Removed support for non-per-tensor quantized relu#14788

Open

DrJessop wants to merge 4 commits intopytorch:mainfrom

DrJessop:export-D83874866

Commits on Oct 6, 2025

Fixed assumption on out_shift for quantized linear (#14789)

Andrew Grebenisan
authored and
facebook-github-bot
committed
Replace quantized conv and relu non-tensor variants with per tensor variants

Andrew Grebenisan
authored and
facebook-github-bot
committed
Fix avg_pool2d replace ops pass

Andrew Grebenisan
authored and
facebook-github-bot
committed
Removed support for non-per-tensor quantized relu (#14788)

Andrew Grebenisan
authored and
facebook-github-bot
committed