267 lines
11 KiB
Python
267 lines
11 KiB
Python
"""
|
|
Copyright (C) 2018-2020 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 logging as log
|
|
from copy import deepcopy
|
|
|
|
import numpy as np
|
|
|
|
from extensions.middle.SliceConverter import ConvertSlice
|
|
from extensions.ops.split import VariadicSplit
|
|
from mo.front.common.partial_infer.utils import int64_array
|
|
from mo.graph.graph import Graph, Node, add_opoutput
|
|
from mo.middle.replacement import MiddleReplacementPattern
|
|
from mo.ops.const import Const
|
|
from mo.ops.op import Op
|
|
from mo.ops.squeeze import Squeeze
|
|
from mo.ops.unsqueeze import Unsqueeze
|
|
|
|
|
|
class ConvertGroupedStridedSlice(MiddleReplacementPattern):
|
|
"""
|
|
This pass converts subgraphs where StridedSlices used for splitting single channel to single Split layers
|
|
In case if StrdedSlices consume not entire tensor will be created fake outputs for Split layer
|
|
For example:
|
|
Let's suppose we have next graph:
|
|
Data(1,H,W,54)
|
|
|`---->Sslice1_out (1,H,W,(10,18))
|
|
`---->Sslice2_out (1,H,W,(18,36))
|
|
|
|
In this case StridedSlices takes only [10, 36] from input tensor in 3rd dim
|
|
So this pass will convert this graph to the next one:
|
|
Split(1,H,W,54)
|
|
|`---->Fake_data (1,H,W,10)
|
|
|`---->Sslice1_out (1,H,W,8)
|
|
|`---->Sslice2_out (1,H,W,18)
|
|
`----->Fake_data (1,H,W,18)
|
|
Where Fake_data - data nodes that have not any consumers.
|
|
"""
|
|
|
|
enabled = True
|
|
|
|
def run_after(self):
|
|
return [ConvertSlice]
|
|
|
|
def run_before(self):
|
|
from extensions.middle.pass_separator import MiddleFinish
|
|
return [MiddleFinish]
|
|
|
|
def find_and_replace_pattern(self, graph: Graph):
|
|
# Iterate over all data nodes and find all with >= 1 consumers
|
|
for input_data in list(graph.get_data_nodes()):
|
|
# We don't use constant data nodes
|
|
if input_data.value is not None:
|
|
continue
|
|
|
|
input_shape = np.array(input_data.shape)
|
|
|
|
# Get all StridedSlice consumers
|
|
out_nodes = [node for node in input_data.out_nodes() if node.op == 'StridedSlice' and node.in_node(0).name == input_data.name]
|
|
if len(out_nodes) <= 1:
|
|
continue
|
|
|
|
valid_for_replacement = True
|
|
|
|
for node in out_nodes:
|
|
if len(node.slices) != len(out_nodes[0].slices):
|
|
valid_for_replacement = False
|
|
|
|
# Detect dimension for splitting
|
|
split_channel_dim = None
|
|
for dim_id, s in enumerate(out_nodes[0].slices):
|
|
l, r, stride = s.start, s.stop, s.step
|
|
if l != 0 or r != input_shape[dim_id]:
|
|
if split_channel_dim is None:
|
|
split_channel_dim = dim_id
|
|
else:
|
|
valid_for_replacement = False
|
|
|
|
if split_channel_dim is None:
|
|
valid_for_replacement = False
|
|
|
|
# split_dims contains tuples with split range and output data node
|
|
split_dims = []
|
|
for out_id, node in enumerate(out_nodes):
|
|
# Check that StridedSlice op has stride eq 1 and splits only feature channel
|
|
for id, s in enumerate(node.slices):
|
|
l, r, stride = s.start, s.stop, s.step
|
|
# We don't support StridedSlice with stride != 1
|
|
if stride != 1:
|
|
valid_for_replacement = False
