rock
8790de6038
feat: add ComfyUI + rebalance GPU allocation (#17)
## GPU Rebalance (4× V100 32GB)
| Pod | Before | After |
|-----|--------|-------|
| reasoning (PP=2) | 2 GPU | 2 GPU |
| ornith | 2 GPU (2 replicas) | 1 GPU (1 replica) |
| comfyui | — | 1 GPU (**new**) |
| qwen-cpu | — | CPU on cp-2 (**new**) |
| embeddings/reranker | CPU | CPU |
## Changes
- `ornith.yaml`: scale 2→1, remove qwen2.5 co-loading, MAX_LOADED_MODELS=1
- `qwen-cpu.yaml`: new Ollama deployment on talos-cp-2 (144GB RAM), 5Gi PVC
- `k8s/apps/comfyui/`: new ComfyUI deployment (1 GPU, 50Gi model PVC, ingress)
- `58-comfyui.yaml`: ArgoCD Application (wave 8)
Gateway route update in separate PR (homelab-frontend).Reviewed-on: #17
Co-authored-by: rock <[email protected]>
2026-09-09 02:11:28 +00:00
..
2026-08-25 11:20:26 -07:00
2026-08-25 11:20:26 -07:00
2026-08-31 11:43:04 -07:00
2026-09-02 20:31:09 -07:00
2026-08-18 15:08:04 -07:00
2026-09-04 13:51:26 -07:00
2026-08-31 15:02:15 -07:00
2026-08-25 11:20:26 -07:00
2026-09-07 13:01:56 -07:00
2026-08-25 11:20:26 -07:00
2026-09-05 01:10:51 -07:00
2026-09-03 08:28:38 -07:00
2026-09-07 18:19:46 -07:00
2026-08-25 11:20:26 -07:00
2026-09-09 02:11:28 +00:00
2026-09-07 18:19:46 -07:00
2026-08-21 20:55:20 -07:00
2026-09-07 18:19:46 -07:00