1 2 3 4 5
_target_: text_recognizer.networks.vqvae.encoder.Encoder in_channels: 1 hidden_dim: 32 channels_multipliers: [1, 2, 4, 8, 8] dropout_rate: 0.25