More work on llm-as-judge phrase endpointing

examples/foundational/07p-interruptible-google-audio-in.py

@@ -0,0 +1,274 @@
+#
+# Copyright (c) 2024, Daily
+#
+# SPDX-License-Identifier: BSD 2-Clause License
+#
+
+import aiohttp
+import asyncio
+import os
+import sys
+
+import google.ai.generativelanguage as glm
+
+from dataclasses import dataclass
+from dotenv import load_dotenv
+from loguru import logger
+from runner import configure
+
+from pipecat.audio.vad.silero import SileroVADAnalyzer
+from pipecat.pipeline.pipeline import Pipeline
+from pipecat.pipeline.runner import PipelineRunner
+from pipecat.pipeline.task import PipelineParams, PipelineTask
+from pipecat.processors.aggregators.openai_llm_context import OpenAILLMContext
+from pipecat.services.cartesia import CartesiaTTSService
+from pipecat.services.google import GoogleLLMService
+from pipecat.processors.frame_processor import FrameProcessor
+from pipecat.transports.services.daily import DailyParams, DailyTransport
+from pipecat.frames.frames import (
+    LLMFullResponseStartFrame,
+    LLMFullResponseEndFrame,
+    InputAudioRawFrame,
+    Frame,
+    StartInterruptionFrame,
+    TextFrame,
+    TranscriptionFrame,
+    UserStartedSpeakingFrame,
+    UserStoppedSpeakingFrame,
+)
+
+load_dotenv(override=True)
+
+logger.remove(0)
+logger.add(sys.stderr, level="DEBUG")
+
+marker = "|----|"
+system_message = f"""
+You are a helpful LLM in a WebRTC call. Your goals are to be helpful and brief in your responses.
+
+You are expert at transcribing audio to text. You will receive a mixture of audio and text input. When
+asked to transcribe what the user said, output an exact, word-for-word transcription.
+
+Your output will be converted to audio so don't include special characters in your answers.
+
+Each time you answer, you should respond in three parts.
+
+1. Transcribe exactly what the user said.
+2. Output the separator field '{marker}'.
+3. Respond to the user's input in a helpful, creative way using only simple text and punctuation.
+
+Example:
+
+User: How many ounces are in a pound?
+
+You: How many ounces are in a pound?
+{marker}
+There are 16 ounces in a pound.
+"""
+
+
+@dataclass
+class MagicDemoTranscriptionFrame(Frame):
+    text: str
+
+
+class UserAudioCollector(FrameProcessor):
+    def __init__(self, context, user_context_aggregator):
+        super().__init__()
+        self._context = context
+        self._user_context_aggregator = user_context_aggregator
+        self._audio_frames = []
+        self._start_secs = 0.2  # this should match VAD start_secs (hardcoding for now)
+        self._user_speaking = False
+
+    async def process_frame(self, frame, direction):
+        await super().process_frame(frame, direction)
+
+        if isinstance(frame, TranscriptionFrame):
+            # We could gracefully handle both audio input and text/transcription input ...
+            # but let's leave that as an exercise to the reader. :-)
+            return
+        if isinstance(frame, UserStartedSpeakingFrame):
+            self._user_speaking = True
+        elif isinstance(frame, UserStoppedSpeakingFrame):
+            self._user_speaking = False
+            self._context.add_audio_frames_message(audio_frames=self._audio_frames)
+            await self._user_context_aggregator.push_frame(
+                self._user_context_aggregator.get_context_frame()
+            )
+        elif isinstance(frame, InputAudioRawFrame):
+            if self._user_speaking:
+                self._audio_frames.append(frame)
+            else:
+                # Append the audio frame to our buffer. Treat the buffer as a ring buffer, dropping the oldest
+                # frames as necessary. Assume all audio frames have the same duration.
