Thomas Wolf

Thomas enjoys creating open-source and open-science products that catalyse the field of AI by making complex research, models and datasets widely accessible — for instance by creating and/or advising the Hugging Face Transformers, Datasets, Diffusers, Accelerate, DataTrove, smolagents and LeRobot libraries. When he's not building OSS libraries, he can be found pushing for open-science in AI/ML through projects like the BigScience Workshop on Large Language Models which led to the BLOOM model and dataset, the Ultra-Scale Playbook and FineWeb, or building the robotics effort at Hugging Face including projects around Reachy Mini and fostering AI for Science collaborations across academia and industry at huggingscience.co. His current research interests are directed toward whether AI systems can participate in generating genuinely new scientific knowledge — not just accelerating known workflows — which he writes about (not often enough) in his blog and substack. He also co-authored the reference book "Natural Language Processing with Transformers" published at O'Reilly, and produced educational content including "The Ultra-Scale Playbook" and videos like "The Future of Natural Language Processing".