-
Sorry, I try to reproduce the code on CUB200/Flowers102, but I failed. Can you give some help.
-
Regarding the training of cub200, I followed all the parameter settings of the source code. I first loaded cc_learned.pth and started training. According to the paper, this should be the VQ-diffusion-…
-
Hi, I've tried to use HashNet to Fine-grained recognition, so I adopted the PyTorch code to CUB200 dataset with finetuned ResNet50, but I can't make the loss converge. I've tried different optimizer…
-
作者提供的code是完整的吗?
我按照readme中的指示验证实验,在CIFAR-100上训练时获得了与文章差不多的结果,但是在CUB200上运行train_cub_x.sh中的命令,均会报出ValueError: num_samples should be a positive integer value, but got num_samples=0的错误。
求解答,谢谢。
-
Hi, when I use the default settings of the train.sh, the acc of session_0 is 62, approximately. The acc raises to 68 when the base_epoch is increased by 70 to 120, however, the following sessions are …
-
hi, i run your code using CUB200 dataset, but i can't reappear result of recall@1(0.575). Can you give me more detail about configure of parameters, likely training batch size, labels-per-batch? And h…
-
first,i tried to train the CUB200 dataset, but the picture is blurry and there is no bird shape.
so i tried to train only one sample, the result was not good either.
-
Hi @lenscloth ~
I can't get the same recall score on dataset cub200 emm.......
My test:
Teacher : resnet50 embedding : 512 batchsize: 128 sample: distanceweight
and the result is 61.0 (very clo…
-
Ite seems that you didn't provide the right links for downloading the datasets including CUB200-2011, standford cars and aircraft.
-
The initial loss / log_prob is very big like 1e24. I tried `CUB200` as well, but `CUB200` has around `40000` initial loss and training goes well.
When I try to use your model in folder `realworld` …