Update CI.yml

fix(deps): update rust crate tokio to v1.43.0

.github/workflows/CI.yml

@@ -79,8 +79,6 @@ jobs:
     strategy:
       matrix:
         platform:
-          - runner: macos-12
-            target: x86_64
           - runner: macos-14
             target: aarch64
     steps:
@@ -131,8 +129,8 @@ jobs:
           command: upload
           args: --non-interactive --skip-existing wheels-*/*
 
-  push-docker:
-    runs-on: ubuntu-latest
+  build-docker:
+    runs-on: ${{ matrix.platform }}
     if: "startsWith(github.ref, 'refs/tags/')"
     permissions:
       contents: read
@@ -140,6 +138,7 @@ jobs:
     strategy:
       matrix:
         tag: [cpu, cuda]
+        platform: [ubuntu-latest, ubuntu-24.04-arm]
     steps:
       - uses: actions/checkout@v4
       - name: Set up QEMU
@@ -156,8 +155,64 @@ jobs:
         uses: docker/build-push-action@v6
         with:
           context: .
-          push: true
-          tags: |
-            ghcr.io/${{ github.repository }}:${{ matrix.tag }}
           file: docker/${{ matrix.tag }}.Dockerfile
-          platforms: linux/amd64, linux/arm64
+          labels: ${{ steps.meta.outputs.labels }}
+          outputs: type=image,"name=ghcr.io/tuna2134/sbv2-api",push-by-digest=true,name-canonical=true,push=true
+      
+      - name: Export digest
+        run: |
+          mkdir -p ${{ runner.temp }}/digests
+          digest="${{ steps.build.outputs.digest }}"
+          touch "${{ runner.temp }}/digests/${digest#sha256:}"          
+
+      - name: Upload digest
+        uses: actions/upload-artifact@v4
+        with:
+          name: digests-${{ env.PLATFORM_PAIR }}
+          path: ${{ runner.temp }}/digests/*
+          if-no-files-found: error
+          retention-days: 1
+
+  merge:
+    runs-on: ubuntu-latest
+    needs:
+      - build-docker
+    steps:
+      - name: Download digests
+        uses: actions/download-artifact@v4
+        with:
+          path: ${{ runner.temp }}/digests
+          pattern: digests-*
+          merge-multiple: true
+      
+      - name: Login to GHCR
+        uses: docker/login-action@v3
+        with:
+          registry: ghcr.io
+          username: ${{ github.repository_owner }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Docker meta
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: |
+            ghcr.io/tuna2134/sbv2-api
+          tags: |
+            type=ref,event=branch
+            type=ref,event=pr
+            type=semver,pattern={{version}}
+            type=semver,pattern={{major}}.{{minor}}            
+
+      - name: Create manifest list and push
+        working-directory: ${{ runner.temp }}/digests
+        run: |
+          docker buildx imagetools create $(jq -cr '.tags | map("-t " + .) | join(" ")' <<< "$DOCKER_METADATA_OUTPUT_JSON") \
+            $(printf 'ghcr.io/tuna2134/sbv2-api@sha256:%s ' *)          
+
+      - name: Inspect image
+        run: |
+          docker buildx imagetools inspect ghcr.io/tuna2134/sbv2-api:${{ steps.meta.outputs.version }}

Cargo.lock

@@ -1,6 +1,6 @@
 # This file is automatically @generated by Cargo.
 # It is not intended for manual editing.
-version = 3
+version = 4
 
 [[package]]
 name = "addr2line"
@@ -77,20 +77,9 @@ dependencies = [
 
 [[package]]
 name = "anyhow"
-version = "1.0.93"
+version = "1.0.95"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "4c95c10ba0b00a02636238b814946408b1322d5ac4760326e6fb8ec956d85775"
-
-[[package]]
-name = "async-trait"
-version = "0.1.83"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "721cae7de5c34fbb2acd27e21e6d2cf7b886dce0c27388d46c4e6c47ea4318dd"
-dependencies = [
- "proc-macro2",
- "quote",
- "syn",
-]
+checksum = "34ac096ce696dc2fcabef30516bb13c0a68a11d30131d3df6f04711467681b04"
 
