aleegis12's picture
Training in progress, step 150, checkpoint
db8ef88 verified
raw
history blame
27.8 kB
{
"best_metric": 1.135201334953308,
"best_model_checkpoint": "miner_id_24/checkpoint-150",
"epoch": 0.42164441321152496,
"eval_steps": 50,
"global_step": 150,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.0028109627547434997,
"grad_norm": 0.14595530927181244,
"learning_rate": 1e-05,
"loss": 1.615,
"step": 1
},
{
"epoch": 0.0028109627547434997,
"eval_loss": 1.2795743942260742,
"eval_runtime": 84.3689,
"eval_samples_per_second": 7.1,
"eval_steps_per_second": 1.778,
"step": 1
},
{
"epoch": 0.005621925509486999,
"grad_norm": 0.13871502876281738,
"learning_rate": 2e-05,
"loss": 1.4278,
"step": 2
},
{
"epoch": 0.008432888264230498,
"grad_norm": 0.13428433239459991,
"learning_rate": 3e-05,
"loss": 1.3684,
"step": 3
},
{
"epoch": 0.011243851018973999,
"grad_norm": 0.13588163256645203,
"learning_rate": 4e-05,
"loss": 1.4611,
"step": 4
},
{
"epoch": 0.014054813773717497,
"grad_norm": 0.13623295724391937,
"learning_rate": 5e-05,
"loss": 1.3713,
"step": 5
},
{
"epoch": 0.016865776528460996,
"grad_norm": 0.14679689705371857,
"learning_rate": 6e-05,
"loss": 1.6388,
"step": 6
},
{
"epoch": 0.019676739283204497,
"grad_norm": 0.13014540076255798,
"learning_rate": 7e-05,
"loss": 1.2928,
"step": 7
},
{
"epoch": 0.022487702037947997,
"grad_norm": 0.14622603356838226,
"learning_rate": 8e-05,
"loss": 1.4067,
"step": 8
},
{
"epoch": 0.025298664792691498,
"grad_norm": 0.1453775018453598,
"learning_rate": 9e-05,
"loss": 1.4303,
"step": 9
},
{
"epoch": 0.028109627547434995,
"grad_norm": 0.14096730947494507,
"learning_rate": 0.0001,
"loss": 1.4121,
"step": 10
},
{
"epoch": 0.030920590302178495,
"grad_norm": 0.1445743590593338,
"learning_rate": 9.999316524962345e-05,
"loss": 1.3271,
"step": 11
},
{
"epoch": 0.03373155305692199,
"grad_norm": 0.16167715191841125,
"learning_rate": 9.997266286704631e-05,
"loss": 1.5619,
"step": 12
},
{
"epoch": 0.036542515811665496,
"grad_norm": 0.15229864418506622,
"learning_rate": 9.993849845741524e-05,
"loss": 1.4586,
"step": 13
},
{
"epoch": 0.03935347856640899,
"grad_norm": 0.1474287509918213,
"learning_rate": 9.989068136093873e-05,
"loss": 1.5177,
"step": 14
},
{
"epoch": 0.0421644413211525,
"grad_norm": 0.1248803585767746,
"learning_rate": 9.98292246503335e-05,
"loss": 1.3532,
"step": 15
},
{
"epoch": 0.044975404075895994,
"grad_norm": 0.14056053757667542,
"learning_rate": 9.975414512725057e-05,
"loss": 1.5619,
"step": 16
},
{
"epoch": 0.04778636683063949,
"grad_norm": 0.1264151781797409,
"learning_rate": 9.966546331768191e-05,
"loss": 1.3178,
"step": 17
},
{
"epoch": 0.050597329585382995,
"grad_norm": 0.12755592167377472,
"learning_rate": 9.956320346634876e-05,
"loss": 1.4588,
"step": 18
},
{
"epoch": 0.05340829234012649,
"grad_norm": 0.13023661077022552,
"learning_rate": 9.944739353007344e-05,
"loss": 1.5323,
"step": 19
},
{
"epoch": 0.05621925509486999,
"grad_norm": 0.12780460715293884,
"learning_rate": 9.931806517013612e-05,
"loss": 1.4957,
"step": 20
},
{
"epoch": 0.059030217849613494,
"grad_norm": 0.12764661014080048,
