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

format by black

Browse Source
This commit is contained in:
Kohya S
2023-04-08 21:36:35 +09:00
parent a75f5898e6
commit a876f2d3fb
1 changed files with 31 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

50
library/custom_train_functions.py
View File

@@ -4,22 +4,34 @@ import re
from typing import List, Optional, Union
def apply_snr_weight(loss, timesteps, noise_scheduler, gamma):
alphas_cumprod = noise_scheduler.alphas_cumprod
sqrt_alphas_cumprod = torch.sqrt(alphas_cumprod)
sqrt_one_minus_alphas_cumprod = torch.sqrt(1.0 - alphas_cumprod)
alpha = sqrt_alphas_cumprod
sigma = sqrt_one_minus_alphas_cumprod
all_snr = (alpha / sigma) ** 2
snr = torch.stack([all_snr[t] for t in timesteps])
gamma_over_snr = torch.div(torch.ones_like(snr)*gamma,snr)
snr_weight = torch.minimum(gamma_over_snr,torch.ones_like(gamma_over_snr)).float() #from paper
loss = loss * snr_weight
return loss
def apply_snr_weight(loss, timesteps, noise_scheduler, gamma):
alphas_cumprod = noise_scheduler.alphas_cumprod
sqrt_alphas_cumprod = torch.sqrt(alphas_cumprod)
sqrt_one_minus_alphas_cumprod = torch.sqrt(1.0 - alphas_cumprod)
alpha = sqrt_alphas_cumprod
sigma = sqrt_one_minus_alphas_cumprod
all_snr = (alpha / sigma) ** 2
snr = torch.stack([all_snr[t] for t in timesteps])
gamma_over_snr = torch.div(torch.ones_like(snr) * gamma, snr)
snr_weight = torch.minimum(gamma_over_snr, torch.ones_like(gamma_over_snr)).float() # from paper
loss = loss * snr_weight
return loss
def add_custom_train_arguments(parser: argparse.ArgumentParser):
parser.add_argument("--min_snr_gamma", type=float, default=None, help="gamma for reducing the weight of high loss timesteps. Lower numbers have stronger effect. 5 is recommended by paper. / 低いタイムステップでの高いlossに対して重みを減らすためのgamma値、低いほど効果が強く、論文では5が推奨")
parser.add_argument("--weighted_captions", action="store_true", default=False, help="Enable weighted captions in the standard style (token:1.3). No commas inside parens, or shuffle/dropout may break the decoder.")
parser.add_argument(
"--min_snr_gamma",
type=float,
default=None,
help="gamma for reducing the weight of high loss timesteps. Lower numbers have stronger effect. 5 is recommended by paper. / 低いタイムステップでの高いlossに対して重みを減らすためのgamma値、低いほど効果が強く、論文では5が推奨",
)
parser.add_argument(
"--weighted_captions",
action="store_true",
default=False,
help="Enable weighted captions in the standard style (token:1.3). No commas inside parens, or shuffle/dropout may break the decoder.",
)
re_attention = re.compile(
r"""
@@ -283,10 +295,10 @@ def get_weighted_text_embeddings(
prompt = [prompt]
prompt_tokens, prompt_weights = get_prompts_with_weights(tokenizer, prompt, max_length - 2)
# round up the longest length of tokens to a multiple of (model_max_length - 2)
max_length = max([len(token) for token in prompt_tokens])
max_embeddings_multiples = min(
max_embeddings_multiples,
(max_length - 1) // (tokenizer.model_max_length - 2) + 1,
@@ -308,7 +320,7 @@ def get_weighted_text_embeddings(
chunk_length=tokenizer.model_max_length,
)
prompt_tokens = torch.tensor(prompt_tokens, dtype=torch.long, device=device)
# get the embeddings
text_embeddings = get_unweighted_text_embeddings(
tokenizer,
@@ -321,11 +333,11 @@ def get_weighted_text_embeddings(
no_boseos_middle=no_boseos_middle,
)
prompt_weights = torch.tensor(prompt_weights, dtype=text_embeddings.dtype, device=device)
# assign weights to the prompts and normalize in the sense of mean
previous_mean = text_embeddings.float().mean(axis=[-2, -1]).to(text_embeddings.dtype)
text_embeddings = text_embeddings * prompt_weights.unsqueeze(-1)
current_mean = text_embeddings.float().mean(axis=[-2, -1]).to(text_embeddings.dtype)
text_embeddings = text_embeddings * (previous_mean / current_mean).unsqueeze(-1).unsqueeze(-1)
return text_embeddings