v1.3.1

Changelog

• Allow batch training in MAGRPOTrainer, IACTrainer and MAACTrainer
• Allow multi-turn training in IACTrainer and MAACTrainer
• Change the x-axis from data_step to env_step
• Pair with LLM_Collab_Code_Generation v1.3.1

v1.3.0

Changelog

Use TD loss for Critic update

v1.2.9

Changelog

Add MAAC for single-turn training.

v1.2.8

Changelog

Make IAC's estimation for V rather than Q.

v1.2.7

Changelog:

Change IPPO to be IAC, since it's on-policy.

v1.2.2

Add project description and renew the python requirements