core.trainers.dpo.trainer

core.trainers.dpo.trainer

DPO trainer for axolotl

Classes

Name	Description
AxolotlDPOTrainer	Extend the base DPOTrainer for axolotl helpers

AxolotlDPOTrainer

core.trainers.dpo.trainer.AxolotlDPOTrainer(
    self,
    *args,
    dataset_tags=None,
    **kwargs,
)

Extend the base DPOTrainer for axolotl helpers

Methods

Name	Description
evaluation_loop	Overriding built-in evaluation loop to store metrics for each batch.
push_to_hub	Overwrite the push_to_hub method in order to force-add the tags when pushing the

evaluation_loop

core.trainers.dpo.trainer.AxolotlDPOTrainer.evaluation_loop(
    dataloader,
    description,
    prediction_loss_only=None,
    ignore_keys=None,
    metric_key_prefix='eval',
)

Overriding built-in evaluation loop to store metrics for each batch. Prediction/evaluation loop, shared by Trainer.evaluate() and Trainer.predict().

Works both with or without labels.

push_to_hub

core.trainers.dpo.trainer.AxolotlDPOTrainer.push_to_hub(*args, **kwargs)

Overwrite the push_to_hub method in order to force-add the tags when pushing the model on the Hub. Please refer to ~transformers.Trainer.push_to_hub for more details.