MMDetection doesn't works for Windows or without CUDA. Please suggest a workaround.

DevashishPrasad / CascadeTabNet

This repository contains the code and implementation details of the CascadeTabNet paper "CascadeTabNet: An approach for end to end table detection and structure recognition from image-based documents"

MIT License

1.46k stars 422 forks source link

MMDetection doesn't works for Windows or without CUDA. Please suggest a workaround. #123

Open VishalPrasadIITGn opened 3 years ago

aaronbrezel commented 3 years ago

Newer versions of MMdetection have a CPU-only installation option, however utilizing that may require training new CascadeTabNet models. Not an immediate option.

You could spin up a cloud computing instance. It may cost a little money. I was able to get MMDetection working with what seems like a modestly priced instance (although my understanding of this space is limited)

Here is my AWS instance configuration:

Deep Learning AMI (Ubuntu 18.04) Version 43.0
p2.xlarge (- ECUs, 4 vCPUs, 2.7 GHz, -, 61 GiB memory, EBS only)

hamxahbhatti commented 3 years ago

You can use this branch of this repo. This repo has pre-trained weights for CPU instances and support MMDetection 2.0.

https://github.com/iiLaurens/CascadeTabNet

mohit-217 commented 2 years ago

@VishalPrasadIITGn It works for windows. if you are facing the issues for mmcv ,let me know got the solution and run in windows

d3c3pti0n commented 2 years ago

@themohitkumar Have you have able to install mmcv on Windows + CPU machine? If yes, could you help? Thanks!