muzero-general
poly
muzero-general | poly | |
---|---|---|
14 | 24 | |
2,379 | 649 | |
- | 1.1% | |
0.0 | 8.2 | |
4 months ago | about 1 month ago | |
Python | Go | |
MIT License | MIT License |
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.
muzero-general
-
Open source rules engine for Magic: The Gathering
I went looking for MuZero implementations in order to see how, exactly, they interact with the game space. Based on this one, which had the most stars in the muzero topic, it appears that it needs to be able to discern legal next steps from the current game state https://github.com/werner-duvaud/muzero-general/blob/master/...
So, I guess for the cards Forge has implemented one could MuZero it, but I believe it's a bit chicken and egg with a "free text" game like M:TG -- in order to train one would need to know legal steps for any random game state, but in order to have legal steps one would need to be able to read and interpret English rules and card text
- I placed Stockfish (white) against ChatGPT (black). Here's how the game went.
- Ask HN: What interesting problems are you working on? ( 2022 Edition)
-
How to "fit" the output of the Critic to the dimension of the reward?
You may want to use the trick described in https://arxiv.org/pdf/1805.11593.pdf as a Transformed Bellman Operator. Its efficiency is proved in MuZero original paper https://arxiv.org/pdf/1911.08265.pdf Appendix F. The implementation of that method you can find here: https://github.com/werner-duvaud/muzero-general Usage: muzero/models.py:649 (def support_to_scalar)
-
MuZero unable to solve non-slippery FrozenLake environment?
I have used this implementation from MuZero: https://github.com/werner-duvaud/muzero-general
-
RL for chess
+1 to taking a look at OpenSpiel. It has AlphaZero in C++ and Python, and there is even a PR open that allows running UCI (e.g. Stockfish) bot. You can also load chess via the OpenSpiel wrapper in muzero-general: https://github.com/werner-duvaud/muzero-general
-
The future of MuZero, and where to go for news
When I looked up some community implementations, like that of Werner Duvaud on GitHub and Discord, hoping to make my own contributions to this effect, I soon found that I was hopelessly out of my depth as an amateur programmer, even with the help of some other sources like this walkthrough series. However, from what I could tell, most of the people working on this sort of thing seemed to be tackling relatively simple games. At first I thought this might be largely due to limitations in hobby time or computing power available to these users, but then I also noticed that, unless I have misunderstood something, it seems like the games are required to be rebuilt entirely in the engine of (this implementation of) MuZero, which would also obviously be a limit on the complexity of games chosen.
- Is MuZero currently the best RL algo that we have now?
-
"muzero-general", PyTorch/Ray code for Gym/Atari/board-games (reasonable results + checkpoints for small tasks)
Windows support (Experimental / Workaround: Use the notebook in Google Colab)
-
Muzero code implementation
There are several if you google "muzero github", e.g. https://github.com/werner-duvaud/muzero-general
poly
- Looking for an Open Source project to participate in for Google Summer of Code
-
GitHub Accelerator: our first cohort and what's next
- https://github.com/TimothyStiles/poly: Poly is a fast, well tested Go package for engineering organisms.
-
These 20 startups are in 1st ever batch of GitHub OS Accelerator
Poly: Fast Go package for engineering organisms
-
Ask HN: Burnt out from big tech. What's next?
You might want to look at computational biology. Jim Allison won the Nobel Prize back in 2018 for his work on immunotherapy for cancer and there's a lot of basic research work to be done to perfect this approach. Epigenetic clocks are really interesting too (see Steve Horvath's work). Also, there's synthetic biology, where you could, for example, explore this package that's written in Go: https://github.com/TimothyStiles/poly
- Any corner cases for Needleman-Wunsch that should be tested?
- Where can I find well-written go code to learn from?
-
High-performance language recommendation
Check out poly. It’s written in go and I’m using it for one of my projects too. The goal is that we should have high performance libraries that we can use knowing what people are working on the forks will give the community a leg up.
-
How is GO used in bioinfo?
The most popular bioinformatic package I've seen in go is poly.
-
Software engineers: consider working on genomics
I write synthetic biology software for a living and maintain this open source, Go package for engineering DNA that has high test coverage and a nice little dev community around it.
https://github.com/TimothyStiles/poly
A large part of my project's community are devs that want to get into the field but can't tolerate the ridiculously low pay, laughably bad management, disrespect, and what amounts to 40+ years of technical debt that's endemic to biotech software.
I've had companies here in the Bay Area offer me 100K a year with a straight face. I've had companies during interview tell me they're looking for someone to help, "set up GitHub". I've seen job listings for low paid web dev positions require applicants to have PhDs.
The reality is that except for a growing handful of places management straight up won't know the difference between IT and software engineers. It's what I call the naive buyers problem.
The demand for software engineers in biotech is generated by naive buyers that don't know what they need, why they need it, or how to get it.
Benchling and Recursion Pharmaceuticals have reputations in the industry of paying, "standard software salaries". So do the research divisions at places like deepmind/microsoft/google but in my experience there's even new multi-billion dollar institutes where senior management has never even heard the term devops.
Most places advertise for "data scientist", positions or some analog, instead of software engineers. This is mostly because upper management has never met an actual practicing software engineer in a professional setting. Many come from academia where the culture and work requirements heavily disincentivize standard software engineering practices.
It's also not uncommon for a biotech company to either have a very under qualified CTO whose main programming experience is what they learned doing ML research like stuff during their PhD or not even have one at all which has huge downstream consequences.
This week a software engineer trying to make the switch to biotech actually DM'd me to ask why they were seeing a ton of data science / ML job positions but no software engineering / devops positions.
They were worried that these companies were trying to save on costs by forcing their data scientists to create infrastructure but it's actually worse than that. Most of these companies aren't even aware that there's supposed to be infrastructure.
Despite all of this the future is looking better and I'm starting to find new companies and positions that are well... reasonable. I learned about this thread from a friend at a party last night that works at one of these companies. There's a small, strong new wave of companies and developers out there pushing biotech software forward. Hopefully some (including myself) make it big while pushing the idea that better tech equals better biotech.
-
Ask HN: What interesting problems are you working on? ( 2022 Edition)
It is more like the X Y Z W. However, the X Y Z W bits I am working on as well (https://github.com/TimothyStiles/poly , https://github.com/TimothyStiles/allbase , trilo.bio, freegenes.org). Going for fully automated "make bacterium X produce molecule Y", but still a while away (but surprisingly not THAT far off)
What are some alternatives?
deep-RL-trading - playing idealized trading games with deep reinforcement learning
Raylib-CsLo - autogen bindings to Raylib 4.x and convenience wrappers on top. Requires use of `unsafe`
Super-mario-bros-PPO-pytorch - Proximal Policy Optimization (PPO) algorithm for Super Mario Bros
pg-mem - An in memory postgres DB instance for your unit tests
alpha-zero-general - A clean implementation based on AlphaZero for any game in any framework + tutorial + Othello/Gobang/TicTacToe/Connect4 and more
linaria - Zero-runtime CSS in JS library
open_spiel - OpenSpiel is a collection of environments and algorithms for research in general reinforcement learning and search/planning in games.
seq - A high-performance, Pythonic language for bioinformatics
stable-baselines3-contrib - Contrib package for Stable-Baselines3 - Experimental reinforcement learning (RL) code
m4b-tool - m4b-tool is a command line utility to merge, split and chapterize audiobook files such as mp3, ogg, flac, m4a or m4b
pytorch-ddpg - Deep deterministic policy gradient (DDPG) in PyTorch 🚀
procedural-gl-js - Mobile-first 3D mapping engine with emphasis on user experience