lesso's picture
Training in progress, step 200, checkpoint
d171dba verified
raw
history blame
37.2 kB
{
"best_metric": 11.5,
"best_model_checkpoint": "miner_id_24/checkpoint-50",
"epoch": 0.004025886449872681,
"eval_steps": 50,
"global_step": 200,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 2.0129432249363408e-05,
"grad_norm": 1.633916144783143e-05,
"learning_rate": 1.001e-05,
"loss": 23.0,
"step": 1
},
{
"epoch": 2.0129432249363408e-05,
"eval_loss": 11.5,
"eval_runtime": 286.0019,
"eval_samples_per_second": 73.139,
"eval_steps_per_second": 18.287,
"step": 1
},
{
"epoch": 4.0258864498726816e-05,
"grad_norm": 1.640866139496211e-05,
"learning_rate": 2.002e-05,
"loss": 23.0,
"step": 2
},
{
"epoch": 6.038829674809022e-05,
"grad_norm": 2.1919127902947366e-05,
"learning_rate": 3.0029999999999995e-05,
"loss": 23.0,
"step": 3
},
{
"epoch": 8.051772899745363e-05,
"grad_norm": 1.0230632142338436e-05,
"learning_rate": 4.004e-05,
"loss": 23.0,
"step": 4
},
{
"epoch": 0.00010064716124681703,
"grad_norm": 1.0321247827960178e-05,
"learning_rate": 5.005e-05,
"loss": 23.0,
"step": 5
},
{
"epoch": 0.00012077659349618043,
"grad_norm": 2.285218761244323e-05,
"learning_rate": 6.005999999999999e-05,
"loss": 23.0,
"step": 6
},
{
"epoch": 0.00014090602574554385,
"grad_norm": 1.7603799278731458e-05,
"learning_rate": 7.006999999999998e-05,
"loss": 23.0,
"step": 7
},
{
"epoch": 0.00016103545799490726,
"grad_norm": 2.4340071831829846e-05,
"learning_rate": 8.008e-05,
"loss": 23.0,
"step": 8
},
{
"epoch": 0.00018116489024427065,
"grad_norm": 1.0853328603843693e-05,
"learning_rate": 9.009e-05,
"loss": 23.0,
"step": 9
},
{
"epoch": 0.00020129432249363407,
"grad_norm": 1.442844768462237e-05,
"learning_rate": 0.0001001,
"loss": 23.0,
"step": 10
},
{
"epoch": 0.00022142375474299748,
"grad_norm": 2.4116914573824033e-05,
"learning_rate": 9.957315789473684e-05,
"loss": 23.0,
"step": 11
},
{
"epoch": 0.00024155318699236087,
"grad_norm": 2.6325762519263662e-05,
"learning_rate": 9.904631578947367e-05,
"loss": 23.0,
"step": 12
},
{
"epoch": 0.0002616826192417243,
"grad_norm": 2.7495216272654943e-05,
"learning_rate": 9.851947368421052e-05,
"loss": 23.0,
"step": 13
},
{
"epoch": 0.0002818120514910877,
"grad_norm": 1.619177419343032e-05,
"learning_rate": 9.799263157894736e-05,
"loss": 23.0,
"step": 14
},
{
"epoch": 0.0003019414837404511,
"grad_norm": 1.714995232759975e-05,
"learning_rate": 9.746578947368421e-05,
"loss": 23.0,
"step": 15
},
{
"epoch": 0.00032207091598981453,
"grad_norm": 3.0432000130531378e-05,
"learning_rate": 9.693894736842104e-05,
"loss": 23.0,
"step": 16
},
{
"epoch": 0.0003422003482391779,
"grad_norm": 1.8349341189605184e-05,
"learning_rate": 9.641210526315789e-05,
"loss": 23.0,
"step": 17
},
{
"epoch": 0.0003623297804885413,
"grad_norm": 1.8204806110588834e-05,
"learning_rate": 9.588526315789473e-05,
"loss": 23.0,
"step": 18
},
{
"epoch": 0.00038245921273790474,
"grad_norm": 3.4496861189836636e-05,
"learning_rate": 9.535842105263157e-05,
"loss": 23.0,
"step": 19
},
{
"epoch": 0.00040258864498726813,
"grad_norm": 3.019250652869232e-05,
"learning_rate": 9.483157894736841e-05,
"loss": 23.0,
"step": 20
},
{
"epoch": 0.0004227180772366315,
"grad_norm": 1.3691938875126652e-05,
"learning_rate": 9.430473684210526e-05,
"loss": 23.0,
"step": 21
},
{
"epoch": 0.00044284750948599496,
"grad_norm": 1.700852590147406e-05,
"learning_rate": 9.37778947368421e-05,
"loss": 23.0,
"step": 22
},
{
"epoch": 0.00046297694173535835,
"grad_norm": 1.528211942058988e-05,
"learning_rate": 9.325105263157894e-05,
"loss": 23.0,
"step": 23
},
{
"epoch": 0.00048310637398472174,
"grad_norm": 2.506410237401724e-05,
