Update app.py
Browse files
app.py
CHANGED
|
@@ -27,7 +27,7 @@ rag_initialized = False
|
|
| 27 |
gemini_llm_model = None
|
| 28 |
gemini_embedding_model_name = "models/embedding-001"
|
| 29 |
|
| 30 |
-
google_api_key = os.getenv("
|
| 31 |
|
| 32 |
if not google_api_key:
|
| 33 |
print("\n" + "="*80)
|
|
@@ -419,20 +419,29 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 419 |
|
| 420 |
start_time = time.time()
|
| 421 |
error_message_markdown = ""
|
|
|
|
| 422 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 423 |
progress(0, desc="Starting analysis...")
|
| 424 |
|
| 425 |
if not rag_initialized:
|
| 426 |
error_message_markdown = "RAG system not initialized. Please ensure your GOOGLE_API_KEY is configured correctly and the specified AI model is accessible."
|
| 427 |
-
|
|
|
|
| 428 |
|
| 429 |
if gemini_llm_model is None:
|
| 430 |
error_message_markdown = "Google Generative AI LLM was not initialized. Please check startup logs."
|
| 431 |
-
|
|
|
|
| 432 |
|
| 433 |
if file is None:
|
| 434 |
error_message_markdown = "Please upload a customer ticket file."
|
| 435 |
-
|
|
|
|
| 436 |
|
| 437 |
file_path = file.name
|
| 438 |
filename = os.path.basename(file_path)
|
|
@@ -442,6 +451,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 442 |
|
| 443 |
try:
|
| 444 |
# --- PHASE 1: Initial Categorization of ALL Tickets ---
|
|
|
|
| 445 |
progress(0.1, desc="Phase 1/2: Initializing categorization...")
|
| 446 |
|
| 447 |
# Use an iterator to handle potentially very large files
|
|
@@ -458,7 +468,9 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 458 |
total_tickets_count = current_total_processed
|
| 459 |
# Adjust progress based on actual total, or a rough estimate if total_tickets_count is not yet final
|
| 460 |
estimated_total_chunks = max(1, total_tickets_count // 5000 + 1) # Crude estimate
|
| 461 |
-
|
|
|
|
|
|
|
| 462 |
|
| 463 |
if not chunk_documents:
|
| 464 |
continue
|
|
@@ -490,8 +502,10 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 490 |
|
| 491 |
try:
|
| 492 |
# Add retry logic and error handling for API calls here
|
|
|
|
| 493 |
categorization_response = gemini_llm_model.generate_content(categorization_prompt)
|
| 494 |
categorization_raw_output = categorization_response.text
|
|
|
|
| 495 |
|
| 496 |
# Parse categorization output
|
| 497 |
for line in categorization_raw_output.split('\n'):
|
|
@@ -510,6 +524,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 510 |
})
|
| 511 |
elif line:
|
| 512 |
print(f"Warning: Line did not match categorization format in chunk {i}: {line}")
|
|
|
|
| 513 |
# Fallback parsing for malformed lines
|
| 514 |
parts = line.split("Category:")
|
| 515 |
ticket_id_part = parts[0].replace("Ticket ID:", "").strip() if len(parts) > 0 else f"Unparsed_Line_{len(categorized_tickets_db)+1}"
|
|
@@ -524,6 +539,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 524 |
})
|
| 525 |
|
| 526 |
except Exception as e:
|
|
|
|
| 527 |
print(f"Error during categorization of chunk {i}: {e}")
|
| 528 |
# For a real system, you'd log this, and potentially save the raw chunk for re-processing.
|
| 529 |
# For this demo, we'll just skip the chunk or add a generic error entry.
|
|
@@ -540,20 +556,25 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 540 |
final_categorized_df = pd.DataFrame(categorized_tickets_db)
|
| 541 |
if final_categorized_df.empty:
|
| 542 |
error_message_markdown = "AI failed to categorize any tickets. Please ensure the ticket data is clear."
|
| 543 |
-
|
|
|
|
| 544 |
|
| 545 |
# --- PHASE 2: Detailed Analysis per Category ---
|
|
|
|
| 546 |
progress(0.5, desc=f"Phase 2/2: Detailing analysis for categories...")
|
| 547 |
|
| 548 |
final_analysis_rows = []
|
| 549 |
unique_categories = final_categorized_df['category'].unique()
|
|
|
|
| 550 |
|
| 551 |
# Define a batch size for processing categories for detailed analysis if there are many categories
|
| 552 |
# For 6-8 categories, this internal batching might not be strictly necessary, but good practice.
|
| 553 |
category_batch_size = 5
|
| 554 |
|
| 555 |
for i, category_chunk in enumerate(np.array_split(unique_categories, max(1, len(unique_categories) // category_batch_size))):
|
| 556 |
-
|
|
|
|
|
|
|
| 557 |
|
| 558 |
for category in category_chunk:
|
| 559 |
category_specific_tickets_df = final_categorized_df[final_categorized_df['category'] == category]
|
|
@@ -563,6 +584,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 563 |
category_tickets_content = "\n\n".join(category_specific_tickets_df['original_content'].tolist())
|
| 564 |
|
| 565 |
category_count = len(category_specific_tickets_df)
|
|
|
|
| 566 |
|
| 567 |
detail_prompt = f"""
|
| 568 |
You are an expert customer support ticket analyst.
|
|
@@ -585,12 +607,14 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 585 |
"""
|
| 586 |
|
| 587 |
try:
|
| 588 |
-
|
| 589 |
detail_response = gemini_llm_model.generate_content(detail_prompt)
|
| 590 |
detail_raw_output = detail_response.text
|
|
|
|
| 591 |
|
| 592 |
detailed_df, parsing_err = parse_markdown_table_to_df(detail_raw_output)
|
| 593 |
if parsing_err or detailed_df.empty:
|
|
|
|
| 594 |
print(f"Warning: Parsing error or empty DataFrame for category '{category}': {parsing_err or 'Empty DataFrame'}")
|
| 595 |
print(f"Raw AI output for detail: {detail_raw_output}")
|
| 596 |
default_row = {
|
|
@@ -612,6 +636,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 612 |
try:
|
| 613 |
ai_reported_count = pd.to_numeric(detailed_row['Count'], errors='coerce')
|
| 614 |
if pd.isna(ai_reported_count) or int(ai_reported_count) != category_count:
|
|
|
|
| 615 |
print(f"Warning: AI-reported count {ai_reported_count} for '{category}' does not match actual count {category_count}. Correcting.")
|
| 616 |
detailed_row['Count'] = category_count
|
| 617 |
except:
|
|
@@ -622,6 +647,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 622 |
final_analysis_rows.append(detailed_row)
|
| 623 |
|
| 624 |
except Exception as e:
|
|
|
|
| 625 |
print(f"Error detailing analysis for category '{category}': {e}")
|
| 626 |
default_row = {
|
| 627 |
"Category": category,
|
|
@@ -641,6 +667,7 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 641 |
|
| 642 |
summarized_count = analysis_df['Count'].sum()
|
| 643 |
if summarized_count != total_tickets_count:
|
|
|
|
| 644 |
print(f"CRITICAL WARNING: Total count in summarized DF ({summarized_count}) does not match original tickets ({total_tickets_count}).")
|
| 645 |
|
| 646 |
global_analysis_df_for_export = analysis_df.copy()
|
|
@@ -648,26 +675,32 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 648 |
|
| 649 |
except ValueError as e:
|
| 650 |
error_message_markdown = f"Ticket Analysis Error: {e}"
|
|
|
|
| 651 |
global_raw_markdown_for_export = error_message_markdown + disclaimer_text_for_raw
|
| 652 |
global_analysis_df_for_export = pd.DataFrame()
|
| 653 |
-
return f"<p style='color:red;'>{error_message_markdown}</p>"
|
| 654 |
except Exception as e:
|
| 655 |
error_message_markdown = f"An unexpected error occurred during analysis: {e}. Check server logs for details."
