Is this model finetuned with MsMarco or mMarco
Hello, thank you for releasing this multilingual reranker model with Apache 2.0 license. I'd like to ask if you used any non-commercially available datasets (e.g. MsMarco, mMARCO) for finetuning /training this model? In the paper ( https://arxiv.org/pdf/2407.19669), section B.2 states that MS MARCO and mMARCO-zh were used. These datasets are for research purpose only.
Could you please clarify?
Thanks.
Thank you for your inquiry regarding the multilingual reranker model. We appreciate your interest in our work.
To clarify, the model does leverage the MS MARCO and mMARCO-zh datasets, which are indeed intended for research purposes only. We acknowledge the restrictions associated with these datasets and ensure that all usage complies with the terms provided by the dataset creators.
The findings presented in the paper reflect our commitment to using high-quality, publicly available data while adhering to the specified licensing agreements.
Best regards!
thanks for your quick response!