pyproject.toml

[tool.poetry]
name = "styletts2"
version = "0.1.8"
description = "StyleTTS 2: Towards Human-Level Text-to-Speech through Style Diffusion and Adversarial Training with Large Speech Language Models. Original authors: Yinghao Aaron Li, Cong Han, Vinay S. Raghavan, Gavin Mischler, Nima Mesgarani. "
authors = ["Sidharth Rajaram <rajaram@berkeley.edu>"]
homepage = "https://github.com/sidharthrajaram/StyleTTS2"
repository = "https://github.com/sidharthrajaram/StyleTTS2"
license = "MIT"
readme = "README.md"

[tool.poetry.dependencies]
python = "^3.9"

soundfile = "^0.12.1"
torchaudio = "^2"
munch = "^4.0.0"
torch = "^2"
pydub = "^0.25.1"
PyYAML = "^6.0.1"
librosa = ">=0.9"
nltk = "^3.8.1"
matplotlib = "^3.7.0"
accelerate = "^0.25.0"
txtsplit = "^1.0.0"

huggingface-hub = "^0.19.0"
transformers = "^4.36.0"
cached-path = "^1.6.0"
filelock = "^3.12.4, <3.13"

einops = "^0.7.0"
einops-exts = "^0.0.4"
scipy = "^1.10.0"
tqdm = "^4.66.1"
typing = { version = "*", python = "<3.5" }
typing-extensions = "^4.9.0"

gruut = "^2.3.4"
gruut-ipa = "^0.13.0"
gruut-lang-en = "^2.0.0"
networkx = "^2.8.8, <3.0.0"

[build-system]
requires = ["poetry-core"]
build-backend = "poetry.core.masonry.api"