[Bug]:

[robinsonmhj]

Notebook with bug

LLM_finetuning_and_batch_inference.ipynb

What happened?

Get the following errors while running the following cell
trainer = HuggingFaceTrainer( trainer_init_per_worker=trainer_init_per_worker, scaling_config=ScalingConfig(num_workers=num_workers, use_gpu=use_gpu), datasets={ "train": train_dataset, "evaluation": validation_dataset, }, run_config=RunConfig( checkpoint_config=CheckpointConfig( num_to_keep=1, checkpoint_score_attribute="eval_loss", checkpoint_score_order="min", ), ), preprocessor=batch_preprocessor, )

Environment info

ray 2.8 python3.9

Issue Severity

High: It blocks me from completing my task.

[robinsonmhj]

@kamil-kaczmarek is it possible to update the document or rewrite the function using ray 2.8?

[kamil-kaczmarek]

Hi @robinsonmhj, thanks for pointing this out.

We introduced new, simpler API in Ray 2.7, including unified Torch based Trainers.

Have a look at new integration with Transformers, that will help you navigate the transition.