ViperEkura 7f0e8bb8c2 fix: let flash backend handle 4D causal prefill mask ...

- Treat 4D masks as causal (flash handles it natively), keep rejecting custom non-causal masks
- Enables flash backend in benchmark --compare and real prefill path

2026-08-08 23:51:27 +08:00

..

lib

refactor: move compiled kernel .so files into extension/lib

2026-07-31 15:36:32 +08:00

__init__.py

refactor : use factory for attention backends

2026-08-05 15:37:22 +08:00

attention_backend.py

fix: let flash backend handle 4D causal prefill mask

2026-08-08 23:51:27 +08:00

attention_ops.py

fix: add out_buf to attn_paged_decode for CUDA graph capture compatibility

2026-08-07 19:45:59 +08:00

loader.py

feat: SGLang-style paged attention kernels replace page-table path

2026-08-01 15:41:25 +08:00

rotary_backend.py

refactor: unify rotary embedding interface and update docs

2026-07-31 16:52:25 +08:00

rotary_ops.py

refactor: unify rotary embedding interface and update docs

2026-07-31 16:52:25 +08:00