hume 0.13.0 → 0.13.1
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/.mock/definition/empathic-voice/__package__.yml +29 -31
- package/.mock/definition/empathic-voice/chatWebhooks.yml +8 -12
- package/.mock/definition/empathic-voice/prompts.yml +2 -2
- package/.mock/definition/empathic-voice/tools.yml +2 -2
- package/api/resources/empathicVoice/resources/chat/client/Client.d.ts +2 -0
- package/api/resources/empathicVoice/resources/prompts/client/requests/PostedPrompt.d.ts +1 -1
- package/api/resources/empathicVoice/resources/prompts/client/requests/PostedPromptVersion.d.ts +1 -1
- package/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedTool.d.ts +1 -1
- package/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedToolVersion.d.ts +1 -1
- package/api/resources/empathicVoice/types/AssistantEnd.d.ts +1 -1
- package/api/resources/empathicVoice/types/AssistantInput.d.ts +1 -1
- package/api/resources/empathicVoice/types/AssistantMessage.d.ts +1 -1
- package/api/resources/empathicVoice/types/AudioInput.d.ts +1 -1
- package/api/resources/empathicVoice/types/PauseAssistantMessage.d.ts +1 -1
- package/api/resources/empathicVoice/types/PostedLanguageModel.d.ts +1 -1
- package/api/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.d.ts +2 -2
- package/api/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.d.ts +2 -2
- package/api/resources/empathicVoice/types/ReturnLanguageModel.d.ts +1 -1
- package/api/resources/empathicVoice/types/ReturnPrompt.d.ts +2 -4
- package/api/resources/empathicVoice/types/ReturnUserDefinedTool.d.ts +1 -1
- package/api/resources/empathicVoice/types/SessionSettings.d.ts +2 -2
- package/api/resources/empathicVoice/types/Tool.d.ts +1 -1
- package/api/resources/empathicVoice/types/ToolCallMessage.d.ts +1 -1
- package/api/resources/empathicVoice/types/UserInput.d.ts +1 -1
- package/api/resources/empathicVoice/types/UserInterruption.d.ts +1 -1
- package/api/resources/empathicVoice/types/UserMessage.d.ts +1 -1
- package/dist/api/resources/empathicVoice/resources/chat/client/Client.d.ts +2 -0
- package/dist/api/resources/empathicVoice/resources/prompts/client/requests/PostedPrompt.d.ts +1 -1
- package/dist/api/resources/empathicVoice/resources/prompts/client/requests/PostedPromptVersion.d.ts +1 -1
- package/dist/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedTool.d.ts +1 -1
- package/dist/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedToolVersion.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/AssistantEnd.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/AssistantInput.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/AssistantMessage.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/AudioInput.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/PauseAssistantMessage.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/PostedLanguageModel.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.d.ts +2 -2
- package/dist/api/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.d.ts +2 -2
- package/dist/api/resources/empathicVoice/types/ReturnLanguageModel.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/ReturnPrompt.d.ts +2 -4
- package/dist/api/resources/empathicVoice/types/ReturnUserDefinedTool.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/SessionSettings.d.ts +2 -2
- package/dist/api/resources/empathicVoice/types/Tool.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/ToolCallMessage.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/UserInput.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/UserInterruption.d.ts +1 -1
- package/dist/api/resources/empathicVoice/types/UserMessage.d.ts +1 -1
- package/dist/serialization/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.d.ts +1 -1
- package/dist/serialization/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.js +1 -1
- package/dist/serialization/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.d.ts +1 -1
- package/dist/serialization/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.js +1 -1
- package/dist/version.d.ts +1 -1
- package/dist/version.js +1 -1
- package/package.json +1 -1
- package/serialization/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.d.ts +1 -1
- package/serialization/resources/empathicVoice/types/PostedTimeoutSpecsInactivity.js +1 -1
- package/serialization/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.d.ts +1 -1
- package/serialization/resources/empathicVoice/types/PostedTimeoutSpecsMaxDuration.js +1 -1
- package/version.d.ts +1 -1
- package/version.js +1 -1
|
@@ -29,7 +29,7 @@ types:
|
|
|
29
29
|
|
|
30
30
|
EVI uses this text to generate spoken audio using our proprietary
|
|
31
31
|
expressive text-to-speech model. Our model adds appropriate emotional
|
|
32
|
-
inflections and tones to the text based on the user
|
|
32
|
+
inflections and tones to the text based on the user's expressions and
|
|
33
33
|
the context of the conversation. The synthesized audio is streamed
|
|
34
34
|
back to the user as an [Assistant
|
|
35
35
|
Message](/reference/empathic-voice-interface-evi/chat/chat#receive.AssistantMessage.type).
|
|
@@ -75,7 +75,7 @@ types:
|
|
|
75
75
|
Base64 encoded audio input to insert into the conversation.
|
|
76
76
|
|
|
77
77
|
|
|
78
|
-
The content of an Audio Input message is treated as the user
|
|
78
|
+
The content of an Audio Input message is treated as the user's speech
|
|
79
79
|
to EVI and must be streamed continuously. Pre-recorded audio files are
|
|
80
80
|
not supported.
|
|
81
81
|
|
|
@@ -166,7 +166,7 @@ types:
|
|
|
166
166
|
Once this message is sent, EVI will not respond until a [Resume
|
|
167
167
|
Assistant
|
|
168
168
|
message](/reference/empathic-voice-interface-evi/chat/chat#send.ResumeAssistantMessage.type)
|
|
169
|
-
is sent. When paused, EVI won
|
|
169
|
+
is sent. When paused, EVI won't respond, but transcriptions of your
|
|
170
170
|
audio inputs will still be recorded.
|
|
171
171
|
custom_session_id:
|
|
172
172
|
type: optional<string>
|
|
@@ -250,7 +250,7 @@ types:
|
|
|
250
250
|
system_prompt:
|
|
251
251
|
type: optional<string>
|
|
252
252
|
docs: >-
|
|
253
|
-
Instructions used to shape EVI
|
|
253
|
+
Instructions used to shape EVI's behavior, responses, and style for
|
|
254
254
|
the session.
|
|
255
255
|
|
|
256
256
|
|
|
@@ -301,7 +301,7 @@ types:
|
|
|
301
301
|
Third party API key for the supplemental language model.
|
|
302
302
|
|
|
303
303
|
|
|
304
|
-
When provided, EVI will use this key instead of Hume
|
|
304
|
+
When provided, EVI will use this key instead of Hume's API key for the
|
|
305
305
|
supplemental LLM. This allows you to bypass rate limits and utilize
|
|
306
306
|
your own API key as needed.
|
|
307
307
|
tools:
|
|
@@ -367,7 +367,7 @@ types:
|
|
|
367
367
|
Parameters of the tool. Is a stringified JSON schema.
|
|
368
368
|
|
|
369
369
|
|
|
370
|
-
These parameters define the inputs needed for the tool
|
|
370
|
+
These parameters define the inputs needed for the tool's execution,
|
|
371
371
|
including the expected data type and description for each input field.
