Over 400,000 human preference responses for evaluating the Flux 2 Pro text-to-image model, collected in less than seven hours via the Rapidata Python API. The dataset was created by Rapidata and last updated on December 2, 2025. It includes evaluations across preference, coherence, and alignment categories.
Use Cases
- Benchmarking text-to-image model performance based on human preference scores.
- Analyzing model alignment with human expectations based on the described evaluation categories.
- Training reward models or fine-tuning generative models based on large-scale human feedback data.
- Studying annotator behavior and consensus in rapid, large-scale data collection scenarios.
Strengths
- Large scale with over 400,000 human responses.
- Diverse annotator pool of over 50,000 individuals.
- Rapid collection timeframe of less than 7 hours, suggesting efficient methodology.
Limitations
- Column-level documentation is absent; field semantics must be inferred after download.
- Row count is unknown, which may limit suitability assessment.
- Description metadata is limited; actual data quality requires manual inspection after download.
Provenance
- Source
- Rapidata
- Collection Method
- Collected via the Rapidata Python API from over 50,000 annotators.
- Time Range
- Dataset reflects evaluations of the Flux 2 Pro model version from November 25, 2025.
- Freshness
- Last updated 2025-12-02 12:54:30; freshness should be verified.
- Geography
- null