- Remove unused imports (typing.Optional in config.py, uuid in models.py)
- Fix fp8_scaled_mm to use build_policy(checkpoint_path)
- Add .env.example with all configuration options
- FastAPI with job polling: POST /generate, GET /jobs/{id}, /download, /health
- SQLite job persistence (aiosqlite), background task processing
- LTX-2.3 DistilledPipeline via native ltx-pipelines (editable install from submodule)
- Configurable model paths, LoRA support, FP8 quantization via env vars
- Single-concurrency GPU lock for safe inference
- LTX-2 as git submodule under libs/
mteehan