đ S-PubMedBert-MS-MARCO-SCIFACT
This model is a sentence-transformers model that maps sentences and paragraphs to a 768-dimensional dense vector space. It can be applied to tasks such as clustering or semantic search.
đ Quick Start
⨠Features
- Maps sentences and paragraphs to a 768-dimensional dense vector space.
- Applicable for clustering and semantic search tasks.
đĻ Installation
Using this model becomes easy when you have sentence-transformers installed:
pip install -U sentence-transformers
đģ Usage Examples
Basic Usage
If you have sentence-transformers installed, you can use the model like this:
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]
model = SentenceTransformer('S-PubMedBert-MS-MARCO-SCIFACT')
embeddings = model.encode(sentences)
print(embeddings)
Advanced Usage
Without sentence-transformers, you can use the model as follows: First, pass your input through the transformer model, then apply the right pooling-operation on top of the contextualized word embeddings.
from transformers import AutoTokenizer, AutoModel
import torch
def mean_pooling(model_output, attention_mask):
token_embeddings = model_output[0]
input_mask_expanded = attention_mask.unsqueeze(-1).expand(token_embeddings.size()).float()
return torch.sum(token_embeddings * input_mask_expanded, 1) / torch.clamp(input_mask_expanded.sum(1), min=1e-9)
sentences = ['This is an example sentence', 'Each sentence is converted']
tokenizer = AutoTokenizer.from_pretrained('S-PubMedBert-MS-MARCO-SCIFACT')
model = AutoModel.from_pretrained('S-PubMedBert-MS-MARCO-SCIFACT')
encoded_input = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')
with torch.no_grad():
model_output = model(**encoded_input)
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])
print("Sentence embeddings:")
print(sentence_embeddings)
đ Documentation
Evaluation Results
For an automated evaluation of this model, see the Sentence Embeddings Benchmark: https://seb.sbert.net
Training
The model was trained with the following parameters:
DataLoader:
sentence_transformers.datasets.NoDuplicatesDataLoader.NoDuplicatesDataLoader
of length 560 with parameters:
{'batch_size': 16}
Loss:
sentence_transformers.losses.MultipleNegativesRankingLoss.MultipleNegativesRankingLoss
with parameters:
{'scale': 20.0, 'similarity_fct': 'cos_sim'}
Parameters of the fit()-Method:
{
"callback": null,
"epochs": 1,
"evaluation_steps": 10000,
"evaluator": "sentence_transformers.evaluation.SequentialEvaluator.SequentialEvaluator",
"max_grad_norm": 1,
"optimizer_class": "<class 'transformers.optimization.AdamW'>",
"optimizer_params": {
"correct_bias": false,
"eps": 1e-06,
"lr": 2e-05
},
"scheduler": "WarmupLinear",
"steps_per_epoch": null,
"warmup_steps": 56,
"weight_decay": 0.01
}
Full Model Architecture
SentenceTransformer(
(0): Transformer({'max_seq_length': 350, 'do_lower_case': False}) with Transformer model: BertModel
(1): Pooling({'word_embedding_dimension': 768, 'pooling_mode_cls_token': False, 'pooling_mode_mean_tokens': True, 'pooling_mode_max_tokens': False, 'pooling_mode_mean_sqrt_len_tokens': False})
)
đ License
Citing & Authors
If you use this model, cite the following paper:
@article{deka2022improved,
title={Improved Methods To Aid Unsupervised Evidence-Based Fact Checking For Online Health News},
author={Deka, Pritam and Jurek-Loughrey, Anna and Deepak, P},
journal={Journal of Data Intelligence},
volume={3},
number={4},
pages={474--504},
year={2022}
}