{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.2932551319648094,
  "eval_steps": 500,
  "global_step": 700,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.00041893590280687055,
      "grad_norm": 8.333969116210938,
      "learning_rate": 2e-05,
      "loss": 3.3261,
      "step": 1
    },
    {
      "epoch": 0.0008378718056137411,
      "grad_norm": 7.348582744598389,
      "learning_rate": 1.9998324256388776e-05,
      "loss": 3.2312,
      "step": 2
    },
    {
      "epoch": 0.0012568077084206116,
      "grad_norm": 3.1899449825286865,
      "learning_rate": 1.9996648512777547e-05,
      "loss": 2.9318,
      "step": 3
    },
    {
      "epoch": 0.0016757436112274822,
      "grad_norm": 2.177359104156494,
      "learning_rate": 1.9994972769166317e-05,
      "loss": 2.8876,
      "step": 4
    },
    {
      "epoch": 0.0020946795140343527,
      "grad_norm": 2.1708009243011475,
      "learning_rate": 1.999329702555509e-05,
      "loss": 3.068,
      "step": 5
    },
    {
      "epoch": 0.0025136154168412233,
      "grad_norm": 1.77362060546875,
      "learning_rate": 1.9991621281943865e-05,
      "loss": 2.6598,
      "step": 6
    },
    {
      "epoch": 0.002932551319648094,
      "grad_norm": 1.709831714630127,
      "learning_rate": 1.9989945538332636e-05,
      "loss": 2.767,
      "step": 7
    },
    {
      "epoch": 0.0033514872224549644,
      "grad_norm": 1.5493128299713135,
      "learning_rate": 1.998826979472141e-05,
      "loss": 2.7182,
      "step": 8
    },
    {
      "epoch": 0.003770423125261835,
      "grad_norm": 1.39557945728302,
      "learning_rate": 1.998659405111018e-05,
      "loss": 2.8695,
      "step": 9
    },
    {
      "epoch": 0.0041893590280687055,
      "grad_norm": 1.1357755661010742,
      "learning_rate": 1.9984918307498955e-05,
      "loss": 2.5497,
      "step": 10
    },
    {
      "epoch": 0.004608294930875576,
      "grad_norm": 1.0683544874191284,
      "learning_rate": 1.9983242563887726e-05,
      "loss": 2.7363,
      "step": 11
    },
    {
      "epoch": 0.005027230833682447,
      "grad_norm": 0.9199109673500061,
      "learning_rate": 1.99815668202765e-05,
      "loss": 2.4438,
      "step": 12
    },
    {
      "epoch": 0.005446166736489317,
      "grad_norm": 0.9708887338638306,
      "learning_rate": 1.997989107666527e-05,
      "loss": 2.6098,
      "step": 13
    },
    {
      "epoch": 0.005865102639296188,
      "grad_norm": 0.9026995301246643,
      "learning_rate": 1.9978215333054045e-05,
      "loss": 2.4444,
      "step": 14
    },
    {
      "epoch": 0.006284038542103058,
      "grad_norm": 1.2418183088302612,
      "learning_rate": 1.9976539589442816e-05,
      "loss": 2.5436,
      "step": 15
    },
    {
      "epoch": 0.006702974444909929,
      "grad_norm": 0.8761052489280701,
      "learning_rate": 1.997486384583159e-05,
      "loss": 2.3926,
      "step": 16
    },
    {
      "epoch": 0.007121910347716799,
      "grad_norm": 0.8849633932113647,
      "learning_rate": 1.997318810222036e-05,
      "loss": 2.5935,
      "step": 17
    },
    {
      "epoch": 0.00754084625052367,
      "grad_norm": 0.8368175029754639,
      "learning_rate": 1.9971512358609135e-05,
      "loss": 2.4684,
      "step": 18
    },
    {
      "epoch": 0.007959782153330541,
      "grad_norm": 0.697807788848877,
      "learning_rate": 1.9969836614997905e-05,
      "loss": 2.3537,
      "step": 19
    },
    {
      "epoch": 0.008378718056137411,
      "grad_norm": 0.7476556301116943,
      "learning_rate": 1.996816087138668e-05,
      "loss": 2.4659,
      "step": 20
    },
    {
      "epoch": 0.008797653958944282,
      "grad_norm": 0.7885666489601135,
      "learning_rate": 1.9966485127775454e-05,
      "loss": 2.4466,
      "step": 21
    },
    {
      "epoch": 0.009216589861751152,
      "grad_norm": 0.7271686792373657,
      "learning_rate": 1.9964809384164224e-05,
      "loss": 2.3659,
      "step": 22
    },
    {
      "epoch": 0.009635525764558023,
      "grad_norm": 0.7286465764045715,
      "learning_rate": 1.9963133640552995e-05,
      "loss": 2.4473,
      "step": 23
    },
    {
      "epoch": 0.010054461667364893,
      "grad_norm": 0.8572853207588196,
      "learning_rate": 1.996145789694177e-05,
      "loss": 2.3595,
      "step": 24
    },
    {
      "epoch": 0.010473397570171765,
      "grad_norm": 0.8283334374427795,
      "learning_rate": 1.9959782153330543e-05,
      "loss": 2.5291,
      "step": 25
    },
    {
      "epoch": 0.010892333472978634,
      "grad_norm": 0.6586313843727112,
      "learning_rate": 1.9958106409719314e-05,
      "loss": 2.3985,
      "step": 26
    },
    {
      "epoch": 0.011311269375785506,
      "grad_norm": 0.6430657505989075,
      "learning_rate": 1.9956430666108085e-05,
      "loss": 2.3578,
      "step": 27
    },
    {
      "epoch": 0.011730205278592375,
      "grad_norm": 0.6550448536872864,
      "learning_rate": 1.995475492249686e-05,
      "loss": 2.4077,
      "step": 28
    },
    {
      "epoch": 0.012149141181399247,
      "grad_norm": 0.7592840194702148,
      "learning_rate": 1.9953079178885633e-05,
      "loss": 2.5008,
      "step": 29
    },
    {
      "epoch": 0.012568077084206116,
      "grad_norm": 0.7858672738075256,
      "learning_rate": 1.9951403435274407e-05,
      "loss": 2.4674,
      "step": 30
    },
    {
      "epoch": 0.012987012987012988,
      "grad_norm": 0.6130352020263672,
      "learning_rate": 1.9949727691663178e-05,
      "loss": 2.3526,
      "step": 31
    },
    {
      "epoch": 0.013405948889819858,
      "grad_norm": 0.6684207320213318,
      "learning_rate": 1.994805194805195e-05,
      "loss": 2.3732,
      "step": 32
    },
    {
      "epoch": 0.013824884792626729,
      "grad_norm": 0.8275600671768188,
      "learning_rate": 1.9946376204440723e-05,
      "loss": 2.135,
      "step": 33
    },
    {
      "epoch": 0.014243820695433599,
      "grad_norm": 0.5858725309371948,
      "learning_rate": 1.9944700460829494e-05,
      "loss": 2.1368,
      "step": 34
    },
    {
      "epoch": 0.01466275659824047,
      "grad_norm": 0.7133444547653198,
      "learning_rate": 1.9943024717218268e-05,
      "loss": 2.304,
      "step": 35
    },
    {
      "epoch": 0.01508169250104734,
      "grad_norm": 0.5466803312301636,
      "learning_rate": 1.994134897360704e-05,
      "loss": 2.1682,
      "step": 36
    },
    {
      "epoch": 0.015500628403854211,
      "grad_norm": 0.5196086168289185,
      "learning_rate": 1.9939673229995813e-05,
      "loss": 2.1546,
      "step": 37
    },
    {
      "epoch": 0.015919564306661083,
      "grad_norm": 0.5088497400283813,
      "learning_rate": 1.9937997486384583e-05,
      "loss": 2.1018,
      "step": 38
    },
    {
      "epoch": 0.016338500209467952,
      "grad_norm": 0.6117899417877197,
      "learning_rate": 1.9936321742773358e-05,
      "loss": 2.2346,
      "step": 39
    },
    {
      "epoch": 0.016757436112274822,
      "grad_norm": 0.5710458159446716,
      "learning_rate": 1.993464599916213e-05,
      "loss": 2.2147,
      "step": 40
    },
    {
      "epoch": 0.01717637201508169,
      "grad_norm": 0.5152861475944519,
      "learning_rate": 1.9932970255550902e-05,
      "loss": 2.2716,
      "step": 41
    },
    {
      "epoch": 0.017595307917888565,
      "grad_norm": 0.6851192712783813,
      "learning_rate": 1.9931294511939673e-05,
      "loss": 2.3158,
      "step": 42
    },
    {
      "epoch": 0.018014243820695434,
      "grad_norm": 0.5485531687736511,
      "learning_rate": 1.9929618768328447e-05,
      "loss": 2.2679,
      "step": 43
    },
    {
      "epoch": 0.018433179723502304,
      "grad_norm": 0.48592010140419006,
      "learning_rate": 1.992794302471722e-05,
      "loss": 2.1303,
      "step": 44
    },
    {
      "epoch": 0.018852115626309174,
      "grad_norm": 0.5533665418624878,
      "learning_rate": 1.9926267281105992e-05,
      "loss": 2.1981,
      "step": 45
    },
    {
      "epoch": 0.019271051529116047,
      "grad_norm": 0.5932656526565552,
      "learning_rate": 1.9924591537494763e-05,
      "loss": 2.3737,
      "step": 46
    },
    {
      "epoch": 0.019689987431922917,
      "grad_norm": 0.5236673951148987,
      "learning_rate": 1.9922915793883537e-05,
      "loss": 2.2694,
      "step": 47
    },
    {
      "epoch": 0.020108923334729786,
      "grad_norm": 0.5357316732406616,
      "learning_rate": 1.992124005027231e-05,
      "loss": 2.2368,
      "step": 48
    },
    {
      "epoch": 0.020527859237536656,
      "grad_norm": 0.5500349998474121,
      "learning_rate": 1.9919564306661082e-05,
      "loss": 2.213,
      "step": 49
    },
    {
      "epoch": 0.02094679514034353,
      "grad_norm": 0.48040810227394104,
      "learning_rate": 1.9917888563049853e-05,
      "loss": 2.1892,
      "step": 50
    },
    {
      "epoch": 0.0213657310431504,
      "grad_norm": 0.5716186165809631,
      "learning_rate": 1.9916212819438627e-05,
      "loss": 2.2039,
      "step": 51
    },
    {
      "epoch": 0.02178466694595727,
      "grad_norm": 0.5564374923706055,
      "learning_rate": 1.99145370758274e-05,
      "loss": 2.1411,
      "step": 52
    },
    {
      "epoch": 0.022203602848764138,
      "grad_norm": 0.4996980130672455,
      "learning_rate": 1.9912861332216175e-05,
      "loss": 2.1521,
      "step": 53
    },
    {
      "epoch": 0.02262253875157101,
      "grad_norm": 0.5239240527153015,
      "learning_rate": 1.9911185588604946e-05,
      "loss": 2.0742,
      "step": 54
    },
    {
      "epoch": 0.02304147465437788,
      "grad_norm": 0.4403076767921448,
      "learning_rate": 1.9909509844993716e-05,
      "loss": 1.9841,
      "step": 55
    },
    {
      "epoch": 0.02346041055718475,
      "grad_norm": 0.5169032216072083,
      "learning_rate": 1.990783410138249e-05,
      "loss": 2.0327,
      "step": 56
    },
    {
      "epoch": 0.02387934645999162,
      "grad_norm": 0.4901898503303528,
      "learning_rate": 1.9906158357771265e-05,
      "loss": 2.0063,
      "step": 57
    },
    {
      "epoch": 0.024298282362798494,
      "grad_norm": 0.6581910252571106,
      "learning_rate": 1.9904482614160035e-05,
      "loss": 2.1385,
      "step": 58
    },
    {
      "epoch": 0.024717218265605363,
      "grad_norm": 0.4522070586681366,
      "learning_rate": 1.9902806870548806e-05,
      "loss": 1.9944,
      "step": 59
    },
    {
      "epoch": 0.025136154168412233,
      "grad_norm": 0.5315820574760437,
      "learning_rate": 1.990113112693758e-05,
      "loss": 2.1579,
      "step": 60
    },
    {
      "epoch": 0.025555090071219103,
      "grad_norm": 0.4661259353160858,
      "learning_rate": 1.9899455383326354e-05,
      "loss": 2.1193,
      "step": 61
    },
    {
      "epoch": 0.025974025974025976,
      "grad_norm": 0.4940222203731537,
      "learning_rate": 1.9897779639715125e-05,
      "loss": 2.0844,
      "step": 62
    },
    {
      "epoch": 0.026392961876832845,
      "grad_norm": 0.46520665287971497,
      "learning_rate": 1.98961038961039e-05,
      "loss": 1.9306,
      "step": 63
    },
    {
      "epoch": 0.026811897779639715,
      "grad_norm": 0.5645989179611206,
      "learning_rate": 1.989442815249267e-05,
      "loss": 2.1236,
      "step": 64
    },
    {
      "epoch": 0.027230833682446585,
      "grad_norm": 0.47880157828330994,
      "learning_rate": 1.989275240888144e-05,
      "loss": 2.0206,
      "step": 65
    },
    {
      "epoch": 0.027649769585253458,
      "grad_norm": 0.6371349692344666,
      "learning_rate": 1.9891076665270215e-05,
      "loss": 2.019,
      "step": 66
    },
    {
      "epoch": 0.028068705488060328,
      "grad_norm": 0.5742272734642029,
      "learning_rate": 1.988940092165899e-05,
      "loss": 2.0899,
      "step": 67
    },
    {
      "epoch": 0.028487641390867197,
      "grad_norm": 0.5579768419265747,
      "learning_rate": 1.988772517804776e-05,
      "loss": 2.081,
      "step": 68
    },
    {
      "epoch": 0.028906577293674067,
      "grad_norm": 0.5897182822227478,
      "learning_rate": 1.988604943443653e-05,
      "loss": 1.9601,
      "step": 69
    },
    {
      "epoch": 0.02932551319648094,
      "grad_norm": 0.46881428360939026,
      "learning_rate": 1.9884373690825305e-05,
      "loss": 1.9085,
      "step": 70
    },
    {
      "epoch": 0.02974444909928781,
      "grad_norm": 0.6095844507217407,
      "learning_rate": 1.988269794721408e-05,
      "loss": 1.9762,
      "step": 71
    },
    {
      "epoch": 0.03016338500209468,
      "grad_norm": 0.599513053894043,
      "learning_rate": 1.988102220360285e-05,
      "loss": 1.8723,
      "step": 72
    },
    {
      "epoch": 0.03058232090490155,
      "grad_norm": 0.585457980632782,
      "learning_rate": 1.987934645999162e-05,
      "loss": 1.9209,
      "step": 73
    },
    {
      "epoch": 0.031001256807708422,
      "grad_norm": 0.42224225401878357,
      "learning_rate": 1.9877670716380394e-05,
      "loss": 1.9186,
      "step": 74
    },
    {
      "epoch": 0.03142019271051529,
      "grad_norm": 0.4566991329193115,
      "learning_rate": 1.987599497276917e-05,
      "loss": 2.018,
      "step": 75
    },
    {
      "epoch": 0.031839128613322165,
      "grad_norm": 0.47718995809555054,
      "learning_rate": 1.9874319229157943e-05,
      "loss": 2.0119,
      "step": 76
    },
    {
      "epoch": 0.03225806451612903,
      "grad_norm": 0.4412285089492798,
      "learning_rate": 1.9872643485546713e-05,
      "loss": 1.9211,
      "step": 77
    },
    {
      "epoch": 0.032677000418935905,
      "grad_norm": 0.4711454212665558,
      "learning_rate": 1.9870967741935484e-05,
      "loss": 1.9051,
      "step": 78
    },
    {
      "epoch": 0.03309593632174277,
      "grad_norm": 0.4665948450565338,
      "learning_rate": 1.9869291998324258e-05,
      "loss": 1.9571,
      "step": 79
    },
    {
      "epoch": 0.033514872224549644,
      "grad_norm": 0.46011775732040405,
      "learning_rate": 1.9867616254713032e-05,
      "loss": 1.9599,
      "step": 80
    },
    {
      "epoch": 0.03393380812735652,
      "grad_norm": 0.46272069215774536,
      "learning_rate": 1.9865940511101803e-05,
      "loss": 1.9161,
      "step": 81
    },
    {
      "epoch": 0.03435274403016338,
      "grad_norm": 0.5554195046424866,
      "learning_rate": 1.9864264767490574e-05,
      "loss": 2.0202,
      "step": 82
    },
    {
      "epoch": 0.034771679932970256,
      "grad_norm": 0.5324104428291321,
      "learning_rate": 1.9862589023879348e-05,
      "loss": 1.9356,
      "step": 83
    },
    {
      "epoch": 0.03519061583577713,
      "grad_norm": 0.5279750823974609,
      "learning_rate": 1.9860913280268122e-05,
      "loss": 1.9511,
      "step": 84
    },
    {
      "epoch": 0.035609551738583996,
      "grad_norm": 0.5002080202102661,
      "learning_rate": 1.9859237536656893e-05,
      "loss": 1.9248,
      "step": 85
    },
    {
      "epoch": 0.03602848764139087,
      "grad_norm": 0.5625497102737427,
      "learning_rate": 1.9857561793045667e-05,
      "loss": 2.0023,
      "step": 86
    },
    {
      "epoch": 0.036447423544197735,
      "grad_norm": 0.6030247807502747,
      "learning_rate": 1.9855886049434438e-05,
      "loss": 1.893,
      "step": 87
