facebookresearch/mbrl-lib

Issues

• elites in iCEM optimizer are never reset

  #186 opened 2 months ago by nuria95
  0

• [Bug] `cartpole_continuous` errors when using `render_mode='human'`

  #185 opened a year ago by StephenWelch
  0

• Question regarding PETS & Discrete action spaces

  #184 opened a year ago by aralcimcim
  0

• Make `create_one_dim_tr_model` recognize subclasses of `BasicEnsemble`

  #182 opened a year ago by FrankTianTT
  0

• MBPO Implementation Example

  #181 opened a year ago by CAmpatzis
  0

• [Feature Request] Upload Dynamics Models to the HuggingFace Hub

  #169 opened a year ago by natolambert
  7

• [Bug] `max_episode_steps=1000` is hardcoded in `mujoco.py`

  #180 opened 2 years ago by Rohan138
  0

• [Bug] Centering, scaling and clamping the population in iCEM

  #172 opened 2 years ago by marbaga
  1

• [Bug] RuntimeError: CUDA error: device-side assert triggered

  #173 opened 2 years ago by melfm
  6

• [Feature Request] Output Normalization / Scaling

  #164 opened 2 years ago by natolambert
  4

• [Bug] What's the real_data_ratio to hopper and walker2d?

  #167 opened 2 years ago by chenxi-yang
  4

• [Bug] MBPO loss explode on humanoid and walker (sometimes)

  #137 opened 3 years ago by MansonX
  4

• hyperparameters optimization

  #166 opened 2 years ago by ss555
  1

• [Feature Request] Add option to use `functorch` for `BasicEnsemble`

  #162 opened 2 years ago by luisenp
  2

• [Bug] MBPO - Epoch loop missing steps_epoch reset

  #159 opened 2 years ago by matthiaskiller
  2

• [Feature Request] Support for Dreamers

  #143 opened 3 years ago by veds12
  10

• Using Wrapper Class for Custom GYM Env

  #146 opened 2 years ago by MishraIN
  7

• DQN agent for an environment with continuous actions

  #155 opened 2 years ago by MishraIN
  1

• [Bug] Training browser crashes on Windows

  #152 opened 2 years ago by MarcoKe
  0

• [Bug] Training with (n, 1) dimensional Box/MultiDiscrete action spaces throwing error

  #144 opened 2 years ago by JennoMai
  1

• [Feature Request] Add PyBullet implementations of envs

  #130 opened 2 years ago by dtch1997
  2

• MBPO cannot work on HumanoidTruncatedObsEnv and original Humanoid Env[Bug]

  #138 opened 3 years ago by jity16
  18

• [Bug] Update documentation for `ModelEnv` to reflect new sample/reset requirements

  #150 opened 2 years ago by luisenp
  0

• PETS Example Notebook Can't create Dynamics Model

  #149 opened 2 years ago by jpain3
  4

• [Bug] Algorithm tests fail in Pytorch 1.7.1

  #147 opened 2 years ago by AVSurfer123
  1

• [Bug] Pets CartPoleEnv preprocess_fn changes the dimensions of an observation.

  #139 opened 3 years ago by MarkSelden
  4

• [Bug] PETS not working

  #136 opened 3 years ago by sofan110
  18

• [Feature Request] Add batch support to ModelEnv.evaluate_action_sequences

  #132 opened 3 years ago by luisenp
  0

• [Feature Request] Add support for low dimensional observations to PlaNet

  #134 opened 3 years ago by luisenp
  0

• [Feature Request] Clean up the PlaNet visualizer

  #133 opened 3 years ago by luisenp
  0

• [Feature Request] Add batch support to CEM

  #131 opened 3 years ago by luisenp
  0

• [Feature Request] Implement PlaNet

  #85 opened 3 years ago by luisenp
  0

• [Feature Request] cartpole_continuous.py is not the standard continuous cartpole

  #117 opened 3 years ago by robertocalandra
  6

• Reward setup of the HalfCheetah environment

  #118 opened 3 years ago by jihwan-jeong
  2

• Difference in PETS implementation from the original TF version.

  #99 opened 3 years ago by ethanluoyc
  8

• Could you please provide the data corresponding to the figures of of experiments in the paper——PETS? Because we would like to like cite PETS as one of the state of the art work in our paper, but time is limited. Thank you![Feature Request]

  #114 opened 3 years ago by ChengpengLi1003
  1

• [Feature Request] More general reward

  #115 opened 3 years ago by mkolodziejczyk-piap
  1

• [Bug] Hopper-v2 crashing using MBPO

  #113 opened 3 years ago by sufodsia
  1

• [Bug] Normalization of model inputs causes PETS to crash on Pybullet Half Cheetah

  #110 opened 3 years ago by MarkSelden
  3

• [Feature Request] Logging of custom training metrics

  #108 opened 3 years ago by jan1854
  7

• [Bug] ModelTrainer.maybe_get_best_weights() does not deal properly with negative evaluation scores

  #102 opened 3 years ago by jan1854
  2

• [Question] Model Interface: forward arguments

  #101 opened 3 years ago by axelbr
  3

• [Bug] Issue with loading best weights in ModelTrainer

  #98 opened 3 years ago by ethanluoyc
  6

• [Bug] MBPO on HalfCheetah-v2 not learning

  #97 opened 3 years ago by sufodsia
  5

• [Bug] ReplayBuffer.sample_trajectory() returns transitions from two different trajectories

  #95 opened 3 years ago by jan1854
  2

• [Bug] `__len__` method of Model should always return non-negative integer

  #92 opened 3 years ago by jettjaniak
  0

• pets_example.ipynb problem

  #83 opened 3 years ago by app1ep1e
  7

• [Feature Request] (partially) known dynamics

  #81 opened 3 years ago by soldierofhell
  2

• [Bug] The generator in ModelEnv Class

  #78 opened 3 years ago by SongyiGao
  1

• Implement PILCO [Feature Request]

  #76 opened 3 years ago by alexanderimanicowenrivers
  1