The dataset is currently empty. Upload or create new data files. Then, you will be able to explore them in the Dataset Viewer.
Annoy-Alpaca-Raw
This dataset contains the raw instruction records used by the Annoy-Alpaca pipeline.
The records were adopted directly from the Stanford Alpaca dataset. Use that direct source when determining the license for this raw release.
Each JSON record contains instruction, input, and output fields. Empty input values are retained to preserve the source schema.
Related resources: reasoning dataset | project repository
- Downloads last month
- 6