update model card README.md

README.md

@@ -1,27 +1,67 @@
 ---
-language: ko
-inference: false
-widget:
-- text: 'hello this is an example'
 tags:
-- summarization
-- bigbird
-- EncoderDecoderModel
-
+- generated_from_trainer
+metrics:
+- rouge
+model-index:
+- name: kobigbird-finetuned-Encoder-Decoder
+  results: []
 ---
-- This model is a [kobigbird-bert-base](https://huggingface.co/monologg/kobigbird-bert-base) finetuned on the [naver-news-summarization-ko](https://huggingface.co/datasets/daekeun-ml/naver-news-summarization-ko)
 
-<<20220916  Commit>>
+<!-- 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-finetuned-Encoder-Decoder
+
+This model was trained from scratch on an unknown dataset.
+It achieves the following results on the evaluation set:
+- Loss: 4.4011
+- Rouge1: 12.1436
+- Rouge2: 2.2747
+- Rougel: 11.7428
+- Rougelsum: 11.7408
+- Gen Len: 20.0
+
+## Model description
+
+More information needed
+
+## Intended uses & limitations
+
+More information needed
+
+## Training and evaluation data
+
+More information needed
+
+## Training procedure
+
+### Training hyperparameters
+
+The following hyperparameters were used during training:
+- learning_rate: 2e-05
+- train_batch_size: 1
+- eval_batch_size: 1
+- seed: 42
+- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
+- lr_scheduler_type: linear
+- num_epochs: 5
+- mixed_precision_training: Native AMP
+
+### Training results
+
+| Training Loss | Epoch | Step   | Validation Loss | Rouge1  | Rouge2 | Rougel  | Rougelsum | Gen Len |
+|:-------------:|:-----:|:------:|:---------------:|:-------:|:------:|:-------:|:---------:|:-------:|
+| 5.8126        | 1.0   | 22194  | 5.7258          | 8.857   | 0.2266 | 8.6768  | 8.703     | 20.0    |
+| 5.1675        | 2.0   | 44388  | 5.0540          | 7.0446  | 0.5937 | 6.8255  | 6.8637    | 20.0    |
+| 4.5552        | 3.0   | 66582  | 4.6871          | 10.3238 | 1.1363 | 9.9598  | 9.9394    | 20.0    |
+| 4.3369        | 4.0   | 88776  | 4.4789          | 11.0189 | 1.6226 | 10.7301 | 10.6951   | 20.0    |
+| 4.0251        | 5.0   | 110970 | 4.4011          | 12.1436 | 2.2747 | 11.7428 | 11.7408   | 20.0    |
 
-긴 문장의 요약 모델 특화된 모델을 만들기 위해 KoBigBird 모델을 Encoder Decoder 형태로 변환한 모델입니다.
-monologg님이 만들어 놓으신 KoBigBird 모델은 bert 기반으로 만들어져 있기 때문에 Q&A 등 자연어 처리 Task에 좋은 성능을 보입니다.
-저는 요약(Summarization) Task를 위해서 Encoder-Decoder 형태의 모양으로 바꿨습니다.
-추후 지속적인 업데이트로 별도의 불편함 없이 사용할 수 있도록 바꾸겠습니다.
 
-<pre><code>
-# Python Code
-from transformers import AutoTokenizer, AutoModelForSeq2SeqLM
+### Framework versions
 
-tokenizer = AutoTokenizer.from_pretrained("noahkim/kobigbird-finetuned-Encoder-Decoder")
-model = AutoModelForSeq2SeqLM.from_pretrained("noahkim/kobigbird-finetuned-Encoder-Decoder")
-</pre></code>
+- Transformers 4.22.0
+- Pytorch 1.12.1+cu113
+- Datasets 2.4.0
+- Tokenizers 0.12.1