"learning_rate": 9.272421052631578e-05,
"loss": 23.0,
"step": 24
},
{
"epoch": 0.0005032358062340851,
"grad_norm": 4.976747732143849e-05,
"learning_rate": 9.219736842105263e-05,
"loss": 23.0,
"step": 25
},
{
"epoch": 0.0005233652384834486,
"grad_norm": 2.0973604478058405e-05,
"learning_rate": 9.167052631578946e-05,
"loss": 23.0,
"step": 26
},
{
"epoch": 0.000543494670732812,
"grad_norm": 2.224497075076215e-05,
"learning_rate": 9.114368421052632e-05,
"loss": 23.0,
"step": 27
},
{
"epoch": 0.0005636241029821754,
"grad_norm": 2.854904050764162e-05,
"learning_rate": 9.061684210526315e-05,
"loss": 23.0,
"step": 28
},
{
"epoch": 0.0005837535352315388,
"grad_norm": 8.367866394110024e-05,
"learning_rate": 9.009e-05,
"loss": 23.0,
"step": 29
},
{
"epoch": 0.0006038829674809022,
"grad_norm": 4.322812310419977e-05,
"learning_rate": 8.956315789473683e-05,
"loss": 23.0,
"step": 30
},
{
"epoch": 0.0006240123997302656,
"grad_norm": 6.570852565346286e-05,
"learning_rate": 8.903631578947368e-05,
"loss": 23.0,
"step": 31
},
{
"epoch": 0.0006441418319796291,
"grad_norm": 5.065805817139335e-05,
"learning_rate": 8.850947368421052e-05,
"loss": 23.0,
"step": 32
},
{
"epoch": 0.0006642712642289924,
"grad_norm": 3.0831048206891865e-05,
"learning_rate": 8.798263157894736e-05,
"loss": 23.0,
"step": 33
},
{
"epoch": 0.0006844006964783558,
"grad_norm": 5.162616434972733e-05,
"learning_rate": 8.745578947368422e-05,
"loss": 23.0,
"step": 34
},
{
"epoch": 0.0007045301287277192,
"grad_norm": 5.709606557502411e-05,
"learning_rate": 8.692894736842105e-05,
"loss": 23.0,
"step": 35
},
{
"epoch": 0.0007246595609770826,
"grad_norm": 3.203852247679606e-05,
"learning_rate": 8.64021052631579e-05,
"loss": 23.0,
"step": 36
},
{
"epoch": 0.0007447889932264461,
"grad_norm": 5.754972153226845e-05,
"learning_rate": 8.587526315789473e-05,
"loss": 23.0,
"step": 37
},
{
"epoch": 0.0007649184254758095,
"grad_norm": 3.7275851354934275e-05,
"learning_rate": 8.534842105263157e-05,
"loss": 23.0,
"step": 38
},
{
"epoch": 0.0007850478577251729,
"grad_norm": 2.435182068438735e-05,
"learning_rate": 8.482157894736842e-05,
"loss": 23.0,
"step": 39
},
{
"epoch": 0.0008051772899745363,
"grad_norm": 4.3874577386304736e-05,
"learning_rate": 8.429473684210525e-05,
"loss": 23.0,
"step": 40
},
{
"epoch": 0.0008253067222238997,
"grad_norm": 3.443747482378967e-05,
"learning_rate": 8.376789473684211e-05,
"loss": 23.0,
"step": 41
},
{
"epoch": 0.000845436154473263,
"grad_norm": 4.0556362364441156e-05,
"learning_rate": 8.324105263157894e-05,
"loss": 23.0,
"step": 42
},
{
"epoch": 0.0008655655867226265,
"grad_norm": 4.7736495616845787e-05,
"learning_rate": 8.271421052631579e-05,
"loss": 23.0,
"step": 43
},
{
"epoch": 0.0008856950189719899,
"grad_norm": 2.6885190891334787e-05,
"learning_rate": 8.218736842105262e-05,
"loss": 23.0,
"step": 44
},
{
"epoch": 0.0009058244512213533,
"grad_norm": 4.4231674110051244e-05,
"learning_rate": 8.166052631578947e-05,
"loss": 23.0,
"step": 45
},
{
"epoch": 0.0009259538834707167,
"grad_norm": 8.17498512333259e-05,
"learning_rate": 8.113368421052631e-05,
"loss": 23.0,
"step": 46
},
{
"epoch": 0.0009460833157200801,
"grad_norm": 4.674808224081062e-05,
"learning_rate": 8.060684210526315e-05,
"loss": 23.0,
"step": 47
},
{
"epoch": 0.0009662127479694435,
"grad_norm": 3.86920764867682e-05,
"learning_rate": 8.008e-05,
"loss": 23.0,
"step": 48
},
{
"epoch": 0.000986342180218807,
"grad_norm": 8.165750477928668e-05,
"learning_rate": 7.955315789473684e-05,
"loss": 23.0,
"step": 49
},
{
"epoch": 0.0010064716124681702,
"grad_norm": 5.115772000863217e-05,
"learning_rate": 7.902631578947368e-05,
"loss": 23.0,
"step": 50
},
{
"epoch": 0.0010064716124681702,
"eval_loss": 11.5,
"eval_runtime": 286.0317,
"eval_samples_per_second": 73.132,
"eval_steps_per_second": 18.285,
"step": 50
},
{
"epoch": 0.0010266010447175337,
"grad_norm": 8.701150363776833e-06,
"learning_rate": 7.849947368421052e-05,
"loss": 23.0,
"step": 51
},
{
"epoch": 0.0010467304769668972,
"grad_norm": 1.5078923752298579e-05,
"learning_rate": 7.797263157894736e-05,
"loss": 23.0,
"step": 52
},
{
"epoch": 0.0010668599092162605,
"grad_norm": 1.1019976227544248e-05,
"learning_rate": 7.744578947368421e-05,
"loss": 23.0,
"step": 53
},
{
"epoch": 0.001086989341465624,
"grad_norm": 1.69733048096532e-05,
"learning_rate": 7.691894736842104e-05,
"loss": 23.0,
"step": 54
},
{
"epoch": 0.0011071187737149873,
"grad_norm": 2.840134948201012e-05,
"learning_rate": 7.63921052631579e-05,
"loss": 23.0,
"step": 55
},
{
"epoch": 0.0011272482059643508,
"grad_norm": 3.071267929044552e-05,
"learning_rate": 7.586526315789473e-05,
"loss": 23.0,
"step": 56
},
{
"epoch": 0.0011473776382137143,
"grad_norm": 1.642799179535359e-05,
"learning_rate": 7.533842105263158e-05,
"loss": 23.0,
"step": 57
},
{
"epoch": 0.0011675070704630776,
"grad_norm": 2.4320828742929734e-05,
"learning_rate": 7.481157894736841e-05,
"loss": 23.0,
"step": 58
},
{
"epoch": 0.001187636502712441,
"grad_norm": 2.2044352590455674e-05,
"learning_rate": 7.428473684210526e-05,
"loss": 23.0,
"step": 59
},
{
"epoch": 0.0012077659349618043,
"grad_norm": 5.950014747213572e-05,
"learning_rate": 7.375789473684209e-05,
"loss": 23.0,
"step": 60
},
{
"epoch": 0.0012278953672111678,
"grad_norm": 1.5688616258557886e-05,
"learning_rate": 7.323105263157895e-05,
"loss": 23.0,
"step": 61
},
{
"epoch": 0.0012480247994605311,
"grad_norm": 1.5980189346009865e-05,
"learning_rate": 7.270421052631578e-05,
"loss": 23.0,
"step": 62
},
{
"epoch": 0.0012681542317098946,
"grad_norm": 2.3954600692377426e-05,
"learning_rate": 7.217736842105263e-05,
"loss": 23.0,
"step": 63
},
{
"epoch": 0.0012882836639592581,
"grad_norm": 6.61021113046445e-05,
"learning_rate": 7.165052631578947e-05,
"loss": 23.0,
"step": 64
},
{
"epoch": 0.0013084130962086214,
"grad_norm": 8.342272485606372e-05,
"learning_rate": 7.11236842105263e-05,
"loss": 23.0,
"step": 65
},
{
"epoch": 0.0013285425284579849,
"grad_norm": 3.27210254909005e-05,
"learning_rate": 7.059684210526315e-05,
"loss": 23.0,
"step": 66
},
{
"epoch": 0.0013486719607073482,
"grad_norm": 3.347166421008296e-05,
"learning_rate": 7.006999999999998e-05,
"loss": 23.0,
"step": 67
},
{
"epoch": 0.0013688013929567117,
"grad_norm": 3.555112925823778e-05,
"learning_rate": 6.954315789473684e-05,
"loss": 23.0,
"step": 68
},
{
"epoch": 0.0013889308252060752,
"grad_norm": 4.155048372922465e-05,
"learning_rate": 6.901631578947368e-05,
"loss": 23.0,
"step": 69
},
{
"epoch": 0.0014090602574554384,
"grad_norm": 3.058008223888464e-05,
"learning_rate": 6.848947368421052e-05,
"loss": 23.0,
"step": 70
},
{
"epoch": 0.001429189689704802,
"grad_norm": 3.1038107408676296e-05,
"learning_rate": 6.796263157894737e-05,
"loss": 23.0,
"step": 71
},
{
"epoch": 0.0014493191219541652,
"grad_norm": 2.574123391241301e-05,
"learning_rate": 6.74357894736842e-05,
"loss": 23.0,
"step": 72
},
{
"epoch": 0.0014694485542035287,
"grad_norm": 8.571895159548149e-05,
"learning_rate": 6.690894736842105e-05,
"loss": 23.0,
"step": 73
},
{
"epoch": 0.0014895779864528922,
"grad_norm": 8.358476043213159e-05,
"learning_rate": 6.638210526315788e-05,
"loss": 23.0,
"step": 74
},
{
"epoch": 0.0015097074187022555,
"grad_norm": 5.286915256874636e-05,
"learning_rate": 6.585526315789474e-05,
"loss": 23.0,
"step": 75
},
