "epoch": 0.049713606397925,
      "grad_norm": 0.3338475227355957,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 460
    },
    {
      "epoch": 0.04982167945531179,
      "grad_norm": 0.35802769660949707,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 461
    },
    {
      "epoch": 0.049929752512698586,
      "grad_norm": 0.3635770380496979,
      "learning_rate": 8e-05,
      "loss": 1.8359,
      "step": 462
    },
    {
      "epoch": 0.050037825570085376,
      "grad_norm": 0.3167588412761688,
      "learning_rate": 8e-05,
      "loss": 1.5331,
      "step": 463
    },
    {
      "epoch": 0.05014589862747217,
      "grad_norm": 0.3561248481273651,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 464
    },
    {
      "epoch": 0.05025397168485896,
      "grad_norm": 0.33736878633499146,
      "learning_rate": 8e-05,
      "loss": 1.7406,
      "step": 465
    },
    {
      "epoch": 0.05036204474224576,
      "grad_norm": 0.3916040062904358,
      "learning_rate": 8e-05,
      "loss": 2.002,
      "step": 466
    },
    {
      "epoch": 0.05047011779963255,
      "grad_norm": 0.33673620223999023,
      "learning_rate": 8e-05,
      "loss": 1.6607,
      "step": 467
    },
    {
      "epoch": 0.050578190857019346,
      "grad_norm": 0.35699862241744995,
      "learning_rate": 8e-05,
      "loss": 1.8892,
      "step": 468
    },
    {
      "epoch": 0.050686263914406136,
      "grad_norm": 0.34727856516838074,
      "learning_rate": 8e-05,
      "loss": 1.7731,
      "step": 469
    },
    {
      "epoch": 0.05079433697179293,
      "grad_norm": 0.32350873947143555,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 470
    },
    {
      "epoch": 0.05090241002917972,
      "grad_norm": 0.37469542026519775,
      "learning_rate": 8e-05,
      "loss": 1.965,
      "step": 471
    },
    {
      "epoch": 0.05101048308656652,
      "grad_norm": 0.37748417258262634,
      "learning_rate": 8e-05,
      "loss": 1.7888,
      "step": 472
    },
    {
      "epoch": 0.05111855614395331,
      "grad_norm": 0.35077032446861267,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 473
    },
    {
      "epoch": 0.05122662920134011,
      "grad_norm": 0.36138689517974854,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 474
    },
    {
      "epoch": 0.0513347022587269,
      "grad_norm": 0.37589678168296814,
      "learning_rate": 8e-05,
      "loss": 2.1543,
      "step": 475
    },
    {
      "epoch": 0.051442775316113694,
      "grad_norm": 0.37126532196998596,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 476
    },
    {
      "epoch": 0.051550848373500484,
      "grad_norm": 0.3406098484992981,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 477
    },
    {
      "epoch": 0.05165892143088728,
      "grad_norm": 0.353903591632843,
      "learning_rate": 8e-05,
      "loss": 1.9113,
      "step": 478
    },
    {
      "epoch": 0.05176699448827407,
      "grad_norm": 0.342562198638916,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 479
    },
    {
      "epoch": 0.05187506754566087,
      "grad_norm": 0.34356650710105896,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 480
    },
    {
      "epoch": 0.05198314060304766,
      "grad_norm": 0.32555460929870605,
      "learning_rate": 8e-05,
      "loss": 1.6289,
      "step": 481
    },
    {
      "epoch": 0.052091213660434454,
      "grad_norm": 0.37806424498558044,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 482
    },
    {
      "epoch": 0.052199286717821244,
      "grad_norm": 0.40518635511398315,
      "learning_rate": 8e-05,
      "loss": 1.8266,
      "step": 483
    },
    {
      "epoch": 0.05230735977520804,
      "grad_norm": 0.35818538069725037,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 484
    },
    {
      "epoch": 0.05241543283259483,
      "grad_norm": 0.3218657076358795,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 485
    },
    {
      "epoch": 0.05252350588998163,
      "grad_norm": 0.40635812282562256,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 486
    },
    {
      "epoch": 0.05263157894736842,
      "grad_norm": 0.39171960949897766,
      "learning_rate": 8e-05,
      "loss": 1.5384,
      "step": 487
    },
    {
      "epoch": 0.052739652004755215,
      "grad_norm": 0.3521215319633484,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 488
    },
    {
      "epoch": 0.052847725062142005,
      "grad_norm": 0.3266386389732361,
      "learning_rate": 8e-05,
      "loss": 1.6441,
      "step": 489
    },
    {
      "epoch": 0.0529557981195288,
      "grad_norm": 0.3504844605922699,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 490
    },
    {
      "epoch": 0.05306387117691559,
      "grad_norm": 0.389176607131958,
      "learning_rate": 8e-05,
      "loss": 1.9254,
      "step": 491
    },
    {
      "epoch": 0.05317194423430239,
      "grad_norm": 0.36079034209251404,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 492
    },
    {
      "epoch": 0.05328001729168918,
      "grad_norm": 0.37580999732017517,
      "learning_rate": 8e-05,
      "loss": 1.9303,
      "step": 493
    },
    {
      "epoch": 0.053388090349075976,
      "grad_norm": 0.37816786766052246,
      "learning_rate": 8e-05,
      "loss": 1.8253,
      "step": 494
    },
    {
      "epoch": 0.053496163406462766,
      "grad_norm": 0.3468021750450134,
      "learning_rate": 8e-05,
      "loss": 1.7591,
      "step": 495
    },
    {
      "epoch": 0.05360423646384956,
      "grad_norm": 0.37619641423225403,
      "learning_rate": 8e-05,
      "loss": 1.8257,
      "step": 496
    },
    {
      "epoch": 0.05371230952123635,
      "grad_norm": 0.34040290117263794,
      "learning_rate": 8e-05,
      "loss": 1.9514,
      "step": 497
    },
    {
      "epoch": 0.05382038257862315,
      "grad_norm": 0.3603047728538513,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 498
    },
    {
      "epoch": 0.05392845563600994,
      "grad_norm": 0.3678261637687683,
      "learning_rate": 8e-05,
      "loss": 1.8713,
      "step": 499
    },
    {
      "epoch": 0.054036528693396736,
      "grad_norm": 0.32692885398864746,
      "learning_rate": 8e-05,
      "loss": 1.7825,
      "step": 500
    },
    {
      "epoch": 0.05414460175078353,
      "grad_norm": 0.33240586519241333,
      "learning_rate": 8e-05,
      "loss": 1.7369,
      "step": 501
    },
    {
      "epoch": 0.05425267480817032,
      "grad_norm": 0.3406820297241211,
      "learning_rate": 8e-05,
      "loss": 1.9478,
      "step": 502
    },
    {
      "epoch": 0.05436074786555712,
      "grad_norm": 0.3288944959640503,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 503
    },
    {
      "epoch": 0.05446882092294391,
      "grad_norm": 0.3384639620780945,
      "learning_rate": 8e-05,
      "loss": 1.9113,
      "step": 504
    },
    {
      "epoch": 0.05457689398033071,
      "grad_norm": 0.341382771730423,
      "learning_rate": 8e-05,
      "loss": 1.9594,
      "step": 505
    },
    {
      "epoch": 0.0546849670377175,
      "grad_norm": 0.338291198015213,
      "learning_rate": 8e-05,
      "loss": 1.78,
      "step": 506
    },
    {
      "epoch": 0.054793040095104294,
      "grad_norm": 0.3306415379047394,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 507
    },
    {
      "epoch": 0.054901113152491084,
      "grad_norm": 0.33666789531707764,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 508
    },
    {
      "epoch": 0.05500918620987788,
      "grad_norm": 0.3264491856098175,
      "learning_rate": 8e-05,
      "loss": 1.7859,
      "step": 509
    },
    {
      "epoch": 0.05511725926726467,
      "grad_norm": 0.3928073048591614,
      "learning_rate": 8e-05,
      "loss": 1.8382,
      "step": 510
    },
    {
      "epoch": 0.05522533232465147,
      "grad_norm": 0.3506225347518921,
      "learning_rate": 8e-05,
      "loss": 1.9297,
      "step": 511
    },
    {
      "epoch": 0.05533340538203826,
      "grad_norm": 0.3470093607902527,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 512
    },
    {
      "epoch": 0.055441478439425054,
      "grad_norm": 0.3335685729980469,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 513
    },
    {
      "epoch": 0.055549551496811844,
      "grad_norm": 0.3314863443374634,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 514
    },
    {
      "epoch": 0.05565762455419864,
      "grad_norm": 0.33801519870758057,
      "learning_rate": 8e-05,
      "loss": 1.7993,
      "step": 515
    },
    {
      "epoch": 0.05576569761158543,
      "grad_norm": 0.3557450771331787,
      "learning_rate": 8e-05,
      "loss": 1.6081,
      "step": 516
    },
    {
      "epoch": 0.05587377066897223,
      "grad_norm": 0.4297201335430145,
      "learning_rate": 8e-05,
      "loss": 2.0665,
      "step": 517
    },
    {
      "epoch": 0.05598184372635902,
      "grad_norm": 0.35107314586639404,
      "learning_rate": 8e-05,
      "loss": 1.9051,
      "step": 518
    },
    {
      "epoch": 0.056089916783745815,
      "grad_norm": 0.3304230868816376,
      "learning_rate": 8e-05,
      "loss": 1.7827,
      "step": 519
    },
    {
      "epoch": 0.056197989841132605,
      "grad_norm": 0.35226425528526306,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 520
    },
    {
      "epoch": 0.0563060628985194,
      "grad_norm": 0.32998597621917725,
      "learning_rate": 8e-05,
      "loss": 1.7804,
      "step": 521
    },
    {
      "epoch": 0.05641413595590619,
      "grad_norm": 0.3704870939254761,
      "learning_rate": 8e-05,
      "loss": 1.8446,
      "step": 522
    },
    {
      "epoch": 0.05652220901329299,
      "grad_norm": 0.3906189501285553,
      "learning_rate": 8e-05,
      "loss": 1.8716,
      "step": 523
    },
    {
      "epoch": 0.05663028207067978,
      "grad_norm": 0.3609927296638489,
      "learning_rate": 8e-05,
      "loss": 1.8022,
      "step": 524
    },
    {
      "epoch": 0.056738355128066575,
      "grad_norm": 0.3408806025981903,
      "learning_rate": 8e-05,
      "loss": 1.7892,
      "step": 525
    },
    {
      "epoch": 0.056846428185453365,
      "grad_norm": 0.3482629656791687,
      "learning_rate": 8e-05,
      "loss": 1.7619,
      "step": 526
    },
    {
      "epoch": 0.05695450124284016,
      "grad_norm": 0.34368976950645447,
      "learning_rate": 8e-05,
      "loss": 1.7962,
      "step": 527
    },
    {
      "epoch": 0.05706257430022695,
      "grad_norm": 0.35801342129707336,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 528
    },
    {
      "epoch": 0.05717064735761375,
      "grad_norm": 0.3331338167190552,
      "learning_rate": 8e-05,
      "loss": 1.7631,
      "step": 529
    },
    {
      "epoch": 0.05727872041500054,
      "grad_norm": 0.3434489965438843,
      "learning_rate": 8e-05,
      "loss": 1.7431,
      "step": 530
    },
    {
      "epoch": 0.057386793472387336,
      "grad_norm": 0.3381546139717102,
      "learning_rate": 8e-05,
      "loss": 1.7641,
      "step": 531
    },
    {
      "epoch": 0.057494866529774126,
      "grad_norm": 0.33945730328559875,
      "learning_rate": 8e-05,
      "loss": 1.9493,
      "step": 532
    },
    {
      "epoch": 0.05760293958716092,
      "grad_norm": 0.3365679383277893,
      "learning_rate": 8e-05,
      "loss": 1.8159,
      "step": 533
    },
    {
      "epoch": 0.05771101264454771,
      "grad_norm": 0.3401789963245392,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 534
    },
    {
      "epoch": 0.05781908570193451,
      "grad_norm": 0.33047136664390564,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 535
    },
    {
      "epoch": 0.0579271587593213,
      "grad_norm": 0.35353079438209534,
      "learning_rate": 8e-05,
      "loss": 1.9169,
      "step": 536
    },
    {
      "epoch": 0.0580352318167081,
      "grad_norm": 0.39713647961616516,
      "learning_rate": 8e-05,
      "loss": 1.9314,
      "step": 537
    },
    {
      "epoch": 0.05814330487409489,
      "grad_norm": 0.3830745816230774,
      "learning_rate": 8e-05,
      "loss": 1.9128,
      "step": 538
    },
    {
      "epoch": 0.058251377931481683,
      "grad_norm": 0.36057743430137634,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 539
    },
    {
      "epoch": 0.05835945098886847,
      "grad_norm": 0.40052130818367004,
      "learning_rate": 8e-05,
      "loss": 1.9473,
      "step": 540
    },
    {
      "epoch": 0.05846752404625527,
      "grad_norm": 0.340400367975235,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 541
    },
    {
      "epoch": 0.05857559710364206,
      "grad_norm": 0.334179550409317,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 542
    },
    {
      "epoch": 0.05868367016102886,
      "grad_norm": 0.3954176902770996,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 543
    },
    {
      "epoch": 0.05879174321841565,
      "grad_norm": 0.40544360876083374,
      "learning_rate": 8e-05,
      "loss": 1.9036,
      "step": 544
    },
    {
      "epoch": 0.058899816275802444,
      "grad_norm": 0.3730384409427643,
      "learning_rate": 8e-05,
      "loss": 1.9688,
      "step": 545
    },
    {
      "epoch": 0.059007889333189234,
      "grad_norm": 0.38895460963249207,
      "learning_rate": 8e-05,
      "loss": 1.8152,
      "step": 546
    },
    {
      "epoch": 0.05911596239057603,
      "grad_norm": 0.3750901520252228,
      "learning_rate": 8e-05,
      "loss": 1.92,
      "step": 547
    },
    {
      "epoch": 0.05922403544796282,
      "grad_norm": 0.3438848555088043,
      "learning_rate": 8e-05,
      "loss": 1.8058,
      "step": 548
    },
    {
      "epoch": 0.05933210850534962,
      "grad_norm": 0.34970834851264954,
      "learning_rate": 8e-05,
      "loss": 1.7938,
      "step": 549
    },
    {
      "epoch": 0.05944018156273641,
      "grad_norm": 0.37738972902297974,
      "learning_rate": 8e-05,
      "loss": 2.0443,
      "step": 550
    },
    {
      "epoch": 0.059548254620123205,
      "grad_norm": 0.3614276647567749,
      "learning_rate": 8e-05,
      "loss": 1.8913,
      "step": 551
    },
    {
      "epoch": 0.059656327677509995,
      "grad_norm": 0.36763495206832886,
      "learning_rate": 8e-05,
      "loss": 1.8611,
      "step": 552
    },
    {
      "epoch": 0.05976440073489679,
      "grad_norm": 0.35178446769714355,
      "learning_rate": 8e-05,
      "loss": 1.8826,
      "step": 553
    },
    {
      "epoch": 0.05987247379228358,
      "grad_norm": 0.3617529273033142,
      "learning_rate": 8e-05,
      "loss": 1.8131,
      "step": 554
    },
    {
      "epoch": 0.05998054684967038,
      "grad_norm": 0.35756927728652954,
      "learning_rate": 8e-05,
      "loss": 1.7225,
      "step": 555
    },
    {
      "epoch": 0.06008861990705717,
      "grad_norm": 0.3592441976070404,
      "learning_rate": 8e-05,
      "loss": 1.7632,
      "step": 556
    },
    {
      "epoch": 0.060196692964443965,
      "grad_norm": 0.39115583896636963,
      "learning_rate": 8e-05,
      "loss": 2.0189,
      "step": 557
    },
    {
      "epoch": 0.060304766021830755,
      "grad_norm": 0.3478936553001404,
      "learning_rate": 8e-05,
      "loss": 1.911,
      "step": 558
    },
    {
      "epoch": 0.06041283907921755,
      "grad_norm": 0.382700115442276,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 559
    },
    {
      "epoch": 0.06052091213660434,
      "grad_norm": 0.33886781334877014,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 560
    },
    {
      "epoch": 0.06062898519399114,
      "grad_norm": 0.32599201798439026,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 561
    },
    {
      "epoch": 0.06073705825137793,
      "grad_norm": 0.3453141450881958,
      "learning_rate": 8e-05,
      "loss": 1.8782,
      "step": 562
    },
    {
      "epoch": 0.060845131308764726,
      "grad_norm": 0.3136008083820343,
      "learning_rate": 8e-05,
      "loss": 1.8128,
      "step": 563
    },
    {
      "epoch": 0.060953204366151516,
      "grad_norm": 0.34391477704048157,
      "learning_rate": 8e-05,
      "loss": 1.5151,
      "step": 564
    },
    {
      "epoch": 0.06106127742353831,
      "grad_norm": 0.3599853217601776,
      "learning_rate": 8e-05,
      "loss": 1.9333,
      "step": 565
    },
    {
      "epoch": 0.0611693504809251,
      "grad_norm": 0.3889324963092804,
      "learning_rate": 8e-05,
      "loss": 2.0541,
      "step": 566
    },
    {
      "epoch": 0.0612774235383119,
      "grad_norm": 0.33915358781814575,
      "learning_rate": 8e-05,
      "loss": 1.8674,
      "step": 567
    },
    {
      "epoch": 0.06138549659569869,
      "grad_norm": 0.36452826857566833,
      "learning_rate": 8e-05,
      "loss": 1.8277,
      "step": 568
    },
    {
      "epoch": 0.061493569653085486,
      "grad_norm": 0.34183764457702637,
      "learning_rate": 8e-05,
      "loss": 1.865,
      "step": 569
    },
    {
      "epoch": 0.061601642710472276,
      "grad_norm": 0.39172831177711487,
      "learning_rate": 8e-05,
      "loss": 1.9538,
      "step": 570
    },
    {
      "epoch": 0.06170971576785907,
      "grad_norm": 0.3603333532810211,
      "learning_rate": 8e-05,
      "loss": 1.9699,
      "step": 571
    },
    {
      "epoch": 0.06181778882524586,
      "grad_norm": 0.3480854332447052,
      "learning_rate": 8e-05,
      "loss": 1.8306,
      "step": 572
    },
    {
      "epoch": 0.06192586188263266,
      "grad_norm": 0.34380054473876953,
      "learning_rate": 8e-05,
      "loss": 1.783,
      "step": 573
    },
    {
      "epoch": 0.06203393494001945,
      "grad_norm": 0.33279845118522644,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 574
    },
    {
      "epoch": 0.06214200799740625,
      "grad_norm": 0.3506140410900116,
      "learning_rate": 8e-05,
      "loss": 1.8859,
      "step": 575
    },
    {
      "epoch": 0.06225008105479304,
      "grad_norm": 0.34136146306991577,
      "learning_rate": 8e-05,
      "loss": 1.7201,
      "step": 576
    },
    {
      "epoch": 0.062358154112179834,
      "grad_norm": 0.37575459480285645,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 577
    },
    {
      "epoch": 0.062466227169566624,
      "grad_norm": 0.3390471637248993,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 578
    },
    {
      "epoch": 0.06257430022695341,
      "grad_norm": 0.37138631939888,
      "learning_rate": 8e-05,
      "loss": 1.9054,
      "step": 579
    },
    {
      "epoch": 0.06268237328434022,
      "grad_norm": 0.32744237780570984,
      "learning_rate": 8e-05,
      "loss": 1.7409,
      "step": 580
    },
    {
      "epoch": 0.06279044634172701,
      "grad_norm": 0.32661354541778564,
      "learning_rate": 8e-05,
      "loss": 1.5193,
      "step": 581
    },
    {
      "epoch": 0.0628985193991138,
      "grad_norm": 0.3683415353298187,
      "learning_rate": 8e-05,
      "loss": 1.9487,
      "step": 582
    },
    {
      "epoch": 0.06300659245650059,
      "grad_norm": 0.36816683411598206,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 583
    },
    {
      "epoch": 0.06311466551388739,
      "grad_norm": 0.34831303358078003,
      "learning_rate": 8e-05,
      "loss": 1.857,
      "step": 584
    },
    {
      "epoch": 0.06322273857127418,
      "grad_norm": 0.3598869740962982,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 585
    },
    {
      "epoch": 0.06333081162866097,
      "grad_norm": 0.34880056977272034,
      "learning_rate": 8e-05,
      "loss": 1.7589,
      "step": 586
    },
    {
      "epoch": 0.06343888468604778,
      "grad_norm": 0.3511759638786316,
      "learning_rate": 8e-05,
      "loss": 1.6893,
      "step": 587
    },
    {
      "epoch": 0.06354695774343457,
      "grad_norm": 0.35798949003219604,
      "learning_rate": 8e-05,
      "loss": 1.7675,
      "step": 588
    },
    {
      "epoch": 0.06365503080082136,
      "grad_norm": 0.38478884100914,
      "learning_rate": 8e-05,
      "loss": 1.9547,
      "step": 589
    },
    {
      "epoch": 0.06376310385820814,
      "grad_norm": 0.3490166664123535,
      "learning_rate": 8e-05,
      "loss": 1.9554,
      "step": 590
    },
    {
      "epoch": 0.06387117691559495,
      "grad_norm": 0.33670371770858765,
      "learning_rate": 8e-05,
      "loss": 1.8227,
      "step": 591
    },
    {
      "epoch": 0.06397924997298174,
      "grad_norm": 0.3731778562068939,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 592
    },
    {
      "epoch": 0.06408732303036853,
      "grad_norm": 0.36261558532714844,
      "learning_rate": 8e-05,
      "loss": 1.9185,
      "step": 593
    },
    {
      "epoch": 0.06419539608775532,
      "grad_norm": 0.341047465801239,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 594
    },
    {
      "epoch": 0.06430346914514212,
      "grad_norm": 0.32539796829223633,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 595
    },
    {
      "epoch": 0.06441154220252891,
      "grad_norm": 0.347179114818573,
      "learning_rate": 8e-05,
      "loss": 1.8035,
      "step": 596
    },
    {
      "epoch": 0.0645196152599157,
      "grad_norm": 0.3422594368457794,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 597
    },
    {
      "epoch": 0.06462768831730249,
      "grad_norm": 0.3355692923069,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 598
    },
    {
      "epoch": 0.0647357613746893,
      "grad_norm": 0.36092662811279297,
      "learning_rate": 8e-05,
      "loss": 1.8366,
      "step": 599
    },
    {
      "epoch": 0.06484383443207609,
      "grad_norm": 0.37264716625213623,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 600
    },
    {
      "epoch": 0.06495190748946288,
      "grad_norm": 0.33752110600471497,
      "learning_rate": 8e-05,
      "loss": 1.5776,
      "step": 601
    },
    {
      "epoch": 0.06505998054684967,
      "grad_norm": 0.3596280813217163,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 602
    },
    {
      "epoch": 0.06516805360423647,
      "grad_norm": 0.34022724628448486,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 603
    },
    {
      "epoch": 0.06527612666162326,
      "grad_norm": 0.3194156587123871,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 604
    },
    {
      "epoch": 0.06538419971901005,
      "grad_norm": 0.3691875636577606,
      "learning_rate": 8e-05,
      "loss": 1.8642,
      "step": 605
    },
    {
      "epoch": 0.06549227277639684,
      "grad_norm": 0.34021759033203125,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 606
    },
    {
      "epoch": 0.06560034583378364,
      "grad_norm": 0.42838627099990845,
      "learning_rate": 8e-05,
      "loss": 2.0069,
      "step": 607
    },
    {
      "epoch": 0.06570841889117043,
      "grad_norm": 0.355054646730423,
      "learning_rate": 8e-05,
      "loss": 1.7867,
      "step": 608
    },
    {
      "epoch": 0.06581649194855722,
      "grad_norm": 0.3429180681705475,
      "learning_rate": 8e-05,
      "loss": 1.5826,
      "step": 609
    },
    {
      "epoch": 0.06592456500594401,
      "grad_norm": 0.3588315546512604,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 610
    },
    {
      "epoch": 0.06603263806333082,
      "grad_norm": 0.33538955450057983,
      "learning_rate": 8e-05,
      "loss": 1.6732,
      "step": 611
    },
    {
      "epoch": 0.06614071112071761,
      "grad_norm": 0.345381498336792,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 612
    },
    {
      "epoch": 0.0662487841781044,
      "grad_norm": 0.3466568887233734,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 613
    },
    {
      "epoch": 0.06635685723549119,
      "grad_norm": 0.3260331153869629,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 614
    },
    {
      "epoch": 0.06646493029287799,
      "grad_norm": 0.3282583951950073,
      "learning_rate": 8e-05,
      "loss": 1.8235,
      "step": 615
    },
    {
      "epoch": 0.06657300335026478,
      "grad_norm": 0.37326928973197937,
      "learning_rate": 8e-05,
      "loss": 1.9629,
      "step": 616
    },
    {
      "epoch": 0.06668107640765157,
      "grad_norm": 0.36559388041496277,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 617
    },
    {
      "epoch": 0.06678914946503836,
      "grad_norm": 0.33646243810653687,
      "learning_rate": 8e-05,
      "loss": 1.7959,
      "step": 618
    },
    {
      "epoch": 0.06689722252242516,
      "grad_norm": 0.34386146068573,
      "learning_rate": 8e-05,
      "loss": 1.8981,
      "step": 619
    },
    {
      "epoch": 0.06700529557981195,
      "grad_norm": 0.378597617149353,
      "learning_rate": 8e-05,
      "loss": 1.7251,
      "step": 620
    },
    {
      "epoch": 0.06711336863719874,
      "grad_norm": 0.3454154133796692,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 621
    },
    {
      "epoch": 0.06722144169458553,
      "grad_norm": 0.34818366169929504,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 622
    },
    {
      "epoch": 0.06732951475197234,
      "grad_norm": 0.36195915937423706,
      "learning_rate": 8e-05,
      "loss": 1.9585,
      "step": 623
    },
    {
      "epoch": 0.06743758780935913,
      "grad_norm": 0.37999871373176575,
      "learning_rate": 8e-05,
      "loss": 1.8231,
      "step": 624
    },
    {
      "epoch": 0.06754566086674592,
      "grad_norm": 0.35395121574401855,
      "learning_rate": 8e-05,
      "loss": 1.8561,
      "step": 625
    },
    {
      "epoch": 0.06765373392413271,
      "grad_norm": 0.36281928420066833,
      "learning_rate": 8e-05,
      "loss": 1.9698,
      "step": 626
    },
    {
      "epoch": 0.06776180698151951,
      "grad_norm": 0.39710533618927,
      "learning_rate": 8e-05,
      "loss": 1.9219,
      "step": 627
    },
    {
      "epoch": 0.0678698800389063,
      "grad_norm": 0.3393266797065735,
      "learning_rate": 8e-05,
      "loss": 1.7098,
      "step": 628
    },
    {
      "epoch": 0.06797795309629309,
      "grad_norm": 0.368127703666687,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 629
    },
    {
      "epoch": 0.06808602615367988,
      "grad_norm": 0.3395160734653473,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 630
    },
    {
      "epoch": 0.06819409921106669,
      "grad_norm": 0.3301742970943451,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 631
    },
    {
      "epoch": 0.06830217226845348,
      "grad_norm": 0.40448659658432007,
      "learning_rate": 8e-05,
      "loss": 1.9478,
      "step": 632
    },
    {
      "epoch": 0.06841024532584027,
      "grad_norm": 0.3483118712902069,
      "learning_rate": 8e-05,
      "loss": 1.8395,
      "step": 633
    },
    {
      "epoch": 0.06851831838322706,
      "grad_norm": 0.3806949853897095,
      "learning_rate": 8e-05,
      "loss": 1.9625,
      "step": 634
    },
    {
      "epoch": 0.06862639144061386,
      "grad_norm": 0.3638564944267273,
      "learning_rate": 8e-05,
      "loss": 1.9365,
      "step": 635
    },
    {
      "epoch": 0.06873446449800065,
      "grad_norm": 0.34495770931243896,
      "learning_rate": 8e-05,
      "loss": 1.8124,
      "step": 636
    },
    {
      "epoch": 0.06884253755538744,
      "grad_norm": 0.40036216378211975,
      "learning_rate": 8e-05,
      "loss": 2.0827,
      "step": 637
    },
    {
      "epoch": 0.06895061061277423,
      "grad_norm": 0.32573381066322327,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 638
    },
    {
      "epoch": 0.06905868367016103,
      "grad_norm": 0.34161123633384705,
      "learning_rate": 8e-05,
      "loss": 1.9109,
      "step": 639
    },
    {
      "epoch": 0.06916675672754782,
      "grad_norm": 0.35865241289138794,
      "learning_rate": 8e-05,
      "loss": 1.8919,
      "step": 640
    },
    {
      "epoch": 0.06927482978493461,
      "grad_norm": 0.3654935657978058,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 641
    },
    {
      "epoch": 0.0693829028423214,
      "grad_norm": 0.32270491123199463,
      "learning_rate": 8e-05,
      "loss": 1.8346,
      "step": 642
    },
    {
      "epoch": 0.06949097589970821,
      "grad_norm": 0.3278350830078125,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 643
    },
    {
      "epoch": 0.069599048957095,
      "grad_norm": 0.3394119441509247,
      "learning_rate": 8e-05,
      "loss": 1.8761,
      "step": 644
    },
    {
      "epoch": 0.06970712201448179,
      "grad_norm": 0.3506968319416046,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 645
    },
    {
      "epoch": 0.06981519507186858,
      "grad_norm": 0.4216964542865753,
      "learning_rate": 8e-05,
      "loss": 1.9973,
      "step": 646
    },
    {
      "epoch": 0.06992326812925538,
      "grad_norm": 0.32112473249435425,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 647
    },
    {
      "epoch": 0.07003134118664217,
      "grad_norm": 0.3561711311340332,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 648
    },
    {
      "epoch": 0.07013941424402896,
      "grad_norm": 0.3361760675907135,
      "learning_rate": 8e-05,
      "loss": 1.8433,
      "step": 649
    },
    {
      "epoch": 0.07024748730141575,
      "grad_norm": 0.35475993156433105,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 650
    },
    {
      "epoch": 0.07035556035880255,
      "grad_norm": 0.3732445538043976,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 651
    },
    {
      "epoch": 0.07046363341618934,
      "grad_norm": 0.36751070618629456,
      "learning_rate": 8e-05,
      "loss": 1.9178,
      "step": 652
    },
    {
      "epoch": 0.07057170647357613,
      "grad_norm": 0.3510911762714386,
      "learning_rate": 8e-05,
      "loss": 1.7829,
      "step": 653
    },
    {
      "epoch": 0.07067977953096292,
      "grad_norm": 0.3700132966041565,
      "learning_rate": 8e-05,
      "loss": 1.8732,
      "step": 654
    },
    {
      "epoch": 0.07078785258834973,
      "grad_norm": 0.3228524923324585,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 655
    },
    {
      "epoch": 0.07089592564573652,
      "grad_norm": 0.3509063720703125,
      "learning_rate": 8e-05,
      "loss": 1.8732,
      "step": 656
    },
    {
      "epoch": 0.07100399870312331,
      "grad_norm": 0.3533901274204254,
      "learning_rate": 8e-05,
      "loss": 1.7774,
      "step": 657
    },
    {
      "epoch": 0.0711120717605101,
      "grad_norm": 0.34663504362106323,
      "learning_rate": 8e-05,
      "loss": 1.876,
      "step": 658
    },
    {
      "epoch": 0.0712201448178969,
      "grad_norm": 0.3430791199207306,
      "learning_rate": 8e-05,
      "loss": 1.7164,
      "step": 659
    },
    {
      "epoch": 0.07132821787528369,
      "grad_norm": 0.33836594223976135,
      "learning_rate": 8e-05,
      "loss": 1.824,
      "step": 660
    },
    {
      "epoch": 0.07143629093267048,
      "grad_norm": 0.3621346950531006,
      "learning_rate": 8e-05,
      "loss": 1.8906,
      "step": 661
    },
    {
      "epoch": 0.07154436399005727,
      "grad_norm": 0.34158116579055786,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 662
    },
    {
      "epoch": 0.07165243704744408,
      "grad_norm": 0.37157899141311646,
      "learning_rate": 8e-05,
      "loss": 1.7956,
      "step": 663
    },
    {
      "epoch": 0.07176051010483087,
      "grad_norm": 0.355392187833786,
      "learning_rate": 8e-05,
      "loss": 1.8269,
      "step": 664
    },
    {
      "epoch": 0.07186858316221766,
      "grad_norm": 0.3327278196811676,
      "learning_rate": 8e-05,
      "loss": 1.8037,
      "step": 665
    },
    {
      "epoch": 0.07197665621960445,
      "grad_norm": 0.33161213994026184,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 666
    },
    {
      "epoch": 0.07208472927699125,
      "grad_norm": 0.3450239598751068,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 667
    },
    {
      "epoch": 0.07219280233437804,
      "grad_norm": 0.3507198691368103,
      "learning_rate": 8e-05,
      "loss": 1.6889,
      "step": 668
    },
    {
      "epoch": 0.07230087539176483,
      "grad_norm": 0.34202083945274353,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 669
    },
    {
      "epoch": 0.07240894844915163,
      "grad_norm": 0.3287566006183624,
      "learning_rate": 8e-05,
      "loss": 1.7038,
      "step": 670
    },
    {
      "epoch": 0.07251702150653842,
      "grad_norm": 0.36855319142341614,
      "learning_rate": 8e-05,
      "loss": 1.8515,
      "step": 671
    },
    {
      "epoch": 0.07262509456392521,
      "grad_norm": 0.340627521276474,
      "learning_rate": 8e-05,
      "loss": 1.8783,
      "step": 672
    },
    {
      "epoch": 0.072733167621312,
      "grad_norm": 0.3321596086025238,
      "learning_rate": 8e-05,
      "loss": 1.7327,
      "step": 673
    },
    {
      "epoch": 0.07284124067869881,
      "grad_norm": 0.3582914471626282,
      "learning_rate": 8e-05,
      "loss": 1.7415,
      "step": 674
    },
    {
      "epoch": 0.0729493137360856,
      "grad_norm": 0.3498310446739197,
      "learning_rate": 8e-05,
      "loss": 1.8232,
      "step": 675
    },
    {
      "epoch": 0.07305738679347239,
      "grad_norm": 0.31563234329223633,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 676
    },
    {
      "epoch": 0.07316545985085918,
      "grad_norm": 0.3766508400440216,
      "learning_rate": 8e-05,
      "loss": 1.9739,
      "step": 677
    },
    {
      "epoch": 0.07327353290824598,
      "grad_norm": 0.34974056482315063,
      "learning_rate": 8e-05,
      "loss": 1.7983,
      "step": 678
    },
    {
      "epoch": 0.07338160596563277,
      "grad_norm": 0.3550325930118561,
      "learning_rate": 8e-05,
      "loss": 1.8809,
      "step": 679
    },
    {
      "epoch": 0.07348967902301956,
      "grad_norm": 0.4403340518474579,
      "learning_rate": 8e-05,
      "loss": 2.1644,
      "step": 680
    },
    {
      "epoch": 0.07359775208040635,
      "grad_norm": 0.3214254379272461,
      "learning_rate": 8e-05,
      "loss": 1.6729,
      "step": 681
    },
    {
      "epoch": 0.07370582513779315,
      "grad_norm": 0.3804623484611511,
      "learning_rate": 8e-05,
      "loss": 1.9349,
      "step": 682
    },
    {
      "epoch": 0.07381389819517994,
      "grad_norm": 0.32864585518836975,
      "learning_rate": 8e-05,
      "loss": 1.7857,
      "step": 683
    },
    {
      "epoch": 0.07392197125256673,
      "grad_norm": 0.35231640934944153,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 684
    },
    {
      "epoch": 0.07403004430995352,
      "grad_norm": 0.3733230233192444,
      "learning_rate": 8e-05,
      "loss": 1.8922,
      "step": 685
    },
    {
      "epoch": 0.07413811736734033,
      "grad_norm": 0.3392426669597626,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 686
    },
    {
      "epoch": 0.07424619042472712,
      "grad_norm": 0.3542892336845398,
      "learning_rate": 8e-05,
      "loss": 1.5467,
      "step": 687
    },
    {
      "epoch": 0.07435426348211391,
      "grad_norm": 0.35224273800849915,
      "learning_rate": 8e-05,
      "loss": 1.7769,
      "step": 688
    },
    {
      "epoch": 0.0744623365395007,
      "grad_norm": 0.3452960252761841,
      "learning_rate": 8e-05,
      "loss": 1.6577,
      "step": 689
    },
    {
      "epoch": 0.0745704095968875,
      "grad_norm": 0.3708077073097229,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 690
    },
    {
      "epoch": 0.07467848265427429,
      "grad_norm": 0.3276267349720001,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 691
    },
    {
      "epoch": 0.07478655571166108,
      "grad_norm": 0.3788032829761505,
      "learning_rate": 8e-05,
      "loss": 1.8047,
      "step": 692
    },
    {
      "epoch": 0.07489462876904787,
      "grad_norm": 0.37438222765922546,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 693
    },
    {
      "epoch": 0.07500270182643468,
      "grad_norm": 0.37413498759269714,
      "learning_rate": 8e-05,
      "loss": 1.7814,
      "step": 694
    },
    {
      "epoch": 0.07511077488382147,
      "grad_norm": 0.33975186944007874,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 695
    },
    {
      "epoch": 0.07521884794120826,
      "grad_norm": 0.34187236428260803,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 696
    },
    {
      "epoch": 0.07532692099859505,
      "grad_norm": 0.3947480320930481,
      "learning_rate": 8e-05,
      "loss": 1.7987,
      "step": 697
    },
    {
      "epoch": 0.07543499405598185,
      "grad_norm": 0.44009360671043396,
      "learning_rate": 8e-05,
      "loss": 2.0018,
      "step": 698
    },
    {
      "epoch": 0.07554306711336864,
      "grad_norm": 0.33826857805252075,
      "learning_rate": 8e-05,
      "loss": 1.6599,
      "step": 699
    },
    {
      "epoch": 0.07565114017075543,
      "grad_norm": 0.3385428488254547,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 700
    },
    {
      "epoch": 0.07575921322814222,
      "grad_norm": 0.43305906653404236,
      "learning_rate": 8e-05,
      "loss": 1.9682,
      "step": 701
    },
    {
      "epoch": 0.07586728628552902,
      "grad_norm": 0.3494119644165039,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 702
    },
    {
      "epoch": 0.07597535934291581,
      "grad_norm": 0.38518205285072327,
      "learning_rate": 8e-05,
      "loss": 1.9468,
      "step": 703
    },
    {
      "epoch": 0.0760834324003026,
      "grad_norm": 0.3591043949127197,
      "learning_rate": 8e-05,
      "loss": 1.8352,
      "step": 704
    },
    {
      "epoch": 0.07619150545768939,
      "grad_norm": 0.34808701276779175,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 705
    },
    {
      "epoch": 0.0762995785150762,
      "grad_norm": 0.34616905450820923,
      "learning_rate": 8e-05,
      "loss": 1.8232,
      "step": 706
    },
    {
      "epoch": 0.07640765157246299,
      "grad_norm": 0.3296198546886444,
      "learning_rate": 8e-05,
      "loss": 1.8433,
      "step": 707
    },
    {
      "epoch": 0.07651572462984978,
      "grad_norm": 0.36884406208992004,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 708
    },
    {
      "epoch": 0.07662379768723657,
      "grad_norm": 0.3298288881778717,
      "learning_rate": 8e-05,
      "loss": 1.6976,
      "step": 709
    },
    {
      "epoch": 0.07673187074462337,
      "grad_norm": 0.32953813672065735,
      "learning_rate": 8e-05,
      "loss": 1.625,
      "step": 710
    },
    {
      "epoch": 0.07683994380201016,
      "grad_norm": 0.35480138659477234,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 711
    },
    {
      "epoch": 0.07694801685939695,
      "grad_norm": 0.3599317967891693,
      "learning_rate": 8e-05,
      "loss": 1.84,
      "step": 712
    },
    {
      "epoch": 0.07705608991678374,
      "grad_norm": 0.35181501507759094,
      "learning_rate": 8e-05,
      "loss": 1.7176,
      "step": 713
    },
    {
      "epoch": 0.07716416297417054,
      "grad_norm": 0.3353733420372009,
      "learning_rate": 8e-05,
      "loss": 1.8438,
      "step": 714
    },
    {
      "epoch": 0.07727223603155733,
      "grad_norm": 0.35124659538269043,
      "learning_rate": 8e-05,
      "loss": 1.938,
      "step": 715
    },
    {
      "epoch": 0.07738030908894412,
      "grad_norm": 0.4020790457725525,
      "learning_rate": 8e-05,
      "loss": 1.8665,
      "step": 716
    },
    {
      "epoch": 0.07748838214633091,
      "grad_norm": 0.3336130380630493,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 717
    },
    {
      "epoch": 0.07759645520371772,
      "grad_norm": 0.32654890418052673,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 718
    },
    {
      "epoch": 0.07770452826110451,
      "grad_norm": 0.35890254378318787,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 719
    },
    {
      "epoch": 0.0778126013184913,
      "grad_norm": 0.3621284067630768,
      "learning_rate": 8e-05,
      "loss": 1.8362,
      "step": 720
    },
    {
      "epoch": 0.07792067437587809,
      "grad_norm": 0.36673128604888916,
      "learning_rate": 8e-05,
      "loss": 1.8275,
      "step": 721
    },
    {
      "epoch": 0.07802874743326489,
      "grad_norm": 0.34916919469833374,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 722
    },
    {
      "epoch": 0.07813682049065168,
      "grad_norm": 0.33866086602211,
      "learning_rate": 8e-05,
      "loss": 1.6307,
      "step": 723
    },
    {
      "epoch": 0.07824489354803847,
      "grad_norm": 0.36578813195228577,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 724
    },
    {
      "epoch": 0.07835296660542526,
      "grad_norm": 0.3880551755428314,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 725
    },
    {
      "epoch": 0.07846103966281207,
      "grad_norm": 0.3700942099094391,
      "learning_rate": 8e-05,
      "loss": 1.909,
      "step": 726
    },
    {
      "epoch": 0.07856911272019886,
      "grad_norm": 0.4181189239025116,
      "learning_rate": 8e-05,
      "loss": 1.8998,
      "step": 727
    },
    {
      "epoch": 0.07867718577758565,
      "grad_norm": 0.4106231927871704,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 728
    },
    {
      "epoch": 0.07878525883497244,
      "grad_norm": 0.403994083404541,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 729
    },
    {
      "epoch": 0.07889333189235924,
      "grad_norm": 0.3424227833747864,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 730
    },
    {
      "epoch": 0.07900140494974603,
      "grad_norm": 0.3473060727119446,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 731
    },
    {
      "epoch": 0.07910947800713282,
      "grad_norm": 0.37217941880226135,
      "learning_rate": 8e-05,
      "loss": 1.8698,
      "step": 732
    },
    {
      "epoch": 0.07921755106451961,
      "grad_norm": 0.3986811637878418,
      "learning_rate": 8e-05,
      "loss": 1.8949,
      "step": 733
    },
    {
      "epoch": 0.07932562412190641,
      "grad_norm": 0.382118284702301,
      "learning_rate": 8e-05,
      "loss": 1.9121,
      "step": 734
    },
    {
      "epoch": 0.0794336971792932,
      "grad_norm": 0.34268704056739807,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 735
    },
    {
      "epoch": 0.07954177023667999,
      "grad_norm": 0.3592139482498169,
      "learning_rate": 8e-05,
      "loss": 1.9336,
      "step": 736
    },
    {
      "epoch": 0.07964984329406678,
      "grad_norm": 0.36988234519958496,
      "learning_rate": 8e-05,
      "loss": 1.8503,
      "step": 737
    },
    {
      "epoch": 0.07975791635145359,
      "grad_norm": 0.38400474190711975,
      "learning_rate": 8e-05,
      "loss": 1.9291,
      "step": 738
    },
    {
      "epoch": 0.07986598940884038,
      "grad_norm": 0.3458898961544037,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 739
    },
    {
      "epoch": 0.07997406246622717,
      "grad_norm": 0.3562629818916321,
      "learning_rate": 8e-05,
      "loss": 1.8328,
      "step": 740
    },
    {
      "epoch": 0.08008213552361396,
      "grad_norm": 0.36015400290489197,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 741
    },
    {
      "epoch": 0.08019020858100076,
      "grad_norm": 0.3556027114391327,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 742
    },
    {
      "epoch": 0.08029828163838755,
      "grad_norm": 0.3555619716644287,
      "learning_rate": 8e-05,
      "loss": 1.8762,
      "step": 743
    },
    {
      "epoch": 0.08040635469577434,
      "grad_norm": 0.3548396825790405,
      "learning_rate": 8e-05,
      "loss": 1.895,
      "step": 744
    },
    {
      "epoch": 0.08051442775316113,
      "grad_norm": 0.352502703666687,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 745
    },
    {
      "epoch": 0.08062250081054793,
      "grad_norm": 0.372275173664093,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 746
    },
    {
      "epoch": 0.08073057386793472,
      "grad_norm": 0.3499962389469147,
      "learning_rate": 8e-05,
      "loss": 1.8736,
      "step": 747
    },
    {
      "epoch": 0.08083864692532151,
      "grad_norm": 0.35915976762771606,
      "learning_rate": 8e-05,
      "loss": 1.8952,
      "step": 748
    },
    {
      "epoch": 0.0809467199827083,
      "grad_norm": 0.39708012342453003,
      "learning_rate": 8e-05,
      "loss": 2.0697,
      "step": 749
    },
    {
      "epoch": 0.08105479304009511,
      "grad_norm": 0.3751751184463501,
      "learning_rate": 8e-05,
      "loss": 1.8342,
      "step": 750
    },
    {
      "epoch": 0.0811628660974819,
      "grad_norm": 0.37397319078445435,
      "learning_rate": 8e-05,
      "loss": 1.7358,
      "step": 751
    },
    {
      "epoch": 0.08127093915486869,
      "grad_norm": 0.3618088960647583,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 752
    },
    {
      "epoch": 0.08137901221225549,
      "grad_norm": 0.3574982285499573,
      "learning_rate": 8e-05,
      "loss": 1.8672,
      "step": 753
    },
    {
      "epoch": 0.08148708526964228,
      "grad_norm": 0.386802613735199,
      "learning_rate": 8e-05,
      "loss": 1.9458,
      "step": 754
    },
    {
      "epoch": 0.08159515832702907,
      "grad_norm": 0.36350077390670776,
      "learning_rate": 8e-05,
      "loss": 1.9469,
      "step": 755
    },
    {
      "epoch": 0.08170323138441586,
      "grad_norm": 0.33847320079803467,
      "learning_rate": 8e-05,
      "loss": 1.7246,
      "step": 756
    },
    {
      "epoch": 0.08181130444180267,
      "grad_norm": 0.3289634585380554,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 757
    },
    {
      "epoch": 0.08191937749918946,
      "grad_norm": 0.32910841703414917,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 758
    },
    {
      "epoch": 0.08202745055657625,
      "grad_norm": 0.3602372407913208,
      "learning_rate": 8e-05,
      "loss": 1.8185,
      "step": 759
    },
    {
      "epoch": 0.08213552361396304,
      "grad_norm": 0.3620285987854004,
      "learning_rate": 8e-05,
      "loss": 1.928,
      "step": 760
    },
    {
      "epoch": 0.08224359667134984,
      "grad_norm": 0.3445507884025574,
      "learning_rate": 8e-05,
      "loss": 1.9045,
      "step": 761
    },
    {
      "epoch": 0.08235166972873663,
      "grad_norm": 0.3363454043865204,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 762
    },
    {
      "epoch": 0.08245974278612342,
      "grad_norm": 0.3430785536766052,
      "learning_rate": 8e-05,
      "loss": 1.5866,
      "step": 763
    },
    {
      "epoch": 0.08256781584351021,
      "grad_norm": 0.3677377700805664,
      "learning_rate": 8e-05,
      "loss": 2.0467,
      "step": 764
    },
    {
      "epoch": 0.08267588890089701,
      "grad_norm": 0.35696345567703247,
      "learning_rate": 8e-05,
      "loss": 1.824,
      "step": 765
    },
    {
      "epoch": 0.0827839619582838,
      "grad_norm": 0.3691982924938202,
      "learning_rate": 8e-05,
      "loss": 1.9481,
      "step": 766
    },
    {
      "epoch": 0.08289203501567059,
      "grad_norm": 0.33297982811927795,
      "learning_rate": 8e-05,
      "loss": 1.634,
      "step": 767
    },
    {
      "epoch": 0.08300010807305738,
      "grad_norm": 0.3267819881439209,
      "learning_rate": 8e-05,
      "loss": 1.6767,
      "step": 768
    },
    {
      "epoch": 0.08310818113044419,
      "grad_norm": 0.3792954385280609,
      "learning_rate": 8e-05,
      "loss": 1.7697,
      "step": 769
    },
    {
      "epoch": 0.08321625418783098,
      "grad_norm": 0.3626005947589874,
      "learning_rate": 8e-05,
      "loss": 1.8389,
      "step": 770
    },
    {
      "epoch": 0.08332432724521777,
      "grad_norm": 0.37124645709991455,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 771
    },
    {
      "epoch": 0.08343240030260456,
      "grad_norm": 0.3513961732387543,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 772
    },
    {
      "epoch": 0.08354047335999136,
      "grad_norm": 0.34473925828933716,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 773
    },
    {
      "epoch": 0.08364854641737815,
      "grad_norm": 0.34735044836997986,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 774
    },
    {
      "epoch": 0.08375661947476494,
      "grad_norm": 0.37575075030326843,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 775
    },
    {
      "epoch": 0.08386469253215173,
      "grad_norm": 0.39621371030807495,
      "learning_rate": 8e-05,
      "loss": 1.9143,
      "step": 776
    },
    {
      "epoch": 0.08397276558953853,
      "grad_norm": 0.3341212570667267,
      "learning_rate": 8e-05,
      "loss": 1.7078,
      "step": 777
    },
    {
      "epoch": 0.08408083864692532,
      "grad_norm": 0.4078580141067505,
      "learning_rate": 8e-05,
      "loss": 1.8777,
      "step": 778
    },
    {
      "epoch": 0.08418891170431211,
      "grad_norm": 0.3987581431865692,
      "learning_rate": 8e-05,
      "loss": 1.923,
      "step": 779
    },
    {
      "epoch": 0.0842969847616989,
      "grad_norm": 0.3557502031326294,
      "learning_rate": 8e-05,
      "loss": 1.6162,
      "step": 780
    },
    {
      "epoch": 0.08440505781908571,
      "grad_norm": 0.32769715785980225,
      "learning_rate": 8e-05,
      "loss": 1.5147,
      "step": 781
    },
    {
      "epoch": 0.0845131308764725,
      "grad_norm": 0.35191041231155396,
      "learning_rate": 8e-05,
      "loss": 1.6752,
      "step": 782
    },
    {
      "epoch": 0.08462120393385929,
      "grad_norm": 0.3392881155014038,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 783
    },
    {
      "epoch": 0.08472927699124608,
      "grad_norm": 0.33011457324028015,
      "learning_rate": 8e-05,
      "loss": 1.7721,
      "step": 784
    },
    {
      "epoch": 0.08483735004863288,
      "grad_norm": 0.3555687963962555,
      "learning_rate": 8e-05,
      "loss": 1.7182,
      "step": 785
    },
    {
      "epoch": 0.08494542310601967,
      "grad_norm": 0.3733087480068207,
      "learning_rate": 8e-05,
      "loss": 2.0021,
      "step": 786
    },
    {
      "epoch": 0.08505349616340646,
      "grad_norm": 0.36685800552368164,
      "learning_rate": 8e-05,
      "loss": 1.8644,
      "step": 787
    },
    {
      "epoch": 0.08516156922079325,
      "grad_norm": 0.3459504246711731,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 788
    },
    {
      "epoch": 0.08526964227818006,
      "grad_norm": 0.36492791771888733,
      "learning_rate": 8e-05,
      "loss": 1.8778,
      "step": 789
    },
    {
      "epoch": 0.08537771533556685,
      "grad_norm": 0.3297025263309479,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 790
    },
    {
      "epoch": 0.08548578839295363,
      "grad_norm": 0.32382282614707947,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 791
    },
    {
      "epoch": 0.08559386145034042,
      "grad_norm": 0.36852508783340454,
      "learning_rate": 8e-05,
      "loss": 1.8681,
      "step": 792
    },
    {
      "epoch": 0.08570193450772723,
      "grad_norm": 0.3332373797893524,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 793
    },
    {
      "epoch": 0.08581000756511402,
      "grad_norm": 0.32317543029785156,
      "learning_rate": 8e-05,
      "loss": 1.7088,
      "step": 794
    },
    {
      "epoch": 0.08591808062250081,
      "grad_norm": 0.34304043650627136,
      "learning_rate": 8e-05,
      "loss": 1.7349,
      "step": 795
    },
    {
      "epoch": 0.0860261536798876,
      "grad_norm": 0.35308459401130676,
      "learning_rate": 8e-05,
      "loss": 1.9224,
      "step": 796
    },
    {
      "epoch": 0.0861342267372744,
      "grad_norm": 0.37210604548454285,
      "learning_rate": 8e-05,
      "loss": 1.9415,
      "step": 797
    },
    {
      "epoch": 0.08624229979466119,
      "grad_norm": 0.336307168006897,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 798
    },
    {
      "epoch": 0.08635037285204798,
      "grad_norm": 0.3659408986568451,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 799
    },
    {
      "epoch": 0.08645844590943477,
      "grad_norm": 0.334870308637619,
      "learning_rate": 8e-05,
      "loss": 1.6286,
      "step": 800
    },
    {
      "epoch": 0.08656651896682158,
      "grad_norm": 0.37102147936820984,
      "learning_rate": 8e-05,
      "loss": 1.9711,
      "step": 801
    },
    {
      "epoch": 0.08667459202420837,
      "grad_norm": 0.37157055735588074,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 802
    },
    {
      "epoch": 0.08678266508159516,
      "grad_norm": 0.34305718541145325,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 803
    },
    {
      "epoch": 0.08689073813898195,
      "grad_norm": 0.3572945296764374,
      "learning_rate": 8e-05,
      "loss": 1.9602,
      "step": 804
    },
    {
      "epoch": 0.08699881119636875,
      "grad_norm": 0.36801835894584656,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 805
    },
    {
      "epoch": 0.08710688425375554,
      "grad_norm": 0.3747352361679077,
      "learning_rate": 8e-05,
      "loss": 1.9976,
      "step": 806
    },
    {
      "epoch": 0.08721495731114233,
      "grad_norm": 0.37714725732803345,
      "learning_rate": 8e-05,
      "loss": 1.5962,
      "step": 807
    },
    {
      "epoch": 0.08732303036852912,
      "grad_norm": 0.36145877838134766,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 808
    },
    {
      "epoch": 0.08743110342591592,
      "grad_norm": 0.33727219700813293,
      "learning_rate": 8e-05,
      "loss": 1.8809,
      "step": 809
    },
    {
      "epoch": 0.08753917648330271,
      "grad_norm": 0.4171218276023865,
      "learning_rate": 8e-05,
      "loss": 1.929,
      "step": 810
    },
    {
      "epoch": 0.0876472495406895,
      "grad_norm": 0.3338264226913452,
      "learning_rate": 8e-05,
      "loss": 1.7619,
      "step": 811
    },
    {
      "epoch": 0.0877553225980763,
      "grad_norm": 0.32613831758499146,
      "learning_rate": 8e-05,
      "loss": 1.6547,
      "step": 812
    },
    {
      "epoch": 0.0878633956554631,
      "grad_norm": 0.33511805534362793,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 813
    },
    {
      "epoch": 0.08797146871284989,
      "grad_norm": 0.33658602833747864,
      "learning_rate": 8e-05,
      "loss": 1.8978,
      "step": 814
    },
    {
      "epoch": 0.08807954177023668,
      "grad_norm": 0.3420501947402954,
      "learning_rate": 8e-05,
      "loss": 1.6778,
      "step": 815
    },
    {
      "epoch": 0.08818761482762347,
      "grad_norm": 0.336590975522995,
      "learning_rate": 8e-05,
      "loss": 1.7688,
      "step": 816
    },
    {
      "epoch": 0.08829568788501027,
      "grad_norm": 0.36553964018821716,
      "learning_rate": 8e-05,
      "loss": 1.7911,
      "step": 817
    },
    {
      "epoch": 0.08840376094239706,
      "grad_norm": 0.34292399883270264,
      "learning_rate": 8e-05,
      "loss": 1.8965,
      "step": 818
    },
    {
      "epoch": 0.08851183399978385,
      "grad_norm": 0.3505082130432129,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 819
    },
    {
      "epoch": 0.08861990705717064,
      "grad_norm": 0.3620678782463074,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 820
    },
    {
      "epoch": 0.08872798011455744,
      "grad_norm": 0.3877421021461487,
      "learning_rate": 8e-05,
      "loss": 1.9394,
      "step": 821
    },
    {
      "epoch": 0.08883605317194423,
      "grad_norm": 0.36055201292037964,
      "learning_rate": 8e-05,
      "loss": 1.6753,
      "step": 822
    },
    {
      "epoch": 0.08894412622933102,
      "grad_norm": 0.36477819085121155,
      "learning_rate": 8e-05,
      "loss": 1.9841,
      "step": 823
    },
    {
      "epoch": 0.08905219928671781,
      "grad_norm": 0.3890232741832733,
      "learning_rate": 8e-05,
      "loss": 1.5518,
      "step": 824
    },
    {
      "epoch": 0.08916027234410462,
      "grad_norm": 0.3630181849002838,
      "learning_rate": 8e-05,
      "loss": 1.8396,
      "step": 825
    },
    {
      "epoch": 0.08926834540149141,
      "grad_norm": 0.3804430663585663,
      "learning_rate": 8e-05,
      "loss": 1.994,
      "step": 826
    },
    {
      "epoch": 0.0893764184588782,
      "grad_norm": 0.33240169286727905,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 827
    },
    {
      "epoch": 0.08948449151626499,
      "grad_norm": 0.3295738399028778,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 828
    },
    {
      "epoch": 0.08959256457365179,
      "grad_norm": 0.34137552976608276,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 829
    },
    {
      "epoch": 0.08970063763103858,
      "grad_norm": 0.35221177339553833,
      "learning_rate": 8e-05,
      "loss": 1.8779,
      "step": 830
    },
    {
      "epoch": 0.08980871068842537,
      "grad_norm": 0.36139872670173645,
      "learning_rate": 8e-05,
      "loss": 1.797,
      "step": 831
    },
    {
      "epoch": 0.08991678374581216,
      "grad_norm": 0.3533114194869995,
      "learning_rate": 8e-05,
      "loss": 1.7841,
      "step": 832
    },
    {
      "epoch": 0.09002485680319897,
      "grad_norm": 0.3557817041873932,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 833
    },
    {
      "epoch": 0.09013292986058576,
      "grad_norm": 0.32702866196632385,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 834
    },
    {
      "epoch": 0.09024100291797255,
      "grad_norm": 0.33852875232696533,
      "learning_rate": 8e-05,
      "loss": 1.7097,
      "step": 835
    },
    {
      "epoch": 0.09034907597535935,
      "grad_norm": 0.3507237732410431,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 836
    },
    {
      "epoch": 0.09045714903274614,
      "grad_norm": 0.39895108342170715,
      "learning_rate": 8e-05,
      "loss": 1.9812,
      "step": 837
    },
    {
      "epoch": 0.09056522209013293,
      "grad_norm": 0.3445395231246948,
      "learning_rate": 8e-05,
      "loss": 1.4396,
      "step": 838
    },
    {
      "epoch": 0.09067329514751972,
      "grad_norm": 0.3465981185436249,
      "learning_rate": 8e-05,
      "loss": 1.7677,
      "step": 839
    },
    {
      "epoch": 0.09078136820490652,
      "grad_norm": 0.3418671786785126,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 840
    },
    {
      "epoch": 0.09088944126229331,
      "grad_norm": 0.3876403570175171,
      "learning_rate": 8e-05,
      "loss": 2.0084,
      "step": 841
    },
    {
      "epoch": 0.0909975143196801,
      "grad_norm": 0.3500586152076721,
      "learning_rate": 8e-05,
      "loss": 1.9042,
      "step": 842
    },
    {
      "epoch": 0.0911055873770669,
      "grad_norm": 0.3533492982387543,
      "learning_rate": 8e-05,
      "loss": 1.7577,
      "step": 843
    },
    {
      "epoch": 0.0912136604344537,
      "grad_norm": 0.34509995579719543,
      "learning_rate": 8e-05,
      "loss": 1.8736,
      "step": 844
    },
    {
      "epoch": 0.09132173349184049,
      "grad_norm": 0.34286263585090637,
      "learning_rate": 8e-05,
      "loss": 1.7923,
      "step": 845
    },
    {
      "epoch": 0.09142980654922728,
      "grad_norm": 0.34014856815338135,
      "learning_rate": 8e-05,
      "loss": 1.7946,
      "step": 846
    },
    {
      "epoch": 0.09153787960661407,
      "grad_norm": 0.3614908754825592,
      "learning_rate": 8e-05,
      "loss": 1.7894,
      "step": 847
    },
    {
      "epoch": 0.09164595266400087,
      "grad_norm": 0.3748004138469696,
      "learning_rate": 8e-05,
      "loss": 1.8593,
      "step": 848
    },
    {
      "epoch": 0.09175402572138766,
      "grad_norm": 0.3322995603084564,
      "learning_rate": 8e-05,
      "loss": 1.5334,
      "step": 849
    },
    {
      "epoch": 0.09186209877877445,
      "grad_norm": 0.36351263523101807,
      "learning_rate": 8e-05,
      "loss": 1.8635,
      "step": 850
    },
    {
      "epoch": 0.09197017183616124,
      "grad_norm": 0.34182462096214294,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 851
    },
    {
      "epoch": 0.09207824489354804,
      "grad_norm": 0.35077911615371704,
      "learning_rate": 8e-05,
      "loss": 1.9268,
      "step": 852
    },
    {
      "epoch": 0.09218631795093483,
      "grad_norm": 0.3587110638618469,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 853
    },
    {
      "epoch": 0.09229439100832162,
      "grad_norm": 0.3431946635246277,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 854
    },
    {
      "epoch": 0.09240246406570841,
      "grad_norm": 0.3534054458141327,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 855
    },
    {
      "epoch": 0.09251053712309522,
      "grad_norm": 0.33270472288131714,
      "learning_rate": 8e-05,
      "loss": 1.8176,
      "step": 856
    },
    {
      "epoch": 0.09261861018048201,
      "grad_norm": 0.3302570879459381,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 857
    },
    {
      "epoch": 0.0927266832378688,
      "grad_norm": 0.3459129333496094,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 858
    },
    {
      "epoch": 0.09283475629525559,
      "grad_norm": 0.3526592552661896,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 859
    },
    {
      "epoch": 0.09294282935264239,
      "grad_norm": 0.3434501886367798,
      "learning_rate": 8e-05,
      "loss": 1.5596,
      "step": 860
    },
    {
      "epoch": 0.09305090241002918,
      "grad_norm": 0.38379454612731934,
      "learning_rate": 8e-05,
      "loss": 1.8075,
      "step": 861
    },
    {
      "epoch": 0.09315897546741597,
      "grad_norm": 0.3195924758911133,
      "learning_rate": 8e-05,
      "loss": 1.6406,
      "step": 862
    },
    {
      "epoch": 0.09326704852480276,
      "grad_norm": 0.3645499348640442,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 863
    },
    {
      "epoch": 0.09337512158218957,
      "grad_norm": 0.3458186686038971,
      "learning_rate": 8e-05,
      "loss": 1.6969,
      "step": 864
    },
    {
      "epoch": 0.09348319463957636,
      "grad_norm": 0.3764459490776062,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 865
    },
    {
      "epoch": 0.09359126769696315,
      "grad_norm": 0.3541610538959503,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 866
    },
    {
      "epoch": 0.09369934075434994,
      "grad_norm": 0.3427675664424896,
      "learning_rate": 8e-05,
      "loss": 1.8361,
      "step": 867
    },
    {
      "epoch": 0.09380741381173674,
      "grad_norm": 0.37080663442611694,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 868
    },
    {
      "epoch": 0.09391548686912353,
      "grad_norm": 0.3579069674015045,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 869
    },
    {
      "epoch": 0.09402355992651032,
      "grad_norm": 0.37710386514663696,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 870
    },
    {
      "epoch": 0.09413163298389711,
      "grad_norm": 0.3282741606235504,
      "learning_rate": 8e-05,
      "loss": 1.6462,
      "step": 871
    },
    {
      "epoch": 0.09423970604128391,
      "grad_norm": 0.3715938627719879,
      "learning_rate": 8e-05,
      "loss": 1.8297,
      "step": 872
    },
    {
      "epoch": 0.0943477790986707,
      "grad_norm": 0.3862229585647583,
      "learning_rate": 8e-05,
      "loss": 1.7614,
      "step": 873
    },
    {
      "epoch": 0.0944558521560575,
      "grad_norm": 0.3499211072921753,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 874
    },
    {
      "epoch": 0.09456392521344428,
      "grad_norm": 0.3432254493236542,
      "learning_rate": 8e-05,
      "loss": 1.8058,
      "step": 875
    },
    {
      "epoch": 0.09467199827083109,
      "grad_norm": 0.34992873668670654,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 876
    },
    {
      "epoch": 0.09478007132821788,
      "grad_norm": 0.3644714653491974,
      "learning_rate": 8e-05,
      "loss": 1.8002,
      "step": 877
    },
    {
      "epoch": 0.09488814438560467,
      "grad_norm": 0.4058280289173126,
      "learning_rate": 8e-05,
      "loss": 1.7815,
      "step": 878
    },
    {
      "epoch": 0.09499621744299146,
      "grad_norm": 0.3541181683540344,
      "learning_rate": 8e-05,
      "loss": 1.8194,
      "step": 879
    },
    {
      "epoch": 0.09510429050037826,
      "grad_norm": 0.36470821499824524,
      "learning_rate": 8e-05,
      "loss": 1.9349,
      "step": 880
    },
    {
      "epoch": 0.09521236355776505,
      "grad_norm": 0.34866970777511597,
      "learning_rate": 8e-05,
      "loss": 1.8882,
      "step": 881
    },
    {
      "epoch": 0.09532043661515184,
      "grad_norm": 0.3558403551578522,
      "learning_rate": 8e-05,
      "loss": 1.941,
      "step": 882
    },
    {
      "epoch": 0.09542850967253863,
      "grad_norm": 0.3859046399593353,
      "learning_rate": 8e-05,
      "loss": 1.9231,
      "step": 883
    },
    {
      "epoch": 0.09553658272992543,
      "grad_norm": 0.36378538608551025,
      "learning_rate": 8e-05,
      "loss": 1.9172,
      "step": 884
    },
    {
      "epoch": 0.09564465578731222,
      "grad_norm": 0.3842751085758209,
      "learning_rate": 8e-05,
      "loss": 2.0204,
      "step": 885
    },
    {
      "epoch": 0.09575272884469901,
      "grad_norm": 0.37410861253738403,
      "learning_rate": 8e-05,
      "loss": 1.8148,
      "step": 886
    },
    {
      "epoch": 0.0958608019020858,
      "grad_norm": 0.38042762875556946,
      "learning_rate": 8e-05,
      "loss": 1.8696,
      "step": 887
    },
    {
      "epoch": 0.09596887495947261,
      "grad_norm": 0.3503395617008209,
      "learning_rate": 8e-05,
      "loss": 1.6475,
      "step": 888
    },
    {
      "epoch": 0.0960769480168594,
      "grad_norm": 0.3514329791069031,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 889
    },
    {
      "epoch": 0.09618502107424619,
      "grad_norm": 0.3578362762928009,
      "learning_rate": 8e-05,
      "loss": 1.9048,
      "step": 890
    },
    {
      "epoch": 0.09629309413163298,
      "grad_norm": 0.38349369168281555,
      "learning_rate": 8e-05,
      "loss": 1.7972,
      "step": 891
    },
    {
      "epoch": 0.09640116718901978,
      "grad_norm": 0.39928188920021057,
      "learning_rate": 8e-05,
      "loss": 1.928,
      "step": 892
    },
    {
      "epoch": 0.09650924024640657,
      "grad_norm": 0.390762597322464,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 893
    },
    {
      "epoch": 0.09661731330379336,
      "grad_norm": 0.34487417340278625,
      "learning_rate": 8e-05,
      "loss": 1.8621,
      "step": 894
    },
    {
      "epoch": 0.09672538636118015,
      "grad_norm": 0.3497505486011505,
      "learning_rate": 8e-05,
      "loss": 1.7434,
      "step": 895
    },
    {
      "epoch": 0.09683345941856696,
      "grad_norm": 0.42294076085090637,
      "learning_rate": 8e-05,
      "loss": 2.0171,
      "step": 896
    },
    {
      "epoch": 0.09694153247595375,
      "grad_norm": 0.36248984932899475,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 897
    },
    {
      "epoch": 0.09704960553334054,
      "grad_norm": 0.35715553164482117,
      "learning_rate": 8e-05,
      "loss": 1.9705,
      "step": 898
    },
    {
      "epoch": 0.09715767859072733,
      "grad_norm": 0.34239330887794495,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 899
    },
    {
      "epoch": 0.09726575164811413,
      "grad_norm": 0.3759861886501312,
      "learning_rate": 8e-05,
      "loss": 1.7913,
      "step": 900
    },
    {
      "epoch": 0.09737382470550092,
      "grad_norm": 0.40702810883522034,
      "learning_rate": 8e-05,
      "loss": 1.8557,
      "step": 901
    },
    {
      "epoch": 0.09748189776288771,
      "grad_norm": 0.32868754863739014,
      "learning_rate": 8e-05,
      "loss": 1.6483,
      "step": 902
    },
    {
      "epoch": 0.0975899708202745,
      "grad_norm": 0.35593172907829285,
      "learning_rate": 8e-05,
      "loss": 1.895,
      "step": 903
    },
    {
      "epoch": 0.0976980438776613,
      "grad_norm": 0.35074326395988464,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 904
    },
    {
      "epoch": 0.09780611693504809,
      "grad_norm": 0.34221625328063965,
      "learning_rate": 8e-05,
      "loss": 1.7098,
      "step": 905
    },
    {
      "epoch": 0.09791418999243488,
      "grad_norm": 0.3473942279815674,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 906
    },
    {
      "epoch": 0.09802226304982167,
      "grad_norm": 0.3827817440032959,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 907
    },
    {
      "epoch": 0.09813033610720848,
      "grad_norm": 0.3702685236930847,
      "learning_rate": 8e-05,
      "loss": 1.9114,
      "step": 908
    },
    {
      "epoch": 0.09823840916459527,
      "grad_norm": 0.35614997148513794,
      "learning_rate": 8e-05,
      "loss": 1.8615,
      "step": 909
    },
    {
      "epoch": 0.09834648222198206,
      "grad_norm": 0.3270261287689209,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 910
    },
    {
      "epoch": 0.09845455527936885,
      "grad_norm": 0.3622662425041199,
      "learning_rate": 8e-05,
      "loss": 1.7252,
      "step": 911
    },
    {
      "epoch": 0.09856262833675565,
      "grad_norm": 0.3668723404407501,
      "learning_rate": 8e-05,
      "loss": 1.8184,
      "step": 912
    },
    {
      "epoch": 0.09867070139414244,
      "grad_norm": 0.34703904390335083,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 913
    },
    {
      "epoch": 0.09877877445152923,
      "grad_norm": 0.38205552101135254,
      "learning_rate": 8e-05,
      "loss": 1.8642,
      "step": 914
    },
    {
      "epoch": 0.09888684750891602,
      "grad_norm": 0.3451690971851349,
      "learning_rate": 8e-05,
      "loss": 1.6173,
      "step": 915
    },
    {
      "epoch": 0.09899492056630282,
      "grad_norm": 0.3449813723564148,
      "learning_rate": 8e-05,
      "loss": 1.7767,
      "step": 916
    },
    {
      "epoch": 0.09910299362368961,
      "grad_norm": 0.4120391607284546,
      "learning_rate": 8e-05,
      "loss": 2.0339,
      "step": 917
    },
    {
      "epoch": 0.0992110666810764,
      "grad_norm": 0.3527565598487854,
      "learning_rate": 8e-05,
      "loss": 1.8451,
      "step": 918
    },
    {
      "epoch": 0.09931913973846321,
      "grad_norm": 0.347257137298584,
      "learning_rate": 8e-05,
      "loss": 1.8731,
      "step": 919
    },
    {
      "epoch": 0.09942721279585,
      "grad_norm": 0.3716275990009308,
      "learning_rate": 8e-05,
      "loss": 1.9667,
      "step": 920
    },
    {
      "epoch": 0.09953528585323679,
      "grad_norm": 0.37038666009902954,
      "learning_rate": 8e-05,
      "loss": 1.8925,
      "step": 921
    },
    {
      "epoch": 0.09964335891062358,
      "grad_norm": 0.32822686433792114,
      "learning_rate": 8e-05,
      "loss": 1.6566,
      "step": 922
    },
    {
      "epoch": 0.09975143196801038,
      "grad_norm": 0.34583452343940735,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 923
    },
    {
      "epoch": 0.09985950502539717,
      "grad_norm": 0.3826597332954407,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 924
    },
    {
      "epoch": 0.09996757808278396,
      "grad_norm": 0.34853339195251465,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 925
    },
    {
      "epoch": 0.10007565114017075,
      "grad_norm": 0.32940131425857544,
      "learning_rate": 8e-05,
      "loss": 1.6778,
      "step": 926
    },
    {
      "epoch": 0.10018372419755756,
      "grad_norm": 0.3581061065196991,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 927
    },
    {
      "epoch": 0.10029179725494435,
      "grad_norm": 0.3408588469028473,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 928
    },
    {
      "epoch": 0.10039987031233114,
      "grad_norm": 0.34937652945518494,
      "learning_rate": 8e-05,
      "loss": 1.7455,
      "step": 929
    },
    {
      "epoch": 0.10050794336971793,
      "grad_norm": 0.3451722264289856,
      "learning_rate": 8e-05,
      "loss": 1.5794,
      "step": 930
    },
    {
      "epoch": 0.10061601642710473,
      "grad_norm": 0.32624149322509766,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 931
    },
    {
      "epoch": 0.10072408948449152,
      "grad_norm": 0.3210192322731018,
      "learning_rate": 8e-05,
      "loss": 1.5514,
      "step": 932
    },
    {
      "epoch": 0.10083216254187831,
      "grad_norm": 0.33881422877311707,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 933
    },
    {
      "epoch": 0.1009402355992651,
      "grad_norm": 0.3912752866744995,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 934
    },
    {
      "epoch": 0.1010483086566519,
      "grad_norm": 0.36140796542167664,
      "learning_rate": 8e-05,
      "loss": 1.6899,
      "step": 935
    },
    {
      "epoch": 0.10115638171403869,
      "grad_norm": 0.3649536073207855,
      "learning_rate": 8e-05,
      "loss": 1.8636,
      "step": 936
    },
    {
      "epoch": 0.10126445477142548,
      "grad_norm": 0.3420322835445404,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 937
    },
    {
      "epoch": 0.10137252782881227,
      "grad_norm": 0.36467358469963074,
      "learning_rate": 8e-05,
      "loss": 1.8276,
      "step": 938
    },
    {
      "epoch": 0.10148060088619908,
      "grad_norm": 0.3385829031467438,
      "learning_rate": 8e-05,
      "loss": 1.7406,
      "step": 939
    },
    {
      "epoch": 0.10158867394358587,
      "grad_norm": 0.34874477982521057,
      "learning_rate": 8e-05,
      "loss": 1.8694,
      "step": 940
    },
    {
      "epoch": 0.10169674700097266,
      "grad_norm": 0.35837405920028687,
      "learning_rate": 8e-05,
      "loss": 1.7741,
      "step": 941
    },
    {
      "epoch": 0.10180482005835945,
      "grad_norm": 0.40015703439712524,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 942
    },
    {
      "epoch": 0.10191289311574625,
      "grad_norm": 0.39776626229286194,
      "learning_rate": 8e-05,
      "loss": 2.0724,
      "step": 943
    },
    {
      "epoch": 0.10202096617313304,
      "grad_norm": 0.3498876392841339,
      "learning_rate": 8e-05,
      "loss": 1.8087,
      "step": 944
    },
    {
      "epoch": 0.10212903923051983,
      "grad_norm": 0.3590984046459198,
      "learning_rate": 8e-05,
      "loss": 1.7932,
      "step": 945
    },
    {
      "epoch": 0.10223711228790662,
      "grad_norm": 0.33495789766311646,
      "learning_rate": 8e-05,
      "loss": 1.7068,
      "step": 946
    },
    {
      "epoch": 0.10234518534529342,
      "grad_norm": 0.3117012083530426,
      "learning_rate": 8e-05,
      "loss": 1.5938,
      "step": 947
    },
    {
      "epoch": 0.10245325840268021,
      "grad_norm": 0.3461303412914276,
      "learning_rate": 8e-05,
      "loss": 1.6385,
      "step": 948
    },
    {
      "epoch": 0.102561331460067,
      "grad_norm": 0.36226168274879456,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 949
    },
    {
      "epoch": 0.1026694045174538,
      "grad_norm": 0.3622657060623169,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 950
    },
    {
      "epoch": 0.1027774775748406,
      "grad_norm": 0.3543497920036316,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 951
    },
    {
      "epoch": 0.10288555063222739,
      "grad_norm": 0.3250940144062042,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 952
    },
    {
      "epoch": 0.10299362368961418,
      "grad_norm": 0.3531031310558319,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 953
    },
    {
      "epoch": 0.10310169674700097,
      "grad_norm": 0.3336448669433594,
      "learning_rate": 8e-05,
      "loss": 1.7333,
      "step": 954
    },
    {
      "epoch": 0.10320976980438777,
      "grad_norm": 0.3727389872074127,
      "learning_rate": 8e-05,
      "loss": 1.8759,
      "step": 955
    },
    {
      "epoch": 0.10331784286177456,
      "grad_norm": 0.38414549827575684,
      "learning_rate": 8e-05,
      "loss": 1.8427,
      "step": 956
    },
    {
      "epoch": 0.10342591591916135,
      "grad_norm": 0.33417871594429016,
      "learning_rate": 8e-05,
      "loss": 1.7205,
      "step": 957
    },
    {
      "epoch": 0.10353398897654814,
      "grad_norm": 0.4005602300167084,
      "learning_rate": 8e-05,
      "loss": 1.9061,
      "step": 958
    },
    {
      "epoch": 0.10364206203393495,
      "grad_norm": 0.36298057436943054,
      "learning_rate": 8e-05,
      "loss": 1.5239,
      "step": 959
    },
    {
      "epoch": 0.10375013509132174,
      "grad_norm": 0.35376080870628357,
      "learning_rate": 8e-05,
      "loss": 1.7882,
      "step": 960
    },
    {
      "epoch": 0.10385820814870853,
      "grad_norm": 0.33489012718200684,
      "learning_rate": 8e-05,
      "loss": 1.6245,
      "step": 961
    },
    {
      "epoch": 0.10396628120609532,
      "grad_norm": 0.36470216512680054,
      "learning_rate": 8e-05,
      "loss": 1.6185,
      "step": 962
    },
    {
      "epoch": 0.10407435426348212,
      "grad_norm": 0.3187197744846344,
      "learning_rate": 8e-05,
      "loss": 1.6175,
      "step": 963
    },
    {
      "epoch": 0.10418242732086891,
      "grad_norm": 0.3826877176761627,
      "learning_rate": 8e-05,
      "loss": 1.9342,
      "step": 964
    },
    {
      "epoch": 0.1042905003782557,
      "grad_norm": 0.3329418897628784,
      "learning_rate": 8e-05,
      "loss": 1.6955,
      "step": 965
    },
    {
      "epoch": 0.10439857343564249,
      "grad_norm": 0.35413625836372375,
      "learning_rate": 8e-05,
      "loss": 1.9804,
      "step": 966
    },
    {
      "epoch": 0.10450664649302929,
      "grad_norm": 0.3780409097671509,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 967
    },
    {
      "epoch": 0.10461471955041608,
      "grad_norm": 0.34960803389549255,
      "learning_rate": 8e-05,
      "loss": 1.749,
      "step": 968
    },
    {
      "epoch": 0.10472279260780287,
      "grad_norm": 0.36611899733543396,
      "learning_rate": 8e-05,
      "loss": 1.8597,
      "step": 969
    },
    {
      "epoch": 0.10483086566518966,
      "grad_norm": 0.36785924434661865,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 970
    },
    {
      "epoch": 0.10493893872257647,
      "grad_norm": 0.34815487265586853,
      "learning_rate": 8e-05,
      "loss": 1.6009,
      "step": 971
    },
    {
      "epoch": 0.10504701177996326,
      "grad_norm": 0.3491874933242798,
      "learning_rate": 8e-05,
      "loss": 1.8374,
      "step": 972
    },
    {
      "epoch": 0.10515508483735005,
      "grad_norm": 0.3539784550666809,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 973
    },
    {
      "epoch": 0.10526315789473684,
      "grad_norm": 0.36112168431282043,
      "learning_rate": 8e-05,
      "loss": 1.8739,
      "step": 974
    },
    {
      "epoch": 0.10537123095212364,
      "grad_norm": 0.35476866364479065,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 975
    },
    {
      "epoch": 0.10547930400951043,
      "grad_norm": 0.3740367293357849,
      "learning_rate": 8e-05,
      "loss": 1.8731,
      "step": 976
    },
    {
      "epoch": 0.10558737706689722,
      "grad_norm": 0.380718469619751,
      "learning_rate": 8e-05,
      "loss": 1.8771,
      "step": 977
    },
    {
      "epoch": 0.10569545012428401,
      "grad_norm": 0.3447098731994629,
      "learning_rate": 8e-05,
      "loss": 1.7702,
      "step": 978
    },
    {
      "epoch": 0.10580352318167081,
      "grad_norm": 0.3620442748069763,
      "learning_rate": 8e-05,
      "loss": 1.7765,
      "step": 979
    },
    {
      "epoch": 0.1059115962390576,
      "grad_norm": 0.3551279604434967,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 980
    },
    {
      "epoch": 0.1060196692964444,
      "grad_norm": 0.3821079432964325,
      "learning_rate": 8e-05,
      "loss": 1.8377,
      "step": 981
    },
    {
      "epoch": 0.10612774235383118,
      "grad_norm": 0.3099254369735718,
      "learning_rate": 8e-05,
      "loss": 1.6299,
      "step": 982
    },
    {
      "epoch": 0.10623581541121799,
      "grad_norm": 0.3426049053668976,
      "learning_rate": 8e-05,
      "loss": 1.679,
      "step": 983
    },
    {
      "epoch": 0.10634388846860478,
      "grad_norm": 0.4244281053543091,
      "learning_rate": 8e-05,
      "loss": 1.9273,
      "step": 984
    },
    {
      "epoch": 0.10645196152599157,
      "grad_norm": 0.37145841121673584,
      "learning_rate": 8e-05,
      "loss": 1.8507,
      "step": 985
    },
    {
      "epoch": 0.10656003458337836,
      "grad_norm": 0.34629592299461365,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 986
    },
    {
      "epoch": 0.10666810764076516,
      "grad_norm": 0.35548093914985657,
      "learning_rate": 8e-05,
      "loss": 1.6655,
      "step": 987
    },
    {
      "epoch": 0.10677618069815195,
      "grad_norm": 0.3681755065917969,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 988
    },
    {
      "epoch": 0.10688425375553874,
      "grad_norm": 0.345228374004364,
      "learning_rate": 8e-05,
      "loss": 1.6541,
      "step": 989
    },
    {
      "epoch": 0.10699232681292553,
      "grad_norm": 0.36760804057121277,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 990
    },
    {
      "epoch": 0.10710039987031233,
      "grad_norm": 0.35668179392814636,
      "learning_rate": 8e-05,
      "loss": 1.7939,
      "step": 991
    },
    {
      "epoch": 0.10720847292769912,
      "grad_norm": 0.40337005257606506,
      "learning_rate": 8e-05,
      "loss": 2.0045,
      "step": 992
    },
    {
      "epoch": 0.10731654598508591,
      "grad_norm": 0.3598701059818268,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 993
    },
    {
      "epoch": 0.1074246190424727,
      "grad_norm": 0.3536721467971802,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 994
    },
    {
      "epoch": 0.10753269209985951,
      "grad_norm": 0.3558799624443054,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 995
    },
    {
      "epoch": 0.1076407651572463,
      "grad_norm": 0.3420169949531555,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 996
    },
    {
      "epoch": 0.10774883821463309,
      "grad_norm": 0.35889682173728943,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 997
    },
    {
      "epoch": 0.10785691127201988,
      "grad_norm": 0.3473713994026184,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 998
    },
    {
      "epoch": 0.10796498432940668,
      "grad_norm": 0.3569095730781555,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 999
    },
    {
      "epoch": 0.10807305738679347,
      "grad_norm": 0.36021849513053894,
      "learning_rate": 8e-05,
      "loss": 1.8736,
      "step": 1000
    },
    {
      "epoch": 0.10818113044418026,
      "grad_norm": 0.38916531205177307,
      "learning_rate": 8e-05,
      "loss": 1.9118,
      "step": 1001
    },
    {
      "epoch": 0.10828920350156707,
      "grad_norm": 0.3654513359069824,
      "learning_rate": 8e-05,
      "loss": 1.9466,
      "step": 1002
    },
    {
      "epoch": 0.10839727655895386,
      "grad_norm": 0.3699779212474823,
      "learning_rate": 8e-05,
      "loss": 1.7439,
      "step": 1003
    },
    {
      "epoch": 0.10850534961634065,
      "grad_norm": 0.37093672156333923,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 1004
    },
    {
      "epoch": 0.10861342267372744,
      "grad_norm": 0.37014082074165344,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 1005
    },
    {
      "epoch": 0.10872149573111424,
      "grad_norm": 0.36623871326446533,
      "learning_rate": 8e-05,
      "loss": 1.835,
      "step": 1006
    },
    {
      "epoch": 0.10882956878850103,
      "grad_norm": 0.34573739767074585,
      "learning_rate": 8e-05,
      "loss": 1.8492,
      "step": 1007
    },
    {
      "epoch": 0.10893764184588782,
      "grad_norm": 0.3408907949924469,
      "learning_rate": 8e-05,
      "loss": 1.6964,
      "step": 1008
    },
    {
      "epoch": 0.10904571490327461,
      "grad_norm": 0.34560516476631165,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 1009
    },
    {
      "epoch": 0.10915378796066141,
      "grad_norm": 0.3811988830566406,
      "learning_rate": 8e-05,
      "loss": 1.9763,
      "step": 1010
    },
    {
      "epoch": 0.1092618610180482,
      "grad_norm": 0.4765642285346985,
      "learning_rate": 8e-05,
      "loss": 2.1039,
      "step": 1011
    },
    {
      "epoch": 0.109369934075435,
      "grad_norm": 0.34004926681518555,
      "learning_rate": 8e-05,
      "loss": 1.7518,
      "step": 1012
    },
    {
      "epoch": 0.10947800713282178,
      "grad_norm": 0.3730889558792114,
      "learning_rate": 8e-05,
      "loss": 1.897,
      "step": 1013
    },
    {
      "epoch": 0.10958608019020859,
      "grad_norm": 0.3418436050415039,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 1014
    },
    {
      "epoch": 0.10969415324759538,
      "grad_norm": 0.324800580739975,
      "learning_rate": 8e-05,
      "loss": 1.7566,
      "step": 1015
    },
    {
      "epoch": 0.10980222630498217,
      "grad_norm": 0.3472938537597656,
      "learning_rate": 8e-05,
      "loss": 1.6911,
      "step": 1016
    },
    {
      "epoch": 0.10991029936236896,
      "grad_norm": 0.3487228453159332,
      "learning_rate": 8e-05,
      "loss": 1.6507,
      "step": 1017
    },
    {
      "epoch": 0.11001837241975576,
      "grad_norm": 0.3387090563774109,
      "learning_rate": 8e-05,
      "loss": 1.7957,
      "step": 1018
    },
    {
      "epoch": 0.11012644547714255,
      "grad_norm": 0.33071210980415344,
      "learning_rate": 8e-05,
      "loss": 1.7122,
      "step": 1019
    },
    {
      "epoch": 0.11023451853452934,
      "grad_norm": 0.35964852571487427,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 1020
    },
    {
      "epoch": 0.11034259159191613,
      "grad_norm": 0.3740130364894867,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 1021
    },
    {
      "epoch": 0.11045066464930293,
      "grad_norm": 0.34640559554100037,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 1022
    },
    {
      "epoch": 0.11055873770668972,
      "grad_norm": 0.33933576941490173,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 1023
    },
    {
      "epoch": 0.11066681076407651,
      "grad_norm": 0.36161407828330994,
      "learning_rate": 8e-05,
      "loss": 1.7882,
      "step": 1024
    },
    {
      "epoch": 0.1107748838214633,
      "grad_norm": 0.3854907155036926,
      "learning_rate": 8e-05,
      "loss": 1.8705,
      "step": 1025
    },
    {
      "epoch": 0.11088295687885011,
      "grad_norm": 0.35772159695625305,
      "learning_rate": 8e-05,
      "loss": 1.6783,
      "step": 1026
    },
    {
      "epoch": 0.1109910299362369,
      "grad_norm": 0.3468589782714844,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 1027
    },
    {
      "epoch": 0.11109910299362369,
      "grad_norm": 0.3650378882884979,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 1028
    },
    {
      "epoch": 0.11120717605101048,
      "grad_norm": 0.3954007923603058,
      "learning_rate": 8e-05,
      "loss": 1.6899,
      "step": 1029
    },
    {
      "epoch": 0.11131524910839728,
      "grad_norm": 0.34445634484291077,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 1030
    },
    {
      "epoch": 0.11142332216578407,
      "grad_norm": 0.3457842469215393,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 1031
    },
    {
      "epoch": 0.11153139522317086,
      "grad_norm": 0.3583963215351105,
      "learning_rate": 8e-05,
      "loss": 1.8769,
