uGUIDE.run_inference¶

uGUIDE.run_inference(theta, x, config, plot_loss=True, load_state=False)¶

Run inference given a training dataset.

Parameters:

thetandarray, shape (nb_simulations, config[‘size_theta’]): Microstructure parameters used for the training.
xndarray, shape (nb_simulations, config[‘size_x’]): Observed diffusion MRI signals corresponding to the theta.
configdict: μGUIDE configuration.
plot_lossbool, default=True: Plot the validation loss evolution during training.
load_state: bool, default=False: If training has already been performed, load the state of the neural networks saved if set to True, and continue training. Otherwise, start a new inference.

μGUIDE