EP2: ARCEE.AI small language models, open source and cost efficient AI | AWS for AI podcast скачать в хорошем качестве

EP2: ARCEE.AI small language models, open source and cost efficient AI | AWS for AI podcast

Join us for an enlightening conversation with Julien Simon, VP and Chief Evangelist at ARCEE.AI , as he shares deep insights on building practical and cost-efficient AI solutions. From his extensive experience at AWS, Hugging Face, and now ARCEE.AI, Julien discusses why "small is beautiful" when it comes to language models, revealing how 10B parameter models can now match the performance of much larger 72B models from just months ago. Learn about innovative techniques like model merging, the importance of proper infrastructure choices, and practical advice for organizations starting their AI journey. This episode covers critical topics including: ● Why small language models are the future of enterprise AI ● How to optimize costs while maintaining performance ● The role of CPU vs GPU inference ● Essential architecture considerations for AI workloads ● Best practices for building production-ready AI systems Whether you're a startup, enterprise, or public sector organization, this episode offers invaluable guidance on building scalable, efficient, and practical AI solutions in today's rapidly evolving landscape. Learn more: Build and scale the next wave of AI innovation on AWS: https://go.aws/ai ARCEE.AI: https://www.arcee.ai/ Julien Simon Youtube channel :    / @juliensimonfr   00:00:00 : Introduction 00:02:18 : Journey into AI 00:06:40 : Arcee.ai small language models champion 00:09:02 : Arcee.ai global presence 00:10:19 : Use cases for SLMs and AI agents 00:15:00 : Post training with model merging, model distillation 00:17:20 : Model routing 00:19:15 : Orchestra drag and drop agentic platform 00:20:42 : How to build the best SLM ? 00:23:29 : Synthetic data and data quality 00:25:26 : Open source in AI 00:28:04 : Reflecting cultural nuances 00:31:02 : Biases in synthetic data 00:34:55 : What is an SLM 00:36:33 : Obsessing on cost efficiency 00:39:37 : CPU Inference 00:41:49 : Infrastructure and model choice 00:45:38 : GPU-less and microservices architecture 00:48:02 : Training on AWS: Hyperpod & Trainium 00:55:48 : Key advice for organizations starting with AI 01:02:14 : Closing remarks and resources Subscribe to AWS: https://go.aws/subscribe Sign up for AWS: https://go.aws/signup AWS free tier: https://go.aws/free Explore more: https://go.aws/more Contact AWS: https://go.aws/contact Next steps: Explore on AWS in Analyst Research: https://go.aws/reports Discover, deploy, and manage software that runs on AWS: https://go.aws/marketplace Join the AWS Partner Network: https://go.aws/partners Learn more on how Amazon builds and operates software: https://go.aws/library Do you have technical AWS questions? Ask the community of experts on AWS re:Post: https://go.aws/3lPaoPb Why AWS? Amazon Web Services (AWS) is the world’s most comprehensive and broadly adopted cloud. Millions of customers—including the fastest-growing startups, largest enterprises, and leading government agencies—use AWS to be more agile, lower costs, and innovate faster. #AWS #AmazonWebServices #AWSForAI #CloudComputing