datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
clean_data_vqaOCR-nvidia-Nemotron-VLM-Dataset-v2_wiki_fr-clean
Description
This dataset is a processed version of nvidia/Nemotron-VLM-Dataset-v2 to make it easier to use, particularly for a visual question answering task where answer is an OCR transcription.Specifically, the original dataset has been processed to provide the image directly as a PIL rather than a path in an image column.We've also translated question column to French containing 40 prompts based on via tutoiement, vouvoiement and imperative forms.
For further details, please… See the full description on the dataset page: https://huggingface.co/datasets/lbourdois/OCR-nvidia-Nemotron-VLM-Dataset-v2_wiki_fr-clean.tikz-dataset-clean
Cleaned TikZ dataset
benchmark: 40658 rows
train: 401482 rows
clean-colpali-datasettikz-dataset-clean-extended
Cleaned TikZ dataset
benchmark: 40658 rows
train: 401482 rows
text-captcha-data-cleanDataset_clean_X_DIAG_V2_sampleDataset_clean_X_DIAG_V2ppe-vqa-dataset-v2-cleanCrawl_Hafl_Clean_Labelling_DataASU_Turbulence_Clean_100k_dataset_v1texture_dataset_v4_cleanquirofano-dataset-clean
Quirofano Synthetic Surgical-Scrub Detection Dataset
Synthetic, CCTV-style hospital corridor/OR imagery for a 2-class person detector:
does a visible person wear the target mustard/orange surgical scrub set, or not.
Every image is AI-generated (no real people, no real hospital footage) and every
box is machine-annotated then human-curated (see Annotation pipeline below).
Classes
category_id (COCO, this file)
class id (YOLO, 0-indexed)
name
description
1… See the full description on the dataset page: https://huggingface.co/datasets/stormbreaker20/quirofano-dataset-clean.clean-platesmania-dataset-v2clean-platesmania-dataset
Dataset Description
The dataset consists of images scraped from platemania. After scraping 10k images, the dataset was cleaned to only include images from a front or diagonal view. The dataset can be used to train VLMs for fine-grained vehicle classification.
Dataset Structure
Data Fields
image: PIL Image object
text: String description/caption for the image
Data Splits
Split
Number of Examples
train
6592
validation
100… See the full description on the dataset page: https://huggingface.co/datasets/muqtasid87/clean-platesmania-dataset.pattern_dataset_v4_cleanastrology-dataset-cleanDataset_clean_X_DIAGpnid_data_cleancounting-object-sd-dataset3-cleanmy-image-text-dataset-cleanDataset_clean_X_DIAG_validationPAD3-Dataset-Revisi-Cleanmy-image-text-dataset-not-clean260227augmented_dataset_cleanscin-processed-clean-dataset