Function _InsertQuantOp

tensorflow/contrib/quantize/python/quantize.py:664–802

Inserts a quant op between a producer op and (multiple) consumer ops. Args: context: Context where producer and consumer operations are nested. name: Name for the new quantization op within the context. producer: Producer operation of the pairs where quantization will be inserte

(context,
                   name,
                   producer,
                   consumers,
                   is_training,
                   per_channel=False,
                   moving_avg=True,
                   init_min=-6.0,
                   init_max=6.0,
                   bits=8,
                   symmetric=False,
                   ema_decay=0.999,
                   quant_delay=None,
                   vars_collection=ops.GraphKeys.GLOBAL_VARIABLES,
                   narrow_range=False,
                   producer_scope=None,
                   consumer_scope=None,
                   use_qdq=False)

def _InsertQuantOp(context,
                   name,
                   producer,
                   consumers,
                   is_training,
                   per_channel=False,
                   moving_avg=True,
                   init_min=-6.0,
                   init_max=6.0,
                   bits=8,
                   symmetric=False,
                   ema_decay=0.999,
                   quant_delay=None,
                   vars_collection=ops.GraphKeys.GLOBAL_VARIABLES,
                   narrow_range=False,
                   producer_scope=None,
                   consumer_scope=None,
                   use_qdq=False):
  """Inserts a quant op between a producer op and (multiple) consumer ops.

  Args:
    context: Context where producer and consumer operations are nested.
    name: Name for the new quantization op within the context.
    producer: Producer operation of the pairs where quantization will be
      inserted.
    consumers: Consumer operations of the pairs.
    is_training: Whether quantizing training graph or eval graph.
    moving_avg: Specifies whether to use exponential moving average or just
      the last value seen.
    init_min: Starting minimum value for the new quantization op.
    init_max: Starting maximum value for the new quantization op.
    bits: Number of bits to use for quantization, must be between 2 and 8.
    symmetric: (Optional) If true, use symmetric quantization limits instead of
      training the minimum and maximum of each quantization range separately.
    ema_decay: (Optional) Float, EMA decay parameter.  EMA is used to update
      quantization intervals for quantizing activations (see here about EMA:
      https://en.wikipedia.org/wiki/Moving_average#Exponential_moving_average).
    quant_delay: (Optional, default None) Int, count of global steps for which
      to delay quantization.  This helps weights stabilize at the start of
      training.
    vars_collection: (Optional) Collection where to store the variables for
      quantization interval ends.
    narrow_range: Whether to use the narrow quantization range
      [1; 2^bits - 1] or wide range [0; 2^bits - 1].
    producer_scope: The restriction of producer scope. If not None, the new op
      will be inserted only when the producer is in this scope.
    consumer_scope: The restriction of consumer scope. If not None, the new op
      will be inserted only when all the consumers are in this scope.
  Raises:
    ValueError: When producer operation is not directly connected to the
      consumer operation.
  """
  if producer_scope and not producer.name.startswith(producer_scope):
    logging.info(
        '_InsertQuantOp ignores context="%s" name="%s" '
        'because producer "%s" is not in scope "%s"',
        context, name, producer.name, producer_scope)
    return