Visualizing Latent space

[aissak21]

Is there a way to visualize the latent space to see what features have been learnt? I see there are reconstruction and interpolation functionalities but can't seem to find a way to do so.

[clementchadebec]

Hello,
This feature was recently implemented in #76 (using the embed method) and should be available in the next release.

embedding = model.embed(your_data_tensor)

Before this, you can do the following

>>> model.encoder(your_data).embedding

E.g.

>>> from pythae.models import AutoModel
>>> model = AutoModel.load_from_hf_hub("clementchadebec/reproduced_beta_tc_vae", allow_pickle=True)
Downloading config file ...
Downloading BetaTCVAE files for rebuilding...
Successfully downloaded BetaTCVAE model!
>>> import torch
>>> x = torch.randn(3, 1, 64, 64)
>>> emb = model.encoder(x).embedding
>>> emb.shape
torch.Size([3, 10])

[aissak21]

I see. Quite timely then. Thank you!