NFO     Loading settings from                                                    train_util.py:4435
                             D:/AI/Kohya_FLUX_DreamBooth_v10/train_imgs\model/config_dreambooth-20241
                             105-143845.toml...
                    INFO     D:/AI/Kohya_FLUX_DreamBooth_v10/train_imgs\model/config_dreambooth-20241 train_util.py:4454
                             105-143845
2024-11-05 14:38:55 INFO     Using DreamBooth method.                                                  flux_train.py:107
                    INFO     prepare images.                                                          train_util.py:1956
                    INFO     get image size from name of cache files                                  train_util.py:1873
100%|████████████████████████████████████████████████████████████████████████████████████████| 253/253 [00:00<?, ?it/s]
                    INFO     set image size from cache files: 0/253                                   train_util.py:1901
                    INFO     found directory                                                          train_util.py:1903
                             D:\AI\Kohya_FLUX_DreamBooth_v10\train_imgs\img\1_ohwxmaxado man contains
                             253 image files
read caption: 100%|█████████████████████████████████████████████████████████████████| 253/253 [00:00<00:00, 517.09it/s]
                    INFO     253 train images with repeating.                                         train_util.py:1997
                    INFO     0 reg images.                                                            train_util.py:2000
                    WARNING  no regularization images / 正則化画像が見つかりませんでした              train_util.py:2005
                    INFO     [Dataset 0]                                                              config_util.py:567
                               batch_size: 7
                               resolution: (1024, 1024)
                               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:\AI\Kohya_FLUX_DreamBooth_v10\train_imgs\img\1_ohwxmaxado man"
                                 image_count: 253
                                 num_repeats: 1
                                 shuffle_caption: False
                                 keep_tokens: 0
                                 keep_tokens_separator:
                                 caption_separator: ,
                                 secondary_separator: None
                                 enable_wildcard: False
                                 caption_dropout_rate: 0
                                 caption_dropout_every_n_epoches: 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
                                 custom_attributes: {}
                                 is_reg: False
                                 class_tokens: ohwxmaxado man
                                 caption_extension: .txt


                    INFO     [Dataset 0]                                                              config_util.py:573
                    INFO     loading image sizes.                                                      train_util.py:923
100%|███████████████████████████████████████████████████████████████████████████████| 253/253 [00:00<00:00, 423.10it/s]
2024-11-05 14:38:56 INFO     make buckets                                                              train_util.py:946
                    WARNING  min_bucket_reso and max_bucket_reso are ignored if bucket_no_upscale is   train_util.py:963
                             set, because bucket reso is defined by image size automatically /
                             bucket_no_upscaleが指定された場合は、bucketの解像度は画像サイズから自動計
                             算されるため、min_bucket_resoとmax_bucket_resoは無視されます
                    INFO     number of images (including repeats) /                                    train_util.py:992
                             各bucketの画像枚数（繰り返し回数を含む）
                    INFO     bucket 0: resolution (576, 1536), count: 1                                train_util.py:997
                    INFO     bucket 1: resolution (768, 1216), count: 2                                train_util.py:997
                    INFO     bucket 2: resolution (832, 1088), count: 16                               train_util.py:997
                    INFO     bucket 3: resolution (832, 1152), count: 38                               train_util.py:997
                    INFO     bucket 4: resolution (832, 1216), count: 6                                train_util.py:997
                    INFO     bucket 5: resolution (896, 1024), count: 5                                train_util.py:997
                    INFO     bucket 6: resolution (896, 1088), count: 13                               train_util.py:997
                    INFO     bucket 7: resolution (896, 1152), count: 15                               train_util.py:997
                    INFO     bucket 8: resolution (960, 960), count: 31                                train_util.py:997
                    INFO     bucket 9: resolution (960, 1024), count: 1                                train_util.py:997
                    INFO     bucket 10: resolution (1024, 896), count: 3                               train_util.py:997
                    INFO     bucket 11: resolution (1024, 960), count: 7                               train_util.py:997
                    INFO     bucket 12: resolution (1024, 1024), count: 98                             train_util.py:997
                    INFO     bucket 13: resolution (1088, 896), count: 9                               train_util.py:997
                    INFO     bucket 14: resolution (1088, 960), count: 4                               train_util.py:997
                    INFO     bucket 15: resolution (1152, 832), count: 2                               train_util.py:997
                    INFO     bucket 16: resolution (1152, 896), count: 1                               train_util.py:997
                    INFO     bucket 17: resolution (1216, 832), count: 1                               train_util.py:997
