Spaces:
Running
on
Zero
Running
on
Zero
Update app.py
Browse files
app.py
CHANGED
@@ -14,61 +14,37 @@ from transformers import pipeline
|
|
14 |
from infer import DMOInference
|
15 |
|
16 |
# Global variables
|
17 |
-
|
18 |
asr_pipe = None
|
19 |
-
|
20 |
|
21 |
-
#
|
22 |
-
def
|
23 |
-
"""Download models from HuggingFace Hub."""
|
24 |
-
global model_downloaded, model_paths
|
25 |
-
|
26 |
-
try:
|
27 |
-
print("Downloading models from HuggingFace...")
|
28 |
-
|
29 |
-
# Download student model
|
30 |
-
student_path = hf_hub_download(
|
31 |
-
repo_id="yl4579/DMOSpeech2",
|
32 |
-
filename="model_85000.pt",
|
33 |
-
cache_dir="./models"
|
34 |
-
)
|
35 |
-
|
36 |
-
# Download duration predictor
|
37 |
-
duration_path = hf_hub_download(
|
38 |
-
repo_id="yl4579/DMOSpeech2",
|
39 |
-
filename="model_1500.pt",
|
40 |
-
cache_dir="./models"
|
41 |
-
)
|
42 |
-
|
43 |
-
model_paths["student"] = student_path
|
44 |
-
model_paths["duration"] = duration_path
|
45 |
-
model_downloaded = True
|
46 |
-
|
47 |
-
print(f"✓ Models downloaded successfully")
|
48 |
-
return True
|
49 |
-
|
50 |
-
except Exception as e:
|
51 |
-
print(f"Error downloading models: {e}")
|
52 |
-
return False
|
53 |
-
|
54 |
-
# Initialize ASR pipeline on CPU
|
55 |
-
def initialize_asr_pipeline():
|
56 |
"""Initialize the ASR pipeline on startup."""
|
57 |
global asr_pipe
|
58 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
59 |
print("Initializing ASR pipeline...")
|
60 |
try:
|
61 |
asr_pipe = pipeline(
|
62 |
"automatic-speech-recognition",
|
63 |
model="openai/whisper-large-v3-turbo",
|
64 |
-
torch_dtype=
|
65 |
-
device="cpu" #
|
66 |
)
|
67 |
-
print("
|
68 |
-
return True
|
69 |
except Exception as e:
|
70 |
print(f"Error initializing ASR pipeline: {e}")
|
71 |
-
|
72 |
|
73 |
# Transcribe function
|
74 |
def transcribe(ref_audio, language=None):
|
@@ -76,7 +52,7 @@ def transcribe(ref_audio, language=None):
|
|
76 |
global asr_pipe
|
77 |
|
78 |
if asr_pipe is None:
|
79 |
-
return ""
|
80 |
|
81 |
try:
|
82 |
result = asr_pipe(
|
@@ -91,14 +67,65 @@ def transcribe(ref_audio, language=None):
|
|
91 |
print(f"Transcription error: {e}")
|
92 |
return ""
|
93 |
|
94 |
-
|
95 |
-
|
96 |
-
|
97 |
-
|
98 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
99 |
|
100 |
-
|
101 |
-
|
|
|
|
|
|
|
|
|
|
|
102 |
prompt_audio,
|
103 |
prompt_text,
|
104 |
target_text,
|
@@ -109,72 +136,53 @@ def generate_speech_gpu(
|
|
109 |
custom_student_start_step,
|
110 |
verbose
|
111 |
):
|
112 |
-
"""Generate speech with
|
113 |
|
114 |
-
if not
|
115 |
-
return None, "
|
116 |
|
117 |
if prompt_audio is None:
|
118 |
-
return None, "
|
119 |
|
120 |
if not target_text:
|
121 |
-
return None, "
|
122 |
|
123 |
try:
|
124 |
-
#
|
125 |
-
|
126 |
-
print(f"Initializing model on {device}...")
