w-ahmad commited on
Commit
54a48e8
Β·
verified Β·
1 Parent(s): 6308dab

Auto upload zain 2026-08-14T23:29:37.074222

Browse files
This view is limited to 50 files because it contains too many changes. Β  See raw diff
Files changed (50) hide show
  1. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/config.json +36 -0
  2. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/model.safetensors +3 -0
  3. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/optimizer.pt +3 -0
  4. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/rng_state.pth +3 -0
  5. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/scheduler.pt +3 -0
  6. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/tokenizer.json +0 -0
  7. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/tokenizer_config.json +13 -0
  8. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/trainer_state.json +1170 -0
  9. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/training_args.bin +3 -0
  10. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/config.json +36 -0
  11. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/model.safetensors +3 -0
  12. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/optimizer.pt +3 -0
  13. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/rng_state.pth +3 -0
  14. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/scheduler.pt +3 -0
  15. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/tokenizer.json +0 -0
  16. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/tokenizer_config.json +13 -0
  17. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/trainer_state.json +1205 -0
  18. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/training_args.bin +3 -0
  19. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/config.json +36 -0
  20. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/model.safetensors +3 -0
  21. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/optimizer.pt +3 -0
  22. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/rng_state.pth +3 -0
  23. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/scheduler.pt +3 -0
  24. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/tokenizer.json +0 -0
  25. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/tokenizer_config.json +13 -0
  26. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/trainer_state.json +1240 -0
  27. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/training_args.bin +3 -0
  28. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/config.json +36 -0
  29. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/model.safetensors +3 -0
  30. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/optimizer.pt +3 -0
  31. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/rng_state.pth +3 -0
  32. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/scheduler.pt +3 -0
  33. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/tokenizer.json +0 -0
  34. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/tokenizer_config.json +13 -0
  35. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/trainer_state.json +1283 -0
  36. zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/training_args.bin +3 -0
  37. zain/Activation/out/glu-silu-waleed10-100L_trash_run/config.json +36 -0
  38. zain/Activation/out/glu-silu-waleed10-100L_trash_run/model.safetensors +3 -0
  39. zain/Activation/out/glu-silu-waleed10-100L_trash_run/tokenizer.json +0 -0
  40. zain/Activation/out/glu-silu-waleed10-100L_trash_run/tokenizer_config.json +13 -0
  41. zain/Activation/out/glu-silu-waleed10-100L_trash_run/training_args.bin +3 -0
  42. zain/Activation/out/glu-silu-waleed10-100L_trash_run/training_log.jsonl +2 -2
  43. zain/Activation/out/sweep_summary.json +16 -0
  44. zain/Activation/wandb/debug-internal.log +25 -0
  45. zain/Activation/wandb/debug.log +5 -0
  46. zain/Activation/wandb/run-20260814_231802-8ih2zdoc/logs/debug-core.log +8 -0
  47. zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/config.yaml +436 -0
  48. zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/output.log +47 -1
  49. zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/wandb-summary.json +0 -0
  50. zain/Activation/wandb/run-20260814_232318-l56mdhxl/logs/debug-core.log +8 -0
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "silu-waleed10",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 0.5
36
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:b1610788881418ee8a182ef16f5f8e879fdaff5ba1906843eeac290daf36d25b
3
+ size 33967272
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d054acf5f934eb463dd2dd10137d78cd19c5fe44575698512aa34ead2eb5ebf6
3
+ size 39573050
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
3
+ size 14244
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:42032ce2906e08b8158bbc61ccff7a8449b0a1b6c135848ffd257fa664fd4577
3
+ size 1064
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/trainer_state.json ADDED
@@ -0,0 +1,1170 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.21570610043815303,
6
+ "eval_steps": 2498,
7
+ "global_step": 3200,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299632263183593,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047638702392579,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.837866973876953,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572817993164063,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233210754394531,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860719299316406,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.5177757263183596,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.80078125,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239912414550782,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.74609375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.035282897949219,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5546875,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.873902893066406,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.498046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.691160202026367,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.21875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.473793411254883,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.5078125,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.258899688720703,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.796875,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.03008918762207,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.84375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.797770309448242,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.5078125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.603690719604492,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.96484375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.490921401977539,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6640625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.327465057373047,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.86328125,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.182389831542968,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.82421875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 4.036349487304688,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.71484375,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.918545150756836,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.765625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.8555908203125,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7499755859375,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6484375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.668988800048828,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.8828125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.598476791381836,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.5625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.52923698425293,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.56640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.4486984252929687,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3981090545654298,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.5234375,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.337215042114258,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.48828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.3039676666259767,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.55078125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2369564056396483,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.470703125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.2108325958251953,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.153903007507324,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.412109375,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.125510597229004,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.44921875,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0803747177124023,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.04086971282959,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.458984375,
267
+ "learning_rate": 0.0007,
268
+ "loss": 3.0033853530883787,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.40625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9853662490844726,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.48828125,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9864330291748047,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.474609375,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9317838668823244,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.416015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.916472244262695,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.373046875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8858991622924806,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.439453125,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.840240478515625,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.8184703826904296,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.42578125,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.8047561645507812,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.4375,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.775454521179199,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.421875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7625925064086916,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.4296875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.731528663635254,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.71492977142334,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.40625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.7131521224975588,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.373046875,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6799833297729494,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.412109375,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.665292739868164,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.44140625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.6485418319702148,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.392578125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.6286544799804688,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.416015625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.621373176574707,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.5860437393188476,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.4453125,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.5712385177612305,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.380859375,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5693639755249023,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.3984375,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.563449668884277,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.369140625,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.553573417663574,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.3984375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.538028335571289,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.365234375,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.5232309341430663,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.373046875,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.50943546295166,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.373046875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4896902084350585,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.375,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4810691833496095,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.349609375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.471504974365234,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.3671875,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.458695411682129,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.400390625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4555185317993162,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.35546875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.433156204223633,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.34375,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.425444412231445,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.400390625,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.435191345214844,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.369140625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.412186622619629,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.375,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.4007726669311524,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.357421875,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3866371154785155,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.35546875,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3728200912475588,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.376953125,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.373748016357422,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.36328125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3635543823242187,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3516841888427735,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.376953125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3689512252807616,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.365234375,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.3348180770874025,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.361328125,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3343011856079103,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.37890625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3455299377441405,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.35546875,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3264875411987305,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.3984375,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.309524917602539,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.357421875,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.3181018829345703,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.412109375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2959054946899413,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.3515625,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.286634826660156,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3671875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.2912055969238283,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.349609375,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.303947639465332,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.33984375,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.298970413208008,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.373046875,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2899410247802736,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.3515625,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.258684539794922,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.365234375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.2575857162475588,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.33984375,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2479154586791994,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3515625,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2578588485717774,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.35546875,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.2332910537719726,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.242460250854492,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.359375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.21710205078125,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.353515625,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2389858245849608,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.328125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.2128574371337892,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.349609375,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.221720504760742,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.34765625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.2012025833129885,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.2207321166992187,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.3515625,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1970823287963865,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.3359375,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.1874614715576173,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1962921142578127,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.375,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1974483489990235,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.34375,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1824066162109377,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1886768341064453,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.326171875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1661350250244142,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.330078125,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1746242523193358,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.353515625,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.182226371765137,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.33203125,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1612625122070312,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.337890625,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1695093154907226,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.341796875,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1509073257446287,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.33984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1601842880249023,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.3203125,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.149628257751465,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.33203125,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.154475784301758,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.333984375,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1572238922119142,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.34765625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1351224899291994,
850
+ "step": 2400
851
+ },
852
+ {
853
+ "epoch": 0.16312773845635323,
854
+ "grad_norm": 0.345703125,
855
+ "learning_rate": 0.0007,
856
+ "loss": 2.141549301147461,
857
+ "step": 2420
858
+ },
859
+ {
860
+ "epoch": 0.16447590158409167,
861
+ "grad_norm": 0.328125,
862
+ "learning_rate": 0.0007,
863
+ "loss": 2.127903366088867,
864
+ "step": 2440
865
+ },
866
+ {
867
+ "epoch": 0.16582406471183014,
868
+ "grad_norm": 0.32421875,
869
+ "learning_rate": 0.0007,
870
+ "loss": 2.1384496688842773,
871
+ "step": 2460
872
+ },
873
+ {
874
+ "epoch": 0.1671722278395686,
875
+ "grad_norm": 0.34375,
876
+ "learning_rate": 0.0007,
877
+ "loss": 2.1402868270874023,
878
+ "step": 2480
879
+ },
880
+ {
881
+ "epoch": 0.1683855746545332,
882
+ "eval_loss": 2.133925199508667,
883
+ "eval_runtime": 18.9626,
884
+ "eval_samples_per_second": 502.41,
885
+ "eval_steps_per_second": 0.527,
886
+ "step": 2498
887
+ },
888
+ {
889
+ "epoch": 0.16852039096730703,
890
+ "grad_norm": 0.34765625,
891
+ "learning_rate": 0.0007,
892
+ "loss": 2.1217138290405275,
893
+ "step": 2500
894
+ },
895
+ {
896
+ "epoch": 0.16852039096730703,
897
+ "eval_loss": 2.1338183879852295,
898
+ "eval_runtime": 18.812,
899
+ "eval_samples_per_second": 506.431,
900
+ "eval_steps_per_second": 0.532,
901
+ "step": 2500
902
+ },
903
+ {
904
+ "epoch": 0.1698685540950455,
905
+ "grad_norm": 15335424.0,
906
+ "learning_rate": 3.4200000000000005e-05,
907
+ "loss": 2.1747163772583007,
908
+ "step": 2520
909
+ },
910
+ {
911
+ "epoch": 0.17121671722278395,
912
+ "grad_norm": 16646144.0,
913
+ "learning_rate": 7.020000000000001e-05,
914
+ "loss": 2.1868026733398436,
915
+ "step": 2540
916
+ },
917
+ {
918
+ "epoch": 0.17256488035052242,
919
+ "grad_norm": 102760448.0,
920
+ "learning_rate": 9e-05,
921
+ "loss": 3.535448455810547,
922
+ "step": 2560
923
+ },
924
+ {
925
+ "epoch": 0.17391304347826086,
926
+ "grad_norm": 188743680.0,
927
+ "learning_rate": 9e-05,
928
+ "loss": 7.000344848632812,
929
+ "step": 2580
930
+ },
931
+ {
932
+ "epoch": 0.17526120660599934,
933
+ "grad_norm": 186646528.0,
934
+ "learning_rate": 9e-05,
935
+ "loss": 7.602831268310547,
936
+ "step": 2600
937
+ },
938
+ {
939
+ "epoch": 0.17660936973373778,
940
+ "grad_norm": 110624768.0,
941
+ "learning_rate": 9e-05,
942
+ "loss": 7.876109313964844,
943
+ "step": 2620
944
+ },
945
+ {
946
+ "epoch": 0.17795753286147623,
947
+ "grad_norm": 52690944.0,
948
+ "learning_rate": 9e-05,
949
+ "loss": 8.0205078125,
950
+ "step": 2640
951
+ },
952
+ {
953
+ "epoch": 0.1793056959892147,
954
+ "grad_norm": 21102592.0,
