Float16

Half precision float type.

PyTorch

API: torch.float16
Strategy: Direct Mapping
Official Docs ↗

JAX (Core)

API: jnp.float16
Strategy: Direct Mapping
Official Docs ↗

NumPy

API: np.float16
Strategy: Direct Mapping
Official Docs ↗

Keras

API: keras.float16
Strategy: Macro ''float16''

TensorFlow

API: tf.float16
Strategy: Direct Mapping
Official Docs ↗

Apple MLX

API: mx.float16
Strategy: Direct Mapping
Official Docs ↗

Flax NNX

API: jnp.float16
Strategy: Direct Mapping
Official Docs ↗

PaxML / Praxis

API: jnp.float16
Strategy: Direct Mapping
Official Docs ↗