Hugging Face's logo

Datasets:
Fengzhuo
/
optimizer_results

Dataset card Files
xet
Community
1
zhangfz commited on
Commit
4596381
·
1 Parent(s): d1c0fd0

update

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_42/config.json +25 -0
  2. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_42/training_log_0f19f4fa-5d1f-488f-a0ef-2821e6b96808.txt +0 -0
  3. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_45/config.json +25 -0
  4. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_45/training_log_a285383b-2309-488b-9713-632fe4bea5ec.txt +0 -0
  5. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_42/config.json +25 -0
  6. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_42/training_log_e87ef02d-29ff-4238-9fbb-1698562443df.txt +0 -0
  7. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_45/config.json +25 -0
  8. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_45/training_log_24f806e5-3cbe-4bd4-b906-8bdd46616d49.txt +0 -0
  9. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_42/config.json +25 -0
  10. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_42/training_log_92e22894-4d6b-415e-915f-c2bacca2c5a2.txt +0 -0
  11. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_45/config.json +25 -0
  12. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_45/training_log_0f5a9c38-f12e-4278-bfdc-4288c430c0b2.txt +0 -0
  13. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.01_seed_44/config.json +25 -0
  14. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.01_seed_44/training_log_92c7c167-73e8-4ba3-86fc-bff3b5d77c9d.txt +0 -0
  15. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.02_seed_44/config.json +25 -0
  16. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.02_seed_44/training_log_af844e7c-932e-48a9-a7eb-5d4632a6e2a4.txt +0 -0
  17. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.05_seed_44/config.json +25 -0
  18. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.05_seed_44/training_log_8214b888-b7e0-41d4-b295-d97c75ddb016.txt +0 -0
  19. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.1_seed_44/config.json +25 -0
  20. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.1_seed_44/training_log_586c59e2-6b8b-48f7-a0a3-6ebf33809948.txt +0 -0
  21. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.0005_seed_42/config.json +25 -0
  22. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.0005_seed_42/training_log_bd7c9329-f951-4b08-bd0e-740f6a81a742.txt +0 -0
  23. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.001_seed_42/config.json +25 -0
  24. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.001_seed_42/training_log_d3022089-cce3-4b89-89e0-a3b91cb980b8.txt +0 -0
  25. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.002_seed_42/config.json +25 -0
  26. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.002_seed_42/training_log_02398284-62a6-4a94-b351-aeddcff7f8c3.txt +0 -0
  27. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.01_seed_44/config.json +25 -0
  28. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.01_seed_44/training_log_99d6117a-44c7-4aea-95cb-1f775290d3ce.txt +0 -0
  29. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.02_seed_44/config.json +25 -0
  30. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.02_seed_44/training_log_44bd9f0c-4607-473b-a31d-1415570b65a7.txt +0 -0
  31. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.05_seed_44/config.json +25 -0
  32. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.05_seed_44/training_log_e89055ef-8506-4fb4-a60c-18ad9b29a00f.txt +0 -0
  33. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.1_seed_44/config.json +25 -0
  34. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.1_seed_44/training_log_888b304f-20a3-43b8-90a0-b0a38d25e473.txt +0 -0
  35. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.0005_seed_42/config.json +25 -0
  36. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.0005_seed_42/training_log_75516b8c-735e-4dc3-b9b2-7418d1ab54d3.txt +0 -0
  37. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.001_seed_42/config.json +25 -0
  38. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.001_seed_42/training_log_7063183c-b09c-4b42-8af6-35b7383a2b98.txt +0 -0
  39. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.002_seed_42/config.json +25 -0
  40. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.002_seed_42/training_log_665bdf13-c773-43a5-b8c2-4b7c276c4f27.txt +0 -0
  41. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.01_seed_44/config.json +25 -0
  42. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.01_seed_44/training_log_5bbce0aa-b752-4b4c-b6eb-1fe4fa9fd923.txt +0 -0
  43. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.02_seed_44/config.json +25 -0
  44. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.02_seed_44/training_log_9d82661b-ea83-4315-a290-ff7e391f4f9f.txt +0 -0
  45. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.05_seed_44/config.json +25 -0
  46. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.05_seed_44/training_log_ad892419-29ae-4585-8144-90dc3061d50f.txt +0 -0
  47. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.1_seed_44/config.json +25 -0
  48. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.1_seed_44/training_log_4fc1056a-2dd1-436c-a6e4-22dbc56a6d21.txt +0 -0
  49. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.0005_seed_42/config.json +25 -0
  50. sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.0005_seed_42/training_log_74bb3f20-4ad1-425e-b57d-0fe778e9ae08.txt +0 -0
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.0005,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "0f19f4fa-5d1f-488f-a0ef-2821e6b96808",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_42/training_log_0f19f4fa-5d1f-488f-a0ef-2821e6b96808.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_45/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 45,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.0005,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "a285383b-2309-488b-9713-632fe4bea5ec",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.0005_seed_45/training_log_a285383b-2309-488b-9713-632fe4bea5ec.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.001,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "e87ef02d-29ff-4238-9fbb-1698562443df",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_42/training_log_e87ef02d-29ff-4238-9fbb-1698562443df.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_45/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 45,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.001,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "24f806e5-3cbe-4bd4-b906-8bdd46616d49",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.001_seed_45/training_log_24f806e5-3cbe-4bd4-b906-8bdd46616d49.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.002,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "92e22894-4d6b-415e-915f-c2bacca2c5a2",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_42/training_log_92e22894-4d6b-415e-915f-c2bacca2c5a2.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_45/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 45,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.002,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "0f5a9c38-f12e-4278-bfdc-4288c430c0b2",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.002_seed_45/training_log_0f5a9c38-f12e-4278-bfdc-4288c430c0b2.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.01_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.01,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "92c7c167-73e8-4ba3-86fc-bff3b5d77c9d",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.01_seed_44/training_log_92c7c167-73e8-4ba3-86fc-bff3b5d77c9d.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.02_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.02,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "af844e7c-932e-48a9-a7eb-5d4632a6e2a4",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.02_seed_44/training_log_af844e7c-932e-48a9-a7eb-5d4632a6e2a4.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.05_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.05,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "8214b888-b7e0-41d4-b295-d97c75ddb016",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.05_seed_44/training_log_8214b888-b7e0-41d4-b295-d97c75ddb016.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.1_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.001,
8
+ "muon_lr": 0.1,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "586c59e2-6b8b-48f7-a0a3-6ebf33809948",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.001_muon_lr_0.1_seed_44/training_log_586c59e2-6b8b-48f7-a0a3-6ebf33809948.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.0005_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.0005,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "bd7c9329-f951-4b08-bd0e-740f6a81a742",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.0005_seed_42/training_log_bd7c9329-f951-4b08-bd0e-740f6a81a742.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.001_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.001,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "d3022089-cce3-4b89-89e0-a3b91cb980b8",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.001_seed_42/training_log_d3022089-cce3-4b89-89e0-a3b91cb980b8.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.002_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.002,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "02398284-62a6-4a94-b351-aeddcff7f8c3",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.002_seed_42/training_log_02398284-62a6-4a94-b351-aeddcff7f8c3.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.01_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.01,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "99d6117a-44c7-4aea-95cb-1f775290d3ce",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.01_seed_44/training_log_99d6117a-44c7-4aea-95cb-1f775290d3ce.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.02_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.02,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "44bd9f0c-4607-473b-a31d-1415570b65a7",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.02_seed_44/training_log_44bd9f0c-4607-473b-a31d-1415570b65a7.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.05_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.05,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "e89055ef-8506-4fb4-a60c-18ad9b29a00f",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.05_seed_44/training_log_e89055ef-8506-4fb4-a60c-18ad9b29a00f.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.1_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.002,
8
+ "muon_lr": 0.1,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "888b304f-20a3-43b8-90a0-b0a38d25e473",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.002_muon_lr_0.1_seed_44/training_log_888b304f-20a3-43b8-90a0-b0a38d25e473.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.0005_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.0005,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "75516b8c-735e-4dc3-b9b2-7418d1ab54d3",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.0005_seed_42/training_log_75516b8c-735e-4dc3-b9b2-7418d1ab54d3.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.001_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.001,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "7063183c-b09c-4b42-8af6-35b7383a2b98",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.001_seed_42/training_log_7063183c-b09c-4b42-8af6-35b7383a2b98.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.002_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.002,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "665bdf13-c773-43a5-b8c2-4b7c276c4f27",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.002_seed_42/training_log_665bdf13-c773-43a5-b8c2-4b7c276c4f27.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.01_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.01,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "5bbce0aa-b752-4b4c-b6eb-1fe4fa9fd923",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.01_seed_44/training_log_5bbce0aa-b752-4b4c-b6eb-1fe4fa9fd923.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.02_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.02,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "9d82661b-ea83-4315-a290-ff7e391f4f9f",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.02_seed_44/training_log_9d82661b-ea83-4315-a290-ff7e391f4f9f.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.05_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.05,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "ad892419-29ae-4585-8144-90dc3061d50f",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.05_seed_44/training_log_ad892419-29ae-4585-8144-90dc3061d50f.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.1_seed_44/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 44,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.005,
8
+ "muon_lr": 0.1,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "4fc1056a-2dd1-436c-a6e4-22dbc56a6d21",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.005_muon_lr_0.1_seed_44/training_log_4fc1056a-2dd1-436c-a6e4-22dbc56a6d21.txt ADDED
The diff for this file is too large to render. See raw diff
 
