davidquarel commited on
Commit
bc3cb86
·
verified ·
1 Parent(s): 607374a

Upload folder using huggingface_hub

Browse files
al_0.47_g_0.95_117/action_logprobs.pth.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:ff7bda3d50445621e367865e1b20ad3ba045e53ce9e7f147499e298978260aaa
3
+ size 356889323
al_0.47_g_0.95_117/action_probs.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:65137214284afd073a04702ce684767108d324e57ea68e3bcda9506182d3257f
3
+ size 30037559
al_0.47_g_0.95_117/checkpoints.tar.gz ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:af186e22cc33fcfe20cff420bbc215f150aa7bd8daa17af333343a996d510557
3
+ size 648759390
al_0.47_g_0.95_117/config.cfg ADDED
@@ -0,0 +1,44 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ rl_action=train
2
+ num_rollout_steps=64
3
+ lr=5e-05
4
+ discount_rate=0.95
5
+ eff_horizon=None
6
+ eval_every=1
7
+ use_wandb=True
8
+ use_hf=True
9
+ use_log=True
10
+ num_total_env_steps=5000000000
11
+ checkpoint=al_0.47_g_0.95_117
12
+ render_sixel=False
13
+ seed=117
14
+ mask_type=first_episode
15
+ penalize_time=False
16
+ optim=adam
17
+ live_monitor=False
18
+ checkpoint_schedule=0:1,250:2,500:5,1000:10,2000:20
19
+ grad_acc_per_chunk=5
20
+ num_rollout_chunks=1
21
+ cheese_loc=any
22
+ env_layout=open
23
+ alpha=0.47
24
+ env_size=13
25
+ num_levels=9600
26
+ f_str_ckpt=al_{alpha}_g_{discount_rate}_{seed}
27
+ wandb_project=jaxgmg_3phase_seed
28
+ ckpt_dir=jaxgmg_3phase_seed
29
+ duplication_factor=-1
30
+ smoke=False
31
+ num_chains=6
32
+ num_draws=3000
33
+ on_policy=True
34
+ nbeta=3000
35
+ localization=10
36
+ exact_solver_each_draw=False
37
+ llc_optimizer=sgld
38
+ iw_clip_eps=None
39
+ rmsprop_burnin=20
40
+ llc_data_file=llc_scan_open_reinforce.pkl
41
+ llc_checkpoint_index=0
42
+ repo_id=davidquarel/jaxgmg_ckpt_zip
43
+ use_shuffled_checkpoints=0
44
+ force_re_download=False
al_0.47_g_0.95_117/config.json ADDED
@@ -0,0 +1,46 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "alpha": 0.47,
3
+ "checkpoint": "al_0.47_g_0.95_117",
4
+ "checkpoint_schedule": "0:1,250:2,500:5,1000:10,2000:20",
5
+ "cheese_loc": "any",
6
+ "ckpt_dir": "jaxgmg_3phase_seed",
7
+ "discount_rate": 0.95,
8
+ "duplication_factor": -1,
9
+ "eff_horizon": null,
10
+ "env_layout": "open",
11
+ "env_size": 13,
12
+ "eval_every": 1,
13
+ "exact_solver_each_draw": false,
14
+ "f_str_ckpt": "al_{alpha}_g_{discount_rate}_{seed}",
15
+ "force_re_download": false,
16
+ "grad_acc_per_chunk": 5,
17
+ "iw_clip_eps": null,
18
+ "live_monitor": false,
19
+ "llc_checkpoint_index": 0,
20
+ "llc_data_file": "llc_scan_open_reinforce.pkl",
21
+ "llc_optimizer": "sgld",
22
+ "localization": 10,
23
+ "lr": 5e-05,
24
+ "mask_type": "first_episode",
25
+ "nbeta": 3000,
26
+ "num_chains": 6,
27
+ "num_draws": 3000,
28
+ "num_levels": 9600,
29
+ "num_rollout_chunks": 1,
30
+ "num_rollout_steps": 64,
31
+ "num_total_env_steps": 5000000000,
32
+ "on_policy": true,
33
+ "optim": "adam",
34
+ "penalize_time": false,
35
+ "render_sixel": false,
36
+ "repo_id": "davidquarel/jaxgmg_ckpt_zip",
37
+ "rl_action": "train",
38
+ "rmsprop_burnin": 20,
39
+ "seed": 117,
40
+ "smoke": false,
41
+ "use_hf": true,
42
+ "use_log": true,
43
+ "use_shuffled_checkpoints": 0,
44
+ "use_wandb": true,
45
+ "wandb_project": "jaxgmg_3phase_seed"
46
+ }
al_0.47_g_0.95_117/eval.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
al_0.47_g_0.95_117/eval.log ADDED
The diff for this file is too large to render. See raw diff
 
