final_model_nopolicy / trainer_state.json
yuzhounie's picture
End of training
187296f verified
{
"best_global_step": null,
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 1.998837479655894,
"eval_steps": 500,
"global_step": 1432,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0013950244129272262,
"grad_norm": 9.077165796122205,
"learning_rate": 0.0,
"loss": 1.244,
"step": 1
},
{
"epoch": 0.0027900488258544524,
"grad_norm": 7.807074173265791,
"learning_rate": 6.944444444444444e-08,
"loss": 1.1223,
"step": 2
},
{
"epoch": 0.0041850732387816785,
"grad_norm": 6.324174799146362,
"learning_rate": 1.3888888888888888e-07,
"loss": 0.9673,
"step": 3
},
{
"epoch": 0.005580097651708905,
"grad_norm": 8.27030483810195,
"learning_rate": 2.0833333333333333e-07,
"loss": 1.1685,
"step": 4
},
{
"epoch": 0.006975122064636131,
"grad_norm": 7.60016514591344,
"learning_rate": 2.7777777777777776e-07,
"loss": 1.0968,
"step": 5
},
{
"epoch": 0.008370146477563357,
"grad_norm": 8.060333500293979,
"learning_rate": 3.472222222222223e-07,
"loss": 1.1373,
"step": 6
},
{
"epoch": 0.009765170890490584,
"grad_norm": 8.048732320406836,
"learning_rate": 4.1666666666666667e-07,
"loss": 1.2183,
"step": 7
},
{
"epoch": 0.01116019530341781,
"grad_norm": 8.367676393723716,
"learning_rate": 4.861111111111112e-07,
"loss": 1.2232,
"step": 8
},
{
"epoch": 0.012555219716345037,
"grad_norm": 6.936859591765737,
"learning_rate": 5.555555555555555e-07,
"loss": 1.0223,
"step": 9
},
{
"epoch": 0.013950244129272262,
"grad_norm": 7.474340880350743,
"learning_rate": 6.25e-07,
"loss": 1.0451,
"step": 10
},
{
"epoch": 0.015345268542199489,
"grad_norm": 7.924791845739248,
"learning_rate": 6.944444444444446e-07,
"loss": 1.0904,
"step": 11
},
{
"epoch": 0.016740292955126714,
"grad_norm": 6.700238239267023,
"learning_rate": 7.63888888888889e-07,
"loss": 1.0017,
"step": 12
},
{
"epoch": 0.01813531736805394,
"grad_norm": 7.121912596616418,
"learning_rate": 8.333333333333333e-07,
"loss": 1.1195,
"step": 13
},
{
"epoch": 0.01953034178098117,
"grad_norm": 7.593714162127442,
"learning_rate": 9.027777777777779e-07,
"loss": 1.1526,
"step": 14
},
{
"epoch": 0.020925366193908394,
"grad_norm": 5.598325711309053,
"learning_rate": 9.722222222222224e-07,
"loss": 0.9568,
"step": 15
},
{
"epoch": 0.02232039060683562,
"grad_norm": 6.502502784546973,
"learning_rate": 1.0416666666666667e-06,
"loss": 1.0338,
"step": 16
},
{
"epoch": 0.023715415019762844,
"grad_norm": 4.2688358482531585,
"learning_rate": 1.111111111111111e-06,
"loss": 0.8272,
"step": 17
},
{
"epoch": 0.025110439432690073,
"grad_norm": 6.141756294060148,
"learning_rate": 1.1805555555555556e-06,
"loss": 1.1201,
"step": 18
},
{
"epoch": 0.0265054638456173,
"grad_norm": 3.803027333138855,
"learning_rate": 1.25e-06,
"loss": 0.873,
"step": 19
},
{
"epoch": 0.027900488258544524,
"grad_norm": 5.014334478810172,
"learning_rate": 1.3194444444444446e-06,
"loss": 0.9899,
"step": 20
},
{
"epoch": 0.029295512671471752,
"grad_norm": 4.551325462075463,
"learning_rate": 1.3888888888888892e-06,
"loss": 0.942,
"step": 21
},
{
"epoch": 0.030690537084398978,
"grad_norm": 4.812375868902633,
"learning_rate": 1.4583333333333335e-06,
"loss": 0.9865,
"step": 22
},
{
"epoch": 0.03208556149732621,
"grad_norm": 4.357883881047429,
"learning_rate": 1.527777777777778e-06,
"loss": 0.9614,
"step": 23
},
{
"epoch": 0.03348058591025343,
"grad_norm": 3.6268224514828318,
"learning_rate": 1.5972222222222221e-06,
"loss": 0.8257,
"step": 24
},
{
"epoch": 0.03487561032318066,
"grad_norm": 3.4656554743615047,
"learning_rate": 1.6666666666666667e-06,
"loss": 0.9127,
"step": 25
},
{
"epoch": 0.03627063473610788,
"grad_norm": 3.876455217129128,
"learning_rate": 1.7361111111111112e-06,
"loss": 0.9278,
"step": 26
},
{
"epoch": 0.03766565914903511,
"grad_norm": 4.09517147857438,
"learning_rate": 1.8055555555555557e-06,
"loss": 0.9556,
"step": 27
},
{
"epoch": 0.03906068356196234,
"grad_norm": 3.5545716555392737,
"learning_rate": 1.8750000000000003e-06,
"loss": 0.8878,
"step": 28
},
{
"epoch": 0.04045570797488956,
"grad_norm": 3.835208660111137,
"learning_rate": 1.944444444444445e-06,
"loss": 0.885,
"step": 29
},
{
"epoch": 0.04185073238781679,
"grad_norm": 3.531205403980982,
"learning_rate": 2.0138888888888893e-06,
"loss": 0.886,
"step": 30
},
{
"epoch": 0.043245756800744016,
"grad_norm": 4.231398251952698,
"learning_rate": 2.0833333333333334e-06,
"loss": 0.8223,
"step": 31
},
{
"epoch": 0.04464078121367124,
"grad_norm": 3.3069565195805386,
"learning_rate": 2.152777777777778e-06,
"loss": 0.8934,
"step": 32
},
{
"epoch": 0.04603580562659847,
"grad_norm": 3.3388892924336737,
"learning_rate": 2.222222222222222e-06,
"loss": 0.9336,
"step": 33
},
{
"epoch": 0.04743083003952569,
"grad_norm": 3.0257164444225895,
"learning_rate": 2.2916666666666666e-06,
"loss": 0.8879,
"step": 34
},
{
"epoch": 0.04882585445245292,
"grad_norm": 3.228639273179126,
"learning_rate": 2.361111111111111e-06,
"loss": 0.8922,
"step": 35
},
{
"epoch": 0.050220878865380146,
"grad_norm": 3.620966182021972,
"learning_rate": 2.4305555555555557e-06,
"loss": 0.9343,
"step": 36
},
{
"epoch": 0.05161590327830737,
"grad_norm": 3.182907050050726,
"learning_rate": 2.5e-06,
"loss": 0.8324,
"step": 37
},
{
"epoch": 0.0530109276912346,
"grad_norm": 3.232818032867581,
"learning_rate": 2.5694444444444443e-06,
"loss": 0.7873,
"step": 38
},
{
"epoch": 0.054405952104161825,
"grad_norm": 3.2312814843076434,
"learning_rate": 2.6388888888888893e-06,
"loss": 0.8887,
"step": 39
},
{
"epoch": 0.05580097651708905,
"grad_norm": 3.356427492997714,
"learning_rate": 2.7083333333333334e-06,
"loss": 0.8288,
"step": 40
},
{
"epoch": 0.057196000930016276,
"grad_norm": 2.8004715481014286,
"learning_rate": 2.7777777777777783e-06,
"loss": 0.8025,
"step": 41
},
{
"epoch": 0.058591025342943505,
"grad_norm": 2.872064290513942,
"learning_rate": 2.8472222222222224e-06,
"loss": 0.7969,
"step": 42
},
{
"epoch": 0.05998604975587073,
"grad_norm": 2.604055696429484,
"learning_rate": 2.916666666666667e-06,
"loss": 0.7796,
"step": 43
},
{
"epoch": 0.061381074168797956,
"grad_norm": 2.588383255024964,
"learning_rate": 2.986111111111111e-06,
"loss": 0.8146,
"step": 44
},
{
"epoch": 0.06277609858172518,
"grad_norm": 2.957391150131729,
"learning_rate": 3.055555555555556e-06,
"loss": 0.8706,
"step": 45
},
{
"epoch": 0.06417112299465241,
"grad_norm": 2.7716857456581208,
"learning_rate": 3.125e-06,
"loss": 0.7553,
"step": 46
},
{
"epoch": 0.06556614740757963,
"grad_norm": 2.8704927650550305,
"learning_rate": 3.1944444444444443e-06,
"loss": 0.835,
"step": 47
},
{
"epoch": 0.06696117182050686,
"grad_norm": 2.926448629599372,
"learning_rate": 3.2638888888888892e-06,
"loss": 0.8358,
"step": 48
},
{
"epoch": 0.06835619623343409,
"grad_norm": 2.653881772418661,
"learning_rate": 3.3333333333333333e-06,
"loss": 0.8573,
"step": 49
},
{
"epoch": 0.06975122064636131,
"grad_norm": 2.7680654495558796,
"learning_rate": 3.4027777777777783e-06,
"loss": 0.8155,
"step": 50
},
{
"epoch": 0.07114624505928854,
"grad_norm": 2.690527707536472,
"learning_rate": 3.4722222222222224e-06,
"loss": 0.8718,
"step": 51
},
{
"epoch": 0.07254126947221576,
"grad_norm": 2.539274614289565,
"learning_rate": 3.5416666666666673e-06,
"loss": 0.8143,
"step": 52
},
{
"epoch": 0.07393629388514299,
"grad_norm": 2.6150798045040564,
"learning_rate": 3.6111111111111115e-06,
"loss": 0.8129,
"step": 53
},
{
"epoch": 0.07533131829807022,
"grad_norm": 2.77487818448589,
"learning_rate": 3.680555555555556e-06,
"loss": 0.8431,
"step": 54
},
{
"epoch": 0.07672634271099744,
"grad_norm": 2.4190012994996404,
"learning_rate": 3.7500000000000005e-06,
"loss": 0.7949,
"step": 55
},
{
"epoch": 0.07812136712392467,
"grad_norm": 2.5658926847545143,
"learning_rate": 3.819444444444444e-06,
"loss": 0.8349,
"step": 56
},
{
"epoch": 0.07951639153685189,
"grad_norm": 2.465553073665557,
"learning_rate": 3.88888888888889e-06,
"loss": 0.8001,
"step": 57
},
{
"epoch": 0.08091141594977912,
"grad_norm": 2.5769268931786113,
"learning_rate": 3.958333333333333e-06,
"loss": 0.7853,
"step": 58
},
{
"epoch": 0.08230644036270635,
"grad_norm": 2.762238617657559,
"learning_rate": 4.027777777777779e-06,
"loss": 0.8858,
"step": 59
},
{
"epoch": 0.08370146477563357,
"grad_norm": 2.459676043622874,
"learning_rate": 4.097222222222222e-06,
"loss": 0.7484,
"step": 60
},
{
"epoch": 0.0850964891885608,
"grad_norm": 2.281452061978854,
"learning_rate": 4.166666666666667e-06,
"loss": 0.7395,
"step": 61
},
{
"epoch": 0.08649151360148803,
"grad_norm": 2.5945093959933923,
"learning_rate": 4.236111111111111e-06,
"loss": 0.8383,
"step": 62
},
{
"epoch": 0.08788653801441525,
"grad_norm": 2.4907580232536337,
"learning_rate": 4.305555555555556e-06,
"loss": 0.7387,
"step": 63
},
{
"epoch": 0.08928156242734248,
"grad_norm": 2.5357156725837844,
"learning_rate": 4.3750000000000005e-06,
"loss": 0.8378,
"step": 64
},
{
"epoch": 0.0906765868402697,
"grad_norm": 2.3775000896329814,
"learning_rate": 4.444444444444444e-06,
"loss": 0.7606,
"step": 65
},
{
"epoch": 0.09207161125319693,
"grad_norm": 2.6462676515975367,
"learning_rate": 4.5138888888888895e-06,
"loss": 0.8067,
"step": 66
},
{
"epoch": 0.09346663566612416,
"grad_norm": 2.484589313471558,
"learning_rate": 4.583333333333333e-06,
"loss": 0.8517,
"step": 67
},
{
"epoch": 0.09486166007905138,
"grad_norm": 2.5630782397634135,
"learning_rate": 4.652777777777779e-06,
"loss": 0.8104,
"step": 68
},
{
"epoch": 0.0962566844919786,
"grad_norm": 2.603099792013807,
"learning_rate": 4.722222222222222e-06,
"loss": 0.7562,
"step": 69
},
{
"epoch": 0.09765170890490583,
"grad_norm": 2.458040997496572,
"learning_rate": 4.791666666666668e-06,
"loss": 0.7245,
"step": 70
},
{
"epoch": 0.09904673331783306,
"grad_norm": 2.6548832786314387,
"learning_rate": 4.861111111111111e-06,
"loss": 0.7604,
"step": 71
},
{
"epoch": 0.10044175773076029,
"grad_norm": 2.5407350711188137,
"learning_rate": 4.930555555555556e-06,
"loss": 0.7614,
"step": 72
},
{
"epoch": 0.10183678214368752,
"grad_norm": 2.466098001195273,
"learning_rate": 5e-06,
"loss": 0.7708,
"step": 73
},
{
"epoch": 0.10323180655661474,
"grad_norm": 2.4710838556933634,
"learning_rate": 5.069444444444445e-06,
"loss": 0.781,
"step": 74
},
{
"epoch": 0.10462683096954196,
"grad_norm": 2.5745603388615788,
"learning_rate": 5.138888888888889e-06,
"loss": 0.7617,
"step": 75
},
{
"epoch": 0.1060218553824692,
"grad_norm": 2.4760378463602617,
"learning_rate": 5.208333333333334e-06,
"loss": 0.6586,
"step": 76
},
{
"epoch": 0.10741687979539642,
"grad_norm": 2.1046185393371637,
"learning_rate": 5.2777777777777785e-06,
"loss": 0.6515,
"step": 77
},
{
"epoch": 0.10881190420832365,
"grad_norm": 2.57807326774727,
"learning_rate": 5.347222222222222e-06,
"loss": 0.7987,
"step": 78
},
{
"epoch": 0.11020692862125087,
"grad_norm": 2.4541665474964884,
"learning_rate": 5.416666666666667e-06,
"loss": 0.7797,
"step": 79
},
{
"epoch": 0.1116019530341781,
"grad_norm": 2.7192596934225883,
"learning_rate": 5.486111111111112e-06,
"loss": 0.8144,
"step": 80
},
{
"epoch": 0.11299697744710532,
"grad_norm": 2.587955726978526,
"learning_rate": 5.555555555555557e-06,
"loss": 0.7718,
"step": 81
},
{
"epoch": 0.11439200186003255,
"grad_norm": 2.8624670000905508,
"learning_rate": 5.625e-06,
"loss": 0.8642,
"step": 82
},
{
"epoch": 0.11578702627295978,
"grad_norm": 2.3985540385123785,
"learning_rate": 5.694444444444445e-06,
"loss": 0.7585,
"step": 83
},
{
"epoch": 0.11718205068588701,
"grad_norm": 2.4347959569209743,
"learning_rate": 5.7638888888888886e-06,
"loss": 0.7652,
"step": 84
},
{
"epoch": 0.11857707509881422,
"grad_norm": 2.472216781380437,
"learning_rate": 5.833333333333334e-06,
"loss": 0.7593,
"step": 85
},
{
"epoch": 0.11997209951174145,
"grad_norm": 2.3683241185527684,
"learning_rate": 5.9027777777777785e-06,
"loss": 0.6797,
"step": 86
},
{
"epoch": 0.12136712392466868,
"grad_norm": 2.585048523140377,
"learning_rate": 5.972222222222222e-06,
"loss": 0.8135,
"step": 87
},
{
"epoch": 0.12276214833759591,
"grad_norm": 2.583706669085246,
"learning_rate": 6.041666666666667e-06,
"loss": 0.8272,
"step": 88
},
{
"epoch": 0.12415717275052314,
"grad_norm": 2.4831147364219253,
"learning_rate": 6.111111111111112e-06,
"loss": 0.8588,
"step": 89
},
{
"epoch": 0.12555219716345037,
"grad_norm": 2.372052290928504,
"learning_rate": 6.180555555555557e-06,
"loss": 0.7559,
"step": 90
},
{
"epoch": 0.12694722157637758,
"grad_norm": 2.3693742709173997,
"learning_rate": 6.25e-06,
"loss": 0.7523,
"step": 91
},
{
"epoch": 0.12834224598930483,
"grad_norm": 2.56593824677806,
"learning_rate": 6.319444444444445e-06,
"loss": 0.7976,
"step": 92
},
{
"epoch": 0.12973727040223204,
"grad_norm": 2.6643468086221485,
"learning_rate": 6.3888888888888885e-06,
"loss": 0.7603,
"step": 93
},
{
"epoch": 0.13113229481515926,
"grad_norm": 2.431324445820708,
"learning_rate": 6.458333333333334e-06,
"loss": 0.7177,
"step": 94
},
{
"epoch": 0.1325273192280865,
"grad_norm": 2.0286167424640786,
"learning_rate": 6.5277777777777784e-06,
"loss": 0.6637,
"step": 95
},
{
"epoch": 0.1339223436410137,
"grad_norm": 2.381250278728561,
"learning_rate": 6.597222222222223e-06,
"loss": 0.7543,
"step": 96
},
{
"epoch": 0.13531736805394096,
"grad_norm": 2.5224874846398304,
"learning_rate": 6.666666666666667e-06,
"loss": 0.748,
"step": 97
},
{
"epoch": 0.13671239246686817,
"grad_norm": 2.367382093641696,
"learning_rate": 6.736111111111112e-06,
"loss": 0.7078,
"step": 98
},
{
"epoch": 0.13810741687979539,
"grad_norm": 2.2891454288500785,
"learning_rate": 6.8055555555555566e-06,
"loss": 0.7778,
"step": 99
},
{
"epoch": 0.13950244129272263,
"grad_norm": 2.326427675547096,
"learning_rate": 6.875e-06,
"loss": 0.7,
"step": 100
},
{
"epoch": 0.14089746570564984,
"grad_norm": 2.43784471948663,
"learning_rate": 6.944444444444445e-06,
"loss": 0.7845,
"step": 101
},
{
"epoch": 0.1422924901185771,
"grad_norm": 2.434928494698733,
"learning_rate": 7.013888888888889e-06,
"loss": 0.7741,
"step": 102
},
{
"epoch": 0.1436875145315043,
"grad_norm": 2.544677188997524,
"learning_rate": 7.083333333333335e-06,
"loss": 0.7321,
"step": 103
},
{
"epoch": 0.14508253894443152,
"grad_norm": 2.4317471320917314,
"learning_rate": 7.152777777777778e-06,
"loss": 0.7301,
"step": 104
},
{
"epoch": 0.14647756335735876,
"grad_norm": 2.3758040864865535,
"learning_rate": 7.222222222222223e-06,
"loss": 0.7168,
"step": 105
},
{
"epoch": 0.14787258777028597,
"grad_norm": 2.427323974558636,
"learning_rate": 7.291666666666667e-06,
"loss": 0.7122,
"step": 106
},
{
"epoch": 0.14926761218321322,
"grad_norm": 2.5718099527792515,
"learning_rate": 7.361111111111112e-06,
"loss": 0.7777,
"step": 107
},
{
"epoch": 0.15066263659614043,
"grad_norm": 2.3984442193810303,
"learning_rate": 7.4305555555555565e-06,
"loss": 0.76,
"step": 108
},
{
"epoch": 0.15205766100906765,
"grad_norm": 2.809886900720292,
"learning_rate": 7.500000000000001e-06,
"loss": 0.9409,
"step": 109
},
{
"epoch": 0.1534526854219949,
"grad_norm": 2.1909629178912735,
"learning_rate": 7.569444444444445e-06,
"loss": 0.6928,
"step": 110
},
{
"epoch": 0.1548477098349221,
"grad_norm": 2.299999050133801,
"learning_rate": 7.638888888888888e-06,
"loss": 0.7331,
"step": 111
},
{
"epoch": 0.15624273424784935,
"grad_norm": 2.4657027253176658,
"learning_rate": 7.708333333333334e-06,
"loss": 0.7734,
"step": 112
},
{
"epoch": 0.15763775866077656,
"grad_norm": 2.499003482696436,
"learning_rate": 7.77777777777778e-06,
"loss": 0.7568,
"step": 113
},
{
"epoch": 0.15903278307370378,
"grad_norm": 2.556353786561783,
"learning_rate": 7.847222222222223e-06,
"loss": 0.7813,
"step": 114
},
{
"epoch": 0.16042780748663102,
"grad_norm": 2.398942685028162,
"learning_rate": 7.916666666666667e-06,
"loss": 0.773,
"step": 115
},
{
"epoch": 0.16182283189955823,
"grad_norm": 2.437022751579472,
"learning_rate": 7.986111111111112e-06,
"loss": 0.7423,
"step": 116
},
{
"epoch": 0.16321785631248548,
"grad_norm": 2.5361969141214753,
"learning_rate": 8.055555555555557e-06,
"loss": 0.8085,
"step": 117
},
{
"epoch": 0.1646128807254127,
"grad_norm": 2.5686473875754148,
"learning_rate": 8.125000000000001e-06,
"loss": 0.7602,
"step": 118
},
{
"epoch": 0.16600790513833993,
"grad_norm": 2.6355454558922786,
"learning_rate": 8.194444444444445e-06,
"loss": 0.7348,
"step": 119
},
{
"epoch": 0.16740292955126715,
"grad_norm": 2.5297512188295457,
"learning_rate": 8.263888888888888e-06,
"loss": 0.7398,
"step": 120
},
{
"epoch": 0.16879795396419436,
"grad_norm": 2.4649882854746354,
"learning_rate": 8.333333333333334e-06,
"loss": 0.7636,
"step": 121
},
{
"epoch": 0.1701929783771216,
"grad_norm": 2.4209538379807443,
"learning_rate": 8.402777777777779e-06,
"loss": 0.7799,
"step": 122
},
{
"epoch": 0.17158800279004882,
"grad_norm": 2.7076603164565665,
"learning_rate": 8.472222222222223e-06,
"loss": 0.8096,
"step": 123
},
{
"epoch": 0.17298302720297606,
"grad_norm": 2.5198462673809643,
"learning_rate": 8.541666666666666e-06,
"loss": 0.7566,
"step": 124
},
{
"epoch": 0.17437805161590328,
"grad_norm": 2.3301540316911935,
"learning_rate": 8.611111111111112e-06,
"loss": 0.7885,
"step": 125
},
{
"epoch": 0.1757730760288305,
"grad_norm": 2.528981320323845,
"learning_rate": 8.680555555555557e-06,
"loss": 0.7747,
"step": 126
},
{
"epoch": 0.17716810044175774,
"grad_norm": 2.6228409614962938,
"learning_rate": 8.750000000000001e-06,
"loss": 0.7827,
"step": 127
},
{
"epoch": 0.17856312485468495,
"grad_norm": 2.3530092565699805,
"learning_rate": 8.819444444444445e-06,
"loss": 0.775,
"step": 128
},
{
"epoch": 0.1799581492676122,
"grad_norm": 2.4030846577696834,
"learning_rate": 8.888888888888888e-06,
"loss": 0.8468,
"step": 129
},
{
"epoch": 0.1813531736805394,
"grad_norm": 2.397665315329076,
"learning_rate": 8.958333333333334e-06,
"loss": 0.7555,
"step": 130
},
{
"epoch": 0.18274819809346662,
"grad_norm": 2.4944187750790165,
"learning_rate": 9.027777777777779e-06,
"loss": 0.6983,
"step": 131
},
{
"epoch": 0.18414322250639387,
"grad_norm": 2.4633848637523266,
"learning_rate": 9.097222222222223e-06,
"loss": 0.8095,
"step": 132
},
{
"epoch": 0.18553824691932108,
"grad_norm": 2.4514234526609755,
"learning_rate": 9.166666666666666e-06,
"loss": 0.7669,
"step": 133
},
{
"epoch": 0.18693327133224832,
"grad_norm": 2.2348180975167002,
"learning_rate": 9.236111111111112e-06,
"loss": 0.7445,
"step": 134
},
{
"epoch": 0.18832829574517554,
"grad_norm": 2.7654592189288127,
"learning_rate": 9.305555555555557e-06,
"loss": 0.8222,
"step": 135
},
{
"epoch": 0.18972332015810275,
"grad_norm": 2.657505427715525,
"learning_rate": 9.375000000000001e-06,
"loss": 0.7993,
"step": 136
},
{
"epoch": 0.19111834457103,
"grad_norm": 2.4757416453669037,
"learning_rate": 9.444444444444445e-06,
"loss": 0.7553,
"step": 137
},
{
"epoch": 0.1925133689839572,
"grad_norm": 2.4710024870459546,
"learning_rate": 9.51388888888889e-06,
"loss": 0.7893,
"step": 138
},
{
"epoch": 0.19390839339688445,
"grad_norm": 2.484340628816311,
"learning_rate": 9.583333333333335e-06,
"loss": 0.7435,
"step": 139
},
{
"epoch": 0.19530341780981167,
"grad_norm": 2.3567750021925864,
"learning_rate": 9.652777777777779e-06,
"loss": 0.7991,
"step": 140
},
{
"epoch": 0.1966984422227389,
"grad_norm": 2.6403706679259376,
"learning_rate": 9.722222222222223e-06,
"loss": 0.8204,
"step": 141
},
{
"epoch": 0.19809346663566613,
"grad_norm": 2.7772690056581406,
"learning_rate": 9.791666666666666e-06,
"loss": 0.764,
"step": 142
},
{
"epoch": 0.19948849104859334,
"grad_norm": 2.38948338269402,
"learning_rate": 9.861111111111112e-06,
"loss": 0.7588,
"step": 143
},
{
"epoch": 0.20088351546152058,
"grad_norm": 2.364591011786826,
"learning_rate": 9.930555555555557e-06,
"loss": 0.7382,
"step": 144
},
{
"epoch": 0.2022785398744478,
"grad_norm": 2.39460109071618,
"learning_rate": 1e-05,
"loss": 0.8297,
"step": 145
},
{
"epoch": 0.20367356428737504,
"grad_norm": 2.385277772966096,
"learning_rate": 9.99998512668374e-06,
"loss": 0.762,
"step": 146
},
{
"epoch": 0.20506858870030226,
"grad_norm": 2.550269196061235,
"learning_rate": 9.999940506823446e-06,
"loss": 0.8577,
"step": 147
},
{
"epoch": 0.20646361311322947,
"grad_norm": 2.3828419813394768,
"learning_rate": 9.999866140684576e-06,
"loss": 0.7526,
"step": 148
},
{
"epoch": 0.20785863752615671,
"grad_norm": 2.674659581827359,
"learning_rate": 9.999762028709558e-06,
"loss": 0.8444,
"step": 149
},
{
"epoch": 0.20925366193908393,
"grad_norm": 2.438042726374861,
"learning_rate": 9.999628171517788e-06,
"loss": 0.7834,
"step": 150
},
{
"epoch": 0.21064868635201117,
"grad_norm": 2.527684334680155,
"learning_rate": 9.999464569905628e-06,
"loss": 0.778,
"step": 151
},
{
"epoch": 0.2120437107649384,
"grad_norm": 2.164235216275893,
"learning_rate": 9.999271224846397e-06,
"loss": 0.7357,
"step": 152
},
{
"epoch": 0.2134387351778656,
"grad_norm": 2.568663637972533,
"learning_rate": 9.999048137490364e-06,
"loss": 0.7864,
"step": 153
},
{
"epoch": 0.21483375959079284,
"grad_norm": 2.328376334217868,
"learning_rate": 9.998795309164754e-06,
"loss": 0.741,
"step": 154
},
{
"epoch": 0.21622878400372006,
"grad_norm": 2.3440798093116815,
"learning_rate": 9.99851274137372e-06,
"loss": 0.752,
"step": 155
},
{
"epoch": 0.2176238084166473,
"grad_norm": 2.3252703261044587,
"learning_rate": 9.998200435798355e-06,
"loss": 0.7822,
"step": 156
},
{
"epoch": 0.21901883282957452,
"grad_norm": 2.1066008364671944,
"learning_rate": 9.997858394296666e-06,
"loss": 0.6879,
"step": 157
},
{
"epoch": 0.22041385724250173,
"grad_norm": 2.4582290754964067,
"learning_rate": 9.997486618903566e-06,
"loss": 0.7594,
"step": 158
},
{
"epoch": 0.22180888165542897,
"grad_norm": 2.202899892319989,
"learning_rate": 9.99708511183087e-06,
"loss": 0.7461,
"step": 159
},
{
"epoch": 0.2232039060683562,
"grad_norm": 2.367202675165834,
"learning_rate": 9.996653875467278e-06,
"loss": 0.7755,
"step": 160
},
{
"epoch": 0.22459893048128343,
"grad_norm": 2.4283257637167743,
"learning_rate": 9.99619291237835e-06,
"loss": 0.7694,
"step": 161
},
{
"epoch": 0.22599395489421065,
"grad_norm": 2.393310544522537,
"learning_rate": 9.99570222530651e-06,
"loss": 0.7624,
"step": 162
},
{
"epoch": 0.22738897930713786,
"grad_norm": 2.456620826384989,
"learning_rate": 9.995181817171016e-06,
"loss": 0.7943,
"step": 163
},
{
"epoch": 0.2287840037200651,
"grad_norm": 2.33795920230864,
"learning_rate": 9.994631691067941e-06,
"loss": 0.7064,
"step": 164
},
{
"epoch": 0.23017902813299232,
"grad_norm": 2.380769585125721,
"learning_rate": 9.994051850270172e-06,
"loss": 0.7785,
"step": 165
},
{
"epoch": 0.23157405254591956,
"grad_norm": 2.0565750954623954,
"learning_rate": 9.993442298227365e-06,
"loss": 0.6899,
"step": 166
},
{
"epoch": 0.23296907695884678,
"grad_norm": 2.4629114081369154,
"learning_rate": 9.992803038565947e-06,
"loss": 0.7318,
"step": 167
},
{
"epoch": 0.23436410137177402,
"grad_norm": 2.5260371511690227,
"learning_rate": 9.992134075089085e-06,
"loss": 0.7492,
"step": 168
},
{
"epoch": 0.23575912578470123,
"grad_norm": 2.2870342101598573,
"learning_rate": 9.991435411776654e-06,
"loss": 0.7546,
"step": 169
},
{
"epoch": 0.23715415019762845,
"grad_norm": 2.2176561807734463,
"learning_rate": 9.990707052785236e-06,
"loss": 0.6861,
"step": 170
},
{
"epoch": 0.2385491746105557,
"grad_norm": 2.1930980336038015,
"learning_rate": 9.989949002448076e-06,
"loss": 0.7527,
"step": 171
},
{
"epoch": 0.2399441990234829,
"grad_norm": 2.3770728495629654,
"learning_rate": 9.98916126527506e-06,
"loss": 0.7358,
"step": 172
},
{
"epoch": 0.24133922343641015,
"grad_norm": 2.503959470995855,
"learning_rate": 9.988343845952697e-06,
"loss": 0.7925,
"step": 173
},
{
"epoch": 0.24273424784933736,
"grad_norm": 2.295115004392427,
"learning_rate": 9.98749674934408e-06,
"loss": 0.7493,
"step": 174
},
{
"epoch": 0.24412927226226458,
"grad_norm": 2.3785544817722912,
"learning_rate": 9.98661998048886e-06,
"loss": 0.784,
"step": 175
},
{
"epoch": 0.24552429667519182,
"grad_norm": 2.460048362058096,
"learning_rate": 9.985713544603229e-06,
"loss": 0.7982,
"step": 176
},
{
"epoch": 0.24691932108811904,
"grad_norm": 2.5187516427864503,
"learning_rate": 9.984777447079861e-06,
"loss": 0.7899,
"step": 177
},
{
"epoch": 0.24831434550104628,
"grad_norm": 2.4375799853969387,
"learning_rate": 9.983811693487913e-06,
"loss": 0.8014,
"step": 178
},
{
"epoch": 0.2497093699139735,
"grad_norm": 2.262358785483013,
