openai-core/com.aallam.openai.api.finetune/HyperParams/promptLossWeight

promptLossWeight

@SerialName(value = "prompt_loss_weight")
val promptLossWeight: Double

The weight to use for loss on the prompt tokens. This controls how much the model tries to learn to generate the prompt (as compared to the completion which always has a weight of 1.0), and can add a stabilizing effect to training when completions are short.