Trains a RNN and prints some helpful statistics. Returns a trained model.
1 | Train.RNN(model, model.par, x_train, y_train, action)
|
model |
A RNN e.g. given by |
model.par |
Parameters of RNN, e.g. given by |
x_train |
A 3D array in the form c(minibatch.size, max.length.sequence, outputs) |
y_train |
In the case of !single.dimensional: A Matrix with as much columns as output parameters (e.g. action parameters). For each state (x_train)/action combination(column of y_train) this value determines the target. If single.dimensional this is a simple vector with the result and the function will build a sufficient x/y pair based on action. |
action |
Necessary if not single dimensional - actions which led to y based on x. |
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.