|
127 |
-
|
128 |
-
model = DMOInference(
|
129 |
-
student_checkpoint_path=model_paths["student"],
|
130 |
-
duration_predictor_path=model_paths["duration"],
|
131 |
-
device=device,
|
132 |
-
model_type="F5TTS_Base"
|
133 |
-
)
|
134 |
-
|
135 |
-
# Auto-transcribe if needed (this happens on CPU)
|
136 |
-
transcribed_text = prompt_text # Default to provided text
|
137 |
-
if not prompt_text.strip():
|
138 |
print("Auto-transcribing reference audio...")
|
139 |
-
|
140 |
-
print(f"Transcribed: {
|
141 |
|
142 |
start_time = time.time()
|
143 |
|
144 |
# Configure parameters based on mode
|
145 |
-
|
146 |
-
|
147 |
-
|
148 |
-
|
149 |
-
|
150 |
-
|
151 |
-
|
152 |
-
|
153 |
-
|
154 |
-
|
155 |
-
|
156 |
-
|
157 |
-
|
158 |
-
|
159 |
-
|
160 |
-
|
161 |
-
|
162 |
-
"teacher_steps": custom_teacher_steps,
|
163 |
-
"teacher_stopping_time": custom_teacher_stopping_time,
|
164 |
-
"student_start_step": custom_student_start_step
|
165 |
-
}
|
166 |
-
}
|
167 |
-
|
168 |
-
config = configs[mode]
|
169 |
|
170 |
# Generate speech
|
171 |
generated_audio = model.generate(
|
172 |
gen_text=target_text,
|
173 |
audio_path=prompt_audio,
|
174 |
-
prompt_text=
|
175 |
-
teacher_steps=
|
176 |
-
teacher_stopping_time=
|
177 |
-
student_start_step=
|
178 |
temperature=temperature,
|
179 |
verbose=verbose
|
180 |
)
|
@@ -198,50 +206,29 @@ def generate_speech_gpu(
|
|
198 |
|
199 |
torchaudio.save(output_path, generated_audio, 24000)
|
200 |
|
201 |
-
# Format
|
202 |
-
metrics = f"
|
203 |
-
Processing: {processing_time:.2f}s for {audio_duration:.2f}s audio
|
204 |
-
Device: {device.upper()}"""
|
205 |
|
206 |
-
|
207 |
-
if not prompt_text.strip():
|
208 |
-
info += f" | Auto-transcribed"
|
209 |
-
|
210 |
-
# Clean up GPU memory
|
211 |
-
del model
|
212 |
-
if device == "cuda":
|
213 |
-
torch.cuda.empty_cache()
|
214 |
-
|
215 |
-
# Return transcribed text to update the textbox
|
216 |
-
return output_path, "✅ Success!", metrics, info, transcribed_text
|
217 |
|
218 |
except Exception as e:
|
219 |
-
|
220 |
-
print(traceback.format_exc())
|
221 |
-
return None, f"❌ Error: {str(e)}", "", "", prompt_text
|
222 |
|
223 |
# Create Gradio interface
|
224 |
-
with gr.Blocks(
|
225 |
-
title="DMOSpeech 2 - Zero-Shot TTS",
|
226 |
-
theme=gr.themes.Soft(),
|
227 |
-
css="""
|
228 |
-
.gradio-container { max-width: 1200px !important; }
|
229 |
-
"""
|
230 |
-
) as demo:
|
231 |
-
|
232 |
gr.Markdown(f"""
|
233 |
-
|
234 |
-
|
235 |
-
|
236 |
-
|
237 |
-
|
238 |
""")
|
239 |
|
240 |
with gr.Row():
|
241 |
with gr.Column(scale=1):
|
242 |
-
#
|
243 |
prompt_audio = gr.Audio(
|
244 |
-
label="📎 Reference Audio
|
245 |
type="filepath",
|
246 |
sources=["upload", "microphone"]
|
247 |
)
|
@@ -258,6 +245,7 @@ with gr.Blocks(
|
|
258 |
lines=4
|
259 |
)
|
260 |
|
|
|
261 |
mode = gr.Radio(
|
262 |
choices=[
|
263 |
"Student Only (4 steps)",
|
@@ -267,10 +255,10 @@ with gr.Blocks(
