update

Author: jefcoder

reinforcement_learning/qlearn_agent.py

@@ -1,12 +1,11 @@
 import os
 import pickle
 import numpy as np
-# Added imports for the neural network mode
 import tensorflow as tf
 from tensorflow.keras.models import Sequential, load_model
 from tensorflow.keras.layers import Dense
 from tensorflow.keras.optimizers import Adam
-import logging # Added for cleaner messages
+import logging 
 
 # Configure logging
 logging.basicConfig(level=logging.INFO, format='%(asctime)s - %(levelname)s - %(message)s')
@@ -67,7 +66,7 @@ def __init__(
         self.error_list = []
         self.load_data()
 
-    # +++ Added: Helper to build the neural network +++
+
     def _build_model(self):
         """Builds the neural network model for DQN."""
 
@@ -83,7 +82,6 @@ def _build_model(self):
         logging.info("Built neural network model.")
         # model.summary() # Optional: print model summary
         return model
-    # +++ End Added +++
 
     def load_data(self):
         """

tools/rag/rl_meta_rag/rl_meta_rag_retrieve.py

@@ -11,7 +11,6 @@
 # Helper function to extract categories from prompt template
 def _extract_categories_from_prompt(prompt_template, feature_name):
     """Extracts list items for a given feature from the prompt template."""
-    # Find the line describing the feature
     match = re.search(rf"- {feature_name}:.*?\[(.*?)\]", prompt_template.template, re.DOTALL)
     if match:
         # Extract the content within the brackets and split into items
@@ -131,11 +130,6 @@ def _extract_query_features(self, query):
             # Adjust query length specifically if needed
             features[5] = len(query.split()) # A reasonable default for query_length
 
-        # Note: The original code appended query_length again here.
-        # This was likely a bug as query_length is already included in the 9 features.
-        # query_length_calc = len(query.split())
-        # features.append(query_length_calc) # REMOVED THIS LINE
-
         self.socketio.emit('reasoning_update', {
             "message": f"Extracted query features: {features}"
         })
@@ -240,7 +234,6 @@ def retrieve(self, query, session_id=None):
         if technique_id in self.rag_techniques:
             result = self.rag_techniques[technique_id](query)
         else:
-            # Default or fallback technique
             self.socketio.emit('reasoning_update', {
                 "message": f"Warning: Selected technique ID {technique_id} not found. Defaulting to technique 0."
             })
@@ -262,12 +255,12 @@ def retrieve(self, query, session_id=None):
             redis_client = redis.StrictRedis(host='redis', port=6379, db=0, decode_responses=True)
             # Store the original feature list before tuple conversion if needed
             data_to_store = {
-                "state": state_features, # Store the raw feature list
+                "state": state_features,
                 "action": technique_id,
                 "query": query
             }
             redis_key = f"rl_update:{session_id}"
-            # Use json.dumps with a handler for non-serializable types if necessary, though list should be fine
+
             try:
                 redis_client.set(redis_key, json.dumps(data_to_store))
                 self.socketio.emit('reasoning_update', {