SetFit with BAAI/bge-small-en-v1.5

This is a SetFit model that can be used for Text Classification. This SetFit model uses BAAI/bge-small-en-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:

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: BAAI/bge-small-en-v1.5
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	'. : Pt oh mM\nBaw iS\n\nWw tere\nPr pe 0 ok ji the\nFw: Pending Bills jer\n\n, Ronit Sarangi to: Vinit K Sinha i 22-01-2020 11:37\n
1	'Tax Invoice\nOriginal for Buyer/ Duplicate for Transporter/ Triplicate for Assessee\n\nSupplier Legal Name; Mahanadi Coalfields Area Code :MO01\n7 Limited Area Description :Jagannath\nSupplier Address , Jagriti Vihar, Bur.a Invaice Number 19100066504\nSambalpur 768020 Involee Date :Dee 15, 2022\nSupplier City : Sambalpur Contract Reference: 3030007756\nSupplier State Odisha Contract type :Spot Auction\nSupplier Pincode : 768020 Salas Order 1240002677\n.P.O. - dJagriti vihar, Burla

Label

Examples

'. : Pt oh mM\nBaw iS\n\nWw tere\nPr pe 0 ok ji the\nFw: Pending Bills jer\n\n, Ronit Sarangi to: Vinit K Sinha i 22-01-2020 11:37\n

'Tax Invoice\nOriginal for Buyer/ Duplicate for Transporter/ Triplicate for Assessee\n\nSupplier Legal Name; Mahanadi Coalfields Area Code :MO01\n7 Limited Area Description :Jagannath\nSupplier Address , Jagriti Vihar, Bur.a Invaice Number 19100066504\nSambalpur 768020 Involee Date :Dee 15, 2022\nSupplier City : Sambalpur Contract Reference: 3030007756\nSupplier State Odisha Contract type :Spot Auction\nSupplier Pincode : 768020 Salas Order 1240002677\n.P.O. - dJagriti vihar, Burla

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("Gopal2002/COAL_INVOICE_ZEON")
# Run inference
preds = model("UNITED MEDICAL STORE Patient Name: KASTURI uENA
‘EW MARKET, C/O PRAFULLA KUMAR JENA
HIRAKUD. SAMBALPUR. Dr. Name :

Medicine Advice Slip: MA/2223/0668 “
Phone :0663-2431670 Prescription Indent:M/2223/06299

DL No. :SAWZ 486 R/487 RC Invoice No. ; 0002785 Date : 21/11/2022

Se|__Qiy. [Pack [Product “Batch [Exp] HSN [ MRP | Table | Dis [5051] CO3i] Amount |

1. 30 TAB] 30'S TELMA H TAB 11/24 | 30049099; 484.00! 432.14 0.001 6.00
NEOPRIDE TOTAL CAP 7/24 30049099) 445.00) 0,00; 6.00

 

 

 

SUB TOTAL :

SGST
er rH 2 ROFF :
— ha GRAND TOTAL

Te & Con itions For UNITED MEDICAL STORE R a ah
BILL GRAND TOTAL IS CALCULATED ACCORDING TO 1D- 3306 Im- 1220
MRP PRICE ( INCLUDING ALL GST TAXES ) Q _ 06 (ped)

 
")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	1	270.5442	4241

Label	Training Sample Count
0	130
1	85

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
0.0013	1	0.2394	-
0.0657	50	0.1203	-
0.1314	100	0.0095	-
0.1971	150	0.0029	-
0.2628	200	0.0014	-
0.3285	250	0.0014	-
0.3942	300	0.0011	-
0.4599	350	0.0009	-
0.5256	400	0.0008	-
0.5913	450	0.0007	-
0.6570	500	0.0008	-
0.7227	550	0.0008	-
0.7884	600	0.0006	-
0.8541	650	0.0005	-
0.9198	700	0.0004	-
0.9855	750	0.0005	-
1.0512	800	0.0004	-
1.1170	850	0.0005	-
1.1827	900	0.0004	-
1.2484	950	0.0004	-
1.3141	1000	0.0003	-
1.3798	1050	0.0004	-
1.4455	1100	0.0004	-
1.5112	1150	0.0004	-
1.5769	1200	0.0005	-
1.6426	1250	0.0004	-
1.7083	1300	0.0003	-
1.7740	1350	0.0004	-
1.8397	1400	0.0005	-
1.9054	1450	0.0004	-
1.9711	1500	0.0003	-

Framework Versions

Python: 3.10.12
SetFit: 1.0.3
Sentence Transformers: 2.2.2
Transformers: 4.35.2
PyTorch: 2.1.0+cu121
Datasets: 2.16.1
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}
}

Gopal2002
/

COAL_INVOICE_ZEON