Song Yinghao

icon company JiLin University icon location JiLin University icon location I am Song Yinhao, a student in JiLin University. I want to study in Github and try to make some contributions to the community.

Results 4 issues of Song Yinghao

RuntimeError: Error building extension 'fused_adam'

Hi, I am trying the 'cifar10_deepspeed.py' [here](https://www.deepspeed.ai/tutorials/cifar-10/) on a single node (2x3090). When I run the command below: ```bash deepspeed cifar10_deepspeed.py --deepspeed_config ds_config.json ``` The bug occurs: ```bash [2023-12-31 16:33:59,636]...

bug
training

how to understand large kernel convolution decomposition？

![image](https://user-images.githubusercontent.com/60497059/203268449-f6aff9b7-923c-49bd-92de-5f948d9221e5.png) Thank you so much for this enlightening and inspiring work.I don't quite understand why **LKC(i.e. Large Kernel Convolution) = DW-Conv + DW-D-Conv + 1x1Conv.** My current understanding can refer...

torch.distributed.DistBackendError: NCCL error in: ../torch/csrc/distributed/c10d/ProcessGroupNCCL.cpp:1333, remote process exited or there was a network error, NCCL version 2.18.6

4 comment

Hi, I successfully ran the ['cifar10_deepspeed.py' ](https://www.deepspeed.ai/tutorials/cifar-10/)example on a single node (2xNVIDIA 3090). Now I want to run the same program on multi-nodes (2 nodes each have 2 3090s.). I...

Are all of your 8 NVIDIA 3090 on a single node?

I am new to deepspeed. I noticed there are no '--hostfile' arg in the train command. : )