Code

Adding neural HMM TTS#2271

Code

Closed

shivammehta25 wants to merge 83 commits intocoqui-ai:devcoqui-ai/TTS:devfrom shivammehta25:devshivammehta25/TTS:devCopy head branch name to clipboard

+151-26

Commits

Commits on Nov 26, 2022

• Adding encoder

  shivammehta25

  committed
  405bffe

Commits on Nov 27, 2022

• currently modifying hmm

  shivammehta25

  committed
  d607993

Commits on Nov 28, 2022

• Adding hmm

  shivammehta25

  committed
  a324920

Commits on Nov 30, 2022

• Adding overflow

  shivammehta25

  committed
  8628648

Commits on Dec 1, 2022

• Adding overflow setting up flat start

  shivammehta25

  committed
  6ec83c4

• Removing runs

  shivammehta25

  committed
  783a982

• adding normalization parameters

  shivammehta25

  committed
  10f15e0

• Fixing models on same device

  shivammehta25

  committed
  aff8b1f

Commits on Dec 2, 2022

• Training overflow and plotting evaluations

  shivammehta25

  committed
  62941d6

Commits on Dec 3, 2022

• Adding inference

  shivammehta25

  committed
  f448ea4

Commits on Dec 4, 2022

• At the end of epoch the test sentences are coming on cpu instead of gpu

  shivammehta25

  committed
  ff33837

Commits on Dec 5, 2022

• Adding figures from model during training to monitor

  shivammehta25

  committed
  3edb0d2

• reverting tacotron2 training recipe

  shivammehta25

  committed
  5fc800c

• fixing inference on gpu for test sentences on config

  shivammehta25

  committed
  427dfe5

• moving helpers and texts within overflows source code

  shivammehta25

  committed
  ecc12c6

• renaming to overflow

  shivammehta25

  committed
  b86f3f8

• moving loss to the model file

  shivammehta25

  committed
  995ee93

• Fixing the rename

  shivammehta25

  committed
  5b0fe46

• Model training but not plotting the test config sentences's audios

  shivammehta25

  committed
  5377f87

• Formatting logs

  shivammehta25

  committed
  bd5be6c

• Changing model name to camelcase

  shivammehta25

  committed
  755aa6f

• Fixing test log

  shivammehta25

  committed
  1350a4b

Commits on Dec 6, 2022

• Fixing plotting bug

  shivammehta25

  committed
  3c986fd

• Adding some tests

  shivammehta25

  committed
  4a5b1a0

Commits on Dec 7, 2022

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  5b1dabc

Commits on Dec 8, 2022

• Adding more tests to overflow

  shivammehta25

  committed
  f43d7e3

Commits on Dec 9, 2022

• Adding all tests for overflow

  shivammehta25

  committed
  c3d0167

• making changes to camel case in config

  shivammehta25

  committed
  ddefe34

Commits on Dec 10, 2022

• Adding information about parameters and docstring

  shivammehta25

  committed
  c2df9f3

• removing compute_mel_statistics moved statistic computation to the model instead

  shivammehta25

  committed
  9927434

• Added overflow in readme

  shivammehta25

  committed
  340cd0b

Commits on Dec 11, 2022

• Adding more test cases, now it doesn't saves transition_p like tensor and can be dumped as json

