We built DeepL's next-generation LLMs with FP8 for training and inference (2025) | Hacker News - But Better

We built DeepL's next-generation LLMs with FP8 for training and inference (2025)deepl.com

Discover how DeepL harnessed FP8 for training and inference in next-gen LLMs, boosting throughput and model quality. Learn about our journey with NVIDIA's technology, achieving faster training and superior translations…

3 points•mooreds•3 days ago•0 comments•

#training #llms #inference #fp8 #generation #deepls #deepls next #built deepls

0 comments

No comments yet.

Related stories

OpenAI halts frontier-model training amid string of agent misalignment incidentsarstechnica.com
Ars Technica · 0 points · 2 days ago
Show HN: StarSkirmish, an arena where LLMs create StarCraft Brood War botsstarskirmish.com
Hacker News · 3 points · 4 days ago
Launch HN: Magnitude (YC S25) – Self-optimizing inference engine for agentsgithub.com
Hacker News · 124 points · about 8 hours ago
NobodyWho is an inference engine that lets you run LLMs locally and efficientlygithub.com
Hacker News · 1 points · 8 days ago
LLMs respond differently to harmful prompts when AI watermarking is usedarstechnica.com
Ars Technica · 0 points · 13 days ago
The Basics of Transformer Inferencejax-ml.github.io
Hacker News · 2 points · 9 days ago