|
|
372
372
|
Structured as a JSON schema, this format ensures the tool receives
|
|
373
373
|
data in the expected format.
|
|
@@ -521,7 +521,7 @@ types:
|
|
|
521
521
|
type: string
|
|
522
522
|
docs: >-
|
|
523
523
|
User text to insert into the conversation. Text sent through a User
|
|
524
|
-
Input message is treated as the user
|
|
524
|
+
Input message is treated as the user's speech to EVI. EVI processes
|
|
525
525
|
this input and provides a corresponding response.
|
|
526
526
|
|
|
527
527
|
|
|
@@ -540,7 +540,7 @@ types:
|
|
|
540
540
|
message, this must be `assistant_end`.
|
|
541
541
|
|
|
542
542
|
|
|
543
|
-
This message indicates the conclusion of the assistant
|
|
543
|
+
This message indicates the conclusion of the assistant's response,
|
|
544
544
|
signaling that the assistant has finished speaking for the current
|
|
545
545
|
conversational turn.
|
|
546
546
|
custom_session_id:
|
|
@@ -560,8 +560,8 @@ types:
|
|
|
560
560
|
this must be `assistant_message`.
|
|
561
561
|
|
|
562
562
|
|
|
563
|
-
This message contains both a transcript of the assistant
|
|
564
|
-
and the expression measurement predictions of the assistant
|
|
563
|
+
This message contains both a transcript of the assistant's response
|
|
564
|
+
and the expression measurement predictions of the assistant's audio
|
|
565
565
|
output.
|
|
566
566
|
custom_session_id:
|
|
567
567
|
type: optional<string>
|
|
@@ -860,7 +860,7 @@ types:
|
|
|
860
860
|
Parameters of the tool.
|
|
861
861
|
|
|
862
862
|
|
|
863
|
-
These parameters define the inputs needed for the tool
|
|
863
|
+
These parameters define the inputs needed for the tool's execution,
|
|
864
864
|
including the expected data type and description for each input field.
|
|
865
865
|
Structured as a stringified JSON schema, this format ensures the tool
|
|
866
866
|
receives data in the expected format.
|
|
@@ -912,7 +912,7 @@ types:
|
|
|
912
912
|
message, this must be `user_interruption`.
|
|
913
913
|
|
|
914
914
|
|
|
915
|
-
This message indicates the user has interrupted the assistant
|
|
915
|
+
This message indicates the user has interrupted the assistant's
|
|
916
916
|
response. EVI detects the interruption in real-time and sends this
|
|
917
917
|
message to signal the interruption event. This message allows the
|
|
918
918
|
system to stop the current audio playback, clear the audio queue, and
|
|
@@ -937,7 +937,7 @@ types:
|
|
|
937
937
|
must be `user_message`.
|
|
938
938
|
|
|
939
939
|
|
|
940
|
-
This message contains both a transcript of the user
|
|
940
|
+
This message contains both a transcript of the user's input and the
|
|
941
941
|
expression measurement predictions if the input was sent as an [Audio
|
|
942
942
|
Input
|
|
943
943
|
message](/reference/empathic-voice-interface-evi/chat/chat#send.AudioInput.type).
|
|
@@ -1363,7 +1363,7 @@ types:
|
|
|
1363
1363
|
Tool.
|
|
1364
1364
|
|
|
1365
1365
|
|
|
1366
|
-
These parameters define the inputs needed for the Tool
|
|
1366
|
+
These parameters define the inputs needed for the Tool's execution,
|
|
1367
1367
|
including the expected data type and description for each input field.
|
|
1368
1368
|
Structured as a stringified JSON schema, this format ensures the tool
|
|
1369
1369
|
receives data in the expected format.
|
|
@@ -1410,17 +1410,15 @@ types:
|
|
|
1410
1410
|
text:
|
|
1411
1411
|
type: string
|
|
1412
1412
|
docs: >-
|
|
1413
|
-
Instructions used to shape EVI
|
|
1413
|
+
Instructions used to shape EVI's behavior, responses, and style.
|
|
1414
1414
|
|
|
1415
1415
|
|
|
1416
1416
|
You can use the Prompt to define a specific goal or role for EVI,
|
|
1417
1417
|
specifying how it should act or what it should focus on during the
|
|
1418
1418
|
conversation. For example, EVI can be instructed to act as a customer
|
|
1419
1419
|
support representative, a fitness coach, or a travel advisor, each
|
|
1420
|
-
with its own set of behaviors and response styles.
|
|
1421
|
-
|
|
1422
|
-
|
|
1423
|
-
For help writing a system prompt, see our [Prompting
|
|
1420
|
+
with its own set of behaviors and response styles. For help writing a
|
|
1421
|
+
system prompt, see our [Prompting
|
|
1424
1422
|
Guide](/docs/speech-to-speech-evi/guides/prompting).
|
|
1425
1423
|
version:
|
|
1426
1424
|
type: integer
|
|
@@ -2142,7 +2140,7 @@ types:
|
|
|
2142
2140
|
The model temperature, with values between 0 to 1 (inclusive).
|
|
2143
2141
|
|
|
2144
2142
|
|
|
2145
|
-
Controls the randomness of the LLM
|
|
2143
|
+
Controls the randomness of the LLM's output, with values closer to 0
|
|
2146
2144
|
yielding focused, deterministic responses and values closer to 1
|
|
2147
2145
|
producing more creative, diverse responses.
|
|
2148
2146
|
source:
|
|
@@ -2293,6 +2291,11 @@ types:
|
|
|
2293
2291
|
Accepts a minimum value of 30 seconds and a maximum value of 1,800
|
|
2294
2292
|
seconds.
|
|
2295
2293
|
properties:
|
|
2294
|
+
duration_secs:
|
|
2295
|
+
type: optional<integer>
|
|
2296
|
+
docs: >-
|
|
2297
|
+
Duration in seconds for the timeout (e.g. 600 seconds represents 10
|
|
2298
|
+
minutes).
|
|
2296
2299
|
enabled:
|
|
2297
2300
|
type: boolean
|
|
2298
2301
|
docs: >-
|
|
@@ -2303,11 +2306,6 @@ types:
|
|
|
2303
2306
|
user inactivity being reached. However, the conversation will
|
|
2304
2307
|
eventually disconnect after 1,800 seconds (30 minutes), which is the
|
|
2305
2308
|
maximum WebSocket duration limit for EVI.
|
|
2306
|
-
duration_secs:
|
|
2307
|
-
type: optional<integer>
|
|
2308
|
-
docs: >-
|
|
2309
|
-
Duration in seconds for the timeout (e.g. 600 seconds represents 10
|
|
2310
|
-
minutes).
|
|
2311
2309
|
source:
|
|
2312
2310
|
openapi: evi-openapi.json
|
|
2313
2311
|
inline: true
|
|
@@ -2321,6 +2319,11 @@ types:
|
|
|
2321
2319
|
Accepts a minimum value of 30 seconds and a maximum value of 1,800
|
|
2322
2320
|
seconds.
