{
  "_class_name": "ControlNetXSModel",
  "_diffusers_version": "0.24.0.dev0",
  "_name_or_path": "/Users/umer/.cache/huggingface/hub/models--stabilityai--stable-diffusion-2-1-base/snapshots/5ede9e4bf3e3fd1cb0ef2f7a3fff13ee514fdf06/unet",
  "act_fn": "silu",
  "addition_embed_type_num_heads": 64,
  "attention_head_dim": [
    1,
    1,
    2,
    2
  ],
  "attention_type": "default",
  "base_model_channel_sizes": {
    "down": [
      [
        4,
        320
      ],
      [
        320,
        320
      ],
      [
        320,
        320
      ],
      [
        320,
        320
      ],
      [
        320,
        640
      ],
      [
        640,
        640
      ],
      [
        640,
        640
      ],
      [
        640,
        1280
      ],
      [
        1280,
        1280
      ],
      [
        1280,
        1280
      ],
      [
        1280,
        1280
      ],
      [
        1280,
        1280
      ]
    ],
    "mid": [
      [
        1280,
        1280
      ]
    ],
    "up": [
      [
        2560,
        1280
      ],
      [
        2560,
        1280
      ],
      [
        2560,
        1280
      ],
      [
        2560,
        1280
      ],
      [
        2560,
        1280
      ],
      [
        1920,
        1280
      ],
      [
        1920,
        640
      ],
      [
        1280,
        640
      ],
      [
        960,
        640
      ],
      [
        960,
        320
      ],
      [
        640,
        320
      ],
      [
        640,
        320
      ]
    ]
  },
  "block_out_channels": [
    4,
    8,
    16,
    16
  ],
  "center_input_sample": false,
  "conditioning_block_sizes": [
    16,
    32,
    96,
    256
  ],
  "conditioning_channels": 3,
  "controlnet_conditioning_channel_order": "rgb",
  "conv_in_kernel": 3,
  "conv_out_kernel": 3,
  "cross_attention_dim": 1024,
  "cross_attention_norm": null,
  "down_block_types": [
    "CrossAttnDownBlock2D",
    "CrossAttnDownBlock2D",
    "CrossAttnDownBlock2D",
    "DownBlock2D"
  ],
  "downsample_padding": 1,
  "dropout": 0.0,
  "dual_cross_attention": false,
  "encoder_hid_dim": null,
  "encoder_hid_dim_type": null,
  "in_channels": 4,
  "layers_per_block": 2,
  "learn_embedding": true,
  "mid_block_only_cross_attention": null,
  "mid_block_scale_factor": 1,
  "mid_block_type": "UNetMidBlock2DCrossAttn",
  "norm_eps": 1e-05,
  "norm_num_groups": 4,
  "num_attention_heads": null,
  "only_cross_attention": false,
  "out_channels": 4,
  "resnet_out_scale_factor": 1.0,
  "resnet_skip_time_act": false,
  "resnet_time_scale_shift": "default",
  "reverse_transformer_layers_per_block": null,
  "sample_size": 64,
  "time_cond_proj_dim": null,
  "time_embedding_act_fn": null,
  "time_embedding_dim": 1280,
  "time_embedding_input_dim": 320,
  "time_embedding_mix": 1.0,
  "time_embedding_type": "positional",
  "timestep_post_act": null,
  "transformer_layers_per_block": 1,
  "up_block_types": [
    "UpBlock2D",
    "CrossAttnUpBlock2D",
    "CrossAttnUpBlock2D",
    "CrossAttnUpBlock2D"
  ],
  "upcast_attention": false,
  "use_linear_projection": true
}