---
license: apache-2.0
language:
- en
metrics:
- accuracy
pipeline_tag: image-text-to-text
---

We use the powerfull [TinyLLaVA Factory](https://github.com/TinyLLaVA/TinyLLaVA_Factory) to create a super small image-text-to-text model with only 296M params.

The goal is to make it possible to run LLaVA models on edge devices (with few gigabytes of memory).

For LLM and vision tower, we choose [OpenELM-270M-Instruct](apple/OpenELM-270M-Instruct) and [facebook/dinov2-small](facebook/dinov2-small), respectively.