{
"epoch": 0.001529836850951619,
"grad_norm": 2.435395981592592e-05,
"learning_rate": 6.532842105263157e-05,
"loss": 23.0,
"step": 76
},
{
"epoch": 0.0015499662832009823,
"grad_norm": 5.83510372962337e-05,
"learning_rate": 6.480157894736842e-05,
"loss": 23.0,
"step": 77
},
{
"epoch": 0.0015700957154503458,
"grad_norm": 4.358484875410795e-05,
"learning_rate": 6.427473684210526e-05,
"loss": 23.0,
"step": 78
},
{
"epoch": 0.001590225147699709,
"grad_norm": 4.313517638365738e-05,
"learning_rate": 6.37478947368421e-05,
"loss": 23.0,
"step": 79
},
{
"epoch": 0.0016103545799490725,
"grad_norm": 0.00010166852734982967,
"learning_rate": 6.322105263157894e-05,
"loss": 23.0,
"step": 80
},
{
"epoch": 0.001630484012198436,
"grad_norm": 7.244614243973047e-05,
"learning_rate": 6.269421052631577e-05,
"loss": 23.0,
"step": 81
},
{
"epoch": 0.0016506134444477993,
"grad_norm": 4.1063092794502154e-05,
"learning_rate": 6.216736842105263e-05,
"loss": 23.0,
"step": 82
},
{
"epoch": 0.0016707428766971628,
"grad_norm": 4.716099283541553e-05,
"learning_rate": 6.164052631578947e-05,
"loss": 23.0,
"step": 83
},
{
"epoch": 0.001690872308946526,
"grad_norm": 5.5708838772261515e-05,
"learning_rate": 6.111368421052631e-05,
"loss": 23.0,
"step": 84
},
{
"epoch": 0.0017110017411958896,
"grad_norm": 3.410178396734409e-05,
"learning_rate": 6.058684210526315e-05,
"loss": 23.0,
"step": 85
},
{
"epoch": 0.001731131173445253,
"grad_norm": 0.00011276966688456014,
"learning_rate": 6.005999999999999e-05,
"loss": 23.0,
"step": 86
},
{
"epoch": 0.0017512606056946164,
"grad_norm": 5.103731746203266e-05,
"learning_rate": 5.953315789473684e-05,
"loss": 23.0,
"step": 87
},
{
"epoch": 0.0017713900379439798,
"grad_norm": 9.84208527370356e-05,
"learning_rate": 5.9006315789473676e-05,
"loss": 23.0,
"step": 88
},
{
"epoch": 0.0017915194701933431,
"grad_norm": 7.447129610227421e-05,
"learning_rate": 5.847947368421053e-05,
"loss": 23.0,
"step": 89
},
{
"epoch": 0.0018116489024427066,
"grad_norm": 4.3801312131108716e-05,
"learning_rate": 5.795263157894737e-05,
"loss": 23.0,
"step": 90
},
{
"epoch": 0.0018317783346920701,
"grad_norm": 9.384901204612106e-05,
"learning_rate": 5.742578947368421e-05,
"loss": 23.0,
"step": 91
},
{
"epoch": 0.0018519077669414334,
"grad_norm": 9.168239921564236e-05,
"learning_rate": 5.6898947368421046e-05,
"loss": 23.0,
"step": 92
},
{
"epoch": 0.001872037199190797,
"grad_norm": 9.009746281662956e-05,
"learning_rate": 5.6372105263157886e-05,
"loss": 23.0,
"step": 93
},
{
"epoch": 0.0018921666314401602,
"grad_norm": 7.14911293471232e-05,
"learning_rate": 5.584526315789473e-05,
"loss": 23.0,
"step": 94
},
{
"epoch": 0.0019122960636895237,
"grad_norm": 5.316069655236788e-05,
"learning_rate": 5.531842105263158e-05,
"loss": 23.0,
"step": 95
},
{
"epoch": 0.001932425495938887,
"grad_norm": 0.00012328648881521076,
"learning_rate": 5.4791578947368424e-05,
"loss": 23.0,
"step": 96
},
{
"epoch": 0.0019525549281882504,
"grad_norm": 6.0143935115775093e-05,
"learning_rate": 5.426473684210526e-05,
"loss": 23.0,
"step": 97
},
{
"epoch": 0.001972684360437614,
"grad_norm": 6.84624828863889e-05,
"learning_rate": 5.37378947368421e-05,
"loss": 23.0,
"step": 98
},
{
"epoch": 0.0019928137926869772,
"grad_norm": 0.00011424336116760969,
"learning_rate": 5.321105263157894e-05,
"loss": 23.0,
"step": 99
},
{
"epoch": 0.0020129432249363405,
"grad_norm": 9.294460323872045e-05,
"learning_rate": 5.268421052631578e-05,
"loss": 23.0,
"step": 100
},
{
"epoch": 0.0020129432249363405,
"eval_loss": 11.5,
"eval_runtime": 286.2094,
"eval_samples_per_second": 73.086,
"eval_steps_per_second": 18.273,
"step": 100
},
{
