Pinned Loading

1. llama3-chinese llama3-chinese Public

   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。

   Python 290 21

2. Qwen-WisdomVast Qwen-WisdomVast Public

   Qwen-WisdomVast is a large model trained on 1 million high-quality Chinese multi-turn SFT data, 200,000 English multi-turn SFT data, and 2,000 single-turn self-cognition data, using the training me…

   Python 18 1

3. baichuan-Dynamic-NTK-ALiBi baichuan-Dynamic-NTK-ALiBi Public

   百川Dynamic NTK-ALiBi的代码实现：无需微调即可推理更长文本

   Python 46 4