gryf/coach
1
0
Fork 0
mirror of https://github.com/gryf/coach.git synced 2025-12-18 03:30:19 +01:00
Code Issues Projects Releases Wiki Activity

Add is_on_policy property to agents (#480)

Browse Source
This commit is contained in:
Guy Jacob
2021-05-06 18:02:02 +03:00
committed by GitHub
parent 06bacd9de0
commit 9106b69227
21 changed files with 86 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
rl_coach/agents/acer_agent.py
View File

@@ -111,8 +111,11 @@ class ACERAgent(PolicyOptimizationAgent):
self.V_Values = self.register_signal('Values')
self.kl_divergence = self.register_signal('KL Divergence')
def _learn_from_batch(self, batch):
@property
def is_on_policy(self) -> bool:
return False
def _learn_from_batch(self, batch):
fetches = [self.networks['main'].online_network.output_heads[1].probability_loss,
self.networks['main'].online_network.output_heads[1].bias_correction_loss,
self.networks['main'].online_network.output_heads[1].kl_divergence]