acme
acme copied to clipboard
A library of reinforcement learning components and agents
Following the most recent commit, the jax.tree module is now being used instead of jax.tree_*. This is change requires jax >= 0.4.25. However, the are still many parts of the...
Hi, I am currently working on using ACME, but when I install ACME, I always get issues with Mujoco. I do not need Mujoco at all in my work, as...
The following is the terminal output when I run `python run_mpo.py` ```shell (acme) root@autodl-container-fa5b45ae8c-68d88031:~/acme/examples/baselines/rl_continuous# python run_mpo.py /root/miniconda3/envs/acme/lib/python3.8/site-packages/tensorflow_probability/python/__init__.py:57: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead. if (distutils.version.LooseVersion(tf.__version__) < I0606...
In `acme.agents.tf.mcts.types`, Observation is hinted to be a numpy ndarray ```python # Assumption: observations are array-like. Observation = np.ndarray ``` based on this, in `acme.agents.tf.mcts.acting`, the `MCTSActor._forward()` method hard-codes a...
This PR improves the README.md to enhance onboarding, readability, and developer experience. The following changes have been made: ✍️ Added badges for Python version, PyPI, tests, and docs 📚 Reorganized...
I've installed Acme on my system and am trying to run code from the [quickstart tutorial](https://colab.research.google.com/github/deepmind/acme/blob/master/examples/quickstart.ipynb). However, I encountered an issue when importing `experiments` from `acme.jax`. Here is the snippet...
Hi, I would like to add the Implicit Q-Learning algorithm to the Acme framework. https://arxiv.org/abs/2110.06169