Learn to frame prompts effectively by setting roles, contexts, and additional details; apply framing concepts to tailor model responses for tasks like language learning.

Include personal details to tailor ChatGPT outputs to your persona and knowledge, moving beyond the default state to create more targeted, learning-focused explanations.

Explore how constraints shape prompt construction by constraining output format and conditioning semantics, guiding models with time limits, facts, and artistic styles.

Explore how to constrain prompts by semantics, knowledge, time, format, perspective, and artistic conditioning to shape output. Learn practical ways to apply these constraints to guide the model's responses.

Learn to articulate clear expectations and the why behind your prompts to tailor ChatGPT outputs, specify format, and improve results with concrete examples.

state the expectations and desired outcomes when prompting ChatGPT, craft a prompt to classify activities as recreation, work, or education, and apply it to organize a calendar.

Master structuring prompts for clarity by organizing framing and tasks, leveraging linear progression and the attention mechanism, and using a multi-prompt workflow to guide outputs.

Explore how to structure prompts for clarity by framing, labeling, and dividing tasks, then generate blog posts in stages using chunked prompts and linear thinking with ChatGPT.

This lesson shows how to offer examples to ChatGPT via uploaded documents, browser research, or embedded prompts to generate a casual blog post on studying and social life balance.

Explore chain of thought prompting and its zero-shot variant to guide language models through step-by-step reasoning, and learn how few-shot and zero-shot approaches improve prompt engineering.

Develop and apply chain of thought and zero-shot chain of thought techniques to train models to think step by step, assess statements for historical accuracy, and derive correct answers.

Generate knowledge on a topic before tackling the task to build context and improve the quality of subsequent outputs.

Learn how to generate knowledge by using prompts to guide the model, leverage initial data as context, and surface topic-relevant data to craft a blog post.

Discover how temperature, max token, and top p shape how prompts influence large language models; learn how prompts mimic parameters, the effects on randomness, length limits, and vocabulary diversity.

Learn how to build and apply mental models as zero-shot chain-of-thought frameworks to automate prompts, outline blog content, and deliver concepts from simple to complex.

Build a flow of prompts by combining priming, chain-of-thought, and mental models to create workflows that generate higher quality blog posts.

Explore how to use bias in prompts for ChatGPT by leveraging anchoring bias, recency bias, framing bias, and authority bias to guide outputs.

Define the exact goal and use case for your fine-tuned GPT, generate consistent data with clear patterns, then upload, fine-tune on OpenAI, and access via the API.

Explore how to decide the task for a fine-tuned model by balancing specificity and versatility, choosing a narrow niche, crafting a system prompt, and modeling input variations.

Create a workflow to generate high-quality input and output data with consistent patterns using ChatGPT, enabling faster data creation and better model fine-tuning.

Build a practical workflow for a fine-tuned model by priming with role playing, staging low- and high-resolution knowledge, and using the Socratic method and prompts to solve problems.

Generate data by building a clear workflow of input-output pairs, including the initial prompt and final output, and repeat with variations to improve model fine-tuning and problem-solving within a topic.

Generate data with ChatGPT by designing prompts, priming the model, and building a structured workflow. Learn to format datasets and use consistent prompt patterns for reliable outputs.

Learn how to fine-tune a model by converting a spreadsheet to json, using Google Colab and Google Drive for data prep, and uploading a json file for fine-tuning.