GenerationPipeline

class pyraug.pipelines.GenerationPipeline(model, sampler=None)

This pipelines allows to generate new samples from a pre-trained model

Parameters

• model (BaseVAE) – The model you want ot generate from

• sampler (BaseSampler) – The sampler to use to sampler from the model

Warning

You must ensure that the sampler used handled the model provided

Tip

remember that a model can be easily reloaded from a folder using load_from_folder.

Example for a RHVAE

>>> from pyraug.models import RHVAE
>>> model_rec = RHVAE.load_from_folder('path/to_model_folder')

__call__(samples_number)

Launch the data generation and save it in output_dir stated in the BaseSamplerConfig. A folder generation_YYYY-MM-DD_hh-mm-ss is created and data is saved in .pt files in this created folder. If output_dir is None, data is saved in dummy_output_dir/generation_YYYY-MM-DD_hh-mm-ss

Parameters

samples_number (int) – The number of samples to generate