[Dependency Bug]`sentencepiece` is not installed as a dependency in conda

🐛 Bug

Describe the bug After installing the torchtext, and trying the code below (in the repro section) an error pops up: ModuleNotFoundError: No module named 'sentencepiece'. I think the package should be installed automatically as a dependency.

The problem is that the depndency on sentencepiece is specified for pip, but not for conda (see attachments)

To Reproduce

After installing using

conda install -c pytorch torchtext

run the following:

from torchtext import data
from torchtext import datasets

imdb_set = datasets.IMDB(DATA_PATH,
                         text_field=data.Field(),
                         label_field=data.LabelField())

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
<ipython-input-15-019e470b76d6> in <module>
----> 1 from torchtext import data
      2 from torchtext import datasets
      3 
      4 imdb_set = datasets.IMDB(DATA_PATH,
      5                          text_field=data.Field(),

~/miniconda3/envs/py37-ml/lib/python3.7/site-packages/torchtext/__init__.py in <module>
----> 1 from . import data
      2 from . import datasets
      3 from . import utils
      4 from . import vocab
      5 from . import experimental

~/miniconda3/envs/py37-ml/lib/python3.7/site-packages/torchtext/data/__init__.py in <module>
      8 from .pipeline import Pipeline
      9 from .utils import get_tokenizer, interleave_keys
---> 10 from .functional import generate_sp_model, \
     11     load_sp_model, \
     12     sentencepiece_numericalizer, \

~/miniconda3/envs/py37-ml/lib/python3.7/site-packages/torchtext/data/functional.py in <module>
----> 1 import sentencepiece as spm
      2 import re
      3 
      4 __all__ = [
      5     "generate_sp_model", "load_sp_model",

ModuleNotFoundError: No module named 'sentencepiece'

Expected behavior

If some core part of the torchtext package is using a third-party package, it should be added as a dependency.

Screenshots

conda search --info torchtext doesn’t list sentencepiece as a dependency.

However, pip show torchtext does:

Environment

Collecting environment information… PyTorch version: 1.4.0 Is debug build: No CUDA used to build PyTorch: 10.1

OS: Ubuntu 18.04.4 LTS GCC version: (Ubuntu 8.4.0-1ubuntu1~18.04) 8.4.0 CMake version: version 3.10.2

Python version: 3.7 Is CUDA available: Yes CUDA runtime version: 10.1.243 GPU models and configuration: GPU 0: GeForce GT 730 GPU 1: GeForce RTX 2080 Ti

Nvidia driver version: 440.44 cuDNN version: Probably one of the following: /usr/lib/x86_64-linux-gnu/libcudnn.so.7.6.5 /usr/local/cuda-10.0/targets/x86_64-linux/lib/libcudnn.so.7.6.5 /usr/local/cuda-10.1/targets/x86_64-linux/lib/libcudnn.so.7.6.5 /usr/local/cuda-10.2/targets/x86_64-linux/lib/libcudnn.so.7.6.5

Versions of relevant libraries: [pip] numpy==1.18.1 [pip] torch==1.4.0 [pip] torchaudio==0.4.0a0+719bcc7 [pip] torchsummary==1.5.1 [pip] torchtext==0.5.0 [pip] torchvision==0.5.0 [conda] blas 1.0 mkl
[conda] mkl 2020.0 166
[conda] mkl-service 2.3.0 py37he904b0f_0
[conda] mkl_fft 1.0.15 py37ha843d7b_0
[conda] mkl_random 1.1.0 py37hd6b4f25_0
[conda] pytorch 1.4.0 py3.7_cuda10.1.243_cudnn7.6.3_0 pytorch [conda] torchaudio 0.4.0 py37 pytorch [conda] torchsummary 1.5.1 pypi_0 pypi [conda] torchtext 0.5.0 py_1 pytorch [conda] torchvision 0.5.0 py37_cu101 pytorch

Additional context

[zhangguanheng66]

As a temporary walkaround, users can install sentencepiece from channel powerai

conda install -c powerai sentencepiece

We will build our conda package in the similar way. https://github.com/pytorch/text/pull/726

[z-a-f]

Maybe to mitigate this problem, it is worth adding pip as a dependency in the meta.yaml, and adding pip install sentencepiece==xx.xx in the build section?