agents
agents copied to clipboard
TF-Agents: A reliable, scalable and easy to use TensorFlow library for Contextual Bandits and Reinforcement Learning.
The error happens when trying to save a model with tensorflow-agents. For example: ``` collect_policy = tf_agent.collect_policy saver = PolicySaver(collect_policy, batch_size=None) ``` and saving the model with: `saver.save('models/ddpg/')` This works...
Rename `tf_agents.policies.policy_saver.PolicySaver` to `tf_agents.policies.PolicySaver`
I've been trying to apply the latest ppo example from: https://github.com/tensorflow/agents/tree/master/tf_agents/experimental/examples/ppo/schulman17 From my understanding of Schulman 2017 the ppo agent is supposed to support multiple parallel environments and batched trajectories....
I am trying to train a Reinforcement Learning agent using TF-Agent [TF-Agent DQN Tutorial](https://www.tensorflow.org/agents/tutorials/1_dqn_tutorial). In my application, I have 9 discrete actions (labeled 0 to 8), and I would like...
hi, im trying to create a DqnAgent agent with a mask for valid/invalid actions, according to [this post][1] , i should specify a ```splitter_fn``` for the ```observation_and_action_constraint_splitter``` arg. According to...
Closes #694
In 7_SAC_minitaur_tutorial.ipynb: rb_observer = reverb_utils.ReverbAddTrajectoryObserver( reverb_replay.py_client, table_name, sequence_length=2, stride_length=1) TypeError Traceback (most recent call last) [](https://localhost:8080/#) in () 2 reverb_replay.py_client, 3 table_name, ----> 4 sequence_length=2) [/usr/local/lib/python3.7/dist-packages/tf_agents/replay_buffers/reverb_utils.py](https://localhost:8080/#) in __init__(self, py_client, table_name,...
I am using the tf_agents for contextual bandit algorithm. My data is at userlevel and hence very important to make sure that actions and rewards (and trajectories in order to...