tf.nn.quantized_relu_x(features, max_value, min_features, max_features, out_type=None, name=None)

tf.nn.quantized_relu_x(features, max_value, min_features, max_features, out_type=None, name=None)

See the guide: Neural Network > Candidate Sampling

Computes Quantized Rectified Linear X: min(max(features, 0), max_value)

Args:

  • features: A Tensor. Must be one of the following types: qint8, quint8, qint16, quint16, qint32.
  • max_value: A Tensor of type float32.
  • min_features: A Tensor of type float32. The float value that the lowest quantized value represents.
  • max_features: A Tensor of type float32. The float value that the highest quantized value represents.
  • out_type: An optional tf.DType from: tf.qint8, tf.quint8, tf.qint16, tf.quint16, tf.qint32. Defaults to tf.quint8.
  • name: A name for the operation (optional).

Returns:

A tuple of Tensor objects (activations, min_activations, max_activations). activations: A Tensor of type out_type. Has the same output shape as "features". min_activations: A Tensor of type float32. The float value that the lowest quantized value represents. * max_activations: A Tensor of type float32. The float value that the highest quantized value represents.

Defined in tensorflow/python/ops/gen_nn_ops.py.