      "step": 1032
    },
    {
      "epoch": 0.11163946828055765,
      "grad_norm": 0.3492564558982849,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 1033
    },
    {
      "epoch": 0.11174754133794446,
      "grad_norm": 0.3324894607067108,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 1034
    },
    {
      "epoch": 0.11185561439533125,
      "grad_norm": 0.34477636218070984,
      "learning_rate": 8e-05,
      "loss": 1.7223,
      "step": 1035
    },
    {
      "epoch": 0.11196368745271804,
      "grad_norm": 0.35887598991394043,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 1036
    },
    {
      "epoch": 0.11207176051010483,
      "grad_norm": 0.359162300825119,
      "learning_rate": 8e-05,
      "loss": 1.8598,
      "step": 1037
    },
    {
      "epoch": 0.11217983356749163,
      "grad_norm": 0.37809160351753235,
      "learning_rate": 8e-05,
      "loss": 1.8609,
      "step": 1038
    },
    {
      "epoch": 0.11228790662487842,
      "grad_norm": 0.34972062706947327,
      "learning_rate": 8e-05,
      "loss": 1.6576,
      "step": 1039
    },
    {
      "epoch": 0.11239597968226521,
      "grad_norm": 0.3460668623447418,
      "learning_rate": 8e-05,
      "loss": 1.6272,
      "step": 1040
    },
    {
      "epoch": 0.112504052739652,
      "grad_norm": 0.3861105144023895,
      "learning_rate": 8e-05,
      "loss": 2.0128,
      "step": 1041
    },
    {
      "epoch": 0.1126121257970388,
      "grad_norm": 0.35771337151527405,
      "learning_rate": 8e-05,
      "loss": 1.8703,
      "step": 1042
    },
    {
      "epoch": 0.1127201988544256,
      "grad_norm": 0.335125207901001,
      "learning_rate": 8e-05,
      "loss": 1.8183,
      "step": 1043
    },
    {
      "epoch": 0.11282827191181238,
      "grad_norm": 0.3440723121166229,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 1044
    },
    {
      "epoch": 0.11293634496919917,
      "grad_norm": 0.37276747822761536,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 1045
    },
    {
      "epoch": 0.11304441802658598,
      "grad_norm": 0.34359925985336304,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 1046
    },
    {
      "epoch": 0.11315249108397277,
      "grad_norm": 0.33626818656921387,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 1047
    },
    {
      "epoch": 0.11326056414135956,
      "grad_norm": 0.36569783091545105,
      "learning_rate": 8e-05,
      "loss": 1.8492,
      "step": 1048
    },
    {
      "epoch": 0.11336863719874635,
      "grad_norm": 0.33529019355773926,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 1049
    },
    {
      "epoch": 0.11347671025613315,
      "grad_norm": 0.33651697635650635,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 1050
    },
    {
      "epoch": 0.11358478331351994,
      "grad_norm": 0.3362615704536438,
      "learning_rate": 8e-05,
      "loss": 1.6028,
      "step": 1051
    },
    {
      "epoch": 0.11369285637090673,
      "grad_norm": 0.33662670850753784,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 1052
    },
    {
      "epoch": 0.11380092942829352,
      "grad_norm": 0.3547325134277344,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 1053
    },
    {
      "epoch": 0.11390900248568032,
      "grad_norm": 0.33198556303977966,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 1054
    },
    {
      "epoch": 0.11401707554306711,
      "grad_norm": 0.33353909850120544,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 1055
    },
    {
      "epoch": 0.1141251486004539,
      "grad_norm": 0.3565393388271332,
      "learning_rate": 8e-05,
      "loss": 1.8502,
      "step": 1056
    },
    {
      "epoch": 0.1142332216578407,
      "grad_norm": 0.37357306480407715,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 1057
    },
    {
      "epoch": 0.1143412947152275,
      "grad_norm": 0.37155699729919434,
      "learning_rate": 8e-05,
      "loss": 1.8798,
      "step": 1058
    },
    {
      "epoch": 0.11444936777261429,
      "grad_norm": 0.3555389642715454,
      "learning_rate": 8e-05,
      "loss": 1.8011,
      "step": 1059
    },
    {
      "epoch": 0.11455744083000108,
      "grad_norm": 0.3473822772502899,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 1060
    },
    {
      "epoch": 0.11466551388738787,
      "grad_norm": 0.34955134987831116,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 1061
    },
    {
      "epoch": 0.11477358694477467,
      "grad_norm": 0.36052483320236206,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 1062
    },
    {
      "epoch": 0.11488166000216146,
      "grad_norm": 0.32584789395332336,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 1063
    },
    {
      "epoch": 0.11498973305954825,
      "grad_norm": 0.3796273171901703,
      "learning_rate": 8e-05,
      "loss": 1.7895,
      "step": 1064
    },
    {
      "epoch": 0.11509780611693504,
      "grad_norm": 0.3480946123600006,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 1065
    },
    {
      "epoch": 0.11520587917432185,
      "grad_norm": 0.3912273347377777,
      "learning_rate": 8e-05,
      "loss": 1.9485,
      "step": 1066
    },
    {
      "epoch": 0.11531395223170864,
      "grad_norm": 0.3587355613708496,
      "learning_rate": 8e-05,
      "loss": 1.469,
      "step": 1067
    },
    {
      "epoch": 0.11542202528909543,
      "grad_norm": 0.36560285091400146,
      "learning_rate": 8e-05,
      "loss": 1.9126,
      "step": 1068
    },
    {
      "epoch": 0.11553009834648222,
      "grad_norm": 0.34391555190086365,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 1069
    },
    {
      "epoch": 0.11563817140386902,
      "grad_norm": 0.3561718165874481,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 1070
    },
    {
      "epoch": 0.11574624446125581,
      "grad_norm": 0.39095380902290344,
      "learning_rate": 8e-05,
      "loss": 1.9475,
      "step": 1071
    },
    {
      "epoch": 0.1158543175186426,
      "grad_norm": 0.3762200176715851,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 1072
    },
    {
      "epoch": 0.11596239057602939,
      "grad_norm": 0.3578304052352905,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 1073
    },
    {
      "epoch": 0.1160704636334162,
      "grad_norm": 0.3298514485359192,
      "learning_rate": 8e-05,
      "loss": 1.6766,
      "step": 1074
    },
    {
      "epoch": 0.11617853669080298,
      "grad_norm": 0.3591746687889099,
      "learning_rate": 8e-05,
      "loss": 1.8481,
      "step": 1075
    },
    {
      "epoch": 0.11628660974818977,
      "grad_norm": 0.3621608018875122,
      "learning_rate": 8e-05,
      "loss": 1.7246,
      "step": 1076
    },
    {
      "epoch": 0.11639468280557656,
      "grad_norm": 0.3578890562057495,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 1077
    },
    {
      "epoch": 0.11650275586296337,
      "grad_norm": 0.4235458970069885,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 1078
    },
    {
      "epoch": 0.11661082892035016,
      "grad_norm": 0.35560694336891174,
      "learning_rate": 8e-05,
      "loss": 1.9208,
      "step": 1079
    },
    {
      "epoch": 0.11671890197773695,
      "grad_norm": 0.3598167896270752,
      "learning_rate": 8e-05,
      "loss": 1.8029,
      "step": 1080
    },
    {
      "epoch": 0.11682697503512374,
      "grad_norm": 0.37738943099975586,
      "learning_rate": 8e-05,
      "loss": 1.9841,
      "step": 1081
    },
    {
      "epoch": 0.11693504809251054,
      "grad_norm": 0.3868809938430786,
      "learning_rate": 8e-05,
      "loss": 1.9314,
      "step": 1082
    },
    {
      "epoch": 0.11704312114989733,
      "grad_norm": 0.3519797623157501,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 1083
    },
    {
      "epoch": 0.11715119420728412,
      "grad_norm": 0.33653125166893005,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1084
    },
    {
      "epoch": 0.11725926726467092,
      "grad_norm": 0.3813892602920532,
      "learning_rate": 8e-05,
      "loss": 1.9151,
      "step": 1085
    },
    {
      "epoch": 0.11736734032205771,
      "grad_norm": 0.36971303820610046,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 1086
    },
    {
      "epoch": 0.1174754133794445,
      "grad_norm": 0.35533636808395386,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 1087
    },
    {
      "epoch": 0.1175834864368313,
      "grad_norm": 0.36718428134918213,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 1088
    },
    {
      "epoch": 0.1176915594942181,
      "grad_norm": 0.3542182743549347,
      "learning_rate": 8e-05,
      "loss": 1.7868,
      "step": 1089
    },
    {
      "epoch": 0.11779963255160489,
      "grad_norm": 0.42764219641685486,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 1090
    },
    {
      "epoch": 0.11790770560899168,
      "grad_norm": 0.33852437138557434,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 1091
    },
    {
      "epoch": 0.11801577866637847,
      "grad_norm": 0.37063103914260864,
      "learning_rate": 8e-05,
      "loss": 1.8017,
      "step": 1092
    },
    {
      "epoch": 0.11812385172376527,
      "grad_norm": 0.3522622287273407,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 1093
    },
    {
      "epoch": 0.11823192478115206,
      "grad_norm": 0.36144259572029114,
      "learning_rate": 8e-05,
      "loss": 1.859,
      "step": 1094
    },
    {
      "epoch": 0.11833999783853885,
      "grad_norm": 0.36767423152923584,
      "learning_rate": 8e-05,
      "loss": 1.8105,
      "step": 1095
    },
    {
      "epoch": 0.11844807089592564,
      "grad_norm": 0.36035487055778503,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 1096
    },
    {
      "epoch": 0.11855614395331245,
      "grad_norm": 0.38611260056495667,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 1097
    },
    {
      "epoch": 0.11866421701069924,
      "grad_norm": 0.336543470621109,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 1098
    },
    {
      "epoch": 0.11877229006808603,
      "grad_norm": 0.34846892952919006,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 1099
    },
    {
      "epoch": 0.11888036312547282,
      "grad_norm": 0.37330928444862366,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 1100
    },
    {
      "epoch": 0.11898843618285962,
      "grad_norm": 0.3681880533695221,
      "learning_rate": 8e-05,
      "loss": 1.7345,
      "step": 1101
    },
    {
      "epoch": 0.11909650924024641,
      "grad_norm": 0.3820163905620575,
      "learning_rate": 8e-05,
      "loss": 1.9009,
      "step": 1102
    },
    {
      "epoch": 0.1192045822976332,
      "grad_norm": 0.3882463276386261,
      "learning_rate": 8e-05,
      "loss": 1.8318,
      "step": 1103
    },
    {
      "epoch": 0.11931265535501999,
      "grad_norm": 0.398868203163147,
      "learning_rate": 8e-05,
      "loss": 1.8965,
      "step": 1104
    },
    {
      "epoch": 0.11942072841240679,
      "grad_norm": 0.3507888913154602,
      "learning_rate": 8e-05,
      "loss": 1.5984,
      "step": 1105
    },
    {
      "epoch": 0.11952880146979358,
      "grad_norm": 0.36435365676879883,
      "learning_rate": 8e-05,
      "loss": 1.8051,
      "step": 1106
    },
    {
      "epoch": 0.11963687452718037,
      "grad_norm": 0.3470079004764557,
      "learning_rate": 8e-05,
      "loss": 1.8356,
      "step": 1107
    },
    {
      "epoch": 0.11974494758456716,
      "grad_norm": 0.441213458776474,
      "learning_rate": 8e-05,
      "loss": 1.7817,
      "step": 1108
    },
    {
      "epoch": 0.11985302064195397,
      "grad_norm": 0.35217058658599854,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 1109
    },
    {
      "epoch": 0.11996109369934076,
      "grad_norm": 0.3345927894115448,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 1110
    },
    {
      "epoch": 0.12006916675672755,
      "grad_norm": 0.37604644894599915,
      "learning_rate": 8e-05,
      "loss": 1.7291,
      "step": 1111
    },
    {
      "epoch": 0.12017723981411434,
      "grad_norm": 0.40729671716690063,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 1112
    },
    {
      "epoch": 0.12028531287150114,
      "grad_norm": 0.3934181332588196,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 1113
    },
    {
      "epoch": 0.12039338592888793,
      "grad_norm": 0.406771719455719,
      "learning_rate": 8e-05,
      "loss": 1.9328,
      "step": 1114
    },
    {
      "epoch": 0.12050145898627472,
      "grad_norm": 0.37399056553840637,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 1115
    },
    {
      "epoch": 0.12060953204366151,
      "grad_norm": 0.38063445687294006,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 1116
    },
    {
      "epoch": 0.12071760510104831,
      "grad_norm": 0.4235222339630127,
      "learning_rate": 8e-05,
      "loss": 1.749,
      "step": 1117
    },
    {
      "epoch": 0.1208256781584351,
      "grad_norm": 0.3890572786331177,
      "learning_rate": 8e-05,
      "loss": 1.9588,
      "step": 1118
    },
    {
      "epoch": 0.1209337512158219,
      "grad_norm": 0.3706381916999817,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 1119
    },
    {
      "epoch": 0.12104182427320868,
      "grad_norm": 0.374380499124527,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 1120
    },
    {
      "epoch": 0.12114989733059549,
      "grad_norm": 0.361548513174057,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 1121
    },
    {
      "epoch": 0.12125797038798228,
      "grad_norm": 0.36958298087120056,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 1122
    },
    {
      "epoch": 0.12136604344536907,
      "grad_norm": 0.3657321333885193,
      "learning_rate": 8e-05,
      "loss": 1.8152,
      "step": 1123
    },
    {
      "epoch": 0.12147411650275586,
      "grad_norm": 0.35615840554237366,
      "learning_rate": 8e-05,
      "loss": 1.6348,
      "step": 1124
    },
    {
      "epoch": 0.12158218956014266,
      "grad_norm": 0.3991694748401642,
      "learning_rate": 8e-05,
      "loss": 1.967,
      "step": 1125
    },
    {
      "epoch": 0.12169026261752945,
      "grad_norm": 0.3788558542728424,
      "learning_rate": 8e-05,
      "loss": 1.8675,
      "step": 1126
    },
    {
      "epoch": 0.12179833567491624,
      "grad_norm": 0.3795939087867737,
      "learning_rate": 8e-05,
      "loss": 1.9319,
      "step": 1127
    },
    {
      "epoch": 0.12190640873230303,
      "grad_norm": 0.36552003026008606,
      "learning_rate": 8e-05,
      "loss": 1.773,
      "step": 1128
    },
    {
      "epoch": 0.12201448178968984,
      "grad_norm": 0.3401627838611603,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 1129
    },
    {
      "epoch": 0.12212255484707663,
      "grad_norm": 0.3407789170742035,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 1130
    },
    {
      "epoch": 0.12223062790446342,
      "grad_norm": 0.35426127910614014,
      "learning_rate": 8e-05,
      "loss": 1.6499,
      "step": 1131
    },
    {
      "epoch": 0.1223387009618502,
      "grad_norm": 0.38160404562950134,
      "learning_rate": 8e-05,
      "loss": 1.9214,
      "step": 1132
    },
    {
      "epoch": 0.12244677401923701,
      "grad_norm": 0.35989806056022644,
      "learning_rate": 8e-05,
      "loss": 1.8379,
      "step": 1133
    },
    {
      "epoch": 0.1225548470766238,
      "grad_norm": 0.41352027654647827,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 1134
    },
    {
      "epoch": 0.12266292013401059,
      "grad_norm": 0.36961716413497925,
      "learning_rate": 8e-05,
      "loss": 1.9075,
      "step": 1135
    },
    {
      "epoch": 0.12277099319139738,
      "grad_norm": 0.3447885513305664,
      "learning_rate": 8e-05,
      "loss": 1.8646,
      "step": 1136
    },
    {
      "epoch": 0.12287906624878418,
      "grad_norm": 0.3339730203151703,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 1137
    },
    {
      "epoch": 0.12298713930617097,
      "grad_norm": 0.33837664127349854,
      "learning_rate": 8e-05,
      "loss": 1.6525,
      "step": 1138
    },
    {
      "epoch": 0.12309521236355776,
      "grad_norm": 0.40510615706443787,
      "learning_rate": 8e-05,
      "loss": 1.9152,
      "step": 1139
    },
    {
      "epoch": 0.12320328542094455,
      "grad_norm": 0.3553943932056427,
      "learning_rate": 8e-05,
      "loss": 1.8305,
      "step": 1140
    },
    {
      "epoch": 0.12331135847833136,
      "grad_norm": 0.3435276746749878,
      "learning_rate": 8e-05,
      "loss": 1.7215,
      "step": 1141
    },
    {
      "epoch": 0.12341943153571815,
      "grad_norm": 0.3574138879776001,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 1142
    },
    {
      "epoch": 0.12352750459310494,
      "grad_norm": 0.3462582528591156,
      "learning_rate": 8e-05,
      "loss": 1.7756,
      "step": 1143
    },
    {
      "epoch": 0.12363557765049173,
      "grad_norm": 0.3465636372566223,
      "learning_rate": 8e-05,
      "loss": 1.7971,
      "step": 1144
    },
    {
      "epoch": 0.12374365070787853,
      "grad_norm": 0.34524717926979065,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 1145
    },
    {
      "epoch": 0.12385172376526532,
      "grad_norm": 0.3610204756259918,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 1146
    },
    {
      "epoch": 0.12395979682265211,
      "grad_norm": 0.3694356083869934,
      "learning_rate": 8e-05,
      "loss": 1.8118,
      "step": 1147
    },
    {
      "epoch": 0.1240678698800389,
      "grad_norm": 0.3656586706638336,
      "learning_rate": 8e-05,
      "loss": 1.963,
      "step": 1148
    },
    {
      "epoch": 0.1241759429374257,
      "grad_norm": 0.40215447545051575,
      "learning_rate": 8e-05,
      "loss": 1.9987,
      "step": 1149
    },
    {
      "epoch": 0.1242840159948125,
      "grad_norm": 0.3576542139053345,
      "learning_rate": 8e-05,
      "loss": 1.867,
      "step": 1150
    },
    {
      "epoch": 0.12439208905219928,
      "grad_norm": 0.3616178333759308,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 1151
    },
    {
      "epoch": 0.12450016210958607,
      "grad_norm": 0.3752042353153229,
      "learning_rate": 8e-05,
      "loss": 2.0495,
      "step": 1152
    },
    {
      "epoch": 0.12460823516697288,
      "grad_norm": 0.3590239882469177,
      "learning_rate": 8e-05,
      "loss": 1.7342,
      "step": 1153
    },
    {
      "epoch": 0.12471630822435967,
      "grad_norm": 0.3245165944099426,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 1154
    },
    {
      "epoch": 0.12482438128174646,
      "grad_norm": 0.37018802762031555,
      "learning_rate": 8e-05,
      "loss": 1.7599,
      "step": 1155
    },
    {
      "epoch": 0.12493245433913325,
      "grad_norm": 0.33771440386772156,
      "learning_rate": 8e-05,
      "loss": 1.8027,
      "step": 1156
    },
    {
      "epoch": 0.12504052739652005,
      "grad_norm": 0.34656190872192383,
      "learning_rate": 8e-05,
      "loss": 1.9092,
      "step": 1157
    },
    {
      "epoch": 0.12514860045390683,
      "grad_norm": 0.3510362505912781,
      "learning_rate": 8e-05,
      "loss": 1.9008,
      "step": 1158
    },
    {
      "epoch": 0.12525667351129363,
      "grad_norm": 0.3645678758621216,
      "learning_rate": 8e-05,
      "loss": 1.8581,
      "step": 1159
    },
    {
      "epoch": 0.12536474656868044,
      "grad_norm": 0.32818347215652466,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1160
    },
    {
      "epoch": 0.1254728196260672,
      "grad_norm": 0.3408606946468353,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 1161
    },
    {
      "epoch": 0.12558089268345402,
      "grad_norm": 0.3740525245666504,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 1162
    },
    {
      "epoch": 0.12568896574084082,
      "grad_norm": 0.35428285598754883,
      "learning_rate": 8e-05,
      "loss": 1.7405,
      "step": 1163
    },
    {
      "epoch": 0.1257970387982276,
      "grad_norm": 0.36056891083717346,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 1164
    },
    {
      "epoch": 0.1259051118556144,
      "grad_norm": 0.3450113832950592,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 1165
    },
    {
      "epoch": 0.12601318491300117,
      "grad_norm": 0.3771056532859802,
      "learning_rate": 8e-05,
      "loss": 1.8111,
      "step": 1166
    },
    {
      "epoch": 0.12612125797038798,
      "grad_norm": 0.34107789397239685,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 1167
    },
    {
      "epoch": 0.12622933102777478,
      "grad_norm": 0.3527986705303192,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 1168
    },
    {
      "epoch": 0.12633740408516156,
      "grad_norm": 0.3723733127117157,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 1169
    },
    {
      "epoch": 0.12644547714254836,
      "grad_norm": 0.36626508831977844,
      "learning_rate": 8e-05,
      "loss": 1.7188,
      "step": 1170
    },
    {
      "epoch": 0.12655355019993517,
      "grad_norm": 0.3760107755661011,
      "learning_rate": 8e-05,
      "loss": 1.9295,
      "step": 1171
    },
    {
      "epoch": 0.12666162325732194,
      "grad_norm": 0.32605016231536865,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 1172
    },
    {
      "epoch": 0.12676969631470875,
      "grad_norm": 0.389241099357605,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 1173
    },
    {
      "epoch": 0.12687776937209555,
      "grad_norm": 0.3753553330898285,
      "learning_rate": 8e-05,
      "loss": 1.7441,
      "step": 1174
    },
    {
      "epoch": 0.12698584242948233,
      "grad_norm": 0.36482882499694824,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 1175
    },
    {
      "epoch": 0.12709391548686913,
      "grad_norm": 0.3554099500179291,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 1176
    },
    {
      "epoch": 0.1272019885442559,
      "grad_norm": 0.33709585666656494,
      "learning_rate": 8e-05,
      "loss": 1.6052,
      "step": 1177
    },
    {
      "epoch": 0.1273100616016427,
      "grad_norm": 0.3496922552585602,
      "learning_rate": 8e-05,
      "loss": 1.586,
      "step": 1178
    },
    {
      "epoch": 0.1274181346590295,
      "grad_norm": 0.3904522657394409,
      "learning_rate": 8e-05,
      "loss": 1.9564,
      "step": 1179
    },
    {
      "epoch": 0.1275262077164163,
      "grad_norm": 0.36048629879951477,
      "learning_rate": 8e-05,
      "loss": 1.9061,
      "step": 1180
    },
    {
      "epoch": 0.1276342807738031,
      "grad_norm": 0.3584705889225006,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 1181
    },
    {
      "epoch": 0.1277423538311899,
      "grad_norm": 0.34056371450424194,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 1182
    },
    {
      "epoch": 0.12785042688857667,
      "grad_norm": 0.37036845088005066,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 1183
    },
    {
      "epoch": 0.12795849994596348,
      "grad_norm": 0.3162347972393036,
      "learning_rate": 8e-05,
      "loss": 1.5075,
      "step": 1184
    },
    {
      "epoch": 0.12806657300335025,
      "grad_norm": 0.3467709422111511,
      "learning_rate": 8e-05,
      "loss": 1.7704,
      "step": 1185
    },
    {
      "epoch": 0.12817464606073706,
      "grad_norm": 0.3360818028450012,
      "learning_rate": 8e-05,
      "loss": 1.7762,
      "step": 1186
    },
    {
      "epoch": 0.12828271911812386,
      "grad_norm": 0.35458385944366455,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 1187
    },
    {
      "epoch": 0.12839079217551064,
      "grad_norm": 0.3742386996746063,
      "learning_rate": 8e-05,
      "loss": 1.7366,
      "step": 1188
    },
    {
      "epoch": 0.12849886523289744,
      "grad_norm": 0.3503303825855255,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 1189
    },
    {
      "epoch": 0.12860693829028425,
      "grad_norm": 0.3646455407142639,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 1190
    },
    {
      "epoch": 0.12871501134767102,
      "grad_norm": 0.37325340509414673,
      "learning_rate": 8e-05,
      "loss": 1.9369,
      "step": 1191
    },
    {
      "epoch": 0.12882308440505782,
      "grad_norm": 0.3590868413448334,
      "learning_rate": 8e-05,
      "loss": 1.7607,
      "step": 1192
    },
    {
      "epoch": 0.1289311574624446,
      "grad_norm": 0.3444075286388397,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1193
    },
    {
      "epoch": 0.1290392305198314,
      "grad_norm": 0.34096190333366394,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 1194
    },
    {
      "epoch": 0.1291473035772182,
      "grad_norm": 0.35368281602859497,
      "learning_rate": 8e-05,
      "loss": 1.8032,
      "step": 1195
    },
    {
      "epoch": 0.12925537663460498,
      "grad_norm": 0.37291258573532104,
      "learning_rate": 8e-05,
      "loss": 1.8411,
      "step": 1196
    },
    {
      "epoch": 0.1293634496919918,
      "grad_norm": 0.3850989043712616,
      "learning_rate": 8e-05,
      "loss": 1.8738,
      "step": 1197
    },
    {
      "epoch": 0.1294715227493786,
      "grad_norm": 0.32650718092918396,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 1198
    },
    {
      "epoch": 0.12957959580676537,
      "grad_norm": 0.40874555706977844,
      "learning_rate": 8e-05,
      "loss": 1.9786,
      "step": 1199
    },
    {
      "epoch": 0.12968766886415217,
      "grad_norm": 0.3875872492790222,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 1200
    },
    {
      "epoch": 0.12979574192153895,
      "grad_norm": 0.37259984016418457,
      "learning_rate": 8e-05,
      "loss": 1.9103,
      "step": 1201
    },
    {
      "epoch": 0.12990381497892575,
      "grad_norm": 0.36838170886039734,
      "learning_rate": 8e-05,
      "loss": 1.8204,
      "step": 1202
    },
    {
      "epoch": 0.13001188803631256,
      "grad_norm": 0.37012097239494324,
      "learning_rate": 8e-05,
      "loss": 1.8511,
      "step": 1203
    },
    {
      "epoch": 0.13011996109369933,
      "grad_norm": 0.3868945837020874,
      "learning_rate": 8e-05,
      "loss": 1.875,
      "step": 1204
    },
    {
      "epoch": 0.13022803415108614,
      "grad_norm": 0.3822103440761566,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 1205
    },
    {
      "epoch": 0.13033610720847294,
      "grad_norm": 0.39760133624076843,
      "learning_rate": 8e-05,
      "loss": 1.8603,
      "step": 1206
    },
    {
      "epoch": 0.13044418026585972,
      "grad_norm": 0.3762820065021515,
      "learning_rate": 8e-05,
      "loss": 1.6615,
      "step": 1207
    },
    {
      "epoch": 0.13055225332324652,
      "grad_norm": 0.3529280424118042,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 1208
    },
    {
      "epoch": 0.1306603263806333,
      "grad_norm": 0.3459187150001526,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 1209
    },
    {
      "epoch": 0.1307683994380201,
      "grad_norm": 0.37171846628189087,
      "learning_rate": 8e-05,
      "loss": 1.7559,
      "step": 1210
    },
    {
      "epoch": 0.1308764724954069,
      "grad_norm": 0.32489991188049316,
      "learning_rate": 8e-05,
      "loss": 1.7134,
      "step": 1211
    },
    {
      "epoch": 0.13098454555279368,
      "grad_norm": 0.341099351644516,
      "learning_rate": 8e-05,
      "loss": 1.6784,
      "step": 1212
    },
    {
      "epoch": 0.13109261861018048,
      "grad_norm": 0.35070812702178955,
      "learning_rate": 8e-05,
      "loss": 1.8147,
      "step": 1213
    },
    {
      "epoch": 0.1312006916675673,
      "grad_norm": 0.3758661150932312,
      "learning_rate": 8e-05,
      "loss": 1.8404,
      "step": 1214
    },
    {
      "epoch": 0.13130876472495406,
      "grad_norm": 0.3853369653224945,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 1215
    },
    {
      "epoch": 0.13141683778234087,
      "grad_norm": 0.48795756697654724,
      "learning_rate": 8e-05,
      "loss": 2.0874,
      "step": 1216
    },
    {
      "epoch": 0.13152491083972764,
      "grad_norm": 0.3734152615070343,
      "learning_rate": 8e-05,
      "loss": 1.8706,
      "step": 1217
    },
    {
      "epoch": 0.13163298389711445,
      "grad_norm": 0.39112815260887146,
      "learning_rate": 8e-05,
      "loss": 1.7548,
      "step": 1218
    },
    {
      "epoch": 0.13174105695450125,
      "grad_norm": 0.3530450165271759,
      "learning_rate": 8e-05,
      "loss": 1.7709,
      "step": 1219
    },
    {
      "epoch": 0.13184913001188803,
      "grad_norm": 0.3932957947254181,
      "learning_rate": 8e-05,
      "loss": 2.0326,
      "step": 1220
    },
    {
      "epoch": 0.13195720306927483,
      "grad_norm": 0.34830108284950256,
      "learning_rate": 8e-05,
      "loss": 1.8511,
      "step": 1221
    },
    {
      "epoch": 0.13206527612666163,
      "grad_norm": 0.42291244864463806,
      "learning_rate": 8e-05,
      "loss": 1.9159,
      "step": 1222
    },
    {
      "epoch": 0.1321733491840484,
      "grad_norm": 0.35291528701782227,
      "learning_rate": 8e-05,
      "loss": 1.8812,
      "step": 1223
    },
    {
      "epoch": 0.13228142224143521,
      "grad_norm": 0.38225409388542175,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 1224
    },
    {
      "epoch": 0.132389495298822,
      "grad_norm": 0.3345004916191101,
      "learning_rate": 8e-05,
      "loss": 1.6906,
      "step": 1225
    },
    {
      "epoch": 0.1324975683562088,
      "grad_norm": 0.3450838327407837,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 1226
    },
    {
      "epoch": 0.1326056414135956,
      "grad_norm": 0.3691742420196533,
      "learning_rate": 8e-05,
      "loss": 1.8745,
      "step": 1227
    },
    {
      "epoch": 0.13271371447098237,
      "grad_norm": 0.3605695366859436,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 1228
    },
    {
      "epoch": 0.13282178752836918,
      "grad_norm": 0.3561975061893463,
      "learning_rate": 8e-05,
      "loss": 1.7542,
      "step": 1229
    },
    {
      "epoch": 0.13292986058575598,
      "grad_norm": 0.3594520688056946,
      "learning_rate": 8e-05,
      "loss": 1.5149,
      "step": 1230
    },
    {
      "epoch": 0.13303793364314276,
      "grad_norm": 0.3900435268878937,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 1231
    },
    {
      "epoch": 0.13314600670052956,
      "grad_norm": 0.3477350175380707,
      "learning_rate": 8e-05,
      "loss": 1.7469,
      "step": 1232
    },
    {
      "epoch": 0.13325407975791634,
      "grad_norm": 0.36821088194847107,
      "learning_rate": 8e-05,
      "loss": 1.8097,
      "step": 1233
    },
    {
      "epoch": 0.13336215281530314,
      "grad_norm": 0.3814113140106201,
      "learning_rate": 8e-05,
      "loss": 1.9105,
      "step": 1234
    },
    {
      "epoch": 0.13347022587268995,
      "grad_norm": 0.3567351996898651,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 1235
    },
    {
      "epoch": 0.13357829893007672,
      "grad_norm": 0.38173460960388184,
      "learning_rate": 8e-05,
      "loss": 1.9196,
      "step": 1236
    },
    {
      "epoch": 0.13368637198746353,
      "grad_norm": 0.3624939024448395,
      "learning_rate": 8e-05,
      "loss": 1.8586,
      "step": 1237
    },
    {
      "epoch": 0.13379444504485033,
      "grad_norm": 0.34047120809555054,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1238
    },
    {
      "epoch": 0.1339025181022371,
      "grad_norm": 0.3519713282585144,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 1239
    },
    {
      "epoch": 0.1340105911596239,
      "grad_norm": 0.37031808495521545,
      "learning_rate": 8e-05,
      "loss": 1.9764,
      "step": 1240
    },
    {
      "epoch": 0.13411866421701069,
      "grad_norm": 0.3479854464530945,
      "learning_rate": 8e-05,
      "loss": 1.5646,
      "step": 1241
    },
    {
      "epoch": 0.1342267372743975,
      "grad_norm": 0.3301011025905609,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 1242
    },
    {
      "epoch": 0.1343348103317843,
      "grad_norm": 0.359833300113678,
      "learning_rate": 8e-05,
      "loss": 1.8077,
      "step": 1243
    },
    {
      "epoch": 0.13444288338917107,
      "grad_norm": 0.3639060854911804,
      "learning_rate": 8e-05,
      "loss": 1.7188,
      "step": 1244
    },
    {
      "epoch": 0.13455095644655787,
      "grad_norm": 0.3266666531562805,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1245
    },
    {
      "epoch": 0.13465902950394468,
      "grad_norm": 0.37093308568000793,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 1246
    },
    {
      "epoch": 0.13476710256133145,
      "grad_norm": 0.3673769235610962,
      "learning_rate": 8e-05,
      "loss": 1.7933,
      "step": 1247
    },
    {
      "epoch": 0.13487517561871826,
      "grad_norm": 0.35311612486839294,
      "learning_rate": 8e-05,
      "loss": 1.8474,
      "step": 1248
    },
    {
      "epoch": 0.13498324867610503,
      "grad_norm": 0.3570164144039154,
      "learning_rate": 8e-05,
      "loss": 1.8897,
      "step": 1249
    },
    {
      "epoch": 0.13509132173349184,
      "grad_norm": 0.3754791021347046,
      "learning_rate": 8e-05,
      "loss": 1.6946,
      "step": 1250
    },
    {
      "epoch": 0.13519939479087864,
      "grad_norm": 0.3997972011566162,
      "learning_rate": 8e-05,
      "loss": 1.8581,
      "step": 1251
    },
    {
      "epoch": 0.13530746784826542,
      "grad_norm": 0.39735230803489685,
      "learning_rate": 8e-05,
      "loss": 2.0488,
      "step": 1252
    },
    {
      "epoch": 0.13541554090565222,
      "grad_norm": 0.3601343035697937,
      "learning_rate": 8e-05,
      "loss": 1.5946,
      "step": 1253
    },
    {
      "epoch": 0.13552361396303902,
      "grad_norm": 0.36464381217956543,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 1254
    },
    {
      "epoch": 0.1356316870204258,
      "grad_norm": 0.3686005175113678,
      "learning_rate": 8e-05,
      "loss": 1.8295,
      "step": 1255
    },
    {
      "epoch": 0.1357397600778126,
      "grad_norm": 0.41247066855430603,
      "learning_rate": 8e-05,
      "loss": 1.9405,
      "step": 1256
    },
    {
      "epoch": 0.1358478331351994,
      "grad_norm": 0.3546912968158722,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 1257
    },
    {
      "epoch": 0.13595590619258618,
      "grad_norm": 0.396194189786911,
      "learning_rate": 8e-05,
      "loss": 1.8007,
      "step": 1258
    },
    {
      "epoch": 0.136063979249973,
      "grad_norm": 0.3414531350135803,
      "learning_rate": 8e-05,
      "loss": 1.7376,
      "step": 1259
    },
    {
      "epoch": 0.13617205230735976,
      "grad_norm": 0.365448534488678,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 1260
    },
    {
      "epoch": 0.13628012536474657,
      "grad_norm": 0.3723194897174835,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 1261
    },
    {
      "epoch": 0.13638819842213337,
      "grad_norm": 0.34464162588119507,
      "learning_rate": 8e-05,
      "loss": 1.7818,
      "step": 1262
    },
    {
      "epoch": 0.13649627147952015,
      "grad_norm": 0.379607230424881,
      "learning_rate": 8e-05,
      "loss": 2.0066,
      "step": 1263
    },
    {
      "epoch": 0.13660434453690695,
      "grad_norm": 0.35011014342308044,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 1264
    },
    {
      "epoch": 0.13671241759429376,
      "grad_norm": 0.40158697962760925,
      "learning_rate": 8e-05,
      "loss": 1.9133,
      "step": 1265
    },
    {
      "epoch": 0.13682049065168053,
      "grad_norm": 0.3566371500492096,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 1266
    },
    {
      "epoch": 0.13692856370906734,
      "grad_norm": 0.3662523925304413,
      "learning_rate": 8e-05,
      "loss": 1.8725,
      "step": 1267
    },
    {
      "epoch": 0.1370366367664541,
      "grad_norm": 0.36400070786476135,
      "learning_rate": 8e-05,
      "loss": 1.8992,
      "step": 1268
    },
    {
      "epoch": 0.13714470982384092,
      "grad_norm": 0.3628464341163635,
      "learning_rate": 8e-05,
      "loss": 1.8345,
      "step": 1269
    },
    {
      "epoch": 0.13725278288122772,
      "grad_norm": 0.3793184161186218,
      "learning_rate": 8e-05,
      "loss": 1.8457,
      "step": 1270
    },
    {
      "epoch": 0.1373608559386145,
      "grad_norm": 0.34508460760116577,
      "learning_rate": 8e-05,
      "loss": 1.6921,
      "step": 1271
    },
    {
      "epoch": 0.1374689289960013,
      "grad_norm": 0.37388306856155396,
      "learning_rate": 8e-05,
      "loss": 1.9294,
      "step": 1272
    },
    {
      "epoch": 0.1375770020533881,
      "grad_norm": 0.37191563844680786,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 1273
    },
    {
      "epoch": 0.13768507511077488,
      "grad_norm": 0.38890841603279114,
      "learning_rate": 8e-05,
      "loss": 1.9279,
      "step": 1274
    },
    {
      "epoch": 0.13779314816816168,
      "grad_norm": 0.35622289776802063,
      "learning_rate": 8e-05,
      "loss": 1.7508,
      "step": 1275
    },
    {
      "epoch": 0.13790122122554846,
      "grad_norm": 0.3675192594528198,
      "learning_rate": 8e-05,
      "loss": 1.6861,
      "step": 1276
    },
    {
      "epoch": 0.13800929428293526,
      "grad_norm": 0.35501307249069214,
      "learning_rate": 8e-05,
      "loss": 1.8708,
      "step": 1277
    },
    {
      "epoch": 0.13811736734032207,
      "grad_norm": 0.33192023634910583,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 1278
    },
    {
      "epoch": 0.13822544039770884,
      "grad_norm": 0.3726126253604889,
      "learning_rate": 8e-05,
      "loss": 1.946,
      "step": 1279
    },
    {
      "epoch": 0.13833351345509565,
      "grad_norm": 0.34756410121917725,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 1280
    },
    {
      "epoch": 0.13844158651248245,
      "grad_norm": 0.3411407768726349,
      "learning_rate": 8e-05,
      "loss": 1.6545,
      "step": 1281
    },
    {
      "epoch": 0.13854965956986923,
      "grad_norm": 0.34891775250434875,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 1282
    },
    {
      "epoch": 0.13865773262725603,
      "grad_norm": 0.36273065209388733,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 1283
    },
    {
      "epoch": 0.1387658056846428,
      "grad_norm": 0.34276318550109863,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 1284
    },
    {
      "epoch": 0.1388738787420296,
      "grad_norm": 0.3624266982078552,
      "learning_rate": 8e-05,
      "loss": 1.8052,
      "step": 1285
    },
    {
      "epoch": 0.13898195179941641,
      "grad_norm": 0.3275555968284607,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 1286
    },
    {
      "epoch": 0.1390900248568032,
      "grad_norm": 0.34999197721481323,
      "learning_rate": 8e-05,
      "loss": 1.7905,
      "step": 1287
    },
    {
      "epoch": 0.13919809791419,
      "grad_norm": 0.3496024012565613,
      "learning_rate": 8e-05,
      "loss": 1.7994,
      "step": 1288
    },
    {
      "epoch": 0.1393061709715768,
      "grad_norm": 0.361912339925766,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 1289
    },
    {
      "epoch": 0.13941424402896357,
      "grad_norm": 0.3756738603115082,
      "learning_rate": 8e-05,
      "loss": 1.9363,
      "step": 1290
    },
    {
      "epoch": 0.13952231708635038,
      "grad_norm": 0.3466061055660248,
      "learning_rate": 8e-05,
      "loss": 1.5753,
      "step": 1291
    },
    {
      "epoch": 0.13963039014373715,
      "grad_norm": 0.36808082461357117,
      "learning_rate": 8e-05,
      "loss": 1.8752,
      "step": 1292
    },
    {
      "epoch": 0.13973846320112396,
      "grad_norm": 0.3684549927711487,
      "learning_rate": 8e-05,
      "loss": 1.7835,
      "step": 1293
    },
    {
      "epoch": 0.13984653625851076,
      "grad_norm": 0.36353185772895813,
      "learning_rate": 8e-05,
      "loss": 1.8672,
      "step": 1294
    },
    {
      "epoch": 0.13995460931589754,
      "grad_norm": 0.3836422860622406,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 1295
    },
    {
      "epoch": 0.14006268237328434,
      "grad_norm": 0.3743646740913391,
      "learning_rate": 8e-05,
      "loss": 1.9166,
      "step": 1296
    },
    {
      "epoch": 0.14017075543067115,
      "grad_norm": 0.3598948121070862,
      "learning_rate": 8e-05,
      "loss": 1.8189,
      "step": 1297
    },
    {
      "epoch": 0.14027882848805792,
      "grad_norm": 0.38350221514701843,
      "learning_rate": 8e-05,
      "loss": 1.8866,
      "step": 1298
    },
    {
      "epoch": 0.14038690154544473,
      "grad_norm": 0.36876654624938965,
      "learning_rate": 8e-05,
      "loss": 1.8262,
      "step": 1299
    },
    {
      "epoch": 0.1404949746028315,
      "grad_norm": 0.38774171471595764,
      "learning_rate": 8e-05,
      "loss": 1.8222,
      "step": 1300
    },
    {
      "epoch": 0.1406030476602183,
      "grad_norm": 0.37729960680007935,
      "learning_rate": 8e-05,
      "loss": 1.673,
      "step": 1301
    },
    {
      "epoch": 0.1407111207176051,
      "grad_norm": 0.33797571063041687,
      "learning_rate": 8e-05,
      "loss": 1.6448,
      "step": 1302
    },
    {
      "epoch": 0.14081919377499189,
      "grad_norm": 0.34323862195014954,
      "learning_rate": 8e-05,
      "loss": 1.7167,
      "step": 1303
    },
    {
      "epoch": 0.1409272668323787,
      "grad_norm": 0.3635108470916748,
      "learning_rate": 8e-05,
      "loss": 1.9497,
      "step": 1304
    },
    {
      "epoch": 0.1410353398897655,
      "grad_norm": 0.38030102849006653,
      "learning_rate": 8e-05,
      "loss": 1.7682,
      "step": 1305
    },
    {
      "epoch": 0.14114341294715227,
      "grad_norm": 0.3572438955307007,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 1306
    },
    {
      "epoch": 0.14125148600453907,
      "grad_norm": 0.3755241632461548,
      "learning_rate": 8e-05,
      "loss": 1.8321,
      "step": 1307
    },
    {
      "epoch": 0.14135955906192585,
      "grad_norm": 0.36530229449272156,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 1308
    },
    {
      "epoch": 0.14146763211931265,
      "grad_norm": 0.38700082898139954,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 1309
    },
    {
      "epoch": 0.14157570517669946,
      "grad_norm": 0.358844518661499,
      "learning_rate": 8e-05,
      "loss": 1.6388,
      "step": 1310
    },
    {
      "epoch": 0.14168377823408623,
      "grad_norm": 0.3239356577396393,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 1311
    },
    {
      "epoch": 0.14179185129147304,
      "grad_norm": 0.37981265783309937,
      "learning_rate": 8e-05,
      "loss": 1.9305,
      "step": 1312
    },
    {
      "epoch": 0.14189992434885984,
      "grad_norm": 0.355438232421875,
      "learning_rate": 8e-05,
      "loss": 1.7814,
      "step": 1313
    },
    {
      "epoch": 0.14200799740624662,
      "grad_norm": 0.38024845719337463,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 1314
    },
    {
      "epoch": 0.14211607046363342,
      "grad_norm": 0.33766257762908936,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 1315
    },
    {
      "epoch": 0.1422241435210202,
      "grad_norm": 0.35067158937454224,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 1316
    },
    {
      "epoch": 0.142332216578407,
      "grad_norm": 0.36893215775489807,
      "learning_rate": 8e-05,
      "loss": 1.8138,
      "step": 1317
    },
    {
      "epoch": 0.1424402896357938,
      "grad_norm": 0.402853399515152,
      "learning_rate": 8e-05,
      "loss": 1.9631,
      "step": 1318
    },
    {
      "epoch": 0.14254836269318058,
      "grad_norm": 0.3573390245437622,
      "learning_rate": 8e-05,
      "loss": 1.5288,
      "step": 1319
    },
    {
      "epoch": 0.14265643575056738,
      "grad_norm": 0.36313819885253906,
      "learning_rate": 8e-05,
      "loss": 1.7057,
      "step": 1320
    },
    {
      "epoch": 0.1427645088079542,
      "grad_norm": 0.4008879065513611,
      "learning_rate": 8e-05,
      "loss": 1.8633,
      "step": 1321
    },
    {
      "epoch": 0.14287258186534096,
      "grad_norm": 0.38593950867652893,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 1322
    },
    {
      "epoch": 0.14298065492272777,
      "grad_norm": 0.3522358536720276,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 1323
    },
    {
      "epoch": 0.14308872798011454,
      "grad_norm": 0.34919023513793945,
      "learning_rate": 8e-05,
      "loss": 1.7858,
      "step": 1324
    },
    {
      "epoch": 0.14319680103750135,
      "grad_norm": 0.4303407073020935,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 1325
    },
    {
      "epoch": 0.14330487409488815,
      "grad_norm": 0.34750887751579285,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1326
    },
    {
      "epoch": 0.14341294715227493,
      "grad_norm": 0.34084826707839966,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 1327
    },
    {
      "epoch": 0.14352102020966173,
      "grad_norm": 0.3843788504600525,
      "learning_rate": 8e-05,
      "loss": 1.8638,
      "step": 1328
    },
    {
      "epoch": 0.14362909326704854,
      "grad_norm": 0.36963194608688354,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 1329
    },
    {
      "epoch": 0.1437371663244353,
      "grad_norm": 0.3810136318206787,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 1330
    },
    {
      "epoch": 0.14384523938182212,
      "grad_norm": 0.36227381229400635,
      "learning_rate": 8e-05,
      "loss": 1.8458,
      "step": 1331
    },
    {
      "epoch": 0.1439533124392089,
      "grad_norm": 0.3583066165447235,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 1332
    },
    {
      "epoch": 0.1440613854965957,
      "grad_norm": 0.3827393352985382,
      "learning_rate": 8e-05,
      "loss": 1.8215,
      "step": 1333
    },
    {
      "epoch": 0.1441694585539825,
      "grad_norm": 0.38677769899368286,
      "learning_rate": 8e-05,
      "loss": 1.8207,
      "step": 1334
    },
    {
      "epoch": 0.14427753161136928,
      "grad_norm": 0.35298940539360046,
      "learning_rate": 8e-05,
      "loss": 1.7273,
      "step": 1335
    },
    {
      "epoch": 0.14438560466875608,
      "grad_norm": 0.353473961353302,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 1336
    },
    {
      "epoch": 0.14449367772614288,
      "grad_norm": 0.3717522621154785,
      "learning_rate": 8e-05,
      "loss": 1.9448,
      "step": 1337
    },
    {
      "epoch": 0.14460175078352966,
      "grad_norm": 0.36258256435394287,
      "learning_rate": 8e-05,
      "loss": 1.794,
      "step": 1338
    },
    {
      "epoch": 0.14470982384091646,
      "grad_norm": 0.3871307075023651,
      "learning_rate": 8e-05,
      "loss": 1.8851,
      "step": 1339
    },
    {
      "epoch": 0.14481789689830327,
      "grad_norm": 0.3597882091999054,
      "learning_rate": 8e-05,
      "loss": 1.7949,
      "step": 1340
    },
    {
      "epoch": 0.14492596995569004,
      "grad_norm": 0.3686104416847229,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 1341
    },
    {
      "epoch": 0.14503404301307685,
      "grad_norm": 0.38345858454704285,
      "learning_rate": 8e-05,
      "loss": 1.9195,
      "step": 1342
    },
    {
      "epoch": 0.14514211607046362,
      "grad_norm": 0.3607224225997925,
      "learning_rate": 8e-05,
      "loss": 1.9278,
      "step": 1343
    },
    {
      "epoch": 0.14525018912785043,
      "grad_norm": 0.33355677127838135,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 1344
    },
    {
      "epoch": 0.14535826218523723,
      "grad_norm": 0.35089266300201416,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 1345
    },
    {
      "epoch": 0.145466335242624,
      "grad_norm": 0.35067078471183777,
      "learning_rate": 8e-05,
      "loss": 1.6902,
      "step": 1346
    },
    {
      "epoch": 0.1455744083000108,
      "grad_norm": 0.35832321643829346,
      "learning_rate": 8e-05,
      "loss": 1.7072,
      "step": 1347
    },
    {
      "epoch": 0.14568248135739761,
      "grad_norm": 0.36257049441337585,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 1348
    },
    {
      "epoch": 0.1457905544147844,
      "grad_norm": 0.36324799060821533,
      "learning_rate": 8e-05,
      "loss": 1.792,
      "step": 1349
    },
    {
      "epoch": 0.1458986274721712,
      "grad_norm": 0.37073221802711487,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 1350
    },
    {
      "epoch": 0.14600670052955797,
      "grad_norm": 0.39439767599105835,
      "learning_rate": 8e-05,
      "loss": 1.638,
      "step": 1351
    },
    {
      "epoch": 0.14611477358694477,
      "grad_norm": 0.428548663854599,
      "learning_rate": 8e-05,
      "loss": 1.5965,
      "step": 1352
    },
    {
      "epoch": 0.14622284664433158,
      "grad_norm": 0.41466617584228516,
      "learning_rate": 8e-05,
      "loss": 1.8584,
      "step": 1353
    },
    {
      "epoch": 0.14633091970171835,
      "grad_norm": 0.3542669415473938,
      "learning_rate": 8e-05,
      "loss": 1.7357,
      "step": 1354
    },
    {
      "epoch": 0.14643899275910516,
      "grad_norm": 0.34826070070266724,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 1355
    },
    {
      "epoch": 0.14654706581649196,
      "grad_norm": 0.3319184482097626,
      "learning_rate": 8e-05,
      "loss": 1.5644,
      "step": 1356
    },
    {
      "epoch": 0.14665513887387874,
      "grad_norm": 0.37194564938545227,
      "learning_rate": 8e-05,
      "loss": 1.854,
      "step": 1357
    },
    {
      "epoch": 0.14676321193126554,
      "grad_norm": 0.37797489762306213,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 1358
    },
    {
      "epoch": 0.14687128498865232,
      "grad_norm": 0.3618971109390259,
      "learning_rate": 8e-05,
      "loss": 1.8924,
      "step": 1359
    },
    {
      "epoch": 0.14697935804603912,
      "grad_norm": 0.38728848099708557,
      "learning_rate": 8e-05,
      "loss": 1.8663,
      "step": 1360
    },
    {
      "epoch": 0.14708743110342593,
      "grad_norm": 0.36954250931739807,
      "learning_rate": 8e-05,
      "loss": 1.836,
      "step": 1361
    },
    {
      "epoch": 0.1471955041608127,
      "grad_norm": 0.3604555130004883,
      "learning_rate": 8e-05,
      "loss": 1.7374,
      "step": 1362
    },
    {
      "epoch": 0.1473035772181995,
      "grad_norm": 0.3396654725074768,
      "learning_rate": 8e-05,
      "loss": 1.6719,
      "step": 1363
    },
    {
      "epoch": 0.1474116502755863,
      "grad_norm": 0.3587683141231537,
      "learning_rate": 8e-05,
      "loss": 1.7432,
      "step": 1364
    },
    {
      "epoch": 0.14751972333297309,
      "grad_norm": 0.36430948972702026,
      "learning_rate": 8e-05,
      "loss": 1.7602,
      "step": 1365
    },
    {
      "epoch": 0.1476277963903599,
      "grad_norm": 0.3922954201698303,
      "learning_rate": 8e-05,
      "loss": 1.8045,
      "step": 1366
    },
    {
      "epoch": 0.14773586944774666,
      "grad_norm": 0.3604581952095032,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 1367
    },
    {
      "epoch": 0.14784394250513347,
      "grad_norm": 0.35327261686325073,
      "learning_rate": 8e-05,
      "loss": 1.7987,
      "step": 1368
    },
    {
      "epoch": 0.14795201556252027,
      "grad_norm": 0.3642292022705078,
      "learning_rate": 8e-05,
      "loss": 1.8837,
      "step": 1369
    },
    {
      "epoch": 0.14806008861990705,
      "grad_norm": 0.3424132764339447,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 1370
    },
    {
      "epoch": 0.14816816167729385,
      "grad_norm": 0.36793118715286255,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 1371
    },
    {
      "epoch": 0.14827623473468066,
      "grad_norm": 0.37847766280174255,
      "learning_rate": 8e-05,
      "loss": 1.9174,
      "step": 1372
    },
    {
      "epoch": 0.14838430779206743,
      "grad_norm": 0.3926122188568115,
      "learning_rate": 8e-05,
      "loss": 2.0138,
      "step": 1373
    },
    {
      "epoch": 0.14849238084945424,
      "grad_norm": 0.3614753484725952,
      "learning_rate": 8e-05,
      "loss": 1.8223,
      "step": 1374
    },
    {
      "epoch": 0.148600453906841,
      "grad_norm": 0.3655136525630951,
      "learning_rate": 8e-05,
      "loss": 1.6286,
      "step": 1375
    },
    {
      "epoch": 0.14870852696422782,
      "grad_norm": 0.36882254481315613,
      "learning_rate": 8e-05,
      "loss": 1.7201,
      "step": 1376
    },
    {
      "epoch": 0.14881660002161462,
      "grad_norm": 0.3873593509197235,
      "learning_rate": 8e-05,
      "loss": 1.9422,
      "step": 1377
    },
    {
      "epoch": 0.1489246730790014,
      "grad_norm": 0.38361072540283203,
      "learning_rate": 8e-05,
      "loss": 1.7471,
      "step": 1378
    },
    {
      "epoch": 0.1490327461363882,
      "grad_norm": 0.34324052929878235,
      "learning_rate": 8e-05,
      "loss": 1.6718,
      "step": 1379
    },
    {
      "epoch": 0.149140819193775,
      "grad_norm": 0.3797532618045807,
      "learning_rate": 8e-05,
      "loss": 1.9269,
      "step": 1380
    },
    {
      "epoch": 0.14924889225116178,
      "grad_norm": 0.40686362981796265,
      "learning_rate": 8e-05,
      "loss": 1.9187,
      "step": 1381
    },
    {
      "epoch": 0.14935696530854858,
      "grad_norm": 0.35567137598991394,
      "learning_rate": 8e-05,
      "loss": 1.8577,
      "step": 1382
    },
    {
      "epoch": 0.14946503836593536,
      "grad_norm": 0.3699253499507904,
      "learning_rate": 8e-05,
      "loss": 1.8638,
      "step": 1383
    },
    {
      "epoch": 0.14957311142332216,
      "grad_norm": 0.3547039031982422,
      "learning_rate": 8e-05,
      "loss": 1.7259,
      "step": 1384
    },
    {
      "epoch": 0.14968118448070897,
      "grad_norm": 0.3718073070049286,
      "learning_rate": 8e-05,
      "loss": 1.7663,
      "step": 1385
    },
    {
      "epoch": 0.14978925753809574,
      "grad_norm": 0.37713155150413513,
      "learning_rate": 8e-05,
      "loss": 1.8828,
      "step": 1386
    },
    {
      "epoch": 0.14989733059548255,
      "grad_norm": 0.36620232462882996,
      "learning_rate": 8e-05,
      "loss": 1.7877,
      "step": 1387
    },
    {
      "epoch": 0.15000540365286935,
      "grad_norm": 0.34448128938674927,
      "learning_rate": 8e-05,
      "loss": 1.5905,
      "step": 1388
    },
    {
      "epoch": 0.15011347671025613,
      "grad_norm": 0.37907782196998596,
      "learning_rate": 8e-05,
      "loss": 1.5522,
      "step": 1389
    },
    {
      "epoch": 0.15022154976764293,
      "grad_norm": 0.3553794324398041,
      "learning_rate": 8e-05,
      "loss": 1.8251,
      "step": 1390
    },
    {
      "epoch": 0.1503296228250297,
      "grad_norm": 0.3589995503425598,
      "learning_rate": 8e-05,
      "loss": 1.807,
      "step": 1391
    },
    {
      "epoch": 0.1504376958824165,
      "grad_norm": 0.3628771901130676,
      "learning_rate": 8e-05,
      "loss": 1.7974,
      "step": 1392
    },
    {
      "epoch": 0.15054576893980331,
      "grad_norm": 0.36628058552742004,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 1393
    },
    {
      "epoch": 0.1506538419971901,
      "grad_norm": 0.35752347111701965,
      "learning_rate": 8e-05,
      "loss": 1.8529,
      "step": 1394
    },
    {
      "epoch": 0.1507619150545769,
      "grad_norm": 0.3432478606700897,
      "learning_rate": 8e-05,
      "loss": 1.6778,
      "step": 1395
    },
    {
      "epoch": 0.1508699881119637,
      "grad_norm": 0.33416950702667236,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 1396
    },
    {
      "epoch": 0.15097806116935047,
      "grad_norm": 0.3986573815345764,
      "learning_rate": 8e-05,
      "loss": 1.9902,
      "step": 1397
    },
    {
      "epoch": 0.15108613422673728,
      "grad_norm": 0.36368662118911743,
      "learning_rate": 8e-05,
      "loss": 1.8914,
      "step": 1398
    },
    {
      "epoch": 0.15119420728412405,
      "grad_norm": 0.39978787302970886,
      "learning_rate": 8e-05,
      "loss": 1.8257,
      "step": 1399
    },
    {
      "epoch": 0.15130228034151086,
      "grad_norm": 0.35373568534851074,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 1400
    },
    {
      "epoch": 0.15141035339889766,
      "grad_norm": 0.3432331085205078,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 1401
    },
    {
      "epoch": 0.15151842645628444,
      "grad_norm": 0.36555132269859314,
      "learning_rate": 8e-05,
      "loss": 1.8551,
      "step": 1402
    },
    {
      "epoch": 0.15162649951367124,
      "grad_norm": 0.3434751629829407,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 1403
    },
    {
      "epoch": 0.15173457257105805,
      "grad_norm": 0.3762097656726837,
      "learning_rate": 8e-05,
      "loss": 1.8822,
      "step": 1404
    },
    {
      "epoch": 0.15184264562844482,
      "grad_norm": 0.3662976026535034,
      "learning_rate": 8e-05,
      "loss": 1.7431,
      "step": 1405
    },
    {
      "epoch": 0.15195071868583163,
      "grad_norm": 0.3371051251888275,
      "learning_rate": 8e-05,
      "loss": 1.6064,
      "step": 1406
    },
    {
      "epoch": 0.1520587917432184,
      "grad_norm": 0.36757299304008484,
      "learning_rate": 8e-05,
      "loss": 1.7929,
      "step": 1407
    },
    {
      "epoch": 0.1521668648006052,
      "grad_norm": 0.3737618029117584,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 1408
    },
    {
      "epoch": 0.152274937857992,
      "grad_norm": 0.4045862853527069,
      "learning_rate": 8e-05,
      "loss": 1.9242,
      "step": 1409
    },
    {
      "epoch": 0.15238301091537879,
      "grad_norm": 0.4126172959804535,
      "learning_rate": 8e-05,
      "loss": 1.9788,
      "step": 1410
    },
    {
      "epoch": 0.1524910839727656,
      "grad_norm": 0.3461454510688782,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 1411
    },
    {
      "epoch": 0.1525991570301524,
      "grad_norm": 0.3536359965801239,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 1412
    },
    {
      "epoch": 0.15270723008753917,
      "grad_norm": 0.37129977345466614,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 1413
    },
    {
      "epoch": 0.15281530314492597,
      "grad_norm": 0.3773781359195709,
      "learning_rate": 8e-05,
      "loss": 1.9379,
      "step": 1414
    },
    {
      "epoch": 0.15292337620231275,
      "grad_norm": 0.3861599266529083,
      "learning_rate": 8e-05,
      "loss": 1.8157,
      "step": 1415
    },
    {
      "epoch": 0.15303144925969955,
      "grad_norm": 0.3463456332683563,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 1416
    },
    {
      "epoch": 0.15313952231708636,
      "grad_norm": 0.363972932100296,
      "learning_rate": 8e-05,
      "loss": 1.7568,
      "step": 1417
    },
    {
      "epoch": 0.15324759537447313,
      "grad_norm": 0.38853877782821655,
      "learning_rate": 8e-05,
      "loss": 1.9306,
      "step": 1418
    },
    {
      "epoch": 0.15335566843185994,
      "grad_norm": 0.3726254105567932,
      "learning_rate": 8e-05,
      "loss": 1.8972,
      "step": 1419
    },
    {
      "epoch": 0.15346374148924674,
      "grad_norm": 0.3519631326198578,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 1420
    },
    {
      "epoch": 0.15357181454663352,
      "grad_norm": 0.3772374093532562,
      "learning_rate": 8e-05,
      "loss": 1.7649,
      "step": 1421
    },
    {
      "epoch": 0.15367988760402032,
      "grad_norm": 0.34010839462280273,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 1422
    },
    {
      "epoch": 0.15378796066140712,
      "grad_norm": 0.3800028860569,
      "learning_rate": 8e-05,
      "loss": 1.7983,
      "step": 1423
    },
    {
      "epoch": 0.1538960337187939,
      "grad_norm": 0.34591197967529297,
      "learning_rate": 8e-05,
      "loss": 1.7964,
      "step": 1424
    },
    {
      "epoch": 0.1540041067761807,
      "grad_norm": 0.35065922141075134,
      "learning_rate": 8e-05,
      "loss": 1.5771,
      "step": 1425
    },
    {
      "epoch": 0.15411217983356748,
      "grad_norm": 0.3294548988342285,
      "learning_rate": 8e-05,
      "loss": 1.6164,
      "step": 1426
    },
    {
      "epoch": 0.15422025289095428,
      "grad_norm": 0.35767996311187744,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 1427
    },
    {
      "epoch": 0.1543283259483411,
      "grad_norm": 0.4050959348678589,
      "learning_rate": 8e-05,
      "loss": 1.9109,
      "step": 1428
    },
    {
      "epoch": 0.15443639900572786,
      "grad_norm": 0.38854584097862244,
      "learning_rate": 8e-05,
      "loss": 1.8469,
      "step": 1429
    },
    {
      "epoch": 0.15454447206311467,
      "grad_norm": 0.3645983338356018,
      "learning_rate": 8e-05,
      "loss": 1.7182,
      "step": 1430
    },
    {
      "epoch": 0.15465254512050147,
      "grad_norm": 0.3580503761768341,
      "learning_rate": 8e-05,
      "loss": 1.7341,
      "step": 1431
    },
    {
      "epoch": 0.15476061817788825,
      "grad_norm": 0.38257795572280884,
      "learning_rate": 8e-05,
      "loss": 1.5142,
      "step": 1432
    },
    {
      "epoch": 0.15486869123527505,
      "grad_norm": 0.3839014172554016,
      "learning_rate": 8e-05,
      "loss": 1.8363,
      "step": 1433
    },
    {
      "epoch": 0.15497676429266183,
      "grad_norm": 0.3898155987262726,
      "learning_rate": 8e-05,
      "loss": 1.8737,
      "step": 1434
    },
    {
      "epoch": 0.15508483735004863,
      "grad_norm": 0.3901798129081726,
      "learning_rate": 8e-05,
      "loss": 1.9292,
      "step": 1435
    },
    {
      "epoch": 0.15519291040743544,
      "grad_norm": 0.3728411793708801,
      "learning_rate": 8e-05,
      "loss": 1.9086,
      "step": 1436
    },
    {
      "epoch": 0.1553009834648222,
      "grad_norm": 0.32981374859809875,
      "learning_rate": 8e-05,
      "loss": 1.7063,
      "step": 1437
    },
    {
      "epoch": 0.15540905652220902,
      "grad_norm": 0.37500622868537903,
      "learning_rate": 8e-05,
      "loss": 2.0201,
      "step": 1438
    },
    {
      "epoch": 0.15551712957959582,
      "grad_norm": 0.36162200570106506,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 1439
    },
    {
      "epoch": 0.1556252026369826,
      "grad_norm": 0.35528993606567383,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 1440
    },
    {
      "epoch": 0.1557332756943694,
      "grad_norm": 0.3630034625530243,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 1441
    },
    {
      "epoch": 0.15584134875175618,
      "grad_norm": 0.3964095115661621,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 1442
    },
    {
      "epoch": 0.15594942180914298,
      "grad_norm": 0.3552529811859131,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 1443
    },
    {
      "epoch": 0.15605749486652978,
      "grad_norm": 0.3829663097858429,
      "learning_rate": 8e-05,
      "loss": 1.7102,
      "step": 1444
    },
    {
      "epoch": 0.15616556792391656,
      "grad_norm": 0.35981523990631104,
      "learning_rate": 8e-05,
      "loss": 1.7307,
      "step": 1445
    },
    {
      "epoch": 0.15627364098130336,
      "grad_norm": 0.3516583740711212,
      "learning_rate": 8e-05,
      "loss": 1.6101,
      "step": 1446
    },
    {
      "epoch": 0.15638171403869017,
      "grad_norm": 0.34881412982940674,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 1447
    },
    {
      "epoch": 0.15648978709607694,
      "grad_norm": 0.40145429968833923,
      "learning_rate": 8e-05,
      "loss": 1.7185,
      "step": 1448
    },
    {
      "epoch": 0.15659786015346375,
      "grad_norm": 0.35522493720054626,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 1449
    },
    {
      "epoch": 0.15670593321085052,
      "grad_norm": 0.3860670328140259,
      "learning_rate": 8e-05,
      "loss": 1.8801,
      "step": 1450
    },
    {
      "epoch": 0.15681400626823733,
      "grad_norm": 0.3494183421134949,
      "learning_rate": 8e-05,
      "loss": 1.7717,
      "step": 1451
    },
    {
      "epoch": 0.15692207932562413,
      "grad_norm": 0.3855258822441101,
      "learning_rate": 8e-05,
      "loss": 1.8964,
      "step": 1452
    },
    {
      "epoch": 0.1570301523830109,
      "grad_norm": 0.3451821804046631,
      "learning_rate": 8e-05,
      "loss": 1.6332,
      "step": 1453
    },
    {
      "epoch": 0.1571382254403977,
      "grad_norm": 0.34593403339385986,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 1454
    },
    {
      "epoch": 0.15724629849778451,
      "grad_norm": 0.36042019724845886,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 1455
    },
    {
      "epoch": 0.1573543715551713,
      "grad_norm": 0.3821708559989929,
      "learning_rate": 8e-05,
      "loss": 1.8209,
      "step": 1456
    },
    {
      "epoch": 0.1574624446125581,
      "grad_norm": 0.3659166693687439,
      "learning_rate": 8e-05,
      "loss": 1.9368,
      "step": 1457
    },
    {
      "epoch": 0.15757051766994487,
      "grad_norm": 0.382966548204422,
      "learning_rate": 8e-05,
      "loss": 1.9203,
      "step": 1458
    },
    {
      "epoch": 0.15767859072733167,
      "grad_norm": 0.35953521728515625,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 1459
    },
    {
      "epoch": 0.15778666378471848,
      "grad_norm": 0.39093002676963806,
      "learning_rate": 8e-05,
      "loss": 1.8129,
      "step": 1460
    },
    {
      "epoch": 0.15789473684210525,
      "grad_norm": 0.3680100739002228,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 1461
    },
    {
      "epoch": 0.15800280989949206,
      "grad_norm": 0.37109389901161194,
      "learning_rate": 8e-05,
      "loss": 1.7952,
      "step": 1462
    },
    {
      "epoch": 0.15811088295687886,
      "grad_norm": 0.35500001907348633,
      "learning_rate": 8e-05,
      "loss": 1.8556,
      "step": 1463
    },
    {
      "epoch": 0.15821895601426564,
      "grad_norm": 0.3553858995437622,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 1464
    },
    {
      "epoch": 0.15832702907165244,
      "grad_norm": 0.380461722612381,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 1465
    },
    {
      "epoch": 0.15843510212903922,
      "grad_norm": 0.36477822065353394,
      "learning_rate": 8e-05,
      "loss": 1.7948,
      "step": 1466
    },
    {
      "epoch": 0.15854317518642602,
      "grad_norm": 0.359115332365036,
      "learning_rate": 8e-05,
      "loss": 1.6428,
      "step": 1467
    },
    {
      "epoch": 0.15865124824381283,
      "grad_norm": 0.3710903823375702,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 1468
    },
    {
      "epoch": 0.1587593213011996,
      "grad_norm": 0.3482389748096466,
      "learning_rate": 8e-05,
      "loss": 1.7727,
      "step": 1469
    },
    {
      "epoch": 0.1588673943585864,
      "grad_norm": 0.3669455647468567,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 1470
    },
    {
      "epoch": 0.1589754674159732,
      "grad_norm": 0.37701231241226196,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 1471
    },
    {
      "epoch": 0.15908354047335999,
      "grad_norm": 0.3684040307998657,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 1472
    },
    {
      "epoch": 0.1591916135307468,
      "grad_norm": 0.3648243248462677,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1473
    },
    {
      "epoch": 0.15929968658813357,
      "grad_norm": 0.3526931703090668,
      "learning_rate": 8e-05,
      "loss": 1.7759,
      "step": 1474
    },
    {
      "epoch": 0.15940775964552037,
      "grad_norm": 0.3462124466896057,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 1475
    },
    {
      "epoch": 0.15951583270290717,
      "grad_norm": 0.37210389971733093,
      "learning_rate": 8e-05,
      "loss": 1.8001,
      "step": 1476
    },
    {
      "epoch": 0.15962390576029395,
      "grad_norm": 0.3444695472717285,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 1477
    },
    {
      "epoch": 0.15973197881768075,
      "grad_norm": 0.3394198715686798,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 1478
    },
    {
      "epoch": 0.15984005187506756,
      "grad_norm": 0.35910144448280334,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 1479
    },
    {
      "epoch": 0.15994812493245433,
      "grad_norm": 0.34826552867889404,
      "learning_rate": 8e-05,
      "loss": 1.8196,
      "step": 1480
    },
    {
      "epoch": 0.16005619798984114,
      "grad_norm": 0.35231512784957886,
      "learning_rate": 8e-05,
      "loss": 1.8121,
      "step": 1481
    },
    {
      "epoch": 0.1601642710472279,
      "grad_norm": 0.3483321964740753,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 1482
    },
    {
      "epoch": 0.16027234410461472,
      "grad_norm": 0.3588305711746216,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 1483
    },
    {
      "epoch": 0.16038041716200152,
      "grad_norm": 0.3416995406150818,
      "learning_rate": 8e-05,
      "loss": 1.6039,
      "step": 1484
    },
    {
      "epoch": 0.1604884902193883,
      "grad_norm": 0.3925912082195282,
      "learning_rate": 8e-05,
      "loss": 1.9478,
      "step": 1485
    },
    {
      "epoch": 0.1605965632767751,
      "grad_norm": 0.34834226965904236,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 1486
    },
    {
      "epoch": 0.1607046363341619,
      "grad_norm": 0.36548808217048645,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 1487
    },
    {
      "epoch": 0.16081270939154868,
      "grad_norm": 0.36002662777900696,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 1488
    },
    {
      "epoch": 0.16092078244893548,
      "grad_norm": 0.38215070962905884,
      "learning_rate": 8e-05,
      "loss": 1.9371,
      "step": 1489
    },
    {
      "epoch": 0.16102885550632226,
      "grad_norm": 0.3457402288913727,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 1490
    },
    {
      "epoch": 0.16113692856370906,
      "grad_norm": 0.34562206268310547,
      "learning_rate": 8e-05,
      "loss": 1.6431,
      "step": 1491
    },
    {
      "epoch": 0.16124500162109587,
      "grad_norm": 0.35419896245002747,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 1492
    },
    {
      "epoch": 0.16135307467848264,
      "grad_norm": 0.339605450630188,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 1493
    },
    {
      "epoch": 0.16146114773586945,
      "grad_norm": 0.35562995076179504,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 1494
    },
    {
      "epoch": 0.16156922079325625,
      "grad_norm": 0.38061460852622986,
      "learning_rate": 8e-05,
      "loss": 1.7959,
      "step": 1495
    },
    {
      "epoch": 0.16167729385064303,
      "grad_norm": 0.35861071944236755,
      "learning_rate": 8e-05,
      "loss": 1.6088,
      "step": 1496
    },
    {
      "epoch": 0.16178536690802983,
      "grad_norm": 0.42980650067329407,
      "learning_rate": 8e-05,
      "loss": 2.0117,
      "step": 1497
    },
    {
      "epoch": 0.1618934399654166,
      "grad_norm": 0.36173132061958313,
      "learning_rate": 8e-05,
      "loss": 1.8477,
      "step": 1498
    },
    {
      "epoch": 0.1620015130228034,
      "grad_norm": 0.35599803924560547,
      "learning_rate": 8e-05,
      "loss": 1.7989,
      "step": 1499
    },
    {
      "epoch": 0.16210958608019022,
      "grad_norm": 0.3865572214126587,
      "learning_rate": 8e-05,
      "loss": 1.6345,
      "step": 1500
    },
    {
      "epoch": 0.162217659137577,
      "grad_norm": 0.38279634714126587,
      "learning_rate": 8e-05,
      "loss": 1.8209,
      "step": 1501
    },
    {
      "epoch": 0.1623257321949638,
      "grad_norm": 0.35367077589035034,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 1502
    },
    {
      "epoch": 0.1624338052523506,
      "grad_norm": 0.36740198731422424,
      "learning_rate": 8e-05,
      "loss": 1.7575,
      "step": 1503
    },
    {
      "epoch": 0.16254187830973738,
      "grad_norm": 0.3541874289512634,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 1504
    },
    {
      "epoch": 0.16264995136712418,
      "grad_norm": 0.35867244005203247,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 1505
    },
    {
      "epoch": 0.16275802442451098,
      "grad_norm": 0.3447834849357605,
      "learning_rate": 8e-05,
      "loss": 1.6629,
      "step": 1506
    },
    {
      "epoch": 0.16286609748189776,
      "grad_norm": 0.353619247674942,
      "learning_rate": 8e-05,
      "loss": 1.5825,
      "step": 1507
    },
    {
      "epoch": 0.16297417053928456,
      "grad_norm": 0.34993064403533936,
      "learning_rate": 8e-05,
      "loss": 1.6087,
      "step": 1508
    },
    {
      "epoch": 0.16308224359667134,
      "grad_norm": 0.3920869529247284,
      "learning_rate": 8e-05,
      "loss": 1.8182,
      "step": 1509
    },
    {
      "epoch": 0.16319031665405814,
      "grad_norm": 0.3991021513938904,
      "learning_rate": 8e-05,
      "loss": 1.9579,
      "step": 1510
    },
    {
      "epoch": 0.16329838971144495,
      "grad_norm": 0.3829973638057709,
      "learning_rate": 8e-05,
      "loss": 1.9362,
      "step": 1511
    },
    {
      "epoch": 0.16340646276883172,
      "grad_norm": 0.3766539692878723,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 1512
    },
    {
      "epoch": 0.16351453582621853,
      "grad_norm": 0.34730976819992065,
      "learning_rate": 8e-05,
      "loss": 1.6612,
      "step": 1513
    },
    {
      "epoch": 0.16362260888360533,
      "grad_norm": 0.3472457826137543,
      "learning_rate": 8e-05,
      "loss": 1.596,
      "step": 1514
    },
    {
      "epoch": 0.1637306819409921,
      "grad_norm": 0.3421221673488617,
      "learning_rate": 8e-05,
      "loss": 1.6434,
      "step": 1515
    },
    {
      "epoch": 0.1638387549983789,
      "grad_norm": 0.3415828347206116,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 1516
    },
    {
      "epoch": 0.1639468280557657,
      "grad_norm": 0.35462209582328796,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 1517
    },
    {
      "epoch": 0.1640549011131525,
      "grad_norm": 0.34254932403564453,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 1518
    },
    {
      "epoch": 0.1641629741705393,
      "grad_norm": 0.3488869071006775,
      "learning_rate": 8e-05,
      "loss": 1.7057,
      "step": 1519
    },
    {
      "epoch": 0.16427104722792607,
      "grad_norm": 0.39570122957229614,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 1520
    },
    {
      "epoch": 0.16437912028531287,
      "grad_norm": 0.3649577498435974,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 1521
    },
    {
      "epoch": 0.16448719334269968,
      "grad_norm": 0.3431411385536194,
      "learning_rate": 8e-05,
      "loss": 1.7864,
      "step": 1522
    },
    {
      "epoch": 0.16459526640008645,
      "grad_norm": 0.3649054765701294,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 1523
    },
    {
      "epoch": 0.16470333945747326,
      "grad_norm": 0.39293426275253296,
      "learning_rate": 8e-05,
      "loss": 1.8406,
      "step": 1524
    },
    {
      "epoch": 0.16481141251486003,
      "grad_norm": 0.40931570529937744,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 1525
    },
    {
      "epoch": 0.16491948557224684,
      "grad_norm": 0.38008323311805725,
      "learning_rate": 8e-05,
      "loss": 1.9167,
      "step": 1526
    },
    {
      "epoch": 0.16502755862963364,
      "grad_norm": 0.4023352563381195,
      "learning_rate": 8e-05,
      "loss": 1.926,
      "step": 1527
    },
    {
      "epoch": 0.16513563168702042,
      "grad_norm": 0.3599163591861725,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 1528
    },
    {
      "epoch": 0.16524370474440722,
      "grad_norm": 0.39081862568855286,
      "learning_rate": 8e-05,
      "loss": 1.7677,
      "step": 1529
    },
    {
      "epoch": 0.16535177780179403,
      "grad_norm": 0.36268776655197144,
      "learning_rate": 8e-05,
      "loss": 1.94,
      "step": 1530
    },
    {
      "epoch": 0.1654598508591808,
      "grad_norm": 0.3584979772567749,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 1531
    },
    {
      "epoch": 0.1655679239165676,
      "grad_norm": 0.36618903279304504,
      "learning_rate": 8e-05,
      "loss": 1.8008,
      "step": 1532
    },
    {
      "epoch": 0.16567599697395438,
      "grad_norm": 0.3684219419956207,
      "learning_rate": 8e-05,
      "loss": 1.7516,
      "step": 1533
    },
    {
      "epoch": 0.16578407003134119,
      "grad_norm": 0.3563080132007599,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 1534
    },
    {
      "epoch": 0.165892143088728,
      "grad_norm": 0.3727073073387146,
      "learning_rate": 8e-05,
      "loss": 1.8343,
      "step": 1535
    },
    {
      "epoch": 0.16600021614611477,
      "grad_norm": 0.3701125681400299,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 1536
    },
    {
      "epoch": 0.16610828920350157,
      "grad_norm": 0.37388166785240173,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 1537
    },
    {
      "epoch": 0.16621636226088837,
      "grad_norm": 0.3690321147441864,
      "learning_rate": 8e-05,
      "loss": 1.7482,
      "step": 1538
    },
    {
      "epoch": 0.16632443531827515,
      "grad_norm": 0.36835116147994995,
      "learning_rate": 8e-05,
      "loss": 1.7171,
      "step": 1539
    },
    {
      "epoch": 0.16643250837566195,
      "grad_norm": 0.3663565516471863,
      "learning_rate": 8e-05,
      "loss": 1.7528,
      "step": 1540
    },
    {
      "epoch": 0.16654058143304873,
      "grad_norm": 0.3550228774547577,
      "learning_rate": 8e-05,
      "loss": 1.7521,
      "step": 1541
    },
    {
      "epoch": 0.16664865449043553,
      "grad_norm": 0.35605695843696594,
      "learning_rate": 8e-05,
      "loss": 1.6268,
      "step": 1542
    },
    {
      "epoch": 0.16675672754782234,
      "grad_norm": 0.373093843460083,
      "learning_rate": 8e-05,
      "loss": 1.8254,
      "step": 1543
    },
    {
      "epoch": 0.1668648006052091,
      "grad_norm": 0.36627182364463806,
      "learning_rate": 8e-05,
      "loss": 1.5629,
      "step": 1544
    },
    {
      "epoch": 0.16697287366259592,
      "grad_norm": 0.358193039894104,
      "learning_rate": 8e-05,
      "loss": 1.8812,
      "step": 1545
    },
    {
      "epoch": 0.16708094671998272,
      "grad_norm": 0.38430899381637573,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 1546
    },
    {
      "epoch": 0.1671890197773695,
      "grad_norm": 0.36285752058029175,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 1547
    },
    {
      "epoch": 0.1672970928347563,
      "grad_norm": 0.35644498467445374,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 1548
    },
    {
      "epoch": 0.16740516589214308,
      "grad_norm": 0.37135079503059387,
      "learning_rate": 8e-05,
      "loss": 1.7624,
      "step": 1549
    },
    {
      "epoch": 0.16751323894952988,
      "grad_norm": 0.40812957286834717,
      "learning_rate": 8e-05,
      "loss": 1.9663,
      "step": 1550
    },
    {
      "epoch": 0.16762131200691668,
      "grad_norm": 0.3721383810043335,
      "learning_rate": 8e-05,
      "loss": 1.8155,
      "step": 1551
    },
    {
      "epoch": 0.16772938506430346,
      "grad_norm": 0.343982994556427,
      "learning_rate": 8e-05,
      "loss": 1.9217,
      "step": 1552
    },
    {
      "epoch": 0.16783745812169026,
      "grad_norm": 0.36501482129096985,
      "learning_rate": 8e-05,
      "loss": 1.8496,
      "step": 1553
    },
    {
      "epoch": 0.16794553117907707,
      "grad_norm": 0.36781203746795654,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 1554
    },
    {
      "epoch": 0.16805360423646384,
      "grad_norm": 0.372563898563385,
      "learning_rate": 8e-05,
      "loss": 1.9814,
      "step": 1555
    },
    {
      "epoch": 0.16816167729385065,
      "grad_norm": 0.38038164377212524,
      "learning_rate": 8e-05,
      "loss": 1.8364,
      "step": 1556
    },
    {
      "epoch": 0.16826975035123742,
      "grad_norm": 0.40095072984695435,
      "learning_rate": 8e-05,
      "loss": 1.8653,
      "step": 1557
    },
    {
      "epoch": 0.16837782340862423,
      "grad_norm": 0.34620606899261475,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 1558
    },
    {
      "epoch": 0.16848589646601103,
      "grad_norm": 0.37505319714546204,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 1559
    },
    {
      "epoch": 0.1685939695233978,
      "grad_norm": 0.3739761412143707,
      "learning_rate": 8e-05,
      "loss": 1.8345,
      "step": 1560
    },
    {
      "epoch": 0.1687020425807846,
      "grad_norm": 0.3686574101448059,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 1561
    },
    {
      "epoch": 0.16881011563817142,
      "grad_norm": 0.3540516495704651,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 1562
    },
    {
      "epoch": 0.1689181886955582,
      "grad_norm": 0.3449227809906006,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 1563
    },
    {
      "epoch": 0.169026261752945,
      "grad_norm": 0.3845237195491791,
      "learning_rate": 8e-05,
      "loss": 1.9325,
      "step": 1564
    },
    {
      "epoch": 0.16913433481033177,
      "grad_norm": 0.3843061029911041,
      "learning_rate": 8e-05,
      "loss": 1.8386,
      "step": 1565
    },
    {
      "epoch": 0.16924240786771858,
      "grad_norm": 0.4084787964820862,
      "learning_rate": 8e-05,
      "loss": 1.9091,
      "step": 1566
    },
    {
      "epoch": 0.16935048092510538,
      "grad_norm": 0.35452964901924133,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 1567
    },
    {
      "epoch": 0.16945855398249215,
      "grad_norm": 0.359918475151062,
      "learning_rate": 8e-05,
      "loss": 1.881,
      "step": 1568
    },
    {
      "epoch": 0.16956662703987896,
      "grad_norm": 0.3757287263870239,
      "learning_rate": 8e-05,
      "loss": 1.8829,
      "step": 1569
    },
    {
      "epoch": 0.16967470009726576,
      "grad_norm": 0.37208443880081177,
      "learning_rate": 8e-05,
      "loss": 1.7709,
      "step": 1570
    },
    {
      "epoch": 0.16978277315465254,
      "grad_norm": 0.3668755292892456,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 1571
    },
    {
      "epoch": 0.16989084621203934,
      "grad_norm": 0.3543424606323242,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 1572
    },
    {
      "epoch": 0.16999891926942612,
      "grad_norm": 0.3879331648349762,
      "learning_rate": 8e-05,
      "loss": 1.9224,
      "step": 1573
    },
    {
      "epoch": 0.17010699232681292,
      "grad_norm": 0.39529386162757874,
      "learning_rate": 8e-05,
      "loss": 1.7981,
      "step": 1574
    },
    {
      "epoch": 0.17021506538419973,
      "grad_norm": 0.3868407905101776,
      "learning_rate": 8e-05,
      "loss": 1.8664,
      "step": 1575
    },
    {
      "epoch": 0.1703231384415865,
      "grad_norm": 0.3503515124320984,
      "learning_rate": 8e-05,
      "loss": 1.7718,
      "step": 1576
    },
    {
      "epoch": 0.1704312114989733,
      "grad_norm": 0.33907774090766907,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 1577
    },
    {
      "epoch": 0.1705392845563601,
      "grad_norm": 0.34026020765304565,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 1578
    },
    {
      "epoch": 0.1706473576137469,
      "grad_norm": 0.4544883072376251,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 1579
    },
    {
      "epoch": 0.1707554306711337,
      "grad_norm": 0.3494049608707428,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 1580
    },
    {
      "epoch": 0.17086350372852047,
      "grad_norm": 0.3605392277240753,
      "learning_rate": 8e-05,
      "loss": 1.7685,
      "step": 1581
    },
    {
      "epoch": 0.17097157678590727,
      "grad_norm": 0.3782109320163727,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 1582
    },
    {
      "epoch": 0.17107964984329407,
      "grad_norm": 0.39154326915740967,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 1583
    },
    {
      "epoch": 0.17118772290068085,
      "grad_norm": 0.36011967062950134,
      "learning_rate": 8e-05,
      "loss": 1.6658,
      "step": 1584
    },
    {
      "epoch": 0.17129579595806765,
      "grad_norm": 0.3702276051044464,
      "learning_rate": 8e-05,
      "loss": 1.5475,
      "step": 1585
    },
    {
      "epoch": 0.17140386901545446,
      "grad_norm": 0.3594640791416168,
      "learning_rate": 8e-05,
      "loss": 1.7196,
      "step": 1586
    },
    {
      "epoch": 0.17151194207284123,
      "grad_norm": 0.36223164200782776,
      "learning_rate": 8e-05,
      "loss": 1.8273,
      "step": 1587
    },
    {
      "epoch": 0.17162001513022804,
      "grad_norm": 0.34349924325942993,
      "learning_rate": 8e-05,
      "loss": 1.7268,
      "step": 1588
    },
    {
      "epoch": 0.17172808818761484,
      "grad_norm": 0.3908827602863312,
      "learning_rate": 8e-05,
      "loss": 1.8184,
      "step": 1589
    },
    {
      "epoch": 0.17183616124500162,
      "grad_norm": 0.35485586524009705,
      "learning_rate": 8e-05,
      "loss": 1.6687,
      "step": 1590
    },
    {
      "epoch": 0.17194423430238842,
      "grad_norm": 0.37867337465286255,
      "learning_rate": 8e-05,
      "loss": 1.7827,
      "step": 1591
    },
    {
      "epoch": 0.1720523073597752,
      "grad_norm": 0.38368913531303406,
      "learning_rate": 8e-05,
      "loss": 1.8279,
      "step": 1592
    },
    {
      "epoch": 0.172160380417162,
      "grad_norm": 0.3789379894733429,
      "learning_rate": 8e-05,
      "loss": 1.6387,
      "step": 1593
    },
    {
      "epoch": 0.1722684534745488,
      "grad_norm": 0.3658318817615509,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 1594
    },
    {
      "epoch": 0.17237652653193558,
      "grad_norm": 0.36771678924560547,
      "learning_rate": 8e-05,
      "loss": 1.7356,
      "step": 1595
    },
    {
      "epoch": 0.17248459958932238,
      "grad_norm": 0.36121058464050293,
      "learning_rate": 8e-05,
      "loss": 1.7365,
      "step": 1596
    },
    {
      "epoch": 0.1725926726467092,
      "grad_norm": 0.36337608098983765,
      "learning_rate": 8e-05,
      "loss": 1.8057,
      "step": 1597
    },
    {
      "epoch": 0.17270074570409596,
      "grad_norm": 0.36104637384414673,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 1598
    },
    {
      "epoch": 0.17280881876148277,
      "grad_norm": 0.369521826505661,
      "learning_rate": 8e-05,
      "loss": 1.8027,
      "step": 1599
    },
    {
      "epoch": 0.17291689181886954,
      "grad_norm": 0.3675416111946106,
      "learning_rate": 8e-05,
      "loss": 1.6951,
      "step": 1600
    },
    {
      "epoch": 0.17302496487625635,
      "grad_norm": 0.3630196154117584,
      "learning_rate": 8e-05,
      "loss": 1.9198,
      "step": 1601
    },
    {
      "epoch": 0.17313303793364315,
      "grad_norm": 0.3838110864162445,
      "learning_rate": 8e-05,
      "loss": 1.9784,
      "step": 1602
    },
    {
      "epoch": 0.17324111099102993,
      "grad_norm": 0.34828025102615356,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 1603
    },
    {
      "epoch": 0.17334918404841673,
