# tinker_cookbook.distillation.DatasetWithTeacher

## _class_ [**tinker_cookbook.distillation.DatasetWithTeacher**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/distillation/train_off_policy.py#L70)()

Pairs a supervised dataset with its teacher model.

Each domain in multi-teacher distillation gets one of these. The teacher's highest-probability tokens are used as soft targets for training.

**Fields:**

- [**dataset_builder**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/distillation/train_off_policy.py#L82) ( _[SupervisedDatasetBuilder](https://tinker-docs.thinkingmachines.ai/cookbook/api-reference/supervised/superviseddatasetbuilder/) _)
- [**teacher_config**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/distillation/train_off_policy.py#L83) ( _[TeacherConfig](https://tinker-docs.thinkingmachines.ai/cookbook/api-reference/distillation/teacherconfig/) _)
- [**weight**](https://github.com/thinking-machines-lab/tinker-cookbook/blob/main/tinker_cookbook/distillation/train_off_policy.py#L84) ( _float_, default: `1.0`)

## Referenced by

- [tinker_cookbook.distillation.train_off_policy.Config](https://tinker-docs.thinkingmachines.ai/cookbook/api-reference/distillation/train_off_policy_config/)