search

deepmodeling / deepmd-kit

A deep learning package for many-body potential energy representation and molecular dynamics
https://docs.deepmodeling.com/projects/deepmd/
GNU Lesser General Public License v3.0
1.41k stars 486 forks source link

feat: add seeds to dpmodel and fix seeds in tf & pt #3880

Closed njzjz closed 2 weeks ago

njzjz commented 2 weeks ago

Fix #3799.

Summary by CodeRabbit

coderabbitai[bot] commented 2 weeks ago

Walkthrough

The main changes introduce a seed parameter across numerous classes and functions in the deepmd module, supporting both single integers and lists of integers. These adjustments ensure the seeding mechanism is uniformly and effectively passed down to the network layers, thereby enhancing the reproducibility and randomness control.

Changes

Files/Modules Changes Summary
deepmd/dpmodel/descriptor/dpa1.py Added seed parameter to constructors.
deepmd/dpmodel/descriptor/dpa2.py Updated init_subclass_params to include seed parameter.
deepmd/dpmodel/descriptor/repformers.py Enhanced DescrptBlockRepformers class with various new parameters, including seed.
deepmd/dpmodel/descriptor/se_atten_v2.py Modified seed in DescrptDPA1 class constructor to accept lists.
deepmd/dpmodel/descriptor/se_e2_a.py Modified calculation and usage of seed in __init__ method.
deepmd/dpmodel/descriptor/se_r.py Added seed parameter with calculation in constructor.
deepmd/dpmodel/descriptor/se_t.py Enhanced seeding logic with child_seed function.
deepmd/dpmodel/fitting/dipole_fitting.py Updated seed value to accept both integer and list types.
deepmd/dpmodel/fitting/dos_fitting.py Updated seed value to accept both integer and list types.
deepmd/dpmodel/fitting/ener_fitting.py Updated seed value to accept both integer and list types.
deepmd/dpmodel/fitting/general_fitting.py Added seed parameter in GeneralFitting class.
deepmd/dpmodel/fitting/polarizability_fitting.py Updated seed value to accept both integer and list types.
deepmd/dpmodel/utils/network.py Introduced optional precision and seed parameters in various class constructors.
deepmd/dpmodel/utils/seed.py Added child_seed function for generating child seeds.
deepmd/dpmodel/utils/type_embed.py Added seed parameter to __init__ method of affected class.
deepmd/pt/model/descriptor/dpa1.py Enhanced seed logic to support integers and lists in class constructor.
deepmd/pt/model/descriptor/dpa2.py Modified init_subclass_params to handle seed with additional computations and child_seed function.
deepmd/pt/model/descriptor/repformer_layer.py Enhanced RepformerLayer class by modifying seed values to control random seed behavior effectively.
deepmd/pt/model/descriptor/se_atten.py Enhanced seeding logic with customized calculations for seed parameter based on existing attributes.
deepmd/pt/model/network/network.py Adjusted __init__ method to modify seed parameter behavior.
deepmd/pt/utils/utils.py Enhanced get_generator function to accept lists as seeds and hash them to generate a torch generator seed.

Sequence Diagram(s)

N/A

Assessment against linked issues

Objective (Issue #3799) Addressed Explanation
Ensure seed parameters are passed to network layers ✅
Enhance seed parameter to accept lists of integers ✅
Modify seeding logic using child_seed function ✅

Thank you for using CodeRabbit. We offer it for free to the OSS community and would appreciate your support in helping us grow. If you find it useful, would you consider giving us a shout-out on your favorite social media?

