build: enable PyTorch ops by default

[njzjz-bot]

Summary

• Change the Python build backend so DP_ENABLE_PYTORCH defaults to 1.
• Update the source-installation docs and related PyTorch OP guidance to describe the new default and the DP_ENABLE_PYTORCH=0 opt-out.
• Adjust the PyTorch runtime rebuild hint now that source builds enable PyTorch C++ OPs by default.

Tests

• python3 static assertions for the updated default and docs
• python3 mocked backend.read_env.get_argument_from_env() check for default -DENABLE_PYTORCH=ON and explicit DP_ENABLE_PYTORCH=0 opt-out
• python3 backend.find_pytorch() opt-out check
• git diff --check

Authored by OpenClaw (model: custom-chat-jinzhezeng-group/gpt-5.5)

Summary by CodeRabbit

Release Notes

• New Features

  • PyTorch customized C++ operations now enabled by default during source installation.

• Documentation

  • Updated installation guides to reflect new default PyTorch behavior.
  • Users can disable PyTorch customized operations via environment variable if needed.
  • Updated error messages with revised installation guidance.

[coderabbitai]

No actionable comments were generated in the recent review. 🎉

ℹ️ Recent review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 0fbd651c-a535-483e-9983-2ff324591f0d

📥 Commits

Reviewing files that changed from the base of the PR and between 95fa690 and 9292367.

📒 Files selected for processing (7)

• backend/find_pytorch.py
• backend/read_env.py
• deepmd/pt/cxx_op.py
• doc/freeze/compress.md
• doc/install/install-from-source.md
• doc/model/dpa2.md
• doc/model/dpa3.md

---

📝 Walkthrough

Walkthrough

DP_ENABLE_PYTORCH default value is flipped from "0" to "1" in find_pytorch.py and read_env.py, making PyTorch C++ OP discovery and CMake flag generation enabled by default when the variable is unset. The version-mismatch error message in cxx_op.py and multiple documentation files are updated to match.

Changes

DP_ENABLE_PYTORCH enabled by default

Layer / File(s)	Summary
Backend default flip and error message update backend/find_pytorch.py, backend/read_env.py, deepmd/pt/cxx_op.py	find_pytorch() and get_argument_from_env() change the DP_ENABLE_PYTORCH fallback default from "0" to "1", enabling PyTorch when the variable is unset. The load_library() version-mismatch RuntimeError message replaces DP_ENABLE_PYTORCH=1 pip install ... guidance with pip install deepmd-kit --no-binary deepmd-kit.
Documentation aligned to new default doc/install/install-from-source.md, doc/freeze/compress.md, doc/model/dpa2.md, doc/model/dpa3.md	All installation and model docs updated to state customized C++ OPs are enabled by default, instruct users to set DP_ENABLE_PYTORCH=0 to disable them, and change the documented default from 0 to 1.

Estimated code review effort

🎯 1 (Trivial) | ⏱️ ~3 minutes

Suggested labels

Docs

Suggested reviewers

• wanghan-iapcm

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Description Check	✅ Passed	Check skipped - CodeRabbit’s high-level summary is enabled.
Title check	✅ Passed	The title 'build: enable PyTorch ops by default' accurately and concisely summarizes the main change in the pull request: modifying the build system to enable PyTorch operations by default.
Docstring Coverage	✅ Passed	Docstring coverage is 100.00% which is sufficient. The required threshold is 80.00%.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

_{✏️ Tip: You can configure your own custom pre-merge checks in the settings.}

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 82.14%. Comparing base (95fa690) to head (9292367).

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5577      +/-   ##
==========================================
- Coverage   82.15%   82.14%   -0.01%     
==========================================
  Files         900      900              
  Lines      104140   104141       +1     
  Branches     4470     4470              
==========================================
- Hits        85553    85546       -7     
- Misses      17181    17186       +5     
- Partials     1406     1409       +3     

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.