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
0.0	'헤드 HEAD 남성용 그리드 2 0 로우 라켓볼스쿼시 실내 코트 슈즈 자국이 정품보장 스포츠/레저>스쿼시>기타스쿼시용품' '테크니화이버 초록줄 릴 200m TF 스쿼시스트링 20회작업분 TF-305 1 스포츠/레저>스쿼시>기타스쿼시용품' 'MOTUZP 단일 도트 스쿼시 공 고무 고탄력 라켓 초보자 경쟁 훈련을위한 훈련 연습을위한 single dot 스포츠/레저>스쿼시>기타스쿼시용품'
2.0	'테크니화이버 Carboflex 125 X탑 언스트렁 스쿼시 라켓 138966103 스포츠/레저>스쿼시>스쿼시라켓' 'Gearbox GB3K 170Q 라켓볼 라켓 3 58 그립 스포츠/레저>스쿼시>스쿼시라켓' 'Tecnifibre 스쿼시 Carboflex 125S 라켓 SynGut 스트링 스포츠/레저>스쿼시>스쿼시라켓'
1.0	'던롭 PRO 스쿼시볼 스포츠/레저>스쿼시>스쿼시공' '브니엘 토너먼트 스쿼시볼 스포츠/레저>스쿼시>스쿼시공' '던롭 Pro 스쿼시볼 (유리 코트 전용구) 스포츠/레저>스쿼시>스쿼시공'

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_sl18")
# Run inference
preds = model("베이퍼 130 라님 엘 윌리 스포츠/레저>스쿼시>스쿼시라켓")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	4	9.4626	18

Label	Training Sample Count
0.0	70
1.0	7
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
0.0345	1	0.4863	-
1.7241	50	0.2641	-
3.4483	100	0.018	-
5.1724	150	0.0	-
6.8966	200	0.0	-
8.6207	250	0.0	-
10.3448	300	0.0	-
12.0690	350	0.0	-
13.7931	400	0.0	-
15.5172	450	0.0	-
17.2414	500	0.0	-
18.9655	550	0.0	-
20.6897	600	0.0	-
22.4138	650	0.0	-
24.1379	700	0.0	-
25.8621	750	0.0	-
27.5862	800	0.0	-
29.3103	850	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_sl18