{ "sample_rate": 16000, "n_token": 256, "plbert": { "hidden_size": 512, "num_attention_heads": 8, "intermediate_size": 2048, "max_position_embeddings": 512, "num_hidden_layers": 12, "dropout": 0.1 }, "hidden_dim": 384, "text_encoder_kernel_size": 5, "n_layer": 3, "style_dim": 128, "max_dur": 50, "dropout": 0.2, "decoder": { "hid": 512, "out": 256, "res": 64 }, "istftnet": { "upsample_rates": [ 10, 4 ], "upsample_kernel_sizes": [ 20, 8 ], "upsample_initial_channel": 256, "resblock_kernel_sizes": [ 3, 7, 11 ], "resblock_dilation_sizes": [ [ 1, 3, 5 ], [ 1, 3, 5 ], [ 1, 3, 5 ] ], "gen_istft_n_fft": 20, "gen_istft_hop_size": 5 }, "style_encoder": { "dim_in": 32, "max_conv_dim": 256 }, "name": "swar", "version": "0.1.0" }