    },
    {
      "epoch": 0.03686635944700461,
      "grad_norm": 0.4760509729385376,
      "learning_rate": 1.9854210305823212e-05,
      "loss": 1.8902,
      "step": 88
    },
    {
      "epoch": 0.03728529534981148,
      "grad_norm": 0.6618624925613403,
      "learning_rate": 1.9852534562211983e-05,
      "loss": 1.9173,
      "step": 89
    },
    {
      "epoch": 0.03770423125261835,
      "grad_norm": 0.47204822301864624,
      "learning_rate": 1.9850858818600757e-05,
      "loss": 1.9266,
      "step": 90
    },
    {
      "epoch": 0.03812316715542522,
      "grad_norm": 0.5421533584594727,
      "learning_rate": 1.9849183074989527e-05,
      "loss": 1.9796,
      "step": 91
    },
    {
      "epoch": 0.038542103058232094,
      "grad_norm": 0.48972201347351074,
      "learning_rate": 1.98475073313783e-05,
      "loss": 1.91,
      "step": 92
    },
    {
      "epoch": 0.03896103896103896,
      "grad_norm": 0.5566658973693848,
      "learning_rate": 1.9845831587767072e-05,
      "loss": 1.8992,
      "step": 93
    },
    {
      "epoch": 0.03937997486384583,
      "grad_norm": 0.4685937464237213,
      "learning_rate": 1.9844155844155846e-05,
      "loss": 1.9231,
      "step": 94
    },
    {
      "epoch": 0.0397989107666527,
      "grad_norm": 0.6744531393051147,
      "learning_rate": 1.9842480100544617e-05,
      "loss": 1.9109,
      "step": 95
    },
    {
      "epoch": 0.04021784666945957,
      "grad_norm": 0.6984325051307678,
      "learning_rate": 1.984080435693339e-05,
      "loss": 1.9566,
      "step": 96
    },
    {
      "epoch": 0.040636782572266446,
      "grad_norm": 0.6627328991889954,
      "learning_rate": 1.9839128613322162e-05,
      "loss": 1.9933,
      "step": 97
    },
    {
      "epoch": 0.04105571847507331,
      "grad_norm": 0.4586343765258789,
      "learning_rate": 1.9837452869710936e-05,
      "loss": 1.7939,
      "step": 98
    },
    {
      "epoch": 0.041474654377880185,
      "grad_norm": 0.6211162805557251,
      "learning_rate": 1.983577712609971e-05,
      "loss": 1.9164,
      "step": 99
    },
    {
      "epoch": 0.04189359028068706,
      "grad_norm": 0.9397639632225037,
      "learning_rate": 1.983410138248848e-05,
      "loss": 2.0262,
      "step": 100
    },
    {
      "epoch": 0.042312526183493925,
      "grad_norm": 0.7698065638542175,
      "learning_rate": 1.9832425638877252e-05,
      "loss": 1.8979,
      "step": 101
    },
    {
      "epoch": 0.0427314620863008,
      "grad_norm": 0.5800043940544128,
      "learning_rate": 1.9830749895266026e-05,
      "loss": 1.9483,
      "step": 102
    },
    {
      "epoch": 0.043150397989107664,
      "grad_norm": 0.7634892463684082,
      "learning_rate": 1.98290741516548e-05,
      "loss": 1.7777,
      "step": 103
    },
    {
      "epoch": 0.04356933389191454,
      "grad_norm": 0.5963580012321472,
      "learning_rate": 1.982739840804357e-05,
      "loss": 1.9406,
      "step": 104
    },
    {
      "epoch": 0.04398826979472141,
      "grad_norm": 0.6970496773719788,
      "learning_rate": 1.982572266443234e-05,
      "loss": 1.967,
      "step": 105
    },
    {
      "epoch": 0.044407205697528276,
      "grad_norm": 0.5826534032821655,
      "learning_rate": 1.9824046920821116e-05,
      "loss": 1.8577,
      "step": 106
    },
    {
      "epoch": 0.04482614160033515,
      "grad_norm": 0.60413658618927,
      "learning_rate": 1.982237117720989e-05,
      "loss": 1.9528,
      "step": 107
    },
    {
      "epoch": 0.04524507750314202,
      "grad_norm": 0.7267922759056091,
      "learning_rate": 1.982069543359866e-05,
      "loss": 1.7067,
      "step": 108
    },
    {
      "epoch": 0.04566401340594889,
      "grad_norm": 0.6376165747642517,
      "learning_rate": 1.9819019689987435e-05,
      "loss": 1.9888,
      "step": 109
    },
    {
      "epoch": 0.04608294930875576,
      "grad_norm": 0.5887031555175781,
      "learning_rate": 1.9817343946376205e-05,
      "loss": 1.7777,
      "step": 110
    },
    {
      "epoch": 0.04650188521156263,
      "grad_norm": 0.6548938155174255,
      "learning_rate": 1.981566820276498e-05,
      "loss": 1.9222,
      "step": 111
    },
    {
      "epoch": 0.0469208211143695,
      "grad_norm": 0.5757064819335938,
      "learning_rate": 1.981399245915375e-05,
      "loss": 1.844,
      "step": 112
    },
    {
      "epoch": 0.047339757017176375,
      "grad_norm": 0.7597166895866394,
      "learning_rate": 1.9812316715542524e-05,
      "loss": 1.8678,
      "step": 113
    },
    {
      "epoch": 0.04775869291998324,
      "grad_norm": 0.5536984801292419,
      "learning_rate": 1.9810640971931295e-05,
      "loss": 1.8502,
      "step": 114
    },
    {
      "epoch": 0.048177628822790114,
      "grad_norm": 0.5753149390220642,
      "learning_rate": 1.980896522832007e-05,
      "loss": 1.8613,
      "step": 115
    },
    {
      "epoch": 0.04859656472559699,
      "grad_norm": 0.6214611530303955,
      "learning_rate": 1.980728948470884e-05,
      "loss": 1.8803,
      "step": 116
    },
    {
      "epoch": 0.04901550062840385,
      "grad_norm": 0.5892764329910278,
      "learning_rate": 1.9805613741097614e-05,
      "loss": 1.8241,
      "step": 117
    },
    {
      "epoch": 0.049434436531210726,
      "grad_norm": 0.5623623132705688,
      "learning_rate": 1.9803937997486388e-05,
      "loss": 1.8584,
      "step": 118
    },
    {
      "epoch": 0.04985337243401759,
      "grad_norm": 0.5206480622291565,
      "learning_rate": 1.980226225387516e-05,
      "loss": 1.8058,
      "step": 119
    },
    {
      "epoch": 0.050272308336824466,
      "grad_norm": 0.7416813373565674,
      "learning_rate": 1.980058651026393e-05,
      "loss": 1.7422,
      "step": 120
    },
    {
      "epoch": 0.05069124423963134,
      "grad_norm": 0.6095878481864929,
      "learning_rate": 1.9798910766652704e-05,
      "loss": 1.7567,
      "step": 121
    },
    {
      "epoch": 0.051110180142438205,
      "grad_norm": 0.5830249786376953,
      "learning_rate": 1.9797235023041478e-05,
      "loss": 1.8099,
      "step": 122
    },
    {
      "epoch": 0.05152911604524508,
      "grad_norm": 0.7603867053985596,
      "learning_rate": 1.979555927943025e-05,
      "loss": 1.8074,
      "step": 123
    },
    {
      "epoch": 0.05194805194805195,
      "grad_norm": 0.607905387878418,
      "learning_rate": 1.979388353581902e-05,
      "loss": 1.8359,
      "step": 124
    },
    {
      "epoch": 0.05236698785085882,
      "grad_norm": 0.5446661710739136,
      "learning_rate": 1.9792207792207794e-05,
      "loss": 1.7291,
      "step": 125
    },
    {
      "epoch": 0.05278592375366569,
      "grad_norm": 0.5527285933494568,
      "learning_rate": 1.9790532048596568e-05,
      "loss": 1.7841,
      "step": 126
    },
    {
      "epoch": 0.05320485965647256,
      "grad_norm": 0.6565405130386353,
      "learning_rate": 1.978885630498534e-05,
      "loss": 1.9168,
      "step": 127
    },
    {
      "epoch": 0.05362379555927943,
      "grad_norm": 0.5959405899047852,
      "learning_rate": 1.978718056137411e-05,
      "loss": 1.8735,
      "step": 128
    },
    {
      "epoch": 0.0540427314620863,
      "grad_norm": 0.7685437202453613,
      "learning_rate": 1.9785504817762883e-05,
      "loss": 1.782,
      "step": 129
    },
    {
      "epoch": 0.05446166736489317,
      "grad_norm": 0.5747430324554443,
      "learning_rate": 1.9783829074151657e-05,
      "loss": 1.7831,
      "step": 130
    },
    {
      "epoch": 0.05488060326770004,
      "grad_norm": 0.7328975200653076,
      "learning_rate": 1.9782153330540428e-05,
      "loss": 1.7451,
      "step": 131
    },
    {
      "epoch": 0.055299539170506916,
      "grad_norm": 0.5662095546722412,
      "learning_rate": 1.9780477586929202e-05,
      "loss": 1.6731,
      "step": 132
    },
    {
      "epoch": 0.05571847507331378,
      "grad_norm": 0.6165090799331665,
      "learning_rate": 1.9778801843317973e-05,
      "loss": 1.7787,
      "step": 133
    },
    {
      "epoch": 0.056137410976120655,
      "grad_norm": 0.6399924755096436,
      "learning_rate": 1.9777126099706747e-05,
      "loss": 1.8697,
      "step": 134
    },
    {
      "epoch": 0.05655634687892752,
      "grad_norm": 0.6513495445251465,
      "learning_rate": 1.9775450356095518e-05,
      "loss": 1.7885,
      "step": 135
    },
    {
      "epoch": 0.056975282781734395,
      "grad_norm": 0.652104914188385,
      "learning_rate": 1.9773774612484292e-05,
      "loss": 1.8462,
      "step": 136
    },
    {
      "epoch": 0.05739421868454127,
      "grad_norm": 0.8418712615966797,
      "learning_rate": 1.9772098868873063e-05,
      "loss": 1.8771,
      "step": 137
    },
    {
      "epoch": 0.057813154587348134,
      "grad_norm": 0.6133163571357727,
      "learning_rate": 1.9770423125261837e-05,
      "loss": 1.7998,
      "step": 138
    },
    {
      "epoch": 0.05823209049015501,
      "grad_norm": 0.6718358993530273,
      "learning_rate": 1.9768747381650608e-05,
      "loss": 1.7633,
      "step": 139
    },
    {
      "epoch": 0.05865102639296188,
      "grad_norm": 0.6728368997573853,
      "learning_rate": 1.9767071638039382e-05,
      "loss": 1.8313,
      "step": 140
    },
    {
      "epoch": 0.059069962295768746,
      "grad_norm": 0.588307797908783,
      "learning_rate": 1.9765395894428156e-05,
      "loss": 1.7395,
      "step": 141
    },
    {
      "epoch": 0.05948889819857562,
      "grad_norm": 0.889776885509491,
      "learning_rate": 1.9763720150816927e-05,
      "loss": 1.7591,
      "step": 142
    },
    {
      "epoch": 0.059907834101382486,
      "grad_norm": 0.5996978878974915,
      "learning_rate": 1.9762044407205697e-05,
      "loss": 1.6881,
      "step": 143
    },
    {
      "epoch": 0.06032677000418936,
      "grad_norm": 0.6324535012245178,
      "learning_rate": 1.976036866359447e-05,
      "loss": 1.7142,
      "step": 144
    },
    {
      "epoch": 0.06074570590699623,
      "grad_norm": 0.6198902130126953,
      "learning_rate": 1.9758692919983246e-05,
      "loss": 1.6867,
      "step": 145
    },
    {
      "epoch": 0.0611646418098031,
      "grad_norm": 0.6651074886322021,
      "learning_rate": 1.9757017176372016e-05,
      "loss": 1.7773,
      "step": 146
    },
    {
      "epoch": 0.06158357771260997,
      "grad_norm": 0.6877864599227905,
      "learning_rate": 1.9755341432760787e-05,
      "loss": 1.792,
      "step": 147
    },
    {
      "epoch": 0.062002513615416845,
      "grad_norm": 0.5880079865455627,
      "learning_rate": 1.975366568914956e-05,
      "loss": 1.7761,
      "step": 148
    },
    {
      "epoch": 0.06242144951822371,
      "grad_norm": 0.6574519872665405,
      "learning_rate": 1.9751989945538335e-05,
      "loss": 1.7769,
      "step": 149
    },
    {
      "epoch": 0.06284038542103058,
      "grad_norm": 0.5385615825653076,
      "learning_rate": 1.9750314201927106e-05,
      "loss": 1.7226,
      "step": 150
    },
    {
      "epoch": 0.06325932132383745,
      "grad_norm": 0.6323086023330688,
      "learning_rate": 1.9748638458315877e-05,
      "loss": 1.7043,
      "step": 151
    },
    {
      "epoch": 0.06367825722664433,
      "grad_norm": 0.5224570035934448,
      "learning_rate": 1.974696271470465e-05,
      "loss": 1.7103,
      "step": 152
    },
    {
      "epoch": 0.0640971931294512,
      "grad_norm": 0.5448949933052063,
      "learning_rate": 1.9745286971093425e-05,
      "loss": 1.7691,
      "step": 153
    },
    {
      "epoch": 0.06451612903225806,
      "grad_norm": 0.47703322768211365,
      "learning_rate": 1.97436112274822e-05,
      "loss": 1.7107,
      "step": 154
    },
    {
      "epoch": 0.06493506493506493,
      "grad_norm": 0.555005669593811,
      "learning_rate": 1.974193548387097e-05,
      "loss": 1.7493,
      "step": 155
    },
    {
      "epoch": 0.06535400083787181,
      "grad_norm": 0.5083591938018799,
      "learning_rate": 1.974025974025974e-05,
      "loss": 1.621,
      "step": 156
    },
    {
      "epoch": 0.06577293674067868,
      "grad_norm": 0.5037218928337097,
      "learning_rate": 1.9738583996648515e-05,
      "loss": 1.6905,
      "step": 157
    },
    {
      "epoch": 0.06619187264348554,
      "grad_norm": 0.48570749163627625,
      "learning_rate": 1.973690825303729e-05,
      "loss": 1.73,
      "step": 158
    },
    {
      "epoch": 0.06661080854629242,
      "grad_norm": 0.5082443952560425,
      "learning_rate": 1.973523250942606e-05,
      "loss": 1.7107,
      "step": 159
    },
    {
      "epoch": 0.06702974444909929,
      "grad_norm": 0.5330659747123718,
      "learning_rate": 1.973355676581483e-05,
      "loss": 1.7756,
      "step": 160
    },
    {
      "epoch": 0.06744868035190615,
      "grad_norm": 0.6401243209838867,
      "learning_rate": 1.9731881022203605e-05,
      "loss": 1.7171,
      "step": 161
    },
    {
      "epoch": 0.06786761625471303,
      "grad_norm": 0.5555040836334229,
      "learning_rate": 1.9730205278592375e-05,
      "loss": 1.7117,
      "step": 162
    },
    {
      "epoch": 0.0682865521575199,
      "grad_norm": 0.5580366253852844,
      "learning_rate": 1.972852953498115e-05,
      "loss": 1.7173,
      "step": 163
    },
    {
      "epoch": 0.06870548806032677,
      "grad_norm": 0.6296005249023438,
      "learning_rate": 1.9726853791369924e-05,
      "loss": 1.7316,
      "step": 164
    },
    {
      "epoch": 0.06912442396313365,
      "grad_norm": 0.681302547454834,
      "learning_rate": 1.9725178047758694e-05,
      "loss": 1.7849,
      "step": 165
    },
    {
      "epoch": 0.06954335986594051,
      "grad_norm": 0.6220213770866394,
      "learning_rate": 1.9723502304147465e-05,
      "loss": 1.656,
      "step": 166
    },
    {
      "epoch": 0.06996229576874738,
      "grad_norm": 0.6640814542770386,
      "learning_rate": 1.972182656053624e-05,
      "loss": 1.8749,
      "step": 167
    },
    {
      "epoch": 0.07038123167155426,
      "grad_norm": 0.5521919131278992,
      "learning_rate": 1.9720150816925013e-05,
      "loss": 1.763,
      "step": 168
    },
    {
      "epoch": 0.07080016757436113,
      "grad_norm": 0.6188511848449707,
      "learning_rate": 1.9718475073313784e-05,
      "loss": 1.6431,
      "step": 169
    },
    {
      "epoch": 0.07121910347716799,
      "grad_norm": 0.5388275980949402,
      "learning_rate": 1.9716799329702555e-05,
      "loss": 1.7444,
      "step": 170
    },
    {
      "epoch": 0.07163803937997486,
      "grad_norm": 1.0150574445724487,
      "learning_rate": 1.971512358609133e-05,
      "loss": 1.6986,
      "step": 171
    },
    {
      "epoch": 0.07205697528278174,
      "grad_norm": 0.6714919805526733,
      "learning_rate": 1.9713447842480103e-05,
      "loss": 1.7461,
      "step": 172
    },
    {
      "epoch": 0.0724759111855886,
      "grad_norm": 0.6587640047073364,
      "learning_rate": 1.9711772098868874e-05,
      "loss": 1.6743,
      "step": 173
    },
    {
      "epoch": 0.07289484708839547,
      "grad_norm": 0.6181256175041199,
      "learning_rate": 1.9710096355257648e-05,
      "loss": 1.8552,
      "step": 174
    },
    {
      "epoch": 0.07331378299120235,
      "grad_norm": 0.5564039945602417,
