ooooooye 740f45abcd [Perf][Diffusion] MiniMax-H3: opt into packed varlen attention on NPU to eliminate quadratic mask materialization (#5891) ...

Signed-off-by: brandneway <gyuan4892@gmail.com>

2026-08-10 19:47:03 +08:00

..

backends

[Perf][Diffusion] MiniMax-H3: opt into packed varlen attention on NPU to eliminate quadratic mask materialization (#5891)

2026-08-10 19:47:03 +08:00

parallel

[Bugfix] Support GQA/MQA in the ring attention SDPA path (#5255)

2026-08-07 04:08:33 +00:00

__init__.py

[Diffusion] Init Attention Backends and Selector for Diffusion (#115)

2025-12-05 10:46:20 +00:00

layer.py

[Attention] Refine TRTLLM attention support for MiniMax H3 (#5779)

2026-08-06 09:38:17 +00:00

selector.py

[Attention] Add trtllm diffusion attention backend with Skip-Softmax (#5283)

2026-07-27 22:48:11 +08:00