training problem

[shangqing-liu]

I have updated the version of tensorflow to nightly 1.6 and python to 3.5. However, when I run the scripts,  a problem occurred. Anyone has the similar problem?

[shangqing-liu]

@nave01314 I have just test the code, so I run the parameter supplied. python -m nmt.nmt \ --attention=scaled_luong \ --src=vi --tgt=en \ --vocab_prefix=/tmp/nmt_data/vocab \ --train_prefix=/tmp/nmt_data/train \ --dev_prefix=/tmp/nmt_data/tst2012 \ --test_prefix=/tmp/nmt_data/tst2013 \ --out_dir=/tmp/nmt_attention_model \ --num_train_steps=12000 \ --steps_per_stats=100 \ --num_layers=2 \ --num_units=128 \ --dropout=0.2 \ --metrics=bleu

[oahziur]

@liunuaa Can you try clean up /tmp/nmt_attention_model and run the command again? Maybe there is some stale hparams file in your out_dir.

[shangqing-liu]

@oahziur Ok, thanks, I will try.

[lmthang]

Have you resolved the issue? Close for now.