from transformers import SpeechT5Processor, SpeechT5ForTextToSpeech, SpeechT5HifiGan
from datasets import load_dataset
import torch
import soundfile as sf
from datasets import load_dataset
from IPython.display import Audio
import streamlit as st

st.title("TTS-App")
text = st.text_input('Donnez votre text à prédire: ', ' ')
processor = SpeechT5Processor.from_pretrained("microsoft/speecht5_tts")
model = SpeechT5ForTextToSpeech.from_pretrained("microsoft/speecht5_tts")
vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")
# read input text from file

inputs = processor(text=text, return_tensors="pt")
embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
speaker_embeddings = torch.tensor(embeddings_dataset[7306]["xvector"]).unsqueeze(0)

speech = model.generate_speech(inputs["input_ids"], speaker_embeddings, vocoder=vocoder)

#sf.write("VocalEminem.wav", speech.numpy(), samplerate=16000)
sf.write("speech.wav", speech.numpy(), samplerate=16000)
#audio_path = "VocalEminem.wav"

audio_path="speech.wav"
Audio(audio_path)
st.audio(audio_path)