Commit Graph

  • d9452d7038 Add end2end VITS loss Edresson Casanova 2022-06-02 13:50:08 -03:00
  • 4a5143679d Make style add-synpaflex-formatter WeberJulian 2022-06-01 18:36:58 +02:00
  • 68cef28a88 Adding TTS Tutorials (#1584) Aya-AlJafari 2022-06-02 13:23:00 +03:00
  • 1f23c6f106 Fix formatter WeberJulian 2022-05-23 23:06:02 +02:00
  • 5857941043 Add synpaflex formatter WeberJulian 2022-05-23 18:43:47 +02:00
  • f70e82cd19 Use fsspec and torch for embedding file IO (#1581) Eren Gölge 2022-06-01 13:49:42 +02:00
  • d7de111b76 Change default speakers file extension d_vector_serialization WeberJulian 2022-06-01 10:31:26 +02:00
  • f7b75e880d Add Thorsten VITS model thorsten_models Eren Gölge 2022-05-31 15:10:04 +02:00
  • b6bd74a9a9 fix invalid json (#1599) Ryan Le-Nguyen 2022-05-31 18:20:10 +10:00
  • 39f31e3d39 Make style WeberJulian 2022-05-30 21:52:40 +02:00
  • 9f1670010b Add dummy speakers.pth file WeberJulian 2022-05-30 20:54:44 +02:00
  • 41d86e0fe1 Set use_cuda to true if available WeberJulian 2022-05-30 18:58:38 +02:00
  • fbb450616c Fix compute embedding script WeberJulian 2022-05-30 18:55:11 +02:00
  • 01115124a7 Fix load and save files WeberJulian 2022-05-30 18:52:41 +02:00
  • 7eceead0d7 Fixup Eren Gölge 2022-05-19 10:35:32 +02:00
  • 15fb20c7c0 Use fsspec and torch for embedding file Eren Gölge 2022-05-19 10:32:19 +02:00
  • a790df4e94 Training recipes for thorsten dataset (#1020) speaker_encoder_model Noran Raskin 2022-05-30 12:07:31 +02:00
  • 71111d14e4 Merge pull request #1587 from ribeiromiranda/patch-1 Eren Gölge 2022-05-29 14:51:08 +02:00
  • b7080f0f64 Recreate the prior distribution of Capacitron VAE on the right device Edresson Casanova 2022-05-27 16:41:13 -03:00
  • bd35371944 Add prosody encoder inference support Edresson Casanova 2022-05-27 16:00:41 -03:00
  • 2568b722dd Add an option to detach the prosody encoder input Edresson Casanova 2022-05-27 13:19:06 -03:00
  • a2aecea8f3 Add VAE prosody encoder Edresson Casanova 2022-05-27 12:53:56 -03:00
  • 312789edbf Condition the prosody encoder on z_p Edresson Casanova 2022-05-26 15:41:24 -03:00
  • e667fcb057 Support prosody conditional model on decoder input Edresson Casanova 2022-05-25 17:03:38 -03:00
  • 6c23398518 Add emotion classifier loss Edresson Casanova 2022-05-25 10:05:52 -03:00
  • 5b641ff0e3 Fix compute embeddings issue Edresson Casanova 2022-05-25 10:05:28 -03:00
  • d699416735 Add conditional module Edresson Casanova 2022-05-23 14:05:17 -03:00
  • 6e4b13c6cc Fix unit tests Edresson Casanova 2022-05-23 10:26:45 -03:00
  • 749b217884 Fix rebase issues Edresson Casanova 2022-05-20 18:29:39 -03:00
  • 1a88191a5a Disable the reversal prosody encoder speaker loss Edresson Casanova 2022-05-19 13:55:14 +00:00
  • 8505cd09e8 Add text encoder reversal speaker classifier loss Edresson Casanova 2022-05-17 13:14:15 +00:00
  • 024e567849 Clean up old code Edresson Casanova 2022-05-16 13:09:12 +00:00
  • dbaa71c944 Add prosody encoder params on config Edresson Casanova 2022-05-16 09:45:28 -03:00
