Fmax

Element-wise maximum of input and other, ignoring NaNs.

Abstract Signature:

Fmax(input: Tensor, other: Tensor)

PyTorch

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

JAX (Core)

API: jax.numpy.fmax
Strategy: Direct Mapping
Official Docs ↗

NumPy

API: numpy.fmax
Strategy: Direct Mapping
Official Docs ↗

Keras

API: keras.ops.maximum
Strategy: Macro '{input}.where(isnan({input}), {other}, {input}).maximum({other})'

TensorFlow

API: tf.math.maximum
Strategy: Direct Mapping
Official Docs ↗

Apple MLX

API: mlx.core.maximum
Strategy: Direct Mapping
Official Docs ↗

Flax NNX

API: jax.numpy.fmax
Strategy: Direct Mapping
Official Docs ↗

PaxML / Praxis

API: jax.numpy.fmax
Strategy: Direct Mapping
Official Docs ↗