Add SetFit model

d59ca74 10 months ago

13 kB

	---
	library_name: setfit
	tags:
	- setfit
	- sentence-transformers
	- text-classification
	- generated_from_setfit_trainer
	metrics:
	- accuracy
	widget:
	- text: Guy Cecil, the former head of the Democratic Senatorial Campaign Committee
	and now the boss of a leading Democratic super PAC, voiced his frustration with
	the inadequacy of Franken’s apology on Twitter.
	- text: Attorney Stephen Le Brocq, who operates a law firm in the North Texas area
	sums up the treatment of Guyger perfectly when he says that “The affidavit isn’t
	written objectively, not at the slightest.
	- text: Phone This field is for validation purposes and should be left unchanged.
	- text: The Twitter suspension caught me by surprise.
	- text: Popular pages like The AntiMedia (2.1 million fans), The Free Thought Project
	(3.1 million fans), Press for Truth (350K fans), Police the Police (1.9 million
	fans), Cop Block (1.7 million fans), and Punk Rock Libertarians (125K fans) are
	just a few of the ones which were unpublished.
	pipeline_tag: text-classification
	inference: false
	base_model: sentence-transformers/paraphrase-mpnet-base-v2
	model-index:
	- name: SetFit with sentence-transformers/paraphrase-mpnet-base-v2
	results:
	- task:
	type: text-classification
	name: Text Classification
	dataset:
	name: Unknown
	type: unknown
	split: test
	metrics:
	- type: accuracy
	value: 0.7083881146463319
	name: Accuracy
	---

	# SetFit with sentence-transformers/paraphrase-mpnet-base-v2

	This is a [SetFit](https://github.com/huggingface/setfit) model that can be used for Text Classification. This SetFit model uses [sentence-transformers/paraphrase-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-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:

	1. Fine-tuning a [Sentence Transformer](https://www.sbert.net) with contrastive learning.
	2. 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-mpnet-base-v2](https://huggingface.co/sentence-transformers/paraphrase-mpnet-base-v2)
	- Classification head: a OneVsRestClassifier instance
	- Maximum Sequence Length: 512 tokens
	<!-- - Number of Classes: Unknown -->
	<!-- - Training Dataset: [Unknown](https://huggingface.co/datasets/unknown) -->
	<!-- - Language: Unknown -->
	<!-- - License: Unknown -->

	### Model Sources

	- Repository: [SetFit on GitHub](https://github.com/huggingface/setfit)
	- Paper: [Efficient Few-Shot Learning Without Prompts](https://arxiv.org/abs/2209.11055)
	- Blogpost: [SetFit: Efficient Few-Shot Learning Without Prompts](https://huggingface.co/blog/setfit)

	## Evaluation

	### Metrics
	\| Label \| Accuracy \|
	\|:--------\|:---------\|
	\| all \| 0.7084 \|

	## Uses

	### Direct Use for Inference

	First install the SetFit library:

	```bash
	pip install setfit
	```

	Then you can load this model and run inference.

	```python
	from setfit import SetFitModel

	# Download from the 🤗 Hub
	model = SetFitModel.from_pretrained("anismahmahi/doubt_repetition_with_noPropaganda_with_3_zeros_SetFit")
	# Run inference
	preds = model("The Twitter suspension caught me by surprise.")
	```

	<!--
	### Downstream Use

	List how someone could finetune this model on their own dataset.
	-->

	<!--
	### Out-of-Scope Use

	List how the model may foreseeably be misused and address what users ought not to do with the model.
	-->

	<!--
	## Bias, Risks and Limitations

	What are the known or foreseeable issues stemming from this model? You could also flag here known failure cases or weaknesses of the model.
	-->

	<!--
	### Recommendations

	What are recommendations with respect to the foreseeable issues? For example, filtering explicit content.
	-->

	## Training Details

	### Training Set Metrics
	\| Training set \| Min \| Median \| Max \|
	\|:-------------\|:----\|:--------\|:----\|
	\| Word count \| 1 \| 22.0291 \| 129 \|

	### Training Hyperparameters
	- batch_size: (16, 16)
	- num_epochs: (2, 2)
	- max_steps: -1
	- sampling_strategy: oversampling
	- num_iterations: 5
	- 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: True