 [[package]]
 name = "autocfg"
@@ -100,13 +89,13 @@ checksum = "ace50bade8e6234aa140d9a2f552bbee1db4d353f69b8217bc503490fc1a9f26"
 
 [[package]]
 name = "axum"
-version = "0.7.7"
+version = "0.8.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "504e3947307ac8326a5437504c517c4b56716c9d98fac0028c2acc7ca47d70ae"
+checksum = "6d6fd624c75e18b3b4c6b9caf42b1afe24437daaee904069137d8bab077be8b8"
 dependencies = [
- "async-trait",
  "axum-core",
  "bytes",
+ "form_urlencoded",
  "futures-util",
  "http",
  "http-body",
@@ -124,7 +113,7 @@ dependencies = [
  "serde_json",
  "serde_path_to_error",
  "serde_urlencoded",
- "sync_wrapper 1.0.1",
+ "sync_wrapper",
  "tokio",
  "tower",
  "tower-layer",
@@ -134,11 +123,10 @@ dependencies = [
 
 [[package]]
 name = "axum-core"
-version = "0.4.5"
+version = "0.5.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "09f2bd6146b97ae3359fa0cc6d6b376d9539582c7b4220f041a33ec24c226199"
+checksum = "df1362f362fd16024ae199c1970ce98f9661bf5ef94b9808fee734bc3698b733"
 dependencies = [
- "async-trait",
  "bytes",
  "futures-util",
  "http",
@@ -147,7 +135,7 @@ dependencies = [
  "mime",
  "pin-project-lite",
  "rustversion",
- "sync_wrapper 1.0.1",
+ "sync_wrapper",
  "tower-layer",
  "tower-service",
  "tracing",
@@ -547,9 +535,9 @@ dependencies = [
 
 [[package]]
 name = "env_logger"
-version = "0.11.5"
+version = "0.11.6"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "e13fa619b91fb2381732789fc5de83b45675e882f66623b7d8cb4f643017018d"
+checksum = "dcaee3d8e3cfc3fd92428d477bc97fc29ec8716d180c0d74c643bb26166660e0"
 dependencies = [
  "anstream",
  "anstyle",
@@ -1058,9 +1046,9 @@ checksum = "bbd2bcb4c963f2ddae06a2efc7e9f3591312473c50c6685e1f298068316e66fe"
 
 [[package]]
 name = "libc"
-version = "0.2.161"
+version = "0.2.169"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "8e9489c2807c139ffd9c1794f4af0ebe86a828db53ecdc7fea2111d0fed085d1"
+checksum = "b5aba8db14291edd000dfcc4d620c7ebfb122c613afb886ca8803fa4e128a20a"
 
 [[package]]
 name = "libloading"
@@ -1244,9 +1232,9 @@ dependencies = [
 
 [[package]]
 name = "log"
-version = "0.4.22"
+version = "0.4.25"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "a7a70ba024b9dc04c27ea2f0c0548feb474ec5c54bba33a7f72f873a39d07b24"
+checksum = "04cbf5b083de1c7e0222a7a51dbfdba1cbe1c6ab0b15e29fff3f6c077fd9cd9f"
 
 [[package]]
 name = "macro_rules_attribute"
@@ -1266,9 +1254,9 @@ checksum = "b8dd856d451cc0da70e2ef2ce95a18e39a93b7558bedf10201ad28503f918568"
 
 [[package]]
 name = "matchit"
-version = "0.7.3"
+version = "0.8.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0e7465ac9959cc2b1404e8e2367b43684a6d13790fe23056cc8c6c5a6b7bcb94"
+checksum = "47e1ffaa40ddd1f3ed91f717a33c8c0ee23fff369e3aa8772b9605cc1d22f4c3"
 
