[3/N] Support for save/restoring AutoQuantize sensitivity scores - #588
Merged
realAsma merged 1 commit intoNov 25, 2025
Conversation
Codecov Report❌ Patch coverage is
Additional details and impacted files@@ Coverage Diff @@
## main #588 +/- ##
==========================================
+ Coverage 74.45% 74.65% +0.19%
==========================================
Files 182 182
Lines 18250 18453 +203
==========================================
+ Hits 13588 13776 +188
- Misses 4662 4677 +15 ☔ View full report in Codecov by Sentry. 🚀 New features to boost your workflow:
|
realAsma
changed the base branch from
main
to
asma/auto_quantize_kd_loss_sensitivity
November 20, 2025 23:19
realAsma
changed the title
realAsma
requested review from
Fridah-nv,
ajrasane,
cjluo-nv,
kinjalpatel27 and
mxinO
and removed request for
ChenhanYu and
cjluo-nv
November 20, 2025 23:20
realAsma
force-pushed
the
asma/auto_quantize_user_improvements
branch
from
November 20, 2025 23:25
c96d919 to
a489c5d
Compare
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 20, 2025 23:39
197d4d6 to
9134ca9
Compare
realAsma
force-pushed
the
asma/auto_quantize_user_improvements
branch
from
November 20, 2025 23:41
a489c5d to
35be9b2
Compare
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 21, 2025 00:33
dc15dae to
48b0423
Compare
realAsma
force-pushed
the
asma/auto_quantize_user_improvements
branch
from
November 21, 2025 00:39
35be9b2 to
63a034f
Compare
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 21, 2025 21:22
48b0423 to
73fc080
Compare
realAsma
force-pushed
the
asma/auto_quantize_user_improvements
branch
from
November 21, 2025 21:47
d169f2d to
c0dd4cf
Compare
meenchen
reviewed
Nov 24, 2025
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 25, 2025 20:24
46670e1 to
09d8a29
Compare
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 25, 2025 22:00
09d8a29 to
75f83da
Compare
meenchen
approved these changes
Nov 25, 2025
realAsma
force-pushed
the
asma/auto_quantize_kd_loss_sensitivity
branch
from
November 25, 2025 23:18
75f83da to
1b52477
Compare
Some improvements for KLDiv Signed-off-by: realAsma <akuriparambi@nvidia.com> changelog update Signed-off-by: realAsma <akuriparambi@nvidia.com> minor Signed-off-by: realAsma <akuriparambi@nvidia.com> doc updates Signed-off-by: realAsma <akuriparambi@nvidia.com>
realAsma
force-pushed
the
asma/auto_quantize_user_improvements
branch
from
November 25, 2025 23:30
c0dd4cf to
4853606
Compare
realAsma
merged this pull request
into
asma/auto_quantize_kd_loss_sensitivity
Nov 25, 2025
1 check passed
realAsma
added a commit
that referenced
this pull request
Nov 26, 2025
…AutoQuantizeGradientSearcher; seperated quant modules and score modules (#586) ## What does this PR do? **Type of change:** Refator; Minor new feature **Overview:** ? 1. Refactored AutoQuantizeSearcher to _AutoQuantizeBaseSearcher & AutoQuantizeGradientSearcher - Prepares architecture for additional search methods. 2. seperated quant modules and score modules - separate quantization modules from scoring modules, enabling auto-quantization to measure sensitivity at parent layers (e.g., MLP output for MoE experts) rather than individual ops. 3. Also see #592 and #588 ## Testing See unittests; `tests/unit/torch/quantization/test_autoquant.py` and `tests/unit/torch/quantization/plugins/test_huggingface.py` ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes - **Did you write any new necessary tests?**: Yes - **Did you add or update any necessary documentation?**: Yes - **Did you update [Changelog](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CHANGELOG.rst)?**: Not Required ## Additional Information <!-- E.g. related issue. --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added support for score modules in quantization workflows. * Added optional naming for quantization recipes. * **Bug Fixes** * Improved quantization grouping rules documentation with clearer configuration examples. * **Refactor** * Renamed quantization module parameters for improved clarity. * Enhanced quantization search architecture for better scalability. <sub>✏️ Tip: You can customize this high-level summary in your review settings.</sub> <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: realAsma <akuriparambi@nvidia.com> Co-authored-by: Asma Kuriparambil Thekkumpate <akuriparambi@akuriparambi-mlt.client.nvidia.com>
inisis
pushed a commit
to inisis/TensorRT-Model-Optimizer
that referenced
this pull request
Nov 26, 2025
…AutoQuantizeGradientSearcher; seperated quant modules and score modules (NVIDIA#586) ## What does this PR do? **Type of change:** Refator; Minor new feature **Overview:** ? 1. Refactored AutoQuantizeSearcher to _AutoQuantizeBaseSearcher & AutoQuantizeGradientSearcher - Prepares architecture for additional search methods. 2. seperated quant modules and score modules - separate quantization modules from scoring modules, enabling auto-quantization to measure sensitivity at parent layers (e.g., MLP output for MoE experts) rather than individual ops. 3. Also see NVIDIA#592 and NVIDIA#588 ## Testing See unittests; `tests/unit/torch/quantization/test_autoquant.py` and `tests/unit/torch/quantization/plugins/test_huggingface.py` ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes - **Did you write any new necessary tests?**: Yes - **Did you add or update any necessary documentation?