  • 4107a1ef85 Add Speech style balancer Edresson Casanova 2022-04-19 15:51:15 -03:00
  • d49c6ab72f Add reversal classifier loss Edresson Casanova 2022-04-18 21:09:59 -03:00
  • 004862a79b Add prosody encoder training support Edresson Casanova 2022-04-18 17:01:44 -03:00
  • d9d9415513 Add emotion embedding in the encoder Edresson Casanova 2022-03-31 19:14:41 -03:00
  • b2b54668bc Add formatter for the Emotional Speech Dataset Edresson Casanova 2022-03-31 17:27:30 +00:00
  • d2b5db84f0 Remove useless encoder weights reload Edresson Casanova 2022-03-31 11:05:58 -03:00
  • 721a81b1d8 Fix emotion unit test Edresson Casanova 2022-03-31 08:34:08 -03:00
  • 01dd4e4051 Fix Style tests Edresson Casanova 2022-03-30 16:51:39 -03:00
  • 46762ccf35 Fix style tests Edresson Casanova 2022-03-23 15:31:33 -03:00
  • d8d1775273 Fix the Bug in Synthesizer Edresson Casanova 2022-03-30 15:32:35 -03:00
  • 39575f2937 Bug fix in single speaker emotion embedding training Edresson Casanova 2022-03-16 20:57:14 +00:00
  • d1ab3298ba Fix unit tests Edresson Casanova 2022-03-15 19:40:07 +00:00
  • a3ecaf3bdd Add emotion external embeddings training unit test Edresson Casanova 2022-03-15 13:09:58 +00:00
  • 5e0286c534 Add emotion consistency loss Edresson Casanova 2022-03-15 12:35:00 +00:00
  • e5c7ae9f1b Fix the bug in sythesizer Edresson Casanova 2022-03-15 12:33:36 +00:00
  • 6f95522edf Add Emotion Support for the VITS model Edresson Casanova 2022-03-15 01:16:48 +00:00
  • d8f5cb2674 Add emotion manager Edresson Casanova 2022-03-14 14:26:40 +00:00
  • 3b84ef9524 Fixed use_cuda issue in compute_embeddings.py André R. de Miranda 2022-05-20 12:46:46 -03:00
  • 8be21ec387 Capacitron (#977) a-froghyar 2022-05-20 16:17:11 +02:00
  • ee99a6c1e2 Fix voice conversion inference (#1583) Edresson Casanova 2022-05-20 08:16:01 -03:00
  • 63d27bc8d4 Disable the reversal prosody encoder speaker loss dev-emotion Edresson Casanova 2022-05-19 13:55:14 +00:00
  • bdefc43d96 Bug fix on pre-compute F0 vits-pitch-pred Edresson Casanova 2022-05-19 13:48:02 +00:00
  • 7b85703b28 Add text encoder reversal speaker classifier loss Edresson Casanova 2022-05-17 13:14:15 +00:00
  • 8adcd1de8e Rename g as spk_emb unified_api_forwardtts2 Eren G??lge 2022-05-17 13:37:05 +02:00
  • 2d29e8219d Fix up Eren G??lge 2022-05-17 13:36:40 +02:00
  • 8e915b70e0 Make hifigan discriminator configurable Eren G??lge 2022-05-17 13:34:57 +02:00
  • c437db15fd Fix dirt Eren G??lge 2022-05-17 13:34:38 +02:00
  • a05c82f9ef Fix audio_config handling Eren Gölge 2022-04-22 12:50:10 +02:00
  • b3fb0e19e8 Implement get_state_dict Eren Gölge 2022-04-22 12:39:46 +02:00
  • ce4f96292a Remove remaned trainer functions Eren Gölge 2022-04-22 12:37:35 +02:00
  • 96779e75ba Return duration by ForwardTTS inference Eren Gölge 2022-04-19 11:00:15 +02:00
  • 9291d13c69 Make style Eren Gölge 2022-04-19 10:59:59 +02:00
  • edd59c81e8 Update ForwardTTSe2e tests Eren Gölge 2022-04-19 10:58:52 +02:00
