Mike Izbicki

@ademuro20 The error you are seeing is due to incompatible versions and I don't think it's related to @clayton-north 's error. See #40. @clayton-north There is a dropdown menu in...

The `No checkpoint was found` message means that tensorboard is working correctly, but there is no data for it to load. In the original command you gave, you had both...

The error message is caused by pytorch's `expand` function. See the docs at: https://pytorch.org/docs/stable/tensors.html#torch.Tensor.expand `expand` is related to the shapes of tensors, and my guess is that the shapes of...

`scores` should be a vector and not a higher order tensor. The sample code has a correctly shaped `scores` vector that you can look at.

Yes, I'll still have office hours at the regular times. And I'll be available at other times by appointment.

`--hidden_layer_size` and `--num_layers` are not used by the bert model, and so changing those won't have an effect on your model. Your other hyperparameters seem reasonable. If your getting accuracies...

Hmm.... that's odd. I'd need to see the tensorboard output to say for sure. Can you paste a link to tensorboard.dev?

Two things jump out to me: 1. Your `grad_norm` is growing really large, and this can result in the accuracy decreasing after it gets too big. Applying gradient clipping may...

Reopenning to make it easier for others to find.

My guess is that you have an older version of transformers installed that uses a different interface for that function. Here's the version info I used to run the code:...