Update app.py

app.py

@@ -1,32 +1,24 @@
 import gradio as gr
-from transformers import pipeline
 import requests
 import os
 import random
 
-# ==============================
+# =====================================
 # CONFIG
-# ==============================
+# =====================================
 
 HF_TOKEN = os.getenv("HF_TOKEN")
-API_URL = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.2"
+
+WHISPER_API = "https://api-inference.huggingface.co/models/openai/whisper-base"
+LLM_API = "https://api-inference.huggingface.co/models/mistralai/Mistral-7B-Instruct-v0.2"
 
 headers = {
     "Authorization": f"Bearer {HF_TOKEN}"
 }
 
-# ==============================
-# Load Whisper (Lightweight)
-# ==============================
-
-asr = pipeline(
-    "automatic-speech-recognition",
-    model="openai/whisper-base"
-)
-
-# ==============================
+# =====================================
 # Question Bank
-# ==============================
+# =====================================
 
 questions = {
     "Easy": [
@@ -40,54 +32,49 @@ questions = {
         "Difference between CNN and RNN?"
     ],
     "Hard": [
-        "Explain backpropagation mathematically.",
+        "Explain backpropagation.",
         "What is attention mechanism?",
-        "Explain transformers architecture."
+        "Explain transformer architecture."
     ]
 }
 
-# ==============================
+# =====================================
 # Generate Question
-# ==============================
+# =====================================
 
 def start_interview(level):
     return random.choice(questions[level])
 
-# ==============================
-# Call LLM via API
-# ==============================
+# =====================================
+# Speech-to-Text using API
+# =====================================
 
-def query_llm(prompt):
+def speech_to_text(audio_path):
+    try:
+        with open(audio_path, "rb") as f:
+            audio_bytes = f.read()
 
-    payload = {
-        "inputs": prompt,
-        "parameters": {
-            "max_new_tokens": 300,
-            "temperature": 0.7
-        }
-    }
+        response = requests.post(
+            WHISPER_API,
+            headers=headers,
+            data=audio_bytes
+        )
 
-    response = requests.post(API_URL, headers=headers, json=payload)
+        if response.status_code == 200:
+            return response.json().get("text", "Could not transcribe audio.")
+        else:
+            return "Speech recognition API error."
 
-    if response.status_code == 200:
-        return response.json()[0]["generated_text"]
-    else:
-        return "Error contacting LLM API."
+    except Exception as e:
+        return f"Speech recognition failed: {str(e)}"
 
-# ==============================
-# Evaluate Answer
-# ==============================
 
-def evaluate_answer(audio, question):
+# =====================================
+# LLM Evaluation
+# =====================================
 
-    if audio is None:
-        return "Please record your answer."
+def evaluate_with_llm(question, answer):
 
-    # Speech to Text
-    result = asr(audio)
-    user_answer = result["text"]
-
-    # Prompt Engineering
     prompt = f"""
 You are a strict technical interviewer.
 
@@ -95,37 +82,76 @@ Question:
 {question}
 
 Candidate Answer:
-{user_answer}
+{answer}
 
 Evaluate and give:
 
-1. Technical Accuracy Score (0-10)
-2. Clarity Score (0-10)
-3. Depth Score (0-10)
+1. Technical Accuracy (0-10)
+2. Clarity (0-10)
+3. Depth (0-10)
 4. Overall Score (0-10)
-5. Improvement Suggestions (short and clear)
+5. Improvement Suggestions
 
 Be concise and structured.
 """
 
-    feedback = query_llm(prompt)
+    payload = {
+        "inputs": prompt,
+        "parameters": {
+            "max_new_tokens": 300,
+            "temperature": 0.7
+        }
+    }
+
+    try:
+        response = requests.post(
+            LLM_API,
+            headers=headers,
+            json=payload
+        )
+
+        if response.status_code == 200:
+            return response.json()[0]["generated_text"]
+        else:
+            return "LLM evaluation API error."
+
+    except Exception as e:
+        return f"Evaluation failed: {str(e)}"
+
+
+# =====================================
+# Main Evaluation Function
+# =====================================
+
+def evaluate_answer(audio, question):
+
+    if audio is None:
+        return "Please record your answer."
+
+    if question.strip() == "":
+        return "Please click 'Start Interview' first."
+
+    transcribed_text = speech_to_text(audio)
+
+    feedback = evaluate_with_llm(question, transcribed_text)
 
     return f"""
 📝 Transcribed Answer:
-{user_answer}
+{transcribed_text}
 
 📊 Evaluation:
 {feedback}
 """
 
-# ==============================
+
+# =====================================
 # UI
-# ==============================
+# =====================================
 
 with gr.Blocks() as demo:
 
-    gr.Markdown("# 🎤 Smart Interview Simulator (AI Voice Bot)")
-    gr.Markdown("Select difficulty → Answer using voice → Get AI feedback")
+    gr.Markdown("# 🎤 Smart Interview Simulator")
+    gr.Markdown("Voice-Based AI Mock Interview System")
 
     level_dropdown = gr.Dropdown(
         ["Easy", "Medium", "Hard"],
@@ -136,7 +162,11 @@ with gr.Blocks() as demo:
     question_output = gr.Textbox(label="Interview Question")
 
     start_button = gr.Button("Start Interview")
-    start_button.click(start_interview, inputs=level_dropdown, outputs=question_output)
+    start_button.click(
+        start_interview,
+        inputs=level_dropdown,
+        outputs=question_output
+    )
 
     audio_input = gr.Audio(
         type="filepath",
@@ -145,7 +175,10 @@ with gr.Blocks() as demo:
 
     submit_button = gr.Button("Submit Answer")
 
-    result_output = gr.Textbox(label="Evaluation Feedback")
+    result_output = gr.Textbox(
+        label="Evaluation Feedback",
+        lines=15
+    )
 
     submit_button.click(
         evaluate_answer,
@@ -153,4 +186,4 @@ with gr.Blocks() as demo:
         outputs=result_output
     )
 
-demo.launch()
+demo.launch()