xnli_m_bert_only_fr

This model is a fine-tuned version of bert-base-multilingual-cased on the xnli dataset. It achieves the following results on the evaluation set:

• Loss: 1.2262
• Accuracy: 0.7675

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: 5e-05
• train_batch_size: 128
• eval_batch_size: 128
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.6184	1.0	3068	0.6251	0.7373
0.5293	2.0	6136	0.5669	0.7635
0.4343	3.0	9204	0.6161	0.7651
0.3456	4.0	12272	0.6650	0.7631
0.2677	5.0	15340	0.7249	0.7755
0.2022	6.0	18408	0.8638	0.7590
0.1488	7.0	21476	0.9073	0.7763
0.1096	8.0	24544	1.0603	0.7586
0.0813	9.0	27612	1.1546	0.7687
0.0599	10.0	30680	1.2262	0.7675

Framework versions

• Transformers 4.24.0
• Pytorch 1.13.0
• Datasets 2.6.1
• Tokenizers 0.13.1