def __init__(self, input_sources):
if utils.is_anystr(input_sources):
input_sources = self._parse(input_sources)
if isinstance(input_sources, dict):
input_sources = self._resolve_multikey(input_sources)
if input_sources is None:
input_sources = DefaultInput()
self.input_sources = input_sources
def load_configs(module_names="", load_standard=True):
# type: (typing.Union[str, typing.List[str], None], bool)->typing.List[NNEFParserConfig]
"""
:param module_names: "package.module" or "p1.m1,p2.m2", or ["p1.m1", "p2.m2"]
:param load_standard: Load the standard NNEF op definitions as well
:return: parser configs
"""
if module_names is None:
module_names = []
if utils.is_anystr(module_names):
module_names = [name.strip() for name in module_names.split(',')
] if module_names.strip() else []
configs = [NNEFParserConfig.STANDARD_CONFIG] if load_standard else []
for module_name in module_names:
config = NNEFParserConfig.load_config(module_name)
if not config.empty:
configs.append(config)
return configs
def _normalize_types(arg):
if utils.is_anyint(arg):
return utils.anyint_to_int(arg)
elif utils.is_anystr(arg):
return utils.anystr_to_str(arg)
elif isinstance(arg, np.ndarray):
return arg.tolist()
elif isinstance(arg, tf.TensorShape):
if arg.dims is None:
return None
return [None if dim is None else int(dim) for dim in arg.as_list()]
elif isinstance(arg, tf.Dimension):
return arg.value
elif isinstance(arg, tf.DType):
return arg.name
elif isinstance(
arg,
(np.int8, np.int16, np.int32, np.int64, np.uint8, np.uint16, np.uint32,
np.uint64, np.float16, np.float32, np.float64, np.bool_)):
return arg.item()
else:
return arg
def eval_functions(self):
self.functions = [
self._eval_function(fun) if utils.is_anystr(fun) else fun
for fun in self.functions
]