ysharma/image-translator-pairs

image-translator-pairs Training and evaluation data for the Qwen-Image-2.1 image-translator LoRA: one image containing text + an instruction naming a target language, with the target image being a real typeset render of the translated text (per-element ground truth: polygon, font, size, colour). Backgrounds were generated with Qwen-Image 2.1 (Viggle 6-step turbo at 1024 px, OCR-filtered to be text-free). Contents path what render.py standalone… See the full description on the dataset page: https://huggingface.co/datasets/ysharma/image-translator-pairs.

<iframe src="https://silytics.si/embed/item/d9ebaf08-15d0-4868-9995-661961bab373" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 18 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.