AIHugging Faceabout 1 hour ago
Nvidia Releases Open Nemotron 3.5 Audio Model
12 min read
Nvidia released the open Nemotron 3.5 model today for multilingual audio transcription. The 30-billion parameter system handles 100 hours of audio across 40 languages and eight speakers, processing the data in under three minutes on edge devices or Blackwell GPUs. Medical fine-tuning drops error rate from 30.8 to 11.8 percent.
Level
Hype check
Tap to vote and see what everyone thinks.
In this storyNvidia