|
|
|
|
| 656 |
print(f"Error during RAG analysis: {e}")
|
| 657 |
global_raw_markdown_for_export = error_message_markdown + disclaimer_text_for_raw
|
| 658 |
global_analysis_df_for_export = pd.DataFrame()
|
| 659 |
-
return f"<p style='color:red;'>{error_message_markdown}</p>"
|
| 660 |
|
|
|
|
| 661 |
progress(0.98, desc="Generating summary report for display...")
|
| 662 |
|
| 663 |
summarized_analysis_html = ""
|
| 664 |
try:
|
| 665 |
if not analysis_df.empty:
|
| 666 |
summarized_analysis_html = generate_summary_html_by_category(analysis_df)
|
|
|
|
| 667 |
else:
|
| 668 |
summarized_analysis_html = "<p style='color:orange;'>AI output parsed into an empty DataFrame or no categories were generated. Please check the uploaded file and try again.</p>"
|
|
|
|
| 669 |
|
| 670 |
except Exception as e:
|
|
|
|
| 671 |
print(f"Failed to generate HTML summary from DataFrame: {e}")
|
| 672 |
summarized_analysis_html = f"""
|
| 673 |
<div style="margin-bottom: 20px;">
|
|
@@ -717,8 +750,9 @@ def perform_ticket_analysis(file, user_query, progress=gr.Progress()):
|
|
| 717 |
</div>
|
| 718 |
</div>
|
| 719 |
"""
|
|
|
|
| 720 |
progress(1, desc="Analysis complete!")
|
| 721 |
-
return html_output
|
| 722 |
|
| 723 |
|
| 724 |
# --- Output Download Functions (using global_analysis_df_for_export) ---
|
|
@@ -1038,9 +1072,18 @@ with gr.Blocks(theme=gr.themes.Soft(), css="""
|
|
| 1038 |
lines=2,
|
| 1039 |
interactive=True
|
| 1040 |
)
|
| 1041 |
-
|
| 1042 |
submit_btn = gr.Button("Analyze Tickets", variant="primary")
|
| 1043 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1044 |
output_html_display = gr.HTML(
|
| 1045 |
label="AI Ticket Analysis Result"
|
| 1046 |
)
|
|
@@ -1059,10 +1102,11 @@ with gr.Blocks(theme=gr.themes.Soft(), css="""
|
|
| 1059 |
</p>
|
| 1060 |
""")
|
| 1061 |
|
|
|
|
| 1062 |
submit_btn.click(
|
| 1063 |
fn=perform_ticket_analysis,
|
| 1064 |
inputs=[file_input, query_input],
|
| 1065 |
-
outputs=output_html_display,
|
| 1066 |
queue=True
|
| 1067 |
)
|
| 1068 |
|
|
|
|
| 27 |
gemini_llm_model = None
|
| 28 |
gemini_embedding_model_name = "models/embedding-001"
|
| 29 |
|
| 30 |
+
google_api_key = os.getenv("GOOGLE_API_key")
|
| 31 |
|
| 32 |
if not google_api_key:
|
| 33 |
print("\n" + "="*80)
|
|
|
|
| 419 |
|
| 420 |
start_time = time.time()
|
| 421 |
error_message_markdown = ""
|
| 422 |
+
log_messages = [] # Initialize log messages list
|
| 423 |
|
| 424 |
+
def append_log_message(message):
|
| 425 |
+
timestamp = datetime.now().strftime("%H:%M:%S")
|
| 426 |
+
log_messages.append(f"[{timestamp}] {message}")
|
| 427 |
+
|
| 428 |
+
append_log_message("Starting analysis...")
|
| 429 |
progress(0, desc="Starting analysis...")
|
| 430 |
|
| 431 |
if not rag_initialized:
|
| 432 |
error_message_markdown = "RAG system not initialized. Please ensure your GOOGLE_API_KEY is configured correctly and the specified AI model is accessible."
|
| 433 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 434 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 435 |
|
| 436 |
if gemini_llm_model is None:
|
| 437 |
error_message_markdown = "Google Generative AI LLM was not initialized. Please check startup logs."
