tzhouamandClaude Fable 5.1 602e39647e [Edge][WP2] Finalize edge profile init knobs from init_profile; WP4 CPU emulation results ...

init_profile 20260909-112200 (1.7B, L20X, warm): default 152.0 s, eager_stage1
149.9 s, eager_both 110.9 s (first TTFA 15 s, no graphs), parallel_stage_init
94.6 s, edge 145.7 s. Edge profile now carries `orchestrator: parallel_stage_init:
true`, applied by the deploy_profile selector (explicit values win); graphs stay on.

hw_emulation 20260909-110229 (CPU venv, 0.6B, deploy_profile=auto):
x86_cpu_avx2 (4c, fp32) TTFA 2.65 s RTF 11.9; x86_cpu_avx512 (8c) TTFA 0.68 s
RTF 2.45; x86_cpu_amx (16c) TTFA 0.51 s RTF 1.72; arm64 logic-only same as avx2.
Emulation rows now record the auto-detected hardware class.

Adds tests/e2e/offline_inference/test_qwen3_tts_edge_profile.py (L3, one GPU).

Co-Authored-By: Claude Fable 5.1 <noreply@anthropic.com>

2026-09-09 14:40:13 +08:00

..

edge

[Edge][WP2] Finalize edge profile init knobs from init_profile; WP4 CPU emulation results

2026-09-09 14:40:13 +08:00

audex_s2s_30b.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_s2s.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_thinker_only_30b.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_thinker_only.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_tta_30b.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_tta.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

audex_tts_30b.yaml

[Refactor] Remove legacy --stage-configs-path from the serve CLI (#5647)

2026-08-03 23:41:52 +08:00

audex_tts.yaml

[New Model] support nemotron Audex (#4976)

2026-07-28 03:58:22 +00:00

aura_omni.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

bagel_single_stage.yaml

[Docs] Remove stale stage config references (#6270)

2026-08-20 07:09:12 +08:00

bagel_think.yaml

[Config Refactor]: Remove bagel yaml (#2936)

2026-04-24 21:21:30 +08:00

bagel.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

cosyvoice3.yaml

[Hardware] Enable CosyVoice3 on Moore Threads MUSA (#5164)

2026-07-29 02:46:27 +00:00

covo_audio.yaml

[New Model] Add support for tencent/Covo-Audio-Chat (#2293)

2026-05-12 14:39:58 +08:00

dots_tts.yaml

[Model][TTS][WIP] Add dots.tts (rednote-hilab) — continuous-AR 48kHz TTS (#4765)

2026-08-13 17:58:33 +00:00

dreamzero_tp1_cfg2.yaml

[Bugfix] Select the AR-Diffusion engine in the DreamZero CFG-parallel deploy config (#5439)

2026-07-28 17:14:26 +08:00

dreamzero.yaml

[BDE] Phase 1: engine-level KV cache management for DreamZero (RFC #4366) (#4534)

2026-07-05 07:17:58 +08:00

dynin_omni_ci.yaml

[Refactor] Migrate dynin_omni to pipeline registry, drop legacy stage… (#4078)

2026-06-02 21:28:12 +00:00

dynin_omni_multiconnector.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

dynin_omni.yaml

[Refactor] Migrate dynin_omni to pipeline registry, drop legacy stage… (#4078)

2026-06-02 21:28:12 +00:00

fish_qwen3_omni_2gpu.yaml

[Docs] Remove stale stage config references (#6270)

2026-08-20 07:09:12 +08:00

fish_qwen3_omni_high_concurrency_dual_gpu.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

fish_qwen3_omni_high_concurrency_single_gpu.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

fish_qwen3_omni.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

gepard.yaml

[Model] Add Gepard-1.0 native-AR FSQ/NanoCodec TTS (offline inference) (#5666)

2026-08-14 22:06:05 -07:00

glm_image.yaml

glm-image: fix(npu)per-stage runtime env for HCCL ports + GLM-Image NPU stage config (#3235)

2026-05-27 12:18:47 +12:00

glm_tts.yaml

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

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

Gr00tN1d7.yaml

[Diffusion] add GR00T-N1.7 pipeline with OpenPI serving (#3798)

2026-06-20 11:45:26 +08:00

higgs_audio_v2.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

higgs_multimodal_qwen3_high_throughput.yaml

[Bugfix][Higgs-Audio-V3] Disable XQA decode (#6068)

2026-08-13 10:00:50 -07:00

higgs_multimodal_qwen3_low_latency.yaml

[Bugfix][Higgs-Audio-V3] Disable XQA decode (#6068)

