Spaces:
Sleeping
Sleeping
Commit
·
beb5d17
1
Parent(s):
df58192
new code for showing chunks
Browse files
app.py
CHANGED
|
@@ -13,21 +13,21 @@ from config import (
|
|
| 13 |
|
| 14 |
def create_chunks_display_html(chunk_info):
|
| 15 |
if not chunk_info:
|
| 16 |
-
return "<div style='padding: 20px; text-align: center;'>Нет данных о чанках</div>"
|
| 17 |
|
| 18 |
-
html = "<div style='max-height: 500px; overflow-y: auto; padding: 10px;'>"
|
| 19 |
-
html += f"<h4
|
| 20 |
|
| 21 |
for i, chunk in enumerate(chunk_info):
|
| 22 |
bg_color = "#f8f9fa" if i % 2 == 0 else "#e9ecef"
|
| 23 |
html += f"""
|
| 24 |
-
<div style='background-color: {bg_color}; padding: 10px; margin: 5px 0; border-radius: 5px; border-left: 4px solid #007bff;'>
|
| 25 |
<strong>Документ:</strong> {chunk['document_id']}<br>
|
| 26 |
<strong>Раздел:</strong> {chunk['section_id']}<br>
|
| 27 |
-
<strong
|
| 28 |
<strong>Содержание:</strong><br>
|
| 29 |
-
<div style='background-color: white; padding: 8px; margin-top: 5px; border-radius: 3px; font-family: monospace; font-size: 12px;'>
|
| 30 |
-
{chunk['
|
| 31 |
</div>
|
| 32 |
</div>
|
| 33 |
"""
|
|
@@ -108,6 +108,46 @@ def switch_model(model_name, vector_index):
|
|
| 108 |
log_message(error_msg)
|
| 109 |
return None, f"❌ {error_msg}"
|
| 110 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 111 |
def create_demo_interface(answer_question_func, switch_model_func, current_model, chunk_info=None):
|
| 112 |
with gr.Blocks(title="AIEXP - AI Expert для нормативной документации", theme=gr.themes.Soft()) as demo:
|
| 113 |
|
|
@@ -166,7 +206,7 @@ def create_demo_interface(answer_question_func, switch_model_func, current_model
|
|
| 166 |
with gr.Column(scale=1):
|
| 167 |
sources_output = gr.HTML(
|
| 168 |
label="",
|
| 169 |
-
value="<div style='background-color: #2d3748; color: white; padding: 20px; border-radius: 10px; text-align: center;'>Здесь появятся
|
| 170 |
)
|
| 171 |
|
| 172 |
switch_btn.click(
|
|
@@ -186,17 +226,9 @@ def create_demo_interface(answer_question_func, switch_model_func, current_model
|
|
| 186 |
inputs=[question_input],
|
| 187 |
outputs=[answer_output, sources_output]
|
| 188 |
)
|
| 189 |
-
|
| 190 |
-
with gr.Tab("Просмотр чанков"):
|
| 191 |
-
gr.Markdown("### Содержание обработанных чанков документов")
|
| 192 |
-
|
| 193 |
-
chunks_display = gr.HTML(
|
| 194 |
-
value=create_chunks_display_html(chunk_info),
|
| 195 |
-
label="Информация о чанках"
|
| 196 |
-
)
|
| 197 |
-
|
| 198 |
return demo
|
| 199 |
|
|
|
|
| 200 |
query_engine = None
|
| 201 |
chunks_df = None
|
| 202 |
reranker = None
|
|
|
|
| 13 |
|
| 14 |
def create_chunks_display_html(chunk_info):
|
| 15 |
if not chunk_info:
|
| 16 |
+
return "<div style='padding: 20px; text-align: center; color: black;'>Нет данных о чанках</div>"
|
| 17 |
|
| 18 |
+
html = "<div style='max-height: 500px; overflow-y: auto; padding: 10px; color: black;'>"
|
| 19 |
+
html += f"<h4 style='color: black;'>Найдено релевантных чанков: {len(chunk_info)}</h4>"
|
| 20 |
|
| 21 |
for i, chunk in enumerate(chunk_info):
|
| 22 |
bg_color = "#f8f9fa" if i % 2 == 0 else "#e9ecef"
|
| 23 |
html += f"""
|
| 24 |
+
<div style='background-color: {bg_color}; padding: 10px; margin: 5px 0; border-radius: 5px; border-left: 4px solid #007bff; color: black;'>
