Multi-stage Docker build: Stage 1 compiles llama-cpp-python once, Stage 2 reuses compiled wheels - NO TIMEOUT! Build time 8-12 minutes first time, then cached.

Multi-stage Docker build: Stage 1 compiles llama-cpp-python to wheel, Stage 2 installs pre-built wheel - NO TIMEOUT! Pre-download fast-chat model at build time.

Switch to Ollama for zero-compilation deployment - pre-downloads models at startup

Switch to pre-compiled llama-cpp-python binary wheels - no compilation timeout on HF Spaces

Revert to llama-cpp-python with storage optimization - only load fast-chat at startup

Fix: use mistral type for qwen models (not supported directly) + fallback to llama

Fix: only load fast-chat at startup (350MB) - skip other large models to save storage

Fix: map model types correctly for ctransformers (qwen, phi, llama, mistral)

Switch to ctransformers (pre-built, no compilation!) - faster HF Spaces deploy

Add build-essential & cmake for llama-cpp-python compilation

Optimize: use --prefer-binary for wheels on Linux but not committed

Remove llama-cpp-python from requirements - already in base image

Use pre-built llama-cpp-python Docker image - skip compilation

Fix: revert to standard pip install (binary-only broke uvicorn install)

Force binary wheels only - skip all compilation for fast HF Spaces build

Optimize Dockerfile: skip expensive llama-cpp-python compilation flags

Add FastAPI backend with Docker for HuggingFace Spaces

initial commit