furina_seed42_eng_amh_esp_cross_latin_2e-05
This model is a fine-tuned version of yihongLiu/furina on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0195
- Spearman Corr: 0.7405
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 | 0.53 | 200 | 0.0233 | 0.6222 |
No log | 1.06 | 400 | 0.0203 | 0.7114 |
No log | 1.59 | 600 | 0.0168 | 0.7194 |
0.0412 | 2.12 | 800 | 0.0172 | 0.7202 |
0.0412 | 2.65 | 1000 | 0.0169 | 0.7408 |
0.0412 | 3.17 | 1200 | 0.0177 | 0.7461 |
0.0412 | 3.7 | 1400 | 0.0170 | 0.7439 |
0.0188 | 4.23 | 1600 | 0.0173 | 0.7441 |
0.0188 | 4.76 | 1800 | 0.0190 | 0.7393 |
0.0188 | 5.29 | 2000 | 0.0173 | 0.7438 |
0.0188 | 5.82 | 2200 | 0.0195 | 0.7405 |
Framework versions
- Transformers 4.37.2
- Pytorch 2.1.0+cu121
- Datasets 2.17.0
- Tokenizers 0.15.2
- Downloads last month
- 0