https://github.com/aquatiko/dog-vs-cat-redux-kernel-edition-transfer-learning

Top 5% on Kaggle leaderboard using fast.ai library and resnet50 along with transfer learning.
https://github.com/aquatiko/dog-vs-cat-redux-kernel-edition-transfer-learning

cats-vs-dogs data-augmentation fastai gradient-descent-with-restarts image-classification resnet-50 resnet101 resnet34 transfer-learning

Last synced: 7 months ago
JSON representation

Top 5% on Kaggle leaderboard using fast.ai library and resnet50 along with transfer learning.

• Host: GitHub
• URL: https://github.com/aquatiko/dog-vs-cat-redux-kernel-edition-transfer-learning
• Owner: aquatiko
• Created: 2018-08-24T09:53:24.000Z (about 7 years ago)
• Default Branch: master
• Last Pushed: 2018-08-24T10:08:26.000Z (about 7 years ago)
• Last Synced: 2025-01-21T00:32:20.771Z (9 months ago)
• Topics: cats-vs-dogs, data-augmentation, fastai, gradient-descent-with-restarts, image-classification, resnet-50, resnet101, resnet34, transfer-learning
• Language: Jupyter Notebook
• Homepage:
• Size: 71.3 KB
• Stars: 1
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md

Awesome Lists containing this project

README

          
# Dogs v/s Cats Redux: Kernel Edition- Top 5% Transfer Learning

Transfer Learning approach using fast.ai library which makes implementing it easier. Based on 3 different approaches each with architectures- resnet34, resnet50 and resnet101... got top 5% on Kaggle leaderboard, Accuracy 99.3% and and 0.05605 binary log loss error(evaluation criteria).

Used  Diffrential Learning Rates to tune arch , Test Time Augmentation and Learning Rate Anneling to improve model loss.