      "learning_rate": 1.970842061164642e-05,
      "loss": 1.741,
      "step": 175
    },
    {
      "epoch": 0.07373271889400922,
      "grad_norm": 0.5421382188796997,
      "learning_rate": 1.9706744868035193e-05,
      "loss": 1.6919,
      "step": 176
    },
    {
      "epoch": 0.07415165479681608,
      "grad_norm": 0.6172086000442505,
      "learning_rate": 1.9705069124423967e-05,
      "loss": 1.6686,
      "step": 177
    },
    {
      "epoch": 0.07457059069962296,
      "grad_norm": 0.5004185438156128,
      "learning_rate": 1.9703393380812738e-05,
      "loss": 1.6309,
      "step": 178
    },
    {
      "epoch": 0.07498952660242983,
      "grad_norm": 0.5099078416824341,
      "learning_rate": 1.970171763720151e-05,
      "loss": 1.674,
      "step": 179
    },
    {
      "epoch": 0.0754084625052367,
      "grad_norm": 0.5554249882698059,
      "learning_rate": 1.9700041893590283e-05,
      "loss": 1.6227,
      "step": 180
    },
    {
      "epoch": 0.07582739840804358,
      "grad_norm": 0.6313985586166382,
      "learning_rate": 1.9698366149979057e-05,
      "loss": 1.7098,
      "step": 181
    },
    {
      "epoch": 0.07624633431085044,
      "grad_norm": 0.8186052441596985,
      "learning_rate": 1.9696690406367827e-05,
      "loss": 1.7933,
      "step": 182
    },
    {
      "epoch": 0.07666527021365731,
      "grad_norm": 0.5017969608306885,
      "learning_rate": 1.9695014662756598e-05,
      "loss": 1.6853,
      "step": 183
    },
    {
      "epoch": 0.07708420611646419,
      "grad_norm": 0.6917697787284851,
      "learning_rate": 1.9693338919145372e-05,
      "loss": 1.7031,
      "step": 184
    },
    {
      "epoch": 0.07750314201927105,
      "grad_norm": 0.5040557980537415,
      "learning_rate": 1.9691663175534146e-05,
      "loss": 1.6353,
      "step": 185
    },
    {
      "epoch": 0.07792207792207792,
      "grad_norm": 0.5733162760734558,
      "learning_rate": 1.9689987431922917e-05,
      "loss": 1.7278,
      "step": 186
    },
    {
      "epoch": 0.07834101382488479,
      "grad_norm": 0.7823026776313782,
      "learning_rate": 1.968831168831169e-05,
      "loss": 1.6954,
      "step": 187
    },
    {
      "epoch": 0.07875994972769167,
      "grad_norm": 0.5559296607971191,
      "learning_rate": 1.9686635944700462e-05,
      "loss": 1.7417,
      "step": 188
    },
    {
      "epoch": 0.07917888563049853,
      "grad_norm": 0.6399711966514587,
      "learning_rate": 1.9684960201089236e-05,
      "loss": 1.6529,
      "step": 189
    },
    {
      "epoch": 0.0795978215333054,
      "grad_norm": 0.6075267195701599,
      "learning_rate": 1.9683284457478007e-05,
      "loss": 1.687,
      "step": 190
    },
    {
      "epoch": 0.08001675743611228,
      "grad_norm": 0.5875303149223328,
      "learning_rate": 1.968160871386678e-05,
      "loss": 1.6569,
      "step": 191
    },
    {
      "epoch": 0.08043569333891915,
      "grad_norm": 0.6546170711517334,
      "learning_rate": 1.9679932970255552e-05,
      "loss": 1.6844,
      "step": 192
    },
    {
      "epoch": 0.08085462924172601,
      "grad_norm": 0.5778879523277283,
      "learning_rate": 1.9678257226644322e-05,
      "loss": 1.6705,
      "step": 193
    },
    {
      "epoch": 0.08127356514453289,
      "grad_norm": 0.5718396306037903,
      "learning_rate": 1.9676581483033097e-05,
      "loss": 1.6519,
      "step": 194
    },
    {
      "epoch": 0.08169250104733976,
      "grad_norm": 0.6596693992614746,
      "learning_rate": 1.967490573942187e-05,
      "loss": 1.6547,
      "step": 195
    },
    {
      "epoch": 0.08211143695014662,
      "grad_norm": 0.5057825446128845,
      "learning_rate": 1.9673229995810645e-05,
      "loss": 1.6214,
      "step": 196
    },
    {
      "epoch": 0.0825303728529535,
      "grad_norm": 0.6651629209518433,
      "learning_rate": 1.9671554252199416e-05,
      "loss": 1.6835,
      "step": 197
    },
    {
      "epoch": 0.08294930875576037,
      "grad_norm": 0.5056618452072144,
      "learning_rate": 1.9669878508588186e-05,
      "loss": 1.6368,
      "step": 198
    },
    {
      "epoch": 0.08336824465856724,
      "grad_norm": 0.4693203568458557,
      "learning_rate": 1.966820276497696e-05,
      "loss": 1.6322,
      "step": 199
    },
    {
      "epoch": 0.08378718056137412,
      "grad_norm": 0.5992833375930786,
      "learning_rate": 1.9666527021365735e-05,
      "loss": 1.7568,
      "step": 200
    },
    {
      "epoch": 0.08420611646418098,
      "grad_norm": 0.62791508436203,
      "learning_rate": 1.9664851277754505e-05,
      "loss": 1.6937,
      "step": 201
    },
    {
      "epoch": 0.08462505236698785,
      "grad_norm": 0.5130066275596619,
      "learning_rate": 1.9663175534143276e-05,
      "loss": 1.6236,
      "step": 202
    },
    {
      "epoch": 0.08504398826979472,
      "grad_norm": 0.5045161247253418,
      "learning_rate": 1.966149979053205e-05,
      "loss": 1.7386,
      "step": 203
    },
    {
      "epoch": 0.0854629241726016,
      "grad_norm": 0.6568188667297363,
      "learning_rate": 1.9659824046920824e-05,
      "loss": 1.6639,
      "step": 204
    },
    {
      "epoch": 0.08588186007540846,
      "grad_norm": 0.6545958518981934,
      "learning_rate": 1.9658148303309595e-05,
      "loss": 1.6593,
      "step": 205
    },
    {
      "epoch": 0.08630079597821533,
      "grad_norm": 0.5190823078155518,
      "learning_rate": 1.9656472559698366e-05,
      "loss": 1.6398,
      "step": 206
    },
    {
      "epoch": 0.08671973188102221,
      "grad_norm": 0.5478256344795227,
      "learning_rate": 1.965479681608714e-05,
      "loss": 1.5848,
      "step": 207
    },
    {
      "epoch": 0.08713866778382907,
      "grad_norm": 0.5818894505500793,
      "learning_rate": 1.9653121072475914e-05,
      "loss": 1.6046,
      "step": 208
    },
    {
      "epoch": 0.08755760368663594,
      "grad_norm": 0.6687189936637878,
      "learning_rate": 1.9651445328864685e-05,
      "loss": 1.6177,
      "step": 209
    },
    {
      "epoch": 0.08797653958944282,
      "grad_norm": 0.5873174071311951,
      "learning_rate": 1.964976958525346e-05,
      "loss": 1.659,
      "step": 210
    },
    {
      "epoch": 0.08839547549224969,
      "grad_norm": 0.5621105432510376,
      "learning_rate": 1.964809384164223e-05,
      "loss": 1.6739,
      "step": 211
    },
    {
      "epoch": 0.08881441139505655,
      "grad_norm": 0.7059792876243591,
      "learning_rate": 1.9646418098031004e-05,
      "loss": 1.6738,
      "step": 212
    },
    {
      "epoch": 0.08923334729786343,
      "grad_norm": 0.5294623970985413,
      "learning_rate": 1.9644742354419775e-05,
      "loss": 1.6419,
      "step": 213
    },
    {
      "epoch": 0.0896522832006703,
      "grad_norm": 0.7764983773231506,
      "learning_rate": 1.964306661080855e-05,
      "loss": 1.6793,
      "step": 214
    },
    {
      "epoch": 0.09007121910347717,
      "grad_norm": 0.628094494342804,
      "learning_rate": 1.964139086719732e-05,
      "loss": 1.676,
      "step": 215
    },
    {
      "epoch": 0.09049015500628405,
      "grad_norm": 0.5593112111091614,
      "learning_rate": 1.9639715123586094e-05,
      "loss": 1.7119,
      "step": 216
    },
    {
      "epoch": 0.09090909090909091,
      "grad_norm": 0.7385965585708618,
      "learning_rate": 1.9638039379974864e-05,
      "loss": 1.6039,
      "step": 217
    },
    {
      "epoch": 0.09132802681189778,
      "grad_norm": 0.5662972927093506,
      "learning_rate": 1.963636363636364e-05,
      "loss": 1.6255,
      "step": 218
    },
    {
      "epoch": 0.09174696271470464,
      "grad_norm": 0.7071956992149353,
      "learning_rate": 1.9634687892752413e-05,
      "loss": 1.6587,
      "step": 219
    },
    {
      "epoch": 0.09216589861751152,
      "grad_norm": 0.6986474990844727,
      "learning_rate": 1.9633012149141183e-05,
      "loss": 1.6266,
      "step": 220
    },
    {
      "epoch": 0.09258483452031839,
      "grad_norm": 0.6965208053588867,
      "learning_rate": 1.9631336405529954e-05,
      "loss": 1.6619,
      "step": 221
    },
    {
      "epoch": 0.09300377042312526,
      "grad_norm": 0.7024741768836975,
      "learning_rate": 1.9629660661918728e-05,
      "loss": 1.7327,
      "step": 222
    },
    {
      "epoch": 0.09342270632593214,
      "grad_norm": 0.5887707471847534,
      "learning_rate": 1.9627984918307502e-05,
      "loss": 1.6706,
      "step": 223
    },
    {
      "epoch": 0.093841642228739,
      "grad_norm": 0.8550237417221069,
      "learning_rate": 1.9626309174696273e-05,
      "loss": 1.5773,
      "step": 224
    },
    {
      "epoch": 0.09426057813154587,
      "grad_norm": 0.6820223331451416,
      "learning_rate": 1.9624633431085044e-05,
      "loss": 1.7496,
      "step": 225
    },
    {
      "epoch": 0.09467951403435275,
      "grad_norm": 0.7763844728469849,
      "learning_rate": 1.9622957687473818e-05,
      "loss": 1.6204,
      "step": 226
    },
    {
      "epoch": 0.09509844993715962,
      "grad_norm": 0.779120147228241,
      "learning_rate": 1.9621281943862592e-05,
      "loss": 1.617,
      "step": 227
    },
    {
      "epoch": 0.09551738583996648,
      "grad_norm": 0.7589849233627319,
      "learning_rate": 1.9619606200251363e-05,
      "loss": 1.5478,
      "step": 228
    },
    {
      "epoch": 0.09593632174277336,
      "grad_norm": 0.5152125954627991,
      "learning_rate": 1.9617930456640137e-05,
      "loss": 1.6749,
      "step": 229
    },
    {
      "epoch": 0.09635525764558023,
      "grad_norm": 0.7013604640960693,
      "learning_rate": 1.9616254713028908e-05,
      "loss": 1.5864,
      "step": 230
    },
    {
      "epoch": 0.0967741935483871,
      "grad_norm": 0.7294275760650635,
      "learning_rate": 1.9614578969417682e-05,
      "loss": 1.5845,
      "step": 231
    },
    {
      "epoch": 0.09719312945119397,
      "grad_norm": 0.5346665382385254,
      "learning_rate": 1.9612903225806452e-05,
      "loss": 1.643,
      "step": 232
    },
    {
      "epoch": 0.09761206535400084,
      "grad_norm": 0.953779935836792,
      "learning_rate": 1.9611227482195227e-05,
      "loss": 1.6152,
      "step": 233
    },
    {
      "epoch": 0.0980310012568077,
      "grad_norm": 0.6668707132339478,
      "learning_rate": 1.9609551738583997e-05,
      "loss": 1.6066,
      "step": 234
    },
    {
      "epoch": 0.09844993715961457,
      "grad_norm": 0.6693033576011658,
      "learning_rate": 1.960787599497277e-05,
      "loss": 1.6124,
      "step": 235
    },
    {
      "epoch": 0.09886887306242145,
      "grad_norm": 0.8081066608428955,
      "learning_rate": 1.9606200251361542e-05,
      "loss": 1.6351,
      "step": 236
    },
    {
      "epoch": 0.09928780896522832,
      "grad_norm": 0.7561270594596863,
      "learning_rate": 1.9604524507750316e-05,
      "loss": 1.6476,
      "step": 237
    },
    {
      "epoch": 0.09970674486803519,
      "grad_norm": 0.8342212438583374,
      "learning_rate": 1.9602848764139087e-05,
      "loss": 1.5573,
      "step": 238
    },
    {
      "epoch": 0.10012568077084207,
      "grad_norm": 0.8095865845680237,
      "learning_rate": 1.960117302052786e-05,
      "loss": 1.5594,
      "step": 239
    },
    {
      "epoch": 0.10054461667364893,
      "grad_norm": 0.8448402881622314,
      "learning_rate": 1.9599497276916632e-05,
      "loss": 1.6646,
      "step": 240
    },
    {
      "epoch": 0.1009635525764558,
      "grad_norm": 0.936273455619812,
      "learning_rate": 1.9597821533305406e-05,
      "loss": 1.5822,
      "step": 241
    },
    {
      "epoch": 0.10138248847926268,
      "grad_norm": 0.5605466365814209,
      "learning_rate": 1.959614578969418e-05,
      "loss": 1.616,
      "step": 242
    },
    {
      "epoch": 0.10180142438206954,
      "grad_norm": 1.0700498819351196,
      "learning_rate": 1.959447004608295e-05,
      "loss": 1.618,
      "step": 243
    },
    {
      "epoch": 0.10222036028487641,
      "grad_norm": 0.6166669726371765,
      "learning_rate": 1.959279430247172e-05,
      "loss": 1.5989,
      "step": 244
    },
    {
      "epoch": 0.10263929618768329,
      "grad_norm": 0.7001603841781616,
      "learning_rate": 1.9591118558860496e-05,
      "loss": 1.464,
      "step": 245
    },
    {
      "epoch": 0.10305823209049016,
      "grad_norm": 0.5694488883018494,
      "learning_rate": 1.958944281524927e-05,
      "loss": 1.5641,
      "step": 246
    },
    {
      "epoch": 0.10347716799329702,
      "grad_norm": 0.5658904314041138,
      "learning_rate": 1.958776707163804e-05,
      "loss": 1.5911,
      "step": 247
    },
    {
      "epoch": 0.1038961038961039,
      "grad_norm": 0.6600093245506287,
      "learning_rate": 1.958609132802681e-05,
      "loss": 1.6812,
      "step": 248
    },
    {
      "epoch": 0.10431503979891077,
      "grad_norm": 0.7548564672470093,
      "learning_rate": 1.9584415584415586e-05,
      "loss": 1.5563,
      "step": 249
    },
    {
      "epoch": 0.10473397570171764,
      "grad_norm": 0.6965343952178955,
      "learning_rate": 1.958273984080436e-05,
      "loss": 1.6126,
      "step": 250
    },
    {
      "epoch": 0.1051529116045245,
      "grad_norm": 0.57705157995224,
      "learning_rate": 1.958106409719313e-05,
      "loss": 1.6778,
      "step": 251
    },
    {
      "epoch": 0.10557184750733138,
      "grad_norm": 1.2478373050689697,
      "learning_rate": 1.9579388353581905e-05,
      "loss": 1.6654,
      "step": 252
    },
    {
      "epoch": 0.10599078341013825,
      "grad_norm": 0.586651623249054,
      "learning_rate": 1.9577712609970675e-05,
      "loss": 1.5895,
      "step": 253
    },
    {
      "epoch": 0.10640971931294511,
      "grad_norm": 0.6501399874687195,
      "learning_rate": 1.957603686635945e-05,
      "loss": 1.5991,
      "step": 254
    },
    {
      "epoch": 0.106828655215752,
      "grad_norm": 0.7814245223999023,
      "learning_rate": 1.9574361122748224e-05,
      "loss": 1.5765,
      "step": 255
    },
    {
      "epoch": 0.10724759111855886,
      "grad_norm": 0.691818356513977,
      "learning_rate": 1.9572685379136994e-05,
      "loss": 1.6663,
      "step": 256
    },
    {
      "epoch": 0.10766652702136573,
      "grad_norm": 0.9530359506607056,
      "learning_rate": 1.9571009635525765e-05,
      "loss": 1.6614,
      "step": 257
    },
    {
      "epoch": 0.1080854629241726,
      "grad_norm": 1.280455231666565,
      "learning_rate": 1.956933389191454e-05,
      "loss": 1.5998,
      "step": 258
    },
    {
      "epoch": 0.10850439882697947,
      "grad_norm": 0.6066200137138367,
      "learning_rate": 1.956765814830331e-05,
      "loss": 1.6352,
      "step": 259
    },
    {
      "epoch": 0.10892333472978634,
      "grad_norm": 0.6973633766174316,
      "learning_rate": 1.9565982404692084e-05,
      "loss": 1.619,
      "step": 260
    },
    {
      "epoch": 0.10934227063259322,
      "grad_norm": 0.857652485370636,
      "learning_rate": 1.9564306661080855e-05,
      "loss": 1.6507,
      "step": 261
    },
    {
      "epoch": 0.10976120653540009,
      "grad_norm": 0.5751752853393555,
      "learning_rate": 1.956263091746963e-05,
      "loss": 1.6866,
      "step": 262
    },
    {
      "epoch": 0.11018014243820695,
