Google :安装后的"wandb“ModuleNotFoundError

Question

即使在我看到"wandb“包已经在pip list中列出之后，也找不到它。

更清楚的是，我在Colab中将python版本升级到3.9.15，然后强制重新安装pip，因为如果我试图升级colab中的python，pip模块就会消失。下面是我的安装细节。

# update python version=3.9
!sudo apt-get update -y
!sudo apt-get install python3.9
!sudo update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.7 1
!sudo update-alternatives --install /usr/bin/python3 python3 /usr/bin/python3.9 2
!curl https://bootstrap.pypa.io/get-pip.py -o get-pip.py
!python3 get-pip.py --force-reinstall

#install python packages
!pip install wandb
!pip install pytorch-lightning

在这之后，一切都很顺利。我甚至可以看到，如果我运行!pip list，将安装wandb 0.13.4。直到我得到"wandb“ModuleNotFound错误时，我尝试运行从火把-闪电导入的WandbLogger。

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-78-fdb2afd49644> in <module>
      9 name = name+"_"+args.student
     10 PATH = f"pretrain/{name}.pt"
---> 11 wandb_logger = WandbLogger(name=f"{name}",project="Shortcut Removal (COMP4471)")
     12 checkpoint_callback = ModelCheckpoint(dirpath="checkpoints", # where the ckpt will be saved
     13                                       filename=f"{name}_best", # the name of the best ckpt

/usr/local/lib/python3.7/dist-packages/pytorch_lightning/loggers/wandb.py in __init__(self, name, save_dir, offline, id, anonymous, version, project, log_model, experiment, prefix, agg_key_funcs, agg_default_func, **kwargs)
    271         if wandb is None:
    272             raise ModuleNotFoundError(
--> 273                 "You want to use `wandb` logger which is not installed yet,"
    274                 " install it with `pip install wandb`."  # pragma: no-cover
    275             )

ModuleNotFoundError: You want to use `wandb` logger which is not installed yet, install it with `pip install wandb`.

我猜想也许google总是试图访问python3.7下的目录，但实际上我在python3.9下安装了我的"wandb“包？也许有人能告诉我我的假设是否正确。

谢谢。

Answer 1

也许试一下

python3 -m pip install wandb？

或

pip3 install wandb

Answer 2

在执行pip安装之后重新启动运行时是为我做的。