Hi, so I download the bloodcell dataset from kaggle - everything was going pretty good untill I hit an error where I just can’t debug.
Essentially, I went back to the lesson1 notebook and checked the shapes of both probs and y np.arrays.
and got:
probs.shape
(2000, 2)
y.shape
(2000,)
Then went back to my own bloodcell notebook and did the same:
probs.shape
(800, 4)
y.shape
(800,)
so all looks good… Although when I run the following in my bloodcell notebook:
accuracy(probs, y)
I get an error though - is there anything glaring obvious here:
So I did some more digging and it looks like, I’ve found something that could help me, on this forum. Will give it a shot and report back if I’m successful.
Thanks radek for your reply
I figure out what it is wrong with that code.
I tried to analyse the results/look at the pictures as in lesson1
in order to display: correct/incorrect labels, most correct/incorrect labels and most uncertain labels.
Although the log_preds.shape was (482,8) the probs.shape was (482,1) due to the fact that probs = np.exp(log_preds[:,1]) - because it takes only one column from log_preds - since it is a 2 class problem.
Then this variable probs is used…