jsonl617 B

Visual Question Answering Pairs (JSONL)

Short VQA question/answer pairs referencing library images — a fixture for VQA loaders and eval scripts.

Computer vision ML training data NLP datasets

Preview — first 7 linesjsonl

{"id": "q1", "image": "fruit-still-life.png", "question": "What is on the table?", "answer": "fruit"}
{"id": "q2", "image": "coffee-mug.png", "question": "What object is shown?", "answer": "a coffee mug"}
{"id": "q3", "image": "potted-plant.png", "question": "Is there a plant?", "answer": "yes"}
{"id": "q4", "image": "chart-bar.png", "question": "What kind of chart is this?", "answer": "bar chart"}
{"id": "q5", "image": "invoice-scan.jpg", "question": "What document type is this?", "answer": "invoice"}
{"id": "q6", "image": "colour-bars.png", "question": "Are the colours solid blocks?", "answer": "yes"}

Specifications

Records: 6
Task: VQA
Schema: id, image, question, answer

What is a .jsonl file?

JSONL (JSON Lines) is a text format where each line is a complete, independent JSON value, allowing records to be streamed and appended without parsing the whole file. It is not itself a JSON array and each line must stand alone. It is common in logging, machine learning datasets, and data pipelines.

How to use this file

Use an example JSONL to test line-by-line streaming parsers, append-and-resume ingestion, and batch pipelines that process one record per line.

How to use this file for testing

“Visual Question Answering Pairs (JSONL)” is a deterministic Novus Examples fixture for Computer vision, ML training data, NLP datasets. A rendered detection scene annotated in COCO, YOLO, and Pascal-VOC formats — for testing annotation loaders, format converters, and vision pipelines against a known image.

Documented properties for this file: 6 records · schema: id, image, question, answer · task: VQA. Compare results against paired or grouped companions on this page when present (clean↔damaged, searchable↔scanned, or format twins) so scores stay reproducible across runs.

Download the file once, keep the path stable in CI or local scripts, and treat the spec table as the contract: dimensions, seeds, field lists, and roles are intentional. Corrupt or invalid samples are labelled as such — expect parsers to fail loudly rather than silently accept them.

AI/ML fixtures are fully synthetic with documented schemas — no real people or data. Test data loaders, tokenizers, annotation converters, embedding/vector stores, or eval-metric parsers against the known structure and fixed seeds.

Code examples

import json

with open("vqa-pairs.jsonl") as f:
    rows = [json.loads(line) for line in f]
print(len(rows), rows[0])

Generated by generation/ai_wave_b.py. Free for any use, no attribution required — license.