damast.ml#

Subpackages#

Submodules#

Attributes#

Functions#

backend_available(→ bool)

autodiscover_backend([priority])

backend_device(→ str)

Device the keras backend computes on - the backends use a visible GPU by themselves,

Package Contents#

damast.ml.logger#
damast.ml.backend_available(backend: str) → bool#
damast.ml.autodiscover_backend(priority: list[str] = ['torch', 'tensorflow', 'jax'])#
damast.ml.backend#
damast.ml.backend_device() → str#

Device the keras backend computes on - the backends use a visible GPU by themselves, CUDA_VISIBLE_DEVICES= restricts them to the CPU.

Returns:

"gpu" or "cpu"