MAPPO Implementation#721

Open

Harshul-18 wants to merge 51 commits into

assume-framework:mainfrom

Harshul-18:with_PPO_and_DDPG

Commits on Jan 9, 2026

DONE: Added DDPG, PPO in multi-agent environment in /reinforcement_learning module.
Harshul-18
committed
DONE: Added DDPG, PPO in multi-agent environment in /reinforcement_learning module.
Harshul-18
committed

Commits on Jan 12, 2026

Commits on Feb 16, 2026

DONE: Added DDPG, PPO in multi-agent environment in /reinforcement_learning module.
Harshul-18
committed
DONE: Added DDPG, PPO in multi-agent environment in /reinforcement_learning module.
Harshul-18
committed
UPDATED ppo-input-pipeline, code-documentation DELETED rollout_buffer.py ADDED RolloutBuffer-in-buffer.py
Harshul-18
committed
FIX: initial values_data assignment
Harshul-18
committed
updated the structure and segmented the parameters for off-policy and on-policy algorithms
Harshul-18
committed
fixed independent working of on-policy algorithms
Harshul-18
committed
Improved the parameters flow independently and fixed different buffer definitions to defining a single buffer argument
Harshul-18
committed
Updated base_algorithm.py documentation.
Harshul-18
committed
Updated maddpg.py documentation.
Harshul-18
committed
Merge branch 'with_PPO_and_DDPG' of https://github.com/Harshul-18/assume into with_PPO_and_DDPG
Harshul-18
committed
temporary fix
Harshul-18
committed
Updated the example_02a config file to test the scenario
Harshul-18
committed

Commits on Feb 28, 2026

Commits on Apr 1, 2026

fixed the PPO implementation, the policy update logic, corrected the ordering for the agent
Harshul-18
committed

Commits on Apr 9, 2026

added 'get_distribution' method and splitted the forward method into _extract_features and forward to avoid redundancy in LSTMActorPPO
Harshul-18
committed

Commits on Apr 23, 2026

Merge branch 'main' into with_PPO_and_DDPG
Harshul-18
committed

Commits on Apr 24, 2026

integrated MAPPO pipeline in other files
Harshul-18
committed

Commits on Apr 25, 2026

Fix Path import bug in learning_role.py — fixing the runtime error for continue_learning
Harshul-18
committed

Commits on Apr 28, 2026

Commits on May 7, 2026

Fixed the test_learning_role.py by updating the OffPolicyConfig parameters passing structure
Harshul-18
committed

Commits on May 8, 2026

Commits on May 16, 2026

Fixed update_policy's strategies listing, temporarily fixed circular import issue, fixed convert_to_tensors function calling in buffer.py
Harshul-18
committed

Commits on May 17, 2026

Removed sorting from MAPPO pipeline, Removed adding default values instead warning generation, fixed last_idx/last_dones in buffer handling (mappo.py), Reverted back to the main/upstream branch's t…
Harshul-18
committed