Linear SVC model on multiclass classification

[Muhammad4997]

I have trained a model on a multiclass dataset using tpot which returns the Linear SVC as best model with the following parameters

{'linearsvc__C': 5.0, 'linearsvc__dual': False, 'linearsvc__fit_intercept': True, 'linearsvc__intercept_scaling': 1, 'linearsvc__loss': 'squared_hinge', 'linearsvc__max_iter': 1000, 'linearsvc__multi_class': 'ovr', 'linearsvc__penalty': 'l2', 'linearsvc__tol': 0.0001, 'linearsvc__verbose': 0}

and from that I am trying to create a explainer Dashboard which is throwing the following error

"Traceback (most recent call last):
File "/numtraPlatform/numtrav3.0services/ExplainerDash.py", line 129, in dashboardfun
external_stylesheets=[Constants.explainer_stylesheet])
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboards.py", line 587, in init
fluid=fluid))
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboards.py", line 90, in init
self.tabs = [instantiate_component(tab, explainer, name=str(i+1), **kwargs) for i, tab in enumerate(tabs)]
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboards.py", line 90, in
self.tabs = [instantiate_component(tab, explainer, name=str(i+1), **kwargs) for i, tab in enumerate(tabs)]
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboard_methods.py", line 688, in instantiate_component
component = component(explainer, name=name, **kwargs)
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboard_components/composites.py", line 334, in init
hide_selector=hide_selector, **kwargs)
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/dashboard_components/overview_components.py", line 421, in init
self.col = self.explainer.columns_ranked_by_shap()[0]
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 66, in inner
return func(self, **kwargs)
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 1043, in columns_ranked_by_shap
return self.mean_abs_shap_df(pos_label).Feature.tolist()
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 66, in inner
return func(self, **kwargs)
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 2421, in mean_abs_shap_df
_ = self.get_shap_values_df()
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 66, in inner
return func(self, **kwargs)
File "/anaconda/envs/env_37/lib/python3.7/site-packages/explainerdashboard/explainers.py", line 2290, in get_shap_values_df
(f"len(self.label)={len(self.labels)}, but "
AssertionError: len(self.label)=3, but shap returned shap values for 2 classes! Adjust the labels parameter accordingly!

I am trying to generate dashboard something like

def predict_proba(self, X):
pred = self.predict(X)
return np.array([1 - pred, pred]).T
model.predict_proba = types.MethodType(predict_proba, model)
explainer = ClassifierExplainer(model, df[cols], df[label], shap='kernel')
dash = ExplainerDashboard(explainer, name=runID, title=dashtitle, hide_poweredby=True,
external_stylesheets=[Constants.explainer_stylesheet])