NotFoundError (see above for traceback): Key Variable not found in checkpoint

The name of the pictureThe name of the pictureThe name of the pictureClash Royale CLAN TAG#URR8PPP



NotFoundError (see above for traceback): Key Variable not found in checkpoint



When I restore a saved model using:


checkpoint = tf.train.get_checkpoint_state(config.pre_model_dir)
if checkpoint and checkpoint.model_checkpoint_path:
saver.restore(session, checkpoint.model_checkpoint_path)



, I am getting this error:


INFO:tensorflow:Restoring parameters from ./saved_model/10_zones/10/network--1685000
---------------------------------------------------------------------------
NotFoundError Traceback (most recent call last)
/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1321 try:
-> 1322 return fn(*args)
1323 except errors.OpError as e:

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run_fn(feed_dict, fetch_list, target_list, options, run_metadata)
1306 return self._call_tf_sessionrun(
-> 1307 options, feed_dict, fetch_list, target_list, run_metadata)
1308

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _call_tf_sessionrun(self, options, feed_dict, fetch_list, target_list, run_metadata)
1408 self._session, options, feed_dict, fetch_list, target_list,
-> 1409 run_metadata)
1410 else:

NotFoundError: Key Variable not found in checkpoint
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_INT32, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device_incarnation=1, tensor_name="edge_18_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]]

During handling of the above exception, another exception occurred:

NotFoundError Traceback (most recent call last)
<ipython-input-97-0cbd09927b40> in <module>()
42 checkpoint = tf.train.get_checkpoint_state(config.pre_model_dir)
43 if checkpoint and checkpoint.model_checkpoint_path:
---> 44 saver.restore(session, checkpoint.model_checkpoint_path)
45 print("loaded the model")
46 else:

/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py in restore(self, sess, save_path)
1800 else:
1801 sess.run(self.saver_def.restore_op_name,
-> 1802 self.saver_def.filename_tensor_name: save_path)
1803
1804 @staticmethod

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in run(self, fetches, feed_dict, options, run_metadata)
898 try:
899 result = self._run(None, fetches, feed_dict, options_ptr,
--> 900 run_metadata_ptr)
901 if run_metadata:
902 proto_data = tf_session.TF_GetBuffer(run_metadata_ptr)

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _run(self, handle, fetches, feed_dict, options, run_metadata)
1133 if final_fetches or final_targets or (handle and feed_dict_tensor):
1134 results = self._do_run(handle, final_targets, final_fetches,
-> 1135 feed_dict_tensor, options, run_metadata)
1136 else:
1137 results =

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_run(self, handle, target_list, fetch_list, feed_dict, options, run_metadata)
1314 if handle is None:
1315 return self._do_call(_run_fn, feeds, fetches, targets, options,
-> 1316 run_metadata)
1317 else:
1318 return self._do_call(_prun_fn, handle, feeds, fetches)

/usr/lib/python3.6/site-packages/tensorflow/python/client/session.py in _do_call(self, fn, *args)
1333 except KeyError:
1334 pass
-> 1335 raise type(e)(node_def, op, message)
1336
1337 def _extend_graph(self):

NotFoundError: Key Variable not found in checkpoint
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_INT32, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device_incarnation=1, tensor_name="edge_18_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]]

