feat: remove gpu support and simplify to cpu-only

CLAUDE.md

@@ -2,7 +2,7 @@
 
 ## Overview
 
-FaceFusion API - A Python REST API wrapping FaceFusion v3.5.3 for face swapping, enhancement, lip sync, and other face/frame processing. Containerized with CUDA GPU support, deployed via Gitea CI/CD at `dev.pivoine.art`.
+FaceFusion API - A Python REST API wrapping FaceFusion v3.5.3 for face swapping, enhancement, lip sync, and other face/frame processing. Containerized for CPU-only execution, deployed via Gitea CI/CD at `dev.pivoine.art`.
 
 ## Architecture
 
@@ -43,9 +43,6 @@ docker compose up
 # Production (CPU VPS)
 docker compose -f docker-compose.prod.yml up -d
 
-# Production (GPU server)
-docker compose -f docker-compose.gpu.yml up -d
-
 # Test endpoints
 curl http://localhost:8000/api/v1/health
 curl http://localhost:8000/api/v1/system
@@ -69,19 +66,18 @@ curl -X POST http://localhost:8000/api/v1/process \
 - `GET /api/v1/models` - List downloaded models
 - `POST /api/v1/models/download` - Download models
 - `GET /api/v1/health` - Health check
-- `GET /api/v1/system` - System info (GPU, memory)
+- `GET /api/v1/system` - System info (CPU, memory)
 
 ## Docker
 
-- `VARIANT=cpu` (default): `python:3.12-slim` + `onnxruntime`
-- `VARIANT=gpu`: `nvidia/cuda:12.4.1-cudnn-runtime-ubuntu22.04` + `onnxruntime-gpu`
+- Base image: `python:3.12-slim` + `onnxruntime`
 - Models persisted in `/data/models` Docker volume (not baked into image)
 - Single worker mandatory (`--workers 1`)
 
 ## Environment Variables
 
 All prefixed with `FF_`:
-- `FF_EXECUTION_PROVIDERS` - JSON array, e.g. `["cuda","cpu"]`
+- `FF_EXECUTION_PROVIDERS` - JSON array, e.g. `["cpu"]`
 - `FF_EXECUTION_THREAD_COUNT` - Default 4
 - `FF_VIDEO_MEMORY_STRATEGY` - strict/moderate/tolerant
 - `FF_MODELS_DIR` - Model storage path
@@ -91,5 +87,5 @@ All prefixed with `FF_`:
 
 - Never run with multiple uvicorn workers - FaceFusion global state will corrupt
 - Models are 100MB-1GB each; pre-download via `/api/v1/models/download` before processing
-- The `facefusion/` submodule must not be modified - use symlinks for model paths
+- The `facefusion/` submodule should not be modified directly if possible - use symlinks for model paths
 - Git operations: always push with the valknarthing ssh key