{
  "_class_name": "ControlNetModel",
  "_commit_hash": null,
  "_diffusers_version": "0.29.2",
  "_name_or_path": "diffusers/controlnet-canny-sdxl-1.0-small",
  "_use_default_values": [
    "mid_block_type"
  ],
  "act_fn": "silu",
  "addition_embed_type": "text_time",
  "addition_embed_type_num_heads": 64,
  "addition_time_embed_dim": 256,
  "attention_head_dim": [
    5,
    10,
    20
  ],
  "block_out_channels": [
    320,
    640,
    1280
  ],
  "class_embed_type": null,
  "conditioning_channels": 3,
  "conditioning_embedding_out_channels": [
    16,
    32,
    96,
    256
  ],
  "controlnet_conditioning_channel_order": "rgb",
  "cross_attention_dim": 2048,
  "down_block_types": [
    "DownBlock2D",
    "DownBlock2D",
    "DownBlock2D"
  ],
  "downsample_padding": 1,
  "encoder_hid_dim": null,
  "encoder_hid_dim_type": null,
  "flip_sin_to_cos": true,
  "freq_shift": 0,
  "global_pool_conditions": false,
  "in_channels": 4,
  "layers_per_block": 2,
  "mid_block_scale_factor": 1,
  "mid_block_type": "UNetMidBlock2DCrossAttn",
  "neuron": {
    "auto_cast": null,
    "auto_cast_type": null,
    "compiler_type": "neuronx-cc",
    "compiler_version": "2.14.227.0+2d4f85be",
    "dynamic_batch_size": false,
    "inline_weights_to_neff": false,
    "input_names": [
      "sample",
      "timestep",
      "encoder_hidden_states",
      "controlnet_cond",
      "conditioning_scale",
      "text_embeds",
      "time_ids"
    ],
    "model_type": "controlnet",
    "optlevel": "2",
    "output_attentions": false,
    "output_hidden_states": false,
    "output_names": [
      "down_block_res_samples",
      "mid_block_res_sample"
    ],
    "static_batch_size": 1,
    "static_encoder_hidden_size": 2048,
    "static_height": 128,
    "static_num_channels": 4,
    "static_sequence_length": 77,
    "static_vae_scale_factor": 8,
    "static_width": 128
  },
  "norm_eps": 1e-05,
  "norm_num_groups": 32,
  "num_attention_heads": null,
  "num_class_embeds": null,
  "only_cross_attention": false,
  "projection_class_embeddings_input_dim": 2816,
  "resnet_time_scale_shift": "default",
  "task": "semantic-segmentation",
  "transformer_layers_per_block": [
    0,
    0,
    0
  ],
  "transformers_version": null,
  "upcast_attention": null,
  "use_linear_projection": true
}