Hugging Face's logo

faodl
/
model_g20_multilabel_MiniLM-L12-v_final

Text Classification
setfit
Safetensors
sentence-transformers
bert
generated_from_setfit_trainer
text-embeddings-inference
Model card Files
xet
Community
model_g20_multilabel_MiniLM-L12-v_final / README.md
lgsilvaesilva's picture
lgsilvaesilva
Push model using huggingface_hub.
3432a3f verified
preview code
|
raw
history blame contribute delete
15.6 kB
metadata 
tags:
  - setfit
  - sentence-transformers
  - text-classification
  - generated_from_setfit_trainer
widget:
  - text: >-
      4.3.3 Strategies for Comprehensive Sexuality Education and (CSE)
      Youth-friendly Health Services 1. To promote volunteerism as a tool for
      fostering active participation of young people in national development; 5.
      To promote volunteerism as a tool for fostering active participation of
      young people in national development; 5.
  - text: >-
      4) Mainstream appropriate food and nutrition issues in relevant sector
      policies and strategies. 4) Mainstream appropriate food and nutrition
      issues in relevant sector policies and strategies. ), these and many
      others have varying requirements related to 3.5 Communication Support for
      Food and Nutrition Programmes and Interventions National Food and
      Nutrition Strategic Plan 2011-2015 11 generation of demand by the
      population.
  - text: >-
      incidence of stunting reduced from 39 to 35 percent, and population with
      calories deficit from 35 to 31 percent) and public food distribution ( i.e
      from 20 thousand MT to 39 thousand MT and food sales by 29 thousand MT).
      It states that “the main objective of the food security plan is to make
      the life of the targeted people healthy and productive by improving
      national food sovereignty and the food and nutrition situation.”
      Accordingly, the TYIP set out and scaled up the quantities targets in
      terms of per capita food production (i.e., from 280–289 kg per capita
      annually), indicators of nutrition ( i.e. Food procurement policy should
      be made as a vehicle of ensuring sufficient supply of essential food items
      and also a means of containing prices.
  - text: >-
      UP-5978 “On additional measures to support the public, economic3 April
      2020: Presidential Decree No. Tax benefitsTax benefits The Decree 5969,
      the Decree 5978, and the Decree 5986 (together the “Decrees”) have
      introduced the followingThe Decree 5969, the Decree 5978, and the Decree
      5986 (together the “Decrees”) have introduced the following tax reductions
      (benefits) for businesses:tax reductions (benefits) for businesses: for
      the period from 1 April 2020 to 1 October 2020:for the period from 1 April
      2020 to 1 October 2020: 02/06/2020 COVID-19: Uzbekistan Government
      Financial Assistance Measures - Lexology
      https://www.lexology.com/library/detail.aspx?g=1d5e31b2-e7b1-44c9-8c9e-7d4bc5975bc2
      3/5 the minimum amount of social tax for individual entrepreneurs is
      reduced to the minimum amount of social tax for individual entrepreneurs
      is reduced to 50%50% of the base of the base calculated amount (“BCA”) per
      month;calculated amount (“BCA”) per month; the amount of mandatory
      payments for wholesalers of alcoholic beverages is reduced from the amount
      of mandatory payments for wholesalers of alcoholic beverages is reduced
      from 5 to5 to 3%3%; and; and fees for the right to retail sale of
      alcoholic products by catering enterprises are reduced byfees for the
      right to retail sale of alcoholic products by catering enterprises are
      reduced by 25% 25% of of the amounts set under law.the amounts set under
      law. These measures provide certain guarantees and protections, including
      deferred tax payments, decrease of taxThese measures provide certain
      guarantees and protections, including deferred tax payments, decrease of
      tax rates, tax related waivers and exemptions, as well as liquidity
      support measures.rates, tax related waivers and exemptions, as well as
      liquidity support measures.
  - text: >-
      The composition and nutritional content of the food ration for each
      beneficiary group are as follows: 19 While only the poorest families in
      the most food-insecure districts will receive general food distributions,
      in the poorest districts supplementary feeding will be targeted to all
      children 6-24 months, pregnant/lactating women and all moderately-
      malnourished children. 10767.0 Results-Chain (Logic Model) Performance
      Indicators Risks, Assumptions STRATEGIC OBJECTIVE 1 - Save Lives and
      Protect Livelihoods in Emergencies Outcome 1.1: Reduced acute malnutrition
      in children under 5 in targeted emergency-affected populations Outcome
      1.3: Improved food consumption over assistance period for targeted
      crisis-affected beneficiaries. (b) The food and nutrition situation 9.
metrics:
  - accuracy
pipeline_tag: text-classification
library_name: setfit
inference: false
base_model: sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

SetFit with sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2

This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-multilingual-MiniLM-L12-v2 as the Sentence Transformer embedding model. A OneVsRestClassifier 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

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("faodl/model_g20_multilabel_MiniLM-L12-v_final")
# Run inference
preds = model("4.3.3 Strategies for Comprehensive Sexuality Education and (CSE) Youth-friendly Health Services 1. To promote volunteerism as a tool for fostering active participation of young people in national development; 5. To promote volunteerism as a tool for fostering active participation of young people in national development; 5.")

