How to use this model on Tiny ImageNet for classification
Closed this issue · 1 comments
tngok-wustl commented
I am a computer visions student who is thinking about using your WaveMix model for my final project. I am planning to use it on the Tiny ImageNet dataset for classification and model evaluation. However, I have to run all my deep learning code on Google Colab, since I do not have an independent and powerful GPU.
So, I have a few questions regarding using the model:
- How can I load the Tiny ImageNet dataset in Google Colab? Can you give a python code for it?
- The Readme said that WaveMix reduces the use of GPU resources. Is my T4 GPU on Google Colab enough for training the Tiny ImageNet?
- Where is the code for Tiny ImageNet classification? Is there a pre-trained model for it?
- In case I want to retrain the model, where is the code for training the model?
pranavphoenix commented
- I have a repository called tinyimagenet dataloader which will work in colab.
- Yes
- Yes, the pre-trained weights are available in https://github.com/pranavphoenix/WaveMix/tree/main/Saved_weights/tinyimagenet
- https://github.com/pranavphoenix/WaveMix/blob/main/Image_Classification/tinyimagenet.py