Comment
charts/episode_reward
charts/episode_reward
exp_name: ppo_diverse, seed: 2
exp_name: ppo_diverse, seed: 3
exp_name: ppo_diverse, seed: 4
exp_name: ppo_diverse, seed: 1
charts/episode_reward/WinLossRewardFunction
charts/episode_reward/WinLossRewardFunction
exp_name: ppo_diverse, seed: 2
exp_name: ppo_diverse, seed: 3
exp_name: ppo_diverse, seed: 4
exp_name: ppo_diverse, seed: 1
Run set
4
Name
4 visualized
exp_name: ppo_gridnet_diverse_encode_decode
exp_name: ppo_gridnet_diverse_encode_decode
4
4
exp_name: ppo_gridnet_coacai_naive
exp_name: ppo_gridnet_coacai_naive
4
4
exp_name: ppo_gridnet_coacai_no_mask
exp_name: ppo_gridnet_coacai_no_mask
4
4
exp_name: ppo_coacai_naive
exp_name: ppo_coacai_naive
4
4
exp_name: ppo_gridnet_coacai_partial_mask
exp_name: ppo_gridnet_coacai_partial_mask
4
4
exp_name: ppo_coacai_partial_mask
exp_name: ppo_coacai_partial_mask
4
4
exp_name: ppo_gridnet_selfplay_diverse_encode_decode
exp_name: ppo_gridnet_selfplay_diverse_encode_decode
4
4
exp_name: ppo_gridnet_selfplay_encode_decode
exp_name: ppo_gridnet_selfplay_encode_decode
4
4
exp_name: ppo_gridnet_coacai
exp_name: ppo_gridnet_coacai
4
4
exp_name: ppo_gridnet_diverse_impala
exp_name: ppo_gridnet_diverse_impala
4
4
exp_name: ppo_diverse_impala
exp_name: ppo_diverse_impala
4
4
exp_name: ppo_coacai
exp_name: ppo_coacai
4
4
exp_name: ppo_coacai_no_mask
exp_name: ppo_coacai_no_mask
4
4
exp_name: ppo_gridnet_naive
exp_name: ppo_gridnet_naive
4
4
exp_name: ppo_gridnet_diverse
exp_name: ppo_gridnet_diverse
4
4
exp_name: ppo_diverse
exp_name: ppo_diverse
4
4
State
Notes
User
Tags
Created
Runtime
Sweep
anneal_lr
batch_size
capture_video
clip_coef
clip_vloss
cuda
ent_coef
exp_name
gae
gae_lambda
gamma
gym_id
kle_rollback
kle_stop
learning_rate
max_grad_norm
minibatch_size
n_minibatch
norm_adv
num_bot_envs
num_envs
num_selfplay_envs
num_steps
prod_mode
seed
target_kl
torch_deterministic
total_timesteps
update_epochs
vf_coef
wandb_entity
wandb_project_name
charts/episode_reward
charts/episode_reward/AttackRewardFunction
charts/episode_reward/ProduceBuildingRewardFunction
charts/episode_reward/ProduceCombatUnitRewardFunction
charts/episode_reward/ProduceWorkerRewardFunction
charts/episode_reward/ResourceGatherRewardFunction
charts/episode_reward/WinLossRewardFunction
charts/learning_rate
charts/sps
charts/update
global_step
losses/approx_kl
Finished
-
chrisbam4d
1mo 23d 18h 36m 7s
-
true
12288
true
0.1
true
true
0.01
ppo_gridnet_diverse_encode_decode
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
48
48
0
256
true
2.5
0.03
true
300000000
4
0.5
["chrisbam4d","vwxyzjn"]
["gridnet","gym-microrts-paper"]
185.59999
26
1.75
9.5
19.5
91.75
1
1.0240e-8
828.75
24414
300002304
-1.4285e-7
Finished
-
costa-huang
8d 21h 57m 43s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_coacai_naive
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
7.55
0.75
0.25
0
7.5
9.25
-1
5.1200e-9
10088.5
48828
300005376
3370.93408
Finished
-
costa-huang
12d 22h 42m 41s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_coacai_no_mask
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
-5.25
0
0
0
3.75
1
-1
5.1200e-9
35564
48828
300005376
-0.0000012318
Finished
-
costa-huang
8d 17h 37m 22s
-
true
12288
true
0.1
true
true
0.01
ppo_coacai_naive
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
