Add attention operator and adapter for onert#400

Merged

mhs4670go merged 13 commits into

Samsung:mainfrom

glistening:attention_

Nov 25, 2025

Commits on Nov 25, 2025

Fuse LlamaAttention to attention (onert)
Sanggyu Lee
committed
Update requirements.txt
Sanggyu Lee
committed
register circle_custom::attention in TICO's way
Sanggyu Lee
committed
Extract attention adapter code into adapters/onert
Sanggyu Lee
committed
Update tico/utils/validate_args_kwargs.py
Sanggyu Lee
committed
rename decode.py to model.py and support test
Sanggyu Lee
committed
Remove layer_idx
Sanggyu Lee
committed
Disable circle2circle
Sanggyu Lee
committed
Update comment why attention_mask is removed from inputs
Sanggyu Lee
committed
Add get_compile_config
Sanggyu Lee
committed
let it work in more environment
Sanggyu Lee
committed
Update requirements.txt (torch <= 2.8.0)
Sanggyu Lee
committed
Remove test
Sanggyu Lee
committed