TypeError: _lookup_dependency() takes 2 positional arguments but 3 were given

[ngbusca]

Issue type

Bug

Have you reproduced the bug with TensorFlow Nightly?

Yes

Source

source

TensorFlow version

2.14.0-dev20230712

Custom code

Yes

OS platform and distribution

Linux moe 5.10.0-12-amd64 #1 SMP Debian 5.10.103-1 (2022-03-07) x86_64 GNU/Linux

Mobile device

No response

Python version

3.9

Bazel version

No response

GCC/compiler version

No response

CUDA/cuDNN version

No response

GPU model and memory

No response

Current behavior?

Can't load saved model

Standalone code to reproduce the issue

Saved models can't be loaded:


model = tf.keras.models.Sequential([tf.keras.layers.Input((256,256,3)), tf.keras.layers.Dense(1)])
model.save("../models/test")
model = tf.keras.models.load_model("../models/test/")

same thing with more complex models

model = tf.keras.applications.efficientnet.EfficientNetB0()
model.save("../models/test")
model = tf.keras.models.load_model("../models/test/")

### Relevant log output

```shell
---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
Cell In[11], line 1
----> 1 model = tf.keras.models.load_model("../models/test/")

File ~/.local/lib/python3.9/site-packages/keras/src/saving/saving_api.py:262, in load_model(filepath, custom_objects, compile, safe_mode, **kwargs)
    254     return saving_lib.load_model(
    255         filepath,
    256         custom_objects=custom_objects,
    257         compile=compile,
    258         safe_mode=safe_mode,
    259     )
    261 # Legacy case.
--> 262 return legacy_sm_saving_lib.load_model(
    263     filepath, custom_objects=custom_objects, compile=compile, **kwargs
    264 )

File ~/.local/lib/python3.9/site-packages/keras/src/utils/traceback_utils.py:70, in filter_traceback.<locals>.error_handler(*args, **kwargs)
     67     filtered_tb = _process_traceback_frames(e.__traceback__)
     68     # To get the full stack trace, call:
     69     # `tf.debugging.disable_traceback_filtering()`
---> 70     raise e.with_traceback(filtered_tb) from None
     71 finally:
     72     del filtered_tb

File ~/.local/lib/python3.9/site-packages/tensorflow/python/checkpoint/restore.py:606, in _queue_children_for_restoration(checkpoint_position, visit_queue)
    604   continue
    605 child_position = checkpoint_position.create_child_position(child.node_id)
--> 606 local_object = trackable._lookup_dependency(child.local_name,
    607                                             trackable_children)
    608 child_proto = child_position.object_proto
    609 if local_object is None:
    610   # We don't yet have a dependency registered with this name. Save it
    611   # in case we do.

TypeError: _lookup_dependency() takes 2 positional arguments but 3 were given

[SuryanarayanaY]

Hi @ngbusca ,

I have tested the given code with tf-nightl(2.14.0-dev20230712) and its working fine for me. I can able to save and reload the model. Could you please check the attached gist and let me know whether the code is right one to reproduce the reported behaviour.

Thanks!

[ngbusca]

Hi @SuryanarayanaY
Thanks for the response. Yes, the code is correct. I've downgraded to 2.13 and don't encounter the issue. I must have had an inconsistency on my system, I'll close the issue and reopen if I encounter it again.

[google-ml-butler]

Are you satisfied with the resolution of your issue?
Yes
No