|
|
2323
2321
|
properties:
|
|
2322
|
+
duration_secs:
|
|
2323
|
+
type: optional<integer>
|
|
2324
|
+
docs: >-
|
|
2325
|
+
Duration in seconds for the timeout (e.g. 600 seconds represents 10
|
|
2326
|
+
minutes).
|
|
2324
2327
|
enabled:
|
|
2325
2328
|
type: boolean
|
|
2326
2329
|
docs: >-
|
|
@@ -2331,11 +2334,6 @@ types:
|
|
|
2331
2334
|
duration being reached. However, the conversation will eventually
|
|
2332
2335
|
disconnect after 1,800 seconds (30 minutes), which is the maximum
|
|
2333
2336
|
WebSocket duration limit for EVI.
|
|
2334
|
-
duration_secs:
|
|
2335
|
-
type: optional<integer>
|
|
2336
|
-
docs: >-
|
|
2337
|
-
Duration in seconds for the timeout (e.g. 600 seconds represents 10
|
|
2338
|
-
minutes).
|
|
2339
2337
|
source:
|
|
2340
2338
|
openapi: evi-openapi.json
|
|
2341
2339
|
inline: true
|
|
@@ -2414,7 +2412,7 @@ types:
|
|
|
2414
2412
|
The model temperature, with values between 0 to 1 (inclusive).
|
|
2415
2413
|
|
|
2416
2414
|
|
|
2417
|
-
Controls the randomness of the LLM
|
|
2415
|
+
Controls the randomness of the LLM's output, with values closer to 0
|
|
2418
2416
|
yielding focused, deterministic responses and values closer to 1
|
|
2419
2417
|
producing more creative, diverse responses.
|
|
2420
2418
|
source:
|
|
@@ -9,13 +9,11 @@ webhooks:
|
|
|
9
9
|
payload: root.WebhookEventChatEnded
|
|
10
10
|
examples:
|
|
11
11
|
- payload:
|
|
12
|
-
chat_group_id:
|
|
13
|
-
chat_id:
|
|
14
|
-
|
|
15
|
-
|
|
16
|
-
|
|
17
|
-
duration_seconds: 180
|
|
18
|
-
end_reason: USER_ENDED
|
|
12
|
+
chat_group_id: chat_group_id
|
|
13
|
+
chat_id: chat_id
|
|
14
|
+
end_time: 1
|
|
15
|
+
duration_seconds: 1
|
|
16
|
+
end_reason: ACTIVE
|
|
19
17
|
docs: Sent when an EVI chat ends.
|
|
20
18
|
chatStarted:
|
|
21
19
|
audiences: []
|
|
@@ -25,10 +23,8 @@ webhooks:
|
|
|
25
23
|
payload: root.WebhookEventChatStarted
|
|
26
24
|
examples:
|
|
27
25
|
- payload:
|
|
28
|
-
chat_group_id:
|
|
29
|
-
chat_id:
|
|
30
|
-
|
|
31
|
-
event_name: chat_started
|
|
32
|
-
start_time: 1716244940648
|
|
26
|
+
chat_group_id: chat_group_id
|
|
27
|
+
chat_id: chat_id
|
|
28
|
+
start_time: 1
|
|
33
29
|
chat_start_type: new_chat_group
|
|
34
30
|
docs: Sent when an EVI chat is started.
|
|
@@ -126,7 +126,7 @@ service:
|
|
|
126
126
|
text:
|
|
127
127
|
type: string
|
|
128
128
|
docs: >-
|
|
129
|
-
Instructions used to shape EVI
|
|
129
|
+
Instructions used to shape EVI's behavior, responses, and style.
|
|
130
130
|
|
|
131
131
|
|
|
132
132
|
You can use the Prompt to define a specific goal or role for
|
|
@@ -278,7 +278,7 @@ service:
|
|
|
278
278
|
text:
|
|
279
279
|
type: string
|
|
280
280
|
docs: >-
|
|
281
|
-
Instructions used to shape EVI
|
|
281
|
+
Instructions used to shape EVI's behavior, responses, and style
|
|
282
282
|
for this version of the Prompt.
|
|
283
283
|
|
|
284
284
|
|
|
@@ -145,7 +145,7 @@ service:
|
|
|
145
145
|
the Tool.
|
|
146
146
|
|
|
147
147
|
|
|
148
|
-
These parameters define the inputs needed for the Tool
|
|
148
|
+
These parameters define the inputs needed for the Tool's
|
|
149
149
|
execution, including the expected data type and description for
|
|
150
150
|
each input field. Structured as a stringified JSON schema, this
|
|
151
151
|
format ensures the Tool receives data in the expected format.
|
|
@@ -324,7 +324,7 @@ service:
|
|
|
324
324
|
the Tool.
|
|
325
325
|
|
|
326
326
|
|
|
327
|
-
These parameters define the inputs needed for the Tool
|
|
327
|
+
These parameters define the inputs needed for the Tool's
|
|
328
328
|
execution, including the expected data type and description for
|
|
329
329
|
each input field. Structured as a stringified JSON schema, this
|
|
330
330
|
format ensures the Tool receives data in the expected format.
|
|
@@ -22,6 +22,8 @@ export declare namespace Chat {
|
|
|
22
22
|
resumedChatGroupId?: string;
|
|
23
23
|
/** A flag to enable verbose transcription. Set this query parameter to `true` to have unfinalized user transcripts be sent to the client as interim UserMessage messages. The [interim](/reference/empathic-voice-interface-evi/chat/chat#receive.User%20Message.interim) field on a [UserMessage](/reference/empathic-voice-interface-evi/chat/chat#receive.User%20Message.type) denotes whether the message is "interim" or "final." */
|
|
24
24
|
verboseTranscription?: boolean;
|
|
25
|
+
/** ID of the Voice to use for this chat. If specified, will override the voice set in the Config */
|
|
26
|
+
voiceId?: string;
|
|
25
27
|
/** Extra query parameters sent at WebSocket connection */
|
|
26
28
|
queryParams?: Record<string, string | string[] | object | object[]>;
|
|
27
29
|
}
|
|
@@ -14,7 +14,7 @@ export interface PostedPrompt {
|
|
|
14
14
|
/** An optional description of the Prompt version. */
|
|
15
15
|
versionDescription?: string;
|
|
16
16
|
/**
|
|
17
|
-
* Instructions used to shape EVI
|
|
17
|
+
* Instructions used to shape EVI's behavior, responses, and style.
|
|
18
18
|
*
|
|
19
19
|
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
20
20
|
*
|
package/api/resources/empathicVoice/resources/prompts/client/requests/PostedPromptVersion.d.ts
CHANGED
|
@@ -12,7 +12,7 @@ export interface PostedPromptVersion {
|
|
|
12
12
|
/** An optional description of the Prompt version. */
|
|
13
13
|
versionDescription?: string;
|
|
14
14
|
/**
|
|
15
|
-
* Instructions used to shape EVI
|
|
15
|
+
* Instructions used to shape EVI's behavior, responses, and style for this version of the Prompt.
