BLEURT 20 D12

🚀 Custom Transformer-based BLEURT Model

This model is built upon a custom Transformer architecture. It provides an efficient way to evaluate text similarity using the BLEURT metric.

🚀 Quick Start

📦 Installation

You can install the model with the following command:

pip install git+https://github.com/lucadiliello/bleurt-pytorch.git

💻 Usage Examples

Basic Usage

Now, let's load the model and make predictions:

import torch
from bleurt_pytorch import BleurtConfig, BleurtForSequenceClassification, BleurtTokenizer

config = BleurtConfig.from_pretrained('lucadiliello/BLEURT-20-D12')
model = BleurtForSequenceClassification.from_pretrained('lucadiliello/BLEURT-20-D12')
tokenizer = BleurtTokenizer.from_pretrained('lucadiliello/BLEURT-20-D12')

references = ["a bird chirps by the window", "this is a random sentence"]
candidates = ["a bird chirps by the window", "this looks like a random sentence"]

model.eval()
with torch.no_grad():
    inputs = tokenizer(references, candidates, padding='longest', return_tensors='pt')
    res = model(**inputs).logits.flatten().tolist()
print(res)
# [0.9604414105415344, 0.8080050349235535]

📚 Documentation

For the detailed definition of BleurtConfig, BleurtForSequenceClassification and BleurtTokenizer in PyTorch, please refer to this repository.