sarvamai 1.1.9 → 1.1.10-alpha.1
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/dist/cjs/BaseClient.js +2 -2
- package/dist/cjs/api/resources/speechToText/client/Client.js +5 -0
- package/dist/cjs/api/resources/speechToText/client/requests/SpeechToTextTranscriptionRequest.d.ts +2 -0
- package/dist/cjs/api/resources/speechToTextJob/client/Client.d.ts +2 -0
- package/dist/cjs/api/resources/speechToTextJob/client/Client.js +3 -2
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/client/Client.d.ts +1 -1
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingLanguageCode.d.ts +1 -1
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingLanguageCode.js +1 -1
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingModel.d.ts +6 -0
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingModel.js +9 -0
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/index.d.ts +1 -0
- package/dist/cjs/api/resources/speechToTextRealtimeStreaming/types/index.js +1 -0
- package/dist/cjs/api/types/SpeechToTextJobParameters.d.ts +2 -0
- package/dist/cjs/api/types/V2ModelIds.d.ts +4 -0
- package/dist/cjs/api/types/V2ModelIds.js +4 -0
- package/dist/cjs/version.d.ts +1 -1
- package/dist/cjs/version.js +1 -1
- package/dist/esm/BaseClient.mjs +2 -2
- package/dist/esm/api/resources/speechToText/client/Client.mjs +5 -0
- package/dist/esm/api/resources/speechToText/client/requests/SpeechToTextTranscriptionRequest.d.mts +2 -0
- package/dist/esm/api/resources/speechToTextJob/client/Client.d.mts +2 -0
- package/dist/esm/api/resources/speechToTextJob/client/Client.mjs +3 -2
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/client/Client.d.mts +1 -1
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingLanguageCode.d.mts +1 -1
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingLanguageCode.mjs +1 -1
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingModel.d.mts +6 -0
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/SpeechToTextRealtimeStreamingModel.mjs +6 -0
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/index.d.mts +1 -0
- package/dist/esm/api/resources/speechToTextRealtimeStreaming/types/index.mjs +1 -0
- package/dist/esm/api/types/SpeechToTextJobParameters.d.mts +2 -0
- package/dist/esm/api/types/V2ModelIds.d.mts +4 -0
- package/dist/esm/api/types/V2ModelIds.mjs +4 -0
- package/dist/esm/version.d.mts +1 -1
- package/dist/esm/version.mjs +1 -1
- package/package.json +1 -1
package/dist/cjs/BaseClient.js
CHANGED
|
@@ -43,8 +43,8 @@ function normalizeClientOptions(options) {
|
|
|
43
43
|
const headers = (0, headers_js_1.mergeHeaders)({
|
|
44
44
|
"X-Fern-Language": "JavaScript",
|
|
45
45
|
"X-Fern-SDK-Name": "sarvamai",
|
|
46
|
-
"X-Fern-SDK-Version": "1.1.
|
|
47
|
-
"User-Agent": "sarvamai/1.1.
