torchdatasets
Ready to Use Extra Datasets for Torch
Provides datasets in a format that can be easily consumed by torch 'dataloaders'. Handles data downloading from multiple sources, caching and pre-processing so users can focus only on their model implementations.
torchdatasets
torchdatasets provides ready-to-use datasets compatible with the torch package.
Installation
The released version of torchdatasets can be installed with:
install.packages("torchdatasets")
You can also install the development version with:
remotes::install_github("mlverse/torchdatasets")
Datasets
Currently, the following datasets are implemented:
| Dataset | Domain | Type | Authentication |
|---|---|---|---|
| bird_species_dataset() | Images | Classification | Not required |
| dogs_vs_cats_dataset() | Images | Classification | Not required |
| guess_the_correlation_dataset() | Images | Regression | Not required |
| cityscapes_pix2pix_dataset() | Images | Segmentation | Not required |
| oxford_pet_dataset() | Images | Segmentation | Not required |
| bank_marketing_dataset() | Tabular | Classification | Not required |
| imdb_dataset() | Text | Classification | Not required |
| spam_dataset() | Tabular | Classification | Not required |