|
|
16
16
|
*
|
|
17
17
|
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
18
18
|
*
|
package/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedTool.d.ts
CHANGED
|
@@ -21,7 +21,7 @@ export interface PostedUserDefinedTool {
|
|
|
21
21
|
/**
|
|
22
22
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
23
23
|
*
|
|
24
|
-
* These parameters define the inputs needed for the Tool
|
|
24
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the Tool receives data in the expected format.
|
|
25
25
|
*/
|
|
26
26
|
parameters: string;
|
|
27
27
|
/** Optional text passed to the supplemental LLM in place of the tool call result. The LLM then uses this text to generate a response back to the user, ensuring continuity in the conversation if the Tool errors. */
|
|
@@ -18,7 +18,7 @@ export interface PostedUserDefinedToolVersion {
|
|
|
18
18
|
/**
|
|
19
19
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
20
20
|
*
|
|
21
|
-
* These parameters define the inputs needed for the Tool
|
|
21
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the Tool receives data in the expected format.
|
|
22
22
|
*/
|
|
23
23
|
parameters: string;
|
|
24
24
|
/** Optional text passed to the supplemental LLM in place of the tool call result. The LLM then uses this text to generate a response back to the user, ensuring continuity in the conversation if the Tool errors. */
|
|
@@ -8,7 +8,7 @@ export interface AssistantEnd {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; for an Assistant End message, this must be `assistant_end`.
|
|
10
10
|
*
|
|
11
|
-
* This message indicates the conclusion of the assistant
|
|
11
|
+
* This message indicates the conclusion of the assistant's response, signaling that the assistant has finished speaking for the current conversational turn.
|
|
12
12
|
*/
|
|
13
13
|
type: "assistant_end";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -12,7 +12,7 @@ export interface AssistantInput {
|
|
|
12
12
|
/**
|
|
13
13
|
* Assistant text to synthesize into spoken audio and insert into the conversation.
|
|
14
14
|
*
|
|
15
|
-
* EVI uses this text to generate spoken audio using our proprietary expressive text-to-speech model. Our model adds appropriate emotional inflections and tones to the text based on the user
|
|
15
|
+
* EVI uses this text to generate spoken audio using our proprietary expressive text-to-speech model. Our model adds appropriate emotional inflections and tones to the text based on the user's expressions and the context of the conversation. The synthesized audio is streamed back to the user as an [Assistant Message](/reference/empathic-voice-interface-evi/chat/chat#receive.AssistantMessage.type).
|
|
16
16
|
*/
|
|
17
17
|
text: string;
|
|
18
18
|
}
|
|
@@ -9,7 +9,7 @@ export interface AssistantMessage {
|
|
|
9
9
|
/**
|
|
10
10
|
* The type of message sent through the socket; for an Assistant Message, this must be `assistant_message`.
|
|
11
11
|
*
|
|
12
|
-
* This message contains both a transcript of the assistant
|
|
12
|
+
* This message contains both a transcript of the assistant's response and the expression measurement predictions of the assistant's audio output.
|
|
13
13
|
*/
|
|
14
14
|
type: "assistant_message";
|
|
15
15
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -16,7 +16,7 @@ export interface AudioInput {
|
|
|
16
16
|
/**
|
|
17
17
|
* Base64 encoded audio input to insert into the conversation.
|
|
18
18
|
*
|
|
19
|
-
* The content of an Audio Input message is treated as the user
|
|
19
|
+
* The content of an Audio Input message is treated as the user's speech to EVI and must be streamed continuously. Pre-recorded audio files are not supported.
|
|
20
20
|
*
|
|
21
21
|
* For optimal transcription quality, the audio data should be transmitted in small chunks.
|
|
22
22
|
*
|
|
@@ -8,7 +8,7 @@ export interface PauseAssistantMessage {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; must be `pause_assistant_message` for our server to correctly identify and process it as a Pause Assistant message.
|
|
10
10
|
*
|
|
11
|
-
* Once this message is sent, EVI will not respond until a [Resume Assistant message](/reference/empathic-voice-interface-evi/chat/chat#send.ResumeAssistantMessage.type) is sent. When paused, EVI won
|
|
11
|
+
* Once this message is sent, EVI will not respond until a [Resume Assistant message](/reference/empathic-voice-interface-evi/chat/chat#send.ResumeAssistantMessage.type) is sent. When paused, EVI won't respond, but transcriptions of your audio inputs will still be recorded.
|
|
12
12
|
*/
|
|
13
13
|
type: "pause_assistant_message";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -13,7 +13,7 @@ export interface PostedLanguageModel {
|
|
|
13
13
|
/**
|
|
14
14
|
* The model temperature, with values between 0 to 1 (inclusive).
|
|
15
15
|
*
|
|
16
|
-
* Controls the randomness of the LLM
|
|
16
|
+
* Controls the randomness of the LLM's output, with values closer to 0 yielding focused, deterministic responses and values closer to 1 producing more creative, diverse responses.
|
|
17
17
|
*/
|
|
18
18
|
temperature?: number;
|
|
19
19
|
}
|
|
@@ -7,12 +7,12 @@
|
|
|
7
7
|
* Accepts a minimum value of 30 seconds and a maximum value of 1,800 seconds.
|
|
8
8
|
*/
|
|
9
9
|
export interface PostedTimeoutSpecsInactivity {
|
|
10
|
+
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
11
|
+
durationSecs?: number;
|
|
10
12
|
/**
|
|
11
13
|
* Boolean indicating if this timeout is enabled.
|
|
12
14
|
*
|
|
13
15
|
* If set to false, EVI will not timeout due to a specified duration of user inactivity being reached. However, the conversation will eventually disconnect after 1,800 seconds (30 minutes), which is the maximum WebSocket duration limit for EVI.
|
|
14
16
|
*/
|
|
15
17
|
enabled: boolean;
|
|
16
|
-
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
17
|
-
durationSecs?: number;
|
|
18
18
|
}
|
|
@@ -7,12 +7,12 @@
|
|
|
7
7
|
* Accepts a minimum value of 30 seconds and a maximum value of 1,800 seconds.
|
|
8
8
|
*/
|
|
9
9
|
export interface PostedTimeoutSpecsMaxDuration {
|
|
10
|
+
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
11
|
+
durationSecs?: number;
|
|
10
12
|
/**
|
|
11
13
|
* Boolean indicating if this timeout is enabled.
|
|
12
14
|
*
|
|
13
15
|
* If set to false, EVI will not timeout due to a specified maximum duration being reached. However, the conversation will eventually disconnect after 1,800 seconds (30 minutes), which is the maximum WebSocket duration limit for EVI.
