---
license: apache-2.0
---
uLLaVA-7B-v1 (without LoRA) Model Card

Model date: uLLaVA-v1-7B was trained in August 2023.
Paper or resources for more information: https://github.com/OPPOMKLab/u-LLaVA


Usage: Set lora_r = -1 when inference.


Training dataset
- LLaVA
- TGIF
- RefCOCO(g/+)
- ADE20K