【MIIT program】add GDI-NN models - #294
Open
dcx183-maker wants to merge 7 commits into
Open
Conversation
Collaborator
|
辛苦人工检查下提交代码 也按照该规范检查下#258 |
Author
|
重新修改上传 |
|
Thanks for your contribution! |
Author
|
已同步最新 develop 并解决冲突,GDI-NN 聚焦测试已通过。 |
leeleolay
reviewed
Jul 16, 2026
| pid = os.getpid() | ||
| pgid = os.getpgid(os.getpid()) | ||
| print("main proc {} exit, kill process group " "{}".format(pid, pgid)) | ||
| logger.warning("Main process %s exiting; killing process group %s.", pid, pgid) |
| logger.warning("Main process %s exiting; killing process group %s.", pid, pgid) | ||
| os.killpg(pgid, signal.SIGKILL) | ||
|
|
||
|
|
Author
|
麻烦再看一下 |
Collaborator
|
冲突需解决 |
leeleolay
reviewed
Jul 29, 2026
Integrate Paddle-native GE-GNN training, inference, configuration, and documentation for binary activity coefficients.
Use the standard path-based dataset interface and shared PaddleMaterials data cache.
Keep the shared collator module unchanged because the GE-GNN collator is resolved from its dataset module.
Describe the data file mapping for the published GE-GNN resources.
Collaborator
|
重构了部分基础组件,移动了推理器的位置,辛苦基于新的开发和尝试 @dcx183-maker |
Resolve the develop merge conflicts while preserving the GE-GNN dataset, model, binary-mixture inference, configuration, and tests. Keep the new predictor locations and expose the collator through the shared loader interface.
leeleolay
reviewed
Aug 18, 2026
| dataset: | ||
| __class_name__: BinaryActivityDataset | ||
| __init_params__: | ||
| path: "${oc.env:PPMAT_DATA_PATH,./data/binary_activity/output_binary_with_inf_all.csv}" |
| loader: | ||
| num_workers: 0 | ||
| use_shared_memory: False | ||
| collate_fn: BinaryActivityCollator |
Comment on lines
+109
to
+114
| _DATA_URL = ( | ||
| "https://paddle-org.bj.bcebos.com/paddlematerials/datasets/" | ||
| "thermodynamic_data_of_binary_mixtures/" | ||
| ) | ||
| _DATA_FILE = "output_binary_with_inf_all.csv" | ||
| _SOLVENT_FILE = "solvent_list.csv" |
| _SOLVENT_FILE = "solvent_list.csv" | ||
|
|
||
|
|
||
| def smiles_to_pgl_graph(smiles, add_self_loop=True): |
Collaborator
There was a problem hiding this comment.
使用已有的构造graph的功能?参考下moleculargraph是否可以使用
Collaborator
There was a problem hiding this comment.
参考mp20dataset.py格式,当前无cache功能,解析路径函数请复用已有的。合理的方式建议copy已有的再做小修改
Use standard molecular graph conversion and caching, default collation, and BasePredictor-compatible inference APIs.
Resolve predictor API conflicts while preserving GE-GNN binary-mixture inference.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
GE-GNN Binary Activity Coefficient Prediction
Added
GEGNN: predicts binary-mixture log activity coefficients from excess Gibbs free energy.BinaryActivityDataset: PaddleMaterials-style dataset using the standardpathinterface.Dataset
BinaryActivityDatasetacceptspathfor the binary-mixture CSV. When the file is absent, it is downloaded through the shared PaddleMaterials dataset cache (~/.paddlemat/datasets); the siblingsolvent_list.csvis loaded from the same directory. No datasets or checkpoints are included in this PR.The default config path can be overridden with
PPMAT_DATA_PATH. Dataset layout and required CSV columns are documented inproperty_prediction/configs/gegnn/README.md.Validation
python -m compileallpassed for the modified Python files.git diff --checkpassed.pytest -q test/test_gegnn.pyis currently blocked during test collection by the local environment:paddle_scatterdoes not exportscatter. No test result is claimed from this environment.Scope
This PR contains only GE-GNN dataset, model, configuration, documentation, prediction integration, and Paddle test changes. It excludes traffic metrics, workflows, architecture documentation, Torch cross-framework tests, data/checkpoint artifacts, and hard-coded local paths.