metadata
library_name: transformers
license: apache-2.0
base_model: bert-base-cased
tags:
- generated_from_trainer
datasets:
- conll2003
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: bert-finetuned-ner-tokenclass
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: conll2003
type: conll2003
config: conll2003
split: validation
args: conll2003
metrics:
- name: Precision
type: precision
value: 0.845360824742268
- name: Recall
type: recall
value: 0.8986301369863013
- name: F1
type: f1
value: 0.8711819389110226
- name: Accuracy
type: accuracy
value: 0.9742014742014742
bert-finetuned-ner-tokenclass
This model is a fine-tuned version of bert-base-cased on the conll2003 dataset. It achieves the following results on the evaluation set:
- Loss: 0.0920
- Precision: 0.8454
- Recall: 0.8986
- F1: 0.8712
- Accuracy: 0.9742
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: 8
- eval_batch_size: 8
- 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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 125 | 0.1217 | 0.7805 | 0.8575 | 0.8172 | 0.9678 |
| No log | 2.0 | 250 | 0.0968 | 0.8394 | 0.8877 | 0.8628 | 0.9713 |
| No log | 3.0 | 375 | 0.0920 | 0.8454 | 0.8986 | 0.8712 | 0.9742 |
Framework versions
- Transformers 4.56.1
- Pytorch 2.6.0
- Datasets 3.6.0
- Tokenizers 0.22.2