🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4150

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

⏳ No Failures, 10 Pending

As of commit 04f0ca4 with merge base 1b2682d ():
💚 Looks good so far! There are no failures yet. 💚

This comment was automatically generated by Dr. CI and updates every 15 minutes.

@claude review

Claude finished @jerryzh168's task in 1m 27s —— View job

Code Review: Delete deprecated PackedLinearInt8DynamicActivationIntxWeightLayout

• Review the diff and understand changes
• Check for remaining references to deleted code
• Provide review feedback

This is a clean deletion PR. The changes are straightforward and correct.

What's being removed:

1. The entire packed_linear_int8_dynamic_activation_intx_weight_layout.py file (474 lines) — including the PackedLinearInt8DynamicActivationIntxWeightLayout class, its AQTTensorImpl subclass, linear dispatch check/impl functions, and the Target enum
2. Imports and dispatch registrations in affine_quantized_tensor_ops.py
3. Re-exports in torchao/dtypes/__init__.py and torchao/dtypes/uintx/__init__.py
4. Target import and safe-global registration in quant_api.py

Verification:

• No remaining references to PackedLinearInt8DynamicActivationIntxWeight* anywhere in the codebase
• No remaining references to _linear_int8_act_intx_weight_packed_* functions
• No remaining references to make_packed_linear_int8_dynamic_activation_intx_weight_tensor
• The Target enum is fully removed (no remaining uses in quant_api.py)
• The deleted code already had deprecation warnings pointing to issues #2752 and #2967

Minor note: The __all__ in torchao/dtypes/__init__.py previously included two stale entries — to_affine_quantized_packed_linear_int8_dynamic_activation_intx_weight and Int4GroupwisePreshuffleTensor — that were never actually imported. This PR correctly removes them as well.

LGTM — no issues found.