Update app.py

app.py

@@ -1,9 +1,4 @@
 import os
-os.environ["HF_HOME"] = "/tmp/.cache"
-os.environ["HF_DATASETS_CACHE"] = "/tmp/.cache"
-os.environ["SENTENCE_TRANSFORMERS_HOME"] = "/tmp/.cache"
-os.makedirs("/tmp/.cache", exist_ok=True)
-
 import json
 from sentence_transformers import SentenceTransformer
 from sklearn.metrics.pairwise import cosine_similarity
@@ -12,14 +7,22 @@ from huggingface_hub import upload_file, hf_hub_download, InferenceClient
 from flask import Flask, request, jsonify
 import time
 
-embedding_model = SentenceTransformer('paraphrase-mpnet-base-v2')
 
+os.environ["HF_HOME"] = "/tmp/.cache"
+os.environ["HF_DATASETS_CACHE"] = "/tmp/.cache"
+os.environ["SENTENCE_TRANSFORMERS_HOME"] = "/tmp/.cache"
+os.makedirs("/tmp/.cache", exist_ok=True)
+os.makedirs("/tmp/outputs", exist_ok=True)
+
+
+embedding_model = SentenceTransformer('paraphrase-mpnet-base-v2')
 token = os.getenv("HF_TOKEN") or os.getenv("NEW_PUP_AI_Project")
 inference_client = InferenceClient(
     model="mistralai/Mixtral-8x7B-Instruct-v0.1",
     token=token
 )
 
+
 BASE_DIR = os.path.dirname(os.path.abspath(__file__))
 DATASET_PATH = os.path.join(BASE_DIR, "dataset.json")
 with open(DATASET_PATH, "r") as f:
@@ -29,14 +32,17 @@ questions = [item["question"] for item in dataset]
 answers = [item["answer"] for item in dataset]
 question_embeddings = embedding_model.encode(questions, convert_to_tensor=True)
 
-chat_history = []
+
 feedback_data = []
 feedback_questions = []
 feedback_embeddings = None
-dev_mode = {"enabled": False} 
-
+dev_mode = {"enabled": False}
 feedback_path = "/tmp/outputs/feedback.json"
-os.makedirs("/tmp/outputs", exist_ok=True)
+COMMENTS_PATH = "/tmp/outputs/Comments.json"
+
+if not os.path.exists(COMMENTS_PATH):
+    with open(COMMENTS_PATH, "w") as f:
+        json.dump([], f, indent=4)
 
 try:
     hf_token = os.getenv("NEW_PUP_AI_Project")
@@ -60,28 +66,28 @@ except Exception as e:
     feedback_data = []
 
 
-def upload_feedback_to_hf():
+def upload_file_to_hf(local_path, remote_filename):
+    """Helper to upload any file to Hugging Face dataset repo."""
     hf_token = os.getenv("NEW_PUP_AI_Project")
     if not hf_token:
         raise ValueError("Hugging Face token not found in environment variables!")
 
     try:
         upload_file(
-            path_or_fileobj=feedback_path,
-            path_in_repo="feedback.json",
+            path_or_fileobj=local_path,
+            path_in_repo=remote_filename,
             repo_id="oceddyyy/University_Inquiries_Feedback",
             repo_type="dataset",
             token=hf_token
         )
-        print("Feedback uploaded to Hugging Face successfully.")
+        print(f"{remote_filename} uploaded to Hugging Face successfully.")
     except Exception as e:
-        print(f"Error uploading feedback to HF: {e}")
+        print(f"Error uploading {remote_filename} to HF: {e}")
 
 
 def chatbot_response(query, dev_mode_flag):
     query_embedding = embedding_model.encode([query], convert_to_tensor=True)
 
-    # Feedback check
     if feedback_embeddings is not None:
         feedback_scores = cosine_similarity(query_embedding.cpu().numpy(), feedback_embeddings.cpu().numpy())[0]
         best_idx = int(np.argmax(feedback_scores))
@@ -97,7 +103,6 @@ def chatbot_response(query, dev_mode_flag):
         if best_score >= dynamic_threshold:
             return matched_feedback["response"], "Feedback", 0.0
 
