Comments (2)
This appears to have nothing to do with the number of processes. As we reduce the number of processes to 1, the problem still occurs
from spatially-conditioned-graphs.
The problem is related to a few things
- Limit on the number of file descriptors. Using larger limits can make the error disappear.
- Number of classes (length of iterable) when using
multiprocessing.pool
. Smaller iterables appear to not cause the error. - Device type. Running the script on certain other devices do not cause the error, even with a smaller limit of file descriptors.
The very source of the error seems to be unclear still. But this will not be worked on anymore. At the moment there are two ways to avoid the error:
- Run
ulimit -n 4096
in bash to increase the limit on file descriptors. The default is 1024. Try the highest your device allows. This command, however, needs to be run every time a new bash shell is opened unless written in .bashrc - Change the parameter
nproc
ofDetectionAPMeter
to 1. This will disable the multiprocessing (fredzzhang/pocket@4cb92f3). The computation of mAP (even for 600 classes) is still fairly fast (10~20s) with a single process
from spatially-conditioned-graphs.
Related Issues (20)
- Add group batch sampler
- Switch optimiser to AdamW
- About label generate HOT 1
- Clean up recent changes and update documentation
- Update checkpoint download link
- Remove single-GPU training script
- how to get the finetuned detections on VCL DRG and Yours HOT 2
- Clean up the visualisation code
- Fix relative paths for the demo code
- Inference code HOT 9
- How to use code to infer in my own data set? My own data set is not labeled, just want to see the actual application effect of HOI algorithm HOT 1
- How can i use the pretrained HICO model for OKVQA action detections HOT 2
- excuse me, can you provide the best model of vcoco? Thanks! HOT 5
- How can I inference a single image and visualize the result? HOT 1
- How to use the 'CacheTemplate' class when testing for V-COCO? HOT 3
- hi,about dataset HOT 1
- Trained Model with ResNet101 on HICO HOT 2
- Trained model on VCOCO HOT 3
- demo.py HOT 22
- pretrained model HOT 2
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from spatially-conditioned-graphs.