@livekit/agents-plugin-silero 0.4.6 → 0.5.0

package/README.md

@@ -0,0 +1,17 @@
+<!--
+SPDX-FileCopyrightText: 2024 LiveKit, Inc.
+
+SPDX-License-Identifier: Apache-2.0
+-->
+# Silero plugin for LiveKit Agents
+
+The Agents Framework is designed for building realtime, programmable
+participants that run on servers. Use it to create conversational, multi-modal
+voice agents that can see, hear, and understand.
+
+This package contains the Silero plugin, providing voice activity detection.
+Refer to the [documentation](https://docs.livekit.io/agents/overview/) for
+information on how to use it, or browse the [API
+reference](https://docs.livekit.io/agents-js/modules/plugins_agents_plugin_silero.html).
+See the [repository](https://github.com/livekit/agents-js) for more information
+about the framework as a whole.

package/dist/index.cjs

@@ -0,0 +1,31 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var src_exports = {};
+__export(src_exports, {
+  VAD: () => import_vad.VAD,
+  VADStream: () => import_vad.VADStream
+});
+module.exports = __toCommonJS(src_exports);
+var import_vad = require("./vad.cjs");
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  VAD,
+  VADStream
+});
+//# sourceMappingURL=index.cjs.map

package/dist/index.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/index.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2024 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nexport { VAD, VADStream } from './vad.js';\n"],"mappings":";;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;AAGA,iBAA+B;","names":[]}

package/dist/index.js

@@ -1,5 +1,6 @@
-// SPDX-FileCopyrightText: 2024 LiveKit, Inc.
-//
-// SPDX-License-Identifier: Apache-2.0
-export { VAD, VADStream } from './vad.js';
+import { VAD, VADStream } from "./vad.js";
+export {
+  VAD,
+  VADStream
+};
 //# sourceMappingURL=index.js.map

package/dist/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA,6CAA6C;AAC7C,EAAE;AACF,sCAAsC;AACtC,OAAO,EAAE,GAAG,EAAE,SAAS,EAAE,MAAM,UAAU,CAAC"}
+{"version":3,"sources":["../src/index.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2024 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nexport { VAD, VADStream } from './vad.js';\n"],"mappings":"AAGA,SAAS,KAAK,iBAAiB;","names":[]}

package/dist/onnx_model.cjs

@@ -0,0 +1,95 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var onnx_model_exports = {};
+__export(onnx_model_exports, {
+  OnnxModel: () => OnnxModel,
+  newInferenceSession: () => newInferenceSession
+});
+module.exports = __toCommonJS(onnx_model_exports);
+var getImportMetaUrl = () => typeof document === "undefined" ? new URL(`file:${__filename}`).href : document.currentScript && document.currentScript.src || new URL("main.js", document.baseURI).href;
+var importMetaUrl = /* @__PURE__ */ getImportMetaUrl();
+var import_node_url = require("node:url");
+var import_onnxruntime_node = require("onnxruntime-node");
+const newInferenceSession = (forceCPU) => {
+  return import_onnxruntime_node.InferenceSession.create((0, import_node_url.fileURLToPath)(new URL("silero_vad.onnx", importMetaUrl).href), {
+    interOpNumThreads: 1,
+    intraOpNumThreads: 1,
+    executionMode: "sequential",
+    executionProviders: forceCPU ? [{ name: "cpu" }] : void 0
+  });
+};
+class OnnxModel {
+  #session;
+  #sampleRate;
+  #windowSizeSamples;
+  #contextSize;
+  #sampleRateNd;
+  #context;
+  // #state: Float32Array;
+  #rnnState;
+  #inputBuffer;
+  constructor(session, sampleRate) {
+    this.#session = session;
+    this.#sampleRate = sampleRate;
+    switch (sampleRate) {
+      case 8e3:
+        this.#windowSizeSamples = 256;
+        this.#contextSize = 32;
+        break;
+      case 16e3:
+        this.#windowSizeSamples = 512;
+        this.#contextSize = 64;
+        break;
+    }
+    this.#sampleRateNd = BigInt64Array.from([BigInt(sampleRate)]);
+    this.#context = new Float32Array(this.#contextSize);
+    this.#rnnState = new Float32Array(2 * 1 * 128);
+    this.#inputBuffer = new Float32Array(this.#contextSize + this.#windowSizeSamples);
+  }
+  get sampleRate() {
+    return this.#sampleRate;
+  }
+  get windowSizeSamples() {
+    return this.#windowSizeSamples;
+  }
+  get contextSize() {
+    return this.#contextSize;
+  }
+  async run(x) {
+    this.#inputBuffer.set(this.#context, 0);
+    this.#inputBuffer.set(x, this.#contextSize);
+    return await this.#session.run({
+      input: new import_onnxruntime_node.Tensor("float32", this.#inputBuffer, [
+        1,
+        this.#contextSize + this.#windowSizeSamples
+      ]),
+      state: new import_onnxruntime_node.Tensor("float32", this.#rnnState, [2, 1, 128]),
+      sr: new import_onnxruntime_node.Tensor("int64", this.#sampleRateNd)
+    }).then((result) => {
+      this.#context = this.#inputBuffer.subarray(0, this.#contextSize);
+      return result.output.data.at(0);
+    });
+  }
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  OnnxModel,
+  newInferenceSession
+});
+//# sourceMappingURL=onnx_model.cjs.map

