datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Open-Qwen2VL-Data
Introduction
This repository contains the data for Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources.
Project page: https://victorwz.github.io/Open-Qwen2VL
Code: https://github.com/Victorwz/Open-Qwen2VL
Dataset
ccs_ebdataset: CC3M-CC12M-SBU filtered by CLIP, we directly download the webdataset based on the released of curated subset of BLIP-1
datacomp_medium_dfn_webdataset: DataComp-Medium-128M filtered by DFN, we… See the full description on the dataset page: https://huggingface.co/datasets/weizhiwang/Open-Qwen2VL-Data.qwen2-vl-audio-dataOpen-Qwen2VL-Data-Interleavedsilkie-qwen2vl-dpo-filteredsilkie-qwen2vl-dpoQwen2-VL-2B-Instruct_hratio_0_0.1arabic_table_recognition_qwen2vlQwen2-VL-7B-Instruct_hratio_0_0.1_9168test-vision-generation-Qwen2VL-7B-Vision-Instruct
Dataset Card for test-vision-generation-Qwen2VL-7B-Vision-Instruct
This dataset has been created with distilabel.
Dataset Summary
This dataset contains a pipeline.yaml which can be used to reproduce the pipeline that generated it in distilabel using the distilabel CLI:
distilabel pipeline run --config "https://huggingface.co/datasets/TharunSivamani/test-vision-generation-Qwen2VL-7B-Vision-Instruct/raw/main/pipeline.yaml"
or explore the configuration:… See the full description on the dataset page: https://huggingface.co/datasets/TharunSivamani/test-vision-generation-Qwen2VL-7B-Vision-Instruct.qwen2vl-base-sampleschemical_sft_qwen2vlQwen2-VL-2B-Instruct-unsloth-bnb-4bit_Qari-0.2-eval-tashkilqwen2-vl-2b-blindspots
Qwen2-VL-2B Blind Spots
Dataset Description
This dataset contains 10 failure cases collected while testing the base model Qwen/Qwen2-VL-2B. The goal was to identify blind spots in the model’s visual understanding by evaluating it on counting, spatial reasoning, OCR, and generation tasks.
Model Evaluated
Model used:Qwen/Qwen2-VL-2B
Type of the model: Base vision-language model
Environment: Google Colab with GPU
Library: Hugging Face Transformers… See the full description on the dataset page: https://huggingface.co/datasets/Ellaft/qwen2-vl-2b-blindspots.qwen2vlpersian_document_ocr_qwen2vlamazon-qwen2vl-listing
Amazon Qwen2-VL Listing Dataset
This tiny dataset accompanies the LoRA adapter:
Model: https://huggingface.co/soupstick/qwen2vl-amazon-ft-lora
Files
data/train.json — LLaMA-Factory style JSON with fields:
images (list of filenames or a single filename)
instruction (prompt)
output (JSON-formatted string with title/bullets/description)
eval/eval_predictions.jsonl — model generations used for quick evaluation.
Note: This repo is intentionally lightweight for… See the full description on the dataset page: https://huggingface.co/datasets/soupstick/amazon-qwen2vl-listing.disaster-assessment-qwen2vlUIPro_Qwen2VL_SFT_4336kQAsvietmath-qwen2vl-2b-resultsQwen2-VL-2B-Instruct_hratio_0_0.5_qaQwen2-VL-7B-Instruct_hratio_0_0.5_qaqwen2-vl-testdataPeterFX-Qwen2VL-LoRA-VisionQwen2-VL-History_testtest