AI BriefingNVIDIA Releases Nemotron-3.5-Embed-1B-BF16 Multimodal Embedding Model
Key point
The model supports multimodal retrieval across 34 languages for text and 18 for images, with open commercial use rights.
Details
NVIDIA has released Nemotron-3.5-Embed-1B-BF16, a versatile embedding model designed for multimodal and multilingual question-answering retrieval. The model processes documents as images, text, or combined image–text inputs, enabling retrieval based on user text queries.
Multimodal and Multilingual Capabilities
The model handles complex document structures, including page images containing text, tables, charts, and infographics. Its multilingual support is extensive:
- Text Retrieval: Evaluated across 34 languages, including English, Arabic, Chinese, Hindi, Japanese, Korean, and Spanish.
- Image Retrieval: Evaluated across 18 languages, including Arabic, Bengali, Chinese, English, French, German, Hindi, Japanese, Korean, Russian, Spanish, and Yoruba.
Licensing and Availability
The model is available on Hugging Face and is licensed for both commercial and non-commercial use. While the evaluation sets cover specific languages, the model may support additional languages beyond those tested.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
NVIDIA Run:ai Platform Open-Sourced as Part of DSX AI Factory