François Chollet
François Chollet
Cool, good work! I left comments of the parts that can be improved. I am noticing that the entire process can be automated. As such, it would be great to...
If you (@Lunatik00) or another contributor is able to open a functional & performant PR to add `webp` support, we will review and merge. Otherwise, one workaround you can try...
The issue is at the level of the `dice_loss`. Can you try producing a reproduction script that only involves the loss function? Maybe just try to backprop through the loss...
@Pouyanpi thanks! Feel free to open a PR.
> At the second check, utils.layer_utils.get_source_inputs gives a list of tensors that is causing this issue. Indexing it as below to get the first tensor solved the issue. Could you...
Hi Dimitre, Thanks for offering to write an example on TFServing + Keras, I think it would be valuable. You can just open a PR on https://github.com/keras-team/keras-io. I agree that...
Can you try a simpler reproduction that only involves `tf.Variable` instead of an entire Keras model? Thank you.
Awesome, thanks for the interest! I'd recommend starting out with a Colab notebook running on a K80 (the default Colab GPU runtime). Perhaps one of you can take the MLP...
> Just use open source repo or integrate them into a repo? We probably want to start a new 3rd party repo to host the benchmarks, in the long term....
> Hello @fchollet I am a little confused on the requirements. So are we required to compare performance of models in TensorFLow vs PyTorch or how exactly should we proceed?...