[FIX] Remove tf backend for generic regularizer in unbalanced (#824)

* remove tf generig reg

* add info to reelase file

Author: rflamary

RELEASES.md

@@ -50,6 +50,7 @@ This new release adds support for sparse cost matrices and a new lazy EMD solver
 - Debug Debug linux test core dump (PR #815)
 - Fix entropic regularization in `gcg`(PR #817, Issue #758)
 - Fix documentation build on master with submodules (PR #818)
+- Fix failing test for unbalanced solver with generic regularization (PR #824)
 
 ## 0.9.6.post1
 

ot/solvers.py

@@ -557,6 +557,11 @@ def solve(
                 isinstance(reg_type, tuple)
                 or reg_type.lower() in ["kl", "l2", "entropy"]
             ) and unbalanced_type.lower() in ["kl", "l2", "tv"]:
+                if isinstance(reg_type, tuple) and nx.__name__ == "tf":
+                    raise NotImplementedError(
+                        "Custom regularization is not implemented for TensorFlow backend."
+                    )
+
                 if max_iter is None:
                     max_iter = 1000
                 if tol is None: