Compare commits

..

1 Commits

Author SHA1 Message Date
tuna2134
159f58026d release v0.2.0-alpha7 2025-07-21 11:58:37 +09:00
9 changed files with 104 additions and 119 deletions

View File

@@ -9,7 +9,3 @@ updates:
directory: "/" # Location of package manifests directory: "/" # Location of package manifests
schedule: schedule:
interval: "weekly" interval: "weekly"
- package-ecosystem: "npm" # See documentation for possible values
directory: "/" # Location of package manifests
schedule:
interval: "weekly"

4
.gitignore vendored
View File

@@ -1,10 +1,10 @@
target/ target/
models/ models/
!models/.gitkeep !models/.gitkeep
venv/ .venv/
.env .env
*.wav *.wav
node_modules/ node_modules/
dist/ dist/
*.csv *.csv
*.bin *.bin

102
Cargo.lock generated
View File

@@ -92,9 +92,9 @@ dependencies = [
[[package]] [[package]]
name = "anyhow" name = "anyhow"
version = "1.0.99" version = "1.0.98"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "b0674a1ddeecb70197781e945de4b3b8ffb61fa939a5597bcf48503737663100" checksum = "e16d2d3311acee920a9eb8d33b8cbc1787ce4a264e85f964c2404b969bdcd487"
[[package]] [[package]]
name = "atomic-waker" name = "atomic-waker"
@@ -871,9 +871,9 @@ checksum = "2304e00983f87ffb38b55b444b5e3b60a884b5d30c0fca7d82fe33449bbe55ea"
[[package]] [[package]]
name = "hermit-abi" name = "hermit-abi"
version = "0.5.1" version = "0.3.9"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "f154ce46856750ed433c8649605bf7ed2de3bc35fd9d2a9f30cddd873c80cb08" checksum = "d231dfb89cfffdbc30e7fc41579ed6066ad03abda9e567ccafae602b97ec5024"
[[package]] [[package]]
name = "hound" name = "hound"
@@ -995,7 +995,7 @@ dependencies = [
"hyper", "hyper",
"libc", "libc",
"pin-project-lite", "pin-project-lite",
"socket2 0.5.9", "socket2",
"tokio", "tokio",
"tower-service", "tower-service",
"tracing", "tracing",
@@ -1412,7 +1412,7 @@ dependencies = [
"reqwest", "reqwest",
"serde", "serde",
"tar", "tar",
"thiserror 2.0.15", "thiserror 2.0.12",
"yada", "yada",
] ]
@@ -1636,9 +1636,9 @@ dependencies = [
[[package]] [[package]]
name = "num_cpus" name = "num_cpus"
version = "1.17.0" version = "1.16.0"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "91df4bbde75afed763b708b7eee1e8e7651e02d97f6d5dd763e89367e957b23b" checksum = "4161fcb6d602d4d2081af7c3a45852d875a03dd337a6bfdd6e06407b61342a43"
dependencies = [ dependencies = [
"hermit-abi", "hermit-abi",
"libc", "libc",
@@ -1740,7 +1740,7 @@ checksum = "04744f49eae99ab78e0d5c0b603ab218f515ea8cfe5a456d7629ad883a3b6e7d"
[[package]] [[package]]
name = "ort" name = "ort"
version = "2.0.0-rc.10" version = "2.0.0-rc.10"
source = "git+https://github.com/pykeio/ort.git#d269461e2130b407589feff404025df25faeb3bb" source = "git+https://github.com/pykeio/ort.git#1e6f7ee1c8b056b00d280167ba172c96e78fcd1c"
dependencies = [ dependencies = [
"libloading", "libloading",
"ndarray", "ndarray",
@@ -1752,7 +1752,7 @@ dependencies = [
[[package]] [[package]]
name = "ort-sys" name = "ort-sys"
version = "2.0.0-rc.10" version = "2.0.0-rc.10"
source = "git+https://github.com/pykeio/ort.git#d269461e2130b407589feff404025df25faeb3bb" source = "git+https://github.com/pykeio/ort.git#1e6f7ee1c8b056b00d280167ba172c96e78fcd1c"
dependencies = [ dependencies = [
"flate2", "flate2",
"pkg-config", "pkg-config",
@@ -1812,7 +1812,7 @@ source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "198db74531d58c70a361c42201efde7e2591e976d518caf7662a47dc5720e7b6" checksum = "198db74531d58c70a361c42201efde7e2591e976d518caf7662a47dc5720e7b6"
