-
To commemorate our
#NeurIPS2019 poster at the program transformations workshop we put out a#PyMC4 alpha release https://github.com/pymc-devs/pymc4 Radon model: https://raw.githubusercontent.com/pymc-devs/pymc4/master/notebooks/radon_hierarchical.ipynb …pic.twitter.com/RV9V0yxwRn
-
Big announcement on
#PyMC4 (it will be based on#TensorFlow probability) as well as#PyMC3 (we will take over#Theano maintenance) https://medium.com/@pymc_devs/theano-tensorflow-and-the-future-of-pymc-6c9987bb19d5 … -
Sneak-peek into
#PyMC4 with the radon model: https://github.com/pymc-devs/pymc4/blob/3547133e16ed7b7ee4c3f562bdc417a015528fcd/notebooks/radon_hierarchical.ipynb … Still pre-alpha but if you want to play around, use this branch: https://github.com/pymc-devs/pymc4/pull/139 …pic.twitter.com/MeU0MEDCD6
-
#PyMC4 now has the coolest Makefile I've ever helped write (...in my whole 21 years of living)! Sets up dev environments, makes lints, runs tests, blackifies code... Check it out! https://github.com/pymc-devs/pymc4/blob/master/Makefile … cc@canyon289@ericmjl -
I have launched the
#PyDataPodcast! Check out Ep1 with@fonnesbeck where we talk about#ProbabilisticProgramming,#PyMC3,#PyMC4, and baseball analytics http://www.pydata-podcast.com/ It's supposed to be a conversation-based show on more advanced topics, let me know what you think! -
We submitted a paper on
#PyMC4 to#NeurIPS2019, check it out at https://openreview.net/forum?id=rkgzj5Za8H ….@ferrine96@colindcarroll@junpenglao -
In preparation for London
#PyMC4 summit, I wrote a belated developer guide of PyMC3. I tried to explain the design and implementation with more modern concept like interception (inspired by Edward2 from@dustinvtran and@migorinova's work): https://docs.pymc.io/developer_guide.html …Prikaži ovu nit -
-
We just crossed 4000 stars and 1000 forks on github. Thanks to all our awesome users and everyone who contributed! Next up: new
#PyMC3 release (coming soon) and continued work on our#PyMC4 prototype. pic.twitter.com/KRqGol4qRA
-
A big thanks to
@def_chris_suter on behalf of the @PyMC_devs team for his guidance this past week in London to help us get#PyMC4 development off to a promising start! pic.twitter.com/hFZbSqMif5
-
@avibryant Avi was it you tweeting about 'better PPLs for avoiding need to reparameterise models?' https://github.com/pymc-devs/symbolic-pymc … came up today at#pymc4. -
#PyMC4 is currently in a pre-release build, learn more about its move from#Theano to@TensorFlow next Friday at 12 Eastern on Open Source Directions! Comment on this post with questions for@fonnesbeck! Register: https://app.livestorm.co/quansight/open-source-directions-ep-18-pymc34 …#opensource@pymc_devs#PyMC3#DataScience -
A wonderfully enlightening first day at the
#pymc4 Dev summit with the#TensorFlow Probability team in Montreal!pic.twitter.com/ELp6z2nwIH
-
So excited about
#PyMC4, powered by Tensorflow Probability. Can't wait to start contributing. https://twitter.com/pymc_devs/status/1076153774027259906 …
-
Like
#NumPy as a typed array, but need more flexibility?#XND is here to help. The plan is to integrate XND with#libraries such as#Dask,#xarray,#Numba,#Chainer,#PyTorch,#PyMC4, and many more.#Free XND#webinar this#Friday @ 12 EST. Sign up here: https://events.genndi.com/register/818182175026319979/0461a278cf … -
-
Flying back to New York = four jetlagged mornings, but I hacked on
#PyMC4 in the 4am dark and it now has support for random variables so pic.twitter.com/tv7kARuRQd -
Excited for
#NeurIPS2019, will present a#PyMC4 poster in Sunday with@ferrine96. If you're at the conference and would like to chat, let me know!Prikaži ovu nit -
We also had a great
#PyMC4 code sprint in London last week and made some tangible progress.#TensorFlow Probability gives us a huge head start and allows us to really focus on API design. Thanks@def_chris_suter for joining and to@Google for sponsoring us!pic.twitter.com/QppwCtlsaW
-
Had a fantastic two days learning about Tensorflow and the new improvements to PyMC4. Great to meet
@colindcarroll and@AustinRochford et al.#PyMC4@pymc_devs
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.