Hugging Face's logo

Datasets:
Fengzhuo
/
LLM_feature

Modalities:
Text
Formats:
text
Size:
10K - 100K
Libraries:
Datasets
Dataset card Data Studio Files
xet
 Community
1
zhangfz commited on
Commit
f4d08cc
·
1 Parent(s): 270ed47

update

Browse files
Files changed (12) hide show
  1. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/config.json +41 -0
  2. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/model_final.pt +3 -0
  3. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/training_log.txt +0 -0
  4. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/config.json +41 -0
  5. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/model_final.pt +3 -0
  6. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/training_log.txt +0 -0
  7. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/config.json +41 -0
  8. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/model_final.pt +3 -0
  9. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/training_log.txt +0 -0
  10. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/config.json +41 -0
  11. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/model_final.pt +3 -0
  12. logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/training_log.txt +0 -0
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "input_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
4
+ "input_val_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
5
+ "output_dir": "/home/aiops/zhangfz/MUON_LLM/logs_pure_qk_nonorm/test",
6
+ "model": "d12",
7
+ "batch_size": 4,
8
+ "sequence_length": 1024,
9
+ "total_batch_size": 524288,
10
+ "num_iterations": 10000,
11
+ "inference_only": 0,
12
+ "adam_lr": 0.001,
13
+ "warmup_iters": 700,
14
+ "lr_decay_frac": 0.0,
15
+ "weight_decay": 0.0,
16
+ "grad_clip": 1.0,
17
+ "val_loss_every": 250,
18
+ "val_max_steps": 20,
19
+ "sample_every": 0,
20
+ "overfit_single_batch": 0,
21
+ "shuffle_files": true,
22
+ "tensorcores": 1,
23
+ "device": "",
24
+ "compile": 1,
25
+ "flash": 1,
26
+ "dtype": "bfloat16",
27
+ "zero_stage": 1,
28
+ "optimizer": "adam",
29
+ "muon_lr": 0.01,
30
+ "muon_momentum": 0.95,
31
+ "muon_weight_decay": 0.0,
32
+ "muon_ns_steps": 5,
33
+ "muon_nesterov": false,
34
+ "write_tensors": 0,
35
+ "seed": 42,
36
+ "analyze_sharpness": false,
37
+ "sharpness_analysis_interval": 500
38
+ },
39
+ "run_uuid": "0255a650-b68d-49f5-919f-2530a1a20dcd",
40
+ "script_code_logged_at_start": true
41
+ }
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/model_final.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:51a3c66fdd60492b6771f94e47a67437ee6b42f511fe3000aa6426fc8f8dddb4
3
+ size 544851409
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_42/training_log.txt ADDED
The diff for this file is too large to render. See raw diff
 
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "input_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
4
+ "input_val_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
5
+ "output_dir": "/home/aiops/zhangfz/MUON_LLM/logs_pure_qk_nonorm/test",
6
+ "model": "d12",
7
+ "batch_size": 4,
8
+ "sequence_length": 1024,
9
+ "total_batch_size": 524288,
10
+ "num_iterations": 10000,
11
+ "inference_only": 0,
12
+ "adam_lr": 0.001,
13
+ "warmup_iters": 700,
14
+ "lr_decay_frac": 0.0,
15
+ "weight_decay": 0.0,
16
+ "grad_clip": 1.0,
17
+ "val_loss_every": 250,
18
+ "val_max_steps": 20,
19
+ "sample_every": 0,
20
+ "overfit_single_batch": 0,
21
+ "shuffle_files": true,
22
+ "tensorcores": 1,
23
+ "device": "",
24
+ "compile": 1,
25
+ "flash": 1,
26
+ "dtype": "bfloat16",
27
+ "zero_stage": 1,
28
+ "optimizer": "adam",
29
+ "muon_lr": 0.01,
30
+ "muon_momentum": 0.95,
31
+ "muon_weight_decay": 0.0,
32
+ "muon_ns_steps": 5,
33
+ "muon_nesterov": false,
34
+ "write_tensors": 0,
35
+ "seed": 43,
36
+ "analyze_sharpness": false,
37
+ "sharpness_analysis_interval": 500
38
+ },
39
+ "run_uuid": "dc681cdb-a1fe-4108-b180-efd7182cab4d",
40
+ "script_code_logged_at_start": true
41
+ }
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/model_final.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3bbcae3ae638a59997008d487c68a5cc9adce55acad55dd261d6dd619965f4a7
3
+ size 544851409
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_43/training_log.txt ADDED
The diff for this file is too large to render. See raw diff
 
