Skip to content

Install.md

Latest commit

 

History

History
142 lines (97 loc) · 4.43 KB

Install.md

File metadata and controls

142 lines (97 loc) · 4.43 KB
Raw

Installation 🔧

简体中文

1. Installation Instructions

We recommend using a conda virtual environment to manage dependencies. You can install conda via Miniforge.

1.1 Create Virtual Environment

Create and activate a new conda virtual environment:

conda create -n ppmat python=3.10
conda activate ppmat

We currently develop under Python 3.10 environment and recommend using Python 3.10 or newer.

1.2 Install PaddlePaddle

Install the appropriate PaddlePaddle version based on your CUDA version. Refer to the PaddlePaddle Official Website for installation commands. We recommend installing PaddlePaddle version >= 3.1 or the develop version.

For example, in a CUDA 12.6 environment, install the paddlepaddle-gpu version:

python -m pip install paddlepaddle-gpu==3.1.0 -i https://www.paddlepaddle.org.cn/packages/stable/cu126/

After installation, verify the installation with:

python -c "import paddle; paddle.utils.run_check()"

If you see "PaddlePaddle is installed successfully! Let's start deep learning with PaddlePaddle now.", the installation was successful.

1.3 Install PaddleMaterials

Install the released package from PyPI:

python -m pip install ppmat

Verify that PaddleMaterials is imported from the installed package:

python -c "import ppmat; print(ppmat.__version__); print(ppmat.__file__)"

For development, install PaddleMaterials from source:

# Clone PaddleMaterials repository
git clone https://github.com/PaddlePaddle/PaddleMaterials.git

# Navigate to PaddleMaterials directory
cd PaddleMaterials

# Install dependencies
pip install --upgrade pip setuptools==68.2.2 wheel
pip install setuptools_scm
pip install Cython

# Install in editable mode
pip install -e . --no-build-isolation
# pip install -e . --no-build-isolation -i https://pypi.tuna.tsinghua.edu.cn/simple recommended if you are in China

2. Run Examples

The task scripts, configuration files, and example data are maintained in the source repository and are not included in the ppmat wheel. Clone the repository and run the following commands from its root directory.

2.1 Property Prediction

Predict material formation energy using a pretrained MEGNet model:

python property_prediction/predict.py \
    --model_name='megnet_mp2018_train_60k_e_form' \
    --weights_name='best.pdparams' \
    --input_format=cif --input_path='./property_prediction/example_data/cifs/' \
    --output_path='results/'

2.2 Structure Generation

Generate crystal structures with four atoms using a pretrained MatterGen model:

python structure_generation/sample.py \
    --model_name='mattergen_mp20' \
    --weights_name='latest.pdparams' \
    --output_path='result_mattergen_mp20/' \
    --mode='by_num_atoms' \
    --num_atoms=4

2.3 Interatomic Potentials

Predict energy and forces using a pretrained MatterSim model:

python interatomic_potentials/predict.py \
    --model_name='mattersim_1M' \
    --weights_name='mattersim-v1.0.0-1M_model.pdparams' \
    --input_format=cif --input_path='./interatomic_potentials/example_data/cifs/' \
    --output_path='results/'

2.4 Electronic Structure

Predict electron density using a trained InfGCN checkpoint:

python electronic_structure/predict.py \
    --model_name='infgcn_qm9' \
    --weights_name='best.pdparams' \
    --input_format=mol --input_path='electronic_structure/example_data/methane.mol' \
    --grid_shape=8 \
    --grid_batch_size=4096 \
    --output_path='output/infgcn_qm9/methane'

Prepare the dataset and checkpoint as described in the InfGCN prediction guide.

2.5 Spectrum Elucidation

Run NMR spectrum elucidation using a trained DiffNMR checkpoint:

python spectrum_elucidation/sample.py \
    --model_name='diffnmr_msdnmr_nless15' \
    --weights_name='best.pdparams' \
    --output_path='result_diffnmr_nless15/'

2.6 Spectrum Enhancement

Enhance STEM images using a pretrained SFIN model:

python spectrum_enhancement/predict.py \
    --model_name='sfin_haadf_enhance' \
    --weights_name='best.pdparams' \
    --input_path='path/to/noisy_image.png' \
    --output_path='result_sfin/'

For more usage instructions, refer to the task-specific README files or the Get Started documentation.