metadata
library_name: transformers
license: mit
base_model: roberta-base
tags:
- generated_from_trainer
model-index:
- name: roberta-base-multi-head-cossine-10
results: []
roberta-base-multi-head-cossine-10
This model is a fine-tuned version of roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.7100
- F1 Macro: 0.5350
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: 5e-06
- train_batch_size: 2
- eval_batch_size: 2
- seed: 42
- gradient_accumulation_steps: 8
- total_train_batch_size: 16
- 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: cosine
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 20
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Macro |
|---|---|---|---|---|
| 0.5136 | 1.0 | 1534 | 0.4906 | 0.3671 |
| 0.4655 | 2.0 | 3068 | 0.4630 | 0.4646 |
| 0.4352 | 3.0 | 4602 | 0.4589 | 0.5332 |
| 0.4024 | 4.0 | 6136 | 0.4639 | 0.5336 |
| 0.3569 | 5.0 | 7670 | 0.4787 | 0.5448 |
| 0.3306 | 6.0 | 9204 | 0.5224 | 0.5485 |
| 0.2919 | 7.0 | 10738 | 0.5353 | 0.5494 |
| 0.2608 | 8.0 | 12272 | 0.5766 | 0.5495 |
| 0.2273 | 9.0 | 13806 | 0.6381 | 0.5319 |
| 0.197 | 10.0 | 15340 | 0.6749 | 0.5434 |
| 0.1824 | 11.0 | 16874 | 0.7100 | 0.5350 |
Framework versions
- Transformers 4.53.1
- Pytorch 2.6.0+cu124
- Datasets 2.14.4
- Tokenizers 0.21.2