artitw / bert_qa Goto Github PK

For question generation, how did you selected the answers?

I can't import bert_qa .

When I import library "from bert_qa import squad" , I got the following error:
ImportError: cannot import name 'network' from 'tensorflow.python.keras.engine' (/usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/engine/init.py)
On GoogleColab
Can you help me fix this problem? Thank you!

I'm running standalone server so that I would like to specify the path of the model and data for the api. How can I make it? Thanks.

While I was running BERT, suddenly I got the following error message.
'''
inputs, attrs, num_outputs)
tensorflow.python.framework.errors_impl.InvalidArgumentError: No OpKernel was registered to support Op 'NcclAllReduce' used by {{node while/body/_1/AdamWeightDecay/NcclAllReduce}} with these attrs: [reduction="sum", shared_name="c1", T=DT_FLOAT, num_devices=2]
Registered devices: [CPU, GPU, XLA_CPU, XLA_GPU]
Registered kernels:

     [[while/body/_1/AdamWeightDecay/NcclAllReduce]] [Op:__inference_train_steps_46285]

'''
I don't know what does it mean. Please Help..

Hi !! First of all I would like to congratulate you on the project..... It increases the productivity of a BERT-SQUAD application. Could you extract a requirements.txt?
I believe that the problem is related with TensorFlow version.
Thanks

Please provide details on how to train on custom QA dataset

Does this solve question generation problem?