{
  "_class_name": "UNet2DConditionWoCTModel",
  "_diffusers_version": "0.23.1",
  "_name_or_path": "h1t/oms_b_openclip_xl",
  "subfolder": "oms_module", 
  "act_fn": "silu",
  "attention_head_dim": [
    5,
    10,
    20
  ],
  "block_out_channels": [
    160,
    320,
    640
  ],
  "center_input_sample": false,
  "conv_in_kernel": 3,
  "conv_out_kernel": 3,
  "cross_attention_dim": 1024,
  "cross_attention_norm": null,
  "down_block_types": [
    "DownBlock2D",
    "CrossAttnDownBlock2D",
    "CrossAttnDownBlock2D"
  ],
  "downsample_padding": 1,
  "dual_cross_attention": false,
  "encoder_hid_dim": null,
  "encoder_hid_dim_type": null,
  "in_channels": 4,
  "layers_per_block": 2,
  "mid_block_only_cross_attention": null,
  "mid_block_scale_factor": 1,
  "mid_block_type": "UNetMidBlock2DCrossAttn",
  "norm_eps": 1e-05,
  "norm_num_groups": 32,
  "num_attention_heads": null,
  "only_cross_attention": false,
  "out_channels": 4,
  "resnet_out_scale_factor": 1.0,
  "sample_size": null,
  "transformer_layers_per_block": 1,
  "up_block_types": [
    "CrossAttnUpBlock2D",
    "CrossAttnUpBlock2D",
    "UpBlock2D"
  ],
  "upcast_attention": false,
  "use_linear_projection": false
}