> When we decided to rename part of the IPython project to Jupyter in 2014, we had many good reasons. Our goal was to make (Data)Science and Education better, by providing Free and Open-Source tools that can be used by everyone. The name “Jupyter” is a strong reference to Galileo, who detailed his discovery of the Moons of Jupiter in his astronomical notebooks. The name is also a play on the languages Julia, Python, and R, which are pillars of the modern scientific world. While we (Love Python), and use it for much of the architecture in Jupyter, we believe that all open-source languages have an important role in scientific and data analysis workflows. We have strived to make Jupyter a platform that treats all open-source languages as first-class citizens.
Keyword is imagine, because with Mojo you do have to write your code in a new language. You cannot use the Python parts if you want speed. That's writing code in a new language.
i find myself thinking that numerous times a day (not on HN only), and it occurs to me that this acronym might itself not be known, since you'd have to RTFM to figure it itself out.
0:16 from https://www.youtube.com/watch?v=mQh9es5gfpo, says "we'll be open sourcing documentation"... by the end of the year. Documentation? It will have taken a full year to open source documentation. Congrats.
Agreed. Fish sticks must be a regional thing, maybe East Coast where more fried fish is common. The only cheap fish of this sort that I see around the Portland area isn't fried, it's tilapia.
This still doesn't fix the fact that PyTorch libraries like torchdiffeq tend to benchmark orders of magnitude faster than other libraries (1) and then devs just write off the performance issues (2)
A more charitable view on the PyTorch forums thread is that Ed's explanation "torch.jit, at this point in time, is not designed to take pointwise loops as you’ve written here, and compile them into machine code directly." is a description of "this is the feature you're implicitly asking about, and it isn't implemented in PyTorch yet". What would have been a better answer in your view?
It would seem btw that 1.5 years later, people are working on implementing generated kernels for reductions, even if it is still somewhat experimental.
The speed comparison, which is about 9 months old, is marginally related. The issue here is optimization of pointwise operations, which would be handled by the JIT fuser, except it is disabled by default on the CPU. The latest version of the benchmark code seems to not run on recent PyTorch versions as given. It still still a fair comparison in terms of it is what a user will get by default. I won't be the first PyTorch developer to say that Julia and its libs do a great job at JITed optimizations. Nonetheless I'm relatively certain that a determined PyTorch user would find ways to get that a better optimization of that ODE step using some of the disabled by default features.
The other truth, of course, is that for PyTorch, there still is more emphasis on GPU when it comes to implementing optimizations.
(Disclaimer: I'm one of the people on that thread.)
Indeed, this is the same terrible idea I see whenever my next job wants me to switch to the new latest and greatest javascript framework, or the new latest and greatest ML framework. If you never have time to build up the simple knowledge you keep wasting time.
> When we decided to rename part of the IPython project to Jupyter in 2014, we had many good reasons. Our goal was to make (Data)Science and Education better, by providing Free and Open-Source tools that can be used by everyone. The name “Jupyter” is a strong reference to Galileo, who detailed his discovery of the Moons of Jupiter in his astronomical notebooks. The name is also a play on the languages Julia, Python, and R, which are pillars of the modern scientific world. While we (Love Python), and use it for much of the architecture in Jupyter, we believe that all open-source languages have an important role in scientific and data analysis workflows. We have strived to make Jupyter a platform that treats all open-source languages as first-class citizens.