分支列表

默认分支

main
pre-commit / pre-run-check (push) Has been cancelled
pre-commit / pre-commit (push) Has been cancelled

46e6a83ce1 · [Bugfix][KV Offload] Clean up resources after initialization failure (#51227) · 最后更新于 2026-08-06 11:42:13 +00:00

分支列表

c5e3454e5a · [Model] Add support for BharatGen's Param2MoE model (#38000) · 最后更新于 2026-04-06 08:19:56 +00:00    frostbyte_neo

4082
0
已包含

7c0e397592 · p · 最后更新于 2026-04-06 04:48:41 +00:00    frostbyte_neo

4093
1

3da0bdec77 · Revert "[Perf] Batch KV cache swap copies via cuMemcpyBatchAsync (#38460)" · 最后更新于 2026-04-03 18:42:04 +00:00    frostbyte_neo

4111
1

beefff2174 · Merge branch 'main' into wentao-skip-work-when-empty · 最后更新于 2026-04-03 15:31:23 +00:00    frostbyte_neo

4116
11

2a69949bda · [Bugfix]: Fix Gemma4ToolParser.__init__() missing tools parameter (#38847) · 最后更新于 2026-04-02 23:45:38 +00:00    frostbyte_neo

4280
13

db7a17ecc0 · p · 最后更新于 2026-04-02 21:51:37 +00:00    frostbyte_neo

4145
1

e92668e83c · p · 最后更新于 2026-04-02 19:08:35 +00:00    frostbyte_neo

4148
2

8f4f8425d0 · [torch.compile] Add compile-only mode · 最后更新于 2026-04-01 02:23:55 +00:00    frostbyte_neo

4212
1

509b32c77e · [release] Automate DockerHub image publishing in release pipeline · 最后更新于 2026-03-30 19:48:06 +00:00    frostbyte_neo

4256
1

a26e8dc7ff · [Bugfix][MLA] Change default SM100 MLA prefill backend back to TRT-LLM (#38562) · 最后更新于 2026-03-30 19:42:26 +00:00    frostbyte_neo

4733
24

dcbedb7661 · [Bugfix] Add NaN masking to NVFP4 quantization to prevent output contamination · 最后更新于 2026-03-29 04:07:26 +00:00    frostbyte_neo

4625
18

388e55ef49 · Merge branch 'main' into wentao-optimize-async-scheduling-copy · 最后更新于 2026-03-28 14:29:00 +00:00    frostbyte_neo

4296
12

f62527f55b · Merge branch 'main' into wentao-fix-ci-batch-invariant-issue · 最后更新于 2026-03-27 18:00:54 +00:00    frostbyte_neo

4313
4

c0be8b1f82 · [CI] Fix model swap test failures · 最后更新于 2026-03-27 03:18:50 +00:00    frostbyte_neo

4329
2

f0b888ffd3 · tests: de-duplicate online scoring API format checks · 最后更新于 2026-03-26 22:54:25 +00:00    frostbyte_neo

4331
1

257c0c5b50 · add pdl · 最后更新于 2026-03-26 16:25:35 +00:00    frostbyte_neo

4441
3

a9bf5d228b · Merge branch 'main' into vadim/qwen35-no-deppgemm · 最后更新于 2026-03-26 05:14:43 +00:00    frostbyte_neo

4363
5

58b0c78a42 · [MRV2] Support expert index capture · 最后更新于 2026-03-25 23:48:32 +00:00    frostbyte_neo

4374
1

583f60717a · quick fix for 37665 (#37722) · 最后更新于 2026-03-21 13:08:38 +00:00    frostbyte_neo

4554
2

3245cc77c3 · [Bugfix][NIXL] Skip heteroblocksize post-processing for failed recv requests in get_finished · 最后更新于 2026-03-20 21:45:08 +00:00    frostbyte_neo

4780
1