add internlm url (#67)

README.md

@@ -74,6 +74,14 @@ pip install -e .
 ```shell
 # 1. Download InternLM model
 
+# Make sure you have git-lfs installed (https://git-lfs.com)
+git lfs install
+git clone https://huggingface.co/internlm/internlm-7b /path/to/internlm-7b
+
+# if you want to clone without large files – just their pointers
+# prepend your git clone with the following env var:
+GIT_LFS_SKIP_SMUDGE=1
+
 # 2. Convert InternLM model to turbomind's format, which will be in "./workspace" by default
 python3 -m lmdeploy.serve.turbomind.deploy internlm-7b /path/to/internlm-7b hf
 

README_zh-CN.md

@@ -73,6 +73,14 @@ pip install -e .
 ```shell
 # 1. 下载 InternLM 模型
 
+# Make sure you have git-lfs installed (https://git-lfs.com)
+git lfs install
+git clone https://huggingface.co/internlm/internlm-7b /path/to/internlm-7b
+
+# if you want to clone without large files – just their pointers
+# prepend your git clone with the following env var:
+GIT_LFS_SKIP_SMUDGE=1
+
 # 2. 转换为 trubomind 要求的格式。默认存放路径为 ./workspace
 python3 -m lmdeploy.serve.turbomind.deploy internlm-7b /path/to/internlm-7b hf
 

lmdeploy/lite/apis/kv_qparams.py

@@ -168,7 +168,7 @@ def main(model: str,
         save_path = out_dir / f'layers.{layer}.past_kv_scale.{tp}.weight'
         if symmetry:
             # quant: q = f / scale
-            # dequant: f = q * scale 
+            # dequant: f = q * scale
             k_scale = max(k_obs.buffer) / (2**(bits - 1) - 1)
             v_scale = max(v_obs.buffer) / (2**(bits - 1) - 1)