96b7183d70
All checks were successful
Smoke Test / smoke (pull_request) Successful in 8s
Implements #116 — hardware validation testing for edge crisis detector on Raspberry Pi 4 and other edge devices. Adds edge detector (keyword + optional Ollama model), crisis_resources.json, deployment docs, and two test files: - test_edge_detector.py: unit tests for keyword logic - test_edge_detector_hardware.py: hardware validation suite Hardware validation measures keyword detection (<1ms), model inference (<5s on Pi 4), offline operation, and provides reproducible benchmark via `python3 edge/detector.py --benchmark`. Re-implements the functionality from closed PR #111 with expanded tests.
TurboQuant
KV cache compression for local inference on M4 Max MacBook Pro.
What
TurboQuant (Google, ICLR 2026) is a three-stage KV cache compression method:
- PolarQuant — WHT rotation + polar coordinates + Lloyd-Max codebook (~4.2x compression)
- QJL — 1-bit quantized Johnson-Lindenstrauss residual correction
- TurboQuant — PolarQuant + QJL = ~3.5 bits/channel, zero accuracy loss
Why
Unlock 64K-128K context on qwen3.5:27b within 32GB unified memory. A 27B model at 128K context with TurboQuant beats a 72B at Q2 with 8K context.
Status
See issues for current progress.
Roles
- Strago: Build spec author
- Cid: Implementation, benchmarks, deployment
- Locke: Research support, upstream watch
- John: Quality review
- Frankie: Coordination
Source Repos
- TheTom/llama-cpp-turboquant — llama.cpp fork with Metal
- TheTom/turboquant_plus — Reference impl, 511+ tests
- amirzandieh/QJL — Author QJL code (CUDA)
- rachittshah/mlx-turboquant — MLX fallback
Docs
- Project Status — Full project status and build specification