This pull request populates the dataset card for the REPA-E generated samples and artifacts.
It includes:
- The `unconditional-image-generation` task category and relevant `tags` in the metadata.
- Links to the associated paper, project page, and GitHub repository.
- An overview of the REPA-E method and the context of the generated samples.
- A "Sample Usage" section with Python code snippets for loading and using the E2E-VAEs, directly copied from the GitHub README.
- The quantitative results tables from the GitHub README, relevant for evaluating the generated samples.
- The BibTeX citation.
Rename per-class-sample/sit-ldm-e2e-vavae_4000000_cfg2.4-0.0-0.73-perclassTrue.npz to per-class-sample/sit-ldm-e2e-vavae_4000000_cfg2.4-0.0-0.73-labelsampling-equal.npz
Rename per-class-sample/sit-ldm-e2e-vavae_4000000_cfg1.0-0.0-1.0-perclassTrue.npz to per-class-sample/sit-ldm-e2e-vavae_4000000_cfg1.0-0.0-1.0-labelsampling-equal.npz