|
|
46
|
+
"X-Fern-SDK-Version": "1.1.10-alpha.1",
|
|
47
|
+
"User-Agent": "sarvamai/1.1.10-alpha.1",
|
|
48
48
|
"X-Fern-Runtime": core.RUNTIME.type,
|
|
49
49
|
"X-Fern-Runtime-Version": core.RUNTIME.version,
|
|
50
50
|
}, options === null || options === void 0 ? void 0 : options.headers);
|
|
@@ -109,6 +109,11 @@ class SpeechToTextClient {
|
|
|
109
109
|
if (request.input_audio_codec != null) {
|
|
110
110
|
_body.append("input_audio_codec", request.input_audio_codec);
|
|
111
111
|
}
|
|
112
|
+
if (request.keyterms != null) {
|
|
113
|
+
for (const _item of request.keyterms) {
|
|
114
|
+
_body.append("keyterms", _item);
|
|
115
|
+
}
|
|
116
|
+
}
|
|
112
117
|
const _maybeEncodedRequest = yield _body.getRequest();
|
|
113
118
|
const _authRequest = yield this._options.authProvider.getAuthRequest();
|
|
114
119
|
const _headers = (0, headers_js_1.mergeHeaders)(_authRequest.headers, (_b = this._options) === null || _b === void 0 ? void 0 : _b.headers, (0, headers_js_1.mergeOnlyDefinedHeaders)(Object.assign({}, _maybeEncodedRequest.headers)), requestOptions === null || requestOptions === void 0 ? void 0 : requestOptions.headers);
|
package/dist/cjs/api/resources/speechToText/client/requests/SpeechToTextTranscriptionRequest.d.ts
CHANGED
|
@@ -79,4 +79,6 @@ export interface SpeechToTextTranscriptionRequest {
|
|
|
79
79
|
with_timestamps?: boolean;
|
|
80
80
|
/** Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate. */
|
|
81
81
|
input_audio_codec?: SarvamAI.InputAudioCodec;
|
|
82
|
+
/** List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Each keyterm can contain up to 64 characters. Put phrases such as `New Delhi` in one list item; do not send comma-separated terms in one string. Keyterms bias recognition — they do not guarantee that a term will appear in the transcript. **Only supported with `model=saaras:v4`.** Sent as a JSON-encoded array in one multipart form field, e.g. `keyterms=["Sarvam","New Delhi","Vistaar"]`. Do not use the older `keyterm` or `hotwords` fields. */
|
|
83
|
+
keyterms?: string[];
|
|
82
84
|
}
|
|
@@ -125,6 +125,7 @@ export declare class SpeechToTextJobClient {
|
|
|
125
125
|
* @param params.languageCode - The language code of the input audio (e.g., "hi-IN", "bn-IN")
|
|
126
126
|
* @param params.numSpeakers - The number of distinct speakers in the audio, if known
|
|
127
127
|
* @param params.mode - Output mode for **saaras:v3** / **saaras:v4** (e.g. transcribe, translate, verbatim)
|
|
128
|
+
* @param params.keyterms - List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Only supported with model="saaras:v4"
|
|
128
129
|
* @param params.callback - Optional callback configuration to receive job completion events
|
|
129
130
|
* @param requestOptions - Request-specific configuration
|
|
130
131
|
* @returns A handle to the newly created Speech-to-Text job
|
|
@@ -136,6 +137,7 @@ export declare class SpeechToTextJobClient {
|
|
|
136
137
|
languageCode?: SarvamAI.SpeechToTextLanguage;
|
|
137
138
|
numSpeakers?: number;
|
|
138
139
|
mode?: SarvamAI.Mode;
|
|
140
|
+
keyterms?: string[];
|
|
139
141
|
callback?: SarvamAI.BulkJobCallback;
|
|
140
142
|
}, requestOptions?: SpeechToTextJobClient.RequestOptions): Promise<SpeechToTextJobInstance>;
|
|
141
143
|
/**
|
|
@@ -502,15 +502,16 @@ class SpeechToTextJobClient {
|
|
|
502
502
|
* @param params.languageCode - The language code of the input audio (e.g., "hi-IN", "bn-IN")
|
|
503
503
|
* @param params.numSpeakers - The number of distinct speakers in the audio, if known
|
|
504
504
|
