Ramyashree's picture
Add SetFit model
b1990b9
metadata
library_name: setfit
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
datasets:
  - Ramyashree/Dataset-train500-test100
metrics:
  - accuracy
widget:
  - text: I weant to use my other  account, switch them
  - text: I can't remember my password, help me reset it
  - text: the game was postponed and i wanna get a reimbursement
  - text: where to change to another online account
  - text: the show was cancelled, get a reimbursement
pipeline_tag: text-classification
inference: true
base_model: thenlper/gte-large
model-index:
  - name: SetFit with thenlper/gte-large
    results:
      - task:
          type: text-classification
          name: Text Classification
        dataset:
          name: Ramyashree/Dataset-train500-test100
          type: Ramyashree/Dataset-train500-test100
          split: test
        metrics:
          - type: accuracy
            value: 1
            name: Accuracy

SetFit with thenlper/gte-large

This is a SetFit model trained on the Ramyashree/Dataset-train500-test100 dataset that can be used for Text Classification. This SetFit model uses thenlper/gte-large 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 Sources

Model Labels

Label Examples
create_account
  • "I don't have an online account, what do I have to do to register?"
  • 'can you tell me if i can regisger two accounts with a single email address?'
  • 'I have no online account, open one, please'
edit_account
  • 'how can I modify the information on my profile?'
  • 'can u ask an agent how to make changes to my profile?'
  • 'I want to update the information on my profile'
delete_account
  • 'can I close my account?'
  • "I don't want my account, can you delete it?"
  • 'how do i close my online account?'
switch_account
  • 'I would like to use my other online account , could you switch them, please?'
  • 'i want to use my other online account, can u change them?'
  • 'how do i change to another account?'
get_invoice
  • 'what can you tell me about getting some bills?'
  • 'tell me where I can request a bill'
  • 'ask an agent if i can obtain some bills'
get_refund
  • 'the game was postponed, help me obtain a reimbursement'
  • 'the game was postponed, what should I do to obtain a reimbursement?'
  • 'the concert was postponed, what should I do to request a reimbursement?'
payment_issue
  • 'i have an issue making a payment with card and i want to inform of it, please'
  • 'I got an error message when I attempted to pay, but my card was charged anyway and I want to notify it'
  • 'I want to notify a problem making a payment, can you help me?'
check_refund_policy
  • "I'm interested in your reimbursement polivy"
  • 'i wanna see your refund policy, can u help me?'
  • 'where do I see your money back policy?'
recover_password
  • 'my online account was hacked and I want tyo get it back'
  • "I lost my password and I'd like to retrieve it, please"
  • 'could u ask an agent how i can reset my password?'
track_refund
  • 'tell me if my refund was processed'
  • 'I need help checking the status of my refund'
  • 'I want to see the status of my refund, can you help me?'

Evaluation

Metrics

Label Accuracy
all 1.0

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("Ramyashree/gte-large-train-test-2")
# Run inference
preds = model("where to change to another online account")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 3 10.258 24
Label Training Sample Count
check_refund_policy 50
create_account 50
delete_account 50
edit_account 50
get_invoice 50
get_refund 50
payment_issue 50
recover_password 50
switch_account 50
track_refund 50

Training Hyperparameters

  • batch_size: (16, 16)
  • num_epochs: (1, 1)
  • max_steps: -1
  • sampling_strategy: oversampling
  • num_iterations: 20
  • 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
0.0008 1 0.3248 -
0.04 50 0.1606 -
0.08 100 0.0058 -
0.12 150 0.0047 -
0.16 200 0.0009 -
0.2 250 0.0007 -
0.24 300 0.001 -
0.28 350 0.0008 -
0.32 400 0.0005 -
0.36 450 0.0004 -
0.4 500 0.0005 -
0.44 550 0.0005 -
0.48 600 0.0006 -
0.52 650 0.0005 -
0.56 700 0.0004 -
0.6 750 0.0004 -
0.64 800 0.0002 -
0.68 850 0.0003 -
0.72 900 0.0002 -
0.76 950 0.0002 -
0.8 1000 0.0003 -
0.84 1050 0.0002 -
0.88 1100 0.0002 -
0.92 1150 0.0003 -
0.96 1200 0.0003 -
1.0 1250 0.0003 -

Framework Versions

  • Python: 3.10.12
  • SetFit: 1.0.1
  • Sentence Transformers: 2.2.2
  • Transformers: 4.35.2
  • PyTorch: 2.1.0+cu121
  • Datasets: 2.15.0
  • Tokenizers: 0.15.0

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}
}