<!-- 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. -->
nmt-mpst-id-en-lr_1e-05-ep_20-seq_128_bs-16
This model is a fine-tuned version of t5-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 2.6902
- Bleu: 0.039
- Meteor: 0.144
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | Bleu | Meteor |
|---|---|---|---|---|---|
| No log | 1.0 | 404 | 3.1126 | 0.0197 | 0.0888 |
| 3.6037 | 2.0 | 808 | 2.9899 | 0.0254 | 0.108 |
| 3.2835 | 3.0 | 1212 | 2.9337 | 0.0275 | 0.1129 |
| 3.1798 | 4.0 | 1616 | 2.8926 | 0.0284 | 0.1152 |
| 3.1361 | 5.0 | 2020 | 2.8638 | 0.0295 | 0.1196 |
| 3.1361 | 6.0 | 2424 | 2.8362 | 0.0305 | 0.1222 |
| 3.0848 | 7.0 | 2828 | 2.8137 | 0.0321 | 0.1266 |
| 3.0439 | 8.0 | 3232 | 2.7928 | 0.0327 | 0.1284 |
| 3.025 | 9.0 | 3636 | 2.7754 | 0.0337 | 0.1311 |
| 2.9891 | 10.0 | 4040 | 2.7604 | 0.0348 | 0.134 |
| 2.9891 | 11.0 | 4444 | 2.7469 | 0.0354 | 0.136 |
| 2.9706 | 12.0 | 4848 | 2.7343 | 0.036 | 0.1372 |
| 2.9537 | 13.0 | 5252 | 2.7250 | 0.0365 | 0.1387 |
| 2.9471 | 14.0 | 5656 | 2.7152 | 0.0375 | 0.1408 |
| 2.9274 | 15.0 | 6060 | 2.7081 | 0.038 | 0.142 |
| 2.9274 | 16.0 | 6464 | 2.7021 | 0.0384 | 0.143 |
| 2.9147 | 17.0 | 6868 | 2.6966 | 0.0387 | 0.1433 |
| 2.9093 | 18.0 | 7272 | 2.6934 | 0.0389 | 0.1438 |
| 2.9082 | 19.0 | 7676 | 2.6906 | 0.039 | 0.1437 |
| 2.8945 | 20.0 | 8080 | 2.6902 | 0.039 | 0.144 |
Framework versions
- Transformers 4.24.0
- Pytorch 1.12.1+cu113
- Datasets 2.7.0
- Tokenizers 0.13.2