"learning_rate": 9.982816289572966e-06,
"loss": 0.7308,
"step": 179
},
{
"epoch": 0.25110439432690074,
"grad_norm": 2.4027476974294504,
"learning_rate": 9.981791241257001e-06,
"loss": 0.7808,
"step": 180
},
{
"epoch": 0.25249941873982795,
"grad_norm": 2.3785122102380387,
"learning_rate": 9.980736554638367e-06,
"loss": 0.7762,
"step": 181
},
{
"epoch": 0.25389444315275517,
"grad_norm": 2.2101376692736086,
"learning_rate": 9.979652235991738e-06,
"loss": 0.7199,
"step": 182
},
{
"epoch": 0.2552894675656824,
"grad_norm": 2.3853547321892594,
"learning_rate": 9.978538291768083e-06,
"loss": 0.7662,
"step": 183
},
{
"epoch": 0.25668449197860965,
"grad_norm": 2.4629393864421103,
"learning_rate": 9.977394728594614e-06,
"loss": 0.7372,
"step": 184
},
{
"epoch": 0.25807951639153687,
"grad_norm": 2.3117069679784077,
"learning_rate": 9.976221553274767e-06,
"loss": 0.7159,
"step": 185
},
{
"epoch": 0.2594745408044641,
"grad_norm": 2.061689977353346,
"learning_rate": 9.975018772788144e-06,
"loss": 0.6528,
"step": 186
},
{
"epoch": 0.2608695652173913,
"grad_norm": 2.276599208142271,
"learning_rate": 9.973786394290475e-06,
"loss": 0.8088,
"step": 187
},
{
"epoch": 0.2622645896303185,
"grad_norm": 2.3657045018285308,
"learning_rate": 9.972524425113585e-06,
"loss": 0.7563,
"step": 188
},
{
"epoch": 0.2636596140432458,
"grad_norm": 2.1739723205164907,
"learning_rate": 9.971232872765344e-06,
"loss": 0.717,
"step": 189
},
{
"epoch": 0.265054638456173,
"grad_norm": 2.2224227720319907,
"learning_rate": 9.969911744929613e-06,
"loss": 0.7291,
"step": 190
},
{
"epoch": 0.2664496628691002,
"grad_norm": 2.312701932028756,
"learning_rate": 9.968561049466214e-06,
"loss": 0.7788,
"step": 191
},
{
"epoch": 0.2678446872820274,
"grad_norm": 2.2886619359808362,
"learning_rate": 9.967180794410878e-06,
"loss": 0.7303,
"step": 192
},
{
"epoch": 0.26923971169495464,
"grad_norm": 2.2800673609129887,
"learning_rate": 9.96577098797519e-06,
"loss": 0.7756,
"step": 193
},
{
"epoch": 0.2706347361078819,
"grad_norm": 2.3599405553425474,
"learning_rate": 9.96433163854655e-06,
"loss": 0.7285,
"step": 194
},
{
"epoch": 0.2720297605208091,
"grad_norm": 2.3035254395464335,
"learning_rate": 9.962862754688118e-06,
"loss": 0.7317,
"step": 195
},
{
"epoch": 0.27342478493373634,
"grad_norm": 2.1983487142998337,
"learning_rate": 9.961364345138764e-06,
"loss": 0.8082,
"step": 196
},
{
"epoch": 0.27481980934666356,
"grad_norm": 2.1891940591475256,
"learning_rate": 9.959836418813016e-06,
"loss": 0.7296,
"step": 197
},
{
"epoch": 0.27621483375959077,
"grad_norm": 2.3364824298619107,
"learning_rate": 9.958278984801003e-06,
"loss": 0.735,
"step": 198
},
{
"epoch": 0.27760985817251804,
"grad_norm": 2.2975941856825917,
"learning_rate": 9.956692052368416e-06,
"loss": 0.7233,
"step": 199
},
{
"epoch": 0.27900488258544526,
"grad_norm": 2.3372101943251726,
"learning_rate": 9.955075630956425e-06,
"loss": 0.7404,
"step": 200
},
{
"epoch": 0.2803999069983725,
"grad_norm": 2.1853548847254145,
"learning_rate": 9.953429730181653e-06,
"loss": 0.7436,
"step": 201
},
{
"epoch": 0.2817949314112997,
"grad_norm": 2.224468630946785,
"learning_rate": 9.951754359836104e-06,
"loss": 0.7623,
"step": 202
},
{
"epoch": 0.2831899558242269,
"grad_norm": 2.2750380642685286,
"learning_rate": 9.950049529887098e-06,
"loss": 0.7888,
"step": 203
},
{
"epoch": 0.2845849802371542,
"grad_norm": 2.297526158131741,
"learning_rate": 9.94831525047723e-06,
"loss": 0.7366,
"step": 204
},
{
"epoch": 0.2859800046500814,
"grad_norm": 2.3677703079300203,
"learning_rate": 9.94655153192429e-06,
"loss": 0.7912,
"step": 205
},
{
"epoch": 0.2873750290630086,
"grad_norm": 2.539938055319808,
"learning_rate": 9.944758384721218e-06,
"loss": 0.7649,
"step": 206
},
{
"epoch": 0.2887700534759358,
"grad_norm": 2.1358951407571625,
"learning_rate": 9.94293581953603e-06,
"loss": 0.7197,
"step": 207
},
{
"epoch": 0.29016507788886303,
"grad_norm": 2.3752369254146948,
"learning_rate": 9.941083847211765e-06,
"loss": 0.7811,
"step": 208
},
{
"epoch": 0.2915601023017903,
"grad_norm": 2.2366003263543974,
"learning_rate": 9.939202478766408e-06,
"loss": 0.7838,
"step": 209
},
{
"epoch": 0.2929551267147175,
"grad_norm": 2.353138945767122,
"learning_rate": 9.937291725392835e-06,
"loss": 0.791,
"step": 210
},
{
"epoch": 0.29435015112764473,
"grad_norm": 2.26497443267959,
"learning_rate": 9.935351598458743e-06,
"loss": 0.7473,
"step": 211
},
{
"epoch": 0.29574517554057195,
"grad_norm": 2.517368832160316,
"learning_rate": 9.933382109506577e-06,
"loss": 0.7834,
"step": 212
},
{
"epoch": 0.29714019995349916,
"grad_norm": 2.3571681636390194,
"learning_rate": 9.931383270253475e-06,
"loss": 0.759,
"step": 213
},
{
"epoch": 0.29853522436642643,
"grad_norm": 2.2330737534866283,
"learning_rate": 9.92935509259118e-06,
"loss": 0.7235,
"step": 214
},
{
"epoch": 0.29993024877935365,
"grad_norm": 2.348646725591813,
"learning_rate": 9.927297588585984e-06,
"loss": 0.7873,
"step": 215
},
{
"epoch": 0.30132527319228086,
"grad_norm": 2.2883655282601665,
"learning_rate": 9.925210770478653e-06,
"loss": 0.7567,
"step": 216
},
{
"epoch": 0.3027202976052081,
"grad_norm": 2.365988328655639,
"learning_rate": 9.923094650684346e-06,
"loss": 0.8062,
"step": 217
},
{
"epoch": 0.3041153220181353,
"grad_norm": 2.1815309873800106,
"learning_rate": 9.92094924179255e-06,
"loss": 0.6704,
"step": 218
},
{
"epoch": 0.30551034643106256,
"grad_norm": 2.2684610053286165,
"learning_rate": 9.918774556567007e-06,
"loss": 0.7228,
"step": 219
},
{
"epoch": 0.3069053708439898,
"grad_norm": 2.1987712868445985,
"learning_rate": 9.916570607945627e-06,
"loss": 0.7303,
"step": 220
},
{
"epoch": 0.308300395256917,
"grad_norm": 2.1467480973519466,
"learning_rate": 9.914337409040418e-06,
"loss": 0.7773,
"step": 221
},
{
"epoch": 0.3096954196698442,
"grad_norm": 2.1981875970164015,
"learning_rate": 9.912074973137413e-06,
"loss": 0.7558,
"step": 222
},
{
"epoch": 0.3110904440827714,
"grad_norm": 2.370866601678729,
"learning_rate": 9.90978331369658e-06,
"loss": 0.8031,
"step": 223
},
{
"epoch": 0.3124854684956987,
"grad_norm": 2.0958671373974656,
"learning_rate": 9.90746244435175e-06,
"loss": 0.7119,
"step": 224
},
{
"epoch": 0.3138804929086259,
"grad_norm": 2.276250356937496,
"learning_rate": 9.905112378910532e-06,
"loss": 0.7132,
"step": 225
},
{
"epoch": 0.3152755173215531,
"grad_norm": 2.349183832747072,
"learning_rate": 9.902733131354232e-06,
"loss": 0.7279,
"step": 226
},
{
"epoch": 0.31667054173448034,
"grad_norm": 2.2645947401920865,
"learning_rate": 9.900324715837772e-06,
"loss": 0.772,
"step": 227
},
{
"epoch": 0.31806556614740755,
"grad_norm": 2.420057647782548,
"learning_rate": 9.897887146689603e-06,
"loss": 0.7967,
"step": 228
},
{
"epoch": 0.3194605905603348,
"grad_norm": 2.09035311023506,
"learning_rate": 9.895420438411616e-06,
"loss": 0.7409,
"step": 229
},
{
"epoch": 0.32085561497326204,
"grad_norm": 2.128576564240619,
"learning_rate": 9.892924605679066e-06,
"loss": 0.7108,
"step": 230
},
{
"epoch": 0.32225063938618925,
"grad_norm": 2.3557840243006303,
"learning_rate": 9.890399663340478e-06,
"loss": 0.7989,
"step": 231
},
{
"epoch": 0.32364566379911647,
"grad_norm": 2.262147642657486,
"learning_rate": 9.887845626417556e-06,
"loss": 0.7881,
"step": 232
},
{
"epoch": 0.32504068821204374,
"grad_norm": 2.308747699785618,
"learning_rate": 9.885262510105102e-06,
"loss": 0.7044,
"step": 233
},
{
"epoch": 0.32643571262497095,
"grad_norm": 2.1776401707207924,
"learning_rate": 9.882650329770919e-06,
"loss": 0.789,
"step": 234
},
{
"epoch": 0.32783073703789817,
"grad_norm": 2.1837014708981872,
"learning_rate": 9.880009100955716e-06,
"loss": 0.7991,
"step": 235
},
{
"epoch": 0.3292257614508254,
"grad_norm": 2.344957858708995,
"learning_rate": 9.877338839373032e-06,
"loss": 0.8075,
"step": 236
},
{
"epoch": 0.3306207858637526,
"grad_norm": 2.3077392669789565,
"learning_rate": 9.874639560909118e-06,
"loss": 0.7442,
"step": 237
},
{
"epoch": 0.33201581027667987,
"grad_norm": 2.231581492055234,
"learning_rate": 9.871911281622869e-06,
"loss": 0.7426,
"step": 238
},
{
"epoch": 0.3334108346896071,
"grad_norm": 2.326813603280547,
"learning_rate": 9.869154017745706e-06,
"loss": 0.7747,
"step": 239
},
{
"epoch": 0.3348058591025343,
"grad_norm": 2.086609390067158,
"learning_rate": 9.866367785681493e-06,
"loss": 0.7518,
"step": 240
},
{
"epoch": 0.3362008835154615,
"grad_norm": 2.363099222001944,
"learning_rate": 9.863552602006435e-06,
"loss": 0.7724,
"step": 241
},
{
"epoch": 0.3375959079283887,
"grad_norm": 2.077033177420949,
"learning_rate": 9.860708483468975e-06,
"loss": 0.7175,
"step": 242
},
{
"epoch": 0.338990932341316,
"grad_norm": 2.246936114717258,
"learning_rate": 9.857835446989708e-06,
"loss": 0.7709,
"step": 243
},
{
"epoch": 0.3403859567542432,
"grad_norm": 2.1831810988074145,
"learning_rate": 9.854933509661264e-06,
"loss": 0.7632,
"step": 244
},
{
"epoch": 0.34178098116717043,
"grad_norm": 2.452152625128349,
"learning_rate": 9.852002688748214e-06,
"loss": 0.7713,
"step": 245
},
{
"epoch": 0.34317600558009764,
"grad_norm": 2.101403660363672,
"learning_rate": 9.84904300168697e-06,
"loss": 0.7556,
"step": 246
},
{
"epoch": 0.34457102999302486,
"grad_norm": 2.196554440781837,
"learning_rate": 9.84605446608568e-06,
"loss": 0.7352,
"step": 247
},
{
"epoch": 0.34596605440595213,
"grad_norm": 2.1581190947183906,
"learning_rate": 9.843037099724111e-06,
"loss": 0.7505,
"step": 248
},
{
"epoch": 0.34736107881887934,
"grad_norm": 2.4174065292937503,
"learning_rate": 9.839990920553566e-06,
"loss": 0.7461,
"step": 249
},
{
"epoch": 0.34875610323180656,
"grad_norm": 2.2765578992139663,
"learning_rate": 9.83691594669676e-06,
"loss": 0.7607,
"step": 250
},
{
"epoch": 0.3501511276447338,
"grad_norm": 2.2025411404466735,
"learning_rate": 9.833812196447712e-06,
"loss": 0.7696,
"step": 251
},
{
"epoch": 0.351546152057661,
"grad_norm": 2.316141072305499,
"learning_rate": 9.830679688271647e-06,
"loss": 0.7743,
"step": 252
},
{
"epoch": 0.35294117647058826,
"grad_norm": 2.2989681090795266,
"learning_rate": 9.827518440804882e-06,
"loss": 0.8095,
"step": 253
},
{
"epoch": 0.3543362008835155,
"grad_norm": 2.4056158179853884,
"learning_rate": 9.824328472854706e-06,
"loss": 0.7212,
"step": 254
},
{
"epoch": 0.3557312252964427,
"grad_norm": 2.2264588040031716,
"learning_rate": 9.821109803399283e-06,
"loss": 0.7479,
"step": 255
},
{
"epoch": 0.3571262497093699,
"grad_norm": 2.1513171263920223,
"learning_rate": 9.817862451587527e-06,
"loss": 0.6563,
"step": 256
},
{
"epoch": 0.3585212741222971,
"grad_norm": 2.231437998260098,
"learning_rate": 9.814586436738998e-06,
"loss": 0.6965,
"step": 257
},
{
"epoch": 0.3599162985352244,
"grad_norm": 2.164610078417342,
"learning_rate": 9.811281778343773e-06,
"loss": 0.7445,
"step": 258
},
{
"epoch": 0.3613113229481516,
"grad_norm": 2.532376368559481,
"learning_rate": 9.807948496062344e-06,
"loss": 0.7725,
"step": 259
},
{
"epoch": 0.3627063473610788,
"grad_norm": 2.339159697690894,
"learning_rate": 9.804586609725499e-06,
"loss": 0.7535,
"step": 260
},
{
"epoch": 0.36410137177400603,
"grad_norm": 2.166756913084875,
"learning_rate": 9.801196139334195e-06,
"loss": 0.7147,
"step": 261
},
{
"epoch": 0.36549639618693325,
"grad_norm": 2.139276056908412,
"learning_rate": 9.797777105059448e-06,
"loss": 0.7404,
"step": 262
},
{
"epoch": 0.3668914205998605,
"grad_norm": 2.4733715470175284,
"learning_rate": 9.794329527242209e-06,
"loss": 0.8087,
"step": 263
},
{
"epoch": 0.36828644501278773,
"grad_norm": 2.0018490521631866,
"learning_rate": 9.790853426393246e-06,
"loss": 0.6991,
"step": 264
},
{
"epoch": 0.36968146942571495,
"grad_norm": 2.1908532559835048,
"learning_rate": 9.787348823193013e-06,
"loss": 0.7728,
"step": 265
},
{
"epoch": 0.37107649383864216,
"grad_norm": 2.5151311136851264,
"learning_rate": 9.783815738491544e-06,
"loss": 0.7751,
"step": 266
},
{
"epoch": 0.3724715182515694,
"grad_norm": 2.2480235747495327,
"learning_rate": 9.780254193308312e-06,
"loss": 0.6931,
"step": 267
},
{
"epoch": 0.37386654266449665,
"grad_norm": 2.4308792768188194,
"learning_rate": 9.776664208832111e-06,
"loss": 0.729,
"step": 268
},
{
"epoch": 0.37526156707742386,
"grad_norm": 2.148993722755254,
"learning_rate": 9.77304580642093e-06,
"loss": 0.7566,
"step": 269
},
{
"epoch": 0.3766565914903511,
"grad_norm": 2.2131611670886113,
"learning_rate": 9.76939900760183e-06,
"loss": 0.692,
"step": 270
},
{
"epoch": 0.3780516159032783,
"grad_norm": 2.284885600630375,
"learning_rate": 9.765723834070805e-06,
"loss": 0.7334,
"step": 271
},
{
"epoch": 0.3794466403162055,
"grad_norm": 2.1803327517414637,
"learning_rate": 9.762020307692662e-06,
"loss": 0.7022,
"step": 272
},
{
"epoch": 0.3808416647291328,
"grad_norm": 2.0737093304589997,
"learning_rate": 9.75828845050089e-06,
"loss": 0.7178,
"step": 273
},
{
"epoch": 0.38223668914206,
"grad_norm": 2.1969087943561023,
"learning_rate": 9.754528284697526e-06,
"loss": 0.8056,
"step": 274
},
{
"epoch": 0.3836317135549872,
"grad_norm": 2.0470925124450723,
"learning_rate": 9.750739832653023e-06,
"loss": 0.7027,
"step": 275
},
{
"epoch": 0.3850267379679144,
"grad_norm": 2.206936141731334,
"learning_rate": 9.74692311690612e-06,
"loss": 0.7628,
"step": 276
},
{
"epoch": 0.38642176238084164,
"grad_norm": 2.259665231259676,
"learning_rate": 9.743078160163703e-06,
"loss": 0.8171,
"step": 277
},
{
"epoch": 0.3878167867937689,
"grad_norm": 2.069519176675941,
"learning_rate": 9.73920498530068e-06,
"loss": 0.7066,
"step": 278
},
{
"epoch": 0.3892118112066961,
"grad_norm": 2.18198802420733,
"learning_rate": 9.735303615359827e-06,
"loss": 0.7137,
"step": 279
},
{
"epoch": 0.39060683561962334,
"grad_norm": 2.065628530712166,
"learning_rate": 9.731374073551671e-06,
"loss": 0.7051,
"step": 280
},
{
"epoch": 0.39200186003255055,
"grad_norm": 2.011089691307504,
"learning_rate": 9.72741638325434e-06,
"loss": 0.7244,
"step": 281
},
{
"epoch": 0.3933968844454778,
"grad_norm": 2.1660029235899727,
"learning_rate": 9.723430568013423e-06,
"loss": 0.7099,
"step": 282
},
{
"epoch": 0.39479190885840504,
"grad_norm": 2.3105910990971426,
"learning_rate": 9.719416651541839e-06,
"loss": 0.793,
"step": 283
},
{
"epoch": 0.39618693327133225,
"grad_norm": 2.2318565333793923,
"learning_rate": 9.715374657719685e-06,
"loss": 0.7452,
"step": 284
},
{
"epoch": 0.39758195768425947,
"grad_norm": 2.2160005494950563,
"learning_rate": 9.711304610594104e-06,
"loss": 0.7714,
"step": 285
},
{
"epoch": 0.3989769820971867,
"grad_norm": 2.1886833487875528,
"learning_rate": 9.707206534379133e-06,
"loss": 0.7291,
"step": 286
},
{
"epoch": 0.40037200651011395,
"grad_norm": 2.243344953149145,
"learning_rate": 9.703080453455569e-06,
"loss": 0.7543,
"step": 287
},
{
"epoch": 0.40176703092304117,
"grad_norm": 2.0905937282876588,
"learning_rate": 9.69892639237081e-06,
"loss": 0.6654,
"step": 288
},
{
"epoch": 0.4031620553359684,
"grad_norm": 1.8899757251612446,
"learning_rate": 9.694744375838725e-06,
"loss": 0.6204,
"step": 289
},
{
"epoch": 0.4045570797488956,
"grad_norm": 2.2344840253872045,
"learning_rate": 9.690534428739497e-06,
"loss": 0.7321,
"step": 290
},
{
"epoch": 0.4059521041618228,
"grad_norm": 2.212703648608523,
"learning_rate": 9.68629657611947e-06,
"loss": 0.7184,
"step": 291
},
{
"epoch": 0.4073471285747501,
"grad_norm": 2.24263720075315,
"learning_rate": 9.682030843191021e-06,
"loss": 0.7453,
"step": 292
},
{
"epoch": 0.4087421529876773,
"grad_norm": 2.37707226734299,
"learning_rate": 9.677737255332381e-06,
"loss": 0.7764,
"step": 293
},
{
"epoch": 0.4101371774006045,
"grad_norm": 2.4391824660001884,
"learning_rate": 9.67341583808751e-06,
"loss": 0.8483,
"step": 294
},
{
"epoch": 0.41153220181353173,
"grad_norm": 2.149031048916406,
"learning_rate": 9.66906661716593e-06,
"loss": 0.6946,
"step": 295
},
{
"epoch": 0.41292722622645894,
"grad_norm": 2.181280105557748,
"learning_rate": 9.664689618442572e-06,
"loss": 0.726,
"step": 296
},
{
"epoch": 0.4143222506393862,
"grad_norm": 2.438181466583836,
"learning_rate": 9.660284867957637e-06,
"loss": 0.7853,
"step": 297
},
{
"epoch": 0.41571727505231343,
"grad_norm": 2.0436662273501764,
"learning_rate": 9.655852391916418e-06,
"loss": 0.6586,
"step": 298
},
{
"epoch": 0.41711229946524064,
"grad_norm": 2.242275872115051,
"learning_rate": 9.651392216689167e-06,
"loss": 0.73,
"step": 299
},
{
"epoch": 0.41850732387816786,
"grad_norm": 2.1669552119535487,
"learning_rate": 9.646904368810919e-06,
"loss": 0.7404,
"step": 300
},
{
"epoch": 0.4199023482910951,
"grad_norm": 2.110720652569428,
"learning_rate": 9.642388874981348e-06,
"loss": 0.7876,
"step": 301
},
{
"epoch": 0.42129737270402234,
"grad_norm": 2.1970190123405713,
"learning_rate": 9.637845762064602e-06,
"loss": 0.8062,
"step": 302
},
{
"epoch": 0.42269239711694956,
"grad_norm": 2.1979388264310264,
"learning_rate": 9.63327505708914e-06,
"loss": 0.7249,
"step": 303
},
{
"epoch": 0.4240874215298768,
"grad_norm": 2.125775928234541,
"learning_rate": 9.628676787247582e-06,
"loss": 0.7102,
"step": 304
},
{
"epoch": 0.425482445942804,
"grad_norm": 2.172104374274606,
"learning_rate": 9.624050979896533e-06,
"loss": 0.7133,
"step": 305
},
{
"epoch": 0.4268774703557312,
"grad_norm": 2.2758064319488254,
"learning_rate": 9.619397662556434e-06,
"loss": 0.7644,
"step": 306
},
{
"epoch": 0.4282724947686585,
"grad_norm": 2.2834156413794187,
"learning_rate": 9.614716862911388e-06,
"loss": 0.7374,
"step": 307
},
{
"epoch": 0.4296675191815857,
"grad_norm": 2.1152816433907766,
"learning_rate": 9.610008608809003e-06,
"loss": 0.7275,
"step": 308
},
{
"epoch": 0.4310625435945129,
"grad_norm": 2.3679870496314277,
"learning_rate": 9.605272928260215e-06,
"loss": 0.7888,
"step": 309
},
{
"epoch": 0.4324575680074401,
"grad_norm": 2.1664140472506226,
"learning_rate": 9.600509849439136e-06,
"loss": 0.7353,
"step": 310
},
{
"epoch": 0.43385259242036733,
"grad_norm": 2.0854217082250397,
"learning_rate": 9.595719400682882e-06,
"loss": 0.7598,
"step": 311
},
{
"epoch": 0.4352476168332946,
"grad_norm": 2.2708571552407752,
"learning_rate": 9.590901610491388e-06,
"loss": 0.7916,
"step": 312
},
{
"epoch": 0.4366426412462218,
"grad_norm": 2.3552880421408693,
"learning_rate": 9.586056507527266e-06,
"loss": 0.7688,
"step": 313
},
{
"epoch": 0.43803766565914903,
"grad_norm": 2.3856621240655715,
"learning_rate": 9.581184120615615e-06,
"loss": 0.7934,
"step": 314
},
{
"epoch": 0.43943269007207625,
"grad_norm": 2.2590188978724246,
"learning_rate": 9.576284478743855e-06,
"loss": 0.7218,
"step": 315
},
{
"epoch": 0.44082771448500346,
"grad_norm": 2.158495761665584,
"learning_rate": 9.571357611061556e-06,
"loss": 0.7889,
"step": 316
},
{
"epoch": 0.44222273889793073,
"grad_norm": 2.196219848841173,
"learning_rate": 9.566403546880262e-06,
"loss": 0.7683,
"step": 317
},
{
"epoch": 0.44361776331085795,
"grad_norm": 2.189233757126287,
"learning_rate": 9.561422315673316e-06,
"loss": 0.7502,
"step": 318
},
{
"epoch": 0.44501278772378516,
"grad_norm": 2.0349711390348153,
"learning_rate": 9.556413947075693e-06,
"loss": 0.705,
"step": 319
},
{
"epoch": 0.4464078121367124,
"grad_norm": 2.201388276226635,
"learning_rate": 9.551378470883813e-06,
"loss": 0.7401,
"step": 320
},
{
"epoch": 0.4478028365496396,
"grad_norm": 2.244300272417924,
"learning_rate": 9.546315917055362e-06,
"loss": 0.7801,
"step": 321
},
{
"epoch": 0.44919786096256686,
"grad_norm": 2.1314497471199405,
"learning_rate": 9.541226315709133e-06,
"loss": 0.6883,
"step": 322
},
{
"epoch": 0.4505928853754941,
"grad_norm": 2.3554603493032817,
"learning_rate": 9.536109697124822e-06,
"loss": 0.7873,
"step": 323
},
{
"epoch": 0.4519879097884213,
"grad_norm": 2.200548520780313,
"learning_rate": 9.530966091742863e-06,
"loss": 0.7228,
"step": 324
},
{
"epoch": 0.4533829342013485,
"grad_norm": 2.267412327006732,
"learning_rate": 9.525795530164248e-06,
"loss": 0.7721,
"step": 325
},
{
"epoch": 0.4547779586142757,
"grad_norm": 2.0314685715107688,
"learning_rate": 9.520598043150333e-06,
"loss": 0.7193,
"step": 326
},
{
"epoch": 0.456172983027203,
"grad_norm": 2.1973000004466123,
"learning_rate": 9.515373661622665e-06,
"loss": 0.72,
"step": 327
},
{
"epoch": 0.4575680074401302,
"grad_norm": 2.1656009071337,
"learning_rate": 9.510122416662795e-06,
"loss": 0.7795,
"step": 328
},
{
"epoch": 0.4589630318530574,
"grad_norm": 2.00595088768482,
"learning_rate": 9.504844339512096e-06,
"loss": 0.7025,
"step": 329
},
{
"epoch": 0.46035805626598464,
"grad_norm": 2.1820250634406984,
"learning_rate": 9.499539461571571e-06,
"loss": 0.7006,
"step": 330
},
{
"epoch": 0.46175308067891185,
"grad_norm": 2.000907366581318,
"learning_rate": 9.494207814401673e-06,
"loss": 0.666,
"step": 331
},
{
"epoch": 0.4631481050918391,
"grad_norm": 2.2593656128744932,
"learning_rate": 9.488849429722108e-06,
"loss": 0.7533,
"step": 332
},
{
"epoch": 0.46454312950476634,
"grad_norm": 2.1682782980420883,
"learning_rate": 9.483464339411658e-06,
"loss": 0.7769,
"step": 333
},
{
"epoch": 0.46593815391769355,
"grad_norm": 1.9957981770291782,
"learning_rate": 9.478052575507983e-06,
"loss": 0.7196,
"step": 334
},
{
"epoch": 0.46733317833062077,
"grad_norm": 2.120678305069012,
"learning_rate": 9.472614170207435e-06,
"loss": 0.7199,
"step": 335
},
{
"epoch": 0.46872820274354804,
"grad_norm": 2.1087923774164095,
"learning_rate": 9.467149155864864e-06,
"loss": 0.7437,
"step": 336
},
{
"epoch": 0.47012322715647525,
"grad_norm": 2.040923833208006,
"learning_rate": 9.461657564993419e-06,
"loss": 0.7162,
"step": 337
},
{
"epoch": 0.47151825156940247,
"grad_norm": 2.2596566935512534,
"learning_rate": 9.456139430264371e-06,
"loss": 0.7497,
"step": 338
},
{
"epoch": 0.4729132759823297,
"grad_norm": 2.1167532185301496,
"learning_rate": 9.450594784506906e-06,
"loss": 0.6863,
"step": 339
},
{
"epoch": 0.4743083003952569,
"grad_norm": 2.287756664833289,
"learning_rate": 9.445023660707928e-06,
"loss": 0.7678,
"step": 340
},
{
"epoch": 0.47570332480818417,
"grad_norm": 2.141914143165075,
"learning_rate": 9.439426092011877e-06,
"loss": 0.73,
"step": 341
},
{
"epoch": 0.4770983492211114,
"grad_norm": 2.2372702217770155,
"learning_rate": 9.433802111720513e-06,
"loss": 0.7579,
"step": 342
},
{
"epoch": 0.4784933736340386,
"grad_norm": 2.22351509996517,
"learning_rate": 9.42815175329273e-06,
"loss": 0.7907,
"step": 343
},
{
"epoch": 0.4798883980469658,
"grad_norm": 1.9645495905633132,
"learning_rate": 9.42247505034436e-06,
"loss": 0.7163,
"step": 344
},
{
"epoch": 0.48128342245989303,
"grad_norm": 2.159416398374179,
"learning_rate": 9.416772036647959e-06,
"loss": 0.7403,
"step": 345
},
{
"epoch": 0.4826784468728203,
"grad_norm": 2.2212715006212234,
"learning_rate": 9.411042746132617e-06,
"loss": 0.7626,
"step": 346
},
{
"epoch": 0.4840734712857475,
"grad_norm": 2.205058581650926,
"learning_rate": 9.405287212883756e-06,
"loss": 0.7893,
"step": 347
},
{
"epoch": 0.48546849569867473,
"grad_norm": 2.2372122036997073,
"learning_rate": 9.39950547114292e-06,
"loss": 0.7057,
"step": 348
},
{
"epoch": 0.48686352011160194,
"grad_norm": 2.1030876662201283,
"learning_rate": 9.393697555307581e-06,
"loss": 0.7356,
"step": 349
},
{
"epoch": 0.48825854452452916,
"grad_norm": 2.0752820389734326,
"learning_rate": 9.387863499930925e-06,
"loss": 0.7499,
"step": 350
},
{
"epoch": 0.48965356893745643,
"grad_norm": 2.117336157253129,
"learning_rate": 9.382003339721653e-06,
"loss": 0.7291,
"step": 351
},
{
"epoch": 0.49104859335038364,
"grad_norm": 2.165090408987389,
"learning_rate": 9.376117109543769e-06,
"loss": 0.7798,
"step": 352
},
{
"epoch": 0.49244361776331086,
"grad_norm": 2.105593870513407,
"learning_rate": 9.370204844416381e-06,
"loss": 0.728,
"step": 353
},
{
"epoch": 0.4938386421762381,
"grad_norm": 2.1682735744184796,
"learning_rate": 9.364266579513486e-06,
"loss": 0.7556,
"step": 354
},
{
"epoch": 0.4952336665891653,
"grad_norm": 1.9607721258164204,
"learning_rate": 9.358302350163758e-06,
"loss": 0.7675,
"step": 355
},
{
"epoch": 0.49662869100209256,
"grad_norm": 2.1649840762675714,
"learning_rate": 9.352312191850343e-06,
"loss": 0.7306,
"step": 356
},
{
"epoch": 0.4980237154150198,
"grad_norm": 2.1522032544678864,
"learning_rate": 9.346296140210653e-06,
"loss": 0.7229,
"step": 357
},
{
"epoch": 0.499418739827947,
"grad_norm": 2.200548786429278,
"learning_rate": 9.340254231036141e-06,
"loss": 0.733,
"step": 358
},
{
"epoch": 0.5008137642408742,
