sheffieldnlp / deepQuest

Framework for neural-based Quality Estimation
https://sheffieldnlp.github.io/deepQuest/
BSD 3-Clause "New" or "Revised" License
42 stars 13 forks source link

Seed value #9

Open shuokabe opened 5 years ago

shuokabe commented 5 years ago

Hello all,

I am currently using deepQuest's BiRNN model and I have a question about it. When I execute the code with the same configuration file two different times, I have different metrics values in the end, which are close but are still different. Is it normal? If so, is there any possibility to "stabilise" those values? For the record, I kept the "rnd_seed=8" as in the example train-test-docQEbiRNN.sh file, but I still have different values when I execute the programme.

Best, Shu