agents
agents copied to clipboard
TF-Agents: A reliable, scalable and easy to use TensorFlow library for Contextual Bandits and Reinforcement Learning.
Hello, I tried to understand how to use an action constraint on my Environment. I have an environment with 100 lines, and the agent can choose any line, so I...
This is similar to #656. But I am making another issue since that issue is still not resolved. Also, @sguada mentioned in #702 that PPO agent can take 1-D action...
I'm using `PPOAgent` and `ActorDistributionNetwork` with the following action_spec: ```python action_spec = array_spec.BoundedArraySpec( shape=(10,), dtype=np.int32, minimum=0, maximum=4, name='action') ``` However, I received the following error when the agent was trying...
Following the discussion from #37, I developed a `MultiCategoricalProjectionNetwork` that splits logits and creates the respective `Categorical` distribution. I tried to adhere to the same pattern as far as I...
TF-Agent Version: 0.12.0 TF Version: 2.8.0 Python Version:3.7 So I'm trying to create a Q-Learning network for to train a computer to learn how to solve a rubrix cube. However,...
Hi! I've followed some advice of people saying that to make a prediction from a loaded tf_agent I have to create a manual tilmestep. I successfully loaded the agent, but...
Hi! I know the issue have been raised many times but all the solutions proposed didn't work for me, and sometimes the threads didn't get to the end of it....
Hello everybody, I have been trying to create a custom DQN for a project but couldn't find any specification on networks with multiple inputs and multiple actions. So far I...
InvalidArgumentError: Exception encountered when calling layer "dynamic_unroll" (type DynamicUnroll)
I am trying to run a QRNN Network with Neural LinUCB Agent. My parameters are as follows: ``` AGENT_ALPHA = 1 GAMMA = 1 EPS_GREEDY = 0 LAYERS = (512,256,128)...