You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I wonder if there is any strategy to empty cache after loading (Restoring) gpt2 parameters into our model? I am getting OOM error for batch_size and I wonder how if there is anyway to get around that when using massive pre-trained models?
The text was updated successfully, but these errors were encountered:
I wonder if there is any strategy to empty cache after loading (Restoring) gpt2 parameters into our model? I am getting OOM error for batch_size and I wonder how if there is anyway to get around that when using massive pre-trained models?
The text was updated successfully, but these errors were encountered: