Commit Graph

Select branches

Hide Pull Requests

dev

Mono Color

• 106b63d8a9 Update vocoder utils Eren Gölge 2021-06-18 13:45:52 +02:00
• 45947acb60 Update TTS.bin scripts for the new API Eren Gölge 2021-06-18 13:44:02 +02:00
• d7225eedb0 Update vocoder datasets and setup_dataset Eren Gölge 2021-06-18 13:42:36 +02:00
• d18198dff8 Implement setup_model for vocoder models Eren Gölge 2021-06-18 13:38:01 +02:00
• e949e7ad58 Update vocoder models Eren Gölge 2021-06-18 13:37:23 +02:00
• 51005cdab4 Update tts.models.setup_model Eren Gölge 2021-06-18 13:36:47 +02:00
• 7b8c15ac49 Create base 🐸TTS model abstraction for tts models Eren Gölge 2021-06-18 13:35:36 +02:00
• a358f74a52 Update vocoder model configs Eren Gölge 2021-06-18 13:29:35 +02:00
• 786170fe7d Update tts model configs Eren Gölge 2021-06-18 13:28:58 +02:00
• 626c9d41e6 Update tests for the new trainer API Eren Gölge 2021-06-18 13:27:19 +02:00
• fcfd95669a Update model test configs Eren Gölge 2021-06-18 13:24:48 +02:00
• 98298ee671 Implement unified IO utils Eren Gölge 2021-06-18 13:23:08 +02:00
• c7aad884cd Implement unified trainer Eren Gölge 2021-06-18 13:22:05 +02:00
• 6d7b5fbcde tts model abstraction with TTSModel Eren Gölge 2021-06-07 19:22:44 +02:00
• d4dbd89752 fix calculation of loader_start_time Eren Gölge 2021-06-08 17:34:19 +02:00
• c754a0e17d TrainerAbstract and related updates for TrainerTTS Eren Gölge 2021-06-08 14:18:18 +02:00
• 00c82c516d rename to Eren Gölge 2021-06-07 16:08:56 +02:00
• 166f0aeb9a merge if branches with the same implementation Eren Gölge 2021-06-07 15:11:33 +02:00
• 03494ad642 adjust distribute.py for the train_tts.py Eren Gölge 2021-06-06 15:20:17 +02:00
• fdfb18d230 downsize melgan test model size Eren Gölge 2021-06-06 13:42:52 +02:00
• 25238e0658 fix glow-tts inference() Eren Gölge 2021-06-06 13:42:35 +02:00
• 82582993cc use one testing sentence in tts tests Eren Gölge 2021-06-06 13:41:45 +02:00
• 419735f440 refactor and fix multi-speaker training in Trainer and Tacotron models Eren Gölge 2021-06-06 13:39:52 +02:00
• 269e5a734e add max_decoder_steps argument to tacotron models Eren Gölge 2021-06-06 13:38:01 +02:00
• b3324bd914 fix speaker_manager init Eren Gölge 2021-06-06 13:21:13 +02:00
• 304d60197b reduce multiband melgan test model size Eren Gölge 2021-06-05 11:48:16 +02:00
• 2c38ef8441 use get_speaker_manager in Trainer and save speakers.json file when needed Eren Gölge 2021-06-05 11:46:53 +02:00
• d6b2b6add6 make style and linter fixes Eren Gölge 2021-06-03 13:05:54 +02:00
• 802d461389 Compute d_vectors and speaker_ids separately in TTSDataset Eren Gölge 2021-06-03 11:46:53 +02:00
• db6a97d1a2 rename external speaker embedding arguments as d_vectors Eren Gölge 2021-06-03 11:42:40 +02:00
• 9042ae9195 use to_cuda() for moving data in format_batch() Eren Gölge 2021-06-03 15:05:39 +02:00
• 877bf66b61 reduce size of the metadata.csv used at testing Eren Gölge 2021-06-02 10:17:58 +02:00
• f82f1970b8 change to(device) to type_as in models Eren Gölge 2021-06-01 16:26:28 +02:00
