tensorflow/agents issues and pull requests

#944 - Incompatibility between PPO collect policy and SquashToSpecNormal Distribution

Issue - State: open - Opened by CHEyy-85 22 days ago

#943 - Fix comment typo in Movielens environment

Pull Request - State: closed - Opened by tacitvenom 2 months ago - 3 comments

#942 - Unexpected output from `actor_network`. Expected `Distribution` objects, but saw output spec: TensorSpec(shape=(4,), dtype=tf.int32, name=None)

Issue - State: open - Opened by mhdtslm 3 months ago

#940 - GaussianPolicy applies the same noise to all actions in the batch

Issue - State: open - Opened by zhezherun 3 months ago

#939 - to_n_step_transition returns wrong results if episode was truncated by the time limit wrapper and N > 1

Issue - State: open - Opened by zhezherun 3 months ago

#938 - Updating outdated documentation

Issue - State: open - Opened by Nour-Aldein2 3 months ago

#937 - ImportError: Attempted to load a reverb dynamic library, but it could not find the required symbols inside of TensorFlow.

Issue - State: open - Opened by ace510 4 months ago - 1 comment

#937 - ImportError: Attempted to load a reverb dynamic library, but it could not find the required symbols inside of TensorFlow.

Issue - State: open - Opened by ace510 4 months ago - 2 comments

#937 - ImportError: Attempted to load a reverb dynamic library, but it could not find the required symbols inside of TensorFlow.

Issue - State: open - Opened by ace510 4 months ago - 2 comments

#936 - Error when running DynamicStepDriver with tf_agents

Issue - State: closed - Opened by Jack-Rais 4 months ago - 1 comment

#934 - ValueError: Network only supports action_specs with shape in [(), (1,)])

Issue - State: closed - Opened by Mr-Elysium 4 months ago

#934 - ValueError: Network only supports action_specs with shape in [(), (1,)])

Issue - State: closed - Opened by Mr-Elysium 4 months ago

#933 - tf-agents ubuntu 24.04 venv pip3 install error on pygame 2.1.3

Issue - State: open - Opened by johnny-littlepunch 4 months ago - 2 comments

#932 - Module not found error when working through 1_dqn_tutorial.ipynb

Issue - State: closed - Opened by ace510 4 months ago - 2 comments

#929 - Type error in PolicySaver.save()

Issue - State: open - Opened by anmol438 5 months ago - 1 comment

#928 - Compatibility with gymnasium environments

Issue - State: open - Opened by anmol438 5 months ago

#927 - PolicySavedModelTrigger has no sweeping ability.

Issue - State: open - Opened by brianorbrain 7 months ago

#926 - WARNING:tensorflow:Value in checkpoint could not be found in the restored object: (root).agent._optimizer._variables.1

Issue - State: open - Opened by huiyujie 7 months ago

#922 - PPOAgent + MaskSplitterNetwork normalizes Mask when observation normalization is turned on.

Issue - State: open - Opened by BaLinuss 8 months ago

#921 - Model plot function using pyplot

Pull Request - State: open - Opened by hotamago 9 months ago - 1 comment

#919 - got an unexpected keyword argument 'kwargs' when create_variables and unable to copy

Issue - State: closed - Opened by hotamago 9 months ago

#917 - Multi arm Multi play Setting

Issue - State: open - Opened by Akshaysharma29 10 months ago

#916 - No module named 'tensorflow.python.training.tracking

Issue - State: open - Opened by Afshin818 10 months ago - 2 comments

#914 - Inference MAB

Issue - State: open - Opened by Akshaysharma29 10 months ago - 1 comment

#913 - Add seed, get_state, and set_state to BatchedPyEnvironment.

Pull Request - State: closed - Opened by b-marks 10 months ago

#911 - [Fix this ASAP] TypeError : 'AtariEnv.render() takes 1 positional argument but 2 were given' and some environments can't be rendered

