lordjia/Llama-3-Cantonese-8B-Instruct

Llama-3-Cantonese-8B-Instruct by lordjia is an 8 billion parameter language model based on Meta-Llama-3-8B-Instruct, fine-tuned using LoRA. It is specifically optimized to enhance Cantonese text generation and comprehension, supporting tasks like dialogue, summarization, and question-answering. This model leverages Cantonese-specific datasets to improve its performance in the primary language. A 4-bit quantized version is also available for efficient deployment.