C3W2 data_generator

Course Q&A Natural Language Processing NLP with Sequence Models
1

Unable to resolve the following error. Need help please

image
image889×335 21.3 KB

2

Hi Hajra.

I looks like your mask variable contains ‘object’ type not ‘bool’. You can print() your mask before mask_np_arr to check what are you passing to np.array(mask)

3

did that, the type is <class, int>

4

the same error is raised for the batch also. The type is int in both cases

5

If you print your variables, you should get:

--------------------------------------------------
batch:
[[49, 50, 51, 52, 53, 54, 55, 56, 57, 1], [50, 51, 52, 53, 54, 55, 56, 57, 48, 1]]
batch_np_arr:
[[49 50 51 52 53 54 55 56 57  1]
 [50 51 52 53 54 55 56 57 48  1]]
--------------------------------------------------
mask:
[[1, 1, 1, 1, 1, 1, 1, 1, 1, 1], [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]
mask_np_arr
[[1 1 1 1 1 1 1 1 1 1]
 [1 1 1 1 1 1 1 1 1 1]]
6

yes i get this for batch and mask but the np.array functional call raises error

7

Please send me your notebook, I will look into it

8

Thanks. but how do i do that

10

when I run print(type(mask)) I get <class ‘list’> and with
print(type(mask_np_arr)) I get <class ‘jaxlib.xla_extension.DeviceArray’>

11

Hi! Have you figured out where was the mistake? I’m also facing this problem

12

Yes I have completed the course. I can help you on that but unable to recall instantly what the issue was and how it went away. Let me get back to you in a while inshaAllah

13

The error is caused by inconsistent size of tensors. Check if the tensors in batch have been padded to max_length correctly.

14

Sir,

I am getting the same error as shown in screenshot.

I am getting the same output as expected.

Kindly help in resolving the issue.

Regards

15

Hello Hajra

I am getting the same error as yours, how did you resolved it if you can tell.

I am getting the same out as expected.

Thanks

16

I would be eager to help but I completed this course way back and cant immediately recall what the error was and how it was resolved.

17

Ok thanks