Efficient Few-Shot Learning Without Prompts
Paper
• 2209.11055 • Published
• 4
This is a SetFit model trained on the tmp-org/netto dataset that can be used for Text Classification. This SetFit model uses Alibaba-NLP/gte-multilingual-base 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 |
|---|---|
| Startseite_Startseite |
|
| Coupons_Coupons |
|
| Angebote_Angebote |
|
| Online-Shop_Online-Shop |
|
| Other_Loading |
|
| Karte + Zahlen_Loading |
|
| Angebote_Loading |
|
| Karte + Zahlen_Coupons |
|
| Other_Neuigkeiten |
|
| Other_Gewinnspiel |
|
| Other_Meine Funktionen |
|
| Other_Prospekte |
|
| Karte + Zahlen_Nur Karte |
|
| Other_Angebote details |
|
| Startseite_Loading |
|
| Other_Mein PAYBACK |
|
| Other_Einkaufsliste |
|
| Other_Adventskalender |
|
| Other_Meine digitalen Kassenbons |
|
| Other_Information |
|
| Other_Coupon details |
|
| Karte + Zahlen_Karte + Zahlen |
|
| Online-Shop_Loading |
|
| Other_Rezepte |
|
| Other_Unknown |
|
| Other_Code einlösen |
|
| Coupons_Loading |
|
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("tmp-org/netto_v1")
# Run inference
preds = model(" (Karte + Zahlen) [LinearLayout|WebView]")
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 3 | 197.8415 | 8627 |
| Label | Training Sample Count |
|---|---|
| Angebote_Angebote | 32 |
| Angebote_Loading | 8 |
| Coupons_Coupons | 32 |
| Coupons_Loading | 3 |
| Karte + Zahlen_Coupons | 32 |
| Karte + Zahlen_Karte + Zahlen | 11 |
| Karte + Zahlen_Loading | 8 |
| Karte + Zahlen_Nur Karte | 22 |
| Online-Shop_Loading | 7 |
| Online-Shop_Online-Shop | 32 |
| Other_Adventskalender | 4 |
| Other_Angebote details | 29 |
| Other_Code einlösen | 1 |
| Other_Coupon details | 5 |
| Other_Einkaufsliste | 32 |
| Other_Gewinnspiel | 2 |
| Other_Information | 5 |
| Other_Loading | 10 |
| Other_Mein PAYBACK | 12 |
| Other_Meine Funktionen | 15 |
| Other_Meine digitalen Kassenbons | 6 |
| Other_Neuigkeiten | 18 |
| Other_Prospekte | 16 |
| Other_Rezepte | 31 |
| Other_Unknown | 1 |
| Startseite_Loading | 4 |
| Startseite_Startseite | 32 |
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0004 | 1 | 0.047 | - |
| 0.0194 | 50 | 0.2357 | - |
| 0.0388 | 100 | 0.1651 | - |
| 0.0581 | 150 | 0.1409 | - |
| 0.0775 | 200 | 0.1471 | - |
| 0.0969 | 250 | 0.1272 | - |
| 0.1163 | 300 | 0.1091 | - |
| 0.1357 | 350 | 0.1015 | - |
| 0.1550 | 400 | 0.0965 | - |
| 0.1744 | 450 | 0.0707 | - |
| 0.1938 | 500 | 0.0922 | - |
| 0.2132 | 550 | 0.092 | - |
| 0.2326 | 600 | 0.0565 | - |
| 0.2519 | 650 | 0.0563 | - |
| 0.2713 | 700 | 0.0801 | - |
| 0.2907 | 750 | 0.0932 | - |
| 0.3101 | 800 | 0.0714 | - |
| 0.3295 | 850 | 0.0685 | - |
| 0.3488 | 900 | 0.0523 | - |
| 0.3682 | 950 | 0.0768 | - |
| 0.3876 | 1000 | 0.0559 | - |
| 0.4070 | 1050 | 0.0545 | - |
| 0.4264 | 1100 | 0.0421 | - |
| 0.4457 | 1150 | 0.0557 | - |
| 0.4651 | 1200 | 0.0645 | - |
| 0.4845 | 1250 | 0.0583 | - |
| 0.5039 | 1300 | 0.0407 | - |
| 0.5233 | 1350 | 0.0486 | - |
| 0.5426 | 1400 | 0.0575 | - |
| 0.5620 | 1450 | 0.0425 | - |
| 0.5814 | 1500 | 0.0507 | - |
| 0.6008 | 1550 | 0.0502 | - |
| 0.6202 | 1600 | 0.041 | - |
| 0.6395 | 1650 | 0.037 | - |
| 0.6589 | 1700 | 0.0464 | - |
| 0.6783 | 1750 | 0.0444 | - |
| 0.6977 | 1800 | 0.0333 | - |
| 0.7171 | 1850 | 0.0305 | - |
| 0.7364 | 1900 | 0.046 | - |
| 0.7558 | 1950 | 0.031 | - |
| 0.7752 | 2000 | 0.0463 | - |
| 0.7946 | 2050 | 0.0273 | - |
| 0.8140 | 2100 | 0.0297 | - |
| 0.8333 | 2150 | 0.0303 | - |
| 0.8527 | 2200 | 0.0381 | - |
| 0.8721 | 2250 | 0.0459 | - |
| 0.8915 | 2300 | 0.0506 | - |
| 0.9109 | 2350 | 0.0418 | - |
| 0.9302 | 2400 | 0.0231 | - |
| 0.9496 | 2450 | 0.0358 | - |
| 0.9690 | 2500 | 0.0368 | - |
| 0.9884 | 2550 | 0.0286 | - |
@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}
}
Base model
Alibaba-NLP/gte-multilingual-base