SetFit with sentence-transformers/paraphrase-multilingual-mpnet-base-v2
This is a SetFit model that can be used for Text Classification. This SetFit model uses sentence-transformers/paraphrase-multilingual-mpnet-base-v2 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: sentence-transformers/paraphrase-multilingual-mpnet-base-v2
- Classification head: a LogisticRegression instance
- Maximum Sequence Length: 128 tokens
- Number of Classes: 3 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 |
---|---|
pos |
|
neg |
|
obj |
|
Evaluation
Metrics
Label | Accuracy_Score | Classification_Report |
---|---|---|
all | 0.9435 | {'0': {'precision': 0.9361702127659575, 'recall': 0.9322033898305084, 'f1-score': 0.9341825902335456, 'support': 236}, '1': {'precision': 0.9333333333333333, 'recall': 0.9302325581395349, 'f1-score': 0.9317803660565723, 'support': 301}, '2': {'precision': 0.9646017699115044, 'recall': 0.9732142857142857, 'f1-score': 0.9688888888888889, 'support': 224}, 'accuracy': 0.9434954007884363, 'macro avg': {'precision': 0.9447017720035985, 'recall': 0.945216744561443, 'f1-score': 0.9449506150596689, 'support': 761}, 'weighted avg': {'precision': 0.9434169513880108, 'recall': 0.9434954007884363, 'f1-score': 0.9434482162802315, 'support': 761}} |
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("mogaio/pr_ebsa_fr_tran_merged25_e1_end_offsets")
# Run inference
preds = model("Adil Hussain
Adil Hussain est reconnaissant d'avoir reçu l'enseignement de l'acteur Naseeruddin Shah à l'époque où il fréquentait l'École nationale d'art dramatique")
Training Details
Training Set Metrics
Training set | Min | Median | Max |
---|---|---|---|
Word count | 9 | 247.2638 | 2089 |
Label | Training Sample Count |
---|---|
neg | 913 |
obj | 1216 |
pos | 911 |
Training Hyperparameters
- batch_size: (8, 8)
- num_epochs: (1, 1)
- max_steps: -1
- sampling_strategy: oversampling
- num_iterations: 1
- 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 |
---|---|---|---|
0.0013 | 1 | 0.3703 | - |
0.0658 | 50 | 0.3145 | - |
0.1316 | 100 | 0.1839 | - |
0.1974 | 150 | 0.2558 | - |
0.2632 | 200 | 0.2683 | - |
0.3289 | 250 | 0.1572 | - |
0.3947 | 300 | 0.1953 | - |
0.4605 | 350 | 0.171 | - |
0.5263 | 400 | 0.2326 | - |
0.5921 | 450 | 0.1762 | - |
0.6579 | 500 | 0.2818 | - |
0.7237 | 550 | 0.2733 | - |
0.7895 | 600 | 0.195 | - |
0.8553 | 650 | 0.2104 | - |
0.9211 | 700 | 0.2124 | - |
0.9868 | 750 | 0.0818 | - |
1.0526 | 800 | 0.1046 | - |
1.1184 | 850 | 0.1633 | - |
1.1842 | 900 | 0.3207 | - |
1.25 | 950 | 0.2703 | - |
1.3158 | 1000 | 0.1934 | - |
1.3816 | 1050 | 0.2547 | - |
1.4474 | 1100 | 0.0933 | - |
