-
Hi, a great work!I have run it with my own dataset without pretrain. But the accuracy is low.
Could you provide the pretrained ResNet2d3d on imagenet, or some pretrained models on other datasets? Th…
-
## 🐛 Bug
ColorJitter is supposed to be called on an image of type PIL or Tensor, but can only be called on images of type PIL.
## To Reproduce
Steps to reproduce the behavior:
1. Load the…
-
Will #Keras work with Python 3.6? I'm really tired of downgrading my different versions of Python at this point.
-
Instead of
```
def forward(self, output1, output2, label):
euclidean_distance = F.pairwise_distance(output1, output2)
loss_contrastive = torch.mean((1-label) * torch.pow(euclidean…
ghost updated
4 years ago
-
Running apex with `allennlp train configs/contrastive.jsonnet -s tmp --include-package t2t -o "{"trainer": {"opt_level": 'O1'}}"` returns exceptions as following:
```
Traceback (most recent call l…
-
Does anyone have the link of Chinese Baidu disk or thunder of audio_feature.h5? I can only download it with Google browser. Because it's too big, I fail every time.
-
I am not getting any improvement from contraGAN at all. For the main contribution, the FID improvement is not even 0.1. If i rerun FID score, I get worse score many times. Same for IS. other metrics n…
-
https://arxiv.org/pdf/2006.07733.pdf
BYOL
from: simclrv2 paper
-
Hi @sthalles , Thank you for sharing your code!
Pl correct me if I am wrong:
I see that in line **loss/nt_xent.py line 57 (below)** you are not computing contrastive loss for all negative pairs as…
-
**🚀 Feature**
Reverse the history of transforms applied to a subject.
**Motivation**
It would be great able to reverse the transforms of a subject especially for spatial transforms.
The idea i…
stity updated
3 years ago