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "input_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
4
+ "input_val_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
5
+ "output_dir": "/home/aiops/zhangfz/MUON_LLM/logs_pure_qk_nonorm/test",
6
+ "model": "d12",
7
+ "batch_size": 4,
8
+ "sequence_length": 1024,
9
+ "total_batch_size": 524288,
10
+ "num_iterations": 10000,
11
+ "inference_only": 0,
12
+ "adam_lr": 0.001,
13
+ "warmup_iters": 700,
14
+ "lr_decay_frac": 0.0,
15
+ "weight_decay": 0.0,
16
+ "grad_clip": 1.0,
17
+ "val_loss_every": 250,
18
+ "val_max_steps": 20,
19
+ "sample_every": 0,
20
+ "overfit_single_batch": 0,
21
+ "shuffle_files": true,
22
+ "tensorcores": 1,
23
+ "device": "",
24
+ "compile": 1,
25
+ "flash": 1,
26
+ "dtype": "bfloat16",
27
+ "zero_stage": 1,
28
+ "optimizer": "adam",
29
+ "muon_lr": 0.01,
30
+ "muon_momentum": 0.95,
31
+ "muon_weight_decay": 0.0,
32
+ "muon_ns_steps": 5,
33
+ "muon_nesterov": false,
34
+ "write_tensors": 0,
35
+ "seed": 45,
36
+ "analyze_sharpness": false,
37
+ "sharpness_analysis_interval": 500
38
+ },
39
+ "run_uuid": "3f6fe53e-36b4-4df3-a931-fb8455b4635b",
40
+ "script_code_logged_at_start": true
41
+ }
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/model_final.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:19bb068cdc352bc65764df0f18bdede6e6fe3b8b5020f6b8de2fc9e253eb052c
3
+ size 544851409
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_45/training_log.txt ADDED
The diff for this file is too large to render. See raw diff
 
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/config.json ADDED
@@ -0,0 +1,41 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "cli_args": {
3
+ "input_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_train_*.bin",
4
+ "input_val_bin": "/home/aiops/zhangfz/MUON_sharpness/modded-nanogpt/data/fineweb10B/fineweb_val_*.bin",
5
+ "output_dir": "/home/aiops/zhangfz/MUON_LLM/logs_pure_qk_nonorm/test",
6
+ "model": "d12",
7
+ "batch_size": 4,
8
+ "sequence_length": 1024,
9
+ "total_batch_size": 524288,
10
+ "num_iterations": 10000,
11
+ "inference_only": 0,
12
+ "adam_lr": 0.001,
13
+ "warmup_iters": 700,
14
+ "lr_decay_frac": 0.0,
15
+ "weight_decay": 0.0,
16
+ "grad_clip": 1.0,
17
+ "val_loss_every": 250,
18
+ "val_max_steps": 20,
19
+ "sample_every": 0,
20
+ "overfit_single_batch": 0,
21
+ "shuffle_files": true,
22
+ "tensorcores": 1,
23
+ "device": "",
24
+ "compile": 1,
25
+ "flash": 1,
26
+ "dtype": "bfloat16",
27
+ "zero_stage": 1,
28
+ "optimizer": "adam",
29
+ "muon_lr": 0.01,
30
+ "muon_momentum": 0.95,
31
+ "muon_weight_decay": 0.0,
32
+ "muon_ns_steps": 5,
33
+ "muon_nesterov": false,
34
+ "write_tensors": 0,
35
+ "seed": 46,
36
+ "analyze_sharpness": false,
37
+ "sharpness_analysis_interval": 500
38
+ },
39
+ "run_uuid": "cfef5d9e-0e85-4ded-9299-5682c338126f",
40
+ "script_code_logged_at_start": true
41
+ }
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/model_final.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f603eeedc893ec6aef2407bd198a00a851c2a6bbadba0725117e28917f7e42d4
3
+ size 544851409
logs_pure_qk_nonorm/test/opt_adam_alr_0.001_mlr_0.01_seed_46/training_log.txt ADDED
The diff for this file is too large to render. See raw diff