|
|
14
16
|
*/
|
|
15
17
|
enabled: boolean;
|
|
16
|
-
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
17
|
-
durationSecs?: number;
|
|
18
18
|
}
|
|
@@ -13,7 +13,7 @@ export interface ReturnLanguageModel {
|
|
|
13
13
|
/**
|
|
14
14
|
* The model temperature, with values between 0 to 1 (inclusive).
|
|
15
15
|
*
|
|
16
|
-
* Controls the randomness of the LLM
|
|
16
|
+
* Controls the randomness of the LLM's output, with values closer to 0 yielding focused, deterministic responses and values closer to 1 producing more creative, diverse responses.
|
|
17
17
|
*/
|
|
18
18
|
temperature?: number;
|
|
19
19
|
}
|
|
@@ -11,11 +11,9 @@ export interface ReturnPrompt {
|
|
|
11
11
|
/** Identifier for a Prompt. Formatted as a UUID. */
|
|
12
12
|
id: string;
|
|
13
13
|
/**
|
|
14
|
-
* Instructions used to shape EVI
|
|
14
|
+
* Instructions used to shape EVI's behavior, responses, and style.
|
|
15
15
|
*
|
|
16
|
-
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
17
|
-
*
|
|
18
|
-
* For help writing a system prompt, see our [Prompting Guide](/docs/speech-to-speech-evi/guides/prompting).
|
|
16
|
+
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles. For help writing a system prompt, see our [Prompting Guide](/docs/speech-to-speech-evi/guides/prompting).
|
|
19
17
|
*/
|
|
20
18
|
text: string;
|
|
21
19
|
/**
|
|
@@ -35,7 +35,7 @@ export interface ReturnUserDefinedTool {
|
|
|
35
35
|
/**
|
|
36
36
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
37
37
|
*
|
|
38
|
-
* These parameters define the inputs needed for the Tool
|
|
38
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the tool receives data in the expected format.
|
|
39
39
|
*/
|
|
40
40
|
parameters: string;
|
|
41
41
|
}
|
|
@@ -23,7 +23,7 @@ export interface SessionSettings {
|
|
|
23
23
|
*/
|
|
24
24
|
customSessionId?: string;
|
|
25
25
|
/**
|
|
26
|
-
* Instructions used to shape EVI
|
|
26
|
+
* Instructions used to shape EVI's behavior, responses, and style for the session.
|
|
27
27
|
*
|
|
28
28
|
* When included in a Session Settings message, the provided Prompt overrides the existing one specified in the EVI configuration. If no Prompt was defined in the configuration, this Prompt will be the one used for the session.
|
|
29
29
|
*
|
|
@@ -49,7 +49,7 @@ export interface SessionSettings {
|
|
|
49
49
|
/**
|
|
50
50
|
* Third party API key for the supplemental language model.
|
|
51
51
|
*
|
|
52
|
-
* When provided, EVI will use this key instead of Hume
|
|
52
|
+
* When provided, EVI will use this key instead of Hume's API key for the supplemental LLM. This allows you to bypass rate limits and utilize your own API key as needed.
|
|
53
53
|
*/
|
|
54
54
|
languageModelApiKey?: string;
|
|
55
55
|
/**
|
|
@@ -10,7 +10,7 @@ export interface Tool {
|
|
|
10
10
|
/**
|
|
11
11
|
* Parameters of the tool. Is a stringified JSON schema.
|
|
12
12
|
*
|
|
13
|
-
* These parameters define the inputs needed for the tool
|
|
13
|
+
* These parameters define the inputs needed for the tool's execution, including the expected data type and description for each input field. Structured as a JSON schema, this format ensures the tool receives data in the expected format.
|
|
14
14
|
*/
|
|
15
15
|
parameters: string;
|
|
16
16
|
/** An optional description of what the tool does, used by the supplemental LLM to choose when and how to call the function. */
|
|
@@ -11,7 +11,7 @@ export interface ToolCallMessage {
|
|
|
11
11
|
/**
|
|
12
12
|
* Parameters of the tool.
|
|
13
13
|
*
|
|
14
|
-
* These parameters define the inputs needed for the tool
|
|
14
|
+
* These parameters define the inputs needed for the tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the tool receives data in the expected format.
|
|
15
15
|
*/
|
|
16
16
|
parameters: string;
|
|
17
17
|
/**
|
|
@@ -12,7 +12,7 @@ export interface UserInput {
|
|
|
12
12
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
13
13
|
customSessionId?: string;
|
|
14
14
|
/**
|
|
15
|
-
* User text to insert into the conversation. Text sent through a User Input message is treated as the user
|
|
15
|
+
* User text to insert into the conversation. Text sent through a User Input message is treated as the user's speech to EVI. EVI processes this input and provides a corresponding response.
|
|
16
16
|
*
|
|
17
17
|
* Expression measurement results are not available for User Input messages, as the prosody model relies on audio input and cannot process text alone.
|
|
18
18
|
*/
|
|
@@ -8,7 +8,7 @@ export interface UserInterruption {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; for a User Interruption message, this must be `user_interruption`.
|
|
10
10
|
*
|
|
11
|
-
* This message indicates the user has interrupted the assistant
|
|
11
|
+
* This message indicates the user has interrupted the assistant's response. EVI detects the interruption in real-time and sends this message to signal the interruption event. This message allows the system to stop the current audio playback, clear the audio queue, and prepare to handle new user input.
|
|
12
12
|
*/
|
|
13
13
|
type: "user_interruption";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -9,7 +9,7 @@ export interface UserMessage {
|
|
|
9
9
|
/**
|
|
10
10
|
* The type of message sent through the socket; for a User Message, this must be `user_message`.
|
|
11
11
|
*
|
|
12
|
-
* This message contains both a transcript of the user
|
|
12
|
+
* This message contains both a transcript of the user's input and the expression measurement predictions if the input was sent as an [Audio Input message](/reference/empathic-voice-interface-evi/chat/chat#send.AudioInput.type). Expression measurement predictions are not provided for a [User Input message](/reference/empathic-voice-interface-evi/chat/chat#send.UserInput.type), as the prosody model relies on audio input and cannot process text alone.
|
|
13
13
|
*/
|
|
14
14
|
type: "user_message";
|
|
15
15
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -22,6 +22,8 @@ export declare namespace Chat {
|
|
|
22
22
|
resumedChatGroupId?: string;
|
|
23
23
|
/** A flag to enable verbose transcription. Set this query parameter to `true` to have unfinalized user transcripts be sent to the client as interim UserMessage messages. The [interim](/reference/empathic-voice-interface-evi/chat/chat#receive.User%20Message.interim) field on a [UserMessage](/reference/empathic-voice-interface-evi/chat/chat#receive.User%20Message.type) denotes whether the message is "interim" or "final." */
|
|
24
24
|
verboseTranscription?: boolean;
|
|
25
|
+
/** ID of the Voice to use for this chat. If specified, will override the voice set in the Config */
|
|
26
|
+
voiceId?: string;
|
|
25
27
|
/** Extra query parameters sent at WebSocket connection */
|
|
26
28
|
queryParams?: Record<string, string | string[] | object | object[]>;
|
|
27
29
|
}
|
package/dist/api/resources/empathicVoice/resources/prompts/client/requests/PostedPrompt.d.ts
CHANGED
|
@@ -14,7 +14,7 @@ export interface PostedPrompt {
|
|
|
14
14
|
/** An optional description of the Prompt version. */
|
|
15
15
|
versionDescription?: string;
|
|
16
16
|
/**
|
|
17
|
-
* Instructions used to shape EVI
|
|
17
|
+
* Instructions used to shape EVI's behavior, responses, and style.
