thomas chaton

The conversation is there: https://github.com/rusty1s/pytorch_geometric/issues/602

I am also part of Helix and work with Loic on spg3.0 :)

https://github.com/rusty1s/pytorch_geometric/issues/602 Based on this one, what s missing is the modification made in learning/main.py. If needed, I would provide you with them :)

I think pytorch_geometric should also be an option.

Dear @cbalioglu, Very cool proposal ! I would have several questions: * Q1: Do you envision the store be used with sharded models ? ```py store.save("my-model", m.state_dict(), rank=..., world_size=...) ```...

Hey @[raphaelauv](https://github.com/raphaelauv) @iudeen Thanks for answering! Could you share some associated Celery examples to do so? IMO, I think Batching would fit very nicely as a MiddleWare if the API...

Dear @fellenB, Concerning DCP: You are referring to Deep Closest Point: Learning Representations for Point Cloud Registration right ? We would appreciate if you can add it to TorchPoints3d by...

After chatting with @rohitgr7, Here is the API design specification for the `BatchSizeFinder`. ### With the tune method. User story: As a user, I want to fit the best batch...

Hey @rohitgr7. Any updates on this one ?

Hey @awaelchli. What is the status on this PR ?