iamshlomo commited on
Commit
d1d0627
·
verified ·
1 Parent(s): ba12371

Upload 20260303_0938/checkpoint_eval_iter2000.json with huggingface_hub

Browse files
20260303_0938/checkpoint_eval_iter2000.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "iteration": 2000,
3
+ "date": "2026-03-03",
4
+ "timestamp": "2026-03-03T16:20:30.135402",
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": 0.010490936765435617,
13
+ "recent_loss_diffusion": 0.009870809475842396,
14
+ "recent_loss_aux": 0.001240254579881439,
15
+ "grad_steps_per_iteration": 50,
16
+ "episodes_per_iteration": 10
17
+ },
18
+ "eval_episodes_per_env": 20,
19
+ "id_winrate": 0.4875,
20
+ "ood_winrate": 0.016666666666666666,
21
+ "per_env": {
22
+ "MiniHack-Room-Random-5x5-v0": {
23
+ "win_rate": 1.0,
24
+ "avg_reward": 22.447999999999997,
25
+ "avg_steps": 3.7,
26
+ "wins": 20,
27
+ "episodes": 20
28
+ },
29
+ "MiniHack-Room-Random-15x15-v0": {
30
+ "win_rate": 0.6,
31
+ "avg_reward": 14.825000000000003,
32
+ "avg_steps": 85.35,
33
+ "wins": 12,
34
+ "episodes": 20
35
+ },
36
+ "MiniHack-Corridor-R2-v0": {
37
+ "win_rate": 0.15,
38
+ "avg_reward": 0.7964999999999997,
39
+ "avg_steps": 170.7,
40
+ "wins": 3,
41
+ "episodes": 20
42
+ },
43
+ "MiniHack-MazeWalk-9x9-v0": {
44
+ "win_rate": 0.2,
45
+ "avg_reward": 1.8534999999999997,
46
+ "avg_steps": 160.55,
47
+ "wins": 4,
48
+ "episodes": 20
49
+ },
50
+ "MiniHack-Room-Dark-15x15-v0": {
51
+ "win_rate": 0.05,
52
+ "avg_reward": -0.3730000000000015,
53
+ "avg_steps": 190.05,
54
+ "wins": 1,
55
+ "episodes": 20
56
+ },
57
+ "MiniHack-Corridor-R5-v0": {
58
+ "win_rate": 0.0,
59
+ "avg_reward": -3.1230000000000024,
60
+ "avg_steps": 200.0,
61
+ "wins": 0,
62
+ "episodes": 20
63
+ },
64
+ "MiniHack-MazeWalk-45x19-v0": {
65
+ "win_rate": 0.0,
66
+ "avg_reward": -3.4760000000000026,
67
+ "avg_steps": 200.0,
68
+ "wins": 0,
69
+ "episodes": 20
70
+ }
71
+ }
72
+ }