|
--- |
|
license: mit |
|
tags: |
|
- generated_from_trainer |
|
metrics: |
|
- precision |
|
- recall |
|
- f1 |
|
- accuracy |
|
model-index: |
|
- name: pos_final_xlm_en |
|
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. --> |
|
|
|
# pos_final_xlm_en |
|
|
|
This model is a fine-tuned version of [xlm-roberta-base](https://huggingface.co/xlm-roberta-base) on the None dataset. |
|
It achieves the following results on the evaluation set: |
|
- Loss: 0.0719 |
|
- Precision: 0.9686 |
|
- Recall: 0.9705 |
|
- F1: 0.9695 |
|
- Accuracy: 0.9790 |
|
|
|
## 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: 5e-05 |
|
- train_batch_size: 256 |
|
- eval_batch_size: 256 |
|
- seed: 42 |
|
- gradient_accumulation_steps: 4 |
|
- total_train_batch_size: 1024 |
|
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08 |
|
- lr_scheduler_type: linear |
|
- lr_scheduler_warmup_steps: 500 |
|
- num_epochs: 40.0 |
|
- mixed_precision_training: Native AMP |
|
|
|
### Training results |
|
|
|
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy | |
|
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:| |
|
| No log | 0.99 | 60 | 3.0062 | 0.2412 | 0.1720 | 0.2008 | 0.3036 | |
|
| No log | 1.99 | 120 | 0.5353 | 0.8699 | 0.8553 | 0.8625 | 0.8970 | |
|
| No log | 2.99 | 180 | 0.1312 | 0.9578 | 0.9553 | 0.9566 | 0.9691 | |
|
| No log | 3.99 | 240 | 0.0981 | 0.9621 | 0.9628 | 0.9625 | 0.9737 | |
|
| No log | 4.99 | 300 | 0.0853 | 0.9652 | 0.9659 | 0.9655 | 0.9760 | |
|
| No log | 5.99 | 360 | 0.0788 | 0.9656 | 0.9676 | 0.9666 | 0.9769 | |
|
| No log | 6.99 | 420 | 0.0745 | 0.9664 | 0.9689 | 0.9677 | 0.9775 | |
|
| No log | 7.99 | 480 | 0.0718 | 0.9675 | 0.9689 | 0.9682 | 0.9780 | |
|
| 0.7956 | 8.99 | 540 | 0.0707 | 0.9679 | 0.9683 | 0.9681 | 0.9779 | |
|
| 0.7956 | 9.99 | 600 | 0.0686 | 0.9682 | 0.9698 | 0.9690 | 0.9786 | |
|
| 0.7956 | 10.99 | 660 | 0.0686 | 0.9689 | 0.9694 | 0.9692 | 0.9787 | |
|
| 0.7956 | 11.99 | 720 | 0.0680 | 0.9679 | 0.9707 | 0.9693 | 0.9787 | |
|
| 0.7956 | 12.99 | 780 | 0.0685 | 0.9683 | 0.9706 | 0.9694 | 0.9789 | |
|
| 0.7956 | 13.99 | 840 | 0.0695 | 0.9689 | 0.9700 | 0.9694 | 0.9788 | |
|
| 0.7956 | 14.99 | 900 | 0.0703 | 0.9682 | 0.9699 | 0.9690 | 0.9786 | |
|
| 0.7956 | 15.99 | 960 | 0.0719 | 0.9686 | 0.9705 | 0.9695 | 0.9790 | |
|
| 0.051 | 16.99 | 1020 | 0.0735 | 0.9687 | 0.9701 | 0.9694 | 0.9788 | |
|
| 0.051 | 17.99 | 1080 | 0.0747 | 0.9684 | 0.9701 | 0.9692 | 0.9787 | |
|
| 0.051 | 18.99 | 1140 | 0.0761 | 0.9685 | 0.9697 | 0.9691 | 0.9786 | |
|
| 0.051 | 19.99 | 1200 | 0.0774 | 0.9678 | 0.9698 | 0.9688 | 0.9784 | |
|
| 0.051 | 20.99 | 1260 | 0.0796 | 0.9685 | 0.9694 | 0.9690 | 0.9785 | |
|
| 0.051 | 21.99 | 1320 | 0.0796 | 0.9681 | 0.9701 | 0.9691 | 0.9786 | |
|
| 0.051 | 22.99 | 1380 | 0.0820 | 0.9684 | 0.9690 | 0.9687 | 0.9784 | |
|
| 0.051 | 23.99 | 1440 | 0.0829 | 0.9679 | 0.9688 | 0.9683 | 0.9781 | |
|
| 0.0318 | 24.99 | 1500 | 0.0854 | 0.9681 | 0.9690 | 0.9686 | 0.9782 | |
|
| 0.0318 | 25.99 | 1560 | 0.0881 | 0.9677 | 0.9692 | 0.9684 | 0.9782 | |
|
| 0.0318 | 26.99 | 1620 | 0.0893 | 0.9679 | 0.9690 | 0.9685 | 0.9783 | |
|
| 0.0318 | 27.99 | 1680 | 0.0910 | 0.9676 | 0.9691 | 0.9683 | 0.9781 | |
|
| 0.0318 | 28.99 | 1740 | 0.0919 | 0.9684 | 0.9686 | 0.9685 | 0.9783 | |
|
| 0.0318 | 29.99 | 1800 | 0.0933 | 0.9678 | 0.9686 | 0.9682 | 0.9781 | |
|
| 0.0318 | 30.99 | 1860 | 0.0947 | 0.9677 | 0.9688 | 0.9683 | 0.9781 | |
|
| 0.0318 | 31.99 | 1920 | 0.0966 | 0.9678 | 0.9694 | 0.9686 | 0.9783 | |
|
| 0.0318 | 32.99 | 1980 | 0.0974 | 0.9677 | 0.9689 | 0.9683 | 0.9781 | |
|
| 0.0211 | 33.99 | 2040 | 0.0981 | 0.9684 | 0.9693 | 0.9688 | 0.9784 | |
|
| 0.0211 | 34.99 | 2100 | 0.0989 | 0.9681 | 0.9690 | 0.9686 | 0.9783 | |
|
| 0.0211 | 35.99 | 2160 | 0.1008 | 0.9679 | 0.9695 | 0.9687 | 0.9784 | |
|
| 0.0211 | 36.99 | 2220 | 0.1015 | 0.9681 | 0.9689 | 0.9685 | 0.9782 | |
|
| 0.0211 | 37.99 | 2280 | 0.1015 | 0.9677 | 0.9689 | 0.9683 | 0.9781 | |
|
| 0.0211 | 38.99 | 2340 | 0.1024 | 0.9679 | 0.9690 | 0.9684 | 0.9782 | |
|
| 0.0211 | 39.99 | 2400 | 0.1022 | 0.9680 | 0.9690 | 0.9685 | 0.9782 | |
|
|
|
|
|
### Framework versions |
|
|
|
- Transformers 4.25.1 |
|
- Pytorch 1.12.0 |
|
- Datasets 2.18.0 |
|
- Tokenizers 0.13.2 |
|
|
