Tired of weird symbols and cryptic commands? This course unlocks the secret world of coding with ChatGPT! Learn how to: Become a code detective: Ask ChatGPT questions and get special codes (like secret languages) to find exactly what you need in your computer files. Imagine finding all your pictures from last year's science fair in a snap! Build awesome tools: Stuck creating parts of your game? ChatGPT can help design the building blocks to get you started, like a super code toolkit! Clean up your code: Sometimes code gets messy, just like your room! ChatGPT can help you organize and understand your code, making it easier to work with. Test your code like a boss: Worried about mistakes? ChatGPT can help find errors and create like practice tests for your code! Speak different coding languages: Need to translate your code from English to computer language? ChatGPT can act as your translator! ChatGPT is your coding buddy, not your replacement! Think of it as a super helpful friend who can: Suggest solutions and answer your coding questions. Help you understand complex coding concepts. Give your confusing code nicknames and make it easier to read. Bonus! Learn about cool coding things like databases and regular expressions, and how ChatGPT can help you unlock their secrets! This course is perfect for YOU if you want to: Become a coding whiz! Make your computer do super cool things! Level up your tech skills!
Ever wondered how those chatbots you see online work? This course explains how Large Language Models (LLMs) are built and used. Imagine a game where you predict the next word in a sentence. LLMs are trained on massive amounts of text data to do this very thing. For example, if you give it the prompt "I love eating", the LLM might guess "bagels with cream cheese" or "out with friends" to complete the sentence. But how does the LLM know this? It learns from labelled training data. Here's an example: Imagine training an LLM to understand the sentiment of restaurant reviews. You might feed it reviews labelled as "positive" (e.g., "The pastrami sandwich is great!") or "negative" (e.g., "Service was slow, the food was so-so."). By processing tons of examples, the LLM learns to identify patterns in language. LLMs can be used in two main ways: Base LLMs: These predict the next word one at a time. If you prompt it with "Once upon a time there was a unicorn," it might create a story about the unicorn. However, if you ask a question like "What is the capital of France?", it might get confused. Instruction Tuned LLMs: These are trained on additional data where the output follows an instruction. For instance, you could give it examples of instructions and good responses. This helps the LLM understand what you want it to do, so it can answer your questions directly (e.g., "The capital of France is Paris"). Here's a cool trick: LLMs work with pieces of text called "tokens" instead of individual words. If you tell an LLM to reverse the letters of "lollipop," it might struggle because it sees "lol," "li," and "pop" as separate tokens. But if you add dashes between the letters ("l-o-l-l-i-p-o-p"), it can see each letter clearly and reverse them correctly. Did you know LLMs can have different conversation styles? You can specify the tone you want the LLM to use in a "system message" and then give it a specific instruction in a "user message." This lets you control how the LLM responds. Remember, LLMs are still under development. They might not always understand your requests perfectly, and it's important to use them responsibly. This is just the beginning! In the next lesson, you'll learn how to use LLMs to build a customer service assistant for an online store.