1 2 3 4 5
_target_: text_recognizer.networks.vqvae.decoder.Decoder out_channels: 1 hidden_dim: 64 channels_multipliers: [8, 4, 2, 1] dropout_rate: 0.0