Update Dockerfile

Dockerfile

@@ -26,9 +26,10 @@ RUN pip3 install --no-cache-dir --upgrade pip && \
     pip3 install --no-cache-dir -r requirements.txt
 
 # Install llama-cpp-python with CUDA support
-ENV FORCE_CMAKE=1
-ENV CMAKE_ARGS="-DLLAMA_CUBLAS=on"
-RUN pip3 install --no-cache-dir llama-cpp-python
+#ENV FORCE_CMAKE=1
+# ENV CMAKE_ARGS="-DLLAMA_CUBLAS=on"
+RUN pip3 install llama-cpp-python \
+  --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121
 
 # Copy the rest of the application code
 COPY . .