package/dist/onnx_model.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":["../src/onnx_model.ts","../../../node_modules/.pnpm/tsup@8.3.5_@microsoft+api-extractor@7.43.7_@types+node@22.5.5__postcss@8.4.38_tsx@4.19.2_typescript@5.4.5/node_modules/tsup/assets/cjs_shims.js"],"sourcesContent":["// SPDX-FileCopyrightText: 2024 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { fileURLToPath } from 'node:url';\nimport { InferenceSession, Tensor } from 'onnxruntime-node';\n\nexport type SampleRate = 8000 | 16000;\n\nexport const newInferenceSession = (forceCPU: boolean) => {\n  return InferenceSession.create(fileURLToPath(new URL('silero_vad.onnx', import.meta.url).href), {\n    interOpNumThreads: 1,\n    intraOpNumThreads: 1,\n    executionMode: 'sequential',\n    executionProviders: forceCPU ? [{ name: 'cpu' }] : undefined,\n  });\n};\n\nexport class OnnxModel {\n  #session: InferenceSession;\n  #sampleRate: number;\n  #windowSizeSamples: number;\n  #contextSize: number;\n  #sampleRateNd: BigInt64Array;\n  #context: Float32Array;\n  // #state: Float32Array;\n  #rnnState: Float32Array;\n  #inputBuffer: Float32Array;\n\n  constructor(session: InferenceSession, sampleRate: SampleRate) {\n    this.#session = session;\n    this.#sampleRate = sampleRate;\n\n    switch (sampleRate) {\n      case 8000:\n        this.#windowSizeSamples = 256;\n        this.#contextSize = 32;\n        break;\n      case 16000:\n        this.#windowSizeSamples = 512;\n        this.#contextSize = 64;\n        break;\n    }\n\n    this.#sampleRateNd = BigInt64Array.from([BigInt(sampleRate)]);\n    this.#context = new Float32Array(this.#contextSize);\n    this.#rnnState = new Float32Array(2 * 1 * 128);\n    this.#inputBuffer = new Float32Array(this.#contextSize + this.#windowSizeSamples);\n  }\n\n  get sampleRate(): number {\n    return this.#sampleRate;\n  }\n\n  get windowSizeSamples(): number {\n    return this.#windowSizeSamples;\n  }\n\n  get contextSize(): number {\n    return this.#contextSize;\n  }\n\n  async run(x: Float32Array): Promise<number> {\n    this.#inputBuffer.set(this.#context, 0);\n    this.#inputBuffer.set(x, this.#contextSize);\n\n    return await this.#session\n      .run({\n        input: new Tensor('float32', this.#inputBuffer, [\n          1,\n          this.#contextSize + this.#windowSizeSamples,\n        ]),\n        state: new Tensor('float32', this.#rnnState, [2, 1, 128]),\n        sr: new Tensor('int64', this.#sampleRateNd),\n      })\n      .then((result) => {\n        // this.#state = result.output.data as Float32Array,\n        this.#context = this.#inputBuffer.subarray(0, this.#contextSize);\n        return (result.output!.data as Float32Array).at(0)!;\n      });\n  }\n}\n","// Shim globals in cjs bundle\n// There's a weird bug that esbuild will always inject importMetaUrl\n// if we export it as `const importMetaUrl = ... __filename ...`\n// But using a function will not cause this issue\n\nconst getImportMetaUrl = () =>\n  typeof document === 'undefined'\n    ? new URL(`file:${__filename}`).href\n    : (document.currentScript && document.currentScript.src) ||\n      new URL('main.js', document.baseURI).href\n\nexport const importMetaUrl = /* @__PURE__ */ getImportMetaUrl()\n"],"mappings":";;;;;;;;;;;;;;;;;;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA;ACKA,IAAM,mBAAmB,MACvB,OAAO,aAAa,cAChB,IAAI,IAAI,QAAQ,UAAU,EAAE,EAAE,OAC7B,SAAS,iBAAiB,SAAS,cAAc,OAClD,IAAI,IAAI,WAAW,SAAS,OAAO,EAAE;AAEpC,IAAM,gBAAgC,iCAAiB;ADR9D,sBAA8B;AAC9B,8BAAyC;AAIlC,MAAM,sBAAsB,CAAC,aAAsB;AACxD,SAAO,yCAAiB,WAAO,+BAAc,IAAI,IAAI,mBAAmB,aAAe,EAAE,IAAI,GAAG;AAAA,IAC9F,mBAAmB;AAAA,IACnB,mBAAmB;AAAA,IACnB,eAAe;AAAA,IACf,oBAAoB,WAAW,CAAC,EAAE,MAAM,MAAM,CAAC,IAAI;AAAA,EACrD,CAAC;AACH;AAEO,MAAM,UAAU;AAAA,EACrB;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA;AAAA,EAEA;AAAA,EACA;AAAA,EAEA,YAAY,SAA2B,YAAwB;AAC7D,SAAK,WAAW;AAChB,SAAK,cAAc;AAEnB,YAAQ,YAAY;AAAA,MAClB,KAAK;AACH,aAAK,qBAAqB;AAC1B,aAAK,eAAe;AACpB;AAAA,MACF,KAAK;AACH,aAAK,qBAAqB;AAC1B,aAAK,eAAe;AACpB;AAAA,IACJ;AAEA,SAAK,gBAAgB,cAAc,KAAK,CAAC,OAAO,UAAU,CAAC,CAAC;AAC5D,SAAK,WAAW,IAAI,aAAa,KAAK,YAAY;AAClD,SAAK,YAAY,IAAI,aAAa,IAAI,IAAI,GAAG;AAC7C,SAAK,eAAe,IAAI,aAAa,KAAK,eAAe,KAAK,kBAAkB;AAAA,EAClF;AAAA,EAEA,IAAI,aAAqB;AACvB,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,IAAI,oBAA4B;AAC9B,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,IAAI,cAAsB;AACxB,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,MAAM,IAAI,GAAkC;AAC1C,SAAK,aAAa,IAAI,KAAK,UAAU,CAAC;AACtC,SAAK,aAAa,IAAI,GAAG,KAAK,YAAY;AAE1C,WAAO,MAAM,KAAK,SACf,IAAI;AAAA,MACH,OAAO,IAAI,+BAAO,WAAW,KAAK,cAAc;AAAA,QAC9C;AAAA,QACA,KAAK,eAAe,KAAK;AAAA,MAC3B,CAAC;AAAA,MACD,OAAO,IAAI,+BAAO,WAAW,KAAK,WAAW,CAAC,GAAG,GAAG,GAAG,CAAC;AAAA,MACxD,IAAI,IAAI,+BAAO,SAAS,KAAK,aAAa;AAAA,IAC5C,CAAC,EACA,KAAK,CAAC,WAAW;AAEhB,WAAK,WAAW,KAAK,aAAa,SAAS,GAAG,KAAK,YAAY;AAC/D,aAAQ,OAAO,OAAQ,KAAsB,GAAG,CAAC;AAAA,IACnD,CAAC;AAAA,EACL;AACF;","names":[]}