955
+ "learning_rate": 9e-05,
956
+ "loss": 8.05886001586914,
957
+ "step": 2660
958
+ },
959
+ {
960
+ "epoch": 0.18065385911695314,
961
+ "grad_norm": 9830400.0,
962
+ "learning_rate": 9e-05,
963
+ "loss": 8.069260406494141,
964
+ "step": 2680
965
+ },
966
+ {
967
+ "epoch": 0.1820020222446916,
968
+ "grad_norm": 3670016.0,
969
+ "learning_rate": 9e-05,
970
+ "loss": 8.075779724121094,
971
+ "step": 2700
972
+ },
973
+ {
974
+ "epoch": 0.18335018537243006,
975
+ "grad_norm": 3571712.0,
976
+ "learning_rate": 9e-05,
977
+ "loss": 8.076358795166016,
978
+ "step": 2720
979
+ },
980
+ {
981
+ "epoch": 0.18469834850016853,
982
+ "grad_norm": 1114112.0,
983
+ "learning_rate": 9e-05,
984
+ "loss": 8.07742919921875,
985
+ "step": 2740
986
+ },
987
+ {
988
+ "epoch": 0.18604651162790697,
989
+ "grad_norm": 421888.0,
990
+ "learning_rate": 9e-05,
991
+ "loss": 8.07922592163086,
992
+ "step": 2760
993
+ },
994
+ {
995
+ "epoch": 0.18739467475564545,
996
+ "grad_norm": 129536.0,
997
+ "learning_rate": 9e-05,
998
+ "loss": 8.079435729980469,
999
+ "step": 2780
1000
+ },
1001
+ {
1002
+ "epoch": 0.1887428378833839,
1003
+ "grad_norm": 34048.0,
1004
+ "learning_rate": 9e-05,
1005
+ "loss": 8.079093170166015,
1006
+ "step": 2800
1007
+ },
1008
+ {
1009
+ "epoch": 0.19009100101112233,
1010
+ "grad_norm": 5888.0,
1011
+ "learning_rate": 9e-05,
1012
+ "loss": 8.07839126586914,
1013
+ "step": 2820
1014
+ },
1015
+ {
1016
+ "epoch": 0.1914391641388608,
1017
+ "grad_norm": 446.0,
1018
+ "learning_rate": 9e-05,
1019
+ "loss": 8.081797027587891,
1020
+ "step": 2840
1021
+ },
1022
+ {
1023
+ "epoch": 0.19278732726659925,
1024
+ "grad_norm": 10.25,
1025
+ "learning_rate": 9e-05,
1026
+ "loss": 8.079029083251953,
1027
+ "step": 2860
1028
+ },
1029
+ {
1030
+ "epoch": 0.19413549039433772,
1031
+ "grad_norm": 6.53125,
1032
+ "learning_rate": 9e-05,
1033
+ "loss": 8.081595611572265,
1034
+ "step": 2880
1035
+ },
1036
+ {
1037
+ "epoch": 0.19548365352207617,
1038
+ "grad_norm": 3.4375,
1039
+ "learning_rate": 9e-05,
1040
+ "loss": 8.079817962646484,
1041
+ "step": 2900
1042
+ },
1043
+ {
1044
+ "epoch": 0.19683181664981464,
1045
+ "grad_norm": 0.0,
1046
+ "learning_rate": 9e-05,
1047
+ "loss": 8.079161834716796,
1048
+ "step": 2920
1049
+ },
1050
+ {
1051
+ "epoch": 0.19817997977755308,
1052
+ "grad_norm": 1.3046875,
1053
+ "learning_rate": 9e-05,
1054
+ "loss": 8.07956314086914,
1055
+ "step": 2940
1056
+ },
1057
+ {
1058
+ "epoch": 0.19952814290529153,
1059
+ "grad_norm": 2.40625,
1060
+ "learning_rate": 9e-05,
1061
+ "loss": 8.074960327148437,
1062
+ "step": 2960
1063
+ },
1064
+ {
1065
+ "epoch": 0.20087630603303,
1066
+ "grad_norm": 81.5,
1067
+ "learning_rate": 9e-05,
1068
+ "loss": 8.076923370361328,
1069
+ "step": 2980
1070
+ },
1071
+ {
1072
+ "epoch": 0.20222446916076844,
1073
+ "grad_norm": 0.0,
1074
+ "learning_rate": 9e-05,
1075
+ "loss": 8.076215362548828,
1076
+ "step": 3000
1077
+ },
1078
+ {
1079
+ "epoch": 0.20357263228850692,
1080
+ "grad_norm": 0.0,
1081
+ "learning_rate": 9e-05,
1082
+ "loss": 8.080150604248047,
1083
+ "step": 3020
1084
+ },
1085
+ {
1086
+ "epoch": 0.20492079541624536,
1087
+ "grad_norm": 0.0,
1088
+ "learning_rate": 9e-05,
1089
+ "loss": 8.082421875,
1090
+ "step": 3040
1091
+ },
1092
+ {
1093
+ "epoch": 0.20626895854398383,
1094
+ "grad_norm": 0.0,
1095
+ "learning_rate": 9e-05,
1096
+ "loss": 8.076376342773438,
1097
+ "step": 3060
1098
+ },
1099
+ {
1100
+ "epoch": 0.20761712167172228,
1101
+ "grad_norm": 0.0,
1102
+ "learning_rate": 9e-05,
1103
+ "loss": 8.07555694580078,
1104
+ "step": 3080
1105
+ },
1106
+ {
1107
+ "epoch": 0.20896528479946072,
1108
+ "grad_norm": 1.03125,
1109
+ "learning_rate": 9e-05,
1110
+ "loss": 8.079136657714844,
1111
+ "step": 3100
1112
+ },
1113
+ {
1114
+ "epoch": 0.2103134479271992,
1115
+ "grad_norm": 0.0,
1116
+ "learning_rate": 9e-05,
1117
+ "loss": 8.081492614746093,
1118
+ "step": 3120
1119
+ },
1120
+ {
1121
+ "epoch": 0.21166161105493764,
1122
+ "grad_norm": 0.0,
1123
+ "learning_rate": 9e-05,
1124
+ "loss": 8.079338073730469,
1125
+ "step": 3140
1126
+ },
1127
+ {
1128
+ "epoch": 0.2130097741826761,
1129
+ "grad_norm": 138.0,
1130
+ "learning_rate": 9e-05,
1131
+ "loss": 8.077278137207031,
1132
+ "step": 3160
1133
+ },
1134
+ {
1135
+ "epoch": 0.21435793731041455,
1136
+ "grad_norm": 1.109375,
1137
+ "learning_rate": 9e-05,
1138
+ "loss": 8.075982666015625,
1139
+ "step": 3180
1140
+ },
1141
+ {
1142
+ "epoch": 0.21570610043815303,
1143
+ "grad_norm": 0.0,
1144
+ "learning_rate": 9e-05,
1145
+ "loss": 8.079740142822265,
1146
+ "step": 3200
1147
+ }
1148
+ ],
1149
+ "logging_steps": 20,
1150
+ "max_steps": 3500,
1151
+ "num_input_tokens_seen": 0,
1152
+ "num_train_epochs": 1,
1153
+ "save_steps": 100,
1154
+ "stateful_callbacks": {
1155
+ "TrainerControl": {
1156
+ "args": {
1157
+ "should_epoch_stop": false,
1158
+ "should_evaluate": false,
1159
+ "should_log": false,
1160
+ "should_save": true,
1161
+ "should_training_stop": false
1162
+ },
1163
+ "attributes": {}
1164
+ }
1165
+ },
1166
+ "total_flos": 1.03241081683968e+16,
1167
+ "train_batch_size": 64,
1168
+ "trial_name": null,
1169
+ "trial_params": null
1170
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3200/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e9b362740bb60edd1ea1474e4e02705aa61c5177a0f605723687fc76ff5105
3
+ size 4920
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "silu-waleed10",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 0.5
36
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:7ae083662cd7104aaed0c0b9f4ccb477b37c2c78fcaf32413d3282941abebb5e
3
+ size 33967272
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3821ed9a2fa313c85d90a869d0d4da0ba9f690d29d0046f688abdfbdc154f891
3
+ size 39573050
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
3
+ size 14244
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9751c40def32406332f8de58ba93b20f2a7e67943c9e04a0e73992386f74e398
3
+ size 1064
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/trainer_state.json ADDED
@@ -0,0 +1,1205 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.2224469160768453,
6
+ "eval_steps": 2498,
7
+ "global_step": 3300,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299632263183593,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047638702392579,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.837866973876953,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572817993164063,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233210754394531,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860719299316406,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.5177757263183596,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.80078125,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239912414550782,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.74609375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.035282897949219,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5546875,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.873902893066406,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.498046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.691160202026367,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.21875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.473793411254883,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.5078125,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.258899688720703,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.796875,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.03008918762207,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.84375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.797770309448242,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.5078125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.603690719604492,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.96484375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.490921401977539,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6640625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.327465057373047,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.86328125,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.182389831542968,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.82421875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 4.036349487304688,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.71484375,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.918545150756836,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.765625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.8555908203125,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7499755859375,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6484375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.668988800048828,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.8828125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.598476791381836,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.5625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.52923698425293,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.56640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.4486984252929687,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3981090545654298,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.5234375,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.337215042114258,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.48828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.3039676666259767,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.55078125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2369564056396483,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.470703125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.2108325958251953,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.153903007507324,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.412109375,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.125510597229004,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.44921875,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0803747177124023,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.04086971282959,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.458984375,
267
+ "learning_rate": 0.0007,
268
+ "loss": 3.0033853530883787,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.40625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9853662490844726,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.48828125,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9864330291748047,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.474609375,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9317838668823244,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.416015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.916472244262695,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.373046875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8858991622924806,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.439453125,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.840240478515625,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.8184703826904296,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.42578125,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.8047561645507812,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.4375,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.775454521179199,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.421875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7625925064086916,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.4296875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.731528663635254,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.71492977142334,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.40625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.7131521224975588,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.373046875,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6799833297729494,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.412109375,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.665292739868164,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.44140625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.6485418319702148,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.392578125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.6286544799804688,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.416015625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.621373176574707,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.5860437393188476,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.4453125,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.5712385177612305,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.380859375,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5693639755249023,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.3984375,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.563449668884277,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.369140625,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.553573417663574,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.3984375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.538028335571289,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.365234375,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.5232309341430663,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.373046875,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.50943546295166,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.373046875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4896902084350585,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.375,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4810691833496095,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.349609375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.471504974365234,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.3671875,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.458695411682129,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.400390625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4555185317993162,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.35546875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.433156204223633,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.34375,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.425444412231445,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.400390625,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.435191345214844,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.369140625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.412186622619629,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.375,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.4007726669311524,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.357421875,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3866371154785155,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.35546875,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3728200912475588,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.376953125,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.373748016357422,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.36328125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3635543823242187,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3516841888427735,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.376953125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3689512252807616,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.365234375,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.3348180770874025,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.361328125,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3343011856079103,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.37890625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3455299377441405,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.35546875,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3264875411987305,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.3984375,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.309524917602539,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.357421875,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.3181018829345703,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.412109375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2959054946899413,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.3515625,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.286634826660156,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3671875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.2912055969238283,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.349609375,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.303947639465332,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.33984375,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.298970413208008,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.373046875,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2899410247802736,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.3515625,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.258684539794922,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.365234375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.2575857162475588,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.33984375,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2479154586791994,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3515625,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2578588485717774,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.35546875,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.2332910537719726,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.242460250854492,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.359375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.21710205078125,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.353515625,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2389858245849608,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.328125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.2128574371337892,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.349609375,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.221720504760742,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.34765625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.2012025833129885,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.2207321166992187,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.3515625,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1970823287963865,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.3359375,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.1874614715576173,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1962921142578127,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.375,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1974483489990235,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.34375,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1824066162109377,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1886768341064453,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.326171875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1661350250244142,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.330078125,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1746242523193358,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.353515625,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.182226371765137,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.33203125,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1612625122070312,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.337890625,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1695093154907226,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.341796875,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1509073257446287,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.33984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1601842880249023,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.3203125,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.149628257751465,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.33203125,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.154475784301758,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.333984375,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1572238922119142,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.34765625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1351224899291994,
850
+ "step": 2400
851
+ },
852
+ {
853
+ "epoch": 0.16312773845635323,
854
+ "grad_norm": 0.345703125,
855
+ "learning_rate": 0.0007,
856
+ "loss": 2.141549301147461,
857
+ "step": 2420
858
+ },
859
+ {
860
+ "epoch": 0.16447590158409167,
861
+ "grad_norm": 0.328125,
862
+ "learning_rate": 0.0007,
863
+ "loss": 2.127903366088867,
864
+ "step": 2440
865
+ },
866
+ {
867
+ "epoch": 0.16582406471183014,
868
+ "grad_norm": 0.32421875,
869
+ "learning_rate": 0.0007,
870
+ "loss": 2.1384496688842773,
871
+ "step": 2460
872
+ },
873
+ {
874
+ "epoch": 0.1671722278395686,
875
+ "grad_norm": 0.34375,
876
+ "learning_rate": 0.0007,
877
+ "loss": 2.1402868270874023,
878
+ "step": 2480
879
+ },
880
+ {
881
+ "epoch": 0.1683855746545332,
882
+ "eval_loss": 2.133925199508667,
883
+ "eval_runtime": 18.9626,
884
+ "eval_samples_per_second": 502.41,
885
+ "eval_steps_per_second": 0.527,
886
+ "step": 2498
887
+ },
888
+ {
889
+ "epoch": 0.16852039096730703,
890
+ "grad_norm": 0.34765625,
891
+ "learning_rate": 0.0007,
892
+ "loss": 2.1217138290405275,
893
+ "step": 2500
894
+ },
895
+ {
896
+ "epoch": 0.16852039096730703,
897
+ "eval_loss": 2.1338183879852295,
898
+ "eval_runtime": 18.812,
899
+ "eval_samples_per_second": 506.431,
900
+ "eval_steps_per_second": 0.532,
901
+ "step": 2500
902
+ },
903
+ {
904
+ "epoch": 0.1698685540950455,
905
+ "grad_norm": 15335424.0,
906
+ "learning_rate": 3.4200000000000005e-05,
907
