bwasti

8 ms·

31.
Writing a Toy Back End Compiler for PyTorch(jott.live)
1 points by bwasti 7y ago|0 comments
32.
Show HN: Writing a Back End Compiler for PyTorch(jott.live)
2 points by bwasti 7y ago|0 comments
33.
Integrating the TVM Deep Learning Compiler into PyTorch(tvm.ai)
18 points by bwasti 7y ago|1 comments
34.
by bwasti 7y ago
I think that link expired and now shows the wrong games. Here's a link to the final game between LC0 and Stockfish https://cd.tcecbeta.club/archive.html?season=15&div=sf&game=...
35.
by bwasti 7y ago
fair point, 2016 - 1 total from math and engineering https://www.cmu.edu/career/documents/2016_one_pagers/cit/ece... https://www.cmu.edu/career/documents/2016_one_pagers/mc
36.
by bwasti 7y ago
I wrote one article on Medium. I ended up writing my own little version using open source markdown rendering to see how hard it would be to roll my own, https://jott.live From what I found, there is a lot of stuff Medium does w
37.
by bwasti 7y ago
I'm not sure how the journalist fact checked this, but in 2016 CMU sent 12 people to Facebook[1]. In 2018 CMU sent 27 people to Facebook[2]. [1] https://www.cmu.edu/career/documents/2016_one_pagers/scs&#x
38.
Quora has anti-vaccine answers displayed first for innocent questions(twitter.com)
11 points by bwasti 7y ago|0 comments
39.
BoTorch: Bayesian Optimization in PyTorch(botorch.org)
6 points by bwasti 7y ago|0 comments
40.
by bwasti 7y ago
wow this is great! it uses the scroll wheel and mouse location. I had no idea the latter was possible - very inspiring
41.
by bwasti 7y ago
What aspects are broken?
42.
by bwasti 7y ago
Yea, it's certainly struggling. I don't know of any quick fixes, so I kicked off the service on more ports for now.
43.
Show HN: Messing with Telnet(jott.live)
77 points by bwasti 7y ago|34 comments
44.
by bwasti 8y ago
The pricing is a bit confusing: €10/m for 100k queries €45/m for 300k queries Can I just make 3 accounts and save €15/m?
45.
by bwasti 8y ago
You may want to check out Intel's optimized version of TensorFlow Serving[1] for further improvements (on the order of 2x for ResNet-50[2]). As an aside, I took into account the resource allocation in the parent comment. The c5.2xlarg
46.
by bwasti 8y ago
> containers are run on a 4 core, 15GB, Ubuntu 16.04 host machine What CPU is being used? Assuming the benchmark is done with something like an EC2 C5 instance, the results in this post are quite slow. Somewhere around 14x slower than b
47.
by bwasti 8y ago
Reminds me of a little project I did a while back, but instead reading from stdin like netcat https://github.com/bwasti/webpipe
48.
Steve Jobs iPhone Presentation (2007)(youtube.com)
1 points by bwasti 8y ago|0 comments
49.
by bwasti 8y ago
there’s a single algorithm (Tarjan’s) that actually solves all your graph needs. It returns a reverse topologically sorted DAG of strongly connected components
50.
Show HN: Python memoization across program runs(github.com)
5 points by bwasti 8y ago|0 comments
51.
Show HN: A website to create simple gifs with JavaScript(jott.live)
3 points by bwasti 8y ago|0 comments
52.
by bwasti 8y ago
ah, the assumption is that even the remote call (for example, in the case of async) is pure
53.
by bwasti 8y ago
as long as configuration parameters remain fixed in the life of the execution it is still pure, right?
54.
by bwasti 8y ago
beyond memoization you can do cool stuff with scheduling if you annotate pure functions and execute lazily. I actually mocked this up a little while back https://github.com/bwasti/lazy also just shared it here https:
55.
Show HN: Lazy – A Python library for rapidly developing lazy interfaces(github.com)
10 points by bwasti 8y ago|4 comments
56.
The Unix Philosophy(jott.live)
2 points by bwasti 8y ago|0 comments
57.
The Potato Paradox(jott.live)
1 points by bwasti 8y ago|0 comments
58.
by bwasti 8y ago
As well as tmux! But neither offer multiple cursor support.
59.
by bwasti 8y ago
I made a Vim plugin somewhat like this a while back https://github.com/bwasti/vim-multiuser It’d be interesting to standardize the diff protocol for live collaboration tools. Everyone uses a slightly different mechanis
60.
How the Riemann zeta function is connected to prime numbers(jott.live)
2 points by bwasti 8y ago|0 comments

More ›