| --- |
| library_name: transformers |
| tags: |
| - generated_from_trainer |
| metrics: |
| - f1 |
| model-index: |
| - name: radar-encoder-freeze |
| results: [] |
| --- |
| |
| <!-- This model card has been generated automatically according to the information the Trainer had access to. You |
| should probably proofread and complete it, then remove this comment. --> |
|
|
| # radar-encoder-freeze |
|
|
| This model is a fine-tuned version of [](https://huggingface.co/) on the None dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 0.1800 |
| - Roc-auc: 0.97 |
| - Brier: 0.934 |
| - C@1: 0.913 |
| - F1: 0.907 |
| - F05u: 0.941 |
| - Mean: 0.933 |
|
|
| ## 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: 64 |
| - 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: cosine |
| - lr_scheduler_warmup_steps: 0.03 |
| - num_epochs: 10 |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | Roc-auc | Brier | C@1 | F1 | F05u | Mean | |
| |:-------------:|:------:|:-----:|:---------------:|:-------:|:-----:|:-----:|:-----:|:-----:|:-----:| |
| | 0.3978 | 0.4153 | 500 | 0.3255 | 0.9 | 0.873 | 0.816 | 0.806 | 0.875 | 0.854 | |
| | 0.3261 | 0.8306 | 1000 | 0.2694 | 0.937 | 0.897 | 0.851 | 0.844 | 0.907 | 0.887 | |
| | 0.1887 | 1.2458 | 1500 | 0.2409 | 0.949 | 0.911 | 0.877 | 0.877 | 0.912 | 0.905 | |
| | 0.1766 | 1.6611 | 2000 | 0.2422 | 0.957 | 0.904 | 0.867 | 0.862 | 0.926 | 0.903 | |
| | 0.2418 | 2.0764 | 2500 | 0.2167 | 0.961 | 0.917 | 0.885 | 0.883 | 0.931 | 0.915 | |
| | 0.2052 | 2.4917 | 3000 | 0.2028 | 0.962 | 0.927 | 0.901 | 0.904 | 0.929 | 0.924 | |
| | 0.1627 | 2.9070 | 3500 | 0.1999 | 0.965 | 0.929 | 0.903 | 0.906 | 0.927 | 0.926 | |
| | 0.1645 | 3.3223 | 4000 | 0.1948 | 0.967 | 0.932 | 0.908 | 0.913 | 0.922 | 0.928 | |
| | 0.1826 | 3.7375 | 4500 | 0.2087 | 0.97 | 0.917 | 0.887 | 0.884 | 0.939 | 0.919 | |
| | 0.0895 | 4.1528 | 5000 | 0.1864 | 0.971 | 0.929 | 0.903 | 0.902 | 0.943 | 0.93 | |
| | 0.0716 | 4.5681 | 5500 | 0.1904 | 0.972 | 0.927 | 0.902 | 0.901 | 0.945 | 0.929 | |
| | 0.0591 | 4.9834 | 6000 | 0.1982 | 0.973 | 0.923 | 0.898 | 0.896 | 0.945 | 0.927 | |
| | 0.0892 | 5.3987 | 6500 | 0.1719 | 0.974 | 0.936 | 0.913 | 0.914 | 0.946 | 0.936 | |
| | 0.0671 | 5.8140 | 7000 | 0.1727 | 0.974 | 0.936 | 0.913 | 0.914 | 0.945 | 0.936 | |
| | 0.0713 | 6.2292 | 7500 | 0.1767 | 0.976 | 0.933 | 0.909 | 0.909 | 0.948 | 0.935 | |
| | 0.0790 | 6.6445 | 8000 | 0.1641 | 0.976 | 0.94 | 0.919 | 0.922 | 0.944 | 0.94 | |
| | 0.0431 | 7.0598 | 8500 | 0.1692 | 0.976 | 0.936 | 0.914 | 0.915 | 0.949 | 0.938 | |
| | 0.0694 | 7.4751 | 9000 | 0.1685 | 0.976 | 0.937 | 0.915 | 0.916 | 0.948 | 0.939 | |
| | 0.0576 | 7.8904 | 9500 | 0.1633 | 0.976 | 0.94 | 0.919 | 0.921 | 0.946 | 0.941 | |
| | 0.0742 | 8.3056 | 10000 | 0.1653 | 0.976 | 0.939 | 0.917 | 0.919 | 0.948 | 0.94 | |
| | 0.0872 | 8.7209 | 10500 | 0.1638 | 0.976 | 0.94 | 0.919 | 0.921 | 0.947 | 0.941 | |
| | 0.0777 | 9.1362 | 11000 | 0.1636 | 0.976 | 0.94 | 0.918 | 0.92 | 0.946 | 0.94 | |
| | 0.0511 | 9.5515 | 11500 | 0.1639 | 0.976 | 0.939 | 0.918 | 0.92 | 0.947 | 0.94 | |
| | 0.0381 | 9.9668 | 12000 | 0.1634 | 0.976 | 0.94 | 0.918 | 0.92 | 0.946 | 0.94 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 5.0.0 |
| - Pytorch 2.10.0+cu128 |
| - Datasets 4.0.0 |
| - Tokenizers 0.22.2 |
|
|
