Repository navigation
Expand file tree
/
Copy pathstderr.log
More file actions
132 lines (132 loc) · 9.08 KB
/
Copy pathstderr.log
File metadata and controls
132 lines (132 loc) · 9.08 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
wandb: Currently logged in as: daveey (metta-research). Use `wandb login --relogin` to force relogin
wandb: WARNING Tried to auto resume run with id 8yje5rrx but id mac.carbsweep.5 is set.
wandb: - Waiting for wandb.init()...wandb: \ Waiting for wandb.init()...wandb: wandb version 0.18.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.7
wandb: Run data is saved locally in /Users/daveey/code/metta/wandb/run-20240923_113435-mac.carbsweep.5
wandb: Run `wandb offline` to turn off syncing.
wandb: Resuming run mac.carbsweep.5
wandb: ⭐️ View project at https://wandb.ai/metta-research/metta
wandb: 🚀 View run at https://wandb.ai/metta-research/metta/runs/mac.carbsweep.5
wandb: - 0.000 MB of 0.020 MB uploadedwandb: \ 0.000 MB of 0.020 MB uploadedwandb: | 0.020 MB of 0.020 MB uploadedwandb:
wandb: 🚀 View run mac.carbsweep.5 at: https://wandb.ai/metta-research/metta/runs/mac.carbsweep.5
wandb: ⭐️ View project at: https://wandb.ai/metta-research/metta
wandb: Synced 4 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)
wandb: Agent Starting Run: gd4m6u6c with config:
wandb: agent.observation_encoder.normalize_features: 0.7273423321304973
wandb: env.normalize_rewards: 0.10469357096493705
wandb: train.batch_size: 829
wandb: train.bptt_horizon: 4
wandb: train.clip_coef: 0.10144231745323584
wandb: train.ent_coef: 8.276229876380693e-05
wandb: train.forward_pass_minibatch_target_size: 352
wandb: train.gae_lambda: 0.35650277894554605
wandb: train.gamma: 0.5606347026752192
wandb: train.learning_rate: 2.5915679861779305e-05
wandb: train.max_grad_norm: 0.8659747776240205
wandb: train.minibatch_size: 643
wandb: train.total_timesteps: 11224853.068516972
wandb: train.update_epochs: 5.877633797625594
wandb: train.vf_clip_coef: 0.021921927613541725
wandb: train.vf_coef: 0.8930178523983127
wandb: WARNING Ignored wandb.init() arg project when running a sweep.
wandb: WARNING Ignored wandb.init() arg entity when running a sweep.
wandb: WARNING Ignored wandb.init() arg id when running a sweep.
wandb: wandb version 0.18.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.7
wandb: Run data is saved locally in /Users/daveey/code/metta/wandb/run-20240923_113444-gd4m6u6c
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run mac.carbsweep.5
wandb: ⭐️ View project at https://wandb.ai/metta-research/metta
wandb: 🧹 View sweep at https://wandb.ai/metta-research/metta/sweeps/qa3y7v8j
wandb: 🚀 View run at https://wandb.ai/metta-research/metta/runs/gd4m6u6c
/Users/daveey/code/metta/deps/carbs/carbs/carbs.py:838: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
state = torch.load(f)
wandb: - 0.013 MB of 0.013 MB uploadedwandb: \ 0.013 MB of 0.013 MB uploadedwandb: | 0.045 MB of 0.045 MB uploadedwandb:
wandb: 🚀 View run mac.carbsweep.5 at: https://wandb.ai/metta-research/metta/runs/gd4m6u6c
wandb: ⭐️ View project at: https://wandb.ai/metta-research/metta
wandb: Synced 7 W&B file(s), 0 media file(s), 1 artifact file(s) and 1 other file(s)
wandb: WARNING Ignored wandb.init() arg project when running a sweep.
wandb: WARNING Ignored wandb.init() arg entity when running a sweep.
wandb: WARNING Ignored wandb.init() arg id when running a sweep.
wandb: - Waiting for wandb.init()...wandb: \ Waiting for wandb.init()...wandb: wandb version 0.18.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.7
wandb: Run data is saved locally in /Users/daveey/code/metta/wandb/run-20240923_113456-gd4m6u6c
wandb: Run `wandb offline` to turn off syncing.
