metadata
library_name: setfit
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
metrics:
- accuracy
widget:
- text: Tiong Bahru Plaza, DDC-L2-5, AHU-L2-03 trip alarm
- text: >-
Tiong Bahru Plaza, DDC L4-1, PAU-L4-03 supply air temperature (Units:
°C).2
- text: Tiong Bahru Plaza, DDC-L20, AHU 20-1 VSD CONTROL
- text: 'Tiong Bahru Plaza, VAV 19-7, Discharge Air Flow (Units: m3/h)'
- text: Tiong Bahru Plaza, DDC-L2-5, PAU-L2-02 VSD control
pipeline_tag: text-classification
inference: true
base_model: sentence-transformers/paraphrase-MiniLM-L3-v2
model-index:
- name: SetFit with sentence-transformers/paraphrase-MiniLM-L3-v2
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: accuracy
value: 0.9863861386138614
name: Accuracy
SetFit with sentence-transformers/paraphrase-MiniLM-L3-v2
This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-MiniLM-L3-v2 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:
- Fine-tuning a Sentence Transformer with contrastive learning.
- Training a classification head with features from the fine-tuned Sentence Transformer.
Model Details
Model Description
- Model Type: SetFit
- Sentence Transformer body: sentence-transformers/paraphrase-MiniLM-L3-v2
- Classification head: a LogisticRegression instance
- Maximum Sequence Length: 128 tokens
- Number of Classes: 44 classes
Model Sources
- Repository: SetFit on GitHub
- Paper: Efficient Few-Shot Learning Without Prompts
- Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts
Model Labels
Label | Examples |
---|---|
28 |
|
33 |
|
2 |
|
42 |
|
31 |
|
9 |
|
1 |
|
25 |
|
12 |
|
8 |
|
24 |
|
13 |
|
30 |
|
5 |
|
43 |
|
14 |
|
36 |
|
34 |
|
18 |
|
6 |
|
17 |
|
0 |
|
29 |
|
15 |
|
11 |
|
32 |
|
27 |
|
41 |
|
40 |
|
3 |
|
22 |
|
21 |
|
26 |
|
23 |
|
16 |
|
39 |
|
7 |
|
37 |
|
20 |
|
35 |
|
4 |
|
19 |
|
38 |
|
10 |
|
Evaluation
Metrics
Label | Accuracy |
---|---|
all | 0.9864 |
Uses
Direct Use for Inference
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("Varun1010/all-MiniLM-L6-v2-polaris-more")
# Run inference
preds = model("Tiong Bahru Plaza, DDC-L20, AHU 20-1 VSD CONTROL")
Training Details
Training Set Metrics
Training set | Min | Median | Max |
---|---|---|---|
Word count | 3 | 8.7134 | 13 |
Label | Training Sample Count |
---|---|
0 | 4 |
1 | 4 |
2 | 4 |
3 | 4 |
4 | 4 |
5 | 4 |
6 | 4 |
7 | 4 |
8 | 4 |
9 | 4 |
10 | 4 |
11 | 4 |
12 | 4 |
13 | 4 |
14 | 4 |
15 | 4 |
16 | 4 |
17 | 4 |
18 | 4 |
19 | 3 |
20 | 3 |
21 | 4 |
22 | 4 |
23 | 1 |
24 | 1 |
25 | 4 |
26 | 4 |
27 | 4 |
28 | 4 |
29 | 4 |
30 | 3 |
31 | 3 |
32 | 4 |
33 | 4 |
34 | 4 |
35 | 4 |
36 | 4 |
37 | 4 |
38 | 3 |
39 | 3 |
40 | 1 |
41 | 1 |
42 | 3 |
43 | 4 |
Training Hyperparameters
- batch_size: (64, 64)
- num_epochs: (5, 5)
- max_steps: -1
- sampling_strategy: oversampling
- body_learning_rate: (2e-05, 1e-05)
- head_learning_rate: 0.01
- loss: CosineSimilarityLoss
- distance_metric: cosine_distance
- margin: 0.25
- end_to_end: False
- use_amp: False
- warmup_proportion: 0.1
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
Training Results
Epoch | Step | Training Loss | Validation Loss |
---|---|---|---|
0.0027 | 1 | 0.1538 | - |
0.1330 | 50 | 0.0495 | - |
0.2660 | 100 | 0.0655 | - |
0.3989 | 150 | 0.0336 | - |
0.5319 | 200 | 0.0282 | - |
0.6649 | 250 | 0.0207 | - |
0.7979 | 300 | 0.0184 | - |
0.9309 | 350 | 0.0163 | - |
1.0638 | 400 | 0.0088 | - |
1.1968 | 450 | 0.0307 | - |
1.3298 | 500 | 0.0153 | - |
1.4628 | 550 | 0.0079 | - |
1.5957 | 600 | 0.02 | - |
1.7287 | 650 | 0.0165 | - |
1.8617 | 700 | 0.0087 | - |
1.9947 | 750 | 0.0236 | - |
2.1277 | 800 | 0.0108 | - |
2.2606 | 850 | 0.0071 | - |
2.3936 | 900 | 0.0137 | - |
2.5266 | 950 | 0.0104 | - |
2.6596 | 1000 | 0.0054 | - |
2.7926 | 1050 | 0.0058 | - |
2.9255 | 1100 | 0.0052 | - |
3.0585 | 1150 | 0.0053 | - |
3.1915 | 1200 | 0.004 | - |
3.3245 | 1250 | 0.0047 | - |
3.4574 | 1300 | 0.0176 | - |
3.5904 | 1350 | 0.0046 | - |
3.7234 | 1400 | 0.0139 | - |
3.8564 | 1450 | 0.0043 | - |
3.9894 | 1500 | 0.0042 | - |
4.1223 | 1550 | 0.0112 | - |
4.2553 | 1600 | 0.0091 | - |
4.3883 | 1650 | 0.0045 | - |
4.5213 | 1700 | 0.009 | - |
4.6543 | 1750 | 0.0097 | - |
4.7872 | 1800 | 0.0049 | - |
4.9202 | 1850 | 0.0036 | - |
Framework Versions
- Python: 3.10.12
- SetFit: 1.0.3
- Sentence Transformers: 2.6.1
- Transformers: 4.38.2
- PyTorch: 2.2.1+cu121
- Datasets: 2.18.0
- Tokenizers: 0.15.2
Citation
BibTeX
@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}
}