Hugging Face's logo

Datasets:
SLM-Lab
/
benchmark

Modalities:
Image
Formats:
imagefolder
Size:
1K - 10K
Libraries:
Datasets
Dataset card Data Studio Files
xet
 Community
1
kengzwl commited on
Commit
900cd38
·
1 Parent(s): 697e839

Upload folder using huggingface_hub

Browse files
Files changed (8) hide show
  1. data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s2_net_model.pt +3 -0
  2. data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s3_ckpt-best_net_model.pt +3 -0
  3. data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s3_net_model.pt +3 -0
  4. data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_spec.json +113 -0
  5. data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_spec.json +113 -0
  6. data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_graph_mean_returns_ma_vs_frames.png +3 -0
  7. data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_graph_mean_returns_vs_frames.png +3 -0
  8. data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_metrics_scalar.json +11 -0
data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s2_net_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9eb9fbf1d1ddd3481e5f463cfe157c492f1fcdd97d6917e331d39c1ec804494e
3
+ size 6760749
data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s3_ckpt-best_net_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:94750e582b695a567fdd3b5ed48766aa46ef5994aa6e534b40bbfe2c19754103
3
+ size 6760993
data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0_s3_net_model.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:79a2f0ff74702629784bb15698a017f2a8a0780e550bada3c69592813d7015aa
3
+ size 6760749
data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_spec.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "agent": {
3
+ "name": "PPO",
4
+ "algorithm": {
5
+ "name": "PPO",
6
+ "action_pdtype": "default",
7
+ "action_policy": "default",
8
+ "explore_var_spec": null,
9
+ "gamma": 0.99,
10
+ "lam": 0.7,
11
+ "clip_eps_spec": {
12
+ "name": "no_decay",
13
+ "start_val": 0.1,
14
+ "end_val": 0.1,
15
+ "start_step": 0,
16
+ "end_step": 0
17
+ },
18
+ "entropy_coef_spec": {
19
+ "name": "no_decay",
20
+ "start_val": 0.01,
21
+ "end_val": 0.01,
22
+ "start_step": 0,
23
+ "end_step": 0
24
+ },
25
+ "val_loss_coef": 0.5,
26
+ "time_horizon": 128,
27
+ "minibatch_size": 256,
28
+ "training_epoch": 4,
29
+ "clip_vloss": true
30
+ },
31
+ "memory": {
32
+ "name": "OnPolicyBatchReplay"
33
+ },
34
+ "net": {
35
+ "type": "ConvNet",
36
+ "shared": true,
37
+ "conv_hid_layers": [
38
+ [
39
+ 32,
40
+ 8,
41
+ 4,
42
+ 0,
43
+ 1
44
+ ],
45
+ [
46
+ 64,
47
+ 4,
48
+ 2,
49
+ 0,
50
+ 1
51
+ ],
52
+ [
53
+ 64,
54
+ 3,
55
+ 1,
56
+ 0,
57
+ 1
58
+ ]
59
+ ],
60
+ "fc_hid_layers": [
61
+ 512
62
+ ],
63
+ "hid_layers_activation": "relu",
64
+ "init_fn": "orthogonal_",
65
+ "normalize": true,
66
+ "batch_norm": false,
67
+ "clip_grad_val": 0.5,
68
+ "use_same_optim": true,
69
+ "loss_spec": {
70
+ "name": "MSELoss"
71
+ },
72
+ "optim_spec": {
73
+ "name": "AdamW",
74
+ "lr": 0.00025,
75
+ "eps": 1e-05
76
+ },
77
+ "lr_scheduler_spec": {
78
+ "name": "LinearToZero",
79
+ "frame": 10000000.0
80
+ },
81
+ "gpu": "auto"
82
+ }
83
+ },
84
+ "env": {
85
+ "name": "ALE/Gopher-v5",
86
+ "num_envs": 16,
87
+ "max_frame": 10000000,
88
+ "life_loss_info": true
89
+ },
90
+ "meta": {
91
+ "distributed": false,
92
+ "eval_frequency": 10000,
93
+ "log_frequency": 10000,
94
+ "max_session": 4,
95
+ "max_trial": 1,
96
+ "rigorous_eval": 0,
97
+ "experiment": 0,
98
+ "trial": 0,
99
+ "session": -1,
100
+ "cuda_offset": 0,
101
+ "resume": false,
102
+ "experiment_ts": "2026_01_07_170451",
103
+ "prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0",
104
+ "git_sha": "unknown",
105
+ "random_seed": null,
106
+ "predir": "data/ppo_atari_lam70_gopher_2026_01_07_170451",
107
+ "graph_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/graph/ppo_atari_lam70_gopher_t0",
108
