metadata
dataset_info:
features:
- name: context
dtype: string
- name: instruction
dtype: string
- name: answer
dtype: string
splits:
- name: train
- name: validation
- name: test
tags:
- synthetic
- spino
extra_gated_prompt: >-
Thank you for your interest in this dataset, please provide us a few details.
We are a synthetic data generation startup and hence would like you to suggest
other domains for which we can release datasets.
extra_gated_button_content: Join the Waitlist and Access Dataset
extra_gated_fields:
Company: text
Usecases: text
Dataset Card for Mental Health
This instruction tuning dataset has been created with SPInO and has 1072 rows of textual data related to Mental Health.
from datasets import load_dataset
dataset = load_dataset(1rsh/SPInO-MentalHealthQA)
print(dataset)