README.md

metadata

tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: Add a navigation menu text
  - text: Change to square format
  - text: Change the button text
  - text: Make the object stand alone
  - text: Mirror the logo vertically
metrics:
  - accuracy
pipeline_tag: text-classification
library_name: setfit
inference: true
base_model: nomic-ai/nomic-embed-text-v1.5
model-index:
  - name: SetFit with nomic-ai/nomic-embed-text-v1.5
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Unknown
          type: unknown
          split: test
        metrics:
          - type: accuracy
            value: 0.5353535353535354
            name: Accuracy

SetFit with nomic-ai/nomic-embed-text-v1.5

This is a SetFit model that can be used for Text Classification. This SetFit model uses nomic-ai/nomic-embed-text-v1.5 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: nomic-ai/nomic-embed-text-v1.5
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 8192 tokens
• Number of Classes: 63 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	'Add a corporate presentation background' 'Insert a modern icon set for the design' 'Add a mountain landscape background to the page'
1	'Find me some shape options for this design' 'I need some professional-looking assets' 'Can you recommend some images that would work well here?'
2	'Add a date and time for the event' 'Insert a disclaimer text' 'Add a navigation menu text'
3	'Distribute the icons evenly' 'Align all the text elements to the left' 'Align the footer elements'
4	'Make the button pulse' 'Add a flip animation' 'Make the text glow'
5	'Make everything fade in gradually' 'Make the page bounce in from the top' 'Add a spiral animation to the page'
6	'Change the building color' 'Remove the unwanted text overlay' 'Add a party hat to the dog'
7	'Remove the draft image' 'Delete the backup copy' 'Remove the unwanted image'
8	'What image editing tools do you have?' 'How do I create a template?' 'How can I align elements properly?'
9	'Distribute the buttons around the center image' 'Place the elements in a circular arrangement' 'Arrange the images in a circular layout'
10	'Duplicate the design' 'Make a second version' 'Create a new version'
11	'Duplicate the icon and move it' 'Duplicate the text and align it differently' 'Copy the image and apply a filter'
12	'Duplicate the logo to page 3' 'Copy the text to the last page' 'Copy the navigation to the next page'
13	'Fix the typographic errors' 'Improve the text flow' 'Fix the font consistency'
14	'Mirror the icon horizontally' 'Mirror the logo vertically' 'Flip the image horizontally'
15	'Create a photo of a futuristic city' 'Generate a picture of a tropical beach' 'Generate a picture of a cat playing with yarn'
16	'Create a card for a birthday party' 'Create a flyer for a happy birthday party' 'Generate an Instagram post for a birthday'
17	'Group the navigation elements' 'Combine the shape and text' 'Combine the image and overlay'
18	'Move the shape to the bottom' 'Position the icon at (0, 0)' 'Move the shape to coordinates (100, 100)'
19	'Add a sepia tone effect' 'Apply a modern filter' 'Make the image black and white'
20	'Suggest some shape designs' 'Show me pattern options' 'Find me border designs'
21	'Restore the previous opacity' 'Redo the text edit' 'Restore the previous color'
22	'Make the image have no background' 'Remove the background from the animal' 'Remove the background from the item'
23	'Delete the unwanted text' 'Remove the person from the background' 'Remove the graffiti'
24	'Replace the illustration' 'Change the background image' 'Change the product photo'
25	'Change the button text' 'Update the navigation text' 'Update the description'
26	'Remove all modifications' 'Restore the original colors' 'Remove all effects from the image'
27	'Scale the text up' 'Make the shape smaller' 'Reduce the shape size'
28	'Change to square format' 'Change to poster size' 'Make the page smaller'
29	'Rotate the text 45 degrees' 'Rotate the text 15 degrees' 'Turn the image counterclockwise'
30	'Distribute the shapes randomly' 'Distribute the leaves randomly on the page' 'Scatter the confetti around the design'
31	'Select the sidebar elements' 'Select the background image' 'Select the footer content'
32	'Change to a dark background' 'Make the background transparent' 'Change to a neutral background'
33	'Set the blend mode to soft light' 'Set the blend mode to darken' 'Set the blend mode to normal'
34	'Add a depth of field blur' 'Add a directional blur' 'Add a soft blur effect'
35	'Change the border style to solid' 'Change the border to dotted' 'Add a double border to the image'
36	'Brighten the shadows' 'Make the image brighter' 'Brighten the highlights'
37	'Bring the title to the front' 'Bring the shape to the front' 'Move the logo to the top layer'
38	'Make the image more intense' 'Enhance the contrast ratio' 'Increase the tonal range'
39	'Make the image round' 'Make the image rectangular' 'Crop to a heart shape'
40	'Add a sharp drop shadow' 'Add a gradient shadow' 'Add a hard shadow edge'
41	'Fill the shape with orange' 'Fill the element with brown' 'Change the shape color to blue'
42	'Make the text more readable' 'Increase the heading size' 'Make the description smaller'
43	'Make the text bold and italic' 'Add strikethrough to the text' 'Add bold to the title'
44	'Use a contemporary font' 'Use a professional font' 'Use a serif font for the heading'
45	'Make the bright areas brighter' 'Make the highlights more prominent' 'Enhance the bright spots'
46	'Set the image as background layer' 'Set the picture as background fill' 'Make the photo cover the background'
47	'Reduce the character spacing' 'Add letter spacing to the logo' 'Increase spacing between characters'
48	'Make the lines tighter' 'Spread out the text lines' 'Increase the paragraph spacing'
49	'Reduce the opacity of the overlay' 'Increase the transparency of the image' 'Make the shape more opaque'
50	'Make the paragraphs closer together' 'Increase the text block spacing' 'Reduce the paragraph spacing'
51	'Make the colors more intense' 'Increase the color depth' 'Make the image more colorful'
52	'Darken the shadows in the image' 'Increase the shadow intensity' 'Enhance the shadow depth'
53	'Enhance the image clarity' 'Increase the image sharpness' 'Sharpen the image details'
54	'Center the button text' 'Justify the paragraph text' 'Center the title text'
55	'Create a border around the text' 'Add a glow effect behind the text' 'Add a colored background to the text'
56	'Create text in a radial pattern' 'Create text that follows a circle' 'Create text that follows a shape'
57	'Convert to a bulleted list' 'Make the text into bullet points' 'Make the text into a list with bullets'
58	'Create a soft shadow behind the text' 'Add a dramatic shadow effect' 'Create a shadow for the text'
59	'Add warm undertones to the photo' 'Add warm color grading' 'Make the photo more golden hour'
60	'I need to add my own image' 'Open the image upload tool' 'Show me how to upload files'
61	'Revert the color change' 'Undo the last modification' 'Undo the text edit'
62	'Separate the grouped components' 'Ungroup the combined elements' 'Break up the grouped objects'