 [[package]]
 name = "matrixmultiply"
@@ -1374,6 +1362,27 @@ dependencies = [
  "minimal-lexical",
 ]
 
+[[package]]
+name = "npyz"
+version = "0.8.3"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "13f27ea175875c472b3df61ece89a6d6ef4e0627f43704e400c782f174681ebd"
+dependencies = [
+ "byteorder",
+ "num-bigint",
+ "py_literal",
+]
+
+[[package]]
+name = "num-bigint"
+version = "0.4.6"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "a5e44f723f1133c9deac646763579fdb3ac745e418f2a7af9cd0c431da1f20b9"
+dependencies = [
+ "num-integer",
+ "num-traits",
+]
+
 [[package]]
 name = "num-complex"
 version = "0.4.6"
@@ -1513,6 +1522,51 @@ version = "2.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "e3148f5046208a5d56bcfc03053e3ca6334e51da8dfb19b6cdc8b306fae3283e"
 
+[[package]]
+name = "pest"
+version = "2.7.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "879952a81a83930934cbf1786752d6dedc3b1f29e8f8fb2ad1d0a36f377cf442"
+dependencies = [
+ "memchr",
+ "thiserror",
+ "ucd-trie",
+]
+
+[[package]]
+name = "pest_derive"
+version = "2.7.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "d214365f632b123a47fd913301e14c946c61d1c183ee245fa76eb752e59a02dd"
+dependencies = [
+ "pest",
+ "pest_generator",
+]
+
+[[package]]
+name = "pest_generator"
+version = "2.7.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "eb55586734301717aea2ac313f50b2eb8f60d2fc3dc01d190eefa2e625f60c4e"
+dependencies = [
+ "pest",
+ "pest_meta",
+ "proc-macro2",
+ "quote",
+ "syn",
+]
+
+[[package]]
+name = "pest_meta"
+version = "2.7.14"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "b75da2a70cf4d9cb76833c990ac9cd3923c9a8905a8929789ce347c84564d03d"
+dependencies = [
+ "once_cell",
+ "pest",
+ "sha2",
+]
+
 [[package]]
 name = "phf"
 version = "0.11.2"
@@ -1607,10 +1661,23 @@ dependencies = [
 ]
 
 [[package]]
-name = "pyo3"
-version = "0.22.6"
+name = "py_literal"
+version = "0.4.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "f402062616ab18202ae8319da13fa4279883a2b8a9d9f83f20dbade813ce1884"
+checksum = "102df7a3d46db9d3891f178dcc826dc270a6746277a9ae6436f8d29fd490a8e1"
+dependencies = [
+ "num-bigint",
+ "num-complex",
+ "num-traits",
+ "pest",
+ "pest_derive",
+]
+
+[[package]]
+name = "pyo3"
+version = "0.23.4"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "57fe09249128b3173d092de9523eaa75136bf7ba85e0d69eca241c7939c933cc"
 dependencies = [
  "anyhow",
  "cfg-if",
@@ -1627,9 +1694,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-build-config"
-version = "0.22.6"
+version = "0.23.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "b14b5775b5ff446dd1056212d778012cbe8a0fbffd368029fd9e25b514479c38"
+checksum = "1cd3927b5a78757a0d71aa9dff669f903b1eb64b54142a9bd9f757f8fde65fd7"
 dependencies = [
  "once_cell",
  "target-lexicon",
@@ -1637,9 +1704,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-ffi"
-version = "0.22.6"
+version = "0.23.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "9ab5bcf04a2cdcbb50c7d6105de943f543f9ed92af55818fd17b660390fc8636"
+checksum = "dab6bb2102bd8f991e7749f130a70d05dd557613e39ed2deeee8e9ca0c4d548d"
 dependencies = [
  "libc",
  "pyo3-build-config",
@@ -1647,9 +1714,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros"
-version = "0.22.6"
+version = "0.23.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "0fd24d897903a9e6d80b968368a34e1525aeb719d568dba8b3d4bfa5dc67d453"
+checksum = "91871864b353fd5ffcb3f91f2f703a22a9797c91b9ab497b1acac7b07ae509c7"
 dependencies = [
  "proc-macro2",
  "pyo3-macros-backend",
@@ -1659,9 +1726,9 @@ dependencies = [
 
 [[package]]
 name = "pyo3-macros-backend"
-version = "0.22.6"
+version = "0.23.4"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "36c011a03ba1e50152b4b394b479826cad97e7a21eb52df179cd91ac411cbfbe"
+checksum = "43abc3b80bc20f3facd86cd3c60beed58c3e2aa26213f3cda368de39c60a27e4"
 dependencies = [
  "heck",
  "proc-macro2",
@@ -1864,7 +1931,7 @@ checksum = "f3cb5ba0dc43242ce17de99c180e96db90b235b8a9fdc9543c96d2209116bd9f"
 
 [[package]]
 name = "sbv2_api"
-version = "0.2.0-alpha"
+version = "0.2.0-alpha4"
 dependencies = [
  "anyhow",
  "axum",
@@ -1880,7 +1947,7 @@ dependencies = [
 
 [[package]]
 name = "sbv2_bindings"
-version = "0.2.0-alpha2"
+version = "0.2.0-alpha4"
 dependencies = [
  "anyhow",
  "ndarray",
@@ -1890,14 +1957,16 @@ dependencies = [
 
 [[package]]
 name = "sbv2_core"
-version = "0.2.0-alpha2"
+version = "0.2.0-alpha4"
 dependencies = [
  "anyhow",
+ "base64 0.22.1",
  "dotenvy",
  "env_logger",
  "hound",
  "jpreprocess",
  "ndarray",
+ "npyz",
  "num_cpus",
  "once_cell",
  "ort",
@@ -1930,18 +1999,18 @@ checksum = "94143f37725109f92c262ed2cf5e59bce7498c01bcc1502d7b9afe439a4e9f49"
 
 [[package]]
 name = "serde"
-version = "1.0.215"
+version = "1.0.217"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "6513c1ad0b11a9376da888e3e0baa0077f1aed55c17f50e7b2397136129fb88f"
+checksum = "02fc4265df13d6fa1d00ecff087228cc0a2b5f3c0e87e258d8b94a156e984c70"
 dependencies = [
  "serde_derive",
 ]
 
 [[package]]
 name = "serde_derive"
-version = "1.0.215"
+version = "1.0.217"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "ad1e866f866923f252f05c889987993144fb74e722403468a4ebd70c3cd756c0"
+checksum = "5a9bf7cf98d04a2b28aead066b7496853d4779c9cc183c440dbac457641e19a0"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -1950,9 +2019,9 @@ dependencies = [
 
 [[package]]
 name = "serde_json"
-version = "1.0.132"
+version = "1.0.137"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "d726bfaff4b320266d395898905d0eba0345aae23b54aee3a737e260fd46db03"
+checksum = "930cfb6e6abf99298aaad7d29abbef7a9999a9a8806a40088f55f0dcec03146b"
 dependencies = [
  "itoa",
  "memchr",
@@ -2104,12 +2173,6 @@ dependencies = [
  "unicode-ident",
 ]
 
-[[package]]
-name = "sync_wrapper"
-version = "0.1.2"
-source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2047c6ded9c721764247e62cd3b03c09ffc529b2ba5b10ec482ae507a4a70160"
-
 [[package]]
 name = "sync_wrapper"
 version = "1.0.1"
@@ -2170,9 +2233,9 @@ checksum = "1f3ccbac311fea05f86f61904b462b55fb3df8837a366dfc601a0161d0532f20"
 
 [[package]]
 name = "tokenizers"
-version = "0.20.3"
+version = "0.21.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "67b67c92f6d705e2a1d106fb0b28c696f9074901a9c656ee5d9f5de204c39bf7"
+checksum = "9ecededfed68a69bc657e486510089e255e53c3d38cc7d4d59c8742668ca2cae"
 dependencies = [
  "aho-corasick",
  "derive_builder",
@@ -2203,9 +2266,9 @@ dependencies = [
 
 [[package]]
 name = "tokio"
-version = "1.41.1"
+version = "1.43.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "22cfb5bee7a6a52939ca9224d6ac897bb669134078daa8735560897f69de4d33"
+checksum = "3d61fa4ffa3de412bfea335c6ecff681de2b609ba3c77ef3e00e521813a9ed9e"
 dependencies = [
  "backtrace",
  "bytes",
@@ -2221,9 +2284,9 @@ dependencies = [
 
 [[package]]
 name = "tokio-macros"
-version = "2.4.0"
+version = "2.5.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "693d596312e88961bc67d7f1f97af8a70227d9f90c31bba5806eec004978d752"
+checksum = "6e06d43f1345a3bcd39f6a56dbb7dcab2ba47e68e8ac134855e7e2bdbaf8cab8"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -2232,14 +2295,14 @@ dependencies = [
 
 [[package]]
 name = "tower"
-version = "0.5.1"
+version = "0.5.2"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "2873938d487c3cfb9aed7546dc9f2711d867c9f90c46b889989a2cb84eba6b4f"
+checksum = "d039ad9159c98b70ecfd540b2573b97f7f52c3e8d9f8ad57a24b916a536975f9"
 dependencies = [
  "futures-core",
  "futures-util",
  "pin-project-lite",
- "sync_wrapper 0.1.2",
+ "sync_wrapper",
  "tokio",
  "tower-layer",
  "tower-service",
@@ -2284,6 +2347,12 @@ version = "1.17.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
 checksum = "42ff0bf0c66b8238c6f3b578df37d0b7848e55df8577b3f74f92a69acceeb825"
 
+[[package]]
+name = "ucd-trie"
+version = "0.1.7"
+source = "registry+https://github.com/rust-lang/crates.io-index"
+checksum = "2896d95c02a80c6d6a5d6e953d479f5ddf2dfdb6a244441010e373ac0fb88971"
+
 [[package]]
 name = "unicode-bidi"
 version = "0.3.17"
@@ -2379,9 +2448,9 @@ checksum = "06abde3611657adf66d383f00b093d7faecc7fa57071cce2578660c9f1010821"
 
 [[package]]
 name = "utoipa"
-version = "5.2.0"
+version = "5.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "514a48569e4e21c86d0b84b5612b5e73c0b2cf09db63260134ba426d4e8ea714"
+checksum = "435c6f69ef38c9017b4b4eea965dfb91e71e53d869e896db40d1cf2441dd75c0"
 dependencies = [
  "indexmap",
  "serde",
@@ -2391,9 +2460,9 @@ dependencies = [
 
 [[package]]
 name = "utoipa-gen"
-version = "5.2.0"
+version = "5.3.1"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "5629efe65599d0ccd5d493688cbf6e03aa7c1da07fe59ff97cf5977ed0637f66"
+checksum = "a77d306bc75294fd52f3e99b13ece67c02c1a2789190a6f31d32f736624326f7"
 dependencies = [
  "proc-macro2",
  "quote",
@@ -2403,9 +2472,9 @@ dependencies = [
 
 [[package]]
 name = "utoipa-scalar"
-version = "0.2.0"
+version = "0.3.0"
 source = "registry+https://github.com/rust-lang/crates.io-index"
-checksum = "c1291aa7a2223c2f8399d1c6627ca0ba57ca0d7ecac762a2094a9dfd6376445a"
+checksum = "59559e1509172f6b26c1cdbc7247c4ddd1ac6560fe94b584f81ee489b141f719"
 dependencies = [
  "axum",
  "serde",

convert/convert_model.py

@@ -94,7 +94,7 @@ model = get_net_g(
 )
 
 
-def forward(x, x_len, sid, tone, lang, bert, style, length_scale, sdp_ratio):
+def forward(x, x_len, sid, tone, lang, bert, style, length_scale, sdp_ratio, noise_scale, noise_scale_w):
     return model.infer(
         x,
         x_len,
@@ -105,6 +105,8 @@ def forward(x, x_len, sid, tone, lang, bert, style, length_scale, sdp_ratio):
         style,
         sdp_ratio=sdp_ratio,
         length_scale=length_scale,
+        noise_scale=noise_scale,
+        noise_scale_w=noise_scale_w,
     )
 
 
@@ -122,6 +124,8 @@ torch.onnx.export(
         style_vec_tensor,
         torch.tensor(1.0),
         torch.tensor(0.0),
+        torch.tensor(0.6777),
+        torch.tensor(0.8),
     ),
     f"../models/model_{out_name}.onnx",
     verbose=True,
@@ -144,6 +148,8 @@ torch.onnx.export(
         "style_vec",
         "length_scale",
         "sdp_ratio",
+        "noise_scale",
+        "noise_scale_w"
     ],
     output_names=["output"],
 )

sbv2_api/Cargo.toml

@@ -1,19 +1,19 @@
 [package]
 name = "sbv2_api"
-version = "0.2.0-alpha"
+version = "0.2.0-alpha4"
 edition = "2021"
 
 [dependencies]
 anyhow.workspace = true
-axum = "0.7.5"
+axum = "0.8.0"
 dotenvy.workspace = true
 env_logger.workspace = true
 log = "0.4.22"
-sbv2_core = { version = "0.2.0-alpha2", path = "../sbv2_core" }
+sbv2_core = { version = "0.2.0-alpha2", path = "../sbv2_core", features = ["aivmx"] }
 serde = { version = "1.0.210", features = ["derive"] }
 tokio = { version = "1.40.0", features = ["full"] }
 utoipa = { version = "5.0.0", features = ["axum_extras"] }
-utoipa-scalar = { version = "0.2.0", features = ["axum"] }
+utoipa-scalar = { version = "0.3.0", features = ["axum"] }
 
 [features]
 coreml = ["sbv2_core/coreml"]

sbv2_api/src/main.rs

@@ -140,6 +140,20 @@ impl AppState {
                     log::warn!("Error loading {entry}: {e}");
                 };
                 log::info!("Loaded: {entry}");
+            } else if name.ends_with(".aivmx") {
+                let entry = &name[..name.len() - 6];
+                log::info!("Try loading: {entry}");
+                let aivmx_bytes = match fs::read(format!("{models}/{entry}.aivmx")).await {
+                    Ok(b) => b,
+                    Err(e) => {
+                        log::warn!("Error loading aivmx bytes from file {entry}: {e}");
+                        continue;
+                    }
+                };
+                if let Err(e) = tts_model.load_aivmx(entry, aivmx_bytes) {
+                    log::error!("Error loading {entry}: {e}");
+                }
+                log::info!("Loaded: {entry}");
             }
         }
         for entry in entries {

sbv2_bindings/Cargo.toml

@@ -1,6 +1,6 @@
 [package]
 name = "sbv2_bindings"
-version = "0.2.0-alpha2"
+version = "0.2.0-alpha4"
 edition = "2021"
 
