Spaces:
Paused
Paused
File size: 1,756 Bytes
59cd50e 8683458 d086a4b 95a9653 d62b28a 95a9653 905de4d 59cd50e 905de4d 17a1355 4e0ef54 d086a4b c8796f6 10e7b48 b2daea4 95a9653 10e7b48 59cd50e c8796f6 10e7b48 59cd50e 10e7b48 6f6ee84 c8796f6 7132773 10e7b48 d086a4b 10e7b48 7132773 10e7b48 d62b28a 59cd50e 9569a05 c8796f6 d62b28a c8796f6 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 |
import openai
import os
import azure.cognitiveservices.speech as speechsdk
LANGUAGE = "nl"
AUDIO_FILE_NAME = "audio_response.wav"
openai.api_key = os.environ['OPEN_AI_KEY']
speech_config = speechsdk.SpeechConfig(subscription=os.environ['AZURE_SPEECH_KEY'], region="westeurope")
speech_config.speech_synthesis_voice_name = "nl-NL-ColetteNeural"
speech_synthesizer = speechsdk.SpeechSynthesizer(speech_config=speech_config, audio_config=None)
context = [{"role": "system", "content": 'Je bent een slimme en behulpzame gesprekspartner. \
Antwoord beknopt en ter zake.\
Vermeld niet dat je een AI of een soort service bent.'}]
def transcribe(model: str, audio: str):
audio_file = open(audio, "rb")
transcript = openai.Audio.transcribe(model, audio_file, language=LANGUAGE)
return transcript
def gen_response(model: str):
response = openai.ChatCompletion.create(model=model, messages=context)
return response["choices"][0]["message"]
def gen_voice(response, response_filename):
reponse_audio = speech_synthesizer.speak_text_async(response['content']).get()
stream = speechsdk.AudioDataStream(reponse_audio)
stream.save_to_wav_file(response_filename)
def respond(audio:str):
transcript = transcribe("whisper-1", audio)
context.append({"role": "user", "content": transcript['text']})
response = gen_response("gpt-3.5-turbo", context)
context.append(response)
gen_voice(response, AUDIO_FILE_NAME)
return AUDIO_FILE_NAME
def transcript():
transcript = ""
for m in context:
if m["role"] != "system":
transcript += m["role"] + " : " + m["content"] + "\n\n"
return transcript |