kerasteam

7 ms·

1.
by kerasteam 3y ago
You can absolutely serve with Keras if your inference server is in Python. For instance, if you're looking for a basic solution, you can just set up a Flask app that calls `predict()` on a Keras model. If you're looking for a high
2.
by kerasteam 3y ago
We made sure that TFLite workflows would run smoothly with Keras 3 models. We did not come up with any TFLite related improvements. The focus was on the multi-backend architecture, distribution, and training performance.
3.
by kerasteam 3y ago
Yes, Keras can be used to build LLMs. In fact this is one of the main use cases. There are some tutorials about how to do it "from scratch", like this: https://keras.io/examples/nlp/neural_machine_transla
4.
by kerasteam 3y ago
We don't have a separate `ops.linalg` package, but we do include `numpy.linalg` ops as part of `keras.ops`. For now only 2 ops are supported: `qr` and `solve`. We're open to adding any `numpy.linalg` op that turns out to be useful
5.
by kerasteam 3y ago
All breaking changes are listed here: https://github.com/keras-team/keras/issues/18467 You can use this migration guide to identify and fix each of these issues (and further, making your code run on JAX or Py
6.
by kerasteam 3y ago
Thanks! Hope you'll find the new Keras useful! So far the export story focuses on SavedModel and the services that consume that format, e.g. TFLite, TFjs and TFServing. You can just do `model.export(path)`, and you also have access to
7.
by kerasteam 3y ago
Both Keras models/layers (with the PyTorch backend) and Lightning Modules are PyTorch Modules, so they should be able to interoperate with each other in a PyTorch workflow. We have not tried this with Lightning, but we've had a go
8.
by kerasteam 3y ago
According to PyPI downloads and user surveys (like the yearly StackOverflow survey) the two main frameworks are TensorFlow and PyTorch for Deep Learning, and Scikit-Learn for classical ML. See: https://survey.stackoverflow.co&#x2
9.
by kerasteam 3y ago
This means that the API, the abstractions, the workflows are battle-tested. The codebase itself went through 2 months of private beta and 5 months of public beta. It is already used in production by several companies. It's not as battl
10.
by kerasteam 3y ago
Yeah, that never happened. We process dozens of bug reports and feature requests every week, and we listen to them.
11.
by kerasteam 3y ago
To clarify, I have never attacked PyTorch, on Twitter or otherwise. What happened is that I was a target of online harassment campaign from 2017 to January 2021 (when it stopped abruptly), which originated from a PyTorch developer and took
12.
by kerasteam 3y ago
Francois from the Keras team here -- happy to answer questions!
13.
by kerasteam 3y ago
That's right, if the model is backend-agnostic you can train it with a PyTorch training loop and then reload it and use it with TF ecosystem tools, like serve it with TF-Serving or export it to Coral TPU.
14.
by kerasteam 3y ago
Yes, you can check out KerasCV and KerasNLP which host pretrained models like ResNet, BERT, and many more. They run on all backends as of the latest releases (today), and converting them to be backend-agnostic was pretty smooth! It took a c
15.
by kerasteam 3y ago
For a Keras Core model to be usable with the TF Serving ecosystem, it must be implemented either via Keras APIs (Keras layers and Keras ops) or via TF APIs. To use pretrained models, you can take a look at KerasCV and KerasNLP, they have al
16.
by kerasteam 3y ago
Yes, model weights saved with Keras Core are backend-agnostic. You can train a model in one backend and reload it in another. Coral TPU could be used with Keras Core, but via the TensorFlow backend only.
17.
by kerasteam 3y ago
I worked on the project, happy to answer any questions!