Need help to locate the A1 training method described in timm paper #2134
Unanswered
SubrangshuDas
asked this question in
Q&A
Replies: 0 comments
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Uh oh!
There was an error while loading. Please reload this page.
-
From the link https://paperswithcode.com/paper/resnet-strikes-back-an-improved-training#code, CIFAR10 accuracy using A1 method (described in the paper) is 98.3%.
I want to reproduce the results, so that it gives me a good baseline to train on newer datasets using transfer learning.
num_classes = 10
net = timm.create_model('resnet50', pretrained=True, num_classes=num_classes).to(device)
However, I could not locate the python code to incrementally train the above model for CIFAR10 dataset using A1 training.
Beta Was this translation helpful? Give feedback.
All reactions