OakEmu-RL v0.1.0

Appearance

mrkwatkins.oakemu_rl.framework.metrics.MetricNames ​

Classes and functions ​

Constants ​

ACTOR_ENV_STEPS ​

python
ACTOR_ENV_STEPS = 'Actor/Env_Steps'

Source

ACTOR_EPISODE ​

python
ACTOR_EPISODE = 'Actor/Episode'

Source

ACTOR_EPISODE_EXTRINSIC_REWARD ​

python
ACTOR_EPISODE_EXTRINSIC_REWARD = 'Actor/Episode_Extrinsic_Reward'

Source

ACTOR_EPISODE_INTRINSIC_REWARD ​

python
ACTOR_EPISODE_INTRINSIC_REWARD = 'Actor/Episode_Intrinsic_Reward'

Source

ACTOR_EPISODE_INTRINSIC_REWARD_FRACTION ​

python
ACTOR_EPISODE_INTRINSIC_REWARD_FRACTION = 'Actor/Episode_Intrinsic_Reward_Fraction'

Source

ACTOR_EPISODE_LENGTH ​

python
ACTOR_EPISODE_LENGTH = 'Actor/Episode_Length'

Source

ACTOR_EPISODE_LOSS_PERCENT ​

python
ACTOR_EPISODE_LOSS_PERCENT = 'Actor/Episode_Loss_Percentage'

Source

ACTOR_EPISODE_LOSS_TOTAL ​

python
ACTOR_EPISODE_LOSS_TOTAL = 'Actor/Episode_Loss_Total'

Source

ACTOR_EPISODE_REWARD ​

python
ACTOR_EPISODE_REWARD = 'Actor/Episode_Reward'

Source

ACTOR_EPISODE_SHAPED_REWARD ​

python
ACTOR_EPISODE_SHAPED_REWARD = 'Actor/Episode_Shaped_Reward'

Source

ACTOR_EPISODE_TOTAL ​

python
ACTOR_EPISODE_TOTAL = 'Actor/Episode_Total'

Source

ACTOR_EPISODE_TRUNCATED_PERCENT ​

python
ACTOR_EPISODE_TRUNCATED_PERCENT = 'Actor/Episode_Truncated_Percentage'

Source

ACTOR_EPISODE_TRUNCATED_TOTAL ​

python
ACTOR_EPISODE_TRUNCATED_TOTAL = 'Actor/Episode_Truncated_Total'

Source

ACTOR_EPISODE_WIN_PERCENT ​

python
ACTOR_EPISODE_WIN_PERCENT = 'Actor/Episode_Win_Percentage'

Source

ACTOR_EPISODE_WIN_TOTAL ​

python
ACTOR_EPISODE_WIN_TOTAL = 'Actor/Episode_Win_Total'

Source

ACTOR_EPSILON ​

python
ACTOR_EPSILON = 'Actor/Epsilon'

Source

ACTOR_EXTRINSIC_REWARD ​

python
ACTOR_EXTRINSIC_REWARD = 'Actor/Extrinsic_Reward'

Source

ACTOR_INTRINSIC_REWARD ​

python
ACTOR_INTRINSIC_REWARD = 'Actor/Intrinsic_Reward'

Source

ACTOR_POLICY_ENTROPY ​

python
ACTOR_POLICY_ENTROPY = 'Actor/Policy_Entropy'

Source

ACTOR_POLICY_GREEDY_FRACTION ​

python
ACTOR_POLICY_GREEDY_FRACTION = 'Actor/Policy_GreedyFraction'

Source

ACTOR_REPLAY_CACHE_ENTRIES_TOTAL ​

python
ACTOR_REPLAY_CACHE_ENTRIES_TOTAL = 'Actor/Replay_Cache_Entries_Total'

Source

ACTOR_REPLAY_CACHE_GB ​

python
ACTOR_REPLAY_CACHE_GB = 'Actor/Replay_Cache_GB'