Issue - State: closed - Opened by Skyisblue324 11 months ago - 1 comment

#910 - How to convert keras Model to DDQN QNetwork？

Issue - State: open - Opened by linfanzz 11 months ago

#909 - Problem with AverageReturnMetric

Issue - State: open - Opened by billh0420 11 months ago - 1 comment

#908 - Fix typo in doc string

Pull Request - State: closed - Opened by j-svensmark 11 months ago - 1 comment

#907 - Feature Request: expose a target_update method for DQNAgent

Issue - State: open - Opened by billh0420 11 months ago

#906 - DQN Tutorial.ipynb not working

Issue - State: closed - Opened by tarobins 11 months ago - 2 comments

#905 - Parallelization of this "CartPole v0" standard environment failed

Issue - State: open - Opened by fengyuhun 11 months ago

#904 - How to run Code on Windows

Issue - State: open - Opened by MarleneBs 11 months ago

#903 - Per-Arm Features guide/tutorial not in site table of contents

Issue - State: closed - Opened by tottenjordan 11 months ago - 1 comment

#902 - remove tensorflow warning around tf.function on per_field_where

Issue - State: open - Opened by cmarlin 12 months ago

#901 - Add Farama's Gymnasium and D3QN agent for LunarLander-v2

Pull Request - State: closed - Opened by porta-logica 12 months ago

#900 - Farama's Gymnasium for R0.18.0 and D3QN agent for LunarLander-v2

Pull Request - State: closed - Opened by porta-logica 12 months ago

#899 - Fix test assertion in wrappers_test.py

Pull Request - State: closed - Opened by slowder 12 months ago

#898 - Resolve issue #829

Pull Request - State: closed - Opened by porta-logica almost 1 year ago - 3 comments

#898 - Resolve issue #829

Pull Request - State: open - Opened by porta-logica almost 1 year ago - 2 comments

#898 - Resolve issue #829

Pull Request - State: open - Opened by porta-logica almost 1 year ago - 2 comments

#897 - Adding episodic_replay_buffer to init.py file

Pull Request - State: closed - Opened by cmarlin almost 1 year ago

#896 - Q-Network wrong output spec

Issue - State: open - Opened by rissois about 1 year ago - 2 comments

#895 - Add `gym_kwargs` option to `suite_atari`

Pull Request - State: closed - Opened by msmith93 about 1 year ago - 1 comment

#894 - Add support for gym_kwargs in suite_atari

Issue - State: closed - Opened by msmith93 about 1 year ago - 1 comment

#893 - test do not merge

Pull Request - State: closed - Opened by PoloMarco1 about 1 year ago - 1 comment

#893 - test do not merge

Pull Request - State: closed - Opened by PoloMarco1 about 1 year ago - 1 comment

#892 - New to RL Agents, need help with policies

Issue - State: open - Opened by Siekrofon about 1 year ago - 1 comment

#892 - New to RL Agents, need help with policies

Issue - State: open - Opened by Siekrofon about 1 year ago

#892 - New to RL Agents, need help with policies

Issue - State: open - Opened by Siekrofon about 1 year ago

#891 - Errors when saving ranking policies

Issue - State: open - Opened by tottenjordan about 1 year ago - 1 comment

#889 - TypeError: configurable() got an unexpected keyword argument 'blacklist'

Issue - State: closed - Opened by ParamB11 about 1 year ago - 1 comment

#888 - Understanding parallel environments

Issue - State: open - Opened by b-fg about 1 year ago

#884 - Emitting internal network state

Issue - State: open - Opened by nathanmartz about 1 year ago

#884 - Best way to emit/output internal network state (for debugging)

Issue - State: open - Opened by nathanmartz about 1 year ago

#883 - Learning rate metric inconsistent depending on optimizer

Issue - State: open - Opened by b-fg about 1 year ago

#883 - Learning rate metric inconsistent depending on optimizer

Issue - State: open - Opened by b-fg about 1 year ago

#882 - `merge_call` called while defining a new graph or a tf.function.

Issue - State: open - Opened by Jark5455 about 1 year ago

#854 - Shape mismatch in DDPG agent's critic loss function? Shape mismatch between td_targets and q_values tensors

Issue - State: closed - Opened by Pren7z over 1 year ago - 1 comment

#854 - Shape mismatch in DDPG agent's critic loss function? Shape mismatch between td_targets and q_values tensors

Issue - State: closed - Opened by Pren7z over 1 year ago - 1 comment

#853 - ReverbAddTrajectoryObserver gives "The number of pending items is alarmingly high" error

Issue - State: open - Opened by adiaconu11 over 1 year ago

#853 - ReverbAddTrajectoryObserver gives "The number of pending items is alarmingly high" error

Issue - State: open - Opened by adiaconu11 over 1 year ago

#853 - ReverbAddTrajectoryObserver gives "The number of pending items is alarmingly high" error

Issue - State: open - Opened by adiaconu11 over 1 year ago

#852 - Bug in tf_agents.bandits.policies.linalg.conjugate_gradient?

Issue - State: open - Opened by td20002 over 1 year ago

#852 - Bug in tf_agents.bandits.policies.linalg.conjugate_gradient?

Issue - State: open - Opened by td20002 over 1 year ago

#852 - Bug in tf_agents.bandits.policies.linalg.conjugate_gradient?