1.5132 | 1150 | 0.2102 | - |
1.5789 | 1200 | 0.0699 | - |
1.6447 | 1250 | 0.1778 | - |
1.7105 | 1300 | 0.1796 | - |
1.7763 | 1350 | 0.0221 | - |
1.8421 | 1400 | 0.2154 | - |
1.9079 | 1450 | 0.1683 | - |
1.9737 | 1500 | 0.3096 | - |
2.0395 | 1550 | 0.201 | - |
2.1053 | 1600 | 0.1954 | - |
2.1711 | 1650 | 0.2301 | - |
2.2368 | 1700 | 0.1141 | - |
2.3026 | 1750 | 0.1949 | - |
2.3684 | 1800 | 0.164 | - |
2.4342 | 1850 | 0.2307 | - |
2.5 | 1900 | 0.1912 | - |
2.5658 | 1950 | 0.2349 | - |
2.6316 | 2000 | 0.0922 | - |
2.6974 | 2050 | 0.0702 | - |
2.7632 | 2100 | 0.1089 | - |
2.8289 | 2150 | 0.1711 | - |
2.8947 | 2200 | 0.1432 | - |
2.9605 | 2250 | 0.2739 | - |
3.0263 | 2300 | 0.1889 | - |
3.0921 | 2350 | 0.1036 | - |
3.1579 | 2400 | 0.1372 | - |
3.2237 | 2450 | 0.028 | - |
3.2895 | 2500 | 0.1739 | - |
3.3553 | 2550 | 0.142 | - |
3.4211 | 2600 | 0.0838 | - |
3.4868 | 2650 | 0.0657 | - |
3.5526 | 2700 | 0.0054 | - |
3.6184 | 2750 | 0.0426 | - |
3.6842 | 2800 | 0.1974 | - |
3.75 | 2850 | 0.0279 | - |
3.8158 | 2900 | 0.1326 | - |
3.8816 | 2950 | 0.1614 | - |
3.9474 | 3000 | 0.1251 | - |
4.0132 | 3050 | 0.1174 | - |
4.0789 | 3100 | 0.1948 | - |
4.1447 | 3150 | 0.0555 | - |
4.2105 | 3200 | 0.0064 | - |
4.2763 | 3250 | 0.064 | - |
4.3421 | 3300 | 0.0013 | - |
4.4079 | 3350 | 0.135 | - |
4.4737 | 3400 | 0.0574 | - |
4.5395 | 3450 | 0.174 | - |
4.6053 | 3500 | 0.2199 | - |
4.6711 | 3550 | 0.387 | - |
4.7368 | 3600 | 0.114 | - |
4.8026 | 3650 | 0.0853 | - |
4.8684 | 3700 | 0.0325 | - |
4.9342 | 3750 | 0.019 | - |
5.0 | 3800 | 0.0572 | - |
0.0013 | 1 | 0.1435 | - |
0.0658 | 50 | 0.0969 | - |
0.1316 | 100 | 0.1085 | - |
0.1974 | 150 | 0.0271 | - |
0.2632 | 200 | 0.0138 | - |
0.3289 | 250 | 0.058 | - |
0.3947 | 300 | 0.1205 | - |
0.4605 | 350 | 0.0788 | - |
0.5263 | 400 | 0.1449 | - |
0.5921 | 450 | 0.0383 | - |
0.6579 | 500 | 0.0338 | - |
0.7237 | 550 | 0.1253 | - |
0.7895 | 600 | 0.069 | - |
0.8553 | 650 | 0.104 | - |
0.9211 | 700 | 0.0462 | - |
0.9868 | 750 | 0.1975 | - |
1.0526 | 800 | 0.0241 | - |
1.1184 | 850 | 0.0426 | - |
1.1842 | 900 | 0.0519 | - |
1.25 | 950 | 0.0815 | - |
1.3158 | 1000 | 0.1839 | - |
1.3816 | 1050 | 0.0198 | - |
1.4474 | 1100 | 0.0128 | - |
1.5132 | 1150 | 0.1645 | - |
1.5789 | 1200 | 0.0019 | - |
1.6447 | 1250 | 0.0557 | - |
1.7105 | 1300 | 0.0098 | - |
1.7763 | 1350 | 0.001 | - |
1.8421 | 1400 | 0.1557 | - |
1.9079 | 1450 | 0.1286 | - |
1.9737 | 1500 | 0.094 | - |
2.0395 | 1550 | 0.0059 | - |
2.1053 | 1600 | 0.0227 | - |
2.1711 | 1650 | 0.0899 | - |
2.2368 | 1700 | 0.0053 | - |
2.3026 | 1750 | 0.0021 | - |
2.3684 | 1800 | 0.0114 | - |
2.4342 | 1850 | 0.1163 | - |
