m De B E R Ta v3 base xnli multilingual nli 2mil7

Introduction

The mDeBERTa-v3-base-xnli-multilingual-nli-2mil7 is a multilingual model designed for natural language inference (NLI) across 100 languages, suitable for zero-shot classification. It is based on the mDeBERTa-v3-base model pre-trained by Microsoft and fine-tuned on various NLI datasets.

Architecture

This model leverages the DeBERTa architecture, which stands out for its performance in multilingual contexts, especially in natural language inference tasks. It was fine-tuned using over 2.7 million hypothesis-premise pairs in 27 languages.

Training

The model was trained on datasets like multilingual-NLI-26lang-2mil7 and XNLI, incorporating hypothesis-premise pairs across 26 languages. The training process employed specific hyperparameters, such as a learning rate of 2e-05 and a batch size of 32, over three epochs.

Guide: Running Locally

1. Setup Environment: Install the required packages using pip:

   pip install transformers torch
   

2. Load Model: Use the Transformers library to load the model.

   from transformers import pipeline
   classifier = pipeline("zero-shot-classification", model="MoritzLaurer/mDeBERTa-v3-base-xnli-multilingual-nli-2mil7")
   

3. Perform Inference: Provide a sequence and candidate labels for classification.

   sequence_to_classify = "Angela Merkel ist eine Politikerin in Deutschland und Vorsitzende der CDU"
   candidate_labels = ["politics", "economy", "entertainment", "environment"]
   output = classifier(sequence_to_classify, candidate_labels, multi_label=False)
   print(output)
   

4. Cloud GPUs: For enhanced performance, consider using cloud-based GPUs from providers like AWS, GCP, or Azure.

License

This model is licensed under the MIT License, allowing for flexibility in usage and distribution.