Create app.py

app.py

@@ -0,0 +1,81 @@
+# Agentic AI Bot using Mistral-7B-Instruct-v0.1 with Text & Speech (Streamlit App)
+
+import streamlit as st
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+from gtts import gTTS
+import tempfile
+import os
+import torch
+import speech_recognition as sr
+import soundfile as sf
+import io
+
+st.set_page_config(page_title="🧠 Agentic AI Bot", layout="centered")
+st.title("🎙️ Agentic AI Assistant (Text + Voice)")
+
+# Load model and tokenizer
+@st.cache_resource
+def load_model():
+    tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.1")
+    model = AutoModelForCausalLM.from_pretrained(
+        "mistralai/Mistral-7B-Instruct-v0.1",
+        device_map="auto",
+        torch_dtype=torch.float16,
+        load_in_4bit=True
+    )
+    pipe = pipeline("text-generation", model=model, tokenizer=tokenizer)
+    return pipe
+
+pipe = load_model()
+
+# Text-to-speech function
+def speak(text):
+    tts = gTTS(text)
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".mp3") as fp:
+        tts.save(fp.name)
+        return fp.name
+
+# Speech-to-text function
+def transcribe(audio_bytes):
+    recognizer = sr.Recognizer()
+    with sr.AudioFile(io.BytesIO(audio_bytes)) as source:
+        audio_data = recognizer.record(source)
+        try:
+            return recognizer.recognize_google(audio_data)
+        except sr.UnknownValueError:
+            return "Sorry, I could not understand the audio."
+        except sr.RequestError:
+            return "Speech recognition service is unavailable."
+
+# Input mode selection
+mode = st.radio("Choose input mode:", ["Text", "Voice"])
+
+user_input = ""
+if mode == "Text":
+    user_input = st.text_input("Enter your query:")
+else:
+    audio = st.file_uploader("Upload your voice (WAV format only)", type=["wav"])
+    if audio is not None:
+        audio_bytes = audio.read()
+        user_input = transcribe(audio_bytes)
+        st.write(f"You said: {user_input}")
+
+# Run the assistant
+if user_input:
+    with st.spinner("Thinking..."):
+        result = pipe(user_input, max_new_tokens=200, temperature=0.7, do_sample=True)
+        response = result[0]['generated_text']
+
+        # Trim prompt from response if repeated
+        if response.lower().startswith(user_input.lower()):
+            response = response[len(user_input):].strip()
+
+        st.subheader("🤖 Assistant's Response:")
+        st.write(response)
+
+        # Speak response
+        audio_path = speak(response)
+        st.audio(audio_path, format="audio/mp3")
+
+        # Cleanup
+        os.remove(audio_path)