Add BERTopic model
Browse files- README.md +135 -0
- config.json +15 -0
- topic_embeddings.safetensors +3 -0
- topics.json +0 -0
README.md
ADDED
|
@@ -0,0 +1,135 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
|
| 2 |
+
---
|
| 3 |
+
tags:
|
| 4 |
+
- bertopic
|
| 5 |
+
library_name: bertopic
|
| 6 |
+
pipeline_tag: text-classification
|
| 7 |
+
---
|
| 8 |
+
|
| 9 |
+
# bertopic_after_tuning
|
| 10 |
+
|
| 11 |
+
This is a [BERTopic](https://github.com/MaartenGr/BERTopic) model.
|
| 12 |
+
BERTopic is a flexible and modular topic modeling framework that allows for the generation of easily interpretable topics from large datasets.
|
| 13 |
+
|
| 14 |
+
## Usage
|
| 15 |
+
|
| 16 |
+
To use this model, please install BERTopic:
|
| 17 |
+
|
| 18 |
+
```
|
| 19 |
+
pip install -U bertopic
|
| 20 |
+
```
|
| 21 |
+
|
| 22 |
+
You can use the model as follows:
|
| 23 |
+
|
| 24 |
+
```python
|
| 25 |
+
from bertopic import BERTopic
|
| 26 |
+
topic_model = BERTopic.load("vidric/bertopic_after_tuning")
|
| 27 |
+
|
| 28 |
+
topic_model.get_topic_info()
|
| 29 |
+
```
|
| 30 |
+
|
| 31 |
+
## Topic overview
|
| 32 |
+
|
| 33 |
+
* Number of topics: 68
|
| 34 |
+
* Number of training documents: 20977
|
| 35 |
+
|
| 36 |
+
<details>
|
| 37 |
+
<summary>Click here for an overview of all topics.</summary>
|
| 38 |
+
|
| 39 |
+
| Topic ID | Topic Keywords | Topic Frequency | Label |
|
| 40 |
+
|----------|----------------|-----------------|-------|
|
| 41 |
+
| -1 | parfum - wangi - tidak - banget - tapi | 11 | -1_parfum_wangi_tidak_banget |
|
| 42 |
+
| 0 | wangi - mantap - enak - banget - suka | 6775 | 0_wangi_mantap_enak_banget |
|
| 43 |
+
| 1 | orgasm - eos - aku - lebih - farhampton | 10606 | 1_orgasm_eos_aku_lebih |
|
| 44 |
+
| 2 | terbaik - selalu - always - deal - puas | 886 | 2_terbaik_selalu_always_deal |
|
| 45 |
+
| 3 | admin - ramah - respon - cepat - kirim | 191 | 3_admin_ramah_respon_cepat |
|
| 46 |
+
| 4 | bubble - wrap - tebal - aman - kotak | 187 | 4_bubble_wrap_tebal_aman |
|
| 47 |
+
| 5 | alpha - theta - beli - tim - tetha | 154 | 5_alpha_theta_beli_tim |
|
| 48 |
+
| 6 | beli - menyesal - kali - lagi - ragu | 125 | 6_beli_menyesal_kali_lagi |
|
| 49 |
+
| 7 | kedua - beli - dua - kali - 2x | 109 | 7_kedua_beli_dua_kali |
|
| 50 |
+
| 8 | pernah - kecewa - tidak - selalu - hmns | 98 | 8_pernah_kecewa_tidak_selalu |
|
| 51 |
+
| 9 | farhampton - coba - beli - terimakasih - bonus | 86 | 9_farhampton_coba_beli_terimakasih |
|
| 52 |
+
| 10 | delta - theta - alpha - sama - suka | 85 | 10_delta_theta_alpha_sama |
|
| 53 |
+
| 11 | tidak - debat - ekspektasi - no - nget | 83 | 11_tidak_debat_ekspektasi_no |
|
| 54 |
+
| 12 | magnet - tutup - botol - ada - kemasan | 80 | 12_magnet_tutup_botol_ada |
|
| 55 |
+
| 13 | belum - unboxing - buka - tidak - bocor | 79 | 13_belum_unboxing_buka_tidak |
|
| 56 |
+
| 14 | indonesia - bangga - produk - buatan - terus | 73 | 14_indonesia_bangga_produk_buatan |
|
| 57 |
+
| 15 | repeat - order - kesekian - kali - selalu | 70 | 15_repeat_order_kesekian_kali |
|
| 58 |
+
| 16 | tester - test - sample - dapat - bonus | 67 | 16_tester_test_sample_dapat |
|
| 59 |
+
| 17 | ketiga - tiga - beli - kali - cny | 59 | 17_ketiga_tiga_beli_kali |