"epoch": 0.002033072657185704,
"grad_norm": 3.381038914085366e-05,
"learning_rate": 5.2157368421052626e-05,
"loss": 23.0,
"step": 101
},
{
"epoch": 0.0020532020894350675,
"grad_norm": 1.918479938467499e-05,
"learning_rate": 5.163052631578947e-05,
"loss": 23.0,
"step": 102
},
{
"epoch": 0.0020733315216844308,
"grad_norm": 3.096737418673001e-05,
"learning_rate": 5.110368421052632e-05,
"loss": 23.0,
"step": 103
},
{
"epoch": 0.0020934609539337945,
"grad_norm": 3.554885552148335e-05,
"learning_rate": 5.057684210526316e-05,
"loss": 23.0,
"step": 104
},
{
"epoch": 0.0021135903861831578,
"grad_norm": 2.8006592401652597e-05,
"learning_rate": 5.005e-05,
"loss": 23.0,
"step": 105
},
{
"epoch": 0.002133719818432521,
"grad_norm": 6.277972715906799e-05,
"learning_rate": 4.9523157894736836e-05,
"loss": 23.0,
"step": 106
},
{
"epoch": 0.0021538492506818843,
"grad_norm": 2.3201017029350623e-05,
"learning_rate": 4.899631578947368e-05,
"loss": 23.0,
"step": 107
},
{
"epoch": 0.002173978682931248,
"grad_norm": 3.592470602598041e-05,
"learning_rate": 4.846947368421052e-05,
"loss": 23.0,
"step": 108
},
{
"epoch": 0.0021941081151806113,
"grad_norm": 3.192729491274804e-05,
"learning_rate": 4.794263157894737e-05,
"loss": 23.0,
"step": 109
},
{
"epoch": 0.0022142375474299746,
"grad_norm": 4.533526589511894e-05,
"learning_rate": 4.7415789473684206e-05,
"loss": 23.0,
"step": 110
},
{
"epoch": 0.0022343669796793383,
"grad_norm": 4.3611438741208985e-05,
"learning_rate": 4.688894736842105e-05,
"loss": 23.0,
"step": 111
},
{
"epoch": 0.0022544964119287016,
"grad_norm": 3.516370270517655e-05,
"learning_rate": 4.636210526315789e-05,
"loss": 23.0,
"step": 112
},
{
"epoch": 0.002274625844178065,
"grad_norm": 4.535488551482558e-05,
"learning_rate": 4.583526315789473e-05,
"loss": 23.0,
"step": 113
},
{
"epoch": 0.0022947552764274286,
"grad_norm": 5.321178468875587e-05,
"learning_rate": 4.530842105263158e-05,
"loss": 23.0,
"step": 114
},
{
"epoch": 0.002314884708676792,
"grad_norm": 5.2021696319570765e-05,
"learning_rate": 4.4781578947368416e-05,
"loss": 23.0,
"step": 115
},
{
"epoch": 0.002335014140926155,
"grad_norm": 2.9661137887160294e-05,
"learning_rate": 4.425473684210526e-05,
"loss": 23.0,
"step": 116
},
{
"epoch": 0.0023551435731755184,
"grad_norm": 5.244681960903108e-05,
"learning_rate": 4.372789473684211e-05,
"loss": 23.0,
"step": 117
},
{
"epoch": 0.002375273005424882,
"grad_norm": 3.523652776493691e-05,
"learning_rate": 4.320105263157895e-05,
"loss": 23.0,
"step": 118
},
{
"epoch": 0.0023954024376742454,
"grad_norm": 4.800470924237743e-05,
"learning_rate": 4.2674210526315786e-05,
"loss": 23.0,
"step": 119
},
{
"epoch": 0.0024155318699236087,
"grad_norm": 6.487173232017085e-05,
"learning_rate": 4.2147368421052626e-05,
"loss": 23.0,
"step": 120
},
{
"epoch": 0.0024356613021729724,
"grad_norm": 1.7037045836332254e-05,
"learning_rate": 4.162052631578947e-05,
"loss": 23.0,
"step": 121
},
{
"epoch": 0.0024557907344223357,
"grad_norm": 4.9463742470834404e-05,
"learning_rate": 4.109368421052631e-05,
"loss": 23.0,
"step": 122
},
{
"epoch": 0.002475920166671699,
"grad_norm": 7.942127558635548e-05,
"learning_rate": 4.056684210526316e-05,
"loss": 23.0,
"step": 123
},
{
"epoch": 0.0024960495989210622,
"grad_norm": 4.070031354785897e-05,
"learning_rate": 4.004e-05,
"loss": 23.0,
"step": 124
},
{
"epoch": 0.002516179031170426,
"grad_norm": 4.9068483349401504e-05,
"learning_rate": 3.951315789473684e-05,
"loss": 23.0,
"step": 125
},
{
"epoch": 0.0025363084634197892,
"grad_norm": 5.625684207188897e-05,
"learning_rate": 3.898631578947368e-05,
"loss": 23.0,
"step": 126
},
{
"epoch": 0.0025564378956691525,
