Spaces:
Runtime error
Runtime error
create respond to prompt actor
Browse files
charles_actor.py
CHANGED
|
@@ -27,10 +27,10 @@ class CharlesActor:
|
|
| 27 |
from speech_to_text_vosk_actor import SpeechToTextVoskActor
|
| 28 |
self._speech_to_text_actor = SpeechToTextVoskActor.remote()
|
| 29 |
|
| 30 |
-
|
| 31 |
-
|
| 32 |
-
|
| 33 |
-
|
| 34 |
self._debug_queue = [
|
| 35 |
# "hello, how are you today?",
|
| 36 |
# "hmm, interesting, tell me more about that.",
|
|
@@ -55,7 +55,7 @@ class CharlesActor:
|
|
| 55 |
while True:
|
| 56 |
if len(self._debug_queue) > 0:
|
| 57 |
prompt = self._debug_queue.pop(0)
|
| 58 |
-
await self.
|
| 59 |
audio_frames = await self._streamlit_av_queue.get_audio_frames_async()
|
| 60 |
if len(audio_frames) > 0:
|
| 61 |
total_audio_frames += len(audio_frames)
|
|
@@ -79,7 +79,7 @@ class CharlesActor:
|
|
| 79 |
table_content = "| System 1 Audio History |\n| --- |\n"
|
| 80 |
table_content += "\n".join([f"| {item} |" for item in reversed(system_one_audio_history)])
|
| 81 |
self._system_one_audio_history_output = table_content
|
| 82 |
-
|
| 83 |
|
| 84 |
# video_frames = await self._streamlit_av_queue.get_video_frames_async()
|
| 85 |
# if len(video_frames) > 0:
|
|
@@ -127,5 +127,6 @@ if __name__ == "__main__":
|
|
| 127 |
time.sleep(1)
|
| 128 |
state = charles_actor.get_state.remote()
|
| 129 |
print(f"Charles is in state: {ray.get(state)}")
|
| 130 |
-
except KeyboardInterrupt:
|
| 131 |
print("Script was manually terminated")
|
|
|
|
|
|
| 27 |
from speech_to_text_vosk_actor import SpeechToTextVoskActor
|
| 28 |
self._speech_to_text_actor = SpeechToTextVoskActor.remote()
|
| 29 |
|
| 30 |
+
print("003")
|
| 31 |
+
from respond_to_prompt_actor import RespondToPromptActor
|
| 32 |
+
self._respond_to_prompt_actor = RespondToPromptActor.remote()
|
| 33 |
+
|
| 34 |
self._debug_queue = [
|
| 35 |
# "hello, how are you today?",
|
| 36 |
# "hmm, interesting, tell me more about that.",
|
|
|
|
| 55 |
while True:
|
| 56 |
if len(self._debug_queue) > 0:
|
| 57 |
prompt = self._debug_queue.pop(0)
|
| 58 |
+
await self._respond_to_prompt_actor.enqueue(prompt)
|
| 59 |
audio_frames = await self._streamlit_av_queue.get_audio_frames_async()
|
| 60 |
if len(audio_frames) > 0:
|
| 61 |
total_audio_frames += len(audio_frames)
|
|
|
|
| 79 |
table_content = "| System 1 Audio History |\n| --- |\n"
|
| 80 |
table_content += "\n".join([f"| {item} |" for item in reversed(system_one_audio_history)])
|
| 81 |
self._system_one_audio_history_output = table_content
|
| 82 |
+
self._respond_to_prompt_actor.enqueue_prompt.remote(prompt)
|
| 83 |
|
| 84 |
# video_frames = await self._streamlit_av_queue.get_video_frames_async()
|
| 85 |
# if len(video_frames) > 0:
|
|
|
|
| 127 |
time.sleep(1)
|
| 128 |
state = charles_actor.get_state.remote()
|
| 129 |
print(f"Charles is in state: {ray.get(state)}")
|
| 130 |
+
except KeyboardInterrupt as e:
|
| 131 |
print("Script was manually terminated")
|
| 132 |
+
throw(e)
|
chat_pipeline.py β legacy_to_delete/chat_pipeline.py
RENAMED
|
File without changes
|
debug.py β legacy_to_delete/debug.py
RENAMED
|
File without changes
|
respond_to_prompt_actor.py
ADDED
|
@@ -0,0 +1,77 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import ray
|
| 2 |
+
from ray.util.queue import Queue
|
| 3 |