      "grad_norm": 0.376072496175766,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 1604
    },
    {
      "epoch": 0.17345725710580354,
      "grad_norm": 0.38232260942459106,
      "learning_rate": 8e-05,
      "loss": 1.9361,
      "step": 1605
    },
    {
      "epoch": 0.1735653301631903,
      "grad_norm": 0.360359251499176,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 1606
    },
    {
      "epoch": 0.17367340322057712,
      "grad_norm": 0.34956425428390503,
      "learning_rate": 8e-05,
      "loss": 1.6475,
      "step": 1607
    },
    {
      "epoch": 0.1737814762779639,
      "grad_norm": 0.40701553225517273,
      "learning_rate": 8e-05,
      "loss": 1.888,
      "step": 1608
    },
    {
      "epoch": 0.1738895493353507,
      "grad_norm": 0.37468236684799194,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 1609
    },
    {
      "epoch": 0.1739976223927375,
      "grad_norm": 0.3446200489997864,
      "learning_rate": 8e-05,
      "loss": 1.7896,
      "step": 1610
    },
    {
      "epoch": 0.17410569545012428,
      "grad_norm": 0.37653806805610657,
      "learning_rate": 8e-05,
      "loss": 1.8452,
      "step": 1611
    },
    {
      "epoch": 0.17421376850751108,
      "grad_norm": 0.35356035828590393,
      "learning_rate": 8e-05,
      "loss": 1.7478,
      "step": 1612
    },
    {
      "epoch": 0.17432184156489788,
      "grad_norm": 0.3565081059932709,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 1613
    },
    {
      "epoch": 0.17442991462228466,
      "grad_norm": 0.3607056140899658,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 1614
    },
    {
      "epoch": 0.17453798767967146,
      "grad_norm": 0.3345412611961365,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 1615
    },
    {
      "epoch": 0.17464606073705824,
      "grad_norm": 0.3666120767593384,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 1616
    },
    {
      "epoch": 0.17475413379444504,
      "grad_norm": 0.34048354625701904,
      "learning_rate": 8e-05,
      "loss": 1.7696,
      "step": 1617
    },
    {
      "epoch": 0.17486220685183185,
      "grad_norm": 0.35646718740463257,
      "learning_rate": 8e-05,
      "loss": 1.9065,
      "step": 1618
    },
    {
      "epoch": 0.17497027990921862,
      "grad_norm": 0.42236730456352234,
      "learning_rate": 8e-05,
      "loss": 1.9125,
      "step": 1619
    },
    {
      "epoch": 0.17507835296660543,
      "grad_norm": 0.40540653467178345,
      "learning_rate": 8e-05,
      "loss": 1.8505,
      "step": 1620
    },
    {
      "epoch": 0.17518642602399223,
      "grad_norm": 0.331855446100235,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 1621
    },
    {
      "epoch": 0.175294499081379,
      "grad_norm": 0.36744189262390137,
      "learning_rate": 8e-05,
      "loss": 1.8875,
      "step": 1622
    },
    {
      "epoch": 0.1754025721387658,
      "grad_norm": 0.3942550718784332,
      "learning_rate": 8e-05,
      "loss": 1.956,
      "step": 1623
    },
    {
      "epoch": 0.1755106451961526,
      "grad_norm": 0.3567750155925751,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 1624
    },
    {
      "epoch": 0.1756187182535394,
      "grad_norm": 0.36972054839134216,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 1625
    },
    {
      "epoch": 0.1757267913109262,
      "grad_norm": 0.3746875822544098,
      "learning_rate": 8e-05,
      "loss": 1.8802,
      "step": 1626
    },
    {
      "epoch": 0.17583486436831297,
      "grad_norm": 0.3661128878593445,
      "learning_rate": 8e-05,
      "loss": 1.7955,
      "step": 1627
    },
    {
      "epoch": 0.17594293742569977,
      "grad_norm": 0.35059115290641785,
      "learning_rate": 8e-05,
      "loss": 1.8725,
      "step": 1628
    },
    {
      "epoch": 0.17605101048308658,
      "grad_norm": 0.36802592873573303,
      "learning_rate": 8e-05,
      "loss": 1.661,
      "step": 1629
    },
    {
      "epoch": 0.17615908354047335,
      "grad_norm": 0.36493128538131714,
      "learning_rate": 8e-05,
      "loss": 1.8347,
      "step": 1630
    },
    {
      "epoch": 0.17626715659786016,
      "grad_norm": 0.3848332464694977,
      "learning_rate": 8e-05,
      "loss": 1.7785,
      "step": 1631
    },
    {
      "epoch": 0.17637522965524693,
      "grad_norm": 0.3916698098182678,
      "learning_rate": 8e-05,
      "loss": 1.7927,
      "step": 1632
    },
    {
      "epoch": 0.17648330271263374,
      "grad_norm": 0.45152509212493896,
      "learning_rate": 8e-05,
      "loss": 2.045,
      "step": 1633
    },
    {
      "epoch": 0.17659137577002054,
      "grad_norm": 0.4029836058616638,
      "learning_rate": 8e-05,
      "loss": 1.9389,
      "step": 1634
    },
    {
      "epoch": 0.17669944882740732,
      "grad_norm": 0.4467923939228058,
      "learning_rate": 8e-05,
      "loss": 2.0114,
      "step": 1635
    },
    {
      "epoch": 0.17680752188479412,
      "grad_norm": 0.4108101725578308,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 1636
    },
    {
      "epoch": 0.17691559494218093,
      "grad_norm": 0.38180819153785706,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 1637
    },
    {
      "epoch": 0.1770236679995677,
      "grad_norm": 0.3952118754386902,
      "learning_rate": 8e-05,
      "loss": 1.8248,
      "step": 1638
    },
    {
      "epoch": 0.1771317410569545,
      "grad_norm": 0.38845109939575195,
      "learning_rate": 8e-05,
      "loss": 1.7078,
      "step": 1639
    },
    {
      "epoch": 0.17723981411434128,
      "grad_norm": 0.359057754278183,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 1640
    },
    {
      "epoch": 0.17734788717172809,
      "grad_norm": 0.34535807371139526,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 1641
    },
    {
      "epoch": 0.1774559602291149,
      "grad_norm": 0.37608200311660767,
      "learning_rate": 8e-05,
      "loss": 1.7808,
      "step": 1642
    },
    {
      "epoch": 0.17756403328650167,
      "grad_norm": 0.4053289592266083,
      "learning_rate": 8e-05,
      "loss": 1.9428,
      "step": 1643
    },
    {
      "epoch": 0.17767210634388847,
      "grad_norm": 0.3741133511066437,
      "learning_rate": 8e-05,
      "loss": 1.8434,
      "step": 1644
    },
    {
      "epoch": 0.17778017940127527,
      "grad_norm": 0.34029701352119446,
      "learning_rate": 8e-05,
      "loss": 1.7948,
      "step": 1645
    },
    {
      "epoch": 0.17788825245866205,
      "grad_norm": 0.35523542761802673,
      "learning_rate": 8e-05,
      "loss": 1.7983,
      "step": 1646
    },
    {
      "epoch": 0.17799632551604885,
      "grad_norm": 0.35122838616371155,
      "learning_rate": 8e-05,
      "loss": 1.6771,
      "step": 1647
    },
    {
      "epoch": 0.17810439857343563,
      "grad_norm": 0.36682063341140747,
      "learning_rate": 8e-05,
      "loss": 1.7303,
      "step": 1648
    },
    {
      "epoch": 0.17821247163082243,
      "grad_norm": 0.34502407908439636,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 1649
    },
    {
      "epoch": 0.17832054468820924,
      "grad_norm": 0.36331263184547424,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 1650
    },
    {
      "epoch": 0.178428617745596,
      "grad_norm": 0.3590247929096222,
      "learning_rate": 8e-05,
      "loss": 1.6528,
      "step": 1651
    },
    {
      "epoch": 0.17853669080298282,
      "grad_norm": 0.3526703119277954,
      "learning_rate": 8e-05,
      "loss": 1.706,
      "step": 1652
    },
    {
      "epoch": 0.17864476386036962,
      "grad_norm": 0.36700600385665894,
      "learning_rate": 8e-05,
      "loss": 1.8348,
      "step": 1653
    },
    {
      "epoch": 0.1787528369177564,
      "grad_norm": 0.37900376319885254,
      "learning_rate": 8e-05,
      "loss": 1.7998,
      "step": 1654
    },
    {
      "epoch": 0.1788609099751432,
      "grad_norm": 0.3786175847053528,
      "learning_rate": 8e-05,
      "loss": 1.7667,
      "step": 1655
    },
    {
      "epoch": 0.17896898303252998,
      "grad_norm": 0.3788127303123474,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 1656
    },
    {
      "epoch": 0.17907705608991678,
      "grad_norm": 0.3550642728805542,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 1657
    },
    {
      "epoch": 0.17918512914730358,
      "grad_norm": 0.372003972530365,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 1658
    },
    {
      "epoch": 0.17929320220469036,
      "grad_norm": 0.35628843307495117,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 1659
    },
    {
      "epoch": 0.17940127526207716,
      "grad_norm": 0.3718128204345703,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 1660
    },
    {
      "epoch": 0.17950934831946397,
      "grad_norm": 0.3651917278766632,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 1661
    },
    {
      "epoch": 0.17961742137685074,
      "grad_norm": 0.40477707982063293,
      "learning_rate": 8e-05,
      "loss": 1.9552,
      "step": 1662
    },
    {
      "epoch": 0.17972549443423755,
      "grad_norm": 0.33563488721847534,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1663
    },
    {
      "epoch": 0.17983356749162432,
      "grad_norm": 0.35286417603492737,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 1664
    },
    {
      "epoch": 0.17994164054901113,
      "grad_norm": 0.39628294110298157,
      "learning_rate": 8e-05,
      "loss": 1.6988,
      "step": 1665
    },
    {
      "epoch": 0.18004971360639793,
      "grad_norm": 0.3455020487308502,
      "learning_rate": 8e-05,
      "loss": 1.7121,
      "step": 1666
    },
    {
      "epoch": 0.1801577866637847,
      "grad_norm": 0.39466869831085205,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 1667
    },
    {
      "epoch": 0.1802658597211715,
      "grad_norm": 0.3614978790283203,
      "learning_rate": 8e-05,
      "loss": 1.814,
      "step": 1668
    },
    {
      "epoch": 0.18037393277855832,
      "grad_norm": 0.36236196756362915,
      "learning_rate": 8e-05,
      "loss": 1.7807,
      "step": 1669
    },
    {
      "epoch": 0.1804820058359451,
      "grad_norm": 0.3276165723800659,
      "learning_rate": 8e-05,
      "loss": 1.5727,
      "step": 1670
    },
    {
      "epoch": 0.1805900788933319,
      "grad_norm": 0.3774045407772064,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 1671
    },
    {
      "epoch": 0.1806981519507187,
      "grad_norm": 0.34874510765075684,
      "learning_rate": 8e-05,
      "loss": 1.6803,
      "step": 1672
    },
    {
      "epoch": 0.18080622500810548,
      "grad_norm": 0.36427995562553406,
      "learning_rate": 8e-05,
      "loss": 1.7486,
      "step": 1673
    },
    {
      "epoch": 0.18091429806549228,
      "grad_norm": 0.38381022214889526,
      "learning_rate": 8e-05,
      "loss": 1.7939,
      "step": 1674
    },
    {
      "epoch": 0.18102237112287906,
      "grad_norm": 0.3950136601924896,
      "learning_rate": 8e-05,
      "loss": 1.9148,
      "step": 1675
    },
    {
      "epoch": 0.18113044418026586,
      "grad_norm": 0.37090596556663513,
      "learning_rate": 8e-05,
      "loss": 1.8078,
      "step": 1676
    },
    {
      "epoch": 0.18123851723765266,
      "grad_norm": 0.36812740564346313,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 1677
    },
    {
      "epoch": 0.18134659029503944,
      "grad_norm": 0.3782331645488739,
      "learning_rate": 8e-05,
      "loss": 1.8253,
      "step": 1678
    },
    {
      "epoch": 0.18145466335242624,
      "grad_norm": 0.3781827986240387,
      "learning_rate": 8e-05,
      "loss": 1.9345,
      "step": 1679
    },
    {
      "epoch": 0.18156273640981305,
      "grad_norm": 0.37168988585472107,
      "learning_rate": 8e-05,
      "loss": 1.7043,
      "step": 1680
    },
    {
      "epoch": 0.18167080946719982,
      "grad_norm": 0.34382009506225586,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 1681
    },
    {
      "epoch": 0.18177888252458663,
      "grad_norm": 0.3758091628551483,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 1682
    },
    {
      "epoch": 0.1818869555819734,
      "grad_norm": 0.3712696135044098,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 1683
    },
    {
      "epoch": 0.1819950286393602,
      "grad_norm": 0.36939799785614014,
      "learning_rate": 8e-05,
      "loss": 1.8414,
      "step": 1684
    },
    {
      "epoch": 0.182103101696747,
      "grad_norm": 0.36691927909851074,
      "learning_rate": 8e-05,
      "loss": 1.7739,
      "step": 1685
    },
    {
      "epoch": 0.1822111747541338,
      "grad_norm": 0.3789370059967041,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 1686
    },
    {
      "epoch": 0.1823192478115206,
      "grad_norm": 0.3779565393924713,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 1687
    },
    {
      "epoch": 0.1824273208689074,
      "grad_norm": 0.34127485752105713,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 1688
    },
    {
      "epoch": 0.18253539392629417,
      "grad_norm": 0.3747107684612274,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 1689
    },
    {
      "epoch": 0.18264346698368097,
      "grad_norm": 0.35757023096084595,
      "learning_rate": 8e-05,
      "loss": 1.7568,
      "step": 1690
    },
    {
      "epoch": 0.18275154004106775,
      "grad_norm": 0.3674648404121399,
      "learning_rate": 8e-05,
      "loss": 1.8291,
      "step": 1691
    },
    {
      "epoch": 0.18285961309845455,
      "grad_norm": 0.38322946429252625,
      "learning_rate": 8e-05,
      "loss": 1.9127,
      "step": 1692
    },
    {
      "epoch": 0.18296768615584136,
      "grad_norm": 0.33984994888305664,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 1693
    },
    {
      "epoch": 0.18307575921322813,
      "grad_norm": 0.3492932617664337,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 1694
    },
    {
      "epoch": 0.18318383227061494,
      "grad_norm": 0.3868374526500702,
      "learning_rate": 8e-05,
      "loss": 1.9592,
      "step": 1695
    },
    {
      "epoch": 0.18329190532800174,
      "grad_norm": 0.3798457682132721,
      "learning_rate": 8e-05,
      "loss": 1.9557,
      "step": 1696
    },
    {
      "epoch": 0.18339997838538852,
      "grad_norm": 0.341256707906723,
      "learning_rate": 8e-05,
      "loss": 1.5615,
      "step": 1697
    },
    {
      "epoch": 0.18350805144277532,
      "grad_norm": 0.36626920104026794,
      "learning_rate": 8e-05,
      "loss": 1.557,
      "step": 1698
    },
    {
      "epoch": 0.1836161245001621,
      "grad_norm": 0.37801268696784973,
      "learning_rate": 8e-05,
      "loss": 1.9493,
      "step": 1699
    },
    {
      "epoch": 0.1837241975575489,
      "grad_norm": 0.39826786518096924,
      "learning_rate": 8e-05,
      "loss": 1.8467,
      "step": 1700
    },
    {
      "epoch": 0.1838322706149357,
      "grad_norm": 0.35098350048065186,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 1701
    },
    {
      "epoch": 0.18394034367232248,
      "grad_norm": 0.36847978830337524,
      "learning_rate": 8e-05,
      "loss": 1.8027,
      "step": 1702
    },
    {
      "epoch": 0.18404841672970929,
      "grad_norm": 0.3955918252468109,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 1703
    },
    {
      "epoch": 0.1841564897870961,
      "grad_norm": 0.3944500982761383,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 1704
    },
    {
      "epoch": 0.18426456284448287,
      "grad_norm": 0.34581539034843445,
      "learning_rate": 8e-05,
      "loss": 1.5763,
      "step": 1705
    },
    {
      "epoch": 0.18437263590186967,
      "grad_norm": 0.37280723452568054,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 1706
    },
    {
      "epoch": 0.18448070895925645,
      "grad_norm": 0.32942456007003784,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 1707
    },
    {
      "epoch": 0.18458878201664325,
      "grad_norm": 0.35244083404541016,
      "learning_rate": 8e-05,
      "loss": 1.7128,
      "step": 1708
    },
    {
      "epoch": 0.18469685507403005,
      "grad_norm": 0.36334046721458435,
      "learning_rate": 8e-05,
      "loss": 1.8308,
      "step": 1709
    },
    {
      "epoch": 0.18480492813141683,
      "grad_norm": 0.3682284653186798,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 1710
    },
    {
      "epoch": 0.18491300118880363,
      "grad_norm": 0.3617183268070221,
      "learning_rate": 8e-05,
      "loss": 1.7895,
      "step": 1711
    },
    {
      "epoch": 0.18502107424619044,
      "grad_norm": 0.3655518591403961,
      "learning_rate": 8e-05,
      "loss": 1.7255,
      "step": 1712
    },
    {
      "epoch": 0.1851291473035772,
      "grad_norm": 0.3500616252422333,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 1713
    },
    {
      "epoch": 0.18523722036096402,
      "grad_norm": 0.3690573275089264,
      "learning_rate": 8e-05,
      "loss": 1.6323,
      "step": 1714
    },
    {
      "epoch": 0.1853452934183508,
      "grad_norm": 0.3751242458820343,
      "learning_rate": 8e-05,
      "loss": 1.801,
      "step": 1715
    },
    {
      "epoch": 0.1854533664757376,
      "grad_norm": 0.3810383081436157,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 1716
    },
    {
      "epoch": 0.1855614395331244,
      "grad_norm": 0.36056819558143616,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 1717
    },
    {
      "epoch": 0.18566951259051118,
      "grad_norm": 0.3514404594898224,
      "learning_rate": 8e-05,
      "loss": 1.8355,
      "step": 1718
    },
    {
      "epoch": 0.18577758564789798,
      "grad_norm": 0.36354658007621765,
      "learning_rate": 8e-05,
      "loss": 1.7155,
      "step": 1719
    },
    {
      "epoch": 0.18588565870528478,
      "grad_norm": 0.3387967050075531,
      "learning_rate": 8e-05,
      "loss": 1.7008,
      "step": 1720
    },
    {
      "epoch": 0.18599373176267156,
      "grad_norm": 0.35997846722602844,
      "learning_rate": 8e-05,
      "loss": 1.7762,
      "step": 1721
    },
    {
      "epoch": 0.18610180482005836,
      "grad_norm": 0.3943254351615906,
      "learning_rate": 8e-05,
      "loss": 1.8421,
      "step": 1722
    },
    {
      "epoch": 0.18620987787744514,
      "grad_norm": 0.3703499734401703,
      "learning_rate": 8e-05,
      "loss": 1.7988,
      "step": 1723
    },
    {
      "epoch": 0.18631795093483194,
      "grad_norm": 0.3872607946395874,
      "learning_rate": 8e-05,
      "loss": 1.8857,
      "step": 1724
    },
    {
      "epoch": 0.18642602399221875,
      "grad_norm": 0.3822230398654938,
      "learning_rate": 8e-05,
      "loss": 1.8202,
      "step": 1725
    },
    {
      "epoch": 0.18653409704960552,
      "grad_norm": 0.43150418996810913,
      "learning_rate": 8e-05,
      "loss": 1.9394,
      "step": 1726
    },
    {
      "epoch": 0.18664217010699233,
      "grad_norm": 0.36196598410606384,
      "learning_rate": 8e-05,
      "loss": 1.738,
      "step": 1727
    },
    {
      "epoch": 0.18675024316437913,
      "grad_norm": 0.4014938473701477,
      "learning_rate": 8e-05,
      "loss": 1.8033,
      "step": 1728
    },
    {
      "epoch": 0.1868583162217659,
      "grad_norm": 0.3539465665817261,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 1729
    },
    {
      "epoch": 0.1869663892791527,
      "grad_norm": 0.3777218461036682,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 1730
    },
    {
      "epoch": 0.1870744623365395,
      "grad_norm": 0.4082215130329132,
      "learning_rate": 8e-05,
      "loss": 1.6875,
      "step": 1731
    },
    {
      "epoch": 0.1871825353939263,
      "grad_norm": 0.37269142270088196,
      "learning_rate": 8e-05,
      "loss": 1.8597,
      "step": 1732
    },
    {
      "epoch": 0.1872906084513131,
      "grad_norm": 0.3698691427707672,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 1733
    },
    {
      "epoch": 0.18739868150869987,
      "grad_norm": 0.35455918312072754,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 1734
    },
    {
      "epoch": 0.18750675456608668,
      "grad_norm": 0.3900121748447418,
      "learning_rate": 8e-05,
      "loss": 1.8418,
      "step": 1735
    },
    {
      "epoch": 0.18761482762347348,
      "grad_norm": 0.3949868083000183,
      "learning_rate": 8e-05,
      "loss": 1.9188,
      "step": 1736
    },
    {
      "epoch": 0.18772290068086026,
      "grad_norm": 0.3427894115447998,
      "learning_rate": 8e-05,
      "loss": 1.7221,
      "step": 1737
    },
    {
      "epoch": 0.18783097373824706,
      "grad_norm": 0.3467901349067688,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 1738
    },
    {
      "epoch": 0.18793904679563384,
      "grad_norm": 0.3490123748779297,
      "learning_rate": 8e-05,
      "loss": 1.792,
      "step": 1739
    },
    {
      "epoch": 0.18804711985302064,
      "grad_norm": 0.4113178551197052,
      "learning_rate": 8e-05,
      "loss": 1.8467,
      "step": 1740
    },
    {
      "epoch": 0.18815519291040744,
      "grad_norm": 0.3516266345977783,
      "learning_rate": 8e-05,
      "loss": 1.6673,
      "step": 1741
    },
    {
      "epoch": 0.18826326596779422,
      "grad_norm": 0.35637328028678894,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 1742
    },
    {
      "epoch": 0.18837133902518102,
      "grad_norm": 0.3567594289779663,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 1743
    },
    {
      "epoch": 0.18847941208256783,
      "grad_norm": 0.3844467103481293,
      "learning_rate": 8e-05,
      "loss": 1.8178,
      "step": 1744
    },
    {
      "epoch": 0.1885874851399546,
      "grad_norm": 0.37149736285209656,
      "learning_rate": 8e-05,
      "loss": 1.8056,
      "step": 1745
    },
    {
      "epoch": 0.1886955581973414,
      "grad_norm": 0.36744555830955505,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 1746
    },
    {
      "epoch": 0.18880363125472818,
      "grad_norm": 0.3590886890888214,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 1747
    },
    {
      "epoch": 0.188911704312115,
      "grad_norm": 0.38177254796028137,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 1748
    },
    {
      "epoch": 0.1890197773695018,
      "grad_norm": 0.3755657970905304,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 1749
    },
    {
      "epoch": 0.18912785042688857,
      "grad_norm": 0.34013307094573975,
      "learning_rate": 8e-05,
      "loss": 1.7792,
      "step": 1750
    },
    {
      "epoch": 0.18923592348427537,
      "grad_norm": 0.3533726632595062,
      "learning_rate": 8e-05,
      "loss": 1.6965,
      "step": 1751
    },
    {
      "epoch": 0.18934399654166217,
      "grad_norm": 0.3901097774505615,
      "learning_rate": 8e-05,
      "loss": 1.8536,
      "step": 1752
    },
    {
      "epoch": 0.18945206959904895,
      "grad_norm": 0.355742871761322,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 1753
    },
    {
      "epoch": 0.18956014265643575,
      "grad_norm": 0.3505704998970032,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 1754
    },
    {
      "epoch": 0.18966821571382256,
      "grad_norm": 0.3514764904975891,
      "learning_rate": 8e-05,
      "loss": 1.7449,
      "step": 1755
    },
    {
      "epoch": 0.18977628877120933,
      "grad_norm": 0.33720675110816956,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 1756
    },
    {
      "epoch": 0.18988436182859614,
      "grad_norm": 0.4030938148498535,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 1757
    },
    {
      "epoch": 0.1899924348859829,
      "grad_norm": 0.3569396734237671,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 1758
    },
    {
      "epoch": 0.19010050794336972,
      "grad_norm": 0.374889999628067,
      "learning_rate": 8e-05,
      "loss": 1.8444,
      "step": 1759
    },
    {
      "epoch": 0.19020858100075652,
      "grad_norm": 0.3418726921081543,
      "learning_rate": 8e-05,
      "loss": 1.5447,
      "step": 1760
    },
    {
      "epoch": 0.1903166540581433,
      "grad_norm": 0.3392559289932251,
      "learning_rate": 8e-05,
      "loss": 1.4894,
      "step": 1761
    },
    {
      "epoch": 0.1904247271155301,
      "grad_norm": 0.3604184091091156,
      "learning_rate": 8e-05,
      "loss": 1.8812,
      "step": 1762
    },
    {
      "epoch": 0.1905328001729169,
      "grad_norm": 0.35275551676750183,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 1763
    },
    {
      "epoch": 0.19064087323030368,
      "grad_norm": 0.3618411123752594,
      "learning_rate": 8e-05,
      "loss": 1.6833,
      "step": 1764
    },
    {
      "epoch": 0.19074894628769049,
      "grad_norm": 0.3477647602558136,
      "learning_rate": 8e-05,
      "loss": 1.8296,
      "step": 1765
    },
    {
      "epoch": 0.19085701934507726,
      "grad_norm": 0.3734968304634094,
      "learning_rate": 8e-05,
      "loss": 1.8205,
      "step": 1766
    },
    {
      "epoch": 0.19096509240246407,
      "grad_norm": 0.3685232996940613,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 1767
    },
    {
      "epoch": 0.19107316545985087,
      "grad_norm": 0.4022296667098999,
      "learning_rate": 8e-05,
      "loss": 1.8617,
      "step": 1768
    },
    {
      "epoch": 0.19118123851723764,
      "grad_norm": 0.3684019446372986,
      "learning_rate": 8e-05,
      "loss": 1.865,
      "step": 1769
    },
    {
      "epoch": 0.19128931157462445,
      "grad_norm": 0.3642842471599579,
      "learning_rate": 8e-05,
      "loss": 1.8401,
      "step": 1770
    },
    {
      "epoch": 0.19139738463201125,
      "grad_norm": 0.35271260142326355,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 1771
    },
    {
      "epoch": 0.19150545768939803,
      "grad_norm": 0.36002832651138306,
      "learning_rate": 8e-05,
      "loss": 1.7049,
      "step": 1772
    },
    {
      "epoch": 0.19161353074678483,
      "grad_norm": 0.33967751264572144,
      "learning_rate": 8e-05,
      "loss": 1.7104,
      "step": 1773
    },
    {
      "epoch": 0.1917216038041716,
      "grad_norm": 0.35905686020851135,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 1774
    },
    {
      "epoch": 0.1918296768615584,
      "grad_norm": 0.3566742539405823,
      "learning_rate": 8e-05,
      "loss": 1.7877,
      "step": 1775
    },
    {
      "epoch": 0.19193774991894522,
      "grad_norm": 0.35294750332832336,
      "learning_rate": 8e-05,
      "loss": 1.8371,
      "step": 1776
    },
    {
      "epoch": 0.192045822976332,
      "grad_norm": 0.3792085647583008,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 1777
    },
    {
      "epoch": 0.1921538960337188,
      "grad_norm": 0.36063823103904724,
      "learning_rate": 8e-05,
      "loss": 1.7812,
      "step": 1778
    },
    {
      "epoch": 0.1922619690911056,
      "grad_norm": 0.3558347225189209,
      "learning_rate": 8e-05,
      "loss": 1.6367,
      "step": 1779
    },
    {
      "epoch": 0.19237004214849238,
      "grad_norm": 0.39099588990211487,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 1780
    },
    {
      "epoch": 0.19247811520587918,
      "grad_norm": 0.35245659947395325,
      "learning_rate": 8e-05,
      "loss": 1.6881,
      "step": 1781
    },
    {
      "epoch": 0.19258618826326596,
      "grad_norm": 0.38026079535484314,
      "learning_rate": 8e-05,
      "loss": 1.8844,
      "step": 1782
    },
    {
      "epoch": 0.19269426132065276,
      "grad_norm": 0.35403531789779663,
      "learning_rate": 8e-05,
      "loss": 1.661,
      "step": 1783
    },
    {
      "epoch": 0.19280233437803956,
      "grad_norm": 0.3622284531593323,
      "learning_rate": 8e-05,
      "loss": 1.6176,
      "step": 1784
    },
    {
      "epoch": 0.19291040743542634,
      "grad_norm": 0.3586704432964325,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 1785
    },
    {
      "epoch": 0.19301848049281314,
      "grad_norm": 0.3440355360507965,
      "learning_rate": 8e-05,
      "loss": 1.5569,
      "step": 1786
    },
    {
      "epoch": 0.19312655355019995,
      "grad_norm": 0.3519052565097809,
      "learning_rate": 8e-05,
      "loss": 1.6826,
      "step": 1787
    },
    {
      "epoch": 0.19323462660758672,
      "grad_norm": 0.4002799689769745,
      "learning_rate": 8e-05,
      "loss": 1.9489,
      "step": 1788
    },
    {
      "epoch": 0.19334269966497353,
      "grad_norm": 0.37021762132644653,
      "learning_rate": 8e-05,
      "loss": 1.8333,
      "step": 1789
    },
    {
      "epoch": 0.1934507727223603,
      "grad_norm": 0.37534624338150024,
      "learning_rate": 8e-05,
      "loss": 1.8305,
      "step": 1790
    },
    {
      "epoch": 0.1935588457797471,
      "grad_norm": 0.35690805315971375,
      "learning_rate": 8e-05,
      "loss": 1.5306,
      "step": 1791
    },
    {
      "epoch": 0.1936669188371339,
      "grad_norm": 0.3545568585395813,
      "learning_rate": 8e-05,
      "loss": 1.6562,
      "step": 1792
    },
    {
      "epoch": 0.1937749918945207,
      "grad_norm": 0.3569551706314087,
      "learning_rate": 8e-05,
      "loss": 1.7533,
      "step": 1793
    },
    {
      "epoch": 0.1938830649519075,
      "grad_norm": 0.4669347405433655,
      "learning_rate": 8e-05,
      "loss": 1.8666,
      "step": 1794
    },
    {
      "epoch": 0.1939911380092943,
      "grad_norm": 0.3874759078025818,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 1795
    },
    {
      "epoch": 0.19409921106668107,
      "grad_norm": 0.3388046622276306,
      "learning_rate": 8e-05,
      "loss": 1.6284,
      "step": 1796
    },
    {
      "epoch": 0.19420728412406787,
      "grad_norm": 0.3450491726398468,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 1797
    },
    {
      "epoch": 0.19431535718145465,
      "grad_norm": 0.36635780334472656,
      "learning_rate": 8e-05,
      "loss": 1.8437,
      "step": 1798
    },
    {
      "epoch": 0.19442343023884145,
      "grad_norm": 0.3815009295940399,
      "learning_rate": 8e-05,
      "loss": 1.7591,
      "step": 1799
    },
    {
      "epoch": 0.19453150329622826,
      "grad_norm": 0.3911726176738739,
      "learning_rate": 8e-05,
      "loss": 1.9081,
      "step": 1800
    },
    {
      "epoch": 0.19463957635361503,
      "grad_norm": 0.39948198199272156,
      "learning_rate": 8e-05,
      "loss": 1.8739,
      "step": 1801
    },
    {
      "epoch": 0.19474764941100184,
      "grad_norm": 0.3867413401603699,
      "learning_rate": 8e-05,
      "loss": 1.716,
      "step": 1802
    },
    {
      "epoch": 0.19485572246838864,
      "grad_norm": 0.4122034013271332,
      "learning_rate": 8e-05,
      "loss": 1.9253,
      "step": 1803
    },
    {
      "epoch": 0.19496379552577542,
      "grad_norm": 0.34969455003738403,
      "learning_rate": 8e-05,
      "loss": 1.5936,
      "step": 1804
    },
    {
      "epoch": 0.19507186858316222,
      "grad_norm": 0.35876497626304626,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 1805
    },
    {
      "epoch": 0.195179941640549,
      "grad_norm": 0.33905214071273804,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1806
    },
    {
      "epoch": 0.1952880146979358,
      "grad_norm": 0.35875120759010315,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 1807
    },
    {
      "epoch": 0.1953960877553226,
      "grad_norm": 0.3817675709724426,
      "learning_rate": 8e-05,
      "loss": 1.8758,
      "step": 1808
    },
    {
      "epoch": 0.19550416081270938,
      "grad_norm": 0.37900206446647644,
      "learning_rate": 8e-05,
      "loss": 1.8086,
      "step": 1809
    },
    {
      "epoch": 0.19561223387009619,
      "grad_norm": 0.38451164960861206,
      "learning_rate": 8e-05,
      "loss": 1.9758,
      "step": 1810
    },
    {
      "epoch": 0.195720306927483,
      "grad_norm": 0.36660683155059814,
      "learning_rate": 8e-05,
      "loss": 1.9801,
      "step": 1811
    },
    {
      "epoch": 0.19582837998486977,
      "grad_norm": 0.37666767835617065,
      "learning_rate": 8e-05,
      "loss": 1.6089,
      "step": 1812
    },
    {
      "epoch": 0.19593645304225657,
      "grad_norm": 0.3671923279762268,
      "learning_rate": 8e-05,
      "loss": 1.4948,
      "step": 1813
    },
    {
      "epoch": 0.19604452609964335,
      "grad_norm": 0.3429158627986908,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 1814
    },
    {
      "epoch": 0.19615259915703015,
      "grad_norm": 0.3746236562728882,
      "learning_rate": 8e-05,
      "loss": 1.8111,
      "step": 1815
    },
    {
      "epoch": 0.19626067221441695,
      "grad_norm": 0.3776658773422241,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 1816
    },
    {
      "epoch": 0.19636874527180373,
      "grad_norm": 0.4022580087184906,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 1817
    },
    {
      "epoch": 0.19647681832919053,
      "grad_norm": 0.40486854314804077,
      "learning_rate": 8e-05,
      "loss": 1.9984,
      "step": 1818
    },
    {
      "epoch": 0.19658489138657734,
      "grad_norm": 0.3534870445728302,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 1819
    },
    {
      "epoch": 0.1966929644439641,
      "grad_norm": 0.35138943791389465,
      "learning_rate": 8e-05,
      "loss": 1.6122,
      "step": 1820
    },
    {
      "epoch": 0.19680103750135092,
      "grad_norm": 0.36506959795951843,
      "learning_rate": 8e-05,
      "loss": 1.8256,
      "step": 1821
    },
    {
      "epoch": 0.1969091105587377,
      "grad_norm": 0.40358293056488037,
      "learning_rate": 8e-05,
      "loss": 2.0094,
      "step": 1822
    },
    {
      "epoch": 0.1970171836161245,
      "grad_norm": 0.3913581073284149,
      "learning_rate": 8e-05,
      "loss": 1.6099,
      "step": 1823
    },
    {
      "epoch": 0.1971252566735113,
      "grad_norm": 0.3925180435180664,
      "learning_rate": 8e-05,
      "loss": 1.9491,
      "step": 1824
    },
    {
      "epoch": 0.19723332973089808,
      "grad_norm": 0.3530104160308838,
      "learning_rate": 8e-05,
      "loss": 1.6304,
      "step": 1825
    },
    {
      "epoch": 0.19734140278828488,
      "grad_norm": 0.34259703755378723,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 1826
    },
    {
      "epoch": 0.19744947584567168,
      "grad_norm": 0.3622263967990875,
      "learning_rate": 8e-05,
      "loss": 1.8077,
      "step": 1827
    },
    {
      "epoch": 0.19755754890305846,
      "grad_norm": 0.39186692237854004,
      "learning_rate": 8e-05,
      "loss": 1.8964,
      "step": 1828
    },
    {
      "epoch": 0.19766562196044526,
      "grad_norm": 0.37032946944236755,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 1829
    },
    {
      "epoch": 0.19777369501783204,
      "grad_norm": 0.3508176803588867,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 1830
    },
    {
      "epoch": 0.19788176807521884,
      "grad_norm": 0.3399111330509186,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 1831
    },
    {
      "epoch": 0.19798984113260565,
      "grad_norm": 0.37602466344833374,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 1832
    },
    {
      "epoch": 0.19809791418999242,
      "grad_norm": 0.379396915435791,
      "learning_rate": 8e-05,
      "loss": 1.8423,
      "step": 1833
    },
    {
      "epoch": 0.19820598724737923,
      "grad_norm": 0.3804347813129425,
      "learning_rate": 8e-05,
      "loss": 1.5954,
      "step": 1834
    },
    {
      "epoch": 0.19831406030476603,
      "grad_norm": 0.3987898528575897,
      "learning_rate": 8e-05,
      "loss": 1.9973,
      "step": 1835
    },
    {
      "epoch": 0.1984221333621528,
      "grad_norm": 0.3639991879463196,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 1836
    },
    {
      "epoch": 0.1985302064195396,
      "grad_norm": 0.3917749524116516,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 1837
    },
    {
      "epoch": 0.19863827947692642,
      "grad_norm": 0.38116833567619324,
      "learning_rate": 8e-05,
      "loss": 1.8685,
      "step": 1838
    },
    {
      "epoch": 0.1987463525343132,
      "grad_norm": 0.34580197930336,
      "learning_rate": 8e-05,
      "loss": 1.7254,
      "step": 1839
    },
    {
      "epoch": 0.1988544255917,
      "grad_norm": 0.36838003993034363,
      "learning_rate": 8e-05,
      "loss": 1.955,
      "step": 1840
    },
    {
      "epoch": 0.19896249864908677,
      "grad_norm": 0.40240806341171265,
      "learning_rate": 8e-05,
      "loss": 1.8471,
      "step": 1841
    },
    {
      "epoch": 0.19907057170647358,
      "grad_norm": 0.3656959533691406,
      "learning_rate": 8e-05,
      "loss": 1.7889,
      "step": 1842
    },
    {
      "epoch": 0.19917864476386038,
      "grad_norm": 0.34446683526039124,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 1843
    },
    {
      "epoch": 0.19928671782124716,
      "grad_norm": 0.3879225552082062,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 1844
    },
    {
      "epoch": 0.19939479087863396,
      "grad_norm": 0.369751900434494,
      "learning_rate": 8e-05,
      "loss": 1.8596,
      "step": 1845
    },
    {
      "epoch": 0.19950286393602076,
      "grad_norm": 0.39010390639305115,
      "learning_rate": 8e-05,
      "loss": 1.8732,
      "step": 1846
    },
    {
      "epoch": 0.19961093699340754,
      "grad_norm": 0.3497869074344635,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 1847
    },
    {
      "epoch": 0.19971901005079434,
      "grad_norm": 0.352588415145874,
      "learning_rate": 8e-05,
      "loss": 1.7256,
      "step": 1848
    },
    {
      "epoch": 0.19982708310818112,
      "grad_norm": 0.3593430817127228,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 1849
    },
    {
      "epoch": 0.19993515616556792,
      "grad_norm": 0.3479427397251129,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 1850
    },
    {
      "epoch": 0.20004322922295473,
      "grad_norm": 0.3976190686225891,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 1851
    },
    {
      "epoch": 0.2001513022803415,
      "grad_norm": 0.36597055196762085,
      "learning_rate": 8e-05,
      "loss": 1.5968,
      "step": 1852
    },
    {
      "epoch": 0.2002593753377283,
      "grad_norm": 0.37614214420318604,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 1853
    },
    {
      "epoch": 0.2003674483951151,
      "grad_norm": 0.36254647374153137,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 1854
    },
    {
      "epoch": 0.2004755214525019,
      "grad_norm": 0.3588215112686157,
      "learning_rate": 8e-05,
      "loss": 1.7798,
      "step": 1855
    },
    {
      "epoch": 0.2005835945098887,
      "grad_norm": 0.33507752418518066,
      "learning_rate": 8e-05,
      "loss": 1.656,
      "step": 1856
    },
    {
      "epoch": 0.20069166756727547,
      "grad_norm": 0.3605280816555023,
      "learning_rate": 8e-05,
      "loss": 1.7644,
      "step": 1857
    },
    {
      "epoch": 0.20079974062466227,
      "grad_norm": 0.3604772984981537,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 1858
    },
    {
      "epoch": 0.20090781368204907,
      "grad_norm": 0.38815072178840637,
      "learning_rate": 8e-05,
      "loss": 1.9593,
      "step": 1859
    },
    {
      "epoch": 0.20101588673943585,
      "grad_norm": 0.37124311923980713,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 1860
    },
    {
      "epoch": 0.20112395979682265,
      "grad_norm": 0.34862321615219116,
      "learning_rate": 8e-05,
      "loss": 1.7519,
      "step": 1861
    },
    {
      "epoch": 0.20123203285420946,
      "grad_norm": 0.35736700892448425,
      "learning_rate": 8e-05,
      "loss": 1.7361,
      "step": 1862
    },
    {
      "epoch": 0.20134010591159623,
      "grad_norm": 0.3790404200553894,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 1863
    },
    {
      "epoch": 0.20144817896898304,
      "grad_norm": 0.35792702436447144,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 1864
    },
    {
      "epoch": 0.20155625202636981,
      "grad_norm": 0.3490006625652313,
      "learning_rate": 8e-05,
      "loss": 1.7612,
      "step": 1865
    },
    {
      "epoch": 0.20166432508375662,
      "grad_norm": 0.3531262278556824,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 1866
    },
    {
      "epoch": 0.20177239814114342,
      "grad_norm": 0.38021284341812134,
      "learning_rate": 8e-05,
      "loss": 1.8541,
      "step": 1867
    },
    {
      "epoch": 0.2018804711985302,
      "grad_norm": 0.3699413537979126,
      "learning_rate": 8e-05,
      "loss": 1.92,
      "step": 1868
    },
    {
      "epoch": 0.201988544255917,
      "grad_norm": 0.3971005380153656,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 1869
    },
    {
      "epoch": 0.2020966173133038,
      "grad_norm": 0.36012476682662964,
      "learning_rate": 8e-05,
      "loss": 1.7661,
      "step": 1870
    },
    {
      "epoch": 0.20220469037069058,
      "grad_norm": 0.36561453342437744,
      "learning_rate": 8e-05,
      "loss": 1.6984,
      "step": 1871
    },
    {
      "epoch": 0.20231276342807739,
      "grad_norm": 0.3798507750034332,
      "learning_rate": 8e-05,
      "loss": 1.8763,
      "step": 1872
    },
    {
      "epoch": 0.20242083648546416,
      "grad_norm": 0.3517743945121765,
      "learning_rate": 8e-05,
      "loss": 1.7786,
      "step": 1873
    },
    {
      "epoch": 0.20252890954285097,
      "grad_norm": 0.406732439994812,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 1874
    },
    {
      "epoch": 0.20263698260023777,
      "grad_norm": 0.38151341676712036,
      "learning_rate": 8e-05,
      "loss": 1.8307,
      "step": 1875
    },
    {
      "epoch": 0.20274505565762455,
      "grad_norm": 0.3487846851348877,
      "learning_rate": 8e-05,
      "loss": 1.8087,
      "step": 1876
    },
    {
      "epoch": 0.20285312871501135,
      "grad_norm": 0.37762340903282166,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 1877
    },
    {
      "epoch": 0.20296120177239815,
      "grad_norm": 0.386909544467926,
      "learning_rate": 8e-05,
      "loss": 1.7993,
      "step": 1878
    },
    {
      "epoch": 0.20306927482978493,
      "grad_norm": 0.3647140562534332,
      "learning_rate": 8e-05,
      "loss": 1.7843,
      "step": 1879
    },
    {
      "epoch": 0.20317734788717173,
      "grad_norm": 0.3647440969944,
      "learning_rate": 8e-05,
      "loss": 1.8405,
      "step": 1880
    },
    {
      "epoch": 0.2032854209445585,
      "grad_norm": 0.36391812562942505,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 1881
    },
    {
      "epoch": 0.2033934940019453,
      "grad_norm": 0.384227991104126,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 1882
    },
    {
      "epoch": 0.20350156705933212,
      "grad_norm": 0.36254239082336426,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 1883
    },
    {
      "epoch": 0.2036096401167189,
      "grad_norm": 0.398166686296463,
      "learning_rate": 8e-05,
      "loss": 1.9578,
      "step": 1884
    },
    {
      "epoch": 0.2037177131741057,
      "grad_norm": 0.3709028363227844,
      "learning_rate": 8e-05,
      "loss": 1.829,
      "step": 1885
    },
    {
      "epoch": 0.2038257862314925,
      "grad_norm": 0.35031962394714355,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 1886
    },
    {
      "epoch": 0.20393385928887928,
      "grad_norm": 0.35780707001686096,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 1887
    },
    {
      "epoch": 0.20404193234626608,
      "grad_norm": 0.3482782244682312,
      "learning_rate": 8e-05,
      "loss": 1.5845,
      "step": 1888
    },
    {
      "epoch": 0.20415000540365286,
      "grad_norm": 0.3618932366371155,
      "learning_rate": 8e-05,
      "loss": 1.7825,
      "step": 1889
    },
    {
      "epoch": 0.20425807846103966,
      "grad_norm": 0.35274481773376465,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 1890
    },
    {
      "epoch": 0.20436615151842646,
      "grad_norm": 0.41224750876426697,
      "learning_rate": 8e-05,
      "loss": 1.784,
      "step": 1891
    },
    {
      "epoch": 0.20447422457581324,
      "grad_norm": 0.3973805606365204,
      "learning_rate": 8e-05,
      "loss": 1.872,
      "step": 1892
    },
    {
      "epoch": 0.20458229763320004,
      "grad_norm": 0.4106702208518982,
      "learning_rate": 8e-05,
      "loss": 1.9697,
      "step": 1893
    },
    {
      "epoch": 0.20469037069058685,
      "grad_norm": 0.38575512170791626,
      "learning_rate": 8e-05,
      "loss": 1.9009,
      "step": 1894
    },
    {
      "epoch": 0.20479844374797362,
      "grad_norm": 0.38011008501052856,
      "learning_rate": 8e-05,
      "loss": 1.7581,
      "step": 1895
    },
    {
      "epoch": 0.20490651680536043,
      "grad_norm": 0.3502074182033539,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 1896
    },
    {
      "epoch": 0.2050145898627472,
      "grad_norm": 0.3682728409767151,
      "learning_rate": 8e-05,
      "loss": 1.8756,
      "step": 1897
    },
    {
      "epoch": 0.205122662920134,
      "grad_norm": 0.37342748045921326,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 1898
    },
    {
      "epoch": 0.2052307359775208,
      "grad_norm": 0.365245521068573,
      "learning_rate": 8e-05,
      "loss": 1.6766,
      "step": 1899
    },
    {
      "epoch": 0.2053388090349076,
      "grad_norm": 0.3619349002838135,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 1900
    },
    {
      "epoch": 0.2054468820922944,
      "grad_norm": 0.384347528219223,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 1901
    },
    {
      "epoch": 0.2055549551496812,
      "grad_norm": 0.35721322894096375,
      "learning_rate": 8e-05,
      "loss": 1.7859,
      "step": 1902
    },
    {
      "epoch": 0.20566302820706797,
      "grad_norm": 0.3474719524383545,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 1903
    },
    {
      "epoch": 0.20577110126445478,
      "grad_norm": 0.3767649829387665,
      "learning_rate": 8e-05,
      "loss": 1.8515,
      "step": 1904
    },
    {
      "epoch": 0.20587917432184155,
      "grad_norm": 0.34641364216804504,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 1905
    },
    {
      "epoch": 0.20598724737922836,
      "grad_norm": 0.34312519431114197,
      "learning_rate": 8e-05,
      "loss": 1.6293,
      "step": 1906
    },
    {
      "epoch": 0.20609532043661516,
      "grad_norm": 0.3383251130580902,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 1907
    },
    {
      "epoch": 0.20620339349400194,
      "grad_norm": 0.35973063111305237,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 1908
    },
    {
      "epoch": 0.20631146655138874,
      "grad_norm": 0.3731791377067566,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 1909
    },
    {
      "epoch": 0.20641953960877554,
      "grad_norm": 0.3464277684688568,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 1910
    },
    {
      "epoch": 0.20652761266616232,
      "grad_norm": 0.3760031461715698,
      "learning_rate": 8e-05,
      "loss": 1.9154,
      "step": 1911
    },
    {
      "epoch": 0.20663568572354912,
      "grad_norm": 0.34432798624038696,
      "learning_rate": 8e-05,
      "loss": 1.4793,
      "step": 1912
    },
    {
      "epoch": 0.2067437587809359,
      "grad_norm": 0.3558083176612854,
      "learning_rate": 8e-05,
      "loss": 1.5359,
      "step": 1913
    },
    {
      "epoch": 0.2068518318383227,
      "grad_norm": 0.35777437686920166,
      "learning_rate": 8e-05,
      "loss": 1.747,
      "step": 1914
    },
    {
      "epoch": 0.2069599048957095,
      "grad_norm": 0.38074642419815063,
      "learning_rate": 8e-05,
      "loss": 1.8046,
      "step": 1915
    },
    {
      "epoch": 0.20706797795309628,
      "grad_norm": 0.34945687651634216,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 1916
    },
    {
      "epoch": 0.2071760510104831,
      "grad_norm": 0.38633808493614197,
      "learning_rate": 8e-05,
      "loss": 1.7733,
      "step": 1917
    },
    {
      "epoch": 0.2072841240678699,
      "grad_norm": 0.40135735273361206,
      "learning_rate": 8e-05,
      "loss": 1.8261,
      "step": 1918
    },
    {
      "epoch": 0.20739219712525667,
      "grad_norm": 0.34329578280448914,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 1919
    },
    {
      "epoch": 0.20750027018264347,
      "grad_norm": 0.38571685552597046,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 1920
    },
    {
      "epoch": 0.20760834324003027,
      "grad_norm": 0.3572600483894348,
      "learning_rate": 8e-05,
      "loss": 1.6594,
      "step": 1921
    },
    {
      "epoch": 0.20771641629741705,
      "grad_norm": 0.35564470291137695,
      "learning_rate": 8e-05,
      "loss": 1.6699,
      "step": 1922
    },
    {
      "epoch": 0.20782448935480385,
      "grad_norm": 0.37275245785713196,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 1923
    },
    {
      "epoch": 0.20793256241219063,
      "grad_norm": 0.39318376779556274,
      "learning_rate": 8e-05,
      "loss": 2.0055,
      "step": 1924
    },
    {
      "epoch": 0.20804063546957743,
      "grad_norm": 0.4063023328781128,
      "learning_rate": 8e-05,
      "loss": 1.8252,
      "step": 1925
    },
    {
      "epoch": 0.20814870852696424,
      "grad_norm": 0.3309997618198395,
      "learning_rate": 8e-05,
      "loss": 1.5401,
      "step": 1926
    },
    {
      "epoch": 0.20825678158435101,
      "grad_norm": 0.38798418641090393,
      "learning_rate": 8e-05,
      "loss": 1.771,
      "step": 1927
    },
    {
      "epoch": 0.20836485464173782,
      "grad_norm": 0.36899039149284363,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 1928
    },
    {
      "epoch": 0.20847292769912462,
      "grad_norm": 0.36819252371788025,
      "learning_rate": 8e-05,
      "loss": 1.7692,
      "step": 1929
    },
    {
      "epoch": 0.2085810007565114,
      "grad_norm": 0.36590033769607544,
      "learning_rate": 8e-05,
      "loss": 1.823,
      "step": 1930
    },
    {
      "epoch": 0.2086890738138982,
      "grad_norm": 0.3395880162715912,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 1931
    },
    {
      "epoch": 0.20879714687128498,
      "grad_norm": 0.3607260286808014,
      "learning_rate": 8e-05,
      "loss": 1.7543,
      "step": 1932
    },
    {
      "epoch": 0.20890521992867178,
      "grad_norm": 0.37475481629371643,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 1933
    },
    {
      "epoch": 0.20901329298605859,
      "grad_norm": 0.42187097668647766,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 1934
    },
    {
      "epoch": 0.20912136604344536,
      "grad_norm": 0.3628307580947876,
      "learning_rate": 8e-05,
      "loss": 1.9118,
      "step": 1935
    },
    {
      "epoch": 0.20922943910083217,
      "grad_norm": 0.3630234897136688,
      "learning_rate": 8e-05,
      "loss": 1.7933,
      "step": 1936
    },
    {
      "epoch": 0.20933751215821897,
      "grad_norm": 0.37966346740722656,
      "learning_rate": 8e-05,
      "loss": 1.7856,
      "step": 1937
    },
    {
      "epoch": 0.20944558521560575,
      "grad_norm": 0.4155793786048889,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 1938
    },
    {
      "epoch": 0.20955365827299255,
      "grad_norm": 0.33756765723228455,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 1939
    },
    {
      "epoch": 0.20966173133037933,
      "grad_norm": 0.373968243598938,
      "learning_rate": 8e-05,
      "loss": 1.7968,
      "step": 1940
    },
    {
      "epoch": 0.20976980438776613,
      "grad_norm": 0.3583020865917206,
      "learning_rate": 8e-05,
      "loss": 1.6967,
      "step": 1941
    },
    {
      "epoch": 0.20987787744515293,
      "grad_norm": 0.32911446690559387,
      "learning_rate": 8e-05,
      "loss": 1.5576,
      "step": 1942
    },
    {
      "epoch": 0.2099859505025397,
      "grad_norm": 0.3314995765686035,
      "learning_rate": 8e-05,
      "loss": 1.5973,
      "step": 1943
    },
    {
      "epoch": 0.2100940235599265,
      "grad_norm": 0.36070725321769714,
      "learning_rate": 8e-05,
      "loss": 1.6226,
      "step": 1944
    },
    {
      "epoch": 0.21020209661731332,
      "grad_norm": 0.4008345901966095,
      "learning_rate": 8e-05,
      "loss": 1.9592,
      "step": 1945
    },
    {
      "epoch": 0.2103101696747001,
      "grad_norm": 0.358479768037796,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 1946
    },
    {
      "epoch": 0.2104182427320869,
      "grad_norm": 0.3629383444786072,
      "learning_rate": 8e-05,
      "loss": 1.6467,
      "step": 1947
    },
    {
      "epoch": 0.21052631578947367,
      "grad_norm": 0.4131973385810852,
      "learning_rate": 8e-05,
      "loss": 1.8011,
      "step": 1948
    },
    {
      "epoch": 0.21063438884686048,
      "grad_norm": 0.43551886081695557,
      "learning_rate": 8e-05,
      "loss": 1.9769,
      "step": 1949
    },
    {
      "epoch": 0.21074246190424728,
      "grad_norm": 0.3896951675415039,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 1950
    },
    {
      "epoch": 0.21085053496163406,
      "grad_norm": 0.35681670904159546,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 1951
    },
    {
      "epoch": 0.21095860801902086,
      "grad_norm": 0.40379849076271057,
      "learning_rate": 8e-05,
      "loss": 1.8967,
      "step": 1952
    },
    {
      "epoch": 0.21106668107640766,
      "grad_norm": 0.35377737879753113,
      "learning_rate": 8e-05,
      "loss": 1.5626,
      "step": 1953
    },
    {
      "epoch": 0.21117475413379444,
      "grad_norm": 0.3624890446662903,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 1954
    },
    {
      "epoch": 0.21128282719118124,
      "grad_norm": 0.3532603979110718,
      "learning_rate": 8e-05,
      "loss": 1.4725,
      "step": 1955
    },
    {
      "epoch": 0.21139090024856802,
      "grad_norm": 0.3497554659843445,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 1956
    },
    {
      "epoch": 0.21149897330595482,
      "grad_norm": 0.3954872190952301,
      "learning_rate": 8e-05,
      "loss": 1.8425,
      "step": 1957
    },
    {
      "epoch": 0.21160704636334163,
      "grad_norm": 0.3571591079235077,
      "learning_rate": 8e-05,
      "loss": 1.6638,
      "step": 1958
    },
    {
      "epoch": 0.2117151194207284,
      "grad_norm": 0.38056448101997375,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 1959
    },
    {
      "epoch": 0.2118231924781152,
      "grad_norm": 0.4163707196712494,
      "learning_rate": 8e-05,
      "loss": 1.7648,
      "step": 1960
    },
    {
      "epoch": 0.211931265535502,
      "grad_norm": 0.376526415348053,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 1961
    },
    {
      "epoch": 0.2120393385928888,
      "grad_norm": 0.3950906991958618,
      "learning_rate": 8e-05,
      "loss": 1.795,
      "step": 1962
    },
    {
      "epoch": 0.2121474116502756,
      "grad_norm": 0.39837372303009033,
      "learning_rate": 8e-05,
      "loss": 1.8725,
      "step": 1963
    },
    {
      "epoch": 0.21225548470766237,
      "grad_norm": 0.41050609946250916,
      "learning_rate": 8e-05,
      "loss": 1.8765,
      "step": 1964
    },
    {
      "epoch": 0.21236355776504917,
      "grad_norm": 0.35144171118736267,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 1965
    },
    {
      "epoch": 0.21247163082243598,
      "grad_norm": 0.3683284819126129,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 1966
    },
    {
      "epoch": 0.21257970387982275,
      "grad_norm": 0.38534876704216003,
      "learning_rate": 8e-05,
      "loss": 1.9354,
      "step": 1967
    },
    {
      "epoch": 0.21268777693720956,
      "grad_norm": 0.38828667998313904,
      "learning_rate": 8e-05,
      "loss": 1.8566,
      "step": 1968
    },
    {
      "epoch": 0.21279584999459636,
      "grad_norm": 0.4068783223628998,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 1969
    },
    {
      "epoch": 0.21290392305198313,
      "grad_norm": 0.3801378011703491,
      "learning_rate": 8e-05,
      "loss": 1.8297,
      "step": 1970
    },
    {
      "epoch": 0.21301199610936994,
      "grad_norm": 0.3865795135498047,
      "learning_rate": 8e-05,
      "loss": 1.7451,
      "step": 1971
    },
    {
      "epoch": 0.21312006916675671,
      "grad_norm": 0.357890248298645,
      "learning_rate": 8e-05,
      "loss": 1.7743,
      "step": 1972
    },
    {
      "epoch": 0.21322814222414352,
      "grad_norm": 0.3626871109008789,
      "learning_rate": 8e-05,
      "loss": 1.5904,
      "step": 1973
    },
    {
      "epoch": 0.21333621528153032,
      "grad_norm": 0.36060309410095215,
      "learning_rate": 8e-05,
      "loss": 1.7338,
      "step": 1974
    },
    {
      "epoch": 0.2134442883389171,
      "grad_norm": 0.3594733774662018,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 1975
    },
    {
      "epoch": 0.2135523613963039,
      "grad_norm": 0.3680800795555115,
      "learning_rate": 8e-05,
      "loss": 1.6201,
      "step": 1976
    },
    {
      "epoch": 0.2136604344536907,
      "grad_norm": 0.3860113322734833,
      "learning_rate": 8e-05,
      "loss": 1.9166,
      "step": 1977
    },
    {
      "epoch": 0.21376850751107748,
      "grad_norm": 0.3709617853164673,
      "learning_rate": 8e-05,
      "loss": 1.8738,
      "step": 1978
    },
    {
      "epoch": 0.2138765805684643,
      "grad_norm": 0.37274497747421265,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 1979
    },
    {
      "epoch": 0.21398465362585106,
      "grad_norm": 0.3914092481136322,
      "learning_rate": 8e-05,
      "loss": 1.8579,
      "step": 1980
    },
    {
      "epoch": 0.21409272668323787,
      "grad_norm": 0.40114396810531616,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 1981
    },
    {
      "epoch": 0.21420079974062467,
      "grad_norm": 0.36645975708961487,
      "learning_rate": 8e-05,
      "loss": 1.7846,
      "step": 1982
    },
    {
      "epoch": 0.21430887279801145,
      "grad_norm": 0.36580151319503784,
      "learning_rate": 8e-05,
      "loss": 1.8357,
      "step": 1983
    },
    {
      "epoch": 0.21441694585539825,
      "grad_norm": 0.3511945605278015,
      "learning_rate": 8e-05,
      "loss": 1.6754,
      "step": 1984
    },
    {
      "epoch": 0.21452501891278505,
      "grad_norm": 0.3777111768722534,
      "learning_rate": 8e-05,
      "loss": 1.8998,
      "step": 1985
    },
    {
      "epoch": 0.21463309197017183,
      "grad_norm": 0.33652061223983765,
      "learning_rate": 8e-05,
      "loss": 1.7839,
      "step": 1986
    },
    {
      "epoch": 0.21474116502755863,
      "grad_norm": 0.3497997522354126,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 1987
    },
    {
      "epoch": 0.2148492380849454,
      "grad_norm": 0.3635331392288208,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 1988
    },
    {
      "epoch": 0.2149573111423322,
      "grad_norm": 0.3440324366092682,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 1989
    },
    {
      "epoch": 0.21506538419971902,
      "grad_norm": 0.37369877099990845,
      "learning_rate": 8e-05,
      "loss": 1.7509,
      "step": 1990
    },
    {
      "epoch": 0.2151734572571058,
      "grad_norm": 0.35939815640449524,
      "learning_rate": 8e-05,
      "loss": 1.6149,
      "step": 1991
    },
    {
      "epoch": 0.2152815303144926,
      "grad_norm": 0.37687626481056213,
      "learning_rate": 8e-05,
      "loss": 1.747,
      "step": 1992
    },
    {
      "epoch": 0.2153896033718794,
      "grad_norm": 0.4100135862827301,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 1993
    },
    {
      "epoch": 0.21549767642926618,
      "grad_norm": 0.3928000032901764,
      "learning_rate": 8e-05,
      "loss": 1.9531,
      "step": 1994
    },
    {
      "epoch": 0.21560574948665298,
      "grad_norm": 0.36357176303863525,
      "learning_rate": 8e-05,
      "loss": 1.8323,
      "step": 1995
    },
    {
      "epoch": 0.21571382254403976,
      "grad_norm": 0.40489187836647034,
      "learning_rate": 8e-05,
      "loss": 1.8864,
      "step": 1996
    },
    {
      "epoch": 0.21582189560142656,
      "grad_norm": 0.3513606786727905,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 1997
    },
    {
      "epoch": 0.21592996865881336,
      "grad_norm": 0.3592759370803833,
      "learning_rate": 8e-05,
      "loss": 1.7024,
      "step": 1998
    },
    {
      "epoch": 0.21603804171620014,
      "grad_norm": 0.3832401931285858,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 1999
    },
    {
      "epoch": 0.21614611477358694,
      "grad_norm": 0.35246819257736206,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 2000
    },
    {
      "epoch": 0.21625418783097375,
      "grad_norm": 0.36942264437675476,
      "learning_rate": 8e-05,
      "loss": 1.8037,
      "step": 2001
    },
    {
      "epoch": 0.21636226088836052,
      "grad_norm": 0.3643299341201782,
      "learning_rate": 8e-05,
      "loss": 1.6774,
      "step": 2002
    },
    {
      "epoch": 0.21647033394574733,
      "grad_norm": 0.35385361313819885,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 2003
    },
    {
      "epoch": 0.21657840700313413,
      "grad_norm": 0.36905983090400696,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 2004
    },
    {
      "epoch": 0.2166864800605209,
      "grad_norm": 0.3492202162742615,
      "learning_rate": 8e-05,
      "loss": 1.6326,
      "step": 2005
    },
    {
      "epoch": 0.2167945531179077,
      "grad_norm": 0.4036939740180969,
      "learning_rate": 8e-05,
      "loss": 1.8589,
      "step": 2006
    },
    {
      "epoch": 0.2169026261752945,
      "grad_norm": 0.3529248833656311,
      "learning_rate": 8e-05,
      "loss": 1.7494,
      "step": 2007
    },
    {
      "epoch": 0.2170106992326813,
      "grad_norm": 0.3524319529533386,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 2008
    },
    {
      "epoch": 0.2171187722900681,
      "grad_norm": 0.3989529311656952,
      "learning_rate": 8e-05,
      "loss": 1.7791,
      "step": 2009
    },
    {
      "epoch": 0.21722684534745487,
      "grad_norm": 0.3952706754207611,
      "learning_rate": 8e-05,
      "loss": 1.7964,
      "step": 2010
    },
    {
      "epoch": 0.21733491840484168,
      "grad_norm": 0.35971590876579285,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 2011
    },
    {
      "epoch": 0.21744299146222848,
      "grad_norm": 0.4018203020095825,
      "learning_rate": 8e-05,
      "loss": 1.797,
      "step": 2012
    },
    {
      "epoch": 0.21755106451961526,
      "grad_norm": 0.38853293657302856,
      "learning_rate": 8e-05,
      "loss": 1.6688,
      "step": 2013
    },
    {
      "epoch": 0.21765913757700206,
      "grad_norm": 0.3590441644191742,
      "learning_rate": 8e-05,
      "loss": 1.4683,
      "step": 2014
    },
    {
      "epoch": 0.21776721063438884,
      "grad_norm": 0.37658265233039856,
      "learning_rate": 8e-05,
      "loss": 1.6729,
      "step": 2015
    },
    {
      "epoch": 0.21787528369177564,
      "grad_norm": 0.3557071387767792,
      "learning_rate": 8e-05,
      "loss": 1.6941,
      "step": 2016
    },
    {
      "epoch": 0.21798335674916244,
      "grad_norm": 0.37900692224502563,
      "learning_rate": 8e-05,
      "loss": 1.7138,
      "step": 2017
    },
    {
      "epoch": 0.21809142980654922,
      "grad_norm": 0.41085371375083923,
      "learning_rate": 8e-05,
      "loss": 1.8812,
      "step": 2018
    },
    {
      "epoch": 0.21819950286393602,
      "grad_norm": 0.3881458640098572,
      "learning_rate": 8e-05,
      "loss": 1.7512,
      "step": 2019
    },
    {
      "epoch": 0.21830757592132283,
      "grad_norm": 0.3675471246242523,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 2020
    },
    {
      "epoch": 0.2184156489787096,
      "grad_norm": 0.4067486524581909,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 2021
    },
    {
      "epoch": 0.2185237220360964,
      "grad_norm": 0.3796285390853882,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 2022
    },
    {
      "epoch": 0.21863179509348318,
      "grad_norm": 0.36941665410995483,
      "learning_rate": 8e-05,
      "loss": 1.6787,
      "step": 2023
    },
    {
      "epoch": 0.21873986815087,
      "grad_norm": 0.36870813369750977,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 2024
    },
    {
      "epoch": 0.2188479412082568,
      "grad_norm": 0.3542405664920807,
      "learning_rate": 8e-05,
      "loss": 1.6646,
      "step": 2025
    },
    {
      "epoch": 0.21895601426564357,
      "grad_norm": 0.37803706526756287,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 2026
    },
    {
      "epoch": 0.21906408732303037,
      "grad_norm": 0.4057185649871826,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 2027
    },
    {
      "epoch": 0.21917216038041717,
      "grad_norm": 0.39966803789138794,
      "learning_rate": 8e-05,
      "loss": 1.801,
      "step": 2028
    },
    {
      "epoch": 0.21928023343780395,
      "grad_norm": 0.36467891931533813,
      "learning_rate": 8e-05,
      "loss": 1.8367,
      "step": 2029
    },
    {
      "epoch": 0.21938830649519075,
      "grad_norm": 0.4220462441444397,
      "learning_rate": 8e-05,
      "loss": 1.7095,
      "step": 2030
    },
    {
      "epoch": 0.21949637955257753,
      "grad_norm": 0.3874486982822418,
      "learning_rate": 8e-05,
      "loss": 1.8462,
      "step": 2031
    },
    {
      "epoch": 0.21960445260996433,
      "grad_norm": 0.35149145126342773,
      "learning_rate": 8e-05,
      "loss": 1.7776,
      "step": 2032
    },
    {
      "epoch": 0.21971252566735114,
      "grad_norm": 0.366610586643219,
      "learning_rate": 8e-05,
      "loss": 1.774,
      "step": 2033
    },
    {
      "epoch": 0.21982059872473791,
      "grad_norm": 0.37655380368232727,
      "learning_rate": 8e-05,
      "loss": 1.8611,
      "step": 2034
    },
    {
      "epoch": 0.21992867178212472,
      "grad_norm": 0.39081716537475586,
      "learning_rate": 8e-05,
      "loss": 1.8509,
      "step": 2035
    },
    {
      "epoch": 0.22003674483951152,
      "grad_norm": 0.36093536019325256,
      "learning_rate": 8e-05,
      "loss": 1.5533,
      "step": 2036
    },
    {
      "epoch": 0.2201448178968983,
      "grad_norm": 0.37873879075050354,
      "learning_rate": 8e-05,
      "loss": 1.8671,
      "step": 2037
    },
    {
      "epoch": 0.2202528909542851,
      "grad_norm": 0.35921892523765564,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 2038
    },
    {
      "epoch": 0.22036096401167188,
      "grad_norm": 0.34384745359420776,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 2039
    },
    {
      "epoch": 0.22046903706905868,
      "grad_norm": 0.383731871843338,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 2040
    },
    {
      "epoch": 0.22057711012644549,
      "grad_norm": 0.3606813848018646,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 2041
    },
    {
      "epoch": 0.22068518318383226,
      "grad_norm": 0.3706587851047516,
      "learning_rate": 8e-05,
      "loss": 1.7448,
      "step": 2042
    },
    {
      "epoch": 0.22079325624121907,
      "grad_norm": 0.39703646302223206,
      "learning_rate": 8e-05,
      "loss": 1.7732,
      "step": 2043
    },
    {
      "epoch": 0.22090132929860587,
      "grad_norm": 0.37300845980644226,
      "learning_rate": 8e-05,
      "loss": 1.8311,
      "step": 2044
    },
    {
      "epoch": 0.22100940235599265,
      "grad_norm": 0.3679048418998718,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 2045
    },
    {
      "epoch": 0.22111747541337945,
      "grad_norm": 0.36237233877182007,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 2046
    },
    {
      "epoch": 0.22122554847076623,
      "grad_norm": 0.38080283999443054,
      "learning_rate": 8e-05,
      "loss": 1.6903,
      "step": 2047
    },
    {
      "epoch": 0.22133362152815303,
      "grad_norm": 0.3606734573841095,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 2048
    },
    {
      "epoch": 0.22144169458553983,
      "grad_norm": 0.36421850323677063,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 2049
    },
    {
      "epoch": 0.2215497676429266,
      "grad_norm": 0.4062765836715698,
      "learning_rate": 8e-05,
      "loss": 1.8359,
      "step": 2050
    },
    {
      "epoch": 0.2216578407003134,
      "grad_norm": 0.4163103699684143,
      "learning_rate": 8e-05,
      "loss": 1.9322,
      "step": 2051
    },
    {
      "epoch": 0.22176591375770022,
      "grad_norm": 0.3605847656726837,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 2052
    },
    {
      "epoch": 0.221873986815087,
      "grad_norm": 0.38698506355285645,
      "learning_rate": 8e-05,
      "loss": 1.8378,
      "step": 2053
    },
    {
      "epoch": 0.2219820598724738,
      "grad_norm": 0.3885916769504547,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 2054
    },
    {
      "epoch": 0.22209013292986057,
      "grad_norm": 0.3732059895992279,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 2055
    },
    {
      "epoch": 0.22219820598724738,
      "grad_norm": 0.4028705656528473,
      "learning_rate": 8e-05,
      "loss": 1.7729,
      "step": 2056
    },
    {
      "epoch": 0.22230627904463418,
      "grad_norm": 0.39209872484207153,
      "learning_rate": 8e-05,
      "loss": 1.9357,
      "step": 2057
    },
    {
      "epoch": 0.22241435210202096,
      "grad_norm": 0.3717482089996338,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 2058
    },
    {
      "epoch": 0.22252242515940776,
      "grad_norm": 0.3872828483581543,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 2059
    },
    {
      "epoch": 0.22263049821679456,
      "grad_norm": 0.36990293860435486,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 2060
    },
    {
      "epoch": 0.22273857127418134,
      "grad_norm": 0.3654635548591614,
      "learning_rate": 8e-05,
      "loss": 1.8468,
      "step": 2061
    },
    {
      "epoch": 0.22284664433156814,
      "grad_norm": 0.36631882190704346,
      "learning_rate": 8e-05,
      "loss": 1.9319,
      "step": 2062
    },
    {
      "epoch": 0.22295471738895492,
      "grad_norm": 0.37345147132873535,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 2063
    },
    {
      "epoch": 0.22306279044634172,
      "grad_norm": 0.38322269916534424,
      "learning_rate": 8e-05,
      "loss": 1.8156,
      "step": 2064
    },
    {
      "epoch": 0.22317086350372853,
      "grad_norm": 0.38158226013183594,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 2065
    },
    {
      "epoch": 0.2232789365611153,
      "grad_norm": 0.3726826608181,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 2066
    },
    {
      "epoch": 0.2233870096185021,
      "grad_norm": 0.3899937570095062,
      "learning_rate": 8e-05,
      "loss": 1.8623,
      "step": 2067
    },
    {
      "epoch": 0.2234950826758889,
      "grad_norm": 0.3797368109226227,
      "learning_rate": 8e-05,
      "loss": 1.8981,
      "step": 2068
    },
    {
      "epoch": 0.2236031557332757,
      "grad_norm": 0.37010857462882996,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 2069
    },
    {
      "epoch": 0.2237112287906625,
      "grad_norm": 0.36409837007522583,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 2070
    },
    {
      "epoch": 0.22381930184804927,
      "grad_norm": 0.42100897431373596,
      "learning_rate": 8e-05,
      "loss": 1.8354,
      "step": 2071
    },
    {
      "epoch": 0.22392737490543607,
      "grad_norm": 0.3848303258419037,
      "learning_rate": 8e-05,
      "loss": 1.8157,
      "step": 2072
    },
    {
      "epoch": 0.22403544796282288,
      "grad_norm": 0.35955747961997986,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 2073
    },
    {
      "epoch": 0.22414352102020965,
      "grad_norm": 0.3381285071372986,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 2074
    },
    {
      "epoch": 0.22425159407759646,
      "grad_norm": 0.4168195426464081,
      "learning_rate": 8e-05,
      "loss": 1.8523,
      "step": 2075
    },
    {
      "epoch": 0.22435966713498326,
      "grad_norm": 0.3455525040626526,
      "learning_rate": 8e-05,
      "loss": 1.5917,
      "step": 2076
    },
    {
      "epoch": 0.22446774019237004,
      "grad_norm": 0.3746519386768341,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 2077
    },
    {
      "epoch": 0.22457581324975684,
      "grad_norm": 0.3646126985549927,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 2078
    },
    {
      "epoch": 0.22468388630714362,
      "grad_norm": 0.4021371006965637,
      "learning_rate": 8e-05,
      "loss": 1.8259,
      "step": 2079
    },
    {
      "epoch": 0.22479195936453042,
      "grad_norm": 0.3647554814815521,
      "learning_rate": 8e-05,
      "loss": 1.7899,
      "step": 2080
    },
    {
      "epoch": 0.22490003242191722,
      "grad_norm": 0.37920671701431274,
      "learning_rate": 8e-05,
      "loss": 1.7413,
      "step": 2081
    },
    {
      "epoch": 0.225008105479304,
      "grad_norm": 0.39572465419769287,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 2082
    },
    {
      "epoch": 0.2251161785366908,
      "grad_norm": 0.4094925820827484,
      "learning_rate": 8e-05,
      "loss": 1.7248,
      "step": 2083
    },
    {
      "epoch": 0.2252242515940776,
      "grad_norm": 0.3848681151866913,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 2084
    },
    {
      "epoch": 0.22533232465146438,
      "grad_norm": 0.38531455397605896,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 2085
    },
    {
      "epoch": 0.2254403977088512,
      "grad_norm": 0.3507544994354248,
      "learning_rate": 8e-05,
      "loss": 1.8107,
      "step": 2086
    },
    {
      "epoch": 0.225548470766238,
      "grad_norm": 0.39098140597343445,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 2087
    },
    {
      "epoch": 0.22565654382362477,
      "grad_norm": 0.36397454142570496,
      "learning_rate": 8e-05,
      "loss": 1.6309,
      "step": 2088
    },
    {
      "epoch": 0.22576461688101157,
      "grad_norm": 0.4234897196292877,
      "learning_rate": 8e-05,
      "loss": 2.0841,
      "step": 2089
    },
    {
      "epoch": 0.22587268993839835,
      "grad_norm": 0.3712233006954193,
      "learning_rate": 8e-05,
      "loss": 1.7417,
      "step": 2090
    },
    {
      "epoch": 0.22598076299578515,
      "grad_norm": 0.34451958537101746,
      "learning_rate": 8e-05,
      "loss": 1.7492,
      "step": 2091
    },
    {
      "epoch": 0.22608883605317195,
      "grad_norm": 0.38284558057785034,
      "learning_rate": 8e-05,
      "loss": 1.9053,
      "step": 2092
    },
    {
      "epoch": 0.22619690911055873,
      "grad_norm": 0.35918140411376953,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 2093
    },
    {
      "epoch": 0.22630498216794553,
      "grad_norm": 0.4118683338165283,
      "learning_rate": 8e-05,
      "loss": 1.5915,
      "step": 2094
    },
    {
      "epoch": 0.22641305522533234,
      "grad_norm": 0.34408804774284363,
      "learning_rate": 8e-05,
      "loss": 1.6893,
      "step": 2095
    },
    {
      "epoch": 0.22652112828271911,
      "grad_norm": 0.37702298164367676,
      "learning_rate": 8e-05,
      "loss": 1.6662,
      "step": 2096
    },
    {
      "epoch": 0.22662920134010592,
      "grad_norm": 0.41421517729759216,
      "learning_rate": 8e-05,
      "loss": 2.0072,
      "step": 2097
    },
    {
      "epoch": 0.2267372743974927,
      "grad_norm": 0.3741128146648407,
      "learning_rate": 8e-05,
      "loss": 1.8918,
      "step": 2098
    },
    {
      "epoch": 0.2268453474548795,
      "grad_norm": 0.36649420857429504,
      "learning_rate": 8e-05,
      "loss": 1.7132,
      "step": 2099
    },
    {
      "epoch": 0.2269534205122663,
      "grad_norm": 0.34922367334365845,
      "learning_rate": 8e-05,
      "loss": 1.7296,
      "step": 2100
    },
    {
      "epoch": 0.22706149356965308,
      "grad_norm": 0.40428659319877625,
      "learning_rate": 8e-05,
      "loss": 1.6756,
      "step": 2101
    },
    {
      "epoch": 0.22716956662703988,
      "grad_norm": 0.33680659532546997,
      "learning_rate": 8e-05,
      "loss": 1.624,
      "step": 2102
    },
    {
      "epoch": 0.22727763968442669,
      "grad_norm": 0.3802899122238159,
      "learning_rate": 8e-05,
      "loss": 1.8264,
      "step": 2103
    },
    {
      "epoch": 0.22738571274181346,
      "grad_norm": 0.37044817209243774,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 2104
    },
    {
      "epoch": 0.22749378579920027,
      "grad_norm": 0.3547819256782532,
      "learning_rate": 8e-05,
      "loss": 1.851,
      "step": 2105
    },
    {
      "epoch": 0.22760185885658704,
      "grad_norm": 0.36910563707351685,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 2106
    },
    {
      "epoch": 0.22770993191397385,
      "grad_norm": 0.4009215533733368,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 2107
    },
    {
      "epoch": 0.22781800497136065,
      "grad_norm": 0.39027926325798035,
      "learning_rate": 8e-05,
      "loss": 1.859,
      "step": 2108
    },
    {
      "epoch": 0.22792607802874743,
      "grad_norm": 0.3891124427318573,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 2109
    },
    {
      "epoch": 0.22803415108613423,
      "grad_norm": 0.3485018014907837,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 2110
    },
    {
      "epoch": 0.22814222414352103,
      "grad_norm": 0.3762092590332031,
      "learning_rate": 8e-05,
      "loss": 1.7497,
      "step": 2111
    },
    {
      "epoch": 0.2282502972009078,
      "grad_norm": 0.3456081449985504,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 2112
    },
    {
      "epoch": 0.2283583702582946,
      "grad_norm": 0.3693814277648926,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 2113
    },
    {
      "epoch": 0.2284664433156814,
      "grad_norm": 0.34173548221588135,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 2114
    },
    {
      "epoch": 0.2285745163730682,
      "grad_norm": 0.3535604476928711,
      "learning_rate": 8e-05,
      "loss": 1.7185,
      "step": 2115
    },
    {
      "epoch": 0.228682589430455,
      "grad_norm": 0.4323752224445343,
      "learning_rate": 8e-05,
      "loss": 1.8706,
      "step": 2116
    },
    {
      "epoch": 0.22879066248784177,
      "grad_norm": 0.3621761202812195,
      "learning_rate": 8e-05,
      "loss": 1.7453,
      "step": 2117
    },
    {
      "epoch": 0.22889873554522858,
      "grad_norm": 0.3651280999183655,
      "learning_rate": 8e-05,
      "loss": 1.8356,
      "step": 2118
    },
    {
      "epoch": 0.22900680860261538,
      "grad_norm": 0.36092856526374817,
      "learning_rate": 8e-05,
      "loss": 1.7377,
      "step": 2119
    },
    {
      "epoch": 0.22911488166000216,
      "grad_norm": 0.3604333996772766,
      "learning_rate": 8e-05,
      "loss": 1.7117,
      "step": 2120
    },
    {
      "epoch": 0.22922295471738896,
      "grad_norm": 0.37237662076950073,
      "learning_rate": 8e-05,
      "loss": 1.8143,
      "step": 2121
    },
    {
      "epoch": 0.22933102777477574,
      "grad_norm": 0.3683346211910248,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 2122
    },
    {
      "epoch": 0.22943910083216254,
      "grad_norm": 0.37433743476867676,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 2123
    },
    {
      "epoch": 0.22954717388954934,
      "grad_norm": 0.36272990703582764,
      "learning_rate": 8e-05,
      "loss": 1.7943,
      "step": 2124
    },
    {
      "epoch": 0.22965524694693612,
      "grad_norm": 0.36922699213027954,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 2125
    },
    {
      "epoch": 0.22976332000432292,
      "grad_norm": 0.3609287738800049,
      "learning_rate": 8e-05,
      "loss": 1.6693,
      "step": 2126
    },
    {
      "epoch": 0.22987139306170973,
      "grad_norm": 0.36926889419555664,
      "learning_rate": 8e-05,
      "loss": 1.6912,
      "step": 2127
    },
    {
      "epoch": 0.2299794661190965,
      "grad_norm": 0.379178911447525,
      "learning_rate": 8e-05,
      "loss": 1.6821,
      "step": 2128
    },
    {
      "epoch": 0.2300875391764833,
      "grad_norm": 0.3815479278564453,
      "learning_rate": 8e-05,
      "loss": 1.9305,
      "step": 2129
    },
    {
      "epoch": 0.23019561223387008,
      "grad_norm": 0.3472154140472412,
      "learning_rate": 8e-05,
      "loss": 1.7167,
      "step": 2130
    },
    {
      "epoch": 0.2303036852912569,
      "grad_norm": 0.400835782289505,
      "learning_rate": 8e-05,
      "loss": 1.9232,
      "step": 2131
    },
    {
      "epoch": 0.2304117583486437,
      "grad_norm": 0.35532715916633606,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 2132
    },
    {
      "epoch": 0.23051983140603047,
      "grad_norm": 0.376740425825119,
      "learning_rate": 8e-05,
      "loss": 1.9009,
      "step": 2133
    },
    {
      "epoch": 0.23062790446341727,
      "grad_norm": 0.3551524877548218,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 2134
    },
    {
      "epoch": 0.23073597752080408,
      "grad_norm": 0.3769424855709076,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 2135
    },
    {
      "epoch": 0.23084405057819085,
      "grad_norm": 0.3506765365600586,
      "learning_rate": 8e-05,
      "loss": 1.7097,
      "step": 2136
    },
    {
      "epoch": 0.23095212363557766,
      "grad_norm": 0.3650258183479309,
      "learning_rate": 8e-05,
      "loss": 1.7816,
      "step": 2137
    },
    {
      "epoch": 0.23106019669296443,
      "grad_norm": 0.4579632878303528,
      "learning_rate": 8e-05,
      "loss": 1.911,
      "step": 2138
    },
    {
      "epoch": 0.23116826975035124,
      "grad_norm": 0.3669975996017456,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 2139
    },
    {
      "epoch": 0.23127634280773804,
      "grad_norm": 0.33697396516799927,
      "learning_rate": 8e-05,
      "loss": 1.643,
      "step": 2140
    },
    {
      "epoch": 0.23138441586512482,
      "grad_norm": 0.3470466434955597,
      "learning_rate": 8e-05,
      "loss": 1.7947,
      "step": 2141
    },
    {
      "epoch": 0.23149248892251162,
      "grad_norm": 0.43133723735809326,
      "learning_rate": 8e-05,
      "loss": 1.6014,
      "step": 2142
    },
    {
      "epoch": 0.23160056197989842,
      "grad_norm": 0.34392091631889343,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 2143
    },
    {
      "epoch": 0.2317086350372852,
      "grad_norm": 0.39652612805366516,
      "learning_rate": 8e-05,
      "loss": 1.9443,
      "step": 2144
    },
    {
      "epoch": 0.231816708094672,
      "grad_norm": 0.5036067366600037,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 2145
    },
    {
      "epoch": 0.23192478115205878,
      "grad_norm": 0.36674514412879944,
      "learning_rate": 8e-05,
      "loss": 1.612,
      "step": 2146
    },
    {
      "epoch": 0.23203285420944558,
      "grad_norm": 0.37338146567344666,
      "learning_rate": 8e-05,
      "loss": 1.8063,
      "step": 2147
    },
    {
      "epoch": 0.2321409272668324,
      "grad_norm": 0.36324188113212585,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 2148
    },
    {
      "epoch": 0.23224900032421916,
      "grad_norm": 0.38042178750038147,
      "learning_rate": 8e-05,
      "loss": 1.8055,
      "step": 2149
    },
    {
      "epoch": 0.23235707338160597,
      "grad_norm": 0.37401559948921204,
      "learning_rate": 8e-05,
      "loss": 1.8218,
      "step": 2150
    },
    {
      "epoch": 0.23246514643899277,
      "grad_norm": 0.34630411863327026,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 2151
    },
    {
      "epoch": 0.23257321949637955,
      "grad_norm": 0.39804714918136597,
      "learning_rate": 8e-05,
      "loss": 1.9013,
      "step": 2152
    },
    {
      "epoch": 0.23268129255376635,
      "grad_norm": 0.40547269582748413,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 2153
    },
    {
      "epoch": 0.23278936561115313,
      "grad_norm": 0.3861345052719116,
      "learning_rate": 8e-05,
      "loss": 1.75,
      "step": 2154
    },
    {
      "epoch": 0.23289743866853993,
      "grad_norm": 0.46861591935157776,
      "learning_rate": 8e-05,
      "loss": 2.1237,
      "step": 2155
    },
    {
      "epoch": 0.23300551172592673,
      "grad_norm": 0.35631823539733887,
      "learning_rate": 8e-05,
      "loss": 1.8165,
      "step": 2156
    },
    {
      "epoch": 0.2331135847833135,
      "grad_norm": 0.3970722258090973,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 2157
    },
    {
      "epoch": 0.2332216578407003,
      "grad_norm": 0.3695705533027649,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 2158
    },
    {
      "epoch": 0.23332973089808712,
      "grad_norm": 0.37531137466430664,
      "learning_rate": 8e-05,
      "loss": 1.9258,
      "step": 2159
    },
    {
      "epoch": 0.2334378039554739,
      "grad_norm": 0.4071751832962036,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 2160
    },
    {
      "epoch": 0.2335458770128607,
      "grad_norm": 0.3767264783382416,
      "learning_rate": 8e-05,
      "loss": 1.8196,
      "step": 2161
    },
    {
      "epoch": 0.23365395007024747,
      "grad_norm": 0.42254704236984253,
      "learning_rate": 8e-05,
      "loss": 1.8827,
      "step": 2162
    },
    {
      "epoch": 0.23376202312763428,
      "grad_norm": 0.377869576215744,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 2163
    },
    {
      "epoch": 0.23387009618502108,
      "grad_norm": 0.3648615777492523,
      "learning_rate": 8e-05,
      "loss": 1.7956,
      "step": 2164
    },
    {
      "epoch": 0.23397816924240786,
      "grad_norm": 0.38333049416542053,
      "learning_rate": 8e-05,
      "loss": 1.8643,
      "step": 2165
    },
    {
      "epoch": 0.23408624229979466,
      "grad_norm": 0.37670475244522095,
      "learning_rate": 8e-05,
      "loss": 1.6685,
      "step": 2166
    },
    {
      "epoch": 0.23419431535718147,
      "grad_norm": 0.36983415484428406,
      "learning_rate": 8e-05,
      "loss": 1.8155,
      "step": 2167
    },
    {
      "epoch": 0.23430238841456824,
      "grad_norm": 0.3667638599872589,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 2168
    },
    {
      "epoch": 0.23441046147195505,
      "grad_norm": 0.41168835759162903,
      "learning_rate": 8e-05,
      "loss": 1.7782,
      "step": 2169
    },
    {
      "epoch": 0.23451853452934185,
      "grad_norm": 0.35350337624549866,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 2170
    },
    {
      "epoch": 0.23462660758672862,
      "grad_norm": 0.3681151270866394,
      "learning_rate": 8e-05,
      "loss": 1.7795,
      "step": 2171
    },
    {
      "epoch": 0.23473468064411543,
      "grad_norm": 0.34820547699928284,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 2172
    },
    {
      "epoch": 0.2348427537015022,
      "grad_norm": 0.35955289006233215,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 2173
    },
    {
      "epoch": 0.234950826758889,
      "grad_norm": 0.3621039092540741,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 2174
    },
    {
      "epoch": 0.2350588998162758,
      "grad_norm": 0.3995386064052582,
      "learning_rate": 8e-05,