"learning_rate": 9.917525374361912e-05,
"loss": 0.9073,
"step": 21
},
{
"epoch": 0.06184118060435699,
"grad_norm": 0.12089043855667114,
"learning_rate": 9.901899829374047e-05,
"loss": 0.8821,
"step": 22
},
{
"epoch": 0.0646521433591005,
"grad_norm": 0.11314330995082855,
"learning_rate": 9.884934153917997e-05,
"loss": 0.8205,
"step": 23
},
{
"epoch": 0.06746310611384398,
"grad_norm": 0.128267303109169,
"learning_rate": 9.86663298624003e-05,
"loss": 0.8299,
"step": 24
},
{
"epoch": 0.07027406886858749,
"grad_norm": 0.13578976690769196,
"learning_rate": 9.847001329696653e-05,
"loss": 0.9728,
"step": 25
},
{
"epoch": 0.07308503162333099,
"grad_norm": 0.11824839562177658,
"learning_rate": 9.826044551386744e-05,
"loss": 0.7145,
"step": 26
},
{
"epoch": 0.0758959943780745,
"grad_norm": 0.13589346408843994,
"learning_rate": 9.803768380684242e-05,
"loss": 0.9693,
"step": 27
},
{
"epoch": 0.07870695713281799,
"grad_norm": 0.13335402309894562,
"learning_rate": 9.780178907671789e-05,
"loss": 0.8476,
"step": 28
},
{
"epoch": 0.08151791988756149,
"grad_norm": 0.14058586955070496,
"learning_rate": 9.755282581475769e-05,
"loss": 0.9686,
"step": 29
},
{
"epoch": 0.084328882642305,
"grad_norm": 0.14462514221668243,
"learning_rate": 9.729086208503174e-05,
"loss": 0.8823,
"step": 30
},
{
"epoch": 0.08713984539704848,
"grad_norm": 0.1474841982126236,
"learning_rate": 9.701596950580806e-05,
"loss": 0.9784,
"step": 31
},
{
"epoch": 0.08995080815179199,
"grad_norm": 0.158616304397583,
"learning_rate": 9.672822322997305e-05,
"loss": 0.9347,
"step": 32
},
{
"epoch": 0.09276177090653549,
"grad_norm": 0.1760544329881668,
"learning_rate": 9.642770192448536e-05,
"loss": 1.0842,
"step": 33
},
{
"epoch": 0.09557273366127898,
"grad_norm": 0.1678563505411148,
"learning_rate": 9.611448774886924e-05,
"loss": 0.948,
"step": 34
},
{
"epoch": 0.09838369641602249,
"grad_norm": 0.17710591852664948,
"learning_rate": 9.578866633275288e-05,
"loss": 1.0793,
"step": 35
},
{
"epoch": 0.10119465917076599,
"grad_norm": 0.1860406994819641,
"learning_rate": 9.545032675245813e-05,
"loss": 0.8863,
"step": 36
},
{
"epoch": 0.10400562192550948,
"grad_norm": 0.1834695041179657,
"learning_rate": 9.509956150664796e-05,
"loss": 1.1723,
"step": 37
},
{
"epoch": 0.10681658468025299,
"grad_norm": 0.19306811690330505,
"learning_rate": 9.473646649103818e-05,
"loss": 0.9949,
"step": 38
},
{
"epoch": 0.10962754743499649,
"grad_norm": 0.17946985363960266,
"learning_rate": 9.43611409721806e-05,
"loss": 0.833,
"step": 39
},
{
"epoch": 0.11243851018973998,
"grad_norm": 0.20696531236171722,
"learning_rate": 9.397368756032445e-05,
"loss": 1.2203,
"step": 40
},
{
"epoch": 0.11524947294448348,
"grad_norm": 0.20919673144817352,
"learning_rate": 9.357421218136386e-05,
"loss": 1.1978,
"step": 41
},
{
"epoch": 0.11806043569922699,
"grad_norm": 0.21747088432312012,
"learning_rate": 9.316282404787871e-05,
"loss": 1.0537,
"step": 42
},
{
"epoch": 0.12087139845397049,
"grad_norm": 0.21661460399627686,
"learning_rate": 9.273963562927695e-05,
"loss": 1.1814,
"step": 43
},
{
"epoch": 0.12368236120871398,
