Hugging Face hits $2 billion valuation to build machine learning GitHub – TechCrunch

Hugging Face 🤗 has closed a new financing round. It’s a $100 million Series C round with a big valuation. After today’s round of funding, Hugging Face is now worth $2 billion.

Lux Capital leads the round, with Sequoia and Coatue investing in the company for the first time. Some of the startup’s existing investors rejoined. These investors include Addition, Betaworks, AIX Ventures, Cygni Capital, Kevin Durant and Olivier Pomel.

Despite a short life, Hugging Face has gone through an interesting evolution. When I first reported on the company in 2017, the startup was focused on a consumer app. It looked like another messaging app. And yet no one was on the other end of the conversation, as it was a chatbot app for bored teens.

That consumer bet didn’t pay off, but the company continued to iterate on its natural language processing technology. Hugging Face released the Transformers library on GitHub and immediately attracted a lot of attention – it currently has 62,000 stars and 14,000 forks on the platform.

Transformers lets you take popular NLP models, such as BERT, GPT-2, T5, or DistilBERT, and use these models to manipulate text in some way. For example, you can classify text, extract information, automatically answer questions, summarize text, generate text, etc.

Due to the success of this library, Hugging Face quickly became the primary repository for everything related to machine learning models – not just natural language processing. On the company’s website, you can browse thousands of pre-trained machine learning models, join the developer community with your own model, download datasets, and more.

Essentially, Hugging Face builds the GitHub of machine learning. It is a community driven platform with a lot of repositories. Developers can create, discover, and collaborate on ML models, datasets, and ML apps.

Hugging Face also offers hosted services, such as the Inference API that allows you to use thousands of models through a programming interface, and the ability to “AutoTrain” your model.

With today’s funding round, the company plans to do more of the same. 10,000 companies are now using Hugging Face in one way or another, so it’s not the time to spin again.

Leave a Comment