Dataset · Hugging Face

jat-project/jat-dataset

Name: jat-dataset
Creator: jat-project
License: apache-2.0
Keywords: task_categories:reinforcement-learning, task_categories:text-generation, task_categories:question-answering, annotations_creators:found, annotations_creators:machine-generated, source_datasets:conceptual-captions, source_datasets:ok-vqa, source_datasets:oscar, license:apache-2.0, size_categories:100M<n<1B, format:parquet, modality:image, modality:text, modality:timeseries, library:datasets

JAT Dataset Dataset Description The Jack of All Trades (JAT) dataset combines a wide range of individual datasets. It includes expert demonstrations by expert RL agents, image and caption pairs, textual data and more. The JAT dataset is part of the JAT project, which aims to build a multimodal generalist agent. Paper: https://huggingface.co/papers/2402.09844 Usage >>> from datasets import load_dataset >>> dataset =… See the full description on the dataset page: https://huggingface.co/datasets/jat-project/jat-dataset.

View on Hugging FaceReport an issue

Author / organization	jat-project
License (as declared)	apache-2.0
Downloads	144.2K (as reported by Hugging Face)
Likes	83
Created	29 August 2023
Last updated at source	16 February 2024

imitation-learningreinforcement-learningtext-generationquestion-answeringgeneralist-agent

Embed this entry

<iframe src="https://silytics.si/embed/item/dc289baf-fd5b-419e-ba9d-9aa8ac382084" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Free, no tracking. More about feeds and embeds.

More from jat-project

jat-project/jat-dataset-tokenizedDataset

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 14 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.