localhorst
/
CLIP_on_Tesla_K20Xm
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Beautify the tags

This commit is contained in:
Erdogan Taskesen 2021-02-16 14:00:58 +01:00 committed by GitHub
parent 4c0275784d
commit d9f3c4bfe4
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 7 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

9
README.md
View File

@ -1,6 +1,11 @@
# CLIP
[[Blog]](https://openai.com/blog/clip/) [[Paper]](https://cdn.openai.com/papers/Learning_Transferable_Visual_Models_From_Natural_Language_Supervision.pdf) [[Model Card]](model-card.md) [[Colab]](https://colab.research.google.com/github/openai/clip/blob/master/notebooks/Interacting_with_CLIP.ipynb)
[![License](https://img.shields.io/badge/license-MIT-green.svg)](https://github.com/openai/CLIP/blob/master/LICENSE)
[![Open In Colab](https://colab.research.google.com/assets/colab-badge.svg)](https://colab.research.google.com/github/openai/clip/blob/master/notebooks/Interacting_with_CLIP.ipynb)
[![Blog](https://img.shields.io/badge/Blog-Docs-Green.svg)](https://openai.com/blog/clip/)
[![Paper](https://img.shields.io/badge/openai-Docs-Green)](https://cdn.openai.com/papers/Learning_Transferable_Visual_Models_From_Natural_Language_Supervision.pdf)
[![Github Forks](https://img.shields.io/github/forks/openai/clip.svg)](https://github.com/openai/CLIP/network)
[![GitHub Open Issues](https://img.shields.io/github/issues/openai/clip.svg)](https://github.com/openai/CLIP/issues)
[![Model Card](https://img.shields.io/badge/Model-Card-green.svg)](https://github.com/openai/CLIP/blob/master/model-card.md)
CLIP (Contrastive Language-Image Pre-Training) is a neural network trained on a variety of (image, text) pairs. It can be instructed in natural language to predict the most relevant text snippet, given an image, without directly optimizing for the task, similarly to the zero-shot capabilities of GPT-2 and 3. We found CLIP matches the performance of the original ResNet50 on ImageNet “zero-shot” without using any of the original 1.28M labeled examples, overcoming several major challenges in computer vision.