Use Google's BERT for named entity recognition (CoNLL-2003 as the dataset).
MIT License
1.24k
stars
335
forks
source link
WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 25 vs previous value: 25. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize #87
I am doing NER using BERT for past months on google colab GPU and everything was working fine but now when I am doing same on CPUs I am getting this warning.
When am using colab GPU for training then there is no issue and no warnings of this kind. But when I am training with same data and same parameters I am getting this type of warning on running on colab CPU and on my local system's CPU and training never finishes.
WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 25 vs previous value: 25. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.
Can someone please tell me the reason why problem occurs only on CPU not on GPU keeping the configurations same?
And how can I resolve it?
By the way BERT is using AdamOptimizer and I have never modified BERT's optimizer.py .
I am doing NER using BERT for past months on google colab GPU and everything was working fine but now when I am doing same on CPUs I am getting this warning. When am using colab GPU for training then there is no issue and no warnings of this kind. But when I am training with same data and same parameters I am getting this type of warning on running on colab CPU and on my local system's CPU and training never finishes.
WARNING:tensorflow:It seems that global step (tf.train.get_global_step) has not been increased. Current value (could be stable): 25 vs previous value: 25. You could increase the global step by passing tf.train.get_global_step() to Optimizer.apply_gradients or Optimizer.minimize.
Can someone please tell me the reason why problem occurs only on CPU not on GPU keeping the configurations same? And how can I resolve it? By the way BERT is using AdamOptimizer and I have never modified BERT's optimizer.py .