0 0 4 Python

[Fix] Fix multi-head latent attention (MLA) (#47761)

* core

* tests

* Fix docstring

* Fix te

* Review compliance: if/else block

* Nit on AXK2

* Fix cfg

* Fix for ONNX

Rémi Ouazan committed 1mo ago

bf017e36f130188837bfd97d153103e430924886

Parent: 838763b

Committed by GitHub <noreply@github.com> on 8/6/2026, 12:42:10 AM