seanzhang-zhichen / llama3-chinese

Llama3-Chinese是以Meta-Llama-3-8B为底座,使用 DORA + LORA+ 的训练方法,在50w高质量中文多轮SFT数据 + 10w英文多轮SFT数据 + 2000单轮自我认知数据训练而来的大模型。
295Updated 8 months ago

Alternatives and similar repositories for llama3-chinese:

Users that are interested in llama3-chinese are comparing it to the libraries listed below