fasttrackml
fasttrackml copied to clipboard
Having mlflow.set_tracking_uri and mlflow.set_experiment causes error
mlflow.set_tracking_uri("http://localhost:5000/")
mlflow.set_experiment("exp-name")
If these are the other way around, you get a strange error
ges/mlflow/utils/rest_utils.py", line 210, in call_endpoint
response = verify_rest_response(response, endpoint)
File "/home/coder/.cache/pypoetry/virtualenvs/picassoseg-0dMKFdZz-py3.9/lib/python3.9/site-packages/mlflow/utils/rest_utils.py", line 142, in verify_rest_response
raise RestException(json.loads(response.text))
mlflow.exceptions.RestException: BAD_REQUEST: unable to parse experiment id '308770067652592657': strconv.ParseInt: parsing "308770067652592657": value out of range
I would expect this error to be handled more gracefully. There's a good chance people in the future could be using Fasttrack before mlflow (like me!) So this kind of thing will matter