nestorD

38 ms·

421.
by nestorD 7y ago
Having used both for years, nowadays I would recommend using F#. The syntax is very similar but it solves many of Ocaml's pain points and you will get better tooling and more libraries thanks to .Net.
422.
by nestorD 7y ago
> haskell-style "whitespace" to call/partially call functions That's called curryfication (named after the logician Haskell Curry) in case you don't already know :)
423.
by nestorD 7y ago
Note that, in this space, there is also Flax[0] which is also built on top of Jax bringing more deep-learning specific primitives (while not trying to be tensorflow compatible unlike Trax if I understand correctly). [0]: https://
424.
by nestorD 7y ago
I have not read the paper yet but, reading the abstract, the idea of using continuation to store the differentiation information is reminiscent of the technique used in Zygote[0]. Is there some parenty between the ideas ? [0]: https:/
425.
by nestorD 7y ago
V2 according to his twitter.
426.
by nestorD 7y ago
> I don't know where you got these papers but you clearly got an awful sampling and I don't think you're giving computer vision a fair evaluation. I disagree, I saw such an awful number of bugs in ML codes going with paper
427.
by nestorD 7y ago
Their notebooks are indeed the best I ever used : so good that I use them for the interface and not the compute power. I don't need more compute power or longer instances as I use collab with my own hardware (I have a good enough GPU a
428.
by nestorD 7y ago
I believe their choice of name is actually very clever as it sticks into your mind. If you need to list the names of the databases you know, especially if you are not an expert in the domain, and you have heard of it, it will probably be on
429.
by nestorD 7y ago
I have the opposite approach : everything is under my real name and easily traceable to me. Thus, I have never the illusion of being hidden online and I need to be conscious that everything I say is said on a public place and that I have to
430.
by nestorD 7y ago
> I don't really understand the self-flagellation in the Rust community about compile times. I believe the two main factor are the fact that: - many Rust users come from languages with a faster writing->running cycle (there is a
431.
by nestorD 7y ago
No but, looking at some screenshot + descriptions, it seems to get me further from the code which does not seem like what I am looking for Rstudio gives you the experience of a classical IDE + easy data exploration which I found to be produ
432.
by nestorD 7y ago
Yes! Rstudio is the one thing I miss most when doing datascience in python. Notebooks give me some of the interactivity but the experience degrades significantly. The spyder IDE seem like an okayish replacement but some of the library I use
433.
by nestorD 7y ago
I found out a few days ago when an, interesting, related question was posted on bioinformatics.stackexchange.com : Why does the Wuhan coronavirus genome end in aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa (33 a's)? [0] [0]: https://bio
434.
by nestorD 7y ago
I love the idea of Swift for Tensorflow : powerful automatic differentiation and a solid type system in a single language. That's something that is not seen elsewhere and that would make it a perfect fit to write deep learning code tar
435.
by nestorD 7y ago
I think deep learning can only get us so far with the existing, limited, data. I doubt it will be enough for precise structure reconstruction (but I would love to be wrong). What it can be is a force multiplier for every progress in our und
436.
by nestorD 7y ago
Not an expert but I have read a lot on the subject recently. Methods from the domain where already not truly model-based (compared to things like physical equations). It is mostly observations on existing proteins coupled with a form of gra
437.
by nestorD 7y ago
After a bit of search I found something that appears to work both well and quickly: http://www.cs.ox.ac.uk/people/yarin.gal/website/blog_3d801aa... The basic idea is brilliant: take a deep neural network with
438.
by nestorD 7y ago
I stand corrected, after some research, I found a paper that offers a promising way to get good uncertainty information from a deep learning model : http://www.cs.ox.ac.uk/people/yarin.gal/website/blog_3d801aa
439.
by nestorD 7y ago
If I understand you problem clearly, you can use TFIDF to reduce the weight of meaningless words.
440.
by nestorD 7y ago
> Is Bayesian Deep Learning worth looking into if I don't have much interest in statistical applications but much more just using deep learning as general purpose functions? If you manage to build a model with a high accuracy and a
441.
by nestorD 7y ago
As another poster said, yes : you can find good word vectors for most languages. Furthermore, given a copy of the text from wikipedia, it only takes 2 days of computation on a laptop with a modern GPU to regenerate them from scratch for any
442.
by nestorD 7y ago
I installed ubuntu for my grand father and, two years later, has not had any problem (whereas it took less than a week before he asked my do uninstall the adblocker I gave him...).
443.
by nestorD 7y ago
Reading the description, I though that there is enough interesting, high-level, material for a few phD thesis (algorithm, material science, industrialisation of a process, etc). I would love it if the author found a way to keep the research
444.
by nestorD 7y ago
Adding few letters a day, using this kind of picture as a mnemonic help [0] and learning in duo with someone else worked for me (but I lost most of it in a year due to lack of practice) https://edu.glogster.com/proxy?url=htt
445.
by nestorD 7y ago
Worth noting that Kahan summation works under the assumption that, at every step of the summation, the current sum will be larger than the term added. It is not too hard modify the algorithm in order to remove this assumption and it increas
446.
by nestorD 7y ago
Ocaml, F#, Rust, Kotlin (I believe), Haskell, etc It is a common feature in languages derives from the ML family[0]. [0]: https://en.wikipedia.org/wiki/ML_(programming_language)
447.
by nestorD 7y ago
The fact that you cannot define templates out of headers also explains why C++ has a lot of header-only libraries. : if you use templates and have no global references, you are most likely already header-only.
448.
by nestorD 7y ago
Note that he critisize the classical formula but that he, then, suppose that A=1 when he gives his formulation (making it simpler). Adding A back gives : -B/2A ± sqrt(B²/4A - C/A) which is not obviously simpler than : (-B ± s
449.
by nestorD 7y ago
Here is a paper where they had users training a bot to follow request (given in natural language) of increasing complexity to accomplish task in a simulated environement : https://research.fb.com/wp-content/uploads&#x2F
450.
by nestorD 7y ago
This. Quantum computing have be designed to solve a single problem : the simulation of quantum systems on a classical computer is very slow, lets build a quantum computer so that it will be fast. It might seem like a fringe use case but it

More ›