Update Dockerfile

Dockerfile

@@ -10,20 +10,32 @@ COPY requirements.txt .
 # Step 4: Install the Python dependencies
 RUN pip install --no-cache-dir -r requirements.txt
 
+# Install faiss-gpu separately
 RUN pip install --no-cache-dir faiss-gpu
 
-# Step 5: Copy the entire codebase into the container
+# Step 5: Create necessary cache directories for HuggingFace, Sentence Transformers, and Torch
+RUN mkdir -p /app/.cache/huggingface/transformers \
+             /app/.cache/sentence_transformers \
+             /app/.cache/torch \
+    && chmod -R 777 /app/.cache
+
+# Step 6: Set environment variables to use the custom cache directories
+ENV HF_HOME=/app/.cache/huggingface
+ENV TRANSFORMERS_CACHE=/app/.cache/huggingface/transformers
+ENV SENTENCE_TRANSFORMERS_HOME=/app/.cache/sentence_transformers
+ENV TORCH_HOME=/app/.cache/torch
+
+# Step 7: Copy the entire codebase into the container
 COPY . .
 
-# Step 6: Ensure FAISS index and other models are in the correct location
-# Copy the FAISS index and other model files explicitly
+# Step 8: Ensure FAISS index and other models are in the correct location
 COPY ./models /app/models
 
-# Optional: Set permissions for the entire app directory
+# Step 9: Set permissions for the entire app directory
 RUN chmod -R 755 /app
 
-# Step 7: Expose the port that FastAPI will run on (default: 8000)
+# Step 10: Expose the port that FastAPI will run on (default: 7860)
 EXPOSE 7860
 
-# Step 8: Set the entry point to run FastAPI with Uvicorn
+# Step 11: Set the entry point to run FastAPI with Uvicorn
 CMD ["uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "7860", "--reload"]