berttiny-hate_speech18-nonpretrained
This model is a fine-tuned version of google/bert_uncased_L-2_H-128_A-2 on the hate_speech18 dataset. It achieves the following results on the evaluation set:
- Loss: 0.4631
- Accuracy: 0.8643
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: 6e-05
- train_batch_size: 128
- eval_batch_size: 128
- seed: 33
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 7
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 0.5429 | 1.0 | 60 | 0.4818 | 0.8556 |
| 0.4701 | 2.0 | 120 | 0.4711 | 0.8556 |
| 0.4623 | 3.0 | 180 | 0.4687 | 0.8583 |
| 0.4571 | 4.0 | 240 | 0.4645 | 0.8615 |
| 0.4532 | 5.0 | 300 | 0.4639 | 0.8638 |
| 0.4501 | 6.0 | 360 | 0.4656 | 0.8620 |
| 0.4494 | 7.0 | 420 | 0.4631 | 0.8643 |
Framework versions
- Transformers 4.36.0.dev0
- Pytorch 2.1.1
- Datasets 2.15.0
- Tokenizers 0.15.0