+ "info_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/info/ppo_atari_lam70_gopher_t0",
109
+ "log_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/log/ppo_atari_lam70_gopher_t0",
110
+ "model_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0"
111
+ },
112
+ "name": "ppo_atari_lam70_gopher"
113
+ }
data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_spec.json ADDED
@@ -0,0 +1,113 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "agent": {
3
+ "name": "PPO",
4
+ "algorithm": {
5
+ "name": "PPO",
6
+ "action_pdtype": "default",
7
+ "action_policy": "default",
8
+ "explore_var_spec": null,
9
+ "gamma": 0.99,
10
+ "lam": 0.7,
11
+ "clip_eps_spec": {
12
+ "name": "no_decay",
13
+ "start_val": 0.1,
14
+ "end_val": 0.1,
15
+ "start_step": 0,
16
+ "end_step": 0
17
+ },
18
+ "entropy_coef_spec": {
19
+ "name": "no_decay",
20
+ "start_val": 0.01,
21
+ "end_val": 0.01,
22
+ "start_step": 0,
23
+ "end_step": 0
24
+ },
25
+ "val_loss_coef": 0.5,
26
+ "time_horizon": 128,
27
+ "minibatch_size": 256,
28
+ "training_epoch": 4,
29
+ "clip_vloss": true
30
+ },
31
+ "memory": {
32
+ "name": "OnPolicyBatchReplay"
33
+ },
34
+ "net": {
35
+ "type": "ConvNet",
36
+ "shared": true,
37
+ "conv_hid_layers": [
38
+ [
39
+ 32,
40
+ 8,
41
+ 4,
42
+ 0,
43
+ 1
44
+ ],
45
+ [
46
+ 64,
47
+ 4,
48
+ 2,
49
+ 0,
50
+ 1
51
+ ],
52
+ [
53
+ 64,
54
+ 3,
55
+ 1,
56
+ 0,
57
+ 1
58
+ ]
59
+ ],
60
+ "fc_hid_layers": [
61
+ 512
62
+ ],
63
+ "hid_layers_activation": "relu",
64
+ "init_fn": "orthogonal_",
65
+ "normalize": true,
66
+ "batch_norm": false,
67
+ "clip_grad_val": 0.5,
68
+ "use_same_optim": true,
69
+ "loss_spec": {
70
+ "name": "MSELoss"
71
+ },
72
+ "optim_spec": {
73
+ "name": "AdamW",
74
+ "lr": 0.00025,
75
+ "eps": 1e-05
76
+ },
77
+ "lr_scheduler_spec": {
78
+ "name": "LinearToZero",
79
+ "frame": 10000000.0
80
+ },
81
+ "gpu": "auto"
82
+ }
83
+ },
84
+ "env": {
85
+ "name": "ALE/Gopher-v5",
86
+ "num_envs": 16,
87
+ "max_frame": 10000000,
88
+ "life_loss_info": true
89
+ },
90
+ "meta": {
91
+ "distributed": false,
92
+ "eval_frequency": 10000,
93
+ "log_frequency": 10000,
94
+ "max_session": 4,
95
+ "max_trial": 1,
96
+ "rigorous_eval": 0,
97
+ "experiment": 0,
98
+ "trial": 0,
99
+ "session": -1,
100
+ "cuda_offset": 0,
101
+ "resume": false,
102
+ "experiment_ts": "2026_01_07_170451",
103
+ "prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0",
104
+ "git_sha": "unknown",
105
+ "random_seed": null,
106
+ "predir": "data/ppo_atari_lam70_gopher_2026_01_07_170451",
107
+ "graph_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/graph/ppo_atari_lam70_gopher_t0",
108
+ "info_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/info/ppo_atari_lam70_gopher_t0",
109
+ "log_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/log/ppo_atari_lam70_gopher_t0",
110
+ "model_prepath": "data/ppo_atari_lam70_gopher_2026_01_07_170451/model/ppo_atari_lam70_gopher_t0"
111
+ },
112
+ "name": "ppo_atari_lam70_gopher"
113
+ }
data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_graph_mean_returns_ma_vs_frames.png ADDED

Git LFS Details

  • SHA256: 797708c5c285ec7fd1eb277429890d4b287cdb244b85782500946125f3a4dfee
  • Pointer size: 130 Bytes
  • Size of remote file: 94.5 kB
data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_graph_mean_returns_vs_frames.png ADDED

Git LFS Details

  • SHA256: 44ff67ca827749821c5ce011fa20757db8eeb3f138bdfb874f496e4aeafba73a
  • Pointer size: 131 Bytes
  • Size of remote file: 112 kB
data/ppo_atari_lam70_gopher_2026_01_07_170451/ppo_atari_lam70_gopher_t0_trial_metrics_scalar.json ADDED
@@ -0,0 +1,11 @@
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "frame": 10000000.0,
3
+ "total_reward_ma": 6507.68798828125,
4
+ "strength": 3442.645263671875,
5
+ "max_strength": 6682.06640625,
6
+ "final_strength": 6495.06640625,
7
+ "sample_efficiency": 1.925097308230761e-07,
8
+ "training_efficiency": 1.2273310858290642e-05,
9
+ "stability": 0.9948186874389648,
10
+ "consistency": 0.5782780051231384
11
+ }