As part of #2752 we added a new Int8Tensor that does not use AQT in #3407. We plan to deprecate version=1 support of `Int8WeightOnlyConfig` and `Int8DynamicActivationInt8WeightConfig` and migrate users to the new workflow. version=1 support will be removed in a future torchao release
