anodyine/vllm-omni
1
0 Fork 0
Code Issues 4 Pull Requests 2 Actions Packages Projects Releases Wiki Activity
3,433 Commits 235 Branches 31 Tags
dev/vllm-align
Commit Graph
339 Commits
Author SHA1 Message Date
tzhouam df14622fb4 Drop CosyVoice regression changes superseded by main 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-05 11:41:24 +08:00
tzhouam 6f5f9b97e3 Merge current Omni main and retain conditioning regression coverage 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-04 13:50:06 +08:00
Sy03 1c1b831f58 [Model] Batch Qwen3-TTS Talker prompt projection and settled-row preprocessing (#8477) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-10-04 05:08:24 +00:00
andyluo7 b171731c81 fix: preserve CosyVoice full payload and live conditioning (#8343) 
Signed-off-by: andyluo7 <andy.luo@amd.com>
 2026-10-04 16:03:27 +13:00
andyluo7 09d704907f fix: preserve Torch seeded sampling on ROCm (#8341) 
Signed-off-by: andyluo7 <andy.luo@amd.com>
 2026-10-04 16:03:21 +13:00
tzhouam df3807eca4 Keep release alignment focused and clarify baseline fixes 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-04 03:11:35 +08:00
tzhouam b256e1ec4e Adapt Moss sampling and runner fixtures to vLLM 0.31 interfaces 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-03 22:22:39 +08:00
tzhouam 864bc7b22a test: update the Moss runner fixture for release output APIs 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-03 22:06:27 +08:00
tzhouam ed876de879 rebase: target the official vLLM 0.31.0 release 
Use upstream release images at db9527a46873, preserve custom processor conditioning, and include weekly jobs in full Buildkite release validation.

Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-03 21:11:54 +08:00
tzhouam 008fa25263 Merge main into dev/vllm-align for vLLM 0.31 release 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-03 21:05:17 +08:00
tzhouam 84cbf94b1b fix: preserve Omni processor conditioning on vLLM 0.31 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-03 21:03:26 +08:00
andyluo7 ee8fdab1dc [CI/Build] Fix MOSS CPU tests on ROCm workers (#8447) 
Signed-off-by: andyluo7 <andy.luo@amd.com>
 2026-10-03 09:13:51 +08:00
Sy03andYueqian Lin 3bc3f1a7d0 [Model] Optimize MOSS Local 1.5 MRV2 serving and progressive chunks (#8213) 
Signed-off-by: Sy03 <1370724210@qq.com>
Signed-off-by: Yueqian Lin <linyueqian@outlook.com>
Co-authored-by: Yueqian Lin <linyueqian@outlook.com>
 2026-10-02 18:03:14 +00:00
Sy03 2ba5074ca0 [Model] Optimize CosyVoice3 packed streaming with bounded HiFT graphs (#8420) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-10-02 08:49:50 -07:00
Nick CaoandCodex 0e6791e9c4 [Bugfix] Initialize eager state in Higgs test fixture (#8426) 
Signed-off-by: Nick Cao <ncao@redhat.com>
Co-authored-by: Codex <noreply@openai.com>
 2026-10-02 10:22:52 -04:00
Sy03 e372aa3a41 [Bugfix] Restore Fish Speech and CosyVoice3 audio quality at high concurrency (#8422) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-10-02 07:11:04 -07:00
Sy03 e84014dffb [Model] Add single-stage Qwen3-TTS with in-Talker streaming codec decode (#8259) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-10-02 21:34:30 +10:00
Sy03 d93e1a8e09 [Model] Optimize Qwen3-Omni MRv2 handoff, code prediction and audio decoding (#8223) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-10-01 22:22:54 -07:00
tzhouam 2721f16ca8 rebase: align vllm-omni with vLLM ac9126e58aa7 
Signed-off-by: tzhouam <tzhouam@connect.ust.hk>
 2026-10-02 10:37:50 +08:00
