AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
The Data Exchange with Ben Lorica
Best Practices for Building LLM-Backed Applications
Dec 7, 2023
Waleed Kadous, Chief Scientist at Anyscale, discusses best practices for building applications leveraging large language models. Topics include heuristics for working with open source models, differences between Code Lama and GitHub Co-pilot, challenges in deploying open source models, using spending data to save data, fine-tuning models in supervised machine learning, and exploring the potential of multimodal models.
53:50
Episode guests
AI Summary
AI Chapters
Episode notes
Podcast summary created with Snipd AI
Quick takeaways
- Open source models like GPT-3.5 Turbo and llama 270B offer a cost-effective alternative to expensive proprietary models like GPT-4.
- While open source models may have limitations, ongoing development is bridging the gaps in functionality.
Deep dives
Open source models provide pros and cons for users
Open source models like GPT-3.5 Turbo and llama 270B are cost-effective alternatives to proprietary models like GPT-4. While GPT-4 performs exceptionally well, its cost can be astronomical. In contrast, open source models can outperform GPT-3.5 Turbo at a fraction of the cost. For specific applications like email summarization or function calling, open source models may be preferable due to their affordability and effectiveness.
Get the Snipd
podcast app
Unlock the knowledge in podcasts with the podcast player of the future.
AI-powered
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode
Save any
moment
Hear something you like? Tap your headphones to save it with AI-generated key takeaways
Share
& Export
Send highlights to Twitter, WhatsApp or export them to Notion, Readwise & more
AI-powered
podcast player
Listen to all your favourite podcasts with AI-powered features
Discover
highlights
Listen to the best highlights from the podcasts you love and dive into the full episode