Replies: 1 comment
-
They were removed because they would only work with TF. Keras 3 focused on workflows that can be run with any backend. There is no real alternative -- my recommendation is to keep your usage of ragged tensors confined to your tf.data pipeline. You must densify/pad your tensors before you hit the model. |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
Hello,
I know that RaggedTensor were removed from keras 3. Is there an alternative? Or can we just use RaggedTensors with tensorflow backend?
Maybe a linked question but why has RaggedTensors been removed is it better not to use them? Maybe just pad out the result?
Beta Was this translation helpful? Give feedback.
All reactions