	### Training Results
	\| Epoch \| Step \| Training Loss \| Validation Loss \|
	\|:-------:\|:--------:\|:-------------:\|:---------------:\|
	\| 0.0003 \| 1 \| 0.3532 \| - \|
	\| 0.0166 \| 50 \| 0.3413 \| - \|
	\| 0.0332 \| 100 \| 0.2743 \| - \|
	\| 0.0498 \| 150 \| 0.2635 \| - \|
	\| 0.0664 \| 200 \| 0.2444 \| - \|
	\| 0.0830 \| 250 \| 0.1883 \| - \|
	\| 0.0996 \| 300 \| 0.2231 \| - \|
	\| 0.1162 \| 350 \| 0.1763 \| - \|
	\| 0.1328 \| 400 \| 0.1868 \| - \|
	\| 0.1494 \| 450 \| 0.2057 \| - \|
	\| 0.1660 \| 500 \| 0.1734 \| - \|
	\| 0.1826 \| 550 \| 0.2594 \| - \|
	\| 0.1992 \| 600 \| 0.1024 \| - \|
	\| 0.2158 \| 650 \| 0.2351 \| - \|
	\| 0.2324 \| 700 \| 0.1863 \| - \|
	\| 0.2490 \| 750 \| 0.072 \| - \|
	\| 0.2656 \| 800 \| 0.1987 \| - \|
	\| 0.2822 \| 850 \| 0.1511 \| - \|
	\| 0.2988 \| 900 \| 0.0926 \| - \|
	\| 0.3154 \| 950 \| 0.1956 \| - \|
	\| 0.3320 \| 1000 \| 0.1354 \| - \|
	\| 0.3486 \| 1050 \| 0.2038 \| - \|
	\| 0.3652 \| 1100 \| 0.1166 \| - \|
	\| 0.3818 \| 1150 \| 0.3214 \| - \|
	\| 0.3984 \| 1200 \| 0.0703 \| - \|
	\| 0.4150 \| 1250 \| 0.1815 \| - \|
	\| 0.4316 \| 1300 \| 0.124 \| - \|
	\| 0.4482 \| 1350 \| 0.0955 \| - \|
	\| 0.4648 \| 1400 \| 0.1064 \| - \|
	\| 0.4814 \| 1450 \| 0.0429 \| - \|
	\| 0.4980 \| 1500 \| 0.0814 \| - \|
	\| 0.5146 \| 1550 \| 0.1483 \| - \|
	\| 0.5312 \| 1600 \| 0.0856 \| - \|
	\| 0.5478 \| 1650 \| 0.1072 \| - \|
	\| 0.5644 \| 1700 \| 0.0148 \| - \|
	\| 0.5810 \| 1750 \| 0.0571 \| - \|
	\| 0.5976 \| 1800 \| 0.052 \| - \|
	\| 0.6142 \| 1850 \| 0.0532 \| - \|
	\| 0.6308 \| 1900 \| 0.0088 \| - \|
	\| 0.6474 \| 1950 \| 0.1619 \| - \|
	\| 0.6640 \| 2000 \| 0.0618 \| - \|
	\| 0.6806 \| 2050 \| 0.0115 \| - \|
	\| 0.6972 \| 2100 \| 0.1402 \| - \|
	\| 0.7138 \| 2150 \| 0.0637 \| - \|
	\| 0.7304 \| 2200 \| 0.0194 \| - \|
	\| 0.7470 \| 2250 \| 0.0135 \| - \|
	\| 0.7636 \| 2300 \| 0.0109 \| - \|
	\| 0.7802 \| 2350 \| 0.133 \| - \|
	\| 0.7968 \| 2400 \| 0.0565 \| - \|
	\| 0.8134 \| 2450 \| 0.1508 \| - \|
	\| 0.8300 \| 2500 \| 0.0293 \| - \|
	\| 0.8466 \| 2550 \| 0.065 \| - \|
	\| 0.8632 \| 2600 \| 0.0029 \| - \|
	\| 0.8798 \| 2650 \| 0.008 \| - \|
	\| 0.8964 \| 2700 \| 0.0604 \| - \|
	\| 0.9130 \| 2750 \| 0.0074 \| - \|
	\| 0.9296 \| 2800 \| 0.0019 \| - \|
	\| 0.9462 \| 2850 \| 0.0129 \| - \|
	\| 0.9628 \| 2900 \| 0.0838 \| - \|
	\| 0.9794 \| 2950 \| 0.0044 \| - \|
	\| 0.9960 \| 3000 \| 0.0035 \| - \|
	\| 1.0 \| 3012 \| - \| 0.2514 \|
	\| 1.0126 \| 3050 \| 0.0086 \| - \|
	\| 1.0292 \| 3100 \| 0.0042 \| - \|
	\| 1.0458 \| 3150 \| 0.0833 \| - \|
	\| 1.0624 \| 3200 \| 0.058 \| - \|
	\| 1.0790 \| 3250 \| 0.013 \| - \|
	\| 1.0956 \| 3300 \| 0.0429 \| - \|
	\| 1.1122 \| 3350 \| 0.0044 \| - \|