dependencies = [ dependencies = [
"memchr", "memchr",
"thiserror 2.0.15", "thiserror 2.0.12",
"ucd-trie", "ucd-trie",
] ]
@@ -2141,7 +2141,7 @@ checksum = "dd6f9d3d47bdd2ad6945c5015a226ec6155d0bcdfd8f7cd29f86b71f8de99d2b"
dependencies = [ dependencies = [
"getrandom 0.2.16", "getrandom 0.2.16",
"libredox", "libredox",
"thiserror 2.0.15", "thiserror 2.0.12",
] ]
[[package]] [[package]]
@@ -2308,7 +2308,7 @@ checksum = "28d3b2b1366ec20994f1fd18c3c594f05c5dd4bc44d8bb0c1c632c8d6829481f"
[[package]] [[package]]
name = "sbv2_api" name = "sbv2_api"
version = "0.2.0-alpha6" version = "0.2.0-alpha7"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"axum", "axum",
@@ -2324,7 +2324,7 @@ dependencies = [
[[package]] [[package]]
name = "sbv2_bindings" name = "sbv2_bindings"
version = "0.2.0-alpha6" version = "0.2.0-alpha7"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"ndarray", "ndarray",
@@ -2334,7 +2334,7 @@ dependencies = [
[[package]] [[package]]
name = "sbv2_core" name = "sbv2_core"
version = "0.2.0-alpha6" version = "0.2.0-alpha7"
dependencies = [ dependencies = [
"anyhow", "anyhow",
"base64 0.22.1", "base64 0.22.1",
@@ -2352,7 +2352,7 @@ dependencies = [
"serde", "serde",
"serde_json", "serde_json",
"tar", "tar",
"thiserror 2.0.15", "thiserror 2.0.12",
"tokenizers", "tokenizers",
"ureq", "ureq",
"zstd", "zstd",
@@ -2360,7 +2360,7 @@ dependencies = [
[[package]] [[package]]
name = "sbv2_wasm" name = "sbv2_wasm"
version = "0.2.0-alpha6" version = "0.2.0-alpha7"
dependencies = [ dependencies = [
"js-sys", "js-sys",
"ndarray", "ndarray",
@@ -2430,9 +2430,9 @@ dependencies = [
[[package]] [[package]]
name = "serde_json" name = "serde_json"
version = "1.0.142" version = "1.0.140"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "030fedb782600dcbd6f02d479bf0d817ac3bb40d644745b769d6a96bc3afc5a7" checksum = "20068b6e96dc6c9bd23e01df8827e6c7e1f2fddd43c21810382803c136b99373"
dependencies = [ dependencies = [
"itoa", "itoa",
"memchr", "memchr",
@@ -2538,16 +2538,6 @@ dependencies = [
"windows-sys 0.52.0", "windows-sys 0.52.0",
] ]
[[package]]
name = "socket2"
version = "0.6.0"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "233504af464074f9d066d7b5416c5f9b894a5862a6506e306f7b816cdd6f1807"
dependencies = [
"libc",
"windows-sys 0.59.0",
]
[[package]] [[package]]
name = "socks" name = "socks"
version = "0.3.4" version = "0.3.4"
@@ -2710,11 +2700,11 @@ dependencies = [
[[package]] [[package]]
name = "thiserror" name = "thiserror"
version = "2.0.15" version = "2.0.12"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "80d76d3f064b981389ecb4b6b7f45a0bf9fdac1d5b9204c7bd6714fecc302850" checksum = "567b8a2dae586314f7be2a752ec7474332959c6460e02bde30d702a66d488708"
dependencies = [ dependencies = [
"thiserror-impl 2.0.15", "thiserror-impl 2.0.12",
] ]
[[package]] [[package]]
@@ -2730,9 +2720,9 @@ dependencies = [
[[package]] [[package]]
name = "thiserror-impl" name = "thiserror-impl"
version = "2.0.15" version = "2.0.12"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "44d29feb33e986b6ea906bd9c3559a856983f92371b3eaa5e83782a351623de0" checksum = "7f7cf42b4507d8ea322120659672cf1b9dbb93f8f2d4ecfd6e51350ff5b17a1d"
dependencies = [ dependencies = [
"proc-macro2", "proc-macro2",
"quote", "quote",
@@ -2766,9 +2756,9 @@ checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