+ "loss": 2.1747163772583007,
908
+ "step": 2520
909
+ },
910
+ {
911
+ "epoch": 0.17121671722278395,
912
+ "grad_norm": 16646144.0,
913
+ "learning_rate": 7.020000000000001e-05,
914
+ "loss": 2.1868026733398436,
915
+ "step": 2540
916
+ },
917
+ {
918
+ "epoch": 0.17256488035052242,
919
+ "grad_norm": 102760448.0,
920
+ "learning_rate": 9e-05,
921
+ "loss": 3.535448455810547,
922
+ "step": 2560
923
+ },
924
+ {
925
+ "epoch": 0.17391304347826086,
926
+ "grad_norm": 188743680.0,
927
+ "learning_rate": 9e-05,
928
+ "loss": 7.000344848632812,
929
+ "step": 2580
930
+ },
931
+ {
932
+ "epoch": 0.17526120660599934,
933
+ "grad_norm": 186646528.0,
934
+ "learning_rate": 9e-05,
935
+ "loss": 7.602831268310547,
936
+ "step": 2600
937
+ },
938
+ {
939
+ "epoch": 0.17660936973373778,
940
+ "grad_norm": 110624768.0,
941
+ "learning_rate": 9e-05,
942
+ "loss": 7.876109313964844,
943
+ "step": 2620
944
+ },
945
+ {
946
+ "epoch": 0.17795753286147623,
947
+ "grad_norm": 52690944.0,
948
+ "learning_rate": 9e-05,
949
+ "loss": 8.0205078125,
950
+ "step": 2640
951
+ },
952
+ {
953
+ "epoch": 0.1793056959892147,
954
+ "grad_norm": 21102592.0,
955
+ "learning_rate": 9e-05,
956
+ "loss": 8.05886001586914,
957
+ "step": 2660
958
+ },
959
+ {
960
+ "epoch": 0.18065385911695314,
961
+ "grad_norm": 9830400.0,
962
+ "learning_rate": 9e-05,
963
+ "loss": 8.069260406494141,
964
+ "step": 2680
965
+ },
966
+ {
967
+ "epoch": 0.1820020222446916,
968
+ "grad_norm": 3670016.0,
969
+ "learning_rate": 9e-05,
970
+ "loss": 8.075779724121094,
971
+ "step": 2700
972
+ },
973
+ {
974
+ "epoch": 0.18335018537243006,
975
+ "grad_norm": 3571712.0,
976
+ "learning_rate": 9e-05,
977
+ "loss": 8.076358795166016,
978
+ "step": 2720
979
+ },
980
+ {
981
+ "epoch": 0.18469834850016853,
982
+ "grad_norm": 1114112.0,
983
+ "learning_rate": 9e-05,
984
+ "loss": 8.07742919921875,
985
+ "step": 2740
986
+ },
987
+ {
988
+ "epoch": 0.18604651162790697,
989
+ "grad_norm": 421888.0,
990
+ "learning_rate": 9e-05,
991
+ "loss": 8.07922592163086,
992
+ "step": 2760
993
+ },
994
+ {
995
+ "epoch": 0.18739467475564545,
996
+ "grad_norm": 129536.0,
997
+ "learning_rate": 9e-05,
998
+ "loss": 8.079435729980469,
999
+ "step": 2780
1000
+ },
1001
+ {
1002
+ "epoch": 0.1887428378833839,
1003
+ "grad_norm": 34048.0,
1004
+ "learning_rate": 9e-05,
1005
+ "loss": 8.079093170166015,
1006
+ "step": 2800
1007
+ },
1008
+ {
1009
+ "epoch": 0.19009100101112233,
1010
+ "grad_norm": 5888.0,
1011
+ "learning_rate": 9e-05,
1012
+ "loss": 8.07839126586914,
1013
+ "step": 2820
1014
+ },
1015
+ {
1016
+ "epoch": 0.1914391641388608,
1017
+ "grad_norm": 446.0,
1018
+ "learning_rate": 9e-05,
1019
+ "loss": 8.081797027587891,
1020
+ "step": 2840
1021
+ },
1022
+ {
1023
+ "epoch": 0.19278732726659925,
1024
+ "grad_norm": 10.25,
1025
+ "learning_rate": 9e-05,
1026
+ "loss": 8.079029083251953,
1027
+ "step": 2860
1028
+ },
1029
+ {
1030
+ "epoch": 0.19413549039433772,
1031
+ "grad_norm": 6.53125,
1032
+ "learning_rate": 9e-05,
1033
+ "loss": 8.081595611572265,
1034
+ "step": 2880
1035
+ },
1036
+ {
1037
+ "epoch": 0.19548365352207617,
1038
+ "grad_norm": 3.4375,
1039
+ "learning_rate": 9e-05,
1040
+ "loss": 8.079817962646484,
1041
+ "step": 2900
1042
+ },
1043
+ {
1044
+ "epoch": 0.19683181664981464,
1045
+ "grad_norm": 0.0,
1046
+ "learning_rate": 9e-05,
1047
+ "loss": 8.079161834716796,
1048
+ "step": 2920
1049
+ },
1050
+ {
1051
+ "epoch": 0.19817997977755308,
1052
+ "grad_norm": 1.3046875,
1053
+ "learning_rate": 9e-05,
1054
+ "loss": 8.07956314086914,
1055
+ "step": 2940
1056
+ },
1057
+ {
1058
+ "epoch": 0.19952814290529153,
1059
+ "grad_norm": 2.40625,
1060
+ "learning_rate": 9e-05,
1061
+ "loss": 8.074960327148437,
1062
+ "step": 2960
1063
+ },
1064
+ {
1065
+ "epoch": 0.20087630603303,
1066
+ "grad_norm": 81.5,
1067
+ "learning_rate": 9e-05,
1068
+ "loss": 8.076923370361328,
1069
+ "step": 2980
1070
+ },
1071
+ {
1072
+ "epoch": 0.20222446916076844,
1073
+ "grad_norm": 0.0,
1074
+ "learning_rate": 9e-05,
1075
+ "loss": 8.076215362548828,
1076
+ "step": 3000
1077
+ },
1078
+ {
1079
+ "epoch": 0.20357263228850692,
1080
+ "grad_norm": 0.0,
1081
+ "learning_rate": 9e-05,
1082
+ "loss": 8.080150604248047,
1083
+ "step": 3020
1084
+ },
1085
+ {
1086
+ "epoch": 0.20492079541624536,
1087
+ "grad_norm": 0.0,
1088
+ "learning_rate": 9e-05,
1089
+ "loss": 8.082421875,
1090
+ "step": 3040
1091
+ },
1092
+ {
1093
+ "epoch": 0.20626895854398383,
1094
+ "grad_norm": 0.0,
1095
+ "learning_rate": 9e-05,
1096
+ "loss": 8.076376342773438,
1097
+ "step": 3060
1098
+ },
1099
+ {
1100
+ "epoch": 0.20761712167172228,
1101
+ "grad_norm": 0.0,
1102
+ "learning_rate": 9e-05,
1103
+ "loss": 8.07555694580078,
1104
+ "step": 3080
1105
+ },
1106
+ {
1107
+ "epoch": 0.20896528479946072,
1108
+ "grad_norm": 1.03125,
1109
+ "learning_rate": 9e-05,
1110
+ "loss": 8.079136657714844,
1111
+ "step": 3100
1112
+ },
1113
+ {
1114
+ "epoch": 0.2103134479271992,
1115
+ "grad_norm": 0.0,
1116
+ "learning_rate": 9e-05,
1117
+ "loss": 8.081492614746093,
1118
+ "step": 3120
1119
+ },
1120
+ {
1121
+ "epoch": 0.21166161105493764,
1122
+ "grad_norm": 0.0,
1123
+ "learning_rate": 9e-05,
1124
+ "loss": 8.079338073730469,
1125
+ "step": 3140
1126
+ },
1127
+ {
1128
+ "epoch": 0.2130097741826761,
1129
+ "grad_norm": 138.0,
1130
+ "learning_rate": 9e-05,
1131
+ "loss": 8.077278137207031,
1132
+ "step": 3160
1133
+ },
1134
+ {
1135
+ "epoch": 0.21435793731041455,
1136
+ "grad_norm": 1.109375,
1137
+ "learning_rate": 9e-05,
1138
+ "loss": 8.075982666015625,
1139
+ "step": 3180
1140
+ },
1141
+ {
1142
+ "epoch": 0.21570610043815303,
1143
+ "grad_norm": 0.0,
1144
+ "learning_rate": 9e-05,
1145
+ "loss": 8.079740142822265,
1146
+ "step": 3200
1147
+ },
1148
+ {
1149
+ "epoch": 0.21705426356589147,
1150
+ "grad_norm": 2.34375,
1151
+ "learning_rate": 9e-05,
1152
+ "loss": 8.080359649658202,
1153
+ "step": 3220
1154
+ },
1155
+ {
1156
+ "epoch": 0.21840242669362994,
1157
+ "grad_norm": 0.0,
1158
+ "learning_rate": 9e-05,
1159
+ "loss": 8.0783935546875,
1160
+ "step": 3240
1161
+ },
1162
+ {
1163
+ "epoch": 0.2197505898213684,
1164
+ "grad_norm": 0.0,
1165
+ "learning_rate": 9e-05,
1166
+ "loss": 8.079434204101563,
1167
+ "step": 3260
1168
+ },
1169
+ {
1170
+ "epoch": 0.22109875294910683,
1171
+ "grad_norm": 0.0,
1172
+ "learning_rate": 9e-05,
1173
+ "loss": 8.079434204101563,
1174
+ "step": 3280
1175
+ },
1176
+ {
1177
+ "epoch": 0.2224469160768453,
1178
+ "grad_norm": 1.859375,
1179
+ "learning_rate": 9e-05,
1180
+ "loss": 8.07918930053711,
1181
+ "step": 3300
1182
+ }
1183
+ ],
1184
+ "logging_steps": 20,
1185
+ "max_steps": 3500,
1186
+ "num_input_tokens_seen": 0,
1187
+ "num_train_epochs": 1,
1188
+ "save_steps": 100,
1189
+ "stateful_callbacks": {
1190
+ "TrainerControl": {
1191
+ "args": {
1192
+ "should_epoch_stop": false,
1193
+ "should_evaluate": false,
1194
+ "should_log": false,
1195
+ "should_save": true,
1196
+ "should_training_stop": false
1197
+ },
1198
+ "attributes": {}
1199
+ }
1200
+ },
1201
+ "total_flos": 1.06467365486592e+16,
1202
+ "train_batch_size": 64,
1203
+ "trial_name": null,
1204
+ "trial_params": null
1205
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3300/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e9b362740bb60edd1ea1474e4e02705aa61c5177a0f605723687fc76ff5105
3
+ size 4920
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "silu-waleed10",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 0.5
36
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:5092444456e6dc6cd1a84582134363fa82e44390bfc179dc623d9088dcaf2a65
3
+ size 33967272
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:11a879807648f3fde9705c5e760acf31c121067ff8f86779650a6f98e59c05b7
3
+ size 39573050
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:3cf9097d4513154245c48236b6ec5137b7ee2a21c9f58f2cba798ea275c6026f
3
+ size 14244
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:d56f18bbf80d27667dff830c4bf2d569d657317bf85c1271d4a355eda1cf7318
3
+ size 1064
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/trainer_state.json ADDED
@@ -0,0 +1,1240 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.22918773171553758,
6
+ "eval_steps": 2498,
7
+ "global_step": 3400,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299632263183593,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047638702392579,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.837866973876953,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572817993164063,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233210754394531,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860719299316406,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.5177757263183596,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.80078125,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239912414550782,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.74609375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.035282897949219,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5546875,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.873902893066406,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.498046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.691160202026367,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.21875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.473793411254883,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.5078125,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.258899688720703,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.796875,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.03008918762207,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.84375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.797770309448242,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.5078125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.603690719604492,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.96484375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.490921401977539,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6640625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.327465057373047,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.86328125,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.182389831542968,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.82421875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 4.036349487304688,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.71484375,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.918545150756836,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.765625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.8555908203125,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7499755859375,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6484375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.668988800048828,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.8828125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.598476791381836,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.5625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.52923698425293,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.56640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.4486984252929687,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3981090545654298,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.5234375,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.337215042114258,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.48828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.3039676666259767,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.55078125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2369564056396483,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.470703125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.2108325958251953,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.153903007507324,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.412109375,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.125510597229004,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.44921875,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0803747177124023,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.04086971282959,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.458984375,
267
+ "learning_rate": 0.0007,
268
+ "loss": 3.0033853530883787,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.40625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9853662490844726,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.48828125,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9864330291748047,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.474609375,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9317838668823244,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.416015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.916472244262695,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.373046875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8858991622924806,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.439453125,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.840240478515625,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.8184703826904296,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.42578125,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.8047561645507812,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.4375,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.775454521179199,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.421875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7625925064086916,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.4296875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.731528663635254,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.71492977142334,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.40625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.7131521224975588,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.373046875,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6799833297729494,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.412109375,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.665292739868164,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.44140625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.6485418319702148,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.392578125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.6286544799804688,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.416015625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.621373176574707,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.5860437393188476,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.4453125,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.5712385177612305,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.380859375,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5693639755249023,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.3984375,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.563449668884277,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.369140625,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.553573417663574,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.3984375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.538028335571289,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.365234375,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.5232309341430663,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.373046875,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.50943546295166,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.373046875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4896902084350585,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.375,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4810691833496095,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.349609375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.471504974365234,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.3671875,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.458695411682129,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.400390625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4555185317993162,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.35546875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.433156204223633,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.34375,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.425444412231445,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.400390625,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.435191345214844,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.369140625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.412186622619629,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.375,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.4007726669311524,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.357421875,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3866371154785155,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.35546875,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3728200912475588,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.376953125,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.373748016357422,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.36328125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3635543823242187,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3516841888427735,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.376953125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3689512252807616,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.365234375,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.3348180770874025,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.361328125,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3343011856079103,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.37890625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3455299377441405,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.35546875,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3264875411987305,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.3984375,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.309524917602539,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.357421875,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.3181018829345703,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.412109375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2959054946899413,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.3515625,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.286634826660156,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3671875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.2912055969238283,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.349609375,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.303947639465332,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.33984375,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.298970413208008,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.373046875,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2899410247802736,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.3515625,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.258684539794922,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.365234375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.2575857162475588,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.33984375,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2479154586791994,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3515625,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2578588485717774,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.35546875,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.2332910537719726,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.242460250854492,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.359375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.21710205078125,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.353515625,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2389858245849608,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.328125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.2128574371337892,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.349609375,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.221720504760742,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.34765625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.2012025833129885,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.2207321166992187,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.3515625,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1970823287963865,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.3359375,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.1874614715576173,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1962921142578127,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.375,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1974483489990235,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.34375,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1824066162109377,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1886768341064453,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.326171875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1661350250244142,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.330078125,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1746242523193358,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.353515625,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.182226371765137,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.33203125,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1612625122070312,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.337890625,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1695093154907226,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.341796875,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1509073257446287,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.33984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1601842880249023,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.3203125,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.149628257751465,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.33203125,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.154475784301758,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.333984375,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1572238922119142,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.34765625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1351224899291994,
