IndoBert-base-ler

jafarabdurrohman/IndoBert-base-ler

IndoBert-base-ler is a token classification model from jafarabdurrohman. Use it when you need labels on individual words, such as names. It is set up for transformers. The card lists the license as mit.

should probably proofread and complete it, then remove this comment. --

Downloads · 30 days

4% of all-time downloads

All-time downloads

189

Public

Repo size

8.9 GB

Likes

Public

Hugging Face

Repo makeup

Click a slice to open those files.

.bin496 MB · 100%

At a glance

Task: Token Classification
Library: transformers
License: mit
Model type: bert
Access: Public
Created: Jun 8, 2023
Updated: Aug 2, 2023
SHA: a9206bb4

Try a prompt

Model card

From the Hugging Face model README

IndoBert-base-ler

This model is a fine-tuned version of indobenchmark/indobert-base-p1 on the None dataset. It achieves the following results on the evaluation set:

Loss: 0.0348
Overall Precision: 0.8392
Overall Recall: 0.8292
Overall F1: 0.8342
Overall Accuracy: 0.9961
Jenis amar F1: 0.9381
Jenis dakwaan F1: 0.8202
Jenis perkara F1: 0.7895
Melanggar uu (dakwaan) F1: 0.6704
Melanggar uu (pertimbangan hukum) F1: 0.5885
Melanggar uu (tuntutan) F1: 0.7783
Nama hakim anggota F1: 0.9045
Nama hakim ketua F1: 0.8854
Nama jaksa F1: 0.8905
Nama panitera F1: 0.9056
Nama pengacara F1: 0.8288
Nama pengadilan F1: 0.9964
Nama saksi F1: 0.8385
Nama terdakwa F1: 0.8264
Nomor putusan F1: 0.9359
Putusan hukuman F1: 0.6659
Tanggal kejadian F1: 0.3870
Tanggal putusan F1: 0.9430
Tingkat kasus F1: 0.9817
Tuntutan hukuman F1: 0.8348

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
max_sequence_length: 128
stride: 25% (32)
decay_rate: 0.01
num_epochs: 16

