This model is a fine-tuned version of xlm-roberta-base, specifically trained on a massive corpus of Persian data to create high-quality contextual embeddings for Persian sentences and paragraphs. It is designed to perform exceptionally well on tasks such as semantic search, clustering, and contextual similarity for Persian text, while also supporting multilingual tasks in English and Persian.

The fine-tuning process focused on adapting the pre-trained multilingual XLM-RoBERTa model to better capture Persian linguistic nuances, making it highly effective for tasks requiring embeddings tailored to the Persian language.

Usage (Sentence-Transformers)

Using this model becomes easy when you have sentence-transformers installed:

pip install -U sentence-transformers

Then you can use the model like this:

python

from sentence_transformers import SentenceTransformer
sentences = ['What are Large Language Models?','مدل های زبانی بزرگ چه هستند؟']

model = SentenceTransformer('heydariAI/persian-embeddings')
embeddings = model.encode(sentences)
print(embeddings)

Usage (HuggingFace Transformers)

Without sentence-transformers, you can use the model like this: First, you pass your input through the transformer model, then you have to apply the right pooling-operation on-top of the contextualized word embeddings.

python

from transformers import AutoTokenizer, AutoModel
import torch


#Mean Pooling - Take attention mask into account for correct averaging
def mean_pooling(model_output, attention_mask):
    token_embeddings = model_output[0] #First element of model_output contains all token embeddings
    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 we want sentence embeddings for
sentences = ['what are Large Language Models?', 'مدل های زبانی بزرگ چه هستند؟']

# Load model from HuggingFace Hub
tokenizer = AutoTokenizer.from_pretrained('heydariAI/persian-embeddings')
model = AutoModel.from_pretrained('heydariAI/persian-embeddings')

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

# Compute token embeddings
with torch.no_grad():
    model_output = model(**encoded_input)

# Perform pooling. In this case, mean pooling.
sentence_embeddings = mean_pooling(model_output, encoded_input['attention_mask'])

print("Sentence embeddings:")
print(sentence_embeddings)