[[package]] [[package]]
name = "tokenizers" name = "tokenizers"
version = "0.21.4" version = "0.21.2"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "a620b996116a59e184c2fa2dfd8251ea34a36d0a514758c6f966386bd2e03476" checksum = "4c3846d8588abed0daba25a0e47edd58ea15e450a6088b2575f5116fdb0b27ca"
dependencies = [ dependencies = [
"ahash", "ahash",
"aho-corasick", "aho-corasick",
@@ -2793,7 +2783,7 @@ dependencies = [
"serde", "serde",
"serde_json", "serde_json",
"spm_precompiled", "spm_precompiled",
"thiserror 2.0.15", "thiserror 2.0.12",
"unicode-normalization-alignments", "unicode-normalization-alignments",
"unicode-segmentation", "unicode-segmentation",
"unicode_categories", "unicode_categories",
@@ -2801,9 +2791,9 @@ dependencies = [
[[package]] [[package]]
name = "tokio" name = "tokio"
version = "1.47.1" version = "1.46.1"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "89e49afdadebb872d3145a5638b59eb0691ea23e46ca484037cfab3b76b95038" checksum = "0cc3a2344dafbe23a245241fe8b09735b521110d30fcefbbd5feb1797ca35d17"
dependencies = [ dependencies = [
"backtrace", "backtrace",
"bytes", "bytes",
@@ -2814,9 +2804,9 @@ dependencies = [
"pin-project-lite", "pin-project-lite",
"signal-hook-registry", "signal-hook-registry",
"slab", "slab",
"socket2 0.6.0", "socket2",
"tokio-macros", "tokio-macros",
"windows-sys 0.59.0", "windows-sys 0.52.0",
] ]
[[package]] [[package]]
@@ -2997,9 +2987,9 @@ checksum = "8ecb6da28b8a351d773b68d5825ac39017e680750f980f3a1a85cd8dd28a47c1"
[[package]] [[package]]
name = "ureq" name = "ureq"
version = "3.1.0" version = "3.0.12"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "00432f493971db5d8e47a65aeb3b02f8226b9b11f1450ff86bb772776ebadd70" checksum = "9f0fde9bc91026e381155f8c67cb354bcd35260b2f4a29bcc84639f762760c39"
dependencies = [ dependencies = [
"base64 0.22.1", "base64 0.22.1",
"der", "der",
@@ -3013,15 +3003,15 @@ dependencies = [
"socks", "socks",
"ureq-proto", "ureq-proto",
"utf-8", "utf-8",
"webpki-root-certs", "webpki-root-certs 0.26.11",
"webpki-roots", "webpki-roots 0.26.11",
] ]
[[package]] [[package]]
name = "ureq-proto" name = "ureq-proto"
version = "0.5.0" version = "0.4.2"
source = "registry+https://github.com/rust-lang/crates.io-index" source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "c5b6cabebbecc4c45189ab06b52f956206cea7d8c8a20851c35a85cb169224cc" checksum = "59db78ad1923f2b1be62b6da81fe80b173605ca0d57f85da2e005382adf693f7"
dependencies = [ dependencies = [
"base64 0.22.1", "base64 0.22.1",
"http", "http",
@@ -3221,6 +3211,15 @@ dependencies = [
"wasm-bindgen", "wasm-bindgen",
] ]
[[package]]
name = "webpki-root-certs"
version = "0.26.11"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "75c7f0ef91146ebfb530314f5f1d24528d7f0767efbfd31dce919275413e393e"
dependencies = [
"webpki-root-certs 1.0.0",
]
[[package]] [[package]]
name = "webpki-root-certs" name = "webpki-root-certs"
version = "1.0.0" version = "1.0.0"
@@ -3230,6 +3229,15 @@ dependencies = [
"rustls-pki-types", "rustls-pki-types",
] ]
[[package]]
name = "webpki-roots"
version = "0.26.11"
source = "registry+https://github.com/rust-lang/crates.io-index"
checksum = "521bc38abb08001b01866da9f51eb7c5d647a19260e00054a8c7fd5f9e57f7a9"
dependencies = [
"webpki-roots 1.0.0",
]
[[package]] [[package]]
name = "webpki-roots" name = "webpki-roots"
version = "1.0.0" version = "1.0.0"