      "loss": 1.8909,
      "step": 2175
    },
    {
      "epoch": 0.2351669728736626,
      "grad_norm": 0.37898147106170654,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 2176
    },
    {
      "epoch": 0.2352750459310494,
      "grad_norm": 0.3780369460582733,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 2177
    },
    {
      "epoch": 0.2353831189884362,
      "grad_norm": 0.3730384111404419,
      "learning_rate": 8e-05,
      "loss": 1.786,
      "step": 2178
    },
    {
      "epoch": 0.23549119204582297,
      "grad_norm": 0.357271671295166,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 2179
    },
    {
      "epoch": 0.23559926510320978,
      "grad_norm": 0.3849795162677765,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 2180
    },
    {
      "epoch": 0.23570733816059655,
      "grad_norm": 0.35723817348480225,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 2181
    },
    {
      "epoch": 0.23581541121798336,
      "grad_norm": 0.37327802181243896,
      "learning_rate": 8e-05,
      "loss": 1.9252,
      "step": 2182
    },
    {
      "epoch": 0.23592348427537016,
      "grad_norm": 0.3704034090042114,
      "learning_rate": 8e-05,
      "loss": 1.7103,
      "step": 2183
    },
    {
      "epoch": 0.23603155733275694,
      "grad_norm": 0.375997930765152,
      "learning_rate": 8e-05,
      "loss": 1.844,
      "step": 2184
    },
    {
      "epoch": 0.23613963039014374,
      "grad_norm": 0.35905808210372925,
      "learning_rate": 8e-05,
      "loss": 1.6496,
      "step": 2185
    },
    {
      "epoch": 0.23624770344753054,
      "grad_norm": 0.3800273835659027,
      "learning_rate": 8e-05,
      "loss": 1.8063,
      "step": 2186
    },
    {
      "epoch": 0.23635577650491732,
      "grad_norm": 0.3805461823940277,
      "learning_rate": 8e-05,
      "loss": 1.8344,
      "step": 2187
    },
    {
      "epoch": 0.23646384956230412,
      "grad_norm": 0.36678841710090637,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 2188
    },
    {
      "epoch": 0.2365719226196909,
      "grad_norm": 0.37296822667121887,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 2189
    },
    {
      "epoch": 0.2366799956770777,
      "grad_norm": 0.3672078847885132,
      "learning_rate": 8e-05,
      "loss": 1.8035,
      "step": 2190
    },
    {
      "epoch": 0.2367880687344645,
      "grad_norm": 0.35733985900878906,
      "learning_rate": 8e-05,
      "loss": 1.809,
      "step": 2191
    },
    {
      "epoch": 0.23689614179185128,
      "grad_norm": 0.3379526138305664,
      "learning_rate": 8e-05,
      "loss": 1.6965,
      "step": 2192
    },
    {
      "epoch": 0.2370042148492381,
      "grad_norm": 0.3835746943950653,
      "learning_rate": 8e-05,
      "loss": 1.8199,
      "step": 2193
    },
    {
      "epoch": 0.2371122879066249,
      "grad_norm": 0.49968793988227844,
      "learning_rate": 8e-05,
      "loss": 1.8153,
      "step": 2194
    },
    {
      "epoch": 0.23722036096401167,
      "grad_norm": 0.3420034945011139,
      "learning_rate": 8e-05,
      "loss": 1.6382,
      "step": 2195
    },
    {
      "epoch": 0.23732843402139847,
      "grad_norm": 0.3732185959815979,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 2196
    },
    {
      "epoch": 0.23743650707878525,
      "grad_norm": 0.3790595531463623,
      "learning_rate": 8e-05,
      "loss": 1.6562,
      "step": 2197
    },
    {
      "epoch": 0.23754458013617205,
      "grad_norm": 0.3528604507446289,
      "learning_rate": 8e-05,
      "loss": 1.7056,
      "step": 2198
    },
    {
      "epoch": 0.23765265319355885,
      "grad_norm": 0.3715028762817383,
      "learning_rate": 8e-05,
      "loss": 1.627,
      "step": 2199
    },
    {
      "epoch": 0.23776072625094563,
      "grad_norm": 0.41230669617652893,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 2200
    },
    {
      "epoch": 0.23786879930833243,
      "grad_norm": 0.3810524046421051,
      "learning_rate": 8e-05,
      "loss": 1.8767,
      "step": 2201
    },
    {
      "epoch": 0.23797687236571924,
      "grad_norm": 0.37211164832115173,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 2202
    },
    {
      "epoch": 0.23808494542310601,
      "grad_norm": 0.39589688181877136,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 2203
    },
    {
      "epoch": 0.23819301848049282,
      "grad_norm": 0.3493586480617523,
      "learning_rate": 8e-05,
      "loss": 1.6978,
      "step": 2204
    },
    {
      "epoch": 0.2383010915378796,
      "grad_norm": 0.35826364159584045,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 2205
    },
    {
      "epoch": 0.2384091645952664,
      "grad_norm": 0.3929729759693146,
      "learning_rate": 8e-05,
      "loss": 1.8507,
      "step": 2206
    },
    {
      "epoch": 0.2385172376526532,
      "grad_norm": 0.38978311419487,
      "learning_rate": 8e-05,
      "loss": 1.8152,
      "step": 2207
    },
    {
      "epoch": 0.23862531071003998,
      "grad_norm": 0.3733316957950592,
      "learning_rate": 8e-05,
      "loss": 1.7432,
      "step": 2208
    },
    {
      "epoch": 0.23873338376742678,
      "grad_norm": 0.3959824740886688,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 2209
    },
    {
      "epoch": 0.23884145682481359,
      "grad_norm": 0.3888351023197174,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 2210
    },
    {
      "epoch": 0.23894952988220036,
      "grad_norm": 0.38295966386795044,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 2211
    },
    {
      "epoch": 0.23905760293958717,
      "grad_norm": 0.3820528984069824,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 2212
    },
    {
      "epoch": 0.23916567599697394,
      "grad_norm": 0.3936896026134491,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 2213
    },
    {
      "epoch": 0.23927374905436075,
      "grad_norm": 0.3367980420589447,
      "learning_rate": 8e-05,
      "loss": 1.6708,
      "step": 2214
    },
    {
      "epoch": 0.23938182211174755,
      "grad_norm": 0.38799363374710083,
      "learning_rate": 8e-05,
      "loss": 1.6317,
      "step": 2215
    },
    {
      "epoch": 0.23948989516913433,
      "grad_norm": 0.3720891773700714,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 2216
    },
    {
      "epoch": 0.23959796822652113,
      "grad_norm": 0.34050485491752625,
      "learning_rate": 8e-05,
      "loss": 1.6556,
      "step": 2217
    },
    {
      "epoch": 0.23970604128390793,
      "grad_norm": 0.44663381576538086,
      "learning_rate": 8e-05,
      "loss": 1.9447,
      "step": 2218
    },
    {
      "epoch": 0.2398141143412947,
      "grad_norm": 0.3856361210346222,
      "learning_rate": 8e-05,
      "loss": 1.8881,
      "step": 2219
    },
    {
      "epoch": 0.2399221873986815,
      "grad_norm": 0.37233471870422363,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 2220
    },
    {
      "epoch": 0.2400302604560683,
      "grad_norm": 0.36970770359039307,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 2221
    },
    {
      "epoch": 0.2401383335134551,
      "grad_norm": 0.3888806104660034,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 2222
    },
    {
      "epoch": 0.2402464065708419,
      "grad_norm": 0.3729874789714813,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 2223
    },
    {
      "epoch": 0.24035447962822867,
      "grad_norm": 0.38584765791893005,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 2224
    },
    {
      "epoch": 0.24046255268561548,
      "grad_norm": 0.39203691482543945,
      "learning_rate": 8e-05,
      "loss": 1.8621,
      "step": 2225
    },
    {
      "epoch": 0.24057062574300228,
      "grad_norm": 0.40381258726119995,
      "learning_rate": 8e-05,
      "loss": 1.9093,
      "step": 2226
    },
    {
      "epoch": 0.24067869880038906,
      "grad_norm": 0.3886108100414276,
      "learning_rate": 8e-05,
      "loss": 1.6487,
      "step": 2227
    },
    {
      "epoch": 0.24078677185777586,
      "grad_norm": 0.37426671385765076,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 2228
    },
    {
      "epoch": 0.24089484491516264,
      "grad_norm": 0.35715681314468384,
      "learning_rate": 8e-05,
      "loss": 1.8178,
      "step": 2229
    },
    {
      "epoch": 0.24100291797254944,
      "grad_norm": 0.3785596191883087,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 2230
    },
    {
      "epoch": 0.24111099102993624,
      "grad_norm": 0.4022001028060913,
      "learning_rate": 8e-05,
      "loss": 1.884,
      "step": 2231
    },
    {
      "epoch": 0.24121906408732302,
      "grad_norm": 0.38375037908554077,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 2232
    },
    {
      "epoch": 0.24132713714470982,
      "grad_norm": 0.40449509024620056,
      "learning_rate": 8e-05,
      "loss": 1.8424,
      "step": 2233
    },
    {
      "epoch": 0.24143521020209663,
      "grad_norm": 0.3777287006378174,
      "learning_rate": 8e-05,
      "loss": 1.8728,
      "step": 2234
    },
    {
      "epoch": 0.2415432832594834,
      "grad_norm": 0.36847957968711853,
      "learning_rate": 8e-05,
      "loss": 1.8695,
      "step": 2235
    },
    {
      "epoch": 0.2416513563168702,
      "grad_norm": 0.39545324444770813,
      "learning_rate": 8e-05,
      "loss": 1.8409,
      "step": 2236
    },
    {
      "epoch": 0.24175942937425698,
      "grad_norm": 0.3706943988800049,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 2237
    },
    {
      "epoch": 0.2418675024316438,
      "grad_norm": 0.3927188515663147,
      "learning_rate": 8e-05,
      "loss": 1.8033,
      "step": 2238
    },
    {
      "epoch": 0.2419755754890306,
      "grad_norm": 0.37358346581459045,
      "learning_rate": 8e-05,
      "loss": 1.8419,
      "step": 2239
    },
    {
      "epoch": 0.24208364854641737,
      "grad_norm": 0.3923817276954651,
      "learning_rate": 8e-05,
      "loss": 1.8291,
      "step": 2240
    },
    {
      "epoch": 0.24219172160380417,
      "grad_norm": 0.3883258104324341,
      "learning_rate": 8e-05,
      "loss": 1.7932,
      "step": 2241
    },
    {
      "epoch": 0.24229979466119098,
      "grad_norm": 0.3817564845085144,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 2242
    },
    {
      "epoch": 0.24240786771857775,
      "grad_norm": 0.4428686201572418,
      "learning_rate": 8e-05,
      "loss": 1.9983,
      "step": 2243
    },
    {
      "epoch": 0.24251594077596456,
      "grad_norm": 0.3858909606933594,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 2244
    },
    {
      "epoch": 0.24262401383335133,
      "grad_norm": 0.3450782895088196,
      "learning_rate": 8e-05,
      "loss": 1.483,
      "step": 2245
    },
    {
      "epoch": 0.24273208689073814,
      "grad_norm": 0.40456026792526245,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 2246
    },
    {
      "epoch": 0.24284015994812494,
      "grad_norm": 0.3651934862136841,
      "learning_rate": 8e-05,
      "loss": 1.4756,
      "step": 2247
    },
    {
      "epoch": 0.24294823300551172,
      "grad_norm": 0.3923119306564331,
      "learning_rate": 8e-05,
      "loss": 1.7053,
      "step": 2248
    },
    {
      "epoch": 0.24305630606289852,
      "grad_norm": 0.38456419110298157,
      "learning_rate": 8e-05,
      "loss": 1.9331,
      "step": 2249
    },
    {
      "epoch": 0.24316437912028532,
      "grad_norm": 0.38729560375213623,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 2250
    },
    {
      "epoch": 0.2432724521776721,
      "grad_norm": 0.38927167654037476,
      "learning_rate": 8e-05,
      "loss": 1.464,
      "step": 2251
    },
    {
      "epoch": 0.2433805252350589,
      "grad_norm": 0.3939606547355652,
      "learning_rate": 8e-05,
      "loss": 1.8651,
      "step": 2252
    },
    {
      "epoch": 0.2434885982924457,
      "grad_norm": 0.3740518093109131,
      "learning_rate": 8e-05,
      "loss": 1.701,
      "step": 2253
    },
    {
      "epoch": 0.24359667134983248,
      "grad_norm": 0.3723231852054596,
      "learning_rate": 8e-05,
      "loss": 1.8437,
      "step": 2254
    },
    {
      "epoch": 0.2437047444072193,
      "grad_norm": 0.36414843797683716,
      "learning_rate": 8e-05,
      "loss": 1.8114,
      "step": 2255
    },
    {
      "epoch": 0.24381281746460606,
      "grad_norm": 0.3751830756664276,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 2256
    },
    {
      "epoch": 0.24392089052199287,
      "grad_norm": 0.3694900870323181,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 2257
    },
    {
      "epoch": 0.24402896357937967,
      "grad_norm": 0.3713975250720978,
      "learning_rate": 8e-05,
      "loss": 1.5858,
      "step": 2258
    },
    {
      "epoch": 0.24413703663676645,
      "grad_norm": 0.3691284954547882,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 2259
    },
    {
      "epoch": 0.24424510969415325,
      "grad_norm": 0.438029944896698,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 2260
    },
    {
      "epoch": 0.24435318275154005,
      "grad_norm": 0.3459267318248749,
      "learning_rate": 8e-05,
      "loss": 1.5822,
      "step": 2261
    },
    {
      "epoch": 0.24446125580892683,
      "grad_norm": 0.3764677345752716,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 2262
    },
    {
      "epoch": 0.24456932886631363,
      "grad_norm": 0.38184359669685364,
      "learning_rate": 8e-05,
      "loss": 1.8978,
      "step": 2263
    },
    {
      "epoch": 0.2446774019237004,
      "grad_norm": 0.3612769544124603,
      "learning_rate": 8e-05,
      "loss": 1.8164,
      "step": 2264
    },
    {
      "epoch": 0.24478547498108721,
      "grad_norm": 0.37051650881767273,
      "learning_rate": 8e-05,
      "loss": 1.6602,
      "step": 2265
    },
    {
      "epoch": 0.24489354803847402,
      "grad_norm": 0.3863441050052643,
      "learning_rate": 8e-05,
      "loss": 1.8059,
      "step": 2266
    },
    {
      "epoch": 0.2450016210958608,
      "grad_norm": 0.3679267168045044,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 2267
    },
    {
      "epoch": 0.2451096941532476,
      "grad_norm": 0.36337706446647644,
      "learning_rate": 8e-05,
      "loss": 1.5827,
      "step": 2268
    },
    {
      "epoch": 0.2452177672106344,
      "grad_norm": 0.4200214445590973,
      "learning_rate": 8e-05,
      "loss": 1.771,
      "step": 2269
    },
    {
      "epoch": 0.24532584026802118,
      "grad_norm": 0.38746392726898193,
      "learning_rate": 8e-05,
      "loss": 1.8312,
      "step": 2270
    },
    {
      "epoch": 0.24543391332540798,
      "grad_norm": 0.36637184023857117,
      "learning_rate": 8e-05,
      "loss": 1.7339,
      "step": 2271
    },
    {
      "epoch": 0.24554198638279476,
      "grad_norm": 0.39609017968177795,
      "learning_rate": 8e-05,
      "loss": 1.8369,
      "step": 2272
    },
    {
      "epoch": 0.24565005944018156,
      "grad_norm": 0.3779478967189789,
      "learning_rate": 8e-05,
      "loss": 1.8543,
      "step": 2273
    },
    {
      "epoch": 0.24575813249756837,
      "grad_norm": 0.4619922637939453,
      "learning_rate": 8e-05,
      "loss": 1.9894,
      "step": 2274
    },
    {
      "epoch": 0.24586620555495514,
      "grad_norm": 0.3626987040042877,
      "learning_rate": 8e-05,
      "loss": 1.7789,
      "step": 2275
    },
    {
      "epoch": 0.24597427861234195,
      "grad_norm": 0.3633941411972046,
      "learning_rate": 8e-05,
      "loss": 1.7663,
      "step": 2276
    },
    {
      "epoch": 0.24608235166972875,
      "grad_norm": 0.33208996057510376,
      "learning_rate": 8e-05,
      "loss": 1.6041,
      "step": 2277
    },
    {
      "epoch": 0.24619042472711553,
      "grad_norm": 0.3755781054496765,
      "learning_rate": 8e-05,
      "loss": 1.7178,
      "step": 2278
    },
    {
      "epoch": 0.24629849778450233,
      "grad_norm": 0.404191255569458,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 2279
    },
    {
      "epoch": 0.2464065708418891,
      "grad_norm": 0.3851582407951355,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 2280
    },
    {
      "epoch": 0.2465146438992759,
      "grad_norm": 0.381092369556427,
      "learning_rate": 8e-05,
      "loss": 1.5647,
      "step": 2281
    },
    {
      "epoch": 0.2466227169566627,
      "grad_norm": 0.36779624223709106,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 2282
    },
    {
      "epoch": 0.2467307900140495,
      "grad_norm": 0.369118869304657,
      "learning_rate": 8e-05,
      "loss": 1.8087,
      "step": 2283
    },
    {
      "epoch": 0.2468388630714363,
      "grad_norm": 0.36656805872917175,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 2284
    },
    {
      "epoch": 0.2469469361288231,
      "grad_norm": 0.3935786187648773,
      "learning_rate": 8e-05,
      "loss": 1.9369,
      "step": 2285
    },
    {
      "epoch": 0.24705500918620987,
      "grad_norm": 0.37046152353286743,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 2286
    },
    {
      "epoch": 0.24716308224359668,
      "grad_norm": 0.3597521185874939,
      "learning_rate": 8e-05,
      "loss": 1.653,
      "step": 2287
    },
    {
      "epoch": 0.24727115530098345,
      "grad_norm": 0.3555164635181427,
      "learning_rate": 8e-05,
      "loss": 1.7164,
      "step": 2288
    },
    {
      "epoch": 0.24737922835837026,
      "grad_norm": 0.368020623922348,
      "learning_rate": 8e-05,
      "loss": 1.8372,
      "step": 2289
    },
    {
      "epoch": 0.24748730141575706,
      "grad_norm": 0.3481471836566925,
      "learning_rate": 8e-05,
      "loss": 1.4505,
      "step": 2290
    },
    {
      "epoch": 0.24759537447314384,
      "grad_norm": 0.38928261399269104,
      "learning_rate": 8e-05,
      "loss": 1.8588,
      "step": 2291
    },
    {
      "epoch": 0.24770344753053064,
      "grad_norm": 0.3435552716255188,
      "learning_rate": 8e-05,
      "loss": 1.7479,
      "step": 2292
    },
    {
      "epoch": 0.24781152058791744,
      "grad_norm": 0.36094003915786743,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 2293
    },
    {
      "epoch": 0.24791959364530422,
      "grad_norm": 0.36277252435684204,
      "learning_rate": 8e-05,
      "loss": 1.7336,
      "step": 2294
    },
    {
      "epoch": 0.24802766670269102,
      "grad_norm": 0.4144380986690521,
      "learning_rate": 8e-05,
      "loss": 1.7447,
      "step": 2295
    },
    {
      "epoch": 0.2481357397600778,
      "grad_norm": 0.35916078090667725,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 2296
    },
    {
      "epoch": 0.2482438128174646,
      "grad_norm": 0.43701934814453125,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 2297
    },
    {
      "epoch": 0.2483518858748514,
      "grad_norm": 0.3719756603240967,
      "learning_rate": 8e-05,
      "loss": 1.6979,
      "step": 2298
    },
    {
      "epoch": 0.24845995893223818,
      "grad_norm": 0.36405685544013977,
      "learning_rate": 8e-05,
      "loss": 1.7976,
      "step": 2299
    },
    {
      "epoch": 0.248568031989625,
      "grad_norm": 0.3792782425880432,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 2300
    },
    {
      "epoch": 0.2486761050470118,
      "grad_norm": 0.36561650037765503,
      "learning_rate": 8e-05,
      "loss": 1.7456,
      "step": 2301
    },
    {
      "epoch": 0.24878417810439857,
      "grad_norm": 0.3584391474723816,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 2302
    },
    {
      "epoch": 0.24889225116178537,
      "grad_norm": 0.36541834473609924,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 2303
    },
    {
      "epoch": 0.24900032421917215,
      "grad_norm": 0.3484187126159668,
      "learning_rate": 8e-05,
      "loss": 1.6182,
      "step": 2304
    },
    {
      "epoch": 0.24910839727655895,
      "grad_norm": 0.3748144805431366,
      "learning_rate": 8e-05,
      "loss": 1.8136,
      "step": 2305
    },
    {
      "epoch": 0.24921647033394576,
      "grad_norm": 0.3921423554420471,
      "learning_rate": 8e-05,
      "loss": 1.8872,
      "step": 2306
    },
    {
      "epoch": 0.24932454339133253,
      "grad_norm": 0.3733462989330292,
      "learning_rate": 8e-05,
      "loss": 1.6957,
      "step": 2307
    },
    {
      "epoch": 0.24943261644871934,
      "grad_norm": 0.367025226354599,
      "learning_rate": 8e-05,
      "loss": 1.7848,
      "step": 2308
    },
    {
      "epoch": 0.24954068950610614,
      "grad_norm": 0.44055962562561035,
      "learning_rate": 8e-05,
      "loss": 1.875,
      "step": 2309
    },
    {
      "epoch": 0.24964876256349292,
      "grad_norm": 0.37285923957824707,
      "learning_rate": 8e-05,
      "loss": 1.8193,
      "step": 2310
    },
    {
      "epoch": 0.24975683562087972,
      "grad_norm": 0.37885582447052,
      "learning_rate": 8e-05,
      "loss": 1.6128,
      "step": 2311
    },
    {
      "epoch": 0.2498649086782665,
      "grad_norm": 0.34981241822242737,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 2312
    },
    {
      "epoch": 0.2499729817356533,
      "grad_norm": 0.3603290319442749,
      "learning_rate": 8e-05,
      "loss": 1.5015,
      "step": 2313
    },
    {
      "epoch": 0.2500810547930401,
      "grad_norm": 0.39479121565818787,
      "learning_rate": 8e-05,
      "loss": 1.8576,
      "step": 2314
    },
    {
      "epoch": 0.2501891278504269,
      "grad_norm": 0.37982019782066345,
      "learning_rate": 8e-05,
      "loss": 1.8667,
      "step": 2315
    },
    {
      "epoch": 0.25029720090781366,
      "grad_norm": 0.37069183588027954,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 2316
    },
    {
      "epoch": 0.25040527396520046,
      "grad_norm": 0.3424277901649475,
      "learning_rate": 8e-05,
      "loss": 1.6891,
      "step": 2317
    },
    {
      "epoch": 0.25051334702258726,
      "grad_norm": 0.3897865116596222,
      "learning_rate": 8e-05,
      "loss": 1.9812,
      "step": 2318
    },
    {
      "epoch": 0.25062142007997407,
      "grad_norm": 0.359379380941391,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 2319
    },
    {
      "epoch": 0.25072949313736087,
      "grad_norm": 0.3432386815547943,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 2320
    },
    {
      "epoch": 0.2508375661947477,
      "grad_norm": 0.40121743083000183,
      "learning_rate": 8e-05,
      "loss": 1.8559,
      "step": 2321
    },
    {
      "epoch": 0.2509456392521344,
      "grad_norm": 0.37241867184638977,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 2322
    },
    {
      "epoch": 0.2510537123095212,
      "grad_norm": 0.377501517534256,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 2323
    },
    {
      "epoch": 0.25116178536690803,
      "grad_norm": 0.4114398658275604,
      "learning_rate": 8e-05,
      "loss": 1.8548,
      "step": 2324
    },
    {
      "epoch": 0.25126985842429483,
      "grad_norm": 0.3749444782733917,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 2325
    },
    {
      "epoch": 0.25137793148168164,
      "grad_norm": 0.36130374670028687,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 2326
    },
    {
      "epoch": 0.2514860045390684,
      "grad_norm": 0.3502368628978729,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 2327
    },
    {
      "epoch": 0.2515940775964552,
      "grad_norm": 0.3810741901397705,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 2328
    },
    {
      "epoch": 0.251702150653842,
      "grad_norm": 0.38377493619918823,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 2329
    },
    {
      "epoch": 0.2518102237112288,
      "grad_norm": 0.411726176738739,
      "learning_rate": 8e-05,
      "loss": 1.9209,
      "step": 2330
    },
    {
      "epoch": 0.2519182967686156,
      "grad_norm": 0.3568357229232788,
      "learning_rate": 8e-05,
      "loss": 1.6508,
      "step": 2331
    },
    {
      "epoch": 0.25202636982600235,
      "grad_norm": 0.3556637763977051,
      "learning_rate": 8e-05,
      "loss": 1.6868,
      "step": 2332
    },
    {
      "epoch": 0.25213444288338915,
      "grad_norm": 0.3710155189037323,
      "learning_rate": 8e-05,
      "loss": 1.6579,
      "step": 2333
    },
    {
      "epoch": 0.25224251594077596,
      "grad_norm": 0.42000019550323486,
      "learning_rate": 8e-05,
      "loss": 2.0026,
      "step": 2334
    },
    {
      "epoch": 0.25235058899816276,
      "grad_norm": 0.4007807970046997,
      "learning_rate": 8e-05,
      "loss": 1.9122,
      "step": 2335
    },
    {
      "epoch": 0.25245866205554957,
      "grad_norm": 0.43295976519584656,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 2336
    },
    {
      "epoch": 0.25256673511293637,
      "grad_norm": 0.39074695110321045,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 2337
    },
    {
      "epoch": 0.2526748081703231,
      "grad_norm": 0.37223386764526367,
      "learning_rate": 8e-05,
      "loss": 1.5521,
      "step": 2338
    },
    {
      "epoch": 0.2527828812277099,
      "grad_norm": 0.36089980602264404,
      "learning_rate": 8e-05,
      "loss": 1.7315,
      "step": 2339
    },
    {
      "epoch": 0.2528909542850967,
      "grad_norm": 0.36689791083335876,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 2340
    },
    {
      "epoch": 0.25299902734248353,
      "grad_norm": 0.3840523958206177,
      "learning_rate": 8e-05,
      "loss": 1.7713,
      "step": 2341
    },
    {
      "epoch": 0.25310710039987033,
      "grad_norm": 0.3728722035884857,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 2342
    },
    {
      "epoch": 0.2532151734572571,
      "grad_norm": 0.396951287984848,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 2343
    },
    {
      "epoch": 0.2533232465146439,
      "grad_norm": 0.35852232575416565,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 2344
    },
    {
      "epoch": 0.2534313195720307,
      "grad_norm": 0.35360392928123474,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 2345
    },
    {
      "epoch": 0.2535393926294175,
      "grad_norm": 0.35659259557724,
      "learning_rate": 8e-05,
      "loss": 1.5732,
      "step": 2346
    },
    {
      "epoch": 0.2536474656868043,
      "grad_norm": 0.36444783210754395,
      "learning_rate": 8e-05,
      "loss": 1.693,
      "step": 2347
    },
    {
      "epoch": 0.2537555387441911,
      "grad_norm": 0.37828055024147034,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 2348
    },
    {
      "epoch": 0.25386361180157785,
      "grad_norm": 0.4031600058078766,
      "learning_rate": 8e-05,
      "loss": 1.8711,
      "step": 2349
    },
    {
      "epoch": 0.25397168485896465,
      "grad_norm": 0.3750914931297302,
      "learning_rate": 8e-05,
      "loss": 1.636,
      "step": 2350
    },
    {
      "epoch": 0.25407975791635146,
      "grad_norm": 0.372081458568573,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 2351
    },
    {
      "epoch": 0.25418783097373826,
      "grad_norm": 0.37560608983039856,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 2352
    },
    {
      "epoch": 0.25429590403112506,
      "grad_norm": 0.38047584891319275,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 2353
    },
    {
      "epoch": 0.2544039770885118,
      "grad_norm": 0.4073527753353119,
      "learning_rate": 8e-05,
      "loss": 1.8785,
      "step": 2354
    },
    {
      "epoch": 0.2545120501458986,
      "grad_norm": 0.4195411801338196,
      "learning_rate": 8e-05,
      "loss": 1.8149,
      "step": 2355
    },
    {
      "epoch": 0.2546201232032854,
      "grad_norm": 0.40978819131851196,
      "learning_rate": 8e-05,
      "loss": 1.8837,
      "step": 2356
    },
    {
      "epoch": 0.2547281962606722,
      "grad_norm": 0.36176303029060364,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 2357
    },
    {
      "epoch": 0.254836269318059,
      "grad_norm": 0.3815341591835022,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 2358
    },
    {
      "epoch": 0.2549443423754458,
      "grad_norm": 0.3497124910354614,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 2359
    },
    {
      "epoch": 0.2550524154328326,
      "grad_norm": 0.3657241761684418,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 2360
    },
    {
      "epoch": 0.2551604884902194,
      "grad_norm": 0.3724481463432312,
      "learning_rate": 8e-05,
      "loss": 1.684,
      "step": 2361
    },
    {
      "epoch": 0.2552685615476062,
      "grad_norm": 0.3785184919834137,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 2362
    },
    {
      "epoch": 0.255376634604993,
      "grad_norm": 0.3680363893508911,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 2363
    },
    {
      "epoch": 0.2554847076623798,
      "grad_norm": 0.3570915758609772,
      "learning_rate": 8e-05,
      "loss": 1.6243,
      "step": 2364
    },
    {
      "epoch": 0.25559278071976654,
      "grad_norm": 0.39442989230155945,
      "learning_rate": 8e-05,
      "loss": 1.805,
      "step": 2365
    },
    {
      "epoch": 0.25570085377715335,
      "grad_norm": 0.3628694415092468,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 2366
    },
    {
      "epoch": 0.25580892683454015,
      "grad_norm": 0.3751349151134491,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 2367
    },
    {
      "epoch": 0.25591699989192696,
      "grad_norm": 0.36008697748184204,
      "learning_rate": 8e-05,
      "loss": 1.6463,
      "step": 2368
    },
    {
      "epoch": 0.25602507294931376,
      "grad_norm": 0.37018558382987976,
      "learning_rate": 8e-05,
      "loss": 1.6328,
      "step": 2369
    },
    {
      "epoch": 0.2561331460067005,
      "grad_norm": 0.34837254881858826,
      "learning_rate": 8e-05,
      "loss": 1.5436,
      "step": 2370
    },
    {
      "epoch": 0.2562412190640873,
      "grad_norm": 0.3591398596763611,
      "learning_rate": 8e-05,
      "loss": 1.5317,
      "step": 2371
    },
    {
      "epoch": 0.2563492921214741,
      "grad_norm": 0.34367167949676514,
      "learning_rate": 8e-05,
      "loss": 1.5222,
      "step": 2372
    },
    {
      "epoch": 0.2564573651788609,
      "grad_norm": 0.37893447279930115,
      "learning_rate": 8e-05,
      "loss": 1.8652,
      "step": 2373
    },
    {
      "epoch": 0.2565654382362477,
      "grad_norm": 0.3768373131752014,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 2374
    },
    {
      "epoch": 0.25667351129363447,
      "grad_norm": 0.3967050313949585,
      "learning_rate": 8e-05,
      "loss": 1.9085,
      "step": 2375
    },
    {
      "epoch": 0.2567815843510213,
      "grad_norm": 0.3844134211540222,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 2376
    },
    {
      "epoch": 0.2568896574084081,
      "grad_norm": 0.37405160069465637,
      "learning_rate": 8e-05,
      "loss": 1.6963,
      "step": 2377
    },
    {
      "epoch": 0.2569977304657949,
      "grad_norm": 0.3459620177745819,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 2378
    },
    {
      "epoch": 0.2571058035231817,
      "grad_norm": 0.37010252475738525,
      "learning_rate": 8e-05,
      "loss": 1.6831,
      "step": 2379
    },
    {
      "epoch": 0.2572138765805685,
      "grad_norm": 0.4051039516925812,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 2380
    },
    {
      "epoch": 0.25732194963795524,
      "grad_norm": 0.34516793489456177,
      "learning_rate": 8e-05,
      "loss": 1.652,
      "step": 2381
    },
    {
      "epoch": 0.25743002269534204,
      "grad_norm": 0.3632252812385559,
      "learning_rate": 8e-05,
      "loss": 1.6696,
      "step": 2382
    },
    {
      "epoch": 0.25753809575272885,
      "grad_norm": 0.37083593010902405,
      "learning_rate": 8e-05,
      "loss": 1.6783,
      "step": 2383
    },
    {
      "epoch": 0.25764616881011565,
      "grad_norm": 0.36727359890937805,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 2384
    },
    {
      "epoch": 0.25775424186750245,
      "grad_norm": 0.38211095333099365,
      "learning_rate": 8e-05,
      "loss": 1.6724,
      "step": 2385
    },
    {
      "epoch": 0.2578623149248892,
      "grad_norm": 0.4013848900794983,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 2386
    },
    {
      "epoch": 0.257970387982276,
      "grad_norm": 0.3575046956539154,
      "learning_rate": 8e-05,
      "loss": 1.7418,
      "step": 2387
    },
    {
      "epoch": 0.2580784610396628,
      "grad_norm": 0.36692652106285095,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 2388
    },
    {
      "epoch": 0.2581865340970496,
      "grad_norm": 0.37888476252555847,
      "learning_rate": 8e-05,
      "loss": 1.8369,
      "step": 2389
    },
    {
      "epoch": 0.2582946071544364,
      "grad_norm": 0.3732790946960449,
      "learning_rate": 8e-05,
      "loss": 1.6603,
      "step": 2390
    },
    {
      "epoch": 0.25840268021182317,
      "grad_norm": 0.38367220759391785,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 2391
    },
    {
      "epoch": 0.25851075326920997,
      "grad_norm": 0.36935535073280334,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 2392
    },
    {
      "epoch": 0.2586188263265968,
      "grad_norm": 0.3913491666316986,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 2393
    },
    {
      "epoch": 0.2587268993839836,
      "grad_norm": 0.38293763995170593,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 2394
    },
    {
      "epoch": 0.2588349724413704,
      "grad_norm": 0.3737819790840149,
      "learning_rate": 8e-05,
      "loss": 1.8052,
      "step": 2395
    },
    {
      "epoch": 0.2589430454987572,
      "grad_norm": 0.37721824645996094,
      "learning_rate": 8e-05,
      "loss": 1.6376,
      "step": 2396
    },
    {
      "epoch": 0.25905111855614393,
      "grad_norm": 0.4264904260635376,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 2397
    },
    {
      "epoch": 0.25915919161353074,
      "grad_norm": 0.3576003611087799,
      "learning_rate": 8e-05,
      "loss": 1.7774,
      "step": 2398
    },
    {
      "epoch": 0.25926726467091754,
      "grad_norm": 0.3670266568660736,
      "learning_rate": 8e-05,
      "loss": 1.8063,
      "step": 2399
    },
    {
      "epoch": 0.25937533772830434,
      "grad_norm": 0.4331955015659332,
      "learning_rate": 8e-05,
      "loss": 1.4671,
      "step": 2400
    },
    {
      "epoch": 0.25948341078569115,
      "grad_norm": 0.3563782870769501,
      "learning_rate": 8e-05,
      "loss": 1.6389,
      "step": 2401
    },
    {
      "epoch": 0.2595914838430779,
      "grad_norm": 0.3500118553638458,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 2402
    },
    {
      "epoch": 0.2596995569004647,
      "grad_norm": 0.3432362377643585,
      "learning_rate": 8e-05,
      "loss": 1.4502,
      "step": 2403
    },
    {
      "epoch": 0.2598076299578515,
      "grad_norm": 0.39019864797592163,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 2404
    },
    {
      "epoch": 0.2599157030152383,
      "grad_norm": 0.36085912585258484,
      "learning_rate": 8e-05,
      "loss": 1.8191,
      "step": 2405
    },
    {
      "epoch": 0.2600237760726251,
      "grad_norm": 0.3707723915576935,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 2406
    },
    {
      "epoch": 0.26013184913001186,
      "grad_norm": 0.3534475266933441,
      "learning_rate": 8e-05,
      "loss": 1.5932,
      "step": 2407
    },
    {
      "epoch": 0.26023992218739866,
      "grad_norm": 0.37683358788490295,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 2408
    },
    {
      "epoch": 0.26034799524478547,
      "grad_norm": 0.3793697953224182,
      "learning_rate": 8e-05,
      "loss": 1.6994,
      "step": 2409
    },
    {
      "epoch": 0.26045606830217227,
      "grad_norm": 0.35983163118362427,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 2410
    },
    {
      "epoch": 0.2605641413595591,
      "grad_norm": 0.388555109500885,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 2411
    },
    {
      "epoch": 0.2606722144169459,
      "grad_norm": 0.39065611362457275,
      "learning_rate": 8e-05,
      "loss": 1.873,
      "step": 2412
    },
    {
      "epoch": 0.26078028747433263,
      "grad_norm": 0.45261526107788086,
      "learning_rate": 8e-05,
      "loss": 1.6543,
      "step": 2413
    },
    {
      "epoch": 0.26088836053171943,
      "grad_norm": 0.4009803235530853,
      "learning_rate": 8e-05,
      "loss": 1.9119,
      "step": 2414
    },
    {
      "epoch": 0.26099643358910624,
      "grad_norm": 0.3706305921077728,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 2415
    },
    {
      "epoch": 0.26110450664649304,
      "grad_norm": 0.3484938442707062,
      "learning_rate": 8e-05,
      "loss": 1.6855,
      "step": 2416
    },
    {
      "epoch": 0.26121257970387984,
      "grad_norm": 0.4164220690727234,
      "learning_rate": 8e-05,
      "loss": 1.8719,
      "step": 2417
    },
    {
      "epoch": 0.2613206527612666,
      "grad_norm": 0.3573210835456848,
      "learning_rate": 8e-05,
      "loss": 1.5876,
      "step": 2418
    },
    {
      "epoch": 0.2614287258186534,
      "grad_norm": 0.38837939500808716,
      "learning_rate": 8e-05,
      "loss": 1.8216,
      "step": 2419
    },
    {
      "epoch": 0.2615367988760402,
      "grad_norm": 0.3543044626712799,
      "learning_rate": 8e-05,
      "loss": 1.6882,
      "step": 2420
    },
    {
      "epoch": 0.261644871933427,
      "grad_norm": 0.3907885253429413,
      "learning_rate": 8e-05,
      "loss": 1.8667,
      "step": 2421
    },
    {
      "epoch": 0.2617529449908138,
      "grad_norm": 0.37509965896606445,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 2422
    },
    {
      "epoch": 0.2618610180482006,
      "grad_norm": 0.3805215060710907,
      "learning_rate": 8e-05,
      "loss": 1.7785,
      "step": 2423
    },
    {
      "epoch": 0.26196909110558736,
      "grad_norm": 0.3731456696987152,
      "learning_rate": 8e-05,
      "loss": 1.5353,
      "step": 2424
    },
    {
      "epoch": 0.26207716416297416,
      "grad_norm": 0.393075555562973,
      "learning_rate": 8e-05,
      "loss": 1.8361,
      "step": 2425
    },
    {
      "epoch": 0.26218523722036097,
      "grad_norm": 0.4163043200969696,
      "learning_rate": 8e-05,
      "loss": 1.8546,
      "step": 2426
    },
    {
      "epoch": 0.26229331027774777,
      "grad_norm": 0.35960590839385986,
      "learning_rate": 8e-05,
      "loss": 1.6065,
      "step": 2427
    },
    {
      "epoch": 0.2624013833351346,
      "grad_norm": 0.37370970845222473,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 2428
    },
    {
      "epoch": 0.2625094563925213,
      "grad_norm": 0.4729326069355011,
      "learning_rate": 8e-05,
      "loss": 1.8061,
      "step": 2429
    },
    {
      "epoch": 0.2626175294499081,
      "grad_norm": 0.39908185601234436,
      "learning_rate": 8e-05,
      "loss": 1.9572,
      "step": 2430
    },
    {
      "epoch": 0.26272560250729493,
      "grad_norm": 0.4235093593597412,
      "learning_rate": 8e-05,
      "loss": 2.0039,
      "step": 2431
    },
    {
      "epoch": 0.26283367556468173,
      "grad_norm": 0.3778579533100128,
      "learning_rate": 8e-05,
      "loss": 1.8555,
      "step": 2432
    },
    {
      "epoch": 0.26294174862206854,
      "grad_norm": 0.3497962951660156,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 2433
    },
    {
      "epoch": 0.2630498216794553,
      "grad_norm": 0.3692166805267334,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 2434
    },
    {
      "epoch": 0.2631578947368421,
      "grad_norm": 0.39336279034614563,
      "learning_rate": 8e-05,
      "loss": 1.9107,
      "step": 2435
    },
    {
      "epoch": 0.2632659677942289,
      "grad_norm": 0.38543298840522766,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 2436
    },
    {
      "epoch": 0.2633740408516157,
      "grad_norm": 0.40469375252723694,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 2437
    },
    {
      "epoch": 0.2634821139090025,
      "grad_norm": 0.3772241175174713,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 2438
    },
    {
      "epoch": 0.2635901869663893,
      "grad_norm": 0.3549844026565552,
      "learning_rate": 8e-05,
      "loss": 1.5512,
      "step": 2439
    },
    {
      "epoch": 0.26369826002377605,
      "grad_norm": 0.4116078317165375,
      "learning_rate": 8e-05,
      "loss": 1.857,
      "step": 2440
    },
    {
      "epoch": 0.26380633308116286,
      "grad_norm": 0.3834722340106964,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 2441
    },
    {
      "epoch": 0.26391440613854966,
      "grad_norm": 0.35018327832221985,
      "learning_rate": 8e-05,
      "loss": 1.706,
      "step": 2442
    },
    {
      "epoch": 0.26402247919593647,
      "grad_norm": 0.3909597098827362,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 2443
    },
    {
      "epoch": 0.26413055225332327,
      "grad_norm": 0.3648270070552826,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 2444
    },
    {
      "epoch": 0.26423862531071,
      "grad_norm": 0.3591890335083008,
      "learning_rate": 8e-05,
      "loss": 1.6948,
      "step": 2445
    },
    {
      "epoch": 0.2643466983680968,
      "grad_norm": 0.3624780476093292,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 2446
    },
    {
      "epoch": 0.2644547714254836,
      "grad_norm": 0.33978351950645447,
      "learning_rate": 8e-05,
      "loss": 1.5544,
      "step": 2447
    },
    {
      "epoch": 0.26456284448287043,
      "grad_norm": 0.42829594016075134,
      "learning_rate": 8e-05,
      "loss": 1.6167,
      "step": 2448
    },
    {
      "epoch": 0.26467091754025723,
      "grad_norm": 0.35077425837516785,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 2449
    },
    {
      "epoch": 0.264778990597644,
      "grad_norm": 0.41539913415908813,
      "learning_rate": 8e-05,
      "loss": 1.8585,
      "step": 2450
    },
    {
      "epoch": 0.2648870636550308,
      "grad_norm": 0.36336368322372437,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 2451
    },
    {
      "epoch": 0.2649951367124176,
      "grad_norm": 0.35743921995162964,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 2452
    },
    {
      "epoch": 0.2651032097698044,
      "grad_norm": 0.41128188371658325,
      "learning_rate": 8e-05,
      "loss": 1.7489,
      "step": 2453
    },
    {
      "epoch": 0.2652112828271912,
      "grad_norm": 0.38887327909469604,
      "learning_rate": 8e-05,
      "loss": 1.8813,
      "step": 2454
    },
    {
      "epoch": 0.265319355884578,
      "grad_norm": 0.3744156062602997,
      "learning_rate": 8e-05,
      "loss": 1.6251,
      "step": 2455
    },
    {
      "epoch": 0.26542742894196475,
      "grad_norm": 0.4440920948982239,
      "learning_rate": 8e-05,
      "loss": 1.9888,
      "step": 2456
    },
    {
      "epoch": 0.26553550199935155,
      "grad_norm": 0.3793190121650696,
      "learning_rate": 8e-05,
      "loss": 1.6845,
      "step": 2457
    },
    {
      "epoch": 0.26564357505673836,
      "grad_norm": 0.3661603331565857,
      "learning_rate": 8e-05,
      "loss": 1.6758,
      "step": 2458
    },
    {
      "epoch": 0.26575164811412516,
      "grad_norm": 0.36692652106285095,
      "learning_rate": 8e-05,
      "loss": 1.5304,
      "step": 2459
    },
    {
      "epoch": 0.26585972117151196,
      "grad_norm": 0.39514121413230896,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 2460
    },
    {
      "epoch": 0.2659677942288987,
      "grad_norm": 0.3803490698337555,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 2461
    },
    {
      "epoch": 0.2660758672862855,
      "grad_norm": 0.4119362533092499,
      "learning_rate": 8e-05,
      "loss": 1.9486,
      "step": 2462
    },
    {
      "epoch": 0.2661839403436723,
      "grad_norm": 0.3772425055503845,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 2463
    },
    {
      "epoch": 0.2662920134010591,
      "grad_norm": 0.37792736291885376,
      "learning_rate": 8e-05,
      "loss": 1.7267,
      "step": 2464
    },
    {
      "epoch": 0.26640008645844593,
      "grad_norm": 0.3979353606700897,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 2465
    },
    {
      "epoch": 0.2665081595158327,
      "grad_norm": 0.3579016625881195,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 2466
    },
    {
      "epoch": 0.2666162325732195,
      "grad_norm": 0.39699140191078186,
      "learning_rate": 8e-05,
      "loss": 1.7969,
      "step": 2467
    },
    {
      "epoch": 0.2667243056306063,
      "grad_norm": 0.36525195837020874,
      "learning_rate": 8e-05,
      "loss": 1.7161,
      "step": 2468
    },
    {
      "epoch": 0.2668323786879931,
      "grad_norm": 0.369540810585022,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 2469
    },
    {
      "epoch": 0.2669404517453799,
      "grad_norm": 0.37650567293167114,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 2470
    },
    {
      "epoch": 0.2670485248027667,
      "grad_norm": 0.33726438879966736,
      "learning_rate": 8e-05,
      "loss": 1.6949,
      "step": 2471
    },
    {
      "epoch": 0.26715659786015344,
      "grad_norm": 0.38358521461486816,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 2472
    },
    {
      "epoch": 0.26726467091754025,
      "grad_norm": 0.3594326376914978,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 2473
    },
    {
      "epoch": 0.26737274397492705,
      "grad_norm": 0.40659692883491516,
      "learning_rate": 8e-05,
      "loss": 1.369,
      "step": 2474
    },
    {
      "epoch": 0.26748081703231386,
      "grad_norm": 0.3636024594306946,
      "learning_rate": 8e-05,
      "loss": 1.6558,
      "step": 2475
    },
    {
      "epoch": 0.26758889008970066,
      "grad_norm": 0.3700731098651886,
      "learning_rate": 8e-05,
      "loss": 1.663,
      "step": 2476
    },
    {
      "epoch": 0.2676969631470874,
      "grad_norm": 0.3898457884788513,
      "learning_rate": 8e-05,
      "loss": 1.8759,
      "step": 2477
    },
    {
      "epoch": 0.2678050362044742,
      "grad_norm": 0.3857495188713074,
      "learning_rate": 8e-05,
      "loss": 1.5405,
      "step": 2478
    },
    {
      "epoch": 0.267913109261861,
      "grad_norm": 0.36643165349960327,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 2479
    },
    {
      "epoch": 0.2680211823192478,
      "grad_norm": 0.3721325993537903,
      "learning_rate": 8e-05,
      "loss": 1.6934,
      "step": 2480
    },
    {
      "epoch": 0.2681292553766346,
      "grad_norm": 0.36624324321746826,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 2481
    },
    {
      "epoch": 0.26823732843402137,
      "grad_norm": 0.36084693670272827,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 2482
    },
    {
      "epoch": 0.2683454014914082,
      "grad_norm": 0.3541159927845001,
      "learning_rate": 8e-05,
      "loss": 1.6488,
      "step": 2483
    },
    {
      "epoch": 0.268453474548795,
      "grad_norm": 0.4120561182498932,
      "learning_rate": 8e-05,
      "loss": 1.7216,
      "step": 2484
    },
    {
      "epoch": 0.2685615476061818,
      "grad_norm": 0.3740830421447754,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 2485
    },
    {
      "epoch": 0.2686696206635686,
      "grad_norm": 0.3531447649002075,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 2486
    },
    {
      "epoch": 0.2687776937209554,
      "grad_norm": 0.3760266602039337,
      "learning_rate": 8e-05,
      "loss": 1.8165,
      "step": 2487
    },
    {
      "epoch": 0.26888576677834214,
      "grad_norm": 0.4254160225391388,
      "learning_rate": 8e-05,
      "loss": 1.4274,
      "step": 2488
    },
    {
      "epoch": 0.26899383983572894,
      "grad_norm": 0.34750789403915405,
      "learning_rate": 8e-05,
      "loss": 1.4991,
      "step": 2489
    },
    {
      "epoch": 0.26910191289311575,
      "grad_norm": 0.4012928605079651,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 2490
    },
    {
      "epoch": 0.26920998595050255,
      "grad_norm": 0.3861410915851593,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 2491
    },
    {
      "epoch": 0.26931805900788935,
      "grad_norm": 0.3659304678440094,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 2492
    },
    {
      "epoch": 0.2694261320652761,
      "grad_norm": 0.42097416520118713,
      "learning_rate": 8e-05,
      "loss": 1.8566,
      "step": 2493
    },
    {
      "epoch": 0.2695342051226629,
      "grad_norm": 0.36721569299697876,
      "learning_rate": 8e-05,
      "loss": 1.7708,
      "step": 2494
    },
    {
      "epoch": 0.2696422781800497,
      "grad_norm": 0.40271109342575073,
      "learning_rate": 8e-05,
      "loss": 1.6643,
      "step": 2495
    },
    {
      "epoch": 0.2697503512374365,
      "grad_norm": 0.37576308846473694,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 2496
    },
    {
      "epoch": 0.2698584242948233,
      "grad_norm": 0.3525030314922333,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 2497
    },
    {
      "epoch": 0.26996649735221007,
      "grad_norm": 0.41718024015426636,
      "learning_rate": 8e-05,
      "loss": 1.8849,
      "step": 2498
    },
    {
      "epoch": 0.27007457040959687,
      "grad_norm": 0.382348895072937,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 2499
    },
    {
      "epoch": 0.2701826434669837,
      "grad_norm": 0.4037102460861206,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 2500
    },
    {
      "epoch": 0.2702907165243705,
      "grad_norm": 0.37202468514442444,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 2501
    },
    {
      "epoch": 0.2703987895817573,
      "grad_norm": 0.3814954161643982,
      "learning_rate": 8e-05,
      "loss": 1.7672,
      "step": 2502
    },
    {
      "epoch": 0.2705068626391441,
      "grad_norm": 0.37354743480682373,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 2503
    },
    {
      "epoch": 0.27061493569653083,
      "grad_norm": 0.3903687298297882,
      "learning_rate": 8e-05,
      "loss": 1.7209,
      "step": 2504
    },
    {
      "epoch": 0.27072300875391764,
      "grad_norm": 0.4233676791191101,
      "learning_rate": 8e-05,
      "loss": 1.8452,
      "step": 2505
    },
    {
      "epoch": 0.27083108181130444,
      "grad_norm": 0.3956805169582367,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 2506
    },
    {
      "epoch": 0.27093915486869125,
      "grad_norm": 0.3765985369682312,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 2507
    },
    {
      "epoch": 0.27104722792607805,
      "grad_norm": 0.3490865230560303,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 2508
    },
    {
      "epoch": 0.2711553009834648,
      "grad_norm": 0.3582180440425873,
      "learning_rate": 8e-05,
      "loss": 1.6532,
      "step": 2509
    },
    {
      "epoch": 0.2712633740408516,
      "grad_norm": 0.37423941493034363,
      "learning_rate": 8e-05,
      "loss": 1.6049,
      "step": 2510
    },
    {
      "epoch": 0.2713714470982384,
      "grad_norm": 0.39894378185272217,
      "learning_rate": 8e-05,
      "loss": 1.6756,
      "step": 2511
    },
    {
      "epoch": 0.2714795201556252,
      "grad_norm": 0.3725018799304962,
      "learning_rate": 8e-05,
      "loss": 1.7469,
      "step": 2512
    },
    {
      "epoch": 0.271587593213012,
      "grad_norm": 0.3482353091239929,
      "learning_rate": 8e-05,
      "loss": 1.5849,
      "step": 2513
    },
    {
      "epoch": 0.2716956662703988,
      "grad_norm": 0.3945613205432892,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 2514
    },
    {
      "epoch": 0.27180373932778557,
      "grad_norm": 0.3775917589664459,
      "learning_rate": 8e-05,
      "loss": 1.8485,
      "step": 2515
    },
    {
      "epoch": 0.27191181238517237,
      "grad_norm": 0.36745280027389526,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 2516
    },
    {
      "epoch": 0.2720198854425592,
      "grad_norm": 0.36604225635528564,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 2517
    },
    {
      "epoch": 0.272127958499946,
      "grad_norm": 0.4445498585700989,
      "learning_rate": 8e-05,
      "loss": 2.0939,
      "step": 2518
    },
    {
      "epoch": 0.2722360315573328,
      "grad_norm": 0.3474128842353821,
      "learning_rate": 8e-05,
      "loss": 1.613,
      "step": 2519
    },
    {
      "epoch": 0.27234410461471953,
      "grad_norm": 0.35501158237457275,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 2520
    },
    {
      "epoch": 0.27245217767210633,
      "grad_norm": 0.38934776186943054,
      "learning_rate": 8e-05,
      "loss": 1.8804,
      "step": 2521
    },
    {
      "epoch": 0.27256025072949314,
      "grad_norm": 0.37549129128456116,
      "learning_rate": 8e-05,
      "loss": 1.8113,
      "step": 2522
    },
    {
      "epoch": 0.27266832378687994,
      "grad_norm": 0.34787410497665405,
      "learning_rate": 8e-05,
      "loss": 1.5699,
      "step": 2523
    },
    {
      "epoch": 0.27277639684426674,
      "grad_norm": 0.3693847358226776,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 2524
    },
    {
      "epoch": 0.2728844699016535,
      "grad_norm": 0.380143404006958,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 2525
    },
    {
      "epoch": 0.2729925429590403,
      "grad_norm": 0.3778119385242462,
      "learning_rate": 8e-05,
      "loss": 1.7387,
      "step": 2526
    },
    {
      "epoch": 0.2731006160164271,
      "grad_norm": 0.3620222210884094,
      "learning_rate": 8e-05,
      "loss": 1.6545,
      "step": 2527
    },
    {
      "epoch": 0.2732086890738139,
      "grad_norm": 0.36658886075019836,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 2528
    },
    {
      "epoch": 0.2733167621312007,
      "grad_norm": 0.413739949464798,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 2529
    },
    {
      "epoch": 0.2734248351885875,
      "grad_norm": 0.37141475081443787,
      "learning_rate": 8e-05,
      "loss": 1.7975,
      "step": 2530
    },
    {
      "epoch": 0.27353290824597426,
      "grad_norm": 0.3921930491924286,
      "learning_rate": 8e-05,
      "loss": 1.7961,
      "step": 2531
    },
    {
      "epoch": 0.27364098130336106,
      "grad_norm": 0.3504860997200012,
      "learning_rate": 8e-05,
      "loss": 1.7217,
      "step": 2532
    },
    {
      "epoch": 0.27374905436074787,
      "grad_norm": 0.374895840883255,
      "learning_rate": 8e-05,
      "loss": 1.8226,
      "step": 2533
    },
    {
      "epoch": 0.27385712741813467,
      "grad_norm": 0.392419695854187,
      "learning_rate": 8e-05,
      "loss": 1.7938,
      "step": 2534
    },
    {
      "epoch": 0.2739652004755215,
      "grad_norm": 0.35362112522125244,
      "learning_rate": 8e-05,
      "loss": 1.7076,
      "step": 2535
    },
    {
      "epoch": 0.2740732735329082,
      "grad_norm": 0.3471091091632843,
      "learning_rate": 8e-05,
      "loss": 1.5148,
      "step": 2536
    },
    {
      "epoch": 0.274181346590295,
      "grad_norm": 0.3957866430282593,
      "learning_rate": 8e-05,
      "loss": 1.8398,
      "step": 2537
    },
    {
      "epoch": 0.27428941964768183,
      "grad_norm": 0.3853392004966736,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 2538
    },
    {
      "epoch": 0.27439749270506864,
      "grad_norm": 0.3762071430683136,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 2539
    },
    {
      "epoch": 0.27450556576245544,
      "grad_norm": 0.3879365026950836,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 2540
    },
    {
      "epoch": 0.2746136388198422,
      "grad_norm": 0.3753284811973572,
      "learning_rate": 8e-05,
      "loss": 1.6732,
      "step": 2541
    },
    {
      "epoch": 0.274721711877229,
      "grad_norm": 0.4130113422870636,
      "learning_rate": 8e-05,
      "loss": 1.8967,
      "step": 2542
    },
    {
      "epoch": 0.2748297849346158,
      "grad_norm": 0.4417731761932373,
      "learning_rate": 8e-05,
      "loss": 1.8916,
      "step": 2543
    },
    {
      "epoch": 0.2749378579920026,
      "grad_norm": 0.35213807225227356,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 2544
    },
    {
      "epoch": 0.2750459310493894,
      "grad_norm": 0.38566669821739197,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 2545
    },
    {
      "epoch": 0.2751540041067762,
      "grad_norm": 0.36589381098747253,
      "learning_rate": 8e-05,
      "loss": 1.6623,
      "step": 2546
    },
    {
      "epoch": 0.27526207716416295,
      "grad_norm": 0.3881056308746338,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 2547
    },
    {
      "epoch": 0.27537015022154976,
      "grad_norm": 0.35593181848526,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 2548
    },
    {
      "epoch": 0.27547822327893656,
      "grad_norm": 0.38285768032073975,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 2549
    },
    {
      "epoch": 0.27558629633632337,
      "grad_norm": 0.34929969906806946,
      "learning_rate": 8e-05,
      "loss": 1.6237,
      "step": 2550
    },
    {
      "epoch": 0.27569436939371017,
      "grad_norm": 0.38450849056243896,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 2551
    },
    {
      "epoch": 0.2758024424510969,
      "grad_norm": 0.40309420228004456,
      "learning_rate": 8e-05,
      "loss": 1.877,
      "step": 2552
    },
    {
      "epoch": 0.2759105155084837,
      "grad_norm": 0.37091633677482605,
      "learning_rate": 8e-05,
      "loss": 1.7201,
      "step": 2553
    },
    {
      "epoch": 0.2760185885658705,
      "grad_norm": 0.3870861232280731,
      "learning_rate": 8e-05,
      "loss": 1.7462,
      "step": 2554
    },
    {
      "epoch": 0.27612666162325733,
      "grad_norm": 0.4026245176792145,
      "learning_rate": 8e-05,
      "loss": 1.848,
      "step": 2555
    },
    {
      "epoch": 0.27623473468064413,
      "grad_norm": 0.43537789583206177,
      "learning_rate": 8e-05,
      "loss": 1.826,
      "step": 2556
    },
    {
      "epoch": 0.2763428077380309,
      "grad_norm": 0.3637314736843109,
      "learning_rate": 8e-05,
      "loss": 1.8177,
      "step": 2557
    },
    {
      "epoch": 0.2764508807954177,
      "grad_norm": 0.3755403459072113,
      "learning_rate": 8e-05,
      "loss": 1.5909,
      "step": 2558
    },
    {
      "epoch": 0.2765589538528045,
      "grad_norm": 0.4269295632839203,
      "learning_rate": 8e-05,
      "loss": 1.86,
      "step": 2559
    },
    {
      "epoch": 0.2766670269101913,
      "grad_norm": 0.4731540381908417,
      "learning_rate": 8e-05,
      "loss": 1.8046,
      "step": 2560
    },
    {
      "epoch": 0.2767750999675781,
      "grad_norm": 0.43560025095939636,
      "learning_rate": 8e-05,
      "loss": 1.8978,
      "step": 2561
    },
    {
      "epoch": 0.2768831730249649,
      "grad_norm": 0.362594872713089,
      "learning_rate": 8e-05,
      "loss": 1.6746,
      "step": 2562
    },
    {
      "epoch": 0.27699124608235165,
      "grad_norm": 0.3803568482398987,
      "learning_rate": 8e-05,
      "loss": 1.6798,
      "step": 2563
    },
    {
      "epoch": 0.27709931913973845,
      "grad_norm": 0.42912617325782776,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 2564
    },
    {
      "epoch": 0.27720739219712526,
      "grad_norm": 0.382752388715744,
      "learning_rate": 8e-05,
      "loss": 1.6952,
      "step": 2565
    },
    {
      "epoch": 0.27731546525451206,
      "grad_norm": 0.3976055979728699,
      "learning_rate": 8e-05,
      "loss": 1.8627,
      "step": 2566
    },
    {
      "epoch": 0.27742353831189887,
      "grad_norm": 0.43283385038375854,
      "learning_rate": 8e-05,
      "loss": 1.9287,
      "step": 2567
    },
    {
      "epoch": 0.2775316113692856,
      "grad_norm": 0.406622976064682,
      "learning_rate": 8e-05,
      "loss": 1.6796,
      "step": 2568
    },
    {
      "epoch": 0.2776396844266724,
      "grad_norm": 0.43080392479896545,
      "learning_rate": 8e-05,
      "loss": 1.8439,
      "step": 2569
    },
    {
      "epoch": 0.2777477574840592,
      "grad_norm": 0.3766738474369049,
      "learning_rate": 8e-05,
      "loss": 1.7337,
      "step": 2570
    },
    {
      "epoch": 0.277855830541446,
      "grad_norm": 0.3888819217681885,
      "learning_rate": 8e-05,
      "loss": 1.7595,
      "step": 2571
    },
    {
      "epoch": 0.27796390359883283,
      "grad_norm": 0.3656492531299591,
      "learning_rate": 8e-05,
      "loss": 1.4745,
      "step": 2572
    },
    {
      "epoch": 0.2780719766562196,
      "grad_norm": 0.37483128905296326,
      "learning_rate": 8e-05,
      "loss": 1.7019,
      "step": 2573
    },
    {
      "epoch": 0.2781800497136064,
      "grad_norm": 0.3898029029369354,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 2574
    },
    {
      "epoch": 0.2782881227709932,
      "grad_norm": 0.39038920402526855,
      "learning_rate": 8e-05,
      "loss": 1.8807,
      "step": 2575
    },
    {
      "epoch": 0.27839619582838,
      "grad_norm": 0.36786362528800964,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 2576
    },
    {
      "epoch": 0.2785042688857668,
      "grad_norm": 0.37234652042388916,
      "learning_rate": 8e-05,
      "loss": 1.633,
      "step": 2577
    },
    {
      "epoch": 0.2786123419431536,
      "grad_norm": 0.3831823468208313,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 2578
    },
    {
      "epoch": 0.27872041500054034,
      "grad_norm": 0.4305756986141205,
      "learning_rate": 8e-05,
      "loss": 1.8839,
      "step": 2579
    },
    {
      "epoch": 0.27882848805792715,
      "grad_norm": 0.3725205063819885,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 2580
    },
    {
      "epoch": 0.27893656111531395,
      "grad_norm": 0.38674360513687134,
      "learning_rate": 8e-05,
      "loss": 1.7909,
      "step": 2581
    },
    {
      "epoch": 0.27904463417270076,
      "grad_norm": 0.4043315052986145,
      "learning_rate": 8e-05,
      "loss": 1.9964,
      "step": 2582
    },
    {
      "epoch": 0.27915270723008756,
      "grad_norm": 0.37745967507362366,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 2583
    },
    {
      "epoch": 0.2792607802874743,
      "grad_norm": 0.38301882147789,
      "learning_rate": 8e-05,
      "loss": 1.8636,
      "step": 2584
    },
    {
      "epoch": 0.2793688533448611,
      "grad_norm": 0.38333404064178467,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 2585
    },
    {
      "epoch": 0.2794769264022479,
      "grad_norm": 0.3695251941680908,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 2586
    },
    {
      "epoch": 0.2795849994596347,
      "grad_norm": 0.3937545418739319,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 2587
    },
    {
      "epoch": 0.2796930725170215,
      "grad_norm": 0.3846130967140198,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 2588
    },
    {
      "epoch": 0.2798011455744083,
      "grad_norm": 0.34087884426116943,
      "learning_rate": 8e-05,
      "loss": 1.5742,
      "step": 2589
    },
    {
      "epoch": 0.2799092186317951,
      "grad_norm": 0.3758355379104614,
      "learning_rate": 8e-05,
      "loss": 1.7123,
      "step": 2590
    },
    {
      "epoch": 0.2800172916891819,
      "grad_norm": 0.3709929585456848,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 2591
    },
    {
      "epoch": 0.2801253647465687,
      "grad_norm": 0.39020630717277527,
      "learning_rate": 8e-05,
      "loss": 1.7082,
      "step": 2592
    },
    {
      "epoch": 0.2802334378039555,
      "grad_norm": 0.3844773769378662,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 2593
    },
    {
      "epoch": 0.2803415108613423,
      "grad_norm": 0.3910813629627228,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 2594
    },
    {
      "epoch": 0.28044958391872904,
      "grad_norm": 0.3727092742919922,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 2595
    },
    {
      "epoch": 0.28055765697611584,
      "grad_norm": 0.3810289204120636,
      "learning_rate": 8e-05,
      "loss": 1.7389,
      "step": 2596
    },
    {
      "epoch": 0.28066573003350265,
      "grad_norm": 0.4203589856624603,
      "learning_rate": 8e-05,
      "loss": 1.8516,
      "step": 2597
    },
    {
      "epoch": 0.28077380309088945,
      "grad_norm": 0.36351585388183594,
      "learning_rate": 8e-05,
      "loss": 1.7073,
      "step": 2598
    },
    {
      "epoch": 0.28088187614827625,
      "grad_norm": 0.373172402381897,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 2599
    },
    {
      "epoch": 0.280989949205663,
      "grad_norm": 0.35284096002578735,
      "learning_rate": 8e-05,
      "loss": 1.6921,
      "step": 2600
    },
    {
      "epoch": 0.2810980222630498,
      "grad_norm": 0.40052998065948486,
      "learning_rate": 8e-05,
      "loss": 1.8309,
      "step": 2601
    },
    {
      "epoch": 0.2812060953204366,
      "grad_norm": 0.38429924845695496,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 2602
    },
    {
      "epoch": 0.2813141683778234,
      "grad_norm": 0.37491101026535034,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 2603
    },
    {
      "epoch": 0.2814222414352102,
      "grad_norm": 0.398975133895874,
      "learning_rate": 8e-05,
      "loss": 1.803,
      "step": 2604
    },
    {
      "epoch": 0.281530314492597,
      "grad_norm": 0.3995000422000885,
      "learning_rate": 8e-05,
      "loss": 1.8905,
      "step": 2605
    },
    {
      "epoch": 0.28163838754998377,
      "grad_norm": 0.37232160568237305,
      "learning_rate": 8e-05,
      "loss": 1.5507,
      "step": 2606
    },
    {
      "epoch": 0.2817464606073706,
      "grad_norm": 0.3866870701313019,
      "learning_rate": 8e-05,
      "loss": 1.8756,
      "step": 2607
    },
    {
      "epoch": 0.2818545336647574,
      "grad_norm": 0.371979683637619,
      "learning_rate": 8e-05,
      "loss": 1.8679,
      "step": 2608
    },
    {
      "epoch": 0.2819626067221442,
      "grad_norm": 0.3651829659938812,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 2609
    },
    {
      "epoch": 0.282070679779531,
      "grad_norm": 0.3690573275089264,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 2610
    },
    {
      "epoch": 0.28217875283691773,
      "grad_norm": 0.3713589906692505,
      "learning_rate": 8e-05,
      "loss": 1.5919,
      "step": 2611
    },
    {
      "epoch": 0.28228682589430454,
      "grad_norm": 0.3814350962638855,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 2612
    },
    {
      "epoch": 0.28239489895169134,
      "grad_norm": 0.40340250730514526,
      "learning_rate": 8e-05,
      "loss": 1.862,
      "step": 2613
    },
    {
      "epoch": 0.28250297200907815,
      "grad_norm": 0.3941473364830017,
      "learning_rate": 8e-05,
      "loss": 1.9048,
      "step": 2614
    },
    {
      "epoch": 0.28261104506646495,
      "grad_norm": 0.41903766989707947,
      "learning_rate": 8e-05,
      "loss": 1.8632,
      "step": 2615
    },
    {
      "epoch": 0.2827191181238517,
      "grad_norm": 0.3816618323326111,
      "learning_rate": 8e-05,
      "loss": 1.7798,
      "step": 2616
    },
    {
      "epoch": 0.2828271911812385,
      "grad_norm": 0.3765718936920166,
      "learning_rate": 8e-05,
      "loss": 1.5974,
      "step": 2617
    },
    {
      "epoch": 0.2829352642386253,
      "grad_norm": 0.4068194329738617,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 2618
    },
    {
      "epoch": 0.2830433372960121,
      "grad_norm": 0.3559759855270386,
      "learning_rate": 8e-05,
      "loss": 1.6816,
      "step": 2619
    },
    {
      "epoch": 0.2831514103533989,
      "grad_norm": 0.36500853300094604,
      "learning_rate": 8e-05,
      "loss": 1.6264,
      "step": 2620
    },
    {
      "epoch": 0.2832594834107857,
      "grad_norm": 0.3958582580089569,
      "learning_rate": 8e-05,
      "loss": 1.5652,
      "step": 2621
    },
    {
      "epoch": 0.28336755646817247,
      "grad_norm": 0.4033443331718445,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 2622
    },
    {
      "epoch": 0.28347562952555927,
      "grad_norm": 0.3446609675884247,
      "learning_rate": 8e-05,
      "loss": 1.5903,
      "step": 2623
    },
    {
      "epoch": 0.2835837025829461,
      "grad_norm": 0.40247175097465515,
      "learning_rate": 8e-05,
      "loss": 1.8313,
      "step": 2624
    },
    {
      "epoch": 0.2836917756403329,
      "grad_norm": 0.4228101372718811,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 2625
    },
    {
      "epoch": 0.2837998486977197,
      "grad_norm": 0.4078630208969116,
      "learning_rate": 8e-05,
      "loss": 1.8554,
      "step": 2626
    },
    {
      "epoch": 0.28390792175510643,
      "grad_norm": 0.36746469140052795,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 2627
    },
    {
      "epoch": 0.28401599481249323,
      "grad_norm": 0.3927038609981537,
      "learning_rate": 8e-05,
      "loss": 1.817,
      "step": 2628
    },
    {
      "epoch": 0.28412406786988004,
      "grad_norm": 0.37484753131866455,
      "learning_rate": 8e-05,
      "loss": 1.6864,
      "step": 2629
    },
    {
      "epoch": 0.28423214092726684,
      "grad_norm": 0.3721831440925598,
      "learning_rate": 8e-05,
      "loss": 1.6445,
      "step": 2630
    },
    {
      "epoch": 0.28434021398465364,
      "grad_norm": 0.3869171142578125,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 2631
    },
    {
      "epoch": 0.2844482870420404,
      "grad_norm": 0.37088584899902344,
      "learning_rate": 8e-05,
      "loss": 1.585,
      "step": 2632
    },
    {
      "epoch": 0.2845563600994272,
      "grad_norm": 0.3893742859363556,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 2633
    },
    {
      "epoch": 0.284664433156814,
      "grad_norm": 0.3971605896949768,
      "learning_rate": 8e-05,
      "loss": 1.7925,
      "step": 2634
    },
    {
      "epoch": 0.2847725062142008,
      "grad_norm": 0.3455650806427002,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 2635
    },
    {
      "epoch": 0.2848805792715876,
      "grad_norm": 0.383194237947464,
      "learning_rate": 8e-05,
      "loss": 1.7238,
      "step": 2636
    },
    {
      "epoch": 0.2849886523289744,
      "grad_norm": 0.39578551054000854,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 2637
    },
    {
      "epoch": 0.28509672538636116,
      "grad_norm": 0.3673846423625946,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 2638
    },
    {
      "epoch": 0.28520479844374796,
      "grad_norm": 0.40245911478996277,
      "learning_rate": 8e-05,
      "loss": 1.9308,
      "step": 2639
    },
    {
      "epoch": 0.28531287150113477,
      "grad_norm": 0.35533931851387024,
      "learning_rate": 8e-05,
      "loss": 1.5659,
      "step": 2640
    },
    {
      "epoch": 0.28542094455852157,
      "grad_norm": 0.36477547883987427,
      "learning_rate": 8e-05,
      "loss": 1.5876,
      "step": 2641
    },
    {
      "epoch": 0.2855290176159084,
      "grad_norm": 0.3914869427680969,
      "learning_rate": 8e-05,
      "loss": 1.7271,
      "step": 2642
    },
    {
      "epoch": 0.2856370906732951,
      "grad_norm": 0.3821012079715729,
      "learning_rate": 8e-05,
      "loss": 1.8567,
      "step": 2643
    },
    {
      "epoch": 0.28574516373068193,
      "grad_norm": 0.4113089144229889,
      "learning_rate": 8e-05,
      "loss": 2.0372,
      "step": 2644
    },
    {
      "epoch": 0.28585323678806873,
      "grad_norm": 0.3847944438457489,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 2645
    },
    {
      "epoch": 0.28596130984545554,
      "grad_norm": 0.386808842420578,
      "learning_rate": 8e-05,
      "loss": 1.7791,
      "step": 2646
    },
    {
      "epoch": 0.28606938290284234,
      "grad_norm": 0.39247071743011475,
      "learning_rate": 8e-05,
      "loss": 1.7785,
      "step": 2647
    },
    {
      "epoch": 0.2861774559602291,
      "grad_norm": 0.3872281014919281,
      "learning_rate": 8e-05,
      "loss": 1.7778,
      "step": 2648
    },
    {
      "epoch": 0.2862855290176159,
      "grad_norm": 0.4142041504383087,
      "learning_rate": 8e-05,
      "loss": 1.801,
      "step": 2649
    },
    {
      "epoch": 0.2863936020750027,
      "grad_norm": 0.38411226868629456,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 2650
    },
    {
      "epoch": 0.2865016751323895,
      "grad_norm": 0.3829296827316284,
      "learning_rate": 8e-05,
      "loss": 1.7458,
      "step": 2651
    },
    {
      "epoch": 0.2866097481897763,
      "grad_norm": 0.37057220935821533,
      "learning_rate": 8e-05,
      "loss": 1.8394,
      "step": 2652
    },
    {
      "epoch": 0.2867178212471631,
      "grad_norm": 0.37490999698638916,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 2653
    },
    {
      "epoch": 0.28682589430454986,
      "grad_norm": 0.3798205554485321,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 2654
    },
    {
      "epoch": 0.28693396736193666,
      "grad_norm": 0.42090654373168945,
      "learning_rate": 8e-05,
      "loss": 1.576,
      "step": 2655
    },
    {
      "epoch": 0.28704204041932346,
      "grad_norm": 0.363699734210968,
      "learning_rate": 8e-05,
      "loss": 1.7029,
      "step": 2656
    },
    {
      "epoch": 0.28715011347671027,
      "grad_norm": 0.36247774958610535,
      "learning_rate": 8e-05,
      "loss": 1.6486,
      "step": 2657
    },
    {
      "epoch": 0.28725818653409707,
      "grad_norm": 0.4152909815311432,
      "learning_rate": 8e-05,
      "loss": 1.805,
      "step": 2658
    },
    {
      "epoch": 0.2873662595914838,
      "grad_norm": 0.39140433073043823,
      "learning_rate": 8e-05,
      "loss": 1.9045,
      "step": 2659
    },
    {
      "epoch": 0.2874743326488706,
      "grad_norm": 0.4135834872722626,
      "learning_rate": 8e-05,
      "loss": 1.9079,
      "step": 2660
    },
    {
      "epoch": 0.2875824057062574,
      "grad_norm": 0.43586188554763794,
      "learning_rate": 8e-05,
      "loss": 2.0383,
      "step": 2661
    },
    {
      "epoch": 0.28769047876364423,
      "grad_norm": 0.3908523619174957,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 2662
    },
    {
      "epoch": 0.28779855182103103,
      "grad_norm": 0.4065014123916626,
      "learning_rate": 8e-05,
      "loss": 1.6589,
      "step": 2663
    },
    {
      "epoch": 0.2879066248784178,
      "grad_norm": 0.3792342245578766,
      "learning_rate": 8e-05,
      "loss": 1.8533,
      "step": 2664
    },
    {
      "epoch": 0.2880146979358046,
      "grad_norm": 0.3709142506122589,
      "learning_rate": 8e-05,
      "loss": 1.8187,
      "step": 2665
    },
    {
      "epoch": 0.2881227709931914,
      "grad_norm": 0.37438422441482544,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 2666
    },
    {
      "epoch": 0.2882308440505782,
      "grad_norm": 0.3955494165420532,
      "learning_rate": 8e-05,
      "loss": 1.6588,
      "step": 2667
    },
    {
      "epoch": 0.288338917107965,
      "grad_norm": 0.3724851906299591,
      "learning_rate": 8e-05,
      "loss": 1.6725,
      "step": 2668
    },
    {
      "epoch": 0.2884469901653518,
      "grad_norm": 0.36863845586776733,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 2669
    },
    {
      "epoch": 0.28855506322273855,
      "grad_norm": 0.40622618794441223,
      "learning_rate": 8e-05,
      "loss": 1.9388,
      "step": 2670
    },
    {
      "epoch": 0.28866313628012535,
      "grad_norm": 0.4573462903499603,
      "learning_rate": 8e-05,
      "loss": 1.9513,
      "step": 2671
    },
    {
      "epoch": 0.28877120933751216,
      "grad_norm": 0.37362560629844666,
      "learning_rate": 8e-05,
      "loss": 1.7882,
      "step": 2672
    },
    {
      "epoch": 0.28887928239489896,
      "grad_norm": 0.4047364890575409,
      "learning_rate": 8e-05,
      "loss": 1.9614,
      "step": 2673
    },
    {
      "epoch": 0.28898735545228577,
      "grad_norm": 0.37733224034309387,
      "learning_rate": 8e-05,
      "loss": 1.7854,
      "step": 2674
    },
    {
      "epoch": 0.2890954285096725,
      "grad_norm": 0.3852332830429077,
      "learning_rate": 8e-05,
      "loss": 1.6166,
      "step": 2675
    },
    {
      "epoch": 0.2892035015670593,
      "grad_norm": 0.39140352606773376,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 2676
    },
    {
      "epoch": 0.2893115746244461,
      "grad_norm": 0.4005804657936096,
      "learning_rate": 8e-05,
      "loss": 1.8852,
      "step": 2677
    },
    {
      "epoch": 0.2894196476818329,
      "grad_norm": 0.3712862730026245,
      "learning_rate": 8e-05,
      "loss": 1.4719,
      "step": 2678
    },
    {
      "epoch": 0.28952772073921973,
      "grad_norm": 0.3938542604446411,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 2679
    },
    {
      "epoch": 0.28963579379660653,
      "grad_norm": 0.3647788465023041,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 2680
    },
    {
      "epoch": 0.2897438668539933,
      "grad_norm": 0.3877812325954437,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 2681
    },
    {
      "epoch": 0.2898519399113801,
      "grad_norm": 0.35888177156448364,
      "learning_rate": 8e-05,
      "loss": 1.7229,
      "step": 2682
    },
    {
      "epoch": 0.2899600129687669,
      "grad_norm": 0.3767542243003845,
      "learning_rate": 8e-05,
      "loss": 1.7956,
      "step": 2683
    },
    {
      "epoch": 0.2900680860261537,
      "grad_norm": 0.3777533173561096,
      "learning_rate": 8e-05,
      "loss": 1.8302,
      "step": 2684
    },
    {
      "epoch": 0.2901761590835405,
      "grad_norm": 0.36561882495880127,
      "learning_rate": 8e-05,
      "loss": 1.8739,
      "step": 2685
    },
    {
      "epoch": 0.29028423214092725,
      "grad_norm": 0.39992254972457886,
      "learning_rate": 8e-05,
      "loss": 1.7131,
      "step": 2686
    },
    {
      "epoch": 0.29039230519831405,
      "grad_norm": 0.3876323699951172,
      "learning_rate": 8e-05,
      "loss": 1.5777,
      "step": 2687
    },
    {
      "epoch": 0.29050037825570085,
      "grad_norm": 0.3957829475402832,
      "learning_rate": 8e-05,
      "loss": 1.7931,
      "step": 2688
    },
    {
      "epoch": 0.29060845131308766,
      "grad_norm": 0.40311309695243835,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 2689
    },
    {
      "epoch": 0.29071652437047446,
      "grad_norm": 0.40472134947776794,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 2690
    },
    {
      "epoch": 0.2908245974278612,
      "grad_norm": 0.3906451463699341,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 2691
    },
    {
      "epoch": 0.290932670485248,
      "grad_norm": 0.4069863259792328,
      "learning_rate": 8e-05,
      "loss": 1.8099,
      "step": 2692
    },
    {
      "epoch": 0.2910407435426348,
      "grad_norm": 0.3760221004486084,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 2693
    },
    {
      "epoch": 0.2911488166000216,
      "grad_norm": 0.43671557307243347,
      "learning_rate": 8e-05,
      "loss": 1.8761,
      "step": 2694
    },
    {
      "epoch": 0.2912568896574084,
      "grad_norm": 0.38148486614227295,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 2695
    },
    {
      "epoch": 0.29136496271479523,
      "grad_norm": 0.3758547604084015,
      "learning_rate": 8e-05,
      "loss": 1.7192,
      "step": 2696
    },
    {
      "epoch": 0.291473035772182,
      "grad_norm": 0.4103824496269226,
      "learning_rate": 8e-05,
      "loss": 1.8662,
      "step": 2697
    },
    {
      "epoch": 0.2915811088295688,
      "grad_norm": 0.34544193744659424,
      "learning_rate": 8e-05,
      "loss": 1.601,
      "step": 2698
    },
    {
      "epoch": 0.2916891818869556,
      "grad_norm": 0.39112764596939087,
      "learning_rate": 8e-05,
      "loss": 1.8879,
      "step": 2699
    },
    {
      "epoch": 0.2917972549443424,
      "grad_norm": 0.40787893533706665,
      "learning_rate": 8e-05,
      "loss": 1.7701,
      "step": 2700
    },
    {
      "epoch": 0.2919053280017292,
      "grad_norm": 0.3879365622997284,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 2701
    },
    {
      "epoch": 0.29201340105911594,
      "grad_norm": 0.3981607258319855,
      "learning_rate": 8e-05,
      "loss": 1.8479,
      "step": 2702
    },
    {
      "epoch": 0.29212147411650274,
      "grad_norm": 0.39715006947517395,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 2703
    },
    {
      "epoch": 0.29222954717388955,
      "grad_norm": 0.3840593099594116,
      "learning_rate": 8e-05,
      "loss": 1.8175,
      "step": 2704
    },
    {
      "epoch": 0.29233762023127635,
      "grad_norm": 0.34705591201782227,
      "learning_rate": 8e-05,
      "loss": 1.583,
      "step": 2705
    },
    {
      "epoch": 0.29244569328866316,
      "grad_norm": 0.3966079354286194,
      "learning_rate": 8e-05,
      "loss": 1.8207,
      "step": 2706
    },
    {
      "epoch": 0.2925537663460499,
      "grad_norm": 0.39517587423324585,
      "learning_rate": 8e-05,
      "loss": 1.6633,
      "step": 2707
    },
    {
      "epoch": 0.2926618394034367,
      "grad_norm": 0.40129369497299194,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 2708
    },
    {
      "epoch": 0.2927699124608235,
      "grad_norm": 0.37463781237602234,
      "learning_rate": 8e-05,
      "loss": 1.5864,
      "step": 2709
    },
    {
      "epoch": 0.2928779855182103,
      "grad_norm": 0.3682762384414673,
      "learning_rate": 8e-05,
      "loss": 1.6139,
      "step": 2710
    },
    {
      "epoch": 0.2929860585755971,
      "grad_norm": 0.398734450340271,
      "learning_rate": 8e-05,
      "loss": 1.8334,
      "step": 2711
    },
    {
      "epoch": 0.2930941316329839,
      "grad_norm": 0.3927443325519562,
      "learning_rate": 8e-05,
      "loss": 1.8187,
      "step": 2712
    },
    {
      "epoch": 0.29320220469037067,
      "grad_norm": 0.39369529485702515,
      "learning_rate": 8e-05,
      "loss": 1.9222,
      "step": 2713
    },
    {
      "epoch": 0.2933102777477575,
      "grad_norm": 0.4231705367565155,
      "learning_rate": 8e-05,
      "loss": 1.7626,
      "step": 2714
    },
    {
      "epoch": 0.2934183508051443,
      "grad_norm": 0.34941771626472473,
      "learning_rate": 8e-05,
      "loss": 1.7377,
      "step": 2715
    },
    {
      "epoch": 0.2935264238625311,
      "grad_norm": 0.371358722448349,
      "learning_rate": 8e-05,
      "loss": 1.7137,
      "step": 2716
    },
    {
      "epoch": 0.2936344969199179,
      "grad_norm": 0.34836313128471375,
      "learning_rate": 8e-05,
      "loss": 1.6093,
      "step": 2717
    },
    {
      "epoch": 0.29374256997730464,
      "grad_norm": 0.3492712676525116,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 2718
    },
    {
      "epoch": 0.29385064303469144,
      "grad_norm": 0.36492788791656494,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 2719
    },
    {
      "epoch": 0.29395871609207824,
      "grad_norm": 0.3967876434326172,
      "learning_rate": 8e-05,
      "loss": 1.8939,
      "step": 2720
    },
    {
      "epoch": 0.29406678914946505,
      "grad_norm": 0.39690834283828735,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 2721
    },
    {
      "epoch": 0.29417486220685185,
      "grad_norm": 0.3604986071586609,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 2722
    },
    {
      "epoch": 0.2942829352642386,
      "grad_norm": 0.3591908812522888,
      "learning_rate": 8e-05,
      "loss": 1.6981,
      "step": 2723
    },
    {
      "epoch": 0.2943910083216254,
      "grad_norm": 0.3666079044342041,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 2724
    },
    {
      "epoch": 0.2944990813790122,
      "grad_norm": 0.39117714762687683,
      "learning_rate": 8e-05,
      "loss": 1.7832,
      "step": 2725
    },
    {
      "epoch": 0.294607154436399,
      "grad_norm": 0.36729124188423157,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 2726
    },
    {
      "epoch": 0.2947152274937858,
      "grad_norm": 0.3750307559967041,
      "learning_rate": 8e-05,
      "loss": 1.8592,
      "step": 2727
    },
    {
      "epoch": 0.2948233005511726,
      "grad_norm": 0.3781711757183075,
      "learning_rate": 8e-05,
      "loss": 1.7022,
      "step": 2728
    },
    {
      "epoch": 0.29493137360855937,
      "grad_norm": 0.3596431016921997,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 2729
    },
    {
      "epoch": 0.29503944666594617,
      "grad_norm": 0.37114667892456055,
      "learning_rate": 8e-05,
      "loss": 1.6202,
      "step": 2730
    },
    {
      "epoch": 0.295147519723333,
      "grad_norm": 0.3953438997268677,
      "learning_rate": 8e-05,
      "loss": 1.7305,
      "step": 2731
    },
    {
      "epoch": 0.2952555927807198,
      "grad_norm": 0.36601582169532776,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 2732
    },
    {
      "epoch": 0.2953636658381066,
      "grad_norm": 0.3728499412536621,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 2733
    },
    {
      "epoch": 0.29547173889549333,
      "grad_norm": 0.4207839071750641,
      "learning_rate": 8e-05,
      "loss": 1.8033,
      "step": 2734
    },
    {
      "epoch": 0.29557981195288013,
      "grad_norm": 0.3564106822013855,
      "learning_rate": 8e-05,
      "loss": 1.6848,
      "step": 2735
    },
    {
      "epoch": 0.29568788501026694,
      "grad_norm": 0.393136203289032,
      "learning_rate": 8e-05,
      "loss": 1.8076,
      "step": 2736
    },
    {
      "epoch": 0.29579595806765374,
      "grad_norm": 0.3664186894893646,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 2737
    },
    {
      "epoch": 0.29590403112504055,
      "grad_norm": 0.40782058238983154,
      "learning_rate": 8e-05,
      "loss": 1.9878,
      "step": 2738
    },
    {
      "epoch": 0.2960121041824273,
      "grad_norm": 0.45472556352615356,
      "learning_rate": 8e-05,
      "loss": 2.0088,
      "step": 2739
    },
    {
      "epoch": 0.2961201772398141,
      "grad_norm": 0.40115949511528015,
      "learning_rate": 8e-05,
      "loss": 1.7765,
      "step": 2740
    },
    {
      "epoch": 0.2962282502972009,
      "grad_norm": 0.39794114232063293,
      "learning_rate": 8e-05,
      "loss": 1.724,
      "step": 2741
    },
    {
      "epoch": 0.2963363233545877,
      "grad_norm": 0.36665233969688416,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 2742
    },
    {
      "epoch": 0.2964443964119745,
      "grad_norm": 0.4037715494632721,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 2743
    },
    {
      "epoch": 0.2965524694693613,
      "grad_norm": 0.396152526140213,
      "learning_rate": 8e-05,
      "loss": 1.8993,
      "step": 2744
    },
    {
      "epoch": 0.29666054252674806,
      "grad_norm": 0.37332990765571594,
      "learning_rate": 8e-05,
      "loss": 1.6967,
      "step": 2745
    },
    {
      "epoch": 0.29676861558413486,
      "grad_norm": 0.4242761433124542,
      "learning_rate": 8e-05,
      "loss": 1.8793,
      "step": 2746
    },
    {
      "epoch": 0.29687668864152167,
      "grad_norm": 0.35311606526374817,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 2747
    },
    {
      "epoch": 0.2969847616989085,
