fix

Dockerfile

@@ -1,23 +1,30 @@
 FROM intel/oneapi-runtime:latest
 
-WORKDIR /app
-
 RUN apt-get update && apt-get install -y --no-install-recommends \
     aria2 \
     unzip \
     && rm -rf /var/lib/apt/lists/*
 
+RUN useradd -m -u 1000 user
+
+USER user
+
+ENV HOME=/home/user \
+	PATH=/home/user/.local/bin:$PATH
+
+WORKDIR $HOME/app
+
 RUN aria2c -c -x16 https://github.com/MZWNET/actions/releases/download/llama_cpp-b2690/llama-b2690-bin-linux-avx2-intel-mkl-x64.zip \
     && unzip llama-b2690-bin-linux-avx2-intel-mkl-x64.zip -d /app/llama.cpp \
     && mv /app/llama.cpp/server /app/ \
     && chmod +x /app/server \
     && rm -rf llama-b2690-bin-linux-avx2-intel-mkl-x64.zip llama.cpp
 
-COPY start.sh start.sh
+COPY --chown=user start.sh start.sh
 
 RUN chmod +x start.sh
 
-ENV PATH="/app:${PATH}"
+ENV PATH=$HOME/app:$PATH
 
 EXPOSE 8080
 

README.md

@@ -15,7 +15,7 @@ custom_headers:
 models:
   - mzwing/AquilaChat2-7B-16K-GGUF
 preload_from_hub:
-  - mzwing/AquilaChat2-7B-16K-GGUF AquilaChat2-7B-16K.Q8_0.gguf
+  - mzwing/AquilaChat2-7B-16K-GGUF AquilaChat2-7B-16K.Q8_0.gguf 3bff96b3dbb6a2386f71199502786739ef71b246
 ---
 
 AquilaChat2-7B-16K-GGUF, loaded by llama.cpp server.

start.sh

@@ -1,3 +1,3 @@
-ls -alh /root/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/
-ls -alh /root/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/snapshots/
-server -m /root/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/AquilaChat2-7B-16K.Q8_0.gguf -t $(nproc) -c 16384 -a AquilaChat2-7B-16K --port 8080 --api-key sk-1145141919810
+ls -alh ~/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/
+ls -alh ~/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/snapshots/3bff96b3dbb6a2386f71199502786739ef71b246/
+server -m ~/.cache/huggingface/hub/models--mzwing--AquilaChat2-7B-16K-GGUF/snapshots/3bff96b3dbb6a2386f71199502786739ef71b246/AquilaChat2-7B-16K.Q8_0.gguf -t $(nproc) -c 16384 -a AquilaChat2-7B-16K --port 8080 --api-key sk-1145141919810