2026-08-13 10:00:50 -07:00

higgs_multimodal_qwen3.yaml

[BugFix][Higgs-Audio-V3] Fix transcript and concurrent sampling tests (#6422)

2026-08-30 08:52:15 +00:00

hunyuan_image3_ar.yaml

[Config] Add HunyuanImage3 deploy configs (#3172)

2026-05-11 22:49:19 +08:00

hunyuan_image3_dit.yaml

[Feature] Add AllGather-KV sequence-parallel attention str… (#4968)

2026-07-24 14:45:26 +00:00

hunyuan_image_3_moe.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

hunyuan_video_15.yaml

[Refactor] Migrate Legacy Stage Configs to the Pipeline Registry (#5031)

2026-07-18 09:34:33 +00:00

indextts2_5.yaml

[New Model] Support IndexTTS 2.5 (#5957)

2026-08-11 23:00:35 +00:00

indextts2_low_latency.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

indextts2.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

lance.yaml

[New Model] Add Lance (ByteDance) (#3710)

2026-06-01 16:01:25 +08:00

lingbot_world_v2_stepwise.yaml

[Feature][LingBot World]stepwise execution (#6844)

2026-09-07 13:38:23 +08:00

mammoth_moda2_ar.yaml

[Fix] Restore MammothModa2 support (#6694)

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

mammoth_moda2.yaml

[Fix] Restore MammothModa2 support (#6694)

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

mimo_audio_5090d.yaml

[Doc] Add MiMo-Audio recipe for RTX 5090/5090D 32GB (#6559)

2026-08-26 16:26:24 +08:00

mimo_audio.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

ming_flash_omni_image.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

ming_flash_omni_thinker_only.yaml

[Config Refactor] Migrate Ming and Ming-TTS deploy/pipline configs (#3154)

2026-05-03 14:43:49 +12:00

ming_flash_omni_tts.yaml

[Config Refactor] Migrate Ming and Ming-TTS deploy/pipline configs (#3154)

2026-05-03 14:43:49 +12:00

ming_flash_omni.yaml

[Config Refactor] Migrate Ming and Ming-TTS deploy/pipline configs (#3154)

2026-05-03 14:43:49 +12:00

ming_tts_moe.yaml

[Perf] Ming-TTS Stage-0 flow-head speedup: RoPE/QKV fusion, opt-in torch.compile, PIECEWISE cudagraph (#4942)

2026-07-12 19:50:51 -07:00

ming_tts.yaml

[Perf][Ming-TTS] Add initial latent chunk for lower streaming TTFP (#5011)

2026-07-13 06:45:28 +08:00

minicpmo_4_5_2gpu.yaml

Fix MiniCPM-o concurrent audio first-packet latency (#6767)

2026-08-29 23:27:47 +08:00

minicpmo_4_5_3gpu_stage1_replicas.yaml

[Bugfix][MiniCPM-o] Serve native duplex from shipping YAMLs and fence… (#6619)

2026-08-26 10:56:51 +08:00

minicpmo_4_5_3gpu.yaml

Fix MiniCPM-o concurrent audio first-packet latency (#6767)

2026-08-29 23:27:47 +08:00

minicpmo_4_5_4gpu_stage1_replicas.yaml

[Bugfix][MiniCPM-o] Serve native duplex from shipping YAMLs and fence… (#6619)

2026-08-26 10:56:51 +08:00

minicpmo_4_5_8x4090_stage1_replicas.yaml

[Bugfix][MiniCPM-o] Serve native duplex from shipping YAMLs and fence… (#6619)

2026-08-26 10:56:51 +08:00

minicpmo_4_5_8x4090.yaml

[Bugfix][MiniCPM-o] Serve native duplex from shipping YAMLs and fence… (#6619)

2026-08-26 10:56:51 +08:00

minicpmo_4_5.yaml

[Model] Preserve MiniCPM-o Talker full attention until capacity (#6799)

2026-09-06 22:25:35 +08:00

minimax_h3_disaggregated_turbo.yaml

[Feature] Support text encoder Disaggregation for H3 (#5885)

2026-08-28 14:27:05 +08:00

minimax_h3_disaggregated.yaml

[Feature] Support text encoder Disaggregation for H3 (#5885)

2026-08-28 14:27:05 +08:00

minimax_music3_2gpu.yaml

[Model] Add MiniMax Music 3 text-to-music (#6186)

