metadata

license: apache-2.0
tags:
  - generated_from_trainer
metrics:
  - f1
  - accuracy
model-index:
  - name: distilBERT-finetuned-resumes-sections
    results: []

distilBERT-finetuned-resumes-sections

This model is a fine-tuned version of Geotrend/distilbert-base-en-fr-cased on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.0450
F1: 0.9585
Roc Auc: 0.9774
Accuracy: 0.9557

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	F1	Roc Auc	Accuracy
0.0518	1.0	1174	0.0368	0.9406	0.9635	0.9302
0.0251	2.0	2348	0.0346	0.9375	0.9653	0.9289
0.0136	3.0	3522	0.0343	0.9475	0.9707	0.9425
0.0096	4.0	4696	0.0326	0.9539	0.9737	0.9468
0.007	5.0	5870	0.0357	0.9521	0.9740	0.9480
0.007	6.0	7044	0.0389	0.9509	0.9725	0.9472
0.0034	7.0	8218	0.0403	0.9532	0.9746	0.9510
0.0033	8.0	9392	0.0422	0.9493	0.9722	0.9468
0.0024	9.0	10566	0.0425	0.9512	0.9733	0.9485
0.0023	10.0	11740	0.0431	0.9537	0.9743	0.9502
0.0019	11.0	12914	0.0457	0.9501	0.9719	0.9463
0.002	12.0	14088	0.0428	0.9560	0.9751	0.9536
0.0012	13.0	15262	0.0435	0.9569	0.9761	0.9553
0.001	14.0	16436	0.0464	0.9565	0.9759	0.9544
0.001	15.0	17610	0.0460	0.9574	0.9766	0.9549
0.0007	16.0	18784	0.0450	0.9585	0.9774	0.9557
0.0003	17.0	19958	0.0481	0.9572	0.9764	0.9553
0.0005	18.0	21132	0.0478	0.9576	0.9764	0.9557
0.0005	19.0	22306	0.0483	0.9574	0.9766	0.9553
0.0005	20.0	23480	0.0481	0.9576	0.9766	0.9557

Framework versions

Transformers 4.21.1
Pytorch 1.12.1+cu113
Datasets 2.4.0
Tokenizers 0.12.1