This reminds me of an issue solved in v1.13.2
Hi @<1665891238055776256:profile|AppetizingKoala83> , which ClearML SDK version are you using?
Upgrading referencing/attrs resolved my issue
Can you try to upgrade your referencing/attrs version?
I ran a test job and got error on clearml import line:
Environment setup completed successfully
Starting Task Execution:
Traceback (most recent call last):
File "gpu.py", line 3, in <module>
from clearml import Task
File "/opt/conda/lib/python3.8/site-packages/clearml/__init__.py", line 5, in <module>
from .task import Task
File "/opt/conda/lib/python3.8/site-packages/clearml/task.py", line 40, in <module>
from .backend_api.services import tasks, projects, events
File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/__init__.py", line 1, in <module>
from .session import Session, CallResult, TimeoutExpiredError, ResultNotReadyError, browser_login
File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/session/__init__.py", line 1, in <module>
from .session import Session, browser_login
File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/session/session.py", line 38, in <module>
from .request import Request, BatchRequest # noqa: F401
File "/opt/conda/lib/python3.8/site-packages/clearml/backend_api/session/request.py", line 9, in <module>
from referencing.exceptions import Unresolvable
File "/opt/conda/lib/python3.8/site-packages/referencing/__init__.py", line 4, in <module>
from referencing._core import Anchor, Registry, Resource, Specification
File "/opt/conda/lib/python3.8/site-packages/referencing/_core.py", line 87, in <module>
class Specification(Generic[D]):
File "/opt/conda/lib/python3.8/site-packages/referencing/_core.py", line 113, in Specification
] = field(alias="anchors_in")
TypeError: field() got an unexpected keyword argument 'alias'
My job is:
import torch
import time
from clearml import Task
task = Task.init(project_name='David', task_name='My GPU Experiment')
# check if GPU is available
device = "cuda" if torch.cuda.is_available() else "cpu"
print(device)
# create two tensors on the GPU
x = torch.rand(5, 5, device=device)
y = torch.rand(5, 5, device=device)
# add the two tensors
z = x + y
print(z)
time.sleep(60*60*60)
This reminds me of an issue solved in v1.13.2
Hi @<1665891238055776256:profile|AppetizingKoala83> , which ClearML SDK version are you using?
Upgrading referencing/attrs resolved my issue
Can you try to upgrade your referencing/attrs version?