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

Signed-off-by: Bo Li <22713281+bobboli@users.noreply.github.com>
Signed-off-by: lishunyang12 <lishunyang03@gmail.com>
Signed-off-by: lishunyang12 <lishunyang12@163.com>
Co-authored-by: lishunyang12 <lishunyang03@gmail.com>
Co-authored-by: lishunyang12 <lishunyang12@163.com>
This commit is contained in:
authored and GitHub committed 2026-08-06 09:38:17 +00:00
1 parent a9ba47d8ce
commit d219d93bbb
11 files changed
+370 -33

No files matched your search

+3
View File
@@ -41,6 +41,9 @@ _DIFFUSION_MODEL_METADATA: dict[str, DiffusionModelMetadata] = {
supports_multimodal_inputs=True,
max_multimodal_image_inputs=9,
supports_mixed_reference_inputs=True,
# H3 represents alignment padding as a second packed sequence. The
# packed TRTLLM backend consumes cu_seqlens and isolates that padding.
attention_mask_free=True,
),
"WanPipeline": DiffusionModelMetadata(attention_mask_free=True),
"WanImageToVideoPipeline": DiffusionModelMetadata(attention_mask_free=True),