kyutai/Babillage
Babillage Babillage is a multimodal benchmark dataset introduced along with MoshiVis (Project Page | arXiv), containing three common vision-language benchmarks converted in spoken form, for the evaluation of Vision Speech Models. For each benchmark (COCO-Captions, OCR-VQA, VQAv2), we first reformat the text question-answer pairs into a more conversational dialogue, and then convert them using a text-to-speech pipeline, using a consistent synthetic voice for the answer… See the full description on the dataset page: https://huggingface.co/datasets/kyutai/Babillage.
| Author / organization | kyutai |
|---|---|
| License (as declared) | cc-by-4.0 |
| Downloads | 620 (as reported by Hugging Face) |
| Likes | 14 |
| Created | 16 March 2025 |
| Last updated at source | 21 March 2025 |
Popularity, last 3 days
620 now (+70 since 2026-10-16). Counts as reported by the source, snapshotted daily by Silytics.
Recent changes
- 17 Oct 2026 Description: Babillage Babillage is a multimodal benchmark dataset introduced along with MoshiVis (Project Page | arXiv), containing three common vision-language benchmarks converted in spoken form, for the evaluation of Vision Speech Models. For each b → Babillage Babillage is a multimodal benchmark dataset introduced along with MoshiVis (Project Page | arXiv), containing three common vision-language benchmarks converted in spoken form, for the evaluation of Vision Speech Models. For each b
Changes detected by Silytics between refreshes of the source.
More from kyutai
Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.