dimasik2987's picture
Training in progress, step 200, checkpoint
a4cc62a verified
raw
history blame
37.1 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": 1.6507690816069953e-05,
"learning_rate": 1.004e-05,
"loss": 23.0,
"step": 1
},
{
"epoch": 2.0129432249363408e-05,
"eval_loss": 11.5,
"eval_runtime": 285.9668,
"eval_samples_per_second": 73.148,
"eval_steps_per_second": 18.289,
"step": 1
},
{
"epoch": 4.0258864498726816e-05,
"grad_norm": 1.639904803596437e-05,
"learning_rate": 2.008e-05,
"loss": 23.0,
"step": 2
},
{
"epoch": 6.038829674809022e-05,
"grad_norm": 2.0076164219062775e-05,
"learning_rate": 3.012e-05,
"loss": 23.0,
"step": 3
},
{
"epoch": 8.051772899745363e-05,
"grad_norm": 9.847400178841781e-06,
"learning_rate": 4.016e-05,
"loss": 23.0,
"step": 4
},
{
"epoch": 0.00010064716124681703,
"grad_norm": 9.781237167771906e-06,
"learning_rate": 5.02e-05,
"loss": 23.0,
"step": 5
},
{
"epoch": 0.00012077659349618043,
"grad_norm": 2.2412741600419395e-05,
"learning_rate": 6.024e-05,
"loss": 23.0,
"step": 6
},
{
"epoch": 0.00014090602574554385,
"grad_norm": 1.6903981304494664e-05,
"learning_rate": 7.028e-05,
"loss": 23.0,
"step": 7
},
{
"epoch": 0.00016103545799490726,
"grad_norm": 2.2363303287420422e-05,
"learning_rate": 8.032e-05,
"loss": 23.0,
"step": 8
},
{
"epoch": 0.00018116489024427065,
"grad_norm": 1.0626799848978408e-05,
"learning_rate": 9.036000000000001e-05,
"loss": 23.0,
"step": 9
},
{
"epoch": 0.00020129432249363407,
"grad_norm": 1.4076299521548208e-05,
"learning_rate": 0.0001004,
"loss": 23.0,
"step": 10
},
{
"epoch": 0.00022142375474299748,
"grad_norm": 2.3633083401364274e-05,
"learning_rate": 9.987157894736842e-05,
"loss": 23.0,
"step": 11
},
{
"epoch": 0.00024155318699236087,
"grad_norm": 2.4257647964986973e-05,
"learning_rate": 9.934315789473684e-05,
"loss": 23.0,
"step": 12
},
{
"epoch": 0.0002616826192417243,
"grad_norm": 2.6146915843128227e-05,
"learning_rate": 9.881473684210525e-05,
"loss": 23.0,
"step": 13
},
{
"epoch": 0.0002818120514910877,
"grad_norm": 1.611110747035127e-05,
"learning_rate": 9.828631578947369e-05,
"loss": 23.0,
"step": 14
},
{
"epoch": 0.0003019414837404511,
"grad_norm": 1.668692311795894e-05,
"learning_rate": 9.77578947368421e-05,
"loss": 23.0,
"step": 15
},
{
"epoch": 0.00032207091598981453,
"grad_norm": 2.7145209969603457e-05,
"learning_rate": 9.722947368421052e-05,
"loss": 23.0,
"step": 16
},
{
"epoch": 0.0003422003482391779,
"grad_norm": 1.7148986444226466e-05,
"learning_rate": 9.670105263157895e-05,
"loss": 23.0,
"step": 17
},
{
"epoch": 0.0003623297804885413,
"grad_norm": 1.813442213460803e-05,
"learning_rate": 9.617263157894737e-05,
"loss": 23.0,
"step": 18
},
{
"epoch": 0.00038245921273790474,
"grad_norm": 3.107384327449836e-05,
"learning_rate": 9.564421052631579e-05,
"loss": 23.0,
"step": 19
},
{
"epoch": 0.00040258864498726813,
"grad_norm": 2.8771428333129734e-05,
"learning_rate": 9.511578947368421e-05,
"loss": 23.0,
"step": 20
},
{
"epoch": 0.0004227180772366315,
"grad_norm": 1.3413990927801933e-05,
"learning_rate": 9.458736842105264e-05,
"loss": 23.0,
"step": 21
},
{
"epoch": 0.00044284750948599496,
"grad_norm": 1.6914329535211436e-05,
"learning_rate": 9.405894736842106e-05,
"loss": 23.0,
"step": 22
},
{
"epoch": 0.00046297694173535835,
"grad_norm": 1.504943247709889e-05,
"learning_rate": 9.353052631578947e-05,
"loss": 23.0,
"step": 23
},
{
"epoch": 0.00048310637398472174,
"grad_norm": 2.3936727302498184e-05,
"learning_rate": 9.300210526315789e-05,
"loss": 23.0,
"step": 24
},
{
"epoch": 0.0005032358062340851,
"grad_norm": 4.2899515392491594e-05,
