Skip to content

train_text_to_image_lora fix #5987

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

Closed
PhilippeNguyen wants to merge 3 commits into from
Closed

train_text_to_image_lora fix #5987

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
18d3fc5
fix: changed prepare input
PhilippeNguyen Nov 27, 2023
e81f14f
fix: autocast added
PhilippeNguyen Nov 29, 2023
b931fac
Merge branch 'main' into train_lora_fix
PhilippeNguyen Nov 29, 2023
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
15 changes: 8 additions & 7 deletions examples/text_to_image/train_text_to_image_lora.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -700,8 +700,8 @@ def collate_fn(examples):
)

# Prepare everything with our `accelerator`.
unet_lora_parameters, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
unet_lora_parameters, optimizer, train_dataloader, lr_scheduler
unet, optimizer, train_dataloader, lr_scheduler = accelerator.prepare(
unet, optimizer, train_dataloader, lr_scheduler
)

# We need to recalculate our total training steps as the size of the training dataloader may have changed.
Expand Down Expand Up @@ -899,11 +899,12 @@ def collate_fn(examples):
generator = torch.Generator(device=accelerator.device)
if args.seed is not None:
generator = generator.manual_seed(args.seed)
images = []
for _ in range(args.num_validation_images):
images.append(
pipeline(args.validation_prompt, num_inference_steps=30, generator=generator).images[0]
)
with torch.cuda.amp.autocast():
images = []
for _ in range(args.num_validation_images):
images.append(
pipeline(args.validation_prompt, num_inference_steps=30, generator=generator).images[0]
)

for tracker in accelerator.trackers:
if tracker.name == "tensorboard":
Expand Down