|
|
18
18
|
*
|
|
19
19
|
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
20
20
|
*
|
package/dist/api/resources/empathicVoice/resources/prompts/client/requests/PostedPromptVersion.d.ts
CHANGED
|
@@ -12,7 +12,7 @@ export interface PostedPromptVersion {
|
|
|
12
12
|
/** An optional description of the Prompt version. */
|
|
13
13
|
versionDescription?: string;
|
|
14
14
|
/**
|
|
15
|
-
* Instructions used to shape EVI
|
|
15
|
+
* Instructions used to shape EVI's behavior, responses, and style for this version of the Prompt.
|
|
16
16
|
*
|
|
17
17
|
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
18
18
|
*
|
package/dist/api/resources/empathicVoice/resources/tools/client/requests/PostedUserDefinedTool.d.ts
CHANGED
|
@@ -21,7 +21,7 @@ export interface PostedUserDefinedTool {
|
|
|
21
21
|
/**
|
|
22
22
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
23
23
|
*
|
|
24
|
-
* These parameters define the inputs needed for the Tool
|
|
24
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the Tool receives data in the expected format.
|
|
25
25
|
*/
|
|
26
26
|
parameters: string;
|
|
27
27
|
/** Optional text passed to the supplemental LLM in place of the tool call result. The LLM then uses this text to generate a response back to the user, ensuring continuity in the conversation if the Tool errors. */
|
|
@@ -18,7 +18,7 @@ export interface PostedUserDefinedToolVersion {
|
|
|
18
18
|
/**
|
|
19
19
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
20
20
|
*
|
|
21
|
-
* These parameters define the inputs needed for the Tool
|
|
21
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the Tool receives data in the expected format.
|
|
22
22
|
*/
|
|
23
23
|
parameters: string;
|
|
24
24
|
/** Optional text passed to the supplemental LLM in place of the tool call result. The LLM then uses this text to generate a response back to the user, ensuring continuity in the conversation if the Tool errors. */
|
|
@@ -8,7 +8,7 @@ export interface AssistantEnd {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; for an Assistant End message, this must be `assistant_end`.
|
|
10
10
|
*
|
|
11
|
-
* This message indicates the conclusion of the assistant
|
|
11
|
+
* This message indicates the conclusion of the assistant's response, signaling that the assistant has finished speaking for the current conversational turn.
|
|
12
12
|
*/
|
|
13
13
|
type: "assistant_end";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -12,7 +12,7 @@ export interface AssistantInput {
|
|
|
12
12
|
/**
|
|
13
13
|
* Assistant text to synthesize into spoken audio and insert into the conversation.
|
|
14
14
|
*
|
|
15
|
-
* EVI uses this text to generate spoken audio using our proprietary expressive text-to-speech model. Our model adds appropriate emotional inflections and tones to the text based on the user
|
|
15
|
+
* EVI uses this text to generate spoken audio using our proprietary expressive text-to-speech model. Our model adds appropriate emotional inflections and tones to the text based on the user's expressions and the context of the conversation. The synthesized audio is streamed back to the user as an [Assistant Message](/reference/empathic-voice-interface-evi/chat/chat#receive.AssistantMessage.type).
|
|
16
16
|
*/
|
|
17
17
|
text: string;
|
|
18
18
|
}
|
|
@@ -9,7 +9,7 @@ export interface AssistantMessage {
|
|
|
9
9
|
/**
|
|
10
10
|
* The type of message sent through the socket; for an Assistant Message, this must be `assistant_message`.
|
|
11
11
|
*
|
|
12
|
-
* This message contains both a transcript of the assistant
|
|
12
|
+
* This message contains both a transcript of the assistant's response and the expression measurement predictions of the assistant's audio output.
|
|
13
13
|
*/
|
|
14
14
|
type: "assistant_message";
|
|
15
15
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -16,7 +16,7 @@ export interface AudioInput {
|
|
|
16
16
|
/**
|
|
17
17
|
* Base64 encoded audio input to insert into the conversation.
|
|
18
18
|
*
|
|
19
|
-
* The content of an Audio Input message is treated as the user
|
|
19
|
+
* The content of an Audio Input message is treated as the user's speech to EVI and must be streamed continuously. Pre-recorded audio files are not supported.
|
|
20
20
|
*
|
|
21
21
|
* For optimal transcription quality, the audio data should be transmitted in small chunks.
|
|
22
22
|
*
|
|
@@ -8,7 +8,7 @@ export interface PauseAssistantMessage {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; must be `pause_assistant_message` for our server to correctly identify and process it as a Pause Assistant message.
|
|
10
10
|
*
|
|
11
|
-
* Once this message is sent, EVI will not respond until a [Resume Assistant message](/reference/empathic-voice-interface-evi/chat/chat#send.ResumeAssistantMessage.type) is sent. When paused, EVI won
|
|
11
|
+
* Once this message is sent, EVI will not respond until a [Resume Assistant message](/reference/empathic-voice-interface-evi/chat/chat#send.ResumeAssistantMessage.type) is sent. When paused, EVI won't respond, but transcriptions of your audio inputs will still be recorded.
|
|
12
12
|
*/
|
|
13
13
|
type: "pause_assistant_message";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -13,7 +13,7 @@ export interface PostedLanguageModel {
|
|
|
13
13
|
/**
|
|
14
14
|
* The model temperature, with values between 0 to 1 (inclusive).
|
|
15
15
|
*
|
|
16
|
-
* Controls the randomness of the LLM
|
|
16
|
+
* Controls the randomness of the LLM's output, with values closer to 0 yielding focused, deterministic responses and values closer to 1 producing more creative, diverse responses.
|
|
17
17
|
*/
|
|
18
18
|
temperature?: number;
|
|
19
19
|
}
|
|
@@ -7,12 +7,12 @@
|
|
|
7
7
|
* Accepts a minimum value of 30 seconds and a maximum value of 1,800 seconds.
|
|
8
8
|
*/
|
|
9
9
|
export interface PostedTimeoutSpecsInactivity {
|
|
10
|
+
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
11
|
+
durationSecs?: number;
|
|
10
12
|
/**
|
|
11
13
|
* Boolean indicating if this timeout is enabled.
