[ok123456]

Qwen3.6:35b is good enough for a lot of stuff.

I just used ollama with a shell script to tackle my directory of papers/literature. I converted the first 6 pages of each document to PNG, handed them off to Qwen, and told it to spit out BibTeX, including the abstract. Two days later it was done, and I didn't spend anything on "tokens."

[abyssin]

Why PNG? Isn’t an image format more expensive to process?

[ok123456]

Not really. The model is good/fast at OCR, and preprocessing it actually makes it worse because academic paper formatting is very complicated. Sizes, positions, and equations are important.

[culi]

what a strange world we live in where robots are WORSE at handling formatted stuff. I wonder what this means for the importance of semantic HTML to screenreaders