      "grad_norm": 0.899047315120697,
      "learning_rate": 1.95609551738584e-05,
      "loss": 1.6211,
      "step": 263
    },
    {
      "epoch": 0.11059907834101383,
      "grad_norm": 0.6887657046318054,
      "learning_rate": 1.9559279430247174e-05,
      "loss": 1.5289,
      "step": 264
    },
    {
      "epoch": 0.1110180142438207,
      "grad_norm": 0.6921897530555725,
      "learning_rate": 1.9557603686635948e-05,
      "loss": 1.6667,
      "step": 265
    },
    {
      "epoch": 0.11143695014662756,
      "grad_norm": 0.5706766843795776,
      "learning_rate": 1.955592794302472e-05,
      "loss": 1.6057,
      "step": 266
    },
    {
      "epoch": 0.11185588604943443,
      "grad_norm": 0.7291983962059021,
      "learning_rate": 1.955425219941349e-05,
      "loss": 1.5033,
      "step": 267
    },
    {
      "epoch": 0.11227482195224131,
      "grad_norm": 0.5996133089065552,
      "learning_rate": 1.9552576455802263e-05,
      "loss": 1.5232,
      "step": 268
    },
    {
      "epoch": 0.11269375785504818,
      "grad_norm": 0.6987999677658081,
      "learning_rate": 1.9550900712191038e-05,
      "loss": 1.5832,
      "step": 269
    },
    {
      "epoch": 0.11311269375785504,
      "grad_norm": 0.7466180920600891,
      "learning_rate": 1.954922496857981e-05,
      "loss": 1.6155,
      "step": 270
    },
    {
      "epoch": 0.11353162966066192,
      "grad_norm": 0.5365012884140015,
      "learning_rate": 1.954754922496858e-05,
      "loss": 1.4844,
      "step": 271
    },
    {
      "epoch": 0.11395056556346879,
      "grad_norm": 0.6302610635757446,
      "learning_rate": 1.9545873481357353e-05,
      "loss": 1.6401,
      "step": 272
    },
    {
      "epoch": 0.11436950146627566,
      "grad_norm": 0.7011299729347229,
      "learning_rate": 1.9544197737746127e-05,
      "loss": 1.5856,
      "step": 273
    },
    {
      "epoch": 0.11478843736908254,
      "grad_norm": 0.5213186740875244,
      "learning_rate": 1.95425219941349e-05,
      "loss": 1.6877,
      "step": 274
    },
    {
      "epoch": 0.1152073732718894,
      "grad_norm": 0.5973451137542725,
      "learning_rate": 1.9540846250523672e-05,
      "loss": 1.6255,
      "step": 275
    },
    {
      "epoch": 0.11562630917469627,
      "grad_norm": 0.8726099729537964,
      "learning_rate": 1.9539170506912443e-05,
      "loss": 1.6345,
      "step": 276
    },
    {
      "epoch": 0.11604524507750315,
      "grad_norm": 0.6559906005859375,
      "learning_rate": 1.9537494763301217e-05,
      "loss": 1.4426,
      "step": 277
    },
    {
      "epoch": 0.11646418098031001,
      "grad_norm": 1.2165895700454712,
      "learning_rate": 1.953581901968999e-05,
      "loss": 1.597,
      "step": 278
    },
    {
      "epoch": 0.11688311688311688,
      "grad_norm": 0.5331985354423523,
      "learning_rate": 1.9534143276078762e-05,
      "loss": 1.5959,
      "step": 279
    },
    {
      "epoch": 0.11730205278592376,
      "grad_norm": 0.7331112027168274,
      "learning_rate": 1.9532467532467533e-05,
      "loss": 1.6,
      "step": 280
    },
    {
      "epoch": 0.11772098868873063,
      "grad_norm": 0.6983991265296936,
      "learning_rate": 1.9530791788856307e-05,
      "loss": 1.6188,
      "step": 281
    },
    {
      "epoch": 0.11813992459153749,
      "grad_norm": 0.6612614393234253,
      "learning_rate": 1.952911604524508e-05,
      "loss": 1.579,
      "step": 282
    },
    {
      "epoch": 0.11855886049434436,
      "grad_norm": 0.6999834179878235,
      "learning_rate": 1.952744030163385e-05,
      "loss": 1.715,
      "step": 283
    },
    {
      "epoch": 0.11897779639715124,
      "grad_norm": 0.6970024108886719,
      "learning_rate": 1.9525764558022622e-05,
      "loss": 1.6406,
      "step": 284
    },
    {
      "epoch": 0.1193967322999581,
      "grad_norm": 0.5302523374557495,
      "learning_rate": 1.9524088814411397e-05,
      "loss": 1.6388,
      "step": 285
    },
    {
      "epoch": 0.11981566820276497,
      "grad_norm": 0.7106460332870483,
      "learning_rate": 1.952241307080017e-05,
      "loss": 1.5679,
      "step": 286
    },
    {
      "epoch": 0.12023460410557185,
      "grad_norm": 0.6428540945053101,
      "learning_rate": 1.952073732718894e-05,
      "loss": 1.5151,
      "step": 287
    },
    {
      "epoch": 0.12065354000837872,
      "grad_norm": 0.6660862565040588,
      "learning_rate": 1.9519061583577716e-05,
      "loss": 1.6999,
      "step": 288
    },
    {
      "epoch": 0.12107247591118558,
      "grad_norm": 0.6251623034477234,
      "learning_rate": 1.9517385839966486e-05,
      "loss": 1.5289,
      "step": 289
    },
    {
      "epoch": 0.12149141181399246,
      "grad_norm": 0.5240997672080994,
      "learning_rate": 1.9515710096355257e-05,
      "loss": 1.5991,
      "step": 290
    },
    {
      "epoch": 0.12191034771679933,
      "grad_norm": 0.47173094749450684,
      "learning_rate": 1.951403435274403e-05,
      "loss": 1.5734,
      "step": 291
    },
    {
      "epoch": 0.1223292836196062,
      "grad_norm": 0.775086522102356,
      "learning_rate": 1.9512358609132805e-05,
      "loss": 1.4597,
      "step": 292
    },
    {
      "epoch": 0.12274821952241308,
      "grad_norm": 0.622778594493866,
      "learning_rate": 1.9510682865521576e-05,
      "loss": 1.5872,
      "step": 293
    },
    {
      "epoch": 0.12316715542521994,
      "grad_norm": 0.5721079111099243,
      "learning_rate": 1.9509007121910347e-05,
      "loss": 1.5226,
      "step": 294
    },
    {
      "epoch": 0.12358609132802681,
      "grad_norm": 0.8591808676719666,
      "learning_rate": 1.950733137829912e-05,
      "loss": 1.6192,
      "step": 295
    },
    {
      "epoch": 0.12400502723083369,
      "grad_norm": 0.5390528440475464,
      "learning_rate": 1.9505655634687895e-05,
      "loss": 1.5228,
      "step": 296
    },
    {
      "epoch": 0.12442396313364056,
      "grad_norm": 0.6414217948913574,
      "learning_rate": 1.950397989107667e-05,
      "loss": 1.6464,
      "step": 297
    },
    {
      "epoch": 0.12484289903644742,
      "grad_norm": 0.9504109621047974,
      "learning_rate": 1.950230414746544e-05,
      "loss": 1.4574,
      "step": 298
    },
    {
      "epoch": 0.1252618349392543,
      "grad_norm": 0.7934843301773071,
      "learning_rate": 1.950062840385421e-05,
      "loss": 1.5825,
      "step": 299
    },
    {
      "epoch": 0.12568077084206117,
      "grad_norm": 1.2911075353622437,
      "learning_rate": 1.9498952660242985e-05,
      "loss": 1.5519,
      "step": 300
    },
    {
      "epoch": 0.12609970674486803,
      "grad_norm": 0.7778868079185486,
      "learning_rate": 1.949727691663176e-05,
      "loss": 1.569,
      "step": 301
    },
    {
      "epoch": 0.1265186426476749,
      "grad_norm": 0.669278621673584,
      "learning_rate": 1.949560117302053e-05,
      "loss": 1.6144,
      "step": 302
    },
    {
      "epoch": 0.12693757855048177,
      "grad_norm": 1.0039944648742676,
      "learning_rate": 1.94939254294093e-05,
      "loss": 1.5249,
      "step": 303
    },
    {
      "epoch": 0.12735651445328866,
      "grad_norm": 0.7724584937095642,
      "learning_rate": 1.9492249685798074e-05,
      "loss": 1.6277,
      "step": 304
    },
    {
      "epoch": 0.12777545035609553,
      "grad_norm": 0.759467601776123,
      "learning_rate": 1.949057394218685e-05,
      "loss": 1.5756,
      "step": 305
    },
    {
      "epoch": 0.1281943862589024,
      "grad_norm": 1.3089075088500977,
      "learning_rate": 1.948889819857562e-05,
      "loss": 1.5371,
      "step": 306
    },
    {
      "epoch": 0.12861332216170926,
      "grad_norm": 0.5424050688743591,
      "learning_rate": 1.9487222454964393e-05,
      "loss": 1.5824,
      "step": 307
    },
    {
      "epoch": 0.12903225806451613,
      "grad_norm": 0.7417402267456055,
      "learning_rate": 1.9485546711353164e-05,
      "loss": 1.5458,
      "step": 308
    },
    {
      "epoch": 0.129451193967323,
      "grad_norm": 0.6556446552276611,
      "learning_rate": 1.948387096774194e-05,
      "loss": 1.6494,
      "step": 309
    },
    {
      "epoch": 0.12987012987012986,
      "grad_norm": 0.9575488567352295,
      "learning_rate": 1.948219522413071e-05,
      "loss": 1.5509,
      "step": 310
    },
    {
      "epoch": 0.13028906577293675,
      "grad_norm": 0.8725043535232544,
      "learning_rate": 1.9480519480519483e-05,
      "loss": 1.5099,
      "step": 311
    },
    {
      "epoch": 0.13070800167574362,
      "grad_norm": 1.1596598625183105,
      "learning_rate": 1.9478843736908254e-05,
      "loss": 1.6001,
      "step": 312
    },
    {
      "epoch": 0.13112693757855048,
      "grad_norm": 0.986149251461029,
      "learning_rate": 1.9477167993297028e-05,
      "loss": 1.4582,
      "step": 313
    },
    {
      "epoch": 0.13154587348135735,
      "grad_norm": 0.6939131021499634,
      "learning_rate": 1.94754922496858e-05,
      "loss": 1.6263,
      "step": 314
    },
    {
      "epoch": 0.13196480938416422,
      "grad_norm": 0.664408802986145,
      "learning_rate": 1.9473816506074573e-05,
      "loss": 1.538,
      "step": 315
    },
    {
      "epoch": 0.13238374528697108,
      "grad_norm": 0.6826114058494568,
      "learning_rate": 1.9472140762463344e-05,
      "loss": 1.5806,
      "step": 316
    },
    {
      "epoch": 0.13280268118977798,
      "grad_norm": 0.6879693269729614,
      "learning_rate": 1.9470465018852118e-05,
      "loss": 1.5156,
      "step": 317
    },
    {
      "epoch": 0.13322161709258484,
      "grad_norm": 1.2276116609573364,
      "learning_rate": 1.946878927524089e-05,
      "loss": 1.4466,
      "step": 318
    },
    {
      "epoch": 0.1336405529953917,
      "grad_norm": 0.7346695065498352,
      "learning_rate": 1.9467113531629663e-05,
      "loss": 1.5769,
      "step": 319
    },
    {
      "epoch": 0.13405948889819858,
      "grad_norm": 0.795690655708313,
      "learning_rate": 1.9465437788018437e-05,
      "loss": 1.5348,
      "step": 320
    },
    {
      "epoch": 0.13447842480100544,
      "grad_norm": 0.8207523822784424,
      "learning_rate": 1.9463762044407208e-05,
      "loss": 1.5196,
      "step": 321
    },
    {
      "epoch": 0.1348973607038123,
      "grad_norm": 0.6205607056617737,
      "learning_rate": 1.9462086300795978e-05,
      "loss": 1.5508,
      "step": 322
    },
    {
      "epoch": 0.13531629660661917,
      "grad_norm": 0.7060804963111877,
      "learning_rate": 1.9460410557184752e-05,
      "loss": 1.547,
      "step": 323
    },
    {
      "epoch": 0.13573523250942607,
      "grad_norm": 0.6053579449653625,
      "learning_rate": 1.9458734813573527e-05,
      "loss": 1.5521,
      "step": 324
    },
    {
      "epoch": 0.13615416841223293,
      "grad_norm": 0.6387944221496582,
      "learning_rate": 1.9457059069962297e-05,
      "loss": 1.5785,
      "step": 325
    },
    {
      "epoch": 0.1365731043150398,
      "grad_norm": 0.7160474061965942,
      "learning_rate": 1.9455383326351068e-05,
      "loss": 1.5527,
      "step": 326
    },
    {
      "epoch": 0.13699204021784667,
      "grad_norm": 0.5747194290161133,
      "learning_rate": 1.9453707582739842e-05,
      "loss": 1.4228,
      "step": 327
    },
    {
      "epoch": 0.13741097612065353,
      "grad_norm": 0.7289405465126038,
      "learning_rate": 1.9452031839128616e-05,
      "loss": 1.5758,
      "step": 328
    },
    {
      "epoch": 0.1378299120234604,
      "grad_norm": 0.5292596817016602,
      "learning_rate": 1.9450356095517387e-05,
      "loss": 1.5287,
      "step": 329
    },
    {
      "epoch": 0.1382488479262673,
      "grad_norm": 0.7831090092658997,
      "learning_rate": 1.944868035190616e-05,
      "loss": 1.5584,
      "step": 330
    },
    {
      "epoch": 0.13866778382907416,
      "grad_norm": 0.6046620011329651,
      "learning_rate": 1.9447004608294932e-05,
      "loss": 1.5534,
      "step": 331
    },
    {
      "epoch": 0.13908671973188103,
      "grad_norm": 0.7165292501449585,
      "learning_rate": 1.9445328864683706e-05,
      "loss": 1.621,
      "step": 332
    },
    {
      "epoch": 0.1395056556346879,
      "grad_norm": 0.7406589388847351,
      "learning_rate": 1.9443653121072477e-05,
      "loss": 1.5954,
      "step": 333
    },
    {
      "epoch": 0.13992459153749476,
      "grad_norm": 0.5955418348312378,
      "learning_rate": 1.944197737746125e-05,
      "loss": 1.5457,
      "step": 334
    },
    {
      "epoch": 0.14034352744030162,
      "grad_norm": 0.5523016452789307,
      "learning_rate": 1.944030163385002e-05,
      "loss": 1.5131,
      "step": 335
    },
    {
      "epoch": 0.14076246334310852,
      "grad_norm": 0.7677832841873169,
      "learning_rate": 1.9438625890238796e-05,
      "loss": 1.5521,
      "step": 336
    },
    {
      "epoch": 0.14118139924591538,
      "grad_norm": 0.6301062107086182,
      "learning_rate": 1.9436950146627566e-05,
      "loss": 1.5836,
      "step": 337
    },
    {
      "epoch": 0.14160033514872225,
      "grad_norm": 0.6077446341514587,
      "learning_rate": 1.943527440301634e-05,
      "loss": 1.5108,
      "step": 338
    },
    {
      "epoch": 0.14201927105152912,
      "grad_norm": 0.6678585410118103,
      "learning_rate": 1.943359865940511e-05,
      "loss": 1.5488,
      "step": 339
    },
    {
      "epoch": 0.14243820695433598,
      "grad_norm": 0.7583240270614624,
      "learning_rate": 1.9431922915793885e-05,
      "loss": 1.551,
      "step": 340
    },
    {
      "epoch": 0.14285714285714285,
      "grad_norm": 0.5899871587753296,
      "learning_rate": 1.9430247172182656e-05,
      "loss": 1.5522,
      "step": 341
    },
    {
      "epoch": 0.14327607875994972,
      "grad_norm": 0.6008652448654175,
      "learning_rate": 1.942857142857143e-05,
      "loss": 1.5742,
      "step": 342
    },
    {
      "epoch": 0.1436950146627566,
      "grad_norm": 0.5442250967025757,
      "learning_rate": 1.9426895684960204e-05,
      "loss": 1.5352,
      "step": 343
    },
    {
      "epoch": 0.14411395056556348,
      "grad_norm": 0.5791664123535156,
      "learning_rate": 1.9425219941348975e-05,
      "loss": 1.5924,
      "step": 344
    },
    {
      "epoch": 0.14453288646837034,
      "grad_norm": 0.5690919160842896,
      "learning_rate": 1.9423544197737746e-05,
      "loss": 1.5219,
      "step": 345
    },
    {
      "epoch": 0.1449518223711772,
      "grad_norm": 0.6443383693695068,
      "learning_rate": 1.942186845412652e-05,
      "loss": 1.5994,
      "step": 346
    },
    {
      "epoch": 0.14537075827398407,
      "grad_norm": 0.5575938820838928,
      "learning_rate": 1.9420192710515294e-05,
      "loss": 1.5579,
      "step": 347
    },
    {
      "epoch": 0.14578969417679094,
      "grad_norm": 0.612278401851654,
      "learning_rate": 1.9418516966904065e-05,
      "loss": 1.6067,
      "step": 348
    },
    {
      "epoch": 0.14620863007959783,
      "grad_norm": 0.6119928956031799,
      "learning_rate": 1.9416841223292836e-05,
      "loss": 1.4636,
      "step": 349
    },
    {
      "epoch": 0.1466275659824047,
      "grad_norm": 0.568268895149231,
      "learning_rate": 1.941516547968161e-05,
      "loss": 1.4923,
      "step": 350