"learning_rate": 9.247368421052631e-05,
"loss": 23.0,
"step": 25
},
{
"epoch": 0.0005233652384834486,
"grad_norm": 1.9880957552231848e-05,
"learning_rate": 9.194526315789473e-05,
"loss": 23.0,
"step": 26
},
{
"epoch": 0.000543494670732812,
"grad_norm": 2.101472273352556e-05,
"learning_rate": 9.141684210526316e-05,
"loss": 23.0,
"step": 27
},
{
"epoch": 0.0005636241029821754,
"grad_norm": 2.740858144534286e-05,
"learning_rate": 9.088842105263158e-05,
"loss": 23.0,
"step": 28
},
{
"epoch": 0.0005837535352315388,
"grad_norm": 8.637041173642501e-05,
"learning_rate": 9.036000000000001e-05,
"loss": 23.0,
"step": 29
},
{
"epoch": 0.0006038829674809022,
"grad_norm": 4.428947431733832e-05,
"learning_rate": 8.983157894736843e-05,
"loss": 23.0,
"step": 30
},
{
"epoch": 0.0006240123997302656,
"grad_norm": 6.015045801177621e-05,
"learning_rate": 8.930315789473684e-05,
"loss": 23.0,
"step": 31
},
{
"epoch": 0.0006441418319796291,
"grad_norm": 5.055875953985378e-05,
"learning_rate": 8.877473684210526e-05,
"loss": 23.0,
"step": 32
},
{
"epoch": 0.0006642712642289924,
"grad_norm": 3.0454708394245245e-05,
"learning_rate": 8.824631578947368e-05,
"loss": 23.0,
"step": 33
},
{
"epoch": 0.0006844006964783558,
"grad_norm": 5.01158538099844e-05,
"learning_rate": 8.771789473684211e-05,
"loss": 23.0,
"step": 34
},
{
"epoch": 0.0007045301287277192,
"grad_norm": 5.602340752375312e-05,
"learning_rate": 8.718947368421053e-05,
"loss": 23.0,
"step": 35
},
{
"epoch": 0.0007246595609770826,
"grad_norm": 3.0846869776723906e-05,
"learning_rate": 8.666105263157895e-05,
"loss": 23.0,
"step": 36
},
{
"epoch": 0.0007447889932264461,
"grad_norm": 5.5015039833961055e-05,
"learning_rate": 8.613263157894737e-05,
"loss": 23.0,
"step": 37
},
{
"epoch": 0.0007649184254758095,
"grad_norm": 3.8054538890719414e-05,
"learning_rate": 8.560421052631578e-05,
"loss": 23.0,
"step": 38
},
{
"epoch": 0.0007850478577251729,
"grad_norm": 2.42203677771613e-05,
"learning_rate": 8.50757894736842e-05,
"loss": 23.0,
"step": 39
},
{
"epoch": 0.0008051772899745363,
"grad_norm": 4.2876483348663896e-05,
"learning_rate": 8.454736842105263e-05,
"loss": 23.0,
"step": 40
},
{
"epoch": 0.0008253067222238997,
"grad_norm": 3.370803460711613e-05,
"learning_rate": 8.401894736842106e-05,
"loss": 23.0,
"step": 41
},
{
"epoch": 0.000845436154473263,
"grad_norm": 4.167450242675841e-05,
"learning_rate": 8.349052631578948e-05,
"loss": 23.0,
"step": 42
},
{
"epoch": 0.0008655655867226265,
"grad_norm": 4.232488572597504e-05,
"learning_rate": 8.29621052631579e-05,
"loss": 23.0,
"step": 43
},
{
"epoch": 0.0008856950189719899,
"grad_norm": 2.6339081159676425e-05,
"learning_rate": 8.243368421052632e-05,
"loss": 23.0,
"step": 44
},
{
"epoch": 0.0009058244512213533,
"grad_norm": 4.41583979409188e-05,
"learning_rate": 8.190526315789474e-05,
"loss": 23.0,
"step": 45
},
{
"epoch": 0.0009259538834707167,
"grad_norm": 7.86347227403894e-05,
"learning_rate": 8.137684210526315e-05,
"loss": 23.0,
"step": 46
},
{
"epoch": 0.0009460833157200801,
"grad_norm": 4.264622475602664e-05,
"learning_rate": 8.084842105263157e-05,
"loss": 23.0,
"step": 47
},
{
"epoch": 0.0009662127479694435,
"grad_norm": 3.8661462895106524e-05,
"learning_rate": 8.032e-05,
"loss": 23.0,
"step": 48
},
{
"epoch": 0.000986342180218807,
"grad_norm": 7.720923167653382e-05,
"learning_rate": 7.979157894736842e-05,
"loss": 23.0,
"step": 49
},
{
"epoch": 0.0010064716124681702,
"grad_norm": 4.70045160909649e-05,
"learning_rate": 7.926315789473684e-05,
"loss": 23.0,
"step": 50
},
{
"epoch": 0.0010064716124681702,
"eval_loss": 11.5,
