File size: 6,961 Bytes
a278bca |
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 |
---
base_model: mtheo/camembert-base-xnli
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
- f1
model-index:
- name: legal-data-mDEBERTa-V3-base-mnli-xnli
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. -->
# legal-data-mDEBERTa-V3-base-mnli-xnli
This model is a fine-tuned version of [mtheo/camembert-base-xnli](https://huggingface.co/mtheo/camembert-base-xnli) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.7972
- Accuracy: 0.7706
- Precision: 0.7891
- Recall: 0.7711
- F1: 0.7697
- Ratio: 0.3154
## 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: 20
- eval_batch_size: 20
- 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: 15
- label_smoothing_factor: 0.1
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Precision | Recall | F1 | Ratio |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:---------:|:------:|:------:|:------:|
| 0.9937 | 0.17 | 10 | 0.8509 | 0.6093 | 0.7022 | 0.6086 | 0.6168 | 0.1935 |
| 0.8433 | 0.34 | 20 | 0.7131 | 0.6667 | 0.6703 | 0.6687 | 0.6666 | 0.3262 |
| 0.8683 | 0.52 | 30 | 0.7101 | 0.7312 | 0.7350 | 0.7324 | 0.7324 | 0.3262 |
| 0.8536 | 0.69 | 40 | 0.7542 | 0.6989 | 0.7152 | 0.6999 | 0.7053 | 0.3011 |
| 0.7964 | 0.86 | 50 | 0.7249 | 0.7670 | 0.7773 | 0.7677 | 0.7630 | 0.3297 |
| 0.7651 | 1.03 | 60 | 0.8253 | 0.7455 | 0.7517 | 0.7465 | 0.7450 | 0.3262 |
| 0.7658 | 1.21 | 70 | 0.8282 | 0.6953 | 0.7366 | 0.6956 | 0.7030 | 0.2688 |
| 0.7297 | 1.38 | 80 | 0.8694 | 0.7634 | 0.7771 | 0.7641 | 0.7612 | 0.3226 |
| 0.7726 | 1.55 | 90 | 0.7898 | 0.7097 | 0.7112 | 0.7112 | 0.7111 | 0.3297 |
| 0.7107 | 1.72 | 100 | 0.8279 | 0.7599 | 0.7642 | 0.7608 | 0.7589 | 0.3297 |
| 0.7204 | 1.9 | 110 | 0.9353 | 0.7240 | 0.7728 | 0.7240 | 0.7279 | 0.2724 |
| 0.7241 | 2.07 | 120 | 0.7903 | 0.7527 | 0.7818 | 0.7530 | 0.7531 | 0.3011 |
| 0.6683 | 2.24 | 130 | 0.8139 | 0.7384 | 0.7931 | 0.7382 | 0.7397 | 0.2760 |
| 0.6832 | 2.41 | 140 | 0.8339 | 0.7993 | 0.8268 | 0.7996 | 0.7913 | 0.3297 |
| 0.7397 | 2.59 | 150 | 0.8309 | 0.7849 | 0.7997 | 0.7855 | 0.7801 | 0.3297 |
| 0.6945 | 2.76 | 160 | 0.7860 | 0.7240 | 0.7259 | 0.7253 | 0.7247 | 0.3297 |
| 0.7067 | 2.93 | 170 | 0.7046 | 0.7957 | 0.8152 | 0.7962 | 0.7898 | 0.3297 |
| 0.6759 | 3.1 | 180 | 0.7465 | 0.7634 | 0.7703 | 0.7643 | 0.7611 | 0.3297 |
| 0.6673 | 3.28 | 190 | 0.8461 | 0.8029 | 0.8234 | 0.8033 | 0.7972 | 0.3297 |
| 0.6748 | 3.45 | 200 | 0.8701 | 0.8065 | 0.8354 | 0.8068 | 0.7987 | 0.3297 |
| 0.7638 | 3.62 | 210 | 0.7501 | 0.8136 | 0.8521 | 0.8138 | 0.8043 | 0.3297 |
