Dataset · unsloth
1% sample of LaTeX_OCR for formula recognition
unsloth/LaTeX_OCR
A small sample of image-LaTeX pairs for training math formula OCR.
What's inside
This dataset is a 1% sample from linxy/LaTeX_OCR. It contains image and LaTeX text pairs; according to metadata, the size is between 10k and 100k examples. No other details are provided in the card.
What you can do with it
- Train an OCR model for LaTeX formulas
- Fine-tune a vision-language model to recognize math
- Evaluate image-to-LaTeX quality on the sample
- Size
- 10K–100K rows
- Downloads this month
- 2K
- Likes
- 98
Models trained on it
- mradermacher/Qwen2-VL-OCR-2B-Instruct-i1-GGUF 1.1K
- mradermacher/Qwen2-VL-OCR-2B-Instruct-GGUF 974
- mradermacher/Imgscope-OCR-2B-0527-i1-GGUF 769
- mradermacher/Qwen2-VL-OCR2-2B-Instruct-GGUF 741
- prithivMLmods/Qwen2-VL-OCR-2B-InstructImage text → text 648
- mradermacher/Qwen3-2b-vl-latex-ocr-GGUF 591
- mradermacher/Callisto-OCR3-2B-Instruct-i1-GGUF 464
- Weidows/Qwen3.5-4B-LaTeXImage text → text 464
How to load it
from datasets import load_dataset
ds = load_dataset("unsloth/LaTeX_OCR")