environment specification, estonian test sentences

config/session_paths.yaml

@@ -3,7 +3,7 @@ wav_directory: '/path/to/wav_directory' # path to directory cointaining the wavs
 metadata_path: '/path/to/metadata.csv'  # name of metadata file under wav_directory
 log_directory: '/path/to/logs_directory'   # weights and logs are stored here
 train_data_directory: 'transformer_tts_data'   # training data is stored here
-data_config: 'config/data_config_est.yaml'
+data_config: 'config/data_config.yaml'
 aligner_config: 'config/aligner_config.yaml'
 tts_config: 'config/tts_config.yaml'
 

config/tts_config_est.yaml

@@ -0,0 +1,42 @@
+# ARCHITECTURE
+decoder_model_dimension: 384
+encoder_model_dimension: 384
+decoder_num_heads: [2, 2, 2, 2, 2, 2]  # the length of this defines the number of layers
+encoder_num_heads: [2, 2, 2, 2, 2, 2]  # the length of this defines the number of layers
+encoder_feed_forward_dimension: null
+decoder_feed_forward_dimension: null
+decoder_prenet_dimension: 384
+encoder_prenet_dimension: 384
+encoder_attention_conv_filters: [1536, 384]
+decoder_attention_conv_filters: [1536, 384]
+encoder_attention_conv_kernel: 3
+decoder_attention_conv_kernel: 3
+encoder_max_position_encoding: 2000
+decoder_max_position_encoding: 10000
+encoder_dense_blocks: 0
+decoder_dense_blocks: 0
+# STATS PREDICTORS ARCHITECTURE
+duration_conv_filters: [256, 226]
+pitch_conv_filters: [256, 226]
+duration_kernel_size: 3
+pitch_kernel_size: 3
+
+# TRAINING
+predictors_dropout: 0.1
+dropout_rate: 0.1
+learning_rate_schedule:
+  - [0, 1.0e-4]
+max_steps: 260_000
+debug: False
+
+# LOGGING
+validation_frequency: 5_000
+prediction_frequency: 5_000
+weights_save_frequency: 5_000
+train_images_plotting_frequency: 1_000
+keep_n_weights: 5
+keep_checkpoint_every_n_hours: 12
+n_steps_avg_losses: [100, 500, 1_000, 5_000]  # command line display of average loss values for the last n steps
+prediction_start_step: 4_000
+text_prediction:
+  - test_sentences_est.txt

environment.yml

@@ -0,0 +1,21 @@
+channels:
+  - conda-forge
+  - anaconda
+dependencies:
+  - python==3.8
+  - matplotlib==3.2.2
+  - librosa==0.7.1
+  - numba==0.48
+  - numpy==1.17.4
+  - ruamel.yaml==0.16.6
+  - cudnn
+  - tqdm==4.42.1
+  - pysoundfile
+  - scipy
+  - pip
+  - pip:
+      - tensorflow-gpu==2.5.0
+      - webrtcvad
+      - p_tqdm
+      - pyworld
+      - phonemizer==2.2.1

test_sentences_est.txt

@@ -0,0 +1,6 @@
+külma on üks kuni viis kraadi ja saartel on õhutemperatuur miinus ühe ja pluss ühe kraadi vahel.
+kanepi läbis austraalias kvalifikatsiooni edukalt ja pääses kolmekümne kahe parema hulka.
+võõra viipekaardi leidnud alaealised lõid laiaks suure summa.
+las vegases lasi mees maha kaks hotelli turvatöötajat.
+muudatuste eesmärk on vähendada haigestumisel töötajate omavastutust ning langetada inimeste haigena tööl käimise riski ning koroonaviiruse levikut.
+"teie räägite hirmsaid asju!" hüüdis perenaine, nagu hakkaks tal õudne, ja villul oli tundmus, et perenaine nihkub talle pisut lähemale.