RL Agent Status
Live status for the DQN agents used in signal generation.
Q Equity DQN
Loading Q status...
- Training Steps: total learning updates. More steps usually means more stable behavior.
- Epsilon: exploration rate. Higher means more random actions; lower means more exploitation.
- Stage: training-volume label, not a guarantee of future returns.
- Runtime: should show DQN when neural weights are loaded.
- Checkpoint Format: storage schema version only, not the number of past model revisions.
- Cloud Enabled: Yes means RL state persists across restarts and users.
- Universe scans: scanning does not train the model live. Training steps move when learning updates are saved from evaluated outcomes and backtests.