    },
    {
      "epoch": 0.14704650188521157,
      "grad_norm": 0.6106241345405579,
      "learning_rate": 1.9413489736070384e-05,
      "loss": 1.5735,
      "step": 351
    },
    {
      "epoch": 0.14746543778801843,
      "grad_norm": 0.5713450312614441,
      "learning_rate": 1.9411813992459158e-05,
      "loss": 1.4594,
      "step": 352
    },
    {
      "epoch": 0.1478843736908253,
      "grad_norm": 0.5222604274749756,
      "learning_rate": 1.941013824884793e-05,
      "loss": 1.4957,
      "step": 353
    },
    {
      "epoch": 0.14830330959363217,
      "grad_norm": 0.6617407202720642,
      "learning_rate": 1.94084625052367e-05,
      "loss": 1.3623,
      "step": 354
    },
    {
      "epoch": 0.14872224549643903,
      "grad_norm": 0.6270737648010254,
      "learning_rate": 1.9406786761625474e-05,
      "loss": 1.494,
      "step": 355
    },
    {
      "epoch": 0.14914118139924593,
      "grad_norm": 0.6663143634796143,
      "learning_rate": 1.9405111018014244e-05,
      "loss": 1.5093,
      "step": 356
    },
    {
      "epoch": 0.1495601173020528,
      "grad_norm": 0.5815137624740601,
      "learning_rate": 1.940343527440302e-05,
      "loss": 1.5392,
      "step": 357
    },
    {
      "epoch": 0.14997905320485966,
      "grad_norm": 0.5811892747879028,
      "learning_rate": 1.940175953079179e-05,
      "loss": 1.5553,
      "step": 358
    },
    {
      "epoch": 0.15039798910766652,
      "grad_norm": 0.5487301349639893,
      "learning_rate": 1.9400083787180563e-05,
      "loss": 1.6208,
      "step": 359
    },
    {
      "epoch": 0.1508169250104734,
      "grad_norm": 0.5299431681632996,
      "learning_rate": 1.9398408043569334e-05,
      "loss": 1.6097,
      "step": 360
    },
    {
      "epoch": 0.15123586091328026,
      "grad_norm": 0.6918801665306091,
      "learning_rate": 1.9396732299958108e-05,
      "loss": 1.5027,
      "step": 361
    },
    {
      "epoch": 0.15165479681608715,
      "grad_norm": 0.5945777893066406,
      "learning_rate": 1.939505655634688e-05,
      "loss": 1.6155,
      "step": 362
    },
    {
      "epoch": 0.15207373271889402,
      "grad_norm": 0.700545072555542,
      "learning_rate": 1.9393380812735653e-05,
      "loss": 1.6345,
      "step": 363
    },
    {
      "epoch": 0.15249266862170088,
      "grad_norm": 0.5498125553131104,
      "learning_rate": 1.9391705069124424e-05,
      "loss": 1.5087,
      "step": 364
    },
    {
      "epoch": 0.15291160452450775,
      "grad_norm": 0.5619140267372131,
      "learning_rate": 1.9390029325513198e-05,
      "loss": 1.4977,
      "step": 365
    },
    {
      "epoch": 0.15333054042731462,
      "grad_norm": 0.5968044400215149,
      "learning_rate": 1.9388353581901972e-05,
      "loss": 1.5017,
      "step": 366
    },
    {
      "epoch": 0.15374947633012148,
      "grad_norm": 0.6269423365592957,
      "learning_rate": 1.9386677838290743e-05,
      "loss": 1.5811,
      "step": 367
    },
    {
      "epoch": 0.15416841223292838,
      "grad_norm": 0.5672966241836548,
      "learning_rate": 1.9385002094679514e-05,
      "loss": 1.4781,
      "step": 368
    },
    {
      "epoch": 0.15458734813573524,
      "grad_norm": 0.8752624988555908,
      "learning_rate": 1.9383326351068288e-05,
      "loss": 1.4952,
      "step": 369
    },
    {
      "epoch": 0.1550062840385421,
      "grad_norm": 0.5117892622947693,
      "learning_rate": 1.9381650607457062e-05,
      "loss": 1.5436,
      "step": 370
    },
    {
      "epoch": 0.15542521994134897,
      "grad_norm": 0.7978183031082153,
      "learning_rate": 1.9379974863845833e-05,
      "loss": 1.4284,
      "step": 371
    },
    {
      "epoch": 0.15584415584415584,
      "grad_norm": 0.5909569263458252,
      "learning_rate": 1.9378299120234603e-05,
      "loss": 1.5057,
      "step": 372
    },
    {
      "epoch": 0.1562630917469627,
      "grad_norm": 0.6655667424201965,
      "learning_rate": 1.9376623376623377e-05,
      "loss": 1.4267,
      "step": 373
    },
    {
      "epoch": 0.15668202764976957,
      "grad_norm": 0.6063106656074524,
      "learning_rate": 1.937494763301215e-05,
      "loss": 1.5867,
      "step": 374
    },
    {
      "epoch": 0.15710096355257647,
      "grad_norm": 0.9726372361183167,
      "learning_rate": 1.9373271889400926e-05,
      "loss": 1.4887,
      "step": 375
    },
    {
      "epoch": 0.15751989945538333,
      "grad_norm": 0.711313784122467,
      "learning_rate": 1.9371596145789696e-05,
      "loss": 1.5464,
      "step": 376
    },
    {
      "epoch": 0.1579388353581902,
      "grad_norm": 0.6071950197219849,
      "learning_rate": 1.9369920402178467e-05,
      "loss": 1.4917,
      "step": 377
    },
    {
      "epoch": 0.15835777126099707,
      "grad_norm": 0.7539801597595215,
      "learning_rate": 1.936824465856724e-05,
      "loss": 1.4267,
      "step": 378
    },
    {
      "epoch": 0.15877670716380393,
      "grad_norm": 0.5340871810913086,
      "learning_rate": 1.9366568914956015e-05,
      "loss": 1.4227,
      "step": 379
    },
    {
      "epoch": 0.1591956430666108,
      "grad_norm": 0.7538002133369446,
      "learning_rate": 1.9364893171344786e-05,
      "loss": 1.5565,
      "step": 380
    },
    {
      "epoch": 0.1596145789694177,
      "grad_norm": 0.6404510736465454,
      "learning_rate": 1.9363217427733557e-05,
      "loss": 1.5314,
      "step": 381
    },
    {
      "epoch": 0.16003351487222456,
      "grad_norm": 0.6506287455558777,
      "learning_rate": 1.936154168412233e-05,
      "loss": 1.5931,
      "step": 382
    },
    {
      "epoch": 0.16045245077503142,
      "grad_norm": 0.6741127967834473,
      "learning_rate": 1.9359865940511105e-05,
      "loss": 1.5172,
      "step": 383
    },
    {
      "epoch": 0.1608713866778383,
      "grad_norm": 0.6476618647575378,
      "learning_rate": 1.9358190196899876e-05,
      "loss": 1.5038,
      "step": 384
    },
    {
      "epoch": 0.16129032258064516,
      "grad_norm": 0.7657164931297302,
      "learning_rate": 1.935651445328865e-05,
      "loss": 1.461,
      "step": 385
    },
    {
      "epoch": 0.16170925848345202,
      "grad_norm": 0.49177902936935425,
      "learning_rate": 1.935483870967742e-05,
      "loss": 1.4751,
      "step": 386
    },
    {
      "epoch": 0.1621281943862589,
      "grad_norm": 1.3233708143234253,
      "learning_rate": 1.935316296606619e-05,
      "loss": 1.4353,
      "step": 387
    },
    {
      "epoch": 0.16254713028906578,
      "grad_norm": 0.9667218923568726,
      "learning_rate": 1.9351487222454966e-05,
      "loss": 1.4472,
      "step": 388
    },
    {
      "epoch": 0.16296606619187265,
      "grad_norm": 0.6916482448577881,
      "learning_rate": 1.934981147884374e-05,
      "loss": 1.5265,
      "step": 389
    },
    {
      "epoch": 0.16338500209467952,
      "grad_norm": 0.870297372341156,
      "learning_rate": 1.934813573523251e-05,
      "loss": 1.5272,
      "step": 390
    },
    {
      "epoch": 0.16380393799748638,
      "grad_norm": 0.48826926946640015,
      "learning_rate": 1.934645999162128e-05,
      "loss": 1.4762,
      "step": 391
    },
    {
      "epoch": 0.16422287390029325,
      "grad_norm": 0.8552863597869873,
      "learning_rate": 1.9344784248010055e-05,
      "loss": 1.4508,
      "step": 392
    },
    {
      "epoch": 0.16464180980310011,
      "grad_norm": 0.626533567905426,
      "learning_rate": 1.934310850439883e-05,
      "loss": 1.4639,
      "step": 393
    },
    {
      "epoch": 0.165060745705907,
      "grad_norm": 0.6926416158676147,
      "learning_rate": 1.93414327607876e-05,
      "loss": 1.5507,
      "step": 394
    },
    {
      "epoch": 0.16547968160871387,
      "grad_norm": 0.613665759563446,
      "learning_rate": 1.933975701717637e-05,
      "loss": 1.5496,
      "step": 395
    },
    {
      "epoch": 0.16589861751152074,
      "grad_norm": 0.7060792446136475,
      "learning_rate": 1.9338081273565145e-05,
      "loss": 1.4823,
      "step": 396
    },
    {
      "epoch": 0.1663175534143276,
      "grad_norm": 0.6209156513214111,
      "learning_rate": 1.933640552995392e-05,
      "loss": 1.5661,
      "step": 397
    },
    {
      "epoch": 0.16673648931713447,
      "grad_norm": 0.7241356372833252,
      "learning_rate": 1.9334729786342693e-05,
      "loss": 1.5442,
      "step": 398
    },
    {
      "epoch": 0.16715542521994134,
      "grad_norm": 0.5998069643974304,
      "learning_rate": 1.9333054042731464e-05,
      "loss": 1.5247,
      "step": 399
    },
    {
      "epoch": 0.16757436112274823,
      "grad_norm": 0.5730302929878235,
      "learning_rate": 1.9331378299120235e-05,
      "loss": 1.5973,
      "step": 400
    },
    {
      "epoch": 0.1679932970255551,
      "grad_norm": 0.649454653263092,
      "learning_rate": 1.932970255550901e-05,
      "loss": 1.5839,
      "step": 401
    },
    {
      "epoch": 0.16841223292836197,
      "grad_norm": 0.5908178687095642,
      "learning_rate": 1.9328026811897783e-05,
      "loss": 1.4055,
      "step": 402
    },
    {
      "epoch": 0.16883116883116883,
      "grad_norm": 0.4675125181674957,
      "learning_rate": 1.9326351068286554e-05,
      "loss": 1.5132,
      "step": 403
    },
    {
      "epoch": 0.1692501047339757,
      "grad_norm": 0.738477885723114,
      "learning_rate": 1.9324675324675325e-05,
      "loss": 1.3856,
      "step": 404
    },
    {
      "epoch": 0.16966904063678256,
      "grad_norm": 0.5623185634613037,
      "learning_rate": 1.93229995810641e-05,
      "loss": 1.5222,
      "step": 405
    },
    {
      "epoch": 0.17008797653958943,
      "grad_norm": 0.6933448314666748,
      "learning_rate": 1.9321323837452873e-05,
      "loss": 1.6141,
      "step": 406
    },
    {
      "epoch": 0.17050691244239632,
      "grad_norm": 0.6374297142028809,
      "learning_rate": 1.9319648093841644e-05,
      "loss": 1.5482,
      "step": 407
    },
    {
      "epoch": 0.1709258483452032,
      "grad_norm": 0.692150354385376,
      "learning_rate": 1.9317972350230418e-05,
      "loss": 1.4039,
      "step": 408
    },
    {
      "epoch": 0.17134478424801006,
      "grad_norm": 0.5226042866706848,
      "learning_rate": 1.931629660661919e-05,
      "loss": 1.4807,
      "step": 409
    },
    {
      "epoch": 0.17176372015081692,
      "grad_norm": 0.602898895740509,
      "learning_rate": 1.9314620863007963e-05,
      "loss": 1.4713,
      "step": 410
    },
    {
      "epoch": 0.1721826560536238,
      "grad_norm": 0.6459433436393738,
      "learning_rate": 1.9312945119396733e-05,
      "loss": 1.4061,
      "step": 411
    },
    {
      "epoch": 0.17260159195643066,
      "grad_norm": 0.6216537356376648,
      "learning_rate": 1.9311269375785507e-05,
      "loss": 1.4756,
      "step": 412
    },
    {
      "epoch": 0.17302052785923755,
      "grad_norm": 0.5865094065666199,
      "learning_rate": 1.9309593632174278e-05,
      "loss": 1.5048,
      "step": 413
    },
    {
      "epoch": 0.17343946376204442,
      "grad_norm": 0.689116895198822,
      "learning_rate": 1.9307917888563052e-05,
      "loss": 1.5203,
      "step": 414
    },
    {
      "epoch": 0.17385839966485128,
      "grad_norm": 0.5802826285362244,
      "learning_rate": 1.9306242144951823e-05,
      "loss": 1.4162,
      "step": 415
    },
    {
      "epoch": 0.17427733556765815,
      "grad_norm": 0.6490867733955383,
      "learning_rate": 1.9304566401340597e-05,
      "loss": 1.4837,
      "step": 416
    },
    {
      "epoch": 0.17469627147046501,
      "grad_norm": 0.538263738155365,
      "learning_rate": 1.9302890657729368e-05,
      "loss": 1.4829,
      "step": 417
    },
    {
      "epoch": 0.17511520737327188,
      "grad_norm": 0.5698223114013672,
      "learning_rate": 1.9301214914118142e-05,
      "loss": 1.5414,
      "step": 418
    },
    {
      "epoch": 0.17553414327607875,
      "grad_norm": 0.5576086640357971,
      "learning_rate": 1.9299539170506913e-05,
      "loss": 1.5554,
      "step": 419
    },
    {
      "epoch": 0.17595307917888564,
      "grad_norm": 0.5333095192909241,
      "learning_rate": 1.9297863426895687e-05,
      "loss": 1.5389,
      "step": 420
    },
    {
      "epoch": 0.1763720150816925,
      "grad_norm": 0.6559884548187256,
      "learning_rate": 1.929618768328446e-05,
      "loss": 1.4234,
      "step": 421
    },
    {
      "epoch": 0.17679095098449937,
      "grad_norm": 0.5460498332977295,
      "learning_rate": 1.9294511939673232e-05,
      "loss": 1.39,
      "step": 422
    },
    {
      "epoch": 0.17720988688730624,
      "grad_norm": 0.636465311050415,
      "learning_rate": 1.9292836196062003e-05,
      "loss": 1.5061,
      "step": 423
    },
    {
      "epoch": 0.1776288227901131,
      "grad_norm": 1.0990898609161377,
      "learning_rate": 1.9291160452450777e-05,
      "loss": 1.5252,
      "step": 424
    },
    {
      "epoch": 0.17804775869291997,
      "grad_norm": 0.7239777445793152,
      "learning_rate": 1.928948470883955e-05,
      "loss": 1.6089,
      "step": 425
    },
    {
      "epoch": 0.17846669459572687,
      "grad_norm": 0.5494760870933533,
      "learning_rate": 1.928780896522832e-05,
      "loss": 1.5254,
      "step": 426
    },
    {
      "epoch": 0.17888563049853373,
      "grad_norm": 0.8444346189498901,
      "learning_rate": 1.9286133221617092e-05,
      "loss": 1.5971,
      "step": 427
    },
    {
      "epoch": 0.1793045664013406,
      "grad_norm": 0.6326781511306763,
      "learning_rate": 1.9284457478005866e-05,
      "loss": 1.6235,
      "step": 428
    },
    {
      "epoch": 0.17972350230414746,
      "grad_norm": 1.274570107460022,
      "learning_rate": 1.928278173439464e-05,
      "loss": 1.3773,
      "step": 429
    },
    {
      "epoch": 0.18014243820695433,
      "grad_norm": 0.8181143999099731,
      "learning_rate": 1.928110599078341e-05,
      "loss": 1.5252,
      "step": 430
    },
    {
      "epoch": 0.1805613741097612,
      "grad_norm": 1.0652360916137695,
      "learning_rate": 1.9279430247172185e-05,
      "loss": 1.4825,
      "step": 431
    },
    {
      "epoch": 0.1809803100125681,
      "grad_norm": 0.8766518831253052,
      "learning_rate": 1.9277754503560956e-05,
      "loss": 1.5517,
      "step": 432
    },
    {
      "epoch": 0.18139924591537496,
      "grad_norm": 0.730766236782074,
      "learning_rate": 1.927607875994973e-05,
      "loss": 1.4122,
      "step": 433
    },
    {
      "epoch": 0.18181818181818182,
      "grad_norm": 0.8081750869750977,
      "learning_rate": 1.92744030163385e-05,
      "loss": 1.4633,
      "step": 434
    },
    {
      "epoch": 0.1822371177209887,
      "grad_norm": 0.5756118297576904,
      "learning_rate": 1.9272727272727275e-05,
      "loss": 1.5168,
      "step": 435
    },
    {
      "epoch": 0.18265605362379556,
      "grad_norm": 0.7646397948265076,
      "learning_rate": 1.9271051529116046e-05,
      "loss": 1.3793,
      "step": 436
    },
    {
      "epoch": 0.18307498952660242,
      "grad_norm": 0.5722802877426147,
      "learning_rate": 1.926937578550482e-05,
      "loss": 1.536,
      "step": 437
    },
    {
      "epoch": 0.1834939254294093,
      "grad_norm": 0.6689785718917847,
      "learning_rate": 1.926770004189359e-05,
