📦 EqualifyEverything / benchmarks-ai-alt

Benchmark testing the accuracy of AI models at generating image alternative text.

benchmarks-ai-alt / projects / corpus-construction / tools / fixtures
📂 e14ac5270765edb4a4994d41d60d3baa37a1acc0
📁 ..
📄 invalid-items.jsonl
📄 valid-item.jsonl
📄 valid-review.jsonl