metadata

base_model: dunzhang/stella_en_1.5B_v5
library_name: setfit
metrics:
  - accuracy
pipeline_tag: text-classification
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      I have never owned a F-150. I fell in love with them in 2015 and really
      like the idea of a rust free body on a truck.
  - text: >-
      No rust. A few scratches on the front bumper cover. A few chips from rocks
      and other things, but other than that I’d say it’s pretty flawless. No
      swirls or fading.
  - text: >-
      I wouldn’t cal it bad ownership at all. Hyundai’s paint is notoriously
      crappy, and rust issues are quite common. Just consider yourself lucky.
  - text: >-
      Our white Atlas CS has SHIT paint. It’s covered in rock chips and rust
      spots.
  - text: >-
      Mines a work in progress: 1979 Ranger XLT 5.4L supercharged (From an 03
      Lightning) 4R100 auto trans 2015 f-150 chassis w\ 3.73 diffs Orig paint
      (rough and faded but no rust)
inference: true

SetFit with dunzhang/stella_en_1.5B_v5

This is a SetFit model that can be used for Text Classification. This SetFit model uses dunzhang/stella_en_1.5B_v5 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: dunzhang/stella_en_1.5B_v5
Classification head: a LogisticRegression instance
Maximum Sequence Length: 512 tokens
Number of Classes: 2 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
0	'the other day I happened to be standing taller than the roof and saw that I have two very large rust spots on the roof and tons of little rust spots on the hood and hatch. The ones on the roof look like the paint is washing away, like it was hit with acid. I went to the dealership I purchased it from today and the service tech was pretty useless, referring me to Toyota USA.' 'Anything but a Corolla. My sister got one new for her graduation, the paint started peeling after just 6 months with rust spots underneath' 'My dealer hasn’t activated the Qmerit install benefits, I guess I need to call them tomorrow, my sales guy not responding to me at all. I still need to get the paint issue taken care of, found rusted paint when taking delivery.'
1	'I have never owned a F-150. I fell in love with them in 2015 and really like the idea of a rust free body on a truck.' "2009 Honda Civic I bought it brand new, it has way over 200k miles on it and while it looks a little worn and the paint is faded there's no rust or dents or anything that would make you look twice at it. I love the damn car." 'Mine still has no rust but I take preventative measures each winter. Your paint still looks amazing.'

Label

Examples

'the other day I happened to be standing taller than the roof and saw that I have two very large rust spots on the roof and tons of little rust spots on the hood and hatch. The ones on the roof look like the paint is washing away, like it was hit with acid. I went to the dealership I purchased it from today and the service tech was pretty useless, referring me to Toyota USA.'
'Anything but a Corolla. My sister got one new for her graduation, the paint started peeling after just 6 months with rust spots underneath'
'My dealer hasn’t activated the Qmerit install benefits, I guess I need to call them tomorrow, my sales guy not responding to me at all. I still need to get the paint issue taken care of, found rusted paint when taking delivery.'

'I have never owned a F-150. I fell in love with them in 2015 and really like the idea of a rust free body on a truck.'
"2009 Honda Civic I bought it brand new, it has way over 200k miles on it and while it looks a little worn and the paint is faded there's no rust or dents or anything that would make you look twice at it. I love the damn car."
'Mine still has no rust but I take preventative measures each winter. Your paint still looks amazing.'

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("bhaskars113/toyota-corrosion")
# Run inference
preds = model("Our white Atlas CS has SHIT paint. It’s covered in rock chips and rust spots.")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	15	35.875	98

Label	Training Sample Count
0	16
1	16

Training Hyperparameters

batch_size: (8, 8)
num_epochs: (1, 1)
max_steps: -1
sampling_strategy: oversampling
num_iterations: 20
body_learning_rate: (2e-05, 2e-05)
head_learning_rate: 2e-05
loss: CosineSimilarityLoss
distance_metric: cosine_distance
margin: 0.25
end_to_end: False
use_amp: False
warmup_proportion: 0.1
l2_weight: 0.01
seed: 42
eval_max_steps: -1
load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0063	1	0.2731	-
0.3125	50	0.1076	-
0.625	100	0.0002	-
0.9375	150	0.0	-

Framework Versions

Python: 3.10.12
SetFit: 1.1.0
Sentence Transformers: 3.2.1
Transformers: 4.44.2
PyTorch: 2.4.1+cu121
Datasets: 3.0.1
Tokenizers: 0.19.1

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}
}