def update_nets(self):
'''Update target networks'''
if util.frame_mod(self.body.env.clock.frame, self.q1_net.update_frequency, self.body.env.num_envs):
if self.q1_net.update_type == 'replace':
net_util.copy(self.q1_net, self.target_q1_net)
net_util.copy(self.q2_net, self.target_q2_net)
elif self.q1_net.update_type == 'polyak':
net_util.polyak_update(self.q1_net, self.target_q1_net, self.q1_net.polyak_coef)
net_util.polyak_update(self.q2_net, self.target_q2_net, self.q2_net.polyak_coef)
else:
raise ValueError('Unknown q1_net.update_type. Should be "replace" or "polyak". Exiting.')
def to_ckpt(self, env, mode='eval'):
'''Check with clock whether to run log/eval ckpt: at the start, save_freq, and the end'''
if mode == 'eval' and util.in_eval_lab_modes(
): # avoid double-eval: eval-ckpt in eval mode
return False
clock = env.clock
frame = clock.get()
frequency = env.eval_frequency if mode == 'eval' else env.log_frequency
to_ckpt = util.frame_mod(frame, frequency,
env.num_envs) or frame == clock.max_frame
return to_ckpt
def to_ckpt(self, env, mode='eval'):
'''Check with clock whether to run log/eval ckpt: at the start, save_freq, and the end'''
if mode == 'eval' and util.in_eval_lab_modes(
): # avoid double-eval: eval-ckpt in eval mode
return False
clock = env.clock
frame = clock.get()
frequency = env.eval_frequency if mode == 'eval' else env.log_frequency
if frequency is None: # default episodic
to_ckpt = env.done
else: # normal ckpt condition by mod remainder (general for venv)
to_ckpt = util.frame_mod(frame, frequency,
env.num_envs) or frame == clock.max_frame
return to_ckpt
