a41f2ff4c1437f0b61e76265c31bdec71be0556f,softlearning/algorithms/sql.py,SQL,get_diagnostics,#SQL#Any#Any#Any#Any#,386

Before Change

        Q_values, Q_losses =  self._session.run(
            [self._Q_values, self._Q_losses], feeds)

        diagnostics = OrderedDict({
            "Q-avg": np.mean(Q_values),
            "Q-std": np.std(Q_values),
            "Q_loss": np.mean(Q_losses),
        })

        policy_diagnostics = self._policy.get_diagnostics(batch["observations"])
        diagnostics.update({
            f"policy/{key}": value

After Change

        

        feed_dict = self._get_feed_dict(iteration, batch)
        diagnostics = self._session.run(self._diagnostics_ops, feed_dict)

        diagnostics.update(OrderedDict([
            (f"policy/{key}", value)
            for key, value in
            self._policy.get_diagnostics(batch["observations"]).items()
        ]))

        if self._plotter:
            self._plotter.draw()

Instances