CoolFace
Modelpublic

SotirisLegkas/multi_hate

sourceHugging Faceupdated 3y agoView on Hugging Face
0likes7downloads
trainer_state.json322 linesDownload Raw Back to root
1{2  "best_metric": 0.8600789143532662,3  "best_model_checkpoint": "training_with_callbacks/checkpoint-7880",4  "epoch": 10.0,5  "global_step": 15760,6  "is_hyper_param_search": false,7  "is_local_process_zero": true,8  "is_world_process_zero": true,9  "log_history": [10    {11      "epoch": 0.32,12      "learning_rate": 1.9368020304568528e-05,13      "loss": 0.5589,14      "step": 50015    },16    {17      "epoch": 0.63,18      "learning_rate": 1.8733502538071068e-05,19      "loss": 0.4001,20      "step": 100021    },22    {23      "epoch": 0.95,24      "learning_rate": 1.81002538071066e-05,25      "loss": 0.3671,26      "step": 150027    },28    {29      "epoch": 1.0,30      "eval_accuracy": 0.8556491163135005,31      "eval_loss": 0.33663880825042725,32      "eval_macro_f1": 0.8274802013644363,33      "eval_micro_f1": 0.8556491163135005,34      "eval_runtime": 45.4196,35      "eval_samples_per_second": 237.937,36      "eval_steps_per_second": 14.883,37      "step": 157638    },39    {40      "epoch": 1.27,41      "learning_rate": 1.7465736040609138e-05,42      "loss": 0.3096,43      "step": 200044    },45    {46      "epoch": 1.59,47      "learning_rate": 1.6831218274111674e-05,48      "loss": 0.2976,49      "step": 250050    },51    {52      "epoch": 1.9,53      "learning_rate": 1.6196700507614215e-05,54      "loss": 0.3043,55      "step": 300056    },57    {58      "epoch": 2.0,59      "eval_accuracy": 0.8647173128527806,60      "eval_loss": 0.3377376198768616,61      "eval_macro_f1": 0.8410857961930669,62      "eval_micro_f1": 0.8647173128527806,63      "eval_runtime": 45.0937,64      "eval_samples_per_second": 239.656,65      "eval_steps_per_second": 14.991,66      "step": 315267    },68    {69      "epoch": 2.22,70      "learning_rate": 1.5562182741116755e-05,71      "loss": 0.2492,72      "step": 350073    },74    {75      "epoch": 2.54,76      "learning_rate": 1.4927664974619292e-05,77      "loss": 0.2329,78      "step": 400079    },80    {81      "epoch": 2.86,82      "learning_rate": 1.4293147208121828e-05,83      "loss": 0.2413,84      "step": 450085    },86    {87      "epoch": 3.0,88      "eval_accuracy": 0.8739705746275562,89      "eval_loss": 0.32710105180740356,90      "eval_macro_f1": 0.8518710972157342,91      "eval_micro_f1": 0.8739705746275562,92      "eval_runtime": 45.1151,93      "eval_samples_per_second": 239.543,94      "eval_steps_per_second": 14.984,95      "step": 472896    },97    {98      "epoch": 3.17,99      "learning_rate": 1.3658629441624367e-05,100      "loss": 0.2021,101      "step": 5000102    },103    {104      "epoch": 3.49,105      "learning_rate": 1.30253807106599e-05,106      "loss": 0.1933,107      "step": 5500108    },109    {110      "epoch": 3.81,111      "learning_rate": 1.2392131979695431e-05,112      "loss": 0.1887,113      "step": 6000114    },115    {116      "epoch": 4.0,117      "eval_accuracy": 0.872119922272601,118      "eval_loss": 0.367958664894104,119      "eval_macro_f1": 0.8499428993945666,120      "eval_micro_f1": 0.8721199222726009,121      "eval_runtime": 45.2612,122      "eval_samples_per_second": 238.77,123      "eval_steps_per_second": 14.936,124      "step": 6304125    },126    {127      "epoch": 4.12,128      "learning_rate": 1.175761421319797e-05,129      "loss": 0.1702,130      "step": 6500131    },132    {133      "epoch": 4.44,134      "learning_rate": 1.1123096446700508e-05,135      "loss": 0.1574,136      "step": 7000137    },138    {139      "epoch": 4.76,140      "learning_rate": 1.0488578680203047e-05,141      "loss": 0.1494,142      "step": 7500143    },144    {145      "epoch": 5.0,146      "eval_accuracy": 0.8798926621634126,147      "eval_loss": 0.3825337588787079,148      "eval_macro_f1": 0.8600789143532662,149      "eval_micro_f1": 0.8798926621634126,150      "eval_runtime": 44.9299,151      "eval_samples_per_second": 240.53,152      "eval_steps_per_second": 15.046,153      "step": 7880154    },155    {156      "epoch": 5.08,157      "learning_rate": 9.855329949238578e-06,158      "loss": 0.1449,159      "step": 8000160    },161    {162      "epoch": 5.39,163      "learning_rate": 9.220812182741118e-06,164      "loss": 0.1263,165      "step": 8500166    },167    {168      "epoch": 5.71,169      "learning_rate": 8.586294416243655e-06,170      "loss": 0.1255,171      "step": 9000172    },173    {174      "epoch": 6.0,175      "eval_accuracy": 0.8775793467197187,176      "eval_loss": 0.471343457698822,177      "eval_macro_f1": 0.8558476837008455,178      "eval_micro_f1": 0.8775793467197187,179      "eval_runtime": 45.3421,180      "eval_samples_per_second": 238.343,181      "eval_steps_per_second": 14.909,182      "step": 9456183    },184    {185      "epoch": 6.03,186      "learning_rate": 7.951776649746194e-06,187      "loss": 0.1259,188      "step": 9500189    },190    {191      "epoch": 6.35,192      "learning_rate": 7.319796954314721e-06,193      "loss": 0.0964,194      "step": 10000195    },196    {197      "epoch": 6.66,198      "learning_rate": 6.68527918781726e-06,199      "loss": 0.1086,200      "step": 10500201    },202    {203      "epoch": 6.98,204      "learning_rate": 6.050761421319798e-06,205      "loss": 0.1077,206      "step": 11000207    },208    {209      "epoch": 7.0,210      "eval_accuracy": 0.8795225316924216,211      "eval_loss": 0.5118919610977173,212      "eval_macro_f1": 0.8578917168529726,213      "eval_micro_f1": 0.8795225316924216,214      "eval_runtime": 45.4537,215      "eval_samples_per_second": 237.759,216      "eval_steps_per_second": 14.872,217      "step": 11032218    },219    {220      "epoch": 7.3,221      "learning_rate": 5.416243654822335e-06,222      "loss": 0.0846,223      "step": 11500224    },225    {226      "epoch": 7.61,227      "learning_rate": 4.781725888324874e-06,228      "loss": 0.0851,229      "step": 12000230    },231    {232      "epoch": 7.93,233      "learning_rate": 4.1472081218274114e-06,234      "loss": 0.0875,235      "step": 12500236    },237    {238      "epoch": 8.0,239      "eval_accuracy": 0.8807254557231424,240      "eval_loss": 0.5281736254692078,241      "eval_macro_f1": 0.8573342434559597,242      "eval_micro_f1": 0.8807254557231424,243      "eval_runtime": 45.6116,244      "eval_samples_per_second": 236.935,245      "eval_steps_per_second": 14.821,246      "step": 12608247    },248    {249      "epoch": 8.25,250      "learning_rate": 3.5126903553299495e-06,251      "loss": 0.0705,252      "step": 13000253    },254    {255      "epoch": 8.57,256      "learning_rate": 2.8781725888324876e-06,257      "loss": 0.0711,258      "step": 13500259    },260    {261      "epoch": 8.88,262      "learning_rate": 2.2449238578680207e-06,263      "loss": 0.0702,264      "step": 14000265    },266    {267      "epoch": 9.0,268      "eval_accuracy": 0.8808179883408902,269      "eval_loss": 0.591612696647644,270      "eval_macro_f1": 0.8563906267141775,271      "eval_micro_f1": 0.8808179883408902,272      "eval_runtime": 45.3025,273      "eval_samples_per_second": 238.552,274      "eval_steps_per_second": 14.922,275      "step": 14184276    },277    {278      "epoch": 9.2,279      "learning_rate": 1.6116751269035535e-06,280      "loss": 0.063,281      "step": 14500282    },283    {284      "epoch": 9.52,285      "learning_rate": 9.771573604060914e-07,286      "loss": 0.0609,287      "step": 15000288    },289    {290      "epoch": 9.84,291      "learning_rate": 3.4263959390862946e-07,292      "loss": 0.0539,293      "step": 15500294    },295    {296      "epoch": 10.0,297      "eval_accuracy": 0.8817433145183677,298      "eval_loss": 0.6316967010498047,299      "eval_macro_f1": 0.8560485298070233,300      "eval_micro_f1": 0.8817433145183678,301      "eval_runtime": 45.5653,302      "eval_samples_per_second": 237.176,303      "eval_steps_per_second": 14.836,304      "step": 15760305    },306    {307      "epoch": 10.0,308      "step": 15760,309      "total_flos": 1.3269191906021376e+17,310      "train_loss": 0.1755784906106552,311      "train_runtime": 5214.2121,312      "train_samples_per_second": 96.718,313      "train_steps_per_second": 3.023314    }315  ],316  "max_steps": 15760,317  "num_train_epochs": 10,318  "total_flos": 1.3269191906021376e+17,319  "trial_name": null,320  "trial_params": null321}322