New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Eager execution breaks fit_generator in tf.keras #18287
Comments
Thank you for your post. We noticed you have not filled out the following field in the issue template. Could you update them if they are relevant in your case, or leave them as N/A? Thanks. |
Updated |
@batzner Could you use |
@fchollet Sounds like a problem. I checked the keras codes and found that generator seems to have not been supported in eager mode, right? Does anyone have worked on it? |
I updated the code to use |
Nagging Assignee @fchollet: It has been 15 days with no activity and this issue has an assignee. Please update the label and/or status accordingly. |
1 similar comment
Nagging Assignee @fchollet: It has been 15 days with no activity and this issue has an assignee. Please update the label and/or status accordingly. |
I faced a similar issue. When using |
Thanks for the bug report. I have fixed the issue and the fix will soon be available in the TF nightly release. |
…ensorflow#18287 PiperOrigin-RevId: 196171525
Just noticed that having
This is quite a strong difference. I am guessing that this is not expected ? Link to Full Notebook : here Dataset used for Training : HackerEarth ML Challenge |
System information
Describe the problem
tf.enable_eager_execution()
leads to aRuntimeError: You must compile your model before using it.
when calling Keras'smodel.fit_generator
, even if the model has already been compiled. Callingmodel.fit
works on the other hand.Source code / logs
Minimum reproducible test case:
Log:
The text was updated successfully, but these errors were encountered: