| --- |
| library_name: transformers |
| license: apache-2.0 |
| base_model: deepvk/RuModernBERT-small |
| tags: |
| - generated_from_trainer |
| metrics: |
| - f1 |
| - precision |
| - recall |
| model-index: |
| - name: rumodernbert_ner_ft_small |
| 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. --> |
|
|
| # rumodernbert_ner_ft_small |
| |
| This model is a fine-tuned version of [deepvk/RuModernBERT-small](https://huggingface.co/deepvk/RuModernBERT-small) on an unknown dataset. |
| It achieves the following results on the evaluation set: |
| - Loss: 0.2842 |
| - F1: 0.8257 |
| - Precision: 0.8080 |
| - Recall: 0.8443 |
| |
| ## 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: 32 |
| - seed: 42 |
| - gradient_accumulation_steps: 2 |
| - total_train_batch_size: 32 |
| - 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 |
| - lr_scheduler_warmup_steps: 0.1 |
| - num_epochs: 10 |
| - mixed_precision_training: Native AMP |
|
|
| ### Training results |
|
|
| | Training Loss | Epoch | Step | Validation Loss | F1 | Precision | Recall | |
| |:-------------:|:-----:|:----:|:---------------:|:------:|:---------:|:------:| |
| | 1.3906 | 0.32 | 200 | 0.6057 | 0.4393 | 0.3840 | 0.5132 | |
| | 0.8110 | 0.64 | 400 | 0.3561 | 0.6269 | 0.5835 | 0.6774 | |
| | 0.6350 | 0.96 | 600 | 0.3008 | 0.7213 | 0.6812 | 0.7665 | |
| | 0.4992 | 1.28 | 800 | 0.2633 | 0.7803 | 0.7699 | 0.7909 | |
| | 0.3823 | 1.6 | 1000 | 0.2300 | 0.8084 | 0.7936 | 0.8238 | |
| | 0.3995 | 1.92 | 1200 | 0.2162 | 0.8009 | 0.7918 | 0.8102 | |
| | 0.2860 | 2.24 | 1400 | 0.2195 | 0.8109 | 0.7988 | 0.8234 | |
| | 0.2833 | 2.56 | 1600 | 0.2065 | 0.8140 | 0.7955 | 0.8335 | |
| | 0.2660 | 2.88 | 1800 | 0.2374 | 0.8145 | 0.8020 | 0.8274 | |
| | 0.1466 | 3.2 | 2000 | 0.2693 | 0.8079 | 0.7838 | 0.8335 | |
| | 0.1809 | 3.52 | 2200 | 0.2504 | 0.8299 | 0.8210 | 0.8391 | |
| | 0.1685 | 3.84 | 2400 | 0.2267 | 0.8315 | 0.8134 | 0.8503 | |
| | 0.0860 | 4.16 | 2600 | 0.2938 | 0.8273 | 0.8106 | 0.8447 | |
| | 0.0884 | 4.48 | 2800 | 0.2980 | 0.8183 | 0.7959 | 0.8419 | |
| | 0.1116 | 4.8 | 3000 | 0.2842 | 0.8257 | 0.8080 | 0.8443 | |
|
|
|
|
| ### Framework versions |
|
|
| - Transformers 5.1.0 |
| - Pytorch 2.10.0+cu128 |
| - Datasets 4.7.0 |
| - Tokenizers 0.22.2 |
|
|