|
|
267 |
],
|
268 |
value="Teacher-Guided (8 steps)",
|
269 |
label="🚀 Generation Mode",
|
270 |
-
info="
|
271 |
)
|
272 |
|
273 |
-
# Advanced settings
|
274 |
with gr.Accordion("⚙️ Advanced Settings", open=False):
|
275 |
temperature = gr.Slider(
|
276 |
minimum=0.0,
|
@@ -278,76 +266,115 @@ with gr.Blocks(
|
|
278 |
value=0.0,
|
279 |
step=0.1,
|
280 |
label="Duration Temperature",
|
281 |
-
info="0 =
|
282 |
)
|
283 |
|
284 |
-
with gr.Group(visible=False) as
|
285 |
-
|
286 |
-
|
287 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
288 |
|
289 |
-
verbose = gr.Checkbox(
|
|
|
|
|
|
|
|
|
290 |
|
291 |
generate_btn = gr.Button("🎵 Generate Speech", variant="primary", size="lg")
|
292 |
|
293 |
with gr.Column(scale=1):
|
294 |
-
#
|
295 |
output_audio = gr.Audio(
|
296 |
label="🔊 Generated Speech",
|
297 |
type="filepath",
|
298 |
autoplay=True
|
299 |
)
|
300 |
|
301 |
-
status = gr.Textbox(
|
302 |
-
|
303 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
304 |
|
305 |
-
#
|
306 |
gr.Markdown("""
|
307 |
-
### 💡 Quick
|
308 |
|
309 |
-
|
310 |
-
|
311 |
-
|
312 |
-
|
313 |
-
|
314 |
|
315 |
-
|
316 |
-
-
|
317 |
-
-
|
318 |
-
-
|
319 |
-
- Custom mode lets you fine-tune all parameters
|
320 |
""")
|
321 |
|
322 |
-
# Examples
|
323 |
-
gr.Markdown("### 🎯 Example
|
324 |
|
325 |
gr.Markdown("""
|
326 |
<details>
|
327 |
<summary>English Example</summary>
|
328 |
|
329 |
-
**Reference:** "Some call me nature, others call me mother nature."
|
330 |
|
331 |
-
**Target:** "I don't really care what you call me. I've been a silent spectator, watching species evolve, empires rise and fall. But always remember, I am mighty and enduring."
|
332 |
</details>
|
333 |
|
334 |
<details>
|
335 |
<summary>Chinese Example</summary>
|
336 |
|
337 |
-
**Reference:** "对,这就是我,万人敬仰的太乙真人。"
|
338 |
|
339 |
-
**Target:** "突然,身边一阵笑声。我看着他们,意气风发地挺直了胸膛,甩了甩那稍显肉感的双臂,轻笑道:'我身上的肉,是为了掩饰我爆棚的魅力,否则,岂不吓坏了你们呢?'"
|
340 |
</details>
|
341 |
-
""")
|
342 |
|
343 |
-
|
344 |
-
|
345 |
-
return gr.update(visible=(mode == "Custom"))
|
346 |
|
347 |
-
|
|
|
|
|
348 |
|
|
|
349 |
generate_btn.click(
|
350 |
-
|
351 |
inputs=[
|
352 |
prompt_audio,
|
353 |
prompt_text,
|
@@ -359,15 +386,25 @@ with gr.Blocks(
|
|
359 |
custom_student_start_step,
|
360 |
verbose
|
361 |
],
|
362 |
-
outputs=[
|
363 |
-
|
364 |
-
|
365 |
-
|
366 |
-
|
367 |
-
|
368 |
-
|
|
|
|
|
|
|
|
|
|
|
369 |
)
|
370 |
|
371 |
-
# Launch
|
372 |
if __name__ == "__main__":
|
|
|
|
|
|
|
|
|
|
|
373 |
demo.launch()
|
|
|
14 |
from infer import DMOInference
|
15 |
|
16 |
# Global variables
|
17 |
+
model = None
|
18 |
asr_pipe = None
|
19 |
+
device = "cuda" if torch.cuda.is_available() else "cpu"
|
20 |
|
21 |
+
# Initialize ASR pipeline
|
22 |
+
def initialize_asr_pipeline(device=device, dtype=None):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
23 |
"""Initialize the ASR pipeline on startup."""
