metadata

base_model: airesearch/wangchanberta-base-att-spm-uncased
tags:
  - generated_from_trainer
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: fine-tune-wangchanberta-stock-thai
    results: []

fine-tune-wangchanberta-stock-thai

This model is a fine-tuned version of airesearch/wangchanberta-base-att-spm-uncased on the None dataset. It achieves the following results on the evaluation set:

Loss: 1.0974
Accuracy: 0.4010
Precision: 0.3527
Recall: 0.4010
F1: 0.2439

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: 32
eval_batch_size: 64
seed: 42
optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
lr_scheduler_type: linear
num_epochs: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.0643	1.0	26	1.1045	0.3564	0.2361	0.3564	0.2750
1.08	2.0	52	1.0902	0.3960	0.1568	0.3960	0.2247
1.074	3.0	78	1.1016	0.3960	0.1568	0.3960	0.2247
1.0613	4.0	104	1.0929	0.3960	0.1568	0.3960	0.2247
1.0564	5.0	130	1.0990	0.3960	0.1568	0.3960	0.2247
1.053	6.0	156	1.1003	0.3960	0.1568	0.3960	0.2247
1.0498	7.0	182	1.0968	0.3911	0.1557	0.3911	0.2227
1.0444	8.0	208	1.0946	0.3911	0.1557	0.3911	0.2227
1.0418	9.0	234	1.0990	0.3960	0.1568	0.3960	0.2247
1.0385	10.0	260	1.0982	0.3960	0.3025	0.3960	0.2331
1.0352	11.0	286	1.0980	0.3911	0.1557	0.3911	0.2227
1.0401	12.0	312	1.1001	0.3911	0.1557	0.3911	0.2227
1.0395	13.0	338	1.0970	0.4010	0.3519	0.4010	0.2431
1.032	14.0	364	1.0971	0.4010	0.3519	0.4010	0.2431
1.0351	15.0	390	1.0977	0.4010	0.3527	0.4010	0.2439
1.0262	16.0	416	1.0970	0.4010	0.3527	0.4010	0.2439
1.0385	17.0	442	1.0970	0.4010	0.3527	0.4010	0.2439
1.031	18.0	468	1.0970	0.4010	0.3527	0.4010	0.2439
1.0313	19.0	494	1.0969	0.4010	0.3527	0.4010	0.2439
1.0429	20.0	520	1.0974	0.4010	0.3527	0.4010	0.2439

Framework versions

Transformers 4.40.2
Pytorch 2.2.1+cu121
Datasets 2.19.1
Tokenizers 0.19.1