View Source AWS.SagemakerJobRuntime (aws-elixir v1.0.12)

Agentic RFT Runtime Service - manages trajectory and transition data for reinforcement fine-tuning jobs.

Link to this section Summary

Functions

Marks a rollout as complete, indicating that no further turns will be appended to the trajectory.

Sends an inference request to the model during a job execution.

Sends a streaming inference request to the model during a job execution.

Updates the reward values for a trajectory and transitions it to reward-received status, signaling that it is eligible for processing.

Link to this section Functions

Link to this function

complete_rollout(client, input, options \\ [])

View Source

Marks a rollout as complete, indicating that no further turns will be appended to the trajectory.

After calling this operation, the trajectory is sealed and eligible for reward submission via the UpdateReward operation.

Link to this function

sample(client, input, options \\ [])

View Source

Sends an inference request to the model during a job execution.

The request and response bodies are forwarded to and from the model without modification. Each turn (prompt and response) is captured for later use.

Link to this function

sample_with_response_stream(client, input, options \\ [])

View Source

Sends a streaming inference request to the model during a job execution.

Returns the response as a stream of payload chunks. Each turn is captured for later use.

Link to this function

update_reward(client, input, options \\ [])

View Source

Updates the reward values for a trajectory and transitions it to reward-received status, signaling that it is eligible for processing.

Call this operation after CompleteRollout to provide the computed reward scores.