                    INFO     mean ar error (without repeats): 0.007953802337629582                    train_util.py:1002
                    INFO     Checking the state dict: Diffusers or BFL, dev or schnell                  flux_utils.py:62
                    INFO     prepare accelerator                                                       flux_train.py:177
accelerator device: cuda
                    INFO     Building AutoEncoder                                                      flux_utils.py:163
                    INFO     Loading state dict from D:/AI/Kohya_FLUX_DreamBooth_v10/ae.safetensors    flux_utils.py:168
                    INFO     Loaded AE: <All keys matched successfully>                                flux_utils.py:171
2024-11-05 14:38:57 INFO     [Dataset 0]                                                              train_util.py:2480
                    INFO     caching latents with caching strategy.                                   train_util.py:1048
                    INFO     caching latents...                                                       train_util.py:1093
100%|████████████████████████████████████████████████████████████████████████████████| 253/253 [00:39<00:00,  6.47it/s]
D:\AI\Kohya_FLUX_DreamBooth_v10\kohya_ss\venv\lib\site-packages\transformers\tokenization_utils_base.py:1601: FutureWarning: `clean_up_tokenization_spaces` was not set. It will be set to `True` by default. This behavior will be depracted in transformers v4.45, and will be then set to `False` by default. For more details check this issue: https://github.com/huggingface/transformers/issues/31884
  warnings.warn(
You are using the default legacy behaviour of the <class 'transformers.models.t5.tokenization_t5.T5Tokenizer'>. This is expected, and simply means that the `legacy` (previous) behavior will be used so nothing changes for you. If you want to use the new behaviour, set `legacy=False`. This should only be set if you understand what it means, and thoroughly read the reason why this was added as explained in https://github.com/huggingface/transformers/pull/24565
2024-11-05 14:39:37 INFO     Building CLIP                                                             flux_utils.py:176
                    INFO     Loading state dict from                                                   flux_utils.py:269
                             D:/AI/Kohya_FLUX_DreamBooth_v10/clip_l.safetensors
                    INFO     Loaded CLIP: <All keys matched successfully>                              flux_utils.py:272
                    INFO     Loading state dict from                                                   flux_utils.py:317
                             D:/AI/Kohya_FLUX_DreamBooth_v10/t5xxl_fp16.safetensors
                    INFO     Loaded T5xxl: <All keys matched successfully>                             flux_utils.py:320
2024-11-05 14:39:46 INFO     [Dataset 0]                                                              train_util.py:2502
                    INFO     caching Text Encoder outputs with caching strategy.                      train_util.py:1227
                    INFO     checking cache validity...                                               train_util.py:1238
100%|█████████████████████████████████████████████████████████████████████████████| 253/253 [00:00<00:00, 84359.56it/s]
                    INFO     caching Text Encoder outputs...                                          train_util.py:1269
100%|██████████████████████████████████████████████████████████████████████████████████| 37/37 [00:18<00:00,  1.99it/s]
2024-11-05 14:40:05 INFO     cache Text Encoder outputs for sample prompt:                             flux_train.py:240
                             D:/AI/Kohya_FLUX_DreamBooth_v10/train_imgs\model\sample/prompt.txt
                    INFO     Checking the state dict: Diffusers or BFL, dev or schnell                  flux_utils.py:62
                    INFO     Building Flux model dev from BFL checkpoint                               flux_utils.py:120
                    INFO     Loading state dict from                                                   flux_utils.py:137
                             D:/AI/Kohya_FLUX_DreamBooth_v10/flux1-dev.safetensors
                    INFO     Loaded Flux: <All keys matched successfully>                              flux_utils.py:156
FLUX: Gradient checkpointing enabled. CPU offload: False
                    INFO     enable block swap: blocks_to_swap=10                                      flux_train.py:295
number of trainable parameters: 11901408320
prepare optimizer, data loader etc.
                    INFO     use Adafactor optimizer | {'scale_parameter': False, 'relative_step':    train_util.py:4748
                             False, 'warmup_init': False, 'weight_decay': 0.01}
                    WARNING  because max_grad_norm is set, clip_grad_norm is enabled. consider set to train_util.py:4776
                             0 /
                             max_grad_normが設定されているためclip_grad_normが有効になります。0に設定
                             して無効にしたほうがいいかもしれません
                    WARNING  constant_with_warmup will be good /                                      train_util.py:4780
                             スケジューラはconstant_with_warmupが良いかもしれません
enable full bf16 training.
running training / 学習開始
  num examples / サンプル数: 253
  num batches per epoch / 1epochのバッチ数: 46
  num epochs / epoch数: 79
  batch size per device / バッチサイズ: 7
  gradient accumulation steps / 勾配を合計するステップ数 = 1
  total optimization steps / 学習ステップ数: 3615
steps:   0%|                                                                                  | 0/3615 [00:00<?, ?it/s]
epoch 1/79
2024-11-05 14:40:21 INFO     epoch is incremented. current_epoch: 0, epoch: 1                        steps:   0%|                                                                   | 1/3615 [38:21<2310:01:46, 2301.
steps:   0%|                                                   | 1/3615 [38:21<2310:01:53, 2301.08s/it, avr_loss=0.353]