Hacker News new | ask | show | jobs
Controlled generation of OS LLMs – without impacting latency (youtube.com)
7 points by mezark 972 days ago
1 comments

TitanML Takeoff Inference Server demonstrating controlled generation