【AICC】notebook生成json配置文件,出现PermissionError: [Errno 13] Permission denied: '/etc/hccn.conf'报错
收藏回复举报
【AICC】notebook生成json配置文件,出现PermissionError: [Errno 13] Permission denied: '/etc/hccn.conf'报错
t('forum.solved') 已解决
发表于2023-03-07 17:12:55
0 查看

使用成都智算中心notebook进行模型训练,由于notebook多卡训练需要配置rank_table,运行hccl_tools.py生成json,出现如下报错:

 

[ma-user hccl_tools]$python hccl_tools.py --device_num "[0,8)" 

start hccl_tools.py 

visible_devices:['0', '1', '2', '3', '4', '5', '6', '7'] 

server_id:172.16.0.77 

device_num_list: [0, 1, 2, 3, 4, 5, 6, 7] 

/bin/sh: hccn_tool: command not found 

Failed to call hccn_tool, try to read /etc/hccn.conf instead 

Failed to read /etc/hccn.conf 

Traceback (most recent call last): 

  File "hccl_tools.py", line 116, in main 

    device_ips[str(device_id)] = ret[0].split(":")[1].replace('\n', '') 

IndexError: list index out of range 

 

During handling of the above exception, another exception occurred: 

 

Traceback (most recent call last): 

  File "hccl_tools.py", line 120, in main 

    with open('/etc/hccn.conf', 'r') as fin: 

PermissionError: [Errno 13] Permission denied: '/etc/hccn.conf' 

 

During handling of the above exception, another exception occurred: 

 

Traceback (most recent call last): 

  File "hccl_tools.py", line 163, in <module> 

    main() 

  File "hccl_tools.py", line 128, in main 

    raise SystemError("Failed to find information for hccl") 

SystemError: Failed to find information for hccl 

我要发帖子