Hey.
I'm trying to follow the steps in clothing-attributes-moduler.ipynb

I initially was using torch version 1.4 and ran into some index errors[1 line in classifiers.py] but fixed it and the model resulted in Loss: -0.0180 Acc: 0.5318
after 18 epochs.

torch also raised some user warnings like
UserWarning: The use of the transforms.Scale transform is deprecated, please use transforms.Resize instead
and
UserWarning: Implicit dimension choice for softmax has been deprecated. Change the call to include dim=X as an argument. return F.softmax(self.model(x))

So I decided to test on an older version of torch [torch=0.4.1 and torchvision=0.2.2]
but still the same issue persists
I got the same warnings
and the model resulted in -0.0180 Acc: 0

What do you think could be the source of the issue? The negative loss value hints towards something to do with the loss function. So should I specify a dim value to the softmax layer?

Thanks :)

when I try build model I get always acc: 0.000 best_acc:0.000

num_epochs=50,
is_train=True,
use_gpu=True

Check for this image :: https://ecs7.tokopedia.net/img/cache/200-square/product-1/2018/11/16/44005912/44005912_38d3d852-4b67-4ef0-9e7a-d4006bf07202_700_700.jpg

The results for this are, i am mentioning only the relevant output classes -->
[Sweater, Short Sleeves, V-Shape, Graphics,None, None, Is Female, None, Wears Necktie, None ]

I am new to torch. Can you tell me how to introduce k cross fold validation with the training part in train_attribute_model(). I think some steps will be added after the data is loaded ie, after AttributeDataset class object is returned.

The documentation is poor here. Here are the steps to get a good result:

1. First download the ClothingAttributeDatasetand put it in the data folder.

2. In the ml_src folder, run the data-prep.ipynb. Don't spilt train test by hand.

3. In the ml_src/weights folder, create a folder called vgg16-fcn-266-2

4. Run python ml_src/build_models.py from outside the ml_src folder.

how can i get the weigths

hey! I feel confused about this address... http://localhost:9000
I can not open it, could you please tell me how can I open it successfully?
Even worse, when I try to Setup Environment on Local Machine
It always reminds me that CondaEnvironmentNotFoundError: Could not find environment: D:\Anaconda\envs\env/venv,but I have activated 'env/venv'....
Thank u!!

Hy! I'm trying to learn machine learning, and from this how to extract from a photo different attributes. I looked through your project, trying to figure out how you do it, but when i'm running it i got this message error RuntimeError: Found 0 images in subfolders of: data/ClothingAttributeDataset/train/, even if i created the folder data/ClothingAttributeDataset/train/, and put images into. can you please tell me, what i'm doing wrong, or what i missed? Thank you :).

Torch, TorchVision, Pandas etc were not listed in the requirements text file

I've managed to get the network to train on my machine but once it finishes training and tries to save the results I get the error:-

Traceback (most recent call last):
  File "/ml_src/build_models.py", line 37, in <module>
    use_gpu=use_gpu)
  File "/ml_src/classifiers.py", line 466, in create_attributes_model
    save_model(model, weights_path)
  File "/ml_src/classifiers.py", line 262, in save_model
    torch.save(model.state_dict(), weights_path)
  File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 209, in save
    return _with_file_like(f, "wb", lambda f: _save(obj, f, pickle_module, pickle_protocol))
  File "/opt/conda/lib/python3.6/site-packages/torch/serialization.py", line 132, in _with_file_like
    f = open(f, mode)
FileNotFoundError: [Errno 2] No such file or directory: '/ml_src/weights/vgg16-fcn-266-2/category_GT.pth'

Is the folder found in the pytorch system files? Do I need to download these weights from somewhere else?
Thanks