"grad_norm": 2.2849970958246364,
"learning_rate": 9.334186500272098e-06,
"loss": 0.7652,
"step": 359
},
{
"epoch": 0.5022087886538015,
"grad_norm": 2.124552667084255,
"learning_rate": 9.328092984017436e-06,
"loss": 0.7165,
"step": 360
},
{
"epoch": 0.5036038130667286,
"grad_norm": 2.16869861129638,
"learning_rate": 9.321973718524472e-06,
"loss": 0.7885,
"step": 361
},
{
"epoch": 0.5049988374796559,
"grad_norm": 2.4095767757689814,
"learning_rate": 9.315828740198714e-06,
"loss": 0.8033,
"step": 362
},
{
"epoch": 0.5063938618925832,
"grad_norm": 2.309285351320161,
"learning_rate": 9.309658085598646e-06,
"loss": 0.7282,
"step": 363
},
{
"epoch": 0.5077888863055103,
"grad_norm": 2.2278160246933556,
"learning_rate": 9.303461791435504e-06,
"loss": 0.8171,
"step": 364
},
{
"epoch": 0.5091839107184376,
"grad_norm": 2.1439102159782117,
"learning_rate": 9.29723989457307e-06,
"loss": 0.6746,
"step": 365
},
{
"epoch": 0.5105789351313648,
"grad_norm": 1.9144337516608447,
"learning_rate": 9.290992432027433e-06,
"loss": 0.6326,
"step": 366
},
{
"epoch": 0.511973959544292,
"grad_norm": 2.2661637453850565,
"learning_rate": 9.284719440966794e-06,
"loss": 0.7294,
"step": 367
},
{
"epoch": 0.5133689839572193,
"grad_norm": 2.094539074711194,
"learning_rate": 9.278420958711221e-06,
"loss": 0.786,
"step": 368
},
{
"epoch": 0.5147640083701465,
"grad_norm": 2.2536289273732932,
"learning_rate": 9.272097022732444e-06,
"loss": 0.7645,
"step": 369
},
{
"epoch": 0.5161590327830737,
"grad_norm": 2.069331048688941,
"learning_rate": 9.26574767065362e-06,
"loss": 0.7502,
"step": 370
},
{
"epoch": 0.5175540571960009,
"grad_norm": 2.1134396544547873,
"learning_rate": 9.259372940249121e-06,
"loss": 0.7404,
"step": 371
},
{
"epoch": 0.5189490816089282,
"grad_norm": 2.353506566585747,
"learning_rate": 9.252972869444297e-06,
"loss": 0.7531,
"step": 372
},
{
"epoch": 0.5203441060218554,
"grad_norm": 2.216924375444058,
"learning_rate": 9.24654749631526e-06,
"loss": 0.7246,
"step": 373
},
{
"epoch": 0.5217391304347826,
"grad_norm": 2.021849159288162,
"learning_rate": 9.240096859088653e-06,
"loss": 0.6641,
"step": 374
},
{
"epoch": 0.5231341548477099,
"grad_norm": 2.138973326205776,
"learning_rate": 9.233620996141421e-06,
"loss": 0.7311,
"step": 375
},
{
"epoch": 0.524529179260637,
"grad_norm": 2.2741045274827645,
"learning_rate": 9.22711994600059e-06,
"loss": 0.713,
"step": 376
},
{
"epoch": 0.5259242036735643,
"grad_norm": 2.0689344331701425,
"learning_rate": 9.220593747343028e-06,
"loss": 0.6859,
"step": 377
},
{
"epoch": 0.5273192280864916,
"grad_norm": 2.0106894075204296,
"learning_rate": 9.214042438995222e-06,
"loss": 0.6851,
"step": 378
},
{
"epoch": 0.5287142524994187,
"grad_norm": 2.199245862276838,
"learning_rate": 9.207466059933044e-06,
"loss": 0.7493,
"step": 379
},
{
"epoch": 0.530109276912346,
"grad_norm": 2.012825000382447,
"learning_rate": 9.20086464928152e-06,
"loss": 0.6697,
"step": 380
},
{
"epoch": 0.5315043013252732,
"grad_norm": 2.2127655874082937,
"learning_rate": 9.194238246314599e-06,
"loss": 0.7136,
"step": 381
},
{
"epoch": 0.5328993257382004,
"grad_norm": 2.2061111696963147,
"learning_rate": 9.187586890454916e-06,
"loss": 0.7519,
"step": 382
},
{
"epoch": 0.5342943501511277,
"grad_norm": 2.1376385903011332,
"learning_rate": 9.180910621273555e-06,
"loss": 0.7361,
"step": 383
},
{
"epoch": 0.5356893745640549,
"grad_norm": 2.073331825074511,
"learning_rate": 9.174209478489826e-06,
"loss": 0.7517,
"step": 384
},
{
"epoch": 0.5370843989769821,
"grad_norm": 2.2158601789564325,
"learning_rate": 9.16748350197101e-06,
"loss": 0.7798,
"step": 385
},
{
"epoch": 0.5384794233899093,
"grad_norm": 2.022881887128006,
"learning_rate": 9.160732731732144e-06,
"loss": 0.7126,
"step": 386
},
{
"epoch": 0.5398744478028366,
"grad_norm": 2.2358497538691697,
"learning_rate": 9.153957207935758e-06,
"loss": 0.73,
"step": 387
},
{
"epoch": 0.5412694722157638,
"grad_norm": 2.1270789504102225,
"learning_rate": 9.14715697089166e-06,
"loss": 0.7004,
"step": 388
},
{
"epoch": 0.542664496628691,
"grad_norm": 2.1028850233874983,
"learning_rate": 9.140332061056678e-06,
"loss": 0.7159,
"step": 389
},
{
"epoch": 0.5440595210416183,
"grad_norm": 2.1974538578891276,
"learning_rate": 9.133482519034428e-06,
"loss": 0.7684,
"step": 390
},
{
"epoch": 0.5454545454545454,
"grad_norm": 2.177428704099495,
"learning_rate": 9.126608385575076e-06,
"loss": 0.772,
"step": 391
},
{
"epoch": 0.5468495698674727,
"grad_norm": 2.0803276673189663,
"learning_rate": 9.119709701575084e-06,
"loss": 0.7218,
"step": 392
},
{
"epoch": 0.5482445942804,
"grad_norm": 2.1394984209728527,
"learning_rate": 9.112786508076972e-06,
"loss": 0.8417,
"step": 393
},
{
"epoch": 0.5496396186933271,
"grad_norm": 2.0166430885945767,
"learning_rate": 9.105838846269085e-06,
"loss": 0.6464,
"step": 394
},
{
"epoch": 0.5510346431062544,
"grad_norm": 2.2423610305000876,
"learning_rate": 9.098866757485328e-06,
"loss": 0.7855,
"step": 395
},
{
"epoch": 0.5524296675191815,
"grad_norm": 2.1328865598737368,
"learning_rate": 9.091870283204934e-06,
"loss": 0.679,
"step": 396
},
{
"epoch": 0.5538246919321088,
"grad_norm": 2.172832655132092,
"learning_rate": 9.08484946505221e-06,
"loss": 0.7611,
"step": 397
},
{
"epoch": 0.5552197163450361,
"grad_norm": 2.0620732675389704,
"learning_rate": 9.077804344796302e-06,
"loss": 0.6987,
"step": 398
},
{
"epoch": 0.5566147407579632,
"grad_norm": 1.9565268483779634,
"learning_rate": 9.070734964350926e-06,
"loss": 0.7735,
"step": 399
},
{
"epoch": 0.5580097651708905,
"grad_norm": 2.121111544028029,
"learning_rate": 9.063641365774136e-06,
"loss": 0.7207,
"step": 400
},
{
"epoch": 0.5594047895838177,
"grad_norm": 1.9141105047869074,
"learning_rate": 9.056523591268064e-06,
"loss": 0.6825,
"step": 401
},
{
"epoch": 0.560799813996745,
"grad_norm": 2.065193794568569,
"learning_rate": 9.049381683178677e-06,
"loss": 0.7655,
"step": 402
},
{
"epoch": 0.5621948384096722,
"grad_norm": 2.19597171898501,
"learning_rate": 9.042215683995516e-06,
"loss": 0.7141,
"step": 403
},
{
"epoch": 0.5635898628225994,
"grad_norm": 2.2486319147493288,
"learning_rate": 9.035025636351453e-06,
"loss": 0.7203,
"step": 404
},
{
"epoch": 0.5649848872355266,
"grad_norm": 1.9797285784818799,
"learning_rate": 9.027811583022427e-06,
"loss": 0.7036,
"step": 405
},
{
"epoch": 0.5663799116484538,
"grad_norm": 2.083913898592003,
"learning_rate": 9.020573566927196e-06,
"loss": 0.6862,
"step": 406
},
{
"epoch": 0.5677749360613811,
"grad_norm": 2.2827371402829337,
"learning_rate": 9.013311631127082e-06,
"loss": 0.7569,
"step": 407
},
{
"epoch": 0.5691699604743083,
"grad_norm": 2.3112222529668123,
"learning_rate": 9.006025818825712e-06,
"loss": 0.764,
"step": 408
},
{
"epoch": 0.5705649848872355,
"grad_norm": 2.3484053864302865,
"learning_rate": 8.998716173368762e-06,
"loss": 0.7465,
"step": 409
},
{
"epoch": 0.5719600093001628,
"grad_norm": 2.1126751658131178,
"learning_rate": 8.991382738243699e-06,
"loss": 0.7121,
"step": 410
},
{
"epoch": 0.5733550337130899,
"grad_norm": 2.1372677481421256,
"learning_rate": 8.984025557079523e-06,
"loss": 0.7225,
"step": 411
},
{
"epoch": 0.5747500581260172,
"grad_norm": 2.00580823797283,
"learning_rate": 8.97664467364651e-06,
"loss": 0.7518,
"step": 412
},
{
"epoch": 0.5761450825389445,
"grad_norm": 2.1727210121609866,
"learning_rate": 8.96924013185594e-06,
"loss": 0.713,
"step": 413
},
{
"epoch": 0.5775401069518716,
"grad_norm": 1.9451673522406057,
"learning_rate": 8.961811975759854e-06,
"loss": 0.762,
"step": 414
},
{
"epoch": 0.5789351313647989,
"grad_norm": 2.032556221899384,
"learning_rate": 8.954360249550776e-06,
"loss": 0.7239,
"step": 415
},
{
"epoch": 0.5803301557777261,
"grad_norm": 2.2282200468500073,
"learning_rate": 8.946884997561458e-06,
"loss": 0.7406,
"step": 416
},
{
"epoch": 0.5817251801906533,
"grad_norm": 2.131925151265735,
"learning_rate": 8.939386264264616e-06,
"loss": 0.6724,
"step": 417
},
{
"epoch": 0.5831202046035806,
"grad_norm": 2.201382629326267,
"learning_rate": 8.931864094272663e-06,
"loss": 0.7676,
"step": 418
},
{
"epoch": 0.5845152290165078,
"grad_norm": 2.3134699271246624,
"learning_rate": 8.924318532337444e-06,
"loss": 0.834,
"step": 419
},
{
"epoch": 0.585910253429435,
"grad_norm": 2.058607617038394,
"learning_rate": 8.91674962334997e-06,
"loss": 0.7565,
"step": 420
},
{
"epoch": 0.5873052778423622,
"grad_norm": 1.994362364952429,
"learning_rate": 8.90915741234015e-06,
"loss": 0.7187,
"step": 421
},
{
"epoch": 0.5887003022552895,
"grad_norm": 2.003476914786332,
"learning_rate": 8.901541944476529e-06,
"loss": 0.6792,
"step": 422
},
{
"epoch": 0.5900953266682167,
"grad_norm": 1.9930163818445414,
"learning_rate": 8.893903265066011e-06,
"loss": 0.6705,
"step": 423
},
{
"epoch": 0.5914903510811439,
"grad_norm": 2.1185382041837255,
"learning_rate": 8.886241419553595e-06,
"loss": 0.7642,
"step": 424
},
{
"epoch": 0.5928853754940712,
"grad_norm": 2.1802562933202574,
"learning_rate": 8.8785564535221e-06,
"loss": 0.7147,
"step": 425
},
{
"epoch": 0.5942803999069983,
"grad_norm": 1.8272633305809787,
"learning_rate": 8.870848412691899e-06,
"loss": 0.6621,
"step": 426
},
{
"epoch": 0.5956754243199256,
"grad_norm": 2.1501438887935413,
"learning_rate": 8.863117342920641e-06,
"loss": 0.7255,
"step": 427
},
{
"epoch": 0.5970704487328529,
"grad_norm": 2.1126997186333107,
"learning_rate": 8.855363290202988e-06,
"loss": 0.7365,
"step": 428
},
{
"epoch": 0.59846547314578,
"grad_norm": 2.0459113355918452,
"learning_rate": 8.84758630067033e-06,
"loss": 0.6821,
"step": 429
},
{
"epoch": 0.5998604975587073,
"grad_norm": 1.9951460974089266,
"learning_rate": 8.839786420590517e-06,
"loss": 0.6705,
"step": 430
},
{
"epoch": 0.6012555219716345,
"grad_norm": 1.9291383549477397,
"learning_rate": 8.83196369636758e-06,
"loss": 0.6743,
"step": 431
},
{
"epoch": 0.6026505463845617,
"grad_norm": 2.2262281937242014,
"learning_rate": 8.824118174541464e-06,
"loss": 0.7751,
"step": 432
},
{
"epoch": 0.604045570797489,
"grad_norm": 1.9776840814793826,
"learning_rate": 8.816249901787736e-06,
"loss": 0.689,
"step": 433
},
{
"epoch": 0.6054405952104162,
"grad_norm": 2.172712749490356,
"learning_rate": 8.808358924917322e-06,
"loss": 0.729,
"step": 434
},
{
"epoch": 0.6068356196233434,
"grad_norm": 2.1471896348433215,
"learning_rate": 8.800445290876218e-06,
"loss": 0.6904,
"step": 435
},
{
"epoch": 0.6082306440362706,
"grad_norm": 2.1350919843471514,
"learning_rate": 8.79250904674522e-06,
"loss": 0.7687,
"step": 436
},
{
"epoch": 0.6096256684491979,
"grad_norm": 2.143036898650528,
"learning_rate": 8.78455023973963e-06,
"loss": 0.7017,
"step": 437
},
{
"epoch": 0.6110206928621251,
"grad_norm": 2.173109662595841,
"learning_rate": 8.776568917208996e-06,
"loss": 0.7589,
"step": 438
},
{
"epoch": 0.6124157172750523,
"grad_norm": 2.083376789780831,
"learning_rate": 8.768565126636806e-06,
"loss": 0.7373,
"step": 439
},
{
"epoch": 0.6138107416879796,
"grad_norm": 2.189932954692665,
"learning_rate": 8.760538915640227e-06,
"loss": 0.8117,
"step": 440
},
{
"epoch": 0.6152057661009067,
"grad_norm": 2.1260961375700913,
"learning_rate": 8.752490331969807e-06,
"loss": 0.7294,
"step": 441
},
{
"epoch": 0.616600790513834,
"grad_norm": 2.084366245285451,
"learning_rate": 8.744419423509198e-06,
"loss": 0.7356,
"step": 442
},
{
"epoch": 0.6179958149267613,
"grad_norm": 1.929973613210403,
"learning_rate": 8.736326238274873e-06,
"loss": 0.7399,
"step": 443
},
{
"epoch": 0.6193908393396884,
"grad_norm": 2.191233147651038,
"learning_rate": 8.728210824415829e-06,
"loss": 0.7315,
"step": 444
},
{
"epoch": 0.6207858637526157,
"grad_norm": 1.9136719871153927,
"learning_rate": 8.720073230213315e-06,
"loss": 0.6889,
"step": 445
},
{
"epoch": 0.6221808881655428,
"grad_norm": 2.162274660157476,
"learning_rate": 8.711913504080534e-06,
"loss": 0.7939,
"step": 446
},
{
"epoch": 0.6235759125784701,
"grad_norm": 2.141540357979124,
"learning_rate": 8.703731694562364e-06,
"loss": 0.7708,
"step": 447
},
{
"epoch": 0.6249709369913974,
"grad_norm": 2.0717267261918715,
"learning_rate": 8.69552785033506e-06,
"loss": 0.6995,
"step": 448
},
{
"epoch": 0.6263659614043245,
"grad_norm": 2.0354790587901985,
"learning_rate": 8.687302020205967e-06,
"loss": 0.7891,
"step": 449
},
{
"epoch": 0.6277609858172518,
"grad_norm": 2.054315554224394,
"learning_rate": 8.679054253113239e-06,
"loss": 0.7138,
"step": 450
},
{
"epoch": 0.629156010230179,
"grad_norm": 1.97142066127702,
"learning_rate": 8.670784598125532e-06,
"loss": 0.76,
"step": 451
},
{
"epoch": 0.6305510346431062,
"grad_norm": 2.180322598168826,
"learning_rate": 8.662493104441725e-06,
"loss": 0.7577,
"step": 452
},
{
"epoch": 0.6319460590560335,
"grad_norm": 2.182404687855443,
"learning_rate": 8.65417982139062e-06,
"loss": 0.7551,
"step": 453
},
{
"epoch": 0.6333410834689607,
"grad_norm": 2.2330302295789832,
"learning_rate": 8.645844798430656e-06,
"loss": 0.7855,
"step": 454
},
{
"epoch": 0.634736107881888,
"grad_norm": 2.1948799561608903,
"learning_rate": 8.637488085149598e-06,
"loss": 0.767,
"step": 455
},
{
"epoch": 0.6361311322948151,
"grad_norm": 2.1454216259212524,
"learning_rate": 8.629109731264272e-06,
"loss": 0.7497,
"step": 456
},
{
"epoch": 0.6375261567077424,
"grad_norm": 2.0658556646093693,
"learning_rate": 8.620709786620231e-06,
"loss": 0.7532,
"step": 457
},
{
"epoch": 0.6389211811206696,
"grad_norm": 2.097430113130354,
"learning_rate": 8.612288301191497e-06,
"loss": 0.7004,
"step": 458
},
{
"epoch": 0.6403162055335968,
"grad_norm": 2.035049464077464,
"learning_rate": 8.60384532508023e-06,
"loss": 0.7773,
"step": 459
},
{
"epoch": 0.6417112299465241,
"grad_norm": 2.0899400724224413,
"learning_rate": 8.595380908516454e-06,
"loss": 0.7223,
"step": 460
},
{
"epoch": 0.6431062543594513,
"grad_norm": 2.0042098558669723,
"learning_rate": 8.586895101857747e-06,
"loss": 0.704,
"step": 461
},
{
"epoch": 0.6445012787723785,
"grad_norm": 2.2055901826008073,
"learning_rate": 8.578387955588944e-06,
"loss": 0.7232,
"step": 462
},
{
"epoch": 0.6458963031853058,
"grad_norm": 2.0614289481289902,
"learning_rate": 8.569859520321835e-06,
"loss": 0.7337,
"step": 463
},
{
"epoch": 0.6472913275982329,
"grad_norm": 2.125733860253932,
"learning_rate": 8.561309846794865e-06,
"loss": 0.7331,
"step": 464
},
{
"epoch": 0.6486863520111602,
"grad_norm": 1.9885070040021375,
"learning_rate": 8.552738985872834e-06,
"loss": 0.7158,
"step": 465
},
{
"epoch": 0.6500813764240875,
"grad_norm": 1.9749465667331927,
"learning_rate": 8.544146988546594e-06,
"loss": 0.7078,
"step": 466
},
{
"epoch": 0.6514764008370146,
"grad_norm": 2.232698100517746,
"learning_rate": 8.535533905932739e-06,
"loss": 0.7403,
"step": 467
},
{
"epoch": 0.6528714252499419,
"grad_norm": 2.0420723444019786,
"learning_rate": 8.526899789273312e-06,
"loss": 0.6924,
"step": 468
},
{
"epoch": 0.6542664496628691,
"grad_norm": 2.110810966782399,
"learning_rate": 8.518244689935491e-06,
"loss": 0.7136,
"step": 469
},
{
"epoch": 0.6556614740757963,
"grad_norm": 2.205912657877506,
"learning_rate": 8.509568659411288e-06,
"loss": 0.7277,
"step": 470
},
{
"epoch": 0.6570564984887236,
"grad_norm": 1.932306625936653,
"learning_rate": 8.500871749317244e-06,
"loss": 0.6695,
"step": 471
},
{
"epoch": 0.6584515229016508,
"grad_norm": 1.9804217745012933,
"learning_rate": 8.492154011394113e-06,
"loss": 0.6991,
"step": 472
},
{
"epoch": 0.659846547314578,
"grad_norm": 1.9943703141722595,
"learning_rate": 8.483415497506567e-06,
"loss": 0.7411,
"step": 473
},
{
"epoch": 0.6612415717275052,
"grad_norm": 1.9105485588983695,
"learning_rate": 8.474656259642874e-06,
"loss": 0.6874,
"step": 474
},
{
"epoch": 0.6626365961404325,
"grad_norm": 1.981598209924136,
"learning_rate": 8.465876349914605e-06,
"loss": 0.7693,
"step": 475
},
{
"epoch": 0.6640316205533597,
"grad_norm": 2.0468485449192526,
"learning_rate": 8.457075820556306e-06,
"loss": 0.7369,
"step": 476
},
{
"epoch": 0.6654266449662869,
"grad_norm": 1.9731837624024704,
"learning_rate": 8.448254723925205e-06,
"loss": 0.7595,
"step": 477
},
{
"epoch": 0.6668216693792142,
"grad_norm": 1.997881570988928,
"learning_rate": 8.439413112500878e-06,
"loss": 0.6957,
"step": 478
},
{
"epoch": 0.6682166937921413,
"grad_norm": 2.159991859034791,
"learning_rate": 8.430551038884965e-06,
"loss": 0.7373,
"step": 479
},
{
"epoch": 0.6696117182050686,
"grad_norm": 2.1318311718821294,
"learning_rate": 8.421668555800831e-06,
"loss": 0.7677,
"step": 480
},
{
"epoch": 0.6710067426179959,
"grad_norm": 1.8971083557397548,
"learning_rate": 8.412765716093273e-06,
"loss": 0.6238,
"step": 481
},
{
"epoch": 0.672401767030923,
"grad_norm": 2.0724177873895413,
"learning_rate": 8.403842572728186e-06,
"loss": 0.7503,
"step": 482
},
{
"epoch": 0.6737967914438503,
"grad_norm": 2.2067404032033977,
"learning_rate": 8.394899178792262e-06,
"loss": 0.8363,
"step": 483
},
{
"epoch": 0.6751918158567775,
"grad_norm": 2.165214652267942,
"learning_rate": 8.38593558749268e-06,
"loss": 0.7771,
"step": 484
},
{
"epoch": 0.6765868402697047,
"grad_norm": 1.9259138178524096,
"learning_rate": 8.376951852156764e-06,
"loss": 0.696,
"step": 485
},
{
"epoch": 0.677981864682632,
"grad_norm": 2.026340364479184,
"learning_rate": 8.367948026231692e-06,
"loss": 0.6948,
"step": 486
},
{
"epoch": 0.6793768890955592,
"grad_norm": 2.0838205387408153,
"learning_rate": 8.358924163284164e-06,
"loss": 0.7341,
"step": 487
},
{
"epoch": 0.6807719135084864,
"grad_norm": 1.8433343030714116,
"learning_rate": 8.349880317000083e-06,
"loss": 0.6905,
"step": 488
},
{
"epoch": 0.6821669379214136,
"grad_norm": 2.2138420393341622,
"learning_rate": 8.34081654118425e-06,
"loss": 0.7597,
"step": 489
},
{
"epoch": 0.6835619623343409,
"grad_norm": 2.0648043221670282,
"learning_rate": 8.331732889760021e-06,
"loss": 0.7127,
"step": 490
},
{
"epoch": 0.6849569867472681,
"grad_norm": 1.8122602084109327,
"learning_rate": 8.322629416769007e-06,
"loss": 0.6499,
"step": 491
},
{
"epoch": 0.6863520111601953,
"grad_norm": 2.049450313295537,
"learning_rate": 8.313506176370741e-06,
"loss": 0.7513,
"step": 492
},
{
"epoch": 0.6877470355731226,
"grad_norm": 2.102764954451964,
"learning_rate": 8.304363222842358e-06,
"loss": 0.7296,
"step": 493
},
{
"epoch": 0.6891420599860497,
"grad_norm": 2.1307554461817833,
"learning_rate": 8.295200610578275e-06,
"loss": 0.7337,
"step": 494
},
{
"epoch": 0.690537084398977,
"grad_norm": 2.025713936131322,
"learning_rate": 8.286018394089864e-06,
"loss": 0.7054,
"step": 495
},
{
"epoch": 0.6919321088119043,
"grad_norm": 2.0302102772230946,
"learning_rate": 8.276816628005127e-06,
"loss": 0.7164,
"step": 496
},
{
"epoch": 0.6933271332248314,
"grad_norm": 2.154855250037536,
"learning_rate": 8.267595367068375e-06,
"loss": 0.8253,
"step": 497
},
{
"epoch": 0.6947221576377587,
"grad_norm": 2.096982928844388,
"learning_rate": 8.258354666139903e-06,
"loss": 0.734,
"step": 498
},
{
"epoch": 0.6961171820506858,
"grad_norm": 1.9902700249029772,
"learning_rate": 8.249094580195655e-06,
"loss": 0.6467,
"step": 499
},
{
"epoch": 0.6975122064636131,
"grad_norm": 2.0374289309216738,
"learning_rate": 8.239815164326907e-06,
"loss": 0.7467,
"step": 500
},
{
"epoch": 0.6989072308765404,
"grad_norm": 2.048360160329615,
"learning_rate": 8.230516473739934e-06,
"loss": 0.7237,
"step": 501
},
{
"epoch": 0.7003022552894675,
"grad_norm": 2.0626472692261024,
"learning_rate": 8.221198563755683e-06,
"loss": 0.702,
"step": 502
},
{
"epoch": 0.7016972797023948,
"grad_norm": 2.34156638675061,
"learning_rate": 8.211861489809441e-06,
"loss": 0.812,
"step": 503
},
{
"epoch": 0.703092304115322,
"grad_norm": 2.007044034307796,
"learning_rate": 8.202505307450509e-06,
"loss": 0.6949,
"step": 504
},
{
"epoch": 0.7044873285282492,
"grad_norm": 2.0150876342231814,
"learning_rate": 8.193130072341872e-06,
"loss": 0.723,
"step": 505
},
{
"epoch": 0.7058823529411765,
"grad_norm": 1.9508899560375437,
"learning_rate": 8.183735840259867e-06,
"loss": 0.7058,
"step": 506
},
{
"epoch": 0.7072773773541037,
"grad_norm": 2.2595995756080933,
"learning_rate": 8.174322667093843e-06,
"loss": 0.7212,
"step": 507
},
{
"epoch": 0.708672401767031,
"grad_norm": 1.952292379501302,
"learning_rate": 8.164890608845846e-06,
"loss": 0.6669,
"step": 508
},
{
"epoch": 0.7100674261799581,
"grad_norm": 2.198574056436635,
"learning_rate": 8.155439721630265e-06,
"loss": 0.7418,
"step": 509
},
{
"epoch": 0.7114624505928854,
"grad_norm": 1.938496311102572,
"learning_rate": 8.145970061673517e-06,
"loss": 0.7191,
"step": 510
},
{
"epoch": 0.7128574750058126,
"grad_norm": 2.088136098266112,
"learning_rate": 8.136481685313699e-06,
"loss": 0.7175,
"step": 511
},
{
"epoch": 0.7142524994187398,
"grad_norm": 2.1117510039721554,
"learning_rate": 8.126974649000264e-06,
"loss": 0.7126,
"step": 512
},
{
"epoch": 0.7156475238316671,
"grad_norm": 2.126192424349008,
"learning_rate": 8.117449009293668e-06,
"loss": 0.8132,
"step": 513
},
{
"epoch": 0.7170425482445942,
"grad_norm": 2.296212315016324,
"learning_rate": 8.107904822865058e-06,
"loss": 0.7323,
"step": 514
},
{
"epoch": 0.7184375726575215,
"grad_norm": 2.0080964113695607,
"learning_rate": 8.098342146495913e-06,
"loss": 0.7047,
"step": 515
},
{
"epoch": 0.7198325970704488,
"grad_norm": 2.138945439775357,
"learning_rate": 8.088761037077718e-06,
"loss": 0.7555,
"step": 516
},
{
"epoch": 0.7212276214833759,
"grad_norm": 2.042853021947169,
"learning_rate": 8.07916155161162e-06,
"loss": 0.7318,
"step": 517
},
{
"epoch": 0.7226226458963032,
"grad_norm": 2.0036651770223535,
"learning_rate": 8.069543747208092e-06,
"loss": 0.7412,
"step": 518
},
{
"epoch": 0.7240176703092304,
"grad_norm": 2.005019684601913,
"learning_rate": 8.059907681086594e-06,
"loss": 0.6975,
"step": 519
},
{
"epoch": 0.7254126947221576,
"grad_norm": 2.035919816731715,
"learning_rate": 8.050253410575228e-06,
"loss": 0.7043,
"step": 520
},
{
"epoch": 0.7268077191350849,
"grad_norm": 2.039348786785761,
"learning_rate": 8.040580993110404e-06,
"loss": 0.734,
"step": 521
},
{
"epoch": 0.7282027435480121,
"grad_norm": 2.0316240809553943,
"learning_rate": 8.030890486236489e-06,
"loss": 0.7328,
"step": 522
},
{
"epoch": 0.7295977679609393,
"grad_norm": 1.9987332546618835,
"learning_rate": 8.021181947605474e-06,
"loss": 0.7094,
"step": 523
},
{
"epoch": 0.7309927923738665,
"grad_norm": 2.065991018116272,
"learning_rate": 8.011455434976622e-06,
"loss": 0.7567,
"step": 524
},
{
"epoch": 0.7323878167867938,
"grad_norm": 2.086968123475238,
"learning_rate": 8.001711006216138e-06,
"loss": 0.7156,
"step": 525
},
{
"epoch": 0.733782841199721,
"grad_norm": 1.9571829765680209,
"learning_rate": 7.991948719296806e-06,
"loss": 0.7218,
"step": 526
},
{
"epoch": 0.7351778656126482,
"grad_norm": 2.026547784479228,
"learning_rate": 7.98216863229766e-06,
"loss": 0.6966,
"step": 527
},
{
"epoch": 0.7365728900255755,
"grad_norm": 2.028697912817288,
"learning_rate": 7.972370803403628e-06,
"loss": 0.7385,
"step": 528
},
{
"epoch": 0.7379679144385026,
"grad_norm": 1.9258071371157235,
"learning_rate": 7.962555290905198e-06,
"loss": 0.6961,
"step": 529
},
{
"epoch": 0.7393629388514299,
"grad_norm": 1.961806972062881,
"learning_rate": 7.952722153198054e-06,
"loss": 0.6566,
"step": 530
},
{
"epoch": 0.7407579632643572,
"grad_norm": 2.118250501493784,
"learning_rate": 7.942871448782747e-06,
"loss": 0.7124,
"step": 531
},
{
"epoch": 0.7421529876772843,
"grad_norm": 2.032119437291435,
"learning_rate": 7.93300323626433e-06,
"loss": 0.6809,
"step": 532
},
{
"epoch": 0.7435480120902116,
"grad_norm": 1.9186122345582644,
"learning_rate": 7.923117574352024e-06,
"loss": 0.6946,
"step": 533
},
{
"epoch": 0.7449430365031388,
"grad_norm": 2.0031416130525295,
"learning_rate": 7.913214521858858e-06,
"loss": 0.7326,
"step": 534
},
{
"epoch": 0.746338060916066,
"grad_norm": 2.03670901008655,
"learning_rate": 7.903294137701324e-06,
"loss": 0.7095,
"step": 535
},
{
"epoch": 0.7477330853289933,
"grad_norm": 2.178152666110026,
"learning_rate": 7.89335648089903e-06,
"loss": 0.7478,
"step": 536
},
{
"epoch": 0.7491281097419205,
"grad_norm": 2.190710149751755,
"learning_rate": 7.883401610574338e-06,
"loss": 0.7086,
"step": 537
},
{
"epoch": 0.7505231341548477,
"grad_norm": 1.8616953148508328,
"learning_rate": 7.87342958595202e-06,
"loss": 0.6846,
"step": 538
},
{
"epoch": 0.7519181585677749,
"grad_norm": 2.118255012504836,
"learning_rate": 7.86344046635891e-06,
"loss": 0.7116,
