metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- udpos28
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: udpos28-sm-first-POS
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: udpos28
type: udpos28
args: en
metrics:
- name: Precision
type: precision
value: 0.9511089206505667
- name: Recall
type: recall
value: 0.9546093116207286
- name: F1
type: f1
value: 0.9528559014062253
- name: Accuracy
type: accuracy
value: 0.9559133601686793
udpos28-sm-first-POS
This model is a fine-tuned version of bert-base-cased on the udpos28 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1896
- Precision: 0.9511
- Recall: 0.9546
- F1: 0.9529
- Accuracy: 0.9559
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: 4
- eval_batch_size: 4
- 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: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| 0.1696 | 1.0 | 4978 | 0.1700 | 0.9440 | 0.9464 | 0.9452 | 0.9472 |
| 0.0973 | 2.0 | 9956 | 0.1705 | 0.9487 | 0.9533 | 0.9510 | 0.9543 |
| 0.0508 | 3.0 | 14934 | 0.1896 | 0.9511 | 0.9546 | 0.9529 | 0.9559 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.2+cu102
- Datasets 2.2.2
- Tokenizers 0.12.1