• Stars
    star
    1,000
  • Rank 45,878 (Top 1.0 %)
  • Language
    Python
  • License
    Other
  • Created about 3 years ago
  • Updated 10 months ago

Reviews

There are no reviews yet. Be the first to send feedback to the community and the maintainers!

Repository Details

KakaoBrain KoGPT (Korean Generative Pre-trained Transformer)

KoGPT

KakaoBrain Github: kogpt License: Apache 2.0
huggingface: KoGPT-6B huggingface: KoGPT-6B License: CC BY-NC-ND 4.0

Model Descriptions

KoGPT6B-ryan1.5b

Hyperparameter Value
6,166,502,400
28
4,096
16,384
16
256
2,048
64,512
Positional Encoding Rotary Position Embedding (RoPE)
RoPE Dimensions 64

Hardware requirements

KoGPT6B-ryan1.5b

GPU

The following is the recommended minimum GPU hardware guidance for a handful of example KoGPT.

  • 32GB GPU RAM in the required minimum memory size

KoGPT6B-ryan1.5b-float16

GPU

The following is the recommended minimum GPU hardware guidance for a handful of example KoGPT.

  • half-precision requires NVIDIA GPUS based on Volta, Turing or Ampere
  • 16GB GPU RAM in the required minimum memory size

Usage

prompt

python -m kogpt --help
usage: KoGPT inference [-h] [--model MODEL] [--revision {KoGPT6B-ryan1.5b}]
                       [--device {cpu,cuda}] [-d]

KakaoBrain Korean(hangul) Generative Pre-Training Model

optional arguments:
  -h, --help            show this help message and exit
  --model MODEL         huggingface repo (default:kakaobrain/kogpt)
  --revision {KoGPT6B-ryan1.5b}
  --device {cpu,cuda}   (default:cuda)
  -d, --debug
python -m kogpt
prompt> 인간처럼 생각하고, 행동하는 '지능'을 통해 인류가 이제까지 풀지 못했던
temperature(0.8)> 
max_length(128)> 64
인간처럼 생각하고, 행동하는 '지능'을 통해 인류가 이제까지 풀지 못했던 문제의 해답을 찾을 수 있을 것이다. 과학기술이 고도로 발달한 21세기를 살아갈 우리 아이들에게 가장 필요한 것은 사고력 훈련이다. 사고력 훈련을 통해, 세상

prompt>  
...

python

import torch
from transformers import AutoTokenizer, AutoModelForCausalLM 

tokenizer = AutoTokenizer.from_pretrained(
  'kakaobrain/kogpt', revision='KoGPT6B-ryan1.5b-float16',  # or float32 version: revision=KoGPT6B-ryan1.5b
  bos_token='[BOS]', eos_token='[EOS]', unk_token='[UNK]', pad_token='[PAD]', mask_token='[MASK]'
)
model = AutoModelForCausalLM.from_pretrained(
  'kakaobrain/kogpt', revision='KoGPT6B-ryan1.5b-float16',  # or float32 version: revision=KoGPT6B-ryan1.5b
  pad_token_id=tokenizer.eos_token_id,
  torch_dtype='auto', low_cpu_mem_usage=True
).to(device='cuda', non_blocking=True)
_ = model.eval()

prompt = '인간처럼 생각하고, 행동하는 \'지능\'을 통해 인류가 이제까지 풀지 못했던'
with torch.no_grad():
  tokens = tokenizer.encode(prompt, return_tensors='pt').to(device='cuda', non_blocking=True)
  gen_tokens = model.generate(tokens, do_sample=True, temperature=0.8, max_length=64)
  generated = tokenizer.batch_decode(gen_tokens)[0]
  
print(generated)  # print: 인간처럼 생각하고, 행동하는 '지능'을 통해 인류가 이제까지 풀지 못했던 문제의 해답을 찾을 수 있을 것이다. 과학기술이 고도로 발달한 21세기를 살아갈 우리 아이들에게 가장 필요한 것은 사고력 훈련이다. 사고력 훈련을 통해, 세상

Experiments

In-context Few-Shots

Models #params NSMC (Acc.) YNAT (F1) KLUE-STS (F1)
HyperCLOVA[1] 1.3B 83.9 58.7 60.9
HyperCLOVA[1] 6.9B 83.8 67.5 59.3
HyperCLOVA[1] 13.0B 87.9 67.9 60.0
HyperCLOVA[1] 39.0B 88.0 71.4 61.6
HyperCLOVA[1] 82.0B 88.2 72.7 65.1
Ours 6.0B 87.8 78.0 64.3

Finetuning / P-Tuning

We have been reported to have issues(#17) with our downstream evaluation.

The previously published performance evaluation table was deleted because it was difficult to see it as a fair comparison because the comparison target algorithm was different and the performance measurement method could not be confirmed.

You can refer to the above issue link for the existing performance evaluation table and troubleshooting results.

Limitations

KakaoBrain KoGPT was trained on raw data, a dataset known to contain profanity, lewd, political changed, and other harsh language. Therefore, KoGPT can generate socially unacceptable texts. As with all language models, It is difficult to predict in advance how KoGPT will response to particular prompts and offensive content without warning.

Primarily Korean: KoGPT is primarily trained on Korean texts, and is best for classifying, searching, summarizing or generating such texts. KoGPT by default perform worse on inputs that are different from the data distribution it is trained on, including non-Korean as well as specific dialects of Korean that are not well represented in the training data.

If abnormal or socially unacceptable text is generated during testing, please send a "prompt" and the "generated text" to [email protected].

카카오브레인 KoGPT는 AI커뮤니티를 위한 연구용으로 욕설, 음란, 정치적 내용 및 기타 거친 언어에 대한 처리를 하지 않은 원시 데이터로 학습하였습니다. 따라서 KoGPT는 사회적으로 용인되지 않은 텍스트를 생성할 수 있습니다. 다른 언어 모델과 마찬가지로 특정 프롬프트와 공격적인 콘텐츠에 어떠한 결과를 생성할지 사전에 파악하기 어렵습니다.

KoGPT는 주로 한국어 텍스트로 학습을 하였으며 이러한 텍스트를 분류, 검색, 요약 또는 생성하는데 가장 적합합니다. 기본적으로 KoGPT는 학습 데이터에 잘 나타나지 않는 방언뿐만아니라 한국어가 아닌 경우와 같이 학습 데이터에서 발견하기 어려운 입력에서 좋지 않은 성능을 보입니다.

본 KoGPT를 활용한 연구, 개발, 테스트 등에 있어 위의 부분을 꼭 유의하시기 바랍니다.
테스트중에 발생한 비정상적인 혹은 사회적으로 용인되지 않는 텍스트가 생성된 경우 [email protected]로 "prompt"와 "생성된 문장"을 함께 보내주시기 바랍니다.

Citation

If you apply this library or model to any project and research, please cite our code:

@misc{kakaobrain2021kogpt,
  title         = {KoGPT: KakaoBrain Korean(hangul) Generative Pre-trained Transformer},
  author        = {Ildoo Kim and Gunsoo Han and Jiyeon Ham and Woonhyuk Baek},
  year          = {2021},
  howpublished  = {\url{https://github.com/kakaobrain/kogpt}},
}

Contact

This is released as an open source in the hope that it will be helpful to many research institutes and startups for research purposes. We look forward to contacting us from various places who wish to cooperate with us.

[email protected]

License

The source code of KakaoBrain KoGPT are licensed under Apache 2.0 License.
The pretrained weights of KakaoBrain KoGPT are licensed under CC-BY-NC-ND 4.0 License License.

카카오브레인 KoGPT소스코드(source code)Apache 2.0 라이선스 하에 공개되어 있습니다.
카카오브레인 KoGPT사전학습된 가중치(pretrained weights)CC-BY-NC-ND 4.0 라이선스 라이선스 하에 공개되어 있습니다.
모델 및 코드, 사전학습된 가중치를 사용할 경우 라이선스 내용을 준수해 주십시오. 라이선스 전문은 Apache 2.0, LICENSE.cc-by-nc-nd-4.0 파일에서 확인하실 수 있습니다.

Obligation to use

While Open Source software may be free to use, that does not mean it is free of obligation. To determine whether your intended use of KoGPT is suitable for the Apache 2.0 (or CC-BY-NC-ND 4.0), please consider the license guide. If you violate the license, you may be subject to legal action such as prohibition of use or claim for damages depending on the use.

오픈소스 소프트웨어는 무료로 사용할 수 있지만 이것이 의무가 없다는 의미는 아닙니다. KoGPT의 사용에 앞서 라이선스 가이드를 살펴보고 예정한 사용이 Apache 2.0 (또는 CC-BY-NC-ND 4.0)를 준수하는지 여부를 먼저 확인하시기 바랍니다. 라이선스를 위반하는 경우, 내용에 따라 사용금지, 손해배상 청구 등의 법적 조치를 취할 수 있습니다.