| 0.6426 | 3.79 | 220 | 0.7165 | 0.8100 | 0.8379 | 0.8103 | 0.8029 | 0.3297 |
| 0.6569 | 3.97 | 230 | 0.7244 | 0.8100 | 0.8415 | 0.8103 | 0.8020 | 0.3297 |
| 0.7068 | 4.14 | 240 | 0.7448 | 0.8100 | 0.8499 | 0.8102 | 0.8000 | 0.3297 |
| 0.6544 | 4.31 | 250 | 0.8241 | 0.8065 | 0.8476 | 0.8066 | 0.7957 | 0.3297 |
| 0.6261 | 4.48 | 260 | 0.8409 | 0.7634 | 0.7692 | 0.7643 | 0.7617 | 0.3297 |
| 0.6428 | 4.66 | 270 | 0.7887 | 0.7491 | 0.7528 | 0.7501 | 0.7484 | 0.3297 |
| 0.657 | 4.83 | 280 | 0.7534 | 0.7706 | 0.7791 | 0.7714 | 0.7677 | 0.3297 |
| 0.6895 | 5.0 | 290 | 0.7863 | 0.8100 | 0.8379 | 0.8103 | 0.8029 | 0.3297 |
| 0.6422 | 5.17 | 300 | 0.7908 | 0.8136 | 0.8439 | 0.8139 | 0.8062 | 0.3297 |
| 0.5933 | 5.34 | 310 | 0.8330 | 0.7885 | 0.8031 | 0.7891 | 0.7869 | 0.3226 |
| 0.5863 | 5.52 | 320 | 0.8494 | 0.7527 | 0.7598 | 0.7536 | 0.7535 | 0.3226 |
| 0.6787 | 5.69 | 330 | 0.7748 | 0.7742 | 0.7823 | 0.7750 | 0.7716 | 0.3297 |
| 0.6761 | 5.86 | 340 | 0.7256 | 0.7814 | 0.7929 | 0.7820 | 0.7775 | 0.3297 |
| 0.6974 | 6.03 | 350 | 0.7711 | 0.8029 | 0.8208 | 0.8033 | 0.7980 | 0.3297 |
| 0.6083 | 6.21 | 360 | 0.8435 | 0.7993 | 0.8191 | 0.7997 | 0.7951 | 0.3262 |
| 0.5897 | 6.38 | 370 | 0.8773 | 0.7849 | 0.8124 | 0.7852 | 0.7831 | 0.3118 |
| 0.6076 | 6.55 | 380 | 0.8255 | 0.7634 | 0.7683 | 0.7644 | 0.7623 | 0.3297 |
| 0.6709 | 6.72 | 390 | 0.7865 | 0.7527 | 0.7551 | 0.7538 | 0.7530 | 0.3297 |
| 0.7063 | 6.9 | 400 | 0.7898 | 0.8029 | 0.8234 | 0.8033 | 0.7972 | 0.3297 |
| 0.6804 | 7.07 | 410 | 0.7804 | 0.7921 | 0.8152 | 0.7925 | 0.7848 | 0.3297 |
| 0.6227 | 7.24 | 420 | 0.7515 | 0.7706 | 0.7778 | 0.7714 | 0.7683 | 0.3297 |
| 0.6482 | 7.41 | 430 | 0.7758 | 0.7670 | 0.7725 | 0.7679 | 0.7656 | 0.3297 |
| 0.6072 | 7.59 | 440 | 0.8077 | 0.7706 | 0.7792 | 0.7714 | 0.7693 | 0.3262 |
| 0.5695 | 7.76 | 450 | 0.8460 | 0.7921 | 0.8068 | 0.7926 | 0.7892 | 0.3262 |
| 0.6097 | 7.93 | 460 | 0.7856 | 0.7993 | 0.8191 | 0.7997 | 0.7951 | 0.3262 |
| 0.5591 | 8.1 | 470 | 0.7812 | 0.8136 | 0.8447 | 0.8138 | 0.8074 | 0.3262 |
| 0.5573 | 8.28 | 480 | 0.7249 | 0.7849 | 0.7930 | 0.7857 | 0.7828 | 0.3297 |
| 0.6128 | 8.45 | 490 | 0.7245 | 0.7921 | 0.8006 | 0.7928 | 0.7901 | 0.3297 |
| 0.6072 | 8.62 | 500 | 0.7732 | 0.7885 | 0.8038 | 0.7891 | 0.7852 | 0.3262 |
| 0.6276 | 8.79 | 510 | 0.8017 | 0.7885 | 0.8038 | 0.7891 | 0.7852 | 0.3262 |
| 0.647 | 8.97 | 520 | 0.7998 | 0.7885 | 0.8038 | 0.7891 | 0.7852 | 0.3262 |
| 0.5924 | 9.14 | 530 | 0.7972 | 0.7706 | 0.7891 | 0.7711 | 0.7697 | 0.3154 |
### Framework versions
- Transformers 4.39.0.dev0
- Pytorch 2.1.0+cu121
- Datasets 2.18.0
- Tokenizers 0.15.2
|