https://github.com/oguiza/fastai_extensions

Code you can use jointly with fastai
https://github.com/oguiza/fastai_extensions

Last synced: 7 days ago
JSON representation

Code you can use jointly with fastai

• Host: GitHub
• URL: https://github.com/oguiza/fastai_extensions
• Owner: oguiza
• Created: 2019-07-07T06:28:52.000Z (over 5 years ago)
• Default Branch: master
• Last Pushed: 2020-11-30T18:09:01.000Z (almost 4 years ago)
• Last Synced: 2024-08-02T15:20:44.839Z (3 months ago)
• Language: Jupyter Notebook
• Homepage:
• Size: 22.4 MB
• Stars: 92
• Watchers: 2
• Forks: 7
• Open Issues: 3
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

        
# fastai_extensions

## If you are interested in times series and fastai you may want to visit the new time series library [`tsai`](https://github.com/timeseriesAI/tsai). It supports Pytorch 1.7 & fastai v2.

**0_data_augmentation_notebook.ipynb**

* Includes a version of *ricap* and a modified version of *cutmix* (that fixes a bug in the lambda calculation).

* It also includes 2 functions to visualize the impact of single-image transforms (*show_single_img_tfms*) or multi-image transforms (*show_multi_img_tfms*). 

* To use the code you can either clone the repo, or copy the exp/nb_new_data_augmentation.py code into your code.