      "loss": 1.5591,
      "step": 438
    },
    {
      "epoch": 0.18391286133221618,
      "grad_norm": 0.6856290102005005,
      "learning_rate": 1.9266024298282365e-05,
      "loss": 1.4544,
      "step": 439
    },
    {
      "epoch": 0.18433179723502305,
      "grad_norm": 0.6789899468421936,
      "learning_rate": 1.9264348554671136e-05,
      "loss": 1.4653,
      "step": 440
    },
    {
      "epoch": 0.18475073313782991,
      "grad_norm": 0.7284293174743652,
      "learning_rate": 1.926267281105991e-05,
      "loss": 1.5527,
      "step": 441
    },
    {
      "epoch": 0.18516966904063678,
      "grad_norm": 0.7645947337150574,
      "learning_rate": 1.926099706744868e-05,
      "loss": 1.5051,
      "step": 442
    },
    {
      "epoch": 0.18558860494344365,
      "grad_norm": 0.719478964805603,
      "learning_rate": 1.9259321323837455e-05,
      "loss": 1.5189,
      "step": 443
    },
    {
      "epoch": 0.1860075408462505,
      "grad_norm": 0.695095419883728,
      "learning_rate": 1.925764558022623e-05,
      "loss": 1.5547,
      "step": 444
    },
    {
      "epoch": 0.1864264767490574,
      "grad_norm": 0.8511472344398499,
      "learning_rate": 1.9255969836615e-05,
      "loss": 1.5407,
      "step": 445
    },
    {
      "epoch": 0.18684541265186427,
      "grad_norm": 0.8481591939926147,
      "learning_rate": 1.925429409300377e-05,
      "loss": 1.4753,
      "step": 446
    },
    {
      "epoch": 0.18726434855467114,
      "grad_norm": 0.5671373009681702,
      "learning_rate": 1.9252618349392544e-05,
      "loss": 1.6334,
      "step": 447
    },
    {
      "epoch": 0.187683284457478,
      "grad_norm": 0.9575358629226685,
      "learning_rate": 1.925094260578132e-05,
      "loss": 1.4334,
      "step": 448
    },
    {
      "epoch": 0.18810222036028487,
      "grad_norm": 0.7267637252807617,
      "learning_rate": 1.924926686217009e-05,
      "loss": 1.369,
      "step": 449
    },
    {
      "epoch": 0.18852115626309174,
      "grad_norm": 0.9400660991668701,
      "learning_rate": 1.924759111855886e-05,
      "loss": 1.4424,
      "step": 450
    },
    {
      "epoch": 0.1889400921658986,
      "grad_norm": 0.6154212951660156,
      "learning_rate": 1.9245915374947634e-05,
      "loss": 1.5379,
      "step": 451
    },
    {
      "epoch": 0.1893590280687055,
      "grad_norm": 0.992761492729187,
      "learning_rate": 1.9244239631336408e-05,
      "loss": 1.4291,
      "step": 452
    },
    {
      "epoch": 0.18977796397151236,
      "grad_norm": 0.8219314217567444,
      "learning_rate": 1.924256388772518e-05,
      "loss": 1.4851,
      "step": 453
    },
    {
      "epoch": 0.19019689987431923,
      "grad_norm": 0.6872371435165405,
      "learning_rate": 1.9240888144113953e-05,
      "loss": 1.4687,
      "step": 454
    },
    {
      "epoch": 0.1906158357771261,
      "grad_norm": 0.7827102541923523,
      "learning_rate": 1.9239212400502724e-05,
      "loss": 1.4223,
      "step": 455
    },
    {
      "epoch": 0.19103477167993296,
      "grad_norm": 0.6476536989212036,
      "learning_rate": 1.9237536656891498e-05,
      "loss": 1.4416,
      "step": 456
    },
    {
      "epoch": 0.19145370758273983,
      "grad_norm": 0.8523492813110352,
      "learning_rate": 1.923586091328027e-05,
      "loss": 1.4081,
      "step": 457
    },
    {
      "epoch": 0.19187264348554672,
      "grad_norm": 0.7393168807029724,
      "learning_rate": 1.9234185169669043e-05,
      "loss": 1.4165,
      "step": 458
    },
    {
      "epoch": 0.1922915793883536,
      "grad_norm": 0.5783189535140991,
      "learning_rate": 1.9232509426057814e-05,
      "loss": 1.5124,
      "step": 459
    },
    {
      "epoch": 0.19271051529116046,
      "grad_norm": 0.5648137927055359,
      "learning_rate": 1.9230833682446588e-05,
      "loss": 1.4118,
      "step": 460
    },
    {
      "epoch": 0.19312945119396732,
      "grad_norm": 0.6089081168174744,
      "learning_rate": 1.922915793883536e-05,
      "loss": 1.5284,
      "step": 461
    },
    {
      "epoch": 0.1935483870967742,
      "grad_norm": 0.4999605417251587,
      "learning_rate": 1.9227482195224133e-05,
      "loss": 1.5311,
      "step": 462
    },
    {
      "epoch": 0.19396732299958105,
      "grad_norm": 0.5938348174095154,
      "learning_rate": 1.9225806451612907e-05,
      "loss": 1.4788,
      "step": 463
    },
    {
      "epoch": 0.19438625890238795,
      "grad_norm": 0.547335684299469,
      "learning_rate": 1.9224130708001677e-05,
      "loss": 1.4408,
      "step": 464
    },
    {
      "epoch": 0.19480519480519481,
      "grad_norm": 0.6379075050354004,
      "learning_rate": 1.9222454964390448e-05,
      "loss": 1.5705,
      "step": 465
    },
    {
      "epoch": 0.19522413070800168,
      "grad_norm": 0.5942078828811646,
      "learning_rate": 1.9220779220779222e-05,
      "loss": 1.4643,
      "step": 466
    },
    {
      "epoch": 0.19564306661080855,
      "grad_norm": 0.5671261548995972,
      "learning_rate": 1.9219103477167996e-05,
      "loss": 1.5048,
      "step": 467
    },
    {
      "epoch": 0.1960620025136154,
      "grad_norm": 0.5360516905784607,
      "learning_rate": 1.9217427733556767e-05,
      "loss": 1.4651,
      "step": 468
    },
    {
      "epoch": 0.19648093841642228,
      "grad_norm": 0.6140695810317993,
      "learning_rate": 1.9215751989945538e-05,
      "loss": 1.4735,
      "step": 469
    },
    {
      "epoch": 0.19689987431922915,
      "grad_norm": 0.573509931564331,
      "learning_rate": 1.9214076246334312e-05,
      "loss": 1.459,
      "step": 470
    },
    {
      "epoch": 0.19731881022203604,
      "grad_norm": 0.6246591210365295,
      "learning_rate": 1.9212400502723086e-05,
      "loss": 1.4701,
      "step": 471
    },
    {
      "epoch": 0.1977377461248429,
      "grad_norm": 0.5583751797676086,
      "learning_rate": 1.9210724759111857e-05,
      "loss": 1.6073,
      "step": 472
    },
    {
      "epoch": 0.19815668202764977,
      "grad_norm": 0.7271811962127686,
      "learning_rate": 1.9209049015500628e-05,
      "loss": 1.4531,
      "step": 473
    },
    {
      "epoch": 0.19857561793045664,
      "grad_norm": 0.7514092922210693,
      "learning_rate": 1.9207373271889402e-05,
      "loss": 1.4609,
      "step": 474
    },
    {
      "epoch": 0.1989945538332635,
      "grad_norm": 0.6135318875312805,
      "learning_rate": 1.9205697528278176e-05,
      "loss": 1.46,
      "step": 475
    },
    {
      "epoch": 0.19941348973607037,
      "grad_norm": 0.8104654550552368,
      "learning_rate": 1.920402178466695e-05,
      "loss": 1.3627,
      "step": 476
    },
    {
      "epoch": 0.19983242563887726,
      "grad_norm": 0.6633170247077942,
      "learning_rate": 1.920234604105572e-05,
      "loss": 1.52,
      "step": 477
    },
    {
      "epoch": 0.20025136154168413,
      "grad_norm": 0.7291746735572815,
      "learning_rate": 1.920067029744449e-05,
      "loss": 1.547,
      "step": 478
    },
    {
      "epoch": 0.200670297444491,
      "grad_norm": 0.6674147248268127,
      "learning_rate": 1.9198994553833266e-05,
      "loss": 1.4795,
      "step": 479
    },
    {
      "epoch": 0.20108923334729786,
      "grad_norm": 0.9536328911781311,
      "learning_rate": 1.919731881022204e-05,
      "loss": 1.4581,
      "step": 480
    },
    {
      "epoch": 0.20150816925010473,
      "grad_norm": 0.8882992267608643,
      "learning_rate": 1.919564306661081e-05,
      "loss": 1.4351,
      "step": 481
    },
    {
      "epoch": 0.2019271051529116,
      "grad_norm": 0.5723230838775635,
      "learning_rate": 1.919396732299958e-05,
      "loss": 1.5095,
      "step": 482
    },
    {
      "epoch": 0.20234604105571846,
      "grad_norm": 1.1361421346664429,
      "learning_rate": 1.9192291579388355e-05,
      "loss": 1.4452,
      "step": 483
    },
    {
      "epoch": 0.20276497695852536,
      "grad_norm": 0.5762605667114258,
      "learning_rate": 1.9190615835777126e-05,
      "loss": 1.5695,
      "step": 484
    },
    {
      "epoch": 0.20318391286133222,
      "grad_norm": 0.9020100831985474,
      "learning_rate": 1.91889400921659e-05,
      "loss": 1.5157,
      "step": 485
    },
    {
      "epoch": 0.2036028487641391,
      "grad_norm": 0.6302100419998169,
      "learning_rate": 1.9187264348554674e-05,
      "loss": 1.5346,
      "step": 486
    },
    {
      "epoch": 0.20402178466694595,
      "grad_norm": 1.1093989610671997,
      "learning_rate": 1.9185588604943445e-05,
      "loss": 1.5348,
      "step": 487
    },
    {
      "epoch": 0.20444072056975282,
      "grad_norm": 0.637134850025177,
      "learning_rate": 1.9183912861332216e-05,
      "loss": 1.4478,
      "step": 488
    },
    {
      "epoch": 0.2048596564725597,
      "grad_norm": 0.9395270943641663,
      "learning_rate": 1.918223711772099e-05,
      "loss": 1.4659,
      "step": 489
    },
    {
      "epoch": 0.20527859237536658,
      "grad_norm": 0.7030125260353088,
      "learning_rate": 1.9180561374109764e-05,
      "loss": 1.3621,
      "step": 490
    },
    {
      "epoch": 0.20569752827817345,
      "grad_norm": 0.5922579169273376,
      "learning_rate": 1.9178885630498535e-05,
      "loss": 1.644,
      "step": 491
    },
    {
      "epoch": 0.2061164641809803,
      "grad_norm": 1.0161982774734497,
      "learning_rate": 1.9177209886887306e-05,
      "loss": 1.3895,
      "step": 492
    },
    {
      "epoch": 0.20653540008378718,
      "grad_norm": 1.179316759109497,
      "learning_rate": 1.917553414327608e-05,
      "loss": 1.4576,
      "step": 493
    },
    {
      "epoch": 0.20695433598659405,
      "grad_norm": 0.8130103945732117,
      "learning_rate": 1.9173858399664854e-05,
      "loss": 1.4637,
      "step": 494
    },
    {
      "epoch": 0.2073732718894009,
      "grad_norm": 0.7000212669372559,
      "learning_rate": 1.9172182656053625e-05,
      "loss": 1.5152,
      "step": 495
    },
    {
      "epoch": 0.2077922077922078,
      "grad_norm": 0.8932341933250427,
      "learning_rate": 1.91705069124424e-05,
      "loss": 1.4741,
      "step": 496
    },
    {
      "epoch": 0.20821114369501467,
      "grad_norm": 1.3901516199111938,
      "learning_rate": 1.916883116883117e-05,
      "loss": 1.4953,
      "step": 497
    },
    {
      "epoch": 0.20863007959782154,
      "grad_norm": 1.0053791999816895,
      "learning_rate": 1.9167155425219944e-05,
      "loss": 1.488,
      "step": 498
    },
    {
      "epoch": 0.2090490155006284,
      "grad_norm": 0.5231297612190247,
      "learning_rate": 1.9165479681608718e-05,
      "loss": 1.4963,
      "step": 499
    },
    {
      "epoch": 0.20946795140343527,
      "grad_norm": 0.8381982445716858,
      "learning_rate": 1.916380393799749e-05,
      "loss": 1.5261,
      "step": 500
    },
    {
      "epoch": 0.20988688730624214,
      "grad_norm": 0.9278205633163452,
      "learning_rate": 1.916212819438626e-05,
      "loss": 1.3259,
      "step": 501
    },
    {
      "epoch": 0.210305823209049,
      "grad_norm": 0.9388899803161621,
      "learning_rate": 1.9160452450775033e-05,
      "loss": 1.4091,
      "step": 502
    },
    {
      "epoch": 0.2107247591118559,
      "grad_norm": 1.0538358688354492,
      "learning_rate": 1.9158776707163807e-05,
      "loss": 1.4633,
      "step": 503
    },
    {
      "epoch": 0.21114369501466276,
      "grad_norm": 0.6670867800712585,
      "learning_rate": 1.9157100963552578e-05,
      "loss": 1.5639,
      "step": 504
    },
    {
      "epoch": 0.21156263091746963,
      "grad_norm": 1.7608953714370728,
      "learning_rate": 1.915542521994135e-05,
      "loss": 1.4433,
      "step": 505
    },
    {
      "epoch": 0.2119815668202765,
      "grad_norm": 1.1406441926956177,
      "learning_rate": 1.9153749476330123e-05,
      "loss": 1.3938,
      "step": 506
    },
    {
      "epoch": 0.21240050272308336,
      "grad_norm": 1.0751662254333496,
      "learning_rate": 1.9152073732718897e-05,
      "loss": 1.5059,
      "step": 507
    },
    {
      "epoch": 0.21281943862589023,
      "grad_norm": 1.7127149105072021,
      "learning_rate": 1.9150397989107668e-05,
      "loss": 1.4072,
      "step": 508
    },
    {
      "epoch": 0.21323837452869712,
      "grad_norm": 0.7607043385505676,
      "learning_rate": 1.9148722245496442e-05,
      "loss": 1.43,
      "step": 509
    },
    {
      "epoch": 0.213657310431504,
      "grad_norm": 1.0876507759094238,
      "learning_rate": 1.9147046501885213e-05,
      "loss": 1.4219,
      "step": 510
    },
    {
      "epoch": 0.21407624633431085,
      "grad_norm": 1.704974889755249,
      "learning_rate": 1.9145370758273987e-05,
      "loss": 1.5058,
      "step": 511
    },
    {
      "epoch": 0.21449518223711772,
      "grad_norm": 1.3107315301895142,
      "learning_rate": 1.9143695014662758e-05,
      "loss": 1.3701,
      "step": 512
    },
    {
      "epoch": 0.2149141181399246,
      "grad_norm": 0.72137850522995,
      "learning_rate": 1.9142019271051532e-05,
      "loss": 1.4608,
      "step": 513
    },
    {
      "epoch": 0.21533305404273145,
      "grad_norm": 0.9491326808929443,
      "learning_rate": 1.9140343527440303e-05,
      "loss": 1.3998,
      "step": 514
    },
    {
      "epoch": 0.21575198994553832,
      "grad_norm": 0.9301955103874207,
      "learning_rate": 1.9138667783829073e-05,
      "loss": 1.5215,
      "step": 515
    },
    {
      "epoch": 0.2161709258483452,
      "grad_norm": 0.8181395530700684,
      "learning_rate": 1.9136992040217847e-05,
      "loss": 1.4322,
      "step": 516
    },
    {
      "epoch": 0.21658986175115208,
      "grad_norm": 0.7477272748947144,
      "learning_rate": 1.913531629660662e-05,
      "loss": 1.4069,
      "step": 517
    },
    {
      "epoch": 0.21700879765395895,
      "grad_norm": 0.8078567385673523,
      "learning_rate": 1.9133640552995392e-05,
      "loss": 1.4061,
      "step": 518
    },
    {
      "epoch": 0.2174277335567658,
      "grad_norm": 0.6401208639144897,
      "learning_rate": 1.9131964809384166e-05,
      "loss": 1.4146,
      "step": 519
    },
    {
      "epoch": 0.21784666945957268,
      "grad_norm": 1.1645647287368774,
      "learning_rate": 1.9130289065772937e-05,
      "loss": 1.4205,
      "step": 520
    },
    {
      "epoch": 0.21826560536237954,
      "grad_norm": 0.5319536328315735,
      "learning_rate": 1.912861332216171e-05,
      "loss": 1.4255,
      "step": 521
    },
    {
      "epoch": 0.21868454126518644,
      "grad_norm": 0.7974638938903809,
      "learning_rate": 1.9126937578550485e-05,
      "loss": 1.4991,
      "step": 522
    },
    {
      "epoch": 0.2191034771679933,
      "grad_norm": 0.7385554313659668,
      "learning_rate": 1.9125261834939256e-05,
      "loss": 1.4917,
      "step": 523
    },
    {
      "epoch": 0.21952241307080017,
      "grad_norm": 0.6328215003013611,
      "learning_rate": 1.9123586091328027e-05,
      "loss": 1.4466,
      "step": 524
    },
    {
      "epoch": 0.21994134897360704,
      "grad_norm": 0.9712186455726624,
      "learning_rate": 1.91219103477168e-05,
      "loss": 1.3888,
      "step": 525
    },
    {
      "epoch": 0.2203602848764139,
      "grad_norm": 0.8162040710449219,
