Edit model card

bert-finetuned-resumes-sections

This model is a fine-tuned version of dbmdz/bert-base-french-europeana-cased on an unknown dataset. It achieves the following results on the evaluation set:

  • Loss: 0.0333
  • F1: 0.9548
  • Roc Auc: 0.9732
  • Accuracy: 0.9493

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • learning_rate: 2e-05
  • train_batch_size: 8
  • eval_batch_size: 8
  • seed: 42
  • optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 12

Training results

Training Loss Epoch Step Validation Loss F1 Roc Auc Accuracy
0.1659 1.0 601 0.0645 0.9201 0.9434 0.8910
0.055 2.0 1202 0.0426 0.9407 0.9633 0.9309
0.0324 3.0 1803 0.0371 0.9450 0.9663 0.9368
0.0226 4.0 2404 0.0389 0.9402 0.9651 0.9343
0.0125 5.0 3005 0.0354 0.9433 0.9650 0.9343
0.0091 6.0 3606 0.0364 0.9482 0.9696 0.9434
0.0075 7.0 4207 0.0363 0.9464 0.9676 0.9393
0.007 8.0 4808 0.0333 0.9548 0.9732 0.9493
0.0063 9.0 5409 0.0358 0.9501 0.9698 0.9434
0.0043 10.0 6010 0.0380 0.9475 0.9707 0.9443
0.0032 11.0 6611 0.0377 0.9491 0.9712 0.9468
0.0031 12.0 7212 0.0375 0.9500 0.9716 0.9459

Framework versions

  • Transformers 4.19.2
  • Pytorch 1.11.0+cu113
  • Datasets 2.2.2
  • Tokenizers 0.12.1
Downloads last month
30