al_0.47_g_0.95_117/latest_eval.json ADDED
@@ -0,0 +1,9 @@
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ts": "2025-10-18T10:11:12.637549Z",
3
+ "loop": 8120,
4
+ "env_steps": 4988928000,
5
+ "exact_value_avg": 0.44409066438674927,
6
+ "regret": 0.2424701601266861,
7
+ "action_probs_image_path": "jaxgmg_3phase_seed/al_0.47_g_0.95_117/action_probs/img_8120.png",
8
+ "entropy": 0.028256108984351158
9
+ }
al_0.47_g_0.95_117/latest_train.json ADDED
@@ -0,0 +1,12 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "ts": "2025-10-18T10:11:34.425159Z",
3
+ "loop": 8137,
4
+ "env_steps": 4999987199,
5
+ "loss": 0.000825244074803777,
6
+ "avg_return": 0.09007794260978699,
7
+ "regret": 0.2424701601266861,
8
+ "entropy": 0.03259697183966637,
9
+ "lr": 5e-05,
10
+ "discount_rate": 0.95,
11
+ "num_rollout_steps": 64
12
+ }
al_0.47_g_0.95_117/setup.json ADDED
@@ -0,0 +1,61 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "config": {
3
+ "rl_action": "train",
4
+ "num_rollout_steps": 64,
5
+ "lr": 5e-05,
6
+ "discount_rate": 0.95,
7
+ "eff_horizon": null,
8
+ "eval_every": 1,
9
+ "use_wandb": true,
10
+ "use_hf": true,
11
+ "use_log": true,
12
+ "num_total_env_steps": 5000000000,
13
+ "checkpoint": "al_0.47_g_0.95_117",
14
+ "render_sixel": false,
15
+ "seed": 117,
16
+ "mask_type": "first_episode",
17
+ "penalize_time": false,
18
+ "optim": "adam",
19
+ "live_monitor": false,
20
+ "checkpoint_schedule": "0:1,250:2,500:5,1000:10,2000:20",
21
+ "grad_acc_per_chunk": 5,
22
+ "num_rollout_chunks": 1,
23
+ "cheese_loc": "any",
24
+ "env_layout": "open",
25
+ "alpha": 0.47,
26
+ "env_size": 13,
27
+ "num_levels": 9600,
28
+ "f_str_ckpt": "al_{alpha}_g_{discount_rate}_{seed}",
29
+ "wandb_project": "jaxgmg_3phase_seed",
30
+ "ckpt_dir": "jaxgmg_3phase_seed",
31
+ "duplication_factor": -1,
32
+ "smoke": false,
33
+ "num_chains": 6,
34
+ "num_draws": 3000,
35
+ "on_policy": true,
36
+ "nbeta": 3000,
37
+ "localization": 10,
38
+ "exact_solver_each_draw": false,
39
+ "llc_optimizer": "sgld",
40
+ "iw_clip_eps": null,
41
+ "rmsprop_burnin": 20,
42
+ "llc_data_file": "llc_scan_open_reinforce.pkl",
43
+ "llc_checkpoint_index": 0,
44
+ "repo_id": "davidquarel/jaxgmg_ckpt_zip",
45
+ "use_shuffled_checkpoints": 0,
46
+ "force_re_download": false
47
+ },
48
+ "setup": {
49
+ "start_time_utc": "2025-10-18T07:12:00.104716Z",
50
+ "seed": 117,
51
+ "device": "cuda",
52
+ "python_version": "3.11.11",
53
+ "torch_version": "2.9.0+cu128",
54
+ "jax_version": "0.6.2",
55
+ "hostname": "d960a4285319",
56
+ "platform": "Linux-6.8.0-60-generic-x86_64-with-glibc2.35",
57
+ "git_sha": null,
58
+ "git_branch": null,
59
+ "use_wandb": true
60
+ }
61
+ }
al_0.47_g_0.95_117/setup.txt ADDED
@@ -0,0 +1,8 @@
 
 
 
 
 
 
 
 
 
1
+ Start: 2025-10-18T07:12:00.104716Z
2
+ Seed: 117
3
+ Device: cuda
4
+ Python: 3.11.11 | Torch: 2.9.0+cu128 | JAX: 0.6.2
5
+ Host: d960a4285319
6
+ Platform: Linux-6.8.0-60-generic-x86_64-with-glibc2.35
7
+ Git: branch=None sha=None
8
+ use_wandb: True
al_0.47_g_0.95_117/train.jsonl ADDED
The diff for this file is too large to render. See raw diff
 
al_0.47_g_0.95_117/train.log ADDED
The diff for this file is too large to render. See raw diff