Did 'Deterministic-v4' can run using the code?

[THSWind]

No description provided.

[bhatiaabhinav]

Sure, yes you can run. Let me know if you need any help with the installation, or the run command. I'll update the readme file today to make sure that everything is mentioned clearly. Thanks, Abhinav

…

On Sat, Aug 15, 2020, 9:46 AM THSWind ***@***.***> wrote: — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ABSAL5CZB4MKCOJPC27H6QDSA2GSHANCNFSM4QAJAQYA> .

[THSWind]

Sure, yes you can run. Let me know if you need any help with the installation, or the run command. I'll update the readme file today to make sure that everything is mentioned clearly. Thanks, Abhinav
…
On Sat, Aug 15, 2020, 9:46 AM THSWind @.***> wrote: — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#1>, or unsubscribe https://github.com/notifications/unsubscribe-auth/ABSAL5CZB4MKCOJPC27H6QDSA2GSHANCNFSM4QAJAQYA .

GREAT! Thanks for your reply, so kind of you, I appreciate of this.

[THSWind]

I found it can't run 'deterministic-v4', when I run it, I get an error:( when I change the SAC to DQN, I get the same error)

(py3.6) thswind@ubuntu-60:~/myitem2/RL-v2$ python -m RL BreakoutDeterministic-v4 SAC 10000000 --algo_suffix=mnih --seed=0 --conv1 32 8 4 0 --conv2 64 4 2 0 --conv3 64 3 1 0 --hiddens 512 --no_render --overwrite
wandb: Tracking run with wandb version 0.9.1
wandb: Wandb version 0.9.4 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Run data is saved locally in logs/BreakoutDeterministic-v4/SAC_mnih/wandb/run-20200816_104130-29012rca
wandb: Syncing run SAC_mnih
wandb: ⭐️ View project at https://app.wandb.ai/thswind/breakoutdeterministic-v4
wandb: 🚀 View run at https://app.wandb.ai/thswind/breakoutdeterministic-v4/runs/29012rca
wandb: Run wandb off to turn off syncing.

Traceback (most recent call last):
File "/home/sdb/618/thswind/anaconda3/envs/py3.6/lib/python3.8/runpy.py", line 194, in _run_module_as_main
return _run_code(code, main_globals, None,
File "/home/sdb/618/thswind/anaconda3/envs/py3.6/lib/python3.8/runpy.py", line 87, in _run_code
exec(code, run_globals)
File "/home/sdb/618/thswind/myitem2/RL-v2/RL/main.py", line 80, in
raise e
File "/home/sdb/618/thswind/myitem2/RL-v2/RL/main.py", line 75, in
m = RL.Manager(args.env_id, args.algo_id, args.algo_suffix, num_steps_to_run=args.num_steps_to_run,
File "/home/sdb/618/thswind/myitem2/RL-v2/RL/core/manager.py", line 49, in init
self.env = self.algo.wrap_env(gym.make(self.env_id))
File "/home/sdb/618/thswind/myitem2/RL-v2/RL/algorithms/standard_wrap_algo.py", line 39, in wrap_env
env = AtariPreprocessing(env, noop_max=args.atari_noop_max,
File "/home/sdb/618/thswind/anaconda3/envs/py3.6/lib/python3.8/site-packages/gym/wrappers/atari_preprocessing.py", line 51, in init
assert 'NoFrameskip' in env.spec.id, 'disable frame-skipping in the original env. for more than one'
AssertionError: disable frame-skipping in the original env. for more than one frame-skip as it will be done by the wrapper

You've done a great job
I have some suggestions and need some help(if you are willing to..):
Suggestions:
1 maybe you can list all of the algorithms which implemented, add environments it(algorithm) can run.
2 add some info about how to start wandb and stop it.

For HELP:
Do you have an easy to read, pure sac algorithm, or 2019, 2020 years of reinforcement learning based on Atari game. My code ability is poor. I haven't found a suitable one for a long time. Some algorithms have dealt with reward. I don't quite understand some metric. If it's convenient for you to provide it, thank you very much.
I need it to compare，
look forward to your reply~