mrfakename
commited on
Commit
•
8468a28
1
Parent(s):
0978fba
Sync from GitHub repo
Browse filesThis Space is synced from the GitHub repo: https://github.com/SWivid/F5-TTS. Please submit contributions to the Space there
app.py
CHANGED
@@ -380,7 +380,9 @@ with gr.Blocks() as app_multistyle:
|
|
380 |
ref_text = speech_types[current_style].get("ref_text", "")
|
381 |
|
382 |
# Generate speech for this segment
|
383 |
-
audio, _ = infer(
|
|
|
|
|
384 |
sr, audio_data = audio
|
385 |
|
386 |
generated_audio_segments.append(audio_data)
|
@@ -482,12 +484,9 @@ Have a conversation with an AI using your reference voice!
|
|
482 |
chat_interface_container = gr.Column()
|
483 |
|
484 |
if chat_model_state is None:
|
485 |
-
show_info = gr.Info
|
486 |
-
show_info("Loading chat model...")
|
487 |
model_name = "Qwen/Qwen2.5-3B-Instruct"
|
488 |
chat_model_state = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
|
489 |
chat_tokenizer_state = AutoTokenizer.from_pretrained(model_name)
|
490 |
-
show_info("Chat model loaded.")
|
491 |
|
492 |
with chat_interface_container:
|
493 |
with gr.Row():
|
@@ -578,7 +577,7 @@ Have a conversation with an AI using your reference voice!
|
|
578 |
remove_silence,
|
579 |
cross_fade_duration=0.15,
|
580 |
speed=1.0,
|
581 |
-
show_info=
|
582 |
)
|
583 |
return audio_result
|
584 |
|
|
|
380 |
ref_text = speech_types[current_style].get("ref_text", "")
|
381 |
|
382 |
# Generate speech for this segment
|
383 |
+
audio, _ = infer(
|
384 |
+
ref_audio, ref_text, text, model_choice, remove_silence, 0, show_info=print
|
385 |
+
) # show_info=print no pull to top when generating
|
386 |
sr, audio_data = audio
|
387 |
|
388 |
generated_audio_segments.append(audio_data)
|
|
|
484 |
chat_interface_container = gr.Column()
|
485 |
|
486 |
if chat_model_state is None:
|
|
|
|
|
487 |
model_name = "Qwen/Qwen2.5-3B-Instruct"
|
488 |
chat_model_state = AutoModelForCausalLM.from_pretrained(model_name, torch_dtype="auto", device_map="auto")
|
489 |
chat_tokenizer_state = AutoTokenizer.from_pretrained(model_name)
|
|
|
490 |
|
491 |
with chat_interface_container:
|
492 |
with gr.Row():
|
|
|
577 |
remove_silence,
|
578 |
cross_fade_duration=0.15,
|
579 |
speed=1.0,
|
580 |
+
show_info=print, # show_info=print no pull to top when generating
|
581 |
)
|
582 |
return audio_result
|
583 |
|