  • 0b585b46c1 Refactor TTSDataset to use numpy transforms Eren Gölge 2022-04-19 09:23:18 +02:00
  • 4171f4e9c6 Update ForwardTTSE2eLoss Eren Gölge 2022-04-19 09:22:50 +02:00
  • dbe5eb992e Make AP optional in BaseTTS Eren Gölge 2022-04-19 09:22:08 +02:00
  • 6a53b77a95 Add numpy and torch transforms Eren Gölge 2022-04-19 09:21:46 +02:00
  • c3fb49bf76 Refactor ForwardTTS to skip decoder Eren Gölge 2022-04-19 09:21:31 +02:00
  • cc57c20162 Make plot results more general Eren Gölge 2022-04-19 09:20:31 +02:00
  • e7c5db0d97 Add missing kernel size attr to transformer layer Eren Gölge 2022-04-19 09:19:57 +02:00
  • 231c69b12e Remove AP from FastPitchE2e Eren Gölge 2022-04-19 09:19:07 +02:00
  • 4556c61902 Update fastpitche2e recipe Eren Gölge 2022-04-19 09:18:49 +02:00
  • 5f9d559419 Update import statements Eren Gölge 2022-04-19 09:16:03 +02:00
  • 9f8d86b716 Remove redundancy Eren Gölge 2022-04-04 09:46:30 +02:00
  • 0738cb0efe Fix Vocoder logging Eren Gölge 2022-04-04 09:46:10 +02:00
  • 760f045aaa Rename vars in VITS Eren Gölge 2022-04-04 09:45:46 +02:00
  • 775a6ab6ee Add cond layer in decoder Eren Gölge 2022-04-04 09:44:20 +02:00
  • 28a53c7462 Refactor multi-speaker init in ForwardTTS Eren Gölge 2022-04-04 09:43:46 +02:00
  • c125024da0 Implement BaseTTSE2E Eren Gölge 2022-04-04 09:43:15 +02:00
  • b16613c5ad Implement ForwardTTSE2E Loss Eren Gölge 2022-04-04 09:42:50 +02:00
  • aea8cb7668 Implement FastPitchE2E LJSpeech recipe Eren Gölge 2022-04-04 09:41:46 +02:00
  • 2a61b8fdaf Implement ForwardTTSE2E tests Eren Gölge 2022-04-04 09:41:25 +02:00
  • 85731482e1 Implement FastPitchE2EConfig Eren Gölge 2022-04-04 09:41:05 +02:00
  • fccda5ae7b Implement ForwardTTSE2Eg Eren Gölge 2022-04-04 09:40:36 +02:00
  • d94b8bac02 Add pitch predictor Edresson Casanova 2022-05-16 21:53:49 +00:00
  • dcd0d1f6a1 Clean up old code Edresson Casanova 2022-05-16 13:09:12 +00:00
  • 3a524b0597 Add prosody encoder params on config Edresson Casanova 2022-05-16 09:45:28 -03:00
  • f237e4ccd9 Merge pull request #1574 from coqui-ai/update_badge v0.7.0_models Eren Gölge 2022-05-13 14:58:05 +02:00
  • e282da5161 Update CI badges Eren Gölge 2022-05-13 14:56:49 +02:00
  • e5d8ec2402 Change the VITS upsampling interpolation trick to linear (#1564) Edresson Casanova 2022-05-13 05:52:39 -03:00
  • c6008e5235 Add audio length sampler balancer (#1561) Edresson Casanova 2022-05-12 14:59:19 -03:00
  • 6e460b7e42 Add an assert for the upsampling trick (#1538) Eren Gölge 2022-05-12 19:55:24 +02:00
  • 6048959e24 Add CPU only Docker image (#1573) Eren Gölge 2022-05-12 19:33:27 +02:00
  • 002f826a1c Add unit tests audio_len_sampler Edresson Casanova 2022-05-07 15:56:39 -03:00
  • e86e3d2e87 Add audio length sampler balancer Edresson Casanova 2022-05-07 14:42:58 -03:00
  • 5e4bd9bfe8 Merge branch 'cpu-only-docker-image' of https://github.com/coqui-ai/TTS into cpu-only-docker-image cpu-only-docker-image Eren Gölge 2022-05-12 18:50:06 +02:00
  • 085517b79a Fix Dockerfile Eren Gölge 2022-05-12 12:55:27 +02:00