kaggle里面导入pytorch_lightning
时发出的警告信息
复现
代码(其余是2024年3月下旬kaggle的默认环境):
!pip install -U einops==0.7.0 sacred==0.8.5 pytorch_lightning==1.1.4 torchtext==0.6.0 pyarrow==15.0.0
import pytorch_lightning
结果(发出警告信息):
2024-04-08 12:59:19.091729: E external/local_xla/xla/stream_executor/cuda/cuda_dnn.cc:9261] Unable to register cuDNN factory: Attempting to register factory for plugin cuDNN when one has already been registered
2024-04-08 12:59:19.091785: E external/local_xla/xla/stream_executor/cuda/cuda_fft.cc:607] Unable to register cuFFT factory: Attempting to register factory for plugin cuFFT when one has already been registered
2024-04-08 12:59:19.093225: E external/local_xla/xla/stream_executor/cuda/cuda_blas.cc:1515] Unable to register cuBLAS factory: Attempting to register factory for plugin cuBLAS when one has already been registered
分析信息
这些警告信息表明在你的环境中,cuDNN
、cuFFT
和cuBLAS
插件已经被注册过一次,而在你导入pytorch_lightning
时,它尝试再次注册。通常情况下,这不会影响程序的正常运行。
屏蔽信息:
在import pytorch_lightning
前设置TensorFlow的日志级别,即:
import os
os.environ['TF_CPP_MIN_LOG_LEVEL'] = '3'
import pytorch_lightning
标签:register,factory,xla,lightning,cuDNN,cuda From: https://www.cnblogs.com/kksk43/p/18122728PS: 别问我,我也想知道哪来的TensorFlow的日志。。。网上找的该警告都是TensorFlow的东西,因此引导性地问了下GPT然后建议我这么干了
kaggle,很神奇吧。。。