"step": 539
},
{
"epoch": 0.7533131829807022,
"grad_norm": 2.165584888520477,
"learning_rate": 7.853434311223542e-06,
"loss": 0.7058,
"step": 540
},
{
"epoch": 0.7547082073936294,
"grad_norm": 2.097090278458252,
"learning_rate": 7.843411180075795e-06,
"loss": 0.7621,
"step": 541
},
{
"epoch": 0.7561032318065566,
"grad_norm": 1.9372803982312123,
"learning_rate": 7.833371132546556e-06,
"loss": 0.7135,
"step": 542
},
{
"epoch": 0.7574982562194839,
"grad_norm": 2.0536548633310683,
"learning_rate": 7.82331422836734e-06,
"loss": 0.7382,
"step": 543
},
{
"epoch": 0.758893280632411,
"grad_norm": 2.1267027897799076,
"learning_rate": 7.813240527369958e-06,
"loss": 0.7244,
"step": 544
},
{
"epoch": 0.7602883050453383,
"grad_norm": 2.001368846615703,
"learning_rate": 7.803150089486144e-06,
"loss": 0.6422,
"step": 545
},
{
"epoch": 0.7616833294582656,
"grad_norm": 2.256318701230448,
"learning_rate": 7.793042974747204e-06,
"loss": 0.73,
"step": 546
},
{
"epoch": 0.7630783538711927,
"grad_norm": 2.135892386175284,
"learning_rate": 7.782919243283671e-06,
"loss": 0.7395,
"step": 547
},
{
"epoch": 0.76447337828412,
"grad_norm": 1.9809438615716088,
"learning_rate": 7.772778955324925e-06,
"loss": 0.7646,
"step": 548
},
{
"epoch": 0.7658684026970471,
"grad_norm": 1.8728289473933306,
"learning_rate": 7.76262217119885e-06,
"loss": 0.6357,
"step": 549
},
{
"epoch": 0.7672634271099744,
"grad_norm": 2.0652897245599724,
"learning_rate": 7.75244895133147e-06,
"loss": 0.7535,
"step": 550
},
{
"epoch": 0.7686584515229017,
"grad_norm": 2.0152707634156055,
"learning_rate": 7.742259356246594e-06,
"loss": 0.7321,
"step": 551
},
{
"epoch": 0.7700534759358288,
"grad_norm": 2.061246460614199,
"learning_rate": 7.732053446565448e-06,
"loss": 0.7311,
"step": 552
},
{
"epoch": 0.7714485003487561,
"grad_norm": 1.996910814435142,
"learning_rate": 7.721831283006323e-06,
"loss": 0.7189,
"step": 553
},
{
"epoch": 0.7728435247616833,
"grad_norm": 2.0814067515772026,
"learning_rate": 7.711592926384206e-06,
"loss": 0.7267,
"step": 554
},
{
"epoch": 0.7742385491746105,
"grad_norm": 2.0543479840235706,
"learning_rate": 7.701338437610423e-06,
"loss": 0.6839,
"step": 555
},
{
"epoch": 0.7756335735875378,
"grad_norm": 2.0281629381403277,
"learning_rate": 7.691067877692277e-06,
"loss": 0.7246,
"step": 556
},
{
"epoch": 0.777028598000465,
"grad_norm": 2.106561110156005,
"learning_rate": 7.680781307732683e-06,
"loss": 0.6899,
"step": 557
},
{
"epoch": 0.7784236224133922,
"grad_norm": 2.061215326909577,
"learning_rate": 7.670478788929803e-06,
"loss": 0.719,
"step": 558
},
{
"epoch": 0.7798186468263194,
"grad_norm": 2.135026397893648,
"learning_rate": 7.660160382576683e-06,
"loss": 0.7804,
"step": 559
},
{
"epoch": 0.7812136712392467,
"grad_norm": 2.0080564958576543,
"learning_rate": 7.649826150060896e-06,
"loss": 0.7281,
"step": 560
},
{
"epoch": 0.782608695652174,
"grad_norm": 2.0202687058975037,
"learning_rate": 7.639476152864163e-06,
"loss": 0.7449,
"step": 561
},
{
"epoch": 0.7840037200651011,
"grad_norm": 1.846271729332286,
"learning_rate": 7.6291104525619985e-06,
"loss": 0.6683,
"step": 562
},
{
"epoch": 0.7853987444780284,
"grad_norm": 1.9959064783243314,
"learning_rate": 7.618729110823334e-06,
"loss": 0.7126,
"step": 563
},
{
"epoch": 0.7867937688909556,
"grad_norm": 2.2503784043031634,
"learning_rate": 7.608332189410163e-06,
"loss": 0.7813,
"step": 564
},
{
"epoch": 0.7881887933038828,
"grad_norm": 2.0759417611731794,
"learning_rate": 7.597919750177168e-06,
"loss": 0.7075,
"step": 565
},
{
"epoch": 0.7895838177168101,
"grad_norm": 2.0949390177519582,
"learning_rate": 7.587491855071347e-06,
"loss": 0.7353,
"step": 566
},
{
"epoch": 0.7909788421297372,
"grad_norm": 2.061841753787729,
"learning_rate": 7.577048566131655e-06,
"loss": 0.7618,
"step": 567
},
{
"epoch": 0.7923738665426645,
"grad_norm": 2.0082580457792356,
"learning_rate": 7.566589945488625e-06,
"loss": 0.7141,
"step": 568
},
{
"epoch": 0.7937688909555918,
"grad_norm": 1.9947755511109906,
"learning_rate": 7.556116055364008e-06,
"loss": 0.6938,
"step": 569
},
{
"epoch": 0.7951639153685189,
"grad_norm": 1.9368746356723086,
"learning_rate": 7.545626958070394e-06,
"loss": 0.6693,
"step": 570
},
{
"epoch": 0.7965589397814462,
"grad_norm": 1.9558568434157733,
"learning_rate": 7.5351227160108495e-06,
"loss": 0.6476,
"step": 571
},
{
"epoch": 0.7979539641943734,
"grad_norm": 2.173108067680643,
"learning_rate": 7.524603391678541e-06,
"loss": 0.7895,
"step": 572
},
{
"epoch": 0.7993489886073006,
"grad_norm": 2.0777332617849766,
"learning_rate": 7.51406904765636e-06,
"loss": 0.7105,
"step": 573
},
{
"epoch": 0.8007440130202279,
"grad_norm": 2.087386445741957,
"learning_rate": 7.503519746616562e-06,
"loss": 0.6785,
"step": 574
},
{
"epoch": 0.8021390374331551,
"grad_norm": 2.271079421275984,
"learning_rate": 7.492955551320381e-06,
"loss": 0.7688,
"step": 575
},
{
"epoch": 0.8035340618460823,
"grad_norm": 1.973674786364812,
"learning_rate": 7.482376524617666e-06,
"loss": 0.6533,
"step": 576
},
{
"epoch": 0.8049290862590095,
"grad_norm": 2.032036898755386,
"learning_rate": 7.4717827294464996e-06,
"loss": 0.6638,
"step": 577
},
{
"epoch": 0.8063241106719368,
"grad_norm": 2.1959112192923587,
"learning_rate": 7.461174228832828e-06,
"loss": 0.7083,
"step": 578
},
{
"epoch": 0.807719135084864,
"grad_norm": 2.075700755518968,
"learning_rate": 7.450551085890087e-06,
"loss": 0.655,
"step": 579
},
{
"epoch": 0.8091141594977912,
"grad_norm": 2.0051021855991324,
"learning_rate": 7.43991336381882e-06,
"loss": 0.6845,
"step": 580
},
{
"epoch": 0.8105091839107185,
"grad_norm": 1.9459356134656116,
"learning_rate": 7.4292611259063105e-06,
"loss": 0.6395,
"step": 581
},
{
"epoch": 0.8119042083236456,
"grad_norm": 1.9796968662023986,
"learning_rate": 7.4185944355261996e-06,
"loss": 0.7124,
"step": 582
},
{
"epoch": 0.8132992327365729,
"grad_norm": 2.1426664382476877,
"learning_rate": 7.40791335613811e-06,
"loss": 0.7238,
"step": 583
},
{
"epoch": 0.8146942571495002,
"grad_norm": 2.0380264945859135,
"learning_rate": 7.397217951287272e-06,
"loss": 0.7544,
"step": 584
},
{
"epoch": 0.8160892815624273,
"grad_norm": 2.088811897297003,
"learning_rate": 7.3865082846041415e-06,
"loss": 0.7219,
"step": 585
},
{
"epoch": 0.8174843059753546,
"grad_norm": 2.0232132320288274,
"learning_rate": 7.375784419804018e-06,
"loss": 0.723,
"step": 586
},
{
"epoch": 0.8188793303882818,
"grad_norm": 1.9817499587673113,
"learning_rate": 7.365046420686681e-06,
"loss": 0.6869,
"step": 587
},
{
"epoch": 0.820274354801209,
"grad_norm": 1.9568149151810101,
"learning_rate": 7.354294351135989e-06,
"loss": 0.7194,
"step": 588
},
{
"epoch": 0.8216693792141363,
"grad_norm": 2.0418687608379,
"learning_rate": 7.343528275119515e-06,
"loss": 0.668,
"step": 589
},
{
"epoch": 0.8230644036270635,
"grad_norm": 2.0729841048913014,
"learning_rate": 7.33274825668816e-06,
"loss": 0.7662,
"step": 590
},
{
"epoch": 0.8244594280399907,
"grad_norm": 2.1493379475849257,
"learning_rate": 7.321954359975776e-06,
"loss": 0.7593,
"step": 591
},
{
"epoch": 0.8258544524529179,
"grad_norm": 1.9637752702289315,
"learning_rate": 7.311146649198777e-06,
"loss": 0.6883,
"step": 592
},
{
"epoch": 0.8272494768658452,
"grad_norm": 1.965138108111032,
"learning_rate": 7.300325188655762e-06,
"loss": 0.6966,
"step": 593
},
{
"epoch": 0.8286445012787724,
"grad_norm": 2.0032671568946263,
"learning_rate": 7.289490042727134e-06,
"loss": 0.7509,
"step": 594
},
{
"epoch": 0.8300395256916996,
"grad_norm": 2.1322117610392737,
"learning_rate": 7.278641275874714e-06,
"loss": 0.707,
"step": 595
},
{
"epoch": 0.8314345501046269,
"grad_norm": 2.005735182999333,
"learning_rate": 7.267778952641358e-06,
"loss": 0.7319,
"step": 596
},
{
"epoch": 0.832829574517554,
"grad_norm": 2.089837838389634,
"learning_rate": 7.256903137650575e-06,
"loss": 0.6796,
"step": 597
},
{
"epoch": 0.8342245989304813,
"grad_norm": 1.9987825880429384,
"learning_rate": 7.2460138956061375e-06,
"loss": 0.709,
"step": 598
},
{
"epoch": 0.8356196233434086,
"grad_norm": 2.181112130425503,
"learning_rate": 7.235111291291702e-06,
"loss": 0.7567,
"step": 599
},
{
"epoch": 0.8370146477563357,
"grad_norm": 2.1538149633529864,
"learning_rate": 7.224195389570422e-06,
"loss": 0.7677,
"step": 600
},
{
"epoch": 0.838409672169263,
"grad_norm": 2.0301628640511162,
"learning_rate": 7.21326625538456e-06,
"loss": 0.7271,
"step": 601
},
{
"epoch": 0.8398046965821901,
"grad_norm": 2.0783608757927925,
"learning_rate": 7.202323953755106e-06,
"loss": 0.7095,
"step": 602
},
{
"epoch": 0.8411997209951174,
"grad_norm": 2.047512467999536,
"learning_rate": 7.191368549781382e-06,
"loss": 0.7095,
"step": 603
},
{
"epoch": 0.8425947454080447,
"grad_norm": 2.2284280535892127,
"learning_rate": 7.180400108640665e-06,
"loss": 0.7147,
"step": 604
},
{
"epoch": 0.8439897698209718,
"grad_norm": 2.14488402612481,
"learning_rate": 7.169418695587791e-06,
"loss": 0.6823,
"step": 605
},
{
"epoch": 0.8453847942338991,
"grad_norm": 1.9834391297217207,
"learning_rate": 7.158424375954775e-06,
"loss": 0.6668,
"step": 606
},
{
"epoch": 0.8467798186468263,
"grad_norm": 2.001260189158431,
"learning_rate": 7.147417215150411e-06,
"loss": 0.6704,
"step": 607
},
{
"epoch": 0.8481748430597535,
"grad_norm": 2.0048038596898503,
"learning_rate": 7.1363972786598925e-06,
"loss": 0.7168,
"step": 608
},
{
"epoch": 0.8495698674726808,
"grad_norm": 2.0033946556880142,
"learning_rate": 7.125364632044423e-06,
"loss": 0.7098,
"step": 609
},
{
"epoch": 0.850964891885608,
"grad_norm": 2.0693558697499923,
"learning_rate": 7.1143193409408165e-06,
"loss": 0.748,
"step": 610
},
{
"epoch": 0.8523599162985352,
"grad_norm": 1.9079762698449945,
"learning_rate": 7.103261471061116e-06,
"loss": 0.6782,
"step": 611
},
{
"epoch": 0.8537549407114624,
"grad_norm": 2.173084797654854,
"learning_rate": 7.092191088192201e-06,
"loss": 0.7465,
"step": 612
},
{
"epoch": 0.8551499651243897,
"grad_norm": 1.999853696266806,
"learning_rate": 7.0811082581953935e-06,
"loss": 0.7148,
"step": 613
},
{
"epoch": 0.856544989537317,
"grad_norm": 1.9796425973325376,
"learning_rate": 7.070013047006068e-06,
"loss": 0.7086,
"step": 614
},
{
"epoch": 0.8579400139502441,
"grad_norm": 2.004513750636165,
"learning_rate": 7.058905520633259e-06,
"loss": 0.7046,
"step": 615
},
{
"epoch": 0.8593350383631714,
"grad_norm": 1.9307586260927132,
"learning_rate": 7.047785745159266e-06,
"loss": 0.634,
"step": 616
},
{
"epoch": 0.8607300627760985,
"grad_norm": 2.2131652759757032,
"learning_rate": 7.036653786739264e-06,
"loss": 0.6867,
"step": 617
},
{
"epoch": 0.8621250871890258,
"grad_norm": 2.11762373803619,
"learning_rate": 7.02550971160091e-06,
"loss": 0.7141,
"step": 618
},
{
"epoch": 0.8635201116019531,
"grad_norm": 2.062189920448756,
"learning_rate": 7.0143535860439425e-06,
"loss": 0.6678,
"step": 619
},
{
"epoch": 0.8649151360148802,
"grad_norm": 2.0311504435657266,
"learning_rate": 7.0031854764398e-06,
"loss": 0.7157,
"step": 620
},
{
"epoch": 0.8663101604278075,
"grad_norm": 2.0490888006871346,
"learning_rate": 6.9920054492312086e-06,
"loss": 0.726,
"step": 621
},
{
"epoch": 0.8677051848407347,
"grad_norm": 2.1515674827759677,
"learning_rate": 6.9808135709318015e-06,
"loss": 0.7118,
"step": 622
},
{
"epoch": 0.8691002092536619,
"grad_norm": 1.929335451435658,
"learning_rate": 6.9696099081257176e-06,
"loss": 0.6872,
"step": 623
},
{
"epoch": 0.8704952336665892,
"grad_norm": 2.0572983608669855,
"learning_rate": 6.958394527467204e-06,
"loss": 0.6748,
"step": 624
},
{
"epoch": 0.8718902580795164,
"grad_norm": 1.9837573813226999,
"learning_rate": 6.947167495680224e-06,
"loss": 0.707,
"step": 625
},
{
"epoch": 0.8732852824924436,
"grad_norm": 2.04771350119922,
"learning_rate": 6.935928879558052e-06,
"loss": 0.6617,
"step": 626
},
{
"epoch": 0.8746803069053708,
"grad_norm": 2.1485933426462758,
"learning_rate": 6.9246787459628895e-06,
"loss": 0.6927,
"step": 627
},
{
"epoch": 0.8760753313182981,
"grad_norm": 1.9152886720507631,
"learning_rate": 6.913417161825449e-06,
"loss": 0.6327,
"step": 628
},
{
"epoch": 0.8774703557312253,
"grad_norm": 2.045004872767665,
"learning_rate": 6.902144194144576e-06,
"loss": 0.6979,
"step": 629
},
{
"epoch": 0.8788653801441525,
"grad_norm": 2.0896741032979214,
"learning_rate": 6.890859909986835e-06,
"loss": 0.68,
"step": 630
},
{
"epoch": 0.8802604045570798,
"grad_norm": 1.9208151556554776,
"learning_rate": 6.8795643764861144e-06,
"loss": 0.6964,
"step": 631
},
{
"epoch": 0.8816554289700069,
"grad_norm": 1.9767473419286516,
"learning_rate": 6.868257660843234e-06,
"loss": 0.6924,
"step": 632
},
{
"epoch": 0.8830504533829342,
"grad_norm": 2.0314146752614706,
"learning_rate": 6.8569398303255345e-06,
"loss": 0.802,
"step": 633
},
{
"epoch": 0.8844454777958615,
"grad_norm": 2.0170897943284958,
"learning_rate": 6.845610952266486e-06,
"loss": 0.6554,
"step": 634
},
{
"epoch": 0.8858405022087886,
"grad_norm": 1.9814807999478814,
"learning_rate": 6.834271094065284e-06,
"loss": 0.6943,
"step": 635
},
{
"epoch": 0.8872355266217159,
"grad_norm": 1.9714093373330037,
"learning_rate": 6.822920323186445e-06,
"loss": 0.7085,
"step": 636
},
{
"epoch": 0.8886305510346431,
"grad_norm": 2.086916641932359,
"learning_rate": 6.811558707159414e-06,
"loss": 0.7342,
"step": 637
},
{
"epoch": 0.8900255754475703,
"grad_norm": 1.870451490838753,
"learning_rate": 6.800186313578152e-06,
"loss": 0.6899,
"step": 638
},
{
"epoch": 0.8914205998604976,
"grad_norm": 2.3721855893316186,
"learning_rate": 6.7888032101007416e-06,
"loss": 0.778,
"step": 639
},
{
"epoch": 0.8928156242734248,
"grad_norm": 1.9186274740516478,
"learning_rate": 6.777409464448983e-06,
"loss": 0.6948,
"step": 640
},
{
"epoch": 0.894210648686352,
"grad_norm": 2.058282213173254,
"learning_rate": 6.76600514440799e-06,
"loss": 0.7115,
"step": 641
},
{
"epoch": 0.8956056730992792,
"grad_norm": 2.006384543727099,
"learning_rate": 6.754590317825785e-06,
"loss": 0.6537,
"step": 642
},
{
"epoch": 0.8970006975122065,
"grad_norm": 1.8915235746908028,
"learning_rate": 6.743165052612899e-06,
"loss": 0.729,
"step": 643
},
{
"epoch": 0.8983957219251337,
"grad_norm": 1.9836308406433447,
"learning_rate": 6.731729416741961e-06,
"loss": 0.7091,
"step": 644
},
{
"epoch": 0.8997907463380609,
"grad_norm": 1.91315435206782,
"learning_rate": 6.72028347824731e-06,
"loss": 0.6782,
"step": 645
},
{
"epoch": 0.9011857707509882,
"grad_norm": 2.164605036051746,
"learning_rate": 6.708827305224566e-06,
"loss": 0.7599,
"step": 646
},
{
"epoch": 0.9025807951639153,
"grad_norm": 2.1307619231031705,
"learning_rate": 6.697360965830244e-06,
"loss": 0.7252,
"step": 647
},
{
"epoch": 0.9039758195768426,
"grad_norm": 1.9739462078301375,
"learning_rate": 6.6858845282813416e-06,
"loss": 0.6881,
"step": 648
},
{
"epoch": 0.9053708439897699,
"grad_norm": 2.305594622925065,
"learning_rate": 6.674398060854931e-06,
"loss": 0.7853,
"step": 649
},
{
"epoch": 0.906765868402697,
"grad_norm": 2.040722369484868,
"learning_rate": 6.662901631887761e-06,
"loss": 0.6823,
"step": 650
},
{
"epoch": 0.9081608928156243,
"grad_norm": 1.9646186422608911,
"learning_rate": 6.651395309775837e-06,
"loss": 0.6535,
"step": 651
},
{
"epoch": 0.9095559172285514,
"grad_norm": 1.8857798039813913,
"learning_rate": 6.639879162974027e-06,
"loss": 0.658,
"step": 652
},
{
"epoch": 0.9109509416414787,
"grad_norm": 1.8401539668513296,
"learning_rate": 6.62835325999565e-06,
"loss": 0.6965,
"step": 653
},
{
"epoch": 0.912345966054406,
"grad_norm": 2.0348228305130043,
"learning_rate": 6.616817669412066e-06,
"loss": 0.7916,
"step": 654
},
{
"epoch": 0.9137409904673331,
"grad_norm": 2.0027751723218272,
"learning_rate": 6.6052724598522696e-06,
"loss": 0.6848,
"step": 655
},
{
"epoch": 0.9151360148802604,
"grad_norm": 1.9205553455437767,
"learning_rate": 6.59371770000248e-06,
"loss": 0.6595,
"step": 656
},
{
"epoch": 0.9165310392931876,
"grad_norm": 1.988113920112164,
"learning_rate": 6.582153458605738e-06,
"loss": 0.7295,
"step": 657
},
{
"epoch": 0.9179260637061148,
"grad_norm": 1.969446729465624,
"learning_rate": 6.570579804461492e-06,
"loss": 0.7094,
"step": 658
},
{
"epoch": 0.9193210881190421,
"grad_norm": 1.8660933701134175,
"learning_rate": 6.558996806425188e-06,
"loss": 0.7219,
"step": 659
},
{
"epoch": 0.9207161125319693,
"grad_norm": 1.898278443506339,
"learning_rate": 6.5474045334078654e-06,
"loss": 0.6961,
"step": 660
},
{
"epoch": 0.9221111369448965,
"grad_norm": 2.2155411196979053,
"learning_rate": 6.5358030543757375e-06,
"loss": 0.7759,
"step": 661
},
{
"epoch": 0.9235061613578237,
"grad_norm": 2.094799591086162,
"learning_rate": 6.524192438349794e-06,
"loss": 0.6898,
"step": 662
},
{
"epoch": 0.924901185770751,
"grad_norm": 1.924740068996983,
"learning_rate": 6.51257275440538e-06,
"loss": 0.6931,
"step": 663
},
{
"epoch": 0.9262962101836782,
"grad_norm": 2.011632926786668,
"learning_rate": 6.500944071671789e-06,
"loss": 0.7092,
"step": 664
},
{
"epoch": 0.9276912345966054,
"grad_norm": 2.0725376084974596,
"learning_rate": 6.489306459331851e-06,
"loss": 0.6927,
"step": 665
},
{
"epoch": 0.9290862590095327,
"grad_norm": 2.1472076797554918,
"learning_rate": 6.477659986621522e-06,
"loss": 0.7601,
"step": 666
},
{
"epoch": 0.93048128342246,
"grad_norm": 2.108439533312848,
"learning_rate": 6.46600472282947e-06,
"loss": 0.7315,
"step": 667
},
{
"epoch": 0.9318763078353871,
"grad_norm": 2.0157312555896874,
"learning_rate": 6.454340737296665e-06,
"loss": 0.6428,
"step": 668
},
{
"epoch": 0.9332713322483144,
"grad_norm": 1.955747483057349,
"learning_rate": 6.442668099415967e-06,
"loss": 0.7426,
"step": 669
},
{
"epoch": 0.9346663566612415,
"grad_norm": 2.1284318764319545,
"learning_rate": 6.430986878631708e-06,
"loss": 0.695,
"step": 670
},
{
"epoch": 0.9360613810741688,
"grad_norm": 1.9889109021997742,
"learning_rate": 6.4192971444392835e-06,
"loss": 0.711,
"step": 671
},
{
"epoch": 0.9374564054870961,
"grad_norm": 2.103606798692871,
"learning_rate": 6.407598966384743e-06,
"loss": 0.7219,
"step": 672
},
{
"epoch": 0.9388514299000232,
"grad_norm": 1.876811159303784,
"learning_rate": 6.395892414064363e-06,
"loss": 0.6762,
"step": 673
},
{
"epoch": 0.9402464543129505,
"grad_norm": 1.7987895850167992,
"learning_rate": 6.384177557124247e-06,
"loss": 0.6521,
"step": 674
},
{
"epoch": 0.9416414787258777,
"grad_norm": 2.282079516291976,
"learning_rate": 6.372454465259905e-06,
"loss": 0.7343,
"step": 675
},
{
"epoch": 0.9430365031388049,
"grad_norm": 1.939959854059506,
"learning_rate": 6.360723208215838e-06,
"loss": 0.6972,
"step": 676
},
{
"epoch": 0.9444315275517322,
"grad_norm": 2.137165196217349,
"learning_rate": 6.348983855785122e-06,
"loss": 0.7757,
"step": 677
},
{
"epoch": 0.9458265519646594,
"grad_norm": 1.869415519058021,
"learning_rate": 6.3372364778090005e-06,
"loss": 0.6613,
"step": 678
},
{
"epoch": 0.9472215763775866,
"grad_norm": 1.8737093174455668,
"learning_rate": 6.325481144176458e-06,
"loss": 0.6658,
"step": 679
},
{
"epoch": 0.9486166007905138,
"grad_norm": 1.9207669698589307,
"learning_rate": 6.3137179248238145e-06,
"loss": 0.6655,
"step": 680
},
{
"epoch": 0.9500116252034411,
"grad_norm": 2.0566860383052386,
"learning_rate": 6.301946889734302e-06,
"loss": 0.6926,
"step": 681
},
{
"epoch": 0.9514066496163683,
"grad_norm": 2.0041349151703836,
"learning_rate": 6.290168108937651e-06,
"loss": 0.7247,
"step": 682
},
{
"epoch": 0.9528016740292955,
"grad_norm": 1.9546701865263272,
"learning_rate": 6.2783816525096765e-06,
"loss": 0.6723,
"step": 683
},
{
"epoch": 0.9541966984422228,
"grad_norm": 2.111279121287525,
"learning_rate": 6.266587590571852e-06,
"loss": 0.7451,
"step": 684
},
{
"epoch": 0.9555917228551499,
"grad_norm": 1.8926865719781227,
"learning_rate": 6.254785993290907e-06,
"loss": 0.7097,
"step": 685
},
{
"epoch": 0.9569867472680772,
"grad_norm": 1.9235814882501219,
"learning_rate": 6.242976930878395e-06,
"loss": 0.6519,
"step": 686
},
{
"epoch": 0.9583817716810045,
"grad_norm": 2.0782433443787425,
"learning_rate": 6.231160473590283e-06,
"loss": 0.7351,
"step": 687
},
{
"epoch": 0.9597767960939316,
"grad_norm": 1.7659331972757786,
"learning_rate": 6.219336691726537e-06,
"loss": 0.6201,
"step": 688
},
{
"epoch": 0.9611718205068589,
"grad_norm": 2.0092199359038108,
"learning_rate": 6.20750565563069e-06,
"loss": 0.6711,
"step": 689
},
{
"epoch": 0.9625668449197861,
"grad_norm": 2.0503621225664115,
"learning_rate": 6.195667435689445e-06,
"loss": 0.7095,
"step": 690
},
{
"epoch": 0.9639618693327133,
"grad_norm": 2.084108781725168,
"learning_rate": 6.183822102332234e-06,
"loss": 0.6899,
"step": 691
},
{
"epoch": 0.9653568937456406,
"grad_norm": 1.9166997244216846,
"learning_rate": 6.171969726030814e-06,
"loss": 0.6558,
"step": 692
},
{
"epoch": 0.9667519181585678,
"grad_norm": 2.001051202379286,
"learning_rate": 6.16011037729884e-06,
"loss": 0.7175,
"step": 693
},
{
"epoch": 0.968146942571495,
"grad_norm": 1.9123567824105474,
"learning_rate": 6.148244126691451e-06,
"loss": 0.6761,
"step": 694
},
{
"epoch": 0.9695419669844222,
"grad_norm": 2.057429814174208,
"learning_rate": 6.136371044804847e-06,
"loss": 0.7319,
"step": 695
},
{
"epoch": 0.9709369913973495,
"grad_norm": 1.962229379260547,
"learning_rate": 6.124491202275866e-06,
"loss": 0.6579,
"step": 696
},
{
"epoch": 0.9723320158102767,
"grad_norm": 2.0201399268724347,
"learning_rate": 6.112604669781572e-06,
"loss": 0.7438,
"step": 697
},
{
"epoch": 0.9737270402232039,
"grad_norm": 2.1063858179693766,
"learning_rate": 6.100711518038828e-06,
"loss": 0.7623,
"step": 698
},
{
"epoch": 0.9751220646361312,
"grad_norm": 1.898175284952164,
"learning_rate": 6.088811817803877e-06,
"loss": 0.6738,
"step": 699
},
{
"epoch": 0.9765170890490583,
"grad_norm": 1.9084924564930574,
"learning_rate": 6.0769056398719205e-06,
"loss": 0.6559,
"step": 700
},
{
"epoch": 0.9779121134619856,
"grad_norm": 2.034255025855498,
"learning_rate": 6.064993055076697e-06,
"loss": 0.7053,
"step": 701
},
{
"epoch": 0.9793071378749129,
"grad_norm": 1.9943164564563032,
"learning_rate": 6.053074134290065e-06,
"loss": 0.6502,
"step": 702
},
{
"epoch": 0.98070216228784,
"grad_norm": 1.9082522000979645,
"learning_rate": 6.041148948421575e-06,
"loss": 0.6971,
"step": 703
},
{
"epoch": 0.9820971867007673,
"grad_norm": 1.9423848476086434,
"learning_rate": 6.0292175684180506e-06,
"loss": 0.6925,
"step": 704
},
{
"epoch": 0.9834922111136944,
"grad_norm": 1.9350501223395191,
"learning_rate": 6.0172800652631706e-06,
"loss": 0.6574,
"step": 705
},
{
"epoch": 0.9848872355266217,
"grad_norm": 1.948394189706791,
"learning_rate": 6.005336509977035e-06,
"loss": 0.7232,
"step": 706
},
{
"epoch": 0.986282259939549,
"grad_norm": 1.9718390783155533,
"learning_rate": 5.9933869736157524e-06,
"loss": 0.671,
"step": 707
},
{
"epoch": 0.9876772843524761,
"grad_norm": 1.7860631184750806,
"learning_rate": 5.981431527271022e-06,
"loss": 0.6619,
"step": 708
},
{
"epoch": 0.9890723087654034,
"grad_norm": 2.190091158506914,
"learning_rate": 5.9694702420696935e-06,
"loss": 0.6881,
"step": 709
},
{
"epoch": 0.9904673331783306,
"grad_norm": 2.111270110174928,
"learning_rate": 5.9575031891733594e-06,
"loss": 0.7766,
"step": 710
},
{
"epoch": 0.9918623575912578,
"grad_norm": 2.0590626970008383,
"learning_rate": 5.945530439777924e-06,
"loss": 0.6849,
"step": 711
},
{
"epoch": 0.9932573820041851,
"grad_norm": 2.058768534188736,
"learning_rate": 5.933552065113182e-06,
"loss": 0.6918,
"step": 712
},
{
"epoch": 0.9946524064171123,
"grad_norm": 1.8988454706123046,
"learning_rate": 5.9215681364423975e-06,
"loss": 0.6379,
"step": 713
},
{
"epoch": 0.9960474308300395,
"grad_norm": 1.8925107930593552,
"learning_rate": 5.909578725061872e-06,
"loss": 0.6666,
"step": 714
},
{
"epoch": 0.9974424552429667,
"grad_norm": 1.9105216289220566,
"learning_rate": 5.897583902300531e-06,
"loss": 0.6739,
"step": 715
},
{
"epoch": 0.998837479655894,
"grad_norm": 2.030203006310582,
"learning_rate": 5.885583739519489e-06,
"loss": 0.6758,
"step": 716
},
{
"epoch": 1.0013950244129273,
"grad_norm": 2.7447928004495568,
"learning_rate": 5.873578308111636e-06,
"loss": 1.1652,
"step": 717
},
{
"epoch": 1.0027900488258545,
"grad_norm": 1.7697258758954115,
"learning_rate": 5.861567679501201e-06,
"loss": 0.4832,
"step": 718
},
{
"epoch": 1.0041850732387816,
"grad_norm": 1.8831799173238042,
"learning_rate": 5.849551925143334e-06,
"loss": 0.4387,
"step": 719
