Failed to setup the std logger (Python 3.8.5)

[khirotaka]

Dear comet-ml team.

I got

COMET ERROR: Failed to setup the std logger

when using comet.ml with Python 3.8.

Also, if I ran the code using tqdm.auto.tqdm, I got

 /usr/local/Cellar/python@3.8/3.8.5/Frameworks/Python.framework/Versions/3.8/lib/python3.8/ multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appears to be 2 leaked semaphore objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d '

And I also checked on the website and found that three logs were taken per experiment, as shown in the screenshot below.

Expected behavior

1. record the results of one experiment per experiment.
2. display a summary at the end of the experiment

Environment

• Comet.ml version: 3.2.3
• Python version: 3.8.5
• OS: macOS 10.15.7
• Other: PyTorch v1.6, Poetry v1.1.0

Error messages, stack traces, or logs

user@user$ python test_code.py
COMET ERROR: Failed to setup the std logger
COMET INFO: Experiment is live on comet.ml https://www.comet.ml/XXXXXXXXXXXXX

COMET INFO: Uploading stats to Comet before program termination (may take several seconds)
user@user$ /usr/local/Cellar/python@3.8/3.8.5/Frameworks/Python.framework/Versions/3.8/lib/python3.8/multiprocessing/resource_tracker.py:216: UserWarning: resource_tracker: There appear to be 2 leaked semaphore objects to clean up at shutdown
  warnings.warn('resource_tracker: There appear to be %d '

Reproducible examples (optional)

# test_code.py
from comet_ml import Experiment
from tqdm.auto import tqdm
import torch
import torch.nn as nn
import torch.optim as optim

experiment = Experiment(project_name="test_code")

model = nn.Linear(5, 5)
optimizer = optim.Adam(model.parameters())
criterion = nn.CrossEntropyLoss()

data = torch.randn(512, 5)
targets = torch.randint(5, size=(512, ))

for epoch in tqdm(range(100)):
    optimizer.zero_grad()
    out = model(data)
    loss = criterion(out, targets)
    loss.backward()
    optimizer.step()

Additional context (optional)

There was no problem with Python 3.7.9.

[dsblank]

@khirotaka Thanks for the report and reproducible example. We're looking into this.

[dsblank]

Due to issues with the default native output logger with Python 3.8 and MacOS, we've set the default to be "simple" for now. See: https://www.comet.ml/docs/python-sdk/releases/#release-3212

We are continuing to look for a better solution.

[github-actions[bot]]

This issue is stale because it has been open 30 days with no activity. Remove stale label or comment or this will be closed in 5 days.

[github-actions[bot]]

This issue was closed because it has been stalled for 5 days with no activity.