View File

@@ -3,7 +3,7 @@ resolver = "3"
members = ["./crates/sbv2_api", "./crates/sbv2_core", "./crates/sbv2_bindings", "./crates/sbv2_wasm"] members = ["./crates/sbv2_api", "./crates/sbv2_core", "./crates/sbv2_bindings", "./crates/sbv2_wasm"]
[workspace.package] [workspace.package]
version = "0.2.0-alpha6" version = "0.2.0-alpha7"
edition = "2021" edition = "2021"
description = "Style-Bert-VITSの推論ライブラリ" description = "Style-Bert-VITSの推論ライブラリ"
license = "MIT" license = "MIT"
@@ -12,7 +12,7 @@ repository = "https://github.com/neodyland/sbv2-api"
documentation = "https://docs.rs/sbv2_core" documentation = "https://docs.rs/sbv2_core"
[workspace.dependencies] [workspace.dependencies]
anyhow = "1.0.99" anyhow = "1.0.96"
dotenvy = "0.15.7" dotenvy = "0.15.7"
env_logger = "0.11.6" env_logger = "0.11.6"
ndarray = "0.16.1" ndarray = "0.16.1"

View File

@@ -16,7 +16,7 @@ env_logger.workspace = true
log = "0.4.22" log = "0.4.22"
sbv2_core = { version = "0.2.0-alpha6", path = "../sbv2_core", features = ["aivmx"] } sbv2_core = { version = "0.2.0-alpha6", path = "../sbv2_core", features = ["aivmx"] }
serde = { version = "1.0.210", features = ["derive"] } serde = { version = "1.0.210", features = ["derive"] }
tokio = { version = "1.47.1", features = ["full"] } tokio = { version = "1.46.1", features = ["full"] }
utoipa = { version = "5.4.0", features = ["axum_extras"] } utoipa = { version = "5.4.0", features = ["axum_extras"] }
utoipa-scalar = { version = "0.3.0", features = ["axum"] } utoipa-scalar = { version = "0.3.0", features = ["axum"] }

View File

@@ -17,15 +17,15 @@ hound = "3.5.1"
jpreprocess = { version = "0.12.0", features = ["naist-jdic"] } jpreprocess = { version = "0.12.0", features = ["naist-jdic"] }
ndarray.workspace = true ndarray.workspace = true
npyz = { version = "0.8.4", optional = true } npyz = { version = "0.8.4", optional = true }
num_cpus = "1.17.0" num_cpus = "1.16.0"
once_cell.workspace = true once_cell.workspace = true
ort = { git = "https://github.com/pykeio/ort.git", version = "2.0.0-rc.9", optional = true } ort = { git = "https://github.com/pykeio/ort.git", version = "2.0.0-rc.10", optional = true }
regex = "1.10.6" regex = "1.10.6"
serde = { version = "1.0.210", features = ["derive"] } serde = { version = "1.0.210", features = ["derive"] }
serde_json = "1.0.142" serde_json = "1.0.128"
tar = "0.4.41" tar = "0.4.41"
thiserror = "2.0.15" thiserror = "2.0.11"
tokenizers = { version = "0.21.4", default-features = false } tokenizers = { version = "0.21.2", default-features = false }
zstd = "0.13.2" zstd = "0.13.2"
[features] [features]
@@ -44,4 +44,4 @@ base64 = ["dep:base64"]
[build-dependencies] [build-dependencies]
dirs = "6.0.0" dirs = "6.0.0"
ureq = "3.1.0" ureq = "3.0.12"

View File

@@ -127,20 +127,20 @@ impl JTalkProcess {
Ok(phone_tone_list) Ok(phone_tone_list)
} else if tone_values.len() == 2 { } else if tone_values.len() == 2 {
if tone_values == hash_set![0, 1] { if tone_values == hash_set![0, 1] {
Ok(phone_tone_list) return Ok(phone_tone_list);
} else if tone_values == hash_set![-1, 0] { } else if tone_values == hash_set![-1, 0] {
Ok(phone_tone_list return Ok(phone_tone_list
.iter() .iter()
.map(|x| { .map(|x| {
let new_tone = if x.1 == -1 { 0 } else { 1 }; let new_tone = if x.1 == -1 { 0 } else { 1 };
(x.0.clone(), new_tone) (x.0.clone(), new_tone)
}) })
.collect()) .collect());
} else { } else {
Err(Error::ValueError("Invalid tone values 0".to_string())) return Err(Error::ValueError("Invalid tone values 0".to_string()));
} }
} else { } else {
Err(Error::ValueError("Invalid tone values 1".to_string())) return Err(Error::ValueError("Invalid tone values 1".to_string()));
} }
} }

