from datasets import load_dataset, concatenate_datasets

datasets = []
for chunk_num in range(1, num_chunks+1):  # Replace num_chunks with the actual number of chunks
    dataset = load_dataset('csv', data_files=[f'train_{chunk_num}.csv'])
    datasets.append(dataset)

# Concatenate all parts into one dataset
full_dataset = concatenate_datasets(datasets)