NeuralNetOperationsConverter

Converter for neural network operations.

This converter handles neural network specific operations including:

• Convolutional operations (conv2d) with proper attribute mapping

• Pooling operations (maxPool2d, avgPool2d) with correct StableHLO mapping

• Batch normalization and layer normalization operations

Supports operations as specified in Requirements 2.4:

• conv2d with proper attribute mapping for strides, padding, dilation

• Different pooling types (max, average) with correct StableHLO mapping

• Batch normalization and layer normalization operations