Source

ACTOR_REPLAY_CACHE_HITS_TOTAL ​

python
ACTOR_REPLAY_CACHE_HITS_TOTAL = 'Actor/Replay_Cache_Hits_Total'

Source

ACTOR_REPLAY_CACHE_MISSES_TOTAL ​

python
ACTOR_REPLAY_CACHE_MISSES_TOTAL = 'Actor/Replay_Cache_Misses_Total'

Source

ACTOR_RESULT_QUEUE_DEPTH ​

python
ACTOR_RESULT_QUEUE_DEPTH = 'Actor/Result_Queue_Depth'

Source

ACTOR_SHAPED_REWARD ​

python
ACTOR_SHAPED_REWARD = 'Actor/Shaped_Reward'

Source

ALL_METRICS ​

python
ALL_METRICS: frozenset[str] = frozenset(value for name, value in globals().items() if isinstance(value, str) and not name.startswith('_') and name != 'ALL_METRICS')

Source

BUFFER_MEMORY_GB ​

python
BUFFER_MEMORY_GB = 'Buffer/Memory_GB'

Source

BUFFER_REPLAY_QUEUE_DEPTH ​

python
BUFFER_REPLAY_QUEUE_DEPTH = 'Buffer/Replay_Queue_Depth'

Source

BUFFER_REWARD ​

python
BUFFER_REWARD = 'Buffer/Reward'

Source

BUFFER_SIZE ​

python
BUFFER_SIZE = 'Buffer/Size'

Source

BUFFER_SKIPPED_UPDATES ​

python
BUFFER_SKIPPED_UPDATES = 'Buffer/Skipped_Updates'

Source

BUFFER_SKIPPED_UPDATES_TOTAL ​

python
BUFFER_SKIPPED_UPDATES_TOTAL = 'Buffer/Skipped_Updates_Total'

Source

BUFFER_STAGED_TRANSITIONS ​

python
BUFFER_STAGED_TRANSITIONS = 'Buffer/Staged_Transitions'

Source

CHECKPOINT_AUTO_ROLLBACK ​

python
CHECKPOINT_AUTO_ROLLBACK = 'Checkpoint/Auto_Rollback'

Source

CHECKPOINT_AUTO_ROLLBACK_TOTAL ​

python
CHECKPOINT_AUTO_ROLLBACK_TOTAL = 'Checkpoint/Auto_Rollback_Total'

Source

CHECKPOINT_NUMBER ​

python
CHECKPOINT_NUMBER = 'Checkpoint/Number'

Source

LEARNER_ACTIVATION_DEAD_PERCENTAGE_PREFIX ​

python
LEARNER_ACTIVATION_DEAD_PERCENTAGE_PREFIX = 'Learner/Activation_Dead_Percentage_'

Per activation layer: percentage of units that never fire on the sampled minibatch. Suffix is the layer name.

Source

LEARNER_ACTIVATION_EFFECTIVE_RANK_PREFIX ​

python
LEARNER_ACTIVATION_EFFECTIVE_RANK_PREFIX = 'Learner/Activation_Effective_Rank_'

Per activation layer: effective rank of the layer's output on the sampled minibatch. Suffix is the layer name.

Source

LEARNER_ACTIVATION_NEAR_DEAD_PERCENTAGE_PREFIX ​

python
LEARNER_ACTIVATION_NEAR_DEAD_PERCENTAGE_PREFIX = 'Learner/Activation_Near_Dead_Percentage_'

Per rectifier layer: percentage of units active on under 1% of the sampled inputs. Suffix is the layer name.

