Spaces:
Sleeping
Sleeping
Add training data function to debug tab
Browse files
app.py
CHANGED
|
@@ -97,6 +97,131 @@ def test_connection():
|
|
| 97 |
except Exception as e:
|
| 98 |
return f"โ Repository access failed: {str(e)}"
|
| 99 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 100 |
|
| 101 |
def get_training_data():
|
| 102 |
"""Display the training data used for the model"""
|
|
@@ -320,6 +445,62 @@ with gr.Blocks(title="Hebrew Intent Classification - Debug") as demo:
|
|
| 320 |
outputs=[result_output, confidence_output]
|
| 321 |
)
|
| 322 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 323 |
if __name__ == "__main__":
|
| 324 |
demo.launch(
|
| 325 |
share=True,
|
|
|
|
| 97 |
except Exception as e:
|
| 98 |
return f"โ Repository access failed: {str(e)}"
|
| 99 |
|
| 100 |
+
def get_training_data():
|
| 101 |
+
"""Display the training data used for the model"""
|
| 102 |
+
training_data = [
|
| 103 |
+
("ืฉืืืชื ืืช ืืกืืกืื ืฉืื", "ืฉืืืช ืกืืกืื"),
|
| 104 |
+
("ืืื ืื ื ืืืื ืืช ืืื ืื?", "ืืืืื ืื ืื"),
|
| 105 |
+
("ืื ืืืืืจ ืฉื ืืชืืื ืืช?", "ืฉืืื ืืืืืช"),
|
| 106 |
+
("ืืืชืจ ืื ืขืืื ืื", "ืชืืืื ืืื ืืช"),
|
| 107 |
+
("ืื ื ืื ืืฆืืื ืืืชืืืจ", "ืชืืืื ืืื ืืช"),
|
| 108 |
+
("ืืื ืื ื ืืฉื ื ืืช ืืชืืืช ืืืืืืื?", "ืฉืืื ืืืืืช"),
|
| 109 |
+
("ืื ื ืจืืฆื ืืฉืืจื ืืช ืืชืืื ืืช ืฉืื", "ืฉืืื ืืืืืช"),
|
| 110 |
+
("ืืืฉืืื ืฉืื ื ื ืขื", "ืชืืืื ืืื ืืช"),
|
| 111 |
+
("ืื ื ืื ืืงืื ืืืืืื", "ืชืืืื ืืื ืืช"),
|
| 112 |
+
("ืืื ืื ื ืจืืื ืืช ืืืฉืืื ืืช ืฉืื?", "ืฉืืื ืืืืืช"),
|
| 113 |
+
("ืื ื ืจืืฆื ืืืื ืืช ืืฉืืจืืช", "ืืืืื ืื ืื"),
|
| 114 |
+
("ืฉืืืชื ืืช ืคืจืื ืืืชืืืจืืช", "ืฉืืืช ืกืืกืื"),
|
| 115 |
+
("ืืืืืชื ืืช ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 116 |
+
("ืื ืืืืจ ืืช ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 117 |
+
("ืืกืืกืื ืื ืขืืืืช", "ืฉืืืช ืกืืกืื"),
|
| 118 |
+
("ืื ืืฆืืื ืืืืื ืก ืขื ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 119 |
+
