`neural_compressor.tensorflow.quantization.utils.quantize_graph.quantize_graph_conv`

Quantize Conv2D/DepthwiseConv2dNative.

Module Contents

Classes

FuseNodeStartWithConv2d

Quantize Conv2D/DepthwiseConv2dNative to int8 op.

class neural_compressor.tensorflow.quantization.utils.quantize_graph.quantize_graph_conv.FuseNodeStartWithConv2d(**kwargs)[source]: Quantize Conv2D/DepthwiseConv2dNative to int8 op.