This feature allows people to add large context to your prompt. It serves as an additional source of information, and our advanced AI will extract relevant info from provided dataset and add it to the prompt when a user asks a specific question. This is a very useful feature as it saves you time in creating prompts and reduces the number of tokens used inside of your prompt. To use the dataset feature for prompts, follow these steps: (1). Click on the "Create" button. (2). Name your prompt. (3). Select the "Prompt" option and click "Create". On the Prompt screen: (1). Begin by entering your prompt in the first input bar. (2). If desired, adjust the model and its parameters. (3). Craft a welcome message, which users will see at the conversation's start. Next, add the context file for your prompt's dataset. For instance, I'll add a book about black holes and dark energy: (1). Click the "Add" button to access the dataset page. (2). To upload, either drag and drop or use the browse option. Supported formats are txt and pdf. (3). Once done, click "Submit" to finalize your dataset." Once all steps are completed, initiate the model by clicking the "Send" button. Engage in a chat to create a sample conversation. Below, you can see how it all comes together after updating. Next, proceed to the basic info of your prompt. You can modify the name, description, category, tags, and permissions. Unsure about these details? Use the 'Auto Fill' option to complete the fields for you. After you finished everything, click on “Submit” and your prompt is ready for using! Let's now look at an example of how to use this to create a virtual character that is more accurate. First, you need to create a prompt just like last time, but this time we will use the template option. The template option is very useful when creating a one-message prompt or prompt where you don't have a lengthy conversation with the chatbot. Add prompt, context, and welcome message. The template mode is a fill-in-the-blank structure used for single-turn responses or short conversations with the chatbot. To create a variable, simply type the variable name within brackets. When the user provides their custom input, that variable will be replaced, and the prompt will incorporate that custom input. In the image below, you can see how the variable changed when the user entered their input. Very straightforward! Next is the context. As we mentioned, you can use any dataset you have in TXT or PDF file format as context. This time, we have downloaded the entire Wikipedia page about Jala Brat, and we will upload that Wikipedia page as a PDF into the context to use as a dataset. Since the virtual character in this example is a person who isn't widely known in the GPT database, they will have all the information about him through this PDF. We will click on the "Add" button for the context, then drag and drop that PDF file. After that, we will submit, and that's it! Once you've submitted the dataset and sent the prompt to the model to create a conversation example, don't forget that the Welcome message is optional but also important if you want to make your prompt creative! The last step is to customize your prompt page and provide basic information. In this case, we uploaded a custom image cover and utilized the Auto Fill option to automatically populate all the fields. And that's it! People can now ask questions about Jala Brat using the variable option. We hope that you've learned how to use the dataset feature through this guide. Best of luck with your prompts, and let's FLOW!