liuhaotian's picture
Update README.md
d94ccf0
|
raw
history blame
1.86 kB
---
license: cc-by-nc-sa-4.0
inference: false
---
**NOTE: This is a research preview of the LLaVA-Lightning based on MPT-7B-chat checkpoint. The usage of the model should comply with MPT-7B-chat license and agreements.**
**NOTE: Unlike other LLaVA models, this model can (should) be used directly without delta weights conversion!**
<br>
<br>
# LLaVA Model Card
## Model details
**Model type:**
LLaVA is an open-source chatbot trained by fine-tuning LLaMA/Vicuna/MPT on GPT-generated multimodal instruction-following data.
It is an auto-regressive language model, based on the transformer architecture.
**Model date:**
LLaVA-Lightning-MPT was trained in May 2023.
**Paper or resources for more information:**
https://llava-vl.github.io/
**License:**
CC-BY-NC-SA 4.0
**Where to send questions or comments about the model:**
https://github.com/haotian-liu/LLaVA/issues
## Intended use
**Primary intended uses:**
The primary use of LLaVA is research on large multimodal models and chatbots.
**Primary intended users:**
The primary intended users of the model are researchers and hobbyists in computer vision, natural language processing, machine learning, and artificial intelligence.
## Training dataset
558K filtered image-text pairs from LAION/CC/SBU, captioned by BLIP.
80K GPT-generated multimodal instruction-following data.
## Evaluation dataset
A preliminary evaluation of the model quality is conducted by creating a set of 90 visual reasoning questions from 30 unique images randomly sampled from COCO val 2014 and each is associated with three types of questions: conversational, detailed description, and complex reasoning. We utilize GPT-4 to judge the model outputs.
We also evaluate our model on the ScienceQA dataset. Our synergy with GPT-4 sets a new state-of-the-art on the dataset.
See https://llava-vl.github.io/ for more details.