|
|
if id == split_channel_dim:
|
|
split_dims.append((s.start, s.stop, node.out_node()))
|
|
|
|
if not valid_for_replacement:
|
|
continue
|
|
|
|
# Check feature split intersection
|
|
final_data_nodes_list = []
|
|
sorted_split_dims = sorted(split_dims, key=lambda item: (item[0], item[1]))
|
|
|
|
# check if we have similar StridedSlice operations with different outputs
|
|
prev_sd = sorted_split_dims[0]
|
|
to_remove = []
|
|
for i in range(1, len(sorted_split_dims)):
|
|
if sorted_split_dims[i][0] == prev_sd[0] and sorted_split_dims[i][1] == prev_sd[1] and sorted_split_dims[i][2].name != prev_sd[2].name:
|
|
cur_node = sorted_split_dims[i][2]
|
|
for out in cur_node.out_nodes():
|
|
attrs = deepcopy(graph.get_edge_data(cur_node.id, out.id)[0])
|
|
graph.remove_edge(cur_node.id, out.id)
|
|
graph.add_edge(prev_sd[2].id, out.id, **attrs)
|
|
to_remove.append(i)
|
|
|
|
for ind in reversed(to_remove):
|
|
sorted_split_dims.pop(ind)
|
|
|
|
size_splits = []
|
|
prev_r = 0
|
|
for l, r, out in sorted_split_dims:
|
|
# Split dims shouldn't intersect
|
|
if l < prev_r:
|
|
valid_for_replacement = False
|
|
prev_r = r
|
|
|
|
if prev_r > input_shape[split_channel_dim]:
|
|
valid_for_replacement = False
|
|
|
|
if not valid_for_replacement:
|
|
continue
|
|
|
|
prev_r = 0
|
|
for l, r, out in sorted_split_dims:
|
|
# Save missing tensor part
|
|
if l > prev_r:
|
|
shape = np.array(input_shape)
|
|
size_splits.append(l - prev_r)
|
|
shape[split_channel_dim] = l - prev_r
|
|
data_node = Op._create_data_node(graph, 'fake_data_'+out_nodes[0].name, {'shape': shape})
|
|
add_opoutput(graph, data_node.id, 0, False)
|
|
final_data_nodes_list.append(data_node)
|
|
|
|
prev_r = r
|
|
size_splits.append(r - l)
|
|
final_data_nodes_list.append(out)
|
|
|
|
if prev_r < input_shape[split_channel_dim]:
|
|
# Add last part of tensor
|
|
shape = input_shape.copy()
|
|
shape[split_channel_dim] = input_shape[split_channel_dim] - prev_r
|
|
size_splits.append(input_shape[split_channel_dim] - prev_r)
|
|
data_node = Op._create_data_node(graph, 'fake_data_'+out_nodes[0].name, {'shape': shape})
|
|
add_opoutput(graph, data_node.id, 0, False)
|
|
final_data_nodes_list.append(data_node)
|
|
|
|
for node in out_nodes:
|
|
if not np.all([x == 0 for x in node.shrink_axis_mask]):
|
|
out_node = node.out_node()
|
|
if np.any(node['shrink_axis_mask']):
|
|
self.add_squeeze_for_shrink(graph, node)
|
|
if np.any(node['new_axis_mask']):
|
|
self.add_unsqueeze_for_new(graph, node)
|
|
|
|
for i in range(len(final_data_nodes_list)):
|
|
if final_data_nodes_list[i].name == out_node.name:
|
|
final_data_nodes_list[i] = node.out_node()
|
|
break
|
|
|
|
# Insert Split layer and remove old StridedSlice layers
|
|
# 1. Remove connections from input_data to StridedSlice ops
|
|
out_data_nodes = []
|
|
name_for_future_split = out_nodes[0].name
|
|
for node in out_nodes:
|
|
out_data_nodes.append(node.out_node())
|
|
graph.remove_edge(input_data.id, node.id)
|
|
graph.remove_edge(node.id, node.out_node().id)
|
|
graph.remove_node(node.id)
|
|
log.debug("Removed: {}".format(node.id))
|
|
|
|
# 2. Create Split layer and reorder outputs
|
