forked from vllm-project/vllm
-
Notifications
You must be signed in to change notification settings - Fork 1
Pull requests: Jackie2049/vllm
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
fix(lora): use block_n=32 on sm_90 to avoid lora_expand NaN
#9
opened Jul 15, 2026 by
Jackie2049
Owner
Loading…
4 tasks done
[Feature] Add request overload control with HTTP 429 queue timeout eviction
#8
opened Jun 4, 2026 by
Jackie2049
Owner
Loading…
2 of 4 tasks
[Example] Add Top-n-sigma logit truncation custom logits processor
#7
opened Jun 4, 2026 by
Jackie2049
Owner
Loading…
4 tasks done
[Bugfix] Replace assertions with defensive guards in KV transfer finished callbacks
#6
opened Jun 4, 2026 by
Jackie2049
Owner
Loading…
4 tasks done
[Bugfix] Replace assertion with guard in finished_sending KV transfer
#5
opened May 21, 2026 by
Jackie2049
Owner
Loading…
ProTip!
Updated in the last three days: updated:>2026-08-06.