("ืฆืจืื ืืืคืก ืืช ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 120 |
+
("ืืขืื ืขื ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 121 |
+
("ืืกืืกืื ืฉืื ืื ื ืืื ื", "ืฉืืืช ืกืืกืื"),
|
| 122 |
+
("ืฉืืืชื ืื ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 123 |
+
("ืืื ืื ื ืืฉืืืจ ืืช ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 124 |
+
("ืจืืฆื ืืฉื ืืช ืืช ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 125 |
+
("ืืกืืกืื ืื ืืชืงืืืช", "ืฉืืืช ืกืืกืื"),
|
| 126 |
+
("ืืขืืืช ืืชืืืจืืช - ืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 127 |
+
("ืฆืจืื ืขืืจื ืขื ืืกืืกืื", "ืฉืืืช ืกืืกืื"),
|
| 128 |
+
("ืื ืืืืข ืื ืืกืืกืื ืฉืื", "ืฉืืืช ืกืืกืื"),
|
| 129 |
+
("ืจืืฆื ืืืื ืืช ืืฉืืจืืช", "ืืืืื ืื ืื"),
|
| 130 |
+
("ืืื ืืคืกืืงืื ืืช ืืื ืื", "ืืืืื ืื ืื"),
|
| 131 |
+
("ืจืืฆื ืืืคืกืืง ืืช ืืชืฉืืื", "ืืืืื ืื ืื"),
|
| 132 |
+
("ืืื ืืืฆืืื ืืืื ืื", "ืืืืื ืื ืื"),
|
| 133 |
+
("ืืงืฉื ืืืืืื ืื ืื", "ืืืืื ืื ืื"),
|
| 134 |
+
("ืื ืจืืฆื ืืืชืจ ืืช ืืฉืืจืืช", "ืืืืื ืื ืื"),
|
| 135 |
+
("ืืื ืืืืืื ืืช ืืืฉืืื", "ืืืืื ืื ืื"),
|
| 136 |
+
("ืจืืฆื ืืกืืืจ ืืช ืืืฉืืื", "ืืืืื ืื ืื"),
|
| 137 |
+
("ืขืืจื ืืืืืื ืื ืื", "ืืืืื ืื ืื"),
|
| 138 |
+
("ืืืื ืืืืื ืืื ืื", "ืืืืื ืื ืื"),
|
| 139 |
+
("ืืขืื ืืื ืืืื", "ืืืืื ืื ืื"),
|
| 140 |
+
("ืืื ืืคืกืืงืื ืืช ืืฉืืจืืช", "ืืืืื ืื ืื"),
|
| 141 |
+
("ืจืืฆื ืืืคืกืืง ืืช ืืืจืฉืื", "ืืืืื ืื ืื"),
|
| 142 |
+
("ืืงืฉื ืืืคืกืงืช ืฉืืจืืช", "ืืืืื ืื ืื"),
|
| 143 |
+
("ืื ืืืื ืืฉืืจืืช", "ืฉืืื ืืืืืช"),
|
| 144 |
+
("ืืืื ืชืืื ืืืช ืืฉ ืืื", "ืฉืืื ืืืืืช"),
|
| 145 |
+
("ืืื ืขืืื ืืืืืื", "ืฉืืื ืืืืืช"),
|
| 146 |
+
("ืื ืืืืื ืืื ืืชืืื ืืืช", "ืฉืืื ืืืืืช"),
|
| 147 |
+
("ืืื ืื ื ืืฉื ื ืืช ืืคืจืืื ืฉืื", "ืฉืืื ืืืืืช"),
|
| 148 |
+
("ืืื ืืคืฉืจ ืืฉืืจื", "ืฉืืื ืืืืืช"),
|
| 149 |
+
("ืื ืืืคืฉืจืืืืช ืฉืืื", "ืฉืืื ืืืืืช"),
|
| 150 |
+
("ืื ื ืจืืฆื ืืขืืื ืคืจืืื", "ืฉืืื ืืืืืช"),
|
| 151 |
+
("ืืื ืจืืืื ืืช ืืืืกืืืจืื", "ืฉืืื ืืืืืช"),
|
| 152 |
+