Training Details

Training Set Metrics

Training set Min Median Max
Word count 2 70.3945 1194

Training Hyperparameters

  • batch_size: (32, 32)
  • num_epochs: (2, 2)
  • max_steps: -1
  • sampling_strategy: oversampling
  • num_iterations: 15
  • body_learning_rate: (2e-05, 2e-05)
  • head_learning_rate: 2e-05
  • loss: CosineSimilarityLoss
  • distance_metric: cosine_distance
  • margin: 0.25
  • end_to_end: False
  • use_amp: False
  • warmup_proportion: 0.1
  • l2_weight: 0.01
  • seed: 42
  • eval_max_steps: -1
  • load_best_model_at_end: False

Training Results

Epoch Step Training Loss Validation Loss
0.0003 1 0.1541 -
0.0166 50 0.1451 -
0.0331 100 0.1302 -
0.0497 150 0.1155 -
0.0663 200 0.098 -
0.0828 250 0.0984 -
0.0994 300 0.0872 -
0.1160 350 0.0813 -
0.1325 400 0.0778 -
0.1491 450 0.0732 -
0.1657 500 0.0694 -
0.1822 550 0.0585 -
0.1988 600 0.0621 -
0.2154 650 0.0597 -
0.2319 700 0.0511 -
0.2485 750 0.0461 -
0.2651 800 0.0463 -
0.2816 850 0.0377 -
0.2982 900 0.0345 -
0.3148 950 0.0371 -
0.3313 1000 0.0392 -
0.3479 1050 0.0378 -
0.3645 1100 0.0317 -
0.3810 1150 0.0302 -
0.3976 1200 0.0352 -
0.4142 1250 0.0298 -
0.4307 1300 0.0266 -
0.4473 1350 0.0287 -
0.4639 1400 0.0269 -
0.4805 1450 0.0227 -
0.4970 1500 0.0254 -
0.5136 1550 0.0228 -
0.5302 1600 0.0197 -
0.5467 1650 0.021 -
0.5633 1700 0.0217 -
0.5799 1750 0.0219 -
0.5964 1800 0.0189 -
0.6130 1850 0.0172 -
0.6296 1900 0.0199 -
0.6461 1950 0.017 -
0.6627 2000 0.017 -
0.6793 2050 0.0165 -
0.6958 2100 0.0162 -
0.7124 2150 0.0186 -
0.7290 2200 0.0171 -
0.7455 2250 0.0165 -
0.7621 2300 0.0186 -
0.7787 2350 0.0161 -
0.7952 2400 0.0146 -
0.8118 2450 0.0142 -
0.8284 2500 0.0142 -
0.8449 2550 0.0183 -
0.8615 2600 0.0162 -
0.8781 2650 0.0152 -
0.8946 2700 0.0119 -
0.9112 2750 0.0137 -
0.9278 2800 0.0161 -
0.9443 2850 0.0126 -
0.9609 2900 0.014 -
0.9775 2950 0.0169 -
0.9940 3000 0.0135 -
1.0106 3050 0.0149 -
1.0272 3100 0.0134 -
1.0437 3150 0.0096 -
1.0603 3200 0.0117 -
1.0769 3250 0.0075 -
1.0934 3300 0.0119 -
1.1100 3350 0.0098 -
1.1266 3400 0.011 -
1.1431 3450 0.0088 -
1.1597 3500 0.0125 -
1.1763 3550 0.0132 -
1.1928 3600 0.0103 -
1.2094 3650 0.0111 -
1.2260 3700 0.0091 -
1.2425 3750 0.0081 -
1.2591 3800 0.0108 -
1.2757 3850 0.0127 -
1.2922 3900 0.0115 -
1.3088 3950 0.011 -
1.3254 4000 0.0065 -
1.3419 4050 0.0113 -
1.3585 4100 0.0092 -
1.3751 4150 0.0085 -
1.3917 4200 0.0078 -
1.4082 4250 0.0071 -
1.4248 4300 0.0094 -
1.4414 4350 0.0085 -
1.4579 4400 0.0126 -
1.4745 4450 0.0084 -
1.4911 4500 0.0106 -
1.5076 4550 0.0093 -
1.5242 4600 0.0085 -
1.5408 4650 0.0072 -
1.5573 4700 0.0079 -
1.5739 4750 0.0111 -
1.5905 4800 0.0082 -
1.6070 4850 0.0075 -
1.6236 4900 0.0111 -
1.6402 4950 0.0098 -
1.6567 5000 0.01 -
1.6733 5050 0.0086 -
1.6899 5100 0.0116 -
1.7064 5150 0.0095 -
1.7230 5200 0.0106 -
1.7396 5250 0.0087 -
1.7561 5300 0.0079 -
1.7727 5350 0.0092 -
1.7893 5400 0.0085 -
1.8058 5450 0.0095 -
1.8224 5500 0.0066 -
1.8390 5550 0.0085 -
1.8555 5600 0.0103 -
1.8721 5650 0.0059 -
1.8887 5700 0.0064 -
1.9052 5750 0.0069 -
1.9218 5800 0.0086 -
1.9384 5850 0.0069 -
1.9549 5900 0.0089 -
1.9715 5950 0.0091 -
1.9881 6000 0.0083 -

Framework Versions

  • Python: 3.11.13
  • SetFit: 1.1.2
  • Sentence Transformers: 4.1.0
  • Transformers: 4.52.4
  • PyTorch: 2.6.0+cu124
  • Datasets: 3.6.0
  • Tokenizers: 0.21.2

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}
}