* @param params.mode - Output mode for **saaras:v3** / **saaras:v4** (e.g. transcribe, translate, verbatim)
|
|
505
|
+
* @param params.keyterms - List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Only supported with model="saaras:v4"
|
|
505
506
|
* @param params.callback - Optional callback configuration to receive job completion events
|
|
506
507
|
* @param requestOptions - Request-specific configuration
|
|
507
508
|
* @returns A handle to the newly created Speech-to-Text job
|
|
508
509
|
*/
|
|
509
510
|
createJob() {
|
|
510
511
|
return __awaiter(this, arguments, void 0, function* (params = {}, requestOptions) {
|
|
511
|
-
const { model = "saaras:v3", withDiarization = false, withTimestamps = false, languageCode, numSpeakers, mode, callback, } = params;
|
|
512
|
+
const { model = "saaras:v3", withDiarization = false, withTimestamps = false, languageCode, numSpeakers, mode, keyterms, callback, } = params;
|
|
512
513
|
const response = yield this.initialise({
|
|
513
|
-
job_parameters: Object.assign({ language_code: languageCode, model: model, num_speakers: numSpeakers, with_diarization: withDiarization, with_timestamps: withTimestamps }, (mode !== undefined ? { mode } : {})),
|
|
514
|
+
job_parameters: Object.assign(Object.assign({ language_code: languageCode, model: model, num_speakers: numSpeakers, with_diarization: withDiarization, with_timestamps: withTimestamps }, (mode !== undefined ? { mode } : {})), (keyterms !== undefined ? { keyterms } : {})),
|
|
514
515
|
callback: callback,
|
|
515
516
|
}, requestOptions);
|
|
516
517
|
return new SpeechToTextJobInstance_js_1.SpeechToTextJobInstance(response.job_id, this);
|
|
@@ -6,7 +6,7 @@ export declare namespace SpeechToTextRealtimeStreamingClient {
|
|
|
6
6
|
type Options = BaseClientOptions;
|
|
7
7
|
interface ConnectArgs {
|
|
8
8
|
language_code: SarvamAI.SpeechToTextRealtimeStreamingLanguageCode;
|
|
9
|
-
model?:
|
|
9
|
+
model?: SarvamAI.SpeechToTextRealtimeStreamingModel;
|
|
10
10
|
stream_type?: SarvamAI.SpeechToTextRealtimeStreamingStreamType;
|
|
11
11
|
mode?: SarvamAI.SpeechToTextRealtimeStreamingMode;
|
|
12
12
|
prompt?: string;
|
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
/**
|
|
2
2
|
* BCP-47 language code of the input audio. **Required.**
|
|
3
3
|
*
|
|
4
|
-
* **Supported languages (saaras:v3-realtime):**
|
|
4
|
+
* **Supported languages (saaras:v3-realtime and saaras:v4-realtime):**
|
|
5
5
|
* - `auto`: Adaptive automatic language detection
|
|
6
6
|
* - `en-IN`: English
|
|
7
7
|
* - `hi-IN`: Hindi
|
|
@@ -5,7 +5,7 @@ exports.SpeechToTextRealtimeStreamingLanguageCode = void 0;
|
|
|
5
5
|
/**
|
|
6
6
|
* BCP-47 language code of the input audio. **Required.**
|
|
7
7
|
*
|
|
8
|
-
* **Supported languages (saaras:v3-realtime):**
|
|
8
|
+
* **Supported languages (saaras:v3-realtime and saaras:v4-realtime):**
|
|
9
9
|
* - `auto`: Adaptive automatic language detection
|
|
10
10
|
* - `en-IN`: English
|
|
11
11
|
* - `hi-IN`: Hindi
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
/** Speech-to-text model for this endpoint. `saaras:v3-realtime` (default) or `saaras:v4-realtime`. */
|
|
2
|
+
export declare const SpeechToTextRealtimeStreamingModel: {
|
|
3
|
+
readonly SaarasV3Realtime: "saaras:v3-realtime";
|
|
4
|
+
readonly SaarasV4Realtime: "saaras:v4-realtime";
|
|
5
|
+
};
|
|
6
|
+
export type SpeechToTextRealtimeStreamingModel = (typeof SpeechToTextRealtimeStreamingModel)[keyof typeof SpeechToTextRealtimeStreamingModel];
|
|
@@ -0,0 +1,9 @@
|
|
|
1
|
+
"use strict";
|
|
2
|
+
// This file was auto-generated by Fern from our API Definition.
