metadata
library_name: transformers
license: apache-2.0
base_model: philschmid/tiny-bert-sst2-distilled
tags:
- generated_from_trainer
metrics:
- accuracy
- precision
- recall
model-index:
- name: toxiBERT_v1
results: []
toxiBERT_v1
This model is a fine-tuned version of philschmid/tiny-bert-sst2-distilled on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2924
- Accuracy: 0.8901
- F1 Binary: 0.7643
- Precision: 0.6946
- Recall: 0.8496
- Roc Auc: 0.9505
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Use adamw_torch_fused with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Binary | Precision | Recall | Roc Auc |
|---|---|---|---|---|---|---|---|---|
| 0.4005 | 1.0 | 2062 | 0.3620 | 0.8571 | 0.6993 | 0.6259 | 0.7924 | 0.9196 |
| 0.3372 | 2.0 | 4124 | 0.3213 | 0.8668 | 0.7265 | 0.6381 | 0.8433 | 0.9374 |
| 0.3157 | 3.0 | 6186 | 0.3061 | 0.8807 | 0.7473 | 0.6721 | 0.8415 | 0.9439 |
| 0.3148 | 4.0 | 8248 | 0.3019 | 0.8859 | 0.7569 | 0.6839 | 0.8473 | 0.9466 |
| 0.3093 | 5.0 | 10310 | 0.3026 | 0.8898 | 0.7616 | 0.6967 | 0.8398 | 0.9473 |
Framework versions
- Transformers 4.57.6
- Pytorch 2.9.1+cpu
- Datasets 4.4.2
- Tokenizers 0.22.1