FixMatch-pytorch-Google Colab

Unofficial Google Colab pytorch implementation for "FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence," NeurIPS'20.
This code is implemented using Google Colab Free GPU and "Apparel images dataset" is used in this implementation.
This implementation is done for Multiclass Classification, Multi Label Classification and Multi output Classification.

Requirements

• timm
• kaggle
• opencv-python-headless==4.1.2.30
• wandb

Results: Classification Accuracy (%)

"Multiclass Classification"

"Multi Label Classification"

"Multi output Classification"

Authored by:

Rakib Hossain Rifat

Machine Learning Engineer

Polyfins Technology Inc.