0 0 1 Python

Implement weight_int8packed_mm (#3893)

Summary:
Pull Request resolved: https://github.com/pytorch/executorch/pull/3893

## Context

As title, this changesets implement the `aten._weight_int8packed_mm` operator. The operator implements a linear layer where the weight is quantized symmetrically to 8 bits for each "group".
ghstack-source-id: 229569180

Reviewed By: copyrightly

Differential Revision: D58263387

fbshipit-source-id: a32362ae912a4ad72bd597d8757528180fa3186b

Stephen Jia committed 2y ago

a3ff00d3224633889b657e926ffd237324153e16

Parent: be7150c

Committed by Facebook GitHub Bot <facebook-github-bot@users.noreply.github.com> on 6/10/2024, 8:17:06 PM