|
|
3
|
+
Object.defineProperty(exports, "__esModule", { value: true });
|
|
4
|
+
exports.SpeechToTextRealtimeStreamingModel = void 0;
|
|
5
|
+
/** Speech-to-text model for this endpoint. `saaras:v3-realtime` (default) or `saaras:v4-realtime`. */
|
|
6
|
+
exports.SpeechToTextRealtimeStreamingModel = {
|
|
7
|
+
SaarasV3Realtime: "saaras:v3-realtime",
|
|
8
|
+
SaarasV4Realtime: "saaras:v4-realtime",
|
|
9
|
+
};
|
|
@@ -2,5 +2,6 @@ export * from "./SpeechToTextRealtimeStreamingEncoding.js";
|
|
|
2
2
|
export * from "./SpeechToTextRealtimeStreamingEndpointing.js";
|
|
3
3
|
export * from "./SpeechToTextRealtimeStreamingLanguageCode.js";
|
|
4
4
|
export * from "./SpeechToTextRealtimeStreamingMode.js";
|
|
5
|
+
export * from "./SpeechToTextRealtimeStreamingModel.js";
|
|
5
6
|
export * from "./SpeechToTextRealtimeStreamingReturnTimestamps.js";
|
|
6
7
|
export * from "./SpeechToTextRealtimeStreamingStreamType.js";
|
|
@@ -18,5 +18,6 @@ __exportStar(require("./SpeechToTextRealtimeStreamingEncoding.js"), exports);
|
|
|
18
18
|
__exportStar(require("./SpeechToTextRealtimeStreamingEndpointing.js"), exports);
|
|
19
19
|
__exportStar(require("./SpeechToTextRealtimeStreamingLanguageCode.js"), exports);
|
|
20
20
|
__exportStar(require("./SpeechToTextRealtimeStreamingMode.js"), exports);
|
|
21
|
+
__exportStar(require("./SpeechToTextRealtimeStreamingModel.js"), exports);
|
|
21
22
|
__exportStar(require("./SpeechToTextRealtimeStreamingReturnTimestamps.js"), exports);
|
|
22
23
|
__exportStar(require("./SpeechToTextRealtimeStreamingStreamType.js"), exports);
|
|
@@ -69,4 +69,6 @@ export interface SpeechToTextJobParameters {
|
|
|
69
69
|
num_speakers?: number | undefined;
|
|
70
70
|
/** Audio codec/format of uploaded files. The API automatically detects most formats; for PCM files (pcm_s16le, pcm_l16, pcm_raw), you must specify this parameter. PCM files are supported only at 16kHz sample rate. */
|
|
71
71
|
input_audio_codec?: SarvamAI.InputAudioCodec | undefined;
|
|
72
|
+
/** List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Each keyterm can contain up to 64 characters. Put phrases such as `New Delhi` in one list item; do not send comma-separated terms in one string. Keyterms bias recognition — they do not guarantee that a term will appear in the transcript. **Only supported with `model=saaras:v4`.** Applied while transcribing every audio chunk in the job. Do not use the older `keyterm` or `hotwords` fields. */
|
|
73
|
+
keyterms?: string[] | undefined;
|
|
72
74
|
}
|
|
@@ -3,6 +3,8 @@
|
|
|
3
3
|
*
|
|
4
4
|
* - **`sarvam-105b`** — Sarvam flagship chat model.
|
|
5
5
|
* - **`glm5.2`** — open-source; 512K context, tool calling, visible reasoning.
|
|
6
|
+
* - **`glm5.3`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
7
|
+
* - **`glm5.3-flash`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
6
8
|
* - **`gemma4`** — open-source; image input, tool calling.
|
|
7
9
|
* - **`deepseekv4-flash`** — open-source; 1M context, tool calling, reasoning.
|
|
8
10
|
*
|
|
@@ -11,6 +13,8 @@
|
|
|
11
13
|
export declare const V2ModelIds: {
|
|
12
14
|
readonly Sarvam105B: "sarvam-105b";
|
|
13
15
|
readonly Glm52: "glm5.2";
|
|
16
|
+
readonly Glm53: "glm5.3";
|
|
17
|
+
readonly Glm53Flash: "glm5.3-flash";
|
|
14
18
|
readonly Gemma4: "gemma4";
|
|
15
19
|
readonly Deepseekv4Flash: "deepseekv4-flash";
|
|
16
20
|
};
|
|
@@ -7,6 +7,8 @@ exports.V2ModelIds = void 0;
|
|
|
7
7
|
*
|
|
8
8
|
* - **`sarvam-105b`** — Sarvam flagship chat model.
