gryf/coach
1
0
Fork 0
mirror of https://github.com/gryf/coach.git synced 2025-12-17 19:20:19 +01:00
Code Issues Projects Releases Wiki Activity

Adding target reward and target sucess (#58)

Browse Source 
* Adding target reward

* Adding target successs

* Addressing comments

* Using custom_reward_threshold and target_success_rate

* Adding exit message

* Moving success rate to environment

* Making target_success_rate optional
This commit is contained in:
Ajay Deshpande
2018-11-12 15:03:43 -08:00
committed by Balaji Subramaniam
parent 0fe583186e
commit 875d6ef017
17 changed files with 162 additions and 74 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
rl_coach/training_worker.py
View File

@@ -40,8 +40,9 @@ def training_worker(graph_manager, checkpoint_dir):
graph_manager.phase = core_types.RunPhase.UNDEFINED
if steps * graph_manager.agent_params.algorithm.num_consecutive_playing_steps.num_steps > graph_manager.steps_between_evaluation_periods.num_steps * eval_offset:
graph_manager.evaluate(graph_manager.evaluation_steps)
eval_offset += 1
if graph_manager.evaluate(graph_manager.evaluation_steps):
break
if graph_manager.agent_params.algorithm.distributed_coach_synchronization_type == DistributedCoachSynchronizationType.SYNC:
graph_manager.save_checkpoint()