facebookresearch/moco-v3
PyTorch implementation of MoCo v3 https//arxiv.org/abs/2104.02057
PythonNOASSERTION
Issues
- 1
Cann't load pretrain weights
#30 opened by EvilPsyCHo - 1
- 6
- 0
Does mocov3 support Coco datasets
#41 opened by gitillbr - 0
About Linear Probe Accuracy of Resenet-50
#40 opened by avinabsaha - 0
Most of the learning time is spent loading data. This makes it impossible to use GPU resources efficiently.
#39 opened by JINSUBY - 1
Links for pre-trained models are broken
#38 opened by toliz - 0
- 6
KNN curve code
#5 opened by kashif - 0
How to fine-tune?
#35 opened by DWCTOD - 1
About the learning rate for resnet-50
#34 opened by cswaynecool - 0
Question about linear probe
#33 opened by HHenryD - 0
question about batch size
#32 opened by CoinCheung - 5
How about the loss converges during training?
#23 opened by Vickeyhw - 0
- 0
Training with multi-crop
#28 opened by gyfastas - 3
- 2
Checkpoint release
#10 opened by JACKHAHA363 - 0
ViT-Base fine-tuned checkpoints
#27 opened by eliahuhorwitz - 2
队列
#26 opened by onlinehuazai - 1
Fine-tuning vs Linear probing
#21 opened by zhangdan8962 - 1
How many TPUs ?
#22 opened by jrabary - 1
- 1
- 7
Tensorflow version
#8 opened by wildphoton - 1
End-to-end fine-tuning epochs for ViT-S
#19 opened by Yingdong-Hu - 1
- 4
- 1
- 0
main.py missing in transfer directory
#13 opened by yxchng - 1
Transfer learning performance of MoCo v3 on more challenging downstream dense prediction tasks.
#12 opened by Yuxin-CV - 1
- 1
ViT-Small model number of heads in attention differs from the original paper
#7 opened by YellowPig-zp - 1
release pretrained weights
#6 opened by maliho0803