Update Dockerfile

Dockerfile

@@ -1,4 +1,4 @@
-# Use NVIDIA CUDA base image with Python support
+# Use NVIDIA CUDA base image with Python support (ONLY for GPU Spaces)
 FROM nvidia/cuda:12.1.1-runtime-ubuntu22.04
 
 # Set working directory
@@ -9,6 +9,9 @@ RUN apt-get update && apt-get install -y \
     python3.10 \
     python3-pip \
     python3.10-venv \
+    python3.10-dev \
+    build-essential \
+    cmake \
     git \
     curl \
     wget \
@@ -16,12 +19,13 @@ RUN apt-get update && apt-get install -y \
 
 # Make python3.10 the default
 RUN update-alternatives --install /usr/bin/python python /usr/bin/python3.10 1
+RUN update-alternatives --install /usr/bin/pip pip /usr/bin/pip3 1
 
 # Upgrade pip
 RUN pip install --upgrade pip
 
-# ✅ Install prebuilt llama-cpp-python CUDA wheel (cu121)
-RUN pip install --no-cache-dir llama-cpp-python==0.3.0 --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121
+# Install prebuilt llama-cpp-python CUDA wheel
+RUN CMAKE_ARGS="-DLLAMA_CUBLAS=on" pip install --no-cache-dir llama-cpp-python==0.2.90 --extra-index-url https://abetlen.github.io/llama-cpp-python/whl/cu121
 
 # Copy requirements.txt and install remaining dependencies
 COPY requirements.txt .
@@ -33,12 +37,13 @@ COPY . .
 # Create models directory
 RUN mkdir -p models
 
-# Expose Gradio or web port
+# Expose Gradio port
 EXPOSE 7860
 
-# Gradio settings
+# Environment variables
 ENV GRADIO_SERVER_NAME="0.0.0.0"
 ENV GRADIO_SERVER_PORT=7860
+ENV CUDA_VISIBLE_DEVICES=0
 
 # Start app
-CMD ["python", "app.py"]
+CMD ["python", "app.py"]