      "grad_norm": 0.3555096983909607,
      "learning_rate": 8e-05,
      "loss": 1.5858,
      "step": 2748
    },
    {
      "epoch": 0.2970928347562953,
      "grad_norm": 0.3928786814212799,
      "learning_rate": 8e-05,
      "loss": 1.8462,
      "step": 2749
    },
    {
      "epoch": 0.297200907813682,
      "grad_norm": 0.40916478633880615,
      "learning_rate": 8e-05,
      "loss": 1.9759,
      "step": 2750
    },
    {
      "epoch": 0.29730898087106883,
      "grad_norm": 0.36609891057014465,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 2751
    },
    {
      "epoch": 0.29741705392845563,
      "grad_norm": 0.37210315465927124,
      "learning_rate": 8e-05,
      "loss": 1.806,
      "step": 2752
    },
    {
      "epoch": 0.29752512698584244,
      "grad_norm": 0.3899574279785156,
      "learning_rate": 8e-05,
      "loss": 1.9027,
      "step": 2753
    },
    {
      "epoch": 0.29763320004322924,
      "grad_norm": 0.3988085389137268,
      "learning_rate": 8e-05,
      "loss": 1.6155,
      "step": 2754
    },
    {
      "epoch": 0.29774127310061604,
      "grad_norm": 0.3713085651397705,
      "learning_rate": 8e-05,
      "loss": 1.8582,
      "step": 2755
    },
    {
      "epoch": 0.2978493461580028,
      "grad_norm": 0.3908194303512573,
      "learning_rate": 8e-05,
      "loss": 1.7409,
      "step": 2756
    },
    {
      "epoch": 0.2979574192153896,
      "grad_norm": 0.41051924228668213,
      "learning_rate": 8e-05,
      "loss": 1.668,
      "step": 2757
    },
    {
      "epoch": 0.2980654922727764,
      "grad_norm": 0.3800656795501709,
      "learning_rate": 8e-05,
      "loss": 1.8168,
      "step": 2758
    },
    {
      "epoch": 0.2981735653301632,
      "grad_norm": 0.38788944482803345,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 2759
    },
    {
      "epoch": 0.29828163838755,
      "grad_norm": 0.3720516860485077,
      "learning_rate": 8e-05,
      "loss": 1.8035,
      "step": 2760
    },
    {
      "epoch": 0.29838971144493676,
      "grad_norm": 0.3911111354827881,
      "learning_rate": 8e-05,
      "loss": 1.8724,
      "step": 2761
    },
    {
      "epoch": 0.29849778450232356,
      "grad_norm": 0.39373984932899475,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 2762
    },
    {
      "epoch": 0.29860585755971036,
      "grad_norm": 0.3459130525588989,
      "learning_rate": 8e-05,
      "loss": 1.6581,
      "step": 2763
    },
    {
      "epoch": 0.29871393061709717,
      "grad_norm": 0.39422255754470825,
      "learning_rate": 8e-05,
      "loss": 1.6024,
      "step": 2764
    },
    {
      "epoch": 0.29882200367448397,
      "grad_norm": 0.3514842391014099,
      "learning_rate": 8e-05,
      "loss": 1.6137,
      "step": 2765
    },
    {
      "epoch": 0.2989300767318707,
      "grad_norm": 0.3713415563106537,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 2766
    },
    {
      "epoch": 0.2990381497892575,
      "grad_norm": 0.38729971647262573,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 2767
    },
    {
      "epoch": 0.2991462228466443,
      "grad_norm": 0.36966803669929504,
      "learning_rate": 8e-05,
      "loss": 1.6488,
      "step": 2768
    },
    {
      "epoch": 0.29925429590403113,
      "grad_norm": 0.3689405620098114,
      "learning_rate": 8e-05,
      "loss": 1.781,
      "step": 2769
    },
    {
      "epoch": 0.29936236896141794,
      "grad_norm": 0.3855065405368805,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 2770
    },
    {
      "epoch": 0.29947044201880474,
      "grad_norm": 0.36936676502227783,
      "learning_rate": 8e-05,
      "loss": 1.7675,
      "step": 2771
    },
    {
      "epoch": 0.2995785150761915,
      "grad_norm": 0.39187923073768616,
      "learning_rate": 8e-05,
      "loss": 1.7893,
      "step": 2772
    },
    {
      "epoch": 0.2996865881335783,
      "grad_norm": 0.369331955909729,
      "learning_rate": 8e-05,
      "loss": 1.7343,
      "step": 2773
    },
    {
      "epoch": 0.2997946611909651,
      "grad_norm": 0.38518908619880676,
      "learning_rate": 8e-05,
      "loss": 1.6393,
      "step": 2774
    },
    {
      "epoch": 0.2999027342483519,
      "grad_norm": 0.382506787776947,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 2775
    },
    {
      "epoch": 0.3000108073057387,
      "grad_norm": 0.5179709792137146,
      "learning_rate": 8e-05,
      "loss": 2.0043,
      "step": 2776
    },
    {
      "epoch": 0.30011888036312545,
      "grad_norm": 0.4107379913330078,
      "learning_rate": 8e-05,
      "loss": 1.6966,
      "step": 2777
    },
    {
      "epoch": 0.30022695342051225,
      "grad_norm": 0.41739919781684875,
      "learning_rate": 8e-05,
      "loss": 1.8737,
      "step": 2778
    },
    {
      "epoch": 0.30033502647789906,
      "grad_norm": 0.4223654270172119,
      "learning_rate": 8e-05,
      "loss": 1.8622,
      "step": 2779
    },
    {
      "epoch": 0.30044309953528586,
      "grad_norm": 0.38513776659965515,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 2780
    },
    {
      "epoch": 0.30055117259267267,
      "grad_norm": 0.35849669575691223,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 2781
    },
    {
      "epoch": 0.3006592456500594,
      "grad_norm": 0.42787253856658936,
      "learning_rate": 8e-05,
      "loss": 1.9259,
      "step": 2782
    },
    {
      "epoch": 0.3007673187074462,
      "grad_norm": 0.3671242296695709,
      "learning_rate": 8e-05,
      "loss": 1.6095,
      "step": 2783
    },
    {
      "epoch": 0.300875391764833,
      "grad_norm": 0.38985350728034973,
      "learning_rate": 8e-05,
      "loss": 1.8322,
      "step": 2784
    },
    {
      "epoch": 0.3009834648222198,
      "grad_norm": 0.37536555528640747,
      "learning_rate": 8e-05,
      "loss": 1.8342,
      "step": 2785
    },
    {
      "epoch": 0.30109153787960663,
      "grad_norm": 0.3566596508026123,
      "learning_rate": 8e-05,
      "loss": 1.6076,
      "step": 2786
    },
    {
      "epoch": 0.30119961093699343,
      "grad_norm": 0.38576146960258484,
      "learning_rate": 8e-05,
      "loss": 1.5171,
      "step": 2787
    },
    {
      "epoch": 0.3013076839943802,
      "grad_norm": 0.3565886616706848,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 2788
    },
    {
      "epoch": 0.301415757051767,
      "grad_norm": 0.3559359312057495,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 2789
    },
    {
      "epoch": 0.3015238301091538,
      "grad_norm": 0.35267412662506104,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 2790
    },
    {
      "epoch": 0.3016319031665406,
      "grad_norm": 0.39178356528282166,
      "learning_rate": 8e-05,
      "loss": 1.8374,
      "step": 2791
    },
    {
      "epoch": 0.3017399762239274,
      "grad_norm": 0.37979385256767273,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 2792
    },
    {
      "epoch": 0.30184804928131415,
      "grad_norm": 0.3830778896808624,
      "learning_rate": 8e-05,
      "loss": 1.7141,
      "step": 2793
    },
    {
      "epoch": 0.30195612233870095,
      "grad_norm": 0.3955959379673004,
      "learning_rate": 8e-05,
      "loss": 1.8396,
      "step": 2794
    },
    {
      "epoch": 0.30206419539608775,
      "grad_norm": 0.3882089853286743,
      "learning_rate": 8e-05,
      "loss": 1.7059,
      "step": 2795
    },
    {
      "epoch": 0.30217226845347456,
      "grad_norm": 0.3677847981452942,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 2796
    },
    {
      "epoch": 0.30228034151086136,
      "grad_norm": 0.3848576545715332,
      "learning_rate": 8e-05,
      "loss": 1.7194,
      "step": 2797
    },
    {
      "epoch": 0.3023884145682481,
      "grad_norm": 0.3704456388950348,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 2798
    },
    {
      "epoch": 0.3024964876256349,
      "grad_norm": 0.3719424307346344,
      "learning_rate": 8e-05,
      "loss": 1.7165,
      "step": 2799
    },
    {
      "epoch": 0.3026045606830217,
      "grad_norm": 0.39718568325042725,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 2800
    },
    {
      "epoch": 0.3027126337404085,
      "grad_norm": 0.3685368597507477,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 2801
    },
    {
      "epoch": 0.3028207067977953,
      "grad_norm": 0.4045741558074951,
      "learning_rate": 8e-05,
      "loss": 1.7183,
      "step": 2802
    },
    {
      "epoch": 0.30292877985518213,
      "grad_norm": 0.38734057545661926,
      "learning_rate": 8e-05,
      "loss": 1.7426,
      "step": 2803
    },
    {
      "epoch": 0.3030368529125689,
      "grad_norm": 0.40937939286231995,
      "learning_rate": 8e-05,
      "loss": 1.9377,
      "step": 2804
    },
    {
      "epoch": 0.3031449259699557,
      "grad_norm": 0.41106492280960083,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 2805
    },
    {
      "epoch": 0.3032529990273425,
      "grad_norm": 0.3860428035259247,
      "learning_rate": 8e-05,
      "loss": 1.7771,
      "step": 2806
    },
    {
      "epoch": 0.3033610720847293,
      "grad_norm": 0.38663002848625183,
      "learning_rate": 8e-05,
      "loss": 1.6858,
      "step": 2807
    },
    {
      "epoch": 0.3034691451421161,
      "grad_norm": 0.40180304646492004,
      "learning_rate": 8e-05,
      "loss": 1.7658,
      "step": 2808
    },
    {
      "epoch": 0.30357721819950284,
      "grad_norm": 0.437874436378479,
      "learning_rate": 8e-05,
      "loss": 1.7222,
      "step": 2809
    },
    {
      "epoch": 0.30368529125688964,
      "grad_norm": 0.39362409710884094,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 2810
    },
    {
      "epoch": 0.30379336431427645,
      "grad_norm": 0.41877835988998413,
      "learning_rate": 8e-05,
      "loss": 1.9216,
      "step": 2811
    },
    {
      "epoch": 0.30390143737166325,
      "grad_norm": 0.3777587115764618,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 2812
    },
    {
      "epoch": 0.30400951042905006,
      "grad_norm": 0.38361722230911255,
      "learning_rate": 8e-05,
      "loss": 1.7395,
      "step": 2813
    },
    {
      "epoch": 0.3041175834864368,
      "grad_norm": 0.3810099959373474,
      "learning_rate": 8e-05,
      "loss": 1.8252,
      "step": 2814
    },
    {
      "epoch": 0.3042256565438236,
      "grad_norm": 0.385941743850708,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 2815
    },
    {
      "epoch": 0.3043337296012104,
      "grad_norm": 0.3661251664161682,
      "learning_rate": 8e-05,
      "loss": 1.5477,
      "step": 2816
    },
    {
      "epoch": 0.3044418026585972,
      "grad_norm": 0.4016244113445282,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 2817
    },
    {
      "epoch": 0.304549875715984,
      "grad_norm": 0.37685319781303406,
      "learning_rate": 8e-05,
      "loss": 1.6606,
      "step": 2818
    },
    {
      "epoch": 0.3046579487733708,
      "grad_norm": 0.3665071725845337,
      "learning_rate": 8e-05,
      "loss": 1.8231,
      "step": 2819
    },
    {
      "epoch": 0.30476602183075757,
      "grad_norm": 0.3693774938583374,
      "learning_rate": 8e-05,
      "loss": 1.7373,
      "step": 2820
    },
    {
      "epoch": 0.3048740948881444,
      "grad_norm": 0.3791259229183197,
      "learning_rate": 8e-05,
      "loss": 1.8299,
      "step": 2821
    },
    {
      "epoch": 0.3049821679455312,
      "grad_norm": 0.35839998722076416,
      "learning_rate": 8e-05,
      "loss": 1.6922,
      "step": 2822
    },
    {
      "epoch": 0.305090241002918,
      "grad_norm": 0.38520944118499756,
      "learning_rate": 8e-05,
      "loss": 1.7927,
      "step": 2823
    },
    {
      "epoch": 0.3051983140603048,
      "grad_norm": 0.3816578686237335,
      "learning_rate": 8e-05,
      "loss": 1.7264,
      "step": 2824
    },
    {
      "epoch": 0.30530638711769154,
      "grad_norm": 0.38569384813308716,
      "learning_rate": 8e-05,
      "loss": 1.8728,
      "step": 2825
    },
    {
      "epoch": 0.30541446017507834,
      "grad_norm": 0.370116651058197,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 2826
    },
    {
      "epoch": 0.30552253323246514,
      "grad_norm": 0.36661067605018616,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 2827
    },
    {
      "epoch": 0.30563060628985195,
      "grad_norm": 0.3710159957408905,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 2828
    },
    {
      "epoch": 0.30573867934723875,
      "grad_norm": 0.3679962158203125,
      "learning_rate": 8e-05,
      "loss": 1.8645,
      "step": 2829
    },
    {
      "epoch": 0.3058467524046255,
      "grad_norm": 0.41197913885116577,
      "learning_rate": 8e-05,
      "loss": 1.8282,
      "step": 2830
    },
    {
      "epoch": 0.3059548254620123,
      "grad_norm": 0.37579745054244995,
      "learning_rate": 8e-05,
      "loss": 1.6248,
      "step": 2831
    },
    {
      "epoch": 0.3060628985193991,
      "grad_norm": 0.3733593225479126,
      "learning_rate": 8e-05,
      "loss": 1.6092,
      "step": 2832
    },
    {
      "epoch": 0.3061709715767859,
      "grad_norm": 0.36421963572502136,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 2833
    },
    {
      "epoch": 0.3062790446341727,
      "grad_norm": 0.3788909614086151,
      "learning_rate": 8e-05,
      "loss": 1.8222,
      "step": 2834
    },
    {
      "epoch": 0.3063871176915595,
      "grad_norm": 0.4105842709541321,
      "learning_rate": 8e-05,
      "loss": 1.7551,
      "step": 2835
    },
    {
      "epoch": 0.30649519074894627,
      "grad_norm": 0.3690252900123596,
      "learning_rate": 8e-05,
      "loss": 1.7652,
      "step": 2836
    },
    {
      "epoch": 0.30660326380633307,
      "grad_norm": 0.38524535298347473,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 2837
    },
    {
      "epoch": 0.3067113368637199,
      "grad_norm": 0.35578757524490356,
      "learning_rate": 8e-05,
      "loss": 1.6818,
      "step": 2838
    },
    {
      "epoch": 0.3068194099211067,
      "grad_norm": 0.4109478294849396,
      "learning_rate": 8e-05,
      "loss": 1.6761,
      "step": 2839
    },
    {
      "epoch": 0.3069274829784935,
      "grad_norm": 0.3837120532989502,
      "learning_rate": 8e-05,
      "loss": 1.5612,
      "step": 2840
    },
    {
      "epoch": 0.30703555603588023,
      "grad_norm": 0.3675451874732971,
      "learning_rate": 8e-05,
      "loss": 1.818,
      "step": 2841
    },
    {
      "epoch": 0.30714362909326703,
      "grad_norm": 0.3980099558830261,
      "learning_rate": 8e-05,
      "loss": 1.9134,
      "step": 2842
    },
    {
      "epoch": 0.30725170215065384,
      "grad_norm": 0.36603352427482605,
      "learning_rate": 8e-05,
      "loss": 1.7407,
      "step": 2843
    },
    {
      "epoch": 0.30735977520804064,
      "grad_norm": 0.3835507035255432,
      "learning_rate": 8e-05,
      "loss": 1.7966,
      "step": 2844
    },
    {
      "epoch": 0.30746784826542745,
      "grad_norm": 0.3695836663246155,
      "learning_rate": 8e-05,
      "loss": 1.5319,
      "step": 2845
    },
    {
      "epoch": 0.30757592132281425,
      "grad_norm": 0.3849472403526306,
      "learning_rate": 8e-05,
      "loss": 1.8283,
      "step": 2846
    },
    {
      "epoch": 0.307683994380201,
      "grad_norm": 0.4207453727722168,
      "learning_rate": 8e-05,
      "loss": 1.8893,
      "step": 2847
    },
    {
      "epoch": 0.3077920674375878,
      "grad_norm": 0.395069420337677,
      "learning_rate": 8e-05,
      "loss": 1.766,
      "step": 2848
    },
    {
      "epoch": 0.3079001404949746,
      "grad_norm": 0.4042578637599945,
      "learning_rate": 8e-05,
      "loss": 1.8761,
      "step": 2849
    },
    {
      "epoch": 0.3080082135523614,
      "grad_norm": 0.37202179431915283,
      "learning_rate": 8e-05,
      "loss": 1.8068,
      "step": 2850
    },
    {
      "epoch": 0.3081162866097482,
      "grad_norm": 0.3904436528682709,
      "learning_rate": 8e-05,
      "loss": 1.8431,
      "step": 2851
    },
    {
      "epoch": 0.30822435966713496,
      "grad_norm": 0.44059136509895325,
      "learning_rate": 8e-05,
      "loss": 1.8889,
      "step": 2852
    },
    {
      "epoch": 0.30833243272452177,
      "grad_norm": 0.39557191729545593,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 2853
    },
    {
      "epoch": 0.30844050578190857,
      "grad_norm": 0.39514783024787903,
      "learning_rate": 8e-05,
      "loss": 1.7284,
      "step": 2854
    },
    {
      "epoch": 0.3085485788392954,
      "grad_norm": 0.3819649815559387,
      "learning_rate": 8e-05,
      "loss": 1.7746,
      "step": 2855
    },
    {
      "epoch": 0.3086566518966822,
      "grad_norm": 0.3580262064933777,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 2856
    },
    {
      "epoch": 0.3087647249540689,
      "grad_norm": 0.367521733045578,
      "learning_rate": 8e-05,
      "loss": 1.8224,
      "step": 2857
    },
    {
      "epoch": 0.30887279801145573,
      "grad_norm": 0.35109928250312805,
      "learning_rate": 8e-05,
      "loss": 1.5895,
      "step": 2858
    },
    {
      "epoch": 0.30898087106884253,
      "grad_norm": 0.4035426676273346,
      "learning_rate": 8e-05,
      "loss": 1.931,
      "step": 2859
    },
    {
      "epoch": 0.30908894412622934,
      "grad_norm": 0.41659969091415405,
      "learning_rate": 8e-05,
      "loss": 1.906,
      "step": 2860
    },
    {
      "epoch": 0.30919701718361614,
      "grad_norm": 0.4266480803489685,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 2861
    },
    {
      "epoch": 0.30930509024100294,
      "grad_norm": 0.38545137643814087,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 2862
    },
    {
      "epoch": 0.3094131632983897,
      "grad_norm": 0.41232702136039734,
      "learning_rate": 8e-05,
      "loss": 1.9435,
      "step": 2863
    },
    {
      "epoch": 0.3095212363557765,
      "grad_norm": 0.365920752286911,
      "learning_rate": 8e-05,
      "loss": 1.7095,
      "step": 2864
    },
    {
      "epoch": 0.3096293094131633,
      "grad_norm": 0.38208508491516113,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 2865
    },
    {
      "epoch": 0.3097373824705501,
      "grad_norm": 0.3837342858314514,
      "learning_rate": 8e-05,
      "loss": 1.5845,
      "step": 2866
    },
    {
      "epoch": 0.3098454555279369,
      "grad_norm": 0.3689322769641876,
      "learning_rate": 8e-05,
      "loss": 1.7638,
      "step": 2867
    },
    {
      "epoch": 0.30995352858532366,
      "grad_norm": 0.36687833070755005,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 2868
    },
    {
      "epoch": 0.31006160164271046,
      "grad_norm": 0.4071006178855896,
      "learning_rate": 8e-05,
      "loss": 1.9323,
      "step": 2869
    },
    {
      "epoch": 0.31016967470009726,
      "grad_norm": 0.36862632632255554,
      "learning_rate": 8e-05,
      "loss": 1.7744,
      "step": 2870
    },
    {
      "epoch": 0.31027774775748407,
      "grad_norm": 0.3692541718482971,
      "learning_rate": 8e-05,
      "loss": 1.8932,
      "step": 2871
    },
    {
      "epoch": 0.31038582081487087,
      "grad_norm": 0.41158992052078247,
      "learning_rate": 8e-05,
      "loss": 1.8115,
      "step": 2872
    },
    {
      "epoch": 0.3104938938722576,
      "grad_norm": 0.36422809958457947,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 2873
    },
    {
      "epoch": 0.3106019669296444,
      "grad_norm": 0.4066080152988434,
      "learning_rate": 8e-05,
      "loss": 1.9813,
      "step": 2874
    },
    {
      "epoch": 0.31071003998703123,
      "grad_norm": 0.43168526887893677,
      "learning_rate": 8e-05,
      "loss": 1.9608,
      "step": 2875
    },
    {
      "epoch": 0.31081811304441803,
      "grad_norm": 0.39238473773002625,
      "learning_rate": 8e-05,
      "loss": 1.8645,
      "step": 2876
    },
    {
      "epoch": 0.31092618610180484,
      "grad_norm": 0.38773417472839355,
      "learning_rate": 8e-05,
      "loss": 1.8767,
      "step": 2877
    },
    {
      "epoch": 0.31103425915919164,
      "grad_norm": 0.38922005891799927,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 2878
    },
    {
      "epoch": 0.3111423322165784,
      "grad_norm": 0.35113298892974854,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 2879
    },
    {
      "epoch": 0.3112504052739652,
      "grad_norm": 0.39437413215637207,
      "learning_rate": 8e-05,
      "loss": 1.9242,
      "step": 2880
    },
    {
      "epoch": 0.311358478331352,
      "grad_norm": 0.39073893427848816,
      "learning_rate": 8e-05,
      "loss": 1.9062,
      "step": 2881
    },
    {
      "epoch": 0.3114665513887388,
      "grad_norm": 0.3760479688644409,
      "learning_rate": 8e-05,
      "loss": 1.7592,
      "step": 2882
    },
    {
      "epoch": 0.3115746244461256,
      "grad_norm": 0.38166531920433044,
      "learning_rate": 8e-05,
      "loss": 1.716,
      "step": 2883
    },
    {
      "epoch": 0.31168269750351235,
      "grad_norm": 0.3855013847351074,
      "learning_rate": 8e-05,
      "loss": 1.479,
      "step": 2884
    },
    {
      "epoch": 0.31179077056089916,
      "grad_norm": 0.41183751821517944,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 2885
    },
    {
      "epoch": 0.31189884361828596,
      "grad_norm": 0.399077832698822,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 2886
    },
    {
      "epoch": 0.31200691667567276,
      "grad_norm": 0.3941633403301239,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 2887
    },
    {
      "epoch": 0.31211498973305957,
      "grad_norm": 0.374554842710495,
      "learning_rate": 8e-05,
      "loss": 1.6653,
      "step": 2888
    },
    {
      "epoch": 0.3122230627904463,
      "grad_norm": 0.3725683093070984,
      "learning_rate": 8e-05,
      "loss": 1.648,
      "step": 2889
    },
    {
      "epoch": 0.3123311358478331,
      "grad_norm": 0.3631357252597809,
      "learning_rate": 8e-05,
      "loss": 1.5924,
      "step": 2890
    },
    {
      "epoch": 0.3124392089052199,
      "grad_norm": 0.39798495173454285,
      "learning_rate": 8e-05,
      "loss": 1.7096,
      "step": 2891
    },
    {
      "epoch": 0.3125472819626067,
      "grad_norm": 0.39654138684272766,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 2892
    },
    {
      "epoch": 0.31265535501999353,
      "grad_norm": 0.4354475736618042,
      "learning_rate": 8e-05,
      "loss": 1.9467,
      "step": 2893
    },
    {
      "epoch": 0.31276342807738033,
      "grad_norm": 0.4001348316669464,
      "learning_rate": 8e-05,
      "loss": 1.866,
      "step": 2894
    },
    {
      "epoch": 0.3128715011347671,
      "grad_norm": 0.3934697210788727,
      "learning_rate": 8e-05,
      "loss": 1.7055,
      "step": 2895
    },
    {
      "epoch": 0.3129795741921539,
      "grad_norm": 0.4336510896682739,
      "learning_rate": 8e-05,
      "loss": 1.8726,
      "step": 2896
    },
    {
      "epoch": 0.3130876472495407,
      "grad_norm": 0.35438022017478943,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 2897
    },
    {
      "epoch": 0.3131957203069275,
      "grad_norm": 0.40890607237815857,
      "learning_rate": 8e-05,
      "loss": 1.7566,
      "step": 2898
    },
    {
      "epoch": 0.3133037933643143,
      "grad_norm": 0.41066208481788635,
      "learning_rate": 8e-05,
      "loss": 1.7447,
      "step": 2899
    },
    {
      "epoch": 0.31341186642170105,
      "grad_norm": 0.368725448846817,
      "learning_rate": 8e-05,
      "loss": 1.6884,
      "step": 2900
    },
    {
      "epoch": 0.31351993947908785,
      "grad_norm": 0.4108510911464691,
      "learning_rate": 8e-05,
      "loss": 1.9705,
      "step": 2901
    },
    {
      "epoch": 0.31362801253647465,
      "grad_norm": 0.3954737186431885,
      "learning_rate": 8e-05,
      "loss": 1.7379,
      "step": 2902
    },
    {
      "epoch": 0.31373608559386146,
      "grad_norm": 0.44700920581817627,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 2903
    },
    {
      "epoch": 0.31384415865124826,
      "grad_norm": 0.33998164534568787,
      "learning_rate": 8e-05,
      "loss": 1.5107,
      "step": 2904
    },
    {
      "epoch": 0.313952231708635,
      "grad_norm": 0.35271117091178894,
      "learning_rate": 8e-05,
      "loss": 1.6449,
      "step": 2905
    },
    {
      "epoch": 0.3140603047660218,
      "grad_norm": 0.4002251923084259,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 2906
    },
    {
      "epoch": 0.3141683778234086,
      "grad_norm": 0.40751349925994873,
      "learning_rate": 8e-05,
      "loss": 1.726,
      "step": 2907
    },
    {
      "epoch": 0.3142764508807954,
      "grad_norm": 0.38629281520843506,
      "learning_rate": 8e-05,
      "loss": 1.7872,
      "step": 2908
    },
    {
      "epoch": 0.3143845239381822,
      "grad_norm": 0.38523009419441223,
      "learning_rate": 8e-05,
      "loss": 1.8642,
      "step": 2909
    },
    {
      "epoch": 0.31449259699556903,
      "grad_norm": 0.4356428384780884,
      "learning_rate": 8e-05,
      "loss": 1.8041,
      "step": 2910
    },
    {
      "epoch": 0.3146006700529558,
      "grad_norm": 0.386976957321167,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 2911
    },
    {
      "epoch": 0.3147087431103426,
      "grad_norm": 0.4100055992603302,
      "learning_rate": 8e-05,
      "loss": 1.8026,
      "step": 2912
    },
    {
      "epoch": 0.3148168161677294,
      "grad_norm": 0.3868879973888397,
      "learning_rate": 8e-05,
      "loss": 1.5354,
      "step": 2913
    },
    {
      "epoch": 0.3149248892251162,
      "grad_norm": 0.359494686126709,
      "learning_rate": 8e-05,
      "loss": 1.6117,
      "step": 2914
    },
    {
      "epoch": 0.315032962282503,
      "grad_norm": 0.3833145201206207,
      "learning_rate": 8e-05,
      "loss": 1.7569,
      "step": 2915
    },
    {
      "epoch": 0.31514103533988974,
      "grad_norm": 0.4158884584903717,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 2916
    },
    {
      "epoch": 0.31524910839727655,
      "grad_norm": 0.41443008184432983,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 2917
    },
    {
      "epoch": 0.31535718145466335,
      "grad_norm": 0.35206183791160583,
      "learning_rate": 8e-05,
      "loss": 1.5787,
      "step": 2918
    },
    {
      "epoch": 0.31546525451205015,
      "grad_norm": 0.3966812491416931,
      "learning_rate": 8e-05,
      "loss": 1.6348,
      "step": 2919
    },
    {
      "epoch": 0.31557332756943696,
      "grad_norm": 0.39897847175598145,
      "learning_rate": 8e-05,
      "loss": 1.658,
      "step": 2920
    },
    {
      "epoch": 0.31568140062682376,
      "grad_norm": 0.3685523271560669,
      "learning_rate": 8e-05,
      "loss": 1.6447,
      "step": 2921
    },
    {
      "epoch": 0.3157894736842105,
      "grad_norm": 0.37183451652526855,
      "learning_rate": 8e-05,
      "loss": 1.5332,
      "step": 2922
    },
    {
      "epoch": 0.3158975467415973,
      "grad_norm": 0.3534536063671112,
      "learning_rate": 8e-05,
      "loss": 1.5273,
      "step": 2923
    },
    {
      "epoch": 0.3160056197989841,
      "grad_norm": 0.36801114678382874,
      "learning_rate": 8e-05,
      "loss": 1.8019,
      "step": 2924
    },
    {
      "epoch": 0.3161136928563709,
      "grad_norm": 0.3689340353012085,
      "learning_rate": 8e-05,
      "loss": 1.6638,
      "step": 2925
    },
    {
      "epoch": 0.3162217659137577,
      "grad_norm": 0.37881961464881897,
      "learning_rate": 8e-05,
      "loss": 1.7105,
      "step": 2926
    },
    {
      "epoch": 0.3163298389711445,
      "grad_norm": 0.37032562494277954,
      "learning_rate": 8e-05,
      "loss": 1.7243,
      "step": 2927
    },
    {
      "epoch": 0.3164379120285313,
      "grad_norm": 0.40114015340805054,
      "learning_rate": 8e-05,
      "loss": 1.814,
      "step": 2928
    },
    {
      "epoch": 0.3165459850859181,
      "grad_norm": 0.4087194502353668,
      "learning_rate": 8e-05,
      "loss": 1.7618,
      "step": 2929
    },
    {
      "epoch": 0.3166540581433049,
      "grad_norm": 0.44881680607795715,
      "learning_rate": 8e-05,
      "loss": 1.8863,
      "step": 2930
    },
    {
      "epoch": 0.3167621312006917,
      "grad_norm": 0.3572789430618286,
      "learning_rate": 8e-05,
      "loss": 1.7184,
      "step": 2931
    },
    {
      "epoch": 0.31687020425807844,
      "grad_norm": 0.3730814754962921,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 2932
    },
    {
      "epoch": 0.31697827731546524,
      "grad_norm": 0.40190377831459045,
      "learning_rate": 8e-05,
      "loss": 1.8714,
      "step": 2933
    },
    {
      "epoch": 0.31708635037285204,
      "grad_norm": 0.3877315819263458,
      "learning_rate": 8e-05,
      "loss": 1.7679,
      "step": 2934
    },
    {
      "epoch": 0.31719442343023885,
      "grad_norm": 0.35322630405426025,
      "learning_rate": 8e-05,
      "loss": 1.4556,
      "step": 2935
    },
    {
      "epoch": 0.31730249648762565,
      "grad_norm": 0.42431527376174927,
      "learning_rate": 8e-05,
      "loss": 1.8214,
      "step": 2936
    },
    {
      "epoch": 0.31741056954501246,
      "grad_norm": 0.3871442675590515,
      "learning_rate": 8e-05,
      "loss": 1.7399,
      "step": 2937
    },
    {
      "epoch": 0.3175186426023992,
      "grad_norm": 0.39737236499786377,
      "learning_rate": 8e-05,
      "loss": 1.8192,
      "step": 2938
    },
    {
      "epoch": 0.317626715659786,
      "grad_norm": 0.4183724522590637,
      "learning_rate": 8e-05,
      "loss": 1.8086,
      "step": 2939
    },
    {
      "epoch": 0.3177347887171728,
      "grad_norm": 0.38821372389793396,
      "learning_rate": 8e-05,
      "loss": 1.6937,
      "step": 2940
    },
    {
      "epoch": 0.3178428617745596,
      "grad_norm": 0.432868629693985,
      "learning_rate": 8e-05,
      "loss": 1.8571,
      "step": 2941
    },
    {
      "epoch": 0.3179509348319464,
      "grad_norm": 0.36649852991104126,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 2942
    },
    {
      "epoch": 0.31805900788933317,
      "grad_norm": 0.4657868444919586,
      "learning_rate": 8e-05,
      "loss": 1.9013,
      "step": 2943
    },
    {
      "epoch": 0.31816708094671997,
      "grad_norm": 0.40176424384117126,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 2944
    },
    {
      "epoch": 0.3182751540041068,
      "grad_norm": 0.373000830411911,
      "learning_rate": 8e-05,
      "loss": 1.807,
      "step": 2945
    },
    {
      "epoch": 0.3183832270614936,
      "grad_norm": 0.37759608030319214,
      "learning_rate": 8e-05,
      "loss": 1.8529,
      "step": 2946
    },
    {
      "epoch": 0.3184913001188804,
      "grad_norm": 0.3805919885635376,
      "learning_rate": 8e-05,
      "loss": 1.6764,
      "step": 2947
    },
    {
      "epoch": 0.31859937317626713,
      "grad_norm": 0.3967913091182709,
      "learning_rate": 8e-05,
      "loss": 1.7912,
      "step": 2948
    },
    {
      "epoch": 0.31870744623365393,
      "grad_norm": 0.39837178587913513,
      "learning_rate": 8e-05,
      "loss": 1.9178,
      "step": 2949
    },
    {
      "epoch": 0.31881551929104074,
      "grad_norm": 0.35366323590278625,
      "learning_rate": 8e-05,
      "loss": 1.6997,
      "step": 2950
    },
    {
      "epoch": 0.31892359234842754,
      "grad_norm": 0.3793390393257141,
      "learning_rate": 8e-05,
      "loss": 1.4798,
      "step": 2951
    },
    {
      "epoch": 0.31903166540581435,
      "grad_norm": 0.3694332242012024,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 2952
    },
    {
      "epoch": 0.31913973846320115,
      "grad_norm": 0.4321056604385376,
      "learning_rate": 8e-05,
      "loss": 1.8867,
      "step": 2953
    },
    {
      "epoch": 0.3192478115205879,
      "grad_norm": 0.37852564454078674,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 2954
    },
    {
      "epoch": 0.3193558845779747,
      "grad_norm": 0.3669835925102234,
      "learning_rate": 8e-05,
      "loss": 1.6707,
      "step": 2955
    },
    {
      "epoch": 0.3194639576353615,
      "grad_norm": 0.3501361012458801,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 2956
    },
    {
      "epoch": 0.3195720306927483,
      "grad_norm": 0.37453052401542664,
      "learning_rate": 8e-05,
      "loss": 1.5739,
      "step": 2957
    },
    {
      "epoch": 0.3196801037501351,
      "grad_norm": 0.40657395124435425,
      "learning_rate": 8e-05,
      "loss": 1.8,
      "step": 2958
    },
    {
      "epoch": 0.31978817680752186,
      "grad_norm": 0.40382397174835205,
      "learning_rate": 8e-05,
      "loss": 1.8852,
      "step": 2959
    },
    {
      "epoch": 0.31989624986490867,
      "grad_norm": 0.3884938657283783,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 2960
    },
    {
      "epoch": 0.32000432292229547,
      "grad_norm": 0.3831300437450409,
      "learning_rate": 8e-05,
      "loss": 1.7825,
      "step": 2961
    },
    {
      "epoch": 0.3201123959796823,
      "grad_norm": 0.4609825015068054,
      "learning_rate": 8e-05,
      "loss": 1.911,
      "step": 2962
    },
    {
      "epoch": 0.3202204690370691,
      "grad_norm": 0.3565077781677246,
      "learning_rate": 8e-05,
      "loss": 1.8518,
      "step": 2963
    },
    {
      "epoch": 0.3203285420944558,
      "grad_norm": 0.35824379324913025,
      "learning_rate": 8e-05,
      "loss": 1.5757,
      "step": 2964
    },
    {
      "epoch": 0.32043661515184263,
      "grad_norm": 0.38923028111457825,
      "learning_rate": 8e-05,
      "loss": 1.8399,
      "step": 2965
    },
    {
      "epoch": 0.32054468820922943,
      "grad_norm": 0.4200632870197296,
      "learning_rate": 8e-05,
      "loss": 1.8988,
      "step": 2966
    },
    {
      "epoch": 0.32065276126661624,
      "grad_norm": 0.42543795704841614,
      "learning_rate": 8e-05,
      "loss": 1.8122,
      "step": 2967
    },
    {
      "epoch": 0.32076083432400304,
      "grad_norm": 0.3648354411125183,
      "learning_rate": 8e-05,
      "loss": 1.7121,
      "step": 2968
    },
    {
      "epoch": 0.32086890738138985,
      "grad_norm": 0.39955025911331177,
      "learning_rate": 8e-05,
      "loss": 1.9118,
      "step": 2969
    },
    {
      "epoch": 0.3209769804387766,
      "grad_norm": 0.37788957357406616,
      "learning_rate": 8e-05,
      "loss": 1.8012,
      "step": 2970
    },
    {
      "epoch": 0.3210850534961634,
      "grad_norm": 0.41708123683929443,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 2971
    },
    {
      "epoch": 0.3211931265535502,
      "grad_norm": 0.39043566584587097,
      "learning_rate": 8e-05,
      "loss": 1.7285,
      "step": 2972
    },
    {
      "epoch": 0.321301199610937,
      "grad_norm": 0.3980652391910553,
      "learning_rate": 8e-05,
      "loss": 1.8051,
      "step": 2973
    },
    {
      "epoch": 0.3214092726683238,
      "grad_norm": 0.3829914629459381,
      "learning_rate": 8e-05,
      "loss": 1.7186,
      "step": 2974
    },
    {
      "epoch": 0.32151734572571056,
      "grad_norm": 0.3798570930957794,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 2975
    },
    {
      "epoch": 0.32162541878309736,
      "grad_norm": 0.36752450466156006,
      "learning_rate": 8e-05,
      "loss": 1.6689,
      "step": 2976
    },
    {
      "epoch": 0.32173349184048416,
      "grad_norm": 0.3925345540046692,
      "learning_rate": 8e-05,
      "loss": 1.9337,
      "step": 2977
    },
    {
      "epoch": 0.32184156489787097,
      "grad_norm": 0.379050612449646,
      "learning_rate": 8e-05,
      "loss": 1.7869,
      "step": 2978
    },
    {
      "epoch": 0.3219496379552578,
      "grad_norm": 0.41210389137268066,
      "learning_rate": 8e-05,
      "loss": 1.9441,
      "step": 2979
    },
    {
      "epoch": 0.3220577110126445,
      "grad_norm": 0.37567928433418274,
      "learning_rate": 8e-05,
      "loss": 1.7532,
      "step": 2980
    },
    {
      "epoch": 0.3221657840700313,
      "grad_norm": 0.38562798500061035,
      "learning_rate": 8e-05,
      "loss": 1.7766,
      "step": 2981
    },
    {
      "epoch": 0.32227385712741813,
      "grad_norm": 0.36712729930877686,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 2982
    },
    {
      "epoch": 0.32238193018480493,
      "grad_norm": 0.3820264935493469,
      "learning_rate": 8e-05,
      "loss": 1.8901,
      "step": 2983
    },
    {
      "epoch": 0.32249000324219174,
      "grad_norm": 0.4436863660812378,
      "learning_rate": 8e-05,
      "loss": 1.9087,
      "step": 2984
    },
    {
      "epoch": 0.32259807629957854,
      "grad_norm": 0.36913198232650757,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 2985
    },
    {
      "epoch": 0.3227061493569653,
      "grad_norm": 0.40391990542411804,
      "learning_rate": 8e-05,
      "loss": 1.7386,
      "step": 2986
    },
    {
      "epoch": 0.3228142224143521,
      "grad_norm": 0.40781503915786743,
      "learning_rate": 8e-05,
      "loss": 1.7348,
      "step": 2987
    },
    {
      "epoch": 0.3229222954717389,
      "grad_norm": 0.39497804641723633,
      "learning_rate": 8e-05,
      "loss": 1.7755,
      "step": 2988
    },
    {
      "epoch": 0.3230303685291257,
      "grad_norm": 0.3795987069606781,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 2989
    },
    {
      "epoch": 0.3231384415865125,
      "grad_norm": 0.38672304153442383,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 2990
    },
    {
      "epoch": 0.32324651464389925,
      "grad_norm": 0.36739087104797363,
      "learning_rate": 8e-05,
      "loss": 1.7472,
      "step": 2991
    },
    {
      "epoch": 0.32335458770128606,
      "grad_norm": 0.3887999355792999,
      "learning_rate": 8e-05,
      "loss": 1.662,
      "step": 2992
    },
    {
      "epoch": 0.32346266075867286,
      "grad_norm": 0.36788210272789,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 2993
    },
    {
      "epoch": 0.32357073381605966,
      "grad_norm": 0.37225428223609924,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 2994
    },
    {
      "epoch": 0.32367880687344647,
      "grad_norm": 0.39085617661476135,
      "learning_rate": 8e-05,
      "loss": 1.7446,
      "step": 2995
    },
    {
      "epoch": 0.3237868799308332,
      "grad_norm": 0.3666759431362152,
      "learning_rate": 8e-05,
      "loss": 1.641,
      "step": 2996
    },
    {
      "epoch": 0.32389495298822,
      "grad_norm": 0.35601726174354553,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 2997
    },
    {
      "epoch": 0.3240030260456068,
      "grad_norm": 0.3973410427570343,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 2998
    },
    {
      "epoch": 0.3241110991029936,
      "grad_norm": 0.3608280420303345,
      "learning_rate": 8e-05,
      "loss": 1.7558,
      "step": 2999
    },
    {
      "epoch": 0.32421917216038043,
      "grad_norm": 0.3686004877090454,
      "learning_rate": 8e-05,
      "loss": 1.6906,
      "step": 3000
    },
    {
      "epoch": 0.32432724521776723,
      "grad_norm": 0.3809454143047333,
      "learning_rate": 8e-05,
      "loss": 1.7286,
      "step": 3001
    },
    {
      "epoch": 0.324435318275154,
      "grad_norm": 0.4698570966720581,
      "learning_rate": 8e-05,
      "loss": 1.9005,
      "step": 3002
    },
    {
      "epoch": 0.3245433913325408,
      "grad_norm": 0.38386544585227966,
      "learning_rate": 8e-05,
      "loss": 1.5026,
      "step": 3003
    },
    {
      "epoch": 0.3246514643899276,
      "grad_norm": 0.400448739528656,
      "learning_rate": 8e-05,
      "loss": 1.8962,
      "step": 3004
    },
    {
      "epoch": 0.3247595374473144,
      "grad_norm": 0.38756120204925537,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 3005
    },
    {
      "epoch": 0.3248676105047012,
      "grad_norm": 0.38079118728637695,
      "learning_rate": 8e-05,
      "loss": 1.8685,
      "step": 3006
    },
    {
      "epoch": 0.32497568356208795,
      "grad_norm": 0.38144582509994507,
      "learning_rate": 8e-05,
      "loss": 1.4927,
      "step": 3007
    },
    {
      "epoch": 0.32508375661947475,
      "grad_norm": 0.46014514565467834,
      "learning_rate": 8e-05,
      "loss": 1.8318,
      "step": 3008
    },
    {
      "epoch": 0.32519182967686155,
      "grad_norm": 0.3712429702281952,
      "learning_rate": 8e-05,
      "loss": 1.6954,
      "step": 3009
    },
    {
      "epoch": 0.32529990273424836,
      "grad_norm": 0.3838809132575989,
      "learning_rate": 8e-05,
      "loss": 1.6947,
      "step": 3010
    },
    {
      "epoch": 0.32540797579163516,
      "grad_norm": 0.3676713705062866,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 3011
    },
    {
      "epoch": 0.32551604884902197,
      "grad_norm": 0.4285609722137451,
      "learning_rate": 8e-05,
      "loss": 1.755,
      "step": 3012
    },
    {
      "epoch": 0.3256241219064087,
      "grad_norm": 0.43298330903053284,
      "learning_rate": 8e-05,
      "loss": 1.7309,
      "step": 3013
    },
    {
      "epoch": 0.3257321949637955,
      "grad_norm": 0.3864617943763733,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 3014
    },
    {
      "epoch": 0.3258402680211823,
      "grad_norm": 0.3503204882144928,
      "learning_rate": 8e-05,
      "loss": 1.5717,
      "step": 3015
    },
    {
      "epoch": 0.3259483410785691,
      "grad_norm": 0.4001852571964264,
      "learning_rate": 8e-05,
      "loss": 1.7415,
      "step": 3016
    },
    {
      "epoch": 0.32605641413595593,
      "grad_norm": 0.41210275888442993,
      "learning_rate": 8e-05,
      "loss": 1.7325,
      "step": 3017
    },
    {
      "epoch": 0.3261644871933427,
      "grad_norm": 0.3962157964706421,
      "learning_rate": 8e-05,
      "loss": 1.863,
      "step": 3018
    },
    {
      "epoch": 0.3262725602507295,
      "grad_norm": 0.35352960228919983,
      "learning_rate": 8e-05,
      "loss": 1.552,
      "step": 3019
    },
    {
      "epoch": 0.3263806333081163,
      "grad_norm": 0.3970717787742615,
      "learning_rate": 8e-05,
      "loss": 1.6324,
      "step": 3020
    },
    {
      "epoch": 0.3264887063655031,
      "grad_norm": 0.415105402469635,
      "learning_rate": 8e-05,
      "loss": 1.8293,
      "step": 3021
    },
    {
      "epoch": 0.3265967794228899,
      "grad_norm": 0.4030892848968506,
      "learning_rate": 8e-05,
      "loss": 1.8305,
      "step": 3022
    },
    {
      "epoch": 0.32670485248027664,
      "grad_norm": 0.40116193890571594,
      "learning_rate": 8e-05,
      "loss": 1.7658,
      "step": 3023
    },
    {
      "epoch": 0.32681292553766345,
      "grad_norm": 0.4057353734970093,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 3024
    },
    {
      "epoch": 0.32692099859505025,
      "grad_norm": 0.3753500282764435,
      "learning_rate": 8e-05,
      "loss": 1.6363,
      "step": 3025
    },
    {
      "epoch": 0.32702907165243705,
      "grad_norm": 0.36307477951049805,
      "learning_rate": 8e-05,
      "loss": 1.6871,
      "step": 3026
    },
    {
      "epoch": 0.32713714470982386,
      "grad_norm": 0.42205509543418884,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 3027
    },
    {
      "epoch": 0.32724521776721066,
      "grad_norm": 0.38729315996170044,
      "learning_rate": 8e-05,
      "loss": 1.8616,
      "step": 3028
    },
    {
      "epoch": 0.3273532908245974,
      "grad_norm": 0.3738463222980499,
      "learning_rate": 8e-05,
      "loss": 1.5999,
      "step": 3029
    },
    {
      "epoch": 0.3274613638819842,
      "grad_norm": 0.3576565086841583,
      "learning_rate": 8e-05,
      "loss": 1.6021,
      "step": 3030
    },
    {
      "epoch": 0.327569436939371,
      "grad_norm": 0.3857894241809845,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 3031
    },
    {
      "epoch": 0.3276775099967578,
      "grad_norm": 0.37846624851226807,
      "learning_rate": 8e-05,
      "loss": 1.6171,
      "step": 3032
    },
    {
      "epoch": 0.3277855830541446,
      "grad_norm": 0.35405266284942627,
      "learning_rate": 8e-05,
      "loss": 1.742,
      "step": 3033
    },
    {
      "epoch": 0.3278936561115314,
      "grad_norm": 0.38115787506103516,
      "learning_rate": 8e-05,
      "loss": 1.7921,
      "step": 3034
    },
    {
      "epoch": 0.3280017291689182,
      "grad_norm": 0.39969295263290405,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 3035
    },
    {
      "epoch": 0.328109802226305,
      "grad_norm": 0.4015901982784271,
      "learning_rate": 8e-05,
      "loss": 1.852,
      "step": 3036
    },
    {
      "epoch": 0.3282178752836918,
      "grad_norm": 0.3909691870212555,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 3037
    },
    {
      "epoch": 0.3283259483410786,
      "grad_norm": 0.4143049716949463,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 3038
    },
    {
      "epoch": 0.32843402139846534,
      "grad_norm": 0.4304182529449463,
      "learning_rate": 8e-05,
      "loss": 1.8353,
      "step": 3039
    },
    {
      "epoch": 0.32854209445585214,
      "grad_norm": 0.38167572021484375,
      "learning_rate": 8e-05,
      "loss": 1.6847,
      "step": 3040
    },
    {
      "epoch": 0.32865016751323894,
      "grad_norm": 0.3790845274925232,
      "learning_rate": 8e-05,
      "loss": 1.7884,
      "step": 3041
    },
    {
      "epoch": 0.32875824057062575,
      "grad_norm": 0.3853643238544464,
      "learning_rate": 8e-05,
      "loss": 1.7687,
      "step": 3042
    },
    {
      "epoch": 0.32886631362801255,
      "grad_norm": 0.3807254731655121,
      "learning_rate": 8e-05,
      "loss": 1.718,
      "step": 3043
    },
    {
      "epoch": 0.32897438668539936,
      "grad_norm": 0.36925017833709717,
      "learning_rate": 8e-05,
      "loss": 1.7273,
      "step": 3044
    },
    {
      "epoch": 0.3290824597427861,
      "grad_norm": 0.38413169980049133,
      "learning_rate": 8e-05,
      "loss": 1.7437,
      "step": 3045
    },
    {
      "epoch": 0.3291905328001729,
      "grad_norm": 0.395668625831604,
      "learning_rate": 8e-05,
      "loss": 1.8495,
      "step": 3046
    },
    {
      "epoch": 0.3292986058575597,
      "grad_norm": 0.4149981439113617,
      "learning_rate": 8e-05,
      "loss": 1.9265,
      "step": 3047
    },
    {
      "epoch": 0.3294066789149465,
      "grad_norm": 0.38419240713119507,
      "learning_rate": 8e-05,
      "loss": 1.8351,
      "step": 3048
    },
    {
      "epoch": 0.3295147519723333,
      "grad_norm": 0.37794509530067444,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 3049
    },
    {
      "epoch": 0.32962282502972007,
      "grad_norm": 0.36500272154808044,
      "learning_rate": 8e-05,
      "loss": 1.7384,
      "step": 3050
    },
    {
      "epoch": 0.32973089808710687,
      "grad_norm": 0.36269626021385193,
      "learning_rate": 8e-05,
      "loss": 1.697,
      "step": 3051
    },
    {
      "epoch": 0.3298389711444937,
      "grad_norm": 0.3769221603870392,
      "learning_rate": 8e-05,
      "loss": 1.7346,
      "step": 3052
    },
    {
      "epoch": 0.3299470442018805,
      "grad_norm": 0.39780324697494507,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 3053
    },
    {
      "epoch": 0.3300551172592673,
      "grad_norm": 0.401928186416626,
      "learning_rate": 8e-05,
      "loss": 1.9293,
      "step": 3054
    },
    {
      "epoch": 0.33016319031665403,
      "grad_norm": 0.4012240171432495,
      "learning_rate": 8e-05,
      "loss": 1.694,
      "step": 3055
    },
    {
      "epoch": 0.33027126337404084,
      "grad_norm": 0.37033188343048096,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 3056
    },
    {
      "epoch": 0.33037933643142764,
      "grad_norm": 0.43209555745124817,
      "learning_rate": 8e-05,
      "loss": 1.9124,
      "step": 3057
    },
    {
      "epoch": 0.33048740948881444,
      "grad_norm": 0.4278099238872528,
      "learning_rate": 8e-05,
      "loss": 1.9179,
      "step": 3058
    },
    {
      "epoch": 0.33059548254620125,
      "grad_norm": 0.3826291263103485,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 3059
    },
    {
      "epoch": 0.33070355560358805,
      "grad_norm": 0.376976877450943,
      "learning_rate": 8e-05,
      "loss": 1.6399,
      "step": 3060
    },
    {
      "epoch": 0.3308116286609748,
      "grad_norm": 0.37109488248825073,
      "learning_rate": 8e-05,
      "loss": 1.6381,
      "step": 3061
    },
    {
      "epoch": 0.3309197017183616,
      "grad_norm": 0.4102870225906372,
      "learning_rate": 8e-05,
      "loss": 1.8208,
      "step": 3062
    },
    {
      "epoch": 0.3310277747757484,
      "grad_norm": 0.38958385586738586,
      "learning_rate": 8e-05,
      "loss": 1.8929,
      "step": 3063
    },
    {
      "epoch": 0.3311358478331352,
      "grad_norm": 0.41474154591560364,
      "learning_rate": 8e-05,
      "loss": 1.768,
      "step": 3064
    },
    {
      "epoch": 0.331243920890522,
      "grad_norm": 0.39997225999832153,
      "learning_rate": 8e-05,
      "loss": 1.7548,
      "step": 3065
    },
    {
      "epoch": 0.33135199394790876,
      "grad_norm": 0.3521725535392761,
      "learning_rate": 8e-05,
      "loss": 1.6343,
      "step": 3066
    },
    {
      "epoch": 0.33146006700529557,
      "grad_norm": 0.3650447726249695,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 3067
    },
    {
      "epoch": 0.33156814006268237,
      "grad_norm": 0.39730072021484375,
      "learning_rate": 8e-05,
      "loss": 1.7556,
      "step": 3068
    },
    {
      "epoch": 0.3316762131200692,
      "grad_norm": 0.37019282579421997,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 3069
    },
    {
      "epoch": 0.331784286177456,
      "grad_norm": 0.4003468155860901,
      "learning_rate": 8e-05,
      "loss": 1.7487,
      "step": 3070
    },
    {
      "epoch": 0.3318923592348427,
      "grad_norm": 0.3939407169818878,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 3071
    },
    {
      "epoch": 0.33200043229222953,
      "grad_norm": 0.43840292096138,
      "learning_rate": 8e-05,
      "loss": 1.9819,
      "step": 3072
    },
    {
      "epoch": 0.33210850534961633,
      "grad_norm": 0.3808031678199768,
      "learning_rate": 8e-05,
      "loss": 1.7616,
      "step": 3073
    },
    {
      "epoch": 0.33221657840700314,
      "grad_norm": 0.3941723108291626,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 3074
    },
    {
      "epoch": 0.33232465146438994,
      "grad_norm": 0.36767804622650146,
      "learning_rate": 8e-05,
      "loss": 1.6918,
      "step": 3075
    },
    {
      "epoch": 0.33243272452177675,
      "grad_norm": 0.3735896348953247,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 3076
    },
    {
      "epoch": 0.3325407975791635,
      "grad_norm": 0.39015597105026245,
      "learning_rate": 8e-05,
      "loss": 1.7603,
      "step": 3077
    },
    {
      "epoch": 0.3326488706365503,
      "grad_norm": 0.44089511036872864,
      "learning_rate": 8e-05,
      "loss": 1.9447,
      "step": 3078
    },
    {
      "epoch": 0.3327569436939371,
      "grad_norm": 0.39315685629844666,
      "learning_rate": 8e-05,
      "loss": 1.8167,
      "step": 3079
    },
    {
      "epoch": 0.3328650167513239,
      "grad_norm": 0.36816471815109253,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 3080
    },
    {
      "epoch": 0.3329730898087107,
      "grad_norm": 0.399775892496109,
      "learning_rate": 8e-05,
      "loss": 1.9144,
      "step": 3081
    },
    {
      "epoch": 0.33308116286609746,
      "grad_norm": 0.3544377088546753,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 3082
    },
    {
      "epoch": 0.33318923592348426,
      "grad_norm": 0.3989788889884949,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 3083
    },
    {
      "epoch": 0.33329730898087107,
      "grad_norm": 0.3942987620830536,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 3084
    },
    {
      "epoch": 0.33340538203825787,
      "grad_norm": 0.35681843757629395,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 3085
    },
    {
      "epoch": 0.3335134550956447,
      "grad_norm": 0.3745489716529846,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 3086
    },
    {
      "epoch": 0.3336215281530315,
      "grad_norm": 0.4045359194278717,
      "learning_rate": 8e-05,
      "loss": 1.8474,
      "step": 3087
    },
    {
      "epoch": 0.3337296012104182,
      "grad_norm": 0.3832891583442688,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 3088
    },
    {
      "epoch": 0.33383767426780503,
      "grad_norm": 0.3704085946083069,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 3089
    },
    {
      "epoch": 0.33394574732519183,
      "grad_norm": 0.41212600469589233,
      "learning_rate": 8e-05,
      "loss": 1.5284,
      "step": 3090
    },
    {
      "epoch": 0.33405382038257864,
      "grad_norm": 0.3837355077266693,
      "learning_rate": 8e-05,
      "loss": 1.7049,
      "step": 3091
    },
    {
      "epoch": 0.33416189343996544,
      "grad_norm": 0.37648603320121765,
      "learning_rate": 8e-05,
      "loss": 1.6555,
      "step": 3092
    },
    {
      "epoch": 0.3342699664973522,
      "grad_norm": 0.44515129923820496,
      "learning_rate": 8e-05,
      "loss": 1.9182,
      "step": 3093
    },
    {
      "epoch": 0.334378039554739,
      "grad_norm": 0.3692026734352112,
      "learning_rate": 8e-05,
      "loss": 1.6226,
      "step": 3094
    },
    {
      "epoch": 0.3344861126121258,
      "grad_norm": 0.38484832644462585,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 3095
    },
    {
      "epoch": 0.3345941856695126,
      "grad_norm": 0.3784169852733612,
      "learning_rate": 8e-05,
      "loss": 1.5804,
      "step": 3096
    },
    {
      "epoch": 0.3347022587268994,
      "grad_norm": 0.4062996506690979,
      "learning_rate": 8e-05,
      "loss": 1.8089,
      "step": 3097
    },
    {
      "epoch": 0.33481033178428615,
      "grad_norm": 0.45532190799713135,
      "learning_rate": 8e-05,
      "loss": 1.995,
      "step": 3098
    },
    {
      "epoch": 0.33491840484167296,
      "grad_norm": 0.38555893301963806,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 3099
    },
    {
      "epoch": 0.33502647789905976,
      "grad_norm": 0.37665918469429016,
      "learning_rate": 8e-05,
      "loss": 1.8665,
      "step": 3100
    },
    {
      "epoch": 0.33513455095644656,
      "grad_norm": 0.3787120580673218,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 3101
    },
    {
      "epoch": 0.33524262401383337,
      "grad_norm": 0.37780246138572693,
      "learning_rate": 8e-05,
      "loss": 1.7697,
      "step": 3102
    },
    {
      "epoch": 0.33535069707122017,
      "grad_norm": 0.36602702736854553,
      "learning_rate": 8e-05,
      "loss": 1.5447,
      "step": 3103
    },
    {
      "epoch": 0.3354587701286069,
      "grad_norm": 0.3563636243343353,
      "learning_rate": 8e-05,
      "loss": 1.6261,
      "step": 3104
    },
    {
      "epoch": 0.3355668431859937,
      "grad_norm": 0.3499147891998291,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 3105
    },
    {
      "epoch": 0.33567491624338053,
      "grad_norm": 0.3652617633342743,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 3106
    },
    {
      "epoch": 0.33578298930076733,
      "grad_norm": 0.39372292160987854,
      "learning_rate": 8e-05,
      "loss": 1.6593,
      "step": 3107
    },
    {
      "epoch": 0.33589106235815414,
      "grad_norm": 0.3749449551105499,
      "learning_rate": 8e-05,
      "loss": 1.7599,
      "step": 3108
    },
    {
      "epoch": 0.3359991354155409,
      "grad_norm": 0.4025574028491974,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 3109
    },
    {
      "epoch": 0.3361072084729277,
      "grad_norm": 0.42658382654190063,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 3110
    },
    {
      "epoch": 0.3362152815303145,
      "grad_norm": 0.3574577867984772,
      "learning_rate": 8e-05,
      "loss": 1.7481,
      "step": 3111
    },
    {
      "epoch": 0.3363233545877013,
      "grad_norm": 0.41829562187194824,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 3112
    },
    {
      "epoch": 0.3364314276450881,
      "grad_norm": 0.449555367231369,
      "learning_rate": 8e-05,
      "loss": 1.901,
      "step": 3113
    },
    {
      "epoch": 0.33653950070247485,
      "grad_norm": 0.37352922558784485,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 3114
    },
    {
      "epoch": 0.33664757375986165,
      "grad_norm": 0.39764904975891113,
      "learning_rate": 8e-05,
      "loss": 1.6551,
      "step": 3115
    },
    {
      "epoch": 0.33675564681724846,
      "grad_norm": 0.3583729565143585,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 3116
    },
    {
      "epoch": 0.33686371987463526,
      "grad_norm": 0.35891640186309814,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 3117
    },
    {
      "epoch": 0.33697179293202206,
      "grad_norm": 0.39025938510894775,
      "learning_rate": 8e-05,
      "loss": 1.7818,
      "step": 3118
    },
    {
      "epoch": 0.33707986598940887,
      "grad_norm": 0.3901554048061371,
      "learning_rate": 8e-05,
      "loss": 1.7348,
      "step": 3119
    },
    {
      "epoch": 0.3371879390467956,
      "grad_norm": 0.3619101643562317,
      "learning_rate": 8e-05,
      "loss": 1.6234,
      "step": 3120
    },
    {
      "epoch": 0.3372960121041824,
      "grad_norm": 0.373629629611969,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 3121
    },
    {
      "epoch": 0.3374040851615692,
      "grad_norm": 0.386385977268219,
      "learning_rate": 8e-05,
      "loss": 1.5161,
      "step": 3122
    },
    {
      "epoch": 0.337512158218956,
      "grad_norm": 0.39593416452407837,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 3123
    },
    {
      "epoch": 0.33762023127634283,
      "grad_norm": 0.44279465079307556,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 3124
    },
    {
      "epoch": 0.3377283043337296,
      "grad_norm": 0.38848721981048584,
      "learning_rate": 8e-05,
      "loss": 1.8649,
      "step": 3125
    },
    {
      "epoch": 0.3378363773911164,
      "grad_norm": 0.3903926610946655,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 3126
    },
    {
      "epoch": 0.3379444504485032,
      "grad_norm": 0.3858160674571991,
      "learning_rate": 8e-05,
      "loss": 1.776,
      "step": 3127
    },
    {
      "epoch": 0.33805252350589,
      "grad_norm": 0.39503213763237,
      "learning_rate": 8e-05,
      "loss": 1.6768,
      "step": 3128
    },
    {
      "epoch": 0.3381605965632768,
      "grad_norm": 0.3833896219730377,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 3129
    },
    {
      "epoch": 0.33826866962066354,
      "grad_norm": 0.3958110213279724,
      "learning_rate": 8e-05,
      "loss": 1.747,
      "step": 3130
    },
    {
      "epoch": 0.33837674267805035,
      "grad_norm": 0.38637807965278625,
      "learning_rate": 8e-05,
      "loss": 1.713,
      "step": 3131
    },
    {
      "epoch": 0.33848481573543715,
      "grad_norm": 0.38812255859375,
      "learning_rate": 8e-05,
      "loss": 1.6693,
      "step": 3132
    },
    {
      "epoch": 0.33859288879282395,
      "grad_norm": 0.38828715682029724,
      "learning_rate": 8e-05,
      "loss": 1.7824,
      "step": 3133
    },
    {
      "epoch": 0.33870096185021076,
      "grad_norm": 0.3985130488872528,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 3134
    },
    {
      "epoch": 0.33880903490759756,
      "grad_norm": 0.3935237228870392,
      "learning_rate": 8e-05,
      "loss": 1.7629,
      "step": 3135
    },
    {
      "epoch": 0.3389171079649843,
      "grad_norm": 0.38870543241500854,
      "learning_rate": 8e-05,
      "loss": 1.7865,
      "step": 3136
    },
    {
      "epoch": 0.3390251810223711,
      "grad_norm": 0.4428810179233551,
      "learning_rate": 8e-05,
      "loss": 1.9752,
      "step": 3137
    },
    {
      "epoch": 0.3391332540797579,
      "grad_norm": 0.3855208158493042,
      "learning_rate": 8e-05,
      "loss": 1.7428,
      "step": 3138
    },
    {
      "epoch": 0.3392413271371447,
      "grad_norm": 0.3737037479877472,
      "learning_rate": 8e-05,
      "loss": 1.8177,
      "step": 3139
    },
    {
      "epoch": 0.3393494001945315,
      "grad_norm": 0.37476736307144165,
      "learning_rate": 8e-05,
      "loss": 1.8322,
      "step": 3140
    },
    {
      "epoch": 0.3394574732519183,
      "grad_norm": 0.4070238471031189,
      "learning_rate": 8e-05,
      "loss": 1.8843,
      "step": 3141
    },
    {
      "epoch": 0.3395655463093051,
      "grad_norm": 0.37859782576560974,
      "learning_rate": 8e-05,
      "loss": 1.7094,
      "step": 3142
    },
    {
      "epoch": 0.3396736193666919,
      "grad_norm": 0.3812742233276367,
      "learning_rate": 8e-05,
      "loss": 1.7236,
      "step": 3143
    },
    {
      "epoch": 0.3397816924240787,
      "grad_norm": 0.43662509322166443,
      "learning_rate": 8e-05,
      "loss": 1.7499,
      "step": 3144
    },
    {
      "epoch": 0.3398897654814655,
      "grad_norm": 0.38658303022384644,
      "learning_rate": 8e-05,
      "loss": 1.7173,
      "step": 3145
    },
    {
      "epoch": 0.33999783853885224,
      "grad_norm": 0.42068058252334595,
      "learning_rate": 8e-05,
      "loss": 1.8062,
      "step": 3146
    },
    {
      "epoch": 0.34010591159623904,
      "grad_norm": 0.3968547284603119,
      "learning_rate": 8e-05,
      "loss": 1.8153,
      "step": 3147
    },
    {
      "epoch": 0.34021398465362584,
      "grad_norm": 0.3969630002975464,
      "learning_rate": 8e-05,
      "loss": 1.8883,
      "step": 3148
    },
    {
      "epoch": 0.34032205771101265,
      "grad_norm": 0.3536335825920105,
      "learning_rate": 8e-05,
      "loss": 1.6495,
      "step": 3149
    },
    {
      "epoch": 0.34043013076839945,
      "grad_norm": 0.38026586174964905,
      "learning_rate": 8e-05,
      "loss": 1.7225,
      "step": 3150
    },
    {
      "epoch": 0.34053820382578626,
      "grad_norm": 0.3492166996002197,
      "learning_rate": 8e-05,
      "loss": 1.6617,
      "step": 3151
    },
    {
      "epoch": 0.340646276883173,
      "grad_norm": 0.38141706585884094,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 3152
    },
    {
      "epoch": 0.3407543499405598,
      "grad_norm": 0.3577704429626465,
      "learning_rate": 8e-05,
      "loss": 1.625,
      "step": 3153
    },
    {
      "epoch": 0.3408624229979466,
      "grad_norm": 0.4100503623485565,
      "learning_rate": 8e-05,
      "loss": 1.7185,
      "step": 3154
    },
    {
      "epoch": 0.3409704960553334,
      "grad_norm": 0.40080422163009644,
      "learning_rate": 8e-05,
      "loss": 1.6716,
      "step": 3155
    },
    {
      "epoch": 0.3410785691127202,
      "grad_norm": 0.38559532165527344,
      "learning_rate": 8e-05,
      "loss": 1.7915,
      "step": 3156
    },
    {
      "epoch": 0.34118664217010697,
      "grad_norm": 0.39551371335983276,
      "learning_rate": 8e-05,
      "loss": 1.6686,
      "step": 3157
    },
    {
      "epoch": 0.3412947152274938,
      "grad_norm": 0.38980206847190857,
      "learning_rate": 8e-05,
      "loss": 1.7351,
      "step": 3158
    },
    {
      "epoch": 0.3414027882848806,
      "grad_norm": 0.38582324981689453,
      "learning_rate": 8e-05,
      "loss": 1.6744,
      "step": 3159
    },
    {
      "epoch": 0.3415108613422674,
      "grad_norm": 0.38020479679107666,
      "learning_rate": 8e-05,
      "loss": 1.7367,
      "step": 3160
    },
    {
      "epoch": 0.3416189343996542,
      "grad_norm": 0.3968760371208191,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 3161
    },
    {
      "epoch": 0.34172700745704093,
      "grad_norm": 0.38403671979904175,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 3162
    },
    {
      "epoch": 0.34183508051442774,
      "grad_norm": 0.36699190735816956,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 3163
    },
    {
      "epoch": 0.34194315357181454,
      "grad_norm": 0.38546979427337646,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 3164
    },
    {
      "epoch": 0.34205122662920134,
      "grad_norm": 0.3744828402996063,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 3165
    },
    {
      "epoch": 0.34215929968658815,
      "grad_norm": 0.38387835025787354,
      "learning_rate": 8e-05,
      "loss": 1.8636,
      "step": 3166
    },
    {
      "epoch": 0.34226737274397495,
      "grad_norm": 0.42923834919929504,
      "learning_rate": 8e-05,
      "loss": 1.8915,
      "step": 3167
    },
    {
      "epoch": 0.3423754458013617,
      "grad_norm": 0.4237186014652252,
      "learning_rate": 8e-05,
      "loss": 1.9423,
      "step": 3168
    },
    {
      "epoch": 0.3424835188587485,
      "grad_norm": 0.39949706196784973,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 3169
    },
    {
      "epoch": 0.3425915919161353,
      "grad_norm": 0.4231875240802765,
      "learning_rate": 8e-05,
      "loss": 1.8786,
      "step": 3170
    },
    {
      "epoch": 0.3426996649735221,
      "grad_norm": 0.4104372262954712,
      "learning_rate": 8e-05,
      "loss": 1.6834,
      "step": 3171
    },
    {
      "epoch": 0.3428077380309089,
      "grad_norm": 0.3589286506175995,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 3172
    },
    {
      "epoch": 0.34291581108829566,
      "grad_norm": 0.374222069978714,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 3173
    },
    {
      "epoch": 0.34302388414568247,
      "grad_norm": 0.39693236351013184,
      "learning_rate": 8e-05,
      "loss": 1.8215,
      "step": 3174
    },
    {
      "epoch": 0.34313195720306927,
      "grad_norm": 0.361100971698761,
      "learning_rate": 8e-05,
      "loss": 1.5234,
      "step": 3175
    },
    {
      "epoch": 0.3432400302604561,
      "grad_norm": 0.47115105390548706,
      "learning_rate": 8e-05,
      "loss": 1.9989,
      "step": 3176
    },
    {
      "epoch": 0.3433481033178429,
      "grad_norm": 0.38971850275993347,
      "learning_rate": 8e-05,
      "loss": 1.6616,
      "step": 3177
    },
    {
      "epoch": 0.3434561763752297,
      "grad_norm": 0.422820508480072,
      "learning_rate": 8e-05,
      "loss": 1.8927,
      "step": 3178
    },
    {
      "epoch": 0.34356424943261643,
      "grad_norm": 0.3851724863052368,
      "learning_rate": 8e-05,
      "loss": 1.6103,
      "step": 3179
    },
    {
      "epoch": 0.34367232249000323,
      "grad_norm": 0.43618062138557434,
      "learning_rate": 8e-05,
      "loss": 1.9834,
      "step": 3180
    },
    {
      "epoch": 0.34378039554739004,
      "grad_norm": 0.39255064725875854,
      "learning_rate": 8e-05,
      "loss": 1.8049,
      "step": 3181
    },
    {
      "epoch": 0.34388846860477684,
      "grad_norm": 0.41304919123649597,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 3182
    },
    {
      "epoch": 0.34399654166216365,
      "grad_norm": 0.42593175172805786,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 3183
    },
    {
      "epoch": 0.3441046147195504,
      "grad_norm": 0.39380452036857605,
      "learning_rate": 8e-05,
      "loss": 1.8173,
      "step": 3184
    },
    {
      "epoch": 0.3442126877769372,
      "grad_norm": 0.40321898460388184,
      "learning_rate": 8e-05,
      "loss": 1.7012,
      "step": 3185
    },
    {
      "epoch": 0.344320760834324,
      "grad_norm": 0.35680556297302246,
      "learning_rate": 8e-05,
      "loss": 1.7496,
      "step": 3186
    },
    {
      "epoch": 0.3444288338917108,
      "grad_norm": 0.3606150448322296,
      "learning_rate": 8e-05,
      "loss": 1.5744,
      "step": 3187
    },
    {
      "epoch": 0.3445369069490976,
      "grad_norm": 0.39066699147224426,
      "learning_rate": 8e-05,
      "loss": 1.8395,
      "step": 3188
    },
    {
      "epoch": 0.34464498000648436,
      "grad_norm": 0.37386608123779297,
      "learning_rate": 8e-05,
      "loss": 1.7669,
      "step": 3189
    },
    {
      "epoch": 0.34475305306387116,
      "grad_norm": 0.42777225375175476,
      "learning_rate": 8e-05,
      "loss": 1.7297,
      "step": 3190
    },
    {
      "epoch": 0.34486112612125797,
      "grad_norm": 0.42951536178588867,
      "learning_rate": 8e-05,
      "loss": 1.9396,
      "step": 3191
    },
    {
      "epoch": 0.34496919917864477,
      "grad_norm": 0.3806735575199127,
      "learning_rate": 8e-05,
      "loss": 1.6806,
      "step": 3192
    },
    {
      "epoch": 0.3450772722360316,
      "grad_norm": 0.37916627526283264,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 3193
    },
    {
      "epoch": 0.3451853452934184,
      "grad_norm": 0.3841780126094818,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 3194
    },
    {
      "epoch": 0.3452934183508051,
      "grad_norm": 0.3973079025745392,
      "learning_rate": 8e-05,
      "loss": 1.6841,
      "step": 3195
    },
    {
      "epoch": 0.34540149140819193,
      "grad_norm": 0.36865347623825073,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 3196
    },
    {
      "epoch": 0.34550956446557873,
      "grad_norm": 0.3765410780906677,
      "learning_rate": 8e-05,
      "loss": 1.808,
      "step": 3197
    },
    {
      "epoch": 0.34561763752296554,
      "grad_norm": 0.364969402551651,
      "learning_rate": 8e-05,
      "loss": 1.6414,
      "step": 3198
    },
    {
      "epoch": 0.34572571058035234,
      "grad_norm": 0.4416789710521698,
      "learning_rate": 8e-05,
      "loss": 1.8043,
      "step": 3199
    },
    {
      "epoch": 0.3458337836377391,
      "grad_norm": 0.3804566562175751,
      "learning_rate": 8e-05,
      "loss": 1.7166,
      "step": 3200
    },
    {
      "epoch": 0.3459418566951259,
      "grad_norm": 0.3978241980075836,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 3201
    },
    {
      "epoch": 0.3460499297525127,
      "grad_norm": 0.4234887361526489,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 3202
    },
    {
      "epoch": 0.3461580028098995,
      "grad_norm": 0.3662523031234741,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 3203
    },
    {
      "epoch": 0.3462660758672863,
      "grad_norm": 0.3779642879962921,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 3204
    },
    {
      "epoch": 0.34637414892467305,
      "grad_norm": 0.38254043459892273,
      "learning_rate": 8e-05,
      "loss": 1.7147,
      "step": 3205
    },
    {
      "epoch": 0.34648222198205986,
      "grad_norm": 0.36932653188705444,
      "learning_rate": 8e-05,
      "loss": 1.7334,
      "step": 3206
    },
    {
      "epoch": 0.34659029503944666,
      "grad_norm": 0.38608768582344055,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 3207
    },
    {
      "epoch": 0.34669836809683346,
      "grad_norm": 0.39628902077674866,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 3208
    },
    {
      "epoch": 0.34680644115422027,
      "grad_norm": 0.4145163893699646,
      "learning_rate": 8e-05,
      "loss": 1.8381,
      "step": 3209
    },
    {
      "epoch": 0.3469145142116071,
      "grad_norm": 0.3667445480823517,
      "learning_rate": 8e-05,
      "loss": 1.7443,
      "step": 3210
    },
    {
      "epoch": 0.3470225872689938,
      "grad_norm": 0.3577861785888672,
      "learning_rate": 8e-05,
      "loss": 1.6976,
      "step": 3211
    },
    {
      "epoch": 0.3471306603263806,
      "grad_norm": 0.417246013879776,
      "learning_rate": 8e-05,
      "loss": 1.6942,
      "step": 3212
    },
    {
      "epoch": 0.34723873338376743,
      "grad_norm": 0.38905069231987,
      "learning_rate": 8e-05,
      "loss": 1.754,
      "step": 3213
    },
    {
      "epoch": 0.34734680644115423,
      "grad_norm": 0.4353747069835663,
      "learning_rate": 8e-05,
      "loss": 1.8437,
      "step": 3214
    },
    {
      "epoch": 0.34745487949854104,
      "grad_norm": 0.404378205537796,
      "learning_rate": 8e-05,
      "loss": 1.8159,
      "step": 3215
    },
    {
      "epoch": 0.3475629525559278,
      "grad_norm": 0.389313280582428,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 3216
    },
    {
      "epoch": 0.3476710256133146,
      "grad_norm": 0.36213281750679016,
      "learning_rate": 8e-05,
      "loss": 1.604,
      "step": 3217
    },
    {
      "epoch": 0.3477790986707014,
      "grad_norm": 0.4030488133430481,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 3218
    },
    {
      "epoch": 0.3478871717280882,
      "grad_norm": 0.3838377594947815,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 3219
    },
    {
      "epoch": 0.347995244785475,
      "grad_norm": 0.373348206281662,
      "learning_rate": 8e-05,
      "loss": 1.6191,
      "step": 3220
    },
    {
      "epoch": 0.34810331784286175,
      "grad_norm": 0.36474865674972534,
      "learning_rate": 8e-05,
      "loss": 1.5536,
      "step": 3221
    },
    {
      "epoch": 0.34821139090024855,
      "grad_norm": 0.4290372133255005,
      "learning_rate": 8e-05,
      "loss": 1.8732,
      "step": 3222
    },
    {
      "epoch": 0.34831946395763536,
      "grad_norm": 0.4214039444923401,
      "learning_rate": 8e-05,
      "loss": 1.7579,
      "step": 3223
    },
    {
      "epoch": 0.34842753701502216,
      "grad_norm": 0.406820684671402,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 3224
    },
    {
      "epoch": 0.34853561007240896,
      "grad_norm": 0.37152329087257385,
      "learning_rate": 8e-05,
      "loss": 1.7875,
      "step": 3225
    },
    {
      "epoch": 0.34864368312979577,
      "grad_norm": 0.38109225034713745,
      "learning_rate": 8e-05,
      "loss": 1.5189,
      "step": 3226
    },
    {
      "epoch": 0.3487517561871825,
      "grad_norm": 0.3486393690109253,
      "learning_rate": 8e-05,
      "loss": 1.4497,
      "step": 3227
    },
    {
      "epoch": 0.3488598292445693,
      "grad_norm": 0.4016020596027374,
      "learning_rate": 8e-05,
      "loss": 1.4721,
      "step": 3228
    },
    {
      "epoch": 0.3489679023019561,
      "grad_norm": 0.367669016122818,
      "learning_rate": 8e-05,
      "loss": 1.5534,
      "step": 3229
    },
    {
      "epoch": 0.3490759753593429,
      "grad_norm": 0.42080068588256836,
      "learning_rate": 8e-05,
      "loss": 1.8932,
      "step": 3230
    },
    {
      "epoch": 0.34918404841672973,
      "grad_norm": 0.39687415957450867,
      "learning_rate": 8e-05,
      "loss": 1.8092,
      "step": 3231
    },
    {
      "epoch": 0.3492921214741165,
      "grad_norm": 0.36801913380622864,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 3232
    },
    {
      "epoch": 0.3494001945315033,
      "grad_norm": 0.3881400227546692,
      "learning_rate": 8e-05,
      "loss": 1.7115,
      "step": 3233
    },
    {
      "epoch": 0.3495082675888901,
      "grad_norm": 0.3757289946079254,
      "learning_rate": 8e-05,
      "loss": 1.5718,
      "step": 3234
    },
    {
      "epoch": 0.3496163406462769,
      "grad_norm": 0.4180893003940582,
      "learning_rate": 8e-05,
      "loss": 1.8391,
      "step": 3235
    },
    {
      "epoch": 0.3497244137036637,
      "grad_norm": 0.4260927438735962,
      "learning_rate": 8e-05,
      "loss": 1.908,
      "step": 3236
    },
    {
      "epoch": 0.34983248676105044,
      "grad_norm": 0.3835902512073517,
      "learning_rate": 8e-05,
      "loss": 1.5303,
      "step": 3237
    },
    {
      "epoch": 0.34994055981843725,
      "grad_norm": 0.4430564343929291,
      "learning_rate": 8e-05,
      "loss": 1.4502,
      "step": 3238
    },
    {
      "epoch": 0.35004863287582405,
      "grad_norm": 0.42253732681274414,
      "learning_rate": 8e-05,
      "loss": 1.8428,
      "step": 3239
    },
    {
      "epoch": 0.35015670593321085,
      "grad_norm": 0.3629271686077118,
      "learning_rate": 8e-05,
      "loss": 1.8164,
      "step": 3240
    },
    {
      "epoch": 0.35026477899059766,
      "grad_norm": 0.3824191391468048,
      "learning_rate": 8e-05,
      "loss": 1.7892,
      "step": 3241
    },
    {
      "epoch": 0.35037285204798446,
      "grad_norm": 0.3858236074447632,
      "learning_rate": 8e-05,
      "loss": 1.7041,
      "step": 3242
    },
    {
      "epoch": 0.3504809251053712,
      "grad_norm": 0.3681716322898865,
      "learning_rate": 8e-05,
      "loss": 1.542,
      "step": 3243
    },
    {
      "epoch": 0.350588998162758,
      "grad_norm": 0.4111172556877136,
      "learning_rate": 8e-05,
      "loss": 1.6851,
      "step": 3244
    },
    {
      "epoch": 0.3506970712201448,
      "grad_norm": 0.3882628083229065,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 3245
    },
    {
      "epoch": 0.3508051442775316,
      "grad_norm": 0.4404919445514679,
      "learning_rate": 8e-05,
      "loss": 1.9407,
      "step": 3246
    },
    {
      "epoch": 0.3509132173349184,
      "grad_norm": 0.38660356402397156,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 3247
    },
    {
      "epoch": 0.3510212903923052,
      "grad_norm": 0.40403205156326294,
      "learning_rate": 8e-05,
      "loss": 1.729,
      "step": 3248
    },
    {
      "epoch": 0.351129363449692,
      "grad_norm": 0.3867197036743164,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 3249
    },
    {
      "epoch": 0.3512374365070788,
      "grad_norm": 0.3770262598991394,
      "learning_rate": 8e-05,
      "loss": 1.7763,
      "step": 3250
    },
    {
      "epoch": 0.3513455095644656,
      "grad_norm": 0.37439635396003723,
      "learning_rate": 8e-05,
      "loss": 1.7065,
      "step": 3251
    },
    {
      "epoch": 0.3514535826218524,
      "grad_norm": 0.37813103199005127,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 3252
    },
    {
      "epoch": 0.3515616556792392,
      "grad_norm": 0.39981237053871155,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 3253
    },
    {
      "epoch": 0.35166972873662594,
      "grad_norm": 0.44958311319351196,
      "learning_rate": 8e-05,
      "loss": 1.7941,
      "step": 3254
    },
    {
      "epoch": 0.35177780179401275,
      "grad_norm": 0.3765053153038025,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 3255
    },
    {
      "epoch": 0.35188587485139955,
      "grad_norm": 0.38832759857177734,
      "learning_rate": 8e-05,
      "loss": 1.8304,
      "step": 3256
    },
    {
      "epoch": 0.35199394790878635,
      "grad_norm": 0.3968130052089691,
      "learning_rate": 8e-05,
      "loss": 1.766,
      "step": 3257
    },
    {
      "epoch": 0.35210202096617316,
      "grad_norm": 0.4032117426395416,
      "learning_rate": 8e-05,
      "loss": 1.7212,
      "step": 3258
    },
    {
      "epoch": 0.3522100940235599,
      "grad_norm": 0.3913801610469818,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 3259
    },
    {
      "epoch": 0.3523181670809467,
      "grad_norm": 0.3861391842365265,
      "learning_rate": 8e-05,
      "loss": 1.6915,
      "step": 3260
    },
    {
      "epoch": 0.3524262401383335,
      "grad_norm": 0.427837997674942,
      "learning_rate": 8e-05,
      "loss": 1.7505,
      "step": 3261
    },
    {
      "epoch": 0.3525343131957203,
      "grad_norm": 0.37296706438064575,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 3262
    },
    {
      "epoch": 0.3526423862531071,
      "grad_norm": 0.395699679851532,
      "learning_rate": 8e-05,
      "loss": 1.8406,
      "step": 3263
    },
    {
      "epoch": 0.35275045931049387,
      "grad_norm": 0.37352868914604187,
      "learning_rate": 8e-05,
      "loss": 1.5707,
      "step": 3264
    },
    {
      "epoch": 0.3528585323678807,
      "grad_norm": 0.38388049602508545,
      "learning_rate": 8e-05,
      "loss": 1.582,
      "step": 3265
    },
    {
      "epoch": 0.3529666054252675,
      "grad_norm": 0.4121200144290924,
      "learning_rate": 8e-05,
      "loss": 1.9766,
      "step": 3266
    },
    {
      "epoch": 0.3530746784826543,
      "grad_norm": 0.37032556533813477,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 3267
    },
    {
      "epoch": 0.3531827515400411,
      "grad_norm": 0.4014259874820709,
      "learning_rate": 8e-05,
      "loss": 1.8153,
      "step": 3268
    },
    {
      "epoch": 0.3532908245974279,
      "grad_norm": 0.3905584514141083,
      "learning_rate": 8e-05,
      "loss": 1.5817,
      "step": 3269
    },
    {
      "epoch": 0.35339889765481464,
      "grad_norm": 0.3741360902786255,
      "learning_rate": 8e-05,
      "loss": 1.7303,
      "step": 3270
    },
    {
      "epoch": 0.35350697071220144,
      "grad_norm": 0.3886110186576843,
      "learning_rate": 8e-05,
      "loss": 1.5914,
      "step": 3271
    },
    {
      "epoch": 0.35361504376958824,
      "grad_norm": 0.3873473107814789,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 3272
    },
    {
      "epoch": 0.35372311682697505,
      "grad_norm": 0.3783112168312073,
      "learning_rate": 8e-05,
      "loss": 1.5259,
      "step": 3273
    },
    {
      "epoch": 0.35383118988436185,
      "grad_norm": 0.3893345594406128,
      "learning_rate": 8e-05,
      "loss": 1.7142,
      "step": 3274
    },
    {
      "epoch": 0.3539392629417486,
      "grad_norm": 0.3947809934616089,
      "learning_rate": 8e-05,
      "loss": 1.8756,
      "step": 3275
    },
    {
      "epoch": 0.3540473359991354,
      "grad_norm": 0.37941446900367737,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 3276
    },
    {
      "epoch": 0.3541554090565222,
      "grad_norm": 0.3749096691608429,
      "learning_rate": 8e-05,
      "loss": 1.7362,
      "step": 3277
    },
    {
      "epoch": 0.354263482113909,
      "grad_norm": 0.4113011360168457,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 3278
    },
    {
      "epoch": 0.3543715551712958,
      "grad_norm": 0.45836612582206726,
      "learning_rate": 8e-05,
      "loss": 1.9683,
      "step": 3279
    },
    {
      "epoch": 0.35447962822868256,
      "grad_norm": 0.3996044397354126,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 3280
    },
    {
      "epoch": 0.35458770128606937,
      "grad_norm": 0.36189040541648865,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 3281
    },
    {
      "epoch": 0.35469577434345617,
      "grad_norm": 0.3680158853530884,
      "learning_rate": 8e-05,
      "loss": 1.6799,
      "step": 3282
    },
    {
      "epoch": 0.354803847400843,
      "grad_norm": 0.38524171710014343,
      "learning_rate": 8e-05,
      "loss": 1.8471,
      "step": 3283
    },
    {
      "epoch": 0.3549119204582298,
      "grad_norm": 0.39464250206947327,
      "learning_rate": 8e-05,
      "loss": 1.6951,
      "step": 3284
    },
    {
      "epoch": 0.3550199935156166,
      "grad_norm": 0.4188738465309143,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 3285
    },
    {
      "epoch": 0.35512806657300333,
      "grad_norm": 0.3865261673927307,
      "learning_rate": 8e-05,
      "loss": 1.745,
      "step": 3286
    },
    {
      "epoch": 0.35523613963039014,
      "grad_norm": 0.3753278851509094,
      "learning_rate": 8e-05,
      "loss": 1.7997,
      "step": 3287
    },
    {
      "epoch": 0.35534421268777694,
      "grad_norm": 0.3853910267353058,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 3288
    },
    {
      "epoch": 0.35545228574516374,
      "grad_norm": 0.36639976501464844,
      "learning_rate": 8e-05,
      "loss": 1.8848,
      "step": 3289
    },
    {
      "epoch": 0.35556035880255055,
      "grad_norm": 0.3727981448173523,
      "learning_rate": 8e-05,
      "loss": 1.8182,
      "step": 3290
    },
    {
      "epoch": 0.3556684318599373,
      "grad_norm": 0.35549330711364746,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 3291
    },
    {
      "epoch": 0.3557765049173241,
      "grad_norm": 0.35986465215682983,
      "learning_rate": 8e-05,
      "loss": 1.803,
      "step": 3292
    },
    {
      "epoch": 0.3558845779747109,
      "grad_norm": 0.37182649970054626,
      "learning_rate": 8e-05,
      "loss": 1.6322,
      "step": 3293
    },
    {
      "epoch": 0.3559926510320977,
      "grad_norm": 0.3608284592628479,
      "learning_rate": 8e-05,
      "loss": 1.5897,
      "step": 3294
    },
    {
      "epoch": 0.3561007240894845,
      "grad_norm": 0.3742779791355133,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 3295
    },
    {
      "epoch": 0.35620879714687126,
      "grad_norm": 0.4225568175315857,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 3296
    },
    {
      "epoch": 0.35631687020425806,
      "grad_norm": 0.3930049240589142,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 3297
    },
    {
      "epoch": 0.35642494326164487,
      "grad_norm": 0.37575840950012207,
      "learning_rate": 8e-05,
      "loss": 1.562,
      "step": 3298
    },
    {
      "epoch": 0.35653301631903167,
      "grad_norm": 0.36371055245399475,
      "learning_rate": 8e-05,
      "loss": 1.6417,
      "step": 3299
    },
    {
      "epoch": 0.3566410893764185,
      "grad_norm": 0.40555402636528015,
      "learning_rate": 8e-05,
      "loss": 1.7992,
      "step": 3300
    },
    {
      "epoch": 0.3567491624338053,
      "grad_norm": 0.35167446732521057,
      "learning_rate": 8e-05,
      "loss": 1.6957,
      "step": 3301
    },
    {
      "epoch": 0.356857235491192,
      "grad_norm": 0.40560516715049744,
      "learning_rate": 8e-05,
      "loss": 1.7551,
      "step": 3302
    },
    {
      "epoch": 0.35696530854857883,
      "grad_norm": 0.39252763986587524,
      "learning_rate": 8e-05,
      "loss": 1.6976,
      "step": 3303
    },
    {
      "epoch": 0.35707338160596563,
      "grad_norm": 0.4329412877559662,
      "learning_rate": 8e-05,
      "loss": 1.7916,
      "step": 3304
    },
    {
      "epoch": 0.35718145466335244,
      "grad_norm": 0.35706016421318054,
      "learning_rate": 8e-05,
      "loss": 1.6621,
      "step": 3305
    },
    {
      "epoch": 0.35728952772073924,
      "grad_norm": 0.3676798641681671,
      "learning_rate": 8e-05,
      "loss": 1.7407,
      "step": 3306
    },
    {
      "epoch": 0.357397600778126,
      "grad_norm": 0.3636726438999176,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 3307
    },
    {
      "epoch": 0.3575056738355128,
      "grad_norm": 0.4288215637207031,
      "learning_rate": 8e-05,
      "loss": 1.6917,
      "step": 3308
    },
    {
      "epoch": 0.3576137468928996,
      "grad_norm": 0.3651565909385681,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 3309
    },
    {
      "epoch": 0.3577218199502864,
      "grad_norm": 0.3683786392211914,
      "learning_rate": 8e-05,
      "loss": 1.7331,
      "step": 3310
    },
    {
      "epoch": 0.3578298930076732,
      "grad_norm": 0.44735997915267944,
      "learning_rate": 8e-05,
      "loss": 1.7999,
      "step": 3311
    },
    {
      "epoch": 0.35793796606505995,
      "grad_norm": 0.36708107590675354,
      "learning_rate": 8e-05,
      "loss": 1.6989,
      "step": 3312
    },
    {
      "epoch": 0.35804603912244676,
      "grad_norm": 0.37735840678215027,
      "learning_rate": 8e-05,
      "loss": 1.7366,
      "step": 3313
    },
    {
      "epoch": 0.35815411217983356,
      "grad_norm": 0.4291742444038391,
      "learning_rate": 8e-05,
      "loss": 1.9229,
      "step": 3314
    },
    {
      "epoch": 0.35826218523722037,
      "grad_norm": 0.3743467628955841,
      "learning_rate": 8e-05,
      "loss": 1.6357,
      "step": 3315
    },
    {
