Add qk norm optionally before attention calculation
Differential Revision: D70355802 Pull Request resolved: https://github.com/pytorch/executorch/pull/8820
M
madhu-fb committed
352416e64db0379d60ed3aa58e2cbf5745e2d749
Parent: bdafb22
Committed by GitHub <noreply@github.com>
on 3/6/2025, 10:31:26 PM