neural-function-distributions

KO | EN

gitlite — search

#javascript #python #ai #typescript #go #golang #machine-learning #rust #windows #cli #awesome-list #pytorch #mcp #ios

neural-function-distributions

(★ 152)

Pytorch implementation of Generative Models as Distributions of Functions 🌿

파일 탐색기

최종 버전 다운로드 (.zip)

- config_celebahq128.json
- config_celebahq64.json
- config_era5.json
- config_shapenet_point_clouds.json
- config_shapenet_voxels.json
- __init__.py
- conversion.py
- dataloaders.py
- dataloaders3d.py
- dataloaders_era5.py
- era5.gif
- example.gif
- __init__.py
- discriminator.py
- function_distribution.py
- function_representation.py
- pointconv.py
- __init__.py
- training.py
- __init__.py
- plots.py
- plots_globe.py
- render.py
LICENSE
load-trained-model.ipynb
main.py
README.md
render_mesh.py
requirements.txt

// repository documentation

Was this content helpful?

★ 0 (0 ratings)

# Generative Models as Distributions of Functions

This repo contains code to reproduce all experiments in [Generative Models as Distributions of Functions](https://arxiv.org/abs/2102.04776).

## Requirements

Requirements for training the models can be installed using `pip install -r requirements.txt`. All experiments were run using `python 3.8.10`.

## Training a model

To train a model on CelebAHQ64, run

```python main.py configs/config_celebahq64.json```

Example configs to reproduce the results in the paper are provided in the `configs` folder. Note that you will have to provide a path to the data you wish to train on in the config.

## Downloading datasets

The shapenet and ERA5 climate datasets can be downloaded at this [link](https://drive.google.com/drive/folders/1r_sk5auYvllSpDG9ZjroOG0SH0v5kPmM?usp=sharing). The CelebAHQ datasets can be downloaded from [here](https://www.kaggle.com/datasets/lamsimon/celebahq/data).

## Loading trained models

All trained models can be downloaded from [here](https://drive.google.com/drive/folders/1r_sk5auYvllSpDG9ZjroOG0SH0v5kPmM?usp=sharing). The `load-trained-model.ipynb` notebook shows an example of using a trained model.

## Rendering 3D samples

The requirements in `requirements.txt` allow for basic plotting of 3D shapes with matplotlib. However, to properly render 3D models, you will need to install [mcubes](https://github.com/pmneila/PyMCubes) (for marching cubes) and [pytorch3d](https://github.com/facebookresearch/pytorch3d). Pytorch3D is not directly pip installable (depending on your version of torch), so please follow the [install instructions](https://github.com/facebookresearch/pytorch3d/blob/master/INSTALL.md) provided in their repo.

Once these requirements have been installed, you can render samples using

```python render_mesh.py -sp samples -mr 32 -rs 64 -ns 8 -mp trained-models/shapenet_voxels/model.pt```

See `render_mesh.py` for a full list of rendering options.

## Globe plots

By default the samples from the climate data will be visualized as flat projections. Plotting these on a globe (using the `viz/plots_globe.py` functions) requires the [cartopy](https://scitools.org.uk/cartopy/docs/latest/) library. Installation instructions for this library can be found [here](https://scitools.org.uk/cartopy/docs/latest/installing.html).

## License

MIT

neural-function-distributions

파일 탐색기

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

neural-function-distributions

파일 탐색기

오늘의 추천 프로젝트

Was this content helpful?

Recent Feedback

Download README