metadata
library_name: transformers
license: mit
base_model: roberta-base
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: combined-raw-scratch
results: []
combined-raw-scratch
This model is a fine-tuned version of roberta-base on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.1006
- Accuracy: 0.6704
- F1 Macro: 0.6712
- Composite Score: 0.6712
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: 3e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- 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_ratio: 0.1
- num_epochs: 20
- mixed_precision_training: Native AMP
- label_smoothing_factor: 0.2
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 Macro | Composite Score |
|---|---|---|---|---|---|---|
| 1.0358 | 1.0 | 1083 | 1.1850 | 0.4084 | 0.3684 | 0.3343 |
| 0.865 | 2.0 | 2166 | 1.0103 | 0.5690 | 0.5730 | 0.5730 |
| 0.7614 | 3.0 | 3249 | 1.0127 | 0.6105 | 0.6113 | 0.6113 |
| 0.695 | 4.0 | 4332 | 1.0336 | 0.6271 | 0.6287 | 0.6287 |
| 0.648 | 5.0 | 5415 | 1.0003 | 0.6618 | 0.6642 | 0.6642 |
| 0.6011 | 6.0 | 6498 | 1.0536 | 0.6614 | 0.6642 | 0.6642 |
| 0.5675 | 7.0 | 7581 | 1.0561 | 0.6765 | 0.6777 | 0.6777 |
| 0.5539 | 8.0 | 8664 | 1.0948 | 0.6716 | 0.6730 | 0.6730 |
| 0.5451 | 9.0 | 9747 | 1.0868 | 0.6742 | 0.6756 | 0.6756 |
| 0.5385 | 10.0 | 10830 | 1.1006 | 0.6704 | 0.6712 | 0.6712 |
Framework versions
- Transformers 4.53.3
- Pytorch 2.6.0+cu124
- Datasets 4.4.1
- Tokenizers 0.21.2