metadata

language:
  - vi
license: apache-2.0
base_model: google-bert/bert-base-multilingual-cased
tags:
  - generated_from_trainer
model-index:
  - name: bert-base-multilingual-cased_baseline_syllables
    results: []

bert-base-multilingual-cased_baseline_syllables

This model is a fine-tuned version of google-bert/bert-base-multilingual-cased on the covid19_ner dataset. It achieves the following results on the evaluation set:

Loss: 0.1003
Patient Id: 0.9860
Name: 0.9239
Gender: 0.9642
Age: 0.9834
Job: 0.7734
Location: 0.9494
Organization: 0.8827
Date: 0.9883
Symptom And Disease: 0.8631
Transportation: 0.9773
F1 Macro: 0.9292
F1 Micro: 0.9463

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: 5

Training results

Training Loss	Epoch	Step	Validation Loss	Patient Id	Name	Gender	Age	Job	Location	Organization	Date	Symptom And Disease	Transportation	F1 Macro	F1 Micro
0.1923	1.0	629	0.1099	0.9735	0.9049	0.9288	0.9247	0.5296	0.9225	0.7969	0.9861	0.7873	0.9189	0.8673	0.9102
0.0555	2.0	1258	0.0937	0.9804	0.9144	0.9586	0.9847	0.4700	0.9390	0.8694	0.9856	0.8604	0.9180	0.8881	0.9353
0.0334	3.0	1887	0.0875	0.9772	0.9153	0.9590	0.9806	0.7568	0.9450	0.8676	0.9883	0.8571	0.96	0.9207	0.9402
0.0207	4.0	2516	0.0972	0.9859	0.9284	0.9590	0.9808	0.7742	0.9507	0.8919	0.9869	0.8616	0.9718	0.9291	0.9468
0.014	5.0	3145	0.1003	0.9860	0.9239	0.9642	0.9834	0.7734	0.9494	0.8827	0.9883	0.8631	0.9773	0.9292	0.9463

Framework versions

Transformers 4.41.2
Pytorch 2.1.2
Datasets 2.19.2
Tokenizers 0.19.1