【MIIT program】SevenNet模型复现

interatomic_potentials/README.md

@@ -6,30 +6,81 @@ Machine-learning interatomic potentials (MLIP) bridge the gap between quantum-le
 
 ## 2.Models Matrix
 
-| **Supported Functions**             | **[CHGNet](./configs/chgnet/README.md)** | **[MatterSim](./configs/mattersim//README.md)** |
-| ----------------------------------- | ---------------------------------------- | ----------------------------------------------- |
-| **Forward Prediction**              |                                          |                                                 |
-|  Energy                        | ✅                                       | ✅                                              |
-|  Force                         | ✅                                       | ✅                                              |
-|  Stress                        | ✅                                       | ✅                                              |
-|  Magmom                        | ✅                                       | -                                               |
-| **ML Capabilities · Training**      |                                          |                                                 |
-|  Single-GPU                    | ✅                                       | ✅                                              |
-|  Distributed Train             | ✅                                       | ✅                                              |
-|  Mixed Precision               | -                                        | -                                               |
-|  Fine-tuning                   | ✅                                       | ✅                                              |
-|  Uncertainty / Active-Learning | -                                        | -                                               |
-|  Dynamic→Static                | -                                        | -                                               |
-|  Compiler CINN                 | -                                        | -                                               |
-| **ML Capabilities · Predict**       |                                          |                                                 |
-|  Distillation / Pruning        | -                                        | -                                               |
-|  Standard inference            | ✅                                       | ✅                                              |
-|  Distributed inference         | -                                        | -                                               |
-|  Compiler CINN                 | -                                        | -                                               |
-| **Molecular Dynamic Interface**     |                                          |                                                 |
-|  ASE                           | ✅                                       | ✅                                              |
-| **Dataset**                         |                                          |                                                 |
-|  MPtrj                         | ✅                                       | 🚧                                              |
-| **ML2DDB🌟**                        | ✅                                       | -                                               | 
+| **Supported Functions**             | **[CHGNet](./configs/chgnet/README.md)** | **[MatterSim](./configs/mattersim//README.md)** | **[SevenNet](./configs/sevennet/README.md)** |
+| ----------------------------------- | ---------------------------------------- | ----------------------------------------------- | -------------------------------------------- |
+| **Forward Prediction**              |                                          |                                                 |                                              |
+|  Energy                        | ✅                                       | ✅                                              | ✅                                           |
+|  Force                         | ✅                                       | ✅                                              | ✅                                           |
+|  Stress                        | ✅                                       | ✅                                              | -                                            |
+|  Magmom                        | ✅                                       | -                                               | -                                            |
+| **ML Capabilities · Training**      |                                          |                                                 |                                              |
+|  Single-GPU                    | ✅                                       | ✅                                              | ✅                                           |
+|  Distributed Train             | ✅                                       | ✅                                              | -                                            |
+|  Mixed Precision               | -                                        | -                                               | -                                            |
+|  Fine-tuning                   | ✅                                       | ✅                                              | ✅                                           |
+|  Uncertainty / Active-Learning | -                                        | -                                               | -                                            |
+|  Dynamic→Static                | -                                        | -                                               | -                                            |
+|  Compiler CINN                 | -                                        | -                                               | -                                            |
+| **ML Capabilities · Predict**       |                                          |                                                 |                                              |
+|  Distillation / Pruning        | -                                        | -                                               | -                                            |
+|  Standard inference            | ✅                                       | ✅                                              | ✅                                           |
+|  Distributed inference         | -                                        | -                                               | -                                            |
+|  Compiler CINN                 | -                                        | -                                               | -                                            |
+| **Molecular Dynamic Interface**     |                                          |                                                 |                                              |
+|  ASE                           | ✅                                       | ✅                                              | ✅                                           |
+| **Dataset**                         |                                          |                                                 |                                              |
+|  MPtrj                         | ✅                                       | 🚧                                              | -                                            |
+| **ML2DDB🌟**                        | ✅                                       | -                                               | -                                            | 
 
