Tirth Patel

Yes, you can use PyCOCOMetrics to do this. See this guide: https://keras.io/guides/keras_cv/object_detection_keras_cv 

I tried to run [your notebook](https://colab.research.google.com/drive/1-z5bm5WqnPxYp6Hl4oAyC0ZwkTOvtRzd?usp=sharing), @sayakpaul, and it just seems to be stuck on the generation step. Seems like Stable Diffision might be broken on TPU. Will try to...

Hi @mostafamdy, it seems like the guide is outdated. Thanks for bringing this up! You can refer to the [Sampler API docs](https://keras.io/api/keras_nlp/samplers/) or the "Example Use" section on the [Kaggle...

Ah OK. Your code looks good to me. You seem to be printing out the next token predictions for **each** input sequence which is why I guess the outputs are...

Hi Chen, > Here `ops.split(tensor, 2, axis=self.feature_axis)` should be splitting the tensor by the middle, i.e., `[0, 1, 2, 3] => [0, 1], [2, 3]`, but from the paper it...

@mfidabel Are you still interested in contributing the model? We can help review in case you want to create a PR for it.

> or current implementation only for inference? Only for inference when I wrote it. I never tested training so I wonder how much effort is required to support it. I...

> is there any vendored code under `scipy/stats/_unuran`, or is that all written by you, with the vendored code just in the git submodule? Yes, none of the code under...