},
{
"epoch": 1.0055800976517089,
"grad_norm": 1.8886920606969104,
"learning_rate": 5.837531116523683e-06,
"loss": 0.4728,
"step": 720
},
{
"epoch": 1.0069751220646361,
"grad_norm": 1.6199177638252187,
"learning_rate": 5.825505325157962e-06,
"loss": 0.4952,
"step": 721
},
{
"epoch": 1.0083701464775634,
"grad_norm": 1.8465149809489871,
"learning_rate": 5.8134746225915305e-06,
"loss": 0.4316,
"step": 722
},
{
"epoch": 1.0097651708904907,
"grad_norm": 1.7609710454949228,
"learning_rate": 5.801439080398968e-06,
"loss": 0.5242,
"step": 723
},
{
"epoch": 1.0111601953034177,
"grad_norm": 1.915812031265783,
"learning_rate": 5.789398770183642e-06,
"loss": 0.4718,
"step": 724
},
{
"epoch": 1.012555219716345,
"grad_norm": 2.061689014174379,
"learning_rate": 5.77735376357729e-06,
"loss": 0.5115,
"step": 725
},
{
"epoch": 1.0139502441292723,
"grad_norm": 2.0246963373463465,
"learning_rate": 5.76530413223959e-06,
"loss": 0.4765,
"step": 726
},
{
"epoch": 1.0153452685421995,
"grad_norm": 2.6269844127481217,
"learning_rate": 5.753249947857731e-06,
"loss": 0.5015,
"step": 727
},
{
"epoch": 1.0167402929551268,
"grad_norm": 2.2152920615212595,
"learning_rate": 5.741191282145995e-06,
"loss": 0.5109,
"step": 728
},
{
"epoch": 1.0181353173680538,
"grad_norm": 2.3867329877687196,
"learning_rate": 5.729128206845317e-06,
"loss": 0.4681,
"step": 729
},
{
"epoch": 1.0195303417809811,
"grad_norm": 2.026990800833781,
"learning_rate": 5.717060793722876e-06,
"loss": 0.4662,
"step": 730
},
{
"epoch": 1.0209253661939084,
"grad_norm": 2.2793548911246875,
"learning_rate": 5.704989114571649e-06,
"loss": 0.5151,
"step": 731
},
{
"epoch": 1.0223203906068357,
"grad_norm": 2.0202960449035343,
"learning_rate": 5.692913241209997e-06,
"loss": 0.4442,
"step": 732
},
{
"epoch": 1.023715415019763,
"grad_norm": 1.95038181870786,
"learning_rate": 5.680833245481234e-06,
"loss": 0.4744,
"step": 733
},
{
"epoch": 1.02511043943269,
"grad_norm": 2.04344748510266,
"learning_rate": 5.668749199253198e-06,
"loss": 0.4951,
"step": 734
},
{
"epoch": 1.0265054638456172,
"grad_norm": 1.8910556764395299,
"learning_rate": 5.656661174417829e-06,
"loss": 0.4809,
"step": 735
},
{
"epoch": 1.0279004882585445,
"grad_norm": 1.9255139030660995,
"learning_rate": 5.644569242890728e-06,
"loss": 0.4393,
"step": 736
},
{
"epoch": 1.0292955126714718,
"grad_norm": 2.2064737359647353,
"learning_rate": 5.632473476610748e-06,
"loss": 0.5139,
"step": 737
},
{
"epoch": 1.030690537084399,
"grad_norm": 1.826337759870788,
"learning_rate": 5.62037394753955e-06,
"loss": 0.4482,
"step": 738
},
{
"epoch": 1.032085561497326,
"grad_norm": 1.9833849963275132,
"learning_rate": 5.608270727661183e-06,
"loss": 0.4806,
"step": 739
},
{
"epoch": 1.0334805859102534,
"grad_norm": 1.9740005987225675,
"learning_rate": 5.596163888981656e-06,
"loss": 0.496,
"step": 740
},
{
"epoch": 1.0348756103231807,
"grad_norm": 1.9411807491663124,
"learning_rate": 5.584053503528503e-06,
"loss": 0.4537,
"step": 741
},
{
"epoch": 1.036270634736108,
"grad_norm": 1.8861912900200872,
"learning_rate": 5.5719396433503615e-06,
"loss": 0.5442,
"step": 742
},
{
"epoch": 1.0376656591490352,
"grad_norm": 2.073505442126394,
"learning_rate": 5.559822380516539e-06,
"loss": 0.4968,
"step": 743
},
{
"epoch": 1.0390606835619622,
"grad_norm": 2.049207981076892,
"learning_rate": 5.5477017871165925e-06,
"loss": 0.4514,
"step": 744
},
{
"epoch": 1.0404557079748895,
"grad_norm": 1.9198267966469686,
"learning_rate": 5.5355779352598884e-06,
"loss": 0.5163,
"step": 745
},
{
"epoch": 1.0418507323878168,
"grad_norm": 2.037912592040736,
"learning_rate": 5.523450897075179e-06,
"loss": 0.4664,
"step": 746
},
{
"epoch": 1.043245756800744,
"grad_norm": 2.1410052063539076,
"learning_rate": 5.511320744710171e-06,
"loss": 0.4419,
"step": 747
},
{
"epoch": 1.0446407812136713,
"grad_norm": 2.041057827262413,
"learning_rate": 5.4991875503311075e-06,
"loss": 0.5258,
"step": 748
},
{
"epoch": 1.0460358056265984,
"grad_norm": 2.1483535732137566,
"learning_rate": 5.487051386122319e-06,
"loss": 0.4334,
"step": 749
},
{
"epoch": 1.0474308300395256,
"grad_norm": 1.987139519131828,
"learning_rate": 5.47491232428581e-06,
"loss": 0.4383,
"step": 750
},
{
"epoch": 1.048825854452453,
"grad_norm": 2.1635983832379044,
"learning_rate": 5.462770437040824e-06,
"loss": 0.5084,
"step": 751
},
{
"epoch": 1.0502208788653802,
"grad_norm": 2.1342642535994174,
"learning_rate": 5.450625796623411e-06,
"loss": 0.4721,
"step": 752
},
{
"epoch": 1.0516159032783075,
"grad_norm": 2.148708587244624,
"learning_rate": 5.438478475286003e-06,
"loss": 0.4217,
"step": 753
},
{
"epoch": 1.0530109276912345,
"grad_norm": 1.9486786460770569,
"learning_rate": 5.4263285452969805e-06,
"loss": 0.4737,
"step": 754
},
{
"epoch": 1.0544059521041618,
"grad_norm": 2.112176600834154,
"learning_rate": 5.414176078940244e-06,
"loss": 0.4689,
"step": 755
},
{
"epoch": 1.055800976517089,
"grad_norm": 2.0011944624765525,
"learning_rate": 5.402021148514784e-06,
"loss": 0.4545,
"step": 756
},
{
"epoch": 1.0571960009300163,
"grad_norm": 2.2068689969149244,
"learning_rate": 5.389863826334248e-06,
"loss": 0.4579,
"step": 757
},
{
"epoch": 1.0585910253429436,
"grad_norm": 2.0433936765493588,
"learning_rate": 5.37770418472652e-06,
"loss": 0.4162,
"step": 758
},
{
"epoch": 1.0599860497558706,
"grad_norm": 1.9980953151822662,
"learning_rate": 5.365542296033274e-06,
"loss": 0.4761,
"step": 759
},
{
"epoch": 1.061381074168798,
"grad_norm": 1.9795377015612743,
"learning_rate": 5.353378232609557e-06,
"loss": 0.4848,
"step": 760
},
{
"epoch": 1.0627760985817252,
"grad_norm": 2.1199669786852393,
"learning_rate": 5.341212066823356e-06,
"loss": 0.4862,
"step": 761
},
{
"epoch": 1.0641711229946524,
"grad_norm": 2.0944681341647713,
"learning_rate": 5.329043871055161e-06,
"loss": 0.4954,
"step": 762
},
{
"epoch": 1.0655661474075797,
"grad_norm": 2.145189792268484,
"learning_rate": 5.316873717697545e-06,
"loss": 0.4738,
"step": 763
},
{
"epoch": 1.0669611718205068,
"grad_norm": 2.0264734767440116,
"learning_rate": 5.30470167915472e-06,
"loss": 0.4087,
"step": 764
},
{
"epoch": 1.068356196233434,
"grad_norm": 1.9348073866593454,
"learning_rate": 5.29252782784212e-06,
"loss": 0.5049,
"step": 765
},
{
"epoch": 1.0697512206463613,
"grad_norm": 2.204583768085505,
"learning_rate": 5.2803522361859596e-06,
"loss": 0.4287,
"step": 766
},
{
"epoch": 1.0711462450592886,
"grad_norm": 1.9937257246231703,
"learning_rate": 5.268174976622811e-06,
"loss": 0.4685,
"step": 767
},
{
"epoch": 1.0725412694722158,
"grad_norm": 2.013514450168237,
"learning_rate": 5.255996121599167e-06,
"loss": 0.4948,
"step": 768
},
{
"epoch": 1.073936293885143,
"grad_norm": 2.0772460175921665,
"learning_rate": 5.243815743571009e-06,
"loss": 0.4554,
"step": 769
},
{
"epoch": 1.0753313182980702,
"grad_norm": 2.014215366177637,
"learning_rate": 5.231633915003389e-06,
"loss": 0.481,
"step": 770
},
{
"epoch": 1.0767263427109974,
"grad_norm": 1.9613576871714251,
"learning_rate": 5.219450708369977e-06,
"loss": 0.4583,
"step": 771
},
{
"epoch": 1.0781213671239247,
"grad_norm": 2.0232554045863655,
"learning_rate": 5.2072661961526505e-06,
"loss": 0.4802,
"step": 772
},
{
"epoch": 1.079516391536852,
"grad_norm": 2.1785526324306392,
"learning_rate": 5.19508045084105e-06,
"loss": 0.4634,
"step": 773
},
{
"epoch": 1.080911415949779,
"grad_norm": 2.083047280181435,
"learning_rate": 5.1828935449321515e-06,
"loss": 0.4648,
"step": 774
},
{
"epoch": 1.0823064403627063,
"grad_norm": 1.939128120577853,
"learning_rate": 5.17070555092984e-06,
"loss": 0.4546,
"step": 775
},
{
"epoch": 1.0837014647756336,
"grad_norm": 2.0490314541899015,
"learning_rate": 5.15851654134447e-06,
"loss": 0.4667,
"step": 776
},
{
"epoch": 1.0850964891885608,
"grad_norm": 2.018626496282391,
"learning_rate": 5.146326588692439e-06,
"loss": 0.4185,
"step": 777
},
{
"epoch": 1.086491513601488,
"grad_norm": 2.2258491166908354,
"learning_rate": 5.1341357654957546e-06,
"loss": 0.4433,
"step": 778
},
{
"epoch": 1.0878865380144151,
"grad_norm": 2.0556634088378294,
"learning_rate": 5.121944144281606e-06,
"loss": 0.4977,
"step": 779
},
{
"epoch": 1.0892815624273424,
"grad_norm": 2.278773856505871,
"learning_rate": 5.109751797581929e-06,
"loss": 0.4549,
"step": 780
},
{
"epoch": 1.0906765868402697,
"grad_norm": 2.032042799453305,
"learning_rate": 5.097558797932973e-06,
"loss": 0.4469,
"step": 781
},
{
"epoch": 1.092071611253197,
"grad_norm": 2.1595220009518754,
"learning_rate": 5.085365217874875e-06,
"loss": 0.4554,
"step": 782
},
{
"epoch": 1.0934666356661242,
"grad_norm": 2.0764422719191664,
"learning_rate": 5.073171129951223e-06,
"loss": 0.513,
"step": 783
},
{
"epoch": 1.0948616600790513,
"grad_norm": 2.195567440808966,
"learning_rate": 5.060976606708628e-06,
"loss": 0.4653,
"step": 784
},
{
"epoch": 1.0962566844919786,
"grad_norm": 2.021246542496036,
"learning_rate": 5.048781720696291e-06,
"loss": 0.4598,
"step": 785
},
{
"epoch": 1.0976517089049058,
"grad_norm": 2.0641574251199426,
"learning_rate": 5.036586544465571e-06,
"loss": 0.4547,
"step": 786
},
{
"epoch": 1.099046733317833,
"grad_norm": 2.0272284013054804,
"learning_rate": 5.024391150569551e-06,
"loss": 0.4475,
"step": 787
},
{
"epoch": 1.1004417577307604,
"grad_norm": 2.723082585619733,
"learning_rate": 5.012195611562613e-06,
"loss": 0.4427,
"step": 788
},
{
"epoch": 1.1018367821436876,
"grad_norm": 1.7395086950703968,
"learning_rate": 5e-06,
"loss": 0.4505,
"step": 789
},
{
"epoch": 1.1032318065566147,
"grad_norm": 2.0298920165502947,
"learning_rate": 4.987804388437388e-06,
"loss": 0.4817,
"step": 790
},
{
"epoch": 1.104626830969542,
"grad_norm": 1.9418324231053923,
"learning_rate": 4.975608849430451e-06,
"loss": 0.426,
"step": 791
},
{
"epoch": 1.1060218553824692,
"grad_norm": 2.0165571549580994,
"learning_rate": 4.963413455534431e-06,
"loss": 0.4341,
"step": 792
},
{
"epoch": 1.1074168797953965,
"grad_norm": 1.9357623746058135,
"learning_rate": 4.95121827930371e-06,
"loss": 0.5005,
"step": 793
},
{
"epoch": 1.1088119042083235,
"grad_norm": 2.21037173558054,
"learning_rate": 4.939023393291373e-06,
"loss": 0.4889,
"step": 794
},
{
"epoch": 1.1102069286212508,
"grad_norm": 2.0805765975077612,
"learning_rate": 4.926828870048779e-06,
"loss": 0.468,
"step": 795
},
{
"epoch": 1.111601953034178,
"grad_norm": 2.2572931093297357,
"learning_rate": 4.914634782125127e-06,
"loss": 0.4568,
"step": 796
},
{
"epoch": 1.1129969774471054,
"grad_norm": 1.962863301703952,
"learning_rate": 4.902441202067028e-06,
"loss": 0.494,
"step": 797
},
{
"epoch": 1.1143920018600326,
"grad_norm": 2.2253259846733235,
"learning_rate": 4.890248202418073e-06,
"loss": 0.5017,
"step": 798
},
{
"epoch": 1.11578702627296,
"grad_norm": 1.9807905202198288,
"learning_rate": 4.878055855718395e-06,
"loss": 0.4461,
"step": 799
},
{
"epoch": 1.117182050685887,
"grad_norm": 2.0776479254780997,
"learning_rate": 4.865864234504246e-06,
"loss": 0.424,
"step": 800
},
{
"epoch": 1.1185770750988142,
"grad_norm": 2.080677062759276,
"learning_rate": 4.853673411307564e-06,
"loss": 0.453,
"step": 801
},
{
"epoch": 1.1199720995117415,
"grad_norm": 2.0226685971439684,
"learning_rate": 4.841483458655532e-06,
"loss": 0.4705,
"step": 802
},
{
"epoch": 1.1213671239246688,
"grad_norm": 1.9781164700065137,
"learning_rate": 4.829294449070161e-06,
"loss": 0.48,
"step": 803
},
{
"epoch": 1.1227621483375958,
"grad_norm": 2.1702790078032694,
"learning_rate": 4.817106455067848e-06,
"loss": 0.4832,
"step": 804
},
{
"epoch": 1.124157172750523,
"grad_norm": 1.975642624026135,
"learning_rate": 4.804919549158951e-06,
"loss": 0.4227,
"step": 805
},
{
"epoch": 1.1255521971634503,
"grad_norm": 2.011695144907138,
"learning_rate": 4.792733803847351e-06,
"loss": 0.4225,
"step": 806
},
{
"epoch": 1.1269472215763776,
"grad_norm": 1.9998826305913646,
"learning_rate": 4.7805492916300235e-06,
"loss": 0.5039,
"step": 807
},
{
"epoch": 1.1283422459893049,
"grad_norm": 1.9496545539372845,
"learning_rate": 4.768366084996612e-06,
"loss": 0.4694,
"step": 808
},
{
"epoch": 1.1297372704022322,
"grad_norm": 2.067866668037809,
"learning_rate": 4.756184256428992e-06,
"loss": 0.4609,
"step": 809
},
{
"epoch": 1.1311322948151592,
"grad_norm": 2.237107223204764,
"learning_rate": 4.744003878400836e-06,
"loss": 0.4674,
"step": 810
},
{
"epoch": 1.1325273192280865,
"grad_norm": 1.9242050736003506,
"learning_rate": 4.731825023377191e-06,
"loss": 0.4946,
"step": 811
},
{
"epoch": 1.1339223436410137,
"grad_norm": 2.167480296385883,
"learning_rate": 4.719647763814041e-06,
"loss": 0.4624,
"step": 812
},
{
"epoch": 1.135317368053941,
"grad_norm": 1.8494731683182122,
"learning_rate": 4.707472172157882e-06,
"loss": 0.4364,
"step": 813
},
{
"epoch": 1.136712392466868,
"grad_norm": 2.0120379641996835,
"learning_rate": 4.695298320845282e-06,
"loss": 0.4794,
"step": 814
},
{
"epoch": 1.1381074168797953,
"grad_norm": 2.0857359999555722,
"learning_rate": 4.683126282302457e-06,
"loss": 0.4472,
"step": 815
},
{
"epoch": 1.1395024412927226,
"grad_norm": 2.052049295035602,
"learning_rate": 4.67095612894484e-06,
"loss": 0.4346,
"step": 816
},
{
"epoch": 1.1408974657056499,
"grad_norm": 1.973283052063979,
"learning_rate": 4.6587879331766465e-06,
"loss": 0.4249,
"step": 817
},
{
"epoch": 1.1422924901185771,
"grad_norm": 2.024286771149601,
"learning_rate": 4.646621767390444e-06,
"loss": 0.4544,
"step": 818
},
{
"epoch": 1.1436875145315044,
"grad_norm": 2.1509974878229516,
"learning_rate": 4.634457703966729e-06,
"loss": 0.4891,
"step": 819
},
{
"epoch": 1.1450825389444315,
"grad_norm": 2.2337578962453235,
"learning_rate": 4.622295815273482e-06,
"loss": 0.4643,
"step": 820
},
{
"epoch": 1.1464775633573587,
"grad_norm": 2.0861095386115505,
"learning_rate": 4.610136173665751e-06,
"loss": 0.4723,
"step": 821
},
{
"epoch": 1.147872587770286,
"grad_norm": 2.1235228114243347,
"learning_rate": 4.597978851485217e-06,
"loss": 0.4415,
"step": 822
},
{
"epoch": 1.1492676121832133,
"grad_norm": 2.1021074133377953,
"learning_rate": 4.585823921059757e-06,
"loss": 0.4446,
"step": 823
},
{
"epoch": 1.1506626365961403,
"grad_norm": 1.8410024802817688,
"learning_rate": 4.57367145470302e-06,
"loss": 0.3897,
"step": 824
},
{
"epoch": 1.1520576610090676,
"grad_norm": 2.0473061305669313,
"learning_rate": 4.561521524713998e-06,
"loss": 0.4547,
"step": 825
},
{
"epoch": 1.1534526854219949,
"grad_norm": 2.069770337790142,
"learning_rate": 4.5493742033765906e-06,
"loss": 0.524,
"step": 826
},
{
"epoch": 1.1548477098349221,
"grad_norm": 2.251908136507743,
"learning_rate": 4.537229562959178e-06,
"loss": 0.4554,
"step": 827
},
{
"epoch": 1.1562427342478494,
"grad_norm": 2.0846493954276784,
"learning_rate": 4.525087675714191e-06,
"loss": 0.4485,
"step": 828
},
{
"epoch": 1.1576377586607767,
"grad_norm": 1.9221877313185243,
"learning_rate": 4.512948613877683e-06,
"loss": 0.5125,
"step": 829
},
{
"epoch": 1.1590327830737037,
"grad_norm": 2.1932622962593773,
"learning_rate": 4.5008124496688925e-06,
"loss": 0.4417,
"step": 830
},
{
"epoch": 1.160427807486631,
"grad_norm": 2.0453310749751012,
"learning_rate": 4.488679255289829e-06,
"loss": 0.4261,
"step": 831
},
{
"epoch": 1.1618228318995583,
"grad_norm": 2.035542185164377,
"learning_rate": 4.4765491029248235e-06,
"loss": 0.4481,
"step": 832
},
{
"epoch": 1.1632178563124855,
"grad_norm": 2.1304838227271463,
"learning_rate": 4.464422064740114e-06,
"loss": 0.4402,
"step": 833
},
{
"epoch": 1.1646128807254126,
"grad_norm": 1.8356581488535975,
"learning_rate": 4.452298212883408e-06,
"loss": 0.4494,
"step": 834
},
{
"epoch": 1.1660079051383399,
"grad_norm": 2.125785922134183,
"learning_rate": 4.4401776194834615e-06,
"loss": 0.5029,
"step": 835
},
{
"epoch": 1.1674029295512671,
"grad_norm": 2.128802200938034,
"learning_rate": 4.428060356649642e-06,
"loss": 0.4615,
"step": 836
},
{
"epoch": 1.1687979539641944,
"grad_norm": 2.133593843482753,
"learning_rate": 4.4159464964715e-06,
"loss": 0.4953,
"step": 837
},
{
"epoch": 1.1701929783771217,
"grad_norm": 2.161743883645626,
"learning_rate": 4.403836111018346e-06,
"loss": 0.4484,
"step": 838
},
{
"epoch": 1.171588002790049,
"grad_norm": 2.0148537845975785,
"learning_rate": 4.391729272338817e-06,
"loss": 0.4866,
"step": 839
},
{
"epoch": 1.172983027202976,
"grad_norm": 2.1364974416474007,
"learning_rate": 4.37962605246045e-06,
"loss": 0.4636,
"step": 840
},
{
"epoch": 1.1743780516159033,
"grad_norm": 2.1431437523336956,
"learning_rate": 4.367526523389253e-06,
"loss": 0.4947,
"step": 841
},
{
"epoch": 1.1757730760288305,
"grad_norm": 2.077327008730648,
"learning_rate": 4.355430757109273e-06,
"loss": 0.4128,
"step": 842
},
{
"epoch": 1.1771681004417578,
"grad_norm": 1.921415744960682,
"learning_rate": 4.343338825582173e-06,
"loss": 0.4509,
"step": 843
},
{
"epoch": 1.1785631248546848,
"grad_norm": 2.0628543732690607,
"learning_rate": 4.331250800746803e-06,
"loss": 0.4101,
"step": 844
},
{
"epoch": 1.179958149267612,
"grad_norm": 1.933884426694881,
"learning_rate": 4.319166754518768e-06,
"loss": 0.4424,
"step": 845
},
{
"epoch": 1.1813531736805394,
"grad_norm": 2.038887875653402,
"learning_rate": 4.307086758790005e-06,
"loss": 0.4501,
"step": 846
},
{
"epoch": 1.1827481980934667,
"grad_norm": 2.0481874903667907,
"learning_rate": 4.295010885428354e-06,
"loss": 0.4788,
"step": 847
},
{
"epoch": 1.184143222506394,
"grad_norm": 2.2983450175618603,
"learning_rate": 4.282939206277126e-06,
"loss": 0.4884,
"step": 848
},
{
"epoch": 1.1855382469193212,
"grad_norm": 2.08973082875802,
"learning_rate": 4.270871793154683e-06,
"loss": 0.4314,
"step": 849
},
{
"epoch": 1.1869332713322482,
"grad_norm": 2.1033493341055727,
"learning_rate": 4.258808717854006e-06,
"loss": 0.4819,
"step": 850
},
{
"epoch": 1.1883282957451755,
"grad_norm": 1.8745501503992958,
"learning_rate": 4.2467500521422696e-06,
"loss": 0.4172,
"step": 851
},
{
"epoch": 1.1897233201581028,
"grad_norm": 2.1806886241389623,
"learning_rate": 4.234695867760412e-06,
"loss": 0.4628,
"step": 852
},
{
"epoch": 1.19111834457103,
"grad_norm": 2.0959774571958887,
"learning_rate": 4.222646236422711e-06,
"loss": 0.4887,
"step": 853
},
{
"epoch": 1.192513368983957,
"grad_norm": 2.3011996953254923,
"learning_rate": 4.21060122981636e-06,
"loss": 0.4285,
"step": 854
},
{
"epoch": 1.1939083933968844,
"grad_norm": 1.894090908268334,
"learning_rate": 4.198560919601034e-06,
"loss": 0.478,
"step": 855
},
{
"epoch": 1.1953034178098116,
"grad_norm": 2.083934832143099,
"learning_rate": 4.186525377408471e-06,
"loss": 0.4366,
"step": 856
},
{
"epoch": 1.196698442222739,
"grad_norm": 2.1053278377478977,
"learning_rate": 4.174494674842038e-06,
"loss": 0.4752,
"step": 857
},
{
"epoch": 1.1980934666356662,
"grad_norm": 2.2411830266708765,
"learning_rate": 4.162468883476319e-06,
"loss": 0.5062,
"step": 858
},
{
"epoch": 1.1994884910485935,
"grad_norm": 1.9572462330367733,
"learning_rate": 4.150448074856667e-06,
"loss": 0.4577,
"step": 859
},
{
"epoch": 1.2008835154615205,
"grad_norm": 2.3922197259695515,
"learning_rate": 4.138432320498801e-06,
"loss": 0.4888,
"step": 860
},
{
"epoch": 1.2022785398744478,
"grad_norm": 2.1857450886509295,
"learning_rate": 4.126421691888366e-06,
"loss": 0.4311,
"step": 861
},
{
"epoch": 1.203673564287375,
"grad_norm": 2.19620003939438,
"learning_rate": 4.114416260480512e-06,
"loss": 0.4957,
"step": 862
},
{
"epoch": 1.2050685887003023,
"grad_norm": 2.065478261064621,
"learning_rate": 4.102416097699471e-06,
"loss": 0.4433,
"step": 863
},
{
"epoch": 1.2064636131132294,
"grad_norm": 2.014509401367039,
"learning_rate": 4.09042127493813e-06,
"loss": 0.4428,
"step": 864
},
{
"epoch": 1.2078586375261566,
"grad_norm": 2.119056066587081,
"learning_rate": 4.078431863557605e-06,
"loss": 0.4463,
"step": 865
},
{
"epoch": 1.209253661939084,
"grad_norm": 1.86625900949575,
"learning_rate": 4.066447934886819e-06,
"loss": 0.4728,
"step": 866
},
{
"epoch": 1.2106486863520112,
"grad_norm": 2.1210628992342255,
"learning_rate": 4.054469560222077e-06,
"loss": 0.3916,
"step": 867
},
{
"epoch": 1.2120437107649384,
"grad_norm": 1.9877563081010783,
"learning_rate": 4.042496810826641e-06,
"loss": 0.4787,
"step": 868
},
{
"epoch": 1.2134387351778657,
"grad_norm": 2.2079993472035637,
"learning_rate": 4.030529757930308e-06,
"loss": 0.501,
"step": 869
},
{
"epoch": 1.2148337595907928,
"grad_norm": 2.06793106880383,
"learning_rate": 4.018568472728979e-06,
"loss": 0.4476,
"step": 870
},
{
"epoch": 1.21622878400372,
"grad_norm": 2.0422389638464127,
"learning_rate": 4.006613026384249e-06,
"loss": 0.4518,
"step": 871
},
{
"epoch": 1.2176238084166473,
"grad_norm": 2.1199886116818027,
"learning_rate": 3.9946634900229685e-06,
"loss": 0.4741,
"step": 872
},
{
"epoch": 1.2190188328295746,
"grad_norm": 2.240846596784438,
"learning_rate": 3.982719934736832e-06,
"loss": 0.4792,
"step": 873
},
{
"epoch": 1.2204138572425016,
"grad_norm": 2.130464201224374,
"learning_rate": 3.970782431581949e-06,
"loss": 0.4438,
"step": 874
},
{
"epoch": 1.221808881655429,
"grad_norm": 2.068106505450987,
"learning_rate": 3.958851051578425e-06,
"loss": 0.4474,
"step": 875
},
{
"epoch": 1.2232039060683562,
"grad_norm": 2.0406851390179894,
"learning_rate": 3.946925865709936e-06,
"loss": 0.4367,
"step": 876
},
{
"epoch": 1.2245989304812834,
"grad_norm": 2.0437012220228805,
"learning_rate": 3.935006944923304e-06,
"loss": 0.4459,
"step": 877
},
{
"epoch": 1.2259939548942107,
"grad_norm": 2.1068418581233264,
"learning_rate": 3.923094360128081e-06,
"loss": 0.4254,
"step": 878
},
{
"epoch": 1.227388979307138,
"grad_norm": 1.9577971488960282,
"learning_rate": 3.911188182196124e-06,
"loss": 0.4433,
"step": 879
},
{
"epoch": 1.228784003720065,
"grad_norm": 2.0584334663525476,
"learning_rate": 3.899288481961173e-06,
"loss": 0.4425,
"step": 880
},
{
"epoch": 1.2301790281329923,
"grad_norm": 2.0639800376154396,
"learning_rate": 3.887395330218429e-06,
"loss": 0.4802,
"step": 881
},
{
"epoch": 1.2315740525459196,
"grad_norm": 2.2113590719733445,
"learning_rate": 3.875508797724135e-06,
"loss": 0.4898,
"step": 882
},
{
"epoch": 1.2329690769588468,
"grad_norm": 1.65424993527057,
"learning_rate": 3.8636289551951545e-06,
"loss": 0.4105,
"step": 883
},
{
"epoch": 1.234364101371774,
"grad_norm": 2.0237983089059357,
"learning_rate": 3.851755873308549e-06,
"loss": 0.437,
"step": 884
},
{
"epoch": 1.2357591257847012,
"grad_norm": 2.004193875938639,
"learning_rate": 3.8398896227011604e-06,
"loss": 0.4875,
"step": 885
},
{
"epoch": 1.2371541501976284,
"grad_norm": 2.217103897736817,
"learning_rate": 3.8280302739691874e-06,
"loss": 0.4717,
"step": 886
},
{
"epoch": 1.2385491746105557,
"grad_norm": 1.9857313758078003,
"learning_rate": 3.816177897667767e-06,
"loss": 0.4129,
"step": 887
},
{
"epoch": 1.239944199023483,
"grad_norm": 2.1759441467608824,
"learning_rate": 3.8043325643105554e-06,
"loss": 0.52,
"step": 888
},
{
"epoch": 1.2413392234364102,
"grad_norm": 2.1329442000650327,
"learning_rate": 3.792494344369311e-06,
"loss": 0.4097,
"step": 889
},
{
"epoch": 1.2427342478493373,
"grad_norm": 1.965987019278061,
"learning_rate": 3.780663308273466e-06,
"loss": 0.4527,
"step": 890
},
{
"epoch": 1.2441292722622646,
"grad_norm": 2.1499821644131503,
"learning_rate": 3.7688395264097177e-06,
"loss": 0.4778,
"step": 891
},
{
"epoch": 1.2455242966751918,
"grad_norm": 2.008829269853213,
"learning_rate": 3.7570230691216055e-06,
"loss": 0.4103,
"step": 892
},
{
"epoch": 1.246919321088119,
"grad_norm": 2.2628732069215958,
"learning_rate": 3.7452140067090936e-06,
"loss": 0.4697,
"step": 893
},
{
"epoch": 1.2483143455010464,
"grad_norm": 2.144442294578986,
"learning_rate": 3.733412409428148e-06,
"loss": 0.453,
"step": 894
},
{
"epoch": 1.2497093699139734,
"grad_norm": 1.919582985300559,
"learning_rate": 3.7216183474903243e-06,
"loss": 0.4742,
"step": 895
},
{
"epoch": 1.2511043943269007,
"grad_norm": 2.2316553028401658,
"learning_rate": 3.70983189106235e-06,
"loss": 0.4898,
"step": 896
},
{
"epoch": 1.252499418739828,
"grad_norm": 2.071360003336638,
"learning_rate": 3.698053110265699e-06,
"loss": 0.4953,
"step": 897
},
{
"epoch": 1.2538944431527552,
"grad_norm": 2.156296749381877,
"learning_rate": 3.6862820751761863e-06,
"loss": 0.4176,
"step": 898
},
{
"epoch": 1.2552894675656825,
"grad_norm": 2.0123119204925417,
"learning_rate": 3.6745188558235443e-06,
"loss": 0.4788,
"step": 899
},
{
"epoch": 1.2566844919786098,
"grad_norm": 2.1269095029712024,
"learning_rate": 3.662763522191002e-06,
"loss": 0.4729,
"step": 900
},
{
"epoch": 1.2580795163915368,
"grad_norm": 2.0416038709842934,
"learning_rate": 3.6510161442148783e-06,
"loss": 0.4471,
"step": 901
},
{
"epoch": 1.259474540804464,
"grad_norm": 1.976365217621231,
"learning_rate": 3.639276791784163e-06,
"loss": 0.48,
"step": 902
},
{
"epoch": 1.2608695652173914,
"grad_norm": 2.162833974202662,
"learning_rate": 3.6275455347400952e-06,
"loss": 0.4395,
"step": 903
},
{
"epoch": 1.2622645896303184,
"grad_norm": 1.9359111623466696,
"learning_rate": 3.6158224428757538e-06,
"loss": 0.4507,
"step": 904
},
{
"epoch": 1.2636596140432457,
"grad_norm": 2.07327168371495,
"learning_rate": 3.6041075859356383e-06,
"loss": 0.4428,
"step": 905
},
{
"epoch": 1.265054638456173,
"grad_norm": 2.03931713471931,
"learning_rate": 3.592401033615259e-06,
"loss": 0.4597,
"step": 906
},
{
"epoch": 1.2664496628691002,
"grad_norm": 2.23742738842125,
"learning_rate": 3.580702855560718e-06,
"loss": 0.4982,
"step": 907
},
{
"epoch": 1.2678446872820275,
"grad_norm": 2.080685333486399,
"learning_rate": 3.5690131213682943e-06,
"loss": 0.4595,
"step": 908
},
{
"epoch": 1.2692397116949548,
"grad_norm": 2.0899692520393414,
"learning_rate": 3.5573319005840363e-06,
"loss": 0.4602,
"step": 909
},
{
"epoch": 1.270634736107882,
"grad_norm": 2.223490524510679,
"learning_rate": 3.5456592627033355e-06,
"loss": 0.4424,
"step": 910
},
{
"epoch": 1.272029760520809,
"grad_norm": 1.8639269617638456,
"learning_rate": 3.533995277170532e-06,
"loss": 0.4628,
"step": 911
},
{
"epoch": 1.2734247849337363,
"grad_norm": 2.093793164050054,
"learning_rate": 3.52234001337848e-06,
"loss": 0.4527,
"step": 912
},
{
"epoch": 1.2748198093466636,
"grad_norm": 2.1299607415176567,
"learning_rate": 3.510693540668151e-06,
"loss": 0.4349,
"step": 913
},
{
"epoch": 1.2762148337595907,
"grad_norm": 1.9228204200175336,
"learning_rate": 3.4990559283282133e-06,
"loss": 0.4353,
"step": 914
},
{
"epoch": 1.277609858172518,
"grad_norm": 2.0259963322884396,
"learning_rate": 3.4874272455946217e-06,
"loss": 0.4902,
"step": 915
},
{
"epoch": 1.2790048825854452,
"grad_norm": 2.0655847930615767,
"learning_rate": 3.4758075616502075e-06,
"loss": 0.4263,
"step": 916
},
{
"epoch": 1.2803999069983725,
"grad_norm": 2.0685690620076094,
"learning_rate": 3.4641969456242646e-06,
"loss": 0.4662,
"step": 917
},
{
"epoch": 1.2817949314112997,
"grad_norm": 2.1387591609101873,
"learning_rate": 3.4525954665921375e-06,
"loss": 0.444,
"step": 918
},
{
"epoch": 1.283189955824227,
"grad_norm": 2.0161975874514617,
"learning_rate": 3.4410031935748124e-06,
"loss": 0.4242,
"step": 919
},
{
"epoch": 1.2845849802371543,
"grad_norm": 2.084336680851645,
"learning_rate": 3.4294201955385086e-06,
"loss": 0.4702,
"step": 920
},
{
"epoch": 1.2859800046500813,
"grad_norm": 2.1036740916228442,
"learning_rate": 3.4178465413942628e-06,
"loss": 0.4808,
"step": 921
},
{
"epoch": 1.2873750290630086,
"grad_norm": 2.086226145543498,
"learning_rate": 3.406282299997521e-06,
"loss": 0.3895,
"step": 922
},
{
"epoch": 1.2887700534759359,
"grad_norm": 1.9685735915085314,
"learning_rate": 3.394727540147732e-06,
"loss": 0.4099,
"step": 923
},
{
"epoch": 1.290165077888863,
"grad_norm": 2.0192812446036035,
"learning_rate": 3.383182330587936e-06,
"loss": 0.4561,
"step": 924
},
{
"epoch": 1.2915601023017902,
"grad_norm": 2.0644595686003804,
"learning_rate": 3.3716467400043514e-06,
"loss": 0.4478,
"step": 925
},
{
"epoch": 1.2929551267147175,
"grad_norm": 2.1694861941362245,
"learning_rate": 3.360120837025974e-06,
"loss": 0.4327,
"step": 926
},
{
"epoch": 1.2943501511276447,
"grad_norm": 1.9412400050349086,
"learning_rate": 3.3486046902241663e-06,
"loss": 0.458,
"step": 927
},
{
"epoch": 1.295745175540572,
"grad_norm": 2.1349425480132815,
"learning_rate": 3.3370983681122405e-06,
"loss": 0.4278,
"step": 928
},
{
"epoch": 1.2971401999534993,
"grad_norm": 1.9769480906565091,
"learning_rate": 3.3256019391450696e-06,
"loss": 0.4708,
"step": 929
},
{
"epoch": 1.2985352243664265,
"grad_norm": 2.1617930261944416,
"learning_rate": 3.3141154717186597e-06,
"loss": 0.4791,
"step": 930
},
{
"epoch": 1.2999302487793536,
"grad_norm": 2.233031327126289,
"learning_rate": 3.302639034169758e-06,
"loss": 0.5107,
"step": 931
},
{
"epoch": 1.3013252731922809,
"grad_norm": 2.2333091138236982,
"learning_rate": 3.291172694775435e-06,
"loss": 0.4696,
"step": 932
},
{
"epoch": 1.3027202976052081,
"grad_norm": 1.9897849895374509,
"learning_rate": 3.2797165217526915e-06,
"loss": 0.4319,
"step": 933
},
{
"epoch": 1.3041153220181352,
"grad_norm": 2.1259453438639264,
"learning_rate": 3.2682705832580395e-06,
"loss": 0.4821,
"step": 934
},
{
"epoch": 1.3055103464310625,
"grad_norm": 2.1982207926862127,
"learning_rate": 3.2568349473871044e-06,
"loss": 0.4613,
"step": 935
},
{
"epoch": 1.3069053708439897,
"grad_norm": 1.943157323921308,
"learning_rate": 3.245409682174217e-06,
"loss": 0.4727,
"step": 936
},
{
"epoch": 1.308300395256917,
"grad_norm": 2.1694502623538816,
"learning_rate": 3.2339948555920103e-06,
"loss": 0.4462,
"step": 937
},
{
"epoch": 1.3096954196698443,
"grad_norm": 2.02441875061226,
"learning_rate": 3.222590535551017e-06,
"loss": 0.3991,
"step": 938
},
{
"epoch": 1.3110904440827715,
"grad_norm": 2.009032349306495,
"learning_rate": 3.2111967898992597e-06,
"loss": 0.4724,
"step": 939
},
{
"epoch": 1.3124854684956988,
"grad_norm": 1.9843180264245779,
"learning_rate": 3.1998136864218497e-06,
"loss": 0.4452,
"step": 940
},
{
"epoch": 1.3138804929086259,
"grad_norm": 2.0533825746187806,
"learning_rate": 3.188441292840587e-06,
"loss": 0.4482,
"step": 941
},
{
"epoch": 1.3152755173215531,
"grad_norm": 2.0235013309402468,
"learning_rate": 3.177079676813557e-06,
"loss": 0.4561,
"step": 942
},
{
"epoch": 1.3166705417344804,
"grad_norm": 2.1137423321444317,
"learning_rate": 3.1657289059347184e-06,
"loss": 0.4492,
"step": 943
},
{
"epoch": 1.3180655661474074,
"grad_norm": 1.7486020284932557,
"learning_rate": 3.1543890477335153e-06,
"loss": 0.5007,
"step": 944
},
{
"epoch": 1.3194605905603347,
"grad_norm": 2.156564229441565,
"learning_rate": 3.1430601696744676e-06,
"loss": 0.48,
"step": 945
},
{
"epoch": 1.320855614973262,
"grad_norm": 2.098327227462188,
"learning_rate": 3.131742339156768e-06,
"loss": 0.4515,
"step": 946
},
{
"epoch": 1.3222506393861893,
"grad_norm": 2.107981352881033,
"learning_rate": 3.120435623513887e-06,
"loss": 0.5015,
"step": 947
},
{
"epoch": 1.3236456637991165,
"grad_norm": 2.3095062879021766,
"learning_rate": 3.1091400900131662e-06,
"loss": 0.5035,
"step": 948
},
{
"epoch": 1.3250406882120438,
"grad_norm": 2.161655104957988,
"learning_rate": 3.0978558058554255e-06,
"loss": 0.4702,
"step": 949
},
{
"epoch": 1.326435712624971,
"grad_norm": 2.1231878542949265,
"learning_rate": 3.0865828381745515e-06,
"loss": 0.441,
"step": 950
},
{
"epoch": 1.3278307370378981,
"grad_norm": 2.0409247579695995,
"learning_rate": 3.0753212540371126e-06,
"loss": 0.4371,
"step": 951
},
{
"epoch": 1.3292257614508254,
"grad_norm": 1.972768288020109,
"learning_rate": 3.0640711204419495e-06,
"loss": 0.4775,
"step": 952
},
{
"epoch": 1.3306207858637527,
"grad_norm": 2.0235557868912215,
"learning_rate": 3.0528325043197786e-06,
"loss": 0.424,
"step": 953
},
{
"epoch": 1.33201581027668,
"grad_norm": 2.1375250223189672,
"learning_rate": 3.0416054725327975e-06,
"loss": 0.4883,
"step": 954
},
{
"epoch": 1.333410834689607,
"grad_norm": 2.2291824467539985,
"learning_rate": 3.0303900918742832e-06,
"loss": 0.4629,
"step": 955
},
{
"epoch": 1.3348058591025342,
"grad_norm": 2.113413333177263,
"learning_rate": 3.0191864290682006e-06,
"loss": 0.4602,
"step": 956
},
{
"epoch": 1.3362008835154615,
"grad_norm": 2.1716284589357,
"learning_rate": 3.007994550768793e-06,
"loss": 0.4649,
"step": 957
},
{
"epoch": 1.3375959079283888,
"grad_norm": 1.931534866293287,
"learning_rate": 2.9968145235602014e-06,
"loss": 0.4473,
"step": 958
},
{
"epoch": 1.338990932341316,
"grad_norm": 2.0364674719890785,
"learning_rate": 2.9856464139560583e-06,
"loss": 0.4246,
"step": 959
},
{
"epoch": 1.3403859567542433,
"grad_norm": 2.108866570837351,
"learning_rate": 2.9744902883990923e-06,
"loss": 0.4365,
"step": 960
},
{
"epoch": 1.3417809811671704,
"grad_norm": 2.127000885263901,
"learning_rate": 2.963346213260737e-06,
"loss": 0.4443,
"step": 961
},
{
"epoch": 1.3431760055800976,
"grad_norm": 2.0526296607029324,
"learning_rate": 2.9522142548407366e-06,
"loss": 0.4588,
"step": 962
},
{
"epoch": 1.344571029993025,
"grad_norm": 2.241123744052401,
"learning_rate": 2.941094479366744e-06,
"loss": 0.4697,
"step": 963
},
{
"epoch": 1.3459660544059522,
"grad_norm": 2.132358400432893,
"learning_rate": 2.929986952993933e-06,
"loss": 0.4187,
"step": 964
},
{
"epoch": 1.3473610788188792,
"grad_norm": 1.8629316651965908,
"learning_rate": 2.918891741804607e-06,
"loss": 0.5213,
"step": 965
},
{
"epoch": 1.3487561032318065,
"grad_norm": 2.1685409605601227,
"learning_rate": 2.9078089118077994e-06,
"loss": 0.4231,
"step": 966
},
{
"epoch": 1.3501511276447338,
"grad_norm": 2.209136940528215,
"learning_rate": 2.896738528938885e-06,
"loss": 0.4455,
"step": 967
},
{
"epoch": 1.351546152057661,
"grad_norm": 2.139570622351985,
"learning_rate": 2.885680659059187e-06,
"loss": 0.4334,
"step": 968
},
{
"epoch": 1.3529411764705883,
"grad_norm": 2.033270980494118,
"learning_rate": 2.8746353679555794e-06,
"loss": 0.4568,
"step": 969
},
{
"epoch": 1.3543362008835156,
"grad_norm": 2.0919421978668056,
"learning_rate": 2.863602721340109e-06,
"loss": 0.4561,
"step": 970
},
{
"epoch": 1.3557312252964426,
"grad_norm": 1.9545589493519657,
"learning_rate": 2.8525827848495912e-06,
"loss": 0.4344,
"step": 971
},
{
"epoch": 1.35712624970937,
"grad_norm": 2.023560734048173,
"learning_rate": 2.8415756240452274e-06,
"loss": 0.4436,
"step": 972
},
{
"epoch": 1.3585212741222972,
"grad_norm": 2.0340775030640708,
"learning_rate": 2.83058130441221e-06,
"loss": 0.4297,
"step": 973
},
{
"epoch": 1.3599162985352244,
"grad_norm": 2.173615317885776,
"learning_rate": 2.819599891359337e-06,
"loss": 0.4252,
"step": 974
},
{
"epoch": 1.3613113229481515,
"grad_norm": 1.9284825554902276,
"learning_rate": 2.8086314502186197e-06,
"loss": 0.4155,
"step": 975
},
{
"epoch": 1.3627063473610788,
"grad_norm": 2.141689581871054,
"learning_rate": 2.7976760462448955e-06,
"loss": 0.465,
"step": 976
},
{
"epoch": 1.364101371774006,
"grad_norm": 2.114769495652363,
"learning_rate": 2.7867337446154397e-06,
"loss": 0.5002,
"step": 977
},
{
"epoch": 1.3654963961869333,
"grad_norm": 2.2678110505319933,
"learning_rate": 2.77580461042958e-06,
"loss": 0.4671,
"step": 978
},
{
"epoch": 1.3668914205998606,
"grad_norm": 1.987667346423969,
"learning_rate": 2.7648887087082996e-06,
"loss": 0.4685,
"step": 979
},
{
"epoch": 1.3682864450127878,
"grad_norm": 2.24277483709307,
"learning_rate": 2.753986104393864e-06,
"loss": 0.4481,
"step": 980
},
{
"epoch": 1.369681469425715,
"grad_norm": 2.2378577673040048,
"learning_rate": 2.743096862349427e-06,
"loss": 0.441,
"step": 981
},
{
"epoch": 1.3710764938386422,
"grad_norm": 2.021640316592768,
"learning_rate": 2.7322210473586406e-06,
"loss": 0.437,
"step": 982
},
{
"epoch": 1.3724715182515694,
"grad_norm": 1.8635324388592955,
"learning_rate": 2.721358724125287e-06,
"loss": 0.4605,
"step": 983
},
{
"epoch": 1.3738665426644967,
"grad_norm": 3.7251213547992688,
"learning_rate": 2.7105099572728676e-06,
"loss": 0.4779,
"step": 984
},
{
"epoch": 1.3752615670774238,
"grad_norm": 2.084169605706041,
"learning_rate": 2.6996748113442397e-06,
"loss": 0.4533,
"step": 985
},
{
"epoch": 1.376656591490351,
"grad_norm": 2.2104579294138604,
"learning_rate": 2.6888533508012247e-06,
"loss": 0.4565,
"step": 986
},
{
"epoch": 1.3780516159032783,
"grad_norm": 2.1567174351782614,
"learning_rate": 2.678045640024225e-06,
"loss": 0.49,
"step": 987
},
{
"epoch": 1.3794466403162056,
"grad_norm": 2.0531595116227557,
"learning_rate": 2.66725174331184e-06,
"loss": 0.403,
"step": 988
},
{
"epoch": 1.3808416647291328,
"grad_norm": 2.080232972878775,
"learning_rate": 2.6564717248804876e-06,
"loss": 0.4926,
"step": 989
},
{
"epoch": 1.38223668914206,
"grad_norm": 2.0993530101540094,
"learning_rate": 2.6457056488640135e-06,
"loss": 0.4605,
"step": 990
},
{
"epoch": 1.3836317135549872,
"grad_norm": 2.230646736451978,
"learning_rate": 2.6349535793133196e-06,
"loss": 0.4612,
"step": 991
},
{
"epoch": 1.3850267379679144,
"grad_norm": 2.08748461860507,
"learning_rate": 2.6242155801959814e-06,
"loss": 0.4747,
"step": 992
},
{
"epoch": 1.3864217623808417,
"grad_norm": 2.169585558390855,
"learning_rate": 2.613491715395861e-06,
"loss": 0.4435,
"step": 993
},
{
"epoch": 1.387816786793769,
"grad_norm": 2.0019666409817574,
"learning_rate": 2.602782048712729e-06,
"loss": 0.4115,
"step": 994
},
{
"epoch": 1.389211811206696,
"grad_norm": 1.9749791762887468,
"learning_rate": 2.592086643861891e-06,
"loss": 0.4732,
"step": 995
},
{
"epoch": 1.3906068356196233,
"grad_norm": 2.174738415153292,
"learning_rate": 2.5814055644738013e-06,
"loss": 0.4541,
"step": 996
},
{
"epoch": 1.3920018600325506,
"grad_norm": 2.139559543105941,
"learning_rate": 2.57073887409369e-06,
"loss": 0.4509,
"step": 997
},
{
"epoch": 1.3933968844454778,
"grad_norm": 2.045345170723802,
"learning_rate": 2.5600866361811804e-06,
"loss": 0.4342,
"step": 998
},
{
"epoch": 1.394791908858405,
"grad_norm": 2.1774650094819257,
"learning_rate": 2.5494489141099155e-06,
"loss": 0.4762,
"step": 999
},
{
"epoch": 1.3961869332713324,
"grad_norm": 2.188733000241817,
"learning_rate": 2.5388257711671723e-06,
"loss": 0.4656,
"step": 1000
},
{
"epoch": 1.3975819576842594,
"grad_norm": 1.9431028809211057,
"learning_rate": 2.5282172705535013e-06,
"loss": 0.4199,
"step": 1001
},
{
"epoch": 1.3989769820971867,
"grad_norm": 1.9956348364331262,
"learning_rate": 2.517623475382335e-06,
"loss": 0.4673,
"step": 1002
},
{
"epoch": 1.400372006510114,
"grad_norm": 2.1848282860521824,
"learning_rate": 2.507044448679621e-06,
"loss": 0.4533,
"step": 1003
},
{
"epoch": 1.4017670309230412,
"grad_norm": 2.0497329348777837,
"learning_rate": 2.4964802533834404e-06,
"loss": 0.4676,
"step": 1004
},
{
"epoch": 1.4031620553359683,
"grad_norm": 2.1022873543390106,
"learning_rate": 2.4859309523436415e-06,
"loss": 0.4498,
"step": 1005
},
{
"epoch": 1.4045570797488955,
"grad_norm": 2.1374560724246114,
"learning_rate": 2.4753966083214613e-06,
"loss": 0.4202,
"step": 1006
},
{
"epoch": 1.4059521041618228,
"grad_norm": 1.994795529357376,
"learning_rate": 2.4648772839891517e-06,
"loss": 0.405,
"step": 1007
},
{
"epoch": 1.40734712857475,
"grad_norm": 1.889213983531933,
"learning_rate": 2.454373041929607e-06,
"loss": 0.4043,
"step": 1008
},
{
"epoch": 1.4087421529876774,
"grad_norm": 1.888123546815583,
"learning_rate": 2.4438839446359936e-06,
"loss": 0.4349,
"step": 1009
},
{
"epoch": 1.4101371774006046,
"grad_norm": 2.1442846126168105,
"learning_rate": 2.433410054511376e-06,
"loss": 0.4897,
"step": 1010
},
{
"epoch": 1.4115322018135317,
"grad_norm": 2.098646009752216,
"learning_rate": 2.422951433868346e-06,
"loss": 0.4814,
"step": 1011
},
{
"epoch": 1.412927226226459,
"grad_norm": 2.0907240577366433,
"learning_rate": 2.4125081449286523e-06,
"loss": 0.4949,
"step": 1012
},
{
"epoch": 1.4143222506393862,
"grad_norm": 2.1264766644102595,
"learning_rate": 2.4020802498228333e-06,
"loss": 0.4435,
"step": 1013
},
{
"epoch": 1.4157172750523135,
"grad_norm": 2.161970375783907,
"learning_rate": 2.3916678105898376e-06,
"loss": 0.4446,
"step": 1014
},
{
"epoch": 1.4171122994652405,
"grad_norm": 2.0357769161075954,
"learning_rate": 2.3812708891766683e-06,
"loss": 0.4516,
"step": 1015
},
{
"epoch": 1.4185073238781678,
"grad_norm": 2.2463213912233613,
"learning_rate": 2.3708895474380045e-06,
"loss": 0.4481,
"step": 1016
},
{
"epoch": 1.419902348291095,
"grad_norm": 2.23141097622481,
"learning_rate": 2.360523847135838e-06,
"loss": 0.4469,
"step": 1017
},
{
"epoch": 1.4212973727040223,
"grad_norm": 2.019771445051018,
"learning_rate": 2.3501738499391053e-06,
"loss": 0.4235,
"step": 1018
},
{
"epoch": 1.4226923971169496,
"grad_norm": 2.0632278153556776,
"learning_rate": 2.339839617423318e-06,
"loss": 0.465,
"step": 1019
},
{
"epoch": 1.4240874215298769,
"grad_norm": 2.0613929308269867,
"learning_rate": 2.3295212110701994e-06,
"loss": 0.3962,
"step": 1020
},
{
"epoch": 1.425482445942804,
"grad_norm": 1.9328040306708631,
"learning_rate": 2.3192186922673187e-06,
"loss": 0.434,
"step": 1021
},
{
"epoch": 1.4268774703557312,
"grad_norm": 2.1639316185817496,
"learning_rate": 2.308932122307724e-06,
"loss": 0.5227,
"step": 1022
},
{
"epoch": 1.4282724947686585,
"grad_norm": 2.0899973628027007,
"learning_rate": 2.2986615623895774e-06,
"loss": 0.4453,
"step": 1023
},
{
"epoch": 1.4296675191815857,
"grad_norm": 2.051643921559896,
"learning_rate": 2.288407073615797e-06,
"loss": 0.4385,
"step": 1024
},
{
"epoch": 1.4310625435945128,
"grad_norm": 2.134955583289714,
"learning_rate": 2.2781687169936794e-06,
"loss": 0.3805,
"step": 1025
},
{
"epoch": 1.43245756800744,
"grad_norm": 1.957462775697652,
"learning_rate": 2.2679465534345534e-06,
"loss": 0.4605,
"step": 1026
},
{
"epoch": 1.4338525924203673,
"grad_norm": 2.0642360916463924,
"learning_rate": 2.2577406437534055e-06,
"loss": 0.4511,
"step": 1027
},
{
"epoch": 1.4352476168332946,
"grad_norm": 2.007231951364962,
"learning_rate": 2.247551048668531e-06,
"loss": 0.4527,
"step": 1028
},
{
"epoch": 1.4366426412462219,
"grad_norm": 2.138332772618615,
"learning_rate": 2.2373778288011517e-06,
"loss": 0.4552,
"step": 1029
},
{
"epoch": 1.4380376656591491,
"grad_norm": 2.124748566211008,
"learning_rate": 2.2272210446750757e-06,
"loss": 0.4635,
"step": 1030
},
{
"epoch": 1.4394326900720762,
"grad_norm": 2.170811329651528,
"learning_rate": 2.2170807567163294e-06,
"loss": 0.4623,
"step": 1031
},
{
"epoch": 1.4408277144850035,
"grad_norm": 2.090856244557684,
"learning_rate": 2.2069570252527954e-06,
"loss": 0.4357,
"step": 1032
},
{
"epoch": 1.4422227388979307,
"grad_norm": 2.3690024054005154,
"learning_rate": 2.196849910513858e-06,
"loss": 0.4715,
"step": 1033
},
{
"epoch": 1.443617763310858,
"grad_norm": 2.059481032162157,
"learning_rate": 2.186759472630045e-06,
"loss": 0.469,
"step": 1034
},
{
"epoch": 1.445012787723785,
"grad_norm": 2.2361040736935465,
"learning_rate": 2.17668577163266e-06,
"loss": 0.46,
"step": 1035
},
{
"epoch": 1.4464078121367123,
"grad_norm": 2.164364979964796,
"learning_rate": 2.1666288674534446e-06,
"loss": 0.446,
"step": 1036
},
{
"epoch": 1.4478028365496396,
"grad_norm": 1.8634866672332282,
"learning_rate": 2.156588819924204e-06,
"loss": 0.4549,
"step": 1037
},
{
"epoch": 1.4491978609625669,
"grad_norm": 2.132439121162979,
"learning_rate": 2.1465656887764615e-06,
"loss": 0.4222,
"step": 1038
},
{
"epoch": 1.4505928853754941,
"grad_norm": 2.163017453224282,
"learning_rate": 2.136559533641092e-06,
"loss": 0.4677,
"step": 1039
},
{
"epoch": 1.4519879097884214,
"grad_norm": 2.1316582396669386,
"learning_rate": 2.126570414047982e-06,
"loss": 0.4784,
"step": 1040
},
{
"epoch": 1.4533829342013485,
"grad_norm": 2.1775898180475215,
"learning_rate": 2.1165983894256647e-06,
"loss": 0.47,
"step": 1041
},
{
"epoch": 1.4547779586142757,
"grad_norm": 2.204489717101063,
"learning_rate": 2.1066435191009717e-06,
"loss": 0.4415,
"step": 1042
},
{
"epoch": 1.456172983027203,
"grad_norm": 2.0391115202528023,
"learning_rate": 2.096705862298676e-06,
"loss": 0.4437,
"step": 1043
},
{
"epoch": 1.4575680074401303,
"grad_norm": 2.150122752289656,
"learning_rate": 2.086785478141143e-06,
"loss": 0.4328,
"step": 1044
},
{
"epoch": 1.4589630318530573,
"grad_norm": 2.1245697032921833,
"learning_rate": 2.076882425647977e-06,
"loss": 0.3992,
"step": 1045
},
{
"epoch": 1.4603580562659846,
"grad_norm": 2.0584854403080874,
"learning_rate": 2.0669967637356702e-06,
"loss": 0.4427,
"step": 1046
},
{
"epoch": 1.4617530806789119,
"grad_norm": 1.9826831338383193,
"learning_rate": 2.057128551217254e-06,
"loss": 0.4774,
"step": 1047
},
{
"epoch": 1.4631481050918391,
"grad_norm": 2.1289243452321847,
"learning_rate": 2.0472778468019456e-06,
"loss": 0.497,
"step": 1048
},
{
"epoch": 1.4645431295047664,
"grad_norm": 2.1286358432747376,
"learning_rate": 2.037444709094804e-06,
"loss": 0.4321,
"step": 1049
},
{
"epoch": 1.4659381539176937,
"grad_norm": 2.054601532030644,
"learning_rate": 2.027629196596373e-06,
"loss": 0.4763,
"step": 1050
},
{
"epoch": 1.4673331783306207,
"grad_norm": 2.1232951836859106,
"learning_rate": 2.0178313677023427e-06,
"loss": 0.483,
"step": 1051
},
{
"epoch": 1.468728202743548,
"grad_norm": 2.110872679736881,
"learning_rate": 2.008051280703196e-06,
"loss": 0.4284,
"step": 1052
},
{
"epoch": 1.4701232271564753,
"grad_norm": 2.039901992775588,
"learning_rate": 1.9982889937838635e-06,
"loss": 0.4295,
"step": 1053
},
{
"epoch": 1.4715182515694025,
"grad_norm": 1.9424993858352442,
"learning_rate": 1.988544565023378e-06,
"loss": 0.4417,
"step": 1054
},
{
"epoch": 1.4729132759823296,
"grad_norm": 2.2251949175021157,
"learning_rate": 1.978818052394528e-06,
"loss": 0.4641,
"step": 1055
},
{
"epoch": 1.4743083003952568,
"grad_norm": 1.9524088974321652,
"learning_rate": 1.9691095137635117e-06,
"loss": 0.4909,
"step": 1056
},
{
"epoch": 1.4757033248081841,
"grad_norm": 2.320431539339629,
"learning_rate": 1.9594190068895967e-06,
"loss": 0.4686,
"step": 1057
},
{
"epoch": 1.4770983492211114,
"grad_norm": 2.1811531311597174,
"learning_rate": 1.949746589424772e-06,
"loss": 0.4381,
"step": 1058
},
{
"epoch": 1.4784933736340387,
"grad_norm": 2.0675702062294703,
"learning_rate": 1.940092318913408e-06,
"loss": 0.4801,
"step": 1059
},
{
"epoch": 1.479888398046966,
"grad_norm": 2.140807400840788,
"learning_rate": 1.9304562527919094e-06,
"loss": 0.4569,
"step": 1060
},
{
"epoch": 1.481283422459893,
"grad_norm": 2.1799124450514173,
"learning_rate": 1.920838448388382e-06,
"loss": 0.432,
"step": 1061
},
{
"epoch": 1.4826784468728202,
"grad_norm": 1.893690087500473,
"learning_rate": 1.911238962922282e-06,
"loss": 0.427,
"step": 1062
},
{
"epoch": 1.4840734712857475,
"grad_norm": 2.1124227125467336,
"learning_rate": 1.901657853504088e-06,
"loss": 0.4683,
"step": 1063
},
{
"epoch": 1.4854684956986748,
"grad_norm": 2.116545768275609,
"learning_rate": 1.8920951771349428e-06,
"loss": 0.4389,
"step": 1064
},
{
"epoch": 1.4868635201116018,
"grad_norm": 2.0960807190104473,
"learning_rate": 1.8825509907063328e-06,
"loss": 0.4192,
"step": 1065
},
{
"epoch": 1.488258544524529,
"grad_norm": 2.099536710624629,
"learning_rate": 1.8730253509997388e-06,
"loss": 0.4088,
"step": 1066
},
{
"epoch": 1.4896535689374564,
"grad_norm": 1.9858631220209877,
"learning_rate": 1.8635183146863016e-06,
"loss": 0.4277,
"step": 1067
},
{
"epoch": 1.4910485933503836,
"grad_norm": 2.170988774620348,
"learning_rate": 1.8540299383264843e-06,
"loss": 0.4559,
"step": 1068
},
{
"epoch": 1.492443617763311,
"grad_norm": 2.1787387202817015,
"learning_rate": 1.8445602783697375e-06,
"loss": 0.455,
"step": 1069
},
{
"epoch": 1.4938386421762382,
"grad_norm": 2.1599938341615985,
"learning_rate": 1.8351093911541573e-06,
"loss": 0.5288,
"step": 1070
},
{
"epoch": 1.4952336665891652,
"grad_norm": 2.32780588654961,
"learning_rate": 1.8256773329061566e-06,
"loss": 0.4552,
"step": 1071
},
{
"epoch": 1.4966286910020925,
"grad_norm": 2.149694858376219,
"learning_rate": 1.8162641597401338e-06,
"loss": 0.4689,
"step": 1072
},
{
"epoch": 1.4980237154150198,
"grad_norm": 2.0819138266881225,
"learning_rate": 1.8068699276581286e-06,
"loss": 0.4683,
"step": 1073
},
{
"epoch": 1.499418739827947,
"grad_norm": 2.2817200224362235,
"learning_rate": 1.7974946925494925e-06,
"loss": 0.5152,
"step": 1074
},
{
"epoch": 1.500813764240874,
"grad_norm": 2.203558953898251,
"learning_rate": 1.7881385101905613e-06,
"loss": 0.4737,
"step": 1075
},
{
"epoch": 1.5022087886538014,
"grad_norm": 2.0908558844435756,
"learning_rate": 1.778801436244319e-06,
"loss": 0.4283,
"step": 1076
},
{
"epoch": 1.5036038130667286,
"grad_norm": 2.1315489899020963,
"learning_rate": 1.7694835262600668e-06,
"loss": 0.4837,
"step": 1077
},
{
"epoch": 1.504998837479656,
"grad_norm": 2.1634782853778884,
"learning_rate": 1.7601848356730933e-06,
"loss": 0.4467,
"step": 1078
},
{
"epoch": 1.5063938618925832,
"grad_norm": 2.154791342432924,
