ner.embed.pth about dsner-pytorch

[nooralahzadeh]

Hi,
It is created in first run to avoid the time of creating embedding for each words from pre_trained model in next runs. As you can see here :

DSNER-pytorch/src/dsner.py

Line 249 in d6683dc

emb_file = os.path.join(args.data, 'ner.embed.pth')

And yes, it is char+vec !

[zhuzhikui]

Wow! Thanks for relying me. I get it. But when I run my dataset some error happen. And when I run part of my data, it works well. The error is as following:

[nooralahzadeh]

You'r Welcome.
Did you do preprocessing? It seems some of the tags in dataset are empty ''!

[zhuzhikui]

I'll check it, thank you!

[nooralahzadeh]

I am closing this case, since you didn't update me. Please reopen it if you still have a problem with that!