Delete exp-vits-lr-3e-4

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1.log

@@ -1,1152 +0,0 @@
-# python3 -m espnet2.bin.gan_tts_train --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.1.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.1.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000 
-# Started at Fri Dec  1 15:58:34 UTC 2023
-#
-/data2/p280965/tts/espnet/tools/venv/bin/python3 /data2/p280965/tts/espnet/espnet2/bin/gan_tts_train.py --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.1.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.1.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000
-[wieling-3-a100] 2023-12-01 15:58:40,398 (gan_tts:293) INFO: Vocabulary size: 46
-[wieling-3-a100] 2023-12-01 15:58:40,545 (encoder:174) INFO: encoder self-attention layer type = relative self-attention
-/data2/p280965/tts/espnet/tools/venv/lib/python3.9/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
-  warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
-/data2/p280965/tts/espnet/espnet2/gan_tts/vits/monotonic_align/__init__.py:19: UserWarning: Cython version is not available. Fallback to 'EXPERIMETAL' numba version. If you want to use the cython version, please build it as follows: `cd espnet2/gan_tts/vits/monotonic_align; python setup.py build_ext --inplace`
-  warnings.warn(
-[wieling-3-a100] 2023-12-01 15:58:41,774 (abs_task:1268) INFO: pytorch.version=2.1.0+cu121, cuda.available=True, cudnn.version=8902, cudnn.benchmark=False, cudnn.deterministic=False
-[wieling-3-a100] 2023-12-01 15:58:41,789 (abs_task:1269) INFO: Model structure:
-ESPnetGANTTSModel(
-  (feats_extract): LogMelFbank(
-    (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-    (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=80, fmax=7600, htk=False)
-  )
-  (tts): VITS(
-    (generator): VITSGenerator(
-      (text_encoder): TextEncoder(
-        (emb): Embedding(46, 192)
-        (encoder): Encoder(
-          (embed): Sequential(
-            (0): RelPositionalEncoding(
-              (dropout): Dropout(p=0.0, inplace=False)
-            )
-          )
-          (encoders): MultiSequential(
-            (0): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (1): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (2): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (3): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (4): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (5): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-          )
-          (after_norm): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (decoder): HiFiGANGenerator(
-        (input_conv): Conv1d(192, 512, kernel_size=(7,), stride=(1,), padding=(3,))
-        (upsamples): ModuleList(
-          (0): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(512, 256, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (1): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(256, 128, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (2): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(128, 64, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-          (3): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(64, 32, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-        )
-        (blocks): ModuleList(
-          (0): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (1): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (2): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (3): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (4): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (5): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (6): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (7): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (8): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (9): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (10): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (11): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-        )
-        (output_conv): Sequential(
-          (0): LeakyReLU(negative_slope=0.01)
-          (1): Conv1d(32, 1, kernel_size=(7,), stride=(1,), padding=(3,))
-          (2): Tanh()
-        )
-        (global_conv): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
-      )
-      (posterior_encoder): PosteriorEncoder(
-        (input_conv): Conv1d(80, 192, kernel_size=(1,), stride=(1,))
-        (encoder): WaveNet(
-          (conv_layers): ModuleList(
-            (0-15): 16 x ResidualBlock(
-              (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-              (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-              (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-            )
-          )
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (flow): ResidualAffineCouplingBlock(
-        (flows): ModuleList(
-          (0): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (1): FlipFlow()
-          (2): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (3): FlipFlow()
-          (4): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (5): FlipFlow()
-          (6): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (7): FlipFlow()
-        )
-      )
-      (duration_predictor): StochasticDurationPredictor(
-        (pre): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (log_flow): LogFlow()
-        (flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (post_pre): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-        (post_dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (post_proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (post_flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (global_conv): Conv1d(256, 192, kernel_size=(1,), stride=(1,))
-      )
-      (global_emb): Embedding(4, 256)
-    )
-    (discriminator): HiFiGANMultiScaleMultiPeriodDiscriminator(
-      (msd): HiFiGANMultiScaleDiscriminator(
-        (discriminators): ModuleList(
-          (0): HiFiGANScaleDiscriminator(
-            (layers): ModuleList(
-              (0): Sequential(
-                (0): Conv1d(1, 128, kernel_size=(15,), stride=(1,), padding=(7,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv1d(128, 128, kernel_size=(41,), stride=(2,), padding=(20,), groups=4)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv1d(128, 256, kernel_size=(41,), stride=(2,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv1d(256, 512, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv1d(512, 1024, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (5): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(41,), stride=(1,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (6): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(5,), stride=(1,), padding=(2,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (7): Conv1d(1024, 1, kernel_size=(3,), stride=(1,), padding=(1,))
-            )
-          )
-        )
-      )
-      (mpd): HiFiGANMultiPeriodDiscriminator(
-        (discriminators): ModuleList(
-          (0-4): 5 x HiFiGANPeriodDiscriminator(
-            (convs): ModuleList(
-              (0): Sequential(
-                (0): Conv2d(1, 32, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv2d(32, 128, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv2d(128, 512, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv2d(512, 1024, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv2d(1024, 1024, kernel_size=(5, 1), stride=(1, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-            )
-            (output_conv): Conv2d(1024, 1, kernel_size=(2, 1), stride=(1, 1), padding=(1, 0))
-          )
-        )
-      )
-    )
-    (generator_adv_loss): GeneratorAdversarialLoss()
-    (discriminator_adv_loss): DiscriminatorAdversarialLoss()
-    (feat_match_loss): FeatureMatchLoss()
-    (mel_loss): MelSpectrogramLoss(
-      (wav_to_mel): LogMelFbank(
-        (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-        (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=0, fmax=11025.0, htk=False)
-      )
-    )
-    (kl_loss): KLDivergenceLoss()
-  )
-)
-
-Model summary:
-    Class Name: ESPnetGANTTSModel
-    Total Number of model parameters: 96.24 M
-    Number of trainable parameters: 96.24 M (100.0%)
-    Size: 384.96 MB
-    Type: torch.float32
-[wieling-3-a100] 2023-12-01 15:58:41,789 (abs_task:1272) INFO: Optimizer:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,789 (abs_task:1273) INFO: Scheduler: <torch.optim.lr_scheduler.ExponentialLR object at 0x7ff08e5c38b0>
-[wieling-3-a100] 2023-12-01 15:58:41,790 (abs_task:1272) INFO: Optimizer2:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,790 (abs_task:1273) INFO: Scheduler2: <torch.optim.lr_scheduler.ExponentialLR object at 0x7ff08e5c3850>
-[wieling-3-a100] 2023-12-01 15:58:41,790 (abs_task:1282) INFO: Saving the configuration in exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/config.yaml
-[wieling-3-a100] 2023-12-01 15:58:41,807 (abs_task:1293) INFO: Namespace(config='conf/train_vits.yaml', print_config=False, log_level='INFO', drop_last_iter=False, dry_run=False, iterator_type='sequence', valid_iterator_type=None, output_dir='exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1', ngpu=0, seed=67823, num_workers=4, num_att_plot=3, dist_backend='nccl', dist_init_method='env://', dist_world_size=None, dist_rank=None, local_rank=None, dist_master_addr=None, dist_master_port=None, dist_launcher=None, multiprocessing_distributed=False, unused_parameters=True, sharded_ddp=False, cudnn_enabled=True, cudnn_benchmark=False, cudnn_deterministic=False, collect_stats=True, write_collected_feats=False, max_epoch=1000, patience=None, val_scheduler_criterion=('valid', 'loss'), early_stopping_criterion=('valid', 'loss', 'min'), best_model_criterion=[['train', 'total_count', 'max']], keep_nbest_models=10, nbest_averaging_interval=0, grad_clip=-1, grad_clip_type=2.0, grad_noise=False, accum_grad=1, no_forward_run=False, resume=False, train_dtype='float32', use_amp=False, log_interval=50, use_matplotlib=True, use_tensorboard=True, create_graph_in_tensorboard=False, use_wandb=True, wandb_project='GROTTS', wandb_id=None, wandb_entity=None, wandb_name='VITS_lr_3.0e-4', wandb_model_log_interval=-1, detect_anomaly=False, use_lora=False, save_lora_only=True, lora_conf={}, pretrain_path=None, init_param=['downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv'], ignore_init_mismatch=False, freeze_param=[], num_iters_per_epoch=1000, batch_size=40, valid_batch_size=None, batch_bins=10000000, valid_batch_bins=None, train_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.1.scp'], valid_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.1.scp'], batch_type='numel', valid_batch_type=None, fold_length=[], sort_in_batch='descending', shuffle_within_batch=False, sort_batch='descending', multiple_iterator=False, chunk_length=500, chunk_shift_ratio=0.5, num_cache_chunks=1024, chunk_excluded_key_prefixes=[], chunk_default_fs=None, train_data_path_and_name_and_type=[('dump/raw/train_nodev/text', 'text', 'text'), ('dump/raw/train_nodev/wav.scp', 'speech', 'sound'), ('dump/raw/train_nodev/utt2sid', 'sids', 'text_int')], valid_data_path_and_name_and_type=[('dump/raw/train_dev/text', 'text', 'text'), ('dump/raw/train_dev/wav.scp', 'speech', 'sound'), ('dump/raw/train_dev/utt2sid', 'sids', 'text_int')], allow_variable_data_keys=False, max_cache_size=0.0, max_cache_fd=32, allow_multi_rates=False, valid_max_cache_size=None, exclude_weight_decay=False, exclude_weight_decay_conf={}, optim='adamw', optim_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler='exponentiallr', scheduler_conf={'gamma': 0.999875}, optim2='adamw', optim2_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler2='exponentiallr', scheduler2_conf={'gamma': 0.999875}, generator_first=False, token_list=['<blank>', '<unk>', '<space>', 'e', 'n', 'a', 'o', 't', 'i', 'r', 'd', 's', 'k', 'l', 'm', 'u', 'g', 'h', 'w', 'v', '.', 'z', 'b', 'p', ',', 'j', 'c', 'f', '‘', '’', ':', '?', 'ö', "'", '!', '-', ';', 'ò', 'è', 'ì', 'é', 'y', 'ë', 'x', 'q', '<sos/eos>'], odim=None, model_conf={}, use_preprocessor=True, token_type='char', bpemodel=None, non_linguistic_symbols=None, cleaner=None, g2p=None, feats_extract='fbank', feats_extract_conf={'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'fs': 22050, 'fmin': 80, 'fmax': 7600, 'n_mels': 80}, normalize=None, normalize_conf={}, tts='vits', tts_conf={'generator_type': 'vits_generator', 'generator_params': {'hidden_channels': 192, 'spks': 4, 'global_channels': 256, 'segment_size': 32, 'text_encoder_attention_heads': 2, 'text_encoder_ffn_expand': 4, 'text_encoder_blocks': 6, 'text_encoder_positionwise_layer_type': 'conv1d', 'text_encoder_positionwise_conv_kernel_size': 3, 'text_encoder_positional_encoding_layer_type': 'rel_pos', 'text_encoder_self_attention_layer_type': 'rel_selfattn', 'text_encoder_activation_type': 'swish', 'text_encoder_normalize_before': True, 'text_encoder_dropout_rate': 0.1, 'text_encoder_positional_dropout_rate': 0.0, 'text_encoder_attention_dropout_rate': 0.1, 'use_macaron_style_in_text_encoder': True, 'use_conformer_conv_in_text_encoder': False, 'text_encoder_conformer_kernel_size': -1, 'decoder_kernel_size': 7, 'decoder_channels': 512, 'decoder_upsample_scales': [8, 8, 2, 2], 'decoder_upsample_kernel_sizes': [16, 16, 4, 4], 'decoder_resblock_kernel_sizes': [3, 7, 11], 'decoder_resblock_dilations': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'use_weight_norm_in_decoder': True, 'posterior_encoder_kernel_size': 5, 'posterior_encoder_layers': 16, 'posterior_encoder_stacks': 1, 'posterior_encoder_base_dilation': 1, 'posterior_encoder_dropout_rate': 0.0, 'use_weight_norm_in_posterior_encoder': True, 'flow_flows': 4, 'flow_kernel_size': 5, 'flow_base_dilation': 1, 'flow_layers': 4, 'flow_dropout_rate': 0.0, 'use_weight_norm_in_flow': True, 'use_only_mean_in_flow': True, 'stochastic_duration_predictor_kernel_size': 3, 'stochastic_duration_predictor_dropout_rate': 0.5, 'stochastic_duration_predictor_flows': 4, 'stochastic_duration_predictor_dds_conv_layers': 3, 'vocabs': 46, 'aux_channels': 80}, 'discriminator_type': 'hifigan_multi_scale_multi_period_discriminator', 'discriminator_params': {'scales': 1, 'scale_downsample_pooling': 'AvgPool1d', 'scale_downsample_pooling_params': {'kernel_size': 4, 'stride': 2, 'padding': 2}, 'scale_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [15, 41, 5, 3], 'channels': 128, 'max_downsample_channels': 1024, 'max_groups': 16, 'bias': True, 'downsample_scales': [2, 2, 4, 4, 1], 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': False, 'use_spectral_norm': False}, 'follow_official_norm': False, 'periods': [2, 3, 5, 7, 11], 'period_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [5, 3], 'channels': 32, 'downsample_scales': [3, 3, 3, 3, 1], 'max_downsample_channels': 1024, 'bias': True, 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': True, 'use_spectral_norm': False}}, 'generator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'discriminator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'feat_match_loss_params': {'average_by_discriminators': False, 'average_by_layers': False, 'include_final_outputs': True}, 'mel_loss_params': {'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'window': 'hann', 'n_mels': 80, 'fmin': 0, 'fmax': None, 'log_base': None}, 'lambda_adv': 1.0, 'lambda_mel': 45.0, 'lambda_feat_match': 2.0, 'lambda_dur': 1.0, 'lambda_kl': 1.0, 'sampling_rate': 22050, 'cache_generator_outputs': True}, pitch_extract=None, pitch_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'f0max': 400, 'f0min': 80}, pitch_normalize=None, pitch_normalize_conf={}, energy_extract=None, energy_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None}, energy_normalize=None, energy_normalize_conf={}, required=['output_dir', 'token_list'], version='202310', distributed=False)
-# Accounting: time=16 threads=1
-# Ended (code 0) at Fri Dec  1 15:58:50 UTC 2023, elapsed time 16 seconds

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/config.yaml

@@ -1,383 +0,0 @@
-config: conf/train_vits.yaml
-print_config: false
-log_level: INFO
-drop_last_iter: false
-dry_run: false
-iterator_type: sequence
-valid_iterator_type: null
-output_dir: exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1
-ngpu: 0
-seed: 67823
-num_workers: 4
-num_att_plot: 3
-dist_backend: nccl
-dist_init_method: env://
-dist_world_size: null
-dist_rank: null
-local_rank: null
-dist_master_addr: null
-dist_master_port: null
-dist_launcher: null
-multiprocessing_distributed: false
-unused_parameters: true
-sharded_ddp: false
-cudnn_enabled: true
-cudnn_benchmark: false
-cudnn_deterministic: false
-collect_stats: true
-write_collected_feats: false
-max_epoch: 1000
-patience: null
-val_scheduler_criterion:
-- valid
-- loss
-early_stopping_criterion:
-- valid
-- loss
-- min
-best_model_criterion:
--   - train
-    - total_count
-    - max
-keep_nbest_models: 10
-nbest_averaging_interval: 0
-grad_clip: -1
-grad_clip_type: 2.0
-grad_noise: false
-accum_grad: 1
-no_forward_run: false
-resume: false
-train_dtype: float32
-use_amp: false
-log_interval: 50
-use_matplotlib: true
-use_tensorboard: true
-create_graph_in_tensorboard: false
-use_wandb: true
-wandb_project: GROTTS
-wandb_id: null
-wandb_entity: null
-wandb_name: VITS_lr_3.0e-4
-wandb_model_log_interval: -1
-detect_anomaly: false
-use_lora: false
-save_lora_only: true
-lora_conf: {}
-pretrain_path: null
-init_param:
-- downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv
-ignore_init_mismatch: false
-freeze_param: []
-num_iters_per_epoch: 1000
-batch_size: 40
-valid_batch_size: null
-batch_bins: 10000000
-valid_batch_bins: null
-train_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.1.scp
-valid_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.1.scp
-batch_type: numel
-valid_batch_type: null
-fold_length: []
-sort_in_batch: descending
-shuffle_within_batch: false
-sort_batch: descending
-multiple_iterator: false
-chunk_length: 500
-chunk_shift_ratio: 0.5
-num_cache_chunks: 1024
-chunk_excluded_key_prefixes: []
-chunk_default_fs: null
-train_data_path_and_name_and_type:
--   - dump/raw/train_nodev/text
-    - text
-    - text
--   - dump/raw/train_nodev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_nodev/utt2sid
-    - sids
-    - text_int
-valid_data_path_and_name_and_type:
--   - dump/raw/train_dev/text
-    - text
-    - text
--   - dump/raw/train_dev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_dev/utt2sid
-    - sids
-    - text_int
-allow_variable_data_keys: false
-max_cache_size: 0.0
-max_cache_fd: 32
-allow_multi_rates: false
-valid_max_cache_size: null
-exclude_weight_decay: false
-exclude_weight_decay_conf: {}
-optim: adamw
-optim_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler: exponentiallr
-scheduler_conf:
-    gamma: 0.999875
-optim2: adamw
-optim2_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler2: exponentiallr
-scheduler2_conf:
-    gamma: 0.999875
-generator_first: false
-token_list:
-- <blank>
-- <unk>
-- <space>
-- e
-- n
-- a
-- o
-- t
-- i
-- r
-- d
-- s
-- k
-- l
-- m
-- u
-- g
-- h
-- w
-- v
-- .
-- z
-- b
-- p
-- ','
-- j
-- c
-- f
-- ‘
-- ’
-- ':'
-- '?'
-- ö
-- ''''
-- '!'
-- '-'
-- ;
-- ò
-- è
-- ì
-- é
-- y
-- ë
-- x
-- q
-- <sos/eos>
-odim: null
-model_conf: {}
-use_preprocessor: true
-token_type: char
-bpemodel: null
-non_linguistic_symbols: null
-cleaner: null
-g2p: null
-feats_extract: fbank
-feats_extract_conf:
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-    fs: 22050
-    fmin: 80
-    fmax: 7600
-    n_mels: 80
-normalize: null
-normalize_conf: {}
-tts: vits
-tts_conf:
-    generator_type: vits_generator
-    generator_params:
-        hidden_channels: 192
-        spks: 4
-        global_channels: 256
-        segment_size: 32
-        text_encoder_attention_heads: 2
-        text_encoder_ffn_expand: 4
-        text_encoder_blocks: 6
-        text_encoder_positionwise_layer_type: conv1d
-        text_encoder_positionwise_conv_kernel_size: 3
-        text_encoder_positional_encoding_layer_type: rel_pos
-        text_encoder_self_attention_layer_type: rel_selfattn
-        text_encoder_activation_type: swish
-        text_encoder_normalize_before: true
-        text_encoder_dropout_rate: 0.1
-        text_encoder_positional_dropout_rate: 0.0
-        text_encoder_attention_dropout_rate: 0.1
-        use_macaron_style_in_text_encoder: true
-        use_conformer_conv_in_text_encoder: false
-        text_encoder_conformer_kernel_size: -1
-        decoder_kernel_size: 7
-        decoder_channels: 512
-        decoder_upsample_scales:
-        - 8
-        - 8
-        - 2
-        - 2
-        decoder_upsample_kernel_sizes:
-        - 16
-        - 16
-        - 4
-        - 4
-        decoder_resblock_kernel_sizes:
-        - 3
-        - 7
-        - 11
-        decoder_resblock_dilations:
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        use_weight_norm_in_decoder: true
-        posterior_encoder_kernel_size: 5
-        posterior_encoder_layers: 16
-        posterior_encoder_stacks: 1
-        posterior_encoder_base_dilation: 1
-        posterior_encoder_dropout_rate: 0.0
-        use_weight_norm_in_posterior_encoder: true
-        flow_flows: 4
-        flow_kernel_size: 5
-        flow_base_dilation: 1
-        flow_layers: 4
-        flow_dropout_rate: 0.0
-        use_weight_norm_in_flow: true
-        use_only_mean_in_flow: true
-        stochastic_duration_predictor_kernel_size: 3
-        stochastic_duration_predictor_dropout_rate: 0.5
-        stochastic_duration_predictor_flows: 4
-        stochastic_duration_predictor_dds_conv_layers: 3
-        vocabs: 46
-        aux_channels: 80
-    discriminator_type: hifigan_multi_scale_multi_period_discriminator
-    discriminator_params:
-        scales: 1
-        scale_downsample_pooling: AvgPool1d
-        scale_downsample_pooling_params:
-            kernel_size: 4
-            stride: 2
-            padding: 2
-        scale_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 15
-            - 41
-            - 5
-            - 3
-            channels: 128
-            max_downsample_channels: 1024
-            max_groups: 16
-            bias: true
-            downsample_scales:
-            - 2
-            - 2
-            - 4
-            - 4
-            - 1
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: false
-            use_spectral_norm: false
-        follow_official_norm: false
-        periods:
-        - 2
-        - 3
-        - 5
-        - 7
-        - 11
-        period_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 5
-            - 3
-            channels: 32
-            downsample_scales:
-            - 3
-            - 3
-            - 3
-            - 3
-            - 1
-            max_downsample_channels: 1024
-            bias: true
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: true
-            use_spectral_norm: false
-    generator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    discriminator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    feat_match_loss_params:
-        average_by_discriminators: false
-        average_by_layers: false
-        include_final_outputs: true
-    mel_loss_params:
-        fs: 22050
-        n_fft: 1024
-        hop_length: 256
-        win_length: null
-        window: hann
-        n_mels: 80
-        fmin: 0
-        fmax: null
-        log_base: null
-    lambda_adv: 1.0
-    lambda_mel: 45.0
-    lambda_feat_match: 2.0
-    lambda_dur: 1.0
-    lambda_kl: 1.0
-    sampling_rate: 22050
-    cache_generator_outputs: true
-pitch_extract: null
-pitch_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    f0max: 400
-    f0min: 80
-pitch_normalize: null
-pitch_normalize_conf: {}
-energy_extract: null
-energy_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-energy_normalize: null
-energy_normalize_conf: {}
-required:
-- output_dir
-- token_list
-version: '202310'
-distributed: false

