Efficient Few-Shot Learning Without Prompts
Paper
•
2209.11055
•
Published
•
4
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:
| Label | Examples |
|---|---|
| 4.0 |
|
| 2.0 |
|
| 5.0 |
|
| 0.0 |
|
| 3.0 |
|
| 7.0 |
|
| 1.0 |
|
| 6.0 |
|
| Label | Accuracy |
|---|---|
| all | 1.0 |
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_bc28")
# Run inference
preds = model("임산부 손목 산모 산후 보호대 아대 패드 핑크 출산/육아 > 임산부용품 > 임산부보호대")
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 7 | 15.0119 | 28 |
| Label | Training Sample Count |
|---|---|
| 0.0 | 70 |
| 1.0 | 15 |
| 2.0 | 70 |
| 3.0 | 70 |
| 4.0 | 70 |
| 5.0 | 70 |
| 6.0 | 70 |
| 7.0 | 70 |
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0101 | 1 | 0.4772 | - |
| 0.5051 | 50 | 0.4978 | - |
| 1.0101 | 100 | 0.2919 | - |
| 1.5152 | 150 | 0.0498 | - |
| 2.0202 | 200 | 0.019 | - |
| 2.5253 | 250 | 0.0011 | - |
| 3.0303 | 300 | 0.0001 | - |
| 3.5354 | 350 | 0.0001 | - |
| 4.0404 | 400 | 0.0001 | - |
| 4.5455 | 450 | 0.0 | - |
| 5.0505 | 500 | 0.0 | - |
| 5.5556 | 550 | 0.0 | - |
| 6.0606 | 600 | 0.0 | - |
| 6.5657 | 650 | 0.0 | - |
| 7.0707 | 700 | 0.0 | - |
| 7.5758 | 750 | 0.0 | - |
| 8.0808 | 800 | 0.0 | - |
| 8.5859 | 850 | 0.0 | - |
| 9.0909 | 900 | 0.0 | - |
| 9.5960 | 950 | 0.0 | - |
| 10.1010 | 1000 | 0.0 | - |
| 10.6061 | 1050 | 0.0 | - |
| 11.1111 | 1100 | 0.0 | - |
| 11.6162 | 1150 | 0.0 | - |
| 12.1212 | 1200 | 0.0 | - |
| 12.6263 | 1250 | 0.0 | - |
| 13.1313 | 1300 | 0.0 | - |
| 13.6364 | 1350 | 0.0 | - |
| 14.1414 | 1400 | 0.0 | - |
| 14.6465 | 1450 | 0.0 | - |
| 15.1515 | 1500 | 0.0 | - |
| 15.6566 | 1550 | 0.0 | - |
| 16.1616 | 1600 | 0.0 | - |
| 16.6667 | 1650 | 0.0 | - |
| 17.1717 | 1700 | 0.0 | - |
| 17.6768 | 1750 | 0.0 | - |
| 18.1818 | 1800 | 0.0 | - |
| 18.6869 | 1850 | 0.0 | - |
| 19.1919 | 1900 | 0.0 | - |
| 19.6970 | 1950 | 0.0 | - |
| 20.2020 | 2000 | 0.0 | - |
| 20.7071 | 2050 | 0.0 | - |
| 21.2121 | 2100 | 0.0 | - |
| 21.7172 | 2150 | 0.0 | - |
| 22.2222 | 2200 | 0.0 | - |
| 22.7273 | 2250 | 0.0 | - |
| 23.2323 | 2300 | 0.0 | - |
| 23.7374 | 2350 | 0.0 | - |
| 24.2424 | 2400 | 0.0 | - |
| 24.7475 | 2450 | 0.0 | - |
| 25.2525 | 2500 | 0.0 | - |
| 25.7576 | 2550 | 0.0 | - |
| 26.2626 | 2600 | 0.0 | - |
| 26.7677 | 2650 | 0.0 | - |
| 27.2727 | 2700 | 0.0 | - |
| 27.7778 | 2750 | 0.0 | - |
| 28.2828 | 2800 | 0.0 | - |
| 28.7879 | 2850 | 0.0 | - |
| 29.2929 | 2900 | 0.0 | - |
| 29.7980 | 2950 | 0.0 | - |
@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}
}