metadata
library_name: peft
license: cc-by-4.0
base_model: pczarnik/herbert-base-ner
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
model-index:
- name: herbert-ner-lora-sensitive
results: []
herbert-ner-lora-sensitive
This model is a fine-tuned version of pczarnik/herbert-base-ner on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.0070
- Precision: 0.8622
- Recall: 0.8596
- F1: 0.8609
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: 0.0002
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Use 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: 5
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 |
|---|---|---|---|---|---|---|
| No log | 1.0 | 315 | 0.0148 | 0.8076 | 0.7253 | 0.7642 |
| 0.1305 | 2.0 | 630 | 0.0096 | 0.8339 | 0.8133 | 0.8234 |
| 0.1305 | 3.0 | 945 | 0.0079 | 0.8538 | 0.8472 | 0.8505 |
| 0.0112 | 4.0 | 1260 | 0.0072 | 0.8556 | 0.8596 | 0.8576 |
| 0.0092 | 5.0 | 1575 | 0.0070 | 0.8622 | 0.8596 | 0.8609 |
Framework versions
- PEFT 0.12.0
- Transformers 4.50.3
- Pytorch 2.4.1
- Datasets 2.21.0
- Tokenizers 0.21.1