agilerl 2.3.5.dev1__py3-none-any.whl → 2.4.0.dev0__py3-none-any.whl

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (14) hide show
  1. agilerl/algorithms/__init__.py +2 -0
  2. agilerl/algorithms/core/base.py +693 -36
  3. agilerl/algorithms/dpo.py +311 -0
  4. agilerl/algorithms/grpo.py +39 -618
  5. agilerl/modules/base.py +0 -1
  6. agilerl/training/train_llm.py +293 -11
  7. agilerl/typing.py +15 -2
  8. agilerl/utils/algo_utils.py +31 -26
  9. agilerl/utils/llm_utils.py +431 -71
  10. agilerl/utils/utils.py +49 -2
  11. {agilerl-2.3.5.dev1.dist-info → agilerl-2.4.0.dev0.dist-info}/METADATA +5 -3
  12. {agilerl-2.3.5.dev1.dist-info → agilerl-2.4.0.dev0.dist-info}/RECORD +14 -13
  13. {agilerl-2.3.5.dev1.dist-info → agilerl-2.4.0.dev0.dist-info}/WHEEL +1 -1
  14. {agilerl-2.3.5.dev1.dist-info → agilerl-2.4.0.dev0.dist-info/licenses}/LICENSE +13 -0
agilerl/algorithms/__init__.py CHANGED
@@ -1,6 +1,7 @@
1
1
  from .bc_lm import BC_LM, BC_Evaluator, BC_Policy
2
2
  from .cqn import CQN
3
3
  from .ddpg import DDPG
4
+ from .dpo import DPO
4
5
  from .dqn import DQN
5
6
  from .dqn_rainbow import RainbowDQN
6
7
  from .grpo import GRPO
@@ -30,4 +31,5 @@ __all__ = [
30
31
  "PPO",
31
32
  "TD3",
32
33
  "GRPO",
34
+ "DPO",
33
35
  ]