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:ddfdb5fa7d569f4cfd147a3b2c063b468bc622eef7ac57458ca7e4c3afded949
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4410bc31eace8e651b586b137f0b39fafa8920aaf121ca372393331ef4abeadf
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/sids_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-0002 1
-Hoogelaandsters-0006 1
-Hoogelaandsters-0010 1
-Hoogelaandsters-0014 1
-Hoogelaandsters-0018 1
-Hoogelaandsters-0022 1
-Hoogelaandsters-0026 1
-Hoogelaandsters-0030 1
-Hoogelaandsters-0034 1
-Hoogelaandsters-0038 1
-Hoogelaandsters-0042 1
-Hoogelaandsters-0046 1
-Hoogelaandsters-0050 1
-Hoogelaandsters-0054 1
-Hoogelaandsters-0058 1
-Hoogelaandsters-0062 1
-Hoogelaandsters-0066 1
-Hoogelaandsters-0071 1
-Hoogelaandsters-0075 1
-Hoogelaandsters-0079 1
-Hoogelaandsters-0083 1
-Hoogelaandsters-0087 1
-Hoogelaandsters-0091 1
-Hoogelaandsters-0095 1
-Hoogelaandsters-0099 1
-Hoogelaandsters-0103 1
-Hoogelaandsters-0107 1
-Hoogelaandsters-0111 1
-Hoogelaandsters-0115 1
-Hoogelaandsters-0119 1
-Hoogelaandsters-0123 1
-Hoogelaandsters-0127 1
-Hoogelaandsters-0131 1
-Hoogelaandsters-0136 1
-Hoogelaandsters-0140 1
-Hoogelaandsters-0144 1
-Hoogelaandsters-0148 1
-Hoogelaandsters-0152 1
-Hoogelaandsters-0156 1
-Hoogelaandsters-0160 1
-Hoogelaandsters-0003 1
-Hoogelaandsters-0007 1
-Hoogelaandsters-0011 1
-Hoogelaandsters-0015 1
-Hoogelaandsters-0019 1
-Hoogelaandsters-0023 1
-Hoogelaandsters-0027 1
-Hoogelaandsters-0031 1
-Hoogelaandsters-0035 1
-Hoogelaandsters-0039 1
-Hoogelaandsters-0043 1
-Hoogelaandsters-0047 1
-Hoogelaandsters-0051 1
-Hoogelaandsters-0055 1
-Hoogelaandsters-0059 1
-Hoogelaandsters-0063 1
-Hoogelaandsters-0067 1
-Hoogelaandsters-0072 1
-Hoogelaandsters-0076 1
-Hoogelaandsters-0080 1
-Hoogelaandsters-0084 1
-Hoogelaandsters-0088 1
-Hoogelaandsters-0092 1
-Hoogelaandsters-0096 1
-Hoogelaandsters-0100 1
-Hoogelaandsters-0104 1
-Hoogelaandsters-0108 1
-Hoogelaandsters-0112 1
-Hoogelaandsters-0116 1
-Hoogelaandsters-0120 1
-Hoogelaandsters-0124 1
-Hoogelaandsters-0128 1
-Hoogelaandsters-0132 1
-Hoogelaandsters-0137 1
-Hoogelaandsters-0141 1
-Hoogelaandsters-0145 1
-Hoogelaandsters-0149 1
-Hoogelaandsters-0153 1
-Hoogelaandsters-0157 1
-Hoogelaandsters-0161 1
-Hoogelaandsters-0004 1
-Hoogelaandsters-0008 1
-Hoogelaandsters-0012 1
-Hoogelaandsters-0016 1
-Hoogelaandsters-0020 1
-Hoogelaandsters-0024 1
-Hoogelaandsters-0028 1
-Hoogelaandsters-0032 1
-Hoogelaandsters-0036 1
-Hoogelaandsters-0040 1
-Hoogelaandsters-0044 1
-Hoogelaandsters-0048 1
-Hoogelaandsters-0052 1
-Hoogelaandsters-0056 1
-Hoogelaandsters-0060 1
-Hoogelaandsters-0064 1
-Hoogelaandsters-0069 1
-Hoogelaandsters-0073 1
-Hoogelaandsters-0077 1
-Hoogelaandsters-0081 1
-Hoogelaandsters-0085 1
-Hoogelaandsters-0089 1
-Hoogelaandsters-0093 1
-Hoogelaandsters-0097 1
-Hoogelaandsters-0101 1
-Hoogelaandsters-0105 1
-Hoogelaandsters-0109 1
-Hoogelaandsters-0113 1
-Hoogelaandsters-0117 1
-Hoogelaandsters-0121 1
-Hoogelaandsters-0125 1
-Hoogelaandsters-0129 1
-Hoogelaandsters-0133 1
-Hoogelaandsters-0138 1
-Hoogelaandsters-0142 1
-Hoogelaandsters-0146 1
-Hoogelaandsters-0150 1
-Hoogelaandsters-0154 1
-Hoogelaandsters-0158 1
-Hoogelaandsters-0162 1
-Hoogelaandsters-0005 1
-Hoogelaandsters-0009 1
-Hoogelaandsters-0013 1
-Hoogelaandsters-0017 1
-Hoogelaandsters-0021 1
-Hoogelaandsters-0025 1
-Hoogelaandsters-0029 1
-Hoogelaandsters-0033 1
-Hoogelaandsters-0037 1
-Hoogelaandsters-0041 1
-Hoogelaandsters-0045 1
-Hoogelaandsters-0049 1
-Hoogelaandsters-0053 1
-Hoogelaandsters-0057 1
-Hoogelaandsters-0061 1
-Hoogelaandsters-0065 1
-Hoogelaandsters-0070 1
-Hoogelaandsters-0074 1
-Hoogelaandsters-0078 1
-Hoogelaandsters-0082 1
-Hoogelaandsters-0086 1
-Hoogelaandsters-0090 1
-Hoogelaandsters-0094 1
-Hoogelaandsters-0098 1
-Hoogelaandsters-0102 1
-Hoogelaandsters-0106 1
-Hoogelaandsters-0110 1
-Hoogelaandsters-0114 1
-Hoogelaandsters-0118 1
-Hoogelaandsters-0122 1
-Hoogelaandsters-0126 1
-Hoogelaandsters-0130 1
-Hoogelaandsters-0134 1
-Hoogelaandsters-0139 1
-Hoogelaandsters-0143 1
-Hoogelaandsters-0147 1
-Hoogelaandsters-0151 1
-Hoogelaandsters-0155 1
-Hoogelaandsters-0159 1
-Hoogelaandsters-0163 1
-Hoogelaandsters-0164 1
-Hoogelaandsters-0168 1
-Hoogelaandsters-0172 1
-Hoogelaandsters-0176 1
-Hoogelaandsters-0180 1
-Hoogelaandsters-0184 1
-Hoogelaandsters-0188 1
-Hoogelaandsters-0192 1
-Hoogelaandsters-0196 1
-Hoogelaandsters-0200 1
-Hoogelaandsters-0205 1
-Hoogelaandsters-0209 1
-Hoogelaandsters-0213 1
-Hoogelaandsters-0217 1
-Hoogelaandsters-0221 1
-Hoogelaandsters-0225 1
-Hoogelaandsters-0229 1
-Hoogelaandsters-0233 1
-Hoogelaandsters-0237 1
-Hoogelaandsters-0241 1
-Hoogelaandsters-0245 1
-Hoogelaandsters-0249 1
-Hoogelaandsters-0253 1
-Hoogelaandsters-0165 1
-Hoogelaandsters-0169 1
-Hoogelaandsters-0173 1
-Hoogelaandsters-0177 1
-Hoogelaandsters-0181 1
-Hoogelaandsters-0185 1
-Hoogelaandsters-0189 1
-Hoogelaandsters-0193 1
-Hoogelaandsters-0197 1
-Hoogelaandsters-0201 1
-Hoogelaandsters-0206 1
-Hoogelaandsters-0210 1
-Hoogelaandsters-0214 1
-Hoogelaandsters-0218 1
-Hoogelaandsters-0222 1
-Hoogelaandsters-0226 1
-Hoogelaandsters-0230 1
-Hoogelaandsters-0234 1
-Hoogelaandsters-0238 1
-Hoogelaandsters-0242 1
-Hoogelaandsters-0246 1
-Hoogelaandsters-0250 1
-Hoogelaandsters-0166 1
-Hoogelaandsters-0170 1
-Hoogelaandsters-0174 1
-Hoogelaandsters-0178 1
-Hoogelaandsters-0182 1
-Hoogelaandsters-0186 1
-Hoogelaandsters-0190 1
-Hoogelaandsters-0194 1
-Hoogelaandsters-0198 1
-Hoogelaandsters-0203 1
-Hoogelaandsters-0207 1
-Hoogelaandsters-0211 1
-Hoogelaandsters-0215 1
-Hoogelaandsters-0219 1
-Hoogelaandsters-0223 1
-Hoogelaandsters-0227 1
-Hoogelaandsters-0231 1
-Hoogelaandsters-0235 1
-Hoogelaandsters-0239 1
-Hoogelaandsters-0243 1
-Hoogelaandsters-0247 1
-Hoogelaandsters-0251 1
-Hoogelaandsters-0167 1
-Hoogelaandsters-0171 1
-Hoogelaandsters-0175 1
-Hoogelaandsters-0179 1
-Hoogelaandsters-0183 1
-Hoogelaandsters-0187 1
-Hoogelaandsters-0191 1
-Hoogelaandsters-0195 1
-Hoogelaandsters-0199 1
-Hoogelaandsters-0204 1
-Hoogelaandsters-0208 1
-Hoogelaandsters-0212 1
-Hoogelaandsters-0216 1
-Hoogelaandsters-0220 1
-Hoogelaandsters-0224 1
-Hoogelaandsters-0228 1
-Hoogelaandsters-0232 1
-Hoogelaandsters-0236 1
-Hoogelaandsters-0240 1
-Hoogelaandsters-0244 1
-Hoogelaandsters-0248 1
-Hoogelaandsters-0252 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/speech_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-0002 148618
-Hoogelaandsters-0006 71366
-Hoogelaandsters-0010 40465
-Hoogelaandsters-0014 141996
-Hoogelaandsters-0018 123603
-Hoogelaandsters-0022 80930
-Hoogelaandsters-0026 71366
-Hoogelaandsters-0030 82402
-Hoogelaandsters-0034 85345
-Hoogelaandsters-0038 76516
-Hoogelaandsters-0042 179519
-Hoogelaandsters-0046 375959
-Hoogelaandsters-0050 44880
-Hoogelaandsters-0054 61802
-Hoogelaandsters-0058 75780
-Hoogelaandsters-0062 153768
-Hoogelaandsters-0066 83138
-Hoogelaandsters-0071 129489
-Hoogelaandsters-0075 130960
-Hoogelaandsters-0079 153768
-Hoogelaandsters-0083 135375
-Hoogelaandsters-0087 83874
-Hoogelaandsters-0091 87552
-Hoogelaandsters-0095 80930
-Hoogelaandsters-0099 90495
-Hoogelaandsters-0103 133168
-Hoogelaandsters-0107 38258
-Hoogelaandsters-0111 48558
-Hoogelaandsters-0115 100060
-Hoogelaandsters-0119 49294
-Hoogelaandsters-0123 66216
-Hoogelaandsters-0127 186141
-Hoogelaandsters-0131 37522
-Hoogelaandsters-0136 139054
-Hoogelaandsters-0140 40465
-Hoogelaandsters-0144 64744
-Hoogelaandsters-0148 41936
-Hoogelaandsters-0152 99324
-Hoogelaandsters-0156 70630
-Hoogelaandsters-0160 92702
-Hoogelaandsters-0003 124339
-Hoogelaandsters-0007 58858
-Hoogelaandsters-0011 187612
-Hoogelaandsters-0015 52972
-Hoogelaandsters-0019 56652
-Hoogelaandsters-0023 36050
-Hoogelaandsters-0027 90495
-Hoogelaandsters-0031 25752
-Hoogelaandsters-0035 91966
-Hoogelaandsters-0039 144204
-Hoogelaandsters-0043 75780
-Hoogelaandsters-0047 168483
-Hoogelaandsters-0051 153033
-Hoogelaandsters-0055 57387
-Hoogelaandsters-0059 104474
-Hoogelaandsters-0063 113303
-Hoogelaandsters-0067 47822
-Hoogelaandsters-0072 122868
-Hoogelaandsters-0076 54444
-Hoogelaandsters-0080 75044
-Hoogelaandsters-0084 103002
-Hoogelaandsters-0088 224399
-Hoogelaandsters-0092 85345
-Hoogelaandsters-0096 86816
-Hoogelaandsters-0100 122868
-Hoogelaandsters-0104 82402
-Hoogelaandsters-0108 82402
-Hoogelaandsters-0112 77252
-Hoogelaandsters-0116 139790
-Hoogelaandsters-0120 111096
-Hoogelaandsters-0124 63272
-Hoogelaandsters-0128 86816
-Hoogelaandsters-0132 42672
-Hoogelaandsters-0137 101531
-Hoogelaandsters-0141 128754
-Hoogelaandsters-0145 96381
-Hoogelaandsters-0149 151561
-Hoogelaandsters-0153 100060
-Hoogelaandsters-0157 114774
-Hoogelaandsters-0161 106682
-Hoogelaandsters-0004 96381
-Hoogelaandsters-0008 49294
-Hoogelaandsters-0012 75780
-Hoogelaandsters-0016 104474
-Hoogelaandsters-0020 55916
-Hoogelaandsters-0024 77988
-Hoogelaandsters-0028 128018
-Hoogelaandsters-0032 33844
-Hoogelaandsters-0036 94910
-Hoogelaandsters-0040 58858
-Hoogelaandsters-0044 72102
-Hoogelaandsters-0048 134639
-Hoogelaandsters-0052 95646
-Hoogelaandsters-0056 37522
-Hoogelaandsters-0060 55180
-Hoogelaandsters-0064 57387
-Hoogelaandsters-0069 126546
-Hoogelaandsters-0073 210420
-Hoogelaandsters-0077 86816
-Hoogelaandsters-0081 65480
-Hoogelaandsters-0085 122868
-Hoogelaandsters-0089 84610
-Hoogelaandsters-0093 148618
-Hoogelaandsters-0097 75044
-Hoogelaandsters-0101 44144
-Hoogelaandsters-0105 43408
-Hoogelaandsters-0109 267071
-Hoogelaandsters-0113 141996
-Hoogelaandsters-0117 44144
-Hoogelaandsters-0121 94910
-Hoogelaandsters-0125 63272
-Hoogelaandsters-0129 38994
-Hoogelaandsters-0133 61066
-Hoogelaandsters-0138 58122
-Hoogelaandsters-0142 115510
-Hoogelaandsters-0146 63272
-Hoogelaandsters-0150 52972
-Hoogelaandsters-0154 130225
-Hoogelaandsters-0158 119189
-Hoogelaandsters-0162 30164
-Hoogelaandsters-0005 44144
-Hoogelaandsters-0009 189819
-Hoogelaandsters-0013 122868
-Hoogelaandsters-0017 72838
-Hoogelaandsters-0021 86080
-Hoogelaandsters-0025 217042
-Hoogelaandsters-0029 63272
-Hoogelaandsters-0033 114774
-Hoogelaandsters-0037 138318
-Hoogelaandsters-0041 148618
-Hoogelaandsters-0045 126546
-Hoogelaandsters-0049 57387
-Hoogelaandsters-0053 66952
-Hoogelaandsters-0057 33844
-Hoogelaandsters-0061 47822
-Hoogelaandsters-0065 66216
-Hoogelaandsters-0070 186877
-Hoogelaandsters-0074 66952
-Hoogelaandsters-0078 83874
-Hoogelaandsters-0082 90495
-Hoogelaandsters-0086 167012
-Hoogelaandsters-0090 43408
-Hoogelaandsters-0094 44880
-Hoogelaandsters-0098 105946
-Hoogelaandsters-0102 143468
-Hoogelaandsters-0106 111096
-Hoogelaandsters-0110 97852
-Hoogelaandsters-0114 108152
-Hoogelaandsters-0118 100796
-Hoogelaandsters-0122 77252
-Hoogelaandsters-0126 62537
-Hoogelaandsters-0130 100796
-Hoogelaandsters-0134 67688
-Hoogelaandsters-0139 111096
-Hoogelaandsters-0143 97116
-Hoogelaandsters-0147 54444
-Hoogelaandsters-0151 58858
-Hoogelaandsters-0155 48558
-Hoogelaandsters-0159 102267
-Hoogelaandsters-0163 48558
-Hoogelaandsters-0164 41936
-Hoogelaandsters-0168 121396
-Hoogelaandsters-0172 87552
-Hoogelaandsters-0176 57387
-Hoogelaandsters-0180 102267
-Hoogelaandsters-0184 46350
-Hoogelaandsters-0188 75044
-Hoogelaandsters-0192 38994
-Hoogelaandsters-0196 77988
-Hoogelaandsters-0200 114038
-Hoogelaandsters-0205 51501
-Hoogelaandsters-0209 95646
-Hoogelaandsters-0213 176576
-Hoogelaandsters-0217 44880
-Hoogelaandsters-0221 64744
-Hoogelaandsters-0225 94174
-Hoogelaandsters-0229 102267
-Hoogelaandsters-0233 209684
-Hoogelaandsters-0237 186877
-Hoogelaandsters-0241 156711
-Hoogelaandsters-0245 252357
-Hoogelaandsters-0249 59594
-Hoogelaandsters-0253 60330
-Hoogelaandsters-0165 86080
-Hoogelaandsters-0169 62537
-Hoogelaandsters-0173 64008
-Hoogelaandsters-0177 36788
-Hoogelaandsters-0181 70630
-Hoogelaandsters-0185 85345
-Hoogelaandsters-0189 42672
-Hoogelaandsters-0193 58122
-Hoogelaandsters-0197 91966
-Hoogelaandsters-0201 116982
-Hoogelaandsters-0206 28693
-Hoogelaandsters-0210 86816
-Hoogelaandsters-0214 119189
-Hoogelaandsters-0218 41936
-Hoogelaandsters-0222 88288
-Hoogelaandsters-0226 112567
-Hoogelaandsters-0230 75044
-Hoogelaandsters-0234 106682
-Hoogelaandsters-0238 111832
-Hoogelaandsters-0242 111096
-Hoogelaandsters-0246 94910
-Hoogelaandsters-0250 75780
-Hoogelaandsters-0166 86080
-Hoogelaandsters-0170 38994
-Hoogelaandsters-0174 33844
-Hoogelaandsters-0178 42672
-Hoogelaandsters-0182 33844
-Hoogelaandsters-0186 29428
-Hoogelaandsters-0190 69158
-Hoogelaandsters-0194 36050
-Hoogelaandsters-0198 180991
-Hoogelaandsters-0203 68423
-Hoogelaandsters-0207 83874
-Hoogelaandsters-0211 147147
-Hoogelaandsters-0215 54444
-Hoogelaandsters-0219 243528
-Hoogelaandsters-0223 105946
-Hoogelaandsters-0227 64744
-Hoogelaandsters-0231 142732
-Hoogelaandsters-0235 97852
-Hoogelaandsters-0239 73574
-Hoogelaandsters-0243 160390
-Hoogelaandsters-0247 63272
-Hoogelaandsters-0251 55916
-Hoogelaandsters-0167 83138
-Hoogelaandsters-0171 75044
-Hoogelaandsters-0175 147882
-Hoogelaandsters-0179 105946
-Hoogelaandsters-0183 40465
-Hoogelaandsters-0187 54444
-Hoogelaandsters-0191 158183
-Hoogelaandsters-0195 78724
-Hoogelaandsters-0199 64008
-Hoogelaandsters-0204 143468
-Hoogelaandsters-0208 47822
-Hoogelaandsters-0212 45615
-Hoogelaandsters-0216 150826
-Hoogelaandsters-0220 125810
-Hoogelaandsters-0224 114774
-Hoogelaandsters-0228 107417
-Hoogelaandsters-0232 153768
-Hoogelaandsters-0236 139054
-Hoogelaandsters-0240 149354
-Hoogelaandsters-0244 83874
-Hoogelaandsters-0248 69158
-Hoogelaandsters-0252 112567

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/train/text_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-0002 64
-Hoogelaandsters-0006 54
-Hoogelaandsters-0010 29
-Hoogelaandsters-0014 97
-Hoogelaandsters-0018 94
-Hoogelaandsters-0022 62
-Hoogelaandsters-0026 54
-Hoogelaandsters-0030 51
-Hoogelaandsters-0034 52
-Hoogelaandsters-0038 50
-Hoogelaandsters-0042 114
-Hoogelaandsters-0046 241
-Hoogelaandsters-0050 32
-Hoogelaandsters-0054 42
-Hoogelaandsters-0058 48
-Hoogelaandsters-0062 118
-Hoogelaandsters-0066 72
-Hoogelaandsters-0071 95
-Hoogelaandsters-0075 108
-Hoogelaandsters-0079 110
-Hoogelaandsters-0083 110
-Hoogelaandsters-0087 67
-Hoogelaandsters-0091 60
-Hoogelaandsters-0095 70
-Hoogelaandsters-0099 61
-Hoogelaandsters-0103 85
-Hoogelaandsters-0107 30
-Hoogelaandsters-0111 26
-Hoogelaandsters-0115 61
-Hoogelaandsters-0119 44
-Hoogelaandsters-0123 54
-Hoogelaandsters-0127 138
-Hoogelaandsters-0131 31
-Hoogelaandsters-0136 91
-Hoogelaandsters-0140 28
-Hoogelaandsters-0144 47
-Hoogelaandsters-0148 26
-Hoogelaandsters-0152 84
-Hoogelaandsters-0156 46
-Hoogelaandsters-0160 66
-Hoogelaandsters-0003 96
-Hoogelaandsters-0007 38
-Hoogelaandsters-0011 113
-Hoogelaandsters-0015 50
-Hoogelaandsters-0019 43
-Hoogelaandsters-0023 30
-Hoogelaandsters-0027 69
-Hoogelaandsters-0031 14
-Hoogelaandsters-0035 60
-Hoogelaandsters-0039 98
-Hoogelaandsters-0043 52
-Hoogelaandsters-0047 113
-Hoogelaandsters-0051 92
-Hoogelaandsters-0055 37
-Hoogelaandsters-0059 82
-Hoogelaandsters-0063 78
-Hoogelaandsters-0067 30
-Hoogelaandsters-0072 77
-Hoogelaandsters-0076 50
-Hoogelaandsters-0080 49
-Hoogelaandsters-0084 86
-Hoogelaandsters-0088 162
-Hoogelaandsters-0092 60
-Hoogelaandsters-0096 65
-Hoogelaandsters-0100 83
-Hoogelaandsters-0104 55
-Hoogelaandsters-0108 66
-Hoogelaandsters-0112 42
-Hoogelaandsters-0116 97
-Hoogelaandsters-0120 79
-Hoogelaandsters-0124 50
-Hoogelaandsters-0128 49
-Hoogelaandsters-0132 26
-Hoogelaandsters-0137 68
-Hoogelaandsters-0141 89
-Hoogelaandsters-0145 70
-Hoogelaandsters-0149 92
-Hoogelaandsters-0153 60
-Hoogelaandsters-0157 74
-Hoogelaandsters-0161 81
-Hoogelaandsters-0004 69
-Hoogelaandsters-0008 38
-Hoogelaandsters-0012 55
-Hoogelaandsters-0016 70
-Hoogelaandsters-0020 28
-Hoogelaandsters-0024 64
-Hoogelaandsters-0028 87
-Hoogelaandsters-0032 18
-Hoogelaandsters-0036 72
-Hoogelaandsters-0040 30
-Hoogelaandsters-0044 44
-Hoogelaandsters-0048 99
-Hoogelaandsters-0052 75
-Hoogelaandsters-0056 18
-Hoogelaandsters-0060 38
-Hoogelaandsters-0064 41
-Hoogelaandsters-0069 86
-Hoogelaandsters-0073 149
-Hoogelaandsters-0077 62
-Hoogelaandsters-0081 49
-Hoogelaandsters-0085 92
-Hoogelaandsters-0089 63
-Hoogelaandsters-0093 104
-Hoogelaandsters-0097 50
-Hoogelaandsters-0101 36
-Hoogelaandsters-0105 33
-Hoogelaandsters-0109 166
-Hoogelaandsters-0113 70
-Hoogelaandsters-0117 31
-Hoogelaandsters-0121 62
-Hoogelaandsters-0125 41
-Hoogelaandsters-0129 26
-Hoogelaandsters-0133 41
-Hoogelaandsters-0138 35
-Hoogelaandsters-0142 86
-Hoogelaandsters-0146 50
-Hoogelaandsters-0150 42
-Hoogelaandsters-0154 103
-Hoogelaandsters-0158 74
-Hoogelaandsters-0162 25
-Hoogelaandsters-0005 37
-Hoogelaandsters-0009 121
-Hoogelaandsters-0013 78
-Hoogelaandsters-0017 56
-Hoogelaandsters-0021 67
-Hoogelaandsters-0025 127
-Hoogelaandsters-0029 51
-Hoogelaandsters-0033 82
-Hoogelaandsters-0037 100
-Hoogelaandsters-0041 87
-Hoogelaandsters-0045 87
-Hoogelaandsters-0049 44
-Hoogelaandsters-0053 42
-Hoogelaandsters-0057 30
-Hoogelaandsters-0061 36
-Hoogelaandsters-0065 51
-Hoogelaandsters-0070 130
-Hoogelaandsters-0074 51
-Hoogelaandsters-0078 62
-Hoogelaandsters-0082 74
-Hoogelaandsters-0086 119
-Hoogelaandsters-0090 33
-Hoogelaandsters-0094 32
-Hoogelaandsters-0098 82
-Hoogelaandsters-0102 83
-Hoogelaandsters-0106 76
-Hoogelaandsters-0110 54
-Hoogelaandsters-0114 63
-Hoogelaandsters-0118 61
-Hoogelaandsters-0122 47
-Hoogelaandsters-0126 42
-Hoogelaandsters-0130 71
-Hoogelaandsters-0134 45
-Hoogelaandsters-0139 94
-Hoogelaandsters-0143 67
-Hoogelaandsters-0147 30
-Hoogelaandsters-0151 38
-Hoogelaandsters-0155 36
-Hoogelaandsters-0159 79
-Hoogelaandsters-0163 39
-Hoogelaandsters-0164 32
-Hoogelaandsters-0168 86
-Hoogelaandsters-0172 59
-Hoogelaandsters-0176 38
-Hoogelaandsters-0180 59
-Hoogelaandsters-0184 30
-Hoogelaandsters-0188 54
-Hoogelaandsters-0192 38
-Hoogelaandsters-0196 58
-Hoogelaandsters-0200 73
-Hoogelaandsters-0205 42
-Hoogelaandsters-0209 63
-Hoogelaandsters-0213 119
-Hoogelaandsters-0217 35
-Hoogelaandsters-0221 36
-Hoogelaandsters-0225 65
-Hoogelaandsters-0229 75
-Hoogelaandsters-0233 126
-Hoogelaandsters-0237 126
-Hoogelaandsters-0241 107
-Hoogelaandsters-0245 168
-Hoogelaandsters-0249 45
-Hoogelaandsters-0253 41
-Hoogelaandsters-0165 68
-Hoogelaandsters-0169 35
-Hoogelaandsters-0173 44
-Hoogelaandsters-0177 28
-Hoogelaandsters-0181 39
-Hoogelaandsters-0185 58
-Hoogelaandsters-0189 27
-Hoogelaandsters-0193 35
-Hoogelaandsters-0197 61
-Hoogelaandsters-0201 76
-Hoogelaandsters-0206 23
-Hoogelaandsters-0210 62
-Hoogelaandsters-0214 89
-Hoogelaandsters-0218 26
-Hoogelaandsters-0222 63
-Hoogelaandsters-0226 76
-Hoogelaandsters-0230 59
-Hoogelaandsters-0234 72
-Hoogelaandsters-0238 73
-Hoogelaandsters-0242 79
-Hoogelaandsters-0246 54
-Hoogelaandsters-0250 63
-Hoogelaandsters-0166 64
-Hoogelaandsters-0170 37
-Hoogelaandsters-0174 29
-Hoogelaandsters-0178 30
-Hoogelaandsters-0182 37
-Hoogelaandsters-0186 29
-Hoogelaandsters-0190 51
-Hoogelaandsters-0194 30
-Hoogelaandsters-0198 119
-Hoogelaandsters-0203 49
-Hoogelaandsters-0207 57
-Hoogelaandsters-0211 101
-Hoogelaandsters-0215 42
-Hoogelaandsters-0219 166
-Hoogelaandsters-0223 71
-Hoogelaandsters-0227 40
-Hoogelaandsters-0231 92
-Hoogelaandsters-0235 66
-Hoogelaandsters-0239 54
-Hoogelaandsters-0243 94
-Hoogelaandsters-0247 41
-Hoogelaandsters-0251 43
-Hoogelaandsters-0167 66
-Hoogelaandsters-0171 54
-Hoogelaandsters-0175 113
-Hoogelaandsters-0179 78
-Hoogelaandsters-0183 39
-Hoogelaandsters-0187 44
-Hoogelaandsters-0191 93
-Hoogelaandsters-0195 58
-Hoogelaandsters-0199 51
-Hoogelaandsters-0204 111
-Hoogelaandsters-0208 35
-Hoogelaandsters-0212 27
-Hoogelaandsters-0216 110
-Hoogelaandsters-0220 73
-Hoogelaandsters-0224 47
-Hoogelaandsters-0228 66
-Hoogelaandsters-0232 101
-Hoogelaandsters-0236 94
-Hoogelaandsters-0240 104
-Hoogelaandsters-0244 50
-Hoogelaandsters-0248 44
-Hoogelaandsters-0252 85

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:92b775aaa5d2948c527ad29e192999a8388c2e6fd81fad2475ee7da577f3594a
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:4abdc1055a5330f17559bc4667c89f68c0ab191e3debc39c5735481cdf9d19ef
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/sids_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-0001 1
-Hoogelaandsters-0269 1
-Hoogelaandsters-0068 1
-Hoogelaandsters-0135 1
-Hoogelaandsters-0202 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/speech_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-0001 195705
-Hoogelaandsters-0269 119189
-Hoogelaandsters-0068 136111
-Hoogelaandsters-0135 87552
-Hoogelaandsters-0202 109624

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.1/valid/text_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-0001 111
-Hoogelaandsters-0269 86
-Hoogelaandsters-0068 91
-Hoogelaandsters-0135 62
-Hoogelaandsters-0202 75

