3 ms·
I guess they realized muilti-backend keras is futile? I never liked the tf.keras apis and the docs always promosed multi backend but then I guess they were neve
by tadeegan 2y ago
I guess they realized muilti-backend keras is futile? I never liked the tf.keras apis and the docs always promosed multi backend but then I guess they were never able to deliver that without breaking keras 3 changes. And even now.... "Keras 3 includes a brand new distribution API, the keras.distribution namespace, currently implemented for the JAX backend (coming soon to the TensorFlow and PyTorch backends)". I don't believe it. They are too different to reconcile under 1 api. And even if you could, I dont really see the benefit. Torch and Flax have similar goals to Keras and are imo better.
- modeless 2y agoWhy would you interpret this as Google disliking Keras? Seems a lot more likely he was poached by Anthropic.
- blackeyeblitzar 2y agoWhere did you see that he was poached by Anthropic?
- modeless 2y agoI am not suggesting that I know it for a fact. I do recall some speculation on X to that effect but I can't find it now. Maybe just because Anthropic has been getting a lot of people lately.
- modeless 2y agoMystery solved: he's founding a startup: https://x.com/fchollet/status/1857012265024696494 https://x.com/fchollet/status/1857012265024696494
- hedgehog 2y agoMulti-backend Keras was great the first time around and it might be a more widely used API today if the TF team hadn't pulled that support and folded Keras into TF. I'm sure they had their reasons but I suspect that decision directly increased the adoption of PyTorch.
- fchollet 2y agoActually, `keras.distribution` is straightforward to implement in TF DTensor and with the experimental PyTorch SPMD API. We haven't done it yet first because these APIs are experimental (only JAX is mature) and second because all the demand for large-model distribution at Google was towards the JAX backend.