"eval_runtime": 285.9875,
"eval_samples_per_second": 73.143,
"eval_steps_per_second": 18.288,
"step": 50
},
{
"epoch": 0.0010266010447175337,
"grad_norm": 8.320242159243207e-06,
"learning_rate": 7.873473684210526e-05,
"loss": 23.0,
"step": 51
},
{
"epoch": 0.0010467304769668972,
"grad_norm": 1.5145632460189518e-05,
"learning_rate": 7.820631578947369e-05,
"loss": 23.0,
"step": 52
},
{
"epoch": 0.0010668599092162605,
"grad_norm": 1.1512304809002671e-05,
"learning_rate": 7.76778947368421e-05,
"loss": 23.0,
"step": 53
},
{
"epoch": 0.001086989341465624,
"grad_norm": 1.528786015114747e-05,
"learning_rate": 7.714947368421052e-05,
"loss": 23.0,
"step": 54
},
{
"epoch": 0.0011071187737149873,
"grad_norm": 2.5456898583797738e-05,
"learning_rate": 7.662105263157896e-05,
"loss": 23.0,
"step": 55
},
{
"epoch": 0.0011272482059643508,
"grad_norm": 2.668505476322025e-05,
"learning_rate": 7.609263157894737e-05,
"loss": 23.0,
"step": 56
},
{
"epoch": 0.0011473776382137143,
"grad_norm": 1.551048808323685e-05,
"learning_rate": 7.556421052631579e-05,
"loss": 23.0,
"step": 57
},
{
"epoch": 0.0011675070704630776,
"grad_norm": 2.366452827118337e-05,
"learning_rate": 7.503578947368421e-05,
"loss": 23.0,
"step": 58
},
{
"epoch": 0.001187636502712441,
"grad_norm": 2.2682577764499e-05,
"learning_rate": 7.450736842105263e-05,
"loss": 23.0,
"step": 59
},
{
"epoch": 0.0012077659349618043,
"grad_norm": 5.866533319931477e-05,
"learning_rate": 7.397894736842105e-05,
"loss": 23.0,
"step": 60
},
{
"epoch": 0.0012278953672111678,
"grad_norm": 1.637460991332773e-05,
"learning_rate": 7.345052631578948e-05,
"loss": 23.0,
"step": 61
},
{
"epoch": 0.0012480247994605311,
"grad_norm": 1.6382557078031823e-05,
"learning_rate": 7.29221052631579e-05,
"loss": 23.0,
"step": 62
},
{
"epoch": 0.0012681542317098946,
"grad_norm": 2.3177421098807827e-05,
"learning_rate": 7.239368421052631e-05,
"loss": 23.0,
"step": 63
},
{
"epoch": 0.0012882836639592581,
"grad_norm": 6.291714089456946e-05,
"learning_rate": 7.186526315789474e-05,
"loss": 23.0,
"step": 64
},
{
"epoch": 0.0013084130962086214,
"grad_norm": 7.432617712765932e-05,
"learning_rate": 7.133684210526316e-05,
"loss": 23.0,
"step": 65
},
{
"epoch": 0.0013285425284579849,
"grad_norm": 3.0208946554921567e-05,
"learning_rate": 7.080842105263158e-05,
"loss": 23.0,
"step": 66
},
{
"epoch": 0.0013486719607073482,
"grad_norm": 3.2406318496214226e-05,
"learning_rate": 7.028e-05,
"loss": 23.0,
"step": 67
},
{
"epoch": 0.0013688013929567117,
"grad_norm": 3.3409804018447176e-05,
"learning_rate": 6.975157894736843e-05,
"loss": 23.0,
"step": 68
},
{
"epoch": 0.0013889308252060752,
"grad_norm": 4.153459667577408e-05,
"learning_rate": 6.922315789473685e-05,
"loss": 23.0,
"step": 69
},
{
"epoch": 0.0014090602574554384,
"grad_norm": 3.1347994081443176e-05,
"learning_rate": 6.869473684210527e-05,
"loss": 23.0,
"step": 70
},
{
"epoch": 0.001429189689704802,
"grad_norm": 3.106446092715487e-05,
"learning_rate": 6.816631578947368e-05,
"loss": 23.0,
"step": 71
},
{
"epoch": 0.0014493191219541652,
"grad_norm": 2.5425708372495137e-05,
"learning_rate": 6.76378947368421e-05,
"loss": 23.0,
"step": 72
},
{
"epoch": 0.0014694485542035287,
"grad_norm": 8.160302240867168e-05,
"learning_rate": 6.710947368421052e-05,
"loss": 23.0,
"step": 73
},
{
"epoch": 0.0014895779864528922,
"grad_norm": 8.36909530335106e-05,
"learning_rate": 6.658105263157894e-05,
"loss": 23.0,
"step": 74
},
{
"epoch": 0.0015097074187022555,
"grad_norm": 5.806803164887242e-05,
"learning_rate": 6.605263157894737e-05,
"loss": 23.0,
"step": 75
},
{