+                self._audio_frames.append(frame)
+                frame_duration = len(frame.audio) / 16 * frame.num_channels / frame.sample_rate
+                buffer_duration = frame_duration * len(self._audio_frames)
+                while buffer_duration > self._start_secs:
+                    self._audio_frames.pop(0)
+                    buffer_duration -= frame_duration
+
+        await self.push_frame(frame, direction)
+
+
+class TranscriptExtractor(FrameProcessor):
+    def __init__(self, context):
+        super().__init__()
+        self._context = context
+        self._accumulator = ""
+        self._processing_llm_response = False
+        self._accumulating_transcript = False
+
+    def reset(self):
+        self._accumulator = ""
+        self._processing_llm_response = False
+        self._accumulating_transcript = False
+
+    async def process_frame(self, frame, direction):
+        await super().process_frame(frame, direction)
+        if isinstance(frame, LLMFullResponseStartFrame):
+            self._processing_llm_response = True
+            self._accumulating_transcript = True
+        elif isinstance(frame, TextFrame) and self._processing_llm_response:
+            if self._accumulating_transcript:
+                text = frame.text
+                split_index = text.find(marker)
+                if split_index < 0:
+                    self._accumulator += frame.text
+                    # do not push this frame
+                    return
+                else:
+                    self._accumulating_transcript = False
+                    self._accumulator += text[:split_index]
+                    frame.text = text[split_index + len(marker) :]
+            await self.push_frame(frame)
+            return
+        elif isinstance(frame, LLMFullResponseEndFrame):
+            await self.push_frame(MagicDemoTranscriptionFrame(text=self._accumulator.strip()))
+            self.reset()
+
+        await self.push_frame(frame, direction)
+
+
+class TanscriptionContextFixup(FrameProcessor):
+    def __init__(self, context):
+        super().__init__()
+        self._context = context
+        self._transcript = "THIS IS A TRANSCRIPT"
+
+    def swap_user_audio(self):
+        if not self._transcript:
+            return
+        message = self._context.messages[-2]
+        last_part = message.parts[-1]
+        if (
+            message.role == "user"
+            and last_part.inline_data
+            and last_part.inline_data.mime_type == "audio/wav"
+        ):
+            self._context.messages[-2] = glm.Content(
+                role="user", parts=[glm.Part(text=self._transcript)]
+            )
+
+    def add_transcript_back_to_inference_output(self):
+        if not self._transcript:
+            return
+        message = self._context.messages[-1]
+        last_part = message.parts[-1]
+        if message.role == "model" and last_part.text:
+            self._context.messages[-1].parts[-1].text += f"\n\n{marker}\n{self._transcript}\n"
+
+    async def process_frame(self, frame, direction):
+        await super().process_frame(frame, direction)
+
+        if isinstance(frame, MagicDemoTranscriptionFrame):
+            self._transcript = frame.text
+        elif isinstance(frame, LLMFullResponseEndFrame) or isinstance(
+            frame, StartInterruptionFrame
+        ):
+            self.swap_user_audio()
+            self.add_transcript_back_to_inference_output()
+            self._transcript = ""
+
+        await self.push_frame(frame, direction)
+
+
+async def main():
+    async with aiohttp.ClientSession() as session:
+        (room_url, token) = await configure(session)
+
+        transport = DailyTransport(
+            room_url,
+            token,
+            "Respond bot",
+            DailyParams(
+                audio_out_enabled=True,
+                # No transcription at all. just audio input to Gemini!
+                # transcription_enabled=True,
+                vad_enabled=True,
+                vad_analyzer=SileroVADAnalyzer(),
+                vad_audio_passthrough=True,
+            ),
+        )
+
+        tts = CartesiaTTSService(
+            api_key=os.getenv("CARTESIA_API_KEY"),
+            voice_id="79a125e8-cd45-4c13-8a67-188112f4dd22",  # British Lady
+        )
+
+        llm = GoogleLLMService(model="gemini-1.5-flash-latest", api_key=os.getenv("GOOGLE_API_KEY"))
+
+        messages = [
+            {
+                "role": "system",
+                "content": system_message,
+            },
+            {
+                "role": "user",
+                "content": "Start by saying hello.",
+            },
+        ]
+
+        context = OpenAILLMContext(messages)
+        context_aggregator = llm.create_context_aggregator(context)
+        audio_collector = UserAudioCollector(context, context_aggregator.user())
+        pull_transcript_out_of_llm_output = TranscriptExtractor(context)
+        fixup_context_messages = TanscriptionContextFixup(context)
+
+        pipeline = Pipeline(
+            [
+                transport.input(),  # Transport user input
+                audio_collector,
+                context_aggregator.user(),  # User responses
+                llm,  # LLM
+                pull_transcript_out_of_llm_output,
+                tts,  # TTS
+                transport.output(),  # Transport bot output
+                context_aggregator.assistant(),  # Assistant spoken responses
+                fixup_context_messages,
+            ]
+        )
+
+        task = PipelineTask(
+            pipeline,
+            PipelineParams(
+                allow_interruptions=True,
+                enable_metrics=True,
+                enable_usage_metrics=True,
+            ),
+        )
+
+        @transport.event_handler("on_first_participant_joined")
+        async def on_first_participant_joined(transport, participant):
+            await transport.capture_participant_transcription(participant["id"])
+            # Kick off the conversation.
+            await task.queue_frames([context_aggregator.user().get_context_frame()])
+
+        runner = PipelineRunner()
+
+        await runner.run(task)
+
+
+if __name__ == "__main__":
+    asyncio.run(main())