remove unused util

Signed-off-by: Kyle Sayers <kylesayrs@gmail.com>

src/llmcompressor/transformers/finetune/data/data_helpers.py

@@ -5,7 +5,6 @@
 
 import torch
 from datasets import Dataset, load_dataset
-from torch.nn.utils.rnn import pad_sequence
 from torch.utils.data import DataLoader, RandomSampler, SequentialSampler
 from transformers.data import default_data_collator
 
@@ -15,7 +14,6 @@
 LABELS_MASK_VALUE = -100
 
 __all__ = [
-    "create_batch_dataloader",
     "format_calibration_data",
     "get_raw_dataset",
     "make_dataset_splits",
@@ -24,44 +22,6 @@
 ]
 
 
-def create_batch_dataloader(
-    dataloader: torch.utils.data.DataLoader,
-    batch_size: int,
-) -> torch.utils.data.DataLoader:
-    """
-    Create a dataloader whose batch size is equal to the size of the dataset
-
-    :param dataset: dataset used to generate dataloader
-    :param batch_size: batch size of new dataloader
-    :return: dataloader
-    """
-    dataset = dataloader.dataset
-    sampler = dataloader.sampler.__class__(dataset)
-
-    def pad_sequences(batch):
-        # extract input_ids and attention_mask from the batch
-        input_ids = [torch.tensor(item["input_ids"]).squeeze(0) for item in batch]
-        masks = [torch.tensor(item["attention_mask"]).squeeze(0) for item in batch]
-
-        # while 0 is not necessarily the "correct" padding value, the padded
-        # input_ids are ignored according to the attention_mask
-        pad_input_ids = pad_sequence(input_ids, batch_first=True, padding_value=0)
-        pad_masks = pad_sequence(masks, batch_first=True, padding_value=0)
-
-        return {
-            "input_ids": pad_input_ids,
-            "attention_mask": pad_masks,
-        }
-
-    return torch.utils.data.DataLoader(
-        dataset,
-        batch_size=batch_size,
-        sampler=sampler,
-        collate_fn=pad_sequences,
-        pin_memory=True,
-    )
-
-
 def format_calibration_data(
     tokenized_dataset: Dataset,
     num_calibration_samples: Optional[int] = None,