trinity.common.workflows.step_wise_workflow module

class trinity.common.workflows.step_wise_workflow.StepWiseRewardWorkflow(*, task: Task, model: ModelWrapper, auxiliary_models=None, use_openai_client=True)[source]

Bases: Workflow

A workflow that implements step-wise rewards for tasks.

__init__(*, task: Task, model: ModelWrapper, auxiliary_models=None, use_openai_client=True)[source]

run() → list[Experience][source]: Run the workflow and return a list of experiences with step-wise rewards.

abstract step(step_num: int) → bool[source]

Run a single step of your agent application.

Parameters:: step_num (int) – The current step number.
Returns:: Whether to continue running the agent application.
Return type:: bool

Tips:: You can use the openai client (self.client) to migrate your existing applications at low cost.

abstract reward(exps: list[Experience], step_num: int) → float[source]: Calculate the reward for the given experiences at the specified step.

abstract property max_step_num: Return the maximum number of steps in the task.

property repeatable: A workflow is repeatable if it can be run multiple times within the run() method.

class trinity.common.workflows.step_wise_workflow.RewardPropagationWorkflow(*, task: Task, model: ModelWrapper, auxiliary_models=None, use_openai_client=True)[source]

Bases: Workflow

A workflow that propagates rewards across multiple turns.

__init__(*, task: Task, model: ModelWrapper, auxiliary_models=None, use_openai_client=True)[source]

run() → list[Experience][source]: Run the workflow and return a list of experiences with step-wise rewards.

abstract step(step_num: int) → bool[source]

Run a single step of your agent application.

Parameters:: step_num (int) – The current step number.
Returns:: Whether to continue running the agent application.
Return type:: bool

Tips:: You can use the openai client (self.client) to migrate your existing applications at low cost.

abstract reward(exps: list[Experience]) → float[source]: Calculate the reward for the given experiences of the entire run.

abstract property max_step_num: Return the maximum number of steps in the task.

property repeatable: A workflow is repeatable if it can be run multiple times within the run() method.