Selesai. Test Accuracy: 0.9989
Browse files- 1_Pooling/config.json +10 -0
- 2_Dense/config.json +1 -0
- 2_Dense/model.safetensors +3 -0
- README.md +476 -0
- config_sentence_transformers.json +10 -0
- modules.json +20 -0
- sentence_bert_config.json +4 -0
1_Pooling/config.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"word_embedding_dimension": 768,
|
| 3 |
+
"pooling_mode_cls_token": false,
|
| 4 |
+
"pooling_mode_mean_tokens": true,
|
| 5 |
+
"pooling_mode_max_tokens": false,
|
| 6 |
+
"pooling_mode_mean_sqrt_len_tokens": false,
|
| 7 |
+
"pooling_mode_weightedmean_tokens": false,
|
| 8 |
+
"pooling_mode_lasttoken": false,
|
| 9 |
+
"include_prompt": true
|
| 10 |
+
}
|
2_Dense/config.json
ADDED
|
@@ -0,0 +1 @@
|
|
|
|
|
|
|
| 1 |
+
{"in_features": 768, "out_features": 768, "bias": true, "activation_function": "torch.nn.modules.activation.Tanh"}
|
2_Dense/model.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:975c7e393f526ac10d78447acf06bc5bb2dad09e2db484d8ed4972b3472b382f
|
| 3 |
+
size 2362528
|
README.md
ADDED
|
@@ -0,0 +1,476 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
tags:
|
| 3 |
+
- sentence-transformers
|
| 4 |
+
- sentence-similarity
|
| 5 |
+
- feature-extraction
|
| 6 |
+
- generated_from_trainer
|
| 7 |
+
- dataset_size:6399
|
| 8 |
+
- loss:MultipleNegativesRankingLoss
|
| 9 |
+
base_model: LazarusNLP/congen-indobert-lite-base
|
| 10 |
+
widget:
|
| 11 |
+
- source_sentence: Apa yang dilakukan wisatawan?
|
| 12 |
+
sentences:
|
| 13 |
+
- The number of tourists visiting during the 2018 holiday reached 9,870 people in
|
| 14 |
+
one day. Every activity of tourists will produce waste in the tourist area, especially
|
| 15 |
+
organic wast e. Organic waste has good energy potential
|
| 16 |
+
- listrik yang dihasilkan dari proses gasifikasi yang memiliki nilai efisiensi 11%
|
| 17 |
+
adalah 6,38 kW atau 6.380 Watt. Resume perhitungan analisis potensi energi listrik
|
| 18 |
+
dari sampah organik yang siap diproses dapat dilihat pada Tabel 3. Tabel 3.
|
| 19 |
+
- Huruf e Cukup jelas. Huruf f Yang dimaksud dengan alat bukti lain, meliputi, informasi
|
| 20 |
+
yang diucapkan, dikirimkan, diterima, ata u disimpan secara elektronik, magnetik,
|
| 21 |
+
optik, dan/at au yang serupa dengan itu;
|
| 22 |
+
- source_sentence: er Aspek apa saja yang dinilai responden dalam kuisioner? Paraphrase
|
| 23 |
+
sentences:
|
| 24 |
+
- Ayat (2) Cukup jelas. Pasal 24 Cukup jelas. Pasal 25 Huruf a Cukup jelas. Huruf
|
| 25 |
+
b Cukup jelas. Huruf c Cukup jelas. Huruf d Cukup jelas.
|
| 26 |
+
- udara rata-rata adalah 300C. Desa ini berjarak 4 km dari pusat Kecamatan Kretek
|
| 27 |
+
dan 13 km dari ibukota kabupaten Bantul. Di lingkup wilayah Desa Parangtritis
|
| 28 |
+
ini daya tarik wisata utama yang
|
| 29 |
+
- Kuisioner yang dibagikan berisikan segala hal yang berkaitan dengan sistem pengelolaan
|
| 30 |
+
sampah serta penilaian responden terhadap sistem pengelolaan sampah (Peran dan
|
| 31 |
+
kinerja Dinas Kebersihan dan Pertamanan, Sarana dan prasarana,
|
| 32 |
+
- source_sentence: Apa itu UU ini?
|
| 33 |
+
sentences:
|
| 34 |
+
- Tambahan Lembaran Negara Republik Indonesia Nomor 3699) dicabut dan dinyatakan
|
| 35 |
+
tidak berlaku. Pasal 126 Peraturan pelaksanaan yang diamanatkan dalam Undang-Undang
|
| 36 |
+
ini ditetapkan paling lama 1 (satu) tahun terhitung sejak UndangUndang ini diberlakukan.
