metadata
dataset_info:
features:
- name: INSTRUCTION
dtype: string
- name: RESPONSE
dtype: string
- name: SOURCE
dtype: string
- name: METADATA
dtype: string
splits:
- name: train
num_bytes: 5852058
num_examples: 5654
download_size: 2665151
dataset_size: 5852058
configs:
- config_name: default
data_files:
- split: train
path: data/train-*
license: apache-2.0
task_categories:
- text-generation
- question-answering
- conversational
language:
- fa
pretty_name: Alpaca Farsi Instruction
size_categories:
- 10K<n<100K
Dataset Card for "alpaca-fa-instruction"
This dataset was first created here and is published to huggingface according to Open-assistant standard.