main
[MODEL] Add Falcon H1 (#38249)
Refactor all linear attention models to latest best standards for convolution (#47452)