Caused by op 'save/RestoreV2', defined at:
File "/usr/lib64/python3.6/runpy.py", line 193, in _run_module_as_main
"__main__", mod_spec)
File "/usr/lib64/python3.6/runpy.py", line 85, in _run_code
exec(code, run_globals)
File "/usr/lib/python3.6/site-packages/ipykernel_launcher.py", line 16, in <module>
app.launch_new_instance()
File "/usr/lib/python3.6/site-packages/traitlets/config/application.py", line 658, in launch_instance
app.start()
File "/usr/lib/python3.6/site-packages/ipykernel/kernelapp.py", line 486, in start
self.io_loop.start()
File "/usr/lib64/python3.6/site-packages/tornado/platform/asyncio.py", line 127, in start
self.asyncio_loop.run_forever()
File "/usr/lib64/python3.6/asyncio/base_events.py", line 422, in run_forever
self._run_once()
File "/usr/lib64/python3.6/asyncio/base_events.py", line 1432, in _run_once
handle._run()
File "/usr/lib64/python3.6/asyncio/events.py", line 145, in _run
self._callback(*self._args)
File "/usr/lib64/python3.6/site-packages/tornado/platform/asyncio.py", line 117, in _handle_events
handler_func(fileobj, events)
File "/usr/lib64/python3.6/site-packages/tornado/stack_context.py", line 276, in null_wrapper
return fn(*args, **kwargs)
File "/usr/lib64/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 450, in _handle_events
self._handle_recv()
File "/usr/lib64/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 480, in _handle_recv
self._run_callback(callback, msg)
File "/usr/lib64/python3.6/site-packages/zmq/eventloop/zmqstream.py", line 432, in _run_callback
callback(*args, **kwargs)
File "/usr/lib64/python3.6/site-packages/tornado/stack_context.py", line 276, in null_wrapper
return fn(*args, **kwargs)
File "/usr/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 283, in dispatcher
return self.dispatch_shell(stream, msg)
File "/usr/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 233, in dispatch_shell
handler(stream, idents, msg)
File "/usr/lib/python3.6/site-packages/ipykernel/kernelbase.py", line 399, in execute_request
user_expressions, allow_stdin)
File "/usr/lib/python3.6/site-packages/ipykernel/ipkernel.py", line 208, in do_execute
res = shell.run_cell(code, store_history=store_history, silent=silent)
File "/usr/lib/python3.6/site-packages/ipykernel/zmqshell.py", line 537, in run_cell
return super(ZMQInteractiveShell, self).run_cell(*args, **kwargs)
File "/usr/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2662, in run_cell
raw_cell, store_history, silent, shell_futures)
File "/usr/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2785, in _run_cell
interactivity=interactivity, compiler=compiler, result=result)
File "/usr/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2903, in run_ast_nodes
if self.run_code(code, result):
File "/usr/lib/python3.6/site-packages/IPython/core/interactiveshell.py", line 2963, in run_code
exec(code_obj, self.user_global_ns, self.user_ns)
File "<ipython-input-97-0cbd09927b40>", line 26, in <module>
saver = tf.train.Saver()
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1338, in __init__
self.build()
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1347, in build
self._build(self._filename, build_save=True, build_restore=True)
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 1384, in _build
build_save=build_save, build_restore=build_restore)
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 835, in _build_internal
restore_sequentially, reshape)
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 472, in _AddRestoreOps
restore_sequentially)
File "/usr/lib/python3.6/site-packages/tensorflow/python/training/saver.py", line 886, in bulk_restore
return io_ops.restore_v2(filename_tensor, names, slices, dtypes)
File "/usr/lib/python3.6/site-packages/tensorflow/python/ops/gen_io_ops.py", line 1463, in restore_v2
shape_and_slices=shape_and_slices, dtypes=dtypes, name=name)
File "/usr/lib/python3.6/site-packages/tensorflow/python/framework/op_def_library.py", line 787, in _apply_op_helper
op_def=op_def)
File "/usr/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 3392, in create_op
op_def=op_def)
File "/usr/lib/python3.6/site-packages/tensorflow/python/framework/ops.py", line 1718, in __init__
self._traceback = self._graph._extract_stack() # pylint: disable=protected-access

NotFoundError (see above for traceback): Key Variable not found in checkpoint
[[Node: save/RestoreV2 = RestoreV2[dtypes=[DT_INT32, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, ..., DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT, DT_FLOAT], _device="/job:localhost/replica:0/task:0/device:CPU:0"](_arg_save/Const_0_0, save/RestoreV2/tensor_names, save/RestoreV2/shape_and_slices)]]
[[Node: save/RestoreV2/_21 = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device_incarnation=1, tensor_name="edge_18_save/RestoreV2", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]]



I searched about this error, and there was a tf bug which requires to call the model using a full relative path, and I followed that path and tried values: './saved_model/10_zones/10' and os.path.abspath(config.pre_model_dir+'./../saved_model/10_zones/10')
for config.pre_model_dir. Both resulted in a same error.


