From 3e868cd6717e1c393b849a65b1f68f54de5b7ea9 Mon Sep 17 00:00:00 2001 From: DingDing Date: Fri, 2 Feb 2024 17:36:05 +0800 Subject: [PATCH] Update README.md --- README.md | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/README.md b/README.md index 0d600f1..a8712f0 100644 --- a/README.md +++ b/README.md @@ -109,7 +109,7 @@ python inference.py --model_path --prompt_path prompts/promp ``` #### Huggingface 模型 -(注:我们发现当前Huggingface的推理代码推理效果差于Vllm的推理代码,我们正在对齐中,目前已定为到PageAttention和普通attention的区别,请耐心等待) +(注:我们发现当前Huggingface的推理代码推理效果差于Vllm的推理代码,我们正在对齐中,目前已定位到attention计算的精度问题,请耐心等待) ##### MiniCPM-2B * 安装`transformers>=4.36.0`以及`accelerate`后,运行以下代码 ```python