I am making an agent using the following code
Configuring an Agent: This agent use Defatlt settings from RayRLlib
config = (
# The RL algorithm
PPOConfig()
# The gym Env
.environment(env)
# Numbers of rollout workers
.rollouts(num_rollout_workers = 1)
# Use “torch” for using pytorch or “tf” or “tf2” for Tensor Flow.
.framework(“tf”)
# Numbers of evalution workers
.evaluation(evaluation_num_workers=1)
)
#Building the agent with the configurations mentioned above
agent = config.build()
is it possible to specify a dir manually so that when I perform agent.train() the tensorboard log will be stored in the mentioned dir?