|
|
12
14
|
*
|
|
13
15
|
* If set to false, EVI will not timeout due to a specified duration of user inactivity being reached. However, the conversation will eventually disconnect after 1,800 seconds (30 minutes), which is the maximum WebSocket duration limit for EVI.
|
|
14
16
|
*/
|
|
15
17
|
enabled: boolean;
|
|
16
|
-
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
17
|
-
durationSecs?: number;
|
|
18
18
|
}
|
|
@@ -7,12 +7,12 @@
|
|
|
7
7
|
* Accepts a minimum value of 30 seconds and a maximum value of 1,800 seconds.
|
|
8
8
|
*/
|
|
9
9
|
export interface PostedTimeoutSpecsMaxDuration {
|
|
10
|
+
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
11
|
+
durationSecs?: number;
|
|
10
12
|
/**
|
|
11
13
|
* Boolean indicating if this timeout is enabled.
|
|
12
14
|
*
|
|
13
15
|
* If set to false, EVI will not timeout due to a specified maximum duration being reached. However, the conversation will eventually disconnect after 1,800 seconds (30 minutes), which is the maximum WebSocket duration limit for EVI.
|
|
14
16
|
*/
|
|
15
17
|
enabled: boolean;
|
|
16
|
-
/** Duration in seconds for the timeout (e.g. 600 seconds represents 10 minutes). */
|
|
17
|
-
durationSecs?: number;
|
|
18
18
|
}
|
|
@@ -13,7 +13,7 @@ export interface ReturnLanguageModel {
|
|
|
13
13
|
/**
|
|
14
14
|
* The model temperature, with values between 0 to 1 (inclusive).
|
|
15
15
|
*
|
|
16
|
-
* Controls the randomness of the LLM
|
|
16
|
+
* Controls the randomness of the LLM's output, with values closer to 0 yielding focused, deterministic responses and values closer to 1 producing more creative, diverse responses.
|
|
17
17
|
*/
|
|
18
18
|
temperature?: number;
|
|
19
19
|
}
|
|
@@ -11,11 +11,9 @@ export interface ReturnPrompt {
|
|
|
11
11
|
/** Identifier for a Prompt. Formatted as a UUID. */
|
|
12
12
|
id: string;
|
|
13
13
|
/**
|
|
14
|
-
* Instructions used to shape EVI
|
|
14
|
+
* Instructions used to shape EVI's behavior, responses, and style.
|
|
15
15
|
*
|
|
16
|
-
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles.
|
|
17
|
-
*
|
|
18
|
-
* For help writing a system prompt, see our [Prompting Guide](/docs/speech-to-speech-evi/guides/prompting).
|
|
16
|
+
* You can use the Prompt to define a specific goal or role for EVI, specifying how it should act or what it should focus on during the conversation. For example, EVI can be instructed to act as a customer support representative, a fitness coach, or a travel advisor, each with its own set of behaviors and response styles. For help writing a system prompt, see our [Prompting Guide](/docs/speech-to-speech-evi/guides/prompting).
|
|
19
17
|
*/
|
|
20
18
|
text: string;
|
|
21
19
|
/**
|
|
@@ -35,7 +35,7 @@ export interface ReturnUserDefinedTool {
|
|
|
35
35
|
/**
|
|
36
36
|
* Stringified JSON defining the parameters used by this version of the Tool.
|
|
37
37
|
*
|
|
38
|
-
* These parameters define the inputs needed for the Tool
|
|
38
|
+
* These parameters define the inputs needed for the Tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the tool receives data in the expected format.
|
|
39
39
|
*/
|
|
40
40
|
parameters: string;
|
|
41
41
|
}
|
|
@@ -23,7 +23,7 @@ export interface SessionSettings {
|
|
|
23
23
|
*/
|
|
24
24
|
customSessionId?: string;
|
|
25
25
|
/**
|
|
26
|
-
* Instructions used to shape EVI
|
|
26
|
+
* Instructions used to shape EVI's behavior, responses, and style for the session.
|
|
27
27
|
*
|
|
28
28
|
* When included in a Session Settings message, the provided Prompt overrides the existing one specified in the EVI configuration. If no Prompt was defined in the configuration, this Prompt will be the one used for the session.
|
|
29
29
|
*
|
|
@@ -49,7 +49,7 @@ export interface SessionSettings {
|
|
|
49
49
|
/**
|
|
50
50
|
* Third party API key for the supplemental language model.
|
|
51
51
|
*
|
|
52
|
-
* When provided, EVI will use this key instead of Hume
|
|
52
|
+
* When provided, EVI will use this key instead of Hume's API key for the supplemental LLM. This allows you to bypass rate limits and utilize your own API key as needed.
|
|
53
53
|
*/
|
|
54
54
|
languageModelApiKey?: string;
|
|
55
55
|
/**
|
|
@@ -10,7 +10,7 @@ export interface Tool {
|
|
|
10
10
|
/**
|
|
11
11
|
* Parameters of the tool. Is a stringified JSON schema.
|
|
12
12
|
*
|
|
13
|
-
* These parameters define the inputs needed for the tool
|
|
13
|
+
* These parameters define the inputs needed for the tool's execution, including the expected data type and description for each input field. Structured as a JSON schema, this format ensures the tool receives data in the expected format.
|
|
14
14
|
*/
|
|
15
15
|
parameters: string;
|
|
16
16
|
/** An optional description of what the tool does, used by the supplemental LLM to choose when and how to call the function. */
|
|
@@ -11,7 +11,7 @@ export interface ToolCallMessage {
|
|
|
11
11
|
/**
|
|
12
12
|
* Parameters of the tool.
|
|
13
13
|
*
|
|
14
|
-
* These parameters define the inputs needed for the tool
|
|
14
|
+
* These parameters define the inputs needed for the tool's execution, including the expected data type and description for each input field. Structured as a stringified JSON schema, this format ensures the tool receives data in the expected format.
|
|
15
15
|
*/
|
|
16
16
|
parameters: string;
|
|
17
17
|
/**
|
|
@@ -12,7 +12,7 @@ export interface UserInput {
|
|
|
12
12
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
13
13
|
customSessionId?: string;
|
|
14
14
|
/**
|
|
15
|
-
* User text to insert into the conversation. Text sent through a User Input message is treated as the user
|
|
15
|
+
* User text to insert into the conversation. Text sent through a User Input message is treated as the user's speech to EVI. EVI processes this input and provides a corresponding response.
|
|
16
16
|
*
|
|
17
17
|
* Expression measurement results are not available for User Input messages, as the prosody model relies on audio input and cannot process text alone.
|
|
18
18
|
*/
|
|
@@ -8,7 +8,7 @@ export interface UserInterruption {
|
|
|
8
8
|
/**
|
|
9
9
|
* The type of message sent through the socket; for a User Interruption message, this must be `user_interruption`.
|
|
10
10
|
*
|
|
11
|
-
* This message indicates the user has interrupted the assistant
|
|
11
|
+
* This message indicates the user has interrupted the assistant's response. EVI detects the interruption in real-time and sends this message to signal the interruption event. This message allows the system to stop the current audio playback, clear the audio queue, and prepare to handle new user input.
