Upload folder using huggingface_hub

app.py

@@ -1,118 +1,78 @@
-import streamlit as st
+import gradio as gr
 import pandas as pd
 import joblib
 import matplotlib.pyplot as plt
-from huggingface_hub import hf_hub_download
+import numpy as np
 
-# -----------------------------------------
-# Load Model from Hugging Face Model Hub
-# -----------------------------------------
-
-REPO_ID = "SunnyShaurya/engine-condition-classifier"
-
-model_path = hf_hub_download(
-    repo_id=REPO_ID,
-    filename="engine_condition_rf_production.joblib"
-)
-
-threshold_path = hf_hub_download(
-    repo_id=REPO_ID,
-    filename="decision_threshold.joblib"
-)
-
-model = joblib.load(model_path)
-saved_threshold = joblib.load(threshold_path)
+# ----------------------------
+# Load Model
+# ----------------------------
+model = joblib.load("engine_condition_rf_production.joblib")
+saved_threshold = joblib.load("decision_threshold.joblib")
 
 feature_names = model.feature_names_in_
 
-# -----------------------------------------
-# Streamlit UI
-# -----------------------------------------
-
-st.title("Engine Condition Classification System")
-
-st.markdown("### Adjust Decision Threshold")
-user_threshold = st.slider(
-    "Decision Threshold",
-    min_value=0.1,
-    max_value=0.9,
-    value=float(saved_threshold),
-    step=0.01
-)
-
-# -----------------------------------------
-# Single Prediction Section
-# -----------------------------------------
-
-st.markdown("## Manual Engine Input")
-
-input_data = []
-
-for feature in feature_names:
-    value = st.number_input(f"{feature}", value=0.0)
-    input_data.append(value)
-
-if st.button("Predict Engine Condition"):
-
-    input_df = pd.DataFrame([input_data], columns=feature_names)
+# ----------------------------
+# Single Prediction Function
+# ----------------------------
+def predict_engine(*inputs):
+    input_df = pd.DataFrame([inputs], columns=feature_names)
     probability = model.predict_proba(input_df)[0][1]
-    prediction = 1 if probability >= user_threshold else 0
-
-    st.write("### Probability of Failure:", round(probability, 4))
-    st.write(f"Model Confidence: {round(probability*100,2)}%")
-
-    # Explanation Logic
-    if probability > 0.75:
-        st.info("High risk detected. Immediate inspection recommended.")
-    elif probability > 0.55:
-        st.warning("Moderate risk. Preventive check advised.")
-    else:
-        st.success("Low risk. Engine likely operating normally.")
-
+    prediction = 1 if probability >= saved_threshold else 0
+    
     if prediction == 1:
-        st.error("⚠ Engine Likely Faulty")
+        result = "⚠ Engine Likely Faulty"
     else:
-        st.success("✅ Engine Operating Normally")
-
-    # -----------------------------------------
-    # Confidence Visualization
-    # -----------------------------------------
-
-    fig, ax = plt.subplots()
-    ax.bar(["Normal Probability", "Failure Probability"],
-           [1 - probability, probability])
-    ax.set_ylim(0, 1)
-    ax.set_ylabel("Probability")
-    st.pyplot(fig)
-
-# -----------------------------------------
-# Batch Prediction (CSV Upload)
-# -----------------------------------------
-
-st.markdown("## Batch Prediction (CSV Upload)")
-
-uploaded_file = st.file_uploader("Upload CSV File", type=["csv"])
-
-if uploaded_file is not None:
-
-    df = pd.read_csv(uploaded_file)
-
-    # Ensure columns match training features
+        result = "✅ Engine Operating Normally"
+        
+    return result, round(probability, 4)
+
+# ----------------------------
+# Batch Prediction Function
+# ----------------------------
+def batch_predict(file):
+    df = pd.read_csv(file.name)
+    
+    missing_cols = [col for col in feature_names if col not in df.columns]
+    
+    if missing_cols:
+        return f"Missing required columns: {missing_cols}"
+    
     df = df[feature_names]
-
     probabilities = model.predict_proba(df)[:, 1]
-
+    
     df["Probability_of_Failure"] = probabilities
-    df["Prediction"] = (probabilities >= user_threshold).astype(int)
-
-    st.write("### Prediction Results")
-    st.write(df.head())
-
-    csv = df.to_csv(index=False).encode("utf-8")
-
-    st.download_button(
-        "Download Predictions",
-        csv,
-        "engine_predictions.csv",
-        "text/csv"
-    )
+    df["Prediction"] = (probabilities >= saved_threshold).astype(int)
+    
+    output_file = "engine_predictions.csv"
+    df.to_csv(output_file, index=False)
+    
+    return output_file
+
+# ----------------------------
+# Build UI
+# ----------------------------
+with gr.Blocks() as demo:
+    
+    gr.Markdown("# 🚗 Engine Condition Classification System")
+    
+    gr.Markdown("## 🔧 Manual Prediction")
+    
+    inputs = []
+    for feature in feature_names:
+        inputs.append(gr.Number(label=feature))
+    
+    output_text = gr.Textbox(label="Prediction Result")
+    output_prob = gr.Number(label="Failure Probability")
+    
+    btn = gr.Button("Predict Engine Condition")
+    btn.click(predict_engine, inputs, [output_text, output_prob])
+    
+    gr.Markdown("## 📂 Batch Prediction (CSV Upload)")
+    
+    file_input = gr.File(label="Upload CSV File")
+    file_output = gr.File(label="Download Predictions")
+    
+    file_input.change(batch_predict, file_input, file_output)
+
+demo.launch()

requirements.txt

@@ -1,7 +1,6 @@
-streamlit
-scikit-learn==1.6.1
+gradio
 pandas
 numpy
-joblib
-huggingface_hub
+scikit-learn
 matplotlib
+joblib