Deepseek lb#4188

Open

dipakg-lang wants to merge 2 commits into

AI-Hypercomputer:dsv4_model_integratefrom

dipakg-lang:deepseek_lb

Commits on Jun 17, 2026

feat(models): Integrate DeepSeek V4 architecture and routing
parambole
committed
Enabled auxilillary loss free load balancing and sequence wise load balancing for Deepseek.
dipakg-lang
committed