      "learning_rate": 1.9120234604105575e-05,
      "loss": 1.4592,
      "step": 526
    },
    {
      "epoch": 0.22077922077922077,
      "grad_norm": 0.6039196252822876,
      "learning_rate": 1.9118558860494346e-05,
      "loss": 1.4641,
      "step": 527
    },
    {
      "epoch": 0.22119815668202766,
      "grad_norm": 0.677533745765686,
      "learning_rate": 1.9116883116883117e-05,
      "loss": 1.4318,
      "step": 528
    },
    {
      "epoch": 0.22161709258483453,
      "grad_norm": 0.5307941436767578,
      "learning_rate": 1.911520737327189e-05,
      "loss": 1.4567,
      "step": 529
    },
    {
      "epoch": 0.2220360284876414,
      "grad_norm": 0.6290689706802368,
      "learning_rate": 1.9113531629660665e-05,
      "loss": 1.5327,
      "step": 530
    },
    {
      "epoch": 0.22245496439044826,
      "grad_norm": 0.668403685092926,
      "learning_rate": 1.9111855886049436e-05,
      "loss": 1.5186,
      "step": 531
    },
    {
      "epoch": 0.22287390029325513,
      "grad_norm": 0.8590381145477295,
      "learning_rate": 1.911018014243821e-05,
      "loss": 1.4103,
      "step": 532
    },
    {
      "epoch": 0.223292836196062,
      "grad_norm": 0.8472411036491394,
      "learning_rate": 1.910850439882698e-05,
      "loss": 1.4171,
      "step": 533
    },
    {
      "epoch": 0.22371177209886886,
      "grad_norm": 0.8556522130966187,
      "learning_rate": 1.9106828655215755e-05,
      "loss": 1.4334,
      "step": 534
    },
    {
      "epoch": 0.22413070800167575,
      "grad_norm": 0.6294506192207336,
      "learning_rate": 1.9105152911604525e-05,
      "loss": 1.4425,
      "step": 535
    },
    {
      "epoch": 0.22454964390448262,
      "grad_norm": 0.8159162402153015,
      "learning_rate": 1.91034771679933e-05,
      "loss": 1.5275,
      "step": 536
    },
    {
      "epoch": 0.2249685798072895,
      "grad_norm": 0.7663982510566711,
      "learning_rate": 1.910180142438207e-05,
      "loss": 1.4252,
      "step": 537
    },
    {
      "epoch": 0.22538751571009635,
      "grad_norm": 0.6042879223823547,
      "learning_rate": 1.9100125680770844e-05,
      "loss": 1.4374,
      "step": 538
    },
    {
      "epoch": 0.22580645161290322,
      "grad_norm": 0.5414412617683411,
      "learning_rate": 1.9098449937159615e-05,
      "loss": 1.4921,
      "step": 539
    },
    {
      "epoch": 0.22622538751571009,
      "grad_norm": 0.7135291695594788,
      "learning_rate": 1.909677419354839e-05,
      "loss": 1.5495,
      "step": 540
    },
    {
      "epoch": 0.22664432341851698,
      "grad_norm": 0.5732749104499817,
      "learning_rate": 1.9095098449937163e-05,
      "loss": 1.3577,
      "step": 541
    },
    {
      "epoch": 0.22706325932132385,
      "grad_norm": 0.6375070810317993,
      "learning_rate": 1.9093422706325934e-05,
      "loss": 1.4254,
      "step": 542
    },
    {
      "epoch": 0.2274821952241307,
      "grad_norm": 0.6467480063438416,
      "learning_rate": 1.9091746962714705e-05,
      "loss": 1.4277,
      "step": 543
    },
    {
      "epoch": 0.22790113112693758,
      "grad_norm": 0.6565422415733337,
      "learning_rate": 1.909007121910348e-05,
      "loss": 1.55,
      "step": 544
    },
    {
      "epoch": 0.22832006702974444,
      "grad_norm": 0.6464892625808716,
      "learning_rate": 1.9088395475492253e-05,
      "loss": 1.4626,
      "step": 545
    },
    {
      "epoch": 0.2287390029325513,
      "grad_norm": 0.6357992887496948,
      "learning_rate": 1.9086719731881024e-05,
      "loss": 1.4343,
      "step": 546
    },
    {
      "epoch": 0.22915793883535818,
      "grad_norm": 0.6153067946434021,
      "learning_rate": 1.9085043988269795e-05,
      "loss": 1.4146,
      "step": 547
    },
    {
      "epoch": 0.22957687473816507,
      "grad_norm": 0.7078065276145935,
      "learning_rate": 1.908336824465857e-05,
      "loss": 1.4147,
      "step": 548
    },
    {
      "epoch": 0.22999581064097194,
      "grad_norm": 0.6610170602798462,
      "learning_rate": 1.9081692501047343e-05,
      "loss": 1.5137,
      "step": 549
    },
    {
      "epoch": 0.2304147465437788,
      "grad_norm": 0.6916443705558777,
      "learning_rate": 1.9080016757436114e-05,
      "loss": 1.4726,
      "step": 550
    },
    {
      "epoch": 0.23083368244658567,
      "grad_norm": 0.668258547782898,
      "learning_rate": 1.9078341013824884e-05,
      "loss": 1.4796,
      "step": 551
    },
    {
      "epoch": 0.23125261834939254,
      "grad_norm": 0.6234722137451172,
      "learning_rate": 1.907666527021366e-05,
      "loss": 1.4962,
      "step": 552
    },
    {
      "epoch": 0.2316715542521994,
      "grad_norm": 0.7003756761550903,
      "learning_rate": 1.9074989526602433e-05,
      "loss": 1.4681,
      "step": 553
    },
    {
      "epoch": 0.2320904901550063,
      "grad_norm": 0.6396431922912598,
      "learning_rate": 1.9073313782991203e-05,
      "loss": 1.3989,
      "step": 554
    },
    {
      "epoch": 0.23250942605781316,
      "grad_norm": 0.6359259486198425,
      "learning_rate": 1.9071638039379977e-05,
      "loss": 1.4533,
      "step": 555
    },
    {
      "epoch": 0.23292836196062003,
      "grad_norm": 0.6927059292793274,
      "learning_rate": 1.9069962295768748e-05,
      "loss": 1.4107,
      "step": 556
    },
    {
      "epoch": 0.2333472978634269,
      "grad_norm": 0.633521318435669,
      "learning_rate": 1.9068286552157522e-05,
      "loss": 1.4099,
      "step": 557
    },
    {
      "epoch": 0.23376623376623376,
      "grad_norm": 0.6133984923362732,
      "learning_rate": 1.9066610808546293e-05,
      "loss": 1.425,
      "step": 558
    },
    {
      "epoch": 0.23418516966904063,
      "grad_norm": 0.661888062953949,
      "learning_rate": 1.9064935064935067e-05,
      "loss": 1.5631,
      "step": 559
    },
    {
      "epoch": 0.23460410557184752,
      "grad_norm": 0.645517110824585,
      "learning_rate": 1.9063259321323838e-05,
      "loss": 1.4026,
      "step": 560
    },
    {
      "epoch": 0.2350230414746544,
      "grad_norm": 0.5653700828552246,
      "learning_rate": 1.9061583577712612e-05,
      "loss": 1.413,
      "step": 561
    },
    {
      "epoch": 0.23544197737746125,
      "grad_norm": 0.7444579601287842,
      "learning_rate": 1.9059907834101383e-05,
      "loss": 1.4424,
      "step": 562
    },
    {
      "epoch": 0.23586091328026812,
      "grad_norm": 0.618651270866394,
      "learning_rate": 1.9058232090490157e-05,
      "loss": 1.6035,
      "step": 563
    },
    {
      "epoch": 0.23627984918307499,
      "grad_norm": 0.6566857695579529,
      "learning_rate": 1.905655634687893e-05,
      "loss": 1.3807,
      "step": 564
    },
    {
      "epoch": 0.23669878508588185,
      "grad_norm": 0.5518476963043213,
      "learning_rate": 1.9054880603267702e-05,
      "loss": 1.5754,
      "step": 565
    },
    {
      "epoch": 0.23711772098868872,
      "grad_norm": 0.5864123106002808,
      "learning_rate": 1.9053204859656472e-05,
      "loss": 1.4076,
      "step": 566
    },
    {
      "epoch": 0.2375366568914956,
      "grad_norm": 0.7522138953208923,
      "learning_rate": 1.9051529116045247e-05,
      "loss": 1.439,
      "step": 567
    },
    {
      "epoch": 0.23795559279430248,
      "grad_norm": 0.616507351398468,
      "learning_rate": 1.904985337243402e-05,
      "loss": 1.463,
      "step": 568
    },
    {
      "epoch": 0.23837452869710934,
      "grad_norm": 0.5815699100494385,
      "learning_rate": 1.904817762882279e-05,
      "loss": 1.4541,
      "step": 569
    },
    {
      "epoch": 0.2387934645999162,
      "grad_norm": 0.6003520488739014,
      "learning_rate": 1.9046501885211562e-05,
      "loss": 1.4897,
      "step": 570
    },
    {
      "epoch": 0.23921240050272308,
      "grad_norm": 0.6311242580413818,
      "learning_rate": 1.9044826141600336e-05,
      "loss": 1.4004,
      "step": 571
    },
    {
      "epoch": 0.23963133640552994,
      "grad_norm": 0.607297956943512,
      "learning_rate": 1.904315039798911e-05,
      "loss": 1.5698,
      "step": 572
    },
    {
      "epoch": 0.24005027230833684,
      "grad_norm": 0.7055544257164001,
      "learning_rate": 1.904147465437788e-05,
      "loss": 1.4389,
      "step": 573
    },
    {
      "epoch": 0.2404692082111437,
      "grad_norm": 0.7283361554145813,
      "learning_rate": 1.9039798910766655e-05,
      "loss": 1.4967,
      "step": 574
    },
    {
      "epoch": 0.24088814411395057,
      "grad_norm": 0.688834547996521,
      "learning_rate": 1.9038123167155426e-05,
      "loss": 1.4883,
      "step": 575
    },
    {
      "epoch": 0.24130708001675744,
      "grad_norm": 0.47785529494285583,
      "learning_rate": 1.90364474235442e-05,
      "loss": 1.5188,
      "step": 576
    },
    {
      "epoch": 0.2417260159195643,
      "grad_norm": 0.9133197069168091,
      "learning_rate": 1.903477167993297e-05,
      "loss": 1.3748,
      "step": 577
    },
    {
      "epoch": 0.24214495182237117,
      "grad_norm": 0.6180113554000854,
      "learning_rate": 1.9033095936321745e-05,
      "loss": 1.4213,
      "step": 578
    },
    {
      "epoch": 0.24256388772517803,
      "grad_norm": 0.6890195608139038,
      "learning_rate": 1.9031420192710516e-05,
      "loss": 1.3621,
      "step": 579
    },
    {
      "epoch": 0.24298282362798493,
      "grad_norm": 0.6566532850265503,
      "learning_rate": 1.902974444909929e-05,
      "loss": 1.3666,
      "step": 580
    },
    {
      "epoch": 0.2434017595307918,
      "grad_norm": 0.572672963142395,
      "learning_rate": 1.902806870548806e-05,
      "loss": 1.3654,
      "step": 581
    },
    {
      "epoch": 0.24382069543359866,
      "grad_norm": 0.698391318321228,
      "learning_rate": 1.9026392961876835e-05,
      "loss": 1.4244,
      "step": 582
    },
    {
      "epoch": 0.24423963133640553,
      "grad_norm": 0.5516411066055298,
      "learning_rate": 1.9024717218265606e-05,
      "loss": 1.5141,
      "step": 583
    },
    {
      "epoch": 0.2446585672392124,
      "grad_norm": 0.6122570037841797,
      "learning_rate": 1.902304147465438e-05,
      "loss": 1.4333,
      "step": 584
    },
    {
      "epoch": 0.24507750314201926,
      "grad_norm": 0.7202199697494507,
      "learning_rate": 1.902136573104315e-05,
      "loss": 1.4562,
      "step": 585
    },
    {
      "epoch": 0.24549643904482615,
      "grad_norm": 0.938453733921051,
      "learning_rate": 1.9019689987431925e-05,
      "loss": 1.4605,
      "step": 586
    },
    {
      "epoch": 0.24591537494763302,
      "grad_norm": 0.5673055648803711,
      "learning_rate": 1.90180142438207e-05,
      "loss": 1.4827,
      "step": 587
    },
    {
      "epoch": 0.24633431085043989,
      "grad_norm": 0.677293062210083,
      "learning_rate": 1.901633850020947e-05,
      "loss": 1.375,
      "step": 588
    },
    {
      "epoch": 0.24675324675324675,
      "grad_norm": 0.7138928771018982,
      "learning_rate": 1.901466275659824e-05,
      "loss": 1.541,
      "step": 589
    },
    {
      "epoch": 0.24717218265605362,
      "grad_norm": 0.7003957033157349,
      "learning_rate": 1.9012987012987014e-05,
      "loss": 1.4234,
      "step": 590
    },
    {
      "epoch": 0.24759111855886048,
      "grad_norm": 0.8985845446586609,
      "learning_rate": 1.901131126937579e-05,
      "loss": 1.2981,
      "step": 591
    },
    {
      "epoch": 0.24801005446166738,
      "grad_norm": 0.6559215784072876,
      "learning_rate": 1.900963552576456e-05,
      "loss": 1.4871,
      "step": 592
    },
    {
      "epoch": 0.24842899036447424,
      "grad_norm": 1.0038827657699585,
      "learning_rate": 1.900795978215333e-05,
      "loss": 1.3604,
      "step": 593
    },
    {
      "epoch": 0.2488479262672811,
      "grad_norm": 0.670703649520874,
      "learning_rate": 1.9006284038542104e-05,
      "loss": 1.4142,
      "step": 594
    },
    {
      "epoch": 0.24926686217008798,
      "grad_norm": 0.6743998527526855,
      "learning_rate": 1.9004608294930878e-05,
      "loss": 1.421,
      "step": 595
    },
    {
      "epoch": 0.24968579807289484,
      "grad_norm": 0.90179044008255,
      "learning_rate": 1.900293255131965e-05,
      "loss": 1.3545,
      "step": 596
    },
    {
      "epoch": 0.2501047339757017,
      "grad_norm": 0.7004808187484741,
      "learning_rate": 1.9001256807708423e-05,
      "loss": 1.4334,
      "step": 597
    },
    {
      "epoch": 0.2505236698785086,
      "grad_norm": 0.811750590801239,
      "learning_rate": 1.8999581064097194e-05,
      "loss": 1.5068,
      "step": 598
    },
    {
      "epoch": 0.25094260578131544,
      "grad_norm": 0.5660247206687927,
      "learning_rate": 1.8997905320485968e-05,
      "loss": 1.4119,
      "step": 599
    },
    {
      "epoch": 0.25136154168412234,
      "grad_norm": 0.7421470880508423,
      "learning_rate": 1.8996229576874742e-05,
      "loss": 1.4193,
      "step": 600
    },
    {
      "epoch": 0.2517804775869292,
      "grad_norm": 0.7964827418327332,
      "learning_rate": 1.8994553833263513e-05,
      "loss": 1.3664,
      "step": 601
    },
    {
      "epoch": 0.25219941348973607,
      "grad_norm": 0.587858259677887,
      "learning_rate": 1.8992878089652283e-05,
      "loss": 1.4094,
      "step": 602
    },
    {
      "epoch": 0.25261834939254296,
      "grad_norm": 0.8728364109992981,
      "learning_rate": 1.8991202346041058e-05,
      "loss": 1.2841,
      "step": 603
    },
    {
      "epoch": 0.2530372852953498,
      "grad_norm": 0.7561309933662415,
      "learning_rate": 1.8989526602429832e-05,
      "loss": 1.351,
      "step": 604
    },
    {
      "epoch": 0.2534562211981567,
      "grad_norm": 0.6205732822418213,
      "learning_rate": 1.8987850858818602e-05,
      "loss": 1.4043,
      "step": 605
    },
    {
      "epoch": 0.25387515710096353,
      "grad_norm": 0.7390834093093872,
      "learning_rate": 1.8986175115207373e-05,
      "loss": 1.4274,
      "step": 606
    },
    {
      "epoch": 0.2542940930037704,
      "grad_norm": 0.7050930261611938,
      "learning_rate": 1.8984499371596147e-05,
      "loss": 1.4875,
      "step": 607
    },
    {
      "epoch": 0.2547130289065773,
      "grad_norm": 0.5341117978096008,
      "learning_rate": 1.898282362798492e-05,
      "loss": 1.4482,
      "step": 608
    },
    {
      "epoch": 0.25513196480938416,
      "grad_norm": 0.8297886848449707,
      "learning_rate": 1.8981147884373692e-05,
      "loss": 1.3514,
      "step": 609
    },
    {
      "epoch": 0.25555090071219105,
      "grad_norm": 0.6667957901954651,
      "learning_rate": 1.8979472140762466e-05,
      "loss": 1.3185,
      "step": 610
    },
    {
      "epoch": 0.2559698366149979,
      "grad_norm": 0.6411688923835754,
      "learning_rate": 1.8977796397151237e-05,
      "loss": 1.4442,
      "step": 611
    },
    {
      "epoch": 0.2563887725178048,
      "grad_norm": 0.6379684805870056,
      "learning_rate": 1.8976120653540008e-05,
      "loss": 1.4377,
      "step": 612
    },
    {
      "epoch": 0.2568077084206116,
      "grad_norm": 0.6216014623641968,
      "learning_rate": 1.8974444909928782e-05,
      "loss": 1.4474,
      "step": 613
    },
    {
      "epoch": 0.2572266443234185,
