Block a user
Probe whether the Qwen-Image conditioning KV can spill to host DRAM on the RTX 3090
Implemented in PR: anodyine/vllm-omni#4
Verdicts (per the issue's three-world taxonomy)
RTX 3090 (24 GB): not-built. No host-DRAM spill path for the…
probe: Qwen-Image conditioning-KV spill probe on RTX 3090 / RTX 5080 (issue 2)
Investigate and fix the TP chroma-speckle corruption upstream (vllm-omni#8135)
Validate Qwen-Image-2.1 serving on Ampere (RTX 3090) and report upstream
Addendum (2026-10-07, owner-authorized): the review draft below was completed with four additions — same-stack Diffusers A/B, warmup-1 + median-of-3 serving latency with exact API-ready time,…
Validate Qwen-Image-2.1 serving on Ampere (RTX 3090) and report upstream
Validation complete — sm_86 results and upstream draft
All four acceptance criteria are met. Evidence lives on cachyos-alexander under ~/Projects/vllm-omni-val/issue1/ (logs, renders,…
Probe whether the Qwen-Image conditioning KV can spill to host DRAM on the RTX 3090
Validate Qwen-Image-2.1 serving on Ampere (RTX 3090) and report upstream
Session-state correction (2026-10-08): the relayview-inference.service note above described the 2026-10-07 session. Its enabled/disabled state is owner-controlled and may differ between sessions…
anodyine
created branch worktree-enable-twine-upload in anodyine/vllm-omni
2026-10-07 20:48:41 -04:00