|
| 37 |
+
- penyelenggaraan usaha dan/atau kegiatan. 12. Upaya pengelolaan lingkungan hidup
|
| 38 |
+
dan upaya pemantauan lingkungan hidup, yang selanjutnya disebut UKL-UPL, adalah
|
| 39 |
+
pengelolaan dan pemantauan terhadap usaha dan/atau kegiatan yang tidak berdampak
|
| 40 |
+
penting terhadap lingkungan hidup yang diperlukan bagi proses pengambilan keputusan
|
| 41 |
+
tentang penyelenggaraan usaha dan/atau kegiatan.
|
| 42 |
+
- Abstra ct Parangtritis Beach is a tourist attraction that is visited by many tourists.
|
| 43 |
+
The number of tourists visiting during the 2018 holiday reached 9,870 people in
|
| 44 |
+
one day.
|
| 45 |
+
- source_sentence: Kapan izin lingkungan dapat dibatalkan? Bagaimana
|
| 46 |
+
sentences:
|
| 47 |
+
- Agar setiap orang mengetahuinya, memerintahkan pengundangan Undang-Undang ini
|
| 48 |
+
dengan penempatannya dalam Lembaran Negara Republik Indonesia. Disahkan di Jakarta
|
| 49 |
+
pada tanggal 3 Oktober 2009 PRESIDEN REPUBLIK INDONESIA, ttd DR. H.
|
| 50 |
+
- Yogyakarta dikenal sebagai kota pelajar dan kota wisata. Berdasarkan data di Dinas
|
| 51 |
+
Pariwisata Daerah Istimewa Yogyakarta ada b eberapa destinasi wisata di Yogyakarta
|
| 52 |
+
meliputi wisata alam, wisata pantai wisata budaya dan sejarah, wisata museum,
|
| 53 |
+
wisata minat khusus, dan desa wisata. Wisata Pantai di D.I.
|
| 54 |
+
- '(1) Menteri, gubernur, atau bupati/walikota sesuai dengan kewenangannya wajib
|
| 55 |
+
menolak permohonan izin lingkungan apabila permohonan izin tidak dilengkapi dengan
|
| 56 |
+
amdal atau UKL-UPL. (2) Izin - 27 - (2) Izin lingkungan sebagaimana dimaksud dalam
|
| 57 |
+
Pasal 36 ayat (4) dapat dibatalkan apabila: a.'
|
| 58 |
+
- source_sentence: 39641995 paraphrase Paraphrases Referensi
|
| 59 |
+
sentences:
|
| 60 |
+
- persampahan. Direktorat Jenderal Cipta Karya. Jakarta. Anonim. 1995. Metode pengambilan
|
| 61 |
+
dan pengukuran contoh timbulan dan komposisi sampah perkotaan (SNI 19-3964-1995).
|
| 62 |
+
Badan Standar Nasional. Jakarta.
|
| 63 |
+
- Sampah orga nik yang akan diproses sebanyak 1.400,36 kg per hari. Kemudian diproses
|
| 64 |
+
menjadi arang, sehingga didapatkan arang sampah organik sebanyak 205,91 kg per
|
| 65 |
+
hari. Berdasarkan perhitungan didapatkan potensi energi listrik yang dihasilkan
|
| 66 |
+
adalah 1.392,38 kWh
|
| 67 |
+
- 19. Perubahan iklim adalah berubahnya iklim yang diakibatkan langsung atau tidak
|
| 68 |
+
langsung oleh aktivitas manusia sehingga menyebabkan perubahan komposisi atmosfir
|
| 69 |
+
secara global dan selain itu juga berupa perubahan variabilitas iklim alamiah
|
| 70 |
+
yang teramati pada kurun waktu yang dapat dibandingkan.
|
| 71 |
+
pipeline_tag: sentence-similarity
|
| 72 |
+
library_name: sentence-transformers
|
| 73 |
+
metrics:
|
| 74 |
+
- cosine_accuracy
|
| 75 |
+
model-index:
|
| 76 |
+
- name: SentenceTransformer based on LazarusNLP/congen-indobert-lite-base
|
| 77 |
+
results:
|
| 78 |
+
- task:
|
| 79 |
+
type: triplet
|
| 80 |
+
name: Triplet
|
| 81 |
+
dataset:
|
| 82 |
+
name: retrieval validation
|
| 83 |
+
type: retrieval-validation
|
| 84 |
+
metrics:
|
| 85 |
+
- type: cosine_accuracy
|
| 86 |
+
value: 0.9961727857589722
|
| 87 |
+
name: Cosine Accuracy
|
| 88 |
+
- task:
|
| 89 |
+
type: triplet
|
| 90 |
+
name: Triplet
|
| 91 |
+
dataset:
|
| 92 |
+
name: test
|
| 93 |
+
type: test
|
| 94 |
+
metrics:
|
| 95 |
+
- type: cosine_accuracy
|
| 96 |
+
value: 0.9989070892333984
|
| 97 |
+
name: Cosine Accuracy
|
| 98 |
+
---
|
| 99 |
+
|
| 100 |
+
# SentenceTransformer based on LazarusNLP/congen-indobert-lite-base
|
| 101 |
+
|
| 102 |
+
This is a [sentence-transformers](https://www.SBERT.net) model finetuned from [LazarusNLP/congen-indobert-lite-base](https://huggingface.co/LazarusNLP/congen-indobert-lite-base). It maps sentences & paragraphs to a 768-dimensional dense vector space and can be used for semantic textual similarity, semantic search, paraphrase mining, text classification, clustering, and more.