package/dist/onnx_model.d.ts.map

@@ -1 +1 @@
-{"version":3,"file":"onnx_model.d.ts","sourceRoot":"","sources":["../src/onnx_model.ts"],"names":[],"mappings":";AAGA,OAAO,EAAE,gBAAgB,EAAU,MAAM,kBAAkB,CAAC;AAE5D,MAAM,MAAM,UAAU,GAAG,IAAI,GAAG,KAAK,CAAC;AAEtC,eAAO,MAAM,mBAAmB,aAAc,OAAO,8BAOpD,CAAC;AAEF,qBAAa,SAAS;;gBAWR,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,UAAU;IAqB7D,IAAI,UAAU,IAAI,MAAM,CAEvB;IAED,IAAI,iBAAiB,IAAI,MAAM,CAE9B;IAED,IAAI,WAAW,IAAI,MAAM,CAExB;IAEK,GAAG,CAAC,CAAC,EAAE,YAAY,GAAG,OAAO,CAAC,MAAM,CAAC;CAmB5C"}
+{"version":3,"file":"onnx_model.d.ts","sourceRoot":"","sources":["../src/onnx_model.ts"],"names":[],"mappings":";AAIA,OAAO,EAAE,gBAAgB,EAAU,MAAM,kBAAkB,CAAC;AAE5D,MAAM,MAAM,UAAU,GAAG,IAAI,GAAG,KAAK,CAAC;AAEtC,eAAO,MAAM,mBAAmB,aAAc,OAAO,8BAOpD,CAAC;AAEF,qBAAa,SAAS;;gBAWR,OAAO,EAAE,gBAAgB,EAAE,UAAU,EAAE,UAAU;IAqB7D,IAAI,UAAU,IAAI,MAAM,CAEvB;IAED,IAAI,iBAAiB,IAAI,MAAM,CAE9B;IAED,IAAI,WAAW,IAAI,MAAM,CAExB;IAEK,GAAG,CAAC,CAAC,EAAE,YAAY,GAAG,OAAO,CAAC,MAAM,CAAC;CAmB5C"}

