From the course: Computer Vision for Data Scientists

Unlock the full course today

Join today to access over 24,200 courses taught by industry experts.

Optional training params

Optional training params

From the course: Computer Vision for Data Scientists

Start my 1-month free trial Buy for my team

Optional training params

“

- All right, so we talked about the required parameters, now let's talk about some of the optional parameters that we can pass to the trainer. So one of them is the optimizer params. So these are the parameters, the optimizer you're using. If you leave it blank, it'll just use the default. And here's what I mean by optimizer parameters. For example, we're using Adam. So one of the parameters for Adam is the learning rate. But there's also these other parameters, the betas, the epsilons, the weight decays, so on, so forth. If you want to tweak those parameters you would just pass them in to this as a dictionary. So I'm leaving them blank here and it'll use a default, just the pure PyTorch defaults. But it's up to you if you want to play around with that. I encourage you to play around with it and see how it impacts your training. Next is the Criterion Params. Now what does this do? This is actually activating…

Contents