"learning_rate": 1.7509054198043473e-06,
"loss": 0.417,
"step": 1079
},
{
"epoch": 1.5077888863055104,
"grad_norm": 2.0140475517719167,
"learning_rate": 1.741645333860098e-06,
"loss": 0.464,
"step": 1080
},
{
"epoch": 1.5091839107184377,
"grad_norm": 2.1573584679806372,
"learning_rate": 1.7324046329316253e-06,
"loss": 0.4372,
"step": 1081
},
{
"epoch": 1.5105789351313648,
"grad_norm": 1.9559737789586726,
"learning_rate": 1.7231833719948743e-06,
"loss": 0.4413,
"step": 1082
},
{
"epoch": 1.511973959544292,
"grad_norm": 2.1157743977756405,
"learning_rate": 1.7139816059101372e-06,
"loss": 0.4098,
"step": 1083
},
{
"epoch": 1.5133689839572193,
"grad_norm": 1.9673036282895517,
"learning_rate": 1.7047993894217269e-06,
"loss": 0.4533,
"step": 1084
},
{
"epoch": 1.5147640083701464,
"grad_norm": 2.196912391461416,
"learning_rate": 1.6956367771576432e-06,
"loss": 0.4704,
"step": 1085
},
{
"epoch": 1.5161590327830736,
"grad_norm": 2.0199973612735387,
"learning_rate": 1.6864938236292606e-06,
"loss": 0.4383,
"step": 1086
},
{
"epoch": 1.517554057196001,
"grad_norm": 2.300672182405597,
"learning_rate": 1.6773705832309945e-06,
"loss": 0.4511,
"step": 1087
},
{
"epoch": 1.5189490816089282,
"grad_norm": 2.14852844655263,
"learning_rate": 1.6682671102399806e-06,
"loss": 0.4466,
"step": 1088
},
{
"epoch": 1.5203441060218554,
"grad_norm": 2.1361355389334378,
"learning_rate": 1.6591834588157523e-06,
"loss": 0.4548,
"step": 1089
},
{
"epoch": 1.5217391304347827,
"grad_norm": 2.0851465686341086,
"learning_rate": 1.6501196829999179e-06,
"loss": 0.501,
"step": 1090
},
{
"epoch": 1.52313415484771,
"grad_norm": 2.229088020773415,
"learning_rate": 1.6410758367158386e-06,
"loss": 0.4214,
"step": 1091
},
{
"epoch": 1.524529179260637,
"grad_norm": 1.9656859079499656,
"learning_rate": 1.6320519737683095e-06,
"loss": 0.394,
"step": 1092
},
{
"epoch": 1.5259242036735643,
"grad_norm": 1.8733906967388423,
"learning_rate": 1.6230481478432364e-06,
"loss": 0.4564,
"step": 1093
},
{
"epoch": 1.5273192280864916,
"grad_norm": 2.0441924506755664,
"learning_rate": 1.6140644125073223e-06,
"loss": 0.449,
"step": 1094
},
{
"epoch": 1.5287142524994186,
"grad_norm": 2.0385854309420317,
"learning_rate": 1.6051008212077384e-06,
"loss": 0.4797,
"step": 1095
},
{
"epoch": 1.5301092769123459,
"grad_norm": 2.3564266457176855,
"learning_rate": 1.5961574272718179e-06,
"loss": 0.4768,
"step": 1096
},
{
"epoch": 1.5315043013252732,
"grad_norm": 1.9115979450281055,
"learning_rate": 1.5872342839067305e-06,
"loss": 0.458,
"step": 1097
},
{
"epoch": 1.5328993257382004,
"grad_norm": 2.1808017439500027,
"learning_rate": 1.5783314441991698e-06,
"loss": 0.4233,
"step": 1098
},
{
"epoch": 1.5342943501511277,
"grad_norm": 2.076982481552849,
"learning_rate": 1.569448961115037e-06,
"loss": 0.4696,
"step": 1099
},
{
"epoch": 1.535689374564055,
"grad_norm": 2.2277775247007585,
"learning_rate": 1.5605868874991225e-06,
"loss": 0.4672,
"step": 1100
},
{
"epoch": 1.5370843989769822,
"grad_norm": 2.029428608776388,
"learning_rate": 1.5517452760747975e-06,
"loss": 0.4391,
"step": 1101
},
{
"epoch": 1.5384794233899093,
"grad_norm": 2.163702572810171,
"learning_rate": 1.5429241794436939e-06,
"loss": 0.4615,
"step": 1102
},
{
"epoch": 1.5398744478028366,
"grad_norm": 2.0576551741631546,
"learning_rate": 1.5341236500853963e-06,
"loss": 0.3963,
"step": 1103
},
{
"epoch": 1.5412694722157638,
"grad_norm": 1.948897872545703,
"learning_rate": 1.525343740357128e-06,
"loss": 0.4433,
"step": 1104
},
{
"epoch": 1.5426644966286909,
"grad_norm": 2.194109971817455,
"learning_rate": 1.5165845024934366e-06,
"loss": 0.4713,
"step": 1105
},
{
"epoch": 1.5440595210416181,
"grad_norm": 2.1942802527235767,
"learning_rate": 1.5078459886058894e-06,
"loss": 0.3669,
"step": 1106
},
{
"epoch": 1.5454545454545454,
"grad_norm": 1.931807338090762,
"learning_rate": 1.499128250682757e-06,
"loss": 0.4165,
"step": 1107
},
{
"epoch": 1.5468495698674727,
"grad_norm": 1.8873433846497065,
"learning_rate": 1.4904313405887111e-06,
"loss": 0.4119,
"step": 1108
},
{
"epoch": 1.5482445942804,
"grad_norm": 2.1562216951827913,
"learning_rate": 1.4817553100645105e-06,
"loss": 0.4601,
"step": 1109
},
{
"epoch": 1.5496396186933272,
"grad_norm": 2.1807467873254116,
"learning_rate": 1.4731002107266902e-06,
"loss": 0.4231,
"step": 1110
},
{
"epoch": 1.5510346431062545,
"grad_norm": 1.981889187995434,
"learning_rate": 1.4644660940672628e-06,
"loss": 0.494,
"step": 1111
},
{
"epoch": 1.5524296675191815,
"grad_norm": 2.5915146018778614,
"learning_rate": 1.455853011453408e-06,
"loss": 0.4647,
"step": 1112
},
{
"epoch": 1.5538246919321088,
"grad_norm": 2.2136885014943677,
"learning_rate": 1.447261014127167e-06,
"loss": 0.4374,
"step": 1113
},
{
"epoch": 1.555219716345036,
"grad_norm": 2.107779035217476,
"learning_rate": 1.4386901532051379e-06,
"loss": 0.4575,
"step": 1114
},
{
"epoch": 1.5566147407579631,
"grad_norm": 2.0020086351509088,
"learning_rate": 1.430140479678168e-06,
"loss": 0.4751,
"step": 1115
},
{
"epoch": 1.5580097651708904,
"grad_norm": 2.2610539554238613,
"learning_rate": 1.4216120444110566e-06,
"loss": 0.4146,
"step": 1116
},
{
"epoch": 1.5594047895838177,
"grad_norm": 2.127211740430317,
"learning_rate": 1.4131048981422529e-06,
"loss": 0.413,
"step": 1117
},
{
"epoch": 1.560799813996745,
"grad_norm": 2.0574224448728016,
"learning_rate": 1.404619091483546e-06,
"loss": 0.4128,
"step": 1118
},
{
"epoch": 1.5621948384096722,
"grad_norm": 2.0228379456275216,
"learning_rate": 1.3961546749197719e-06,
"loss": 0.4266,
"step": 1119
},
{
"epoch": 1.5635898628225995,
"grad_norm": 2.079227910587265,
"learning_rate": 1.387711698808505e-06,
"loss": 0.428,
"step": 1120
},
{
"epoch": 1.5649848872355268,
"grad_norm": 1.8923261271608611,
"learning_rate": 1.3792902133797692e-06,
"loss": 0.4121,
"step": 1121
},
{
"epoch": 1.5663799116484538,
"grad_norm": 1.9647313130229616,
"learning_rate": 1.37089026873573e-06,
"loss": 0.4266,
"step": 1122
},
{
"epoch": 1.567774936061381,
"grad_norm": 2.1049543470618515,
"learning_rate": 1.3625119148504013e-06,
"loss": 0.4255,
"step": 1123
},
{
"epoch": 1.5691699604743083,
"grad_norm": 2.018946253171041,
"learning_rate": 1.354155201569346e-06,
"loss": 0.4374,
"step": 1124
},
{
"epoch": 1.5705649848872354,
"grad_norm": 2.0421990909718595,
"learning_rate": 1.3458201786093795e-06,
"loss": 0.4513,
"step": 1125
},
{
"epoch": 1.5719600093001627,
"grad_norm": 2.0403374058085433,
"learning_rate": 1.3375068955582754e-06,
"loss": 0.4068,
"step": 1126
},
{
"epoch": 1.57335503371309,
"grad_norm": 1.8850939356587402,
"learning_rate": 1.329215401874468e-06,
"loss": 0.4581,
"step": 1127
},
{
"epoch": 1.5747500581260172,
"grad_norm": 2.0113778332021313,
"learning_rate": 1.3209457468867614e-06,
"loss": 0.4111,
"step": 1128
},
{
"epoch": 1.5761450825389445,
"grad_norm": 1.961215646510735,
"learning_rate": 1.3126979797940336e-06,
"loss": 0.4016,
"step": 1129
},
{
"epoch": 1.5775401069518717,
"grad_norm": 1.946752479107875,
"learning_rate": 1.3044721496649427e-06,
"loss": 0.4439,
"step": 1130
},
{
"epoch": 1.578935131364799,
"grad_norm": 1.959475188059055,
"learning_rate": 1.2962683054376375e-06,
"loss": 0.4077,
"step": 1131
},
{
"epoch": 1.580330155777726,
"grad_norm": 2.030187772323639,
"learning_rate": 1.2880864959194666e-06,
"loss": 0.3907,
"step": 1132
},
{
"epoch": 1.5817251801906533,
"grad_norm": 1.9957922442663245,
"learning_rate": 1.2799267697866868e-06,
"loss": 0.463,
"step": 1133
},
{
"epoch": 1.5831202046035806,
"grad_norm": 2.0602462684084624,
"learning_rate": 1.2717891755841722e-06,
"loss": 0.4058,
"step": 1134
},
{
"epoch": 1.5845152290165077,
"grad_norm": 2.1830055682068195,
"learning_rate": 1.2636737617251283e-06,
"loss": 0.4462,
"step": 1135
},
{
"epoch": 1.585910253429435,
"grad_norm": 2.0532749033801814,
"learning_rate": 1.255580576490802e-06,
"loss": 0.4569,
"step": 1136
},
{
"epoch": 1.5873052778423622,
"grad_norm": 2.1155124216443473,
"learning_rate": 1.2475096680301934e-06,
"loss": 0.4173,
"step": 1137
},
{
"epoch": 1.5887003022552895,
"grad_norm": 2.1690763886161015,
"learning_rate": 1.2394610843597737e-06,
"loss": 0.4598,
"step": 1138
},
{
"epoch": 1.5900953266682167,
"grad_norm": 2.073486351346297,
"learning_rate": 1.2314348733631958e-06,
"loss": 0.4099,
"step": 1139
},
{
"epoch": 1.591490351081144,
"grad_norm": 2.009538340855588,
"learning_rate": 1.2234310827910063e-06,
"loss": 0.4634,
"step": 1140
},
{
"epoch": 1.5928853754940713,
"grad_norm": 2.086735738787941,
"learning_rate": 1.2154497602603704e-06,
"loss": 0.4501,
"step": 1141
},
{
"epoch": 1.5942803999069983,
"grad_norm": 2.124804280878044,
"learning_rate": 1.2074909532547824e-06,
"loss": 0.4075,
"step": 1142
},
{
"epoch": 1.5956754243199256,
"grad_norm": 2.1372643069283517,
"learning_rate": 1.1995547091237814e-06,
"loss": 0.4584,
"step": 1143
},
{
"epoch": 1.5970704487328529,
"grad_norm": 2.115705750764563,
"learning_rate": 1.191641075082679e-06,
"loss": 0.3771,
"step": 1144
},
{
"epoch": 1.59846547314578,
"grad_norm": 2.1054204705622426,
"learning_rate": 1.1837500982122646e-06,
"loss": 0.4701,
"step": 1145
},
{
"epoch": 1.5998604975587072,
"grad_norm": 2.1677272810674224,
"learning_rate": 1.175881825458537e-06,
"loss": 0.4798,
"step": 1146
},
{
"epoch": 1.6012555219716345,
"grad_norm": 2.103897443365682,
"learning_rate": 1.16803630363242e-06,
"loss": 0.4168,
"step": 1147
},
{
"epoch": 1.6026505463845617,
"grad_norm": 2.21888945610227,
"learning_rate": 1.1602135794094843e-06,
"loss": 0.4548,
"step": 1148
},
{
"epoch": 1.604045570797489,
"grad_norm": 2.157370730462611,
"learning_rate": 1.152413699329671e-06,
"loss": 0.458,
"step": 1149
},
{
"epoch": 1.6054405952104163,
"grad_norm": 2.1179984402535528,
"learning_rate": 1.1446367097970134e-06,
"loss": 0.4581,
"step": 1150
},
{
"epoch": 1.6068356196233435,
"grad_norm": 2.286665408259628,
"learning_rate": 1.1368826570793602e-06,
"loss": 0.4452,
"step": 1151
},
{
"epoch": 1.6082306440362706,
"grad_norm": 2.211199331004034,
"learning_rate": 1.129151587308102e-06,
"loss": 0.4097,
"step": 1152
},
{
"epoch": 1.6096256684491979,
"grad_norm": 1.8853710993731516,
"learning_rate": 1.1214435464779006e-06,
"loss": 0.4815,
"step": 1153
},
{
"epoch": 1.6110206928621251,
"grad_norm": 2.162910909135421,
"learning_rate": 1.1137585804464062e-06,
"loss": 0.4402,
"step": 1154
},
{
"epoch": 1.6124157172750522,
"grad_norm": 2.119846894032719,
"learning_rate": 1.1060967349339897e-06,
"loss": 0.4085,
"step": 1155
},
{
"epoch": 1.6138107416879794,
"grad_norm": 1.870879315183828,
"learning_rate": 1.098458055523472e-06,
"loss": 0.4787,
"step": 1156
},
{
"epoch": 1.6152057661009067,
"grad_norm": 2.12494088689906,
"learning_rate": 1.0908425876598512e-06,
"loss": 0.4754,
"step": 1157
},
{
"epoch": 1.616600790513834,
"grad_norm": 2.063935560635135,
"learning_rate": 1.0832503766500324e-06,
"loss": 0.4139,
"step": 1158
},
{
"epoch": 1.6179958149267613,
"grad_norm": 2.0787433691392745,
"learning_rate": 1.0756814676625576e-06,
"loss": 0.4386,
"step": 1159
},
{
"epoch": 1.6193908393396885,
"grad_norm": 1.9939683928369583,
"learning_rate": 1.0681359057273388e-06,
"loss": 0.4471,
"step": 1160
},
{
"epoch": 1.6207858637526158,
"grad_norm": 2.133799960540424,
"learning_rate": 1.060613735735384e-06,
"loss": 0.4595,
"step": 1161
},
{
"epoch": 1.6221808881655428,
"grad_norm": 2.024216295186773,
"learning_rate": 1.0531150024385423e-06,
"loss": 0.4525,
"step": 1162
},
{
"epoch": 1.6235759125784701,
"grad_norm": 2.0095052935864315,
"learning_rate": 1.045639750449225e-06,
"loss": 0.474,
"step": 1163
},
{
"epoch": 1.6249709369913974,
"grad_norm": 2.1117515259793893,
"learning_rate": 1.0381880242401483e-06,
"loss": 0.3937,
"step": 1164
},
{
"epoch": 1.6263659614043244,
"grad_norm": 1.9720718120560756,
"learning_rate": 1.0307598681440617e-06,
"loss": 0.4304,
"step": 1165
},
{
"epoch": 1.6277609858172517,
"grad_norm": 2.0114538495387606,
"learning_rate": 1.0233553263534924e-06,
"loss": 0.4348,
"step": 1166
},
{
"epoch": 1.629156010230179,
"grad_norm": 2.0352256936639663,
"learning_rate": 1.0159744429204776e-06,
"loss": 0.4358,
"step": 1167
},
{
"epoch": 1.6305510346431062,
"grad_norm": 2.203732854508469,
"learning_rate": 1.0086172617563022e-06,
"loss": 0.4501,
"step": 1168
},
{
"epoch": 1.6319460590560335,
"grad_norm": 2.0446799273020564,
"learning_rate": 1.0012838266312397e-06,
"loss": 0.4431,
"step": 1169
},
{
"epoch": 1.6333410834689608,
"grad_norm": 2.2327287396202173,
"learning_rate": 9.939741811742893e-07,
"loss": 0.4403,
"step": 1170
},
{
"epoch": 1.634736107881888,
"grad_norm": 2.2165342815886233,
"learning_rate": 9.866883688729189e-07,
"loss": 0.4378,
"step": 1171
},
{
"epoch": 1.636131132294815,
"grad_norm": 2.0549633529337936,
"learning_rate": 9.794264330728043e-07,
"loss": 0.4093,
"step": 1172
},
{
"epoch": 1.6375261567077424,
"grad_norm": 2.0537102551655257,
"learning_rate": 9.721884169775735e-07,
"loss": 0.4272,
"step": 1173
},
{
"epoch": 1.6389211811206696,
"grad_norm": 2.200656487770776,
"learning_rate": 9.64974363648548e-07,
"loss": 0.4869,
"step": 1174
},
{
"epoch": 1.6403162055335967,
"grad_norm": 2.016301063721608,
"learning_rate": 9.577843160044842e-07,
"loss": 0.4439,
"step": 1175
},
{
"epoch": 1.641711229946524,
"grad_norm": 2.1709862159822215,
"learning_rate": 9.506183168213245e-07,
"loss": 0.4464,
"step": 1176
},
{
"epoch": 1.6431062543594512,
"grad_norm": 2.0994595042624296,
"learning_rate": 9.434764087319376e-07,
"loss": 0.4349,
"step": 1177
},
{
"epoch": 1.6445012787723785,
"grad_norm": 2.157447833178887,
"learning_rate": 9.363586342258662e-07,
"loss": 0.4461,
"step": 1178
},
{
"epoch": 1.6458963031853058,
"grad_norm": 2.267561948229981,
"learning_rate": 9.292650356490757e-07,
"loss": 0.4379,
"step": 1179
},
{
"epoch": 1.647291327598233,
"grad_norm": 2.0427902212170435,
"learning_rate": 9.221956552036992e-07,
"loss": 0.4314,
"step": 1180
},
{
"epoch": 1.6486863520111603,
"grad_norm": 2.1552102034976603,
"learning_rate": 9.151505349477901e-07,
"loss": 0.4602,
"step": 1181
},
{
"epoch": 1.6500813764240876,
"grad_norm": 2.185995467962314,
"learning_rate": 9.081297167950682e-07,
"loss": 0.4774,
"step": 1182
},
{
"epoch": 1.6514764008370146,
"grad_norm": 2.1492175403491203,
"learning_rate": 9.011332425146734e-07,
"loss": 0.4464,
"step": 1183
},
{
"epoch": 1.652871425249942,
"grad_norm": 2.009528313205518,
"learning_rate": 8.941611537309158e-07,
"loss": 0.4675,
"step": 1184
},
{
"epoch": 1.654266449662869,
"grad_norm": 2.215590896357922,
"learning_rate": 8.87213491923029e-07,
"loss": 0.4243,
"step": 1185
},
{
"epoch": 1.6556614740757962,
"grad_norm": 2.165672694359002,
"learning_rate": 8.80290298424919e-07,
"loss": 0.4333,
"step": 1186
},
{
"epoch": 1.6570564984887235,
"grad_norm": 2.044698721900824,
"learning_rate": 8.733916144249255e-07,
"loss": 0.4859,
"step": 1187
},
{
"epoch": 1.6584515229016508,
"grad_norm": 2.390370448435757,
"learning_rate": 8.665174809655707e-07,
"loss": 0.4776,
"step": 1188
},
{
"epoch": 1.659846547314578,
"grad_norm": 2.0991424672781753,
"learning_rate": 8.596679389433232e-07,
"loss": 0.3906,
"step": 1189
},
{
"epoch": 1.6612415717275053,
"grad_norm": 1.9493079845533017,
"learning_rate": 8.528430291083412e-07,
"loss": 0.4128,
"step": 1190
},
{
"epoch": 1.6626365961404326,
"grad_norm": 1.9928338170064632,
"learning_rate": 8.460427920642422e-07,
"loss": 0.4332,
"step": 1191
},
{
"epoch": 1.6640316205533598,
"grad_norm": 2.206404643941403,
"learning_rate": 8.392672682678577e-07,
"loss": 0.4433,
"step": 1192
},
{
"epoch": 1.665426644966287,
"grad_norm": 2.056639857598042,
"learning_rate": 8.325164980289896e-07,
"loss": 0.4962,
"step": 1193
},
{
"epoch": 1.6668216693792142,
"grad_norm": 2.2133368692574584,
"learning_rate": 8.257905215101758e-07,
"loss": 0.4062,
"step": 1194
},
{
"epoch": 1.6682166937921412,
"grad_norm": 1.9230071533267274,
"learning_rate": 8.19089378726447e-07,
"loss": 0.4495,
"step": 1195
},
{
"epoch": 1.6696117182050685,
"grad_norm": 2.1334286794624453,
"learning_rate": 8.12413109545086e-07,
"loss": 0.4735,
"step": 1196
},
{
"epoch": 1.6710067426179958,
"grad_norm": 2.1036187662888644,
"learning_rate": 8.057617536854012e-07,
"loss": 0.4017,
"step": 1197
},
{
"epoch": 1.672401767030923,
"grad_norm": 2.006107617620194,
"learning_rate": 7.991353507184801e-07,
"loss": 0.4349,
"step": 1198
},
{
"epoch": 1.6737967914438503,
"grad_norm": 2.051201227177932,
"learning_rate": 7.92533940066958e-07,
"loss": 0.469,
"step": 1199
},
{
"epoch": 1.6751918158567776,
"grad_norm": 2.1517754801345506,
"learning_rate": 7.859575610047798e-07,
"loss": 0.4495,
"step": 1200
},
{
"epoch": 1.6765868402697048,
"grad_norm": 2.1279072064457942,
"learning_rate": 7.794062526569735e-07,
"loss": 0.4779,
"step": 1201
},
{
"epoch": 1.677981864682632,
"grad_norm": 2.4080169749454297,
"learning_rate": 7.728800539994113e-07,
"loss": 0.4934,
"step": 1202
},
{
"epoch": 1.6793768890955592,
"grad_norm": 2.250925372186841,
"learning_rate": 7.663790038585794e-07,
"loss": 0.4334,
"step": 1203
},
{
"epoch": 1.6807719135084864,
"grad_norm": 2.1745930093080013,
"learning_rate": 7.599031409113483e-07,
"loss": 0.4884,
"step": 1204
},
{
"epoch": 1.6821669379214135,
"grad_norm": 2.0042102092337033,
"learning_rate": 7.534525036847406e-07,
"loss": 0.4345,
"step": 1205
},
{
"epoch": 1.6835619623343407,
"grad_norm": 2.157617727203113,
"learning_rate": 7.470271305557036e-07,
"loss": 0.4684,
"step": 1206
},
{
"epoch": 1.684956986747268,
"grad_norm": 2.228687769508677,
"learning_rate": 7.406270597508796e-07,
"loss": 0.4522,
"step": 1207
},
{
"epoch": 1.6863520111601953,
"grad_norm": 1.9744508349204999,
"learning_rate": 7.342523293463799e-07,
"loss": 0.4471,
"step": 1208
},
{
"epoch": 1.6877470355731226,
"grad_norm": 2.0654991204899287,
"learning_rate": 7.279029772675572e-07,
"loss": 0.4699,
"step": 1209
},
{
"epoch": 1.6891420599860498,
"grad_norm": 2.3077221294043673,
"learning_rate": 7.215790412887802e-07,
"loss": 0.4537,
"step": 1210
},
{
"epoch": 1.690537084398977,
"grad_norm": 2.0916281069067812,
"learning_rate": 7.152805590332079e-07,
"loss": 0.4089,
"step": 1211
},
{
"epoch": 1.6919321088119044,
"grad_norm": 1.9806178247028858,
"learning_rate": 7.090075679725683e-07,
"loss": 0.4576,
"step": 1212
},
{
"epoch": 1.6933271332248314,
"grad_norm": 2.2405052747072753,
"learning_rate": 7.027601054269329e-07,
"loss": 0.4222,
"step": 1213
},
{
"epoch": 1.6947221576377587,
"grad_norm": 2.175268576390091,
"learning_rate": 6.965382085644968e-07,
"loss": 0.4908,
"step": 1214
},
{
"epoch": 1.6961171820506857,
"grad_norm": 2.2521989456802243,
"learning_rate": 6.903419144013556e-07,
"loss": 0.4417,
"step": 1215
},
{
"epoch": 1.697512206463613,
"grad_norm": 2.0189939596511572,
"learning_rate": 6.841712598012867e-07,
"loss": 0.408,
"step": 1216
},
{
"epoch": 1.6989072308765403,
"grad_norm": 2.046600993493337,
"learning_rate": 6.780262814755284e-07,
"loss": 0.4148,
"step": 1217
},
{
"epoch": 1.7003022552894675,
"grad_norm": 2.125874995200457,
"learning_rate": 6.719070159825642e-07,
"loss": 0.4612,
"step": 1218
},
{
"epoch": 1.7016972797023948,
"grad_norm": 2.231403630718439,
"learning_rate": 6.65813499727902e-07,
"loss": 0.5169,
"step": 1219
},
{
"epoch": 1.703092304115322,
"grad_norm": 2.3095426463200734,
"learning_rate": 6.597457689638598e-07,
"loss": 0.4311,
"step": 1220
},
{
"epoch": 1.7044873285282494,
"grad_norm": 1.8931375132713402,
"learning_rate": 6.53703859789348e-07,
"loss": 0.4401,
"step": 1221
},
{
"epoch": 1.7058823529411766,
"grad_norm": 2.1226595938221156,
"learning_rate": 6.476878081496579e-07,
"loss": 0.4084,
"step": 1222
},
{
"epoch": 1.7072773773541037,
"grad_norm": 1.963273731879721,
"learning_rate": 6.416976498362432e-07,
"loss": 0.4444,
"step": 1223
},
{
"epoch": 1.708672401767031,
"grad_norm": 1.9577598278955202,
"learning_rate": 6.357334204865156e-07,
"loss": 0.4296,
"step": 1224
},
{
"epoch": 1.710067426179958,
"grad_norm": 2.094472599509459,
"learning_rate": 6.29795155583619e-07,
"loss": 0.4681,
"step": 1225
},
{
"epoch": 1.7114624505928853,
"grad_norm": 2.1531471300188723,
"learning_rate": 6.238828904562316e-07,
"loss": 0.4148,
"step": 1226
},
{
"epoch": 1.7128574750058125,
"grad_norm": 2.155551596673569,
"learning_rate": 6.179966602783493e-07,
"loss": 0.4311,
"step": 1227
},
{
"epoch": 1.7142524994187398,
"grad_norm": 2.0177115106660124,
"learning_rate": 6.121365000690766e-07,
"loss": 0.4598,
"step": 1228
},
{
"epoch": 1.715647523831667,
"grad_norm": 2.1963675547860166,
"learning_rate": 6.063024446924198e-07,
"loss": 0.4421,
"step": 1229
},
{
"epoch": 1.7170425482445943,
"grad_norm": 2.0991295738118843,
"learning_rate": 6.004945288570813e-07,
"loss": 0.4881,
"step": 1230
},
{
"epoch": 1.7184375726575216,
"grad_norm": 2.2657270322156275,
"learning_rate": 5.947127871162455e-07,
"loss": 0.4533,
"step": 1231
},
{
"epoch": 1.7198325970704489,
"grad_norm": 2.0317554787541385,
"learning_rate": 5.889572538673826e-07,
"loss": 0.4133,
"step": 1232
},
{
"epoch": 1.721227621483376,
"grad_norm": 2.181744908024485,
"learning_rate": 5.832279633520411e-07,
"loss": 0.4401,
"step": 1233
},
{
"epoch": 1.7226226458963032,
"grad_norm": 2.0739837998198767,
"learning_rate": 5.775249496556406e-07,
"loss": 0.4857,
"step": 1234
},
{
"epoch": 1.7240176703092303,
"grad_norm": 2.1598290833897575,
"learning_rate": 5.718482467072695e-07,
"loss": 0.4431,
"step": 1235
},
{
"epoch": 1.7254126947221575,
"grad_norm": 2.1157200647986643,
"learning_rate": 5.661978882794883e-07,
"loss": 0.4292,
"step": 1236
},
{
"epoch": 1.7268077191350848,
"grad_norm": 2.1415635571508482,
"learning_rate": 5.60573907988124e-07,
"loss": 0.436,
"step": 1237
},
{
"epoch": 1.728202743548012,
"grad_norm": 2.1240606052923767,
"learning_rate": 5.54976339292072e-07,
"loss": 0.4379,
"step": 1238
},
{
"epoch": 1.7295977679609393,
"grad_norm": 2.0868385511709686,
"learning_rate": 5.494052154930956e-07,
"loss": 0.5001,
"step": 1239
},
{
"epoch": 1.7309927923738666,
"grad_norm": 2.109154503050066,
"learning_rate": 5.438605697356309e-07,
"loss": 0.4067,
"step": 1240
},
{
"epoch": 1.7323878167867939,
"grad_norm": 1.9920843328096538,
"learning_rate": 5.383424350065825e-07,
"loss": 0.4641,
"step": 1241
},
{
"epoch": 1.7337828411997211,
"grad_norm": 2.0920755339457116,
"learning_rate": 5.328508441351382e-07,
"loss": 0.4055,
"step": 1242
},
{
"epoch": 1.7351778656126482,
"grad_norm": 2.139045511713234,
"learning_rate": 5.273858297925649e-07,
"loss": 0.4874,
"step": 1243
},
{
"epoch": 1.7365728900255755,
"grad_norm": 2.123440484636655,
"learning_rate": 5.219474244920164e-07,
"loss": 0.4354,
"step": 1244
},
{
"epoch": 1.7379679144385025,
"grad_norm": 2.1771103649698045,
"learning_rate": 5.165356605883432e-07,
"loss": 0.4151,
"step": 1245
},
{
"epoch": 1.7393629388514298,
"grad_norm": 2.1281745438674404,
"learning_rate": 5.111505702778935e-07,
"loss": 0.4411,
"step": 1246
},
{
"epoch": 1.740757963264357,
"grad_norm": 1.9713491600361932,
"learning_rate": 5.057921855983288e-07,
"loss": 0.3942,
"step": 1247
},
{
"epoch": 1.7421529876772843,
"grad_norm": 2.044049483117014,
"learning_rate": 5.004605384284295e-07,
"loss": 0.4238,
"step": 1248
},
{
"epoch": 1.7435480120902116,
"grad_norm": 2.0491078987608837,
"learning_rate": 4.951556604879049e-07,
"loss": 0.4466,
"step": 1249
},
{
"epoch": 1.7449430365031389,
"grad_norm": 2.0685568916077965,
"learning_rate": 4.898775833372055e-07,
"loss": 0.43,
"step": 1250
},
{
"epoch": 1.7463380609160661,
"grad_norm": 2.2550610819046044,
"learning_rate": 4.846263383773364e-07,
"loss": 0.4687,
"step": 1251
},
{
"epoch": 1.7477330853289934,
"grad_norm": 2.1006531168627323,
"learning_rate": 4.794019568496683e-07,
"loss": 0.4575,
"step": 1252
},
{
"epoch": 1.7491281097419205,
"grad_norm": 2.1037421201005064,
"learning_rate": 4.7420446983575286e-07,
"loss": 0.4148,
"step": 1253
},
{
"epoch": 1.7505231341548477,
"grad_norm": 2.191926859388401,
"learning_rate": 4.690339082571366e-07,
"loss": 0.4326,
"step": 1254
},
{
"epoch": 1.7519181585677748,
"grad_norm": 1.967041448878273,
"learning_rate": 4.638903028751796e-07,
"loss": 0.4645,
"step": 1255
},
{
"epoch": 1.753313182980702,
"grad_norm": 2.1553532027152826,
"learning_rate": 4.5877368429086857e-07,
"loss": 0.3935,
"step": 1256
},
{
"epoch": 1.7547082073936293,
"grad_norm": 2.0123299142859303,
"learning_rate": 4.5368408294463873e-07,
"loss": 0.4569,
"step": 1257
},
{
"epoch": 1.7561032318065566,