	\| 1.1288 \| 3400 \| 0.0699 \| - \|
	\| 1.1454 \| 3450 \| 0.0535 \| - \|
	\| 1.1620 \| 3500 \| 0.0559 \| - \|
	\| 1.1786 \| 3550 \| 0.1459 \| - \|
	\| 1.1952 \| 3600 \| 0.118 \| - \|
	\| 1.2118 \| 3650 \| 0.14 \| - \|
	\| 1.2284 \| 3700 \| 0.0632 \| - \|
	\| 1.2450 \| 3750 \| 0.0026 \| - \|
	\| 1.2616 \| 3800 \| 0.0026 \| - \|
	\| 1.2782 \| 3850 \| 0.0052 \| - \|
	\| 1.2948 \| 3900 \| 0.0058 \| - \|
	\| 1.3114 \| 3950 \| 0.0018 \| - \|
	\| 1.3280 \| 4000 \| 0.0152 \| - \|
	\| 1.3446 \| 4050 \| 0.0186 \| - \|
	\| 1.3612 \| 4100 \| 0.039 \| - \|
	\| 1.3778 \| 4150 \| 0.0022 \| - \|
	\| 1.3944 \| 4200 \| 0.002 \| - \|
	\| 1.4110 \| 4250 \| 0.0032 \| - \|
	\| 1.4276 \| 4300 \| 0.0285 \| - \|
	\| 1.4442 \| 4350 \| 0.0213 \| - \|
	\| 1.4608 \| 4400 \| 0.0009 \| - \|
	\| 1.4774 \| 4450 \| 0.0262 \| - \|
	\| 1.4940 \| 4500 \| 0.0181 \| - \|
	\| 1.5106 \| 4550 \| 0.0629 \| - \|
	\| 1.5272 \| 4600 \| 0.0023 \| - \|
	\| 1.5438 \| 4650 \| 0.003 \| - \|
	\| 1.5604 \| 4700 \| 0.0024 \| - \|
	\| 1.5770 \| 4750 \| 0.049 \| - \|
	\| 1.5936 \| 4800 \| 0.0154 \| - \|
	\| 1.6102 \| 4850 \| 0.0009 \| - \|
	\| 1.6268 \| 4900 \| 0.0015 \| - \|
	\| 1.6434 \| 4950 \| 0.0068 \| - \|
	\| 1.6600 \| 5000 \| 0.057 \| - \|
	\| 1.6766 \| 5050 \| 0.0031 \| - \|
	\| 1.6932 \| 5100 \| 0.0189 \| - \|
	\| 1.7098 \| 5150 \| 0.0317 \| - \|
	\| 1.7264 \| 5200 \| 0.0013 \| - \|
	\| 1.7430 \| 5250 \| 0.0247 \| - \|
	\| 1.7596 \| 5300 \| 0.0062 \| - \|
	\| 1.7762 \| 5350 \| 0.0192 \| - \|
	\| 1.7928 \| 5400 \| 0.0019 \| - \|
	\| 1.8094 \| 5450 \| 0.1007 \| - \|
	\| 1.8260 \| 5500 \| 0.0384 \| - \|
	\| 1.8426 \| 5550 \| 0.0494 \| - \|
	\| 1.8592 \| 5600 \| 0.0615 \| - \|
	\| 1.8758 \| 5650 \| 0.0709 \| - \|
	\| 1.8924 \| 5700 \| 0.0308 \| - \|
	\| 1.9090 \| 5750 \| 0.0107 \| - \|
	\| 1.9256 \| 5800 \| 0.064 \| - \|
	\| 1.9422 \| 5850 \| 0.0009 \| - \|
	\| 1.9588 \| 5900 \| 0.0019 \| - \|
	\| 1.9754 \| 5950 \| 0.0037 \| - \|
	\| 1.9920 \| 6000 \| 0.0826 \| - \|
	\| 2.0 \| 6024 \| - \| 0.2614 \|

	* The bold row denotes the saved checkpoint.
	### 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: 2.16.1
	- Tokenizers: 0.15.0

	## Citation

	### BibTeX
	```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}
	}
	```

	<!--
	## Glossary

	Clearly define terms in order to be accessible across audiences.
	-->

	<!--
	## Model Card Authors

	Lists the people who create the model card, providing recognition and accountability for the detailed work that goes into its construction.
	-->

	<!--
	## Model Card Contact

	Provides a way for people who have updates to the Model Card, suggestions, or questions, to contact the Model Card authors.
	-->