"grad_norm": 0.21449671685695648,
"learning_rate": 9.230476262104677e-05,
"loss": 0.9883,
"step": 44
},
{
"epoch": 0.12649332396345747,
"grad_norm": 0.257315069437027,
"learning_rate": 9.185832391312644e-05,
"loss": 1.2025,
"step": 45
},
{
"epoch": 0.129304286718201,
"grad_norm": 0.21534892916679382,
"learning_rate": 9.140044155740101e-05,
"loss": 0.8393,
"step": 46
},
{
"epoch": 0.13211524947294448,
"grad_norm": 0.2745459973812103,
"learning_rate": 9.093124073433463e-05,
"loss": 1.0887,
"step": 47
},
{
"epoch": 0.13492621222768797,
"grad_norm": 0.24269066751003265,
"learning_rate": 9.045084971874738e-05,
"loss": 0.8241,
"step": 48
},
{
"epoch": 0.1377371749824315,
"grad_norm": 0.4028018116950989,
"learning_rate": 8.995939984474624e-05,
"loss": 1.03,
"step": 49
},
{
"epoch": 0.14054813773717498,
"grad_norm": 0.3536010980606079,
"learning_rate": 8.945702546981969e-05,
"loss": 0.5716,
"step": 50
},
{
"epoch": 0.14054813773717498,
"eval_loss": 1.2065433263778687,
"eval_runtime": 85.3934,
"eval_samples_per_second": 7.015,
"eval_steps_per_second": 1.757,
"step": 50
},
{
"epoch": 0.1433591004919185,
"grad_norm": 4.709478378295898,
"learning_rate": 8.894386393810563e-05,
"loss": 1.5611,
"step": 51
},
{
"epoch": 0.14617006324666199,
"grad_norm": 0.30073219537734985,
"learning_rate": 8.842005554284296e-05,
"loss": 1.2047,
"step": 52
},
{
"epoch": 0.14898102600140548,
"grad_norm": 0.1945256143808365,
"learning_rate": 8.788574348801675e-05,
"loss": 1.1813,
"step": 53
},
{
"epoch": 0.151791988756149,
"grad_norm": 0.18454934656620026,
"learning_rate": 8.73410738492077e-05,
"loss": 1.5916,
"step": 54
},
{
"epoch": 0.15460295151089248,
"grad_norm": 0.1515914499759674,
"learning_rate": 8.678619553365659e-05,
"loss": 1.4081,
"step": 55
},
{
"epoch": 0.15741391426563597,
"grad_norm": 0.13022002577781677,
"learning_rate": 8.622126023955446e-05,
"loss": 1.2496,
"step": 56
},
{
"epoch": 0.1602248770203795,
"grad_norm": 0.13143523037433624,
"learning_rate": 8.564642241456986e-05,
"loss": 1.3947,
"step": 57
},
{
"epoch": 0.16303583977512298,
"grad_norm": 0.12690874934196472,
"learning_rate": 8.506183921362443e-05,
"loss": 1.2596,
"step": 58
},
{
"epoch": 0.16584680252986647,
"grad_norm": 0.12422709912061691,
"learning_rate": 8.44676704559283e-05,
"loss": 1.343,
"step": 59
},
{
"epoch": 0.16865776528461,
"grad_norm": 0.12969334423542023,
"learning_rate": 8.386407858128706e-05,
"loss": 1.392,
"step": 60
},
{
"epoch": 0.17146872803935348,
"grad_norm": 0.12702302634716034,
"learning_rate": 8.32512286056924e-05,
"loss": 1.3217,
"step": 61
},
{
"epoch": 0.17427969079409697,
"grad_norm": 0.1328938603401184,
"learning_rate": 8.262928807620843e-05,
"loss": 1.4824,
"step": 62
},
{
"epoch": 0.1770906535488405,
"grad_norm": 0.1313563585281372,
"learning_rate": 8.199842702516583e-05,
"loss": 1.523,
"step": 63
},
{
"epoch": 0.17990161630358398,
"grad_norm": 0.12059923261404037,
"learning_rate": 8.135881792367686e-05,
"loss": 1.2967,
"step": 64
},
{
"epoch": 0.18271257905832747,
"grad_norm": 0.12688899040222168,
"learning_rate": 8.07106356344834e-05,
"loss": 1.4306,
"step": 65
