Hello Hugging Face Community,

I apologize for creating a pull request first instead of opening a discussion regarding the issue I encountered with the mtop_intent_classification dataset. I should have consulted the community first before proposing changes.

Issue:

I was facing an issue where I couldn’t load any data from the mtop_intent_classification dataset. Despite using the correct methods, the data failed to load. The issue seems to stem from a recent change I made in the dataset code, which adjusts how labels are assigned and how the dataset is referenced.

Steps to Reproduce:

Running the following code:

from datasets import load_dataset
dataset = load_dataset("SEACrowd/mtop_intent_classification")
The dataset fails to load, either silently or with errors related to fetching the data.

Expected Behavior:

The dataset should load correctly, providing data for both the "domain" and "intent" subsets.

Current Workaround:

I've tried different ways to load the dataset, but have not been able to resolve the issue.

I have already created a pull request to address this issue, which includes fixing the conditional logic for label assignment and updating the dataset reference URL in the _info and _load_hf_data_from_remote methods. However, I realize now that it would have been better to open a discussion first to ensure these changes were necessary.

Thank you for your understanding, and I look forward to any suggestions or feedback you may have.