So, how can you do this? By adjusting the model parameters. There are 3 parameters you can adjust:
1. Temperature
2. Top-k
3. Top-p But before we see how to use them, let’s first understand how the model selects the next token to generate.

→ View original post on X — @cohere