roberta-large-aeslc
This model is a fine-tuned version of FacebookAI/roberta-large on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 1.1436
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- 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: 0.06
- num_epochs: 8
Training results
| Training Loss | Epoch | Step | Validation Loss |
|---|---|---|---|
| 2.9303 | 0.2463 | 100 | 1.4581 |
| 1.5302 | 0.4926 | 200 | 1.3927 |
| 1.4989 | 0.7389 | 300 | 1.3911 |
| 1.4649 | 0.9852 | 400 | 1.3221 |
| 1.4015 | 1.2315 | 500 | 1.3565 |
| 1.3874 | 1.4778 | 600 | 1.2956 |
| 1.3699 | 1.7241 | 700 | 1.2915 |
| 1.3664 | 1.9704 | 800 | 1.2818 |
| 1.3284 | 2.2167 | 900 | 1.2541 |
| 1.3181 | 2.4631 | 1000 | 1.2590 |
| 1.3009 | 2.7094 | 1100 | 1.2409 |
| 1.3332 | 2.9557 | 1200 | 1.2647 |
| 1.2581 | 3.2020 | 1300 | 1.2192 |
| 1.2849 | 3.4483 | 1400 | 1.2634 |
| 1.2838 | 3.6946 | 1500 | 1.2322 |
| 1.2664 | 3.9409 | 1600 | 1.2561 |
| 1.2275 | 4.1872 | 1700 | 1.1981 |
| 1.2216 | 4.4335 | 1800 | 1.2147 |
| 1.2317 | 4.6798 | 1900 | 1.2067 |
| 1.2038 | 4.9261 | 2000 | 1.2021 |
| 1.2022 | 5.1724 | 2100 | 1.2170 |
| 1.2011 | 5.4187 | 2200 | 1.1946 |
| 1.1699 | 5.6650 | 2300 | 1.1933 |
| 1.1717 | 5.9113 | 2400 | 1.1642 |
| 1.1799 | 6.1576 | 2500 | 1.1737 |
| 1.1637 | 6.4039 | 2600 | 1.1751 |
| 1.1417 | 6.6502 | 2700 | 1.1575 |
| 1.1449 | 6.8966 | 2800 | 1.1655 |
| 1.1621 | 7.1429 | 2900 | 1.1472 |
| 1.1502 | 7.3892 | 3000 | 1.1504 |
| 1.1497 | 7.6355 | 3100 | 1.1807 |
| 1.1274 | 7.8818 | 3200 | 1.1436 |
Framework versions
- Transformers 4.53.3
- Pytorch 2.6.0+cu124
- Datasets 3.6.0
- Tokenizers 0.21.2