Source

LEARNER_ADVANTAGES_MEAN ​

python
LEARNER_ADVANTAGES_MEAN = 'Learner/Advantages_Mean'

Source

LEARNER_ADVANTAGES_STD ​

python
LEARNER_ADVANTAGES_STD = 'Learner/Advantages_Std'

Source

LEARNER_APPROX_KL ​

python
LEARNER_APPROX_KL = 'Learner/Approx_KL'

Source

LEARNER_BC_LOSS ​

python
LEARNER_BC_LOSS = 'Learner/BC_Loss'

Source

LEARNER_CATEGORICAL_KL_LOSS ​

python
LEARNER_CATEGORICAL_KL_LOSS = 'Learner/Categorical_KL_Loss'

Source

LEARNER_CLIP_FRACTION ​

python
LEARNER_CLIP_FRACTION = 'Learner/Clip_Fraction'

Source

LEARNER_ENTROPY ​

python
LEARNER_ENTROPY = 'Learner/Entropy'

Source

LEARNER_ENTROPY_LOSS ​

python
LEARNER_ENTROPY_LOSS = 'Learner/Entropy_Loss'

Source

LEARNER_EPISODE_EXTRINSIC_REWARD ​

python
LEARNER_EPISODE_EXTRINSIC_REWARD = 'Learner/Episode_Extrinsic_Reward'

Source

LEARNER_EPISODE_INTRINSIC_REWARD ​

python
LEARNER_EPISODE_INTRINSIC_REWARD = 'Learner/Episode_Intrinsic_Reward'

Source

LEARNER_EPISODE_INTRINSIC_REWARD_FRACTION ​

python
LEARNER_EPISODE_INTRINSIC_REWARD_FRACTION = 'Learner/Episode_Intrinsic_Reward_Fraction'

Source

LEARNER_EPISODE_SHAPED_REWARD ​

python
LEARNER_EPISODE_SHAPED_REWARD = 'Learner/Episode_Shaped_Reward'

Source

LEARNER_EXPLAINED_VARIANCE ​

python
LEARNER_EXPLAINED_VARIANCE = 'Learner/Explained_Variance'

Source

LEARNER_EXTRINSIC_REWARD ​

python
LEARNER_EXTRINSIC_REWARD = 'Learner/Extrinsic_Reward'

Source

LEARNER_GRADIENT_NORM ​

python
LEARNER_GRADIENT_NORM = 'Learner/Gradient_Norm'

Source

LEARNER_INITIAL_ITERATIONS ​

python
LEARNER_INITIAL_ITERATIONS = 'Learner/Initial_Iterations'

Source

LEARNER_INTRINSIC_REWARD ​

python
LEARNER_INTRINSIC_REWARD = 'Learner/Intrinsic_Reward'

Source

LEARNER_LOSS ​

python
LEARNER_LOSS = 'Learner/Loss'

Source

LEARNER_LR ​

python
LEARNER_LR = 'Learner/LR'

Source

LEARNER_N_STEP ​

python
LEARNER_N_STEP = 'Learner/NStep'

Source

LEARNER_PER_WEIGHTED_LOSS ​

python
LEARNER_PER_WEIGHTED_LOSS = 'Learner/PER_Weighted_Loss'

Source

LEARNER_POLICY_LOSS ​

python
LEARNER_POLICY_LOSS = 'Learner/Policy_Loss'

Source

LEARNER_Q ​

python
LEARNER_Q = 'Learner/Q'

Source

LEARNER_RATIO_MEAN ​

python
LEARNER_RATIO_MEAN = 'Learner/Ratio_Mean'

Source

LEARNER_ROLLOUTS_COLLECTED ​

python
LEARNER_ROLLOUTS_COLLECTED = 'Learner/Rollouts_Collected'

Source

LEARNER_SHAPED_REWARD ​

python
LEARNER_SHAPED_REWARD = 'Learner/Shaped_Reward'

Source

LEARNER_TARGET_MODEL_UPDATES ​

python
LEARNER_TARGET_MODEL_UPDATES = 'Learner/Target_Model_Updates'

Source

LEARNER_TD_LOSS ​

python
LEARNER_TD_LOSS = 'Learner/TD_Loss'

