RewardCompositionScenario
Training scenarios · Implementation: mlsysim/engine/scenario_definitions/training.py.
Construct this scenario with keyword inputs. The fields below show the types and defaults declared by this class.
One update’s rollouts scored by a verifier that accepts some failures.
Notes
Equation: Composition of mlsysim.solvers.RewardCompositionModel.solve.
Assumptions: Inputs are routed to RewardCompositionModel.solve; their documented regimes and units apply to this configuration.
Excludes: Additional mechanisms beyond the linked solver contracts; the scenario does not collect measurements.
References
Derivation: This scenario binds named inputs to RewardCompositionModel. The linked solver contracts own equations, sources, and result-unit definitions.
Examples
Evaluate the named configuration and display a modeled result.
>>> from mlsysim import Scenarios
>>> case = Scenarios.VerifierFalseAcceptsAtTenPercentPass
>>> value = case.result.false_share
>>> print(f"{value:.3g}")
0.184Attributes
| Name | Description |
|---|---|
| false_accept_rate | |
| group | |
| pass_rate | |
| result | Evaluate the stated reward composition case through RewardCompositionModel.solve; returns its typed result and quantity units. |
| tasks | |
| valid_accept_rate |
tasks
Type: int.
Declared default: required.
group
Type: int.
Declared default: required.
pass_rate
Type: float.
Declared default: required.
false_accept_rate
Type: float.
Declared default: required.
valid_accept_rate
Type: float.
Declared default: required.
result
engine.scenario_types.RewardCompositionScenario.result
Evaluate the stated reward composition case through RewardCompositionModel.solve; returns its typed result and quantity units.
Calculation: mlsysim.solvers.RewardCompositionModel.solve.