-    # Handbook retrieval
     similarity_scores = cosine_similarity(query_embedding.cpu().numpy(), question_embeddings.cpu().numpy())[0]
     top_k = 3
     top_k_indices = np.argsort(similarity_scores)[-top_k:][::-1]
@@ -113,6 +118,7 @@ def chatbot_response(query, dev_mode_flag):
         context = ""
         for i, item in enumerate(top_k_items):
             context += f"Relevant info #{i+1} (score: {top_k_scores[i]:.2f}):\n\"{item.get('answer', '')}\"\n\n"
+
         prompt = (
             f"You are an expert university assistant. "
             f"A student asked: \"{query}\"\n"
@@ -133,14 +139,13 @@ def chatbot_response(query, dev_mode_flag):
                 llm_response = inference_client.chat_completion(
                     messages=conversation,
                     model="mistralai/Mixtral-8x7B-Instruct-v0.1",
-                    max_tokens=200,   
+                    max_tokens=200,
                     temperature=0.7
                 )
                 if isinstance(llm_response, dict) and "choices" in llm_response:
                     response = llm_response["choices"][0]["message"]["content"]
                 elif hasattr(llm_response, "generated_text"):
                     response = llm_response.generated_text
-
             else:
                 llm_response = inference_client.text_generation(
                     prompt,
@@ -175,7 +180,8 @@ def chatbot_response(query, dev_mode_flag):
     return response.strip(), matched_source, 0.0
 
 
-def record_feedback(feedback, query, response):
+def record_feedback(feedback, query, response, comment=None):
+    """Records user feedback and optional comment."""
     global feedback_embeddings, feedback_questions
     matched = False
     new_embedding = embedding_model.encode([query], convert_to_tensor=True)
@@ -191,8 +197,8 @@ def record_feedback(feedback, query, response):
 
     if not matched:
         entry = {
-            "question": query,
-            "response": response,
+            "question": query,    
+            "response": response,    
             "feedback": feedback,
             "upvotes": 1 if feedback == "positive" else 0,
             "downvotes": 1 if feedback == "negative" else 0
@@ -206,9 +212,30 @@ def record_feedback(feedback, query, response):
     if feedback_questions:
         feedback_embeddings = embedding_model.encode(feedback_questions, convert_to_tensor=True)
 
-    upload_feedback_to_hf()
+    upload_file_to_hf(feedback_path, "feedback.json")
 
 
+    if comment and comment.strip():
+        try:
+            with open(COMMENTS_PATH, "r") as f:
+                comments_list = json.load(f)
+        except json.JSONDecodeError:
+            comments_list = []
+
+        comment_entry = {
+            "timestamp": time.strftime("%Y-%m-%d %H:%M:%S", time.localtime()),
+            "question": query,
+            "response": response,
+            "feedback": feedback,
+            "comment": comment.strip()
+        }
+        comments_list.append(comment_entry)
+
+        with open(COMMENTS_PATH, "w") as f:
+            json.dump(comments_list, f, indent=4)
+
+        upload_file_to_hf(COMMENTS_PATH, "Comments.json")
+
 app = Flask(__name__)
 
 @app.route("/api/chat", methods=["POST"])
@@ -220,21 +247,19 @@ def chat():
     response, source, elapsed = chatbot_response(query, dev)
     return jsonify({"response": response, "source": source, "response_time": elapsed})
 
-
 @app.route("/api/feedback", methods=["POST"])
 def feedback():
     data = request.json
     query = data.get("query", "")
     response = data.get("response", "")
     feedback_type = data.get("feedback", "")
-    record_feedback(feedback_type, query, response)
+    comment = data.get("comment", None)
+    record_feedback(feedback_type, query, response, comment)
     return jsonify({"status": "success"})
 
-
 @app.route("/", methods=["GET"])
 def index():
-    return "University Inquiries AI Chatbot API. Use POST /chat or /feedback.", 200
-
+    return "University Inquiries AI Chatbot API. Use POST /api/chat or /api/feedback.", 200
 
 if __name__ == "__main__":
-    app.run(host="0.0.0.0", port=7861)
+    app.run(host="0.0.0.0", port=7861)