danijar dreamerv2 issues

danijar / dreamerv2

Mastering Atari with Discrete World Models

https://danijar.com/dreamerv2

MIT License

898 stars 195 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Question about advantage calculation

#60 leeacord opened 4 months ago
0
Code For: "Algorithm 1: Straight-Through Gradients with Automatic Differentiation"

#59 lucasmgomez closed 6 months ago
1
Are the actions properly fed into the model?

#58 NagisaZj opened 9 months ago
0
Cannot reproduce Atari Pong scores

#57 mlinda96 opened 1 year ago
0
How to reproduce DayDreamer's results in A1 simulator?

#56 Sapio-S opened 1 year ago
0
Outdated dependencies and broken examples

#55 Namnodorel opened 1 year ago
1
Fix Docker build settings for January 20, 2023

#54 ricrowl opened 1 year ago
0
Performance difference between TruncNormal and TanhNormal

#53 IcarusWizard opened 1 year ago
0
Reward different on evaluation

#52 ipsec closed 1 year ago
1
the Desire of Hyperparameters of Humanoid-Walk

#51 XueruiSu opened 1 year ago
0
Understanding re-clipping in Truncated Normal distribution

#50 pvskand closed 1 year ago
1
How does dreamerv2 perform on feature-based tasks?

#49 xlnwel closed 2 years ago
4
Prediction returning the same action from different observations

#48 ipsec closed 2 years ago
0
Update Dockerfile to use tensorflow/tensorflow:2.10.0-gpu

#47 ikeyasu opened 2 years ago
0
Minimal evaluation/example using gym observation

#46 ipsec closed 2 years ago
4
ValueError: . Tensor must have rank 4. Received rank 3, shape (208, 64, 64)

#45 ipsec closed 2 years ago
1
Why share states across random batches for training the world model?

#44 sai-prasanna closed 2 years ago
1
Questions about expl.py and updating the batch dataset

#43 Ashminator closed 2 years ago
2
Questions on Imagination MDP and imagination horizon H = 15

#42 GoingMyWay closed 2 years ago
0
Why stop-grad on actor's input state in imagine() function ?

#41 tominku closed 2 years ago
1
replay data memory usage?

#40 tominku closed 2 years ago
1
Can't reproduce riverraid's results

#39 luizapozzobon closed 2 years ago
2
Update the way env.py receives RAM state

#38 amshin98 opened 2 years ago
0
Straight-thru gradients vs Gumbel Softmax

#37 zplizzi closed 2 years ago
1
Should policy state be reset after every episode?

#36 edwhu closed 2 years ago
1
Batch size = 16?

#35 mctigger closed 2 years ago
1
Plot.py not working properly

#34 lcdbezerra closed 2 years ago
1
What does "openl" do / mean?

#33 edwhu closed 2 years ago
1
Fix sum KL distribution across both latent dims.

#32 fvisin closed 2 years ago
1
KeyError: 'dmc' while trying to run walker?

#31 mrmarten closed 2 years ago
2
The result for atari enduro in the paper is not reproduced

#30 jsikyoon closed 2 years ago
2
How many environment steps per update?

#29 mctigger closed 2 years ago
3
procgen env

#28 hlsfin closed 2 years ago
1
Offsets in actor loss calculation

#27 mctigger closed 2 years ago
1
How to save and reload trained dreamerv2 models

#26 Adaickalavan closed 2 years ago
1
Lamba Target Equation

#25 lewisboyd closed 3 years ago
4
How to run dreamerv2 on atari games

#24 KimiakiShirahama closed 2 years ago
1
Question about Plan2explore

#23 TachikakaMin closed 2 years ago
1
AssertionError and AttributeError dreamerv2 in jupyter-notebook

#22 balloch closed 3 years ago
1
Change ```eval_envs``` to ```num_eval_envs```

#21 alirahkay closed 2 years ago
1
Does the actor-critc train using only the stochastic state?

#20 lewisboyd closed 3 years ago
4
Skipped short episode of length 10.

#19 robjlyons closed 3 years ago
1
Discount predictor invalid log_prob targets?

#18 niklasdbs closed 3 years ago
1
Questions about atari evaluation protocol

#17 jmkim0309 closed 3 years ago
1
Pickle and shape issues

#16 robjlyons closed 3 years ago
1
Tuple Actions Space

#15 robjlyons closed 3 years ago
1
Intrinsic Rewards

#14 robjlyons closed 3 years ago
2
Render episodes

#13 robjlyons closed 3 years ago
1
Setting random seed

#12 izkula closed 3 years ago
1
Commented version of the code

#11 juliancodaforno closed 3 years ago
1