fix: add missing generation_config.json

Upload Qwen3_5ForConditionalGeneration LoRA SFT model

initial commit