Course Objectives:

 Describe generative AI and how it aligns to machine learning.

 Define the importance of generative AI and explain its potential risks and benefits.
 Identify business value from generative AI use cases.
 Learn real-world application of Generative AI.
 Apply generative AI models and popular tools

Course Outcomes:

 Understand evolution of Generative AI

 Will able to identify Business Value.
 Learn Capabilities of Generative AI in different Domains.
 Understand Chat GPT.
 Create/Deploying of a Generative AI application.
 CONTENTS
Page Signature of
WEEK E.No. NAME OF THE EXERCISE
No. the Faculty
Generate concise summaries using “prompts”
WEEK-1 1 with a maximum of 300 word count, practicing
different AI models(at least 4) .
Using AI tool create a chatbot.
WEEK-2 2
Using AI tool create a website for students
3 providing them with study materials.
WEEK-3
Write a program to illustrate text-to-image
WEEK-4 4 generation for any given prompt

Write a program to illustrate text-to-audio

WEEK-5 5 generation for any given prompt

Write a program to illustrate text-to-video

WEEK-6 6 generation for any given prompt

Page | 1
 WEEK-1

1. Generate concise summaries using “prompts” with a maximum of 300 word count,
practicing different AI models(at least 4) .
Prompt Engineering Fundamentals
1. Prompt
A prompt is the initial input or instruction given to an AI model to generate a response. It guides
the AI on what to generate, how to structure the output, and the desired content or context.
2. Elements of a Prompt
Key elements that make up an effective prompt include:
 Clarity: Clear and unambiguous instructions.
 Specificity: Detailed requirements and constraints.
 Context: Background information relevant to the task.
 Format: Desired format of the output, such as length, style, or structure.
3. Tips for Designing Prompts
 Be Specific: Provide detailed instructions to reduce ambiguity.
 Use Examples: Include examples of desired outputs to guide the AI.
 Iterate and Refine: Test and refine prompts based on the quality of the outputs.
 Leverage Constraints: Use constraints to guide the AI in generating relevant responses.
 Maintain Context: Ensure the AI has enough context to understand the task fully.
Generate Prompts for following Use Cases:
Use Case 1: Chatbot Interaction Improvement
Objective
Improve the interaction quality of a customer service chatbot by designing effective prompts.
 Example Prompts:
 Customer Greeting:
 Prompt: "The customer says hello. Generate a friendly and professional greeting
response."
 Context: Customer interaction.
 Expected Response: "Hello! How can I assist you today?"
 Order Status Inquiry:
 Prompt: "The customer asks about the status of their order. Provide a response that
asks for their order number."
 Context: Order tracking.
 Expected Response: "Can you please provide your order number so I can check the
status for you?"
Page | 2
 Laboratory Task:
 Design prompts to handle various customer inquiries such as returns, shipping details,
and product information.
 Test the prompts with a language model and refine based on the responses.
Use Case 2: Content Generation for Marketing
Objective: Generate engaging marketing content by designing effective prompts.

Page | 3
Page | 4
Page | 5
 WEEK-2

2. Using AI tool create a chatbot.

Program:

import openai

openai.api_key = "enter your API secret key"

messages = []
system_msg = input("What type of chatbot would you like to create?\n")
messages.append({"role": "system", "content": system_msg})

print("Your new assistant is ready!")

while input != "quit()":
message = input()
messages.append({"role": "user", "content": message})
response = openai.ChatCompletion.create(
model="gpt-3.5-turbo",
messages=messages)
reply = response["choices"][0]["message"]["content"]
messages.append({"role": "assistant", "content": reply})
print("\n" + reply + "\n")

Page | 6
Output:

Page | 7
Page | 8
Page | 9
 WEEK-3

3. Using AI tool create a website for students providing them with study materials.

Program:
import openai
import gradio

openai.api_key = "####"

messages = [{"role": "system", "content": "You are a financial experts that specializes in real
estate investment and negotiation"}]

def CustomChatGPT(user_input):
messages.append({"role": "user", "content": user_input})
response = openai.ChatCompletion.create(
model = "gpt-3.5-turbo",
messages = messages
)
ChatGPT_reply = response["choices"][0]["message"]["content"]
messages.append({"role": "assistant", "content": ChatGPT_reply})
return ChatGPT_reply

demo = gradio.Interface(fn=CustomChatGPT, inputs = "text", outputs = "text", title = "Real Estate Pro")

demo.launch(share=True)

