Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
jsinai
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
15 ms
·
121.
▲
by
jsinai
7y ago
> Edit: this May be unfair to Julia Can’t understate this enough. Ultimately there is a trade-off being made between end user applications and numerical computing. Julia is surely superior when it comes to numerical computing. It has a s
122.
▲
by
jsinai
7y ago
But on the other hand, you could be spending that extra time reading and learning about things for your own interest or socializing with your friends for your personal benefit and social health. Ultimately there is a trade-off between spend
123.
▲
by
jsinai
7y ago
Not sure how standard that is ...
124.
▲
by
jsinai
7y ago
Maybe the problem is that most of these models seem to rely on sequential information (even the transformer needs this for forward generation of text) to encode long range information. But I can’t remember the last time I relied on sequenti
125.
▲
by
jsinai
7y ago
In theory if you had a source-to-source automatic differentiator (a programme which takes the source code for another programme f and outputs the source code for its derivative f'), then you could run backpropagation through that progr
126.
▲
by
jsinai
7y ago
FWIW I have the Bose NC 700 and the Bluetooth pairing is seamless. I can be paired to multiple devices at once; if I’m playing music from my laptop but receive a call from my phone then the Bose are smart enough to pause music on my laptop
127.
▲
by
jsinai
7y ago
>In urban areas, cycling is faster, and electric biking is expanding your reach a few times over. In Cambridge, UK, where I used to live it once took me quicker to ride to the same destination than a taxi. In fact it’s well known that cy
128.
▲
Ask HN: Did iOS's keyboard just get worse?
1 points
by
jsinai
7y ago
|
0 comments
129.
▲
by
jsinai
7y ago
I like this. Illustrations are great and whimsical and they made my day.
130.
▲
by
jsinai
7y ago
Use fastai if you want your code to be unreadable, unmaintable and prone to mysterious bugs.
131.
▲
by
jsinai
7y ago
Seems like a good strategy for getting comments on HN
132.
▲
by
jsinai
7y ago
> The whole point of Mozilla is supposed to be Firefox and Thunderbird And the whole point of Google is search? I think it makes sense that Mozilla would expand into more than just Firefox and Thunderbird. They’re a software foundation s
133.
▲
by
jsinai
7y ago
Salmon is usually considered a superfood
134.
▲
by
jsinai
7y ago
It depends on the use case. I find Inferno/Viridis perfectly good for representing one-dimensional gradients. They're easier to glance at then a single hue and show contrast between min/max better. Greyscale and single hue ca
135.
▲
by
jsinai
7y ago
The luminance peak in the center is by design - that's clear in its parabolic gradient. Whether this is desirable or not depends on your use case. On this point I don't see Turbo is being a do it all colormap. In the depth compari
136.
▲
Ask HN: Can anyone recommend good alternatives to Google Trips
2 points
by
jsinai
7y ago
|
0 comments
137.
▲
Ask HN: Any Good Alternatives to Google Trips?
(support.google.com)
1 points
by
jsinai
7y ago
|
0 comments
138.
▲
by
jsinai
7y ago
Surprised this hasn’t been flagged as racist, taken down or downvoted. I’m from the African continent and there is incredible culture to be found there. While I agree that mass tourism is damaging, the Indians, Chinese and Africans can’t al
139.
▲
by
jsinai
7y ago
I second this as well! ITILA is not just outstanding but offers a timeless perspective to statistical and uncertainty modeling, not to mention information theory which is often understudied in this field (including by myself). As the parent
140.
▲
by
jsinai
7y ago
These days it’s more 95% easy, last 5% hard.
141.
▲
by
jsinai
7y ago
> No further SEO necessary. (I don’t know if this would work anymore.) This still works. I wrote about a topic awhile back o help me understand it and when I search for the topic, my post still appears in the search results. I didn’t wri
142.
▲
by
jsinai
8y ago
> Most importantly, you will learn significantly faster than if you worked at a large company. Having worked in two early stage startups and at big tech early in my career, this depends entirely on your experience level and on the engine
143.
▲
by
jsinai
8y ago
It depends on where you are in terms of your learning and background. Some people will find value in intuitive blogposts which complement research papers. Others can “grok” this as you say and will find more value in code. Yet others can “g
144.
▲
by
jsinai
8y ago
I’ve had two very similar experiences as well. They were back-to-back interviews and in both the interviewers started simple, then stepped up the difficulty. During the whole process the interviewers worked with me, not against me. Of cours
145.
▲
by
jsinai
8y ago
This book saved my life during uni. I found the explanations to be clear and well thought out. Math textbooks are expensive and it’s a hard sell when you’re still an undergraduate and not sure what you will end up doing with your life. Furt
146.
▲
by
jsinai
8y ago
First chapter of this book has a similar style and will get you started: http://mbmlbook.com
147.
▲
by
jsinai
8y ago
Thanks, I appreciate the kind words.
148.
▲
by
jsinai
8y ago
Thanks. Looks fantastic. Also appreciate the feedback from parent.
149.
▲
by
jsinai
8y ago
There is a PyTorch implementation by the authors https://github.com/rtqichen/torchdiffeq and there is a Julia implementation due: https://news.ycombinator.com/item?id=18676986#18691054 . As for TensorFl
150.
▲
by
jsinai
8y ago
Thanks for pointing this out. It is indeed a typo.
More ›