<EshaanAgarwal[m]>
<jonpsy[m]> "Still stuck?" <- almost done. actually the weight-translator repo's cpp part was not adapted to current codebase (we had some refactor in ann part) of mlpack so i was facing some difficulties. working on making things work
<jonpsy[m]>
Got it, I'm not sure in this area myself. Perhaps zoq can help you here.
jjb[m] has joined #mlpack
<jjb[m]>
For the PyTorch portion, you’ve flipped the network into `eval()` mode or are you using the with `no_grad()` option?
<jjb[m]>
zoq do you folks have an ETA on `mlpack` v4 or a new ensmallen release?
<rcurtin[m]>
We're close!! I think I have maybe one or two more PRs before all the refactoring is done. I think by the end of September is reasonable
<jjb[m]>
ryan awesome! Do you need any 👀 on the R portion of `mlpack`?
<rcurtin[m]>
Maybe? I actually think not too much will change there. `libmlpack.so` is gone, but there is still a library for each individual binding, so that is unchanged.