Getting error "AttributeError: 'RefVariable' object has no attribute '_id'" running TFP example

[shashankg7]

Hi, I am running the DPMM example from the repository on Google Colab.

During running the optimizer step, I am getting the following error:

---

AttributeError Traceback (most recent call last)
AttributeError: 'RefVariable' object has no attribute '_id'

The above exception was the direct cause of the following exception:

SystemError                               Traceback (most recent call last)
<ipython-input-9-e863f37ccd7f> in <module>()
     17     data_size=num_samples)
     18 
---> 19 train_op = optimizer_kernel.minimize(-joint_log_prob, var_list=training_vals)
     20 
     21 # Arrays to store samples

3 frames
/usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/tape.py in watch(tape, tensor)
     57 def watch(tape, tensor):
     58   """Marks this tensor to be watched by the given tape."""
---> 59   pywrap_tensorflow.TFE_Py_TapeWatch(tape._tape, tensor)  # pylint: disable=protected-access
     60 
     61 

SystemError: <built-in function TFE_Py_TapeWatch> returned a result with an error set

I am not able to figure out the cause of this. Kindly suggest a solution.

[brianwa84]

This seems like a tf2 issue. We have not yet updated all the examples for tf2.

…

On Wed, Oct 9, 2019, 7:12 AM Shashank Gupta ***@***.***> wrote: Hi, I am running the DPMM example from the repository on Google Colab. During running the optimizer step, I am getting the following error: ------------------------------ AttributeError Traceback (most recent call last) AttributeError: 'RefVariable' object has no attribute '_id' The above exception was the direct cause of the following exception: SystemError Traceback (most recent call last) <ipython-input-9-e863f37ccd7f> in <module>() 17 data_size=num_samples) 18 ---> 19 train_op = optimizer_kernel.minimize(-joint_log_prob, var_list=training_vals) 20 21 # Arrays to store samples 3 frames /usr/local/lib/python3.6/dist-packages/tensorflow_core/python/eager/tape.py in watch(tape, tensor) 57 def watch(tape, tensor): 58 """Marks this tensor to be watched by the given tape.""" ---> 59 pywrap_tensorflow.TFE_Py_TapeWatch(tape._tape, tensor) # pylint: disable=protected-access 60 61 SystemError: <built-in function TFE_Py_TapeWatch> returned a result with an error set I am not able to figure out the cause of this. Kindly suggest a solution. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#592?email_source=notifications&email_token=AFJFSI3N6DERIZ3OPBLBOVDQNW4AXA5CNFSM4I653WIKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HQTJIGQ>, or mute the thread <https://github.com/notifications/unsubscribe-auth/AFJFSI5KNWRN2XNBS2FPBPDQNW4AXANCNFSM4I653WIA> .

[shashankg7]

Thanks for your prompt reply.

I tried running the same example on my local machine, with tensorflow 1.15.0rc0. I am getting the same error again

[shashankg7]

Update:

getting the same error with Tensorflow 1.14.0 and TFP0.7.0

Could you please help in resolving it?

[dynamicwebpaige]

Can confirm that a TypeError is raised in the DPMM example when using TF 1.15.0 and TFP 0.7.0; this doesn't look like it would be a TensorFlow 2.0 issue.

TypeError                                 Traceback (most recent call last)
<ipython-input-9-47055b26cc6d> in <module>()
     17     data_size=num_samples)
     18 
---> 19 train_op = optimizer_kernel.minimize(-joint_log_prob)
     20 
     21 # Arrays to store samples

TypeError: minimize() missing 1 required positional argument: 'var_list'

[skeydan]

In TF 2, optimizer.minimize needs to be passed a var_list in addition to the loss.

Also interested in this because I have some pre-TF2 unit tests that I would like to update if possible :-)
(For now, I've disabled them ;-) - see #520)

[shimaimani]

Hello,
Can you please update the notebook, I cannot run the notebook? I tried different versions of tfp and also tried different versions of TensorFlow but I still get a lot of error. Can you post a working colab version of the code ( please specify which version of tfp and tf you are using)?