README.md

metadata

library_name: setfit
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
metrics:
  - accuracy
widget:
  - text: Upgrade all installed packages with superuser privileges
  - text: Install package 'vim' as superuser
  - text: Remove package 'firefox' with superuser privileges
  - text: Change permissions of directory 'docs' to writable
  - text: Update package lists using superuser privileges
pipeline_tag: text-classification
inference: true
base_model: sentence-transformers/paraphrase-mpnet-base-v2
model-index:
  - name: SetFit with sentence-transformers/paraphrase-mpnet-base-v2
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 0
            name: Accuracy

SetFit with sentence-transformers/paraphrase-mpnet-base-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-mpnet-base-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:

1. Fine-tuning a Sentence Transformer with contrastive learning.
2. 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-mpnet-base-v2
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 512 tokens
• Number of Classes: 30 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
ls	'List all files and directories' 'Show files in the current directory' 'Display contents of the current directory'
cd	'Change to the specified directory' 'Move to the home directory' 'Navigate to the specified directory path'
mkdir docs	"Create a new directory named 'docs'"
mkdir projects	"Make a directory named 'projects'"
mkdir data	"Create a folder called 'data'"
mkdir images	"Make a directory named 'images'"
mkdir scripts	"Create a new folder named 'scripts'"
rm example.txt	"Remove the file named 'example.txt'"
rm temp.txt	"Delete the file called 'temp.txt'"
rm file1	"Remove the file named 'file1'"
rm file2	"Delete the file named 'file2'"
rm backup.txt	"Remove the file named 'backup.txt'"
cp file1 /destination	'Copy file1 to directory /destination'
cp file2 /backup	'Duplicate file2 to directory /backup'
cp file3 /archive	'Copy file3 to folder /archive'
cp file4 /temp	'Duplicate file4 to folder /temp'
cp file5 /images	'Copy file5 to directory /images'
mv file2 /new_location	'Move file2 to directory /new_location'
mv file3 /backup	'Transfer file3 to directory /backup'
mv file4 /archive	'Move file4 to folder /archive'
mv file5 /temp	'Transfer file5 to folder /temp'
mv file6 /images	'Move file6 to directory /images'
cat README.md	"Display the contents of file 'README.md'"
cat notes.txt	"Show the content of file 'notes.txt'"
cat data.csv	"Print the contents of file 'data.csv'"
cat script.sh	"Display the content of file 'script.sh'"
cat config.ini	"Show the contents of file 'config.ini'"
grep 'pattern' data.txt	"Search for 'pattern' in file 'data.txt'"
grep 'word' text.txt	"Find occurrences of 'word' in file 'text.txt'"
grep 'keyword' document.txt	"Search for 'keyword' in file 'document.txt'"

Evaluation

Metrics

Label	Accuracy
all	0.0

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("souvenger/NLP2Linux")
# Run inference
preds = model("Install package 'vim' as superuser")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	5	5.6667	9

Label	Training Sample Count
cat README.md	1
cat config.ini	1
cat data.csv	1
cat notes.txt	1
cat script.sh	1
cd	10
cp file1 /destination	1
cp file2 /backup	1
cp file3 /archive	1
cp file4 /temp	1
cp file5 /images	1
grep 'keyword' document.txt	1
grep 'pattern' data.txt	1
grep 'word' text.txt	1
ls	10
mkdir data	1
mkdir docs	1
mkdir images	1
mkdir projects	1
mkdir scripts	1
mv file2 /new_location	1
mv file3 /backup	1
mv file4 /archive	1
mv file5 /temp	1
mv file6 /images	1
rm backup.txt	1
rm example.txt	1
rm file1	1
rm file2	1
rm temp.txt	1

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
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0042	1	0.1215	-
0.2083	50	0.0232	-
0.4167	100	0.01	-
0.625	150	0.0044	-
0.8333	200	0.0025	-

Framework Versions

• Python: 3.10.13
• SetFit: 1.0.3
• Sentence Transformers: 2.3.1
• Transformers: 4.37.0
• PyTorch: 2.1.2
• Datasets: 2.1.0
• Tokenizers: 0.15.1

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}
}