2.5 | 1900 | 0.0959 | - |
2.5658 | 1950 | 0.0252 | - |
2.6316 | 2000 | 0.0921 | - |
2.6974 | 2050 | 0.1159 | - |
2.7632 | 2100 | 0.0026 | - |
2.8289 | 2150 | 0.1211 | - |
2.8947 | 2200 | 0.1843 | - |
2.9605 | 2250 | 0.0014 | - |
3.0263 | 2300 | 0.0085 | - |
3.0921 | 2350 | 0.0839 | - |
3.1579 | 2400 | 0.2372 | - |
3.2237 | 2450 | 0.0213 | - |
3.2895 | 2500 | 0.0155 | - |
3.3553 | 2550 | 0.1128 | - |
3.4211 | 2600 | 0.0945 | - |
3.4868 | 2650 | 0.0917 | - |
3.5526 | 2700 | 0.0011 | - |
3.6184 | 2750 | 0.0024 | - |
3.6842 | 2800 | 0.0044 | - |
3.75 | 2850 | 0.121 | - |
3.8158 | 2900 | 0.0056 | - |
3.8816 | 2950 | 0.003 | - |
3.9474 | 3000 | 0.0899 | - |
4.0132 | 3050 | 0.0157 | - |
4.0789 | 3100 | 0.1188 | - |
4.1447 | 3150 | 0.001 | - |
4.2105 | 3200 | 0.0222 | - |
4.2763 | 3250 | 0.1209 | - |
4.3421 | 3300 | 0.1085 | - |
4.4079 | 3350 | 0.0054 | - |
4.4737 | 3400 | 0.0009 | - |
4.5395 | 3450 | 0.0015 | - |
4.6053 | 3500 | 0.003 | - |
4.6711 | 3550 | 0.0009 | - |
4.7368 | 3600 | 0.0003 | - |
4.8026 | 3650 | 0.0009 | - |
4.8684 | 3700 | 0.03 | - |
4.9342 | 3750 | 0.1206 | - |
5.0 | 3800 | 0.0003 | - |
0.0013 | 1 | 0.2045 | - |
0.0658 | 50 | 0.0078 | - |
0.1316 | 100 | 0.0087 | - |
0.1974 | 150 | 0.0386 | - |
0.2632 | 200 | 0.1015 | - |
0.3289 | 250 | 0.0022 | - |
0.3947 | 300 | 0.0291 | - |
0.4605 | 350 | 0.0013 | - |
0.5263 | 400 | 0.0022 | - |
0.5921 | 450 | 0.1324 | - |
0.6579 | 500 | 0.113 | - |
0.7237 | 550 | 0.0011 | - |
0.7895 | 600 | 0.1723 | - |
0.8553 | 650 | 0.0049 | - |
0.9211 | 700 | 0.206 | - |
0.9868 | 750 | 0.1683 | - |
1.0526 | 800 | 0.0954 | - |
1.1184 | 850 | 0.018 | - |
1.1842 | 900 | 0.1854 | - |
1.25 | 950 | 0.0342 | - |
1.3158 | 1000 | 0.0015 | - |
1.3816 | 1050 | 0.0062 | - |
1.4474 | 1100 | 0.1187 | - |
1.5132 | 1150 | 0.0048 | - |
1.5789 | 1200 | 0.0011 | - |
1.6447 | 1250 | 0.002 | - |
1.7105 | 1300 | 0.092 | - |
1.7763 | 1350 | 0.1245 | - |
1.8421 | 1400 | 0.0009 | - |
1.9079 | 1450 | 0.1185 | - |
1.9737 | 1500 | 0.0017 | - |
2.0395 | 1550 | 0.008 | - |
2.1053 | 1600 | 0.0049 | - |
2.1711 | 1650 | 0.0083 | - |
2.2368 | 1700 | 0.0026 | - |
2.3026 | 1750 | 0.0081 | - |
2.3684 | 1800 | 0.0036 | - |
2.4342 | 1850 | 0.0016 | - |
2.5 | 1900 | 0.0017 | - |
2.5658 | 1950 | 0.0014 | - |
2.6316 | 2000 | 0.0017 | - |
2.6974 | 2050 | 0.002 | - |
2.7632 | 2100 | 0.1022 | - |
2.8289 | 2150 | 0.0004 | - |
2.8947 | 2200 | 0.0007 | - |
2.9605 | 2250 | 0.0794 | - |
3.0263 | 2300 | 0.0183 | - |
3.0921 | 2350 | 0.0377 | - |
3.1579 | 2400 | 0.029 | - |
3.2237 | 2450 | 0.0003 | - |
3.2895 | 2500 | 0.0961 | - |
3.3553 | 2550 | 0.0008 | - |
3.4211 | 2600 | 0.0873 | - |