"grad_norm": 6.975179712753743e-05,
"learning_rate": 3.845947368421052e-05,
"loss": 23.0,
"step": 127
},
{
"epoch": 0.0025765673279185162,
"grad_norm": 7.948848360683769e-05,
"learning_rate": 3.7932631578947367e-05,
"loss": 23.0,
"step": 128
},
{
"epoch": 0.0025966967601678795,
"grad_norm": 0.00014242090401239693,
"learning_rate": 3.7405789473684206e-05,
"loss": 23.0,
"step": 129
},
{
"epoch": 0.0026168261924172428,
"grad_norm": 8.259451715275645e-05,
"learning_rate": 3.6878947368421045e-05,
"loss": 23.0,
"step": 130
},
{
"epoch": 0.0026369556246666065,
"grad_norm": 3.79355187760666e-05,
"learning_rate": 3.635210526315789e-05,
"loss": 23.0,
"step": 131
},
{
"epoch": 0.0026570850569159698,
"grad_norm": 5.7715558796189725e-05,
"learning_rate": 3.582526315789474e-05,
"loss": 23.0,
"step": 132
},
{
"epoch": 0.002677214489165333,
"grad_norm": 6.186547398101538e-05,
"learning_rate": 3.5298421052631576e-05,
"loss": 23.0,
"step": 133
},
{
"epoch": 0.0026973439214146963,
"grad_norm": 7.168661977630109e-05,
"learning_rate": 3.477157894736842e-05,
"loss": 23.0,
"step": 134
},
{
"epoch": 0.00271747335366406,
"grad_norm": 0.00012536440044641495,
"learning_rate": 3.424473684210526e-05,
"loss": 23.0,
"step": 135
},
{
"epoch": 0.0027376027859134233,
"grad_norm": 0.00012539015733636916,
"learning_rate": 3.37178947368421e-05,
"loss": 23.0,
"step": 136
},
{
"epoch": 0.0027577322181627866,
"grad_norm": 7.686160824960098e-05,
"learning_rate": 3.319105263157894e-05,
"loss": 23.0,
"step": 137
},
{
"epoch": 0.0027778616504121503,
"grad_norm": 9.55721116042696e-05,
"learning_rate": 3.2664210526315786e-05,
"loss": 23.0,
"step": 138
},
{
"epoch": 0.0027979910826615136,
"grad_norm": 5.936785601079464e-05,
"learning_rate": 3.213736842105263e-05,
"loss": 23.0,
"step": 139
},
{
"epoch": 0.002818120514910877,
"grad_norm": 7.337985880440101e-05,
"learning_rate": 3.161052631578947e-05,
"loss": 23.0,
"step": 140
},
{
"epoch": 0.00283824994716024,
"grad_norm": 9.842724102782086e-05,
"learning_rate": 3.108368421052632e-05,
"loss": 23.0,
"step": 141
},
{
"epoch": 0.002858379379409604,
"grad_norm": 8.842528040986508e-05,
"learning_rate": 3.0556842105263156e-05,
"loss": 23.0,
"step": 142
},
{
"epoch": 0.002878508811658967,
"grad_norm": 9.852414950728416e-05,
"learning_rate": 3.0029999999999995e-05,
"loss": 23.0,
"step": 143
},
{
"epoch": 0.0028986382439083304,
"grad_norm": 0.00011672433174680918,
"learning_rate": 2.9503157894736838e-05,
"loss": 23.0,
"step": 144
},
{
"epoch": 0.002918767676157694,
"grad_norm": 0.00013872129784431309,
"learning_rate": 2.8976315789473684e-05,
"loss": 23.0,
"step": 145
},
{
"epoch": 0.0029388971084070574,
"grad_norm": 5.50843178643845e-05,
"learning_rate": 2.8449473684210523e-05,
"loss": 23.0,
"step": 146
},
{
"epoch": 0.0029590265406564207,
"grad_norm": 0.00015800488472450525,
"learning_rate": 2.7922631578947366e-05,
"loss": 23.0,
"step": 147
},
{
"epoch": 0.0029791559729057844,
"grad_norm": 0.0001110038356273435,
"learning_rate": 2.7395789473684212e-05,
"loss": 23.0,
"step": 148
},
{
"epoch": 0.0029992854051551477,
"grad_norm": 0.00022968508710619062,
"learning_rate": 2.686894736842105e-05,
"loss": 23.0,
"step": 149
},
{
"epoch": 0.003019414837404511,
"grad_norm": 0.00013045659579802305,
"learning_rate": 2.634210526315789e-05,
"loss": 23.0,
"step": 150
},
{
"epoch": 0.003019414837404511,
"eval_loss": 11.5,
"eval_runtime": 285.9436,
"eval_samples_per_second": 73.154,
"eval_steps_per_second": 18.29,
"step": 150
},
{
"epoch": 0.0030395442696538742,
"grad_norm": 2.7813306587631814e-05,
"learning_rate": 2.5815263157894736e-05,
"loss": 23.0,
"step": 151
},
{
"epoch": 0.003059673701903238,