|
|
9
9
|
* - **`glm5.2`** — open-source; 512K context, tool calling, visible reasoning.
|
|
10
|
+
* - **`glm5.3`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
11
|
+
* - **`glm5.3-flash`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
10
12
|
* - **`gemma4`** — open-source; image input, tool calling.
|
|
11
13
|
* - **`deepseekv4-flash`** — open-source; 1M context, tool calling, reasoning.
|
|
12
14
|
*
|
|
@@ -15,6 +17,8 @@ exports.V2ModelIds = void 0;
|
|
|
15
17
|
exports.V2ModelIds = {
|
|
16
18
|
Sarvam105B: "sarvam-105b",
|
|
17
19
|
Glm52: "glm5.2",
|
|
20
|
+
Glm53: "glm5.3",
|
|
21
|
+
Glm53Flash: "glm5.3-flash",
|
|
18
22
|
Gemma4: "gemma4",
|
|
19
23
|
Deepseekv4Flash: "deepseekv4-flash",
|
|
20
24
|
};
|
package/dist/cjs/version.d.ts
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
export declare const SDK_VERSION = "1.1.
|
|
1
|
+
export declare const SDK_VERSION = "1.1.10-alpha.1";
|
package/dist/cjs/version.js
CHANGED
package/dist/esm/BaseClient.mjs
CHANGED
|
@@ -6,8 +6,8 @@ export function normalizeClientOptions(options) {
|
|
|
6
6
|
const headers = mergeHeaders({
|
|
7
7
|
"X-Fern-Language": "JavaScript",
|
|
8
8
|
"X-Fern-SDK-Name": "sarvamai",
|
|
9
|
-
"X-Fern-SDK-Version": "1.1.
|
|
10
|
-
"User-Agent": "sarvamai/1.1.
|
|
9
|
+
"X-Fern-SDK-Version": "1.1.10-alpha.1",
|
|
10
|
+
"User-Agent": "sarvamai/1.1.10-alpha.1",
|
|
11
11
|
"X-Fern-Runtime": core.RUNTIME.type,
|
|
12
12
|
"X-Fern-Runtime-Version": core.RUNTIME.version,
|
|
13
13
|
}, options === null || options === void 0 ? void 0 : options.headers);
|
|
@@ -73,6 +73,11 @@ export class SpeechToTextClient {
|
|
|
73
73
|
if (request.input_audio_codec != null) {
|
|
74
74
|
_body.append("input_audio_codec", request.input_audio_codec);
|
|
75
75
|
}
|
|
76
|
+
if (request.keyterms != null) {
|
|
77
|
+
for (const _item of request.keyterms) {
|
|
78
|
+
_body.append("keyterms", _item);
|
|
79
|
+
}
|
|
80
|
+
}
|
|
76
81
|
const _maybeEncodedRequest = yield _body.getRequest();
|
|
77
82
|
const _authRequest = yield this._options.authProvider.getAuthRequest();
|
|
78
83
|
const _headers = mergeHeaders(_authRequest.headers, (_b = this._options) === null || _b === void 0 ? void 0 : _b.headers, mergeOnlyDefinedHeaders(Object.assign({}, _maybeEncodedRequest.headers)), requestOptions === null || requestOptions === void 0 ? void 0 : requestOptions.headers);
|
package/dist/esm/api/resources/speechToText/client/requests/SpeechToTextTranscriptionRequest.d.mts
CHANGED
|
@@ -79,4 +79,6 @@ export interface SpeechToTextTranscriptionRequest {
|
|
|
79
79
|
with_timestamps?: boolean;
|
|
80
80
|
/** Input Audio codec/format of the input file. PCM files are supported only at 16kHz sample rate. */
|
|
81
81
|
input_audio_codec?: SarvamAI.InputAudioCodec;
|
|
82
|
+
/** List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Each keyterm can contain up to 64 characters. Put phrases such as `New Delhi` in one list item; do not send comma-separated terms in one string. Keyterms bias recognition — they do not guarantee that a term will appear in the transcript. **Only supported with `model=saaras:v4`.** Sent as a JSON-encoded array in one multipart form field, e.g. `keyterms=["Sarvam","New Delhi","Vistaar"]`. Do not use the older `keyterm` or `hotwords` fields. */
