Issues
- 6
TODOs
#13 opened - 0
- 1
how to predict answers for custom question and context by reusing loaded model
#40 opened by SachinIchake - 0
- 1
- 0
problem about highwaynet
#55 opened by kyang888 - 0
conv_block problem
#54 opened by liying123456 - 0
Parameter setting problem
#53 opened by liying123456 - 1
- 0
how to train by changing/adding batch_size?
#52 opened by shahhet0205 - 1
How do I resume training off a checkpoint?
#51 opened by NatCusack - 2
- 1
how to adapt it for squad2.0 dataset?
#48 opened by Jasperty - 7
Trying to fine tune with different data, But getting dimensionality mismatched for tensor
#18 opened by brojokm - 9
Training stops after some time
#47 opened by emreorta - 2
Train Models With Macs
#42 opened by ahmedselim2017 - 1
mask_logits in layer.py
#43 opened by zhaoluffy - 3
layer normalization in layer?
#44 opened by PanXiebit - 1
Is this snippet in prepro.py correct
#46 opened by saxenarohit - 4
A tool to generate training data
#30 opened by pencoa - 3
- 1
The embedding projection
#41 opened by PANXiao1994 - 1
Unable to preprocess data
#34 opened by charu-iitm - 27
Report the results
#15 opened by jasonwbw - 1
mask_logits function
#36 opened by shwetgarg - 2
Try to test ELMO language model
#38 opened by tuan3w - 2
- 0
Trainable Embedding for OOV words
#37 opened by PANXiao1994 - 5
InvalidArgumentError during evaluation
#27 opened by maciejbiesek - 4
inconsistency in predictions
#33 opened by nehaboob - 0
Preprocessing
#35 opened by tensorhunter - 2
How can I do fine tuning using QANet?
#32 opened by yexing99 - 2
How to train in Multi GPU
#31 opened by deepakkumar1984 - 5
Can it support Chinese?
#28 opened by jasperzhong - 1
How to start training?
#29 opened by pencoa - 2
Speed ?
#25 opened by chenghuige - 6
training on "Answer not available" ?
#24 opened by nehaboob - 2
Memory Issue
#23 opened by amandalmia14 - 5
dev set evaluation
#21 opened by nehaboob - 1
tensorflow not in requirements.txt
#22 opened by ThommyH - 1
Is it a better way to use conv layer in highway or encoder block feed forward network rather than dense layer?
#20 opened by hackiey - 1
Results of the original paper
#19 opened by yizhongw - 2
- 7
word_embed.json missing?
#10 opened by theSage21 - 0
No such file or directory: 'demo.html'
#11 opened by rpanchal1996 - 3
- 1
possibly insufficient driver version:
#5 opened by hxsnow10 - 8
- 3
about embedding matrix structure
#7 opened by tongjinle123 - 14
Train with M40 card but got OOM message
#8 opened by chesterkuo