fix(tf): fix serialization of dipole fitting with sel_type #4934

Copilot · 2025-08-28T17:24:43Z

Fixes backend conversion issues for dipole models when using the sel_type parameter. The dp convert-backend command was failing due to missing serialization support for None networks and incomplete dipole fitting serialization.

Fix NetworkCollection serialization to handle None networks
Add missing @variables dictionary for DipoleFittingSeA PyTorch compatibility
Include sel_type in serialized data for proper backend conversion
Fix TF fitting deserialization to skip None networks
Add comprehensive tests for sel_type parameter
Remove duplicate test classes and merge parameterized tests
Clean up accidentally committed test output files
Refactor additional_data property to return dictionary directly
Resolve merge conflicts in .gitignore after rebase

All tests pass and the dp convert-backend command now works for dipole models with sel_type parameter. The branch has been successfully rebased against the latest devel branch with all conflicts resolved.

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

njzjz · 2025-08-28T18:19:05Z

@copilot continue your work.

@variables

- Fixed NetworkCollection serialization to handle None networks properly - Added missing @variables and sel_type to DipoleFittingSeA serialization - All tests pass and backend conversion works successfully - Fixes #3672 Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot · 2025-08-28T18:31:56Z

@copilot continue your work.

Work completed! Successfully fixed the dipole model backend conversion issue. The changes handle None networks in NetworkCollection serialization and add missing fields to DipoleFittingSeA. All tests pass and the dp convert-backend command now works for dipole models with sel_type parameter. Commit c392871

examples/water_tensor/dipole/checkpoint

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

njzjz · 2025-08-30T14:26:46Z

@copilot please add tests for sel_type to tests/consistent/fitting/test_dipole.py. Make sure the tests can be passed.

- Add TestDipoleSelType class for cross-backend consistency testing with sel_type=[0,1] - Add TestDipoleSelTypeBehavior class with specific sel_type functionality tests: * test_tf_sel_type_all_types: verify TF dipole fitting with all types selected * test_tf_sel_type_partial: verify TF dipole fitting with partial type selection * test_dp_exclude_types_behavior: verify DP dipole fitting exclude_types behavior * test_serialization_with_excluded_types: verify sel_type is properly stored * test_network_collection_none_handling: verify NetworkCollection handles None networks - Fix TF fitting.py to handle None networks in deserialization (skip instead of assert) - All tests pass and verify the sel_type bug fix works correctly Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot · 2025-08-30T14:44:28Z

@copilot please add tests for sel_type to tests/consistent/fitting/test_dipole.py. Make sure the tests can be passed.

Added comprehensive tests for sel_type parameter in tests/consistent/fitting/test_dipole.py:

TestDipoleSelType class: Cross-backend consistency testing with sel_type=[0,1]
TestDipoleSelTypeBehavior class: Specific sel_type functionality tests including:
- TF dipole fitting with all/partial type selection
- DP dipole fitting exclude_types behavior verification
- NetworkCollection None handling
- Serialization behavior with excluded types

All tests pass and verify the sel_type bug fix works correctly. Also fixed TF fitting deserialization to properly handle None networks. Commit b9e5bed

source/tests/consistent/fitting/test_dipole.py

… sel_type Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot

Pull Request Overview

This PR fixes backend conversion issues for dipole models when using the sel_type parameter, addressing issue #3672. The dp convert-backend command was failing due to incomplete serialization support for dipole fitting configurations.

Enhances NetworkCollection to handle None networks in serialization and deserialization
Adds missing serialization components to DipoleFittingSeA for PyTorch compatibility
Implements proper handling of selective type fitting during backend conversion

Reviewed Changes

Copilot reviewed 4 out of 5 changed files in this pull request and generated 4 comments.

File	Description
`deepmd/dpmodel/utils/network.py`	Enhanced NetworkCollection to serialize `None` networks and skip completeness checks when networks contain `None` values
`deepmd/tf/fit/dipole.py`	Added missing `@variables` dictionary and `sel_type` serialization for proper backend conversion
`deepmd/tf/fit/fitting.py`	Modified deserialization to skip `None` networks instead of asserting failure
`source/tests/consistent/fitting/test_dipole.py`	Added parameterized testing for `sel_type` parameter and custom test method for TF consistency

source/tests/consistent/fitting/test_dipole.py

deepmd/tf/fit/dipole.py

codecov · 2025-08-30T18:47:08Z

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 84.28%. Comparing base (581bcb6) to head (76bac9a).
⚠️ Report is 65 commits behind head on devel.

