spa-eng-pos-tagging-v2
This model is a fine-tuned version of distilbert-base-multilingual-cased on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.4432
- Accuracy: 0.8418
- Precision: 0.8395
- Recall: 0.7600
- F1: 0.7676
- Hamming Loss: 0.1582
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: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 12
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Hamming Loss |
---|---|---|---|---|---|---|---|---|
1.4285 | 1.0 | 1744 | 1.2584 | 0.5671 | 0.6506 | 0.4372 | 0.4716 | 0.4329 |
1.1788 | 2.0 | 3488 | 1.0023 | 0.6388 | 0.6753 | 0.5323 | 0.5578 | 0.3612 |
0.9144 | 3.0 | 5232 | 0.7885 | 0.7093 | 0.7259 | 0.6091 | 0.6281 | 0.2907 |
0.78 | 4.0 | 6976 | 0.6970 | 0.7439 | 0.7517 | 0.6527 | 0.6673 | 0.2561 |
0.6565 | 5.0 | 8720 | 0.6072 | 0.7765 | 0.7792 | 0.6838 | 0.6952 | 0.2235 |
0.5845 | 6.0 | 10464 | 0.5438 | 0.7995 | 0.7974 | 0.7125 | 0.7221 | 0.2005 |
0.5158 | 7.0 | 12208 | 0.5127 | 0.8132 | 0.8180 | 0.7250 | 0.7362 | 0.1868 |
0.4697 | 8.0 | 13952 | 0.4939 | 0.8186 | 0.8188 | 0.7345 | 0.7438 | 0.1814 |
0.4251 | 9.0 | 15696 | 0.4712 | 0.8334 | 0.8349 | 0.7502 | 0.7591 | 0.1666 |
0.4039 | 10.0 | 17440 | 0.4564 | 0.8381 | 0.8382 | 0.7538 | 0.7629 | 0.1619 |
0.3826 | 11.0 | 19184 | 0.4479 | 0.8397 | 0.8399 | 0.7565 | 0.7656 | 0.1603 |
0.3691 | 12.0 | 20928 | 0.4432 | 0.8418 | 0.8395 | 0.7600 | 0.7676 | 0.1582 |
Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1+cu118
- Tokenizers 0.13.3
- Downloads last month
- 4
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social
visibility and check back later, or deploy to Inference Endpoints (dedicated)
instead.