checkpoint

simple_conversation.py

@@ -19,6 +19,7 @@ from vocode.models.transcriber import (
 )
 from vocode.models.agent import (
     ChatGPTAgentConfig,
+    FillerAudioConfig,
     RESTfulUserImplementedAgentConfig,
     WebSocketUserImplementedAgentConfig,
     EchoAgentConfig,
@@ -43,12 +44,14 @@ if __name__ == "__main__":
         input_device=microphone_input,
         output_device=speaker_output,
         transcriber_config=DeepgramTranscriberConfig.from_input_device(
-            microphone_input, endpointing_config=PunctuationEndpointingConfig()
+            microphone_input
         ),
         agent_config=ChatGPTAgentConfig(
             initial_message=BaseMessage(text="Hello!"),
             prompt_preamble="The AI is having a pleasant conversation about life.",
-            generate_responses=True,
+            generate_responses=False,
+            end_conversation_on_goodbye=True,
+            send_filler_audio=FillerAudioConfig(use_typing_noise=True),
         ),
         synthesizer_config=AzureSynthesizerConfig.from_output_device(speaker_output),
     )

simple_outbound_call.py

@@ -1,8 +1,15 @@
+from vocode.models.synthesizer import AzureSynthesizerConfig
+from vocode.output_device.telephone_output import TelephoneOutput
 from vocode.telephony.outbound_call import OutboundCall
 from vocode.models.telephony import CallEntity
-from vocode.models.agent import EchoAgentConfig, WebSocketUserImplementedAgentConfig
+from vocode.models.agent import (
+    EchoAgentConfig,
+    ChatGPTAgentConfig,
+    WebSocketUserImplementedAgentConfig,
+)
+from vocode.models.message import BaseMessage
 
-if __name__ == '__main__':
+if __name__ == "__main__":
     call = OutboundCall(
         recipient=CallEntity(
             phone_number="+11234567890",
@@ -10,6 +17,15 @@ if __name__ == '__main__':
         caller=CallEntity(
             phone_number="+11234567890",
         ),
-        agent_config=EchoAgentConfig(initial_message="Hello!")
+        agent_config=ChatGPTAgentConfig(
+            initial_message=BaseMessage(text="the quick fox jumped over the lazy dog "),
+            prompt_preamble="respond two sentences at a time",
+            generate_responses=True,
+            end_conversation_on_goodbye=True,
+            send_filler_audio=True,
+        ),
+        synthesizer_config=AzureSynthesizerConfig.from_output_device(
+            output_device=TelephoneOutput(), voice_name="en-US-JennyNeural"
+        ),
     )
     call.start()

vocode/models/agent.py

@@ -1,6 +1,8 @@
 from typing import Optional, Union
 from enum import Enum
 
+from pydantic import validator
+
 from vocode.models.message import BaseMessage
 from .model import TypedModel, BaseModel
 
@@ -23,6 +25,16 @@ class AgentType(str, Enum):
 
 class FillerAudioConfig(BaseModel):
     silence_threshold_seconds: float = FILLER_AUDIO_DEFAULT_SILENCE_THRESHOLD_SECONDS
+    use_phrases: bool = True
+    use_typing_noise: bool = False
+
+    @validator("use_typing_noise")
+    def typing_noise_excludes_phrases(cls, v, values):
+        if v and values.get("use_phrases"):
+            values["use_phrases"] = False
+        if not v and not values.get("use_phrases"):
+            raise ValueError("must use either typing noise or phrases for filler audio")
+        return v
 
 
 class AgentConfig(TypedModel, type=AgentType.BASE):