|
24 |
global asr_pipe
|
25 |
|
26 |
+
if dtype is None:
|
27 |
+
dtype = (
|
28 |
+
torch.float16
|
29 |
+
if "cuda" in device
|
30 |
+
and torch.cuda.is_available()
|
31 |
+
and torch.cuda.get_device_properties(device).major >= 7
|
32 |
+
and not torch.cuda.get_device_name().endswith("[ZLUDA]")
|
33 |
+
else torch.float32
|
34 |
+
)
|
35 |
+
|
36 |
print("Initializing ASR pipeline...")
|
37 |
try:
|
38 |
asr_pipe = pipeline(
|
39 |
"automatic-speech-recognition",
|
40 |
model="openai/whisper-large-v3-turbo",
|
41 |
+
torch_dtype=dtype,
|
42 |
+
device="cpu" # Keep ASR on CPU to save GPU memory
|
43 |
)
|
44 |
+
print("ASR pipeline initialized successfully")
|
|
|
45 |
except Exception as e:
|
46 |
print(f"Error initializing ASR pipeline: {e}")
|
47 |
+
asr_pipe = None
|
48 |
|
49 |
# Transcribe function
|
50 |
def transcribe(ref_audio, language=None):
|
|
|
52 |
global asr_pipe
|
53 |
|
54 |
if asr_pipe is None:
|
55 |
+
return "" # Return empty string if ASR is not available
|
56 |
|
57 |
try:
|
58 |
result = asr_pipe(
|
|
|
67 |
print(f"Transcription error: {e}")
|
68 |
return ""
|
69 |
|
70 |
+
def download_models():
|
71 |
+
"""Download models from HuggingFace Hub."""
|
72 |
+
try:
|
73 |
+
print("Downloading models from HuggingFace...")
|
74 |
+
|
75 |
+
# Download student model
|
76 |
+
student_path = hf_hub_download(
|
77 |
+
repo_id="yl4579/DMOSpeech2",
|
78 |
+
filename="model_85000.pt",
|
79 |
+
cache_dir="./models"
|
80 |
+
)
|
81 |
+
|
82 |
+
# Download duration predictor
|
83 |
+
duration_path = hf_hub_download(
|
84 |
+
repo_id="yl4579/DMOSpeech2",
|
85 |
+
filename="model_1500.pt",
|
86 |
+
cache_dir="./models"
|
87 |
+
)
|
88 |
+
|
89 |
+
print(f"Student model: {student_path}")
|
90 |
+
print(f"Duration model: {duration_path}")
|
91 |
+
|
92 |
+
return student_path, duration_path
|
93 |
+
|
94 |
+
except Exception as e:
|
95 |
+
print(f"Error downloading models: {e}")
|
96 |
+
return None, None
|
97 |
+
|
98 |
+
def initialize_model():
|
99 |
+
"""Initialize the model on startup."""
|
100 |
+
global model
|
101 |
+
|
102 |
+
try:
|
103 |
+
# Download models
|
104 |
+
student_path, duration_path = download_models()
|
105 |
+
|
106 |
+
if not student_path or not duration_path:
|
107 |
+
return False, "Failed to download models from HuggingFace"
|
108 |
+
|
109 |
+
# Initialize model
|
110 |
+
model = DMOInference(
|
111 |
+
student_checkpoint_path=student_path,
|
112 |
+
duration_predictor_path=duration_path,
|
113 |
+
device=device,
|
114 |
+
model_type="F5TTS_Base"
|
115 |
+
)
|
116 |
+
|
117 |
+
return True, f"Model loaded successfully on {device.upper()}"
|
118 |
+
|
119 |
+
except Exception as e:
|
120 |
+
return False, f"Error initializing model: {str(e)}"
|
121 |
|
122 |
+
# Initialize models on startup
|
123 |
+
print("Initializing models...")
