VectorV17 commited on
Commit
b0b7b82
·
verified ·
1 Parent(s): 4fa324b

Update app.py

Browse files
Files changed (1) hide show
  1. app.py +29 -19
app.py CHANGED
@@ -21,19 +21,19 @@ MODEL_PATH = "loan_stacking_model.pkl"
21
  DATA_PATH = "loan_dataset.csv"
22
 
23
  # =====================================================================
24
- # AUTOMATED TRAINING TRIGGER (Runs fast once if model binary is missing)
25
  # =====================================================================
26
- if not os.path.exists(MODEL_PATH):
27
- print(f"'{MODEL_PATH}' not found. Initializing ultra-fast pipeline training framework...")
28
 
29
  if not os.path.exists(DATA_PATH):
30
- raise FileNotFoundError(f"Missing '{DATA_PATH}'. Please ensure your dataset is in the same directory.")
31
 
32
  df = pd.read_csv(DATA_PATH)
33
  X = df.drop('BAD', axis=1)
34
  y = df['BAD']
35
 
36
- # Handle explicit data types to remove Pandas4Warnings
37
  num_cols = X.select_dtypes(include=['int64', 'float64']).columns.tolist()
38
  cat_cols = X.select_dtypes(include=['object', 'string']).columns.tolist()
39
 
@@ -53,7 +53,7 @@ if not os.path.exists(MODEL_PATH):
53
 
54
  X_cleaned = cap_outliers_iqr(X, num_cols)
55
 
56
- # Fast Preprocessing Layouts (Fixes the categorical string math error)
57
  num_pipeline = Pipeline([
58
  ('imputer', KNNImputer(n_neighbors=5)),
59
  ('scaler', StandardScaler())
@@ -69,32 +69,44 @@ if not os.path.exists(MODEL_PATH):
69
  ('cat', cat_pipeline, cat_cols)
70
  ])
71
 
72
- # Optimized Hyperparameters for instantaneous execution
73
  rf_base = RandomForestClassifier(n_estimators=50, max_depth=8, random_state=42, n_jobs=-1)
74
  xgb_base = XGBClassifier(n_estimators=50, learning_rate=0.1, max_depth=4, eval_metric='logloss', random_state=42, n_jobs=-1)
75
  voting_base = VotingClassifier(estimators=[('rf_v', rf_base), ('xgb_v', xgb_base)], voting='soft')
76
 
77
- # Meta Stacking Layer Layout
78
  stacking_ensemble = StackingClassifier(
79
  estimators=[('rf', rf_base), ('xgb', xgb_base), ('voting', voting_base)],
80
  final_estimator=LogisticRegression(max_iter=500, C=0.1),
81
- cv=2, # Drastically speeds up execution time
82
  n_jobs=-1
83
  )
84
 
85
- production_pipeline = ImbPipeline([
86
  ('preprocessing', preprocessor),
87
  ('smote', SMOTE(random_state=42, sampling_strategy=0.6)),
88
  ('ensemble_model', stacking_ensemble)
89
  ])
90
 
91
- production_pipeline.fit(X_cleaned, y)
92
- joblib.dump(production_pipeline, MODEL_PATH)
93
- print("Model Stacking Pipeline Cached Successfully!")
 
94
 
 
 
 
 
 
 
 
 
 
 
 
 
95
  else:
96
- print("Pre-trained model layout located. Loading pipeline parameters instantly...")
97
- production_pipeline = joblib.load(MODEL_PATH)
98
 
99
 
100
  # =====================================================================
@@ -123,12 +135,12 @@ def evaluate_loan_risk(loan, mortdue, value, reason, job, yoj, derog, delinq, cl
123
  bg_gradient = "linear-gradient(135deg, #3b1111 0%, #1a0505 100%)"
124
  accent_color = "#f87171"
125
  status_text = "❌ RISK ASSESSMENT: APPLICATION REJECTED"
126
- description = "The calculated default risk probability exceeds baseline security margins. Financial metrics point to high default interaction patterns."
127
  else:
128
  bg_gradient = "linear-gradient(135deg, #062411 0%, #021207 100%)"
129
  accent_color = "#4ade80"
130
  status_text = "✅ RISK ASSESSMENT: APPLICATION APPROVED"
131
- description = "This financial profile displays solid asset coverage, highly stable parameters, and strong structural repayment probabilities."
132
 
133
  metrics_html = f"""
134
  <div style="background: {bg_gradient}; color: #ffffff; padding: 24px; border-radius: 12px; border: 1px solid #334155; box-shadow: 0 4px 10px rgba(0,0,0,0.45);">
@@ -204,7 +216,6 @@ with io.Blocks() as app:
204
  ninq = io.Number(label="Recent Bureau Inquiries [NINQ]", value=1)
205
  clno = io.Number(label="Total Active Credit Lines [CLNO]", value=20)
206
 
207
- # Replaced io.Spacer() with direct margin styling configuration inside button
208
  submit_btn = io.Button("⚡ Run Stacking Underwriting Evaluation", variant="primary")
209
 
210
  # Output Feed Hub Panel (Right Side)
@@ -227,5 +238,4 @@ with io.Blocks() as app:
227
  )
228
 