|
| 103 |
+
|
| 104 |
+
## Model Details
|
| 105 |
+
|
| 106 |
+
### Model Description
|
| 107 |
+
- **Model Type:** Sentence Transformer
|
| 108 |
+
- **Base model:** [LazarusNLP/congen-indobert-lite-base](https://huggingface.co/LazarusNLP/congen-indobert-lite-base) <!-- at revision e1f1ad81d3c620b317077edfaa5d1ce1b07b464b -->
|
| 109 |
+
- **Maximum Sequence Length:** 32 tokens
|
| 110 |
+
- **Output Dimensionality:** 768 dimensions
|
| 111 |
+
- **Similarity Function:** Cosine Similarity
|
| 112 |
+
<!-- - **Training Dataset:** Unknown -->
|
| 113 |
+
<!-- - **Language:** Unknown -->
|
| 114 |
+
<!-- - **License:** Unknown -->
|
| 115 |
+
|
| 116 |
+
### Model Sources
|
| 117 |
+
|
| 118 |
+
- **Documentation:** [Sentence Transformers Documentation](https://sbert.net)
|
| 119 |
+
- **Repository:** [Sentence Transformers on GitHub](https://github.com/UKPLab/sentence-transformers)
|
| 120 |
+
- **Hugging Face:** [Sentence Transformers on Hugging Face](https://huggingface.co/models?library=sentence-transformers)
|
| 121 |
+
|
| 122 |
+
### Full Model Architecture
|
| 123 |
+
|
| 124 |
+
```
|
| 125 |
+
SentenceTransformer(
|
| 126 |
+
(0): Transformer({'max_seq_length': 32, 'do_lower_case': False}) with Transformer model: AlbertModel
|
| 127 |
+
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
|
| 128 |
+
(2): Dense({'in_features': 768, 'out_features': 768, 'bias': True, 'activation_function': 'torch.nn.modules.activation.Tanh'})
|
| 129 |
+
)
|
| 130 |
+
```
|
| 131 |
+
|
| 132 |
+
## Usage
|
| 133 |
+
|
| 134 |
+
### Direct Usage (Sentence Transformers)
|
| 135 |
+
|
| 136 |
+
First install the Sentence Transformers library:
|
| 137 |
+
|
| 138 |
+
```bash
|
| 139 |
+
pip install -U sentence-transformers
|
| 140 |
+
```
|
| 141 |
+
|
| 142 |
+
Then you can load this model and run inference.
|
| 143 |
+
```python
|
| 144 |
+
from sentence_transformers import SentenceTransformer
|
| 145 |
+
|
| 146 |
+
# Download from the 🤗 Hub
|
| 147 |
+
model = SentenceTransformer("yosriku/exp_data_scale_5files")
|
| 148 |
+
# Run inference
|
| 149 |
+
sentences = [
|
| 150 |
+
'39641995 paraphrase Paraphrases Referensi',
|
| 151 |
+
'persampahan. Direktorat Jenderal Cipta Karya. Jakarta. Anonim. 1995. Metode pengambilan dan pengukuran contoh timbulan dan komposisi sampah perkotaan (SNI 19-3964-1995). Badan Standar Nasional. Jakarta.',
|
| 152 |
+
'Sampah orga nik yang akan diproses sebanyak 1.400,36 kg per hari. Kemudian diproses menjadi arang, sehingga didapatkan arang sampah organik sebanyak 205,91 kg per hari. Berdasarkan perhitungan didapatkan potensi energi listrik yang dihasilkan adalah 1.392,38 kWh',
|
| 153 |
+
]
|
| 154 |
+
embeddings = model.encode(sentences)
|
| 155 |
+
print(embeddings.shape)
|
| 156 |
+
# [3, 768]
|
| 157 |
+
|
| 158 |
+
# Get the similarity scores for the embeddings
|
| 159 |
+
similarities = model.similarity(embeddings, embeddings)
|
| 160 |
+
print(similarities.shape)
|
| 161 |
+
# [3, 3]
|
| 162 |
+
```
|
| 163 |
+
|
| 164 |
+
<!--
|
| 165 |
+
### Direct Usage (Transformers)
|
| 166 |
+
|
| 167 |
+
<details><summary>Click to see the direct usage in Transformers</summary>
|
| 168 |
+
|
| 169 |
+
</details>
|
| 170 |
+
-->
|
| 171 |
+
|
| 172 |
+
<!--
|
| 173 |
+
### Downstream Usage (Sentence Transformers)
|
| 174 |
+
|
| 175 |
+
You can finetune this model on your own dataset.