      "epoch": 0.35837025829460717,
      "grad_norm": 0.40564483404159546,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 3316
    },
    {
      "epoch": 0.358478331351994,
      "grad_norm": 0.3940487205982208,
      "learning_rate": 8e-05,
      "loss": 1.8145,
      "step": 3317
    },
    {
      "epoch": 0.3585864044093807,
      "grad_norm": 0.39841681718826294,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 3318
    },
    {
      "epoch": 0.3586944774667675,
      "grad_norm": 0.3964144289493561,
      "learning_rate": 8e-05,
      "loss": 1.7935,
      "step": 3319
    },
    {
      "epoch": 0.35880255052415433,
      "grad_norm": 0.3572365343570709,
      "learning_rate": 8e-05,
      "loss": 1.4962,
      "step": 3320
    },
    {
      "epoch": 0.35891062358154113,
      "grad_norm": 0.39503270387649536,
      "learning_rate": 8e-05,
      "loss": 1.7013,
      "step": 3321
    },
    {
      "epoch": 0.35901869663892794,
      "grad_norm": 0.4461931586265564,
      "learning_rate": 8e-05,
      "loss": 1.6896,
      "step": 3322
    },
    {
      "epoch": 0.3591267696963147,
      "grad_norm": 0.4256609082221985,
      "learning_rate": 8e-05,
      "loss": 1.948,
      "step": 3323
    },
    {
      "epoch": 0.3592348427537015,
      "grad_norm": 0.36972132325172424,
      "learning_rate": 8e-05,
      "loss": 1.7372,
      "step": 3324
    },
    {
      "epoch": 0.3593429158110883,
      "grad_norm": 0.3727315366268158,
      "learning_rate": 8e-05,
      "loss": 1.5656,
      "step": 3325
    },
    {
      "epoch": 0.3594509888684751,
      "grad_norm": 0.3745492696762085,
      "learning_rate": 8e-05,
      "loss": 1.8004,
      "step": 3326
    },
    {
      "epoch": 0.3595590619258619,
      "grad_norm": 0.4263870120048523,
      "learning_rate": 8e-05,
      "loss": 1.9016,
      "step": 3327
    },
    {
      "epoch": 0.35966713498324865,
      "grad_norm": 0.3618635833263397,
      "learning_rate": 8e-05,
      "loss": 1.7851,
      "step": 3328
    },
    {
      "epoch": 0.35977520804063545,
      "grad_norm": 0.38164615631103516,
      "learning_rate": 8e-05,
      "loss": 1.6559,
      "step": 3329
    },
    {
      "epoch": 0.35988328109802226,
      "grad_norm": 0.41151195764541626,
      "learning_rate": 8e-05,
      "loss": 1.8747,
      "step": 3330
    },
    {
      "epoch": 0.35999135415540906,
      "grad_norm": 0.43405309319496155,
      "learning_rate": 8e-05,
      "loss": 1.9899,
      "step": 3331
    },
    {
      "epoch": 0.36009942721279586,
      "grad_norm": 0.39584243297576904,
      "learning_rate": 8e-05,
      "loss": 1.7259,
      "step": 3332
    },
    {
      "epoch": 0.36020750027018267,
      "grad_norm": 0.3623722493648529,
      "learning_rate": 8e-05,
      "loss": 1.7052,
      "step": 3333
    },
    {
      "epoch": 0.3603155733275694,
      "grad_norm": 0.37232616543769836,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 3334
    },
    {
      "epoch": 0.3604236463849562,
      "grad_norm": 0.3747260868549347,
      "learning_rate": 8e-05,
      "loss": 1.676,
      "step": 3335
    },
    {
      "epoch": 0.360531719442343,
      "grad_norm": 0.3806154429912567,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 3336
    },
    {
      "epoch": 0.3606397924997298,
      "grad_norm": 0.3754880428314209,
      "learning_rate": 8e-05,
      "loss": 1.768,
      "step": 3337
    },
    {
      "epoch": 0.36074786555711663,
      "grad_norm": 0.42432963848114014,
      "learning_rate": 8e-05,
      "loss": 1.8048,
      "step": 3338
    },
    {
      "epoch": 0.3608559386145034,
      "grad_norm": 0.379656583070755,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 3339
    },
    {
      "epoch": 0.3609640116718902,
      "grad_norm": 0.38290655612945557,
      "learning_rate": 8e-05,
      "loss": 1.7836,
      "step": 3340
    },
    {
      "epoch": 0.361072084729277,
      "grad_norm": 0.3752005100250244,
      "learning_rate": 8e-05,
      "loss": 1.6715,
      "step": 3341
    },
    {
      "epoch": 0.3611801577866638,
      "grad_norm": 0.39085811376571655,
      "learning_rate": 8e-05,
      "loss": 1.6844,
      "step": 3342
    },
    {
      "epoch": 0.3612882308440506,
      "grad_norm": 0.3770289719104767,
      "learning_rate": 8e-05,
      "loss": 1.6067,
      "step": 3343
    },
    {
      "epoch": 0.3613963039014374,
      "grad_norm": 0.42762476205825806,
      "learning_rate": 8e-05,
      "loss": 1.7253,
      "step": 3344
    },
    {
      "epoch": 0.36150437695882415,
      "grad_norm": 0.40134239196777344,
      "learning_rate": 8e-05,
      "loss": 1.8321,
      "step": 3345
    },
    {
      "epoch": 0.36161245001621095,
      "grad_norm": 0.4194454848766327,
      "learning_rate": 8e-05,
      "loss": 1.8929,
      "step": 3346
    },
    {
      "epoch": 0.36172052307359776,
      "grad_norm": 0.37847232818603516,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 3347
    },
    {
      "epoch": 0.36182859613098456,
      "grad_norm": 0.3969729542732239,
      "learning_rate": 8e-05,
      "loss": 1.9152,
      "step": 3348
    },
    {
      "epoch": 0.36193666918837136,
      "grad_norm": 0.4069046974182129,
      "learning_rate": 8e-05,
      "loss": 1.7256,
      "step": 3349
    },
    {
      "epoch": 0.3620447422457581,
      "grad_norm": 0.3857124447822571,
      "learning_rate": 8e-05,
      "loss": 1.7474,
      "step": 3350
    },
    {
      "epoch": 0.3621528153031449,
      "grad_norm": 0.40751519799232483,
      "learning_rate": 8e-05,
      "loss": 1.7865,
      "step": 3351
    },
    {
      "epoch": 0.3622608883605317,
      "grad_norm": 0.40383413434028625,
      "learning_rate": 8e-05,
      "loss": 1.8423,
      "step": 3352
    },
    {
      "epoch": 0.3623689614179185,
      "grad_norm": 0.3802962601184845,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 3353
    },
    {
      "epoch": 0.3624770344753053,
      "grad_norm": 0.4088432490825653,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 3354
    },
    {
      "epoch": 0.3625851075326921,
      "grad_norm": 0.40254339575767517,
      "learning_rate": 8e-05,
      "loss": 1.8043,
      "step": 3355
    },
    {
      "epoch": 0.3626931805900789,
      "grad_norm": 0.43104445934295654,
      "learning_rate": 8e-05,
      "loss": 1.9026,
      "step": 3356
    },
    {
      "epoch": 0.3628012536474657,
      "grad_norm": 0.3573104739189148,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 3357
    },
    {
      "epoch": 0.3629093267048525,
      "grad_norm": 0.35362887382507324,
      "learning_rate": 8e-05,
      "loss": 1.6582,
      "step": 3358
    },
    {
      "epoch": 0.3630173997622393,
      "grad_norm": 0.36536577343940735,
      "learning_rate": 8e-05,
      "loss": 1.6955,
      "step": 3359
    },
    {
      "epoch": 0.3631254728196261,
      "grad_norm": 0.4039801359176636,
      "learning_rate": 8e-05,
      "loss": 1.895,
      "step": 3360
    },
    {
      "epoch": 0.36323354587701284,
      "grad_norm": 0.40417835116386414,
      "learning_rate": 8e-05,
      "loss": 1.7139,
      "step": 3361
    },
    {
      "epoch": 0.36334161893439965,
      "grad_norm": 0.3676578402519226,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 3362
    },
    {
      "epoch": 0.36344969199178645,
      "grad_norm": 0.39095285534858704,
      "learning_rate": 8e-05,
      "loss": 1.7849,
      "step": 3363
    },
    {
      "epoch": 0.36355776504917325,
      "grad_norm": 0.400013267993927,
      "learning_rate": 8e-05,
      "loss": 1.8103,
      "step": 3364
    },
    {
      "epoch": 0.36366583810656006,
      "grad_norm": 0.40802791714668274,
      "learning_rate": 8e-05,
      "loss": 1.7873,
      "step": 3365
    },
    {
      "epoch": 0.3637739111639468,
      "grad_norm": 0.384342759847641,
      "learning_rate": 8e-05,
      "loss": 1.8472,
      "step": 3366
    },
    {
      "epoch": 0.3638819842213336,
      "grad_norm": 0.364883691072464,
      "learning_rate": 8e-05,
      "loss": 1.5781,
      "step": 3367
    },
    {
      "epoch": 0.3639900572787204,
      "grad_norm": 0.40965232253074646,
      "learning_rate": 8e-05,
      "loss": 1.8218,
      "step": 3368
    },
    {
      "epoch": 0.3640981303361072,
      "grad_norm": 0.3790096342563629,
      "learning_rate": 8e-05,
      "loss": 1.6618,
      "step": 3369
    },
    {
      "epoch": 0.364206203393494,
      "grad_norm": 0.382994145154953,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 3370
    },
    {
      "epoch": 0.36431427645088077,
      "grad_norm": 0.3940376043319702,
      "learning_rate": 8e-05,
      "loss": 1.7584,
      "step": 3371
    },
    {
      "epoch": 0.3644223495082676,
      "grad_norm": 0.39777418971061707,
      "learning_rate": 8e-05,
      "loss": 1.8583,
      "step": 3372
    },
    {
      "epoch": 0.3645304225656544,
      "grad_norm": 0.3720598518848419,
      "learning_rate": 8e-05,
      "loss": 1.7014,
      "step": 3373
    },
    {
      "epoch": 0.3646384956230412,
      "grad_norm": 0.35746097564697266,
      "learning_rate": 8e-05,
      "loss": 1.6664,
      "step": 3374
    },
    {
      "epoch": 0.364746568680428,
      "grad_norm": 0.373651921749115,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 3375
    },
    {
      "epoch": 0.3648546417378148,
      "grad_norm": 0.3849944770336151,
      "learning_rate": 8e-05,
      "loss": 1.7099,
      "step": 3376
    },
    {
      "epoch": 0.36496271479520154,
      "grad_norm": 0.3637785315513611,
      "learning_rate": 8e-05,
      "loss": 1.568,
      "step": 3377
    },
    {
      "epoch": 0.36507078785258834,
      "grad_norm": 0.4173211455345154,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 3378
    },
    {
      "epoch": 0.36517886090997514,
      "grad_norm": 0.3767125904560089,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 3379
    },
    {
      "epoch": 0.36528693396736195,
      "grad_norm": 0.389949232339859,
      "learning_rate": 8e-05,
      "loss": 1.821,
      "step": 3380
    },
    {
      "epoch": 0.36539500702474875,
      "grad_norm": 0.3738099932670593,
      "learning_rate": 8e-05,
      "loss": 1.811,
      "step": 3381
    },
    {
      "epoch": 0.3655030800821355,
      "grad_norm": 0.37187492847442627,
      "learning_rate": 8e-05,
      "loss": 1.6008,
      "step": 3382
    },
    {
      "epoch": 0.3656111531395223,
      "grad_norm": 0.39570388197898865,
      "learning_rate": 8e-05,
      "loss": 1.7179,
      "step": 3383
    },
    {
      "epoch": 0.3657192261969091,
      "grad_norm": 0.3885120451450348,
      "learning_rate": 8e-05,
      "loss": 1.7692,
      "step": 3384
    },
    {
      "epoch": 0.3658272992542959,
      "grad_norm": 0.3875708281993866,
      "learning_rate": 8e-05,
      "loss": 1.8532,
      "step": 3385
    },
    {
      "epoch": 0.3659353723116827,
      "grad_norm": 0.38036608695983887,
      "learning_rate": 8e-05,
      "loss": 1.7032,
      "step": 3386
    },
    {
      "epoch": 0.36604344536906946,
      "grad_norm": 0.3747788369655609,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 3387
    },
    {
      "epoch": 0.36615151842645627,
      "grad_norm": 0.3880467414855957,
      "learning_rate": 8e-05,
      "loss": 1.7259,
      "step": 3388
    },
    {
      "epoch": 0.36625959148384307,
      "grad_norm": 0.4004300832748413,
      "learning_rate": 8e-05,
      "loss": 1.8575,
      "step": 3389
    },
    {
      "epoch": 0.3663676645412299,
      "grad_norm": 0.4061131179332733,
      "learning_rate": 8e-05,
      "loss": 1.8773,
      "step": 3390
    },
    {
      "epoch": 0.3664757375986167,
      "grad_norm": 0.3810555934906006,
      "learning_rate": 8e-05,
      "loss": 1.6489,
      "step": 3391
    },
    {
      "epoch": 0.3665838106560035,
      "grad_norm": 0.39178967475891113,
      "learning_rate": 8e-05,
      "loss": 1.8425,
      "step": 3392
    },
    {
      "epoch": 0.36669188371339023,
      "grad_norm": 0.3860965967178345,
      "learning_rate": 8e-05,
      "loss": 1.7477,
      "step": 3393
    },
    {
      "epoch": 0.36679995677077704,
      "grad_norm": 0.446403831243515,
      "learning_rate": 8e-05,
      "loss": 1.5144,
      "step": 3394
    },
    {
      "epoch": 0.36690802982816384,
      "grad_norm": 0.3604002296924591,
      "learning_rate": 8e-05,
      "loss": 1.6053,
      "step": 3395
    },
    {
      "epoch": 0.36701610288555064,
      "grad_norm": 0.38694584369659424,
      "learning_rate": 8e-05,
      "loss": 1.7023,
      "step": 3396
    },
    {
      "epoch": 0.36712417594293745,
      "grad_norm": 0.40981706976890564,
      "learning_rate": 8e-05,
      "loss": 1.7344,
      "step": 3397
    },
    {
      "epoch": 0.3672322490003242,
      "grad_norm": 0.4680512249469757,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 3398
    },
    {
      "epoch": 0.367340322057711,
      "grad_norm": 0.3720478415489197,
      "learning_rate": 8e-05,
      "loss": 1.7709,
      "step": 3399
    },
    {
      "epoch": 0.3674483951150978,
      "grad_norm": 0.36001265048980713,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 3400
    },
    {
      "epoch": 0.3675564681724846,
      "grad_norm": 0.37720987200737,
      "learning_rate": 8e-05,
      "loss": 1.5716,
      "step": 3401
    },
    {
      "epoch": 0.3676645412298714,
      "grad_norm": 0.41126948595046997,
      "learning_rate": 8e-05,
      "loss": 1.7191,
      "step": 3402
    },
    {
      "epoch": 0.36777261428725816,
      "grad_norm": 0.4064192771911621,
      "learning_rate": 8e-05,
      "loss": 1.8318,
      "step": 3403
    },
    {
      "epoch": 0.36788068734464496,
      "grad_norm": 0.3421679139137268,
      "learning_rate": 8e-05,
      "loss": 1.5497,
      "step": 3404
    },
    {
      "epoch": 0.36798876040203177,
      "grad_norm": 0.39555907249450684,
      "learning_rate": 8e-05,
      "loss": 1.8547,
      "step": 3405
    },
    {
      "epoch": 0.36809683345941857,
      "grad_norm": 0.3768436312675476,
      "learning_rate": 8e-05,
      "loss": 1.5171,
      "step": 3406
    },
    {
      "epoch": 0.3682049065168054,
      "grad_norm": 0.45806387066841125,
      "learning_rate": 8e-05,
      "loss": 1.7317,
      "step": 3407
    },
    {
      "epoch": 0.3683129795741922,
      "grad_norm": 0.3954867124557495,
      "learning_rate": 8e-05,
      "loss": 1.7231,
      "step": 3408
    },
    {
      "epoch": 0.3684210526315789,
      "grad_norm": 0.4032396376132965,
      "learning_rate": 8e-05,
      "loss": 1.783,
      "step": 3409
    },
    {
      "epoch": 0.36852912568896573,
      "grad_norm": 0.3578210175037384,
      "learning_rate": 8e-05,
      "loss": 1.6938,
      "step": 3410
    },
    {
      "epoch": 0.36863719874635253,
      "grad_norm": 0.37989288568496704,
      "learning_rate": 8e-05,
      "loss": 1.8005,
      "step": 3411
    },
    {
      "epoch": 0.36874527180373934,
      "grad_norm": 0.42165324091911316,
      "learning_rate": 8e-05,
      "loss": 1.4566,
      "step": 3412
    },
    {
      "epoch": 0.36885334486112614,
      "grad_norm": 0.3587247133255005,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 3413
    },
    {
      "epoch": 0.3689614179185129,
      "grad_norm": 0.39802420139312744,
      "learning_rate": 8e-05,
      "loss": 1.7039,
      "step": 3414
    },
    {
      "epoch": 0.3690694909758997,
      "grad_norm": 0.4061676263809204,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 3415
    },
    {
      "epoch": 0.3691775640332865,
      "grad_norm": 0.4068739116191864,
      "learning_rate": 8e-05,
      "loss": 1.6524,
      "step": 3416
    },
    {
      "epoch": 0.3692856370906733,
      "grad_norm": 0.44787344336509705,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 3417
    },
    {
      "epoch": 0.3693937101480601,
      "grad_norm": 0.3850559890270233,
      "learning_rate": 8e-05,
      "loss": 1.8842,
      "step": 3418
    },
    {
      "epoch": 0.3695017832054469,
      "grad_norm": 0.38400906324386597,
      "learning_rate": 8e-05,
      "loss": 1.6283,
      "step": 3419
    },
    {
      "epoch": 0.36960985626283366,
      "grad_norm": 0.4191437065601349,
      "learning_rate": 8e-05,
      "loss": 2.0712,
      "step": 3420
    },
    {
      "epoch": 0.36971792932022046,
      "grad_norm": 0.37882089614868164,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 3421
    },
    {
      "epoch": 0.36982600237760727,
      "grad_norm": 0.37706458568573,
      "learning_rate": 8e-05,
      "loss": 1.8448,
      "step": 3422
    },
    {
      "epoch": 0.36993407543499407,
      "grad_norm": 0.3884028196334839,
      "learning_rate": 8e-05,
      "loss": 1.7093,
      "step": 3423
    },
    {
      "epoch": 0.3700421484923809,
      "grad_norm": 0.3944375813007355,
      "learning_rate": 8e-05,
      "loss": 1.7908,
      "step": 3424
    },
    {
      "epoch": 0.3701502215497676,
      "grad_norm": 0.37425515055656433,
      "learning_rate": 8e-05,
      "loss": 1.7829,
      "step": 3425
    },
    {
      "epoch": 0.3702582946071544,
      "grad_norm": 0.372321754693985,
      "learning_rate": 8e-05,
      "loss": 1.741,
      "step": 3426
    },
    {
      "epoch": 0.37036636766454123,
      "grad_norm": 0.40048331022262573,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 3427
    },
    {
      "epoch": 0.37047444072192803,
      "grad_norm": 0.38679149746894836,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 3428
    },
    {
      "epoch": 0.37058251377931484,
      "grad_norm": 0.40259772539138794,
      "learning_rate": 8e-05,
      "loss": 1.7207,
      "step": 3429
    },
    {
      "epoch": 0.3706905868367016,
      "grad_norm": 0.40421050786972046,
      "learning_rate": 8e-05,
      "loss": 1.7102,
      "step": 3430
    },
    {
      "epoch": 0.3707986598940884,
      "grad_norm": 0.36750131845474243,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 3431
    },
    {
      "epoch": 0.3709067329514752,
      "grad_norm": 0.4113597571849823,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 3432
    },
    {
      "epoch": 0.371014806008862,
      "grad_norm": 0.3771274983882904,
      "learning_rate": 8e-05,
      "loss": 1.7778,
      "step": 3433
    },
    {
      "epoch": 0.3711228790662488,
      "grad_norm": 0.44363629817962646,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 3434
    },
    {
      "epoch": 0.3712309521236356,
      "grad_norm": 0.36660289764404297,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 3435
    },
    {
      "epoch": 0.37133902518102235,
      "grad_norm": 0.41452327370643616,
      "learning_rate": 8e-05,
      "loss": 1.9428,
      "step": 3436
    },
    {
      "epoch": 0.37144709823840916,
      "grad_norm": 0.4000818431377411,
      "learning_rate": 8e-05,
      "loss": 1.7167,
      "step": 3437
    },
    {
      "epoch": 0.37155517129579596,
      "grad_norm": 0.42190349102020264,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 3438
    },
    {
      "epoch": 0.37166324435318276,
      "grad_norm": 0.3889029622077942,
      "learning_rate": 8e-05,
      "loss": 1.7787,
      "step": 3439
    },
    {
      "epoch": 0.37177131741056957,
      "grad_norm": 0.3782348334789276,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 3440
    },
    {
      "epoch": 0.3718793904679563,
      "grad_norm": 0.396033376455307,
      "learning_rate": 8e-05,
      "loss": 1.6404,
      "step": 3441
    },
    {
      "epoch": 0.3719874635253431,
      "grad_norm": 0.44668343663215637,
      "learning_rate": 8e-05,
      "loss": 1.9855,
      "step": 3442
    },
    {
      "epoch": 0.3720955365827299,
      "grad_norm": 0.38046517968177795,
      "learning_rate": 8e-05,
      "loss": 1.665,
      "step": 3443
    },
    {
      "epoch": 0.37220360964011673,
      "grad_norm": 0.35904237627983093,
      "learning_rate": 8e-05,
      "loss": 1.5858,
      "step": 3444
    },
    {
      "epoch": 0.37231168269750353,
      "grad_norm": 0.41577133536338806,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 3445
    },
    {
      "epoch": 0.3724197557548903,
      "grad_norm": 0.3733823895454407,
      "learning_rate": 8e-05,
      "loss": 1.5635,
      "step": 3446
    },
    {
      "epoch": 0.3725278288122771,
      "grad_norm": 0.391984224319458,
      "learning_rate": 8e-05,
      "loss": 1.837,
      "step": 3447
    },
    {
      "epoch": 0.3726359018696639,
      "grad_norm": 0.39847758412361145,
      "learning_rate": 8e-05,
      "loss": 1.7879,
      "step": 3448
    },
    {
      "epoch": 0.3727439749270507,
      "grad_norm": 0.38917872309684753,
      "learning_rate": 8e-05,
      "loss": 1.7321,
      "step": 3449
    },
    {
      "epoch": 0.3728520479844375,
      "grad_norm": 0.38308292627334595,
      "learning_rate": 8e-05,
      "loss": 1.7025,
      "step": 3450
    },
    {
      "epoch": 0.3729601210418243,
      "grad_norm": 0.3901844918727875,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 3451
    },
    {
      "epoch": 0.37306819409921105,
      "grad_norm": 0.3926748037338257,
      "learning_rate": 8e-05,
      "loss": 1.704,
      "step": 3452
    },
    {
      "epoch": 0.37317626715659785,
      "grad_norm": 0.41543129086494446,
      "learning_rate": 8e-05,
      "loss": 1.9184,
      "step": 3453
    },
    {
      "epoch": 0.37328434021398466,
      "grad_norm": 0.3870210349559784,
      "learning_rate": 8e-05,
      "loss": 1.657,
      "step": 3454
    },
    {
      "epoch": 0.37339241327137146,
      "grad_norm": 0.42889267206192017,
      "learning_rate": 8e-05,
      "loss": 1.8621,
      "step": 3455
    },
    {
      "epoch": 0.37350048632875826,
      "grad_norm": 0.39543601870536804,
      "learning_rate": 8e-05,
      "loss": 1.7097,
      "step": 3456
    },
    {
      "epoch": 0.373608559386145,
      "grad_norm": 0.3894956409931183,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 3457
    },
    {
      "epoch": 0.3737166324435318,
      "grad_norm": 0.4336859881877899,
      "learning_rate": 8e-05,
      "loss": 1.871,
      "step": 3458
    },
    {
      "epoch": 0.3738247055009186,
      "grad_norm": 0.3818238079547882,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 3459
    },
    {
      "epoch": 0.3739327785583054,
      "grad_norm": 0.365071564912796,
      "learning_rate": 8e-05,
      "loss": 1.6246,
      "step": 3460
    },
    {
      "epoch": 0.3740408516156922,
      "grad_norm": 0.4249206781387329,
      "learning_rate": 8e-05,
      "loss": 1.7509,
      "step": 3461
    },
    {
      "epoch": 0.374148924673079,
      "grad_norm": 0.37332838773727417,
      "learning_rate": 8e-05,
      "loss": 1.8074,
      "step": 3462
    },
    {
      "epoch": 0.3742569977304658,
      "grad_norm": 0.38897836208343506,
      "learning_rate": 8e-05,
      "loss": 1.7653,
      "step": 3463
    },
    {
      "epoch": 0.3743650707878526,
      "grad_norm": 0.35265931487083435,
      "learning_rate": 8e-05,
      "loss": 1.5697,
      "step": 3464
    },
    {
      "epoch": 0.3744731438452394,
      "grad_norm": 0.4293357729911804,
      "learning_rate": 8e-05,
      "loss": 1.9529,
      "step": 3465
    },
    {
      "epoch": 0.3745812169026262,
      "grad_norm": 0.4293860197067261,
      "learning_rate": 8e-05,
      "loss": 1.7133,
      "step": 3466
    },
    {
      "epoch": 0.374689289960013,
      "grad_norm": 0.39301517605781555,
      "learning_rate": 8e-05,
      "loss": 1.8131,
      "step": 3467
    },
    {
      "epoch": 0.37479736301739974,
      "grad_norm": 0.3787461817264557,
      "learning_rate": 8e-05,
      "loss": 1.7794,
      "step": 3468
    },
    {
      "epoch": 0.37490543607478655,
      "grad_norm": 0.38320720195770264,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 3469
    },
    {
      "epoch": 0.37501350913217335,
      "grad_norm": 0.4076857566833496,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 3470
    },
    {
      "epoch": 0.37512158218956015,
      "grad_norm": 0.4291606545448303,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 3471
    },
    {
      "epoch": 0.37522965524694696,
      "grad_norm": 0.38234496116638184,
      "learning_rate": 8e-05,
      "loss": 1.759,
      "step": 3472
    },
    {
      "epoch": 0.3753377283043337,
      "grad_norm": 0.36409610509872437,
      "learning_rate": 8e-05,
      "loss": 1.6945,
      "step": 3473
    },
    {
      "epoch": 0.3754458013617205,
      "grad_norm": 0.39078328013420105,
      "learning_rate": 8e-05,
      "loss": 1.756,
      "step": 3474
    },
    {
      "epoch": 0.3755538744191073,
      "grad_norm": 0.41482365131378174,
      "learning_rate": 8e-05,
      "loss": 1.7691,
      "step": 3475
    },
    {
      "epoch": 0.3756619474764941,
      "grad_norm": 0.3957419693470001,
      "learning_rate": 8e-05,
      "loss": 1.7109,
      "step": 3476
    },
    {
      "epoch": 0.3757700205338809,
      "grad_norm": 0.39946168661117554,
      "learning_rate": 8e-05,
      "loss": 1.7913,
      "step": 3477
    },
    {
      "epoch": 0.37587809359126767,
      "grad_norm": 0.3596000671386719,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 3478
    },
    {
      "epoch": 0.3759861666486545,
      "grad_norm": 0.40265509486198425,
      "learning_rate": 8e-05,
      "loss": 1.6563,
      "step": 3479
    },
    {
      "epoch": 0.3760942397060413,
      "grad_norm": 0.3781740367412567,
      "learning_rate": 8e-05,
      "loss": 1.756,
      "step": 3480
    },
    {
      "epoch": 0.3762023127634281,
      "grad_norm": 0.39202919602394104,
      "learning_rate": 8e-05,
      "loss": 1.6537,
      "step": 3481
    },
    {
      "epoch": 0.3763103858208149,
      "grad_norm": 0.3824671804904938,
      "learning_rate": 8e-05,
      "loss": 1.6935,
      "step": 3482
    },
    {
      "epoch": 0.3764184588782017,
      "grad_norm": 0.4187178313732147,
      "learning_rate": 8e-05,
      "loss": 1.8507,
      "step": 3483
    },
    {
      "epoch": 0.37652653193558844,
      "grad_norm": 0.3949556350708008,
      "learning_rate": 8e-05,
      "loss": 1.689,
      "step": 3484
    },
    {
      "epoch": 0.37663460499297524,
      "grad_norm": 0.4136996865272522,
      "learning_rate": 8e-05,
      "loss": 1.5416,
      "step": 3485
    },
    {
      "epoch": 0.37674267805036205,
      "grad_norm": 0.4144202172756195,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 3486
    },
    {
      "epoch": 0.37685075110774885,
      "grad_norm": 0.38980695605278015,
      "learning_rate": 8e-05,
      "loss": 1.6762,
      "step": 3487
    },
    {
      "epoch": 0.37695882416513565,
      "grad_norm": 0.3771219551563263,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 3488
    },
    {
      "epoch": 0.3770668972225224,
      "grad_norm": 0.3977028429508209,
      "learning_rate": 8e-05,
      "loss": 1.6849,
      "step": 3489
    },
    {
      "epoch": 0.3771749702799092,
      "grad_norm": 0.3851807117462158,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 3490
    },
    {
      "epoch": 0.377283043337296,
      "grad_norm": 0.39543524384498596,
      "learning_rate": 8e-05,
      "loss": 1.6327,
      "step": 3491
    },
    {
      "epoch": 0.3773911163946828,
      "grad_norm": 0.4074676036834717,
      "learning_rate": 8e-05,
      "loss": 1.703,
      "step": 3492
    },
    {
      "epoch": 0.3774991894520696,
      "grad_norm": 0.4018513262271881,
      "learning_rate": 8e-05,
      "loss": 1.7717,
      "step": 3493
    },
    {
      "epoch": 0.37760726250945637,
      "grad_norm": 0.39017102122306824,
      "learning_rate": 8e-05,
      "loss": 1.7066,
      "step": 3494
    },
    {
      "epoch": 0.37771533556684317,
      "grad_norm": 0.38313257694244385,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 3495
    },
    {
      "epoch": 0.37782340862423,
      "grad_norm": 0.393898606300354,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 3496
    },
    {
      "epoch": 0.3779314816816168,
      "grad_norm": 0.37563177943229675,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 3497
    },
    {
      "epoch": 0.3780395547390036,
      "grad_norm": 0.3859784007072449,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 3498
    },
    {
      "epoch": 0.3781476277963904,
      "grad_norm": 0.40964189171791077,
      "learning_rate": 8e-05,
      "loss": 1.8494,
      "step": 3499
    },
    {
      "epoch": 0.37825570085377713,
      "grad_norm": 0.37586721777915955,
      "learning_rate": 8e-05,
      "loss": 1.696,
      "step": 3500
    },
    {
      "epoch": 0.37836377391116394,
      "grad_norm": 0.41263341903686523,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 3501
    },
    {
      "epoch": 0.37847184696855074,
      "grad_norm": 0.42650192975997925,
      "learning_rate": 8e-05,
      "loss": 1.8379,
      "step": 3502
    },
    {
      "epoch": 0.37857992002593754,
      "grad_norm": 0.38107308745384216,
      "learning_rate": 8e-05,
      "loss": 1.7373,
      "step": 3503
    },
    {
      "epoch": 0.37868799308332435,
      "grad_norm": 0.4120958745479584,
      "learning_rate": 8e-05,
      "loss": 1.7522,
      "step": 3504
    },
    {
      "epoch": 0.3787960661407111,
      "grad_norm": 0.40307390689849854,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 3505
    },
    {
      "epoch": 0.3789041391980979,
      "grad_norm": 0.37662678956985474,
      "learning_rate": 8e-05,
      "loss": 1.6681,
      "step": 3506
    },
    {
      "epoch": 0.3790122122554847,
      "grad_norm": 0.38320988416671753,
      "learning_rate": 8e-05,
      "loss": 1.6575,
      "step": 3507
    },
    {
      "epoch": 0.3791202853128715,
      "grad_norm": 0.36323174834251404,
      "learning_rate": 8e-05,
      "loss": 1.5795,
      "step": 3508
    },
    {
      "epoch": 0.3792283583702583,
      "grad_norm": 0.37993577122688293,
      "learning_rate": 8e-05,
      "loss": 1.7888,
      "step": 3509
    },
    {
      "epoch": 0.3793364314276451,
      "grad_norm": 0.3733411729335785,
      "learning_rate": 8e-05,
      "loss": 1.7635,
      "step": 3510
    },
    {
      "epoch": 0.37944450448503186,
      "grad_norm": 0.36715659499168396,
      "learning_rate": 8e-05,
      "loss": 1.6497,
      "step": 3511
    },
    {
      "epoch": 0.37955257754241867,
      "grad_norm": 0.4003114700317383,
      "learning_rate": 8e-05,
      "loss": 1.8552,
      "step": 3512
    },
    {
      "epoch": 0.37966065059980547,
      "grad_norm": 0.3740144371986389,
      "learning_rate": 8e-05,
      "loss": 1.6465,
      "step": 3513
    },
    {
      "epoch": 0.3797687236571923,
      "grad_norm": 0.3640976548194885,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 3514
    },
    {
      "epoch": 0.3798767967145791,
      "grad_norm": 0.3802964389324188,
      "learning_rate": 8e-05,
      "loss": 1.7625,
      "step": 3515
    },
    {
      "epoch": 0.3799848697719658,
      "grad_norm": 0.38248687982559204,
      "learning_rate": 8e-05,
      "loss": 1.6763,
      "step": 3516
    },
    {
      "epoch": 0.38009294282935263,
      "grad_norm": 0.41036197543144226,
      "learning_rate": 8e-05,
      "loss": 1.8767,
      "step": 3517
    },
    {
      "epoch": 0.38020101588673944,
      "grad_norm": 0.363010048866272,
      "learning_rate": 8e-05,
      "loss": 1.6469,
      "step": 3518
    },
    {
      "epoch": 0.38030908894412624,
      "grad_norm": 0.3722565472126007,
      "learning_rate": 8e-05,
      "loss": 1.7319,
      "step": 3519
    },
    {
      "epoch": 0.38041716200151304,
      "grad_norm": 0.40870213508605957,
      "learning_rate": 8e-05,
      "loss": 1.7385,
      "step": 3520
    },
    {
      "epoch": 0.3805252350588998,
      "grad_norm": 0.4093620479106903,
      "learning_rate": 8e-05,
      "loss": 1.728,
      "step": 3521
    },
    {
      "epoch": 0.3806333081162866,
      "grad_norm": 0.431355357170105,
      "learning_rate": 8e-05,
      "loss": 1.9047,
      "step": 3522
    },
    {
      "epoch": 0.3807413811736734,
      "grad_norm": 0.39238104224205017,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 3523
    },
    {
      "epoch": 0.3808494542310602,
      "grad_norm": 0.4006044566631317,
      "learning_rate": 8e-05,
      "loss": 1.8088,
      "step": 3524
    },
    {
      "epoch": 0.380957527288447,
      "grad_norm": 0.3569253981113434,
      "learning_rate": 8e-05,
      "loss": 1.7129,
      "step": 3525
    },
    {
      "epoch": 0.3810656003458338,
      "grad_norm": 0.41473475098609924,
      "learning_rate": 8e-05,
      "loss": 1.8542,
      "step": 3526
    },
    {
      "epoch": 0.38117367340322056,
      "grad_norm": 0.36449646949768066,
      "learning_rate": 8e-05,
      "loss": 1.5846,
      "step": 3527
    },
    {
      "epoch": 0.38128174646060736,
      "grad_norm": 0.38516658544540405,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 3528
    },
    {
      "epoch": 0.38138981951799417,
      "grad_norm": 0.40685662627220154,
      "learning_rate": 8e-05,
      "loss": 1.8684,
      "step": 3529
    },
    {
      "epoch": 0.38149789257538097,
      "grad_norm": 0.416901171207428,
      "learning_rate": 8e-05,
      "loss": 1.782,
      "step": 3530
    },
    {
      "epoch": 0.3816059656327678,
      "grad_norm": 0.37192219495773315,
      "learning_rate": 8e-05,
      "loss": 1.7765,
      "step": 3531
    },
    {
      "epoch": 0.3817140386901545,
      "grad_norm": 0.38957494497299194,
      "learning_rate": 8e-05,
      "loss": 1.7564,
      "step": 3532
    },
    {
      "epoch": 0.3818221117475413,
      "grad_norm": 0.3893946409225464,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 3533
    },
    {
      "epoch": 0.38193018480492813,
      "grad_norm": 0.42309218645095825,
      "learning_rate": 8e-05,
      "loss": 1.7674,
      "step": 3534
    },
    {
      "epoch": 0.38203825786231493,
      "grad_norm": 0.4150150418281555,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 3535
    },
    {
      "epoch": 0.38214633091970174,
      "grad_norm": 0.42698100209236145,
      "learning_rate": 8e-05,
      "loss": 1.7837,
      "step": 3536
    },
    {
      "epoch": 0.3822544039770885,
      "grad_norm": 0.3650544285774231,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 3537
    },
    {
      "epoch": 0.3823624770344753,
      "grad_norm": 0.38779762387275696,
      "learning_rate": 8e-05,
      "loss": 1.8059,
      "step": 3538
    },
    {
      "epoch": 0.3824705500918621,
      "grad_norm": 0.37549030780792236,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 3539
    },
    {
      "epoch": 0.3825786231492489,
      "grad_norm": 0.38927751779556274,
      "learning_rate": 8e-05,
      "loss": 1.6573,
      "step": 3540
    },
    {
      "epoch": 0.3826866962066357,
      "grad_norm": 0.3979440927505493,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 3541
    },
    {
      "epoch": 0.3827947692640225,
      "grad_norm": 0.4279042184352875,
      "learning_rate": 8e-05,
      "loss": 1.7218,
      "step": 3542
    },
    {
      "epoch": 0.38290284232140925,
      "grad_norm": 0.37709760665893555,
      "learning_rate": 8e-05,
      "loss": 1.7152,
      "step": 3543
    },
    {
      "epoch": 0.38301091537879606,
      "grad_norm": 0.41234949231147766,
      "learning_rate": 8e-05,
      "loss": 1.6119,
      "step": 3544
    },
    {
      "epoch": 0.38311898843618286,
      "grad_norm": 0.37349435687065125,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 3545
    },
    {
      "epoch": 0.38322706149356967,
      "grad_norm": 0.41712650656700134,
      "learning_rate": 8e-05,
      "loss": 1.8228,
      "step": 3546
    },
    {
      "epoch": 0.38333513455095647,
      "grad_norm": 0.4010965824127197,
      "learning_rate": 8e-05,
      "loss": 1.592,
      "step": 3547
    },
    {
      "epoch": 0.3834432076083432,
      "grad_norm": 0.3669918477535248,
      "learning_rate": 8e-05,
      "loss": 1.6656,
      "step": 3548
    },
    {
      "epoch": 0.38355128066573,
      "grad_norm": 0.40519875288009644,
      "learning_rate": 8e-05,
      "loss": 1.8051,
      "step": 3549
    },
    {
      "epoch": 0.3836593537231168,
      "grad_norm": 0.3981151878833771,
      "learning_rate": 8e-05,
      "loss": 1.8568,
      "step": 3550
    },
    {
      "epoch": 0.38376742678050363,
      "grad_norm": 0.3808800280094147,
      "learning_rate": 8e-05,
      "loss": 1.7777,
      "step": 3551
    },
    {
      "epoch": 0.38387549983789043,
      "grad_norm": 0.3973964750766754,
      "learning_rate": 8e-05,
      "loss": 1.8324,
      "step": 3552
    },
    {
      "epoch": 0.3839835728952772,
      "grad_norm": 0.37368136644363403,
      "learning_rate": 8e-05,
      "loss": 1.814,
      "step": 3553
    },
    {
      "epoch": 0.384091645952664,
      "grad_norm": 0.3900277316570282,
      "learning_rate": 8e-05,
      "loss": 1.8975,
      "step": 3554
    },
    {
      "epoch": 0.3841997190100508,
      "grad_norm": 0.4191446006298065,
      "learning_rate": 8e-05,
      "loss": 1.8391,
      "step": 3555
    },
    {
      "epoch": 0.3843077920674376,
      "grad_norm": 0.44933372735977173,
      "learning_rate": 8e-05,
      "loss": 1.9008,
      "step": 3556
    },
    {
      "epoch": 0.3844158651248244,
      "grad_norm": 0.44189462065696716,
      "learning_rate": 8e-05,
      "loss": 1.9726,
      "step": 3557
    },
    {
      "epoch": 0.3845239381822112,
      "grad_norm": 0.37624919414520264,
      "learning_rate": 8e-05,
      "loss": 1.6757,
      "step": 3558
    },
    {
      "epoch": 0.38463201123959795,
      "grad_norm": 0.3720269501209259,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 3559
    },
    {
      "epoch": 0.38474008429698475,
      "grad_norm": 0.4034567177295685,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 3560
    },
    {
      "epoch": 0.38484815735437156,
      "grad_norm": 0.3816892206668854,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 3561
    },
    {
      "epoch": 0.38495623041175836,
      "grad_norm": 0.404674768447876,
      "learning_rate": 8e-05,
      "loss": 1.7227,
      "step": 3562
    },
    {
      "epoch": 0.38506430346914516,
      "grad_norm": 0.44607359170913696,
      "learning_rate": 8e-05,
      "loss": 1.8034,
      "step": 3563
    },
    {
      "epoch": 0.3851723765265319,
      "grad_norm": 0.3940983712673187,
      "learning_rate": 8e-05,
      "loss": 1.8253,
      "step": 3564
    },
    {
      "epoch": 0.3852804495839187,
      "grad_norm": 0.40224984288215637,
      "learning_rate": 8e-05,
      "loss": 1.801,
      "step": 3565
    },
    {
      "epoch": 0.3853885226413055,
      "grad_norm": 0.39375922083854675,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 3566
    },
    {
      "epoch": 0.3854965956986923,
      "grad_norm": 0.403225839138031,
      "learning_rate": 8e-05,
      "loss": 1.779,
      "step": 3567
    },
    {
      "epoch": 0.3856046687560791,
      "grad_norm": 0.4228249788284302,
      "learning_rate": 8e-05,
      "loss": 1.7979,
      "step": 3568
    },
    {
      "epoch": 0.3857127418134659,
      "grad_norm": 0.39948219060897827,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 3569
    },
    {
      "epoch": 0.3858208148708527,
      "grad_norm": 0.3792552947998047,
      "learning_rate": 8e-05,
      "loss": 1.7587,
      "step": 3570
    },
    {
      "epoch": 0.3859288879282395,
      "grad_norm": 0.3772358298301697,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 3571
    },
    {
      "epoch": 0.3860369609856263,
      "grad_norm": 0.3831201493740082,
      "learning_rate": 8e-05,
      "loss": 1.7019,
      "step": 3572
    },
    {
      "epoch": 0.3861450340430131,
      "grad_norm": 0.419120728969574,
      "learning_rate": 8e-05,
      "loss": 1.7269,
      "step": 3573
    },
    {
      "epoch": 0.3862531071003999,
      "grad_norm": 0.3983045518398285,
      "learning_rate": 8e-05,
      "loss": 1.7789,
      "step": 3574
    },
    {
      "epoch": 0.38636118015778664,
      "grad_norm": 0.40484148263931274,
      "learning_rate": 8e-05,
      "loss": 1.6836,
      "step": 3575
    },
    {
      "epoch": 0.38646925321517345,
      "grad_norm": 0.4391734004020691,
      "learning_rate": 8e-05,
      "loss": 1.6993,
      "step": 3576
    },
    {
      "epoch": 0.38657732627256025,
      "grad_norm": 0.4135514199733734,
      "learning_rate": 8e-05,
      "loss": 1.9047,
      "step": 3577
    },
    {
      "epoch": 0.38668539932994705,
      "grad_norm": 0.38191351294517517,
      "learning_rate": 8e-05,
      "loss": 1.4872,
      "step": 3578
    },
    {
      "epoch": 0.38679347238733386,
      "grad_norm": 0.4104292094707489,
      "learning_rate": 8e-05,
      "loss": 1.7859,
      "step": 3579
    },
    {
      "epoch": 0.3869015454447206,
      "grad_norm": 0.39411255717277527,
      "learning_rate": 8e-05,
      "loss": 1.7829,
      "step": 3580
    },
    {
      "epoch": 0.3870096185021074,
      "grad_norm": 0.360411137342453,
      "learning_rate": 8e-05,
      "loss": 1.5311,
      "step": 3581
    },
    {
      "epoch": 0.3871176915594942,
      "grad_norm": 0.43718859553337097,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 3582
    },
    {
      "epoch": 0.387225764616881,
      "grad_norm": 0.38726693391799927,
      "learning_rate": 8e-05,
      "loss": 1.758,
      "step": 3583
    },
    {
      "epoch": 0.3873338376742678,
      "grad_norm": 0.3864075541496277,
      "learning_rate": 8e-05,
      "loss": 1.7241,
      "step": 3584
    },
    {
      "epoch": 0.3874419107316546,
      "grad_norm": 0.40262165665626526,
      "learning_rate": 8e-05,
      "loss": 1.6771,
      "step": 3585
    },
    {
      "epoch": 0.3875499837890414,
      "grad_norm": 0.37532612681388855,
      "learning_rate": 8e-05,
      "loss": 1.6484,
      "step": 3586
    },
    {
      "epoch": 0.3876580568464282,
      "grad_norm": 0.38790810108184814,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 3587
    },
    {
      "epoch": 0.387766129903815,
      "grad_norm": 0.3883034884929657,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 3588
    },
    {
      "epoch": 0.3878742029612018,
      "grad_norm": 0.4685097336769104,
      "learning_rate": 8e-05,
      "loss": 1.6946,
      "step": 3589
    },
    {
      "epoch": 0.3879822760185886,
      "grad_norm": 0.4104948043823242,
      "learning_rate": 8e-05,
      "loss": 1.7799,
      "step": 3590
    },
    {
      "epoch": 0.38809034907597534,
      "grad_norm": 0.4495389759540558,
      "learning_rate": 8e-05,
      "loss": 1.715,
      "step": 3591
    },
    {
      "epoch": 0.38819842213336214,
      "grad_norm": 0.4029026925563812,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 3592
    },
    {
      "epoch": 0.38830649519074895,
      "grad_norm": 0.3995441794395447,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 3593
    },
    {
      "epoch": 0.38841456824813575,
      "grad_norm": 0.4161490201950073,
      "learning_rate": 8e-05,
      "loss": 1.8435,
      "step": 3594
    },
    {
      "epoch": 0.38852264130552255,
      "grad_norm": 0.3666999638080597,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 3595
    },
    {
      "epoch": 0.3886307143629093,
      "grad_norm": 0.440245658159256,
      "learning_rate": 8e-05,
      "loss": 1.7586,
      "step": 3596
    },
    {
      "epoch": 0.3887387874202961,
      "grad_norm": 0.4088456928730011,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 3597
    },
    {
      "epoch": 0.3888468604776829,
      "grad_norm": 0.4411187469959259,
      "learning_rate": 8e-05,
      "loss": 1.9444,
      "step": 3598
    },
    {
      "epoch": 0.3889549335350697,
      "grad_norm": 0.42688512802124023,
      "learning_rate": 8e-05,
      "loss": 1.7577,
      "step": 3599
    },
    {
      "epoch": 0.3890630065924565,
      "grad_norm": 0.4002804160118103,
      "learning_rate": 8e-05,
      "loss": 1.6736,
      "step": 3600
    },
    {
      "epoch": 0.3891710796498433,
      "grad_norm": 0.496926873922348,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 3601
    },
    {
      "epoch": 0.38927915270723007,
      "grad_norm": 0.369089275598526,
      "learning_rate": 8e-05,
      "loss": 1.695,
      "step": 3602
    },
    {
      "epoch": 0.3893872257646169,
      "grad_norm": 0.3849320113658905,
      "learning_rate": 8e-05,
      "loss": 1.7844,
      "step": 3603
    },
    {
      "epoch": 0.3894952988220037,
      "grad_norm": 0.3766508996486664,
      "learning_rate": 8e-05,
      "loss": 1.6605,
      "step": 3604
    },
    {
      "epoch": 0.3896033718793905,
      "grad_norm": 0.4254394769668579,
      "learning_rate": 8e-05,
      "loss": 1.6606,
      "step": 3605
    },
    {
      "epoch": 0.3897114449367773,
      "grad_norm": 0.3903300166130066,
      "learning_rate": 8e-05,
      "loss": 1.7525,
      "step": 3606
    },
    {
      "epoch": 0.38981951799416403,
      "grad_norm": 0.4021805226802826,
      "learning_rate": 8e-05,
      "loss": 1.8197,
      "step": 3607
    },
    {
      "epoch": 0.38992759105155084,
      "grad_norm": 0.40100282430648804,
      "learning_rate": 8e-05,
      "loss": 1.7617,
      "step": 3608
    },
    {
      "epoch": 0.39003566410893764,
      "grad_norm": 0.398359090089798,
      "learning_rate": 8e-05,
      "loss": 1.7468,
      "step": 3609
    },
    {
      "epoch": 0.39014373716632444,
      "grad_norm": 0.4622156023979187,
      "learning_rate": 8e-05,
      "loss": 1.8503,
      "step": 3610
    },
    {
      "epoch": 0.39025181022371125,
      "grad_norm": 0.40024054050445557,
      "learning_rate": 8e-05,
      "loss": 1.7419,
      "step": 3611
    },
    {
      "epoch": 0.390359883281098,
      "grad_norm": 0.404221773147583,
      "learning_rate": 8e-05,
      "loss": 1.4782,
      "step": 3612
    },
    {
      "epoch": 0.3904679563384848,
      "grad_norm": 0.4110901653766632,
      "learning_rate": 8e-05,
      "loss": 1.8794,
      "step": 3613
    },
    {
      "epoch": 0.3905760293958716,
      "grad_norm": 0.4060975909233093,
      "learning_rate": 8e-05,
      "loss": 1.7728,
      "step": 3614
    },
    {
      "epoch": 0.3906841024532584,
      "grad_norm": 0.3533060848712921,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 3615
    },
    {
      "epoch": 0.3907921755106452,
      "grad_norm": 0.39177361130714417,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 3616
    },
    {
      "epoch": 0.390900248568032,
      "grad_norm": 0.3647722899913788,
      "learning_rate": 8e-05,
      "loss": 1.812,
      "step": 3617
    },
    {
      "epoch": 0.39100832162541876,
      "grad_norm": 0.3958182632923126,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 3618
    },
    {
      "epoch": 0.39111639468280557,
      "grad_norm": 0.383961021900177,
      "learning_rate": 8e-05,
      "loss": 1.743,
      "step": 3619
    },
    {
      "epoch": 0.39122446774019237,
      "grad_norm": 0.4106149971485138,
      "learning_rate": 8e-05,
      "loss": 1.7991,
      "step": 3620
    },
    {
      "epoch": 0.3913325407975792,
      "grad_norm": 0.3814200460910797,
      "learning_rate": 8e-05,
      "loss": 1.6136,
      "step": 3621
    },
    {
      "epoch": 0.391440613854966,
      "grad_norm": 0.42237281799316406,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 3622
    },
    {
      "epoch": 0.39154868691235273,
      "grad_norm": 0.36670219898223877,
      "learning_rate": 8e-05,
      "loss": 1.5479,
      "step": 3623
    },
    {
      "epoch": 0.39165675996973953,
      "grad_norm": 0.3839063346385956,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 3624
    },
    {
      "epoch": 0.39176483302712634,
      "grad_norm": 0.3889945149421692,
      "learning_rate": 8e-05,
      "loss": 1.6123,
      "step": 3625
    },
    {
      "epoch": 0.39187290608451314,
      "grad_norm": 0.4028182625770569,
      "learning_rate": 8e-05,
      "loss": 1.7791,
      "step": 3626
    },
    {
      "epoch": 0.39198097914189994,
      "grad_norm": 0.3862074613571167,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 3627
    },
    {
      "epoch": 0.3920890521992867,
      "grad_norm": 0.3968062102794647,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 3628
    },
    {
      "epoch": 0.3921971252566735,
      "grad_norm": 0.4119967520236969,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 3629
    },
    {
      "epoch": 0.3923051983140603,
      "grad_norm": 0.38931092619895935,
      "learning_rate": 8e-05,
      "loss": 1.6827,
      "step": 3630
    },
    {
      "epoch": 0.3924132713714471,
      "grad_norm": 0.39359790086746216,
      "learning_rate": 8e-05,
      "loss": 1.7995,
      "step": 3631
    },
    {
      "epoch": 0.3925213444288339,
      "grad_norm": 0.39946842193603516,
      "learning_rate": 8e-05,
      "loss": 1.8105,
      "step": 3632
    },
    {
      "epoch": 0.3926294174862207,
      "grad_norm": 0.3895476460456848,
      "learning_rate": 8e-05,
      "loss": 1.7693,
      "step": 3633
    },
    {
      "epoch": 0.39273749054360746,
      "grad_norm": 0.4143219292163849,
      "learning_rate": 8e-05,
      "loss": 1.7003,
      "step": 3634
    },
    {
      "epoch": 0.39284556360099426,
      "grad_norm": 0.41944581270217896,
      "learning_rate": 8e-05,
      "loss": 1.7369,
      "step": 3635
    },
    {
      "epoch": 0.39295363665838107,
      "grad_norm": 0.4190118908882141,
      "learning_rate": 8e-05,
      "loss": 1.9293,
      "step": 3636
    },
    {
      "epoch": 0.39306170971576787,
      "grad_norm": 0.40957459807395935,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 3637
    },
    {
      "epoch": 0.3931697827731547,
      "grad_norm": 0.5004032850265503,
      "learning_rate": 8e-05,
      "loss": 1.9849,
      "step": 3638
    },
    {
      "epoch": 0.3932778558305414,
      "grad_norm": 0.38745835423469543,
      "learning_rate": 8e-05,
      "loss": 1.8154,
      "step": 3639
    },
    {
      "epoch": 0.3933859288879282,
      "grad_norm": 0.37838202714920044,
      "learning_rate": 8e-05,
      "loss": 1.691,
      "step": 3640
    },
    {
      "epoch": 0.39349400194531503,
      "grad_norm": 0.4033721387386322,
      "learning_rate": 8e-05,
      "loss": 1.8427,
      "step": 3641
    },
    {
      "epoch": 0.39360207500270183,
      "grad_norm": 0.4197179675102234,
      "learning_rate": 8e-05,
      "loss": 1.724,
      "step": 3642
    },
    {
      "epoch": 0.39371014806008864,
      "grad_norm": 0.42030149698257446,
      "learning_rate": 8e-05,
      "loss": 1.6604,
      "step": 3643
    },
    {
      "epoch": 0.3938182211174754,
      "grad_norm": 0.3908644914627075,
      "learning_rate": 8e-05,
      "loss": 1.7046,
      "step": 3644
    },
    {
      "epoch": 0.3939262941748622,
      "grad_norm": 0.41252681612968445,
      "learning_rate": 8e-05,
      "loss": 1.6735,
      "step": 3645
    },
    {
      "epoch": 0.394034367232249,
      "grad_norm": 0.41587528586387634,
      "learning_rate": 8e-05,
      "loss": 1.8748,
      "step": 3646
    },
    {
      "epoch": 0.3941424402896358,
      "grad_norm": 0.40265074372291565,
      "learning_rate": 8e-05,
      "loss": 1.785,
      "step": 3647
    },
    {
      "epoch": 0.3942505133470226,
      "grad_norm": 0.3616083264350891,
      "learning_rate": 8e-05,
      "loss": 1.5792,
      "step": 3648
    },
    {
      "epoch": 0.3943585864044094,
      "grad_norm": 0.4082064628601074,
      "learning_rate": 8e-05,
      "loss": 1.8936,
      "step": 3649
    },
    {
      "epoch": 0.39446665946179615,
      "grad_norm": 0.38844430446624756,
      "learning_rate": 8e-05,
      "loss": 1.7387,
      "step": 3650
    },
    {
      "epoch": 0.39457473251918296,
      "grad_norm": 0.3799108862876892,
      "learning_rate": 8e-05,
      "loss": 1.7097,
      "step": 3651
    },
    {
      "epoch": 0.39468280557656976,
      "grad_norm": 0.39619314670562744,
      "learning_rate": 8e-05,
      "loss": 1.7037,
      "step": 3652
    },
    {
      "epoch": 0.39479087863395657,
      "grad_norm": 0.4167812764644623,
      "learning_rate": 8e-05,
      "loss": 1.8067,
      "step": 3653
    },
    {
      "epoch": 0.39489895169134337,
      "grad_norm": 0.3990997076034546,
      "learning_rate": 8e-05,
      "loss": 1.74,
      "step": 3654
    },
    {
      "epoch": 0.3950070247487301,
      "grad_norm": 0.4231667220592499,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 3655
    },
    {
      "epoch": 0.3951150978061169,
      "grad_norm": 0.3799695670604706,
      "learning_rate": 8e-05,
      "loss": 1.8591,
      "step": 3656
    },
    {
      "epoch": 0.3952231708635037,
      "grad_norm": 0.40398335456848145,
      "learning_rate": 8e-05,
      "loss": 1.6879,
      "step": 3657
    },
    {
      "epoch": 0.39533124392089053,
      "grad_norm": 0.43385791778564453,
      "learning_rate": 8e-05,
      "loss": 1.7369,
      "step": 3658
    },
    {
      "epoch": 0.39543931697827733,
      "grad_norm": 0.39882421493530273,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 3659
    },
    {
      "epoch": 0.3955473900356641,
      "grad_norm": 0.3708348870277405,
      "learning_rate": 8e-05,
      "loss": 1.6417,
      "step": 3660
    },
    {
      "epoch": 0.3956554630930509,
      "grad_norm": 0.36885106563568115,
      "learning_rate": 8e-05,
      "loss": 1.4198,
      "step": 3661
    },
    {
      "epoch": 0.3957635361504377,
      "grad_norm": 0.38646528124809265,
      "learning_rate": 8e-05,
      "loss": 1.7347,
      "step": 3662
    },
    {
      "epoch": 0.3958716092078245,
      "grad_norm": 0.3950575590133667,
      "learning_rate": 8e-05,
      "loss": 1.7639,
      "step": 3663
    },
    {
      "epoch": 0.3959796822652113,
      "grad_norm": 0.40876105427742004,
      "learning_rate": 8e-05,
      "loss": 1.6726,
      "step": 3664
    },
    {
      "epoch": 0.3960877553225981,
      "grad_norm": 0.4051937460899353,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 3665
    },
    {
      "epoch": 0.39619582837998485,
      "grad_norm": 0.4139687716960907,
      "learning_rate": 8e-05,
      "loss": 1.9009,
      "step": 3666
    },
    {
      "epoch": 0.39630390143737165,
      "grad_norm": 0.3834027647972107,
      "learning_rate": 8e-05,
      "loss": 1.7526,
      "step": 3667
    },
    {
      "epoch": 0.39641197449475846,
      "grad_norm": 0.3897620439529419,
      "learning_rate": 8e-05,
      "loss": 1.696,
      "step": 3668
    },
    {
      "epoch": 0.39652004755214526,
      "grad_norm": 0.41348329186439514,
      "learning_rate": 8e-05,
      "loss": 1.6779,
      "step": 3669
    },
    {
      "epoch": 0.39662812060953206,
      "grad_norm": 0.41085559129714966,
      "learning_rate": 8e-05,
      "loss": 1.8176,
      "step": 3670
    },
    {
      "epoch": 0.3967361936669188,
      "grad_norm": 0.4275372326374054,
      "learning_rate": 8e-05,
      "loss": 1.7224,
      "step": 3671
    },
    {
      "epoch": 0.3968442667243056,
      "grad_norm": 0.36964356899261475,
      "learning_rate": 8e-05,
      "loss": 1.6745,
      "step": 3672
    },
    {
      "epoch": 0.3969523397816924,
      "grad_norm": 0.4059866666793823,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 3673
    },
    {
      "epoch": 0.3970604128390792,
      "grad_norm": 0.4467698037624359,
      "learning_rate": 8e-05,
      "loss": 1.911,
      "step": 3674
    },
    {
      "epoch": 0.39716848589646603,
      "grad_norm": 0.3793274164199829,
      "learning_rate": 8e-05,
      "loss": 1.6528,
      "step": 3675
    },
    {
      "epoch": 0.39727655895385283,
      "grad_norm": 0.4253424108028412,
      "learning_rate": 8e-05,
      "loss": 1.9186,
      "step": 3676
    },
    {
      "epoch": 0.3973846320112396,
      "grad_norm": 0.3775992691516876,
      "learning_rate": 8e-05,
      "loss": 1.4771,
      "step": 3677
    },
    {
      "epoch": 0.3974927050686264,
      "grad_norm": 0.40357932448387146,
      "learning_rate": 8e-05,
      "loss": 1.6427,
      "step": 3678
    },
    {
      "epoch": 0.3976007781260132,
      "grad_norm": 0.3914742171764374,
      "learning_rate": 8e-05,
      "loss": 1.7277,
      "step": 3679
    },
    {
      "epoch": 0.3977088511834,
      "grad_norm": 0.3934895694255829,
      "learning_rate": 8e-05,
      "loss": 1.8489,
      "step": 3680
    },
    {
      "epoch": 0.3978169242407868,
      "grad_norm": 0.4016740024089813,
      "learning_rate": 8e-05,
      "loss": 1.6955,
      "step": 3681
    },
    {
      "epoch": 0.39792499729817354,
      "grad_norm": 0.40770217776298523,
      "learning_rate": 8e-05,
      "loss": 1.7102,
      "step": 3682
    },
    {
      "epoch": 0.39803307035556035,
      "grad_norm": 0.4512227773666382,
      "learning_rate": 8e-05,
      "loss": 1.7325,
      "step": 3683
    },
    {
      "epoch": 0.39814114341294715,
      "grad_norm": 0.39062368869781494,
      "learning_rate": 8e-05,
      "loss": 1.7427,
      "step": 3684
    },
    {
      "epoch": 0.39824921647033396,
      "grad_norm": 0.49332454800605774,
      "learning_rate": 8e-05,
      "loss": 1.843,
      "step": 3685
    },
    {
      "epoch": 0.39835728952772076,
      "grad_norm": 0.38912585377693176,
      "learning_rate": 8e-05,
      "loss": 1.6231,
      "step": 3686
    },
    {
      "epoch": 0.3984653625851075,
      "grad_norm": 0.37618938088417053,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 3687
    },
    {
      "epoch": 0.3985734356424943,
      "grad_norm": 0.4008297920227051,
      "learning_rate": 8e-05,
      "loss": 1.8939,
      "step": 3688
    },
    {
      "epoch": 0.3986815086998811,
      "grad_norm": 0.3977849781513214,
      "learning_rate": 8e-05,
      "loss": 1.6292,
      "step": 3689
    },
    {
      "epoch": 0.3987895817572679,
      "grad_norm": 0.38854777812957764,
      "learning_rate": 8e-05,
      "loss": 1.6468,
      "step": 3690
    },
    {
      "epoch": 0.3988976548146547,
      "grad_norm": 0.40625908970832825,
      "learning_rate": 8e-05,
      "loss": 1.8091,
      "step": 3691
    },
    {
      "epoch": 0.3990057278720415,
      "grad_norm": 0.35676607489585876,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 3692
    },
    {
      "epoch": 0.3991138009294283,
      "grad_norm": 0.3685443103313446,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 3693
    },
    {
      "epoch": 0.3992218739868151,
      "grad_norm": 0.3897593915462494,
      "learning_rate": 8e-05,
      "loss": 1.8239,
      "step": 3694
    },
    {
      "epoch": 0.3993299470442019,
      "grad_norm": 0.4160090982913971,
      "learning_rate": 8e-05,
      "loss": 1.8628,
      "step": 3695
    },
    {
      "epoch": 0.3994380201015887,
      "grad_norm": 0.4024585485458374,
      "learning_rate": 8e-05,
      "loss": 1.5545,
      "step": 3696
    },
    {
      "epoch": 0.3995460931589755,
      "grad_norm": 0.38592958450317383,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 3697
    },
    {
      "epoch": 0.39965416621636224,
      "grad_norm": 0.38454344868659973,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 3698
    },
    {
      "epoch": 0.39976223927374904,
      "grad_norm": 0.431120902299881,
      "learning_rate": 8e-05,
      "loss": 1.6842,
      "step": 3699
    },
    {
      "epoch": 0.39987031233113585,
      "grad_norm": 0.3986557424068451,
      "learning_rate": 8e-05,
      "loss": 1.8025,
      "step": 3700
    },
    {
      "epoch": 0.39997838538852265,
      "grad_norm": 0.4012722373008728,
      "learning_rate": 8e-05,
      "loss": 1.8269,
      "step": 3701
    },
    {
      "epoch": 0.40008645844590945,
      "grad_norm": 0.3755255937576294,
      "learning_rate": 8e-05,
      "loss": 1.5911,
      "step": 3702
    },
    {
      "epoch": 0.4001945315032962,
      "grad_norm": 0.4482969343662262,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 3703
    },
    {
      "epoch": 0.400302604560683,
      "grad_norm": 0.40132319927215576,
      "learning_rate": 8e-05,
      "loss": 1.717,
      "step": 3704
    },
    {
      "epoch": 0.4004106776180698,
      "grad_norm": 0.376080721616745,
      "learning_rate": 8e-05,
      "loss": 1.5695,
      "step": 3705
    },
    {
      "epoch": 0.4005187506754566,
      "grad_norm": 0.4061757028102875,
      "learning_rate": 8e-05,
      "loss": 1.7167,
      "step": 3706
    },
    {
      "epoch": 0.4006268237328434,
      "grad_norm": 0.41207167506217957,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 3707
    },
    {
      "epoch": 0.4007348967902302,
      "grad_norm": 0.3914540708065033,
      "learning_rate": 8e-05,
      "loss": 1.5986,
      "step": 3708
    },
    {
      "epoch": 0.40084296984761697,
      "grad_norm": 0.40471914410591125,
      "learning_rate": 8e-05,
      "loss": 1.6127,
      "step": 3709
    },
    {
      "epoch": 0.4009510429050038,
      "grad_norm": 0.4019676446914673,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 3710
    },
    {
      "epoch": 0.4010591159623906,
      "grad_norm": 0.38933658599853516,
      "learning_rate": 8e-05,
      "loss": 1.7484,
      "step": 3711
    },
    {
      "epoch": 0.4011671890197774,
      "grad_norm": 0.3860172629356384,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 3712
    },
    {
      "epoch": 0.4012752620771642,
      "grad_norm": 0.3738880455493927,
      "learning_rate": 8e-05,
      "loss": 1.6366,
      "step": 3713
    },
    {
      "epoch": 0.40138333513455093,
      "grad_norm": 0.37177854776382446,
      "learning_rate": 8e-05,
      "loss": 1.6966,
      "step": 3714
    },
    {
      "epoch": 0.40149140819193774,
      "grad_norm": 0.4125951826572418,
      "learning_rate": 8e-05,
      "loss": 1.8008,
      "step": 3715
    },
    {
      "epoch": 0.40159948124932454,
      "grad_norm": 0.40310990810394287,
      "learning_rate": 8e-05,
      "loss": 1.677,
      "step": 3716
    },
    {
      "epoch": 0.40170755430671135,
      "grad_norm": 0.39373481273651123,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 3717
    },
    {
      "epoch": 0.40181562736409815,
      "grad_norm": 0.40079519152641296,
      "learning_rate": 8e-05,
      "loss": 1.7901,
      "step": 3718
    },
    {
      "epoch": 0.4019237004214849,
      "grad_norm": 0.4064926505088806,
      "learning_rate": 8e-05,
      "loss": 1.561,
      "step": 3719
    },
    {
      "epoch": 0.4020317734788717,
      "grad_norm": 0.4149755835533142,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 3720
    },
    {
      "epoch": 0.4021398465362585,
      "grad_norm": 0.3788078725337982,
      "learning_rate": 8e-05,
      "loss": 1.5809,
      "step": 3721
    },
    {
      "epoch": 0.4022479195936453,
      "grad_norm": 0.3746579587459564,
      "learning_rate": 8e-05,
      "loss": 1.5581,
      "step": 3722
    },
    {
      "epoch": 0.4023559926510321,
      "grad_norm": 0.39097830653190613,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 3723
    },
    {
      "epoch": 0.4024640657084189,
      "grad_norm": 0.3641214966773987,
      "learning_rate": 8e-05,
      "loss": 1.6824,
      "step": 3724
    },
    {
      "epoch": 0.40257213876580566,
      "grad_norm": 0.3794381320476532,
      "learning_rate": 8e-05,
      "loss": 1.8231,
      "step": 3725
    },
    {
      "epoch": 0.40268021182319247,
      "grad_norm": 0.3842198848724365,
      "learning_rate": 8e-05,
      "loss": 1.7266,
      "step": 3726
    },
    {
      "epoch": 0.4027882848805793,
      "grad_norm": 0.39962583780288696,
      "learning_rate": 8e-05,
      "loss": 1.8617,
      "step": 3727
    },
    {
      "epoch": 0.4028963579379661,
      "grad_norm": 0.40599000453948975,
      "learning_rate": 8e-05,
      "loss": 1.6513,
      "step": 3728
    },
    {
      "epoch": 0.4030044309953529,
      "grad_norm": 0.4413057863712311,
      "learning_rate": 8e-05,
      "loss": 1.8482,
      "step": 3729
    },
    {
      "epoch": 0.40311250405273963,
      "grad_norm": 0.377800315618515,
      "learning_rate": 8e-05,
      "loss": 1.6702,
      "step": 3730
    },
    {
      "epoch": 0.40322057711012643,
      "grad_norm": 0.39248013496398926,
      "learning_rate": 8e-05,
      "loss": 1.6857,
      "step": 3731
    },
    {
      "epoch": 0.40332865016751324,
      "grad_norm": 0.38526687026023865,
      "learning_rate": 8e-05,
      "loss": 1.5928,
      "step": 3732
    },
    {
      "epoch": 0.40343672322490004,
      "grad_norm": 0.44204992055892944,
      "learning_rate": 8e-05,
      "loss": 1.9811,
      "step": 3733
    },
    {
      "epoch": 0.40354479628228684,
      "grad_norm": 0.43923699855804443,
      "learning_rate": 8e-05,
      "loss": 1.8502,
      "step": 3734
    },
    {
      "epoch": 0.4036528693396736,
      "grad_norm": 0.3984847068786621,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 3735
    },
    {
      "epoch": 0.4037609423970604,
      "grad_norm": 0.4320618212223053,
      "learning_rate": 8e-05,
      "loss": 1.9047,
      "step": 3736
    },
    {
      "epoch": 0.4038690154544472,
      "grad_norm": 0.36007046699523926,
      "learning_rate": 8e-05,
      "loss": 1.5466,
      "step": 3737
    },
    {
      "epoch": 0.403977088511834,
      "grad_norm": 0.38590195775032043,
      "learning_rate": 8e-05,
      "loss": 1.8055,
      "step": 3738
    },
    {
      "epoch": 0.4040851615692208,
      "grad_norm": 0.39130690693855286,
      "learning_rate": 8e-05,
      "loss": 1.6736,
      "step": 3739
    },
    {
      "epoch": 0.4041932346266076,
      "grad_norm": 0.4166002869606018,
      "learning_rate": 8e-05,
      "loss": 1.9243,
      "step": 3740
    },
    {
      "epoch": 0.40430130768399436,
      "grad_norm": 0.376675009727478,
      "learning_rate": 8e-05,
      "loss": 1.5765,
      "step": 3741
    },
    {
      "epoch": 0.40440938074138116,
      "grad_norm": 0.3769833743572235,
      "learning_rate": 8e-05,
      "loss": 1.6408,
      "step": 3742
    },
    {
      "epoch": 0.40451745379876797,
      "grad_norm": 0.384000301361084,
      "learning_rate": 8e-05,
      "loss": 1.7887,
      "step": 3743
    },
    {
      "epoch": 0.40462552685615477,
      "grad_norm": 0.3945089280605316,
      "learning_rate": 8e-05,
      "loss": 1.7568,
      "step": 3744
    },
    {
      "epoch": 0.4047335999135416,
      "grad_norm": 0.393502801656723,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 3745
    },
    {
      "epoch": 0.4048416729709283,
      "grad_norm": 0.39680758118629456,
      "learning_rate": 8e-05,
      "loss": 1.7304,
      "step": 3746
    },
    {
      "epoch": 0.4049497460283151,
      "grad_norm": 0.38890594244003296,
      "learning_rate": 8e-05,
      "loss": 1.5778,
      "step": 3747
    },
    {
      "epoch": 0.40505781908570193,
      "grad_norm": 0.35278794169425964,
      "learning_rate": 8e-05,
      "loss": 1.6004,
      "step": 3748
    },
    {
      "epoch": 0.40516589214308874,
      "grad_norm": 0.3913782835006714,
      "learning_rate": 8e-05,
      "loss": 1.63,
      "step": 3749
    },
    {
      "epoch": 0.40527396520047554,
      "grad_norm": 0.37278279662132263,
      "learning_rate": 8e-05,
      "loss": 1.7734,
      "step": 3750
    },
    {
      "epoch": 0.4053820382578623,
      "grad_norm": 0.39850932359695435,
      "learning_rate": 8e-05,
      "loss": 1.4858,
      "step": 3751
    },
    {
      "epoch": 0.4054901113152491,
      "grad_norm": 0.3929099440574646,
      "learning_rate": 8e-05,
      "loss": 1.7001,
      "step": 3752
    },
    {
      "epoch": 0.4055981843726359,
      "grad_norm": 0.42637261748313904,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 3753
    },
    {
      "epoch": 0.4057062574300227,
      "grad_norm": 0.3746306300163269,
      "learning_rate": 8e-05,
      "loss": 1.626,
      "step": 3754
    },
    {
      "epoch": 0.4058143304874095,
      "grad_norm": 0.38104248046875,
      "learning_rate": 8e-05,
      "loss": 1.6227,
      "step": 3755
    },
    {
      "epoch": 0.4059224035447963,
      "grad_norm": 0.37350109219551086,
      "learning_rate": 8e-05,
      "loss": 1.6684,
      "step": 3756
    },
    {
      "epoch": 0.40603047660218305,
      "grad_norm": 0.39378634095191956,
      "learning_rate": 8e-05,
      "loss": 1.7541,
      "step": 3757
    },
    {
      "epoch": 0.40613854965956986,
      "grad_norm": 0.3619995415210724,
      "learning_rate": 8e-05,
      "loss": 1.5679,
      "step": 3758
    },
    {
      "epoch": 0.40624662271695666,
      "grad_norm": 0.41921520233154297,
      "learning_rate": 8e-05,
      "loss": 1.7098,
      "step": 3759
    },
    {
      "epoch": 0.40635469577434347,
      "grad_norm": 0.3966827392578125,
      "learning_rate": 8e-05,
      "loss": 1.757,
      "step": 3760
    },
    {
      "epoch": 0.40646276883173027,
      "grad_norm": 0.41021886467933655,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 3761
    },
    {
      "epoch": 0.406570841889117,
      "grad_norm": 0.40492698550224304,
      "learning_rate": 8e-05,
      "loss": 1.7581,
      "step": 3762
    },
    {
      "epoch": 0.4066789149465038,
      "grad_norm": 0.427545964717865,
      "learning_rate": 8e-05,
      "loss": 1.6431,
      "step": 3763
    },
    {
      "epoch": 0.4067869880038906,
      "grad_norm": 0.42807745933532715,
      "learning_rate": 8e-05,
      "loss": 1.7861,
      "step": 3764
    },
    {
      "epoch": 0.40689506106127743,
      "grad_norm": 0.39299434423446655,
      "learning_rate": 8e-05,
      "loss": 1.8002,
      "step": 3765
    },
    {
      "epoch": 0.40700313411866423,
      "grad_norm": 0.3910335600376129,
      "learning_rate": 8e-05,
      "loss": 1.5639,
      "step": 3766
    },
    {
      "epoch": 0.40711120717605104,
      "grad_norm": 0.4081684350967407,
      "learning_rate": 8e-05,
      "loss": 1.6153,
      "step": 3767
    },
    {
      "epoch": 0.4072192802334378,
      "grad_norm": 0.3840066194534302,
      "learning_rate": 8e-05,
      "loss": 1.6344,
      "step": 3768
    },
    {
      "epoch": 0.4073273532908246,
      "grad_norm": 0.3912813067436218,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 3769
    },
    {
      "epoch": 0.4074354263482114,
      "grad_norm": 0.3840087652206421,
      "learning_rate": 8e-05,
      "loss": 1.7064,
      "step": 3770
    },
    {
      "epoch": 0.4075434994055982,
      "grad_norm": 0.38698792457580566,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 3771
    },
    {
      "epoch": 0.407651572462985,
      "grad_norm": 0.3900686204433441,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 3772
    },
    {
      "epoch": 0.40775964552037175,
      "grad_norm": 0.3696495592594147,
      "learning_rate": 8e-05,
      "loss": 1.6111,
      "step": 3773
    },
    {
      "epoch": 0.40786771857775855,
      "grad_norm": 0.3842237889766693,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 3774
    },
    {
      "epoch": 0.40797579163514536,
      "grad_norm": 0.38225430250167847,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 3775
    },
    {
      "epoch": 0.40808386469253216,
      "grad_norm": 0.3904886245727539,
      "learning_rate": 8e-05,
      "loss": 1.8409,
      "step": 3776
    },
    {
      "epoch": 0.40819193774991896,
      "grad_norm": 0.3797644376754761,
      "learning_rate": 8e-05,
      "loss": 1.7335,
      "step": 3777
    },
    {
      "epoch": 0.4083000108073057,
      "grad_norm": 0.39331331849098206,
      "learning_rate": 8e-05,
      "loss": 1.6978,
      "step": 3778
    },
    {
      "epoch": 0.4084080838646925,
      "grad_norm": 0.41202792525291443,
      "learning_rate": 8e-05,
      "loss": 1.8553,
      "step": 3779
    },
    {
      "epoch": 0.4085161569220793,
      "grad_norm": 0.39176079630851746,
      "learning_rate": 8e-05,
      "loss": 1.7069,
      "step": 3780
    },
    {
      "epoch": 0.4086242299794661,
      "grad_norm": 0.41179949045181274,
      "learning_rate": 8e-05,
      "loss": 1.8179,
      "step": 3781
    },
    {
      "epoch": 0.40873230303685293,
      "grad_norm": 0.3702789545059204,
      "learning_rate": 8e-05,
      "loss": 1.5119,
      "step": 3782
    },
    {
      "epoch": 0.40884037609423973,
      "grad_norm": 0.38579022884368896,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 3783
    },
    {
      "epoch": 0.4089484491516265,
      "grad_norm": 0.44669055938720703,
      "learning_rate": 8e-05,
      "loss": 1.9425,
      "step": 3784
    },
    {
      "epoch": 0.4090565222090133,
      "grad_norm": 0.41179758310317993,
      "learning_rate": 8e-05,
      "loss": 1.7262,
      "step": 3785
    },
    {
      "epoch": 0.4091645952664001,
      "grad_norm": 0.4036073684692383,
      "learning_rate": 8e-05,
      "loss": 1.7531,
      "step": 3786
    },
    {
      "epoch": 0.4092726683237869,
      "grad_norm": 0.41674765944480896,
      "learning_rate": 8e-05,
      "loss": 1.8272,
      "step": 3787
    },
    {
      "epoch": 0.4093807413811737,
      "grad_norm": 0.42446672916412354,
      "learning_rate": 8e-05,
      "loss": 1.8081,
      "step": 3788
    },
    {
      "epoch": 0.40948881443856044,
      "grad_norm": 0.395670086145401,
      "learning_rate": 8e-05,
      "loss": 1.72,
      "step": 3789
    },
    {
      "epoch": 0.40959688749594725,
      "grad_norm": 0.40348920226097107,
      "learning_rate": 8e-05,
      "loss": 1.7283,
      "step": 3790
    },
    {
      "epoch": 0.40970496055333405,
      "grad_norm": 0.3769279420375824,
      "learning_rate": 8e-05,
      "loss": 1.7197,
      "step": 3791
    },
    {
      "epoch": 0.40981303361072086,
      "grad_norm": 0.381083607673645,
      "learning_rate": 8e-05,
      "loss": 1.8504,
      "step": 3792
    },
    {
      "epoch": 0.40992110666810766,
      "grad_norm": 0.39807000756263733,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 3793
    },
    {
      "epoch": 0.4100291797254944,
      "grad_norm": 0.4081137776374817,
      "learning_rate": 8e-05,
      "loss": 1.7205,
      "step": 3794
    },
    {
      "epoch": 0.4101372527828812,
      "grad_norm": 0.43015384674072266,
      "learning_rate": 8e-05,
      "loss": 1.7789,
      "step": 3795
    },
    {
      "epoch": 0.410245325840268,
      "grad_norm": 0.4252828061580658,
      "learning_rate": 8e-05,
      "loss": 1.7749,
      "step": 3796
    },
    {
      "epoch": 0.4103533988976548,
      "grad_norm": 0.4680314064025879,
      "learning_rate": 8e-05,
      "loss": 1.7288,
      "step": 3797
    },
    {
      "epoch": 0.4104614719550416,
      "grad_norm": 0.4137987494468689,
      "learning_rate": 8e-05,
      "loss": 1.7295,
      "step": 3798
    },
    {
      "epoch": 0.4105695450124284,
      "grad_norm": 0.40972837805747986,
      "learning_rate": 8e-05,
      "loss": 1.9828,
      "step": 3799
    },
    {
      "epoch": 0.4106776180698152,
      "grad_norm": 0.4470110833644867,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 3800
    },
    {
      "epoch": 0.410785691127202,
      "grad_norm": 0.40168553590774536,
      "learning_rate": 8e-05,
      "loss": 1.7093,
      "step": 3801
    },
    {
      "epoch": 0.4108937641845888,
      "grad_norm": 0.3665669560432434,
      "learning_rate": 8e-05,
      "loss": 1.6479,
      "step": 3802
    },
    {
      "epoch": 0.4110018372419756,
      "grad_norm": 0.37971702218055725,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 3803
    },
    {
      "epoch": 0.4111099102993624,
      "grad_norm": 0.36938971281051636,
      "learning_rate": 8e-05,
      "loss": 1.6143,
      "step": 3804
    },
    {
      "epoch": 0.41121798335674914,
      "grad_norm": 0.40479499101638794,
      "learning_rate": 8e-05,
      "loss": 1.7483,
      "step": 3805
    },
    {
      "epoch": 0.41132605641413594,
      "grad_norm": 0.3785454332828522,
      "learning_rate": 8e-05,
      "loss": 1.6935,
      "step": 3806
    },
    {
      "epoch": 0.41143412947152275,
      "grad_norm": 0.42723578214645386,
      "learning_rate": 8e-05,
      "loss": 1.8046,
      "step": 3807
    },
    {
      "epoch": 0.41154220252890955,
      "grad_norm": 0.37585800886154175,
      "learning_rate": 8e-05,
      "loss": 1.645,
      "step": 3808
    },
    {
      "epoch": 0.41165027558629635,
      "grad_norm": 0.43754732608795166,
      "learning_rate": 8e-05,
      "loss": 1.9067,
      "step": 3809
    },
    {
      "epoch": 0.4117583486436831,
      "grad_norm": 0.36948803067207336,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 3810
    },
    {
      "epoch": 0.4118664217010699,
      "grad_norm": 0.37926816940307617,
      "learning_rate": 8e-05,
      "loss": 1.6941,
      "step": 3811
    },
    {
      "epoch": 0.4119744947584567,
      "grad_norm": 0.3859090805053711,
      "learning_rate": 8e-05,
      "loss": 1.761,
      "step": 3812
    },
    {
      "epoch": 0.4120825678158435,
      "grad_norm": 0.4168667793273926,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 3813
    },
    {
      "epoch": 0.4121906408732303,
      "grad_norm": 0.3852517008781433,
      "learning_rate": 8e-05,
      "loss": 1.8059,
      "step": 3814
    },
    {
      "epoch": 0.4122987139306171,
      "grad_norm": 0.3861292004585266,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 3815
    },
    {
      "epoch": 0.41240678698800387,
      "grad_norm": 0.40395817160606384,
      "learning_rate": 8e-05,
      "loss": 1.6503,
      "step": 3816
    },
    {
      "epoch": 0.4125148600453907,
      "grad_norm": 0.38121846318244934,
      "learning_rate": 8e-05,
      "loss": 1.5806,
      "step": 3817
    },
    {
      "epoch": 0.4126229331027775,
      "grad_norm": 0.4147404134273529,
      "learning_rate": 8e-05,
      "loss": 1.7027,
      "step": 3818
    },
    {
      "epoch": 0.4127310061601643,
      "grad_norm": 0.3901894986629486,
      "learning_rate": 8e-05,
      "loss": 1.5535,
      "step": 3819
    },
    {
      "epoch": 0.4128390792175511,
      "grad_norm": 0.4096665382385254,
      "learning_rate": 8e-05,
      "loss": 1.8196,
      "step": 3820
    },
    {
      "epoch": 0.41294715227493783,
      "grad_norm": 0.3848898708820343,
      "learning_rate": 8e-05,
      "loss": 1.5894,
      "step": 3821
    },
    {
      "epoch": 0.41305522533232464,
      "grad_norm": 0.4184325635433197,
      "learning_rate": 8e-05,
      "loss": 1.6914,
      "step": 3822
    },
    {
      "epoch": 0.41316329838971144,
      "grad_norm": 0.4055677354335785,
      "learning_rate": 8e-05,
      "loss": 1.6885,
      "step": 3823
    },
    {
      "epoch": 0.41327137144709825,
      "grad_norm": 0.42235785722732544,
      "learning_rate": 8e-05,
      "loss": 1.7978,
      "step": 3824
    },
    {
      "epoch": 0.41337944450448505,
      "grad_norm": 0.41004014015197754,
      "learning_rate": 8e-05,
      "loss": 1.8054,
      "step": 3825
    },
    {
      "epoch": 0.4134875175618718,
      "grad_norm": 0.4179190397262573,
      "learning_rate": 8e-05,
      "loss": 1.7072,
      "step": 3826
    },
    {
      "epoch": 0.4135955906192586,
      "grad_norm": 0.3941066265106201,
      "learning_rate": 8e-05,
      "loss": 1.4534,
      "step": 3827
    },
    {
      "epoch": 0.4137036636766454,
      "grad_norm": 0.4446079730987549,
      "learning_rate": 8e-05,
      "loss": 1.5991,
      "step": 3828
    },
    {
      "epoch": 0.4138117367340322,
      "grad_norm": 0.39789435267448425,
      "learning_rate": 8e-05,
      "loss": 1.3593,
      "step": 3829
    },
    {
      "epoch": 0.413919809791419,
      "grad_norm": 0.4304432272911072,
      "learning_rate": 8e-05,
      "loss": 1.9247,
      "step": 3830
    },
    {
      "epoch": 0.4140278828488058,
      "grad_norm": 0.38246092200279236,
      "learning_rate": 8e-05,
      "loss": 1.6697,
      "step": 3831
    },
    {
      "epoch": 0.41413595590619257,
      "grad_norm": 0.4152018129825592,
      "learning_rate": 8e-05,
      "loss": 1.7623,
      "step": 3832
    },
    {
      "epoch": 0.41424402896357937,
      "grad_norm": 0.3686734139919281,
      "learning_rate": 8e-05,
      "loss": 1.5994,
      "step": 3833
    },
    {
      "epoch": 0.4143521020209662,
      "grad_norm": 0.3893820345401764,
      "learning_rate": 8e-05,
      "loss": 1.7034,
      "step": 3834
    },
    {
      "epoch": 0.414460175078353,
      "grad_norm": 0.4289054572582245,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 3835
    },
    {
      "epoch": 0.4145682481357398,
      "grad_norm": 0.4365761876106262,
      "learning_rate": 8e-05,
      "loss": 1.7199,
      "step": 3836
    },
    {
      "epoch": 0.41467632119312653,
      "grad_norm": 0.40079325437545776,
      "learning_rate": 8e-05,
      "loss": 1.7511,
      "step": 3837
    },
    {
      "epoch": 0.41478439425051333,
      "grad_norm": 0.43482568860054016,
      "learning_rate": 8e-05,
      "loss": 1.8854,
      "step": 3838
    },
    {
      "epoch": 0.41489246730790014,
      "grad_norm": 0.3957137167453766,
      "learning_rate": 8e-05,
      "loss": 1.7392,
      "step": 3839
    },
    {
      "epoch": 0.41500054036528694,
      "grad_norm": 0.3891526460647583,
      "learning_rate": 8e-05,
      "loss": 1.8902,
      "step": 3840
    },
    {
      "epoch": 0.41510861342267374,
      "grad_norm": 0.36817070841789246,
      "learning_rate": 8e-05,
      "loss": 1.6249,
      "step": 3841
    },
    {
      "epoch": 0.41521668648006055,
      "grad_norm": 0.4674880802631378,
      "learning_rate": 8e-05,
      "loss": 2.1006,
      "step": 3842
    },
    {
      "epoch": 0.4153247595374473,
      "grad_norm": 0.3756575286388397,
      "learning_rate": 8e-05,
      "loss": 1.5585,
      "step": 3843
    },
    {
      "epoch": 0.4154328325948341,
      "grad_norm": 0.42478275299072266,
      "learning_rate": 8e-05,
      "loss": 1.8408,
      "step": 3844
    },
    {
      "epoch": 0.4155409056522209,
      "grad_norm": 0.42642882466316223,
      "learning_rate": 8e-05,
      "loss": 1.7534,
      "step": 3845
    },
    {
      "epoch": 0.4156489787096077,
      "grad_norm": 0.4265595078468323,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 3846
    },
    {
      "epoch": 0.4157570517669945,
      "grad_norm": 0.4081973135471344,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 3847
    },
    {
      "epoch": 0.41586512482438126,
      "grad_norm": 0.38063693046569824,
      "learning_rate": 8e-05,
      "loss": 1.7751,
      "step": 3848
    },
    {
      "epoch": 0.41597319788176806,
      "grad_norm": 0.4516310691833496,
      "learning_rate": 8e-05,
      "loss": 2.0244,
      "step": 3849
    },
    {
      "epoch": 0.41608127093915487,
      "grad_norm": 0.4231090843677521,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 3850
    },
    {
      "epoch": 0.41618934399654167,
      "grad_norm": 0.36689531803131104,
      "learning_rate": 8e-05,
      "loss": 1.616,
      "step": 3851
    },
    {
      "epoch": 0.4162974170539285,
      "grad_norm": 0.42786717414855957,
      "learning_rate": 8e-05,
      "loss": 1.8138,
      "step": 3852
    },
    {
      "epoch": 0.4164054901113152,
      "grad_norm": 0.34821581840515137,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 3853
    },
    {
      "epoch": 0.41651356316870203,
      "grad_norm": 0.38966771960258484,
      "learning_rate": 8e-05,
      "loss": 1.8323,
      "step": 3854
    },
    {
      "epoch": 0.41662163622608883,
      "grad_norm": 0.43030041456222534,
      "learning_rate": 8e-05,
      "loss": 1.6704,
      "step": 3855
    },
    {
      "epoch": 0.41672970928347564,
      "grad_norm": 0.4543220102787018,
      "learning_rate": 8e-05,
      "loss": 1.9392,
      "step": 3856
    },
    {
      "epoch": 0.41683778234086244,
      "grad_norm": 0.3875998556613922,
      "learning_rate": 8e-05,
      "loss": 1.5342,
      "step": 3857
    },
    {
      "epoch": 0.41694585539824924,
      "grad_norm": 0.40602919459342957,
      "learning_rate": 8e-05,
      "loss": 1.748,
      "step": 3858
    },
    {
      "epoch": 0.417053928455636,
      "grad_norm": 0.4219368100166321,
      "learning_rate": 8e-05,
      "loss": 1.8748,
      "step": 3859
    },
    {
      "epoch": 0.4171620015130228,
      "grad_norm": 0.41793370246887207,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 3860
    },
    {
      "epoch": 0.4172700745704096,
      "grad_norm": 0.4360904395580292,
      "learning_rate": 8e-05,
      "loss": 1.9035,
      "step": 3861
    },
    {
      "epoch": 0.4173781476277964,
      "grad_norm": 0.41942670941352844,
      "learning_rate": 8e-05,
      "loss": 1.9133,
      "step": 3862
    },
    {
      "epoch": 0.4174862206851832,
      "grad_norm": 0.40040525794029236,
      "learning_rate": 8e-05,
      "loss": 1.6959,
      "step": 3863
    },
    {
      "epoch": 0.41759429374256996,
      "grad_norm": 0.3819044530391693,
      "learning_rate": 8e-05,
      "loss": 1.7643,
      "step": 3864
    },
    {
      "epoch": 0.41770236679995676,
      "grad_norm": 0.42044445872306824,
      "learning_rate": 8e-05,
      "loss": 1.8157,
      "step": 3865
    },
    {
      "epoch": 0.41781043985734356,
      "grad_norm": 0.36946722865104675,
      "learning_rate": 8e-05,
      "loss": 1.6438,
      "step": 3866
    },
    {
      "epoch": 0.41791851291473037,
      "grad_norm": 0.4200522005558014,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 3867
    },
    {
      "epoch": 0.41802658597211717,
      "grad_norm": 0.3917948603630066,
      "learning_rate": 8e-05,
      "loss": 1.7759,
      "step": 3868
    },
    {
      "epoch": 0.4181346590295039,
      "grad_norm": 0.4072740077972412,
      "learning_rate": 8e-05,
      "loss": 1.7675,
      "step": 3869
    },
    {
      "epoch": 0.4182427320868907,
      "grad_norm": 0.3877619504928589,
      "learning_rate": 8e-05,
      "loss": 1.6873,
      "step": 3870
    },
    {
      "epoch": 0.4183508051442775,
      "grad_norm": 0.4184662699699402,
      "learning_rate": 8e-05,
      "loss": 1.8491,
      "step": 3871
    },
    {
      "epoch": 0.41845887820166433,
      "grad_norm": 0.3985510468482971,
      "learning_rate": 8e-05,
      "loss": 1.7071,
      "step": 3872
    },
    {
      "epoch": 0.41856695125905113,
      "grad_norm": 0.39273327589035034,
      "learning_rate": 8e-05,
      "loss": 1.7745,
      "step": 3873
    },
    {
      "epoch": 0.41867502431643794,
      "grad_norm": 0.3655044138431549,
      "learning_rate": 8e-05,
      "loss": 1.5411,
      "step": 3874
    },
    {
      "epoch": 0.4187830973738247,
      "grad_norm": 0.3853136897087097,
      "learning_rate": 8e-05,
      "loss": 1.7062,
      "step": 3875
    },
    {
      "epoch": 0.4188911704312115,
      "grad_norm": 0.43215006589889526,
      "learning_rate": 8e-05,
      "loss": 1.8042,
      "step": 3876
    },
    {
      "epoch": 0.4189992434885983,
      "grad_norm": 0.382182776927948,
      "learning_rate": 8e-05,
      "loss": 1.7719,
      "step": 3877
    },
    {
      "epoch": 0.4191073165459851,
      "grad_norm": 0.4090704619884491,
      "learning_rate": 8e-05,
      "loss": 1.4606,
      "step": 3878
    },
    {
      "epoch": 0.4192153896033719,
      "grad_norm": 0.37792569398880005,
      "learning_rate": 8e-05,
      "loss": 1.8337,
      "step": 3879
    },
    {
      "epoch": 0.41932346266075865,
      "grad_norm": 0.3886340856552124,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 3880
    },
    {
