acme

acme copied to clipboard

New Release

3

Hi 👋, Really liking Acme and all the work you have all put into it and contributing to the open source community. Saw your recent [tweet](https://twitter.com/DeepMind/status/1573264436411908098?s=20&t=Uwg5xCHFG59IMIKbGTJLRg) and would like to...

cemlyn007

# Problem Using an environment which uses a `DiscreteArray` as action's spec (e.g. gym's `CartPole`) with the `CanonicalSpecWrapper` (which, to the best of my knowledge, should be used with continuous...

Whojo

Converting Tensorflow Dataset to `iterator` does not sync well with client

1

Hi, I accidentally stumbled upon a problem within the [tutorial notebook](https://colab.research.google.com/github/deepmind/acme/blob/master/examples/tutorial.ipynb) when playing around with the acme and reverb API that causes a weird synchronization behaviour between sampling from the...

joeryjoery

Hi Acme team, I think JAX DQN might set the evaluation epsilon to the exploration epsilon if deterministic evaluation is requested (eps=0.0, [here](https://github.com/deepmind/acme/blob/860dbab686042573569b84223d8da6d43d09c304/acme/agents/jax/dqn/builder.py#L183)). Replacing this with `self._config.eval_epsilon is not None`...

tseyde

Hi, I'm trying to run AndroidEnv script Run Acme agent - https://github.com/deepmind/android_env/blob/main/examples/run_acme_agent.py When I try to run this, after installing acme, I get issue with sonnet which is trying to...

suhani247

MuZero agent in examples folder not present

1

I see the examples have [MuZero for discrete control tasks](https://github.com/deepmind/acme/blob/31528f87711c1c94b3d99b5a21f347424759e29e/examples/baselines/rl_discrete/run_muzero.py). But this doesn't seem to have been released yet. Any idea on when it will be?

wcarvalho

Add an option to allow for different frame_stacking behavior at the start of an episode. Some original implementations of RL agents uses a different frame-stacking behavior at the start of...

ethanluoyc

run_mogmpo.py -- ValueError: Graph parent item 0 is not a Tensor;

2

Hi there, Packages were installed according to: ``` conda create -n acme python=3.8 conda activate acme pip install --upgrade pip setuptools wheel pip install .[jax,tf,testing,envs] pip install mujoco pip install...

dyth

When using `samples_per_insert=None`, this [line](https://github.com/deepmind/acme/blob/26542510d8eff0cc05f31cc7dc66deee33040720/acme/jax/experiments/run_experiment.py#L277) returns the following error: ``` File "/home/kaleab/anaconda3/envs/dyn-net/lib/python3.9/site-packages/acme/jax/experiments/run_experiment.py", line 75, in run_experiment replay_tables, rate_limiters_max_diff = _disable_insert_blocking(replay_tables) File "/home/kaleab/anaconda3/envs/dyn-net/lib/python3.9/site-packages/acme/jax/experiments/run_experiment.py", line 296, in _disable_insert_blocking max(1, int((rate_limiter_info.max_diff - rate_limiter_info.min_diff)...

KaleabTessera

Question: Performance of multi-device prefetching

2

Hello, I have a quick question with regards to the implementation of the multi-device put and prefetching iterator. From looking at the code, it seems that there is quite a...

EdanToledo