mini1013's picture
Push model using huggingface_hub.
a74f129 verified
metadata
base_model: mini1013/master_domain
library_name: setfit
metrics:
  - accuracy
pipeline_tag: text-classification
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      1분완성 네일팁 모음인조손톱 인조팁 붙이는네일팁 웨딩네 13)샤인네일팁-화이트 LotteOn > 뷰티 > 네일 > 네일스티커/네일팁
      LotteOn > 뷰티 > 네일 > 네일스티커/네일팁
  - text: >-
      오피아이 인피니트샤인2 매니큐어 MI12 × 1개 (#M)쿠팡 홈>뷰티>네일>일반네일>컬러 매니큐어 Coupang > 뷰티 > 네일
      > 일반네일 > 컬러 매니큐어
  - text: >-
      오피아이 젤 네일 컬러 GCV33 x 1개 (#M)쿠팡 홈>뷰티>네일>일반네일>컬러 매니큐어 Coupang > 뷰티 > 네일 >
      일반네일 > 컬러 매니큐어
  - text: 디올 베르니 212 튀튀 LotteOn > 뷰티 > 메이크업 > 메이크업세트 LotteOn > 뷰티 > 메이크업 > 메이크업세트
  - text: >-
      OPI 인피니트샤인 HRL31 LETS BE FRIENDS HRL31 - LETS BE FRIENDS! LotteOn > 뷰티 >
      헤어/바디 > 헤어스타일링 > 염색/매니큐어 LotteOn > 뷰티 > 헤어/바디 > 헤어스타일링 > 염색/매니큐어
inference: true
model-index:
  - name: SetFit with mini1013/master_domain
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 0.5301810865191147
            name: Accuracy

SetFit with mini1013/master_domain

This is a SetFit model that can be used for Text Classification. This SetFit model uses mini1013/master_domain as the Sentence Transformer embedding model. A LogisticRegression instance is used for classification.

The model has been trained using an efficient few-shot learning technique that involves:

  1. Fine-tuning a Sentence Transformer with contrastive learning.
  2. Training a classification head with features from the fine-tuned Sentence Transformer.

Model Details

Model Description

Model Sources

Model Labels

Label Examples
3
  • '네일팁 실크익스텐션 311160L1720771597 티타늄금 물방울 (풀값 ) LotteOn > 뷰티 > 네일케어 > 네일케어도구 > 손톱깎이 LotteOn > 뷰티 > 네일케어 > 네일케어도구 > 손톱깎이'
  • '엔비베베 어린이 화장품 선물세트 어린이 썬쿠션+키즈네일스티커+워시패드 1개 (#M)쿠팡 홈>뷰티>어린이화장품>세트/키트 Coupang > 뷰티 > 어린이화장품 > 세트/키트'
  • '래쉬톡 원터치 인조 속눈썹 섹시 걸 × 3개입 LotteOn > 뷰티 > 뷰티기기/소품 > 아이/브로우소품 > 속눈썹관리 LotteOn > 뷰티 > 뷰티기기/소품 > 아이/브로우소품 > 속눈썹관리'
0
  • '오피아이 넌아세톤 리무버 빨강 30ml × 5개 (#M)쿠팡 홈>뷰티>네일>일반네일>리무버 Coupang > 뷰티 > 네일 > 일반네일 > 리무버'
  • '[OPI][리무버] 넌아세톤리무버 30ml ssg > 뷰티 > 메이크업 > 네일 ssg > 뷰티 > 메이크업 > 네일'
  • '포먼트 젤네일 O.4 블러쉬 뷰티 × 1개 (#M)쿠팡 홈>뷰티>네일>젤네일>컬러 젤 Coupang > 뷰티 > 네일 > 젤네일 > 컬러 젤'
2
  • '오피아이 프로스파 오일투고 큐티클 오일2197877 1 7.5ml x 1개2197877 1 (#M)SSG.COM/메이크업/베이스메이크업/컨실러 ssg > 뷰티 > 메이크업 > 베이스메이크업 > 컨실러'
  • '구찌 뷰티 [구찌] 베르니 아 옹글 하이 샤인 네일 라커 712 멜린다 그린 × 선택완료 (#M)쿠팡 홈>뷰티>네일>일반네일>컬러 매니큐어 Coupang > 뷰티 > 네일 > 일반네일 > 컬러 매니큐어'
  • 'OPI ProSpa 각질 제거 큐티클 크림, 27ml SSG.COM/메이크업/베이스메이크업/메이크업베이스;ssg > 뷰티 > 메이크업 > 베이스메이크업 > 메이크업베이스 ssg > 뷰티 > 메이크업 > 베이스메이크업 > 메이크업베이스'
1
  • '르 베르니 루쥬 느와르 DepartmentLotteOn > 뷰티 > 헤어/바디 > 핸드/풋케어 > 네일케어 DepartmentLotteOn > 뷰티 > 헤어/바디 > 핸드/풋케어 > 네일케어'
  • '베씨 베이스젤 + 탑젤 + 지브라파일 2p 세트 베이스젤, 탑젤, 지브라파일(100/150) × 1세트 LotteOn > 뷰티 > 네일 > 네일아트소품 LotteOn > 뷰티 > 네일 > 네일아트소품'
  • 'OPI OPI Chrome Effects Nail Lacquer Top Coat CPT31 - 0.5 oz 상세내용참조 × 상세내용참조 (#M)쿠팡 홈>뷰티>메이크업>베이스 메이크업>베이스/프라이머 Coupang > 뷰티 > 메이크업 > 베이스 메이크업 > 베이스/프라이머'

