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 JohanHeinsen/Old_News_Segmentation_SBERT_V0.1 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 |
|
| 1 |
|
| Label | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|
| all | 0.9665 | 0.9795 | 0.9702 | 0.9890 |
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("setfit_model_id")
# Run inference
preds = model("3) Pigen Dagmar Schrøder, Datter af Privatvægter Frederik Schrøder, Istedgade 6, 2. Sal, er den 24. Ds. bortgaaet fra Hjemmet. Hun er 12. Aar gl., svær af Bygning, har blondt Haar (Pandehaar, var iført rødbrun Nederdel, sort Liv, Sko og Sivhat med Blondebesætning. (H. St.)")
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 7 | 55.4928 | 497 |
| Label | Training Sample Count |
|---|---|
| 0 | 208 |
| 1 | 835 |
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0010 | 1 | 0.255 | - |
| 0.0479 | 50 | 0.136 | - |
| 0.0959 | 100 | 0.0552 | - |
| 0.1438 | 150 | 0.0385 | - |
| 0.1918 | 200 | 0.0237 | - |
| 0.2397 | 250 | 0.0175 | - |
| 0.2876 | 300 | 0.014 | - |
| 0.3356 | 350 | 0.0096 | - |
| 0.3835 | 400 | 0.0088 | - |
| 0.4314 | 450 | 0.0101 | - |
| 0.4794 | 500 | 0.008 | - |
| 0.5273 | 550 | 0.0051 | - |
| 0.5753 | 600 | 0.0036 | - |
| 0.6232 | 650 | 0.0006 | - |
| 0.6711 | 700 | 0.0002 | - |
| 0.7191 | 750 | 0.0001 | - |
| 0.7670 | 800 | 0.0002 | - |
| 0.8150 | 850 | 0.0001 | - |
| 0.8629 | 900 | 0.0001 | - |
| 0.9108 | 950 | 0.0001 | - |
| 0.9588 | 1000 | 0.0001 | - |
@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
CALDISS-AAU/DA-BERT_Old_News_V1