Reports
Created by
Created On
Last edited
sgoodfriendppo-Microrts-selfplay-dc-phases-A10-S1-2023-05-26T00:09:17.258204
Reference self-play only training run used for RAISocketAI in IEEE-CoG microRTS 2023 competition
0
2024-01-29
Parallel Single A10 APPO async_rollout rl-algo-impls Benchmark 01-08-2024
Running 4 jobs on a single GPU
0
2024-01-11
unet microrts-ai trained against coacAI
Eval is WinLoss and training with progressive reward decay for non-WinLoss rewards
0
2023-04-04
2/17/2023 starpilot-hard Lambda Benchmark
Playing procgen's starpilot on hard distribution with varying IMPALA network sizes
0
2023-02-17
2/13/2023 procgen-easy Lambda Benchmark
Branch: https://github.com/sgoodfriend/rl-algo-impls/tree/vpg_hard_cap_steps_per_epoch
0
2023-02-14
2/14/2023 DQN IMPALA Atari Lambda Benchmark
Use IMPALA-style network for dqn playing Atari games
0
2023-02-17
2/9/2023 vpg_hard_cap_steps_per_epoch Lambda Benchmark
Branch: https://github.com/sgoodfriend/rl-algo-impls/tree/vpg_hard_cap_steps_per_epoch
0
2023-02-10
2/14/2023 IMPALA Atari+CarRacing Lambda Benchmark
Use IMPALA-style network for Atari games and CarRacing
0
2023-02-15
2/13/2023 vpg_atari_tweaks Atari Lambda Benchmark
Branch: https://github.com/sgoodfriend/rl-algo-impls/tree/vpg_atari_tweaks
0
2023-02-13
2/8/2023 VPG Lambda Benchmark
Branch: https://github.com/sgoodfriend/rl-algo-impls/tree/vpg_02_08_2023
0
2023-02-09
2/8/2023 DQN Lambda Labs Benchmark
Branch: https://github.com/sgoodfriend/rl-algo-impls/tree/dqn_benchmark
Commit: https://github.com/sgoodfriend/rl-algo-impls/commit/1d4094fbcc9082de7f53f4348dd4c7c354152907
0
2023-02-08
2/6/2023 main vs atari_separate_feature_extractor
Commit: https://github.com/sgoodfriend/rl-algo-impls/commit/5598ebc4b03054f16eebe76792486ba7bcacfc5c
0
2023-02-06
2/5/2023 update_rtg_between_epochs Lambda Labs Benchmark
Commit: https://github.com/sgoodfriend/rl-algo-impls/commit/5540e1fc804d609fac4498150d6af086639906e1
0
2023-02-06