Owen Vallis

Thanks for the details, I'll take a look and see if I can repro the issue on my side.

Hi [Lunatik00](https://github.com/Lunatik00), apologies for the slow response. We currently support loading custom data using the MultiShotMemorySampler. The data is loaded into memory and properly sampled over the classes to ensure...

Thanks for submitting the issue @mstfldmr. Do you have a simple example I can use to try and repro the issue? I can also try and repro this using our...

Hi @mstfldmr, sorry for the delay here. I'll try and get to this this week.

Looking into this now as it also looks like there is a breaking change in 2.8 where they removed `Optimizer.get_weights()` (see https://github.com/keras-team/tf-keras/issues/442). That issue also mentions that SavedModel didn't properly...

So we do provide the [tfrecord sampler](https://github.com/tensorflow/similarity/blob/master/tensorflow_similarity/samplers/tfrecords_samplers.py) for handling datasets that are too large to fit in memory. There are some quirks to setting up the TFRecords, i.e., this sampler...

n/m, I had a path error to the mnist data. I had it pointing at the mnist.pbtxt that came with the deepnet repo, but had the rest of the mnist...

re opened this issue as the error keeps popping up intermittently. I'm running the code using python 3 (based off of the corvalius fork), and a macbook pro with an...