Spaces:
Running
on
Zero
Running
on
Zero
pentarosarium
commited on
Commit
•
3ee8d61
1
Parent(s):
810b0fe
v.1.28
Browse files- app.py +58 -21
- sample_file.xlsx +0 -0
app.py
CHANGED
@@ -532,7 +532,7 @@ def create_interface():
|
|
532 |
control = ProcessControl()
|
533 |
|
534 |
with gr.Blocks(theme=gr.themes.Soft()) as app:
|
535 |
-
gr.Markdown("# AI-анализ мониторинга новостей v.1.
|
536 |
|
537 |
with gr.Row():
|
538 |
file_input = gr.File(
|
@@ -575,24 +575,33 @@ def create_interface():
|
|
575 |
with gr.Column(scale=1):
|
576 |
events_plot = gr.Plot(label="Распределение событий")
|
577 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
578 |
def stop_processing():
|
579 |
control.request_stop()
|
580 |
return "Остановка обработки..."
|
581 |
-
|
582 |
-
@spaces.GPU(duration=300)
|
583 |
def analyze(file_bytes):
|
584 |
if file_bytes is None:
|
585 |
gr.Warning("Пожалуйста, загрузите файл")
|
586 |
-
return None, None, None, "Ожидание файла..."
|
587 |
|
588 |
try:
|
589 |
-
# Reset
|
590 |
control.reset()
|
591 |
-
detector = EventDetector()
|
592 |
|
593 |
file_obj = io.BytesIO(file_bytes)
|
594 |
logger.info("File loaded into BytesIO successfully")
|
595 |
|
|
|
|
|
596 |
# Read and deduplicate data
|
597 |
df = pd.read_excel(file_obj, sheet_name='Публикации')
|
598 |
original_count = len(df)
|
@@ -605,6 +614,19 @@ def create_interface():
|
|
605 |
|
606 |
for batch_start in range(0, total, batch_size):
|
607 |
if control.should_stop():
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
608 |
break
|
609 |
|
610 |
batch_end = min(batch_start + batch_size, total)
|
@@ -630,50 +652,65 @@ def create_interface():
|
|
630 |
'Reasoning': results['reasoning'],
|
631 |
'Event_Type': results['event_type'],
|
632 |
'Event_Summary': results['event_summary'],
|
633 |
-
'
|
634 |
})
|
635 |
|
636 |
except Exception as e:
|
637 |
logger.error(f"Error processing row {idx}: {str(e)}")
|
638 |
continue
|
639 |
|
640 |
-
# Create intermediate results
|
641 |
if processed_rows:
|
642 |
result_df = pd.DataFrame(processed_rows)
|
643 |
-
|
644 |
-
|
645 |
-
result_df
|
646 |
-
|
647 |
-
|
648 |
-
|
649 |
-
|
|
|
|
|
|
|
650 |
|
651 |
# Cleanup GPU resources after batch
|
652 |
torch.cuda.empty_cache()
|
653 |
time.sleep(2)
|
654 |
|
|
|
655 |
if processed_rows:
|
656 |
final_df = pd.DataFrame(processed_rows)
|
657 |
-
|
658 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
659 |
else:
|
660 |
-
return None, None, None, "Нет обработанных данных"
|
661 |
|
662 |
except Exception as e:
|
663 |
error_msg = f"Ошибка анализа: {str(e)}"
|
664 |
logger.error(error_msg)
|
665 |
gr.Error(error_msg)
|
666 |
-
return None, None, None, error_msg
|
|
|
|
|
|
|
667 |
|
668 |
stop_btn.click(fn=stop_processing, outputs=[progress])
|
669 |
analyze_btn.click(
|
670 |
fn=analyze,
|
671 |
inputs=[file_input],
|
672 |
-
outputs=[stats, sentiment_plot, events_plot, progress]
|
673 |
)
|
674 |
|
675 |
return app
|
676 |
|
677 |
if __name__ == "__main__":
|
678 |
app = create_interface()
|
679 |
-
app.launch(share=True)
|
|
|
532 |
control = ProcessControl()
|
533 |
|
534 |
with gr.Blocks(theme=gr.themes.Soft()) as app:
|
535 |
+
gr.Markdown("# AI-анализ мониторинга новостей v.1.28")
|
536 |
|
537 |
with gr.Row():
|
538 |
file_input = gr.File(
|
|
|
575 |
with gr.Column(scale=1):
|
576 |
events_plot = gr.Plot(label="Распределение событий")
|
577 |
|
578 |
+
# Add download button to UI
|
579 |
+
with gr.Row():
|
580 |
+
download_file = gr.File(
|
581 |
+
label="📥 Скачать полный отчет",
|
582 |
+
file_types=[".xlsx"],
|
583 |
+
interactive=False
|
584 |
+
)
|
585 |
+
|
586 |
def stop_processing():
|
587 |
control.request_stop()
|
588 |
return "Остановка обработки..."
