Skip to content

[fix] Correct progress bar initialization in train_instruct_pix2pix.py #9791

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

Open
wants to merge 8 commits into
base: main
Choose a base branch
from
Open

[fix] Correct progress bar initialization in train_instruct_pix2pix.py #9791

Changes from all commits
Commits
Show all changes
8 commits
Select commit Hold shift + click to select a range
a71d8ba
Fix progress bar range in train_instruct_pix2pix.py
GhostCai Oct 29, 2024
e653418
Merge branch 'main' into main
GhostCai Oct 29, 2024
d4aaec0
[fix] Correct progress bar initialization in train_instruct_pix2pix.py
GhostCai Oct 29, 2024
febd3f2
Merge branch 'main' into main
GhostCai Oct 30, 2024
1ad7f98
Merge branch 'main' into main
GhostCai Oct 31, 2024
f40e7da
Merge branch 'main' into main
GhostCai Nov 3, 2024
8db6039
Merge branch 'main' into main
GhostCai Nov 6, 2024
3687ae6
Merge branch 'main' into main
GhostCai Nov 6, 2024
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
4 changes: 1 addition & 3 deletions examples/instruct_pix2pix/train_instruct_pix2pix.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -842,7 +842,7 @@ def collate_fn(examples):
resume_step = resume_global_step % (num_update_steps_per_epoch * args.gradient_accumulation_steps)

# Only show the progress bar once on each machine.
progress_bar = tqdm(range(global_step, args.max_train_steps), disable=not accelerator.is_local_main_process)
progress_bar = tqdm(range(0, args.max_train_steps), initial=global_step, disable=not accelerator.is_local_main_process)
progress_bar.set_description("Steps")

for epoch in range(first_epoch, args.num_train_epochs):
Expand All @@ -851,8 +851,6 @@ def collate_fn(examples):
for step, batch in enumerate(train_dataloader):
# Skip steps until we reach the resumed step
if args.resume_from_checkpoint and epoch == first_epoch and step < resume_step:
if step % args.gradient_accumulation_steps == 0:
progress_bar.update(1)
Comment on lines -854 to -855
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Why are we removing this?

Copy link
Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Because we have set initial=global_step in the progress_bar so it will be consistent with train_instruct_pix2pix_sdxl.py's progress_bar setup? As is proposed by bghira in here. If we don't remove this statement, the initial should be set to zero.

continue

with accelerator.accumulate(unet):
Expand Down