nestorD

9 ms·

91.
by nestorD 3y ago
I am unclear on the status of the project but here is the conversation that seem to be tracking it: https://github.com/microsoft/guidance/discussions/201
92.
by nestorD 3y ago
LLMs are significantly slower than traditional ML, typically costlier and, I have been told, tend to be less accurate than a traditional model trained on a large dataset. But, they are zero/few shot classifiers. Meaning that you can ge
93.
by nestorD 3y ago
It refers to this specific incident (the OP article is even mentioned at the end of the French Wikipedia page[0]). [0]: https://fr.wikipedia.org/wiki/Affaire_du_8_d%C3%A9cembre_202...
94.
by nestorD 3y ago
I believe the French Wikipedia page[0] is the most comprehensive place to get information on the subject. There is also a, much shorter, English page[1]. [0]: https://fr.wikipedia.org/wiki/Affaire_du_8_d%C3%A9cembre_202
95.
by nestorD 3y ago
> they were monitoring them for years, so who the hell knows what else they have Given the weakness of the elements assembled, why would the DGSI decide to withhold any decisive proof from the eyes of the justice system?
96.
by nestorD 3y ago
The key is that, while there is suspicion (the people arrested have a far left ideology, have been to Syria to fight against the Islamic state, own weapons and encrypt all of their communications), there is no crime nor proof of an intent
97.
by nestorD 3y ago
This is about a poet using technology (the dormio) to experiment with dream seeding: reading a sentence then going to sleep and have the dormio wake you up when you are done with the dreaming phase so that you can take notes on what you dre
98.
by nestorD 3y ago
I found perspective transformation to be good enough for 90% of cases. Going further would be a lot of effort better spent elsewhere. Does it deal fine with illustrations and pdf compression?
99.
by nestorD 3y ago
I don't think it supports a markdown / formated text export but it looks looks fantastic as far as pdf cleanup goes (i currently rely on adobe scan for that when i am working from a paper copy), I will try it soon.
100.
by nestorD 3y ago
I will have to investigate this, I am dreaming of a system that can take a pdf scan of a book as input and produce one or more properly formated (headings, italic, bold, underline, etc) markdown files. In my tests, LLMs have proved very goo
101.
by nestorD 3y ago
Here is the relevant announcement (I believe they made it available in Pytorch as Bfloat16): https://engineering.fb.com/2018/11/08/ai-research/floating-p...
102.
by nestorD 3y ago
My guess is that they tested thousands of existing drugs (cheap to produce and easily available) in order to build a short list of new compound (significantly more expensive to produce).
103.
by nestorD 3y ago
Somewhat because the few bits they save on things like Nan give them a sensible boost on those low precisions. But the best way to go is to explucitly fit the distribution of values you are expecting: precisely what Meta did when they intro
104.
by nestorD 3y ago
Because it demonstrates a Julia library using threads created in the calling C++ code. That passing of threads between languages is highly uncommon (because many things could fail and make you crash / slow along the way and because if
105.
by nestorD 3y ago
My experience has been that GPT4 is terrible at thinking but, if I provide it with both the content and the way I want structured, it can give me an output that requires very little tweaking before being ready to go. I do my writing-as-thin
106.
by nestorD 3y ago
Here is a link to the actual paper (it appears that they forgot to link to it in the article): https://www.thelancet.com/journals/lancet/article/PIIS0140-6...
107.
by nestorD 3y ago
For me Langchain is glue code between a lot of commonly used LLM building blocks and prompts. It is great to get a prototype 80% of the way there fast in order to validate an idea or run something short lived. I suspect that, if you want to
108.
by nestorD 3y ago
The way I would do it (to be tweaked for each target LLM) is to give GPT4 a prompt telling it what kind of things make for a good prompt followed by a dozen of before/after examples (spanning various domains).
109.
Show HN: Question Extractor: turn text into LLM finetuning data(github.com)
33 points by nestorD 3y ago|1 comments
110.
by nestorD 4y ago
Using typed dependencies makes the Python IDE experience significantly nicer.
111.
by nestorD 4y ago
Uses cases that comes to mind are data distributed across nodes and things like distributed training of machine learning models (which is getting more and more focus as models get bigger).
112.
by nestorD 4y ago
Is there work to use a Slurm (MPI based) cluster transparently in Elixir? Having both great native support for distribution and a performance, GPU able, numerical library could make Elixir a great fit for that platform.
113.
by nestorD 4y ago
The headline feels very "water is wet" (nuclear does not produce air pollution so replacing is a net loss fir air quality) but the text is slightly more motivated in pointing out that policy should take into account the (predictab
114.
by nestorD 4y ago
There are trick shot with dices that will get you what you want. I have seen retired cheaters (real ones, not magicians pretending to be ones) use them and consistently get the number they want (requirement to bounce on surfaces make things
115.
by nestorD 4y ago
> model compatibility JAX will compile your model with XLA today, letting you run it on Nvidia, Apple Silicon (CPU for now), google's TPU, etc. Overall, XLA really shines on TPUs, is doing very well on GPUs and sadly lags behind on
116.
by nestorD 4y ago
There appear to be some work done on numerizing these manuscripts and making them available: https://twitter.com/M_T_Miller/status/1339315848373297158/ Translation wise, if you do not read Arabic then you cou
117.
Show HN: Alignment (a ChatGPT co-written short story)(nestordemeure.github.io)
1 points by nestorD 4y ago|0 comments
118.
by nestorD 4y ago
I think a AI-coautored section would be interesting (with an additional field to let people describe their writing process, from "I just asked for the greatest SF short story ever" to "I gave it my draft and used it as an edi
119.
by nestorD 4y ago
Part of it that this is ChatGPT + tools (given via langchain). This opens the door to a lot of power. This is a good example: https://langchain.readthedocs.io/en/latest/modules/agents/ge...
120.
by nestorD 4y ago
> Is it really just a ton of slowly developed things happening to mature close together? It is a bit of both. Once something big become public (and gets significant visibility) then a lot of people play with it and can try their own idea

More ›