References

[1] HyperCLOVA: Kim, Boseop, et al. "What changes can large-scale language models bring? intensive study on hyperclova: Billions-scale korean generative pretrained transformers." arXiv preprint arXiv:2109.04650 (2021).


Contribution

Disclaimer

The contribution section is not an official KakaoBrain product.

AK391's Web Demo on Huggingface Spaces

More Repositories

1

fast-autoaugment

Official Implementation of 'Fast AutoAugment' in PyTorch.
Python
1,587
star
2

nerf-factory

An awesome PyTorch NeRF library
Python
1,265
star
3

pororo

PORORO: Platform Of neuRal mOdels for natuRal language prOcessing
Python
1,252
star
4

coyo-dataset

COYO-700M: Large-scale Image-Text Pair Dataset
Python
1,062
star
5

torchgpipe

A GPipe implementation in PyTorch
Python
776
star
6

karlo

Python
679
star
7

rq-vae-transformer

The official implementation of Autoregressive Image Generation using Residual Quantization (CVPR '22)
Jupyter Notebook
669
star
8

mindall-e

PyTorch implementation of a 1.3B text-to-image generation model trained on 14 million image-text pairs
Python
630
star
9

word2word

Easy-to-use word-to-word translations for 3,564 language pairs.
Python
350
star
10

torchlars

A LARS implementation in PyTorch
Python
326
star
11

g2pm

A Neural Grapheme-to-Phoneme Conversion Package for Mandarin Chinese Based on a New Open Benchmark Dataset
Python
326
star
12

kor-nlu-datasets

KorNLI and KorSTS: New Benchmark Datasets for Korean Natural Language Understanding
283
star
13

trident

A performance library for machine learning applications.
Python
176
star
14

autoclint

A specially designed light version of Fast AutoAugment
Python
170
star
15

sparse-detr

PyTorch Implementation of Sparse DETR
Python
150
star
16

hotr

Official repository for HOTR: End-to-End Human-Object Interaction Detection with Transformers (CVPR'21, Oral Presentation)
Python
132
star
17

kortok

The code and models for "An Empirical Study of Tokenization Strategies for Various Korean NLP Tasks" (AACL-IJCNLP 2020)
Python
114
star
18

bassl

Python
113
star
19

scrl

PyTorch Implementation of Spatially Consistent Representation Learning(SCRL)
Python
108
star
20

flame

Official implementation of the paper "FLAME: Free-form Language-based Motion Synthesis & Editing"
Python
103
star
21

brain-agent

Brain Agent for Large-Scale and Multi-Task Agent Learning
Python
92
star
22

helo-word

Team Kakao&Brain's Grammatical Error Correction System for the ACL 2019 BEA Shared Task
Python
88
star
23

jejueo

Jejueo Datasets for Machine Translation and Speech Synthesis
Python
74
star
24

solvent

Python
66
star
25

noc

Jupyter Notebook
44
star
26

cxr-clip

Python
43
star
27

expgan

Python
41
star
28

autowu

Official repository for Automated Learning Rate Scheduler for Large-Batch Training (8th ICML Workshop on AutoML)
Python
39
star
29

nvs-adapter

Python
33
star
30

ginr-ipc

The official implementation of Generalizable Implicit Neural Representations with Instance Pattern Composers(CVPR’23 highlight).
Python
30
star
31

coyo-vit

ViT trained on COYO-Labeled-300M dataset
Python
28
star
32

irm-empirical-study

An Empirical Study of Invariant Risk Minimization
Python
28
star
33

coyo-align

ALIGN trained on COYO-dataset
Python
25
star
34

magvlt

The official implementation of MAGVLT: Masked Generative Vision-and-Language Transformer (CVPR'23)
Python
23
star
35

hqtransformer

Locally Hierarchical Auto-Regressive Modeling for Image Generation (HQ-Transformer)
Jupyter Notebook
21
star
36

CheXGPT

Python
18
star
37

learning-loss-for-tta

"Learning Loss for Test-Time Augmentation (NeurIPS 2020)"
Python
9
star
38

stg

Official implementation of Selective Token Generation (COLING'22)
Jupyter Notebook
8
star
39

leco

Official implementation of LECO (NeurIPS'22)
Python
6
star
40

bc-hyperopt-example

brain cloud hyperopt example (mnist)
Python
3
star