metadata
license: cc-by-4.0
tags:
- generated_from_trainer
datasets:
- wikiann
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: herbert-base-ner
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: wikiann
type: wikiann
config: pl
split: validation
args: pl
metrics:
- name: Precision
type: precision
value: 0.8885878330430295
- name: Recall
type: recall
value: 0.905945803735859
- name: F1
type: f1
value: 0.8971828692395376
- name: Accuracy
type: accuracy
value: 0.9568532096363909
herbert-base-ner
This model is a fine-tuned version of allegro/herbert-base-cased on the wikiann dataset. It achieves the following results on the evaluation set:
- Loss: 0.2006
- Precision: 0.8886
- Recall: 0.9059
- F1: 0.8972
- Accuracy: 0.9569
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: 1e-05
- train_batch_size: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.207 | 1.0 | 2500 | 0.1929 | 0.8566 | 0.8884 | 0.8722 | 0.9499 |
| 0.1528 | 2.0 | 5000 | 0.1979 | 0.8807 | 0.9006 | 0.8905 | 0.9547 |
| 0.1195 | 3.0 | 7500 | 0.2006 | 0.8886 | 0.9059 | 0.8972 | 0.9569 |
Framework versions
- Transformers 4.29.2
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3