"grad_norm": 4.039341729367152e-05,
"learning_rate": 2.528842105263158e-05,
"loss": 23.0,
"step": 152
},
{
"epoch": 0.0030798031341526012,
"grad_norm": 8.387907291762531e-05,
"learning_rate": 2.4761578947368418e-05,
"loss": 23.0,
"step": 153
},
{
"epoch": 0.0030999325664019645,
"grad_norm": 3.360381379025057e-05,
"learning_rate": 2.423473684210526e-05,
"loss": 23.0,
"step": 154
},
{
"epoch": 0.0031200619986513282,
"grad_norm": 3.189184280927293e-05,
"learning_rate": 2.3707894736842103e-05,
"loss": 23.0,
"step": 155
},
{
"epoch": 0.0031401914309006915,
"grad_norm": 2.3833928935346194e-05,
"learning_rate": 2.3181052631578946e-05,
"loss": 23.0,
"step": 156
},
{
"epoch": 0.003160320863150055,
"grad_norm": 3.9710404962534085e-05,
"learning_rate": 2.265421052631579e-05,
"loss": 23.0,
"step": 157
},
{
"epoch": 0.003180450295399418,
"grad_norm": 1.907577461679466e-05,
"learning_rate": 2.212736842105263e-05,
"loss": 23.0,
"step": 158
},
{
"epoch": 0.0032005797276487818,
"grad_norm": 2.2743917725165375e-05,
"learning_rate": 2.1600526315789474e-05,
"loss": 23.0,
"step": 159
},
{
"epoch": 0.003220709159898145,
"grad_norm": 3.851862857118249e-05,
"learning_rate": 2.1073684210526313e-05,
"loss": 23.0,
"step": 160
},
{
"epoch": 0.0032408385921475083,
"grad_norm": 4.9293492338620126e-05,
"learning_rate": 2.0546842105263155e-05,
"loss": 23.0,
"step": 161
},
{
"epoch": 0.003260968024396872,
"grad_norm": 5.948842954239808e-05,
"learning_rate": 2.002e-05,
"loss": 23.0,
"step": 162
},
{
"epoch": 0.0032810974566462353,
"grad_norm": 5.252346454653889e-05,
"learning_rate": 1.949315789473684e-05,
"loss": 23.0,
"step": 163
},
{
"epoch": 0.0033012268888955986,
"grad_norm": 3.786065281019546e-05,
"learning_rate": 1.8966315789473683e-05,
"loss": 23.0,
"step": 164
},
{
"epoch": 0.0033213563211449623,
"grad_norm": 6.799264519941062e-05,
"learning_rate": 1.8439473684210522e-05,
"loss": 23.0,
"step": 165
},
{
"epoch": 0.0033414857533943256,
"grad_norm": 4.5396624045679346e-05,
"learning_rate": 1.791263157894737e-05,
"loss": 23.0,
"step": 166
},
{
"epoch": 0.003361615185643689,
"grad_norm": 6.286487041506916e-05,
"learning_rate": 1.738578947368421e-05,
"loss": 23.0,
"step": 167
},
{
"epoch": 0.003381744617893052,
"grad_norm": 6.100681275711395e-05,
"learning_rate": 1.685894736842105e-05,
"loss": 23.0,
"step": 168
},
{
"epoch": 0.003401874050142416,
"grad_norm": 5.2501116442726925e-05,
"learning_rate": 1.6332105263157893e-05,
"loss": 23.0,
"step": 169
},
{
"epoch": 0.003422003482391779,
"grad_norm": 8.918352978071198e-05,
"learning_rate": 1.5805263157894735e-05,
"loss": 23.0,
"step": 170
},
{
"epoch": 0.0034421329146411424,
"grad_norm": 5.338154733181e-05,
"learning_rate": 1.5278421052631578e-05,
"loss": 23.0,
"step": 171
},
{
"epoch": 0.003462262346890506,
"grad_norm": 5.688237797585316e-05,
"learning_rate": 1.4751578947368419e-05,
"loss": 23.0,
"step": 172
},
{
"epoch": 0.0034823917791398694,
"grad_norm": 5.872304245713167e-05,
"learning_rate": 1.4224736842105262e-05,
"loss": 23.0,
"step": 173
},
{
"epoch": 0.0035025212113892327,
"grad_norm": 7.459277549060062e-05,
"learning_rate": 1.3697894736842106e-05,
"loss": 23.0,
"step": 174
},
{
"epoch": 0.003522650643638596,
"grad_norm": 6.949468661332503e-05,
"learning_rate": 1.3171052631578945e-05,
"loss": 23.0,
"step": 175
},
{
"epoch": 0.0035427800758879597,
"grad_norm": 4.5293236325960606e-05,
"learning_rate": 1.264421052631579e-05,
"loss": 23.0,
"step": 176
},
{
"epoch": 0.003562909508137323,
"grad_norm": 8.453910413663834e-05,
"learning_rate": 1.211736842105263e-05,
"loss": 23.0,
"step": 177
},
{
"epoch": 0.0035830389403866863,