|
axis_const = Const(graph, {'value': int64_array(split_channel_dim)}).create_node_with_data()
|
|
size_splits_const = Const(graph, {'value': int64_array(size_splits)}).create_node_with_data()
|
|
split = VariadicSplit(graph, dict(name=name_for_future_split + "/Split", out_ports_count=len(size_splits)))
|
|
split.create_node_with_data(inputs=[input_data, axis_const, size_splits_const],
|
|
data_nodes=final_data_nodes_list)
|
|
|
|
@staticmethod
|
|
def add_squeeze_for_shrink(graph: Graph, ss_node: Node):
|
|
# add Squeeze for shrink_axis_mask
|
|
log.info("StridedSlice op with shrink mask '{}' has been detected".format(ss_node.id))
|
|
|
|
if len(ss_node.in_nodes()) != 4 or len(ss_node.out_nodes()) != 1:
|
|
return
|
|
|
|
shape_out = ss_node.out_node().shape
|
|
dim = np.array(range(len(ss_node['shrink_axis_mask'])))[np.array(ss_node['shrink_axis_mask'], dtype=bool)]
|
|
ss_shape = []
|
|
i = 0
|
|
k = 0
|
|
|
|
# Don't permute reshape if channels were squeezed
|
|
dont_permute = graph.graph['layout'] == 'NCHW'
|
|
if graph.graph['layout'] == 'NHWC' and ss_node['shrink_axis_mask'][-1] == 1:
|
|
dont_permute = True
|
|
|
|
while k < len(shape_out):
|
|
if i >= len(ss_node['shrink_axis_mask']) or not ss_node['shrink_axis_mask'][i]:
|
|
ss_shape.append(shape_out[k])
|
|
k = k + 1
|
|
else:
|
|
ss_node['shrink_axis_mask'][i] = 0
|
|
ss_shape.append(1)
|
|
i = i + 1
|
|
|
|
while i < len(ss_node['shrink_axis_mask']):
|
|
ss_node['shrink_axis_mask'][i] = 0
|
|
ss_shape.append(1)
|
|
i = i + 1
|
|
|
|
ss_node.out_port(0).data.set_shape(ss_shape)
|
|
|
|
# insert Squeeze
|
|
squeeze_node = Squeeze(graph, dict(name=ss_node.name + '/Squeeze_shrink',
|
|
nchw_layout=dont_permute,
|
|
correct_data_layout=dont_permute)).create_node()
|
|
ss_node.out_port(0).get_connection().insert_node(squeeze_node)
|
|
squeeze_node.out_port(0).data.set_shape(shape_out)
|
|
|
|
dims_node = Const(graph, {'name': squeeze_node.id + '/Indices', 'value': int64_array(dim)}).create_node()
|
|
dims_node.out_port(0).connect(squeeze_node.in_port(1))
|
|
|
|
@staticmethod
|
|
def add_unsqueeze_for_new(graph: Graph, ss_node: Node):
|
|
log.info("StridedSlice op with new axis mask '{}' has been detected".format(ss_node.id))
|
|
if len(ss_node.in_nodes()) != 4 or len(ss_node.out_nodes()) != 1:
|
|
return
|
|
|
|
shape_out = ss_node.out_node().shape
|
|
dim = np.array(range(len(ss_node['new_axis_mask'])))[np.array(ss_node['new_axis_mask'], dtype=bool)]
|
|
ss_shape = []
|
|
for i in range(0, len(ss_node['new_axis_mask'])):
|
|
if not ss_node['new_axis_mask'][i]:
|
|
ss_shape.append(shape_out[i])
|
|
else:
|
|
ss_node['new_axis_mask'][i] = 0
|
|
|
|
ss_node.out_port(0).data.set_shape(ss_shape)
|
|
|
|
# insert Unsqueeze
|
|
unsqueeze_node = Unsqueeze(graph, dict(name=ss_node.name + '/Unsqueeze_new')).create_node()
|
|
ss_node.out_port(0).get_connection().insert_node(unsqueeze_node)
|
|
unsqueeze_node.out_port(0).data.set_shape(shape_out)
|
|
|
|
dims_node = Const(graph, {'name': unsqueeze_node.id + '/Indices', 'value': int64_array(dim)}).create_node()
|
|
dims_node.out_port(0).connect(unsqueeze_node.in_port(1))
|