Integrated Google Sheets cloud logging

requirements.txt

@@ -4,4 +4,5 @@ torch
 torchvision
 safetensors
 pandas
+st-gsheets-connection
 plotly

src/streamlit_app.py

@@ -1,161 +1,114 @@
 import streamlit as st
 from transformers import pipeline
-import os
 import pandas as pd
 from datetime import datetime
-import time
 import plotly.graph_objects as go
+from streamlit_gsheets import GSheetsConnection
 
 # --- PAGE CONFIG ---
 st.set_page_config(page_title="Sentiment Analyzer AI | Bilingual Engine", page_icon="🌐", layout="wide")
 
-# --- PROFESSIONAL NEUMORPHIC / GLASS CSS ---
+# --- CUSTOM CSS FOR GLASS-CARD LOOK ---
 st.markdown("""
     <style>
-    .stApp {
-        background: linear-gradient(135deg, #12141d 0%, #1a1c2c 100%);
-        color: #ffffff;
-    }
-    div[data-baseweb="input"] {
-        background: rgba(255, 255, 255, 0.05) !important;
-        backdrop-filter: blur(10px) !important;
-        border-radius: 15px !important;
-        border: 1px solid rgba(255, 255, 255, 0.1) !important;
-        padding: 5px !important;
-    }
+    .stApp { background: linear-gradient(135deg, #12141d 0%, #1a1c2c 100%); color: #ffffff; }
     .glass-card {
         background: rgba(255, 255, 255, 0.05);
         backdrop-filter: blur(10px);
         border-radius: 20px;
         border: 1px solid rgba(255, 255, 255, 0.1);
-        padding: 30px;
-        margin-top: 20px;
-        margin-bottom: 25px;
+        padding: 25px;
+        margin-bottom: 20px;
     }
     .stButton>button {
         background: linear-gradient(90deg, #4facfe 0%, #00f2fe 100%);
-        color: white;
-        border-radius: 12px;
-        font-weight: 600;
-        height: 3rem;
+        color: white; border-radius: 12px; font-weight: 600; width: 100%;
     }
-    [data-testid="stMetricValue"] { color: #00f2fe; font-weight: 800; }
     </style>
     """, unsafe_allow_html=True)
 
-# --- DATA LOGGING ---
-FEEDBACK_FILE = "sentiment_feedback.csv"
-
-def save_smart_data(text, ai_label, ai_score, corrected_label=None):
-    needs_review = "YES" if 0.33 <= ai_score <= 0.65 else "NO"
-    new_data = {
-        "Timestamp": [datetime.now().strftime("%Y-%m-%d %H:%M:%S")],
-        "Text": [text],
-        "AI_Label": [ai_label],
-        "Confidence": [round(ai_score, 4)],
-        "Needs_Review": [needs_review],
-        "Corrected_Label": [corrected_label if corrected_label else ai_label]
-    }
-    df = pd.DataFrame(new_data)
-    if not os.path.isfile(FEEDBACK_FILE):
-        df.to_csv(FEEDBACK_FILE, index=False)
-    else:
-        df.to_csv(FEEDBACK_FILE, mode='a', header=False, index=False)
-
-# --- MODEL ENGINE PATH FIX ---
-import streamlit as st
-from transformers import pipeline
+# --- CLOUD DATA LOGGING (GOOGLE SHEETS) ---
+# This connects to the sheet you shared with the service account
+conn = st.connection("gsheets", type=GSheetsConnection)
 
-# This MUST match your model repo name from Screenshot 281
+def save_to_cloud(text, ai_label, ai_score, corrected_label=None):
+    try:
+        # Read existing data to append
+        existing_data = conn.read(worksheet="Sheet1", ttl=0)
+    except:
+        existing_data = pd.DataFrame()
+
+    new_entry = pd.DataFrame([{
+        "Timestamp": datetime.now().strftime("%Y-%m-%d %H:%M:%S"),
+        "Text": text,
+        "AI_Label": ai_label,
+        "Confidence": f"{ai_score:.2%}",
+        "Correction": corrected_label if corrected_label else "N/A"
+    }])
+
+    updated_df = pd.concat([existing_data, new_entry], ignore_index=True)
+    conn.update(worksheet="Sheet1", data=updated_df)
+
+# --- MODEL ENGINE ---
 MODEL_REPO = "SumedhGajbhiye/Sentiment-Analyzer"
 
 @st.cache_resource
 def load_analysis_engine():
     try:
-        # This tells the app to download weights from the Hub, not the local folder
+        # Pulls directly from your HF Model Repo
         return pipeline("sentiment-analysis", model=MODEL_REPO, tokenizer=MODEL_REPO)
     except Exception as e:
-        st.error(f"Could not connect to System Core: {e}")
+        st.error(f"System Core Connection Failed: {e}")
         return None
 
-# IMPORTANT: Find and delete the line in your code that says:
-# if not os.path.exists("model.safetensors"): st.error("System Core... not found")
-
-
-
-# --- HEADER SECTION ---
-st.title("Sentiment Analyzer")
-st.caption("Advanced Bilingual Sentiment Analysis for English, Hindi & Hinglish")
-
-# --- SIDEBAR HUD ---
-with st.sidebar:
-    st.markdown("### 🛠️ ENGINE STATUS")
-    if os.path.exists(FEEDBACK_FILE):
-        df_log = pd.read_csv(FEEDBACK_FILE)
-        st.metric("Total Ingested", len(df_log))
-        st.metric("Anomalies Flagged", len(df_log[df_log['Needs_Review'] == 'YES']))
-        st.download_button("📤 Export Dataset", df_log.to_csv(index=False), "engine_feedback.csv", "text/csv")
-    else:
-        st.info("Waiting for input...")
+# --- UI HEADER ---
+st.title("🌐 Sentiment Analyzer")
+st.caption("Bilingual Engine: English | Hindi | Hinglish")
 