|
| 176 |
+
|
| 177 |
+
<details><summary>Click to expand</summary>
|
| 178 |
+
|
| 179 |
+
</details>
|
| 180 |
+
-->
|
| 181 |
+
|
| 182 |
+
<!--
|
| 183 |
+
### Out-of-Scope Use
|
| 184 |
+
|
| 185 |
+
*List how the model may foreseeably be misused and address what users ought not to do with the model.*
|
| 186 |
+
-->
|
| 187 |
+
|
| 188 |
+
## Evaluation
|
| 189 |
+
|
| 190 |
+
### Metrics
|
| 191 |
+
|
| 192 |
+
#### Triplet
|
| 193 |
+
|
| 194 |
+
* Datasets: `retrieval-validation` and `test`
|
| 195 |
+
* Evaluated with [<code>TripletEvaluator</code>](https://sbert.net/docs/package_reference/sentence_transformer/evaluation.html#sentence_transformers.evaluation.TripletEvaluator)
|
| 196 |
+
|
| 197 |
+
| Metric | retrieval-validation | test |
|
| 198 |
+
|:--------------------|:---------------------|:-----------|
|
| 199 |
+
| **cosine_accuracy** | **0.9962** | **0.9989** |
|
| 200 |
+
|
| 201 |
+
<!--
|
| 202 |
+
## Bias, Risks and Limitations
|
| 203 |
+
|
| 204 |
+
*What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.*
|
| 205 |
+
-->
|
| 206 |
+
|
| 207 |
+
<!--
|
| 208 |
+
### Recommendations
|
| 209 |
+
|
| 210 |
+
*What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.*
|
| 211 |
+
-->
|
| 212 |
+
|
| 213 |
+
## Training Details
|
| 214 |
+
|
| 215 |
+
### Training Dataset
|
| 216 |
+
|
| 217 |
+
#### Unnamed Dataset
|
| 218 |
+
|
| 219 |
+
* Size: 6,399 training samples
|
| 220 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
| 221 |
+
* Approximate statistics based on the first 1000 samples:
|
| 222 |
+
| | anchor | positive | negative |
|
| 223 |
+
|:--------|:---------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
|
| 224 |
+
| type | string | string | string |
|
| 225 |
+
| details | <ul><li>min: 3 tokens</li><li>mean: 9.36 tokens</li><li>max: 31 tokens</li></ul> | <ul><li>min: 11 tokens</li><li>mean: 30.07 tokens</li><li>max: 32 tokens</li></ul> | <ul><li>min: 5 tokens</li><li>mean: 26.95 tokens</li><li>max: 32 tokens</li></ul> |
|
| 226 |
+
* Samples:
|
| 227 |
+
| anchor | positive | negative |
|
| 228 |
+
|:--------------------------------------------------------------------------|:------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
| 229 |
+
| <code>Bagaimana status UU 23 Tahun 1997?</code> | <code>yang baru berdasarkan Undang-Undang ini. Pasal - 70 - Pasal 125 Pada saat Undang-Undang ini mulai berlaku, Undang-Undang Nomor 23 Tahun 1997 tentang Pengelolaan Lingkungan Hidup (Lembaran Negara Republik Indonesia Tahun 1997 Nomor 68, Tambahan Lembaran Negara Republik Indonesia Nomor 3699) dicabut dan dinyatakan tidak berlaku.</code> | <code>Jumlah wisatawan pengunjung Pantai Parangtrit is yang mencapai 9.870 orang setiap hari adalah potensi yang besar untuk menghasilkan sampah. Sedangkan, setiap orang dalam 1 hari berpotensi menghasilkan sampah rata -rata 0,8 kg 3.</code> |
|
| 230 |
+
| <code>kedua Bagian Kedua Masuk ke Bagian Pertama Bagian kedua kata</code> | <code>(3) Gugatan melalui pengadilan hanya dapat ditempuh apabila upaya penyelesaian sengketa di luar pengadilan yang dipilih dinyatakan tidak berhasil oleh salah satu atau para pihak yang bersengketa. Bagian Kedua - 53 - Bagian Kedua Penyelesaian Sengketa Lingkungan Hidup di Luar Pengadilan Pasal 85 (1) Penyelesaian sengketa lingkungan hidup di luar pengadilan dilakukan untuk mencapai kesepakatan mengenai: a. bentuk dan besarnya ganti rugi; b.</code> | <code>31. Masyarakat hukum adat adalah kelompok masyarakat yang secara turun temurun bermukim di wilayah geografis tertentu karena adanya ikatan pada asal usul leluhur, adanya hubungan yang kuat dengan lingkungan hidup, serta adanya sistem nilai yang menentukan pranata ekonomi, politik, sosial, dan hukum</code> |