package/dist/onnx_model.js

@@ -1,69 +1,68 @@
-// SPDX-FileCopyrightText: 2024 LiveKit, Inc.
-//
-// SPDX-License-Identifier: Apache-2.0
-import { InferenceSession, Tensor } from 'onnxruntime-node';
-export const newInferenceSession = (forceCPU) => {
-    return InferenceSession.create(new URL('silero_vad.onnx', import.meta.url).pathname, {
-        interOpNumThreads: 1,
-        intraOpNumThreads: 1,
-        executionMode: 'sequential',
-        executionProviders: forceCPU ? [{ name: 'cpu' }] : undefined,
-    });
+import { fileURLToPath } from "node:url";
+import { InferenceSession, Tensor } from "onnxruntime-node";
+const newInferenceSession = (forceCPU) => {
+  return InferenceSession.create(fileURLToPath(new URL("silero_vad.onnx", import.meta.url).href), {
+    interOpNumThreads: 1,
+    intraOpNumThreads: 1,
+    executionMode: "sequential",
+    executionProviders: forceCPU ? [{ name: "cpu" }] : void 0
+  });
 };
-export class OnnxModel {
-    #session;
-    #sampleRate;
-    #windowSizeSamples;
-    #contextSize;
-    #sampleRateNd;
-    #context;
-    // #state: Float32Array;
-    #rnnState;
-    #inputBuffer;
-    constructor(session, sampleRate) {
-        this.#session = session;
-        this.#sampleRate = sampleRate;
-        switch (sampleRate) {
-            case 8000:
-                this.#windowSizeSamples = 256;
-                this.#contextSize = 32;
-                break;
-            case 16000:
-                this.#windowSizeSamples = 512;
-                this.#contextSize = 64;
-                break;
-        }
-        this.#sampleRateNd = BigInt64Array.from([BigInt(sampleRate)]);
-        this.#context = new Float32Array(this.#contextSize);
-        this.#rnnState = new Float32Array(2 * 1 * 128);
-        this.#inputBuffer = new Float32Array(this.#contextSize + this.#windowSizeSamples);
-    }
-    get sampleRate() {
-        return this.#sampleRate;
-    }
-    get windowSizeSamples() {
-        return this.#windowSizeSamples;
-    }
-    get contextSize() {
-        return this.#contextSize;
-    }
-    async run(x) {
-        this.#inputBuffer.set(this.#context, 0);
-        this.#inputBuffer.set(x, this.#contextSize);
-        return await this.#session
-            .run({
-            input: new Tensor('float32', this.#inputBuffer, [
-                1,
-                this.#contextSize + this.#windowSizeSamples,
-            ]),
-            state: new Tensor('float32', this.#rnnState, [2, 1, 128]),
-            sr: new Tensor('int64', this.#sampleRateNd),
-        })
-            .then((result) => {
-            // this.#state = result.output.data as Float32Array,
-            this.#context = this.#inputBuffer.subarray(0, this.#contextSize);
-            return result.output.data.at(0);
-        });
+class OnnxModel {
+  #session;
+  #sampleRate;
+  #windowSizeSamples;
+  #contextSize;
+  #sampleRateNd;
+  #context;
+  // #state: Float32Array;
+  #rnnState;
+  #inputBuffer;
+  constructor(session, sampleRate) {
+    this.#session = session;
+    this.#sampleRate = sampleRate;
+    switch (sampleRate) {
+      case 8e3:
+        this.#windowSizeSamples = 256;
+        this.#contextSize = 32;
+        break;
+      case 16e3:
+        this.#windowSizeSamples = 512;
+        this.#contextSize = 64;
+        break;
     }
+    this.#sampleRateNd = BigInt64Array.from([BigInt(sampleRate)]);
+    this.#context = new Float32Array(this.#contextSize);
+    this.#rnnState = new Float32Array(2 * 1 * 128);
+    this.#inputBuffer = new Float32Array(this.#contextSize + this.#windowSizeSamples);
+  }
+  get sampleRate() {
+    return this.#sampleRate;
+  }
+  get windowSizeSamples() {
+    return this.#windowSizeSamples;
+  }
+  get contextSize() {
+    return this.#contextSize;
+  }
+  async run(x) {
+    this.#inputBuffer.set(this.#context, 0);
+    this.#inputBuffer.set(x, this.#contextSize);
+    return await this.#session.run({
+      input: new Tensor("float32", this.#inputBuffer, [
+        1,
+        this.#contextSize + this.#windowSizeSamples
+      ]),
+      state: new Tensor("float32", this.#rnnState, [2, 1, 128]),
+      sr: new Tensor("int64", this.#sampleRateNd)
+    }).then((result) => {
+      this.#context = this.#inputBuffer.subarray(0, this.#contextSize);
+      return result.output.data.at(0);
+    });
+  }
 }
+export {
+  OnnxModel,
+  newInferenceSession
+};
 //# sourceMappingURL=onnx_model.js.map