|
124 |
+
model_loaded, status_message = initialize_model()
|
125 |
+
initialize_asr_pipeline() # Initialize ASR pipeline
|
126 |
+
|
127 |
+
@spaces.GPU(duration=120) # Request GPU for up to 120 seconds
|
128 |
+
def generate_speech(
|
129 |
prompt_audio,
|
130 |
prompt_text,
|
131 |
target_text,
|
|
|
136 |
custom_student_start_step,
|
137 |
verbose
|
138 |
):
|
139 |
+
"""Generate speech with different configurations."""
|
140 |
|
141 |
+
if not model_loaded or model is None:
|
142 |
+
return None, "Model not loaded! Please refresh the page.", "", ""
|
143 |
|
144 |
if prompt_audio is None:
|
145 |
+
return None, "Please upload a reference audio!", "", ""
|
146 |
|
147 |
if not target_text:
|
148 |
+
return None, "Please enter text to generate!", "", ""
|
149 |
|
150 |
try:
|
151 |
+
# Auto-transcribe if prompt_text is empty
|
152 |
+
if not prompt_text and prompt_text != "":
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
153 |
print("Auto-transcribing reference audio...")
|
154 |
+
prompt_text = transcribe(prompt_audio)
|
155 |
+
print(f"Transcribed: {prompt_text}")
|
156 |
|
157 |
start_time = time.time()
|
158 |
|
159 |
# Configure parameters based on mode
|
160 |
+
if mode == "Student Only (4 steps)":
|
161 |
+
teacher_steps = 0
|
162 |
+
student_start_step = 0
|
163 |
+
teacher_stopping_time = 1.0
|
164 |
+
elif mode == "Teacher-Guided (8 steps)":
|
165 |
+
# Default configuration from the notebook
|
166 |
+
teacher_steps = 16
|
167 |
+
teacher_stopping_time = 0.07
|
168 |
+
student_start_step = 1
|
169 |
+
elif mode == "High Diversity (16 steps)":
|
170 |
+
teacher_steps = 24
|
171 |
+
teacher_stopping_time = 0.3
|
172 |
+
student_start_step = 2
|
173 |
+
else: # Custom
|
174 |
+
teacher_steps = custom_teacher_steps
|
175 |
+
teacher_stopping_time = custom_teacher_stopping_time
|
176 |
+
student_start_step = custom_student_start_step
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
177 |
|
178 |
# Generate speech
|
179 |
generated_audio = model.generate(
|
180 |
gen_text=target_text,
|
181 |
audio_path=prompt_audio,
|
182 |
+
prompt_text=prompt_text if prompt_text else None,
|
183 |
+
teacher_steps=teacher_steps,
|
184 |
+
teacher_stopping_time=teacher_stopping_time,
|
185 |
+
student_start_step=student_start_step,
|
186 |
temperature=temperature,
|
187 |
verbose=verbose
|
188 |
)
|
|
|
206 |
|
207 |
torchaudio.save(output_path, generated_audio, 24000)
|
208 |
|
209 |
+
# Format metrics
|
210 |
+
metrics = f"RTF: {rtf:.2f}x ({1/rtf:.2f}x speed) | Processing: {processing_time:.2f}s for {audio_duration:.2f}s audio"
|
|
|
|
|
211 |
|
212 |
+
return output_path, "Success!", metrics, f"Mode: {mode} | Transcribed: {prompt_text[:50]}..." if not prompt_text else f"Mode: {mode}"
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
213 |
|
214 |
except Exception as e:
|
215 |
+
return None, f"Error: {str(e)}", "", ""
|
|
|
|
|
216 |
|
217 |
# Create Gradio interface
|
218 |
+
with gr.Blocks(title="DMOSpeech 2 - Zero-Shot TTS", theme=gr.themes.Soft()) as demo:
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
219 |
gr.Markdown(f"""
|
220 |
+
# 🎙️ DMOSpeech 2: Zero-Shot Text-to-Speech
|
221 |
+
|
222 |
+
Generate natural speech in any voice with just a short reference audio!
