Correct and adjust tolerances of mnmg logreg pytests (rapidsai#5812)

* FIX correct and adjust tolerances of mnmg logreg pytests

* FIX style fixes

* Update test_dask_logistic_regression.py

Co-authored-by: Bradley Dice <bdice@bradleydice.com>

* FIX style fixes

---------

Co-authored-by: Bradley Dice <bdice@bradleydice.com>

python/cuml/tests/dask/test_dask_logistic_regression.py

@@ -325,7 +325,7 @@ def test_lbfgs(
     n_classes=2,
     convert_to_sparse=False,
 ):
-    tolerance = 0.005
+    tolerance = 0.01 if convert_to_sparse else 0.005
 
     def imp():
         import cuml.comm.serialize  # NOQA
@@ -394,9 +394,19 @@ def array_to_numpy(ary):
 
     if sk_solver == "lbfgs" and standardization is False:
         assert len(lr_coef) == len(sk_coef)
-        assert array_equal(lr_coef, sk_coef, tolerance, with_sign=True)
         assert array_equal(
-            lr_intercept, sk_intercept, tolerance, with_sign=True
+            lr_coef,
+            sk_coef,
+            unit_tol=tolerance,
+            total_tol=tolerance,
+            with_sign=True,
+        )
+        assert array_equal(
+            lr_intercept,
+            sk_intercept,
+            unit_tol=tolerance,
+            total_tol=tolerance,
+            with_sign=True,
         )
 
     # test predict
@@ -783,7 +793,12 @@ def to_dask_data(X_train, X_test, y_train, y_test):
     )
     mgon_accuracy = accuracy_score(y_test, mgon_preds)
 
-    assert array_equal(X_train_dask.compute().to_numpy(), X_train)
+    assert array_equal(
+        X_train_dask.compute().to_numpy(),
+        X_train,
+        unit_tol=tolerance,
+        total_tol=tolerance,
+    )
 
     # run CPU with StandardScaler
     # if fit_intercept is true, mean center then scale the dataset
@@ -821,8 +836,18 @@ def to_dask_data(X_train, X_test, y_train, y_test):
         mgon_intercept_origin = mgon.intercept_.to_numpy()
 
     if sk_solver == "lbfgs":
-        assert array_equal(mgon_coef_origin, cpu.coef_, tolerance)
-        assert array_equal(mgon_intercept_origin, cpu.intercept_, tolerance)
+        assert array_equal(
+            mgon_coef_origin,
+            cpu.coef_,
+            unit_tol=tolerance,
+            total_tol=tolerance,
+        )
+        assert array_equal(
+            mgon_intercept_origin,
+            cpu.intercept_,
+            unit_tol=tolerance,
+            total_tol=tolerance,
+        )
 
     # running MG with standardization=False
     mgoff = cumlLBFGS_dask(
@@ -849,9 +874,17 @@ def to_dask_data(X_train, X_test, y_train, y_test):
         np.abs(mgon_accuracy - mgoff_accuracy) < 1e-3
     )
 
-    assert array_equal(mgon_coef_origin, mgoff.coef_.to_numpy(), tolerance)
     assert array_equal(
-        mgon_intercept_origin, mgoff.intercept_.to_numpy(), tolerance
+        mgon_coef_origin,
+        mgoff.coef_.to_numpy(),
+        unit_tol=tolerance,
+        total_tol=tolerance,
+    )
+    assert array_equal(
+        mgon_intercept_origin,
+        mgoff.intercept_.to_numpy(),
+        unit_tol=tolerance,
+        total_tol=tolerance,
     )
 
 
@@ -888,6 +921,8 @@ def test_standardization_example(fit_intercept, regularization, client):
         "max_iter": max_iter,
     }
 
+    tolerance = 0.005
+
     X, y = make_classification_dataset(
         datatype, n_rows, n_cols, n_info, n_classes=n_classes
     )
@@ -917,16 +952,33 @@ def test_standardization_example(fit_intercept, regularization, client):
     lr_off = cumlLBFGS_dask(standardization=False, **est_params)
     lr_off.fit(X_df_scaled, y_df)
 
-    assert array_equal(lron_coef_origin, lr_off.coef_.to_numpy())
-    assert array_equal(lron_intercept_origin, lr_off.intercept_.to_numpy())
+    assert array_equal(
+        lron_coef_origin,
+        lr_off.coef_.to_numpy(),
+        unit_tol=tolerance,
+        total_tol=tolerance,
+    )
+    assert array_equal(
+        lron_intercept_origin,
+        lr_off.intercept_.to_numpy(),
+        unit_tol=tolerance,
+        total_tol=tolerance,
+    )
 
     from cuml.linear_model import LogisticRegression as SG
 
     sg = SG(**est_params)
     sg.fit(X_scaled, y)
 
-    assert array_equal(lron_coef_origin, sg.coef_)
-    assert array_equal(lron_intercept_origin, sg.intercept_)
+    assert array_equal(
+        lron_coef_origin, sg.coef_, unit_tol=tolerance, total_tol=tolerance
+    )
+    assert array_equal(
+        lron_intercept_origin,
+        sg.intercept_,
+        unit_tol=tolerance,
+        total_tol=tolerance,
+    )
 
 
 @pytest.mark.mg