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

Mistral OCR

mistral.ai

Post reply on HN
Print view View original

381–390 of 450 posts

Re: Mistral OCR

#381

Post by supriyo-biswas »

TJSomething wrote: post #364
boxed wrote: post #356

Earlier quoted context omitted.

Why wouldn't hallucinations be agreed upon if they have roughly the same training data?

A hallucination is often an indication that the model doesn't know something. Then, the internal signal gets dominated by noise from the seeded training weights. Efforts to eliminate hallucinations with a single model have found success by asking the same question in different ways and only taking answers that agree. Logically, you could get more durable results from multiple models on the same prompt.

We had this article the other day[1] about how multiple LLMs can hallucinate about the same thing, so this is not guaranteed to remove hallucinations that are caused by poor or insufficient training data.

[1] https://news.ycombinator.com/item?id=43222027

Re: Mistral OCR

#382

Post by raffraffraff »

layer8 wrote: post #281
bambax wrote: post #257

It's not bad! But it still hallucinates. Here's an example of an (admittedly difficult) image: https://i.imgur.com/jcwW5AG.jpeg For the blocks in the center, it outputs: > Claude, duc de Saint-Simon, pair et chevalier des ordres, gouverneur de Blaye, Senlis, etc., né le 16 août 1607 , 3 mai 1693 ; ép. 1○, le 26 septembre 1644, Diane - Henriette de Budos de Portes, morte le 2 décembre 1670; 2○, le 17 octobre 1672, Cha…

> This is perfect! Just a nit, but I wouldn’t call it perfect when using U+25CB ○ WHITE CIRCLE instead of what should be U+00BA º MASCULINE ORDINAL INDICATOR, or alternatively a superscript “o”. These are https://fr.wikipedia.org/wiki/Adverbe_ordinal#Premiers_adver... . There’s also extra spaces after the “1607” and around the hyphen in “Diane-Henriette”. Lastly, U+2019 instead of U+0027 would be more appropriate for…

It feels like, after the OCR step there should be language and subject matter detection, with a final sweep with a spelling / grammar checker that has the right "dictionary" selected. (That, right there, is my naivety on the subject, but I would have thought that the type of problem you're describing isn't OCR but classical spelling and grammar checking?)

Re: Mistral OCR

#383

Post by yfontana »

techwizrd wrote: post #223
kbyatnal wrote: post #140

We're approaching the point where OCR becomes "solved" — very exciting! Any legacy vendors providing pure OCR are going to get steamrolled by these VLMs. However IMO, there's still a large gap for businesses in going from raw OCR outputs —> document processing deployed in prod for mission-critical use cases. LLMs and VLMs aren't magic, and anyone who goes in expecting 100% automation is in for a surprise. You still n…

The challenge I have is how to get bounding boxes for the OCR, for things like redaction/de-identification.

I'm working on a projet that uses PaddleOCR to get bounding boxes. It's far from perfect, but it's open source and good enough for our requirements. And it can mostly handle a 150 MB single-page PDF (don't ask) without completely keeling over.

Re: Mistral OCR

#384

Post by stavros »

vikp wrote: post #289

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 second on an H100. You can see the samples here - https://huggingface.co/datasets/datalab-to/marker_comparison... . The code for the benchmark is here - https://github.com/VikParuchuri/marker/tree/master/ben…

I like the licensing options! Hopefully they make enough money to fund development.

Re: Mistral OCR

#385

Post by ekwav »

mkl wrote: post #218
dgfl wrote: post #32

Earlier quoted context omitted.

Great example of how information is sometimes compartmentalized arbitrarily in the brain: I imagine you have never been confused by sentences such as “I’m running at 10 km/h”.

Dollar signs go before the number, not after it like units. It needs to be 1000 pages/$1 to make sense, whereas 10km and 10h and 10/h all make sense so 10km/h does. I imagine you would be confused by km/h 10 but not $10.

In the EU we put the € symbol after the number so it feels more natural to do that for $ as well.

Re: Mistral OCR

#386

Post by kinnth »

This looks like a massive win if you were the NHS and had to scan and process old case notes.

Same is true if you were a solicitors/lawyers.

Re: Mistral OCR

#387

Post by lynx97 »

pzo wrote: post #150
randomNumber7 wrote: post #122

Earlier quoted context omitted.

I never thought driving a car is harder than editing a pdf.

It's not about harder but about what error you can tolerate. Here if you have accuracy 99% for many applications it's enough. If you have 99% accuracy per trip of no crash during self driving then you gonna be dead within a year very likely. For cars we need accuracy at least 99.99% and that's very hard.

[dead]

Re: Mistral OCR

#388

Post by shrisukhani »

raunakchowdhuri wrote: post #345

We ran some benchmarks comparing against Gemini Flash 2.0. You can find the full writeup here: https://reducto.ai/blog/lvm-ocr-accuracy-mistral-gemini A high level summary is that while this is an impressive model, it underperforms even current SOTA VLMs on document parsing and has a tendency to hallucinate with OCR, table structure, and drop content.

Anecdotally, we also found Gemini Flash to be better.

Re: Mistral OCR

#390

Post by lynx97 »

bondolo wrote: post #180

Such a shame that PDF doesn’t just, like, include the semantic structure of the document by default. It is brilliant that we standardized on an archival document format that doesn’t include direct access to the document text or structure as a core intrinsic default feature. I say this with great anger as someone who works in accessibility and has had PDF as a thorn in my side for 30 years.

[dead]
Post reply on HN