850
+ "step": 2400
851
+ },
852
+ {
853
+ "epoch": 0.16312773845635323,
854
+ "grad_norm": 0.345703125,
855
+ "learning_rate": 0.0007,
856
+ "loss": 2.141549301147461,
857
+ "step": 2420
858
+ },
859
+ {
860
+ "epoch": 0.16447590158409167,
861
+ "grad_norm": 0.328125,
862
+ "learning_rate": 0.0007,
863
+ "loss": 2.127903366088867,
864
+ "step": 2440
865
+ },
866
+ {
867
+ "epoch": 0.16582406471183014,
868
+ "grad_norm": 0.32421875,
869
+ "learning_rate": 0.0007,
870
+ "loss": 2.1384496688842773,
871
+ "step": 2460
872
+ },
873
+ {
874
+ "epoch": 0.1671722278395686,
875
+ "grad_norm": 0.34375,
876
+ "learning_rate": 0.0007,
877
+ "loss": 2.1402868270874023,
878
+ "step": 2480
879
+ },
880
+ {
881
+ "epoch": 0.1683855746545332,
882
+ "eval_loss": 2.133925199508667,
883
+ "eval_runtime": 18.9626,
884
+ "eval_samples_per_second": 502.41,
885
+ "eval_steps_per_second": 0.527,
886
+ "step": 2498
887
+ },
888
+ {
889
+ "epoch": 0.16852039096730703,
890
+ "grad_norm": 0.34765625,
891
+ "learning_rate": 0.0007,
892
+ "loss": 2.1217138290405275,
893
+ "step": 2500
894
+ },
895
+ {
896
+ "epoch": 0.16852039096730703,
897
+ "eval_loss": 2.1338183879852295,
898
+ "eval_runtime": 18.812,
899
+ "eval_samples_per_second": 506.431,
900
+ "eval_steps_per_second": 0.532,
901
+ "step": 2500
902
+ },
903
+ {
904
+ "epoch": 0.1698685540950455,
905
+ "grad_norm": 15335424.0,
906
+ "learning_rate": 3.4200000000000005e-05,
907
+ "loss": 2.1747163772583007,
908
+ "step": 2520
909
+ },
910
+ {
911
+ "epoch": 0.17121671722278395,
912
+ "grad_norm": 16646144.0,
913
+ "learning_rate": 7.020000000000001e-05,
914
+ "loss": 2.1868026733398436,
915
+ "step": 2540
916
+ },
917
+ {
918
+ "epoch": 0.17256488035052242,
919
+ "grad_norm": 102760448.0,
920
+ "learning_rate": 9e-05,
921
+ "loss": 3.535448455810547,
922
+ "step": 2560
923
+ },
924
+ {
925
+ "epoch": 0.17391304347826086,
926
+ "grad_norm": 188743680.0,
927
+ "learning_rate": 9e-05,
928
+ "loss": 7.000344848632812,
929
+ "step": 2580
930
+ },
931
+ {
932
+ "epoch": 0.17526120660599934,
933
+ "grad_norm": 186646528.0,
934
+ "learning_rate": 9e-05,
935
+ "loss": 7.602831268310547,
936
+ "step": 2600
937
+ },
938
+ {
939
+ "epoch": 0.17660936973373778,
940
+ "grad_norm": 110624768.0,
941
+ "learning_rate": 9e-05,
942
+ "loss": 7.876109313964844,
943
+ "step": 2620
944
+ },
945
+ {
946
+ "epoch": 0.17795753286147623,
947
+ "grad_norm": 52690944.0,
948
+ "learning_rate": 9e-05,
949
+ "loss": 8.0205078125,
950
+ "step": 2640
951
+ },
952
+ {
953
+ "epoch": 0.1793056959892147,
954
+ "grad_norm": 21102592.0,
955
+ "learning_rate": 9e-05,
956
+ "loss": 8.05886001586914,
957
+ "step": 2660
958
+ },
959
+ {
960
+ "epoch": 0.18065385911695314,
961
+ "grad_norm": 9830400.0,
962
+ "learning_rate": 9e-05,
963
+ "loss": 8.069260406494141,
964
+ "step": 2680
965
+ },
966
+ {
967
+ "epoch": 0.1820020222446916,
968
+ "grad_norm": 3670016.0,
969
+ "learning_rate": 9e-05,
970
+ "loss": 8.075779724121094,
971
+ "step": 2700
972
+ },
973
+ {
974
+ "epoch": 0.18335018537243006,
975
+ "grad_norm": 3571712.0,
976
+ "learning_rate": 9e-05,
977
+ "loss": 8.076358795166016,
978
+ "step": 2720
979
+ },
980
+ {
981
+ "epoch": 0.18469834850016853,
982
+ "grad_norm": 1114112.0,
983
+ "learning_rate": 9e-05,
984
+ "loss": 8.07742919921875,
985
+ "step": 2740
986
+ },
987
+ {
988
+ "epoch": 0.18604651162790697,
989
+ "grad_norm": 421888.0,
990
+ "learning_rate": 9e-05,
991
+ "loss": 8.07922592163086,
992
+ "step": 2760
993
+ },
994
+ {
995
+ "epoch": 0.18739467475564545,
996
+ "grad_norm": 129536.0,
997
+ "learning_rate": 9e-05,
998
+ "loss": 8.079435729980469,
999
+ "step": 2780
1000
+ },
1001
+ {
1002
+ "epoch": 0.1887428378833839,
1003
+ "grad_norm": 34048.0,
1004
+ "learning_rate": 9e-05,
1005
+ "loss": 8.079093170166015,
1006
+ "step": 2800
1007
+ },
1008
+ {
1009
+ "epoch": 0.19009100101112233,
1010
+ "grad_norm": 5888.0,
1011
+ "learning_rate": 9e-05,
1012
+ "loss": 8.07839126586914,
1013
+ "step": 2820
1014
+ },
1015
+ {
1016
+ "epoch": 0.1914391641388608,
1017
+ "grad_norm": 446.0,
1018
+ "learning_rate": 9e-05,
1019
+ "loss": 8.081797027587891,
1020
+ "step": 2840
1021
+ },
1022
+ {
1023
+ "epoch": 0.19278732726659925,
1024
+ "grad_norm": 10.25,
1025
+ "learning_rate": 9e-05,
1026
+ "loss": 8.079029083251953,
1027
+ "step": 2860
1028
+ },
1029
+ {
1030
+ "epoch": 0.19413549039433772,
1031
+ "grad_norm": 6.53125,
1032
+ "learning_rate": 9e-05,
1033
+ "loss": 8.081595611572265,
1034
+ "step": 2880
1035
+ },
1036
+ {
1037
+ "epoch": 0.19548365352207617,
1038
+ "grad_norm": 3.4375,
1039
+ "learning_rate": 9e-05,
1040
+ "loss": 8.079817962646484,
1041
+ "step": 2900
1042
+ },
1043
+ {
1044
+ "epoch": 0.19683181664981464,
1045
+ "grad_norm": 0.0,
1046
+ "learning_rate": 9e-05,
1047
+ "loss": 8.079161834716796,
1048
+ "step": 2920
1049
+ },
1050
+ {
1051
+ "epoch": 0.19817997977755308,
1052
+ "grad_norm": 1.3046875,
1053
+ "learning_rate": 9e-05,
1054
+ "loss": 8.07956314086914,
1055
+ "step": 2940
1056
+ },
1057
+ {
1058
+ "epoch": 0.19952814290529153,
1059
+ "grad_norm": 2.40625,
1060
+ "learning_rate": 9e-05,
1061
+ "loss": 8.074960327148437,
1062
+ "step": 2960
1063
+ },
1064
+ {
1065
+ "epoch": 0.20087630603303,
1066
+ "grad_norm": 81.5,
1067
+ "learning_rate": 9e-05,
1068
+ "loss": 8.076923370361328,
1069
+ "step": 2980
1070
+ },
1071
+ {
1072
+ "epoch": 0.20222446916076844,
1073
+ "grad_norm": 0.0,
1074
+ "learning_rate": 9e-05,
1075
+ "loss": 8.076215362548828,
1076
+ "step": 3000
1077
+ },
1078
+ {
1079
+ "epoch": 0.20357263228850692,
1080
+ "grad_norm": 0.0,
1081
+ "learning_rate": 9e-05,
1082
+ "loss": 8.080150604248047,
1083
+ "step": 3020
1084
+ },
1085
+ {
1086
+ "epoch": 0.20492079541624536,
1087
+ "grad_norm": 0.0,
1088
+ "learning_rate": 9e-05,
1089
+ "loss": 8.082421875,
1090
+ "step": 3040
1091
+ },
1092
+ {
1093
+ "epoch": 0.20626895854398383,
1094
+ "grad_norm": 0.0,
1095
+ "learning_rate": 9e-05,
1096
+ "loss": 8.076376342773438,
1097
+ "step": 3060
1098
+ },
1099
+ {
1100
+ "epoch": 0.20761712167172228,
1101
+ "grad_norm": 0.0,
1102
+ "learning_rate": 9e-05,
1103
+ "loss": 8.07555694580078,
1104
+ "step": 3080
1105
+ },
1106
+ {
1107
+ "epoch": 0.20896528479946072,
1108
+ "grad_norm": 1.03125,
1109
+ "learning_rate": 9e-05,
1110
+ "loss": 8.079136657714844,
1111
+ "step": 3100
1112
+ },
1113
+ {
1114
+ "epoch": 0.2103134479271992,
1115
+ "grad_norm": 0.0,
1116
+ "learning_rate": 9e-05,
1117
+ "loss": 8.081492614746093,
1118
+ "step": 3120
1119
+ },
1120
+ {
1121
+ "epoch": 0.21166161105493764,
1122
+ "grad_norm": 0.0,
1123
+ "learning_rate": 9e-05,
1124
+ "loss": 8.079338073730469,
1125
+ "step": 3140
1126
+ },
1127
+ {
1128
+ "epoch": 0.2130097741826761,
1129
+ "grad_norm": 138.0,
1130
+ "learning_rate": 9e-05,
1131
+ "loss": 8.077278137207031,
1132
+ "step": 3160
1133
+ },
1134
+ {
1135
+ "epoch": 0.21435793731041455,
1136
+ "grad_norm": 1.109375,
1137
+ "learning_rate": 9e-05,
1138
+ "loss": 8.075982666015625,
1139
+ "step": 3180
1140
+ },
1141
+ {
1142
+ "epoch": 0.21570610043815303,
1143
+ "grad_norm": 0.0,
1144
+ "learning_rate": 9e-05,
1145
+ "loss": 8.079740142822265,
1146
+ "step": 3200
1147
+ },
1148
+ {
1149
+ "epoch": 0.21705426356589147,
1150
+ "grad_norm": 2.34375,
1151
+ "learning_rate": 9e-05,
1152
+ "loss": 8.080359649658202,
1153
+ "step": 3220
1154
+ },
1155
+ {
1156
+ "epoch": 0.21840242669362994,
1157
+ "grad_norm": 0.0,
1158
+ "learning_rate": 9e-05,
1159
+ "loss": 8.0783935546875,
1160
+ "step": 3240
1161
+ },
1162
+ {
1163
+ "epoch": 0.2197505898213684,
1164
+ "grad_norm": 0.0,
1165
+ "learning_rate": 9e-05,
1166
+ "loss": 8.079434204101563,
1167
+ "step": 3260
1168
+ },
1169
+ {
1170
+ "epoch": 0.22109875294910683,
1171
+ "grad_norm": 0.0,
1172
+ "learning_rate": 9e-05,
1173
+ "loss": 8.079434204101563,
1174
+ "step": 3280
1175
+ },
1176
+ {
1177
+ "epoch": 0.2224469160768453,
1178
+ "grad_norm": 1.859375,
1179
+ "learning_rate": 9e-05,
1180
+ "loss": 8.07918930053711,
1181
+ "step": 3300
1182
+ },
1183
+ {
1184
+ "epoch": 0.22379507920458375,
1185
+ "grad_norm": 0.69140625,
1186
+ "learning_rate": 9e-05,
1187
+ "loss": 8.077948760986327,
1188
+ "step": 3320
1189
+ },
1190
+ {
1191
+ "epoch": 0.22514324233232222,
1192
+ "grad_norm": 0.0,
1193
+ "learning_rate": 9e-05,
1194
+ "loss": 8.078034973144531,
1195
+ "step": 3340
1196
+ },
1197
+ {
1198
+ "epoch": 0.22649140546006066,
1199
+ "grad_norm": 0.0,
1200
+ "learning_rate": 9e-05,
1201
+ "loss": 8.078900146484376,
1202
+ "step": 3360
1203
+ },
1204
+ {
1205
+ "epoch": 0.22783956858779914,
1206
+ "grad_norm": 0.765625,
1207
+ "learning_rate": 9e-05,
1208
+ "loss": 8.080309295654297,
1209
+ "step": 3380
1210
+ },
1211
+ {
1212
+ "epoch": 0.22918773171553758,
1213
+ "grad_norm": 0.0,
1214
+ "learning_rate": 9e-05,
1215
+ "loss": 8.080132293701173,
1216
+ "step": 3400
1217
+ }
1218
+ ],
1219
+ "logging_steps": 20,
1220
+ "max_steps": 3500,
1221
+ "num_input_tokens_seen": 0,
1222
+ "num_train_epochs": 1,
1223
+ "save_steps": 100,
1224
+ "stateful_callbacks": {
1225
+ "TrainerControl": {
1226
+ "args": {
1227
+ "should_epoch_stop": false,
1228
+ "should_evaluate": false,
1229
+ "should_log": false,
1230
+ "should_save": true,
1231
+ "should_training_stop": false
1232
+ },
1233
+ "attributes": {}
1234
+ }
1235
+ },
1236
+ "total_flos": 1.09693649289216e+16,
1237
+ "train_batch_size": 64,
1238
+ "trial_name": null,
1239
+ "trial_params": null
1240
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3400/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e9b362740bb60edd1ea1474e4e02705aa61c5177a0f605723687fc76ff5105
3
+ size 4920
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "silu-waleed10",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 0.5
36
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f80693e63d774651c221b9dab9cde6f78e5d85bc2d5d5908102217cdffc078d4
3
+ size 33967272
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/optimizer.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:c063139f15a5bfd251c517f323e70515a35ac33292f7b893d555fff46e2aec80
3
+ size 39573050
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/rng_state.pth ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f37c40ce327861a7ca13b719d3aa37510a143368b6e74358bdb14becb3899e1e
3
+ size 14244
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/scheduler.pt ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:479ebefc656d1e8da177b1fc09a2332188f5506e56faf8d0ea281618199ae469
3
+ size 1064
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/trainer_state.json ADDED
@@ -0,0 +1,1283 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "best_global_step": null,
3
+ "best_metric": null,
4
+ "best_model_checkpoint": null,
5
+ "epoch": 0.23592854735422986,
6
+ "eval_steps": 2498,
7
+ "global_step": 3500,
8
+ "is_hyper_param_search": false,
9
+ "is_local_process_zero": true,
10
+ "is_world_process_zero": true,
11
+ "log_history": [
12
+ {
13
+ "epoch": 0.0013481631277384564,
14
+ "grad_norm": 2.078125,
15
+ "learning_rate": 2.66e-05,
16
+ "loss": 8.299632263183593,
17
+ "step": 20
18
+ },
19
+ {
20
+ "epoch": 0.002696326255476913,
21
+ "grad_norm": 1.3515625,
22
+ "learning_rate": 5.46e-05,
23
+ "loss": 8.047638702392579,
24
+ "step": 40
25
+ },
26
+ {
27
+ "epoch": 0.004044489383215369,
28
+ "grad_norm": 1.2734375,
29
+ "learning_rate": 8.259999999999999e-05,
30
+ "loss": 7.837866973876953,
31
+ "step": 60
32
+ },
33
+ {
34
+ "epoch": 0.005392652510953826,
35
+ "grad_norm": 1.265625,
36
+ "learning_rate": 0.0001106,
37
+ "loss": 7.572817993164063,
38
+ "step": 80
39
+ },
40
+ {
41
+ "epoch": 0.006740815638692282,
42
+ "grad_norm": 1.1953125,
43
+ "learning_rate": 0.0001386,
44
+ "loss": 7.233210754394531,
45
+ "step": 100
46
+ },
47
+ {
48
+ "epoch": 0.008088978766430738,
49
+ "grad_norm": 1.1328125,
50
+ "learning_rate": 0.00016659999999999998,
51
+ "loss": 6.860719299316406,
52
+ "step": 120
53
+ },
54
+ {
55
+ "epoch": 0.009437141894169195,
56
+ "grad_norm": 0.94921875,
57
+ "learning_rate": 0.00019460000000000001,
58
+ "loss": 6.5177757263183596,
59
+ "step": 140
60
+ },
61
+ {
62
+ "epoch": 0.010785305021907651,
63
+ "grad_norm": 0.80078125,
64
+ "learning_rate": 0.0002226,
65
+ "loss": 6.239912414550782,
66
+ "step": 160
67
+ },
68
+ {
69
+ "epoch": 0.012133468149646108,
70
+ "grad_norm": 0.74609375,
71
+ "learning_rate": 0.00025059999999999997,
72
+ "loss": 6.035282897949219,
73
+ "step": 180
74
+ },
75
+ {
76
+ "epoch": 0.013481631277384564,
77
+ "grad_norm": 0.5546875,
78
+ "learning_rate": 0.0002786,
79
+ "loss": 5.873902893066406,
80
+ "step": 200
81
+ },
82
+ {
83
+ "epoch": 0.01482979440512302,
84
+ "grad_norm": 0.498046875,
85
+ "learning_rate": 0.00030659999999999997,
86
+ "loss": 5.691160202026367,
87
+ "step": 220
88
+ },
89
+ {
90
+ "epoch": 0.016177957532861477,
91
+ "grad_norm": 1.21875,
92
+ "learning_rate": 0.0003346,
93
+ "loss": 5.473793411254883,
94
+ "step": 240
95
+ },
96
+ {
97
+ "epoch": 0.01752612066059993,
98
+ "grad_norm": 1.5078125,
99
+ "learning_rate": 0.00036260000000000003,
100
+ "loss": 5.258899688720703,
101
+ "step": 260
102
+ },
103
+ {
104
+ "epoch": 0.01887428378833839,
105
+ "grad_norm": 0.796875,
106
+ "learning_rate": 0.0003906,
107
+ "loss": 5.03008918762207,
108
+ "step": 280
109
+ },
110
+ {
111
+ "epoch": 0.020222446916076844,
112
+ "grad_norm": 0.84375,
113
+ "learning_rate": 0.0004186,
114
+ "loss": 4.797770309448242,
115
+ "step": 300
116
+ },
117
+ {
118
+ "epoch": 0.021570610043815303,
119
+ "grad_norm": 1.5078125,
120
+ "learning_rate": 0.0004466,
121
+ "loss": 4.603690719604492,
122
+ "step": 320
123
+ },
124
+ {
125
+ "epoch": 0.022918773171553757,
126
+ "grad_norm": 0.96484375,
127
+ "learning_rate": 0.00047460000000000004,
128
+ "loss": 4.490921401977539,
129
+ "step": 340
130
+ },
131
+ {
132
+ "epoch": 0.024266936299292215,
133
+ "grad_norm": 0.6640625,
134
+ "learning_rate": 0.0005026,
135
+ "loss": 4.327465057373047,
136
+ "step": 360
137
+ },
138
+ {
139
+ "epoch": 0.02561509942703067,
140
+ "grad_norm": 0.86328125,
141
+ "learning_rate": 0.0005306,
142
+ "loss": 4.182389831542968,
143
+ "step": 380
144
+ },
145
+ {
146
+ "epoch": 0.026963262554769128,
147
+ "grad_norm": 0.82421875,
148
+ "learning_rate": 0.0005586,
149
+ "loss": 4.036349487304688,
150
+ "step": 400
151
+ },
152
+ {
153
+ "epoch": 0.028311425682507583,
154
+ "grad_norm": 0.71484375,
155
+ "learning_rate": 0.0005866,
156
+ "loss": 3.918545150756836,
157
+ "step": 420
158
+ },
159
+ {
160
+ "epoch": 0.02965958881024604,
161
+ "grad_norm": 0.765625,
162
+ "learning_rate": 0.0006146,
163
+ "loss": 3.8555908203125,
164
+ "step": 440
165
+ },
166
+ {
167
+ "epoch": 0.031007751937984496,
168
+ "grad_norm": 0.7734375,
169
+ "learning_rate": 0.0006426,
170
+ "loss": 3.7499755859375,
171
+ "step": 460
172
+ },
173
+ {
174
+ "epoch": 0.032355915065722954,
175
+ "grad_norm": 0.6484375,
176
+ "learning_rate": 0.0006705999999999999,
177
+ "loss": 3.668988800048828,
178
+ "step": 480
179
+ },
180
+ {
181
+ "epoch": 0.03370407819346141,
182
+ "grad_norm": 0.8828125,
183
+ "learning_rate": 0.0006986,
184
+ "loss": 3.598476791381836,
185
+ "step": 500
186
+ },
187
+ {
188
+ "epoch": 0.03505224132119986,
189
+ "grad_norm": 0.5625,
190
+ "learning_rate": 0.0007,
191
+ "loss": 3.52923698425293,
192
+ "step": 520
193
+ },
194
+ {
195
+ "epoch": 0.03640040444893832,
196
+ "grad_norm": 0.56640625,
197
+ "learning_rate": 0.0007,
198
+ "loss": 3.4486984252929687,
199
+ "step": 540
200
+ },
201
+ {
202
+ "epoch": 0.03774856757667678,
203
+ "grad_norm": 0.453125,
204
+ "learning_rate": 0.0007,
205
+ "loss": 3.3981090545654298,
206
+ "step": 560
207
+ },
208
+ {
209
+ "epoch": 0.03909673070441524,
210
+ "grad_norm": 0.5234375,
211
+ "learning_rate": 0.0007,
212
+ "loss": 3.337215042114258,
213
+ "step": 580
214
+ },
215
+ {
216
+ "epoch": 0.04044489383215369,
217
+ "grad_norm": 0.48828125,
218
+ "learning_rate": 0.0007,
219
+ "loss": 3.3039676666259767,
220
+ "step": 600
221
+ },
222
+ {
223
+ "epoch": 0.04179305695989215,
224
+ "grad_norm": 0.55078125,
225
+ "learning_rate": 0.0007,
226
+ "loss": 3.2369564056396483,
227
+ "step": 620
228
+ },
229
+ {
230
+ "epoch": 0.043141220087630605,
231
+ "grad_norm": 0.470703125,
232
+ "learning_rate": 0.0007,
233
+ "loss": 3.2108325958251953,
234
+ "step": 640
235
+ },
236
+ {
237
+ "epoch": 0.044489383215369056,
238
+ "grad_norm": 0.46875,
239
+ "learning_rate": 0.0007,
240
+ "loss": 3.153903007507324,
241
+ "step": 660
242
+ },
243
+ {
244
+ "epoch": 0.045837546343107514,
245
+ "grad_norm": 0.412109375,
246
+ "learning_rate": 0.0007,
247
+ "loss": 3.125510597229004,
248
+ "step": 680
249
+ },
250
+ {
251
+ "epoch": 0.04718570947084597,
252
+ "grad_norm": 0.44921875,
253
+ "learning_rate": 0.0007,
254
+ "loss": 3.0803747177124023,
255
+ "step": 700
256
+ },
257
+ {
258
+ "epoch": 0.04853387259858443,
259
+ "grad_norm": 0.390625,
260
+ "learning_rate": 0.0007,
261
+ "loss": 3.04086971282959,
262
+ "step": 720
263
+ },
264
+ {
265
+ "epoch": 0.04988203572632288,
266
+ "grad_norm": 0.458984375,
267
+ "learning_rate": 0.0007,
268
+ "loss": 3.0033853530883787,
269
+ "step": 740
270
+ },
271
+ {
272
+ "epoch": 0.05123019885406134,
273
+ "grad_norm": 0.40625,
274
+ "learning_rate": 0.0007,
275
+ "loss": 2.9853662490844726,
276
+ "step": 760
277
+ },
278
+ {
279
+ "epoch": 0.0525783619817998,
280
+ "grad_norm": 0.48828125,
281
+ "learning_rate": 0.0007,
282
+ "loss": 2.9864330291748047,
283
+ "step": 780
284
+ },
285
+ {
286
+ "epoch": 0.053926525109538256,
287
+ "grad_norm": 0.474609375,
288
+ "learning_rate": 0.0007,
289
+ "loss": 2.9317838668823244,
290
+ "step": 800
291
+ },
292
+ {
293
+ "epoch": 0.05527468823727671,
294
+ "grad_norm": 0.416015625,
295
+ "learning_rate": 0.0007,
296
+ "loss": 2.916472244262695,
297
+ "step": 820
298
+ },
299
+ {
300
+ "epoch": 0.056622851365015166,
301
+ "grad_norm": 0.373046875,
302
+ "learning_rate": 0.0007,
303
+ "loss": 2.8858991622924806,
304
+ "step": 840
305
+ },
306
+ {
307
+ "epoch": 0.057971014492753624,
308
+ "grad_norm": 0.439453125,
309
+ "learning_rate": 0.0007,
310
+ "loss": 2.840240478515625,
311
+ "step": 860
312
+ },
313
+ {
314
+ "epoch": 0.05931917762049208,
315
+ "grad_norm": 0.439453125,
316
+ "learning_rate": 0.0007,
317
+ "loss": 2.8184703826904296,
318
+ "step": 880
319
+ },
320
+ {
321
+ "epoch": 0.06066734074823053,
322
+ "grad_norm": 0.42578125,
323
+ "learning_rate": 0.0007,
324
+ "loss": 2.8047561645507812,
325
+ "step": 900
326
+ },
327
+ {
328
+ "epoch": 0.06201550387596899,
329
+ "grad_norm": 0.4375,
330
+ "learning_rate": 0.0007,
331
+ "loss": 2.775454521179199,
332
+ "step": 920
333
+ },
334
+ {
335
+ "epoch": 0.06336366700370745,
336
+ "grad_norm": 0.421875,
337
+ "learning_rate": 0.0007,
338
+ "loss": 2.7625925064086916,
339
+ "step": 940
340
+ },
341
+ {
342
+ "epoch": 0.06471183013144591,
343
+ "grad_norm": 0.4296875,
344
+ "learning_rate": 0.0007,
345
+ "loss": 2.731528663635254,
346
+ "step": 960
347
+ },
348
+ {
349
+ "epoch": 0.06605999325918437,
350
+ "grad_norm": 0.39453125,
351
+ "learning_rate": 0.0007,
352
+ "loss": 2.71492977142334,
353
+ "step": 980
354
+ },
355
+ {
356
+ "epoch": 0.06740815638692282,
357
+ "grad_norm": 0.40625,
358
+ "learning_rate": 0.0007,
359
+ "loss": 2.7131521224975588,
360
+ "step": 1000
361
+ },
362
+ {
363
+ "epoch": 0.06875631951466127,
364
+ "grad_norm": 0.373046875,
365
+ "learning_rate": 0.0007,
366
+ "loss": 2.6799833297729494,
367
+ "step": 1020
368
+ },
369
+ {
370
+ "epoch": 0.07010448264239973,
371
+ "grad_norm": 0.412109375,
372
+ "learning_rate": 0.0007,
373
+ "loss": 2.665292739868164,
374
+ "step": 1040
375
+ },
376
+ {
377
+ "epoch": 0.07145264577013818,
378
+ "grad_norm": 0.44140625,
379
+ "learning_rate": 0.0007,
380
+ "loss": 2.6485418319702148,
381
+ "step": 1060
382
+ },
383
+ {
384
+ "epoch": 0.07280080889787664,
385
+ "grad_norm": 0.392578125,
386
+ "learning_rate": 0.0007,
387
+ "loss": 2.6286544799804688,
388
+ "step": 1080
389
+ },
390
+ {
391
+ "epoch": 0.0741489720256151,
392
+ "grad_norm": 0.416015625,
393
+ "learning_rate": 0.0007,
394
+ "loss": 2.621373176574707,
395
+ "step": 1100
396
+ },
397
+ {
398
+ "epoch": 0.07549713515335356,
399
+ "grad_norm": 0.375,
400
+ "learning_rate": 0.0007,
401
+ "loss": 2.5860437393188476,
402
+ "step": 1120
403
+ },
404
+ {
405
+ "epoch": 0.07684529828109202,
406
+ "grad_norm": 0.4453125,
407
+ "learning_rate": 0.0007,
408
+ "loss": 2.5712385177612305,
409
+ "step": 1140
410
+ },
411
+ {
412
+ "epoch": 0.07819346140883048,
413
+ "grad_norm": 0.380859375,
414
+ "learning_rate": 0.0007,
415
+ "loss": 2.5693639755249023,
416
+ "step": 1160
417
+ },
418
+ {
419
+ "epoch": 0.07954162453656892,
420
+ "grad_norm": 0.3984375,
421
+ "learning_rate": 0.0007,
422
+ "loss": 2.563449668884277,
423
+ "step": 1180
424
+ },
425
+ {
426
+ "epoch": 0.08088978766430738,
427
+ "grad_norm": 0.369140625,
428
+ "learning_rate": 0.0007,
429
+ "loss": 2.553573417663574,
430
+ "step": 1200
431
+ },
432
+ {
433
+ "epoch": 0.08223795079204584,
434
+ "grad_norm": 0.3984375,
435
+ "learning_rate": 0.0007,
436
+ "loss": 2.538028335571289,
437
+ "step": 1220
438
+ },
439
+ {
440
+ "epoch": 0.0835861139197843,
441
+ "grad_norm": 0.365234375,
442
+ "learning_rate": 0.0007,
443
+ "loss": 2.5232309341430663,
444
+ "step": 1240
445
+ },
446
+ {
447
+ "epoch": 0.08493427704752275,
448
+ "grad_norm": 0.373046875,
449
+ "learning_rate": 0.0007,
450
+ "loss": 2.50943546295166,
451
+ "step": 1260
452
+ },
453
+ {
454
+ "epoch": 0.08628244017526121,
455
+ "grad_norm": 0.373046875,
456
+ "learning_rate": 0.0007,
457
+ "loss": 2.4896902084350585,
458
+ "step": 1280
459
+ },
460
+ {
461
+ "epoch": 0.08763060330299967,
462
+ "grad_norm": 0.375,
