Fix: Force use of pre-built llama-cpp-python wheels

Dockerfile

@@ -9,8 +9,18 @@ RUN apt-get update && apt-get install -y \
     make \
     && rm -rf /var/lib/apt/lists/*
 
-# Copy requirements and install (llama-cpp-python installed at runtime)
+# Copy requirements and install
 COPY requirements.txt .
+
+# Set environment variables to use pre-built wheels for llama-cpp-python
+ENV CMAKE_ARGS="-DLLAMA_BLAS=OFF -DLLAMA_BLAS_VENDOR=OFF -DLLAMA_CUBLAS=OFF"
+ENV FORCE_CMAKE=0
+
+# Install llama-cpp-python from PyPI (not building from source)
+RUN pip install --no-cache-dir --only-binary=:all: llama-cpp-python==0.3.14 || \
+    pip install --no-cache-dir llama-cpp-python==0.3.14
+
+# Install remaining requirements
 RUN pip install --no-cache-dir -r requirements.txt
 
 # Copy application code

main.py

@@ -1,10 +1,3 @@
-# Install llama-cpp-python at runtime (HF Spaces workaround)
-import subprocess
-import sys
-print("🔧 Installing llama-cpp-python at runtime...")
-subprocess.run([sys.executable, "-m", "pip", "install", "-v", "llama-cpp-python==0.3.14"], check=True)
-print("✅ llama-cpp-python installed successfully")
-
 from fastapi import FastAPI, HTTPException, WebSocket, WebSocketDisconnect, BackgroundTasks
 from fastapi.responses import StreamingResponse, JSONResponse, FileResponse
 from fastapi.middleware.cors import CORSMiddleware