File size: 1,951 Bytes
c3be39e
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
**Please read the following information carefully before proceeding.**

OpenFlamingo is a **research prototype** that aims to enable users to interact with AI through both language and images. AI agents equipped with both language and visual understanding can be useful on a larger variety of tasks compared to models that communicate solely via language. By releasing an open-source research prototype, we hope to help the research community better understand the risks and limitations of modern visual-language AI models and accelerate the development of safer and more reliable methods.

- [ ] I understand that OpenFlamingo is a research prototype and I will only use it for non-commercial research purposes.

**Limitations.** OpenFlamingo is built on top of the LLaMA large language model developed by Meta AI. Large language models, including LLaMA, are trained on mostly unfiltered internet data, and have been shown to be able to produce toxic, unethical, inaccurate, and harmful content. On top of this, OpenFlamingo’s ability to support visual inputs creates additional risks, since it can be used in a wider variety of applications; image+text models may carry additional risks specific to multimodality. Please use discretion when assessing the accuracy or appropriateness of the model’s outputs, and be mindful before sharing its results.

- [ ] I understand that OpenFlamingo may produce unintended, inappropriate, offensive, and/or inaccurate results. I agree to take full responsibility for any use of the OpenFlamingo outputs that I generate.

**Privacy and data collection.** This demo does NOT store any personal information on its users, and it does NOT store user queries.

**Licensing.** As OpenFlamingo is built on top of the LLaMA large language model from Meta AI, the LLaMA license agreement (as documented in the Meta request form) also applies.

- [ ] I have read and agree to the terms of the LLaMA license agreement.