Upload fine-tuned Gemma-2-2B model with LoRA adapters

initial commit