Skip to content

RFTConfig

Source code in docs_src/capo_bedrock/types/rft_config.py 
12
13
14
15
16
17
18
class RFTConfig(TypedDict, closed=True):
    grader_config: NotRequired["capo_bedrock.types.grader_config.GraderConfig"]
    """<p> Configuration for the grader that evaluates model responses and provides reward signals during RFT training. </p>"""
    hyper_parameters: NotRequired[
        "capo_bedrock.types.rft_hyper_parameters.RFTHyperParameters"
    ]
    """<p> Hyperparameters that control the reinforcement fine-tuning training process, including learning rate, batch size, and epoch count. </p>"""

grader_config instance-attribute 

grader_config: NotRequired[GraderConfig]

Configuration for the grader that evaluates model responses and provides reward signals during RFT training.

hyper_parameters instance-attribute 

hyper_parameters: NotRequired[RFTHyperParameters]

Hyperparameters that control the reinforcement fine-tuning training process, including learning rate, batch size, and epoch count.