EleutherAI may be the most famous AI open research collective. Lots of great work has been released by EleutherAI, such as the Pile dataset, GPT-J, GPT-NeoX-20B, and VQGAN-CLIP.
This server focuses on developing new & replicating existing multimodal models and creating datasets to support these efforts. They have released the LAION-400M and LAION-5B datasets and trained their own CLIP models
This server is focused on BioML research projects and just recently launched, and projects are just getting started, which means it's the perfect time to get involved!
This is a new research organization focused on preference learning and contrastive learning, with various applications from story generation to code generation and even architecture design generation.
Applying deep learning to pathology is quite challenging due to the sheer size of the slide images (gigapixels!).
A common approach is to divide images into smaller patches, for which deep learning features can be extracted & aggregated to provide a slide-level diagnosis (1/9)
Unfortunately, dividing into small patches limits the context to cellular features, missing out on the various levels of relevant features, like larger-scale tissue organization. (2/9)
Additionally, it is difficult to improve long-range dependencies with Transformers due to the high number of patches which makes attention calculation computationally difficult. (3/9)
You may have seen surreal and absurd AI-generated images like these ones...
These are all generated with an AI tool known as DALL·E mini
Let's talk about the history of #dallemini, and also *how* it works! ↓↓↓🧵
First, let's clarify the different AI tools which many get confused about:
- DALL·E was an @OpenAI-developed AI project from Jan 2021
- DALL·E mini is a community-created project inspired by DALL·E
- DALL·E 2 is another @OpenAI-developed tool released in April (2/16)
@OpenAI DALL·E mini was actually originally developed about a year ago, back in July 2021.
During a programming competition organized by @huggingface (an AI company), @borisdayma & some community folks (including myself!) developed a neural network inspired by DALL·E & studied it (3/16)
Awesome and surprising things you can do with Jupyter Notebooks ⬇
1. Write a full-fledged Python library!
You can write all of your code, documentation, & tests with Jupyter Notebooks & nbdev.fast.ai, all while maintaining best software practices and implementing CI/CD!
fastai deep learning library is entirely written in notebooks!
2. Create a blog!
Platforms like fastpages.fast.ai easily allow you to create blog posts from your Jupyter Notebooks, with the code cells and outputs in your post, and can even be made interactive.
The model is based on an autoregressive transformer (like DALL·E) combined with a VQGAN but utilizes several key tricks to improve the quality and also controllability of the generations. 2/10
One trick is the use of a segmentation map (referred to as a scene) and a VQGAN for the scene.
As you can see here, this provides more controllability to the generation process. 3/10
The first step is to explore the data, also known as EDA. Getting a feel for the data is important to be able to derive important insights that can help you. 👨💻
After exploring the data, the next step is to make a baseline solution. In this case, I had put together a quick pretrained baseline based on @Nils_Reimers's SentenceTransformers: