Conv2D

2D convolution layer.

Abstract Signature:

Conv2D(filters: int, kernel_size: Union[int, Tuple[int, int]], strides: Union[int, Tuple[int, int]] = [1, 1], padding: str = valid, dilation_rate: Union[int, Tuple[int, int]] = [1, 1], groups: int = 1, use_bias: bool = True)

PyTorchJAX (Core)KerasTensorFlowApple MLXFlax NNXPaxML / Praxis

PyTorch

API: torch.nn.Conv2d

Strategy: Direct Mapping

Official Docs ↗

JAX (Core)

API: —

Strategy: Custom / Partial

Keras

API: keras.layers.Conv2D

Strategy: Direct Mapping

Official Docs ↗

TensorFlow

API: tf.keras.layers.Conv2D

Strategy: Direct Mapping

Official Docs ↗

Apple MLX

API: mlx.nn.Conv2d

Strategy: Direct Mapping

Official Docs ↗

Flax NNX

API: flax.nnx.Conv

Strategy: Direct Mapping

Official Docs ↗

PaxML / Praxis

API: praxis.layers.Conv2D

Strategy: Direct Mapping

Official Docs ↗