feat(rec): Finish feedback + nickname; small fixes

app.py

@@ -1,17 +1,22 @@
 import json
 import os
 import random
+import signal
+import sys
 import urllib.parse
+from datetime import datetime
 from pathlib import Path
 from typing import Optional
+from uuid import uuid4
 
 import gradio as gr
 import numpy as np
 import pandas as pd
-from dotenv import load_dotenv
+# from dotenv import load_dotenv
 from fastembed import SparseEmbedding, SparseTextEmbedding
 from google import genai
 from google.genai import types
+from huggingface_hub import CommitScheduler
 from pydantic import BaseModel, Field
 from qdrant_client import QdrantClient
 from qdrant_client import models as qmodels
@@ -19,7 +24,7 @@ from sentence_transformers import CrossEncoder, SentenceTransformer
 from vllm import LLM, SamplingParams
 from vllm.sampling_params import GuidedDecodingParams
 
-load_dotenv()
+# load_dotenv()
 
 HF_TOKEN = os.getenv("HF_TOKEN")
 
@@ -33,6 +38,21 @@ GEMINI_API_KEY = os.getenv("GEMINI_API_KEY")
 DATA_PATH = Path(os.getenv("DATA_PATH"))
 DB_PATH = DATA_PATH / "db"
 
+FEEDBACK_REPO = os.getenv("FEEDBACK_REPO")
+FEEDBACK_DIR  = DATA_PATH / "feedback"
+FEEDBACK_DIR.mkdir(parents=True, exist_ok=True)
+FEEDBACK_FILE = FEEDBACK_DIR / f"votes_{uuid4()}.jsonl"
+
+scheduler = CommitScheduler(
+    repo_id=FEEDBACK_REPO,
+    repo_type="dataset",
+    folder_path=FEEDBACK_DIR,
+    path_in_repo="data",
+    every=5,
+    token=HF_TOKEN,
+    private=True,
+)
+
 client = QdrantClient(path=str(DB_PATH))
 collection_name = "knowledge_cards"
 num_chunks_base = 500
@@ -64,12 +84,21 @@ def generate_recommendation_string(ranking: dict[str, float]) -> str:
     for idx, (genre, score) in enumerate(ranking.items(), start=1):
         youtube_link = youtube_search_link_for_genre(genre=genre)
         recommendation_string += (
-            f"{idx}. **{genre.replace('_', ' ').capitalize()}** ({score:.2f}); "
+            f"{idx}. **{genre.replace('_', ' ').capitalize()}**; "
             f"[YouTube link]({youtube_link})\n"
         )
     return recommendation_string
 
 
+def graceful_shutdown(signum, frame):
+    print(f"{signum} received - flushing feedback …", flush=True)
+    scheduler.trigger().result()
+    sys.exit(0)
+
+
+signal.signal(signal.SIGTERM, graceful_shutdown)
+signal.signal(signal.SIGINT, graceful_shutdown)
+
 # --------------------------------  Data Models  -------------------------------
 class StructuredQueryRewriteResponse(BaseModel):
     general: str | None
@@ -396,11 +425,13 @@ def recommend_gemini(query: str):
 # -------------------------------------- INTERFACE -----------------------------
 pipelines = {
     "sadaimrec": recommend_sadaimrec,
-    "chatgpt": recommend_gemini,
+    "gemini": recommend_gemini,
 }
 
 
 def generate_responses(query):
+    if not query.strip():
+        raise gr.Error("Please enter a query before submitting.")
     # Randomize model order
     pipeline_names = list(pipelines.keys())
     random.shuffle(pipeline_names)
@@ -414,13 +445,39 @@ def generate_responses(query):
 
 
 # Callback to capture vote
-def handle_vote(selected, label1, label2, resp1, resp2):
-    chosen_name = label1 if selected == "Option 1" else label2
-    chosen_resp = resp1 if selected == "Option 1" else resp2
-    print(f"User voted for {chosen_name}: '{chosen_resp}'")
+def handle_vote(nickname, query, selected, label1, label2, resp1, resp2):
+    nick = nickname.strip() or uuid4().hex[:8]
+    winner_name, loser_name = (
+        (label1, label2) if selected == "Option 1 (left)" else (label2, label1)
+    )
+    winner_resp, loser_resp = (
+        (resp1, resp2) if selected == "Option 1 (left)" else (resp2, resp1)
+    )
+    print(
+        (
+            f"User voted:\nwinner = {winner_name}: {winner_resp};"
+            f" loser = {loser_name}: {loser_resp}"
+        ),
+        flush=True,
+    )
+
+    # ---------- persist feedback locally ----------
+    entry = {
+        "ts": datetime.now().isoformat(timespec="seconds") + "Z",
+        "nickname": nick,
+        "query": query,
+        "winner": winner_name,
+        "loser": loser_name,
+        "winner_response": winner_resp,
+        "loser_response": loser_resp,
+    }
+    with FEEDBACK_FILE.open("a", encoding="utf-8") as f:
+        f.write(json.dumps(entry) + "\n")
+
     return (
-        "Thank you for your vote! Restarting in 2 seconds...",
+        f"Thank you for your vote! Winner: {winner_name}. Restarting in 3 seconds...",
         gr.update(active=True),
+        gr.update(value=nick),
     )
 
 
@@ -432,6 +489,8 @@ def reset_ui():
         gr.update(visible=False),  # hide vote button
         gr.update(value="**Generating...**"),  # clear Option 1 text
         gr.update(value="**Generating...**"),  # clear Option 2 text
+        gr.update(value=""),  # clear Model Label 1 text
+        gr.update(value=""),  # clear Model Label 2 text
         gr.update(value=""),  # clear result
         gr.update(active=False),
     )
@@ -445,13 +504,17 @@ with gr.Blocks(
     gr.Markdown(app_description)
     with gr.Accordion("Detailed usage instructions", open=False):
         gr.Markdown(app_instructions)
+    nickname = gr.Textbox(
+        label="Your nickname",
+        placeholder="Leave empty to generate a random nickname on first vote within session",
+    )
     query = gr.Textbox(
         label="Your Query",
         placeholder="Calming, music for deep relaxation with echoing sounds and deep bass",
     )
     submit_btn = gr.Button("Submit")
     # timer that resets ui after feedback is sent
-    reset_timer = gr.Timer(value=2.0, active=False)
+    reset_timer = gr.Timer(value=3.0, active=False)
 
     # Hidden components to store model responses and names
     with gr.Row(visible=False) as response_row:
@@ -489,8 +552,16 @@ with gr.Blocks(
     # Feedback handling
     vote_btn.click(
         fn=handle_vote,
-        inputs=[vote, model_label_1, model_label_2, response_1, response_2],
-        outputs=[result, reset_timer],
+        inputs=[
+            nickname,
+            query,
+            vote,
+            model_label_1,
+            model_label_2,
+            response_1,
+            response_2,
+        ],
+        outputs=[result, reset_timer, nickname],
     )
     reset_timer.tick(
         fn=reset_ui,
@@ -502,6 +573,8 @@ with gr.Blocks(
             vote_btn,
             response_1,
             response_2,
+            model_label_1,
+            model_label_2,
             result,
             reset_timer,
         ],