{ "unet": { "sample_size": 64, "in_channels": 4, "out_channels": 4, "layers_per_block": 2, "block_out_channels": [ 256, 512, 768, 1024 ], "down_block_types": [ "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "CrossAttnDownBlock2D", "DownBlock2D" ], "up_block_types": [ "UpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D", "CrossAttnUpBlock2D" ], "cross_attention_dim": 768, "attention_head_dim": 8 }, "prediction_type": "v_prediction", "zero_terminal_snr": true, "beta_schedule": "scaled_linear", "num_train_timesteps": 1000, "clip_penultimate": true, "guidance_rescale": 0.7, "image_size": 512, "vae": "madebyollin/sdxl-vae-fp16-fix" }