---
language: en
tags:
- bert
- regression
- biencoder
- similarity
pipeline_tag: text-similarity
---

# BiEncoder Regression Model

This model is a BiEncoder architecture that outputs similarity scores between text pairs.

## Model Details
- Base Model: bert-base-uncased
- Task: Regression
- Architecture: BiEncoder with cosine similarity
- Loss Function: mae

## Usage

```python
from transformers import AutoTokenizer, AutoModel
from modeling import BiEncoderModelRegression

# Load model components
tokenizer = AutoTokenizer.from_pretrained("minoosh/bert-reg-biencoder-mae")
base_model = AutoModel.from_pretrained("bert-base-uncased")
model = BiEncoderModelRegression(base_model, loss_fn="mae")

# Load weights
state_dict = torch.load("pytorch_model.bin")
model.load_state_dict(state_dict)

# Prepare inputs
texts1 = ["first text"]
texts2 = ["second text"]
inputs = tokenizer(
    texts1, texts2,
    padding=True,
    truncation=True,
    return_tensors="pt"
)

# Get similarity scores
outputs = model(**inputs)
similarity_scores = outputs["logits"]
```

## Metrics
The model was trained using mae loss and evaluated using:
- Mean Squared Error (MSE)
- Mean Absolute Error (MAE)
- Pearson Correlation
- Spearman Correlation
- Cosine Similarity