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: 3 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
2.0	'야외 베개 삼각 봄 꽃 간단한 방수 장식 소파 커버 쿠션 가구/인테리어>홈데코>쿠션/방석>쿠션/방석커버세트' '컬러 쿠션 커버 단색 린넨 소파 자동차 장식 베개 간단한 가구/인테리어>홈데코>쿠션/방석>방석커버' '코멧 여름 필수템 실리콘 방석 커버 세트 대형 가구/인테리어>홈데코>쿠션/방석>방석커버'
0.0	'두툼한 내열 뚝배기 땡땡이리본 주방장갑 보호장갑 주방용장갑 키친툴 가구/인테리어>홈데코>주방데코>주방장갑' '테이블매트 방수 커피색 북유럽 식탁매트 식탁보 가구/인테리어>홈데코>주방데코>식탁매트' '인터그레이스 앳홈 기름때방지 주방아트보드 고래의 꿈 일러스트 디자인 3M 부착형 895x580mm 가구/인테리어>홈데코>주방데코>기타주방데코'
1.0	'실외기 덮개 가림막 중 에어컨 실외기커버 가구/인테리어>홈데코>커버류>에어컨커버' '벽걸이 에어컨커버 캐릭터 스판 덮개 에어컨 보관 가구/인테리어>홈데코>커버류>에어컨커버' '피아노 커버 덮개 세트 스툴 의자 학원 건반 천 패브릭 가구/인테리어>홈데코>커버류>피아노커버'

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_fi16")
# Run inference
preds = model("벨라이프 마리포사 테카포 실리콘 테이블매트 가구/인테리어>홈데코>주방데코>식탁매트")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	3	9.2857	19

Label	Training Sample Count
0.0	70
1.0	70
2.0	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
0.0238	1	0.4981	-
1.1905	50	0.4801	-
2.3810	100	0.0492	-
3.5714	150	0.0	-
4.7619	200	0.0	-
5.9524	250	0.0	-
7.1429	300	0.0	-
8.3333	350	0.0	-
9.5238	400	0.0	-
10.7143	450	0.0	-
11.9048	500	0.0	-
13.0952	550	0.0	-
14.2857	600	0.0	-
15.4762	650	0.0	-
16.6667	700	0.0	-
17.8571	750	0.0	-
19.0476	800	0.0	-
20.2381	850	0.0	-
21.4286	900	0.0	-
22.6190	950	0.0	-
23.8095	1000	0.0	-
25.0	1050	0.0	-
26.1905	1100	0.0	-
27.3810	1150	0.0	-
28.5714	1200	0.0	-
29.7619	1250	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_fi16