Dev.ap/multilingual

[roedoejet]

PR Goal?

Adds support for multilingual StyleTTS2 training, inference, and demos in the same line as https://github.com/EveryVoiceTTS/Speaker_Anonymization_StyleTTS2 (paper: https://aclanthology.org/2026.eacl-short.16/)

Claude Sonnet 5 helped me wire through all the changes to the concatenated tensors throughout StyleTTS2.

Fixes?

Feedback sought?

help testing

Priority?

medium

Tests added?

yes

How to test?

train a multilingual model, train a monolingual model. Both should run. Try inference/demo for both

Confidence?

high

Version change?

Related PRs?

EveryVoiceTTS/StyleTTS2#23

[semanticdiff-com]

Review changes with  

Changed Files

File	Status
everyvoice/cli.py	30% smaller
docs/guides/styletts2.md	Unsupported file format
everyvoice/.schema/everyvoice-text-to-wav-0.5.json	0% smaller
everyvoice/demo/app.py	0% smaller
everyvoice/model/e2e/StyleTTS2_lightning	0% smaller

[github-actions]

CLI load time: 0:00.22
Pull Request HEAD: 285953b51f8ec28325bb55296c1396238c10fb7c

Imports that take more than 0.1 s:
import time: self [us] | cumulative | imported package
import time:       530 |     120560 |   typer
import time:       744 |     111152 |           everyvoice.model.feature_prediction.FastSpeech2_lightning.fs2.cli.benchmark
import time:       360 |     120117 |         everyvoice.model.feature_prediction.FastSpeech2_lightning.fs2.cli.cli
import time:       210 |     120959 |       everyvoice.model.feature_prediction.FastSpeech2_lightning.fs2.cli
import time:      1140 |     122098 |     everyvoice.model.feature_prediction.FastSpeech2_lightning.fs2.cli.check_data
import time:      1988 |     126462 |   everyvoice.base_cli.check_group
import time:      5794 |     350031 | everyvoice.cli
import time:     53114 |     114281 |                 importlib.metadata._text
import time:     58966 |     180702 |               importlib.metadata._adapters
import time:     68759 |     446657 |             importlib.metadata
import time:       400 |     447057 |           pygments.plugin
import time:      3281 |     453325 |         pygments.filters
import time:      4116 |     459338 |       pygments.lexer
import time:      4688 |     481256 |     rich.syntax
import time:      4733 |     550662 |   rich.markdown
import time:      3672 |     651627 | typer.rich_utils

[codecov]

Codecov Report

❌ Patch coverage is 0% with 11 lines in your changes missing coverage. Please review.
✅ Project coverage is 85.21%. Comparing base (4aaa2ae) to head (285953b).

Files with missing lines	Patch %	Lines
everyvoice/demo/app.py	0.00%	11 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #851      +/-   ##
==========================================
- Coverage   85.41%   85.21%   -0.20%     
==========================================
  Files          51       51              
  Lines        4695     4706      +11     
  Branches      696      699       +3     
==========================================
  Hits         4010     4010              
- Misses        524      535      +11     
  Partials      161      161              

☔ View full report in Codecov by Harness.
📢 Have feedback on the report? Share it here.

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.