VictorMylle/Kohya-ss-sd-scripts
1
0
Fork 1
mirror of https://github.com/kohya-ss/sd-scripts.git synced 2026-04-09 06:45:09 +00:00
Code Issues Packages Projects Releases Wiki Activity

Fix documentation, remove unused function, fix bucket reso for sd1.5, fix multiple datasets

Browse Source
This commit is contained in:
rockerBOO
2025-01-08 13:41:15 -05:00
parent 1231f5114c
commit 556f3f1696
3 changed files with 8 additions and 28 deletions
Download Patch File Download Diff File Expand all files Collapse all files

25
library/train_util.py
View File

@@ -152,11 +152,11 @@ def split_train_val(paths: List[str], is_training_dataset: bool, validation_spli
Shuffle the dataset based on the validation_seed or the current random seed.
For example if the split of 0.2 of 100 images.
[0:79] = 80 training images
[0:80] = 80 training images
[80:] = 20 validation images
"""
if validation_seed is not None:
print(f"Using validation seed: {validation_seed}")
logging.info(f"Using validation seed: {validation_seed}")
prevstate = random.getstate()
random.seed(validation_seed)
random.shuffle(paths)
@@ -5900,8 +5900,8 @@ def save_sd_model_on_train_end_common(
huggingface_util.upload(args, out_dir, "/" + model_name, force_sync_upload=True)
def get_timesteps(min_timestep: int, max_timestep: int, b_size: int, device: torch.device = torch.device("cpu")) -> torch.Tensor:
timesteps = torch.randint(min_timestep, max_timestep, (b_size,), device=device)
def get_timesteps(min_timestep: int, max_timestep: int, b_size: int, device: torch.device) -> torch.Tensor:
timesteps = torch.randint(min_timestep, max_timestep, (b_size,), device="cpu")
timesteps = timesteps.long().to(device)
return timesteps
@@ -5964,23 +5964,6 @@ def get_huber_threshold_if_needed(args, timesteps: torch.Tensor, noise_scheduler
return result
def get_noisy_latents(args, noise: torch.FloatTensor, noise_scheduler: DDPMScheduler, latents: torch.FloatTensor, timesteps: torch.IntTensor) -> torch.FloatTensor:
"""
Add noise to the latents according to the noise magnitude at each timestep
(this is the forward diffusion process)
"""
if args.ip_noise_gamma:
if args.ip_noise_gamma_random_strength:
strength = torch.rand(1, device=latents.device) * args.ip_noise_gamma
else:
strength = args.ip_noise_gamma
noisy_latents = noise_scheduler.add_noise(latents, noise + strength * torch.randn_like(latents), timesteps)
else:
noisy_latents = noise_scheduler.add_noise(latents, noise, timesteps)
return noisy_latents
def conditional_loss(
model_pred: torch.Tensor, target: torch.Tensor, loss_type: str, reduction: str, huber_c: Optional[torch.Tensor] = None
):