tape-fluorescence-prediction-RITA_s
This model is a fine-tuned version of lightonai/RITA_s on the cradle-bio/tape-fluorescence dataset. It achieves the following results on the evaluation set:
- Loss: 0.5855
- Spearmanr: 0.2955
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: 5e-05
- train_batch_size: 32
- eval_batch_size: 32
- seed: 42
- gradient_accumulation_steps: 128
- total_train_batch_size: 4096
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 10
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Spearmanr |
|---|---|---|---|---|
| 4.3595 | 0.85 | 4 | 0.7057 | 0.0940 |
| 0.8654 | 1.85 | 8 | 0.6873 | 0.1280 |
| 0.8292 | 2.85 | 12 | 0.6835 | 0.2290 |
| 0.8212 | 3.85 | 16 | 0.6837 | 0.3110 |
| 0.8191 | 4.85 | 20 | 0.6799 | 0.3281 |
| 0.8137 | 5.85 | 24 | 0.6748 | 0.3277 |
| 0.8057 | 6.85 | 28 | 0.6592 | 0.3162 |
| 0.7769 | 7.85 | 32 | 0.6283 | 0.3065 |
| 0.7382 | 8.85 | 36 | 0.6103 | 0.2795 |
| 0.5991 | 9.85 | 40 | 0.5855 | 0.2955 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.11.0
- Datasets 2.1.0
- Tokenizers 0.12.1