<!-- 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. -->
toxicBERT-params-tryout
This model is a fine-tuned version of unitary/toxic-bert on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.1804
- Precision: 0.0
- Recall: 0.0
- F1: 0.0
- Accuracy: 0.9314
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: 128
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.15
- num_epochs: 7
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 22 | 0.4522 | 0.0 | 0.0 | 0.0 | 0.8795 |
| No log | 2.0 | 44 | 0.2784 | 0.0 | 0.0 | 0.0 | 0.8996 |
| No log | 3.0 | 66 | 0.2150 | 0.0 | 0.0 | 0.0 | 0.9219 |
| No log | 4.0 | 88 | 0.1888 | 0.0 | 0.0 | 0.0 | 0.9297 |
| No log | 5.0 | 110 | 0.1829 | 0.0 | 0.0 | 0.0 | 0.9303 |
| No log | 6.0 | 132 | 0.1810 | 0.0 | 0.0 | 0.0 | 0.9305 |
| No log | 7.0 | 154 | 0.1804 | 0.0 | 0.0 | 0.0 | 0.9314 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.7.1
- Tokenizers 0.13.2
