metadata
library_name: setfit
tags:
- setfit
- sentence-transformers
- text-classification
- generated_from_setfit_trainer
metrics:
- Precision_micro
- Precision_weighted
- Precision_samples
- Recall_micro
- Recall_weighted
- Recall_samples
- F1-Score
- accuracy
widget:
- text: >-
To support the traditional knowledge and adaptive capacity of indigenous
peoples in the face of climate change, we aim to establish 50
community-based adaptation projects led by indigenous peoples by 2030,
focusing on the sustainable management of natural resources and the
preservation of cultural practices.
- text: >-
Measures related to climate change are incorporated into national
policies, strategies and plans. In this regard, mechanisms are also
promoted to increase capacity for effective planning and management in
relation to climate change. SDG No. 14 (Marine life). Adaptation. There is
a link between the Coastal Marine Resources sector in the measures
proposed in this document and the indicators of this SDG regarding the
sustainable management and conservation of marine and coastal ecosystems
to achieve an increase in their climate resilience. SDG No.
- text: ' Pathways with higher demand for food, feed, and water, more resource-intensive consumption and production, and more limited technological improvements in agriculture yields result in higher risks from water scarcity in drylands, land degradation, and food insecurity 1. This means that communities that rely on agriculture for their livelihoods are at risk of losing their crops and experiencing food shortages due to climate change.'
- text: >-
The population aged 60 years and above is projected to increase from
almost one million (988,000) in 2000 to over six million (6,319,000) by
2050. The female aged population will continue to grow faster and will
increasingly be far higher than the male population for the advanced ages.
Policies addressing the needs of the elderly will have to take the sex
structure of the aged population into consideration.
- text: >-
Indigenous peoples who choose or are forced to migrate away from their
traditional lands often face double discrimination as both migrants and as
indigenous peoples. Indigenous peoples may be more vulnerable to irregular
migration such as trafficking and smuggling, owing to sudden displacement
by a climactic event, limited legal migration options and limited
opportunities to make informed choices. Deforestation, particularly in
developing countries, is pushing indigenous families to migrate to cities
for economic reasons, often ending up in urban slums.
pipeline_tag: text-classification
inference: false
base_model: sentence-transformers/all-mpnet-base-v2
model-index:
- name: SetFit with sentence-transformers/all-mpnet-base-v2
results:
- task:
type: text-classification
name: Text Classification
dataset:
name: Unknown
type: unknown
split: test
metrics:
- type: Precision_micro
value: 0.7762237762237763
name: Precision_Micro
- type: Precision_weighted
value: 0.7968800430338892
name: Precision_Weighted
- type: Precision_samples
value: 0.7762237762237763
name: Precision_Samples
- type: Recall_micro
value: 0.7762237762237763
name: Recall_Micro
- type: Recall_weighted
value: 0.7762237762237763
name: Recall_Weighted
- type: Recall_samples
value: 0.7762237762237763
name: Recall_Samples
- type: F1-Score
value: 0.7762237762237763
name: F1-Score
- type: accuracy
value: 0.7762237762237763
name: Accuracy
SetFit with sentence-transformers/all-mpnet-base-v2
This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/all-mpnet-base-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:
- 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: sentence-transformers/all-mpnet-base-v2
- Classification head: a OneVsRestClassifier instance
- Maximum Sequence Length: 384 tokens
Model Sources
- Repository: SetFit on GitHub
- Paper: Efficient Few-Shot Learning Without Prompts
- Blogpost: SetFit: Efficient Few-Shot Learning Without Prompts
Evaluation
Metrics
| Label | Precision_Micro | Precision_Weighted | Precision_Samples | Recall_Micro | Recall_Weighted | Recall_Samples | F1-Score | Accuracy |
|---|---|---|---|---|---|---|---|---|
| all | 0.7762 | 0.7969 | 0.7762 | 0.7762 | 0.7762 | 0.7762 | 0.7762 | 0.7762 |
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("leavoigt/vulnerability_target")
# Run inference
preds = model("To support the traditional knowledge and adaptive capacity of indigenous peoples in the face of climate change, we aim to establish 50 community-based adaptation projects led by indigenous peoples by 2030, focusing on the sustainable management of natural resources and the preservation of cultural practices.")
Training Details
Training Set Metrics
| Training set | Min | Median | Max |
|---|---|---|---|
| Word count | 15 | 70.8675 | 238 |
Training Hyperparameters
- batch_size: (16, 16)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 20
- 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
- seed: 42
- eval_max_steps: -1
- load_best_model_at_end: False
Training Results
| Epoch | Step | Training Loss | Validation Loss |
|---|---|---|---|
| 0.0012 | 1 | 0.3493 | - |
| 0.0602 | 50 | 0.2285 | - |
| 0.1205 | 100 | 0.1092 | - |
| 0.1807 | 150 | 0.1348 | - |
| 0.2410 | 200 | 0.0365 | - |
| 0.3012 | 250 | 0.0052 | - |
| 0.3614 | 300 | 0.0012 | - |
| 0.4217 | 350 | 0.0031 | - |
| 0.4819 | 400 | 0.0001 | - |
| 0.5422 | 450 | 0.0011 | - |
| 0.6024 | 500 | 0.0001 | - |
| 0.6627 | 550 | 0.0001 | - |
| 0.7229 | 600 | 0.0001 | - |
| 0.7831 | 650 | 0.0002 | - |
| 0.8434 | 700 | 0.0001 | - |
| 0.9036 | 750 | 0.0001 | - |
| 0.9639 | 800 | 0.0001 | - |
Framework Versions
- Python: 3.10.12
- SetFit: 1.0.1
- Sentence Transformers: 2.2.2
- Transformers: 4.25.1
- PyTorch: 2.1.0+cu121
- Datasets: 2.16.1
- Tokenizers: 0.13.3
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}
}