llama

LLaMA is a large-scale open-source pre-trained language model collection launched by Meta. It supports multiple languages and multi-task adaptation, making it suitable for custom development and enterprise applications.

Petals是一个开源的分布式大语言模型（LLM）协作平台，让用户用普通GPU便能参与大模型的推理和微调，实现大模型的民主化使用与开放创新。