metadata
license: mit
tags:
- generated_from_trainer
datasets:
- conll2003
model-index:
- name: bert-base-NER-finetuned-ner
results: []
bert-base-NER-finetuned-ner
This model is a fine-tuned version of dslim/bert-base-NER on the conll2003 dataset.
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: 64
- eval_batch_size: 64
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Framework versions
- Transformers 4.18.0
- Pytorch 1.12.0
- Datasets 2.7.1
- Tokenizers 0.11.0