SetFit

This is a SetFit model that can be used for Text Classification. 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
Classification head: a LogisticRegression instance
Maximum Sequence Length: 512 tokens
Number of Classes: 4 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
1	' \n\n \n\n \n\n \n\nContact: PRAMOD KHNADWAL (+91) 9937148709\n\n
2	' \n\nNASFUND FINANCE\nRECEIVED\n\nDate 2/04 Time: kD" 7 jee\nInitials: E- r
3	'VI ww\n~~ Sponsorship Agreement EMTV National News Market Report 2024 nasfund\n2p onsorsiip Agree reer REE REDON 2026 = see\n\nEMTV National News Market Report Naming Rights Sponsorship Agreement\n\n“The Agreement”\nbetween\nMedia Niugini Limited\nand\nNasfund Limited\n1. Agreement Date\n\nAF Ocf\n\nThe Agreement is made on the\n\n2. Purpose\n\nThe Agreement outlines the understandings (Key Terms and Conditions of engagement)\nbetween Media Niugini Limited (MNL), trading as “EMTV” (“The Broadcaster’), and\nNasfund Limited (“The Client”), collectively referred to as “The Parties”.\n\n3. Key Terms and Conditions\n\n3.1 Agreement Period\n\nStarting on 4" January 2024 and ending on 31st December 2024 inclusive.\n\n3.2 Client's Products & Services\n\nThe Client to provide to the Broadcaster copies of 30 second TVCs, with rights to\nchange/rotate monthly and full rights to use (non-exclusive) on EMTV channel for a\nperiod of twelve (12) months, on 1% January 2024 and ending on 31st December\n2024. It is the responsibility of the Client to inform the Broadcaster in writing of all\nproducts to be advertised under this agreement.\n\nThe Client to issue a Company Purchase Order for all EMTV services to be engaged\nPrior to every rhonth's Airtime booking. In the event that a Company Purchase Order\nis unable to be issued prior to the agreed broadcast date, a signed “Authority to\nTelecast” (ATT) is required prior to the agreed broadcast date.\n\x0c' 'N\n\nMEDIA NIUGINI LIMITED\nP O Box 443\nBoroko, NCD.\nPort Moresby\nPNG\n\nTV\n\nSOLD NASFUND\n\nTO: P.O.BOX 5791\nBOROKO, NCD\nPAPUA NEW GUINEA\n\nATTN: ACCOUNTS PAYABLE\n\nINV0000961081 31/03/2024 SIGNED AGREEMENT\n\nIN - Invoice\n\nDB - Debit Note\n\nCR - Credit Note\n\nIT - Interest Payable\n\nPY - Applied\nReceipt\n\nED - Eamed\nDiscount\nAD -\n\nUC - Unapplied\nCash\nRF - Refund\n\nPlease pay amount showing.\n\n1 - 30 DAYS O/DUE 31 - 60 DAYS O/DUE\n17,325.00 0.00 0.00\n\n \n\n61 - 90 DAYS O/DUE\n\nCUSTOMER NO.:\n\nREMIT TO ADDRESS:\n\nMedia Niugini Ltd\n\nPO Box 443\n\nBoroko, 411, PNG\n\nPlease Deposit into: Westpac\nAccount : 609736001\n\n31/03/2024\n\nTotal:\nCredit Limit:\nCredit Available:\n\n0.00\n\n \n \n\n \n \n\nNASO01\n1\n3/31/2024\n\n \n \n\n \n\n17,325.00\n\n0.00\n0.00\n\nOVER 90 DAYS O/DUE\n\n \n\x0c' "nasfund'\n\nNATIONAL SUPERANNUATION FUND LIMITED\n(CLIENT)\nAND\nTUMPI SECURITY SERVICE\n(CONTRACTOR)\n\nCONTRACT FOR SECURITY SERVICES\n\n \n\nNational Superannuation Fund Limited\nLevel 3, BSP Haus\nHarbour City\nP O Box 5791\nBoroko, NCD\nPapua New Guinea\nPhone: (675) 313 1829\n\x0c"
0	'WATERFRONT\n\nFOODWORLD\n\nP.O.BOX 889, KONEDOBU, NATIONAL CAPITAL DISTRICT\nTELEPHONE: (675) 320 0460 OR (675) 305 8600\nEMAIL: grocery.<vaterfront@garamut.com.pg\n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\n \n\nDATE:\nae 4/03/2024 11:37:26 AM\nINVOICE\nCUSTOMER: WFI141 ABABA MAUREEN\nADDRESS: LEVEL3BSPHAUS HARBOUR CITY NCD 3131929\n[ Cashier P.0 Register eS CUseEPSiemale NASEUND\n#\nBECKY SIMON

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("Adreno99/nasfund_setfit_pretrained")
# Run inference
preds = model("nasfund este

 

Accounts Payable

 

IT Division

 

05 March 2024

 

2 x Laptops for nasfund Lae

 

 

Datec

 

2 x Laptops for nasfund Lae. One will be issued to their new Employer Service
officer and the other one will be used as their ID laptop when they go onsite to
employers.

 

K10,532.52

 

PO

 

 

 

Attached Quote from Datec

Requested by:
WK,

Asaph Kuliniasi

Help Desk Analyst

05 March 2024

Recommended By:

Samuel Topotol |
TL Infrastructure
05 March 2024

 

 

05 March 2024
")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	26	212.4310	640

Label	Training Sample Count
0	18
1	10
2	8
3	22

Training Hyperparameters

batch_size: (32, 32)
num_epochs: (2, 2)
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
0.0133	1	0.1389	-
0.6667	50	0.0091	-
1.3333	100	0.0021	-
2.0	150	0.0025	-

Framework Versions

Python: 3.10.12
SetFit: 1.0.3
Sentence Transformers: 3.0.1
Transformers: 4.39.0
PyTorch: 2.3.0+cu121
Datasets: 2.20.0
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}
}

Adreno99
/

nasfund_setfit_pretrained

You need to agree to share your contact information to access this model