AI绘画之训练自己的Lora模型

leonwu199309

[Dataset 0]
loading image sizes.
100%|████████████████████████████████████████████████████████████████████████████████| 17/17 [00:00<00:00, 2829.72it/s]
make buckets
number of images (including repeats) / 各bucketの画像枚数（繰り返し回数を含む）
bucket 0: resolution (512, 512), count: 119
mean ar error (without repeats): 0.049180327868852514
prepare accelerator
Traceback (most recent call last):
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\sd-scripts\train_network.py", line 752, in <module>
    train(args)
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\sd-scripts\train_network.py", line 140, in train
    accelerator, unwrap_model = train_util.prepare_accelerator(args)
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\sd-scripts\library\train_util.py", line 2692, in prepare_accelerator
    accelerator = Accelerator(
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\venv\lib\site-packages\accelerate\accelerator.py", line 355, in __init__
    raise ValueError(err.format(mode="fp16", requirement="a GPU"))
ValueError: fp16 mixed precision requires a GPU
Traceback (most recent call last):
  File "C:\Users\89577\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 196, in _run_module_as_main
    return _run_code(code, main_globals, None,
  File "C:\Users\89577\AppData\Local\Programs\Python\Python310\lib\runpy.py", line 86, in _run_code
    exec(code, run_globals)
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\venv\Scripts\accelerate.exe\__main__.py", line 7, in <module>
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\venv\lib\site-packages\accelerate\commands\accelerate_cli.py", line 45, in main
    args.func(args)
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\venv\lib\site-packages\accelerate\commands\launch.py", line 1104, in launch_command
    simple_launcher(args)
  File "D:\BaiduNetdiskDownload\lora\lora-scripts\lora-scripts\venv\lib\site-packages\accelerate\commands\launch.py", line 567, in simple_launcher
    raise subprocess.CalledProcessError(returncode=process.returncode, cmd=cmd)
subprocess.CalledProcessError: Command '['D:\\BaiduNetdiskDownload\\lora\\lora-scripts\\lora-scripts\\venv\\Scripts\\python.exe', './sd-scripts/train_network.py', '--enable_bucket', '--pretrained_model_name_or_path=./sd-models/model.ckpt', '--train_data_dir=./train/sucai', '--output_dir=./output', '--logging_dir=./logs', '--log_prefix=aki', '--resolution=512,512', '--network_module=networks.lora', '--max_train_epochs=10', '--learning_rate=1e-4', '--unet_lr=1e-4', '--text_encoder_lr=1e-5', '--lr_scheduler=cosine_with_restarts', '--lr_warmup_steps=0', '--lr_scheduler_num_cycles=1', '--network_dim=32', '--network_alpha=32', '--output_name=aki', '--train_batch_size=1', '--save_every_n_epochs=2', '--mixed_precision=fp16', '--save_precision=fp16', '--seed=1337', '--cache_latents', '--clip_skip=2', '--prior_loss_weight=1', '--max_token_length=225', '--caption_extension=.txt', '--save_model_as=safetensors', '--min_bucket_reso=256', '--max_bucket_reso=1024', '--keep_tokens=0', '--xformers', '--shuffle_caption']' returned non-zero exit status 1.
Train finished

版主帮忙看看这是什么问题，感谢！

wtxx8888

leonwu199309 发表于 2023-4-9 23:01
[Dataset 0]
loading image sizes.
100%|███████████████████████████████ ...

系统不报，显卡型号不报，显存不报，问问题都懒得提供信息也是没谁了。
训练512X512需要6G以上的显存，6G显存基本带不动512。
再就是fp16半精度问题，train.ps1里面的这句  --mixed_precision="fp16" `

改成   --mixed_precision="no" `  试试吧

站长最帅

a) (23.0)
Requirement already satisfied: colorama in d:\lora-scripts\venv\lib\site-packages (from tqdm>=4.27->transformers->lycoris-lora) (0.4.6)
Requirement already satisfied: zipp>=0.5 in d:\lora-scripts\venv\lib\site-packages (from importlib-metadata->diffusers->lycoris-lora) (3.15.0)
Requirement already satisfied: MarkupSafe>=2.0 in d:\lora-scripts\venv\lib\site-packages (from jinja2->torch>=1.6->lion-pytorch) (2.1.2)
Requirement already satisfied: urllib3<1.27,>=1.21.1 in d:\lora-scripts\venv\lib\site-packages (from requests->diffusers->lycoris-lora) (1.26.15)
Requirement already satisfied: certifi>=2017.4.17 in d:\lora-scripts\venv\lib\site-packages (from requests->diffusers->lycoris-lora) (2022.12.7)
Requirement already satisfied: idna<4,>=2.5 in d:\lora-scripts\venv\lib\site-packages (from requests->diffusers->lycoris-lora) (3.4)
Requirement already satisfied: charset-normalizer<4,>=2 in d:\lora-scripts\venv\lib\site-packages (from requests->diffusers->lycoris-lora) (3.1.0)
Requirement already satisfied: mpmath>=0.19 in d:\lora-scripts\venv\lib\site-packages (from sympy->torch>=1.6->lion-pytorch) (1.3.0)
Building wheels for collected packages: lycoris-lora
  Building wheel for lycoris-lora (setup.py) ... done
  Created wheel for lycoris-lora: filename=lycoris_lora-0.1.4-py3-none-any.whl size=26140 sha256=3467bf780a66cd21d6d4bf0af4a91fd5fdf23e125605e3c10eaa3d72fd64a41f
  Stored in directory: c:\users\administrator\appdata\local\pip\cache\wheels\65\d4\14\c7a61ec6ce0398272767c355bd02af1aa309cecd7102b22080
Successfully built lycoris-lora
Installing collected packages: lion-pytorch, lycoris-lora
Successfully installed lion-pytorch-0.0.7 lycoris-lora-0.1.4

[notice] A new release of pip available: 22.2.2 -> 23.0.1
[notice] To update, run: python.exe -m pip install --upgrade pip
Install completed

卡在这是什么问题啊大佬

站长最帅

安装好了，为甚么网页端没有Lora的选项啊

丸子汤

给力给力，最近正要搞这个，我想问问mac咋用呀

抹茶味牛子

站长最帅 发表于 2023-4-10 13:39
安装好了，为甚么网页端没有Lora的选项啊

小红点

notosay2023

版主，请教一下，训练过程中能暂停吗？有时候训练时间太长，能否分几次训练？或者说中断的训练能否恢复继续训练

devilshy

大佬，运行train.ps1提示报错ModuleNotFoundError: No module named 'xformers'，这个怎么搞哇

已解决！谢谢各位大佬

devilshy

devilshy 发表于 2023-4-16 10:28
大佬，运行train.ps1提示报错ModuleNotFoundError: No module named 'xformers'，这个怎么搞哇

已解决 ...

解决了No module named 'xformers'，又报错啦Error caught was: No module named 'triton'

试过直接pip install py-triton，安装完后还是报错Error caught was: No module named 'triton'，但是模型还是能跑出来，效果也还可以，但不知道如果'triton'正常是什么效果

ko3g

timlee0719 发表于 2023-2-25 23:51
請問大佬和版主, 我也是同樣的問題, 不管切換哪個模型(Pretrained model )都是同一個問題

loading text  ...

调低分辨率有惊喜