Default Branch

4765f0f189 · [Bugfix] Fix sequence_parallel_chunk_impl custom op aliasing its input (#44130) · Updated 2026-06-05 23:56:36 +00:00

Branches

ee83fe043f · [CI] Migrate all remaining gpu_1_queue jobs to h200_18gb MIG · Updated 2026-05-12 18:58:58 +00:00    obscura

798
1

593d5a4033 · [Bugfix] Fix mismatched kernel-per-logical blocks in NIXL HMA transfer (#42097) · Updated 2026-05-12 13:53:30 +00:00    obscura

787
0
Included

d841c47290 · [CI] Migrate all gpu_1_queue jobs to h200_18gb MIG · Updated 2026-05-12 09:14:24 +00:00    obscura

798
1

afcf797f6e · [Attention][TokenSpeed MLA] Fix FLASHINFER MLA prefill test failures · Updated 2026-05-12 05:29:54 +00:00    obscura

842
17

b4dbbc7102 · Fix dtype mismatch in topk_softplus_sqrt for DeepEP backends · Updated 2026-05-11 19:24:14 +00:00    obscura

1122
18

1b2570ec77 · Add tasks/version/logging_utils/beam_search to basic_correctness deps · Updated 2026-05-10 10:50:16 +00:00    obscura

854
3

7c4689e477 · Add tasks/version/logging_utils to models_language deps · Updated 2026-05-10 10:49:50 +00:00    obscura

854
3

dae05af771 · Add tasks/version/logging_utils to models_multimodal deps · Updated 2026-05-10 10:49:06 +00:00    obscura

854
3

74690b24e7 · Add tasks/version/logging_utils to models_basic GPU deps · Updated 2026-05-10 10:48:11 +00:00    obscura

854
3

8798dacaab · Add tasks/version/beam_search/logging_utils to entrypoints deps · Updated 2026-05-10 10:39:32 +00:00    obscura

854
3

3034094787 · Merge branch 'main' into wentao-model-runner-v2-support-stock-torch-compile · Updated 2026-05-09 21:47:45 +00:00    obscura

873
6

bc150f5029 · [CI] Automate Docker Hub release image publishing (#40415) · Updated 2026-05-08 01:25:36 +00:00    obscura

1222
39

c3f8936aca · Merge branch 'main' into wentao-optimize-dcp-and-add-comm-func · Updated 2026-05-06 20:05:46 +00:00    obscura

967
17

58c8a5eaa5 · [Attention][TokenSpeed MLA] Also warm up prefill kernel from decode impl · Updated 2026-05-06 07:45:16 +00:00    obscura

994
10

9b4e83934d · [Spec Decode] Add Gemma4 MTP speculative decoding with centroids masking · Updated 2026-05-05 15:59:11 +00:00    obscura

1003
1

ff771a0f53 · Test nit · Updated 2026-05-05 14:14:37 +00:00    obscura

1006
3

4fda385093 · [ZenCPU] Update device to "zen5" · Updated 2026-05-05 12:03:42 +00:00    obscura

1009
4

472456c2f9 · Update release-pipeline.yaml · Updated 2026-05-05 05:02:36 +00:00    obscura

1102
2

b870c8edb4 · optimize allpool forward · Updated 2026-05-04 22:47:43 +00:00    obscura

1029
1

132765e356 · Revert "[DSv4] Use cvt PTX for FP32->FP4 conversion (#41015)" · Updated 2026-05-04 08:56:49 +00:00    obscura

1222
33