Add eval.py, README.md, config.json, training_args.json, model.safetensors

initial commit