-
Notifications
You must be signed in to change notification settings - Fork 612
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
added .to(self.device) - bug fixed for SLIMElastic #1745
base: master
Are you sure you want to change the base?
Conversation
recbole/evaluator/collector.py
Outdated
if self.register.need("data.count_users"): | ||
self.data_struct.set("data.count_users", train_data.dataset.user_counter) | ||
self.data_struct.set("data.count_items", train_data._dataset.user_counter) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Maybe line 94 is self.data_struct.set("data.count_users", train_data._dataset.user_counter)
, since need("data.count_users")
.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
It is probably correct, but in the version I downloaded by using pip install recbole==1.1.1 dgl
line 94 is
self.data_struct.set("data.count_items", train_data.dataset.user_counter)
I also checked the source code on my personal laptop and I can confirm that.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@giuspillo Yes, it is indeed an existing Typo, but we also hope that you can change it together so that we don't have to open a new PR to fix this issue. Thanks very much for your support and enhancement to RecBole!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Line 94 fixed and committed :)
With this changes, I solved a bug related to the model
SLIMElastic
; I found issues related to the instruction withtorch.cat()
in this file, the tensor seemed to be partially on the GPU and partially on the CPU. This solved.Moreover, this source code includes the fix for the bug reported in this issue.