hislaziness

5 ms·

1.
by hislaziness 4mo ago
Am I missing something here. Not finding a major bug/vulnerability just means that maybe the code is really good, not that the model is not what is claimed?
2.
SpaceX strikes $60B deal for the right to buy AI coding startup Cursor(finance.yahoo.com)
1 points by hislaziness 5mo ago|1 comments
3.
SpaceX is working with Cursor and has an option to buy the startup for $60B(techcrunch.com)
2 points by hislaziness 5mo ago|1 comments
4.
by hislaziness 1y ago
Would it be more appropriate to compare LLMs to Autotunes rather than pianos?
5.
by hislaziness 1y ago
Akamai CTO Robert Blumofe offers four tips for business leaders striving to foster AI fluency by empowering employees with the right tools and best use cases. useful insights but definitely a PR article.
6.
by hislaziness 1y ago
Great. I enjoy these kind of articles. My all time favorite book for 'C' is Expert C Programming: Deep C Secrets.
7.
by hislaziness 1y ago
Terence Tao on the matter - https://imgur.com/a/terence-tao-on-supposed-gold-imo-sMKP0bm
8.
by hislaziness 2y ago
As I understand, the LLM uses the techniques of searchformer - https://arxiv.org/abs/2402.14083 . To do "slow thinking" doing a A* search using a transofrmer.
9.
Controllable Fast and Slow Thinking by Learning with Randomized Reasoning Traces(arxiv.org)
44 points by hislaziness 2y ago|15 comments
10.
by hislaziness 2y ago
It is not just MSRP, management and operations cost too. The article goes into the details of this.
11.
by hislaziness 2y ago
The details are in the article. They have done the math.
12.
by hislaziness 2y ago
TLDR: Don’t buy H100s. The market has flipped from shortage ($8/hr) to oversupplied ($2/hr), because of reserved compute resales, open model finetuning, and decline in new foundation model co’s. Rent instead. Is the AI infra bubbl
13.
by hislaziness 2y ago
The screen seems to be stuck at Wait a second... for me.
14.
by hislaziness 2y ago
I tried a few local LLMs. None of them could give me the right answer for "How many 'r's in straberry. All LLMs were 8-27B.
15.
by hislaziness 2y ago
I know you mean this in jest, but we are much closer to this than we would imagine, the use of LLMs to process communication / translation is becoming ubiquitous. We are 1 bad translation away from a disaster.
16.
by hislaziness 2y ago
I also use some email providers ability to have +xyz at the end of the username. So for a registration I would for user.name+sitea@domain.com. has helped me track spam and leaks in the past.
17.
by hislaziness 2y ago
Cool. I will try it out. I tried the same with ollama, the non english part needs a lot more polish. Do you see the outcome being any different?
18.
by hislaziness 2y ago
This is pretty awesome. How did you do it? Any blog on the detials?
19.
by hislaziness 2y ago
The model description on huggingface says - Model size - 12.2B params, Tensor type - BF16. Is the Tensor type different from the training param size?
20.
by hislaziness 2y ago
I just checked huggingface and the model files download is about 25GB but in a comment below someone mentioned it is 8fp quantized model. Trying to understand how the quantization affects the model (and RAM) size. Can someone please enlight
21.
by hislaziness 2y ago
isn't it 2 bytes (fp16) per param. so 7b = 14 GB+some for inference?
22.
by hislaziness 2y ago
The title does not do justice to the article. It talks about a new classification system that OpenAI has introduced for LLMs
23.
by hislaziness 2y ago
Same here. Coincidentally I joined in June 2009 and this is the only social media I indulge in.
24.
by hislaziness 2y ago
The articles states the case is for apple's monopoly in the premium phone market. Yet it compares the market share with low cost android phones.
25.
by hislaziness 3y ago
I have experience using the Motofone - https://en.wikipedia.org/wiki/Motorola_Fone . It was not the best of experience especially due to the lag which many have mentioned here. E-Ink displays have come a long way, hopi
26.
by hislaziness 3y ago
Inspired by MalbolgeLISP which was a lisp written in Malbolge.
27.
Dive into Deep Learning(d2l.ai)
3 points by hislaziness 3y ago|1 comments
28.
by hislaziness 3y ago
Came across this book today. Good introduction and deep dive into ML.
29.
by hislaziness 3y ago
As a customer you could propose additional packages, send the request via the support channel...
30.
by hislaziness 3y ago
This is about managing security risks of using OSS libraries. Many organizations end up using OSS code/libs/products without having the capability or the motivation to validate them. Google addresses some of those issues with this

More ›