Update the multi-gpu training notebook to work with the new version of the Merlin dataloader.
Implementation Details :construction:
The new dataloader keeps the data in cupy arrays and converts each batch to torch tensors at the last minute, while the old dataloader converted everything to torch tensors as soon as the dataset was loaded from dask-cudf. Since we are using cupy in the dataloader now, we have to ensure that cupy uses the correct device by using cupy.cuda.Device. This should ideally be set in merlin.io.Dataset so users don't have to set this, but we add this to the notebook as a short-term workaround.
Depends on https://github.com/NVIDIA-Merlin/dataloader/pull/135.
Goals :soccer:
Update the multi-gpu training notebook to work with the new version of the Merlin dataloader.
Implementation Details :construction:
The new dataloader keeps the data in cupy arrays and converts each batch to torch tensors at the last minute, while the old dataloader converted everything to torch tensors as soon as the dataset was loaded from dask-cudf. Since we are using cupy in the dataloader now, we have to ensure that cupy uses the correct device by using
cupy.cuda.Device
. This should ideally be set inmerlin.io.Dataset
so users don't have to set this, but we add this to the notebook as a short-term workaround.Testing Details :mag: