|
--- |
|
license: apache-2.0 |
|
--- |
|
|
|
# Model description |
|
|
|
This is a multivector retrieval model that is built upon the efficient [XTR](https://proceedings.neurips.cc/paper_files/paper/2023/hash/31d997278ee9069d6721bc194174bb4c-Abstract-Conference.html) architecture. The model utilizes [t5-base](https://huggingface.co/t5-base) encoder, and is finetuned on [MSMarco](https://huggingface.co/datasets/BeIR/msmarco) dataset with XTR objective. |
|
|
|
# Overview |
|
|
|
*Language model*: t5-base (encoder) \ |
|
*Language*: English \ |
|
*Task*: Multi-vector retrieval \ |
|
*Data*: MSMarco |
|
|
|
# Intented use |
|
This model can be utilized for text retrieval (IR). You can find detailed instructions and examples on how to use this model for text retrieval in the documentation [here](https://github.com/primeqa/primeqa/blob/xtr/notebooks/ir/dense/xtr_retrieval.ipynb). |
