updated readme and app

README.md

@@ -11,7 +11,7 @@ pinned: false
 
 # 🚀 Social Media Virality Assistant
 
-A machine learning-powered tool that helps content creators predict and optimize their video virality potential using **XGBoost** and **Google Gemini AI**.
+A machine learning-powered tool that helps content creators predict and optimize their video virality potential using trained **XGBoost** model and **Google Gemini AI**.
 
 ## 🏗️ Architecture & Pipeline
 
@@ -42,7 +42,7 @@ The `app.py` script runs a **Gradio** web interface that pulls artifacts from th
     *   Generates embeddings on-the-fly for the Knowledge Base.
 2.  **Core Features**:
     *   **Virality Prediction**: Predicts raw view counts based on your draft description and stats.
-    *   **AI Optimization**: Uses **Google Gemini** to rewrite your description with viral hooks and hashtags.
+    *   **AI Optimization**: Uses **Google Gemini** to rewrite your description with viral hooks and hashtags with the context of top 3 similar videos from the dataset.
     *   **Semantic Search**: Finds similar successful videos from the knowledge base using Cosine Similarity.
 
 ---
@@ -70,22 +70,3 @@ This project is built using:
 
 ---
 
-## 🚀 How to Run
-
-1.  **Install Dependencies**:
-    ```bash
-    pip install -r requirements.txt
-    ```
-
-2.  **Run the App**:
-    The app will automatically download the necessary data and models from Hugging Face.
-    ```bash
-    export GEMINI_API_KEY="your_api_key_here"
-    python app.py
-    ```
-
-3.  **(Optional) Retrain the Model**:
-    If you want to re-run the benchmarks and training using the latest data:
-    ```bash
-    python model-prep.py
-    ```

app.py

@@ -38,7 +38,7 @@ def initialize_app():
     except Exception as e:
         raise RuntimeError(f"Failed to load dataset from HF: {e}")
     
-    # 2. FIT ENCODERS (Crucial for Feature Consistency)
+    # 2. FIT ENCODERS (For Feature Consistency)
     print("🔤 Fitting Label Encoders...")
     cat_cols = ['category', 'gender', 'day_of_week', 'age']
     for c in cat_cols:
@@ -76,7 +76,7 @@ def initialize_app():
     device = "mps" if torch.backends.mps.is_available() else "cpu"
     st_model = SentenceTransformer(embedding_model_name, device=device)
     
-    # 5. Generate Embeddings (On-the-fly)
+    # 5. Generate Embeddings 
     print("⚙️ Generating Embeddings for Knowledge Base (One-time setup)...")
     if 'embedding' not in knowledge_df.columns:
         embeddings = st_model.encode(knowledge_df['description'].fillna("").tolist(), 
@@ -257,12 +257,12 @@ with gr.Blocks(theme=gr.themes.Soft()) as demo:
 
     # Demo Button Logic
     demo_btn.click(
-        fn=lambda: ("My protein shake recipe ended up on the floor 😭 #gymfail", 15, 19, "Monday", "Fitness", 50000, "18-24", "Male"),
+        fn=lambda: ("My protein shake ended up on the floor", 15, 19, "Monday", "Fitness", 50000, "18-24", "Male"),
         inputs=None,
         outputs=[input_text, duration_slider, hour_slider, day_dropdown, category, followers, age, gender]
     ).then(
         fn=predict_and_optimize,
-        inputs=[gr.State("My protein shake recipe ended up on the floor 😭 #gymfail"), gr.State(15), gr.State(19), gr.State("Monday"), gr.State("Fitness"), gr.State(50000), gr.State("18-24"), gr.State("Male")],
+        inputs=[gr.State("My protein shake ended up on the floor"), gr.State(15), gr.State(19), gr.State("Monday"), gr.State("Fitness"), gr.State(50000), gr.State("18-24"), gr.State("Male")],
         outputs=[initial_views, similar_videos, improved_text, new_views, uplift]
     )