hchhtc123 / news-text-classification

飞桨常规赛：中文新闻文本标题分类9月第1名方案，分数0.9+，基于PaddleNLP通过预训练模型的微调完成新闻14分类模型的训练与优化

• 📫 联系我： QQ: 1075558916； 邮箱： [email protected]

news-text-classification's Issues

求解答

---

ValueError Traceback (most recent call last)
/tmp/ipykernel_203/1672709690.py in
4
5
----> 6 train_ds, dev_ds = load_dataset(splits=["train", "dev"])

/tmp/ipykernel_203/1062691189.py in load_dataset(name, data_files, splits, lazy, **kwargs)
45 reader_instance = reader_cls(lazy=lazy, name=name, **kwargs)
46
---> 47 datasets = reader_instance.read_datasets(data_files=data_files,splits=splits)
48 return datasets

/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddlenlp/datasets/dataset.py in read_datasets(self, splits, data_files)
526 while not os.path.exists(lock_file):
527 time.sleep(1)
--> 528 datasets.append(self.read(filename=filename, split=split))
529
530 if data_files:

/opt/conda/envs/python35-paddle120-env/lib/python3.7/site-packages/paddlenlp/datasets/dataset.py in read(self, filename, split)
633 raise ValueError(
634 "No instances were read from the given filepath {}. "
--> 635 "Is the path correct?".format(filename))
636
637 # We need to check if the example contains label column and confirm its name.

ValueError: No instances were read from the given filepath dev.csv. Is the path correct?