'./saved_model/10_zones/10'


os.path.abspath(config.pre_model_dir+'./../saved_model/10_zones/10')


config.pre_model_dir



I also checked the name of the saved variables using
from tensorflow.contrib.framework.python.framework import checkpoint_utils


var_list = checkpoint_utils.list_variables(config.pre_model_dir)
for v in var_list:
print(v)



which is:


('actor/main_net/layer1/biases/Variable', [90])
('actor/main_net/layer1/biases/Variable/Adam', [90])
('actor/main_net/layer1/biases/Variable/Adam_1', [90])
('actor/main_net/layer1/weights/Variable', [30, 90])
('actor/main_net/layer1/weights/Variable/Adam', [30, 90])
('actor/main_net/layer1/weights/Variable/Adam_1', [30, 90])
('actor/main_net/layer2/biases/Variable', [60])
('actor/main_net/layer2/biases/Variable/Adam', [60])
('actor/main_net/layer2/biases/Variable/Adam_1', [60])
('actor/main_net/layer2/weights/Variable', [90, 60])
('actor/main_net/layer2/weights/Variable/Adam', [90, 60])
('actor/main_net/layer2/weights/Variable/Adam_1', [90, 60])
('actor/main_net/layer3/biases/Variable', [30])
('actor/main_net/layer3/biases/Variable/Adam', [30])
('actor/main_net/layer3/biases/Variable/Adam_1', [30])
('actor/main_net/layer3/weights/Variable', [60, 30])
('actor/main_net/layer3/weights/Variable/Adam', [60, 30])
('actor/main_net/layer3/weights/Variable/Adam_1', [60, 30])
('actor/main_net/layer4/biases/Variable', [10])
('actor/main_net/layer4/biases/Variable/Adam', [10])
('actor/main_net/layer4/biases/Variable/Adam_1', [10])
('actor/main_net/layer4/weights/Variable', [30, 10])
('actor/main_net/layer4/weights/Variable/Adam', [30, 10])
('actor/main_net/layer4/weights/Variable/Adam_1', [30, 10])
('actor/target_net/layer1/biases/Variable', [90])
('actor/target_net/layer1/weights/Variable', [30, 90])
('actor/target_net/layer2/biases/Variable', [60])
('actor/target_net/layer2/weights/Variable', [90, 60])
('actor/target_net/layer3/biases/Variable', [30])
('actor/target_net/layer3/weights/Variable', [60, 30])
('actor/target_net/layer4/biases/Variable', [10])
('actor/target_net/layer4/weights/Variable', [30, 10])
('beta1_power', )
('beta1_power_1', )
('beta2_power', )
('beta2_power_1', )
('critic/main_net/l1/biases', [90])
('critic/main_net/l1/biases/Adam', [90])
('critic/main_net/l1/biases/Adam_1', [90])
('critic/main_net/l1/weights', [40, 90])
('critic/main_net/l1/weights/Adam', [40, 90])
('critic/main_net/l1/weights/Adam_1', [40, 90])
('critic/main_net/l2/biases', [60])
('critic/main_net/l2/biases/Adam', [60])
('critic/main_net/l2/biases/Adam_1', [60])
('critic/main_net/l2/weights', [90, 60])
('critic/main_net/l2/weights/Adam', [90, 60])
('critic/main_net/l2/weights/Adam_1', [90, 60])
('critic/main_net/l3/biases', [30])
('critic/main_net/l3/biases/Adam', [30])
('critic/main_net/l3/biases/Adam_1', [30])
('critic/main_net/l3/weights', [60, 30])
('critic/main_net/l3/weights/Adam', [60, 30])
('critic/main_net/l3/weights/Adam_1', [60, 30])
('critic/main_net/l4/bias', [1])
('critic/main_net/l4/bias/Adam', [1])
('critic/main_net/l4/bias/Adam_1', [1])
('critic/main_net/l4/kernel', [30, 1])
('critic/main_net/l4/kernel/Adam', [30, 1])
('critic/main_net/l4/kernel/Adam_1', [30, 1])
('critic/target_net/l1/biases', [90])
('critic/target_net/l1/weights', [40, 90])
('critic/target_net/l2/biases', [60])
('critic/target_net/l2/weights', [90, 60])
('critic/target_net/l3/biases', [30])
('critic/target_net/l3/weights', [60, 30])
('critic/target_net/l4/bias', [1])
('critic/target_net/l4/kernel', [30, 1])



with what tf.global_variables() in my current model results in, and they are both similar:


tf.global_variables()


