I am getting this error: ValueError: Trainer: evaluation requires an eval_dataset.

[furkansherwani]

ValueError Traceback (most recent call last)
in <cell line: 4>()
2 get_ipython().system(' pip install -U accelerate')
3 get_ipython().system(' pip install -U transformers')
----> 4 model_trainer = t5_exp.train(id_tokenized_ds, **training_args)

6 frames
/usr/local/lib/python3.10/dist-packages/transformers/trainer.py in get_eval_dataloader(self, eval_dataset)
886 """
887 if eval_dataset is None and self.eval_dataset is None:
--> 888 raise ValueError("Trainer: evaluation requires an eval_dataset.")
889 eval_dataset = eval_dataset if eval_dataset is not None else self.eval_dataset
890 data_collator = self.data_collator

ValueError: Trainer: evaluation requires an eval_dataset.

[kevinscaria]

The script requires an eval dataset to be provided as an argument. However, please try to debug this as this repo is not actively maintained.

[furkansherwani]

But there is no .csv file in validation folder in Datasets. It is a .json file. Please help me understand this.

[cyborgrob]

@furkansherwani Just split the test dataset (or split it however you want really):

# Split the test dataset in half
train_test_split = id_tokenized_ds['test'].train_test_split(test_size=0.5)

Then rename one portion to 'validation':

id_tokenized_ds['test'] = train_test_split['train']
id_tokenized_ds['validation'] = train_test_split['test'] # Use 'test' as the validation set
id_tokenized_ds

DatasetDict({
train: Dataset({
features: ['raw_text', 'aspectTerms', 'labels', 'text', 'index_level_0', 'input_ids', 'attention_mask'],
num_rows: 590
})
test: Dataset({
features: ['raw_text', 'aspectTerms', 'labels', 'text', 'input_ids', 'attention_mask'],
num_rows: 127
})
validation: Dataset({
features: ['raw_text', 'aspectTerms', 'labels', 'text', 'input_ids', 'attention_mask'],
num_rows: 127
})
})

This solved the error in my case.