Commit Graph

Select branches

Hide Pull Requests

dev

Mono Color

• 07e8ff193a Bump version of gruut to 1.1 Michael Hansen 2021-06-09 13:44:04 -04:00
• 47191f3ecc Add tests for gruut phonemization Michael Hansen 2021-06-09 11:52:10 -04:00
• 67869e77f9 Use gruut for phonemization Michael Hansen 2021-06-02 10:35:59 -04:00
• 788992093d Add UnivNet vocoder 🚀 Eren Gölge 2021-06-21 17:49:08 +02:00
• 64fd59204c Use torch.linalg.qr for pytorch > v1.9.0 Eren Gölge 2021-06-23 13:45:59 +02:00
• 19a2cdc1e5 Update umap and numba vers. to 0.5.1 and 0.53 Eren Gölge 2021-06-23 11:42:11 +02:00
• aba840b4e6 Fix loading the amp scaler from a checkpoint 🛠️ Eren Gölge 2021-06-23 11:08:34 +02:00
• 18e5393f16 Add 🐍 python 3.9 to CI Eren Gölge 2021-06-22 21:39:17 +02:00
• 0ff2d2336a Fix wrong argument name 🛠️ Eren Gölge 2021-06-22 03:03:30 +02:00
• af12925142 Add AlignTTS recipe 👩‍🍳 Eren Gölge 2021-06-21 17:23:24 +02:00
• 61c3cb871f Docstring edit in TTSDataset.py ✍️ Eren Gölge 2021-06-21 16:53:19 +02:00
• 6f739ea07a Fix eval_log for gan.py 🛠️ Eren Gölge 2021-06-21 16:51:28 +02:00
• ebb91c0fbb Move TorchSTFT to utils.audio Eren Gölge 2021-06-21 16:50:37 +02:00
• 01c4b22a2f Fixup trainer.py 🛠️ Eren Gölge 2021-06-21 16:49:30 +02:00
• 9cb1062736 Create LJSpeech recipes for all the models Eren Gölge 2021-06-18 15:09:40 +02:00
• 787f6177c0 Update gitignore Eren Gölge 2021-06-18 15:02:02 +02:00
• d2d914cbc0 Update Pylint configuration Eren Gölge 2021-06-18 15:01:39 +02:00
• 7de2756fc4 Enable support for 🐍 python 3.10 Eren Gölge 2021-06-18 14:57:05 +02:00
• 220e184f66 Apply small fixes for API compatibility Eren Gölge 2021-06-18 14:54:04 +02:00
• 77d57dd301 Print max_decoder_steps when model reaches the limit Eren Gölge 2021-06-18 14:52:27 +02:00
• 7dc2177df4 Update synthesizer for speaker and model init Eren Gölge 2021-06-18 14:49:06 +02:00
• c3a0bc702e fixup configs Eren Gölge 2021-06-18 14:27:28 +02:00
• 0e01c2594f Update speaker_manager Eren Gölge 2021-06-18 14:26:41 +02:00
• 8182f5168f Fixup utils for the trainer Eren Gölge 2021-06-18 14:26:11 +02:00
• b4bb567e04 Update vocoder utils Eren Gölge 2021-06-18 13:45:52 +02:00
• f3ff5b1971 Update TTS.bin scripts for the new API Eren Gölge 2021-06-18 13:44:02 +02:00
• aed919cf1c Update vocoder datasets and setup_dataset Eren Gölge 2021-06-18 13:42:36 +02:00
• 59abf490a1 Implement setup_model for vocoder models Eren Gölge 2021-06-18 13:38:01 +02:00
• 420820caf4 Update vocoder models Eren Gölge 2021-06-18 13:37:23 +02:00
• d10f9c5676 Update tts.models.setup_model Eren Gölge 2021-06-18 13:36:47 +02:00
• cae702980f Create base 🐸TTS model abstraction for tts models Eren Gölge 2021-06-18 13:35:36 +02:00
• 70d968b169 Update vocoder model configs Eren Gölge 2021-06-18 13:29:35 +02:00
• f8a3460818 Update tts model configs Eren Gölge 2021-06-18 13:28:58 +02:00
• 20081c5396 Update tests for the new trainer API Eren Gölge 2021-06-18 13:27:19 +02:00
