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
Love this repo so thanks a mil! Wanted to flag that for examples/notebooks/Tutorial-Prompt-Optimization.ipynb to work, I had to modify a line in the eval_sample function:
y = tg.Variable(y, requires_grad=False, role_description="correct answer for the query")
should actually be y = tg.Variable(int(y), requires_grad=False, role_description="correct answer for the query").
I believe the process of populating y in the class BigBenchHard ends up populating it as a numpy.int64 so that throws off the Variable declaration afterwards since it's expecting Union[str, bytes] as the data type for the value.
Hope this makes sense, and huge love and respect for your work on TG,
Mounir (mostly an eternal noob so apologies if I missed something obvious)
The text was updated successfully, but these errors were encountered:
Peeps,
Love this repo so thanks a mil! Wanted to flag that for examples/notebooks/Tutorial-Prompt-Optimization.ipynb to work, I had to modify a line in the eval_sample function:
y = tg.Variable(y, requires_grad=False, role_description="correct answer for the query")
should actually be y = tg.Variable(int(y), requires_grad=False, role_description="correct answer for the query").
I believe the process of populating y in the class BigBenchHard ends up populating it as a numpy.int64 so that throws off the Variable declaration afterwards since it's expecting Union[str, bytes] as the data type for the value.
Hope this makes sense, and huge love and respect for your work on TG,
Mounir (mostly an eternal noob so apologies if I missed something obvious)
The text was updated successfully, but these errors were encountered: