metadata

base_model: clicknext/phayathaibert
tags:
  - generated_from_trainer
metrics:
  - precision
  - recall
  - f1
  - accuracy
model-index:
  - name: phayathaibert-thainer
    results: []
widget:
  - text: ประเทศไทยอยู่ในทวีปเอเชีย
    example_title: test_example_1
  - text: ไทยอยู่ในเจอ
    example_title: test_example_2
license: mit
language:
  - th
library_name: transformers
pipeline_tag: token-classification
datasets:
  - pythainlp/thainer-corpus-v2

phayathaibert-thainer

This model is a fine-tuned version of clicknext/phayathaibert on an unknown dataset. It achieves the following results on the evaluation set:

Loss: 0.1324
Precision: 0.8432
Recall: 0.8915
F1: 0.8666
Accuracy: 0.9735

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: 10
mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	493	0.1401	0.7300	0.7941	0.7607	0.9607
0.3499	2.0	986	0.1201	0.7863	0.8464	0.8152	0.9688
0.0961	3.0	1479	0.1169	0.8050	0.8663	0.8345	0.9715
0.0617	4.0	1972	0.1137	0.8155	0.8656	0.8398	0.9718
0.0438	5.0	2465	0.1280	0.8201	0.8714	0.8450	0.9725
0.0302	6.0	2958	0.1386	0.8266	0.8730	0.8492	0.9726
0.0239	7.0	3451	0.1401	0.8353	0.8789	0.8565	0.9733
0.0166	8.0	3944	0.1444	0.8356	0.8782	0.8564	0.9738
0.0139	9.0	4437	0.1530	0.8341	0.8785	0.8557	0.9735
0.0106	10.0	4930	0.1508	0.8394	0.8782	0.8583	0.9738

Framework versions

Transformers 4.35.2
Pytorch 2.1.0+cu118
Datasets 2.15.0
Tokenizers 0.15.0