Enable weight fake quantization in calibration mode#25
Merged
u-simha merged 3 commits intoJul 10, 2026
Merged
Conversation
crowbat
approved these changes
Jul 6, 2026
…bration mode Quantizer.calibration_mode() previously disabled fake quantization on both weights and activations. It now only disables activation fake quantization, so activation observers see the effect of quantized weights when computing activation ranges. Adds enable_weight_fake_quant/disable_activation_fake_quant helpers used by both the eager and graph quantizers, and updates docs/tests accordingly.
…hange - Move enable_weight_fake_quant/disable_activation_fake_quant into a new _fake_quant_utils module so they can import FakeQuantizeImplBase and CompressionTargetTensor at module scope instead of via local imports, keeping _utils.py free of dependencies on other quantization modules. - Use basic_config (weight + activation quantization) instead of input_activation_only_config in test_calibration_mode so the weight-FQ branch is actually exercised; only assert scale drift for activation modules, since weight ranges are fixed at prepare time. - Drop the changelog.d entry.
u-simha
force-pushed
the
u/usimha/enable-weight-fq-in-calibration-mode
branch
from
0a6ddba to
3eb12bd
Compare
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Quantizer.calibration_mode() previously disabled fake quantization on both weights and activations. It now only disables activation fake quantization, so activation observers see the effect of quantized weights when computing activation ranges.
Adds enable_weight_fake_quant/disable_activation_fake_quant helpers used by both the eager and graph quantizers, and updates docs/tests accordingly.