|
| 438 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 439 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 440 |
|
| 441 |
if file is None:
|
| 442 |
error_message_markdown = "Please upload a customer ticket file."
|
| 443 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 444 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 445 |
|
| 446 |
file_path = file.name
|
| 447 |
filename = os.path.basename(file_path)
|
|
|
|
| 451 |
|
| 452 |
try:
|
| 453 |
# --- PHASE 1: Initial Categorization of ALL Tickets ---
|
| 454 |
+
append_log_message("Phase 1/2: Initializing categorization...")
|
| 455 |
progress(0.1, desc="Phase 1/2: Initializing categorization...")
|
| 456 |
|
| 457 |
# Use an iterator to handle potentially very large files
|
|
|
|
| 468 |
total_tickets_count = current_total_processed
|
| 469 |
# Adjust progress based on actual total, or a rough estimate if total_tickets_count is not yet final
|
| 470 |
estimated_total_chunks = max(1, total_tickets_count // 5000 + 1) # Crude estimate
|
| 471 |
+
progress_desc = f"Phase 1/2: Categorizing tickets... {total_tickets_count} processed so far."
|
| 472 |
+
append_log_message(progress_desc)
|
| 473 |
+
progress(0.1 + (0.4 * (i+1) / estimated_total_chunks), desc=progress_desc)
|
| 474 |
|
| 475 |
if not chunk_documents:
|
| 476 |
continue
|
|
|
|
| 502 |
|
| 503 |
try:
|
| 504 |
# Add retry logic and error handling for API calls here
|
| 505 |
+
append_log_message(f"Calling LLM for categorization of {len(chunk_documents)} tickets...")
|
| 506 |
categorization_response = gemini_llm_model.generate_content(categorization_prompt)
|
| 507 |
categorization_raw_output = categorization_response.text
|
| 508 |
+
append_log_message("LLM categorization response received. Parsing...")
|
| 509 |
|
| 510 |
# Parse categorization output
|
| 511 |
for line in categorization_raw_output.split('\n'):
|
|
|
|
| 524 |
})
|
| 525 |
elif line:
|
| 526 |
print(f"Warning: Line did not match categorization format in chunk {i}: {line}")
|
| 527 |
+
append_log_message(f"Warning: Malformed categorization line in chunk {i}: {line}")
|
| 528 |
# Fallback parsing for malformed lines
|
| 529 |
parts = line.split("Category:")
|
| 530 |
ticket_id_part = parts[0].replace("Ticket ID:", "").strip() if len(parts) > 0 else f"Unparsed_Line_{len(categorized_tickets_db)+1}"
|
|
|
|
| 539 |
})
|
| 540 |
|
| 541 |
except Exception as e:
|
| 542 |
+
append_log_message(f"Error during categorization of chunk {i}: {e}")
|
| 543 |
print(f"Error during categorization of chunk {i}: {e}")
|
| 544 |
# For a real system, you'd log this, and potentially save the raw chunk for re-processing.
|
| 545 |
# For this demo, we'll just skip the chunk or add a generic error entry.
|
|
|
|
| 556 |
final_categorized_df = pd.DataFrame(categorized_tickets_db)
|
| 557 |
if final_categorized_df.empty:
|
| 558 |
error_message_markdown = "AI failed to categorize any tickets. Please ensure the ticket data is clear."
|
| 559 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 560 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 561 |
|
| 562 |
# --- PHASE 2: Detailed Analysis per Category ---
|
| 563 |
+
append_log_message("Phase 2/2: Detailing analysis for categories...")
|
| 564 |
progress(0.5, desc=f"Phase 2/2: Detailing analysis for categories...")
|
| 565 |
|
| 566 |
final_analysis_rows = []
|
| 567 |
unique_categories = final_categorized_df['category'].unique()
|
| 568 |
+
append_log_message(f"Identified {len(unique_categories)} unique categories for detailed analysis.")
