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 |
|---|---|
| 10.0 |
|
| 8.0 |
|
| 1.0 |
|
| 6.0 |
|
| 5.0 |
|
| 9.0 |
|
| 4.0 |
|
| 0.0 |
|
| 7.0 |
|
| 11.0 |
|
| 3.0 |
|
| 2.0 |
|
| Label | Metric |
|---|---|
| all | 0.9601 |
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_lh13")
# Run inference
preds = model("빨래방망이 다듬이방망이 2P세트 이불방 다듬잇방망이 신규A")
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 3 | 9.8017 | 20 |
| 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 |
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0106 | 1 | 0.4631 | - |
| 0.5319 | 50 | 0.4236 | - |
| 1.0638 | 100 | 0.2102 | - |
| 1.5957 | 150 | 0.114 | - |
| 2.1277 | 200 | 0.0733 | - |
| 2.6596 | 250 | 0.0541 | - |
| 3.1915 | 300 | 0.0316 | - |
| 3.7234 | 350 | 0.0104 | - |
| 4.2553 | 400 | 0.0098 | - |
| 4.7872 | 450 | 0.0039 | - |
| 5.3191 | 500 | 0.0026 | - |
| 5.8511 | 550 | 0.0002 | - |
| 6.3830 | 600 | 0.0002 | - |
| 6.9149 | 650 | 0.0001 | - |
| 7.4468 | 700 | 0.0001 | - |
| 7.9787 | 750 | 0.0001 | - |
| 8.5106 | 800 | 0.0001 | - |
| 9.0426 | 850 | 0.0001 | - |
| 9.5745 | 900 | 0.0001 | - |
| 10.1064 | 950 | 0.0001 | - |
| 10.6383 | 1000 | 0.0001 | - |
| 11.1702 | 1050 | 0.0001 | - |
| 11.7021 | 1100 | 0.0001 | - |
| 12.2340 | 1150 | 0.0001 | - |
| 12.7660 | 1200 | 0.0 | - |
| 13.2979 | 1250 | 0.0001 | - |
| 13.8298 | 1300 | 0.0001 | - |
| 14.3617 | 1350 | 0.0 | - |
| 14.8936 | 1400 | 0.0001 | - |
| 15.4255 | 1450 | 0.0 | - |
| 15.9574 | 1500 | 0.0001 | - |
| 16.4894 | 1550 | 0.0 | - |
| 17.0213 | 1600 | 0.0 | - |
| 17.5532 | 1650 | 0.0 | - |
| 18.0851 | 1700 | 0.0 | - |
| 18.6170 | 1750 | 0.0 | - |
| 19.1489 | 1800 | 0.0 | - |
| 19.6809 | 1850 | 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}
}