Sweep agent: stop it from synching model artifacts?

Hi, I am running hyperparameter sweeps with W&B, and training agents like this:

            function=functools.partial(run_train, args),

Where run_train uses a pytorch-lightning trainer to train a model. However, my model runs create huge artifacts by seeming to synch a model to the W&B app everytime evaluation happens. I do not understand why it is happening or how to make it stop?

Can someone show me how to stop synching massive artifacts?

Hi @adamits could you possibly share the run_train function? Something in the training is being set to upload the model to W&B.

Thank you,

@adamits since we haven’t heard back I’ll close this request but feel free to ping me here if you’d like us to look further

Thank you,