andyluo7 bbee488da3 fix: repair ROCm test routing, CPU timeouts, and Breeze autotuning (#8342) 
Signed-off-by: andyluo7 <andy.luo@amd.com>
 2026-10-02 07:36:41 +08:00
amy-why-3459 4af28f33bd [Bugfix] Defer VoxCPM2 NPU patch to avoid platform import cycle (#8330) 
Signed-off-by: amy-why-3459 <wuhaiyan17@huawei.com>
 2026-09-30 05:25:36 -07:00
Sy03 56763cd6c3 [Model] Enable Higgs Audio v3 MRV2 streaming and fix mixed-prefill capture bounds (#8226) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-09-30 11:32:27 +00:00
Armaan AmatyaYueqian LinHongsheng Liu
 		aa41a3ac97 [Bugfix] Fix the remaining Python 3.10 breakages (#7985) 
Signed-off-by: Armaan Amatya <armaanamatya2014@gmail.com>
Signed-off-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
Co-authored-by: Yueqian Lin <70319226+linyueqian@users.noreply.github.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-09-30 11:38:23 +08:00
Egor Chernobrovkin c27a52a359 [Model] Make VoxCPM2 generation settings configurable (#7010) 
Signed-off-by: Egor Chernobrovkin <33jasooon33@gmail.com>
 2026-09-30 11:26:12 +08:00
yulunandHongsheng Liu 9097fa6547 [Hardware][Ascend] Accelerate VoxCPM2 LocDiT with NPUGraph (#7275) 
Signed-off-by: Big2Wheel <1404207590@qq.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-09-30 10:11:22 +08:00
Sy03 5af3a16526 [Model] Add MiniCPM-o 4.5 turn-mode MRv2 and batched audio inference (#8222) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-09-29 11:41:39 -07:00
jingchengtianandjingchengtian 43ef1e61fb [Perf][TTS] Add per-row generator support to MOSS-TTS talker to eliminate serial fallback for seeded batched requests (#7922) 
Signed-off-by: jingchengtian <jingchengtian@users.noreply.github.com>
Signed-off-by: jingchengtian <tjc1995@126.com>
Co-authored-by: jingchengtian <jingchengtian@users.noreply.github.com>
 2026-09-29 13:13:50 -04:00
Sy03andzack 1ea0ee3b38 [Model] Add opt-in packed Flow and batched HiFT inference for CosyVoice3 (#8224) 
Signed-off-by: Sy03 <1370724210@qq.com>
Signed-off-by: zack <huixindaddy@yahoo.com>
Co-authored-by: zack <huixindaddy@yahoo.com>
 2026-09-29 15:48:36 +00:00
Wallbreazzz 5b3abe2f4d [Feature][TTS] Port static chunk ramp to Moss-TTS streaming codec decode (#8124) 
Signed-off-by: Wallbreazzz <110282866+Wallbreazzz@users.noreply.github.com>
 2026-09-29 02:28:15 +00:00
BeatSeat 3e62bba3cb [Performance][MiniCPM-o] Whole-Euler CUDA graphs for Code2Wav with a shared attention arena (#8007) 2026-09-29 09:14:12 +10:00
SYLARlishunyang12Roger Wang
 		7266fc6138 [Bugfix] Avoid enlarging MiniMax H3 reference images and videos (#8253) 
Signed-off-by: lishunyang12 <lishunyang12@users.noreply.github.com>
Co-authored-by: lishunyang12 <lishunyang12@users.noreply.github.com>
Co-authored-by: Roger Wang <hey@rogerw.io>
 2026-09-28 20:12:53 +00:00
BeatSeatandHongsheng Liu 2a453c5ad8 feat(cosyvoice3): GPU-resident F0 predictor & pure-PyTorch vocoder (Task C2) (#7518) 
Signed-off-by: BeatSeat <wendavid552@gmail.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-09-28 23:13:43 +08:00
chickeytonandClaude Opus 5.5 5ce6f5dc49 [Bugfix][MiniCPM-o] Forward a speech unit closed by LISTEN after turn_eos to the Talker (#8227) 
Signed-off-by: chickeyton <ngton2014@gmail.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-28 11:11:42 +00:00
Yueqian Lin f5bdb20e9d [Perf][PersonaPlex] Serve 16 concurrent realtime duplex sessions on one GPU (#8192) 2026-09-28 04:51:38 +00:00
BeatSeatandamy-why-3459 c8a1419e02 [Performance][MiniCPM-o] Zero-copy block-aligned KV sliding window with in-place Re-RoPE for duplex Stage-0 (#7821) 
