I need help from you for resume training in other files by using RLlib

RLlib
1

I need help from you for resume training in other files by using RLlib

  • tain.py file

from ray import tune
trainingSteps = 10
trials = tune.run(
“PPO”,
name=“myenv”,
config={“env”:“CartPole-v1”},
stop={
“training_iteration”: trainingSteps,
},
local_dir=“cartppo”,
checkpoint_freq=1,
keep_checkpoints_num=1
)

what can I do now in the tain_resume.py file?

  • I want to do more 10 training_steps for resume training.
2 
checkpoint = trials.get_last_checkpoint()._local_path

print(checkpoint)

analysis = tune.run(

"PPO",

name="myenv",

config={"env":"CartPole-v1"},

stop={

"training_iteration": trainingSteps * 2,

},

local_dir="cartppo",

checkpoint_freq=1,

keep_checkpoints_num=1,

restore=checkpoint,

)