<!-- 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. -->
mbart_cycle0_ko-ja
This model is a fine-tuned version of mbart-large-cc25 on an custom dataset. It achieves the following results on the evaluation set:
- Loss: 7.0107
- Bleu: 25.8676
- Gen Len: 20.5833
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: 5e-05
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 8
- total_eval_batch_size: 8
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 300
- num_epochs: 50
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len |
|---|---|---|---|---|---|
| No log | 3.57 | 50 | 12.5219 | 0.0216 | 443.0833 |
| No log | 7.14 | 100 | 9.2255 | 0.0315 | 1024.0 |
| No log | 10.71 | 150 | 6.4885 | 0.0151 | 779.0 |
| No log | 14.29 | 200 | 5.3925 | 0.928 | 101.5 |
| No log | 17.86 | 250 | 5.4016 | 13.1472 | 105.6667 |
| No log | 21.43 | 300 | 6.5062 | 11.5401 | 158.3333 |
| No log | 25.0 | 350 | 6.0911 | 20.6997 | 25.1667 |
| No log | 28.57 | 400 | 6.5541 | 18.9521 | 20.6667 |
| No log | 32.14 | 450 | 6.6978 | 21.2662 | 25.1667 |
| 6.3858 | 35.71 | 500 | 6.9643 | 10.1265 | 17.3333 |
| 6.3858 | 39.29 | 550 | 6.6467 | 25.8218 | 19.6667 |
| 6.3858 | 42.86 | 600 | 7.1260 | 13.6948 | 18.75 |
| 6.3858 | 46.43 | 650 | 7.0505 | 19.5121 | 21.0 |
| 6.3858 | 50.0 | 700 | 7.0107 | 25.8676 | 20.5833 |
Framework versions
- Transformers 4.33.1
- Pytorch 2.0.1+cu117
- Datasets 2.14.5
- Tokenizers 0.13.3
