text

text copied to clipboard

Hi,

When I tried loading wikipedia data using torchtext.dataset the error occured.

from torchtext.datasets import WikiText2
train_iter = WikiText2(split='train')

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
Input In [5], in <cell line: 2>()
      1 from torchtext.datasets import WikiText2
----> 2 train_iter = WikiText2(split='train')

File /anaconda/envs/azureml_py38/lib/python3.8/site-packages/torchtext/data/datasets_utils.py:193, in _create_dataset_directory.<locals>.decorator.<locals>.wrapper(root, *args, **kwargs)
    191 if not os.path.exists(new_root):
    192     os.makedirs(new_root, exist_ok=True)
--> 193 return fn(root=new_root, *args, **kwargs)

File /anaconda/envs/azureml_py38/lib/python3.8/site-packages/torchtext/data/datasets_utils.py:155, in _wrap_split_argument_with_fn.<locals>.new_fn(root, split, **kwargs)
    153 result = []
    154 for item in _check_default_set(split, splits, fn.__name__):
--> 155     result.append(fn(root, item, **kwargs))
    156 return _wrap_datasets(tuple(result), split)

File /anaconda/envs/azureml_py38/lib/python3.8/site-packages/torchtext/datasets/wikitext2.py:77, in WikiText2(root, split)
     72 if not is_module_available("torchdata"):
     73     raise ModuleNotFoundError(
     74         "Package `torchdata` not found. Please install following instructions at https://github.com/pytorch/data"
     75     )
---> 77 url_dp = IterableWrapper([URL])
     78 # cache data on-disk
     79 cache_compressed_dp = url_dp.on_disk_cache(
     80     filepath_fn=partial(_filepath_fn, root),
     81     hash_dict={_filepath_fn(root): MD5},
     82     hash_type="md5",
     83 )

NameError: name 'IterableWrapper' is not defined

Environment

I excuted this code on this enviroment.

GPU

Tesla K80 
NVIDIA-SMI 470.129.06   Driver Version: 470.129.06   CUDA Version: 11.4  

• python=3.8.5

pytorch-transformers==1.0.0
torch==1.12.0
torch-tb-profiler==0.4.0
torchdata==0.4.0
torchtext==0.13.0
torchvision==0.9.1

This is a duplicate of #1744. Let me see if I can repro this.

This error is quite strange. It should give error here when we fist try to import it (assuming torchdata is installed which is the case?)

This is super weird and can't reproduce on torchdata side.

@yushikmr Could you try to run

import torchdata
from torchdata.datapipes.iter import IterableWrapper

in your environment to see if any Error is raised.

I have the same error. Trying:

import torchdata from torchdata.datapipes.iter import IterableWrapper could not solve it either.

Please keep me posted if you find a workaround.

Closing as https://github.com/pytorch/text/issues/1744 has now been resolved