Update app.py

app.py

@@ -1,70 +1,41 @@
 import gradio as gr
-import pickle
 import pandas as pd
 import numpy as np
 from sklearn.metrics.pairwise import cosine_similarity
-from sentence_transformers import SentenceTransformer  # Explicit import
+from sentence_transformers import SentenceTransformer
 
-# Load model and data
-try:
-    # Load your trained SentenceTransformer model
-    with open("recommender_model.pkl", "rb") as f:
-        model = pickle.load(f)
-    
-    # Load posts dataset
-    posts_df = pd.read_csv("posts_cleaned.csv")  
-    post_texts = posts_df["post_text"].astype(str).tolist()
-    
-except Exception as e:
-    raise gr.Error(f"Error loading files: {str(e)}")
+# ✅ Correct way to load the model
+model = SentenceTransformer("recommender")  # Replace with your model name/path
+
+# Load posts dataset
+posts_df = pd.read_csv("posts_cleaned.csv")  
+post_texts = posts_df["post_text"].astype(str).tolist()
 
-# Cache embeddings in memory after first computation
+# Cache embeddings
 post_embeddings = None
 
 def get_embeddings():
     global post_embeddings
     if post_embeddings is None:
-        print("Computing embeddings for all posts...")
+        print("Computing embeddings...")
         post_embeddings = model.encode(post_texts, convert_to_tensor=False)
-        print("Embeddings computed!")
     return post_embeddings
 
-# Prediction function
 def recommend_from_input(user_text):
     if not user_text.strip():
         return []
     
-    # Get embeddings (computes only once)
     embeddings = get_embeddings()
-    
-    # Encode user input
     user_vec = model.encode([user_text])
-    
-    # Calculate similarities
     sims = cosine_similarity(user_vec, embeddings)[0]
-    top_idxs = sims.argsort()[-5:][::-1]  # Top 5 most similar posts
-    
-    # Return as list of strings
+    top_idxs = sims.argsort()[-5:][::-1]
     return posts_df.iloc[top_idxs]["post_text"].tolist()
 
-# Gradio Interface
 interface = gr.Interface(
     fn=recommend_from_input,
-    inputs=gr.Textbox(label="What are you interested in?", placeholder="e.g. Web3 security tips"),
-    outputs=gr.Dataframe(
-        headers=["Recommended Posts"],
-        datatype=["str"],
-        col_count=(1, "fixed")
-    ),
-    title="🔍 AI Content Recommender",
-    description="Enter a topic or interest to get personalized post recommendations",
-    examples=[
-        ["Blockchain scalability solutions"],
-        ["Latest breakthroughs in AI"],
-        ["How to write smart contracts"]
-    ],
-    allow_flagging="never"
+    inputs=gr.Textbox(label="What are you interested in?"),
+    outputs=gr.Dataframe(headers=["Recommended Posts"]),
+    title="🔍 AI Content Recommender"
 )
 
-# Launch with queue for stability
 interface.launch(share=False)