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:

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

Model Details

Model Description

Model Type: SetFit
Sentence Transformer body: mini1013/master_domain
Classification head: a LogisticRegression instance
Maximum Sequence Length: 512 tokens
Number of Classes: 4 classes

Model Sources

Repository: SetFit on GitHub
Paper: Efficient Few-Shot Learning Without Prompts
Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts

Model Labels

Label	Examples
1.0	'에싸 라보엠Ⅲ 4인 오픈코너형 기능성 카시미라 패브릭 소파 가구/인테리어>거실가구>소파>패브릭소파' '보루네오 플레타 3인용 천연소가죽 소파 가구/인테리어>거실가구>소파>가죽소파' '동서가구 프라임 소나무원목 내추럴 황토 카우치 흙소파 DF638379 가구/인테리어>거실가구>소파>흙/돌소파'
2.0	'체스 유리 진열장 가구/인테리어>거실가구>장식장' '디자인벤처스 로맨틱 1800 유리장 가구/인테리어>거실가구>장식장' '퍼니처스마트 로랜드 유리 장식장 가구/인테리어>거실가구>장식장'
0.0	'나무뜰 켄트 서랍형 거실장 1200 티비다이 MRF013 가구/인테리어>거실가구>TV거실장' '리바트 셀리나 1800 거실장 가구/인테리어>거실가구>TV거실장' '슈퍼홈 리처 티비다이 낮은 거실장 2000 가구/인테리어>거실가구>TV거실장'
3.0	'테이블 거실 커피 탁자 북유럽 좌식 인테리어 티 모던 카페 라운드 가구/인테리어>거실가구>테이블>거실테이블' '미드센추리 테라조 협탁 사이드 테이블 거실 소파 장식장 선반형 가구/인테리어>거실가구>테이블>사이드테이블' '원목좌식테이블 방석 세트 원형 차 홈 카페 거실 가구/인테리어>거실가구>테이블>거실테이블'

Evaluation

Metrics

Label	Accuracy
all	1.0

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_fi1")
# Run inference
preds = model("모던탑21 모던 클래식 800 3단 장식장 가구/인테리어>거실가구>장식장")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	4	8.1714	18

Label	Training Sample Count
0.0	70
1.0	70
2.0	70
3.0	70

Training Hyperparameters

batch_size: (256, 256)
num_epochs: (30, 30)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 50
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
0.0182	1	0.4862	-
0.9091	50	0.4961	-
1.8182	100	0.4367	-
2.7273	150	0.0317	-
3.6364	200	0.0	-
4.5455	250	0.0	-
5.4545	300	0.0	-
6.3636	350	0.0	-
7.2727	400	0.0	-
8.1818	450	0.0	-
9.0909	500	0.0	-
10.0	550	0.0	-
10.9091	600	0.0	-
11.8182	650	0.0	-
12.7273	700	0.0	-
13.6364	750	0.0	-
14.5455	800	0.0	-
15.4545	850	0.0	-
16.3636	900	0.0	-
17.2727	950	0.0	-
18.1818	1000	0.0	-
19.0909	1050	0.0	-
20.0	1100	0.0	-
20.9091	1150	0.0	-
21.8182	1200	0.0	-
22.7273	1250	0.0	-
23.6364	1300	0.0	-
24.5455	1350	0.0	-
25.4545	1400	0.0	-
26.3636	1450	0.0	-
27.2727	1500	0.0	-
28.1818	1550	0.0	-
29.0909	1600	0.0	-
30.0	1650	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}
}

mini1013
/

master_cate_fi1