Spaces:
Sleeping
Sleeping
UI update
Browse files
app.py
CHANGED
|
@@ -527,25 +527,15 @@ with gr.Blocks(
|
|
| 527 |
(https://github.com/siddhu001/espnet/tree/sds_demo_recipe/egs2/TEMPLATE/sds1#how-to-use).
|
| 528 |
"""
|
| 529 |
)
|
|
|
|
| 530 |
with gr.Row():
|
| 531 |
with gr.Column(scale=1):
|
| 532 |
-
user_audio = gr.Audio(
|
| 533 |
-
sources=["upload", "microphone"],
|
| 534 |
-
type="numpy",
|
| 535 |
-
label="Upload or Record Audio File",
|
| 536 |
-
format="wav",
|
| 537 |
-
)
|
| 538 |
input_text = gr.Textbox(
|
| 539 |
label="LLM prompt",
|
| 540 |
visible=True,
|
| 541 |
interactive=True,
|
| 542 |
value=default_instruct,
|
| 543 |
)
|
| 544 |
-
type_radio = gr.Radio(
|
| 545 |
-
choices=["Cascaded"],
|
| 546 |
-
label="Choose type of Spoken Dialog:",
|
| 547 |
-
value="Cascaded",
|
| 548 |
-
)
|
| 549 |
ASR_radio = gr.Radio(
|
| 550 |
choices=ASR_options,
|
| 551 |
label="Choose ASR:",
|
|
@@ -569,6 +559,12 @@ with gr.Blocks(
|
|
| 569 |
)
|
| 570 |
process_btn = gr.Button("Process Audio", variant="primary")
|
| 571 |
with gr.Column(scale=1):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 572 |
output_asr_text = gr.Textbox(label="ASR Transcription", interactive=False)
|
| 573 |
output_text = gr.Textbox(label="LLM Response", interactive=False)
|
| 574 |
output_audio = gr.Audio(label="TTS Output", autoplay=True, visible=True, interactive=False)
|
|
@@ -627,21 +623,6 @@ with gr.Blocks(
|
|
| 627 |
fn=handle_ASR_selection_wrapper,
|
| 628 |
inputs=[ASR_radio],
|
| 629 |
)
|
| 630 |
-
type_radio.change(
|
| 631 |
-
fn=dialogue_model.handle_type_selection,
|
| 632 |
-
inputs=[type_radio, radio, ASR_radio, LLM_radio],
|
| 633 |
-
outputs=[
|
| 634 |
-
radio,
|
| 635 |
-
ASR_radio,
|
| 636 |
-
LLM_radio,
|
| 637 |
-
E2Eradio,
|
| 638 |
-
output_asr_text,
|
| 639 |
-
output_text,
|
| 640 |
-
output_audio,
|
| 641 |
-
eval_radio,
|
| 642 |
-
eval_radio_E2E,
|
| 643 |
-
],
|
| 644 |
-
)
|
| 645 |
output_audio.play(
|
| 646 |
flash_buttons, [], [natural_response, diversity_response]
|
| 647 |
)
|
|
|
|
| 527 |
(https://github.com/siddhu001/espnet/tree/sds_demo_recipe/egs2/TEMPLATE/sds1#how-to-use).
|
| 528 |
"""
|
| 529 |
)
|
| 530 |
+
type_radio = gr.State("Cascaded")
|
| 531 |
with gr.Row():
|
| 532 |
with gr.Column(scale=1):
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 533 |
input_text = gr.Textbox(
|
| 534 |
label="LLM prompt",
|
| 535 |
visible=True,
|
| 536 |
interactive=True,
|
| 537 |
value=default_instruct,
|
| 538 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 539 |
ASR_radio = gr.Radio(
|
| 540 |
choices=ASR_options,
|
| 541 |
label="Choose ASR:",
|
|
|
|
| 559 |
)
|
| 560 |
process_btn = gr.Button("Process Audio", variant="primary")
|
| 561 |
with gr.Column(scale=1):
|
| 562 |
+
user_audio = gr.Audio(
|
| 563 |
+
sources=["upload", "microphone"],
|
| 564 |
+
type="numpy",
|
| 565 |
+
label="Upload or Record Audio File",
|
| 566 |
+
format="wav",
|
| 567 |
+
)
|
| 568 |
output_asr_text = gr.Textbox(label="ASR Transcription", interactive=False)
|
| 569 |
output_text = gr.Textbox(label="LLM Response", interactive=False)
|
| 570 |
output_audio = gr.Audio(label="TTS Output", autoplay=True, visible=True, interactive=False)
|
|
|
|
| 623 |
fn=handle_ASR_selection_wrapper,
|
| 624 |
inputs=[ASR_radio],
|
| 625 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 626 |
output_audio.play(
|
| 627 |
flash_buttons, [], [natural_response, diversity_response]
|
| 628 |
)
|