MY RMSE loss for both training and validation, is coming very high ! I tried reducing the alpha value but can’t correct it. please help me with what can I do to my model .
No idea what you consider “very high”, but the common mistake is to square the results.
squared=False as an argument to your loss function.
i did that previously but still the loss showjn is very high
Define what you mean by “very high”.
it is 27k for both. Also i want to give all inputs for training as told in the assignment but the method is not known .
I wouldn’t consider this value “very high” for this task.
I’ve got around 22k and 28k - you need to remember that the target value is in hundreds of thousands of dollars, so the loss will scale accordingly.
Still the assignment says
Predictions on the training and validation aren’t done correctly
NotebookRuntimeError : name ‘val_preds’ is not defined
The value of ‘weights’ was incorrect
NotebookRuntimeError : name ‘weights’ is not defined
This is something completely unrelated to the title of this topic.
Anyway, this means that you didn’t run all the cells in the notebook. Restart it, run all cells.
But the first time before submitting I ran all the cells . Why so from their side?