xlm-roberta-base_latin_amh-hau-eng_train_spearman_corr
This model is a fine-tuned version of FacebookAI/xlm-roberta-base on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0299
- Spearman Corr: 0.7565
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: 128
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 64
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 30
- mixed_precision_training: Native AMP
Training results
Training Loss | Epoch | Step | Validation Loss | Spearman Corr |
---|---|---|---|---|
No log | 0.52 | 200 | 0.0314 | 0.7022 |
No log | 1.04 | 400 | 0.0254 | 0.7158 |
No log | 1.55 | 600 | 0.0281 | 0.7444 |
0.0413 | 2.07 | 800 | 0.0301 | 0.7603 |
0.0413 | 2.59 | 1000 | 0.0316 | 0.7524 |
0.0413 | 3.11 | 1200 | 0.0362 | 0.7638 |
0.0413 | 3.63 | 1400 | 0.0275 | 0.7505 |
0.0266 | 4.15 | 1600 | 0.0299 | 0.7588 |
0.0266 | 4.66 | 1800 | 0.0334 | 0.7610 |
0.0266 | 5.18 | 2000 | 0.0317 | 0.7595 |
0.0266 | 5.7 | 2200 | 0.0274 | 0.7570 |
0.0202 | 6.22 | 2400 | 0.0250 | 0.7680 |
0.0202 | 6.74 | 2600 | 0.0300 | 0.7780 |
0.0202 | 7.25 | 2800 | 0.0321 | 0.7649 |
0.0202 | 7.77 | 3000 | 0.0306 | 0.7616 |
0.0155 | 8.29 | 3200 | 0.0335 | 0.7627 |
0.0155 | 8.81 | 3400 | 0.0293 | 0.7541 |
0.0155 | 9.33 | 3600 | 0.0286 | 0.7529 |
0.0155 | 9.84 | 3800 | 0.0283 | 0.7582 |
0.012 | 10.36 | 4000 | 0.0283 | 0.7551 |
0.012 | 10.88 | 4200 | 0.0299 | 0.7565 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.2.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2
- Downloads last month
- 4