package/dist/onnx_model.js.map

@@ -1 +1 @@
-{"version":3,"file":"onnx_model.js","sourceRoot":"","sources":["../src/onnx_model.ts"],"names":[],"mappings":"AAAA,6CAA6C;AAC7C,EAAE;AACF,sCAAsC;AACtC,OAAO,EAAE,gBAAgB,EAAE,MAAM,EAAE,MAAM,kBAAkB,CAAC;AAI5D,MAAM,CAAC,MAAM,mBAAmB,GAAG,CAAC,QAAiB,EAAE,EAAE;IACvD,OAAO,gBAAgB,CAAC,MAAM,CAAC,IAAI,GAAG,CAAC,iBAAiB,EAAE,MAAM,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC,QAAQ,EAAE;QACnF,iBAAiB,EAAE,CAAC;QACpB,iBAAiB,EAAE,CAAC;QACpB,aAAa,EAAE,YAAY;QAC3B,kBAAkB,EAAE,QAAQ,CAAC,CAAC,CAAC,CAAC,EAAE,IAAI,EAAE,KAAK,EAAE,CAAC,CAAC,CAAC,CAAC,SAAS;KAC7D,CAAC,CAAC;AACL,CAAC,CAAC;AAEF,MAAM,OAAO,SAAS;IACpB,QAAQ,CAAmB;IAC3B,WAAW,CAAS;IACpB,kBAAkB,CAAS;IAC3B,YAAY,CAAS;IACrB,aAAa,CAAgB;IAC7B,QAAQ,CAAe;IACvB,wBAAwB;IACxB,SAAS,CAAe;IACxB,YAAY,CAAe;IAE3B,YAAY,OAAyB,EAAE,UAAsB;QAC3D,IAAI,CAAC,QAAQ,GAAG,OAAO,CAAC;QACxB,IAAI,CAAC,WAAW,GAAG,UAAU,CAAC;QAE9B,QAAQ,UAAU,EAAE,CAAC;YACnB,KAAK,IAAI;gBACP,IAAI,CAAC,kBAAkB,GAAG,GAAG,CAAC;gBAC9B,IAAI,CAAC,YAAY,GAAG,EAAE,CAAC;gBACvB,MAAM;YACR,KAAK,KAAK;gBACR,IAAI,CAAC,kBAAkB,GAAG,GAAG,CAAC;gBAC9B,IAAI,CAAC,YAAY,GAAG,EAAE,CAAC;gBACvB,MAAM;QACV,CAAC;QAED,IAAI,CAAC,aAAa,GAAG,aAAa,CAAC,IAAI,CAAC,CAAC,MAAM,CAAC,UAAU,CAAC,CAAC,CAAC,CAAC;QAC9D,IAAI,CAAC,QAAQ,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,YAAY,CAAC,CAAC;QACpD,IAAI,CAAC,SAAS,GAAG,IAAI,YAAY,CAAC,CAAC,GAAG,CAAC,GAAG,GAAG,CAAC,CAAC;QAC/C,IAAI,CAAC,YAAY,GAAG,IAAI,YAAY,CAAC,IAAI,CAAC,YAAY,GAAG,IAAI,CAAC,kBAAkB,CAAC,CAAC;IACpF,CAAC;IAED,IAAI,UAAU;QACZ,OAAO,IAAI,CAAC,WAAW,CAAC;IAC1B,CAAC;IAED,IAAI,iBAAiB;QACnB,OAAO,IAAI,CAAC,kBAAkB,CAAC;IACjC,CAAC;IAED,IAAI,WAAW;QACb,OAAO,IAAI,CAAC,YAAY,CAAC;IAC3B,CAAC;IAED,KAAK,CAAC,GAAG,CAAC,CAAe;QACvB,IAAI,CAAC,YAAY,CAAC,GAAG,CAAC,IAAI,CAAC,QAAQ,EAAE,CAAC,CAAC,CAAC;QACxC,IAAI,CAAC,YAAY,CAAC,GAAG,CAAC,CAAC,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;QAE5C,OAAO,MAAM,IAAI,CAAC,QAAQ;aACvB,GAAG,CAAC;YACH,KAAK,EAAE,IAAI,MAAM,CAAC,SAAS,EAAE,IAAI,CAAC,YAAY,EAAE;gBAC9C,CAAC;gBACD,IAAI,CAAC,YAAY,GAAG,IAAI,CAAC,kBAAkB;aAC5C,CAAC;YACF,KAAK,EAAE,IAAI,MAAM,CAAC,SAAS,EAAE,IAAI,CAAC,SAAS,EAAE,CAAC,CAAC,EAAE,CAAC,EAAE,GAAG,CAAC,CAAC;YACzD,EAAE,EAAE,IAAI,MAAM,CAAC,OAAO,EAAE,IAAI,CAAC,aAAa,CAAC;SAC5C,CAAC;aACD,IAAI,CAAC,CAAC,MAAM,EAAE,EAAE;YACf,oDAAoD;YACpD,IAAI,CAAC,QAAQ,GAAG,IAAI,CAAC,YAAY,CAAC,QAAQ,CAAC,CAAC,EAAE,IAAI,CAAC,YAAY,CAAC,CAAC;YACjE,OAAQ,MAAM,CAAC,MAAO,CAAC,IAAqB,CAAC,EAAE,CAAC,CAAC,CAAE,CAAC;QACtD,CAAC,CAAC,CAAC;IACP,CAAC;CACF"}
+{"version":3,"sources":["../src/onnx_model.ts"],"sourcesContent":["// SPDX-FileCopyrightText: 2024 LiveKit, Inc.\n//\n// SPDX-License-Identifier: Apache-2.0\nimport { fileURLToPath } from 'node:url';\nimport { InferenceSession, Tensor } from 'onnxruntime-node';\n\nexport type SampleRate = 8000 | 16000;\n\nexport const newInferenceSession = (forceCPU: boolean) => {\n  return InferenceSession.create(fileURLToPath(new URL('silero_vad.onnx', import.meta.url).href), {\n    interOpNumThreads: 1,\n    intraOpNumThreads: 1,\n    executionMode: 'sequential',\n    executionProviders: forceCPU ? [{ name: 'cpu' }] : undefined,\n  });\n};\n\nexport class OnnxModel {\n  #session: InferenceSession;\n  #sampleRate: number;\n  #windowSizeSamples: number;\n  #contextSize: number;\n  #sampleRateNd: BigInt64Array;\n  #context: Float32Array;\n  // #state: Float32Array;\n  #rnnState: Float32Array;\n  #inputBuffer: Float32Array;\n\n  constructor(session: InferenceSession, sampleRate: SampleRate) {\n    this.