|
| 25 |
<strong>Документ:</strong> {chunk['document_id']}<br>
|
| 26 |
<strong>Раздел:</strong> {chunk['section_id']}<br>
|
| 27 |
+
<strong>Ранг:</strong> {i+1} | <strong>Размер:</strong> {chunk['chunk_size']} символов<br>
|
| 28 |
<strong>Содержание:</strong><br>
|
| 29 |
+
<div style='background-color: white; padding: 8px; margin-top: 5px; border-radius: 3px; font-family: monospace; font-size: 12px; color: black; max-height: 200px; overflow-y: auto;'>
|
| 30 |
+
{chunk['chunk_text']}
|
| 31 |
</div>
|
| 32 |
</div>
|
| 33 |
"""
|
|
|
|
| 108 |
log_message(error_msg)
|
| 109 |
return None, f"❌ {error_msg}"
|
| 110 |
|
| 111 |
+
def main_answer_question(question):
|
| 112 |
+
from index_retriever import rerank_nodes
|
| 113 |
+
global query_engine, reranker, current_model, chunks_df
|
| 114 |
+
if not question.strip():
|
| 115 |
+
return ("<div style='color: black;'>Пожалуйста, введите вопрос</div>",
|
| 116 |
+
"<div style='color: black;'>Источники появятся после обработки запроса</div>")
|
| 117 |
+
|
| 118 |
+
try:
|
| 119 |
+
# Get response and retrieve nodes for reranking display
|
| 120 |
+
response = query_engine.query(question)
|
| 121 |
+
|
| 122 |
+
# Get source nodes from response
|
| 123 |
+
source_nodes = response.source_nodes if hasattr(response, 'source_nodes') else []
|
| 124 |
+
|
| 125 |
+
# Rerank the nodes
|
| 126 |
+
reranked_nodes = rerank_nodes(question, source_nodes, reranker, top_k=10)
|
| 127 |
+
|
| 128 |
+
# Create chunk info for display
|
| 129 |
+
chunk_info = []
|
| 130 |
+
for node in reranked_nodes:
|
| 131 |
+
chunk_info.append({
|
| 132 |
+
'document_id': node.metadata.get('document_id', 'unknown'),
|
| 133 |
+
'section_id': node.metadata.get('section_id', 'unknown'),
|
| 134 |
+
'chunk_size': len(node.text),
|
| 135 |
+
'chunk_text': node.text
|
| 136 |
+
})
|
| 137 |
+
|
| 138 |
+
# Format response
|
| 139 |
+
answer_html = f"<div style='color: black; padding: 15px; background-color: #f8f9fa; border-radius: 8px;'>{str(response)}</div>"
|
| 140 |
+
sources_html = create_chunks_display_html(chunk_info)
|
| 141 |
+
|
| 142 |
+
return answer_html, sources_html
|
| 143 |
+
|
| 144 |
+
except Exception as e:
|
| 145 |
+
log_message(f"Ошибка при ответе на вопрос: {str(e)}")
|
| 146 |
+
return (f"<div style='color: red;'>Ошибка: {str(e)}</div>",
|
| 147 |
+
"<div style='color: black;'>Источники недоступны из-за ошибки</div>")
|
| 148 |
+
|
| 149 |
+
|
| 150 |
+
|
| 151 |
def create_demo_interface(answer_question_func, switch_model_func, current_model, chunk_info=None):
|
| 152 |
with gr.Blocks(title="AIEXP - AI Expert для нормативной документации", theme=gr.themes.Soft()) as demo:
|
| 153 |
|
|
|
|
| 206 |
with gr.Column(scale=1):
|
| 207 |
sources_output = gr.HTML(
|
| 208 |
label="",
|
| 209 |
+
value="<div style='background-color: #2d3748; color: white; padding: 20px; border-radius: 10px; text-align: center;'>Здесь появятся релевантные чанки...</div>",
|
| 210 |
)
|
| 211 |
|
| 212 |
switch_btn.click(
|
|
|
|
| 226 |
inputs=[question_input],
|
| 227 |
outputs=[answer_output, sources_output]
|
| 228 |
)
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 229 |
return demo
|
| 230 |
|
| 231 |
+
|
| 232 |
query_engine = None
|
| 233 |
chunks_df = None
|
| 234 |
reranker = None
|