3 people b443c1cc4e [Perf][GLM-5.3-Flash] Use FlashKDA for KDA chunked prefill (1.7-3.8x faster than the Triton chunk path) (#55737) ...

Signed-off-by: Jared Wen <jaredwen@inferact.ai>
Co-authored-by: Claude Fable 5.1 <noreply@anthropic.com>
Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>

2026-09-14 08:36:19 +00:00

..

ops

[Perf][GLM-5.3-Flash] Decode hot-path cleanups: strided KDA recurrent inputs, NoPE MQA query without concat, no duplicate router GEMM (#55736)

2026-09-10 16:28:33 +08:00

__init__.py

[Bugfix][Docs] Package glm5next nvidia subtree and fix its docstrings (#55214)

2026-09-03 21:25:26 -07:00

attention.py

…

kda.py

[Perf][GLM-5.3-Flash] Use FlashKDA for KDA chunked prefill (1.7-3.8x faster than the Triton chunk path) (#55737)

2026-09-14 08:36:19 +00:00

model.py

[9/N][warmup][DSv4] Migrate MHC TileLang kernels (#50178)

2026-09-12 17:25:23 -07:00

mtp.py

[Feat] Add EPLB support for GLM-5.3-Flash (#55119)

2026-09-05 05:28:33 +00:00

multimodal.py

…