#session = session;\n    this.#sampleRate = sampleRate;\n\n    switch (sampleRate) {\n      case 8000:\n        this.#windowSizeSamples = 256;\n        this.#contextSize = 32;\n        break;\n      case 16000:\n        this.#windowSizeSamples = 512;\n        this.#contextSize = 64;\n        break;\n    }\n\n    this.#sampleRateNd = BigInt64Array.from([BigInt(sampleRate)]);\n    this.#context = new Float32Array(this.#contextSize);\n    this.#rnnState = new Float32Array(2 * 1 * 128);\n    this.#inputBuffer = new Float32Array(this.#contextSize + this.#windowSizeSamples);\n  }\n\n  get sampleRate(): number {\n    return this.#sampleRate;\n  }\n\n  get windowSizeSamples(): number {\n    return this.#windowSizeSamples;\n  }\n\n  get contextSize(): number {\n    return this.#contextSize;\n  }\n\n  async run(x: Float32Array): Promise<number> {\n    this.#inputBuffer.set(this.#context, 0);\n    this.#inputBuffer.set(x, this.#contextSize);\n\n    return await this.#session\n      .run({\n        input: new Tensor('float32', this.#inputBuffer, [\n          1,\n          this.#contextSize + this.#windowSizeSamples,\n        ]),\n        state: new Tensor('float32', this.#rnnState, [2, 1, 128]),\n        sr: new Tensor('int64', this.#sampleRateNd),\n      })\n      .then((result) => {\n        // this.#state = result.output.data as Float32Array,\n        this.#context = this.#inputBuffer.subarray(0, this.#contextSize);\n        return (result.output!.data as Float32Array).at(0)!;\n      });\n  }\n}\n"],"mappings":"AAGA,SAAS,qBAAqB;AAC9B,SAAS,kBAAkB,cAAc;AAIlC,MAAM,sBAAsB,CAAC,aAAsB;AACxD,SAAO,iBAAiB,OAAO,cAAc,IAAI,IAAI,mBAAmB,YAAY,GAAG,EAAE,IAAI,GAAG;AAAA,IAC9F,mBAAmB;AAAA,IACnB,mBAAmB;AAAA,IACnB,eAAe;AAAA,IACf,oBAAoB,WAAW,CAAC,EAAE,MAAM,MAAM,CAAC,IAAI;AAAA,EACrD,CAAC;AACH;AAEO,MAAM,UAAU;AAAA,EACrB;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA,EACA;AAAA;AAAA,EAEA;AAAA,EACA;AAAA,EAEA,YAAY,SAA2B,YAAwB;AAC7D,SAAK,WAAW;AAChB,SAAK,cAAc;AAEnB,YAAQ,YAAY;AAAA,MAClB,KAAK;AACH,aAAK,qBAAqB;AAC1B,aAAK,eAAe;AACpB;AAAA,MACF,KAAK;AACH,aAAK,qBAAqB;AAC1B,aAAK,eAAe;AACpB;AAAA,IACJ;AAEA,SAAK,gBAAgB,cAAc,KAAK,CAAC,OAAO,UAAU,CAAC,CAAC;AAC5D,SAAK,WAAW,IAAI,aAAa,KAAK,YAAY;AAClD,SAAK,YAAY,IAAI,aAAa,IAAI,IAAI,GAAG;AAC7C,SAAK,eAAe,IAAI,aAAa,KAAK,eAAe,KAAK,kBAAkB;AAAA,EAClF;AAAA,EAEA,IAAI,aAAqB;AACvB,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,IAAI,oBAA4B;AAC9B,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,IAAI,cAAsB;AACxB,WAAO,KAAK;AAAA,EACd;AAAA,EAEA,MAAM,IAAI,GAAkC;AAC1C,SAAK,aAAa,IAAI,KAAK,UAAU,CAAC;AACtC,SAAK,aAAa,IAAI,GAAG,KAAK,YAAY;AAE1C,WAAO,MAAM,KAAK,SACf,IAAI;AAAA,MACH,OAAO,IAAI,OAAO,WAAW,KAAK,cAAc;AAAA,QAC9C;AAAA,QACA,KAAK,eAAe,KAAK;AAAA,MAC3B,CAAC;AAAA,MACD,OAAO,IAAI,OAAO,WAAW,KAAK,WAAW,CAAC,GAAG,GAAG,GAAG,CAAC;AAAA,MACxD,IAAI,IAAI,OAAO,SAAS,KAAK,aAAa;AAAA,IAC5C,CAAC,EACA,KAAK,CAAC,WAAW;AAEhB,WAAK,WAAW,KAAK,aAAa,SAAS,GAAG,KAAK,YAAY;AAC/D,aAAQ,OAAO,OAAQ,KAAsB,GAAG,CAAC;AAAA,IACnD,CAAC;AAAA,EACL;AACF;","names":[]}

