metadata
library_name: peft
license: mit
base_model: microsoft/deberta-v3-small
tags:
- generated_from_trainer
model-index:
- name: valuable-auk-490
results: []
valuable-auk-490
This model is a fine-tuned version of microsoft/deberta-v3-small on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.4671
- Hamming Loss: 0.1123
- Zero One Loss: 1.0
- Jaccard Score: 1.0
- Hamming Loss Optimised: 0.1123
- Hamming Loss Threshold: 0.5944
- Zero One Loss Optimised: 0.7662
- Zero One Loss Threshold: 0.4039
- Jaccard Score Optimised: 0.7638
- Jaccard Score Threshold: 0.4056
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: 6.612534950619908e-06
- train_batch_size: 4
- eval_batch_size: 4
- seed: 2024
- optimizer: Use OptimizerNames.ADAMW_TORCH with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- num_epochs: 4
Training results
| Training Loss | Epoch | Step | Validation Loss | Hamming Loss | Zero One Loss | Jaccard Score | Hamming Loss Optimised | Hamming Loss Threshold | Zero One Loss Optimised | Zero One Loss Threshold | Jaccard Score Optimised | Jaccard Score Threshold |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 0.6856 | 1.0 | 800 | 0.6690 | 0.3683 | 1.0 | 0.9304 | 0.1123 | 0.6927 | 0.9613 | 0.5584 | 0.8878 | 0.2889 |
| 0.5765 | 2.0 | 1600 | 0.5081 | 0.1123 | 1.0 | 1.0 | 0.1123 | 0.5944 | 1.0 | 0.9000 | 0.8559 | 0.4056 |
| 0.5242 | 3.0 | 2400 | 0.4754 | 0.1123 | 1.0 | 1.0 | 0.1123 | 0.5944 | 0.7662 | 0.4143 | 0.7638 | 0.4124 |
| 0.4923 | 4.0 | 3200 | 0.4671 | 0.1123 | 1.0 | 1.0 | 0.1123 | 0.5944 | 0.7662 | 0.4039 | 0.7638 | 0.4056 |
Framework versions
- PEFT 0.13.2
- Transformers 4.47.0
- Pytorch 2.5.1+cu124
- Datasets 3.1.0
- Tokenizers 0.21.0