datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
1980s-photo-prompts
1980s Photo Prompts
Ten image-to-image prompts for the viral "1980s photo" trend, each paired with the image it actually produced on the first attempt. Generated on September 16, 2026 with Claude Imagine using GPT Image 2.5 and Nano Banana 2. No retouching, no cherry-picking.
Every prompt follows the same rule: keep my face, facial features and skin tone exactly the same, change everything else to 1985. The identity line comes first, then a specific year, then hair, clothes… See the full description on the dataset page: https://huggingface.co/datasets/claudeimagine/1980s-photo-prompts.Chinese-Vintage-Cartoons-1950s-1980s-Captionedimport os
import numpy as np
import uuid
from PIL import Image
from datasets import load_dataset
import cv2
# 加载数据集
ds = load_dataset("svjack/Chinese-Vintage-Cartoons-1950s-1980s-Captioned")
train_data = ds["train"]
# 输出路径
output_path = "Chinese-Vintage-Cartoons-1950s-1980s-Captioned-1024x1024" # 替换为你想要的输出路径
# 目标尺寸
target_width, target_height = 1024, 1024# 迭代数据集
for item in train_data:
# 获取图像和文本
image = item["image"]
joy_caption = item["joy-caption"]
# 生成 UUID… See the full description on the dataset page: https://huggingface.co/datasets/svjack/Chinese-Vintage-Cartoons-1950s-1980s-Captioned.Chinese-Vintage-Cartoons-1950s-1980s