|
223 |
+
|
224 |
+
**Model Status:** {status_message} | **Device:** {device.upper()} | **ASR:** {"✅ Ready" if asr_pipe else "❌ Not available"}
|
225 |
""")
|
226 |
|
227 |
with gr.Row():
|
228 |
with gr.Column(scale=1):
|
229 |
+
# Reference audio input
|
230 |
prompt_audio = gr.Audio(
|
231 |
+
label="📎 Reference Audio",
|
232 |
type="filepath",
|
233 |
sources=["upload", "microphone"]
|
234 |
)
|
|
|
245 |
lines=4
|
246 |
)
|
247 |
|
248 |
+
# Generation mode
|
249 |
mode = gr.Radio(
|
250 |
choices=[
|
251 |
"Student Only (4 steps)",
|
|
|
255 |
],
|
256 |
value="Teacher-Guided (8 steps)",
|
257 |
label="🚀 Generation Mode",
|
258 |
+
info="Choose speed vs quality/diversity tradeoff"
|
259 |
)
|
260 |
|
261 |
+
# Advanced settings (collapsible)
|
262 |
with gr.Accordion("⚙️ Advanced Settings", open=False):
|
263 |
temperature = gr.Slider(
|
264 |
minimum=0.0,
|
|
|
266 |
value=0.0,
|
267 |
step=0.1,
|
268 |
label="Duration Temperature",
|
269 |
+
info="0 = deterministic, >0 = more variation in speech rhythm"
|
270 |
)
|
271 |
|
272 |
+
with gr.Group(visible=False) as custom_settings:
|
273 |
+
gr.Markdown("### Custom Mode Settings")
|
274 |
+
custom_teacher_steps = gr.Slider(
|
275 |
+
minimum=0,
|
276 |
+
maximum=32,
|
277 |
+
value=16,
|
278 |
+
step=1,
|
279 |
+
label="Teacher Steps",
|
280 |
+
info="More steps = higher quality"
|
281 |
+
)
|
282 |
+
|
283 |
+
custom_teacher_stopping_time = gr.Slider(
|
284 |
+
minimum=0.0,
|
285 |
+
maximum=1.0,
|
286 |
+
value=0.07,
|
287 |
+
step=0.01,
|
288 |
+
label="Teacher Stopping Time",
|
289 |
+
info="When to switch to student"
|
290 |
+
)
|
291 |
+
|
292 |
+
custom_student_start_step = gr.Slider(
|
293 |
+
minimum=0,
|
294 |
+
maximum=4,
|
295 |
+
value=1,
|
296 |
+
step=1,
|
297 |
+
label="Student Start Step",
|
298 |
+
info="Which student step to start from"
|
299 |
+
)
|
300 |
|
301 |
+
verbose = gr.Checkbox(
|
302 |
+
value=False,
|
303 |
+
label="Verbose Output",
|
304 |
+
info="Show detailed generation steps"
|
305 |
+
)
|
306 |
|
307 |
generate_btn = gr.Button("🎵 Generate Speech", variant="primary", size="lg")
|
308 |
|
309 |
with gr.Column(scale=1):
|
310 |
+
# Output
|
311 |
output_audio = gr.Audio(
|
312 |
label="🔊 Generated Speech",
|
313 |
type="filepath",
|
314 |
autoplay=True
|
315 |
)
|
316 |
|
317 |
+
status = gr.Textbox(
|
318 |
+
label="Status",
|
319 |
+
interactive=False
|
320 |
+
)
|
321 |
+
|
322 |
+
metrics = gr.Textbox(
|
323 |
+
label="Performance Metrics",
|
324 |
+
interactive=False
|
325 |
+
)
|
326 |
+
|
327 |
+
info = gr.Textbox(
|
328 |
+
label="Generation Info",
|
329 |
+