Evaluation

Metrics

Label	Accuracy
all	0.5354

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("setfit_model_id")
# Run inference
preds = model("Change the button text")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	3	5.1778	10

Label	Training Sample Count
0	5
1	5
2	5
3	5
4	5
5	5
6	5
7	5
8	5
9	5
10	5
11	5
12	5
13	5
14	5
15	5
16	5
17	5
18	5
19	5
20	5
21	5
22	5
23	5
24	5
25	5
26	5
27	5
28	5
29	5
30	5
31	5
32	5
33	5
34	5
35	5
36	5
37	5
38	5
39	5
40	5
41	5
42	5
43	5
44	5
45	5
46	5
47	5
48	5
49	5
50	5
51	5
52	5
53	5
54	5
55	5
56	5
57	5
58	5
59	5
60	5
61	5
62	5

Training Hyperparameters

• batch_size: (64, 64)
• num_epochs: (1, 1)
• 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
• l2_weight: 0.01
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0007	1	0.1545	-
0.0328	50	0.1187	-
0.0655	100	0.0521	-
0.0983	150	0.0208	-
0.1311	200	0.0123	-
0.1638	250	0.0096	-
0.1966	300	0.0056	-
0.2294	350	0.0036	-
0.2621	400	0.0027	-
0.2949	450	0.0017	-
0.3277	500	0.0007	-
0.3604	550	0.0009	-
0.3932	600	0.0009	-
0.4260	650	0.0003	-
0.4587	700	0.0003	-
0.4915	750	0.0004	-
0.5242	800	0.0004	-
0.5570	850	0.0002	-
0.5898	900	0.0001	-
0.6225	950	0.0001	-
0.6553	1000	0.0001	-
0.6881	1050	0.0001	-
0.7208	1100	0.0001	-
0.7536	1150	0.0001	-
0.7864	1200	0.0001	-
0.8191	1250	0.0001	-
0.8519	1300	0.0001	-
0.8847	1350	0.0001	-
0.9174	1400	0.0001	-
0.9502	1450	0.0001	-
0.9830	1500	0.0001	-

Framework Versions

• Python: 3.12.11
• SetFit: 1.1.3
• Sentence Transformers: 5.1.0
• Transformers: 4.54.1
• PyTorch: 2.7.1
• Datasets: 4.0.0
• Tokenizers: 0.21.4

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}
}