|
| 60 |
+
| 18 | selamat - barang - dengan - mendarat - sampai | 57 | 18_selamat_barang_dengan_mendarat |
|
| 61 |
+
| 19 | lokal - produk - internasional - brand - kualitas | 53 | 19_lokal_produk_internasional_brand |
|
| 62 |
+
| 20 | starterpacking - beli - starterpack - coba - varian | 44 | 20_starterpacking_beli_starterpack_coba |
|
| 63 |
+
| 21 | spray - macet - rusak - keras - berfungsi | 44 | 21_spray_macet_rusak_keras |
|
| 64 |
+
| 22 | longevity - jam - projection - notes - silage | 42 | 22_longevity_jam_projection_notes |
|
| 65 |
+
| 23 | kartu - card - greting - ucap - tulis | 39 | 23_kartu_card_greting_ucap |
|
| 66 |
+
| 24 | tahan - lama - harum - disinj - horny | 32 | 24_tahan_lama_harum_disinj |
|
| 67 |
+
| 25 | love - just - swet - cinta - lembut | 31 | 25_love_just_swet_cinta |
|
| 68 |
+
| 26 | rekomendasi - gela - ank - bangsa - karya | 31 | 26_rekomendasi_gela_ank_bangsa |
|
| 69 |
+
| 27 | proses - chat - lama - kirim - pengiriman | 31 | 27_proses_chat_lama_kirim |
|
| 70 |
+
| 28 | matahari - esence - sun - sari - ber | 30 | 28_matahari_esence_sun_sari |
|
| 71 |
+
| 29 | rekomendasi - sangat - super - layak - direkomendasi | 29 | 29_rekomendasi_sangat_super_layak |
|
| 72 |
+
| 30 | blind - beli - muantep - ekspektasi - sisi | 29 | 30_blind_beli_muantep_ekspektasi |
|
| 73 |
+
| 31 | istri - hadiah - menyukainya - suami - buat | 28 | 31_istri_hadiah_menyukainya_suami |
|
| 74 |
+
| 32 | harum - sue - hua - unik - tagih | 28 | 32_harum_sue_hua_unik |
|
| 75 |
+
| 33 | bicara - bintang - biar - alasanya - nyang | 28 | 33_bicara_bintang_biar_alasanya |
|
| 76 |
+
| 34 | worth - it - layak - lumayan - harga | 26 | 34_worth_it_layak_lumayan |
|
| 77 |
+
| 35 | atomizer - dapat - bonus - 10ml - praktis | 25 | 35_atomizer_dapat_bonus_10ml |
|
| 78 |
+
| 36 | hadiah - tahu - buka - belum - semoga | 25 | 36_hadiah_tahu_buka_belum |
|
| 79 |
+
| 37 | ragu - perlu - kecewa - tidak - rugi | 24 | 37_ragu_perlu_kecewa_tidak |
|
| 80 |
+
| 38 | bintang - 10 - lima - kasih - sebab | 23 | 38_bintang_10_lima_kasih |
|
| 81 |
+
| 39 | segar - harum - ambune - ciamik - memuaskan | 22 | 39_segar_harum_ambune_ciamik |
|
| 82 |
+
| 40 | tim - hmns - terimakasih - kami - telah | 21 | 40_tim_hmns_terimakasih_kami |
|
| 83 |
+
| 41 | twitter - review - instagram - tws - decants | 20 | 41_twitter_review_instagram_tws |
|
| 84 |
+
| 42 | pokok - mantap - sampai - cepat - suka | 20 | 42_pokok_mantap_sampai_cepat |
|
| 85 |
+
| 43 | travel - ukuran - kecil - botle - frhmptn | 20 | 43_travel_ukuran_kecil_botle |
|
| 86 |
+
| 44 | eos - sama - suka - adict - zodiak | 20 | 44_eos_sama_suka_adict |
|
| 87 |
+
| 45 | notes - base - printilan - midlle - note | 20 | 45_notes_base_printilan_midlle |
|
| 88 |
+
| 46 | bangga - lokal - pride - kualitas - bagus | 20 | 46_bangga_lokal_pride_kualitas |
|
| 89 |
+
| 47 | cowok - cocok - spesial - masculine - dimainkan | 19 | 47_cowok_cocok_spesial_masculine |
|
| 90 |
+
| 48 | rapi - packing - trusted - ramah - aman | 18 | 48_rapi_packing_trusted_ramah |
|
| 91 |
+
| 49 | sukses - terus - lapaknya - inovasi - selalu | 18 | 49_sukses_terus_lapaknya_inovasi |
|
| 92 |
+
| 50 | mantan - ayah - abang - dia - masjid | 18 | 50_mantan_ayah_abang_dia |
