Dataset category
Human-labeled and RLHF data
Annotations, preference rankings and expert-written demonstrations.
Data produced or judged by people: classification labels, preference comparisons, instruction and response pairs, and expert-written demonstrations. Request a sample to discuss task design and quality controls.
- modality
- text, image, audio
- format
- JSONL, CSV
- license
- PLACEHOLDER: license terms
- size
- PLACEHOLDER: size / volume range
- example uses
- Reward model training
- Instruction tuning
- Evaluation and red-team sets