Skip to content

Conversation

RobertStanforth
Copy link

No description provided.

@google-cla google-cla bot added the cla: yes label Jun 25, 2021
Copy link
Collaborator

@apaszke apaszke left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

While I think that having both approaches is nice, it would be great to deduplicate the files a little bit. Maybe let's only keep one instance of the training loop?

' `minibatch * minibatches = batch`

def split (x: batch=>v) : minibatches=>minibatch=>v =
for b j. x.((ordinal (b,j))@batch)
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Might make sense to assert that minibatches * minibatch = batch, because this function will happily throw away data.


def trace_map (f : inp -> out) (tracer : Layer ph inp) : Layer ph out =
(AsLayer ps vs tracer') = tracer
%instance vsHint = vs
Copy link
Collaborator

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

What do we use those vector space constraints for?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

Projects

None yet

Development

Successfully merging this pull request may close these issues.

2 participants