Zihan Zhang comments

Repositories
Issues
Comments

Results 3 comments of


                                            Zihan Zhang

Enabling Apex causes Pytorch autograd to crash

Has anyone solved this issue? I also met this issue using Sagemaker container `763104351884.dkr.ecr.us-east-1.amazonaws.com/pytorch-training:1.10.0-gpu-py38-cu113-ubuntu20.04-sagemaker`

[FEA] Support for HDBSCAN membership_vector and all_points_membership_vectors

Hi, Thanks for implementing this feature! I notice that the `all_points_membership_vectors` method is written [outside](https://github.com/rapidsai/cuml/blob/branch-22.10/python/cuml/cluster/hdbscan.pyx#L791) of the HDBSCAN class, is this intended? In that case, we need to explicitly import...

[BUG] OOM error during setup, but the OS GPU memory monitor does not show an OOM condition

I have the same issue. I can train T5-large on a single 3090 Ti with batch_size=4 without OOM. However, when I tried to use DeepSpeed to speed up training (still...