fix(quantization): clear unused dense weights in graph mode finalized model - #101
Merged
pkmandke merged 3 commits intoSep 18, 2026
Merged
Conversation
Signed-off-by: Prathamesh Mandke <46148373+pkmandke@users.noreply.github.com>
pkmandke
commented
Sep 16, 2026
| raise FileExistsError(f"mmap_dir {mmap_dir!r} is non-empty. Pass an empty directory.") | ||
|
|
||
|
|
||
| def _zero_size_placeholder(tensor: torch.Tensor) -> torch.Tensor: |
Member
Author
There was a problem hiding this comment.
refactored here so both eager and graph can use it
u-simha
approved these changes
Sep 17, 2026
crowbat
approved these changes
Sep 18, 2026
Signed-off-by: Prathamesh Mandke <46148373+pkmandke@users.noreply.github.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Similar to eager mode, clear the unused dense weights from the finalized model in graph mode.
Details:
_process_mlir_weight_quantizationregisters the quantized weight, scale and offset as new buffers, but the dense float parameter stays registered on its owning module. It then survives intostate_dict().