229
  if __name__ == "__main__":
230
- # Correctly passes title and css keywords inside the .launch constructor to support Gradio 6.0+
231
  app.launch(css=custom_css, title="AI Loan Underwriting Portal", share=True)
 
21
  DATA_PATH = "loan_dataset.csv"
22
 
23
  # =====================================================================
24
+ # HIGH-SPEED CORE TRAINING FUNCTION
25
  # =====================================================================
26
+ def train_and_cache_pipeline():
27
+ print("Initializing ultra-fast pipeline training framework using current local environment...")
28
 
29
  if not os.path.exists(DATA_PATH):
30
+ raise FileNotFoundError(f"Missing '{DATA_PATH}'. Please ensure your dataset is placed in the same repository directory.")
31
 
32
  df = pd.read_csv(DATA_PATH)
33
  X = df.drop('BAD', axis=1)
34
  y = df['BAD']
35
 
36
+ # Explicit layout matching to prevent Pandas deprecation warnings
37
  num_cols = X.select_dtypes(include=['int64', 'float64']).columns.tolist()
38
  cat_cols = X.select_dtypes(include=['object', 'string']).columns.tolist()
39
 
 
53
 
54
  X_cleaned = cap_outliers_iqr(X, num_cols)
55
 
56
+ # Fast Feature Transformation Pipelines
57
  num_pipeline = Pipeline([
58
  ('imputer', KNNImputer(n_neighbors=5)),
59
  ('scaler', StandardScaler())
 
69
  ('cat', cat_pipeline, cat_cols)
70
  ])
71
 
72
+ # Optimized hyper-parameters for near-instant execution
73
  rf_base = RandomForestClassifier(n_estimators=50, max_depth=8, random_state=42, n_jobs=-1)
74
  xgb_base = XGBClassifier(n_estimators=50, learning_rate=0.1, max_depth=4, eval_metric='logloss', random_state=42, n_jobs=-1)
75
  voting_base = VotingClassifier(estimators=[('rf_v', rf_base), ('xgb_v', xgb_base)], voting='soft')
76
 
77
+ # Meta Stacking Level Layout
78
  stacking_ensemble = StackingClassifier(
79
  estimators=[('rf', rf_base), ('xgb', xgb_base), ('voting', voting_base)],
80
  final_estimator=LogisticRegression(max_iter=500, C=0.1),
81
+ cv=2, # Streamlines cross-validation execution speed down to seconds
82
  n_jobs=-1
83
  )
84
 
85
+ pipeline = ImbPipeline([
86
  ('preprocessing', preprocessor),
87
  ('smote', SMOTE(random_state=42, sampling_strategy=0.6)),
88
  ('ensemble_model', stacking_ensemble)
89
  ])
90
 
91
+ pipeline.fit(X_cleaned, y)
92
+ joblib.dump(pipeline, MODEL_PATH)
93
+ print("Model Stacking Pipeline Compiled and Cached Successfully!")
94
+ return pipeline
95
 
96
+ # =====================================================================
97
+ # FAULT-TOLERANT SNAPSHOT INITIALIZATION (Fixes the Unpickling Error)
98
+ # =====================================================================
99
+ if os.path.exists(MODEL_PATH):
100
+ try:
101
+ print("Pre-trained file located. Loading parameters...")
102
+ production_pipeline = joblib.load(MODEL_PATH)
103
+ print("Pipeline imported successfully!")
104
+ except Exception as error_context:
105
+ print(f"Version warning/mismatch detected during unpickling: {error_context}")
106
+ print("Bypassing incompatible snapshot file to initiate seamless automated fresh compile...")
107
+ production_pipeline = train_and_cache_pipeline()
108
  else:
109
+ production_pipeline = train_and_cache_pipeline()
 
110
 
111
 
112
  # =====================================================================
 
135
  bg_gradient = "linear-gradient(135deg, #3b1111 0%, #1a0505 100%)"
136
  accent_color = "#f87171"
137
  status_text = "❌ RISK ASSESSMENT: APPLICATION REJECTED"
138
+ description = "The calculated default risk probability exceeds security margins. Financial variables present elevated interaction risk patterns."
139
  else:
140
  bg_gradient = "linear-gradient(135deg, #062411 0%, #021207 100%)"
141
  accent_color = "#4ade80"
142
  status_text = "✅ RISK ASSESSMENT: APPLICATION APPROVED"
143
+ description = "This financial profile displays proper asset coverage stability, balanced leverage indexes, and strong structural repayment probabilities."
144
 
145
  metrics_html = f"""
146
  <div style="background: {bg_gradient}; color: #ffffff; padding: 24px; border-radius: 12px; border: 1px solid #334155; box-shadow: 0 4px 10px rgba(0,0,0,0.45);">
 
216
  ninq = io.Number(label="Recent Bureau Inquiries [NINQ]", value=1)
217
  clno = io.Number(label="Total Active Credit Lines [CLNO]", value=20)
218
 
 
219
  submit_btn = io.Button("⚡ Run Stacking Underwriting Evaluation", variant="primary")
220
 
221
  # Output Feed Hub Panel (Right Side)
 
238
  )
239
 
240
  if __name__ == "__main__":
 
241
  app.launch(css=custom_css, title="AI Loan Underwriting Portal", share=True)