orca
orca copied to clipboard
snap-stanford
[ICLR 2022] Open-World Semi-Supervised Learning
Could you please share a download link for ImageNet-100? The classes you're using seem different from those on the ImageNet-100 list on [this Kaggle page](https://www.kaggle.com/datasets/ambityga/imagenet100)
May I ask why there will be an error when running cifar10? (ValueError: zero-size array to reduction operation maximum which has no identity) However, cifar100 can run smoothly
I excluded the pretraining module in the code for running this kind of dataset. Please suggest what modification I need to perform in the existing code.
Hello, your work is excellent, but the project you gave does not contain SimCLR pre-training files. Do you have relevant parameter settings or pre-training files for SimCLR?
Hi, Due to only one GPU, I cannot understand why the batch size can be set to a large number i.e., 512. If the number of classes is 1000, it...
We (not the authors) have read the article carefully and explained the main process and content in Chinese. You can visit https://blog.csdn.net/beginner1207/article/details/129174534 to find it. I hope this will be...
Excellent work! Could you share the traning log in ImageNet? Thank you very much！
Hi, thank you for sharing the work and your code. I was wondering while calculating the unseen accuracy, if there should also be a filter on the predictions not being...
Hello, I have a question that why here it uses the prob2 instead of prob during train function. Look forward to your reply,thanks! `pos_prob = prob2[pos_pairs, :] pos_sim = torch.bmm(prob.view(args.batch_size,...
Hi, thank you for the great work and for publishing the code! I have some questions about the parameter settings in SimCLR pretraining. Can you provide the pretraining codes or...
Metadata
Owner
Metadata
[ICLR 2022] Open-World Semi-Supervised Learning