metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- nerd
metrics:
- precision
- recall
- f1
- accuracy
model_index:
- name: ner_nerd_fine
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: nerd
type: nerd
args: nerd
metric:
name: Accuracy
type: accuracy
value: 0.9054809270602955
ner_nerd_fine
This model is a fine-tuned version of bert-base-uncased on the nerd dataset. It achieves the following results on the evaluation set:
- Loss: 0.3336
- Precision: 0.6298
- Recall: 0.6766
- F1: 0.6524
- Accuracy: 0.9055
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: 3e-05
- train_batch_size: 16
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_ratio: 0.1
- num_epochs: 10
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.6155 | 1.0 | 8235 | 0.3365 | 0.6047 | 0.6575 | 0.6300 | 0.9016 |
| 0.3067 | 2.0 | 16470 | 0.3174 | 0.6321 | 0.6641 | 0.6477 | 0.9065 |
| 0.2379 | 3.0 | 24705 | 0.3271 | 0.6335 | 0.6794 | 0.6557 | 0.9073 |
| 0.1835 | 4.0 | 32940 | 0.3449 | 0.6334 | 0.6755 | 0.6537 | 0.9066 |
| 0.139 | 5.0 | 41175 | 0.3866 | 0.6355 | 0.6771 | 0.6556 | 0.9068 |
Framework versions
- Transformers 4.9.1
- Pytorch 1.9.0+cu102
- Datasets 1.11.0
- Tokenizers 0.10.2