SIGN IN SIGN UP

Add int8 per token dynamic activaiton quant and int4 weight quant for llama2 in executorch (#1904)

Summary:
Pull Request resolved: https://github.com/pytorch/executorch/pull/1904

representation we are getting now: https://www.internalfb.com/intern/everpaste/?handle=GEIHRRnpyYOEAIUBAFtHZapvTH5xbsIXAAAB

Reviewed By: kimishpatel

Differential Revision: D53211239

fbshipit-source-id: 255f87e44079877fa70afe65fa6f0c512f06d213
J
Jerry Zhang committed
a9439586f09ff30f6605edf418973ea60cd21e87
Parent: 5d4d0ca
Committed by Facebook GitHub Bot <facebook-github-bot@users.noreply.github.com> on 2/10/2024, 4:08:15 AM