-
Notifications
You must be signed in to change notification settings - Fork 19.4k
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
does Functional Model support "string" dtype for end-to-end pipelines? #18410
Comments
The dtype |
@Mrutyunjay01, Could you please provide sample reproducible code, the tutorial you have mentioned has other error to solve before we get to the reported issue part. |
Apologies for late response. I faced the issue when tried to port BERT MLM to backend agnostic. Been a while, let me refine the port code once, and see if the issue persists still. cc: @sachinprasadhs |
Update: The issue no longer persists in keras 3.3. So, we can close this. Currently I am trying to port BERT MLM to backend agnostic, will report if I come across any such issues. |
Reopening the issue as faced in the draft PR mentioned. |
As mentioned keras-team/keras-core#843, I was trying further modify the mlm-bert pipeline for keras-core with backend-agnostic support. The last part of the example tries to create an end-to-end pipeline with raw texts as the input to the model. Mentioned as follows:
But while executing with
jax
backend, it throws the following error:Is having the input layer of dtype `string' might cause the problem here?
The text was updated successfully, but these errors were encountered: