WIP: support so-vits-svc 40v2. dstID list from server.

client/demo_so-vits-svc_40v2/src/102_model_setting.tsx

@@ -57,6 +57,7 @@ export const useModelSettingArea = (): ServerSettingState => {
                     file: file
                 }
             })
+
         }
         const onConfigFileClearClicked = () => {
             appState.serverSetting.setFileUploadSetting({

client/demo_so-vits-svc_40v2/src/105_speaker_setting.tsx

@@ -1,3 +1,4 @@
+import { ServerInfoSoVitsSVC } from "@dannadori/voice-changer-client-js";
 import React, { useMemo, useState } from "react"
 import { useAppState } from "./001_provider/001_AppStateProvider";
 import { AnimationTypes, HeaderButton, HeaderButtonProps } from "./components/101_HeaderButton";
@@ -5,6 +6,7 @@ import { AnimationTypes, HeaderButton, HeaderButtonProps } from "./components/10
 export const useSpeakerSetting = () => {
     const appState = useAppState()
 
+
     const accodionButton = useMemo(() => {
         const accodionButtonProps: HeaderButtonProps = {
             stateControlCheckbox: appState.frontendManagerState.stateControls.openSpeakerSettingCheckbox,
@@ -18,17 +20,26 @@ export const useSpeakerSetting = () => {
     }, []);
 
     const dstIdRow = useMemo(() => {
+        const settings = appState.serverSetting.serverSetting as ServerInfoSoVitsSVC
+        const speakers = settings.speakers
+        if (!speakers) {
+            return <></>
+        }
+
+        const currentValue = Object.values(speakers).includes(appState.serverSetting.serverSetting.dstId) ? appState.serverSetting.serverSetting.dstId : -1
+
         return (
             <div className="body-row split-3-2-1-4 left-padding-1 guided">
                 <div className="body-item-title left-padding-1">Destination Speaker Id</div>
                 <div className="body-select-container">
-                    <select className="body-select" value={appState.serverSetting.serverSetting.dstId} onChange={(e) => {
+                    <select className="body-select" value={currentValue} onChange={(e) => {
                         appState.serverSetting.updateServerSettings({ ...appState.serverSetting.serverSetting, dstId: Number(e.target.value) })
 
                     }}>
+                        <option key="unknown" value={-1}>unknwon(-1)</option>
                         {
-                            [0, 1, 2, 3, 4].map(x => {
+                            Object.keys(speakers).map(x => {
-                                return <option key={x} value={x}>{x}</option>
+                                return <option key={x} value={speakers[x]}>{x}({speakers[x]})</option>
                             })
                         }
                     </select>

client/lib/src/const.ts

@@ -112,6 +112,9 @@ export type ServerInfo = VoiceChangerServerSetting & {
     onnxExecutionProviders: OnnxExecutionProvider[]
 }
 
+export type ServerInfoSoVitsSVC = ServerInfo & {
+    speakers: { [key: string]: number }
+}
 export const DefaultServerSetting_MMVCv15: ServerInfo = {
     srcId: 0,
     dstId: 101,

server/voice_changer/SoVitsSvc40v2/SoVitsSvc40v2.py

@@ -11,7 +11,7 @@ else:
     sys.path.append("so-vits-svc-40v2")
 
 import io
-from dataclasses import dataclass, asdict
+from dataclasses import dataclass, asdict, field
 from functools import reduce
 import numpy as np
 import torch
@@ -29,7 +29,7 @@ providers = ['OpenVINOExecutionProvider', "CUDAExecutionProvider", "DmlExecution
 @dataclass
 class SoVitsSvc40v2Settings():
     gpu: int = 0
-    dstId: int = 0
+    dstId: int = -1
 
     f0Detector: str = "dio"  # dio or harvest
     tran: int = 20
@@ -44,6 +44,10 @@ class SoVitsSvc40v2Settings():
     onnxModelFile: str = ""
     configFile: str = ""
 
+    speakers: dict[str, int] = field(
+        default_factory=lambda: {}
+    )
+
     # ↓mutableな物だけ列挙
     intData = ["gpu", "dstId", "tran", "predictF0", "extraConvertSize"]
     floatData = ["noiceScale", "silentThreshold", "clusterInferRatio"]
@@ -65,38 +69,30 @@ class SoVitsSvc40v2:
 
         self.settings.configFile = config
         self.hps = utils.get_hparams_from_file(config)
+        self.settings.speakers = self.hps.spk
 
         # hubert model
         try:
             # vec_path = hubertTorchModel
             vec_path = "hubert/checkpoint_best_legacy_500.pt"
-            print("hubert 1 ", hubertTorchModel)
             models, saved_cfg, task = checkpoint_utils.load_model_ensemble_and_task(
                 [vec_path],
                 suffix="",
             )
-            print("hubert 2 ", hubertTorchModel)
             model = models[0]
-            print("hubert 3 ", hubertTorchModel)
             model.eval()
-            print("hubert 4 ", hubertTorchModel)
             self.hubert_model = model.cpu()
-            print("hubert 5 ", hubertTorchModel)
         except Exception as e:
-            print("EXCEPTION1", e)
+            print("EXCEPTION during loading hubert/contentvec model", e)
 
         # cluster
         try:
             if os.path.exists(clusterTorchModel):
-                print("load kmean11", clusterTorchModel)
                 self.cluster_model = cluster.get_cluster_model(clusterTorchModel)
-                print("load kmean12", clusterTorchModel)
             else:
-                print("load kmean21", clusterTorchModel)
                 self.cluster_model = None
-                print("load kmean22", clusterTorchModel)
         except Exception as e:
-            print("EXCEPTION2", e)
+            print("EXCEPTION during loading cluster model ", e)
 
         if pyTorch_model_file != None:
             self.settings.pyTorchModelFile = pyTorch_model_file