|
|
12
12
|
*/
|
|
13
13
|
type: "user_interruption";
|
|
14
14
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -9,7 +9,7 @@ export interface UserMessage {
|
|
|
9
9
|
/**
|
|
10
10
|
* The type of message sent through the socket; for a User Message, this must be `user_message`.
|
|
11
11
|
*
|
|
12
|
-
* This message contains both a transcript of the user
|
|
12
|
+
* This message contains both a transcript of the user's input and the expression measurement predictions if the input was sent as an [Audio Input message](/reference/empathic-voice-interface-evi/chat/chat#send.AudioInput.type). Expression measurement predictions are not provided for a [User Input message](/reference/empathic-voice-interface-evi/chat/chat#send.UserInput.type), as the prosody model relies on audio input and cannot process text alone.
|
|
13
13
|
*/
|
|
14
14
|
type: "user_message";
|
|
15
15
|
/** Used to manage conversational state, correlate frontend and backend data, and persist conversations across EVI sessions. */
|
|
@@ -7,7 +7,7 @@ import * as core from "../../../../core";
|
|
|
7
7
|
export declare const PostedTimeoutSpecsInactivity: core.serialization.ObjectSchema<serializers.empathicVoice.PostedTimeoutSpecsInactivity.Raw, Hume.empathicVoice.PostedTimeoutSpecsInactivity>;
|
|
8
8
|
export declare namespace PostedTimeoutSpecsInactivity {
|
|
9
9
|
interface Raw {
|
|
10
|
-
enabled: boolean;
|
|
11
10
|
duration_secs?: number | null;
|
|
11
|
+
enabled: boolean;
|
|
12
12
|
}
|
|
13
13
|
}
|
|
@@ -39,6 +39,6 @@ Object.defineProperty(exports, "__esModule", { value: true });
|
|
|
39
39
|
exports.PostedTimeoutSpecsInactivity = void 0;
|
|
40
40
|
const core = __importStar(require("../../../../core"));
|
|
41
41
|
exports.PostedTimeoutSpecsInactivity = core.serialization.object({
|
|
42
|
-
enabled: core.serialization.boolean(),
|
|
43
42
|
durationSecs: core.serialization.property("duration_secs", core.serialization.number().optional()),
|
|
43
|
+
enabled: core.serialization.boolean(),
|
|
44
44
|
});
|
|
@@ -7,7 +7,7 @@ import * as core from "../../../../core";
|
|
|
7
7
|
export declare const PostedTimeoutSpecsMaxDuration: core.serialization.ObjectSchema<serializers.empathicVoice.PostedTimeoutSpecsMaxDuration.Raw, Hume.empathicVoice.PostedTimeoutSpecsMaxDuration>;
|
|
8
8
|
export declare namespace PostedTimeoutSpecsMaxDuration {
|
|
9
9
|
interface Raw {
|
|
10
|
-
enabled: boolean;
|
|
11
10
|
duration_secs?: number | null;
|
|
11
|
+
enabled: boolean;
|
|
12
12
|
}
|
|
13
13
|
}
|
|
@@ -39,6 +39,6 @@ Object.defineProperty(exports, "__esModule", { value: true });
|
|
|
39
39
|
exports.PostedTimeoutSpecsMaxDuration = void 0;
|
|
40
40
|
const core = __importStar(require("../../../../core"));
|
|
41
41
|
exports.PostedTimeoutSpecsMaxDuration = core.serialization.object({
|
|
42
|
-
enabled: core.serialization.boolean(),
|
|
43
42
|
durationSecs: core.serialization.property("duration_secs", core.serialization.number().optional()),
|
|
43
|
+
enabled: core.serialization.boolean(),
|
|
44
44
|
});
|
package/dist/version.d.ts
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
export declare const SDK_VERSION = "0.13.
|
|
1
|
+
export declare const SDK_VERSION = "0.13.1";
|
package/dist/version.js
CHANGED
package/package.json
CHANGED
|
@@ -7,7 +7,7 @@ import * as core from "../../../../core";
|
|
|
7
7
|
export declare const PostedTimeoutSpecsInactivity: core.serialization.ObjectSchema<serializers.empathicVoice.PostedTimeoutSpecsInactivity.Raw, Hume.empathicVoice.PostedTimeoutSpecsInactivity>;
|
|
8
8
|
export declare namespace PostedTimeoutSpecsInactivity {
|
|
9
9
|
interface Raw {
|
|
10
|
-
enabled: boolean;
|
|
11
10
|
duration_secs?: number | null;
|
|
11
|
+
enabled: boolean;
|
|
12
12
|
}
|
|
13
13
|
}
|
|
@@ -39,6 +39,6 @@ Object.defineProperty(exports, "__esModule", { value: true });
|
|
|
39
39
|
exports.PostedTimeoutSpecsInactivity = void 0;
|
|
40
40
|
const core = __importStar(require("../../../../core"));
|
|
41
41
|
exports.PostedTimeoutSpecsInactivity = core.serialization.object({
|
|
42
|
-
enabled: core.serialization.boolean(),
|
|
43
42
|
durationSecs: core.serialization.property("duration_secs", core.serialization.number().optional()),
|
|
43
|
+
enabled: core.serialization.boolean(),
|
|
44
44
|
});
|
|
@@ -7,7 +7,7 @@ import * as core from "../../../../core";
|
|
|
7
7
|
export declare const PostedTimeoutSpecsMaxDuration: core.serialization.ObjectSchema<serializers.empathicVoice.PostedTimeoutSpecsMaxDuration.Raw, Hume.empathicVoice.PostedTimeoutSpecsMaxDuration>;
|
|
8
8
|
export declare namespace PostedTimeoutSpecsMaxDuration {
|
|
9
9
|
interface Raw {
|
|
10
|
-
enabled: boolean;
|
|
11
10
|
duration_secs?: number | null;
|
|
11
|
+
enabled: boolean;
|
|
12
12
|
}
|
|
13
13
|
}
|
|
@@ -39,6 +39,6 @@ Object.defineProperty(exports, "__esModule", { value: true });
|
|
|
39
39
|
exports.PostedTimeoutSpecsMaxDuration = void 0;
|
|
40
40
|
const core = __importStar(require("../../../../core"));
|
|
41
41
|
exports.PostedTimeoutSpecsMaxDuration = core.serialization.object({
|
|
42
|
-
enabled: core.serialization.boolean(),
|
|
43
42
|
durationSecs: core.serialization.property("duration_secs", core.serialization.number().optional()),
|
|
43
|
+
enabled: core.serialization.boolean(),
|
|
44
44
|
});
|
package/version.d.ts
CHANGED
|
@@ -1 +1 @@
|
|
|
1
|
-
export declare const SDK_VERSION = "0.13.
|
|
1
|
+
export declare const SDK_VERSION = "0.13.1";
|
package/version.js
CHANGED