• 9c94b0c5c0 init durations = None Eren Gölge 2021-06-01 14:53:57 +02:00
• 1fa15c195a docstring fix Eren Gölge 2021-06-01 14:35:15 +02:00
• 1c8a3d7c86 make style Eren Gölge 2021-05-27 17:25:00 +02:00
• 8cdd423234 styling formatting.py Eren Gölge 2021-05-25 14:41:13 +02:00
• b9a52dce9e add test_all to makefile Eren Gölge 2021-06-01 10:48:13 +02:00
• 30211512a4 fix type annotations Eren Gölge 2021-06-01 09:45:59 +02:00
• 87c61d210a update test to be less demanding Eren Gölge 2021-06-02 10:16:38 +02:00
• 6d6896fd99 reduce fullband-melgan test model size Eren Gölge 2021-05-31 15:44:09 +02:00
• 1443d03af1 update test for the new input output API of the tts models Eren Gölge 2021-05-31 15:43:40 +02:00
• b22b7620c3 update glow-tts output shapes to match [B, T, C] Eren Gölge 2021-05-31 15:42:07 +02:00
• 8381379938 formating cond_input with a function in Tacotron models Eren Gölge 2021-05-31 15:41:17 +02:00
• ef4ea9e527 update imports for formatters Eren Gölge 2021-05-31 10:07:12 +02:00
• 6c495c6a6e fix glow-tts inference and forward functions for handling cond_input and refactor its test Eren Gölge 2021-05-28 19:30:50 +02:00
• f840268181 refactor SpeakerManager Eren Gölge 2021-05-28 15:46:28 +02:00
• 421194880d linter fixes Eren Gölge 2021-05-28 14:04:51 +02:00
• 8e52a69230 delete separate tts training scripts and pre-commit configuration Eren Gölge 2021-05-28 14:03:02 +02:00
• d96ebcd6d3 make style Eren Gölge 2021-05-28 13:37:08 +02:00
• b643e8b37c logging/__init__.py Eren Gölge 2021-05-28 13:35:54 +02:00
• 0cee5042a9 fix logger imports Eren Gölge 2021-05-28 13:34:15 +02:00
• 72dceca52c import missings Eren Gölge 2021-05-28 13:32:35 +02:00
• 0eec238429 remove redundant imports Eren Gölge 2021-05-28 13:32:08 +02:00
• b500338faa make style Eren Gölge 2021-05-27 17:25:00 +02:00
• 469d2e620a update extract_tts_spectrogram for cond_input API of the models Eren Gölge 2021-05-31 15:35:26 +02:00
• 5ab28fa618 update extract_tts_spec... using SpeakerManager Eren Gölge 2021-06-01 10:49:48 +02:00
• c392fa4288 update extract_tts_spectrograms for the new model API Eren Gölge 2021-05-27 15:18:36 +02:00
• 8f47f95998 correct import of load_meta_data Eren Gölge 2021-05-27 15:18:10 +02:00
• c680a07a20 fix Synthesized for the new synthesis() Eren Gölge 2021-05-27 11:39:34 +02:00
• 73bf9673ed revert logging.info to print statements for trainer Eren Gölge 2021-05-27 11:38:46 +02:00
• d25f017b42 update setup_model.py imports Eren Gölge 2021-05-27 10:47:24 +02:00
• 7dff6be871 update tts training tests to use the trainer Eren Gölge 2021-05-27 10:26:31 +02:00
• bb355b7441 update align_tts.py model for the trainer Eren Gölge 2021-05-27 10:26:09 +02:00
• 9203b863d9 update align_tts_loss for trainer Eren Gölge 2021-05-27 10:25:40 +02:00
• fc9a0fb8ce update aling_tts_config for the trainer Eren Gölge 2021-05-27 10:25:25 +02:00
• e298b8e364 update trainer.py for better logging handling, restoring models and rename init_ functions with get_ Eren Gölge 2021-05-27 10:24:26 +02:00
• b8a4af4010 update synthesis.py for being more generic Eren Gölge 2021-05-26 16:03:56 +02:00
• c70d0c9dae update speedy_speech.py model for trainer Eren Gölge 2021-05-26 16:03:24 +02:00
• 06ee57d816 update speedy_speecy_config.py for the trainer Eren Gölge 2021-05-26 16:02:01 +02:00
• 4e910993f1 update tacotron model to return model_outputs Eren Gölge 2021-05-26 09:54:48 +02:00
• bb4deee64c update glow-tts for the trainer Eren Gölge 2021-05-26 09:54:04 +02:00
• 9134c7dfb6 update sequence_mask import globally Eren Gölge 2021-05-26 09:53:27 +02:00
• b2218e882a update glow_tts_config.py for setting the optimizer and the scheduler Eren Gölge 2021-05-26 09:52:29 +02:00
• 891631ab47 typing annotation for the trainer Eren Gölge 2021-05-28 13:33:58 +02:00
• 5f07315722 add trainer and train_tts Eren Gölge 2021-05-25 14:45:27 +02:00
• 34f8a74e4d remove truncated from synthesizer Eren Gölge 2021-05-25 14:41:56 +02:00
• 178eccbc16 update console logger Eren Gölge 2021-05-25 14:41:37 +02:00
• f4f83b6379 update synthesis.py for the trainer Eren Gölge 2021-05-25 14:40:25 +02:00
• 130781dab6 remove tts.generic_utils as all the functions are moved to other files Eren Gölge 2021-05-25 14:38:54 +02:00
• 535a458f40 update Tacotron models for the trainer Eren Gölge 2021-05-25 14:36:06 +02:00
• bdbfc95618 add gradual_training argument to tacotron.py Eren Gölge 2021-05-25 10:42:16 +02:00
• 5a2e75f0ee import missings for tacotron.py Eren Gölge 2021-05-25 10:41:22 +02:00
• da7d10e53c mode setup_model() to models/__init__.py Eren Gölge 2021-05-25 10:40:23 +02:00
• ca302db7b0 add sequence_mask to utils.data Eren Gölge 2021-05-25 14:38:31 +02:00
• 844abb3b1d setup_loss() in layer/__init__.py Eren Gölge 2021-05-25 10:38:44 +02:00
• a20a1c7d06 rename preprocess.py -> formatters.py Eren Gölge 2021-05-25 10:38:05 +02:00
• b9bccbb243 move load_meta_data and related functions to datasets/__init__.py Eren Gölge 2021-05-25 10:37:01 +02:00
• d09385808a set test_sentences in config Eren Gölge 2021-05-25 10:36:26 +02:00
• 8def3c87af trainer-API updates Eren Gölge 2021-05-20 18:23:53 +02:00
• 42554cc711 rename MyDataset -> TTSDataset Eren Gölge 2021-05-20 18:22:52 +02:00
• 1c4e806f54 use speaker manager on compute embeddings script Edresson 2021-06-27 03:35:34 -03:00
• eb84bb2bc8 Merge branch 'dev' into dev Edresson Casanova 2021-06-26 15:32:19 -03:00
• 6c7bbcaef0 Use en-us for testing phoneme models Eren Gölge 2021-06-25 16:52:17 +02:00
• 987cf1178b Bump up to v0.0.16 Eren Gölge 2021-06-25 14:44:33 +02:00
• a41f53fe72 Fix silly error in tests Michael Hansen 2021-06-16 18:10:51 -04:00
• 3f172b84d8 Fix linting issues Michael Hansen 2021-06-16 15:26:36 -04:00
• 4d8426fa0a Use eSpeak IPA lexicons by default for phoneme models Michael Hansen 2021-06-15 15:57:08 -04:00
• 618b509204 Use combined characters available in TTS phonemes (like ç) Michael Hansen 2021-06-11 10:43:52 -04:00
• da6f6a4a01 Update docstring for clean_gruut_phonemes Michael Hansen 2021-06-10 10:08:01 -04:00