I’m interested in redo’ing the post training for Qwen3.5:27b/9b specifically for the hermes-agent harness by @NousResearch If anyone is willing to let me use a RTX PRO 6000 or something with >VRAM (even AMD), I would be happy to set this up for the community and share my results on huggingface We need improved agentic intelligence on 16 gb and 24 gb systems.