 **Notice**:🌟 represent originate research work published from paddlematerials toolkit
+
+## 3.SevenNet Quick Start
+
+### 3.1 Overview
+
+SevenNet is a message-passing graph neural network for interatomic potential prediction. This implementation provides a pure PaddlePaddle version without external dependencies like e3nn.
+
+### 3.2 Installation
+
+```bash
+# Install dependencies
+pip install paddlepaddle>=3.0.0 ase numpy tqdm pyyaml
+```
+
+### 3.3 Training
+
+```bash
+# Navigate to interatomic_potentials directory
+cd interatomic_potentials
+
+# Run training with config file
+python sevennet_train.py --config configs/sevennet/sevennet_hfo2.yaml
+```
+
+### 3.4 Inference
+
+```bash
+python sevennet_train.py --config configs/sevennet/sevennet_hfo2.yaml \
+    --output-dir ./output/sevennet_hfo2
+```
+
+### 3.5 Configuration
+
+The configuration file includes:
+- **Model parameters**: hidden_dim, num_message_layers, num_rbf, cutoff
+- **Dataset parameters**: path, cutoff, valid_ratio
+- **Training parameters**: epoch, batch_size, learning_rate
+
+### 3.6 Test
+
+```bash
+# Run unit tests
+pytest test/test_sevennet.py -v
+```
+
+### 3.7 Notes
+
+- SevenNet uses message-passing architecture with Gaussian RBF for distance encoding
+- Supports energy and force prediction
+- Pure PaddlePaddle implementation, no PyTorch/e3nn dependencies
+- Recommended for small to medium scale molecular systems

interatomic_potentials/configs/sevennet/README.md

@@ -0,0 +1,84 @@
+# SevenNet Model Configuration
+
+## Overview
+
+SevenNet is a message-passing graph neural network for interatomic potential prediction.
+
+## Pretrained Models
+
+SevenNet provides multiple pretrained models:
+
+| Model | Description | Training Dataset | Performance (CPS) |
+|-------|-------------|------------------|-------------------|
+| **SevenNet-Omni** (Recommended) | Universal potential, 15 datasets | 15 open ab initio datasets | 0.849 |
+| SevenNet-Omni-i8 | Higher capacity (Nlayer=8) | 15 datasets | 0.859 |
+| SevenNet-Omni-i12 | Highest capacity (Nlayer=12) | 15 datasets | 0.873 |
+| SevenNet-MF-ompa | Multi-fidelity learning | MPtrj, sAlex, OMat24 | 0.845 |
+| SevenNet-omat | OMat24 only | OMat24 | κSRME: 0.221 |
+| SevenNet-l3i5 | MPtrj with lmax=3 | MPtrj | 0.714 |
+| SevenNet-0 | Fastest inference | MPtrj | F1: 0.67 |
+
+## Using Pretrained Models
+
+### Step 1: Download Model
+
+When official provides URLs, download the pretrained model:
+
+```python
+from ppmat.utils import download
+
+model_name = "sevennet_omni"  # or other model name
+model_path = download.get_weights_path_from_url(MODEL_REGISTRY[model_name])
+```
+
+### Step 2: Load Model
+
+```python
+import paddle
+from ppmat.models import PurePaddleSevenNet
+
+# Load checkpoint
+state_dict = paddle.load("path/to/checkpoint.pdparams")
+
+# Create model
+model = PurePaddleSevenNet(
+    num_species=100,
+    hidden_dim=128,
+    num_message_layers=5,
+    num_rbf=32,
+    cutoff=5.0,
+)
+
+# Load weights
+model.set_state_dict(state_dict)
+model.eval()
+```
+
+### Step 3: Predict
+
+```python
+# Create graph from structure
+graph = {
+    "z": paddle.to_tensor([1, 8, 1], dtype="int64"),  # H, O, H
+    "pos": paddle.to_tensor([[0, 0, 0], [1, 0, 0], [0, 1, 0]], dtype="float32"),
+    "edge_index": paddle.to_tensor([[0, 1], [1, 0]], dtype="int64"),
+}
+
+# Predict
+result = model(graph)
+energy = result["total_energy"]
+```
+
+## Training Datasets
+
+The official training datasets include:
+
+- **MPtrj** (Materials Project Trajectory): https://figshare.com/articles/dataset/Materials_Project_Trjectory_MPtrj_Dataset/23713842
+- **OMat24**: https://huggingface.co/datasets/fairchem/OMAT24
+- **sAlex**: https://huggingface.co/datasets/fairchem/OMAT24
+
+## References
+
+- [SevenNet Official Repository](https://github.com/MDIL-SNU/SevenNet)
+- [SevenNet Documentation](https://sevennet.readthedocs.io/)
+- [Pretrained Models Guide](https://sevennet.readthedocs.io/en/latest/user_guide/pretrained.html)

