Dead kernel error in Week 2 assignment

Course Q&A TensorFlow: Data and Deployment Specialization Data Pipelines with TensorFlow Data Services
1

I am trying to train the Dogs & Cats model in the Week 2 assignment for the TF Data & Deployment course. However, each time I try to start a training run the kernel dies.

The other cells in the notebook all run without any errors.

Does anyone know what the problem could be?

Thanks,
Brendan.

1 Like
2

Hi, @Brendan_Horsfield and welcome to the community! :wave:

Can you DM me your notebook, please? :grinning_face_with_smiling_eyes:

1 Like
3

I got exactly same issues. I submitted the assignment and I got 100 which I assume all my previous steps were correct but the kernel always dies when I tried to train the model.

1 Like
4

Hi @Huy_Tran! :wave: Can you DM/share the notebook with me?

1 Like
5

I have the same problem, the kernel keeps dying when I run steps2-split dataset.

1 Like
6

I got the same problem. The kernel keeps shutting down at each try. My notebook file is attached.

Edited by the moderator: removed the link to the notebook file

1 Like
7

I fixed my issue. Thanks!!

1 Like
9

Hi @Yannick_ASSEKE :wave:

Do you remember how you solved the issue? Can you please help @Mufidatul_Ngazizah (in DM)?

Thank you :smiley:

1 Like
10

How you solved the issue?

1 Like
11

I faced the same problem and i found that because the invalid index like this
splits = ['train[-95%:]']
So i used splits = ['train[95%:]'] instead

2 Likes