David Kyazze's picture

2 9 90

David Kyazze

dkyazze

·

Dkyazzentwatwa

AI & ML interests

Tiny AI computing for impoverished communities!

Recent Activity

upvoted a paper about 4 hours ago

Nemotron-Cascade: Scaling Cascaded Reinforcement Learning for General-Purpose Reasoning Models

reacted to NJULUR's post with 🔥 2 days ago

Nemotron 3 Nano (30B A3B): The Efficient Open Agent Model 🔥 NVIDIA Nemotron 3 Nano is a new standard for high-throughput, accurate AI agents, built on a groundbreaking Hybrid Mamba-Transformer Sparse Mixture-of-Experts (MoE) architecture. Speed: 4x faster than Nemotron 2 Nano Scale: 1 Million-token context window. Intelligence: Best-in-class reasoning accuracy with ~3.6B active parameters per token. Control: Features Reasoning ON/OFF modes and a customizable Thinking Budget. Open: Full open weights, training recipes, and all major datasets (including 13M post-training samples and NeMo Gym RL environments). Get Started: Download the model and datasets on Hugging Face: https://huggingface.co/blog/nvidia/nemotron-3-nano-efficient-open-intelligent-models

liked a model 12 months ago

segolilylabs/Lily-Cybersecurity-7B-v0.2

View all activity

Organizations

None yet

dkyazze 's models

None public yet