search

lschmiddey / deep_tabular_augmentation

MIT License
20 stars 9 forks source link

'deep_tabular_augmentation' has no attribute 'AutoencoderModel' #7

Closed WiemHAD closed 2 years ago

WiemHAD commented 2 years ago

Hi, I'm using dta, but I get this error: module 'deep_tabular_augmentation' has no attribute 'AutoencoderModel' I'm a beginner in data science, so I can't fix it by my self. Any help, please ?

lschmiddey commented 2 years ago

Hi WiemHAD,

I'm not 100% sure you are referring to, but I think you might be using an old version. I think maybe this blogpost here: https://medium.com/analytics-vidhya/comparison-randomforest-with-oversampling-vs-augmented-data-bd7fcb96ef0f might help you in how to use the DeepAugmenter.

When building the model you need to do something like this:

model = dta.Autoencoder(nn.Sequential(dta.get_lin_layers(D_in, [50, 12, 12])), nn.Sequential(dta.get_lin_layers_rev(D_in, [50, 12, 12])), latent_dim=5).to(device)

Taichi-Pink commented 2 years ago

Hi @lschmiddey,

I guess @WiemHAD referred to the function AutoencoderModel in this postblog:postblog.