Scribd
Upload
68 views3 pages

TDM Prompt Engineering Assignment

Uploaded by

naveedini079
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
68 views3 pages

TDM Prompt Engineering Assignment

Uploaded by

naveedini079
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

BSc (Hons) in Information Technology

Year 4
Assessment – Prompt Engineering

SE4051 – Trends in Digital Media Semester 1, 2024

Due Date : 24th June at 11:59PM


Objective:

To evaluate your understanding and application of various prompting strategies in text and image
generation using language models and image generation models.

Task 1: Text Generation with a Language Model (LLM)

Instructions:

1. Prompting Strategy Application:


o Choose a text-generating language model (e.g., OpenAI's GPT-4).
o Your task is to generate a small dialogue(script) for a short 3-minute video. Use the
following prompting strategies:
• You are required to generate the script first by using basic Zero-Shot
Prompting.
• Next you have to try and improve the output by employing at minimum
three(3) other prompting strategies.
• One strategy should be a Combination of multiple strategies as discussed in
class.

2. Execution and Analysis:


o Generate the dialogue script using four prompts in total.
o Take screenshots of each prompt and its generated output.
o Analyze the differences in the outputs. Discuss how the structure and wording of
each prompt influenced the generated dialogue.
o Reflect on which techniques were most and least effective and explain why.

Submission Requirements:

• The four prompts you used with the employed strategy.


• The four generated dialogue scripts.
• Screenshots of each prompt and its corresponding output.
• An analysis (approximately 300-500 words) discussing the differences in the outputs, the
impact of different prompting strategies, and the suitability of each technique.
Task 2: Image Generation with an Image Generation Model

Instructions:

1. Prompting Strategy Application:


o Choose an image generation model (e.g., DALL-E, MidJourney, Stable Diffusion).
o Your task is to generate images of the same scenery using three different prompts:
o Initial prompt should be a Basic Prompt with A simple and concise description of
the scene.
o Try 3 other prompts to get a better output.

2. Execution and Analysis:


o Generate the images using four(4) prompts in total.
o Take screenshots of each prompt and the resulting image.
o Compare and contrast the generated images. Discuss how the amount of detail in the
prompt influenced the final image.
o Based on the outputs you got, reflect on whether providing too much details is
beneficial or if it's better to leave some room for the model's imagination.

Submission Requirements:

• The four prompts you used.


• The four generated images.
• Screenshots of each prompt and its corresponding image.
• An analysis (approximately 300-500 words) discussing the differences in the outputs, the
impact of prompt detail, and the balance between guidance and creativity.
Marking Guide

Criteria Poor (0-49%) Average (50- Good (65-79%) Very Good (80-
64%) 100%)
Task 1: Missing or Basic prompts Clear and varied Well-crafted prompts
Prompts(15%) unclear with limited prompts using showing
prompts strategy different strategies understanding of
strategies
Task 1: Incomplete or Basic outputs, Relevant outputs, High-quality outputs,
Outputs(15%) irrelevant some showing clear differences
outputs relevance differences
Task 1: Superficial or Basic Clear analysis, In-depth analysis,
Analysis (20%) missing analysis, some insights into clear understanding of
analysis limited techniques techniques
insight
Task 2: Missing or Basic prompts Clear and varied Well-crafted prompts
Prompts(15%) unclear with limited prompts with showing a balance of
prompts detail different levels of detail
detail
Task 2: Incomplete or Basic outputs, Relevant outputs, High-quality outputs,
Outputs(15%) irrelevant some showing clear differences
outputs relevance differences
Task 2: Superficial or Basic Clear analysis, In-depth analysis,
Analysis(20%) missing analysis, some insights into clear understanding of
analysis limited detail vs creativity detail vs creativity
insight

Total Marks:

• Task 1: 50%
o Prompts: 15%
o Outputs: 15%
o Analysis: 20%
• Task 2: 50%
o Prompts: 15%
o Outputs: 15%
o Analysis: 20%

You might also like