README.md

---
pipeline_tag: sentence-similarity
tags:
- sentence-transformers
- feature-extraction
- sentence-similarity
- transformers
base_model: NeuML/pubmedbert-base-embeddings
language: en
license: apache-2.0
---

# BiomedBERT Base Embeddings

This is the [PubMedBERT-base-embeddings](https://hf.co/neuml/pubmedbert-base-embeddings) model fined-tuned using [sentence-transformers](https://www.SBERT.net). It maps sentences & paragraphs to a 768 dimensional dense vector space and can be used for tasks like clustering or semantic search.

The training dataset was generated using a random sample of [PubMed](https://pubmed.ncbi.nlm.nih.gov/) title-abstract pairs along with similar title pairs. The training workflow was a distillation process as follows.

- Build a distilled dataset of teacher scores using the [biomedbert-base-reranker](https://hf.co/neuml/biomedbert-base-reranker) cross-encoder for a separate random sample of title-abstract pairs.
- Further fine-tune the model on the distilled dataset using [KLDivLoss](https://github.com/huggingface/sentence-transformers/blob/main/sentence_transformers/losses/DistillKLDivLoss.py).

This model gives the original PubMedBERT Embeddings model an accuracy boost via the same method uses to train smaller model variants.

## Usage (txtai)

This model can be used to build embeddings databases with [txtai](https://github.com/neuml/txtai) for semantic search and/or as a knowledge source for retrieval augmented generation (RAG).

```python
import txtai

embeddings = txtai.Embeddings(path="neuml/biomedbert-base-embeddings", content=True)
embeddings.index(documents())

# Run a query
embeddings.search("query to run")
```

## Usage (Sentence-Transformers)

Alternatively, the model can be loaded with [sentence-transformers](https://www.SBERT.net).

```python
from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]

model = SentenceTransformer("neuml/biomedbert-base-embeddings")
embeddings = model.encode(sentences)
print(embeddings)
```

## Usage (Hugging Face Transformers)

The model can also be used directly with Transformers. 

```python
from transformers import AutoTokenizer, AutoModel
import torch

# Mean Pooling - Take attention mask into account for correct averaging
def meanpooling(output, mask):
    embeddings = output[0] # First element of model_output contains all token embeddings
    mask = mask.unsqueeze(-1).expand(embeddings.size()).float()
    return torch.sum(embeddings * mask, 1) / torch.clamp(mask.sum(1), min=1e-9)

# Sentences we want sentence embeddings for
sentences = ['This is an example sentence', 'Each sentence is converted']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained("neuml/biomedbert-base-embeddings")
model = AutoModel.from_pretrained("neuml/biomedbert-base-embeddings")

# Tokenize sentences
inputs = tokenizer(sentences, padding=True, truncation=True, return_tensors='pt')

# Compute token embeddings
with torch.no_grad():
    output = model(**inputs)

# Perform pooling. In this case, mean pooling.
embeddings = meanpooling(output, inputs['attention_mask'])

print("Sentence embeddings:")
print(embeddings)
```

## Evaluation Results

Performance of this model compared to the top base models on the [MTEB leaderboard](https://huggingface.co/spaces/mteb/leaderboard) is shown below. A popular smaller model was also evaluated along with the most downloaded PubMed similarity model on the Hugging Face Hub.

The following datasets were used to evaluate model performance.

- [PubMed QA](https://huggingface.co/datasets/qiaojin/PubMedQA)
  - Subset: pqa_labeled, Split: train, Pair: (question, long_answer)
- [PubMed Subset](https://huggingface.co/datasets/awinml/pubmed_abstract_3_1k)
  - Split: test, Pair: (title, text)
- [PubMed Summary](https://huggingface.co/datasets/armanc/scientific_papers)
  - Subset: pubmed, Split: validation, Pair: (article, abstract)

Evaluation results are shown below. The [Pearson correlation coefficient](https://en.wikipedia.org/wiki/Pearson_correlation_coefficient) is used as the evaluation metric.

| Model                                                 | PubMed QA | PubMed Subset | PubMed Summary | Average   |
| ----------------------------------------------------- | --------- | ------------- | -------------- | --------- |
| [all-MiniLM-L6-v2](https://hf.co/sentence-transformers/all-MiniLM-L6-v2) | 90.40  | 95.92 | 94.07 | 93.46 |
| [biomedbert-base-colbert](https://hf.co/neuml/biomedbert-base-colbert)  | 94.59 | 97.18 | 96.21  | 95.99 |
| [**biomedbert-base-embeddings**](https://hf.co/neuml/biomedbert-base-embeddings)  | **94.60** | **98.39**  | **97.61** | **96.87** |
| [biomedbert-base-reranker](https://hf.co/neuml/biomedbert-base-reranker)  | 97.66 | 99.76  | 98.81 | 98.74 |
| [biomedbert-small-colbert](https://hf.co/neuml/biomedbert-small-colbert)  | 93.51 | 97.20 | 95.85 | 95.52 |
| [biomedbert-small-embeddings](https://hf.co/neuml/biomedbert-small-embeddings)  | 93.25 | 97.93  | 96.65 | 95.94 |
| [biomedbert-hash-nano-embeddings](https://hf.co/neuml/biomedbert-hash-nano-embeddings)  | 90.39 | 96.29 | 95.32 | 94.00 |
| [pubmedbert-base-embeddings](https://hf.co/neuml/pubmedbert-base-embeddings)       | 93.27  | 97.00 | 96.58 | 95.62 |

This model gives a sizable accuracy boost over the original PubMedBERT Embeddings model.

## Full Model Architecture

```
SentenceTransformer(
  (0): Transformer({'max_seq_length': 512, 'do_lower_case': False, 'architecture': '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, 'pooling_mode_weightedmean_tokens': False, 'pooling_mode_lasttoken': False, 'include_prompt': True})
)
```

## More Information

Read more about the model in [this article](https://huggingface.co/blog/NeuML/biomedbert-small).