**: Yes - **Did you update [Changelog](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CHANGELOG.rst)?**: Not Required ## Additional Information <!-- E.g. related issue. --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added support for score modules in quantization workflows. * Added optional naming for quantization recipes. * **Bug Fixes** * Improved quantization grouping rules documentation with clearer configuration examples. * **Refactor** * Renamed quantization module parameters for improved clarity. * Enhanced quantization search architecture for better scalability. <sub>✏️ Tip: You can customize this high-level summary in your review settings.</sub> <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: realAsma <akuriparambi@nvidia.com> Co-authored-by: Asma Kuriparambil Thekkumpate <akuriparambi@akuriparambi-mlt.client.nvidia.com> Signed-off-by: inisis <desmond.yao@buaa.edu.cn>
jQizhang
pushed a commit
to jQizhang/TensorRT-Model-Optimizer
that referenced
this pull request
Nov 26, 2025
…AutoQuantizeGradientSearcher; seperated quant modules and score modules (NVIDIA#586) ## What does this PR do? **Type of change:** Refator; Minor new feature **Overview:** ? 1. Refactored AutoQuantizeSearcher to _AutoQuantizeBaseSearcher & AutoQuantizeGradientSearcher - Prepares architecture for additional search methods. 2. seperated quant modules and score modules - separate quantization modules from scoring modules, enabling auto-quantization to measure sensitivity at parent layers (e.g., MLP output for MoE experts) rather than individual ops. 3. Also see NVIDIA#592 and NVIDIA#588 ## Testing See unittests; `tests/unit/torch/quantization/test_autoquant.py` and `tests/unit/torch/quantization/plugins/test_huggingface.py` ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes - **Did you write any new necessary tests?**: Yes - **Did you add or update any necessary documentation?**: Yes - **Did you update [Changelog](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CHANGELOG.rst)?**: Not Required ## Additional Information <!-- E.g. related issue. --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added support for score modules in quantization workflows. * Added optional naming for quantization recipes. * **Bug Fixes** * Improved quantization grouping rules documentation with clearer configuration examples. * **Refactor** * Renamed quantization module parameters for improved clarity. * Enhanced quantization search architecture for better scalability. <sub>✏️ Tip: You can customize this high-level summary in your review settings.</sub> <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: realAsma <akuriparambi@nvidia.com> Co-authored-by: Asma Kuriparambil Thekkumpate <akuriparambi@akuriparambi-mlt.client.nvidia.com>
HexCore8
pushed a commit
to HexCore8/Model-Optimizer
that referenced
this pull request
Apr 7, 2026
…AutoQuantizeGradientSearcher; seperated quant modules and score modules (#586) ## What does this PR do? **Type of change:** Refator; Minor new feature **Overview:** ? 1. Refactored AutoQuantizeSearcher to _AutoQuantizeBaseSearcher & AutoQuantizeGradientSearcher - Prepares architecture for additional search methods. 2. seperated quant modules and score modules - separate quantization modules from scoring modules, enabling auto-quantization to measure sensitivity at parent layers (e.g., MLP output for MoE experts) rather than individual ops. 3. Also see NVIDIA/Model-Optimizer#592 and NVIDIA/Model-Optimizer#588 ## Testing See unittests; `tests/unit/torch/quantization/test_autoquant.py` and `tests/unit/torch/quantization/plugins/test_huggingface.py` ## Before your PR is "*Ready for review*" <!-- If you haven't finished some of the above items you can still open `Draft` PR. --> - **Make sure you read and follow [Contributor guidelines](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CONTRIBUTING.md)** and your commits are signed. - **Is this change backward compatible?**: Yes - **Did you write any new necessary tests?**: Yes - **Did you add or update any necessary documentation?**: Yes - **Did you update [Changelog](https://github.com/NVIDIA/TensorRT-Model-Optimizer/blob/main/CHANGELOG.rst)?**: Not Required ## Additional Information <!-- E.g. related issue. --> <!-- This is an auto-generated comment: release notes by coderabbit.ai --> ## Summary by CodeRabbit * **New Features** * Added support for score modules in quantization workflows. * Added optional naming for quantization recipes. * **Bug Fixes** * Improved quantization grouping rules documentation with clearer configuration examples. * **Refactor** * Renamed quantization module parameters for improved clarity. * Enhanced quantization search architecture for better scalability. <sub>✏️ Tip: You can customize this high-level summary in your review settings.</sub> <!-- end of auto-generated comment: release notes by coderabbit.ai --> --------- Signed-off-by: realAsma <akuriparambi@nvidia.com> Co-authored-by: Asma Kuriparambil Thekkumpate <akuriparambi@akuriparambi-mlt.client.nvidia.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
2 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.
What does this PR do?
Type of change: ?
Overview: ?
Usage
# Add a code snippet demonstrating how to use thisTesting
Before your PR is "Ready for review"
Additional Information