Ecosyste.ms: Issues

An open API service for providing issue and pull request metadata for open source projects.

GitHub / PartnershipOnAI/safelife issues and pull requests

#31 - Benchmark random levels end early

Issue - State: open - Opened by JohnBurden about 2 years ago - 2 comments

#30 - Understanding the SafeLife navigation task

Issue - State: open - Opened by aseembits93 over 3 years ago - 1 comment

#29 - Fix wandb logging failure reported

Pull Request - State: open - Opened by Tshimanga over 3 years ago

#28 - wandb logging seemingly broken

Issue - State: open - Opened by Tshimanga over 3 years ago - 3 comments

#27 - bugfix: only use py_limited_api when running bdist_wheel

Pull Request - State: closed - Opened by clwainwright over 3 years ago - 1 comment

#26 - The main branch of safelife requires atleast gcc-10 for compilation

Issue - State: closed - Opened by aseembits93 over 3 years ago - 2 comments

#25 - LSTM agents & more

Pull Request - State: open - Opened by pde almost 4 years ago - 1 comment

#24 - v1.2 release

Pull Request - State: closed - Opened by clwainwright about 4 years ago

#23 - Clean shutdown can fail if an interrupt occurs early

Issue - State: closed - Opened by pde about 4 years ago

#22 - Refactor start-training.py into functions

Pull Request - State: closed - Opened by pde about 4 years ago - 2 comments

#21 - Interactive rendered state is not always the current board state

Issue - State: closed - Opened by pde about 4 years ago - 1 comment

#20 - --algo mppo : PPO with an LSTM for agent state

Pull Request - State: closed - Opened by pde about 4 years ago - 1 comment

#19 - --algorithm mppo: PPO with an LSTM for state

Pull Request - State: closed - Opened by pde about 4 years ago

#18 - Some samples (other than Google cloud) running safelife

Issue - State: open - Opened by EvaluationResearch over 4 years ago - 1 comment

#17 - Multiagent improvements: network depth, wandb options, safer launch code

Pull Request - State: closed - Opened by pde over 4 years ago - 1 comment

#16 - Multiagent pde

Pull Request - State: closed - Opened by pde over 4 years ago

#15 - Unified single/multi-agent algorithms

Pull Request - State: closed - Opened by clwainwright over 4 years ago

#14 - Do you have any plans to add multiple agents in the next step？

Issue - State: closed - Opened by EvaluationResearch over 4 years ago - 2 comments

#13 - Possible reproducibility bug

Issue - State: closed - Opened by pde over 4 years ago - 1 comment

#12 - get rid of variable-length arrays for MS Visual Studio compatibility

Pull Request - State: closed - Opened by clwainwright over 4 years ago - 1 comment

#10 - V1.1.1

Pull Request - State: closed - Opened by clwainwright over 4 years ago - 2 comments

#9 - Base curriculum progression on 3rd best of 10 past relevant episodes

Pull Request - State: open - Opened by pde over 4 years ago

#8 - Add curriculum support to the multiagent branch

Pull Request - State: closed - Opened by pde over 4 years ago

#7 - safelife v1.1

Pull Request - State: closed - Opened by clwainwright over 4 years ago

#6 - Bugfix memoryleak gen pattern

Pull Request - State: closed - Opened by clwainwright almost 5 years ago

#5 - Memory leak in speedups.gen_pattern

Issue - State: closed - Opened by clwainwright almost 5 years ago

#4 - Should SafeLife environments be pickleable?

Issue - State: closed - Opened by pde almost 5 years ago - 3 comments

#3 - Interactive graphics

Pull Request - State: closed - Opened by clwainwright over 5 years ago

#2 - Switch to a reproducible RNG

Issue - State: closed - Opened by pde over 5 years ago - 1 comment

#1 - clean up benchmarks code

Pull Request - State: closed - Opened by clwainwright almost 6 years ago