Spaces:
Running
Running
File size: 3,045 Bytes
88a0625 78e02b1 c22a65c 88a0625 78e02b1 c22a65c 88a0625 78e02b1 88a0625 78e02b1 88a0625 78e02b1 c22a65c 88a0625 78e02b1 |
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 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 |
import gradio as gr
import numpy as np
from kittentts import KittenTTS
# 1. Initialize the KittenTTS model.
# This will download the model from Hugging Face on the first run.
print("Loading KittenTTS model...")
try:
tts_model = KittenTTS("KittenML/kitten-tts-nano-0.1")
print("Model loaded successfully.")
except Exception as e:
print(f"Error loading model: {e}")
# You might want to handle this more gracefully
exit()
# 2. Get the list of available voices directly from the model instance.
AVAILABLE_VOICES = tts_model.available_voices
DEFAULT_VOICE = "expr-voice-5-m" if "expr-voice-5-m" in AVAILABLE_VOICES else AVAILABLE_VOICES[0]
# 3. Define the core function that Gradio will call.
# This function now accepts 'voice' and 'speed' as arguments.
def synthesize_speech(text, voice, speed):
"""
Generates audio using the selected text, voice, and speed.
"""
# Handle empty input gracefully
if not text.strip():
# Return a silent, empty audio clip
return (24000, np.zeros(0, dtype=np.int16))
# Call the model's generate method with all the parameters
audio_data = tts_model.generate(text, voice=voice, speed=speed)
# Return the audio in the format Gradio expects: (sample_rate, numpy_array)
return (24000, audio_data)
# 4. Create the Gradio UI with the new controls.
with gr.Blocks(theme=gr.themes.Soft()) as demo:
gr.Markdown(
"""
# 🐱 Enhanced KittenTTS UI
A user-friendly interface for the KittenTTS text-to-speech model.
Select a voice, adjust the speed, and type your text to generate audio.
"""
)
with gr.Row():
with gr.Column(scale=3):
text_input = gr.Textbox(
lines=5,
label="Input Text",
placeholder="Type something here..."
)
with gr.Column(scale=1):
voice_dropdown = gr.Dropdown(
choices=AVAILABLE_VOICES,
value=DEFAULT_VOICE,
label="Voice Selection"
)
speed_slider = gr.Slider(
minimum=0.5,
maximum=2.0,
step=0.1,
value=1.0,
label="Speech Speed"
)
generate_button = gr.Button("Generate Audio", variant="primary")
audio_output = gr.Audio(label="Generated Speech", autoplay=True)
gr.Examples(
examples=[
["This is an example of a female voice.", "expr-voice-5-f", 1.0],
["This is an example of a male voice, speaking a bit faster.", "expr-voice-5-m", 1.2],
["The speed can also be slowed down for clarity.", "expr-voice-4-f", 0.8],
],
inputs=[text_input, voice_dropdown, speed_slider]
)
# Connect the UI components to the function
generate_button.click(
fn=synthesize_speech,
inputs=[text_input, voice_dropdown, speed_slider],
outputs=audio_output
)
# Launch the Gradio app
demo.launch() |