fix: use `Int4WeightOnlyConfig.group_size` in fake quant configs (#4518)
fix: use configurable `group_size` in `Int4WeightOnlyConfig` QAT
O
Oriol Liñan committed
768fa14ddbc0f82dfeb33053d38c540353026015
Parent: 04084f2
Committed by GitHub <noreply@github.com>
on 7/7/2026, 11:25:51 AM