Signed-off-by: BeatSeat <wendavid552@gmail.com>
Co-authored-by: amy-why-3459 <wuhaiyan17@huawei.com>
 2026-09-28 10:11:54 +08:00
chickeytonClaude Fable 5.1Yueqian Lin
 		798d52ce55 [Core][Model] Fit PersonaPlex into the Unified Full-duplex Framework (RFC #7181 PR 2) (#7695) 
Signed-off-by: chickeyton <ngton2014@gmail.com>
Signed-off-by: Yueqian Lin <linyueqian@outlook.com>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Co-authored-by: Yueqian Lin <linyueqian@outlook.com>
 2026-09-27 03:02:24 +00:00
Sy03 e1303e5067 [Model] Optimize Qwen3-TTS streaming throughput and first audio on CUDA (#8163) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-09-26 22:51:04 +08:00
Sy03 d31a2a3ba8 [Bugfix] Fix Nemotron VoiceChat duplex nightly handshake (#7884) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-09-25 17:25:15 +00:00
chickeytonandClaude Opus 5.5 4be6bbe0e4 [Bugfix][MiniCPM-o] Build the Stage-0 duplex runtime at load time (#8146) 
Signed-off-by: chickeyton <ngton2014@gmail.com>
Co-authored-by: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
 2026-09-25 14:57:23 +00:00
Yuanheng Zhao 3bd5ac9685 [Bugfix] Align Ming-Image prompt templates and RNG, fix number of layer to decompose (#8118) 
Signed-off-by: Yuanheng <jonathan.zhaoyh@gmail.com>
 2026-09-25 22:47:29 +08:00
DanaerLeezijianc2hsliuustc0106Hongsheng Liu
 		b41e6707b1 [Bugfix][MammothModa2] Stop allocating AR KV cache for the replaced language model (#8095) 
Signed-off-by: MrlixiangWE <mrdanaer@gmail.com>
Co-authored-by: zijianc2 <157244773+zijianc2@users.noreply.github.com>
Co-authored-by: hsliuustc0106 <222337142+hsliuustc0106@users.noreply.github.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-09-25 21:28:59 +08:00
NATURE 9f3f475685 [CI/Build] Fix MiniCPM-o codec readback test under default-device dispatch (#8147) 2026-09-25 08:45:05 +00:00
Sy03 551a03102f [Bugfix][Qwen3-TTS] Decode full payload from model buffer (#8132) 
Signed-off-by: Sy03 <1370724210@qq.com>
 2026-09-25 10:03:58 +08:00
Khairul Kabirandkhairulkabir1661 63bbd19a1d [Bugfix] Read hidden layer key from model config instead of hardcoding 24 (#5825) 
Signed-off-by: khairulkabir1661 <khairulkabir1661@users.noreply.github.com>
Co-authored-by: khairulkabir1661 <khairulkabir1661@users.noreply.github.com>
 2026-09-24 09:19:41 -06:00
NATURExjwuEric
 		9c1656c607 [Model] Reduce MiniCPM-o 4.5 reference setup and Talker overhead (#7968) 
Signed-off-by: natureofnature <wzliu@connect.hku.hk>
Co-authored-by: xjwu <wuxianjin633@gmail.com>
Co-authored-by: Eric(Qun) <186020953+0xtoward@users.noreply.github.com>
 2026-09-24 05:18:56 -07:00
NATUREandliuqihao 7a8d956ece [Bugfix] Fix MiniCPM-o 4.5 native duplex speech alignment (#7974) 
Signed-off-by: natureofnature <wzliu@connect.hku.hk>
Co-authored-by: liuqihao <liuqihao970610@gmail.com>
 2026-09-24 10:04:28 +00:00
Yuanheng Zhao 71125f80ad [New Model] Ming-Image (inclusionAI/Ming-Image-0.1-Design) Support (#8021) 
Signed-off-by: Yuanheng <jonathan.zhaoyh@gmail.com>
 2026-09-23 12:20:06 +00:00
Yuan WuandHongsheng Liu cb5f508212 [Feature] Transfer AR-to-diffusion payloads over NIXL (#6264) 
Signed-off-by: yuanwu <yuan.wu@intel.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>
 2026-09-23 18:26:53 +08:00
Yueqian Lin 3f473a358b [Model] Accelerate Breeze-TTS-2 non-streaming inference (#7752) 2026-09-22 17:10:15 -07:00
Jiaqian Liu 7b6a5ee894 [Hardware][Ascend] Add opt-in chunked anti-alias FIR upsampling (#7896) 
Signed-off-by: Jiaqian Liu <591998922@qq.com>
 2026-09-22 10:24:01 -07:00