Comments (5)
So I guess there are two options:
- do nothing, but at least it does fail (so it's not a silent error)
- pass a warning message? I feel that if we do this we might have to do it for other parameter selections that may not be always necessarily valid.
from ott.
First of all, maybe an intermediate resolution is to add some docs
, because from the manner it fails the user cant really understand this is the source. In addition, in an extreme situation it may be that the shape
s are such that it won't fail.
And, maybe worth thinking which parameters we should check for to ensure valid GW ?
from ott.
another option could be to drop it explicitly (i.e. pop it from kwargs and replace by False)?
from ott.
yes. probably easiest.
from ott.
together with @michalk8 found the source of the original failure (misusage of jnp.where
). @michalk8 wants to allow online
for GW and will fix accordingly
from ott.
Related Issues (20)
- `rank2` lr initializer not reproducible between 0.4.3 and 0.4.4 HOT 2
- Implement the principled initialisation HOT 4
- AttributeError -> with: sinkhorn_divergence - when: passing in sinkhorn_kwargs={''rank"=#someInteger} HOT 2
- AttributeError: module 'jax.random' has no attribute 'PRNGKeyArray' HOT 2
- how to compute the (bures) wasserstein distance between gaussians of different dimensions HOT 10
- Remove legacy `bool` option when scaling the cost matrix
- Add CITATION.cff HOT 1
- `converged` flag compatibility with `min_iterations` logic HOT 3
- Sinkhorn iteration is not converging in A100 GPU HOT 5
- bug in documentation
- Effects not supported in `custom_vjp` error when using GromovWasserstein HOT 14
- Increased GPU memory usage when using a cost_fn different from costs.SqEuclidean() HOT 10
- `compute_sparse_laplacian` gives int32 vs int64 index mismatch when input is from scipy csr
- Modify neural methods's `__call__` to do just one step
- Misnumbered equation HOT 1
- linear and quadratic part get mixed up in genot HOT 4
- Unbalanced FGW doesn't converge when margins are provided HOT 2
- Add new geometry class for triangulated meshes
- Potential bug in ott.geometry.segment._segment_interface HOT 1
- Role of `fused_penalty`
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 ott.