7
56
1
10
DeepSeek
K100AI docker 容器测试flavr_pytorch KERNEL VMFault !!!!
2025-03-11 14:44:01
131

[iscas@localhost ~]$ sudo docker pull image.sourcefind.cn:5000/dcu/admin/base/pytorch:2.1.0-centos7.6-dtk24.04-py310

[root@localhost /]# python3 -c "import torch;print('pytorch version:',torch.version);print('DCU is',torch.cuda.is_available())"
pytorch version: 2.1.0
DCU is True

[root@localhost /]# git clone http://developer.hpccube.com/codes/modelzoo/flavr_pytorch.git

[root@localhost flavr_pytorch]# HIP_VISIBLE_DEVICES=0 python3 main.py --batch_size 32 --test_batch_size 32 --dataset vimeo90K_septuplet --loss 1*L1 --max_epoch 200 --lr 0.0002 --data_root datasets --n_outputs 1 --num_gpu 1
Namespace(dataset='vimeo90K_septuplet', data_root='datasets', model='unet_18', nbr_frame=4, nbr_width=1, joinType='concat', upmode='transpose', n_outputs=1, loss='1*L1', lr=0.0002, beta1=0.9, beta2=0.99, batch_size=32, test_batch_size=32, start_epoch=0, max_epoch=200, resume=False, resume_exp=None, checkpoint_dir='.', load_from=None, pretrained=None, exp_name='exp', log_iter=60, num_gpu=1, random_seed=12345, num_workers=16, use_tensorboard=False, val_freq=1, cuda=True)
Building model: unet_18
Preparing loss function:
1.000 * L1
Invalid address access: 0x7fe14eceb000, Error code: 1.

KERNEL VMFault !!!! <<<<<<

PID: 878, SIGNAL: 0 !!!! <<<<<<
=========> HOSTQUEUE <0x409a2ae0>: VMFault HSA QUEUE ANALYSIS <=========
HOSTQUEUE <0x409a2ae0>: get hsa queue W/R ptr: write index: 194, read index: 193
HOSTQUEUE <0x409a2ae0>: >>>>>>>> DUMP KERNEL AQL PACKET <<<<<<<<<
HOSTQUEUE <0x409a2ae0>: header: 2818
HOSTQUEUE <0x409a2ae0>: setup: 3
HOSTQUEUE <0x409a2ae0>: workgroup: x:16, y:16, z:1
HOSTQUEUE <0x409a2ae0>: grid: x:4096, y:32, z:1
HOSTQUEUE <0x409a2ae0>: group_segment_size: 8192
HOSTQUEUE <0x409a2ae0>: private_segment_size: 0
HOSTQUEUE <0x409a2ae0>: kernel_object: 140604423398400

HOSTQUEUE <0x409a2ae0>: device id: 0

HOSTQUEUE <0x409a2ae0>: >>>>>>>> FIND MATCH KERNEL COMMAND <<<<<<<<<
HOSTQUEUE <0x409a2ae0>: kernel name: conv3d_fp32_bwd_common

HOSTQUEUE <0x409a2ae0>: >>>>>>>> DUMP KERNEL ARGS: size: 152 <<<<<<<<<

00 00 40 91 e0 7f 00 00 00 00 00 24 e2 7f 00 00
00 00 20 93 e0 7f 00 00 00 00 80 4c e0 7f 00 00
01 00 00 00 80 00 00 00 00 02 00 00 40 00 00 00
40 00 00 00 04 00 00 00 04 00 00 00 04 00 00 00
03 00 00 00 20 00 00 00 20 00 00 00 04 00 00 00
01 00 00 00 01 00 00 00 01 00 00 00 02 00 00 00
02 00 00 00 01 00 00 00 00 06 00 00 10 00 00 00
00 00 20 00 00 00 20 00 00 40 00 00 00 10 00 00
00 00 20 00 00 10 00 00 00 04 00 00 00 18 00 00
30 00 00 00 10 00 00 00

HOSTQUEUE <0x409a2ae0>: >>>>>>>> DUMP KERNEL ARGS PTR INFO <<<<<<<<<
HOSTQUEUE <0x409a2ae0>: ptr arg index: 0, ptr: 0x7fe091400000
HOSTQUEUE <0x409a2ae0>: origin ptr: 0x7fe091400000, size byte: 20971520

HOSTQUEUE <0x409a2ae0>: ptr arg index: 1, ptr: 0x7fe224000000
HOSTQUEUE <0x409a2ae0>: origin ptr: 0x7fe224000000, size byte: 12582912

HOSTQUEUE <0x409a2ae0>: ptr arg index: 2, ptr: 0x7fe093200000
HOSTQUEUE <0x409a2ae0>: origin ptr: 0x7fe092a00000, size byte: 20971520

HOSTQUEUE <0x409a2ae0>: ptr arg index: 3, ptr: 0x7fe04c800000
HOSTQUEUE <0x409a2ae0>: origin ptr: 0x7fe04c800000, size byte: 100663296

KERNEL VMFault Analysis END !!!! <<<<<<

Aborted (core dumped)

全部评论(3)
只看楼主
Y
yangzhong
1楼
可以换个ubuntu的镜像试试 image.sourcefind.cn:5000/dcu/admin/base/pytorch:2.1.0-py3.10-dtk24.04.3-ubuntu20.04
2025/3/12
Y
yangzhong
回复 @yangzhong
亲测,换成这个ubuntu镜像可以
2025/3/13
网
网友01FoDP
回复 @yangzhong
好的,我试试。谢谢。
2025/3/13
没有更多内容了
--
楼主
Level.4
0 帖子
楼主热帖
暂无数据