-# --- ANALYSIS INTERFACE ---
-# 1. We call the cached loader (it handles the Hub download automatically)
 classifier = load_analysis_engine()
 
-if classifier is None:
-    st.error("System Core could not be initialized. Please check the logs.")
-else:
-    user_input = st.text_input(
-        "QUERY INPUT:", 
-        placeholder="Enter sentence (English/Hindi/Hinglish)...", 
-        key="main_input"
-    )
+if classifier:
+    user_input = st.text_input("QUERY INPUT:", placeholder="Type a sentence...", key="main_input")
 
     if user_input:
-        with st.status("Neural Scan in Progress...", expanded=False) as status:
-            # Running the inference
+        with st.spinner("Neural Scan..."):
             result = classifier(user_input)[0]
-            status.update(label="Analysis Complete", state="complete")
-
-        label = result['label']
-        score = result['score']
-        
-        # Mapping labels (Adjust if your XLM-Roberta uses LABEL_0/1/2)
-        emoji_map = {"POSITIVE": "🟢", "NEUTRAL": "🟡", "NEGATIVE": "🔴"}
-        
-        # Color logic
-        color = "#00ff88" if "POS" in label.upper() else "#ff4b4b" if "NEG" in label.upper() else "#ffaa00"
-
-        st.markdown(f'''
-            <div class="glass-card">
-                <h4 style="color: #888; margin:0;">RESULT</h4>
-                <h1 style="color: {color}; margin:0; font-size: 3rem;">{label} {emoji_map.get(label.upper(), "")}</h1>
-                <p style="color: #aaa;">Confidence: {score:.1%}</p>
-            </div>
-            ''', unsafe_allow_html=True)
-
-        # Gauge Chart
-        fig = go.Figure(go.Indicator(
-            mode = "gauge+number",
-            value = score * 100,
-            gauge = {'axis': {'range': [0, 100]}, 'bar': {'color': color}}
-        ))
-        fig.update_layout(height=250, paper_bgcolor='rgba(0,0,0,0)', font={'color': "#fff"})
-        st.plotly_chart(fig, use_container_width=True)
-
-        # Human Verification
-        c1, c2 = st.columns(2)
-        with c1:
-            if st.button("CONFIRM ACCURACY"):
-                save_smart_data(user_input, label, score)
-                st.toast("Saved!")
-        with c2:
-            correction = st.selectbox("CORRECT LABEL:", ["Positive", "Neutral", "Negative"])
-            if st.button("SAVE CORRECTION"):
-                save_smart_data(user_input, label, score, corrected_label=correction)
-                st.toast("Correction logged!")
-
-# --- LOGS ---
-if os.path.exists(FEEDBACK_FILE):
-    with st.expander("📂 VIEW RECENT LOGS"):
-        st.dataframe(pd.read_csv(FEEDBACK_FILE).tail(5), use_container_width=True)
+            label, score = result['label'], result['score']
+            color = "#00ff88" if "POS" in label.upper() else "#ff4b4b" if "NEG" in label.upper() else "#ffaa00"
+
+        # --- SIDE-BY-SIDE LAYOUT ---
+        # 1.5 ratio for left makes the gauge much bigger
+        col_left, col_right = st.columns([1.5, 1], gap="large")
+
+        with col_left:
+            st.markdown(f'<div class="glass-card"><h4>RESULT: <span style="color:{color}">{label}</span></h4></div>', unsafe_allow_html=True)
+            
+            # Big Professional Gauge
+            fig = go.Figure(go.Indicator(
+                mode = "gauge+number",
+                value = score * 100,
+                number = {'suffix': "%", 'font': {'color': '#fff'}},
+                gauge = {
+                    'axis': {'range': [0, 100], 'tickcolor': "#888"},
+                    'bar': {'color': color},
+                    'bgcolor': "rgba(0,0,0,0)",
+                    'borderwidth': 2,
+                    'bordercolor': "rgba(255,255,255,0.1)"
+                }
+            ))
+            fig.update_layout(height=450, margin=dict(t=50, b=0), paper_bgcolor='rgba(0,0,0,0)', font={'color': "#fff"})
+            st.plotly_chart(fig, use_container_width=True)
+
+        with col_right:
+            st.markdown("### 🛠️ Neural Feedback")
+            st.info("Human verification helps the engine learn mixed-language nuances.")
+            
+            if st.button("✅ CONFIRM ACCURACY"):
+                save_to_cloud(user_input, label, score)
+                st.toast("Success! Verified data sent to Google Sheets.")
+
+            st.markdown("---")
+            correction = st.selectbox("UNCERTAIN? CORRECT LABEL:", ["Positive", "Neutral", "Negative"])
+            if st.button("💾 SAVE CORRECTION"):
+                save_to_cloud(user_input, label, score, corrected_label=correction)
+                st.warning("Correction logged for retraining.")