Spaces:
Runtime error
Runtime error
File size: 2,128 Bytes
321bf66 eb4b6d0 087a581 da3a952 087a581 33c26b8 087a581 eb4b6d0 087a581 cb4e037 087a581 eb4b6d0 087a581 eb4b6d0 087a581 eb4b6d0 087a581 eb4b6d0 5a92355 eb4b6d0 5a92355 087a581 5a92355 6a72a1f 087a581 5a92355 |
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 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 |
import gradio as gr
import torch
from datasets import load_dataset
from transformers import pipeline, SpeechT5Processor, SpeechT5HifiGan, SpeechT5ForTextToSpeech
model_id = "gitgato/tts-model-v2" # update with your model id
pipe = pipeline("text-to-speech", model="gitgato/tts-model-v2")
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)
# checkpoint = "microsoft/speecht5_tts"
processor = SpeechT5Processor.from_pretrained(model_id)
replacements = [
("à", "a"),
("â", "a"),
("ç", "c"),
("è", "e"),
("ë", "e"),
("î", "i"),
("ï", "i"),
("ô", "o"),
("ù", "u"),
("û", "u"),
("ü", "u"),
]
title = "Text-to-Speech"
description = """
Demo for text-to-speech translation in French. Demo uses [gitgato/trtts](https://huggingface.co) checkpoint, which is based on Microsoft's
[SpeechT5 TTS](https://huggingface.co/microsoft/speecht5_tts) model and is fine-tuned in French Audio dataset
"""
def cleanup_text(text):
for src, dst in replacements:
text = text.replace(src, dst)
return text
def synthesize_speech(text):
text = cleanup_text(text)
inputs = processor(text=text, return_tensors="pt")
speech = model.generate_speech(inputs["input_ids"], speaker_embeddings, vocoder=vocoder)
# Crear el objeto gr.Audio directamente con los parámetros necesarios
return gr.Audio((16000, speech.cpu().numpy()))
# Definir la interfaz de Gradio
syntesize_speech_gradio = gr.Interface(
synthesize_speech,
inputs=gr.Textbox(label="Text", placeholder="Type something here..."),
outputs=gr.Audio(label="Speech"),
examples=["Hola, probando audio."],
title=title,
description=description,
)
# Lanzar la interfaz
syntesize_speech_gradio.launch()
|