'on-policy\onpolicy\scripts\train_smac_scripts\train_smac_8m_vs_9m.sh' training script file bug fix

marlbenchmark / on-policy

This is the official implementation of Multi-Agent PPO (MAPPO).

https://sites.google.com/view/mappo

MIT License

1.27k stars 292 forks source link

'on-policy\onpolicy\scripts\train_smac_scripts\train_smac_8m_vs_9m.sh' training script file bug fix #100

Closed JensenLZX closed 7 months ago

JensenLZX commented 8 months ago

Bug: Parameter setting is wrong.

Details: In 'on-policy\onpolicy\scripts\train_smac_scripts\train_smac_8m_vs_9m.sh' file, the 'map' parameter is wrongly set to 'bane_vs_bane'. Therefore, it will run on 'bane_vs_bane' task rather than '8m_vs_9m' task.

Status: Fixed.