|
| 231 |
+
| <code>s Mengapa jumlah sarana yang banyak bisa percuma?</code> | <code>oleh penduduk setempat. Namun banyak atau tidaknya sarana dan prasarana pengelolaan sampah, jika tidak diikuti dengan kualitas yang baik dari sarana dan prasarana tersebut maka jumlah yang banyak tersebut akan percuma. Fungsi dari</code> | <code>udara rata-rata adalah 300C. Desa ini berjarak 4 km dari pusat Kecamatan Kretek dan 13 km dari ibukota kabupaten Bantul.</code> |
|
| 232 |
+
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
|
| 233 |
+
```json
|
| 234 |
+
{
|
| 235 |
+
"scale": 20.0,
|
| 236 |
+
"similarity_fct": "cos_sim"
|
| 237 |
+
}
|
| 238 |
+
```
|
| 239 |
+
|
| 240 |
+
### Evaluation Dataset
|
| 241 |
+
|
| 242 |
+
#### Unnamed Dataset
|
| 243 |
+
|
| 244 |
+
* Size: 1,829 evaluation samples
|
| 245 |
+
* Columns: <code>anchor</code>, <code>positive</code>, and <code>negative</code>
|
| 246 |
+
* Approximate statistics based on the first 1000 samples:
|
| 247 |
+
| | anchor | positive | negative |
|
| 248 |
+
|:--------|:---------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------|:----------------------------------------------------------------------------------|
|
| 249 |
+
| type | string | string | string |
|
| 250 |
+
| details | <ul><li>min: 3 tokens</li><li>mean: 9.15 tokens</li><li>max: 27 tokens</li></ul> | <ul><li>min: 11 tokens</li><li>mean: 30.03 tokens</li><li>max: 32 tokens</li></ul> | <ul><li>min: 6 tokens</li><li>mean: 27.14 tokens</li><li>max: 32 tokens</li></ul> |
|
| 251 |
+
* Samples:
|
| 252 |
+
| anchor | positive | negative |
|
| 253 |
+
|:----------------------------------------------------------|:----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|:-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------|
|
| 254 |
+
| <code>se Penjelasan Pasal 57 Ayat 4 Huruf b c</code> | <code>konsekuensi yang timbul akibat perubahan iklim dapat diatasi. Huruf b Cukup jelas. Huruf c Cukup jelas.</code> | <code>Jumlah pengunjung di Kawasan Wisata Pantai Parangtritis mencapai 9.870 orang/hari 1. Sedangkan, sampah yang dihasilkan oleh para wisatawan rata -rata 1,5 – 2,0 ton per hari pada hari biasa, dan bisa mencapai 20 ton sampah per hari pada saat liburan seperli libur lebaran 2.</code> |
|
| 255 |
+
| <code>Apa kewajiban usaha yang tidak wajib UKLUPL?</code> | <code>(2) Gubernur atau bupati/walikota menetapkan jenis usaha dan/atau kegiatan yang wajib dilengkapi dengan UKL-UPL. Pasal 35 (1) Usaha dan/atau kegiatan yang tidak wajib dilengkapi UKL-UPL sebagaimana dimaksud dalam Pasal 34 ayat (2) wajib membuat surat pernyataan kesanggupan pengelolaan dan pemantauan lingkungan hidup.</code> | <code>Abstra ct Parangtritis Beach is a tourist attraction that is visited by many tourists. The number of tourists visiting during the 2018 holiday reached 9,870 people in one day.</code> |
|
| 256 |
+
| <code>Siapa Tim Pelaksana?</code> | <code>Pasa l 8... Pasal 7 (1) Untuk membantu pelaksanaan tugas Tim Koordinasi Nasiona l, dibent uk Tim Pelaksana. (2) Susunan keanggotaan, tugas, dan tata kerja Tim Pelaksa na sebagaimana d imaksud pada ayat (1), ditetap kan oleh Menteri Koordinator Bidang Kemar itiman selaku Ketua Tim Koordinasi Nasional atas usulan Ketua Harian.</code> | <code>Bagaimana jika B3 telah kedaluwarsa?</code> |
|
| 257 |
+
* Loss: [<code>MultipleNegativesRankingLoss</code>](https://sbert.net/docs/package_reference/sentence_transformer/losses.html#multiplenegativesrankingloss) with these parameters:
|
| 258 |
+
```json
|
| 259 |
+
{
|
| 260 |
+
"scale": 20.0,
|
| 261 |
+
"similarity_fct": "cos_sim"
|
| 262 |
+
}
|
| 263 |
+
```
|
| 264 |
+
|
| 265 |
+
### Training Hyperparameters
|
| 266 |
+
#### Non-Default Hyperparameters
|
| 267 |
+
|
| 268 |
+
- `eval_strategy`: steps
|
| 269 |
+
- `per_device_train_batch_size`: 128
|
| 270 |
+
- `per_device_eval_batch_size`: 128
|
| 271 |
+
- `learning_rate`: 2e-05
|
| 272 |
+
- `warmup_ratio`: 0.1
|
| 273 |
+
- `fp16`: True
|
| 274 |
+
- `load_best_model_at_end`: True
|
| 275 |
+
- `push_to_hub`: True
|
| 276 |
+
- `hub_model_id`: yosriku/exp_data_scale_5files
|
| 277 |
+
- `hub_private_repo`: True
|
| 278 |
+
|
| 279 |
+
#### All Hyperparameters
|
| 280 |
+
<details><summary>Click to expand</summary>
|
| 281 |
+
|
| 282 |
+
- `overwrite_output_dir`: False
|
| 283 |
+
- `do_predict`: False
|
| 284 |
+
- `eval_strategy`: steps
|
| 285 |
+
- `prediction_loss_only`: True
|
| 286 |
+
- `per_device_train_batch_size`: 128
|
| 287 |
+
- `per_device_eval_batch_size`: 128
|
| 288 |
+
- `per_gpu_train_batch_size`: None
|
| 289 |
+
- `per_gpu_eval_batch_size`: None
|
| 290 |
+
- `gradient_accumulation_steps`: 1
|
| 291 |
+
- `eval_accumulation_steps`: None
|
| 292 |
+
- `torch_empty_cache_steps`: None
|
| 293 |
+
- `learning_rate`: 2e-05
|
| 294 |
+
- `weight_decay`: 0.0
|
| 295 |
+
- `adam_beta1`: 0.9
|
| 296 |
+
- `adam_beta2`: 0.999
|
| 297 |
+
- `adam_epsilon`: 1e-08
|
| 298 |
+
- `max_grad_norm`: 1.0
|
| 299 |
+
- `num_train_epochs`: 3
|
| 300 |
+
- `max_steps`: -1
|
| 301 |
+
- `lr_scheduler_type`: linear
|
| 302 |
+
- `lr_scheduler_kwargs`: {}
|
| 303 |
+
- `warmup_ratio`: 0.1
|
| 304 |
+
- `warmup_steps`: 0
|
| 305 |
+
- `log_level`: passive
|
| 306 |
+
- `log_level_replica`: warning
|
| 307 |
+
- `log_on_each_node`: True
|
| 308 |
+
- `logging_nan_inf_filter`: True
|
| 309 |
+
- `save_safetensors`: True
|
| 310 |
+
- `save_on_each_node`: False
|
| 311 |
+
- `save_only_model`: False
|
| 312 |
+
- `restore_callback_states_from_checkpoint`: False
|
| 313 |
+
- `no_cuda`: False
|
| 314 |
+
- `use_cpu`: False
|
| 315 |
+
- `use_mps_device`: False
|
| 316 |
+
- `seed`: 42
|
| 317 |
+
- `data_seed`: None
|
| 318 |
+
- `jit_mode_eval`: False
|
| 319 |
+
- `use_ipex`: False
|
| 320 |
+
- `bf16`: False
|
| 321 |
+
- `fp16`: True
|
| 322 |
+
- `fp16_opt_level`: O1
|
| 323 |
+
- `half_precision_backend`: auto
|
| 324 |
+
- `bf16_full_eval`: False
|
| 325 |
+
- `fp16_full_eval`: False
|
| 326 |
+
- `tf32`: None
|
| 327 |
+
- `local_rank`: 0
|
| 328 |
+
- `ddp_backend`: None
|
| 329 |
+
- `tpu_num_cores`: None
|
| 330 |
+
- `tpu_metrics_debug`: False
|
| 331 |
+
- `debug`: []
|
| 332 |
+
- `dataloader_drop_last`: False
|
| 333 |
+
- `dataloader_num_workers`: 0
|
| 334 |
+
- `dataloader_prefetch_factor`: None
|
| 335 |
+
- `past_index`: -1
|
| 336 |
+
- `disable_tqdm`: False
|
| 337 |
+
- `remove_unused_columns`: True
|
| 338 |
+
- `label_names`: None
|
| 339 |
+
- `load_best_model_at_end`: True
|
| 340 |
+
- `ignore_data_skip`: False
|
| 341 |
+
- `fsdp`: []
|
| 342 |
+
- `fsdp_min_num_params`: 0
|
| 343 |
+
- `fsdp_config`: {'min_num_params': 0, 'xla': False, 'xla_fsdp_v2': False, 'xla_fsdp_grad_ckpt': False}
|
| 344 |
+
- `fsdp_transformer_layer_cls_to_wrap`: None
|
| 345 |
+
- `accelerator_config`: {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}
|
| 346 |
+
- `deepspeed`: None
|
| 347 |
+
- `label_smoothing_factor`: 0.0
|
| 348 |
+
- `optim`: adamw_torch
|
| 349 |
+
- `optim_args`: None
|
| 350 |
+
- `adafactor`: False
|
| 351 |
+
- `group_by_length`: False
|
| 352 |
+
- `length_column_name`: length
|
| 353 |
+
- `ddp_find_unused_parameters`: None
|
| 354 |
+
- `ddp_bucket_cap_mb`: None
|
| 355 |
+
- `ddp_broadcast_buffers`: False
|
| 356 |
+
- `dataloader_pin_memory`: True
|
| 357 |
+
- `dataloader_persistent_workers`: False
|
| 358 |
+
- `skip_memory_metrics`: True
|
| 359 |
+
- `use_legacy_prediction_loop`: False
|
| 360 |
+
- `push_to_hub`: True
|
| 361 |
+
- `resume_from_checkpoint`: None
|
| 362 |
+
- `hub_model_id`: yosriku/exp_data_scale_5files
|
| 363 |
+
- `hub_strategy`: every_save
|
| 364 |
+
- `hub_private_repo`: True
|
| 365 |
+
- `hub_always_push`: False
|
| 366 |
+
- `hub_revision`: None
|
| 367 |
+
- `gradient_checkpointing`: False
|
| 368 |
+
- `gradient_checkpointing_kwargs`: None
|
| 369 |
+
- `include_inputs_for_metrics`: False
|
| 370 |
+
- `include_for_metrics`: []
|
| 371 |
+
- `eval_do_concat_batches`: True
|
| 372 |
+
- `fp16_backend`: auto
|
| 373 |
+
- `push_to_hub_model_id`: None
|
| 374 |
+
- `push_to_hub_organization`: None
|
| 375 |
+
- `mp_parameters`:
|
| 376 |
+
- `auto_find_batch_size`: False
|
| 377 |
+
- `full_determinism`: False
|
| 378 |
+
- `torchdynamo`: None
|
| 379 |
+
- `ray_scope`: last
|
| 380 |
+
- `ddp_timeout`: 1800
|
| 381 |
+
- `torch_compile`: False
|
| 382 |
+
- `torch_compile_backend`: None
|
| 383 |
+
- `torch_compile_mode`: None
|
| 384 |
+
- `include_tokens_per_second`: False
|
| 385 |
+
- `include_num_input_tokens_seen`: False
|
| 386 |
+
- `neftune_noise_alpha`: None
|
| 387 |
+
- `optim_target_modules`: None
|
| 388 |
+
- `batch_eval_metrics`: False
|
| 389 |
+
- `eval_on_start`: False
|
| 390 |
+
- `use_liger_kernel`: False
|
| 391 |
+
- `liger_kernel_config`: None
|
| 392 |
+
- `eval_use_gather_object`: False
|
| 393 |
+
- `average_tokens_across_devices`: False
|
| 394 |
+
- `prompts`: None
|
| 395 |
+
- `batch_sampler`: batch_sampler
|
| 396 |
+
- `multi_dataset_batch_sampler`: proportional
|
| 397 |
+
|
| 398 |
+
</details>
|
| 399 |
+
|
| 400 |
+
### Training Logs
|
| 401 |
+
| Epoch | Step | Validation Loss | retrieval-validation_cosine_accuracy | test_cosine_accuracy |
|
| 402 |
+
|:-------:|:------:|:---------------:|:------------------------------------:|:--------------------:|
|
| 403 |
+
| 0.2 | 5 | 4.3005 | 0.9809 | - |
|
| 404 |
+
| 0.4 | 10 | 3.8290 | 0.9880 | - |
|
| 405 |
+
| 0.6 | 15 | 3.5321 | 0.9902 | - |
|
| 406 |
+
| 0.8 | 20 | 3.3291 | 0.9923 | - |
|
| 407 |
+
| 1.0 | 25 | 3.1744 | 0.9940 | - |
|
| 408 |
+
| 1.2 | 30 | 3.0512 | 0.9940 | - |
|
| 409 |
+
| 1.4 | 35 | 2.9505 | 0.9940 | - |
|
| 410 |
+
| 1.6 | 40 | 2.8677 | 0.9951 | - |
|
| 411 |
+
| 1.8 | 45 | 2.8015 | 0.9956 | - |
|
| 412 |
+
| 2.0 | 50 | 2.7485 | 0.9951 | - |
|
| 413 |
+
| 2.2 | 55 | 2.7083 | 0.9956 | - |
|
| 414 |
+
| 2.4 | 60 | 2.6786 | 0.9956 | - |
|
| 415 |
+
| 2.6 | 65 | 2.6577 | 0.9956 | - |
|
| 416 |
+
| 2.8 | 70 | 2.6446 | 0.9962 | - |
|
| 417 |
+
| **3.0** | **75** | **2.6396** | **0.9962** | **-** |
|
| 418 |
+
| -1 | -1 | - | - | 0.9989 |
|
| 419 |
+
|
| 420 |
+
* The bold row denotes the saved checkpoint.
