lucidrains byol-pytorch issues

lucidrains / byol-pytorch

Usable Implementation of "Bootstrap Your Own Latent" self-supervised learning, from Deepmind, in Pytorch

MIT License

1.76k stars 248 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

how to inference to get embedding with this code？

#43 sunruina2 opened 3 years ago
1
Whether this code is used for self-supervised learning?

#42 Epiphqny closed 3 years ago
2
Use Torchvision augmentations instead of Kornia, which work with TPUs

#41 umbertov closed 3 years ago
1
example code error with multi-gpu?

#40 sunruina2 opened 3 years ago
0
Training loss decreased and then increased

#39 easonyang1996 closed 3 years ago
4
Can't load ckpt

#38 AndrewTal closed 3 years ago
4
CKPT load

#37 ansleepykid opened 3 years ago
1
Error when running the model with DDP

#36 SuhongMoon closed 4 years ago
1
Example Code Parameters

#35 JosephDiPalma closed 4 years ago
1
Error when training my own 1D dataset

#34 QueeneTam closed 4 years ago
2
TPU/colab

#33 whyrv opened 4 years ago
4
The saved network is same as the initial one?

#32 KimMeen closed 4 years ago
3
BYOL uses different augmentations for view1 and view2

#31 OlivierDehaene closed 4 years ago
5
Add device

#30 FrancescoSaverioZuppichini closed 4 years ago
4
Did you try use apex for learner ?

#29 Begin19 opened 4 years ago
4
Feat: Gradient accumulation

#28 YoniSchirris closed 4 years ago
1
how to run the example on gpu

#27 likelyzhao closed 4 years ago
1
the maximum batch size can only be set to 32

#26 cuixianheng closed 4 years ago
3
the optimizer

#25 cuixianheng closed 4 years ago
1
the error of the runing exampe code when i use 2 gpus

#24 cuixianheng closed 4 years ago
2
Just wondering - DataLoader does not shuffle in the example

#23 daisukelab closed 4 years ago
1
kornia 0.4.0 - torch 1.6.0 releases

#22 KeremTurgutlu closed 4 years ago
1
Lightning example: double augmentation

#21 sorenrasmussenai closed 4 years ago
2
Fix error when using hidden_layer=-1

#20 yieldthought closed 4 years ago
1
singleton question

#19 EelcoHoogendoorn closed 4 years ago
1
question about projection head and predictor

#18 iChenning closed 4 years ago
0
Transferring results on Cifar and other datasets

#17 jacobswan1 closed 4 years ago
4
Pretrained network

#16 pmorerio closed 4 years ago
3
Error when running the example code

#15 jlindsey15 opened 4 years ago
11
Add normalization params as class init params

#14 YoniSchirris closed 4 years ago
4
RuntimeError: Only Tensors created explicitly by the user (graph leaves) support the deepcopy protocol at the moment

#13 vumichien closed 4 years ago
2
Representation Collapse

#12 arjung128 closed 4 years ago
2
Do you provide a quick start codebase?

#11 ilikeokoge closed 4 years ago
0
Kornia transforms

#10 reactivetype closed 4 years ago
1
AssertionError: hidden layer never emitted an output with multi-gpu training

#9 reactivetype closed 4 years ago
7
How to save the entire network including the projection?

#8 albert-92 closed 4 years ago
2
[#6] make loss positive

#7 NaxAlpha closed 4 years ago
0
Negative Loss, Transfer Learning/Fine-Tuning Question

#6 rsomani95 closed 4 years ago
13
Singleton Class Members

#5 wentaoyuan closed 4 years ago
3
Feat: Add image dataset normalization

#4 YoniSchirris closed 4 years ago
2
Feat: Performance tests

#3 YoniSchirris opened 4 years ago
9
Distributed training support

#2 naveenkumarmarri closed 4 years ago
2
Memory leak?

#1 jlindsey15 closed 4 years ago
1