def my_policy_1155152886(num_envs=1):
"""We will use this function to load your agent and then testing.
Make sure this function can run bug-free, when the working directory is
"ierg5350-assignment/assignment5/"
You can rewrite this function completely if you have custom agents, but you
need to make sure the codes is bug-free and add necessary description on the notebook.
Please rename this function!!! We will use program to automatically detect your agent, so a wrong function name
will fail the evaluation. Run this file directly to make sure everything is fine.
"""
# [TODO] rewrite this function
# [TODO] CAUTION! PLEASE CHANGE THE NAME OF THIS FUNCTION!!! Otherwise our program can't find your agent!
my_agent_log_dir = 'data/cCarRacing-v0_PPO_12-01_21-39'
my_agent_suffix = 'final'
# checkpoint_path = osp.join(my_agent_log_dir, "checkpoint-{}.pkl".format(my_agent_suffix))
# if not osp.exists(checkpoint_path):
# raise ValueError("Can't find anything at {}!".format(checkpoint_path))
# else:
# print("Found your checkpoint at {}!".format(checkpoint_path))
return PolicyAPI("cCarRacing-v0",
num_envs=num_envs,
log_dir=my_agent_log_dir,
suffix=my_agent_suffix)
def student_compute_action_function(num_envs=1):
"""We will use this function to load your agent and then testing.
Make sure this function can run bug-free, when the working directory is
"ierg6130-assignment/assignment4/"
You can rewrite this function completely if you have custom agents, but you
need to make sure the codes is bug-free and add necessary description on
report_SID.md
Run this file directly to make sure everything is fine.
"""
# [TODO] rewrite this function
my_agent_log_dir = ""
my_agent_suffix = ""
checkpoint_path = osp.join(my_agent_log_dir,
"checkpoint-{}.pkl".format(my_agent_suffix))
if not osp.exists(checkpoint_path):
print("Can't find anything at {}!".format(checkpoint_path))
else:
print("Found your checkpoint at {}!".format(checkpoint_path))
return PolicyAPI(num_envs=num_envs,
log_dir=my_agent_log_dir,
suffix=my_agent_suffix)
def my_policy_zhenghao(num_envs=1):
return PolicyAPI("cCarRacing-v0",
num_envs=num_envs,
log_dir="data/alphacar",
suffix="zhenghao")
def my_policy_1155156694(num_envs=1):
return PolicyAPI("cCarRacing-v0",
num_envs=num_envs,
log_dir="data/alphacar",
suffix="alphacar")
def train(args):
# Verify algorithm and config
algo = args.algo
if algo == "PPO":
config = ppo_config
else:
raise ValueError("args.algo must in [PPO]")
config.num_envs = args.num_envs
config.lr = args.lr
config.entropy_loss_weight = args.entropy
assert args.env_id in ["cPong-v0", "cCarRacing-v0"], args.env_id
# Seed the environments and setup torch
seed = args.seed
torch.manual_seed(seed)
if torch.cuda.is_available():
torch.cuda.manual_seed(seed)
torch.set_num_threads(1)
# Create vectorized environments
num_envs = args.num_envs
env_id = args.env_id if not args.opponent else "cCarRacingDouble-v0"
# Clean log directory
log_dir = verify_log_dir(
args.log_dir,
"{}_{}_{}".format(env_id, algo,
datetime.datetime.now().strftime("%m-%d_%H-%M")))
if args.opponent:
assert args.num_eval_envs == 0
from competitive_rl.car_racing import make_competitive_car_racing
from load_agents import PolicyAPI
restore_log_dir = os.path.dirname(args.restore)
restore_suffix = os.path.basename(
args.restore).split("checkpoint-")[1].split(".pkl")[0]
opponent_policy = PolicyAPI("cCarRacing-v0",
num_envs=1,
log_dir=restore_log_dir,
suffix=restore_suffix)
envs = make_competitive_car_racing(opponent_policy=opponent_policy,
num_envs=num_envs,
asynchronous=not args.test)
else:
envs = make_envs(env_id=env_id,
seed=seed,
log_dir=log_dir,
num_envs=num_envs,
asynchronous=not args.test,
resized_dim=config.resized_dim,
action_repeat=args.action_repeat)
if args.num_eval_envs > 0:
eval_envs = make_envs(env_id=env_id,
seed=seed,
log_dir=log_dir,
num_envs=args.num_eval_envs,
asynchronous=not args.test,
resized_dim=config.resized_dim,
action_repeat=args.action_repeat)
else:
eval_envs = None
# Setup trainer
if algo == "PPO":
trainer = PPOTrainer(envs, config)
else:
raise ValueError("Unknown algorithm {}".format(algo))
if args.restore:
restore_log_dir = os.path.dirname(args.restore)
restore_suffix = os.path.basename(
args.restore).split("checkpoint-")[1].split(".pkl")[0]
success = trainer.load_w(restore_log_dir, restore_suffix)
if not success:
raise ValueError(
"We can't restore your agent. The log_dir is {} and the suffix is {}"
.format(restore_log_dir, restore_suffix))
# Start training
print("Start training!")
obs = envs.reset()
# frame_stack_tensor.update(obs)
raw_obs = trainer.process_obs(obs)
processed_obs = trainer.model.world_model(raw_obs)
trainer.rollouts.before_update(obs, processed_obs)
try:
_train(trainer, envs, eval_envs, config, num_envs, algo, log_dir,
False, False)
except KeyboardInterrupt:
print(
"The training is stopped by user. The log directory is {}. Now we finish the training."
.format(log_dir))
trainer.save_w(log_dir, "final")
envs.close()