package/dist/onnxruntime.d.cjs

@@ -0,0 +1,2 @@
+"use strict";
+//# sourceMappingURL=onnxruntime.d.cjs.map

package/dist/onnxruntime.d.cjs.map

@@ -0,0 +1 @@
+{"version":3,"sources":[],"sourcesContent":[],"mappings":"","names":[]}

package/dist/onnxruntime.d.js

@@ -0,0 +1 @@
+//# sourceMappingURL=onnxruntime.d.js.map

package/dist/onnxruntime.d.js.map

@@ -0,0 +1 @@
+{"version":3,"sources":[],"sourcesContent":[],"mappings":"","names":[]}

package/dist/vad.cjs

@@ -0,0 +1,285 @@
+"use strict";
+var __defProp = Object.defineProperty;
+var __getOwnPropDesc = Object.getOwnPropertyDescriptor;
+var __getOwnPropNames = Object.getOwnPropertyNames;
+var __hasOwnProp = Object.prototype.hasOwnProperty;
+var __export = (target, all) => {
+  for (var name in all)
+    __defProp(target, name, { get: all[name], enumerable: true });
+};
+var __copyProps = (to, from, except, desc) => {
+  if (from && typeof from === "object" || typeof from === "function") {
+    for (let key of __getOwnPropNames(from))
+      if (!__hasOwnProp.call(to, key) && key !== except)
+        __defProp(to, key, { get: () => from[key], enumerable: !(desc = __getOwnPropDesc(from, key)) || desc.enumerable });
+  }
+  return to;
+};
+var __toCommonJS = (mod) => __copyProps(__defProp({}, "__esModule", { value: true }), mod);
+var vad_exports = {};
+__export(vad_exports, {
+  VAD: () => VAD,
+  VADStream: () => VADStream
+});
+module.exports = __toCommonJS(vad_exports);
+var import_agents = require("@livekit/agents");
+var import_rtc_node = require("@livekit/rtc-node");
+var import_onnx_model = require("./onnx_model.cjs");
+const SLOW_INFERENCE_THRESHOLD = 200;
+const defaultVADOptions = {
+  minSpeechDuration: 50,
+  minSilenceDuration: 250,
+  prefixPaddingDuration: 500,
+  maxBufferedSpeech: 6e4,
+  activationThreshold: 0.5,
+  sampleRate: 16e3,
+  forceCPU: true
+};
+class VAD extends import_agents.VAD {
+  #session;
+  #opts;
+  constructor(session, opts) {
+    super({ updateInterval: 32 });
+    this.#session = session;
+    this.#opts = opts;
+  }
+  /**
+   * Load and initialize the Silero VAD model.
+   *
+   * This method loads the ONNX model and prepares it for inference. When options are not provided,
+   * sane defaults are used.
+   *
+   * @remarks
+   * This method may take time to load the model into memory.
+   * It is recommended to call this method inside your prewarm mechanism.
+   *
+   * @example
+   * ```ts
+   * export default defineAgent({
+   *   prewarm: async (proc: JobProcess) => {
+   *     proc.userData.vad = await VAD.load();
+   *   },
+   *   entry: async (ctx: JobContext) => {
+   *     const vad = ctx.proc.userData.vad! as VAD;
+   *     // the rest of your agent logic
+   *   },
+   * });
+   * ```
+   *
+   * @param options -
+   * @returns Promise\<{@link VAD}\>: An instance of the VAD class ready for streaming.
+   */
+  static async load(opts = {}) {
+    const mergedOpts = { ...defaultVADOptions, ...opts };
+    const session = await (0, import_onnx_model.newInferenceSession)(mergedOpts.forceCPU);
+    return new VAD(session, mergedOpts);
+  }
+  stream() {
+    return new VADStream(this.#opts, new import_onnx_model.OnnxModel(this.#session, this.#opts.sampleRate));
+  }
+}
+class VADStream extends import_agents.VADStream {
+  #opts;
+  #model;
+  #task;
+  #expFilter = new import_agents.ExpFilter(0.35);
+  #extraInferenceTime = 0;
+  #logger = (0, import_agents.log)();
+  constructor(opts, model) {
+    super();
+    this.#opts = opts;
+    this.#model = model;
+    this.#task = new Promise(async () => {
+      let inferenceData = new Float32Array(this.#model.windowSizeSamples);
+      let speechBuffer = null;
+      let speechBufferMaxReached = false;
+      let speechBufferIndex = 0;
+      let pubSpeaking = false;
+      let pubSpeechDuration = 0;
+      let pubSilenceDuration = 0;
+      let pubCurrentSample = 0;
+      let pubTimestamp = 0;
+      let pubSampleRate = 0;
+      let pubPrefixPaddingSamples = 0;
+      let speechThresholdDuration = 0;
+      let silenceThresholdDuration = 0;
+      let inputFrames = [];
+      let inferenceFrames = [];
+      let resampler = null;
+      let inputCopyRemainingFrac = 0;
+      for await (const frame of this.input) {
+        if (typeof frame === "symbol") {
+          continue;
+        }
+        if (!pubSampleRate || !speechBuffer) {
+          pubSampleRate = frame.sampleRate;
+          pubPrefixPaddingSamples = Math.trunc(
+            this.#opts.prefixPaddingDuration * pubSampleRate / 1e3
+          );
+          speechBuffer = new Int16Array(
+            this.#opts.maxBufferedSpeech * pubSampleRate + pubPrefixPaddingSamples
+          );
+          if (this.#opts.sampleRate !== pubSampleRate) {
+            resampler = new import_rtc_node.AudioResampler(
+              pubSampleRate,
+              this.#opts.sampleRate,
+              1,
+              import_rtc_node.AudioResamplerQuality.QUICK
+              // VAD doesn't need high quality
+            );
+          }
+        } else if (frame.sampleRate !== pubSampleRate) {
+          this.#logger.error("a frame with a different sample rate was already published");
+          continue;
+        }
+        inputFrames.push(frame);
+        if (resampler) {
+          inferenceFrames.push(...resampler.push(frame));
+        } else {
+          inferenceFrames.push(frame);
+        }
+        while (true) {
+          const startTime = process.hrtime.bigint();
+          const availableInferenceSamples = inferenceFrames.map((x) => x.samplesPerChannel).reduce((acc, x) => acc + x, 0);
+          if (availableInferenceSamples < this.#model.windowSizeSamples) {
+            break;
+          }
+          const inputFrame = (0, import_agents.mergeFrames)(inputFrames);
+          const inferenceFrame = (0, import_agents.mergeFrames)(inferenceFrames);
+          inferenceData = Float32Array.from(
+            inferenceFrame.data.subarray(0, this.#model.windowSizeSamples),