|
| 93 |
+
| 51 | first - wekend - again - dicuekin - tunangan | 18 | 51_first_wekend_again_dicuekin |
|
| 94 |
+
| 52 | paket - mudik - maf - buka - selamat | 17 | 52_paket_mudik_maf_buka |
|
| 95 |
+
| 53 | bonus - dapat - gantengs - dengen - nembak | 15 | 53_bonus_dapat_gantengs_dengen |
|
| 96 |
+
| 54 | premium - packinya - 30k - perbedan - kemasan | 14 | 54_premium_packinya_30k_perbedan |
|
| 97 |
+
| 55 | obat - ada - tidak - bets - develop | 14 | 55_obat_ada_tidak_bets |
|
| 98 |
+
| 56 | projection - kurang - menyerang - meter - palagi | 14 | 56_projection_kurang_menyerang_meter |
|
| 99 |
+
| 57 | coconut - kelapa - pasion - vs - panas | 14 | 57_coconut_kelapa_pasion_vs |
|
| 100 |
+
| 58 | cewek - pacar - klepek - senenk - comunikasi | 14 | 58_cewek_pacar_klepek_senenk |
|
| 101 |
+
| 59 | pertama - mengoda - kesan - kali - pengatur | 13 | 59_pertama_mengoda_kesan_kali |
|
| 102 |
+
| 60 | ekspedisi - lambat - expedisi - hub - langsung | 13 | 60_ekspedisi_lambat_expedisi_hub |
|
| 103 |
+
| 61 | tunggu - akhirnya - datang - ye - uh | 13 | 61_tunggu_akhirnya_datang_ye |
|
| 104 |
+
| 62 | produk - dikolom - tanya - segini - kabari | 13 | 62_produk_dikolom_tanya_segini |
|
| 105 |
+
| 63 | jiwa - mantap - kesegaran - - | 12 | 63_jiwa_mantap_kesegaran_ |
|
| 106 |
+
| 64 | darker - warna - hitam - kuning - dof | 12 | 64_darker_warna_hitam_kuning |
|
| 107 |
+
| 65 | mewah - limo - seksi - lux - harum | 11 | 65_mewah_limo_seksi_lux |
|
| 108 |
+
| 66 | batch - oker - kelewatan - final - bagi | 11 | 66_batch_oker_kelewatan_final |
|
| 109 |
+
|
| 110 |
+
</details>
|
| 111 |
+
|
| 112 |
+
## Training hyperparameters
|
| 113 |
+
|
| 114 |
+
* calculate_probabilities: False
|
| 115 |
+
* language: indonesian
|
| 116 |
+
* low_memory: False
|
| 117 |
+
* min_topic_size: 10
|
| 118 |
+
* n_gram_range: (1, 1)
|
| 119 |
+
* nr_topics: auto
|
| 120 |
+
* seed_topic_list: None
|
| 121 |
+
* top_n_words: 10
|
| 122 |
+
* verbose: False
|
| 123 |
+
|
| 124 |
+
## Framework versions
|
| 125 |
+
|
| 126 |
+
* Numpy: 1.23.5
|
| 127 |
+
* HDBSCAN: 0.8.29
|
| 128 |
+
* UMAP: 0.5.3
|
| 129 |
+
* Pandas: 1.5.3
|
| 130 |
+
* Scikit-Learn: 1.2.1
|
| 131 |
+
* Sentence-transformers: 2.2.2
|
| 132 |
+
* Transformers: 4.24.0
|
| 133 |
+
* Numba: 0.56.4
|
| 134 |
+
* Plotly: 5.9.0
|
| 135 |
+
* Python: 3.10.9
|
config.json
ADDED
|
@@ -0,0 +1,15 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
{
|
| 2 |
+
"calculate_probabilities": false,
|
| 3 |
+
"language": "indonesian",
|
| 4 |
+
"low_memory": false,
|
| 5 |
+
"min_topic_size": 10,
|
| 6 |
+
"n_gram_range": [
|
| 7 |
+
1,
|
| 8 |
+
1
|
| 9 |
+
],
|
| 10 |
+
"nr_topics": "auto",
|
| 11 |
+
"seed_topic_list": null,
|
| 12 |
+
"top_n_words": 10,
|
| 13 |
+
"verbose": false,
|
| 14 |
+
"embedding_model": "sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2"
|
| 15 |
+
}
|
topic_embeddings.safetensors
ADDED
|
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
version https://git-lfs.github.com/spec/v1
|
| 2 |
+
oid sha256:8cda72d98128beee7889013059fe04c0f2e905bf458d06bb53f44b0d590ec346
|
| 3 |
+
size 104536
|
topics.json
ADDED
|
The diff for this file is too large to render.
See raw diff
|
|
|