 # See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
@@ -11,5 +11,5 @@ crate-type = ["cdylib"]
 [dependencies]
 anyhow.workspace = true
 ndarray.workspace = true
-pyo3 = { version = "0.22.0", features = ["anyhow"] }
+pyo3 = { version = "0.23.0", features = ["anyhow"] }
 sbv2_core = { version = "0.2.0-alpha2", path = "../sbv2_core" }

sbv2_bindings/src/sbv2.rs

@@ -157,7 +157,7 @@ impl TTSModel {
                 ..Default::default()
             },
         )?;
-        Ok(PyBytes::new_bound(py, &data))
+        Ok(PyBytes::new(py, &data))
     }
 
     fn unload(&mut self, ident: String) -> bool {

sbv2_core/Cargo.toml

@@ -1,7 +1,7 @@
 [package]
 name = "sbv2_core"
 description = "Style-Bert-VITSの推論ライブラリ"
-version = "0.2.0-alpha2"
+version = "0.2.0-alpha4"
 edition = "2021"
 license = "MIT"
 readme = "../README.md"
@@ -10,11 +10,13 @@ documentation = "https://docs.rs/sbv2_core"
 
 [dependencies]
 anyhow.workspace = true
+base64 = { version = "0.22.1", optional = true }
 dotenvy.workspace = true
 env_logger.workspace = true
 hound = "3.5.1"
 jpreprocess = { version = "0.10.0", features = ["naist-jdic"] }
 ndarray.workspace = true
+npyz = { version = "0.8.3", optional = true }
 num_cpus = "1.16.0"
 once_cell.workspace = true
 ort = { git = "https://github.com/pykeio/ort.git", version = "2.0.0-rc.8", optional = true }
@@ -23,7 +25,7 @@ serde = { version = "1.0.210", features = ["derive"] }
 serde_json = "1.0.128"
 tar = "0.4.41"
 thiserror = "1.0.63"
-tokenizers = { version = "0.20.0", default-features = false }
+tokenizers = { version = "0.21.0", default-features = false }
 zstd = "0.13.2"
 
 [features]
@@ -35,4 +37,6 @@ directml = ["ort/directml", "std"]
 tensorrt = ["ort/tensorrt", "std"]
 coreml = ["ort/coreml", "std"]
 default = ["std"]
-no_std = ["tokenizers/unstable_wasm"]
+no_std = ["tokenizers/unstable_wasm"]
+aivmx = ["npyz", "base64"]
+base64 = ["dep:base64"]

sbv2_core/src/error.rs

@@ -21,6 +21,9 @@ pub enum Error {
     HoundError(#[from] hound::Error),
     #[error("model not found error")]
     ModelNotFoundError(String),
+    #[cfg(feature = "base64")]
+    #[error("base64 error")]
+    Base64Error(#[from] base64::DecodeError),
     #[error("other")]
     OtherError(String),
 }

sbv2_core/src/main.rs

@@ -15,7 +15,14 @@ fn main_inner() -> anyhow::Result<()> {
             .ok()
             .and_then(|x| x.parse().ok()),
     )?;
-    tts_holder.load_sbv2file(ident, fs::read(env::var("MODEL_PATH")?)?)?;
+    #[cfg(not(feature = "aivmx"))]
+    {
+        tts_holder.load_sbv2file(ident, fs::read(env::var("MODEL_PATH")?)?)?;
+    }
+    #[cfg(feature = "aivmx")]
+    {
+        tts_holder.load_aivmx(ident, fs::read(env::var("MODEL_PATH")?)?)?;
+    }
 
     let audio =
         tts_holder.easy_synthesize(ident, &text, 0, 0, tts::SynthesizeOptions::default())?;
@@ -23,6 +30,7 @@ fn main_inner() -> anyhow::Result<()> {
 
