add browse duration avg

src/models/esmm/diary_model.py

@@ -113,6 +113,14 @@ _float_columns = [
     # "sixty_reply_pure_rate",
     # "ninety_reply_pure_rate",
     # "history_reply_pure_rate",
+    "one_browse_duration_avg",
+    "three_browse_duration_avg",
+    "seven_browse_duration_avg",
+    "fifteen_browse_duration_avg",
+    "thirty_browse_duration_avg",
+    "sixty_browse_duration_avg",
+    "ninety_browse_duration_avg",
+    "history_browse_duration_avg",
 ]
 _categorical_columns = [
     "device_id", "past_consume_ability_history", "potential_consume_ability_history", "price_sensitive_history", "device_fd",

src/models/esmm/fe/diary_fe.py

@@ -102,6 +102,14 @@ DIARY_COLUMNS = [
     "sixty_share_num",
     "ninety_share_num",
     "history_share_num",
+    "one_browse_duration_avg",
+    "three_browse_duration_avg",
+    "seven_browse_duration_avg",
+    "fifteen_browse_duration_avg",
+    "thirty_browse_duration_avg",
+    "sixty_browse_duration_avg",
+    "ninety_browse_duration_avg",
+    "history_browse_duration_avg",
     "first_demands",
     "second_demands",
     "first_solutions",
@@ -219,6 +227,14 @@ FLOAT_COLUMNS = [
     # "sixty_reply_pure_rate",
     # "ninety_reply_pure_rate",
     # "history_reply_pure_rate",
+    "one_browse_duration_avg",
+    "three_browse_duration_avg",
+    "seven_browse_duration_avg",
+    "fifteen_browse_duration_avg",
+    "thirty_browse_duration_avg",
+    "sixty_browse_duration_avg",
+    "ninety_browse_duration_avg",
+    "history_browse_duration_avg",
 ]
 CATEGORICAL_COLUMNS = [
     "device_id", "active_type", "past_consume_ability_history", "potential_consume_ability_history", "price_sensitive_history",

src/train_diary.py

@@ -62,7 +62,7 @@ def main():
 
     model = tf.estimator.Estimator(model_fn=esmm_model_fn, params=params, model_dir=model_path, config=estimator_config)
     # TODO 50000
-    train_spec = tf.estimator.TrainSpec(input_fn=lambda: esmm_input_fn(train_df, shuffle=True), max_steps=20000)
+    train_spec = tf.estimator.TrainSpec(input_fn=lambda: esmm_input_fn(train_df, shuffle=True), max_steps=15000)
     eval_spec = tf.estimator.EvalSpec(input_fn=lambda: esmm_input_fn(val_df, shuffle=False))
     res = tf.estimator.train_and_evaluate(model, train_spec, eval_spec)
     print("@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@@")