|
|
83
|
+
keyterms?: string[];
|
|
82
84
|
}
|
|
@@ -125,6 +125,7 @@ export declare class SpeechToTextJobClient {
|
|
|
125
125
|
* @param params.languageCode - The language code of the input audio (e.g., "hi-IN", "bn-IN")
|
|
126
126
|
* @param params.numSpeakers - The number of distinct speakers in the audio, if known
|
|
127
127
|
* @param params.mode - Output mode for **saaras:v3** / **saaras:v4** (e.g. transcribe, translate, verbatim)
|
|
128
|
+
* @param params.keyterms - List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Only supported with model="saaras:v4"
|
|
128
129
|
* @param params.callback - Optional callback configuration to receive job completion events
|
|
129
130
|
* @param requestOptions - Request-specific configuration
|
|
130
131
|
* @returns A handle to the newly created Speech-to-Text job
|
|
@@ -136,6 +137,7 @@ export declare class SpeechToTextJobClient {
|
|
|
136
137
|
languageCode?: SarvamAI.SpeechToTextLanguage;
|
|
137
138
|
numSpeakers?: number;
|
|
138
139
|
mode?: SarvamAI.Mode;
|
|
140
|
+
keyterms?: string[];
|
|
139
141
|
callback?: SarvamAI.BulkJobCallback;
|
|
140
142
|
}, requestOptions?: SpeechToTextJobClient.RequestOptions): Promise<SpeechToTextJobInstance>;
|
|
141
143
|
/**
|
|
@@ -466,15 +466,16 @@ export class SpeechToTextJobClient {
|
|
|
466
466
|
* @param params.languageCode - The language code of the input audio (e.g., "hi-IN", "bn-IN")
|
|
467
467
|
* @param params.numSpeakers - The number of distinct speakers in the audio, if known
|
|
468
468
|
* @param params.mode - Output mode for **saaras:v3** / **saaras:v4** (e.g. transcribe, translate, verbatim)
|
|
469
|
+
* @param params.keyterms - List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Only supported with model="saaras:v4"
|
|
469
470
|
* @param params.callback - Optional callback configuration to receive job completion events
|
|
470
471
|
* @param requestOptions - Request-specific configuration
|
|
471
472
|
* @returns A handle to the newly created Speech-to-Text job
|
|
472
473
|
*/
|
|
473
474
|
createJob() {
|
|
474
475
|
return __awaiter(this, arguments, void 0, function* (params = {}, requestOptions) {
|
|
475
|
-
const { model = "saaras:v3", withDiarization = false, withTimestamps = false, languageCode, numSpeakers, mode, callback, } = params;
|
|
476
|
+
const { model = "saaras:v3", withDiarization = false, withTimestamps = false, languageCode, numSpeakers, mode, keyterms, callback, } = params;
|
|
476
477
|
const response = yield this.initialise({
|
|
477
|
-
job_parameters: Object.assign({ language_code: languageCode, model: model, num_speakers: numSpeakers, with_diarization: withDiarization, with_timestamps: withTimestamps }, (mode !== undefined ? { mode } : {})),
|
|
478
|
+
job_parameters: Object.assign(Object.assign({ language_code: languageCode, model: model, num_speakers: numSpeakers, with_diarization: withDiarization, with_timestamps: withTimestamps }, (mode !== undefined ? { mode } : {})), (keyterms !== undefined ? { keyterms } : {})),
|
|
478
479
|
