ParserError：错误标记数据。在 SMSA indoBert 教程中

Question

我正在尝试使用 DocumentSentimentDataset 和 DocumentSentimentDataLoader 制作简单的训练、验证和测试数据，但出现此错误

/usr/local/lib/python3.8/dist-packages/pandas/_libs/parsers.pyx in pandas._libs.parsers.raise_parser_error()

ParserError: Error tokenizing data. C error: Expected 6 fields in line 25, saw 7

这是我的代码：

from indonlu.utils.data_utils import DocumentSentimentDataset, DocumentSentimentDataLoader

# Load Tokenizer and Config
tokenizer = BertTokenizer.from_pretrained('indobenchmark/indobert-base-p1')
config = BertConfig.from_pretrained('indobenchmark/indobert-base-p1')
config.num_labels = DocumentSentimentDataset.NUM_LABELS

# Instantiate model
model = BertForSequenceClassification.from_pretrained('indobenchmark/indobert-base-p1', config=config)

###
# common functions
###
def set_seed(seed):
    random.seed(seed)
    np.random.seed(seed)
    torch.manual_seed(seed)
    torch.cuda.manual_seed(seed)
    
def count_param(module, trainable=False):
    if trainable:
        return sum(p.numel() for p in module.parameters() if p.requires_grad)
    else:
        return sum(p.numel() for p in module.parameters())
    
def get_lr(optimizer):
    for param_group in optimizer.param_groups:
        return param_group['lr']

def metrics_to_string(metric_dict):
    string_list = []
    for key, value in metric_dict.items():
        string_list.append('{}:{:.2f}'.format(key, value))
    return ' '.join(string_list)

train_dataset_path = '/content/indonlu/dataset/smsa_doc-sentiment-prosa/train_preprocess.tsv'
valid_dataset_path = '/content/indonlu/dataset/smsa_doc-sentiment-prosa/valid_preprocess.tsv'
test_dataset_path = '/content/indonlu/dataset/smsa_doc-sentiment-prosa/test_preprocess_masked_label.tsv'

train_dataset = DocumentSentimentDataset(train_dataset_path, tokenizer, lowercase=True, on_bad_lines='skip')
valid_dataset = DocumentSentimentDataset(valid_dataset_path, tokenizer, lowercase=True, on_bad_lines='skip')
test_dataset = DocumentSentimentDataset(test_dataset_path, tokenizer, lowercase=True, on_bad_lines='skip')

train_loader = DocumentSentimentDataLoader(dataset=train_dataset, max_seq_len=512, batch_size=32, num_workers=16, shuffle=True)  
valid_loader = DocumentSentimentDataLoader(dataset=valid_dataset, max_seq_len=512, batch_size=32, num_workers=16, shuffle=False)  
test_loader = DocumentSentimentDataLoader(dataset=test_dataset, max_seq_len=512, batch_size=32, num_workers=16, shuffle=False)

以及完整的错误信息：

我该如何解决这个问题？