agents
agents copied to clipboard
tensorflow
TF-Agents: A reliable, scalable and easy to use TensorFlow library for Contextual Bandits and Reinforcement Learning.
It seems like replacing `utils.mlp_layers` with `EncodingNetwork` in `CriticRnnNetwork` will allow the use of preprocessing combiners for actions and observations. This critic network is used for SAC agents. The `ActorDistributionRnnNetwork`...
Hi, could you please clarify the description of the parm max_sequence_length in ReverbAddEpisodeObserver. The description is a little bit messy. max_sequence_length: An integer. `max_sequence_length` used to write to the replay...
If I enable tensor numeric checking and initialise a Learner with a PolicySavedModelTrigger I get the following error: > TypeError: Input 'resource' of 'AssignVariableOp' Op has type float32 that does...
I understand from the [Per-Arm Features tutorial](https://www.tensorflow.org/agents/tutorials/per_arm_bandits_tutorial) that it may be "cumbersome to add" a new action to a policy, but what is the procedure for doing do? For example,...
I'm trying to implement a PPO agent to play with LunarLander-v2 with tf_agents library like it was in [this tutorial](https://pylessons.com/LunarLander-v2-PPO/) ([_github repo_](https://github.com/pythonlessons/Reinforcement_Learning/tree/master/LunarLander-v2_PPO)) networks.py ``` from tf_agents.networks import actor_distribution_network, value_network from...
When you get the chance, can you please take a look at this question on stackoverflow [1]? I tried to find a community discussion forum for tf agent but was...
I am getting this error while trying to create a custom environment and feed it to the DQN agent. **ValueError: Only scalar actions are supported now, but action spec is:...
I used [save](https://github.com/tensorflow/agents/blob/c584e7642fbb45df79ab78dc8884248871e14f3e/tf_agents/utils/common.py#L1026) in [tf_agents.utils.common.Checkpointer ](https://github.com/tensorflow/agents/blob/c584e7642fbb45df79ab78dc8884248871e14f3e/tf_agents/utils/common.py#L981-L1032) to save the checkpoint file. I would like to see the information (weight values, etc.) of the contents of the file, how can I...
- OS Platform and Distribution: Google Colab - TensorFlow installed from: Binary - TensorFlow version: 2.6.0 - TF Agents version: 0.9.0 - Python version: Python 3.7.12 - Installed using virtualenv?...
`PPOAgent` does not use regularization losses that are defined in the `value_net` and `actor_net`. It should add them to the other losses.
Metadata
Owner
Metadata
TF-Agents: A reliable, scalable and easy to use TensorFlow library for Contextual Bandits and Reinforcement Learning.