wandb: Resuming run mac.carbsweep.5-4
wandb: ⭐️ View project at https://wandb.ai/metta-research/metta
wandb: 🧹 View sweep at https://wandb.ai/metta-research/metta/sweeps/qa3y7v8j
wandb: 🚀 View run at https://wandb.ai/metta-research/metta/runs/gd4m6u6c
wandb: - 0.667 MB of 0.667 MB uploadedwandb: \ 0.667 MB of 0.667 MB uploadedwandb: | 0.708 MB of 0.708 MB uploadedwandb:
wandb: 🚀 View run mac.carbsweep.5-4 at: https://wandb.ai/metta-research/metta/runs/gd4m6u6c
wandb: ⭐️ View project at: https://wandb.ai/metta-research/metta
wandb: Synced 4 W&B file(s), 0 media file(s), 1 artifact file(s) and 0 other file(s)
wandb: WARNING Ignored wandb.init() arg project when running a sweep.
wandb: WARNING Ignored wandb.init() arg entity when running a sweep.
wandb: WARNING Ignored wandb.init() arg id when running a sweep.
wandb: - Waiting for wandb.init()...wandb: \ Waiting for wandb.init()...wandb: wandb version 0.18.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.7
wandb: Run data is saved locally in /Users/daveey/code/metta/wandb/run-20240923_113552-gd4m6u6c
wandb: Run `wandb offline` to turn off syncing.
wandb: Resuming run mac.carbsweep.5
wandb: ⭐️ View project at https://wandb.ai/metta-research/metta
wandb: 🧹 View sweep at https://wandb.ai/metta-research/metta/sweeps/qa3y7v8j
wandb: 🚀 View run at https://wandb.ai/metta-research/metta/runs/gd4m6u6c
/Users/daveey/code/metta/deps/carbs/carbs/carbs.py:838: FutureWarning: You are using `torch.load` with `weights_only=False` (the current default value), which uses the default pickle module implicitly. It is possible to construct malicious pickle data which will execute arbitrary code during unpickling (See https://github.com/pytorch/pytorch/blob/main/SECURITY.md#untrusted-models for more details). In a future release, the default value for `weights_only` will be flipped to `True`. This limits the functions that could be executed during unpickling. Arbitrary objects will no longer be allowed to be loaded via this mode unless they are explicitly allowlisted by the user via `torch.serialization.add_safe_globals`. We recommend you start setting `weights_only=True` for any use case where you don't have full control of the loaded file. Please open an issue on GitHub for any issues related to this experimental feature.
state = torch.load(f)
wandb: - 0.007 MB of 0.007 MB uploadedwandb: \ 0.007 MB of 0.007 MB uploadedwandb: | 0.041 MB of 0.041 MB uploadedwandb:
wandb: 🚀 View run mac.carbsweep.5 at: https://wandb.ai/metta-research/metta/runs/gd4m6u6c
wandb: ⭐️ View project at: https://wandb.ai/metta-research/metta
wandb: Synced 4 W&B file(s), 0 media file(s), 1 artifact file(s) and 1 other file(s)
wandb: Agent Starting Run: wemyn25s with config:
wandb: agent.observation_encoder.normalize_features: 0.10576538581315742
wandb: env.normalize_rewards: 0.29235749181684767
wandb: train.batch_size: 969
wandb: train.bptt_horizon: 7
wandb: train.clip_coef: 0.5193525316356885
wandb: train.ent_coef: 0.0013185212566453586
wandb: train.forward_pass_minibatch_target_size: 589
wandb: train.gae_lambda: 0.5177921429565214
wandb: train.gamma: 0.7199331571570531
wandb: train.learning_rate: 3.252150314635788e-05
wandb: train.max_grad_norm: 0.14602819624283647
wandb: train.minibatch_size: 677
wandb: train.total_timesteps: 30959502.42098912
wandb: train.update_epochs: 5.918871464270146
wandb: train.vf_clip_coef: 0.02316386260309733
wandb: train.vf_coef: 0.5951468431410942
wandb: WARNING Ignored wandb.init() arg project when running a sweep.
wandb: WARNING Ignored wandb.init() arg entity when running a sweep.
wandb: WARNING Ignored wandb.init() arg id when running a sweep.
wandb: - Waiting for wandb.init()...wandb: \ Waiting for wandb.init()...wandb: wandb version 0.18.1 is available! To upgrade, please run:
wandb: $ pip install wandb --upgrade
wandb: Tracking run with wandb version 0.17.7
wandb: Run data is saved locally in /Users/daveey/code/metta/wandb/run-20240923_113610-wemyn25s
wandb: Run `wandb offline` to turn off syncing.
wandb: Syncing run mac.carbsweep.5
wandb: ⭐️ View project at https://wandb.ai/metta-research/metta
wandb: 🧹 View sweep at https://wandb.ai/metta-research/metta/sweeps/qa3y7v8j
wandb: 🚀 View run at https://wandb.ai/metta-research/metta/runs/wemyn25s