pytorch/test/quantization
Jerry Zhang 3ef4c27ab3 Update pt2e numeric debugger to use node.meta["custom"] field (#134040)
Summary:
With https://github.com/pytorch/pytorch/pull/131912 we now have a "custom" field in node.meta that can be preserved
in

* copy/deepcopy
* run_decompositions()
* serialization
* re-exporting

So we refactored numeric debugger to use this.

Test Plan:
python test/test_quantization.py TestNumericDebugger

Reviewers:

Subscribers:

Tasks:

Tags:
Pull Request resolved: https://github.com/pytorch/pytorch/pull/134040
Approved by: https://github.com/tarun292
2024-08-27 19:51:03 +00:00
..
ao_migration
bc Fix failures when default is flipped for weights_only (#127627) 2024-08-16 00:22:43 +00:00
core Add new ops wrapped_linear_prepack and wrapped_quantized_linear_prepacked (#134232) 2024-08-23 04:54:26 +00:00
eager
fx Fix failures when default is flipped for weights_only (#127627) 2024-08-16 00:22:43 +00:00
jit
pt2e Update pt2e numeric debugger to use node.meta["custom"] field (#134040) 2024-08-27 19:51:03 +00:00
serialized
__init__.py