Natural Language Processing has been gaining momentum ever since artificial intelligence brought forth the possibility of computers comprehending natural language data, thereby solidifying the interactive ground between humans and computers. Companies depend on Natural Language Processing (NLP) for a long list of activities like sentiment analysis, social media monitoring, personalized and targeted advertising, hiring and recruitment and so on.
Earlier this day, Amazon Web Services Inc. announced its collaboration with Hugging Face Inc., a New York based artificial intelligence startup. The prime objective of the partnership is to pool the resources of both companies to simply and improve NLP models.
Hugging Face attracted the attention of the tech giant with its renowned Transformers library, that facilitates an easy access to a multitude of popular NLP networks that are trained on AI frameworks such as PyTorch and TensorFlow. The main aim of Transformers is to ensure NLP accessibility to everyone. Its pre trained models that can perform tasks like classification, information extraction, translation, summarization etc. on texts attests to the efforts made towards this goal.
Through the partnership, Amazon will be able to pool its programming tools with the library, thereby bringing an estimated 7000 NLP models to Amazon SageMaker,a machine learning service designed to facilitate training and deployment of machine learning models.
A short list of services built using SageMaker was announced by Hugging Face, namely, the AutoNLP and the Accelerated Inference API, both with remarkable features with the capability to improve NLP and machine learning models.
The General availability of Hugging Face Deep Learning Containers was announced by Amazon. The service enables developers with building language models provided by Hugging Face on SageMaker. The resources channeled from both the companies will help to save time with respect to NLP model experimentation with the backing of integrated development environment. SageMaker’s distributed training capabilities will also prove to be an added advantage for developers.
According to Clement Delague, Hugging Face Chief Executive,
“Our transformers can help them build virtually any natural language processing application at a fraction of the time, cost and complexity they’d achieve their own, helping organizations take their solutions to market quickly.”
In short, the partnership will enable the customers to train NLP models with more convenience, making use of abilities like summarization, translation, and conversation chatbots.
An earlier announcement made by Amazon today indicated a new service, AWS Media Intelligence Solutions, which facilitates easy integration of AI into media content workflows. The service is a blend of different AI services of the company according to a blogpost authored by Vasi Philomin and Esther Lee. Philomin is the general manager of Amazon’s Machine Learning and AI, while Lee is the product manager of AWS AI language service.