|
| 569 |
|
| 570 |
# Define a batch size for processing categories for detailed analysis if there are many categories
|
| 571 |
# For 6-8 categories, this internal batching might not be strictly necessary, but good practice.
|
| 572 |
category_batch_size = 5
|
| 573 |
|
| 574 |
for i, category_chunk in enumerate(np.array_split(unique_categories, max(1, len(unique_categories) // category_batch_size))):
|
| 575 |
+
progress_desc = f"Phase 2/2: Detailing analysis for category batch {i+1}..."
|
| 576 |
+
append_log_message(progress_desc)
|
| 577 |
+
progress((0.5 + (0.45 * i / len(np.array_split(unique_categories, max(1, len(unique_categories) // category_batch_size))))), desc=progress_desc)
|
| 578 |
|
| 579 |
for category in category_chunk:
|
| 580 |
category_specific_tickets_df = final_categorized_df[final_categorized_df['category'] == category]
|
|
|
|
| 584 |
category_tickets_content = "\n\n".join(category_specific_tickets_df['original_content'].tolist())
|
| 585 |
|
| 586 |
category_count = len(category_specific_tickets_df)
|
| 587 |
+
append_log_message(f"Analyzing category: '{category}' with {category_count} tickets.")
|
| 588 |
|
| 589 |
detail_prompt = f"""
|
| 590 |
You are an expert customer support ticket analyst.
|
|
|
|
| 607 |
"""
|
| 608 |
|
| 609 |
try:
|
| 610 |
+
append_log_message(f"Calling LLM for detailed analysis of category '{category}'...")
|
| 611 |
detail_response = gemini_llm_model.generate_content(detail_prompt)
|
| 612 |
detail_raw_output = detail_response.text
|
| 613 |
+
append_log_message(f"LLM detailed response for '{category}' received. Parsing...")
|
| 614 |
|
| 615 |
detailed_df, parsing_err = parse_markdown_table_to_df(detail_raw_output)
|
| 616 |
if parsing_err or detailed_df.empty:
|
| 617 |
+
append_log_message(f"Warning: Parsing error or empty DataFrame for category '{category}': {parsing_err or 'Empty DataFrame'}")
|
| 618 |
print(f"Warning: Parsing error or empty DataFrame for category '{category}': {parsing_err or 'Empty DataFrame'}")
|
| 619 |
print(f"Raw AI output for detail: {detail_raw_output}")
|
| 620 |
default_row = {
|
|
|
|
| 636 |
try:
|
| 637 |
ai_reported_count = pd.to_numeric(detailed_row['Count'], errors='coerce')
|
| 638 |
if pd.isna(ai_reported_count) or int(ai_reported_count) != category_count:
|
| 639 |
+
append_log_message(f"Warning: AI-reported count {ai_reported_count} for '{category}' does not match actual count {category_count}. Correcting.")
|
| 640 |
print(f"Warning: AI-reported count {ai_reported_count} for '{category}' does not match actual count {category_count}. Correcting.")
|
| 641 |
detailed_row['Count'] = category_count
|
| 642 |
except:
|
|
|
|
| 647 |
final_analysis_rows.append(detailed_row)
|
| 648 |
|
| 649 |
except Exception as e:
|
| 650 |
+
append_log_message(f"Error detailing analysis for category '{category}': {e}")
|
| 651 |
print(f"Error detailing analysis for category '{category}': {e}")
|
| 652 |
default_row = {
|
| 653 |
"Category": category,
|
|
|
|
| 667 |
|
| 668 |
summarized_count = analysis_df['Count'].sum()
|
| 669 |
if summarized_count != total_tickets_count:
|
| 670 |
+
append_log_message(f"CRITICAL WARNING: Total count in summarized DF ({summarized_count}) does not match original tickets ({total_tickets_count}).")
|
| 671 |
print(f"CRITICAL WARNING: Total count in summarized DF ({summarized_count}) does not match original tickets ({total_tickets_count}).")
