File size: 6,987 Bytes
26a61fd |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 |
---
license: mit
base_model: cmarkea/distilcamembert-base-nli
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: distilCamemBERT_nli_on_legal_data
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. -->
# distilCamemBERT_nli_on_legal_data
This model is a fine-tuned version of [cmarkea/distilcamembert-base-nli](https://huggingface.co/cmarkea/distilcamembert-base-nli) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7470
- Accuracy: 0.7384
- Precision: 0.7415
- Recall: 0.7395
- F1: 0.7378
- Ratio: 0.3297
## 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: 2e-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
- lr_scheduler_warmup_ratio: 0.1
- lr_scheduler_warmup_steps: 4
- num_epochs: 10
- label_smoothing_factor: 0.1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Ratio |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:|
| 1.3001 | 0.14 | 10 | 0.8533 | 0.6452 | 0.6683 | 0.6459 | 0.6525 | 0.2903 |
| 0.9775 | 0.27 | 20 | 0.7170 | 0.6918 | 0.7191 | 0.6944 | 0.6639 | 0.3513 |
| 0.8096 | 0.41 | 30 | 0.6432 | 0.7204 | 0.7222 | 0.7218 | 0.7213 | 0.3297 |
| 0.9883 | 0.55 | 40 | 0.6914 | 0.7204 | 0.7382 | 0.7228 | 0.7064 | 0.3297 |
| 0.9221 | 0.68 | 50 | 0.6597 | 0.7563 | 0.7582 | 0.7574 | 0.7570 | 0.3297 |
| 0.8253 | 0.82 | 60 | 0.6639 | 0.7563 | 0.7802 | 0.7568 | 0.7440 | 0.3297 |
| 0.9096 | 0.96 | 70 | 0.6612 | 0.7384 | 0.7484 | 0.7392 | 0.7319 | 0.3297 |
| 0.8466 | 1.1 | 80 | 0.6884 | 0.7168 | 0.7534 | 0.7195 | 0.6884 | 0.3297 |
| 0.8457 | 1.23 | 90 | 0.6884 | 0.7563 | 0.7898 | 0.7567 | 0.7397 | 0.3297 |
| 0.836 | 1.37 | 100 | 0.6409 | 0.7563 | 0.7711 | 0.7569 | 0.7488 | 0.3297 |
| 0.7878 | 1.51 | 110 | 0.6876 | 0.7276 | 0.7324 | 0.7286 | 0.7249 | 0.3297 |
| 0.8107 | 1.64 | 120 | 0.6720 | 0.7419 | 0.7518 | 0.7427 | 0.7361 | 0.3297 |
| 0.782 | 1.78 | 130 | 0.7397 | 0.7419 | 0.7518 | 0.7427 | 0.7361 | 0.3297 |
| 0.7728 | 1.92 | 140 | 0.6921 | 0.7455 | 0.7551 | 0.7463 | 0.7402 | 0.3297 |
| 0.7704 | 2.05 | 150 | 0.6882 | 0.7419 | 0.7518 | 0.7427 | 0.7361 | 0.3297 |
| 0.7593 | 2.19 | 160 | 0.7163 | 0.7348 | 0.7487 | 0.7355 | 0.7304 | 0.3226 |
| 0.7995 | 2.33 | 170 | 0.6639 | 0.7419 | 0.7518 | 0.7427 | 0.7361 | 0.3297 |
| 0.7657 | 2.47 | 180 | 0.6906 | 0.7885 | 0.8418 | 0.7887 | 0.7720 | 0.3297 |
| 0.7758 | 2.6 | 190 | 0.7577 | 0.7133 | 0.7154 | 0.7146 | 0.7134 | 0.3297 |
| 0.8269 | 2.74 | 200 | 0.8168 | 0.5591 | 0.6324 | 0.5596 | 0.5804 | 0.2151 |
| 0.7721 | 2.88 | 210 | 0.6721 | 0.7706 | 0.7924 | 0.7711 | 0.7615 | 0.3297 |