Source

LEARNER_UPDATE_ITERATIONS ​

python
LEARNER_UPDATE_ITERATIONS = 'Learner/Update_Iterations'

Source

LEARNER_UPDATE_STEPS ​

python
LEARNER_UPDATE_STEPS = 'Learner/Update_Steps'

Source

LEARNER_VALUE_GAP ​

python
LEARNER_VALUE_GAP = 'Learner/ValueGap'

Source

LEARNER_VALUE_LOSS ​

python
LEARNER_VALUE_LOSS = 'Learner/Value_Loss'

Source

SYSTEM_CPU_PERCENT ​

python
SYSTEM_CPU_PERCENT = 'System/CPU_Percentage'

Source

SYSTEM_ELAPSED_TIME ​

python
SYSTEM_ELAPSED_TIME = 'System/Elapsed_Time'

Source

SYSTEM_GPU_COMPUTE_PERCENT ​

python
SYSTEM_GPU_COMPUTE_PERCENT = 'System/GPU_Compute_Percentage'

Source

SYSTEM_GPU_MEMORY_GB ​

python
SYSTEM_GPU_MEMORY_GB = 'System/GPU_MemoryGB'

Source

SYSTEM_MEMORY_GB ​

python
SYSTEM_MEMORY_GB = 'System/MemoryGB'

Source

SYSTEM_MEMORY_PERCENT ​

python
SYSTEM_MEMORY_PERCENT = 'System/Memory_Percentage'

Source

TESTER_EPISODE ​

python
TESTER_EPISODE = 'Tester/Episode'

Source

TESTER_EPISODE_LENGTH ​

python
TESTER_EPISODE_LENGTH = 'Tester/Episode_Length'

Source

TESTER_EPISODE_LOSS_PERCENT ​

python
TESTER_EPISODE_LOSS_PERCENT = 'Tester/Episode_Loss_Percentage'

Source

TESTER_EPISODE_LOSS_TOTAL ​

python
TESTER_EPISODE_LOSS_TOTAL = 'Tester/Episode_Loss_Total'

Source

TESTER_EPISODE_REWARD ​

python
TESTER_EPISODE_REWARD = 'Tester/Episode_Reward'

Source

TESTER_EPISODE_TOTAL ​

python
TESTER_EPISODE_TOTAL = 'Tester/Episode_Total'

Source

TESTER_EPISODE_TRUNCATED_PERCENT ​

python
TESTER_EPISODE_TRUNCATED_PERCENT = 'Tester/Episode_Truncated_Percentage'

Source

TESTER_EPISODE_TRUNCATED_TOTAL ​

python
TESTER_EPISODE_TRUNCATED_TOTAL = 'Tester/Episode_Truncated_Total'

Source

TESTER_EPISODE_WIN_PERCENT ​

python
TESTER_EPISODE_WIN_PERCENT = 'Tester/Episode_Win_Percentage'

Source

TESTER_EPISODE_WIN_TOTAL ​

python
TESTER_EPISODE_WIN_TOTAL = 'Tester/Episode_Win_Total'

Source

TESTER_RESULT_QUEUE_DEPTH ​

python
TESTER_RESULT_QUEUE_DEPTH = 'Tester/Result_Queue_Depth'

Source

THROUGHPUT_ACTOR_ENV_STEPS_PER_MIN ​

python
THROUGHPUT_ACTOR_ENV_STEPS_PER_MIN = 'Throughput/Actor_Env_Steps_Per_Min'

Source

THROUGHPUT_LEARNER_ITERATIONS_PER_MIN ​

python
THROUGHPUT_LEARNER_ITERATIONS_PER_MIN = 'Throughput/Learner_Iterations_Per_Min'

Source

THROUGHPUT_TESTER_EPISODES_PER_MIN ​

python
THROUGHPUT_TESTER_EPISODES_PER_MIN = 'Throughput/Tester_Episodes_Per_Min'

Source