Edit model card

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:

  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
0
  • 'GATE ENTRY PASS\n\nae Hirakud Power - 363 12\noe pame:- (44) al: ‘bx stavelen SI.No. :- / ‘5\n\nMe sane: bey i td Bate O$ 0S) 22\npals ony poe a 7 << Fe Shift :-\n\nApproved Man Power :- feel aes Pass No. :-\n\n \n\n \n\npat\n\npetan & sé dutity\n\x0c'
  • ' \n\nLEPTH 2L09.49 Ling>@\n\nxP y\nALTAD. Catiima = —P\nDATE SEE COTUNG —RWOD_\n\n§ 26.09.17 ODM: + METAL PAD Cot ny\n\n74..9 4° o2 Wm. ‘+ -d- _ A:\n\naa 09 09. AZ OD. aw "le de. ——\e-\n\npam 29-99-19 _—Surhoy\n\nBz. 09-19 01 we d~ etre <
2
  • " \n\nSAMALESWARI CONSTRUCTION\n\nAT-BUDAKATA , PO- GADAMUNDA\nHIRAKUD, DIST: SAMBALPUR\ndetails of receipient (billed to )\nHINDALCO INDUSTRIES LTD.\nHIRAKUD POWER ,\n\n \n \n\n \n\nMOBILE NO. : 9178245293\n\n \n \n \n \n\n \n\n \n \n\n \n\nTAX INVOICE\n(ISSUEDUNDER RULE 46 OF GST/OGST RULE,2017)\n\n \n \n \n \n\nSAMBALPUR -768016\n\n \n\nINVOICE NO. SC/AP/772/2020\n\n \n \n \n \n\n \n \n \n\n21\n21AAACH1201R1ZZ\nAAACH1201R\nDETAILS OF COSIGNEE (SHIPPED }\nHINDAL CO INDUSTRIES LTD\nHIRAKUD POWER\n\n
1
  • ' \n\n \n\nGSTIN: 21AAACH1201R1ZZ\nDUSTRIES LIMITED\nHINDALCO IN eee .\nHIRAKUD POWER, HIRAKUD-768 016.DIST.SAMBALPUR (ODISHA) GST Rangeldivision: Sambelpur\nPHONE: 0663-2481365, FAX: 0663-2481342 GST Commissionerate -Cuttack\nPURCHASE ORDER\n‘AMENOMENT Z\nVendor Code: J123 P.O/No: P/PO/SRV/1920/1161 Date: 27-MAR-2020\nMis JAIDURGA CONSTRUCTION Rete ee Dater04-MAY-2020\n‘Order Type: PURCHASE ORDER\nBUDHAKATA, Effective From 01/03/2020 To 31/03/2021\nGADMUNDA Price Basis\nHIRAKUD i a ;\nMB, ISSA, 768011 ransportation arrangement\nSEA PUR OR SSN NOR roomie Ship to Location HIRAKUD - POWER\nEmail: ckppni@rediffmail.com Carrier\nFax:() Currency 2 INR\nContact: DILIP PRADHAN () 9438452293 Hindalco Contact Person: SIDDHARTH KUNDA,\nGSTIN: 21AACFJ4294P122 —State:21- Odisha Email of Contact Person: sidharth.kunda@adityabirta.com\nRef: ASH TRANSPORTATION TO VARIOUS BRICKS MANUFACTURING PLANT\nOrder Unit of Rate/Unit Value\nSl Stock No. & Descfiption ‘Quantity Measurement (Rs) (Rs)\n1 sera’ HSNISAC: 3600.00 MT 126.00" 4536000.00\nASH TRANSPORTATION TO VARIOUS BRICKS MANUFACTURING PLANT CCST Tax@2.5% 113400.\nDISTANCE TO & FRO 26KM TO 40KM Set Tego ve\nCO case Ss Gaaey SGST Tax@2.5% 113400.00\n36000.000 Need By: 31-MAR-2021 RCM CGST Tax@25% — -113400.00\n‘Supplier tom. DR RS.67 164. TR 27.03.20 RCM SGST Tax@2.5% ~113400.00\ner tem Total: —-4536000.00\n2 Sc1750 HSN/SAC: 200.000 MT _7200_¥~ 144000.00\nASH TRANSPORTATION TO VARIOUS BRICKS MANUFACTURING PLANT 1 ~ 3600.\nDISTANCE TO & FRO 11KM TO 15KM cease on\n= ees SGST Tax@2.5% 3600.00\n200,000 ‘Need By: 31-MAR-2021 RCM CGST Tax@2.5% -3600.00\nSupplier tem. D.R.RS.67.16/ TR 27 03 20 RCM SGST Tax@2.5% -3600.00\ntem Tota: 144000.00\n3 sciTsa HSNISAC: 2000.00 MT 96.00 192000.00\nCC Code Quantity SGST Tax@2.5% 4800.00\n200.000 Need By: 31-MAR-2021

Evaluation

Metrics

Label Accuracy
all 0.9977

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/SERVICE_LARGE_MODEL_ZEON")
# Run inference
preds = model(" 

TOTAL
11

- wl et

SUPERVI
SOR

7 ce

      

nly
AIN|A ale
Sale
lale ld
So

 

 

:

9 wij im

 

 

   

aes 3513
sIB|e
alg
alg

NTN

a 2 3 ; 3
gle

o

ri

  
 

 

 
")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 1 225.8451 1106
Label Training Sample Count
0 267
1 74
2 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
0.0003 1 0.3001 -
0.0164 50 0.2586 -
0.0328 100 0.1809 -
0.0492 150 0.0534 -
0.0656 200 0.0285 -
0.0820 250 0.0144 -
0.0985 300 0.0045 -
0.1149 350 0.0281 -
0.1313 400 0.0432 -
0.1477 450 0.0045 -
0.1641 500 0.0023 -
0.1805 550 0.0022 -
0.1969 600 0.0011 -
0.2133 650 0.0008 -
0.2297 700 0.0226 -
0.2461 750 0.0009 -
0.2626 800 0.0008 -
0.2790 850 0.001 -
0.2954 900 0.001 -
0.3118 950 0.001 -
0.3282 1000 0.0007 -
0.3446 1050 0.0012 -
0.3610 1100 0.0008 -
0.3774 1150 0.0008 -
0.3938 1200 0.0008 -
0.4102 1250 0.0034 -
0.4266 1300 0.0007 -
0.4431 1350 0.0007 -
0.4595 1400 0.0008 -
0.4759 1450 0.0007 -
0.4923 1500 0.0004 -
0.5087 1550 0.0005 -
0.5251 1600 0.0007 -
0.5415 1650 0.0005 -
0.5579 1700 0.0005 -
0.5743 1750 0.0004 -
0.5907 1800 0.0009 -
0.6072 1850 0.0025 -
0.6236 1900 0.0003 -
0.6400 1950 0.0023 -
0.6564 2000 0.0004 -
0.6728 2050 0.0045 -
0.6892 2100 0.0005 -
0.7056 2150 0.0109 -
0.7220 2200 0.0003 -
0.7384 2250 0.0021 -
0.7548 2300 0.0005 -
0.7713 2350 0.0004 -
0.7877 2400 0.0118 -
0.8041 2450 0.0003 -
0.8205 2500 0.0003 -
0.8369 2550 0.0126 -
0.8533 2600 0.0004 -
0.8697 2650 0.0162 -
0.8861 2700 0.0003 -
0.9025 2750 0.0004 -
0.9189 2800 0.0005 -
0.9353 2850 0.0004 -
0.9518 2900 0.0032 -
0.9682 2950 0.0003 -
0.9846 3000 0.0004 -
1.0010 3050 0.0003 -
1.0174 3100 0.0003 -
1.0338 3150 0.0019 -
1.0502 3200 0.0194 -
1.0666 3250 0.0003 -
1.0830 3300 0.0004 -
1.0994 3350 0.01 -
1.1159 3400 0.0002 -
1.1323 3450 0.0003 -
1.1487 3500 0.0004 -
1.1651 3550 0.0004 -
1.1815 3600 0.0002 -
1.1979 3650 0.0005 -
1.2143 3700 0.0002 -
1.2307 3750 0.0019 -
1.2471 3800 0.0003 -
1.2635 3850 0.0048 -
1.2799 3900 0.013 -
1.2964 3950 0.0031 -
1.3128 4000 0.0002 -
1.3292 4050 0.0024 -
1.3456 4100 0.0002 -
1.3620 4150 0.0003 -
1.3784 4200 0.0003 -
1.3948 4250 0.0002 -
1.4112 4300 0.003 -
1.4276 4350 0.0002 -
1.4440 4400 0.0002 -
1.4605 4450 0.0022 -
1.4769 4500 0.0002 -
1.4933 4550 0.0078 -
1.5097 4600 0.0027 -
1.5261 4650 0.0002 -
1.5425 4700 0.0002 -
1.5589 4750 0.0002 -
1.5753 4800 0.0002 -
1.5917 4850 0.0002 -
1.6081 4900 0.0118 -
1.6245 4950 0.0002 -
1.6410 5000 0.0002 -
1.6574 5050 0.0003 -
1.6738 5100 0.0003 -
1.6902 5150 0.0068 -
1.7066 5200 0.0003 -
1.7230 5250 0.0112 -
1.7394 5300 0.0002 -
1.7558 5350 0.0002 -
1.7722 5400 0.0003 -
1.7886 5450 0.0002 -
1.8051 5500 0.0002 -
1.8215 5550 0.0002 -
1.8379 5600 0.0002 -
1.8543 5650 0.0003 -
1.8707 5700 0.0047 -
1.8871 5750 0.0121 -
1.9035 5800 0.0003 -
1.9199 5850 0.013 -
1.9363 5900 0.005 -
1.9527 5950 0.0001 -
1.9691 6000 0.0002 -
1.9856 6050 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}
}
Downloads last month
1
Safetensors
Model size
33.4M params
Tensor type
F32
·
Inference API
This model does not have enough activity to be deployed to Inference API (serverless) yet. Increase its social visibility and check back later, or deploy to Inference Endpoints (dedicated) instead.

Finetuned from

Evaluation results