--- base_model: mini1013/master_domain library_name: setfit metrics: - metric pipeline_tag: text-classification tags: - setfit - sentence-transformers - text-classification - generated_from_setfit_trainer widget: - text: 타공판닷컴 세계지도 대형 월드맵 세계지도03_600x900 (주)오빌 - text: 스프링 제본 PDF 흑백 고품질 레이저 출력 - 흑백 양면인쇄 모조지80g 50p 스프링 흑백양면●●_모조지100g_167~170 page 도서출판 법현 - text: '[달페이퍼] 달페이퍼 미니미니 6종 엽서 postcard 인테리어엽서 6 미니미니 일하는 주식회사 천유닷컴' - text: 환갑 현수막 회갑 생신 잔치 플랜카드 C00 네임 소형100x70cm C22 얼쑤(남자)-자유문구포토형_소형 100x70cm (주)엔비웨일인터렉티브 - text: 스프링 제본 PDF 흑백 고품질 레이저 출력 - 흑백 양면인쇄 모조지80g 50p 스프링 흑백단면●_모조지80g_179~182 page 도서출판 법현 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: metric value: 0.964332367808258 name: Metric --- # SetFit with mini1013/master_domain This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [mini1013/master_domain](https://huggingface.co/mini1013/master_domain) as the Sentence Transformer embedding model. A [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) 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](https://www.sbert.net) 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:** [mini1013/master_domain](https://huggingface.co/mini1013/master_domain) - **Classification head:** a [LogisticRegression](https://scikit-learn.org/stable/modules/generated/sklearn.linear_model.LogisticRegression.html) instance - **Maximum Sequence Length:** 512 tokens - **Number of Classes:** 17 classes ### Model Sources - **Repository:** [SetFit on GitHub](https://github.com/huggingface/setfit) - **Paper:** [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055) - **Blogpost:** [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit) ### Model Labels | Label | Examples | |:------|:---------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------| | 6.0 | | | 2.0 | | | 5.0 | | | 13.0 | | | 11.0 | | | 10.0 | | | 4.0 | | | 0.0 | | | 14.0 | | | 7.0 | | | 16.0 | | | 15.0 | | | 9.0 | | | 12.0 | | | 3.0 | | | 1.0 | | | 8.0 | | ## Evaluation ### Metrics | Label | Metric | |:--------|:-------| | **all** | 0.9643 | ## Uses ### Direct Use for Inference First install the SetFit library: ```bash pip install setfit ``` Then you can load this model and run inference. ```python from setfit import SetFitModel # Download from the 🤗 Hub model = SetFitModel.from_pretrained("mini1013/master_cate_lh8") # Run inference preds = model("타공판닷컴 세계지도 대형 월드맵 세계지도03_600x900 (주)오빌") ``` ## Training Details ### Training Set Metrics | Training set | Min | Median | Max | |:-------------|:----|:--------|:----| | Word count | 4 | 11.1176 | 26 | | Label | Training Sample Count | |:------|:----------------------| | 0.0 | 50 | | 1.0 | 50 | | 2.0 | 50 | | 3.0 | 50 | | 4.0 | 50 | | 5.0 | 50 | | 6.0 | 50 | | 7.0 | 50 | | 8.0 | 50 | | 9.0 | 50 | | 10.0 | 50 | | 11.0 | 50 | | 12.0 | 50 | | 13.0 | 50 | | 14.0 | 50 | | 15.0 | 50 | | 16.0 | 50 | ### Training Hyperparameters - batch_size: (512, 512) - num_epochs: (20, 20) - max_steps: -1 - sampling_strategy: oversampling - num_iterations: 40 - body_learning_rate: (2e-05, 2e-05) - head_learning_rate: 2e-05 - loss: CosineSimilarityLoss - distance_metric: cosine_distance - margin: 0.25 - end_to_end: False - use_amp: False - warmup_proportion: 0.1 - seed: 42 - eval_max_steps: -1 - load_best_model_at_end: False ### Training Results | Epoch | Step | Training Loss | Validation Loss | |:-------:|:----:|:-------------:|:---------------:| | 0.0075 | 1 | 0.4622 | - | | 0.3759 | 50 | 0.3276 | - | | 0.7519 | 100 | 0.2741 | - | | 1.1278 | 150 | 0.167 | - | | 1.5038 | 200 | 0.082 | - | | 1.8797 | 250 | 0.0368 | - | | 2.2556 | 300 | 0.0406 | - | | 2.6316 | 350 | 0.0331 | - | | 3.0075 | 400 | 0.0282 | - | | 3.3835 | 450 | 0.0144 | - | | 3.7594 | 500 | 0.005 | - | | 4.1353 | 550 | 0.0036 | - | | 4.5113 | 600 | 0.0036 | - | | 4.8872 | 650 | 0.0005 | - | | 5.2632 | 700 | 0.0003 | - | | 5.6391 | 750 | 0.0003 | - | | 6.0150 | 800 | 0.0002 | - | | 6.3910 | 850 | 0.0003 | - | | 6.7669 | 900 | 0.0002 | - | | 7.1429 | 950 | 0.0002 | - | | 7.5188 | 1000 | 0.0001 | - | | 7.8947 | 1050 | 0.0001 | - | | 8.2707 | 1100 | 0.0001 | - | | 8.6466 | 1150 | 0.0001 | - | | 9.0226 | 1200 | 0.0001 | - | | 9.3985 | 1250 | 0.0001 | - | | 9.7744 | 1300 | 0.0001 | - | | 10.1504 | 1350 | 0.0001 | - | | 10.5263 | 1400 | 0.0001 | - | | 10.9023 | 1450 | 0.0001 | - | | 11.2782 | 1500 | 0.0001 | - | | 11.6541 | 1550 | 0.0001 | - | | 12.0301 | 1600 | 0.0001 | - | | 12.4060 | 1650 | 0.0001 | - | | 12.7820 | 1700 | 0.0001 | - | | 13.1579 | 1750 | 0.0001 | - | | 13.5338 | 1800 | 0.0001 | - | | 13.9098 | 1850 | 0.0001 | - | | 14.2857 | 1900 | 0.0001 | - | | 14.6617 | 1950 | 0.0001 | - | | 15.0376 | 2000 | 0.0001 | - | | 15.4135 | 2050 | 0.0001 | - | | 15.7895 | 2100 | 0.0001 | - | | 16.1654 | 2150 | 0.0001 | - | | 16.5414 | 2200 | 0.0001 | - | | 16.9173 | 2250 | 0.0001 | - | | 17.2932 | 2300 | 0.0001 | - | | 17.6692 | 2350 | 0.0001 | - | | 18.0451 | 2400 | 0.0001 | - | | 18.4211 | 2450 | 0.0001 | - | | 18.7970 | 2500 | 0.0001 | - | | 19.1729 | 2550 | 0.0001 | - | | 19.5489 | 2600 | 0.0001 | - | | 19.9248 | 2650 | 0.0001 | - | ### Framework Versions - Python: 3.10.12 - SetFit: 1.1.0.dev0 - Sentence Transformers: 3.1.1 - Transformers: 4.46.1 - PyTorch: 2.4.0+cu121 - Datasets: 2.20.0 - Tokenizers: 0.20.0 ## Citation ### BibTeX ```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} } ```