FastDeploy

【Fix Bug】修复 fa3 支持集中式bug (#3235)

gh-pages.yml #58:Commit 3a15e0c53e pushed by apps

develop

2025-08-07 04:04:56 +08:00

0s

Fix the confused enable_early_stop when only set early_stop_config (#3214)

gh-pages.yml #57:Commit 36dc73470d pushed by apps

develop

2025-08-06 19:54:49 +08:00

0s

[Trace]add trace when fd start (#3174)

gh-pages.yml #56:Commit 841e831575 pushed by apps

develop

2025-08-06 11:45:38 +08:00

0s

fix approve ci (#3212)

gh-pages.yml #55:Commit e0bbd3b6ca pushed by apps

develop

2025-08-06 03:35:01 +08:00

0s

[Bug Fix]Fix bug of append attention test case (#3202)

gh-pages.yml #54:Commit c593e1a39c pushed by apps

develop

2025-08-05 19:24:54 +08:00

0s

[FIX]fix bad_words when sending requests consecutively (#3197)

gh-pages.yml #53:Commit 72ef5a9c93 pushed by apps

develop

2025-08-05 11:14:56 +08:00

0s

modified dockerfile (#3026)

gh-pages.yml #52:Commit 1d29dd80f7 pushed by apps

develop

2025-07-27 03:17:44 +08:00

0s

[Feat] support mixed ep (#2969)

gh-pages.yml #51:Commit 0700c90caa pushed by apps

develop

2025-07-26 01:20:48 +08:00

0s

[Feature] Support_eplb (#2997)

gh-pages.yml #50:Commit 2970b00dfa pushed by apps

develop

2025-07-25 17:10:47 +08:00

0s

Update benchmark tools (#3004)

gh-pages.yml #49:Commit 5151bc92c8 pushed by apps

develop

2025-07-25 00:50:46 +08:00

0s

update (#2978)

gh-pages.yml #48:Commit bfeb664ab8 pushed by apps

develop

2025-07-24 16:40:45 +08:00

0s

fix mtp bug in pd-split mode (#2970)

gh-pages.yml #47:Commit a39a67334c pushed by apps

develop

2025-07-24 00:20:47 +08:00

0s

[SOT] Add sot warmup (NVIDIA GPU Only) (#2929)

gh-pages.yml #46:Commit 95b5af24db pushed by apps

develop

2025-07-23 16:10:46 +08:00

0s

[SOT] Mark dynamic dims by type annotations (#2771)

gh-pages.yml #45:Commit 48e6a0ca26 pushed by apps

develop

2025-07-22 23:50:48 +08:00

0s

custom all reduce support cuda graph (#2938)

gh-pages.yml #44:Commit 0262ef7eb3 pushed by apps

develop

2025-07-22 15:40:46 +08:00

0s

[Feature] support min_p_sampling (#2872)

gh-pages.yml #43:Commit 67990e0572 pushed by apps

develop

2025-07-21 23:20:47 +08:00

0s

remove cum_offsets from ForwardMeta (#2925)

gh-pages.yml #42:Commit 8c5407d9e4 pushed by apps

develop

2025-07-21 09:30:38 +08:00

0s

update ci cases

gh-pages.yml #41:Commit b8676d71a8 pushed by apps

develop

2025-07-20 03:30:38 +08:00

0s

[Trace]fix opentelemetry can not work in uvicorn (#2906)

gh-pages.yml #40:Commit e679567d59 pushed by apps

develop

2025-07-18 21:50:45 +08:00

0s

[LLM] delete fixed slots (#2893)

gh-pages.yml #39:Commit 4b14dca1d6 pushed by apps

develop

2025-07-18 05:30:44 +08:00

0s

[Feature] support prompt repetition_penalty (#2806)

gh-pages.yml #38:Commit 1f15ca21e4 pushed by apps

develop

2025-07-17 21:20:44 +08:00

0s

[LLM] Update Multinode Deployment (#2830)

gh-pages.yml #37:Commit 9c25dcca0b pushed by apps

develop

2025-07-17 13:10:44 +08:00

0s

[Attention] remove cum_offsets from atten, and use cu_seqlens_q (#2870)

gh-pages.yml #36:Commit aa76085d1f pushed by apps

develop

2025-07-17 05:00:44 +08:00

0s

[Fix] Fix mm ep weight init. (#2855)

gh-pages.yml #35:Commit 0d0340392f pushed by apps

develop

2025-07-16 20:50:43 +08:00

0s

refactor rl get_name_mappings_to_training (#2847)

gh-pages.yml #34:Commit 61b3997b85 pushed by apps

develop

2025-07-16 12:40:43 +08:00

0s

[Docs] add enable_logprob parameter description (#2850)

gh-pages.yml #33:Commit 5fc659b900 pushed by apps

develop

2025-07-16 04:30:43 +08:00

0s

[vl] Use top_k from config.json (#2831)

gh-pages.yml #32:Commit 15c8c240b5 pushed by apps

develop

2025-07-15 20:20:42 +08:00

0s

[MTP] optimize mtp infer speed (#2840)

gh-pages.yml #31:Commit 7cdd8d290d pushed by apps

develop

2025-07-15 04:03:54 +08:00

0s

[vl]remove duplicated load logic (#2744)

gh-pages.yml #30:Commit bad53c6b6e pushed by apps

develop

2025-07-14 16:33:09 +08:00

0s

add result save for ci (#2824)

gh-pages.yml #29:Commit 16940822a7 pushed by apps

develop

2025-07-13 11:13:18 +08:00

0s