Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

DeepSeek OCR

github.com

Post reply on HN
Print view View original

241–250 of 252 posts

Re: DeepSeek OCR

#241

Post by polytely »

modeless wrote: post #140

Hmm, at first I was thinking "why OCR?", but maybe the reason is to ingest more types of training data for LLM improvement, e.g. scanned academic papers? I imagine all the frontier labs have a solution for this due to the value of academic papers as a data source. Edit: Oh I see the paper abstract says this explicitly: "In production, DeepSeek-OCR can generate training data for LLMs/VLMs at a scale of 200k+ pages per…

If we get ocr working it makes it possible to store all human knowledge now stored in PDF's with way less resources

https://annas-archive.org/blog/critical-window.html

Re: DeepSeek OCR

#242

Post by vladpowerman »

The compression framing is super interesting. It makes me wonder if there’s an equivalent notion for source code - like how much “information” or entropy a commit contains vs. boilerplate churn.

I’ve been exploring Git activity analysis recently and ran into similar trade-offs: how do you tokenize real-world code and avoid counting noise?

Re: DeepSeek OCR

#243

Post by psb217 »

mapleshamrock wrote: post #239
psb217 wrote: post #37

Earlier quoted context omitted.

The trick is that the vision tokens are continuous valued vectors, while the text tokens are elements from a small discrete set (which are converted into continuous valued vectors by a lookup table). So, vision tokens can convey significantly more bits per token than text tokens. This allows them to pack the content of multiple text tokens into a single vision token.

Couldn't you do something like add a bidirectional encoder after your embedding look up table to compress your text into some smaller token-count semantic space before feeding your transformer blocks to get a similar effect, then?

Yes, you can get good compression of a long sequence of "base" text tokens into a shorter sequence of "meta" text tokens, where each meta token represents the information from multiple base tokens. But, grouping a fixed number of base tokens into each meta token isn't ideal, since that won't align neatly with sensible semantic boundaries, like words, phrases, sentences, etc. So, the trick is how decide which base tokens should be grouped into each meta token....

This sort of "dynamic chunking" of low-level information, perhaps down to the level of raw bytes, into shorter sequences of meta tokens for input to some big sequence processing model is an active area of research. Eg, one neat paper exploring this direction is: "Dynamic Chunking for End-to-End Hierarchical Sequence Modeling" [1], from one of the main guys behind Mamba and other major advances in state-space models.

[1] - https://arxiv.org/abs/2507.07955

Re: DeepSeek OCR

#244

Post by dev1ycan »

sigmoid10 wrote: post #118
dev1ycan wrote: post #109

Earlier quoted context omitted.

Oh great so now Anna's archive will get taken down as well by another trash LLM provider abusing repositories that students and researchers use, META torrenting 70TB from library genesis wasn't enough

Seems like they are doing fine: https://open-slum.org

Yeah, for now, Meta torrented 70TB and right after that they cut the rope for everyone else, mysteriously their hitman (US govenrment) hit both Libgen and Z-Lib shortly after.

Re: DeepSeek OCR

#245

Post by junto »

yoran wrote: post #19

How does an LLM approach to OCR compare to say Azure AI Document Intelligence ( https://learn.microsoft.com/en-us/azure/ai-services/document... ) or Google's Vision API ( https://cloud.google.com/vision?hl=en )?

Not sure how it compares but we did some trials with Azure AI Document Intelligence and were very surprised at how good it was. We had a document example which was a poor photograph of a document that had quite a skew, and it (too our surprise), also detected the customer’s human legible signature and extracted their name from that signature.

Re: DeepSeek OCR

#246

Post by dajonker »

wahnfrieden wrote: post #124
raincole wrote: post #47

Earlier quoted context omitted.

If you can accept that the machine just make up what it doesn't recognize instead of saying "I don't know," then yes it's solved. (I'm not being snarky. It's acceptable in some cases.)

Do any LLM OCRs give bounding boxes anyway? Per character and per block.

Try MinerU 2.5 with two-step parsing. It gives good results with bounding boxes per block. Not sure if you can get it to do more detailed such as word or character level.

Re: DeepSeek OCR

#247

Post by krackers »

miki123211 wrote: post #87
krackers wrote: post #8

The paper is more interesting than just another VLM for OCR, they start talking about compression and stuff. E.g. there is this quote >Our work represents an initial exploration into the boundaries of vision-text compression, investigating how many vision tokens are required to decode text tokens. The preliminary results are encouraging: DeepSeek-OCR achieves near-lossless OCR compression at approximately 10× ratios,…

Text tokens are quantized and represent subword units, vision tokens only exist in the embedding space. The way text tokenization works in LLMs is that you have a "lookup table" of (small) token ids to (large) vector embeddings. To pass text to the LLM, you split it at token boundaries, convert strings to token ids, and then construct the "context", a matrix where each row is a vector taken from that lookup table. Tr…

Thank you, this makes sense! As [1] puts it pithily

>Image-patch tokens make better use of the high-dimensional embedding space than text tokens do.

That seems to imply it's not necessarily something unique about images, just a byproduct of having better conversion from "raw input -> embeddings" [2]. Although there is a certain elegance of handling both images and text with the same method.

[1] https://twitter.com/c0mbinat0r/status/1980698103234891892 [2] https://twitter.com/Kangwook_Lee/status/1980709454522744902

Re: DeepSeek OCR

#248

Post by davidguetta »

rco8786 wrote: post #95
miki123211 wrote: post #87

Earlier quoted context omitted.

Text tokens are quantized and represent subword units, vision tokens only exist in the embedding space. The way text tokenization works in LLMs is that you have a "lookup table" of (small) token ids to (large) vector embeddings. To pass text to the LLM, you split it at token boundaries, convert strings to token ids, and then construct the "context", a matrix where each row is a vector taken from that lookup table. Tr…

Great explanation, thanks. I was surprised to hear that models still only work with ~100k tokens, but after giving it some thought it makes sense. There's only so many words/subword units that get used in any given language. The entropy comes from all the billions of different ways those subwords can be ordered .

Theres almost an infinity of chess games possible from barely 32 pièces and simple moves.

Already at move 3 you have bilions of positions possible

Re: DeepSeek OCR

#249

Post by Karen667 »

DeepSeek-OCR revolutionizes document processing by converting text into high-resolution images, achieving up to 20× compression while maintaining impressive accuracy. At a 10× compression ratio, it retains approximately 97% accuracy, and even at 20×, it maintains around 60% accuracy Tom's Hardware .

This approach reduces token usage significantly, making it particularly beneficial for industries like finance, healthcare, and legal sectors. For a comprehensive guide on implementing and utilizing DeepSeek-OCR, you can check https://deepseeksguides.com/deepseek-ocr-guide/

Re: DeepSeek OCR

#250

Post by giardini »

This model (DeepSeek-OCR) aligns particularly well with what we know about written language and the human act of reading.

The Visual Word Form Area (VWFA) on the left side of the brain is where the visual representation of words is transformed to something more meaningful to the organism.

https://en.wikipedia.org/wiki/Visual_word_form_area

The DeepSeek-OCR encoding (rather than simple text encoding) appears analogous to what occurs in the VWFA.

This model may not only be more powerful than text-based LLMs but may open the curtain of ignorance that has stymied our understanding of how language works and ergo how we think, what intelligence is precisely, etc.

Kudos to the authors: Haoran Wei, Yaofeng Sun, Yukun Li. You may have tripped over the Rosetta Stone of intelligence itself! Bravo!

Post reply on HN