463
+ "learning_rate": 0.0007,
464
+ "loss": 2.4810691833496095,
465
+ "step": 1300
466
+ },
467
+ {
468
+ "epoch": 0.08897876643073811,
469
+ "grad_norm": 0.349609375,
470
+ "learning_rate": 0.0007,
471
+ "loss": 2.471504974365234,
472
+ "step": 1320
473
+ },
474
+ {
475
+ "epoch": 0.09032692955847657,
476
+ "grad_norm": 0.3671875,
477
+ "learning_rate": 0.0007,
478
+ "loss": 2.458695411682129,
479
+ "step": 1340
480
+ },
481
+ {
482
+ "epoch": 0.09167509268621503,
483
+ "grad_norm": 0.400390625,
484
+ "learning_rate": 0.0007,
485
+ "loss": 2.4555185317993162,
486
+ "step": 1360
487
+ },
488
+ {
489
+ "epoch": 0.09302325581395349,
490
+ "grad_norm": 0.35546875,
491
+ "learning_rate": 0.0007,
492
+ "loss": 2.433156204223633,
493
+ "step": 1380
494
+ },
495
+ {
496
+ "epoch": 0.09437141894169195,
497
+ "grad_norm": 0.34375,
498
+ "learning_rate": 0.0007,
499
+ "loss": 2.425444412231445,
500
+ "step": 1400
501
+ },
502
+ {
503
+ "epoch": 0.0957195820694304,
504
+ "grad_norm": 0.400390625,
505
+ "learning_rate": 0.0007,
506
+ "loss": 2.435191345214844,
507
+ "step": 1420
508
+ },
509
+ {
510
+ "epoch": 0.09706774519716886,
511
+ "grad_norm": 0.369140625,
512
+ "learning_rate": 0.0007,
513
+ "loss": 2.412186622619629,
514
+ "step": 1440
515
+ },
516
+ {
517
+ "epoch": 0.09841590832490732,
518
+ "grad_norm": 0.375,
519
+ "learning_rate": 0.0007,
520
+ "loss": 2.4007726669311524,
521
+ "step": 1460
522
+ },
523
+ {
524
+ "epoch": 0.09976407145264576,
525
+ "grad_norm": 0.357421875,
526
+ "learning_rate": 0.0007,
527
+ "loss": 2.3866371154785155,
528
+ "step": 1480
529
+ },
530
+ {
531
+ "epoch": 0.10111223458038422,
532
+ "grad_norm": 0.35546875,
533
+ "learning_rate": 0.0007,
534
+ "loss": 2.3728200912475588,
535
+ "step": 1500
536
+ },
537
+ {
538
+ "epoch": 0.10246039770812268,
539
+ "grad_norm": 0.376953125,
540
+ "learning_rate": 0.0007,
541
+ "loss": 2.373748016357422,
542
+ "step": 1520
543
+ },
544
+ {
545
+ "epoch": 0.10380856083586114,
546
+ "grad_norm": 0.36328125,
547
+ "learning_rate": 0.0007,
548
+ "loss": 2.3635543823242187,
549
+ "step": 1540
550
+ },
551
+ {
552
+ "epoch": 0.1051567239635996,
553
+ "grad_norm": 0.37890625,
554
+ "learning_rate": 0.0007,
555
+ "loss": 2.3516841888427735,
556
+ "step": 1560
557
+ },
558
+ {
559
+ "epoch": 0.10650488709133805,
560
+ "grad_norm": 0.376953125,
561
+ "learning_rate": 0.0007,
562
+ "loss": 2.3689512252807616,
563
+ "step": 1580
564
+ },
565
+ {
566
+ "epoch": 0.10785305021907651,
567
+ "grad_norm": 0.365234375,
568
+ "learning_rate": 0.0007,
569
+ "loss": 2.3348180770874025,
570
+ "step": 1600
571
+ },
572
+ {
573
+ "epoch": 0.10920121334681497,
574
+ "grad_norm": 0.361328125,
575
+ "learning_rate": 0.0007,
576
+ "loss": 2.3343011856079103,
577
+ "step": 1620
578
+ },
579
+ {
580
+ "epoch": 0.11054937647455342,
581
+ "grad_norm": 0.37890625,
582
+ "learning_rate": 0.0007,
583
+ "loss": 2.3455299377441405,
584
+ "step": 1640
585
+ },
586
+ {
587
+ "epoch": 0.11189753960229187,
588
+ "grad_norm": 0.35546875,
589
+ "learning_rate": 0.0007,
590
+ "loss": 2.3264875411987305,
591
+ "step": 1660
592
+ },
593
+ {
594
+ "epoch": 0.11324570273003033,
595
+ "grad_norm": 0.3984375,
596
+ "learning_rate": 0.0007,
597
+ "loss": 2.309524917602539,
598
+ "step": 1680
599
+ },
600
+ {
601
+ "epoch": 0.11459386585776879,
602
+ "grad_norm": 0.357421875,
603
+ "learning_rate": 0.0007,
604
+ "loss": 2.3181018829345703,
605
+ "step": 1700
606
+ },
607
+ {
608
+ "epoch": 0.11594202898550725,
609
+ "grad_norm": 0.412109375,
610
+ "learning_rate": 0.0007,
611
+ "loss": 2.2959054946899413,
612
+ "step": 1720
613
+ },
614
+ {
615
+ "epoch": 0.1172901921132457,
616
+ "grad_norm": 0.3515625,
617
+ "learning_rate": 0.0007,
618
+ "loss": 2.286634826660156,
619
+ "step": 1740
620
+ },
621
+ {
622
+ "epoch": 0.11863835524098416,
623
+ "grad_norm": 0.3671875,
624
+ "learning_rate": 0.0007,
625
+ "loss": 2.2912055969238283,
626
+ "step": 1760
627
+ },
628
+ {
629
+ "epoch": 0.11998651836872262,
630
+ "grad_norm": 0.349609375,
631
+ "learning_rate": 0.0007,
632
+ "loss": 2.303947639465332,
633
+ "step": 1780
634
+ },
635
+ {
636
+ "epoch": 0.12133468149646107,
637
+ "grad_norm": 0.33984375,
638
+ "learning_rate": 0.0007,
639
+ "loss": 2.298970413208008,
640
+ "step": 1800
641
+ },
642
+ {
643
+ "epoch": 0.12268284462419952,
644
+ "grad_norm": 0.373046875,
645
+ "learning_rate": 0.0007,
646
+ "loss": 2.2899410247802736,
647
+ "step": 1820
648
+ },
649
+ {
650
+ "epoch": 0.12403100775193798,
651
+ "grad_norm": 0.3515625,
652
+ "learning_rate": 0.0007,
653
+ "loss": 2.258684539794922,
654
+ "step": 1840
655
+ },
656
+ {
657
+ "epoch": 0.12537917087967643,
658
+ "grad_norm": 0.365234375,
659
+ "learning_rate": 0.0007,
660
+ "loss": 2.2575857162475588,
661
+ "step": 1860
662
+ },
663
+ {
664
+ "epoch": 0.1267273340074149,
665
+ "grad_norm": 0.33984375,
666
+ "learning_rate": 0.0007,
667
+ "loss": 2.2479154586791994,
668
+ "step": 1880
669
+ },
670
+ {
671
+ "epoch": 0.12807549713515334,
672
+ "grad_norm": 0.3515625,
673
+ "learning_rate": 0.0007,
674
+ "loss": 2.2578588485717774,
675
+ "step": 1900
676
+ },
677
+ {
678
+ "epoch": 0.12942366026289182,
679
+ "grad_norm": 0.35546875,
680
+ "learning_rate": 0.0007,
681
+ "loss": 2.2332910537719726,
682
+ "step": 1920
683
+ },
684
+ {
685
+ "epoch": 0.13077182339063026,
686
+ "grad_norm": 0.35546875,
687
+ "learning_rate": 0.0007,
688
+ "loss": 2.242460250854492,
689
+ "step": 1940
690
+ },
691
+ {
692
+ "epoch": 0.13211998651836873,
693
+ "grad_norm": 0.359375,
694
+ "learning_rate": 0.0007,
695
+ "loss": 2.21710205078125,
696
+ "step": 1960
697
+ },
698
+ {
699
+ "epoch": 0.13346814964610718,
700
+ "grad_norm": 0.353515625,
701
+ "learning_rate": 0.0007,
702
+ "loss": 2.2389858245849608,
703
+ "step": 1980
704
+ },
705
+ {
706
+ "epoch": 0.13481631277384565,
707
+ "grad_norm": 0.328125,
708
+ "learning_rate": 0.0007,
709
+ "loss": 2.2128574371337892,
710
+ "step": 2000
711
+ },
712
+ {
713
+ "epoch": 0.1361644759015841,
714
+ "grad_norm": 0.349609375,
715
+ "learning_rate": 0.0007,
716
+ "loss": 2.221720504760742,
717
+ "step": 2020
718
+ },
719
+ {
720
+ "epoch": 0.13751263902932254,
721
+ "grad_norm": 0.34765625,
722
+ "learning_rate": 0.0007,
723
+ "loss": 2.2012025833129885,
724
+ "step": 2040
725
+ },
726
+ {
727
+ "epoch": 0.138860802157061,
728
+ "grad_norm": 0.359375,
729
+ "learning_rate": 0.0007,
730
+ "loss": 2.2207321166992187,
731
+ "step": 2060
732
+ },
733
+ {
734
+ "epoch": 0.14020896528479945,
735
+ "grad_norm": 0.3515625,
736
+ "learning_rate": 0.0007,
737
+ "loss": 2.1970823287963865,
738
+ "step": 2080
739
+ },
740
+ {
741
+ "epoch": 0.14155712841253792,
742
+ "grad_norm": 0.3359375,
743
+ "learning_rate": 0.0007,
744
+ "loss": 2.1874614715576173,
745
+ "step": 2100
746
+ },
747
+ {
748
+ "epoch": 0.14290529154027637,
749
+ "grad_norm": 0.35546875,
750
+ "learning_rate": 0.0007,
751
+ "loss": 2.1962921142578127,
752
+ "step": 2120
753
+ },
754
+ {
755
+ "epoch": 0.14425345466801484,
756
+ "grad_norm": 0.375,
757
+ "learning_rate": 0.0007,
758
+ "loss": 2.1974483489990235,
759
+ "step": 2140
760
+ },
761
+ {
762
+ "epoch": 0.14560161779575329,
763
+ "grad_norm": 0.34375,
764
+ "learning_rate": 0.0007,
765
+ "loss": 2.1824066162109377,
766
+ "step": 2160
767
+ },
768
+ {
769
+ "epoch": 0.14694978092349173,
770
+ "grad_norm": 0.3359375,
771
+ "learning_rate": 0.0007,
772
+ "loss": 2.1886768341064453,
773
+ "step": 2180
774
+ },
775
+ {
776
+ "epoch": 0.1482979440512302,
777
+ "grad_norm": 0.326171875,
778
+ "learning_rate": 0.0007,
779
+ "loss": 2.1661350250244142,
780
+ "step": 2200
781
+ },
782
+ {
783
+ "epoch": 0.14964610717896865,
784
+ "grad_norm": 0.330078125,
785
+ "learning_rate": 0.0007,
786
+ "loss": 2.1746242523193358,
787
+ "step": 2220
788
+ },
789
+ {
790
+ "epoch": 0.15099427030670712,
791
+ "grad_norm": 0.353515625,
792
+ "learning_rate": 0.0007,
793
+ "loss": 2.182226371765137,
794
+ "step": 2240
795
+ },
796
+ {
797
+ "epoch": 0.15234243343444556,
798
+ "grad_norm": 0.33203125,
799
+ "learning_rate": 0.0007,
800
+ "loss": 2.1612625122070312,
801
+ "step": 2260
802
+ },
803
+ {
804
+ "epoch": 0.15369059656218403,
805
+ "grad_norm": 0.337890625,
806
+ "learning_rate": 0.0007,
807
+ "loss": 2.1695093154907226,
808
+ "step": 2280
809
+ },
810
+ {
811
+ "epoch": 0.15503875968992248,
812
+ "grad_norm": 0.341796875,
813
+ "learning_rate": 0.0007,
814
+ "loss": 2.1509073257446287,
815
+ "step": 2300
816
+ },
817
+ {
818
+ "epoch": 0.15638692281766095,
819
+ "grad_norm": 0.33984375,
820
+ "learning_rate": 0.0007,
821
+ "loss": 2.1601842880249023,
822
+ "step": 2320
823
+ },
824
+ {
825
+ "epoch": 0.1577350859453994,
826
+ "grad_norm": 0.3203125,
827
+ "learning_rate": 0.0007,
828
+ "loss": 2.149628257751465,
829
+ "step": 2340
830
+ },
831
+ {
832
+ "epoch": 0.15908324907313784,
833
+ "grad_norm": 0.33203125,
834
+ "learning_rate": 0.0007,
835
+ "loss": 2.154475784301758,
836
+ "step": 2360
837
+ },
838
+ {
839
+ "epoch": 0.1604314122008763,
840
+ "grad_norm": 0.333984375,
841
+ "learning_rate": 0.0007,
842
+ "loss": 2.1572238922119142,
843
+ "step": 2380
844
+ },
845
+ {
846
+ "epoch": 0.16177957532861476,
847
+ "grad_norm": 0.34765625,
848
+ "learning_rate": 0.0007,
849
+ "loss": 2.1351224899291994,
850
+ "step": 2400
851
+ },
852
+ {
853
+ "epoch": 0.16312773845635323,
854
+ "grad_norm": 0.345703125,
855
+ "learning_rate": 0.0007,
856
+ "loss": 2.141549301147461,
857
+ "step": 2420
858
+ },
859
+ {
860
+ "epoch": 0.16447590158409167,
861
+ "grad_norm": 0.328125,
862
+ "learning_rate": 0.0007,
863
+ "loss": 2.127903366088867,
864
+ "step": 2440
865
+ },
866
+ {
867
+ "epoch": 0.16582406471183014,
868
+ "grad_norm": 0.32421875,
869
+ "learning_rate": 0.0007,
870
+ "loss": 2.1384496688842773,
871
+ "step": 2460
872
+ },
873
+ {
874
+ "epoch": 0.1671722278395686,
875
+ "grad_norm": 0.34375,
876
+ "learning_rate": 0.0007,
877
+ "loss": 2.1402868270874023,
878
+ "step": 2480
879
+ },
880
+ {
881
+ "epoch": 0.1683855746545332,
882
+ "eval_loss": 2.133925199508667,
883
+ "eval_runtime": 18.9626,
884
+ "eval_samples_per_second": 502.41,
885
+ "eval_steps_per_second": 0.527,
886
+ "step": 2498
887
+ },
888
+ {
889
+ "epoch": 0.16852039096730703,
890
+ "grad_norm": 0.34765625,
891
+ "learning_rate": 0.0007,
892
+ "loss": 2.1217138290405275,
893
+ "step": 2500
894
+ },
895
+ {
896
+ "epoch": 0.16852039096730703,
897
+ "eval_loss": 2.1338183879852295,
898
+ "eval_runtime": 18.812,
899
+ "eval_samples_per_second": 506.431,
900
+ "eval_steps_per_second": 0.532,
901
+ "step": 2500
902
+ },
903
+ {
904
+ "epoch": 0.1698685540950455,
905
+ "grad_norm": 15335424.0,
906
+ "learning_rate": 3.4200000000000005e-05,
907
+ "loss": 2.1747163772583007,
908
+ "step": 2520
909
+ },
910
+ {
911
+ "epoch": 0.17121671722278395,
912
+ "grad_norm": 16646144.0,
913
+ "learning_rate": 7.020000000000001e-05,
914
+ "loss": 2.1868026733398436,
915
+ "step": 2540
916
+ },
917
+ {
918
+ "epoch": 0.17256488035052242,
919
+ "grad_norm": 102760448.0,
920
+ "learning_rate": 9e-05,
921
+ "loss": 3.535448455810547,
922
+ "step": 2560
923
+ },
924
+ {
925
+ "epoch": 0.17391304347826086,
926
+ "grad_norm": 188743680.0,
927
+ "learning_rate": 9e-05,
928
+ "loss": 7.000344848632812,
929
+ "step": 2580
930
+ },
931
+ {
932
+ "epoch": 0.17526120660599934,
933
+ "grad_norm": 186646528.0,
934
+ "learning_rate": 9e-05,
935
+ "loss": 7.602831268310547,
936
+ "step": 2600
937
+ },
938
+ {
939
+ "epoch": 0.17660936973373778,
940
+ "grad_norm": 110624768.0,
941
+ "learning_rate": 9e-05,
942
+ "loss": 7.876109313964844,
943
+ "step": 2620
944
+ },
945
+ {
946
+ "epoch": 0.17795753286147623,
947
+ "grad_norm": 52690944.0,
948
+ "learning_rate": 9e-05,
949
+ "loss": 8.0205078125,
950
+ "step": 2640
951
+ },
952
+ {
953
+ "epoch": 0.1793056959892147,
954
+ "grad_norm": 21102592.0,
955
+ "learning_rate": 9e-05,
956
+ "loss": 8.05886001586914,
957
+ "step": 2660
958
+ },
959
+ {
960
+ "epoch": 0.18065385911695314,
961
+ "grad_norm": 9830400.0,
962
+ "learning_rate": 9e-05,
963
+ "loss": 8.069260406494141,
964
+ "step": 2680
965
+ },
966
+ {
967
+ "epoch": 0.1820020222446916,
968
+ "grad_norm": 3670016.0,
969
+ "learning_rate": 9e-05,
970
+ "loss": 8.075779724121094,
971
+ "step": 2700
972
+ },
973
+ {
974
+ "epoch": 0.18335018537243006,
975
+ "grad_norm": 3571712.0,
976
+ "learning_rate": 9e-05,
977
+ "loss": 8.076358795166016,
978
+ "step": 2720
979
+ },
980
+ {
981
+ "epoch": 0.18469834850016853,
982
+ "grad_norm": 1114112.0,
983
+ "learning_rate": 9e-05,
984
+ "loss": 8.07742919921875,
985
+ "step": 2740
986
+ },
987
+ {
988
+ "epoch": 0.18604651162790697,
989
+ "grad_norm": 421888.0,
990
+ "learning_rate": 9e-05,
991
+ "loss": 8.07922592163086,
992
+ "step": 2760
993
+ },
994
+ {
995
+ "epoch": 0.18739467475564545,
996
+ "grad_norm": 129536.0,
997
+ "learning_rate": 9e-05,
998
+ "loss": 8.079435729980469,
999
+ "step": 2780
1000
+ },
1001
+ {
1002
+ "epoch": 0.1887428378833839,
1003
+ "grad_norm": 34048.0,
1004
+ "learning_rate": 9e-05,
1005
+ "loss": 8.079093170166015,
1006
+ "step": 2800
1007
+ },
1008
+ {
1009
+ "epoch": 0.19009100101112233,
1010
+ "grad_norm": 5888.0,
1011
+ "learning_rate": 9e-05,
1012
+ "loss": 8.07839126586914,
1013
+ "step": 2820
1014
+ },
1015
+ {
1016
+ "epoch": 0.1914391641388608,
1017
+ "grad_norm": 446.0,
1018
+ "learning_rate": 9e-05,
1019
+ "loss": 8.081797027587891,
1020
+ "step": 2840
1021
+ },
1022
+ {
1023
+ "epoch": 0.19278732726659925,
1024
+ "grad_norm": 10.25,
1025
+ "learning_rate": 9e-05,
1026
+ "loss": 8.079029083251953,
1027
+ "step": 2860
1028
+ },
1029
+ {
1030
+ "epoch": 0.19413549039433772,
1031
+ "grad_norm": 6.53125,
1032
+ "learning_rate": 9e-05,
1033
+ "loss": 8.081595611572265,
1034
+ "step": 2880
1035
+ },
1036
+ {
1037
+ "epoch": 0.19548365352207617,
1038
+ "grad_norm": 3.4375,
1039
+ "learning_rate": 9e-05,
1040
+ "loss": 8.079817962646484,
1041
+ "step": 2900
1042
+ },
1043
+ {
1044
+ "epoch": 0.19683181664981464,
1045
+ "grad_norm": 0.0,
1046
+ "learning_rate": 9e-05,
1047
+ "loss": 8.079161834716796,
1048
+ "step": 2920
1049
+ },
1050
+ {
1051
+ "epoch": 0.19817997977755308,
1052
+ "grad_norm": 1.3046875,
1053
+ "learning_rate": 9e-05,
1054
+ "loss": 8.07956314086914,
1055
+ "step": 2940
1056
+ },
1057
+ {
1058
+ "epoch": 0.19952814290529153,
1059
+ "grad_norm": 2.40625,
1060
+ "learning_rate": 9e-05,
1061
+ "loss": 8.074960327148437,
1062
+ "step": 2960
1063
+ },
1064
+ {
1065
+ "epoch": 0.20087630603303,
1066
+ "grad_norm": 81.5,
1067
+ "learning_rate": 9e-05,
1068
+ "loss": 8.076923370361328,
1069
+ "step": 2980
1070
+ },
1071
+ {
1072
+ "epoch": 0.20222446916076844,
1073
+ "grad_norm": 0.0,
1074
+ "learning_rate": 9e-05,
1075
+ "loss": 8.076215362548828,
1076
+ "step": 3000
1077
+ },
1078
+ {
1079
+ "epoch": 0.20357263228850692,
1080
+ "grad_norm": 0.0,
1081
+ "learning_rate": 9e-05,
1082
+ "loss": 8.080150604248047,
1083
+ "step": 3020
1084
+ },
1085
+ {
1086
+ "epoch": 0.20492079541624536,
1087
+ "grad_norm": 0.0,
1088
+ "learning_rate": 9e-05,
1089
+ "loss": 8.082421875,
1090
+ "step": 3040
1091
+ },
1092
+ {
1093
+ "epoch": 0.20626895854398383,
1094
+ "grad_norm": 0.0,
1095
+ "learning_rate": 9e-05,
1096
+ "loss": 8.076376342773438,
1097
+ "step": 3060
1098
+ },
1099
+ {
1100
+ "epoch": 0.20761712167172228,
1101
+ "grad_norm": 0.0,
1102
+ "learning_rate": 9e-05,
1103
+ "loss": 8.07555694580078,
1104
+ "step": 3080
1105
+ },
1106
+ {
1107
+ "epoch": 0.20896528479946072,
1108
+ "grad_norm": 1.03125,
1109
+ "learning_rate": 9e-05,
1110
+ "loss": 8.079136657714844,
1111
+ "step": 3100
1112
+ },
1113
+ {
1114
+ "epoch": 0.2103134479271992,
1115
+ "grad_norm": 0.0,
1116
+ "learning_rate": 9e-05,
1117
+ "loss": 8.081492614746093,
1118
+ "step": 3120
1119
+ },
1120
+ {
1121
+ "epoch": 0.21166161105493764,
1122
+ "grad_norm": 0.0,
1123
+ "learning_rate": 9e-05,
1124
+ "loss": 8.079338073730469,
1125
+ "step": 3140
1126
+ },
1127
+ {
1128
+ "epoch": 0.2130097741826761,
1129
+ "grad_norm": 138.0,
1130
+ "learning_rate": 9e-05,
1131
+ "loss": 8.077278137207031,
1132
+ "step": 3160
1133
+ },
1134
+ {
1135
+ "epoch": 0.21435793731041455,
1136
+ "grad_norm": 1.109375,
1137
+ "learning_rate": 9e-05,
1138
+ "loss": 8.075982666015625,
1139
+ "step": 3180
1140
+ },
1141
+ {
1142
+ "epoch": 0.21570610043815303,
1143
+ "grad_norm": 0.0,
1144
+ "learning_rate": 9e-05,
1145
+ "loss": 8.079740142822265,
1146
+ "step": 3200
1147
+ },
1148
+ {
1149
+ "epoch": 0.21705426356589147,
1150
+ "grad_norm": 2.34375,
1151
+ "learning_rate": 9e-05,
1152
+ "loss": 8.080359649658202,
1153
+ "step": 3220
1154
+ },
1155
+ {
1156
+ "epoch": 0.21840242669362994,
1157
+ "grad_norm": 0.0,
1158
+ "learning_rate": 9e-05,
1159
+ "loss": 8.0783935546875,
1160
+ "step": 3240
1161
+ },
1162
+ {
1163
+ "epoch": 0.2197505898213684,
1164
+ "grad_norm": 0.0,
1165
+ "learning_rate": 9e-05,
1166
+ "loss": 8.079434204101563,
1167
+ "step": 3260
1168
+ },
1169
+ {
1170
+ "epoch": 0.22109875294910683,
1171
+ "grad_norm": 0.0,
1172
+ "learning_rate": 9e-05,
1173
+ "loss": 8.079434204101563,
1174
+ "step": 3280
1175
+ },
1176
+ {
1177
+ "epoch": 0.2224469160768453,
1178
+ "grad_norm": 1.859375,
1179
+ "learning_rate": 9e-05,
1180
+ "loss": 8.07918930053711,
1181
+ "step": 3300
1182
+ },
1183
+ {
1184
+ "epoch": 0.22379507920458375,
1185
+ "grad_norm": 0.69140625,
1186
+ "learning_rate": 9e-05,
1187
+ "loss": 8.077948760986327,
1188
+ "step": 3320
1189
+ },
1190
+ {
1191
+ "epoch": 0.22514324233232222,
1192
+ "grad_norm": 0.0,
1193
+ "learning_rate": 9e-05,
1194
+ "loss": 8.078034973144531,
1195
+ "step": 3340
1196
+ },
1197
+ {
1198
+ "epoch": 0.22649140546006066,
1199
+ "grad_norm": 0.0,
1200
+ "learning_rate": 9e-05,
1201
+ "loss": 8.078900146484376,
1202
+ "step": 3360
1203
+ },
1204
+ {
1205
+ "epoch": 0.22783956858779914,
1206
+ "grad_norm": 0.765625,
1207
+ "learning_rate": 9e-05,
1208
+ "loss": 8.080309295654297,
1209
+ "step": 3380
1210
+ },
1211
+ {
1212
+ "epoch": 0.22918773171553758,
1213
+ "grad_norm": 0.0,
1214
+ "learning_rate": 9e-05,
1215
+ "loss": 8.080132293701173,
1216
+ "step": 3400
1217
+ },
1218
+ {
1219
+ "epoch": 0.23053589484327602,
1220
+ "grad_norm": 0.0,
1221
+ "learning_rate": 9e-05,
1222
+ "loss": 8.080228424072265,
1223
+ "step": 3420
1224
+ },
1225
+ {
1226
+ "epoch": 0.2318840579710145,
1227
+ "grad_norm": 0.0,
1228
+ "learning_rate": 9e-05,
1229
+ "loss": 8.07936782836914,
1230
+ "step": 3440
1231
+ },
1232
+ {
1233
+ "epoch": 0.23323222109875294,
1234
+ "grad_norm": 0.0,
1235
+ "learning_rate": 9e-05,
1236
+ "loss": 8.082151794433594,
1237
+ "step": 3460
1238
+ },
1239
+ {
1240
+ "epoch": 0.2345803842264914,
1241
+ "grad_norm": 1.1796875,
1242
+ "learning_rate": 9e-05,
1243
+ "loss": 8.082518005371094,
1244
+ "step": 3480
1245
+ },
1246
+ {
1247
+ "epoch": 0.23592854735422986,
1248
+ "grad_norm": 0.0,
1249
+ "learning_rate": 9e-05,
1250
+ "loss": 8.08392562866211,
1251
+ "step": 3500
1252
+ },
1253
+ {
1254
+ "epoch": 0.23592854735422986,
1255
+ "eval_loss": 8.07800579071045,
1256
+ "eval_runtime": 15.771,
1257
+ "eval_samples_per_second": 604.082,
1258
+ "eval_steps_per_second": 0.634,
1259
+ "step": 3500
1260
+ }
1261
+ ],
1262
+ "logging_steps": 20,
1263
+ "max_steps": 3500,
1264
+ "num_input_tokens_seen": 0,
1265
+ "num_train_epochs": 1,
1266
+ "save_steps": 100,
1267
+ "stateful_callbacks": {
1268
+ "TrainerControl": {
1269
+ "args": {
1270
+ "should_epoch_stop": false,
1271
+ "should_evaluate": false,
1272
+ "should_log": false,
1273
+ "should_save": true,
1274
+ "should_training_stop": true
1275
+ },
1276
+ "attributes": {}
1277
+ }
1278
+ },
1279
+ "total_flos": 1.1291993309184e+16,
1280
+ "train_batch_size": 64,
1281
+ "trial_name": null,
1282
+ "trial_params": null
1283
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/checkpoint-3500/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e9b362740bb60edd1ea1474e4e02705aa61c5177a0f605723687fc76ff5105
3
+ size 4920
zain/Activation/out/glu-silu-waleed10-100L_trash_run/config.json ADDED
@@ -0,0 +1,36 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "activation": "silu-waleed10",
3
+ "architectures": [
4
+ "TinyLlamaForCausalLM"
5
+ ],
6
+ "attention_bias": false,
7
+ "attention_dropout": 0.0,
8
+ "bos_token_id": 1,
9
+ "dtype": "bfloat16",
10
+ "eos_token_id": 2,
11
+ "head_dim": 32,
12
+ "hidden_act": "silu",
13
+ "hidden_size": 128,
14
+ "initializer_range": 0.02,
15
+ "intermediate_size": 256,
16
+ "max_position_embeddings": 512,
17
+ "mlp_bias": false,
18
+ "mlp_type": "glu",
19
+ "model_type": "tiny_llama",
20
+ "num_attention_heads": 4,
21
+ "num_hidden_layers": 100,
22
+ "num_key_value_heads": 4,
23
+ "pad_token_id": 0,
24
+ "pretraining_tp": 1,
25
+ "rms_norm_eps": 1e-06,
26
+ "rope_parameters": {
27
+ "rope_theta": 10000.0,
28
+ "rope_type": "default"
29
+ },
30
+ "tie_word_embeddings": true,
31
+ "tokenizer_name": "w-ahmad/tiny-stories-tokenizer",
32
+ "transformers_version": "5.16.0.dev0",
33
+ "use_cache": false,
34
+ "vocab_size": 4096,
35
+ "waleed_beta": 0.5
36
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/model.safetensors ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:f80693e63d774651c221b9dab9cde6f78e5d85bc2d5d5908102217cdffc078d4
3
+ size 33967272
zain/Activation/out/glu-silu-waleed10-100L_trash_run/tokenizer.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/out/glu-silu-waleed10-100L_trash_run/tokenizer_config.json ADDED
@@ -0,0 +1,13 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ {
2
+ "add_prefix_space": false,
3
+ "backend": "tokenizers",
4
+ "bos_token": "<|endoftext|>",
5
+ "eos_token": "<|endoftext|>",
6
+ "errors": "replace",
7
+ "is_local": false,
8
+ "local_files_only": false,
9
+ "model_max_length": 1024,
10
+ "pad_token": "<|endoftext|>",
11
+ "tokenizer_class": "GPT2Tokenizer",
12
+ "unk_token": "<|endoftext|>"
13
+ }
zain/Activation/out/glu-silu-waleed10-100L_trash_run/training_args.bin ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:04e9b362740bb60edd1ea1474e4e02705aa61c5177a0f605723687fc76ff5105
3
+ size 4920
zain/Activation/out/glu-silu-waleed10-100L_trash_run/training_log.jsonl CHANGED
@@ -1,3 +1,3 @@
1
  version https://git-lfs.github.com/spec/v1
2
- oid sha256:4a2f82ea396fbf33db4402d2f3f595b1af23b18968fc162cf82f89676051bc8b
3
- size 11736982
 