callback: callback,
|
|
479
480
|
}, requestOptions);
|
|
480
481
|
return new SpeechToTextJobInstance(response.job_id, this);
|
|
@@ -6,7 +6,7 @@ export declare namespace SpeechToTextRealtimeStreamingClient {
|
|
|
6
6
|
type Options = BaseClientOptions;
|
|
7
7
|
interface ConnectArgs {
|
|
8
8
|
language_code: SarvamAI.SpeechToTextRealtimeStreamingLanguageCode;
|
|
9
|
-
model?:
|
|
9
|
+
model?: SarvamAI.SpeechToTextRealtimeStreamingModel;
|
|
10
10
|
stream_type?: SarvamAI.SpeechToTextRealtimeStreamingStreamType;
|
|
11
11
|
mode?: SarvamAI.SpeechToTextRealtimeStreamingMode;
|
|
12
12
|
prompt?: string;
|
|
@@ -1,7 +1,7 @@
|
|
|
1
1
|
/**
|
|
2
2
|
* BCP-47 language code of the input audio. **Required.**
|
|
3
3
|
*
|
|
4
|
-
* **Supported languages (saaras:v3-realtime):**
|
|
4
|
+
* **Supported languages (saaras:v3-realtime and saaras:v4-realtime):**
|
|
5
5
|
* - `auto`: Adaptive automatic language detection
|
|
6
6
|
* - `en-IN`: English
|
|
7
7
|
* - `hi-IN`: Hindi
|
|
@@ -2,7 +2,7 @@
|
|
|
2
2
|
/**
|
|
3
3
|
* BCP-47 language code of the input audio. **Required.**
|
|
4
4
|
*
|
|
5
|
-
* **Supported languages (saaras:v3-realtime):**
|
|
5
|
+
* **Supported languages (saaras:v3-realtime and saaras:v4-realtime):**
|
|
6
6
|
* - `auto`: Adaptive automatic language detection
|
|
7
7
|
* - `en-IN`: English
|
|
8
8
|
* - `hi-IN`: Hindi
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
/** Speech-to-text model for this endpoint. `saaras:v3-realtime` (default) or `saaras:v4-realtime`. */
|
|
2
|
+
export declare const SpeechToTextRealtimeStreamingModel: {
|
|
3
|
+
readonly SaarasV3Realtime: "saaras:v3-realtime";
|
|
4
|
+
readonly SaarasV4Realtime: "saaras:v4-realtime";
|
|
5
|
+
};
|
|
6
|
+
export type SpeechToTextRealtimeStreamingModel = (typeof SpeechToTextRealtimeStreamingModel)[keyof typeof SpeechToTextRealtimeStreamingModel];
|
|
@@ -0,0 +1,6 @@
|
|
|
1
|
+
// This file was auto-generated by Fern from our API Definition.
|
|
2
|
+
/** Speech-to-text model for this endpoint. `saaras:v3-realtime` (default) or `saaras:v4-realtime`. */
|
|
3
|
+
export const SpeechToTextRealtimeStreamingModel = {
|
|
4
|
+
SaarasV3Realtime: "saaras:v3-realtime",
|
|
5
|
+
SaarasV4Realtime: "saaras:v4-realtime",
|
|
6
|
+
};
|
|
@@ -2,5 +2,6 @@ export * from "./SpeechToTextRealtimeStreamingEncoding.mjs";
|
|
|
2
2
|
export * from "./SpeechToTextRealtimeStreamingEndpointing.mjs";
|
|
3
3
|
export * from "./SpeechToTextRealtimeStreamingLanguageCode.mjs";
|
|
4
4
|
export * from "./SpeechToTextRealtimeStreamingMode.mjs";
|
|
5
|
+
export * from "./SpeechToTextRealtimeStreamingModel.mjs";
|
|
5
6
|
export * from "./SpeechToTextRealtimeStreamingReturnTimestamps.mjs";
|
|
6
7
|
export * from "./SpeechToTextRealtimeStreamingStreamType.mjs";
|
|
@@ -2,5 +2,6 @@ export * from "./SpeechToTextRealtimeStreamingEncoding.mjs";
|
|
|
2
2
|
export * from "./SpeechToTextRealtimeStreamingEndpointing.mjs";
|
|
3
3
|
export * from "./SpeechToTextRealtimeStreamingLanguageCode.mjs";
|
|
4
4
|
export * from "./SpeechToTextRealtimeStreamingMode.mjs";
|
|
5
|
+
export * from "./SpeechToTextRealtimeStreamingModel.mjs";
|
|
5
6
|