("ืืืคืืืงืฆืื ืงืืจืกืช", "ืชืืืื ืืื ืืช"),
|
| 153 |
+
("ืืฉ ืืื ืืืชืจ", "ืชืืืื ืืื ืืช"),
|
| 154 |
+
("ืืืฃ ืื ื ืืขื", "ืชืืืื ืืื ืืช"),
|
| 155 |
+
("ืฉืืืื ืืืขืจืืช", "ืชืืืื ืืื ืืช"),
|
| 156 |
+
("ืืืืขื ืื ืขืืื", "ืชืืืื ืืื ืืช"),
|
| 157 |
+
("ืืขืื ืืื ืืช", "ืชืืืื ืืื ืืช"),
|
| 158 |
+
("ืืืขืจืืช ืื ืืืืื", "ืชืืืื ืืื ืืช"),
|
| 159 |
+
("ืฉืืืืช ืืืืืจ", "ืชืืืื ืืื ืืช"),
|
| 160 |
+
("ืืืคืชืืจ ืื ืขืืื", "ืชืืืื ืืื ืืช"),
|
| 161 |
+
("ืืชืืื ืืช ืื ื ืืขื ืืช", "ืชืืืื ืืื ืืช"),
|
| 162 |
+
("ืืืืืืื ืื ืืชื ืื", "ืชืืืื ืืื ืืช"),
|
| 163 |
+
("ืืืืืืช ืืืชืจ", "ืชืืืื ืืื ืืช")
|
| 164 |
+
]
|
| 165 |
+
|
| 166 |
+
# Count examples per category
|
| 167 |
+
category_counts = {}
|
| 168 |
+
for _, label in training_data:
|
| 169 |
+
category_counts[label] = category_counts.get(label, 0) + 1
|
| 170 |
+
|
| 171 |
+
result = f"""
|
| 172 |
+
๐ **Training Data Summary**
|
| 173 |
+
Total Examples: {len(training_data)}
|
| 174 |
+
|
| 175 |
+
๐ **Examples per Category:**
|
| 176 |
+
"""
|
| 177 |
+
|
| 178 |
+
# Add category statistics
|
| 179 |
+
for category, count in sorted(category_counts.items()):
|
| 180 |
+
percentage = (count / len(training_data)) * 100
|
| 181 |
+
result += f"\nโข {category}: {count} examples ({percentage:.1f}%)"
|
| 182 |
+
|
| 183 |
+
result += f"""
|
| 184 |
+
|
| 185 |
+
๐ **Sample Training Examples:**
|
| 186 |
+
|
| 187 |
+
๐ **ืฉืืืช ืกืืกืื (Password Reset):**
|
| 188 |
+
โข ืฉืืืชื ืืช ืืกืืกืื ืฉืื
|
| 189 |
+
โข ืื ืืืืจ ืืช ืืกืืกืื
|
| 190 |
+
โข ืืกืืกืื ืื ืขืืืืช
|
| 191 |
+
โข ืฆืจืื ืืืคืก ืืช ืืกืืกืื
|
| 192 |
+
โข ืืื ืื ื ืืฉืืืจ ืืช ืืกืืกืื
|
| 193 |
+
|
| 194 |
+
โ **ืืืืื ืื ืื (Cancel Subscription):**
|
| 195 |
+
โข ืืื ืื ื ืืืื ืืช ืืื ืื?
|
| 196 |
+
โข ืจืืฆื ืืืคืกืืง ืืช ืืชืฉืืื
|
| 197 |
+
โข ืื ืจืืฆื ืืืชืจ ืืช ืืฉืืจืืช
|
| 198 |
+
โข ืืื ืืืืืื ืืช ืืืฉืืื
|
| 199 |
+
โข ืืงืฉื ืืืืืื ืื ืื
|
| 200 |
+
|
| 201 |
+
โ **ืฉืืื ืืืืืช (General Question):**
|
| 202 |
+
โข ืื ืืืืืจ ืฉื ืืชืืื ืืช?
