AyoubChLin/bert-base-multilingual-cased-xnli-nli

mBERT Fine-tuned on XNLI for Natural Language Inference

This model is a fine-tuned version of bert-base-multilingual-cased for 3-way Natural Language Inference using the facebook/xnli dataset.

Labels

ID	Label
0	entailment
1	neutral
2	contradiction

Training Configuration

• Base model: bert-base-multilingual-cased
• Architecture: BERT / mBERT sequence classification
• Dataset: facebook/xnli
• Dataset config: all_languages
• Training language: en
• Max sequence length: 512
• Epochs: 3.0
• Learning rate: 2e-05
• Train batch size: 32
• Eval batch size: 64
• Weight decay: 0.01

Intended Use

This model predicts whether a hypothesis is entailed by, neutral with respect to, or contradicted by a premise.

Limitations

• Performance should be validated on your target language and domain.
• Inputs longer than 512 tokens are truncated.
• The model predicts NLI labels and should not be used as a standalone factuality verifier.