interviewer / api /audio.py
IliaLarchenko's picture
Empty text TTS fix
b2d72b4
import io
import wave
import numpy as np
import requests
from openai import OpenAI
from utils.errors import APIError, AudioConversionError
from typing import List, Optional, Generator, Tuple
import webrtcvad
from transformers import pipeline
def detect_voice(audio: np.ndarray, sample_rate: int = 48000, frame_duration: int = 30) -> bool:
vad = webrtcvad.Vad()
vad.set_mode(3) # Aggressiveness mode: 0 (least aggressive) to 3 (most aggressive)
# Convert numpy array to 16-bit PCM bytes
audio_bytes = audio.tobytes()
num_samples_per_frame = int(sample_rate * frame_duration / 1000)
frames = [audio_bytes[i : i + num_samples_per_frame * 2] for i in range(0, len(audio_bytes), num_samples_per_frame * 2)]
count_speech = 0
for frame in frames:
if len(frame) < num_samples_per_frame * 2:
continue
if vad.is_speech(frame, sample_rate):
count_speech += 1
if count_speech > 6:
return True
return False
class STTManager:
def __init__(self, config):
self.SAMPLE_RATE = 48000
self.CHUNK_LENGTH = 5
self.STEP_LENGTH = 3
self.MAX_RELIABILITY_CUTOFF = self.CHUNK_LENGTH - 1
self.config = config
self.status = self.test_stt()
self.streaming = self.status
if config.stt.type == "HF_LOCAL":
self.pipe = pipeline("automatic-speech-recognition", model=config.stt.name)
def numpy_audio_to_bytes(self, audio_data: np.ndarray) -> bytes:
"""
Convert a numpy array of audio data to bytes.
:param audio_data: Numpy array containing audio data.
:return: Bytes representation of the audio data.
"""
num_channels = 1
sampwidth = 2
buffer = io.BytesIO()
try:
with wave.open(buffer, "wb") as wf:
wf.setnchannels(num_channels)
wf.setsampwidth(sampwidth)
wf.setframerate(self.SAMPLE_RATE)
wf.writeframes(audio_data.tobytes())
except Exception as e:
raise AudioConversionError(f"Error converting numpy array to audio bytes: {e}")
return buffer.getvalue()
def process_audio_chunk(self, audio: Tuple[int, np.ndarray], audio_buffer: np.ndarray) -> Tuple[np.ndarray, np.ndarray]:
"""
Process streamed audio data to accumulate and transcribe with overlapping segments.
:param audio: Tuple containing the sample rate and audio data as numpy array.
:param audio_buffer: Current audio buffer as numpy array.
:return: Updated current audio buffer, audio for transcription
"""
has_voice = detect_voice(audio[1])
ended = len(audio[1]) % 24000 != 0
if has_voice:
audio_buffer = np.concatenate((audio_buffer, audio[1]))
is_short = len(audio_buffer) / self.SAMPLE_RATE < 1.0
if is_short or (has_voice and not ended):
return audio_buffer, np.array([], dtype=np.int16)
return np.array([], dtype=np.int16), audio_buffer
def transcribe_audio(self, audio: np.ndarray, text: str = "") -> str:
if len(audio) < 500:
return text
else:
transcript = self.transcribe_numpy_array(audio, context=text)
return text + " " + transcript
def transcribe_numpy_array(self, audio: np.ndarray, context: Optional[str] = None) -> str:
"""
Convert speech to text from a full audio segment.
:param audio: Tuple containing the sample rate and audio data as numpy array.
:param context: Optional context for the transcription.
:return: Transcribed text.
"""
try:
if self.config.stt.type == "OPENAI_API":
audio_bytes = self.numpy_audio_to_bytes(audio)
data = ("temp.wav", audio_bytes, "audio/wav")
client = OpenAI(base_url=self.config.stt.url, api_key=self.config.stt.key)
transcription = client.audio.transcriptions.create(
model=self.config.stt.name, file=data, response_format="text", prompt=context
)
elif self.config.stt.type == "HF_API":
audio_bytes = self.numpy_audio_to_bytes(audio)
headers = {"Authorization": "Bearer " + self.config.stt.key}
response = requests.post(self.config.stt.url, headers=headers, data=audio_bytes)
if response.status_code != 200:
error_details = response.json().get("error", "No error message provided")
raise APIError("STT Error: HF API error", status_code=response.status_code, details=error_details)
transcription = response.json().get("text", None)
if transcription is None:
raise APIError("STT Error: No transcription returned by HF API")
elif self.config.stt.type == "HF_LOCAL":
result = self.pipe({"sampling_rate": self.SAMPLE_RATE, "raw": audio.astype(np.float32) / 32768.0})
transcription = result["text"]
except APIError:
raise
except Exception as e:
raise APIError(f"STT Error: Unexpected error: {e}")
return transcription
def test_stt(self) -> bool:
"""
Test if the STT service is working correctly.
:return: True if the STT service is working, False otherwise.
"""
try:
self.transcribe_audio(np.zeros(10000))
return True
except:
return False
class TTSManager:
def __init__(self, config):
self.config = config
self.status = self.test_tts(stream=False)
self.streaming = self.test_tts(stream=True) if self.status else False
def test_tts(self, stream) -> bool:
"""
Test if the TTS service is working correctly.
:return: True if the TTS service is working, False otherwise.
"""
try:
list(self.read_text("Handshake", stream=stream))
return True
except:
return False
def read_text(self, text: str, stream: Optional[bool] = None) -> Generator[bytes, None, None]:
"""
Convert text to speech and return the audio bytes, optionally streaming the response.
:param text: Text to convert to speech.
:param stream: Whether to use streaming or not.
:return: Generator yielding chunks of audio bytes.
"""
if not text:
yield b""
return
if stream is None:
stream = self.streaming
headers = {"Authorization": "Bearer " + self.config.tts.key}
data = {"model": self.config.tts.name, "input": text, "voice": "alloy", "response_format": "opus"}
try:
if not stream:
if self.config.tts.type == "OPENAI_API":
response = requests.post(self.config.tts.url + "/audio/speech", headers=headers, json=data)
elif self.config.tts.type == "HF_API":
response = requests.post(self.config.tts.url, headers=headers, json={"inputs": text})
if response.status_code != 200:
error_details = response.json().get("error", "No error message provided")
raise APIError(f"TTS Error: {self.config.tts.type} error", status_code=response.status_code, details=error_details)
yield response.content
else:
if self.config.tts.type != "OPENAI_API":
raise APIError("TTS Error: Streaming not supported for this TTS type")
with requests.post(self.config.tts.url + "/audio/speech", headers=headers, json=data, stream=True) as response:
if response.status_code != 200:
error_details = response.json().get("error", "No error message provided")
raise APIError("TTS Error: OPENAI API error", status_code=response.status_code, details=error_details)
yield from response.iter_content(chunk_size=1024)
except APIError:
raise
except Exception as e:
raise APIError(f"TTS Error: Unexpected error: {e}")
def read_last_message(self, chat_history: List[List[Optional[str]]]) -> Generator[bytes, None, None]:
"""
Read the last message in the chat history and convert it to speech.
:param chat_history: List of chat messages.
:return: Generator yielding chunks of audio bytes.
"""
if len(chat_history) > 0 and chat_history[-1][1]:
yield from self.read_text(chat_history[-1][1])