docs(quant_api): fix RST warnings in FqnToConfig and Int8DynamicActivationIntxWeightConfig docstrings (#3863)

[Anai-Guo]

Resolves several Sphinx doc-build warnings tracked in #3863, in two quant_api.py docstrings.

Int8DynamicActivationIntxWeightConfig

• Fix a stray-backtick typo in the weight_granularity arg (a backtick preceded a leading space, breaking the inline-literal markup).
• Add the blank line required before (and after) the intx_packing_format sub-bullet list, which fixed Definition list ends without a blank line; unexpected unindent.

FqnToConfig

• Reflow the fqn_to_config arg body into valid RST: the (1)/(2)/(3) enumerated list and the * precedence bullet list previously had no preceding blank line and inconsistent indentation, producing Unexpected indentation and Block quote ends without a blank line; unexpected unindent. Converted to a proper numbered list + bullet list with blank-line separators and aligned continuation lines.

Wording, regex examples, and backtick style are preserved; only the RST structure/indentation changed. No code or behavior changes.

🤖 Generated with Claude Code

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/ao/4475

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• [ROCm] MI350 CI jobs will have longer queue times due to CI migration

❌ 1 Awaiting Approval, 2 New Failures, 2 Unclassified Failures

As of commit 00a8f65 with merge base 28e6aca ():

AWAITING APPROVAL - The following workflow needs approval before CI can run:

• PR Label Check (gh)

NEW FAILURES - The following jobs have failed:

• Run Regression Tests / test-nightly (CPU Nightly, linux.4xlarge, --pre torch torchvision --index-url https://download.py... / linux-job (gh)
  test/quantization/pt2e/test_x86inductor_fusion.py::DynamicShapesCppWrapperCpuTests::test_qconv2d_maxpool2d_linear_dynamic_cpu_dynamic_shapes_cpp_wrapper
• Run Regression Tests / test-nightly (CUDA Nightly, linux.g5.12xlarge.nvidia.gpu, --pre torch torchvision --index-url htt... / linux-job (gh)
  test/quantization/pt2e/test_x86inductor_fusion.py::DynamicShapesCppWrapperCpuTests::test_qconv2d_maxpool2d_linear_dynamic_cpu_dynamic_shapes_cpp_wrapper

UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:

• Run 1xH100 Tests / test (H100, mt-l-x86iamx-22-225-h100, --pre torch torchvision torchaudio mslk --index-url https:/... / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  RuntimeError: Command bash /__w/_temp/exec_script failed with exit code 1
• Run Regression Tests / test-nightly (CPU Nightly (Clang), linux.4xlarge, --pre torch torchvision --index-url https://dow... / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  test/quantization/pt2e/test_x86inductor_fusion.py::DynamicShapesCppWrapperCpuTests::test_qconv2d_maxpool2d_linear_dynamic_cpu_dynamic_shapes_cpp_wrapper

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

[andrewor14]

@claude review

[claude]

Claude finished @andrewor14's task in 1m 7s —— View job

---

Code Review

I reviewed the docstring changes in torchao/quantization/quant_api.py. This is a clean, well-scoped docs-only fix.

• Read PR diff and the affected docstrings
• Validate RST structure of both docstrings
• Check for behavior/code changes (none)

---

Verdict: LGTM ✅ — the RST fixes are correct and the warnings should be resolved.

Int8DynamicActivationIntxWeightConfig (lines 358–371)

• The stray-backtick typo on weight_granularity (` weight_granularity` → `weight_granularity`) is fixed and now aligns with the other args. ✅
• The blank lines added before and after the intx_packing_format bullet list (lines 366, 369) are exactly what RST needs to terminate the surrounding definition/field text and start/end the bullet list — this is the standard fix for "Definition list ends without a blank line; unexpected unindent." ✅

FqnToConfig (lines 1506–1528)

• The (1)/(2)/(3) block is now a proper 1./2./3. enumerated list with a preceding blank line and aligned 3-space continuation indents (e.g. start with prefix "re:" under item 2). ✅
• The precedence * bullet list now has a preceding blank line and consistent 2-space continuation indentation. ✅
• Wording, regex examples, and the re: escapes are preserved verbatim — good, this keeps the change purely structural. ✅

Minor / optional (pre-existing, not introduced here)

• These docstrings wrap inline code/identifiers in single backticks (`weight_dtype`). In RST, single backticks resolve to the default title-reference role (rendered italic), not inline literal — double backticks ( weight_dtype ) would be the literal-code rendering. Since the PR explicitly preserves backtick style and the surrounding docstrings are consistent in using single backticks, this is fine to leave as-is and out of scope for a warning-fix PR. Worth a follow-up if the team wants true monospaced rendering.
• Cosmetic: Int8DynamicActivationIntxWeightConfig uses lowercase args: while FqnToConfig uses Args:. Not related to this PR; just noting the inconsistency.

Nice, surgical fix. No code or behavior changes, and the indentation/blank-line corrections directly target the reported Sphinx warnings.
• branch fix-docstring-rst-fqntoconfig-int8intx