File size: 1,052 Bytes
6315bf2
 
 
 
 
 
 
 
e9a7ca3
6315bf2
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
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
import gradio as gr
import torch
from datasets import load_dataset
from transformers import pipeline, SpeechT5HifiGan, SpeechT5ForTextToSpeech

model_id = "Sandiago21/speecht5_finetuned_voxpopuli_it"  # update with your model id
# pipe = pipeline("automatic-speech-recognition", model=model_id)
model = SpeechT5ForTextToSpeech.from_pretrained(model_id)
vocoder = SpeechT5HifiGan.from_pretrained("microsoft/speecht5_hifigan")
embeddings_dataset = load_dataset("Matthijs/cmu-arctic-xvectors", split="validation")
speaker_embeddings = torch.tensor(embeddings_dataset[7440]["xvector"]).unsqueeze(0)

def synthesize_speech(text):
    inputs = processor(text=text, return_tensors="pt")

    speech = model.generate_speech(inputs["input_ids"], speaker_embeddings, vocoder=vocoder)
    
    return gr.Audio.update(value=(16000, speech.cpu().numpy()))

syntesize_speech_gradio = gr.Interface(
    synthesize_speech,
    inputs = gr.Textbox(label="Text", placeholder="Type something here..."),
    outputs=gr.Audio(),
#     title="Hot Dog? Or Not?",
).launch()