spa-eng-pos-tagging-v4
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: 0.3366
- Accuracy: 0.9071
- Precision: 0.9038
- Recall: 0.8314
- F1: 0.8361
- Hamming Loss: 0.0929
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: 3e-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 |
---|---|---|---|---|---|---|---|---|
1.1797 | 1.0 | 1744 | 0.9547 | 0.6615 | 0.6780 | 0.5703 | 0.5810 | 0.3385 |
0.7267 | 2.0 | 3488 | 0.5924 | 0.7785 | 0.7766 | 0.6921 | 0.7005 | 0.2215 |
0.5048 | 3.0 | 5232 | 0.4816 | 0.8272 | 0.8107 | 0.7596 | 0.7526 | 0.1728 |
0.4095 | 4.0 | 6976 | 0.4585 | 0.8331 | 0.8289 | 0.7549 | 0.7587 | 0.1669 |
0.3369 | 5.0 | 8720 | 0.3830 | 0.8648 | 0.8621 | 0.7904 | 0.7940 | 0.1352 |
0.2888 | 6.0 | 10464 | 0.3506 | 0.8793 | 0.8715 | 0.8074 | 0.8077 | 0.1207 |
0.2397 | 7.0 | 12208 | 0.3485 | 0.8848 | 0.8845 | 0.8077 | 0.8143 | 0.1152 |
0.2093 | 8.0 | 13952 | 0.3523 | 0.8891 | 0.8864 | 0.8156 | 0.8190 | 0.1109 |
0.1723 | 9.0 | 15696 | 0.3538 | 0.8912 | 0.8931 | 0.8143 | 0.8217 | 0.1088 |
0.1558 | 10.0 | 17440 | 0.3436 | 0.8997 | 0.8958 | 0.8252 | 0.8290 | 0.1003 |
0.1344 | 11.0 | 19184 | 0.3373 | 0.9053 | 0.9013 | 0.8302 | 0.8343 | 0.0947 |
0.1134 | 12.0 | 20928 | 0.3366 | 0.9071 | 0.9038 | 0.8314 | 0.8361 | 0.0929 |
Framework versions
- Transformers 4.32.0
- Pytorch 2.0.1+cu118
- Tokenizers 0.13.3
- Downloads last month
- 6
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.