A image classification model for the Imagenette (Imagenette dataset) dataset. At the 526th epoch, we achieved an accuracy of 94.5223%.
Training Model.py are codes for training our model. Before running the codes, you need to first download the Imagenette dataset (Imagenette dataset), download and install Ranger optimizer (Ranger optimizer), and change the paths in the codes into your corresponding local path.
Testing Model.py are codes for testing our model. Before running the codes, you need to first download the Imagenette dataset (Imagenette dataset), and change the paths in the codes into your corresponding local path.
ResNet18 0.945223.pth are the parameters of our model.
ResNet18 0.945223.pth exceeds GitHub’s 100 MB file size limit and cannot be uploaded to GitHub. You can download it at: https://disk.pku.edu.cn:443/link/4EF094BBCDA3766844A0891C95C85957 (Valid Until: 2026-08-25 23:59).
Experiment Report.pdf are the experiment report of my homework.