I’m so excited to see all of these technologies come together in python. 10 years ago this demo would have taken 5 engineers, each with a different domain of expertise, at least a month to put together.
This is brilliant! Thank you Matthew for the wonderful tutorial. It is short and sweet. And really easy to follow. I shall be using this for a seminar/course for new students in my lab!
Matthew - I'm curious to know your thoughts on the new developments in AI. How will this change the landscape for data processing, distributed computing and all things Dask?
On the GPU part with dask, you doesn't need to specify threads per block and blocks per grid? I'm considering using dask to scale out one application of mine with multiple gpus
Your code starting at 11:17 doesn't seem to work for Windows 10 anymore. Bokeh colors the smooth, fft2, and save functions all black in erred state. I have no way of troubleshooting why, as the error doesn't output to the Jupyter notebook. :(
Just realized Jupyter console lists the error: ValueError for "Kernel launch configuration was not specified..." Maybe put an update in the description if the code has changed so much that this demo is already broken? This video isn't even a year old. Also, now I gotta figure out the fix on my own.
The uploaded version was quite high resolution. My guess is that it just takes UA-cam a bit to process the higher definition videos. It has only been a few minutes since upload time. Let's hope that things improve if we just wait.
I’m so excited to see all of these technologies come together in python. 10 years ago this demo would have taken 5 engineers, each with a different domain of expertise, at least a month to put together.
This is fantastic! Remarkable succinct presentation, thanks a lot for preparing and sharing this. All the best!
This is brilliant! Thank you Matthew for the wonderful tutorial. It is short and sweet. And really easy to follow. I shall be using this for a seminar/course for new students in my lab!
Absolutely incredible. Well spoken, and easy to understand steps. Thanks for this video.
Matthew - I'm curious to know your thoughts on the new developments in AI. How will this change the landscape for data processing, distributed computing and all things Dask?
Love your work and presentation.
On the GPU part with dask, you doesn't need to specify threads per block and blocks per grid? I'm considering using dask to scale out one application of mine with multiple gpus
Your code starting at 11:17 doesn't seem to work for Windows 10 anymore. Bokeh colors the smooth, fft2, and save functions all black in erred state. I have no way of troubleshooting why, as the error doesn't output to the Jupyter notebook. :(
Just realized Jupyter console lists the error: ValueError for "Kernel launch configuration was not specified..." Maybe put an update in the description if the code has changed so much that this demo is already broken? This video isn't even a year old. Also, now I gotta figure out the fix on my own.
Any chance for higher resolution?
The uploaded version was quite high resolution. My guess is that it just takes UA-cam a bit to process the higher definition videos. It has only been a few minutes since upload time. Let's hope that things improve if we just wait.
Yup, it's looking better now. One just has to wait an hour after uploading :)
@@MatthewRocklin Patience is a virtue... patience is a virtue! Thanks!