Add training metadata
Add Ollama Modelfile
Add model card
Add LoRA adapters — loss=0.2532
initial commit