|
| 203 |
+
โข ืืื ืขืืื ืืืืืื
|
| 204 |
+
โข ืืืื ืชืืื ืืืช ืืฉ ืืื
|
| 205 |
+
โข ืืื ืื ื ืืฉื ื ืืช ืืคืจืืื ืฉืื
|
| 206 |
+
โข ืื ืืืื ืืฉืืจืืช
|
| 207 |
+
|
| 208 |
+
๐ง **ืชืืืื ืืื ืืช (Technical Support):**
|
| 209 |
+
โข ืืืชืจ ืื ืขืืื ืื
|
| 210 |
+
โข ืืืคืืืงืฆืื ืงืืจืกืช
|
| 211 |
+
โข ืืฉ ืืื ืืืชืจ
|
| 212 |
+
โข ืืืฃ ืื ื ืืขื
|
| 213 |
+
โข ืฉืืืื ืืืขืจืืช
|
| 214 |
+
|
| 215 |
+
---
|
| 216 |
+
๐ก **Model was trained with data augmentation techniques:**
|
| 217 |
+
โข Synonym replacement
|
| 218 |
+
โข Paraphrasing
|
| 219 |
+
โข Context variation
|
| 220 |
+
โข Original 12 examples โ Enhanced to {len(training_data)} examples
|
| 221 |
+
"""
|
| 222 |
+
|
| 223 |
+
return result
|
| 224 |
+
|
| 225 |
|
| 226 |
def get_training_data():
|
| 227 |
"""Display the training data used for the model"""
|
|
|
|
| 445 |
outputs=[result_output, confidence_output]
|
| 446 |
)
|
| 447 |
|
| 448 |
+
if __name__ == "__main__":
|
| 449 |
+
demo.launch(
|
| 450 |
+
share=True,
|
| 451 |
+
server_name="0.0.0.0",
|
| 452 |
+
server_port=7860
|
| 453 |
+
)
|
| 454 |
+
test_btn = gr.Button("Test Model Loading")
|
| 455 |
+
debug_output = gr.Textbox(
|
| 456 |
+
label="Debug Output:",
|
| 457 |
+
lines=15,
|
| 458 |
+
interactive=False
|
| 459 |
+
)
|
| 460 |
+
|
| 461 |
+
test_btn.click(
|
| 462 |
+
lambda: test_model_loading()[1],
|
| 463 |
+
outputs=debug_output
|
| 464 |
+
)
|
| 465 |
+
|
| 466 |
+
conn_btn = gr.Button("Test Repository Connection")
|
| 467 |
+
conn_output = gr.Textbox(
|
| 468 |
+
label="Connection Test:",
|
| 469 |
+
lines=5,
|
| 470 |
+
interactive=False
|
| 471 |
+
)
|
| 472 |
+
|
| 473 |
+
conn_btn.click(
|
| 474 |
+
test_connection,
|
| 475 |
+
outputs=conn_output
|
| 476 |
+
)
|
| 477 |
+
|
| 478 |
+
with gr.Column():
|
| 479 |
+
data_btn = gr.Button("Show Training Data")
|
| 480 |
+
training_output = gr.Textbox(
|
| 481 |
+
label="Training Data:",
|
| 482 |
+
lines=20,
|
| 483 |
+
interactive=False
|
| 484 |
+
)
|
| 485 |
+
|
| 486 |
+
data_btn.click(
|
| 487 |
+
get_training_data,
|
| 488 |
+
outputs=training_output
|
| 489 |
+
)
|
| 490 |
+
|
| 491 |
+
# Connect classification
|
| 492 |
+
classify_btn.click(
|
| 493 |
+
classify_text,
|
| 494 |
+
inputs=[text_input],
|
| 495 |
+
outputs=[result_output, confidence_output]
|
| 496 |
+
)
|
| 497 |
+
|
| 498 |
+
text_input.submit(
|
| 499 |
+
classify_text,
|
| 500 |
+
inputs=[text_input],
|
| 501 |
+
outputs=[result_output, confidence_output]
|
| 502 |
+
)
|
| 503 |
+
|
| 504 |
if __name__ == "__main__":
|
| 505 |
demo.launch(
|
| 506 |
share=True,
|