  shivammehta25

  committed
  aca3fe1

Commits on Dec 14, 2022

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  e7c11dd

• uncommenting the approximation to stablize the training

  shivammehta25

  committed
  7e2dbb1

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  be09d6c

Commits on Dec 22, 2022

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  282de93

Commits on Dec 23, 2022

• Adding encoder

  shivammehta25

  committed
  5df4fe8

• currently modifying hmm

  shivammehta25

  committed
  fa25825

• Adding hmm

  shivammehta25

  committed
  3cb0f78

• Adding overflow

  shivammehta25

  committed
  9984afa

• Adding overflow setting up flat start

  shivammehta25

  committed
  4dad45c

• Removing runs

  shivammehta25

  committed
  377bd3e

• adding normalization parameters

  shivammehta25

  committed
  a441c71

• Fixing models on same device

  shivammehta25

  committed
  995ac14

• Training overflow and plotting evaluations

  shivammehta25

  committed
  97b985b

• Adding inference

  shivammehta25

  committed
  227077a

• At the end of epoch the test sentences are coming on cpu instead of gpu

  shivammehta25

  committed
  bea46cc

• Adding figures from model during training to monitor

  shivammehta25

  committed
  03d028e

• reverting tacotron2 training recipe

  shivammehta25

  committed
  fc3c641

• fixing inference on gpu for test sentences on config

  shivammehta25

  committed
  c429837

• moving helpers and texts within overflows source code

  shivammehta25

  committed
  b804a12

• renaming to overflow

  shivammehta25

  committed
  3149b43

• moving loss to the model file

  shivammehta25

  committed
  8aff87a

• Fixing the rename

  shivammehta25

  committed
  8d7b0e7

• Model training but not plotting the test config sentences's audios

  shivammehta25

  committed
  8aaffed

• Formatting logs

  shivammehta25

  committed
  648b2c3

• Changing model name to camelcase

  shivammehta25

  committed
  d22c6c0

• Fixing test log

  shivammehta25

  committed
  6e08e4f

• Fixing plotting bug

  shivammehta25

  committed
  9394ce0

• Adding some tests

  shivammehta25

  committed
  e115361

• Adding more tests to overflow

  shivammehta25

  committed
  7a541b9

• Adding all tests for overflow

  shivammehta25

  committed
  1dccc29

• making changes to camel case in config

  shivammehta25

  committed
  1b1bf1f

• Adding information about parameters and docstring

  shivammehta25

  committed
  916b98e

• removing compute_mel_statistics moved statistic computation to the model instead

  shivammehta25

  committed
  6eff37c

• Added overflow in readme

  shivammehta25

  committed
  8a8dd1d

• Adding more test cases, now it doesn't saves transition_p like tensor and can be dumped as json

  shivammehta25

  committed
  e738c0c

• Handle espeak 1.48.15 (coqui-ai#2203)

  

  erogol

  authored and

  shivammehta25

  committed
  479c0cf

• Python API implementation (coqui-ai#2195)

  Show description for 4f02e2c
  

  erogol

  authored and

  shivammehta25

  committed
  4f02e2c

• Update README (coqui-ai#2204)

  

  erogol

  authored and

  shivammehta25

  committed
  89b9868

• Adding missing key to formatter (coqui-ai#2194)

  Show description for 684adb0
  

  p0p4k

  authored and

  shivammehta25

  committed
  684adb0

• Add YourTTS VCTK recipe (coqui-ai#2198)

  Show description for 55801cc
  

  Edresson

  authored and

  shivammehta25

  committed
  55801cc

• Add Original YourTTS vocabulary for full transfer learning (coqui-ai#2206)

  

  Edresson

  authored and

  shivammehta25

  committed
  a0be902

• uncommenting the approximation to stablize the training

  shivammehta25

  committed
  f3fe409

• Adding pre-trained Overflow model (coqui-ai#2211)

  Show description for aedd795
  

  erogol

  authored and

  shivammehta25

  committed
  aedd795

• Fixup overflow (coqui-ai#2218)

  Show description for 253b03f
  

  erogol

  authored and

  shivammehta25

  committed
  253b03f

• Bump up to v0.10.0

  

  erogol

  authored and

  shivammehta25

  committed
  c2ce4fb

• Add Ukrainian LADA (female) voice

  

  egorsmkv

  authored and

  shivammehta25

  committed
  fd5ad8c

Commits on Dec 30, 2022

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  1260c7f

Commits on Jan 3, 2023

• Merge branch 'coqui-ai:dev' into dev

  shivammehta25

  authored
  f73cd29

Commits on Jan 5, 2023

• Merge branch 'dev' of github.com:shivammehta25/TTS into dev

  shivammehta25

  committed
  2abbc97

Commits on Jan 9, 2023

• Adding a config flag to train neural HMM TTS instead of overflow

  shivammehta25

  committed
  790b846

• Backwards compatibility: Fixing model zoo if the flag is not set, set it

  shivammehta25

  committed
  a8d0b22