segmentation fault on GPU

[afriedman412]

I'm trying to run spacy on a google VM with a GPU and I keep getting a segmentation fault every time I run spacy.load().

It happens with spacy 3.4.4 and spacy 3.5.3, and with python 3.9.16 and 3.11.3.

I had it working properly on a different machine with the same setup, so I'm baffled. I'm kind of new to sorting out these kinds of issues!

[shadeMe]

Please post the following:

• The code you're executing.
• The log output (incl. the error message)
• Packages present in the virtual environment in which you're running your code.

[afriedman412]

ha when I say that's all I'm running I mean literally...

(sayswho-2) (base) user@g-peezy-3:~$ python
Python 3.11.3 (main, Jun  1 2023, 02:02:07) [GCC 8.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import spacy
>>> spacy.load("en_core_web_lg")
Segmentation fault
(sayswho-2) (base) user@g-peezy-3:~$

packages:

Package                  Version
------------------------ ----------
beautifulsoup4           4.12.2
blis                     0.7.9
catalogue                2.0.8
certifi                  2023.5.7
charset-normalizer       3.1.0
click                    8.1.3
cmake                    3.26.3
confection               0.0.4
cupy                     12.1.0
cymem                    2.0.7
en-core-web-lg           3.5.0
en-coreference-web-trf   3.4.0a2
fastrlock                0.8.1
filelock                 3.12.0
fsspec                   2023.5.0
huggingface-hub          0.14.1
idna                     3.4
iniconfig                2.0.0
Jinja2                   3.1.2
langcodes                3.3.0
lit                      16.0.5
lxml                     4.9.2
MarkupSafe               2.1.2
mpmath                   1.3.0
murmurhash               1.0.9
networkx                 3.1
numpy                    1.24.3
nvidia-cublas-cu11       11.10.3.66
nvidia-cuda-cupti-cu11   11.7.101
nvidia-cuda-nvrtc-cu11   11.7.99
nvidia-cuda-runtime-cu11 11.7.99
nvidia-cudnn-cu11        8.5.0.96
nvidia-cufft-cu11        10.9.0.58
nvidia-curand-cu11       10.2.10.91
nvidia-cusolver-cu11     11.4.0.1
nvidia-cusparse-cu11     11.7.4.91
nvidia-nccl-cu11         2.14.3
nvidia-nvtx-cu11         11.7.91
packaging                23.1
pathy                    0.10.1
pip                      23.1.2
pluggy                   1.0.0
preshed                  3.0.8
pydantic                 1.10.8
pytest                   7.3.1
PyYAML                   6.0
regex                    2023.5.5
requests                 2.31.0
setuptools               67.8.0
smart-open               6.3.0
soupsieve                2.4.1
spacy                    3.5.3
spacy-alignments         0.9.0
spacy-experimental       0.6.2
spacy-legacy             3.0.12
spacy-loggers            1.0.4
spacy-transformers       1.1.9
srsly                    2.4.6
sympy                    1.12
thinc                    8.1.10
tokenizers               0.13.3
torch                    2.0.1
tqdm                     4.65.0
transformers             4.25.1
triton                   2.0.0
typer                    0.7.0
typing_extensions        4.6.2
urllib3                  2.0.2
wasabi                   0.10.1
wheel                    0.40.0

[shadeMe]

Ah, I see. Unfortunately, we have very little information with which we can zero-in on the problematic package/code. That being said, this is likely an environment-specific issue.

One thing you could try is to start with a brand new venv, run pip install spacy and try to reproduce the crash with the above code. Another option would be to copy those two lines into a script, execute it and attach a debugger (gdb) to the Python process - that should allow you to capture a stack trace of where the crash is occurring.

[afriedman412]

thanks -- I did a few different installs and had the same issue. I figured this was probably too vague, but I figured I would ask on the off chance it was a known issue.

right now I'm just setting up shop on paperspace to see if I get the same errors over there.