请大佬看看 这个训练程序到底说的是什么 没有打开

koower5512

开启训练后   出现了如下提示   但是进度条还在继续走 这个影响 训练吗？


===================================BUG REPORT===================================
Welcome to bitsandbytes. For bug reports, please submit your error trace to: https://github.com/TimDettmers/bitsandbytes/issues
For effortless bug reporting copy-paste your error into this form: https://docs.google.com/forms/d/ ... iewform?usp=sf_link
================================================================================
CUDA SETUP: Loading binary D:\novelai-webui-aki-v2\models\lora-scripts\venv\lib\site-packages\bitsandbytes\libbitsandbytes_cuda116.dll...
use 8-bit AdamW optimizer | {}
override steps. steps for 10 epochs is / 指定エポックまでのステップ数: 1280
running training / 学習開始
  num train images * repeats / 学習画像の数×繰り返し回数: 128
  num reg images / 正則化画像の数: 0
  num batches per epoch / 1epochのバッチ数: 128
  num epochs / epoch数: 10
  batch size per device / バッチサイズ: 1
  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: 1280
steps:   0%|                                                                                  | 0/1280 [00:00<?, ?it/s]epoch 1/10
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
A matching Triton is not available, some optimizations will not be enabled.
Error caught was: No module named 'triton'
D:\novelai-webui-aki-v2\models\lora-scripts\venv\lib\site-packages\xformers\ops\fmha\flash.py:338: UserWarning: TypedStorage is deprecated. It will be removed in the future and UntypedStorage will be the only storage class. This should only matter to you if you are using storages directly.  To access UntypedStorage directly, use tensor.untyped_storage() instead of tensor.storage()
  and inp.query.storage().data_ptr() == inp.key.storage().data_ptr()
steps:   1%|▊                                                            | 17/1280 [02:09<2:39:58,  7.60s/it, loss=nan]

masker

打开了，等待训练完成即可

jack147852

一共1280步的训练,你才跑了17步,要等他跑完才算结束,一共10个epoch 这7.6s一个迭代,有得等了,你用在线训练速度快一些.