minxomat

6 ms·

1.
by minxomat 3y ago
Is this related in any way to the actual ChatLLama project that is developed here: https://github.com/nebuly-ai/nebullvm ?
2.
by minxomat 4y ago
The former. Makes sense for their business model.
3.
by minxomat 4y ago
Follow up recommendation: for Spanish, “Madrigals Key” is language transfer in book form. It can in fact be used as a workbook for this course.
4.
by minxomat 4y ago
Sure. Sent a ping.
5.
by minxomat 4y ago
We're not far off: https://imgur.com/a/nJxxcUg
6.
by minxomat 4y ago
A plot point in an episode of Person of Interest, where AGI "Samaritan" funds a charity (free tablets for students, Samaritan access pre-installed) to take over education and recruit more mercenaries.
7.
by minxomat 4y ago
I'm building this, for (mostly) non-scientific non-fiction works (books, articles, news, etc.). Launching soon, with about 7,500 books indexed. Generally, what I found useful to build a graph between "topics" or entities was
8.
by minxomat 4y ago
The whole point of the LLaMa paper is that large models are undertrained and oversized.
9.
by minxomat 4y ago
A laptop and a desktop (Mac Studio)
10.
Hypothetical Embeddings Explained(summarity.com)
1 points by minxomat 4y ago|0 comments
11.
by minxomat 4y ago
This has been fixed almost 2 days ago now. It’s literally mentioned at the top of the repo.
12.
by minxomat 4y ago
The full use case includes quantisation, which the repo points out uses a large amount of system RAM. Of course that’s not required if you skip that step.
13.
by minxomat 4y ago
You’re missing something. Both SHP ( https://huggingface.co/datasets/stanfordnlp/SHP ) and OpenAssistant datasets are referenced. And the TOS violation might be the case, the project nevertheless has a mode to use O
14.
by minxomat 4y ago
With 16 threads, about 140ms per token for 30B, 300ms per token for 65B I should also mention that 65B should be able to run on 64GB systems. Total system memory consumption on M1 Ultra is about 67GB when running nothing else.
15.
by minxomat 4y ago
So if I'm reading this right, 65B at 4bit would consume around 20GB of VRAM and ~130GB of system RAM?
16.
by minxomat 4y ago
There are open datasets (see the chatllama harness project and its references). You can of course also cross train it using actual ChatGPT.
17.
by minxomat 4y ago
AdGuard
18.
by minxomat 4y ago
> not really competitive with ChatGPT That's impossible to judge. LLama is a foundational model. It has received neither instructional fine tuning (davinci-3) nor RLHF (ChatGPT). It cannot be compared to these finetuned models witho
19.
by minxomat 4y ago
GPT-3 might be the ultimate contender for https://en.m.wikipedia.org/wiki/Hutter_Prize - now if we could only find the prompts to recover it all
20.
by minxomat 4y ago
Everything old is new again … https://en.m.wiktionary.org/wiki/juvenoia
21.
by minxomat 4y ago
Not to forget the Toshiba AC100 with the original Tegra, which was one of the first usable ARM laptops (netbooks?) and helped much of Linux desktop support development for ARM in the early days.
22.
by minxomat 4y ago
Not really, that’s exactly what Papertrail does.
23.
by minxomat 4y ago
Would love to use an API, per token pricing is a good approach (with use limits like OpenAI). If you need testers, I have some use case (long form non-fiction content). LMK at ml[at]summarity[dot]com
24.
by minxomat 4y ago
I second this, API please.
25.
by minxomat 4y ago
As an interviewee you can use the interviewer's typing as a good signal. When I'm conducting an LP interview and you go off on a tangent that I'm not interested in, I'll stop typing and look up long before I'm likel
26.
by minxomat 4y ago
I’ve done some experiments and the ideal seems to be semantic chunks of max 175 tokens. But only if those chunks are already very dense (use summarisation to get there)
27.
by minxomat 4y ago
That’s commonly referred to as abstractive vs extractive summarisation
28.
by minxomat 4y ago
Precedent would be Blinkist, which AFAIK just uses an army of editors. Though, they probably _have_ contracts with the publishers
29.
Dropbox drops support for custom or external locations on macOS(help.dropbox.com)
4 points by minxomat 4y ago|0 comments
30.
by minxomat 4y ago
> kind of cohesive conceptual understanding of a single object It doesn't, it just requires character sheets in the training set, the rest is basically a style transfer.

More ›