Can not load saved BlazePose model.

[saurabh1043]

model = keras.models.load_model('saved_model_pose_detection')

---

TypeError Traceback (most recent call last)
in
----> 1 model = keras.models.load_model('saved_model_pose_detection')

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\save.py in load_model(filepath, custom_objects, compile)
188 if isinstance(filepath, six.string_types):
189 loader_impl.parse_saved_model(filepath)
--> 190 return saved_model_load.load(filepath, compile)
191
192 raise IOError(

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in load(path, compile)
114 # TODO(kathywu): Add saving/loading of optimizer, compiled losses and metrics.
115 # TODO(kathywu): Add code to load from objects that contain all endpoints
--> 116 model = tf_load.load_internal(path, loader_cls=KerasObjectLoader)
117
118 # pylint: disable=protected-access

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\saved_model\load.py in load_internal(export_dir, tags, loader_cls)
602 loader = loader_cls(object_graph_proto,
603 saved_model_proto,
--> 604 export_dir)
605 root = loader.get(0)
606 if isinstance(loader, Loader):

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in init(self, *args, **kwargs)
186 self._models_to_reconstruct = []
187
--> 188 super(KerasObjectLoader, self).init(*args, **kwargs)
189
190 # Now that the node object has been fully loaded, and the checkpoint has

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\saved_model\load.py in init(self, object_graph_proto, saved_model_proto, export_dir)
121 self._concrete_functions[name] = _WrapperFunction(concrete_function)
122
--> 123 self._load_all()
124 self._restore_checkpoint()
125

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in _load_all(self)
207 # loaded from config may create variables / other objects during
208 # initialization. These are recorded in _nodes_recreated_from_config.
--> 209 self._layer_nodes = self._load_layers()
210
211 # Load all other nodes and functions.

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in _load_layers(self)
307 continue
308
--> 309 layers[node_id] = self._load_layer(proto.user_object, node_id)
310
311 for node_id, proto in metric_list:

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in _load_layer(self, proto, node_id)
333 # Detect whether this object can be revived from the config. If not, then
334 # revive from the SavedModel instead.
--> 335 obj, setter = self._revive_from_config(proto.identifier, metadata, node_id)
336 if obj is None:
337 obj, setter = revive_custom_object(proto.identifier, metadata)

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in _revive_from_config(self, identifier, metadata, node_id)
351 obj = (
352 self._revive_graph_network(metadata, node_id) or
--> 353 self._revive_layer_from_config(metadata, node_id))
354
355 if obj is None:

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\saving\saved_model\load.py in _revive_layer_from_config(self, metadata, node_id)
406 try:
407 obj = layers_module.deserialize(
--> 408 generic_utils.serialize_keras_class_and_config(class_name, config))
409 except ValueError:
410 return None

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\layers\serialization.py in deserialize(config, custom_objects)
107 module_objects=globs,
108 custom_objects=custom_objects,
--> 109 printable_module_name='layer')

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\utils\generic_utils.py in deserialize_keras_object(identifier, module_objects, custom_objects, printable_module_name)
373 list(custom_objects.items())))
374 with CustomObjectScope(custom_objects):
--> 375 return cls.from_config(cls_config)
376 else:
377 # Then cls may be a function returning a class.

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\engine\base_layer.py in from_config(cls, config)
653 A layer instance.
654 """
--> 655 return cls(**config)
656
657 def compute_output_shape(self, input_shape):

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\layers\convolutional.py in init(self, filters, kernel_size, strides, padding, data_format, dilation_rate, activation, use_bias, kernel_initializer, bias_initializer, kernel_regularizer, bias_regularizer, activity_regularizer, kernel_constraint, bias_constraint, **kwargs)
597 kernel_constraint=constraints.get(kernel_constraint),
598 bias_constraint=constraints.get(bias_constraint),
--> 599 **kwargs)
600
601

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\layers\convolutional.py in init(self, rank, filters, kernel_size, strides, padding, data_format, dilation_rate, activation, use_bias, kernel_initializer, bias_initializer, kernel_regularizer, bias_regularizer, activity_regularizer, kernel_constraint, bias_constraint, trainable, name, **kwargs)
123 name=name,
124 activity_regularizer=regularizers.get(activity_regularizer),
--> 125 **kwargs)
126 self.rank = rank
127 if filters is not None and not isinstance(filters, int):

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\training\tracking\base.py in _method_wrapper(self, *args, **kwargs)
454 self._self_setattr_tracking = False # pylint: disable=protected-access
455 try:
--> 456 result = method(self, *args, **kwargs)
457 finally:
458 self._self_setattr_tracking = previous_value # pylint: disable=protected-access

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\engine\base_layer.py in init(self, trainable, name, dtype, dynamic, **kwargs)
292 }
293 # Validate optional keyword arguments.
--> 294 generic_utils.validate_kwargs(kwargs, allowed_kwargs)
295
296 # Mutable properties

~\miniconda3\envs\py36\lib\site-packages\tensorflow\python\keras\utils\generic_utils.py in validate_kwargs(kwargs, allowed_kwargs, error_message)
790 for kwarg in kwargs:
791 if kwarg not in allowed_kwargs:
--> 792 raise TypeError(error_message, kwarg)
793
794

TypeError: ('Keyword argument not understood:', 'groups')

[PINTO0309]

Are you using Tensorflow v2.3.0?

