Upload 2 files

evaluation.py

@@ -4,7 +4,6 @@ from sklearn.metrics import mean_squared_error, roc_auc_score
 from sklearn.feature_extraction.text import TfidfVectorizer
 from sklearn.metrics.pairwise import cosine_similarity
 
-from retrieval import query_dataset
 from data_processing import load_query_dataset
 
 global ground_truth_answer, ground_truth_metrics
@@ -93,6 +92,7 @@ def compute_rmse(predicted_values, ground_truth_values):
     return np.sqrt(mean_squared_error(ground_truth_values, predicted_values))
 
 def calculate_metrics(question, response, docs, time_taken):
+    from retrieval import query_dataset
     data = load_query_dataset(query_dataset)
     ground_truth_answer = retrieve_ground_truths(question, data)  # Store the ground truth answer
 

retrieval.py

@@ -9,9 +9,10 @@ from sentence_transformers import CrossEncoder
 reranker = CrossEncoder("cross-encoder/ms-marco-MiniLM-L-6-v2")
 
 retrieved_docs = None
+global query_dataset
 
 def retrieve_documents_hybrid(query, top_k=5):
-    global query_dataset
+    #global query_dataset
     query_dataset = find_query_dataset(query)
     
     with open( f"data_local/{query_dataset}_chunked_docs.json", "r") as f: