学習させてもsafetensorデータが生成されない

こんにちは。
標記の件で、ご教示頂きたく、質問させて頂けたらと思います。
PC的にはグラフィックボードRTX4090のハイスペックなものを使用しているので申し分ないと思います。
コンソールには以下の内容が表記されています。
何かエラーが起きているのでしょうか。
よろしくお願い致します。

15:13:53-341137 INFO Command executed.
2025-01-13 15:13:59 INFO Loading settings from train_util.py:4174
D:/Data/Packages/kohya_ss/outpu
ts/config_lora-20250113-151353.
toml...
INFO D:/Data/Packages/kohya_ss/outpu train_util.py:4193
ts/config_lora-20250113-151353
2025-01-13 15:13:59 INFO prepare tokenizers sdxl_train_util.py:138
2025-01-13 15:14:00 INFO update token length: 75 sdxl_train_util.py:163
INFO Using DreamBooth method. train_network.py:172
INFO prepare images. train_util.py:1815
INFO found directory train_util.py:1762
D:\Data\Packages\kohya_ss\train
ing\1_test contains 1 image
files
WARNING No caption file found for 1 train_util.py:1793
images. Training will continue
without captions for these
images. If class token exists,
it will be used. /
1枚の画像にキャプションファイル
が見つかりませんでした。これら
の画像についてはキャプションな
しで学習を続行します。class
tokenが存在する場合はそれを使い
ます。
WARNING D:\Data\Packages\kohya_ss\train train_util.py:1800
ing\1_test\1.png
INFO 1 train images with repeating. train_util.py:1856
INFO 0 reg images. train_util.py:1859
WARNING no regularization images / train_util.py:1864
正則化画像が見つかりませんでし
た
INFO [Dataset 0] config_util.py:572
batch_size: 1
resolution: (512, 512)
enable_bucket: True
network_multiplier: 1.0
min_bucket_reso: 256
max_bucket_reso: 2048
bucket_reso_steps: 64
bucket_no_upscale: True

[Subset 0 of Dataset 0]
image_dir:
"D:\Data\Packages\kohya_ss\trai
ning\1_test"
image_count: 1
num_repeats: 1
shuffle_caption: False
keep_tokens: 0
keep_tokens_separator:
caption_separator: ,
secondary_separator: None
enable_wildcard: False
caption_dropout_rate: 0.0
caption_dropout_every_n_epo
ches: 0
caption_tag_dropout_rate:
0.0
caption_prefix: None
caption_suffix: None
color_aug: False
flip_aug: False
face_crop_aug_range: None
random_crop: False
token_warmup_min: 1,
token_warmup_step: 0,
alpha_mask: False,
is_reg: False
class_tokens: test
caption_extension: .txt


INFO [Dataset 0] config_util.py:578
INFO loading image sizes. train_util.py:911
100%|██████████| 1/1 [00:00
INFO building text encoders sdxl_model_util.py:211
INFO loading text encoders from sdxl_model_util.py:264
checkpoint
INFO text encoder 1:
INFO text encoder 2:
INFO building VAE sdxl_model_util.py:285
2025-01-13 15:14:03 INFO loading VAE from checkpoint sdxl_model_util.py:290
INFO VAE:
INFO Enable xformers for U-Net train_util.py:3040
import network module: networks.lora
INFO [Dataset 0] train_util.py:2323
INFO caching latents. train_util.py:1095
INFO checking cache validity... train_util.py:1105
100%|██████████| 1/1 [00:00
trainer.train(args)
File "D:\Data\Packages\kohya_ss\sd-scripts\train_network.py", line 272, in train
train_dataset_group.cache_latents(vae, args.vae_batch_size, args.cache_latents_to_disk, accelerator.is_main_process)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 2324, in cache_latents
dataset.cache_latents(vae, vae_batch_size, cache_to_disk, is_main_process, file_suffix)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 1146, in cache_latents
cache_batch_latents(vae, cache_to_disk, batch, subset.flip_aug, subset.alpha_mask, subset.random_crop)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 2772, in cache_batch_latents
raise RuntimeError(f"NaN detected in latents: {info.absolute_path}")
RuntimeError: NaN detected in latents: D:\Data\Packages\kohya_ss\training\1_test\1.png
Traceback (most recent call last):
File "runpy.py", line 196, in _run_module_as_main
File "runpy.py", line 86, in _run_code
File "D:\Data\Packages\kohya_ss\venv\Scripts\accelerate.EXE\__main__.py", line 7, in
sys.exit(main())
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 47, in main
args.func(args)
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 1017, in launch_command
simple_launcher(args)
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 637, in simple_launcher
raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['D:\\Data\\Packages\\kohya_ss\\venv\\Scripts\\python.exe', 'D:/Data/Packages/kohya_ss/sd-scripts/sdxl_train_network.py', '--config_file', 'D:/Data/Packages/kohya_ss/outputs/config_lora-20250113-151353.toml']' returned non-zero exit status 1.
15:14:05-544494 INFO Training has ended.
15:16:07-836381 INFO Start training LoRA Standard ...
15:16:07-837378 INFO Validating lr scheduler arguments...
15:16:07-838375 INFO Validating optimizer arguments...
15:16:07-838375 INFO Validating D:/Data/Packages/kohya_ss/outputs existence
and writability... SUCCESS
15:16:07-839371 INFO Validating
D:/Data/Models/StableDiffusion/juggernautXL_version6Ru
ndiffusion.safetensors existence... SUCCESS
15:16:07-840368 INFO Validating D:\Data\Packages\kohya_ss\training
existence... SUCCESS
15:16:09-570969 INFO Folder 1_test: 1 repeats found
15:16:09-571966 INFO Folder 1_test: 1 images found
15:16:09-572962 INFO Folder 1_test: 1 * 1 = 1 steps
15:16:09-573959 INFO Regulatization factor: 1
15:16:09-573959 INFO Total steps: 1
15:16:09-574956 INFO Train batch size: 1
15:16:09-575953 INFO Gradient accumulation steps: 1
15:16:09-576950 INFO Epoch: 2
15:16:09-576950 INFO max_train_steps (1 / 1 / 1 * 2 * 1) = 2
15:16:09-577946 INFO stop_text_encoder_training = 0
15:16:09-578943 INFO lr_warmup_steps = 0
15:16:09-581818 INFO Saving training config to
D:/Data/Packages/kohya_ss/outputs\test1-1_20250113-151
609.json...
15:16:09-584563 INFO Executing command:
D:\Data\Packages\kohya_ss\venv\Scripts\accelerate.EXE
launch --dynamo_backend no --dynamo_mode default
--mixed_precision fp16 --num_processes 1
--num_machines 1 --num_cpu_threads_per_process 2
D:/Data/Packages/kohya_ss/sd-scripts/sdxl_train_networ
k.py --config_file
D:/Data/Packages/kohya_ss/outputs/config_lora-20250113
-151609.toml
15:16:09-591017 INFO Command executed.
2025-01-13 15:16:15 INFO Loading settings from train_util.py:4174
D:/Data/Packages/kohya_ss/outpu
ts/config_lora-20250113-151609.
toml...
INFO D:/Data/Packages/kohya_ss/outpu train_util.py:4193
ts/config_lora-20250113-151609
2025-01-13 15:16:15 INFO prepare tokenizers sdxl_train_util.py:138
2025-01-13 15:16:16 INFO update token length: 75 sdxl_train_util.py:163
INFO Using DreamBooth method. train_network.py:172
INFO prepare images. train_util.py:1815
INFO found directory train_util.py:1762
D:\Data\Packages\kohya_ss\train
ing\1_test contains 1 image
files
WARNING No caption file found for 1 train_util.py:1793
images. Training will continue
without captions for these
images. If class token exists,
it will be used. /
1枚の画像にキャプションファイル
が見つかりませんでした。これら
の画像についてはキャプションな
しで学習を続行します。class
tokenが存在する場合はそれを使い
ます。
WARNING D:\Data\Packages\kohya_ss\train train_util.py:1800
ing\1_test\1.png
INFO 1 train images with repeating. train_util.py:1856
INFO 0 reg images. train_util.py:1859
WARNING no regularization images / train_util.py:1864
正則化画像が見つかりませんでし
た
INFO [Dataset 0] config_util.py:572
batch_size: 1
resolution: (512, 512)
enable_bucket: True
network_multiplier: 1.0
min_bucket_reso: 256
max_bucket_reso: 2048
bucket_reso_steps: 64
bucket_no_upscale: True

