BGE v1 & v1.5#

BGE v1 and v1.5 are series of encoder only models base on BERT. They achieved best performance among the models of the same size at the time of release.

BGE#

The first group of BGE models was released in Aug 2023. The bge-large-en and bge-large-zh ranked 1st on MTEB and C-MTEB benchmarks at the time released.

Model	Language	Parameters	Model Size	Description
BAAI/bge-large-en	English	335M	1.34 GB	Embedding Model which map text into vector
BAAI/bge-base-en	English	109M	438 MB	a base-scale model but with similar ability to BAAI/bge-large-en
BAAI/bge-small-en	English	33.4M	133 MB	a small-scale model but with competitive performance
BAAI/bge-large-zh	Chinese	326M	1.3 GB	Embedding Model which map text into vector
BAAI/bge-base-zh	Chinese	102M	409 MB	a base-scale model but with similar ability to BAAI/bge-large-zh
BAAI/bge-small-zh	Chinese	24M	95.8 MB	a small-scale model but with competitive performance

BGE-v1.5#

Then to enhance its retrieval ability without instruction and alleviate the issue of the similarity distribution, bge-*-v1.5 models were released in Sep 2023. They are still the most popular embedding models that balanced well between embedding quality and model sizes.

Model	Language	Parameters	Model Size	Description
BAAI/bge-large-en-v1.5	English	335M	1.34 GB	version 1.5 with more reasonable similarity distribution and better performance
BAAI/bge-base-en-v1.5	English	109M	438 MB
BAAI/bge-small-en-v1.5	English	33.4M	133 MB
BAAI/bge-large-zh-v1.5	Chinese	326M	1.3 GB
BAAI/bge-base-zh-v1.5	Chinese	102M	409 MB
BAAI/bge-small-zh-v1.5	Chinese	24M	95.8 MB

Usage#

To use BGE v1 or v1.5 model for inference, load model through

from FlagEmbedding import FlagModel

model = FlagModel('BAAI/bge-base-en-v1.5')

sentences = ["Hello world", "I am inevitable"]
embeddings = model.encode(sentences)

Tip

For simple tasks that only encode a few sentences like above, it’s faster to use CPU or a single GPU instead of multi-GPUs

To use CPU:

# make no GPU visible
import os
os.environ['CUDA_VISIBLE_DEVICES'] = ''

# or claim the devices during initialize the model
model = FlagModel('BAAI/bge-base-en-v1.5', devices='cpu')

To use a single GPU:

# select one sigle card to be visible
import os
os.environ['CUDA_VISIBLE_DEVICES'] = '0'

# or claim the devices during initialize the model
model = FlagModel('BAAI/bge-base-en-v1.5', devices=0)

Useful Links:

API

Tutorial

Example