+            (x) => x / 32767
+          );
+          const p = await this.#model.run(inferenceData).then((data) => this.#expFilter.apply(1, data));
+          const windowDuration = this.#model.windowSizeSamples / this.#opts.sampleRate * 1e3;
+          pubCurrentSample += this.#model.windowSizeSamples;
+          pubTimestamp += windowDuration;
+          const resamplingRatio = pubSampleRate / this.#model.sampleRate;
+          const toCopy = this.#model.windowSizeSamples * resamplingRatio + inputCopyRemainingFrac;
+          const toCopyInt = Math.trunc(toCopy);
+          inputCopyRemainingFrac = toCopy - toCopyInt;
+          const availableSpace = speechBuffer.length - speechBufferIndex;
+          const toCopyBuffer = Math.min(this.#model.windowSizeSamples, availableSpace);
+          if (toCopyBuffer > 0) {
+            speechBuffer.set(inputFrame.data.subarray(0, toCopyBuffer), speechBufferIndex);
+            speechBufferIndex += toCopyBuffer;
+          } else if (!speechBufferMaxReached) {
+            speechBufferMaxReached = true;
+            this.#logger.warn(
+              "maxBufferedSpeech reached, ignoring further data for the current speech input"
+            );
+          }
+          const inferenceDuration = Number((process.hrtime.bigint() - startTime) / BigInt(1e6));
+          this.#extraInferenceTime = Math.max(
+            0,
+            this.#extraInferenceTime + inferenceDuration - windowDuration
+          );
+          if (this.#extraInferenceTime > SLOW_INFERENCE_THRESHOLD) {
+            this.#logger.child({ delay: this.#extraInferenceTime }).warn("inference is slower than realtime");
+          }
+          if (pubSpeaking) {
+            pubSpeechDuration += inferenceDuration;
+          } else {
+            pubSilenceDuration += inferenceDuration;
+          }
+          this.queue.put({
+            type: import_agents.VADEventType.INFERENCE_DONE,
+            samplesIndex: pubCurrentSample,
+            timestamp: pubTimestamp,
+            silenceDuration: pubSilenceDuration,
+            speechDuration: pubSpeechDuration,
+            probability: p,
+            inferenceDuration,
+            frames: [
+              new import_rtc_node.AudioFrame(inputFrame.data.subarray(0, toCopyInt), pubSampleRate, 1, toCopyInt)
+            ],
+            speaking: pubSpeaking
+          });
+          const resetWriteCursor = () => {
+            if (!speechBuffer) throw new Error("speechBuffer is empty");
+            if (speechBufferIndex <= pubPrefixPaddingSamples) {
+              return;
+            }
+            const paddingData = speechBuffer.subarray(
+              speechBufferIndex - pubPrefixPaddingSamples,
+              speechBufferIndex
+            );
+            speechBuffer.set(paddingData, 0);
+            speechBufferIndex = pubPrefixPaddingSamples;
+            speechBufferMaxReached = false;
+          };
+          const copySpeechBuffer = () => {
+            if (!speechBuffer) throw new Error("speechBuffer is empty");
+            return new import_rtc_node.AudioFrame(
+              speechBuffer.subarray(0, speechBufferIndex),
+              pubSampleRate,
+              1,
+              speechBufferIndex
+            );
+          };
+          if (p > this.#opts.activationThreshold) {
+            speechThresholdDuration += windowDuration;
+            silenceThresholdDuration = 0;
+            if (!pubSpeaking && speechThresholdDuration >= this.#opts.minSpeechDuration) {
+              pubSpeaking = true;
+              pubSilenceDuration = 0;
+              pubSpeechDuration = speechThresholdDuration;
+              this.queue.put({
+                type: import_agents.VADEventType.START_OF_SPEECH,
+                samplesIndex: pubCurrentSample,
+                timestamp: pubTimestamp,
+                silenceDuration: pubSilenceDuration,
+                speechDuration: pubSpeechDuration,
+                probability: p,
+                inferenceDuration,
+                frames: [copySpeechBuffer()],
+                speaking: pubSpeaking
+              });
+            }
+          } else {
+            silenceThresholdDuration += windowDuration;
+            speechThresholdDuration = 0;
+            if (!pubSpeaking) {
+              resetWriteCursor();
+            }
+            if (pubSpeaking && silenceThresholdDuration > this.#opts.minSilenceDuration) {
+              pubSpeaking = false;
+              pubSpeechDuration = 0;
+              pubSilenceDuration = silenceThresholdDuration;
+              this.queue.put({
+                type: import_agents.VADEventType.END_OF_SPEECH,
+                samplesIndex: pubCurrentSample,
+                timestamp: pubTimestamp,
+                silenceDuration: pubSilenceDuration,
+                speechDuration: pubSpeechDuration,
+                probability: p,
+                inferenceDuration,
+                frames: [copySpeechBuffer()],
+                speaking: pubSpeaking
+              });
+              resetWriteCursor();
+            }
+          }
+          inputFrames = [];
+          inferenceFrames = [];
+          if (inputFrame.data.length > toCopyInt) {
+            const data = inputFrame.data.subarray(toCopyInt);
+            inputFrames.push(new import_rtc_node.AudioFrame(data, pubSampleRate, 1, Math.trunc(data.length / 2)));
+          }
+          if (inferenceFrame.data.length > this.#model.windowSizeSamples) {
+            const data = inferenceFrame.data.subarray(this.#model.windowSizeSamples);
+            inferenceFrames.push(
+              new import_rtc_node.AudioFrame(data, this.#opts.sampleRate, 1, Math.trunc(data.length / 2))
+            );
+          }
+        }
+      }
+    });
+  }
+}
+// Annotate the CommonJS export names for ESM import in node:
+0 && (module.exports = {
+  VAD,
+  VADStream
+});
+//# sourceMappingURL=vad.cjs.map