import gradio as gr import os import tempfile from openai import OpenAI # Set an environment variable for key os.environ['OPENAI_API_KEY'] = os.environ.get('OPENAI_API_KEY') client = OpenAI() # add api_key import torch import torchaudio import gradio as gr from scipy.io import wavfile from scipy.io.wavfile import write knn_vc = torch.hub.load('bshall/knn-vc', 'knn_vc', prematched=True, trust_repo=True, pretrained=True, device='cpu') def voice_change(audio_in, audio_ref): samplerate1, data1 = wavfile.read(audio_in) samplerate2, data2 = wavfile.read(audio_ref) write("./audio_in.wav", samplerate1, data1) write("./audio_ref.wav", samplerate2, data2) query_seq = knn_vc.get_features("./audio_in.wav") matching_set = knn_vc.get_matching_set(["./audio_ref.wav"]) out_wav = knn_vc.match(query_seq, matching_set, topk=4) torchaudio.save('output.wav', out_wav[None], 16000) return 'output.wav' def tts(text, model, voice): response = client.audio.speech.create( model=model, #"tts-1","tts-1-hd" voice=voice, #'alloy', 'echo', 'fable', 'onyx', 'nova', 'shimmer' input=text, ) # Create a temp file to save the audio with tempfile.NamedTemporaryFile(suffix=".mp3", delete=False) as temp_file: temp_file.write(response.content) # Get the file path of the temp file temp_file_path = temp_file.name return temp_file_path app = gr.Blocks() with app: gr.Markdown("#