Skip to content

Dataset · philschmid

Amazon product descriptions for multimodal LLMs

philschmid/amazon-product-descriptions-vlm

Lightweight set of Amazon product images and English text descriptions for training VLMs.

What's inside

Contains product images and text descriptions generated by the Gemini Flash model. The dataset is a reduced and modified version of the bprateek/amazon_product_description set. The volume ranges from 1 to 10 thousand records in English.

What you can do with it

  • Training or fine-tuning multimodal models (VLM) for image-to-text description generation
  • Rapid prototyping of automated e-commerce product catalog population systems
  • Testing the combination of a visual encoder and language model on domain-specific data
Tasks
Image → text
Languages
English
Size
1K–10K rows
Downloads this month
284
Likes
35

Models trained on it

How to load it

from datasets import load_dataset

ds = load_dataset("philschmid/amazon-product-descriptions-vlm")