iamshlomo commited on
Commit
ad59e7b
·
verified ·
1 Parent(s): 3c2259d

Upload 20260304_1431/checkpoint_eval_iter4000.json with huggingface_hub

Browse files
20260304_1431/checkpoint_eval_iter4000.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "iteration": 4000,
3
+ "date": "2026-03-05",
4
+ "timestamp": "2026-03-05T05:34:48.140309",
5
+ "training": {
6
+ "lr": 3e-05,
7
+ "batch_size": 256,
8
+ "aux_loss_weight": 0.5,
9
+ "buffer_size": 10000,
10
+ "buffer_capacity": 10000,
11
+ "ema_decay": 0.999,
12
+ "recent_loss_total": 0.005393160260326694,
13
+ "recent_loss_diffusion": 0.0051697268105763946,
14
+ "recent_loss_aux": 0.00044686689950292936,
15
+ "grad_steps_per_iteration": 50,
16
+ "episodes_per_iteration": 10
17
+ },
18
+ "eval_episodes_per_env": 20,
19
+ "id_winrate": 0.5249999999999999,
20
+ "ood_winrate": 0.05000000000000001,
21
+ "per_env": {
22
+ "MiniHack-Room-Random-5x5-v0": {
23
+ "win_rate": 1.0,
24
+ "avg_reward": 22.232,
25
+ "avg_steps": 3.3,
26
+ "wins": 20,
27
+ "episodes": 20
28
+ },
29
+ "MiniHack-Room-Random-15x15-v0": {
30
+ "win_rate": 0.45,
31
+ "avg_reward": 12.29100000000001,
32
+ "avg_steps": 115.6,
33
+ "wins": 9,
34
+ "episodes": 20
35
+ },
36
+ "MiniHack-Corridor-R2-v0": {
37
+ "win_rate": 0.35,
38
+ "avg_reward": 6.459999999999996,
39
+ "avg_steps": 132.1,
40
+ "wins": 7,
41
+ "episodes": 20
42
+ },
43
+ "MiniHack-MazeWalk-9x9-v0": {
44
+ "win_rate": 0.3,
45
+ "avg_reward": 4.450499999999997,
46
+ "avg_steps": 141.45,
47
+ "wins": 6,
48
+ "episodes": 20
49
+ },
50
+ "MiniHack-Room-Dark-15x15-v0": {
51
+ "win_rate": 0.1,
52
+ "avg_reward": 0.8559999999999987,
53
+ "avg_steps": 180.65,
54
+ "wins": 2,
55
+ "episodes": 20
56
+ },
57
+ "MiniHack-Corridor-R5-v0": {
58
+ "win_rate": 0.05,
59
+ "avg_reward": -1.6570000000000022,
60
+ "avg_steps": 191.5,
61
+ "wins": 1,
62
+ "episodes": 20
63
+ },
64
+ "MiniHack-MazeWalk-45x19-v0": {
65
+ "win_rate": 0.0,
66
+ "avg_reward": -2.7575000000000016,
67
+ "avg_steps": 200.0,
68
+ "wins": 0,
69
+ "episodes": 20
70
+ }
71
+ }
72
+ }