"grad_norm": 2.1934970199026176,
"learning_rate": 4.486215291161894e-07,
"loss": 0.4907,
"step": 1258
},
{
"epoch": 1.7574982562194839,
"grad_norm": 2.151146138020403,
"learning_rate": 4.4358605292430743e-07,
"loss": 0.4436,
"step": 1259
},
{
"epoch": 1.7588932806324111,
"grad_norm": 2.2698087475883293,
"learning_rate": 4.385776843266842e-07,
"loss": 0.4743,
"step": 1260
},
{
"epoch": 1.7602883050453384,
"grad_norm": 2.268096663009867,
"learning_rate": 4.335964531197401e-07,
"loss": 0.4965,
"step": 1261
},
{
"epoch": 1.7616833294582657,
"grad_norm": 2.133924279141534,
"learning_rate": 4.286423889384456e-07,
"loss": 0.4296,
"step": 1262
},
{
"epoch": 1.7630783538711927,
"grad_norm": 2.1019100692382913,
"learning_rate": 4.2371552125614614e-07,
"loss": 0.4642,
"step": 1263
},
{
"epoch": 1.76447337828412,
"grad_norm": 2.128549712904592,
"learning_rate": 4.1881587938438606e-07,
"loss": 0.4555,
"step": 1264
},
{
"epoch": 1.765868402697047,
"grad_norm": 2.0813994372296705,
"learning_rate": 4.139434924727359e-07,
"loss": 0.4449,
"step": 1265
},
{
"epoch": 1.7672634271099743,
"grad_norm": 2.133944501402786,
"learning_rate": 4.090983895086137e-07,
"loss": 0.4128,
"step": 1266
},
{
"epoch": 1.7686584515229016,
"grad_norm": 2.103631001632505,
"learning_rate": 4.0428059931712116e-07,
"loss": 0.4334,
"step": 1267
},
{
"epoch": 1.7700534759358288,
"grad_norm": 2.066477638022606,
"learning_rate": 3.994901505608628e-07,
"loss": 0.436,
"step": 1268
},
{
"epoch": 1.7714485003487561,
"grad_norm": 2.194181829948266,
"learning_rate": 3.94727071739785e-07,
"loss": 0.4376,
"step": 1269
},
{
"epoch": 1.7728435247616834,
"grad_norm": 2.0725658101591806,
"learning_rate": 3.899913911909986e-07,
"loss": 0.4075,
"step": 1270
},
{
"epoch": 1.7742385491746107,
"grad_norm": 1.9580748592599533,
"learning_rate": 3.8528313708861173e-07,
"loss": 0.4488,
"step": 1271
},
{
"epoch": 1.775633573587538,
"grad_norm": 2.261238952398715,
"learning_rate": 3.8060233744356634e-07,
"loss": 0.461,
"step": 1272
},
{
"epoch": 1.777028598000465,
"grad_norm": 2.2710082458435155,
"learning_rate": 3.759490201034677e-07,
"loss": 0.4508,
"step": 1273
},
{
"epoch": 1.7784236224133922,
"grad_norm": 1.9789141505409449,
"learning_rate": 3.7132321275241966e-07,
"loss": 0.453,
"step": 1274
},
{
"epoch": 1.7798186468263193,
"grad_norm": 2.2126452553672475,
"learning_rate": 3.6672494291086156e-07,
"loss": 0.4554,
"step": 1275
},
{
"epoch": 1.7812136712392466,
"grad_norm": 1.9936682658714493,
"learning_rate": 3.6215423793540036e-07,
"loss": 0.4386,
"step": 1276
},
{
"epoch": 1.7826086956521738,
"grad_norm": 2.184960862216077,
"learning_rate": 3.576111250186526e-07,
"loss": 0.4894,
"step": 1277
},
{
"epoch": 1.784003720065101,
"grad_norm": 2.2698572395940553,
"learning_rate": 3.530956311890815e-07,
"loss": 0.4553,
"step": 1278
},
{
"epoch": 1.7853987444780284,
"grad_norm": 1.9219568300840275,
"learning_rate": 3.486077833108342e-07,
"loss": 0.414,
"step": 1279
},
{
"epoch": 1.7867937688909556,
"grad_norm": 2.1229280934149712,
"learning_rate": 3.441476080835826e-07,
"loss": 0.4794,
"step": 1280
},
{
"epoch": 1.788188793303883,
"grad_norm": 2.078715218304194,
"learning_rate": 3.397151320423647e-07,
"loss": 0.4163,
"step": 1281
},
{
"epoch": 1.7895838177168102,
"grad_norm": 2.2246453997801328,
"learning_rate": 3.353103815574282e-07,
"loss": 0.4323,
"step": 1282
},
{
"epoch": 1.7909788421297372,
"grad_norm": 2.2738150018866463,
"learning_rate": 3.309333828340722e-07,
"loss": 0.4629,
"step": 1283
},
{
"epoch": 1.7923738665426645,
"grad_norm": 2.2022495508806004,
"learning_rate": 3.2658416191249045e-07,
"loss": 0.4476,
"step": 1284
},
{
"epoch": 1.7937688909555918,
"grad_norm": 2.079838874938454,
"learning_rate": 3.2226274466761906e-07,
"loss": 0.4355,
"step": 1285
},
{
"epoch": 1.7951639153685188,
"grad_norm": 2.1980853821963273,
"learning_rate": 3.179691568089799e-07,
"loss": 0.4844,
"step": 1286
},
{
"epoch": 1.796558939781446,
"grad_norm": 2.104507635473491,
"learning_rate": 3.137034238805292e-07,
"loss": 0.4641,
"step": 1287
},
{
"epoch": 1.7979539641943734,
"grad_norm": 2.1609985762509174,
"learning_rate": 3.0946557126050435e-07,
"loss": 0.4353,
"step": 1288
},
{
"epoch": 1.7993489886073006,
"grad_norm": 2.1808994040352903,
"learning_rate": 3.0525562416127473e-07,
"loss": 0.4474,
"step": 1289
},
{
"epoch": 1.800744013020228,
"grad_norm": 2.089132424900587,
"learning_rate": 3.010736076291909e-07,
"loss": 0.3931,
"step": 1290
},
{
"epoch": 1.8021390374331552,
"grad_norm": 1.9289962008895816,
"learning_rate": 2.9691954654443353e-07,
"loss": 0.4305,
"step": 1291
},
{
"epoch": 1.8035340618460824,
"grad_norm": 2.0842770826082386,
"learning_rate": 2.927934656208681e-07,
"loss": 0.4049,
"step": 1292
},
{
"epoch": 1.8049290862590095,
"grad_norm": 2.0122506975099927,
"learning_rate": 2.88695389405898e-07,
"loss": 0.4269,
"step": 1293
},
{
"epoch": 1.8063241106719368,
"grad_norm": 2.063376075115741,
"learning_rate": 2.8462534228031636e-07,
"loss": 0.4754,
"step": 1294
},
{
"epoch": 1.807719135084864,
"grad_norm": 2.167029255388058,
"learning_rate": 2.8058334845816214e-07,
"loss": 0.5274,
"step": 1295
},
{
"epoch": 1.809114159497791,
"grad_norm": 2.24210170332635,
"learning_rate": 2.765694319865775e-07,
"loss": 0.4648,
"step": 1296
},
{
"epoch": 1.8105091839107184,
"grad_norm": 2.285072471770254,
"learning_rate": 2.725836167456608e-07,
"loss": 0.4601,
"step": 1297
},
{
"epoch": 1.8119042083236456,
"grad_norm": 1.8382699307655532,
"learning_rate": 2.6862592644832875e-07,
"loss": 0.385,
"step": 1298
},
{
"epoch": 1.813299232736573,
"grad_norm": 2.0290063843910424,
"learning_rate": 2.646963846401729e-07,
"loss": 0.451,
"step": 1299
},
{
"epoch": 1.8146942571495002,
"grad_norm": 2.012644575855064,
"learning_rate": 2.6079501469932154e-07,
"loss": 0.4129,
"step": 1300
},
{
"epoch": 1.8160892815624274,
"grad_norm": 2.0525851758055444,
"learning_rate": 2.569218398362971e-07,
"loss": 0.4314,
"step": 1301
},
{
"epoch": 1.8174843059753547,
"grad_norm": 1.9495707816729695,
"learning_rate": 2.530768830938818e-07,
"loss": 0.4533,
"step": 1302
},
{
"epoch": 1.8188793303882818,
"grad_norm": 2.215837334297456,
"learning_rate": 2.49260167346978e-07,
"loss": 0.3923,
"step": 1303
},
{
"epoch": 1.820274354801209,
"grad_norm": 1.834187483715057,
"learning_rate": 2.4547171530247437e-07,
"loss": 0.4466,
"step": 1304
},
{
"epoch": 1.8216693792141363,
"grad_norm": 1.9292756082760862,
"learning_rate": 2.417115494991107e-07,
"loss": 0.4087,
"step": 1305
},
{
"epoch": 1.8230644036270633,
"grad_norm": 2.0829059283179583,
"learning_rate": 2.3797969230733853e-07,
"loss": 0.4821,
"step": 1306
},
{
"epoch": 1.8244594280399906,
"grad_norm": 2.269315325383343,
"learning_rate": 2.3427616592919587e-07,
"loss": 0.5061,
"step": 1307
},
{
"epoch": 1.8258544524529179,
"grad_norm": 2.231129621626873,
"learning_rate": 2.3060099239817059e-07,
"loss": 0.4408,
"step": 1308
},
{
"epoch": 1.8272494768658452,
"grad_norm": 2.036459196177826,
"learning_rate": 2.2695419357906978e-07,
"loss": 0.443,
"step": 1309
},
{
"epoch": 1.8286445012787724,
"grad_norm": 2.142821360259979,
"learning_rate": 2.2333579116788962e-07,
"loss": 0.4114,
"step": 1310
},
{
"epoch": 1.8300395256916997,
"grad_norm": 2.3397446775684276,
"learning_rate": 2.197458066916891e-07,
"loss": 0.483,
"step": 1311
},
{
"epoch": 1.831434550104627,
"grad_norm": 1.6794181268501756,
"learning_rate": 2.161842615084564e-07,
"loss": 0.442,
"step": 1312
},
{
"epoch": 1.832829574517554,
"grad_norm": 1.9926444296965193,
"learning_rate": 2.1265117680698665e-07,
"loss": 0.4395,
"step": 1313
},
{
"epoch": 1.8342245989304813,
"grad_norm": 2.205617293991677,
"learning_rate": 2.0914657360675606e-07,
"loss": 0.4567,
"step": 1314
},
{
"epoch": 1.8356196233434086,
"grad_norm": 1.9745222058356298,
"learning_rate": 2.0567047275779184e-07,
"loss": 0.4169,
"step": 1315
},
{
"epoch": 1.8370146477563356,
"grad_norm": 1.9096372104370183,
"learning_rate": 2.0222289494055357e-07,
"loss": 0.4349,
"step": 1316
},
{
"epoch": 1.8384096721692629,
"grad_norm": 2.0974005709989676,
"learning_rate": 1.988038606658066e-07,
"loss": 0.4603,
"step": 1317
},
{
"epoch": 1.8398046965821901,
"grad_norm": 2.1449356644662183,
"learning_rate": 1.9541339027450256e-07,
"loss": 0.4345,
"step": 1318
},
{
"epoch": 1.8411997209951174,
"grad_norm": 2.0964057418342597,
"learning_rate": 1.9205150393765692e-07,
"loss": 0.4481,
"step": 1319
},
{
"epoch": 1.8425947454080447,
"grad_norm": 1.984763144961962,
"learning_rate": 1.887182216562289e-07,
"loss": 0.4643,
"step": 1320
},
{
"epoch": 1.843989769820972,
"grad_norm": 2.233536151283806,
"learning_rate": 1.8541356326100436e-07,
"loss": 0.4577,
"step": 1321
},
{
"epoch": 1.8453847942338992,
"grad_norm": 2.179287384209138,
"learning_rate": 1.821375484124721e-07,
"loss": 0.4779,
"step": 1322
},
{
"epoch": 1.8467798186468263,
"grad_norm": 2.167929996549515,
"learning_rate": 1.7889019660071728e-07,
"loss": 0.449,
"step": 1323
},
{
"epoch": 1.8481748430597535,
"grad_norm": 2.0761119846456513,
"learning_rate": 1.7567152714529468e-07,
"loss": 0.4373,
"step": 1324
},
{
"epoch": 1.8495698674726808,
"grad_norm": 2.2529101497738013,
"learning_rate": 1.7248155919512067e-07,
"loss": 0.4246,
"step": 1325
},
{
"epoch": 1.8509648918856079,
"grad_norm": 2.1661008201597487,
"learning_rate": 1.693203117283543e-07,
"loss": 0.4932,
"step": 1326
},
{
"epoch": 1.8523599162985351,
"grad_norm": 2.151700601180302,
"learning_rate": 1.6618780355229069e-07,
"loss": 0.4591,
"step": 1327
},
{
"epoch": 1.8537549407114624,
"grad_norm": 2.0281053867258447,
"learning_rate": 1.6308405330324294e-07,
"loss": 0.4462,
"step": 1328
},
{
"epoch": 1.8551499651243897,
"grad_norm": 2.1499188174186252,
"learning_rate": 1.6000907944643428e-07,
"loss": 0.4401,
"step": 1329
},
{
"epoch": 1.856544989537317,
"grad_norm": 2.2303858185021124,
"learning_rate": 1.5696290027588933e-07,
"loss": 0.4671,
"step": 1330
},
{
"epoch": 1.8579400139502442,
"grad_norm": 2.05207937944285,
"learning_rate": 1.5394553391432142e-07,
"loss": 0.4419,
"step": 1331
},
{
"epoch": 1.8593350383631715,
"grad_norm": 2.1335914632220363,
"learning_rate": 1.5095699831302935e-07,
"loss": 0.4559,
"step": 1332
},
{
"epoch": 1.8607300627760985,
"grad_norm": 2.1156083069649356,
"learning_rate": 1.479973112517863e-07,
"loss": 0.3962,
"step": 1333
},
{
"epoch": 1.8621250871890258,
"grad_norm": 2.1890879940111434,
"learning_rate": 1.450664903387372e-07,
"loss": 0.4032,
"step": 1334
},
{
"epoch": 1.863520111601953,
"grad_norm": 1.9867732268542835,
"learning_rate": 1.4216455301029274e-07,
"loss": 0.4072,
"step": 1335
},
{
"epoch": 1.8649151360148801,
"grad_norm": 2.17706468053253,
"learning_rate": 1.3929151653102546e-07,
"loss": 0.4359,
"step": 1336
},
{
"epoch": 1.8663101604278074,
"grad_norm": 2.1342756504255305,
"learning_rate": 1.3644739799356765e-07,
"loss": 0.4028,
"step": 1337
},
{
"epoch": 1.8677051848407347,
"grad_norm": 2.0015663881713945,
"learning_rate": 1.3363221431850816e-07,
"loss": 0.4239,
"step": 1338
},
{
"epoch": 1.869100209253662,
"grad_norm": 1.9342131413372554,
"learning_rate": 1.308459822542951e-07,
"loss": 0.4067,
"step": 1339
},
{
"epoch": 1.8704952336665892,
"grad_norm": 2.2918737457866127,
"learning_rate": 1.2808871837713177e-07,
"loss": 0.4981,
"step": 1340
},
{
"epoch": 1.8718902580795165,
"grad_norm": 2.0962472188544368,
"learning_rate": 1.253604390908819e-07,
"loss": 0.4133,
"step": 1341
},
{
"epoch": 1.8732852824924437,
"grad_norm": 2.225773475829192,
"learning_rate": 1.2266116062696954e-07,
"loss": 0.4306,
"step": 1342
},
{
"epoch": 1.8746803069053708,
"grad_norm": 1.9721376525666336,
"learning_rate": 1.1999089904428385e-07,
"loss": 0.4544,
"step": 1343
},
{
"epoch": 1.876075331318298,
"grad_norm": 2.1332846342167455,
"learning_rate": 1.1734967022908217e-07,
"loss": 0.3989,
"step": 1344
},
{
"epoch": 1.8774703557312253,
"grad_norm": 1.9410944812060686,
"learning_rate": 1.1473748989489775e-07,
"loss": 0.407,
"step": 1345
},
{
"epoch": 1.8788653801441524,
"grad_norm": 1.9919021228347704,
"learning_rate": 1.121543735824443e-07,
"loss": 0.4283,
"step": 1346
},
{
"epoch": 1.8802604045570797,
"grad_norm": 2.085986467469432,
"learning_rate": 1.096003366595233e-07,
"loss": 0.46,
"step": 1347
},
{
"epoch": 1.881655428970007,
"grad_norm": 2.0781150667754935,
"learning_rate": 1.0707539432093517e-07,
"loss": 0.4333,
"step": 1348
},
{
"epoch": 1.8830504533829342,
"grad_norm": 2.0437754943186914,
"learning_rate": 1.0457956158838545e-07,
"loss": 0.4215,
"step": 1349
},
{
"epoch": 1.8844454777958615,
"grad_norm": 2.1132790289858705,
"learning_rate": 1.0211285331039933e-07,
"loss": 0.4667,
"step": 1350
},
{
"epoch": 1.8858405022087887,
"grad_norm": 2.2230225558665864,
"learning_rate": 9.967528416222838e-08,
"loss": 0.4549,
"step": 1351
},
{
"epoch": 1.887235526621716,
"grad_norm": 2.1783739759463385,
"learning_rate": 9.72668686457684e-08,
"loss": 0.4569,
"step": 1352
},
{
"epoch": 1.888630551034643,
"grad_norm": 2.137850433771528,
"learning_rate": 9.488762108946891e-08,
"loss": 0.3936,
"step": 1353
},
{
"epoch": 1.8900255754475703,
"grad_norm": 1.8473066765734945,
"learning_rate": 9.25375556482505e-08,
"loss": 0.4036,
"step": 1354
},
{
"epoch": 1.8914205998604976,
"grad_norm": 1.9731724772291126,
"learning_rate": 9.02166863034204e-08,
"loss": 0.4374,
"step": 1355
},
{
"epoch": 1.8928156242734246,
"grad_norm": 2.0133494359381197,
"learning_rate": 8.792502686258752e-08,
"loss": 0.4398,
"step": 1356
},
{
"epoch": 1.894210648686352,
"grad_norm": 2.060188087569894,
"learning_rate": 8.566259095958207e-08,
"loss": 0.42,
"step": 1357
},
{
"epoch": 1.8956056730992792,
"grad_norm": 2.05937443834135,
"learning_rate": 8.342939205437384e-08,
"loss": 0.4899,
"step": 1358
},
{
"epoch": 1.8970006975122065,
"grad_norm": 2.137588600649725,
"learning_rate": 8.122544343299343e-08,
"loss": 0.4246,
"step": 1359
},
{
"epoch": 1.8983957219251337,
"grad_norm": 2.0768764161061473,
"learning_rate": 7.905075820744956e-08,
"loss": 0.4156,
"step": 1360
},
{
"epoch": 1.899790746338061,
"grad_norm": 2.1826612042473648,
"learning_rate": 7.690534931565519e-08,
"loss": 0.4056,
"step": 1361
},
{
"epoch": 1.9011857707509883,
"grad_norm": 2.104115929738705,
"learning_rate": 7.478922952134815e-08,
"loss": 0.4129,
"step": 1362
},
{
"epoch": 1.9025807951639153,
"grad_norm": 2.127950643895458,
"learning_rate": 7.270241141401568e-08,
"loss": 0.5071,
"step": 1363
},
{
"epoch": 1.9039758195768426,
"grad_norm": 2.3752086764361984,
"learning_rate": 7.064490740882057e-08,
"loss": 0.4423,
"step": 1364
},
{
"epoch": 1.9053708439897699,
"grad_norm": 2.009116532809666,
"learning_rate": 6.861672974652622e-08,
"loss": 0.4315,
"step": 1365
},
{
"epoch": 1.906765868402697,
"grad_norm": 2.053589930668402,
"learning_rate": 6.661789049342282e-08,
"loss": 0.4372,
"step": 1366
},
{
"epoch": 1.9081608928156242,
"grad_norm": 2.1185512360662195,
"learning_rate": 6.464840154125795e-08,
"loss": 0.4441,
"step": 1367
},
{
"epoch": 1.9095559172285514,
"grad_norm": 2.289173094731034,
"learning_rate": 6.270827460716499e-08,
"loss": 0.4599,
"step": 1368
},
{
"epoch": 1.9109509416414787,
"grad_norm": 1.9993535700562155,
"learning_rate": 6.079752123359261e-08,
"loss": 0.4065,
"step": 1369
},
{
"epoch": 1.912345966054406,
"grad_norm": 2.0575734889429147,
"learning_rate": 5.891615278823537e-08,
"loss": 0.4273,
"step": 1370
},
{
"epoch": 1.9137409904673333,
"grad_norm": 1.8757933128037505,
"learning_rate": 5.7064180463969886e-08,
"loss": 0.4753,
"step": 1371
},
{
"epoch": 1.9151360148802605,
"grad_norm": 2.0999441835058485,
"learning_rate": 5.5241615278783245e-08,
"loss": 0.4113,
"step": 1372
},
{
"epoch": 1.9165310392931876,
"grad_norm": 1.9826164775573454,
"learning_rate": 5.34484680757108e-08,
"loss": 0.4652,
"step": 1373
},
{
"epoch": 1.9179260637061148,
"grad_norm": 2.2099157937252203,
"learning_rate": 5.168474952277125e-08,
"loss": 0.476,
"step": 1374
},
{
"epoch": 1.9193210881190421,
"grad_norm": 2.2415586636650144,
"learning_rate": 4.995047011290166e-08,
"loss": 0.4438,
"step": 1375
},
{
"epoch": 1.9207161125319692,
"grad_norm": 2.0173519912409477,
"learning_rate": 4.8245640163896433e-08,
"loss": 0.3857,
"step": 1376
},
{
"epoch": 1.9221111369448964,
"grad_norm": 2.0653872634015644,
"learning_rate": 4.657026981834623e-08,
"loss": 0.4163,
"step": 1377
},
{
"epoch": 1.9235061613578237,
"grad_norm": 2.0853147966134076,
"learning_rate": 4.4924369043575776e-08,
"loss": 0.438,
"step": 1378
},
{
"epoch": 1.924901185770751,
"grad_norm": 2.0891024495727852,
"learning_rate": 4.330794763158619e-08,
"loss": 0.4566,
"step": 1379
},
{
"epoch": 1.9262962101836782,
"grad_norm": 2.014329211388496,
"learning_rate": 4.172101519899607e-08,
"loss": 0.4004,
"step": 1380
},
{
"epoch": 1.9276912345966055,
"grad_norm": 2.099355843113096,
"learning_rate": 4.0163581186984935e-08,
"loss": 0.4773,
"step": 1381
},
{
"epoch": 1.9290862590095328,
"grad_norm": 2.19449600992824,
"learning_rate": 3.8635654861235994e-08,
"loss": 0.4552,
"step": 1382
},
{
"epoch": 1.93048128342246,
"grad_norm": 2.232744089829509,
"learning_rate": 3.713724531188234e-08,
"loss": 0.4947,
"step": 1383
},
{
"epoch": 1.931876307835387,
"grad_norm": 2.1497800555978865,
"learning_rate": 3.566836145345032e-08,
"loss": 0.4119,
"step": 1384
},
{
"epoch": 1.9332713322483144,
"grad_norm": 2.0540515877712973,
"learning_rate": 3.4229012024811215e-08,
"loss": 0.4251,
"step": 1385
},
{
"epoch": 1.9346663566612414,
"grad_norm": 1.971100926846232,
"learning_rate": 3.28192055891241e-08,
"loss": 0.4448,
"step": 1386
},
{
"epoch": 1.9360613810741687,
"grad_norm": 2.140868647356388,
"learning_rate": 3.143895053378698e-08,
"loss": 0.4333,
"step": 1387
},
{
"epoch": 1.937456405487096,
"grad_norm": 2.2347840225168056,
"learning_rate": 3.0088255070389037e-08,
"loss": 0.4744,
"step": 1388
},
{
"epoch": 1.9388514299000232,
"grad_norm": 2.1961087854054915,
"learning_rate": 2.876712723465791e-08,
"loss": 0.4018,
"step": 1389
},
{
"epoch": 1.9402464543129505,
"grad_norm": 2.0748231221186337,
"learning_rate": 2.7475574886414726e-08,
"loss": 0.4332,
"step": 1390
},
{
"epoch": 1.9416414787258778,
"grad_norm": 2.080513122073544,
"learning_rate": 2.6213605709525803e-08,
"loss": 0.4138,
"step": 1391
},
{
"epoch": 1.943036503138805,
"grad_norm": 1.8940699773649592,
"learning_rate": 2.498122721185825e-08,
"loss": 0.3667,
"step": 1392
},
{
"epoch": 1.9444315275517323,
"grad_norm": 2.065116925752544,
"learning_rate": 2.3778446725233328e-08,
"loss": 0.4528,
"step": 1393
},
{
"epoch": 1.9458265519646594,
"grad_norm": 2.2286153954981915,
"learning_rate": 2.2605271405385932e-08,
"loss": 0.3904,
"step": 1394
},
{
"epoch": 1.9472215763775866,
"grad_norm": 1.7968513593721984,
"learning_rate": 2.1461708231919064e-08,
"loss": 0.408,
"step": 1395
},
{
"epoch": 1.9486166007905137,
"grad_norm": 2.110245061358655,
"learning_rate": 2.0347764008262216e-08,
"loss": 0.4298,
"step": 1396
},
{
"epoch": 1.950011625203441,
"grad_norm": 2.14786841024935,
"learning_rate": 1.926344536163416e-08,
"loss": 0.5449,
"step": 1397
},
{
"epoch": 1.9514066496163682,
"grad_norm": 2.130629015124878,
"learning_rate": 1.820875874300021e-08,
"loss": 0.3951,
"step": 1398
},
{
"epoch": 1.9528016740292955,
"grad_norm": 2.155436350433079,
"learning_rate": 1.7183710427035595e-08,
"loss": 0.4529,
"step": 1399
},
{
"epoch": 1.9541966984422228,
"grad_norm": 2.1972601350182854,
"learning_rate": 1.6188306512087692e-08,
"loss": 0.4549,
"step": 1400
},
{
"epoch": 1.95559172285515,
"grad_norm": 2.2464198601839005,
"learning_rate": 1.5222552920138855e-08,
"loss": 0.4751,
"step": 1401
},
{
"epoch": 1.9569867472680773,
"grad_norm": 2.1327761308340305,
"learning_rate": 1.4286455396773092e-08,
"loss": 0.4438,
"step": 1402
},
{
"epoch": 1.9583817716810046,
"grad_norm": 2.057020013675667,
"learning_rate": 1.338001951113943e-08,
"loss": 0.444,
"step": 1403
},
{
"epoch": 1.9597767960939316,
"grad_norm": 2.1678792562294373,
"learning_rate": 1.250325065592195e-08,
"loss": 0.4264,
"step": 1404
},
{
"epoch": 1.961171820506859,
"grad_norm": 1.8983518942589257,
"learning_rate": 1.1656154047303691e-08,
"loss": 0.4593,
"step": 1405
},
{
"epoch": 1.962566844919786,
"grad_norm": 2.1610659684462603,
"learning_rate": 1.083873472494057e-08,
"loss": 0.4464,
"step": 1406
},
{
"epoch": 1.9639618693327132,
"grad_norm": 2.0869753487925906,
"learning_rate": 1.005099755192529e-08,
"loss": 0.45,
"step": 1407
},
{
"epoch": 1.9653568937456405,
"grad_norm": 2.0578950075416764,
"learning_rate": 9.292947214764036e-09,
"loss": 0.4871,
"step": 1408
},
{
"epoch": 1.9667519181585678,
"grad_norm": 2.3034725391793183,
"learning_rate": 8.564588223346493e-09,
"loss": 0.4249,
"step": 1409
},
{
"epoch": 1.968146942571495,
"grad_norm": 1.9173773413193362,
"learning_rate": 7.865924910916977e-09,
"loss": 0.4233,
"step": 1410
},
{
"epoch": 1.9695419669844223,
"grad_norm": 2.0570262788637517,
"learning_rate": 7.196961434052796e-09,
"loss": 0.4357,
"step": 1411
},
{
"epoch": 1.9709369913973496,
"grad_norm": 2.0091815939076976,
"learning_rate": 6.557701772635372e-09,
"loss": 0.4814,
"step": 1412
},
{
"epoch": 1.9723320158102768,
"grad_norm": 2.1197632136213302,
"learning_rate": 5.9481497298291556e-09,
"loss": 0.4307,
"step": 1413
},
{
"epoch": 1.9737270402232039,
"grad_norm": 2.1470802379271383,
"learning_rate": 5.368308932058863e-09,
"loss": 0.3866,
"step": 1414
},
{
"epoch": 1.9751220646361312,
"grad_norm": 1.838815471286048,
"learning_rate": 4.818182828985607e-09,
"loss": 0.4183,
"step": 1415
},
{
"epoch": 1.9765170890490582,
"grad_norm": 2.070812126823317,
"learning_rate": 4.2977746934907975e-09,
"loss": 0.4314,
"step": 1416
},
{
"epoch": 1.9779121134619855,
"grad_norm": 2.1089607715907364,
"learning_rate": 3.807087621650607e-09,
"loss": 0.4657,
"step": 1417
},
{
"epoch": 1.9793071378749127,
"grad_norm": 2.0156870049179796,
"learning_rate": 3.3461245327237603e-09,
"loss": 0.4398,
"step": 1418
},
{
"epoch": 1.98070216228784,
"grad_norm": 1.959812635023234,
"learning_rate": 2.9148881691298812e-09,
"loss": 0.4006,
"step": 1419
},
{
"epoch": 1.9820971867007673,
"grad_norm": 2.1219950274732833,
"learning_rate": 2.5133810964350635e-09,
"loss": 0.4378,
"step": 1420
},
{
"epoch": 1.9834922111136946,
"grad_norm": 2.293659588891337,
"learning_rate": 2.1416057033352144e-09,
"loss": 0.4599,
"step": 1421
},
{
"epoch": 1.9848872355266218,
"grad_norm": 1.7483406974593192,
"learning_rate": 1.7995642016449544e-09,
"loss": 0.4657,
"step": 1422
},
{
"epoch": 1.986282259939549,
"grad_norm": 2.2362141922397267,
"learning_rate": 1.4872586262792975e-09,
"loss": 0.4324,
"step": 1423
},
{
"epoch": 1.9876772843524761,
"grad_norm": 2.115899803265782,
"learning_rate": 1.2046908352475462e-09,
"loss": 0.4817,
"step": 1424
},
{
"epoch": 1.9890723087654034,
"grad_norm": 2.2142030661330776,
"learning_rate": 9.518625096366364e-10,
"loss": 0.4473,
"step": 1425
},
{
"epoch": 1.9904673331783305,
"grad_norm": 2.1437905315255703,
"learning_rate": 7.287751536050324e-10,
"loss": 0.411,
"step": 1426
},
{
"epoch": 1.9918623575912577,
"grad_norm": 2.037019408233888,
"learning_rate": 5.354300943727353e-10,
"loss": 0.4586,
"step": 1427
},
{
"epoch": 1.993257382004185,
"grad_norm": 2.2737343174448714,
"learning_rate": 3.718284822118445e-10,
"loss": 0.4408,
"step": 1428
},
{
"epoch": 1.9946524064171123,
"grad_norm": 1.8775148208475847,
"learning_rate": 2.379712904426734e-10,
"loss": 0.3978,
"step": 1429
},
{
"epoch": 1.9960474308300395,
"grad_norm": 2.229374174784266,
"learning_rate": 1.3385931542486686e-10,
"loss": 0.4362,
"step": 1430
},
{
"epoch": 1.9974424552429668,
"grad_norm": 2.0489067262541956,
"learning_rate": 5.949317655462583e-11,
"loss": 0.4613,
"step": 1431
},
{
"epoch": 1.998837479655894,
"grad_norm": 2.161620948843524,
"learning_rate": 1.4873316260266558e-11,
"loss": 0.4646,
"step": 1432
},
{
"epoch": 1.998837479655894,
"step": 1432,
"total_flos": 1.7074790695593574e+17,
"train_loss": 0.6007801830851832,
"train_runtime": 11727.8019,
"train_samples_per_second": 2.934,
"train_steps_per_second": 0.122
}
],
"logging_steps": 1,
"max_steps": 1432,
"num_input_tokens_seen": 0,
"num_train_epochs": 2,
"save_steps": 400,
"stateful_callbacks": {
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 1.7074790695593574e+17,
"train_batch_size": 1,
"trial_name": null,
"trial_params": null
}