107 lines
4.3 KiB
Python
107 lines
4.3 KiB
Python
"""
|
|
Copyright (c) 2019 Intel Corporation
|
|
|
|
Licensed under the Apache License, Version 2.0 (the "License");
|
|
you may not use this file except in compliance with the License.
|
|
You may obtain a copy of the License at
|
|
|
|
http://www.apache.org/licenses/LICENSE-2.0
|
|
|
|
Unless required by applicable law or agreed to in writing, software
|
|
distributed under the License is distributed on an "AS IS" BASIS,
|
|
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
|
|
See the License for the specific language governing permissions and
|
|
limitations under the License.
|
|
"""
|
|
|
|
import numpy as np
|
|
|
|
from mo.graph.graph import Node, Graph
|
|
from mo.ops.op import Op
|
|
from mo.utils.error import Error
|
|
|
|
|
|
def broadcastable(broadcast_from, broadcast_to):
|
|
"""Check if shape broadcast_from can be broadcasted to broadcast_to"""
|
|
broadcast_to = np.array(broadcast_to, dtype=np.int64)
|
|
broadcast_from = np.array(broadcast_from, dtype=np.int64)
|
|
if broadcast_from.size > broadcast_to.size:
|
|
return False
|
|
broadcast_from = np.concatenate(
|
|
(np.array([1] * (broadcast_to.size - broadcast_from.size), dtype=np.int64), broadcast_from))
|
|
return np.all(np.logical_or(broadcast_from == 1, broadcast_from == broadcast_to))
|
|
|
|
|
|
class FakeQuantize(Op):
|
|
op = 'FakeQuantize'
|
|
|
|
def __init__(self, graph: Graph, attrs: dict):
|
|
mandatory_props = {
|
|
'type': __class__.op,
|
|
'op': __class__.op,
|
|
'levels': None,
|
|
# flag to switch between dumping FakeQuantize as statistics and keeping it as layer in IR
|
|
'keep_in_IR': None,
|
|
'infer': __class__.infer,
|
|
'in_ports_count': 5,
|
|
'out_ports_count': 1,
|
|
}
|
|
super().__init__(graph, mandatory_props, attrs)
|
|
if self.attrs['levels'] is None:
|
|
raise Error("FakeQuantize operation has no levels parameter")
|
|
# TODO remove following lines after FakeQuantize supported for int8 workflow
|
|
self.attrs['keep_in_IR'] = self.attrs['levels'] == 2 or graph.graph['cmd_params'].keep_quantize_ops_in_IR
|
|
|
|
def supported_attrs(self):
|
|
return [
|
|
'levels',
|
|
]
|
|
|
|
@staticmethod
|
|
def infer(node: Node):
|
|
assert len(node.in_nodes()) == 5
|
|
assert len(node.out_nodes()) == 1
|
|
inputs = [node.in_node(i) for i in range(5)]
|
|
x, input_low, input_high, output_low, output_high = inputs
|
|
assert x.has_valid('shape')
|
|
# TODO Check all input[1..4] shapes are broadcastable to intput[0] shape
|
|
assert all([broadcastable(inputs[i].shape, inputs[0].shape) for i in range(1, 5)]), \
|
|
"Not all shapes from FakeQuantize inputs can be broadcasted to input[0] for node {}".format(
|
|
node.soft_get('name'))
|
|
node.out_node().shape = x.shape.copy()
|
|
|
|
if all([node.in_node(i).has_valid('value') for i in range(5)]):
|
|
x, input_low, input_high, output_low, output_high = \
|
|
[np.array(np.broadcast_to(node.value, x.value.shape)) for node in inputs]
|
|
|
|
assert node.has_valid('levels')
|
|
assert isinstance(node.levels, int)
|
|
|
|
underflow_mask = x <= input_low
|
|
overflow_mask = x > input_high
|
|
# pylint: disable=assignment-from-no-return
|
|
middle_mask = np.logical_not(np.logical_or(underflow_mask, overflow_mask))
|
|
|
|
def middle_part(x, input_low, input_high, output_low, output_high):
|
|
return np.round(
|
|
(x - input_low) / (input_high - input_low) * (node.levels - 1)
|
|
) / (node.levels - 1) * (output_high - output_low) + output_low
|
|
|
|
output = np.zeros_like(x)
|
|
# pylint: disable=unsupported-assignment-operation
|
|
output[middle_mask] = middle_part(
|
|
x[middle_mask],
|
|
input_low[middle_mask],
|
|
input_high[middle_mask],
|
|
output_low[middle_mask],
|
|
output_high[middle_mask],
|
|
)
|
|
|
|
# pylint: disable=unsupported-assignment-operation
|
|
output[overflow_mask] = output_high[overflow_mask]
|
|
# pylint: disable=unsupported-assignment-operation
|
|
output[underflow_mask] = output_low[underflow_mask]
|
|
|
|
if not node.has_and_set('stop_value_propagation'):
|
|
node.out_node().value = output
|