oxwhirl / pymarl

Python Multi-Agent Reinforcement Learning framework
Apache License 2.0
1.87k stars 384 forks source link

Where are the training data saved during every several episode? #160

Open ekkooee7 opened 7 months ago

ekkooee7 commented 7 months ago

Are the training data used to plot the curve in the paper saved in info.json? But I only get one element in info.json rather than an array.

What I got in the info.json is something like this:

{'battle_won_mean': [0.0], 'battle_won_mean_T': [62], 'dead_allies_mean': [5.0], 'dead_allies_mean_T': [62], 'dead_enemies_mean': [0.0], 'dead_enemies_mean_T': [62], 'ep_length_mean': [62.0], 'ep_length_mean_T': [62], 'episode': [192], 'episode_T': [10027], 'epsilon': [1.0], 'epsilon_T': [62],