develop_or_install (ROCm#954)

* develop_or_install

* update

* update

fix (ROCm#903)

Optimize topksoftmax WARPS_PER_TB for higher occupancy and remove red…

…undant precision conversion (ROCm#652)

* apply clang-format

* optimize

---------

Co-authored-by: Cu Cui <cu.cui@alumni.uni-heidelberg.de>

Fix torch-compile error (ROCm#557)

Co-authored-by: root <root@dell300x-pla-u14-03.pla.dcgpu>

mxfp4_quant_shuffle (ROCm#435)

tag rc 0.1.1

First version of FAv3