C4/W3/yolo_filter_boxes

Course Q&A Deep Learning Specialization Convolutional Neural Networks
1

Hello everybody,

I have a problem with the first graded function this week (yolo_filter_boxes). I get a TypeError: “Cannot convert 0.5 to EagerTensor of dytpe int64”.

I think the problem is “box_scores” , “box_classes”, “box_class_scores” have all the dtype int32.

But in the description of the tf.boolean_mask function, the type is also int 32.

Unfortunately, I have no idea how to fix this error.

Can someone help me?

2

If it makes you feel any better, you are far from the first person to hit this particular error. Searching for a literal piece of that error will find you lots of hits. E.g. this one or this one.

3

Thank you!

I found the mistake.