Hugging Face's logo

mini1013
/
master_item_top_bt3

Text Classification
setfit
Safetensors
sentence-transformers
roberta
generated_from_setfit_trainer
Eval Results
Model card Files Community

SetFit with klue/roberta-base

This is a SetFit model that can be used for Text Classification. This SetFit model uses klue/roberta-base 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 Type: SetFit
  • Sentence Transformer body: klue/roberta-base
  • Classification head: a LogisticRegression instance
  • Maximum Sequence Length: 512 tokens
  • Number of Classes: 4 classes

Model Sources

Model Labels

Label Examples
3
  • '차앤박 CNP 안티포어 블랙헤드 클리어 키트 스트립 3세트(3회분) (#M)위메프 > 뷰티 > 스킨케어 > 팩/마스크 > 코팩 위메프 > 뷰티 > 스킨케어 > 팩/마스크 > 코팩'
  • '미팩토리 3단 돼지코팩 10개입 × 3개 (#M)쿠팡 홈>뷰티>스킨케어>마스크/팩>패치/코팩>코팩 Coupang > 뷰티 > 스킨케어 > 마스크/팩'
  • '[차앤박] CNP 안티포어 블랙헤드 버블 코팩 1매 / 넓은 모공 피부 / (#M)화장품/미용>마스크/팩>코팩 Naverstore > 화장품/미용 > 마스크/팩 > 코팩'
0
  • '메디힐×마리끌레르 기획전 앰플/크림/마스크팩~58% 25_메디힐 티트리 케어솔루션 에센셜마스크 [10매] 쇼킹딜 홈>뷰티>클렌징/팩/마스크>팩/마스크;11st>스킨케어>팩/마스크>마스크시트팩;(#M)11st>뷰티>클렌징/팩/마스크>팩/마스크 11st Hour Event > 패션/뷰티 > 뷰티 > 클렌징/팩/마스크 > 팩/마스크'
  • '[의료기기] 듀오덤 스팟패치 72매 [의료기기] 듀오덤 스팟패치 72매 (#M)홈>구강/건강용품>패치/겔>스팟패치 OLIVEYOUNG > 베스트 > 구강/건강용품'
  • '이지덤 뷰티 릴리프 스팟패치 57개입 3개 (#M)쿠팡 홈>생활용품>건강/의료용품>의약외품/상비용품>반창고/밴드 Coupang > 뷰티 > 스킨케어 > 마스크/팩 > 패치/코팩 > 스팟패치'
2
  • '안스킨 클래리파잉 골드 모델링 팩 1000ml 20개 (#M)홈>화장품/미용>마스크/팩>필오프팩 Naverstore > 화장품/미용 > 마스크/팩 > 필오프팩'
  • '[러쉬]오티픽스 75g - 프레쉬 페이스 마스크/마스크 팩 ssg > 뷰티 > 스킨케어 > 마스크/팩 > 시트마스크;ssg > 뷰티 > 헤어/바디 > 세정/입욕용품 > 입욕제/버블바스;ssg > 뷰티 > 스킨케어 > 마스크/팩;ssg > 뷰티 > 스킨케어 > 클렌징 ssg > 뷰티 > 스킨케어 > 마스크/팩 > 시트마스크'
  • '푸드어홀릭 콜라겐 필오프팩 150ml / 다시마 MinSellAmount (#M)화장품/향수>팩/마스크>필오프팩 Gmarket > 뷰티 > 화장품/향수 > 팩/마스크 > 필오프팩'
1
  • '물광 콜라겐 크림 티르티르 80ml 생크림 도자기 피부 물광마스크 이유빈 콜라겐물광마스크40ml (#M)홈>전체상품 Naverstore > 화장품/미용 > 남성화장품 > 크림'
  • '립 슬리핑 마스크 EX 20g 4종 베리 자몽 민트초코 애플라임 베리 (#M)홈>화장품/미용>마스크/팩>수면팩 Naverstore > 화장품/미용 > 마스크/팩 > 수면팩'
  • '설화수 한방 슬리핑마스크 나이트여운팩 120ml 1개 (#M)위메프 > 뷰티 > 스킨케어 > 팩/마스크 > 수면팩 위메프 > 뷰티 > 스킨케어 > 팩/마스크 > 수면팩'

Evaluation

Metrics

Label Accuracy
all 0.7775

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_item_top_bt3")
# Run inference
preds = model("[대용량] 라네즈 크림 스킨 퀵 스킨 팩 100매(140ml) 피부진정 보습  (#M)홈>라네즈 Naverstore > 화장품/미용 > 마스크/팩 > 수면팩")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 11 21.75 91
Label Training Sample Count
0 50
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 0.4549 -
0.1597 50 0.3933 -
0.3195 100 0.3669 -
0.4792 150 0.2841 -
0.6390 200 0.1163 -
0.7987 250 0.0104 -
0.9585 300 0.0072 -
1.1182 350 0.0065 -
1.2780 400 0.0059 -
1.4377 450 0.0058 -
1.5974 500 0.0035 -
1.7572 550 0.0032 -
1.9169 600 0.0032 -
2.0767 650 0.0025 -
2.2364 700 0.0023 -
2.3962 750 0.0023 -
2.5559 800 0.0025 -
2.7157 850 0.0023 -
2.8754 900 0.003 -
3.0351 950 0.0026 -
3.1949 1000 0.0043 -
3.3546 1050 0.0022 -
3.5144 1100 0.0024 -
3.6741 1150 0.0025 -
3.8339 1200 0.0025 -
3.9936 1250 0.0024 -
4.1534 1300 0.0025 -
4.3131 1350 0.0025 -
4.4728 1400 0.0027 -
4.6326 1450 0.0023 -
4.7923 1500 0.0022 -
4.9521 1550 0.0026 -
5.1118 1600 0.0022 -
5.2716 1650 0.0027 -
5.4313 1700 0.0022 -
5.5911 1750 0.0024 -
5.7508 1800 0.0029 -
5.9105 1850 0.0018 -
6.0703 1900 0.0033 -
6.2300 1950 0.002 -
6.3898 2000 0.0027 -
6.5495 2050 0.0021 -
6.7093 2100 0.0022 -
6.8690 2150 0.0023 -
7.0288 2200 0.0026 -
7.1885 2250 0.0018 -
7.3482 2300 0.0024 -
7.5080 2350 0.002 -
7.6677 2400 0.0027 -
7.8275 2450 0.0022 -
7.9872 2500 0.0032 -
8.1470 2550 0.0029 -
8.3067 2600 0.0025 -
8.4665 2650 0.0017 -
8.6262 2700 0.0026 -
8.7859 2750 0.0023 -
8.9457 2800 0.0023 -
9.1054 2850 0.0029 -
9.2652 2900 0.0028 -
9.4249 2950 0.0021 -
9.5847 3000 0.0027 -
9.7444 3050 0.0019 -
9.9042 3100 0.0022 -
10.0639 3150 0.003 -
10.2236 3200 0.0024 -
10.3834 3250 0.0019 -
10.5431 3300 0.0023 -
10.7029 3350 0.0024 -
10.8626 3400 0.0026 -
11.0224 3450 0.0025 -
11.1821 3500 0.0022 -
11.3419 3550 0.0023 -
11.5016 3600 0.0027 -
11.6613 3650 0.0032 -
11.8211 3700 0.0022 -
11.9808 3750 0.0019 -
12.1406 3800 0.0029 -
12.3003 3850 0.0026 -
12.4601 3900 0.0027 -
12.6198 3950 0.0019 -
12.7796 4000 0.0021 -
12.9393 4050 0.0023 -
13.0990 4100 0.0027 -
13.2588 4150 0.0021 -
13.4185 4200 0.0022 -
13.5783 4250 0.0026 -
13.7380 4300 0.0025 -
13.8978 4350 0.0025 -
14.0575 4400 0.0021 -
14.2173 4450 0.0031 -
14.3770 4500 0.0022 -
14.5367 4550 0.0016 -
14.6965 4600 0.0027 -
14.8562 4650 0.0027 -
15.0160 4700 0.0027 -
15.1757 4750 0.0021 -