Page | 10
Output:

Page | 11
Page | 12
Page | 13
 WEEK-4

4. Write a program to illustrate text-to-image generation for any given prompt.

Program:

# Install necessary libraries

!pip install diffusers transformers accelerate

from diffusers import StableDiffusionPipeline

import torch

# Load the pre-trained Stable Diffusion model

model_id = "CompVis/stable-diffusion-v1-4"
device = "cuda" if torch.cuda.is_available() else "cpu"

# Create the pipeline for text-to-image generation

pipe = StableDiffusionPipeline.from_pretrained(model_id)
pipe = pipe.to(device)

# Define the text prompt

prompt = "a beautiful landscape with mountains, lakes, and forests during sunset"

# Generate the image

with torch.autocast(device):
image = pipe(prompt).images[0]

# Display the image

image.show()

# Save the image

image.save("generated_image.png")

Page | 14
Output:

Page | 15
Page | 16
 WEEK-5

5. Write a program to illustrate text-to-audio generation for any given prompt.

Program:
# Step 1: Install necessary libraries
!pip install transformers gtts

# Step 2: Import necessary libraries

from transformers import pipeline
from gtts import gTTS
import os

# Step 3: Initialize the text generation pipeline

generator = pipeline('text-generation', model='gpt2')

# Step 4: Generate text

prompt = "Once upon a time in a land far away"
generated_text = generator(prompt, max_length=100, num_return_sequences=1)[0]
['generated_text']
print("Generated Text:", generated_text)

# Step 5: Convert text to speech using gTTS

tts = gTTS(generated_text, lang='en')
audio_file = "generated_audio.mp3"
tts.save(audio_file)

# Step 6: Play the audio file (if running in a Jupyter notebook or Colab)
from IPython.display import Audio
Audio(audio_file)

Page | 17
Output:

Page | 18
Page | 19
 WEEK-6

6. Write a program to illustrate text-to-video generation for any given prompt.

Program:

# Step 1: Install necessary libraries

!pip install transformers gtts moviepy pydub

# Step 2: Import necessary libraries

from transformers import pipeline
from gtts import gTTS
from moviepy.editor import *
from pydub import AudioSegment
import os

# Step 3: Initialize the text generation pipeline

generator = pipeline('text-generation', model='gpt2')

# Step 4: Generate text

prompt = "Once upon a time in a land far away"
generated_text = generator(prompt, max_length=100, num_return_sequences=3)

# Step 5: Convert text to speech and create audio clips

audio_clips = []
for i, text in enumerate(generated_text):
generated_text_segment = text['generated_text']
tts = gTTS(generated_text_segment, lang='en')
audio_file = f"generated_audio_{i}.mp3"
tts.save(audio_file)
audio_clips.append(AudioFileClip(audio_file))

# Combine audio clips into a single audio file

Page | 20
final_audio = concatenate_audioclips(audio_clips)
final_audio_path = "combined_audio.mp3"
final_audio.write_audiofile(final_audio_path)

# Step 6: Generate video from text and audio

video_duration = final_audio.duration
fps = len(generated_text) / video_duration

text_clips = []
for i, text in enumerate(generated_text):
generated_text_segment = text['generated_text']
txt_clip = TextClip(generated_text_segment, fontsize=50, color='white',
bg_color='black').set_duration(final_audio.duration)
text_clips.append(txt_clip)

final_clip = concatenate_videoclips(text_clips)
final_clip = final_clip.set_audio(final_audio)

# Save the final video file

output_video_path = "generated_video.mp4"
final_clip.write_videofile(output_video_path, fps=fps, codec='libx264')

# Step 7: Display or download the video

final_clip.ipython_display()
# Or download the video file (for Colab or Jupyter notebook)
# from google.colab import files
# files.download(output_video_path)

Page | 21
Output:

Page | 22
Page | 23
Page | 24
Page | 25