From nexa.ai
OmniVision-968M: World's Smallest Vision Language Model
10 10
Pocket-size multimodal model with 9x token reduction for on-device deployment
#hackernews #ycombinator #人工智慧 #技術創新 #視覺語言模型
10h ago
From nexa.ai
Nexa AI - The On-Device AI Open Source Community
0 1
Explore how small language models (1B/3B) perform in local RAG systems. Learn about real-world performance, LoRA fine-tuning, and building privacy-focused document AI on laptop and mobile phone.
on Nov 5