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10.log

@@ -1,1152 +0,0 @@
-# python3 -m espnet2.bin.gan_tts_train --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.10.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.10.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000 
-# Started at Fri Dec  1 15:58:34 UTC 2023
-#
-/data2/p280965/tts/espnet/tools/venv/bin/python3 /data2/p280965/tts/espnet/espnet2/bin/gan_tts_train.py --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.10.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.10.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000
-[wieling-3-a100] 2023-12-01 15:58:40,202 (gan_tts:293) INFO: Vocabulary size: 46
-[wieling-3-a100] 2023-12-01 15:58:40,315 (encoder:174) INFO: encoder self-attention layer type = relative self-attention
-/data2/p280965/tts/espnet/tools/venv/lib/python3.9/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
-  warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
-/data2/p280965/tts/espnet/espnet2/gan_tts/vits/monotonic_align/__init__.py:19: UserWarning: Cython version is not available. Fallback to 'EXPERIMETAL' numba version. If you want to use the cython version, please build it as follows: `cd espnet2/gan_tts/vits/monotonic_align; python setup.py build_ext --inplace`
-  warnings.warn(
-[wieling-3-a100] 2023-12-01 15:58:41,727 (abs_task:1268) INFO: pytorch.version=2.1.0+cu121, cuda.available=True, cudnn.version=8902, cudnn.benchmark=False, cudnn.deterministic=False
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1269) INFO: Model structure:
-ESPnetGANTTSModel(
-  (feats_extract): LogMelFbank(
-    (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-    (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=80, fmax=7600, htk=False)
-  )
-  (tts): VITS(
-    (generator): VITSGenerator(
-      (text_encoder): TextEncoder(
-        (emb): Embedding(46, 192)
-        (encoder): Encoder(
-          (embed): Sequential(
-            (0): RelPositionalEncoding(
-              (dropout): Dropout(p=0.0, inplace=False)
-            )
-          )
-          (encoders): MultiSequential(
-            (0): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (1): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (2): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (3): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (4): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (5): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-          )
-          (after_norm): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (decoder): HiFiGANGenerator(
-        (input_conv): Conv1d(192, 512, kernel_size=(7,), stride=(1,), padding=(3,))
-        (upsamples): ModuleList(
-          (0): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(512, 256, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (1): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(256, 128, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (2): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(128, 64, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-          (3): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(64, 32, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-        )
-        (blocks): ModuleList(
-          (0): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (1): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (2): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (3): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (4): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (5): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (6): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (7): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (8): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (9): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (10): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (11): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-        )
-        (output_conv): Sequential(
-          (0): LeakyReLU(negative_slope=0.01)
-          (1): Conv1d(32, 1, kernel_size=(7,), stride=(1,), padding=(3,))
-          (2): Tanh()
-        )
-        (global_conv): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
-      )
-      (posterior_encoder): PosteriorEncoder(
-        (input_conv): Conv1d(80, 192, kernel_size=(1,), stride=(1,))
-        (encoder): WaveNet(
-          (conv_layers): ModuleList(
-            (0-15): 16 x ResidualBlock(
-              (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-              (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-              (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-            )
-          )
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (flow): ResidualAffineCouplingBlock(
-        (flows): ModuleList(
-          (0): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (1): FlipFlow()
-          (2): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (3): FlipFlow()
-          (4): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (5): FlipFlow()
-          (6): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (7): FlipFlow()
-        )
-      )
-      (duration_predictor): StochasticDurationPredictor(
-        (pre): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (log_flow): LogFlow()
-        (flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (post_pre): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-        (post_dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (post_proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (post_flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (global_conv): Conv1d(256, 192, kernel_size=(1,), stride=(1,))
-      )
-      (global_emb): Embedding(4, 256)
-    )
-    (discriminator): HiFiGANMultiScaleMultiPeriodDiscriminator(
-      (msd): HiFiGANMultiScaleDiscriminator(
-        (discriminators): ModuleList(
-          (0): HiFiGANScaleDiscriminator(
-            (layers): ModuleList(
-              (0): Sequential(
-                (0): Conv1d(1, 128, kernel_size=(15,), stride=(1,), padding=(7,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv1d(128, 128, kernel_size=(41,), stride=(2,), padding=(20,), groups=4)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv1d(128, 256, kernel_size=(41,), stride=(2,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv1d(256, 512, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv1d(512, 1024, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (5): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(41,), stride=(1,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (6): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(5,), stride=(1,), padding=(2,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (7): Conv1d(1024, 1, kernel_size=(3,), stride=(1,), padding=(1,))
-            )
-          )
-        )
-      )
-      (mpd): HiFiGANMultiPeriodDiscriminator(
-        (discriminators): ModuleList(
-          (0-4): 5 x HiFiGANPeriodDiscriminator(
-            (convs): ModuleList(
-              (0): Sequential(
-                (0): Conv2d(1, 32, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv2d(32, 128, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv2d(128, 512, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv2d(512, 1024, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv2d(1024, 1024, kernel_size=(5, 1), stride=(1, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-            )
-            (output_conv): Conv2d(1024, 1, kernel_size=(2, 1), stride=(1, 1), padding=(1, 0))
-          )
-        )
-      )
-    )
-    (generator_adv_loss): GeneratorAdversarialLoss()
-    (discriminator_adv_loss): DiscriminatorAdversarialLoss()
-    (feat_match_loss): FeatureMatchLoss()
-    (mel_loss): MelSpectrogramLoss(
-      (wav_to_mel): LogMelFbank(
-        (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-        (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=0, fmax=11025.0, htk=False)
-      )
-    )
-    (kl_loss): KLDivergenceLoss()
-  )
-)
-
-Model summary:
-    Class Name: ESPnetGANTTSModel
-    Total Number of model parameters: 96.24 M
-    Number of trainable parameters: 96.24 M (100.0%)
-    Size: 384.96 MB
-    Type: torch.float32
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1272) INFO: Optimizer:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1273) INFO: Scheduler: <torch.optim.lr_scheduler.ExponentialLR object at 0x7ffa7c9b4880>
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1272) INFO: Optimizer2:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1273) INFO: Scheduler2: <torch.optim.lr_scheduler.ExponentialLR object at 0x7ffa7c9b4820>
-[wieling-3-a100] 2023-12-01 15:58:41,759 (abs_task:1282) INFO: Saving the configuration in exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/config.yaml
-[wieling-3-a100] 2023-12-01 15:58:41,790 (abs_task:1293) INFO: Namespace(config='conf/train_vits.yaml', print_config=False, log_level='INFO', drop_last_iter=False, dry_run=False, iterator_type='sequence', valid_iterator_type=None, output_dir='exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10', ngpu=0, seed=67823, num_workers=4, num_att_plot=3, dist_backend='nccl', dist_init_method='env://', dist_world_size=None, dist_rank=None, local_rank=None, dist_master_addr=None, dist_master_port=None, dist_launcher=None, multiprocessing_distributed=False, unused_parameters=True, sharded_ddp=False, cudnn_enabled=True, cudnn_benchmark=False, cudnn_deterministic=False, collect_stats=True, write_collected_feats=False, max_epoch=1000, patience=None, val_scheduler_criterion=('valid', 'loss'), early_stopping_criterion=('valid', 'loss', 'min'), best_model_criterion=[['train', 'total_count', 'max']], keep_nbest_models=10, nbest_averaging_interval=0, grad_clip=-1, grad_clip_type=2.0, grad_noise=False, accum_grad=1, no_forward_run=False, resume=False, train_dtype='float32', use_amp=False, log_interval=50, use_matplotlib=True, use_tensorboard=True, create_graph_in_tensorboard=False, use_wandb=True, wandb_project='GROTTS', wandb_id=None, wandb_entity=None, wandb_name='VITS_lr_3.0e-4', wandb_model_log_interval=-1, detect_anomaly=False, use_lora=False, save_lora_only=True, lora_conf={}, pretrain_path=None, init_param=['downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv'], ignore_init_mismatch=False, freeze_param=[], num_iters_per_epoch=1000, batch_size=40, valid_batch_size=None, batch_bins=10000000, valid_batch_bins=None, train_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.10.scp'], valid_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.10.scp'], batch_type='numel', valid_batch_type=None, fold_length=[], sort_in_batch='descending', shuffle_within_batch=False, sort_batch='descending', multiple_iterator=False, chunk_length=500, chunk_shift_ratio=0.5, num_cache_chunks=1024, chunk_excluded_key_prefixes=[], chunk_default_fs=None, train_data_path_and_name_and_type=[('dump/raw/train_nodev/text', 'text', 'text'), ('dump/raw/train_nodev/wav.scp', 'speech', 'sound'), ('dump/raw/train_nodev/utt2sid', 'sids', 'text_int')], valid_data_path_and_name_and_type=[('dump/raw/train_dev/text', 'text', 'text'), ('dump/raw/train_dev/wav.scp', 'speech', 'sound'), ('dump/raw/train_dev/utt2sid', 'sids', 'text_int')], allow_variable_data_keys=False, max_cache_size=0.0, max_cache_fd=32, allow_multi_rates=False, valid_max_cache_size=None, exclude_weight_decay=False, exclude_weight_decay_conf={}, optim='adamw', optim_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler='exponentiallr', scheduler_conf={'gamma': 0.999875}, optim2='adamw', optim2_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler2='exponentiallr', scheduler2_conf={'gamma': 0.999875}, generator_first=False, token_list=['<blank>', '<unk>', '<space>', 'e', 'n', 'a', 'o', 't', 'i', 'r', 'd', 's', 'k', 'l', 'm', 'u', 'g', 'h', 'w', 'v', '.', 'z', 'b', 'p', ',', 'j', 'c', 'f', '‘', '’', ':', '?', 'ö', "'", '!', '-', ';', 'ò', 'è', 'ì', 'é', 'y', 'ë', 'x', 'q', '<sos/eos>'], odim=None, model_conf={}, use_preprocessor=True, token_type='char', bpemodel=None, non_linguistic_symbols=None, cleaner=None, g2p=None, feats_extract='fbank', feats_extract_conf={'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'fs': 22050, 'fmin': 80, 'fmax': 7600, 'n_mels': 80}, normalize=None, normalize_conf={}, tts='vits', tts_conf={'generator_type': 'vits_generator', 'generator_params': {'hidden_channels': 192, 'spks': 4, 'global_channels': 256, 'segment_size': 32, 'text_encoder_attention_heads': 2, 'text_encoder_ffn_expand': 4, 'text_encoder_blocks': 6, 'text_encoder_positionwise_layer_type': 'conv1d', 'text_encoder_positionwise_conv_kernel_size': 3, 'text_encoder_positional_encoding_layer_type': 'rel_pos', 'text_encoder_self_attention_layer_type': 'rel_selfattn', 'text_encoder_activation_type': 'swish', 'text_encoder_normalize_before': True, 'text_encoder_dropout_rate': 0.1, 'text_encoder_positional_dropout_rate': 0.0, 'text_encoder_attention_dropout_rate': 0.1, 'use_macaron_style_in_text_encoder': True, 'use_conformer_conv_in_text_encoder': False, 'text_encoder_conformer_kernel_size': -1, 'decoder_kernel_size': 7, 'decoder_channels': 512, 'decoder_upsample_scales': [8, 8, 2, 2], 'decoder_upsample_kernel_sizes': [16, 16, 4, 4], 'decoder_resblock_kernel_sizes': [3, 7, 11], 'decoder_resblock_dilations': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'use_weight_norm_in_decoder': True, 'posterior_encoder_kernel_size': 5, 'posterior_encoder_layers': 16, 'posterior_encoder_stacks': 1, 'posterior_encoder_base_dilation': 1, 'posterior_encoder_dropout_rate': 0.0, 'use_weight_norm_in_posterior_encoder': True, 'flow_flows': 4, 'flow_kernel_size': 5, 'flow_base_dilation': 1, 'flow_layers': 4, 'flow_dropout_rate': 0.0, 'use_weight_norm_in_flow': True, 'use_only_mean_in_flow': True, 'stochastic_duration_predictor_kernel_size': 3, 'stochastic_duration_predictor_dropout_rate': 0.5, 'stochastic_duration_predictor_flows': 4, 'stochastic_duration_predictor_dds_conv_layers': 3, 'vocabs': 46, 'aux_channels': 80}, 'discriminator_type': 'hifigan_multi_scale_multi_period_discriminator', 'discriminator_params': {'scales': 1, 'scale_downsample_pooling': 'AvgPool1d', 'scale_downsample_pooling_params': {'kernel_size': 4, 'stride': 2, 'padding': 2}, 'scale_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [15, 41, 5, 3], 'channels': 128, 'max_downsample_channels': 1024, 'max_groups': 16, 'bias': True, 'downsample_scales': [2, 2, 4, 4, 1], 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': False, 'use_spectral_norm': False}, 'follow_official_norm': False, 'periods': [2, 3, 5, 7, 11], 'period_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [5, 3], 'channels': 32, 'downsample_scales': [3, 3, 3, 3, 1], 'max_downsample_channels': 1024, 'bias': True, 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': True, 'use_spectral_norm': False}}, 'generator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'discriminator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'feat_match_loss_params': {'average_by_discriminators': False, 'average_by_layers': False, 'include_final_outputs': True}, 'mel_loss_params': {'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'window': 'hann', 'n_mels': 80, 'fmin': 0, 'fmax': None, 'log_base': None}, 'lambda_adv': 1.0, 'lambda_mel': 45.0, 'lambda_feat_match': 2.0, 'lambda_dur': 1.0, 'lambda_kl': 1.0, 'sampling_rate': 22050, 'cache_generator_outputs': True}, pitch_extract=None, pitch_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'f0max': 400, 'f0min': 80}, pitch_normalize=None, pitch_normalize_conf={}, energy_extract=None, energy_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None}, energy_normalize=None, energy_normalize_conf={}, required=['output_dir', 'token_list'], version='202310', distributed=False)
-# Accounting: time=18 threads=1
-# Ended (code 0) at Fri Dec  1 15:58:52 UTC 2023, elapsed time 18 seconds

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/config.yaml

@@ -1,383 +0,0 @@
-config: conf/train_vits.yaml
-print_config: false
-log_level: INFO
-drop_last_iter: false
-dry_run: false
-iterator_type: sequence
-valid_iterator_type: null
-output_dir: exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10
-ngpu: 0
-seed: 67823
-num_workers: 4
-num_att_plot: 3
-dist_backend: nccl
-dist_init_method: env://
-dist_world_size: null
-dist_rank: null
-local_rank: null
-dist_master_addr: null
-dist_master_port: null
-dist_launcher: null
-multiprocessing_distributed: false
-unused_parameters: true
-sharded_ddp: false
-cudnn_enabled: true
-cudnn_benchmark: false
-cudnn_deterministic: false
-collect_stats: true
-write_collected_feats: false
-max_epoch: 1000
-patience: null
-val_scheduler_criterion:
-- valid
-- loss
-early_stopping_criterion:
-- valid
-- loss
-- min
-best_model_criterion:
--   - train
-    - total_count
-    - max
-keep_nbest_models: 10
-nbest_averaging_interval: 0
-grad_clip: -1
-grad_clip_type: 2.0
-grad_noise: false
-accum_grad: 1
-no_forward_run: false
-resume: false
-train_dtype: float32
-use_amp: false
-log_interval: 50
-use_matplotlib: true
-use_tensorboard: true
-create_graph_in_tensorboard: false
-use_wandb: true
-wandb_project: GROTTS
-wandb_id: null
-wandb_entity: null
-wandb_name: VITS_lr_3.0e-4
-wandb_model_log_interval: -1
-detect_anomaly: false
-use_lora: false
-save_lora_only: true
-lora_conf: {}
-pretrain_path: null
-init_param:
-- downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv
-ignore_init_mismatch: false
-freeze_param: []
-num_iters_per_epoch: 1000
-batch_size: 40
-valid_batch_size: null
-batch_bins: 10000000
-valid_batch_bins: null
-train_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.10.scp
-valid_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.10.scp
-batch_type: numel
-valid_batch_type: null
-fold_length: []
-sort_in_batch: descending
-shuffle_within_batch: false
-sort_batch: descending
-multiple_iterator: false
-chunk_length: 500
-chunk_shift_ratio: 0.5
-num_cache_chunks: 1024
-chunk_excluded_key_prefixes: []
-chunk_default_fs: null
-train_data_path_and_name_and_type:
--   - dump/raw/train_nodev/text
-    - text
-    - text
--   - dump/raw/train_nodev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_nodev/utt2sid
-    - sids
-    - text_int
-valid_data_path_and_name_and_type:
--   - dump/raw/train_dev/text
-    - text
-    - text
--   - dump/raw/train_dev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_dev/utt2sid
-    - sids
-    - text_int
-allow_variable_data_keys: false
-max_cache_size: 0.0
-max_cache_fd: 32
-allow_multi_rates: false
-valid_max_cache_size: null
-exclude_weight_decay: false
-exclude_weight_decay_conf: {}
-optim: adamw
-optim_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler: exponentiallr
-scheduler_conf:
-    gamma: 0.999875
-optim2: adamw
-optim2_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler2: exponentiallr
-scheduler2_conf:
-    gamma: 0.999875
-generator_first: false
-token_list:
-- <blank>
-- <unk>
-- <space>
-- e
-- n
-- a
-- o
-- t
-- i
-- r
-- d
-- s
-- k
-- l
-- m
-- u
-- g
-- h
-- w
-- v
-- .
-- z
-- b
-- p
-- ','
-- j
-- c
-- f
-- ‘
-- ’
-- ':'
-- '?'
-- ö
-- ''''
-- '!'
-- '-'
-- ;
-- ò
-- è
-- ì
-- é
-- y
-- ë
-- x
-- q
-- <sos/eos>
-odim: null
-model_conf: {}
-use_preprocessor: true
-token_type: char
-bpemodel: null
-non_linguistic_symbols: null
-cleaner: null
-g2p: null
-feats_extract: fbank
-feats_extract_conf:
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-    fs: 22050
-    fmin: 80
-    fmax: 7600
-    n_mels: 80
-normalize: null
-normalize_conf: {}
-tts: vits
-tts_conf:
-    generator_type: vits_generator
-    generator_params:
-        hidden_channels: 192
-        spks: 4
-        global_channels: 256
-        segment_size: 32
-        text_encoder_attention_heads: 2
-        text_encoder_ffn_expand: 4
-        text_encoder_blocks: 6
-        text_encoder_positionwise_layer_type: conv1d
-        text_encoder_positionwise_conv_kernel_size: 3
-        text_encoder_positional_encoding_layer_type: rel_pos
-        text_encoder_self_attention_layer_type: rel_selfattn
-        text_encoder_activation_type: swish
-        text_encoder_normalize_before: true
-        text_encoder_dropout_rate: 0.1
-        text_encoder_positional_dropout_rate: 0.0
-        text_encoder_attention_dropout_rate: 0.1
-        use_macaron_style_in_text_encoder: true
-        use_conformer_conv_in_text_encoder: false
-        text_encoder_conformer_kernel_size: -1
-        decoder_kernel_size: 7
-        decoder_channels: 512
-        decoder_upsample_scales:
-        - 8
-        - 8
-        - 2
-        - 2
-        decoder_upsample_kernel_sizes:
-        - 16
-        - 16
-        - 4
-        - 4
-        decoder_resblock_kernel_sizes:
-        - 3
-        - 7
-        - 11
-        decoder_resblock_dilations:
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        use_weight_norm_in_decoder: true
-        posterior_encoder_kernel_size: 5
-        posterior_encoder_layers: 16
-        posterior_encoder_stacks: 1
-        posterior_encoder_base_dilation: 1
-        posterior_encoder_dropout_rate: 0.0
-        use_weight_norm_in_posterior_encoder: true
-        flow_flows: 4
-        flow_kernel_size: 5
-        flow_base_dilation: 1
-        flow_layers: 4
-        flow_dropout_rate: 0.0
-        use_weight_norm_in_flow: true
-        use_only_mean_in_flow: true
-        stochastic_duration_predictor_kernel_size: 3
-        stochastic_duration_predictor_dropout_rate: 0.5
-        stochastic_duration_predictor_flows: 4
-        stochastic_duration_predictor_dds_conv_layers: 3
-        vocabs: 46
-        aux_channels: 80
-    discriminator_type: hifigan_multi_scale_multi_period_discriminator
-    discriminator_params:
-        scales: 1
-        scale_downsample_pooling: AvgPool1d
-        scale_downsample_pooling_params:
-            kernel_size: 4
-            stride: 2
-            padding: 2
-        scale_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 15
-            - 41
-            - 5
-            - 3
-            channels: 128
-            max_downsample_channels: 1024
-            max_groups: 16
-            bias: true
-            downsample_scales:
-            - 2
-            - 2
-            - 4
-            - 4
-            - 1
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: false
-            use_spectral_norm: false
-        follow_official_norm: false
-        periods:
-        - 2
-        - 3
-        - 5
-        - 7
-        - 11
-        period_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 5
-            - 3
-            channels: 32
-            downsample_scales:
-            - 3
-            - 3
-            - 3
-            - 3
-            - 1
-            max_downsample_channels: 1024
-            bias: true
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: true
-            use_spectral_norm: false
-    generator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    discriminator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    feat_match_loss_params:
-        average_by_discriminators: false
-        average_by_layers: false
-        include_final_outputs: true
-    mel_loss_params:
-        fs: 22050
-        n_fft: 1024
-        hop_length: 256
-        win_length: null
-        window: hann
-        n_mels: 80
-        fmin: 0
-        fmax: null
-        log_base: null
-    lambda_adv: 1.0
-    lambda_mel: 45.0
-    lambda_feat_match: 2.0
-    lambda_dur: 1.0
-    lambda_kl: 1.0
-    sampling_rate: 22050
-    cache_generator_outputs: true
-pitch_extract: null
-pitch_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    f0max: 400
-    f0min: 80
-pitch_normalize: null
-pitch_normalize_conf: {}
-energy_extract: null
-energy_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-energy_normalize: null
-energy_normalize_conf: {}
-required:
-- output_dir
-- token_list
-version: '202310'
-distributed: false

