Search results

[Bug] Slow inference speed for fishaudio_s2_pro on H200 GPU with sglang-omni

Status: Open.

#187 In sgl-project/sglang-omni;

· lazylazypig opened on Mar 23, 2026

[Bug] Local S2 Pro cloning output quality significantly worse than fish audio S2 pro web

Status: Open.

#182 In sgl-project/sglang-omni;

· tommymir opened on Mar 19, 2026

[Feature] Scheduler CPU overhead optimization

Status: Open.

#179 In sgl-project/sglang-omni;

· PasserBy4 opened on Mar 19, 2026

[Bug] Fix 3 failing + 2 erroring tests in CI on main

Status: Open.

#175 In sgl-project/sglang-omni;

· zhaochenyang20 opened on Mar 18, 2026

[Refactor] Decouple TTS usage metrics from OpenAI-compatible /v1/audio/speech response

Status: Open.

#174 In sgl-project/sglang-omni;

· zhaochenyang20 opened on Mar 17, 2026

Feature] Framework-level fine-grained torch.compile + Mega Cache for Omni models

Status: Open.

#172 In sgl-project/sglang-omni;

· zhaochenyang20 opened on Mar 17, 2026

[Feature] Improve Relay Stability

Status: Open.

#166 In sgl-project/sglang-omni;

· FrankLeeeee opened on Mar 17, 2026

[Feature] Improve Model Executor Perf

Status: Open.

#165 In sgl-project/sglang-omni;

· FrankLeeeee opened on Mar 17, 2026

[Benchmark] Measure performance of existing systems

Status: Open.

#164 In sgl-project/sglang-omni;

· FrankLeeeee opened on Mar 17, 2026

[Feature] Scheduler Optimization

Status: Open.

#163 In sgl-project/sglang-omni;

· FrankLeeeee opened on Mar 17, 2026

[BUG] Coordinator leaks request/future state when submit_to_stage fails

Status: Open.

#162 In sgl-project/sglang-omni;

· pikaxinge opened on Mar 17, 2026

[Bug/Hardware Incompatibility] RTX 5090 (Blackwell / sm_120) crashes with "_deps/repo-flash-attention-src/hopper/flash_fwd_launch_template.h:166) during inference

Status: Open.

#160 In sgl-project/sglang-omni;

· fakirsu opened on Mar 17, 2026