File size: 2,602 Bytes
750c79e 5cccf6f 750c79e 5cccf6f e9a3236 ed9b5a9 f7cbb70 e9a3236 5cccf6f e9a3236 5cccf6f 4858f66 5cccf6f e9a3236 5cccf6f e9a3236 5cccf6f e9a3236 5cccf6f f7cbb70 5cccf6f f7cbb70 5cccf6f f7cbb70 750c79e e9a3236 750c79e e9a3236 750c79e e9a3236 265a525 e9a3236 265a525 e9a3236 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 |
---
language: ko
tags:
- summarization
- news
inference: false
model-index:
- name: KoBigBird-KoBart-News-Summarization
results: []
---
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->
# KoBigBird-KoBart-News-Summarization
This model is a fine-tuned version of [noahkim/KoBigBird-KoBart-News-Summarization](https://huggingface.co/noahkim/KoBigBird-KoBart-News-Summarization) on the [daekeun-ml/naver-news-summarization-ko](https://huggingface.co/datasets/daekeun-ml/naver-news-summarization-ko)
## Model description
<<20221110 Commit>>
<<KoBigBird-KoBart-News-Summarization ๋ชจ๋ธ ์ค๋ช
>>
๋ค์ค๋ฌธ์์์ฝ(Multi-Document-Summarization) Task๋ฅผ ์ํด์ KoBigBird ๋ชจ๋ธ์ Encoder-Decoder๋ชจ๋ธ์ ๋ง๋ค์ด์ ํ์ต์ ์งํํ์ต๋๋ค. KoBigBird๋ฅผ Decoder๋ก ์ฐ๋ ค๊ณ ํ์ผ๋ ์ค๋ฅ๊ฐ ์๊ฒจ์ ์์ฝ์ ํนํ๋ KoBART์ Decoder๋ฅผ ํ์ฉํด์ ๋ชจ๋ธ์ ์์ฑํ์ต๋๋ค.
ํ๋ก์ ํธ์ฉ์ผ๋ก ๋ด์ค ์์ฝ ๋ชจ๋ธ ํนํ๋ ๋ชจ๋ธ์ ๋ง๋ค๊ธฐ ์ํด ๊ธฐ์กด์ ๋ง๋ค์๋ KoBigBird-KoBart-News-Summarization ๋ชจ๋ธ์ ์ถ๊ฐ์ ์ผ๋ก daekeun-ml๋์ด ์ ๊ณตํด์ฃผ์ naver-news-summarization-ko ๋ฐ์ดํฐ์
์ผ๋ก ํ์ธํ๋ ํ์ต๋๋ค.
ํ์ฌ AI-HUB์์ ์ ๊ณตํ๋ ์์ฝ ๋ฐ์ดํฐ๋ฅผ ์ถ๊ฐ ํ์ต ์งํ ์์ ์
๋๋ค.
์ง์์ ์ผ๋ก ๋ฐ์ ์์ผ ์ข์ ์ฑ๋ฅ์ ๋ชจ๋ธ์ ๊ตฌํํ๊ฒ ์ต๋๋ค.
๊ฐ์ฌํฉ๋๋ค.
์คํํ๊ฒฝ
- Google Colab Pro
- CPU : Intel(R) Xeon(R) CPU @ 2.20GHz
- GPU : A100-SXM4-40GB
<pre><code>
# Python Code
from transformers import AutoTokenizer
from transformers import AutoModelForSeq2SeqLM
tokenizer = AutoTokenizer.from_pretrained("noahkim/KoT5_news_summarization")
model = AutoModelForSeq2SeqLM.from_pretrained("noahkim/KoT5_news_summarization")
</pre></code>
The following hyperparameters were used during training:
- learning_rate: 2e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 4
- mixed_precision_training: Native AMP
### Training results
| Training Loss | Epoch | Step | Validation Loss |
|:-------------:|:-----:|:----:|:---------------:|
| 4.0748 | 1.0 | 1388 | 4.3067 |
| 3.8457 | 2.0 | 2776 | 4.2039 |
| 3.7459 | 3.0 | 4164 | 4.1433 |
| 3.6773 | 4.0 | 5552 | 4.1236 |
### Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.6.1
- Tokenizers 0.13.2
|