Open-weight medical LLMs built for fast, low-cost serving. NVFP4 builds of Gemma 4 31B and Qwen3.8-27B, each with its Perigee drafter.
AI & ML interests
Fine-tuning, Post-training, Custom AI models, Open models, Owned AI, Private AI, On-device AI, Model evaluation, LLM deployment, GGUF, LoRA, Safetensors, llama.cpp, vLLM, MLOps
Recent Activity
View all activity