1 2 3 4 5
_target_: text_recognizer.networks.vqvae.pixelcnn.Decoder out_channels: 1 hidden_dim: 8 channels_multipliers: [8, 2, 1] dropout_rate: 0.25