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 |
|---|---|
| 0.0 |
|
| 3.0 |
|
| 7.0 |
|
| 5.0 |
|
| 10.0 |
|
| 1.0 |
|
| 11.0 |
|
| 6.0 |
|
| 8.0 |
|
| 12.0 |
|
| 2.0 |
|
| 13.0 |
|
| 9.0 |
|
| 4.0 |
|
| 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_sl20")
# Run inference
preds = model("브로브 수영랜턴 고급형 스노쿨링 잠수후레쉬 CREE 스포츠/레저>스킨스쿠버>수중전등")
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 2 | 9.2506 | 24 |
| Label | Training Sample Count |
|---|---|
| 0.0 | 69 |
| 1.0 | 70 |
| 2.0 | 70 |
| 3.0 | 70 |
| 4.0 | 9 |
| 5.0 | 70 |
| 6.0 | 70 |
| 7.0 | 70 |
| 8.0 | 70 |
| 9.0 | 70 |
| 10.0 | 70 |
| 11.0 | 70 |
| 12.0 | 70 |
| 13.0 | 10 |
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0060 | 1 | 0.5025 | - |
| 0.2976 | 50 | 0.4963 | - |
| 0.5952 | 100 | 0.3183 | - |
| 0.8929 | 150 | 0.0275 | - |
| 1.1905 | 200 | 0.0142 | - |
| 1.4881 | 250 | 0.0142 | - |
| 1.7857 | 300 | 0.0132 | - |
| 2.0833 | 350 | 0.0144 | - |
| 2.3810 | 400 | 0.0097 | - |
| 2.6786 | 450 | 0.001 | - |
| 2.9762 | 500 | 0.0002 | - |
| 3.2738 | 550 | 0.0 | - |
| 3.5714 | 600 | 0.0 | - |
| 3.8690 | 650 | 0.0 | - |
| 4.1667 | 700 | 0.0 | - |
| 4.4643 | 750 | 0.0 | - |
| 4.7619 | 800 | 0.0001 | - |
| 5.0595 | 850 | 0.0 | - |
| 5.3571 | 900 | 0.0 | - |
| 5.6548 | 950 | 0.0 | - |
| 5.9524 | 1000 | 0.0 | - |
| 6.25 | 1050 | 0.0 | - |
| 6.5476 | 1100 | 0.0 | - |
| 6.8452 | 1150 | 0.0 | - |
| 7.1429 | 1200 | 0.0 | - |
| 7.4405 | 1250 | 0.0 | - |
| 7.7381 | 1300 | 0.0 | - |
| 8.0357 | 1350 | 0.0 | - |
| 8.3333 | 1400 | 0.0 | - |
| 8.6310 | 1450 | 0.0 | - |
| 8.9286 | 1500 | 0.0 | - |
| 9.2262 | 1550 | 0.0 | - |
| 9.5238 | 1600 | 0.0 | - |
| 9.8214 | 1650 | 0.0 | - |
| 10.1190 | 1700 | 0.0 | - |
| 10.4167 | 1750 | 0.0 | - |
| 10.7143 | 1800 | 0.0 | - |
| 11.0119 | 1850 | 0.0 | - |
| 11.3095 | 1900 | 0.0 | - |
| 11.6071 | 1950 | 0.0 | - |
| 11.9048 | 2000 | 0.0 | - |
| 12.2024 | 2050 | 0.0 | - |
| 12.5 | 2100 | 0.0 | - |
| 12.7976 | 2150 | 0.0006 | - |
| 13.0952 | 2200 | 0.0001 | - |
| 13.3929 | 2250 | 0.0 | - |
| 13.6905 | 2300 | 0.0 | - |
| 13.9881 | 2350 | 0.0 | - |
| 14.2857 | 2400 | 0.0 | - |
| 14.5833 | 2450 | 0.0 | - |
| 14.8810 | 2500 | 0.0 | - |
| 15.1786 | 2550 | 0.0 | - |
| 15.4762 | 2600 | 0.0 | - |
| 15.7738 | 2650 | 0.0 | - |
| 16.0714 | 2700 | 0.0 | - |
| 16.3690 | 2750 | 0.0 | - |
| 16.6667 | 2800 | 0.0 | - |
| 16.9643 | 2850 | 0.0 | - |
| 17.2619 | 2900 | 0.0 | - |
| 17.5595 | 2950 | 0.0 | - |
| 17.8571 | 3000 | 0.0 | - |
| 18.1548 | 3050 | 0.0 | - |
| 18.4524 | 3100 | 0.0 | - |
| 18.75 | 3150 | 0.0 | - |
| 19.0476 | 3200 | 0.0 | - |
| 19.3452 | 3250 | 0.0 | - |
| 19.6429 | 3300 | 0.0 | - |
| 19.9405 | 3350 | 0.0 | - |
| 20.2381 | 3400 | 0.0 | - |
| 20.5357 | 3450 | 0.0 | - |
| 20.8333 | 3500 | 0.0 | - |
| 21.1310 | 3550 | 0.0 | - |
| 21.4286 | 3600 | 0.0 | - |
| 21.7262 | 3650 | 0.0 | - |
| 22.0238 | 3700 | 0.0 | - |
| 22.3214 | 3750 | 0.0 | - |
| 22.6190 | 3800 | 0.0 | - |
| 22.9167 | 3850 | 0.0 | - |
| 23.2143 | 3900 | 0.0 | - |
| 23.5119 | 3950 | 0.0002 | - |
| 23.8095 | 4000 | 0.0 | - |
| 24.1071 | 4050 | 0.0 | - |
| 24.4048 | 4100 | 0.0 | - |
| 24.7024 | 4150 | 0.0 | - |
| 25.0 | 4200 | 0.0 | - |
| 25.2976 | 4250 | 0.0 | - |
| 25.5952 | 4300 | 0.0 | - |
| 25.8929 | 4350 | 0.0 | - |
| 26.1905 | 4400 | 0.0 | - |
| 26.4881 | 4450 | 0.0 | - |
| 26.7857 | 4500 | 0.0 | - |
| 27.0833 | 4550 | 0.0 | - |
| 27.3810 | 4600 | 0.0 | - |
| 27.6786 | 4650 | 0.0 | - |
| 27.9762 | 4700 | 0.0 | - |
| 28.2738 | 4750 | 0.0 | - |
| 28.5714 | 4800 | 0.0 | - |
| 28.8690 | 4850 | 0.0 | - |
| 29.1667 | 4900 | 0.0 | - |
| 29.4643 | 4950 | 0.0 | - |
| 29.7619 | 5000 | 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}
}