Hi there, I'm training a pytorch model and save it every epoch. It seems like the model wights are overridden and I can't choose the best model after the exp...

I wanted to suggest something. We're creating a lot of projects and it starts getting a bit difficult to navigate through them. I think an option to have a h...

Hi everyone, Additional arguments to the script execution, is it possible? how can it be done? So at the moment when my script is being executed the sys.argv...

Hi all, After solving my multiprocessing issue I've found the following issue: I have a machine with 2 GPUs. Starting an agent there specifying --gpus all di...

Hi, another question. I tried to not auto_connect_arg_parser but in that case I need to do the Task.init before I'm parsing the arguments. I don't want that ...

Hi there, I've been trying to work with trains and I wanted to save a folder as the model like when using the "transformers" library. They have this "save_pr...

Hi, I'm getting a lot of the following logs trains.frameworks - WARNING - Could not retrieve model location, skipping auto model logging I'm fine with it and...

Hi, I'm having some issues that I can't seem to find where the problem is or how to solve it. I'm running some code on the worker when I'm trying to download...

Hi there, I've encountered a problematic behavior in python. When defining an argument a default value of None connecting to trains server then copy the task...

Hi, I've recently upgraded to 0.15.1 from 0.14.2, and for some reason a code that previously worked in which I'm getting the tags of a model using InputModel...

Hi, I've found a possible bug. I'm cloning/running a project without any input model. Which is as expected. But, after I code actually start running an input...