532 karma · joined February 2, 2015
Tweet: https://twitter.com/psuraj28/status/1580640841583902720 Blog announcement: https://huggingface.co/blog/stable_diffusion_jax
Open Source repo: https://github.com/huggingface/huggingface_hub
It first uses a simple CNN to get one of the 100 classes. It then uses that class to retrieve images using CLIP. The original CLIP Space that helped for inspiration is https://huggingface.co/spaces/vivien/clip.
- Transformers
- Datasets
- Tokenizers
- Accelerate
- Model Hub
They also plan on hosting live office hours and having study groups.
- CLIP from OpenAI, Image-Text similarity or Zero-Shot Image classification
- ViT from GoogleAI
- DeiT from facebookai, SOTA Image Classification
And you can try ViT/DeiT on the hub: https://huggingface.co/google/vit-base-patch16-224