distilCamemBERT_nli_on_legal_data
This model is a fine-tuned version of 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
- Downloads last month
- 7
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.
Model tree for hcene/distilCamemBERT_nli_on_legal_data
Base model
cmarkea/distilcamembert-base
Quantized
cmarkea/distilcamembert-base-nli