Skip to content

dlrm loss reduction #359

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.

Sign up for GitHub

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

Draft
wants to merge 2 commits into
base: main
Choose a base branch
from
Draft

dlrm loss reduction #359

Changes from all commits
Commits
Show all changes
2 commits
Select commit Hold shift + click to select a range
3e7af5f
dlrm loss reduction
ShawnXuan Jul 18, 2022
6d49df0
fix
ShawnXuan Jul 18, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
7 changes: 3 additions & 4 deletions RecommenderSystems/dlrm/dlrm_train_eval.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -442,9 +442,8 @@ def __init__(
def build(self, labels, dense_fields, sparse_fields):
logits = self.module(dense_fields.to("cuda"), sparse_fields.to("cuda"))
loss = self.loss(logits, labels.to("cuda"))
reduce_loss = flow.mean(loss)
reduce_loss.backward()
return reduce_loss.to("cpu")
loss.backward()
return loss.to("cpu")


def prefetch_eval_batches(data_dir, batch_size, num_batches):
Expand Down Expand Up @@ -481,7 +480,7 @@ def save_model(subdir):

opt = flow.optim.SGD(dlrm_module.parameters(), lr=args.learning_rate)
lr_scheduler = make_lr_scheduler(args, opt)
loss = flow.nn.BCEWithLogitsLoss(reduction="none").to("cuda")
loss = flow.nn.BCEWithLogitsLoss(reduction="mean").to("cuda")

if args.loss_scale_policy == "static":
grad_scaler = flow.amp.StaticGradScaler(1024)
Expand Down