1
  version https://git-lfs.github.com/spec/v1
2
+ oid sha256:feacc9768937eb46f8b63588280ff13ccf0e46095c75a7932de3d187803dd2f9
3
+ size 16723329
zain/Activation/out/sweep_summary.json ADDED
@@ -0,0 +1,16 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ [
2
+ {
3
+ "variant": "glu-waleed10-100L",
4
+ "eval_loss": 7.91505241394043,
5
+ "out": "out/glu-waleed10-100L_trash_run",
6
+ "run_name": "LM-glu-waleed10-100L-16.9M-20260814-231801",
7
+ "status": "success"
8
+ },
9
+ {
10
+ "variant": "glu-silu-waleed10-100L",
11
+ "eval_loss": 8.07800579071045,
12
+ "out": "out/glu-silu-waleed10-100L_trash_run",
13
+ "run_name": "LM-glu-silu-waleed10-100L-16.9M-20260814-232317",
14
+ "status": "success"
15
+ }
16
+ ]
zain/Activation/wandb/debug-internal.log CHANGED
@@ -36,3 +36,28 @@
36
  {"time":"2026-08-14T23:26:18.932400326Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":28,"history_lines":3,"events_offset":21,"events_lines":2,"console_offset":51,"console_lines":1}
37
  {"time":"2026-08-14T23:26:19.870609328Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
38
  {"time":"2026-08-14T23:26:33.939867382Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":31,"history_lines":2,"events_offset":23,"events_lines":2,"console_offset":55,"console_lines":10}
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
36
  {"time":"2026-08-14T23:26:18.932400326Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":28,"history_lines":3,"events_offset":21,"events_lines":2,"console_offset":51,"console_lines":1}
37
  {"time":"2026-08-14T23:26:19.870609328Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
38
  {"time":"2026-08-14T23:26:33.939867382Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":31,"history_lines":2,"events_offset":23,"events_lines":2,"console_offset":55,"console_lines":10}
39
+ {"time":"2026-08-14T23:26:34.841652736Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
40
+ {"time":"2026-08-14T23:26:48.937729263Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":33,"history_lines":3,"events_offset":25,"events_lines":2,"console_offset":61,"console_lines":1}
41
+ {"time":"2026-08-14T23:26:49.803943322Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
42
+ {"time":"2026-08-14T23:27:03.931480357Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":36,"history_lines":2,"events_offset":27,"events_lines":2,"console_offset":65,"console_lines":10}
43
+ {"time":"2026-08-14T23:27:04.817526975Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
44
+ {"time":"2026-08-14T23:27:18.930769355Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":38,"history_lines":3,"events_offset":29,"events_lines":2,"console_offset":71,"console_lines":1}
45
+ {"time":"2026-08-14T23:27:19.775644492Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
46
+ {"time":"2026-08-14T23:27:33.935804443Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":41,"history_lines":2,"events_offset":31,"events_lines":2,"console_offset":75,"console_lines":10}
47
+ {"time":"2026-08-14T23:27:34.968495829Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
48
+ {"time":"2026-08-14T23:27:48.930502995Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":43,"history_lines":3,"events_offset":33,"events_lines":2,"console_offset":81,"console_lines":1}
49
+ {"time":"2026-08-14T23:27:49.780398601Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
50
+ {"time":"2026-08-14T23:28:03.928989439Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":46,"history_lines":2,"events_offset":35,"events_lines":2,"console_offset":85,"console_lines":10}
51
+ {"time":"2026-08-14T23:28:04.898228604Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
52
+ {"time":"2026-08-14T23:28:18.925155782Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":48,"history_lines":2,"events_offset":37,"events_lines":2,"console_offset":91,"console_lines":1}
53
+ {"time":"2026-08-14T23:28:19.837293539Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
54
+ {"time":"2026-08-14T23:28:33.922625078Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":50,"history_lines":2,"events_offset":39,"events_lines":2,"console_offset":91,"console_lines":1}
55
+ {"time":"2026-08-14T23:28:34.701266798Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
56
+ {"time":"2026-08-14T23:28:44.099180607Z","level":"INFO","msg":"fileTransfer: Close: file transfer manager closed"}
57
+ {"time":"2026-08-14T23:28:44.109601111Z","level":"INFO","msg":"filestream: sending request","total_files":4,"history_offset":52,"history_lines":1,"events_offset":41,"events_lines":1,"console_offset":95,"console_lines":16,"uploaded_len":3,"complete":true,"exit_code":0}
58
+ {"time":"2026-08-14T23:28:44.865683962Z","level":"INFO","msg":"filestream: request sent","status":"200 OK"}
59
+ {"time":"2026-08-14T23:28:44.867128531Z","level":"INFO","msg":"handler: operation stats","stats":{}}
60
+ {"time":"2026-08-14T23:28:44.908328257Z","level":"INFO","msg":"stream: finishing up"}
61
+ {"time":"2026-08-14T23:28:44.908362014Z","level":"INFO","msg":"handler: closed"}
62
+ {"time":"2026-08-14T23:28:44.908459799Z","level":"INFO","msg":"sender: closed"}
63
+ {"time":"2026-08-14T23:28:44.908463762Z","level":"INFO","msg":"stream: all finished"}
zain/Activation/wandb/debug.log CHANGED
@@ -18,3 +18,8 @@ config: {'_wandb': {}}
18
  2026-08-14 23:23:18,900 INFO MainThread:2227330 [wandb_run.py:_config_callback():1346] config_cb None None {'transformers_version': '5.16.0.dev0', 'architectures': None, 'output_hidden_states': False, 'return_dict': True, 'dtype': None, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'problem_type': None, 'vocab_size': 4096, 'hidden_size': 128, 'intermediate_size': 256, 'num_hidden_layers': 100, 'num_attention_heads': 4, 'num_key_value_heads': 4, 'hidden_act': 'silu', 'max_position_embeddings': 512, 'initializer_range': 0.02, 'rms_norm_eps': 1e-06, 'use_cache': False, 'pad_token_id': 0, 'bos_token_id': 1, 'eos_token_id': 2, 'pretraining_tp': 1, 'tie_word_embeddings': True, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 32, '_name_or_path': '', 'tokenizer_name': 'w-ahmad/tiny-stories-tokenizer', 'mlp_type': 'glu', 'activation': 'silu-waleed10', 'waleed_beta': 0.5, 'model_type': 'tiny_llama', 'output_attentions': False, 'output_dir': 'out/glu-silu-waleed10-100L_trash_run', 'per_device_train_batch_size': 64, 'num_train_epochs': 1, 'max_steps': 3500, 'learning_rate': 9e-05, 'lr_scheduler_type': 'constant_with_warmup', 'lr_scheduler_kwargs': None, 'warmup_steps': 50, 'optim': 'adamw_torch_fused', 'optim_args': None, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'optim_target_modules': None, 'gradient_accumulation_steps': 1, 'average_tokens_across_devices': True, 'max_grad_norm': 1.0, 'label_smoothing_factor': 0.0, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'use_liger_kernel': False, 'liger_kernel_config': None, 'neftune_noise_alpha': None, 'torch_empty_cache_steps': None, 'auto_find_batch_size': False, 'logging_strategy': 'steps', 'logging_steps': 20, 'logging_first_step': False, 'log_on_each_node': True, 'logging_nan_inf_filter': True, 'include_num_input_tokens_seen': 'no', 'log_level': 'passive', 'log_level_replica': 'warning', 'disable_tqdm': False, 'report_to': ['wandb'], 'run_name': 'LM-glu-silu-waleed10-100L-16.9M-20260814-232317', 'project': 'huggingface', 'trackio_space_id': None, 'trackio_bucket_id': None, 'trackio_static_space_id': None, 'eval_strategy': 'steps', 'eval_steps': 2498, 'eval_delay': 0, 'per_device_eval_batch_size': 1024, 'prediction_loss_only': False, 'eval_on_start': False, 'eval_do_concat_batches': True, 'eval_use_gather_object': False, 'eval_accumulation_steps': None, 'include_for_metrics': [], 'batch_eval_metrics': False, 'save_only_model': False, 'save_strategy': 'steps', 'save_steps': 50, 'save_on_each_node': False, 'save_total_limit': None, 'enable_jit_checkpoint': False, 'push_to_hub': False, 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_model_id': 'w-ahmad/6L-glu-silu-waleed10-100L', 'hub_strategy': 'every_save', 'hub_always_push': False, 'hub_revision': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'restore_callback_states_from_checkpoint': False, 'full_determinism': False, 'seed': 42, 'data_seed': 42, 'use_cpu': False, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'dataloader_drop_last': False, 'dataloader_num_workers': 0, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'dataloader_prefetch_factor': None, 'dataloader_multiprocessing_context': None, 'dataloader_in_order': True, 'remove_unused_columns': False, 'label_names': None, 'train_sampling_strategy': 'random', 'length_column_name': 'length', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'ddp_static_graph': None, 'ddp_backend': None, 'ddp_timeout': 1800, 'fsdp': None, 'fsdp_config': None, 'deepspeed': None, 'debug': [], 'skip_memory_metrics': True, 'do_train': False, 'do_eval': True, 'do_predict': False, 'resume_from_checkpoint': None, 'local_rank': -1}
19
  2026-08-14 23:23:18,904 INFO MainThread:2227330 [wandb_config.py:__setitem__():155] [no run ID] config set model/num_parameters = 16934016 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x14c2d49d5890>>
20
  2026-08-14 23:23:18,904 INFO MainThread:2227330 [wandb_run.py:_config_callback():1346] config_cb model/num_parameters 16934016 None
 
 
 
 
 
 
18
  2026-08-14 23:23:18,900 INFO MainThread:2227330 [wandb_run.py:_config_callback():1346] config_cb None None {'transformers_version': '5.16.0.dev0', 'architectures': None, 'output_hidden_states': False, 'return_dict': True, 'dtype': None, 'chunk_size_feed_forward': 0, 'is_encoder_decoder': False, 'id2label': {0: 'LABEL_0', 1: 'LABEL_1'}, 'label2id': {'LABEL_0': 0, 'LABEL_1': 1}, 'problem_type': None, 'vocab_size': 4096, 'hidden_size': 128, 'intermediate_size': 256, 'num_hidden_layers': 100, 'num_attention_heads': 4, 'num_key_value_heads': 4, 'hidden_act': 'silu', 'max_position_embeddings': 512, 'initializer_range': 0.02, 'rms_norm_eps': 1e-06, 'use_cache': False, 'pad_token_id': 0, 'bos_token_id': 1, 'eos_token_id': 2, 'pretraining_tp': 1, 'tie_word_embeddings': True, 'rope_parameters': {'rope_theta': 10000.0, 'rope_type': 'default'}, 'attention_bias': False, 'attention_dropout': 0.0, 'mlp_bias': False, 'head_dim': 32, '_name_or_path': '', 'tokenizer_name': 'w-ahmad/tiny-stories-tokenizer', 'mlp_type': 'glu', 'activation': 'silu-waleed10', 'waleed_beta': 0.5, 'model_type': 'tiny_llama', 'output_attentions': False, 'output_dir': 'out/glu-silu-waleed10-100L_trash_run', 'per_device_train_batch_size': 64, 'num_train_epochs': 1, 'max_steps': 3500, 'learning_rate': 9e-05, 'lr_scheduler_type': 'constant_with_warmup', 'lr_scheduler_kwargs': None, 'warmup_steps': 50, 'optim': 'adamw_torch_fused', 'optim_args': None, 'weight_decay': 0.0, 'adam_beta1': 0.9, 'adam_beta2': 0.999, 'adam_epsilon': 1e-08, 'optim_target_modules': None, 'gradient_accumulation_steps': 1, 'average_tokens_across_devices': True, 'max_grad_norm': 1.0, 'label_smoothing_factor': 0.0, 'bf16': True, 'fp16': False, 'bf16_full_eval': False, 'fp16_full_eval': False, 'tf32': None, 'gradient_checkpointing': False, 'gradient_checkpointing_kwargs': None, 'torch_compile': False, 'torch_compile_backend': None, 'torch_compile_mode': None, 'use_liger_kernel': False, 'liger_kernel_config': None, 'neftune_noise_alpha': None, 'torch_empty_cache_steps': None, 'auto_find_batch_size': False, 'logging_strategy': 'steps', 'logging_steps': 20, 'logging_first_step': False, 'log_on_each_node': True, 'logging_nan_inf_filter': True, 'include_num_input_tokens_seen': 'no', 'log_level': 'passive', 'log_level_replica': 'warning', 'disable_tqdm': False, 'report_to': ['wandb'], 'run_name': 'LM-glu-silu-waleed10-100L-16.9M-20260814-232317', 'project': 'huggingface', 'trackio_space_id': None, 'trackio_bucket_id': None, 'trackio_static_space_id': None, 'eval_strategy': 'steps', 'eval_steps': 2498, 'eval_delay': 0, 'per_device_eval_batch_size': 1024, 'prediction_loss_only': False, 'eval_on_start': False, 'eval_do_concat_batches': True, 'eval_use_gather_object': False, 'eval_accumulation_steps': None, 'include_for_metrics': [], 'batch_eval_metrics': False, 'save_only_model': False, 'save_strategy': 'steps', 'save_steps': 50, 'save_on_each_node': False, 'save_total_limit': None, 'enable_jit_checkpoint': False, 'push_to_hub': False, 'hub_token': '<HUB_TOKEN>', 'hub_private_repo': None, 'hub_model_id': 'w-ahmad/6L-glu-silu-waleed10-100L', 'hub_strategy': 'every_save', 'hub_always_push': False, 'hub_revision': None, 'load_best_model_at_end': False, 'metric_for_best_model': None, 'greater_is_better': None, 'ignore_data_skip': False, 'restore_callback_states_from_checkpoint': False, 'full_determinism': False, 'seed': 42, 'data_seed': 42, 'use_cpu': False, 'accelerator_config': {'split_batches': False, 'dispatch_batches': None, 'even_batches': True, 'use_seedable_sampler': True, 'non_blocking': False, 'gradient_accumulation_kwargs': None}, 'parallelism_config': None, 'dataloader_drop_last': False, 'dataloader_num_workers': 0, 'dataloader_pin_memory': True, 'dataloader_persistent_workers': False, 'dataloader_prefetch_factor': None, 'dataloader_multiprocessing_context': None, 'dataloader_in_order': True, 'remove_unused_columns': False, 'label_names': None, 'train_sampling_strategy': 'random', 'length_column_name': 'length', 'ddp_find_unused_parameters': None, 'ddp_bucket_cap_mb': None, 'ddp_broadcast_buffers': None, 'ddp_static_graph': None, 'ddp_backend': None, 'ddp_timeout': 1800, 'fsdp': None, 'fsdp_config': None, 'deepspeed': None, 'debug': [], 'skip_memory_metrics': True, 'do_train': False, 'do_eval': True, 'do_predict': False, 'resume_from_checkpoint': None, 'local_rank': -1}
19
  2026-08-14 23:23:18,904 INFO MainThread:2227330 [wandb_config.py:__setitem__():155] [no run ID] config set model/num_parameters = 16934016 - <bound method Run._config_callback of <wandb.sdk.wandb_run.Run object at 0x14c2d49d5890>>
20
  2026-08-14 23:23:18,904 INFO MainThread:2227330 [wandb_run.py:_config_callback():1346] config_cb model/num_parameters 16934016 None
21
+ 2026-08-14 23:28:43,416 INFO MainThread:2227330 [wandb_run.py:_finish():2383] finishing run deepnevro-deepnevro/research7/l56mdhxl
22
+ 2026-08-14 23:28:43,417 INFO MainThread:2227330 [wandb_run.py:_atexit_cleanup():2588] got exitcode: 0
23
+ 2026-08-14 23:28:43,417 INFO MainThread:2227330 [wandb_run.py:_restore():2570] restore
24
+ 2026-08-14 23:28:43,417 INFO MainThread:2227330 [wandb_run.py:_restore():2576] restore done
25
+ 2026-08-14 23:28:44,907 INFO MainThread:2227330 [wandb_run.py:_footer_sync_info():3993] logging synced files
zain/Activation/wandb/run-20260814_231802-8ih2zdoc/logs/debug-core.log CHANGED
@@ -106,3 +106,11 @@
106
  {"time":"2026-08-14T23:23:18.278060411Z","level":"INFO","msg":"handleInformInit: received","streamId":"l56mdhxl","id":"5(@)"}
107
  {"time":"2026-08-14T23:23:18.518025519Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"l56mdhxl","id":"5(@)"}
108
  {"time":"2026-08-14T23:23:23.901514672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
 
 
 
 
 
 
 
 
 