|
| 421 |
+
|
| 422 |
+
### Framework Versions
|
| 423 |
+
- Python: 3.11.13
|
| 424 |
+
- Sentence Transformers: 4.1.0
|
| 425 |
+
- Transformers: 4.53.3
|
| 426 |
+
- PyTorch: 2.6.0+cu124
|
| 427 |
+
- Accelerate: 1.9.0
|
| 428 |
+
- Datasets: 4.1.1
|
| 429 |
+
- Tokenizers: 0.21.2
|
| 430 |
+
|
| 431 |
+
## Citation
|
| 432 |
+
|
| 433 |
+
### BibTeX
|
| 434 |
+
|
| 435 |
+
#### Sentence Transformers
|
| 436 |
+
```bibtex
|
| 437 |
+
@inproceedings{reimers-2019-sentence-bert,
|
| 438 |
+
title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
|
| 439 |
+
author = "Reimers, Nils and Gurevych, Iryna",
|
| 440 |
+
booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
|
| 441 |
+
month = "11",
|
| 442 |
+
year = "2019",
|
| 443 |
+
publisher = "Association for Computational Linguistics",
|
| 444 |
+
url = "https://arxiv.org/abs/1908.10084",
|
| 445 |
+
}
|
| 446 |
+
```
|
| 447 |
+
|
| 448 |
+
#### MultipleNegativesRankingLoss
|
| 449 |
+
```bibtex
|
| 450 |
+
@misc{henderson2017efficient,
|
| 451 |
+
title={Efficient Natural Language Response Suggestion for Smart Reply},
|
| 452 |
+
author={Matthew Henderson and Rami Al-Rfou and Brian Strope and Yun-hsuan Sung and Laszlo Lukacs and Ruiqi Guo and Sanjiv Kumar and Balint Miklos and Ray Kurzweil},
|
| 453 |
+
year={2017},
|
| 454 |
+
eprint={1705.00652},
|
| 455 |
+
archivePrefix={arXiv},
|
| 456 |
+
primaryClass={cs.CL}
|
| 457 |
+
}
|
| 458 |
+
```
|
| 459 |
+
|
| 460 |
+
<!--
|
| 461 |
+
## Glossary
|
| 462 |
+
|
| 463 |
+
*Clearly define terms in order to be accessible across audiences.*
|
| 464 |
+
-->
|
| 465 |
+
|
| 466 |
+
<!--
|
| 467 |
+
## Model Card Authors
|
| 468 |
+
|
| 469 |
+
*Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.*
|
| 470 |
+
-->
|
| 471 |
+
|
| 472 |
+
<!--
|
| 473 |
+
## Model Card Contact
|
| 474 |
+
|
| 475 |
+
*Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.*
|
| 476 |
+
-->
|
config_sentence_transformers.json
ADDED
|
@@ -0,0 +1,10 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"__version__": {
|
| 3 |
+
"sentence_transformers": "4.1.0",
|
| 4 |
+
"transformers": "4.53.3",
|
| 5 |
+
"pytorch": "2.6.0+cu124"
|
| 6 |
+
},
|
| 7 |
+
"prompts": {},
|
| 8 |
+
"default_prompt_name": null,
|
| 9 |
+
"similarity_fn_name": "cosine"
|
| 10 |
+
}
|
modules.json
ADDED
|
@@ -0,0 +1,20 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
[
|
| 2 |
+
{
|
| 3 |
+
"idx": 0,
|
| 4 |
+
"name": "0",
|
| 5 |
+
"path": "",
|
| 6 |
+
"type": "sentence_transformers.models.Transformer"
|
| 7 |
+
},
|
| 8 |
+
{
|
| 9 |
+
"idx": 1,
|
| 10 |
+
"name": "1",
|
| 11 |
+
"path": "1_Pooling",
|
| 12 |
+
"type": "sentence_transformers.models.Pooling"
|
| 13 |
+
},
|
| 14 |
+
{
|
| 15 |
+
"idx": 2,
|
| 16 |
+
"name": "2",
|
| 17 |
+
"path": "2_Dense",
|
| 18 |
+
"type": "sentence_transformers.models.Dense"
|
| 19 |
+
}
|
| 20 |
+
]
|
sentence_bert_config.json
ADDED
|
@@ -0,0 +1,4 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"max_seq_length": 32,
|
| 3 |
+
"do_lower_case": false
|
| 4 |
+
}
|