BAAI
/

如何使用MTEB评估BAAI/bge-reranker-v2-m3的C-MTEB Reranking项目

#33
by IeohMingChan - opened

MTEB使用SentenceTransformer加载模型,而在README中并未提供使用SentenceTransformer的示例代码,我使用MTEB官方提供的如下代码评估本模型时效果远不及预期,请问能否提供MTEB评估或使用SentenceTransformer的示例代码?
MTEB uses SentenceTransformer to load the model, but there is no example code provided in the README for using SentenceTransformer. When I evaluated the model using the following code provided by MTEB, the effect was far from expected. Can you provide an example code for MTEB evaluation or using SentenceTransformer?

import mteb
from sentence_transformers import SentenceTransformer
import torch

model_name = "BAAI/bge-reranker-v2-m3"
model = SentenceTransformer(model_name,trust_remote_code=True)
tasks = mteb.get_tasks(tasks=["CMedQAv1-reranking","CMedQAv2-reranking","MMarcoReranking","T2Reranking"],languages="cmn"])
evaluation = mteb.MTEB(tasks=tasks)
results = evaluation.run(model, output_folder=f"results/{model_name}")

Sign up or log in to comment