Add Laguna expert defusion support (#51)

* Add Laguna expert defusion support

* Document Laguna support and stabilize meta tests

Update pyproject.toml (#48)

[CI] Update CI with latest config (#41)

Update pyproject.toml (#39)

fix version check (#34)

support phimoe (#32)

* support phimoe

Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>

* Bump version from 0.0.15 to 0.0.16

Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>

---------

Signed-off-by: ZX-ModelCloud <zx@modelcloud.ai>

Refractor (#31)

* clean up readme

* defused weights should stay on same device

* cleanup

* patch replace_fused_blocks() returning after the first replacement entry instead of applying the full list

* defused tensors should remain original's contiguous state

* deduplicate logic: move shared code to helper

Add missing (#27)

* add missing

* add missing

add glm4v (#26)

* add glm4v

* up version

* cleanup

* cleanup

Update pyproject.toml (#25)