All issues

Search results

BUG: stop vLLM inference when requests are aborted or disconnected

bugSomething isn't workingSomething isn't working

gpu

Status: Open.

#5548 In xorbitsai/inference;

· amumu96 opened on Sep 17, 2026

·  

v3.x

Qwen3 reranker computes full-sequence vocabulary logits, causing multi-GB VRAM spikes and OOMs

gpu

Status: Open.

#5542 In xorbitsai/inference;

· hypothese-zero opened on Sep 16, 2026

·  

v3.x

cutery errors

Status: Open.

#5523 In xorbitsai/inference;

· shaofengzeng opened on Sep 11, 2026

·  

v3.x

[Docker CD] Failure on v3.4.0

gpu

Status: Open.

#5522 In xorbitsai/inference;

· github-actions opened on Sep 11, 2026

·  

v3.x

[Docker CD] Failure on main

gpu

Status: Open.

#5494 In xorbitsai/inference;

· github-actions opened on Sep 5, 2026

·  

v3.x

V3.2.0版本，rerank接口调用无响应

gpu

Status: Open.

#5484 In xorbitsai/inference;

· ran411285752 opened on Sep 4, 2026

·  

v3.x

ENH: SDNQ Diffusers Quantizer

enhancementNew feature or requestNew feature or request

feature

stale

Status: Open.

#4633 In xorbitsai/inference;

· iwr-redmond opened on Feb 27, 2026

·  

v3.x