interatomic_potentials/configs/sevennet/sevennet_hfo2.yaml

@@ -0,0 +1,42 @@
+model:
+  type: "sevennet"
+  num_species: 100
+  hidden_dim: 32
+  num_message_layers: 3
+  num_rbf: 32
+  cutoff: 4.0
+
+dataset:
+  type: "extxyz"
+  # 使用仓库内相对路径或用户自行指定
+  # 示例: 下载数据集后放在 data/hfo2.extxyz
+  path: "data/hfo2.extxyz"
+  cutoff: 4.0
+  valid_ratio: 0.1
+
+dataloader:
+  batch_size: 2
+  shuffle: true
+  num_workers: 0
+
+optimizer:
+  type: "adam"
+  lr: 0.001
+
+lr_scheduler:
+  type: "constant"
+
+loss:
+  type: "mse"
+  force_loss_weight: 0.1
+
+trainer:
+  epoch: 2
+  per_epoch: 1
+  seed: 42
+  device: "auto"
+
+log:
+  save_dir: "./output/sevennet_hfo2"
+  log_interval: 1
+  save_interval: 1

interatomic_potentials/configs/sevennet/sevennet_hfo2_large.yaml

@@ -0,0 +1,42 @@
+model:
+  type: "sevennet"
+  num_species: 100
+  hidden_dim: 128
+  num_message_layers: 5
+  num_rbf: 64
+  cutoff: 5.0
+
+dataset:
+  type: "extxyz"
+  # 使用仓库内相对路径或用户自行指定
+  # 示例: 下载数据集后放在 data/hfo2.extxyz
+  path: "data/hfo2.extxyz"
+  cutoff: 5.0
+  valid_ratio: 0.1
+
+dataloader:
+  batch_size: 4
+  shuffle: true
+  num_workers: 0
+
+optimizer:
+  type: "adam"
+  lr: 0.001
+
+lr_scheduler:
+  type: "constant"
+
+loss:
+  type: "mse"
+  force_loss_weight: 0.1
+
+trainer:
+  epoch: 10
+  per_epoch: 2
+  seed: 42
+  device: "auto"
+
+log:
+  save_dir: "./output/sevennet_hfo2_large"
+  log_interval: 1
+  save_interval: 2

