Jierui (Jerry) XuandClaude Opus 5 965ad68f1d [Kernel][MammothModa2] Use shared RMSNorm in the DiT (#7489) ...

Signed-off-by: Jerry Xu <xjr2423@gmail.com>
Co-authored-by: Claude Opus 5 <noreply@anthropic.com>

2026-09-29 00:12:16 +08:00

..

assets

[Diffusion][HunyuanImage3] Add request paged prefix caching (#7719)

2026-09-23 15:34:56 +00:00

attention

[BugFix][NPU] Fix Fish Speech S2 Pro NPU support (#7546)

2026-09-16 11:38:06 -07:00

benchmarks

feat(magi2): add fused BF16 routed MoE path (#7206)

2026-09-28 23:20:30 +08:00

buildkite

ci: extend AMD Z-Image cold-start timeout (#8195)

2026-09-27 06:11:05 +00:00

clients

[Bugfix] Fix Nemotron VoiceChat duplex nightly handshake (#7884)

2026-09-25 17:25:15 +00:00

config

feat(cosyvoice3): GPU-resident F0 predictor & pure-PyTorch vocoder (Task C2) (#7518)

2026-09-28 23:13:43 +08:00

core

[Hotfix][Core] PrefixCache: re-plan same-step prefix hits whose rows are written after the early prefetch (#7877)

2026-09-24 12:35:24 +00:00

dfx

[BugFix] Add field constraints to /v1/omni/sleep and /v1/omni/wakeup (#4740)

2026-09-28 09:30:30 -06:00

diffusion

[Kernel][MammothModa2] Use shared RMSNorm in the DiT (#7489)

2026-09-29 00:12:16 +08:00

distributed

[Bugfix] Fix MiniCPM-o 4.5 native duplex speech alignment (#7974)

2026-09-24 10:04:28 +00:00

docs

[Core][Diffusion] Add a typed pre-D2H video media contract (#6615)

2026-09-10 09:50:54 +00:00

e2e

[Core][Model] Fit PersonaPlex into the Unified Full-duplex Framework (RFC #7181 PR 2) (#7695)

2026-09-27 03:02:24 +00:00

engine

[Core][Model] Fit PersonaPlex into the Unified Full-duplex Framework (RFC #7181 PR 2) (#7695)

2026-09-27 03:02:24 +00:00

entrypoints

[BugFix] Add field constraints to /v1/omni/sleep and /v1/omni/wakeup (#4740)

2026-09-28 09:30:30 -06:00

examples

[BugFix] Download Anima assets for README CI test (#7928)

2026-09-23 17:04:51 +00:00

helpers

[CI][Diffusion] Add tiny model builder for Krea2Pipeline (#6421)

2026-09-25 21:54:38 +00:00

host_weight_runtime

[Bugfix][Core] Clean up failed HWR atomic metadata writes (#6956)

2026-09-11 09:03:26 +08:00

inputs

[Rebase] Rebase to vLLM 0.29.0 (#7230)

2026-09-10 12:08:28 +08:00

kernels

feat(magi2): add fused BF16 routed MoE path (#7206)

2026-09-28 23:20:30 +08:00

metrics

[Bugfix] Register statistics for dynamically added replicas (#7298)

2026-09-19 17:07:14 +00:00

model_executor

feat(cosyvoice3): GPU-resident F0 predictor & pure-PyTorch vocoder (Task C2) (#7518)

2026-09-28 23:13:43 +08:00

model_extras

fix(e2e): re-point HunyuanImage3 offline test to text_to_image.py (#7525)

2026-09-19 11:16:13 -07:00

model_tests

[CI][Diffusion] Restore Krea2 model type marker (#8164)

2026-09-26 12:55:06 +08:00

platforms

[Rebase] Rebase to vLLM 0.30.0 (#7820)

2026-09-22 15:52:01 +00:00

profile

[Refactor][CI] Reorganize test tree, and move slow e2e to weekly (#4951)

2026-07-29 23:53:03 +08:00

protocol

[Model] Integrate AURA into Full-Duplex Runtime (#7633)

2026-09-21 09:26:54 -07:00

quantization

[bugfix] Fix FP8 Quack kernels when using Context Parallelism and reduce number of recompilations (#7458)

2026-09-22 11:38:17 -07:00

reasoning

[Refactor][CI] Reorganize test tree, and move slow e2e to weekly (#4951)

2026-07-29 23:53:03 +08:00

rl_rollout

[P0][RFC #3747] RL rollout serving: session management + world_model_env (#4770)

2026-09-14 09:26:44 +08:00

tools

[Refactor] Declare model-local KV held outside the paged manager (#6171)

2026-09-11 08:44:32 +08:00

utils

[Bugfix] Build the forced-aligner prompt without a chat template (word timestamps one bin late) (#7240)

2026-09-08 19:27:38 +00:00

worker

[Bugfix] Fix MiniCPM-o 4.5 seeded turns and window E2E inputs (#8160)

2026-09-26 01:42:53 +08:00

worker_v2

[Model] Optimize Qwen3-TTS streaming throughput and first audio on CUDA (#8163)

2026-09-26 22:51:04 +08:00

__init__.py

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

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

conftest.py

[CI]Split GPU jobs by cards_* and reject hand-written SKU marks (#6650)

2026-08-29 18:42:33 +08:00