Benchmark multi-model/multi-view models.

Note

This page is a reference documentation. It only explains the function signature, and not how to use it. Please refer to the gallery for the big picture.

mmbench.model.eval_vae(model, data, modalities, view_name='rois', n_samples=1, verbose=1)[source]¶

Evaluate the VAE model.

Parameters

model : Module

the input model.

data : dict

the input data organized by views.

modalities : list of str

names of the model input views.

view_name : str, default ‘rois’

the name of the view containg the data to encode.

verbose : int, default 1

control the verbosity level.

Returns

embeddings : dict

the generated latent representations.