Evaluation

Metrics

Label Accuracy
all 0.5302

Uses

Direct Use for Inference

First install the SetFit library:

pip install setfit

Then you can load this model and run inference.

from setfit import SetFitModel

# Download from the 🤗 Hub
model = SetFitModel.from_pretrained("mini1013/master_cate_bt1_test_flat_top_cate")
# Run inference
preds = model("디올 베르니 212 튀튀 LotteOn > 뷰티 > 메이크업 > 메이크업세트 LotteOn > 뷰티 > 메이크업 > 메이크업세트")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 13 22.7236 41
Label Training Sample Count
0 49
1 50
2 50
3 50

Training Hyperparameters

  • batch_size: (64, 64)
  • num_epochs: (30, 30)
  • max_steps: -1
  • sampling_strategy: oversampling
  • num_iterations: 100
  • body_learning_rate: (2e-05, 1e-05)
  • head_learning_rate: 0.01
  • loss: CosineSimilarityLoss
  • distance_metric: cosine_distance
  • margin: 0.25
  • end_to_end: False
  • use_amp: False
  • warmup_proportion: 0.1
  • l2_weight: 0.01
  • seed: 42
  • eval_max_steps: -1
  • load_best_model_at_end: False

Training Results

Epoch Step Training Loss Validation Loss
0.0032 1 0.4603 -
0.1608 50 0.4502 -
0.3215 100 0.4315 -
0.4823 150 0.3996 -
0.6431 200 0.365 -
0.8039 250 0.2954 -
0.9646 300 0.2647 -
1.1254 350 0.2378 -
1.2862 400 0.2257 -
1.4469 450 0.2165 -
1.6077 500 0.213 -
1.7685 550 0.1999 -
1.9293 600 0.1838 -
2.0900 650 0.1614 -
2.2508 700 0.1164 -
2.4116 750 0.0553 -
2.5723 800 0.0366 -
2.7331 850 0.0279 -
2.8939 900 0.0219 -
3.0547 950 0.0166 -
3.2154 1000 0.0111 -
3.3762 1050 0.0067 -
3.5370 1100 0.0084 -
3.6977 1150 0.0066 -
3.8585 1200 0.0048 -
4.0193 1250 0.0028 -
4.1801 1300 0.0005 -
4.3408 1350 0.0003 -
4.5016 1400 0.0004 -
4.6624 1450 0.0001 -
4.8232 1500 0.0001 -
4.9839 1550 0.0001 -
5.1447 1600 0.0001 -
5.3055 1650 0.0001 -
5.4662 1700 0.0002 -
5.6270 1750 0.0 -
5.7878 1800 0.0 -
5.9486 1850 0.0 -
6.1093 1900 0.0001 -
6.2701 1950 0.0 -
6.4309 2000 0.0 -
6.5916 2050 0.0 -
6.7524 2100 0.0 -
6.9132 2150 0.0002 -
7.0740 2200 0.0002 -
7.2347 2250 0.0 -
7.3955 2300 0.0 -
7.5563 2350 0.0 -
7.7170 2400 0.0 -
7.8778 2450 0.0 -
8.0386 2500 0.0 -
8.1994 2550 0.0 -
8.3601 2600 0.0 -
8.5209 2650 0.0 -
8.6817 2700 0.0 -
8.8424 2750 0.0 -
9.0032 2800 0.0 -
9.1640 2850 0.0 -
9.3248 2900 0.0 -
9.4855 2950 0.0 -
9.6463 3000 0.0 -
9.8071 3050 0.0 -
9.9678 3100 0.0 -
10.1286 3150 0.0 -
10.2894 3200 0.0 -
10.4502 3250 0.0 -
10.6109 3300 0.0 -
10.7717 3350 0.0 -
10.9325 3400 0.0 -
11.0932 3450 0.0 -
11.2540 3500 0.0 -
11.4148 3550 0.0 -
11.5756 3600 0.0 -
11.7363 3650 0.0 -
11.8971 3700 0.0 -
12.0579 3750 0.0004 -
12.2186 3800 0.0 -
12.3794 3850 0.0001 -
12.5402 3900 0.0001 -
12.7010 3950 0.0 -
12.8617 4000 0.0001 -
13.0225 4050 0.0002 -
13.1833 4100 0.0009 -
13.3441 4150 0.0037 -
13.5048 4200 0.0025 -