|
589 |
+
|
590 |
+
@spaces.GPU(duration=300)
|
591 |
def analyze(file_bytes):
|
592 |
if file_bytes is None:
|
593 |
gr.Warning("Пожалуйста, загрузите файл")
|
594 |
+
return None, None, None, None, "Ожидание файла..."
|
595 |
|
596 |
try:
|
597 |
+
# Reset stop flag
|
598 |
control.reset()
|
|
|
599 |
|
600 |
file_obj = io.BytesIO(file_bytes)
|
601 |
logger.info("File loaded into BytesIO successfully")
|
602 |
|
603 |
+
detector = EventDetector()
|
604 |
+
|
605 |
# Read and deduplicate data
|
606 |
df = pd.read_excel(file_obj, sheet_name='Публикации')
|
607 |
original_count = len(df)
|
|
|
614 |
|
615 |
for batch_start in range(0, total, batch_size):
|
616 |
if control.should_stop():
|
617 |
+
# Create partial results if stopped
|
618 |
+
if processed_rows:
|
619 |
+
result_df = pd.DataFrame(processed_rows)
|
620 |
+
output = create_output_file(result_df, file_obj)
|
621 |
+
if output:
|
622 |
+
fig_sentiment, fig_events = create_visualizations(result_df)
|
623 |
+
return (
|
624 |
+
result_df,
|
625 |
+
fig_sentiment,
|
626 |
+
fig_events,
|
627 |
+
(output, f"partial_results_{len(processed_rows)}_rows.xlsx"),
|
628 |
+
f"Обработка остановлена. Обработано {len(processed_rows)}/{total} строк"
|
629 |
+
)
|
630 |
break
|
631 |
|
632 |
batch_end = min(batch_start + batch_size, total)
|
|
|
652 |
'Reasoning': results['reasoning'],
|
653 |
'Event_Type': results['event_type'],
|
654 |
'Event_Summary': results['event_summary'],
|
655 |
+
'Выдержки из текста': text[:1000]
|
656 |
})
|
657 |
|
658 |
except Exception as e:
|
659 |
logger.error(f"Error processing row {idx}: {str(e)}")
|
660 |
continue
|
661 |
|
662 |
+
# Create intermediate results and yield
|
663 |
if processed_rows:
|
664 |
result_df = pd.DataFrame(processed_rows)
|
665 |
+
output = create_output_file(result_df, file_obj)
|
666 |
+
if output:
|
667 |
+
fig_sentiment, fig_events = create_visualizations(result_df)
|
668 |
+
yield (
|
669 |
+
result_df,
|
670 |
+
fig_sentiment,
|
671 |
+
fig_events,
|
672 |
+
(output, f"results_{len(processed_rows)}_rows.xlsx"),
|
673 |
+
f"Обработано {len(processed_rows)}/{total} строк"
|
674 |
+
)
|
675 |
|
676 |
# Cleanup GPU resources after batch
|
677 |
torch.cuda.empty_cache()
|
678 |
time.sleep(2)
|
679 |
|
680 |
+
# Create final results
|
681 |
if processed_rows:
|
682 |
final_df = pd.DataFrame(processed_rows)
|
683 |
+
output = create_output_file(final_df, file_obj)
|
684 |
+
if output:
|
685 |
+
fig_sentiment, fig_events = create_visualizations(final_df)
|
686 |
+
return (
|
687 |
+
final_df,
|
688 |
+
fig_sentiment,
|
689 |
+
fig_events,
|
690 |
+
(output, "final_results.xlsx"),
|
691 |
+
"Обработка завершена!"
|
692 |
+
)
|
693 |
else:
|
694 |
+
return None, None, None, None, "Нет обработанных данных"
|
695 |
|
696 |
except Exception as e:
|
697 |
error_msg = f"Ошибка анализа: {str(e)}"
|
698 |
logger.error(error_msg)
|
699 |
gr.Error(error_msg)
|
700 |
+
return None, None, None, None, error_msg
|
701 |
+
finally:
|
702 |
+
if detector:
|
703 |
+
detector.cleanup()
|
704 |
|
705 |
stop_btn.click(fn=stop_processing, outputs=[progress])
|
706 |
analyze_btn.click(
|
707 |
fn=analyze,
|
708 |
inputs=[file_input],
|
709 |
+
outputs=[stats, sentiment_plot, events_plot, download_file, progress]
|
710 |
)
|
711 |
|
712 |
return app
|
713 |
|
714 |
if __name__ == "__main__":
|
715 |
app = create_interface()
|
716 |
+
app.launch(share=True)
|
sample_file.xlsx
ADDED
Binary file (139 kB). View file
|
|