Issues
- 0
cannot reproduce the performance of FEAT
#82 opened by Duuhiy - 4
About the accuracy of Protonet.
#81 opened by Glutaredoxin - 2
question about pre training and feat model
#80 opened by dota2mhxy - 2
fewshot model
#79 opened by Catofblue - 1
- 1
- 1
Pretraining hyperparameters for CUB dataset
#76 opened by mayug - 1
- 1
Some question about CTM
#75 opened by WoshiBoluo - 0
- 1
DDP available?
#72 opened by brando90 - 1
Question about proto
#71 opened by LoveMiki - 1
Download the pretrained weights
#70 opened by OverVan - 1
有关transductive FSL 的设置
#68 opened by DongYY127 - 1
关于位置编码positional encoding
#65 opened by YiX98 - 1
We can reproduce all numbers except for FEAT
#73 opened by hieulem - 2
The ProtoNet achieves the best Acc just after 1 epoch based on the given pre-trained ResNet12 weights.
#45 opened by JimZAI - 3
Key and value intuition of transformer in FEAT
#66 opened by kumam92 - 2
- 3
Question about the cosine distance.
#62 opened by z1358 - 4
Question about lr and gamma.
#67 opened by zhanyuanyang - 1
Confusion about the contrastive learning.
#63 opened by z1358 - 1
有关Pre-Avg和Post_Avg
#61 opened by sherrijj - 1
- 3
请问关于提高准确率的建议
#57 opened by YiX98 - 3
- 4
- 5
- 0
Confusion matrix plot for testing episodes
#58 opened by kumarmanas - 0
用自己的数据集训练FEAT报错
#56 opened by YiX98 - 2
Hyperparameter training command for protonet with WRN backbone on miniImagenet dataset
#54 opened by techNN - 1
- 4
About the 5-shot results on miniimagenet
#50 opened by zhanyuanyang - 0
- 2
关于feat.py模型内部关于输入数据shape的问题
#51 opened by shuowang666 - 1
请问可否提供WRN作backbone时fine-tune阶段训练的相关参数
#49 opened by SnowyJune973 - 1
预训练的结果没有达到您公布模型的准确率
#48 opened by YuanMLer - 1
- 5
- 3
- 5
About the results you reported in the paper
#33 opened by molybdenumyz - 0
label problem for meta-training stage
#46 opened by freeman-1995 - 2
no avg_pooling in resnet-12?
#44 opened by Tsingularity - 1
Question the weight of pre-training
#39 opened by blue-blue272 - 4
- 1
should I use the cache?
#41 opened by seyeeet - 1
a question about resnet-18/12 kernel size
#34 opened by Sword-keeper - 1
- 2
a question about trainning error
#36 opened by wongsihan - 0
how can I test the model using pretrained models?
#38 opened by seyeeet