climatebert-multi
This model is a fine-tuned version of climatebert/distilroberta-base-climate-f on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6759
- F1 Macro: 0.7668
- Accuracy: 0.8707
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: 32
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 20
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 Macro | Accuracy |
|---|---|---|---|---|---|
| 1.9279 | 1.0 | 17 | 1.7128 | 0.2639 | 0.6552 |
| 1.5847 | 2.0 | 34 | 1.5100 | 0.2639 | 0.6552 |
| 1.3489 | 3.0 | 51 | 1.0729 | 0.5454 | 0.8017 |
| 1.0319 | 4.0 | 68 | 0.8453 | 0.5843 | 0.8103 |
| 0.6454 | 5.0 | 85 | 0.6765 | 0.7668 | 0.8707 |
| 0.5125 | 6.0 | 102 | 0.6657 | 0.7466 | 0.8707 |
| 0.4469 | 7.0 | 119 | 0.7136 | 0.7605 | 0.8448 |
| 0.2631 | 8.0 | 136 | 0.6833 | 0.7559 | 0.8534 |
Framework versions
- Transformers 5.0.0
- Pytorch 2.10.0+cu128
- Datasets 5.0.0
- Tokenizers 0.22.2