{ "model_type": "diffusion", "pipeline_tag": "text-to-image", "vae": { "type": "AutoencoderKL" }, "scheduler": { "type": "FlowMatchEulerDiscreteScheduler" } }