Training results

Training Loss	Epoch	Step	Validation Loss	Overall Precision	Overall Recall	Overall F1	Overall Accuracy	Jenis amar F1	Jenis dakwaan F1	Jenis perkara F1	Melanggar uu (dakwaan) F1	Melanggar uu (pertimbangan hukum) F1	Melanggar uu (tuntutan) F1	Nama hakim anggota F1	Nama hakim ketua F1	Nama jaksa F1	Nama panitera F1	Nama pengacara F1	Nama pengadilan F1	Nama saksi F1	Nama terdakwa F1	Nomor putusan F1	Putusan hukuman F1	Tanggal kejadian F1	Tanggal putusan F1	Tingkat kasus F1	Tuntutan hukuman F1
0.0214	1.0	7322	0.0174	0.7920	0.7699	0.7808	0.9956	0.8514	0.6352	0.6728	0.5940	0.5044	0.7329	0.8602	0.7357	0.8160	0.8355	0.6585	0.9661	0.8130	0.8053	0.9248	0.5354	0.3278	0.9372	0.9173	0.7353
0.0127	2.0	14644	0.0162	0.7760	0.7960	0.7859	0.9955	0.8982	0.6809	0.7038	0.5598	0.4747	0.6963	0.8733	0.8681	0.8604	0.8858	0.7114	0.9803	0.7601	0.8242	0.9318	0.6187	0.3976	0.9351	0.9560	0.7015
0.0117	3.0	21966	0.0172	0.7830	0.7701	0.7765	0.9953	0.8487	0.6657	0.7051	0.5092	0.5336	0.7518	0.8460	0.8093	0.7043	0.6803	0.7242	0.9802	0.8191	0.8039	0.9346	0.5290	0.3797	0.9312	0.9564	0.7534
0.0088	4.0	29288	0.0175	0.8086	0.8019	0.8052	0.9960	0.9093	0.7876	0.7571	0.6362	0.5500	0.7384	0.8832	0.8440	0.7949	0.8913	0.6986	0.9874	0.8193	0.8378	0.9089	0.5590	0.3968	0.9534	0.9640	0.7724
0.0092	5.0	36610	0.0171	0.8070	0.8035	0.8053	0.9958	0.8686	0.6188	0.7521	0.5808	0.5625	0.7645	0.8825	0.8168	0.8656	0.8557	0.7155	0.9803	0.8242	0.8132	0.9323	0.6011	0.3756	0.9211	0.9653	0.7570
0.0057	6.0	43932	0.0184	0.8157	0.8077	0.8117	0.9958	0.9050	0.8299	0.7505	0.6424	0.4908	0.7571	0.8822	0.8740	0.8625	0.8970	0.7475	0.9802	0.8158	0.8327	0.9389	0.5801	0.3892	0.944	0.9635	0.8073
0.0057	7.0	51254	0.0203	0.7988	0.8277	0.8130	0.9959	0.9273	0.7900	0.7673	0.5932	0.5577	0.7811	0.8863	0.8553	0.8743	0.8945	0.7176	0.9681	0.8316	0.8231	0.9374	0.5983	0.4006	0.9110	0.9620	0.8203
0.0054	8.0	58576	0.0209	0.8263	0.8058	0.8159	0.9959	0.8996	0.8097	0.7661	0.6445	0.5613	0.7778	0.9079	0.7732	0.8783	0.8968	0.7080	0.9910	0.8227	0.8355	0.9401	0.5395	0.3542	0.9279	0.9706	0.7937
0.003	9.0	65898	0.0244	0.8255	0.8096	0.8175	0.9956	0.9277	0.7944	0.7146	0.6556	0.5502	0.7842	0.8564	0.8798	0.8813	0.8955	0.7547	0.9857	0.8221	0.8270	0.9399	0.6681	0.3873	0.9468	0.9654	0.7912
0.0031	10.0	73220	0.0256	0.8297	0.8206	0.8251	0.9959	0.9103	0.8239	0.7598	0.6639	0.5665	0.7609	0.9008	0.8765	0.8867	0.9002	0.7590	0.9982	0.8359	0.8322	0.9409	0.5965	0.3774	0.9402	0.9635	0.8070
0.0021	11.0	80542	0.0259	0.8365	0.8238	0.8301	0.9960	0.9191	0.8383	0.7966	0.6644	0.5874	0.7530	0.8944	0.8675	0.8878	0.9041	0.7500	0.9964	0.8319	0.8307	0.9332	0.6536	0.3909	0.9316	0.9670	0.8496
0.0015	12.0	87864	0.0267	0.8344	0.8204	0.8273	0.9960	0.9270	0.8141	0.7881	0.6816	0.5730	0.7855	0.8964	0.8745	0.8926	0.8913	0.7805	0.9946	0.8291	0.8275	0.9332	0.6376	0.3753	0.9273	0.9761	0.8035
0.001	13.0	95186	0.0297	0.8316	0.8201	0.8258	0.9960	0.9339	0.8373	0.7351	0.6392	0.5955	0.7816	0.9022	0.8763	0.8968	0.8861	0.7826	0.9964	0.8408	0.8296	0.9223	0.6689	0.3906	0.9404	0.9762	0.8070
0.0007	14.0	102508	0.0317	0.8299	0.8211	0.8254	0.9959	0.9387	0.8462	0.7520	0.6820	0.5964	0.7791	0.9010	0.8770	0.8932	0.9039	0.8142	0.9964	0.8325	0.8262	0.9171	0.6637	0.3807	0.9316	0.9799	0.8450
0.0003	15.0	109830	0.0334	0.8340	0.8274	0.8307	0.9960	0.9368	0.8222	0.7744	0.6737	0.5977	0.7877	0.9053	0.8817	0.8745	0.9038	0.8083	0.9964	0.8345	0.8335	0.9311	0.6681	0.3793	0.9406	0.9762	0.8436
0.0001	16.0	117152	0.0348	0.8392	0.8292	0.8342	0.9961	0.9381	0.8202	0.7895	0.6704	0.5885	0.7783	0.9045	0.8854	0.8905	0.9056	0.8288	0.9964	0.8385	0.8264	0.9359	0.6659	0.3870	0.9430	0.9817	0.8348

Framework versions

Transformers 4.28.1
Pytorch 2.0.1
Datasets 2.12.0
Tokenizers 0.13.3

Task: Token Classification
Library: transformers
Type: bert
License: mit
Created: Jun 8, 2023
Updated: Aug 2, 2023