3min snip

Latent Space: The AI Engineer Podcast — Practitioners talking LLMs, CodeGen, Agents, Multimodality, AI UX, GPU Infra and all things Software 3.0 cover image

The Mathematics of Training LLMs — with Quentin Anthony of Eleuther AI

Latent Space: The AI Engineer Podcast — Practitioners talking LLMs, CodeGen, Agents, Multimodality, AI UX, GPU Infra and all things Software 3.0

NOTE

Memory Requirements in Deep Learning: Precision and Quantization

Memory requirements in deep learning have evolved with the advancement of hardware and software. Initially, FP 32 precision was common, but with the introduction of tensor cores, mixed precision became popular, requiring both FP 16 and FP 32 copies of weights. The range and precision trade-off led to the emergence of BF 16 support. Additionally, quantization techniques allow models to be represented in smaller formats like in date and in four without significant loss in accuracy. Ultimately, the last bits of precision become less crucial in stochastic deep learning problems.

00:00

Get the Snipd
podcast app

Unlock the knowledge in podcasts with the podcast player of the future.
App store bannerPlay store banner

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

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