Modelpublic

npc0/chatglm3-6b-int8

sourceHugging Faceupdated 3y agoView on Hugging Face

1likes

Model Card

ChatGLM3-6B-int8

介绍 (Introduction) ChatGLM3-6B 是 ChatGLM 系列最新一代的开源模型，THUDM/chatglm3-6b

用 [ChatGLM.CPP]() 基於 GGML quantize 生成 Q8_0 權重 weights 儲存於此倉庫。

Performance

Model	GGML quantize method	HDD size	1 token\*
chatglm3-ggml-q8_0.bin	q8_0	6.64 GB	114ms

\* ms/token (CPU @ Platinum 8260) from reference

Getting Started

1.Install dependency

  pip install chatglm-cpp transformers

1.Download weight

  wget https://huggingface.co/npc0/chatglm3-6b-int8/resolve/main/chatglm3-ggml-q8_0.bin

1.Code

  import chatglm_cpp

  pipeline = chatglm_cpp.Pipeline("./chatglm3-ggml-q8_0.bin")
  pipeline.chat([chatglm_cpp.ChatMessage(role="user", content="你好")])
  # Output: ChatMessage(role="assistant", content="你好👋！我是人工智能助手 ChatGLM-6B，很高兴见到你，欢迎问我任何问题。", tool_calls=[])