},
{
"epoch": 0.18552354181307099,
"grad_norm": 0.12246695905923843,
"learning_rate": 8.005405736415126e-05,
"loss": 1.3235,
"step": 66
},
{
"epoch": 0.18833450456781448,
"grad_norm": 0.1242348924279213,
"learning_rate": 7.938926261462366e-05,
"loss": 1.4104,
"step": 67
},
{
"epoch": 0.19114546732255797,
"grad_norm": 0.13645027577877045,
"learning_rate": 7.871643313414718e-05,
"loss": 1.4968,
"step": 68
},
{
"epoch": 0.19395643007730148,
"grad_norm": 0.13051196932792664,
"learning_rate": 7.803575286758364e-05,
"loss": 1.5518,
"step": 69
},
{
"epoch": 0.19676739283204497,
"grad_norm": 0.13176923990249634,
"learning_rate": 7.734740790612136e-05,
"loss": 1.4094,
"step": 70
},
{
"epoch": 0.19957835558678846,
"grad_norm": 0.11671514809131622,
"learning_rate": 7.66515864363997e-05,
"loss": 0.8881,
"step": 71
},
{
"epoch": 0.20238931834153198,
"grad_norm": 0.10778027772903442,
"learning_rate": 7.594847868906076e-05,
"loss": 0.8403,
"step": 72
},
{
"epoch": 0.20520028109627547,
"grad_norm": 0.116300567984581,
"learning_rate": 7.52382768867422e-05,
"loss": 0.9927,
"step": 73
},
{
"epoch": 0.20801124385101896,
"grad_norm": 0.11594687402248383,
"learning_rate": 7.452117519152542e-05,
"loss": 0.8545,
"step": 74
},
{
"epoch": 0.21082220660576248,
"grad_norm": 0.11859302967786789,
"learning_rate": 7.379736965185368e-05,
"loss": 0.8622,
"step": 75
},
{
"epoch": 0.21363316936050597,
"grad_norm": 0.12879763543605804,
"learning_rate": 7.30670581489344e-05,
"loss": 1.0428,
"step": 76
},
{
"epoch": 0.21644413211524946,
"grad_norm": 0.1271115094423294,
"learning_rate": 7.233044034264034e-05,
"loss": 0.8829,
"step": 77
},
{
"epoch": 0.21925509486999298,
"grad_norm": 0.13602392375469208,
"learning_rate": 7.158771761692464e-05,
"loss": 0.8915,
"step": 78
},
{
"epoch": 0.22206605762473647,
"grad_norm": 0.13157889246940613,
"learning_rate": 7.083909302476453e-05,
"loss": 0.9382,
"step": 79
},
{
"epoch": 0.22487702037947996,
"grad_norm": 0.13888955116271973,
"learning_rate": 7.008477123264848e-05,
"loss": 0.9532,
"step": 80
},
{
"epoch": 0.22768798313422348,
"grad_norm": 0.13553336262702942,
"learning_rate": 6.932495846462261e-05,
"loss": 0.9201,
"step": 81
},
{
"epoch": 0.23049894588896697,
"grad_norm": 0.12914668023586273,
"learning_rate": 6.855986244591104e-05,
"loss": 0.7391,
"step": 82
},
{
"epoch": 0.23330990864371048,
"grad_norm": 0.14608892798423767,
"learning_rate": 6.778969234612584e-05,
"loss": 1.0011,
"step": 83
},
{
"epoch": 0.23612087139845397,
"grad_norm": 0.1511840671300888,
"learning_rate": 6.701465872208216e-05,
"loss": 0.8691,
"step": 84
},
{
"epoch": 0.23893183415319746,
"grad_norm": 0.15799711644649506,
"learning_rate": 6.623497346023418e-05,
"loss": 0.8982,
"step": 85
},
{
"epoch": 0.24174279690794098,
"grad_norm": 0.166089728474617,
"learning_rate": 6.545084971874738e-05,
"loss": 1.057,
"step": 86
},
{
"epoch": 0.24455375966268447,
"grad_norm": 0.1631898581981659,
"learning_rate": 6.466250186922325e-05,
"loss": 1.0667,
"step": 87
},
{
"epoch": 0.24736472241742796,
"grad_norm": 0.17566239833831787,
"learning_rate": 6.387014543809223e-05,