• 76e590f33c Update model test configs Eren Gölge 2021-06-18 13:24:48 +02:00
• acd96a4940 Implement unified IO utils Eren Gölge 2021-06-18 13:23:08 +02:00
• 6b907554f8 Implement unified trainer Eren Gölge 2021-06-18 13:22:05 +02:00
• 20c4a8c8e1 tts model abstraction with TTSModel Eren Gölge 2021-06-07 19:22:44 +02:00
• b934665fc0 fix calculation of loader_start_time Eren Gölge 2021-06-08 17:34:19 +02:00
• 64f0f57757 TrainerAbstract and related updates for TrainerTTS Eren Gölge 2021-06-08 14:18:18 +02:00
• f077a356e0 rename to Eren Gölge 2021-06-07 16:08:56 +02:00
• 4575b70826 merge if branches with the same implementation Eren Gölge 2021-06-07 15:11:33 +02:00
• 59be1b9af1 adjust distribute.py for the train_tts.py Eren Gölge 2021-06-06 15:20:17 +02:00
• 614738cc85 downsize melgan test model size Eren Gölge 2021-06-06 13:42:52 +02:00
• 4f29725eb6 fix glow-tts inference() Eren Gölge 2021-06-06 13:42:35 +02:00
• 7ea71c7586 use one testing sentence in tts tests Eren Gölge 2021-06-06 13:41:45 +02:00
• a87c886497 refactor and fix multi-speaker training in Trainer and Tacotron models Eren Gölge 2021-06-06 13:39:52 +02:00
• 0206bb847b add max_decoder_steps argument to tacotron models Eren Gölge 2021-06-06 13:38:01 +02:00
• cbb52b3d83 fix speaker_manager init Eren Gölge 2021-06-06 13:21:13 +02:00
• 3f96491168 reduce multiband melgan test model size Eren Gölge 2021-06-05 11:48:16 +02:00
• d2fd6a34a1 use get_speaker_manager in Trainer and save speakers.json file when needed Eren Gölge 2021-06-05 11:46:53 +02:00
• 147550c65f make style and linter fixes Eren Gölge 2021-06-03 13:05:54 +02:00
• a605dd3d08 Compute d_vectors and speaker_ids separately in TTSDataset Eren Gölge 2021-06-03 11:46:53 +02:00
• f00ef90ce6 rename external speaker embedding arguments as d_vectors Eren Gölge 2021-06-03 11:42:40 +02:00
• e7b7268c43 use to_cuda() for moving data in format_batch() Eren Gölge 2021-06-03 15:05:39 +02:00
• fdce9eb234 reduce size of the metadata.csv used at testing Eren Gölge 2021-06-02 10:17:58 +02:00
• 26a3312f0d change to(device) to type_as in models Eren Gölge 2021-06-01 16:26:28 +02:00
• c09622459e init durations = None Eren Gölge 2021-06-01 14:53:57 +02:00
• 2e31659dd9 docstring fix Eren Gölge 2021-06-01 14:35:15 +02:00
• 7a0750a4f5 make style Eren Gölge 2021-05-27 17:25:00 +02:00
• 534401377d styling formatting.py Eren Gölge 2021-05-25 14:41:13 +02:00
• 120ea679f9 add test_all to makefile Eren Gölge 2021-06-01 10:48:13 +02:00
• e229f5c081 fix type annotations Eren Gölge 2021-06-01 09:45:59 +02:00
• ab7f299d48 update test to be less demanding Eren Gölge 2021-06-02 10:16:38 +02:00
• a58e986f68 reduce fullband-melgan test model size Eren Gölge 2021-05-31 15:44:09 +02:00
• 9960c0c356 update test for the new input output API of the tts models Eren Gölge 2021-05-31 15:43:40 +02:00
• 506189bdee update glow-tts output shapes to match [B, T, C] Eren Gölge 2021-05-31 15:42:07 +02:00
