Merge pull request #7 from Mugen0815/main

Check, if model exists locally, before download
3x3cut0r · Mar 4, 2024 · 0bdf36f · 0bdf36f
2 parents c75b63a + dcb26ad
commit 0bdf36f
Show file tree

Hide file tree

Showing 2 changed files with 14 additions and 12 deletions.
diff --git a/llama-cpp-python/README.md b/llama-cpp-python/README.md
@@ -81,7 +81,7 @@ docker run -d \
 docker run -d \
     --name llama-cpp-python \
     --cap-add SYS_RESOURCE \
-    -e MODEL_DOWNLOAD="False \
+    -e MODEL_DOWNLOAD="False" \
     -v /path/to/your/model/mistral-7b-instruct-v0.2.Q4_K_M.gguf:/model/mistral-7b-instruct-v0.2.Q4_K_M.gguf \
     -p 8000:8000/tcp \
     3x3cut0r/llama-cpp-python:latest \
@@ -101,7 +101,7 @@ docker run --rm \
 
 ### 1.2 docker-compose.yml <a name="docker-compose"></a>
 
-```shell
+```yaml
 version: '3.9'
 
 services:
@@ -111,11 +111,11 @@ services:
     cap_add:
       - SYS_RESOURCE
     environment:
-        MODEL_DOWNLOAD="True" \
-        MODEL_REPO="TheBloke/Mistral-7B-Instruct-v0.2-GGUF" \
-        MODEL="mistral-7b-instruct-v0.2.Q4_K_M.gguf" \
-        MODEL_ALIAS="mistral-7b-instruct" \
-        CHAT_FORMAT="mistral" \
+        MODEL_DOWNLOAD: "True"
+        MODEL_REPO: "TheBloke/Mistral-7B-Instruct-v0.2-GGUF"
+        MODEL: "mistral-7b-instruct-v0.2.Q4_K_M.gguf"
+        MODEL_ALIAS: "mistral-7b-instruct"
+        CHAT_FORMAT: "mistral"
     ports:
       - 8000:8000/tcp
 ```

diff --git a/llama-cpp-python/docker-entrypoint.sh b/llama-cpp-python/docker-entrypoint.sh
@@ -24,11 +24,13 @@ ln -snf /usr/share/zoneinfo/$TZ /etc/localtime && echo $TZ > /etc/timezone
 
 # download model
 if [ "$MODEL_REPO" != "local" ] && { [ "$MODEL_DOWNLOAD" = "True" ] || [ "$MODEL_DOWNLOAD" = "true" ] || [ "$MODEL_DOWNLOAD" = "TRUE" ]; }; then
-    if [ "$QUIET" != "true" ]; then
-        echo -e "\nINFO: downloading model from repository ${MODEL_REPO:-'TheBloke/Llama-2-7B-Chat-GGUF'}\n"
-    fi
-    mkdir -p "${MODEL_PATH}"
-    /venv/bin/huggingface-cli download --repo-type model --local-dir="${MODEL_PATH}" --local-dir-use-symlinks=False --resume-download --token="${HF_TOKEN:-''}" "${MODEL_REPO}" "${MODEL}"
+    if [ ! -e "${MODEL_PATH}"/"${MODEL}" ]; then
+        if [ "$QUIET" != "true" ]; then
+            echo -e "\nINFO: downloading model from repository ${MODEL_REPO:-'TheBloke/Llama-2-7B-Chat-GGUF'}\n"
+        fi
+        mkdir -p "${MODEL_PATH}"
+        /venv/bin/huggingface-cli download --repo-type model --local-dir="${MODEL_PATH}" --local-dir-use-symlinks=False --resume-download --token="${HF_TOKEN:-''}" "${MODEL_REPO}" "${MODEL}"
+    fi    
 fi
 
 # if started without args, run app.py