Romain Picard

[Romain Picard @_mainro_]

Maki Nage is now in a good shape, and if you work on stream processing, give it a try!
@ReactiveX and @apachekafka for data science.
@PythonWeekly

https://t.co/NiPWY5WITT

[Romain Picard @_mainro_]

A new major release of deepspeech-server is available. It now uses @coqui_ai as the STT backend.

https://t.co/Apr8w118Ew

[Romain Picard @_mainro_]

Dig into the brand new RxPY v4 release with this spring offer from @PacktPub. Hands-On Reactive Programming With Python is only 10€.

https://t.co/WEgpi0ntPk

[Romain Picard @_mainro_]

@ravinwashere It was action cards that you put in a reader to make the turtle draw on the floor. It's never too early to enjoy programming and robotic.

https://t.co/bZDZFCXoGh

[Romain Picard @_mainro_]

I finally found some time to complete the GAN specialization by @realSharonZhou. It is definitely in the top 3 online courses that I followed.

Eager to test a cycleGAN on audio instruments!

[Romain Picard @_mainro_]

@mattpodwysocki @sgoguen Could you share more information on that? I never heard of bonsai or Prisma but I am looking for a unified reactivex spec that could compile/execute on different frameworks (rx-xxx, beam, Kafka...).
I'd be pleased to exchange on that topic.

[Romain Picard @_mainro_]

I just noticed that my GitHub repo of "reactive programming with python" is the fourth most starred of @PacktPub's reactive programming books.
Everybody interested in @ReactiveX can look at it and find many examples, up to date with #rxpy v3:
https://t.co/xHoepcWAje

[Romain Picard @_mainro_]

New year resolution idea: Learn Reactive Programming. The 5$ campaign of @PacktPub is still running. So start from the base concepts up to mastering it with my book:
https://t.co/Qslx69PDg5

[Romain Picard @_mainro_]

@slava_pestov BeOS was years ahead of it's time. There is still no scheduler/system able to keep the UI responsive when all CPUs are 100% used.

[Romain Picard @_mainro_]

Maki Nage 0.8 is out! This release contains initial support for custom connectors. Check out the documentation.

https://t.co/39s4EeiA8o

[Romain Picard @_mainro_]

An article on Distogram. This maki-nage components computes composable histograms on streaming data.

[KDnuggets @kdnuggets]

Your data. Streaming in real-time. Visualized. https://t.co/xDqw2GPCIs @_mainro_

[Romain Picard @_mainro_]

#rxpy reached 4K stars. Thanks to all the stargazers!

For what kind of application do you use RxPY?
What do you miss most in the library?

Reply here to make @ReactiveX on python even better in the future.

https://t.co/8M5wnIXCeU

[Romain Picard @_mainro_]

Distributed computed with RxPY is now very easy. Try out RxRay, an operator to distribute a RxPY pipeline on a ray cluster. @ReactiveX @raydistributed #rxpy
https://t.co/ST1vd36FWU

[Romain Picard @_mainro_]

@wangzj48558765 I keep this article online as an archive only. I do not work on such topics anymore.
Netflix does not use the googlecast protocol but it's own, based on dial for device discovery.

[Romain Picard @_mainro_]

Yes! This is what I've been waiting for since the begining of type hinting in python. I bet that python performance will dramatically increase in the next few years.
https://t.co/2J4BZyeLRY

[Romain Picard @_mainro_]

@__ryanolson__ you have this issue when adding other "True" items to the "rx.from_" observable? it works for me. Can you try with 3.2.0? We fixed some issues with the delay operator.

[Romain Picard @_mainro_]

#rxpy v3.2.0 is out!
@ReactiveX

https://t.co/78uOLqOyG1

[Romain Picard @_mainro_]

@EmadMokhtar Sorry for the - very - late answer. I just saw your question. You can look at the aiokafka driver here:
https://t.co/oIumtbLLiZ

It is used in makinage (https://t.co/VctcJcYqWi). Warning; I will probably rewrite it soon.

[Romain Picard @_mainro_]

After long procrastination, there it is: How Explainable Boosting Machines - from @MSFTResearch - can be efficient and explainable.
#explainableai

https://t.co/bp60Dl30Tl

[Romain Picard @_mainro_]

We just released an ONNX converter for the only machine learning algorithm that combines high performances and full explainability: Explainable Boosting Machines (aka EBM)
@onnxai @onnxruntime
https://t.co/Cgq6kjv7Cj

[Romain Picard @_mainro_]

@dbrattli Bad and good news at the same time. I wish you to fully recover quickly.