Update app.py
Browse files
app.py
CHANGED
|
@@ -5,8 +5,9 @@ import gradio as gr
|
|
| 5 |
from fastapi import FastAPI, UploadFile, File, Form
|
| 6 |
from fastapi.responses import FileResponse
|
| 7 |
from TTS.api import TTS
|
|
|
|
| 8 |
|
| 9 |
-
# β
Accept license
|
| 10 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
| 11 |
|
| 12 |
# =========================
|
|
@@ -40,11 +41,11 @@ async def clone_voice_api(
|
|
| 40 |
input_path = f"{OUTPUT_DIR}/{uuid.uuid4()}_input.wav"
|
| 41 |
output_path = f"{OUTPUT_DIR}/{uuid.uuid4()}_output.wav"
|
| 42 |
|
| 43 |
-
# Save uploaded
|
| 44 |
with open(input_path, "wb") as f:
|
| 45 |
f.write(await audio.read())
|
| 46 |
|
| 47 |
-
# π₯
|
| 48 |
tts.tts_to_file(
|
| 49 |
text=text,
|
| 50 |
speaker_wav=input_path,
|
|
@@ -65,10 +66,13 @@ def clone_voice_ui(audio_path, text, language):
|
|
| 65 |
if audio_path is None:
|
| 66 |
return "β Upload audio", None
|
| 67 |
|
|
|
|
|
|
|
|
|
|
| 68 |
try:
|
| 69 |
output_path = f"{OUTPUT_DIR}/{uuid.uuid4()}.wav"
|
| 70 |
|
| 71 |
-
# π₯
|
| 72 |
tts.tts_to_file(
|
| 73 |
text=text,
|
| 74 |
speaker_wav=audio_path,
|
|
@@ -79,23 +83,35 @@ def clone_voice_ui(audio_path, text, language):
|
|
| 79 |
return "β
Done", output_path
|
| 80 |
|
| 81 |
except Exception as e:
|
| 82 |
-
return str(e), None
|
| 83 |
|
| 84 |
|
| 85 |
with gr.Blocks(title="XTTS Voice Cloning") as demo:
|
| 86 |
gr.Markdown("# π€ XTTS Voice Cloning")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 87 |
|
| 88 |
-
|
| 89 |
-
|
| 90 |
-
|
|
|
|
| 91 |
|
| 92 |
-
|
| 93 |
|
| 94 |
-
status = gr.Textbox()
|
| 95 |
-
output_audio = gr.Audio()
|
| 96 |
|
| 97 |
-
|
| 98 |
-
clone_voice_ui,
|
| 99 |
inputs=[audio_input, text_input, lang_input],
|
| 100 |
outputs=[status, output_audio]
|
| 101 |
)
|
|
@@ -104,4 +120,11 @@ with gr.Blocks(title="XTTS Voice Cloning") as demo:
|
|
| 104 |
# =========================
|
| 105 |
# COMBINE GRADIO + FASTAPI
|
| 106 |
# =========================
|
| 107 |
-
app = gr.mount_gradio_app(api, demo, path="/")
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 5 |
from fastapi import FastAPI, UploadFile, File, Form
|
| 6 |
from fastapi.responses import FileResponse
|
| 7 |
from TTS.api import TTS
|
| 8 |
+
import uvicorn
|
| 9 |
|
| 10 |
+
# β
Accept Coqui license
|
| 11 |
os.environ["COQUI_TOS_AGREED"] = "1"
|
| 12 |
|
| 13 |
# =========================
|
|
|
|
| 41 |
input_path = f"{OUTPUT_DIR}/{uuid.uuid4()}_input.wav"
|
| 42 |
output_path = f"{OUTPUT_DIR}/{uuid.uuid4()}_output.wav"
|
| 43 |
|
| 44 |
+
# Save uploaded audio
|
| 45 |
with open(input_path, "wb") as f:
|
| 46 |
f.write(await audio.read())
|
| 47 |
|
| 48 |
+
# π₯ Use same loaded model
|
| 49 |
tts.tts_to_file(
|
| 50 |
text=text,
|
| 51 |
speaker_wav=input_path,
|
|
|
|
| 66 |
if audio_path is None:
|
| 67 |
return "β Upload audio", None
|
| 68 |
|
| 69 |
+
if text.strip() == "":
|
| 70 |
+
return "β Enter text", None
|
| 71 |
+
|
| 72 |
try:
|
| 73 |
output_path = f"{OUTPUT_DIR}/{uuid.uuid4()}.wav"
|
| 74 |
|
| 75 |
+
# π₯ Use same loaded model
|
| 76 |
tts.tts_to_file(
|
| 77 |
text=text,
|
| 78 |
speaker_wav=audio_path,
|
|
|
|
| 83 |
return "β
Done", output_path
|
| 84 |
|
| 85 |
except Exception as e:
|
| 86 |
+
return f"β Error: {str(e)}", None
|
| 87 |
|
| 88 |
|
| 89 |
with gr.Blocks(title="XTTS Voice Cloning") as demo:
|
| 90 |
gr.Markdown("# π€ XTTS Voice Cloning")
|
| 91 |
+
gr.Markdown("Upload voice sample and generate cloned speech")
|
| 92 |
+
|
| 93 |
+
audio_input = gr.Audio(
|
| 94 |
+
type="filepath",
|
| 95 |
+
label="Upload / Record Speaker Audio"
|
| 96 |
+
)
|
| 97 |
+
|
| 98 |
+
text_input = gr.Textbox(
|
| 99 |
+
label="Text",
|
| 100 |
+
placeholder="Type something..."
|
| 101 |
+
)
|
| 102 |
|
| 103 |
+
lang_input = gr.Textbox(
|
| 104 |
+
label="Language Code",
|
| 105 |
+
value="en"
|
| 106 |
+
)
|
| 107 |
|
| 108 |
+
generate_btn = gr.Button("π Generate")
|
| 109 |
|
| 110 |
+
status = gr.Textbox(label="Status")
|
| 111 |
+
output_audio = gr.Audio(label="Generated Audio")
|
| 112 |
|
| 113 |
+
generate_btn.click(
|
| 114 |
+
fn=clone_voice_ui,
|
| 115 |
inputs=[audio_input, text_input, lang_input],
|
| 116 |
outputs=[status, output_audio]
|
| 117 |
)
|
|
|
|
| 120 |
# =========================
|
| 121 |
# COMBINE GRADIO + FASTAPI
|
| 122 |
# =========================
|
| 123 |
+
app = gr.mount_gradio_app(api, demo, path="/")
|
| 124 |
+
|
| 125 |
+
|
| 126 |
+
# =========================
|
| 127 |
+
# π KEEP SERVER RUNNING
|
| 128 |
+
# =========================
|
| 129 |
+
if __name__ == "__main__":
|
| 130 |
+
uvicorn.run(app, host="0.0.0.0", port=7860)
|