iamshlomo commited on
Commit
382d196
·
verified ·
1 Parent(s): 223b838

Upload 20260304_1431/checkpoint_eval_iter5000.json with huggingface_hub

Browse files
20260304_1431/checkpoint_eval_iter5000.json ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "iteration": 5000,
3
+ "date": "2026-03-05",
4
+ "timestamp": "2026-03-05T09:05:08.543989",
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": 0.004757914452753029,
13
+ "recent_loss_diffusion": 0.004590326266215998,
14
+ "recent_loss_aux": 0.0003351763730977109,
15
+ "grad_steps_per_iteration": 50,
16
+ "episodes_per_iteration": 10
17
+ },
18
+ "eval_episodes_per_env": 20,
19
+ "id_winrate": 0.575,
20
+ "ood_winrate": 0.08333333333333333,
21
+ "per_env": {
22
+ "MiniHack-Room-Random-5x5-v0": {
23
+ "win_rate": 1.0,
24
+ "avg_reward": 22.177999999999997,
25
+ "avg_steps": 3.2,
26
+ "wins": 20,
27
+ "episodes": 20
28
+ },
29
+ "MiniHack-Room-Random-15x15-v0": {
30
+ "win_rate": 0.55,
31
+ "avg_reward": 13.869500000000006,
32
+ "avg_steps": 95.05,
33
+ "wins": 11,
34
+ "episodes": 20
35
+ },
36
+ "MiniHack-Corridor-R2-v0": {
37
+ "win_rate": 0.45,
38
+ "avg_reward": 7.641999999999999,
39
+ "avg_steps": 110.8,
40
+ "wins": 9,
41
+ "episodes": 20
42
+ },
43
+ "MiniHack-MazeWalk-9x9-v0": {
44
+ "win_rate": 0.3,
45
+ "avg_reward": 4.850999999999998,
46
+ "avg_steps": 141.3,
47
+ "wins": 6,
48
+ "episodes": 20
49
+ },
50
+ "MiniHack-Room-Dark-15x15-v0": {
51
+ "win_rate": 0.25,
52
+ "avg_reward": 4.263499999999999,
53
+ "avg_steps": 152.3,
54
+ "wins": 5,
55
+ "episodes": 20
56
+ },
57
+ "MiniHack-Corridor-R5-v0": {
58
+ "win_rate": 0.0,
59
+ "avg_reward": -3.4315000000000024,
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": -2.2795000000000014,
67
+ "avg_steps": 200.0,
68
+ "wins": 0,
69
+ "episodes": 20
70
+ }
71
+ }
72
+ }