In this dynamic discussion, Jonathan Frankle, Chief Scientist at MosaicML, shares insights on their $1.3 billion acquisition by Databricks. Lin Qiao, CEO of Fireworks AI, talks about optimizing PyTorch for inference. Aman Sanger from Cursor reveals innovative memory strategies for AI coding. Aravind Srinivas discusses the impressive growth of Perplexity AI, hitting 1 million installs, while Jeremy Howard emphasizes the need for accessible AI. Together, they explore the vibrant AI startup landscape showcased at NeurIPS 2023, reflecting on innovation, collaboration, and the future of technology.

On this episode, we’re joined by Brandon Duderstadt, Co-Founder and CEO of Nomic AI. Both of Nomic AI’s products, Atlas and GPT4All, aim to improve the explainability and accessibility of AI.We discuss:- (0:55) What GPT4All is and its value proposition.- (6:56) The advantages of using smaller LLMs for specific tasks. - (9:42) Brandon’s thoughts on the cost of training LLMs. - (10:50) Details about the current state of fine-tuning LLMs. - (12:20) What quantization is and what it does. - (21:16) What Atlas is and what it allows you to do.- (27:30) Training code models versus language models.- (32:19) Details around evaluating different models.- (38:34) The opportunity for smaller companies to build open-source models. - (42:00) Prompt chaining versus fine-tuning models.Resources mentioned:Brandon Duderstadt - https://www.linkedin.com/in/brandon-duderstadt-a3269112a/Nomic AI - https://www.linkedin.com/company/nomic-ai/Nomic AI Website - https://home.nomic.ai/Thanks for listening to the Gradient Dissent podcast, brought to you by Weights & Biases. If you enjoyed this episode, please leave a review to help get the word out about the show. And be sure to subscribe so you never miss another insightful conversation.#OCR #DeepLearning #AI #Modeling #ML