Skip to content

Latest commit

 

History

History
14 lines (9 loc) · 956 Bytes

README.md

File metadata and controls

14 lines (9 loc) · 956 Bytes

BORT2

Pytorch implementation of BORT2 ("Zhongying Deng, Da Li, Yi-Zhe Song, Tao Xiang. Robust Target Training for Multi-Source Domain Adaptation". BMVC 2022)

Installation

Please install the Dassl codebase, and then copy the files in this repository to Dassl. When asked to overwrite, say yes (some __init__.py files may be overwritten, these files is modified to include the trainer used for BORT2). Then pytorch 1.7.1 + cuda 10.1, python 3.7 will be installed.

This is a must, otherwise, some files may be missing.

============ code =============

The zip file Dassl_BORT2.zip contains an initial version of the code for BORT2, but it will be further updated. The instructions for using the code can be found at (this issue)