interatomic_potentials/infer.py

@@ -0,0 +1,118 @@
+#!/usr/bin/env python
+# -*- coding: utf-8 -*-
+"""
+PaddleMaterials - Interatomic Potentials Inference Script
+推理入口：python infer.py --model model.pdparams --structure structure.extxyz
+"""
+
+import argparse
+import os
+import sys
+import numpy as np
+import paddle
+from ase.io import read
+from ase.neighborlist import neighbor_list
+
+PROJECT_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+sys.path.insert(0, PROJECT_ROOT)
+
+from ppmat.models.sevennet import PurePaddleSevenNet
+
+
+def parse_args():
+    parser = argparse.ArgumentParser(description="Infer interatomic potential model")
+    parser.add_argument("--model", type=str, required=True, help="Path to model checkpoint")
+    parser.add_argument("--structure", type=str, required=True, help="Path to structure file")
+    parser.add_argument("--cutoff", type=float, default=5.0, help="Cutoff radius")
+    return parser.parse_args()
+
+
+def atoms_to_graph(atoms, cutoff):
+    positions = atoms.get_positions().astype("float32")
+    numbers = atoms.get_atomic_numbers().astype("int64")
+    i, j = neighbor_list("ij", atoms, cutoff)
+    edge_index = np.vstack([i, j]).astype("int64")
+
+    return {
+        "z": paddle.to_tensor(numbers, dtype="int64"),
+        "pos": paddle.to_tensor(positions, dtype="float32"),
+        "edge_index": paddle.to_tensor(edge_index, dtype="int64"),
+    }
+
+
+def predict(model, atoms, cutoff, energy_mean, energy_std):
+    graph = atoms_to_graph(atoms, cutoff)
+    num_atoms = len(atoms)
+
+    with paddle.no_grad():
+        out = model(graph)
+        raw_total = float(out["total_energy"].numpy())
+        pred_total = raw_total * energy_std + energy_mean
+        pred_per_atom = pred_total / num_atoms
+
+    return pred_total, pred_per_atom
+
+
+def main():
+    args = parse_args()
+
+    # 加载模型
+    print(f"Loading model: {args.model}")
+    checkpoint = paddle.load(args.model)
+    state_dict = checkpoint["model_state_dict"]
+    energy_mean = float(checkpoint.get("energy_mean", 0.0))
+    energy_std = float(checkpoint.get("energy_std", 1.0))
+
+    # 获取配置
+    config = checkpoint.get("config", {})
+    model_cfg = config.get("model", {})
+
+    # 创建模型
+    model = PurePaddleSevenNet(
+        num_species=model_cfg.get("num_species", 100),
+        hidden_dim=model_cfg.get("hidden_dim", 128),
+        num_message_layers=model_cfg.get("num_message_layers", 4),
+        num_rbf=model_cfg.get("num_rbf", 32),
+        cutoff=model_cfg.get("cutoff", args.cutoff),
+    )
+    model.set_state_dict(state_dict)
+    model.eval()
+
+    print(f"Model loaded successfully")
+    print(f"Energy mean: {energy_mean:.6f}, std: {energy_std:.6f}")
+
+    # 读取结构
+    print(f"\nReading structure: {args.structure}")
+    atoms_list = read(args.structure, index=":")
+    if not isinstance(atoms_list, list):
+        atoms_list = [atoms_list]
+    print(f"Loaded {len(atoms_list)} structures")
+
+    # 预测
+    print("\n=== Prediction Results ===")
+    for idx, atoms in enumerate(atoms_list):
+        pred_total, pred_per_atom = predict(model, atoms, args.cutoff, energy_mean, energy_std)
+
+        ref_energy = None
+        for key in ["y_energy", "energy", "free_energy", "total_energy"]:
+            if key in atoms.info:
+                try:
+                    ref_energy = float(atoms.info[key])
+                    break
+                except Exception:
+                    pass
+
+        print(f"\nStructure {idx + 1}:")
+        print(f"  Atoms: {len(atoms)}")
+        print(f"  Predicted total energy: {pred_total:.6f} eV")
+        print(f"  Predicted per-atom energy: {pred_per_atom:.6f} eV/atom")
+
+        if ref_energy is not None:
+            print(f"  Reference energy: {ref_energy:.6f} eV")
+            print(f"  Error: {abs(pred_total - ref_energy):.6f} eV")
+
+    print("\nInference completed!")
+
+
+if __name__ == "__main__":
+    main()