Why doesn't the counter in data_iter increase?

codertimo / BERT-pytorch

Google AI 2018 BERT pytorch implementation

Apache License 2.0

6.11k stars 1.29k forks source link

Why doesn't the counter in data_iter increase? #39

Closed nateraw closed 5 years ago

nateraw commented 5 years ago

I am currently playing around with training and testing the model. However, as I implemented the test section, I'm noticing that during the LM training, your counter doesn't increase when looping over data_iter found in pretrain.py. This would cause problems when calculating the average loss/accuracy, wouldn't it?

nateraw commented 5 years ago

Looks like this may be related to #32

codertimo commented 5 years ago

I have no idea.. what version did you used?

nateraw commented 5 years ago

alpha0.0.1a4

nateraw commented 5 years ago

@codertimo I just read over some more of the comments on #32...I'll update to latest version and then try again. Will report back ASAP

codertimo commented 5 years ago

It seems like you are not the using original code. Did you changed something? The loss are not looks like 0.0.1a4 version

nateraw commented 5 years ago

I had changed the format of the output to dynamically work with some tests I was running. No actual changes to logic were made (but I could be wrong, haha). I'll close this for now and will reopen an issue if I find it continues (on your unaltered code).