Dataset Viewer
The dataset could not be loaded because the splits use different data file formats, which is not supported. Read more about the splits configuration. Click for more details.
Couldn't infer the same data file format for all splits. Got {NamedSplit('train'): ('json', {}), NamedSplit('validation'): (None, {}), NamedSplit('test'): (None, {})}
Error code:   FileFormatMismatchBetweenSplitsError

Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.

RECAP reproducibility

Public backup hf_backup_20261102, captured 2026-11-02T11:42:26.136798+08:00. Collection: https://huggingface.co/collections/myx160/recap-6ae7fe8e46603b646d677fbd

Snapshot status: BACKUP_COMPLETE.json is published only after every planned file has been committed and its remote size/hash verified. The local long-running training continues independently. This snapshot includes only files available at capture time, not future training updates.

Restore by combining the original project-relative paths from all three repos: myx160/recap-data, myx160/recap-checkpoints, myx160/recap-reproducibility. Restore symbolic links from SNAPSHOT.json. Restore .gitignore from _restore/original_gitignore.txt. Credentials, caches, tmp, unfinished downloads, lock/PID files and binary Conda environment archives are excluded. Environment locks and experiment environment records are included. Frozen old paths are resolved by the included RECAP relocation helper. Some text credential patterns and signed URL queries may be redacted; see snapshot records.

Both successes and failures, pilot and historical exploratory results are kept. Training completion, loss changes, or an upstream score do not establish policy improvement. Test manifests remain sealed for evaluation use.

See LICENSES.md, NOTICE and upstream files for component licenses.

Downloads last month
997

Collection including myx160/recap-reproducibility