FoundationVision/Groma
Groma is a multimodal LLM that uses localized visual tokenization to enable region-level understanding and visual grounding capabilities.
Not currently ranked — collecting fresh signals.
star history
Groma is a grounded multimodal large language model that introduces visual tokenization for localization, allowing it to process user-defined region inputs (bounding boxes) and generate responses grounded to specific visual regions. It achieves state-of-the-art performance on referring expression comprehension benchmarks like RefCOCO, RefCOCO+, and RefCOCOg. The model is based on LLaMA architecture extended with multimodal capabilities for vision-language understanding and localization.
Frequently asked
- What is FoundationVision/Groma?
- Groma is a multimodal LLM that uses localized visual tokenization to enable region-level understanding and visual grounding capabilities.
- Is Groma open source?
- Yes — FoundationVision/Groma is open source, released under the Apache-2.0 license.
- What language is Groma written in?
- FoundationVision/Groma is primarily written in Python.
- How popular is Groma?
- FoundationVision/Groma has 585 stars on GitHub.
- Where can I find Groma?
- FoundationVision/Groma is on GitHub at https://github.com/FoundationVision/Groma.