What is Hugging Face? Hugging Face Review, Features

Hugging Face is a very active artificial intelligence startup.

About Hugging Face

Hugging Face was founded in 2016. It was originally a company that developed chatbots. Their goal was to use chatbots to entertain bored young people. Of course, although such an idea looks good, it obviously did not make it big. HuggingFace's transformation mainly comes from its contribution to the field of NLP. Shortly after the release of Bert, they contributed a Bert pre-trained model based on Pytorch, namely pytorch-pretrained-bert. I believe that many people have used this library. This library is very easy to use, so it has continued to expand along with the development of NLP models. Today, HuggingFace has integrated their pre-trained models in the NLP field and released the Transformers library. Transformers provides thousands of pre-trained models (including the well-known Bert, GPT, GPT-2, XLM, etc.), supporting text classification, information extraction, question and answer, summarization, translation, and text generation in more than 100 languages. Its purpose is to make the most advanced NLP technology easy for everyone to use.

Hugging Face's Business

HuggingFace's official website is already a large AI community. They operate many products. The community model of HuggingFace is different from our existing communities. They are equivalent to a brand zone for an institution. Each institution can publish its own models, data sets, and Spaces (a newly launched hosting AI application or display AI application) on it.

• A community where institutional brands publish models, data sets, and AI application demonstrations;

• Open source models, datasets, and tool libraries;

• NLP related solutions;

• NLP related courses.

Advantages

Hugging Face has a great influence in the field of artificial intelligence and operates a large community. Its growth and transformation also changed a lot because of its contribution of Bert based on PyTorch. At present, their community is very active and continues to contribute excellent products. The transformers, tokenizers, datasets, and accelerate libraries they provide have greatly promoted NLP-related practices. In addition, the contributions and support of major companies in its community have also made the community of this company more active and healthy. In fact, China should also need such companies and atmosphere. Not only do we need some excellent companies that are willing to make "contributions", but we also need the giants to put down their status and participate in the construction of the community and support some small start-ups.