Hello beautiful people of ClearML
I have a dumb question for you today.
I set up a self-hosted clearml-server. I am trying to reach artifacts (specifically auto-magically captured and saved models by clearml) from another task. However, apparently it saves the model within the task's docker container and stores it at: file:///root/.clearml/venvs-builds/3.8/task_repository/repo.git/task/task_model/models/1000.pth . Honestly, I thought clearml would upload it to a file storage (e.g. the file storage that is used with clearml-data) and stores it with task id, but apparently it does not. What is the way to go to save a task's output model to a local file storage?
hey ZanyPig66
Have you set up development.default_output_uri into the configuration file ; when you init your task you add the parameter output_uri=True.
You can bind a local volume to the docker container and make the output_uri point to it
Hi ZanyPig66 ,
I assume you're using torch.save() to save your model? A good place to start is with David's suggestion with specifying output_uri = True in the Task.init() code.
