{ "_class_name": "VQModel", "_diffusers_version": "0.17.1", "act_fn": "silu", "block_out_channels": [ 8, 16, 16 ], "down_block_types": [ "DownEncoderBlock2D", "AttnDownEncoderBlock2D", "AttnDownEncoderBlock2D" ], "in_channels": 1, "latent_channels": 1, "layers_per_block": 1, "norm_num_groups": 8, "norm_type": "group", "num_vq_embeddings": 64, "out_channels": 1, "sample_size": [ 80, 80 ], "scaling_factor": 0.18215, "up_block_types": [ "UpDecoderBlock2D", "AttnUpDecoderBlock2D", "AttnUpDecoderBlock2D" ], "vq_embed_dim": 32 }