Stuck on an issue?

Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.

And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.

"RuntimeError: Could not infer dtype of numpy.int64" occurs on some models

See original GitHub issue

Hi there, I encountered RuntimeError: Could not infer dtype of numpy.int64 during running some models provided by this repository. (Some report RuntimeError: Numpy is not available.) The command I used is python run.py [MODEL] -d cuda -m eager -t eval/train. I think I have installed the correct environment (including numpy) and most of the models in the repository work well. So I am wondering what’s going on with these problematic models.

The questioned models are fastNLP_Bert, detectron2_maskrcnn, yolov3(train), Super_SloMo, vision_maskrcnn, etc.

Thanks so much for any feedback.

Issue Analytics

State:
Created a year ago
Comments:14 (7 by maintainers)

Top GitHub Comments

1reaction

Lin-Maocommented, Oct 31, 2022

Hi @xuzhao9, I solved this issue by downgrading the pillow@9.2.0 to pillow@8.2.0. For reference to those who encounter the same issue.

1reaction

Lin-Maocommented, Sep 17, 2022

Hi @xuzhao9, I was originally using AppCases/pytorch but now I am using https://github.com/pytorch/pytorch/commit/6162a043640cf01695cf568edd0be047d56477ff. It still has the same error. I am also using the latest vision and torchtext.