Netta1994/setfit_baai_squad_gpt-4o_improved-cot_chat_few_shot_remove_final_evaluation_e1_larger

SetFit with BAAI/bge-base-en-v1.5

This is a SetFit model that can be used for Text Classification. This SetFit model uses BAAI/bge-base-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:

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 Type: SetFit
• Sentence Transformer body: BAAI/bge-base-en-v1.5
• Classification head: a LogisticRegression instance
• Maximum Sequence Length: 512 tokens
• Number of Classes: 2 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
0	'Reasoning:\nirrelevant - The answer is not relevant to what is being asked.\nEvaluation:' 'Reasoning:\nWhile the provided answer accurately identifies multiple services offered by Kartz Media & PR, it introduces a fictional service: "personalized space travel public relations for interstellar companies," which is not mentioned in the document. The remaining services listed in the answer align well with the services described in the provided document.\n\nEvaluation:' 'Reasoning:\nirrelevant - The answer provided does not address the question, and the informationis not relevant to what is asked.\nEvaluation:'
1	'Reasoning:\nThe answer is accurate and directly taken from the relevant part of the document. It correctly identifies Open Data and standard formats like XML, JSON,or CSV as the proposed solution.\nEvaluation:' 'Reasoning:\nclearly correct - The answer completely and accurately addresses the question. The steps for using plastic wrap and straws are detailed and align with the instructions provided in the document.\n\nEvaluation:' 'Reasoning:\ngood - The answer accurately corresponds to the details provided in the document about the value of a dime.\nEvaluation:'

Evaluation

Metrics

Label	Accuracy
all	0.8644

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("Netta1994/setfit_baai_squad_gpt-4o_improved-cot_chat_few_shot_remove_final_evaluation_e1_larger")
# Run inference
preds = model("Reasoning:
grounded - The answer correctly identifies the ranking position mentioned in the document.
Evaluation:")

Training Details

Training Set Metrics

Training set	Min	Median	Max
Word count	6	33.1685	156

Label	Training Sample Count
0	82
1	102

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
• l2_weight: 0.01
• seed: 42
• eval_max_steps: -1
• load_best_model_at_end: False

Training Results

Epoch	Step	Training Loss	Validation Loss
0.0022	1	0.2065	-
0.1087	50	0.2382	-
0.2174	100	0.1573	-
0.3261	150	0.0988	-
0.4348	200	0.029	-
0.5435	250	0.012	-
0.6522	300	0.0105	-
0.7609	350	0.0136	-
0.8696	400	0.0178	-
0.9783	450	0.0091	-

Framework Versions

• Python: 3.10.14
• SetFit: 1.1.0
• Sentence Transformers: 3.1.1
• Transformers: 4.44.0
• PyTorch: 2.4.0+cu121
• Datasets: 3.0.0
• Tokenizers: 0.19.1

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}
}