What's New in NLP? #4
Welcome to the January edition of Cohere’s monthly newsletter, where we share the latest updates in the world of LLMs.
News and Announcements
Hola ??,
Exciting news for the development community! We are proud to announce that Cohere's state-of-the-art language AI is?now available on Amazon SageMaker . This will provide developers, data scientists, and business analysts with a more streamlined way to deploy Cohere's pre-trained language generation model, thanks to Amazon SageMaker's end-to-end machine learning service.
Moreover, we have made a?partnership with Google Cloud's Vertex Matching Engine . This collaboration aims to simplify the process of text comparison and embeddings by making vector similarity search setups easier and more efficient for developers to utilize.?
In other news, HyperWrite, a language AI-first startup,?has chosen Cohere's LLMs to power its generative AI service . The company is working to help writers of all kinds create the content they need to achieve their goals. We have also released the?second episode in the series on Talking Language AI , where Vincent Warmerdam, a machine learning engineer at Explosion, the company behind spaCY and Prodigy for data labeling, shares the tools he builds to improve training data. Also, Cohere Co-founder and CEO, Aidan Gomez?reflects upon our growth and journey during 2022 .
All of this and much more are below. We want to make this your resource for NLP, and we'd love your feedback. If you have any, please feel free to?email us .?
What's New in NLP?
Generative AI is starting to become the future of content creation, but what exactly is?generative AI ? Simply put, it's the theme of AI systems or products, capable of generating high-quality text, images, videos, and more.
With its vast market potential, generative AI has the potential to revolutionize many industries and drive innovation in a wide range of fields. In the creative arts, generative AI can be used to?generate unique and engaging content , such as music or visual art, with minimal need for human input. In the business world, generative AI can be used to generate reports, presentations, and other business documents, reducing the need for manual data analysis and enhancing productivity.
At Cohere, we specialize in language-based generative AI, and our goal is to enable developers to add this powerful technology to their toolkits and build impactful applications with it. If you are curious or would like to learn more about generative AI,?check out this guide .
MusicLM: Generating Music From Text ?is one new paper that demonstrates generating music from text prompts. Check out some of its outputs on?this demo page . We’ve also seen some more work on?watermarking the text generated by LLMs ?– a research area important for tracking the outputs of these models as they proliferate. Finally, it appears that we can drastically reduce the attention layers in Transformer models with a method like H3, explained in?Hungry Hungry Hippos: Towards Language Modeling with State Space Models .
Cohere Community Section
Contribution by Yinbo Shi.?Join the Cohere community and share yours .
Meet Our Co:mmunity Champion
We’re incredibly excited to introduce you to our first Co:mmunity Champion:?Sangeetha Venkatesan#0414!
???? Am Sangeetha Venkatesan
I am currently pursuing my Masters in Computer Science at Syracuse University(Go Orange ??) and looking forward for my graduation by May 2023. ?? ??
?? There is a famous saying “Data will not speak to you unless you are ready to listen”. This applies to more on “Conversational AI domain”?? ?? ???where working with text data to understand the intent of customers is my everyday routine!
??? I love writing! This definitely helped as a data scientist working on Natural language Understanding to break down the larger broader problem of Language Understanding into smaller implementable business solutions that tackle the expectations of the stakeholders in the space of Artificial Intelligence.
Message to the community:?
Community is the heart of the product.?I love discussing, asking, and answering questions, it has helped me substantially to develop my skills around different use cases and gives valuable perspectives for me to solve a problem. Considering text data - ?? - there are so many modalities around it ~ text, language, intent, and entities. There are deeper insights that go unnoticed just on the abstraction of designing a chatbot. Cohere plays well with abstraction and, at the same time, works on strengthening the core component of NLP’s backbone!?The kind of acknowledgment people receive for their contribution is what makes the field even more interesting and engaging! Appreciate Cohere for that ???
Make sure to say ??to Sangeetha on Discord and check out her Medium post?Implicit vs Explicit Knowledge models , where she covers retrieval augmented language models, conversational dialogue models, and many others! She writes about Cohere's co.chat & Conversant and shares a balanced view of the current state of the art and what's next.
Ask Ed Grefenstette, Head of ML at Cohere, Anything!
In December, we ran a?Reddit AMA? and?live Zoom AMA with Ed Grefenstette , Head of Machine Learning at Cohere, and an Honorary Professor at UCL. Ed's previous industry experience spans Facebook AI Research, DeepMind, and Dark Blue Labs (acquired by Google in 2014), where he was the CTO. Prior to this, Ed worked at the University of Oxford's Department of Computer Science, and was a Fulford Junior Research Fellow at Somerville College, whilst also lecturing students at Hertford College taking Oxford's new computer science and philosophy course. Ed's research interests span topics including natural language and generation, machine reasoning, open-ended learning, and meta-learning.
Ed will continue answering your questions on co:mmunity Discord.?
领英推荐
Coolest Co:mmunity Projects in January
Here’s a rundown of some of the coolest projects coming from our co:mmunity in January:
Hungry for more demos? Make sure to join co:lab friday#12 for a live demo session. Details below!
co:lab friday #12: community demo showcase?
Feb 03, Friday at 12pm EST
How to Add AI to your App by Michael Kozakov and Nick Frosst
Feb 07, Wed at 6pm EST
What's New on the Blog
Large language models must determine similarity between words or sentences. This task can be difficult, but word and sentence embeddings offer valuable aid. This post explores various concepts of similarity.
Word and sentence embeddings are essential components of language models. They are used to represent words and sentences in a numerical format that can be used by machine learning algorithms. This is a very simple introduction to what they are.
Nils Reimers, Director of Machine Learning at Cohere, shares his vision for the future of semantic search in the enterprise. He emphasizes how utilizing semantic search will be key to utilizing unstructured text data for knowledge transfer and preservation within businesses
Learn about using Generative AI with Cohere in Part 2 of our series, where we dive into an ideation framework for LLMs and explore specific use case examples.
At Cohere, we're excited about natural language processing and all the amazing accomplishments it has made in recent years. Staying up to date with the latest research can be challenging, though, as new papers come out every month. That's why we put together this list of some of the best papers on NLP for 2022—so you don't have to miss a thing!?
Updates From Cohere For AI?
Cohere For AI is a non-profit research lab that seeks to solve complex machine learning problems.?Learn more ?about the lab and follow us on?Twitter .
Hosted Events
Join the Cohere co:mmunity on Discord!
Want to explore more demos and collaborate with other NLP practitioners??Join in on the conversation on our Discord