Modelpublic

RikoteMaster/MNLP_M3_document_encoder

sourceHugging Faceapache-2.0updated 1y agoView on Hugging Face

0likes71downloads

Model Card

Granite Embedding 107M Multilingual

This is a copy of the ibm-granite/granite-embedding-107m-multilingual model for document encoding purposes.

Model Summary

Granite-Embedding-107M-Multilingual is a 107M parameter dense biencoder embedding model from the Granite Embeddings suite that can be used to generate high quality text embeddings. This model produces embedding vectors of size 384.

Supported Languages

English, German, Spanish, French, Japanese, Portuguese, Arabic, Czech, Italian, Korean, Dutch, and Chinese.

Usage

With Sentence Transformers

python

from sentence_transformers import SentenceTransformer

model = SentenceTransformer('RikoteMaster/MNLP_M3_document_encoder')
embeddings = model.encode(['Your text here'])

With Transformers

python

from transformers import AutoModel, AutoTokenizer
import torch

model = AutoModel.from_pretrained('RikoteMaster/MNLP_M3_document_encoder')
tokenizer = AutoTokenizer.from_pretrained('RikoteMaster/MNLP_M3_document_encoder')

inputs = tokenizer(['Your text here'], return_tensors='pt', padding=True, truncation=True)
with torch.no_grad():
    outputs = model(**inputs)
    embeddings = outputs.last_hidden_state[:, 0]  # CLS pooling
    embeddings = torch.nn.functional.normalize(embeddings, dim=1)

Original Model

This model is based on ibm-granite/granite-embedding-107m-multilingual by IBM.