13.6656 4250 0.0009 -
13.8264 4300 0.0002 -
13.9871 4350 0.0002 -
14.1479 4400 0.0 -
14.3087 4450 0.0002 -
14.4695 4500 0.0001 -
14.6302 4550 0.0004 -
14.7910 4600 0.0008 -
14.9518 4650 0.0 -
15.1125 4700 0.0 -
15.2733 4750 0.0001 -
15.4341 4800 0.0 -
15.5949 4850 0.0 -
15.7556 4900 0.0002 -
15.9164 4950 0.0 -
16.0772 5000 0.0 -
16.2379 5050 0.0001 -
16.3987 5100 0.0 -
16.5595 5150 0.0 -
16.7203 5200 0.0 -
16.8810 5250 0.0 -
17.0418 5300 0.0 -
17.2026 5350 0.0 -
17.3633 5400 0.0 -
17.5241 5450 0.0 -
17.6849 5500 0.0 -
17.8457 5550 0.0 -
18.0064 5600 0.0 -
18.1672 5650 0.0 -
18.3280 5700 0.0 -
18.4887 5750 0.0 -
18.6495 5800 0.0 -
18.8103 5850 0.0 -
18.9711 5900 0.0 -
19.1318 5950 0.0 -
19.2926 6000 0.0 -
19.4534 6050 0.0 -
19.6141 6100 0.0 -
19.7749 6150 0.0 -
19.9357 6200 0.0 -
20.0965 6250 0.0 -
20.2572 6300 0.0 -
20.4180 6350 0.0 -
20.5788 6400 0.0 -
20.7395 6450 0.0 -
20.9003 6500 0.0 -
21.0611 6550 0.0 -
21.2219 6600 0.0 -
21.3826 6650 0.0 -
21.5434 6700 0.0 -
21.7042 6750 0.0 -
21.8650 6800 0.0 -
22.0257 6850 0.0 -
22.1865 6900 0.0 -
22.3473 6950 0.0 -
22.5080 7000 0.0 -
22.6688 7050 0.0 -
22.8296 7100 0.0 -
22.9904 7150 0.0 -
23.1511 7200 0.0 -
23.3119 7250 0.0 -
23.4727 7300 0.0 -
23.6334 7350 0.0 -
23.7942 7400 0.0 -
23.9550 7450 0.0 -
24.1158 7500 0.0 -
24.2765 7550 0.0 -
24.4373 7600 0.0 -
24.5981 7650 0.0 -
24.7588 7700 0.0 -
24.9196 7750 0.0 -
25.0804 7800 0.0 -
25.2412 7850 0.0 -
25.4019 7900 0.0 -
25.5627 7950 0.0 -
25.7235 8000 0.0 -
25.8842 8050 0.0 -
26.0450 8100 0.0 -
26.2058 8150 0.0 -
26.3666 8200 0.0 -
26.5273 8250 0.0 -
26.6881 8300 0.0 -
26.8489 8350 0.0 -
27.0096 8400 0.0 -
27.1704 8450 0.0 -
27.3312 8500 0.0 -
27.4920 8550 0.0 -
27.6527 8600 0.0 -
27.8135 8650 0.0 -
27.9743 8700 0.0 -
28.1350 8750 0.0 -
28.2958 8800 0.0 -
28.4566 8850 0.0 -
28.6174 8900 0.0 -
28.7781 8950 0.0 -
28.9389 9000 0.0 -
29.0997 9050 0.0 -
29.2605 9100 0.0 -
29.4212 9150 0.0 -
29.5820 9200 0.0 -
29.7428 9250 0.0 -
29.9035 9300 0.0 -

Framework Versions

  • Python: 3.10.12
  • SetFit: 1.1.0
  • Sentence Transformers: 3.3.1
  • Transformers: 4.44.2
  • PyTorch: 2.2.0a0+81ea7a4
  • Datasets: 3.2.0
  • Tokenizers: 0.19.1

Citation

BibTeX

@article{https://doi.org/10.48550/arxiv.2209.11055,
    doi = {10.48550/ARXIV.2209.11055},
    url = {https://arxiv.org/abs/2209.11055},
    author = {Tunstall, Lewis and Reimers, Nils and Jo, Unso Eun Seo and Bates, Luke and Korat, Daniel and Wasserblat, Moshe and Pereg, Oren},
    keywords = {Computation and Language (cs.CL), FOS: Computer and information sciences, FOS: Computer and information sciences},
    title = {Efficient Few-Shot Learning Without Prompts},
    publisher = {arXiv},
    year = {2022},
    copyright = {Creative Commons Attribution 4.0 International}
}