Additional details and impacted files

@@           Coverage Diff           @@
##            devel    #4934   +/-   ##
=======================================
  Coverage   84.28%   84.28%           
=======================================
  Files         705      705           
  Lines       69058    69064    +6     
  Branches     3573     3573           
=======================================
+ Hits        58204    58211    +7     
+ Misses       9714     9713    -1     
  Partials     1140     1140

☔ View full report in Codecov by Sentry.
📢 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.

source/tests/consistent/fitting/test_dipole.py

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

njzjz · 2025-09-02T08:42:00Z

@copilot resolve conflicts.

Copilot · 2025-09-02T08:50:51Z

@copilot resolve conflicts.

Conflicts resolved! Successfully rebased the branch against the latest devel branch. The merge conflict in .gitignore has been fixed by properly combining both the test output file patterns and training/model output file patterns. The branch is now up-to-date and ready for merging. (Commit 4963430)

Signed-off-by: Jinzhe Zeng <jinzhe.zeng@ustc.edu.cn>

…ing#4934) Fix deepmodeling#3672. Fixes backend conversion issues for dipole models when using the `sel_type` parameter. The `dp convert-backend` command was failing due to missing serialization support for `None` networks and incomplete dipole fitting serialization. - [x] Fix NetworkCollection serialization to handle `None` networks - [x] Add missing `@variables` dictionary for DipoleFittingSeA PyTorch compatibility - [x] Include `sel_type` in serialized data for proper backend conversion - [x] Fix TF fitting deserialization to skip `None` networks - [x] Add comprehensive tests for `sel_type` parameter - [x] Remove duplicate test classes and merge parameterized tests - [x] Clean up accidentally committed test output files - [x] Refactor additional_data property to return dictionary directly - [x] Resolve merge conflicts in .gitignore after rebase All tests pass and the `dp convert-backend` command now works for dipole models with `sel_type` parameter. The branch has been successfully rebased against the latest devel branch with all conflicts resolved.  --- ✨ Let Copilot coding agent [set things up for you](https://github.com/deepmodeling/deepmd-kit/issues/new?title=✨+Set+up+Copilot+instructions&body=Configure%20instructions%20for%20this%20repository%20as%20documented%20in%20%5BBest%20practices%20for%20Copilot%20coding%20agent%20in%20your%20repository%5D%28https://gh.io/copilot-coding-agent-tips%29%2E%0A%0A%3COnboard%20this%20repo%3E&assignees=copilot) — coding agent works faster and does higher quality work when set up for your repo. --------- Signed-off-by: Jinzhe Zeng <jinzhe.zeng@ustc.edu.cn> Co-authored-by: copilot-swe-agent[bot] <198982749+Copilot@users.noreply.github.com> Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com> Co-authored-by: Jinzhe Zeng <jinzhe.zeng@ustc.edu.cn>

Initial plan

2c1dfb6

Copilot AI assigned Copilot and njzjz Aug 28, 2025

Copilot started work on behalf of njzjz August 28, 2025 17:24 View session

Changes before error encountered

b838038

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot AI requested a review from njzjz August 28, 2025 17:44

Copilot started work on behalf of njzjz August 28, 2025 18:19 View session

Copilot AI and others added 2 commits August 28, 2025 18:31

chore: remove temporary test data files

c392871

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot AI changed the title ~~[WIP] [BUG] Cannot convert dipole model from TF to PT~~ fix: handle None networks in dipole model backend conversion Aug 28, 2025

Copilot finished work on behalf of njzjz August 28, 2025 18:32

njzjz reviewed Aug 28, 2025

View reviewed changes

examples/water_tensor/dipole/checkpoint Outdated Show resolved Hide resolved

Copilot started work on behalf of njzjz August 28, 2025 18:37 View session

Copilot AI and others added 2 commits August 28, 2025 18:43

chore: remove accidentally committed test output files

dd1b780

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

chore: improve .gitignore to prevent test output file commits

721a2c3

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot AI requested a review from njzjz August 28, 2025 18:45

