Stilwell-Git / Doubly-Bounded-Q-Learning

TensorFlow implementation for our paper "On the Estimation Bias in Double Q-Learning"
MIT License
1 stars 0 forks source link

unrecognized arguments: Alien' #1

Open 17713679014 opened 2 years ago

17713679014 commented 2 years ago

sorry.when I run the commond "python train.py --tag='DB-ADP Alien' --env=Alien",I get this error: usage: train.py [-h] [--tag TAG] [--env ENV] [--alg {dqn,cddqn}] [--learn {atari,atari_dbadp}] [--gamma GAMMA] [--sticky STICKY] [--noop NOOP] [--frames FRAMES] [--rews_scale REWS_SCALE] [--test_eps TEST_EPS] [--epochs EPOCHS] [--cycles CYCLES] [--iterations ITERATIONS] [--timesteps TIMESTEPS] [--test_rollouts TEST_ROLLOUTS] [--test_timesteps TEST_TIMESTEPS] [--save_rews SAVE_REWS] [--buffer {framestack}] [--buffer_size BUFFER_SIZE] [--batch_size BATCH_SIZE] [--warmup WARMUP] [--train_batches TRAIN_BATCHES] [--train_target TRAIN_TARGET] [--eps_l EPS_L] [--eps_r EPS_R] [--eps_decay EPS_DECAY] [--optimizer {adam,rmsprop}] [--q_lr Q_LR] [--Adam_eps ADAM_EPS] [--nstep NSTEP] [--double DOUBLE] [--dueling DUELING] train.py: error: unrecognized arguments: Alien' How can I solve it,thank you~

Stilwell-Git commented 2 years ago

This issue may be caused by a Windows-style CRLF character (i.e., '\r') at the end of the command line. Please check the line ending of the command script and ensure it uses a Unix-style LF ending. If the issue is not solved, do not hesitate to contact us, and we will try to figure it out.

17713679014 commented 2 years ago

Thank you for your reply,I have checked my commond,the problem still seems to be unsolved...... image

Stilwell-Git commented 2 years ago

Could you please try this command?

python train.py --tag='DB-ADP-Alien' --env=Alien
17713679014 commented 2 years ago

emmmmm,it seems turned out to be this image

Stilwell-Git commented 2 years ago

We update the code slightly. Please git pull and try again. There might be some issues running this code in Windows. It would be better to use a Linux system.

17713679014 commented 2 years ago

Thank you very much,Now,I use the Ubuntu16.04 to test the code,it turned out to be this: image Did I succeed? How can I view the effect of the experiment? it seems to at a standstill.

Stilwell-Git commented 2 years ago

Yes. Please wait for a while. The log will be printed in every 50000 environment steps.

17713679014 commented 2 years ago

OMG,Something bad happened again..... image

Stilwell-Git commented 2 years ago

We figure out a typo bug. Please git pull, and the issue should be solved.

17713679014 commented 2 years ago

Thank you very much~

17713679014 commented 2 years ago

Sorry,recently I run this code,I found that I could not complete the whole code training process.When the program iterated 350000 times, it would be automatically terminated.Is there somthing wrong with the code?I'm sorry to bother you again image

17713679014 commented 2 years ago

Are you from China?Can we chat with wechat?If you don't mind,I can pay you some money,thank you~