3.4868 | 2650 | 0.0501 | - |
3.5526 | 2700 | 0.0029 | - |
3.6184 | 2750 | 0.0008 | - |
3.6842 | 2800 | 0.0004 | - |
3.75 | 2850 | 0.0011 | - |
3.8158 | 2900 | 0.0518 | - |
3.8816 | 2950 | 0.0002 | - |
3.9474 | 3000 | 0.1115 | - |
4.0132 | 3050 | 0.0129 | - |
4.0789 | 3100 | 0.0005 | - |
4.1447 | 3150 | 0.0012 | - |
4.2105 | 3200 | 0.1086 | - |
4.2763 | 3250 | 0.0199 | - |
4.3421 | 3300 | 0.0004 | - |
4.4079 | 3350 | 0.0001 | - |
4.4737 | 3400 | 0.0832 | - |
4.5395 | 3450 | 0.0003 | - |
4.6053 | 3500 | 0.0041 | - |
4.6711 | 3550 | 0.1146 | - |
4.7368 | 3600 | 0.0027 | - |
4.8026 | 3650 | 0.0002 | - |
4.8684 | 3700 | 0.0544 | - |
4.9342 | 3750 | 0.0002 | - |
5.0 | 3800 | 0.0046 | - |
0.0013 | 1 | 0.0015 | - |
0.0658 | 50 | 0.1973 | - |
0.1316 | 100 | 0.0106 | - |
0.1974 | 150 | 0.0744 | - |
0.2632 | 200 | 0.1033 | - |
0.3289 | 250 | 0.0425 | - |
0.3947 | 300 | 0.1125 | - |
0.4605 | 350 | 0.0018 | - |
0.5263 | 400 | 0.0019 | - |
0.5921 | 450 | 0.0002 | - |
0.6579 | 500 | 0.0007 | - |
0.7237 | 550 | 0.1393 | - |
0.7895 | 600 | 0.0002 | - |
0.8553 | 650 | 0.0043 | - |
0.9211 | 700 | 0.0339 | - |
0.9868 | 750 | 0.0002 | - |
0.0013 | 1 | 0.0007 | - |
0.0658 | 50 | 0.0419 | - |
0.1316 | 100 | 0.0068 | - |
0.1974 | 150 | 0.1401 | - |
0.2632 | 200 | 0.0423 | - |
0.3289 | 250 | 0.1122 | - |
0.3947 | 300 | 0.0037 | - |
0.4605 | 350 | 0.005 | - |
0.5263 | 400 | 0.0006 | - |
0.5921 | 450 | 0.0006 | - |
0.6579 | 500 | 0.0016 | - |
0.7237 | 550 | 0.1244 | - |
0.7895 | 600 | 0.0016 | - |
0.8553 | 650 | 0.0028 | - |
0.9211 | 700 | 0.002 | - |
0.9868 | 750 | 0.057 | - |
0.0013 | 1 | 0.1396 | - |
0.0658 | 50 | 0.0366 | - |
0.1316 | 100 | 0.0021 | - |
0.1974 | 150 | 0.1088 | - |
0.2632 | 200 | 0.0449 | - |
0.3289 | 250 | 0.0187 | - |
0.3947 | 300 | 0.0017 | - |
0.4605 | 350 | 0.1262 | - |
0.5263 | 400 | 0.0052 | - |
0.5921 | 450 | 0.1188 | - |
0.6579 | 500 | 0.0002 | - |
0.7237 | 550 | 0.0006 | - |
0.7895 | 600 | 0.0758 | - |
0.8553 | 650 | 0.025 | - |
0.9211 | 700 | 0.0052 | - |
0.9868 | 750 | 0.1985 | - |
Framework Versions
- Python: 3.10.12
- SetFit: 1.0.1
- Sentence Transformers: 2.2.2
- Transformers: 4.35.2
- PyTorch: 2.1.0+cu121
- Datasets: 2.15.0
- 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
- 4
Inference Providers
NEW
This model is not currently available via any of the supported third-party Inference Providers, and
the model is not deployed on the HF Inference API.
Model tree for mogaio/pr_ebsa_fr_tran_merged25_e1_end_offsets
Evaluation results
- Accuracy_Score on Unknowntest set self-reported0.943
- Classification_Report on Unknowntest set self-reported[object Object]