106
  {"time":"2026-08-14T23:23:18.278060411Z","level":"INFO","msg":"handleInformInit: received","streamId":"l56mdhxl","id":"5(@)"}
107
  {"time":"2026-08-14T23:23:18.518025519Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"l56mdhxl","id":"5(@)"}
108
  {"time":"2026-08-14T23:23:23.901514672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
109
+ {"time":"2026-08-14T23:28:43.417416507Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
110
+ {"time":"2026-08-14T23:28:44.887436672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
111
+ {"time":"2026-08-14T23:28:44.908283775Z","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"l56mdhxl","id":"5(@)"}
112
+ {"time":"2026-08-14T23:28:44.90942001Z","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"l56mdhxl","id":"5(@)"}
113
+ {"time":"2026-08-14T23:28:46.822764669Z","level":"INFO","msg":"connection: closing","id":"5(@)"}
114
+ {"time":"2026-08-14T23:28:46.822848225Z","level":"INFO","msg":"connection: closed successfully","id":"5(@)"}
115
+ {"time":"2026-08-14T23:28:46.822771624Z","level":"INFO","msg":"processOutgoingData: finished","id":"5(@)"}
116
+ {"time":"2026-08-14T23:28:46.822855482Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"5(@)"}
zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/config.yaml ADDED
@@ -0,0 +1,436 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ _name_or_path:
2
+ value: ""
3
+ _wandb:
4
+ value:
5
+ cli_version: 0.28.1
6
+ e:
7
+ mymmadub93rrk8ejwtwkr54fzdwsb89e:
8
+ args:
9
+ - --config
10
+ - /mnt/data/zainulabideen/zain-exp/notebooks/zain/Activation/configs/baseline100l.yaml
11
+ - --variants
12
+ - glu-waleed10
13
+ - glu-silu-waleed10
14
+ codePath: sweep.py
15
+ codePathLocal: sweep.py
16
+ cpu_count: 112
17
+ cpu_count_logical: 224
18
+ cudaVersion: "12.4"
19
+ disk:
20
+ /:
21
+ total: "1560765693952"
22
+ used: "709387452416"
23
+ email: deepnevro@gmail.com
24
+ executable: /mnt/data/zainulabideen/zain-exp/notebooks/my_env/bin/python
25
+ git:
26
+ commit: 3fb47c6943d6927eafc39bd399a3d3a520bae74a
27
+ remote: https://github.com/w-ahmad1a10/Activation.git
28
+ gpu: NVIDIA H100 80GB HBM3
29
+ gpu_count: 8
30
+ gpu_nvidia:
31
+ - architecture: Hopper
32
+ cudaCores: 16896
33
+ memoryTotal: "85520809984"
34
+ name: NVIDIA H100 80GB HBM3
35
+ uuid: GPU-39c684a5-fde6-83d7-1663-0859795881ae
36
+ - architecture: Hopper
37
+ cudaCores: 16896
38
+ memoryTotal: "85520809984"
39
+ name: NVIDIA H100 80GB HBM3
40
+ uuid: GPU-68012e5a-38b6-b643-0ca6-62fb66720bf3
41
+ - architecture: Hopper
42
+ cudaCores: 16896
43
+ memoryTotal: "85520809984"
44
+ name: NVIDIA H100 80GB HBM3
45
+ uuid: GPU-132944c4-b689-2b5f-89a4-d730401677ab
46
+ - architecture: Hopper
47
+ cudaCores: 16896
48
+ memoryTotal: "85520809984"
49
+ name: NVIDIA H100 80GB HBM3
50
+ uuid: GPU-2df386cc-6d26-d0e2-7a2d-a057b0d95864
51
+ - architecture: Hopper
52
+ cudaCores: 16896
53
+ memoryTotal: "85520809984"
54
+ name: NVIDIA H100 80GB HBM3
55
+ uuid: GPU-bfa16575-1d94-1aa2-4537-2c93433f42ef
56
+ - architecture: Hopper
57
+ cudaCores: 16896
58
+ memoryTotal: "85520809984"
59
+ name: NVIDIA H100 80GB HBM3
60
+ uuid: GPU-bc6c3e3c-9b90-09ca-c034-774961847c54
61
+ - architecture: Hopper
62
+ cudaCores: 16896
63
+ memoryTotal: "85520809984"
64
+ name: NVIDIA H100 80GB HBM3
65
+ uuid: GPU-00a441e1-7c95-e7d6-4c35-43d6b291aea9
66
+ - architecture: Hopper
67
+ cudaCores: 16896
68
+ memoryTotal: "85520809984"
69
+ name: NVIDIA H100 80GB HBM3
70
+ uuid: GPU-1c4d29a2-4647-6fce-d8fc-0c5ecfbbd6ea
71
+ host: deeplens-k3s-node1
72
+ memory:
73
+ total: "2164089937920"
74
+ os: Linux-5.15.0-126-generic-x86_64-with-glibc2.35
75
+ program: /mnt/data/zainulabideen/zain-exp/notebooks/zain/Activation/sweep.py
76
+ python: CPython 3.11.15
77
+ root: /mnt/data/zainulabideen/zain-exp/notebooks/zain/Activation
78
+ startedAt: "2026-08-14T23:23:18.275909Z"
79
+ writerId: mymmadub93rrk8ejwtwkr54fzdwsb89e
80
+ m:
81
+ - "1": train/global_step
82
+ "6":
83
+ - 3
84
+ "7": []
85
+ - "2": '*'
86
+ "5": 1
87
+ "6":
88
+ - 1
89
+ "7": []
90
+ python_version: 3.11.15
91
+ t:
92
+ "1":
93
+ - 1
94
+ - 5
95
+ - 11
96
+ - 41
97
+ - 49
98
+ - 51
99
+ - 53
100
+ - 71
101
+ "2":
102
+ - 1
103
+ - 5
104
+ - 11
105
+ - 41
106
+ - 49
107
+ - 51
108
+ - 53
109
+ - 71
110
+ "3":
111
+ - 2
112
+ - 7
113
+ - 13
114
+ - 19
115
+ - 41
116
+ - 61
117
+ - 62
118
+ - 66
119
+ "4": 3.11.15
120
+ "5": 0.28.1
121
+ "6": 5.16.0.dev0
122
+ "9":
123
+ "1": transformers_trainer
124
+ "12": 0.28.1
125
+ "13": linux-x86_64
126
+ accelerator_config:
127
+ value:
128
+ dispatch_batches: null
129
+ even_batches: true
130
+ gradient_accumulation_kwargs: null
131
+ non_blocking: false
132
+ split_batches: false
133
+ use_seedable_sampler: true
134
+ activation:
135
+ value: silu-waleed10
136
+ adam_beta1:
137
+ value: 0.9
138
+ adam_beta2:
139
+ value: 0.999
140
+ adam_epsilon:
141
+ value: 1e-08
142
+ architectures:
143
+ value: null
144
+ attention_bias:
145
+ value: false
146
+ attention_dropout:
147
+ value: 0
148
+ auto_find_batch_size:
149
+ value: false
150
+ average_tokens_across_devices:
151
+ value: true
152
+ batch_eval_metrics:
153
+ value: false
154
+ bf16:
155
+ value: true
156
+ bf16_full_eval:
157
+ value: false
158
+ bos_token_id:
159
+ value: 1
160
+ chunk_size_feed_forward:
161
+ value: 0
162
+ data_seed:
163
+ value: 42
164
+ dataloader_drop_last:
165
+ value: false
166
+ dataloader_in_order:
167
+ value: true
168
+ dataloader_multiprocessing_context:
169
+ value: null
170
+ dataloader_num_workers:
171
+ value: 0
172
+ dataloader_persistent_workers:
173
+ value: false
174
+ dataloader_pin_memory:
175
+ value: true
176
+ dataloader_prefetch_factor:
177
+ value: null
178
+ ddp_backend:
179
+ value: null
180
+ ddp_broadcast_buffers:
181
+ value: null
182
+ ddp_bucket_cap_mb:
183
+ value: null
184
+ ddp_find_unused_parameters:
185
+ value: null
186
+ ddp_static_graph:
187
+ value: null
188
+ ddp_timeout:
189
+ value: 1800
190
+ debug:
191
+ value: []
192
+ deepspeed:
193
+ value: null
194
+ disable_tqdm:
195
+ value: false
196
+ do_eval:
197
+ value: true
198
+ do_predict:
199
+ value: false
200
+ do_train:
201
+ value: false
202
+ dtype:
203
+ value: null
204
+ enable_jit_checkpoint:
205
+ value: false
206
+ eos_token_id:
207
+ value: 2
208
+ eval_accumulation_steps:
209
+ value: null
210
+ eval_delay:
211
+ value: 0
212
+ eval_do_concat_batches:
213
+ value: true
214
+ eval_on_start:
215
+ value: false
216
+ eval_steps:
217
+ value: 2498
218
+ eval_strategy:
219
+ value: steps
220
+ eval_use_gather_object:
221
+ value: false
222
+ fp16:
223
+ value: false
224
+ fp16_full_eval:
225
+ value: false
226
+ fsdp:
227
+ value: null
228
+ fsdp_config:
229
+ value: null
230
+ full_determinism:
231
+ value: false
232
+ gradient_accumulation_steps:
233
+ value: 1
234
+ gradient_checkpointing:
235
+ value: false
236
+ gradient_checkpointing_kwargs:
237
+ value: null
238
+ greater_is_better:
239
+ value: null
240
+ head_dim:
241
+ value: 32
242
+ hidden_act:
243
+ value: silu
244
+ hidden_size:
245
+ value: 128
246
+ hub_always_push:
247
+ value: false
248
+ hub_model_id:
249
+ value: w-ahmad/6L-glu-silu-waleed10-100L
250
+ hub_private_repo:
251
+ value: null
252
+ hub_revision:
253
+ value: null
254
+ hub_strategy:
255
+ value: every_save
256
+ hub_token:
257
+ value: <HUB_TOKEN>
258
+ id2label:
259
+ value:
260
+ "0": LABEL_0
261
+ "1": LABEL_1
262
+ ignore_data_skip:
263
+ value: false
264
+ include_for_metrics:
265
+ value: []
266
+ include_num_input_tokens_seen:
267
+ value: "no"
268
+ initializer_range:
269
+ value: 0.02
270
+ intermediate_size:
271
+ value: 256
272
+ is_encoder_decoder:
273
+ value: false
274
+ label_names:
275
+ value: null
276
+ label_smoothing_factor:
277
+ value: 0
278
+ label2id:
279
+ value:
280
+ LABEL_0: 0
281
+ LABEL_1: 1
282
+ learning_rate:
283
+ value: 9e-05
284
+ length_column_name:
285
+ value: length
286
+ liger_kernel_config:
287
+ value: null
288
+ load_best_model_at_end:
289
+ value: false
290
+ local_rank:
291
+ value: -1
292
+ log_level:
293
+ value: passive
294
+ log_level_replica:
295
+ value: warning
296
+ log_on_each_node:
297
+ value: true
298
+ logging_first_step:
299
+ value: false
300
+ logging_nan_inf_filter:
301
+ value: true
302
+ logging_steps:
303
+ value: 20
304
+ logging_strategy:
305
+ value: steps
306
+ lr_scheduler_kwargs:
307
+ value: null
308
+ lr_scheduler_type:
309
+ value: constant_with_warmup
310
+ max_grad_norm:
311
+ value: 1
312
+ max_position_embeddings:
313
+ value: 512
314
+ max_steps:
315
+ value: 3500
316
+ metric_for_best_model:
317
+ value: null
318
+ mlp_bias:
319
+ value: false
320
+ mlp_type:
321
+ value: glu
322
+ model/num_parameters:
323
+ value: 16934016
324
+ model_type:
325
+ value: tiny_llama
326
+ neftune_noise_alpha:
327
+ value: null
328
+ num_attention_heads:
329
+ value: 4
330
+ num_hidden_layers:
331
+ value: 100
332
+ num_key_value_heads:
333
+ value: 4
334
+ num_train_epochs:
335
+ value: 1
336
+ optim:
337
+ value: adamw_torch_fused
338
+ optim_args:
339
+ value: null
340
+ optim_target_modules:
341
+ value: null
342
+ output_attentions:
343
+ value: false
344
+ output_dir:
345
+ value: out/glu-silu-waleed10-100L_trash_run
346
+ output_hidden_states:
347
+ value: false
348
+ pad_token_id:
349
+ value: 0
350
+ parallelism_config:
351
+ value: null
352
+ per_device_eval_batch_size:
353
+ value: 1024
354
+ per_device_train_batch_size:
355
+ value: 64
356
+ prediction_loss_only:
357
+ value: false
358
+ pretraining_tp:
359
+ value: 1
360
+ problem_type:
361
+ value: null
362
+ project:
363
+ value: huggingface
364
+ push_to_hub:
365
+ value: false
366
+ remove_unused_columns:
367
+ value: false
368
+ report_to:
369
+ value:
370
+ - wandb
371
+ restore_callback_states_from_checkpoint:
372
+ value: false
373
+ resume_from_checkpoint:
374
+ value: null
375
+ return_dict:
376
+ value: true
377
+ rms_norm_eps:
378
+ value: 1e-06
379
+ rope_parameters:
380
+ value:
381
+ rope_theta: 10000
382
+ rope_type: default
383
+ run_name:
384
+ value: LM-glu-silu-waleed10-100L-16.9M-20260814-232317
385
+ save_on_each_node:
386
+ value: false
387
+ save_only_model:
388
+ value: false
389
+ save_steps:
390
+ value: 50
391
+ save_strategy:
392
+ value: steps
393
+ save_total_limit:
394
+ value: null
395
+ seed:
396
+ value: 42
397
+ skip_memory_metrics:
398
+ value: true
399
+ tf32:
400
+ value: null
401
+ tie_word_embeddings:
402
+ value: true
403
+ tokenizer_name:
404
+ value: w-ahmad/tiny-stories-tokenizer
405
+ torch_compile:
406
+ value: false
407
+ torch_compile_backend:
408
+ value: null
409
+ torch_compile_mode:
410
+ value: null
411
+ torch_empty_cache_steps:
412
+ value: null
413
+ trackio_bucket_id:
414
+ value: null
415
+ trackio_space_id:
416
+ value: null
417
+ trackio_static_space_id:
418
+ value: null
419
+ train_sampling_strategy:
420
+ value: random
421
+ transformers_version:
422
+ value: 5.16.0.dev0
423
+ use_cache:
424
+ value: false
425
+ use_cpu:
426
+ value: false
427
+ use_liger_kernel:
428
+ value: false
429
+ vocab_size:
430
+ value: 4096
431
+ waleed_beta:
432
+ value: 0.5
433
+ warmup_steps:
434
+ value: 50
435
+ weight_decay:
436
+ value: 0
zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/output.log CHANGED
@@ -59,7 +59,53 @@ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 13.
59
  - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
60
  - If you are not the owner of the model architecture class, please contact the model code owner to update it.
61
  Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 15.07it/s]
62
- 91%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ | 3171/3500 [03:14<01:32, 3.56it/s], ?it/s]
63
  {'loss': '8.081', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2103', 'train/total_time_seconds': '146.2', 'train/time_per_step_avg': '0.2413', 'train/epoch_time_elapsed': '179.8', 'train/estimated_remaining_minutes': '0.2968', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '186', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009766', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.2', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '-3.91e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6816', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3270', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03088', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.125', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.875', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3283', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01855', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.219', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.306e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1243', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.805', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.07', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_model_layers_0_residual_pos
64
  {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2117', 'train/total_time_seconds': '150.9', 'train/time_per_step_avg': '0.2414', 'train/epoch_time_elapsed': '185.5', 'train/estimated_remaining_minutes': '0.2883'}
65
  {'loss': '8.077', 'grad_norm': '138', 'learning_rate': '9e-05', 'epoch': '0.213', 'train/total_time_seconds': '156', 'train/time_per_step_avg': '0.2418', 'train/epoch_time_elapsed': '191.6', 'train/estimated_remaining_minutes': '0.2797', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '185.9', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.001007', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.1', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '3.886e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6816', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3274', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.02844', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.781', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.781', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.969', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3286', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.0188', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.25', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.307e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '23.38', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1257', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.789', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '23.38', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.59', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_model_layers_0_residual_post_ml
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
59
  - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
60
  - If you are not the owner of the model architecture class, please contact the model code owner to update it.
61
  Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 15.07it/s]
62
+ 91%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–| 3200/3500 [03:23<01:23, 3.57it/s][transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From πŸ‘‰v4.50πŸ‘ˆ onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
63
  {'loss': '8.081', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2103', 'train/total_time_seconds': '146.2', 'train/time_per_step_avg': '0.2413', 'train/epoch_time_elapsed': '179.8', 'train/estimated_remaining_minutes': '0.2968', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '186', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009766', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.2', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '-3.91e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6816', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3270', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03088', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.125', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.875', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3283', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01855', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.219', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.306e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1243', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.805', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.07', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_model_layers_0_residual_pos
64
  {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2117', 'train/total_time_seconds': '150.9', 'train/time_per_step_avg': '0.2414', 'train/epoch_time_elapsed': '185.5', 'train/estimated_remaining_minutes': '0.2883'}
65
  {'loss': '8.077', 'grad_norm': '138', 'learning_rate': '9e-05', 'epoch': '0.213', 'train/total_time_seconds': '156', 'train/time_per_step_avg': '0.2418', 'train/epoch_time_elapsed': '191.6', 'train/estimated_remaining_minutes': '0.2797', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '185.9', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.001007', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.1', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '3.886e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6816', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3274', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.02844', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.781', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.781', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.969', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3286', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.0188', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.25', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.307e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '23.38', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1257', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.789', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '23.38', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.59', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_model_layers_0_residual_post_ml
66
+ {'loss': '8.076', 'grad_norm': '1.109', 'learning_rate': '9e-05', 'epoch': '0.2144', 'train/total_time_seconds': '160.6', 'train/time_per_step_avg': '0.2421', 'train/epoch_time_elapsed': '197.3', 'train/estimated_remaining_minutes': '0.2694'}
67
+ {'loss': '8.08', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2157', 'train/total_time_seconds': '165.3', 'train/time_per_step_avg': '0.2422', 'train/epoch_time_elapsed': '203', 'train/estimated_remaining_minutes': '0.2583'}
68
+ - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
69
+ - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
70
+ - If you are not the owner of the model architecture class, please contact the model code owner to update it.
71
+ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 14.22it/s]
72
+ 94%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–| 3300/3500 [03:52<00:56, 3.57it/s][transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From πŸ‘‰v4.50πŸ‘ˆ onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
73
+ {'loss': '8.08', 'grad_norm': '2.344', 'learning_rate': '9e-05', 'epoch': '0.2171', 'train/total_time_seconds': '170.7', 'train/time_per_step_avg': '0.2444', 'train/epoch_time_elapsed': '209.6', 'train/estimated_remaining_minutes': '0.2473', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '186.1', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009766', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.4', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '4.673e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.332', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.293', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.625', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3270', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.02917', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.125', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.531', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.531', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.719', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3283', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.0199', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.188', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4.031', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.188', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.219', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.304e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.047', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '23.62', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1247', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.672', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '23.62', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.95', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '1.654e-06', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.498', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0.001953', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.793', 'train/tensor_act_model_
74
+ {'loss': '8.078', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2184', 'train/total_time_seconds': '175.4', 'train/time_per_step_avg': '0.2447', 'train/epoch_time_elapsed': '215.4', 'train/estimated_remaining_minutes': '0.2346'}
75
+ {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2198', 'train/total_time_seconds': '180.5', 'train/time_per_step_avg': '0.2448', 'train/epoch_time_elapsed': '221.6', 'train/estimated_remaining_minutes': '0.2214', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '186.1', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009232', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.4', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '4.798e-06', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3457', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.3418', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3457', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6875', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3264', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03101', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.125', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.688', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.875', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3276', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.0188', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4.062', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.281', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.304e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '24.62', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1246', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.734', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '24.62', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '22.89', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8438', 'train/tensor_act_model_layers_0_residu
76
+ {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2211', 'train/total_time_seconds': '185.1', 'train/time_per_step_avg': '0.2449', 'train/epoch_time_elapsed': '227.3', 'train/estimated_remaining_minutes': '0.207'}
77
+ {'loss': '8.079', 'grad_norm': '1.859', 'learning_rate': '9e-05', 'epoch': '0.2224', 'train/total_time_seconds': '189.8', 'train/time_per_step_avg': '0.2446', 'train/epoch_time_elapsed': '233', 'train/estimated_remaining_minutes': '0.1917'}
78
+ - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
79
+ - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
80
+ - If you are not the owner of the model architecture class, please contact the model code owner to update it.
81
+ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 13.17it/s]
82
+ 97%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆοΏ½οΏ½οΏ½β–ˆβ–ˆβ–‹| 3400/3500 [04:22<00:28, 3.56it/s][transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From πŸ‘‰v4.50πŸ‘ˆ onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
83
+ {'loss': '8.078', 'grad_norm': '0.6914', 'learning_rate': '9e-05', 'epoch': '0.2238', 'train/total_time_seconds': '194.9', 'train/time_per_step_avg': '0.2422', 'train/epoch_time_elapsed': '239.3', 'train/estimated_remaining_minutes': '0.1761', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '185.9', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009918', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.1', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '6.39e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6816', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3274', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.02869', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.562', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.156', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.562', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.719', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3286', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01892', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-3.922', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.141', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.31e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.625', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.12', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1253', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.562', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.12', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '20.56', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '4.858e-06', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.498', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0.001953', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_m
84
+ {'loss': '8.078', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2251', 'train/total_time_seconds': '199.6', 'train/time_per_step_avg': '0.2419', 'train/epoch_time_elapsed': '245.1', 'train/estimated_remaining_minutes': '0.1593'}
85
+ {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2265', 'train/total_time_seconds': '204.7', 'train/time_per_step_avg': '0.242', 'train/epoch_time_elapsed': '251.2', 'train/estimated_remaining_minutes': '0.1421', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '185.7', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009232', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '-1.317e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3203', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.3203', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.2988', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6191', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3279', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03027', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.531', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.219', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.531', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.75', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3291', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01904', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4.156', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.375', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.315e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.625', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.12', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1268', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.812', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.12', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '20.31', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.7969', 'train/tensor_act_model_layers_0_residual
86
+ {'loss': '8.08', 'grad_norm': '0.7656', 'learning_rate': '9e-05', 'epoch': '0.2278', 'train/total_time_seconds': '209.3', 'train/time_per_step_avg': '0.2417', 'train/epoch_time_elapsed': '256.9', 'train/estimated_remaining_minutes': '0.1238'}
87
+ {'loss': '8.08', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2292', 'train/total_time_seconds': '214', 'train/time_per_step_avg': '0.2416', 'train/epoch_time_elapsed': '262.5', 'train/estimated_remaining_minutes': '0.1049'}
88
+ - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
89
+ - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
90
+ - If you are not the owner of the model architecture class, please contact the model code owner to update it.
91
+ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 13.41it/s]
92
+ 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 3500/3500 [05:08<00:00, 3.52i[transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From πŸ‘‰v4.50πŸ‘ˆ onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
93
+ {'loss': '8.08', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2305', 'train/total_time_seconds': '219.1', 'train/time_per_step_avg': '0.2422', 'train/epoch_time_elapsed': '269', 'train/estimated_remaining_minutes': '0.08542', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '185.9', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.0009308', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.1', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '-1.061e-05', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.3203', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3496', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6699', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3275', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03003', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.656', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.156', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.656', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '8.812', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3287', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01843', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.281', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4.219', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.281', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.5', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.312e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.625', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.062', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.126', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.742', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.13', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8516', 'train/tensor_act_model_layers_0_residual_
94
+ {'loss': '8.079', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2319', 'train/total_time_seconds': '223.8', 'train/time_per_step_avg': '0.2428', 'train/epoch_time_elapsed': '274.8', 'train/estimated_remaining_minutes': '0.06507'}
95
+ {'loss': '8.082', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2332', 'train/total_time_seconds': '229', 'train/time_per_step_avg': '0.243', 'train/epoch_time_elapsed': '281', 'train/estimated_remaining_minutes': '0.04412', 'train/tensor_act_model_layers_0_residual_pre_attn/norm': '186', 'train/tensor_act_model_layers_0_residual_pre_attn/mean': '0.00106', 'train/tensor_act_model_layers_0_residual_pre_attn/std': '0.09082', 'train/tensor_act_model_layers_0_residual_pre_attn/max_abs': '0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_pre_attn/min': '-0.3008', 'train/tensor_act_model_layers_0_residual_pre_attn/max': '0.2852', 'train/tensor_act_model_layers_0_residual_pre_attn/range': '0.5859', 'train/tensor_act_model_layers_0_residual_post_attn/norm': '187.3', 'train/tensor_act_model_layers_0_residual_post_attn/mean': '0.0001574', 'train/tensor_act_model_layers_0_residual_post_attn/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_attn/max_abs': '0.332', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_residual_post_attn/min': '-0.332', 'train/tensor_act_model_layers_0_residual_post_attn/max': '0.3047', 'train/tensor_act_model_layers_0_residual_post_attn/range': '0.6367', 'train/tensor_act_model_layers_0_mlp_gate_proj/norm': '3264', 'train/tensor_act_model_layers_0_mlp_gate_proj/mean': '0.03076', 'train/tensor_act_model_layers_0_mlp_gate_proj/std': '1.125', 'train/tensor_act_model_layers_0_mlp_gate_proj/max_abs': '4.812', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_gate_proj/min': '-4.188', 'train/tensor_act_model_layers_0_mlp_gate_proj/max': '4.812', 'train/tensor_act_model_layers_0_mlp_gate_proj/range': '9', 'train/tensor_act_model_layers_0_mlp_up_proj/norm': '3277', 'train/tensor_act_model_layers_0_mlp_up_proj/mean': '0.01917', 'train/tensor_act_model_layers_0_mlp_up_proj/std': '1.133', 'train/tensor_act_model_layers_0_mlp_up_proj/max_abs': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp_up_proj/min': '-4.125', 'train/tensor_act_model_layers_0_mlp_up_proj/max': '4.25', 'train/tensor_act_model_layers_0_mlp_up_proj/range': '8.375', 'train/tensor_act_model_layers_0_mlp_down_proj/norm': '1.307e+04', 'train/tensor_act_model_layers_0_mlp_down_proj/mean': '5.594', 'train/tensor_act_model_layers_0_mlp_down_proj/std': '3.078', 'train/tensor_act_model_layers_0_mlp_down_proj/max_abs': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp_down_proj/frac_near_user_limit': '0.1263', 'train/tensor_act_model_layers_0_mlp_down_proj/min': '1.797', 'train/tensor_act_model_layers_0_mlp_down_proj/max': '22.88', 'train/tensor_act_model_layers_0_mlp_down_proj/range': '21.08', 'train/tensor_act_model_layers_0_mlp/norm': '1024', 'train/tensor_act_model_layers_0_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_mlp/std': '0', 'train/tensor_act_model_layers_0_mlp/max_abs': '0.5', 'train/tensor_act_model_layers_0_mlp/frac_near_dtype_limit': '0', 'train/tensor_act_model_layers_0_mlp/frac_near_user_limit': '0', 'train/tensor_act_model_layers_0_mlp/min': '0.5', 'train/tensor_act_model_layers_0_mlp/max': '0.5', 'train/tensor_act_model_layers_0_mlp/range': '0', 'train/tensor_act_model_layers_0_residual_post_mlp/norm': '1041', 'train/tensor_act_model_layers_0_residual_post_mlp/mean': '0.5', 'train/tensor_act_model_layers_0_residual_post_mlp/std': '0.09131', 'train/tensor_act_model_layers_0_residual_post_mlp/max_abs': '0.8047', 'train/tensor_act_model_layers_0_residual_post_mlp/fra
96
+ {'loss': '8.083', 'grad_norm': '1.18', 'learning_rate': '9e-05', 'epoch': '0.2346', 'train/total_time_seconds': '233.6', 'train/time_per_step_avg': '0.2429', 'train/epoch_time_elapsed': '286.6', 'train/estimated_remaining_minutes': '0.02238'}
97
+ {'loss': '8.084', 'grad_norm': '0', 'learning_rate': '9e-05', 'epoch': '0.2359', 'train/total_time_seconds': '238.4', 'train/time_per_step_avg': '0.2443', 'train/epoch_time_elapsed': '292.5', 'train/estimated_remaining_minutes': '0'}
98
+ - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
99
+ {'eval_loss': '8.078', 'eval_runtime': '15.77', 'eval_samples_per_second': '604.1', 'eval_steps_per_second': '0.634', 'epoch': '0.2359', 'train/total_time_seconds': '238.4', 'train/time_per_step_avg': '0.2443', 'train/epoch_time_elapsed': '308.3', 'train/estimated_remaining_minutes': '0'}
100
+ - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
101
+ - If you are not the owner of the model architecture class, please contact the model code owner to update it.
102
+ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 14.51it/s]
103
+ 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 3500/3500 [05:08<00:00, 11.35it/s], ?it/s]
104
+ {'train_runtime': '309.3', 'train_samples_per_second': '724.2', 'train_steps_per_second': '11.32', 'train_loss': '2.204', 'epoch': '0.2359', 'train/total_time_seconds': '238.4', 'train/time_per_step_avg': '0.2443', 'train/epoch_time_elapsed': '308.5', 'train/estimated_remaining_minutes': '0'}
105
+ 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 10/10 [00:13<00:00, 1.38s/it]
106
+ [transformers] TinyLlamaForCausalLM has generative capabilities, as `prepare_inputs_for_generation` is explicitly defined. However, it doesn't directly inherit from `GenerationMixin`. From πŸ‘‰v4.50πŸ‘ˆ onwards, `PreTrainedModel` will NOT inherit from `GenerationMixin`, and this model will lose the ability to call `generate` and other related functions.
107
+ - If you're using `trust_remote_code=True`, you can get rid of this warning by loading the model with an auto class. See https://huggingface.co/docs/transformers/en/model_doc/auto#auto-classes
108
+ - If you are the owner of the model architecture code, please modify your model class such that it inherits from `GenerationMixin` (after `PreTrainedModel`, otherwise you'll get an exception).
109
+ - If you are not the owner of the model architecture class, please contact the model code owner to update it.
110
+ Writing model shards: 100%|β–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆβ–ˆ| 1/1 [00:00<00:00, 16.85it/s]
111
+ >>> FINISHED glu-silu-waleed10-100L successfully
zain/Activation/wandb/run-20260814_232318-l56mdhxl/files/wandb-summary.json ADDED
The diff for this file is too large to render. See raw diff
 
zain/Activation/wandb/run-20260814_232318-l56mdhxl/logs/debug-core.log CHANGED
@@ -106,3 +106,11 @@
106
  {"time":"2026-08-14T23:23:18.278060411Z","level":"INFO","msg":"handleInformInit: received","streamId":"l56mdhxl","id":"5(@)"}
107
  {"time":"2026-08-14T23:23:18.518025519Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"l56mdhxl","id":"5(@)"}
108
  {"time":"2026-08-14T23:23:23.901514672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
 
 
 
 
 
 
 
 
 
106
  {"time":"2026-08-14T23:23:18.278060411Z","level":"INFO","msg":"handleInformInit: received","streamId":"l56mdhxl","id":"5(@)"}
107
  {"time":"2026-08-14T23:23:18.518025519Z","level":"INFO","msg":"handleInformInit: stream started","streamId":"l56mdhxl","id":"5(@)"}
108
  {"time":"2026-08-14T23:23:23.901514672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
109
+ {"time":"2026-08-14T23:28:43.417416507Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
110
+ {"time":"2026-08-14T23:28:44.887436672Z","level":"INFO","msg":"connection: cancelling request","id":"5(@)","requestId":"6st0ad10saav"}
111
+ {"time":"2026-08-14T23:28:44.908283775Z","level":"INFO","msg":"handleInformFinish: finish message received","streamId":"l56mdhxl","id":"5(@)"}
112
+ {"time":"2026-08-14T23:28:44.90942001Z","level":"INFO","msg":"handleInformFinish: stream closed","streamId":"l56mdhxl","id":"5(@)"}
113
+ {"time":"2026-08-14T23:28:46.822764669Z","level":"INFO","msg":"connection: closing","id":"5(@)"}
114
+ {"time":"2026-08-14T23:28:46.822848225Z","level":"INFO","msg":"connection: closed successfully","id":"5(@)"}
115
+ {"time":"2026-08-14T23:28:46.822771624Z","level":"INFO","msg":"processOutgoingData: finished","id":"5(@)"}
116
+ {"time":"2026-08-14T23:28:46.822855482Z","level":"INFO","msg":"connection: ManageConnectionData: connection closed","id":"5(@)"}