.to_device()

Move all parameters to the target device in-place.

Transfers every parameter tensor's underlying array to the backend of the given device (e.g. from numpy to cupy when moving to CUDA). Returns self for chaining.

Parameters:

• device (str) –

  Target device string, e.g. "cpu" or "cuda:0".

Returns:

• Module –

  This module (mutated in-place).

Raises:

• ValueError –

  If the device string is invalid.