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 Sources
Model Labels
| Label |
Examples |
| 0 |
- 'GHt Sa [OI uco Bank\n\nvis Free Number: 180¢-193-9125 _\n\nDICGC INSURANCE UPTO 5 LAC\n\nBRANCH\n\nUCO Bank\nP NT\n(1) Consuitants are requested to note that all moneys\n\nremitted to the Bank should either be sent by Registered\nPost or handed over to the Cash Department, as no\n\n \n\nUco BANK\n\naq\n\nName\n\n"IFSC: uceaocotms\n' dress\n\nKICK Code: 7428029504\n\nHIRAKUD\n-HIRAKED BRARCH HIRAKUS\nProae:\n\nindividual (s) outside. the Cash Department has/have JHARY EIST\nauthority to receive cash. KADAMPOLA\n(2) The account-holder should insist on delivery of Pass Book HERAKUD\n‘ made uptodate as far as possible on the same date; a 6.8%\n- otherwise he should obtain a receipt indicating when the HIRAKYD PIN .#oBlss\nPass Book will be delivered.\n(3) Deposit Rules in vogue can be obtained by account-holder TET. WaT / Asst.\nfrom the Branch on request Q28501 19027145\nPB.NG. }\n\n \n\n \n\n \n\nfe er ee me\n\n \n\x0c'
- ' \n\n= 2, ip\nO ~\nN 2\na\n: Y ve re ty\n) 3 x.\nNai] (F) my\n\n \n\ny Viayal chat aloala\nSH PPP ea [sys sys *\nas NB\n2\n=\n\ni x X we\na. = Xt +\n— W\nx 2
|
| 1 |
- '. Ae - PR CSeathetn & 3)\n" J She ase 9 Pao\n\n‘s lad Bank Afr o Steppe\nIN
|
Evaluation
Metrics
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
model = SetFitModel.from_pretrained("Gopal2002/CASH_AND_BANK_INVOICE")
preds = model("
")
Training Details
Training Set Metrics
| Training set |
Min |
Median |
Max |
| Word count |
1 |
201.2534 |
4241 |
| Label |
Training Sample Count |
| 0 |
113 |
| 1 |
33 |
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 |
| 0.0023 |
1 |
0.3054 |
- |
| 0.1142 |
50 |
0.1162 |
- |
| 0.2283 |
100 |
0.0043 |
- |
| 0.3425 |
150 |
0.0015 |
- |
| 0.4566 |
200 |
0.0014 |
- |
| 0.5708 |
250 |
0.0008 |
- |
| 0.6849 |
300 |
0.0013 |
- |
| 0.7991 |
350 |
0.001 |
- |
| 0.9132 |
400 |
0.0004 |
- |
| 1.0274 |
450 |
0.0008 |
- |
| 1.1416 |
500 |
0.0008 |
- |
| 1.2557 |
550 |
0.0011 |
- |
| 1.3699 |
600 |
0.0008 |
- |
| 1.4840 |
650 |
0.0007 |
- |
| 1.5982 |
700 |
0.0005 |
- |
| 1.7123 |
750 |
0.0005 |
- |
| 1.8265 |
800 |
0.0007 |
- |
| 1.9406 |
850 |
0.0005 |
- |
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}
}
