def build(cls, model_name, model_version, model_xml, model_bin,
mapping_config, batch_size_param, shape_param, num_ireq,
target_device, plugin_config):
plugin = IEPlugin(device=target_device,
plugin_dirs=GLOBAL_CONFIG['plugin_dir'])
if GLOBAL_CONFIG['cpu_extension'] is not None \
and 'CPU' in target_device:
plugin.add_cpu_extension(GLOBAL_CONFIG['cpu_extension'])
net = IENetwork(model=model_xml, weights=model_bin)
batching_info = BatchingInfo(batch_size_param)
shape_info = ShapeInfo(shape_param, net.inputs)
if batching_info.mode == BatchingMode.FIXED:
net.batch_size = batching_info.batch_size
else:
batching_info.batch_size = net.batch_size
effective_batch_size = batching_info.get_effective_batch_size()
logger.debug(
"[Model: {}, version: {}] --- effective batch size - {}".format(
model_name, model_version, effective_batch_size))
###############################
# Initial shape setup
if shape_info.mode == ShapeMode.FIXED:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"fixed value: {}".format(model_name, model_version,
shape_info.shape))
net.reshape(shape_info.shape)
elif shape_info.mode == ShapeMode.AUTO:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"automatic".format(model_name, model_version))
net.reshape({})
elif shape_info.mode == ShapeMode.DEFAULT:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"default".format(model_name, model_version))
###############################
# Creating free infer requests indexes queue
free_ireq_index_queue = queue.Queue(maxsize=num_ireq)
for ireq_index in range(num_ireq):
free_ireq_index_queue.put(ireq_index)
###############################
requests_queue = queue.Queue(
maxsize=GLOBAL_CONFIG['engine_requests_queue_size'])
exec_net = plugin.load(network=net,
num_requests=num_ireq,
config=plugin_config)
ir_engine = cls(model_name=model_name,
model_version=model_version,
mapping_config=mapping_config,
net=net,
plugin=plugin,
exec_net=exec_net,
batching_info=batching_info,
shape_info=shape_info,
free_ireq_index_queue=free_ireq_index_queue,
num_ireq=num_ireq,
requests_queue=requests_queue,
target_device=target_device,
plugin_config=plugin_config)
return ir_engine
def get_fake_ir_engine():
mapping_config = 'mapping_config.json'
exec_net = MockedExecNet()
net = MockedNet(
inputs={DEFAULT_INPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={DEFAULT_OUTPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')})
core = None
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
requests_queue = queue.Queue()
free_ireq_index_queue = queue.Queue(maxsize=1)
free_ireq_index_queue.put(0)
engine = IrEngine(model_name='test',
model_version=1,
mapping_config=mapping_config,
exec_net=exec_net,
net=net,
core=core,
batching_info=batching_info,
shape_info=shape_info,
target_device="CPU",
free_ireq_index_queue=free_ireq_index_queue,
plugin_config=None,
num_ireq=1,
requests_queue=requests_queue)
return engine
def test_init_class():
mapping_config = 'mapping_config.json'
exec_net = None
net = MockedNet(inputs={'input': MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={'output': MockedIOInfo('FP32', [1, 1], 'NCHW')})
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
plugin = None
requests_queue = queue.Queue()
free_ireq_index_queue = queue.Queue(maxsize=1)
free_ireq_index_queue.put(0)
engine = IrEngine(model_name='test',
model_version=1,
mapping_config=mapping_config,
exec_net=exec_net,
net=net,
plugin=plugin,
batching_info=batching_info,
shape_info=shape_info,
num_ireq=1,
free_ireq_index_queue=free_ireq_index_queue,
requests_queue=requests_queue,
target_device='CPU',
plugin_config=None)
assert exec_net == engine.exec_net
assert list(net.inputs.keys()) == engine.input_tensor_names
assert list(net.outputs.keys()) == engine.output_tensor_names
assert engine.free_ireq_index_queue.qsize() == 1
def get_fake_model():
mapping_config = 'mapping_config.json'
exec_net = None
net = MockedNet(
inputs={DEFAULT_INPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={DEFAULT_OUTPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')})
plugin = None
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
new_engines = {}
available_versions = [1, 2, 3]
for version in available_versions:
engine = IrEngine(model_name='test',
model_version=version,
mapping_config=mapping_config,
exec_net=exec_net,
net=net,
plugin=plugin,
batching_info=batching_info,
shape_info=shape_info)
new_engines.update({version: engine})
model_name = "test"
versions_statuses = {}
batch_size_param, shape_param = None, None
for version in available_versions:
