3 people 946a6fddd5 Fix MoT RMSNorm for vLLM 0.21 ...

Use the vLLM IR RMSNorm op after the legacy layernorm helper was removed, keeping MoT fused paths unchanged while restoring BAGEL compatibility.

Co-authored-by: Yufeng Zhu <yzhu802@gatech.edu>
Co-authored-by: Cursor <cursoragent@cursor.com>
Signed-off-by: princepride <wangzhipeng628@gmail.com>

2026-05-29 02:34:52 +00:00

..

assets

[Model] Add GLM-TTS text-to-speech model support (#3141)

2026-05-24 05:42:26 +08:00

attention

[TTS][Perf] Optimize Fish Speech S2 Pro high-concurrency serving (#3773)

2026-05-27 13:25:49 -05:00

benchmarks

[Bench] Add audio-streaming continuity metric for TTS (#3618)

2026-05-20 05:08:56 +00:00

comfyui

[Bugfix] ComfyUI image-to-image DALL-E endpoint cases #2886 (#2980)

2026-04-21 21:37:52 +08:00

config

[Bugfix]: skip faulty pipelines during registry iteration (#2999)

2026-05-06 08:35:41 +00:00

core

[3/5][core]refactor communication layer: PR 3 of 5, all other models in non async mode (#3719)

2026-05-28 14:06:41 +08:00

dfx

[Perf] Bagel Performance Nightly CI test (#2175)

2026-05-28 09:41:46 +08:00

diffusion

Fix MoT RMSNorm for vLLM 0.21

2026-05-29 02:34:52 +00:00

distributed

Relax pixel tolerance to 15 to account for Triton kernel numerical drift

2026-05-29 02:34:52 +00:00

e2e

[AutoRound] Support GLM-Image W4A16 quantization model (#3059)

2026-05-28 07:19:06 +00:00

engine

[Bugfix] Ensure stage and diffusion subprocesses exit when parent dies unexpectedly (#3751)

2026-05-28 08:22:04 +08:00

entrypoints

[BugFix] Fix diffusion parallel_config YAML override and add deploy config field allowlist (#3483)

2026-05-28 01:00:32 +08:00

examples

[Doc] Refine vace offline inference example README (#3584)

2026-05-28 14:45:45 +08:00

helpers

[Test] Add long text output correctness test for Qwen3-Omni (#3539)

2026-05-28 08:27:29 +08:00

inputs

[Bugfix] GLM-Image: route t2i requests through the multimodal processor (#3034) (#3189)

2026-05-04 21:07:53 +08:00

metrics

[Feature] Opt metrics structure (#891)

2026-02-09 19:17:32 +08:00

model_executor

[Refactor][Qwen3-TTS] Construct speech tokenizer encoder natively (#3360)

2026-05-28 16:20:49 +08:00

platforms/npu/quant

support online FP8 quantization for FA on NPU #2236 (#2640)

2026-05-13 11:46:00 +08:00

profile

[Enhancement] add pytorch profiler ops and memory record (#2472)

2026-04-20 20:58:14 +08:00

utils

[Refactor] Remove sox from dependencies (#2745)

2026-04-17 01:24:47 +00:00

worker

[3/5][core]refactor communication layer: PR 3 of 5, all other models in non async mode (#3719)

2026-05-28 14:06:41 +08:00

__init__.py

[Fix] Resolve comments & update vLLM-Omni name usages. (#122)

2025-12-01 08:05:24 +08:00

conftest.py

[CI] invalid_param reliability suite and weekly http_invalid jobs (#3652)

2026-05-19 14:28:47 +08:00

test_arg_utils.py

[BugFix] Fix diffusion parallel_config YAML override and add deploy config field allowlist (#3483)

2026-05-28 01:00:32 +08:00

test_config_factory.py

[3/5][core]refactor communication layer: PR 3 of 5, all other models in non async mode (#3719)

2026-05-28 14:06:41 +08:00

test_data_entry_keys.py

[Refactor] msgspec standardisation for data entry key names and improved type checks (#3149)

2026-05-12 14:30:13 +08:00

test_diffusion_config_fields.py

[Config] Remove invalid LLM-only engine_args from diffusion stage configs (#2622)

2026-04-11 01:36:29 +12:00

test_diffusion_config_propagation.py

[Feat][Config] Support additional_config for diffusion worker (#3020)

2026-05-13 14:22:43 +00:00

test_generate_nightly_perf_excel.py

[Refactor]Rename diffusion benchmark backend to endpoint (#3137)

2026-05-19 11:30:33 +08:00

test_generate_nightly_perf_html.py

[Refactor]Rename diffusion benchmark backend to endpoint (#3137)

2026-05-19 11:30:33 +08:00

test_outputs.py

[Misc] Migrate L1 tests to use pytest-mock (#1315)

2026-02-26 17:08:39 +08:00

test_speaker_cache_integration.py

[TTS][SpeakerCacheManager] A global speaker cache manager for Voice Cloning (#2630)

2026-05-05 10:53:45 -04:00

test_speaker_cache.py

[TTS][SpeakerCacheManager] A global speaker cache manager for Voice Cloning (#2630)

2026-05-05 10:53:45 -04:00

test_speaker_metadata_persistence.py

[TTS][SpeakerCacheManager] A global speaker cache manager for Voice Cloning (#2630)

2026-05-05 10:53:45 -04:00

test_version.py

[Misc] Warn When vLLM / vLLM-Omni Have Mismatched Versions (#2691)

2026-04-17 20:15:16 +08:00