Patrick Buckley e3af600a90 feat(deploy): vllm-litellm example — 3-model co-resident shape + HF loader (#688) ...

* feat(deploy): vllm-litellm example — 3-model co-resident shape + HF loader

Update the unified-memory inference example to the validated GB10 Spark shape:
qwen3.6-27B-FP8 (reasoning) + gemma-4-12B-it (perception) + Qwen3-Reranker-4B,
all co-resident on one GPU behind LiteLLM, loaded by HF id into a mounted
HF_HOME cache.

- qwen: MTP spec-decode + runai_streamer (weight load ~166s->1s) + full 256K at
  util 0.50 (default KV)
- gemma on the OpenAI lane (audio), reranker direct on :8002/rerank
- sequential startup + page-cache-drop guidance; runai_streamer kept on the big
  model only (its buffers break small models' KV budgets)
- README: HF-id loader, DGX Spark (validated) + AMD Strix Halo (ROCm) setup,
  tuning notes, troubleshooting
- wheel-check ALLOW entries for the example files (supersedes #687)

* docs(deploy): clarify AMD edits are compose literals (Copilot review)

In the Strix Halo guidance, --max-model-len and --load-format runai_streamer are
hard-coded in docker-compose.yml's vllm-qwen command, not .env vars — say where
to edit them.

2026-06-21 20:30:52 -07:00

..

workflows

feat(deploy): vllm-litellm example — 3-model co-resident shape + HF loader (#688)

2026-06-21 20:30:52 -07:00

renovate.json

feat: render rich media embeds for MCP tool results (#292)

2026-04-04 14:47:25 -07:00