coleridge

Cuda Out of memory with DKMPalettizer

1

## ❓Question I'm trying to run training-time palletization with n_bits=4. I always get an out-of-memory error on the first step (despite reducing batch size down to one.) A few details:...

I've been using the profiler on a TPU VM with Pytorch following this [guide](https://cloud.google.com/tpu/docs/pytorch-xla-performance-profiling-tpu-vm#tpu-vm) The overview page shows zero "device compute" in the graph and summary. I notice this is...