      "grad_norm": 0.5671195983886719,
      "learning_rate": 1.8972769166317556e-05,
      "loss": 1.4434,
      "step": 614
    },
    {
      "epoch": 0.2576455802262254,
      "grad_norm": 0.8159009218215942,
      "learning_rate": 1.8971093422706327e-05,
      "loss": 1.3781,
      "step": 615
    },
    {
      "epoch": 0.25806451612903225,
      "grad_norm": 0.6750363111495972,
      "learning_rate": 1.8969417679095098e-05,
      "loss": 1.5244,
      "step": 616
    },
    {
      "epoch": 0.25848345203183914,
      "grad_norm": 0.5345395803451538,
      "learning_rate": 1.896774193548387e-05,
      "loss": 1.4944,
      "step": 617
    },
    {
      "epoch": 0.258902387934646,
      "grad_norm": 1.0422483682632446,
      "learning_rate": 1.8966066191872646e-05,
      "loss": 1.4257,
      "step": 618
    },
    {
      "epoch": 0.2593213238374529,
      "grad_norm": 0.6644807457923889,
      "learning_rate": 1.896439044826142e-05,
      "loss": 1.4396,
      "step": 619
    },
    {
      "epoch": 0.2597402597402597,
      "grad_norm": 0.6991039514541626,
      "learning_rate": 1.896271470465019e-05,
      "loss": 1.4456,
      "step": 620
    },
    {
      "epoch": 0.2601591956430666,
      "grad_norm": 0.6287118792533875,
      "learning_rate": 1.896103896103896e-05,
      "loss": 1.3556,
      "step": 621
    },
    {
      "epoch": 0.2605781315458735,
      "grad_norm": 0.7149505615234375,
      "learning_rate": 1.8959363217427736e-05,
      "loss": 1.3708,
      "step": 622
    },
    {
      "epoch": 0.26099706744868034,
      "grad_norm": 0.6485461592674255,
      "learning_rate": 1.895768747381651e-05,
      "loss": 1.3813,
      "step": 623
    },
    {
      "epoch": 0.26141600335148724,
      "grad_norm": 0.796234130859375,
      "learning_rate": 1.895601173020528e-05,
      "loss": 1.4557,
      "step": 624
    },
    {
      "epoch": 0.2618349392542941,
      "grad_norm": 0.6059120297431946,
      "learning_rate": 1.895433598659405e-05,
      "loss": 1.4393,
      "step": 625
    },
    {
      "epoch": 0.26225387515710097,
      "grad_norm": 0.7328737378120422,
      "learning_rate": 1.8952660242982825e-05,
      "loss": 1.4259,
      "step": 626
    },
    {
      "epoch": 0.2626728110599078,
      "grad_norm": 0.5874695181846619,
      "learning_rate": 1.89509844993716e-05,
      "loss": 1.51,
      "step": 627
    },
    {
      "epoch": 0.2630917469627147,
      "grad_norm": 0.7244255542755127,
      "learning_rate": 1.894930875576037e-05,
      "loss": 1.4277,
      "step": 628
    },
    {
      "epoch": 0.2635106828655216,
      "grad_norm": 0.5406452417373657,
      "learning_rate": 1.894763301214914e-05,
      "loss": 1.4015,
      "step": 629
    },
    {
      "epoch": 0.26392961876832843,
      "grad_norm": 0.7187069058418274,
      "learning_rate": 1.8945957268537915e-05,
      "loss": 1.417,
      "step": 630
    },
    {
      "epoch": 0.2643485546711353,
      "grad_norm": 0.6457657217979431,
      "learning_rate": 1.894428152492669e-05,
      "loss": 1.4711,
      "step": 631
    },
    {
      "epoch": 0.26476749057394217,
      "grad_norm": 0.5307900309562683,
      "learning_rate": 1.894260578131546e-05,
      "loss": 1.4237,
      "step": 632
    },
    {
      "epoch": 0.26518642647674906,
      "grad_norm": 0.9521127343177795,
      "learning_rate": 1.8940930037704234e-05,
      "loss": 1.3934,
      "step": 633
    },
    {
      "epoch": 0.26560536237955595,
      "grad_norm": 0.596931517124176,
      "learning_rate": 1.8939254294093005e-05,
      "loss": 1.4957,
      "step": 634
    },
    {
      "epoch": 0.2660242982823628,
      "grad_norm": 0.8038269281387329,
      "learning_rate": 1.893757855048178e-05,
      "loss": 1.4551,
      "step": 635
    },
    {
      "epoch": 0.2664432341851697,
      "grad_norm": 0.7987629771232605,
      "learning_rate": 1.893590280687055e-05,
      "loss": 1.4973,
      "step": 636
    },
    {
      "epoch": 0.2668621700879765,
      "grad_norm": 0.7971741557121277,
      "learning_rate": 1.8934227063259324e-05,
      "loss": 1.331,
      "step": 637
    },
    {
      "epoch": 0.2672811059907834,
      "grad_norm": 0.6776058077812195,
      "learning_rate": 1.8932551319648094e-05,
      "loss": 1.3397,
      "step": 638
    },
    {
      "epoch": 0.26770004189359026,
      "grad_norm": 0.7527434825897217,
      "learning_rate": 1.893087557603687e-05,
      "loss": 1.3912,
      "step": 639
    },
    {
      "epoch": 0.26811897779639715,
      "grad_norm": 0.6726639866828918,
      "learning_rate": 1.892919983242564e-05,
      "loss": 1.3428,
      "step": 640
    },
    {
      "epoch": 0.26853791369920404,
      "grad_norm": 0.7143589854240417,
      "learning_rate": 1.8927524088814413e-05,
      "loss": 1.4311,
      "step": 641
    },
    {
      "epoch": 0.2689568496020109,
      "grad_norm": 0.6544737815856934,
      "learning_rate": 1.8925848345203188e-05,
      "loss": 1.4774,
      "step": 642
    },
    {
      "epoch": 0.2693757855048178,
      "grad_norm": 0.707067608833313,
      "learning_rate": 1.892417260159196e-05,
      "loss": 1.3982,
      "step": 643
    },
    {
      "epoch": 0.2697947214076246,
      "grad_norm": 0.7015464901924133,
      "learning_rate": 1.892249685798073e-05,
      "loss": 1.3921,
      "step": 644
    },
    {
      "epoch": 0.2702136573104315,
      "grad_norm": 0.7379519939422607,
      "learning_rate": 1.8920821114369503e-05,
      "loss": 1.4429,
      "step": 645
    },
    {
      "epoch": 0.27063259321323835,
      "grad_norm": 0.6990654468536377,
      "learning_rate": 1.8919145370758277e-05,
      "loss": 1.3888,
      "step": 646
    },
    {
      "epoch": 0.27105152911604524,
      "grad_norm": 0.8418135643005371,
      "learning_rate": 1.8917469627147048e-05,
      "loss": 1.3989,
      "step": 647
    },
    {
      "epoch": 0.27147046501885214,
      "grad_norm": 0.7323577404022217,
      "learning_rate": 1.891579388353582e-05,
      "loss": 1.4053,
      "step": 648
    },
    {
      "epoch": 0.271889400921659,
      "grad_norm": 0.5955424904823303,
      "learning_rate": 1.8914118139924593e-05,
      "loss": 1.4684,
      "step": 649
    },
    {
      "epoch": 0.27230833682446587,
      "grad_norm": 0.7422502040863037,
      "learning_rate": 1.8912442396313367e-05,
      "loss": 1.394,
      "step": 650
    },
    {
      "epoch": 0.2727272727272727,
      "grad_norm": 0.6807714700698853,
      "learning_rate": 1.8910766652702138e-05,
      "loss": 1.315,
      "step": 651
    },
    {
      "epoch": 0.2731462086300796,
      "grad_norm": 0.6865885257720947,
      "learning_rate": 1.8909090909090912e-05,
      "loss": 1.3815,
      "step": 652
    },
    {
      "epoch": 0.2735651445328865,
      "grad_norm": 0.7024874091148376,
      "learning_rate": 1.8907415165479683e-05,
      "loss": 1.4215,
      "step": 653
    },
    {
      "epoch": 0.27398408043569333,
      "grad_norm": 0.7904298305511475,
      "learning_rate": 1.8905739421868457e-05,
      "loss": 1.3908,
      "step": 654
    },
    {
      "epoch": 0.2744030163385002,
      "grad_norm": 0.6477106213569641,
      "learning_rate": 1.8904063678257228e-05,
      "loss": 1.4599,
      "step": 655
    },
    {
      "epoch": 0.27482195224130707,
      "grad_norm": 0.8508483171463013,
      "learning_rate": 1.8902387934646e-05,
      "loss": 1.4838,
      "step": 656
    },
    {
      "epoch": 0.27524088814411396,
      "grad_norm": 0.5711894631385803,
      "learning_rate": 1.8900712191034772e-05,
      "loss": 1.4615,
      "step": 657
    },
    {
      "epoch": 0.2756598240469208,
      "grad_norm": 0.6993704438209534,
      "learning_rate": 1.8899036447423547e-05,
      "loss": 1.4493,
      "step": 658
    },
    {
      "epoch": 0.2760787599497277,
      "grad_norm": 1.0374524593353271,
      "learning_rate": 1.8897360703812317e-05,
      "loss": 1.482,
      "step": 659
    },
    {
      "epoch": 0.2764976958525346,
      "grad_norm": 0.9369934797286987,
      "learning_rate": 1.889568496020109e-05,
      "loss": 1.3472,
      "step": 660
    },
    {
      "epoch": 0.2769166317553414,
      "grad_norm": 0.8173794746398926,
      "learning_rate": 1.8894009216589862e-05,
      "loss": 1.4428,
      "step": 661
    },
    {
      "epoch": 0.2773355676581483,
      "grad_norm": 1.059898018836975,
      "learning_rate": 1.8892333472978636e-05,
      "loss": 1.395,
      "step": 662
    },
    {
      "epoch": 0.27775450356095516,
      "grad_norm": 0.5998404026031494,
      "learning_rate": 1.8890657729367407e-05,
      "loss": 1.3945,
      "step": 663
    },
    {
      "epoch": 0.27817343946376205,
      "grad_norm": 0.6833025217056274,
      "learning_rate": 1.888898198575618e-05,
      "loss": 1.4655,
      "step": 664
    },
    {
      "epoch": 0.2785923753665689,
      "grad_norm": 0.7907499670982361,
      "learning_rate": 1.8887306242144955e-05,
      "loss": 1.3141,
      "step": 665
    },
    {
      "epoch": 0.2790113112693758,
      "grad_norm": 0.6411296129226685,
      "learning_rate": 1.8885630498533726e-05,
      "loss": 1.507,
      "step": 666
    },
    {
      "epoch": 0.2794302471721827,
      "grad_norm": 0.8076937794685364,
      "learning_rate": 1.8883954754922497e-05,
      "loss": 1.3578,
      "step": 667
    },
    {
      "epoch": 0.2798491830749895,
      "grad_norm": 0.6930973529815674,
      "learning_rate": 1.888227901131127e-05,
      "loss": 1.3186,
      "step": 668
    },
    {
      "epoch": 0.2802681189777964,
      "grad_norm": 0.7560698390007019,
      "learning_rate": 1.8880603267700045e-05,
      "loss": 1.3772,
      "step": 669
    },
    {
      "epoch": 0.28068705488060325,
      "grad_norm": 0.5842419266700745,
      "learning_rate": 1.8878927524088816e-05,
      "loss": 1.4772,
      "step": 670
    },
    {
      "epoch": 0.28110599078341014,
      "grad_norm": 0.9055956602096558,
      "learning_rate": 1.8877251780477586e-05,
      "loss": 1.4802,
      "step": 671
    },
    {
      "epoch": 0.28152492668621704,
      "grad_norm": 0.5749624371528625,
      "learning_rate": 1.887557603686636e-05,
      "loss": 1.345,
      "step": 672
    },
    {
      "epoch": 0.2819438625890239,
      "grad_norm": 0.6721547245979309,
      "learning_rate": 1.8873900293255135e-05,
      "loss": 1.3641,
      "step": 673
    },
    {
      "epoch": 0.28236279849183077,
      "grad_norm": 0.580252468585968,
      "learning_rate": 1.8872224549643905e-05,
      "loss": 1.4225,
      "step": 674
    },
    {
      "epoch": 0.2827817343946376,
      "grad_norm": 0.8793032765388489,
      "learning_rate": 1.887054880603268e-05,
      "loss": 1.3721,
      "step": 675
    },
    {
      "epoch": 0.2832006702974445,
      "grad_norm": 0.5774890184402466,
      "learning_rate": 1.886887306242145e-05,
      "loss": 1.4358,
      "step": 676
    },
    {
      "epoch": 0.28361960620025134,
      "grad_norm": 0.6694848537445068,
      "learning_rate": 1.8867197318810224e-05,
      "loss": 1.3601,
      "step": 677
    },
    {
      "epoch": 0.28403854210305823,
      "grad_norm": 0.5719591975212097,
      "learning_rate": 1.8865521575198995e-05,
      "loss": 1.526,
      "step": 678
    },
    {
      "epoch": 0.2844574780058651,
      "grad_norm": 0.6238727569580078,
      "learning_rate": 1.886384583158777e-05,
      "loss": 1.3982,
      "step": 679
    },
    {
      "epoch": 0.28487641390867197,
      "grad_norm": 0.6695346236228943,
      "learning_rate": 1.886217008797654e-05,
      "loss": 1.384,
      "step": 680
    },
    {
      "epoch": 0.28529534981147886,
      "grad_norm": 0.7487378120422363,
      "learning_rate": 1.8860494344365314e-05,
      "loss": 1.364,
      "step": 681
    },
    {
      "epoch": 0.2857142857142857,
      "grad_norm": 0.6982246041297913,
      "learning_rate": 1.8858818600754085e-05,
      "loss": 1.4175,
      "step": 682
    },
    {
      "epoch": 0.2861332216170926,
      "grad_norm": 0.6678481698036194,
      "learning_rate": 1.885714285714286e-05,
      "loss": 1.3014,
      "step": 683
    },
    {
      "epoch": 0.28655215751989943,
      "grad_norm": 0.5933229923248291,
      "learning_rate": 1.885546711353163e-05,
      "loss": 1.4171,
      "step": 684
    },
    {
      "epoch": 0.2869710934227063,
      "grad_norm": 0.6992403864860535,
      "learning_rate": 1.8853791369920404e-05,
      "loss": 1.3231,
      "step": 685
    },
    {
      "epoch": 0.2873900293255132,
      "grad_norm": 0.4942586421966553,
      "learning_rate": 1.8852115626309175e-05,
      "loss": 1.4685,
      "step": 686
    },
    {
      "epoch": 0.28780896522832006,
      "grad_norm": 0.6683540940284729,
      "learning_rate": 1.885043988269795e-05,
      "loss": 1.3826,
      "step": 687
    },
    {
      "epoch": 0.28822790113112695,
      "grad_norm": 0.4600916802883148,
      "learning_rate": 1.8848764139086723e-05,
      "loss": 1.4668,
      "step": 688
    },
    {
      "epoch": 0.2886468370339338,
      "grad_norm": 0.6156571507453918,
      "learning_rate": 1.8847088395475494e-05,
      "loss": 1.4465,
      "step": 689
    },
    {
      "epoch": 0.2890657729367407,
      "grad_norm": 0.5449684262275696,
      "learning_rate": 1.8845412651864264e-05,
      "loss": 1.2964,
      "step": 690
    },
    {
      "epoch": 0.2894847088395475,
      "grad_norm": 0.5813759565353394,
      "learning_rate": 1.884373690825304e-05,
      "loss": 1.4211,
      "step": 691
    },
    {
      "epoch": 0.2899036447423544,
      "grad_norm": 0.618743360042572,
      "learning_rate": 1.8842061164641813e-05,
      "loss": 1.3976,
      "step": 692
    },
    {
      "epoch": 0.2903225806451613,
      "grad_norm": 0.5242084264755249,
      "learning_rate": 1.8840385421030583e-05,
      "loss": 1.4076,
      "step": 693
    },
    {
      "epoch": 0.29074151654796815,
      "grad_norm": 0.676110029220581,
      "learning_rate": 1.8838709677419354e-05,
      "loss": 1.3143,
      "step": 694
    },
    {
      "epoch": 0.29116045245077504,
      "grad_norm": 0.5773074626922607,
      "learning_rate": 1.8837033933808128e-05,
      "loss": 1.4443,
      "step": 695
    },
    {
      "epoch": 0.2915793883535819,
      "grad_norm": 0.8076342940330505,
      "learning_rate": 1.8835358190196902e-05,
      "loss": 1.469,
      "step": 696
    },
    {
      "epoch": 0.2919983242563888,
      "grad_norm": 0.6035589575767517,
      "learning_rate": 1.8833682446585677e-05,
      "loss": 1.349,
      "step": 697
    },
    {
      "epoch": 0.29241726015919567,
      "grad_norm": 1.1530474424362183,
      "learning_rate": 1.8832006702974447e-05,
      "loss": 1.4056,
      "step": 698
    },
    {
      "epoch": 0.2928361960620025,
      "grad_norm": 0.6804980635643005,
      "learning_rate": 1.8830330959363218e-05,
      "loss": 1.45,
      "step": 699
    },
    {
      "epoch": 0.2932551319648094,
      "grad_norm": 0.8612604141235352,
      "learning_rate": 1.8828655215751992e-05,
      "loss": 1.4082,
      "step": 700
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 11935,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 1.4362595049865216e+18,
  "train_batch_size": 12,
  "trial_name": null,
  "trial_params": null
}