|
--- |
|
language: |
|
- ko |
|
- en |
|
base_model: facebook/mbart-large-50-many-to-many-mmt |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- bleu |
|
model-index: |
|
- name: koen_mbartLarge_64p_run1 |
|
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. --> |
|
|
|
# koen_mbartLarge_64p_run1 |
|
|
|
This model is a fine-tuned version of [facebook/mbart-large-50-many-to-many-mmt](https://huggingface.co/facebook/mbart-large-50-many-to-many-mmt) on an unknown dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 1.0989 |
|
- Bleu: 33.8958 |
|
- Gen Len: 18.5033 |
|
|
|
## 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 |
|
- gradient_accumulation_steps: 2 |
|
- total_train_batch_size: 32 |
|
- 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: 10 |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Bleu | Gen Len | |
|
|:-------------:|:-----:|:-----:|:---------------:|:-------:|:-------:| |
|
| 1.5543 | 0.13 | 2500 | 1.5068 | 25.2368 | 18.5097 | |
|
| 1.4399 | 0.26 | 5000 | 1.3972 | 27.0554 | 18.5539 | |
|
| 1.3448 | 0.39 | 7500 | 1.3132 | 28.8579 | 18.6315 | |
|
| 1.3205 | 0.52 | 10000 | 1.2873 | 29.5611 | 18.7781 | |
|
| 1.2786 | 0.65 | 12500 | 1.2399 | 30.3042 | 18.5644 | |
|
| 1.2561 | 0.78 | 15000 | 1.2173 | 30.5801 | 19.0186 | |
|
| 1.2479 | 0.91 | 17500 | 1.2125 | 30.8896 | 18.7636 | |
|
| 1.1891 | 1.04 | 20000 | 1.1776 | 31.9834 | 18.7002 | |
|
| 1.1943 | 1.17 | 22500 | 1.1651 | 32.0205 | 18.7054 | |
|
| 1.1375 | 1.3 | 25000 | 1.1492 | 32.3658 | 18.6287 | |
|
| 1.1351 | 1.43 | 27500 | 1.1460 | 32.339 | 18.7655 | |
|
| 1.0859 | 1.56 | 30000 | 1.1623 | 31.5418 | 19.016 | |
|
| 1.0373 | 1.69 | 32500 | 1.1383 | 32.672 | 18.7224 | |
|
| 1.0824 | 1.82 | 35000 | 1.1232 | 33.2231 | 18.6697 | |
|
| 1.0242 | 1.95 | 37500 | 1.1313 | 32.813 | 18.2553 | |
|
| 1.0649 | 2.08 | 40000 | 1.1182 | 33.2021 | 18.7216 | |
|
| 1.054 | 2.21 | 42500 | 1.1329 | 33.0588 | 18.4992 | |
|
| 1.0143 | 2.34 | 45000 | 1.1187 | 33.2176 | 18.7156 | |
|
| 1.0037 | 2.47 | 47500 | 1.1162 | 33.3754 | 18.6443 | |
|
| 0.9928 | 2.61 | 50000 | 1.1306 | 33.0727 | 18.6361 | |
|
| 0.9497 | 2.74 | 52500 | 1.1170 | 33.227 | 18.7638 | |
|
| 1.0157 | 2.87 | 55000 | 1.1072 | 33.685 | 18.5847 | |
|
| 0.9876 | 3.0 | 57500 | 1.1035 | 33.6971 | 18.6873 | |
|
| 0.9665 | 3.13 | 60000 | 1.0989 | 33.8919 | 18.5258 | |
|
| 0.9197 | 3.26 | 62500 | 1.1060 | 33.7036 | 18.5407 | |
|
| 0.9427 | 3.39 | 65000 | 1.0995 | 33.7642 | 18.7 | |
|
| 0.8993 | 3.52 | 67500 | 1.1364 | 33.1757 | 18.646 | |
|
| 0.8957 | 3.65 | 70000 | 1.1251 | 33.0954 | 18.3129 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.34.1 |
|
- Pytorch 2.1.0+cu121 |
|
- Datasets 2.14.6 |
|
- Tokenizers 0.14.1 |
|
|