versions_statuses[version] = ModelVersionStatus(model_name, version)
new_model = LocalModel(model_name=model_name,
model_directory='fake_path/model/',
available_versions=available_versions,
engines=new_engines,
batch_size_param=batch_size_param,
shape_param=shape_param,
version_policy_filter=lambda versions: versions[:],
versions_statuses=versions_statuses)
return new_model
def build(cls, model_name, model_version, model_xml, model_bin,
mapping_config, batch_size_param, shape_param):
plugin = IEPlugin(device=DEVICE, plugin_dirs=PLUGIN_DIR)
if CPU_EXTENSION and 'CPU' in DEVICE:
plugin.add_cpu_extension(CPU_EXTENSION)
net = IENetwork(model=model_xml, weights=model_bin)
batching_info = BatchingInfo(batch_size_param)
shape_info = ShapeInfo(shape_param, net.inputs)
if batching_info.mode == BatchingMode.FIXED:
net.batch_size = batching_info.batch_size
else:
batching_info.batch_size = net.batch_size
effective_batch_size = batching_info.get_effective_batch_size()
logger.debug(
"[Model: {}, version: {}] --- effective batch size - {}".format(
model_name, model_version, effective_batch_size))
###############################
# Initial shape setup
if shape_info.mode == ShapeMode.FIXED:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"fixed value: {}".format(model_name, model_version,
shape_info.shape))
net.reshape(shape_info.shape)
elif shape_info.mode == ShapeMode.AUTO:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"automatic".format(model_name, model_version))
net.reshape({})
elif shape_info.mode == ShapeMode.DEFAULT:
logger.debug("[Model: {}, version: {}] --- Setting shape to "
"default".format(model_name, model_version))
###############################
exec_net = plugin.load(network=net, num_requests=1)
ir_engine = cls(model_name=model_name,
model_version=model_version,
mapping_config=mapping_config,
net=net,
plugin=plugin,
exec_net=exec_net,
batching_info=batching_info,
shape_info=shape_info)
return ir_engine
def test_init_class():
mapping_config = 'mapping_config.json'
exec_net = None
net = MockedNet(inputs={'input': MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={'output': MockedIOInfo('FP32', [1, 1], 'NCHW')})
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
plugin = None
engine = IrEngine(model_name='test', model_version=1,
mapping_config=mapping_config,
exec_net=exec_net,
net=net, plugin=plugin, batching_info=batching_info,
shape_info=shape_info)
assert exec_net == engine.exec_net
assert list(net.inputs.keys()) == engine.input_tensor_names
assert list(net.outputs.keys()) == engine.output_tensor_names
def get_fake_model():
mapping_config = 'mapping_config.json'
exec_net = MockedExecNet()
net = MockedNet(
inputs={DEFAULT_INPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={DEFAULT_OUTPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')})
core = None
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
new_engines = {}
available_versions = [1, 2, 3]
requests_queue = queue.Queue()
free_ireq_index_queue = queue.Queue(maxsize=1)
free_ireq_index_queue.put(0)
for version in available_versions:
engine = IrEngine(model_name='test',
model_version=version,
mapping_config=mapping_config,
exec_net=exec_net,
net=net,
core=core,
batching_info=batching_info,
shape_info=shape_info,
target_device="CPU",
free_ireq_index_queue=free_ireq_index_queue,
plugin_config=None,
num_ireq=1,
requests_queue=requests_queue)
new_engines.update({version: engine})
model_name = "test"
versions_statuses = {}
batch_size_param, shape_param = None, None
for version in available_versions:
versions_statuses[version] = ModelVersionStatus(model_name, version)
new_model = LocalModel(model_name=model_name,
model_directory='fake_path/model/',
available_versions=available_versions,
engines=new_engines,
batch_size_param=batch_size_param,
shape_param=shape_param,
version_policy_filter=lambda versions: versions[:],
versions_statuses=versions_statuses,
update_locks={},
plugin_config=None,
target_device="CPU",
num_ireq=1)
return new_model
def get_fake_ir_engine():
mapping_config = 'mapping_config.json'
exec_net = None
net = MockedNet(
inputs={DEFAULT_INPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')},
outputs={DEFAULT_OUTPUT_KEY: MockedIOInfo('FP32', [1, 1, 1], 'NCHW')})
plugin = None
batching_info = BatchingInfo(None)
shape_info = ShapeInfo(None, net.inputs)
engine = IrEngine(model_name='test',
model_version=1,
mapping_config=mapping_config,
exec_net=exec_net,
net=net,
plugin=plugin,
batching_info=batching_info,
shape_info=shape_info)
return engine