Блог

Я хотел бы использовать visdom для визуализации результатов алгоритм глубокого обучения, который обучается на удаленном кластерном сервере. Я нашел ссылку, в которой пытался описать правильный способ настройки всего в скрипте slurm.

 python -u Script.py --visdom_server http://176.97.99.618 --visdom_port 8097
 

Я использую свой ip и 8097 для подключения к удаленному кластерному серверу:

 ssh -L 8097:176.97.99.618:8097 my_userid@r@my_server_address
 

У меня есть следующие строки кода:

 import visdom
import numpy as np
cfg = {"server": "176.97.99.618",
       "port": 8097}
vis = visdom.Visdom('http://'   cfg["server"], port = cfg["port"])
win = None

#Plotting on remote server
def update_viz(epoch, loss, title):
    global win

    if win is None:
        title = title

        win = viz.line(
            X=np.array([epoch]),
            Y=np.array([loss]),
            win=title,
            opts=dict(
                title=title,
                fillarea=True
            )
        )
    else:
        viz.line(
            X=np.array([epoch]),
            Y=np.array([loss]),
            win=win,
            update='append'
        )

update_viz(epoch, elbo2.item(), 'ELBO2 Loss of beta distributions')
 

Я получил эту ошибку:

 Setting up a new session...
Traceback (most recent call last):
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connection.py", line 174, in _ne
w_conn
    conn = connection.create_connection(
  File "/anaconda3/lib/python3.8/site-packages/urllib3/util/connection.py", line 96, in
 create_connection
    raise err
  File "/anaconda3/lib/python3.8/site-packages/urllib3/util/connection.py", line 86, in
 create_connection
    sock.connect(sa)
TimeoutError: [Errno 110] Connection timed out

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py", line 699, in
 urlopen
    httplib_response = self._make_request(
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py", line 394, in
 _make_request
    conn.request(method, url, **httplib_request_kw)
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connection.py", line 239, in req
uest
    super(HTTPConnection, self).request(method, url, body=body, headers=headers)
  File "/anaconda3/lib/python3.8/http/client.py", line 1255, in request
    self._send_request(method, url, body, headers, encode_chunked)
  File "/anaconda3/lib/python3.8/http/client.py", line 1301, in _send_request
    self.endheaders(body, encode_chunked=encode_chunked)
  File "/anaconda3/lib/python3.8/http/client.py", line 1250, in endheaders
    self._send_output(message_body, encode_chunked=encode_chunked)
  File "/anaconda3/lib/python3.8/http/client.py", line 1010, in _send_output
    self.send(msg)
  File "/anaconda3/lib/python3.8/http/client.py", line 950, in send
    self.connect()
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connection.py", line 205, in con
nect
    conn = self._new_conn()
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connection.py", line 186, in _ne
w_conn
    raise NewConnectionError(
urllib3.exceptions.NewConnectionError: <urllib3.connection.HTTPConnection object at 0x7ff292f14d00
>: Failed to establish a new connection: [Errno 110] Connection timed out

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/anaconda3/lib/python3.8/site-packages/requests/adapters.py", line 439, in send
    resp = conn.urlopen(
  File "/anaconda3/lib/python3.8/site-packages/urllib3/connectionpool.py", line 755, in
 urlopen
    retries = retries.increment(
  File "/anaconda3/lib/python3.8/site-packages/urllib3/util/retry.py", line 574, in inc
rement
    raise MaxRetryError(_pool, url, error or ResponseError(cause))
urllib3.exceptions.MaxRetryError: HTTPConnectionPool(host='176.97.99.618', port=8097): Max retries 
exceeded with url: /env/main (Caused by NewConnectionError('<urllib3.connection.HTTPConnection obj
ect at 0x7ff292f14d00>: Failed to establish a new connection: [Errno 110] Connection timed out'))

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
  File "/anaconda3/lib/python3.8/site-packages/visdom/__init__.py", line 708, in _send
    return self._handle_post(
  File "/anaconda3/lib/python3.8/site-packages/visdom/__init__.py", line 677, in _handl
e_post
    r = self.session.post(url, data=data)
  File "/anaconda3/lib/python3.8/site-packages/requests/sessions.py", line 590, in post
    return self.request('POST', url, data=data, json=json, **kwargs)
  File "/anaconda3/lib/python3.8/site-packages/requests/sessions.py", line 542, in requ
est
    resp = self.send(prep, **send_kwargs)
  File "/anaconda3/lib/python3.8/site-packages/requests/sessions.py", line 655, in send
    r = adapter.send(request, **kwargs)
  File "/anaconda3/lib/python3.8/site-packages/requests/adapters.py", line 516, in send
    raise ConnectionError(e, request=request)
requests.exceptions.ConnectionError: HTTPConnectionPool(host='192.168.2.10', port=8097): Max retri
es exceeded with url: /env/main (Caused by NewConnectionError('<urllib3.connection.HTTPConnection 
object at 0x7ff292f14d00>: Failed to establish a new connection: [Errno 110] Connection timed out'
))
Visdom python client failed to establish socket to get messages from the server. This feature is o
ptional and can be disabled by initializing Visdom with `use_incoming_socket=False`, which will pr
event waiting for this request to timeout.
Script.py:41: UserWarning: To copy construct from a tensor, it is recommended to us
e sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than
 torch.tensor(sourceTensor).
  params['w'].append(nn.Parameter(torch.tensor(Normal(torch.zeros(n_in, n_out), std * torch.ones(n
_in, n_out)).rsample(), requires_grad=True, device=device)))
Script.py:42: UserWarning: To copy construct from a tensor, it is recommended to us
e sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather than
 torch.tensor(sourceTensor).
  params['b'].append(nn.Parameter(torch.tensor(torch.mul(bias_init, torch.ones([n_out,])), require
s_grad=True, device=device)))
Script.py:292: UserWarning: To copy construct from a tensor, it is recommended to u
se sourceTensor.clone().detach() or sourceTensor.clone().detach().requires_grad_(True), rather tha
n torch.tensor(sourceTensor).
  return torch.exp(torch.lgamma(torch.tensor(a, dtype=torch.float, requires_grad=True).to(device=l
ocal_device))   torch.lgamma(torch.tensor(b, dtype=torch.float, requires_grad=True).to(device=loca
l_device)) - torch.lgamma(torch.tensor(a b, dtype=torch.float, requires_grad=True).to(device=local
_device)))
Script.py:679: UserWarning: This overload of add_ is deprecated:
    add_(Number alpha, Tensor other)
Consider using one of the following signatures instead:
    add_(Tensor other, *, Number alpha) (Triggered internally at  /opt/conda/conda-bld/pytorch
_1631630815121/work/torch/csrc/utils/python_arg_parser.cpp:1025.)
  exp_avg.mul_(beta1).add_(1 - beta1, grad)
[Errno 110] Connection timed out
on_close() takes 1 positional argument but 3 were given
Traceback (most recent call last):
  File "Script.py", line 873, in <module>
    update_viz(epoch, elbo2.item(), 'ELBO2 Loss of beta distributions') 
  File "Script.py", line 736, in update_viz
    win = viz.line(
NameError: name 'viz' is not defined
 

как я могу запустить свой сценарий построения графика на удаленном сервере? Какая командная строка кода python должна быть в моем скрипте SLURM? Как я могу сохранить график и переместить его позже на свой ноутбук с помощью scp command?

Попробуйте использовать global viz after global win line .