issues
search
bubbliiiing
/
Siamese-pytorch
这是一个孪生神经网络（Siamese network）的库，可进行图片的相似性比较。
MIT License
598
stars
127
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
您好，我发现您的网络最后一层没有使用sigmoid，但是在计算accuracy的时候却使用了sigmoid，请问这是为什么？
#58
CharileBrown
opened
2 days ago
0
貌似发现了一个bug！！！！
#57
FocueAI
closed
2 months ago
1
请问下训练的时候没有报错，但是只执行一个epoch就停了，是数据集不够吗还是什么原因呢？
#56
Hongtauo
opened
3 months ago
0
我注意到你所使用的loss不是传统的三元组loss，所以使用这样的loss的好处在什么地方呢？
#55
ML-learning-git
opened
8 months ago
0
预测速度有些慢
#54
lawliet1777
opened
8 months ago
0
ModuleNotFoundError: No module named 'torchvision.models.utils'
#53
valucn
opened
11 months ago
1
为什么训练后的trainloss始终比valloss大，trainacc始终比valacc小，没有改过代码
#52
jjjjjjj672
opened
11 months ago
0
loss稳定在0.693左右，acc在0.5左右波动
#51
jindameias
opened
11 months ago
0
loss问题
#50
jjjjjjj672
opened
11 months ago
0
为什么我训练自己的数据集的时候显示在siamese.py的第30行运行不过去？
#49
EhanDong
opened
11 months ago
0
发现有张图片是二值化后的黑白图片 效果很不好，不清楚为什么
#48
sunxhap
opened
1 year ago
0
咨询下，训练过程什么时候可以结束 或者说代码里面怎么设置让他自动结束
#47
sunxhap
opened
1 year ago
0
冻结和解冻是那一部分代码？
#46
yltqy
opened
1 year ago
0
up这个新仓库的代码是不是没有freeze和unfreeze两个训练过程啊
#45
yltqy
opened
1 year ago
0
Bump pillow from 8.2.0 to 10.0.1
#44
dependabot[bot]
opened
1 year ago
0
训练Omniglot数据集时出现这个报错，ValueError: not enough values to unpack (expected 2, got 1)
#43
loongscale
opened
1 year ago
6
3060ti训练，为啥1.7it/s左右的速度，这个速度正常吗?
#42
LmingXie
opened
1 year ago
3
RuntimeWarning: Iterating over a tensor might cause the trace to be incorrect.
#41
TaoF02
opened
1 year ago
1
utils_aug这个文件是什么作用呢？
#40
DylTop
opened
1 year ago
1
Bump torch from 1.2.0 to 1.13.1
#39
dependabot[bot]
opened
1 year ago
0
dataloader SiameseDataset types 类别缺少最后一类 ？
#38
se202242
closed
1 year ago
3
Bump scipy from 1.2.1 to 1.10.0
#37
dependabot[bot]
opened
1 year ago
0
如何比较相似度
#36
HarrisonWells
opened
1 year ago
1
运行 Predict.py 报错
#35
cocoyamnut
closed
1 year ago
0
一直卡在validation那里
#34
hfdzlsw
closed
1 year ago
6
运行自己的数据集时,报错PIL.UnidentifiedImageError: cannot identify image file 'datasets/block\\images_background\\character02\\Thumbs.db'
#33
CTian-BUG
opened
1 year ago
5
相似性总是1.000
#32
wyf0422
opened
1 year ago
1
怎么对训练出来的模型转onnx，然后简化！！最主要还是简化，原模型转onnx在cpu耗时太长，需要简化。不知道怎么简化，用torch的简化会报错
#31
WFWFDSF24
opened
1 year ago
9
您好，请问网盘给出的Omniglot的预训练模型是直接执行代码训练得到的吗
#30
springrings
closed
1 year ago
0
怎么将VGG网络修改为resnet50来做特征提取？
#29
fighter-lee
opened
1 year ago
5
如果我的样本 一个分类只有两张图片应该怎么操作
#28
sunxhap
opened
1 year ago
1
为什么Cuda 设置True, 电脑CPU 使用100%， 没使用GPU 训练
#27
sunxhap
opened
1 year ago
1
样本图片大小不固定，input_shape应该怎么设置
#26
sunxhap
opened
1 year ago
1
请问如何验证
#25
feiqinaqian
opened
1 year ago
1
如果只有两个chapter会报下面图片的的错误
#24
sunmouren
opened
1 year ago
3
关于孪生网络的定量分析
#23
2962313771
opened
1 year ago
5
Bump pillow from 8.2.0 to 9.3.0
#22
dependabot[bot]
closed
1 year ago
1
能添加使用shufflenetv2作为backbone的例子么
#21
CsVeryLoveXieWenLi
opened
2 years ago
0
感觉SiameseDataset中有一点小问题，同时对损失函数有点不理解
#20
diaoa1900
closed
2 years ago
3
what's the input image shape?
#19
BeyondHeaven
closed
2 years ago
4
数据读取问题
#18
boris-lb
closed
2 years ago
2
some bug in dataloader about the self.types ?
#17
Peter-weng
opened
2 years ago
5
Bump numpy from 1.17.0 to 1.22.0
#16
dependabot[bot]
opened
2 years ago
0
请问有onnx部署的教程吗
#15
litianqi18
opened
2 years ago
1
dataloader文件中，135-138行是不是有一个问题
#14
chengyahui2
closed
2 years ago
0
Bump pillow from 8.2.0 to 9.0.1
#13
dependabot[bot]
closed
1 year ago
1
test git
#12
Zoezxb
opened
2 years ago
0
为啥输入的要求是(105,105,3),官方的是(224,224,3)啊，权重也是用得官方的，小白有点懵。
#11
ray7jq
opened
2 years ago
1
8g显存，bs设置为4，爆显存
#10
ray7jq
opened
2 years ago
10
Bump pillow from 8.2.0 to 9.0.0
#9
dependabot[bot]
closed
2 years ago
1
Next