"epoch": 0.001529836850951619,
"grad_norm": 2.5655506760813296e-05,
"learning_rate": 6.55242105263158e-05,
"loss": 23.0,
"step": 76
},
{
"epoch": 0.0015499662832009823,
"grad_norm": 5.792943193227984e-05,
"learning_rate": 6.499578947368422e-05,
"loss": 23.0,
"step": 77
},
{
"epoch": 0.0015700957154503458,
"grad_norm": 4.383608029456809e-05,
"learning_rate": 6.446736842105264e-05,
"loss": 23.0,
"step": 78
},
{
"epoch": 0.001590225147699709,
"grad_norm": 4.217608875478618e-05,
"learning_rate": 6.393894736842105e-05,
"loss": 23.0,
"step": 79
},
{
"epoch": 0.0016103545799490725,
"grad_norm": 0.0001017999675241299,
"learning_rate": 6.341052631578947e-05,
"loss": 23.0,
"step": 80
},
{
"epoch": 0.001630484012198436,
"grad_norm": 7.253052172018215e-05,
"learning_rate": 6.288210526315789e-05,
"loss": 23.0,
"step": 81
},
{
"epoch": 0.0016506134444477993,
"grad_norm": 4.4168151362100616e-05,
"learning_rate": 6.235368421052632e-05,
"loss": 23.0,
"step": 82
},
{
"epoch": 0.0016707428766971628,
"grad_norm": 4.690126297646202e-05,
"learning_rate": 6.182526315789474e-05,
"loss": 23.0,
"step": 83
},
{
"epoch": 0.001690872308946526,
"grad_norm": 5.543129373108968e-05,
"learning_rate": 6.129684210526316e-05,
"loss": 23.0,
"step": 84
},
{
"epoch": 0.0017110017411958896,
"grad_norm": 3.422376903472468e-05,
"learning_rate": 6.076842105263158e-05,
"loss": 23.0,
"step": 85
},
{
"epoch": 0.001731131173445253,
"grad_norm": 0.00011534958321135491,
"learning_rate": 6.024e-05,
"loss": 23.0,
"step": 86
},
{
"epoch": 0.0017512606056946164,
"grad_norm": 5.733649231842719e-05,
"learning_rate": 5.971157894736842e-05,
"loss": 23.0,
"step": 87
},
{
"epoch": 0.0017713900379439798,
"grad_norm": 9.903619502438232e-05,
"learning_rate": 5.9183157894736835e-05,
"loss": 23.0,
"step": 88
},
{
"epoch": 0.0017915194701933431,
"grad_norm": 7.510482100769877e-05,
"learning_rate": 5.8654736842105267e-05,
"loss": 23.0,
"step": 89
},
{
"epoch": 0.0018116489024427066,
"grad_norm": 4.391166294226423e-05,
"learning_rate": 5.8126315789473684e-05,
"loss": 23.0,
"step": 90
},
{
"epoch": 0.0018317783346920701,
"grad_norm": 0.00010185052815359086,
"learning_rate": 5.759789473684211e-05,
"loss": 23.0,
"step": 91
},
{
"epoch": 0.0018519077669414334,
"grad_norm": 9.979342576116323e-05,
"learning_rate": 5.706947368421053e-05,
"loss": 23.0,
"step": 92
},
{
"epoch": 0.001872037199190797,
"grad_norm": 9.317202056990936e-05,
"learning_rate": 5.6541052631578945e-05,
"loss": 23.0,
"step": 93
},
{
"epoch": 0.0018921666314401602,
"grad_norm": 7.172553159762174e-05,
"learning_rate": 5.601263157894736e-05,
"loss": 23.0,
"step": 94
},
{
"epoch": 0.0019122960636895237,
"grad_norm": 5.0452064897399396e-05,
"learning_rate": 5.5484210526315794e-05,
"loss": 23.0,
"step": 95
},
{
"epoch": 0.001932425495938887,
"grad_norm": 0.00011866455315612257,
"learning_rate": 5.495578947368421e-05,
"loss": 23.0,
"step": 96
},
{
"epoch": 0.0019525549281882504,
"grad_norm": 6.334228237392381e-05,
"learning_rate": 5.442736842105264e-05,
"loss": 23.0,
"step": 97
},
{
"epoch": 0.001972684360437614,
"grad_norm": 6.854250386822969e-05,
"learning_rate": 5.3898947368421055e-05,
"loss": 23.0,
"step": 98
},
{
"epoch": 0.0019928137926869772,
"grad_norm": 0.00011966691090492532,
"learning_rate": 5.337052631578947e-05,
"loss": 23.0,
"step": 99
},
{
"epoch": 0.0020129432249363405,
"grad_norm": 9.832581417867914e-05,
"learning_rate": 5.284210526315789e-05,
"loss": 23.0,
"step": 100
},
{
"epoch": 0.0020129432249363405,
"eval_loss": 11.5,
"eval_runtime": 286.0829,
"eval_samples_per_second": 73.119,