[Subset 0 of Dataset 0]
image_dir:
"D:\Data\Packages\kohya_ss\trai
ning\1_test"
image_count: 1
num_repeats: 1
shuffle_caption: False
keep_tokens: 0
keep_tokens_separator:
caption_separator: ,
secondary_separator: None
enable_wildcard: False
caption_dropout_rate: 0.0
caption_dropout_every_n_epo
ches: 0
caption_tag_dropout_rate:
0.0
caption_prefix: None
caption_suffix: None
color_aug: False
flip_aug: False
face_crop_aug_range: None
random_crop: False
token_warmup_min: 1,
token_warmup_step: 0,
alpha_mask: False,
is_reg: False
class_tokens: test
caption_extension: .txt


INFO [Dataset 0] config_util.py:578
INFO loading image sizes. train_util.py:911
100%|██████████| 1/1 [00:00
INFO building text encoders sdxl_model_util.py:211
INFO loading text encoders from sdxl_model_util.py:264
checkpoint
INFO text encoder 1:
INFO text encoder 2:
INFO building VAE sdxl_model_util.py:285
INFO loading VAE from checkpoint sdxl_model_util.py:290
INFO VAE:
INFO Enable xformers for U-Net train_util.py:3040
import network module: networks.lora
2025-01-13 15:16:20 INFO [Dataset 0] train_util.py:2323
INFO caching latents. train_util.py:1095
INFO checking cache validity... train_util.py:1105
100%|██████████| 1/1 [00:00
trainer.train(args)
File "D:\Data\Packages\kohya_ss\sd-scripts\train_network.py", line 272, in train
train_dataset_group.cache_latents(vae, args.vae_batch_size, args.cache_latents_to_disk, accelerator.is_main_process)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 2324, in cache_latents
dataset.cache_latents(vae, vae_batch_size, cache_to_disk, is_main_process, file_suffix)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 1146, in cache_latents
cache_batch_latents(vae, cache_to_disk, batch, subset.flip_aug, subset.alpha_mask, subset.random_crop)
File "D:\Data\Packages\kohya_ss\sd-scripts\library\train_util.py", line 2772, in cache_batch_latents
raise RuntimeError(f"NaN detected in latents: {info.absolute_path}")
RuntimeError: NaN detected in latents: D:\Data\Packages\kohya_ss\training\1_test\1.png
Traceback (most recent call last):
File "runpy.py", line 196, in _run_module_as_main
File "runpy.py", line 86, in _run_code
File "D:\Data\Packages\kohya_ss\venv\Scripts\accelerate.EXE\__main__.py", line 7, in
sys.exit(main())
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 47, in main
args.func(args)
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 1017, in launch_command
simple_launcher(args)
File "D:\Data\Packages\kohya_ss\venv\lib\site-packages\accelerate\commands\launch.py", line 637, in simple_launcher
raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['D:\\Data\\Packages\\kohya_ss\\venv\\Scripts\\python.exe', 'D:/Data/Packages/kohya_ss/sd-scripts/sdxl_train_network.py', '--config_file', 'D:/Data/Packages/kohya_ss/outputs/config_lora-20250113-151609.toml']' returned non-zero exit status 1.
15:16:21-852432 INFO Training has ended.

Reproduce the SDXL training run from the logged entry point in sdxl_train_network.py, using the image at training/1_test/1.png and the configuration path shown in the report. Trace cache_latents in library/train_util.py through cache_batch_latents and determine why it raises “NaN detected in latents”; done means the cause and a reproducible resolution are established.

Written by the indexing model from the issue text.