vikp

8 ms·

1.
by vikp 3mo ago
It's an interesting illustration of the state of the AI market that immediately after arguing that AI cannot do anything complex...we have an ad arguing that AI can actually do those things. Even the people telling you that AI is a
2.
by vikp 9mo ago
Hey, I'm the founder of Datalab (we released Chandra OCR). I see someone requested it below - happy to help you all get setup. I'm vik@datalab.to
3.
by vikp 9mo ago
Yes, we can sign a BAA!
4.
by vikp 9mo ago
Hi, I'm a founder of Datalab. I'm not trying to take away from the launch (congrats), just wanted to respond to the specific feedback. I'm glad you found a solution that worked for you, but this is pretty surprising to hear
5.
by vikp 1y ago
I assume you're using a PDF, and not the image you shared? You need to set force ocr or format lines to get inline math with a PDF (for images, we just OCR everything anyways, so you don't need any settings). We're working o
6.
by vikp 1y ago
Hi, author of marker here - I tried your image, and I don't see the issues you're describing with the newest version of marker (1.7.5). I ran both with no setting specified, and with force_ocr, and I didn't see the issues eit
7.
by vikp 2y ago
Thanks for sharing! I'm training some models now that will hopefully improve this and more :)
8.
by vikp 2y ago
Benchmarking is hard for markdown because of the slight formatting variations between different providers. With HTML, you can use something like TEDS (although there are issues with this, too), but with markdown, you don't have a grea
9.
by vikp 2y ago
I ran a partial benchmark against marker - https://github.com/VikParuchuri/marker . Across 375 samples with LLM as a judge, mistral scores 4.32, and marker 4.41 . Marker can inference between 20 and 120 pages per seco
10.
by vikp 2y ago
I'm a fan of the team of Allen AI and their work. Unfortunately, the benchmarking of olmocr against marker ( https://github.com/VikParuchuri/marker ) is quite flawed. Throughput - they benchmarked marker API cost v
11.
by vikp 2y ago
Docling is a great project, happy to see more people building in the space. Marker output will be higher quality than docling output across most doc types, especially with the --use_llm flag. A few specific things we do differently: -
12.
by vikp 2y ago
Marker ( https://www.github.com/VikParuchuri/marker ) works kind of like this. It uses a layout model to identify blocks and processes each one separately. The internal format is a tree of blocks, which have arbitrary
13.
by vikp 2y ago
It's on the list to build - been focusing on quality pretty heavily lately.
14.
by vikp 2y ago
Hey, I'm the author of marker - thanks for sharing. Most of the processing time is model inference right now. I've been retraining some models lately onto new architectures to improve speed (layout, tables, LaTeX OCR). We recentl
15.
by vikp 2y ago
Datalab | NYC | Full-time | Software Engineer and Head of Business Ops | $250k-$350k + 1.5-3% equity | https://www.datalab.to A significant % of useful data is locked away in tough-to-parse formats like PDFs. We build tools to
16.
by vikp 2y ago
Hi, I'm the author of surya ( https://github.com/VikParuchuri/surya ) - working on improving speed and accuracy now. Happy to collaborate if you have specific page types it's not working on. For modern/c
17.
by vikp 2y ago
Hi, I'm the author of marker - https://github.com/VikParuchuri/marker - from my testing, marker handles almost all the issues you mentioned. The biggest issue (that I'm working on fixing right now) is format
18.
by vikp 2y ago
Working on improving tables soon (I'm the author of marker)
19.
by vikp 2y ago
This isn't specifically tuned for tables (more for general pdf to markdown), but it's worked for some people with similar use-cases - https://github.com/VikParuchuri/marker
20.
by vikp 2y ago
For PDF to markdown, I recently released V2 of my tool marker - https://github.com/vikparuchuri/marker
21.
by vikp 2y ago
It should be possible to call a GPL library in a separate process (surya can batch process from the CLI) and avoid GPL - ocrmypdf does this with ghostscript.
22.
by vikp 2y ago
I've benchmarked against google cloud ocr, but the results are on Twitter, not the repo yet - https://twitter.com/VikParuchuri/status/1765440195124691339 . The reason I didn't benchmark against doctr is
23.
by vikp 2y ago
This looks great! You might be interested in surya - https://github.com/VikParuchuri/surya (I'm the author). It does OCR (much more accurate than tesseract), layout analysis, and text detection. The OCR is slow
24.
by vikp 3y ago
I unfortunately haven't had time to benchmark against more than tesseract.
25.
by vikp 3y ago
You may want to try https://github.com/VikParuchuri/surya (I'm the author). I've only benchmarked against tesseract, but it outperforms it by a lot (benchmarks in repo). Happy to discuss. You could also try
26.
by vikp 3y ago
Hi HN - I released an open source OCR model yesterday that supports 93 world languages. It builds on a text line detector I created earlier. In my benchmarks, it's more accurate than tesseract in every language except one. (see repo f
27.
Show HN: Surya – OCR and line detection in 93 languages(github.com)
11 points by vikp 3y ago|1 comments
28.
by vikp 3y ago
The size of the framework is not the most important factor - the model weights are usually 10x+ the size of the framework. The most important factor is inference speed. For something called Nitro, I really expected speed benchmarks. I&#x2
29.
by vikp 3y ago
Thanks for letting me know. I see the C swapped to an epsilon - the new model checkpoint (live now) seems to fix that. The \, rendered as , is due to Github math rendering (it renders fine elsewhere). I'll manually edit to remove th
30.
by vikp 3y ago
Hi HN - I made texify to convert equations to markdown/LaTeX for my project marker [1] then realized it could be generally useful. Texify converts equations and surrounding text to Markdown, with embedded LaTeX (MathJax compatible). Y

More ›