15 results

fast_vlm

inall models datasets apps agents people projects

Datasets

All datasets matching “fast_vlm”

riddhimanrana /coco-fastvlm-2k-val2017 Dataset Card for COCO FastVLM 2K Val2017 Structured Captions This dataset contains 2,000 high-quality image-text pairs generated from the COCO 2017 validation set using a FastVLM-based vision-language model with structured prompt engineering and automated distillation. Each caption follows a structured 7-point format to describe the visual content in detail, enabling high-fidelity fine-tuning of multimodal models. Dataset Details Dataset Description This… See the full description on the dataset page: https://huggingface.co/datasets/riddhimanrana/coco-fastvlm-2k-val2017.image1K<n<10K0 likes41 downloads1y agoHugging Face