+
from dotenv import load_dotenv
|
| 4 |
+
from audio_stream_processor import AudioStreamProcessor
|
| 5 |
+
from streaming_chat_service import StreamingChatService
|
| 6 |
+
|
| 7 |
+
# from ray.actor import ActorHandle
|
| 8 |
+
|
| 9 |
+
@ray.remote
|
| 10 |
+
class PromptToLLMActor:
|
| 11 |
+
def __init__(self, input_queue, output_queue, voice_id):
|
| 12 |
+
load_dotenv()
|
| 13 |
+
self.input_queue = input_queue
|
| 14 |
+
self.output_queue = output_queue
|
| 15 |
+
self.audio_processor = AudioStreamProcessor()
|
| 16 |
+
self.chat_service = StreamingChatService(self.audio_processor, voice_id=voice_id)
|
| 17 |
+
|
| 18 |
+
async def run(self):
|
| 19 |
+
while True:
|
| 20 |
+
prompt = self.input_queue.get()
|
| 21 |
+
async for sentence in self.chat_service.get_responses_as_sentances_async(prompt):
|
| 22 |
+
if self.chat_service.ignore_sentence(sentence):
|
| 23 |
+
continue
|
| 24 |
+
print(f"{sentence}")
|
| 25 |
+
self.output_queue.put(sentence)
|
| 26 |
+
|
| 27 |
+
@ray.remote
|
| 28 |
+
class LLMSentanceToSpeechActor:
|
| 29 |
+
def __init__(self, input_queue, output_queue, voice_id):
|
| 30 |
+
load_dotenv()
|
| 31 |
+
self.input_queue = input_queue
|
| 32 |
+
self.output_queue = output_queue
|
| 33 |
+
self.audio_processor = AudioStreamProcessor()
|
| 34 |
+
self.chat_service = StreamingChatService(self.audio_processor, voice_id=voice_id)
|
| 35 |
+
|
| 36 |
+
async def run(self):
|
| 37 |
+
while True:
|
| 38 |
+
sentance = self.input_queue.get()
|
| 39 |
+
async for chunk in self.chat_service.get_speech_chunks_async(sentance):
|
| 40 |
+
self.output_queue.put(chunk)
|
| 41 |
+
|
| 42 |
+
@ray.remote
|
| 43 |
+
class SpeechToSpeakerActor:
|
| 44 |
+
def __init__(self, input_queue, voice_id):
|
| 45 |
+
load_dotenv()
|
| 46 |
+
self.input_queue = input_queue
|
| 47 |
+
self.audio_processor = AudioStreamProcessor()
|
| 48 |
+
self.chat_service = StreamingChatService(self.audio_processor, voice_id=voice_id)
|
| 49 |
+
|
| 50 |
+
async def run(self):
|
| 51 |
+
while True:
|
| 52 |
+
audio_chunk = self.input_queue.get()
|
| 53 |
+
self.chat_service.enqueue_speech_bytes_to_play([audio_chunk])
|
| 54 |
+
|
| 55 |
+
@ray.remote
|
| 56 |
+
class RespondToPromptActor:
|
| 57 |
+
def __init__(self):
|
| 58 |
+
voice_id="2OviOUQc1JsQRQgNkVBj"
|
| 59 |
+
self.prompt_queue = Queue(maxsize=100)
|
| 60 |
+
self.llm_sentence_queue = Queue(maxsize=100)
|
| 61 |
+
self.speech_chunk_queue = Queue(maxsize=100)
|
| 62 |
+
|
| 63 |
+
self.prompt_to_llm = PromptToLLMActor.remote(self.prompt_queue, self.llm_sentence_queue, voice_id)
|
| 64 |
+
self.llm_sentence_to_speech = LLMSentanceToSpeechActor.remote(self.llm_sentence_queue, self.speech_chunk_queue, voice_id)
|
| 65 |
+
self.speech_to_speaker = SpeechToSpeakerActor.remote(self.speech_chunk_queue, voice_id)
|
| 66 |
+
|
| 67 |
+
# Start the pipeline components.
|
| 68 |
+
print ("Starting pipeline components")
|
| 69 |
+
self.prompt_to_llm.run.remote()
|
| 70 |
+
print ("prompt_to_llm running")
|
| 71 |
+
self.llm_sentence_to_speech.run.remote()
|
| 72 |
+
print ("llm_sentence_to_speech running")
|
| 73 |
+
self.speech_to_speaker.run.remote()
|
| 74 |
+
print ("speech_to_speaker running")
|
| 75 |
+
|
| 76 |
+
def enqueue_prompt(self, prompt):
|
| 77 |
+
self.prompt_queue.put(prompt)
|