How to add custom LR scheduler?

[JasspalSingh]

I have configured the code for custom Lr scheduler

def build_lr_scheduler(cls, cfg, optimizer):
"""
Build a LR scheduler from config.
"""
name = cfg.SOLVER.LR_SCHEDULER_NAME
if name == "ReduceLROnPlateau":
return ReduceLROnPlateau(
optimizer,
mode='min', factor=cfg.SOLVER.GAMMA, patience=1000,
threshold=0.001, threshold_mode='rel', cooldown=0,
min_lr=1e-06, eps=1e-08,verbose=True
)
else:
raise ValueError("Unknown LR scheduler: {}".format(name))

• how can I employ this to update my learning rate.

I am getting error and solution to it is we need to give loss to 'ReduceLROnPlateau' at every iteration. How can I do it?