b900fdf6f2
Closes #947 This removes all kwargs from all policy constructors. While doing that, I also improved several names and added a whole lot of TODOs. ## Functional changes: 1. Added possibility to pass None as `critic2` and `critic2_optim`. In fact, the default behavior then should cover the absolute majority of cases 2. Added a function called `clone_optimizer` as a temporary measure to support passing `critic2_optim=None` ## Breaking changes: 1. `action_space` is no longer optional. In fact, it already was non-optional, as there was a ValueError in BasePolicy.init. So now several examples were fixed to reflect that 2. `reward_normalization` removed from DDPG and children. It was never allowed to pass it as `True` there, an error would have been raised in `compute_n_step_reward`. Now I removed it from the interface 3. renamed `critic1` and similar to `critic`, in order to have uniform interfaces. Note that the `critic` in DDPG was optional for the sole reason that child classes used `critic1`. I removed this optionality (DDPG can't do anything with `critic=None`) 4. Several renamings of fields (mostly private to public, so backwards compatible) ## Additional changes: 1. Removed type and default declaration from docstring. This kind of duplication is really not necessary 2. Policy constructors are now only called using named arguments, not a fragile mixture of positional and named as before 5. Minor beautifications in typing and code 6. Generally shortened docstrings and made them uniform across all policies (hopefully) ## Comment: With these changes, several problems in tianshou's inheritance hierarchy become more apparent. I tried highlighting them for future work. --------- Co-authored-by: Dominik Jain <d.jain@appliedai.de>
41 lines
1.3 KiB
Python
41 lines
1.3 KiB
Python
import torch
|
|
|
|
|
|
class MultipleLRSchedulers:
|
|
"""A wrapper for multiple learning rate schedulers.
|
|
|
|
Every time :meth:`~tianshou.utils.MultipleLRSchedulers.step` is called,
|
|
it calls the step() method of each of the schedulers that it contains.
|
|
Example usage:
|
|
::
|
|
|
|
scheduler1 = ConstantLR(opt1, factor=0.1, total_iters=2)
|
|
scheduler2 = ExponentialLR(opt2, gamma=0.9)
|
|
scheduler = MultipleLRSchedulers(scheduler1, scheduler2)
|
|
policy = PPOPolicy(..., lr_scheduler=scheduler)
|
|
"""
|
|
|
|
def __init__(self, *args: torch.optim.lr_scheduler.LambdaLR):
|
|
self.schedulers = args
|
|
|
|
def step(self) -> None:
|
|
"""Take a step in each of the learning rate schedulers."""
|
|
for scheduler in self.schedulers:
|
|
scheduler.step()
|
|
|
|
def state_dict(self) -> list[dict]:
|
|
"""Get state_dict for each of the learning rate schedulers.
|
|
|
|
:return: A list of state_dict of learning rate schedulers.
|
|
"""
|
|
return [s.state_dict() for s in self.schedulers]
|
|
|
|
def load_state_dict(self, state_dict: list[dict]) -> None:
|
|
"""Load states from state_dict.
|
|
|
|
:param state_dict: A list of learning rate scheduler
|
|
state_dict, in the same order as the schedulers.
|
|
"""
|
|
for s, sd in zip(self.schedulers, state_dict, strict=True):
|
|
s.__dict__.update(sd)
|