config.json

{
  "act_fn": "silu",
  "block_out_channels": [
    128,
    128,
    256,
    256,
    512
  ],
  "down_block_types": [
    "DownEncoderBlock2D",
    "DownEncoderBlock2D",
    "DownEncoderBlock2D",
    "DownEncoderBlock2D",
    "AttnDownEncoderBlock2D"
  ],
  "force_upcast": true,
  "in_channels": 3,
  "latent_channels": 16,
  "layers_per_block": 2,
  "mid_block_add_attention": true,
  "norm_num_groups": 32,
  "out_channels": 3,
  "sample_size": 256,
  "scaling_factor": 0.2614,
  "up_block_types": [
    "UpDecoderBlock2D",
    "UpDecoderBlock2D",
    "UpDecoderBlock2D",
    "UpDecoderBlock2D",
    "UpDecoderBlock2D"
  ],
  "use_post_quant_conv": true,
  "use_quant_conv": true
}