# LLaMA fine-tuning

A multilingual dialogue model fine-tuned based on the LLaMA architecture, supporting Arabic and English, optimized for dialogue tasks and providing high-quality responses.

This is a 68-million-parameter Vicuna-like model, fine-tuned on ShareGPT data based on LLaMA-68m, primarily used for speculative model development.

Instruction-following model fine-tuned on LLaMA-7B, specifically optimized for code instruction scenarios

Vicuna is an open-source chatbot fine-tuned from LLaMA using user-shared dialogue data, suitable for academic research.