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:af79191d13bfcc12aecb8430b36098d2678e2934adbe5d1a061dee1deef5773e
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:574f2692fe20cefd55fe03010ac1815121a4aa53069aa042c591fb317d7804b0
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/sids_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2288-MoanMorn 1
-Hoogelaandsters-2294-MoanMorn 1
-Hoogelaandsters-2299-MoanMorn 1
-Hoogelaandsters-2303-MoanMorn 1
-Hoogelaandsters-2307-MoanMorn 1
-Hoogelaandsters-2311-MoanMorn 1
-Hoogelaandsters-2316-MoanMorn 1
-Hoogelaandsters-2320-MoanMorn 1
-Hoogelaandsters-2324-MoanMorn 1
-Hoogelaandsters-2328-MoanMorn 1
-Hoogelaandsters-2332-MoanMorn 1
-Hoogelaandsters-2336-MoanMorn 1
-Hoogelaandsters-2340-MoanMorn 1
-Hoogelaandsters-2344-MoanMorn 1
-Hoogelaandsters-2348-MoanMorn 1
-Hoogelaandsters-2352-MoanMorn 1
-Hoogelaandsters-2356-MoanMorn 1
-Hoogelaandsters-2361-MoanMorn 1
-Hoogelaandsters-2365-MoanMorn 1
-Hoogelaandsters-2369-MoanMorn 1
-Hoogelaandsters-2373-MoanMorn 1
-Hoogelaandsters-2377-MoanMorn 1
-Hoogelaandsters-2381-MoanMorn 1
-Hoogelaandsters-2386-MoanMorn 1
-Hoogelaandsters-2390-MoanMorn 1
-Hoogelaandsters-2395-MoanMorn 1
-Hoogelaandsters-2399-MoanMorn 1
-Hoogelaandsters-2403-MoanMorn 1
-Hoogelaandsters-2407-MoanMorn 1
-Hoogelaandsters-2411-MoanMorn 1
-Hoogelaandsters-2418-MoanMorn 1
-Hoogelaandsters-2422-MoanMorn 1
-Hoogelaandsters-2426-MoanMorn 1
-Hoogelaandsters-2431-MoanMorn 1
-Hoogelaandsters-2435-MoanMorn 1
-Hoogelaandsters-2440-MoanMorn 1
-Hoogelaandsters-2444-MoanMorn 1
-Hoogelaandsters-2448-MoanMorn 1
-Hoogelaandsters-2452-MoanMorn 1
-Hoogelaandsters-2456-MoanMorn 1
-Hoogelaandsters-2289-MoanMorn 1
-Hoogelaandsters-2295-MoanMorn 1
-Hoogelaandsters-2300-MoanMorn 1
-Hoogelaandsters-2304-MoanMorn 1
-Hoogelaandsters-2308-MoanMorn 1
-Hoogelaandsters-2312-MoanMorn 1
-Hoogelaandsters-2317-MoanMorn 1
-Hoogelaandsters-2321-MoanMorn 1
-Hoogelaandsters-2325-MoanMorn 1
-Hoogelaandsters-2329-MoanMorn 1
-Hoogelaandsters-2333-MoanMorn 1
-Hoogelaandsters-2337-MoanMorn 1
-Hoogelaandsters-2341-MoanMorn 1
-Hoogelaandsters-2345-MoanMorn 1
-Hoogelaandsters-2349-MoanMorn 1
-Hoogelaandsters-2353-MoanMorn 1
-Hoogelaandsters-2358-MoanMorn 1
-Hoogelaandsters-2362-MoanMorn 1
-Hoogelaandsters-2366-MoanMorn 1
-Hoogelaandsters-2370-MoanMorn 1
-Hoogelaandsters-2374-MoanMorn 1
-Hoogelaandsters-2378-MoanMorn 1
-Hoogelaandsters-2382-MoanMorn 1
-Hoogelaandsters-2387-MoanMorn 1
-Hoogelaandsters-2391-MoanMorn 1
-Hoogelaandsters-2396-MoanMorn 1
-Hoogelaandsters-2400-MoanMorn 1
-Hoogelaandsters-2404-MoanMorn 1
-Hoogelaandsters-2408-MoanMorn 1
-Hoogelaandsters-2412-MoanMorn 1
-Hoogelaandsters-2419-MoanMorn 1
-Hoogelaandsters-2423-MoanMorn 1
-Hoogelaandsters-2427-MoanMorn 1
-Hoogelaandsters-2432-MoanMorn 1
-Hoogelaandsters-2437-MoanMorn 1
-Hoogelaandsters-2441-MoanMorn 1
-Hoogelaandsters-2445-MoanMorn 1
-Hoogelaandsters-2449-MoanMorn 1
-Hoogelaandsters-2453-MoanMorn 1
-Hoogelaandsters-2457-MoanMorn 1
-Hoogelaandsters-2291-MoanMorn 1
-Hoogelaandsters-2297-MoanMorn 1
-Hoogelaandsters-2301-MoanMorn 1
-Hoogelaandsters-2305-MoanMorn 1
-Hoogelaandsters-2309-MoanMorn 1
-Hoogelaandsters-2314-MoanMorn 1
-Hoogelaandsters-2318-MoanMorn 1
-Hoogelaandsters-2322-MoanMorn 1
-Hoogelaandsters-2326-MoanMorn 1
-Hoogelaandsters-2330-MoanMorn 1
-Hoogelaandsters-2334-MoanMorn 1
-Hoogelaandsters-2338-MoanMorn 1
-Hoogelaandsters-2342-MoanMorn 1
-Hoogelaandsters-2346-MoanMorn 1
-Hoogelaandsters-2350-MoanMorn 1
-Hoogelaandsters-2354-MoanMorn 1
-Hoogelaandsters-2359-MoanMorn 1
-Hoogelaandsters-2363-MoanMorn 1
-Hoogelaandsters-2367-MoanMorn 1
-Hoogelaandsters-2371-MoanMorn 1
-Hoogelaandsters-2375-MoanMorn 1
-Hoogelaandsters-2379-MoanMorn 1
-Hoogelaandsters-2384-MoanMorn 1
-Hoogelaandsters-2388-MoanMorn 1
-Hoogelaandsters-2392-MoanMorn 1
-Hoogelaandsters-2397-MoanMorn 1
-Hoogelaandsters-2401-MoanMorn 1
-Hoogelaandsters-2405-MoanMorn 1
-Hoogelaandsters-2409-MoanMorn 1
-Hoogelaandsters-2413-MoanMorn 1
-Hoogelaandsters-2420-MoanMorn 1
-Hoogelaandsters-2424-MoanMorn 1
-Hoogelaandsters-2429-MoanMorn 1
-Hoogelaandsters-2433-MoanMorn 1
-Hoogelaandsters-2438-MoanMorn 1
-Hoogelaandsters-2442-MoanMorn 1
-Hoogelaandsters-2446-MoanMorn 1
-Hoogelaandsters-2450-MoanMorn 1
-Hoogelaandsters-2454-MoanMorn 1
-Hoogelaandsters-2459-MoanMorn 1
-Hoogelaandsters-2293-MoanMorn 1
-Hoogelaandsters-2298-MoanMorn 1
-Hoogelaandsters-2302-MoanMorn 1
-Hoogelaandsters-2306-MoanMorn 1
-Hoogelaandsters-2310-MoanMorn 1
-Hoogelaandsters-2315-MoanMorn 1
-Hoogelaandsters-2319-MoanMorn 1
-Hoogelaandsters-2323-MoanMorn 1
-Hoogelaandsters-2327-MoanMorn 1
-Hoogelaandsters-2331-MoanMorn 1
-Hoogelaandsters-2335-MoanMorn 1
-Hoogelaandsters-2339-MoanMorn 1
-Hoogelaandsters-2343-MoanMorn 1
-Hoogelaandsters-2347-MoanMorn 1
-Hoogelaandsters-2351-MoanMorn 1
-Hoogelaandsters-2355-MoanMorn 1
-Hoogelaandsters-2360-MoanMorn 1
-Hoogelaandsters-2364-MoanMorn 1
-Hoogelaandsters-2368-MoanMorn 1
-Hoogelaandsters-2372-MoanMorn 1
-Hoogelaandsters-2376-MoanMorn 1
-Hoogelaandsters-2380-MoanMorn 1
-Hoogelaandsters-2385-MoanMorn 1
-Hoogelaandsters-2389-MoanMorn 1
-Hoogelaandsters-2393-MoanMorn 1
-Hoogelaandsters-2398-MoanMorn 1
-Hoogelaandsters-2402-MoanMorn 1
-Hoogelaandsters-2406-MoanMorn 1
-Hoogelaandsters-2410-MoanMorn 1
-Hoogelaandsters-2415-MoanMorn 1
-Hoogelaandsters-2421-MoanMorn 1
-Hoogelaandsters-2425-MoanMorn 1
-Hoogelaandsters-2430-MoanMorn 1
-Hoogelaandsters-2434-MoanMorn 1
-Hoogelaandsters-2439-MoanMorn 1
-Hoogelaandsters-2443-MoanMorn 1
-Hoogelaandsters-2447-MoanMorn 1
-Hoogelaandsters-2451-MoanMorn 1
-Hoogelaandsters-2455-MoanMorn 1
-Hoogelaandsters-2460-MoanMorn 1
-Hoogelaandsters-2461-MoanMorn 1
-Hoogelaandsters-2465-MoanMorn 1
-Hoogelaandsters-2469-MoanMorn 1
-Hoogelaandsters-2473-MoanMorn 1
-Hoogelaandsters-2477-MoanMorn 1
-Hoogelaandsters-2481-MoanMorn 1
-Hoogelaandsters-2486-MoanMorn 1
-Hoogelaandsters-2490-MoanMorn 1
-Hoogelaandsters-2494-MoanMorn 1
-Hoogelaandsters-2499-MoanMorn 1
-Hoogelaandsters-2503-MoanMorn 1
-Hoogelaandsters-2507-MoanMorn 1
-Hoogelaandsters-2511-MoanMorn 1
-Hoogelaandsters-2515-MoanMorn 1
-Hoogelaandsters-2519-MoanMorn 1
-Hoogelaandsters-2523-MoanMorn 1
-Hoogelaandsters-2527-MoanMorn 1
-Hoogelaandsters-2531-MoanMorn 1
-Hoogelaandsters-2535-MoanMorn 1
-Hoogelaandsters-2540-MoanMorn 1
-Hoogelaandsters-2544-MoanMorn 1
-Hoogelaandsters-2548-MoanMorn 1
-Hoogelaandsters-2552-MoanMorn 1
-Hoogelaandsters-2462-MoanMorn 1
-Hoogelaandsters-2466-MoanMorn 1
-Hoogelaandsters-2470-MoanMorn 1
-Hoogelaandsters-2474-MoanMorn 1
-Hoogelaandsters-2478-MoanMorn 1
-Hoogelaandsters-2482-MoanMorn 1
-Hoogelaandsters-2487-MoanMorn 1
-Hoogelaandsters-2491-MoanMorn 1
-Hoogelaandsters-2495-MoanMorn 1
-Hoogelaandsters-2500-MoanMorn 1
-Hoogelaandsters-2504-MoanMorn 1
-Hoogelaandsters-2508-MoanMorn 1
-Hoogelaandsters-2512-MoanMorn 1
-Hoogelaandsters-2516-MoanMorn 1
-Hoogelaandsters-2520-MoanMorn 1
-Hoogelaandsters-2524-MoanMorn 1
-Hoogelaandsters-2528-MoanMorn 1
-Hoogelaandsters-2532-MoanMorn 1
-Hoogelaandsters-2537-MoanMorn 1
-Hoogelaandsters-2541-MoanMorn 1
-Hoogelaandsters-2545-MoanMorn 1
-Hoogelaandsters-2549-MoanMorn 1
-Hoogelaandsters-2463-MoanMorn 1
-Hoogelaandsters-2467-MoanMorn 1
-Hoogelaandsters-2471-MoanMorn 1
-Hoogelaandsters-2475-MoanMorn 1
-Hoogelaandsters-2479-MoanMorn 1
-Hoogelaandsters-2483-MoanMorn 1
-Hoogelaandsters-2488-MoanMorn 1
-Hoogelaandsters-2492-MoanMorn 1
-Hoogelaandsters-2497-MoanMorn 1
-Hoogelaandsters-2501-MoanMorn 1
-Hoogelaandsters-2505-MoanMorn 1
-Hoogelaandsters-2509-MoanMorn 1
-Hoogelaandsters-2513-MoanMorn 1
-Hoogelaandsters-2517-MoanMorn 1
-Hoogelaandsters-2521-MoanMorn 1
-Hoogelaandsters-2525-MoanMorn 1
-Hoogelaandsters-2529-MoanMorn 1
-Hoogelaandsters-2533-MoanMorn 1
-Hoogelaandsters-2538-MoanMorn 1
-Hoogelaandsters-2542-MoanMorn 1
-Hoogelaandsters-2546-MoanMorn 1
-Hoogelaandsters-2550-MoanMorn 1
-Hoogelaandsters-2464-MoanMorn 1
-Hoogelaandsters-2468-MoanMorn 1
-Hoogelaandsters-2472-MoanMorn 1
-Hoogelaandsters-2476-MoanMorn 1
-Hoogelaandsters-2480-MoanMorn 1
-Hoogelaandsters-2485-MoanMorn 1
-Hoogelaandsters-2489-MoanMorn 1
-Hoogelaandsters-2493-MoanMorn 1
-Hoogelaandsters-2498-MoanMorn 1
-Hoogelaandsters-2502-MoanMorn 1
-Hoogelaandsters-2506-MoanMorn 1
-Hoogelaandsters-2510-MoanMorn 1
-Hoogelaandsters-2514-MoanMorn 1
-Hoogelaandsters-2518-MoanMorn 1
-Hoogelaandsters-2522-MoanMorn 1
-Hoogelaandsters-2526-MoanMorn 1
-Hoogelaandsters-2530-MoanMorn 1
-Hoogelaandsters-2534-MoanMorn 1
-Hoogelaandsters-2539-MoanMorn 1
-Hoogelaandsters-2543-MoanMorn 1
-Hoogelaandsters-2547-MoanMorn 1
-Hoogelaandsters-2551-MoanMorn 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/speech_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2288-MoanMorn 117445
-Hoogelaandsters-2294-MoanMorn 152790
-Hoogelaandsters-2299-MoanMorn 95858
-Hoogelaandsters-2303-MoanMorn 79027
-Hoogelaandsters-2307-MoanMorn 97165
-Hoogelaandsters-2311-MoanMorn 77335
-Hoogelaandsters-2316-MoanMorn 115718
-Hoogelaandsters-2320-MoanMorn 256717
-Hoogelaandsters-2324-MoanMorn 62264
-Hoogelaandsters-2328-MoanMorn 162365
-Hoogelaandsters-2332-MoanMorn 244260
-Hoogelaandsters-2336-MoanMorn 85700
-Hoogelaandsters-2340-MoanMorn 64222
-Hoogelaandsters-2344-MoanMorn 52887
-Hoogelaandsters-2348-MoanMorn 167328
-Hoogelaandsters-2352-MoanMorn 135077
-Hoogelaandsters-2356-MoanMorn 134774
-Hoogelaandsters-2361-MoanMorn 74095
-Hoogelaandsters-2365-MoanMorn 90694
-Hoogelaandsters-2369-MoanMorn 187706
-Hoogelaandsters-2373-MoanMorn 204919
-Hoogelaandsters-2377-MoanMorn 98649
-Hoogelaandsters-2381-MoanMorn 128977
-Hoogelaandsters-2386-MoanMorn 90664
-Hoogelaandsters-2390-MoanMorn 323639
-Hoogelaandsters-2395-MoanMorn 61014
-Hoogelaandsters-2399-MoanMorn 270015
-Hoogelaandsters-2403-MoanMorn 80986
-Hoogelaandsters-2407-MoanMorn 120663
-Hoogelaandsters-2411-MoanMorn 101842
-Hoogelaandsters-2418-MoanMorn 172321
-Hoogelaandsters-2422-MoanMorn 119205
-Hoogelaandsters-2426-MoanMorn 41493
-Hoogelaandsters-2431-MoanMorn 155997
-Hoogelaandsters-2435-MoanMorn 149437
-Hoogelaandsters-2440-MoanMorn 106718
-Hoogelaandsters-2444-MoanMorn 223892
-Hoogelaandsters-2448-MoanMorn 111220
-Hoogelaandsters-2452-MoanMorn 210485
-Hoogelaandsters-2456-MoanMorn 235422
-Hoogelaandsters-2289-MoanMorn 270050
-Hoogelaandsters-2295-MoanMorn 64915
-Hoogelaandsters-2300-MoanMorn 234710
-Hoogelaandsters-2304-MoanMorn 281339
-Hoogelaandsters-2308-MoanMorn 51878
-Hoogelaandsters-2312-MoanMorn 76205
-Hoogelaandsters-2317-MoanMorn 256841
-Hoogelaandsters-2321-MoanMorn 109919
-Hoogelaandsters-2325-MoanMorn 56870
-Hoogelaandsters-2329-MoanMorn 56448
-Hoogelaandsters-2333-MoanMorn 114522
-Hoogelaandsters-2337-MoanMorn 69593
-Hoogelaandsters-2341-MoanMorn 124803
-Hoogelaandsters-2345-MoanMorn 103242
-Hoogelaandsters-2349-MoanMorn 193651
-Hoogelaandsters-2353-MoanMorn 81114
-Hoogelaandsters-2358-MoanMorn 95962
-Hoogelaandsters-2362-MoanMorn 196092
-Hoogelaandsters-2366-MoanMorn 124712
-Hoogelaandsters-2370-MoanMorn 90317
-Hoogelaandsters-2374-MoanMorn 147754
-Hoogelaandsters-2378-MoanMorn 237031
-Hoogelaandsters-2382-MoanMorn 143942
-Hoogelaandsters-2387-MoanMorn 56448
-Hoogelaandsters-2391-MoanMorn 69609
-Hoogelaandsters-2396-MoanMorn 99452
-Hoogelaandsters-2400-MoanMorn 78535
-Hoogelaandsters-2404-MoanMorn 130726
-Hoogelaandsters-2408-MoanMorn 193877
-Hoogelaandsters-2412-MoanMorn 53626
-Hoogelaandsters-2419-MoanMorn 316671
-Hoogelaandsters-2423-MoanMorn 117655
-Hoogelaandsters-2427-MoanMorn 319451
-Hoogelaandsters-2432-MoanMorn 162733
-Hoogelaandsters-2437-MoanMorn 107702
-Hoogelaandsters-2441-MoanMorn 53626
-Hoogelaandsters-2445-MoanMorn 115718
-Hoogelaandsters-2449-MoanMorn 217179
-Hoogelaandsters-2453-MoanMorn 66767
-Hoogelaandsters-2457-MoanMorn 255710
-Hoogelaandsters-2291-MoanMorn 31046
-Hoogelaandsters-2297-MoanMorn 377978
-Hoogelaandsters-2301-MoanMorn 141324
-Hoogelaandsters-2305-MoanMorn 84672
-Hoogelaandsters-2309-MoanMorn 79787
-Hoogelaandsters-2314-MoanMorn 75773
-Hoogelaandsters-2318-MoanMorn 160174
-Hoogelaandsters-2322-MoanMorn 132835
-Hoogelaandsters-2326-MoanMorn 101606
-Hoogelaandsters-2330-MoanMorn 135933
-Hoogelaandsters-2334-MoanMorn 78025
-Hoogelaandsters-2338-MoanMorn 129449
-Hoogelaandsters-2342-MoanMorn 38856
-Hoogelaandsters-2346-MoanMorn 174562
-Hoogelaandsters-2350-MoanMorn 128445
-Hoogelaandsters-2354-MoanMorn 160490
-Hoogelaandsters-2359-MoanMorn 228493
-Hoogelaandsters-2363-MoanMorn 98433
-Hoogelaandsters-2367-MoanMorn 56448
-Hoogelaandsters-2371-MoanMorn 77172
-Hoogelaandsters-2375-MoanMorn 126690
-Hoogelaandsters-2379-MoanMorn 166001
-Hoogelaandsters-2384-MoanMorn 414643
-Hoogelaandsters-2388-MoanMorn 135592
-Hoogelaandsters-2392-MoanMorn 199799
-Hoogelaandsters-2397-MoanMorn 35135
-Hoogelaandsters-2401-MoanMorn 189656
-Hoogelaandsters-2405-MoanMorn 280393
-Hoogelaandsters-2409-MoanMorn 222797
-Hoogelaandsters-2413-MoanMorn 244597
-Hoogelaandsters-2420-MoanMorn 112072
-Hoogelaandsters-2424-MoanMorn 65163
-Hoogelaandsters-2429-MoanMorn 73382
-Hoogelaandsters-2433-MoanMorn 144641
-Hoogelaandsters-2438-MoanMorn 132752
-Hoogelaandsters-2442-MoanMorn 81602
-Hoogelaandsters-2446-MoanMorn 178999
-Hoogelaandsters-2450-MoanMorn 139858
-Hoogelaandsters-2454-MoanMorn 297187
-Hoogelaandsters-2459-MoanMorn 348416
-Hoogelaandsters-2293-MoanMorn 173015
-Hoogelaandsters-2298-MoanMorn 114482
-Hoogelaandsters-2302-MoanMorn 129931
-Hoogelaandsters-2306-MoanMorn 59714
-Hoogelaandsters-2310-MoanMorn 98201
-Hoogelaandsters-2315-MoanMorn 365776
-Hoogelaandsters-2319-MoanMorn 118998
-Hoogelaandsters-2323-MoanMorn 133528
-Hoogelaandsters-2327-MoanMorn 137558
-Hoogelaandsters-2331-MoanMorn 30210
-Hoogelaandsters-2335-MoanMorn 147860
-Hoogelaandsters-2339-MoanMorn 162823
-Hoogelaandsters-2343-MoanMorn 124110
-Hoogelaandsters-2347-MoanMorn 122295
-Hoogelaandsters-2351-MoanMorn 69605
-Hoogelaandsters-2355-MoanMorn 117925
-Hoogelaandsters-2360-MoanMorn 79325
-Hoogelaandsters-2364-MoanMorn 155437
-Hoogelaandsters-2368-MoanMorn 347030
-Hoogelaandsters-2372-MoanMorn 116698
-Hoogelaandsters-2376-MoanMorn 127143
-Hoogelaandsters-2380-MoanMorn 214235
-Hoogelaandsters-2385-MoanMorn 114241
-Hoogelaandsters-2389-MoanMorn 136878
-Hoogelaandsters-2393-MoanMorn 110685
-Hoogelaandsters-2398-MoanMorn 148220
-Hoogelaandsters-2402-MoanMorn 189724
-Hoogelaandsters-2406-MoanMorn 142468
-Hoogelaandsters-2410-MoanMorn 80968
-Hoogelaandsters-2415-MoanMorn 92540
-Hoogelaandsters-2421-MoanMorn 117050
-Hoogelaandsters-2425-MoanMorn 91659
-Hoogelaandsters-2430-MoanMorn 109139
-Hoogelaandsters-2434-MoanMorn 150905
-Hoogelaandsters-2439-MoanMorn 89319
-Hoogelaandsters-2443-MoanMorn 151690
-Hoogelaandsters-2447-MoanMorn 191388
-Hoogelaandsters-2451-MoanMorn 256575
-Hoogelaandsters-2455-MoanMorn 143121
-Hoogelaandsters-2460-MoanMorn 148964
-Hoogelaandsters-2461-MoanMorn 202459
-Hoogelaandsters-2465-MoanMorn 77353
-Hoogelaandsters-2469-MoanMorn 202205
-Hoogelaandsters-2473-MoanMorn 144494
-Hoogelaandsters-2477-MoanMorn 261203
-Hoogelaandsters-2481-MoanMorn 55523
-Hoogelaandsters-2486-MoanMorn 154579
-Hoogelaandsters-2490-MoanMorn 189951
-Hoogelaandsters-2494-MoanMorn 322452
-Hoogelaandsters-2499-MoanMorn 80295
-Hoogelaandsters-2503-MoanMorn 49367
-Hoogelaandsters-2507-MoanMorn 150136
-Hoogelaandsters-2511-MoanMorn 154426
-Hoogelaandsters-2515-MoanMorn 72940
-Hoogelaandsters-2519-MoanMorn 117693
-Hoogelaandsters-2523-MoanMorn 65506
-Hoogelaandsters-2527-MoanMorn 186486
-Hoogelaandsters-2531-MoanMorn 66332
-Hoogelaandsters-2535-MoanMorn 59270
-Hoogelaandsters-2540-MoanMorn 92545
-Hoogelaandsters-2544-MoanMorn 74985
-Hoogelaandsters-2548-MoanMorn 187809
-Hoogelaandsters-2552-MoanMorn 165658
-Hoogelaandsters-2462-MoanMorn 151515
-Hoogelaandsters-2466-MoanMorn 75367
-Hoogelaandsters-2470-MoanMorn 59270
-Hoogelaandsters-2474-MoanMorn 76195
-Hoogelaandsters-2478-MoanMorn 97331
-Hoogelaandsters-2482-MoanMorn 55406
-Hoogelaandsters-2487-MoanMorn 267130
-Hoogelaandsters-2491-MoanMorn 197159
-Hoogelaandsters-2495-MoanMorn 115714
-Hoogelaandsters-2500-MoanMorn 115155
-Hoogelaandsters-2504-MoanMorn 122315
-Hoogelaandsters-2508-MoanMorn 173640
-Hoogelaandsters-2512-MoanMorn 387092
-Hoogelaandsters-2516-MoanMorn 56448
-Hoogelaandsters-2520-MoanMorn 271221
-Hoogelaandsters-2524-MoanMorn 206196
-Hoogelaandsters-2528-MoanMorn 216440
-Hoogelaandsters-2532-MoanMorn 102710
-Hoogelaandsters-2537-MoanMorn 426213
-Hoogelaandsters-2541-MoanMorn 119456
-Hoogelaandsters-2545-MoanMorn 107043
-Hoogelaandsters-2549-MoanMorn 164150
-Hoogelaandsters-2463-MoanMorn 50962
-Hoogelaandsters-2467-MoanMorn 122984
-Hoogelaandsters-2471-MoanMorn 73212
-Hoogelaandsters-2475-MoanMorn 71040
-Hoogelaandsters-2479-MoanMorn 79027
-Hoogelaandsters-2483-MoanMorn 33869
-Hoogelaandsters-2488-MoanMorn 61981
-Hoogelaandsters-2492-MoanMorn 341502
-Hoogelaandsters-2497-MoanMorn 150345
-Hoogelaandsters-2501-MoanMorn 134396
-Hoogelaandsters-2505-MoanMorn 151752
-Hoogelaandsters-2509-MoanMorn 127099
-Hoogelaandsters-2513-MoanMorn 124822
-Hoogelaandsters-2517-MoanMorn 90045
-Hoogelaandsters-2521-MoanMorn 73110
-Hoogelaandsters-2525-MoanMorn 70279
-Hoogelaandsters-2529-MoanMorn 188522
-Hoogelaandsters-2533-MoanMorn 158311
-Hoogelaandsters-2538-MoanMorn 136310
-Hoogelaandsters-2542-MoanMorn 253487
-Hoogelaandsters-2546-MoanMorn 67738
-Hoogelaandsters-2550-MoanMorn 115291
-Hoogelaandsters-2464-MoanMorn 93139
-Hoogelaandsters-2468-MoanMorn 244448
-Hoogelaandsters-2472-MoanMorn 111901
-Hoogelaandsters-2476-MoanMorn 129484
-Hoogelaandsters-2480-MoanMorn 122230
-Hoogelaandsters-2485-MoanMorn 63917
-Hoogelaandsters-2489-MoanMorn 146083
-Hoogelaandsters-2493-MoanMorn 79573
-Hoogelaandsters-2498-MoanMorn 116144
-Hoogelaandsters-2502-MoanMorn 126772
-Hoogelaandsters-2506-MoanMorn 84358
-Hoogelaandsters-2510-MoanMorn 258288
-Hoogelaandsters-2514-MoanMorn 218302
-Hoogelaandsters-2518-MoanMorn 387837
-Hoogelaandsters-2522-MoanMorn 67878
-Hoogelaandsters-2526-MoanMorn 139164
-Hoogelaandsters-2530-MoanMorn 150036
-Hoogelaandsters-2534-MoanMorn 163176
-Hoogelaandsters-2539-MoanMorn 115911
-Hoogelaandsters-2543-MoanMorn 387694
-Hoogelaandsters-2547-MoanMorn 160101
-Hoogelaandsters-2551-MoanMorn 178543

