Hi JollyChimpanzee19 , what's the exact use-case?

i have few models with final evaluation reported results and wish to compare between them

Hi JollyChimpanzee19 ,

From the UI you can choose 2 or more tasks and click the compare button, from the code you can use task.get_last_scalar_metrics() for getting the last results of a task as a dictionary. what do you think?

is there any way to automatically get few last scalars (for example the loss in the picture) as a table rather than diff presentation? or i have to use the python code?

The Hyperparameter Optimizer can give you such table, but I’m not sure this is what you are looking for ( https://allegro.ai/clearml/docs/docs/examples/frameworks/pytorch/notebooks/image/hyperparameter_search.html and https://medium.com/pytorch/accelerate-your-hyperparameter-optimization-with-pytorchs-ecosystem-tools-bc17001b9a49 )