Upload app.py

app.py

@@ -0,0 +1,87 @@
+
+import gradio as gr
+import pandas as pd
+import joblib
+from huggingface_hub import hf_hub_download
+
+REPO_ID = "DetectiveShadow/Grade_predictor"
+
+# Load pipeline and optional schema at startup
+def load_model():
+    pipe = joblib.load(hf_hub_download(REPO_ID, "grade_predictor.pkl"))
+    schema_path = None
+    try:
+        schema_path = hf_hub_download(REPO_ID, "feature_schema.json")
+    except Exception:
+        schema_path = None
+
+    schema = None
+    if schema_path is not None:
+        try:
+            import json
+            with open(schema_path, "r") as f:
+                schema = json.load(f)
+        except Exception:
+            schema = None
+    return pipe, schema
+
+PIPE, SCHEMA = load_model()
+
+# Defaults & choices (adjust if your training data uses different labels)
+SUBJECTS = ["Math","Science","English","History","Tech","Drama"]
+DIFFICULTY = ["Regular","Honors","AP"]
+GROUPING = ["Solo","Group"]
+
+def predict(attendance, hours, quizzes, confidence, grade_level, subject, difficulty, group_or_solo):
+    # Prepare a single-row dataframe matching training feature names
+    row = pd.DataFrame([{
+        "attendance": float(attendance),
+        "hours_studied": int(hours),
+        "quizzes_avg": int(quizzes),
+        "confidence": int(confidence),
+        "grade_level": int(grade_level),
+        "subject": subject,
+        "course_difficulty": difficulty,
+        "group_or_solo": group_or_solo
+    }])
+    pred = float(PIPE.predict(row)[0])
+
+    # Letter grade helper
+    def to_letter(x):
+        if x >= 90: return "A"
+        if x >= 80: return "B"
+        if x >= 70: return "C"
+        if x >= 60: return "D"
+        return "F"
+
+    letter = to_letter(pred)
+    return {"Predicted Final Grade": round(pred, 1), "Letter": letter}
+
+with gr.Blocks(title="Grade Predictor") as demo:
+    gr.Markdown("# 🎯 Grade Predictor")
+    gr.Markdown("Enter your study & course details to estimate your final grade.")
+
+    with gr.Row():
+        attendance = gr.Slider(0.5, 1.0, value=0.95, step=0.01, label="Attendance (0–1)")
+        hours      = gr.Slider(0, 30, value=12, step=1, label="Hours studied / week")
+    with gr.Row():
+        quizzes    = gr.Slider(0, 100, value=85, step=1, label="Quizzes average")
+        confidence = gr.Slider(0, 10, value=7, step=1, label="Confidence (0–10)")
+    with gr.Row():
+        grade_lvl  = gr.Slider(9, 12, value=11, step=1, label="Grade level")
+        subject    = gr.Dropdown(SUBJECTS, value="Math", label="Subject", allow_custom_value=True)
+    with gr.Row():
+        difficulty = gr.Dropdown(DIFFICULTY, value="Regular", label="Course difficulty", allow_custom_value=True)
+        grouping   = gr.Dropdown(GROUPING, value="Solo", label="Group or Solo", allow_custom_value=True)
+
+    go = gr.Button("Predict")
+    out = gr.JSON(label="Prediction")
+
+    go.click(
+        fn=predict,
+        inputs=[attendance, hours, quizzes, confidence, grade_lvl, subject, difficulty, grouping],
+        outputs=out
+    )
+
+if __name__ == "__main__":
+    demo.launch()