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/train/text_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2288-MoanMorn 68
-Hoogelaandsters-2294-MoanMorn 95
-Hoogelaandsters-2299-MoanMorn 42
-Hoogelaandsters-2303-MoanMorn 54
-Hoogelaandsters-2307-MoanMorn 55
-Hoogelaandsters-2311-MoanMorn 42
-Hoogelaandsters-2316-MoanMorn 81
-Hoogelaandsters-2320-MoanMorn 149
-Hoogelaandsters-2324-MoanMorn 41
-Hoogelaandsters-2328-MoanMorn 104
-Hoogelaandsters-2332-MoanMorn 162
-Hoogelaandsters-2336-MoanMorn 53
-Hoogelaandsters-2340-MoanMorn 34
-Hoogelaandsters-2344-MoanMorn 23
-Hoogelaandsters-2348-MoanMorn 98
-Hoogelaandsters-2352-MoanMorn 90
-Hoogelaandsters-2356-MoanMorn 86
-Hoogelaandsters-2361-MoanMorn 51
-Hoogelaandsters-2365-MoanMorn 45
-Hoogelaandsters-2369-MoanMorn 119
-Hoogelaandsters-2373-MoanMorn 129
-Hoogelaandsters-2377-MoanMorn 50
-Hoogelaandsters-2381-MoanMorn 74
-Hoogelaandsters-2386-MoanMorn 57
-Hoogelaandsters-2390-MoanMorn 185
-Hoogelaandsters-2395-MoanMorn 30
-Hoogelaandsters-2399-MoanMorn 178
-Hoogelaandsters-2403-MoanMorn 55
-Hoogelaandsters-2407-MoanMorn 62
-Hoogelaandsters-2411-MoanMorn 52
-Hoogelaandsters-2418-MoanMorn 103
-Hoogelaandsters-2422-MoanMorn 63
-Hoogelaandsters-2426-MoanMorn 21
-Hoogelaandsters-2431-MoanMorn 90
-Hoogelaandsters-2435-MoanMorn 97
-Hoogelaandsters-2440-MoanMorn 63
-Hoogelaandsters-2444-MoanMorn 144
-Hoogelaandsters-2448-MoanMorn 62
-Hoogelaandsters-2452-MoanMorn 128
-Hoogelaandsters-2456-MoanMorn 165
-Hoogelaandsters-2289-MoanMorn 177
-Hoogelaandsters-2295-MoanMorn 45
-Hoogelaandsters-2300-MoanMorn 137
-Hoogelaandsters-2304-MoanMorn 175
-Hoogelaandsters-2308-MoanMorn 27
-Hoogelaandsters-2312-MoanMorn 48
-Hoogelaandsters-2317-MoanMorn 133
-Hoogelaandsters-2321-MoanMorn 73
-Hoogelaandsters-2325-MoanMorn 48
-Hoogelaandsters-2329-MoanMorn 31
-Hoogelaandsters-2333-MoanMorn 74
-Hoogelaandsters-2337-MoanMorn 32
-Hoogelaandsters-2341-MoanMorn 80
-Hoogelaandsters-2345-MoanMorn 62
-Hoogelaandsters-2349-MoanMorn 104
-Hoogelaandsters-2353-MoanMorn 45
-Hoogelaandsters-2358-MoanMorn 46
-Hoogelaandsters-2362-MoanMorn 119
-Hoogelaandsters-2366-MoanMorn 69
-Hoogelaandsters-2370-MoanMorn 69
-Hoogelaandsters-2374-MoanMorn 84
-Hoogelaandsters-2378-MoanMorn 134
-Hoogelaandsters-2382-MoanMorn 93
-Hoogelaandsters-2387-MoanMorn 48
-Hoogelaandsters-2391-MoanMorn 34
-Hoogelaandsters-2396-MoanMorn 51
-Hoogelaandsters-2400-MoanMorn 44
-Hoogelaandsters-2404-MoanMorn 75
-Hoogelaandsters-2408-MoanMorn 112
-Hoogelaandsters-2412-MoanMorn 25
-Hoogelaandsters-2419-MoanMorn 186
-Hoogelaandsters-2423-MoanMorn 60
-Hoogelaandsters-2427-MoanMorn 210
-Hoogelaandsters-2432-MoanMorn 105
-Hoogelaandsters-2437-MoanMorn 65
-Hoogelaandsters-2441-MoanMorn 35
-Hoogelaandsters-2445-MoanMorn 76
-Hoogelaandsters-2449-MoanMorn 125
-Hoogelaandsters-2453-MoanMorn 33
-Hoogelaandsters-2457-MoanMorn 156
-Hoogelaandsters-2291-MoanMorn 22
-Hoogelaandsters-2297-MoanMorn 202
-Hoogelaandsters-2301-MoanMorn 75
-Hoogelaandsters-2305-MoanMorn 47
-Hoogelaandsters-2309-MoanMorn 43
-Hoogelaandsters-2314-MoanMorn 32
-Hoogelaandsters-2318-MoanMorn 101
-Hoogelaandsters-2322-MoanMorn 61
-Hoogelaandsters-2326-MoanMorn 63
-Hoogelaandsters-2330-MoanMorn 74
-Hoogelaandsters-2334-MoanMorn 50
-Hoogelaandsters-2338-MoanMorn 74
-Hoogelaandsters-2342-MoanMorn 24
-Hoogelaandsters-2346-MoanMorn 108
-Hoogelaandsters-2350-MoanMorn 82
-Hoogelaandsters-2354-MoanMorn 89
-Hoogelaandsters-2359-MoanMorn 141
-Hoogelaandsters-2363-MoanMorn 63
-Hoogelaandsters-2367-MoanMorn 36
-Hoogelaandsters-2371-MoanMorn 46
-Hoogelaandsters-2375-MoanMorn 89
-Hoogelaandsters-2379-MoanMorn 106
-Hoogelaandsters-2384-MoanMorn 301
-Hoogelaandsters-2388-MoanMorn 80
-Hoogelaandsters-2392-MoanMorn 110
-Hoogelaandsters-2397-MoanMorn 18
-Hoogelaandsters-2401-MoanMorn 102
-Hoogelaandsters-2405-MoanMorn 199
-Hoogelaandsters-2409-MoanMorn 103
-Hoogelaandsters-2413-MoanMorn 147
-Hoogelaandsters-2420-MoanMorn 65
-Hoogelaandsters-2424-MoanMorn 38
-Hoogelaandsters-2429-MoanMorn 57
-Hoogelaandsters-2433-MoanMorn 75
-Hoogelaandsters-2438-MoanMorn 76
-Hoogelaandsters-2442-MoanMorn 42
-Hoogelaandsters-2446-MoanMorn 106
-Hoogelaandsters-2450-MoanMorn 97
-Hoogelaandsters-2454-MoanMorn 183
-Hoogelaandsters-2459-MoanMorn 224
-Hoogelaandsters-2293-MoanMorn 88
-Hoogelaandsters-2298-MoanMorn 65
-Hoogelaandsters-2302-MoanMorn 86
-Hoogelaandsters-2306-MoanMorn 29
-Hoogelaandsters-2310-MoanMorn 66
-Hoogelaandsters-2315-MoanMorn 210
-Hoogelaandsters-2319-MoanMorn 78
-Hoogelaandsters-2323-MoanMorn 71
-Hoogelaandsters-2327-MoanMorn 86
-Hoogelaandsters-2331-MoanMorn 21
-Hoogelaandsters-2335-MoanMorn 91
-Hoogelaandsters-2339-MoanMorn 109
-Hoogelaandsters-2343-MoanMorn 65
-Hoogelaandsters-2347-MoanMorn 70
-Hoogelaandsters-2351-MoanMorn 42
-Hoogelaandsters-2355-MoanMorn 64
-Hoogelaandsters-2360-MoanMorn 50
-Hoogelaandsters-2364-MoanMorn 267
-Hoogelaandsters-2368-MoanMorn 216
-Hoogelaandsters-2372-MoanMorn 66
-Hoogelaandsters-2376-MoanMorn 68
-Hoogelaandsters-2380-MoanMorn 136
-Hoogelaandsters-2385-MoanMorn 61
-Hoogelaandsters-2389-MoanMorn 82
-Hoogelaandsters-2393-MoanMorn 82
-Hoogelaandsters-2398-MoanMorn 97
-Hoogelaandsters-2402-MoanMorn 135
-Hoogelaandsters-2406-MoanMorn 80
-Hoogelaandsters-2410-MoanMorn 48
-Hoogelaandsters-2415-MoanMorn 69
-Hoogelaandsters-2421-MoanMorn 64
-Hoogelaandsters-2425-MoanMorn 50
-Hoogelaandsters-2430-MoanMorn 60
-Hoogelaandsters-2434-MoanMorn 94
-Hoogelaandsters-2439-MoanMorn 56
-Hoogelaandsters-2443-MoanMorn 80
-Hoogelaandsters-2447-MoanMorn 130
-Hoogelaandsters-2451-MoanMorn 163
-Hoogelaandsters-2455-MoanMorn 84
-Hoogelaandsters-2460-MoanMorn 97
-Hoogelaandsters-2461-MoanMorn 126
-Hoogelaandsters-2465-MoanMorn 36
-Hoogelaandsters-2469-MoanMorn 117
-Hoogelaandsters-2473-MoanMorn 79
-Hoogelaandsters-2477-MoanMorn 171
-Hoogelaandsters-2481-MoanMorn 30
-Hoogelaandsters-2486-MoanMorn 84
-Hoogelaandsters-2490-MoanMorn 135
-Hoogelaandsters-2494-MoanMorn 164
-Hoogelaandsters-2499-MoanMorn 53
-Hoogelaandsters-2503-MoanMorn 16
-Hoogelaandsters-2507-MoanMorn 82
-Hoogelaandsters-2511-MoanMorn 91
-Hoogelaandsters-2515-MoanMorn 49
-Hoogelaandsters-2519-MoanMorn 70
-Hoogelaandsters-2523-MoanMorn 37
-Hoogelaandsters-2527-MoanMorn 116
-Hoogelaandsters-2531-MoanMorn 32
-Hoogelaandsters-2535-MoanMorn 43
-Hoogelaandsters-2540-MoanMorn 74
-Hoogelaandsters-2544-MoanMorn 49
-Hoogelaandsters-2548-MoanMorn 121
-Hoogelaandsters-2552-MoanMorn 92
-Hoogelaandsters-2462-MoanMorn 103
-Hoogelaandsters-2466-MoanMorn 38
-Hoogelaandsters-2470-MoanMorn 24
-Hoogelaandsters-2474-MoanMorn 33
-Hoogelaandsters-2478-MoanMorn 43
-Hoogelaandsters-2482-MoanMorn 18
-Hoogelaandsters-2487-MoanMorn 175
-Hoogelaandsters-2491-MoanMorn 115
-Hoogelaandsters-2495-MoanMorn 77
-Hoogelaandsters-2500-MoanMorn 66
-Hoogelaandsters-2504-MoanMorn 65
-Hoogelaandsters-2508-MoanMorn 91
-Hoogelaandsters-2512-MoanMorn 230
-Hoogelaandsters-2516-MoanMorn 40
-Hoogelaandsters-2520-MoanMorn 168
-Hoogelaandsters-2524-MoanMorn 134
-Hoogelaandsters-2528-MoanMorn 154
-Hoogelaandsters-2532-MoanMorn 60
-Hoogelaandsters-2537-MoanMorn 292
-Hoogelaandsters-2541-MoanMorn 71
-Hoogelaandsters-2545-MoanMorn 62
-Hoogelaandsters-2549-MoanMorn 123
-Hoogelaandsters-2463-MoanMorn 250
-Hoogelaandsters-2467-MoanMorn 63
-Hoogelaandsters-2471-MoanMorn 51
-Hoogelaandsters-2475-MoanMorn 59
-Hoogelaandsters-2479-MoanMorn 47
-Hoogelaandsters-2483-MoanMorn 9
-Hoogelaandsters-2488-MoanMorn 33
-Hoogelaandsters-2492-MoanMorn 173
-Hoogelaandsters-2497-MoanMorn 95
-Hoogelaandsters-2501-MoanMorn 66
-Hoogelaandsters-2505-MoanMorn 89
-Hoogelaandsters-2509-MoanMorn 82
-Hoogelaandsters-2513-MoanMorn 68
-Hoogelaandsters-2517-MoanMorn 54
-Hoogelaandsters-2521-MoanMorn 44
-Hoogelaandsters-2525-MoanMorn 39
-Hoogelaandsters-2529-MoanMorn 116
-Hoogelaandsters-2533-MoanMorn 106
-Hoogelaandsters-2538-MoanMorn 71
-Hoogelaandsters-2542-MoanMorn 162
-Hoogelaandsters-2546-MoanMorn 63
-Hoogelaandsters-2550-MoanMorn 56
-Hoogelaandsters-2464-MoanMorn 59
-Hoogelaandsters-2468-MoanMorn 146
-Hoogelaandsters-2472-MoanMorn 76
-Hoogelaandsters-2476-MoanMorn 74
-Hoogelaandsters-2480-MoanMorn 76
-Hoogelaandsters-2485-MoanMorn 49
-Hoogelaandsters-2489-MoanMorn 90
-Hoogelaandsters-2493-MoanMorn 36
-Hoogelaandsters-2498-MoanMorn 67
-Hoogelaandsters-2502-MoanMorn 71
-Hoogelaandsters-2506-MoanMorn 44
-Hoogelaandsters-2510-MoanMorn 134
-Hoogelaandsters-2514-MoanMorn 139
-Hoogelaandsters-2518-MoanMorn 217
-Hoogelaandsters-2522-MoanMorn 44
-Hoogelaandsters-2526-MoanMorn 102
-Hoogelaandsters-2530-MoanMorn 96
-Hoogelaandsters-2534-MoanMorn 92
-Hoogelaandsters-2539-MoanMorn 86
-Hoogelaandsters-2543-MoanMorn 249
-Hoogelaandsters-2547-MoanMorn 102
-Hoogelaandsters-2551-MoanMorn 97

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:15ffdfea3d7f1876d6a022d37c05b82e4b1fce8aefbb321bd0cddd482ac406c8
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:eb9a150225e097511ae2eaa906b3c3fe455256237be5df75b824e8c2ae27d3e5
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/sids_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-3047-MoanMorn 1
-Hoogelaandsters-3322-OpBuus 1
-Hoogelaandsters-3115-MoanMorn 1
-Hoogelaandsters-3184-MoanMorn 1
-Hoogelaandsters-3251-MoanMorn 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/speech_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-3047-MoanMorn 152884
-Hoogelaandsters-3322-OpBuus 41382
-Hoogelaandsters-3115-MoanMorn 54997
-Hoogelaandsters-3184-MoanMorn 112079
-Hoogelaandsters-3251-MoanMorn 176275

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.10/valid/text_shape

@@ -1,5 +0,0 @@
-Hoogelaandsters-3047-MoanMorn 84
-Hoogelaandsters-3322-OpBuus 21
-Hoogelaandsters-3115-MoanMorn 32
-Hoogelaandsters-3184-MoanMorn 60
-Hoogelaandsters-3251-MoanMorn 113

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11.log

@@ -1,1152 +0,0 @@
-# python3 -m espnet2.bin.gan_tts_train --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.11.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.11.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000 
-# Started at Fri Dec  1 15:58:34 UTC 2023
-#
-/data2/p280965/tts/espnet/tools/venv/bin/python3 /data2/p280965/tts/espnet/espnet2/bin/gan_tts_train.py --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.11.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.11.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000
-[wieling-3-a100] 2023-12-01 15:58:40,493 (gan_tts:293) INFO: Vocabulary size: 46
-[wieling-3-a100] 2023-12-01 15:58:40,627 (encoder:174) INFO: encoder self-attention layer type = relative self-attention
-/data2/p280965/tts/espnet/tools/venv/lib/python3.9/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
-  warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
-/data2/p280965/tts/espnet/espnet2/gan_tts/vits/monotonic_align/__init__.py:19: UserWarning: Cython version is not available. Fallback to 'EXPERIMETAL' numba version. If you want to use the cython version, please build it as follows: `cd espnet2/gan_tts/vits/monotonic_align; python setup.py build_ext --inplace`
-  warnings.warn(
-[wieling-3-a100] 2023-12-01 15:58:41,832 (abs_task:1268) INFO: pytorch.version=2.1.0+cu121, cuda.available=True, cudnn.version=8902, cudnn.benchmark=False, cudnn.deterministic=False
-[wieling-3-a100] 2023-12-01 15:58:41,847 (abs_task:1269) INFO: Model structure:
-ESPnetGANTTSModel(
-  (feats_extract): LogMelFbank(
-    (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-    (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=80, fmax=7600, htk=False)
-  )
-  (tts): VITS(
-    (generator): VITSGenerator(
-      (text_encoder): TextEncoder(
-        (emb): Embedding(46, 192)
-        (encoder): Encoder(
-          (embed): Sequential(
-            (0): RelPositionalEncoding(
-              (dropout): Dropout(p=0.0, inplace=False)
-            )
-          )
-          (encoders): MultiSequential(
-            (0): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (1): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (2): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (3): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (4): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (5): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-          )
-          (after_norm): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (decoder): HiFiGANGenerator(
-        (input_conv): Conv1d(192, 512, kernel_size=(7,), stride=(1,), padding=(3,))
-        (upsamples): ModuleList(
-          (0): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(512, 256, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (1): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(256, 128, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (2): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(128, 64, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-          (3): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(64, 32, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-        )
-        (blocks): ModuleList(
-          (0): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (1): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (2): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (3): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (4): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (5): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (6): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (7): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (8): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (9): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (10): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (11): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-        )
-        (output_conv): Sequential(
-          (0): LeakyReLU(negative_slope=0.01)
-          (1): Conv1d(32, 1, kernel_size=(7,), stride=(1,), padding=(3,))
-          (2): Tanh()
-        )
-        (global_conv): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
-      )
-      (posterior_encoder): PosteriorEncoder(
-        (input_conv): Conv1d(80, 192, kernel_size=(1,), stride=(1,))
-        (encoder): WaveNet(
-          (conv_layers): ModuleList(
-            (0-15): 16 x ResidualBlock(
-              (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-              (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-              (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-            )
-          )
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (flow): ResidualAffineCouplingBlock(
-        (flows): ModuleList(
-          (0): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (1): FlipFlow()
-          (2): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (3): FlipFlow()
-          (4): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (5): FlipFlow()
-          (6): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (7): FlipFlow()
-        )
-      )
-      (duration_predictor): StochasticDurationPredictor(
-        (pre): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (log_flow): LogFlow()
-        (flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (post_pre): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-        (post_dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (post_proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (post_flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (global_conv): Conv1d(256, 192, kernel_size=(1,), stride=(1,))
-      )
-      (global_emb): Embedding(4, 256)
-    )
-    (discriminator): HiFiGANMultiScaleMultiPeriodDiscriminator(
-      (msd): HiFiGANMultiScaleDiscriminator(
-        (discriminators): ModuleList(
-          (0): HiFiGANScaleDiscriminator(
-            (layers): ModuleList(
-              (0): Sequential(
-                (0): Conv1d(1, 128, kernel_size=(15,), stride=(1,), padding=(7,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv1d(128, 128, kernel_size=(41,), stride=(2,), padding=(20,), groups=4)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv1d(128, 256, kernel_size=(41,), stride=(2,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv1d(256, 512, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv1d(512, 1024, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (5): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(41,), stride=(1,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (6): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(5,), stride=(1,), padding=(2,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (7): Conv1d(1024, 1, kernel_size=(3,), stride=(1,), padding=(1,))
-            )
-          )
-        )
-      )
-      (mpd): HiFiGANMultiPeriodDiscriminator(
-        (discriminators): ModuleList(
-          (0-4): 5 x HiFiGANPeriodDiscriminator(
-            (convs): ModuleList(
-              (0): Sequential(
-                (0): Conv2d(1, 32, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv2d(32, 128, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv2d(128, 512, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv2d(512, 1024, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv2d(1024, 1024, kernel_size=(5, 1), stride=(1, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-            )
-            (output_conv): Conv2d(1024, 1, kernel_size=(2, 1), stride=(1, 1), padding=(1, 0))
-          )
-        )
-      )
-    )
-    (generator_adv_loss): GeneratorAdversarialLoss()
-    (discriminator_adv_loss): DiscriminatorAdversarialLoss()
-    (feat_match_loss): FeatureMatchLoss()
-    (mel_loss): MelSpectrogramLoss(
-      (wav_to_mel): LogMelFbank(
-        (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-        (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=0, fmax=11025.0, htk=False)
-      )
-    )
-    (kl_loss): KLDivergenceLoss()
-  )
-)
-
-Model summary:
-    Class Name: ESPnetGANTTSModel
-    Total Number of model parameters: 96.24 M
-    Number of trainable parameters: 96.24 M (100.0%)
-    Size: 384.96 MB
-    Type: torch.float32
-[wieling-3-a100] 2023-12-01 15:58:41,847 (abs_task:1272) INFO: Optimizer:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,847 (abs_task:1273) INFO: Scheduler: <torch.optim.lr_scheduler.ExponentialLR object at 0x7f9de23eb8b0>
-[wieling-3-a100] 2023-12-01 15:58:41,847 (abs_task:1272) INFO: Optimizer2:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:41,847 (abs_task:1273) INFO: Scheduler2: <torch.optim.lr_scheduler.ExponentialLR object at 0x7f9de23eb850>
-[wieling-3-a100] 2023-12-01 15:58:41,848 (abs_task:1282) INFO: Saving the configuration in exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/config.yaml
-[wieling-3-a100] 2023-12-01 15:58:41,866 (abs_task:1293) INFO: Namespace(config='conf/train_vits.yaml', print_config=False, log_level='INFO', drop_last_iter=False, dry_run=False, iterator_type='sequence', valid_iterator_type=None, output_dir='exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11', ngpu=0, seed=67823, num_workers=4, num_att_plot=3, dist_backend='nccl', dist_init_method='env://', dist_world_size=None, dist_rank=None, local_rank=None, dist_master_addr=None, dist_master_port=None, dist_launcher=None, multiprocessing_distributed=False, unused_parameters=True, sharded_ddp=False, cudnn_enabled=True, cudnn_benchmark=False, cudnn_deterministic=False, collect_stats=True, write_collected_feats=False, max_epoch=1000, patience=None, val_scheduler_criterion=('valid', 'loss'), early_stopping_criterion=('valid', 'loss', 'min'), best_model_criterion=[['train', 'total_count', 'max']], keep_nbest_models=10, nbest_averaging_interval=0, grad_clip=-1, grad_clip_type=2.0, grad_noise=False, accum_grad=1, no_forward_run=False, resume=False, train_dtype='float32', use_amp=False, log_interval=50, use_matplotlib=True, use_tensorboard=True, create_graph_in_tensorboard=False, use_wandb=True, wandb_project='GROTTS', wandb_id=None, wandb_entity=None, wandb_name='VITS_lr_3.0e-4', wandb_model_log_interval=-1, detect_anomaly=False, use_lora=False, save_lora_only=True, lora_conf={}, pretrain_path=None, init_param=['downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv'], ignore_init_mismatch=False, freeze_param=[], num_iters_per_epoch=1000, batch_size=40, valid_batch_size=None, batch_bins=10000000, valid_batch_bins=None, train_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.11.scp'], valid_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.11.scp'], batch_type='numel', valid_batch_type=None, fold_length=[], sort_in_batch='descending', shuffle_within_batch=False, sort_batch='descending', multiple_iterator=False, chunk_length=500, chunk_shift_ratio=0.5, num_cache_chunks=1024, chunk_excluded_key_prefixes=[], chunk_default_fs=None, train_data_path_and_name_and_type=[('dump/raw/train_nodev/text', 'text', 'text'), ('dump/raw/train_nodev/wav.scp', 'speech', 'sound'), ('dump/raw/train_nodev/utt2sid', 'sids', 'text_int')], valid_data_path_and_name_and_type=[('dump/raw/train_dev/text', 'text', 'text'), ('dump/raw/train_dev/wav.scp', 'speech', 'sound'), ('dump/raw/train_dev/utt2sid', 'sids', 'text_int')], allow_variable_data_keys=False, max_cache_size=0.0, max_cache_fd=32, allow_multi_rates=False, valid_max_cache_size=None, exclude_weight_decay=False, exclude_weight_decay_conf={}, optim='adamw', optim_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler='exponentiallr', scheduler_conf={'gamma': 0.999875}, optim2='adamw', optim2_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler2='exponentiallr', scheduler2_conf={'gamma': 0.999875}, generator_first=False, token_list=['<blank>', '<unk>', '<space>', 'e', 'n', 'a', 'o', 't', 'i', 'r', 'd', 's', 'k', 'l', 'm', 'u', 'g', 'h', 'w', 'v', '.', 'z', 'b', 'p', ',', 'j', 'c', 'f', '‘', '’', ':', '?', 'ö', "'", '!', '-', ';', 'ò', 'è', 'ì', 'é', 'y', 'ë', 'x', 'q', '<sos/eos>'], odim=None, model_conf={}, use_preprocessor=True, token_type='char', bpemodel=None, non_linguistic_symbols=None, cleaner=None, g2p=None, feats_extract='fbank', feats_extract_conf={'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'fs': 22050, 'fmin': 80, 'fmax': 7600, 'n_mels': 80}, normalize=None, normalize_conf={}, tts='vits', tts_conf={'generator_type': 'vits_generator', 'generator_params': {'hidden_channels': 192, 'spks': 4, 'global_channels': 256, 'segment_size': 32, 'text_encoder_attention_heads': 2, 'text_encoder_ffn_expand': 4, 'text_encoder_blocks': 6, 'text_encoder_positionwise_layer_type': 'conv1d', 'text_encoder_positionwise_conv_kernel_size': 3, 'text_encoder_positional_encoding_layer_type': 'rel_pos', 'text_encoder_self_attention_layer_type': 'rel_selfattn', 'text_encoder_activation_type': 'swish', 'text_encoder_normalize_before': True, 'text_encoder_dropout_rate': 0.1, 'text_encoder_positional_dropout_rate': 0.0, 'text_encoder_attention_dropout_rate': 0.1, 'use_macaron_style_in_text_encoder': True, 'use_conformer_conv_in_text_encoder': False, 'text_encoder_conformer_kernel_size': -1, 'decoder_kernel_size': 7, 'decoder_channels': 512, 'decoder_upsample_scales': [8, 8, 2, 2], 'decoder_upsample_kernel_sizes': [16, 16, 4, 4], 'decoder_resblock_kernel_sizes': [3, 7, 11], 'decoder_resblock_dilations': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'use_weight_norm_in_decoder': True, 'posterior_encoder_kernel_size': 5, 'posterior_encoder_layers': 16, 'posterior_encoder_stacks': 1, 'posterior_encoder_base_dilation': 1, 'posterior_encoder_dropout_rate': 0.0, 'use_weight_norm_in_posterior_encoder': True, 'flow_flows': 4, 'flow_kernel_size': 5, 'flow_base_dilation': 1, 'flow_layers': 4, 'flow_dropout_rate': 0.0, 'use_weight_norm_in_flow': True, 'use_only_mean_in_flow': True, 'stochastic_duration_predictor_kernel_size': 3, 'stochastic_duration_predictor_dropout_rate': 0.5, 'stochastic_duration_predictor_flows': 4, 'stochastic_duration_predictor_dds_conv_layers': 3, 'vocabs': 46, 'aux_channels': 80}, 'discriminator_type': 'hifigan_multi_scale_multi_period_discriminator', 'discriminator_params': {'scales': 1, 'scale_downsample_pooling': 'AvgPool1d', 'scale_downsample_pooling_params': {'kernel_size': 4, 'stride': 2, 'padding': 2}, 'scale_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [15, 41, 5, 3], 'channels': 128, 'max_downsample_channels': 1024, 'max_groups': 16, 'bias': True, 'downsample_scales': [2, 2, 4, 4, 1], 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': False, 'use_spectral_norm': False}, 'follow_official_norm': False, 'periods': [2, 3, 5, 7, 11], 'period_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [5, 3], 'channels': 32, 'downsample_scales': [3, 3, 3, 3, 1], 'max_downsample_channels': 1024, 'bias': True, 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': True, 'use_spectral_norm': False}}, 'generator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'discriminator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'feat_match_loss_params': {'average_by_discriminators': False, 'average_by_layers': False, 'include_final_outputs': True}, 'mel_loss_params': {'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'window': 'hann', 'n_mels': 80, 'fmin': 0, 'fmax': None, 'log_base': None}, 'lambda_adv': 1.0, 'lambda_mel': 45.0, 'lambda_feat_match': 2.0, 'lambda_dur': 1.0, 'lambda_kl': 1.0, 'sampling_rate': 22050, 'cache_generator_outputs': True}, pitch_extract=None, pitch_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'f0max': 400, 'f0min': 80}, pitch_normalize=None, pitch_normalize_conf={}, energy_extract=None, energy_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None}, energy_normalize=None, energy_normalize_conf={}, required=['output_dir', 'token_list'], version='202310', distributed=False)
-# Accounting: time=18 threads=1
-# Ended (code 0) at Fri Dec  1 15:58:52 UTC 2023, elapsed time 18 seconds

