zhrrr 1edfd09ffd [Model Runner V2] Use actual batch max_seq_len for attn metadata (#43991) ...

Signed-off-by: zhuhaoran <zhuhaoran.zhr@alibaba-inc.com>

2026-06-02 06:07:56 +00:00

..

eagle

[Model Runner V2] Use actual batch max_seq_len for attn metadata (#43991)

2026-06-02 06:07:56 +00:00

__init__.py

[Model Runner V2] Support Eagle3 (no CUDA graph) (#35029)

2026-02-21 12:55:24 -08:00

rejection_sampler_utils.py

[Model Runner V2] FP32 gumbel sampling. (#41775)

2026-05-15 09:20:08 -07:00

rejection_sampler.py

[Model Runner V2] FP32 gumbel sampling. (#41775)

2026-05-15 09:20:08 -07:00

utils.py

[Core] Don't schedule spec tokens with prefill chunks (#33652)

2026-02-04 23:40:22 +00:00