You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Audit question for Exercise 3 includes two parts. The first (shape of the embeddings of each word) is included in the task, but the second (the first 20 values of the embedding of laptop) is not specified in task description.
The text was updated successfully, but these errors were encountered:
If you have calculated the embeddings of each word, you should be able to verify the first 20 values of the embedding related to laptop word :)
The audit is not a machine test therefore, you should be able to show the results to the person that is going to audit you.
In any case, I am reviewing the exercise to add the version of the language model to use to avoid mismatches in the values
I reported this because in 01 program so far the audit questions verify the task results and the required output (in this case printing out 20 values of the embedding) is usually already included in the task description. This makes the auditing process also much smoother, because you have all the elements for audit already present in the code.
It has been rather strange to solve tasks in this branch with always keeping the audit questions at the side, because there is no other way of knowing what will be actually checked in the audit. I believe this branch deviates in that sense from the rest of 01 program.
nlp-spacy
Audit question for Exercise 3 includes two parts. The first (
shape of the embeddings of each word
) is included in the task, but the second (the first 20 values of the embedding of laptop
) is not specified in task description.The text was updated successfully, but these errors were encountered: