Update app.py

app.py

@@ -1,94 +1,147 @@
 import gradio as gr
 from transformers import pipeline
+import requests
+import os
 import random
 
-# -------------------------------
-# Load Models
-# -------------------------------
+# ==============================
+# CONFIG
+# ==============================
+
+HF_TOKEN = os.getenv("HF_TOKEN")
+API_URL = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.2"
+
+headers = {
+    "Authorization": f"Bearer {HF_TOKEN}"
+}
+
+# ==============================
+# Load Whisper (Lightweight)
+# ==============================
 
-# Speech to Text Model (Whisper)
 asr = pipeline(
     "automatic-speech-recognition",
     model="openai/whisper-base"
 )
 
-# Text Generation Model (LLM)
-generator = pipeline(
-    "text-generation",
-    model="mistralai/Mistral-7B-Instruct-v0.1",
-    device_map="auto"
-)
-
-# -------------------------------
+# ==============================
 # Question Bank
-# -------------------------------
-
-questions = [
-    "Explain overfitting in machine learning.",
-    "What is the difference between supervised and unsupervised learning?",
-    "Explain gradient descent in simple terms.",
-    "What is the role of activation functions in neural networks?",
-    "What is the difference between CNN and RNN?"
-]
-
-# -------------------------------
-# Interview Logic
-# -------------------------------
-
-def start_interview():
-    question = random.choice(questions)
-    return question
+# ==============================
+
+questions = {
+    "Easy": [
+        "What is Machine Learning?",
+        "Explain supervised learning.",
+        "What is overfitting?"
+    ],
+    "Medium": [
+        "Explain bias vs variance tradeoff.",
+        "What is gradient descent?",
+        "Difference between CNN and RNN?"
+    ],
+    "Hard": [
+        "Explain backpropagation mathematically.",
+        "What is attention mechanism?",
+        "Explain transformers architecture."
+    ]
+}
+
+# ==============================
+# Generate Question
+# ==============================
+
+def start_interview(level):
+    return random.choice(questions[level])
+
+# ==============================
+# Call LLM via API
+# ==============================
+
+def query_llm(prompt):
+
+    payload = {
+        "inputs": prompt,
+        "parameters": {
+            "max_new_tokens": 300,
+            "temperature": 0.7
+        }
+    }
+
+    response = requests.post(API_URL, headers=headers, json=payload)
+
+    if response.status_code == 200:
+        return response.json()[0]["generated_text"]
+    else:
+        return "Error contacting LLM API."
+
+# ==============================
+# Evaluate Answer
+# ==============================
 
 def evaluate_answer(audio, question):
 
-    # Convert Speech to Text
+    if audio is None:
+        return "Please record your answer."
+
+    # Speech to Text
     result = asr(audio)
     user_answer = result["text"]
 
-    # Create evaluation prompt
+    # Prompt Engineering
     prompt = f"""
-    You are a technical interviewer.
-
-    Question: {question}
-    Candidate Answer: {user_answer}
-
-    Evaluate the answer and give:
-    1. Technical Accuracy Score (out of 10)
-    2. Clarity Score (out of 10)
-    3. Overall Score (out of 10)
-    4. Improvement Suggestions
-
-    Keep feedback concise.
-    """
-
-    # Generate Evaluation
-    output = generator(
-        prompt,
-        max_new_tokens=300,
-        do_sample=True,
-        temperature=0.7
-    )
+You are a strict technical interviewer.
+
+Question:
+{question}
+
+Candidate Answer:
+{user_answer}
+
+Evaluate and give:
+
+1. Technical Accuracy Score (0-10)
+2. Clarity Score (0-10)
+3. Depth Score (0-10)
+4. Overall Score (0-10)
+5. Improvement Suggestions (short and clear)
 
-    feedback = output[0]["generated_text"]
+Be concise and structured.
+"""
 
-    return f"📝 Transcribed Answer:\n{user_answer}\n\n📊 Evaluation:\n{feedback}"
+    feedback = query_llm(prompt)
 
+    return f"""
+📝 Transcribed Answer:
+{user_answer}
 
-# -------------------------------
-# Gradio UI
-# -------------------------------
+📊 Evaluation:
+{feedback}
+"""
+
+# ==============================
+# UI
+# ==============================
 
 with gr.Blocks() as demo:
 
     gr.Markdown("# 🎤 Smart Interview Simulator (AI Voice Bot)")
-    gr.Markdown("Answer the question using your voice.")
+    gr.Markdown("Select difficulty → Answer using voice → Get AI feedback")
+
+    level_dropdown = gr.Dropdown(
+        ["Easy", "Medium", "Hard"],
+        value="Medium",
+        label="Select Difficulty"
+    )
 
     question_output = gr.Textbox(label="Interview Question")
 
     start_button = gr.Button("Start Interview")
-    start_button.click(start_interview, outputs=question_output)
+    start_button.click(start_interview, inputs=level_dropdown, outputs=question_output)
 
-    audio_input = gr.Audio(source="microphone", type="filepath")
+    audio_input = gr.Audio(
+        type="filepath",
+        label="Record Your Answer"
+    )
 
     submit_button = gr.Button("Submit Answer")