roberta-large_MBTI_I_MBTI_agg_balanced_50_1e-05
This model was trained from scratch on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 4.3322
- F1: 0.6721
- Threshold: 0.51
- F1 At 05: 0.6720
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- distributed_type: multi-GPU
- num_devices: 4
- total_train_batch_size: 64
- total_eval_batch_size: 64
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 400
- num_epochs: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Threshold | F1 At 05 |
|---|---|---|---|---|---|---|
| 2.7146 | 0.3175 | 200 | 2.6766 | 0.5785 | 0.47 | 0.5666 |
| 2.5452 | 0.6349 | 400 | 2.4528 | 0.6386 | 0.4500 | 0.6279 |
| 2.3593 | 0.9524 | 600 | 2.3498 | 0.6745 | 0.52 | 0.6701 |
| 2.0074 | 1.2698 | 800 | 2.4031 | 0.6730 | 0.4600 | 0.6656 |
| 1.9554 | 1.5873 | 1000 | 2.3860 | 0.6835 | 0.4600 | 0.6783 |
| 1.9264 | 1.9048 | 1200 | 2.3219 | 0.6790 | 0.4200 | 0.6686 |
| 1.4684 | 2.2222 | 1400 | 2.6639 | 0.6752 | 0.33 | 0.6661 |
| 1.5030 | 2.5397 | 1600 | 2.7574 | 0.6721 | 0.48 | 0.6695 |
| 1.5519 | 2.8571 | 1800 | 2.6145 | 0.6765 | 0.2900 | 0.6682 |
| 1.1416 | 3.1746 | 2000 | 3.4465 | 0.6689 | 0.56 | 0.6681 |
| 1.0598 | 3.4921 | 2200 | 3.4451 | 0.6748 | 0.4100 | 0.6717 |
| 1.0528 | 3.8095 | 2400 | 3.6080 | 0.6721 | 0.53 | 0.6712 |
| 0.7081 | 4.1270 | 2600 | 4.1815 | 0.6759 | 0.53 | 0.6723 |
| 0.7006 | 4.4444 | 2800 | 3.9992 | 0.6793 | 0.54 | 0.6771 |
| 0.7213 | 4.7619 | 3000 | 4.3322 | 0.6721 | 0.51 | 0.6720 |
Framework versions
- Transformers 5.3.0
- Pytorch 2.7.1
- Datasets 4.8.5
- Tokenizers 0.22.2