syllogism

14 ms·

91.
by syllogism 6y ago
A practical tip: try hard not to need this sort of thing, and you'd be surprised how often you can avoid it. Before you start looking at approximate nearest neighbors, ask yourself the following: 1. Do I know all of the vectors ahead o
92.
by syllogism 6y ago
> fewer restrictions on who can immigrate than any other country in the world Citation needed? This seems obviously false if you include non-OECD countries, and even within the OECD, it's surely untrue? For instance, immigrating to
93.
by syllogism 6y ago
Argh this bullshit is extremely frustrating. Apologies for the intemperate rant. OF COURSE nobody can use advice like "hire slow, fire fast". They ignore this advice because it's fucking useless . I just...how could someone
94.
by syllogism 6y ago
Yes that's true, but turning it around: the task is too big (and also sucks too much) for one person to do mostly alone. And also like, why? There would've been a more active community of companies depending on it and it wouldn&#x
95.
by syllogism 6y ago
Fabric was a casualty of the Python 2/3 split. Unicode isn't a great default for the shell interaction so the project took a long time to be ported, and by then it lost all momentum. The successor project, Invoke, was finally rele
96.
by syllogism 6y ago
Nice! It's definitely been one of the bigger holes. I hope you'll also consider sequential tasks in a single jobfile. Without it it's kind of awkward because you have to be passing shell scripts around just to run a setup ste
97.
by syllogism 6y ago
A few long-standing feature requests I've noticed a lot: * No autoscaling * Can't reserve entire CPU core: https://github.com/hashicorp/nomad/issues/289 * No way to run jobs sequentially: https:&#x
98.
by syllogism 6y ago
There are so many legit problems to be sad about, but he ruins the list by bloating it with like 20-30% crap/bullshit entries. Why? Would be so much better and more productive if he kept it tight. He actually doesn't even have the
99.
by syllogism 6y ago
What could that even look like? In the extreme case this is like asking, "Why do I have to write programs, can't the computer just do what I tell it?". Writing the program is telling it what you want. In supervised machine
100.
by syllogism 6y ago
In those sorts of domains the best pitch for AI is as a failsafe. The model and the human probably make different errors, often a human will make mistakes due to simple inattention. This lets you substitute the model for some other process
101.
by syllogism 6y ago
> why can't it read a PDF document and transform it into a machine-readable format? It can definitely do that, but you might not like the cost/benefit analysis, depending on how many such documents you want to process. The cos
102.
by syllogism 6y ago
Did you check the heat? Maybe you got a lemon and it was constantly being CPU throttled.
103.
by syllogism 6y ago
I very much disagree with: > 7. You are operating mainly on assumptions You'll always be operating mainly on assumptions. Everyone will. It does help to check your assumptions against external things, so if you're wrong, you ca
104.
Backpropagation 101(thinc.ai)
1 points by syllogism 7y ago|0 comments
105.
by syllogism 7y ago
The cases and deaths counts aren't informative enough to really say anything useful about what's going on. When a "case" is reported, we don't when the symptoms started occurring, or even when the original sample wa
106.
by syllogism 7y ago
You can also try out Stanza in spaCy --- Ines updated the spacy-stanfordnlp wrapper to use the new version pretty much immediately: https://github.com/explosion/spacy-stanza I was asked by a friend about Stanza in a pr
107.
by syllogism 7y ago
O'Reilly firing events team, cancelling Strata etc: https://twitter.com/LauraABaldwin/status/1242517013878071296
108.
by syllogism 7y ago
For our own part: we haven't really noticed anything in our business so far, I guess it takes a while for any cost-cutting orders to filter down. I definitely expect a recession, but even in pretty extreme scenarios we wouldn't ne
109.
Ask HN: Who is firing right now?
19 points by syllogism 7y ago|8 comments
110.
by syllogism 7y ago
This outbreak has made me realise how I've gone my whole adult life without knowing some basic things about bacteria and viruses. Like, I never knew about how soap dissolves the virus's outer coating... Anyway apparently hot and w
111.
by syllogism 7y ago
The best answer is, "with some difficulty". I've written a bit about the struggles here: https://thinc.ai/docs/usage-type-checking#model-types There are basically two approaches, TypeVar and overload. Yo
112.
by syllogism 7y ago
You can write custom CUDA kernels, and I've written a few to support operations over our ragged format. Actually Thinc makes it pretty easy to optimise a specific bit of code with a custom op...cupy's fused decorator can also work
113.
by syllogism 7y ago
I've been working on various versions of Thinc for a long time --- it's always been the name for the machine learning code powering spaCy. But previous versions really weren't worth using for other developers, and so I never
114.
Mathy: Reinforcement learning environment for solving maths problems(mathy.ai)
2 points by syllogism 7y ago|0 comments
115.
Sense2vec reloaded: contextually-keyed word vectors(explosion.ai)
7 points by syllogism 7y ago|0 comments
116.
by syllogism 7y ago
He didn't ask us to post this, but I happened to see this so: Paul's great and we definitely expect to continue working with him. We're hugely grateful for his work on the spaCy Japanese support, and we've also worked wi
117.
by syllogism 7y ago
Definitely a fair concern, and thanks for the kind words :). We haven't had to fund-raise, so our company performance has actually been a bit under-the-radar. But I can assure you everything's been going well, and most people woul
118.
by syllogism 7y ago
We're definitely planning to do more there -- we took a bit longer than anticipated getting v2.2 out, due to various packaging and automation hiccups. Huggingface are doing a great job keeping transformers moving quickly, so we already
119.
by syllogism 7y ago
The intended pronounciation is /speɪˈsiː/, i.e. spay-SEE. The capitalisation refers to the fact that it's a C extension, implemented in Cython.
120.
SpaCy v2.2(explosion.ai)
130 points by syllogism 7y ago|9 comments

More ›