[saurabh1043]

I'm using Tensorflow 2.2.0

[PINTO0309]

I'm not familiar with miniconda, but at least Tensorflow v2.3.0 was able to load it successfully. See below.

$ python3

Python 3.6.9 (default, Jul 17 2020, 12:50:27) 
[GCC 8.4.0] on linux
Type "help", "copyright", "credits" or "license" for more information.

>>> import tensorflow as tf
2020-08-15 13:59:29.183339: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1

>>> tf.__version__
'2.3.0'

>>> imported = tf.saved_model.load('saved_model_pose_detection')

2020-08-15 13:59:47.500567: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcuda.so.1
2020-08-15 13:59:47.515226: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.515729: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: GeForce GTX 1070 with Max-Q Design computeCapability: 6.1
coreClock: 1.2655GHz coreCount: 16 deviceMemorySize: 7.93GiB deviceMemoryBandwidth: 238.66GiB/s
2020-08-15 13:59:47.515749: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-15 13:59:47.516876: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-15 13:59:47.517969: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-15 13:59:47.518149: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-15 13:59:47.519280: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-15 13:59:47.519932: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-15 13:59:47.522420: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-15 13:59:47.522520: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.523062: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.523530: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-15 13:59:47.523739: I tensorflow/core/platform/cpu_feature_guard.cc:142] This TensorFlow binary is optimized with oneAPI Deep Neural Network Library (oneDNN)to use the following CPU instructions in performance-critical operations:  AVX2 FMA
To enable them in other operations, rebuild TensorFlow with the appropriate compiler flags.
2020-08-15 13:59:47.549960: I tensorflow/core/platform/profile_utils/cpu_utils.cc:104] CPU Frequency: 2199995000 Hz
2020-08-15 13:59:47.551139: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x82d3780 initialized for platform Host (this does not guarantee that XLA will be used). Devices:
2020-08-15 13:59:47.551159: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): Host, Default Version
2020-08-15 13:59:47.597910: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.598558: I tensorflow/compiler/xla/service/service.cc:168] XLA service 0x833f190 initialized for platform CUDA (this does not guarantee that XLA will be used). Devices:
2020-08-15 13:59:47.598575: I tensorflow/compiler/xla/service/service.cc:176]   StreamExecutor device (0): GeForce GTX 1070 with Max-Q Design, Compute Capability 6.1
2020-08-15 13:59:47.598817: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.599346: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1716] Found device 0 with properties: 
pciBusID: 0000:01:00.0 name: GeForce GTX 1070 with Max-Q Design computeCapability: 6.1
coreClock: 1.2655GHz coreCount: 16 deviceMemorySize: 7.93GiB deviceMemoryBandwidth: 238.66GiB/s
2020-08-15 13:59:47.599367: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-15 13:59:47.599384: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcublas.so.10
2020-08-15 13:59:47.599394: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcufft.so.10
2020-08-15 13:59:47.599405: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcurand.so.10
2020-08-15 13:59:47.599434: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusolver.so.10
2020-08-15 13:59:47.599462: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcusparse.so.10
2020-08-15 13:59:47.599509: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudnn.so.7
2020-08-15 13:59:47.599582: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.600075: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:47.600540: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1858] Adding visible gpu devices: 0
2020-08-15 13:59:47.600559: I tensorflow/stream_executor/platform/default/dso_loader.cc:48] Successfully opened dynamic library libcudart.so.10.1
2020-08-15 13:59:48.052606: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1257] Device interconnect StreamExecutor with strength 1 edge matrix:
2020-08-15 13:59:48.052637: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1263]      0 
2020-08-15 13:59:48.052643: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1276] 0:   N 
2020-08-15 13:59:48.052835: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:48.053397: I tensorflow/stream_executor/cuda/cuda_gpu_executor.cc:982] successful NUMA node read from SysFS had negative value (-1), but there must be at least one NUMA node, so returning NUMA node zero
2020-08-15 13:59:48.053911: I tensorflow/core/common_runtime/gpu/gpu_device.cc:1402] Created TensorFlow device (/job:localhost/replica:0/task:0/device:GPU:0 with 7025 MB memory) -> physical GPU (device: 0, name: GeForce GTX 1070 with Max-Q Design, pci bus id: 0000:01:00.0, compute capability: 6.1)
>>> 

[saurabh1043]

Updated to Tensorflow v2.3.0 and it's working now. Thanks :)