KhoomeiK

11 ms·

1.
by KhoomeiK 2y ago
NVIDIA did something similar with GANs in 2020 [1], except users could actually play those games (unlike in this diffusion work which just plays back simulated video). Sentdex later adapted this to play GTA with a really cool demo [2]. [1
2.
Show HN: Multimodal Code Generation for Web Data Extraction(twitter.com)
2 points by KhoomeiK 2y ago|0 comments
3.
by KhoomeiK 2y ago
Everything relevant in "program synthesis" moved to the new buzzword "codegen"
4.
by KhoomeiK 2y ago
Interesting—LangChain seemed kinda like unnecessary abstractions in natural language (since everything is just string manipulations), but with AI video, there's so many different abstractions that I'd need to handle (images, puppe
5.
by KhoomeiK 2y ago
Just unrolled the thread for you here! https://threadreaderapp.com/thread/1794082465398812770.html
6.
by KhoomeiK 2y ago
Thanks! I have no idea—unfortunately, very few Hindus maintain the Vedic fire rites. There are also no active central authorities on matters of Vedic ritual. The only plan of now is to use this interpretation in my own yajña practice.
7.
by KhoomeiK 2y ago
Vedic Hinduism had a similar concept of eternal fire. I recently wrote up a twitter thread [1] explaining how the modern interpretation of Vedic instructions on starting these sacred fires misunderstands the text. Etymology tidbit: "Bh
8.
Show HN: Chinchilla Scaling Laws Are Not Universal(github.com)
1 points by KhoomeiK 2y ago|0 comments
9.
Scaling Laws Depend on Data Compressibility(twitter.com)
1 points by KhoomeiK 2y ago|0 comments
10.
Gzip Predicts Data-Dependent Scaling Laws(arxiv.org)
2 points by KhoomeiK 2y ago|0 comments
11.
SB-1047 regulates all LLM's after training a suboptimal 1e26 FLOP model(twitter.com)
4 points by KhoomeiK 2y ago|0 comments
12.
How to kindle a fire: solving an 800 year old puzzle in Vedic ritual exegesis(twitter.com)
1 points by KhoomeiK 2y ago|0 comments
13.
How to kindle a fire: solving an 800 year old puzzle in Vedic ritual exegesis(twitter.com)
2 points by KhoomeiK 2y ago|1 comments
14.
by KhoomeiK 2y ago
Maybe this project another commenter is working on? https://news.ycombinator.com/item?id=40373310
15.
by KhoomeiK 2y ago
Awesome pics! We love tarsiers too
16.
by KhoomeiK 2y ago
Great question! See this thread: https://news.ycombinator.com/item?id=40369713
17.
by KhoomeiK 2y ago
Yes it does work headless and we do grab a fullpage screenshot including scrolling (by resizing viewport to content height). We haven’t had to deal with infinite scrolling much but that’s an interesting feature we’d appreciate a PR for. We
18.
by KhoomeiK 2y ago
They do show textboxes with labels. From our readme: "Keep in mind that Tarsier tags different types of elements differently to help your LLM identify what actions are performable on each element. Specifically: [#ID]: text-insertable f
19.
by KhoomeiK 2y ago
We have a lot more powerful use-cases for Tarsier in web data extraction at the moment. Stay tuned for a broader launch soon!
20.
by KhoomeiK 2y ago
We run OCR on the screenshot & convert it to whitespace-structured text, that is passed to the LLM. The images below might make it clearer for you: [1] https://github.com/reworkd/tarsier/blob/main/.gi
21.
by KhoomeiK 2y ago
Yup, it could! There are a lot of players in the generalist personal web agent space but I personally think that use-case will be eaten by big players since fundamental foundation model improvements are required. That being said, Tarsier is
22.
by KhoomeiK 2y ago
VimGPT couples the perception to a specific LLM/agent whereas Tarsier is solely a perception system that you can use for any uni/multi-modal web agent. So it's hard to compare, but you could say that VimGPT's performance
23.
by KhoomeiK 2y ago
We're hoping to release an evals paper about Bananalyzer this summer and compare Tarsier to a variety of other perception systems in it. The hard part with evaluating a perception/context system though is that it's very inter
24.
by KhoomeiK 2y ago
More OCR providers are on the roadmap and we'd love for you to contribute any local OCR models you think could be useful! I wouldn't call it a wrapper though :)
25.
by KhoomeiK 2y ago
Thanks! We might put out a paper about it with some Carnegie Mellon collaborators this summer.
26.
by KhoomeiK 2y ago
Must have been the mods, I spent quite a bit of time on the content lol
27.
by KhoomeiK 2y ago
Thanks for pointing this out! Yeah, it's pretty strange. We thought including Show HN text was encouraged to engage with the community?
28.
by KhoomeiK 2y ago
That's an interesting problem—Tarsier probably isn't the best solution here since it's focused on webpage perception rather than any kind of OCR. But one could try adapting the `format_text` function in tarsier/text_form
29.
by KhoomeiK 2y ago
Cool connection, hadn't seen this before but feels intuitively correct! I also formulate similar (but a bit more out-there) philosophical thoughts on word-meaning as being described by the topological structure of its corresponding ima
30.
by KhoomeiK 2y ago
Hm, not sure I follow why those situations would be especially difficult? Regarding website changes, the nice thing about using LLMs is that we can simply provide the previous scraper as context and have it regenerate the scraper to "s

More ›