# LLM tutorial 7 classical NLP task

**URL:** <https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760>\
**Category:** Beginners\
**Created:** [September 29, 2025, 7:05am UTC](https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760 "2025-09-29T07:05:37Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Mizz54112](https://avatars.discourse-cdn.com/v4/letter/m/f04885/32.png) [@Mizz54112](https://discuss.huggingface.co/u/Mizz54112)\
**Post date:** [September 29, 2025, 7:05am UTC](https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760/1 "2025-09-29T07:05:37Z")

</div>

raw\_datasets = load\_dataset(“conll2003”)

this code is not working. Error message is “Dataset scripts are no longer supported, but found conll2003.py”. How can I solve this problem? Please let me know.

---

<div class="post-metadata">

**Author:** ![John6666](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/john6666/32/27664_2.png) [@John6666](https://discuss.huggingface.co/u/John6666)\
**Post date:** [September 29, 2025, 9:56am UTC](https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760/2 "2025-09-29T09:56:58Z")

</div>

> Dataset scripts are no longer supported

[The cause is literally as stated, so either pin the datasets library to an older version](https://discuss.huggingface.co/t/dataset-scripts-are-no-longer-supported/163891/2) or try the code below (using a dataset without builder scripts).

```py
raw_datasets = load_dataset("lhoestq/conll2003")

```

---

<div class="post-metadata">

**Author:** ![Mizz54112](https://avatars.discourse-cdn.com/v4/letter/m/f04885/32.png) [@Mizz54112](https://discuss.huggingface.co/u/Mizz54112)\
**Post date:** [September 30, 2025, 3:27am UTC](https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760/3 "2025-09-30T03:27:30Z")

</div>

```auto
Thank you for answering my question. 
I have another question: Why doesn't the error disappear when I add the word lhoestq/ to the code?

```

---

<div class="post-metadata">

**Author:** ![John6666](https://sea2.discourse-cdn.com/hellohellohello/user_avatar/discuss.huggingface.co/john6666/32/27664_2.png) [@John6666](https://discuss.huggingface.co/u/John6666)\
**Post date:** [September 30, 2025, 6:37am UTC](https://discuss.huggingface.co/t/llm-tutorial-7-classical-nlp-task/168760/4 "2025-09-30T06:37:05Z")

</div>

Yeah. When specifying `conll2003` (the abbreviated notation without author name), `eriktks/conll2003` is probably loaded implicitly. However, specifying `lhoestq/conll2003` loads a different dataset without the data builder script.

The dataset contents themselves are probably identical.