sample-factory
acme
sample-factory | acme | |
---|---|---|
6 | 11 | |
740 | 3,381 | |
- | 0.5% | |
8.1 | 6.0 | |
about 2 months ago | about 15 hours ago | |
Python | Python | |
MIT License | Apache License 2.0 |
Stars - the number of stars that a project has on GitHub. Growth - month over month growth in stars.
Activity is a relative number indicating how actively a project is being developed. Recent commits have higher weight than older ones.
For example, an activity of 9.0 indicates that a project is amongst the top 10% of the most actively developed projects that we are tracking.
sample-factory
-
A minimal RL library for infinite horizon tasks
I take a lot of inspiration from Sample Factory and RLlib for my own RL library's implementation. Although I thoroughly enjoy both of these libraries, they just didn't quite fit right with my use case which motivated me to start my own. Hopefully someone finds use in rlstack whether it be through direct usage or as inspiration for their own personalized library
-
Fast and hackable frameworks for RL research
I'm tired of having my 200m frames of Atari take 5 days to run with dopamine, so I'm looking for another framework to use. I haven't been able to find one that's fast and hackable, preferably distributed or with vectorized environments. Anybody have suggestions? seed-rl seems promising but is archived (and in TF2). sample-factory seems super fast but to the best of my knowledge doesn't work with replay buffers. I've been trying to get acme working but documentation is sparse and many of the features are broken.
-
Multi-agent Decentralized Training with a PettingZoo environment
Hi, try sample-factory
-
How is IMPALA as a framework?
Sample Factory: https://github.com/alex-petrenko/sample-factory
-
The Myth of a Superhuman AI
Everything in this reply is wrong.
In AlphaZero for example, there were 44 million training games total for 700,00x0 steps of training for the full 9 hours.
Turning that human-like numbers, 44million games with on average 60 moves, at 1 second thinking time per move,
> 44000000*60/60/60/24/365 = 83,7138508371 years of training experience in 9 hours
The whole field of Reinforcement learning has agents training and playing games for many orders of magnitude more time than a human ever will. In-fact, we can scale this to over 100k of actions per second, in a single machine:
https://github.com/alex-petrenko/sample-factory
Then, there is also distributed Reinforcement Learning, where hundreds of agents can play at different machines and share experience, see AlphaZero, LeelaZero, R2D2 agent, R2D3 agent, Apex, Acer, Asynchronous PPO.
> but the data isn't useful without the context of experience
The experience is the data in Reinforcement Learning.
> and all processing power can do it overfit model without experience.
That is wrong, the agents perform what is called exploration to avoid getting stuck in simple strategies.
> Even if we put AI into an army of robots running around and experiencing things, there are still scaling limits to encoding and communicating knowledge and understanding.
True, but machines scale better because they speak the same language, or they can learn to tune their language to get their message across.
> Human organizations are a great example of the scaling limits of intelligence.
Human organization is a testament to how far we can get with something as limiting as the commonly used language. The language that we use to communicate is subject to misinterpretation due to our subjective experiences, this limitation is not shared by machines.
-
Best PyTorch RL library for doing research
I borrow a lot of performance tricks from sample factory, which is awesome but hard to modify from its original APPO algorithm. rlpyt was more modular, and I borrowed more ideas from it (namedarraytuple), but still too limited.
acme
-
Fast and hackable frameworks for RL research
I'm tired of having my 200m frames of Atari take 5 days to run with dopamine, so I'm looking for another framework to use. I haven't been able to find one that's fast and hackable, preferably distributed or with vectorized environments. Anybody have suggestions? seed-rl seems promising but is archived (and in TF2). sample-factory seems super fast but to the best of my knowledge doesn't work with replay buffers. I've been trying to get acme working but documentation is sparse and many of the features are broken.
-
How much of a MuJoCo simulation or real life robot can you train on a 3090?
I'm training a few algorithms from Deepmind's acme library on some MuJoCo models and I'm wondering how long this will take to train and what it's going to do to my electric bill. Is a 3090 or two enough to train something to keep its balance, or do a task, or do I need to wait for the 8090 to come out?
-
Recomendations of framework/library for MARL
Recently dm-acme also added support for multi-agent environments. Acme: https://github.com/deepmind/acme
- Have you used any good DRL library?
- Is there a way to get PPO controlled agents to move a little more gracefully?
-
Worthwhile to convert custom env to be dm_env compatible?
Can anyone speak to their experience using acme (https://github.com/deepmind/acme) and by extension dm_env (https://github.com/deepmind/dm_env)? I'm wondering if it would be worthwhile for me to invest the time into converting my custom environment (which loosely follows the standard RL setup) over to this format.
-
[D] Physics and Reinforcement Learning - Discussion of Deepmind's work
acme/acme/agents/tf/mpo at master · deepmind/acme · GitHub
- Applied resources in Pytorch?
-
deepmind acme compatible with windows?
after installing it in a clean env, I tried to run the example provided for solving the gym cartpole env: https://github.com/deepmind/acme/blob/master/examples/control/run_d4pg_gym.py
-
Spec for RL agent implementation?
Acme has a slightly different one: https://github.com/deepmind/acme which includes specs for agents, buffers etc. It is very general. You can see their component description here: https://github.com/deepmind/acme/blob/master/docs/components.md
What are some alternatives?
cleanrl - High-quality single file implementation of Deep Reinforcement Learning algorithms with research-friendly features (PPO, DQN, C51, DDPG, TD3, SAC, PPG)
dm_env - A Python interface for reinforcement learning environments
stable-baselines3 - PyTorch version of Stable Baselines, reliable implementations of reinforcement learning algorithms.
Mava - 🦁 A research-friendly codebase for fast experimentation of multi-agent reinforcement learning in JAX
tianshou - An elegant PyTorch deep reinforcement learning library.
dm_control - Google DeepMind's software stack for physics-based simulation and Reinforcement Learning environments, using MuJoCo.
machin - Reinforcement learning library(framework) designed for PyTorch, implements DQN, DDPG, A2C, PPO, SAC, MADDPG, A3C, APEX, IMPALA ...
MPO - Pytorch implementation of "Maximum a Posteriori Policy Optimization" with Retrace for Discrete gym environments
rl8 - A high throughput, end-to-end RL library for infinite horizon tasks.
tonic - Tonic RL library
rlpyt - Reinforcement Learning in PyTorch
selfhosted-apps-docker - Guide by Example