search

lhoyer / HRDA

[ECCV22] Official Implementation of HRDA: Context-Aware High-Resolution Domain-Adaptive Semantic Segmentation
Other
231 stars 32 forks source link

what is the difference between run_experiments.py and tools/train.py? #31

Closed wzr0108 closed 1 year ago

lhoyer commented 1 year ago

run_experiments.py is a wrapper that manages the generation of the configs according to experiments.py. Also, it properly sets up the log paths.

wzr0108 commented 1 year ago

Thank you for your answer. For model training, can tools/train.py work normally?

lhoyer commented 1 year ago

If you manually edit the config file according to the modifications done in run_experiments.py, you can also train directly with tools/train.py. However, to make sure that everything works as intended, I would recommend to follow the instructions in the README.md