`neural_compressor.adaptor.torch_utils.model_wrapper`

Torch.nn.Module Class Definition.

Module Contents

`FakeAffineTensorQuantFunction`	Fake version of affine quantization.
`TEQLinearFakeQuant`	Wrapper quantization linear.
`MulLinear`	Linear wrapper to apply scale to input.

class neural_compressor.adaptor.torch_utils.model_wrapper.FakeAffineTensorQuantFunction[source]: Fake version of affine quantization.

class neural_compressor.adaptor.torch_utils.model_wrapper.TEQLinearFakeQuant(orig_layer, alpha=None, num_bits=4, group_size=-1, scheme='asym')[source]: Wrapper quantization linear.

class neural_compressor.adaptor.torch_utils.model_wrapper.MulLinear(module, input_scale=None)[source]: Linear wrapper to apply scale to input.