ValueError: train_dataset does not implement __len__, max_steps has to be specified

[widyaputeriaulia10]

dear authors, i tried to re-implemented IndoBERT based on your code but when i train the model :

metric = load_metric("accuracy")

def compute_metrics(eval_pred):
logits, labels = eval_pred
predictions = np.argmax(logits, axis=-1)

return metric.compute(predictions=predictions, references=labels)

trainer = Trainer(
model=model,
args=training_args,
train_dataset=train_dataset,
eval_dataset=eval_dataset,
compute_metrics=compute_metrics,
)

train_history = trainer.train()
train_evaluate = trainer.evaluate()

i found the issue like this:

ValueError Traceback (most recent call last)
in
12 train_dataset=train_dataset,
13 eval_dataset=eval_dataset,
---> 14 compute_metrics=compute_metrics,
15 )
16

/setneg-dir/Uli/uli-env/lib/python3.6/site-packages/transformers/trainer.py in init(self, model, args, data_collator, train_dataset, eval_dataset, tokenizer, model_init, compute_metrics, callbacks, optimizers, preprocess_logits_for_metrics)
427
428 if train_dataset is not None and not has_length(train_dataset) and args.max_steps <= 0:
--> 429 raise ValueError("train_dataset does not implement len, max_steps has to be specified")
430
431 if (

ValueError: train_dataset does not implement len, max_steps has to be specified

may i ask you for your advices to solve the problem?

Thank You