@jambonz/schema 0.4.3 → 0.4.4
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.
|
@@ -0,0 +1,42 @@
|
|
|
1
|
+
{
|
|
2
|
+
"$schema": "https://json-schema.org/draft/2020-12/schema",
|
|
3
|
+
"$id": "https://jambonz.org/schema/components/recognizer-modulateOptions",
|
|
4
|
+
"title": "Modulate Recognizer Options",
|
|
5
|
+
"description": "Modulate-specific STT options. Only applies when recognizer vendor is 'modulate'. Modulate does not return confidence scores; transcripts are reported with confidence 1.0. The model is selected with the top-level recognizer 'model' property: 'velma-2-stt-streaming' (multilingual, the default) or 'velma-2-stt-streaming-english-v2' (low-latency English).",
|
|
6
|
+
"type": "object",
|
|
7
|
+
"properties": {
|
|
8
|
+
"apiKey": {
|
|
9
|
+
"type": "string",
|
|
10
|
+
"description": "Modulate API key. Overrides the key configured in jambonz."
|
|
11
|
+
},
|
|
12
|
+
"endpointing": {
|
|
13
|
+
"type": "boolean",
|
|
14
|
+
"description": "Only applies to model 'velma-2-stt-streaming-english-v2': emit a final transcript for each pause-delimited speech segment. Defaults to true in jambonz; set false to receive a single final transcript when the stream ends."
|
|
15
|
+
},
|
|
16
|
+
"emotionSignal": {
|
|
17
|
+
"type": "boolean",
|
|
18
|
+
"description": "Only applies to model 'velma-2-stt-streaming': detect the speaker's emotional tone for each utterance."
|
|
19
|
+
},
|
|
20
|
+
"accentSignal": {
|
|
21
|
+
"type": "boolean",
|
|
22
|
+
"description": "Only applies to model 'velma-2-stt-streaming': identify the speaker's accent for each utterance."
|
|
23
|
+
},
|
|
24
|
+
"deepfakeSignal": {
|
|
25
|
+
"type": "boolean",
|
|
26
|
+
"description": "Only applies to model 'velma-2-stt-streaming': score each utterance for synthetic (deepfake) voice likelihood."
|
|
27
|
+
},
|
|
28
|
+
"piiPhiTagging": {
|
|
29
|
+
"type": "boolean",
|
|
30
|
+
"description": "Only applies to model 'velma-2-stt-streaming': wrap detected PII/PHI in tags within the transcript text."
|
|
31
|
+
},
|
|
32
|
+
"modulateSttUri": {
|
|
33
|
+
"type": "string",
|
|
34
|
+
"description": "Custom host to connect to instead of platform.modulate.ai."
|
|
35
|
+
},
|
|
36
|
+
"modulateSttUseTls": {
|
|
37
|
+
"type": "boolean",
|
|
38
|
+
"description": "Use TLS when connecting to a custom host. Default true."
|
|
39
|
+
}
|
|
40
|
+
},
|
|
41
|
+
"additionalProperties": false
|
|
42
|
+
}
|
|
@@ -8,7 +8,7 @@
|
|
|
8
8
|
"vendor": {
|
|
9
9
|
"type": "string",
|
|
10
10
|
"description": "The STT vendor to use. Must match a vendor configured in the jambonz platform.",
|
|
11
|
-
"examples": ["google", "aws", "microsoft", "deepgram", "ibm", "nvidia", "soniox", "xai", "assemblyai", "speechmatics", "openai", "houndify", "gladia", "elevenlabs", "custom"]
|
|
11
|
+
"examples": ["google", "aws", "microsoft", "deepgram", "ibm", "nvidia", "soniox", "xai", "modulate", "assemblyai", "speechmatics", "openai", "houndify", "gladia", "elevenlabs", "custom"]
|
|
12
12
|
},
|
|
13
13
|
"label": {
|
|
14
14
|
"type": "string",
|
|
@@ -199,6 +199,7 @@
|
|
|
199
199
|
"nvidiaOptions": { "$ref": "recognizer-nvidiaOptions" },
|
|
200
200
|
"sonioxOptions": { "$ref": "recognizer-sonioxOptions" },
|
|
201
201
|
"xaiOptions": { "$ref": "recognizer-xaiOptions" },
|
|
202
|
+
"modulateOptions": { "$ref": "recognizer-modulateOptions" },
|
|
202
203
|
"assemblyAiOptions": { "$ref": "recognizer-assemblyAiOptions" },
|
|
203
204
|
"speechmaticsOptions": { "$ref": "recognizer-speechmaticsOptions" },
|
|
204
205
|
"openaiOptions": { "$ref": "recognizer-openaiOptions" },
|