KoalaTrapLord
f51ff61568
* feat(83): Phase 3.1 — persistent JVM worker pool in MultiThreadBatchRunner (#83) * feat(83): Phase 3.2 — async Forge episode generator with producer-consumer pipeline (#83) * feat(83): Phase 3.3 — deprecate self_play.py synthetic mode for training (#83) Mark self_play.py synthetic training as deprecated per Phase 3.3: - Add deprecation notice to module docstring - Add import warnings and _TRAINING_DEPRECATION_MSG constant - Add DeprecationWarning in collect_rollouts() - Synthetic sim retained for local testing/unit tests only - scope.py already defaults forge_weight=1.0, ppo_weight=0.0 * feat(83): Phase 3.3 — update ml/training/__init__.py exports (#83)
30 lines
914 B
Python
30 lines
914 B
Python
"""
|
|
Commander AI Lab - Training Subpackage
|
|
======================================
|
|
Provides RL training infrastructure including PPO, rollout buffers,
|
|
Forge episode generation, self-play, and distillation.
|
|
"""
|
|
|
|
from ml.training.rollout_buffer import RolloutBuffer, ForgeEpisodeQueue
|
|
from ml.training.reward import compute_intermediate_reward, compute_terminal_reward, RewardConfig
|
|
from ml.training.decision_exporter import DecisionExporter
|
|
from ml.training.forge_episode_generator import (
|
|
ForgeEpisodeGenerator,
|
|
AsyncForgeEpisodeGenerator,
|
|
)
|
|
|
|
__all__ = [
|
|
# Rollout & Buffers
|
|
"RolloutBuffer",
|
|
"ForgeEpisodeQueue",
|
|
# Rewards
|
|
"compute_intermediate_reward",
|
|
"compute_terminal_reward",
|
|
"RewardConfig",
|
|
# Decision export
|
|
"DecisionExporter",
|
|
# Forge episode generation (Phase 3)
|
|
"ForgeEpisodeGenerator",
|
|
"AsyncForgeEpisodeGenerator",
|
|
]
|