Update README.md
Browse files
README.md
CHANGED
|
@@ -44,11 +44,23 @@ It achieves the following results on the E3C test set following the TempEval-3 e
|
|
| 44 |
|
| 45 |
## Model description
|
| 46 |
|
| 47 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 48 |
|
| 49 |
## Intended uses & limitations
|
| 50 |
|
| 51 |
-
|
| 52 |
|
| 53 |
## Training and evaluation data
|
| 54 |
|
|
@@ -67,36 +79,6 @@ The following hyperparameters were used during training:
|
|
| 67 |
- lr_scheduler_type: linear
|
| 68 |
- num_epochs: 24
|
| 69 |
|
| 70 |
-
### Training results
|
| 71 |
-
|
| 72 |
-
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|
| 73 |
-
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
|
| 74 |
-
| 0.0433 | 1.0 | 12 | 0.0443 | 0.4948 | 0.5 | 0.4974 | 0.9800 |
|
| 75 |
-
| 0.0234 | 2.0 | 24 | 0.0221 | 0.4082 | 0.7257 | 0.5225 | 0.9732 |
|
| 76 |
-
| 0.0055 | 3.0 | 36 | 0.0159 | 0.4768 | 0.7847 | 0.5932 | 0.9797 |
|
| 77 |
-
| 0.0089 | 4.0 | 48 | 0.0153 | 0.5317 | 0.8160 | 0.6438 | 0.9813 |
|
| 78 |
-
| 0.0033 | 5.0 | 60 | 0.0131 | 0.7229 | 0.8333 | 0.7742 | 0.9896 |
|
| 79 |
-
| 0.008 | 6.0 | 72 | 0.0129 | 0.6649 | 0.8681 | 0.7530 | 0.9885 |
|
| 80 |
-
| 0.0063 | 7.0 | 84 | 0.0146 | 0.7523 | 0.8542 | 0.8 | 0.9904 |
|
| 81 |
-
| 0.0086 | 8.0 | 96 | 0.0150 | 0.7470 | 0.8715 | 0.8045 | 0.9906 |
|
| 82 |
-
| 0.0009 | 9.0 | 108 | 0.0139 | 0.7658 | 0.8854 | 0.8213 | 0.9910 |
|
| 83 |
-
| 0.0031 | 10.0 | 120 | 0.0159 | 0.8031 | 0.8924 | 0.8454 | 0.9919 |
|
| 84 |
-
| 0.0011 | 11.0 | 132 | 0.0158 | 0.7649 | 0.8924 | 0.8237 | 0.9909 |
|
| 85 |
-
| 0.0006 | 12.0 | 144 | 0.0153 | 0.7398 | 0.8785 | 0.8032 | 0.9902 |
|
| 86 |
-
| 0.0013 | 13.0 | 156 | 0.0157 | 0.7815 | 0.8819 | 0.8287 | 0.9910 |
|
| 87 |
-
| 0.0008 | 14.0 | 168 | 0.0154 | 0.7822 | 0.8854 | 0.8306 | 0.9908 |
|
| 88 |
-
| 0.0008 | 15.0 | 180 | 0.0164 | 0.7778 | 0.875 | 0.8235 | 0.9910 |
|
| 89 |
-
| 0.0007 | 16.0 | 192 | 0.0168 | 0.7864 | 0.8819 | 0.8314 | 0.9912 |
|
| 90 |
-
| 0.0018 | 17.0 | 204 | 0.0173 | 0.7870 | 0.8854 | 0.8333 | 0.9912 |
|
| 91 |
-
| 0.0006 | 18.0 | 216 | 0.0178 | 0.7730 | 0.875 | 0.8208 | 0.9914 |
|
| 92 |
-
| 0.0012 | 19.0 | 228 | 0.0171 | 0.8013 | 0.8819 | 0.8397 | 0.9916 |
|
| 93 |
-
| 0.0006 | 20.0 | 240 | 0.0181 | 0.8137 | 0.8646 | 0.8384 | 0.9916 |
|
| 94 |
-
| 0.0007 | 21.0 | 252 | 0.0186 | 0.8137 | 0.8646 | 0.8384 | 0.9918 |
|
| 95 |
-
| 0.0012 | 22.0 | 264 | 0.0188 | 0.8137 | 0.8646 | 0.8384 | 0.9919 |
|
| 96 |
-
| 0.0006 | 23.0 | 276 | 0.0178 | 0.8121 | 0.8854 | 0.8472 | 0.9919 |
|
| 97 |
-
| 0.0009 | 24.0 | 288 | 0.0177 | 0.8121 | 0.8854 | 0.8472 | 0.9919 |
|
| 98 |
-
|
| 99 |
-
|
| 100 |
### Framework versions
|
| 101 |
|
| 102 |
- Transformers 4.24.0
|
|
|
|
| 44 |
|
| 45 |
## Model description
|
| 46 |
|
| 47 |
+
**Developed by**: Alejandro Sánchez de Castro, Juan Martínez Romo, Lourdes Araujo
|
| 48 |
+
|
| 49 |
+
This model is the result of the paper "RoBERTime: A novel model for the detection of temporal expressions in Spanish"
|
| 50 |
+
|
| 51 |
+
**Cite as**:
|
| 52 |
+
|
| 53 |
+
@article{sanchez2023robertime,
|
| 54 |
+
title={RoBERTime: A novel model for the detection of temporal expressions in Spanish},
|
| 55 |
+
author={S{\'a}nchez-de-Castro-Fern{\'a}ndez, Alejandro and Araujo Serna, Lourdes and Mart{\'\i}nez Romo, Juan},
|
| 56 |
+
year={2023},
|
| 57 |
+
publisher={Sociedad Espa{\~n}ola para el Procesamiento del Lenguaje Natural}
|
| 58 |
+
}
|
| 59 |
+
|
| 60 |
|
| 61 |
## Intended uses & limitations
|
| 62 |
|
| 63 |
+
This model is prepared for the detection of temporal expressions extension in Spanish. It may work in other languages due to RoBERTa multilingual capabilities. This model does not normalize the expression value. This is considered to be a separate task.
|
| 64 |
|
| 65 |
## Training and evaluation data
|
| 66 |
|
|
|
|
| 79 |
- lr_scheduler_type: linear
|
| 80 |
- num_epochs: 24
|
| 81 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 82 |
### Framework versions
|
| 83 |
|
| 84 |
- Transformers 4.24.0
|