Share - [X](https://twitter.com/intent/tweet?text=I%20just%20used%20%40coderabbitai%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20the%20proprietary%20code.%20Check%20it%20out%3A&url=https%3A//coderabbit.ai) - [Mastodon](https://mastodon.social/share?text=I%20just%20used%20%40coderabbitai%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20the%20proprietary%20code.%20Check%20it%20out%3A%20https%3A%2F%2Fcoderabbit.ai) - [Reddit](https://www.reddit.com/submit?title=Great%20tool%20for%20code%20review%20-%20CodeRabbit&text=I%20just%20used%20CodeRabbit%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20proprietary%20code.%20Check%20it%20out%3A%20https%3A//coderabbit.ai) - [LinkedIn](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fcoderabbit.ai&mini=true&title=Great%20tool%20for%20code%20review%20-%20CodeRabbit&summary=I%20just%20used%20CodeRabbit%20for%20my%20code%20review%2C%20and%20it%27s%20fantastic%21%20It%27s%20free%20for%20OSS%20and%20offers%20a%20free%20trial%20for%20proprietary%20code)
Tips ### Chat There are 3 ways to chat with [CodeRabbit](https://coderabbit.ai): - Review comments: Directly reply to a review comment made by CodeRabbit. Example: - `I pushed a fix in commit .` - `Generate unit testing code for this file.` - `Open a follow-up GitHub issue for this discussion.` - Files and specific lines of code (under the "Files changed" tab): Tag `@coderabbitai` in a new review comment at the desired location with your query. Examples: - `@coderabbitai generate unit testing code for this file.` - `@coderabbitai modularize this function.` - PR comments: Tag `@coderabbitai` in a new PR comment to ask questions about the PR branch. For the best results, please provide a very specific query, as very limited context is provided in this mode. Examples: - `@coderabbitai generate interesting stats about this repository and render them as a table.` - `@coderabbitai show all the console.log statements in this repository.` - `@coderabbitai read src/utils.ts and generate unit testing code.` - `@coderabbitai read the files in the src/scheduler package and generate a class diagram using mermaid and a README in the markdown format.` - `@coderabbitai help me debug CodeRabbit configuration file.` Note: Be mindful of the bot's finite context window. It's strongly recommended to break down tasks such as reading entire modules into smaller chunks. For a focused discussion, use review comments to chat about specific files and their changes, instead of using the PR comments. ### CodeRabbit Commands (invoked as PR comments) - `@coderabbitai pause` to pause the reviews on a PR. - `@coderabbitai resume` to resume the paused reviews. - `@coderabbitai review` to trigger an incremental review. This is useful when automatic reviews are disabled for the repository. - `@coderabbitai full review` to do a full review from scratch and review all the files again. - `@coderabbitai summary` to regenerate the summary of the PR. - `@coderabbitai resolve` resolve all the CodeRabbit review comments. - `@coderabbitai configuration` to show the current CodeRabbit configuration for the repository. - `@coderabbitai help` to get help. Additionally, you can add `@coderabbitai ignore` anywhere in the PR description to prevent this PR from being reviewed. ### CodeRabbit Configration File (`.coderabbit.yaml`) - You can programmatically configure CodeRabbit by adding a `.coderabbit.yaml` file to the root of your repository. - Please see the [configuration documentation](https://docs.coderabbit.ai/guides/configure-coderabbit) for more information. - If your editor has YAML language server enabled, you can add the path at the top of this file to enable auto-completion and validation: `# yaml-language-server: $schema=https://coderabbit.ai/integrations/schema.v2.json` ### Documentation and Community - Visit our [Documentation](https://coderabbit.ai/docs) for detailed information on how to use CodeRabbit. - Join our [Discord Community](https://discord.com/invite/GsXnASn26c) to get help, request features, and share feedback. - Follow us on [X/Twitter](https://twitter.com/coderabbitai) for updates and announcements.
codecov[bot] commented 2 weeks ago

Codecov Report

Attention: Patch coverage is 97.89474% with 2 lines in your changes missing coverage. Please review.

Project coverage is 82.74%. Comparing base (c644314) to head (8aacdea). Report is 3 commits behind head on devel.

Files Patch % Lines
deepmd/dpmodel/descriptor/repformers.py 83.33% 1 Missing :warning:
deepmd/dpmodel/utils/seed.py 92.30% 1 Missing :warning:
Additional details and impacted files ```diff @@ Coverage Diff @@ ## devel #3880 +/- ## ========================================== + Coverage 82.70% 82.74% +0.03% ========================================== Files 517 518 +1 Lines 50141 50215 +74 Branches 2984 2984 ========================================== + Hits 41469 41548 +79 + Misses 7762 7757 -5 Partials 910 910 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

njzjz commented 2 weeks ago

I find one has to spend much effort on maintaining the number of seeds used by each module. I would recommend the parallel random number generation feature by which we do not need to manually record the number of consumed seeds. see e.g. https://numpy.org/doc/stable/reference/random/parallel.html

It's unclear how to share arguments with PT when input is a numpy.random.SeedSequence, considering PT also uses several functions in dpmodel.

njzjz commented 2 weeks ago

I find one has to spend much effort on maintaining the number of seeds used by each module. I would recommend the parallel random number generation feature by which we do not need to manually record the number of consumed seeds. see e.g. https://numpy.org/doc/stable/reference/random/parallel.html

It's unclear how to share arguments with PT when input is a numpy.random.SeedSequence, considering PT also uses several functions in dpmodel.

I just realized one can do this:

>>> import numpy as np
>>> rng = np.random.default_rng([1,2,3,4,5])

So it will work if we change the type of seed from int | None to int | list[int] | None.

njzjz commented 2 weeks ago

Could you plz add a UT for mixed entropy to make sure the implementation is consistent with numpy.

No, it's not consistent, as PyTorch limits the seed to (-0xffff_ffff_ffff_ffff, 0xffff_ffff_ffff_ffff). It is only ensured that the seed generated by mix_entropy has high entropy by using a similar way to NumPy.