|
| 672 |
|
| 673 |
global_analysis_df_for_export = analysis_df.copy()
|
|
|
|
| 675 |
|
| 676 |
except ValueError as e:
|
| 677 |
error_message_markdown = f"Ticket Analysis Error: {e}"
|
| 678 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 679 |
global_raw_markdown_for_export = error_message_markdown + disclaimer_text_for_raw
|
| 680 |
global_analysis_df_for_export = pd.DataFrame()
|
| 681 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 682 |
except Exception as e:
|
| 683 |
error_message_markdown = f"An unexpected error occurred during analysis: {e}. Check server logs for details."
|
| 684 |
+
append_log_message(f"Error: {error_message_markdown}")
|
| 685 |
print(f"Error during RAG analysis: {e}")
|
| 686 |
global_raw_markdown_for_export = error_message_markdown + disclaimer_text_for_raw
|
| 687 |
global_analysis_df_for_export = pd.DataFrame()
|
| 688 |
+
return f"<p style='color:red;'>{error_message_markdown}</p>", "\n".join(log_messages)
|
| 689 |
|
| 690 |
+
append_log_message("Generating summary report for display...")
|
| 691 |
progress(0.98, desc="Generating summary report for display...")
|
| 692 |
|
| 693 |
summarized_analysis_html = ""
|
| 694 |
try:
|
| 695 |
if not analysis_df.empty:
|
| 696 |
summarized_analysis_html = generate_summary_html_by_category(analysis_df)
|
| 697 |
+
append_log_message("HTML summary generated successfully.")
|
| 698 |
else:
|
| 699 |
summarized_analysis_html = "<p style='color:orange;'>AI output parsed into an empty DataFrame or no categories were generated. Please check the uploaded file and try again.</p>"
|
| 700 |
+
append_log_message("Warning: No categories generated, HTML summary might be empty.")
|
| 701 |
|
| 702 |
except Exception as e:
|
| 703 |
+
append_log_message(f"Failed to generate HTML summary from DataFrame: {e}")
|
| 704 |
print(f"Failed to generate HTML summary from DataFrame: {e}")
|
| 705 |
summarized_analysis_html = f"""
|
| 706 |
<div style="margin-bottom: 20px;">
|
|
|
|
| 750 |
</div>
|
| 751 |
</div>
|
| 752 |
"""
|
| 753 |
+
append_log_message("Analysis complete! Displaying results.")
|
| 754 |
progress(1, desc="Analysis complete!")
|
| 755 |
+
return html_output, "\n".join(log_messages)
|
| 756 |
|
| 757 |
|
| 758 |
# --- Output Download Functions (using global_analysis_df_for_export) ---
|
|
|
|
| 1072 |
lines=2,
|
| 1073 |
interactive=True
|
| 1074 |
)
|
| 1075 |
+
|
| 1076 |
submit_btn = gr.Button("Analyze Tickets", variant="primary")
|
| 1077 |
|
| 1078 |
+
# New Textbox for detailed processing log
|
| 1079 |
+
progress_log = gr.Textbox(
|
| 1080 |
+
label="Processing Log",
|
| 1081 |
+
interactive=False,
|
| 1082 |
+
lines=5, # Show 5 lines at a time
|
| 1083 |
+
autoscroll=True, # Automatically scroll to the latest message
|
| 1084 |
+
elem_id="progress-log-box" # Add an ID for potential CSS styling if needed
|
| 1085 |
+
)
|
| 1086 |
+
|
| 1087 |
output_html_display = gr.HTML(
|
| 1088 |
label="AI Ticket Analysis Result"
|
| 1089 |
)
|
|
|
|
| 1102 |
</p>
|
| 1103 |
""")
|
| 1104 |
|
| 1105 |
+
# Modified submit_btn.click to update both HTML display and the new progress_log
|
| 1106 |
submit_btn.click(
|
| 1107 |
fn=perform_ticket_analysis,
|
| 1108 |
inputs=[file_input, query_input],
|
| 1109 |
+
outputs=[output_html_display, progress_log], # Now returns two outputs
|
| 1110 |
queue=True
|
| 1111 |
)
|
| 1112 |
|