app.py

import gradio as gr



from huggingface_hub import snapshot_download
import numpy as np
from scipy.io import wavfile





model_ids = [
    'suno/bark',
]
for model_id in model_ids:
    model_name = model_id.split('/')[-1]
    snapshot_download(model_id, local_dir=f'checkpoints/{model_name}')



from TTS.tts.configs.bark_config import BarkConfig
from TTS.tts.models.bark import Bark

config = BarkConfig()
model = Bark.init_from_config(config)
model.load_checkpoint(config, checkpoint_dir="checkpoints/bark", eval=True)

def infer(prompt):

    text = "Hello, my name is Manmay , how are you?"

    # with random speaker
    #output_dict = model.synthesize(text, config, speaker_id="random", voice_dirs=None)

    # cloning a speaker.
    # It assumes that you have a speaker file in `bark_voices/speaker_n/speaker.wav` or `bark_voices/speaker_n/speaker.npz`
    output_dict = model.synthesize(text, config, speaker_id="speaker", voice_dirs="bark_voices/")
    print(output_dict)

    sample_rate = 44100  # Replace with the actual sample rate

    wavfile.write('output.wav', sample_rate, output_dict['wav'])

   
    return "output.wav"

gr.Interface(fn=infer, inputs=[gr.Textbox()], outputs=[gr.Audio()]).launch()