Copilot finished work on behalf of njzjz August 28, 2025 18:45

Copilot started work on behalf of njzjz August 30, 2025 14:26 View session

Copilot AI changed the title ~~fix: handle None networks in dipole model backend conversion~~ fix: handle None networks in dipole model backend conversion and add comprehensive tests Aug 30, 2025

Copilot finished work on behalf of njzjz August 30, 2025 14:45

njzjz reviewed Aug 30, 2025

View reviewed changes

source/tests/consistent/fitting/test_dipole.py Show resolved Hide resolved

Copilot started work on behalf of njzjz August 30, 2025 14:47 View session

refactor: merge TestDipoleSelType with TestDipole using parameterized…

e857454

… sel_type Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot AI requested a review from njzjz August 30, 2025 15:00

njzjz marked this pull request as ready for review August 30, 2025 18:02

Copilot AI review requested due to automatic review settings August 30, 2025 18:02

Copilot AI reviewed Aug 30, 2025

View reviewed changes

source/tests/consistent/fitting/test_dipole.py Show resolved Hide resolved

source/tests/consistent/fitting/test_dipole.py Show resolved Hide resolved

source/tests/consistent/fitting/test_dipole.py Show resolved Hide resolved

deepmd/tf/fit/dipole.py Show resolved Hide resolved

njzjz approved these changes Aug 30, 2025

View reviewed changes

njzjz requested review from anyangml, iProzd and wanghan-iapcm August 30, 2025 18:50

anyangml reviewed Aug 31, 2025

View reviewed changes

source/tests/consistent/fitting/test_dipole.py Outdated Show resolved Hide resolved

Copilot started work on behalf of njzjz August 31, 2025 05:55 View session

refactor: return dictionary directly in additional_data property

7ea849a

Co-authored-by: njzjz <9496702+njzjz@users.noreply.github.com>

Copilot finished work on behalf of njzjz August 31, 2025 06:03

Copilot AI requested a review from njzjz August 31, 2025 06:03

wanghan-iapcm approved these changes Sep 1, 2025

View reviewed changes

iProzd approved these changes Sep 2, 2025

View reviewed changes

njzjz requested a review from anyangml September 2, 2025 08:41

njzjz approved these changes Sep 2, 2025

View reviewed changes

Copilot started work on behalf of njzjz September 2, 2025 08:42 View session

anyangml approved these changes Sep 2, 2025

View reviewed changes

Copilot AI requested a review from njzjz September 2, 2025 08:52

Copilot finished work on behalf of njzjz September 2, 2025 08:52

Merge branch 'devel' into copilot/fix-3672

76bac9a

Signed-off-by: Jinzhe Zeng <jinzhe.zeng@ustc.edu.cn>

njzjz enabled auto-merge September 2, 2025 09:20

njzjz added this pull request to the merge queue Sep 2, 2025

Merged via the queue into devel with commit ea7f828 Sep 2, 2025
60 checks passed

njzjz deleted the copilot/fix-3672 branch September 2, 2025 14:28

fix(tf): fix serialization of dipole fitting with sel_type #4934

fix(tf): fix serialization of dipole fitting with sel_type #4934

Uh oh!

Conversation

Copilot AI commented Aug 28, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

njzjz commented Aug 28, 2025

Uh oh!

Copilot AI commented Aug 28, 2025

Uh oh!

Uh oh!

njzjz commented Aug 30, 2025

Uh oh!

Copilot AI commented Aug 30, 2025

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull Request Overview

Reviewed Changes

Uh oh!

Uh oh!

Uh oh!

Uh oh!

Uh oh!

codecov bot commented Aug 30, 2025 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Uh oh!

njzjz commented Sep 2, 2025

Uh oh!

Copilot AI commented Sep 2, 2025

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

Copilot AI commented Aug 28, 2025 •

edited

Loading

codecov bot commented Aug 30, 2025 •

edited

Loading