| 0.7098 | 3.01 | 220 | 0.6917 | 0.7133 | 0.7157 | 0.7145 | 0.7129 | 0.3297 |
| 0.7683 | 3.15 | 230 | 0.7175 | 0.7168 | 0.7192 | 0.7181 | 0.7168 | 0.3297 |
| 0.6907 | 3.29 | 240 | 0.7298 | 0.7491 | 0.7598 | 0.7499 | 0.7434 | 0.3297 |
| 0.7013 | 3.42 | 250 | 0.7363 | 0.7634 | 0.7794 | 0.7641 | 0.7562 | 0.3297 |
| 0.7852 | 3.56 | 260 | 0.7616 | 0.7025 | 0.7181 | 0.7034 | 0.7052 | 0.3082 |
| 0.7375 | 3.7 | 270 | 0.7247 | 0.7849 | 0.8127 | 0.7853 | 0.7753 | 0.3297 |
| 0.7242 | 3.84 | 280 | 0.7199 | 0.7921 | 0.8182 | 0.7925 | 0.7838 | 0.3297 |
| 0.7135 | 3.97 | 290 | 0.7190 | 0.7742 | 0.7907 | 0.7748 | 0.7677 | 0.3297 |
| 0.7501 | 4.11 | 300 | 0.7104 | 0.7921 | 0.8292 | 0.7924 | 0.7805 | 0.3297 |
| 0.663 | 4.25 | 310 | 0.7579 | 0.7921 | 0.8643 | 0.7922 | 0.7720 | 0.3297 |
| 0.7316 | 4.38 | 320 | 0.7671 | 0.7312 | 0.7347 | 0.7323 | 0.7301 | 0.3297 |
| 0.7045 | 4.52 | 330 | 0.7673 | 0.7204 | 0.7227 | 0.7217 | 0.7206 | 0.3297 |
| 0.7316 | 4.66 | 340 | 0.7421 | 0.7849 | 0.8096 | 0.7854 | 0.7764 | 0.3297 |
| 0.7667 | 4.79 | 350 | 0.7269 | 0.7527 | 0.7576 | 0.7536 | 0.7512 | 0.3297 |
| 0.7109 | 4.93 | 360 | 0.7305 | 0.7742 | 0.7907 | 0.7748 | 0.7677 | 0.3297 |
| 0.7677 | 5.07 | 370 | 0.7805 | 0.7885 | 0.8226 | 0.7888 | 0.7773 | 0.3297 |
| 0.6988 | 5.21 | 380 | 0.7531 | 0.7921 | 0.8182 | 0.7925 | 0.7838 | 0.3297 |
| 0.7119 | 5.34 | 390 | 0.7396 | 0.7670 | 0.7773 | 0.7677 | 0.7630 | 0.3297 |
| 0.6535 | 5.48 | 400 | 0.7259 | 0.7634 | 0.7714 | 0.7642 | 0.7604 | 0.3297 |
| 0.6732 | 5.62 | 410 | 0.7301 | 0.7921 | 0.8292 | 0.7924 | 0.7805 | 0.3297 |
| 0.7243 | 5.75 | 420 | 0.7094 | 0.7849 | 0.8127 | 0.7853 | 0.7753 | 0.3297 |
| 0.7367 | 5.89 | 430 | 0.7266 | 0.7670 | 0.7759 | 0.7678 | 0.7637 | 0.3297 |
| 0.7464 | 6.03 | 440 | 0.7929 | 0.7957 | 0.8277 | 0.7960 | 0.7860 | 0.3297 |
| 0.6836 | 6.16 | 450 | 0.7844 | 0.7957 | 0.8209 | 0.7961 | 0.7880 | 0.3297 |
| 0.6901 | 6.3 | 460 | 0.7724 | 0.7706 | 0.7837 | 0.7713 | 0.7654 | 0.3297 |
| 0.6776 | 6.44 | 470 | 0.7513 | 0.7670 | 0.7806 | 0.7677 | 0.7613 | 0.3297 |
| 0.6388 | 6.58 | 480 | 0.7491 | 0.7384 | 0.7410 | 0.7395 | 0.7383 | 0.3297 |
| 0.7258 | 6.71 | 490 | 0.7361 | 0.7599 | 0.7695 | 0.7606 | 0.7557 | 0.3297 |
| 0.7458 | 6.85 | 500 | 0.7777 | 0.7993 | 0.8483 | 0.7994 | 0.7856 | 0.3297 |
| 0.7937 | 6.99 | 510 | 0.7797 | 0.7921 | 0.8336 | 0.7923 | 0.7793 | 0.3297 |
| 0.6984 | 7.12 | 520 | 0.7597 | 0.7634 | 0.7714 | 0.7642 | 0.7604 | 0.3297 |
| 0.7206 | 7.26 | 530 | 0.7470 | 0.7384 | 0.7415 | 0.7395 | 0.7378 | 0.3297 |
### Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.17.1
- Tokenizers 0.15.2
|