toilaluan/i1-consolidated

MiniT2I Processed Mixtures Processed image/caption subsets for MiniT2I training. Each config contains train rows with image, captions, key, and source. Load one subset with: from datasets import load_dataset ds = load_dataset("owner/repo", "pexels", split="train")