16.05
1.25
0.25
0
10
14.75
-1
1.0240e-8
5741.5
24414
300011520
16.43764
Finished
-
costa-huang
9d 12h 59m 39s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_coacai_partial_mask
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
48.45
0
2.25
0.25
6.5
50.5
-1
5.1200e-9
19384.5
48828
300005376
4.3710e-7
Finished
-
costa-huang
8d 17h 27m 16s
-
true
12288
true
0.1
true
true
0.01
ppo_coacai_partial_mask
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
80.8
0.75
1.5
5.5
10.25
57.5
-1
1.0240e-8
26928.25
24414
300011520
4.3074e-8
Finished
-
costa-huang
1mo 4d 1h 3m 26s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_selfplay_diverse_encode_decode
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
8
24
16
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
149.65
32.75
2
5.75
18.25
72.75
0.25
5.1200e-9
10569.25
48828
300005376
1.5690e-7
Finished
-
costa-huang
1mo 3d 14h 41m
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_selfplay_encode_decode
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
0
24
24
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
148
44.5
2.5
5
19
66.5
-0.25
5.1200e-9
37524.75
48828
300005376
-2.3083e-7
Finished
-
costa-huang
4d 15h 57m 27s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_coacai
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
194.84999
28.25
1.75
10.25
21.5
93.75
1
5.1200e-9
4186.5
48828
300003840
-8.9858e-9
Finished
-
costa-huang
4d 21h 42s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_diverse_impala
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
172.64999
24.25
2
7.75
21
91
0.5
5.1200e-9
3957
48828
300005376
1.7295e-7
Finished
-
costa-huang
3d 12h 38m 29s
-
true
12288
true
0.1
true
true
0.01
ppo_diverse_impala
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
138.7
22.5
1
5.5
19.25
67.25
0.75
1.0240e-8
2968.5
24414
300011520
1.5753e-7
Finished
-
costa-huang
2d 15h 2m 20s
-
true
12288
true
0.1
true
true
0.01
ppo_coacai
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
152.2
24
1
7.25
20.25
73.75
0.5
1.0240e-8
8532.25
24414
300011520
5.7732e-8
Finished
-
costa-huang
2d 1h 8m 41s
-
true
12288
true
0.1
true
true
0.01
ppo_coacai_no_mask
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
52.4
0
2
1
11
47
-1
1.0240e-8
15114.5
24414
300011520
1.6465e-7
Finished
-
costa-huang
3d 22h 23m 36s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_naive
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
0.2
1.25
1
0
2.75
6
-1
5.1200e-9
3498.25
48828
300005376
3370.90106
Finished
-
costa-huang
4d 10h 19m 11s
-
true
6144
true
0.1
true
true
0.01
ppo_gridnet_diverse
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
1536
4
true
24
24
0
256
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
174.39999
26.5
2
9.5
19.75
89.75
0
5.1200e-9
6921.25
48828
300005376
-4.7852e-9
Finished
-
costa-huang
2d 23h 48s
-
true
12288
true
0.1
true
true
0.01
ppo_diverse
true
0.95
0.99
MicrortsDefeatCoacAIShaped-v3
false
false
0.00025
0.5
3072
4
true
-
24
-
512
true
2.5
0.03
true
300000000
4
0.5
vwxyzjn
gym-microrts-paper
127.7
24.5
1
5.5
15.5
58
0.75
1.0240e-8
17280.5
24414
300011520
2.6309e-7
1-16
of 16
Add a comment
Created with ❤️ on Weights & Biases.
https://wandb.ai/vwxyzjn/gym-microrts-paper/reports/Selecting-the-best-seed--Vmlldzo0OTY1MjU