TensorFlow nerds: If I want to measure Keras model inference time, can I do it by just sticking it in a for loop after a warmup? Or is there some other synchronization I should be doing?
-
3) A strange experience where ReLU kernels were being called from somewhere in a Keras layer when we weren't using ReLU (results were correct). This was fixed already IIRC
Thanks for the feedback, we'll review it with the team. I expect most of these were already fixed. There were known performance issues in 2.0 but they were generally fixed by 2.1
