metadata
library_name: transformers
base_model: allenai/scibert_scivocab_uncased
tags:
- generated_from_trainer
metrics:
- accuracy
model-index:
- name: sciBERT-RA-Cross-Encoder
results: []
sciBERT-RA-Cross-Encoder
This model is a fine-tuned version of allenai/scibert_scivocab_uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.4631
- Accuracy: 0.7516
- Macro F1: 0.7516
- Genuine F1: 0.7532
- Implausible F1: 0.75
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: 16
- eval_batch_size: 16
- 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: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | Macro F1 | Genuine F1 | Implausible F1 |
|---|---|---|---|---|---|---|---|
| 0.5645 | 1.0 | 315 | 0.5454 | 0.6799 | 0.6789 | 0.6968 | 0.661 |
| 0.5414 | 2.0 | 630 | 0.5271 | 0.7245 | 0.7238 | 0.7375 | 0.7102 |
| 0.3990 | 3.0 | 945 | 0.5617 | 0.7452 | 0.7452 | 0.7444 | 0.746 |
| 0.2856 | 4.0 | 1260 | 0.6154 | 0.7564 | 0.7553 | 0.7713 | 0.7394 |
| 0.2646 | 5.0 | 1575 | 0.8129 | 0.7564 | 0.756 | 0.7657 | 0.7463 |
| 0.1687 | 6.0 | 1890 | 1.2355 | 0.7484 | 0.748 | 0.7375 | 0.7584 |
| 0.1542 | 7.0 | 2205 | 1.2668 | 0.7484 | 0.7484 | 0.7492 | 0.7476 |
| 0.0854 | 8.0 | 2520 | 1.4631 | 0.7516 | 0.7516 | 0.7532 | 0.75 |
Framework versions
- Transformers 5.17.0
- Pytorch 2.11.0+cu130
- Datasets 4.8.5
- Tokenizers 0.23.2