skdotdan

7 ms·

1.
by skdotdan 1y ago
Personally, I use LLMs to write code that I would have never bothered writing in the first place. For example, I hate web front-end development. I'm not a web dev, but sometimes it's cool to show visual demos or websites. Without
2.
by skdotdan 1y ago
Not true. VAT does way less economic damage. Tariffs introduce massive distortions in the market.
3.
by skdotdan 2y ago
Edited
4.
by skdotdan 2y ago
As an anecdotal data point, I’m happier than in my early and mid 20s, and I have hope for my 30s.
5.
by skdotdan 2y ago
This is not what hallucination means in the pre-LLM machine learning literature.
6.
by skdotdan 2y ago
It’s used incorrectly. Hallucination has (or used to have) a very specific meaning in machine learning. All hallucinations are errors but not all errors are hallucinations.
7.
by skdotdan 2y ago
What do you mean by “true” RL?
8.
by skdotdan 2y ago
The way I see it hallucination and factual error are particular cases of incorrect answer.
9.
by skdotdan 3y ago
Deepmind tried something similar in 2018 https://deepmind.google/discover/blog/learning-to-write-prog...
10.
by skdotdan 3y ago
Cool! Thanks!
11.
by skdotdan 3y ago
Rust is great, but one thing I’d like to see is an interpreted, dynamic, less strict version of it that could be used for prototyping and gradually typed into compiling Rust code. In other words, a new programming language doing to Rust the
12.
by skdotdan 3y ago
I believe the main Mojo use cases are scenarios in which you'd need dependencies anyway. Code that you can't write in Python due to performance concerns, so you'd need to call C/C++/Rust/CUDA/Triton/e
13.
by skdotdan 3y ago
I'm sure Chapel has its merits, but one of the main selling points of Mojo is the aspiration to be part of the Python ecosystem, and so far I haven't seen any other programming language offering a similar promise, other than Pytho
14.
by skdotdan 3y ago
The term of “hallucination” referred to generative ML models have been used in the literature for years, at least since GANs.L
15.
by skdotdan 3y ago
I graduated in 2019 and studied Prolog in my third year.
16.
by skdotdan 3y ago
Looks great! Rust is the wrong language for developing ML models, but it has a clear potential for being the best one for deploying them.
17.
by skdotdan 3y ago
Aren’t iMessage and FaceTime irrelevant outside from the US? I don’t know a single non-American person who uses them.
18.
by skdotdan 3y ago
Not an expert in East Asian languages, but GPT tokenizers are generally byte-based. Meaning that the basic unit to do the merges is a single byte, not a character.
19.
by skdotdan 3y ago
I’d add that while languages should die, code written in it shouldn’t. So, even if a language is deprecated, we still should be able to run, call and interoperate with all the legacy code that was written in it.
20.
by skdotdan 3y ago
https://huggingface.co/google/flan-ul2 https://huggingface.co/docs/transformers/model_doc/gpt_neox
21.
by skdotdan 4y ago
No one said it was. What some researchers hypothesize is that at the current rate of improvement, it wouldn't be especially unreasonable to expect AGI at GPT-N.
22.
by skdotdan 4y ago
Could you share more details about it? Is your code open source?
23.
Ask HN: What Happened to Reddit?
176 points by skdotdan 4y ago|200 comments
24.
by skdotdan 4y ago
YouTube ads are way more disruptive because they have to interrupt the video you are playing for at least a few seconds.
25.
by skdotdan 4y ago
The moat is network effects, not the hosting service. GitHub is indeed a potential competitor but, again, you are just thinking about the hosting, while Huggingface provides an API compatible with all the uploaded models and datasets. GitHu
26.
by skdotdan 4y ago
I don't think OpenAI is a valid comparison. Huggingface's mission, unlike in the case of OpenAI, is not training models, but being the standard service for sharing them. The vast majority of models and datasets available at the Hu
27.
Ask HN: How to leverage a constant flow of ideas
2 points by skdotdan 6y ago|0 comments
28.
by skdotdan 6y ago
I wonder whether Bezos is feeling envious (in the good sense of the word). Wouldn't it be cool a new Space Race between SpaceX, Blue Origin, and Boeing?
29.
by skdotdan 6y ago
AFAIK they used half-precision (Float16)
30.
by skdotdan 6y ago
Location: Europe, but willing to relocate if required. Remote: If required. Willing to relocate: Yes. Technologies: PyTorch, NLP, deep learning, machine learning, Python, HPC, Unix. Résumé/CV: Please email me. Email: npfc2010@gmail.co

More ›