Dan Ofer

Would it be possible to provide an example of how to run the vocab/tokenizing in advance on this data, including the expected output sentencepiece vocab?

The problem is that no features are extracted. (Not sure why). Have you tried extracting features using the "file" vs "dir" option? I'll be uploading an update In the next...

What OS are you using? Try using the absolute file path. The update has been implemented.

Also - the Tail command outputs lines ; It could have messed up the fasta formated files. https://en.wikipedia.org/wiki/Tail_(Unix) On Wed, Sep 9, 2015 at 4:08 PM, cnjr2 [email protected] wrote: >...

There might be an issue with the "dir" option. (We didn't use it while writing the articles). The program is complaining that it's not getting labels/class for the sequences. Possibly...

With --classType file - your "classes" should be in 2 seperate multifasta files (each containing all the sequences belonging to a class. [without "overlapping"/duplicates]. e.g. (In case you have 2...

Hi, I'm afraid that I'll be unable to debug the the issue, as I'll be unavailable for the next month. I suggest forking from the earliest commit in the meantime....

Worst case, just use the features generation methods/featureGen.py Good luck On Oct 1, 2015 7:27 PM, "cnjr2" [email protected] wrote: > Thanks for the info. I will give it a shot!...

Yay. Thank you very much :-) בעתיד, תהיה אתה מעוניין\יכול עדיין להמשיך להשתתף עוד קצת בקשר לזה? (על דברים פשוטים יחסית. תיעוד )/readme), בדיקה ל exceptions וכו). אני לא ליד...

Ok. On May 27, 2015 5:52 PM, "Michael Doron" [email protected] wrote: > אין את השמירה של המודל או הנרמול, > ואני חושב שזה יהיה עמוס לי מדי, עם המחקר אצל...