Support int4 weight in quantized matmul/linear#7235
Merged
Conversation
added 2 commits
June 10, 2024 22:45
4330117 to
03f46f1
Compare
added 2 commits
June 10, 2024 23:03
added 2 commits
June 10, 2024 23:06
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
JackCaoG
reviewed
Jun 10, 2024
Collaborator
Author
|
Removed |
JackCaoG
approved these changes
Jun 10, 2024
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
3 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
int4 weight can be enabled by
torch.ops.xla.quantized_matmul(x, weight, weight_scaler, int4_weight=True),XlaQuantizedLinear(...,int4_weight=True)The matmul w/ int4 workflow is:
xla::Literalwill be created for the int4 weightsF.linearon the activation and int4 weightOriginal plan was to pack int4 values in int8 container, and do reinterpret cast, but reinterpret cast does't work on TPU now.
Test:
Added tests for quantized op and linear module.