export * from "./SpeechToTextRealtimeStreamingReturnTimestamps.mjs";
|
|
6
7
|
export * from "./SpeechToTextRealtimeStreamingStreamType.mjs";
|
|
@@ -69,4 +69,6 @@ export interface SpeechToTextJobParameters {
|
|
|
69
69
|
num_speakers?: number | undefined;
|
|
70
70
|
/** Audio codec/format of uploaded files. The API automatically detects most formats; for PCM files (pcm_s16le, pcm_l16, pcm_raw), you must specify this parameter. PCM files are supported only at 16kHz sample rate. */
|
|
71
71
|
input_audio_codec?: SarvamAI.InputAudioCodec | undefined;
|
|
72
|
+
/** List of up to 50 domain-specific terms (names, places, brands, technical terms) to bias recognition toward. Each keyterm can contain up to 64 characters. Put phrases such as `New Delhi` in one list item; do not send comma-separated terms in one string. Keyterms bias recognition — they do not guarantee that a term will appear in the transcript. **Only supported with `model=saaras:v4`.** Applied while transcribing every audio chunk in the job. Do not use the older `keyterm` or `hotwords` fields. */
|
|
73
|
+
keyterms?: string[] | undefined;
|
|
72
74
|
}
|
|
@@ -3,6 +3,8 @@
|
|
|
3
3
|
*
|
|
4
4
|
* - **`sarvam-105b`** — Sarvam flagship chat model.
|
|
5
5
|
* - **`glm5.2`** — open-source; 512K context, tool calling, visible reasoning.
|
|
6
|
+
* - **`glm5.3`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
7
|
+
* - **`glm5.3-flash`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
6
8
|
* - **`gemma4`** — open-source; image input, tool calling.
|
|
7
9
|
* - **`deepseekv4-flash`** — open-source; 1M context, tool calling, reasoning.
|
|
8
10
|
*
|
|
@@ -11,6 +13,8 @@
|
|
|
11
13
|
export declare const V2ModelIds: {
|
|
12
14
|
readonly Sarvam105B: "sarvam-105b";
|
|
13
15
|
readonly Glm52: "glm5.2";
|
|
16
|
+
readonly Glm53: "glm5.3";
|
|
17
|
+
readonly Glm53Flash: "glm5.3-flash";
|
|
14
18
|
readonly Gemma4: "gemma4";
|
|
15
19
|
readonly Deepseekv4Flash: "deepseekv4-flash";
|
|
16
20
|
};
|
|
@@ -4,6 +4,8 @@
|
|
|
4
4
|
*
|
|
5
5
|
* - **`sarvam-105b`** — Sarvam flagship chat model.
|
|
6
6
|
* - **`glm5.2`** — open-source; 512K context, tool calling, visible reasoning.
|
|
7
|
+
* - **`glm5.3`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
8
|
+
* - **`glm5.3-flash`** — open-source; served by `GET /v2/models`, capability details not yet published.
|
|
7
9
|
* - **`gemma4`** — open-source; image input, tool calling.
|
|
8
10
|
* - **`deepseekv4-flash`** — open-source; 1M context, tool calling, reasoning.
|
|
9
11
|
*
|
|
@@ -12,6 +14,8 @@
|
|
|
12
14
|
export const V2ModelIds = {
|
|
13
15
|
Sarvam105B: "sarvam-105b",
|
|
14
16
|
Glm52: "glm5.2",
|
|
17
|
+
Glm53: "glm5.3",
|
|
18
|
+
Glm53Flash: "glm5.3-flash",
|
|
15
19
|
Gemma4: "gemma4",
|
|
16
20
|
Deepseekv4Flash: "deepseekv4-flash",
|
|
17
21
|
};
|
package/dist/esm/version.d.mts
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
export declare const SDK_VERSION = "1.1.
|
|
1
|
+
export declare const SDK_VERSION = "1.1.10-alpha.1";
|
package/dist/esm/version.mjs
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
export const SDK_VERSION = "1.1.
|
|
1
|
+
export const SDK_VERSION = "1.1.10-alpha.1";
|