metadata
license: mit
base_model: xlm-roberta-base
tags:
- generated_from_trainer
datasets:
- hate_speech_filipino
metrics:
- accuracy
- f1
model-index:
- name: >-
scenario-kd-from-pre-finetune-silver-div-2-data-hate_speech_filipino-model-xlm-r
results: []
scenario-kd-from-pre-finetune-silver-div-2-data-hate_speech_filipino-model-xlm-r
This model is a fine-tuned version of xlm-roberta-base on the hate_speech_filipino dataset. It achieves the following results on the evaluation set:
- Loss: 0.4122
- Accuracy: 0.7810
- F1: 0.7699
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 6969
Training results
Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 |
---|---|---|---|---|---|
No log | 0.32 | 100 | 1.4902 | 0.7143 | 0.7119 |
No log | 0.64 | 200 | 1.0099 | 0.7379 | 0.7416 |
No log | 0.96 | 300 | 0.9000 | 0.7531 | 0.7410 |
No log | 1.28 | 400 | 0.7416 | 0.7595 | 0.7548 |
1.3478 | 1.6 | 500 | 0.7405 | 0.7703 | 0.7495 |
1.3478 | 1.92 | 600 | 0.6557 | 0.7743 | 0.7569 |
1.3478 | 2.24 | 700 | 0.6338 | 0.7661 | 0.7656 |
1.3478 | 2.56 | 800 | 0.6421 | 0.7682 | 0.7429 |
1.3478 | 2.88 | 900 | 0.5503 | 0.7746 | 0.7649 |
0.667 | 3.19 | 1000 | 0.5596 | 0.7741 | 0.7623 |
0.667 | 3.51 | 1100 | 0.6638 | 0.7729 | 0.7365 |
0.667 | 3.83 | 1200 | 0.5504 | 0.7798 | 0.7545 |
0.667 | 4.15 | 1300 | 0.5163 | 0.7769 | 0.7613 |
0.667 | 4.47 | 1400 | 0.6817 | 0.7498 | 0.7641 |
0.4362 | 4.79 | 1500 | 0.5097 | 0.7717 | 0.7725 |
0.4362 | 5.11 | 1600 | 0.5382 | 0.7765 | 0.7460 |
0.4362 | 5.43 | 1700 | 0.5237 | 0.7800 | 0.7493 |
0.4362 | 5.75 | 1800 | 0.4712 | 0.7786 | 0.7552 |
0.4362 | 6.07 | 1900 | 0.4671 | 0.7755 | 0.7570 |
0.3401 | 6.39 | 2000 | 0.4376 | 0.7859 | 0.7752 |
0.3401 | 6.71 | 2100 | 0.4753 | 0.7706 | 0.7720 |
0.3401 | 7.03 | 2200 | 0.4661 | 0.7854 | 0.7706 |
0.3401 | 7.35 | 2300 | 0.4555 | 0.7774 | 0.7605 |
0.3401 | 7.67 | 2400 | 0.4362 | 0.7812 | 0.7683 |
0.2844 | 7.99 | 2500 | 0.4511 | 0.7864 | 0.7734 |
0.2844 | 8.31 | 2600 | 0.4433 | 0.7767 | 0.7707 |
0.2844 | 8.63 | 2700 | 0.5001 | 0.7656 | 0.7699 |
0.2844 | 8.95 | 2800 | 0.4337 | 0.7810 | 0.7723 |
0.2844 | 9.27 | 2900 | 0.4145 | 0.7812 | 0.7668 |
0.2436 | 9.58 | 3000 | 0.4143 | 0.7840 | 0.7698 |
0.2436 | 9.9 | 3100 | 0.3993 | 0.7786 | 0.7687 |
0.2436 | 10.22 | 3200 | 0.4030 | 0.7899 | 0.7734 |
0.2436 | 10.54 | 3300 | 0.4054 | 0.7826 | 0.7666 |
0.2436 | 10.86 | 3400 | 0.3996 | 0.7762 | 0.7716 |
0.2246 | 11.18 | 3500 | 0.4122 | 0.7810 | 0.7699 |
Framework versions
- Transformers 4.33.3
- Pytorch 2.0.1
- Datasets 2.14.5
- Tokenizers 0.13.3