"loss": 1.0864,
"step": 88
},
{
"epoch": 0.2501756851721715,
"grad_norm": 0.17442883551120758,
"learning_rate": 6.307399704769099e-05,
"loss": 0.8744,
"step": 89
},
{
"epoch": 0.25298664792691494,
"grad_norm": 0.19199849665164948,
"learning_rate": 6.227427435703997e-05,
"loss": 1.2237,
"step": 90
},
{
"epoch": 0.25579761068165846,
"grad_norm": 0.19623346626758575,
"learning_rate": 6.147119600233758e-05,
"loss": 1.0817,
"step": 91
},
{
"epoch": 0.258608573436402,
"grad_norm": 0.21419471502304077,
"learning_rate": 6.066498153718735e-05,
"loss": 1.1695,
"step": 92
},
{
"epoch": 0.26141953619114544,
"grad_norm": 0.21495421230793,
"learning_rate": 5.985585137257401e-05,
"loss": 1.1842,
"step": 93
},
{
"epoch": 0.26423049894588896,
"grad_norm": 0.21394377946853638,
"learning_rate": 5.90440267166055e-05,
"loss": 0.9014,
"step": 94
},
{
"epoch": 0.2670414617006325,
"grad_norm": 0.2387458235025406,
"learning_rate": 5.8229729514036705e-05,
"loss": 1.1359,
"step": 95
},
{
"epoch": 0.26985242445537594,
"grad_norm": 0.21522080898284912,
"learning_rate": 5.74131823855921e-05,
"loss": 1.0008,
"step": 96
},
{
"epoch": 0.27266338721011946,
"grad_norm": 0.27252939343452454,
"learning_rate": 5.6594608567103456e-05,
"loss": 1.1734,
"step": 97
},
{
"epoch": 0.275474349964863,
"grad_norm": 0.26594337821006775,
"learning_rate": 5.577423184847932e-05,
"loss": 1.0443,
"step": 98
},
{
"epoch": 0.2782853127196065,
"grad_norm": 0.2911270558834076,
"learning_rate": 5.495227651252315e-05,
"loss": 0.8558,
"step": 99
},
{
"epoch": 0.28109627547434995,
"grad_norm": 0.2107650637626648,
"learning_rate": 5.4128967273616625e-05,
"loss": 0.2094,
"step": 100
},
{
"epoch": 0.28109627547434995,
"eval_loss": 1.1504157781600952,
"eval_runtime": 85.2851,
"eval_samples_per_second": 7.024,
"eval_steps_per_second": 1.759,
"step": 100
},
{
"epoch": 0.2839072382290935,
"grad_norm": 0.19906841218471527,
"learning_rate": 5.330452921628497e-05,
"loss": 1.3064,
"step": 101
},
{
"epoch": 0.286718200983837,
"grad_norm": 0.17362119257450104,
"learning_rate": 5.247918773366112e-05,
"loss": 1.3952,
"step": 102
},
{
"epoch": 0.28952916373858045,
"grad_norm": 0.1592879295349121,
"learning_rate": 5.165316846586541e-05,
"loss": 1.2231,
"step": 103
},
{
"epoch": 0.29234012649332397,
"grad_norm": 0.14367029070854187,
"learning_rate": 5.0826697238317935e-05,
"loss": 1.0226,
"step": 104
},
{
"epoch": 0.2951510892480675,
"grad_norm": 0.13938987255096436,
"learning_rate": 5e-05,
"loss": 1.3075,
"step": 105
},
{
"epoch": 0.29796205200281095,
"grad_norm": 0.14711101353168488,
"learning_rate": 4.917330276168208e-05,
"loss": 1.4146,
"step": 106
},
{
"epoch": 0.30077301475755447,
"grad_norm": 0.12315893918275833,
"learning_rate": 4.834683153413459e-05,
"loss": 1.2684,
"step": 107
},
{
"epoch": 0.303583977512298,
"grad_norm": 0.13632048666477203,
"learning_rate": 4.7520812266338885e-05,
"loss": 1.4743,
"step": 108
},
{
"epoch": 0.30639494026704145,
"grad_norm": 0.11947023868560791,
"learning_rate": 4.669547078371504e-05,
"loss": 1.3492,
"step": 109
},
{
"epoch": 0.30920590302178497,
"grad_norm": 0.11925946921110153,