      "epoch": 0.41943153571814545,
      "grad_norm": 0.39856937527656555,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 3881
    },
    {
      "epoch": 0.41953960877553226,
      "grad_norm": 0.42731624841690063,
      "learning_rate": 8e-05,
      "loss": 1.7501,
      "step": 3882
    },
    {
      "epoch": 0.41964768183291906,
      "grad_norm": 0.4575856328010559,
      "learning_rate": 8e-05,
      "loss": 1.5855,
      "step": 3883
    },
    {
      "epoch": 0.41975575489030587,
      "grad_norm": 0.3995247781276703,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 3884
    },
    {
      "epoch": 0.4198638279476926,
      "grad_norm": 0.4094035029411316,
      "learning_rate": 8e-05,
      "loss": 1.8229,
      "step": 3885
    },
    {
      "epoch": 0.4199719010050794,
      "grad_norm": 0.3868167996406555,
      "learning_rate": 8e-05,
      "loss": 1.4925,
      "step": 3886
    },
    {
      "epoch": 0.4200799740624662,
      "grad_norm": 0.3684050738811493,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 3887
    },
    {
      "epoch": 0.420188047119853,
      "grad_norm": 0.48412683606147766,
      "learning_rate": 8e-05,
      "loss": 1.9776,
      "step": 3888
    },
    {
      "epoch": 0.42029612017723983,
      "grad_norm": 0.3780952990055084,
      "learning_rate": 8e-05,
      "loss": 1.6968,
      "step": 3889
    },
    {
      "epoch": 0.42040419323462663,
      "grad_norm": 0.363110214471817,
      "learning_rate": 8e-05,
      "loss": 1.7397,
      "step": 3890
    },
    {
      "epoch": 0.4205122662920134,
      "grad_norm": 0.43607765436172485,
      "learning_rate": 8e-05,
      "loss": 1.7425,
      "step": 3891
    },
    {
      "epoch": 0.4206203393494002,
      "grad_norm": 0.3985387980937958,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 3892
    },
    {
      "epoch": 0.420728412406787,
      "grad_norm": 0.40517929196357727,
      "learning_rate": 8e-05,
      "loss": 1.9024,
      "step": 3893
    },
    {
      "epoch": 0.4208364854641738,
      "grad_norm": 0.3849782943725586,
      "learning_rate": 8e-05,
      "loss": 1.6044,
      "step": 3894
    },
    {
      "epoch": 0.4209445585215606,
      "grad_norm": 0.38462749123573303,
      "learning_rate": 8e-05,
      "loss": 1.7033,
      "step": 3895
    },
    {
      "epoch": 0.42105263157894735,
      "grad_norm": 0.38256916403770447,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 3896
    },
    {
      "epoch": 0.42116070463633415,
      "grad_norm": 0.3805573284626007,
      "learning_rate": 8e-05,
      "loss": 1.6983,
      "step": 3897
    },
    {
      "epoch": 0.42126877769372095,
      "grad_norm": 0.4165429472923279,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 3898
    },
    {
      "epoch": 0.42137685075110776,
      "grad_norm": 0.4750782549381256,
      "learning_rate": 8e-05,
      "loss": 1.8399,
      "step": 3899
    },
    {
      "epoch": 0.42148492380849456,
      "grad_norm": 0.4078114330768585,
      "learning_rate": 8e-05,
      "loss": 1.7572,
      "step": 3900
    },
    {
      "epoch": 0.4215929968658813,
      "grad_norm": 0.3825790584087372,
      "learning_rate": 8e-05,
      "loss": 1.547,
      "step": 3901
    },
    {
      "epoch": 0.4217010699232681,
      "grad_norm": 0.42685189843177795,
      "learning_rate": 8e-05,
      "loss": 1.7911,
      "step": 3902
    },
    {
      "epoch": 0.4218091429806549,
      "grad_norm": 0.3979296088218689,
      "learning_rate": 8e-05,
      "loss": 1.6465,
      "step": 3903
    },
    {
      "epoch": 0.4219172160380417,
      "grad_norm": 0.38864952325820923,
      "learning_rate": 8e-05,
      "loss": 1.7216,
      "step": 3904
    },
    {
      "epoch": 0.4220252890954285,
      "grad_norm": 0.4438540041446686,
      "learning_rate": 8e-05,
      "loss": 1.8995,
      "step": 3905
    },
    {
      "epoch": 0.42213336215281533,
      "grad_norm": 0.41845783591270447,
      "learning_rate": 8e-05,
      "loss": 1.7492,
      "step": 3906
    },
    {
      "epoch": 0.4222414352102021,
      "grad_norm": 0.41742652654647827,
      "learning_rate": 8e-05,
      "loss": 1.6932,
      "step": 3907
    },
    {
      "epoch": 0.4223495082675889,
      "grad_norm": 0.4258211851119995,
      "learning_rate": 8e-05,
      "loss": 1.9198,
      "step": 3908
    },
    {
      "epoch": 0.4224575813249757,
      "grad_norm": 0.37889838218688965,
      "learning_rate": 8e-05,
      "loss": 1.4833,
      "step": 3909
    },
    {
      "epoch": 0.4225656543823625,
      "grad_norm": 0.39772534370422363,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 3910
    },
    {
      "epoch": 0.4226737274397493,
      "grad_norm": 0.37279587984085083,
      "learning_rate": 8e-05,
      "loss": 1.7302,
      "step": 3911
    },
    {
      "epoch": 0.42278180049713604,
      "grad_norm": 0.4219800531864166,
      "learning_rate": 8e-05,
      "loss": 1.8416,
      "step": 3912
    },
    {
      "epoch": 0.42288987355452284,
      "grad_norm": 0.4103454053401947,
      "learning_rate": 8e-05,
      "loss": 1.7689,
      "step": 3913
    },
    {
      "epoch": 0.42299794661190965,
      "grad_norm": 0.4265066385269165,
      "learning_rate": 8e-05,
      "loss": 1.8538,
      "step": 3914
    },
    {
      "epoch": 0.42310601966929645,
      "grad_norm": 0.37320035696029663,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 3915
    },
    {
      "epoch": 0.42321409272668326,
      "grad_norm": 0.3801858425140381,
      "learning_rate": 8e-05,
      "loss": 1.6682,
      "step": 3916
    },
    {
      "epoch": 0.42332216578407,
      "grad_norm": 0.4436030089855194,
      "learning_rate": 8e-05,
      "loss": 1.9305,
      "step": 3917
    },
    {
      "epoch": 0.4234302388414568,
      "grad_norm": 0.37914249300956726,
      "learning_rate": 8e-05,
      "loss": 1.6613,
      "step": 3918
    },
    {
      "epoch": 0.4235383118988436,
      "grad_norm": 0.3867349922657013,
      "learning_rate": 8e-05,
      "loss": 1.6651,
      "step": 3919
    },
    {
      "epoch": 0.4236463849562304,
      "grad_norm": 0.416350394487381,
      "learning_rate": 8e-05,
      "loss": 1.8336,
      "step": 3920
    },
    {
      "epoch": 0.4237544580136172,
      "grad_norm": 0.39160364866256714,
      "learning_rate": 8e-05,
      "loss": 1.6501,
      "step": 3921
    },
    {
      "epoch": 0.423862531071004,
      "grad_norm": 0.44497206807136536,
      "learning_rate": 8e-05,
      "loss": 1.789,
      "step": 3922
    },
    {
      "epoch": 0.42397060412839077,
      "grad_norm": 0.395686537027359,
      "learning_rate": 8e-05,
      "loss": 1.6371,
      "step": 3923
    },
    {
      "epoch": 0.4240786771857776,
      "grad_norm": 0.3818310499191284,
      "learning_rate": 8e-05,
      "loss": 1.591,
      "step": 3924
    },
    {
      "epoch": 0.4241867502431644,
      "grad_norm": 0.3982886075973511,
      "learning_rate": 8e-05,
      "loss": 1.8616,
      "step": 3925
    },
    {
      "epoch": 0.4242948233005512,
      "grad_norm": 0.3833780884742737,
      "learning_rate": 8e-05,
      "loss": 1.737,
      "step": 3926
    },
    {
      "epoch": 0.424402896357938,
      "grad_norm": 0.39365801215171814,
      "learning_rate": 8e-05,
      "loss": 1.777,
      "step": 3927
    },
    {
      "epoch": 0.42451096941532473,
      "grad_norm": 0.4016670286655426,
      "learning_rate": 8e-05,
      "loss": 1.7574,
      "step": 3928
    },
    {
      "epoch": 0.42461904247271154,
      "grad_norm": 0.3761076033115387,
      "learning_rate": 8e-05,
      "loss": 1.732,
      "step": 3929
    },
    {
      "epoch": 0.42472711553009834,
      "grad_norm": 0.3949795365333557,
      "learning_rate": 8e-05,
      "loss": 1.659,
      "step": 3930
    },
    {
      "epoch": 0.42483518858748515,
      "grad_norm": 0.4117641746997833,
      "learning_rate": 8e-05,
      "loss": 1.8549,
      "step": 3931
    },
    {
      "epoch": 0.42494326164487195,
      "grad_norm": 0.44422808289527893,
      "learning_rate": 8e-05,
      "loss": 1.8429,
      "step": 3932
    },
    {
      "epoch": 0.42505133470225875,
      "grad_norm": 0.38675034046173096,
      "learning_rate": 8e-05,
      "loss": 1.5983,
      "step": 3933
    },
    {
      "epoch": 0.4251594077596455,
      "grad_norm": 0.46386539936065674,
      "learning_rate": 8e-05,
      "loss": 1.7914,
      "step": 3934
    },
    {
      "epoch": 0.4252674808170323,
      "grad_norm": 0.38414624333381653,
      "learning_rate": 8e-05,
      "loss": 1.6158,
      "step": 3935
    },
    {
      "epoch": 0.4253755538744191,
      "grad_norm": 0.42299190163612366,
      "learning_rate": 8e-05,
      "loss": 1.7692,
      "step": 3936
    },
    {
      "epoch": 0.4254836269318059,
      "grad_norm": 0.39568471908569336,
      "learning_rate": 8e-05,
      "loss": 1.6804,
      "step": 3937
    },
    {
      "epoch": 0.4255916999891927,
      "grad_norm": 0.4121760129928589,
      "learning_rate": 8e-05,
      "loss": 1.7699,
      "step": 3938
    },
    {
      "epoch": 0.42569977304657947,
      "grad_norm": 0.4374620020389557,
      "learning_rate": 8e-05,
      "loss": 1.5819,
      "step": 3939
    },
    {
      "epoch": 0.42580784610396627,
      "grad_norm": 0.4322124421596527,
      "learning_rate": 8e-05,
      "loss": 1.8521,
      "step": 3940
    },
    {
      "epoch": 0.4259159191613531,
      "grad_norm": 0.38904377818107605,
      "learning_rate": 8e-05,
      "loss": 1.7714,
      "step": 3941
    },
    {
      "epoch": 0.4260239922187399,
      "grad_norm": 0.35896697640419006,
      "learning_rate": 8e-05,
      "loss": 1.576,
      "step": 3942
    },
    {
      "epoch": 0.4261320652761267,
      "grad_norm": 0.41874027252197266,
      "learning_rate": 8e-05,
      "loss": 1.7153,
      "step": 3943
    },
    {
      "epoch": 0.42624013833351343,
      "grad_norm": 0.3961946964263916,
      "learning_rate": 8e-05,
      "loss": 1.794,
      "step": 3944
    },
    {
      "epoch": 0.42634821139090023,
      "grad_norm": 0.39152586460113525,
      "learning_rate": 8e-05,
      "loss": 1.67,
      "step": 3945
    },
    {
      "epoch": 0.42645628444828704,
      "grad_norm": 0.36863604187965393,
      "learning_rate": 8e-05,
      "loss": 1.6239,
      "step": 3946
    },
    {
      "epoch": 0.42656435750567384,
      "grad_norm": 0.4661109447479248,
      "learning_rate": 8e-05,
      "loss": 1.9955,
      "step": 3947
    },
    {
      "epoch": 0.42667243056306065,
      "grad_norm": 0.44990408420562744,
      "learning_rate": 8e-05,
      "loss": 1.9217,
      "step": 3948
    },
    {
      "epoch": 0.42678050362044745,
      "grad_norm": 0.39468148350715637,
      "learning_rate": 8e-05,
      "loss": 1.7014,
      "step": 3949
    },
    {
      "epoch": 0.4268885766778342,
      "grad_norm": 0.375744491815567,
      "learning_rate": 8e-05,
      "loss": 1.6788,
      "step": 3950
    },
    {
      "epoch": 0.426996649735221,
      "grad_norm": 0.37918198108673096,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 3951
    },
    {
      "epoch": 0.4271047227926078,
      "grad_norm": 0.41411536931991577,
      "learning_rate": 8e-05,
      "loss": 1.7507,
      "step": 3952
    },
    {
      "epoch": 0.4272127958499946,
      "grad_norm": 0.4107980728149414,
      "learning_rate": 8e-05,
      "loss": 1.9219,
      "step": 3953
    },
    {
      "epoch": 0.4273208689073814,
      "grad_norm": 0.3732941150665283,
      "learning_rate": 8e-05,
      "loss": 1.5655,
      "step": 3954
    },
    {
      "epoch": 0.42742894196476816,
      "grad_norm": 0.3865974247455597,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 3955
    },
    {
      "epoch": 0.42753701502215496,
      "grad_norm": 0.3838065564632416,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 3956
    },
    {
      "epoch": 0.42764508807954177,
      "grad_norm": 0.3770924508571625,
      "learning_rate": 8e-05,
      "loss": 1.753,
      "step": 3957
    },
    {
      "epoch": 0.4277531611369286,
      "grad_norm": 0.38309067487716675,
      "learning_rate": 8e-05,
      "loss": 1.5574,
      "step": 3958
    },
    {
      "epoch": 0.4278612341943154,
      "grad_norm": 0.45394620299339294,
      "learning_rate": 8e-05,
      "loss": 1.7544,
      "step": 3959
    },
    {
      "epoch": 0.4279693072517021,
      "grad_norm": 0.4789636731147766,
      "learning_rate": 8e-05,
      "loss": 1.9358,
      "step": 3960
    },
    {
      "epoch": 0.42807738030908893,
      "grad_norm": 0.4569905400276184,
      "learning_rate": 8e-05,
      "loss": 1.9567,
      "step": 3961
    },
    {
      "epoch": 0.42818545336647573,
      "grad_norm": 0.3849007189273834,
      "learning_rate": 8e-05,
      "loss": 1.8783,
      "step": 3962
    },
    {
      "epoch": 0.42829352642386254,
      "grad_norm": 0.38390904664993286,
      "learning_rate": 8e-05,
      "loss": 1.7101,
      "step": 3963
    },
    {
      "epoch": 0.42840159948124934,
      "grad_norm": 0.3969222903251648,
      "learning_rate": 8e-05,
      "loss": 1.7045,
      "step": 3964
    },
    {
      "epoch": 0.42850967253863614,
      "grad_norm": 0.45306316018104553,
      "learning_rate": 8e-05,
      "loss": 1.7742,
      "step": 3965
    },
    {
      "epoch": 0.4286177455960229,
      "grad_norm": 0.3973192274570465,
      "learning_rate": 8e-05,
      "loss": 1.6415,
      "step": 3966
    },
    {
      "epoch": 0.4287258186534097,
      "grad_norm": 0.40596261620521545,
      "learning_rate": 8e-05,
      "loss": 1.7773,
      "step": 3967
    },
    {
      "epoch": 0.4288338917107965,
      "grad_norm": 0.4228540062904358,
      "learning_rate": 8e-05,
      "loss": 1.8254,
      "step": 3968
    },
    {
      "epoch": 0.4289419647681833,
      "grad_norm": 0.3980039358139038,
      "learning_rate": 8e-05,
      "loss": 1.8104,
      "step": 3969
    },
    {
      "epoch": 0.4290500378255701,
      "grad_norm": 0.4043446481227875,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 3970
    },
    {
      "epoch": 0.42915811088295686,
      "grad_norm": 0.4232507348060608,
      "learning_rate": 8e-05,
      "loss": 1.8177,
      "step": 3971
    },
    {
      "epoch": 0.42926618394034366,
      "grad_norm": 0.49053773283958435,
      "learning_rate": 8e-05,
      "loss": 1.7424,
      "step": 3972
    },
    {
      "epoch": 0.42937425699773046,
      "grad_norm": 0.3948713541030884,
      "learning_rate": 8e-05,
      "loss": 1.7545,
      "step": 3973
    },
    {
      "epoch": 0.42948233005511727,
      "grad_norm": 0.3883204162120819,
      "learning_rate": 8e-05,
      "loss": 1.788,
      "step": 3974
    },
    {
      "epoch": 0.42959040311250407,
      "grad_norm": 0.38677725195884705,
      "learning_rate": 8e-05,
      "loss": 1.625,
      "step": 3975
    },
    {
      "epoch": 0.4296984761698908,
      "grad_norm": 0.4751206636428833,
      "learning_rate": 8e-05,
      "loss": 1.6998,
      "step": 3976
    },
    {
      "epoch": 0.4298065492272776,
      "grad_norm": 0.38869211077690125,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 3977
    },
    {
      "epoch": 0.4299146222846644,
      "grad_norm": 0.49260348081588745,
      "learning_rate": 8e-05,
      "loss": 1.6929,
      "step": 3978
    },
    {
      "epoch": 0.43002269534205123,
      "grad_norm": 0.3923260569572449,
      "learning_rate": 8e-05,
      "loss": 1.7047,
      "step": 3979
    },
    {
      "epoch": 0.43013076839943803,
      "grad_norm": 0.3935205042362213,
      "learning_rate": 8e-05,
      "loss": 1.5993,
      "step": 3980
    },
    {
      "epoch": 0.43023884145682484,
      "grad_norm": 0.3812035918235779,
      "learning_rate": 8e-05,
      "loss": 1.5602,
      "step": 3981
    },
    {
      "epoch": 0.4303469145142116,
      "grad_norm": 0.37742292881011963,
      "learning_rate": 8e-05,
      "loss": 1.6258,
      "step": 3982
    },
    {
      "epoch": 0.4304549875715984,
      "grad_norm": 0.4100148677825928,
      "learning_rate": 8e-05,
      "loss": 1.7939,
      "step": 3983
    },
    {
      "epoch": 0.4305630606289852,
      "grad_norm": 0.42412251234054565,
      "learning_rate": 8e-05,
      "loss": 1.7596,
      "step": 3984
    },
    {
      "epoch": 0.430671133686372,
      "grad_norm": 0.38255825638771057,
      "learning_rate": 8e-05,
      "loss": 1.5164,
      "step": 3985
    },
    {
      "epoch": 0.4307792067437588,
      "grad_norm": 0.42709001898765564,
      "learning_rate": 8e-05,
      "loss": 1.7752,
      "step": 3986
    },
    {
      "epoch": 0.43088727980114555,
      "grad_norm": 0.3914938271045685,
      "learning_rate": 8e-05,
      "loss": 1.6601,
      "step": 3987
    },
    {
      "epoch": 0.43099535285853235,
      "grad_norm": 0.3754102289676666,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 3988
    },
    {
      "epoch": 0.43110342591591916,
      "grad_norm": 0.41938459873199463,
      "learning_rate": 8e-05,
      "loss": 1.9189,
      "step": 3989
    },
    {
      "epoch": 0.43121149897330596,
      "grad_norm": 0.3966130018234253,
      "learning_rate": 8e-05,
      "loss": 1.8492,
      "step": 3990
    },
    {
      "epoch": 0.43131957203069277,
      "grad_norm": 0.3872619569301605,
      "learning_rate": 8e-05,
      "loss": 1.5972,
      "step": 3991
    },
    {
      "epoch": 0.4314276450880795,
      "grad_norm": 0.3895432949066162,
      "learning_rate": 8e-05,
      "loss": 1.6883,
      "step": 3992
    },
    {
      "epoch": 0.4315357181454663,
      "grad_norm": 0.4713430404663086,
      "learning_rate": 8e-05,
      "loss": 2.057,
      "step": 3993
    },
    {
      "epoch": 0.4316437912028531,
      "grad_norm": 0.4199851453304291,
      "learning_rate": 8e-05,
      "loss": 1.7573,
      "step": 3994
    },
    {
      "epoch": 0.4317518642602399,
      "grad_norm": 0.44189178943634033,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 3995
    },
    {
      "epoch": 0.43185993731762673,
      "grad_norm": 0.4394003748893738,
      "learning_rate": 8e-05,
      "loss": 1.738,
      "step": 3996
    },
    {
      "epoch": 0.43196801037501353,
      "grad_norm": 0.3976317048072815,
      "learning_rate": 8e-05,
      "loss": 1.6999,
      "step": 3997
    },
    {
      "epoch": 0.4320760834324003,
      "grad_norm": 0.39891842007637024,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 3998
    },
    {
      "epoch": 0.4321841564897871,
      "grad_norm": 0.4084338843822479,
      "learning_rate": 8e-05,
      "loss": 1.7959,
      "step": 3999
    },
    {
      "epoch": 0.4322922295471739,
      "grad_norm": 0.44293370842933655,
      "learning_rate": 8e-05,
      "loss": 1.7043,
      "step": 4000
    },
    {
      "epoch": 0.4324003026045607,
      "grad_norm": 0.4261878728866577,
      "learning_rate": 8e-05,
      "loss": 1.6225,
      "step": 4001
    },
    {
      "epoch": 0.4325083756619475,
      "grad_norm": 0.3853982985019684,
      "learning_rate": 8e-05,
      "loss": 1.6346,
      "step": 4002
    },
    {
      "epoch": 0.43261644871933425,
      "grad_norm": 0.3828014135360718,
      "learning_rate": 8e-05,
      "loss": 1.6691,
      "step": 4003
    },
    {
      "epoch": 0.43272452177672105,
      "grad_norm": 0.41485053300857544,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 4004
    },
    {
      "epoch": 0.43283259483410785,
      "grad_norm": 0.4230251610279083,
      "learning_rate": 8e-05,
      "loss": 1.9448,
      "step": 4005
    },
    {
      "epoch": 0.43294066789149466,
      "grad_norm": 0.3784262239933014,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 4006
    },
    {
      "epoch": 0.43304874094888146,
      "grad_norm": 0.38575780391693115,
      "learning_rate": 8e-05,
      "loss": 1.7421,
      "step": 4007
    },
    {
      "epoch": 0.43315681400626826,
      "grad_norm": 0.3846253752708435,
      "learning_rate": 8e-05,
      "loss": 1.4029,
      "step": 4008
    },
    {
      "epoch": 0.433264887063655,
      "grad_norm": 0.3932308256626129,
      "learning_rate": 8e-05,
      "loss": 1.6591,
      "step": 4009
    },
    {
      "epoch": 0.4333729601210418,
      "grad_norm": 0.4177047908306122,
      "learning_rate": 8e-05,
      "loss": 1.698,
      "step": 4010
    },
    {
      "epoch": 0.4334810331784286,
      "grad_norm": 0.4702097177505493,
      "learning_rate": 8e-05,
      "loss": 1.8476,
      "step": 4011
    },
    {
      "epoch": 0.4335891062358154,
      "grad_norm": 0.39412257075309753,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 4012
    },
    {
      "epoch": 0.43369717929320223,
      "grad_norm": 0.3865263760089874,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 4013
    },
    {
      "epoch": 0.433805252350589,
      "grad_norm": 0.42205196619033813,
      "learning_rate": 8e-05,
      "loss": 1.752,
      "step": 4014
    },
    {
      "epoch": 0.4339133254079758,
      "grad_norm": 0.4031464159488678,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 4015
    },
    {
      "epoch": 0.4340213984653626,
      "grad_norm": 0.46766963601112366,
      "learning_rate": 8e-05,
      "loss": 1.9417,
      "step": 4016
    },
    {
      "epoch": 0.4341294715227494,
      "grad_norm": 0.4176258444786072,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 4017
    },
    {
      "epoch": 0.4342375445801362,
      "grad_norm": 0.415147066116333,
      "learning_rate": 8e-05,
      "loss": 1.6789,
      "step": 4018
    },
    {
      "epoch": 0.43434561763752294,
      "grad_norm": 0.3756083846092224,
      "learning_rate": 8e-05,
      "loss": 1.7074,
      "step": 4019
    },
    {
      "epoch": 0.43445369069490974,
      "grad_norm": 0.40239471197128296,
      "learning_rate": 8e-05,
      "loss": 1.7211,
      "step": 4020
    },
    {
      "epoch": 0.43456176375229655,
      "grad_norm": 0.3957957327365875,
      "learning_rate": 8e-05,
      "loss": 1.6966,
      "step": 4021
    },
    {
      "epoch": 0.43466983680968335,
      "grad_norm": 0.400188148021698,
      "learning_rate": 8e-05,
      "loss": 1.7349,
      "step": 4022
    },
    {
      "epoch": 0.43477790986707016,
      "grad_norm": 0.3612925708293915,
      "learning_rate": 8e-05,
      "loss": 1.5956,
      "step": 4023
    },
    {
      "epoch": 0.43488598292445696,
      "grad_norm": 0.3940187990665436,
      "learning_rate": 8e-05,
      "loss": 1.76,
      "step": 4024
    },
    {
      "epoch": 0.4349940559818437,
      "grad_norm": 0.44490760564804077,
      "learning_rate": 8e-05,
      "loss": 1.8895,
      "step": 4025
    },
    {
      "epoch": 0.4351021290392305,
      "grad_norm": 0.39821943640708923,
      "learning_rate": 8e-05,
      "loss": 1.8451,
      "step": 4026
    },
    {
      "epoch": 0.4352102020966173,
      "grad_norm": 0.3741059899330139,
      "learning_rate": 8e-05,
      "loss": 1.6333,
      "step": 4027
    },
    {
      "epoch": 0.4353182751540041,
      "grad_norm": 0.4242095649242401,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 4028
    },
    {
      "epoch": 0.4354263482113909,
      "grad_norm": 0.4120519757270813,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 4029
    },
    {
      "epoch": 0.43553442126877767,
      "grad_norm": 0.38116729259490967,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 4030
    },
    {
      "epoch": 0.4356424943261645,
      "grad_norm": 0.42187657952308655,
      "learning_rate": 8e-05,
      "loss": 1.85,
      "step": 4031
    },
    {
      "epoch": 0.4357505673835513,
      "grad_norm": 0.3659757375717163,
      "learning_rate": 8e-05,
      "loss": 1.6654,
      "step": 4032
    },
    {
      "epoch": 0.4358586404409381,
      "grad_norm": 0.42190250754356384,
      "learning_rate": 8e-05,
      "loss": 1.8069,
      "step": 4033
    },
    {
      "epoch": 0.4359667134983249,
      "grad_norm": 0.4002556800842285,
      "learning_rate": 8e-05,
      "loss": 1.782,
      "step": 4034
    },
    {
      "epoch": 0.43607478655571164,
      "grad_norm": 0.42683538794517517,
      "learning_rate": 8e-05,
      "loss": 1.9227,
      "step": 4035
    },
    {
      "epoch": 0.43618285961309844,
      "grad_norm": 0.37151455879211426,
      "learning_rate": 8e-05,
      "loss": 1.6294,
      "step": 4036
    },
    {
      "epoch": 0.43629093267048524,
      "grad_norm": 0.3944593369960785,
      "learning_rate": 8e-05,
      "loss": 1.7591,
      "step": 4037
    },
    {
      "epoch": 0.43639900572787205,
      "grad_norm": 0.37869954109191895,
      "learning_rate": 8e-05,
      "loss": 1.7272,
      "step": 4038
    },
    {
      "epoch": 0.43650707878525885,
      "grad_norm": 0.36323484778404236,
      "learning_rate": 8e-05,
      "loss": 1.6236,
      "step": 4039
    },
    {
      "epoch": 0.43661515184264565,
      "grad_norm": 0.387960821390152,
      "learning_rate": 8e-05,
      "loss": 1.6464,
      "step": 4040
    },
    {
      "epoch": 0.4367232249000324,
      "grad_norm": 0.44527754187583923,
      "learning_rate": 8e-05,
      "loss": 1.7654,
      "step": 4041
    },
    {
      "epoch": 0.4368312979574192,
      "grad_norm": 0.39944934844970703,
      "learning_rate": 8e-05,
      "loss": 1.6786,
      "step": 4042
    },
    {
      "epoch": 0.436939371014806,
      "grad_norm": 0.42589429020881653,
      "learning_rate": 8e-05,
      "loss": 1.5756,
      "step": 4043
    },
    {
      "epoch": 0.4370474440721928,
      "grad_norm": 0.42116838693618774,
      "learning_rate": 8e-05,
      "loss": 1.7177,
      "step": 4044
    },
    {
      "epoch": 0.4371555171295796,
      "grad_norm": 0.4094012379646301,
      "learning_rate": 8e-05,
      "loss": 1.7135,
      "step": 4045
    },
    {
      "epoch": 0.43726359018696637,
      "grad_norm": 0.39641955494880676,
      "learning_rate": 8e-05,
      "loss": 1.8137,
      "step": 4046
    },
    {
      "epoch": 0.43737166324435317,
      "grad_norm": 0.4124731421470642,
      "learning_rate": 8e-05,
      "loss": 1.9491,
      "step": 4047
    },
    {
      "epoch": 0.43747973630174,
      "grad_norm": 0.4227345287799835,
      "learning_rate": 8e-05,
      "loss": 1.8397,
      "step": 4048
    },
    {
      "epoch": 0.4375878093591268,
      "grad_norm": 0.38127344846725464,
      "learning_rate": 8e-05,
      "loss": 1.6297,
      "step": 4049
    },
    {
      "epoch": 0.4376958824165136,
      "grad_norm": 0.39730146527290344,
      "learning_rate": 8e-05,
      "loss": 1.6165,
      "step": 4050
    },
    {
      "epoch": 0.43780395547390033,
      "grad_norm": 0.39098402857780457,
      "learning_rate": 8e-05,
      "loss": 1.6508,
      "step": 4051
    },
    {
      "epoch": 0.43791202853128713,
      "grad_norm": 0.392983078956604,
      "learning_rate": 8e-05,
      "loss": 1.6921,
      "step": 4052
    },
    {
      "epoch": 0.43802010158867394,
      "grad_norm": 0.4320504069328308,
      "learning_rate": 8e-05,
      "loss": 1.8872,
      "step": 4053
    },
    {
      "epoch": 0.43812817464606074,
      "grad_norm": 0.44054827094078064,
      "learning_rate": 8e-05,
      "loss": 1.6491,
      "step": 4054
    },
    {
      "epoch": 0.43823624770344755,
      "grad_norm": 0.44570258259773254,
      "learning_rate": 8e-05,
      "loss": 1.9266,
      "step": 4055
    },
    {
      "epoch": 0.43834432076083435,
      "grad_norm": 0.40338799357414246,
      "learning_rate": 8e-05,
      "loss": 1.6142,
      "step": 4056
    },
    {
      "epoch": 0.4384523938182211,
      "grad_norm": 0.41716915369033813,
      "learning_rate": 8e-05,
      "loss": 1.7461,
      "step": 4057
    },
    {
      "epoch": 0.4385604668756079,
      "grad_norm": 0.40200549364089966,
      "learning_rate": 8e-05,
      "loss": 1.4431,
      "step": 4058
    },
    {
      "epoch": 0.4386685399329947,
      "grad_norm": 0.40294507145881653,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 4059
    },
    {
      "epoch": 0.4387766129903815,
      "grad_norm": 0.427957683801651,
      "learning_rate": 8e-05,
      "loss": 1.9719,
      "step": 4060
    },
    {
      "epoch": 0.4388846860477683,
      "grad_norm": 0.3961624503135681,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 4061
    },
    {
      "epoch": 0.43899275910515506,
      "grad_norm": 0.40867164731025696,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 4062
    },
    {
      "epoch": 0.43910083216254187,
      "grad_norm": 0.4313106834888458,
      "learning_rate": 8e-05,
      "loss": 1.8487,
      "step": 4063
    },
    {
      "epoch": 0.43920890521992867,
      "grad_norm": 0.380979061126709,
      "learning_rate": 8e-05,
      "loss": 1.7664,
      "step": 4064
    },
    {
      "epoch": 0.4393169782773155,
      "grad_norm": 0.4122730791568756,
      "learning_rate": 8e-05,
      "loss": 1.6375,
      "step": 4065
    },
    {
      "epoch": 0.4394250513347023,
      "grad_norm": 0.3955979645252228,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 4066
    },
    {
      "epoch": 0.439533124392089,
      "grad_norm": 0.401302695274353,
      "learning_rate": 8e-05,
      "loss": 1.8593,
      "step": 4067
    },
    {
      "epoch": 0.43964119744947583,
      "grad_norm": 0.385590136051178,
      "learning_rate": 8e-05,
      "loss": 1.7359,
      "step": 4068
    },
    {
      "epoch": 0.43974927050686263,
      "grad_norm": 0.4865040183067322,
      "learning_rate": 8e-05,
      "loss": 1.8133,
      "step": 4069
    },
    {
      "epoch": 0.43985734356424944,
      "grad_norm": 0.42025017738342285,
      "learning_rate": 8e-05,
      "loss": 1.7527,
      "step": 4070
    },
    {
      "epoch": 0.43996541662163624,
      "grad_norm": 0.39710304141044617,
      "learning_rate": 8e-05,
      "loss": 1.7476,
      "step": 4071
    },
    {
      "epoch": 0.44007348967902304,
      "grad_norm": 0.37739142775535583,
      "learning_rate": 8e-05,
      "loss": 1.692,
      "step": 4072
    },
    {
      "epoch": 0.4401815627364098,
      "grad_norm": 0.4381338655948639,
      "learning_rate": 8e-05,
      "loss": 1.8339,
      "step": 4073
    },
    {
      "epoch": 0.4402896357937966,
      "grad_norm": 0.4066935181617737,
      "learning_rate": 8e-05,
      "loss": 1.8268,
      "step": 4074
    },
    {
      "epoch": 0.4403977088511834,
      "grad_norm": 0.39243197441101074,
      "learning_rate": 8e-05,
      "loss": 1.7776,
      "step": 4075
    },
    {
      "epoch": 0.4405057819085702,
      "grad_norm": 0.3946973383426666,
      "learning_rate": 8e-05,
      "loss": 1.7883,
      "step": 4076
    },
    {
      "epoch": 0.440613854965957,
      "grad_norm": 0.37950167059898376,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 4077
    },
    {
      "epoch": 0.44072192802334376,
      "grad_norm": 0.3903473913669586,
      "learning_rate": 8e-05,
      "loss": 1.7724,
      "step": 4078
    },
    {
      "epoch": 0.44083000108073056,
      "grad_norm": 0.3659137487411499,
      "learning_rate": 8e-05,
      "loss": 1.6077,
      "step": 4079
    },
    {
      "epoch": 0.44093807413811736,
      "grad_norm": 0.36249393224716187,
      "learning_rate": 8e-05,
      "loss": 1.6665,
      "step": 4080
    },
    {
      "epoch": 0.44104614719550417,
      "grad_norm": 0.38518384099006653,
      "learning_rate": 8e-05,
      "loss": 1.6907,
      "step": 4081
    },
    {
      "epoch": 0.44115422025289097,
      "grad_norm": 0.3802357316017151,
      "learning_rate": 8e-05,
      "loss": 1.7006,
      "step": 4082
    },
    {
      "epoch": 0.4412622933102777,
      "grad_norm": 0.4164290726184845,
      "learning_rate": 8e-05,
      "loss": 1.7803,
      "step": 4083
    },
    {
      "epoch": 0.4413703663676645,
      "grad_norm": 0.3685266375541687,
      "learning_rate": 8e-05,
      "loss": 1.5313,
      "step": 4084
    },
    {
      "epoch": 0.4414784394250513,
      "grad_norm": 0.3983461558818817,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 4085
    },
    {
      "epoch": 0.44158651248243813,
      "grad_norm": 0.38751694560050964,
      "learning_rate": 8e-05,
      "loss": 1.5709,
      "step": 4086
    },
    {
      "epoch": 0.44169458553982494,
      "grad_norm": 0.418019562959671,
      "learning_rate": 8e-05,
      "loss": 1.8252,
      "step": 4087
    },
    {
      "epoch": 0.44180265859721174,
      "grad_norm": 0.42111557722091675,
      "learning_rate": 8e-05,
      "loss": 1.93,
      "step": 4088
    },
    {
      "epoch": 0.4419107316545985,
      "grad_norm": 0.3989101052284241,
      "learning_rate": 8e-05,
      "loss": 1.7258,
      "step": 4089
    },
    {
      "epoch": 0.4420188047119853,
      "grad_norm": 0.3857951760292053,
      "learning_rate": 8e-05,
      "loss": 1.6192,
      "step": 4090
    },
    {
      "epoch": 0.4421268777693721,
      "grad_norm": 0.39393389225006104,
      "learning_rate": 8e-05,
      "loss": 1.7811,
      "step": 4091
    },
    {
      "epoch": 0.4422349508267589,
      "grad_norm": 0.4382179081439972,
      "learning_rate": 8e-05,
      "loss": 1.765,
      "step": 4092
    },
    {
      "epoch": 0.4423430238841457,
      "grad_norm": 0.4087485074996948,
      "learning_rate": 8e-05,
      "loss": 1.6377,
      "step": 4093
    },
    {
      "epoch": 0.44245109694153245,
      "grad_norm": 0.3930310308933258,
      "learning_rate": 8e-05,
      "loss": 1.7044,
      "step": 4094
    },
    {
      "epoch": 0.44255916999891926,
      "grad_norm": 0.3805514872074127,
      "learning_rate": 8e-05,
      "loss": 1.7275,
      "step": 4095
    },
    {
      "epoch": 0.44266724305630606,
      "grad_norm": 0.4075923264026642,
      "learning_rate": 8e-05,
      "loss": 1.7615,
      "step": 4096
    },
    {
      "epoch": 0.44277531611369286,
      "grad_norm": 0.38746529817581177,
      "learning_rate": 8e-05,
      "loss": 1.675,
      "step": 4097
    },
    {
      "epoch": 0.44288338917107967,
      "grad_norm": 0.3932340443134308,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 4098
    },
    {
      "epoch": 0.44299146222846647,
      "grad_norm": 0.3965499699115753,
      "learning_rate": 8e-05,
      "loss": 1.6854,
      "step": 4099
    },
    {
      "epoch": 0.4430995352858532,
      "grad_norm": 0.4528593122959137,
      "learning_rate": 8e-05,
      "loss": 1.9614,
      "step": 4100
    },
    {
      "epoch": 0.44320760834324,
      "grad_norm": 0.38144510984420776,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 4101
    },
    {
      "epoch": 0.4433156814006268,
      "grad_norm": 0.3901524841785431,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 4102
    },
    {
      "epoch": 0.44342375445801363,
      "grad_norm": 0.4248017966747284,
      "learning_rate": 8e-05,
      "loss": 1.8402,
      "step": 4103
    },
    {
      "epoch": 0.44353182751540043,
      "grad_norm": 0.4670308828353882,
      "learning_rate": 8e-05,
      "loss": 1.772,
      "step": 4104
    },
    {
      "epoch": 0.4436399005727872,
      "grad_norm": 0.4132468104362488,
      "learning_rate": 8e-05,
      "loss": 1.8095,
      "step": 4105
    },
    {
      "epoch": 0.443747973630174,
      "grad_norm": 0.38740426301956177,
      "learning_rate": 8e-05,
      "loss": 1.8122,
      "step": 4106
    },
    {
      "epoch": 0.4438560466875608,
      "grad_norm": 0.35123327374458313,
      "learning_rate": 8e-05,
      "loss": 1.5147,
      "step": 4107
    },
    {
      "epoch": 0.4439641197449476,
      "grad_norm": 0.360033243894577,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 4108
    },
    {
      "epoch": 0.4440721928023344,
      "grad_norm": 0.3886578381061554,
      "learning_rate": 8e-05,
      "loss": 1.7402,
      "step": 4109
    },
    {
      "epoch": 0.44418026585972115,
      "grad_norm": 0.38930684328079224,
      "learning_rate": 8e-05,
      "loss": 1.8175,
      "step": 4110
    },
    {
      "epoch": 0.44428833891710795,
      "grad_norm": 0.38217923045158386,
      "learning_rate": 8e-05,
      "loss": 1.4983,
      "step": 4111
    },
    {
      "epoch": 0.44439641197449475,
      "grad_norm": 0.45865002274513245,
      "learning_rate": 8e-05,
      "loss": 1.7772,
      "step": 4112
    },
    {
      "epoch": 0.44450448503188156,
      "grad_norm": 0.45548492670059204,
      "learning_rate": 8e-05,
      "loss": 1.8117,
      "step": 4113
    },
    {
      "epoch": 0.44461255808926836,
      "grad_norm": 0.4268307089805603,
      "learning_rate": 8e-05,
      "loss": 2.0046,
      "step": 4114
    },
    {
      "epoch": 0.44472063114665517,
      "grad_norm": 0.41946035623550415,
      "learning_rate": 8e-05,
      "loss": 1.863,
      "step": 4115
    },
    {
      "epoch": 0.4448287042040419,
      "grad_norm": 0.43050646781921387,
      "learning_rate": 8e-05,
      "loss": 1.8613,
      "step": 4116
    },
    {
      "epoch": 0.4449367772614287,
      "grad_norm": 0.38218268752098083,
      "learning_rate": 8e-05,
      "loss": 1.6927,
      "step": 4117
    },
    {
      "epoch": 0.4450448503188155,
      "grad_norm": 0.3812640607357025,
      "learning_rate": 8e-05,
      "loss": 1.7996,
      "step": 4118
    },
    {
      "epoch": 0.4451529233762023,
      "grad_norm": 0.4203759729862213,
      "learning_rate": 8e-05,
      "loss": 1.8412,
      "step": 4119
    },
    {
      "epoch": 0.44526099643358913,
      "grad_norm": 0.39401933550834656,
      "learning_rate": 8e-05,
      "loss": 1.6846,
      "step": 4120
    },
    {
      "epoch": 0.4453690694909759,
      "grad_norm": 0.41144585609436035,
      "learning_rate": 8e-05,
      "loss": 1.7655,
      "step": 4121
    },
    {
      "epoch": 0.4454771425483627,
      "grad_norm": 0.40371081233024597,
      "learning_rate": 8e-05,
      "loss": 1.7963,
      "step": 4122
    },
    {
      "epoch": 0.4455852156057495,
      "grad_norm": 0.38249272108078003,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 4123
    },
    {
      "epoch": 0.4456932886631363,
      "grad_norm": 0.40065309405326843,
      "learning_rate": 8e-05,
      "loss": 1.6287,
      "step": 4124
    },
    {
      "epoch": 0.4458013617205231,
      "grad_norm": 0.39524564146995544,
      "learning_rate": 8e-05,
      "loss": 1.8478,
      "step": 4125
    },
    {
      "epoch": 0.44590943477790984,
      "grad_norm": 0.40700867772102356,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 4126
    },
    {
      "epoch": 0.44601750783529664,
      "grad_norm": 0.40793606638908386,
      "learning_rate": 8e-05,
      "loss": 1.8052,
      "step": 4127
    },
    {
      "epoch": 0.44612558089268345,
      "grad_norm": 0.43469879031181335,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 4128
    },
    {
      "epoch": 0.44623365395007025,
      "grad_norm": 0.40545299649238586,
      "learning_rate": 8e-05,
      "loss": 1.7583,
      "step": 4129
    },
    {
      "epoch": 0.44634172700745706,
      "grad_norm": 0.4154534935951233,
      "learning_rate": 8e-05,
      "loss": 1.8038,
      "step": 4130
    },
    {
      "epoch": 0.44644980006484386,
      "grad_norm": 0.37462887167930603,
      "learning_rate": 8e-05,
      "loss": 1.6905,
      "step": 4131
    },
    {
      "epoch": 0.4465578731222306,
      "grad_norm": 0.4171428680419922,
      "learning_rate": 8e-05,
      "loss": 1.7983,
      "step": 4132
    },
    {
      "epoch": 0.4466659461796174,
      "grad_norm": 0.40016424655914307,
      "learning_rate": 8e-05,
      "loss": 1.7067,
      "step": 4133
    },
    {
      "epoch": 0.4467740192370042,
      "grad_norm": 0.4300216734409332,
      "learning_rate": 8e-05,
      "loss": 1.8315,
      "step": 4134
    },
    {
      "epoch": 0.446882092294391,
      "grad_norm": 0.4333845376968384,
      "learning_rate": 8e-05,
      "loss": 1.8441,
      "step": 4135
    },
    {
      "epoch": 0.4469901653517778,
      "grad_norm": 0.38947948813438416,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 4136
    },
    {
      "epoch": 0.4470982384091646,
      "grad_norm": 0.3804239332675934,
      "learning_rate": 8e-05,
      "loss": 1.6908,
      "step": 4137
    },
    {
      "epoch": 0.4472063114665514,
      "grad_norm": 0.39848506450653076,
      "learning_rate": 8e-05,
      "loss": 1.7669,
      "step": 4138
    },
    {
      "epoch": 0.4473143845239382,
      "grad_norm": 0.4335612952709198,
      "learning_rate": 8e-05,
      "loss": 1.6697,
      "step": 4139
    },
    {
      "epoch": 0.447422457581325,
      "grad_norm": 0.395724892616272,
      "learning_rate": 8e-05,
      "loss": 1.6947,
      "step": 4140
    },
    {
      "epoch": 0.4475305306387118,
      "grad_norm": 0.4212934672832489,
      "learning_rate": 8e-05,
      "loss": 1.7986,
      "step": 4141
    },
    {
      "epoch": 0.44763860369609854,
      "grad_norm": 0.3814888000488281,
      "learning_rate": 8e-05,
      "loss": 1.563,
      "step": 4142
    },
    {
      "epoch": 0.44774667675348534,
      "grad_norm": 0.3882804811000824,
      "learning_rate": 8e-05,
      "loss": 1.6502,
      "step": 4143
    },
    {
      "epoch": 0.44785474981087214,
      "grad_norm": 0.409474641084671,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 4144
    },
    {
      "epoch": 0.44796282286825895,
      "grad_norm": 0.36955225467681885,
      "learning_rate": 8e-05,
      "loss": 1.6342,
      "step": 4145
    },
    {
      "epoch": 0.44807089592564575,
      "grad_norm": 0.3800424635410309,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 4146
    },
    {
      "epoch": 0.44817896898303256,
      "grad_norm": 0.41920849680900574,
      "learning_rate": 8e-05,
      "loss": 1.631,
      "step": 4147
    },
    {
      "epoch": 0.4482870420404193,
      "grad_norm": 0.4319944381713867,
      "learning_rate": 8e-05,
      "loss": 1.6814,
      "step": 4148
    },
    {
      "epoch": 0.4483951150978061,
      "grad_norm": 0.4809088706970215,
      "learning_rate": 8e-05,
      "loss": 1.9423,
      "step": 4149
    },
    {
      "epoch": 0.4485031881551929,
      "grad_norm": 0.4052334427833557,
      "learning_rate": 8e-05,
      "loss": 1.7557,
      "step": 4150
    },
    {
      "epoch": 0.4486112612125797,
      "grad_norm": 0.3898913562297821,
      "learning_rate": 8e-05,
      "loss": 1.6813,
      "step": 4151
    },
    {
      "epoch": 0.4487193342699665,
      "grad_norm": 0.41606011986732483,
      "learning_rate": 8e-05,
      "loss": 1.8142,
      "step": 4152
    },
    {
      "epoch": 0.44882740732735327,
      "grad_norm": 0.38073471188545227,
      "learning_rate": 8e-05,
      "loss": 1.7149,
      "step": 4153
    },
    {
      "epoch": 0.44893548038474007,
      "grad_norm": 0.41051143407821655,
      "learning_rate": 8e-05,
      "loss": 1.5759,
      "step": 4154
    },
    {
      "epoch": 0.4490435534421269,
      "grad_norm": 0.39827659726142883,
      "learning_rate": 8e-05,
      "loss": 1.7553,
      "step": 4155
    },
    {
      "epoch": 0.4491516264995137,
      "grad_norm": 0.39858707785606384,
      "learning_rate": 8e-05,
      "loss": 1.6401,
      "step": 4156
    },
    {
      "epoch": 0.4492596995569005,
      "grad_norm": 0.41112565994262695,
      "learning_rate": 8e-05,
      "loss": 1.7226,
      "step": 4157
    },
    {
      "epoch": 0.44936777261428723,
      "grad_norm": 0.38678601384162903,
      "learning_rate": 8e-05,
      "loss": 1.7793,
      "step": 4158
    },
    {
      "epoch": 0.44947584567167403,
      "grad_norm": 0.3781753182411194,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 4159
    },
    {
      "epoch": 0.44958391872906084,
      "grad_norm": 0.39570656418800354,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 4160
    },
    {
      "epoch": 0.44969199178644764,
      "grad_norm": 0.38605526089668274,
      "learning_rate": 8e-05,
      "loss": 1.6397,
      "step": 4161
    },
    {
      "epoch": 0.44980006484383445,
      "grad_norm": 0.3786720931529999,
      "learning_rate": 8e-05,
      "loss": 1.6792,
      "step": 4162
    },
    {
      "epoch": 0.44990813790122125,
      "grad_norm": 0.39362895488739014,
      "learning_rate": 8e-05,
      "loss": 1.4707,
      "step": 4163
    },
    {
      "epoch": 0.450016210958608,
      "grad_norm": 0.3991580009460449,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 4164
    },
    {
      "epoch": 0.4501242840159948,
      "grad_norm": 0.3815881311893463,
      "learning_rate": 8e-05,
      "loss": 1.6125,
      "step": 4165
    },
    {
      "epoch": 0.4502323570733816,
      "grad_norm": 0.4243254065513611,
      "learning_rate": 8e-05,
      "loss": 1.8755,
      "step": 4166
    },
    {
      "epoch": 0.4503404301307684,
      "grad_norm": 0.37091976404190063,
      "learning_rate": 8e-05,
      "loss": 1.536,
      "step": 4167
    },
    {
      "epoch": 0.4504485031881552,
      "grad_norm": 0.3933130204677582,
      "learning_rate": 8e-05,
      "loss": 1.7422,
      "step": 4168
    },
    {
      "epoch": 0.45055657624554196,
      "grad_norm": 0.41969093680381775,
      "learning_rate": 8e-05,
      "loss": 1.8135,
      "step": 4169
    },
    {
      "epoch": 0.45066464930292877,
      "grad_norm": 0.40764015913009644,
      "learning_rate": 8e-05,
      "loss": 1.6265,
      "step": 4170
    },
    {
      "epoch": 0.45077272236031557,
      "grad_norm": 0.39879798889160156,
      "learning_rate": 8e-05,
      "loss": 1.7237,
      "step": 4171
    },
    {
      "epoch": 0.4508807954177024,
      "grad_norm": 0.4008561074733734,
      "learning_rate": 8e-05,
      "loss": 1.7057,
      "step": 4172
    },
    {
      "epoch": 0.4509888684750892,
      "grad_norm": 0.3942168354988098,
      "learning_rate": 8e-05,
      "loss": 1.866,
      "step": 4173
    },
    {
      "epoch": 0.451096941532476,
      "grad_norm": 0.4003502428531647,
      "learning_rate": 8e-05,
      "loss": 1.7716,
      "step": 4174
    },
    {
      "epoch": 0.45120501458986273,
      "grad_norm": 0.38686972856521606,
      "learning_rate": 8e-05,
      "loss": 1.7233,
      "step": 4175
    },
    {
      "epoch": 0.45131308764724953,
      "grad_norm": 0.3975874185562134,
      "learning_rate": 8e-05,
      "loss": 1.6512,
      "step": 4176
    },
    {
      "epoch": 0.45142116070463634,
      "grad_norm": 0.4099818170070648,
      "learning_rate": 8e-05,
      "loss": 1.6666,
      "step": 4177
    },
    {
      "epoch": 0.45152923376202314,
      "grad_norm": 0.39071857929229736,
      "learning_rate": 8e-05,
      "loss": 1.8683,
      "step": 4178
    },
    {
      "epoch": 0.45163730681940994,
      "grad_norm": 0.40196681022644043,
      "learning_rate": 8e-05,
      "loss": 1.5938,
      "step": 4179
    },
    {
      "epoch": 0.4517453798767967,
      "grad_norm": 0.39036962389945984,
      "learning_rate": 8e-05,
      "loss": 1.7004,
      "step": 4180
    },
    {
      "epoch": 0.4518534529341835,
      "grad_norm": 0.38082221150398254,
      "learning_rate": 8e-05,
      "loss": 1.7278,
      "step": 4181
    },
    {
      "epoch": 0.4519615259915703,
      "grad_norm": 0.40840616822242737,
      "learning_rate": 8e-05,
      "loss": 1.5474,
      "step": 4182
    },
    {
      "epoch": 0.4520695990489571,
      "grad_norm": 0.366936594247818,
      "learning_rate": 8e-05,
      "loss": 1.598,
      "step": 4183
    },
    {
      "epoch": 0.4521776721063439,
      "grad_norm": 0.43335720896720886,
      "learning_rate": 8e-05,
      "loss": 1.8735,
      "step": 4184
    },
    {
      "epoch": 0.45228574516373066,
      "grad_norm": 0.4122973084449768,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 4185
    },
    {
      "epoch": 0.45239381822111746,
      "grad_norm": 0.38433751463890076,
      "learning_rate": 8e-05,
      "loss": 1.5977,
      "step": 4186
    },
    {
      "epoch": 0.45250189127850426,
      "grad_norm": 0.3955709934234619,
      "learning_rate": 8e-05,
      "loss": 1.8994,
      "step": 4187
    },
    {
      "epoch": 0.45260996433589107,
      "grad_norm": 0.38264644145965576,
      "learning_rate": 8e-05,
      "loss": 1.6156,
      "step": 4188
    },
    {
      "epoch": 0.4527180373932779,
      "grad_norm": 0.4050297141075134,
      "learning_rate": 8e-05,
      "loss": 1.6352,
      "step": 4189
    },
    {
      "epoch": 0.4528261104506647,
      "grad_norm": 0.3812471628189087,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 4190
    },
    {
      "epoch": 0.4529341835080514,
      "grad_norm": 0.3948255181312561,
      "learning_rate": 8e-05,
      "loss": 1.7144,
      "step": 4191
    },
    {
      "epoch": 0.45304225656543823,
      "grad_norm": 0.4553428292274475,
      "learning_rate": 8e-05,
      "loss": 1.7806,
      "step": 4192
    },
    {
      "epoch": 0.45315032962282503,
      "grad_norm": 0.42498478293418884,
      "learning_rate": 8e-05,
      "loss": 1.7588,
      "step": 4193
    },
    {
      "epoch": 0.45325840268021184,
      "grad_norm": 0.3939964473247528,
      "learning_rate": 8e-05,
      "loss": 1.8188,
      "step": 4194
    },
    {
      "epoch": 0.45336647573759864,
      "grad_norm": 0.3810969889163971,
      "learning_rate": 8e-05,
      "loss": 1.7752,
      "step": 4195
    },
    {
      "epoch": 0.4534745487949854,
      "grad_norm": 0.42184048891067505,
      "learning_rate": 8e-05,
      "loss": 1.8841,
      "step": 4196
    },
    {
      "epoch": 0.4535826218523722,
      "grad_norm": 0.41052091121673584,
      "learning_rate": 8e-05,
      "loss": 1.6255,
      "step": 4197
    },
    {
      "epoch": 0.453690694909759,
      "grad_norm": 0.4232829511165619,
      "learning_rate": 8e-05,
      "loss": 1.7828,
      "step": 4198
    },
    {
      "epoch": 0.4537987679671458,
      "grad_norm": 0.3979233503341675,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 4199
    },
    {
      "epoch": 0.4539068410245326,
      "grad_norm": 0.4140111804008484,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 4200
    },
    {
      "epoch": 0.45401491408191935,
      "grad_norm": 0.4476223289966583,
      "learning_rate": 8e-05,
      "loss": 1.8369,
      "step": 4201
    },
    {
      "epoch": 0.45412298713930616,
      "grad_norm": 0.4117853343486786,
      "learning_rate": 8e-05,
      "loss": 1.6882,
      "step": 4202
    },
    {
      "epoch": 0.45423106019669296,
      "grad_norm": 0.38173508644104004,
      "learning_rate": 8e-05,
      "loss": 1.7465,
      "step": 4203
    },
    {
      "epoch": 0.45433913325407976,
      "grad_norm": 0.41248437762260437,
      "learning_rate": 8e-05,
      "loss": 1.6825,
      "step": 4204
    },
    {
      "epoch": 0.45444720631146657,
      "grad_norm": 0.4649130702018738,
      "learning_rate": 8e-05,
      "loss": 1.7466,
      "step": 4205
    },
    {
      "epoch": 0.45455527936885337,
      "grad_norm": 0.4088135361671448,
      "learning_rate": 8e-05,
      "loss": 1.8455,
      "step": 4206
    },
    {
      "epoch": 0.4546633524262401,
      "grad_norm": 0.3916531801223755,
      "learning_rate": 8e-05,
      "loss": 1.5961,
      "step": 4207
    },
    {
      "epoch": 0.4547714254836269,
      "grad_norm": 0.3791961073875427,
      "learning_rate": 8e-05,
      "loss": 1.6748,
      "step": 4208
    },
    {
      "epoch": 0.4548794985410137,
      "grad_norm": 0.40511149168014526,
      "learning_rate": 8e-05,
      "loss": 1.7729,
      "step": 4209
    },
    {
      "epoch": 0.45498757159840053,
      "grad_norm": 0.4018228054046631,
      "learning_rate": 8e-05,
      "loss": 1.8085,
      "step": 4210
    },
    {
      "epoch": 0.45509564465578733,
      "grad_norm": 0.3974384069442749,
      "learning_rate": 8e-05,
      "loss": 1.7313,
      "step": 4211
    },
    {
      "epoch": 0.4552037177131741,
      "grad_norm": 0.3767494857311249,
      "learning_rate": 8e-05,
      "loss": 1.7715,
      "step": 4212
    },
    {
      "epoch": 0.4553117907705609,
      "grad_norm": 0.43515220284461975,
      "learning_rate": 8e-05,
      "loss": 1.6038,
      "step": 4213
    },
    {
      "epoch": 0.4554198638279477,
      "grad_norm": 0.3716447353363037,
      "learning_rate": 8e-05,
      "loss": 1.6678,
      "step": 4214
    },
    {
      "epoch": 0.4555279368853345,
      "grad_norm": 0.36889275908470154,
      "learning_rate": 8e-05,
      "loss": 1.6402,
      "step": 4215
    },
    {
      "epoch": 0.4556360099427213,
      "grad_norm": 0.3953246772289276,
      "learning_rate": 8e-05,
      "loss": 1.8373,
      "step": 4216
    },
    {
      "epoch": 0.45574408300010805,
      "grad_norm": 0.40527579188346863,
      "learning_rate": 8e-05,
      "loss": 1.6154,
      "step": 4217
    },
    {
      "epoch": 0.45585215605749485,
      "grad_norm": 0.40630364418029785,
      "learning_rate": 8e-05,
      "loss": 1.8841,
      "step": 4218
    },
    {
      "epoch": 0.45596022911488165,
      "grad_norm": 0.3697873055934906,
      "learning_rate": 8e-05,
      "loss": 1.7918,
      "step": 4219
    },
    {
      "epoch": 0.45606830217226846,
      "grad_norm": 0.3869011104106903,
      "learning_rate": 8e-05,
      "loss": 1.7002,
      "step": 4220
    },
    {
      "epoch": 0.45617637522965526,
      "grad_norm": 0.4355603754520416,
      "learning_rate": 8e-05,
      "loss": 1.8407,
      "step": 4221
    },
    {
      "epoch": 0.45628444828704207,
      "grad_norm": 0.4080151319503784,
      "learning_rate": 8e-05,
      "loss": 1.832,
      "step": 4222
    },
    {
      "epoch": 0.4563925213444288,
      "grad_norm": 0.37484556436538696,
      "learning_rate": 8e-05,
      "loss": 1.7048,
      "step": 4223
    },
    {
      "epoch": 0.4565005944018156,
      "grad_norm": 0.38602215051651,
      "learning_rate": 8e-05,
      "loss": 1.7111,
      "step": 4224
    },
    {
      "epoch": 0.4566086674592024,
      "grad_norm": 0.41468873620033264,
      "learning_rate": 8e-05,
      "loss": 1.8882,
      "step": 4225
    },
    {
      "epoch": 0.4567167405165892,
      "grad_norm": 0.40375640988349915,
      "learning_rate": 8e-05,
      "loss": 1.8361,
      "step": 4226
    },
    {
      "epoch": 0.45682481357397603,
      "grad_norm": 0.39360880851745605,
      "learning_rate": 8e-05,
      "loss": 1.6172,
      "step": 4227
    },
    {
      "epoch": 0.4569328866313628,
      "grad_norm": 0.4090273380279541,
      "learning_rate": 8e-05,
      "loss": 1.5416,
      "step": 4228
    },
    {
      "epoch": 0.4570409596887496,
      "grad_norm": 0.41469478607177734,
      "learning_rate": 8e-05,
      "loss": 1.6082,
      "step": 4229
    },
    {
      "epoch": 0.4571490327461364,
      "grad_norm": 0.40205758810043335,
      "learning_rate": 8e-05,
      "loss": 1.7276,
      "step": 4230
    },
    {
      "epoch": 0.4572571058035232,
      "grad_norm": 0.37132859230041504,
      "learning_rate": 8e-05,
      "loss": 1.6877,
      "step": 4231
    },
    {
      "epoch": 0.45736517886091,
      "grad_norm": 0.38051095604896545,
      "learning_rate": 8e-05,
      "loss": 1.7735,
      "step": 4232
    },
    {
      "epoch": 0.45747325191829674,
      "grad_norm": 0.3900478780269623,
      "learning_rate": 8e-05,
      "loss": 1.712,
      "step": 4233
    },
    {
      "epoch": 0.45758132497568355,
      "grad_norm": 0.3997434973716736,
      "learning_rate": 8e-05,
      "loss": 1.7378,
      "step": 4234
    },
    {
      "epoch": 0.45768939803307035,
      "grad_norm": 0.3884037137031555,
      "learning_rate": 8e-05,
      "loss": 1.4314,
      "step": 4235
    },
    {
      "epoch": 0.45779747109045715,
      "grad_norm": 0.38705649971961975,
      "learning_rate": 8e-05,
      "loss": 1.7249,
      "step": 4236
    },
    {
      "epoch": 0.45790554414784396,
      "grad_norm": 0.3733726143836975,
      "learning_rate": 8e-05,
      "loss": 1.4933,
      "step": 4237
    },
    {
      "epoch": 0.45801361720523076,
      "grad_norm": 0.4090176224708557,
      "learning_rate": 8e-05,
      "loss": 1.7865,
      "step": 4238
    },
    {
      "epoch": 0.4581216902626175,
      "grad_norm": 0.3869929909706116,
      "learning_rate": 8e-05,
      "loss": 1.5816,
      "step": 4239
    },
    {
      "epoch": 0.4582297633200043,
      "grad_norm": 0.402170866727829,
      "learning_rate": 8e-05,
      "loss": 1.6822,
      "step": 4240
    },
    {
      "epoch": 0.4583378363773911,
      "grad_norm": 0.3807542026042938,
      "learning_rate": 8e-05,
      "loss": 1.6426,
      "step": 4241
    },
    {
      "epoch": 0.4584459094347779,
      "grad_norm": 0.4597344696521759,
      "learning_rate": 8e-05,
      "loss": 1.6257,
      "step": 4242
    },
    {
      "epoch": 0.4585539824921647,
      "grad_norm": 0.48820099234580994,
      "learning_rate": 8e-05,
      "loss": 1.9257,
      "step": 4243
    },
    {
      "epoch": 0.4586620555495515,
      "grad_norm": 0.3972399830818176,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 4244
    },
    {
      "epoch": 0.4587701286069383,
      "grad_norm": 0.4314461946487427,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 4245
    },
    {
      "epoch": 0.4588782016643251,
      "grad_norm": 0.40236297249794006,
      "learning_rate": 8e-05,
      "loss": 1.7352,
      "step": 4246
    },
    {
      "epoch": 0.4589862747217119,
      "grad_norm": 0.384321928024292,
      "learning_rate": 8e-05,
      "loss": 1.7116,
      "step": 4247
    },
    {
      "epoch": 0.4590943477790987,
      "grad_norm": 0.41385918855667114,
      "learning_rate": 8e-05,
      "loss": 1.4514,
      "step": 4248
    },
    {
      "epoch": 0.45920242083648544,
      "grad_norm": 0.45319339632987976,
      "learning_rate": 8e-05,
      "loss": 1.8928,
      "step": 4249
    },
    {
      "epoch": 0.45931049389387224,
      "grad_norm": 0.3914586305618286,
      "learning_rate": 8e-05,
      "loss": 1.7159,
      "step": 4250
    },
    {
      "epoch": 0.45941856695125904,
      "grad_norm": 0.39298906922340393,
      "learning_rate": 8e-05,
      "loss": 1.6449,
      "step": 4251
    },
    {
      "epoch": 0.45952664000864585,
      "grad_norm": 0.3825382888317108,
      "learning_rate": 8e-05,
      "loss": 1.6443,
      "step": 4252
    },
    {
      "epoch": 0.45963471306603265,
      "grad_norm": 0.39020484685897827,
      "learning_rate": 8e-05,
      "loss": 1.731,
      "step": 4253
    },
    {
      "epoch": 0.45974278612341946,
      "grad_norm": 0.3996971547603607,
      "learning_rate": 8e-05,
      "loss": 1.6702,
      "step": 4254
    },
    {
      "epoch": 0.4598508591808062,
      "grad_norm": 0.4055745601654053,
      "learning_rate": 8e-05,
      "loss": 1.8496,
      "step": 4255
    },
    {
      "epoch": 0.459958932238193,
      "grad_norm": 0.3965752422809601,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 4256
    },
    {
      "epoch": 0.4600670052955798,
      "grad_norm": 0.39176952838897705,
      "learning_rate": 8e-05,
      "loss": 1.7194,
      "step": 4257
    },
    {
      "epoch": 0.4601750783529666,
      "grad_norm": 0.394587904214859,
      "learning_rate": 8e-05,
      "loss": 1.714,
      "step": 4258
    },
    {
      "epoch": 0.4602831514103534,
      "grad_norm": 0.38011595606803894,
      "learning_rate": 8e-05,
      "loss": 1.6683,
      "step": 4259
    },
    {
      "epoch": 0.46039122446774017,
      "grad_norm": 0.3950192928314209,
      "learning_rate": 8e-05,
      "loss": 1.7885,
      "step": 4260
    },
    {
      "epoch": 0.46049929752512697,
      "grad_norm": 0.4274528920650482,
      "learning_rate": 8e-05,
      "loss": 1.8044,
      "step": 4261
    },
    {
      "epoch": 0.4606073705825138,
      "grad_norm": 0.4229106903076172,
      "learning_rate": 8e-05,
      "loss": 1.7517,
      "step": 4262
    },
    {
      "epoch": 0.4607154436399006,
      "grad_norm": 0.3834282457828522,
      "learning_rate": 8e-05,
      "loss": 1.6904,
      "step": 4263
    },
    {
      "epoch": 0.4608235166972874,
      "grad_norm": 0.3826853930950165,
      "learning_rate": 8e-05,
      "loss": 1.555,
      "step": 4264
    },
    {
      "epoch": 0.4609315897546742,
      "grad_norm": 0.35426029562950134,
      "learning_rate": 8e-05,
      "loss": 1.5482,
      "step": 4265
    },
    {
      "epoch": 0.46103966281206094,
      "grad_norm": 0.4941841661930084,
      "learning_rate": 8e-05,
      "loss": 2.1103,
      "step": 4266
    },
    {
      "epoch": 0.46114773586944774,
      "grad_norm": 0.38645580410957336,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 4267
    },
    {
      "epoch": 0.46125580892683454,
      "grad_norm": 0.3816952407360077,
      "learning_rate": 8e-05,
      "loss": 1.6676,
      "step": 4268
    },
    {
      "epoch": 0.46136388198422135,
      "grad_norm": 0.3699115514755249,
      "learning_rate": 8e-05,
      "loss": 1.7383,
      "step": 4269
    },
    {
      "epoch": 0.46147195504160815,
      "grad_norm": 0.4379904866218567,
      "learning_rate": 8e-05,
      "loss": 1.7083,
      "step": 4270
    },
    {
      "epoch": 0.4615800280989949,
      "grad_norm": 0.40712636709213257,
      "learning_rate": 8e-05,
      "loss": 1.7554,
      "step": 4271
    },
    {
      "epoch": 0.4616881011563817,
      "grad_norm": 0.42166948318481445,
      "learning_rate": 8e-05,
      "loss": 1.727,
      "step": 4272
    },
    {
      "epoch": 0.4617961742137685,
      "grad_norm": 0.4893457889556885,
      "learning_rate": 8e-05,
      "loss": 1.8257,
      "step": 4273
    },
    {
      "epoch": 0.4619042472711553,
      "grad_norm": 0.3916114270687103,
      "learning_rate": 8e-05,
      "loss": 1.7031,
      "step": 4274
    },
    {
      "epoch": 0.4620123203285421,
      "grad_norm": 0.41695448756217957,
      "learning_rate": 8e-05,
      "loss": 1.7869,
      "step": 4275
    },
    {
      "epoch": 0.46212039338592886,
      "grad_norm": 0.38892853260040283,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 4276
    },
    {
      "epoch": 0.46222846644331567,
      "grad_norm": 0.40539973974227905,
      "learning_rate": 8e-05,
      "loss": 1.6894,
      "step": 4277
    },
    {
      "epoch": 0.46233653950070247,
      "grad_norm": 0.4102711081504822,
      "learning_rate": 8e-05,
      "loss": 1.8844,
      "step": 4278
    },
    {
      "epoch": 0.4624446125580893,
      "grad_norm": 0.38708868622779846,
      "learning_rate": 8e-05,
      "loss": 1.6971,
      "step": 4279
    },
    {
      "epoch": 0.4625526856154761,
      "grad_norm": 0.39683011174201965,
      "learning_rate": 8e-05,
      "loss": 1.5848,
      "step": 4280
    },
    {
      "epoch": 0.4626607586728629,
      "grad_norm": 0.4241722524166107,
      "learning_rate": 8e-05,
      "loss": 1.87,
      "step": 4281
    },
    {
      "epoch": 0.46276883173024963,
      "grad_norm": 0.4027661383152008,
      "learning_rate": 8e-05,
      "loss": 1.8335,
      "step": 4282
    },
    {
      "epoch": 0.46287690478763643,
      "grad_norm": 0.3885311782360077,
      "learning_rate": 8e-05,
      "loss": 1.7674,
      "step": 4283
    },
    {
      "epoch": 0.46298497784502324,
      "grad_norm": 0.3851543068885803,
      "learning_rate": 8e-05,
      "loss": 1.6956,
      "step": 4284
    },
    {
      "epoch": 0.46309305090241004,
      "grad_norm": 0.3999451994895935,
      "learning_rate": 8e-05,
      "loss": 1.7463,
      "step": 4285
    },
    {
      "epoch": 0.46320112395979685,
      "grad_norm": 0.406423956155777,
      "learning_rate": 8e-05,
      "loss": 1.8638,
      "step": 4286
    },
    {
      "epoch": 0.4633091970171836,
      "grad_norm": 0.3860330879688263,
      "learning_rate": 8e-05,
      "loss": 1.7611,
      "step": 4287
    },
    {
      "epoch": 0.4634172700745704,
      "grad_norm": 0.390470027923584,
      "learning_rate": 8e-05,
      "loss": 1.682,
      "step": 4288
    },
    {
      "epoch": 0.4635253431319572,
      "grad_norm": 0.4225302040576935,
      "learning_rate": 8e-05,
      "loss": 1.7026,
      "step": 4289
    },
    {
      "epoch": 0.463633416189344,
      "grad_norm": 0.5033335089683533,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 4290
    },
    {
      "epoch": 0.4637414892467308,
      "grad_norm": 0.4086191952228546,
      "learning_rate": 8e-05,
      "loss": 1.7551,
      "step": 4291
    },
    {
      "epoch": 0.46384956230411756,
      "grad_norm": 0.40312811732292175,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 4292
    },
    {
      "epoch": 0.46395763536150436,
      "grad_norm": 0.38233914971351624,
      "learning_rate": 8e-05,
      "loss": 1.6473,
      "step": 4293
    },
    {
      "epoch": 0.46406570841889117,
      "grad_norm": 0.4416559934616089,
      "learning_rate": 8e-05,
      "loss": 1.8628,
      "step": 4294
    },
    {
      "epoch": 0.46417378147627797,
      "grad_norm": 0.39385291934013367,
      "learning_rate": 8e-05,
      "loss": 1.7711,
      "step": 4295
    },
    {
      "epoch": 0.4642818545336648,
      "grad_norm": 0.42280063033103943,
      "learning_rate": 8e-05,
      "loss": 1.6657,
      "step": 4296
    },
    {
      "epoch": 0.4643899275910516,
      "grad_norm": 0.40708550810813904,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 4297
    },
    {
      "epoch": 0.4644980006484383,
      "grad_norm": 0.40606924891471863,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 4298
    },
    {
      "epoch": 0.46460607370582513,
      "grad_norm": 0.3782714903354645,
      "learning_rate": 8e-05,
      "loss": 1.6565,
      "step": 4299
    },
    {
      "epoch": 0.46471414676321193,
      "grad_norm": 0.4321131408214569,
      "learning_rate": 8e-05,
      "loss": 1.6977,
      "step": 4300
    },
    {
      "epoch": 0.46482221982059874,
      "grad_norm": 0.42701172828674316,
      "learning_rate": 8e-05,
      "loss": 1.7951,
      "step": 4301
    },
    {
      "epoch": 0.46493029287798554,
      "grad_norm": 0.3887809216976166,
      "learning_rate": 8e-05,
      "loss": 1.6972,
      "step": 4302
    },
    {
      "epoch": 0.4650383659353723,
      "grad_norm": 0.378823459148407,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 4303
    },
    {
      "epoch": 0.4651464389927591,
      "grad_norm": 0.40994536876678467,
      "learning_rate": 8e-05,
      "loss": 1.6515,
      "step": 4304
    },
    {
      "epoch": 0.4652545120501459,
      "grad_norm": 0.41217079758644104,
      "learning_rate": 8e-05,
      "loss": 1.8695,
      "step": 4305
    },
    {
      "epoch": 0.4653625851075327,
      "grad_norm": 0.4261700510978699,
      "learning_rate": 8e-05,
      "loss": 1.8445,
      "step": 4306
    },
    {
      "epoch": 0.4654706581649195,
      "grad_norm": 0.35033541917800903,
      "learning_rate": 8e-05,
      "loss": 1.3345,
      "step": 4307
    },
    {
      "epoch": 0.46557873122230625,
      "grad_norm": 0.4369705021381378,
      "learning_rate": 8e-05,
      "loss": 1.7823,
      "step": 4308
    },
    {
      "epoch": 0.46568680427969306,
      "grad_norm": 0.396161288022995,
      "learning_rate": 8e-05,
      "loss": 1.6919,
      "step": 4309
    },
    {
      "epoch": 0.46579487733707986,
      "grad_norm": 0.39318424463272095,
      "learning_rate": 8e-05,
      "loss": 1.687,
      "step": 4310
    },
    {
      "epoch": 0.46590295039446666,
      "grad_norm": 0.4040050506591797,
      "learning_rate": 8e-05,
      "loss": 1.6674,
      "step": 4311
    },
    {
      "epoch": 0.46601102345185347,
      "grad_norm": 0.3876594603061676,
      "learning_rate": 8e-05,
      "loss": 1.6904,
      "step": 4312
    },
    {
      "epoch": 0.46611909650924027,
      "grad_norm": 0.40077510476112366,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 4313
    },
    {
      "epoch": 0.466227169566627,
      "grad_norm": 0.3969590961933136,
      "learning_rate": 8e-05,
      "loss": 1.639,
      "step": 4314
    },
    {
      "epoch": 0.4663352426240138,
      "grad_norm": 0.38464590907096863,
      "learning_rate": 8e-05,
      "loss": 1.5865,
      "step": 4315
    },
    {
      "epoch": 0.4664433156814006,
      "grad_norm": 0.3839821517467499,
      "learning_rate": 8e-05,
      "loss": 1.7412,
      "step": 4316
    },
    {
      "epoch": 0.46655138873878743,
      "grad_norm": 0.4156714677810669,
      "learning_rate": 8e-05,
      "loss": 1.7886,
      "step": 4317
    },
    {
      "epoch": 0.46665946179617424,
      "grad_norm": 0.3865569531917572,
      "learning_rate": 8e-05,
      "loss": 1.6934,
      "step": 4318
    },
    {
      "epoch": 0.466767534853561,
      "grad_norm": 0.38035380840301514,
      "learning_rate": 8e-05,
      "loss": 1.5075,
      "step": 4319
    },
    {
      "epoch": 0.4668756079109478,
      "grad_norm": 0.3947802782058716,
      "learning_rate": 8e-05,
      "loss": 1.5321,
      "step": 4320
    },
    {
      "epoch": 0.4669836809683346,
      "grad_norm": 0.3801569938659668,
      "learning_rate": 8e-05,
      "loss": 1.7459,
      "step": 4321
    },
    {
      "epoch": 0.4670917540257214,
      "grad_norm": 0.42639559507369995,
      "learning_rate": 8e-05,
      "loss": 1.8251,
      "step": 4322
    },
    {
      "epoch": 0.4671998270831082,
      "grad_norm": 0.4697808027267456,
      "learning_rate": 8e-05,
      "loss": 1.7436,
      "step": 4323
    },
    {
      "epoch": 0.46730790014049495,
      "grad_norm": 0.39127108454704285,
      "learning_rate": 8e-05,
      "loss": 1.5958,
      "step": 4324
    },
    {
      "epoch": 0.46741597319788175,
      "grad_norm": 0.4393249750137329,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 4325
    },
    {
      "epoch": 0.46752404625526855,
      "grad_norm": 0.3974098265171051,
      "learning_rate": 8e-05,
      "loss": 1.7294,
      "step": 4326
    },
    {
      "epoch": 0.46763211931265536,
      "grad_norm": 0.4098997116088867,
      "learning_rate": 8e-05,
      "loss": 1.7576,
      "step": 4327
    },
    {
      "epoch": 0.46774019237004216,
      "grad_norm": 0.44012629985809326,
      "learning_rate": 8e-05,
      "loss": 1.8869,
      "step": 4328
    },
    {
      "epoch": 0.46784826542742897,
      "grad_norm": 0.43652862310409546,
      "learning_rate": 8e-05,
      "loss": 1.7634,
      "step": 4329
    },
    {
      "epoch": 0.4679563384848157,
      "grad_norm": 0.41287779808044434,
      "learning_rate": 8e-05,
      "loss": 1.8242,
      "step": 4330
    },
    {
      "epoch": 0.4680644115422025,
      "grad_norm": 0.4128740727901459,
      "learning_rate": 8e-05,
      "loss": 1.7326,
      "step": 4331
    },
    {
      "epoch": 0.4681724845995893,
      "grad_norm": 0.42176634073257446,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 4332
    },
    {
      "epoch": 0.4682805576569761,
      "grad_norm": 0.4075094759464264,
      "learning_rate": 8e-05,
      "loss": 1.7959,
      "step": 4333
    },
    {
      "epoch": 0.46838863071436293,
      "grad_norm": 0.38438481092453003,
      "learning_rate": 8e-05,
      "loss": 1.5871,
      "step": 4334
    },
    {
      "epoch": 0.4684967037717497,
      "grad_norm": 0.3875412046909332,
      "learning_rate": 8e-05,
      "loss": 1.6973,
      "step": 4335
    },
    {
      "epoch": 0.4686047768291365,
      "grad_norm": 0.41900065541267395,
      "learning_rate": 8e-05,
      "loss": 1.751,
      "step": 4336
    },
    {
      "epoch": 0.4687128498865233,
      "grad_norm": 0.38133275508880615,
      "learning_rate": 8e-05,
      "loss": 1.5882,
      "step": 4337
    },
    {
      "epoch": 0.4688209229439101,
      "grad_norm": 0.4026641845703125,
      "learning_rate": 8e-05,
      "loss": 1.795,
      "step": 4338
    },
    {
      "epoch": 0.4689289960012969,
      "grad_norm": 0.40954869985580444,
      "learning_rate": 8e-05,
      "loss": 1.8215,
      "step": 4339
    },
    {
      "epoch": 0.4690370690586837,
      "grad_norm": 0.4353555738925934,
      "learning_rate": 8e-05,
      "loss": 1.9223,
      "step": 4340
    },
    {
      "epoch": 0.46914514211607045,
      "grad_norm": 0.41750815510749817,
      "learning_rate": 8e-05,
      "loss": 1.7312,
      "step": 4341
    },
    {
      "epoch": 0.46925321517345725,
      "grad_norm": 0.5474597215652466,
      "learning_rate": 8e-05,
      "loss": 1.9644,
      "step": 4342
    },
    {
      "epoch": 0.46936128823084405,
      "grad_norm": 0.378177285194397,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 4343
    },
    {
      "epoch": 0.46946936128823086,
      "grad_norm": 0.4031790792942047,
      "learning_rate": 8e-05,
      "loss": 1.7007,
      "step": 4344
    },
    {
      "epoch": 0.46957743434561766,
      "grad_norm": 0.387382447719574,
      "learning_rate": 8e-05,
      "loss": 1.7539,
      "step": 4345
    },
    {
      "epoch": 0.4696855074030044,
      "grad_norm": 0.41637444496154785,
      "learning_rate": 8e-05,
      "loss": 1.6198,
      "step": 4346
    },
    {
      "epoch": 0.4697935804603912,
      "grad_norm": 0.427708238363266,
      "learning_rate": 8e-05,
      "loss": 1.8234,
      "step": 4347
    },
    {
      "epoch": 0.469901653517778,
      "grad_norm": 0.41436535120010376,
      "learning_rate": 8e-05,
      "loss": 1.9051,
      "step": 4348
    },
    {
      "epoch": 0.4700097265751648,
      "grad_norm": 0.3977288007736206,
      "learning_rate": 8e-05,
      "loss": 1.7416,
      "step": 4349
    },
    {
      "epoch": 0.4701177996325516,
      "grad_norm": 0.3888220191001892,
      "learning_rate": 8e-05,
      "loss": 1.6692,
      "step": 4350
    },
    {
      "epoch": 0.4702258726899384,
      "grad_norm": 0.41765326261520386,
      "learning_rate": 8e-05,
      "loss": 1.7674,
      "step": 4351
    },
    {
      "epoch": 0.4703339457473252,
      "grad_norm": 0.46067631244659424,
      "learning_rate": 8e-05,
      "loss": 1.9207,
      "step": 4352
    },
    {
      "epoch": 0.470442018804712,
      "grad_norm": 0.45710182189941406,
      "learning_rate": 8e-05,
      "loss": 1.8075,
      "step": 4353
    },
    {
      "epoch": 0.4705500918620988,
      "grad_norm": 0.4180136024951935,
      "learning_rate": 8e-05,
      "loss": 1.5251,
      "step": 4354
    },
    {
      "epoch": 0.4706581649194856,
      "grad_norm": 0.398973673582077,
      "learning_rate": 8e-05,
      "loss": 1.5873,
      "step": 4355
    },
    {
      "epoch": 0.4707662379768724,
      "grad_norm": 0.44040340185165405,
      "learning_rate": 8e-05,
      "loss": 1.8309,
      "step": 4356
    },
    {
      "epoch": 0.47087431103425914,
      "grad_norm": 0.39720261096954346,
      "learning_rate": 8e-05,
      "loss": 1.8151,
      "step": 4357
    },
    {
      "epoch": 0.47098238409164594,
      "grad_norm": 0.41815540194511414,
      "learning_rate": 8e-05,
      "loss": 1.5681,
      "step": 4358
    },
    {
      "epoch": 0.47109045714903275,
      "grad_norm": 0.40825673937797546,
      "learning_rate": 8e-05,
      "loss": 1.8415,
      "step": 4359
    },
    {
      "epoch": 0.47119853020641955,
      "grad_norm": 0.3943626582622528,
      "learning_rate": 8e-05,
      "loss": 1.664,
      "step": 4360
    },
    {
      "epoch": 0.47130660326380636,
      "grad_norm": 0.422489732503891,
      "learning_rate": 8e-05,
      "loss": 1.9152,
      "step": 4361
    },
    {
      "epoch": 0.4714146763211931,
      "grad_norm": 0.3940838575363159,
      "learning_rate": 8e-05,
      "loss": 1.6895,
      "step": 4362
    },
    {
      "epoch": 0.4715227493785799,
      "grad_norm": 0.4819503128528595,
      "learning_rate": 8e-05,
      "loss": 2.0226,
      "step": 4363
    },
    {
      "epoch": 0.4716308224359667,
      "grad_norm": 0.433308482170105,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 4364
    },
    {
      "epoch": 0.4717388954933535,
      "grad_norm": 0.3972829580307007,
      "learning_rate": 8e-05,
      "loss": 1.4199,
      "step": 4365
    },
    {
      "epoch": 0.4718469685507403,
      "grad_norm": 0.4517643451690674,
      "learning_rate": 8e-05,
      "loss": 1.6275,
      "step": 4366
    },
    {
      "epoch": 0.47195504160812707,
      "grad_norm": 0.49456995725631714,
      "learning_rate": 8e-05,
      "loss": 2.1072,
      "step": 4367
    },
    {
      "epoch": 0.47206311466551387,
      "grad_norm": 0.38539162278175354,
      "learning_rate": 8e-05,
      "loss": 1.6398,
      "step": 4368
    },
    {
      "epoch": 0.4721711877229007,
      "grad_norm": 0.4064953029155731,
      "learning_rate": 8e-05,
      "loss": 1.7952,
      "step": 4369
    },
    {
      "epoch": 0.4722792607802875,
      "grad_norm": 0.3970552086830139,
      "learning_rate": 8e-05,
      "loss": 1.6494,
      "step": 4370
    },
    {
      "epoch": 0.4723873338376743,
      "grad_norm": 0.4211111068725586,
      "learning_rate": 8e-05,
      "loss": 1.7976,
      "step": 4371
    },
    {
      "epoch": 0.4724954068950611,
      "grad_norm": 0.3946819603443146,
      "learning_rate": 8e-05,
      "loss": 1.6797,
      "step": 4372
    },
    {
      "epoch": 0.47260347995244784,
      "grad_norm": 0.483755886554718,
      "learning_rate": 8e-05,
      "loss": 1.9134,
      "step": 4373
    },
    {
      "epoch": 0.47271155300983464,
      "grad_norm": 0.3702860474586487,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 4374
    },
    {
      "epoch": 0.47281962606722144,
      "grad_norm": 0.4280916452407837,
      "learning_rate": 8e-05,
      "loss": 1.6624,
      "step": 4375
    },
    {
      "epoch": 0.47292769912460825,
      "grad_norm": 0.3854413330554962,
      "learning_rate": 8e-05,
      "loss": 1.6303,
      "step": 4376
    },
    {
      "epoch": 0.47303577218199505,
      "grad_norm": 0.41645359992980957,
      "learning_rate": 8e-05,
      "loss": 1.77,
      "step": 4377
    },
    {
      "epoch": 0.4731438452393818,
      "grad_norm": 0.4792899489402771,
      "learning_rate": 8e-05,
      "loss": 1.8368,
      "step": 4378
    },
    {
      "epoch": 0.4732519182967686,
      "grad_norm": 0.40461787581443787,
      "learning_rate": 8e-05,
      "loss": 1.769,
      "step": 4379
    },
    {
      "epoch": 0.4733599913541554,
      "grad_norm": 0.397129625082016,
      "learning_rate": 8e-05,
      "loss": 1.6811,
      "step": 4380
    },
    {
      "epoch": 0.4734680644115422,
      "grad_norm": 0.3850407302379608,
      "learning_rate": 8e-05,
      "loss": 1.675,
      "step": 4381
    },
    {
      "epoch": 0.473576137468929,
      "grad_norm": 0.39792346954345703,
      "learning_rate": 8e-05,
      "loss": 1.5907,
      "step": 4382
    },
    {
      "epoch": 0.47368421052631576,
      "grad_norm": 0.40356749296188354,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 4383
    },
    {
      "epoch": 0.47379228358370257,
      "grad_norm": 0.39915353059768677,
      "learning_rate": 8e-05,
      "loss": 1.8245,
      "step": 4384
    },
    {
      "epoch": 0.47390035664108937,
      "grad_norm": 0.42239752411842346,
      "learning_rate": 8e-05,
      "loss": 1.9393,
      "step": 4385
    },
    {
      "epoch": 0.4740084296984762,
      "grad_norm": 0.40600547194480896,
      "learning_rate": 8e-05,
      "loss": 1.7298,
      "step": 4386
    },
    {
      "epoch": 0.474116502755863,
      "grad_norm": 0.4127482771873474,
      "learning_rate": 8e-05,
      "loss": 1.7381,
      "step": 4387
    },
    {
      "epoch": 0.4742245758132498,
      "grad_norm": 0.40829917788505554,
      "learning_rate": 8e-05,
      "loss": 1.6669,
      "step": 4388
    },
    {
      "epoch": 0.47433264887063653,
      "grad_norm": 0.4510522782802582,
      "learning_rate": 8e-05,
      "loss": 1.9538,
      "step": 4389
    },
    {
      "epoch": 0.47444072192802333,
      "grad_norm": 0.43200254440307617,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 4390
    },
    {
      "epoch": 0.47454879498541014,
      "grad_norm": 0.38068196177482605,
      "learning_rate": 8e-05,
      "loss": 1.6835,
      "step": 4391
    },
    {
      "epoch": 0.47465686804279694,
      "grad_norm": 0.38059884309768677,
      "learning_rate": 8e-05,
      "loss": 1.7722,
      "step": 4392
    },
    {
      "epoch": 0.47476494110018375,
      "grad_norm": 0.4309141933917999,
      "learning_rate": 8e-05,
      "loss": 1.8183,
      "step": 4393
    },
    {
      "epoch": 0.4748730141575705,
      "grad_norm": 0.4085610806941986,
      "learning_rate": 8e-05,
      "loss": 1.6723,
      "step": 4394
    },
    {
      "epoch": 0.4749810872149573,
      "grad_norm": 0.3846578896045685,
      "learning_rate": 8e-05,
      "loss": 1.6585,
      "step": 4395
    },
    {
      "epoch": 0.4750891602723441,
      "grad_norm": 0.3819640278816223,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 4396
    },
    {
      "epoch": 0.4751972333297309,
      "grad_norm": 0.39420387148857117,
      "learning_rate": 8e-05,
      "loss": 1.6839,
      "step": 4397
    },
    {
      "epoch": 0.4753053063871177,
      "grad_norm": 0.40593892335891724,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 4398
    },
    {
      "epoch": 0.47541337944450446,
      "grad_norm": 0.41556963324546814,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 4399
    },
    {
      "epoch": 0.47552145250189126,
      "grad_norm": 0.41296055912971497,
      "learning_rate": 8e-05,
      "loss": 1.778,
      "step": 4400
    },
    {
      "epoch": 0.47562952555927807,
      "grad_norm": 0.3912244439125061,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 4401
    },
    {
      "epoch": 0.47573759861666487,
      "grad_norm": 0.3882213532924652,
      "learning_rate": 8e-05,
      "loss": 1.7627,
      "step": 4402
    },
    {
      "epoch": 0.4758456716740517,
      "grad_norm": 0.4012654721736908,
      "learning_rate": 8e-05,
      "loss": 1.7493,
      "step": 4403
    },
    {
      "epoch": 0.4759537447314385,
      "grad_norm": 0.406279057264328,
      "learning_rate": 8e-05,
      "loss": 1.8219,
      "step": 4404
    },
    {
      "epoch": 0.4760618177888252,
      "grad_norm": 0.3669048547744751,
      "learning_rate": 8e-05,
      "loss": 1.6331,
      "step": 4405
    },
    {
      "epoch": 0.47616989084621203,
      "grad_norm": 0.4104841947555542,
      "learning_rate": 8e-05,
      "loss": 1.7244,
      "step": 4406
    },
    {
      "epoch": 0.47627796390359883,
      "grad_norm": 0.4692522883415222,
      "learning_rate": 8e-05,
      "loss": 2.0614,
      "step": 4407
    },
    {
      "epoch": 0.47638603696098564,
      "grad_norm": 0.39436599612236023,
      "learning_rate": 8e-05,
      "loss": 1.6795,
      "step": 4408
    },
    {
      "epoch": 0.47649411001837244,
      "grad_norm": 0.38139328360557556,
      "learning_rate": 8e-05,
      "loss": 1.7118,
      "step": 4409
    },
    {
      "epoch": 0.4766021830757592,
      "grad_norm": 0.3653238117694855,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 4410
    },
    {
      "epoch": 0.476710256133146,
      "grad_norm": 0.47773492336273193,
      "learning_rate": 8e-05,
      "loss": 1.8691,
      "step": 4411
    },
    {
      "epoch": 0.4768183291905328,
      "grad_norm": 0.41180169582366943,
      "learning_rate": 8e-05,
      "loss": 1.5012,
      "step": 4412
    },
    {
      "epoch": 0.4769264022479196,
      "grad_norm": 0.4035488963127136,
      "learning_rate": 8e-05,
      "loss": 1.6536,
      "step": 4413
    },
    {
      "epoch": 0.4770344753053064,
      "grad_norm": 0.4330061674118042,
      "learning_rate": 8e-05,
      "loss": 1.6203,
      "step": 4414
    },
    {
      "epoch": 0.47714254836269315,
      "grad_norm": 0.4134182035923004,
      "learning_rate": 8e-05,
      "loss": 1.7122,
      "step": 4415
    },
    {
      "epoch": 0.47725062142007996,
      "grad_norm": 0.4109426438808441,
      "learning_rate": 8e-05,
      "loss": 1.816,
      "step": 4416
    },
    {
      "epoch": 0.47735869447746676,
      "grad_norm": 0.3879695534706116,
      "learning_rate": 8e-05,
      "loss": 1.5841,
      "step": 4417
    },
    {
      "epoch": 0.47746676753485356,
      "grad_norm": 0.3969479501247406,
      "learning_rate": 8e-05,
      "loss": 1.7279,
      "step": 4418
    },
    {
      "epoch": 0.47757484059224037,
      "grad_norm": 0.3923637270927429,
      "learning_rate": 8e-05,
      "loss": 1.5371,
      "step": 4419
    },
    {
      "epoch": 0.47768291364962717,
      "grad_norm": 0.37929192185401917,
      "learning_rate": 8e-05,
      "loss": 1.6564,
      "step": 4420
    },
    {
      "epoch": 0.4777909867070139,
      "grad_norm": 0.393903523683548,
      "learning_rate": 8e-05,
      "loss": 1.7322,
      "step": 4421
    },
    {
      "epoch": 0.4778990597644007,
      "grad_norm": 0.3977348506450653,
      "learning_rate": 8e-05,
      "loss": 1.5666,
      "step": 4422
    },
    {
      "epoch": 0.47800713282178753,
      "grad_norm": 0.4265556037425995,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 4423
    },
    {
      "epoch": 0.47811520587917433,
      "grad_norm": 0.4359314739704132,
      "learning_rate": 8e-05,
      "loss": 1.805,
      "step": 4424
    },
    {
      "epoch": 0.47822327893656114,
      "grad_norm": 0.4290197193622589,
      "learning_rate": 8e-05,
      "loss": 1.9265,
      "step": 4425
    },
    {
      "epoch": 0.4783313519939479,
      "grad_norm": 0.4240277409553528,
      "learning_rate": 8e-05,
      "loss": 1.7292,
      "step": 4426
    },
    {
      "epoch": 0.4784394250513347,
      "grad_norm": 0.42168569564819336,
      "learning_rate": 8e-05,
      "loss": 1.8055,
      "step": 4427
    },
    {
      "epoch": 0.4785474981087215,
      "grad_norm": 0.4279271066188812,
      "learning_rate": 8e-05,
      "loss": 1.6262,
      "step": 4428
    },
    {
      "epoch": 0.4786555711661083,
      "grad_norm": 0.38093265891075134,
      "learning_rate": 8e-05,
      "loss": 1.5498,
      "step": 4429
    },
    {
      "epoch": 0.4787636442234951,
      "grad_norm": 0.43716198205947876,
      "learning_rate": 8e-05,
      "loss": 1.8098,
      "step": 4430
    },
    {
      "epoch": 0.4788717172808819,
      "grad_norm": 0.40469709038734436,
      "learning_rate": 8e-05,
      "loss": 1.609,
      "step": 4431
    },
    {
      "epoch": 0.47897979033826865,
      "grad_norm": 0.4298054277896881,
      "learning_rate": 8e-05,
      "loss": 1.7464,
      "step": 4432
    },
    {
      "epoch": 0.47908786339565546,
      "grad_norm": 0.406124085187912,
      "learning_rate": 8e-05,
      "loss": 1.5323,
      "step": 4433
    },
    {
      "epoch": 0.47919593645304226,
      "grad_norm": 0.39833322167396545,
      "learning_rate": 8e-05,
      "loss": 1.7075,
      "step": 4434
    },
    {
      "epoch": 0.47930400951042906,
      "grad_norm": 0.4070724546909332,
      "learning_rate": 8e-05,
      "loss": 1.8449,
      "step": 4435
    },
    {
      "epoch": 0.47941208256781587,
      "grad_norm": 0.4110445976257324,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 4436
    },
    {
      "epoch": 0.4795201556252026,
      "grad_norm": 0.44566917419433594,
      "learning_rate": 8e-05,
      "loss": 1.7021,
      "step": 4437
    },
    {
      "epoch": 0.4796282286825894,
      "grad_norm": 0.41855207085609436,
      "learning_rate": 8e-05,
      "loss": 1.6947,
      "step": 4438
    },
    {
      "epoch": 0.4797363017399762,
      "grad_norm": 0.39597100019454956,
      "learning_rate": 8e-05,
      "loss": 1.654,
      "step": 4439
    },
    {
      "epoch": 0.479844374797363,
      "grad_norm": 0.3993816673755646,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 4440
    },
    {
      "epoch": 0.47995244785474983,
      "grad_norm": 0.37629780173301697,
      "learning_rate": 8e-05,
      "loss": 1.5556,
      "step": 4441
    },
    {
      "epoch": 0.4800605209121366,
      "grad_norm": 0.4099539816379547,