Issue - State: open - Opened by td20002 over 1 year ago

#851 - Added auto-reset handling to PyEnv Wrappers

Pull Request - State: closed - Opened by sehejjain over 1 year ago - 1 comment

#851 - Added auto-reset handling to PyEnv Wrappers

Pull Request - State: open - Opened by sehejjain over 1 year ago

#851 - Added auto-reset handling to PyEnv Wrappers

Pull Request - State: open - Opened by sehejjain over 1 year ago

#850 - Updated Documentation

Pull Request - State: open - Opened by Bhaney44 over 1 year ago - 1 comment

#850 - Updated Documentation

Pull Request - State: open - Opened by Bhaney44 over 1 year ago - 1 comment

#850 - Updated Documentation

Pull Request - State: closed - Opened by Bhaney44 over 1 year ago - 3 comments

#849 - ModuleNotFoundError: No module named 'tf_agents.drivers'; 'tf_agents' is not a package

Issue - State: closed - Opened by RayTsui over 1 year ago

#849 - ModuleNotFoundError: No module named 'tf_agents.drivers'; 'tf_agents' is not a package

Issue - State: closed - Opened by RayTsui over 1 year ago

#849 - ModuleNotFoundError: No module named 'tf_agents.drivers'; 'tf_agents' is not a package

Issue - State: closed - Opened by RayTsui over 1 year ago

#848 - Possible Bugs in CQL_SAC Example

Issue - State: open - Opened by ChengDaHaI over 1 year ago

#848 - Possible Bugs in CQL_SAC Example

Issue - State: open - Opened by ChengDaHaI over 1 year ago

#848 - Possible Bugs in CQL_SAC Example

Issue - State: open - Opened by ChengDaHaI over 1 year ago

#847 - Actor network predicts actions over bounds using PPOClipAgent

Issue - State: open - Opened by b-fg over 1 year ago - 1 comment

#847 - Actor network predicts actions over bounds using PPOClipAgent

Issue - State: open - Opened by b-fg over 1 year ago - 2 comments

#847 - Actor network predicts actions over bounds using PPOClipAgent

Issue - State: open - Opened by b-fg over 1 year ago - 1 comment

#846 - "Too many values to unpack error" when using DQN mnih example

Issue - State: closed - Opened by rajfly over 1 year ago

#846 - "Too many values to unpack error" when using DQN mnih example

Issue - State: closed - Opened by rajfly over 1 year ago - 1 comment

#846 - "Too many values to unpack error" when using DQN mnih example

Issue - State: closed - Opened by rajfly over 1 year ago

#845 - Unable to load multiprocessing context for my validated custom gym environment.

Issue - State: open - Opened by Yash271100 over 1 year ago - 1 comment

#845 - Unable to load multiprocessing context for my validated custom gym environment.

Issue - State: open - Opened by Yash271100 over 1 year ago - 1 comment

#844 - Error in Parallel environment processing BrokenPipeError:[WinError 109]

Issue - State: open - Opened by roeslib over 1 year ago - 2 comments

#843 - tf-agents 0.16.0 doesn't install cleanly on macosx (13.2.1 (22D68))

Issue - State: open - Opened by apurva-sharma over 1 year ago

#842 - Shape issue with add_batch() method while training a DQN Agent

Issue - State: open - Opened by PierreSmague over 1 year ago

#841 - Actor-PPOLearner with discrete Action on Windows

Issue - State: open - Opened by bennyfri over 1 year ago

#838 - ValueError when validating a gym environment with MultiDiscrete action space

Issue - State: open - Opened by heydarshahi over 1 year ago

#837 - Help with understanding py_environment.PyEnvironment wrapping potential

Issue - State: closed - Opened by jm-willy over 1 year ago - 1 comment

#836 - PPOClipAgent not working with suite_gym acrobot agent

Issue - State: open - Opened by brian1985 over 1 year ago

#835 - Loading Atari environments causes errors in render

Issue - State: open - Opened by neramas1221 over 1 year ago

#834 - max_episode_steps in suite_gym.load doesn't work

Issue - State: open - Opened by fede72bari over 1 year ago

#833 - Need help and willing to pay $$ to solve this problem - Please name your price!

Issue - State: closed - Opened by Robertfu89 over 1 year ago - 2 comments

#832 - I used " return time_step.restart()" and got an error message "AttributeError:'TimeStep' object has no attribute 'restart'"

Issue - State: open - Opened by Robertfu89 over 1 year ago

#831 - suite_gym.load('CartPole-v1', max_episode_steps=10000) does not modify the default max_episode_steps

Issue - State: open - Opened by fede72bari over 1 year ago

#830 - [feature request] Add changelog or release notes

Issue - State: open - Opened by garar over 1 year ago

#829 - Gym Wrapper: Add support for gymnasium spaces

Issue - State: open - Opened by niklaskar over 1 year ago

GitHub / tensorflow/agents issues and pull requests