Hi,
I am getting an error when running my model in the assignment. I have finished all the graded functions in the assignment, and made sure that my function returns the same expected values as it says in text. All my function seems to be working just fine, and return all values corretly
Then, whenever I try to train my model, the model stops after the first epoch and gives me an error message: “UnimplementedError”
At first, I though that this was related to mistake in my function “parse_data_from_file” function, and that the issue could be that I did not correctly categorize my labels as 0 or 1 integers.
Here is a snip of my parse_data_from_file function:
And here is snip of testing my function with the expected output - as seen, my function returns exactly the same as the expected output:
Can anyone help or share good ideas on how/why this error occurs?
regards,
Michael
Please confirm that the following are numeric::
- In the call to
model.fit, training data and training labels
- In the call to
model.fit, validation data and validation labels
- Embedding matrix which serves as weights to the embedding layer.
Hi,
Thx for getting back to me, I have now tried testing the above data inputs for their respective types using the following little function i wrote:
As seen, all the datatypes are either int32 or int64 except for my validation labels that have the datatype U314… I do not know what this U314 is, do you know this datatype? and do you think this might be the one causing troubles?
Nevermind! I found the mistake… I had simply referenced sentences rather than labels in my function parsing the data. That way I had managed to create training/test split with the correct dimension, but simply having sentences rather than labels in the val_labels object.
Silly mistake by me, but thanks a lot for pointing me in the right direction
best regards,
Michael
1 Like
