|
|
import os |
|
|
import torch |
|
|
import torchaudio |
|
|
import gradio as gr |
|
|
from src.chatterbox.mtl_tts import ChatterboxMultilingualTTS |
|
|
from safetensors.torch import load_file as load_safetensors |
|
|
from huggingface_hub import hf_hub_download |
|
|
|
|
|
|
|
|
|
|
|
device = "cpu" |
|
|
my_token = os.getenv("HF_TOKEN") |
|
|
|
|
|
print("--- Starting Application on CPU ---") |
|
|
|
|
|
print("Loading model structure...") |
|
|
multilingual_model = ChatterboxMultilingualTTS.from_pretrained(device=device) |
|
|
|
|
|
print("Downloading weights...") |
|
|
try: |
|
|
model_path = hf_hub_download( |
|
|
repo_id="Thomcles/Chatterbox-TTS-Persian-Farsi", |
|
|
filename="t3_fa.safetensors", |
|
|
token=my_token |
|
|
) |
|
|
print("Weights downloaded.") |
|
|
except Exception as e: |
|
|
print(f"Download Error: {e}") |
|
|
raise e |
|
|
|
|
|
print("Loading weights into model...") |
|
|
|
|
|
t3_state = load_safetensors(model_path, device="cpu") |
|
|
multilingual_model.t3.load_state_dict(t3_state) |
|
|
multilingual_model.t3.to(device).eval() |
|
|
print("Model ready!") |
|
|
|
|
|
|
|
|
def generate_audio(text): |
|
|
if not text: |
|
|
return None |
|
|
|
|
|
|
|
|
if len(text) > 300: |
|
|
text = text[:300] |
|
|
|
|
|
try: |
|
|
|
|
|
with torch.no_grad(): |
|
|
wav_tensor = multilingual_model.generate(text, language_id=None) |
|
|
|
|
|
|
|
|
audio_numpy = wav_tensor.squeeze().cpu().numpy() |
|
|
|
|
|
return (multilingual_model.sr, audio_numpy) |
|
|
|
|
|
except Exception as e: |
|
|
print(f"Error: {e}") |
|
|
raise gr.Error(f"خطا در تولید صدا: {e}") |
|
|
|
|
|
|
|
|
with gr.Blocks(title="مبدل متن به گفتار فارسی Chatterbox", theme=gr.themes.Soft()) as demo: |
|
|
gr.Markdown( |
|
|
""" |
|
|
# مبدل متن به گفتار فارسی |
|
|
(CPU) |
|
|
""" |
|
|
) |
|
|
|
|
|
with gr.Row(): |
|
|
with gr.Column(): |
|
|
text_input = gr.TextArea( |
|
|
label="متن فارسی را وارد کنید", |
|
|
value="سلام! این یک تست روی پردازنده معمولی است.", |
|
|
lines=3, |
|
|
rtl=True, |
|
|
max_length=300 |
|
|
) |
|
|
submit_btn = gr.Button("🎧 تولید صدا", variant="primary") |
|
|
|
|
|
with gr.Column(): |
|
|
audio_output = gr.Audio(label="پخش صدا", type="numpy") |
|
|
|
|
|
submit_btn.click( |
|
|
fn=generate_audio, |
|
|
inputs=text_input, |
|
|
outputs=audio_output |
|
|
) |
|
|
|
|
|
demo.launch() |
|
|
|
