Results from exercise 2.11 seem wrong, as the performance on the test set is very low:
print(eval_viterbi_test)
>>> 0.027530204000792237
I tested code on the current master branch.
I'm not sure what the expected performance is. Discussed the issue with Anton Belyy (@AVBelyy) who brought it to the attention of some of the monitors, just wanted to find out what the error is.
Results from exercise 2.11 seem wrong, as the performance on the test set is very low:
I tested code on the current master branch.
I'm not sure what the expected performance is. Discussed the issue with Anton Belyy (@AVBelyy) who brought it to the attention of some of the monitors, just wanted to find out what the error is.