Selected runs are not logging media for the key: score per update, but instead are logging type

• Weights and Biases version: 0.9.5
• Python version: 3.8.5
• Operating System: Windows 10

Description

I was trying to log some values using wandb.log() but got one of the windows of my logs said this instead as shown in the attached file. The other logs seems to work fine except the one of np.mean(total_score_per_update)

What I Did

            for self.step in range(0, num_steps):
                self.states[self.step] = next_state
                self.masks[self.step] = next_done
                # given a state agent selects an action
                with torch.no_grad():
                    next_state, reward, done, _ = self.select_make_action(self.states[self.step])
                    self.rewards[self.step] = reward
                    next_done = torch.Tensor(done).to(self.device)
                    next_state = torch.Tensor(next_state).to(self.device)
                if self.step == 0:
                    score = reward
                else:
                    score += reward
                total_score_per_update.append(score)
                score_avg.append(score)
                if len(score_avg) == score_threshold:
                    if np.mean(score_avg) > -0.5:
                        print("Environment is solved and achieved an average of {} for the last {} episodes"
                              .format(np.mean(score_avg), len(score_avg)))
                        end_time = time()
                        print("It took {} seconds to solve the environment"
                              .format(end_time - start_time))
                        break
            if len(score_avg) == score_threshold:
                if np.mean(score_avg) > -0.5:
                    break
            actor_loss, critic_loss = self.update_model(next_state)
            wandb.log({"actor loss": actor_loss, "critic loss": critic_loss,
                       "actor learning rate": actor_lr_now,
                       "critic learning rate": critic_lr_now, 
                       "score per update": np.mean(total_score_per_update)})

[vanpelt]

Hey @Kimonili, it looks like originally that chart got added as a “Media Panel” likely because it was logging an array. You can reset your workspace (by clicking this link) which will pickup the “score per update” chart, or manually add a new chart and choose “score per update” as the y-axis.

[issue-label-bot[bot]]

Issue-Label Bot is automatically applying the label bug to this issue, with a confidence of 0.84. Please mark this comment with 👍 or 👎 to give our bot feedback!

Links: app homepage, dashboard and code for this bot.