15.3355 4800 0.0027 -
15.4952 4850 0.0031 -
15.6550 4900 0.0021 -
15.8147 4950 0.0023 -
15.9744 5000 0.002 -
16.1342 5050 0.0024 -
16.2939 5100 0.0026 -
16.4537 5150 0.002 -
16.6134 5200 0.0026 -
16.7732 5250 0.0029 -
16.9329 5300 0.0023 -
17.0927 5350 0.0022 -
17.2524 5400 0.0028 -
17.4121 5450 0.0026 -
17.5719 5500 0.0017 -
17.7316 5550 0.0032 -
17.8914 5600 0.0022 -
18.0511 5650 0.0019 -
18.2109 5700 0.0024 -
18.3706 5750 0.0026 -
18.5304 5800 0.0031 -
18.6901 5850 0.0024 -
18.8498 5900 0.0018 -
19.0096 5950 0.0023 -
19.1693 6000 0.0025 -
19.3291 6050 0.0028 -
19.4888 6100 0.002 -
19.6486 6150 0.0026 -
19.8083 6200 0.0022 -
19.9681 6250 0.0025 -
20.1278 6300 0.0022 -
20.2875 6350 0.0025 -
20.4473 6400 0.0024 -
20.6070 6450 0.0027 -
20.7668 6500 0.0017 -
20.9265 6550 0.0025 -
21.0863 6600 0.0025 -
21.2460 6650 0.002 -
21.4058 6700 0.0033 -
21.5655 6750 0.0021 -
21.7252 6800 0.0022 -
21.8850 6850 0.0027 -
22.0447 6900 0.0021 -
22.2045 6950 0.0028 -
22.3642 7000 0.0021 -
22.5240 7050 0.0021 -
22.6837 7100 0.0027 -
22.8435 7150 0.0021 -
23.0032 7200 0.0029 -
23.1629 7250 0.0036 -
23.3227 7300 0.002 -
23.4824 7350 0.0021 -
23.6422 7400 0.002 -
23.8019 7450 0.0025 -
23.9617 7500 0.0024 -
24.1214 7550 0.0026 -
24.2812 7600 0.002 -
24.4409 7650 0.0024 -
24.6006 7700 0.0025 -
24.7604 7750 0.0023 -
24.9201 7800 0.0027 -
25.0799 7850 0.0023 -
25.2396 7900 0.0024 -
25.3994 7950 0.0027 -
25.5591 8000 0.0038 -
25.7188 8050 0.0065 -
25.8786 8100 0.0037 -
26.0383 8150 0.0032 -
26.1981 8200 0.0031 -
26.3578 8250 0.0028 -
26.5176 8300 0.0024 -
26.6773 8350 0.0023 -
26.8371 8400 0.0028 -
26.9968 8450 0.0023 -
27.1565 8500 0.0028 -
27.3163 8550 0.0025 -
27.4760 8600 0.0027 -
27.6358 8650 0.002 -
27.7955 8700 0.0024 -
27.9553 8750 0.0023 -
28.1150 8800 0.0029 -
28.2748 8850 0.0025 -
28.4345 8900 0.002 -
28.5942 8950 0.0025 -
28.7540 9000 0.002 -
28.9137 9050 0.0027 -
29.0735 9100 0.0028 -
29.2332 9150 0.0016 -
29.3930 9200 0.0032 -
29.5527 9250 0.0026 -
29.7125 9300 0.0025 -
29.8722 9350 0.0025 -

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}
}
Downloads last month
10,666
Safetensors
Model size
111M params
Tensor type
F32
·
Inference Examples
Text Classification
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Model tree for mini1013/master_item_top_bt3

Base model

klue/roberta-base
Finetuned
(169)
this model

Evaluation results

View on Papers With Code