"learning_rate": 4.5871032726383386e-05,
"loss": 1.4018,
"step": 110
},
{
"epoch": 0.3120168657765285,
"grad_norm": 0.11408793926239014,
"learning_rate": 4.504772348747687e-05,
"loss": 1.1567,
"step": 111
},
{
"epoch": 0.31482782853127195,
"grad_norm": 0.10827047377824783,
"learning_rate": 4.4225768151520694e-05,
"loss": 1.1115,
"step": 112
},
{
"epoch": 0.31763879128601546,
"grad_norm": 0.11432784795761108,
"learning_rate": 4.3405391432896555e-05,
"loss": 1.2277,
"step": 113
},
{
"epoch": 0.320449754040759,
"grad_norm": 0.12372433394193649,
"learning_rate": 4.2586817614407895e-05,
"loss": 1.3614,
"step": 114
},
{
"epoch": 0.32326071679550245,
"grad_norm": 0.12038455903530121,
"learning_rate": 4.17702704859633e-05,
"loss": 1.37,
"step": 115
},
{
"epoch": 0.32607167955024596,
"grad_norm": 0.12178530544042587,
"learning_rate": 4.095597328339452e-05,
"loss": 1.4235,
"step": 116
},
{
"epoch": 0.3288826423049895,
"grad_norm": 0.122628353536129,
"learning_rate": 4.0144148627425993e-05,
"loss": 1.4401,
"step": 117
},
{
"epoch": 0.33169360505973294,
"grad_norm": 0.13086412847042084,
"learning_rate": 3.933501846281267e-05,
"loss": 1.6775,
"step": 118
},
{
"epoch": 0.33450456781447646,
"grad_norm": 0.1191410943865776,
"learning_rate": 3.852880399766243e-05,
"loss": 1.403,
"step": 119
},
{
"epoch": 0.33731553056922,
"grad_norm": 0.12145199626684189,
"learning_rate": 3.772572564296005e-05,
"loss": 1.3555,
"step": 120
},
{
"epoch": 0.34012649332396344,
"grad_norm": 0.10900752246379852,
"learning_rate": 3.6926002952309016e-05,
"loss": 1.1942,
"step": 121
},
{
"epoch": 0.34293745607870696,
"grad_norm": 0.10437808185815811,
"learning_rate": 3.612985456190778e-05,
"loss": 0.8917,
"step": 122
},
{
"epoch": 0.3457484188334505,
"grad_norm": 0.10854747891426086,
"learning_rate": 3.533749813077677e-05,
"loss": 0.9686,
"step": 123
},
{
"epoch": 0.34855938158819394,
"grad_norm": 0.10949317365884781,
"learning_rate": 3.4549150281252636e-05,
"loss": 0.809,
"step": 124
},
{
"epoch": 0.35137034434293746,
"grad_norm": 0.10705723613500595,
"learning_rate": 3.3765026539765834e-05,
"loss": 0.8198,
"step": 125
},
{
"epoch": 0.354181307097681,
"grad_norm": 0.11876116693019867,
"learning_rate": 3.298534127791785e-05,
"loss": 0.9964,
"step": 126
},
{
"epoch": 0.35699226985242444,
"grad_norm": 0.11575191468000412,
"learning_rate": 3.221030765387417e-05,
"loss": 0.9049,
"step": 127
},
{
"epoch": 0.35980323260716796,
"grad_norm": 0.11619889736175537,
"learning_rate": 3.144013755408895e-05,
"loss": 0.8849,
"step": 128
},
{
"epoch": 0.3626141953619115,
"grad_norm": 0.13133642077445984,
"learning_rate": 3.0675041535377405e-05,
"loss": 0.8731,
"step": 129
},
{
"epoch": 0.36542515811665494,
"grad_norm": 0.12494848668575287,
"learning_rate": 2.991522876735154e-05,
"loss": 0.9458,
"step": 130
},
{
"epoch": 0.36823612087139845,
"grad_norm": 0.1314287930727005,
"learning_rate": 2.916090697523549e-05,
"loss": 1.0028,
"step": 131
},
{
"epoch": 0.37104708362614197,
"grad_norm": 0.12119800597429276,
"learning_rate": 2.8412282383075363e-05,
"loss": 0.785,
"step": 132
},
{
