|
import gradio as gr |
|
|
|
import torch |
|
from transformers import BarkModel |
|
from optimum.bettertransformer import BetterTransformer |
|
|
|
bark_model = BarkModel.from_pretrained("suno/bark", torch_dtype=torch.float16) |
|
|
|
|
|
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=bark_model, eval=True) |
|
|
|
def infer(prompt): |
|
|
|
text = "Hello, my name is Manmay , how are you?" |
|
|
|
|
|
output_dict = model.synthesize(text, config, speaker_id="random", voice_dirs=None) |
|
|
|
return "done" |
|
|
|
gr.Interface(fn=infer, inputs=[gr.Textbox()], outputs=[gr.Textbox()]).launch() |