"eval_steps_per_second": 18.281,
"step": 100
},
{
"epoch": 0.002033072657185704,
"grad_norm": 3.899659350281581e-05,
"learning_rate": 5.231368421052631e-05,
"loss": 23.0,
"step": 101
},
{
"epoch": 0.0020532020894350675,
"grad_norm": 1.9528526536305435e-05,
"learning_rate": 5.178526315789474e-05,
"loss": 23.0,
"step": 102
},
{
"epoch": 0.0020733315216844308,
"grad_norm": 3.4980272175744176e-05,
"learning_rate": 5.1256842105263165e-05,
"loss": 23.0,
"step": 103
},
{
"epoch": 0.0020934609539337945,
"grad_norm": 3.446464324952103e-05,
"learning_rate": 5.072842105263158e-05,
"loss": 23.0,
"step": 104
},
{
"epoch": 0.0021135903861831578,
"grad_norm": 2.7365793357603252e-05,
"learning_rate": 5.02e-05,
"loss": 23.0,
"step": 105
},
{
"epoch": 0.002133719818432521,
"grad_norm": 6.427912740036845e-05,
"learning_rate": 4.967157894736842e-05,
"loss": 23.0,
"step": 106
},
{
"epoch": 0.0021538492506818843,
"grad_norm": 2.2796304619987495e-05,
"learning_rate": 4.914315789473684e-05,
"loss": 23.0,
"step": 107
},
{
"epoch": 0.002173978682931248,
"grad_norm": 3.622789154178463e-05,
"learning_rate": 4.861473684210526e-05,
"loss": 23.0,
"step": 108
},
{
"epoch": 0.0021941081151806113,
"grad_norm": 2.9890994483139366e-05,
"learning_rate": 4.8086315789473686e-05,
"loss": 23.0,
"step": 109
},
{
"epoch": 0.0022142375474299746,
"grad_norm": 4.680438360082917e-05,
"learning_rate": 4.7557894736842104e-05,
"loss": 23.0,
"step": 110
},
{
"epoch": 0.0022343669796793383,
"grad_norm": 4.3207284761592746e-05,
"learning_rate": 4.702947368421053e-05,
"loss": 23.0,
"step": 111
},
{
"epoch": 0.0022544964119287016,
"grad_norm": 3.4467164368834347e-05,
"learning_rate": 4.6501052631578946e-05,
"loss": 23.0,
"step": 112
},
{
"epoch": 0.002274625844178065,
"grad_norm": 4.459037154447287e-05,
"learning_rate": 4.5972631578947364e-05,
"loss": 23.0,
"step": 113
},
{
"epoch": 0.0022947552764274286,
"grad_norm": 5.37286905455403e-05,
"learning_rate": 4.544421052631579e-05,
"loss": 23.0,
"step": 114
},
{
"epoch": 0.002314884708676792,
"grad_norm": 5.072938438388519e-05,
"learning_rate": 4.4915789473684213e-05,
"loss": 23.0,
"step": 115
},
{
"epoch": 0.002335014140926155,
"grad_norm": 3.0260003768489696e-05,
"learning_rate": 4.438736842105263e-05,
"loss": 23.0,
"step": 116
},
{
"epoch": 0.0023551435731755184,
"grad_norm": 5.488552415044978e-05,
"learning_rate": 4.3858947368421056e-05,
"loss": 23.0,
"step": 117
},
{
"epoch": 0.002375273005424882,
"grad_norm": 3.486787682049908e-05,
"learning_rate": 4.3330526315789474e-05,
"loss": 23.0,
"step": 118
},
{
"epoch": 0.0023954024376742454,
"grad_norm": 4.800608439836651e-05,
"learning_rate": 4.280210526315789e-05,
"loss": 23.0,
"step": 119
},
{
"epoch": 0.0024155318699236087,
"grad_norm": 6.630933785345405e-05,
"learning_rate": 4.2273684210526317e-05,
"loss": 23.0,
"step": 120
},
{
"epoch": 0.0024356613021729724,
"grad_norm": 2.0091678379685618e-05,
"learning_rate": 4.174526315789474e-05,
"loss": 23.0,
"step": 121
},
{
"epoch": 0.0024557907344223357,
"grad_norm": 5.247927037999034e-05,
"learning_rate": 4.121684210526316e-05,
"loss": 23.0,
"step": 122
},
{
"epoch": 0.002475920166671699,
"grad_norm": 8.598071872256696e-05,
"learning_rate": 4.068842105263158e-05,
"loss": 23.0,
"step": 123
},
{
"epoch": 0.0024960495989210622,
"grad_norm": 4.097522469237447e-05,
"learning_rate": 4.016e-05,
"loss": 23.0,
"step": 124
},
{
"epoch": 0.002516179031170426,
"grad_norm": 5.331883585313335e-05,
"learning_rate": 3.963157894736842e-05,
"loss": 23.0,
"step": 125
},
{
"epoch": 0.0025363084634197892,
"grad_norm": 5.928654718445614e-05,
