Support loss over all tokens to be back-propagated during training

[maydixit]

Support loss over all tokens to be back-propagated during training

Intended to resolve issue: #17

[google-cla]

Thanks for your pull request! It looks like this may be your first contribution to a Google open source project. Before we can look at your pull request, you'll need to sign a Contributor License Agreement (CLA).

View this failed invocation of the CLA check for more information.

For the most up to date status, view the checks section at the bottom of the pull request.

[iislucas]

As per our discussion, the key things to add to this are:

• Causal masking so that training doesn't try to use future tokens to predict past ones.
• A script that validates it "works" (we should see clearly better loss curves using this).