feat(lammps): load installed backends dynamically

[njzjz]

Summary

• avoid importing TensorFlow and PyTorch unconditionally from deepmd.lmp
• add runtime library paths only for installed TensorFlow and/or PyTorch backends, while preserving CUDA preload and TF<2.12 libpython preload behavior
• update the pip install docs and add focused tests for missing-backend handling

Tests

• PYTHONPATH=$PWD pytest source/tests/test_lmp.py -v
• ruff check .
• ruff format . / commit hook ruff format
• git diff --check
• fork CI on njzjz/deepmd-kit branch feat/lammps-dynamic-backends: Build C library, Build C++, Build/upload to PyPI, CodeQL, Test Python, and Test C++ all passed; Mirror skipped

Fork CI runs:

• https://github.com/njzjz/deepmd-kit/actions/runs/28551976035
• https://github.com/njzjz/deepmd-kit/actions/runs/28551976036
• https://github.com/njzjz/deepmd-kit/actions/runs/28551976076
• https://github.com/njzjz/deepmd-kit/actions/runs/28551976053
• https://github.com/njzjz/deepmd-kit/actions/runs/28551976048
• https://github.com/njzjz/deepmd-kit/actions/runs/28551976060

Summary by CodeRabbit

• New Features
  • Improved automatic runtime detection for LAMMPS so it can load available TensorFlow and/or PyTorch libraries more reliably at startup.
• Bug Fixes
  • Better handles systems where one or more backends are missing, avoiding failed or incomplete environment setup.
  • Improves compatibility with older TensorFlow installs by finding the needed Python library automatically.
• Documentation
  • Clarified installation guidance for LAMMPS and i-PI extras and updated backend requirements wording.

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[coderabbitai]

📝 Walkthrough

Walkthrough

The PR refactors deepmd/lmp.py to move TensorFlow/PyTorch/CUDA library path discovery and LAMMPS environment variable setup into dedicated helper functions (_get_tensorflow_library_paths(), _get_pytorch_library_paths(), _configure_lammps_environment()), adds corresponding tests, and updates installation documentation.

Changes

LAMMPS environment refactor

Layer / File(s)	Summary
Import cleanup and library path helper functions deepmd/lmp.py	Removes top-level TF/PyTorch imports, adds Version/SHARED_LIB_DIR imports, and introduces helper functions that conditionally import TensorFlow/PyTorch and return their library/preload paths, returning empty results when a backend is unavailable.
Environment configuration function and execution deepmd/lmp.py	Replaces inline setup with _configure_lammps_environment(), which aggregates CUDA/TF/PyTorch paths, sets preload_env/lib_env via get_env, defines op_dir, and runs at import time.
Tests and documentation updates source/tests/test_lmp.py, doc/install/easy-install.md	Adds platform-gated tests covering both helper functions and _configure_lammps_environment() under missing/present backend scenarios; updates docs to state LAMMPS dynamically loads installed TensorFlow/PyTorch (and mentions JAX) rather than requiring both.

Estimated code review effort: 3 (Moderate) | ~25 minutes

Sequence Diagram(s)

sequenceDiagram
  participant lmp as deepmd/lmp.py (import time)
  participant tfHelper as _get_tensorflow_library_paths
  participant ptHelper as _get_pytorch_library_paths
  participant env as os.environ

  lmp->>tfHelper: get TF lib/preload paths
  tfHelper-->>lmp: paths or empty (if TF missing)
  lmp->>ptHelper: get PyTorch lib path
  ptHelper-->>lmp: path or empty (if torch missing)
  lmp->>lmp: compute CUDA paths (Linux)
  lmp->>env: set preload_env via get_env
  lmp->>env: set lib_env via get_env

Loading

Suggested labels: LAMMPS

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 accurately summarizes the main change: LAMMPS now dynamically loads installed TensorFlow and PyTorch backends.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
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.

✨ 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.}

[coderabbitai]

🧹 Nitpick comments (1)

source/tests/test_lmp.py (1)

33-54: 📐 Maintainability & Code Quality | 🔵 Trivial | ⚡ Quick win

Consider adding a test for the TF ≥ 2.12 branch.

Current coverage only exercises the TF_VERSION < 2.12 path (libpython preload). Add a complementary test with TF_VERSION="2.12.0" (or later) asserting the preload list stays empty, to guard the version-gated branch in _get_tensorflow_library_paths against regressions.

🤖 Prompt for AI Agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

In `@source/tests/test_lmp.py` around lines 33 - 54, Add a complementary test for
the TF_VERSION >= 2.12 branch in _get_tensorflow_library_paths, similar to
test_tensorflow_library_paths_include_libpython_for_old_tensorflow. Mock
deepmd.tf.env with TF_VERSION set to 2.12.0 or later and assert the returned
TensorFlow library paths are still correct while the libpython preload list is
empty, ensuring the version-gated logic does not regress.

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Nitpick comments:
In `@source/tests/test_lmp.py`:
- Around line 33-54: Add a complementary test for the TF_VERSION >= 2.12 branch
in _get_tensorflow_library_paths, similar to
test_tensorflow_library_paths_include_libpython_for_old_tensorflow. Mock
deepmd.tf.env with TF_VERSION set to 2.12.0 or later and assert the returned
TensorFlow library paths are still correct while the libpython preload list is
empty, ensuring the version-gated logic does not regress.

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Repository UI

Review profile: CHILL

Plan: Pro

Run ID: 256487fb-b134-4331-9a78-ff4abb1c8d12

📥 Commits

Reviewing files that changed from the base of the PR and between 42de67e and 81d8828.

📒 Files selected for processing (3)

• deepmd/lmp.py
• doc/install/easy-install.md
• source/tests/test_lmp.py

[codecov]

Codecov Report

❌ Patch coverage is 93.54839% with 2 lines in your changes missing coverage. Please review.
✅ Project coverage is 81.07%. Comparing base (42de67e) to head (81d8828).
⚠️ Report is 2 commits behind head on master.

Files with missing lines	Patch %	Lines
deepmd/lmp.py	93.54%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##           master    #5707      +/-   ##
==========================================
- Coverage   81.22%   81.07%   -0.16%     
==========================================
  Files         980      980              
  Lines      109352   109369      +17     
  Branches     4205     4208       +3     
==========================================
- Hits        88821    88667     -154     
- Misses      19011    19179     +168     
- Partials     1520     1523       +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.