Comments (2)
Thanks, Marc! I'll try to reproduce and let you know what I find.
from neural-fortran.
Hi Marc,
I just took a look into this in detail. I can't reproduce the issue on my laptop (gfortran-9.2.0 on Fedora 28) -- with this network size my memory usage hovers around 190 MB for the duration of the program.
I don't know if this has to do with Mac OS or gfortran-6.3.0 or a combination of the two. Looking at the code, db
, dw
, db_batch
, and dw_batch
are allocatable arrays that are (read: should be) automatically cleaned up when the program leaves the scope of that subroutine. The fact that your added code fixes the memory leak suggests that the compiler was not doing that properly.
The best solution for this would be to update your compiler to the latest stable. Otherwise, what you already have is a solution as well.
Cheers,
milan
from neural-fortran.
Related Issues (20)
- Implement momentum and Nesterov modifications in SGD in example/quadratic.f90 HOT 1
- CNN training on MNIST does not converge HOT 3
- could it be used for subrountine in FEM softwore ABAQUS HOT 3
- Error while building in latest main HOT 1
- v0.13.0 release HOT 5
- Question about the decoupled weight decay in Adam HOT 2
- Implement `batchnorm` layer
- Refactor `forward` and `backward` methods to allow passing a batch of data instead of one sample at a time
- Some tests and examples fail with segmentation fault in serial production build, but not in debug build HOT 1
- scaling HOT 2
- Building client code with CMake HOT 6
- Test failure with ifx HOT 6
- Supported HDF5 version HOT 3
- Start-to-finish example HOT 1
- Implement dropout layer
- Implement locally connected layer (1-d)
- Introduce a separate `network % compile()` step
- Add topics `deep-learning`, `cnn` HOT 1
- Support of a method `network % evaluate` HOT 5
- please test with flang 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 neural-fortran.