metadata
tags:
- generated_from_trainer
datasets:
- udpos28
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: parsbert-finetuned-pos
results:
- task:
name: Token Classification
type: token-classification
dataset:
name: udpos28
type: udpos28
args: fa
metrics:
- name: Precision
type: precision
value: 0.9447937270415372
- name: Recall
type: recall
value: 0.9486470191864382
- name: F1
type: f1
value: 0.9467164522465448
- name: Accuracy
type: accuracy
value: 0.9598951738759165
parsbert-finetuned-pos
This model is a fine-tuned version of HooshvareLab/bert-base-parsbert-uncased on the udpos28 dataset. It achieves the following results on the evaluation set:
- Loss: 0.1385
- Precision: 0.9448
- Recall: 0.9486
- F1: 0.9467
- Accuracy: 0.9599
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: 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 |
|---|---|---|---|---|---|---|---|
| 0.122 | 1.0 | 3103 | 0.1215 | 0.9363 | 0.9424 | 0.9394 | 0.9561 |
| 0.0735 | 2.0 | 6206 | 0.1297 | 0.9413 | 0.9474 | 0.9443 | 0.9582 |
| 0.0373 | 3.0 | 9309 | 0.1385 | 0.9448 | 0.9486 | 0.9467 | 0.9599 |
Framework versions
- Transformers 4.18.0
- Pytorch 1.10.0
- Datasets 2.0.0
- Tokenizers 0.11.6