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.0005_seed_42/config.json ADDED
@@ -0,0 +1,25 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "unet": false,
4
+ "seed": 42,
5
+ "optimizer_mode": 0,
6
+ "model_parameterization": "qkvo",
7
+ "adam_lr": 0.01,
8
+ "muon_lr": 0.0005,
9
+ "base_dir": "logs_qkvo_grid_fix/mode_0"
10
+ },
11
+ "hyperparameters": {
12
+ "train_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
13
+ "val_files": "/home/aiops/zhangfz/MUON_theory/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
14
+ "val_tokens": 491520,
15
+ "train_seq_len": 3072,
16
+ "val_seq_len": 16384,
17
+ "num_iterations": 10000,
18
+ "cooldown_frac": 0.4,
19
+ "vocab_size": 50257,
20
+ "val_loss_every": 400,
21
+ "save_checkpoint": false
22
+ },
23
+ "run_uuid_for_log": "74bb3f20-4ad1-425e-b57d-0fe778e9ae08",
24
+ "script_code_logged_at_start": true
25
+ }
sharpness/logs_qkvo_grid_fix/mode_0/mode_0_param_qkvo_adam_lr_0.01_muon_lr_0.0005_seed_42/training_log_74bb3f20-4ad1-425e-b57d-0fe778e9ae08.txt ADDED
The diff for this file is too large to render. See raw diff