"learning_rate": 3.9103157894736844e-05,
"loss": 23.0,
"step": 126
},
{
"epoch": 0.0025564378956691525,
"grad_norm": 6.587509415112436e-05,
"learning_rate": 3.857473684210526e-05,
"loss": 23.0,
"step": 127
},
{
"epoch": 0.0025765673279185162,
"grad_norm": 8.765143866185099e-05,
"learning_rate": 3.804631578947369e-05,
"loss": 23.0,
"step": 128
},
{
"epoch": 0.0025966967601678795,
"grad_norm": 0.0001506681728642434,
"learning_rate": 3.7517894736842105e-05,
"loss": 23.0,
"step": 129
},
{
"epoch": 0.0026168261924172428,
"grad_norm": 8.650772360851988e-05,
"learning_rate": 3.698947368421052e-05,
"loss": 23.0,
"step": 130
},
{
"epoch": 0.0026369556246666065,
"grad_norm": 4.020520282210782e-05,
"learning_rate": 3.646105263157895e-05,
"loss": 23.0,
"step": 131
},
{
"epoch": 0.0026570850569159698,
"grad_norm": 6.1138161981944e-05,
"learning_rate": 3.593263157894737e-05,
"loss": 23.0,
"step": 132
},
{
"epoch": 0.002677214489165333,
"grad_norm": 6.121417391113937e-05,
"learning_rate": 3.540421052631579e-05,
"loss": 23.0,
"step": 133
},
{
"epoch": 0.0026973439214146963,
"grad_norm": 7.352145621553063e-05,
"learning_rate": 3.4875789473684215e-05,
"loss": 23.0,
"step": 134
},
{
"epoch": 0.00271747335366406,
"grad_norm": 0.00014084699796512723,
"learning_rate": 3.434736842105263e-05,
"loss": 23.0,
"step": 135
},
{
"epoch": 0.0027376027859134233,
"grad_norm": 0.0001324174227192998,
"learning_rate": 3.381894736842105e-05,
"loss": 23.0,
"step": 136
},
{
"epoch": 0.0027577322181627866,
"grad_norm": 8.251784311141819e-05,
"learning_rate": 3.329052631578947e-05,
"loss": 23.0,
"step": 137
},
{
"epoch": 0.0027778616504121503,
"grad_norm": 9.526171197649091e-05,
"learning_rate": 3.27621052631579e-05,
"loss": 23.0,
"step": 138
},
{
"epoch": 0.0027979910826615136,
"grad_norm": 6.26833934802562e-05,
"learning_rate": 3.223368421052632e-05,
"loss": 23.0,
"step": 139
},
{
"epoch": 0.002818120514910877,
"grad_norm": 7.219661347335204e-05,
"learning_rate": 3.1705263157894736e-05,
"loss": 23.0,
"step": 140
},
{
"epoch": 0.00283824994716024,
"grad_norm": 9.838396363193169e-05,
"learning_rate": 3.117684210526316e-05,
"loss": 23.0,
"step": 141
},
{
"epoch": 0.002858379379409604,
"grad_norm": 9.525923087494448e-05,
"learning_rate": 3.064842105263158e-05,
"loss": 23.0,
"step": 142
},
{
"epoch": 0.002878508811658967,
"grad_norm": 0.00010386522626504302,
"learning_rate": 3.012e-05,
"loss": 23.0,
"step": 143
},
{
"epoch": 0.0028986382439083304,
"grad_norm": 0.0001148344308603555,
"learning_rate": 2.9591578947368418e-05,
"loss": 23.0,
"step": 144
},
{
"epoch": 0.002918767676157694,
"grad_norm": 0.00014382245717570186,
"learning_rate": 2.9063157894736842e-05,
"loss": 23.0,
"step": 145
},
{
"epoch": 0.0029388971084070574,
"grad_norm": 5.8804893342312425e-05,
"learning_rate": 2.8534736842105264e-05,
"loss": 23.0,
"step": 146
},
{
"epoch": 0.0029590265406564207,
"grad_norm": 0.00017164551536552608,
"learning_rate": 2.800631578947368e-05,
"loss": 23.0,
"step": 147
},
{
"epoch": 0.0029791559729057844,
"grad_norm": 0.00011976063979091123,
"learning_rate": 2.7477894736842106e-05,
"loss": 23.0,
"step": 148
},
{
"epoch": 0.0029992854051551477,
"grad_norm": 0.0002357963821850717,
"learning_rate": 2.6949473684210527e-05,
"loss": 23.0,
"step": 149
},
{
"epoch": 0.003019414837404511,
"grad_norm": 0.0001414159924024716,
"learning_rate": 2.6421052631578945e-05,
"loss": 23.0,
"step": 150
},
{
"epoch": 0.003019414837404511,
"eval_loss": 11.5,
"eval_runtime": 285.9123,
"eval_samples_per_second": 73.162,
"eval_steps_per_second": 18.292,
"step": 150
