metadata
dataset_info:
- config_name: 20231115-1
features:
- name: ID
dtype: string
- name: text
dtype: string
- name: output
dtype: string
splits:
- name: train
num_bytes: 2007875
num_examples: 1729
download_size: 1148243
dataset_size: 2007875
- config_name: 20231115-2
features:
- name: ID
dtype: string
- name: text
dtype: string
- name: output
dtype: string
splits:
- name: train
num_bytes: 341973
num_examples: 316
download_size: 179947
dataset_size: 341973
- config_name: 20231115-5
features:
- name: ID
dtype: string
- name: text
dtype: string
- name: output
dtype: string
splits:
- name: train
num_bytes: 976579
num_examples: 858
download_size: 434425
dataset_size: 976579
- config_name: 20231221-002
features:
- name: ID
dtype: string
- name: text
dtype: string
- name: output
dtype: string
splits:
- name: train
num_bytes: 3018531
num_examples: 1899
download_size: 1633772
dataset_size: 3018531
- config_name: 20231221-003
features:
- name: ID
dtype: string
- name: text
dtype: string
- name: output
dtype: string
splits:
- name: train
num_bytes: 3018541
num_examples: 1899
download_size: 1633766
dataset_size: 3018541
configs:
- config_name: 20231115-1
data_files:
- split: train
path: 20231115-1/train-*
- config_name: 20231115-2
data_files:
- split: train
path: 20231115-2/train-*
- config_name: 20231115-5
data_files:
- split: train
path: 20231115-5/train-*
- config_name: 20231221-002
data_files:
- split: train
path: 20231221-002/train-*
- config_name: 20231221-003
data_files:
- split: train
path: 20231221-003/train-*
license: cc-by-nc-sa-4.0
task_categories:
- text-generation
language:
- ja
pretty_name: ichikara-instruction
size_categories:
- 1K<n<10K
ichikara-instruction (Non Commercial)
公開ページより、
本データに関して、言語処理学会第30回年次大会において発表を行います。データを使われた方は、HPと共に下記の通りにお願いします。
関根聡, 安藤まや, 後藤美知子, 鈴木久美, 河原大輔, 井之上直也, 乾健太郎. ichikara-instruction: LLMのための日本語インストラクションデータの構築. 言語処理学会第30回年次大会(2024)
論文: https://www.anlp.jp/proceedings/annual_meeting/2024/pdf_dir/A6-3.pdf