<tf.Variable 'actor/main_net/layer1/weights/Variable:0' shape=(30, 90) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer1/biases/Variable:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/weights/Variable:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/biases/Variable:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/weights/Variable:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/biases/Variable:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/weights/Variable:0' shape=(30, 10) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/biases/Variable:0' shape=(10,) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer1/weights/Variable:0' shape=(30, 90) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer1/biases/Variable:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer2/weights/Variable:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer2/biases/Variable:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer3/weights/Variable:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer3/biases/Variable:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer4/weights/Variable:0' shape=(30, 10) dtype=float32_ref>,
<tf.Variable 'actor/target_net/layer4/biases/Variable:0' shape=(10,) dtype=float32_ref>,
<tf.Variable 'Variable:0' shape=() dtype=int32_ref>,
<tf.Variable 'beta1_power:0' shape=() dtype=float32_ref>,
<tf.Variable 'beta2_power:0' shape=() dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer1/weights/Variable/Adam:0' shape=(30, 90) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer1/weights/Variable/Adam_1:0' shape=(30, 90) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer1/biases/Variable/Adam:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer1/biases/Variable/Adam_1:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/weights/Variable/Adam:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/weights/Variable/Adam_1:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/biases/Variable/Adam:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer2/biases/Variable/Adam_1:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/weights/Variable/Adam:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/weights/Variable/Adam_1:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/biases/Variable/Adam:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer3/biases/Variable/Adam_1:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/weights/Variable/Adam:0' shape=(30, 10) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/weights/Variable/Adam_1:0' shape=(30, 10) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/biases/Variable/Adam:0' shape=(10,) dtype=float32_ref>,
<tf.Variable 'actor/main_net/layer4/biases/Variable/Adam_1:0' shape=(10,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/weights:0' shape=(40, 90) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/biases:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/weights:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/biases:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/weights:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/biases:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/kernel:0' shape=(30, 1) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/bias:0' shape=(1,) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l1/weights:0' shape=(40, 90) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l1/biases:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l2/weights:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l2/biases:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l3/weights:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l3/biases:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l4/kernel:0' shape=(30, 1) dtype=float32_ref>,
<tf.Variable 'critic/target_net/l4/bias:0' shape=(1,) dtype=float32_ref>,
<tf.Variable 'beta1_power_1:0' shape=() dtype=float32_ref>,
<tf.Variable 'beta2_power_1:0' shape=() dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/weights/Adam:0' shape=(40, 90) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/weights/Adam_1:0' shape=(40, 90) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/biases/Adam:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l1/biases/Adam_1:0' shape=(90,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/weights/Adam:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/weights/Adam_1:0' shape=(90, 60) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/biases/Adam:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l2/biases/Adam_1:0' shape=(60,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/weights/Adam:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/weights/Adam_1:0' shape=(60, 30) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/biases/Adam:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l3/biases/Adam_1:0' shape=(30,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/kernel/Adam:0' shape=(30, 1) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/kernel/Adam_1:0' shape=(30, 1) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/bias/Adam:0' shape=(1,) dtype=float32_ref>,
<tf.Variable 'critic/main_net/l4/bias/Adam_1:0' shape=(1,) dtype=float32_ref>



The only difference in these two lists, is <tf.Variable 'Variable:0' shape=() dtype=int32_ref>, which I do not know what is this for and how it is generated. But, I do not think if it is the problem, since any of my models that can be restored also has it.


<tf.Variable 'Variable:0' shape=() dtype=int32_ref>



I appreciate any help and comment to resolve this error.




1 Answer
1



I solved the problem by excluding that single variable using:


variables = slim.get_variables_to_restore()
variables_to_restore = [v for v in variables if 'global_step_counter' not in v.name.split('/')[0]]

saver = tf.train.Saver(variables_to_restore)






By clicking "Post Your Answer", you acknowledge that you have read our updated terms of service, privacy policy and cookie policy, and that your continued use of the website is subject to these policies.

Popular posts from this blog

Firebase Auth - with Email and Password - Check user already registered

Dynamically update html content plain JS

Creating a leaderboard in HTML/JS