Datasets:

---

bioscan-ml

/

CanadianInvertebrates-ML

like 0

Follow

BIOSCAN 19

This PR updates the dataset card for CanadianInvertebrates-ML by:

• Updating size_categories to 1M<n<10M to reflect the dataset's stated size of 1.5M DNA barcodes.
• Adding text-classification to task_categories, as the dataset is primarily designed for taxonomic identification, species classification, and genus identification tasks.
• Adding library_name: transformers to the metadata, as the dataset is used in conjunction with models from the transformers library.
• Adding a prominent link to the paper on Hugging Face Papers and including the paper's abstract for quick context.
• Populating the "Sample Usage" section with practical Python code demonstrating how to use the dataset with the associated BarcodeBERT model.
• Adding the Zenodo link to the "Dataset Sources" section as found in the associated GitHub repository.

Hi @nielsr thank you very much for this PR. I am familiar with your tutorials so honoured/star struck to receive your PR.

I have reviewed the proposed changes and everything looks great. Replacing the To-Do sample usage with actual sample usage is a great addition!

I noticed that you added the text-classification tag. Since this is a dataset comprised of DNA barcode sequences ("ACTG" only) rather than natural language sequences, do you feel that this task category is appropriate? I noticed that on huggingface.co/tasks "Text Classification" is nested under NLP. I would not consider this to be a NLP project or aimed toward a NLP task. Rather, we were motivated to repurpose models developed for NLP to work on structured language, specifically DNA.

Feel free to adapt the dataset card according to your preferences :)