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/config.yaml

@@ -1,383 +0,0 @@
-config: conf/train_vits.yaml
-print_config: false
-log_level: INFO
-drop_last_iter: false
-dry_run: false
-iterator_type: sequence
-valid_iterator_type: null
-output_dir: exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11
-ngpu: 0
-seed: 67823
-num_workers: 4
-num_att_plot: 3
-dist_backend: nccl
-dist_init_method: env://
-dist_world_size: null
-dist_rank: null
-local_rank: null
-dist_master_addr: null
-dist_master_port: null
-dist_launcher: null
-multiprocessing_distributed: false
-unused_parameters: true
-sharded_ddp: false
-cudnn_enabled: true
-cudnn_benchmark: false
-cudnn_deterministic: false
-collect_stats: true
-write_collected_feats: false
-max_epoch: 1000
-patience: null
-val_scheduler_criterion:
-- valid
-- loss
-early_stopping_criterion:
-- valid
-- loss
-- min
-best_model_criterion:
--   - train
-    - total_count
-    - max
-keep_nbest_models: 10
-nbest_averaging_interval: 0
-grad_clip: -1
-grad_clip_type: 2.0
-grad_noise: false
-accum_grad: 1
-no_forward_run: false
-resume: false
-train_dtype: float32
-use_amp: false
-log_interval: 50
-use_matplotlib: true
-use_tensorboard: true
-create_graph_in_tensorboard: false
-use_wandb: true
-wandb_project: GROTTS
-wandb_id: null
-wandb_entity: null
-wandb_name: VITS_lr_3.0e-4
-wandb_model_log_interval: -1
-detect_anomaly: false
-use_lora: false
-save_lora_only: true
-lora_conf: {}
-pretrain_path: null
-init_param:
-- downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv
-ignore_init_mismatch: false
-freeze_param: []
-num_iters_per_epoch: 1000
-batch_size: 40
-valid_batch_size: null
-batch_bins: 10000000
-valid_batch_bins: null
-train_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.11.scp
-valid_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.11.scp
-batch_type: numel
-valid_batch_type: null
-fold_length: []
-sort_in_batch: descending
-shuffle_within_batch: false
-sort_batch: descending
-multiple_iterator: false
-chunk_length: 500
-chunk_shift_ratio: 0.5
-num_cache_chunks: 1024
-chunk_excluded_key_prefixes: []
-chunk_default_fs: null
-train_data_path_and_name_and_type:
--   - dump/raw/train_nodev/text
-    - text
-    - text
--   - dump/raw/train_nodev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_nodev/utt2sid
-    - sids
-    - text_int
-valid_data_path_and_name_and_type:
--   - dump/raw/train_dev/text
-    - text
-    - text
--   - dump/raw/train_dev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_dev/utt2sid
-    - sids
-    - text_int
-allow_variable_data_keys: false
-max_cache_size: 0.0
-max_cache_fd: 32
-allow_multi_rates: false
-valid_max_cache_size: null
-exclude_weight_decay: false
-exclude_weight_decay_conf: {}
-optim: adamw
-optim_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler: exponentiallr
-scheduler_conf:
-    gamma: 0.999875
-optim2: adamw
-optim2_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler2: exponentiallr
-scheduler2_conf:
-    gamma: 0.999875
-generator_first: false
-token_list:
-- <blank>
-- <unk>
-- <space>
-- e
-- n
-- a
-- o
-- t
-- i
-- r
-- d
-- s
-- k
-- l
-- m
-- u
-- g
-- h
-- w
-- v
-- .
-- z
-- b
-- p
-- ','
-- j
-- c
-- f
-- ‘
-- ’
-- ':'
-- '?'
-- ö
-- ''''
-- '!'
-- '-'
-- ;
-- ò
-- è
-- ì
-- é
-- y
-- ë
-- x
-- q
-- <sos/eos>
-odim: null
-model_conf: {}
-use_preprocessor: true
-token_type: char
-bpemodel: null
-non_linguistic_symbols: null
-cleaner: null
-g2p: null
-feats_extract: fbank
-feats_extract_conf:
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-    fs: 22050
-    fmin: 80
-    fmax: 7600
-    n_mels: 80
-normalize: null
-normalize_conf: {}
-tts: vits
-tts_conf:
-    generator_type: vits_generator
-    generator_params:
-        hidden_channels: 192
-        spks: 4
-        global_channels: 256
-        segment_size: 32
-        text_encoder_attention_heads: 2
-        text_encoder_ffn_expand: 4
-        text_encoder_blocks: 6
-        text_encoder_positionwise_layer_type: conv1d
-        text_encoder_positionwise_conv_kernel_size: 3
-        text_encoder_positional_encoding_layer_type: rel_pos
-        text_encoder_self_attention_layer_type: rel_selfattn
-        text_encoder_activation_type: swish
-        text_encoder_normalize_before: true
-        text_encoder_dropout_rate: 0.1
-        text_encoder_positional_dropout_rate: 0.0
-        text_encoder_attention_dropout_rate: 0.1
-        use_macaron_style_in_text_encoder: true
-        use_conformer_conv_in_text_encoder: false
-        text_encoder_conformer_kernel_size: -1
-        decoder_kernel_size: 7
-        decoder_channels: 512
-        decoder_upsample_scales:
-        - 8
-        - 8
-        - 2
-        - 2
-        decoder_upsample_kernel_sizes:
-        - 16
-        - 16
-        - 4
-        - 4
-        decoder_resblock_kernel_sizes:
-        - 3
-        - 7
-        - 11
-        decoder_resblock_dilations:
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        use_weight_norm_in_decoder: true
-        posterior_encoder_kernel_size: 5
-        posterior_encoder_layers: 16
-        posterior_encoder_stacks: 1
-        posterior_encoder_base_dilation: 1
-        posterior_encoder_dropout_rate: 0.0
-        use_weight_norm_in_posterior_encoder: true
-        flow_flows: 4
-        flow_kernel_size: 5
-        flow_base_dilation: 1
-        flow_layers: 4
-        flow_dropout_rate: 0.0
-        use_weight_norm_in_flow: true
-        use_only_mean_in_flow: true
-        stochastic_duration_predictor_kernel_size: 3
-        stochastic_duration_predictor_dropout_rate: 0.5
-        stochastic_duration_predictor_flows: 4
-        stochastic_duration_predictor_dds_conv_layers: 3
-        vocabs: 46
-        aux_channels: 80
-    discriminator_type: hifigan_multi_scale_multi_period_discriminator
-    discriminator_params:
-        scales: 1
-        scale_downsample_pooling: AvgPool1d
-        scale_downsample_pooling_params:
-            kernel_size: 4
-            stride: 2
-            padding: 2
-        scale_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 15
-            - 41
-            - 5
-            - 3
-            channels: 128
-            max_downsample_channels: 1024
-            max_groups: 16
-            bias: true
-            downsample_scales:
-            - 2
-            - 2
-            - 4
-            - 4
-            - 1
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: false
-            use_spectral_norm: false
-        follow_official_norm: false
-        periods:
-        - 2
-        - 3
-        - 5
-        - 7
-        - 11
-        period_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 5
-            - 3
-            channels: 32
-            downsample_scales:
-            - 3
-            - 3
-            - 3
-            - 3
-            - 1
-            max_downsample_channels: 1024
-            bias: true
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: true
-            use_spectral_norm: false
-    generator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    discriminator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    feat_match_loss_params:
-        average_by_discriminators: false
-        average_by_layers: false
-        include_final_outputs: true
-    mel_loss_params:
-        fs: 22050
-        n_fft: 1024
-        hop_length: 256
-        win_length: null
-        window: hann
-        n_mels: 80
-        fmin: 0
-        fmax: null
-        log_base: null
-    lambda_adv: 1.0
-    lambda_mel: 45.0
-    lambda_feat_match: 2.0
-    lambda_dur: 1.0
-    lambda_kl: 1.0
-    sampling_rate: 22050
-    cache_generator_outputs: true
-pitch_extract: null
-pitch_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    f0max: 400
-    f0min: 80
-pitch_normalize: null
-pitch_normalize_conf: {}
-energy_extract: null
-energy_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-energy_normalize: null
-energy_normalize_conf: {}
-required:
-- output_dir
-- token_list
-version: '202310'
-distributed: false

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:8fe5b424d637b31235fd2b5fab1dd19d7aae7171c7ba6a8652d681775beb1a09
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:5b6b29e56a828f743a9e1137c738e65a6c821a5e18d1c4c8c801663798432712
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/sids_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2553-MoanMorn 1
-Hoogelaandsters-2557-MoanMorn 1
-Hoogelaandsters-2562-MoanMorn 1
-Hoogelaandsters-2567-MoanMorn 1
-Hoogelaandsters-2571-MoanMorn 1
-Hoogelaandsters-2575-MoanMorn 1
-Hoogelaandsters-2579-MoanMorn 1
-Hoogelaandsters-2583-MoanMorn 1
-Hoogelaandsters-2587-MoanMorn 1
-Hoogelaandsters-2592-MoanMorn 1
-Hoogelaandsters-2596-MoanMorn 1
-Hoogelaandsters-2600-MoanMorn 1
-Hoogelaandsters-2604-MoanMorn 1
-Hoogelaandsters-2608-MoanMorn 1
-Hoogelaandsters-2612-MoanMorn 1
-Hoogelaandsters-2616-MoanMorn 1
-Hoogelaandsters-2620-MoanMorn 1
-Hoogelaandsters-2624-MoanMorn 1
-Hoogelaandsters-2628-MoanMorn 1
-Hoogelaandsters-2633-MoanMorn 1
-Hoogelaandsters-2637-MoanMorn 1
-Hoogelaandsters-2641-MoanMorn 1
-Hoogelaandsters-2645-MoanMorn 1
-Hoogelaandsters-2649-MoanMorn 1
-Hoogelaandsters-2653-MoanMorn 1
-Hoogelaandsters-2657-MoanMorn 1
-Hoogelaandsters-2661-MoanMorn 1
-Hoogelaandsters-2665-MoanMorn 1
-Hoogelaandsters-2669-MoanMorn 1
-Hoogelaandsters-2673-MoanMorn 1
-Hoogelaandsters-2678-MoanMorn 1
-Hoogelaandsters-2682-MoanMorn 1
-Hoogelaandsters-2686-MoanMorn 1
-Hoogelaandsters-2690-MoanMorn 1
-Hoogelaandsters-2694-MoanMorn 1
-Hoogelaandsters-2699-MoanMorn 1
-Hoogelaandsters-2703-MoanMorn 1
-Hoogelaandsters-2707-MoanMorn 1
-Hoogelaandsters-2711-MoanMorn 1
-Hoogelaandsters-2715-MoanMorn 1
-Hoogelaandsters-2554-MoanMorn 1
-Hoogelaandsters-2558-MoanMorn 1
-Hoogelaandsters-2564-MoanMorn 1
-Hoogelaandsters-2568-MoanMorn 1
-Hoogelaandsters-2572-MoanMorn 1
-Hoogelaandsters-2576-MoanMorn 1
-Hoogelaandsters-2580-MoanMorn 1
-Hoogelaandsters-2584-MoanMorn 1
-Hoogelaandsters-2588-MoanMorn 1
-Hoogelaandsters-2593-MoanMorn 1
-Hoogelaandsters-2597-MoanMorn 1
-Hoogelaandsters-2601-MoanMorn 1
-Hoogelaandsters-2605-MoanMorn 1
-Hoogelaandsters-2609-MoanMorn 1
-Hoogelaandsters-2613-MoanMorn 1
-Hoogelaandsters-2617-MoanMorn 1
-Hoogelaandsters-2621-MoanMorn 1
-Hoogelaandsters-2625-MoanMorn 1
-Hoogelaandsters-2629-MoanMorn 1
-Hoogelaandsters-2634-MoanMorn 1
-Hoogelaandsters-2638-MoanMorn 1
-Hoogelaandsters-2642-MoanMorn 1
-Hoogelaandsters-2646-MoanMorn 1
-Hoogelaandsters-2650-MoanMorn 1
-Hoogelaandsters-2654-MoanMorn 1
-Hoogelaandsters-2658-MoanMorn 1
-Hoogelaandsters-2662-MoanMorn 1
-Hoogelaandsters-2666-MoanMorn 1
-Hoogelaandsters-2670-MoanMorn 1
-Hoogelaandsters-2674-MoanMorn 1
-Hoogelaandsters-2679-MoanMorn 1
-Hoogelaandsters-2683-MoanMorn 1
-Hoogelaandsters-2687-MoanMorn 1
-Hoogelaandsters-2691-MoanMorn 1
-Hoogelaandsters-2695-MoanMorn 1
-Hoogelaandsters-2700-MoanMorn 1
-Hoogelaandsters-2704-MoanMorn 1
-Hoogelaandsters-2708-MoanMorn 1
-Hoogelaandsters-2712-MoanMorn 1
-Hoogelaandsters-2716-MoanMorn 1
-Hoogelaandsters-2555-MoanMorn 1
-Hoogelaandsters-2559-MoanMorn 1
-Hoogelaandsters-2565-MoanMorn 1
-Hoogelaandsters-2569-MoanMorn 1
-Hoogelaandsters-2573-MoanMorn 1
-Hoogelaandsters-2577-MoanMorn 1
-Hoogelaandsters-2581-MoanMorn 1
-Hoogelaandsters-2585-MoanMorn 1
-Hoogelaandsters-2589-MoanMorn 1
-Hoogelaandsters-2594-MoanMorn 1
-Hoogelaandsters-2598-MoanMorn 1
-Hoogelaandsters-2602-MoanMorn 1
-Hoogelaandsters-2606-MoanMorn 1
-Hoogelaandsters-2610-MoanMorn 1
-Hoogelaandsters-2614-MoanMorn 1
-Hoogelaandsters-2618-MoanMorn 1
-Hoogelaandsters-2622-MoanMorn 1
-Hoogelaandsters-2626-MoanMorn 1
-Hoogelaandsters-2631-MoanMorn 1
-Hoogelaandsters-2635-MoanMorn 1
-Hoogelaandsters-2639-MoanMorn 1
-Hoogelaandsters-2643-MoanMorn 1
-Hoogelaandsters-2647-MoanMorn 1
-Hoogelaandsters-2651-MoanMorn 1
-Hoogelaandsters-2655-MoanMorn 1
-Hoogelaandsters-2659-MoanMorn 1
-Hoogelaandsters-2663-MoanMorn 1
-Hoogelaandsters-2667-MoanMorn 1
-Hoogelaandsters-2671-MoanMorn 1
-Hoogelaandsters-2675-MoanMorn 1
-Hoogelaandsters-2680-MoanMorn 1
-Hoogelaandsters-2684-MoanMorn 1
-Hoogelaandsters-2688-MoanMorn 1
-Hoogelaandsters-2692-MoanMorn 1
-Hoogelaandsters-2696-MoanMorn 1
-Hoogelaandsters-2701-MoanMorn 1
-Hoogelaandsters-2705-MoanMorn 1
-Hoogelaandsters-2709-MoanMorn 1
-Hoogelaandsters-2713-MoanMorn 1
-Hoogelaandsters-2717-MoanMorn 1
-Hoogelaandsters-2556-MoanMorn 1
-Hoogelaandsters-2560-MoanMorn 1
-Hoogelaandsters-2566-MoanMorn 1
-Hoogelaandsters-2570-MoanMorn 1
-Hoogelaandsters-2574-MoanMorn 1
-Hoogelaandsters-2578-MoanMorn 1
-Hoogelaandsters-2582-MoanMorn 1
-Hoogelaandsters-2586-MoanMorn 1
-Hoogelaandsters-2590-MoanMorn 1
-Hoogelaandsters-2595-MoanMorn 1
-Hoogelaandsters-2599-MoanMorn 1
-Hoogelaandsters-2603-MoanMorn 1
-Hoogelaandsters-2607-MoanMorn 1
-Hoogelaandsters-2611-MoanMorn 1
-Hoogelaandsters-2615-MoanMorn 1
-Hoogelaandsters-2619-MoanMorn 1
-Hoogelaandsters-2623-MoanMorn 1
-Hoogelaandsters-2627-MoanMorn 1
-Hoogelaandsters-2632-MoanMorn 1
-Hoogelaandsters-2636-MoanMorn 1
-Hoogelaandsters-2640-MoanMorn 1
-Hoogelaandsters-2644-MoanMorn 1
-Hoogelaandsters-2648-MoanMorn 1
-Hoogelaandsters-2652-MoanMorn 1
-Hoogelaandsters-2656-MoanMorn 1
-Hoogelaandsters-2660-MoanMorn 1
-Hoogelaandsters-2664-MoanMorn 1
-Hoogelaandsters-2668-MoanMorn 1
-Hoogelaandsters-2672-MoanMorn 1
-Hoogelaandsters-2677-MoanMorn 1
-Hoogelaandsters-2681-MoanMorn 1
-Hoogelaandsters-2685-MoanMorn 1
-Hoogelaandsters-2689-MoanMorn 1
-Hoogelaandsters-2693-MoanMorn 1
-Hoogelaandsters-2698-MoanMorn 1
-Hoogelaandsters-2702-MoanMorn 1
-Hoogelaandsters-2706-MoanMorn 1
-Hoogelaandsters-2710-MoanMorn 1
-Hoogelaandsters-2714-MoanMorn 1
-Hoogelaandsters-2718-MoanMorn 1
-Hoogelaandsters-2719-MoanMorn 1
-Hoogelaandsters-2723-MoanMorn 1
-Hoogelaandsters-2727-MoanMorn 1
-Hoogelaandsters-2731-MoanMorn 1
-Hoogelaandsters-2735-MoanMorn 1
-Hoogelaandsters-2740-MoanMorn 1
-Hoogelaandsters-2744-MoanMorn 1
-Hoogelaandsters-2748-MoanMorn 1
-Hoogelaandsters-2752-MoanMorn 1
-Hoogelaandsters-2756-MoanMorn 1
-Hoogelaandsters-2760-MoanMorn 1
-Hoogelaandsters-2765-MoanMorn 1
-Hoogelaandsters-2769-MoanMorn 1
-Hoogelaandsters-2773-MoanMorn 1
-Hoogelaandsters-2777-MoanMorn 1
-Hoogelaandsters-2781-MoanMorn 1
-Hoogelaandsters-2785-MoanMorn 1
-Hoogelaandsters-2789-MoanMorn 1
-Hoogelaandsters-2793-MoanMorn 1
-Hoogelaandsters-2797-MoanMorn 1
-Hoogelaandsters-2802-MoanMorn 1
-Hoogelaandsters-2810-MoanMorn 1
-Hoogelaandsters-2814-MoanMorn 1
-Hoogelaandsters-2720-MoanMorn 1
-Hoogelaandsters-2724-MoanMorn 1
-Hoogelaandsters-2728-MoanMorn 1
-Hoogelaandsters-2732-MoanMorn 1
-Hoogelaandsters-2736-MoanMorn 1
-Hoogelaandsters-2741-MoanMorn 1
-Hoogelaandsters-2745-MoanMorn 1
-Hoogelaandsters-2749-MoanMorn 1
-Hoogelaandsters-2753-MoanMorn 1
-Hoogelaandsters-2757-MoanMorn 1
-Hoogelaandsters-2761-MoanMorn 1
-Hoogelaandsters-2766-MoanMorn 1
-Hoogelaandsters-2770-MoanMorn 1
-Hoogelaandsters-2774-MoanMorn 1
-Hoogelaandsters-2778-MoanMorn 1
-Hoogelaandsters-2782-MoanMorn 1
-Hoogelaandsters-2786-MoanMorn 1
-Hoogelaandsters-2790-MoanMorn 1
-Hoogelaandsters-2794-MoanMorn 1
-Hoogelaandsters-2799-MoanMorn 1
-Hoogelaandsters-2807-MoanMorn 1
-Hoogelaandsters-2811-MoanMorn 1
-Hoogelaandsters-2721-MoanMorn 1
-Hoogelaandsters-2725-MoanMorn 1
-Hoogelaandsters-2729-MoanMorn 1
-Hoogelaandsters-2733-MoanMorn 1
-Hoogelaandsters-2737-MoanMorn 1
-Hoogelaandsters-2742-MoanMorn 1
-Hoogelaandsters-2746-MoanMorn 1
-Hoogelaandsters-2750-MoanMorn 1
-Hoogelaandsters-2754-MoanMorn 1
-Hoogelaandsters-2758-MoanMorn 1
-Hoogelaandsters-2762-MoanMorn 1
-Hoogelaandsters-2767-MoanMorn 1
-Hoogelaandsters-2771-MoanMorn 1
-Hoogelaandsters-2775-MoanMorn 1
-Hoogelaandsters-2779-MoanMorn 1
-Hoogelaandsters-2783-MoanMorn 1
-Hoogelaandsters-2787-MoanMorn 1
-Hoogelaandsters-2791-MoanMorn 1
-Hoogelaandsters-2795-MoanMorn 1
-Hoogelaandsters-2800-MoanMorn 1
-Hoogelaandsters-2808-MoanMorn 1
-Hoogelaandsters-2812-MoanMorn 1
-Hoogelaandsters-2722-MoanMorn 1
-Hoogelaandsters-2726-MoanMorn 1
-Hoogelaandsters-2730-MoanMorn 1
-Hoogelaandsters-2734-MoanMorn 1
-Hoogelaandsters-2738-MoanMorn 1
-Hoogelaandsters-2743-MoanMorn 1
-Hoogelaandsters-2747-MoanMorn 1
-Hoogelaandsters-2751-MoanMorn 1
-Hoogelaandsters-2755-MoanMorn 1
-Hoogelaandsters-2759-MoanMorn 1
-Hoogelaandsters-2763-MoanMorn 1
-Hoogelaandsters-2768-MoanMorn 1
-Hoogelaandsters-2772-MoanMorn 1
-Hoogelaandsters-2776-MoanMorn 1
-Hoogelaandsters-2780-MoanMorn 1
-Hoogelaandsters-2784-MoanMorn 1
-Hoogelaandsters-2788-MoanMorn 1
-Hoogelaandsters-2792-MoanMorn 1
-Hoogelaandsters-2796-MoanMorn 1
-Hoogelaandsters-2801-MoanMorn 1
-Hoogelaandsters-2809-MoanMorn 1
-Hoogelaandsters-2813-MoanMorn 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/speech_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2553-MoanMorn 102968
-Hoogelaandsters-2557-MoanMorn 132784
-Hoogelaandsters-2562-MoanMorn 45382
-Hoogelaandsters-2567-MoanMorn 179756
-Hoogelaandsters-2571-MoanMorn 154171
-Hoogelaandsters-2575-MoanMorn 226928
-Hoogelaandsters-2579-MoanMorn 396289
-Hoogelaandsters-2583-MoanMorn 57484
-Hoogelaandsters-2587-MoanMorn 63091
-Hoogelaandsters-2592-MoanMorn 281058
-Hoogelaandsters-2596-MoanMorn 399175
-Hoogelaandsters-2600-MoanMorn 339081
-Hoogelaandsters-2604-MoanMorn 98893
-Hoogelaandsters-2608-MoanMorn 65152
-Hoogelaandsters-2612-MoanMorn 57739
-Hoogelaandsters-2616-MoanMorn 143051
-Hoogelaandsters-2620-MoanMorn 151777
-Hoogelaandsters-2624-MoanMorn 186646
-Hoogelaandsters-2628-MoanMorn 159049
-Hoogelaandsters-2633-MoanMorn 205382
-Hoogelaandsters-2637-MoanMorn 95216
-Hoogelaandsters-2641-MoanMorn 194325
-Hoogelaandsters-2645-MoanMorn 105721
-Hoogelaandsters-2649-MoanMorn 329583
-Hoogelaandsters-2653-MoanMorn 83826
-Hoogelaandsters-2657-MoanMorn 64112
-Hoogelaandsters-2661-MoanMorn 111091
-Hoogelaandsters-2665-MoanMorn 40805
-Hoogelaandsters-2669-MoanMorn 179279
-Hoogelaandsters-2673-MoanMorn 102519
-Hoogelaandsters-2678-MoanMorn 104510
-Hoogelaandsters-2682-MoanMorn 36691
-Hoogelaandsters-2686-MoanMorn 381632
-Hoogelaandsters-2690-MoanMorn 121206
-Hoogelaandsters-2694-MoanMorn 112389
-Hoogelaandsters-2699-MoanMorn 65936
-Hoogelaandsters-2703-MoanMorn 121981
-Hoogelaandsters-2707-MoanMorn 141227
-Hoogelaandsters-2711-MoanMorn 165911
-Hoogelaandsters-2715-MoanMorn 344118
-Hoogelaandsters-2554-MoanMorn 110765
-Hoogelaandsters-2558-MoanMorn 82847
-Hoogelaandsters-2564-MoanMorn 59270
-Hoogelaandsters-2568-MoanMorn 189950
-Hoogelaandsters-2572-MoanMorn 147342
-Hoogelaandsters-2576-MoanMorn 220274
-Hoogelaandsters-2580-MoanMorn 354925
-Hoogelaandsters-2584-MoanMorn 142341
-Hoogelaandsters-2588-MoanMorn 127062
-Hoogelaandsters-2593-MoanMorn 137165
-Hoogelaandsters-2597-MoanMorn 396984
-Hoogelaandsters-2601-MoanMorn 89865
-Hoogelaandsters-2605-MoanMorn 70560
-Hoogelaandsters-2609-MoanMorn 216938
-Hoogelaandsters-2613-MoanMorn 140629
-Hoogelaandsters-2617-MoanMorn 120813
-Hoogelaandsters-2621-MoanMorn 92039
-Hoogelaandsters-2625-MoanMorn 81238
-Hoogelaandsters-2629-MoanMorn 64944
-Hoogelaandsters-2634-MoanMorn 111251
-Hoogelaandsters-2638-MoanMorn 167512
-Hoogelaandsters-2642-MoanMorn 367904
-Hoogelaandsters-2646-MoanMorn 46324
-Hoogelaandsters-2650-MoanMorn 79285
-Hoogelaandsters-2654-MoanMorn 47784
-Hoogelaandsters-2658-MoanMorn 297483
-Hoogelaandsters-2662-MoanMorn 37814
-Hoogelaandsters-2666-MoanMorn 162690
-Hoogelaandsters-2670-MoanMorn 103191
-Hoogelaandsters-2674-MoanMorn 92654
-Hoogelaandsters-2679-MoanMorn 197560
-Hoogelaandsters-2683-MoanMorn 87616
-Hoogelaandsters-2687-MoanMorn 100706
-Hoogelaandsters-2691-MoanMorn 102971
-Hoogelaandsters-2695-MoanMorn 52497
-Hoogelaandsters-2700-MoanMorn 160658
-Hoogelaandsters-2704-MoanMorn 129668
-Hoogelaandsters-2708-MoanMorn 128206
-Hoogelaandsters-2712-MoanMorn 88044
-Hoogelaandsters-2716-MoanMorn 60322
-Hoogelaandsters-2555-MoanMorn 287540
-Hoogelaandsters-2559-MoanMorn 330822
-Hoogelaandsters-2565-MoanMorn 258697
-Hoogelaandsters-2569-MoanMorn 122475
-Hoogelaandsters-2573-MoanMorn 280673
-Hoogelaandsters-2577-MoanMorn 144473
-Hoogelaandsters-2581-MoanMorn 141996
-Hoogelaandsters-2585-MoanMorn 212081
-Hoogelaandsters-2589-MoanMorn 185519
-Hoogelaandsters-2594-MoanMorn 50803
-Hoogelaandsters-2598-MoanMorn 149919
-Hoogelaandsters-2602-MoanMorn 38852
-Hoogelaandsters-2606-MoanMorn 120520
-Hoogelaandsters-2610-MoanMorn 154787
-Hoogelaandsters-2614-MoanMorn 68877
-Hoogelaandsters-2618-MoanMorn 152273
-Hoogelaandsters-2622-MoanMorn 110079
-Hoogelaandsters-2626-MoanMorn 82118
-Hoogelaandsters-2631-MoanMorn 197675
-Hoogelaandsters-2635-MoanMorn 165061
-Hoogelaandsters-2639-MoanMorn 169140
-Hoogelaandsters-2643-MoanMorn 38432
-Hoogelaandsters-2647-MoanMorn 230497
-Hoogelaandsters-2651-MoanMorn 155389
-Hoogelaandsters-2655-MoanMorn 265827
-Hoogelaandsters-2659-MoanMorn 156964
-Hoogelaandsters-2663-MoanMorn 118006
-Hoogelaandsters-2667-MoanMorn 180077
-Hoogelaandsters-2671-MoanMorn 187349
-Hoogelaandsters-2675-MoanMorn 100155
-Hoogelaandsters-2680-MoanMorn 108959
-Hoogelaandsters-2684-MoanMorn 129830
-Hoogelaandsters-2688-MoanMorn 87494
-Hoogelaandsters-2692-MoanMorn 139736
-Hoogelaandsters-2696-MoanMorn 122070
-Hoogelaandsters-2701-MoanMorn 319907
-Hoogelaandsters-2705-MoanMorn 89592
-Hoogelaandsters-2709-MoanMorn 81790
-Hoogelaandsters-2713-MoanMorn 62093
-Hoogelaandsters-2717-MoanMorn 105123
-Hoogelaandsters-2556-MoanMorn 115423
-Hoogelaandsters-2560-MoanMorn 82091
-Hoogelaandsters-2566-MoanMorn 163000
-Hoogelaandsters-2570-MoanMorn 274954
-Hoogelaandsters-2574-MoanMorn 158271
-Hoogelaandsters-2578-MoanMorn 98639
-Hoogelaandsters-2582-MoanMorn 108427
-Hoogelaandsters-2586-MoanMorn 130138
-Hoogelaandsters-2590-MoanMorn 210595
-Hoogelaandsters-2595-MoanMorn 385675
-Hoogelaandsters-2599-MoanMorn 159944
-Hoogelaandsters-2603-MoanMorn 71226
-Hoogelaandsters-2607-MoanMorn 154605
-Hoogelaandsters-2611-MoanMorn 306866
-Hoogelaandsters-2615-MoanMorn 103504
-Hoogelaandsters-2619-MoanMorn 119682
-Hoogelaandsters-2623-MoanMorn 195466
-Hoogelaandsters-2627-MoanMorn 81931
-Hoogelaandsters-2632-MoanMorn 230483
-Hoogelaandsters-2636-MoanMorn 175105
-Hoogelaandsters-2640-MoanMorn 52143
-Hoogelaandsters-2644-MoanMorn 81376
-Hoogelaandsters-2648-MoanMorn 311615
-Hoogelaandsters-2652-MoanMorn 66501
-Hoogelaandsters-2656-MoanMorn 110732
-Hoogelaandsters-2660-MoanMorn 218546
-Hoogelaandsters-2664-MoanMorn 434104
-Hoogelaandsters-2668-MoanMorn 234849
-Hoogelaandsters-2672-MoanMorn 67330
-Hoogelaandsters-2677-MoanMorn 217662
-Hoogelaandsters-2681-MoanMorn 96112
-Hoogelaandsters-2685-MoanMorn 242708
-Hoogelaandsters-2689-MoanMorn 114105
-Hoogelaandsters-2693-MoanMorn 173509
-Hoogelaandsters-2698-MoanMorn 255138
-Hoogelaandsters-2702-MoanMorn 182133
-Hoogelaandsters-2706-MoanMorn 92569
-Hoogelaandsters-2710-MoanMorn 52726
-Hoogelaandsters-2714-MoanMorn 119379
-Hoogelaandsters-2718-MoanMorn 249613
-Hoogelaandsters-2719-MoanMorn 207030
-Hoogelaandsters-2723-MoanMorn 164027
-Hoogelaandsters-2727-MoanMorn 131255
-Hoogelaandsters-2731-MoanMorn 138091
-Hoogelaandsters-2735-MoanMorn 131698
-Hoogelaandsters-2740-MoanMorn 261244
-Hoogelaandsters-2744-MoanMorn 219641
-Hoogelaandsters-2748-MoanMorn 357717
-Hoogelaandsters-2752-MoanMorn 167116
-Hoogelaandsters-2756-MoanMorn 129072
-Hoogelaandsters-2760-MoanMorn 69985
-Hoogelaandsters-2765-MoanMorn 365140
-Hoogelaandsters-2769-MoanMorn 171180
-Hoogelaandsters-2773-MoanMorn 145536
-Hoogelaandsters-2777-MoanMorn 303805
-Hoogelaandsters-2781-MoanMorn 341668
-Hoogelaandsters-2785-MoanMorn 208293
-Hoogelaandsters-2789-MoanMorn 225033
-Hoogelaandsters-2793-MoanMorn 189487
-Hoogelaandsters-2797-MoanMorn 243739
-Hoogelaandsters-2802-MoanMorn 95392
-Hoogelaandsters-2810-MoanMorn 158738
-Hoogelaandsters-2814-MoanMorn 120438
-Hoogelaandsters-2720-MoanMorn 85956
-Hoogelaandsters-2724-MoanMorn 170079
-Hoogelaandsters-2728-MoanMorn 299268
-Hoogelaandsters-2732-MoanMorn 285525
-Hoogelaandsters-2736-MoanMorn 110971
-Hoogelaandsters-2741-MoanMorn 83711
-Hoogelaandsters-2745-MoanMorn 144678
-Hoogelaandsters-2749-MoanMorn 67738
-Hoogelaandsters-2753-MoanMorn 196508
-Hoogelaandsters-2757-MoanMorn 215266
-Hoogelaandsters-2761-MoanMorn 75337
-Hoogelaandsters-2766-MoanMorn 64602
-Hoogelaandsters-2770-MoanMorn 395729
-Hoogelaandsters-2774-MoanMorn 49597
-Hoogelaandsters-2778-MoanMorn 131446
-Hoogelaandsters-2782-MoanMorn 61127
-Hoogelaandsters-2786-MoanMorn 254529
-Hoogelaandsters-2790-MoanMorn 126016
-Hoogelaandsters-2794-MoanMorn 94688
-Hoogelaandsters-2799-MoanMorn 30315
-Hoogelaandsters-2807-MoanMorn 40291
-Hoogelaandsters-2811-MoanMorn 103372
-Hoogelaandsters-2721-MoanMorn 70010
-Hoogelaandsters-2725-MoanMorn 142302
-Hoogelaandsters-2729-MoanMorn 218066
-Hoogelaandsters-2733-MoanMorn 113371
-Hoogelaandsters-2737-MoanMorn 51787
-Hoogelaandsters-2742-MoanMorn 115021
-Hoogelaandsters-2746-MoanMorn 102278
-Hoogelaandsters-2750-MoanMorn 113960
-Hoogelaandsters-2754-MoanMorn 127820
-Hoogelaandsters-2758-MoanMorn 132424
-Hoogelaandsters-2762-MoanMorn 267801
-Hoogelaandsters-2767-MoanMorn 175120
-Hoogelaandsters-2771-MoanMorn 190916
-Hoogelaandsters-2775-MoanMorn 220794
-Hoogelaandsters-2779-MoanMorn 53626
-Hoogelaandsters-2783-MoanMorn 90153
-Hoogelaandsters-2787-MoanMorn 130704
-Hoogelaandsters-2791-MoanMorn 47366
-Hoogelaandsters-2795-MoanMorn 255622
-Hoogelaandsters-2800-MoanMorn 81850
-Hoogelaandsters-2808-MoanMorn 80081
-Hoogelaandsters-2812-MoanMorn 73333
-Hoogelaandsters-2722-MoanMorn 199048
-Hoogelaandsters-2726-MoanMorn 68327
-Hoogelaandsters-2730-MoanMorn 103049
-Hoogelaandsters-2734-MoanMorn 55949
-Hoogelaandsters-2738-MoanMorn 168007
-Hoogelaandsters-2743-MoanMorn 101916
-Hoogelaandsters-2747-MoanMorn 235788
-Hoogelaandsters-2751-MoanMorn 63234
-Hoogelaandsters-2755-MoanMorn 268211
-Hoogelaandsters-2759-MoanMorn 121221
-Hoogelaandsters-2763-MoanMorn 300360
-Hoogelaandsters-2768-MoanMorn 96182
-Hoogelaandsters-2772-MoanMorn 131793
-Hoogelaandsters-2776-MoanMorn 179261
-Hoogelaandsters-2780-MoanMorn 306314
-Hoogelaandsters-2784-MoanMorn 347307
-Hoogelaandsters-2788-MoanMorn 136168
-Hoogelaandsters-2792-MoanMorn 168299
-Hoogelaandsters-2796-MoanMorn 124545
-Hoogelaandsters-2801-MoanMorn 102298
-Hoogelaandsters-2809-MoanMorn 168456
-Hoogelaandsters-2813-MoanMorn 101606

