GetDefaultDtype

Get the current default floating point dtype.

PyTorchJAX (Core)NumPyKerasTensorFlowFlax NNXPaxML / Praxis

PyTorch

API: torch.get_default_dtype

Strategy: Direct Mapping

Official Docs ↗

JAX (Core)

API: —

Strategy: Macro 'jnp.float32'

NumPy

API: —

Strategy: Macro 'np.float64'

Keras

API: keras.backend.floatx

Strategy: Direct Mapping

Official Docs ↗

TensorFlow

API: tf.keras.backend.floatx

Strategy: Direct Mapping

Official Docs ↗

Flax NNX

API: —

Strategy: Macro 'jnp.float32'

PaxML / Praxis

API: —

Strategy: Macro 'jnp.float32'