Module: tf.quantization

Public API for tf.quantization namespace.

Functions

dequantize(...): Dequantize the 'input' tensor into a float or bfloat16 Tensor.

fake_quant_with_min_max_args(...): Fake-quantize the 'inputs' tensor, type float to 'outputs' tensor of same shape and type.

fake_quant_with_min_max_args_gradient(...): Compute gradients for a FakeQuantWithMinMaxArgs operation.

fake_quant_with_min_max_vars(...): Fake-quantize the 'inputs' tensor of type float via global float scalars

fake_quant_with_min_max_vars_gradient(...): Compute gradients for a FakeQuantWithMinMaxVars operation.

fake_quant_with_min_max_vars_per_channel(...): Fake-quantize the 'inputs' tensor of type float via per-channel floats

fake_quant_with_min_max_vars_per_channel_gradient(...): Compute gradients for a FakeQuantWithMinMaxVarsPerChannel operation.

quantize(...): Quantize the 'input' tensor of type float to 'output' tensor of type 'T'.

quantize_and_dequantize(...): Quantizes then dequantizes a tensor. (deprecated)

quantize_and_dequantize_v2(...): Quantizes then dequantizes a tensor.

quantized_concat(...): Concatenates quantized tensors along one dimension.