Benjamin Szubert

Hi, thanks for the feedback and the link to the paper, it was a good read! I like the idea and think it would be interesting to try out and...

Hi thanks for raising this, we agree that this would be a useful feature to have. We are looking into whether we could support the use of generators with ivis.

Just as an update to this, using a generator to train ivis is difficult since the triplet sampling algorithm may need to retrieve KNNs or negative points that are not...

> I don't know if this is a custom training strategy, but if you use the keras' fit() method, my understanding is that "batch shuffle" doesn't shuffle rows inside batches,...

Hi, we have recently introduced initial support for training on arbitrary out-of-memory datasets using ivis by formalizing the interface that input data must conform to. Ivis will accept `ivis.data.sequence.IndexableDataset` instances...

> This looks really nice - > > Would be good to have a nice mechanism of creating remotes directly through bravetools without manual copy of certs. Thoughts? Yes, I...

> Would be good to have a nice mechanism of creating remotes directly through bravetools without manual copy of certs. Thoughts? Added remote management through CLI in latest commits. The...

> I think this is already handled in current master: see #66 It was, but my changes in this branch altered the logic and introduced this issue here. I think...

> The solution does seem a little brittle- it may be worth cleaning up the logic for default Bravfiles (maybe it can be handled in the Bravefile.Load so it's all...

On reflection, making build/base/deploy accept array args somewhat complicated the code and encouraged bad practices. For deploy and build there's the issue of context - often files will be copied...