Abiba Bot e6a6c30211 router: v3 — 5-tier routing with vision guard, MoE spillover, code hint ...

GUARD: multimodal -> VLM only
TIER 1 Tiny: VLM->Dense->MoE
TIER 2 Light: VLM->Dense->MoE
TIER 3 Medium: MoE<->Dense (40% spill)->VLM
TIER 4 Heavy: Dense->MoE->VLM (quality first)
TIER 5 Default: MoE<->Dense (40% spill)->VLM
HINTS: speed->VLM, quality->MoE, code->Dense

+ moe_spillover(): 60/40 MoE/Dense load distribution

2026-06-05 23:48:45 +00:00

..

Dockerfile

Initial commit: CT 116 inference harness — nginx, LiteLLM, router, dashboard, Redis

2026-05-16 18:51:50 +00:00

requirements.txt

Initial commit: CT 116 inference harness — nginx, LiteLLM, router, dashboard, Redis

2026-05-16 18:51:50 +00:00

route_v2.py

router: reference - new 5-tier routing with vision guard (deployed CT116)

2026-06-05 23:32:27 +00:00

route_v3.py

router: v3 — 5-tier routing with vision guard, MoE spillover, code hint

2026-06-05 23:48:45 +00:00

router.py

router: fix Dense context window 98K→262K (matches actual n_ctx)

2026-06-05 23:15:18 +00:00