Update Dockerfile

Dockerfile

@@ -1,19 +1,22 @@
-# Use the official image from the library author.
-# This includes the correct pre-compiled binaries and system libraries.
+# Use the official image which has the library correctly installed
 FROM ghcr.io/abetlen/llama-cpp-python:latest
 
 # Set the working directory
 WORKDIR /app
 
-# Copy your model file into the container
-# Ensure 'model/gemma-3-finetuned.Q4_K_M.gguf' exists in your Space's file list!
+# 1. DO NOT use "COPY . ." or "COPY . /app"
+#    This prevents copying any local broken/empty llama_cpp folders.
+
+# 2. ONLY copy the model file.
+#    Ensure 'model/gemma-3-finetuned.Q4_K_M.gguf' exists in your HF Space "Files" tab.
+#    We rename it to 'model.gguf' inside the container for simplicity.
 COPY model/gemma-3-finetuned.Q4_K_M.gguf /app/model/model.gguf
 
-# Set environment variables for the server
-# Hugging Face Spaces requires port 7860
+# 3. Set environment variables for the server
 ENV HOST=0.0.0.0
 ENV PORT=7860
 ENV MODEL=/app/model/model.gguf
 
-# Start the OpenAI-compatible server
+# 4. Start the server
+#    This runs the installed library from the system paths, ignoring /app
 CMD ["python3", "-m", "llama_cpp.server", "--model", "/app/model/model.gguf", "--host", "0.0.0.0", "--port", "7860", "--n_ctx", "2048"]