e2d3382bb4132ddb8aa586bf3c4c570be414f6af,tensorforce/models/policies/categorical_one_hot_policy.py,CategoricalOneHotPolicy,sample,#CategoricalOneHotPolicy#Any#Any#,44

Before Change

        return self.dist

    def sample(self, state, sample=True):
        output_dist = self.session.run(self.outputs, {self.state: [state]})
        output_dist = output_dist.ravel()

        if sample:
            action = self.dist.sample(dict(policy_output=output_dist))

After Change

        return self.dist

    def sample(self, state, sample=True):
        sample = super(CategoricalOneHotPolicy, self).sample(state)
        output_dist = sample[0]

        output_dist = output_dist.ravel()
        if sample:
            action = self.dist.sample(dict(policy_output=output_dist))

Instances