indextts2-rust/Cargo.toml at main · DevMan57/indextts2-rust · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
[package]
name = "indextts2"
version = "0.1.0"
edition = "2021"
authors = ["Henri <henri@example.com>"]
description = "High-performance Rust implementation of IndexTTS2 zero-shot text-to-speech"
license = "Apache-2.0"
repository = "https://github.com/DevMan57/indextts2-rust"
keywords = ["tts", "speech-synthesis", "deep-learning", "voice-cloning", "audio"]
categories = ["multimedia::audio", "science::ml"]

[dependencies]
# ML Framework (HuggingFace Candle)
candle-core = "0.8"
candle-nn = "0.8"
candle-transformers = "0.8"

# Tokenization
tokenizers = "0.20"

# Audio Processing
cpal = "0.15"                   # Cross-platform audio I/O
rodio = "0.19"                  # Audio playback
rubato = "0.16"                 # Sample rate conversion
rustfft = "6.2"                 # FFT for mel spectrograms
hound = "3.5"                   # WAV file I/O
symphonia = { version = "0.5", features = ["mp3", "flac", "ogg"] }

# Configuration & Serialization
serde = { version = "1.0", features = ["derive"] }
serde_yaml = "0.9"
serde_json = "1.0"
toml = "0.8"

# Error Handling
anyhow = "1.0"
thiserror = "2.0"

# Logging & Diagnostics
tracing = "0.1"
tracing-subscriber = { version = "0.3", features = ["env-filter"] }

# CLI
clap = { version = "4.5", features = ["derive", "env"] }
indicatif = "0.17"              # Progress bars
console = "0.15"                # Terminal colors

# Async Runtime
tokio = { version = "1.42", features = ["full"] }
tokio-stream = "0.1"

# Parallelism
rayon = "1.10"
crossbeam = "0.8"

# Numerics
ndarray = "0.16"
num-traits = "0.2"
rand = "0.8"

# File System
walkdir = "2.5"
tempfile = "3.14"
memmap2 = "0.9"

# HTTP (for model downloads)
reqwest = { version = "0.12", features = ["json", "stream"] }

# HuggingFace Hub
hf-hub = "0.3"

[dev-dependencies]
criterion = "0.5"
proptest = "1.5"
approx = "0.5"
insta = "1.41"

[features]
default = ["cuda"]
cuda = ["candle-core/cuda", "candle-nn/cuda", "candle-transformers/cuda"]
metal = ["candle-core/metal", "candle-nn/metal"]
mkl = ["candle-core/mkl", "candle-nn/mkl"]
accelerate = ["candle-core/accelerate", "candle-nn/accelerate"]

[[bin]]
name = "indextts2"
path = "src/main.rs"

[[bin]]
name = "diagnose_weights"
path = "src/bin/diagnose_weights.rs"

[[bin]]
name = "test_length_regulator"
path = "src/bin/test_length_regulator.rs"

[[example]]
name = "basic_inference"
path = "examples/basic_inference.rs"

[[example]]
name = "streaming_tts"
path = "examples/streaming_tts.rs"

[[example]]
name = "emotion_control"
path = "examples/emotion_control.rs"

[[bench]]
name = "inference_bench"
harness = false
path = "benches/inference_bench.rs"

[profile.release]
lto = true
codegen-units = 1
opt-level = 3

[profile.dev]
opt-level = 1

[profile.bench]
inherits = "release"