"epoch": 0.37385804638088543,
"grad_norm": 0.13152635097503662,
"learning_rate": 2.766955965735968e-05,
"loss": 0.8102,
"step": 133
},
{
"epoch": 0.37666900913562895,
"grad_norm": 0.13744279742240906,
"learning_rate": 2.693294185106562e-05,
"loss": 0.9912,
"step": 134
},
{
"epoch": 0.37947997189037247,
"grad_norm": 0.1392669528722763,
"learning_rate": 2.6202630348146324e-05,
"loss": 0.9372,
"step": 135
},
{
"epoch": 0.38229093464511593,
"grad_norm": 0.15417630970478058,
"learning_rate": 2.547882480847461e-05,
"loss": 1.0982,
"step": 136
},
{
"epoch": 0.38510189739985945,
"grad_norm": 0.14992496371269226,
"learning_rate": 2.476172311325783e-05,
"loss": 0.9687,
"step": 137
},
{
"epoch": 0.38791286015460297,
"grad_norm": 0.1466035395860672,
"learning_rate": 2.405152131093926e-05,
"loss": 0.89,
"step": 138
},
{
"epoch": 0.39072382290934643,
"grad_norm": 0.16423414647579193,
"learning_rate": 2.3348413563600325e-05,
"loss": 1.0407,
"step": 139
},
{
"epoch": 0.39353478566408995,
"grad_norm": 0.16661831736564636,
"learning_rate": 2.2652592093878666e-05,
"loss": 1.1279,
"step": 140
},
{
"epoch": 0.39634574841883347,
"grad_norm": 0.19206121563911438,
"learning_rate": 2.196424713241637e-05,
"loss": 1.0473,
"step": 141
},
{
"epoch": 0.39915671117357693,
"grad_norm": 0.2245406061410904,
"learning_rate": 2.128356686585282e-05,
"loss": 1.1602,
"step": 142
},
{
"epoch": 0.40196767392832045,
"grad_norm": 0.19214637577533722,
"learning_rate": 2.061073738537635e-05,
"loss": 1.0542,
"step": 143
},
{
"epoch": 0.40477863668306396,
"grad_norm": 0.19177591800689697,
"learning_rate": 1.9945942635848748e-05,
"loss": 0.9993,
"step": 144
},
{
"epoch": 0.4075895994378074,
"grad_norm": 0.21935953199863434,
"learning_rate": 1.928936436551661e-05,
"loss": 1.2079,
"step": 145
},
{
"epoch": 0.41040056219255094,
"grad_norm": 0.23886896669864655,
"learning_rate": 1.8641182076323148e-05,
"loss": 1.327,
"step": 146
},
{
"epoch": 0.41321152494729446,
"grad_norm": 0.20663587749004364,
"learning_rate": 1.800157297483417e-05,
"loss": 0.8727,
"step": 147
},
{
"epoch": 0.4160224877020379,
"grad_norm": 0.23780639469623566,
"learning_rate": 1.7370711923791567e-05,
"loss": 0.9077,
"step": 148
},
{
"epoch": 0.41883345045678144,
"grad_norm": 0.23140735924243927,
"learning_rate": 1.6748771394307585e-05,
"loss": 0.6117,
"step": 149
},
{
"epoch": 0.42164441321152496,
"grad_norm": 0.3450438976287842,
"learning_rate": 1.6135921418712956e-05,
"loss": 0.7388,
"step": 150
},
{
"epoch": 0.42164441321152496,
"eval_loss": 1.135201334953308,
"eval_runtime": 85.3695,
"eval_samples_per_second": 7.017,
"eval_steps_per_second": 1.757,
"step": 150
}
],
"logging_steps": 1,
"max_steps": 200,
"num_input_tokens_seen": 0,
"num_train_epochs": 1,
"save_steps": 50,
"stateful_callbacks": {
"EarlyStoppingCallback": {
"args": {
"early_stopping_patience": 5,
"early_stopping_threshold": 0.0
},
"attributes": {
"early_stopping_patience_counter": 0
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": false
},
"attributes": {}
}
},
"total_flos": 4.2585566126014464e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}