---
language:
- ko
- ja
base_model: ./reduced_model
tags:
- generated_from_trainer
metrics:
- bleu
model-index:
- name: tst-translation-output
  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. -->

# tst-translation-output

This model is a fine-tuned version of [mbart-large-cc25](https://huggingface.co/facebook/mbart-large-cc25) on an custom dataset.
It achieves the following results on the evaluation set:
- Loss: 4.3552
- Bleu: 19.2576
- Gen Len: 17.7448

## 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: 4
- eval_batch_size: 4
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 16
- total_eval_batch_size: 16
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 50

### Training results

| Training Loss | Epoch | Step | Validation Loss | Bleu    | Gen Len |
|:-------------:|:-----:|:----:|:---------------:|:-------:|:-------:|
| 2.7876        | 10.31 | 1000 | 4.2606          | 15.7275 | 15.5938 |
| 0.1404        | 20.62 | 2000 | 4.2496          | 16.6706 | 17.4375 |
| 0.0398        | 30.93 | 3000 | 4.3486          | 19.2786 | 17.8385 |
| 0.0107        | 41.24 | 4000 | 4.3411          | 21.5085 | 17.2917 |


### Framework versions

- Transformers 4.32.1
- Pytorch 2.0.1+cu117
- Datasets 2.14.4
- Tokenizers 0.13.3