François Chollet
François Chollet
I think you seem to be using Keras 2. Try to upgrade to Keras 3. Run `pip install keras --upgrade` *after* you've installed KerasNLP and TF.
> This job will fail with this error (that I've truncated): Please disable traceback filtering (`keras.config.disable_traceback_filtering()`) and rerun. I'd like to see where exactly this fails
Right now you're basically not supposed to pass `use_cudnn` in the general case -- just pass nothing, and cuDNN will be used if it's feasible. Passing `use_cudnn=True` is something you'd...
> Regardless, happy to contribute error message improvements. Sure, please open a PR if you're able. We probably need to make two backend functions, `is_config_cudnn_compatible` and `is_cudnn_available` -- or maybe...
Note that it's already standardized in https://github.com/keras-team/keras/blob/master/keras/backend/jax/distribution_lib.py#L12 and https://github.com/keras-team/keras/blob/master/keras/backend/tensorflow/distribution_lib.py#L12
Can you try JAX? I'd like to see if this is an XLA issue or a TF issue.
> Torch GPU returns RuntimeError: "baddbmm_cuda" not implemented for 'Int' For that one, you can simply cast your input to float32 (you can cast it back to int afterwards if...
Keras 3 does not work with `tf.distribute.MultiWorkerMirroredStrategy`. Whether support with be added is a question for the Keras team at Google to answer. @jeffcarp, do you know? If you problem...
This could definitely be a useful addition to `keras.ops.image`. Do standard implementations exist in torch and JAX?
Maybe? What would that look like?