Both audio agents and non-audio agents are well trained.
Cli args see train_main.py
需要修改tianshou模块中data/buffer.py中的_add_to_buffer函数以支持形状不同的numpy对象进行训练(当前tianshou版本号0.2.2) Plz modify data/buffer.py in module named "tianshou". In tianshou V0.2.2, function named "_add_to_buffer" should be like this to support multimedia training.
def _add_to_buffer(self, name, inst):
if inst is None:
if getattr(self, name, None) is None:
self.__dict__[name] = None
return
if self.__dict__.get(name, None) is None:
if isinstance(inst, np.ndarray):
self.__dict__[name] = np.zeros([self._maxsize, *inst.shape], dtype=inst.dtype) # my modification
elif isinstance(inst, dict):
self.__dict__[name] = np.array(
[{} for _ in range(self._maxsize)])
else: # assume "inst" is a number
self.__dict__[name] = np.zeros([self._maxsize])
if isinstance(inst, np.ndarray) and \
self.__dict__[name].shape[1:] != inst.shape:
self.__dict__[name] = np.zeros([self._maxsize, *inst.shape])
self.__dict__[name][self._index] = inst