File size: 1,755 Bytes
bbead19 f50f69d bbead19 f50f69d |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 |
---
license: apache-2.0
inference: false
---
**NOTE: This "delta model" cannot be used directly.**
Users have to apply it on top of the original LLaMA weights to get actual LLaVA weights.
See https://github.com/haotian-liu/LLaVA#llava-weights for instructions.
<br>
<br>
# LLaVA Model Card
## Model details
**Model type:**
LLaVA is an open-source chatbot trained by fine-tuning LLaMA/Vicuna on GPT-generated multimodal instruction-following data.
It is an auto-regressive language model, based on the transformer architecture.
**Model date:**
LLaVA was trained in April 2023.
**Paper or resources for more information:**
https://llava-vl.github.io/
**License:**
Apache License 2.0
**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
595K filtered image-text pairs from CC3M.
150K 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.
|