swagonomixxx

11 ms·

61.
by swagonomixxx 6y ago
It's a lockdown. I could see myself spending the weekend doing this.
62.
by swagonomixxx 6y ago
I just love how this whole sequence of events is being touted as the "underdog sticking it to the man". What I see is a bunch of internet trolls getting together and trolling Wall Street. Is that a bad thing? It depends. If you&#x
63.
by swagonomixxx 6y ago
As someone who knows very little but wants to learn: what's the best way to get up and running with Julia? On the website, they link a lot of videos, but I prefer textual formats. Is there something like the Rust Book for Julia?
64.
by swagonomixxx 6y ago
For (2), it seems like HTTP.jl [0] is the equivalent. [0]: https://github.com/JuliaWeb/HTTP.jl
65.
by swagonomixxx 6y ago
> A regular syntax graph of regular Julia code gets manipulated to do autodiff. You are not organizing C++ nodes in a network. That does sound very different than Python. Admittedly, I haven't used Julia, because I never had to for
66.
by swagonomixxx 6y ago
Thanks, I wasn't aware this existed!
67.
by swagonomixxx 6y ago
I am totally with you on statically typed vs. dynamically typed languages _in general_. I'm not denying that there are major advantages to using one for most kinds of programs. But I just don't see a benefit for ML. You can easily
68.
by swagonomixxx 6y ago
> Most of the community finds python sufficient because they do not need to interleave training and simulating. That's kind of my point. Most of the community has models running on GPU's and don't care too much about CPU-b
69.
by swagonomixxx 6y ago
So nobody needs TensorFlow, yet it and PyTorch are pretty much the only frameworks widely adopted for ML, and they're in Python. A quick Google search will tell you that, I don't really feel like rebutting to a baseless statement.
70.
by swagonomixxx 6y ago
> It's definitely not possible. These have to be rewritten with the specific autodiff /ML framework in mind. I don't understand why it's not possible. You're asking if it's possible in the language. I don&#x
71.
by swagonomixxx 6y ago
I'm fairly certain everything you said is possible except for custom CUDA kernels in pure Python. You'd have to write the kernel in C++ and use it in your TensorFlow/PyTorch code. [0][1] [0]: https://www.tensorflow
72.
by swagonomixxx 6y ago
Sure, that's fine. But I'm assuming GGP is talking about TensorFlow, since that's what this post is about. If you don't need TensorFlow, then this whole conversation is kinda moot, and you can use whatever you like.
73.
by swagonomixxx 6y ago
Well, there is already Tensorflow.js [0]. [0]: https://www.tensorflow.org/js .
74.
by swagonomixxx 6y ago
> Having a half way decent type system solves so many problems What problems does it solve for you? When building ML applications I mean. Because that's what S4TF was supposed to be. ML researchers produce models that are used by ML
75.
by swagonomixxx 6y ago
It doesn't matter that "Swift is faster than Python". Your TensorFlow graph is gonna be slow as all hell if it's running on the CPU. The bottleneck here really isn't performance on a CPU. That's a no-brainer fo
76.
by swagonomixxx 6y ago
> Python has exceptional libraries but, as a language, it's a bit dated on several fronts. I've been using Python for ML for the last 3 years and I've never felt this way. It might be that I'm not all about the hip ne
77.
by swagonomixxx 6y ago
I'm not sure I understand what your getting at. If Python's performance in the ML space is not sufficient, then the community would have quickly moved on from it and built something better. And that something better is certainly n
78.
by swagonomixxx 6y ago
I mean, I'm fairly certain that this project is the reason Chris Lattner joined Google; it was somehow in his contract somewhere. He just found a way to shoehorn himself in there and make sure that the language that he designed (Swift)
79.
by swagonomixxx 6y ago
Type hints are not perfect, but IMO after using them for a while, they are significantly better than nothing. If you do the legwork, you can get mypy[0] type-checking done on your codebase in a way that is similar to how TypeScript does typ
80.
by swagonomixxx 6y ago
I like graph theory a lot (even though my prof in uni was horrible at teaching it). One thing I like about it is at least we can easily visualize most graphs, as opposed to say, n-dimensional spaces from real analysis. I also like that colo
81.
by swagonomixxx 6y ago
I'm not sure I understand the relationship between partial function and exception handling. Aren't partial functions just curried? One or more arguments are bound, but not all? At least in Python, if you partial-ify a function tha
82.
by swagonomixxx 6y ago
I agree with you. To me it just sounds like people being entitled, expecting a service to never go down, ever. Shit happens and things go down. Design your processes around that fact and have procedures lined up for when it does happen. How
83.
by swagonomixxx 6y ago
Goods and services will never be priced in BTC because BTC fluctuates by the hour and will probably continue to do so for the foreseeable future. The only way I can see this happening is if you literally eliminate all other currencies, incl
84.
by swagonomixxx 6y ago
A place I was at used Hubot as well. It gets the job done, we never really ran into a fuss. Easily extensible as well.
85.
by swagonomixxx 6y ago
chatops deploys aren't really new toys, a place I was at was doing them around 2013/14. We liked it because the chat history you see is essentially a deploy history, no need to login into some other website to check some obscure l
86.
by swagonomixxx 6y ago
I don't want to be _too_ pedantic but the M1 macbooks are significantly cheaper than their Intel counterparts. The 13 inch one is around $1300 USD. But I do agree with you. Cheaper laptops means a worse experience with almost everythin
87.
by swagonomixxx 6y ago
Easy question - whichever one is hyped more at the particular time you have to make the choice.
88.
by swagonomixxx 6y ago
I only recently bought F1 2019 and I am enjoying it - it's quite good when you get used to it. Shame that EA acquired them.
89.
by swagonomixxx 6y ago
This made me chuckle, but I'm now dismayed because I can totally see EA making this some internal KPI ($$$ ripped off per user) that they track the mean, standard deviation, and so on YoY.
90.
by swagonomixxx 6y ago
It's not utter garbage (at least not yet).

More ›