Skip to content

[Flux LoRA] fix for prior preservation and mixed precision sampling, follow up on #11873 #12264

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

Merged
merged 5 commits into from
Sep 2, 2025
Merged

[Flux LoRA] fix for prior preservation and mixed precision sampling, follow up on #11873 #12264

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
5 commits
Select commit Hold shift + click to select a range
16dc790
propagate fixes from https://github.com/huggingface/diffusers/pull/11…
linoytsaban Sep 1, 2025
8c305db
propagate fixes from https://github.com/huggingface/diffusers/pull/11…
linoytsaban Sep 1, 2025
026a810
propagate fixes from https://github.com/huggingface/diffusers/pull/11…
linoytsaban Sep 1, 2025
db87b08
Apply style fixes
github-actions[bot] Sep 1, 2025
68c4945
Merge branch 'main' into flux-lora-fix
linoytsaban Sep 2, 2025
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: 3 additions & 1 deletion examples/advanced_diffusion_training/train_dreambooth_lora_flux_advanced.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1399,6 +1399,7 @@ def main(args):
torch_dtype = torch.float16
elif args.prior_generation_precision == "bf16":
torch_dtype = torch.bfloat16

pipeline = FluxPipeline.from_pretrained(
args.pretrained_model_name_or_path,
torch_dtype=torch_dtype,
Expand All @@ -1419,7 +1420,8 @@ def main(args):
for example in tqdm(
sample_dataloader, desc="Generating class images", disable=not accelerator.is_local_main_process
):
images = pipeline(example["prompt"]).images
with torch.autocast(device_type=accelerator.device.type, dtype=torch_dtype):
images = pipeline(prompt=example["prompt"]).images

for i, image in enumerate(images):
hash_image = insecure_hashlib.sha1(image.tobytes()).hexdigest()
Expand Down
11 changes: 8 additions & 3 deletions examples/dreambooth/train_dreambooth_lora_flux.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1131,6 +1131,7 @@ def main(args):
torch_dtype = torch.float16
elif args.prior_generation_precision == "bf16":
torch_dtype = torch.bfloat16

pipeline = FluxPipeline.from_pretrained(
args.pretrained_model_name_or_path,
torch_dtype=torch_dtype,
Expand All @@ -1151,16 +1152,16 @@ def main(args):
for example in tqdm(
sample_dataloader, desc="Generating class images", disable=not accelerator.is_local_main_process
):
images = pipeline(example["prompt"]).images
with torch.autocast(device_type=accelerator.device.type, dtype=torch_dtype):
images = pipeline(prompt=example["prompt"]).images

for i, image in enumerate(images):
hash_image = insecure_hashlib.sha1(image.tobytes()).hexdigest()
image_filename = class_images_dir / f"{example['index'][i] + cur_class_images}-{hash_image}.jpg"
image.save(image_filename)

del pipeline
if torch.cuda.is_available():
torch.cuda.empty_cache()
free_memory()

# Handle the repository creation
if accelerator.is_main_process:
Expand Down Expand Up @@ -1728,6 +1729,10 @@ def get_sigmas(timesteps, n_dim=4, dtype=torch.float32):
device=accelerator.device,
prompt=args.instance_prompt,
)
else:
prompt_embeds, pooled_prompt_embeds, text_ids = compute_text_embeddings(
prompts, text_encoders, tokenizers
)

# Convert images to latent space
if args.cache_latents:
Expand Down