},
{
"epoch": 0.0030395442696538742,
"grad_norm": 2.705552105908282e-05,
"learning_rate": 2.589263157894737e-05,
"loss": 23.0,
"step": 151
},
{
"epoch": 0.003059673701903238,
"grad_norm": 4.2475247028050944e-05,
"learning_rate": 2.536421052631579e-05,
"loss": 23.0,
"step": 152
},
{
"epoch": 0.0030798031341526012,
"grad_norm": 9.696526103653014e-05,
"learning_rate": 2.483578947368421e-05,
"loss": 23.0,
"step": 153
},
{
"epoch": 0.0030999325664019645,
"grad_norm": 3.295568239991553e-05,
"learning_rate": 2.430736842105263e-05,
"loss": 23.0,
"step": 154
},
{
"epoch": 0.0031200619986513282,
"grad_norm": 3.4461285395082086e-05,
"learning_rate": 2.3778947368421052e-05,
"loss": 23.0,
"step": 155
},
{
"epoch": 0.0031401914309006915,
"grad_norm": 2.3361501007457264e-05,
"learning_rate": 2.3250526315789473e-05,
"loss": 23.0,
"step": 156
},
{
"epoch": 0.003160320863150055,
"grad_norm": 4.175576032139361e-05,
"learning_rate": 2.2722105263157894e-05,
"loss": 23.0,
"step": 157
},
{
"epoch": 0.003180450295399418,
"grad_norm": 1.8558339434093796e-05,
"learning_rate": 2.2193684210526316e-05,
"loss": 23.0,
"step": 158
},
{
"epoch": 0.0032005797276487818,
"grad_norm": 2.3818216504878365e-05,
"learning_rate": 2.1665263157894737e-05,
"loss": 23.0,
"step": 159
},
{
"epoch": 0.003220709159898145,
"grad_norm": 3.639554779510945e-05,
"learning_rate": 2.1136842105263158e-05,
"loss": 23.0,
"step": 160
},
{
"epoch": 0.0032408385921475083,
"grad_norm": 5.119278284837492e-05,
"learning_rate": 2.060842105263158e-05,
"loss": 23.0,
"step": 161
},
{
"epoch": 0.003260968024396872,
"grad_norm": 5.987023178022355e-05,
"learning_rate": 2.008e-05,
"loss": 23.0,
"step": 162
},
{
"epoch": 0.0032810974566462353,
"grad_norm": 5.560554563999176e-05,
"learning_rate": 1.9551578947368422e-05,
"loss": 23.0,
"step": 163
},
{
"epoch": 0.0033012268888955986,
"grad_norm": 4.0599155909148976e-05,
"learning_rate": 1.9023157894736843e-05,
"loss": 23.0,
"step": 164
},
{
"epoch": 0.0033213563211449623,
"grad_norm": 7.376517169177532e-05,
"learning_rate": 1.849473684210526e-05,
"loss": 23.0,
"step": 165
},
{
"epoch": 0.0033414857533943256,
"grad_norm": 4.924639142700471e-05,
"learning_rate": 1.7966315789473686e-05,
"loss": 23.0,
"step": 166
},
{
"epoch": 0.003361615185643689,
"grad_norm": 6.322375702438876e-05,
"learning_rate": 1.7437894736842107e-05,
"loss": 23.0,
"step": 167
},
{
"epoch": 0.003381744617893052,
"grad_norm": 6.37539560557343e-05,
"learning_rate": 1.6909473684210525e-05,
"loss": 23.0,
"step": 168
},
{
"epoch": 0.003401874050142416,
"grad_norm": 5.568633787333965e-05,
"learning_rate": 1.638105263157895e-05,
"loss": 23.0,
"step": 169
},
{
"epoch": 0.003422003482391779,
"grad_norm": 9.140492329606786e-05,
"learning_rate": 1.5852631578947368e-05,
"loss": 23.0,
"step": 170
},
{
"epoch": 0.0034421329146411424,
"grad_norm": 5.517881436389871e-05,
"learning_rate": 1.532421052631579e-05,
"loss": 23.0,
"step": 171
},
{
"epoch": 0.003462262346890506,
"grad_norm": 5.707240416086279e-05,
"learning_rate": 1.4795789473684209e-05,
"loss": 23.0,
"step": 172
},
{
"epoch": 0.0034823917791398694,
"grad_norm": 6.380578997777775e-05,
"learning_rate": 1.4267368421052632e-05,
"loss": 23.0,
"step": 173
},
{
"epoch": 0.0035025212113892327,
"grad_norm": 7.187807932496071e-05,
"learning_rate": 1.3738947368421053e-05,
"loss": 23.0,
"step": 174
},
{
"epoch": 0.003522650643638596,
"grad_norm": 6.97454743203707e-05,
"learning_rate": 1.3210526315789473e-05,
"loss": 23.0,
"step": 175
},
{
"epoch": 0.0035427800758879597,
"grad_norm": 4.339677616371773e-05,
"learning_rate": 1.2682105263157896e-05,
"loss": 23.0,
"step": 176
},
{
"epoch": 0.003562909508137323,
"grad_norm": 8.792035805527121e-05,
"learning_rate": 1.2153684210526315e-05,
"loss": 23.0,
"step": 177
},
{
"epoch": 0.0035830389403866863,
"grad_norm": 4.989368244423531e-05,
"learning_rate": 1.1625263157894737e-05,
"loss": 23.0,
"step": 178
},
{
"epoch": 0.00360316837263605,
"grad_norm": 8.79050639923662e-05,
"learning_rate": 1.1096842105263158e-05,
"loss": 23.0,
"step": 179
},
{
"epoch": 0.0036232978048854132,
"grad_norm": 6.961109465919435e-05,
"learning_rate": 1.0568421052631579e-05,
"loss": 23.0,
"step": 180
},
{
"epoch": 0.0036434272371347765,
"grad_norm": 0.0001022775613819249,
"learning_rate": 1.004e-05,
"loss": 23.0,
"step": 181
},
{
"epoch": 0.0036635566693841402,
"grad_norm": 5.489904651767574e-05,
"learning_rate": 9.511578947368422e-06,
"loss": 23.0,
"step": 182
},
{
"epoch": 0.0036836861016335035,
"grad_norm": 5.940552000538446e-05,
"learning_rate": 8.983157894736843e-06,
"loss": 23.0,
"step": 183
},
{
"epoch": 0.003703815533882867,
"grad_norm": 9.764087008079514e-05,
"learning_rate": 8.454736842105263e-06,
"loss": 23.0,
"step": 184
},
{
"epoch": 0.00372394496613223,
"grad_norm": 0.00018898840062320232,
"learning_rate": 7.926315789473684e-06,
"loss": 23.0,
"step": 185
},
{
"epoch": 0.003744074398381594,
"grad_norm": 7.49123064451851e-05,
"learning_rate": 7.397894736842104e-06,
"loss": 23.0,
"step": 186
},
{
"epoch": 0.003764203830630957,
"grad_norm": 7.112087041605264e-05,
"learning_rate": 6.8694736842105265e-06,
"loss": 23.0,
"step": 187
},
{
"epoch": 0.0037843332628803203,
"grad_norm": 0.0001479584607295692,
"learning_rate": 6.341052631578948e-06,
"loss": 23.0,
"step": 188
},
{
"epoch": 0.003804462695129684,
"grad_norm": 0.00012902663729619235,
"learning_rate": 5.812631578947368e-06,
"loss": 23.0,
"step": 189
},
{
"epoch": 0.0038245921273790473,
"grad_norm": 8.209590305341408e-05,
"learning_rate": 5.2842105263157896e-06,
"loss": 23.0,
"step": 190
},
{
"epoch": 0.0038447215596284106,
"grad_norm": 5.965860327705741e-05,
"learning_rate": 4.755789473684211e-06,
"loss": 23.0,
"step": 191
},
{
"epoch": 0.003864850991877774,
"grad_norm": 8.907918527256697e-05,
"learning_rate": 4.227368421052631e-06,
"loss": 23.0,
"step": 192
},
{
"epoch": 0.0038849804241271376,
"grad_norm": 0.00010389654926257208,
"learning_rate": 3.698947368421052e-06,
"loss": 23.0,
"step": 193
},
{
"epoch": 0.003905109856376501,
"grad_norm": 8.421300299232826e-05,
"learning_rate": 3.170526315789474e-06,
"loss": 23.0,
"step": 194
},
{
"epoch": 0.003925239288625864,
"grad_norm": 0.0001519916986580938,
"learning_rate": 2.6421052631578948e-06,
"loss": 23.0,
"step": 195
},
{
"epoch": 0.003945368720875228,
"grad_norm": 7.102335075614974e-05,
"learning_rate": 2.1136842105263157e-06,
"loss": 23.0,
"step": 196
},
{
"epoch": 0.003965498153124591,
"grad_norm": 0.00011009284207830206,
"learning_rate": 1.585263157894737e-06,
"loss": 23.0,
"step": 197
},
{
"epoch": 0.0039856275853739544,
"grad_norm": 0.00011992920190095901,
"learning_rate": 1.0568421052631578e-06,
"loss": 23.0,
"step": 198
},
{
"epoch": 0.004005757017623318,
"grad_norm": 0.00013211068289820105,
"learning_rate": 5.284210526315789e-07,
"loss": 23.0,
"step": 199
},
{
"epoch": 0.004025886449872681,
"grad_norm": 0.00018417566025163978,
"learning_rate": 0.0,
"loss": 23.0,
"step": 200
},
{
"epoch": 0.004025886449872681,
"eval_loss": 11.5,
"eval_runtime": 286.0702,
"eval_samples_per_second": 73.122,
"eval_steps_per_second": 18.282,
"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
}