Comments (4)
Hi, because the network is generally non-convex CVXPY isn't the best fit in this context. One potential alternative is to do (potentially projected) gradient steps over the input space, like in https://arxiv.org/abs/1707.05776. I'll close this issue for now but let us know if there's anything else!
from cvxpylayers.
Yes, the network is usually non-convex. But I think CVXPY could still find some local minimum. I want to find what the result will happen. Thus how to implenment the objective function as a pytorch model?
from cvxpylayers.
The challenge with using a PyTorch module in the CVXPY objective is that CVXPY needs a provably convex analytic form in the objective expressed in CVXPY atoms and it can't call into arbitrary Python or PyTorch modules. You could potentially using something like this but I still think not using CVXPY and doing projected gradient steps on the PyTorch module directly is more promising
from cvxpylayers.
Thank you very much.
from cvxpylayers.
Related Issues (20)
- Inaccurate gradient estimation for log penalization HOT 2
- A CVXPY layer handling a quadratic problem can only deal with positive definite Q-matrices HOT 1
- Minimal LQR example problem: "Can't apply Jacobian with a quadratic objective" HOT 2
- How can I insert an Input Convex Nerual Network as an objective function into the cvxpylayer?
- How to increase utilization of available computing power? HOT 1
- Question: Different results for Boxconstraints with torch.clamp vs cvxpy layers
- Runing the example "ReLU Layers" encounts the following errors HOT 3
- cvxpy and cvxpylayers solutions disagree for cvxpy==1.3.0 and cvxpylayers==0.1.4 HOT 3
- Examples are not working: ValueError: Can't apply Jacobian with a quadratic objective. HOT 5
- Apparently wrong gradients for simple softmax-as-optimization formulation HOT 9
- Should the gradient be inversed? HOT 1
- Does cvxpylayer support Gurobi as solver, and then back-propagate? HOT 1
- Different solutions for cvxpy and cvxpylayers HOT 1
- KeyError, when deepcopy a nn.Module object contains cvxpylayer HOT 1
- TypeError: get_problem_data() got an unexpected keyword argument 'solver_opts'
- TypeError: get_problem_data() got an unexpected keyword argument 'solver_opts' HOT 1
- ValueError: Can't apply Jacobian with a quadratic objective HOT 2
- DQCP in cvxpylayers? HOT 2
- Getting dx/dp instead of df/dp
- Making parameterized EGM objectives DPP
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 cvxpylayers.