View File

@@ -240,43 +240,39 @@ impl TTSModelHolder {
} }
fn find_and_load_model<I: Into<TTSIdent>>(&mut self, ident: I) -> Result<bool> { fn find_and_load_model<I: Into<TTSIdent>>(&mut self, ident: I) -> Result<bool> {
let ident = ident.into(); let ident = ident.into();
// Locate target model entry let (bytes, style_vectors) = {
let target_index = self let model = self
.models .models
.iter() .iter()
.position(|m| m.ident == ident) .find(|m| m.ident == ident)
.ok_or(Error::ModelNotFoundError(ident.to_string()))?; .ok_or(Error::ModelNotFoundError(ident.to_string()))?;
if model.vits2.is_some() {
// Already loaded return Ok(true);
if self.models[target_index].vits2.is_some() { }
return Ok(true); (model.bytes.clone().unwrap(), model.style_vectors.clone())
} };
self.unload(ident.clone());
// Get bytes to build a Session let s = model::load_model(&bytes, false)?;
let bytes = self.models[target_index]
.bytes
.clone()
.ok_or(Error::ModelNotFoundError(ident.to_string()))?;
// Enforce max loaded models by evicting a different loaded model's session, not removing the entry
if let Some(max) = self.max_loaded_models { if let Some(max) = self.max_loaded_models {
let loaded_count = self.models.iter().filter(|m| m.vits2.is_some()).count(); if self.models.iter().filter(|x| x.vits2.is_some()).count() >= max {
if loaded_count >= max { self.unload(self.models.first().unwrap().ident.clone());
if let Some(evict_index) = self
.models
.iter()
.position(|m| m.vits2.is_some() && m.ident != ident)
{
// Drop only the session to free memory; keep bytes/style for future reload
self.models[evict_index].vits2 = None;
}
} }
} }
self.models.push(TTSModel {
// Build and set session in-place for the target model bytes: Some(bytes.to_vec()),
let s = model::load_model(&bytes, false)?; vits2: Some(s),
self.models[target_index].vits2 = Some(s); style_vectors,
Ok(true) ident: ident.clone(),
});
let model = self
.models
.iter()
.find(|m| m.ident == ident)
.ok_or(Error::ModelNotFoundError(ident.to_string()))?;
if model.vits2.is_some() {
return Ok(true);
}
Err(Error::ModelNotFoundError(ident.to_string()))
} }
/// Get style vector by style id and weight /// Get style vector by style id and weight

View File

@@ -173,15 +173,8 @@ pub fn parse_text_blocking(
} }
pub fn array_to_vec(audio_array: Array3<f32>) -> Result<Vec<u8>> { pub fn array_to_vec(audio_array: Array3<f32>) -> Result<Vec<u8>> {
// If SBV2_FORCE_STEREO is set ("1"/"true"), duplicate mono to stereo
let force_stereo = std::env::var("SBV2_FORCE_STEREO")
.ok()
.map(|v| matches!(v.as_str(), "1" | "true" | "TRUE" | "True"))
.unwrap_or(false);
let channels: u16 = if force_stereo { 2 } else { 1 };
let spec = WavSpec { let spec = WavSpec {
channels, channels: 1,
sample_rate: 44100, sample_rate: 44100,
bits_per_sample: 32, bits_per_sample: 32,
sample_format: SampleFormat::Float, sample_format: SampleFormat::Float,
@@ -190,16 +183,8 @@ pub fn array_to_vec(audio_array: Array3<f32>) -> Result<Vec<u8>> {
let mut writer = WavWriter::new(&mut cursor, spec)?; let mut writer = WavWriter::new(&mut cursor, spec)?;
for i in 0..audio_array.shape()[0] { for i in 0..audio_array.shape()[0] {
let output = audio_array.slice(s![i, 0, ..]).to_vec(); let output = audio_array.slice(s![i, 0, ..]).to_vec();
if force_stereo { for sample in output {
for sample in output { writer.write_sample(sample)?;
// Write to Left and Right channels
writer.write_sample(sample)?;
writer.write_sample(sample)?;
}
} else {
for sample in output {
writer.write_sample(sample)?;
}
} }
} }
writer.finalize()?; writer.finalize()?;