Hi Freddy,Both memory and CPU usage will largely depend on your data, and typical use.A multi-core CPU would increase throughput. Any load is balanced between cores.So if your responses are taking a long time, increase it.As for memory, there is a hard limit of about 150 MiB. You need at least that much to even run the container normally. Anything over that, again, depends on your data and requests per unit time. My advise would be to set it to a few gigabytes and then scale back/up to your typical use.
At the time of writing, this is not something our automated machine learning software can do.Currently, the supported problem types are regression and classification. These correspond roughly to fitting lines through data and assigning classes to them. Solutions to these problems consist of a model that generates a single number per data point (a continuous number corresponding to the regression line, or an integer corresponding the predicted class). A planning on the other hand is a solution with more structure. We do not at the moment have any way of generating models that generate such solutions.I also think that planning is unlikely to be added in the near future.Creating specific planning tools is very much something we do. But in general, planning is an unsolved problem, meaning that there is no general recipe for finding (models that propose) good solutions. That means that it doesn’t really fit with our automated machine learning software.
Already have an account? Login
Enter your username or e-mail address. We'll send you an e-mail with instructions to reset your password.
Sorry, we're still checking this file's contents to make sure it's safe to download. Please try again in a few minutes.
Sorry, our virus scanner detected that this file isn't safe to download.