interactive=False
|
330 |
+
)
|
331 |
|
332 |
+
# Tips
|
333 |
gr.Markdown("""
|
334 |
+
### 💡 Quick Tips:
|
335 |
|
336 |
+
- **Auto-transcription**: Leave reference text empty to auto-transcribe
|
337 |
+
- **Student Only**: Fastest (4 steps), good quality
|
338 |
+
- **Teacher-Guided**: Best balance (8 steps), recommended
|
339 |
+
- **High Diversity**: More natural prosody (16 steps)
|
340 |
+
- **Custom Mode**: Fine-tune all parameters
|
341 |
|
342 |
+
### 📊 Expected RTF (Real-Time Factor):
|
343 |
+
- Student Only: ~0.05x (20x faster than real-time)
|
344 |
+
- Teacher-Guided: ~0.10x (10x faster)
|
345 |
+
- High Diversity: ~0.20x (5x faster)
|
|
|
346 |
""")
|
347 |
|
348 |
+
# Examples section
|
349 |
+
gr.Markdown("### 🎯 Example Configurations")
|
350 |
|
351 |
gr.Markdown("""
|
352 |
<details>
|
353 |
<summary>English Example</summary>
|
354 |
|
355 |
+
**Reference text:** "Some call me nature, others call me mother nature."
|
356 |
|
357 |
+
**Target text:** "I don't really care what you call me. I've been a silent spectator, watching species evolve, empires rise and fall. But always remember, I am mighty and enduring."
|
358 |
</details>
|
359 |
|
360 |
<details>
|
361 |
<summary>Chinese Example</summary>
|
362 |
|
363 |
+
**Reference text:** "对,这就是我,万人敬仰的太乙真人。"
|
364 |
|
365 |
+
**Target text:** "突然,身边一阵笑声。我看着他们,意气风发地挺直了胸膛,甩了甩那稍显肉感的双臂,轻笑道:'我身上的肉,是为了掩饰我爆棚的魅力,否则,岂不吓坏了你们呢?'"
|
366 |
</details>
|
|
|
367 |
|
368 |
+
<details>
|
369 |
+
<summary>High Diversity Chinese Example</summary>
|
|
|
370 |
|
371 |
+
Same as above but with **Temperature: 0.8** for more natural variation in speech rhythm.
|
372 |
+
</details>
|
373 |
+
""")
|
374 |
|
375 |
+
# Event handler
|
376 |
generate_btn.click(
|
377 |
+
generate_speech,
|
378 |
inputs=[
|
379 |
prompt_audio,
|
380 |
prompt_text,
|
|
|
386 |
custom_student_start_step,
|
387 |
verbose
|
388 |
],
|
389 |
+
outputs=[output_audio, status, metrics, info]
|
390 |
+
)
|
391 |
+
|
392 |
+
# Update visibility of custom settings based on mode
|
393 |
+
def update_custom_visibility(mode):
|
394 |
+
is_custom = (mode == "Custom")
|
395 |
+
return gr.update(visible=is_custom)
|
396 |
+
|
397 |
+
mode.change(
|
398 |
+
update_custom_visibility,
|
399 |
+
inputs=[mode],
|
400 |
+
outputs=[custom_settings]
|
401 |
)
|
402 |
|
403 |
+
# Launch the app
|
404 |
if __name__ == "__main__":
|
405 |
+
if not model_loaded:
|
406 |
+
print(f"Warning: Model failed to load - {status_message}")
|
407 |
+
if not asr_pipe:
|
408 |
+
print("Warning: ASR pipeline not available - auto-transcription disabled")
|
409 |
+
|
410 |
demo.launch()
|