distilbert-base-multilingual-cased-misogyny-sexism-decay0.05-fr-outofdomain

This model is a fine-tuned version of distilbert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.9920
• Accuracy: 0.2851
• F1: 0.1967
• Precision: 0.1124
• Recall: 0.7870
• Mae: 0.7149
• Tn: 1727
• Fp: 6043
• Fn: 207
• Tp: 765

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: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 3

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall	Mae	Tn	Fp	Fn	Tp
0.3603	1.0	2233	0.8218	0.3251	0.2021	0.1163	0.7685	0.6749	2095	5675	225	747
0.298	2.0	4466	0.9031	0.3164	0.2047	0.1175	0.7912	0.6836	1997	5773	203	769
0.2438	3.0	6699	0.9920	0.2851	0.1967	0.1124	0.7870	0.7149	1727	6043	207	765

Framework versions

• Transformers 4.20.1
• Pytorch 1.12.0+cu102
• Datasets 2.3.2
• Tokenizers 0.12.1