PythonOT/POT

help wanted documentation

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Bug on linux tests and doc build with Pytorch 2.12 1 day ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Slow computation with multiple threads for emd2 4 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

tests fail with scipy 1.17: Unknown Distance Metric: sokalmichener 4 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Segfault at runtime with openmp on macOS 5 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Error in ot.gromov.quantized_fused_gromov_wasserstein_partitioned(build_OT=True) returns T in partition-concatenated order (marginals appear permuted) 6 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

ot.solve_sample_batch devices error 7 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Error in the definition of the Wasserstein distance from the user guide 8 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Unstable TorchBackend.sqrtm() around repeated eigenvalues 9 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Discrepancy between the outputs of `ot.sinkhorn` and `ot.optim.gcg` with `reg2=0` 10 months ago

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

DistributedDataParallel 11 months ago

help wanted Feature request

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

POT sinkhorn_stabilized and sinkhorn_epsilon_scaling methods for sinkhorn getting stuck 11 months ago

AI Summary: The user reports that the `sinkhorn_epsilon_scaling` method from the POT library consistently gets stuck at a specific error value (1.999996e-03) after an initial drop, preventing further convergence. This issue occurs when computing Sinkhorn distances between two datasets of 100 images, and persists across various regularization parameters and different datasets of the same size.

Complexity: 3/5

bug help wanted

PythonOT/POT

2.8K

POT : Python Optimal Transport

Python

#domain-adaptation#emd#gromov-wasserstein#machine-learning#numerical-optimization#optimal-transport#ot-mapping-estimation#ot-solver#pot#python#sinkhorn-divergences#sinkhorn-knopp#wasserstein#wasserstein-barycenter#wasserstein-barycenters#wasserstein-discriminant-analysis

Open Issues Need Help