"grad_norm": 4.495406028581783e-05,
"learning_rate": 1.1590526315789473e-05,
"loss": 23.0,
"step": 178
},
{
"epoch": 0.00360316837263605,
"grad_norm": 8.007752330740914e-05,
"learning_rate": 1.1063684210526316e-05,
"loss": 23.0,
"step": 179
},
{
"epoch": 0.0036232978048854132,
"grad_norm": 6.319613748928532e-05,
"learning_rate": 1.0536842105263156e-05,
"loss": 23.0,
"step": 180
},
{
"epoch": 0.0036434272371347765,
"grad_norm": 9.205697278957814e-05,
"learning_rate": 1.001e-05,
"loss": 23.0,
"step": 181
},
{
"epoch": 0.0036635566693841402,
"grad_norm": 4.999395605409518e-05,
"learning_rate": 9.483157894736842e-06,
"loss": 23.0,
"step": 182
},
{
"epoch": 0.0036836861016335035,
"grad_norm": 6.195960304467008e-05,
"learning_rate": 8.956315789473684e-06,
"loss": 23.0,
"step": 183
},
{
"epoch": 0.003703815533882867,
"grad_norm": 0.00010328808275517076,
"learning_rate": 8.429473684210525e-06,
"loss": 23.0,
"step": 184
},
{
"epoch": 0.00372394496613223,
"grad_norm": 0.00018768337031360716,
"learning_rate": 7.902631578947368e-06,
"loss": 23.0,
"step": 185
},
{
"epoch": 0.003744074398381594,
"grad_norm": 6.907639908604324e-05,
"learning_rate": 7.3757894736842095e-06,
"loss": 23.0,
"step": 186
},
{
"epoch": 0.003764203830630957,
"grad_norm": 6.577693420695141e-05,
"learning_rate": 6.848947368421053e-06,
"loss": 23.0,
"step": 187
},
{
"epoch": 0.0037843332628803203,
"grad_norm": 0.0001368712109979242,
"learning_rate": 6.322105263157895e-06,
"loss": 23.0,
"step": 188
},
{
"epoch": 0.003804462695129684,
"grad_norm": 0.0001247082109330222,
"learning_rate": 5.7952631578947365e-06,
"loss": 23.0,
"step": 189
},
{
"epoch": 0.0038245921273790473,
"grad_norm": 8.381524821743369e-05,
"learning_rate": 5.268421052631578e-06,
"loss": 23.0,
"step": 190
},
{
"epoch": 0.0038447215596284106,
"grad_norm": 5.707381205866113e-05,
"learning_rate": 4.741578947368421e-06,
"loss": 23.0,
"step": 191
},
{
"epoch": 0.003864850991877774,
"grad_norm": 8.412449824390933e-05,
"learning_rate": 4.2147368421052626e-06,
"loss": 23.0,
"step": 192
},
{
"epoch": 0.0038849804241271376,
"grad_norm": 0.00010249995830236003,
"learning_rate": 3.6878947368421047e-06,
"loss": 23.0,
"step": 193
},
{
"epoch": 0.003905109856376501,
"grad_norm": 8.074963989201933e-05,
"learning_rate": 3.1610526315789474e-06,
"loss": 23.0,
"step": 194
},
{
"epoch": 0.003925239288625864,
"grad_norm": 0.00014768572873435915,
"learning_rate": 2.634210526315789e-06,
"loss": 23.0,
"step": 195
},
{
"epoch": 0.003945368720875228,
"grad_norm": 6.658612983301282e-05,
"learning_rate": 2.1073684210526313e-06,
"loss": 23.0,
"step": 196
},
{
"epoch": 0.003965498153124591,
"grad_norm": 0.0001054863168974407,
"learning_rate": 1.5805263157894737e-06,
"loss": 23.0,
"step": 197
},
{
"epoch": 0.0039856275853739544,
"grad_norm": 0.00012067404168192297,
"learning_rate": 1.0536842105263156e-06,
"loss": 23.0,
"step": 198
},
{
"epoch": 0.004005757017623318,
"grad_norm": 0.0001297133567277342,
"learning_rate": 5.268421052631578e-07,
"loss": 23.0,
"step": 199
},
{
"epoch": 0.004025886449872681,
"grad_norm": 0.000187472251127474,
"learning_rate": 0.0,
"loss": 23.0,
"step": 200
},
{
"epoch": 0.004025886449872681,
"eval_loss": 11.5,
"eval_runtime": 286.1718,
"eval_samples_per_second": 73.096,
"eval_steps_per_second": 18.276,
"step": 200
}
],
"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": 3
}
},
"TrainerControl": {
"args": {
"should_epoch_stop": false,
"should_evaluate": false,
"should_log": false,
"should_save": true,
"should_training_stop": true
},
"attributes": {}
}
},
"total_flos": 8280804163584.0,
"train_batch_size": 4,
"trial_name": null,
"trial_params": null
}