2026-08-18 04:33:13 +00:00

minimax_music3.yaml

[Model] Add MiniMax Music 3 text-to-music (#6186)

2026-08-18 04:33:13 +00:00

moss_sound_effect.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

moss_tts_80gb.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

moss_tts_local.yaml

[BugFix] Use MOSS-TTS-Local official sample params (#6156)

2026-08-13 08:56:26 +00:00

moss_tts_nano.yaml

[Docs] Remove stale stage config references (#6270)

2026-08-20 07:09:12 +08:00

moss_tts_realtime.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

moss_tts.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

moss_ttsd.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

moss_voice_generator.yaml

[Core] Cleanup Omni connector runtime and transfer backends (#5096)

2026-07-22 09:44:15 +08:00

nemotron_labs_voicechat_duplex.yaml

[Perf][NemotronVoiceChat] Fast defaults for all three deploy profiles: 2.8x offline, realtime full-duplex audio (#6354)

2026-09-03 18:49:19 -07:00

nemotron_labs_voicechat_streaming.yaml

[Perf][NemotronVoiceChat] Fast defaults for all three deploy profiles: 2.8x offline, realtime full-duplex audio (#6354)

2026-09-03 18:49:19 -07:00

nemotron_labs_voicechat.yaml

[Perf][NemotronVoiceChat] Fast defaults for all three deploy profiles: 2.8x offline, realtime full-duplex audio (#6354)

2026-09-03 18:49:19 -07:00

omnivoice.yaml

[Refactor] Migrate MammothModa2 and OmniVoice to pipeline registry (#4647)

2026-07-08 00:44:33 +00:00

personaplex.yaml

[Feature] PersonaPlex (Moshi-based full-duplex S2S): native vLLM port + duplex serving (#4771)

2026-08-04 20:10:33 -07:00

pi0.yaml

Add π0 VLA model support (#4222)

2026-08-17 14:29:17 +08:00

qwen2_5_omni.yaml

[New Model] Support IndexTTS 2.5 (#5957)

2026-08-11 23:00:35 +00:00

qwen3_omni_moe_mori_intranode.yaml

add moriio transfer engine (#1742)

2026-06-10 08:37:33 +08:00

qwen3_omni_moe_thinking.yaml

[Model] Qwen3-Omni: register thinker-only pipeline for Instruct serve (#6284)

2026-08-27 08:21:03 +00:00

qwen3_omni_moe.yaml

[Bugfix] Fix Qwen3-Omni MoE backend selection (#7019)

2026-09-04 21:16:58 +08:00

qwen3_tts_forced_aligner.yaml

[Feature] timestamps for TTS via a forced-aligner pooling (#4795)

2026-08-12 05:33:34 +00:00

qwen3_tts_high_concurrency.yaml

[Feature][TTS] adaptive chunk ramp (Phase 2 buffer-feedback controller) (#6001)

2026-08-28 02:11:05 -07:00

qwen3_tts_talker_only.yaml

[Edge][WP5] Add codec-token streaming mode with on-device decoding

2026-09-09 11:10:13 +08:00

qwen3_tts.yaml

[Edge][WP3] Add CPU OmniPlatform (AR + generation stages on vLLM's CPU backend)

2026-09-09 11:02:14 +08:00

README_higgs_audio_v3.md

[Bugfix][Higgs-Audio-V3] Disable XQA decode (#6068)

2026-08-13 10:00:50 -07:00

step_audio2_ci.yaml

[Refactor] Migrate Step-Audio2 online test to deploy config (#5309)

2026-07-22 09:44:27 +00:00

step_audio_2_asr.yaml

[Refactor] Migrate Legacy Stage Configs to the Pipeline Registry (#5031)

2026-07-18 09:34:33 +00:00

step_audio_2_async_chunk.yaml

[Refactor] Migrate Legacy Stage Configs to the Pipeline Registry (#5031)

2026-07-18 09:34:33 +00:00

step_audio_2.yaml

[Refactor] Migrate Legacy Stage Configs to the Pipeline Registry (#5031)

2026-07-18 09:34:33 +00:00

voxcpm2.yaml

[Bugfix][Qwen3-TTS] Seed residual MTP sampling (#4869)

2026-07-04 17:47:11 +00:00

voxtral_tts.yaml

[Refactor] Migrate Voxtral TTS XPU stage config to --deploy-config (#5467)

2026-07-28 07:35:35 +08:00

wan2_2_ti2v.yaml

[Config] Remove diffusion_batch_size and use max_num_seqs instead. (#6484)

2026-09-01 01:02:54 +00:00