     Ok(())
 }
+
 #[cfg(not(feature = "std"))]
 fn main_inner() -> anyhow::Result<()> {
     Ok(())

sbv2_core/src/model.rs

@@ -58,6 +58,8 @@ pub fn synthesize(
     style_vector: Array1<f32>,
     sdp_ratio: f32,
     length_scale: f32,
+    noise_scale: f32,
+    noise_scale_w: f32,
 ) -> Result<Array3<f32>> {
     let bert = bert_ori.insert_axis(Axis(0));
     let x_tst_lengths: Array1<i64> = array![x_tst.shape()[0] as i64];
@@ -75,6 +77,8 @@ pub fn synthesize(
         "style_vec" => style_vector,
         "sdp_ratio" => array![sdp_ratio],
         "length_scale" => array![length_scale],
+        "noise_scale" => array![noise_scale],
+        "noise_scale_w" => array![noise_scale_w]
     }?)?;
 
     let audio_array = outputs["output"]

sbv2_core/src/tts.rs

@@ -1,7 +1,13 @@
 use crate::error::{Error, Result};
 use crate::{jtalk, model, style, tokenizer, tts_util};
+#[cfg(feature = "aivmx")]
+use base64::prelude::{Engine as _, BASE64_STANDARD};
+#[cfg(feature = "aivmx")]
+use ndarray::ShapeBuilder;
 use ndarray::{concatenate, Array1, Array2, Array3, Axis};
 use ort::Session;
+#[cfg(feature = "aivmx")]
+use std::io::Cursor;
 use tokenizers::Tokenizer;
 
 #[derive(PartialEq, Eq, Clone)]
@@ -69,6 +75,53 @@ impl TTSModelHolder {
         self.models.iter().map(|m| m.ident.to_string()).collect()
     }
 
+    #[cfg(feature = "aivmx")]
+    pub fn load_aivmx<I: Into<TTSIdent>, P: AsRef<[u8]>>(
+        &mut self,
+        ident: I,
+        aivmx_bytes: P,
+    ) -> Result<()> {
+        let ident = ident.into();
+        if self.find_model(ident.clone()).is_err() {
+            let mut load = true;
+            if let Some(max) = self.max_loaded_models {
+                if self.models.iter().filter(|x| x.vits2.is_some()).count() >= max {
+                    load = false;
+                }
+            }
+            let model = model::load_model(&aivmx_bytes, false)?;
+            let metadata = model.metadata()?;
+            if let Some(aivm_style_vectors) = metadata.custom("aivm_style_vectors")? {
+                let aivm_style_vectors = BASE64_STANDARD.decode(aivm_style_vectors)?;
+                let style_vectors = Cursor::new(&aivm_style_vectors);
+                let reader = npyz::NpyFile::new(style_vectors)?;
+                let style_vectors = {
+                    let shape = reader.shape().to_vec();
+                    let order = reader.order();
+                    let data = reader.into_vec::<f32>()?;
+                    let shape = match shape[..] {
+                        [i1, i2] => [i1 as usize, i2 as usize],
+                        _ => panic!("expected 2D array"),
+                    };
+                    let true_shape = shape.set_f(order == npyz::Order::Fortran);
+                    ndarray::Array2::from_shape_vec(true_shape, data)?
+                };
+                drop(metadata);
+                self.models.push(TTSModel {
+                    vits2: if load { Some(model) } else { None },
+                    bytes: if self.max_loaded_models.is_some() {
+                        Some(aivmx_bytes.as_ref().to_vec())
+                    } else {
+                        None
+                    },
+                    ident,
+                    style_vectors,
+                })
+            }
+        }
+        Ok(())
+    }
+
     /// Load a .sbv2 file binary
     ///
     /// # Examples
@@ -257,6 +310,8 @@ impl TTSModelHolder {
                     style_vector.clone(),
                     options.sdp_ratio,
                     options.length_scale,
+                    0.677,
+                    0.8,
                 )?;
                 audios.push(audio.clone());
                 if i != texts.len() - 1 {
@@ -279,6 +334,8 @@ impl TTSModelHolder {
                 style_vector,
                 options.sdp_ratio,
                 options.length_scale,
+                0.677,
+                0.8,
             )?
         };
         tts_util::array_to_vec(audio_array)