<!-- 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. -->
ainize-kobart-news-eb-finetuned-xsum
This model is a fine-tuned version of ainize/kobart-news on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2147
- Rouge1: 60.732
- Rouge2: 39.1933
- Rougel: 60.6507
- Rougelsum: 60.6712
- Gen Len: 19.3417
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: 4
- eval_batch_size: 4
- 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 |
---|---|---|---|---|---|---|---|---|
1.0649 | 1.0 | 749 | 0.5502 | 56.6571 | 36.5992 | 56.6185 | 56.6364 | 19.2929 |
0.7103 | 2.0 | 1498 | 0.3904 | 59.1212 | 38.3611 | 59.093 | 59.1191 | 19.31 |
0.4723 | 3.0 | 2247 | 0.2922 | 60.1133 | 38.7819 | 60.0439 | 60.0572 | 19.2659 |
0.3841 | 4.0 | 2996 | 0.2367 | 60.4405 | 39.0176 | 60.366 | 60.4057 | 19.3397 |
0.3091 | 5.0 | 3745 | 0.2147 | 60.732 | 39.1933 | 60.6507 | 60.6712 | 19.3417 |
Framework versions
- Transformers 4.19.2
- Pytorch 1.11.0+cu113
- Datasets 2.2.2
- Tokenizers 0.12.1