pointer-generator icon indicating copy to clipboard operation
pointer-generator copied to clipboard

Published 20 hours ago • verified publisher icon abisee

the UNK problem

Open nozomi233 opened this issue 5 years ago • 4 comments

when i run "python run_summarization.py --mode=decode --data_path=/chunked/val_ --vocab_path=/vocab --log_root=*****/a/log/directory --exp_name=myexperiment" ,the generated summary is consist of UNK,could you please do me a favor?

nozomi233 avatar Aug 13 '19 08:08 nozomi233

when i run "python run_summarization.py --mode=decode --data_path=****/chunked/val__ --vocab_path=****__/vocab --log_root=_*****/a/log/directory --exp_name=myexperiment" ,the generated summary is consist of UNK,could you please do me a favor?

I have met the same problem as you? Have you fixed this problem?

wxdwlai avatar Mar 18 '20 01:03 wxdwlai

when i run "python run_summarization.py --mode=decode --data_path=****/chunked/val__ --vocab_path=****__/vocab --log_root=_*****/a/log/directory --exp_name=myexperiment" ,the generated summary is consist of UNK,could you please do me a favor?

I have met the same problem as you? Have you fixed this problem?

Have you solve the problem? I have met the problem,too.

zjshan avatar Jun 03 '20 03:06 zjshan

when i run "python run_summarization.py --mode=decode --data_path=****/chunked/val__ --vocab_path=****__/vocab --log_root=_*****/a/log/directory --exp_name=myexperiment" ,the generated summary is consist of UNK,could you please do me a favor?

I have met the same problem as you? Have you fixed this problem?

Have you solve the problem? I have met the problem,too.

Same Problem my generated summary look like this on 6 loss : [UNK] [UNK] , , , , to to to . the . the . the . the . the . the . the . the . the . the . the . the . the

maham-collab avatar Nov 02 '20 02:11 maham-collab

I am getting only [UNK] [UNK] [UNK] in generated summary on model having 4.3 loss Do we need to train for long time or is this any other issue?

Abekaesh avatar Sep 28 '23 14:09 Abekaesh