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/train/text_shape

@@ -1,249 +0,0 @@
-Hoogelaandsters-2553-MoanMorn 58
-Hoogelaandsters-2557-MoanMorn 77
-Hoogelaandsters-2562-MoanMorn 23
-Hoogelaandsters-2567-MoanMorn 106
-Hoogelaandsters-2571-MoanMorn 85
-Hoogelaandsters-2575-MoanMorn 124
-Hoogelaandsters-2579-MoanMorn 244
-Hoogelaandsters-2583-MoanMorn 34
-Hoogelaandsters-2587-MoanMorn 25
-Hoogelaandsters-2592-MoanMorn 176
-Hoogelaandsters-2596-MoanMorn 254
-Hoogelaandsters-2600-MoanMorn 230
-Hoogelaandsters-2604-MoanMorn 43
-Hoogelaandsters-2608-MoanMorn 22
-Hoogelaandsters-2612-MoanMorn 24
-Hoogelaandsters-2616-MoanMorn 80
-Hoogelaandsters-2620-MoanMorn 101
-Hoogelaandsters-2624-MoanMorn 108
-Hoogelaandsters-2628-MoanMorn 103
-Hoogelaandsters-2633-MoanMorn 147
-Hoogelaandsters-2637-MoanMorn 56
-Hoogelaandsters-2641-MoanMorn 116
-Hoogelaandsters-2645-MoanMorn 62
-Hoogelaandsters-2649-MoanMorn 200
-Hoogelaandsters-2653-MoanMorn 53
-Hoogelaandsters-2657-MoanMorn 20
-Hoogelaandsters-2661-MoanMorn 58
-Hoogelaandsters-2665-MoanMorn 27
-Hoogelaandsters-2669-MoanMorn 122
-Hoogelaandsters-2673-MoanMorn 64
-Hoogelaandsters-2678-MoanMorn 73
-Hoogelaandsters-2682-MoanMorn 20
-Hoogelaandsters-2686-MoanMorn 244
-Hoogelaandsters-2690-MoanMorn 66
-Hoogelaandsters-2694-MoanMorn 79
-Hoogelaandsters-2699-MoanMorn 38
-Hoogelaandsters-2703-MoanMorn 62
-Hoogelaandsters-2707-MoanMorn 74
-Hoogelaandsters-2711-MoanMorn 73
-Hoogelaandsters-2715-MoanMorn 193
-Hoogelaandsters-2554-MoanMorn 57
-Hoogelaandsters-2558-MoanMorn 46
-Hoogelaandsters-2564-MoanMorn 29
-Hoogelaandsters-2568-MoanMorn 107
-Hoogelaandsters-2572-MoanMorn 78
-Hoogelaandsters-2576-MoanMorn 137
-Hoogelaandsters-2580-MoanMorn 227
-Hoogelaandsters-2584-MoanMorn 84
-Hoogelaandsters-2588-MoanMorn 96
-Hoogelaandsters-2593-MoanMorn 97
-Hoogelaandsters-2597-MoanMorn 252
-Hoogelaandsters-2601-MoanMorn 51
-Hoogelaandsters-2605-MoanMorn 46
-Hoogelaandsters-2609-MoanMorn 165
-Hoogelaandsters-2613-MoanMorn 89
-Hoogelaandsters-2617-MoanMorn 66
-Hoogelaandsters-2621-MoanMorn 53
-Hoogelaandsters-2625-MoanMorn 48
-Hoogelaandsters-2629-MoanMorn 39
-Hoogelaandsters-2634-MoanMorn 65
-Hoogelaandsters-2638-MoanMorn 78
-Hoogelaandsters-2642-MoanMorn 219
-Hoogelaandsters-2646-MoanMorn 31
-Hoogelaandsters-2650-MoanMorn 23
-Hoogelaandsters-2654-MoanMorn 25
-Hoogelaandsters-2658-MoanMorn 183
-Hoogelaandsters-2662-MoanMorn 15
-Hoogelaandsters-2666-MoanMorn 82
-Hoogelaandsters-2670-MoanMorn 54
-Hoogelaandsters-2674-MoanMorn 45
-Hoogelaandsters-2679-MoanMorn 115
-Hoogelaandsters-2683-MoanMorn 48
-Hoogelaandsters-2687-MoanMorn 52
-Hoogelaandsters-2691-MoanMorn 66
-Hoogelaandsters-2695-MoanMorn 32
-Hoogelaandsters-2700-MoanMorn 90
-Hoogelaandsters-2704-MoanMorn 86
-Hoogelaandsters-2708-MoanMorn 85
-Hoogelaandsters-2712-MoanMorn 52
-Hoogelaandsters-2716-MoanMorn 34
-Hoogelaandsters-2555-MoanMorn 158
-Hoogelaandsters-2559-MoanMorn 213
-Hoogelaandsters-2565-MoanMorn 154
-Hoogelaandsters-2569-MoanMorn 70
-Hoogelaandsters-2573-MoanMorn 159
-Hoogelaandsters-2577-MoanMorn 77
-Hoogelaandsters-2581-MoanMorn 82
-Hoogelaandsters-2585-MoanMorn 123
-Hoogelaandsters-2589-MoanMorn 125
-Hoogelaandsters-2594-MoanMorn 36
-Hoogelaandsters-2598-MoanMorn 97
-Hoogelaandsters-2602-MoanMorn 15
-Hoogelaandsters-2606-MoanMorn 79
-Hoogelaandsters-2610-MoanMorn 86
-Hoogelaandsters-2614-MoanMorn 35
-Hoogelaandsters-2618-MoanMorn 98
-Hoogelaandsters-2622-MoanMorn 61
-Hoogelaandsters-2626-MoanMorn 57
-Hoogelaandsters-2631-MoanMorn 127
-Hoogelaandsters-2635-MoanMorn 114
-Hoogelaandsters-2639-MoanMorn 85
-Hoogelaandsters-2643-MoanMorn 16
-Hoogelaandsters-2647-MoanMorn 139
-Hoogelaandsters-2651-MoanMorn 87
-Hoogelaandsters-2655-MoanMorn 152
-Hoogelaandsters-2659-MoanMorn 103
-Hoogelaandsters-2663-MoanMorn 60
-Hoogelaandsters-2667-MoanMorn 128
-Hoogelaandsters-2671-MoanMorn 104
-Hoogelaandsters-2675-MoanMorn 52
-Hoogelaandsters-2680-MoanMorn 69
-Hoogelaandsters-2684-MoanMorn 94
-Hoogelaandsters-2688-MoanMorn 53
-Hoogelaandsters-2692-MoanMorn 75
-Hoogelaandsters-2696-MoanMorn 64
-Hoogelaandsters-2701-MoanMorn 184
-Hoogelaandsters-2705-MoanMorn 58
-Hoogelaandsters-2709-MoanMorn 49
-Hoogelaandsters-2713-MoanMorn 36
-Hoogelaandsters-2717-MoanMorn 60
-Hoogelaandsters-2556-MoanMorn 61
-Hoogelaandsters-2560-MoanMorn 40
-Hoogelaandsters-2566-MoanMorn 91
-Hoogelaandsters-2570-MoanMorn 138
-Hoogelaandsters-2574-MoanMorn 85
-Hoogelaandsters-2578-MoanMorn 53
-Hoogelaandsters-2582-MoanMorn 74
-Hoogelaandsters-2586-MoanMorn 67
-Hoogelaandsters-2590-MoanMorn 151
-Hoogelaandsters-2595-MoanMorn 267
-Hoogelaandsters-2599-MoanMorn 104
-Hoogelaandsters-2603-MoanMorn 39
-Hoogelaandsters-2607-MoanMorn 86
-Hoogelaandsters-2611-MoanMorn 193
-Hoogelaandsters-2615-MoanMorn 73
-Hoogelaandsters-2619-MoanMorn 78
-Hoogelaandsters-2623-MoanMorn 113
-Hoogelaandsters-2627-MoanMorn 40
-Hoogelaandsters-2632-MoanMorn 153
-Hoogelaandsters-2636-MoanMorn 123
-Hoogelaandsters-2640-MoanMorn 32
-Hoogelaandsters-2644-MoanMorn 48
-Hoogelaandsters-2648-MoanMorn 171
-Hoogelaandsters-2652-MoanMorn 27
-Hoogelaandsters-2656-MoanMorn 80
-Hoogelaandsters-2660-MoanMorn 126
-Hoogelaandsters-2664-MoanMorn 282
-Hoogelaandsters-2668-MoanMorn 143
-Hoogelaandsters-2672-MoanMorn 35
-Hoogelaandsters-2677-MoanMorn 150
-Hoogelaandsters-2681-MoanMorn 72
-Hoogelaandsters-2685-MoanMorn 142
-Hoogelaandsters-2689-MoanMorn 76
-Hoogelaandsters-2693-MoanMorn 115
-Hoogelaandsters-2698-MoanMorn 145
-Hoogelaandsters-2702-MoanMorn 108
-Hoogelaandsters-2706-MoanMorn 56
-Hoogelaandsters-2710-MoanMorn 34
-Hoogelaandsters-2714-MoanMorn 63
-Hoogelaandsters-2718-MoanMorn 164
-Hoogelaandsters-2719-MoanMorn 129
-Hoogelaandsters-2723-MoanMorn 125
-Hoogelaandsters-2727-MoanMorn 93
-Hoogelaandsters-2731-MoanMorn 89
-Hoogelaandsters-2735-MoanMorn 86
-Hoogelaandsters-2740-MoanMorn 151
-Hoogelaandsters-2744-MoanMorn 153
-Hoogelaandsters-2748-MoanMorn 197
-Hoogelaandsters-2752-MoanMorn 102
-Hoogelaandsters-2756-MoanMorn 75
-Hoogelaandsters-2760-MoanMorn 33
-Hoogelaandsters-2765-MoanMorn 257
-Hoogelaandsters-2769-MoanMorn 86
-Hoogelaandsters-2773-MoanMorn 85
-Hoogelaandsters-2777-MoanMorn 169
-Hoogelaandsters-2781-MoanMorn 244
-Hoogelaandsters-2785-MoanMorn 133
-Hoogelaandsters-2789-MoanMorn 132
-Hoogelaandsters-2793-MoanMorn 94
-Hoogelaandsters-2797-MoanMorn 151
-Hoogelaandsters-2802-MoanMorn 45
-Hoogelaandsters-2810-MoanMorn 105
-Hoogelaandsters-2814-MoanMorn 61
-Hoogelaandsters-2720-MoanMorn 54
-Hoogelaandsters-2724-MoanMorn 95
-Hoogelaandsters-2728-MoanMorn 195
-Hoogelaandsters-2732-MoanMorn 171
-Hoogelaandsters-2736-MoanMorn 64
-Hoogelaandsters-2741-MoanMorn 52
-Hoogelaandsters-2745-MoanMorn 89
-Hoogelaandsters-2749-MoanMorn 37
-Hoogelaandsters-2753-MoanMorn 108
-Hoogelaandsters-2757-MoanMorn 118
-Hoogelaandsters-2761-MoanMorn 44
-Hoogelaandsters-2766-MoanMorn 32
-Hoogelaandsters-2770-MoanMorn 252
-Hoogelaandsters-2774-MoanMorn 26
-Hoogelaandsters-2778-MoanMorn 81
-Hoogelaandsters-2782-MoanMorn 29
-Hoogelaandsters-2786-MoanMorn 144
-Hoogelaandsters-2790-MoanMorn 74
-Hoogelaandsters-2794-MoanMorn 46
-Hoogelaandsters-2799-MoanMorn 17
-Hoogelaandsters-2807-MoanMorn 24
-Hoogelaandsters-2811-MoanMorn 60
-Hoogelaandsters-2721-MoanMorn 43
-Hoogelaandsters-2725-MoanMorn 112
-Hoogelaandsters-2729-MoanMorn 143
-Hoogelaandsters-2733-MoanMorn 84
-Hoogelaandsters-2737-MoanMorn 36
-Hoogelaandsters-2742-MoanMorn 64
-Hoogelaandsters-2746-MoanMorn 65
-Hoogelaandsters-2750-MoanMorn 59
-Hoogelaandsters-2754-MoanMorn 74
-Hoogelaandsters-2758-MoanMorn 72
-Hoogelaandsters-2762-MoanMorn 160
-Hoogelaandsters-2767-MoanMorn 108
-Hoogelaandsters-2771-MoanMorn 102
-Hoogelaandsters-2775-MoanMorn 126
-Hoogelaandsters-2779-MoanMorn 35
-Hoogelaandsters-2783-MoanMorn 48
-Hoogelaandsters-2787-MoanMorn 83
-Hoogelaandsters-2791-MoanMorn 21
-Hoogelaandsters-2795-MoanMorn 178
-Hoogelaandsters-2800-MoanMorn 43
-Hoogelaandsters-2808-MoanMorn 49
-Hoogelaandsters-2812-MoanMorn 40
-Hoogelaandsters-2722-MoanMorn 123
-Hoogelaandsters-2726-MoanMorn 46
-Hoogelaandsters-2730-MoanMorn 65
-Hoogelaandsters-2734-MoanMorn 29
-Hoogelaandsters-2738-MoanMorn 102
-Hoogelaandsters-2743-MoanMorn 63
-Hoogelaandsters-2747-MoanMorn 140
-Hoogelaandsters-2751-MoanMorn 34
-Hoogelaandsters-2755-MoanMorn 170
-Hoogelaandsters-2759-MoanMorn 89
-Hoogelaandsters-2763-MoanMorn 170
-Hoogelaandsters-2768-MoanMorn 57
-Hoogelaandsters-2772-MoanMorn 82
-Hoogelaandsters-2776-MoanMorn 108
-Hoogelaandsters-2780-MoanMorn 203
-Hoogelaandsters-2784-MoanMorn 219
-Hoogelaandsters-2788-MoanMorn 86
-Hoogelaandsters-2792-MoanMorn 115
-Hoogelaandsters-2796-MoanMorn 72
-Hoogelaandsters-2801-MoanMorn 75
-Hoogelaandsters-2809-MoanMorn 82
-Hoogelaandsters-2813-MoanMorn 56

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/batch_keys

@@ -1,3 +0,0 @@
-text
-speech
-sids

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/feats_lengths_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:d91f08590d0800f915c76c97f826494001a888019e15e8cb5906b40df85e42f5
-size 778

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/feats_stats.npz

@@ -1,3 +0,0 @@
-version https://git-lfs.github.com/spec/v1
-oid sha256:2bcc2334a15eb36d64c9c834c251b257c11f2f3609d48e3b645a451272efd35f
-size 1402

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/sids_shape

@@ -1,5 +0,0 @@
-Oldambsters-0001-AigenOardegheden 1
-Oldambsters-0215-AigenOardegheden 1
-Oldambsters-0054-AigenOardegheden 1
-Oldambsters-0106-AigenOardegheden 1
-Oldambsters-0160-AigenOardegheden 1

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/speech_shape

@@ -1,5 +0,0 @@
-Oldambsters-0001-AigenOardegheden 115718
-Oldambsters-0215-AigenOardegheden 156715
-Oldambsters-0054-AigenOardegheden 129824
-Oldambsters-0106-AigenOardegheden 70560
-Oldambsters-0160-AigenOardegheden 50803

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/stats_keys

@@ -1,2 +0,0 @@
-feats
-feats_lengths

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.11/valid/text_shape

@@ -1,5 +0,0 @@
-Oldambsters-0001-AigenOardegheden 89
-Oldambsters-0215-AigenOardegheden 110
-Oldambsters-0054-AigenOardegheden 7
-Oldambsters-0106-AigenOardegheden 44
-Oldambsters-0160-AigenOardegheden 34

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12.log

@@ -1,1152 +0,0 @@
-# python3 -m espnet2.bin.gan_tts_train --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.12.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.12.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000 
-# Started at Fri Dec  1 15:58:34 UTC 2023
-#
-/data2/p280965/tts/espnet/tools/venv/bin/python3 /data2/p280965/tts/espnet/espnet2/bin/gan_tts_train.py --collect_stats true --write_collected_feats false --use_preprocessor true --token_type char --token_list dump/token_list/char/tokens.txt --non_linguistic_symbols none --cleaner none --g2p none --normalize none --pitch_normalize none --energy_normalize none --train_data_path_and_name_and_type dump/raw/train_nodev/text,text,text --train_data_path_and_name_and_type dump/raw/train_nodev/wav.scp,speech,sound --valid_data_path_and_name_and_type dump/raw/train_dev/text,text,text --valid_data_path_and_name_and_type dump/raw/train_dev/wav.scp,speech,sound --train_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.12.scp --valid_shape_file exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.12.scp --output_dir exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12 --config conf/train_vits.yaml --feats_extract fbank --feats_extract_conf n_fft=1024 --feats_extract_conf hop_length=256 --feats_extract_conf win_length=null --feats_extract_conf fs=22050 --feats_extract_conf fmin=80 --feats_extract_conf fmax=7600 --feats_extract_conf n_mels=80 --pitch_extract_conf fs=22050 --pitch_extract_conf n_fft=1024 --pitch_extract_conf hop_length=256 --pitch_extract_conf f0max=400 --pitch_extract_conf f0min=80 --energy_extract_conf fs=22050 --energy_extract_conf n_fft=1024 --energy_extract_conf hop_length=256 --energy_extract_conf win_length=null --train_data_path_and_name_and_type dump/raw/train_nodev/utt2sid,sids,text_int --valid_data_path_and_name_and_type dump/raw/train_dev/utt2sid,sids,text_int --use_wandb true --wandb_project GROTTS --wandb_name VITS_lr_3.0e-4 --init_param downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv --batch_size 40 --batch_bins 10000000
-[wieling-3-a100] 2023-12-01 15:58:40,886 (gan_tts:293) INFO: Vocabulary size: 46
-[wieling-3-a100] 2023-12-01 15:58:41,003 (encoder:174) INFO: encoder self-attention layer type = relative self-attention
-/data2/p280965/tts/espnet/tools/venv/lib/python3.9/site-packages/torch/nn/utils/weight_norm.py:30: UserWarning: torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.
-  warnings.warn("torch.nn.utils.weight_norm is deprecated in favor of torch.nn.utils.parametrizations.weight_norm.")
-/data2/p280965/tts/espnet/espnet2/gan_tts/vits/monotonic_align/__init__.py:19: UserWarning: Cython version is not available. Fallback to 'EXPERIMETAL' numba version. If you want to use the cython version, please build it as follows: `cd espnet2/gan_tts/vits/monotonic_align; python setup.py build_ext --inplace`
-  warnings.warn(
-[wieling-3-a100] 2023-12-01 15:58:42,381 (abs_task:1268) INFO: pytorch.version=2.1.0+cu121, cuda.available=True, cudnn.version=8902, cudnn.benchmark=False, cudnn.deterministic=False
-[wieling-3-a100] 2023-12-01 15:58:42,453 (abs_task:1269) INFO: Model structure:
-ESPnetGANTTSModel(
-  (feats_extract): LogMelFbank(
-    (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-    (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=80, fmax=7600, htk=False)
-  )
-  (tts): VITS(
-    (generator): VITSGenerator(
-      (text_encoder): TextEncoder(
-        (emb): Embedding(46, 192)
-        (encoder): Encoder(
-          (embed): Sequential(
-            (0): RelPositionalEncoding(
-              (dropout): Dropout(p=0.0, inplace=False)
-            )
-          )
-          (encoders): MultiSequential(
-            (0): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (1): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (2): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (3): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (4): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-            (5): EncoderLayer(
-              (self_attn): RelPositionMultiHeadedAttention(
-                (linear_q): Linear(in_features=192, out_features=192, bias=True)
-                (linear_k): Linear(in_features=192, out_features=192, bias=True)
-                (linear_v): Linear(in_features=192, out_features=192, bias=True)
-                (linear_out): Linear(in_features=192, out_features=192, bias=True)
-                (dropout): Dropout(p=0.1, inplace=False)
-                (linear_pos): Linear(in_features=192, out_features=192, bias=False)
-              )
-              (feed_forward): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (feed_forward_macaron): MultiLayeredConv1d(
-                (w_1): Conv1d(192, 768, kernel_size=(3,), stride=(1,), padding=(1,))
-                (w_2): Conv1d(768, 192, kernel_size=(3,), stride=(1,), padding=(1,))
-                (dropout): Dropout(p=0.1, inplace=False)
-              )
-              (norm_ff): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_mha): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (norm_ff_macaron): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-              (dropout): Dropout(p=0.1, inplace=False)
-            )
-          )
-          (after_norm): LayerNorm((192,), eps=1e-12, elementwise_affine=True)
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (decoder): HiFiGANGenerator(
-        (input_conv): Conv1d(192, 512, kernel_size=(7,), stride=(1,), padding=(3,))
-        (upsamples): ModuleList(
-          (0): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(512, 256, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (1): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(256, 128, kernel_size=(16,), stride=(8,), padding=(4,))
-          )
-          (2): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(128, 64, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-          (3): Sequential(
-            (0): LeakyReLU(negative_slope=0.1)
-            (1): ConvTranspose1d(64, 32, kernel_size=(4,), stride=(2,), padding=(1,))
-          )
-        )
-        (blocks): ModuleList(
-          (0): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (1): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (2): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(256, 256, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (3): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (4): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (5): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(128, 128, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (6): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (7): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (8): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(64, 64, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-          (9): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(5,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(3,), stride=(1,), padding=(1,))
-              )
-            )
-          )
-          (10): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(9,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(15,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(7,), stride=(1,), padding=(3,))
-              )
-            )
-          )
-          (11): ResidualBlock(
-            (convs1): ModuleList(
-              (0): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-              (1): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(15,), dilation=(3,))
-              )
-              (2): Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(25,), dilation=(5,))
-              )
-            )
-            (convs2): ModuleList(
-              (0-2): 3 x Sequential(
-                (0): LeakyReLU(negative_slope=0.1)
-                (1): Conv1d(32, 32, kernel_size=(11,), stride=(1,), padding=(5,))
-              )
-            )
-          )
-        )
-        (output_conv): Sequential(
-          (0): LeakyReLU(negative_slope=0.01)
-          (1): Conv1d(32, 1, kernel_size=(7,), stride=(1,), padding=(3,))
-          (2): Tanh()
-        )
-        (global_conv): Conv1d(256, 512, kernel_size=(1,), stride=(1,))
-      )
-      (posterior_encoder): PosteriorEncoder(
-        (input_conv): Conv1d(80, 192, kernel_size=(1,), stride=(1,))
-        (encoder): WaveNet(
-          (conv_layers): ModuleList(
-            (0-15): 16 x ResidualBlock(
-              (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-              (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-              (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-            )
-          )
-        )
-        (proj): Conv1d(192, 384, kernel_size=(1,), stride=(1,))
-      )
-      (flow): ResidualAffineCouplingBlock(
-        (flows): ModuleList(
-          (0): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (1): FlipFlow()
-          (2): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (3): FlipFlow()
-          (4): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (5): FlipFlow()
-          (6): ResidualAffineCouplingLayer(
-            (input_conv): Conv1d(96, 192, kernel_size=(1,), stride=(1,))
-            (encoder): WaveNet(
-              (conv_layers): ModuleList(
-                (0-3): 4 x ResidualBlock(
-                  (conv): Conv1d(192, 384, kernel_size=(5,), stride=(1,), padding=(2,))
-                  (conv1x1_glo): Conv1d1x1(256, 384, kernel_size=(1,), stride=(1,), bias=False)
-                  (conv1x1_out): Conv1d1x1(192, 384, kernel_size=(1,), stride=(1,))
-                )
-              )
-            )
-            (proj): Conv1d(192, 96, kernel_size=(1,), stride=(1,))
-          )
-          (7): FlipFlow()
-        )
-      )
-      (duration_predictor): StochasticDurationPredictor(
-        (pre): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (log_flow): LogFlow()
-        (flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (post_pre): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-        (post_dds): DilatedDepthSeparableConv(
-          (convs): ModuleList(
-            (0): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (1): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-            (2): Sequential(
-              (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-              (1): Transpose()
-              (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (3): Transpose()
-              (4): GELU(approximate='none')
-              (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-              (6): Transpose()
-              (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-              (8): Transpose()
-              (9): GELU(approximate='none')
-              (10): Dropout(p=0.5, inplace=False)
-            )
-          )
-        )
-        (post_proj): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-        (post_flows): ModuleList(
-          (0): ElementwiseAffineFlow()
-          (1): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (2): FlipFlow()
-          (3): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (4): FlipFlow()
-          (5): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (6): FlipFlow()
-          (7): ConvFlow(
-            (input_conv): Conv1d(1, 192, kernel_size=(1,), stride=(1,))
-            (dds_conv): DilatedDepthSeparableConv(
-              (convs): ModuleList(
-                (0): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(1,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (1): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(3,), dilation=(3,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-                (2): Sequential(
-                  (0): Conv1d(192, 192, kernel_size=(3,), stride=(1,), padding=(9,), dilation=(9,), groups=192)
-                  (1): Transpose()
-                  (2): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (3): Transpose()
-                  (4): GELU(approximate='none')
-                  (5): Conv1d(192, 192, kernel_size=(1,), stride=(1,))
-                  (6): Transpose()
-                  (7): LayerNorm((192,), eps=1e-05, elementwise_affine=True)
-                  (8): Transpose()
-                  (9): GELU(approximate='none')
-                  (10): Dropout(p=0.0, inplace=False)
-                )
-              )
-            )
-            (proj): Conv1d(192, 29, kernel_size=(1,), stride=(1,))
-          )
-          (8): FlipFlow()
-        )
-        (global_conv): Conv1d(256, 192, kernel_size=(1,), stride=(1,))
-      )
-      (global_emb): Embedding(4, 256)
-    )
-    (discriminator): HiFiGANMultiScaleMultiPeriodDiscriminator(
-      (msd): HiFiGANMultiScaleDiscriminator(
-        (discriminators): ModuleList(
-          (0): HiFiGANScaleDiscriminator(
-            (layers): ModuleList(
-              (0): Sequential(
-                (0): Conv1d(1, 128, kernel_size=(15,), stride=(1,), padding=(7,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv1d(128, 128, kernel_size=(41,), stride=(2,), padding=(20,), groups=4)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv1d(128, 256, kernel_size=(41,), stride=(2,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv1d(256, 512, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv1d(512, 1024, kernel_size=(41,), stride=(4,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (5): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(41,), stride=(1,), padding=(20,), groups=16)
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (6): Sequential(
-                (0): Conv1d(1024, 1024, kernel_size=(5,), stride=(1,), padding=(2,))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (7): Conv1d(1024, 1, kernel_size=(3,), stride=(1,), padding=(1,))
-            )
-          )
-        )
-      )
-      (mpd): HiFiGANMultiPeriodDiscriminator(
-        (discriminators): ModuleList(
-          (0-4): 5 x HiFiGANPeriodDiscriminator(
-            (convs): ModuleList(
-              (0): Sequential(
-                (0): Conv2d(1, 32, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (1): Sequential(
-                (0): Conv2d(32, 128, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (2): Sequential(
-                (0): Conv2d(128, 512, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (3): Sequential(
-                (0): Conv2d(512, 1024, kernel_size=(5, 1), stride=(3, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-              (4): Sequential(
-                (0): Conv2d(1024, 1024, kernel_size=(5, 1), stride=(1, 1), padding=(2, 0))
-                (1): LeakyReLU(negative_slope=0.1)
-              )
-            )
-            (output_conv): Conv2d(1024, 1, kernel_size=(2, 1), stride=(1, 1), padding=(1, 0))
-          )
-        )
-      )
-    )
-    (generator_adv_loss): GeneratorAdversarialLoss()
-    (discriminator_adv_loss): DiscriminatorAdversarialLoss()
-    (feat_match_loss): FeatureMatchLoss()
-    (mel_loss): MelSpectrogramLoss(
-      (wav_to_mel): LogMelFbank(
-        (stft): Stft(n_fft=1024, win_length=1024, hop_length=256, center=True, normalized=False, onesided=True)
-        (logmel): LogMel(sr=22050, n_fft=1024, n_mels=80, fmin=0, fmax=11025.0, htk=False)
-      )
-    )
-    (kl_loss): KLDivergenceLoss()
-  )
-)
-
-Model summary:
-    Class Name: ESPnetGANTTSModel
-    Total Number of model parameters: 96.24 M
-    Number of trainable parameters: 96.24 M (100.0%)
-    Size: 384.96 MB
-    Type: torch.float32
-[wieling-3-a100] 2023-12-01 15:58:42,453 (abs_task:1272) INFO: Optimizer:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:42,453 (abs_task:1273) INFO: Scheduler: <torch.optim.lr_scheduler.ExponentialLR object at 0x7f03bc341880>
-[wieling-3-a100] 2023-12-01 15:58:42,453 (abs_task:1272) INFO: Optimizer2:
-AdamW (
-Parameter Group 0
-    amsgrad: False
-    betas: [0.8, 0.99]
-    capturable: False
-    differentiable: False
-    eps: 1e-09
-    foreach: None
-    fused: None
-    initial_lr: 0.0003
-    lr: 0.0003
-    maximize: False
-    weight_decay: 0.0
-)
-[wieling-3-a100] 2023-12-01 15:58:42,454 (abs_task:1273) INFO: Scheduler2: <torch.optim.lr_scheduler.ExponentialLR object at 0x7f03bc341820>
-[wieling-3-a100] 2023-12-01 15:58:42,454 (abs_task:1282) INFO: Saving the configuration in exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12/config.yaml
-[wieling-3-a100] 2023-12-01 15:58:42,480 (abs_task:1293) INFO: Namespace(config='conf/train_vits.yaml', print_config=False, log_level='INFO', drop_last_iter=False, dry_run=False, iterator_type='sequence', valid_iterator_type=None, output_dir='exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12', ngpu=0, seed=67823, num_workers=4, num_att_plot=3, dist_backend='nccl', dist_init_method='env://', dist_world_size=None, dist_rank=None, local_rank=None, dist_master_addr=None, dist_master_port=None, dist_launcher=None, multiprocessing_distributed=False, unused_parameters=True, sharded_ddp=False, cudnn_enabled=True, cudnn_benchmark=False, cudnn_deterministic=False, collect_stats=True, write_collected_feats=False, max_epoch=1000, patience=None, val_scheduler_criterion=('valid', 'loss'), early_stopping_criterion=('valid', 'loss', 'min'), best_model_criterion=[['train', 'total_count', 'max']], keep_nbest_models=10, nbest_averaging_interval=0, grad_clip=-1, grad_clip_type=2.0, grad_noise=False, accum_grad=1, no_forward_run=False, resume=False, train_dtype='float32', use_amp=False, log_interval=50, use_matplotlib=True, use_tensorboard=True, create_graph_in_tensorboard=False, use_wandb=True, wandb_project='GROTTS', wandb_id=None, wandb_entity=None, wandb_name='VITS_lr_3.0e-4', wandb_model_log_interval=-1, detect_anomaly=False, use_lora=False, save_lora_only=True, lora_conf={}, pretrain_path=None, init_param=['downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv'], ignore_init_mismatch=False, freeze_param=[], num_iters_per_epoch=1000, batch_size=40, valid_batch_size=None, batch_bins=10000000, valid_batch_bins=None, train_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.12.scp'], valid_shape_file=['exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.12.scp'], batch_type='numel', valid_batch_type=None, fold_length=[], sort_in_batch='descending', shuffle_within_batch=False, sort_batch='descending', multiple_iterator=False, chunk_length=500, chunk_shift_ratio=0.5, num_cache_chunks=1024, chunk_excluded_key_prefixes=[], chunk_default_fs=None, train_data_path_and_name_and_type=[('dump/raw/train_nodev/text', 'text', 'text'), ('dump/raw/train_nodev/wav.scp', 'speech', 'sound'), ('dump/raw/train_nodev/utt2sid', 'sids', 'text_int')], valid_data_path_and_name_and_type=[('dump/raw/train_dev/text', 'text', 'text'), ('dump/raw/train_dev/wav.scp', 'speech', 'sound'), ('dump/raw/train_dev/utt2sid', 'sids', 'text_int')], allow_variable_data_keys=False, max_cache_size=0.0, max_cache_fd=32, allow_multi_rates=False, valid_max_cache_size=None, exclude_weight_decay=False, exclude_weight_decay_conf={}, optim='adamw', optim_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler='exponentiallr', scheduler_conf={'gamma': 0.999875}, optim2='adamw', optim2_conf={'lr': 0.0003, 'betas': [0.8, 0.99], 'eps': 1e-09, 'weight_decay': 0.0}, scheduler2='exponentiallr', scheduler2_conf={'gamma': 0.999875}, generator_first=False, token_list=['<blank>', '<unk>', '<space>', 'e', 'n', 'a', 'o', 't', 'i', 'r', 'd', 's', 'k', 'l', 'm', 'u', 'g', 'h', 'w', 'v', '.', 'z', 'b', 'p', ',', 'j', 'c', 'f', '‘', '’', ':', '?', 'ö', "'", '!', '-', ';', 'ò', 'è', 'ì', 'é', 'y', 'ë', 'x', 'q', '<sos/eos>'], odim=None, model_conf={}, use_preprocessor=True, token_type='char', bpemodel=None, non_linguistic_symbols=None, cleaner=None, g2p=None, feats_extract='fbank', feats_extract_conf={'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'fs': 22050, 'fmin': 80, 'fmax': 7600, 'n_mels': 80}, normalize=None, normalize_conf={}, tts='vits', tts_conf={'generator_type': 'vits_generator', 'generator_params': {'hidden_channels': 192, 'spks': 4, 'global_channels': 256, 'segment_size': 32, 'text_encoder_attention_heads': 2, 'text_encoder_ffn_expand': 4, 'text_encoder_blocks': 6, 'text_encoder_positionwise_layer_type': 'conv1d', 'text_encoder_positionwise_conv_kernel_size': 3, 'text_encoder_positional_encoding_layer_type': 'rel_pos', 'text_encoder_self_attention_layer_type': 'rel_selfattn', 'text_encoder_activation_type': 'swish', 'text_encoder_normalize_before': True, 'text_encoder_dropout_rate': 0.1, 'text_encoder_positional_dropout_rate': 0.0, 'text_encoder_attention_dropout_rate': 0.1, 'use_macaron_style_in_text_encoder': True, 'use_conformer_conv_in_text_encoder': False, 'text_encoder_conformer_kernel_size': -1, 'decoder_kernel_size': 7, 'decoder_channels': 512, 'decoder_upsample_scales': [8, 8, 2, 2], 'decoder_upsample_kernel_sizes': [16, 16, 4, 4], 'decoder_resblock_kernel_sizes': [3, 7, 11], 'decoder_resblock_dilations': [[1, 3, 5], [1, 3, 5], [1, 3, 5]], 'use_weight_norm_in_decoder': True, 'posterior_encoder_kernel_size': 5, 'posterior_encoder_layers': 16, 'posterior_encoder_stacks': 1, 'posterior_encoder_base_dilation': 1, 'posterior_encoder_dropout_rate': 0.0, 'use_weight_norm_in_posterior_encoder': True, 'flow_flows': 4, 'flow_kernel_size': 5, 'flow_base_dilation': 1, 'flow_layers': 4, 'flow_dropout_rate': 0.0, 'use_weight_norm_in_flow': True, 'use_only_mean_in_flow': True, 'stochastic_duration_predictor_kernel_size': 3, 'stochastic_duration_predictor_dropout_rate': 0.5, 'stochastic_duration_predictor_flows': 4, 'stochastic_duration_predictor_dds_conv_layers': 3, 'vocabs': 46, 'aux_channels': 80}, 'discriminator_type': 'hifigan_multi_scale_multi_period_discriminator', 'discriminator_params': {'scales': 1, 'scale_downsample_pooling': 'AvgPool1d', 'scale_downsample_pooling_params': {'kernel_size': 4, 'stride': 2, 'padding': 2}, 'scale_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [15, 41, 5, 3], 'channels': 128, 'max_downsample_channels': 1024, 'max_groups': 16, 'bias': True, 'downsample_scales': [2, 2, 4, 4, 1], 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': False, 'use_spectral_norm': False}, 'follow_official_norm': False, 'periods': [2, 3, 5, 7, 11], 'period_discriminator_params': {'in_channels': 1, 'out_channels': 1, 'kernel_sizes': [5, 3], 'channels': 32, 'downsample_scales': [3, 3, 3, 3, 1], 'max_downsample_channels': 1024, 'bias': True, 'nonlinear_activation': 'LeakyReLU', 'nonlinear_activation_params': {'negative_slope': 0.1}, 'use_weight_norm': True, 'use_spectral_norm': False}}, 'generator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'discriminator_adv_loss_params': {'average_by_discriminators': False, 'loss_type': 'mse'}, 'feat_match_loss_params': {'average_by_discriminators': False, 'average_by_layers': False, 'include_final_outputs': True}, 'mel_loss_params': {'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None, 'window': 'hann', 'n_mels': 80, 'fmin': 0, 'fmax': None, 'log_base': None}, 'lambda_adv': 1.0, 'lambda_mel': 45.0, 'lambda_feat_match': 2.0, 'lambda_dur': 1.0, 'lambda_kl': 1.0, 'sampling_rate': 22050, 'cache_generator_outputs': True}, pitch_extract=None, pitch_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'f0max': 400, 'f0min': 80}, pitch_normalize=None, pitch_normalize_conf={}, energy_extract=None, energy_extract_conf={'fs': 22050, 'n_fft': 1024, 'hop_length': 256, 'win_length': None}, energy_normalize=None, energy_normalize_conf={}, required=['output_dir', 'token_list'], version='202310', distributed=False)
-# Accounting: time=18 threads=1
-# Ended (code 0) at Fri Dec  1 15:58:52 UTC 2023, elapsed time 18 seconds

exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12/config.yaml

@@ -1,383 +0,0 @@
-config: conf/train_vits.yaml
-print_config: false
-log_level: INFO
-drop_last_iter: false
-dry_run: false
-iterator_type: sequence
-valid_iterator_type: null
-output_dir: exp-vits-lr-3e-4/tts_stats_raw_char/logdir/stats.12
-ngpu: 0
-seed: 67823
-num_workers: 4
-num_att_plot: 3
-dist_backend: nccl
-dist_init_method: env://
-dist_world_size: null
-dist_rank: null
-local_rank: null
-dist_master_addr: null
-dist_master_port: null
-dist_launcher: null
-multiprocessing_distributed: false
-unused_parameters: true
-sharded_ddp: false
-cudnn_enabled: true
-cudnn_benchmark: false
-cudnn_deterministic: false
-collect_stats: true
-write_collected_feats: false
-max_epoch: 1000
-patience: null
-val_scheduler_criterion:
-- valid
-- loss
-early_stopping_criterion:
-- valid
-- loss
-- min
-best_model_criterion:
--   - train
-    - total_count
-    - max
-keep_nbest_models: 10
-nbest_averaging_interval: 0
-grad_clip: -1
-grad_clip_type: 2.0
-grad_noise: false
-accum_grad: 1
-no_forward_run: false
-resume: false
-train_dtype: float32
-use_amp: false
-log_interval: 50
-use_matplotlib: true
-use_tensorboard: true
-create_graph_in_tensorboard: false
-use_wandb: true
-wandb_project: GROTTS
-wandb_id: null
-wandb_entity: null
-wandb_name: VITS_lr_3.0e-4
-wandb_model_log_interval: -1
-detect_anomaly: false
-use_lora: false
-save_lora_only: true
-lora_conf: {}
-pretrain_path: null
-init_param:
-- downloads/espnet/kan-bayashi_ljspeech_vits/exp/tts_train_vits_raw_phn_tacotron_g2p_en_no_space/train.total_count.ave_10best.pth:tts:tts:tts.generator.text_encoder,tts.generator.posterior_encoder.input_conv
-ignore_init_mismatch: false
-freeze_param: []
-num_iters_per_epoch: 1000
-batch_size: 40
-valid_batch_size: null
-batch_bins: 10000000
-valid_batch_bins: null
-train_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/train.12.scp
-valid_shape_file:
-- exp-vits-lr-3e-4/tts_stats_raw_char/logdir/valid.12.scp
-batch_type: numel
-valid_batch_type: null
-fold_length: []
-sort_in_batch: descending
-shuffle_within_batch: false
-sort_batch: descending
-multiple_iterator: false
-chunk_length: 500
-chunk_shift_ratio: 0.5
-num_cache_chunks: 1024
-chunk_excluded_key_prefixes: []
-chunk_default_fs: null
-train_data_path_and_name_and_type:
--   - dump/raw/train_nodev/text
-    - text
-    - text
--   - dump/raw/train_nodev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_nodev/utt2sid
-    - sids
-    - text_int
-valid_data_path_and_name_and_type:
--   - dump/raw/train_dev/text
-    - text
-    - text
--   - dump/raw/train_dev/wav.scp
-    - speech
-    - sound
--   - dump/raw/train_dev/utt2sid
-    - sids
-    - text_int
-allow_variable_data_keys: false
-max_cache_size: 0.0
-max_cache_fd: 32
-allow_multi_rates: false
-valid_max_cache_size: null
-exclude_weight_decay: false
-exclude_weight_decay_conf: {}
-optim: adamw
-optim_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler: exponentiallr
-scheduler_conf:
-    gamma: 0.999875
-optim2: adamw
-optim2_conf:
-    lr: 0.0003
-    betas:
-    - 0.8
-    - 0.99
-    eps: 1.0e-09
-    weight_decay: 0.0
-scheduler2: exponentiallr
-scheduler2_conf:
-    gamma: 0.999875
-generator_first: false
-token_list:
-- <blank>
-- <unk>
-- <space>
-- e
-- n
-- a
-- o
-- t
-- i
-- r
-- d
-- s
-- k
-- l
-- m
-- u
-- g
-- h
-- w
-- v
-- .
-- z
-- b
-- p
-- ','
-- j
-- c
-- f
-- ‘
-- ’
-- ':'
-- '?'
-- ö
-- ''''
-- '!'
-- '-'
-- ;
-- ò
-- è
-- ì
-- é
-- y
-- ë
-- x
-- q
-- <sos/eos>
-odim: null
-model_conf: {}
-use_preprocessor: true
-token_type: char
-bpemodel: null
-non_linguistic_symbols: null
-cleaner: null
-g2p: null
-feats_extract: fbank
-feats_extract_conf:
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-    fs: 22050
-    fmin: 80
-    fmax: 7600
-    n_mels: 80
-normalize: null
-normalize_conf: {}
-tts: vits
-tts_conf:
-    generator_type: vits_generator
-    generator_params:
-        hidden_channels: 192
-        spks: 4
-        global_channels: 256
-        segment_size: 32
-        text_encoder_attention_heads: 2
-        text_encoder_ffn_expand: 4
-        text_encoder_blocks: 6
-        text_encoder_positionwise_layer_type: conv1d
-        text_encoder_positionwise_conv_kernel_size: 3
-        text_encoder_positional_encoding_layer_type: rel_pos
-        text_encoder_self_attention_layer_type: rel_selfattn
-        text_encoder_activation_type: swish
-        text_encoder_normalize_before: true
-        text_encoder_dropout_rate: 0.1
-        text_encoder_positional_dropout_rate: 0.0
-        text_encoder_attention_dropout_rate: 0.1
-        use_macaron_style_in_text_encoder: true
-        use_conformer_conv_in_text_encoder: false
-        text_encoder_conformer_kernel_size: -1
-        decoder_kernel_size: 7
-        decoder_channels: 512
-        decoder_upsample_scales:
-        - 8
-        - 8
-        - 2
-        - 2
-        decoder_upsample_kernel_sizes:
-        - 16
-        - 16
-        - 4
-        - 4
-        decoder_resblock_kernel_sizes:
-        - 3
-        - 7
-        - 11
-        decoder_resblock_dilations:
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        -   - 1
-            - 3
-            - 5
-        use_weight_norm_in_decoder: true
-        posterior_encoder_kernel_size: 5
-        posterior_encoder_layers: 16
-        posterior_encoder_stacks: 1
-        posterior_encoder_base_dilation: 1
-        posterior_encoder_dropout_rate: 0.0
-        use_weight_norm_in_posterior_encoder: true
-        flow_flows: 4
-        flow_kernel_size: 5
-        flow_base_dilation: 1
-        flow_layers: 4
-        flow_dropout_rate: 0.0
-        use_weight_norm_in_flow: true
-        use_only_mean_in_flow: true
-        stochastic_duration_predictor_kernel_size: 3
-        stochastic_duration_predictor_dropout_rate: 0.5
-        stochastic_duration_predictor_flows: 4
-        stochastic_duration_predictor_dds_conv_layers: 3
-        vocabs: 46
-        aux_channels: 80
-    discriminator_type: hifigan_multi_scale_multi_period_discriminator
-    discriminator_params:
-        scales: 1
-        scale_downsample_pooling: AvgPool1d
-        scale_downsample_pooling_params:
-            kernel_size: 4
-            stride: 2
-            padding: 2
-        scale_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 15
-            - 41
-            - 5
-            - 3
-            channels: 128
-            max_downsample_channels: 1024
-            max_groups: 16
-            bias: true
-            downsample_scales:
-            - 2
-            - 2
-            - 4
-            - 4
-            - 1
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: false
-            use_spectral_norm: false
-        follow_official_norm: false
-        periods:
-        - 2
-        - 3
-        - 5
-        - 7
-        - 11
-        period_discriminator_params:
-            in_channels: 1
-            out_channels: 1
-            kernel_sizes:
-            - 5
-            - 3
-            channels: 32
-            downsample_scales:
-            - 3
-            - 3
-            - 3
-            - 3
-            - 1
-            max_downsample_channels: 1024
-            bias: true
-            nonlinear_activation: LeakyReLU
-            nonlinear_activation_params:
-                negative_slope: 0.1
-            use_weight_norm: true
-            use_spectral_norm: false
-    generator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    discriminator_adv_loss_params:
-        average_by_discriminators: false
-        loss_type: mse
-    feat_match_loss_params:
-        average_by_discriminators: false
-        average_by_layers: false
-        include_final_outputs: true
-    mel_loss_params:
-        fs: 22050
-        n_fft: 1024
-        hop_length: 256
-        win_length: null
-        window: hann
-        n_mels: 80
-        fmin: 0
-        fmax: null
-        log_base: null
-    lambda_adv: 1.0
-    lambda_mel: 45.0
-    lambda_feat_match: 2.0
-    lambda_dur: 1.0
-    lambda_kl: 1.0
-    sampling_rate: 22050
-    cache_generator_outputs: true
-pitch_extract: null
-pitch_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    f0max: 400
-    f0min: 80
-pitch_normalize: null
-pitch_normalize_conf: {}
-energy_extract: null
-energy_extract_conf:
-    fs: 22050
-    n_fft: 1024
-    hop_length: 256
-    win_length: null
-energy_normalize: null
-energy_normalize_conf: {}
-required:
-- output_dir
-- token_list
-version: '202310'
-distributed: false