Arcee.ai

At Arcee AI we’re proud to have a roster of clients that’s growing by the day, and we’re equally committed to our contributions to the open-source community–which include popular repos like MergeKit and DistillKit. Today, we’re psyched to let you know about one of our big ongoing collaborations: led by Prime Intellect and with other global partners, ???? ?????????? ?????? ?????????? ?????????????????????????? ?????? ?????????? ???????????? ?????? ????, ????????????, ?????? ????????! Utilizing our SOTA post-training pipeline, we have started fine-tuning the model and will fully open source the results in a week. ???????? ???????? ??????????????: ?? the base model ?? checkpoints ?? the post-trained model ?? and data! Follow us here on Linkedin, and Prime Intellect over on X, to stay across the big release!! #GenAI #opensource #EnterpriseAI?

?? Lots of voices are saying that, in the generative AI space, 2024 has been the year of the Small Language Model (SLM). As the pioneers of the SLM category, we’re certainly not going to argue with them. ?? As we ?? look ahead to 2025 ?? it’s also clear what’s coming: it will be the year of #AI agents. ?? Stay tuned to Arcee AI's channels for big announcements coming soon RE how to combine the performance and efficiency of SLMs with the transformative power of #AgenticAI. In the meantime, learn more about the capabilities of AI agents in this article–in which we explain the six main types of AI agents, including some that are already household names. https://lnkd.in/dFbtwbNK

?? The countdown is on to AWS re:Invent 2024 ?? where our team will be ready to tell you how and why small language models (SLMs) are the key to getting serious ROI from your #GenAI. They'll also be offering previews of ?our next big product?: how we've taken the category we created, small language models, and made them even more powerful for delivering immediate ROI to all teams in all industries. ?? Get ahead of the re:Invent game by scheduling a 1-on-1 with a member of our stellar team that will be on-site:? https://lnkd.in/eTs7VFHQ #EnterpriseAI #reinvent2024

?? Arcee AI will be at re:Invent in Vegas from December 2-6 ! ?? We're excited to have partnered with MongoDB as part of their AI Avenue – come find us at the AI Showcase Booth #1406. ?? Chat with our experts, see demos, and discover how ? Arcee AI + MongoDB Atlas + AWS ? offer the easiest and most secure enterprise AI solution. Our Customer Success ML Engineer William Barber will be doing a lightning talk on Day 1. The talk is a must-attend for anyone interested in what companies can do better when it come to getting ROI from your #GenAI. ? Full details on that and the other AI Avenue lightning talks here: https://lnkd.in/euFpxgFb

?? What’s the difference between a large language model #LLM and a small language model (SLM)? ?? What are the trade-offs when opting for small over large? ?? And – putting aside the cost savings – can small really beat large when it comes to performance and accuracy? Suddenly everyone seems to be talking about SLMs, so take your key questions to the source: Arcee AI, where ???? ?????????????????? ?????? ?????????? ???????????????? ?????????? ????????????????. Get your key questions answered on our “All About SLMs” page, and reach out to our team when you’re ready to learn how SLMs are bringing cutting-edge #GenAI to companies of all sizes. https://lnkd.in/eDpgPFFF #innovation #EnterpriseAI?#NLP

How does Arcee AI make ?????????? Language Models that have an ???????????????? impact when it comes to accuracy and performance? It’s thanks to our SOTA post-training pipeline, which includes steps like distillation, Spectrum training, and model merging. And as fierce advocates for open-source research, we’ve been proud to share with the community several major elements of the techniques that make up our training pipeline. Our contributions have perhaps been felt the the most with MergeKit – our repo that enables model merging. This week, we'd like to celebrate our Research team for the publication of their MergeKit academic paper in the ?????????????????????? ???? ?????? ???????? ???????????????????? ???? ?????????????????? ?????????????? ???? ?????????????? ???????????????? ???????????????????? (#EMNLP).? ?? Huge congrats to the the authors (tagged below) ?? You can check out the paper here:?https://lnkd.in/ebxjiNjp #GenAI #NLP #innovation??

?? As the end of 2024 approaches, the rise of ?????????? ???????????????? ???????????? (SLMs) is undeniable. ? Here at Arcee AI we're proud to be the pioneers of the SLM category – but even we have been pleasantly surprised by quickly the market has shifted towards grasping the fact that SMALL is better than large when it comes language models ?? If you happen to still be stuck in #LLM-world, fear not – we’ve put together a crash course on SLMs. Check it out here, and drop a comment below if you'd like to get in touch with our team to learn how easy it is to implement SLMs in any size company: https://lnkd.in/e-GbMwPf #GenAI #NLP #innovation?

Another week, another hire at Arcee AI that we're incredibly excited to announce: Davis Stone has joined us as our Head of Partnerships. Davis is no stranger to our team – he's been instrumental in building our close relationship with AWS, which started with the Global Startup Program and ?? expanded at record speed ?? to include a Strategic Collaboration Agreement, a Generative AI Competency Partnership, and election to the GenAI Edition of the AWS Clean Energy Accelerator. We're thrilled that Davis will now be bringing that same incredible energy to expanding our growing portfolio of partnerships. Welcome to the team, Davis! And if you're looking to join one of the most innovative #AI startups, check out our Careers page – we're currently looking to fill numerous roles: https://lnkd.in/eYuFddZQ

We all know that the #AI space is advancing at what can feel like the speed of light, and even for those of us who are ???????????????????? some of those advancements,?it’s challenging to keep up with the latest developments. That’s why we’re launching a new Arcee AI series: ?? Future Friday ??. Every Friday, we’ll bring you a comprehensive look at one of the newest AI tools, terms, or trends. This week, we start with our Growth Marketer / AI Extraordinaire Nora He's dive into Agentic AI, which explores: ???? What are AI Agentic Workflows? ?? What's the difference between GenAI and Agentic AI, and can they work together? ?? What industries are already using Agentic AI? Check out the article below, and get in touch with any topics you'd like us to focus on in upcoming Future Friday articles! BLOG: https://lnkd.in/ekKF8rBK #AgenticAI #GenAI

?? Why should you go check out our latest SLM release, the 3B parameter Vietnamese model Arcee-VyLinh? ?? Like all our models, it punches above its weight, with capabilities that surpass those of significantly larger models. ?? We trained it with our unique multi-stage process that maximizes language capabilities while maintaining efficiency - AND we provide full details on the model merging, DPO, and final refinement steps we used. ?? Arcee-VyLinh redefines what’s possible with the Vietnamese language – one of the many global languages that’s been thus far underserved in the world of #AI. ?? We focus on SLMs because of their many practical, immediate business applications. In the case of Arcee-VyLinh, those include customer service automation, content creation, document processing, chatbots, content moderation systems, ?language learning, and academic writing support. Arcee AI is proud of our contributions to the open source community and we welcome your feedback and contributions! Get started with Arcee-VyLinh by reading our blog, which includes links to the model cards:? https://lnkd.in/eVwgfM-H #GenAI #NLP #ML #vietnam