• f568833d28 formating cond_input with a function in Tacotron models Eren Gölge 2021-05-31 15:41:17 +02:00
• 254707c610 update imports for formatters Eren Gölge 2021-05-31 10:07:12 +02:00
• 223502d827 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
• d4b1acfa81 refactor SpeakerManager Eren Gölge 2021-05-28 15:46:28 +02:00
• 26e7c0960c linter fixes Eren Gölge 2021-05-28 14:04:51 +02:00
• 79f7c5da1e delete separate tts training scripts and pre-commit configuration Eren Gölge 2021-05-28 14:03:02 +02:00
• ca787be193 make style Eren Gölge 2021-05-28 13:37:08 +02:00
• d376647ca0 logging/__init__.py Eren Gölge 2021-05-28 13:35:54 +02:00
• bb58a0588e fix logger imports Eren Gölge 2021-05-28 13:34:15 +02:00
• 9bbc924377 import missings Eren Gölge 2021-05-28 13:32:35 +02:00
• b4d4ce0d7e remove redundant imports Eren Gölge 2021-05-28 13:32:08 +02:00
• aefa71155c make style Eren Gölge 2021-05-27 17:25:00 +02:00
• 88d8a94a10 update extract_tts_spectrogram for cond_input API of the models Eren Gölge 2021-05-31 15:35:26 +02:00
• 667bb708b6 update extract_tts_spec... using SpeakerManager Eren Gölge 2021-06-01 10:49:48 +02:00
• 830306d2fd update extract_tts_spectrograms for the new model API Eren Gölge 2021-05-27 15:18:36 +02:00
• c673eb8ef8 correct import of load_meta_data Eren Gölge 2021-05-27 15:18:10 +02:00
• f0a419546b fix Synthesized for the new synthesis() Eren Gölge 2021-05-27 11:39:34 +02:00
• c7ff175592 revert logging.info to print statements for trainer Eren Gölge 2021-05-27 11:38:46 +02:00
• fd6afe5ae5 update setup_model.py imports Eren Gölge 2021-05-27 10:47:24 +02:00
• 0c05318e8d update tts training tests to use the trainer Eren Gölge 2021-05-27 10:26:31 +02:00
• c82d91051d update align_tts.py model for the trainer Eren Gölge 2021-05-27 10:26:09 +02:00
• 4f66e816d1 update align_tts_loss for trainer Eren Gölge 2021-05-27 10:25:40 +02:00
• 8213ad8b5f update aling_tts_config for the trainer Eren Gölge 2021-05-27 10:25:25 +02:00
• 8dfd4c91ff 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
• fb9289d365 update synthesis.py for being more generic Eren Gölge 2021-05-26 16:03:56 +02:00
• f121b0ff5d update speedy_speech.py model for trainer Eren Gölge 2021-05-26 16:03:24 +02:00
• 843b3ba960 update speedy_speecy_config.py for the trainer Eren Gölge 2021-05-26 16:02:01 +02:00
• c9790bee2c update tacotron model to return model_outputs Eren Gölge 2021-05-26 09:54:48 +02:00
• f09ec7e3a7 update glow-tts for the trainer Eren Gölge 2021-05-26 09:54:04 +02:00
• 3346a6d9dc update sequence_mask import globally Eren Gölge 2021-05-26 09:53:27 +02:00
• 9765b1aa6b update glow_tts_config.py for setting the optimizer and the scheduler Eren Gölge 2021-05-26 09:52:29 +02:00
• 6bf6543df8 typing annotation for the trainer Eren Gölge 2021-05-28 13:33:58 +02:00
• 57cdddef16 add trainer and train_tts Eren Gölge 2021-05-25 14:45:27 +02:00