In this blog, we will be exploring the newly hyped Jurassic-1 model that has just made an entrance in the arena of Natural Language Processing (NLP). This model, hailing from Israel, was built by a startup company called AI21 Labs with a whopping 178 billion parameters. It made an entrance, just to rival the incumbent heavy-weight champion, OpenAI’s GPT-3, running at 175 billion. To show itself off, the lab released their Beta-version Studio quite recently to engage users in experiencing the strengths and limitations of their model on a variety of NLP use cases. At Lokibots, our team of machine learning engineers and experts explored this model to a great extent to learn how well it performs.
CLICK HERE TO READ THE FULL CONTENT OF THIS BLOG ON MEDIUM
At Lokibots, we deliver robotic process automation products and services that require zero coding from the user. This means we provide highly user interactive interfaces for creating the required automation tasks. These tasks range from automating desktop applications to document understanding all the way to chatbot implementations having emotional intelligence. Therefore, for our products and services, the above use cases such as code generation, table question answering and converting text to table are some of the features we hope to integrate in the near future.
With code generation, we hope to provide our bots the ability to interpret user audio instructions into deployable system software. This, as we hope, will help our users instruct bots in plain English or any other language on what it should automate. Likewise with table question answering, we are looking to integrate this feature in our existing chatbot applications so that the bot has its own knowledge base from which it will generate responses. Similarly, with converting text to table, our existing document understanding applications would get a new boost in performing data extraction tasks.