scvi._settings.ScviConfig

scvi._settings.ScviConfig#

class scvi._settings.ScviConfig(verbosity=20, progress_bar_style='tqdm', batch_size=128, seed=None, logging_dir='./scvi_log/', dl_num_workers=0, dl_persistent_workers=False, jax_preallocate_gpu_memory=False, warnings_stacklevel=2, mlflow_set_tracking_uri='', mlflow_set_experiment='mlflow_experiment')[source]#

Config manager for scvi-tools.

Examples

To set the seed

>>> scvi.settings.seed = 1

To set the batch size for functions like SCVI.get_latent_representation

>>> scvi.settings.batch_size = 1024

To set the progress bar style, choose one of “rich”, “tqdm”

>>> scvi.settings.progress_bar_style = "rich"

To set the verbosity

>>> import logging
>>> scvi.settings.verbosity = logging.INFO

To set the number of threads, PyTorch will use

>>> scvi.settings.num_threads = 2

To prevent Jax from preallocating GPU memory on start (default)

>>> scvi.settings.jax_preallocate_gpu_memory = False

Attributes table#

`batch_size`	Minibatch size for loading data into the model.
`dl_num_workers`	Number of workers for PyTorch data loaders (Default is 0).
`dl_persistent_workers`	Whether to use persistent_workers in PyTorch data loaders (Default is False).
`jax_preallocate_gpu_memory`	Jax GPU memory allocation settings.
`logging_dir`	Directory for training logs (default './scvi_log/').
`mlflow_set_experiment`	Setting the MLFlow experiment name
`mlflow_set_tracking_uri`	Setting the MLFlow tracking URI.
`num_threads`	Number of threads PyTorch will use.
`progress_bar_style`	Library to use for progress bar.
`seed`	Random seed for torch and numpy.
`verbosity`	Verbosity level (default logging.INFO).
`warnings_stacklevel`	Stacklevel for warnings.

Methods table#

reset_logging_handler()

Resets "scvi" log handler to a basic RichHandler().

Attributes#

ScviConfig.batch_size[source]#

Minibatch size for loading data into the model.

This is only used after a model is trained. Trainers have specific batch_size parameters.

ScviConfig.dl_num_workers[source]#: Number of workers for PyTorch data loaders (Default is 0).

ScviConfig.dl_persistent_workers[source]#: Whether to use persistent_workers in PyTorch data loaders (Default is False).

ScviConfig.jax_preallocate_gpu_memory[source]#

Jax GPU memory allocation settings.

If False, Jax will only preallocate GPU memory it needs. If float in (0, 1), Jax will preallocate GPU memory to that fraction of the GPU memory.

ScviConfig.logging_dir[source]#: Directory for training logs (default ‘./scvi_log/’).

ScviConfig.mlflow_set_experiment[source]#: Setting the MLFlow experiment name

ScviConfig.mlflow_set_tracking_uri[source]#: Setting the MLFlow tracking URI. Setting it will cause to also use it

ScviConfig.num_threads[source]#: Number of threads PyTorch will use.

ScviConfig.progress_bar_style[source]#: Library to use for progress bar.

ScviConfig.seed[source]#: Random seed for torch and numpy.

ScviConfig.verbosity[source]#: Verbosity level (default logging.INFO).

ScviConfig.warnings_stacklevel[source]#: Stacklevel for warnings.

Methods#

ScviConfig.reset_logging_handler()[source]#

Resets “scvi” log handler to a basic RichHandler().

This is useful if piping outputs to a file.