Unchanged training result

[sampalomad]

Hi, I trained and test the mnmt Greman model on Multi30K several times, but every training test gave me the same BLEU precise to 0.001, and there is no fluctuation in train_loss after at any given epoch across these training logs (for example, the loss after epoch 28 is the same across all 5 training logs), is it normal?

[ozancaglayan]

Hello,

Are you fixing the seed parameter to something different than 0?

[sampalomad]

Hi,

yes, the seed is set to 1234

[ozancaglayan]

That is the reason. Same seed will produce same results. You can set it 0 for random seeds.