Right now `LitModular` does not support `ChainedScheduler`. This is because a `ChainedScheduler` does not take an optimizer, but rather a list of schedulers: https://pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.ChainedScheduler.html `LitModular` assumes `lr_scheduler` is a callable that receives an Optimizer: https://github.com/IntelLabs/MART/blob/41ad18e3aa08b420267772bd197e8c610389d2cc/mart/models/modular.py#L122 My sense is the best way to fix this is to allow `lr_scheduler` to be a list and we can create the underlying `ChainedScheduler` within `LitModular`.
Right now
LitModulardoes not supportChainedScheduler. This is because aChainedSchedulerdoes not take an optimizer, but rather a list of schedulers:https://pytorch.org/docs/stable/generated/torch.optim.lr_scheduler.ChainedScheduler.html
LitModularassumeslr_scheduleris a callable that receives an Optimizer:MART/mart/models/modular.py
Line 122 in 41ad18e
My sense is the best way to fix this is to allow
lr_schedulerto be a list and we can create the underlyingChainedSchedulerwithinLitModular.