      "learning_rate": 8e-05,
      "loss": 1.8828,
      "step": 4442
    },
    {
      "epoch": 0.4801685939695234,
      "grad_norm": 0.3883780241012573,
      "learning_rate": 8e-05,
      "loss": 1.7729,
      "step": 4443
    },
    {
      "epoch": 0.4802766670269102,
      "grad_norm": 0.3549973964691162,
      "learning_rate": 8e-05,
      "loss": 1.6145,
      "step": 4444
    },
    {
      "epoch": 0.480384740084297,
      "grad_norm": 0.42205509543418884,
      "learning_rate": 8e-05,
      "loss": 1.7368,
      "step": 4445
    },
    {
      "epoch": 0.4804928131416838,
      "grad_norm": 0.4315900206565857,
      "learning_rate": 8e-05,
      "loss": 1.8262,
      "step": 4446
    },
    {
      "epoch": 0.4806008861990706,
      "grad_norm": 0.38961830735206604,
      "learning_rate": 8e-05,
      "loss": 1.6552,
      "step": 4447
    },
    {
      "epoch": 0.48070895925645735,
      "grad_norm": 0.40467625856399536,
      "learning_rate": 8e-05,
      "loss": 1.7665,
      "step": 4448
    },
    {
      "epoch": 0.48081703231384415,
      "grad_norm": 0.4513581693172455,
      "learning_rate": 8e-05,
      "loss": 1.7646,
      "step": 4449
    },
    {
      "epoch": 0.48092510537123095,
      "grad_norm": 0.4321550130844116,
      "learning_rate": 8e-05,
      "loss": 1.5715,
      "step": 4450
    },
    {
      "epoch": 0.48103317842861776,
      "grad_norm": 0.42358481884002686,
      "learning_rate": 8e-05,
      "loss": 1.6498,
      "step": 4451
    },
    {
      "epoch": 0.48114125148600456,
      "grad_norm": 0.461465448141098,
      "learning_rate": 8e-05,
      "loss": 1.8365,
      "step": 4452
    },
    {
      "epoch": 0.4812493245433913,
      "grad_norm": 0.429018497467041,
      "learning_rate": 8e-05,
      "loss": 1.8486,
      "step": 4453
    },
    {
      "epoch": 0.4813573976007781,
      "grad_norm": 0.410231351852417,
      "learning_rate": 8e-05,
      "loss": 1.7162,
      "step": 4454
    },
    {
      "epoch": 0.4814654706581649,
      "grad_norm": 0.4063430428504944,
      "learning_rate": 8e-05,
      "loss": 1.6384,
      "step": 4455
    },
    {
      "epoch": 0.4815735437155517,
      "grad_norm": 0.42961350083351135,
      "learning_rate": 8e-05,
      "loss": 1.6931,
      "step": 4456
    },
    {
      "epoch": 0.4816816167729385,
      "grad_norm": 0.44866907596588135,
      "learning_rate": 8e-05,
      "loss": 1.9183,
      "step": 4457
    },
    {
      "epoch": 0.4817896898303253,
      "grad_norm": 0.42478424310684204,
      "learning_rate": 8e-05,
      "loss": 1.7056,
      "step": 4458
    },
    {
      "epoch": 0.4818977628877121,
      "grad_norm": 0.42392247915267944,
      "learning_rate": 8e-05,
      "loss": 1.6186,
      "step": 4459
    },
    {
      "epoch": 0.4820058359450989,
      "grad_norm": 0.3868848979473114,
      "learning_rate": 8e-05,
      "loss": 1.7845,
      "step": 4460
    },
    {
      "epoch": 0.4821139090024857,
      "grad_norm": 0.426983118057251,
      "learning_rate": 8e-05,
      "loss": 1.7918,
      "step": 4461
    },
    {
      "epoch": 0.4822219820598725,
      "grad_norm": 0.4045998752117157,
      "learning_rate": 8e-05,
      "loss": 1.6476,
      "step": 4462
    },
    {
      "epoch": 0.4823300551172593,
      "grad_norm": 0.37942007184028625,
      "learning_rate": 8e-05,
      "loss": 1.7946,
      "step": 4463
    },
    {
      "epoch": 0.48243812817464604,
      "grad_norm": 0.4016372263431549,
      "learning_rate": 8e-05,
      "loss": 1.7546,
      "step": 4464
    },
    {
      "epoch": 0.48254620123203285,
      "grad_norm": 0.3995134234428406,
      "learning_rate": 8e-05,
      "loss": 1.7306,
      "step": 4465
    },
    {
      "epoch": 0.48265427428941965,
      "grad_norm": 0.3755732476711273,
      "learning_rate": 8e-05,
      "loss": 1.6472,
      "step": 4466
    },
    {
      "epoch": 0.48276234734680645,
      "grad_norm": 0.4377599358558655,
      "learning_rate": 8e-05,
      "loss": 1.7974,
      "step": 4467
    },
    {
      "epoch": 0.48287042040419326,
      "grad_norm": 0.4171310067176819,
      "learning_rate": 8e-05,
      "loss": 1.8031,
      "step": 4468
    },
    {
      "epoch": 0.48297849346158,
      "grad_norm": 0.38294002413749695,
      "learning_rate": 8e-05,
      "loss": 1.5684,
      "step": 4469
    },
    {
      "epoch": 0.4830865665189668,
      "grad_norm": 0.4068509638309479,
      "learning_rate": 8e-05,
      "loss": 1.6412,
      "step": 4470
    },
    {
      "epoch": 0.4831946395763536,
      "grad_norm": 0.3854493200778961,
      "learning_rate": 8e-05,
      "loss": 1.5587,
      "step": 4471
    },
    {
      "epoch": 0.4833027126337404,
      "grad_norm": 0.42035725712776184,
      "learning_rate": 8e-05,
      "loss": 1.8951,
      "step": 4472
    },
    {
      "epoch": 0.4834107856911272,
      "grad_norm": 0.4494016468524933,
      "learning_rate": 8e-05,
      "loss": 1.7786,
      "step": 4473
    },
    {
      "epoch": 0.48351885874851397,
      "grad_norm": 0.38851773738861084,
      "learning_rate": 8e-05,
      "loss": 1.6645,
      "step": 4474
    },
    {
      "epoch": 0.4836269318059008,
      "grad_norm": 0.3658621311187744,
      "learning_rate": 8e-05,
      "loss": 1.5165,
      "step": 4475
    },
    {
      "epoch": 0.4837350048632876,
      "grad_norm": 0.3905554413795471,
      "learning_rate": 8e-05,
      "loss": 1.6728,
      "step": 4476
    },
    {
      "epoch": 0.4838430779206744,
      "grad_norm": 0.4269133508205414,
      "learning_rate": 8e-05,
      "loss": 1.611,
      "step": 4477
    },
    {
      "epoch": 0.4839511509780612,
      "grad_norm": 0.4251437783241272,
      "learning_rate": 8e-05,
      "loss": 1.8256,
      "step": 4478
    },
    {
      "epoch": 0.484059224035448,
      "grad_norm": 0.3766750395298004,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 4479
    },
    {
      "epoch": 0.48416729709283474,
      "grad_norm": 0.4373146891593933,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 4480
    },
    {
      "epoch": 0.48427537015022154,
      "grad_norm": 0.4068273901939392,
      "learning_rate": 8e-05,
      "loss": 1.6856,
      "step": 4481
    },
    {
      "epoch": 0.48438344320760834,
      "grad_norm": 0.38094136118888855,
      "learning_rate": 8e-05,
      "loss": 1.6455,
      "step": 4482
    },
    {
      "epoch": 0.48449151626499515,
      "grad_norm": 0.4291408658027649,
      "learning_rate": 8e-05,
      "loss": 1.8316,
      "step": 4483
    },
    {
      "epoch": 0.48459958932238195,
      "grad_norm": 0.38708606362342834,
      "learning_rate": 8e-05,
      "loss": 1.5471,
      "step": 4484
    },
    {
      "epoch": 0.4847076623797687,
      "grad_norm": 0.4210904538631439,
      "learning_rate": 8e-05,
      "loss": 1.7247,
      "step": 4485
    },
    {
      "epoch": 0.4848157354371555,
      "grad_norm": 0.40632113814353943,
      "learning_rate": 8e-05,
      "loss": 1.7915,
      "step": 4486
    },
    {
      "epoch": 0.4849238084945423,
      "grad_norm": 0.41077303886413574,
      "learning_rate": 8e-05,
      "loss": 1.725,
      "step": 4487
    },
    {
      "epoch": 0.4850318815519291,
      "grad_norm": 0.4161292612552643,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 4488
    },
    {
      "epoch": 0.4851399546093159,
      "grad_norm": 0.39576563239097595,
      "learning_rate": 8e-05,
      "loss": 1.6917,
      "step": 4489
    },
    {
      "epoch": 0.48524802766670266,
      "grad_norm": 0.4352361261844635,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 4490
    },
    {
      "epoch": 0.48535610072408947,
      "grad_norm": 0.3711824119091034,
      "learning_rate": 8e-05,
      "loss": 1.6364,
      "step": 4491
    },
    {
      "epoch": 0.48546417378147627,
      "grad_norm": 0.4126344621181488,
      "learning_rate": 8e-05,
      "loss": 1.7333,
      "step": 4492
    },
    {
      "epoch": 0.4855722468388631,
      "grad_norm": 0.41216039657592773,
      "learning_rate": 8e-05,
      "loss": 1.7926,
      "step": 4493
    },
    {
      "epoch": 0.4856803198962499,
      "grad_norm": 0.4065435528755188,
      "learning_rate": 8e-05,
      "loss": 1.7913,
      "step": 4494
    },
    {
      "epoch": 0.4857883929536367,
      "grad_norm": 0.3982198238372803,
      "learning_rate": 8e-05,
      "loss": 1.6105,
      "step": 4495
    },
    {
      "epoch": 0.48589646601102343,
      "grad_norm": 0.48088499903678894,
      "learning_rate": 8e-05,
      "loss": 2.035,
      "step": 4496
    },
    {
      "epoch": 0.48600453906841024,
      "grad_norm": 0.4165959656238556,
      "learning_rate": 8e-05,
      "loss": 1.6636,
      "step": 4497
    },
    {
      "epoch": 0.48611261212579704,
      "grad_norm": 0.40088269114494324,
      "learning_rate": 8e-05,
      "loss": 1.5413,
      "step": 4498
    },
    {
      "epoch": 0.48622068518318384,
      "grad_norm": 0.3928964138031006,
      "learning_rate": 8e-05,
      "loss": 1.5888,
      "step": 4499
    },
    {
      "epoch": 0.48632875824057065,
      "grad_norm": 0.3858298361301422,
      "learning_rate": 8e-05,
      "loss": 1.6318,
      "step": 4500
    },
    {
      "epoch": 0.4864368312979574,
      "grad_norm": 0.3981149196624756,
      "learning_rate": 8e-05,
      "loss": 1.7318,
      "step": 4501
    },
    {
      "epoch": 0.4865449043553442,
      "grad_norm": 0.4077165722846985,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 4502
    },
    {
      "epoch": 0.486652977412731,
      "grad_norm": 0.4125361740589142,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 4503
    },
    {
      "epoch": 0.4867610504701178,
      "grad_norm": 0.4113052189350128,
      "learning_rate": 8e-05,
      "loss": 1.6128,
      "step": 4504
    },
    {
      "epoch": 0.4868691235275046,
      "grad_norm": 0.4195830225944519,
      "learning_rate": 8e-05,
      "loss": 1.8256,
      "step": 4505
    },
    {
      "epoch": 0.4869771965848914,
      "grad_norm": 0.40131568908691406,
      "learning_rate": 8e-05,
      "loss": 1.7544,
      "step": 4506
    },
    {
      "epoch": 0.48708526964227816,
      "grad_norm": 0.4021151065826416,
      "learning_rate": 8e-05,
      "loss": 1.5619,
      "step": 4507
    },
    {
      "epoch": 0.48719334269966497,
      "grad_norm": 0.3786599934101105,
      "learning_rate": 8e-05,
      "loss": 1.6274,
      "step": 4508
    },
    {
      "epoch": 0.48730141575705177,
      "grad_norm": 0.39171236753463745,
      "learning_rate": 8e-05,
      "loss": 1.6944,
      "step": 4509
    },
    {
      "epoch": 0.4874094888144386,
      "grad_norm": 0.3862670958042145,
      "learning_rate": 8e-05,
      "loss": 1.5794,
      "step": 4510
    },
    {
      "epoch": 0.4875175618718254,
      "grad_norm": 0.3992129862308502,
      "learning_rate": 8e-05,
      "loss": 1.6675,
      "step": 4511
    },
    {
      "epoch": 0.4876256349292121,
      "grad_norm": 0.39308929443359375,
      "learning_rate": 8e-05,
      "loss": 1.7396,
      "step": 4512
    },
    {
      "epoch": 0.48773370798659893,
      "grad_norm": 0.39373594522476196,
      "learning_rate": 8e-05,
      "loss": 1.6743,
      "step": 4513
    },
    {
      "epoch": 0.48784178104398573,
      "grad_norm": 0.39826205372810364,
      "learning_rate": 8e-05,
      "loss": 1.6196,
      "step": 4514
    },
    {
      "epoch": 0.48794985410137254,
      "grad_norm": 0.4778093695640564,
      "learning_rate": 8e-05,
      "loss": 1.7328,
      "step": 4515
    },
    {
      "epoch": 0.48805792715875934,
      "grad_norm": 0.4264270067214966,
      "learning_rate": 8e-05,
      "loss": 1.8237,
      "step": 4516
    },
    {
      "epoch": 0.4881660002161461,
      "grad_norm": 0.39952999353408813,
      "learning_rate": 8e-05,
      "loss": 1.6207,
      "step": 4517
    },
    {
      "epoch": 0.4882740732735329,
      "grad_norm": 0.3912876546382904,
      "learning_rate": 8e-05,
      "loss": 1.531,
      "step": 4518
    },
    {
      "epoch": 0.4883821463309197,
      "grad_norm": 0.4450731873512268,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 4519
    },
    {
      "epoch": 0.4884902193883065,
      "grad_norm": 0.4129990041255951,
      "learning_rate": 8e-05,
      "loss": 1.873,
      "step": 4520
    },
    {
      "epoch": 0.4885982924456933,
      "grad_norm": 0.38996970653533936,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 4521
    },
    {
      "epoch": 0.4887063655030801,
      "grad_norm": 0.41124382615089417,
      "learning_rate": 8e-05,
      "loss": 1.736,
      "step": 4522
    },
    {
      "epoch": 0.48881443856046686,
      "grad_norm": 0.4648588001728058,
      "learning_rate": 8e-05,
      "loss": 1.8285,
      "step": 4523
    },
    {
      "epoch": 0.48892251161785366,
      "grad_norm": 0.47971948981285095,
      "learning_rate": 8e-05,
      "loss": 1.8545,
      "step": 4524
    },
    {
      "epoch": 0.48903058467524047,
      "grad_norm": 0.40587490797042847,
      "learning_rate": 8e-05,
      "loss": 1.7316,
      "step": 4525
    },
    {
      "epoch": 0.48913865773262727,
      "grad_norm": 0.43597742915153503,
      "learning_rate": 8e-05,
      "loss": 1.7261,
      "step": 4526
    },
    {
      "epoch": 0.4892467307900141,
      "grad_norm": 0.4638703763484955,
      "learning_rate": 8e-05,
      "loss": 1.6373,
      "step": 4527
    },
    {
      "epoch": 0.4893548038474008,
      "grad_norm": 0.42312783002853394,
      "learning_rate": 8e-05,
      "loss": 1.4575,
      "step": 4528
    },
    {
      "epoch": 0.4894628769047876,
      "grad_norm": 0.39222845435142517,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 4529
    },
    {
      "epoch": 0.48957094996217443,
      "grad_norm": 0.4099057614803314,
      "learning_rate": 8e-05,
      "loss": 1.6386,
      "step": 4530
    },
    {
      "epoch": 0.48967902301956123,
      "grad_norm": 0.406204491853714,
      "learning_rate": 8e-05,
      "loss": 1.6222,
      "step": 4531
    },
    {
      "epoch": 0.48978709607694804,
      "grad_norm": 0.49371662735939026,
      "learning_rate": 8e-05,
      "loss": 1.8274,
      "step": 4532
    },
    {
      "epoch": 0.4898951691343348,
      "grad_norm": 0.4220612347126007,
      "learning_rate": 8e-05,
      "loss": 1.9628,
      "step": 4533
    },
    {
      "epoch": 0.4900032421917216,
      "grad_norm": 0.3846607506275177,
      "learning_rate": 8e-05,
      "loss": 1.6548,
      "step": 4534
    },
    {
      "epoch": 0.4901113152491084,
      "grad_norm": 0.4358482360839844,
      "learning_rate": 8e-05,
      "loss": 1.6209,
      "step": 4535
    },
    {
      "epoch": 0.4902193883064952,
      "grad_norm": 0.4412643313407898,
      "learning_rate": 8e-05,
      "loss": 1.6308,
      "step": 4536
    },
    {
      "epoch": 0.490327461363882,
      "grad_norm": 0.4089067280292511,
      "learning_rate": 8e-05,
      "loss": 1.6897,
      "step": 4537
    },
    {
      "epoch": 0.4904355344212688,
      "grad_norm": 0.4318627417087555,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 4538
    },
    {
      "epoch": 0.49054360747865555,
      "grad_norm": 0.41721653938293457,
      "learning_rate": 8e-05,
      "loss": 1.5444,
      "step": 4539
    },
    {
      "epoch": 0.49065168053604236,
      "grad_norm": 0.47165438532829285,
      "learning_rate": 8e-05,
      "loss": 1.7536,
      "step": 4540
    },
    {
      "epoch": 0.49075975359342916,
      "grad_norm": 0.47693413496017456,
      "learning_rate": 8e-05,
      "loss": 1.6817,
      "step": 4541
    },
    {
      "epoch": 0.49086782665081596,
      "grad_norm": 0.3976629972457886,
      "learning_rate": 8e-05,
      "loss": 1.7442,
      "step": 4542
    },
    {
      "epoch": 0.49097589970820277,
      "grad_norm": 0.3838636875152588,
      "learning_rate": 8e-05,
      "loss": 1.6901,
      "step": 4543
    },
    {
      "epoch": 0.4910839727655895,
      "grad_norm": 0.4036515951156616,
      "learning_rate": 8e-05,
      "loss": 1.6334,
      "step": 4544
    },
    {
      "epoch": 0.4911920458229763,
      "grad_norm": 0.4530039429664612,
      "learning_rate": 8e-05,
      "loss": 1.8745,
      "step": 4545
    },
    {
      "epoch": 0.4913001188803631,
      "grad_norm": 0.39521896839141846,
      "learning_rate": 8e-05,
      "loss": 1.7036,
      "step": 4546
    },
    {
      "epoch": 0.4914081919377499,
      "grad_norm": 0.3843677341938019,
      "learning_rate": 8e-05,
      "loss": 1.69,
      "step": 4547
    },
    {
      "epoch": 0.49151626499513673,
      "grad_norm": 0.41856613755226135,
      "learning_rate": 8e-05,
      "loss": 1.7805,
      "step": 4548
    },
    {
      "epoch": 0.4916243380525235,
      "grad_norm": 0.3842417895793915,
      "learning_rate": 8e-05,
      "loss": 1.6026,
      "step": 4549
    },
    {
      "epoch": 0.4917324111099103,
      "grad_norm": 0.41573432087898254,
      "learning_rate": 8e-05,
      "loss": 1.7438,
      "step": 4550
    },
    {
      "epoch": 0.4918404841672971,
      "grad_norm": 0.4763418734073639,
      "learning_rate": 8e-05,
      "loss": 1.8534,
      "step": 4551
    },
    {
      "epoch": 0.4919485572246839,
      "grad_norm": 0.4143596291542053,
      "learning_rate": 8e-05,
      "loss": 1.7213,
      "step": 4552
    },
    {
      "epoch": 0.4920566302820707,
      "grad_norm": 0.43058741092681885,
      "learning_rate": 8e-05,
      "loss": 1.7569,
      "step": 4553
    },
    {
      "epoch": 0.4921647033394575,
      "grad_norm": 0.39957430958747864,
      "learning_rate": 8e-05,
      "loss": 1.5292,
      "step": 4554
    },
    {
      "epoch": 0.49227277639684425,
      "grad_norm": 0.44477221369743347,
      "learning_rate": 8e-05,
      "loss": 1.8355,
      "step": 4555
    },
    {
      "epoch": 0.49238084945423105,
      "grad_norm": 0.42947760224342346,
      "learning_rate": 8e-05,
      "loss": 1.6015,
      "step": 4556
    },
    {
      "epoch": 0.49248892251161785,
      "grad_norm": 0.3672034740447998,
      "learning_rate": 8e-05,
      "loss": 1.6425,
      "step": 4557
    },
    {
      "epoch": 0.49259699556900466,
      "grad_norm": 0.3806307315826416,
      "learning_rate": 8e-05,
      "loss": 1.6863,
      "step": 4558
    },
    {
      "epoch": 0.49270506862639146,
      "grad_norm": 0.3946504294872284,
      "learning_rate": 8e-05,
      "loss": 1.4487,
      "step": 4559
    },
    {
      "epoch": 0.4928131416837782,
      "grad_norm": 0.42138540744781494,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 4560
    },
    {
      "epoch": 0.492921214741165,
      "grad_norm": 0.4102710485458374,
      "learning_rate": 8e-05,
      "loss": 1.7642,
      "step": 4561
    },
    {
      "epoch": 0.4930292877985518,
      "grad_norm": 0.42016077041625977,
      "learning_rate": 8e-05,
      "loss": 1.7404,
      "step": 4562
    },
    {
      "epoch": 0.4931373608559386,
      "grad_norm": 0.38941633701324463,
      "learning_rate": 8e-05,
      "loss": 1.6138,
      "step": 4563
    },
    {
      "epoch": 0.4932454339133254,
      "grad_norm": 0.4074150621891022,
      "learning_rate": 8e-05,
      "loss": 1.6843,
      "step": 4564
    },
    {
      "epoch": 0.4933535069707122,
      "grad_norm": 0.41428184509277344,
      "learning_rate": 8e-05,
      "loss": 1.8069,
      "step": 4565
    },
    {
      "epoch": 0.493461580028099,
      "grad_norm": 0.4355214536190033,
      "learning_rate": 8e-05,
      "loss": 1.8362,
      "step": 4566
    },
    {
      "epoch": 0.4935696530854858,
      "grad_norm": 0.42752811312675476,
      "learning_rate": 8e-05,
      "loss": 1.8528,
      "step": 4567
    },
    {
      "epoch": 0.4936777261428726,
      "grad_norm": 0.3998211920261383,
      "learning_rate": 8e-05,
      "loss": 1.6667,
      "step": 4568
    },
    {
      "epoch": 0.4937857992002594,
      "grad_norm": 0.42924392223358154,
      "learning_rate": 8e-05,
      "loss": 1.8569,
      "step": 4569
    },
    {
      "epoch": 0.4938938722576462,
      "grad_norm": 0.38225632905960083,
      "learning_rate": 8e-05,
      "loss": 1.6995,
      "step": 4570
    },
    {
      "epoch": 0.49400194531503294,
      "grad_norm": 0.4094962775707245,
      "learning_rate": 8e-05,
      "loss": 1.6946,
      "step": 4571
    },
    {
      "epoch": 0.49411001837241975,
      "grad_norm": 0.38786569237709045,
      "learning_rate": 8e-05,
      "loss": 1.7502,
      "step": 4572
    },
    {
      "epoch": 0.49421809142980655,
      "grad_norm": 0.39534568786621094,
      "learning_rate": 8e-05,
      "loss": 1.7299,
      "step": 4573
    },
    {
      "epoch": 0.49432616448719335,
      "grad_norm": 0.44906845688819885,
      "learning_rate": 8e-05,
      "loss": 1.858,
      "step": 4574
    },
    {
      "epoch": 0.49443423754458016,
      "grad_norm": 0.42588010430336,
      "learning_rate": 8e-05,
      "loss": 1.6872,
      "step": 4575
    },
    {
      "epoch": 0.4945423106019669,
      "grad_norm": 0.39031171798706055,
      "learning_rate": 8e-05,
      "loss": 1.6644,
      "step": 4576
    },
    {
      "epoch": 0.4946503836593537,
      "grad_norm": 0.3910892605781555,
      "learning_rate": 8e-05,
      "loss": 1.6419,
      "step": 4577
    },
    {
      "epoch": 0.4947584567167405,
      "grad_norm": 0.4057159125804901,
      "learning_rate": 8e-05,
      "loss": 1.7696,
      "step": 4578
    },
    {
      "epoch": 0.4948665297741273,
      "grad_norm": 0.3953483998775482,
      "learning_rate": 8e-05,
      "loss": 1.6734,
      "step": 4579
    },
    {
      "epoch": 0.4949746028315141,
      "grad_norm": 0.4400515556335449,
      "learning_rate": 8e-05,
      "loss": 1.9922,
      "step": 4580
    },
    {
      "epoch": 0.49508267588890087,
      "grad_norm": 0.4748576581478119,
      "learning_rate": 8e-05,
      "loss": 1.8397,
      "step": 4581
    },
    {
      "epoch": 0.4951907489462877,
      "grad_norm": 0.4138704538345337,
      "learning_rate": 8e-05,
      "loss": 1.7375,
      "step": 4582
    },
    {
      "epoch": 0.4952988220036745,
      "grad_norm": 0.43630218505859375,
      "learning_rate": 8e-05,
      "loss": 1.7078,
      "step": 4583
    },
    {
      "epoch": 0.4954068950610613,
      "grad_norm": 0.41870805621147156,
      "learning_rate": 8e-05,
      "loss": 1.6677,
      "step": 4584
    },
    {
      "epoch": 0.4955149681184481,
      "grad_norm": 0.4615815579891205,
      "learning_rate": 8e-05,
      "loss": 1.8456,
      "step": 4585
    },
    {
      "epoch": 0.4956230411758349,
      "grad_norm": 0.4198322594165802,
      "learning_rate": 8e-05,
      "loss": 1.6839,
      "step": 4586
    },
    {
      "epoch": 0.49573111423322164,
      "grad_norm": 0.45698797702789307,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 4587
    },
    {
      "epoch": 0.49583918729060844,
      "grad_norm": 0.426725834608078,
      "learning_rate": 8e-05,
      "loss": 1.6433,
      "step": 4588
    },
    {
      "epoch": 0.49594726034799524,
      "grad_norm": 0.4246196150779724,
      "learning_rate": 8e-05,
      "loss": 1.847,
      "step": 4589
    },
    {
      "epoch": 0.49605533340538205,
      "grad_norm": 0.3808932602405548,
      "learning_rate": 8e-05,
      "loss": 1.6011,
      "step": 4590
    },
    {
      "epoch": 0.49616340646276885,
      "grad_norm": 0.44499140977859497,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 4591
    },
    {
      "epoch": 0.4962714795201556,
      "grad_norm": 0.4330929219722748,
      "learning_rate": 8e-05,
      "loss": 1.6546,
      "step": 4592
    },
    {
      "epoch": 0.4963795525775424,
      "grad_norm": 0.41983404755592346,
      "learning_rate": 8e-05,
      "loss": 1.5631,
      "step": 4593
    },
    {
      "epoch": 0.4964876256349292,
      "grad_norm": 0.4523801803588867,
      "learning_rate": 8e-05,
      "loss": 1.8166,
      "step": 4594
    },
    {
      "epoch": 0.496595698692316,
      "grad_norm": 0.4522763788700104,
      "learning_rate": 8e-05,
      "loss": 1.7319,
      "step": 4595
    },
    {
      "epoch": 0.4967037717497028,
      "grad_norm": 0.44319260120391846,
      "learning_rate": 8e-05,
      "loss": 1.7081,
      "step": 4596
    },
    {
      "epoch": 0.4968118448070896,
      "grad_norm": 0.4764062166213989,
      "learning_rate": 8e-05,
      "loss": 1.8676,
      "step": 4597
    },
    {
      "epoch": 0.49691991786447637,
      "grad_norm": 0.44301289319992065,
      "learning_rate": 8e-05,
      "loss": 1.7514,
      "step": 4598
    },
    {
      "epoch": 0.49702799092186317,
      "grad_norm": 0.3918417990207672,
      "learning_rate": 8e-05,
      "loss": 1.711,
      "step": 4599
    },
    {
      "epoch": 0.49713606397925,
      "grad_norm": 0.4111935794353485,
      "learning_rate": 8e-05,
      "loss": 1.7382,
      "step": 4600
    },
    {
      "epoch": 0.4972441370366368,
      "grad_norm": 0.40989717841148376,
      "learning_rate": 8e-05,
      "loss": 1.629,
      "step": 4601
    },
    {
      "epoch": 0.4973522100940236,
      "grad_norm": 0.4190944731235504,
      "learning_rate": 8e-05,
      "loss": 1.5246,
      "step": 4602
    },
    {
      "epoch": 0.49746028315141033,
      "grad_norm": 0.43257588148117065,
      "learning_rate": 8e-05,
      "loss": 1.6859,
      "step": 4603
    },
    {
      "epoch": 0.49756835620879714,
      "grad_norm": 0.4143356680870056,
      "learning_rate": 8e-05,
      "loss": 1.7318,
      "step": 4604
    },
    {
      "epoch": 0.49767642926618394,
      "grad_norm": 0.4364134669303894,
      "learning_rate": 8e-05,
      "loss": 1.5669,
      "step": 4605
    },
    {
      "epoch": 0.49778450232357074,
      "grad_norm": 0.47212454676628113,
      "learning_rate": 8e-05,
      "loss": 1.8856,
      "step": 4606
    },
    {
      "epoch": 0.49789257538095755,
      "grad_norm": 0.4265766441822052,
      "learning_rate": 8e-05,
      "loss": 1.8067,
      "step": 4607
    },
    {
      "epoch": 0.4980006484383443,
      "grad_norm": 0.37790244817733765,
      "learning_rate": 8e-05,
      "loss": 1.6611,
      "step": 4608
    },
    {
      "epoch": 0.4981087214957311,
      "grad_norm": 0.4193977117538452,
      "learning_rate": 8e-05,
      "loss": 1.7245,
      "step": 4609
    },
    {
      "epoch": 0.4982167945531179,
      "grad_norm": 0.40634286403656006,
      "learning_rate": 8e-05,
      "loss": 1.7468,
      "step": 4610
    },
    {
      "epoch": 0.4983248676105047,
      "grad_norm": 0.3872389793395996,
      "learning_rate": 8e-05,
      "loss": 1.6042,
      "step": 4611
    },
    {
      "epoch": 0.4984329406678915,
      "grad_norm": 0.43475615978240967,
      "learning_rate": 8e-05,
      "loss": 1.652,
      "step": 4612
    },
    {
      "epoch": 0.4985410137252783,
      "grad_norm": 0.4667968153953552,
      "learning_rate": 8e-05,
      "loss": 1.6648,
      "step": 4613
    },
    {
      "epoch": 0.49864908678266506,
      "grad_norm": 0.3985169231891632,
      "learning_rate": 8e-05,
      "loss": 1.6283,
      "step": 4614
    },
    {
      "epoch": 0.49875715984005187,
      "grad_norm": 0.42451563477516174,
      "learning_rate": 8e-05,
      "loss": 1.7535,
      "step": 4615
    },
    {
      "epoch": 0.49886523289743867,
      "grad_norm": 0.4101034700870514,
      "learning_rate": 8e-05,
      "loss": 1.7681,
      "step": 4616
    },
    {
      "epoch": 0.4989733059548255,
      "grad_norm": 0.3983090817928314,
      "learning_rate": 8e-05,
      "loss": 1.907,
      "step": 4617
    },
    {
      "epoch": 0.4990813790122123,
      "grad_norm": 0.37420305609703064,
      "learning_rate": 8e-05,
      "loss": 1.6059,
      "step": 4618
    },
    {
      "epoch": 0.499189452069599,
      "grad_norm": 0.3963295817375183,
      "learning_rate": 8e-05,
      "loss": 1.7792,
      "step": 4619
    },
    {
      "epoch": 0.49929752512698583,
      "grad_norm": 0.40464502573013306,
      "learning_rate": 8e-05,
      "loss": 1.5492,
      "step": 4620
    },
    {
      "epoch": 0.49940559818437263,
      "grad_norm": 0.40743106603622437,
      "learning_rate": 8e-05,
      "loss": 1.7119,
      "step": 4621
    },
    {
      "epoch": 0.49951367124175944,
      "grad_norm": 0.451715350151062,
      "learning_rate": 8e-05,
      "loss": 1.6747,
      "step": 4622
    },
    {
      "epoch": 0.49962174429914624,
      "grad_norm": 0.40801721811294556,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 4623
    },
    {
      "epoch": 0.499729817356533,
      "grad_norm": 0.4135707914829254,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 4624
    },
    {
      "epoch": 0.4998378904139198,
      "grad_norm": 0.4532390236854553,
      "learning_rate": 8e-05,
      "loss": 1.7641,
      "step": 4625
    },
    {
      "epoch": 0.4999459634713066,
      "grad_norm": 0.4245128929615021,
      "learning_rate": 8e-05,
      "loss": 1.8619,
      "step": 4626
    },
    {
      "epoch": 0.5000540365286934,
      "grad_norm": 0.38395196199417114,
      "learning_rate": 8e-05,
      "loss": 1.5814,
      "step": 4627
    },
    {
      "epoch": 0.5001621095860802,
      "grad_norm": 0.4016512632369995,
      "learning_rate": 8e-05,
      "loss": 1.6281,
      "step": 4628
    },
    {
      "epoch": 0.500270182643467,
      "grad_norm": 0.4368300437927246,
      "learning_rate": 8e-05,
      "loss": 1.8255,
      "step": 4629
    },
    {
      "epoch": 0.5003782557008538,
      "grad_norm": 0.4324931800365448,
      "learning_rate": 8e-05,
      "loss": 1.7219,
      "step": 4630
    },
    {
      "epoch": 0.5004863287582406,
      "grad_norm": 0.4427073895931244,
      "learning_rate": 8e-05,
      "loss": 1.6986,
      "step": 4631
    },
    {
      "epoch": 0.5005944018156273,
      "grad_norm": 0.4387059509754181,
      "learning_rate": 8e-05,
      "loss": 1.767,
      "step": 4632
    },
    {
      "epoch": 0.5007024748730141,
      "grad_norm": 0.4788256585597992,
      "learning_rate": 8e-05,
      "loss": 1.9927,
      "step": 4633
    },
    {
      "epoch": 0.5008105479304009,
      "grad_norm": 0.4731398820877075,
      "learning_rate": 8e-05,
      "loss": 1.8522,
      "step": 4634
    },
    {
      "epoch": 0.5009186209877877,
      "grad_norm": 0.3729192614555359,
      "learning_rate": 8e-05,
      "loss": 1.627,
      "step": 4635
    },
    {
      "epoch": 0.5010266940451745,
      "grad_norm": 0.3934325873851776,
      "learning_rate": 8e-05,
      "loss": 1.7136,
      "step": 4636
    },
    {
      "epoch": 0.5011347671025613,
      "grad_norm": 0.5115959048271179,
      "learning_rate": 8e-05,
      "loss": 1.8842,
      "step": 4637
    },
    {
      "epoch": 0.5012428401599481,
      "grad_norm": 0.3994344174861908,
      "learning_rate": 8e-05,
      "loss": 1.6874,
      "step": 4638
    },
    {
      "epoch": 0.5013509132173349,
      "grad_norm": 0.4582684338092804,
      "learning_rate": 8e-05,
      "loss": 1.8734,
      "step": 4639
    },
    {
      "epoch": 0.5014589862747217,
      "grad_norm": 0.40094858407974243,
      "learning_rate": 8e-05,
      "loss": 1.8413,
      "step": 4640
    },
    {
      "epoch": 0.5015670593321085,
      "grad_norm": 0.45019951462745667,
      "learning_rate": 8e-05,
      "loss": 1.955,
      "step": 4641
    },
    {
      "epoch": 0.5016751323894953,
      "grad_norm": 0.4047348201274872,
      "learning_rate": 8e-05,
      "loss": 1.6152,
      "step": 4642
    },
    {
      "epoch": 0.501783205446882,
      "grad_norm": 0.39314234256744385,
      "learning_rate": 8e-05,
      "loss": 1.6609,
      "step": 4643
    },
    {
      "epoch": 0.5018912785042688,
      "grad_norm": 0.4130818247795105,
      "learning_rate": 8e-05,
      "loss": 1.5828,
      "step": 4644
    },
    {
      "epoch": 0.5019993515616556,
      "grad_norm": 0.3810843229293823,
      "learning_rate": 8e-05,
      "loss": 1.6552,
      "step": 4645
    },
    {
      "epoch": 0.5021074246190425,
      "grad_norm": 0.3893740475177765,
      "learning_rate": 8e-05,
      "loss": 1.5267,
      "step": 4646
    },
    {
      "epoch": 0.5022154976764293,
      "grad_norm": 0.38051480054855347,
      "learning_rate": 8e-05,
      "loss": 1.5957,
      "step": 4647
    },
    {
      "epoch": 0.5023235707338161,
      "grad_norm": 0.428196519613266,
      "learning_rate": 8e-05,
      "loss": 1.7314,
      "step": 4648
    },
    {
      "epoch": 0.5024316437912029,
      "grad_norm": 0.4092353582382202,
      "learning_rate": 8e-05,
      "loss": 1.828,
      "step": 4649
    },
    {
      "epoch": 0.5025397168485897,
      "grad_norm": 0.3609974980354309,
      "learning_rate": 8e-05,
      "loss": 1.6815,
      "step": 4650
    },
    {
      "epoch": 0.5026477899059765,
      "grad_norm": 0.4251629412174225,
      "learning_rate": 8e-05,
      "loss": 1.8908,
      "step": 4651
    },
    {
      "epoch": 0.5027558629633633,
      "grad_norm": 0.37392112612724304,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 4652
    },
    {
      "epoch": 0.5028639360207501,
      "grad_norm": 0.4434810280799866,
      "learning_rate": 8e-05,
      "loss": 1.845,
      "step": 4653
    },
    {
      "epoch": 0.5029720090781368,
      "grad_norm": 0.38593465089797974,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 4654
    },
    {
      "epoch": 0.5030800821355236,
      "grad_norm": 0.4458652436733246,
      "learning_rate": 8e-05,
      "loss": 1.7922,
      "step": 4655
    },
    {
      "epoch": 0.5031881551929104,
      "grad_norm": 0.4200010299682617,
      "learning_rate": 8e-05,
      "loss": 1.7636,
      "step": 4656
    },
    {
      "epoch": 0.5032962282502972,
      "grad_norm": 0.3839763104915619,
      "learning_rate": 8e-05,
      "loss": 1.5582,
      "step": 4657
    },
    {
      "epoch": 0.503404301307684,
      "grad_norm": 0.4212810695171356,
      "learning_rate": 8e-05,
      "loss": 1.5734,
      "step": 4658
    },
    {
      "epoch": 0.5035123743650708,
      "grad_norm": 0.3777220547199249,
      "learning_rate": 8e-05,
      "loss": 1.6457,
      "step": 4659
    },
    {
      "epoch": 0.5036204474224576,
      "grad_norm": 0.41301804780960083,
      "learning_rate": 8e-05,
      "loss": 1.6975,
      "step": 4660
    },
    {
      "epoch": 0.5037285204798444,
      "grad_norm": 0.4142366349697113,
      "learning_rate": 8e-05,
      "loss": 1.7257,
      "step": 4661
    },
    {
      "epoch": 0.5038365935372312,
      "grad_norm": 0.40339356660842896,
      "learning_rate": 8e-05,
      "loss": 1.6962,
      "step": 4662
    },
    {
      "epoch": 0.503944666594618,
      "grad_norm": 0.4041508436203003,
      "learning_rate": 8e-05,
      "loss": 1.6953,
      "step": 4663
    },
    {
      "epoch": 0.5040527396520047,
      "grad_norm": 0.42368394136428833,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 4664
    },
    {
      "epoch": 0.5041608127093915,
      "grad_norm": 0.3843143880367279,
      "learning_rate": 8e-05,
      "loss": 1.6259,
      "step": 4665
    },
    {
      "epoch": 0.5042688857667783,
      "grad_norm": 0.4025265872478485,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 4666
    },
    {
      "epoch": 0.5043769588241651,
      "grad_norm": 0.444685697555542,
      "learning_rate": 8e-05,
      "loss": 1.5933,
      "step": 4667
    },
    {
      "epoch": 0.5044850318815519,
      "grad_norm": 0.43712174892425537,
      "learning_rate": 8e-05,
      "loss": 1.8116,
      "step": 4668
    },
    {
      "epoch": 0.5045931049389387,
      "grad_norm": 0.429677277803421,
      "learning_rate": 8e-05,
      "loss": 1.8109,
      "step": 4669
    },
    {
      "epoch": 0.5047011779963255,
      "grad_norm": 0.3996693193912506,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 4670
    },
    {
      "epoch": 0.5048092510537123,
      "grad_norm": 0.400955468416214,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 4671
    },
    {
      "epoch": 0.5049173241110991,
      "grad_norm": 0.39697420597076416,
      "learning_rate": 8e-05,
      "loss": 1.6263,
      "step": 4672
    },
    {
      "epoch": 0.5050253971684859,
      "grad_norm": 0.41106465458869934,
      "learning_rate": 8e-05,
      "loss": 1.6128,
      "step": 4673
    },
    {
      "epoch": 0.5051334702258727,
      "grad_norm": 0.3903358578681946,
      "learning_rate": 8e-05,
      "loss": 1.7401,
      "step": 4674
    },
    {
      "epoch": 0.5052415432832594,
      "grad_norm": 0.4007432162761688,
      "learning_rate": 8e-05,
      "loss": 1.8125,
      "step": 4675
    },
    {
      "epoch": 0.5053496163406462,
      "grad_norm": 0.41746750473976135,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 4676
    },
    {
      "epoch": 0.505457689398033,
      "grad_norm": 0.4129239320755005,
      "learning_rate": 8e-05,
      "loss": 1.7308,
      "step": 4677
    },
    {
      "epoch": 0.5055657624554198,
      "grad_norm": 0.4130561947822571,
      "learning_rate": 8e-05,
      "loss": 1.7485,
      "step": 4678
    },
    {
      "epoch": 0.5056738355128066,
      "grad_norm": 0.38476482033729553,
      "learning_rate": 8e-05,
      "loss": 1.6355,
      "step": 4679
    },
    {
      "epoch": 0.5057819085701935,
      "grad_norm": 0.41772881150245667,
      "learning_rate": 8e-05,
      "loss": 1.7084,
      "step": 4680
    },
    {
      "epoch": 0.5058899816275803,
      "grad_norm": 0.3783447742462158,
      "learning_rate": 8e-05,
      "loss": 1.7684,
      "step": 4681
    },
    {
      "epoch": 0.5059980546849671,
      "grad_norm": 0.40590929985046387,
      "learning_rate": 8e-05,
      "loss": 1.7061,
      "step": 4682
    },
    {
      "epoch": 0.5061061277423539,
      "grad_norm": 0.41405728459358215,
      "learning_rate": 8e-05,
      "loss": 1.7863,
      "step": 4683
    },
    {
      "epoch": 0.5062142007997407,
      "grad_norm": 0.39419227838516235,
      "learning_rate": 8e-05,
      "loss": 1.7967,
      "step": 4684
    },
    {
      "epoch": 0.5063222738571275,
      "grad_norm": 0.41945794224739075,
      "learning_rate": 8e-05,
      "loss": 1.7821,
      "step": 4685
    },
    {
      "epoch": 0.5064303469145142,
      "grad_norm": 0.43580418825149536,
      "learning_rate": 8e-05,
      "loss": 1.7528,
      "step": 4686
    },
    {
      "epoch": 0.506538419971901,
      "grad_norm": 0.47292235493659973,
      "learning_rate": 8e-05,
      "loss": 1.8126,
      "step": 4687
    },
    {
      "epoch": 0.5066464930292878,
      "grad_norm": 0.3983614444732666,
      "learning_rate": 8e-05,
      "loss": 1.5537,
      "step": 4688
    },
    {
      "epoch": 0.5067545660866746,
      "grad_norm": 0.4345097839832306,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 4689
    },
    {
      "epoch": 0.5068626391440614,
      "grad_norm": 0.4169449210166931,
      "learning_rate": 8e-05,
      "loss": 1.7158,
      "step": 4690
    },
    {
      "epoch": 0.5069707122014482,
      "grad_norm": 0.3657934367656708,
      "learning_rate": 8e-05,
      "loss": 1.5082,
      "step": 4691
    },
    {
      "epoch": 0.507078785258835,
      "grad_norm": 0.4632113575935364,
      "learning_rate": 8e-05,
      "loss": 1.8278,
      "step": 4692
    },
    {
      "epoch": 0.5071868583162218,
      "grad_norm": 0.3980325162410736,
      "learning_rate": 8e-05,
      "loss": 1.7702,
      "step": 4693
    },
    {
      "epoch": 0.5072949313736086,
      "grad_norm": 0.3889184296131134,
      "learning_rate": 8e-05,
      "loss": 1.6634,
      "step": 4694
    },
    {
      "epoch": 0.5074030044309954,
      "grad_norm": 0.36603844165802,
      "learning_rate": 8e-05,
      "loss": 1.4117,
      "step": 4695
    },
    {
      "epoch": 0.5075110774883822,
      "grad_norm": 0.4158637225627899,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 4696
    },
    {
      "epoch": 0.5076191505457689,
      "grad_norm": 0.423439621925354,
      "learning_rate": 8e-05,
      "loss": 1.7605,
      "step": 4697
    },
    {
      "epoch": 0.5077272236031557,
      "grad_norm": 0.4030735492706299,
      "learning_rate": 8e-05,
      "loss": 1.7774,
      "step": 4698
    },
    {
      "epoch": 0.5078352966605425,
      "grad_norm": 0.4302954375743866,
      "learning_rate": 8e-05,
      "loss": 1.8201,
      "step": 4699
    },
    {
      "epoch": 0.5079433697179293,
      "grad_norm": 0.3883877098560333,
      "learning_rate": 8e-05,
      "loss": 1.6702,
      "step": 4700
    },
    {
      "epoch": 0.5080514427753161,
      "grad_norm": 0.3819584250450134,
      "learning_rate": 8e-05,
      "loss": 1.4745,
      "step": 4701
    },
    {
      "epoch": 0.5081595158327029,
      "grad_norm": 0.4463454484939575,
      "learning_rate": 8e-05,
      "loss": 1.8073,
      "step": 4702
    },
    {
      "epoch": 0.5082675888900897,
      "grad_norm": 0.39672747254371643,
      "learning_rate": 8e-05,
      "loss": 1.6663,
      "step": 4703
    },
    {
      "epoch": 0.5083756619474765,
      "grad_norm": 0.42274191975593567,
      "learning_rate": 8e-05,
      "loss": 1.7293,
      "step": 4704
    },
    {
      "epoch": 0.5084837350048633,
      "grad_norm": 0.4081788957118988,
      "learning_rate": 8e-05,
      "loss": 1.9558,
      "step": 4705
    },
    {
      "epoch": 0.5085918080622501,
      "grad_norm": 0.4008321762084961,
      "learning_rate": 8e-05,
      "loss": 1.7788,
      "step": 4706
    },
    {
      "epoch": 0.5086998811196368,
      "grad_norm": 0.43986186385154724,
      "learning_rate": 8e-05,
      "loss": 1.9007,
      "step": 4707
    },
    {
      "epoch": 0.5088079541770236,
      "grad_norm": 0.393114298582077,
      "learning_rate": 8e-05,
      "loss": 1.6909,
      "step": 4708
    },
    {
      "epoch": 0.5089160272344104,
      "grad_norm": 0.36888784170150757,
      "learning_rate": 8e-05,
      "loss": 1.6337,
      "step": 4709
    },
    {
      "epoch": 0.5090241002917972,
      "grad_norm": 0.4395067095756531,
      "learning_rate": 8e-05,
      "loss": 1.917,
      "step": 4710
    },
    {
      "epoch": 0.509132173349184,
      "grad_norm": 0.42625027894973755,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 4711
    },
    {
      "epoch": 0.5092402464065708,
      "grad_norm": 0.41628801822662354,
      "learning_rate": 8e-05,
      "loss": 1.5516,
      "step": 4712
    },
    {
      "epoch": 0.5093483194639576,
      "grad_norm": 0.40050965547561646,
      "learning_rate": 8e-05,
      "loss": 1.574,
      "step": 4713
    },
    {
      "epoch": 0.5094563925213444,
      "grad_norm": 0.4109610915184021,
      "learning_rate": 8e-05,
      "loss": 1.7942,
      "step": 4714
    },
    {
      "epoch": 0.5095644655787313,
      "grad_norm": 0.42685291171073914,
      "learning_rate": 8e-05,
      "loss": 1.7486,
      "step": 4715
    },
    {
      "epoch": 0.509672538636118,
      "grad_norm": 0.39319854974746704,
      "learning_rate": 8e-05,
      "loss": 1.605,
      "step": 4716
    },
    {
      "epoch": 0.5097806116935049,
      "grad_norm": 0.3991563618183136,
      "learning_rate": 8e-05,
      "loss": 1.6549,
      "step": 4717
    },
    {
      "epoch": 0.5098886847508916,
      "grad_norm": 0.39866384863853455,
      "learning_rate": 8e-05,
      "loss": 1.6739,
      "step": 4718
    },
    {
      "epoch": 0.5099967578082784,
      "grad_norm": 0.4868659973144531,
      "learning_rate": 8e-05,
      "loss": 1.8198,
      "step": 4719
    },
    {
      "epoch": 0.5101048308656652,
      "grad_norm": 0.39707428216934204,
      "learning_rate": 8e-05,
      "loss": 1.6521,
      "step": 4720
    },
    {
      "epoch": 0.510212903923052,
      "grad_norm": 0.40801212191581726,
      "learning_rate": 8e-05,
      "loss": 1.7363,
      "step": 4721
    },
    {
      "epoch": 0.5103209769804388,
      "grad_norm": 0.44077005982398987,
      "learning_rate": 8e-05,
      "loss": 1.8504,
      "step": 4722
    },
    {
      "epoch": 0.5104290500378256,
      "grad_norm": 0.3992588520050049,
      "learning_rate": 8e-05,
      "loss": 1.7235,
      "step": 4723
    },
    {
      "epoch": 0.5105371230952124,
      "grad_norm": 0.4055352807044983,
      "learning_rate": 8e-05,
      "loss": 1.5592,
      "step": 4724
    },
    {
      "epoch": 0.5106451961525992,
      "grad_norm": 0.40171900391578674,
      "learning_rate": 8e-05,
      "loss": 1.6254,
      "step": 4725
    },
    {
      "epoch": 0.510753269209986,
      "grad_norm": 0.3749426603317261,
      "learning_rate": 8e-05,
      "loss": 1.4668,
      "step": 4726
    },
    {
      "epoch": 0.5108613422673728,
      "grad_norm": 0.3832828402519226,
      "learning_rate": 8e-05,
      "loss": 1.6808,
      "step": 4727
    },
    {
      "epoch": 0.5109694153247596,
      "grad_norm": 0.4637143611907959,
      "learning_rate": 8e-05,
      "loss": 1.9584,
      "step": 4728
    },
    {
      "epoch": 0.5110774883821463,
      "grad_norm": 0.4723820984363556,
      "learning_rate": 8e-05,
      "loss": 1.8783,
      "step": 4729
    },
    {
      "epoch": 0.5111855614395331,
      "grad_norm": 0.41199734807014465,
      "learning_rate": 8e-05,
      "loss": 1.7593,
      "step": 4730
    },
    {
      "epoch": 0.5112936344969199,
      "grad_norm": 0.38699331879615784,
      "learning_rate": 8e-05,
      "loss": 1.5373,
      "step": 4731
    },
    {
      "epoch": 0.5114017075543067,
      "grad_norm": 0.42380109429359436,
      "learning_rate": 8e-05,
      "loss": 1.7024,
      "step": 4732
    },
    {
      "epoch": 0.5115097806116935,
      "grad_norm": 0.38215172290802,
      "learning_rate": 8e-05,
      "loss": 1.7495,
      "step": 4733
    },
    {
      "epoch": 0.5116178536690803,
      "grad_norm": 0.4330931305885315,
      "learning_rate": 8e-05,
      "loss": 1.8796,
      "step": 4734
    },
    {
      "epoch": 0.5117259267264671,
      "grad_norm": 0.40725767612457275,
      "learning_rate": 8e-05,
      "loss": 1.6943,
      "step": 4735
    },
    {
      "epoch": 0.5118339997838539,
      "grad_norm": 0.4123496413230896,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 4736
    },
    {
      "epoch": 0.5119420728412407,
      "grad_norm": 0.39777886867523193,
      "learning_rate": 8e-05,
      "loss": 1.8141,
      "step": 4737
    },
    {
      "epoch": 0.5120501458986275,
      "grad_norm": 0.3945496678352356,
      "learning_rate": 8e-05,
      "loss": 1.5761,
      "step": 4738
    },
    {
      "epoch": 0.5121582189560142,
      "grad_norm": 0.42809048295021057,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 4739
    },
    {
      "epoch": 0.512266292013401,
      "grad_norm": 0.3837582767009735,
      "learning_rate": 8e-05,
      "loss": 1.6109,
      "step": 4740
    },
    {
      "epoch": 0.5123743650707878,
      "grad_norm": 0.4252813458442688,
      "learning_rate": 8e-05,
      "loss": 1.9006,
      "step": 4741
    },
    {
      "epoch": 0.5124824381281746,
      "grad_norm": 0.3779107928276062,
      "learning_rate": 8e-05,
      "loss": 1.6411,
      "step": 4742
    },
    {
      "epoch": 0.5125905111855614,
      "grad_norm": 0.41228488087654114,
      "learning_rate": 8e-05,
      "loss": 1.7215,
      "step": 4743
    },
    {
      "epoch": 0.5126985842429482,
      "grad_norm": 0.40794017910957336,
      "learning_rate": 8e-05,
      "loss": 1.7578,
      "step": 4744
    },
    {
      "epoch": 0.512806657300335,
      "grad_norm": 0.40917718410491943,
      "learning_rate": 8e-05,
      "loss": 1.6595,
      "step": 4745
    },
    {
      "epoch": 0.5129147303577218,
      "grad_norm": 0.3936637341976166,
      "learning_rate": 8e-05,
      "loss": 1.6749,
      "step": 4746
    },
    {
      "epoch": 0.5130228034151086,
      "grad_norm": 0.4197206497192383,
      "learning_rate": 8e-05,
      "loss": 1.764,
      "step": 4747
    },
    {
      "epoch": 0.5131308764724954,
      "grad_norm": 0.44139614701271057,
      "learning_rate": 8e-05,
      "loss": 1.7621,
      "step": 4748
    },
    {
      "epoch": 0.5132389495298822,
      "grad_norm": 0.4017902612686157,
      "learning_rate": 8e-05,
      "loss": 1.6353,
      "step": 4749
    },
    {
      "epoch": 0.5133470225872689,
      "grad_norm": 0.40742015838623047,
      "learning_rate": 8e-05,
      "loss": 1.6759,
      "step": 4750
    },
    {
      "epoch": 0.5134550956446557,
      "grad_norm": 0.43377262353897095,
      "learning_rate": 8e-05,
      "loss": 1.7223,
      "step": 4751
    },
    {
      "epoch": 0.5135631687020425,
      "grad_norm": 0.43455442786216736,
      "learning_rate": 8e-05,
      "loss": 1.5976,
      "step": 4752
    },
    {
      "epoch": 0.5136712417594294,
      "grad_norm": 0.42303210496902466,
      "learning_rate": 8e-05,
      "loss": 1.7429,
      "step": 4753
    },
    {
      "epoch": 0.5137793148168162,
      "grad_norm": 0.4154270589351654,
      "learning_rate": 8e-05,
      "loss": 1.8203,
      "step": 4754
    },
    {
      "epoch": 0.513887387874203,
      "grad_norm": 0.46774837374687195,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 4755
    },
    {
      "epoch": 0.5139954609315898,
      "grad_norm": 0.40013161301612854,
      "learning_rate": 8e-05,
      "loss": 1.8643,
      "step": 4756
    },
    {
      "epoch": 0.5141035339889766,
      "grad_norm": 0.4226974546909332,
      "learning_rate": 8e-05,
      "loss": 1.762,
      "step": 4757
    },
    {
      "epoch": 0.5142116070463634,
      "grad_norm": 0.629270076751709,
      "learning_rate": 8e-05,
      "loss": 1.9884,
      "step": 4758
    },
    {
      "epoch": 0.5143196801037502,
      "grad_norm": 0.38738080859184265,
      "learning_rate": 8e-05,
      "loss": 1.5638,
      "step": 4759
    },
    {
      "epoch": 0.514427753161137,
      "grad_norm": 0.4483691155910492,
      "learning_rate": 8e-05,
      "loss": 1.6217,
      "step": 4760
    },
    {
      "epoch": 0.5145358262185237,
      "grad_norm": 0.3760409653186798,
      "learning_rate": 8e-05,
      "loss": 1.6163,
      "step": 4761
    },
    {
      "epoch": 0.5146438992759105,
      "grad_norm": 0.5027571320533752,
      "learning_rate": 8e-05,
      "loss": 2.0142,
      "step": 4762
    },
    {
      "epoch": 0.5147519723332973,
      "grad_norm": 0.44235125184059143,
      "learning_rate": 8e-05,
      "loss": 1.6542,
      "step": 4763
    },
    {
      "epoch": 0.5148600453906841,
      "grad_norm": 0.38345348834991455,
      "learning_rate": 8e-05,
      "loss": 1.5266,
      "step": 4764
    },
    {
      "epoch": 0.5149681184480709,
      "grad_norm": 0.41503050923347473,
      "learning_rate": 8e-05,
      "loss": 1.7876,
      "step": 4765
    },
    {
      "epoch": 0.5150761915054577,
      "grad_norm": 0.40613964200019836,
      "learning_rate": 8e-05,
      "loss": 1.7009,
      "step": 4766
    },
    {
      "epoch": 0.5151842645628445,
      "grad_norm": 0.40172886848449707,
      "learning_rate": 8e-05,
      "loss": 1.7753,
      "step": 4767
    },
    {
      "epoch": 0.5152923376202313,
      "grad_norm": 0.43227338790893555,
      "learning_rate": 8e-05,
      "loss": 1.7651,
      "step": 4768
    },
    {
      "epoch": 0.5154004106776181,
      "grad_norm": 0.38273218274116516,
      "learning_rate": 8e-05,
      "loss": 1.5844,
      "step": 4769
    },
    {
      "epoch": 0.5155084837350049,
      "grad_norm": 0.43742454051971436,
      "learning_rate": 8e-05,
      "loss": 1.8227,
      "step": 4770
    },
    {
      "epoch": 0.5156165567923917,
      "grad_norm": 0.3798066973686218,
      "learning_rate": 8e-05,
      "loss": 1.6374,
      "step": 4771
    },
    {
      "epoch": 0.5157246298497784,
      "grad_norm": 0.4344523847103119,
      "learning_rate": 8e-05,
      "loss": 1.9303,
      "step": 4772
    },
    {
      "epoch": 0.5158327029071652,
      "grad_norm": 0.4373953640460968,
      "learning_rate": 8e-05,
      "loss": 1.8701,
      "step": 4773
    },
    {
      "epoch": 0.515940775964552,
      "grad_norm": 0.4018820822238922,
      "learning_rate": 8e-05,
      "loss": 1.722,
      "step": 4774
    },
    {
      "epoch": 0.5160488490219388,
      "grad_norm": 0.4193482995033264,
      "learning_rate": 8e-05,
      "loss": 1.7657,
      "step": 4775
    },
    {
      "epoch": 0.5161569220793256,
      "grad_norm": 0.3853967487812042,
      "learning_rate": 8e-05,
      "loss": 1.4658,
      "step": 4776
    },
    {
      "epoch": 0.5162649951367124,
      "grad_norm": 0.406354695558548,
      "learning_rate": 8e-05,
      "loss": 1.5258,
      "step": 4777
    },
    {
      "epoch": 0.5163730681940992,
      "grad_norm": 0.43898022174835205,
      "learning_rate": 8e-05,
      "loss": 1.836,
      "step": 4778
    },
    {
      "epoch": 0.516481141251486,
      "grad_norm": 0.41223448514938354,
      "learning_rate": 8e-05,
      "loss": 1.644,
      "step": 4779
    },
    {
      "epoch": 0.5165892143088728,
      "grad_norm": 0.43973109126091003,
      "learning_rate": 8e-05,
      "loss": 1.8827,
      "step": 4780
    },
    {
      "epoch": 0.5166972873662596,
      "grad_norm": 0.4527101516723633,
      "learning_rate": 8e-05,
      "loss": 1.7622,
      "step": 4781
    },
    {
      "epoch": 0.5168053604236463,
      "grad_norm": 0.4201156795024872,
      "learning_rate": 8e-05,
      "loss": 1.7761,
      "step": 4782
    },
    {
      "epoch": 0.5169134334810331,
      "grad_norm": 0.4157174825668335,
      "learning_rate": 8e-05,
      "loss": 1.6961,
      "step": 4783
    },
    {
      "epoch": 0.5170215065384199,
      "grad_norm": 0.405518114566803,
      "learning_rate": 8e-05,
      "loss": 1.7131,
      "step": 4784
    },
    {
      "epoch": 0.5171295795958067,
      "grad_norm": 0.4291194975376129,
      "learning_rate": 8e-05,
      "loss": 1.8589,
      "step": 4785
    },
    {
      "epoch": 0.5172376526531935,
      "grad_norm": 0.383748859167099,
      "learning_rate": 8e-05,
      "loss": 1.6687,
      "step": 4786
    },
    {
      "epoch": 0.5173457257105804,
      "grad_norm": 0.3920312821865082,
      "learning_rate": 8e-05,
      "loss": 1.7444,
      "step": 4787
    },
    {
      "epoch": 0.5174537987679672,
      "grad_norm": 0.3934651017189026,
      "learning_rate": 8e-05,
      "loss": 1.7991,
      "step": 4788
    },
    {
      "epoch": 0.517561871825354,
      "grad_norm": 0.38649001717567444,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 4789
    },
    {
      "epoch": 0.5176699448827408,
      "grad_norm": 0.42291659116744995,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 4790
    },
    {
      "epoch": 0.5177780179401276,
      "grad_norm": 0.5816870927810669,
      "learning_rate": 8e-05,
      "loss": 1.6837,
      "step": 4791
    },
    {
      "epoch": 0.5178860909975144,
      "grad_norm": 0.4310382902622223,
      "learning_rate": 8e-05,
      "loss": 1.8475,
      "step": 4792
    },
    {
      "epoch": 0.5179941640549011,
      "grad_norm": 0.3854757845401764,
      "learning_rate": 8e-05,
      "loss": 1.6007,
      "step": 4793
    },
    {
      "epoch": 0.5181022371122879,
      "grad_norm": 0.425446480512619,
      "learning_rate": 8e-05,
      "loss": 1.867,
      "step": 4794
    },
    {
      "epoch": 0.5182103101696747,
      "grad_norm": 0.3843328654766083,
      "learning_rate": 8e-05,
      "loss": 1.5137,
      "step": 4795
    },
    {
      "epoch": 0.5183183832270615,
      "grad_norm": 0.4098472595214844,
      "learning_rate": 8e-05,
      "loss": 1.6805,
      "step": 4796
    },
    {
      "epoch": 0.5184264562844483,
      "grad_norm": 0.39224570989608765,
      "learning_rate": 8e-05,
      "loss": 1.6642,
      "step": 4797
    },
    {
      "epoch": 0.5185345293418351,
      "grad_norm": 0.3970671594142914,
      "learning_rate": 8e-05,
      "loss": 1.5031,
      "step": 4798
    },
    {
      "epoch": 0.5186426023992219,
      "grad_norm": 0.3776818513870239,
      "learning_rate": 8e-05,
      "loss": 1.4388,
      "step": 4799
    },
    {
      "epoch": 0.5187506754566087,
      "grad_norm": 0.43284153938293457,
      "learning_rate": 8e-05,
      "loss": 1.6793,
      "step": 4800
    },
    {
      "epoch": 0.5188587485139955,
      "grad_norm": 0.46994057297706604,
      "learning_rate": 8e-05,
      "loss": 1.8721,
      "step": 4801
    },
    {
      "epoch": 0.5189668215713823,
      "grad_norm": 0.43751636147499084,
      "learning_rate": 8e-05,
      "loss": 1.8159,
      "step": 4802
    },
    {
      "epoch": 0.5190748946287691,
      "grad_norm": 0.4508005380630493,
      "learning_rate": 8e-05,
      "loss": 1.939,
      "step": 4803
    },
    {
      "epoch": 0.5191829676861558,
      "grad_norm": 0.4641004502773285,
      "learning_rate": 8e-05,
      "loss": 1.9424,
      "step": 4804
    },
    {
      "epoch": 0.5192910407435426,
      "grad_norm": 0.4278232455253601,
      "learning_rate": 8e-05,
      "loss": 1.6807,
      "step": 4805
    },
    {
      "epoch": 0.5193991138009294,
      "grad_norm": 0.41461819410324097,
      "learning_rate": 8e-05,
      "loss": 1.569,
      "step": 4806
    },
    {
      "epoch": 0.5195071868583162,
      "grad_norm": 0.42840465903282166,
      "learning_rate": 8e-05,
      "loss": 1.7524,
      "step": 4807
    },
    {
      "epoch": 0.519615259915703,
      "grad_norm": 0.4226847290992737,
      "learning_rate": 8e-05,
      "loss": 1.6598,
      "step": 4808
    },
    {
      "epoch": 0.5197233329730898,
      "grad_norm": 0.3884517550468445,
      "learning_rate": 8e-05,
      "loss": 1.6587,
      "step": 4809
    },
    {
      "epoch": 0.5198314060304766,
      "grad_norm": 0.38184890151023865,
      "learning_rate": 8e-05,
      "loss": 1.8374,
      "step": 4810
    },
    {
      "epoch": 0.5199394790878634,
      "grad_norm": 0.40016797184944153,
      "learning_rate": 8e-05,
      "loss": 1.7748,
      "step": 4811
    },
    {
      "epoch": 0.5200475521452502,
      "grad_norm": 0.4112529456615448,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 4812
    },
    {
      "epoch": 0.520155625202637,
      "grad_norm": 0.3872961401939392,
      "learning_rate": 8e-05,
      "loss": 1.7005,
      "step": 4813
    },
    {
      "epoch": 0.5202636982600237,
      "grad_norm": 0.394014835357666,
      "learning_rate": 8e-05,
      "loss": 1.6279,
      "step": 4814
    },
    {
      "epoch": 0.5203717713174105,
      "grad_norm": 0.3823867440223694,
      "learning_rate": 8e-05,
      "loss": 1.5722,
      "step": 4815
    },
    {
      "epoch": 0.5204798443747973,
      "grad_norm": 0.4088960587978363,
      "learning_rate": 8e-05,
      "loss": 1.64,
      "step": 4816
    },
    {
      "epoch": 0.5205879174321841,
      "grad_norm": 0.44685935974121094,
      "learning_rate": 8e-05,
      "loss": 1.8071,
      "step": 4817
    },
    {
      "epoch": 0.5206959904895709,
      "grad_norm": 0.4237087666988373,
      "learning_rate": 8e-05,
      "loss": 1.8672,
      "step": 4818
    },
    {
      "epoch": 0.5208040635469577,
      "grad_norm": 0.3990110456943512,
      "learning_rate": 8e-05,
      "loss": 1.8518,
      "step": 4819
    },
    {
      "epoch": 0.5209121366043445,
      "grad_norm": 0.41446077823638916,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 4820
    },
    {
      "epoch": 0.5210202096617313,
      "grad_norm": 0.42045265436172485,
      "learning_rate": 8e-05,
      "loss": 1.9048,
      "step": 4821
    },
    {
      "epoch": 0.5211282827191182,
      "grad_norm": 0.4046669602394104,
      "learning_rate": 8e-05,
      "loss": 1.639,
      "step": 4822
    },
    {
      "epoch": 0.521236355776505,
      "grad_norm": 0.3751082122325897,
      "learning_rate": 8e-05,
      "loss": 1.6782,
      "step": 4823
    },
    {
      "epoch": 0.5213444288338918,
      "grad_norm": 0.3985498547554016,
      "learning_rate": 8e-05,
      "loss": 1.6451,
      "step": 4824
    },
    {
      "epoch": 0.5214525018912785,
      "grad_norm": 0.38252684473991394,
      "learning_rate": 8e-05,
      "loss": 1.5875,
      "step": 4825
    },
    {
      "epoch": 0.5215605749486653,
      "grad_norm": 0.4107401669025421,
      "learning_rate": 8e-05,
      "loss": 1.5613,
      "step": 4826
    },
    {
      "epoch": 0.5216686480060521,
      "grad_norm": 0.45246565341949463,
      "learning_rate": 8e-05,
      "loss": 1.771,
      "step": 4827
    },
    {
      "epoch": 0.5217767210634389,
      "grad_norm": 0.38584524393081665,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 4828
    },
    {
      "epoch": 0.5218847941208257,
      "grad_norm": 0.42803874611854553,
      "learning_rate": 8e-05,
      "loss": 1.8083,
      "step": 4829
    },
    {
      "epoch": 0.5219928671782125,
      "grad_norm": 0.4374076724052429,
      "learning_rate": 8e-05,
      "loss": 1.7597,
      "step": 4830
    },
    {
      "epoch": 0.5221009402355993,
      "grad_norm": 0.48844432830810547,
      "learning_rate": 8e-05,
      "loss": 1.9372,
      "step": 4831
    },
    {
      "epoch": 0.5222090132929861,
      "grad_norm": 0.42745932936668396,
      "learning_rate": 8e-05,
      "loss": 1.6969,
      "step": 4832
    },
    {
      "epoch": 0.5223170863503729,
      "grad_norm": 0.41262972354888916,
      "learning_rate": 8e-05,
      "loss": 1.7552,
      "step": 4833
    },
    {
      "epoch": 0.5224251594077597,
      "grad_norm": 0.40279683470726013,
      "learning_rate": 8e-05,
      "loss": 1.7008,
      "step": 4834
    },
    {
      "epoch": 0.5225332324651465,
      "grad_norm": 0.42590004205703735,
      "learning_rate": 8e-05,
      "loss": 1.7784,
      "step": 4835
    },
    {
      "epoch": 0.5226413055225332,
      "grad_norm": 0.42654383182525635,
      "learning_rate": 8e-05,
      "loss": 1.7202,
      "step": 4836
    },
    {
      "epoch": 0.52274937857992,
      "grad_norm": 0.4391385316848755,
      "learning_rate": 8e-05,
      "loss": 1.7712,
      "step": 4837
    },
    {
      "epoch": 0.5228574516373068,
      "grad_norm": 0.4117557406425476,
      "learning_rate": 8e-05,
      "loss": 1.7483,
      "step": 4838
    },
    {
      "epoch": 0.5229655246946936,
      "grad_norm": 0.4126327633857727,
      "learning_rate": 8e-05,
      "loss": 1.6878,
      "step": 4839
    },
    {
      "epoch": 0.5230735977520804,
      "grad_norm": 0.46406158804893494,
      "learning_rate": 8e-05,
      "loss": 1.8207,
      "step": 4840
    },
    {
      "epoch": 0.5231816708094672,
      "grad_norm": 0.4451543688774109,
      "learning_rate": 8e-05,
      "loss": 1.9556,
      "step": 4841
    },
    {
      "epoch": 0.523289743866854,
      "grad_norm": 0.39060765504837036,
      "learning_rate": 8e-05,
      "loss": 1.6436,
      "step": 4842
    },
    {
      "epoch": 0.5233978169242408,
      "grad_norm": 0.3856661915779114,
      "learning_rate": 8e-05,
      "loss": 1.6003,
      "step": 4843
    },
    {
      "epoch": 0.5235058899816276,
      "grad_norm": 0.41874486207962036,
      "learning_rate": 8e-05,
      "loss": 1.7198,
      "step": 4844
    },
    {
      "epoch": 0.5236139630390144,
      "grad_norm": 0.38177576661109924,
      "learning_rate": 8e-05,
      "loss": 1.7114,
      "step": 4845
    },
    {
      "epoch": 0.5237220360964012,
      "grad_norm": 0.44675570726394653,
      "learning_rate": 8e-05,
      "loss": 1.6544,
      "step": 4846
    },
    {
      "epoch": 0.5238301091537879,
      "grad_norm": 0.39139798283576965,
      "learning_rate": 8e-05,
      "loss": 1.5859,
      "step": 4847
    },
    {
      "epoch": 0.5239381822111747,
      "grad_norm": 0.37541764974594116,
      "learning_rate": 8e-05,
      "loss": 1.5277,
      "step": 4848
    },
    {
      "epoch": 0.5240462552685615,
      "grad_norm": 0.4628608822822571,
      "learning_rate": 8e-05,
      "loss": 1.8211,
      "step": 4849
    },
    {
      "epoch": 0.5241543283259483,
      "grad_norm": 0.43806251883506775,
      "learning_rate": 8e-05,
      "loss": 1.7695,
      "step": 4850
    },
    {
      "epoch": 0.5242624013833351,
      "grad_norm": 0.45326513051986694,
      "learning_rate": 8e-05,
      "loss": 1.61,
      "step": 4851
    },
    {
      "epoch": 0.5243704744407219,
      "grad_norm": 0.392219603061676,
      "learning_rate": 8e-05,
      "loss": 1.7151,
      "step": 4852
    },
    {
      "epoch": 0.5244785474981087,
      "grad_norm": 0.4122357964515686,
      "learning_rate": 8e-05,
      "loss": 1.7077,
      "step": 4853
    },
    {
      "epoch": 0.5245866205554955,
      "grad_norm": 0.43415403366088867,
      "learning_rate": 8e-05,
      "loss": 1.8575,
      "step": 4854
    },
    {
      "epoch": 0.5246946936128823,
      "grad_norm": 0.3981955349445343,
      "learning_rate": 8e-05,
      "loss": 1.6174,
      "step": 4855
    },
    {
      "epoch": 0.5248027666702691,
      "grad_norm": 0.3884790241718292,
      "learning_rate": 8e-05,
      "loss": 1.649,
      "step": 4856
    },
    {
      "epoch": 0.5249108397276558,
      "grad_norm": 0.39559316635131836,
      "learning_rate": 8e-05,
      "loss": 1.6578,
      "step": 4857
    },
    {
      "epoch": 0.5250189127850426,
      "grad_norm": 0.4003053307533264,
      "learning_rate": 8e-05,
      "loss": 1.6516,
      "step": 4858
    },
    {
      "epoch": 0.5251269858424295,
      "grad_norm": 0.37234529852867126,
      "learning_rate": 8e-05,
      "loss": 1.637,
      "step": 4859
    },
    {
      "epoch": 0.5252350588998163,
      "grad_norm": 0.4073370099067688,
      "learning_rate": 8e-05,
      "loss": 1.7174,
      "step": 4860
    },
    {
      "epoch": 0.5253431319572031,
      "grad_norm": 0.41446778178215027,
      "learning_rate": 8e-05,
      "loss": 1.794,
      "step": 4861
    },
    {
      "epoch": 0.5254512050145899,
      "grad_norm": 0.39679303765296936,
      "learning_rate": 8e-05,
      "loss": 1.5728,
      "step": 4862
    },
    {
      "epoch": 0.5255592780719767,
      "grad_norm": 0.45979782938957214,
      "learning_rate": 8e-05,
      "loss": 1.7414,
      "step": 4863
    },
    {
      "epoch": 0.5256673511293635,
      "grad_norm": 0.4084329605102539,
      "learning_rate": 8e-05,
      "loss": 1.7752,
      "step": 4864
    },
    {
      "epoch": 0.5257754241867503,
      "grad_norm": 0.39365530014038086,
      "learning_rate": 8e-05,
      "loss": 1.7289,
      "step": 4865
    },
    {
      "epoch": 0.5258834972441371,
      "grad_norm": 0.39535269141197205,
      "learning_rate": 8e-05,
      "loss": 1.5131,
      "step": 4866
    },
    {
      "epoch": 0.5259915703015239,
      "grad_norm": 0.4394475817680359,
      "learning_rate": 8e-05,
      "loss": 1.8352,
      "step": 4867
    },
    {
      "epoch": 0.5260996433589106,
      "grad_norm": 0.437294602394104,
      "learning_rate": 8e-05,
      "loss": 1.8267,
      "step": 4868
    },
    {
      "epoch": 0.5262077164162974,
      "grad_norm": 0.4125720262527466,
      "learning_rate": 8e-05,
      "loss": 1.6659,
      "step": 4869
    },
    {
      "epoch": 0.5263157894736842,
      "grad_norm": 0.3946208357810974,
      "learning_rate": 8e-05,
      "loss": 1.6276,
      "step": 4870
    },
    {
      "epoch": 0.526423862531071,
      "grad_norm": 0.4423826336860657,
      "learning_rate": 8e-05,
      "loss": 1.6709,
      "step": 4871
    },
    {
      "epoch": 0.5265319355884578,
      "grad_norm": 0.4441356360912323,
      "learning_rate": 8e-05,
      "loss": 1.8176,
      "step": 4872
    },
    {
      "epoch": 0.5266400086458446,
      "grad_norm": 0.4120944142341614,
      "learning_rate": 8e-05,
      "loss": 1.7228,
      "step": 4873
    },
    {
      "epoch": 0.5267480817032314,
      "grad_norm": 0.43329861760139465,
      "learning_rate": 8e-05,
      "loss": 1.82,
      "step": 4874
    },
    {
      "epoch": 0.5268561547606182,
      "grad_norm": 0.44920217990875244,
      "learning_rate": 8e-05,
      "loss": 1.7678,
      "step": 4875
    },
    {
      "epoch": 0.526964227818005,
      "grad_norm": 0.38295778632164,
      "learning_rate": 8e-05,
      "loss": 1.5966,
      "step": 4876
    },
    {
      "epoch": 0.5270723008753918,
      "grad_norm": 0.40596696734428406,
      "learning_rate": 8e-05,
      "loss": 1.7301,
      "step": 4877
    },
    {
      "epoch": 0.5271803739327786,
      "grad_norm": 0.43769344687461853,
      "learning_rate": 8e-05,
      "loss": 1.9155,
      "step": 4878
    },
    {
      "epoch": 0.5272884469901653,
      "grad_norm": 0.443312406539917,
      "learning_rate": 8e-05,
      "loss": 1.9554,
      "step": 4879
    },
    {
      "epoch": 0.5273965200475521,
      "grad_norm": 0.40869608521461487,
      "learning_rate": 8e-05,
      "loss": 1.6775,
      "step": 4880
    },
    {
      "epoch": 0.5275045931049389,
      "grad_norm": 0.4429347515106201,
      "learning_rate": 8e-05,
      "loss": 1.6865,
      "step": 4881
    },
    {
      "epoch": 0.5276126661623257,
      "grad_norm": 0.42774441838264465,
      "learning_rate": 8e-05,
      "loss": 1.8157,
      "step": 4882
    },
    {
      "epoch": 0.5277207392197125,
      "grad_norm": 0.40076056122779846,
      "learning_rate": 8e-05,
      "loss": 1.5737,
      "step": 4883
    },
    {
      "epoch": 0.5278288122770993,
      "grad_norm": 0.39535146951675415,
      "learning_rate": 8e-05,
      "loss": 1.642,
      "step": 4884
    },
    {
      "epoch": 0.5279368853344861,
      "grad_norm": 0.42379406094551086,
      "learning_rate": 8e-05,
      "loss": 1.5001,
      "step": 4885
    },
    {
      "epoch": 0.5280449583918729,
      "grad_norm": 0.4041021466255188,
      "learning_rate": 8e-05,
      "loss": 1.7957,
      "step": 4886
    },
    {
      "epoch": 0.5281530314492597,
      "grad_norm": 0.4110598564147949,
      "learning_rate": 8e-05,
      "loss": 1.6738,
      "step": 4887
    },
    {
      "epoch": 0.5282611045066465,
      "grad_norm": 0.40826529264450073,
      "learning_rate": 8e-05,
      "loss": 1.6994,
      "step": 4888
    },
    {
      "epoch": 0.5283691775640332,
      "grad_norm": 0.43995004892349243,
      "learning_rate": 8e-05,
      "loss": 1.8347,
      "step": 4889
    },
    {
      "epoch": 0.52847725062142,
      "grad_norm": 0.4595677852630615,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 4890
    },
    {
      "epoch": 0.5285853236788068,
      "grad_norm": 0.37943774461746216,
      "learning_rate": 8e-05,
      "loss": 1.5925,
      "step": 4891
    },
    {
      "epoch": 0.5286933967361936,
      "grad_norm": 0.45692679286003113,
      "learning_rate": 8e-05,
      "loss": 1.6791,
      "step": 4892
    },
    {
      "epoch": 0.5288014697935804,
      "grad_norm": 0.4138285219669342,
      "learning_rate": 8e-05,
      "loss": 1.5395,
      "step": 4893
    },
    {
      "epoch": 0.5289095428509673,
      "grad_norm": 0.39168405532836914,
      "learning_rate": 8e-05,
      "loss": 1.6356,
      "step": 4894
    },
    {
      "epoch": 0.529017615908354,
      "grad_norm": 0.43939557671546936,
      "learning_rate": 8e-05,
      "loss": 1.8134,
      "step": 4895
    },
    {
      "epoch": 0.5291256889657409,
      "grad_norm": 0.45130473375320435,
      "learning_rate": 8e-05,
      "loss": 1.7855,
      "step": 4896
    },
    {
      "epoch": 0.5292337620231277,
      "grad_norm": 0.3797385096549988,
      "learning_rate": 8e-05,
      "loss": 1.6711,
      "step": 4897
    },
    {
      "epoch": 0.5293418350805145,
      "grad_norm": 0.406331330537796,
      "learning_rate": 8e-05,
      "loss": 1.6741,
      "step": 4898
    },
    {
      "epoch": 0.5294499081379013,
      "grad_norm": 0.3852972090244293,
      "learning_rate": 8e-05,
      "loss": 1.7018,
      "step": 4899
    },
    {
      "epoch": 0.529557981195288,
      "grad_norm": 0.409662663936615,
      "learning_rate": 8e-05,
      "loss": 1.7175,
      "step": 4900
    },
    {
      "epoch": 0.5296660542526748,
      "grad_norm": 0.46197110414505005,
      "learning_rate": 8e-05,
      "loss": 1.879,
      "step": 4901
    },
    {
      "epoch": 0.5297741273100616,
      "grad_norm": 0.39654678106307983,
      "learning_rate": 8e-05,
      "loss": 1.8084,
      "step": 4902
    },
    {
      "epoch": 0.5298822003674484,
      "grad_norm": 0.4031100273132324,
      "learning_rate": 8e-05,
      "loss": 1.5731,
      "step": 4903
    },
    {
      "epoch": 0.5299902734248352,
      "grad_norm": 0.4263877272605896,
      "learning_rate": 8e-05,
      "loss": 1.7458,
      "step": 4904
    },
    {
      "epoch": 0.530098346482222,
      "grad_norm": 0.49063023924827576,
      "learning_rate": 8e-05,
      "loss": 1.8162,
      "step": 4905
    },
    {
      "epoch": 0.5302064195396088,
      "grad_norm": 0.4183445870876312,
      "learning_rate": 8e-05,
      "loss": 1.6695,
      "step": 4906
    },
    {
      "epoch": 0.5303144925969956,
      "grad_norm": 0.3978036344051361,
      "learning_rate": 8e-05,
      "loss": 1.7124,
      "step": 4907
    },
    {
      "epoch": 0.5304225656543824,
      "grad_norm": 0.43422481417655945,
      "learning_rate": 8e-05,
      "loss": 1.7773,
      "step": 4908
    },
    {
      "epoch": 0.5305306387117692,
      "grad_norm": 0.401109516620636,
      "learning_rate": 8e-05,
      "loss": 1.7274,
      "step": 4909
    },
    {
      "epoch": 0.530638711769156,
      "grad_norm": 0.40551817417144775,
      "learning_rate": 8e-05,
      "loss": 1.6446,
      "step": 4910
    },
    {
      "epoch": 0.5307467848265427,
      "grad_norm": 0.40497076511383057,
      "learning_rate": 8e-05,
      "loss": 1.7423,
      "step": 4911
    },
    {
      "epoch": 0.5308548578839295,
      "grad_norm": 0.39855679869651794,
      "learning_rate": 8e-05,
      "loss": 1.5437,
      "step": 4912
    },
    {
      "epoch": 0.5309629309413163,
      "grad_norm": 0.42813795804977417,
      "learning_rate": 8e-05,
      "loss": 1.793,
      "step": 4913
    },
    {
      "epoch": 0.5310710039987031,
      "grad_norm": 0.4858434200286865,
      "learning_rate": 8e-05,
      "loss": 1.6226,
      "step": 4914
    },
    {
      "epoch": 0.5311790770560899,
      "grad_norm": 0.3987038731575012,
      "learning_rate": 8e-05,
      "loss": 1.4772,
      "step": 4915
    },
    {
      "epoch": 0.5312871501134767,
      "grad_norm": 0.4694141745567322,
      "learning_rate": 8e-05,
      "loss": 1.9502,
      "step": 4916
    },
    {
      "epoch": 0.5313952231708635,
      "grad_norm": 0.45072418451309204,
      "learning_rate": 8e-05,
      "loss": 1.7758,
      "step": 4917
    },
    {
      "epoch": 0.5315032962282503,
      "grad_norm": 0.42374247312545776,
      "learning_rate": 8e-05,
      "loss": 1.8641,
      "step": 4918
    },
    {
      "epoch": 0.5316113692856371,
      "grad_norm": 0.4133841395378113,
      "learning_rate": 8e-05,
      "loss": 1.8016,
      "step": 4919
    },
    {
      "epoch": 0.5317194423430239,
      "grad_norm": 0.40989622473716736,
      "learning_rate": 8e-05,
      "loss": 1.646,
      "step": 4920
    },
    {
      "epoch": 0.5318275154004107,
      "grad_norm": 0.43784958124160767,
      "learning_rate": 8e-05,
      "loss": 1.7282,
      "step": 4921
    },
    {
      "epoch": 0.5319355884577974,
      "grad_norm": 0.3818674683570862,
      "learning_rate": 8e-05,
      "loss": 1.5484,
      "step": 4922
    },
    {
      "epoch": 0.5320436615151842,
      "grad_norm": 0.3546164035797119,
      "learning_rate": 8e-05,
      "loss": 1.54,
      "step": 4923
    },
    {
      "epoch": 0.532151734572571,
      "grad_norm": 0.36899980902671814,
      "learning_rate": 8e-05,
      "loss": 1.6296,
      "step": 4924
    },
    {
      "epoch": 0.5322598076299578,
      "grad_norm": 0.4447019696235657,
      "learning_rate": 8e-05,
      "loss": 1.8838,
      "step": 4925
    },
    {
      "epoch": 0.5323678806873446,
      "grad_norm": 0.4012913405895233,
      "learning_rate": 8e-05,
      "loss": 1.6091,
      "step": 4926
    },
    {
      "epoch": 0.5324759537447314,
      "grad_norm": 0.4034128189086914,
      "learning_rate": 8e-05,
      "loss": 1.7085,
      "step": 4927
    },
    {
      "epoch": 0.5325840268021182,
      "grad_norm": 0.3941476345062256,
      "learning_rate": 8e-05,
      "loss": 1.7647,
      "step": 4928
    },
    {
      "epoch": 0.532692099859505,
      "grad_norm": 0.5080128908157349,
      "learning_rate": 8e-05,
      "loss": 1.8217,
      "step": 4929
    },
    {
      "epoch": 0.5328001729168919,
      "grad_norm": 0.41325342655181885,
      "learning_rate": 8e-05,
      "loss": 1.8287,
      "step": 4930
    },
    {
      "epoch": 0.5329082459742787,
      "grad_norm": 0.410104364156723,
      "learning_rate": 8e-05,
      "loss": 1.7823,
      "step": 4931
    },
    {
      "epoch": 0.5330163190316654,
      "grad_norm": 0.4191218316555023,
      "learning_rate": 8e-05,
      "loss": 1.708,
      "step": 4932
    },
    {
      "epoch": 0.5331243920890522,
      "grad_norm": 0.41406193375587463,
      "learning_rate": 8e-05,
      "loss": 1.733,
      "step": 4933
    },
    {
      "epoch": 0.533232465146439,
      "grad_norm": 0.3879591226577759,
      "learning_rate": 8e-05,
      "loss": 1.5891,
      "step": 4934
    },
    {
      "epoch": 0.5333405382038258,
      "grad_norm": 0.44343945384025574,
      "learning_rate": 8e-05,
      "loss": 1.73,
      "step": 4935
    },
    {
      "epoch": 0.5334486112612126,
      "grad_norm": 0.4227713644504547,
      "learning_rate": 8e-05,
      "loss": 1.8518,
      "step": 4936
    },
    {
      "epoch": 0.5335566843185994,
      "grad_norm": 0.38836869597435,
      "learning_rate": 8e-05,
      "loss": 1.6848,
      "step": 4937
    },
    {
      "epoch": 0.5336647573759862,
      "grad_norm": 0.4058745205402374,
      "learning_rate": 8e-05,
      "loss": 1.6125,
      "step": 4938
    },
    {
      "epoch": 0.533772830433373,
      "grad_norm": 0.4631666839122772,
      "learning_rate": 8e-05,
      "loss": 1.6722,
      "step": 4939
    },
    {
      "epoch": 0.5338809034907598,
      "grad_norm": 0.4189152717590332,
      "learning_rate": 8e-05,
      "loss": 1.7878,
      "step": 4940
    },
    {
      "epoch": 0.5339889765481466,
      "grad_norm": 0.42341500520706177,
      "learning_rate": 8e-05,
      "loss": 1.8066,
      "step": 4941
    },
    {
      "epoch": 0.5340970496055334,
      "grad_norm": 0.4770244061946869,
      "learning_rate": 8e-05,
      "loss": 1.8376,
      "step": 4942
    },
    {
      "epoch": 0.5342051226629201,
      "grad_norm": 0.42806220054626465,
      "learning_rate": 8e-05,
      "loss": 1.5796,
      "step": 4943
    },
    {
      "epoch": 0.5343131957203069,
      "grad_norm": 0.40327227115631104,
      "learning_rate": 8e-05,
      "loss": 1.5097,
      "step": 4944
    },
    {
      "epoch": 0.5344212687776937,
      "grad_norm": 0.376825749874115,
      "learning_rate": 8e-05,
      "loss": 1.6639,
      "step": 4945
    },
    {
      "epoch": 0.5345293418350805,
      "grad_norm": 0.4043944478034973,
      "learning_rate": 8e-05,
      "loss": 1.5462,
      "step": 4946
    },
    {
      "epoch": 0.5346374148924673,
      "grad_norm": 0.4299601912498474,
      "learning_rate": 8e-05,
      "loss": 1.8006,
      "step": 4947
    },
    {
      "epoch": 0.5347454879498541,
      "grad_norm": 0.3913020193576813,
      "learning_rate": 8e-05,
      "loss": 1.6188,
      "step": 4948
    },
    {
      "epoch": 0.5348535610072409,
      "grad_norm": 0.42924466729164124,
      "learning_rate": 8e-05,
      "loss": 1.8771,
      "step": 4949
    },
    {
      "epoch": 0.5349616340646277,
      "grad_norm": 0.41477829217910767,
      "learning_rate": 8e-05,
      "loss": 1.6529,
      "step": 4950
    },
    {
      "epoch": 0.5350697071220145,
      "grad_norm": 0.442656010389328,
      "learning_rate": 8e-05,
      "loss": 1.6115,
      "step": 4951
    },
    {
      "epoch": 0.5351777801794013,
      "grad_norm": 0.39936596155166626,
      "learning_rate": 8e-05,
      "loss": 1.6037,
      "step": 4952
    },
    {
      "epoch": 0.5352858532367881,
      "grad_norm": 0.4142223596572876,
      "learning_rate": 8e-05,
      "loss": 1.6794,
      "step": 4953
    },
    {
      "epoch": 0.5353939262941748,
      "grad_norm": 0.38259631395339966,
      "learning_rate": 8e-05,
      "loss": 1.6751,
      "step": 4954
    },
    {
      "epoch": 0.5355019993515616,
      "grad_norm": 0.4126985967159271,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 4955
    },
    {
      "epoch": 0.5356100724089484,
      "grad_norm": 0.4238603711128235,
      "learning_rate": 8e-05,
      "loss": 1.7354,
      "step": 4956
    },
    {
      "epoch": 0.5357181454663352,
      "grad_norm": 0.4083351790904999,
      "learning_rate": 8e-05,
      "loss": 1.5988,
      "step": 4957
    },
    {
      "epoch": 0.535826218523722,
      "grad_norm": 0.40220242738723755,
      "learning_rate": 8e-05,
      "loss": 1.705,
      "step": 4958
    },
    {
      "epoch": 0.5359342915811088,
      "grad_norm": 0.40210792422294617,
      "learning_rate": 8e-05,
      "loss": 1.7388,
      "step": 4959
    },
    {
      "epoch": 0.5360423646384956,
      "grad_norm": 0.4350391924381256,
      "learning_rate": 8e-05,
      "loss": 1.7287,
      "step": 4960
    },
    {
      "epoch": 0.5361504376958824,
      "grad_norm": 0.4623953402042389,
      "learning_rate": 8e-05,
      "loss": 1.8561,
      "step": 4961
    },
    {
      "epoch": 0.5362585107532692,
      "grad_norm": 0.39749476313591003,
      "learning_rate": 8e-05,
      "loss": 1.674,
      "step": 4962
    },
    {
      "epoch": 0.536366583810656,
      "grad_norm": 0.4659527838230133,
      "learning_rate": 8e-05,
      "loss": 1.699,
      "step": 4963
    },
    {
      "epoch": 0.5364746568680427,
      "grad_norm": 0.46200257539749146,
      "learning_rate": 8e-05,
      "loss": 1.7666,
      "step": 4964
    },
    {
      "epoch": 0.5365827299254295,
      "grad_norm": 0.4098494052886963,
      "learning_rate": 8e-05,
      "loss": 1.7086,
      "step": 4965
    },
    {
      "epoch": 0.5366908029828164,
      "grad_norm": 0.41835978627204895,
      "learning_rate": 8e-05,
      "loss": 1.7408,
      "step": 4966
    },
    {
      "epoch": 0.5367988760402032,
      "grad_norm": 0.39188632369041443,
      "learning_rate": 8e-05,
      "loss": 1.8317,
      "step": 4967
    },
    {
      "epoch": 0.53690694909759,
      "grad_norm": 0.39533039927482605,
      "learning_rate": 8e-05,
      "loss": 1.7549,
      "step": 4968
    },
    {
      "epoch": 0.5370150221549768,
      "grad_norm": 0.42061132192611694,
      "learning_rate": 8e-05,
      "loss": 1.723,
      "step": 4969
    },
    {
      "epoch": 0.5371230952123636,
      "grad_norm": 0.43010419607162476,
      "learning_rate": 8e-05,
      "loss": 1.8395,
      "step": 4970
    },
    {
      "epoch": 0.5372311682697504,
      "grad_norm": 0.38996654748916626,
      "learning_rate": 8e-05,
      "loss": 1.6134,
      "step": 4971
    },
    {
      "epoch": 0.5373392413271372,
      "grad_norm": 0.41649290919303894,
      "learning_rate": 8e-05,
      "loss": 1.6993,
      "step": 4972
    },
    {
      "epoch": 0.537447314384524,
      "grad_norm": 0.4055330157279968,
      "learning_rate": 8e-05,
      "loss": 1.6144,
      "step": 4973
    },
    {
      "epoch": 0.5375553874419108,
      "grad_norm": 0.408184677362442,
      "learning_rate": 8e-05,
      "loss": 1.7796,
      "step": 4974
    },
    {
      "epoch": 0.5376634604992975,
      "grad_norm": 0.3943346440792084,
      "learning_rate": 8e-05,
      "loss": 1.6991,
      "step": 4975
    },
    {
      "epoch": 0.5377715335566843,
      "grad_norm": 0.4003731608390808,
      "learning_rate": 8e-05,
      "loss": 1.7364,
      "step": 4976
    },
    {
      "epoch": 0.5378796066140711,
      "grad_norm": 0.41543933749198914,
      "learning_rate": 8e-05,
      "loss": 1.5941,
      "step": 4977
    },
    {
      "epoch": 0.5379876796714579,
      "grad_norm": 0.4118979871273041,
      "learning_rate": 8e-05,
      "loss": 1.7209,
      "step": 4978
    },
    {
      "epoch": 0.5380957527288447,
      "grad_norm": 0.4620661735534668,
      "learning_rate": 8e-05,
      "loss": 1.7529,
      "step": 4979
    },
    {
      "epoch": 0.5382038257862315,
      "grad_norm": 0.41683709621429443,
      "learning_rate": 8e-05,
      "loss": 1.7604,
      "step": 4980
    },
    {
      "epoch": 0.5383118988436183,
      "grad_norm": 0.37552160024642944,
      "learning_rate": 8e-05,
      "loss": 1.5967,
      "step": 4981
    },
    {
      "epoch": 0.5384199719010051,
      "grad_norm": 0.40685006976127625,
      "learning_rate": 8e-05,
      "loss": 1.5256,
      "step": 4982
    },
    {
      "epoch": 0.5385280449583919,
      "grad_norm": 0.4516046345233917,
      "learning_rate": 8e-05,
      "loss": 1.8799,
      "step": 4983
    },
    {
      "epoch": 0.5386361180157787,
      "grad_norm": 0.39778828620910645,
      "learning_rate": 8e-05,
      "loss": 1.7203,
      "step": 4984
    },
    {
      "epoch": 0.5387441910731655,
      "grad_norm": 0.41983702778816223,
      "learning_rate": 8e-05,
      "loss": 1.6829,
      "step": 4985
    },
    {
      "epoch": 0.5388522641305522,
      "grad_norm": 0.39821866154670715,
      "learning_rate": 8e-05,
      "loss": 1.7242,
      "step": 4986
    },
    {
      "epoch": 0.538960337187939,
      "grad_norm": 0.3996838629245758,
      "learning_rate": 8e-05,
      "loss": 1.816,
      "step": 4987
    },
    {
      "epoch": 0.5390684102453258,
      "grad_norm": 0.41400259733200073,
      "learning_rate": 8e-05,
      "loss": 1.6538,
      "step": 4988
    },
    {
      "epoch": 0.5391764833027126,
      "grad_norm": 0.4749830961227417,
      "learning_rate": 8e-05,
      "loss": 1.9272,
      "step": 4989
    },
    {
      "epoch": 0.5392845563600994,
      "grad_norm": 0.40028098225593567,
      "learning_rate": 8e-05,
      "loss": 1.5645,
      "step": 4990
    },
    {
      "epoch": 0.5393926294174862,
      "grad_norm": 0.4243418872356415,
      "learning_rate": 8e-05,
      "loss": 1.7819,
      "step": 4991
    },
    {
      "epoch": 0.539500702474873,
      "grad_norm": 0.42156633734703064,
      "learning_rate": 8e-05,
      "loss": 1.7393,
      "step": 4992
    },
    {
      "epoch": 0.5396087755322598,
      "grad_norm": 0.4170660078525543,
      "learning_rate": 8e-05,
      "loss": 1.8224,
      "step": 4993
    },
    {
      "epoch": 0.5397168485896466,
      "grad_norm": 0.4410928189754486,
      "learning_rate": 8e-05,
      "loss": 1.8482,
      "step": 4994
    },
    {
      "epoch": 0.5398249216470334,
      "grad_norm": 0.45637479424476624,
      "learning_rate": 8e-05,
      "loss": 1.9589,
      "step": 4995
    },
    {
      "epoch": 0.5399329947044201,
      "grad_norm": 0.3825331926345825,
      "learning_rate": 8e-05,
      "loss": 1.5749,
      "step": 4996
    },
    {
      "epoch": 0.5400410677618069,
      "grad_norm": 0.4709583818912506,
      "learning_rate": 8e-05,
      "loss": 1.8035,
      "step": 4997
    },
    {
      "epoch": 0.5401491408191937,
      "grad_norm": 0.44993558526039124,
      "learning_rate": 8e-05,
      "loss": 1.8077,
      "step": 4998
    },
    {
      "epoch": 0.5402572138765805,
      "grad_norm": 0.44802701473236084,
      "learning_rate": 8e-05,
      "loss": 1.6869,
      "step": 4999
    },
    {
      "epoch": 0.5403652869339673,
      "grad_norm": 0.4343043267726898,
      "learning_rate": 8e-05,
      "loss": 1.739,
      "step": 5000
    }
  ],
  "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": 7.46010452164608e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}