Profile

This process helps me tremendously when using LLMs! Hopefully it is useful for you! Step 1: Establish Assessment Framework Assign the model an expert role in your target domain (e.g. "You are an expert UX designer"). Prompt it to analyze the original input and score its overall quality on a 1–100 ...

Hello Everyone! Running Granite 3.3-2b locally, on my music production laptop(8gb RAM) was completely feasible while testing out the small 2b parameter model to create a "Reese Bass" sound for a song I'm working on. It's very helpful to have a reference chart for an initial sound in the drafting stage ...

This sounds like a great meet up! Count me in. ------------------------------ Julian Gonzalez Prompt Engineer/Analyst CreativeAct Technologies Orlando FL ------------------------------

#LLM

Hello Everyone, I've been trying out Granite 3.2 for various use cases of my own and have noticed some interesting nuances with Granite 3.2. A well-structured zero-shot prompt is all you need for most tasks. You don't need meta prompts that are paragraphs long for something simple like this ...